Merge branch 'linus' into tracing/urgent

Merge reason: Merge up to almost-rc6 to pick up latest perfcounters
              (on which we'll queue up a dependent fix)

Signed-off-by: Ingo Molnar <mingo@elte.hu>

292 files changed, 11053 insertions, 7380 deletions

diff --git a/drivers/acpi/acpi_memhotplug.c b/drivers/acpi/acpi_memhotplug.c
index 7a0f4aa4fa1..9a62224cc27 100644
--- a/drivers/acpi/acpi_memhotplug.c
+++ b/drivers/acpi/acpi_memhotplug.c
@@ -38,6 +38,9 @@
 
 #define _COMPONENT		ACPI_MEMORY_DEVICE_COMPONENT
 
+#undef PREFIX
+#define 	PREFIX		"ACPI:memory_hp:"
+
 ACPI_MODULE_NAME("acpi_memhotplug");
 MODULE_AUTHOR("Naveen B S <naveen.b.s@intel.com>");
 MODULE_DESCRIPTION("Hotplug Mem Driver");
@@ -153,6 +156,7 @@ acpi_memory_get_device(acpi_handle handle,
 	acpi_handle phandle;
 	struct acpi_device *device = NULL;
 	struct acpi_device *pdevice = NULL;
+	int result;
 
 
 	if (!acpi_bus_get_device(handle, &device) && device)
@@ -165,9 +169,9 @@ acpi_memory_get_device(acpi_handle handle,
 	}
 
 	/* Get the parent device */
-	status = acpi_bus_get_device(phandle, &pdevice);
-	if (ACPI_FAILURE(status)) {
-		ACPI_EXCEPTION((AE_INFO, status, "Cannot get acpi bus device"));
+	result = acpi_bus_get_device(phandle, &pdevice);
+	if (result) {
+		printk(KERN_WARNING PREFIX "Cannot get acpi bus device");
 		return -EINVAL;
 	}
 
@@ -175,9 +179,9 @@ acpi_memory_get_device(acpi_handle handle,
 	 * Now add the notified device.  This creates the acpi_device
 	 * and invokes .add function
 	 */
-	status = acpi_bus_add(&device, pdevice, handle, ACPI_BUS_TYPE_DEVICE);
-	if (ACPI_FAILURE(status)) {
-		ACPI_EXCEPTION((AE_INFO, status, "Cannot add acpi bus"));
+	result = acpi_bus_add(&device, pdevice, handle, ACPI_BUS_TYPE_DEVICE);
+	if (result) {
+		printk(KERN_WARNING PREFIX "Cannot add acpi bus");
 		return -EINVAL;
 	}
 
@@ -238,7 +242,12 @@ static int acpi_memory_enable_device(struct acpi_memory_device *mem_device)
 			num_enabled++;
 			continue;
 		}
-
+		/*
+		 * If the memory block size is zero, please ignore it.
+		 * Don't try to do the following memory hotplug flowchart.
+		 */
+		if (!info->length)
+			continue;
 		if (node < 0)
 			node = memory_add_physaddr_to_nid(info->start_addr);
 
@@ -253,8 +262,15 @@ static int acpi_memory_enable_device(struct acpi_memory_device *mem_device)
 		mem_device->state = MEMORY_INVALID_STATE;
 		return -EINVAL;
 	}
-
-	return result;
+	/*
+	 * Sometimes the memory device will contain several memory blocks.
+	 * When one memory block is hot-added to the system memory, it will
+	 * be regarded as a success.
+	 * Otherwise if the last memory block can't be hot-added to the system
+	 * memory, it will be failure and the memory device can't be bound with
+	 * driver.
+	 */
+	return 0;
 }
 
 static int acpi_memory_powerdown_device(struct acpi_memory_device *mem_device)
diff --git a/drivers/acpi/acpica/acobject.h b/drivers/acpi/acpica/acobject.h
index 544dcf83492..eb6f038b03d 100644
--- a/drivers/acpi/acpica/acobject.h
+++ b/drivers/acpi/acpica/acobject.h
@@ -97,6 +97,7 @@
 #define AOPOBJ_OBJECT_INITIALIZED   0x08
 #define AOPOBJ_SETUP_COMPLETE       0x10
 #define AOPOBJ_SINGLE_DATUM         0x20
+#define AOPOBJ_INVALID              0x40	/* Used if host OS won't allow an op_region address */
 
 /******************************************************************************
  *
diff --git a/drivers/acpi/acpica/dsopcode.c b/drivers/acpi/acpica/dsopcode.c
index 584d766e6f1..b79978f7bc7 100644
--- a/drivers/acpi/acpica/dsopcode.c
+++ b/drivers/acpi/acpica/dsopcode.c
@@ -397,6 +397,30 @@ acpi_status acpi_ds_get_region_arguments(union acpi_operand_object *obj_desc)
 	status = acpi_ds_execute_arguments(node, acpi_ns_get_parent_node(node),
 					   extra_desc->extra.aml_length,
 					   extra_desc->extra.aml_start);
+	if (ACPI_FAILURE(status)) {
+		return_ACPI_STATUS(status);
+	}
+
+	/* Validate the region address/length via the host OS */
+
+	status = acpi_os_validate_address(obj_desc->region.space_id,
+					  obj_desc->region.address,
+					  (acpi_size) obj_desc->region.length,
+					  acpi_ut_get_node_name(node));
+
+	if (ACPI_FAILURE(status)) {
+		/*
+		 * Invalid address/length. We will emit an error message and mark
+		 * the region as invalid, so that it will cause an additional error if
+		 * it is ever used. Then return AE_OK.
+		 */
+		ACPI_EXCEPTION((AE_INFO, status,
+				"During address validation of OpRegion [%4.4s]",
+				node->name.ascii));
+		obj_desc->common.flags |= AOPOBJ_INVALID;
+		status = AE_OK;
+	}
+
 	return_ACPI_STATUS(status);
 }
 
diff --git a/drivers/acpi/acpica/exfldio.c b/drivers/acpi/acpica/exfldio.c
index d4075b82102..6687be167f5 100644
--- a/drivers/acpi/acpica/exfldio.c
+++ b/drivers/acpi/acpica/exfldio.c
@@ -113,6 +113,12 @@ acpi_ex_setup_region(union acpi_operand_object *obj_desc,
 		}
 	}
 
+	/* Exit if Address/Length have been disallowed by the host OS */
+
+	if (rgn_desc->common.flags & AOPOBJ_INVALID) {
+		return_ACPI_STATUS(AE_AML_ILLEGAL_ADDRESS);
+	}
+
 	/*
 	 * Exit now for SMBus address space, it has a non-linear address space
 	 * and the request cannot be directly validated
diff --git a/drivers/acpi/osl.c b/drivers/acpi/osl.c
index 71670719d61..5691f165a95 100644
--- a/drivers/acpi/osl.c
+++ b/drivers/acpi/osl.c
@@ -189,11 +189,36 @@ acpi_status __init acpi_os_initialize(void)
 	return AE_OK;
 }
 
+static void bind_to_cpu0(struct work_struct *work)
+{
+	set_cpus_allowed(current, cpumask_of_cpu(0));
+	kfree(work);
+}
+
+static void bind_workqueue(struct workqueue_struct *wq)
+{
+	struct work_struct *work;
+
+	work = kzalloc(sizeof(struct work_struct), GFP_KERNEL);
+	INIT_WORK(work, bind_to_cpu0);
+	queue_work(wq, work);
+}
+
 acpi_status acpi_os_initialize1(void)
 {
+	/*
+	 * On some machines, a software-initiated SMI causes corruption unless
+	 * the SMI runs on CPU 0.  An SMI can be initiated by any AML, but
+	 * typically it's done in GPE-related methods that are run via
+	 * workqueues, so we can avoid the known corruption cases by binding
+	 * the workqueues to CPU 0.
+	 */
 	kacpid_wq = create_singlethread_workqueue("kacpid");
+	bind_workqueue(kacpid_wq);
 	kacpi_notify_wq = create_singlethread_workqueue("kacpi_notify");
+	bind_workqueue(kacpi_notify_wq);
 	kacpi_hotplug_wq = create_singlethread_workqueue("kacpi_hotplug");
+	bind_workqueue(kacpi_hotplug_wq);
 	BUG_ON(!kacpid_wq);
 	BUG_ON(!kacpi_notify_wq);
 	BUG_ON(!kacpi_hotplug_wq);
diff --git a/drivers/acpi/sleep.c b/drivers/acpi/sleep.c
index 01574a06653..42159a28f43 100644
--- a/drivers/acpi/sleep.c
+++ b/drivers/acpi/sleep.c
@@ -397,6 +397,14 @@ static struct dmi_system_id __initdata acpisleep_dmi_table[] = {
 		},
 	},
 	{
+	.callback = init_set_sci_en_on_resume,
+	.ident = "Hewlett-Packard HP G7000 Notebook PC",
+	.matches = {
+		DMI_MATCH(DMI_SYS_VENDOR, "Hewlett-Packard"),
+		DMI_MATCH(DMI_PRODUCT_NAME, "HP G7000 Notebook PC"),
+		},
+	},
+	{
 	.callback = init_old_suspend_ordering,
 	.ident = "Panasonic CF51-2L",
 	.matches = {
diff --git a/drivers/acpi/system.c b/drivers/acpi/system.c
index 0944daec064..9c61ab2177c 100644
--- a/drivers/acpi/system.c
+++ b/drivers/acpi/system.c
@@ -121,7 +121,7 @@ static void acpi_table_attr_init(struct acpi_table_attr *table_attr,
 	table_attr->attr.size = 0;
 	table_attr->attr.read = acpi_table_show;
 	table_attr->attr.attr.name = table_attr->name;
-	table_attr->attr.attr.mode = 0444;
+	table_attr->attr.attr.mode = 0400;
 
 	return;
 }
diff --git a/drivers/ata/ahci.c b/drivers/ata/ahci.c
index 336eb1ed73c..958c1fa4190 100644
--- a/drivers/ata/ahci.c
+++ b/drivers/ata/ahci.c
@@ -515,10 +515,14 @@ static const struct pci_device_id ahci_pci_tbl[] = {
 	{ PCI_VDEVICE(INTEL, 0x3a05), board_ahci }, /* ICH10 */
 	{ PCI_VDEVICE(INTEL, 0x3a22), board_ahci }, /* ICH10 */
 	{ PCI_VDEVICE(INTEL, 0x3a25), board_ahci }, /* ICH10 */
+	{ PCI_VDEVICE(INTEL, 0x3b22), board_ahci }, /* PCH AHCI */
+	{ PCI_VDEVICE(INTEL, 0x3b23), board_ahci }, /* PCH AHCI */
 	{ PCI_VDEVICE(INTEL, 0x3b24), board_ahci }, /* PCH RAID */
 	{ PCI_VDEVICE(INTEL, 0x3b25), board_ahci }, /* PCH RAID */
+	{ PCI_VDEVICE(INTEL, 0x3b29), board_ahci }, /* PCH AHCI */
 	{ PCI_VDEVICE(INTEL, 0x3b2b), board_ahci }, /* PCH RAID */
 	{ PCI_VDEVICE(INTEL, 0x3b2c), board_ahci }, /* PCH RAID */
+	{ PCI_VDEVICE(INTEL, 0x3b2f), board_ahci }, /* PCH AHCI */
 
 	/* JMicron 360/1/3/5/6, match class to avoid IDE function */
 	{ PCI_VENDOR_ID_JMICRON, PCI_ANY_ID, PCI_ANY_ID, PCI_ANY_ID,
diff --git a/drivers/ata/ata_piix.c b/drivers/ata/ata_piix.c
index d0a14cf2bd7..56b8a3ff128 100644
--- a/drivers/ata/ata_piix.c
+++ b/drivers/ata/ata_piix.c
@@ -596,9 +596,12 @@ static const struct ich_laptop ich_laptop[] = {
 	{ 0x27DF, 0x0005, 0x0280 },	/* ICH7 on Acer 5602WLMi */
 	{ 0x27DF, 0x1025, 0x0102 },	/* ICH7 on Acer 5602aWLMi */
 	{ 0x27DF, 0x1025, 0x0110 },	/* ICH7 on Acer 3682WLMi */
+	{ 0x27DF, 0x1028, 0x02b0 },	/* ICH7 on unknown Dell */
 	{ 0x27DF, 0x1043, 0x1267 },	/* ICH7 on Asus W5F */
 	{ 0x27DF, 0x103C, 0x30A1 },	/* ICH7 on HP Compaq nc2400 */
+	{ 0x27DF, 0x103C, 0x361a },	/* ICH7 on unkown HP  */
 	{ 0x27DF, 0x1071, 0xD221 },	/* ICH7 on Hercules EC-900 */
+	{ 0x27DF, 0x152D, 0x0778 },	/* ICH7 on unknown Intel */
 	{ 0x24CA, 0x1025, 0x0061 },	/* ICH4 on ACER Aspire 2023WLMi */
 	{ 0x24CA, 0x1025, 0x003d },	/* ICH4 on ACER TM290 */
 	{ 0x266F, 0x1025, 0x0066 },	/* ICH6 on ACER Aspire 1694WLMi */
diff --git a/drivers/ata/libata-core.c b/drivers/ata/libata-core.c
index 2c6aedaef71..8ac98ff16d7 100644
--- a/drivers/ata/libata-core.c
+++ b/drivers/ata/libata-core.c
@@ -1515,6 +1515,7 @@ static int ata_hpa_resize(struct ata_device *dev)
 
 		return rc;
 	}
+	dev->n_native_sectors = native_sectors;
 
 	/* nothing to do? */
 	if (native_sectors <= sectors || !ata_ignore_hpa) {
@@ -4099,6 +4100,7 @@ int ata_dev_revalidate(struct ata_device *dev, unsigned int new_class,
 		       unsigned int readid_flags)
 {
 	u64 n_sectors = dev->n_sectors;
+	u64 n_native_sectors = dev->n_native_sectors;
 	int rc;
 
 	if (!ata_dev_enabled(dev))
@@ -4128,16 +4130,30 @@ int ata_dev_revalidate(struct ata_device *dev, unsigned int new_class,
 	/* verify n_sectors hasn't changed */
 	if (dev->class == ATA_DEV_ATA && n_sectors &&
 	    dev->n_sectors != n_sectors) {
-		ata_dev_printk(dev, KERN_INFO, "n_sectors mismatch "
+		ata_dev_printk(dev, KERN_WARNING, "n_sectors mismatch "
 			       "%llu != %llu\n",
 			       (unsigned long long)n_sectors,
 			       (unsigned long long)dev->n_sectors);
-
-		/* restore original n_sectors */
-		dev->n_sectors = n_sectors;
-
-		rc = -ENODEV;
-		goto fail;
+		/*
+		 * Something could have caused HPA to be unlocked
+		 * involuntarily.  If n_native_sectors hasn't changed
+		 * and the new size matches it, keep the device.
+		 */
+		if (dev->n_native_sectors == n_native_sectors &&
+		    dev->n_sectors > n_sectors &&
+		    dev->n_sectors == n_native_sectors) {
+			ata_dev_printk(dev, KERN_WARNING,
+				       "new n_sectors matches native, probably "
+				       "late HPA unlock, continuing\n");
+			/* keep using the old n_sectors */
+			dev->n_sectors = n_sectors;
+		} else {
+			/* restore original n_[native]_sectors and fail */
+			dev->n_native_sectors = n_native_sectors;
+			dev->n_sectors = n_sectors;
+			rc = -ENODEV;
+			goto fail;
+		}
 	}
 
 	return 0;
diff --git a/drivers/ata/libata-eh.c b/drivers/ata/libata-eh.c
index 1a07c061f64..79711b64054 100644
--- a/drivers/ata/libata-eh.c
+++ b/drivers/ata/libata-eh.c
@@ -2327,7 +2327,7 @@ int ata_eh_reset(struct ata_link *link, int classify,
 	struct ata_port *ap = link->ap;
 	struct ata_link *slave = ap->slave_link;
 	struct ata_eh_context *ehc = &link->eh_context;
-	struct ata_eh_context *sehc = &slave->eh_context;
+	struct ata_eh_context *sehc = slave ? &slave->eh_context : NULL;
 	unsigned int *classes = ehc->classes;
 	unsigned int lflags = link->flags;
 	int verbose = !(ehc->i.flags & ATA_EHI_QUIET);
diff --git a/drivers/ata/pata_at91.c b/drivers/ata/pata_at91.c
index 8561a9f195c..5702affcb32 100644
--- a/drivers/ata/pata_at91.c
+++ b/drivers/ata/pata_at91.c
@@ -26,9 +26,7 @@
 #include <linux/platform_device.h>
 #include <linux/ata_platform.h>
 
-#include <mach/at91sam9260_matrix.h>
 #include <mach/at91sam9_smc.h>
-#include <mach/at91sam9260.h>
 #include <mach/board.h>
 #include <mach/gpio.h>
 
@@ -44,65 +42,62 @@ struct at91_ide_info {
 	unsigned long mode;
 	unsigned int cs;
 
+	struct clk *mck;
+
 	void __iomem *ide_addr;
 	void __iomem *alt_addr;
 };
 
-const struct ata_timing initial_timing =
+static const struct ata_timing initial_timing =
 	{XFER_PIO_0, 70, 290, 240, 600, 165, 150, 600, 0};
 
-static unsigned int calc_mck_cycles(unsigned int ns, unsigned int mck_hz)
+static unsigned long calc_mck_cycles(unsigned long ns, unsigned long mck_hz)
 {
 	unsigned long mul;
 
-    /*
-     * cycles = x [nsec] * f [Hz] / 10^9 [ns in sec] =
-     *     x * (f / 1_000_000_000) =
-     *     x * ((f * 65536) / 1_000_000_000) / 65536 =
-     *     x * (((f / 10_000) * 65536) / 100_000) / 65536 =
-     */
+	/*
+	* cycles = x [nsec] * f [Hz] / 10^9 [ns in sec] =
+	*     x * (f / 1_000_000_000) =
+	*     x * ((f * 65536) / 1_000_000_000) / 65536 =
+	*     x * (((f / 10_000) * 65536) / 100_000) / 65536 =
+	*/
 
-    mul = (mck_hz / 10000) << 16;
-    mul /= 100000;
+	mul = (mck_hz / 10000) << 16;
+	mul /= 100000;
 
-    return (ns * mul + 65536) >> 16;    /* rounding */
+	return (ns * mul + 65536) >> 16;    /* rounding */
 }
 
 static void set_smc_mode(struct at91_ide_info *info)
 {
-    at91_sys_write(AT91_SMC_MODE(info->cs), info->mode);
-    return;
+	at91_sys_write(AT91_SMC_MODE(info->cs), info->mode);
+	return;
 }
 
 static void set_smc_timing(struct device *dev,
 		struct at91_ide_info *info, const struct ata_timing *ata)
 {
-	int read_cycle, write_cycle, active, recover;
-	int nrd_setup, nrd_pulse, nrd_recover;
-	int nwe_setup, nwe_pulse;
+	unsigned long read_cycle, write_cycle, active, recover;
+	unsigned long nrd_setup, nrd_pulse, nrd_recover;
+	unsigned long nwe_setup, nwe_pulse;
 
-	int ncs_write_setup, ncs_write_pulse;
-	int ncs_read_setup, ncs_read_pulse;
+	unsigned long ncs_write_setup, ncs_write_pulse;
+	unsigned long ncs_read_setup, ncs_read_pulse;
 
-	unsigned int mck_hz;
-	struct clk *mck;
+	unsigned long mck_hz;
 
 	read_cycle  = ata->cyc8b;
 	nrd_setup   = ata->setup;
 	nrd_pulse   = ata->act8b;
 	nrd_recover = ata->rec8b;
 
-	mck = clk_get(NULL, "mck");
-	BUG_ON(IS_ERR(mck));
-	mck_hz = clk_get_rate(mck);
+	mck_hz = clk_get_rate(info->mck);
 
 	read_cycle  = calc_mck_cycles(read_cycle, mck_hz);
 	nrd_setup   = calc_mck_cycles(nrd_setup, mck_hz);
 	nrd_pulse   = calc_mck_cycles(nrd_pulse, mck_hz);
 	nrd_recover = calc_mck_cycles(nrd_recover, mck_hz);
 
-	clk_put(mck);
-
 	active  = nrd_setup + nrd_pulse;
 	recover = read_cycle - active;
 
@@ -121,13 +116,13 @@ static void set_smc_timing(struct device *dev,
 	ncs_write_setup = ncs_read_setup;
 	ncs_write_pulse = ncs_read_pulse;
 
-	dev_dbg(dev, "ATA timings: nrd_setup = %d nrd_pulse = %d nrd_cycle = %d\n",
+	dev_dbg(dev, "ATA timings: nrd_setup = %lu nrd_pulse = %lu nrd_cycle = %lu\n",
 			nrd_setup, nrd_pulse, read_cycle);
-	dev_dbg(dev, "ATA timings: nwe_setup = %d nwe_pulse = %d nwe_cycle = %d\n",
+	dev_dbg(dev, "ATA timings: nwe_setup = %lu nwe_pulse = %lu nwe_cycle = %lu\n",
 			nwe_setup, nwe_pulse, write_cycle);
-	dev_dbg(dev, "ATA timings: ncs_read_setup = %d ncs_read_pulse = %d\n",
+	dev_dbg(dev, "ATA timings: ncs_read_setup = %lu ncs_read_pulse = %lu\n",
 			ncs_read_setup, ncs_read_pulse);
-	dev_dbg(dev, "ATA timings: ncs_write_setup = %d ncs_write_pulse = %d\n",
+	dev_dbg(dev, "ATA timings: ncs_write_setup = %lu ncs_write_pulse = %lu\n",
 			ncs_write_setup, ncs_write_pulse);
 
 	at91_sys_write(AT91_SMC_SETUP(info->cs),
@@ -217,6 +212,7 @@ static int __devinit pata_at91_probe(struct platform_device *pdev)
 	struct resource *mem_res;
 	struct ata_host *host;
 	struct ata_port *ap;
+
 	int irq_flags = 0;
 	int irq = 0;
 	int ret;
@@ -261,6 +257,13 @@ static int __devinit pata_at91_probe(struct platform_device *pdev)
 		return -ENOMEM;
 	}
 
+	info->mck = clk_get(NULL, "mck");
+
+	if (IS_ERR(info->mck)) {
+		dev_err(dev, "failed to get access to mck clock\n");
+		return -ENODEV;
+	}
+
 	info->cs    = board->chipselect;
 	info->mode  = AT91_SMC_READMODE | AT91_SMC_WRITEMODE |
 		AT91_SMC_EXNWMODE_READY | AT91_SMC_BAT_SELECT |
@@ -304,6 +307,7 @@ err_alt_ioremap:
 	devm_iounmap(dev, info->ide_addr);
 
 err_ide_ioremap:
+	clk_put(info->mck);
 	kfree(info);
 
 	return ret;
@@ -326,6 +330,7 @@ static int __devexit pata_at91_remove(struct platform_device *pdev)
 
 	devm_iounmap(dev, info->ide_addr);
 	devm_iounmap(dev, info->alt_addr);
+	clk_put(info->mck);
 
 	kfree(info);
 	return 0;
diff --git a/drivers/ata/pata_octeon_cf.c b/drivers/ata/pata_octeon_cf.c
index 8d9343accf3..abdd19fe990 100644
--- a/drivers/ata/pata_octeon_cf.c
+++ b/drivers/ata/pata_octeon_cf.c
@@ -653,7 +653,8 @@ static irqreturn_t octeon_cf_interrupt(int irq, void *dev_instance)
 
 		ap = host->ports[i];
 		ocd = ap->dev->platform_data;
-		if (!ap || (ap->flags & ATA_FLAG_DISABLED))
+
+		if (ap->flags & ATA_FLAG_DISABLED)
 			continue;
 
 		ocd = ap->dev->platform_data;
diff --git a/drivers/ata/pata_pcmcia.c b/drivers/ata/pata_pcmcia.c
index f4d009ed50a..dc99e26f8e5 100644
--- a/drivers/ata/pata_pcmcia.c
+++ b/drivers/ata/pata_pcmcia.c
@@ -411,6 +411,7 @@ static struct pcmcia_device_id pcmcia_devices[] = {
 	PCMCIA_DEVICE_PROD_ID123("PCMCIA", "IDE CARD", "F1", 0x281f1c5d, 0x1907960c, 0xf7fde8b9),
 	PCMCIA_DEVICE_PROD_ID12("ARGOSY", "CD-ROM", 0x78f308dc, 0x66536591),
 	PCMCIA_DEVICE_PROD_ID12("ARGOSY", "PnPIDE", 0x78f308dc, 0x0c694728),
+	PCMCIA_DEVICE_PROD_ID12("CNF   ", "CD-ROM", 0x46d7db81, 0x66536591),
 	PCMCIA_DEVICE_PROD_ID12("CNF CD-M", "CD-ROM", 0x7d93b852, 0x66536591),
 	PCMCIA_DEVICE_PROD_ID12("Creative Technology Ltd.", "PCMCIA CD-ROM Interface Card", 0xff8c8a45, 0xfe8020c4),
 	PCMCIA_DEVICE_PROD_ID12("Digital Equipment Corporation.", "Digital Mobile Media CD-ROM", 0x17692a66, 0xef1dcbde),
diff --git a/drivers/ata/sata_mv.c b/drivers/ata/sata_mv.c
index 23714aefb82..c19417e0220 100644
--- a/drivers/ata/sata_mv.c
+++ b/drivers/ata/sata_mv.c
@@ -2514,7 +2514,7 @@ static void mv_unexpected_intr(struct ata_port *ap, int edma_was_enabled)
 	char *when = "idle";
 
 	ata_ehi_clear_desc(ehi);
-	if (!ap || (ap->flags & ATA_FLAG_DISABLED)) {
+	if (ap->flags & ATA_FLAG_DISABLED) {
 		when = "disabled";
 	} else if (edma_was_enabled) {
 		when = "EDMA enabled";
diff --git a/drivers/ata/sata_sil.c b/drivers/ata/sata_sil.c
index 030ec079b18..35bd5cc7f28 100644
--- a/drivers/ata/sata_sil.c
+++ b/drivers/ata/sata_sil.c
@@ -532,7 +532,7 @@ static irqreturn_t sil_interrupt(int irq, void *dev_instance)
 		struct ata_port *ap = host->ports[i];
 		u32 bmdma2 = readl(mmio_base + sil_port[ap->port_no].bmdma2);
 
-		if (unlikely(!ap || ap->flags & ATA_FLAG_DISABLED))
+		if (unlikely(ap->flags & ATA_FLAG_DISABLED))
 			continue;
 
 		/* turn off SATA_IRQ if not supported */
diff --git a/drivers/base/firmware_class.c b/drivers/base/firmware_class.c
index f285f441fab..7376367bcb8 100644
--- a/drivers/base/firmware_class.c
+++ b/drivers/base/firmware_class.c
@@ -180,7 +180,6 @@ static ssize_t firmware_loading_store(struct device *dev,
 				goto err;
 			}
 			/* Pages will be freed by vfree() */
-			fw_priv->pages = NULL;
 			fw_priv->page_array_size = 0;
 			fw_priv->nr_pages = 0;
 			complete(&fw_priv->completion);
diff --git a/drivers/base/sys.c b/drivers/base/sys.c
index 79a9ae5238a..0d903909af7 100644
--- a/drivers/base/sys.c
+++ b/drivers/base/sys.c
@@ -275,9 +275,9 @@ int sysdev_register(struct sys_device *sysdev)
 				drv->add(sysdev);
 		}
 		mutex_unlock(&sysdev_drivers_lock);
+		kobject_uevent(&sysdev->kobj, KOBJ_ADD);
 	}
 
-	kobject_uevent(&sysdev->kobj, KOBJ_ADD);
 	return error;
 }
 
diff --git a/drivers/block/mg_disk.c b/drivers/block/mg_disk.c
index f703f547824..6d7fbaa9224 100644
--- a/drivers/block/mg_disk.c
+++ b/drivers/block/mg_disk.c
@@ -36,7 +36,6 @@
 
 /* Register offsets */
 #define MG_BUFF_OFFSET			0x8000
-#define MG_STORAGE_BUFFER_SIZE		0x200
 #define MG_REG_OFFSET			0xC000
 #define MG_REG_FEATURE			(MG_REG_OFFSET + 2)	/* write case */
 #define MG_REG_ERROR			(MG_REG_OFFSET + 2)	/* read case */
@@ -219,6 +218,16 @@ static unsigned int mg_wait(struct mg_host *host, u32 expect, u32 msec)
 	host->error = MG_ERR_NONE;
 	expire = jiffies + msecs_to_jiffies(msec);
 
+	/* These 2 times dummy status read prevents reading invalid
+	 * status. A very little time (3 times of mflash operating clk)
+	 * is required for busy bit is set. Use dummy read instead of
+	 * busy wait, because mflash's PLL is machine dependent.
+	 */
+	if (prv_data->use_polling) {
+		status = inb((unsigned long)host->dev_base + MG_REG_STATUS);
+		status = inb((unsigned long)host->dev_base + MG_REG_STATUS);
+	}
+
 	status = inb((unsigned long)host->dev_base + MG_REG_STATUS);
 
 	do {
@@ -245,8 +254,6 @@ static unsigned int mg_wait(struct mg_host *host, u32 expect, u32 msec)
 			mg_dump_status("not ready", status, host);
 			return MG_ERR_INV_STAT;
 		}
-		if (prv_data->use_polling)
-			msleep(1);
 
 		status = inb((unsigned long)host->dev_base + MG_REG_STATUS);
 	} while (time_before(cur_jiffies, expire));
@@ -469,9 +476,18 @@ static unsigned int mg_out(struct mg_host *host,
 	return MG_ERR_NONE;
 }
 
+static void mg_read_one(struct mg_host *host, struct request *req)
+{
+	u16 *buff = (u16 *)req->buffer;
+	u32 i;
+
+	for (i = 0; i < MG_SECTOR_SIZE >> 1; i++)
+		*buff++ = inw((unsigned long)host->dev_base + MG_BUFF_OFFSET +
+			      (i << 1));
+}
+
 static void mg_read(struct request *req)
 {
-	u32 j;
 	struct mg_host *host = req->rq_disk->private_data;
 
 	if (mg_out(host, blk_rq_pos(req), blk_rq_sectors(req),
@@ -482,49 +498,65 @@ static void mg_read(struct request *req)
 	       blk_rq_sectors(req), blk_rq_pos(req), req->buffer);
 
 	do {
-		u16 *buff = (u16 *)req->buffer;
-
 		if (mg_wait(host, ATA_DRQ,
 			    MG_TMAX_WAIT_RD_DRQ) != MG_ERR_NONE) {
 			mg_bad_rw_intr(host);
 			return;
 		}
-		for (j = 0; j < MG_SECTOR_SIZE >> 1; j++)
-			*buff++ = inw((unsigned long)host->dev_base +
-				      MG_BUFF_OFFSET + (j << 1));
+
+		mg_read_one(host, req);
 
 		outb(MG_CMD_RD_CONF, (unsigned long)host->dev_base +
 				MG_REG_COMMAND);
 	} while (mg_end_request(host, 0, MG_SECTOR_SIZE));
 }
 
+static void mg_write_one(struct mg_host *host, struct request *req)
+{
+	u16 *buff = (u16 *)req->buffer;
+	u32 i;
+
+	for (i = 0; i < MG_SECTOR_SIZE >> 1; i++)
+		outw(*buff++, (unsigned long)host->dev_base + MG_BUFF_OFFSET +
+		     (i << 1));
+}
+
 static void mg_write(struct request *req)
 {
-	u32 j;
 	struct mg_host *host = req->rq_disk->private_data;
+	unsigned int rem = blk_rq_sectors(req);
 
-	if (mg_out(host, blk_rq_pos(req), blk_rq_sectors(req),
+	if (mg_out(host, blk_rq_pos(req), rem,
 		   MG_CMD_WR, NULL) != MG_ERR_NONE) {
 		mg_bad_rw_intr(host);
 		return;
 	}
 
 	MG_DBG("requested %d sects (from %ld), buffer=0x%p\n",
-	       blk_rq_sectors(req), blk_rq_pos(req), req->buffer);
+	       rem, blk_rq_pos(req), req->buffer);
+
+	if (mg_wait(host, ATA_DRQ,
+		    MG_TMAX_WAIT_WR_DRQ) != MG_ERR_NONE) {
+		mg_bad_rw_intr(host);
+		return;
+	}
 
 	do {
-		u16 *buff = (u16 *)req->buffer;
+		mg_write_one(host, req);
 
-	if (mg_wait(host, ATA_DRQ, MG_TMAX_WAIT_WR_DRQ) != MG_ERR_NONE) {
+		outb(MG_CMD_WR_CONF, (unsigned long)host->dev_base +
+				MG_REG_COMMAND);
+
+		rem--;
+		if (rem > 1 && mg_wait(host, ATA_DRQ,
+					MG_TMAX_WAIT_WR_DRQ) != MG_ERR_NONE) {
+			mg_bad_rw_intr(host);
+			return;
+		} else if (mg_wait(host, MG_STAT_READY,
+					MG_TMAX_WAIT_WR_DRQ) != MG_ERR_NONE) {
 			mg_bad_rw_intr(host);
 			return;
 		}
-		for (j = 0; j < MG_SECTOR_SIZE >> 1; j++)
-			outw(*buff++, (unsigned long)host->dev_base +
-				      MG_BUFF_OFFSET + (j << 1));
-
-		outb(MG_CMD_WR_CONF, (unsigned long)host->dev_base +
-				MG_REG_COMMAND);
 	} while (mg_end_request(host, 0, MG_SECTOR_SIZE));
 }
 
@@ -532,7 +564,6 @@ static void mg_read_intr(struct mg_host *host)
 {
 	struct request *req = host->req;
 	u32 i;
-	u16 *buff;
 
 	/* check status */
 	do {
@@ -550,13 +581,7 @@ static void mg_read_intr(struct mg_host *host)
 	return;
 
 ok_to_read:
-	/* get current segment of request */
-	buff = (u16 *)req->buffer;
-
-	/* read 1 sector */
-	for (i = 0; i < MG_SECTOR_SIZE >> 1; i++)
-		*buff++ = inw((unsigned long)host->dev_base + MG_BUFF_OFFSET +
-			      (i << 1));
+	mg_read_one(host, req);
 
 	MG_DBG("sector %ld, remaining=%ld, buffer=0x%p\n",
 	       blk_rq_pos(req), blk_rq_sectors(req) - 1, req->buffer);
@@ -575,8 +600,7 @@ ok_to_read:
 static void mg_write_intr(struct mg_host *host)
 {
 	struct request *req = host->req;
-	u32 i, j;
-	u16 *buff;
+	u32 i;
 	bool rem;
 
 	/* check status */
@@ -597,12 +621,7 @@ static void mg_write_intr(struct mg_host *host)
 ok_to_write:
 	if ((rem = mg_end_request(host, 0, MG_SECTOR_SIZE))) {
 		/* write 1 sector and set handler if remains */
-		buff = (u16 *)req->buffer;
-		for (j = 0; j < MG_STORAGE_BUFFER_SIZE >> 1; j++) {
-			outw(*buff, (unsigned long)host->dev_base +
-					MG_BUFF_OFFSET + (j << 1));
-			buff++;
-		}
+		mg_write_one(host, req);
 		MG_DBG("sector %ld, remaining=%ld, buffer=0x%p\n",
 		       blk_rq_pos(req), blk_rq_sectors(req), req->buffer);
 		host->mg_do_intr = mg_write_intr;
@@ -667,9 +686,6 @@ static unsigned int mg_issue_req(struct request *req,
 		unsigned int sect_num,
 		unsigned int sect_cnt)
 {
-	u16 *buff;
-	u32 i;
-
 	switch (rq_data_dir(req)) {
 	case READ:
 		if (mg_out(host, sect_num, sect_cnt, MG_CMD_RD, &mg_read_intr)
@@ -693,12 +709,7 @@ static unsigned int mg_issue_req(struct request *req,
 			mg_bad_rw_intr(host);
 			return host->error;
 		}
-		buff = (u16 *)req->buffer;
-		for (i = 0; i < MG_SECTOR_SIZE >> 1; i++) {
-			outw(*buff, (unsigned long)host->dev_base +
-					MG_BUFF_OFFSET + (i << 1));
-			buff++;
-		}
+		mg_write_one(host, req);
 		mod_timer(&host->timer, jiffies + 3 * HZ);
 		outb(MG_CMD_WR_CONF, (unsigned long)host->dev_base +
 				MG_REG_COMMAND);
diff --git a/drivers/char/agp/parisc-agp.c b/drivers/char/agp/parisc-agp.c
index f4bb43fb801..e077701ae3d 100644
--- a/drivers/char/agp/parisc-agp.c
+++ b/drivers/char/agp/parisc-agp.c
@@ -225,7 +225,7 @@ static const struct agp_bridge_driver parisc_agp_driver = {
 	.configure		= parisc_agp_configure,
 	.fetch_size		= parisc_agp_fetch_size,
 	.tlb_flush		= parisc_agp_tlbflush,
-	.mask_memory		= parisc_agp_mask_memory,
+	.mask_memory		= parisc_agp_page_mask_memory,
 	.masks			= parisc_agp_masks,
 	.agp_enable		= parisc_agp_enable,
 	.cache_flush		= global_cache_flush,
diff --git a/drivers/char/n_tty.c b/drivers/char/n_tty.c
index ff47907ff1b..973be2f4419 100644
--- a/drivers/char/n_tty.c
+++ b/drivers/char/n_tty.c
@@ -1583,6 +1583,7 @@ static int n_tty_open(struct tty_struct *tty)
 
 static inline int input_available_p(struct tty_struct *tty, int amt)
 {
+	tty_flush_to_ldisc(tty);
 	if (tty->icanon) {
 		if (tty->canon_data)
 			return 1;
diff --git a/drivers/char/pty.c b/drivers/char/pty.c
index 3850a68f265..6e6942c45f5 100644
--- a/drivers/char/pty.c
+++ b/drivers/char/pty.c
@@ -52,7 +52,6 @@ static void pty_close(struct tty_struct *tty, struct file *filp)
 		return;
 	tty->link->packet = 0;
 	set_bit(TTY_OTHER_CLOSED, &tty->link->flags);
-	tty_flip_buffer_push(tty->link);
 	wake_up_interruptible(&tty->link->read_wait);
 	wake_up_interruptible(&tty->link->write_wait);
 	if (tty->driver->subtype == PTY_TYPE_MASTER) {
@@ -208,7 +207,6 @@ static int pty_open(struct tty_struct *tty, struct file *filp)
 	clear_bit(TTY_OTHER_CLOSED, &tty->link->flags);
 	set_bit(TTY_THROTTLED, &tty->flags);
 	retval = 0;
-	tty->low_latency = 1;
 out:
 	return retval;
 }
diff --git a/drivers/char/sysrq.c b/drivers/char/sysrq.c
index 0db35857e4d..5d7a02f63e1 100644
--- a/drivers/char/sysrq.c
+++ b/drivers/char/sysrq.c
@@ -35,7 +35,6 @@
 #include <linux/spinlock.h>
 #include <linux/vt_kern.h>
 #include <linux/workqueue.h>
-#include <linux/kexec.h>
 #include <linux/hrtimer.h>
 #include <linux/oom.h>
 
@@ -124,9 +123,12 @@ static struct sysrq_key_op sysrq_unraw_op = {
 static void sysrq_handle_crash(int key, struct tty_struct *tty)
 {
 	char *killer = NULL;
+
+	panic_on_oops = 1;	/* force panic */
+	wmb();
 	*killer = 1;
 }
-static struct sysrq_key_op sysrq_crashdump_op = {
+static struct sysrq_key_op sysrq_crash_op = {
 	.handler	= sysrq_handle_crash,
 	.help_msg	= "Crash",
 	.action_msg	= "Trigger a crash",
@@ -401,7 +403,7 @@ static struct sysrq_key_op *sysrq_key_table[36] = {
 	 */
 	NULL,				/* a */
 	&sysrq_reboot_op,		/* b */
-	&sysrq_crashdump_op,		/* c & ibm_emac driver debug */
+	&sysrq_crash_op,		/* c & ibm_emac driver debug */
 	&sysrq_showlocks_op,		/* d */
 	&sysrq_term_op,			/* e */
 	&sysrq_moom_op,			/* f */
diff --git a/drivers/char/tty_buffer.c b/drivers/char/tty_buffer.c
index 810ee25d66a..3108991c5c8 100644
--- a/drivers/char/tty_buffer.c
+++ b/drivers/char/tty_buffer.c
@@ -462,6 +462,19 @@ static void flush_to_ldisc(struct work_struct *work)
 }
 
 /**
+ *	tty_flush_to_ldisc
+ *	@tty: tty to push
+ *
+ *	Push the terminal flip buffers to the line discipline.
+ *
+ *	Must not be called from IRQ context.
+ */
+void tty_flush_to_ldisc(struct tty_struct *tty)
+{
+	flush_to_ldisc(&tty->buf.work.work);
+}
+
+/**
  *	tty_flip_buffer_push	-	terminal
  *	@tty: tty to push
  *
diff --git a/drivers/char/tty_ldisc.c b/drivers/char/tty_ldisc.c
index acd76b767d4..1733d3439ad 100644
--- a/drivers/char/tty_ldisc.c
+++ b/drivers/char/tty_ldisc.c
@@ -48,6 +48,41 @@ static DECLARE_WAIT_QUEUE_HEAD(tty_ldisc_wait);
 /* Line disc dispatch table */
 static struct tty_ldisc_ops *tty_ldiscs[NR_LDISCS];
 
+static inline struct tty_ldisc *get_ldisc(struct tty_ldisc *ld)
+{
+	if (ld)
+		atomic_inc(&ld->users);
+	return ld;
+}
+
+static void put_ldisc(struct tty_ldisc *ld)
+{
+	unsigned long flags;
+
+	if (WARN_ON_ONCE(!ld))
+		return;
+
+	/*
+	 * If this is the last user, free the ldisc, and
+	 * release the ldisc ops.
+	 *
+	 * We really want an "atomic_dec_and_lock_irqsave()",
+	 * but we don't have it, so this does it by hand.
+	 */
+	local_irq_save(flags);
+	if (atomic_dec_and_lock(&ld->users, &tty_ldisc_lock)) {
+		struct tty_ldisc_ops *ldo = ld->ops;
+
+		ldo->refcount--;
+		module_put(ldo->owner);
+		spin_unlock_irqrestore(&tty_ldisc_lock, flags);
+
+		kfree(ld);
+		return;
+	}
+	local_irq_restore(flags);
+}
+
 /**
  *	tty_register_ldisc	-	install a line discipline
  *	@disc: ldisc number
@@ -142,7 +177,7 @@ static struct tty_ldisc *tty_ldisc_try_get(int disc)
 			/* lock it */
 			ldops->refcount++;
 			ld->ops = ldops;
-			ld->refcount = 0;
+			atomic_set(&ld->users, 1);
 			err = 0;
 		}
 	}
@@ -181,35 +216,6 @@ static struct tty_ldisc *tty_ldisc_get(int disc)
 	return ld;
 }
 
-/**
- *	tty_ldisc_put		-	drop ldisc reference
- *	@ld: ldisc
- *
- *	Drop a reference to a line discipline. Manage refcounts and
- *	module usage counts. Free the ldisc once the recount hits zero.
- *
- *	Locking:
- *		takes tty_ldisc_lock to guard against ldisc races
- */
-
-static void tty_ldisc_put(struct tty_ldisc *ld)
-{
-	unsigned long flags;
-	int disc = ld->ops->num;
-	struct tty_ldisc_ops *ldo;
-
-	BUG_ON(disc < N_TTY || disc >= NR_LDISCS);
-
-	spin_lock_irqsave(&tty_ldisc_lock, flags);
-	ldo = tty_ldiscs[disc];
-	BUG_ON(ldo->refcount == 0);
-	ldo->refcount--;
-	module_put(ldo->owner);
-	spin_unlock_irqrestore(&tty_ldisc_lock, flags);
-	WARN_ON(ld->refcount);
-	kfree(ld);
-}
-
 static void *tty_ldiscs_seq_start(struct seq_file *m, loff_t *pos)
 {
 	return (*pos < NR_LDISCS) ? pos : NULL;
@@ -234,7 +240,7 @@ static int tty_ldiscs_seq_show(struct seq_file *m, void *v)
 	if (IS_ERR(ld))
 		return 0;
 	seq_printf(m, "%-10s %2d\n", ld->ops->name ? ld->ops->name : "???", i);
-	tty_ldisc_put(ld);
+	put_ldisc(ld);
 	return 0;
 }
 
@@ -288,20 +294,17 @@ static void tty_ldisc_assign(struct tty_struct *tty, struct tty_ldisc *ld)
  *	Locking: takes tty_ldisc_lock
  */
 
-static int tty_ldisc_try(struct tty_struct *tty)
+static struct tty_ldisc *tty_ldisc_try(struct tty_struct *tty)
 {
 	unsigned long flags;
 	struct tty_ldisc *ld;
-	int ret = 0;
 
 	spin_lock_irqsave(&tty_ldisc_lock, flags);
-	ld = tty->ldisc;
-	if (test_bit(TTY_LDISC, &tty->flags)) {
-		ld->refcount++;
-		ret = 1;
-	}
+	ld = NULL;
+	if (test_bit(TTY_LDISC, &tty->flags))
+		ld = get_ldisc(tty->ldisc);
 	spin_unlock_irqrestore(&tty_ldisc_lock, flags);
-	return ret;
+	return ld;
 }
 
 /**
@@ -322,10 +325,11 @@ static int tty_ldisc_try(struct tty_struct *tty)
 
 struct tty_ldisc *tty_ldisc_ref_wait(struct tty_struct *tty)
 {
+	struct tty_ldisc *ld;
+
 	/* wait_event is a macro */
-	wait_event(tty_ldisc_wait, tty_ldisc_try(tty));
-	WARN_ON(tty->ldisc->refcount == 0);
-	return tty->ldisc;
+	wait_event(tty_ldisc_wait, (ld = tty_ldisc_try(tty)) != NULL);
+	return ld;
 }
 EXPORT_SYMBOL_GPL(tty_ldisc_ref_wait);
 
@@ -342,9 +346,7 @@ EXPORT_SYMBOL_GPL(tty_ldisc_ref_wait);
 
 struct tty_ldisc *tty_ldisc_ref(struct tty_struct *tty)
 {
-	if (tty_ldisc_try(tty))
-		return tty->ldisc;
-	return NULL;
+	return tty_ldisc_try(tty);
 }
 EXPORT_SYMBOL_GPL(tty_ldisc_ref);
 
@@ -360,21 +362,15 @@ EXPORT_SYMBOL_GPL(tty_ldisc_ref);
 
 void tty_ldisc_deref(struct tty_ldisc *ld)
 {
-	unsigned long flags;
-
-	BUG_ON(ld == NULL);
-
-	spin_lock_irqsave(&tty_ldisc_lock, flags);
-	if (ld->refcount == 0)
-		printk(KERN_ERR "tty_ldisc_deref: no references.\n");
-	else
-		ld->refcount--;
-	if (ld->refcount == 0)
-		wake_up(&tty_ldisc_wait);
-	spin_unlock_irqrestore(&tty_ldisc_lock, flags);
+	put_ldisc(ld);
 }
 EXPORT_SYMBOL_GPL(tty_ldisc_deref);
 
+static inline void tty_ldisc_put(struct tty_ldisc *ld)
+{
+	put_ldisc(ld);
+}
+
 /**
  *	tty_ldisc_enable	-	allow ldisc use
  *	@tty: terminal to activate ldisc on
@@ -523,31 +519,6 @@ static int tty_ldisc_halt(struct tty_struct *tty)
 }
 
 /**
- *	tty_ldisc_wait_idle	-	wait for the ldisc to become idle
- *	@tty: tty to wait for
- *
- *	Wait for the line discipline to become idle. The discipline must
- *	have been halted for this to guarantee it remains idle.
- *
- *	tty_ldisc_lock protects the ref counts currently.
- */
-
-static int tty_ldisc_wait_idle(struct tty_struct *tty)
-{
-	unsigned long flags;
-	spin_lock_irqsave(&tty_ldisc_lock, flags);
-	while (tty->ldisc->refcount) {
-		spin_unlock_irqrestore(&tty_ldisc_lock, flags);
-		if (wait_event_timeout(tty_ldisc_wait,
-				tty->ldisc->refcount == 0, 5 * HZ) == 0)
-			return -EBUSY;
-		spin_lock_irqsave(&tty_ldisc_lock, flags);
-	}
-	spin_unlock_irqrestore(&tty_ldisc_lock, flags);
-	return 0;
-}
-
-/**
  *	tty_set_ldisc		-	set line discipline
  *	@tty: the terminal to set
  *	@ldisc: the line discipline
@@ -642,14 +613,6 @@ int tty_set_ldisc(struct tty_struct *tty, int ldisc)
 
 	flush_scheduled_work();
 
-	/* Let any existing reference holders finish */
-	retval = tty_ldisc_wait_idle(tty);
-	if (retval < 0) {
-		clear_bit(TTY_LDISC_CHANGING, &tty->flags);
-		tty_ldisc_put(new_ldisc);
-		return retval;
-	}
-
 	mutex_lock(&tty->ldisc_mutex);
 	if (test_bit(TTY_HUPPED, &tty->flags)) {
 		/* We were raced by the hangup method. It will have stomped
@@ -795,7 +758,6 @@ void tty_ldisc_hangup(struct tty_struct *tty)
 		if (tty->ldisc) {	/* Not yet closed */
 			/* Switch back to N_TTY */
 			tty_ldisc_halt(tty);
-			tty_ldisc_wait_idle(tty);
 			tty_ldisc_reinit(tty);
 			/* At this point we have a closed ldisc and we want to
 			   reopen it. We could defer this to the next open but
@@ -860,14 +822,6 @@ void tty_ldisc_release(struct tty_struct *tty, struct tty_struct *o_tty)
 	tty_ldisc_halt(tty);
 	flush_scheduled_work();
 
-	/*
-	 * Wait for any short term users (we know they are just driver
-	 * side waiters as the file is closing so user count on the file
-	 * side is zero.
-	 */
-
-	tty_ldisc_wait_idle(tty);
-
 	mutex_lock(&tty->ldisc_mutex);
 	/*
 	 * Now kill off the ldisc
diff --git a/drivers/char/vr41xx_giu.c b/drivers/char/vr41xx_giu.c
deleted file mode 100644
index e69de29bb2d..00000000000
--- a/drivers/char/vr41xx_giu.c
+++ /dev/null
diff --git a/drivers/cpufreq/cpufreq.c b/drivers/cpufreq/cpufreq.c
index b90eda8b344..fd69086d08d 100644
--- a/drivers/cpufreq/cpufreq.c
+++ b/drivers/cpufreq/cpufreq.c
@@ -858,6 +858,8 @@ static int cpufreq_add_dev(struct sys_device *sys_dev)
 
 		/* Check for existing affected CPUs.
 		 * They may not be aware of it due to CPU Hotplug.
+		 * cpufreq_cpu_put is called when the device is removed
+		 * in __cpufreq_remove_dev()
 		 */
 		managed_policy = cpufreq_cpu_get(j);
 		if (unlikely(managed_policy)) {
@@ -884,7 +886,7 @@ static int cpufreq_add_dev(struct sys_device *sys_dev)
 			ret = sysfs_create_link(&sys_dev->kobj,
 						&managed_policy->kobj,
 						"cpufreq");
-			if (!ret)
+			if (ret)
 				cpufreq_cpu_put(managed_policy);
 			/*
 			 * Success. We only needed to be added to the mask.
@@ -924,6 +926,8 @@ static int cpufreq_add_dev(struct sys_device *sys_dev)
 
 	spin_lock_irqsave(&cpufreq_driver_lock, flags);
 	for_each_cpu(j, policy->cpus) {
+		if (!cpu_online(j))
+			continue;
 		per_cpu(cpufreq_cpu_data, j) = policy;
 		per_cpu(policy_cpu, j) = policy->cpu;
 	}
@@ -1244,13 +1248,22 @@ EXPORT_SYMBOL(cpufreq_get);
 
 static int cpufreq_suspend(struct sys_device *sysdev, pm_message_t pmsg)
 {
-	int cpu = sysdev->id;
 	int ret = 0;
+
+#ifdef __powerpc__
+	int cpu = sysdev->id;
 	unsigned int cur_freq = 0;
 	struct cpufreq_policy *cpu_policy;
 
 	dprintk("suspending cpu %u\n", cpu);
 
+	/*
+	 * This whole bogosity is here because Powerbooks are made of fail.
+	 * No sane platform should need any of the code below to be run.
+	 * (it's entirely the wrong thing to do, as driver->get may
+	 *  reenable interrupts on some architectures).
+	 */
+
 	if (!cpu_online(cpu))
 		return 0;
 
@@ -1309,6 +1322,7 @@ static int cpufreq_suspend(struct sys_device *sysdev, pm_message_t pmsg)
 
 out:
 	cpufreq_cpu_put(cpu_policy);
+#endif	/* __powerpc__ */
 	return ret;
 }
 
@@ -1322,12 +1336,18 @@ out:
  */
 static int cpufreq_resume(struct sys_device *sysdev)
 {
-	int cpu = sysdev->id;
 	int ret = 0;
+
+#ifdef __powerpc__
+	int cpu = sysdev->id;
 	struct cpufreq_policy *cpu_policy;
 
 	dprintk("resuming cpu %u\n", cpu);
 
+	/* As with the ->suspend method, all the code below is
+	 * only necessary because Powerbooks suck.
+	 * See commit 42d4dc3f4e1e for jokes. */
+
 	if (!cpu_online(cpu))
 		return 0;
 
@@ -1391,6 +1411,7 @@ out:
 	schedule_work(&cpu_policy->update);
 fail:
 	cpufreq_cpu_put(cpu_policy);
+#endif	/* __powerpc__ */
 	return ret;
 }
 
diff --git a/drivers/cpufreq/cpufreq_conservative.c b/drivers/cpufreq/cpufreq_conservative.c
index 57490502b21..bdea7e2f94b 100644
--- a/drivers/cpufreq/cpufreq_conservative.c
+++ b/drivers/cpufreq/cpufreq_conservative.c
@@ -63,6 +63,7 @@ struct cpu_dbs_info_s {
 	unsigned int down_skip;
 	unsigned int requested_freq;
 	int cpu;
+	unsigned int enable:1;
 	/*
 	 * percpu mutex that serializes governor limit change with
 	 * do_dbs_timer invocation. We do not want do_dbs_timer to run
@@ -141,6 +142,9 @@ dbs_cpufreq_notifier(struct notifier_block *nb, unsigned long val,
 
 	struct cpufreq_policy *policy;
 
+	if (!this_dbs_info->enable)
+		return 0;
+
 	policy = this_dbs_info->cur_policy;
 
 	/*
@@ -497,6 +501,7 @@ static inline void dbs_timer_init(struct cpu_dbs_info_s *dbs_info)
 	int delay = usecs_to_jiffies(dbs_tuners_ins.sampling_rate);
 	delay -= jiffies % delay;
 
+	dbs_info->enable = 1;
 	INIT_DELAYED_WORK_DEFERRABLE(&dbs_info->work, do_dbs_timer);
 	queue_delayed_work_on(dbs_info->cpu, kconservative_wq, &dbs_info->work,
 				delay);
@@ -504,6 +509,7 @@ static inline void dbs_timer_init(struct cpu_dbs_info_s *dbs_info)
 
 static inline void dbs_timer_exit(struct cpu_dbs_info_s *dbs_info)
 {
+	dbs_info->enable = 0;
 	cancel_delayed_work_sync(&dbs_info->work);
 }
 
diff --git a/drivers/dma/Kconfig b/drivers/dma/Kconfig
index 070357aaedb..81e1020fb51 100644
--- a/drivers/dma/Kconfig
+++ b/drivers/dma/Kconfig
@@ -4,7 +4,7 @@
 
 menuconfig DMADEVICES
 	bool "DMA Engine support"
-	depends on !HIGHMEM64G && HAS_DMA
+	depends on HAS_DMA
 	help
 	  DMA engines can do asynchronous data transfers without
 	  involving the host CPU.  Currently, this framework can be
@@ -46,6 +46,14 @@ config DW_DMAC
 	  Support the Synopsys DesignWare AHB DMA controller.  This
 	  can be integrated in chips such as the Atmel AT32ap7000.
 
+config AT_HDMAC
+	tristate "Atmel AHB DMA support"
+	depends on ARCH_AT91SAM9RL
+	select DMA_ENGINE
+	help
+	  Support the Atmel AHB DMA controller.  This can be integrated in
+	  chips such as the Atmel AT91SAM9RL.
+
 config FSL_DMA
 	tristate "Freescale Elo and Elo Plus DMA support"
 	depends on FSL_SOC
@@ -108,7 +116,7 @@ config NET_DMA
 
 config ASYNC_TX_DMA
 	bool "Async_tx: Offload support for the async_tx api"
-	depends on DMA_ENGINE
+	depends on DMA_ENGINE && !HIGHMEM64G
 	help
 	  This allows the async_tx api to take advantage of offload engines for
 	  memcpy, memset, xor, and raid6 p+q operations.  If your platform has
diff --git a/drivers/dma/Makefile b/drivers/dma/Makefile
index a0b6564800c..40e1e008357 100644
--- a/drivers/dma/Makefile
+++ b/drivers/dma/Makefile
@@ -7,5 +7,6 @@ obj-$(CONFIG_INTEL_IOP_ADMA) += iop-adma.o
 obj-$(CONFIG_FSL_DMA) += fsldma.o
 obj-$(CONFIG_MV_XOR) += mv_xor.o
 obj-$(CONFIG_DW_DMAC) += dw_dmac.o
+obj-$(CONFIG_AT_HDMAC) += at_hdmac.o
 obj-$(CONFIG_MX3_IPU) += ipu/
 obj-$(CONFIG_TXX9_DMAC) += txx9dmac.o
diff --git a/drivers/dma/at_hdmac.c b/drivers/dma/at_hdmac.c
new file mode 100644
index 00000000000..9a1e5fb412e
--- /dev/null
+++ b/drivers/dma/at_hdmac.c
@@ -0,0 +1,1213 @@
+/*
+ * Driver for the Atmel AHB DMA Controller (aka HDMA or DMAC on AT91 systems)
+ *
+ * Copyright (C) 2008 Atmel Corporation
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ *
+ * This supports the Atmel AHB DMA Controller,
+ *
+ * The driver has currently been tested with the Atmel AT91SAM9RL
+ * and AT91SAM9G45 series.
+ */
+
+#include <linux/clk.h>
+#include <linux/dmaengine.h>
+#include <linux/dma-mapping.h>
+#include <linux/dmapool.h>
+#include <linux/interrupt.h>
+#include <linux/module.h>
+#include <linux/platform_device.h>
+
+#include "at_hdmac_regs.h"
+
+/*
+ * Glossary
+ * --------
+ *
+ * at_hdmac		: Name of the ATmel AHB DMA Controller
+ * at_dma_ / atdma	: ATmel DMA controller entity related
+ * atc_	/ atchan	: ATmel DMA Channel entity related
+ */
+
+#define	ATC_DEFAULT_CFG		(ATC_FIFOCFG_HALFFIFO)
+#define	ATC_DEFAULT_CTRLA	(0)
+#define	ATC_DEFAULT_CTRLB	(ATC_SIF(0)	\
+				|ATC_DIF(1))
+
+/*
+ * Initial number of descriptors to allocate for each channel. This could
+ * be increased during dma usage.
+ */
+static unsigned int init_nr_desc_per_channel = 64;
+module_param(init_nr_desc_per_channel, uint, 0644);
+MODULE_PARM_DESC(init_nr_desc_per_channel,
+		 "initial descriptors per channel (default: 64)");
+
+
+/* prototypes */
+static dma_cookie_t atc_tx_submit(struct dma_async_tx_descriptor *tx);
+
+
+/*----------------------------------------------------------------------*/
+
+static struct at_desc *atc_first_active(struct at_dma_chan *atchan)
+{
+	return list_first_entry(&atchan->active_list,
+				struct at_desc, desc_node);
+}
+
+static struct at_desc *atc_first_queued(struct at_dma_chan *atchan)
+{
+	return list_first_entry(&atchan->queue,
+				struct at_desc, desc_node);
+}
+
+/**
+ * atc_alloc_descriptor - allocate and return an initilized descriptor
+ * @chan: the channel to allocate descriptors for
+ * @gfp_flags: GFP allocation flags
+ *
+ * Note: The ack-bit is positioned in the descriptor flag at creation time
+ *       to make initial allocation more convenient. This bit will be cleared
+ *       and control will be given to client at usage time (during
+ *       preparation functions).
+ */
+static struct at_desc *atc_alloc_descriptor(struct dma_chan *chan,
+					    gfp_t gfp_flags)
+{
+	struct at_desc	*desc = NULL;
+	struct at_dma	*atdma = to_at_dma(chan->device);
+	dma_addr_t phys;
+
+	desc = dma_pool_alloc(atdma->dma_desc_pool, gfp_flags, &phys);
+	if (desc) {
+		memset(desc, 0, sizeof(struct at_desc));
+		dma_async_tx_descriptor_init(&desc->txd, chan);
+		/* txd.flags will be overwritten in prep functions */
+		desc->txd.flags = DMA_CTRL_ACK;
+		desc->txd.tx_submit = atc_tx_submit;
+		desc->txd.phys = phys;
+	}
+
+	return desc;
+}
+
+/**
+ * atc_desc_get - get a unsused descriptor from free_list
+ * @atchan: channel we want a new descriptor for
+ */
+static struct at_desc *atc_desc_get(struct at_dma_chan *atchan)
+{
+	struct at_desc *desc, *_desc;
+	struct at_desc *ret = NULL;
+	unsigned int i = 0;
+	LIST_HEAD(tmp_list);
+
+	spin_lock_bh(&atchan->lock);
+	list_for_each_entry_safe(desc, _desc, &atchan->free_list, desc_node) {
+		i++;
+		if (async_tx_test_ack(&desc->txd)) {
+			list_del(&desc->desc_node);
+			ret = desc;
+			break;
+		}
+		dev_dbg(chan2dev(&atchan->chan_common),
+				"desc %p not ACKed\n", desc);
+	}
+	spin_unlock_bh(&atchan->lock);
+	dev_vdbg(chan2dev(&atchan->chan_common),
+		"scanned %u descriptors on freelist\n", i);
+
+	/* no more descriptor available in initial pool: create one more */
+	if (!ret) {
+		ret = atc_alloc_descriptor(&atchan->chan_common, GFP_ATOMIC);
+		if (ret) {
+			spin_lock_bh(&atchan->lock);
+			atchan->descs_allocated++;
+			spin_unlock_bh(&atchan->lock);
+		} else {
+			dev_err(chan2dev(&atchan->chan_common),
+					"not enough descriptors available\n");
+		}
+	}
+
+	return ret;
+}
+
+/**
+ * atc_desc_put - move a descriptor, including any children, to the free list
+ * @atchan: channel we work on
+ * @desc: descriptor, at the head of a chain, to move to free list
+ */
+static void atc_desc_put(struct at_dma_chan *atchan, struct at_desc *desc)
+{
+	if (desc) {
+		struct at_desc *child;
+
+		spin_lock_bh(&atchan->lock);
+		list_for_each_entry(child, &desc->txd.tx_list, desc_node)
+			dev_vdbg(chan2dev(&atchan->chan_common),
+					"moving child desc %p to freelist\n",
+					child);
+		list_splice_init(&desc->txd.tx_list, &atchan->free_list);
+		dev_vdbg(chan2dev(&atchan->chan_common),
+			 "moving desc %p to freelist\n", desc);
+		list_add(&desc->desc_node, &atchan->free_list);
+		spin_unlock_bh(&atchan->lock);
+	}
+}
+
+/**
+ * atc_assign_cookie - compute and assign new cookie
+ * @atchan: channel we work on
+ * @desc: descriptor to asign cookie for
+ *
+ * Called with atchan->lock held and bh disabled
+ */
+static dma_cookie_t
+atc_assign_cookie(struct at_dma_chan *atchan, struct at_desc *desc)
+{
+	dma_cookie_t cookie = atchan->chan_common.cookie;
+
+	if (++cookie < 0)
+		cookie = 1;
+
+	atchan->chan_common.cookie = cookie;
+	desc->txd.cookie = cookie;
+
+	return cookie;
+}
+
+/**
+ * atc_dostart - starts the DMA engine for real
+ * @atchan: the channel we want to start
+ * @first: first descriptor in the list we want to begin with
+ *
+ * Called with atchan->lock held and bh disabled
+ */
+static void atc_dostart(struct at_dma_chan *atchan, struct at_desc *first)
+{
+	struct at_dma	*atdma = to_at_dma(atchan->chan_common.device);
+
+	/* ASSERT:  channel is idle */
+	if (atc_chan_is_enabled(atchan)) {
+		dev_err(chan2dev(&atchan->chan_common),
+			"BUG: Attempted to start non-idle channel\n");
+		dev_err(chan2dev(&atchan->chan_common),
+			"  channel: s0x%x d0x%x ctrl0x%x:0x%x l0x%x\n",
+			channel_readl(atchan, SADDR),
+			channel_readl(atchan, DADDR),
+			channel_readl(atchan, CTRLA),
+			channel_readl(atchan, CTRLB),
+			channel_readl(atchan, DSCR));
+
+		/* The tasklet will hopefully advance the queue... */
+		return;
+	}
+
+	vdbg_dump_regs(atchan);
+
+	/* clear any pending interrupt */
+	while (dma_readl(atdma, EBCISR))
+		cpu_relax();
+
+	channel_writel(atchan, SADDR, 0);
+	channel_writel(atchan, DADDR, 0);
+	channel_writel(atchan, CTRLA, 0);
+	channel_writel(atchan, CTRLB, 0);
+	channel_writel(atchan, DSCR, first->txd.phys);
+	dma_writel(atdma, CHER, atchan->mask);
+
+	vdbg_dump_regs(atchan);
+}
+
+/**
+ * atc_chain_complete - finish work for one transaction chain
+ * @atchan: channel we work on
+ * @desc: descriptor at the head of the chain we want do complete
+ *
+ * Called with atchan->lock held and bh disabled */
+static void
+atc_chain_complete(struct at_dma_chan *atchan, struct at_desc *desc)
+{
+	dma_async_tx_callback		callback;
+	void				*param;
+	struct dma_async_tx_descriptor	*txd = &desc->txd;
+
+	dev_vdbg(chan2dev(&atchan->chan_common),
+		"descriptor %u complete\n", txd->cookie);
+
+	atchan->completed_cookie = txd->cookie;
+	callback = txd->callback;
+	param = txd->callback_param;
+
+	/* move children to free_list */
+	list_splice_init(&txd->tx_list, &atchan->free_list);
+	/* move myself to free_list */
+	list_move(&desc->desc_node, &atchan->free_list);
+
+	/* unmap dma addresses */
+	if (!(txd->flags & DMA_COMPL_SKIP_DEST_UNMAP)) {
+		if (txd->flags & DMA_COMPL_DEST_UNMAP_SINGLE)
+			dma_unmap_single(chan2parent(&atchan->chan_common),
+					desc->lli.daddr,
+					desc->len, DMA_FROM_DEVICE);
+		else
+			dma_unmap_page(chan2parent(&atchan->chan_common),
+					desc->lli.daddr,
+					desc->len, DMA_FROM_DEVICE);
+	}
+	if (!(txd->flags & DMA_COMPL_SKIP_SRC_UNMAP)) {
+		if (txd->flags & DMA_COMPL_SRC_UNMAP_SINGLE)
+			dma_unmap_single(chan2parent(&atchan->chan_common),
+					desc->lli.saddr,
+					desc->len, DMA_TO_DEVICE);
+		else
+			dma_unmap_page(chan2parent(&atchan->chan_common),
+					desc->lli.saddr,
+					desc->len, DMA_TO_DEVICE);
+	}
+
+	/*
+	 * The API requires that no submissions are done from a
+	 * callback, so we don't need to drop the lock here
+	 */
+	if (callback)
+		callback(param);
+
+	dma_run_dependencies(txd);
+}
+
+/**
+ * atc_complete_all - finish work for all transactions
+ * @atchan: channel to complete transactions for
+ *
+ * Eventually submit queued descriptors if any
+ *
+ * Assume channel is idle while calling this function
+ * Called with atchan->lock held and bh disabled
+ */
+static void atc_complete_all(struct at_dma_chan *atchan)
+{
+	struct at_desc *desc, *_desc;
+	LIST_HEAD(list);
+
+	dev_vdbg(chan2dev(&atchan->chan_common), "complete all\n");
+
+	BUG_ON(atc_chan_is_enabled(atchan));
+
+	/*
+	 * Submit queued descriptors ASAP, i.e. before we go through
+	 * the completed ones.
+	 */
+	if (!list_empty(&atchan->queue))
+		atc_dostart(atchan, atc_first_queued(atchan));
+	/* empty active_list now it is completed */
+	list_splice_init(&atchan->active_list, &list);
+	/* empty queue list by moving descriptors (if any) to active_list */
+	list_splice_init(&atchan->queue, &atchan->active_list);
+
+	list_for_each_entry_safe(desc, _desc, &list, desc_node)
+		atc_chain_complete(atchan, desc);
+}
+
+/**
+ * atc_cleanup_descriptors - cleanup up finished descriptors in active_list
+ * @atchan: channel to be cleaned up
+ *
+ * Called with atchan->lock held and bh disabled
+ */
+static void atc_cleanup_descriptors(struct at_dma_chan *atchan)
+{
+	struct at_desc	*desc, *_desc;
+	struct at_desc	*child;
+
+	dev_vdbg(chan2dev(&atchan->chan_common), "cleanup descriptors\n");
+
+	list_for_each_entry_safe(desc, _desc, &atchan->active_list, desc_node) {
+		if (!(desc->lli.ctrla & ATC_DONE))
+			/* This one is currently in progress */
+			return;
+
+		list_for_each_entry(child, &desc->txd.tx_list, desc_node)
+			if (!(child->lli.ctrla & ATC_DONE))
+				/* Currently in progress */
+				return;
+
+		/*
+		 * No descriptors so far seem to be in progress, i.e.
+		 * this chain must be done.
+		 */
+		atc_chain_complete(atchan, desc);
+	}
+}
+
+/**
+ * atc_advance_work - at the end of a transaction, move forward
+ * @atchan: channel where the transaction ended
+ *
+ * Called with atchan->lock held and bh disabled
+ */
+static void atc_advance_work(struct at_dma_chan *atchan)
+{
+	dev_vdbg(chan2dev(&atchan->chan_common), "advance_work\n");
+
+	if (list_empty(&atchan->active_list) ||
+	    list_is_singular(&atchan->active_list)) {
+		atc_complete_all(atchan);
+	} else {
+		atc_chain_complete(atchan, atc_first_active(atchan));
+		/* advance work */
+		atc_dostart(atchan, atc_first_active(atchan));
+	}
+}
+
+
+/**
+ * atc_handle_error - handle errors reported by DMA controller
+ * @atchan: channel where error occurs
+ *
+ * Called with atchan->lock held and bh disabled
+ */
+static void atc_handle_error(struct at_dma_chan *atchan)
+{
+	struct at_desc *bad_desc;
+	struct at_desc *child;
+
+	/*
+	 * The descriptor currently at the head of the active list is
+	 * broked. Since we don't have any way to report errors, we'll
+	 * just have to scream loudly and try to carry on.
+	 */
+	bad_desc = atc_first_active(atchan);
+	list_del_init(&bad_desc->desc_node);
+
+	/* As we are stopped, take advantage to push queued descriptors
+	 * in active_list */
+	list_splice_init(&atchan->queue, atchan->active_list.prev);
+
+	/* Try to restart the controller */
+	if (!list_empty(&atchan->active_list))
+		atc_dostart(atchan, atc_first_active(atchan));
+
+	/*
+	 * KERN_CRITICAL may seem harsh, but since this only happens
+	 * when someone submits a bad physical address in a
+	 * descriptor, we should consider ourselves lucky that the
+	 * controller flagged an error instead of scribbling over
+	 * random memory locations.
+	 */
+	dev_crit(chan2dev(&atchan->chan_common),
+			"Bad descriptor submitted for DMA!\n");
+	dev_crit(chan2dev(&atchan->chan_common),
+			"  cookie: %d\n", bad_desc->txd.cookie);
+	atc_dump_lli(atchan, &bad_desc->lli);
+	list_for_each_entry(child, &bad_desc->txd.tx_list, desc_node)
+		atc_dump_lli(atchan, &child->lli);
+
+	/* Pretend the descriptor completed successfully */
+	atc_chain_complete(atchan, bad_desc);
+}
+
+
+/*--  IRQ & Tasklet  ---------------------------------------------------*/
+
+static void atc_tasklet(unsigned long data)
+{
+	struct at_dma_chan *atchan = (struct at_dma_chan *)data;
+
+	/* Channel cannot be enabled here */
+	if (atc_chan_is_enabled(atchan)) {
+		dev_err(chan2dev(&atchan->chan_common),
+			"BUG: channel enabled in tasklet\n");
+		return;
+	}
+
+	spin_lock(&atchan->lock);
+	if (test_and_clear_bit(0, &atchan->error_status))
+		atc_handle_error(atchan);
+	else
+		atc_advance_work(atchan);
+
+	spin_unlock(&atchan->lock);
+}
+
+static irqreturn_t at_dma_interrupt(int irq, void *dev_id)
+{
+	struct at_dma		*atdma = (struct at_dma *)dev_id;
+	struct at_dma_chan	*atchan;
+	int			i;
+	u32			status, pending, imr;
+	int			ret = IRQ_NONE;
+
+	do {
+		imr = dma_readl(atdma, EBCIMR);
+		status = dma_readl(atdma, EBCISR);
+		pending = status & imr;
+
+		if (!pending)
+			break;
+
+		dev_vdbg(atdma->dma_common.dev,
+			"interrupt: status = 0x%08x, 0x%08x, 0x%08x\n",
+			 status, imr, pending);
+
+		for (i = 0; i < atdma->dma_common.chancnt; i++) {
+			atchan = &atdma->chan[i];
+			if (pending & (AT_DMA_CBTC(i) | AT_DMA_ERR(i))) {
+				if (pending & AT_DMA_ERR(i)) {
+					/* Disable channel on AHB error */
+					dma_writel(atdma, CHDR, atchan->mask);
+					/* Give information to tasklet */
+					set_bit(0, &atchan->error_status);
+				}
+				tasklet_schedule(&atchan->tasklet);
+				ret = IRQ_HANDLED;
+			}
+		}
+
+	} while (pending);
+
+	return ret;
+}
+
+
+/*--  DMA Engine API  --------------------------------------------------*/
+
+/**
+ * atc_tx_submit - set the prepared descriptor(s) to be executed by the engine
+ * @desc: descriptor at the head of the transaction chain
+ *
+ * Queue chain if DMA engine is working already
+ *
+ * Cookie increment and adding to active_list or queue must be atomic
+ */
+static dma_cookie_t atc_tx_submit(struct dma_async_tx_descriptor *tx)
+{
+	struct at_desc		*desc = txd_to_at_desc(tx);
+	struct at_dma_chan	*atchan = to_at_dma_chan(tx->chan);
+	dma_cookie_t		cookie;
+
+	spin_lock_bh(&atchan->lock);
+	cookie = atc_assign_cookie(atchan, desc);
+
+	if (list_empty(&atchan->active_list)) {
+		dev_vdbg(chan2dev(tx->chan), "tx_submit: started %u\n",
+				desc->txd.cookie);
+		atc_dostart(atchan, desc);
+		list_add_tail(&desc->desc_node, &atchan->active_list);
+	} else {
+		dev_vdbg(chan2dev(tx->chan), "tx_submit: queued %u\n",
+				desc->txd.cookie);
+		list_add_tail(&desc->desc_node, &atchan->queue);
+	}
+
+	spin_unlock_bh(&atchan->lock);
+
+	return cookie;
+}
+
+/**
+ * atc_prep_dma_memcpy - prepare a memcpy operation
+ * @chan: the channel to prepare operation on
+ * @dest: operation virtual destination address
+ * @src: operation virtual source address
+ * @len: operation length
+ * @flags: tx descriptor status flags
+ */
+static struct dma_async_tx_descriptor *
+atc_prep_dma_memcpy(struct dma_chan *chan, dma_addr_t dest, dma_addr_t src,
+		size_t len, unsigned long flags)
+{
+	struct at_dma_chan	*atchan = to_at_dma_chan(chan);
+	struct at_desc		*desc = NULL;
+	struct at_desc		*first = NULL;
+	struct at_desc		*prev = NULL;
+	size_t			xfer_count;
+	size_t			offset;
+	unsigned int		src_width;
+	unsigned int		dst_width;
+	u32			ctrla;
+	u32			ctrlb;
+
+	dev_vdbg(chan2dev(chan), "prep_dma_memcpy: d0x%x s0x%x l0x%zx f0x%lx\n",
+			dest, src, len, flags);
+
+	if (unlikely(!len)) {
+		dev_dbg(chan2dev(chan), "prep_dma_memcpy: length is zero!\n");
+		return NULL;
+	}
+
+	ctrla =   ATC_DEFAULT_CTRLA;
+	ctrlb =   ATC_DEFAULT_CTRLB
+		| ATC_SRC_ADDR_MODE_INCR
+		| ATC_DST_ADDR_MODE_INCR
+		| ATC_FC_MEM2MEM;
+
+	/*
+	 * We can be a lot more clever here, but this should take care
+	 * of the most common optimization.
+	 */
+	if (!((src | dest  | len) & 3)) {
+		ctrla |= ATC_SRC_WIDTH_WORD | ATC_DST_WIDTH_WORD;
+		src_width = dst_width = 2;
+	} else if (!((src | dest | len) & 1)) {
+		ctrla |= ATC_SRC_WIDTH_HALFWORD | ATC_DST_WIDTH_HALFWORD;
+		src_width = dst_width = 1;
+	} else {
+		ctrla |= ATC_SRC_WIDTH_BYTE | ATC_DST_WIDTH_BYTE;
+		src_width = dst_width = 0;
+	}
+
+	for (offset = 0; offset < len; offset += xfer_count << src_width) {
+		xfer_count = min_t(size_t, (len - offset) >> src_width,
+				ATC_BTSIZE_MAX);
+
+		desc = atc_desc_get(atchan);
+		if (!desc)
+			goto err_desc_get;
+
+		desc->lli.saddr = src + offset;
+		desc->lli.daddr = dest + offset;
+		desc->lli.ctrla = ctrla | xfer_count;
+		desc->lli.ctrlb = ctrlb;
+
+		desc->txd.cookie = 0;
+		async_tx_ack(&desc->txd);
+
+		if (!first) {
+			first = desc;
+		} else {
+			/* inform the HW lli about chaining */
+			prev->lli.dscr = desc->txd.phys;
+			/* insert the link descriptor to the LD ring */
+			list_add_tail(&desc->desc_node,
+					&first->txd.tx_list);
+		}
+		prev = desc;
+	}
+
+	/* First descriptor of the chain embedds additional information */
+	first->txd.cookie = -EBUSY;
+	first->len = len;
+
+	/* set end-of-link to the last link descriptor of list*/
+	set_desc_eol(desc);
+
+	desc->txd.flags = flags; /* client is in control of this ack */
+
+	return &first->txd;
+
+err_desc_get:
+	atc_desc_put(atchan, first);
+	return NULL;
+}
+
+
+/**
+ * atc_prep_slave_sg - prepare descriptors for a DMA_SLAVE transaction
+ * @chan: DMA channel
+ * @sgl: scatterlist to transfer to/from
+ * @sg_len: number of entries in @scatterlist
+ * @direction: DMA direction
+ * @flags: tx descriptor status flags
+ */
+static struct dma_async_tx_descriptor *
+atc_prep_slave_sg(struct dma_chan *chan, struct scatterlist *sgl,
+		unsigned int sg_len, enum dma_data_direction direction,
+		unsigned long flags)
+{
+	struct at_dma_chan	*atchan = to_at_dma_chan(chan);
+	struct at_dma_slave	*atslave = chan->private;
+	struct at_desc		*first = NULL;
+	struct at_desc		*prev = NULL;
+	u32			ctrla;
+	u32			ctrlb;
+	dma_addr_t		reg;
+	unsigned int		reg_width;
+	unsigned int		mem_width;
+	unsigned int		i;
+	struct scatterlist	*sg;
+	size_t			total_len = 0;
+
+	dev_vdbg(chan2dev(chan), "prep_slave_sg: %s f0x%lx\n",
+			direction == DMA_TO_DEVICE ? "TO DEVICE" : "FROM DEVICE",
+			flags);
+
+	if (unlikely(!atslave || !sg_len)) {
+		dev_dbg(chan2dev(chan), "prep_dma_memcpy: length is zero!\n");
+		return NULL;
+	}
+
+	reg_width = atslave->reg_width;
+
+	sg_len = dma_map_sg(chan2parent(chan), sgl, sg_len, direction);
+
+	ctrla = ATC_DEFAULT_CTRLA | atslave->ctrla;
+	ctrlb = ATC_DEFAULT_CTRLB | ATC_IEN;
+
+	switch (direction) {
+	case DMA_TO_DEVICE:
+		ctrla |=  ATC_DST_WIDTH(reg_width);
+		ctrlb |=  ATC_DST_ADDR_MODE_FIXED
+			| ATC_SRC_ADDR_MODE_INCR
+			| ATC_FC_MEM2PER;
+		reg = atslave->tx_reg;
+		for_each_sg(sgl, sg, sg_len, i) {
+			struct at_desc	*desc;
+			u32		len;
+			u32		mem;
+
+			desc = atc_desc_get(atchan);
+			if (!desc)
+				goto err_desc_get;
+
+			mem = sg_phys(sg);
+			len = sg_dma_len(sg);
+			mem_width = 2;
+			if (unlikely(mem & 3 || len & 3))
+				mem_width = 0;
+
+			desc->lli.saddr = mem;
+			desc->lli.daddr = reg;
+			desc->lli.ctrla = ctrla
+					| ATC_SRC_WIDTH(mem_width)
+					| len >> mem_width;
+			desc->lli.ctrlb = ctrlb;
+
+			if (!first) {
+				first = desc;
+			} else {
+				/* inform the HW lli about chaining */
+				prev->lli.dscr = desc->txd.phys;
+				/* insert the link descriptor to the LD ring */
+				list_add_tail(&desc->desc_node,
+						&first->txd.tx_list);
+			}
+			prev = desc;
+			total_len += len;
+		}
+		break;
+	case DMA_FROM_DEVICE:
+		ctrla |=  ATC_SRC_WIDTH(reg_width);
+		ctrlb |=  ATC_DST_ADDR_MODE_INCR
+			| ATC_SRC_ADDR_MODE_FIXED
+			| ATC_FC_PER2MEM;
+
+		reg = atslave->rx_reg;
+		for_each_sg(sgl, sg, sg_len, i) {
+			struct at_desc	*desc;
+			u32		len;
+			u32		mem;
+
+			desc = atc_desc_get(atchan);
+			if (!desc)
+				goto err_desc_get;
+
+			mem = sg_phys(sg);
+			len = sg_dma_len(sg);
+			mem_width = 2;
+			if (unlikely(mem & 3 || len & 3))
+				mem_width = 0;
+
+			desc->lli.saddr = reg;
+			desc->lli.daddr = mem;
+			desc->lli.ctrla = ctrla
+					| ATC_DST_WIDTH(mem_width)
+					| len >> mem_width;
+			desc->lli.ctrlb = ctrlb;
+
+			if (!first) {
+				first = desc;
+			} else {
+				/* inform the HW lli about chaining */
+				prev->lli.dscr = desc->txd.phys;
+				/* insert the link descriptor to the LD ring */
+				list_add_tail(&desc->desc_node,
+						&first->txd.tx_list);
+			}
+			prev = desc;
+			total_len += len;
+		}
+		break;
+	default:
+		return NULL;
+	}
+
+	/* set end-of-link to the last link descriptor of list*/
+	set_desc_eol(prev);
+
+	/* First descriptor of the chain embedds additional information */
+	first->txd.cookie = -EBUSY;
+	first->len = total_len;
+
+	/* last link descriptor of list is responsible of flags */
+	prev->txd.flags = flags; /* client is in control of this ack */
+
+	return &first->txd;
+
+err_desc_get:
+	dev_err(chan2dev(chan), "not enough descriptors available\n");
+	atc_desc_put(atchan, first);
+	return NULL;
+}
+
+static void atc_terminate_all(struct dma_chan *chan)
+{
+	struct at_dma_chan	*atchan = to_at_dma_chan(chan);
+	struct at_dma		*atdma = to_at_dma(chan->device);
+	struct at_desc		*desc, *_desc;
+	LIST_HEAD(list);
+
+	/*
+	 * This is only called when something went wrong elsewhere, so
+	 * we don't really care about the data. Just disable the
+	 * channel. We still have to poll the channel enable bit due
+	 * to AHB/HSB limitations.
+	 */
+	spin_lock_bh(&atchan->lock);
+
+	dma_writel(atdma, CHDR, atchan->mask);
+
+	/* confirm that this channel is disabled */
+	while (dma_readl(atdma, CHSR) & atchan->mask)
+		cpu_relax();
+
+	/* active_list entries will end up before queued entries */
+	list_splice_init(&atchan->queue, &list);
+	list_splice_init(&atchan->active_list, &list);
+
+	spin_unlock_bh(&atchan->lock);
+
+	/* Flush all pending and queued descriptors */
+	list_for_each_entry_safe(desc, _desc, &list, desc_node)
+		atc_chain_complete(atchan, desc);
+}
+
+/**
+ * atc_is_tx_complete - poll for transaction completion
+ * @chan: DMA channel
+ * @cookie: transaction identifier to check status of
+ * @done: if not %NULL, updated with last completed transaction
+ * @used: if not %NULL, updated with last used transaction
+ *
+ * If @done and @used are passed in, upon return they reflect the driver
+ * internal state and can be used with dma_async_is_complete() to check
+ * the status of multiple cookies without re-checking hardware state.
+ */
+static enum dma_status
+atc_is_tx_complete(struct dma_chan *chan,
+		dma_cookie_t cookie,
+		dma_cookie_t *done, dma_cookie_t *used)
+{
+	struct at_dma_chan	*atchan = to_at_dma_chan(chan);
+	dma_cookie_t		last_used;
+	dma_cookie_t		last_complete;
+	enum dma_status		ret;
+
+	dev_vdbg(chan2dev(chan), "is_tx_complete: %d (d%d, u%d)\n",
+			cookie, done ? *done : 0, used ? *used : 0);
+
+	spin_lock_bh(atchan->lock);
+
+	last_complete = atchan->completed_cookie;
+	last_used = chan->cookie;
+
+	ret = dma_async_is_complete(cookie, last_complete, last_used);
+	if (ret != DMA_SUCCESS) {
+		atc_cleanup_descriptors(atchan);
+
+		last_complete = atchan->completed_cookie;
+		last_used = chan->cookie;
+
+		ret = dma_async_is_complete(cookie, last_complete, last_used);
+	}
+
+	spin_unlock_bh(atchan->lock);
+
+	if (done)
+		*done = last_complete;
+	if (used)
+		*used = last_used;
+
+	return ret;
+}
+
+/**
+ * atc_issue_pending - try to finish work
+ * @chan: target DMA channel
+ */
+static void atc_issue_pending(struct dma_chan *chan)
+{
+	struct at_dma_chan	*atchan = to_at_dma_chan(chan);
+
+	dev_vdbg(chan2dev(chan), "issue_pending\n");
+
+	if (!atc_chan_is_enabled(atchan)) {
+		spin_lock_bh(&atchan->lock);
+		atc_advance_work(atchan);
+		spin_unlock_bh(&atchan->lock);
+	}
+}
+
+/**
+ * atc_alloc_chan_resources - allocate resources for DMA channel
+ * @chan: allocate descriptor resources for this channel
+ * @client: current client requesting the channel be ready for requests
+ *
+ * return - the number of allocated descriptors
+ */
+static int atc_alloc_chan_resources(struct dma_chan *chan)
+{
+	struct at_dma_chan	*atchan = to_at_dma_chan(chan);
+	struct at_dma		*atdma = to_at_dma(chan->device);
+	struct at_desc		*desc;
+	struct at_dma_slave	*atslave;
+	int			i;
+	u32			cfg;
+	LIST_HEAD(tmp_list);
+
+	dev_vdbg(chan2dev(chan), "alloc_chan_resources\n");
+
+	/* ASSERT:  channel is idle */
+	if (atc_chan_is_enabled(atchan)) {
+		dev_dbg(chan2dev(chan), "DMA channel not idle ?\n");
+		return -EIO;
+	}
+
+	cfg = ATC_DEFAULT_CFG;
+
+	atslave = chan->private;
+	if (atslave) {
+		/*
+		 * We need controller-specific data to set up slave
+		 * transfers.
+		 */
+		BUG_ON(!atslave->dma_dev || atslave->dma_dev != atdma->dma_common.dev);
+
+		/* if cfg configuration specified take it instad of default */
+		if (atslave->cfg)
+			cfg = atslave->cfg;
+	}
+
+	/* have we already been set up?
+	 * reconfigure channel but no need to reallocate descriptors */
+	if (!list_empty(&atchan->free_list))
+		return atchan->descs_allocated;
+
+	/* Allocate initial pool of descriptors */
+	for (i = 0; i < init_nr_desc_per_channel; i++) {
+		desc = atc_alloc_descriptor(chan, GFP_KERNEL);
+		if (!desc) {
+			dev_err(atdma->dma_common.dev,
+				"Only %d initial descriptors\n", i);
+			break;
+		}
+		list_add_tail(&desc->desc_node, &tmp_list);
+	}
+
+	spin_lock_bh(&atchan->lock);
+	atchan->descs_allocated = i;
+	list_splice(&tmp_list, &atchan->free_list);
+	atchan->completed_cookie = chan->cookie = 1;
+	spin_unlock_bh(&atchan->lock);
+
+	/* channel parameters */
+	channel_writel(atchan, CFG, cfg);
+
+	dev_dbg(chan2dev(chan),
+		"alloc_chan_resources: allocated %d descriptors\n",
+		atchan->descs_allocated);
+
+	return atchan->descs_allocated;
+}
+
+/**
+ * atc_free_chan_resources - free all channel resources
+ * @chan: DMA channel
+ */
+static void atc_free_chan_resources(struct dma_chan *chan)
+{
+	struct at_dma_chan	*atchan = to_at_dma_chan(chan);
+	struct at_dma		*atdma = to_at_dma(chan->device);
+	struct at_desc		*desc, *_desc;
+	LIST_HEAD(list);
+
+	dev_dbg(chan2dev(chan), "free_chan_resources: (descs allocated=%u)\n",
+		atchan->descs_allocated);
+
+	/* ASSERT:  channel is idle */
+	BUG_ON(!list_empty(&atchan->active_list));
+	BUG_ON(!list_empty(&atchan->queue));
+	BUG_ON(atc_chan_is_enabled(atchan));
+
+	list_for_each_entry_safe(desc, _desc, &atchan->free_list, desc_node) {
+		dev_vdbg(chan2dev(chan), "  freeing descriptor %p\n", desc);
+		list_del(&desc->desc_node);
+		/* free link descriptor */
+		dma_pool_free(atdma->dma_desc_pool, desc, desc->txd.phys);
+	}
+	list_splice_init(&atchan->free_list, &list);
+	atchan->descs_allocated = 0;
+
+	dev_vdbg(chan2dev(chan), "free_chan_resources: done\n");
+}
+
+
+/*--  Module Management  -----------------------------------------------*/
+
+/**
+ * at_dma_off - disable DMA controller
+ * @atdma: the Atmel HDAMC device
+ */
+static void at_dma_off(struct at_dma *atdma)
+{
+	dma_writel(atdma, EN, 0);
+
+	/* disable all interrupts */
+	dma_writel(atdma, EBCIDR, -1L);
+
+	/* confirm that all channels are disabled */
+	while (dma_readl(atdma, CHSR) & atdma->all_chan_mask)
+		cpu_relax();
+}
+
+static int __init at_dma_probe(struct platform_device *pdev)
+{
+	struct at_dma_platform_data *pdata;
+	struct resource		*io;
+	struct at_dma		*atdma;
+	size_t			size;
+	int			irq;
+	int			err;
+	int			i;
+
+	/* get DMA Controller parameters from platform */
+	pdata = pdev->dev.platform_data;
+	if (!pdata || pdata->nr_channels > AT_DMA_MAX_NR_CHANNELS)
+		return -EINVAL;
+
+	io = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	if (!io)
+		return -EINVAL;
+
+	irq = platform_get_irq(pdev, 0);
+	if (irq < 0)
+		return irq;
+
+	size = sizeof(struct at_dma);
+	size += pdata->nr_channels * sizeof(struct at_dma_chan);
+	atdma = kzalloc(size, GFP_KERNEL);
+	if (!atdma)
+		return -ENOMEM;
+
+	/* discover transaction capabilites from the platform data */
+	atdma->dma_common.cap_mask = pdata->cap_mask;
+	atdma->all_chan_mask = (1 << pdata->nr_channels) - 1;
+
+	size = io->end - io->start + 1;
+	if (!request_mem_region(io->start, size, pdev->dev.driver->name)) {
+		err = -EBUSY;
+		goto err_kfree;
+	}
+
+	atdma->regs = ioremap(io->start, size);
+	if (!atdma->regs) {
+		err = -ENOMEM;
+		goto err_release_r;
+	}
+
+	atdma->clk = clk_get(&pdev->dev, "dma_clk");
+	if (IS_ERR(atdma->clk)) {
+		err = PTR_ERR(atdma->clk);
+		goto err_clk;
+	}
+	clk_enable(atdma->clk);
+
+	/* force dma off, just in case */
+	at_dma_off(atdma);
+
+	err = request_irq(irq, at_dma_interrupt, 0, "at_hdmac", atdma);
+	if (err)
+		goto err_irq;
+
+	platform_set_drvdata(pdev, atdma);
+
+	/* create a pool of consistent memory blocks for hardware descriptors */
+	atdma->dma_desc_pool = dma_pool_create("at_hdmac_desc_pool",
+			&pdev->dev, sizeof(struct at_desc),
+			4 /* word alignment */, 0);
+	if (!atdma->dma_desc_pool) {
+		dev_err(&pdev->dev, "No memory for descriptors dma pool\n");
+		err = -ENOMEM;
+		goto err_pool_create;
+	}
+
+	/* clear any pending interrupt */
+	while (dma_readl(atdma, EBCISR))
+		cpu_relax();
+
+	/* initialize channels related values */
+	INIT_LIST_HEAD(&atdma->dma_common.channels);
+	for (i = 0; i < pdata->nr_channels; i++, atdma->dma_common.chancnt++) {
+		struct at_dma_chan	*atchan = &atdma->chan[i];
+
+		atchan->chan_common.device = &atdma->dma_common;
+		atchan->chan_common.cookie = atchan->completed_cookie = 1;
+		atchan->chan_common.chan_id = i;
+		list_add_tail(&atchan->chan_common.device_node,
+				&atdma->dma_common.channels);
+
+		atchan->ch_regs = atdma->regs + ch_regs(i);
+		spin_lock_init(&atchan->lock);
+		atchan->mask = 1 << i;
+
+		INIT_LIST_HEAD(&atchan->active_list);
+		INIT_LIST_HEAD(&atchan->queue);
+		INIT_LIST_HEAD(&atchan->free_list);
+
+		tasklet_init(&atchan->tasklet, atc_tasklet,
+				(unsigned long)atchan);
+		atc_enable_irq(atchan);
+	}
+
+	/* set base routines */
+	atdma->dma_common.device_alloc_chan_resources = atc_alloc_chan_resources;
+	atdma->dma_common.device_free_chan_resources = atc_free_chan_resources;
+	atdma->dma_common.device_is_tx_complete = atc_is_tx_complete;
+	atdma->dma_common.device_issue_pending = atc_issue_pending;
+	atdma->dma_common.dev = &pdev->dev;
+
+	/* set prep routines based on capability */
+	if (dma_has_cap(DMA_MEMCPY, atdma->dma_common.cap_mask))
+		atdma->dma_common.device_prep_dma_memcpy = atc_prep_dma_memcpy;
+
+	if (dma_has_cap(DMA_SLAVE, atdma->dma_common.cap_mask)) {
+		atdma->dma_common.device_prep_slave_sg = atc_prep_slave_sg;
+		atdma->dma_common.device_terminate_all = atc_terminate_all;
+	}
+
+	dma_writel(atdma, EN, AT_DMA_ENABLE);
+
+	dev_info(&pdev->dev, "Atmel AHB DMA Controller ( %s%s), %d channels\n",
+	  dma_has_cap(DMA_MEMCPY, atdma->dma_common.cap_mask) ? "cpy " : "",
+	  dma_has_cap(DMA_SLAVE, atdma->dma_common.cap_mask)  ? "slave " : "",
+	  atdma->dma_common.chancnt);
+
+	dma_async_device_register(&atdma->dma_common);
+
+	return 0;
+
+err_pool_create:
+	platform_set_drvdata(pdev, NULL);
+	free_irq(platform_get_irq(pdev, 0), atdma);
+err_irq:
+	clk_disable(atdma->clk);
+	clk_put(atdma->clk);
+err_clk:
+	iounmap(atdma->regs);
+	atdma->regs = NULL;
+err_release_r:
+	release_mem_region(io->start, size);
+err_kfree:
+	kfree(atdma);
+	return err;
+}
+
+static int __exit at_dma_remove(struct platform_device *pdev)
+{
+	struct at_dma		*atdma = platform_get_drvdata(pdev);
+	struct dma_chan		*chan, *_chan;
+	struct resource		*io;
+
+	at_dma_off(atdma);
+	dma_async_device_unregister(&atdma->dma_common);
+
+	dma_pool_destroy(atdma->dma_desc_pool);
+	platform_set_drvdata(pdev, NULL);
+	free_irq(platform_get_irq(pdev, 0), atdma);
+
+	list_for_each_entry_safe(chan, _chan, &atdma->dma_common.channels,
+			device_node) {
+		struct at_dma_chan	*atchan = to_at_dma_chan(chan);
+
+		/* Disable interrupts */
+		atc_disable_irq(atchan);
+		tasklet_disable(&atchan->tasklet);
+
+		tasklet_kill(&atchan->tasklet);
+		list_del(&chan->device_node);
+	}
+
+	clk_disable(atdma->clk);
+	clk_put(atdma->clk);
+
+	iounmap(atdma->regs);
+	atdma->regs = NULL;
+
+	io = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	release_mem_region(io->start, io->end - io->start + 1);
+
+	kfree(atdma);
+
+	return 0;
+}
+
+static void at_dma_shutdown(struct platform_device *pdev)
+{
+	struct at_dma	*atdma = platform_get_drvdata(pdev);
+
+	at_dma_off(platform_get_drvdata(pdev));
+	clk_disable(atdma->clk);
+}
+
+static int at_dma_suspend_late(struct platform_device *pdev, pm_message_t mesg)
+{
+	struct at_dma	*atdma = platform_get_drvdata(pdev);
+
+	at_dma_off(platform_get_drvdata(pdev));
+	clk_disable(atdma->clk);
+	return 0;
+}
+
+static int at_dma_resume_early(struct platform_device *pdev)
+{
+	struct at_dma	*atdma = platform_get_drvdata(pdev);
+
+	clk_enable(atdma->clk);
+	dma_writel(atdma, EN, AT_DMA_ENABLE);
+	return 0;
+
+}
+
+static struct platform_driver at_dma_driver = {
+	.remove		= __exit_p(at_dma_remove),
+	.shutdown	= at_dma_shutdown,
+	.suspend_late	= at_dma_suspend_late,
+	.resume_early	= at_dma_resume_early,
+	.driver = {
+		.name	= "at_hdmac",
+	},
+};
+
+static int __init at_dma_init(void)
+{
+	return platform_driver_probe(&at_dma_driver, at_dma_probe);
+}
+module_init(at_dma_init);
+
+static void __exit at_dma_exit(void)
+{
+	platform_driver_unregister(&at_dma_driver);
+}
+module_exit(at_dma_exit);
+
+MODULE_DESCRIPTION("Atmel AHB DMA Controller driver");
+MODULE_AUTHOR("Nicolas Ferre <nicolas.ferre@atmel.com>");
+MODULE_LICENSE("GPL");
+MODULE_ALIAS("platform:at_hdmac");
diff --git a/drivers/dma/at_hdmac_regs.h b/drivers/dma/at_hdmac_regs.h
new file mode 100644
index 00000000000..4c972afc49e
--- /dev/null
+++ b/drivers/dma/at_hdmac_regs.h
@@ -0,0 +1,353 @@
+/*
+ * Header file for the Atmel AHB DMA Controller driver
+ *
+ * Copyright (C) 2008 Atmel Corporation
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ */
+#ifndef AT_HDMAC_REGS_H
+#define	AT_HDMAC_REGS_H
+
+#include <mach/at_hdmac.h>
+
+#define	AT_DMA_MAX_NR_CHANNELS	8
+
+
+#define	AT_DMA_GCFG	0x00	/* Global Configuration Register */
+#define		AT_DMA_IF_BIGEND(i)	(0x1 << (i))	/* AHB-Lite Interface i in Big-endian mode */
+#define		AT_DMA_ARB_CFG	(0x1 << 4)	/* Arbiter mode. */
+#define			AT_DMA_ARB_CFG_FIXED		(0x0 << 4)
+#define			AT_DMA_ARB_CFG_ROUND_ROBIN	(0x1 << 4)
+
+#define	AT_DMA_EN	0x04	/* Controller Enable Register */
+#define		AT_DMA_ENABLE	(0x1 << 0)
+
+#define	AT_DMA_SREQ	0x08	/* Software Single Request Register */
+#define		AT_DMA_SSREQ(x)	(0x1 << ((x) << 1))		/* Request a source single transfer on channel x */
+#define		AT_DMA_DSREQ(x)	(0x1 << (1 + ((x) << 1)))	/* Request a destination single transfer on channel x */
+
+#define	AT_DMA_CREQ	0x0C	/* Software Chunk Transfer Request Register */
+#define		AT_DMA_SCREQ(x)	(0x1 << ((x) << 1))		/* Request a source chunk transfer on channel x */
+#define		AT_DMA_DCREQ(x)	(0x1 << (1 + ((x) << 1)))	/* Request a destination chunk transfer on channel x */
+
+#define	AT_DMA_LAST	0x10	/* Software Last Transfer Flag Register */
+#define		AT_DMA_SLAST(x)	(0x1 << ((x) << 1))		/* This src rq is last tx of buffer on channel x */
+#define		AT_DMA_DLAST(x)	(0x1 << (1 + ((x) << 1)))	/* This dst rq is last tx of buffer on channel x */
+
+#define	AT_DMA_SYNC	0x14	/* Request Synchronization Register */
+#define		AT_DMA_SYR(h)	(0x1 << (h))			/* Synchronize handshake line h */
+
+/* Error, Chained Buffer transfer completed and Buffer transfer completed Interrupt registers */
+#define	AT_DMA_EBCIER	0x18	/* Enable register */
+#define	AT_DMA_EBCIDR	0x1C	/* Disable register */
+#define	AT_DMA_EBCIMR	0x20	/* Mask Register */
+#define	AT_DMA_EBCISR	0x24	/* Status Register */
+#define		AT_DMA_CBTC_OFFSET	8
+#define		AT_DMA_ERR_OFFSET	16
+#define		AT_DMA_BTC(x)	(0x1 << (x))
+#define		AT_DMA_CBTC(x)	(0x1 << (AT_DMA_CBTC_OFFSET + (x)))
+#define		AT_DMA_ERR(x)	(0x1 << (AT_DMA_ERR_OFFSET + (x)))
+
+#define	AT_DMA_CHER	0x28	/* Channel Handler Enable Register */
+#define		AT_DMA_ENA(x)	(0x1 << (x))
+#define		AT_DMA_SUSP(x)	(0x1 << ( 8 + (x)))
+#define		AT_DMA_KEEP(x)	(0x1 << (24 + (x)))
+
+#define	AT_DMA_CHDR	0x2C	/* Channel Handler Disable Register */
+#define		AT_DMA_DIS(x)	(0x1 << (x))
+#define		AT_DMA_RES(x)	(0x1 << ( 8 + (x)))
+
+#define	AT_DMA_CHSR	0x30	/* Channel Handler Status Register */
+#define		AT_DMA_EMPT(x)	(0x1 << (16 + (x)))
+#define		AT_DMA_STAL(x)	(0x1 << (24 + (x)))
+
+
+#define	AT_DMA_CH_REGS_BASE	0x3C	/* Channel registers base address */
+#define	ch_regs(x)	(AT_DMA_CH_REGS_BASE + (x) * 0x28) /* Channel x base addr */
+
+/* Hardware register offset for each channel */
+#define	ATC_SADDR_OFFSET	0x00	/* Source Address Register */
+#define	ATC_DADDR_OFFSET	0x04	/* Destination Address Register */
+#define	ATC_DSCR_OFFSET		0x08	/* Descriptor Address Register */
+#define	ATC_CTRLA_OFFSET	0x0C	/* Control A Register */
+#define	ATC_CTRLB_OFFSET	0x10	/* Control B Register */
+#define	ATC_CFG_OFFSET		0x14	/* Configuration Register */
+#define	ATC_SPIP_OFFSET		0x18	/* Src PIP Configuration Register */
+#define	ATC_DPIP_OFFSET		0x1C	/* Dst PIP Configuration Register */
+
+
+/* Bitfield definitions */
+
+/* Bitfields in DSCR */
+#define	ATC_DSCR_IF(i)		(0x3 & (i))	/* Dsc feched via AHB-Lite Interface i */
+
+/* Bitfields in CTRLA */
+#define	ATC_BTSIZE_MAX		0xFFFFUL	/* Maximum Buffer Transfer Size */
+#define	ATC_BTSIZE(x)		(ATC_BTSIZE_MAX & (x)) /* Buffer Transfer Size */
+/* Chunck Tranfer size definitions are in at_hdmac.h */
+#define	ATC_SRC_WIDTH_MASK	(0x3 << 24)	/* Source Single Transfer Size */
+#define		ATC_SRC_WIDTH(x)	((x) << 24)
+#define		ATC_SRC_WIDTH_BYTE	(0x0 << 24)
+#define		ATC_SRC_WIDTH_HALFWORD	(0x1 << 24)
+#define		ATC_SRC_WIDTH_WORD	(0x2 << 24)
+#define	ATC_DST_WIDTH_MASK	(0x3 << 28)	/* Destination Single Transfer Size */
+#define		ATC_DST_WIDTH(x)	((x) << 28)
+#define		ATC_DST_WIDTH_BYTE	(0x0 << 28)
+#define		ATC_DST_WIDTH_HALFWORD	(0x1 << 28)
+#define		ATC_DST_WIDTH_WORD	(0x2 << 28)
+#define	ATC_DONE		(0x1 << 31)	/* Tx Done (only written back in descriptor) */
+
+/* Bitfields in CTRLB */
+#define	ATC_SIF(i)		(0x3 & (i))	/* Src tx done via AHB-Lite Interface i */
+#define	ATC_DIF(i)		((0x3 & (i)) <<  4)	/* Dst tx done via AHB-Lite Interface i */
+#define	ATC_SRC_PIP		(0x1 <<  8)	/* Source Picture-in-Picture enabled */
+#define	ATC_DST_PIP		(0x1 << 12)	/* Destination Picture-in-Picture enabled */
+#define	ATC_SRC_DSCR_DIS	(0x1 << 16)	/* Src Descriptor fetch disable */
+#define	ATC_DST_DSCR_DIS	(0x1 << 20)	/* Dst Descriptor fetch disable */
+#define	ATC_FC_MASK		(0x7 << 21)	/* Choose Flow Controller */
+#define		ATC_FC_MEM2MEM		(0x0 << 21)	/* Mem-to-Mem (DMA) */
+#define		ATC_FC_MEM2PER		(0x1 << 21)	/* Mem-to-Periph (DMA) */
+#define		ATC_FC_PER2MEM		(0x2 << 21)	/* Periph-to-Mem (DMA) */
+#define		ATC_FC_PER2PER		(0x3 << 21)	/* Periph-to-Periph (DMA) */
+#define		ATC_FC_PER2MEM_PER	(0x4 << 21)	/* Periph-to-Mem (Peripheral) */
+#define		ATC_FC_MEM2PER_PER	(0x5 << 21)	/* Mem-to-Periph (Peripheral) */
+#define		ATC_FC_PER2PER_SRCPER	(0x6 << 21)	/* Periph-to-Periph (Src Peripheral) */
+#define		ATC_FC_PER2PER_DSTPER	(0x7 << 21)	/* Periph-to-Periph (Dst Peripheral) */
+#define	ATC_SRC_ADDR_MODE_MASK	(0x3 << 24)
+#define		ATC_SRC_ADDR_MODE_INCR	(0x0 << 24)	/* Incrementing Mode */
+#define		ATC_SRC_ADDR_MODE_DECR	(0x1 << 24)	/* Decrementing Mode */
+#define		ATC_SRC_ADDR_MODE_FIXED	(0x2 << 24)	/* Fixed Mode */
+#define	ATC_DST_ADDR_MODE_MASK	(0x3 << 28)
+#define		ATC_DST_ADDR_MODE_INCR	(0x0 << 28)	/* Incrementing Mode */
+#define		ATC_DST_ADDR_MODE_DECR	(0x1 << 28)	/* Decrementing Mode */
+#define		ATC_DST_ADDR_MODE_FIXED	(0x2 << 28)	/* Fixed Mode */
+#define	ATC_IEN			(0x1 << 30)	/* BTC interrupt enable (active low) */
+#define	ATC_AUTO		(0x1 << 31)	/* Auto multiple buffer tx enable */
+
+/* Bitfields in CFG */
+/* are in at_hdmac.h */
+
+/* Bitfields in SPIP */
+#define	ATC_SPIP_HOLE(x)	(0xFFFFU & (x))
+#define	ATC_SPIP_BOUNDARY(x)	((0x3FF & (x)) << 16)
+
+/* Bitfields in DPIP */
+#define	ATC_DPIP_HOLE(x)	(0xFFFFU & (x))
+#define	ATC_DPIP_BOUNDARY(x)	((0x3FF & (x)) << 16)
+
+
+/*--  descriptors  -----------------------------------------------------*/
+
+/* LLI == Linked List Item; aka DMA buffer descriptor */
+struct at_lli {
+	/* values that are not changed by hardware */
+	dma_addr_t	saddr;
+	dma_addr_t	daddr;
+	/* value that may get written back: */
+	u32		ctrla;
+	/* more values that are not changed by hardware */
+	u32		ctrlb;
+	dma_addr_t	dscr;	/* chain to next lli */
+};
+
+/**
+ * struct at_desc - software descriptor
+ * @at_lli: hardware lli structure
+ * @txd: support for the async_tx api
+ * @desc_node: node on the channed descriptors list
+ * @len: total transaction bytecount
+ */
+struct at_desc {
+	/* FIRST values the hardware uses */
+	struct at_lli			lli;
+
+	/* THEN values for driver housekeeping */
+	struct dma_async_tx_descriptor	txd;
+	struct list_head		desc_node;
+	size_t				len;
+};
+
+static inline struct at_desc *
+txd_to_at_desc(struct dma_async_tx_descriptor *txd)
+{
+	return container_of(txd, struct at_desc, txd);
+}
+
+
+/*--  Channels  --------------------------------------------------------*/
+
+/**
+ * struct at_dma_chan - internal representation of an Atmel HDMAC channel
+ * @chan_common: common dmaengine channel object members
+ * @device: parent device
+ * @ch_regs: memory mapped register base
+ * @mask: channel index in a mask
+ * @error_status: transmit error status information from irq handler
+ *                to tasklet (use atomic operations)
+ * @tasklet: bottom half to finish transaction work
+ * @lock: serializes enqueue/dequeue operations to descriptors lists
+ * @completed_cookie: identifier for the most recently completed operation
+ * @active_list: list of descriptors dmaengine is being running on
+ * @queue: list of descriptors ready to be submitted to engine
+ * @free_list: list of descriptors usable by the channel
+ * @descs_allocated: records the actual size of the descriptor pool
+ */
+struct at_dma_chan {
+	struct dma_chan		chan_common;
+	struct at_dma		*device;
+	void __iomem		*ch_regs;
+	u8			mask;
+	unsigned long		error_status;
+	struct tasklet_struct	tasklet;
+
+	spinlock_t		lock;
+
+	/* these other elements are all protected by lock */
+	dma_cookie_t		completed_cookie;
+	struct list_head	active_list;
+	struct list_head	queue;
+	struct list_head	free_list;
+	unsigned int		descs_allocated;
+};
+
+#define	channel_readl(atchan, name) \
+	__raw_readl((atchan)->ch_regs + ATC_##name##_OFFSET)
+
+#define	channel_writel(atchan, name, val) \
+	__raw_writel((val), (atchan)->ch_regs + ATC_##name##_OFFSET)
+
+static inline struct at_dma_chan *to_at_dma_chan(struct dma_chan *dchan)
+{
+	return container_of(dchan, struct at_dma_chan, chan_common);
+}
+
+
+/*--  Controller  ------------------------------------------------------*/
+
+/**
+ * struct at_dma - internal representation of an Atmel HDMA Controller
+ * @chan_common: common dmaengine dma_device object members
+ * @ch_regs: memory mapped register base
+ * @clk: dma controller clock
+ * @all_chan_mask: all channels availlable in a mask
+ * @dma_desc_pool: base of DMA descriptor region (DMA address)
+ * @chan: channels table to store at_dma_chan structures
+ */
+struct at_dma {
+	struct dma_device	dma_common;
+	void __iomem		*regs;
+	struct clk		*clk;
+
+	u8			all_chan_mask;
+
+	struct dma_pool		*dma_desc_pool;
+	/* AT THE END channels table */
+	struct at_dma_chan	chan[0];
+};
+
+#define	dma_readl(atdma, name) \
+	__raw_readl((atdma)->regs + AT_DMA_##name)
+#define	dma_writel(atdma, name, val) \
+	__raw_writel((val), (atdma)->regs + AT_DMA_##name)
+
+static inline struct at_dma *to_at_dma(struct dma_device *ddev)
+{
+	return container_of(ddev, struct at_dma, dma_common);
+}
+
+
+/*--  Helper functions  ------------------------------------------------*/
+
+static struct device *chan2dev(struct dma_chan *chan)
+{
+	return &chan->dev->device;
+}
+static struct device *chan2parent(struct dma_chan *chan)
+{
+	return chan->dev->device.parent;
+}
+
+#if defined(VERBOSE_DEBUG)
+static void vdbg_dump_regs(struct at_dma_chan *atchan)
+{
+	struct at_dma	*atdma = to_at_dma(atchan->chan_common.device);
+
+	dev_err(chan2dev(&atchan->chan_common),
+		"  channel %d : imr = 0x%x, chsr = 0x%x\n",
+		atchan->chan_common.chan_id,
+		dma_readl(atdma, EBCIMR),
+		dma_readl(atdma, CHSR));
+
+	dev_err(chan2dev(&atchan->chan_common),
+		"  channel: s0x%x d0x%x ctrl0x%x:0x%x cfg0x%x l0x%x\n",
+		channel_readl(atchan, SADDR),
+		channel_readl(atchan, DADDR),
+		channel_readl(atchan, CTRLA),
+		channel_readl(atchan, CTRLB),
+		channel_readl(atchan, CFG),
+		channel_readl(atchan, DSCR));
+}
+#else
+static void vdbg_dump_regs(struct at_dma_chan *atchan) {}
+#endif
+
+static void atc_dump_lli(struct at_dma_chan *atchan, struct at_lli *lli)
+{
+	dev_printk(KERN_CRIT, chan2dev(&atchan->chan_common),
+			"  desc: s0x%x d0x%x ctrl0x%x:0x%x l0x%x\n",
+			lli->saddr, lli->daddr,
+			lli->ctrla, lli->ctrlb, lli->dscr);
+}
+
+
+static void atc_setup_irq(struct at_dma_chan *atchan, int on)
+{
+	struct at_dma	*atdma = to_at_dma(atchan->chan_common.device);
+	u32		ebci;
+
+	/* enable interrupts on buffer chain completion & error */
+	ebci =    AT_DMA_CBTC(atchan->chan_common.chan_id)
+		| AT_DMA_ERR(atchan->chan_common.chan_id);
+	if (on)
+		dma_writel(atdma, EBCIER, ebci);
+	else
+		dma_writel(atdma, EBCIDR, ebci);
+}
+
+static inline void atc_enable_irq(struct at_dma_chan *atchan)
+{
+	atc_setup_irq(atchan, 1);
+}
+
+static inline void atc_disable_irq(struct at_dma_chan *atchan)
+{
+	atc_setup_irq(atchan, 0);
+}
+
+
+/**
+ * atc_chan_is_enabled - test if given channel is enabled
+ * @atchan: channel we want to test status
+ */
+static inline int atc_chan_is_enabled(struct at_dma_chan *atchan)
+{
+	struct at_dma	*atdma = to_at_dma(atchan->chan_common.device);
+
+	return !!(dma_readl(atdma, CHSR) & atchan->mask);
+}
+
+
+/**
+ * set_desc_eol - set end-of-link to descriptor so it will end transfer
+ * @desc: descriptor, signle or at the end of a chain, to end chain on
+ */
+static void set_desc_eol(struct at_desc *desc)
+{
+	desc->lli.ctrlb |= ATC_SRC_DSCR_DIS | ATC_DST_DSCR_DIS;
+	desc->lli.dscr = 0;
+}
+
+#endif /* AT_HDMAC_REGS_H */
diff --git a/drivers/dma/dmatest.c b/drivers/dma/dmatest.c
index fb7da5141e9..d93017fc787 100644
--- a/drivers/dma/dmatest.c
+++ b/drivers/dma/dmatest.c
@@ -38,6 +38,11 @@ module_param(max_channels, uint, S_IRUGO);
 MODULE_PARM_DESC(max_channels,
 		"Maximum number of channels to use (default: all)");
 
+static unsigned int iterations;
+module_param(iterations, uint, S_IRUGO);
+MODULE_PARM_DESC(iterations,
+		"Iterations before stopping test (default: infinite)");
+
 static unsigned int xor_sources = 3;
 module_param(xor_sources, uint, S_IRUGO);
 MODULE_PARM_DESC(xor_sources,
@@ -114,7 +119,7 @@ static void dmatest_init_srcs(u8 **bufs, unsigned int start, unsigned int len)
 			buf[i] = PATTERN_SRC | (~i & PATTERN_COUNT_MASK);
 		for ( ; i < start + len; i++)
 			buf[i] = PATTERN_SRC | PATTERN_COPY
-				| (~i & PATTERN_COUNT_MASK);;
+				| (~i & PATTERN_COUNT_MASK);
 		for ( ; i < test_buf_size; i++)
 			buf[i] = PATTERN_SRC | (~i & PATTERN_COUNT_MASK);
 		buf++;
@@ -270,7 +275,8 @@ static int dmatest_func(void *data)
 
 	flags = DMA_CTRL_ACK | DMA_COMPL_SKIP_DEST_UNMAP | DMA_PREP_INTERRUPT;
 
-	while (!kthread_should_stop()) {
+	while (!kthread_should_stop()
+	       && !(iterations && total_tests >= iterations)) {
 		struct dma_device *dev = chan->device;
 		struct dma_async_tx_descriptor *tx = NULL;
 		dma_addr_t dma_srcs[src_cnt];
@@ -416,6 +422,13 @@ err_srcbuf:
 err_srcs:
 	pr_notice("%s: terminating after %u tests, %u failures (status %d)\n",
 			thread_name, total_tests, failed_tests, ret);
+
+	if (iterations > 0)
+		while (!kthread_should_stop()) {
+			DECLARE_WAIT_QUEUE_HEAD(wait_dmatest_exit);
+			interruptible_sleep_on(&wait_dmatest_exit);
+		}
+
 	return ret;
 }
 
@@ -495,11 +508,11 @@ static int dmatest_add_channel(struct dma_chan *chan)
 
 	if (dma_has_cap(DMA_MEMCPY, dma_dev->cap_mask)) {
 		cnt = dmatest_add_threads(dtc, DMA_MEMCPY);
-		thread_count += cnt > 0 ?: 0;
+		thread_count += cnt > 0 ? cnt : 0;
 	}
 	if (dma_has_cap(DMA_XOR, dma_dev->cap_mask)) {
 		cnt = dmatest_add_threads(dtc, DMA_XOR);
-		thread_count += cnt > 0 ?: 0;
+		thread_count += cnt > 0 ? cnt : 0;
 	}
 
 	pr_info("dmatest: Started %u threads using %s\n",
diff --git a/drivers/dma/fsldma.c b/drivers/dma/fsldma.c
index f18d1bde043..ef87a898414 100644
--- a/drivers/dma/fsldma.c
+++ b/drivers/dma/fsldma.c
@@ -12,6 +12,11 @@
  *   also fit for MPC8560, MPC8555, MPC8548, MPC8641, and etc.
  *   The support for MPC8349 DMA contorller is also added.
  *
+ * This driver instructs the DMA controller to issue the PCI Read Multiple
+ * command for PCI read operations, instead of using the default PCI Read Line
+ * command. Please be aware that this setting may result in read pre-fetching
+ * on some platforms.
+ *
  * This is free software; you can redistribute it and/or modify
  * it under the terms of the GNU General Public License as published by
  * the Free Software Foundation; either version 2 of the License, or
@@ -49,9 +54,10 @@ static void dma_init(struct fsl_dma_chan *fsl_chan)
 	case FSL_DMA_IP_83XX:
 		/* Set the channel to below modes:
 		 * EOTIE - End-of-transfer interrupt enable
+		 * PRC_RM - PCI read multiple
 		 */
-		DMA_OUT(fsl_chan, &fsl_chan->reg_base->mr, FSL_DMA_MR_EOTIE,
-				32);
+		DMA_OUT(fsl_chan, &fsl_chan->reg_base->mr, FSL_DMA_MR_EOTIE
+				| FSL_DMA_MR_PRC_RM, 32);
 		break;
 	}
 
@@ -136,15 +142,16 @@ static int dma_is_idle(struct fsl_dma_chan *fsl_chan)
 
 static void dma_start(struct fsl_dma_chan *fsl_chan)
 {
-	u32 mr_set = 0;;
+	u32 mr_set = 0;
 
 	if (fsl_chan->feature & FSL_DMA_CHAN_PAUSE_EXT) {
 		DMA_OUT(fsl_chan, &fsl_chan->reg_base->bcr, 0, 32);
 		mr_set |= FSL_DMA_MR_EMP_EN;
-	} else
+	} else if ((fsl_chan->feature & FSL_DMA_IP_MASK) == FSL_DMA_IP_85XX) {
 		DMA_OUT(fsl_chan, &fsl_chan->reg_base->mr,
 			DMA_IN(fsl_chan, &fsl_chan->reg_base->mr, 32)
 				& ~FSL_DMA_MR_EMP_EN, 32);
+	}
 
 	if (fsl_chan->feature & FSL_DMA_CHAN_START_EXT)
 		mr_set |= FSL_DMA_MR_EMS_EN;
@@ -871,9 +878,9 @@ static int __devinit fsl_dma_chan_probe(struct fsl_dma_device *fdev,
 
 	switch (new_fsl_chan->feature & FSL_DMA_IP_MASK) {
 	case FSL_DMA_IP_85XX:
-		new_fsl_chan->toggle_ext_start = fsl_chan_toggle_ext_start;
 		new_fsl_chan->toggle_ext_pause = fsl_chan_toggle_ext_pause;
 	case FSL_DMA_IP_83XX:
+		new_fsl_chan->toggle_ext_start = fsl_chan_toggle_ext_start;
 		new_fsl_chan->set_src_loop_size = fsl_chan_set_src_loop_size;
 		new_fsl_chan->set_dest_loop_size = fsl_chan_set_dest_loop_size;
 	}
diff --git a/drivers/dma/fsldma.h b/drivers/dma/fsldma.h
index 4f21a512d84..dc7f2686579 100644
--- a/drivers/dma/fsldma.h
+++ b/drivers/dma/fsldma.h
@@ -38,6 +38,7 @@
 
 /* Special MR definition for MPC8349 */
 #define FSL_DMA_MR_EOTIE	0x00000080
+#define FSL_DMA_MR_PRC_RM	0x00000800
 
 #define FSL_DMA_SR_CH		0x00000020
 #define FSL_DMA_SR_PE		0x00000010
diff --git a/drivers/dma/mv_xor.c b/drivers/dma/mv_xor.c
index ddab94f5122..3f23eabe09f 100644
--- a/drivers/dma/mv_xor.c
+++ b/drivers/dma/mv_xor.c
@@ -1176,7 +1176,7 @@ static int __devinit mv_xor_probe(struct platform_device *pdev)
 	if (dma_has_cap(DMA_MEMSET, dma_dev->cap_mask))
 		dma_dev->device_prep_dma_memset = mv_xor_prep_dma_memset;
 	if (dma_has_cap(DMA_XOR, dma_dev->cap_mask)) {
-		dma_dev->max_xor = 8;                  ;
+		dma_dev->max_xor = 8;
 		dma_dev->device_prep_dma_xor = mv_xor_prep_dma_xor;
 	}
 
diff --git a/drivers/edac/amd64_edac.c b/drivers/edac/amd64_edac.c
index 24964c1d0af..e2a10bcba7a 100644
--- a/drivers/edac/amd64_edac.c
+++ b/drivers/edac/amd64_edac.c
@@ -868,6 +868,8 @@ static void amd64_read_dbam_reg(struct amd64_pvt *pvt)
 			goto err_reg;
 	}
 
+	return;
+
 err_reg:
 	debugf0("Error reading F2x%03x.\n", reg);
 }
@@ -2634,6 +2636,8 @@ static void amd64_read_mc_registers(struct amd64_pvt *pvt)
 
 	amd64_dump_misc_regs(pvt);
 
+	return;
+
 err_reg:
 	debugf0("Reading an MC register failed\n");
 
@@ -2977,6 +2981,9 @@ static int amd64_check_ecc_enabled(struct amd64_pvt *pvt)
 			"ECC is enabled by BIOS, Proceeding "
 			"with EDAC module initialization\n");
 
+		/* Signal good ECC status */
+		ret = 0;
+
 		/* CLEAR the override, since BIOS controlled it */
 		ecc_enable_override = 0;
 	}
diff --git a/drivers/edac/x38_edac.c b/drivers/edac/x38_edac.c
index 2406c2ce284..d4ec6059317 100644
--- a/drivers/edac/x38_edac.c
+++ b/drivers/edac/x38_edac.c
@@ -30,7 +30,7 @@
 /* Intel X38 register addresses - device 0 function 0 - DRAM Controller */
 
 #define X38_MCHBAR_LOW	0x48	/* MCH Memory Mapped Register BAR */
-#define X38_MCHBAR_HIGH	0x4b
+#define X38_MCHBAR_HIGH	0x4c
 #define X38_MCHBAR_MASK	0xfffffc000ULL	/* bits 35:14 */
 #define X38_MMR_WINDOW_SIZE	16384
 
diff --git a/drivers/gpu/drm/drm_crtc.c b/drivers/gpu/drm/drm_crtc.c
index 8fab7890a36..33be210d672 100644
--- a/drivers/gpu/drm/drm_crtc.c
+++ b/drivers/gpu/drm/drm_crtc.c
@@ -1461,7 +1461,7 @@ int drm_mode_setcrtc(struct drm_device *dev, void *data,
 		goto out;
 	}
 
-	if (crtc_req->count_connectors > 0 && !mode && !fb) {
+	if (crtc_req->count_connectors > 0 && (!mode || !fb)) {
 		DRM_DEBUG("Count connectors is %d but no mode or fb set\n",
 			  crtc_req->count_connectors);
 		ret = -EINVAL;
diff --git a/drivers/gpu/drm/drm_crtc_helper.c b/drivers/gpu/drm/drm_crtc_helper.c
index 3da9cfa3184..6aaa2cb2336 100644
--- a/drivers/gpu/drm/drm_crtc_helper.c
+++ b/drivers/gpu/drm/drm_crtc_helper.c
@@ -706,8 +706,8 @@ int drm_crtc_helper_set_config(struct drm_mode_set *set)
 	struct drm_encoder **save_encoders, *new_encoder;
 	struct drm_framebuffer *old_fb = NULL;
 	bool save_enabled;
-	bool mode_changed = false;
-	bool fb_changed = false;
+	bool mode_changed = false; /* if true do a full mode set */
+	bool fb_changed = false; /* if true and !mode_changed just do a flip */
 	struct drm_connector *connector;
 	int count = 0, ro, fail = 0;
 	struct drm_crtc_helper_funcs *crtc_funcs;
@@ -758,6 +758,8 @@ int drm_crtc_helper_set_config(struct drm_mode_set *set)
 		if (set->crtc->fb == NULL) {
 			DRM_DEBUG("crtc has no fb, full mode set\n");
 			mode_changed = true;
+		} else if (set->fb == NULL) {
+			mode_changed = true;
 		} else if ((set->fb->bits_per_pixel !=
 			 set->crtc->fb->bits_per_pixel) ||
 			 set->fb->depth != set->crtc->fb->depth)
diff --git a/drivers/gpu/drm/i915/i915_dma.c b/drivers/gpu/drm/i915/i915_dma.c
index 8c4783180bf..50d1f782768 100644
--- a/drivers/gpu/drm/i915/i915_dma.c
+++ b/drivers/gpu/drm/i915/i915_dma.c
@@ -1186,6 +1186,13 @@ int i915_driver_load(struct drm_device *dev, unsigned long flags)
 	if (ret)
 		goto out_iomapfree;
 
+	dev_priv->wq = create_workqueue("i915");
+	if (dev_priv->wq == NULL) {
+		DRM_ERROR("Failed to create our workqueue.\n");
+		ret = -ENOMEM;
+		goto out_iomapfree;
+	}
+
 	/* enable GEM by default */
 	dev_priv->has_gem = 1;
 
@@ -1211,7 +1218,7 @@ int i915_driver_load(struct drm_device *dev, unsigned long flags)
 	if (!I915_NEED_GFX_HWS(dev)) {
 		ret = i915_init_phys_hws(dev);
 		if (ret != 0)
-			goto out_iomapfree;
+			goto out_workqueue_free;
 	}
 
 	i915_get_mem_freq(dev);
@@ -1245,7 +1252,7 @@ int i915_driver_load(struct drm_device *dev, unsigned long flags)
 		ret = i915_load_modeset_init(dev, prealloc_size, agp_size);
 		if (ret < 0) {
 			DRM_ERROR("failed to init modeset\n");
-			goto out_rmmap;
+			goto out_workqueue_free;
 		}
 	}
 
@@ -1256,6 +1263,8 @@ int i915_driver_load(struct drm_device *dev, unsigned long flags)
 
 	return 0;
 
+out_workqueue_free:
+	destroy_workqueue(dev_priv->wq);
 out_iomapfree:
 	io_mapping_free(dev_priv->mm.gtt_mapping);
 out_rmmap:
@@ -1269,6 +1278,8 @@ int i915_driver_unload(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
 
+	destroy_workqueue(dev_priv->wq);
+
 	io_mapping_free(dev_priv->mm.gtt_mapping);
 	if (dev_priv->mm.gtt_mtrr >= 0) {
 		mtrr_del(dev_priv->mm.gtt_mtrr, dev->agp->base,
diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h
index d0875287588..7537f57d8a8 100644
--- a/drivers/gpu/drm/i915/i915_drv.h
+++ b/drivers/gpu/drm/i915/i915_drv.h
@@ -219,6 +219,7 @@ typedef struct drm_i915_private {
 	unsigned int lvds_vbt:1;
 	unsigned int int_crt_support:1;
 	unsigned int lvds_use_ssc:1;
+	unsigned int edp_support:1;
 	int lvds_ssc_freq;
 
 	struct drm_i915_fence_reg fence_regs[16]; /* assume 965 */
@@ -229,6 +230,8 @@ typedef struct drm_i915_private {
 
 	spinlock_t error_lock;
 	struct drm_i915_error_state *first_error;
+	struct work_struct error_work;
+	struct workqueue_struct *wq;
 
 	/* Register state */
 	u8 saveLBB;
@@ -888,6 +891,7 @@ extern int i915_wait_ring(struct drm_device * dev, int n, const char *caller);
 						      IS_I915GM(dev)))
 #define SUPPORTS_INTEGRATED_HDMI(dev)	(IS_G4X(dev) || IS_IGDNG(dev))
 #define SUPPORTS_INTEGRATED_DP(dev)	(IS_G4X(dev) || IS_IGDNG(dev))
+#define SUPPORTS_EDP(dev)		(IS_IGDNG_M(dev))
 #define I915_HAS_HOTPLUG(dev) (IS_I945G(dev) || IS_I945GM(dev) || IS_I965G(dev))
 /* dsparb controlled by hw only */
 #define DSPARB_HWCONTROL(dev) (IS_G4X(dev) || IS_IGDNG(dev))
diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
index 5bf420378b6..140bee142fc 100644
--- a/drivers/gpu/drm/i915/i915_gem.c
+++ b/drivers/gpu/drm/i915/i915_gem.c
@@ -1570,7 +1570,7 @@ i915_add_request(struct drm_device *dev, struct drm_file *file_priv,
 	}
 
 	if (was_empty && !dev_priv->mm.suspended)
-		schedule_delayed_work(&dev_priv->mm.retire_work, HZ);
+		queue_delayed_work(dev_priv->wq, &dev_priv->mm.retire_work, HZ);
 	return seqno;
 }
 
@@ -1719,7 +1719,7 @@ i915_gem_retire_work_handler(struct work_struct *work)
 	i915_gem_retire_requests(dev);
 	if (!dev_priv->mm.suspended &&
 	    !list_empty(&dev_priv->mm.request_list))
-		schedule_delayed_work(&dev_priv->mm.retire_work, HZ);
+		queue_delayed_work(dev_priv->wq, &dev_priv->mm.retire_work, HZ);
 	mutex_unlock(&dev->struct_mutex);
 }
 
diff --git a/drivers/gpu/drm/i915/i915_gem_debugfs.c b/drivers/gpu/drm/i915/i915_gem_debugfs.c
index 9a44bfcb813..cb3b97405fb 100644
--- a/drivers/gpu/drm/i915/i915_gem_debugfs.c
+++ b/drivers/gpu/drm/i915/i915_gem_debugfs.c
@@ -343,6 +343,8 @@ static int i915_error_state(struct seq_file *m, void *unused)
 
 	error = dev_priv->first_error;
 
+	seq_printf(m, "Time: %ld s %ld us\n", error->time.tv_sec,
+		   error->time.tv_usec);
 	seq_printf(m, "EIR: 0x%08x\n", error->eir);
 	seq_printf(m, "  PGTBL_ER: 0x%08x\n", error->pgtbl_er);
 	seq_printf(m, "  INSTPM: 0x%08x\n", error->instpm);
diff --git a/drivers/gpu/drm/i915/i915_irq.c b/drivers/gpu/drm/i915/i915_irq.c
index 7ba23a69a0c..83aee80e77a 100644
--- a/drivers/gpu/drm/i915/i915_irq.c
+++ b/drivers/gpu/drm/i915/i915_irq.c
@@ -290,6 +290,35 @@ irqreturn_t igdng_irq_handler(struct drm_device *dev)
 	return ret;
 }
 
+/**
+ * i915_error_work_func - do process context error handling work
+ * @work: work struct
+ *
+ * Fire an error uevent so userspace can see that a hang or error
+ * was detected.
+ */
+static void i915_error_work_func(struct work_struct *work)
+{
+	drm_i915_private_t *dev_priv = container_of(work, drm_i915_private_t,
+						    error_work);
+	struct drm_device *dev = dev_priv->dev;
+	char *event_string = "ERROR=1";
+	char *envp[] = { event_string, NULL };
+
+	DRM_DEBUG("generating error event\n");
+
+	kobject_uevent_env(&dev->primary->kdev.kobj, KOBJ_CHANGE, envp);
+}
+
+/**
+ * i915_capture_error_state - capture an error record for later analysis
+ * @dev: drm device
+ *
+ * Should be called when an error is detected (either a hang or an error
+ * interrupt) to capture error state from the time of the error.  Fills
+ * out a structure which becomes available in debugfs for user level tools
+ * to pick up.
+ */
 static void i915_capture_error_state(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
@@ -325,12 +354,137 @@ static void i915_capture_error_state(struct drm_device *dev)
 		error->acthd = I915_READ(ACTHD_I965);
 	}
 
+	do_gettimeofday(&error->time);
+
 	dev_priv->first_error = error;
 
 out:
 	spin_unlock_irqrestore(&dev_priv->error_lock, flags);
 }
 
+/**
+ * i915_handle_error - handle an error interrupt
+ * @dev: drm device
+ *
+ * Do some basic checking of regsiter state at error interrupt time and
+ * dump it to the syslog.  Also call i915_capture_error_state() to make
+ * sure we get a record and make it available in debugfs.  Fire a uevent
+ * so userspace knows something bad happened (should trigger collection
+ * of a ring dump etc.).
+ */
+static void i915_handle_error(struct drm_device *dev)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	u32 eir = I915_READ(EIR);
+	u32 pipea_stats = I915_READ(PIPEASTAT);
+	u32 pipeb_stats = I915_READ(PIPEBSTAT);
+
+	i915_capture_error_state(dev);
+
+	printk(KERN_ERR "render error detected, EIR: 0x%08x\n",
+	       eir);
+
+	if (IS_G4X(dev)) {
+		if (eir & (GM45_ERROR_MEM_PRIV | GM45_ERROR_CP_PRIV)) {
+			u32 ipeir = I915_READ(IPEIR_I965);
+
+			printk(KERN_ERR "  IPEIR: 0x%08x\n",
+			       I915_READ(IPEIR_I965));
+			printk(KERN_ERR "  IPEHR: 0x%08x\n",
+			       I915_READ(IPEHR_I965));
+			printk(KERN_ERR "  INSTDONE: 0x%08x\n",
+			       I915_READ(INSTDONE_I965));
+			printk(KERN_ERR "  INSTPS: 0x%08x\n",
+			       I915_READ(INSTPS));
+			printk(KERN_ERR "  INSTDONE1: 0x%08x\n",
+			       I915_READ(INSTDONE1));
+			printk(KERN_ERR "  ACTHD: 0x%08x\n",
+			       I915_READ(ACTHD_I965));
+			I915_WRITE(IPEIR_I965, ipeir);
+			(void)I915_READ(IPEIR_I965);
+		}
+		if (eir & GM45_ERROR_PAGE_TABLE) {
+			u32 pgtbl_err = I915_READ(PGTBL_ER);
+			printk(KERN_ERR "page table error\n");
+			printk(KERN_ERR "  PGTBL_ER: 0x%08x\n",
+			       pgtbl_err);
+			I915_WRITE(PGTBL_ER, pgtbl_err);
+			(void)I915_READ(PGTBL_ER);
+		}
+	}
+
+	if (IS_I9XX(dev)) {
+		if (eir & I915_ERROR_PAGE_TABLE) {
+			u32 pgtbl_err = I915_READ(PGTBL_ER);
+			printk(KERN_ERR "page table error\n");
+			printk(KERN_ERR "  PGTBL_ER: 0x%08x\n",
+			       pgtbl_err);
+			I915_WRITE(PGTBL_ER, pgtbl_err);
+			(void)I915_READ(PGTBL_ER);
+		}
+	}
+
+	if (eir & I915_ERROR_MEMORY_REFRESH) {
+		printk(KERN_ERR "memory refresh error\n");
+		printk(KERN_ERR "PIPEASTAT: 0x%08x\n",
+		       pipea_stats);
+		printk(KERN_ERR "PIPEBSTAT: 0x%08x\n",
+		       pipeb_stats);
+		/* pipestat has already been acked */
+	}
+	if (eir & I915_ERROR_INSTRUCTION) {
+		printk(KERN_ERR "instruction error\n");
+		printk(KERN_ERR "  INSTPM: 0x%08x\n",
+		       I915_READ(INSTPM));
+		if (!IS_I965G(dev)) {
+			u32 ipeir = I915_READ(IPEIR);
+
+			printk(KERN_ERR "  IPEIR: 0x%08x\n",
+			       I915_READ(IPEIR));
+			printk(KERN_ERR "  IPEHR: 0x%08x\n",
+			       I915_READ(IPEHR));
+			printk(KERN_ERR "  INSTDONE: 0x%08x\n",
+			       I915_READ(INSTDONE));
+			printk(KERN_ERR "  ACTHD: 0x%08x\n",
+			       I915_READ(ACTHD));
+			I915_WRITE(IPEIR, ipeir);
+			(void)I915_READ(IPEIR);
+		} else {
+			u32 ipeir = I915_READ(IPEIR_I965);
+
+			printk(KERN_ERR "  IPEIR: 0x%08x\n",
+			       I915_READ(IPEIR_I965));
+			printk(KERN_ERR "  IPEHR: 0x%08x\n",
+			       I915_READ(IPEHR_I965));
+			printk(KERN_ERR "  INSTDONE: 0x%08x\n",
+			       I915_READ(INSTDONE_I965));
+			printk(KERN_ERR "  INSTPS: 0x%08x\n",
+			       I915_READ(INSTPS));
+			printk(KERN_ERR "  INSTDONE1: 0x%08x\n",
+			       I915_READ(INSTDONE1));
+			printk(KERN_ERR "  ACTHD: 0x%08x\n",
+			       I915_READ(ACTHD_I965));
+			I915_WRITE(IPEIR_I965, ipeir);
+			(void)I915_READ(IPEIR_I965);
+		}
+	}
+
+	I915_WRITE(EIR, eir);
+	(void)I915_READ(EIR);
+	eir = I915_READ(EIR);
+	if (eir) {
+		/*
+		 * some errors might have become stuck,
+		 * mask them.
+		 */
+		DRM_ERROR("EIR stuck: 0x%08x, masking\n", eir);
+		I915_WRITE(EMR, I915_READ(EMR) | eir);
+		I915_WRITE(IIR, I915_RENDER_COMMAND_PARSER_ERROR_INTERRUPT);
+	}
+
+	queue_work(dev_priv->wq, &dev_priv->error_work);
+}
+
 irqreturn_t i915_driver_irq_handler(DRM_IRQ_ARGS)
 {
 	struct drm_device *dev = (struct drm_device *) arg;
@@ -372,6 +526,9 @@ irqreturn_t i915_driver_irq_handler(DRM_IRQ_ARGS)
 		pipea_stats = I915_READ(PIPEASTAT);
 		pipeb_stats = I915_READ(PIPEBSTAT);
 
+		if (iir & I915_RENDER_COMMAND_PARSER_ERROR_INTERRUPT)
+			i915_handle_error(dev);
+
 		/*
 		 * Clear the PIPE(A|B)STAT regs before the IIR
 		 */
@@ -403,86 +560,13 @@ irqreturn_t i915_driver_irq_handler(DRM_IRQ_ARGS)
 			DRM_DEBUG("hotplug event received, stat 0x%08x\n",
 				  hotplug_status);
 			if (hotplug_status & dev_priv->hotplug_supported_mask)
-				schedule_work(&dev_priv->hotplug_work);
+				queue_work(dev_priv->wq,
+					   &dev_priv->hotplug_work);
 
 			I915_WRITE(PORT_HOTPLUG_STAT, hotplug_status);
 			I915_READ(PORT_HOTPLUG_STAT);
 		}
 
-		if (iir & I915_RENDER_COMMAND_PARSER_ERROR_INTERRUPT) {
-			u32 eir = I915_READ(EIR);
-
-			i915_capture_error_state(dev);
-
-			printk(KERN_ERR "render error detected, EIR: 0x%08x\n",
-			       eir);
-			if (eir & I915_ERROR_PAGE_TABLE) {
-				u32 pgtbl_err = I915_READ(PGTBL_ER);
-				printk(KERN_ERR "page table error\n");
-				printk(KERN_ERR "  PGTBL_ER: 0x%08x\n",
-				       pgtbl_err);
-				I915_WRITE(PGTBL_ER, pgtbl_err);
-				(void)I915_READ(PGTBL_ER);
-			}
-			if (eir & I915_ERROR_MEMORY_REFRESH) {
-				printk(KERN_ERR "memory refresh error\n");
-				printk(KERN_ERR "PIPEASTAT: 0x%08x\n",
-				       pipea_stats);
-				printk(KERN_ERR "PIPEBSTAT: 0x%08x\n",
-				       pipeb_stats);
-				/* pipestat has already been acked */
-			}
-			if (eir & I915_ERROR_INSTRUCTION) {
-				printk(KERN_ERR "instruction error\n");
-				printk(KERN_ERR "  INSTPM: 0x%08x\n",
-				       I915_READ(INSTPM));
-				if (!IS_I965G(dev)) {
-					u32 ipeir = I915_READ(IPEIR);
-
-					printk(KERN_ERR "  IPEIR: 0x%08x\n",
-					       I915_READ(IPEIR));
-					printk(KERN_ERR "  IPEHR: 0x%08x\n",
-						   I915_READ(IPEHR));
-					printk(KERN_ERR "  INSTDONE: 0x%08x\n",
-						   I915_READ(INSTDONE));
-					printk(KERN_ERR "  ACTHD: 0x%08x\n",
-						   I915_READ(ACTHD));
-					I915_WRITE(IPEIR, ipeir);
-					(void)I915_READ(IPEIR);
-				} else {
-					u32 ipeir = I915_READ(IPEIR_I965);
-
-					printk(KERN_ERR "  IPEIR: 0x%08x\n",
-					       I915_READ(IPEIR_I965));
-					printk(KERN_ERR "  IPEHR: 0x%08x\n",
-					       I915_READ(IPEHR_I965));
-					printk(KERN_ERR "  INSTDONE: 0x%08x\n",
-					       I915_READ(INSTDONE_I965));
-					printk(KERN_ERR "  INSTPS: 0x%08x\n",
-					       I915_READ(INSTPS));
-					printk(KERN_ERR "  INSTDONE1: 0x%08x\n",
-					       I915_READ(INSTDONE1));
-					printk(KERN_ERR "  ACTHD: 0x%08x\n",
-					       I915_READ(ACTHD_I965));
-					I915_WRITE(IPEIR_I965, ipeir);
-					(void)I915_READ(IPEIR_I965);
-				}
-			}
-
-			I915_WRITE(EIR, eir);
-			(void)I915_READ(EIR);
-			eir = I915_READ(EIR);
-			if (eir) {
-				/*
-				 * some errors might have become stuck,
-				 * mask them.
-				 */
-				DRM_ERROR("EIR stuck: 0x%08x, masking\n", eir);
-				I915_WRITE(EMR, I915_READ(EMR) | eir);
-				I915_WRITE(IIR, I915_RENDER_COMMAND_PARSER_ERROR_INTERRUPT);
-			}
-		}
-
 		I915_WRITE(IIR, iir);
 		new_iir = I915_READ(IIR); /* Flush posted writes */
 
@@ -830,6 +914,7 @@ void i915_driver_irq_preinstall(struct drm_device * dev)
 	atomic_set(&dev_priv->irq_received, 0);
 
 	INIT_WORK(&dev_priv->hotplug_work, i915_hotplug_work_func);
+	INIT_WORK(&dev_priv->error_work, i915_error_work_func);
 
 	if (IS_IGDNG(dev)) {
 		igdng_irq_preinstall(dev);
diff --git a/drivers/gpu/drm/i915/i915_reg.h b/drivers/gpu/drm/i915/i915_reg.h
index 6c085848409..2955083aa47 100644
--- a/drivers/gpu/drm/i915/i915_reg.h
+++ b/drivers/gpu/drm/i915/i915_reg.h
@@ -1395,6 +1395,7 @@
 #define TV_V_CHROMA_42		0x684a8
 
 /* Display Port */
+#define DP_A				0x64000 /* eDP */
 #define DP_B				0x64100
 #define DP_C				0x64200
 #define DP_D				0x64300
@@ -1437,13 +1438,22 @@
 /* Mystic DPCD version 1.1 special mode */
 #define   DP_ENHANCED_FRAMING		(1 << 18)
 
+/* eDP */
+#define   DP_PLL_FREQ_270MHZ		(0 << 16)
+#define   DP_PLL_FREQ_160MHZ		(1 << 16)
+#define   DP_PLL_FREQ_MASK		(3 << 16)
+
 /** locked once port is enabled */
 #define   DP_PORT_REVERSAL		(1 << 15)
 
+/* eDP */
+#define   DP_PLL_ENABLE			(1 << 14)
+
 /** sends the clock on lane 15 of the PEG for debug */
 #define   DP_CLOCK_OUTPUT_ENABLE	(1 << 13)
 
 #define   DP_SCRAMBLING_DISABLE		(1 << 12)
+#define   DP_SCRAMBLING_DISABLE_IGDNG	(1 << 7)
 
 /** limit RGB values to avoid confusing TVs */
 #define   DP_COLOR_RANGE_16_235		(1 << 8)
@@ -1463,6 +1473,13 @@
  * is 20 bytes in each direction, hence the 5 fixed
  * data registers
  */
+#define DPA_AUX_CH_CTL			0x64010
+#define DPA_AUX_CH_DATA1		0x64014
+#define DPA_AUX_CH_DATA2		0x64018
+#define DPA_AUX_CH_DATA3		0x6401c
+#define DPA_AUX_CH_DATA4		0x64020
+#define DPA_AUX_CH_DATA5		0x64024
+
 #define DPB_AUX_CH_CTL			0x64110
 #define DPB_AUX_CH_DATA1		0x64114
 #define DPB_AUX_CH_DATA2		0x64118
@@ -1618,7 +1635,7 @@
 #define I830_FIFO_LINE_SIZE	32
 #define I945_FIFO_SIZE		127 /* 945 & 965 */
 #define I915_FIFO_SIZE		95
-#define I855GM_FIFO_SIZE	255
+#define I855GM_FIFO_SIZE	127 /* In cachelines */
 #define I830_FIFO_SIZE		95
 #define I915_MAX_WM		0x3f
 
@@ -1848,6 +1865,8 @@
 #define PFA_CTL_1               0x68080
 #define PFB_CTL_1               0x68880
 #define  PF_ENABLE              (1<<31)
+#define PFA_WIN_SZ		0x68074
+#define PFB_WIN_SZ		0x68874
 
 /* legacy palette */
 #define LGC_PALETTE_A           0x4a000
@@ -2208,4 +2227,28 @@
 #define PCH_PP_OFF_DELAYS	0xc720c
 #define PCH_PP_DIVISOR		0xc7210
 
+#define PCH_DP_B		0xe4100
+#define PCH_DPB_AUX_CH_CTL	0xe4110
+#define PCH_DPB_AUX_CH_DATA1	0xe4114
+#define PCH_DPB_AUX_CH_DATA2	0xe4118
+#define PCH_DPB_AUX_CH_DATA3	0xe411c
+#define PCH_DPB_AUX_CH_DATA4	0xe4120
+#define PCH_DPB_AUX_CH_DATA5	0xe4124
+
+#define PCH_DP_C		0xe4200
+#define PCH_DPC_AUX_CH_CTL	0xe4210
+#define PCH_DPC_AUX_CH_DATA1	0xe4214
+#define PCH_DPC_AUX_CH_DATA2	0xe4218
+#define PCH_DPC_AUX_CH_DATA3	0xe421c
+#define PCH_DPC_AUX_CH_DATA4	0xe4220
+#define PCH_DPC_AUX_CH_DATA5	0xe4224
+
+#define PCH_DP_D		0xe4300
+#define PCH_DPD_AUX_CH_CTL	0xe4310
+#define PCH_DPD_AUX_CH_DATA1	0xe4314
+#define PCH_DPD_AUX_CH_DATA2	0xe4318
+#define PCH_DPD_AUX_CH_DATA3	0xe431c
+#define PCH_DPD_AUX_CH_DATA4	0xe4320
+#define PCH_DPD_AUX_CH_DATA5	0xe4324
+
 #endif /* _I915_REG_H_ */
diff --git a/drivers/gpu/drm/i915/i915_suspend.c b/drivers/gpu/drm/i915/i915_suspend.c
index 9e1d16e5c3e..1d04e1904ac 100644
--- a/drivers/gpu/drm/i915/i915_suspend.c
+++ b/drivers/gpu/drm/i915/i915_suspend.c
@@ -598,7 +598,7 @@ int i915_restore_state(struct drm_device *dev)
 
 	for (i = 0; i < 16; i++) {
 		I915_WRITE(SWF00 + (i << 2), dev_priv->saveSWF0[i]);
-		I915_WRITE(SWF10 + (i << 2), dev_priv->saveSWF1[i+7]);
+		I915_WRITE(SWF10 + (i << 2), dev_priv->saveSWF1[i]);
 	}
 	for (i = 0; i < 3; i++)
 		I915_WRITE(SWF30 + (i << 2), dev_priv->saveSWF2[i]);
diff --git a/drivers/gpu/drm/i915/intel_bios.c b/drivers/gpu/drm/i915/intel_bios.c
index 7cc44719102..300aee3296c 100644
--- a/drivers/gpu/drm/i915/intel_bios.c
+++ b/drivers/gpu/drm/i915/intel_bios.c
@@ -97,14 +97,13 @@ static void
 parse_lfp_panel_data(struct drm_i915_private *dev_priv,
 			    struct bdb_header *bdb)
 {
-	struct drm_device *dev = dev_priv->dev;
 	struct bdb_lvds_options *lvds_options;
 	struct bdb_lvds_lfp_data *lvds_lfp_data;
 	struct bdb_lvds_lfp_data_ptrs *lvds_lfp_data_ptrs;
 	struct bdb_lvds_lfp_data_entry *entry;
 	struct lvds_dvo_timing *dvo_timing;
 	struct drm_display_mode *panel_fixed_mode;
-	int lfp_data_size;
+	int lfp_data_size, dvo_timing_offset;
 
 	/* Defaults if we can't find VBT info */
 	dev_priv->lvds_dither = 0;
@@ -133,14 +132,16 @@ parse_lfp_panel_data(struct drm_i915_private *dev_priv,
 	entry = (struct bdb_lvds_lfp_data_entry *)
 		((uint8_t *)lvds_lfp_data->data + (lfp_data_size *
 						   lvds_options->panel_type));
+	dvo_timing_offset = lvds_lfp_data_ptrs->ptr[0].dvo_timing_offset -
+		lvds_lfp_data_ptrs->ptr[0].fp_timing_offset;
 
-	/* On IGDNG mobile, LVDS data block removes panel fitting registers.
-	   So dec 2 dword from dvo_timing offset */
-	if (IS_IGDNG(dev))
-		dvo_timing = (struct lvds_dvo_timing *)
-					((u8 *)&entry->dvo_timing - 8);
-	else
-		dvo_timing = &entry->dvo_timing;
+	/*
+	 * the size of fp_timing varies on the different platform.
+	 * So calculate the DVO timing relative offset in LVDS data
+	 * entry to get the DVO timing entry
+	 */
+	dvo_timing = (struct lvds_dvo_timing *)
+			((unsigned char *)entry + dvo_timing_offset);
 
 	panel_fixed_mode = kzalloc(sizeof(*panel_fixed_mode), GFP_KERNEL);
 
@@ -295,6 +296,25 @@ parse_sdvo_device_mapping(struct drm_i915_private *dev_priv,
 	}
 	return;
 }
+
+static void
+parse_driver_features(struct drm_i915_private *dev_priv,
+		       struct bdb_header *bdb)
+{
+	struct drm_device *dev = dev_priv->dev;
+	struct bdb_driver_features *driver;
+
+	/* set default for chips without eDP */
+	if (!SUPPORTS_EDP(dev)) {
+		dev_priv->edp_support = 0;
+		return;
+	}
+
+	driver = find_section(bdb, BDB_DRIVER_FEATURES);
+	if (driver && driver->lvds_config == BDB_DRIVER_FEATURE_EDP)
+		dev_priv->edp_support = 1;
+}
+
 /**
  * intel_init_bios - initialize VBIOS settings & find VBT
  * @dev: DRM device
@@ -345,6 +365,8 @@ intel_init_bios(struct drm_device *dev)
 	parse_lfp_panel_data(dev_priv, bdb);
 	parse_sdvo_panel_data(dev_priv, bdb);
 	parse_sdvo_device_mapping(dev_priv, bdb);
+	parse_driver_features(dev_priv, bdb);
+
 	pci_unmap_rom(pdev, bios);
 
 	return 0;
diff --git a/drivers/gpu/drm/i915/intel_bios.h b/drivers/gpu/drm/i915/intel_bios.h
index fe72e1c225d..0f8e5f69ac7 100644
--- a/drivers/gpu/drm/i915/intel_bios.h
+++ b/drivers/gpu/drm/i915/intel_bios.h
@@ -381,6 +381,51 @@ struct bdb_sdvo_lvds_options {
 } __attribute__((packed));
 
 
+#define BDB_DRIVER_FEATURE_NO_LVDS		0
+#define BDB_DRIVER_FEATURE_INT_LVDS		1
+#define BDB_DRIVER_FEATURE_SDVO_LVDS		2
+#define BDB_DRIVER_FEATURE_EDP			3
+
+struct bdb_driver_features {
+	u8 boot_dev_algorithm:1;
+	u8 block_display_switch:1;
+	u8 allow_display_switch:1;
+	u8 hotplug_dvo:1;
+	u8 dual_view_zoom:1;
+	u8 int15h_hook:1;
+	u8 sprite_in_clone:1;
+	u8 primary_lfp_id:1;
+
+	u16 boot_mode_x;
+	u16 boot_mode_y;
+	u8 boot_mode_bpp;
+	u8 boot_mode_refresh;
+
+	u16 enable_lfp_primary:1;
+	u16 selective_mode_pruning:1;
+	u16 dual_frequency:1;
+	u16 render_clock_freq:1; /* 0: high freq; 1: low freq */
+	u16 nt_clone_support:1;
+	u16 power_scheme_ui:1; /* 0: CUI; 1: 3rd party */
+	u16 sprite_display_assign:1; /* 0: secondary; 1: primary */
+	u16 cui_aspect_scaling:1;
+	u16 preserve_aspect_ratio:1;
+	u16 sdvo_device_power_down:1;
+	u16 crt_hotplug:1;
+	u16 lvds_config:2;
+	u16 tv_hotplug:1;
+	u16 hdmi_config:2;
+
+	u8 static_display:1;
+	u8 reserved2:7;
+	u16 legacy_crt_max_x;
+	u16 legacy_crt_max_y;
+	u8 legacy_crt_max_refresh;
+
+	u8 hdmi_termination;
+	u8 custom_vbt_version;
+} __attribute__((packed));
+
 bool intel_init_bios(struct drm_device *dev);
 
 /*
diff --git a/drivers/gpu/drm/i915/intel_crt.c b/drivers/gpu/drm/i915/intel_crt.c
index d6a1a6e5539..4cf8e2e88a4 100644
--- a/drivers/gpu/drm/i915/intel_crt.c
+++ b/drivers/gpu/drm/i915/intel_crt.c
@@ -156,6 +156,9 @@ static bool intel_igdng_crt_detect_hotplug(struct drm_connector *connector)
 
 	temp = adpa = I915_READ(PCH_ADPA);
 
+	adpa &= ~ADPA_DAC_ENABLE;
+	I915_WRITE(PCH_ADPA, adpa);
+
 	adpa &= ~ADPA_CRT_HOTPLUG_MASK;
 
 	adpa |= (ADPA_CRT_HOTPLUG_PERIOD_128 |
@@ -169,13 +172,14 @@ static bool intel_igdng_crt_detect_hotplug(struct drm_connector *connector)
 	DRM_DEBUG("pch crt adpa 0x%x", adpa);
 	I915_WRITE(PCH_ADPA, adpa);
 
-	/* This might not be needed as not specified in spec...*/
-	udelay(1000);
+	while ((I915_READ(PCH_ADPA) & ADPA_CRT_HOTPLUG_FORCE_TRIGGER) != 0)
+		;
 
 	/* Check the status to see if both blue and green are on now */
 	adpa = I915_READ(PCH_ADPA);
-	if ((adpa & ADPA_CRT_HOTPLUG_MONITOR_MASK) ==
-			ADPA_CRT_HOTPLUG_MONITOR_COLOR)
+	adpa &= ADPA_CRT_HOTPLUG_MONITOR_MASK;
+	if ((adpa == ADPA_CRT_HOTPLUG_MONITOR_COLOR) ||
+		(adpa == ADPA_CRT_HOTPLUG_MONITOR_MONO))
 		ret = true;
 	else
 		ret = false;
diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c
index 508838ee31e..d6fce213341 100644
--- a/drivers/gpu/drm/i915/intel_display.c
+++ b/drivers/gpu/drm/i915/intel_display.c
@@ -34,6 +34,8 @@
 
 #include "drm_crtc_helper.h"
 
+#define HAS_eDP (intel_pipe_has_type(crtc, INTEL_OUTPUT_EDP))
+
 bool intel_pipe_has_type (struct drm_crtc *crtc, int type);
 static void intel_update_watermarks(struct drm_device *dev);
 
@@ -88,7 +90,7 @@ struct intel_limit {
 #define I8XX_P2_SLOW		      4
 #define I8XX_P2_FAST		      2
 #define I8XX_P2_LVDS_SLOW	      14
-#define I8XX_P2_LVDS_FAST	      14 /* No fast option */
+#define I8XX_P2_LVDS_FAST	      7
 #define I8XX_P2_SLOW_LIMIT	 165000
 
 #define I9XX_DOT_MIN		  20000
@@ -268,6 +270,9 @@ intel_igdng_find_best_PLL(const intel_limit_t *limit, struct drm_crtc *crtc,
 static bool
 intel_find_pll_g4x_dp(const intel_limit_t *, struct drm_crtc *crtc,
 		      int target, int refclk, intel_clock_t *best_clock);
+static bool
+intel_find_pll_igdng_dp(const intel_limit_t *, struct drm_crtc *crtc,
+		      int target, int refclk, intel_clock_t *best_clock);
 
 static const intel_limit_t intel_limits_i8xx_dvo = {
         .dot = { .min = I8XX_DOT_MIN,		.max = I8XX_DOT_MAX },
@@ -598,6 +603,23 @@ bool intel_pipe_has_type (struct drm_crtc *crtc, int type)
     return false;
 }
 
+struct drm_connector *
+intel_pipe_get_output (struct drm_crtc *crtc)
+{
+    struct drm_device *dev = crtc->dev;
+    struct drm_mode_config *mode_config = &dev->mode_config;
+    struct drm_connector *l_entry, *ret = NULL;
+
+    list_for_each_entry(l_entry, &mode_config->connector_list, head) {
+	    if (l_entry->encoder &&
+	        l_entry->encoder->crtc == crtc) {
+		    ret = l_entry;
+		    break;
+	    }
+    }
+    return ret;
+}
+
 #define INTELPllInvalid(s)   do { /* DRM_DEBUG(s); */ return false; } while (0)
 /**
  * Returns whether the given set of divisors are valid for a given refclk with
@@ -645,7 +667,7 @@ intel_find_best_PLL(const intel_limit_t *limit, struct drm_crtc *crtc,
 	int err = target;
 
 	if (IS_I9XX(dev) && intel_pipe_has_type(crtc, INTEL_OUTPUT_LVDS) &&
-	    (I915_READ(LVDS) & LVDS_PORT_EN) != 0) {
+	    (I915_READ(LVDS)) != 0) {
 		/*
 		 * For LVDS, if the panel is on, just rely on its current
 		 * settings for dual-channel.  We haven't figured out how to
@@ -752,6 +774,30 @@ intel_g4x_find_best_PLL(const intel_limit_t *limit, struct drm_crtc *crtc,
 }
 
 static bool
+intel_find_pll_igdng_dp(const intel_limit_t *limit, struct drm_crtc *crtc,
+		      int target, int refclk, intel_clock_t *best_clock)
+{
+	struct drm_device *dev = crtc->dev;
+	intel_clock_t clock;
+	if (target < 200000) {
+		clock.n = 1;
+		clock.p1 = 2;
+		clock.p2 = 10;
+		clock.m1 = 12;
+		clock.m2 = 9;
+	} else {
+		clock.n = 2;
+		clock.p1 = 1;
+		clock.p2 = 10;
+		clock.m1 = 14;
+		clock.m2 = 8;
+	}
+	intel_clock(dev, refclk, &clock);
+	memcpy(best_clock, &clock, sizeof(intel_clock_t));
+	return true;
+}
+
+static bool
 intel_igdng_find_best_PLL(const intel_limit_t *limit, struct drm_crtc *crtc,
 			int target, int refclk, intel_clock_t *best_clock)
 {
@@ -763,6 +809,14 @@ intel_igdng_find_best_PLL(const intel_limit_t *limit, struct drm_crtc *crtc,
 	int err_most = 47;
 	found = false;
 
+	/* eDP has only 2 clock choice, no n/m/p setting */
+	if (HAS_eDP)
+		return true;
+
+	if (intel_pipe_has_type(crtc, INTEL_OUTPUT_DISPLAYPORT))
+		return intel_find_pll_igdng_dp(limit, crtc, target,
+					       refclk, best_clock);
+
 	if (intel_pipe_has_type(crtc, INTEL_OUTPUT_LVDS)) {
 		if ((I915_READ(LVDS) & LVDS_CLKB_POWER_MASK) ==
 		    LVDS_CLKB_POWER_UP)
@@ -998,6 +1052,90 @@ intel_pipe_set_base(struct drm_crtc *crtc, int x, int y,
 	return 0;
 }
 
+/* Disable the VGA plane that we never use */
+static void i915_disable_vga (struct drm_device *dev)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	u8 sr1;
+	u32 vga_reg;
+
+	if (IS_IGDNG(dev))
+		vga_reg = CPU_VGACNTRL;
+	else
+		vga_reg = VGACNTRL;
+
+	if (I915_READ(vga_reg) & VGA_DISP_DISABLE)
+		return;
+
+	I915_WRITE8(VGA_SR_INDEX, 1);
+	sr1 = I915_READ8(VGA_SR_DATA);
+	I915_WRITE8(VGA_SR_DATA, sr1 | (1 << 5));
+	udelay(100);
+
+	I915_WRITE(vga_reg, VGA_DISP_DISABLE);
+}
+
+static void igdng_disable_pll_edp (struct drm_crtc *crtc)
+{
+	struct drm_device *dev = crtc->dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	u32 dpa_ctl;
+
+	DRM_DEBUG("\n");
+	dpa_ctl = I915_READ(DP_A);
+	dpa_ctl &= ~DP_PLL_ENABLE;
+	I915_WRITE(DP_A, dpa_ctl);
+}
+
+static void igdng_enable_pll_edp (struct drm_crtc *crtc)
+{
+	struct drm_device *dev = crtc->dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	u32 dpa_ctl;
+
+	dpa_ctl = I915_READ(DP_A);
+	dpa_ctl |= DP_PLL_ENABLE;
+	I915_WRITE(DP_A, dpa_ctl);
+	udelay(200);
+}
+
+
+static void igdng_set_pll_edp (struct drm_crtc *crtc, int clock)
+{
+	struct drm_device *dev = crtc->dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	u32 dpa_ctl;
+
+	DRM_DEBUG("eDP PLL enable for clock %d\n", clock);
+	dpa_ctl = I915_READ(DP_A);
+	dpa_ctl &= ~DP_PLL_FREQ_MASK;
+
+	if (clock < 200000) {
+		u32 temp;
+		dpa_ctl |= DP_PLL_FREQ_160MHZ;
+		/* workaround for 160Mhz:
+		   1) program 0x4600c bits 15:0 = 0x8124
+		   2) program 0x46010 bit 0 = 1
+		   3) program 0x46034 bit 24 = 1
+		   4) program 0x64000 bit 14 = 1
+		   */
+		temp = I915_READ(0x4600c);
+		temp &= 0xffff0000;
+		I915_WRITE(0x4600c, temp | 0x8124);
+
+		temp = I915_READ(0x46010);
+		I915_WRITE(0x46010, temp | 1);
+
+		temp = I915_READ(0x46034);
+		I915_WRITE(0x46034, temp | (1 << 24));
+	} else {
+		dpa_ctl |= DP_PLL_FREQ_270MHZ;
+	}
+	I915_WRITE(DP_A, dpa_ctl);
+
+	udelay(500);
+}
+
 static void igdng_crtc_dpms(struct drm_crtc *crtc, int mode)
 {
 	struct drm_device *dev = crtc->dev;
@@ -1015,6 +1153,7 @@ static void igdng_crtc_dpms(struct drm_crtc *crtc, int mode)
 	int fdi_rx_imr_reg = (pipe == 0) ? FDI_RXA_IMR : FDI_RXB_IMR;
 	int transconf_reg = (pipe == 0) ? TRANSACONF : TRANSBCONF;
 	int pf_ctl_reg = (pipe == 0) ? PFA_CTL_1 : PFB_CTL_1;
+	int pf_win_size = (pipe == 0) ? PFA_WIN_SZ : PFB_WIN_SZ;
 	int cpu_htot_reg = (pipe == 0) ? HTOTAL_A : HTOTAL_B;
 	int cpu_hblank_reg = (pipe == 0) ? HBLANK_A : HBLANK_B;
 	int cpu_hsync_reg = (pipe == 0) ? HSYNC_A : HSYNC_B;
@@ -1028,7 +1167,7 @@ static void igdng_crtc_dpms(struct drm_crtc *crtc, int mode)
 	int trans_vblank_reg = (pipe == 0) ? TRANS_VBLANK_A : TRANS_VBLANK_B;
 	int trans_vsync_reg = (pipe == 0) ? TRANS_VSYNC_A : TRANS_VSYNC_B;
 	u32 temp;
-	int tries = 5, j;
+	int tries = 5, j, n;
 
 	/* XXX: When our outputs are all unaware of DPMS modes other than off
 	 * and on, we should map those modes to DRM_MODE_DPMS_OFF in the CRTC.
@@ -1038,27 +1177,32 @@ static void igdng_crtc_dpms(struct drm_crtc *crtc, int mode)
 	case DRM_MODE_DPMS_STANDBY:
 	case DRM_MODE_DPMS_SUSPEND:
 		DRM_DEBUG("crtc %d dpms on\n", pipe);
-		/* enable PCH DPLL */
-		temp = I915_READ(pch_dpll_reg);
-		if ((temp & DPLL_VCO_ENABLE) == 0) {
-			I915_WRITE(pch_dpll_reg, temp | DPLL_VCO_ENABLE);
-			I915_READ(pch_dpll_reg);
-		}
-
-		/* enable PCH FDI RX PLL, wait warmup plus DMI latency */
-		temp = I915_READ(fdi_rx_reg);
-		I915_WRITE(fdi_rx_reg, temp | FDI_RX_PLL_ENABLE |
-				FDI_SEL_PCDCLK |
-				FDI_DP_PORT_WIDTH_X4); /* default 4 lanes */
-		I915_READ(fdi_rx_reg);
-		udelay(200);
+		if (HAS_eDP) {
+			/* enable eDP PLL */
+			igdng_enable_pll_edp(crtc);
+		} else {
+			/* enable PCH DPLL */
+			temp = I915_READ(pch_dpll_reg);
+			if ((temp & DPLL_VCO_ENABLE) == 0) {
+				I915_WRITE(pch_dpll_reg, temp | DPLL_VCO_ENABLE);
+				I915_READ(pch_dpll_reg);
+			}
 
-		/* Enable CPU FDI TX PLL, always on for IGDNG */
-		temp = I915_READ(fdi_tx_reg);
-		if ((temp & FDI_TX_PLL_ENABLE) == 0) {
-			I915_WRITE(fdi_tx_reg, temp | FDI_TX_PLL_ENABLE);
-			I915_READ(fdi_tx_reg);
-			udelay(100);
+			/* enable PCH FDI RX PLL, wait warmup plus DMI latency */
+			temp = I915_READ(fdi_rx_reg);
+			I915_WRITE(fdi_rx_reg, temp | FDI_RX_PLL_ENABLE |
+					FDI_SEL_PCDCLK |
+					FDI_DP_PORT_WIDTH_X4); /* default 4 lanes */
+			I915_READ(fdi_rx_reg);
+			udelay(200);
+
+			/* Enable CPU FDI TX PLL, always on for IGDNG */
+			temp = I915_READ(fdi_tx_reg);
+			if ((temp & FDI_TX_PLL_ENABLE) == 0) {
+				I915_WRITE(fdi_tx_reg, temp | FDI_TX_PLL_ENABLE);
+				I915_READ(fdi_tx_reg);
+				udelay(100);
+			}
 		}
 
 		/* Enable CPU pipe */
@@ -1077,122 +1221,126 @@ static void igdng_crtc_dpms(struct drm_crtc *crtc, int mode)
 			I915_WRITE(dspbase_reg, I915_READ(dspbase_reg));
 		}
 
-		/* enable CPU FDI TX and PCH FDI RX */
-		temp = I915_READ(fdi_tx_reg);
-		temp |= FDI_TX_ENABLE;
-		temp |= FDI_DP_PORT_WIDTH_X4; /* default */
-		temp &= ~FDI_LINK_TRAIN_NONE;
-		temp |= FDI_LINK_TRAIN_PATTERN_1;
-		I915_WRITE(fdi_tx_reg, temp);
-		I915_READ(fdi_tx_reg);
+		if (!HAS_eDP) {
+			/* enable CPU FDI TX and PCH FDI RX */
+			temp = I915_READ(fdi_tx_reg);
+			temp |= FDI_TX_ENABLE;
+			temp |= FDI_DP_PORT_WIDTH_X4; /* default */
+			temp &= ~FDI_LINK_TRAIN_NONE;
+			temp |= FDI_LINK_TRAIN_PATTERN_1;
+			I915_WRITE(fdi_tx_reg, temp);
+			I915_READ(fdi_tx_reg);
 
-		temp = I915_READ(fdi_rx_reg);
-		temp &= ~FDI_LINK_TRAIN_NONE;
-		temp |= FDI_LINK_TRAIN_PATTERN_1;
-		I915_WRITE(fdi_rx_reg, temp | FDI_RX_ENABLE);
-		I915_READ(fdi_rx_reg);
+			temp = I915_READ(fdi_rx_reg);
+			temp &= ~FDI_LINK_TRAIN_NONE;
+			temp |= FDI_LINK_TRAIN_PATTERN_1;
+			I915_WRITE(fdi_rx_reg, temp | FDI_RX_ENABLE);
+			I915_READ(fdi_rx_reg);
 
-		udelay(150);
+			udelay(150);
 
-		/* Train FDI. */
-		/* umask FDI RX Interrupt symbol_lock and bit_lock bit
-		   for train result */
-		temp = I915_READ(fdi_rx_imr_reg);
-		temp &= ~FDI_RX_SYMBOL_LOCK;
-		temp &= ~FDI_RX_BIT_LOCK;
-		I915_WRITE(fdi_rx_imr_reg, temp);
-		I915_READ(fdi_rx_imr_reg);
-		udelay(150);
+			/* Train FDI. */
+			/* umask FDI RX Interrupt symbol_lock and bit_lock bit
+			   for train result */
+			temp = I915_READ(fdi_rx_imr_reg);
+			temp &= ~FDI_RX_SYMBOL_LOCK;
+			temp &= ~FDI_RX_BIT_LOCK;
+			I915_WRITE(fdi_rx_imr_reg, temp);
+			I915_READ(fdi_rx_imr_reg);
+			udelay(150);
 
-		temp = I915_READ(fdi_rx_iir_reg);
-		DRM_DEBUG("FDI_RX_IIR 0x%x\n", temp);
+			temp = I915_READ(fdi_rx_iir_reg);
+			DRM_DEBUG("FDI_RX_IIR 0x%x\n", temp);
 
-		if ((temp & FDI_RX_BIT_LOCK) == 0) {
-			for (j = 0; j < tries; j++) {
-				temp = I915_READ(fdi_rx_iir_reg);
-				DRM_DEBUG("FDI_RX_IIR 0x%x\n", temp);
-				if (temp & FDI_RX_BIT_LOCK)
-					break;
-				udelay(200);
-			}
-			if (j != tries)
+			if ((temp & FDI_RX_BIT_LOCK) == 0) {
+				for (j = 0; j < tries; j++) {
+					temp = I915_READ(fdi_rx_iir_reg);
+					DRM_DEBUG("FDI_RX_IIR 0x%x\n", temp);
+					if (temp & FDI_RX_BIT_LOCK)
+						break;
+					udelay(200);
+				}
+				if (j != tries)
+					I915_WRITE(fdi_rx_iir_reg,
+							temp | FDI_RX_BIT_LOCK);
+				else
+					DRM_DEBUG("train 1 fail\n");
+			} else {
 				I915_WRITE(fdi_rx_iir_reg,
 						temp | FDI_RX_BIT_LOCK);
-			else
-				DRM_DEBUG("train 1 fail\n");
-		} else {
-			I915_WRITE(fdi_rx_iir_reg,
-					temp | FDI_RX_BIT_LOCK);
-			DRM_DEBUG("train 1 ok 2!\n");
-		}
-		temp = I915_READ(fdi_tx_reg);
-		temp &= ~FDI_LINK_TRAIN_NONE;
-		temp |= FDI_LINK_TRAIN_PATTERN_2;
-		I915_WRITE(fdi_tx_reg, temp);
-
-		temp = I915_READ(fdi_rx_reg);
-		temp &= ~FDI_LINK_TRAIN_NONE;
-		temp |= FDI_LINK_TRAIN_PATTERN_2;
-		I915_WRITE(fdi_rx_reg, temp);
+				DRM_DEBUG("train 1 ok 2!\n");
+			}
+			temp = I915_READ(fdi_tx_reg);
+			temp &= ~FDI_LINK_TRAIN_NONE;
+			temp |= FDI_LINK_TRAIN_PATTERN_2;
+			I915_WRITE(fdi_tx_reg, temp);
+
+			temp = I915_READ(fdi_rx_reg);
+			temp &= ~FDI_LINK_TRAIN_NONE;
+			temp |= FDI_LINK_TRAIN_PATTERN_2;
+			I915_WRITE(fdi_rx_reg, temp);
 
-		udelay(150);
+			udelay(150);
 
-		temp = I915_READ(fdi_rx_iir_reg);
-		DRM_DEBUG("FDI_RX_IIR 0x%x\n", temp);
+			temp = I915_READ(fdi_rx_iir_reg);
+			DRM_DEBUG("FDI_RX_IIR 0x%x\n", temp);
 
-		if ((temp & FDI_RX_SYMBOL_LOCK) == 0) {
-			for (j = 0; j < tries; j++) {
-				temp = I915_READ(fdi_rx_iir_reg);
-				DRM_DEBUG("FDI_RX_IIR 0x%x\n", temp);
-				if (temp & FDI_RX_SYMBOL_LOCK)
-					break;
-				udelay(200);
-			}
-			if (j != tries) {
+			if ((temp & FDI_RX_SYMBOL_LOCK) == 0) {
+				for (j = 0; j < tries; j++) {
+					temp = I915_READ(fdi_rx_iir_reg);
+					DRM_DEBUG("FDI_RX_IIR 0x%x\n", temp);
+					if (temp & FDI_RX_SYMBOL_LOCK)
+						break;
+					udelay(200);
+				}
+				if (j != tries) {
+					I915_WRITE(fdi_rx_iir_reg,
+							temp | FDI_RX_SYMBOL_LOCK);
+					DRM_DEBUG("train 2 ok 1!\n");
+				} else
+					DRM_DEBUG("train 2 fail\n");
+			} else {
 				I915_WRITE(fdi_rx_iir_reg,
 						temp | FDI_RX_SYMBOL_LOCK);
-				DRM_DEBUG("train 2 ok 1!\n");
-			} else
-				DRM_DEBUG("train 2 fail\n");
-		} else {
-			I915_WRITE(fdi_rx_iir_reg, temp | FDI_RX_SYMBOL_LOCK);
-			DRM_DEBUG("train 2 ok 2!\n");
-		}
-		DRM_DEBUG("train done\n");
+				DRM_DEBUG("train 2 ok 2!\n");
+			}
+			DRM_DEBUG("train done\n");
 
-		/* set transcoder timing */
-		I915_WRITE(trans_htot_reg, I915_READ(cpu_htot_reg));
-		I915_WRITE(trans_hblank_reg, I915_READ(cpu_hblank_reg));
-		I915_WRITE(trans_hsync_reg, I915_READ(cpu_hsync_reg));
+			/* set transcoder timing */
+			I915_WRITE(trans_htot_reg, I915_READ(cpu_htot_reg));
+			I915_WRITE(trans_hblank_reg, I915_READ(cpu_hblank_reg));
+			I915_WRITE(trans_hsync_reg, I915_READ(cpu_hsync_reg));
 
-		I915_WRITE(trans_vtot_reg, I915_READ(cpu_vtot_reg));
-		I915_WRITE(trans_vblank_reg, I915_READ(cpu_vblank_reg));
-		I915_WRITE(trans_vsync_reg, I915_READ(cpu_vsync_reg));
+			I915_WRITE(trans_vtot_reg, I915_READ(cpu_vtot_reg));
+			I915_WRITE(trans_vblank_reg, I915_READ(cpu_vblank_reg));
+			I915_WRITE(trans_vsync_reg, I915_READ(cpu_vsync_reg));
 
-		/* enable PCH transcoder */
-		temp = I915_READ(transconf_reg);
-		I915_WRITE(transconf_reg, temp | TRANS_ENABLE);
-		I915_READ(transconf_reg);
+			/* enable PCH transcoder */
+			temp = I915_READ(transconf_reg);
+			I915_WRITE(transconf_reg, temp | TRANS_ENABLE);
+			I915_READ(transconf_reg);
 
-		while ((I915_READ(transconf_reg) & TRANS_STATE_ENABLE) == 0)
-			;
+			while ((I915_READ(transconf_reg) & TRANS_STATE_ENABLE) == 0)
+				;
 
-		/* enable normal */
+			/* enable normal */
 
-		temp = I915_READ(fdi_tx_reg);
-		temp &= ~FDI_LINK_TRAIN_NONE;
-		I915_WRITE(fdi_tx_reg, temp | FDI_LINK_TRAIN_NONE |
-				FDI_TX_ENHANCE_FRAME_ENABLE);
-		I915_READ(fdi_tx_reg);
+			temp = I915_READ(fdi_tx_reg);
+			temp &= ~FDI_LINK_TRAIN_NONE;
+			I915_WRITE(fdi_tx_reg, temp | FDI_LINK_TRAIN_NONE |
+					FDI_TX_ENHANCE_FRAME_ENABLE);
+			I915_READ(fdi_tx_reg);
 
-		temp = I915_READ(fdi_rx_reg);
-		temp &= ~FDI_LINK_TRAIN_NONE;
-		I915_WRITE(fdi_rx_reg, temp | FDI_LINK_TRAIN_NONE |
-				FDI_RX_ENHANCE_FRAME_ENABLE);
-		I915_READ(fdi_rx_reg);
+			temp = I915_READ(fdi_rx_reg);
+			temp &= ~FDI_LINK_TRAIN_NONE;
+			I915_WRITE(fdi_rx_reg, temp | FDI_LINK_TRAIN_NONE |
+					FDI_RX_ENHANCE_FRAME_ENABLE);
+			I915_READ(fdi_rx_reg);
 
-		/* wait one idle pattern time */
-		udelay(100);
+			/* wait one idle pattern time */
+			udelay(100);
+
+		}
 
 		intel_crtc_load_lut(crtc);
 
@@ -1200,8 +1348,7 @@ static void igdng_crtc_dpms(struct drm_crtc *crtc, int mode)
 	case DRM_MODE_DPMS_OFF:
 		DRM_DEBUG("crtc %d dpms off\n", pipe);
 
-		/* Disable the VGA plane that we never use */
-		I915_WRITE(CPU_VGACNTRL, VGA_DISP_DISABLE);
+		i915_disable_vga(dev);
 
 		/* Disable display plane */
 		temp = I915_READ(dspcntr_reg);
@@ -1217,17 +1364,23 @@ static void igdng_crtc_dpms(struct drm_crtc *crtc, int mode)
 		if ((temp & PIPEACONF_ENABLE) != 0) {
 			I915_WRITE(pipeconf_reg, temp & ~PIPEACONF_ENABLE);
 			I915_READ(pipeconf_reg);
+			n = 0;
 			/* wait for cpu pipe off, pipe state */
-			while ((I915_READ(pipeconf_reg) & I965_PIPECONF_ACTIVE) != 0)
-				;
+			while ((I915_READ(pipeconf_reg) & I965_PIPECONF_ACTIVE) != 0) {
+				n++;
+				if (n < 60) {
+					udelay(500);
+					continue;
+				} else {
+					DRM_DEBUG("pipe %d off delay\n", pipe);
+					break;
+				}
+			}
 		} else
 			DRM_DEBUG("crtc %d is disabled\n", pipe);
 
-		/* IGDNG-A : disable cpu panel fitter ? */
-		temp = I915_READ(pf_ctl_reg);
-		if ((temp & PF_ENABLE) != 0) {
-			I915_WRITE(pf_ctl_reg, temp & ~PF_ENABLE);
-			I915_READ(pf_ctl_reg);
+		if (HAS_eDP) {
+			igdng_disable_pll_edp(crtc);
 		}
 
 		/* disable CPU FDI tx and PCH FDI rx */
@@ -1239,6 +1392,8 @@ static void igdng_crtc_dpms(struct drm_crtc *crtc, int mode)
 		I915_WRITE(fdi_rx_reg, temp & ~FDI_RX_ENABLE);
 		I915_READ(fdi_rx_reg);
 
+		udelay(100);
+
 		/* still set train pattern 1 */
 		temp = I915_READ(fdi_tx_reg);
 		temp &= ~FDI_LINK_TRAIN_NONE;
@@ -1250,14 +1405,25 @@ static void igdng_crtc_dpms(struct drm_crtc *crtc, int mode)
 		temp |= FDI_LINK_TRAIN_PATTERN_1;
 		I915_WRITE(fdi_rx_reg, temp);
 
+		udelay(100);
+
 		/* disable PCH transcoder */
 		temp = I915_READ(transconf_reg);
 		if ((temp & TRANS_ENABLE) != 0) {
 			I915_WRITE(transconf_reg, temp & ~TRANS_ENABLE);
 			I915_READ(transconf_reg);
+			n = 0;
 			/* wait for PCH transcoder off, transcoder state */
-			while ((I915_READ(transconf_reg) & TRANS_STATE_ENABLE) != 0)
-				;
+			while ((I915_READ(transconf_reg) & TRANS_STATE_ENABLE) != 0) {
+				n++;
+				if (n < 60) {
+					udelay(500);
+					continue;
+				} else {
+					DRM_DEBUG("transcoder %d off delay\n", pipe);
+					break;
+				}
+			}
 		}
 
 		/* disable PCH DPLL */
@@ -1275,6 +1441,22 @@ static void igdng_crtc_dpms(struct drm_crtc *crtc, int mode)
 			I915_READ(fdi_rx_reg);
 		}
 
+		/* Disable CPU FDI TX PLL */
+		temp = I915_READ(fdi_tx_reg);
+		if ((temp & FDI_TX_PLL_ENABLE) != 0) {
+			I915_WRITE(fdi_tx_reg, temp & ~FDI_TX_PLL_ENABLE);
+			I915_READ(fdi_tx_reg);
+			udelay(100);
+		}
+
+		/* Disable PF */
+		temp = I915_READ(pf_ctl_reg);
+		if ((temp & PF_ENABLE) != 0) {
+			I915_WRITE(pf_ctl_reg, temp & ~PF_ENABLE);
+			I915_READ(pf_ctl_reg);
+		}
+		I915_WRITE(pf_win_size, 0);
+
 		/* Wait for the clocks to turn off. */
 		udelay(150);
 		break;
@@ -1342,7 +1524,7 @@ static void i9xx_crtc_dpms(struct drm_crtc *crtc, int mode)
 		//intel_crtc_dpms_video(crtc, FALSE); TODO
 
 		/* Disable the VGA plane that we never use */
-		I915_WRITE(VGACNTRL, VGA_DISP_DISABLE);
+		i915_disable_vga(dev);
 
 		/* Disable display plane */
 		temp = I915_READ(dspcntr_reg);
@@ -1623,48 +1805,72 @@ static struct intel_watermark_params igd_cursor_hplloff_wm = {
 	IGD_FIFO_LINE_SIZE
 };
 static struct intel_watermark_params i945_wm_info = {
-	I915_FIFO_LINE_SIZE,
+	I945_FIFO_SIZE,
 	I915_MAX_WM,
 	1,
-	0,
-	IGD_FIFO_LINE_SIZE
+	2,
+	I915_FIFO_LINE_SIZE
 };
 static struct intel_watermark_params i915_wm_info = {
-	I945_FIFO_SIZE,
+	I915_FIFO_SIZE,
 	I915_MAX_WM,
 	1,
-	0,
+	2,
 	I915_FIFO_LINE_SIZE
 };
 static struct intel_watermark_params i855_wm_info = {
 	I855GM_FIFO_SIZE,
 	I915_MAX_WM,
 	1,
-	0,
+	2,
 	I830_FIFO_LINE_SIZE
 };
 static struct intel_watermark_params i830_wm_info = {
 	I830_FIFO_SIZE,
 	I915_MAX_WM,
 	1,
-	0,
+	2,
 	I830_FIFO_LINE_SIZE
 };
 
+/**
+ * intel_calculate_wm - calculate watermark level
+ * @clock_in_khz: pixel clock
+ * @wm: chip FIFO params
+ * @pixel_size: display pixel size
+ * @latency_ns: memory latency for the platform
+ *
+ * Calculate the watermark level (the level at which the display plane will
+ * start fetching from memory again).  Each chip has a different display
+ * FIFO size and allocation, so the caller needs to figure that out and pass
+ * in the correct intel_watermark_params structure.
+ *
+ * As the pixel clock runs, the FIFO will be drained at a rate that depends
+ * on the pixel size.  When it reaches the watermark level, it'll start
+ * fetching FIFO line sized based chunks from memory until the FIFO fills
+ * past the watermark point.  If the FIFO drains completely, a FIFO underrun
+ * will occur, and a display engine hang could result.
+ */
 static unsigned long intel_calculate_wm(unsigned long clock_in_khz,
 					struct intel_watermark_params *wm,
 					int pixel_size,
 					unsigned long latency_ns)
 {
-	unsigned long bytes_required, wm_size;
+	long entries_required, wm_size;
+
+	entries_required = (clock_in_khz * pixel_size * latency_ns) / 1000000;
+	entries_required /= wm->cacheline_size;
 
-	bytes_required = (clock_in_khz * pixel_size * latency_ns) / 1000000;
-	bytes_required /= wm->cacheline_size;
-	wm_size = wm->fifo_size - bytes_required - wm->guard_size;
+	DRM_DEBUG("FIFO entries required for mode: %d\n", entries_required);
 
-	if (wm_size > wm->max_wm)
+	wm_size = wm->fifo_size - (entries_required + wm->guard_size);
+
+	DRM_DEBUG("FIFO watermark level: %d\n", wm_size);
+
+	/* Don't promote wm_size to unsigned... */
+	if (wm_size > (long)wm->max_wm)
 		wm_size = wm->max_wm;
-	if (wm_size == 0)
+	if (wm_size <= 0)
 		wm_size = wm->default_wm;
 	return wm_size;
 }
@@ -1799,8 +2005,40 @@ static void igd_enable_cxsr(struct drm_device *dev, unsigned long clock,
 	return;
 }
 
-const static int latency_ns = 5000; /* default for non-igd platforms */
+const static int latency_ns = 3000; /* default for non-igd platforms */
 
+static int intel_get_fifo_size(struct drm_device *dev, int plane)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	uint32_t dsparb = I915_READ(DSPARB);
+	int size;
+
+	if (IS_I9XX(dev)) {
+		if (plane == 0)
+			size = dsparb & 0x7f;
+		else
+			size = ((dsparb >> DSPARB_CSTART_SHIFT) & 0x7f) -
+				(dsparb & 0x7f);
+	} else if (IS_I85X(dev)) {
+		if (plane == 0)
+			size = dsparb & 0x1ff;
+		else
+			size = ((dsparb >> DSPARB_BEND_SHIFT) & 0x1ff) -
+				(dsparb & 0x1ff);
+		size >>= 1; /* Convert to cachelines */
+	} else if (IS_845G(dev)) {
+		size = dsparb & 0x7f;
+		size >>= 2; /* Convert to cachelines */
+	} else {
+		size = dsparb & 0x7f;
+		size >>= 1; /* Convert to cachelines */
+	}
+
+	DRM_DEBUG("FIFO size - (0x%08x) %s: %d\n", dsparb, plane ? "B" : "A",
+		  size);
+
+	return size;
+}
 
 static void i965_update_wm(struct drm_device *dev)
 {
@@ -1817,101 +2055,89 @@ static void i9xx_update_wm(struct drm_device *dev, int planea_clock,
 			   int planeb_clock, int sr_hdisplay, int pixel_size)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	uint32_t fwater_lo = I915_READ(FW_BLC) & MM_FIFO_WATERMARK;
-	uint32_t fwater_hi = I915_READ(FW_BLC2) & LM_FIFO_WATERMARK;
-	int bsize, asize, cwm, bwm = 1, awm = 1, srwm = 1;
-	uint32_t dsparb = I915_READ(DSPARB);
-	int planea_entries, planeb_entries;
-	struct intel_watermark_params *wm_params;
+	uint32_t fwater_lo;
+	uint32_t fwater_hi;
+	int total_size, cacheline_size, cwm, srwm = 1;
+	int planea_wm, planeb_wm;
+	struct intel_watermark_params planea_params, planeb_params;
 	unsigned long line_time_us;
 	int sr_clock, sr_entries = 0;
 
+	/* Create copies of the base settings for each pipe */
 	if (IS_I965GM(dev) || IS_I945GM(dev))
-		wm_params = &i945_wm_info;
+		planea_params = planeb_params = i945_wm_info;
 	else if (IS_I9XX(dev))
-		wm_params = &i915_wm_info;
+		planea_params = planeb_params = i915_wm_info;
 	else
-		wm_params = &i855_wm_info;
-
-	planea_entries = intel_calculate_wm(planea_clock, wm_params,
-					    pixel_size, latency_ns);
-	planeb_entries = intel_calculate_wm(planeb_clock, wm_params,
-					    pixel_size, latency_ns);
-
-	DRM_DEBUG("FIFO entries - A: %d, B: %d\n", planea_entries,
-		  planeb_entries);
+		planea_params = planeb_params = i855_wm_info;
 
-	if (IS_I9XX(dev)) {
-		asize = dsparb & 0x7f;
-		bsize = (dsparb >> DSPARB_CSTART_SHIFT) & 0x7f;
-	} else {
-		asize = dsparb & 0x1ff;
-		bsize = (dsparb >> DSPARB_BEND_SHIFT) & 0x1ff;
-	}
-	DRM_DEBUG("FIFO size - A: %d, B: %d\n", asize, bsize);
+	/* Grab a couple of global values before we overwrite them */
+	total_size = planea_params.fifo_size;
+	cacheline_size = planea_params.cacheline_size;
 
-	/* Two extra entries for padding */
-	awm = asize - (planea_entries + 2);
-	bwm = bsize - (planeb_entries + 2);
+	/* Update per-plane FIFO sizes */
+	planea_params.fifo_size = intel_get_fifo_size(dev, 0);
+	planeb_params.fifo_size = intel_get_fifo_size(dev, 1);
 
-	/* Sanity check against potentially bad FIFO allocations */
-	if (awm <= 0) {
-		/* pipe is on but has too few FIFO entries */
-		if (planea_entries != 0)
-			DRM_DEBUG("plane A needs more FIFO entries\n");
-		awm = 1;
-	}
-	if (bwm <= 0) {
-		if (planeb_entries != 0)
-			DRM_DEBUG("plane B needs more FIFO entries\n");
-		bwm = 1;
-	}
+	planea_wm = intel_calculate_wm(planea_clock, &planea_params,
+				       pixel_size, latency_ns);
+	planeb_wm = intel_calculate_wm(planeb_clock, &planeb_params,
+				       pixel_size, latency_ns);
+	DRM_DEBUG("FIFO watermarks - A: %d, B: %d\n", planea_wm, planeb_wm);
 
 	/*
 	 * Overlay gets an aggressive default since video jitter is bad.
 	 */
 	cwm = 2;
 
-	/* Calc sr entries for one pipe configs */
-	if (!planea_clock || !planeb_clock) {
+	/* Calc sr entries for one plane configs */
+	if (sr_hdisplay && (!planea_clock || !planeb_clock)) {
+		/* self-refresh has much higher latency */
+		const static int sr_latency_ns = 6000;
+
 		sr_clock = planea_clock ? planea_clock : planeb_clock;
-		line_time_us = (sr_hdisplay * 1000) / sr_clock;
-		sr_entries = (((latency_ns / line_time_us) + 1) * pixel_size *
-			      sr_hdisplay) / 1000;
-		sr_entries = roundup(sr_entries / wm_params->cacheline_size, 1);
-		if (sr_entries < wm_params->fifo_size)
-			srwm = wm_params->fifo_size - sr_entries;
+		line_time_us = ((sr_hdisplay * 1000) / sr_clock);
+
+		/* Use ns/us then divide to preserve precision */
+		sr_entries = (((sr_latency_ns / line_time_us) + 1) *
+			      pixel_size * sr_hdisplay) / 1000;
+		sr_entries = roundup(sr_entries / cacheline_size, 1);
+		DRM_DEBUG("self-refresh entries: %d\n", sr_entries);
+		srwm = total_size - sr_entries;
+		if (srwm < 0)
+			srwm = 1;
+		if (IS_I9XX(dev))
+			I915_WRITE(FW_BLC_SELF, (srwm & 0x3f));
 	}
 
 	DRM_DEBUG("Setting FIFO watermarks - A: %d, B: %d, C: %d, SR %d\n",
-		  awm, bwm, cwm, srwm);
+		  planea_wm, planeb_wm, cwm, srwm);
 
-	fwater_lo = fwater_lo | ((bwm & 0x3f) << 16) | (awm & 0x3f);
-	fwater_hi = fwater_hi | (cwm & 0x1f);
+	fwater_lo = ((planeb_wm & 0x3f) << 16) | (planea_wm & 0x3f);
+	fwater_hi = (cwm & 0x1f);
+
+	/* Set request length to 8 cachelines per fetch */
+	fwater_lo = fwater_lo | (1 << 24) | (1 << 8);
+	fwater_hi = fwater_hi | (1 << 8);
 
 	I915_WRITE(FW_BLC, fwater_lo);
 	I915_WRITE(FW_BLC2, fwater_hi);
-	if (IS_I9XX(dev))
-		I915_WRITE(FW_BLC_SELF, FW_BLC_SELF_EN | (srwm & 0x3f));
 }
 
 static void i830_update_wm(struct drm_device *dev, int planea_clock,
 			   int pixel_size)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	uint32_t dsparb = I915_READ(DSPARB);
-	uint32_t fwater_lo = I915_READ(FW_BLC) & MM_FIFO_WATERMARK;
-	unsigned int asize, awm;
-	int planea_entries;
-
-	planea_entries = intel_calculate_wm(planea_clock, &i830_wm_info,
-					    pixel_size, latency_ns);
+	uint32_t fwater_lo = I915_READ(FW_BLC) & ~0xfff;
+	int planea_wm;
 
-	asize = dsparb & 0x7f;
+	i830_wm_info.fifo_size = intel_get_fifo_size(dev, 0);
 
-	awm = asize - planea_entries;
+	planea_wm = intel_calculate_wm(planea_clock, &i830_wm_info,
+				       pixel_size, latency_ns);
+	fwater_lo |= (3<<8) | planea_wm;
 
-	fwater_lo = fwater_lo | awm;
+	DRM_DEBUG("Setting FIFO watermarks - A: %d\n", planea_wm);
 
 	I915_WRITE(FW_BLC, fwater_lo);
 }
@@ -1984,7 +2210,7 @@ static void intel_update_watermarks(struct drm_device *dev)
 	if (enabled <= 0)
 		return;
 
-	/* Single pipe configs can enable self refresh */
+	/* Single plane configs can enable self refresh */
 	if (enabled == 1 && IS_IGD(dev))
 		igd_enable_cxsr(dev, sr_clock, pixel_size);
 	else if (IS_IGD(dev))
@@ -2028,6 +2254,7 @@ static int intel_crtc_mode_set(struct drm_crtc *crtc,
 	u32 dpll = 0, fp = 0, dspcntr, pipeconf;
 	bool ok, is_sdvo = false, is_dvo = false;
 	bool is_crt = false, is_lvds = false, is_tv = false, is_dp = false;
+	bool is_edp = false;
 	struct drm_mode_config *mode_config = &dev->mode_config;
 	struct drm_connector *connector;
 	const intel_limit_t *limit;
@@ -2043,6 +2270,7 @@ static int intel_crtc_mode_set(struct drm_crtc *crtc,
 	int lvds_reg = LVDS;
 	u32 temp;
 	int sdvo_pixel_multiply;
+	int target_clock;
 
 	drm_vblank_pre_modeset(dev, pipe);
 
@@ -2074,6 +2302,9 @@ static int intel_crtc_mode_set(struct drm_crtc *crtc,
 		case INTEL_OUTPUT_DISPLAYPORT:
 			is_dp = true;
 			break;
+		case INTEL_OUTPUT_EDP:
+			is_edp = true;
+			break;
 		}
 
 		num_outputs++;
@@ -2125,11 +2356,29 @@ static int intel_crtc_mode_set(struct drm_crtc *crtc,
 	}
 
 	/* FDI link */
-	if (IS_IGDNG(dev))
-		igdng_compute_m_n(3, 4, /* lane num 4 */
-				adjusted_mode->clock,
-				270000, /* lane clock */
-				&m_n);
+	if (IS_IGDNG(dev)) {
+		int lane, link_bw;
+		/* eDP doesn't require FDI link, so just set DP M/N
+		   according to current link config */
+		if (is_edp) {
+			struct drm_connector *edp;
+			target_clock = mode->clock;
+			edp = intel_pipe_get_output(crtc);
+			intel_edp_link_config(to_intel_output(edp),
+					&lane, &link_bw);
+		} else {
+			/* DP over FDI requires target mode clock
+			   instead of link clock */
+			if (is_dp)
+				target_clock = mode->clock;
+			else
+				target_clock = adjusted_mode->clock;
+			lane = 4;
+			link_bw = 270000;
+		}
+		igdng_compute_m_n(3, lane, target_clock,
+				  link_bw, &m_n);
+	}
 
 	if (IS_IGD(dev))
 		fp = (1 << clock.n) << 16 | clock.m1 << 8 | clock.m2;
@@ -2250,29 +2499,15 @@ static int intel_crtc_mode_set(struct drm_crtc *crtc,
 		dpll_reg = pch_dpll_reg;
 	}
 
-	if (dpll & DPLL_VCO_ENABLE) {
+	if (is_edp) {
+		igdng_disable_pll_edp(crtc);
+	} else if ((dpll & DPLL_VCO_ENABLE)) {
 		I915_WRITE(fp_reg, fp);
 		I915_WRITE(dpll_reg, dpll & ~DPLL_VCO_ENABLE);
 		I915_READ(dpll_reg);
 		udelay(150);
 	}
 
-	if (IS_IGDNG(dev)) {
-		/* enable PCH clock reference source */
-		/* XXX need to change the setting for other outputs */
-		u32 temp;
-		temp = I915_READ(PCH_DREF_CONTROL);
-		temp &= ~DREF_NONSPREAD_SOURCE_MASK;
-		temp |= DREF_NONSPREAD_CK505_ENABLE;
-		temp &= ~DREF_SSC_SOURCE_MASK;
-		temp |= DREF_SSC_SOURCE_ENABLE;
-		temp &= ~DREF_SSC1_ENABLE;
-		/* if no eDP, disable source output to CPU */
-		temp &= ~DREF_CPU_SOURCE_OUTPUT_MASK;
-		temp |= DREF_CPU_SOURCE_OUTPUT_DISABLE;
-		I915_WRITE(PCH_DREF_CONTROL, temp);
-	}
-
 	/* The LVDS pin pair needs to be on before the DPLLs are enabled.
 	 * This is an exception to the general rule that mode_set doesn't turn
 	 * things on.
@@ -2304,23 +2539,25 @@ static int intel_crtc_mode_set(struct drm_crtc *crtc,
 	if (is_dp)
 		intel_dp_set_m_n(crtc, mode, adjusted_mode);
 
-	I915_WRITE(fp_reg, fp);
-	I915_WRITE(dpll_reg, dpll);
-	I915_READ(dpll_reg);
-	/* Wait for the clocks to stabilize. */
-	udelay(150);
-
-	if (IS_I965G(dev) && !IS_IGDNG(dev)) {
-		sdvo_pixel_multiply = adjusted_mode->clock / mode->clock;
-		I915_WRITE(dpll_md_reg, (0 << DPLL_MD_UDI_DIVIDER_SHIFT) |
-			   ((sdvo_pixel_multiply - 1) << DPLL_MD_UDI_MULTIPLIER_SHIFT));
-	} else {
-		/* write it again -- the BIOS does, after all */
+	if (!is_edp) {
+		I915_WRITE(fp_reg, fp);
 		I915_WRITE(dpll_reg, dpll);
+		I915_READ(dpll_reg);
+		/* Wait for the clocks to stabilize. */
+		udelay(150);
+
+		if (IS_I965G(dev) && !IS_IGDNG(dev)) {
+			sdvo_pixel_multiply = adjusted_mode->clock / mode->clock;
+			I915_WRITE(dpll_md_reg, (0 << DPLL_MD_UDI_DIVIDER_SHIFT) |
+					((sdvo_pixel_multiply - 1) << DPLL_MD_UDI_MULTIPLIER_SHIFT));
+		} else {
+			/* write it again -- the BIOS does, after all */
+			I915_WRITE(dpll_reg, dpll);
+		}
+		I915_READ(dpll_reg);
+		/* Wait for the clocks to stabilize. */
+		udelay(150);
 	}
-	I915_READ(dpll_reg);
-	/* Wait for the clocks to stabilize. */
-	udelay(150);
 
 	I915_WRITE(htot_reg, (adjusted_mode->crtc_hdisplay - 1) |
 		   ((adjusted_mode->crtc_htotal - 1) << 16));
@@ -2350,10 +2587,14 @@ static int intel_crtc_mode_set(struct drm_crtc *crtc,
 		I915_WRITE(link_m1_reg, m_n.link_m);
 		I915_WRITE(link_n1_reg, m_n.link_n);
 
-		 /* enable FDI RX PLL too */
-		temp = I915_READ(fdi_rx_reg);
-		I915_WRITE(fdi_rx_reg, temp | FDI_RX_PLL_ENABLE);
-		udelay(200);
+		if (is_edp) {
+			igdng_set_pll_edp(crtc, adjusted_mode->clock);
+		} else {
+			/* enable FDI RX PLL too */
+			temp = I915_READ(fdi_rx_reg);
+			I915_WRITE(fdi_rx_reg, temp | FDI_RX_PLL_ENABLE);
+			udelay(200);
+		}
 	}
 
 	I915_WRITE(pipeconf_reg, pipeconf);
@@ -2951,12 +3192,17 @@ static void intel_setup_outputs(struct drm_device *dev)
 	if (IS_IGDNG(dev)) {
 		int found;
 
+		if (IS_MOBILE(dev) && (I915_READ(DP_A) & DP_DETECTED))
+			intel_dp_init(dev, DP_A);
+
 		if (I915_READ(HDMIB) & PORT_DETECTED) {
 			/* check SDVOB */
 			/* found = intel_sdvo_init(dev, HDMIB); */
 			found = 0;
 			if (!found)
 				intel_hdmi_init(dev, HDMIB);
+			if (!found && (I915_READ(PCH_DP_B) & DP_DETECTED))
+				intel_dp_init(dev, PCH_DP_B);
 		}
 
 		if (I915_READ(HDMIC) & PORT_DETECTED)
@@ -2965,6 +3211,12 @@ static void intel_setup_outputs(struct drm_device *dev)
 		if (I915_READ(HDMID) & PORT_DETECTED)
 			intel_hdmi_init(dev, HDMID);
 
+		if (I915_READ(PCH_DP_C) & DP_DETECTED)
+			intel_dp_init(dev, PCH_DP_C);
+
+		if (I915_READ(PCH_DP_D) & DP_DETECTED)
+			intel_dp_init(dev, PCH_DP_D);
+
 	} else if (IS_I9XX(dev)) {
 		int found;
 		u32 reg;
@@ -3039,6 +3291,10 @@ static void intel_setup_outputs(struct drm_device *dev)
 				     (1 << 1));
 			clone_mask = (1 << INTEL_OUTPUT_DISPLAYPORT);
 			break;
+		case INTEL_OUTPUT_EDP:
+			crtc_mask = (1 << 1);
+			clone_mask = (1 << INTEL_OUTPUT_EDP);
+			break;
 		}
 		encoder->possible_crtcs = crtc_mask;
 		encoder->possible_clones = intel_connector_clones(dev, clone_mask);
@@ -3148,6 +3404,9 @@ void intel_modeset_init(struct drm_device *dev)
 	if (IS_I965G(dev)) {
 		dev->mode_config.max_width = 8192;
 		dev->mode_config.max_height = 8192;
+	} else if (IS_I9XX(dev)) {
+		dev->mode_config.max_width = 4096;
+		dev->mode_config.max_height = 4096;
 	} else {
 		dev->mode_config.max_width = 2048;
 		dev->mode_config.max_height = 2048;
diff --git a/drivers/gpu/drm/i915/intel_dp.c b/drivers/gpu/drm/i915/intel_dp.c
index 6770ae88370..a6ff15ac548 100644
--- a/drivers/gpu/drm/i915/intel_dp.c
+++ b/drivers/gpu/drm/i915/intel_dp.c
@@ -40,6 +40,8 @@
 
 #define DP_LINK_CONFIGURATION_SIZE	9
 
+#define IS_eDP(i) ((i)->type == INTEL_OUTPUT_EDP)
+
 struct intel_dp_priv {
 	uint32_t output_reg;
 	uint32_t DP;
@@ -63,6 +65,19 @@ intel_dp_link_train(struct intel_output *intel_output, uint32_t DP,
 static void
 intel_dp_link_down(struct intel_output *intel_output, uint32_t DP);
 
+void
+intel_edp_link_config (struct intel_output *intel_output,
+		int *lane_num, int *link_bw)
+{
+	struct intel_dp_priv   *dp_priv = intel_output->dev_priv;
+
+	*lane_num = dp_priv->lane_count;
+	if (dp_priv->link_bw == DP_LINK_BW_1_62)
+		*link_bw = 162000;
+	else if (dp_priv->link_bw == DP_LINK_BW_2_7)
+		*link_bw = 270000;
+}
+
 static int
 intel_dp_max_lane_count(struct intel_output *intel_output)
 {
@@ -206,7 +221,13 @@ intel_dp_aux_ch(struct intel_output *intel_output,
 	 * and would like to run at 2MHz. So, take the
 	 * hrawclk value and divide by 2 and use that
 	 */
-	aux_clock_divider = intel_hrawclk(dev) / 2;
+	if (IS_eDP(intel_output))
+		aux_clock_divider = 225; /* eDP input clock at 450Mhz */
+	else if (IS_IGDNG(dev))
+		aux_clock_divider = 62; /* IGDNG: input clock fixed at 125Mhz */
+	else
+		aux_clock_divider = intel_hrawclk(dev) / 2;
+
 	/* Must try at least 3 times according to DP spec */
 	for (try = 0; try < 5; try++) {
 		/* Load the send data into the aux channel data registers */
@@ -236,7 +257,7 @@ intel_dp_aux_ch(struct intel_output *intel_output,
 		}
 	
 		/* Clear done status and any errors */
-		I915_WRITE(ch_ctl, (ctl |
+		I915_WRITE(ch_ctl, (status |
 				DP_AUX_CH_CTL_DONE |
 				DP_AUX_CH_CTL_TIME_OUT_ERROR |
 				DP_AUX_CH_CTL_RECEIVE_ERROR));
@@ -295,7 +316,7 @@ intel_dp_aux_native_write(struct intel_output *intel_output,
 		return -1;
 	msg[0] = AUX_NATIVE_WRITE << 4;
 	msg[1] = address >> 8;
-	msg[2] = address;
+	msg[2] = address & 0xff;
 	msg[3] = send_bytes - 1;
 	memcpy(&msg[4], send, send_bytes);
 	msg_bytes = send_bytes + 4;
@@ -387,8 +408,8 @@ intel_dp_i2c_init(struct intel_output *intel_output, const char *name)
 	memset(&dp_priv->adapter, '\0', sizeof (dp_priv->adapter));
 	dp_priv->adapter.owner = THIS_MODULE;
 	dp_priv->adapter.class = I2C_CLASS_DDC;
-	strncpy (dp_priv->adapter.name, name, sizeof dp_priv->adapter.name - 1);
-	dp_priv->adapter.name[sizeof dp_priv->adapter.name - 1] = '\0';
+	strncpy (dp_priv->adapter.name, name, sizeof(dp_priv->adapter.name) - 1);
+	dp_priv->adapter.name[sizeof(dp_priv->adapter.name) - 1] = '\0';
 	dp_priv->adapter.algo_data = &dp_priv->algo;
 	dp_priv->adapter.dev.parent = &intel_output->base.kdev;
 	
@@ -493,22 +514,40 @@ intel_dp_set_m_n(struct drm_crtc *crtc, struct drm_display_mode *mode,
 	intel_dp_compute_m_n(3, lane_count,
 			     mode->clock, adjusted_mode->clock, &m_n);
 
-	if (intel_crtc->pipe == 0) {
-		I915_WRITE(PIPEA_GMCH_DATA_M,
-		       ((m_n.tu - 1) << PIPE_GMCH_DATA_M_TU_SIZE_SHIFT) |
-		       m_n.gmch_m);
-		I915_WRITE(PIPEA_GMCH_DATA_N,
-		       m_n.gmch_n);
-		I915_WRITE(PIPEA_DP_LINK_M, m_n.link_m);
-		I915_WRITE(PIPEA_DP_LINK_N, m_n.link_n);
+	if (IS_IGDNG(dev)) {
+		if (intel_crtc->pipe == 0) {
+			I915_WRITE(TRANSA_DATA_M1,
+				   ((m_n.tu - 1) << PIPE_GMCH_DATA_M_TU_SIZE_SHIFT) |
+				   m_n.gmch_m);
+			I915_WRITE(TRANSA_DATA_N1, m_n.gmch_n);
+			I915_WRITE(TRANSA_DP_LINK_M1, m_n.link_m);
+			I915_WRITE(TRANSA_DP_LINK_N1, m_n.link_n);
+		} else {
+			I915_WRITE(TRANSB_DATA_M1,
+				   ((m_n.tu - 1) << PIPE_GMCH_DATA_M_TU_SIZE_SHIFT) |
+				   m_n.gmch_m);
+			I915_WRITE(TRANSB_DATA_N1, m_n.gmch_n);
+			I915_WRITE(TRANSB_DP_LINK_M1, m_n.link_m);
+			I915_WRITE(TRANSB_DP_LINK_N1, m_n.link_n);
+		}
 	} else {
-		I915_WRITE(PIPEB_GMCH_DATA_M,
-		       ((m_n.tu - 1) << PIPE_GMCH_DATA_M_TU_SIZE_SHIFT) |
-		       m_n.gmch_m);
-		I915_WRITE(PIPEB_GMCH_DATA_N,
-		       m_n.gmch_n);
-		I915_WRITE(PIPEB_DP_LINK_M, m_n.link_m);
-		I915_WRITE(PIPEB_DP_LINK_N, m_n.link_n);
+		if (intel_crtc->pipe == 0) {
+			I915_WRITE(PIPEA_GMCH_DATA_M,
+				   ((m_n.tu - 1) << PIPE_GMCH_DATA_M_TU_SIZE_SHIFT) |
+				   m_n.gmch_m);
+			I915_WRITE(PIPEA_GMCH_DATA_N,
+				   m_n.gmch_n);
+			I915_WRITE(PIPEA_DP_LINK_M, m_n.link_m);
+			I915_WRITE(PIPEA_DP_LINK_N, m_n.link_n);
+		} else {
+			I915_WRITE(PIPEB_GMCH_DATA_M,
+				   ((m_n.tu - 1) << PIPE_GMCH_DATA_M_TU_SIZE_SHIFT) |
+				   m_n.gmch_m);
+			I915_WRITE(PIPEB_GMCH_DATA_N,
+					m_n.gmch_n);
+			I915_WRITE(PIPEB_DP_LINK_M, m_n.link_m);
+			I915_WRITE(PIPEB_DP_LINK_N, m_n.link_n);
+		}
 	}
 }
 
@@ -556,8 +595,38 @@ intel_dp_mode_set(struct drm_encoder *encoder, struct drm_display_mode *mode,
 
 	if (intel_crtc->pipe == 1)
 		dp_priv->DP |= DP_PIPEB_SELECT;
+
+	if (IS_eDP(intel_output)) {
+		/* don't miss out required setting for eDP */
+		dp_priv->DP |= DP_PLL_ENABLE;
+		if (adjusted_mode->clock < 200000)
+			dp_priv->DP |= DP_PLL_FREQ_160MHZ;
+		else
+			dp_priv->DP |= DP_PLL_FREQ_270MHZ;
+	}
 }
 
+static void igdng_edp_backlight_on (struct drm_device *dev)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	u32 pp;
+
+	DRM_DEBUG("\n");
+	pp = I915_READ(PCH_PP_CONTROL);
+	pp |= EDP_BLC_ENABLE;
+	I915_WRITE(PCH_PP_CONTROL, pp);
+}
+
+static void igdng_edp_backlight_off (struct drm_device *dev)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	u32 pp;
+
+	DRM_DEBUG("\n");
+	pp = I915_READ(PCH_PP_CONTROL);
+	pp &= ~EDP_BLC_ENABLE;
+	I915_WRITE(PCH_PP_CONTROL, pp);
+}
 
 static void
 intel_dp_dpms(struct drm_encoder *encoder, int mode)
@@ -569,11 +638,17 @@ intel_dp_dpms(struct drm_encoder *encoder, int mode)
 	uint32_t dp_reg = I915_READ(dp_priv->output_reg);
 
 	if (mode != DRM_MODE_DPMS_ON) {
-		if (dp_reg & DP_PORT_EN)
+		if (dp_reg & DP_PORT_EN) {
 			intel_dp_link_down(intel_output, dp_priv->DP);
+			if (IS_eDP(intel_output))
+				igdng_edp_backlight_off(dev);
+		}
 	} else {
-		if (!(dp_reg & DP_PORT_EN))
+		if (!(dp_reg & DP_PORT_EN)) {
 			intel_dp_link_train(intel_output, dp_priv->DP, dp_priv->link_configuration);
+			if (IS_eDP(intel_output))
+				igdng_edp_backlight_on(dev);
+		}
 	}
 	dp_priv->dpms_mode = mode;
 }
@@ -935,6 +1010,23 @@ intel_dp_link_down(struct intel_output *intel_output, uint32_t DP)
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_dp_priv *dp_priv = intel_output->dev_priv;
 
+	DRM_DEBUG("\n");
+
+	if (IS_eDP(intel_output)) {
+		DP &= ~DP_PLL_ENABLE;
+		I915_WRITE(dp_priv->output_reg, DP);
+		POSTING_READ(dp_priv->output_reg);
+		udelay(100);
+	}
+
+	DP &= ~DP_LINK_TRAIN_MASK;
+	I915_WRITE(dp_priv->output_reg, DP | DP_LINK_TRAIN_PAT_IDLE);
+	POSTING_READ(dp_priv->output_reg);
+
+	udelay(17000);
+
+	if (IS_eDP(intel_output))
+		DP |= DP_LINK_TRAIN_OFF;
 	I915_WRITE(dp_priv->output_reg, DP & ~DP_PORT_EN);
 	POSTING_READ(dp_priv->output_reg);
 }
@@ -978,6 +1070,24 @@ intel_dp_check_link_status(struct intel_output *intel_output)
 		intel_dp_link_train(intel_output, dp_priv->DP, dp_priv->link_configuration);
 }
 
+static enum drm_connector_status
+igdng_dp_detect(struct drm_connector *connector)
+{
+	struct intel_output *intel_output = to_intel_output(connector);
+	struct intel_dp_priv *dp_priv = intel_output->dev_priv;
+	enum drm_connector_status status;
+
+	status = connector_status_disconnected;
+	if (intel_dp_aux_native_read(intel_output,
+				     0x000, dp_priv->dpcd,
+				     sizeof (dp_priv->dpcd)) == sizeof (dp_priv->dpcd))
+	{
+		if (dp_priv->dpcd[0] != 0)
+			status = connector_status_connected;
+	}
+	return status;
+}
+
 /**
  * Uses CRT_HOTPLUG_EN and CRT_HOTPLUG_STAT to detect DP connection.
  *
@@ -996,6 +1106,9 @@ intel_dp_detect(struct drm_connector *connector)
 
 	dp_priv->has_audio = false;
 
+	if (IS_IGDNG(dev))
+		return igdng_dp_detect(connector);
+
 	temp = I915_READ(PORT_HOTPLUG_EN);
 
 	I915_WRITE(PORT_HOTPLUG_EN,
@@ -1039,11 +1152,27 @@ intel_dp_detect(struct drm_connector *connector)
 static int intel_dp_get_modes(struct drm_connector *connector)
 {
 	struct intel_output *intel_output = to_intel_output(connector);
+	struct drm_device *dev = intel_output->base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	int ret;
 
 	/* We should parse the EDID data and find out if it has an audio sink
 	 */
 
-	return intel_ddc_get_modes(intel_output);
+	ret = intel_ddc_get_modes(intel_output);
+	if (ret)
+		return ret;
+
+	/* if eDP has no EDID, try to use fixed panel mode from VBT */
+	if (IS_eDP(intel_output)) {
+		if (dev_priv->panel_fixed_mode != NULL) {
+			struct drm_display_mode *mode;
+			mode = drm_mode_duplicate(dev, dev_priv->panel_fixed_mode);
+			drm_mode_probed_add(connector, mode);
+			return 1;
+		}
+	}
+	return 0;
 }
 
 static void
@@ -1106,6 +1235,7 @@ intel_dp_init(struct drm_device *dev, int output_reg)
 	struct drm_connector *connector;
 	struct intel_output *intel_output;
 	struct intel_dp_priv *dp_priv;
+	const char *name = NULL;
 
 	intel_output = kcalloc(sizeof(struct intel_output) + 
 			       sizeof(struct intel_dp_priv), 1, GFP_KERNEL);
@@ -1119,7 +1249,10 @@ intel_dp_init(struct drm_device *dev, int output_reg)
 			   DRM_MODE_CONNECTOR_DisplayPort);
 	drm_connector_helper_add(connector, &intel_dp_connector_helper_funcs);
 
-	intel_output->type = INTEL_OUTPUT_DISPLAYPORT;
+	if (output_reg == DP_A)
+		intel_output->type = INTEL_OUTPUT_EDP;
+	else
+		intel_output->type = INTEL_OUTPUT_DISPLAYPORT;
 
 	connector->interlace_allowed = true;
 	connector->doublescan_allowed = 0;
@@ -1139,12 +1272,41 @@ intel_dp_init(struct drm_device *dev, int output_reg)
 	drm_sysfs_connector_add(connector);
 
 	/* Set up the DDC bus. */
-	intel_dp_i2c_init(intel_output,
-			  (output_reg == DP_B) ? "DPDDC-B" :
-			  (output_reg == DP_C) ? "DPDDC-C" : "DPDDC-D");
+	switch (output_reg) {
+		case DP_A:
+			name = "DPDDC-A";
+			break;
+		case DP_B:
+		case PCH_DP_B:
+			name = "DPDDC-B";
+			break;
+		case DP_C:
+		case PCH_DP_C:
+			name = "DPDDC-C";
+			break;
+		case DP_D:
+		case PCH_DP_D:
+			name = "DPDDC-D";
+			break;
+	}
+
+	intel_dp_i2c_init(intel_output, name);
+
 	intel_output->ddc_bus = &dp_priv->adapter;
 	intel_output->hot_plug = intel_dp_hot_plug;
 
+	if (output_reg == DP_A) {
+		/* initialize panel mode from VBT if available for eDP */
+		if (dev_priv->lfp_lvds_vbt_mode) {
+			dev_priv->panel_fixed_mode =
+				drm_mode_duplicate(dev, dev_priv->lfp_lvds_vbt_mode);
+			if (dev_priv->panel_fixed_mode) {
+				dev_priv->panel_fixed_mode->type |=
+					DRM_MODE_TYPE_PREFERRED;
+			}
+		}
+	}
+
 	/* For G4X desktop chip, PEG_BAND_GAP_DATA 3:0 must first be written
 	 * 0xd.  Failure to do so will result in spurious interrupts being
 	 * generated on the port when a cable is not attached.
diff --git a/drivers/gpu/drm/i915/intel_drv.h b/drivers/gpu/drm/i915/intel_drv.h
index 004541c935a..d6f92ea1b55 100644
--- a/drivers/gpu/drm/i915/intel_drv.h
+++ b/drivers/gpu/drm/i915/intel_drv.h
@@ -55,6 +55,7 @@
 #define INTEL_OUTPUT_TVOUT 5
 #define INTEL_OUTPUT_HDMI 6
 #define INTEL_OUTPUT_DISPLAYPORT 7
+#define INTEL_OUTPUT_EDP 8
 
 #define INTEL_DVO_CHIP_NONE 0
 #define INTEL_DVO_CHIP_LVDS 1
@@ -121,6 +122,8 @@ extern void intel_dp_init(struct drm_device *dev, int dp_reg);
 void
 intel_dp_set_m_n(struct drm_crtc *crtc, struct drm_display_mode *mode,
 		 struct drm_display_mode *adjusted_mode);
+extern void intel_edp_link_config (struct intel_output *, int *, int *);
+
 
 extern void intel_crtc_load_lut(struct drm_crtc *crtc);
 extern void intel_encoder_prepare (struct drm_encoder *encoder);
diff --git a/drivers/gpu/drm/i915/intel_hdmi.c b/drivers/gpu/drm/i915/intel_hdmi.c
index 9e30daae37d..1842290cded 100644
--- a/drivers/gpu/drm/i915/intel_hdmi.c
+++ b/drivers/gpu/drm/i915/intel_hdmi.c
@@ -130,16 +130,17 @@ static bool intel_hdmi_mode_fixup(struct drm_encoder *encoder,
 }
 
 static enum drm_connector_status
-intel_hdmi_edid_detect(struct drm_connector *connector)
+intel_hdmi_detect(struct drm_connector *connector)
 {
 	struct intel_output *intel_output = to_intel_output(connector);
 	struct intel_hdmi_priv *hdmi_priv = intel_output->dev_priv;
 	struct edid *edid = NULL;
 	enum drm_connector_status status = connector_status_disconnected;
 
+	hdmi_priv->has_hdmi_sink = false;
 	edid = drm_get_edid(&intel_output->base,
 			    intel_output->ddc_bus);
-	hdmi_priv->has_hdmi_sink = false;
+
 	if (edid) {
 		if (edid->input & DRM_EDID_INPUT_DIGITAL) {
 			status = connector_status_connected;
@@ -148,65 +149,8 @@ intel_hdmi_edid_detect(struct drm_connector *connector)
 		intel_output->base.display_info.raw_edid = NULL;
 		kfree(edid);
 	}
-	return status;
-}
-
-static enum drm_connector_status
-igdng_hdmi_detect(struct drm_connector *connector)
-{
-	struct intel_output *intel_output = to_intel_output(connector);
-	struct intel_hdmi_priv *hdmi_priv = intel_output->dev_priv;
-
-	/* FIXME hotplug detect */
-
-	hdmi_priv->has_hdmi_sink = false;
-	return intel_hdmi_edid_detect(connector);
-}
 
-static enum drm_connector_status
-intel_hdmi_detect(struct drm_connector *connector)
-{
-	struct drm_device *dev = connector->dev;
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct intel_output *intel_output = to_intel_output(connector);
-	struct intel_hdmi_priv *hdmi_priv = intel_output->dev_priv;
-	u32 temp, bit;
-
-	if (IS_IGDNG(dev))
-		return igdng_hdmi_detect(connector);
-
-	temp = I915_READ(PORT_HOTPLUG_EN);
-
-	switch (hdmi_priv->sdvox_reg) {
-	case SDVOB:
-		temp |= HDMIB_HOTPLUG_INT_EN;
-		break;
-	case SDVOC:
-		temp |= HDMIC_HOTPLUG_INT_EN;
-		break;
-	default:
-		return connector_status_unknown;
-	}
-
-	I915_WRITE(PORT_HOTPLUG_EN, temp);
-
-	POSTING_READ(PORT_HOTPLUG_EN);
-
-	switch (hdmi_priv->sdvox_reg) {
-	case SDVOB:
-		bit = HDMIB_HOTPLUG_INT_STATUS;
-		break;
-	case SDVOC:
-		bit = HDMIC_HOTPLUG_INT_STATUS;
-		break;
-	default:
-		return connector_status_unknown;
-	}
-
-	if ((I915_READ(PORT_HOTPLUG_STAT) & bit) != 0)
-		return intel_hdmi_edid_detect(connector);
-	else
-		return connector_status_disconnected;
+	return status;
 }
 
 static int intel_hdmi_get_modes(struct drm_connector *connector)
diff --git a/drivers/gpu/drm/i915/intel_lvds.c b/drivers/gpu/drm/i915/intel_lvds.c
index 9ab38efffec..3f445a80c55 100644
--- a/drivers/gpu/drm/i915/intel_lvds.c
+++ b/drivers/gpu/drm/i915/intel_lvds.c
@@ -780,6 +780,14 @@ static const struct dmi_system_id intel_no_lvds[] = {
 	},
 	{
 		.callback = intel_no_lvds_dmi_callback,
+		.ident = "AOpen Mini PC MP915",
+		.matches = {
+			DMI_MATCH(DMI_BOARD_VENDOR, "AOpen"),
+			DMI_MATCH(DMI_BOARD_NAME, "i915GMx-F"),
+		},
+	},
+	{
+		.callback = intel_no_lvds_dmi_callback,
 		.ident = "Aopen i945GTt-VFA",
 		.matches = {
 			DMI_MATCH(DMI_PRODUCT_VERSION, "AO00001JW"),
@@ -884,6 +892,10 @@ void intel_lvds_init(struct drm_device *dev)
 	if (IS_IGDNG(dev)) {
 		if ((I915_READ(PCH_LVDS) & LVDS_DETECTED) == 0)
 			return;
+		if (dev_priv->edp_support) {
+			DRM_DEBUG("disable LVDS for eDP support\n");
+			return;
+		}
 		gpio = PCH_GPIOC;
 	}
 
diff --git a/drivers/gpu/drm/i915/intel_sdvo.c b/drivers/gpu/drm/i915/intel_sdvo.c
index 4f0c30948bc..5371d933255 100644
--- a/drivers/gpu/drm/i915/intel_sdvo.c
+++ b/drivers/gpu/drm/i915/intel_sdvo.c
@@ -31,6 +31,7 @@
 #include "drm.h"
 #include "drm_crtc.h"
 #include "intel_drv.h"
+#include "drm_edid.h"
 #include "i915_drm.h"
 #include "i915_drv.h"
 #include "intel_sdvo_regs.h"
@@ -55,6 +56,12 @@ struct intel_sdvo_priv {
 	/* Pixel clock limitations reported by the SDVO device, in kHz */
 	int pixel_clock_min, pixel_clock_max;
 
+	/*
+	* For multiple function SDVO device,
+	* this is for current attached outputs.
+	*/
+	uint16_t attached_output;
+
 	/**
 	 * This is set if we're going to treat the device as TV-out.
 	 *
@@ -114,6 +121,9 @@ struct intel_sdvo_priv {
 	u32 save_SDVOX;
 };
 
+static bool
+intel_sdvo_output_setup(struct intel_output *intel_output, uint16_t flags);
+
 /**
  * Writes the SDVOB or SDVOC with the given value, but always writes both
  * SDVOB and SDVOC to work around apparent hardware issues (according to
@@ -1435,41 +1445,96 @@ void intel_sdvo_set_hotplug(struct drm_connector *connector, int on)
 	intel_sdvo_read_response(intel_output, &response, 2);
 }
 
-static void
-intel_sdvo_hdmi_sink_detect(struct drm_connector *connector)
+static bool
+intel_sdvo_multifunc_encoder(struct intel_output *intel_output)
+{
+	struct intel_sdvo_priv *sdvo_priv = intel_output->dev_priv;
+	int caps = 0;
+
+	if (sdvo_priv->caps.output_flags &
+		(SDVO_OUTPUT_TMDS0 | SDVO_OUTPUT_TMDS1))
+		caps++;
+	if (sdvo_priv->caps.output_flags &
+		(SDVO_OUTPUT_RGB0 | SDVO_OUTPUT_RGB1))
+		caps++;
+	if (sdvo_priv->caps.output_flags &
+		(SDVO_OUTPUT_SVID0 | SDVO_OUTPUT_SVID0))
+		caps++;
+	if (sdvo_priv->caps.output_flags &
+		(SDVO_OUTPUT_CVBS0 | SDVO_OUTPUT_CVBS1))
+		caps++;
+	if (sdvo_priv->caps.output_flags &
+		(SDVO_OUTPUT_YPRPB0 | SDVO_OUTPUT_YPRPB1))
+		caps++;
+
+	if (sdvo_priv->caps.output_flags &
+		(SDVO_OUTPUT_SCART0 | SDVO_OUTPUT_SCART1))
+		caps++;
+
+	if (sdvo_priv->caps.output_flags &
+		(SDVO_OUTPUT_LVDS0 | SDVO_OUTPUT_LVDS1))
+		caps++;
+
+	return (caps > 1);
+}
+
+enum drm_connector_status
+intel_sdvo_hdmi_sink_detect(struct drm_connector *connector, u16 response)
 {
 	struct intel_output *intel_output = to_intel_output(connector);
 	struct intel_sdvo_priv *sdvo_priv = intel_output->dev_priv;
+	enum drm_connector_status status = connector_status_connected;
 	struct edid *edid = NULL;
 
 	edid = drm_get_edid(&intel_output->base,
 			    intel_output->ddc_bus);
 	if (edid != NULL) {
-		sdvo_priv->is_hdmi = drm_detect_hdmi_monitor(edid);
+		/* Don't report the output as connected if it's a DVI-I
+		 * connector with a non-digital EDID coming out.
+		 */
+		if (response & (SDVO_OUTPUT_TMDS0 | SDVO_OUTPUT_TMDS1)) {
+			if (edid->input & DRM_EDID_INPUT_DIGITAL)
+				sdvo_priv->is_hdmi =
+					drm_detect_hdmi_monitor(edid);
+			else
+				status = connector_status_disconnected;
+		}
+
 		kfree(edid);
 		intel_output->base.display_info.raw_edid = NULL;
-	}
+
+	} else if (response & (SDVO_OUTPUT_TMDS0 | SDVO_OUTPUT_TMDS1))
+		status = connector_status_disconnected;
+
+	return status;
 }
 
 static enum drm_connector_status intel_sdvo_detect(struct drm_connector *connector)
 {
-	u8 response[2];
+	uint16_t response;
 	u8 status;
 	struct intel_output *intel_output = to_intel_output(connector);
+	struct intel_sdvo_priv *sdvo_priv = intel_output->dev_priv;
 
 	intel_sdvo_write_cmd(intel_output, SDVO_CMD_GET_ATTACHED_DISPLAYS, NULL, 0);
 	status = intel_sdvo_read_response(intel_output, &response, 2);
 
-	DRM_DEBUG("SDVO response %d %d\n", response[0], response[1]);
+	DRM_DEBUG("SDVO response %d %d\n", response & 0xff, response >> 8);
 
 	if (status != SDVO_CMD_STATUS_SUCCESS)
 		return connector_status_unknown;
 
-	if ((response[0] != 0) || (response[1] != 0)) {
-		intel_sdvo_hdmi_sink_detect(connector);
-		return connector_status_connected;
-	} else
+	if (response == 0)
 		return connector_status_disconnected;
+
+	if (intel_sdvo_multifunc_encoder(intel_output) &&
+		sdvo_priv->attached_output != response) {
+		if (sdvo_priv->controlled_output != response &&
+			intel_sdvo_output_setup(intel_output, response) != true)
+			return connector_status_unknown;
+		sdvo_priv->attached_output = response;
+	}
+	return intel_sdvo_hdmi_sink_detect(connector, response);
 }
 
 static void intel_sdvo_get_ddc_modes(struct drm_connector *connector)
@@ -1866,16 +1931,101 @@ intel_sdvo_get_slave_addr(struct drm_device *dev, int output_device)
 		return 0x72;
 }
 
+static bool
+intel_sdvo_output_setup(struct intel_output *intel_output, uint16_t flags)
+{
+	struct drm_connector *connector = &intel_output->base;
+	struct drm_encoder *encoder = &intel_output->enc;
+	struct intel_sdvo_priv *sdvo_priv = intel_output->dev_priv;
+	bool ret = true, registered = false;
+
+	sdvo_priv->is_tv = false;
+	intel_output->needs_tv_clock = false;
+	sdvo_priv->is_lvds = false;
+
+	if (device_is_registered(&connector->kdev)) {
+		drm_sysfs_connector_remove(connector);
+		registered = true;
+	}
+
+	if (flags &
+	    (SDVO_OUTPUT_TMDS0 | SDVO_OUTPUT_TMDS1)) {
+		if (sdvo_priv->caps.output_flags & SDVO_OUTPUT_TMDS0)
+			sdvo_priv->controlled_output = SDVO_OUTPUT_TMDS0;
+		else
+			sdvo_priv->controlled_output = SDVO_OUTPUT_TMDS1;
+
+		encoder->encoder_type = DRM_MODE_ENCODER_TMDS;
+		connector->connector_type = DRM_MODE_CONNECTOR_DVID;
+
+		if (intel_sdvo_get_supp_encode(intel_output,
+					       &sdvo_priv->encode) &&
+		    intel_sdvo_get_digital_encoding_mode(intel_output) &&
+		    sdvo_priv->is_hdmi) {
+			/* enable hdmi encoding mode if supported */
+			intel_sdvo_set_encode(intel_output, SDVO_ENCODE_HDMI);
+			intel_sdvo_set_colorimetry(intel_output,
+						   SDVO_COLORIMETRY_RGB256);
+			connector->connector_type = DRM_MODE_CONNECTOR_HDMIA;
+		}
+	} else if (flags & SDVO_OUTPUT_SVID0) {
+
+		sdvo_priv->controlled_output = SDVO_OUTPUT_SVID0;
+		encoder->encoder_type = DRM_MODE_ENCODER_TVDAC;
+		connector->connector_type = DRM_MODE_CONNECTOR_SVIDEO;
+		sdvo_priv->is_tv = true;
+		intel_output->needs_tv_clock = true;
+	} else if (flags & SDVO_OUTPUT_RGB0) {
+
+		sdvo_priv->controlled_output = SDVO_OUTPUT_RGB0;
+		encoder->encoder_type = DRM_MODE_ENCODER_DAC;
+		connector->connector_type = DRM_MODE_CONNECTOR_VGA;
+	} else if (flags & SDVO_OUTPUT_RGB1) {
+
+		sdvo_priv->controlled_output = SDVO_OUTPUT_RGB1;
+		encoder->encoder_type = DRM_MODE_ENCODER_DAC;
+		connector->connector_type = DRM_MODE_CONNECTOR_VGA;
+	} else if (flags & SDVO_OUTPUT_LVDS0) {
+
+		sdvo_priv->controlled_output = SDVO_OUTPUT_LVDS0;
+		encoder->encoder_type = DRM_MODE_ENCODER_LVDS;
+		connector->connector_type = DRM_MODE_CONNECTOR_LVDS;
+		sdvo_priv->is_lvds = true;
+	} else if (flags & SDVO_OUTPUT_LVDS1) {
+
+		sdvo_priv->controlled_output = SDVO_OUTPUT_LVDS1;
+		encoder->encoder_type = DRM_MODE_ENCODER_LVDS;
+		connector->connector_type = DRM_MODE_CONNECTOR_LVDS;
+		sdvo_priv->is_lvds = true;
+	} else {
+
+		unsigned char bytes[2];
+
+		sdvo_priv->controlled_output = 0;
+		memcpy(bytes, &sdvo_priv->caps.output_flags, 2);
+		DRM_DEBUG_KMS(I915_SDVO,
+				"%s: Unknown SDVO output type (0x%02x%02x)\n",
+				  SDVO_NAME(sdvo_priv),
+				  bytes[0], bytes[1]);
+		ret = false;
+	}
+
+	if (ret && registered)
+		ret = drm_sysfs_connector_add(connector) == 0 ? true : false;
+
+
+	return ret;
+
+}
+
 bool intel_sdvo_init(struct drm_device *dev, int output_device)
 {
 	struct drm_connector *connector;
 	struct intel_output *intel_output;
 	struct intel_sdvo_priv *sdvo_priv;
 
-	int connector_type;
 	u8 ch[0x40];
 	int i;
-	int encoder_type;
 
 	intel_output = kcalloc(sizeof(struct intel_output)+sizeof(struct intel_sdvo_priv), 1, GFP_KERNEL);
 	if (!intel_output) {
@@ -1925,88 +2075,28 @@ bool intel_sdvo_init(struct drm_device *dev, int output_device)
 	intel_output->ddc_bus->algo = &intel_sdvo_i2c_bit_algo;
 
 	/* In defaut case sdvo lvds is false */
-	sdvo_priv->is_lvds = false;
 	intel_sdvo_get_capabilities(intel_output, &sdvo_priv->caps);
 
-	if (sdvo_priv->caps.output_flags &
-	    (SDVO_OUTPUT_TMDS0 | SDVO_OUTPUT_TMDS1)) {
-		if (sdvo_priv->caps.output_flags & SDVO_OUTPUT_TMDS0)
-			sdvo_priv->controlled_output = SDVO_OUTPUT_TMDS0;
-		else
-			sdvo_priv->controlled_output = SDVO_OUTPUT_TMDS1;
-
-		encoder_type = DRM_MODE_ENCODER_TMDS;
-		connector_type = DRM_MODE_CONNECTOR_DVID;
-
-		if (intel_sdvo_get_supp_encode(intel_output,
-					       &sdvo_priv->encode) &&
-		    intel_sdvo_get_digital_encoding_mode(intel_output) &&
-		    sdvo_priv->is_hdmi) {
-			/* enable hdmi encoding mode if supported */
-			intel_sdvo_set_encode(intel_output, SDVO_ENCODE_HDMI);
-			intel_sdvo_set_colorimetry(intel_output,
-						   SDVO_COLORIMETRY_RGB256);
-			connector_type = DRM_MODE_CONNECTOR_HDMIA;
-		}
-	}
-	else if (sdvo_priv->caps.output_flags & SDVO_OUTPUT_SVID0)
-	{
-		sdvo_priv->controlled_output = SDVO_OUTPUT_SVID0;
-		encoder_type = DRM_MODE_ENCODER_TVDAC;
-		connector_type = DRM_MODE_CONNECTOR_SVIDEO;
-		sdvo_priv->is_tv = true;
-		intel_output->needs_tv_clock = true;
-	}
-	else if (sdvo_priv->caps.output_flags & SDVO_OUTPUT_RGB0)
-	{
-		sdvo_priv->controlled_output = SDVO_OUTPUT_RGB0;
-		encoder_type = DRM_MODE_ENCODER_DAC;
-		connector_type = DRM_MODE_CONNECTOR_VGA;
-	}
-	else if (sdvo_priv->caps.output_flags & SDVO_OUTPUT_RGB1)
-	{
-		sdvo_priv->controlled_output = SDVO_OUTPUT_RGB1;
-		encoder_type = DRM_MODE_ENCODER_DAC;
-		connector_type = DRM_MODE_CONNECTOR_VGA;
-	}
-	else if (sdvo_priv->caps.output_flags & SDVO_OUTPUT_LVDS0)
-	{
-		sdvo_priv->controlled_output = SDVO_OUTPUT_LVDS0;
-		encoder_type = DRM_MODE_ENCODER_LVDS;
-		connector_type = DRM_MODE_CONNECTOR_LVDS;
-		sdvo_priv->is_lvds = true;
-	}
-	else if (sdvo_priv->caps.output_flags & SDVO_OUTPUT_LVDS1)
-	{
-		sdvo_priv->controlled_output = SDVO_OUTPUT_LVDS1;
-		encoder_type = DRM_MODE_ENCODER_LVDS;
-		connector_type = DRM_MODE_CONNECTOR_LVDS;
-		sdvo_priv->is_lvds = true;
-	}
-	else
-	{
-		unsigned char bytes[2];
-
-		sdvo_priv->controlled_output = 0;
-		memcpy (bytes, &sdvo_priv->caps.output_flags, 2);
-		DRM_DEBUG_KMS(I915_SDVO,
-				"%s: Unknown SDVO output type (0x%02x%02x)\n",
-				  SDVO_NAME(sdvo_priv),
-				  bytes[0], bytes[1]);
-		encoder_type = DRM_MODE_ENCODER_NONE;
-		connector_type = DRM_MODE_CONNECTOR_Unknown;
+	if (intel_sdvo_output_setup(intel_output,
+				    sdvo_priv->caps.output_flags) != true) {
+		DRM_DEBUG("SDVO output failed to setup on SDVO%c\n",
+			  output_device == SDVOB ? 'B' : 'C');
 		goto err_i2c;
 	}
 
+
 	connector = &intel_output->base;
 	drm_connector_init(dev, connector, &intel_sdvo_connector_funcs,
-			   connector_type);
+			   connector->connector_type);
+
 	drm_connector_helper_add(connector, &intel_sdvo_connector_helper_funcs);
 	connector->interlace_allowed = 0;
 	connector->doublescan_allowed = 0;
 	connector->display_info.subpixel_order = SubPixelHorizontalRGB;
 
-	drm_encoder_init(dev, &intel_output->enc, &intel_sdvo_enc_funcs, encoder_type);
+	drm_encoder_init(dev, &intel_output->enc,
+			&intel_sdvo_enc_funcs, intel_output->enc.encoder_type);
+
 	drm_encoder_helper_add(&intel_output->enc, &intel_sdvo_helper_funcs);
 
 	drm_mode_connector_attach_encoder(&intel_output->base, &intel_output->enc);
diff --git a/drivers/gpu/drm/i915/intel_tv.c b/drivers/gpu/drm/i915/intel_tv.c
index a43c98e3f07..da4ab4dc163 100644
--- a/drivers/gpu/drm/i915/intel_tv.c
+++ b/drivers/gpu/drm/i915/intel_tv.c
@@ -1490,6 +1490,27 @@ static struct input_res {
 	{"1920x1080", 1920, 1080},
 };
 
+/*
+ * Chose preferred mode  according to line number of TV format
+ */
+static void
+intel_tv_chose_preferred_modes(struct drm_connector *connector,
+			       struct drm_display_mode *mode_ptr)
+{
+	struct intel_output *intel_output = to_intel_output(connector);
+	const struct tv_mode *tv_mode = intel_tv_mode_find(intel_output);
+
+	if (tv_mode->nbr_end < 480 && mode_ptr->vdisplay == 480)
+		mode_ptr->type |= DRM_MODE_TYPE_PREFERRED;
+	else if (tv_mode->nbr_end > 480) {
+		if (tv_mode->progressive == true && tv_mode->nbr_end < 720) {
+			if (mode_ptr->vdisplay == 720)
+				mode_ptr->type |= DRM_MODE_TYPE_PREFERRED;
+		} else if (mode_ptr->vdisplay == 1080)
+				mode_ptr->type |= DRM_MODE_TYPE_PREFERRED;
+	}
+}
+
 /**
  * Stub get_modes function.
  *
@@ -1544,6 +1565,7 @@ intel_tv_get_modes(struct drm_connector *connector)
 		mode_ptr->clock = (int) tmp;
 
 		mode_ptr->type = DRM_MODE_TYPE_DRIVER;
+		intel_tv_chose_preferred_modes(connector, mode_ptr);
 		drm_mode_probed_add(connector, mode_ptr);
 		count++;
 	}
diff --git a/drivers/gpu/drm/radeon/Makefile b/drivers/gpu/drm/radeon/Makefile
index 5fae1e074b4..013d3805994 100644
--- a/drivers/gpu/drm/radeon/Makefile
+++ b/drivers/gpu/drm/radeon/Makefile
@@ -13,7 +13,8 @@ radeon-$(CONFIG_DRM_RADEON_KMS) += radeon_device.o radeon_kms.o \
 	radeon_encoders.o radeon_display.o radeon_cursor.o radeon_i2c.o \
 	radeon_clocks.o radeon_fb.o radeon_gem.o radeon_ring.o radeon_irq_kms.o \
 	radeon_cs.o radeon_bios.o radeon_benchmark.o r100.o r300.o r420.o \
-	rs400.o rs600.o rs690.o rv515.o r520.o r600.o rs780.o rv770.o
+	rs400.o rs600.o rs690.o rv515.o r520.o r600.o rs780.o rv770.o \
+	radeon_test.o
 
 radeon-$(CONFIG_COMPAT) += radeon_ioc32.o
 
diff --git a/drivers/gpu/drm/radeon/atombios_crtc.c b/drivers/gpu/drm/radeon/atombios_crtc.c
index c0080cc9bf8..74d034f77c6 100644
--- a/drivers/gpu/drm/radeon/atombios_crtc.c
+++ b/drivers/gpu/drm/radeon/atombios_crtc.c
@@ -31,6 +31,132 @@
 #include "atom.h"
 #include "atom-bits.h"
 
+static void atombios_overscan_setup(struct drm_crtc *crtc,
+				    struct drm_display_mode *mode,
+				    struct drm_display_mode *adjusted_mode)
+{
+	struct drm_device *dev = crtc->dev;
+	struct radeon_device *rdev = dev->dev_private;
+	struct radeon_crtc *radeon_crtc = to_radeon_crtc(crtc);
+	SET_CRTC_OVERSCAN_PS_ALLOCATION args;
+	int index = GetIndexIntoMasterTable(COMMAND, SetCRTC_OverScan);
+	int a1, a2;
+
+	memset(&args, 0, sizeof(args));
+
+	args.usOverscanRight = 0;
+	args.usOverscanLeft = 0;
+	args.usOverscanBottom = 0;
+	args.usOverscanTop = 0;
+	args.ucCRTC = radeon_crtc->crtc_id;
+
+	switch (radeon_crtc->rmx_type) {
+	case RMX_CENTER:
+		args.usOverscanTop = (adjusted_mode->crtc_vdisplay - mode->crtc_vdisplay) / 2;
+		args.usOverscanBottom = (adjusted_mode->crtc_vdisplay - mode->crtc_vdisplay) / 2;
+		args.usOverscanLeft = (adjusted_mode->crtc_hdisplay - mode->crtc_hdisplay) / 2;
+		args.usOverscanRight = (adjusted_mode->crtc_hdisplay - mode->crtc_hdisplay) / 2;
+		atom_execute_table(rdev->mode_info.atom_context, index, (uint32_t *)&args);
+		break;
+	case RMX_ASPECT:
+		a1 = mode->crtc_vdisplay * adjusted_mode->crtc_hdisplay;
+		a2 = adjusted_mode->crtc_vdisplay * mode->crtc_hdisplay;
+
+		if (a1 > a2) {
+			args.usOverscanLeft = (adjusted_mode->crtc_hdisplay - (a2 / mode->crtc_vdisplay)) / 2;
+			args.usOverscanRight = (adjusted_mode->crtc_hdisplay - (a2 / mode->crtc_vdisplay)) / 2;
+		} else if (a2 > a1) {
+			args.usOverscanLeft = (adjusted_mode->crtc_vdisplay - (a1 / mode->crtc_hdisplay)) / 2;
+			args.usOverscanRight = (adjusted_mode->crtc_vdisplay - (a1 / mode->crtc_hdisplay)) / 2;
+		}
+		atom_execute_table(rdev->mode_info.atom_context, index, (uint32_t *)&args);
+		break;
+	case RMX_FULL:
+	default:
+		args.usOverscanRight = 0;
+		args.usOverscanLeft = 0;
+		args.usOverscanBottom = 0;
+		args.usOverscanTop = 0;
+		atom_execute_table(rdev->mode_info.atom_context, index, (uint32_t *)&args);
+		break;
+	}
+}
+
+static void atombios_scaler_setup(struct drm_crtc *crtc)
+{
+	struct drm_device *dev = crtc->dev;
+	struct radeon_device *rdev = dev->dev_private;
+	struct radeon_crtc *radeon_crtc = to_radeon_crtc(crtc);
+	ENABLE_SCALER_PS_ALLOCATION args;
+	int index = GetIndexIntoMasterTable(COMMAND, EnableScaler);
+	/* fixme - fill in enc_priv for atom dac */
+	enum radeon_tv_std tv_std = TV_STD_NTSC;
+
+	if (!ASIC_IS_AVIVO(rdev) && radeon_crtc->crtc_id)
+		return;
+
+	memset(&args, 0, sizeof(args));
+
+	args.ucScaler = radeon_crtc->crtc_id;
+
+	if (radeon_crtc->devices & (ATOM_DEVICE_TV_SUPPORT)) {
+		switch (tv_std) {
+		case TV_STD_NTSC:
+		default:
+			args.ucTVStandard = ATOM_TV_NTSC;
+			break;
+		case TV_STD_PAL:
+			args.ucTVStandard = ATOM_TV_PAL;
+			break;
+		case TV_STD_PAL_M:
+			args.ucTVStandard = ATOM_TV_PALM;
+			break;
+		case TV_STD_PAL_60:
+			args.ucTVStandard = ATOM_TV_PAL60;
+			break;
+		case TV_STD_NTSC_J:
+			args.ucTVStandard = ATOM_TV_NTSCJ;
+			break;
+		case TV_STD_SCART_PAL:
+			args.ucTVStandard = ATOM_TV_PAL; /* ??? */
+			break;
+		case TV_STD_SECAM:
+			args.ucTVStandard = ATOM_TV_SECAM;
+			break;
+		case TV_STD_PAL_CN:
+			args.ucTVStandard = ATOM_TV_PALCN;
+			break;
+		}
+		args.ucEnable = SCALER_ENABLE_MULTITAP_MODE;
+	} else if (radeon_crtc->devices & (ATOM_DEVICE_CV_SUPPORT)) {
+		args.ucTVStandard = ATOM_TV_CV;
+		args.ucEnable = SCALER_ENABLE_MULTITAP_MODE;
+	} else {
+		switch (radeon_crtc->rmx_type) {
+		case RMX_FULL:
+			args.ucEnable = ATOM_SCALER_EXPANSION;
+			break;
+		case RMX_CENTER:
+			args.ucEnable = ATOM_SCALER_CENTER;
+			break;
+		case RMX_ASPECT:
+			args.ucEnable = ATOM_SCALER_EXPANSION;
+			break;
+		default:
+			if (ASIC_IS_AVIVO(rdev))
+				args.ucEnable = ATOM_SCALER_DISABLE;
+			else
+				args.ucEnable = ATOM_SCALER_CENTER;
+			break;
+		}
+	}
+	atom_execute_table(rdev->mode_info.atom_context, index, (uint32_t *)&args);
+	if (radeon_crtc->devices & (ATOM_DEVICE_CV_SUPPORT | ATOM_DEVICE_TV_SUPPORT)
+	    && rdev->family >= CHIP_RV515 && rdev->family <= CHIP_RV570) {
+		atom_rv515_force_tv_scaler(rdev);
+	}
+}
+
 static void atombios_lock_crtc(struct drm_crtc *crtc, int lock)
 {
 	struct radeon_crtc *radeon_crtc = to_radeon_crtc(crtc);
@@ -203,6 +329,12 @@ void atombios_crtc_set_pll(struct drm_crtc *crtc, struct drm_display_mode *mode)
 	if (ASIC_IS_AVIVO(rdev)) {
 		uint32_t ss_cntl;
 
+		if ((rdev->family == CHIP_RS600) ||
+		    (rdev->family == CHIP_RS690) ||
+		    (rdev->family == CHIP_RS740))
+			pll_flags |= (RADEON_PLL_USE_FRAC_FB_DIV |
+				      RADEON_PLL_PREFER_CLOSEST_LOWER);
+
 		if (ASIC_IS_DCE32(rdev) && mode->clock > 200000)	/* range limits??? */
 			pll_flags |= RADEON_PLL_PREFER_HIGH_FB_DIV;
 		else
@@ -321,7 +453,7 @@ int atombios_crtc_set_base(struct drm_crtc *crtc, int x, int y,
 	struct drm_gem_object *obj;
 	struct drm_radeon_gem_object *obj_priv;
 	uint64_t fb_location;
-	uint32_t fb_format, fb_pitch_pixels;
+	uint32_t fb_format, fb_pitch_pixels, tiling_flags;
 
 	if (!crtc->fb)
 		return -EINVAL;
@@ -358,7 +490,14 @@ int atombios_crtc_set_base(struct drm_crtc *crtc, int x, int y,
 		return -EINVAL;
 	}
 
-	/* TODO tiling */
+	radeon_object_get_tiling_flags(obj->driver_private,
+				       &tiling_flags, NULL);
+	if (tiling_flags & RADEON_TILING_MACRO)
+		fb_format |= AVIVO_D1GRPH_MACRO_ADDRESS_MODE;
+
+	if (tiling_flags & RADEON_TILING_MICRO)
+		fb_format |= AVIVO_D1GRPH_TILED;
+
 	if (radeon_crtc->crtc_id == 0)
 		WREG32(AVIVO_D1VGA_CONTROL, 0);
 	else
@@ -509,6 +648,9 @@ int atombios_crtc_mode_set(struct drm_crtc *crtc,
 		radeon_crtc_set_base(crtc, x, y, old_fb);
 		radeon_legacy_atom_set_surface(crtc);
 	}
+	atombios_overscan_setup(crtc, mode, adjusted_mode);
+	atombios_scaler_setup(crtc);
+	radeon_bandwidth_update(rdev);
 	return 0;
 }
 
@@ -516,6 +658,8 @@ static bool atombios_crtc_mode_fixup(struct drm_crtc *crtc,
 				     struct drm_display_mode *mode,
 				     struct drm_display_mode *adjusted_mode)
 {
+	if (!radeon_crtc_scaling_mode_fixup(crtc, mode, adjusted_mode))
+		return false;
 	return true;
 }
 
@@ -548,148 +692,3 @@ void radeon_atombios_init_crtc(struct drm_device *dev,
 		    AVIVO_D2CRTC_H_TOTAL - AVIVO_D1CRTC_H_TOTAL;
 	drm_crtc_helper_add(&radeon_crtc->base, &atombios_helper_funcs);
 }
-
-void radeon_init_disp_bw_avivo(struct drm_device *dev,
-			       struct drm_display_mode *mode1,
-			       uint32_t pixel_bytes1,
-			       struct drm_display_mode *mode2,
-			       uint32_t pixel_bytes2)
-{
-	struct radeon_device *rdev = dev->dev_private;
-	fixed20_12 min_mem_eff;
-	fixed20_12 peak_disp_bw, mem_bw, pix_clk, pix_clk2, temp_ff;
-	fixed20_12 sclk_ff, mclk_ff;
-	uint32_t dc_lb_memory_split, temp;
-
-	min_mem_eff.full = rfixed_const_8(0);
-	if (rdev->disp_priority == 2) {
-		uint32_t mc_init_misc_lat_timer = 0;
-		if (rdev->family == CHIP_RV515)
-			mc_init_misc_lat_timer =
-			    RREG32_MC(RV515_MC_INIT_MISC_LAT_TIMER);
-		else if (rdev->family == CHIP_RS690)
-			mc_init_misc_lat_timer =
-			    RREG32_MC(RS690_MC_INIT_MISC_LAT_TIMER);
-
-		mc_init_misc_lat_timer &=
-		    ~(R300_MC_DISP1R_INIT_LAT_MASK <<
-		      R300_MC_DISP1R_INIT_LAT_SHIFT);
-		mc_init_misc_lat_timer &=
-		    ~(R300_MC_DISP0R_INIT_LAT_MASK <<
-		      R300_MC_DISP0R_INIT_LAT_SHIFT);
-
-		if (mode2)
-			mc_init_misc_lat_timer |=
-			    (1 << R300_MC_DISP1R_INIT_LAT_SHIFT);
-		if (mode1)
-			mc_init_misc_lat_timer |=
-			    (1 << R300_MC_DISP0R_INIT_LAT_SHIFT);
-
-		if (rdev->family == CHIP_RV515)
-			WREG32_MC(RV515_MC_INIT_MISC_LAT_TIMER,
-				  mc_init_misc_lat_timer);
-		else if (rdev->family == CHIP_RS690)
-			WREG32_MC(RS690_MC_INIT_MISC_LAT_TIMER,
-				  mc_init_misc_lat_timer);
-	}
-
-	/*
-	 * determine is there is enough bw for current mode
-	 */
-	temp_ff.full = rfixed_const(100);
-	mclk_ff.full = rfixed_const(rdev->clock.default_mclk);
-	mclk_ff.full = rfixed_div(mclk_ff, temp_ff);
-	sclk_ff.full = rfixed_const(rdev->clock.default_sclk);
-	sclk_ff.full = rfixed_div(sclk_ff, temp_ff);
-
-	temp = (rdev->mc.vram_width / 8) * (rdev->mc.vram_is_ddr ? 2 : 1);
-	temp_ff.full = rfixed_const(temp);
-	mem_bw.full = rfixed_mul(mclk_ff, temp_ff);
-	mem_bw.full = rfixed_mul(mem_bw, min_mem_eff);
-
-	pix_clk.full = 0;
-	pix_clk2.full = 0;
-	peak_disp_bw.full = 0;
-	if (mode1) {
-		temp_ff.full = rfixed_const(1000);
-		pix_clk.full = rfixed_const(mode1->clock);	/* convert to fixed point */
-		pix_clk.full = rfixed_div(pix_clk, temp_ff);
-		temp_ff.full = rfixed_const(pixel_bytes1);
-		peak_disp_bw.full += rfixed_mul(pix_clk, temp_ff);
-	}
-	if (mode2) {
-		temp_ff.full = rfixed_const(1000);
-		pix_clk2.full = rfixed_const(mode2->clock);	/* convert to fixed point */
-		pix_clk2.full = rfixed_div(pix_clk2, temp_ff);
-		temp_ff.full = rfixed_const(pixel_bytes2);
-		peak_disp_bw.full += rfixed_mul(pix_clk2, temp_ff);
-	}
-
-	if (peak_disp_bw.full >= mem_bw.full) {
-		DRM_ERROR
-		    ("You may not have enough display bandwidth for current mode\n"
-		     "If you have flickering problem, try to lower resolution, refresh rate, or color depth\n");
-		printk("peak disp bw %d, mem_bw %d\n",
-		       rfixed_trunc(peak_disp_bw), rfixed_trunc(mem_bw));
-	}
-
-	/*
-	 * Line Buffer Setup
-	 * There is a single line buffer shared by both display controllers.
-	 * DC_LB_MEMORY_SPLIT controls how that line buffer is shared between the display
-	 * controllers.  The paritioning can either be done manually or via one of four
-	 * preset allocations specified in bits 1:0:
-	 * 0 - line buffer is divided in half and shared between each display controller
-	 * 1 - D1 gets 3/4 of the line buffer, D2 gets 1/4
-	 * 2 - D1 gets the whole buffer
-	 * 3 - D1 gets 1/4 of the line buffer, D2 gets 3/4
-	 * Setting bit 2 of DC_LB_MEMORY_SPLIT controls switches to manual allocation mode.
-	 * In manual allocation mode, D1 always starts at 0, D1 end/2 is specified in bits
-	 * 14:4; D2 allocation follows D1.
-	 */
-
-	/* is auto or manual better ? */
-	dc_lb_memory_split =
-	    RREG32(AVIVO_DC_LB_MEMORY_SPLIT) & ~AVIVO_DC_LB_MEMORY_SPLIT_MASK;
-	dc_lb_memory_split &= ~AVIVO_DC_LB_MEMORY_SPLIT_SHIFT_MODE;
-#if 1
-	/* auto */
-	if (mode1 && mode2) {
-		if (mode1->hdisplay > mode2->hdisplay) {
-			if (mode1->hdisplay > 2560)
-				dc_lb_memory_split |=
-				    AVIVO_DC_LB_MEMORY_SPLIT_D1_3Q_D2_1Q;
-			else
-				dc_lb_memory_split |=
-				    AVIVO_DC_LB_MEMORY_SPLIT_D1HALF_D2HALF;
-		} else if (mode2->hdisplay > mode1->hdisplay) {
-			if (mode2->hdisplay > 2560)
-				dc_lb_memory_split |=
-				    AVIVO_DC_LB_MEMORY_SPLIT_D1_1Q_D2_3Q;
-			else
-				dc_lb_memory_split |=
-				    AVIVO_DC_LB_MEMORY_SPLIT_D1HALF_D2HALF;
-		} else
-			dc_lb_memory_split |=
-			    AVIVO_DC_LB_MEMORY_SPLIT_D1HALF_D2HALF;
-	} else if (mode1) {
-		dc_lb_memory_split |= AVIVO_DC_LB_MEMORY_SPLIT_D1_ONLY;
-	} else if (mode2) {
-		dc_lb_memory_split |= AVIVO_DC_LB_MEMORY_SPLIT_D1_1Q_D2_3Q;
-	}
-#else
-	/* manual */
-	dc_lb_memory_split |= AVIVO_DC_LB_MEMORY_SPLIT_SHIFT_MODE;
-	dc_lb_memory_split &=
-	    ~(AVIVO_DC_LB_DISP1_END_ADR_MASK <<
-	      AVIVO_DC_LB_DISP1_END_ADR_SHIFT);
-	if (mode1) {
-		dc_lb_memory_split |=
-		    ((((mode1->hdisplay / 2) + 64) & AVIVO_DC_LB_DISP1_END_ADR_MASK)
-		     << AVIVO_DC_LB_DISP1_END_ADR_SHIFT);
-	} else if (mode2) {
-		dc_lb_memory_split |= (0 << AVIVO_DC_LB_DISP1_END_ADR_SHIFT);
-	}
-#endif
-	WREG32(AVIVO_DC_LB_MEMORY_SPLIT, dc_lb_memory_split);
-}
diff --git a/drivers/gpu/drm/radeon/r100.c b/drivers/gpu/drm/radeon/r100.c
index c550932a108..f1ba8ff4113 100644
--- a/drivers/gpu/drm/radeon/r100.c
+++ b/drivers/gpu/drm/radeon/r100.c
@@ -110,7 +110,7 @@ int r100_pci_gart_set_page(struct radeon_device *rdev, int i, uint64_t addr)
 	if (i < 0 || i > rdev->gart.num_gpu_pages) {
 		return -EINVAL;
 	}
-	rdev->gart.table.ram.ptr[i] = cpu_to_le32((uint32_t)addr);
+	rdev->gart.table.ram.ptr[i] = cpu_to_le32(lower_32_bits(addr));
 	return 0;
 }
 
@@ -173,8 +173,12 @@ void r100_mc_setup(struct radeon_device *rdev)
 		DRM_ERROR("Failed to register debugfs file for R100 MC !\n");
 	}
 	/* Write VRAM size in case we are limiting it */
-	WREG32(RADEON_CONFIG_MEMSIZE, rdev->mc.vram_size);
-	tmp = rdev->mc.vram_location + rdev->mc.vram_size - 1;
+	WREG32(RADEON_CONFIG_MEMSIZE, rdev->mc.real_vram_size);
+	/* Novell bug 204882 for RN50/M6/M7 with 8/16/32MB VRAM,
+	 * if the aperture is 64MB but we have 32MB VRAM
+	 * we report only 32MB VRAM but we have to set MC_FB_LOCATION
+	 * to 64MB, otherwise the gpu accidentially dies */
+	tmp = rdev->mc.vram_location + rdev->mc.mc_vram_size - 1;
 	tmp = REG_SET(RADEON_MC_FB_TOP, tmp >> 16);
 	tmp |= REG_SET(RADEON_MC_FB_START, rdev->mc.vram_location >> 16);
 	WREG32(RADEON_MC_FB_LOCATION, tmp);
@@ -215,7 +219,6 @@ int r100_mc_init(struct radeon_device *rdev)
 	r100_pci_gart_disable(rdev);
 
 	/* Setup GPU memory space */
-	rdev->mc.vram_location = 0xFFFFFFFFUL;
 	rdev->mc.gtt_location = 0xFFFFFFFFUL;
 	if (rdev->flags & RADEON_IS_AGP) {
 		r = radeon_agp_init(rdev);
@@ -719,13 +722,14 @@ int r100_cs_packet_parse(struct radeon_cs_parser *p,
 			 unsigned idx)
 {
 	struct radeon_cs_chunk *ib_chunk = &p->chunks[p->chunk_ib_idx];
-	uint32_t header = ib_chunk->kdata[idx];
+	uint32_t header;
 
 	if (idx >= ib_chunk->length_dw) {
 		DRM_ERROR("Can not parse packet at %d after CS end %d !\n",
 			  idx, ib_chunk->length_dw);
 		return -EINVAL;
 	}
+	header = ib_chunk->kdata[idx];
 	pkt->idx = idx;
 	pkt->type = CP_PACKET_GET_TYPE(header);
 	pkt->count = CP_PACKET_GET_COUNT(header);
@@ -753,6 +757,102 @@ int r100_cs_packet_parse(struct radeon_cs_parser *p,
 }
 
 /**
+ * r100_cs_packet_next_vline() - parse userspace VLINE packet
+ * @parser:		parser structure holding parsing context.
+ *
+ * Userspace sends a special sequence for VLINE waits.
+ * PACKET0 - VLINE_START_END + value
+ * PACKET0 - WAIT_UNTIL +_value
+ * RELOC (P3) - crtc_id in reloc.
+ *
+ * This function parses this and relocates the VLINE START END
+ * and WAIT UNTIL packets to the correct crtc.
+ * It also detects a switched off crtc and nulls out the
+ * wait in that case.
+ */
+int r100_cs_packet_parse_vline(struct radeon_cs_parser *p)
+{
+	struct radeon_cs_chunk *ib_chunk;
+	struct drm_mode_object *obj;
+	struct drm_crtc *crtc;
+	struct radeon_crtc *radeon_crtc;
+	struct radeon_cs_packet p3reloc, waitreloc;
+	int crtc_id;
+	int r;
+	uint32_t header, h_idx, reg;
+
+	ib_chunk = &p->chunks[p->chunk_ib_idx];
+
+	/* parse the wait until */
+	r = r100_cs_packet_parse(p, &waitreloc, p->idx);
+	if (r)
+		return r;
+
+	/* check its a wait until and only 1 count */
+	if (waitreloc.reg != RADEON_WAIT_UNTIL ||
+	    waitreloc.count != 0) {
+		DRM_ERROR("vline wait had illegal wait until segment\n");
+		r = -EINVAL;
+		return r;
+	}
+
+	if (ib_chunk->kdata[waitreloc.idx + 1] != RADEON_WAIT_CRTC_VLINE) {
+		DRM_ERROR("vline wait had illegal wait until\n");
+		r = -EINVAL;
+		return r;
+	}
+
+	/* jump over the NOP */
+	r = r100_cs_packet_parse(p, &p3reloc, p->idx);
+	if (r)
+		return r;
+
+	h_idx = p->idx - 2;
+	p->idx += waitreloc.count;
+	p->idx += p3reloc.count;
+
+	header = ib_chunk->kdata[h_idx];
+	crtc_id = ib_chunk->kdata[h_idx + 5];
+	reg = ib_chunk->kdata[h_idx] >> 2;
+	mutex_lock(&p->rdev->ddev->mode_config.mutex);
+	obj = drm_mode_object_find(p->rdev->ddev, crtc_id, DRM_MODE_OBJECT_CRTC);
+	if (!obj) {
+		DRM_ERROR("cannot find crtc %d\n", crtc_id);
+		r = -EINVAL;
+		goto out;
+	}
+	crtc = obj_to_crtc(obj);
+	radeon_crtc = to_radeon_crtc(crtc);
+	crtc_id = radeon_crtc->crtc_id;
+
+	if (!crtc->enabled) {
+		/* if the CRTC isn't enabled - we need to nop out the wait until */
+		ib_chunk->kdata[h_idx + 2] = PACKET2(0);
+		ib_chunk->kdata[h_idx + 3] = PACKET2(0);
+	} else if (crtc_id == 1) {
+		switch (reg) {
+		case AVIVO_D1MODE_VLINE_START_END:
+			header &= R300_CP_PACKET0_REG_MASK;
+			header |= AVIVO_D2MODE_VLINE_START_END >> 2;
+			break;
+		case RADEON_CRTC_GUI_TRIG_VLINE:
+			header &= R300_CP_PACKET0_REG_MASK;
+			header |= RADEON_CRTC2_GUI_TRIG_VLINE >> 2;
+			break;
+		default:
+			DRM_ERROR("unknown crtc reloc\n");
+			r = -EINVAL;
+			goto out;
+		}
+		ib_chunk->kdata[h_idx] = header;
+		ib_chunk->kdata[h_idx + 3] |= RADEON_ENG_DISPLAY_SELECT_CRTC1;
+	}
+out:
+	mutex_unlock(&p->rdev->ddev->mode_config.mutex);
+	return r;
+}
+
+/**
  * r100_cs_packet_next_reloc() - parse next packet which should be reloc packet3
  * @parser:		parser structure holding parsing context.
  * @data:		pointer to relocation data
@@ -814,6 +914,7 @@ static int r100_packet0_check(struct radeon_cs_parser *p,
 	unsigned idx;
 	bool onereg;
 	int r;
+	u32 tile_flags = 0;
 
 	ib = p->ib->ptr;
 	ib_chunk = &p->chunks[p->chunk_ib_idx];
@@ -825,6 +926,15 @@ static int r100_packet0_check(struct radeon_cs_parser *p,
 	}
 	for (i = 0; i <= pkt->count; i++, idx++, reg += 4) {
 		switch (reg) {
+		case RADEON_CRTC_GUI_TRIG_VLINE:
+			r = r100_cs_packet_parse_vline(p);
+			if (r) {
+				DRM_ERROR("No reloc for ib[%d]=0x%04X\n",
+						idx, reg);
+				r100_cs_dump_packet(p, pkt);
+				return r;
+			}
+			break;
 		/* FIXME: only allow PACKET3 blit? easier to check for out of
 		 * range access */
 		case RADEON_DST_PITCH_OFFSET:
@@ -838,7 +948,20 @@ static int r100_packet0_check(struct radeon_cs_parser *p,
 			}
 			tmp = ib_chunk->kdata[idx] & 0x003fffff;
 			tmp += (((u32)reloc->lobj.gpu_offset) >> 10);
-			ib[idx] = (ib_chunk->kdata[idx] & 0xffc00000) | tmp;
+
+			if (reloc->lobj.tiling_flags & RADEON_TILING_MACRO)
+				tile_flags |= RADEON_DST_TILE_MACRO;
+			if (reloc->lobj.tiling_flags & RADEON_TILING_MICRO) {
+				if (reg == RADEON_SRC_PITCH_OFFSET) {
+					DRM_ERROR("Cannot src blit from microtiled surface\n");
+					r100_cs_dump_packet(p, pkt);
+					return -EINVAL;
+				}
+				tile_flags |= RADEON_DST_TILE_MICRO;
+			}
+
+			tmp |= tile_flags;
+			ib[idx] = (ib_chunk->kdata[idx] & 0x3fc00000) | tmp;
 			break;
 		case RADEON_RB3D_DEPTHOFFSET:
 		case RADEON_RB3D_COLOROFFSET:
@@ -869,6 +992,11 @@ static int r100_packet0_check(struct radeon_cs_parser *p,
 		case R300_TX_OFFSET_0+52:
 		case R300_TX_OFFSET_0+56:
 		case R300_TX_OFFSET_0+60:
+			/* rn50 has no 3D engine so fail on any 3d setup */
+			if (ASIC_IS_RN50(p->rdev)) {
+				DRM_ERROR("attempt to use RN50 3D engine failed\n");
+				return -EINVAL;
+			}
 			r = r100_cs_packet_next_reloc(p, &reloc);
 			if (r) {
 				DRM_ERROR("No reloc for ib[%d]=0x%04X\n",
@@ -878,6 +1006,25 @@ static int r100_packet0_check(struct radeon_cs_parser *p,
 			}
 			ib[idx] = ib_chunk->kdata[idx] + ((u32)reloc->lobj.gpu_offset);
 			break;
+		case R300_RB3D_COLORPITCH0:
+		case RADEON_RB3D_COLORPITCH:
+			r = r100_cs_packet_next_reloc(p, &reloc);
+			if (r) {
+				DRM_ERROR("No reloc for ib[%d]=0x%04X\n",
+					  idx, reg);
+				r100_cs_dump_packet(p, pkt);
+				return r;
+			}
+
+			if (reloc->lobj.tiling_flags & RADEON_TILING_MACRO)
+				tile_flags |= RADEON_COLOR_TILE_ENABLE;
+			if (reloc->lobj.tiling_flags & RADEON_TILING_MICRO)
+				tile_flags |= RADEON_COLOR_MICROTILE_ENABLE;
+
+			tmp = ib_chunk->kdata[idx] & ~(0x7 << 16);
+			tmp |= tile_flags;
+			ib[idx] = tmp;
+			break;
 		default:
 			/* FIXME: we don't want to allow anyothers packet */
 			break;
@@ -1256,29 +1403,100 @@ static void r100_vram_get_type(struct radeon_device *rdev)
 	}
 }
 
-void r100_vram_info(struct radeon_device *rdev)
+static u32 r100_get_accessible_vram(struct radeon_device *rdev)
 {
-	r100_vram_get_type(rdev);
+	u32 aper_size;
+	u8 byte;
+
+	aper_size = RREG32(RADEON_CONFIG_APER_SIZE);
+
+	/* Set HDP_APER_CNTL only on cards that are known not to be broken,
+	 * that is has the 2nd generation multifunction PCI interface
+	 */
+	if (rdev->family == CHIP_RV280 ||
+	    rdev->family >= CHIP_RV350) {
+		WREG32_P(RADEON_HOST_PATH_CNTL, RADEON_HDP_APER_CNTL,
+		       ~RADEON_HDP_APER_CNTL);
+		DRM_INFO("Generation 2 PCI interface, using max accessible memory\n");
+		return aper_size * 2;
+	}
+
+	/* Older cards have all sorts of funny issues to deal with. First
+	 * check if it's a multifunction card by reading the PCI config
+	 * header type... Limit those to one aperture size
+	 */
+	pci_read_config_byte(rdev->pdev, 0xe, &byte);
+	if (byte & 0x80) {
+		DRM_INFO("Generation 1 PCI interface in multifunction mode\n");
+		DRM_INFO("Limiting VRAM to one aperture\n");
+		return aper_size;
+	}
+
+	/* Single function older card. We read HDP_APER_CNTL to see how the BIOS
+	 * have set it up. We don't write this as it's broken on some ASICs but
+	 * we expect the BIOS to have done the right thing (might be too optimistic...)
+	 */
+	if (RREG32(RADEON_HOST_PATH_CNTL) & RADEON_HDP_APER_CNTL)
+		return aper_size * 2;
+	return aper_size;
+}
+
+void r100_vram_init_sizes(struct radeon_device *rdev)
+{
+	u64 config_aper_size;
+	u32 accessible;
+
+	config_aper_size = RREG32(RADEON_CONFIG_APER_SIZE);
 
 	if (rdev->flags & RADEON_IS_IGP) {
 		uint32_t tom;
 		/* read NB_TOM to get the amount of ram stolen for the GPU */
 		tom = RREG32(RADEON_NB_TOM);
-		rdev->mc.vram_size = (((tom >> 16) - (tom & 0xffff) + 1) << 16);
-		WREG32(RADEON_CONFIG_MEMSIZE, rdev->mc.vram_size);
+		rdev->mc.real_vram_size = (((tom >> 16) - (tom & 0xffff) + 1) << 16);
+		/* for IGPs we need to keep VRAM where it was put by the BIOS */
+		rdev->mc.vram_location = (tom & 0xffff) << 16;
+		WREG32(RADEON_CONFIG_MEMSIZE, rdev->mc.real_vram_size);
+		rdev->mc.mc_vram_size = rdev->mc.real_vram_size;
 	} else {
-		rdev->mc.vram_size = RREG32(RADEON_CONFIG_MEMSIZE);
+		rdev->mc.real_vram_size = RREG32(RADEON_CONFIG_MEMSIZE);
 		/* Some production boards of m6 will report 0
 		 * if it's 8 MB
 		 */
-		if (rdev->mc.vram_size == 0) {
-			rdev->mc.vram_size = 8192 * 1024;
-			WREG32(RADEON_CONFIG_MEMSIZE, rdev->mc.vram_size);
+		if (rdev->mc.real_vram_size == 0) {
+			rdev->mc.real_vram_size = 8192 * 1024;
+			WREG32(RADEON_CONFIG_MEMSIZE, rdev->mc.real_vram_size);
 		}
+		/* let driver place VRAM */
+		rdev->mc.vram_location = 0xFFFFFFFFUL;
+		 /* Fix for RN50, M6, M7 with 8/16/32(??) MBs of VRAM - 
+		  * Novell bug 204882 + along with lots of ubuntu ones */
+		if (config_aper_size > rdev->mc.real_vram_size)
+			rdev->mc.mc_vram_size = config_aper_size;
+		else
+			rdev->mc.mc_vram_size = rdev->mc.real_vram_size;
 	}
 
+	/* work out accessible VRAM */
+	accessible = r100_get_accessible_vram(rdev);
+
 	rdev->mc.aper_base = drm_get_resource_start(rdev->ddev, 0);
 	rdev->mc.aper_size = drm_get_resource_len(rdev->ddev, 0);
+
+	if (accessible > rdev->mc.aper_size)
+		accessible = rdev->mc.aper_size;
+
+	if (rdev->mc.mc_vram_size > rdev->mc.aper_size)
+		rdev->mc.mc_vram_size = rdev->mc.aper_size;
+
+	if (rdev->mc.real_vram_size > rdev->mc.aper_size)
+		rdev->mc.real_vram_size = rdev->mc.aper_size;
+}
+
+void r100_vram_info(struct radeon_device *rdev)
+{
+	r100_vram_get_type(rdev);
+
+	r100_vram_init_sizes(rdev);
 }
 
 
@@ -1533,3 +1751,530 @@ int r100_debugfs_mc_info_init(struct radeon_device *rdev)
 	return 0;
 #endif
 }
+
+int r100_set_surface_reg(struct radeon_device *rdev, int reg,
+			 uint32_t tiling_flags, uint32_t pitch,
+			 uint32_t offset, uint32_t obj_size)
+{
+	int surf_index = reg * 16;
+	int flags = 0;
+
+	/* r100/r200 divide by 16 */
+	if (rdev->family < CHIP_R300)
+		flags = pitch / 16;
+	else
+		flags = pitch / 8;
+
+	if (rdev->family <= CHIP_RS200) {
+		if ((tiling_flags & (RADEON_TILING_MACRO|RADEON_TILING_MICRO))
+				 == (RADEON_TILING_MACRO|RADEON_TILING_MICRO))
+			flags |= RADEON_SURF_TILE_COLOR_BOTH;
+		if (tiling_flags & RADEON_TILING_MACRO)
+			flags |= RADEON_SURF_TILE_COLOR_MACRO;
+	} else if (rdev->family <= CHIP_RV280) {
+		if (tiling_flags & (RADEON_TILING_MACRO))
+			flags |= R200_SURF_TILE_COLOR_MACRO;
+		if (tiling_flags & RADEON_TILING_MICRO)
+			flags |= R200_SURF_TILE_COLOR_MICRO;
+	} else {
+		if (tiling_flags & RADEON_TILING_MACRO)
+			flags |= R300_SURF_TILE_MACRO;
+		if (tiling_flags & RADEON_TILING_MICRO)
+			flags |= R300_SURF_TILE_MICRO;
+	}
+
+	DRM_DEBUG("writing surface %d %d %x %x\n", reg, flags, offset, offset+obj_size-1);
+	WREG32(RADEON_SURFACE0_INFO + surf_index, flags);
+	WREG32(RADEON_SURFACE0_LOWER_BOUND + surf_index, offset);
+	WREG32(RADEON_SURFACE0_UPPER_BOUND + surf_index, offset + obj_size - 1);
+	return 0;
+}
+
+void r100_clear_surface_reg(struct radeon_device *rdev, int reg)
+{
+	int surf_index = reg * 16;
+	WREG32(RADEON_SURFACE0_INFO + surf_index, 0);
+}
+
+void r100_bandwidth_update(struct radeon_device *rdev)
+{
+	fixed20_12 trcd_ff, trp_ff, tras_ff, trbs_ff, tcas_ff;
+	fixed20_12 sclk_ff, mclk_ff, sclk_eff_ff, sclk_delay_ff;
+	fixed20_12 peak_disp_bw, mem_bw, pix_clk, pix_clk2, temp_ff, crit_point_ff;
+	uint32_t temp, data, mem_trcd, mem_trp, mem_tras;
+	fixed20_12 memtcas_ff[8] = {
+		fixed_init(1),
+		fixed_init(2),
+		fixed_init(3),
+		fixed_init(0),
+		fixed_init_half(1),
+		fixed_init_half(2),
+		fixed_init(0),
+	};
+	fixed20_12 memtcas_rs480_ff[8] = {
+		fixed_init(0),
+		fixed_init(1),
+		fixed_init(2),
+		fixed_init(3),
+		fixed_init(0),
+		fixed_init_half(1),
+		fixed_init_half(2),
+		fixed_init_half(3),
+	};
+	fixed20_12 memtcas2_ff[8] = {
+		fixed_init(0),
+		fixed_init(1),
+		fixed_init(2),
+		fixed_init(3),
+		fixed_init(4),
+		fixed_init(5),
+		fixed_init(6),
+		fixed_init(7),
+	};
+	fixed20_12 memtrbs[8] = {
+		fixed_init(1),
+		fixed_init_half(1),
+		fixed_init(2),
+		fixed_init_half(2),
+		fixed_init(3),
+		fixed_init_half(3),
+		fixed_init(4),
+		fixed_init_half(4)
+	};
+	fixed20_12 memtrbs_r4xx[8] = {
+		fixed_init(4),
+		fixed_init(5),
+		fixed_init(6),
+		fixed_init(7),
+		fixed_init(8),
+		fixed_init(9),
+		fixed_init(10),
+		fixed_init(11)
+	};
+	fixed20_12 min_mem_eff;
+	fixed20_12 mc_latency_sclk, mc_latency_mclk, k1;
+	fixed20_12 cur_latency_mclk, cur_latency_sclk;
+	fixed20_12 disp_latency, disp_latency_overhead, disp_drain_rate,
+		disp_drain_rate2, read_return_rate;
+	fixed20_12 time_disp1_drop_priority;
+	int c;
+	int cur_size = 16;       /* in octawords */
+	int critical_point = 0, critical_point2;
+/* 	uint32_t read_return_rate, time_disp1_drop_priority; */
+	int stop_req, max_stop_req;
+	struct drm_display_mode *mode1 = NULL;
+	struct drm_display_mode *mode2 = NULL;
+	uint32_t pixel_bytes1 = 0;
+	uint32_t pixel_bytes2 = 0;
+
+	if (rdev->mode_info.crtcs[0]->base.enabled) {
+		mode1 = &rdev->mode_info.crtcs[0]->base.mode;
+		pixel_bytes1 = rdev->mode_info.crtcs[0]->base.fb->bits_per_pixel / 8;
+	}
+	if (rdev->mode_info.crtcs[1]->base.enabled) {
+		mode2 = &rdev->mode_info.crtcs[1]->base.mode;
+		pixel_bytes2 = rdev->mode_info.crtcs[1]->base.fb->bits_per_pixel / 8;
+	}
+
+	min_mem_eff.full = rfixed_const_8(0);
+	/* get modes */
+	if ((rdev->disp_priority == 2) && ASIC_IS_R300(rdev)) {
+		uint32_t mc_init_misc_lat_timer = RREG32(R300_MC_INIT_MISC_LAT_TIMER);
+		mc_init_misc_lat_timer &= ~(R300_MC_DISP1R_INIT_LAT_MASK << R300_MC_DISP1R_INIT_LAT_SHIFT);
+		mc_init_misc_lat_timer &= ~(R300_MC_DISP0R_INIT_LAT_MASK << R300_MC_DISP0R_INIT_LAT_SHIFT);
+		/* check crtc enables */
+		if (mode2)
+			mc_init_misc_lat_timer |= (1 << R300_MC_DISP1R_INIT_LAT_SHIFT);
+		if (mode1)
+			mc_init_misc_lat_timer |= (1 << R300_MC_DISP0R_INIT_LAT_SHIFT);
+		WREG32(R300_MC_INIT_MISC_LAT_TIMER, mc_init_misc_lat_timer);
+	}
+
+	/*
+	 * determine is there is enough bw for current mode
+	 */
+	mclk_ff.full = rfixed_const(rdev->clock.default_mclk);
+	temp_ff.full = rfixed_const(100);
+	mclk_ff.full = rfixed_div(mclk_ff, temp_ff);
+	sclk_ff.full = rfixed_const(rdev->clock.default_sclk);
+	sclk_ff.full = rfixed_div(sclk_ff, temp_ff);
+
+	temp = (rdev->mc.vram_width / 8) * (rdev->mc.vram_is_ddr ? 2 : 1);
+	temp_ff.full = rfixed_const(temp);
+	mem_bw.full = rfixed_mul(mclk_ff, temp_ff);
+
+	pix_clk.full = 0;
+	pix_clk2.full = 0;
+	peak_disp_bw.full = 0;
+	if (mode1) {
+		temp_ff.full = rfixed_const(1000);
+		pix_clk.full = rfixed_const(mode1->clock); /* convert to fixed point */
+		pix_clk.full = rfixed_div(pix_clk, temp_ff);
+		temp_ff.full = rfixed_const(pixel_bytes1);
+		peak_disp_bw.full += rfixed_mul(pix_clk, temp_ff);
+	}
+	if (mode2) {
+		temp_ff.full = rfixed_const(1000);
+		pix_clk2.full = rfixed_const(mode2->clock); /* convert to fixed point */
+		pix_clk2.full = rfixed_div(pix_clk2, temp_ff);
+		temp_ff.full = rfixed_const(pixel_bytes2);
+		peak_disp_bw.full += rfixed_mul(pix_clk2, temp_ff);
+	}
+
+	mem_bw.full = rfixed_mul(mem_bw, min_mem_eff);
+	if (peak_disp_bw.full >= mem_bw.full) {
+		DRM_ERROR("You may not have enough display bandwidth for current mode\n"
+			  "If you have flickering problem, try to lower resolution, refresh rate, or color depth\n");
+	}
+
+	/*  Get values from the EXT_MEM_CNTL register...converting its contents. */
+	temp = RREG32(RADEON_MEM_TIMING_CNTL);
+	if ((rdev->family == CHIP_RV100) || (rdev->flags & RADEON_IS_IGP)) { /* RV100, M6, IGPs */
+		mem_trcd = ((temp >> 2) & 0x3) + 1;
+		mem_trp  = ((temp & 0x3)) + 1;
+		mem_tras = ((temp & 0x70) >> 4) + 1;
+	} else if (rdev->family == CHIP_R300 ||
+		   rdev->family == CHIP_R350) { /* r300, r350 */
+		mem_trcd = (temp & 0x7) + 1;
+		mem_trp = ((temp >> 8) & 0x7) + 1;
+		mem_tras = ((temp >> 11) & 0xf) + 4;
+	} else if (rdev->family == CHIP_RV350 ||
+		   rdev->family <= CHIP_RV380) {
+		/* rv3x0 */
+		mem_trcd = (temp & 0x7) + 3;
+		mem_trp = ((temp >> 8) & 0x7) + 3;
+		mem_tras = ((temp >> 11) & 0xf) + 6;
+	} else if (rdev->family == CHIP_R420 ||
+		   rdev->family == CHIP_R423 ||
+		   rdev->family == CHIP_RV410) {
+		/* r4xx */
+		mem_trcd = (temp & 0xf) + 3;
+		if (mem_trcd > 15)
+			mem_trcd = 15;
+		mem_trp = ((temp >> 8) & 0xf) + 3;
+		if (mem_trp > 15)
+			mem_trp = 15;
+		mem_tras = ((temp >> 12) & 0x1f) + 6;
+		if (mem_tras > 31)
+			mem_tras = 31;
+	} else { /* RV200, R200 */
+		mem_trcd = (temp & 0x7) + 1;
+		mem_trp = ((temp >> 8) & 0x7) + 1;
+		mem_tras = ((temp >> 12) & 0xf) + 4;
+	}
+	/* convert to FF */
+	trcd_ff.full = rfixed_const(mem_trcd);
+	trp_ff.full = rfixed_const(mem_trp);
+	tras_ff.full = rfixed_const(mem_tras);
+
+	/* Get values from the MEM_SDRAM_MODE_REG register...converting its */
+	temp = RREG32(RADEON_MEM_SDRAM_MODE_REG);
+	data = (temp & (7 << 20)) >> 20;
+	if ((rdev->family == CHIP_RV100) || rdev->flags & RADEON_IS_IGP) {
+		if (rdev->family == CHIP_RS480) /* don't think rs400 */
+			tcas_ff = memtcas_rs480_ff[data];
+		else
+			tcas_ff = memtcas_ff[data];
+	} else
+		tcas_ff = memtcas2_ff[data];
+
+	if (rdev->family == CHIP_RS400 ||
+	    rdev->family == CHIP_RS480) {
+		/* extra cas latency stored in bits 23-25 0-4 clocks */
+		data = (temp >> 23) & 0x7;
+		if (data < 5)
+			tcas_ff.full += rfixed_const(data);
+	}
+
+	if (ASIC_IS_R300(rdev) && !(rdev->flags & RADEON_IS_IGP)) {
+		/* on the R300, Tcas is included in Trbs.
+		 */
+		temp = RREG32(RADEON_MEM_CNTL);
+		data = (R300_MEM_NUM_CHANNELS_MASK & temp);
+		if (data == 1) {
+			if (R300_MEM_USE_CD_CH_ONLY & temp) {
+				temp = RREG32(R300_MC_IND_INDEX);
+				temp &= ~R300_MC_IND_ADDR_MASK;
+				temp |= R300_MC_READ_CNTL_CD_mcind;
+				WREG32(R300_MC_IND_INDEX, temp);
+				temp = RREG32(R300_MC_IND_DATA);
+				data = (R300_MEM_RBS_POSITION_C_MASK & temp);
+			} else {
+				temp = RREG32(R300_MC_READ_CNTL_AB);
+				data = (R300_MEM_RBS_POSITION_A_MASK & temp);
+			}
+		} else {
+			temp = RREG32(R300_MC_READ_CNTL_AB);
+			data = (R300_MEM_RBS_POSITION_A_MASK & temp);
+		}
+		if (rdev->family == CHIP_RV410 ||
+		    rdev->family == CHIP_R420 ||
+		    rdev->family == CHIP_R423)
+			trbs_ff = memtrbs_r4xx[data];
+		else
+			trbs_ff = memtrbs[data];
+		tcas_ff.full += trbs_ff.full;
+	}
+
+	sclk_eff_ff.full = sclk_ff.full;
+
+	if (rdev->flags & RADEON_IS_AGP) {
+		fixed20_12 agpmode_ff;
+		agpmode_ff.full = rfixed_const(radeon_agpmode);
+		temp_ff.full = rfixed_const_666(16);
+		sclk_eff_ff.full -= rfixed_mul(agpmode_ff, temp_ff);
+	}
+	/* TODO PCIE lanes may affect this - agpmode == 16?? */
+
+	if (ASIC_IS_R300(rdev)) {
+		sclk_delay_ff.full = rfixed_const(250);
+	} else {
+		if ((rdev->family == CHIP_RV100) ||
+		    rdev->flags & RADEON_IS_IGP) {
+			if (rdev->mc.vram_is_ddr)
+				sclk_delay_ff.full = rfixed_const(41);
+			else
+				sclk_delay_ff.full = rfixed_const(33);
+		} else {
+			if (rdev->mc.vram_width == 128)
+				sclk_delay_ff.full = rfixed_const(57);
+			else
+				sclk_delay_ff.full = rfixed_const(41);
+		}
+	}
+
+	mc_latency_sclk.full = rfixed_div(sclk_delay_ff, sclk_eff_ff);
+
+	if (rdev->mc.vram_is_ddr) {
+		if (rdev->mc.vram_width == 32) {
+			k1.full = rfixed_const(40);
+			c  = 3;
+		} else {
+			k1.full = rfixed_const(20);
+			c  = 1;
+		}
+	} else {
+		k1.full = rfixed_const(40);
+		c  = 3;
+	}
+
+	temp_ff.full = rfixed_const(2);
+	mc_latency_mclk.full = rfixed_mul(trcd_ff, temp_ff);
+	temp_ff.full = rfixed_const(c);
+	mc_latency_mclk.full += rfixed_mul(tcas_ff, temp_ff);
+	temp_ff.full = rfixed_const(4);
+	mc_latency_mclk.full += rfixed_mul(tras_ff, temp_ff);
+	mc_latency_mclk.full += rfixed_mul(trp_ff, temp_ff);
+	mc_latency_mclk.full += k1.full;
+
+	mc_latency_mclk.full = rfixed_div(mc_latency_mclk, mclk_ff);
+	mc_latency_mclk.full += rfixed_div(temp_ff, sclk_eff_ff);
+
+	/*
+	  HW cursor time assuming worst case of full size colour cursor.
+	*/
+	temp_ff.full = rfixed_const((2 * (cur_size - (rdev->mc.vram_is_ddr + 1))));
+	temp_ff.full += trcd_ff.full;
+	if (temp_ff.full < tras_ff.full)
+		temp_ff.full = tras_ff.full;
+	cur_latency_mclk.full = rfixed_div(temp_ff, mclk_ff);
+
+	temp_ff.full = rfixed_const(cur_size);
+	cur_latency_sclk.full = rfixed_div(temp_ff, sclk_eff_ff);
+	/*
+	  Find the total latency for the display data.
+	*/
+	disp_latency_overhead.full = rfixed_const(80);
+	disp_latency_overhead.full = rfixed_div(disp_latency_overhead, sclk_ff);
+	mc_latency_mclk.full += disp_latency_overhead.full + cur_latency_mclk.full;
+	mc_latency_sclk.full += disp_latency_overhead.full + cur_latency_sclk.full;
+
+	if (mc_latency_mclk.full > mc_latency_sclk.full)
+		disp_latency.full = mc_latency_mclk.full;
+	else
+		disp_latency.full = mc_latency_sclk.full;
+
+	/* setup Max GRPH_STOP_REQ default value */
+	if (ASIC_IS_RV100(rdev))
+		max_stop_req = 0x5c;
+	else
+		max_stop_req = 0x7c;
+
+	if (mode1) {
+		/*  CRTC1
+		    Set GRPH_BUFFER_CNTL register using h/w defined optimal values.
+		    GRPH_STOP_REQ <= MIN[ 0x7C, (CRTC_H_DISP + 1) * (bit depth) / 0x10 ]
+		*/
+		stop_req = mode1->hdisplay * pixel_bytes1 / 16;
+
+		if (stop_req > max_stop_req)
+			stop_req = max_stop_req;
+
+		/*
+		  Find the drain rate of the display buffer.
+		*/
+		temp_ff.full = rfixed_const((16/pixel_bytes1));
+		disp_drain_rate.full = rfixed_div(pix_clk, temp_ff);
+
+		/*
+		  Find the critical point of the display buffer.
+		*/
+		crit_point_ff.full = rfixed_mul(disp_drain_rate, disp_latency);
+		crit_point_ff.full += rfixed_const_half(0);
+
+		critical_point = rfixed_trunc(crit_point_ff);
+
+		if (rdev->disp_priority == 2) {
+			critical_point = 0;
+		}
+
+		/*
+		  The critical point should never be above max_stop_req-4.  Setting
+		  GRPH_CRITICAL_CNTL = 0 will thus force high priority all the time.
+		*/
+		if (max_stop_req - critical_point < 4)
+			critical_point = 0;
+
+		if (critical_point == 0 && mode2 && rdev->family == CHIP_R300) {
+			/* some R300 cards have problem with this set to 0, when CRTC2 is enabled.*/
+			critical_point = 0x10;
+		}
+
+		temp = RREG32(RADEON_GRPH_BUFFER_CNTL);
+		temp &= ~(RADEON_GRPH_STOP_REQ_MASK);
+		temp |= (stop_req << RADEON_GRPH_STOP_REQ_SHIFT);
+		temp &= ~(RADEON_GRPH_START_REQ_MASK);
+		if ((rdev->family == CHIP_R350) &&
+		    (stop_req > 0x15)) {
+			stop_req -= 0x10;
+		}
+		temp |= (stop_req << RADEON_GRPH_START_REQ_SHIFT);
+		temp |= RADEON_GRPH_BUFFER_SIZE;
+		temp &= ~(RADEON_GRPH_CRITICAL_CNTL   |
+			  RADEON_GRPH_CRITICAL_AT_SOF |
+			  RADEON_GRPH_STOP_CNTL);
+		/*
+		  Write the result into the register.
+		*/
+		WREG32(RADEON_GRPH_BUFFER_CNTL, ((temp & ~RADEON_GRPH_CRITICAL_POINT_MASK) |
+						       (critical_point << RADEON_GRPH_CRITICAL_POINT_SHIFT)));
+
+#if 0
+		if ((rdev->family == CHIP_RS400) ||
+		    (rdev->family == CHIP_RS480)) {
+			/* attempt to program RS400 disp regs correctly ??? */
+			temp = RREG32(RS400_DISP1_REG_CNTL);
+			temp &= ~(RS400_DISP1_START_REQ_LEVEL_MASK |
+				  RS400_DISP1_STOP_REQ_LEVEL_MASK);
+			WREG32(RS400_DISP1_REQ_CNTL1, (temp |
+						       (critical_point << RS400_DISP1_START_REQ_LEVEL_SHIFT) |
+						       (critical_point << RS400_DISP1_STOP_REQ_LEVEL_SHIFT)));
+			temp = RREG32(RS400_DMIF_MEM_CNTL1);
+			temp &= ~(RS400_DISP1_CRITICAL_POINT_START_MASK |
+				  RS400_DISP1_CRITICAL_POINT_STOP_MASK);
+			WREG32(RS400_DMIF_MEM_CNTL1, (temp |
+						      (critical_point << RS400_DISP1_CRITICAL_POINT_START_SHIFT) |
+						      (critical_point << RS400_DISP1_CRITICAL_POINT_STOP_SHIFT)));
+		}
+#endif
+
+		DRM_DEBUG("GRPH_BUFFER_CNTL from to %x\n",
+			  /* 	  (unsigned int)info->SavedReg->grph_buffer_cntl, */
+			  (unsigned int)RREG32(RADEON_GRPH_BUFFER_CNTL));
+	}
+
+	if (mode2) {
+		u32 grph2_cntl;
+		stop_req = mode2->hdisplay * pixel_bytes2 / 16;
+
+		if (stop_req > max_stop_req)
+			stop_req = max_stop_req;
+
+		/*
+		  Find the drain rate of the display buffer.
+		*/
+		temp_ff.full = rfixed_const((16/pixel_bytes2));
+		disp_drain_rate2.full = rfixed_div(pix_clk2, temp_ff);
+
+		grph2_cntl = RREG32(RADEON_GRPH2_BUFFER_CNTL);
+		grph2_cntl &= ~(RADEON_GRPH_STOP_REQ_MASK);
+		grph2_cntl |= (stop_req << RADEON_GRPH_STOP_REQ_SHIFT);
+		grph2_cntl &= ~(RADEON_GRPH_START_REQ_MASK);
+		if ((rdev->family == CHIP_R350) &&
+		    (stop_req > 0x15)) {
+			stop_req -= 0x10;
+		}
+		grph2_cntl |= (stop_req << RADEON_GRPH_START_REQ_SHIFT);
+		grph2_cntl |= RADEON_GRPH_BUFFER_SIZE;
+		grph2_cntl &= ~(RADEON_GRPH_CRITICAL_CNTL   |
+			  RADEON_GRPH_CRITICAL_AT_SOF |
+			  RADEON_GRPH_STOP_CNTL);
+
+		if ((rdev->family == CHIP_RS100) ||
+		    (rdev->family == CHIP_RS200))
+			critical_point2 = 0;
+		else {
+			temp = (rdev->mc.vram_width * rdev->mc.vram_is_ddr + 1)/128;
+			temp_ff.full = rfixed_const(temp);
+			temp_ff.full = rfixed_mul(mclk_ff, temp_ff);
+			if (sclk_ff.full < temp_ff.full)
+				temp_ff.full = sclk_ff.full;
+
+			read_return_rate.full = temp_ff.full;
+
+			if (mode1) {
+				temp_ff.full = read_return_rate.full - disp_drain_rate.full;
+				time_disp1_drop_priority.full = rfixed_div(crit_point_ff, temp_ff);
+			} else {
+				time_disp1_drop_priority.full = 0;
+			}
+			crit_point_ff.full = disp_latency.full + time_disp1_drop_priority.full + disp_latency.full;
+			crit_point_ff.full = rfixed_mul(crit_point_ff, disp_drain_rate2);
+			crit_point_ff.full += rfixed_const_half(0);
+
+			critical_point2 = rfixed_trunc(crit_point_ff);
+
+			if (rdev->disp_priority == 2) {
+				critical_point2 = 0;
+			}
+
+			if (max_stop_req - critical_point2 < 4)
+				critical_point2 = 0;
+
+		}
+
+		if (critical_point2 == 0 && rdev->family == CHIP_R300) {
+			/* some R300 cards have problem with this set to 0 */
+			critical_point2 = 0x10;
+		}
+
+		WREG32(RADEON_GRPH2_BUFFER_CNTL, ((grph2_cntl & ~RADEON_GRPH_CRITICAL_POINT_MASK) |
+						  (critical_point2 << RADEON_GRPH_CRITICAL_POINT_SHIFT)));
+
+		if ((rdev->family == CHIP_RS400) ||
+		    (rdev->family == CHIP_RS480)) {
+#if 0
+			/* attempt to program RS400 disp2 regs correctly ??? */
+			temp = RREG32(RS400_DISP2_REQ_CNTL1);
+			temp &= ~(RS400_DISP2_START_REQ_LEVEL_MASK |
+				  RS400_DISP2_STOP_REQ_LEVEL_MASK);
+			WREG32(RS400_DISP2_REQ_CNTL1, (temp |
+						       (critical_point2 << RS400_DISP1_START_REQ_LEVEL_SHIFT) |
+						       (critical_point2 << RS400_DISP1_STOP_REQ_LEVEL_SHIFT)));
+			temp = RREG32(RS400_DISP2_REQ_CNTL2);
+			temp &= ~(RS400_DISP2_CRITICAL_POINT_START_MASK |
+				  RS400_DISP2_CRITICAL_POINT_STOP_MASK);
+			WREG32(RS400_DISP2_REQ_CNTL2, (temp |
+						       (critical_point2 << RS400_DISP2_CRITICAL_POINT_START_SHIFT) |
+						       (critical_point2 << RS400_DISP2_CRITICAL_POINT_STOP_SHIFT)));
+#endif
+			WREG32(RS400_DISP2_REQ_CNTL1, 0x105DC1CC);
+			WREG32(RS400_DISP2_REQ_CNTL2, 0x2749D000);
+			WREG32(RS400_DMIF_MEM_CNTL1,  0x29CA71DC);
+			WREG32(RS400_DISP1_REQ_CNTL1, 0x28FBC3AC);
+		}
+
+		DRM_DEBUG("GRPH2_BUFFER_CNTL from to %x\n",
+			  (unsigned int)RREG32(RADEON_GRPH2_BUFFER_CNTL));
+	}
+}
diff --git a/drivers/gpu/drm/radeon/r300.c b/drivers/gpu/drm/radeon/r300.c
index e2ed5bc0817..9c8d41534a5 100644
--- a/drivers/gpu/drm/radeon/r300.c
+++ b/drivers/gpu/drm/radeon/r300.c
@@ -30,6 +30,8 @@
 #include "drm.h"
 #include "radeon_reg.h"
 #include "radeon.h"
+#include "radeon_drm.h"
+#include "radeon_share.h"
 
 /* r300,r350,rv350,rv370,rv380 depends on : */
 void r100_hdp_reset(struct radeon_device *rdev);
@@ -44,6 +46,7 @@ int r100_gui_wait_for_idle(struct radeon_device *rdev);
 int r100_cs_packet_parse(struct radeon_cs_parser *p,
 			 struct radeon_cs_packet *pkt,
 			 unsigned idx);
+int r100_cs_packet_parse_vline(struct radeon_cs_parser *p);
 int r100_cs_packet_next_reloc(struct radeon_cs_parser *p,
 			      struct radeon_cs_reloc **cs_reloc);
 int r100_cs_parse_packet0(struct radeon_cs_parser *p,
@@ -150,8 +153,13 @@ int rv370_pcie_gart_set_page(struct radeon_device *rdev, int i, uint64_t addr)
 	if (i < 0 || i > rdev->gart.num_gpu_pages) {
 		return -EINVAL;
 	}
-	addr = (((u32)addr) >> 8) | ((upper_32_bits(addr) & 0xff) << 4) | 0xC;
-	writel(cpu_to_le32(addr), ((void __iomem *)ptr) + (i * 4));
+	addr = (lower_32_bits(addr) >> 8) |
+	       ((upper_32_bits(addr) & 0xff) << 24) |
+	       0xc;
+	/* on x86 we want this to be CPU endian, on powerpc
+	 * on powerpc without HW swappers, it'll get swapped on way
+	 * into VRAM - so no need for cpu_to_le32 on VRAM tables */
+	writel(addr, ((void __iomem *)ptr) + (i * 4));
 	return 0;
 }
 
@@ -579,10 +587,8 @@ void r300_vram_info(struct radeon_device *rdev)
 	} else {
 		rdev->mc.vram_width = 64;
 	}
-	rdev->mc.vram_size = RREG32(RADEON_CONFIG_MEMSIZE);
 
-	rdev->mc.aper_base = drm_get_resource_start(rdev->ddev, 0);
-	rdev->mc.aper_size = drm_get_resource_len(rdev->ddev, 0);
+	r100_vram_init_sizes(rdev);
 }
 
 
@@ -970,7 +976,7 @@ static inline void r300_cs_track_clear(struct r300_cs_track *track)
 
 static const unsigned r300_reg_safe_bm[159] = {
 	0xFFFFFFFF, 0xFFFFFFFF, 0xFFFFFFFF, 0xFFFFFFFF,
-	0xFFFFFFBF, 0xFFFFFFFF, 0xFFFFFFBF, 0xFFFFFFFF,
+	0xFFFFFFFF, 0xFFFFFFFF, 0xFFFFFFFF, 0xFFFFFFFF,
 	0xFFFFFFFF, 0xFFFFFFFF, 0xFFFFFFFF, 0xFFFFFFFF,
 	0xFFFFFFFF, 0xFFFFFFFF, 0xFFFFFFFF, 0xFFFFFFFF,
 	0xFFFFFFFF, 0xFFFFFFFF, 0xFFFFFFFF, 0xFFFFFFFF,
@@ -1019,7 +1025,7 @@ static int r300_packet0_check(struct radeon_cs_parser *p,
 	struct radeon_cs_reloc *reloc;
 	struct r300_cs_track *track;
 	volatile uint32_t *ib;
-	uint32_t tmp;
+	uint32_t tmp, tile_flags = 0;
 	unsigned i;
 	int r;
 
@@ -1027,6 +1033,16 @@ static int r300_packet0_check(struct radeon_cs_parser *p,
 	ib_chunk = &p->chunks[p->chunk_ib_idx];
 	track = (struct r300_cs_track*)p->track;
 	switch(reg) {
+	case AVIVO_D1MODE_VLINE_START_END:
+	case RADEON_CRTC_GUI_TRIG_VLINE:
+		r = r100_cs_packet_parse_vline(p);
+		if (r) {
+			DRM_ERROR("No reloc for ib[%d]=0x%04X\n",
+					idx, reg);
+			r100_cs_dump_packet(p, pkt);
+			return r;
+		}
+		break;
 	case RADEON_DST_PITCH_OFFSET:
 	case RADEON_SRC_PITCH_OFFSET:
 		r = r100_cs_packet_next_reloc(p, &reloc);
@@ -1038,7 +1054,19 @@ static int r300_packet0_check(struct radeon_cs_parser *p,
 		}
 		tmp = ib_chunk->kdata[idx] & 0x003fffff;
 		tmp += (((u32)reloc->lobj.gpu_offset) >> 10);
-		ib[idx] = (ib_chunk->kdata[idx] & 0xffc00000) | tmp;
+
+		if (reloc->lobj.tiling_flags & RADEON_TILING_MACRO)
+			tile_flags |= RADEON_DST_TILE_MACRO;
+		if (reloc->lobj.tiling_flags & RADEON_TILING_MICRO) {
+			if (reg == RADEON_SRC_PITCH_OFFSET) {
+				DRM_ERROR("Cannot src blit from microtiled surface\n");
+				r100_cs_dump_packet(p, pkt);
+				return -EINVAL;
+			}
+			tile_flags |= RADEON_DST_TILE_MICRO;
+		}
+		tmp |= tile_flags;
+		ib[idx] = (ib_chunk->kdata[idx] & 0x3fc00000) | tmp;
 		break;
 	case R300_RB3D_COLOROFFSET0:
 	case R300_RB3D_COLOROFFSET1:
@@ -1127,6 +1155,23 @@ static int r300_packet0_check(struct radeon_cs_parser *p,
 		/* RB3D_COLORPITCH1 */
 		/* RB3D_COLORPITCH2 */
 		/* RB3D_COLORPITCH3 */
+		r = r100_cs_packet_next_reloc(p, &reloc);
+		if (r) {
+			DRM_ERROR("No reloc for ib[%d]=0x%04X\n",
+				  idx, reg);
+			r100_cs_dump_packet(p, pkt);
+			return r;
+		}
+
+		if (reloc->lobj.tiling_flags & RADEON_TILING_MACRO)
+			tile_flags |= R300_COLOR_TILE_ENABLE;
+		if (reloc->lobj.tiling_flags & RADEON_TILING_MICRO)
+			tile_flags |= R300_COLOR_MICROTILE_ENABLE;
+
+		tmp = ib_chunk->kdata[idx] & ~(0x7 << 16);
+		tmp |= tile_flags;
+		ib[idx] = tmp;
+
 		i = (reg - 0x4E38) >> 2;
 		track->cb[i].pitch = ib_chunk->kdata[idx] & 0x3FFE;
 		switch (((ib_chunk->kdata[idx] >> 21) & 0xF)) {
@@ -1182,6 +1227,23 @@ static int r300_packet0_check(struct radeon_cs_parser *p,
 		break;
 	case 0x4F24:
 		/* ZB_DEPTHPITCH */
+		r = r100_cs_packet_next_reloc(p, &reloc);
+		if (r) {
+			DRM_ERROR("No reloc for ib[%d]=0x%04X\n",
+				  idx, reg);
+			r100_cs_dump_packet(p, pkt);
+			return r;
+		}
+
+		if (reloc->lobj.tiling_flags & RADEON_TILING_MACRO)
+			tile_flags |= R300_DEPTHMACROTILE_ENABLE;
+		if (reloc->lobj.tiling_flags & RADEON_TILING_MICRO)
+			tile_flags |= R300_DEPTHMICROTILE_TILED;;
+
+		tmp = ib_chunk->kdata[idx] & ~(0x7 << 16);
+		tmp |= tile_flags;
+		ib[idx] = tmp;
+
 		track->zb.pitch = ib_chunk->kdata[idx] & 0x3FFC;
 		break;
 	case 0x4104:
diff --git a/drivers/gpu/drm/radeon/r300_reg.h b/drivers/gpu/drm/radeon/r300_reg.h
index 70f48609515..4b7afef35a6 100644
--- a/drivers/gpu/drm/radeon/r300_reg.h
+++ b/drivers/gpu/drm/radeon/r300_reg.h
@@ -27,7 +27,9 @@
 #ifndef _R300_REG_H_
 #define _R300_REG_H_
 
-
+#define R300_SURF_TILE_MACRO (1<<16)
+#define R300_SURF_TILE_MICRO (2<<16)
+#define R300_SURF_TILE_BOTH (3<<16)
 
 
 #define R300_MC_INIT_MISC_LAT_TIMER	0x180
diff --git a/drivers/gpu/drm/radeon/r500_reg.h b/drivers/gpu/drm/radeon/r500_reg.h
index 9070a1c2ce2..036691b38cb 100644
--- a/drivers/gpu/drm/radeon/r500_reg.h
+++ b/drivers/gpu/drm/radeon/r500_reg.h
@@ -445,6 +445,7 @@
 #define AVIVO_D1MODE_DATA_FORMAT                0x6528
 #       define AVIVO_D1MODE_INTERLEAVE_EN       (1 << 0)
 #define AVIVO_D1MODE_DESKTOP_HEIGHT             0x652C
+#define AVIVO_D1MODE_VLINE_START_END            0x6538
 #define AVIVO_D1MODE_VIEWPORT_START             0x6580
 #define AVIVO_D1MODE_VIEWPORT_SIZE              0x6584
 #define AVIVO_D1MODE_EXT_OVERSCAN_LEFT_RIGHT    0x6588
@@ -496,6 +497,7 @@
 #define AVIVO_D2CUR_SIZE                        0x6c10
 #define AVIVO_D2CUR_POSITION                    0x6c14
 
+#define AVIVO_D2MODE_VLINE_START_END            0x6d38
 #define AVIVO_D2MODE_VIEWPORT_START             0x6d80
 #define AVIVO_D2MODE_VIEWPORT_SIZE              0x6d84
 #define AVIVO_D2MODE_EXT_OVERSCAN_LEFT_RIGHT    0x6d88
diff --git a/drivers/gpu/drm/radeon/r520.c b/drivers/gpu/drm/radeon/r520.c
index 570a244bd88..09fb0b6ec7d 100644
--- a/drivers/gpu/drm/radeon/r520.c
+++ b/drivers/gpu/drm/radeon/r520.c
@@ -28,6 +28,7 @@
 #include "drmP.h"
 #include "radeon_reg.h"
 #include "radeon.h"
+#include "radeon_share.h"
 
 /* r520,rv530,rv560,rv570,r580 depends on : */
 void r100_hdp_reset(struct radeon_device *rdev);
@@ -94,8 +95,8 @@ int r520_mc_init(struct radeon_device *rdev)
 		       "programming pipes. Bad things might happen.\n");
 	}
 	/* Write VRAM size in case we are limiting it */
-	WREG32(RADEON_CONFIG_MEMSIZE, rdev->mc.vram_size);
-	tmp = rdev->mc.vram_location + rdev->mc.vram_size - 1;
+	WREG32(RADEON_CONFIG_MEMSIZE, rdev->mc.real_vram_size);
+	tmp = rdev->mc.vram_location + rdev->mc.mc_vram_size - 1;
 	tmp = REG_SET(R520_MC_FB_TOP, tmp >> 16);
 	tmp |= REG_SET(R520_MC_FB_START, rdev->mc.vram_location >> 16);
 	WREG32_MC(R520_MC_FB_LOCATION, tmp);
@@ -226,9 +227,20 @@ static void r520_vram_get_type(struct radeon_device *rdev)
 
 void r520_vram_info(struct radeon_device *rdev)
 {
+	fixed20_12 a;
+
 	r520_vram_get_type(rdev);
-	rdev->mc.vram_size = RREG32(RADEON_CONFIG_MEMSIZE);
 
-	rdev->mc.aper_base = drm_get_resource_start(rdev->ddev, 0);
-	rdev->mc.aper_size = drm_get_resource_len(rdev->ddev, 0);
+	r100_vram_init_sizes(rdev);
+	/* FIXME: we should enforce default clock in case GPU is not in
+	 * default setup
+	 */
+	a.full = rfixed_const(100);
+	rdev->pm.sclk.full = rfixed_const(rdev->clock.default_sclk);
+	rdev->pm.sclk.full = rfixed_div(rdev->pm.sclk, a);
+}
+
+void r520_bandwidth_update(struct radeon_device *rdev)
+{
+	rv515_bandwidth_avivo_update(rdev);
 }
diff --git a/drivers/gpu/drm/radeon/r600.c b/drivers/gpu/drm/radeon/r600.c
index c45559fc97f..538cd907df6 100644
--- a/drivers/gpu/drm/radeon/r600.c
+++ b/drivers/gpu/drm/radeon/r600.c
@@ -67,7 +67,7 @@ int r600_mc_init(struct radeon_device *rdev)
 		       "programming pipes. Bad things might happen.\n");
 	}
 
-	tmp = rdev->mc.vram_location + rdev->mc.vram_size - 1;
+	tmp = rdev->mc.vram_location + rdev->mc.mc_vram_size - 1;
 	tmp = REG_SET(R600_MC_FB_TOP, tmp >> 24);
 	tmp |= REG_SET(R600_MC_FB_BASE, rdev->mc.vram_location >> 24);
 	WREG32(R600_MC_VM_FB_LOCATION, tmp);
@@ -140,7 +140,8 @@ void r600_vram_get_type(struct radeon_device *rdev)
 void r600_vram_info(struct radeon_device *rdev)
 {
 	r600_vram_get_type(rdev);
-	rdev->mc.vram_size = RREG32(R600_CONFIG_MEMSIZE);
+	rdev->mc.real_vram_size = RREG32(R600_CONFIG_MEMSIZE);
+	rdev->mc.mc_vram_size = rdev->mc.real_vram_size;
 
 	/* Could aper size report 0 ? */
 	rdev->mc.aper_base = drm_get_resource_start(rdev->ddev, 0);
diff --git a/drivers/gpu/drm/radeon/r600_cp.c b/drivers/gpu/drm/radeon/r600_cp.c
index 146f3570af8..20f17908b03 100644
--- a/drivers/gpu/drm/radeon/r600_cp.c
+++ b/drivers/gpu/drm/radeon/r600_cp.c
@@ -384,8 +384,9 @@ static void r600_cp_load_microcode(drm_radeon_private_t *dev_priv)
 		DRM_INFO("Loading RV670 PFP Microcode\n");
 		for (i = 0; i < PFP_UCODE_SIZE; i++)
 			RADEON_WRITE(R600_CP_PFP_UCODE_DATA, RV670_pfp_microcode[i]);
-	} else if (((dev_priv->flags & RADEON_FAMILY_MASK) == CHIP_RS780)) {
-		DRM_INFO("Loading RS780 CP Microcode\n");
+	} else if (((dev_priv->flags & RADEON_FAMILY_MASK) == CHIP_RS780) ||
+		   ((dev_priv->flags & RADEON_FAMILY_MASK) == CHIP_RS880)) {
+		DRM_INFO("Loading RS780/RS880 CP Microcode\n");
 		for (i = 0; i < PM4_UCODE_SIZE; i++) {
 			RADEON_WRITE(R600_CP_ME_RAM_DATA,
 				     RS780_cp_microcode[i][0]);
@@ -396,7 +397,7 @@ static void r600_cp_load_microcode(drm_radeon_private_t *dev_priv)
 		}
 
 		RADEON_WRITE(R600_CP_PFP_UCODE_ADDR, 0);
-		DRM_INFO("Loading RS780 PFP Microcode\n");
+		DRM_INFO("Loading RS780/RS880 PFP Microcode\n");
 		for (i = 0; i < PFP_UCODE_SIZE; i++)
 			RADEON_WRITE(R600_CP_PFP_UCODE_DATA, RS780_pfp_microcode[i]);
 	}
@@ -783,6 +784,7 @@ static void r600_gfx_init(struct drm_device *dev,
 		break;
 	case CHIP_RV610:
 	case CHIP_RS780:
+	case CHIP_RS880:
 	case CHIP_RV620:
 		dev_priv->r600_max_pipes = 1;
 		dev_priv->r600_max_tile_pipes = 1;
@@ -917,7 +919,8 @@ static void r600_gfx_init(struct drm_device *dev,
 	    ((dev_priv->flags & RADEON_FAMILY_MASK) == CHIP_RV630) ||
 	    ((dev_priv->flags & RADEON_FAMILY_MASK) == CHIP_RV610) ||
 	    ((dev_priv->flags & RADEON_FAMILY_MASK) == CHIP_RV620) ||
-	    ((dev_priv->flags & RADEON_FAMILY_MASK) == CHIP_RS780))
+	    ((dev_priv->flags & RADEON_FAMILY_MASK) == CHIP_RS780) ||
+	    ((dev_priv->flags & RADEON_FAMILY_MASK) == CHIP_RS880))
 		RADEON_WRITE(R600_DB_DEBUG, R600_PREZ_MUST_WAIT_FOR_POSTZ_DONE);
 	else
 		RADEON_WRITE(R600_DB_DEBUG, 0);
@@ -935,7 +938,8 @@ static void r600_gfx_init(struct drm_device *dev,
 	sq_ms_fifo_sizes = RADEON_READ(R600_SQ_MS_FIFO_SIZES);
 	if (((dev_priv->flags & RADEON_FAMILY_MASK) == CHIP_RV610) ||
 	    ((dev_priv->flags & RADEON_FAMILY_MASK) == CHIP_RV620) ||
-	    ((dev_priv->flags & RADEON_FAMILY_MASK) == CHIP_RS780)) {
+	    ((dev_priv->flags & RADEON_FAMILY_MASK) == CHIP_RS780) ||
+	    ((dev_priv->flags & RADEON_FAMILY_MASK) == CHIP_RS880)) {
 		sq_ms_fifo_sizes = (R600_CACHE_FIFO_SIZE(0xa) |
 				    R600_FETCH_FIFO_HIWATER(0xa) |
 				    R600_DONE_FIFO_HIWATER(0xe0) |
@@ -978,7 +982,8 @@ static void r600_gfx_init(struct drm_device *dev,
 					    R600_NUM_ES_STACK_ENTRIES(0));
 	} else if (((dev_priv->flags & RADEON_FAMILY_MASK) == CHIP_RV610) ||
 		   ((dev_priv->flags & RADEON_FAMILY_MASK) == CHIP_RV620) ||
-		   ((dev_priv->flags & RADEON_FAMILY_MASK) == CHIP_RS780)) {
+		   ((dev_priv->flags & RADEON_FAMILY_MASK) == CHIP_RS780) ||
+		   ((dev_priv->flags & RADEON_FAMILY_MASK) == CHIP_RS880)) {
 		/* no vertex cache */
 		sq_config &= ~R600_VC_ENABLE;
 
@@ -1035,7 +1040,8 @@ static void r600_gfx_init(struct drm_device *dev,
 
 	if (((dev_priv->flags & RADEON_FAMILY_MASK) == CHIP_RV610) ||
 	    ((dev_priv->flags & RADEON_FAMILY_MASK) == CHIP_RV620) ||
-	    ((dev_priv->flags & RADEON_FAMILY_MASK) == CHIP_RS780))
+	    ((dev_priv->flags & RADEON_FAMILY_MASK) == CHIP_RS780) ||
+	    ((dev_priv->flags & RADEON_FAMILY_MASK) == CHIP_RS880))
 		RADEON_WRITE(R600_VGT_CACHE_INVALIDATION, R600_CACHE_INVALIDATION(R600_TC_ONLY));
 	else
 		RADEON_WRITE(R600_VGT_CACHE_INVALIDATION, R600_CACHE_INVALIDATION(R600_VC_AND_TC));
@@ -1078,6 +1084,7 @@ static void r600_gfx_init(struct drm_device *dev,
 		break;
 	case CHIP_RV610:
 	case CHIP_RS780:
+	case CHIP_RS880:
 	case CHIP_RV620:
 		gs_prim_buffer_depth = 32;
 		break;
@@ -1123,6 +1130,7 @@ static void r600_gfx_init(struct drm_device *dev,
 	switch (dev_priv->flags & RADEON_FAMILY_MASK) {
 	case CHIP_RV610:
 	case CHIP_RS780:
+	case CHIP_RS880:
 	case CHIP_RV620:
 		tc_cntl = R600_TC_L2_SIZE(8);
 		break;
diff --git a/drivers/gpu/drm/radeon/radeon.h b/drivers/gpu/drm/radeon/radeon.h
index d61f2fc61df..b1d945b8ed6 100644
--- a/drivers/gpu/drm/radeon/radeon.h
+++ b/drivers/gpu/drm/radeon/radeon.h
@@ -64,6 +64,7 @@ extern int radeon_agpmode;
 extern int radeon_vram_limit;
 extern int radeon_gart_size;
 extern int radeon_benchmarking;
+extern int radeon_testing;
 extern int radeon_connector_table;
 
 /*
@@ -113,6 +114,7 @@ enum radeon_family {
 	CHIP_RV770,
 	CHIP_RV730,
 	CHIP_RV710,
+	CHIP_RS880,
 	CHIP_LAST,
 };
 
@@ -201,6 +203,14 @@ int radeon_fence_wait_last(struct radeon_device *rdev);
 struct radeon_fence *radeon_fence_ref(struct radeon_fence *fence);
 void radeon_fence_unref(struct radeon_fence **fence);
 
+/*
+ * Tiling registers
+ */
+struct radeon_surface_reg {
+	struct radeon_object *robj;
+};
+
+#define RADEON_GEM_MAX_SURFACES 8
 
 /*
  * Radeon buffer.
@@ -213,6 +223,7 @@ struct radeon_object_list {
 	uint64_t		gpu_offset;
 	unsigned		rdomain;
 	unsigned		wdomain;
+	uint32_t                tiling_flags;
 };
 
 int radeon_object_init(struct radeon_device *rdev);
@@ -242,8 +253,15 @@ void radeon_object_list_clean(struct list_head *head);
 int radeon_object_fbdev_mmap(struct radeon_object *robj,
 			     struct vm_area_struct *vma);
 unsigned long radeon_object_size(struct radeon_object *robj);
-
-
+void radeon_object_clear_surface_reg(struct radeon_object *robj);
+int radeon_object_check_tiling(struct radeon_object *robj, bool has_moved,
+			       bool force_drop);
+void radeon_object_set_tiling_flags(struct radeon_object *robj,
+				    uint32_t tiling_flags, uint32_t pitch);
+void radeon_object_get_tiling_flags(struct radeon_object *robj, uint32_t *tiling_flags, uint32_t *pitch);
+void radeon_bo_move_notify(struct ttm_buffer_object *bo,
+			   struct ttm_mem_reg *mem);
+void radeon_bo_fault_reserve_notify(struct ttm_buffer_object *bo);
 /*
  * GEM objects.
  */
@@ -315,8 +333,11 @@ struct radeon_mc {
 	unsigned		gtt_location;
 	unsigned		gtt_size;
 	unsigned		vram_location;
-	unsigned		vram_size;
+	/* for some chips with <= 32MB we need to lie
+	 * about vram size near mc fb location */
+	unsigned		mc_vram_size;
 	unsigned		vram_width;
+	unsigned		real_vram_size;
 	int			vram_mtrr;
 	bool			vram_is_ddr;
 };
@@ -474,6 +495,39 @@ struct radeon_wb {
 	uint64_t		gpu_addr;
 };
 
+/**
+ * struct radeon_pm - power management datas
+ * @max_bandwidth:      maximum bandwidth the gpu has (MByte/s)
+ * @igp_sideport_mclk:  sideport memory clock Mhz (rs690,rs740,rs780,rs880)
+ * @igp_system_mclk:    system clock Mhz (rs690,rs740,rs780,rs880)
+ * @igp_ht_link_clk:    ht link clock Mhz (rs690,rs740,rs780,rs880)
+ * @igp_ht_link_width:  ht link width in bits (rs690,rs740,rs780,rs880)
+ * @k8_bandwidth:       k8 bandwidth the gpu has (MByte/s) (IGP)
+ * @sideport_bandwidth: sideport bandwidth the gpu has (MByte/s) (IGP)
+ * @ht_bandwidth:       ht bandwidth the gpu has (MByte/s) (IGP)
+ * @core_bandwidth:     core GPU bandwidth the gpu has (MByte/s) (IGP)
+ * @sclk:          	GPU clock Mhz (core bandwith depends of this clock)
+ * @needed_bandwidth:   current bandwidth needs
+ *
+ * It keeps track of various data needed to take powermanagement decision.
+ * Bandwith need is used to determine minimun clock of the GPU and memory.
+ * Equation between gpu/memory clock and available bandwidth is hw dependent
+ * (type of memory, bus size, efficiency, ...)
+ */
+struct radeon_pm {
+	fixed20_12		max_bandwidth;
+	fixed20_12		igp_sideport_mclk;
+	fixed20_12		igp_system_mclk;
+	fixed20_12		igp_ht_link_clk;
+	fixed20_12		igp_ht_link_width;
+	fixed20_12		k8_bandwidth;
+	fixed20_12		sideport_bandwidth;
+	fixed20_12		ht_bandwidth;
+	fixed20_12		core_bandwidth;
+	fixed20_12		sclk;
+	fixed20_12		needed_bandwidth;
+};
+
 
 /*
  * Benchmarking
@@ -482,6 +536,12 @@ void radeon_benchmark(struct radeon_device *rdev);
 
 
 /*
+ * Testing
+ */
+void radeon_test_moves(struct radeon_device *rdev);
+
+
+/*
  * Debugfs
  */
 int radeon_debugfs_add_files(struct radeon_device *rdev,
@@ -535,6 +595,11 @@ struct radeon_asic {
 	void (*set_memory_clock)(struct radeon_device *rdev, uint32_t mem_clock);
 	void (*set_pcie_lanes)(struct radeon_device *rdev, int lanes);
 	void (*set_clock_gating)(struct radeon_device *rdev, int enable);
+	int (*set_surface_reg)(struct radeon_device *rdev, int reg,
+			       uint32_t tiling_flags, uint32_t pitch,
+			       uint32_t offset, uint32_t obj_size);
+	int (*clear_surface_reg)(struct radeon_device *rdev, int reg);
+	void (*bandwidth_update)(struct radeon_device *rdev);
 };
 
 union radeon_asic_config {
@@ -566,6 +631,10 @@ int radeon_gem_busy_ioctl(struct drm_device *dev, void *data,
 int radeon_gem_wait_idle_ioctl(struct drm_device *dev, void *data,
 			      struct drm_file *filp);
 int radeon_cs_ioctl(struct drm_device *dev, void *data, struct drm_file *filp);
+int radeon_gem_set_tiling_ioctl(struct drm_device *dev, void *data,
+				struct drm_file *filp);
+int radeon_gem_get_tiling_ioctl(struct drm_device *dev, void *data,
+				struct drm_file *filp);
 
 
 /*
@@ -594,8 +663,8 @@ struct radeon_device {
 	struct radeon_object		*fbdev_robj;
 	struct radeon_framebuffer	*fbdev_rfb;
 	/* Register mmio */
-	unsigned long			rmmio_base;
-	unsigned long			rmmio_size;
+	resource_size_t			rmmio_base;
+	resource_size_t			rmmio_size;
 	void				*rmmio;
 	radeon_rreg_t			mm_rreg;
 	radeon_wreg_t			mm_wreg;
@@ -619,11 +688,14 @@ struct radeon_device {
 	struct radeon_irq		irq;
 	struct radeon_asic		*asic;
 	struct radeon_gem		gem;
+	struct radeon_pm		pm;
 	struct mutex			cs_mutex;
 	struct radeon_wb		wb;
 	bool				gpu_lockup;
 	bool				shutdown;
 	bool				suspend;
+	bool				need_dma32;
+	struct radeon_surface_reg surface_regs[RADEON_GEM_MAX_SURFACES];
 };
 
 int radeon_device_init(struct radeon_device *rdev,
@@ -670,6 +742,8 @@ void r100_pll_errata_after_index(struct radeon_device *rdev);
 /*
  * ASICs helpers.
  */
+#define ASIC_IS_RN50(rdev) ((rdev->pdev->device == 0x515e) || \
+			    (rdev->pdev->device == 0x5969))
 #define ASIC_IS_RV100(rdev) ((rdev->family == CHIP_RV100) || \
 		(rdev->family == CHIP_RV200) || \
 		(rdev->family == CHIP_RS100) || \
@@ -796,5 +870,8 @@ static inline void radeon_ring_write(struct radeon_device *rdev, uint32_t v)
 #define radeon_set_memory_clock(rdev, e) (rdev)->asic->set_engine_clock((rdev), (e))
 #define radeon_set_pcie_lanes(rdev, l) (rdev)->asic->set_pcie_lanes((rdev), (l))
 #define radeon_set_clock_gating(rdev, e) (rdev)->asic->set_clock_gating((rdev), (e))
+#define radeon_set_surface_reg(rdev, r, f, p, o, s) ((rdev)->asic->set_surface_reg((rdev), (r), (f), (p), (o), (s)))
+#define radeon_clear_surface_reg(rdev, r) ((rdev)->asic->clear_surface_reg((rdev), (r)))
+#define radeon_bandwidth_update(rdev) (rdev)->asic->bandwidth_update((rdev))
 
 #endif
diff --git a/drivers/gpu/drm/radeon/radeon_asic.h b/drivers/gpu/drm/radeon/radeon_asic.h
index e2e567395df..9a75876e0c3 100644
--- a/drivers/gpu/drm/radeon/radeon_asic.h
+++ b/drivers/gpu/drm/radeon/radeon_asic.h
@@ -71,6 +71,11 @@ int r100_copy_blit(struct radeon_device *rdev,
 		   uint64_t dst_offset,
 		   unsigned num_pages,
 		   struct radeon_fence *fence);
+int r100_set_surface_reg(struct radeon_device *rdev, int reg,
+			 uint32_t tiling_flags, uint32_t pitch,
+			 uint32_t offset, uint32_t obj_size);
+int r100_clear_surface_reg(struct radeon_device *rdev, int reg);
+void r100_bandwidth_update(struct radeon_device *rdev);
 
 static struct radeon_asic r100_asic = {
 	.init = &r100_init,
@@ -100,6 +105,9 @@ static struct radeon_asic r100_asic = {
 	.set_memory_clock = NULL,
 	.set_pcie_lanes = NULL,
 	.set_clock_gating = &radeon_legacy_set_clock_gating,
+	.set_surface_reg = r100_set_surface_reg,
+	.clear_surface_reg = r100_clear_surface_reg,
+	.bandwidth_update = &r100_bandwidth_update,
 };
 
 
@@ -128,6 +136,7 @@ int r300_copy_dma(struct radeon_device *rdev,
 		  uint64_t dst_offset,
 		  unsigned num_pages,
 		  struct radeon_fence *fence);
+
 static struct radeon_asic r300_asic = {
 	.init = &r300_init,
 	.errata = &r300_errata,
@@ -156,6 +165,9 @@ static struct radeon_asic r300_asic = {
 	.set_memory_clock = NULL,
 	.set_pcie_lanes = &rv370_set_pcie_lanes,
 	.set_clock_gating = &radeon_legacy_set_clock_gating,
+	.set_surface_reg = r100_set_surface_reg,
+	.clear_surface_reg = r100_clear_surface_reg,
+	.bandwidth_update = &r100_bandwidth_update,
 };
 
 /*
@@ -193,6 +205,9 @@ static struct radeon_asic r420_asic = {
 	.set_memory_clock = &radeon_atom_set_memory_clock,
 	.set_pcie_lanes = &rv370_set_pcie_lanes,
 	.set_clock_gating = &radeon_atom_set_clock_gating,
+	.set_surface_reg = r100_set_surface_reg,
+	.clear_surface_reg = r100_clear_surface_reg,
+	.bandwidth_update = &r100_bandwidth_update,
 };
 
 
@@ -237,6 +252,9 @@ static struct radeon_asic rs400_asic = {
 	.set_memory_clock = NULL,
 	.set_pcie_lanes = NULL,
 	.set_clock_gating = &radeon_legacy_set_clock_gating,
+	.set_surface_reg = r100_set_surface_reg,
+	.clear_surface_reg = r100_clear_surface_reg,
+	.bandwidth_update = &r100_bandwidth_update,
 };
 
 
@@ -254,6 +272,7 @@ void rs600_gart_tlb_flush(struct radeon_device *rdev);
 int rs600_gart_set_page(struct radeon_device *rdev, int i, uint64_t addr);
 uint32_t rs600_mc_rreg(struct radeon_device *rdev, uint32_t reg);
 void rs600_mc_wreg(struct radeon_device *rdev, uint32_t reg, uint32_t v);
+void rs600_bandwidth_update(struct radeon_device *rdev);
 static struct radeon_asic rs600_asic = {
 	.init = &r300_init,
 	.errata = &rs600_errata,
@@ -282,6 +301,7 @@ static struct radeon_asic rs600_asic = {
 	.set_memory_clock = &radeon_atom_set_memory_clock,
 	.set_pcie_lanes = NULL,
 	.set_clock_gating = &radeon_atom_set_clock_gating,
+	.bandwidth_update = &rs600_bandwidth_update,
 };
 
 
@@ -294,6 +314,7 @@ int rs690_mc_init(struct radeon_device *rdev);
 void rs690_mc_fini(struct radeon_device *rdev);
 uint32_t rs690_mc_rreg(struct radeon_device *rdev, uint32_t reg);
 void rs690_mc_wreg(struct radeon_device *rdev, uint32_t reg, uint32_t v);
+void rs690_bandwidth_update(struct radeon_device *rdev);
 static struct radeon_asic rs690_asic = {
 	.init = &r300_init,
 	.errata = &rs690_errata,
@@ -322,6 +343,9 @@ static struct radeon_asic rs690_asic = {
 	.set_memory_clock = &radeon_atom_set_memory_clock,
 	.set_pcie_lanes = NULL,
 	.set_clock_gating = &radeon_atom_set_clock_gating,
+	.set_surface_reg = r100_set_surface_reg,
+	.clear_surface_reg = r100_clear_surface_reg,
+	.bandwidth_update = &rs690_bandwidth_update,
 };
 
 
@@ -339,6 +363,7 @@ void rv515_mc_wreg(struct radeon_device *rdev, uint32_t reg, uint32_t v);
 void rv515_ring_start(struct radeon_device *rdev);
 uint32_t rv515_pcie_rreg(struct radeon_device *rdev, uint32_t reg);
 void rv515_pcie_wreg(struct radeon_device *rdev, uint32_t reg, uint32_t v);
+void rv515_bandwidth_update(struct radeon_device *rdev);
 static struct radeon_asic rv515_asic = {
 	.init = &rv515_init,
 	.errata = &rv515_errata,
@@ -367,6 +392,9 @@ static struct radeon_asic rv515_asic = {
 	.set_memory_clock = &radeon_atom_set_memory_clock,
 	.set_pcie_lanes = &rv370_set_pcie_lanes,
 	.set_clock_gating = &radeon_atom_set_clock_gating,
+	.set_surface_reg = r100_set_surface_reg,
+	.clear_surface_reg = r100_clear_surface_reg,
+	.bandwidth_update = &rv515_bandwidth_update,
 };
 
 
@@ -377,6 +405,7 @@ void r520_errata(struct radeon_device *rdev);
 void r520_vram_info(struct radeon_device *rdev);
 int r520_mc_init(struct radeon_device *rdev);
 void r520_mc_fini(struct radeon_device *rdev);
+void r520_bandwidth_update(struct radeon_device *rdev);
 static struct radeon_asic r520_asic = {
 	.init = &rv515_init,
 	.errata = &r520_errata,
@@ -405,6 +434,9 @@ static struct radeon_asic r520_asic = {
 	.set_memory_clock = &radeon_atom_set_memory_clock,
 	.set_pcie_lanes = &rv370_set_pcie_lanes,
 	.set_clock_gating = &radeon_atom_set_clock_gating,
+	.set_surface_reg = r100_set_surface_reg,
+	.clear_surface_reg = r100_clear_surface_reg,
+	.bandwidth_update = &r520_bandwidth_update,
 };
 
 /*
diff --git a/drivers/gpu/drm/radeon/radeon_atombios.c b/drivers/gpu/drm/radeon/radeon_atombios.c
index 1f5a1a49098..fcfe5c02d74 100644
--- a/drivers/gpu/drm/radeon/radeon_atombios.c
+++ b/drivers/gpu/drm/radeon/radeon_atombios.c
@@ -103,7 +103,8 @@ static inline struct radeon_i2c_bus_rec radeon_lookup_gpio(struct drm_device
 static bool radeon_atom_apply_quirks(struct drm_device *dev,
 				     uint32_t supported_device,
 				     int *connector_type,
-				     struct radeon_i2c_bus_rec *i2c_bus)
+				     struct radeon_i2c_bus_rec *i2c_bus,
+				     uint8_t *line_mux)
 {
 
 	/* Asus M2A-VM HDMI board lists the DVI port as HDMI */
@@ -127,8 +128,10 @@ static bool radeon_atom_apply_quirks(struct drm_device *dev,
 	if ((dev->pdev->device == 0x5653) &&
 	    (dev->pdev->subsystem_vendor == 0x1462) &&
 	    (dev->pdev->subsystem_device == 0x0291)) {
-		if (*connector_type == DRM_MODE_CONNECTOR_LVDS)
+		if (*connector_type == DRM_MODE_CONNECTOR_LVDS) {
 			i2c_bus->valid = false;
+			*line_mux = 53;
+		}
 	}
 
 	/* Funky macbooks */
@@ -526,7 +529,7 @@ bool radeon_get_atom_connector_info_from_supported_devices_table(struct
 
 		if (!radeon_atom_apply_quirks
 		    (dev, (1 << i), &bios_connectors[i].connector_type,
-		     &bios_connectors[i].ddc_bus))
+		     &bios_connectors[i].ddc_bus, &bios_connectors[i].line_mux))
 			continue;
 
 		bios_connectors[i].valid = true;
diff --git a/drivers/gpu/drm/radeon/radeon_benchmark.c b/drivers/gpu/drm/radeon/radeon_benchmark.c
index c44403a2ca7..2e938f7496f 100644
--- a/drivers/gpu/drm/radeon/radeon_benchmark.c
+++ b/drivers/gpu/drm/radeon/radeon_benchmark.c
@@ -63,7 +63,7 @@ void radeon_benchmark_move(struct radeon_device *rdev, unsigned bsize,
 		if (r) {
 			goto out_cleanup;
 		}
-		r = radeon_copy_dma(rdev, saddr, daddr, size >> 14, fence);
+		r = radeon_copy_dma(rdev, saddr, daddr, size / 4096, fence);
 		if (r) {
 			goto out_cleanup;
 		}
@@ -88,7 +88,7 @@ void radeon_benchmark_move(struct radeon_device *rdev, unsigned bsize,
 		if (r) {
 			goto out_cleanup;
 		}
-		r = radeon_copy_blit(rdev, saddr, daddr, size >> 14, fence);
+		r = radeon_copy_blit(rdev, saddr, daddr, size / 4096, fence);
 		if (r) {
 			goto out_cleanup;
 		}
diff --git a/drivers/gpu/drm/radeon/radeon_cs.c b/drivers/gpu/drm/radeon/radeon_cs.c
index b843f9bdfb1..a169067efc4 100644
--- a/drivers/gpu/drm/radeon/radeon_cs.c
+++ b/drivers/gpu/drm/radeon/radeon_cs.c
@@ -127,17 +127,23 @@ int radeon_cs_parser_init(struct radeon_cs_parser *p, void *data)
 				       sizeof(struct drm_radeon_cs_chunk))) {
 			return -EFAULT;
 		}
+		p->chunks[i].length_dw = user_chunk.length_dw;
+		p->chunks[i].kdata = NULL;
 		p->chunks[i].chunk_id = user_chunk.chunk_id;
+
 		if (p->chunks[i].chunk_id == RADEON_CHUNK_ID_RELOCS) {
 			p->chunk_relocs_idx = i;
 		}
 		if (p->chunks[i].chunk_id == RADEON_CHUNK_ID_IB) {
 			p->chunk_ib_idx = i;
+			/* zero length IB isn't useful */
+			if (p->chunks[i].length_dw == 0)
+				return -EINVAL;
 		}
+
 		p->chunks[i].length_dw = user_chunk.length_dw;
 		cdata = (uint32_t *)(unsigned long)user_chunk.chunk_data;
 
-		p->chunks[i].kdata = NULL;
 		size = p->chunks[i].length_dw * sizeof(uint32_t);
 		p->chunks[i].kdata = kzalloc(size, GFP_KERNEL);
 		if (p->chunks[i].kdata == NULL) {
diff --git a/drivers/gpu/drm/radeon/radeon_cursor.c b/drivers/gpu/drm/radeon/radeon_cursor.c
index 5232441f119..b13c79e38bc 100644
--- a/drivers/gpu/drm/radeon/radeon_cursor.c
+++ b/drivers/gpu/drm/radeon/radeon_cursor.c
@@ -111,9 +111,11 @@ static void radeon_set_cursor(struct drm_crtc *crtc, struct drm_gem_object *obj,
 
 	if (ASIC_IS_AVIVO(rdev))
 		WREG32(AVIVO_D1CUR_SURFACE_ADDRESS + radeon_crtc->crtc_offset, gpu_addr);
-	else
+	else {
+		radeon_crtc->legacy_cursor_offset = gpu_addr - radeon_crtc->legacy_display_base_addr;
 		/* offset is from DISP(2)_BASE_ADDRESS */
-		WREG32(RADEON_CUR_OFFSET + radeon_crtc->crtc_offset, gpu_addr);
+		WREG32(RADEON_CUR_OFFSET + radeon_crtc->crtc_offset, radeon_crtc->legacy_cursor_offset);
+	}
 }
 
 int radeon_crtc_cursor_set(struct drm_crtc *crtc,
@@ -245,6 +247,9 @@ int radeon_crtc_cursor_move(struct drm_crtc *crtc,
 		       (RADEON_CUR_LOCK
 			| ((xorigin ? 0 : x) << 16)
 			| (yorigin ? 0 : y)));
+		/* offset is from DISP(2)_BASE_ADDRESS */
+		WREG32(RADEON_CUR_OFFSET + radeon_crtc->crtc_offset, (radeon_crtc->legacy_cursor_offset +
+								      (yorigin * 256)));
 	}
 	radeon_lock_cursor(crtc, false);
 
diff --git a/drivers/gpu/drm/radeon/radeon_device.c b/drivers/gpu/drm/radeon/radeon_device.c
index f97563db4e5..9ff6dcb97f9 100644
--- a/drivers/gpu/drm/radeon/radeon_device.c
+++ b/drivers/gpu/drm/radeon/radeon_device.c
@@ -48,6 +48,8 @@ static void radeon_surface_init(struct radeon_device *rdev)
 			       i * (RADEON_SURFACE1_INFO - RADEON_SURFACE0_INFO),
 			       0);
 		}
+		/* enable surfaces */
+		WREG32(RADEON_SURFACE_CNTL, 0);
 	}
 }
 
@@ -119,7 +121,7 @@ int radeon_mc_setup(struct radeon_device *rdev)
 	if (rdev->mc.vram_location != 0xFFFFFFFFUL) {
 		/* vram location was already setup try to put gtt after
 		 * if it fits */
-		tmp = rdev->mc.vram_location + rdev->mc.vram_size;
+		tmp = rdev->mc.vram_location + rdev->mc.mc_vram_size;
 		tmp = (tmp + rdev->mc.gtt_size - 1) & ~(rdev->mc.gtt_size - 1);
 		if ((0xFFFFFFFFUL - tmp) >= rdev->mc.gtt_size) {
 			rdev->mc.gtt_location = tmp;
@@ -134,13 +136,13 @@ int radeon_mc_setup(struct radeon_device *rdev)
 	} else if (rdev->mc.gtt_location != 0xFFFFFFFFUL) {
 		/* gtt location was already setup try to put vram before
 		 * if it fits */
-		if (rdev->mc.vram_size < rdev->mc.gtt_location) {
+		if (rdev->mc.mc_vram_size < rdev->mc.gtt_location) {
 			rdev->mc.vram_location = 0;
 		} else {
 			tmp = rdev->mc.gtt_location + rdev->mc.gtt_size;
-			tmp += (rdev->mc.vram_size - 1);
-			tmp &= ~(rdev->mc.vram_size - 1);
-			if ((0xFFFFFFFFUL - tmp) >= rdev->mc.vram_size) {
+			tmp += (rdev->mc.mc_vram_size - 1);
+			tmp &= ~(rdev->mc.mc_vram_size - 1);
+			if ((0xFFFFFFFFUL - tmp) >= rdev->mc.mc_vram_size) {
 				rdev->mc.vram_location = tmp;
 			} else {
 				printk(KERN_ERR "[drm] vram too big to fit "
@@ -150,12 +152,16 @@ int radeon_mc_setup(struct radeon_device *rdev)
 		}
 	} else {
 		rdev->mc.vram_location = 0;
-		rdev->mc.gtt_location = rdev->mc.vram_size;
+		tmp = rdev->mc.mc_vram_size;
+		tmp = (tmp + rdev->mc.gtt_size - 1) & ~(rdev->mc.gtt_size - 1);
+		rdev->mc.gtt_location = tmp;
 	}
-	DRM_INFO("radeon: VRAM %uM\n", rdev->mc.vram_size >> 20);
+	DRM_INFO("radeon: VRAM %uM\n", rdev->mc.real_vram_size >> 20);
 	DRM_INFO("radeon: VRAM from 0x%08X to 0x%08X\n",
 		 rdev->mc.vram_location,
-		 rdev->mc.vram_location + rdev->mc.vram_size - 1);
+		 rdev->mc.vram_location + rdev->mc.mc_vram_size - 1);
+	if (rdev->mc.real_vram_size != rdev->mc.mc_vram_size)
+		DRM_INFO("radeon: VRAM less than aperture workaround enabled\n");
 	DRM_INFO("radeon: GTT %uM\n", rdev->mc.gtt_size >> 20);
 	DRM_INFO("radeon: GTT from 0x%08X to 0x%08X\n",
 		 rdev->mc.gtt_location,
@@ -450,6 +456,7 @@ int radeon_device_init(struct radeon_device *rdev,
 		       uint32_t flags)
 {
 	int r, ret;
+	int dma_bits;
 
 	DRM_INFO("radeon: Initializing kernel modesetting.\n");
 	rdev->shutdown = false;
@@ -492,8 +499,20 @@ int radeon_device_init(struct radeon_device *rdev,
 		return r;
 	}
 
-	/* Report DMA addressing limitation */
-	r = pci_set_dma_mask(rdev->pdev, DMA_BIT_MASK(32));
+	/* set DMA mask + need_dma32 flags.
+	 * PCIE - can handle 40-bits.
+	 * IGP - can handle 40-bits (in theory)
+	 * AGP - generally dma32 is safest
+	 * PCI - only dma32
+	 */
+	rdev->need_dma32 = false;
+	if (rdev->flags & RADEON_IS_AGP)
+		rdev->need_dma32 = true;
+	if (rdev->flags & RADEON_IS_PCI)
+		rdev->need_dma32 = true;
+
+	dma_bits = rdev->need_dma32 ? 32 : 40;
+	r = pci_set_dma_mask(rdev->pdev, DMA_BIT_MASK(dma_bits));
 	if (r) {
 		printk(KERN_WARNING "radeon: No suitable DMA available.\n");
 	}
@@ -546,27 +565,22 @@ int radeon_device_init(struct radeon_device *rdev,
 			radeon_combios_asic_init(rdev->ddev);
 		}
 	}
+	/* Initialize clocks */
+	r = radeon_clocks_init(rdev);
+	if (r) {
+		return r;
+	}
 	/* Get vram informations */
 	radeon_vram_info(rdev);
-	/* Device is severly broken if aper size > vram size.
-	 * for RN50/M6/M7 - Novell bug 204882 ?
-	 */
-	if (rdev->mc.vram_size < rdev->mc.aper_size) {
-		rdev->mc.aper_size = rdev->mc.vram_size;
-	}
+
 	/* Add an MTRR for the VRAM */
 	rdev->mc.vram_mtrr = mtrr_add(rdev->mc.aper_base, rdev->mc.aper_size,
 				      MTRR_TYPE_WRCOMB, 1);
 	DRM_INFO("Detected VRAM RAM=%uM, BAR=%uM\n",
-		 rdev->mc.vram_size >> 20,
+		 rdev->mc.real_vram_size >> 20,
 		 (unsigned)rdev->mc.aper_size >> 20);
 	DRM_INFO("RAM width %dbits %cDR\n",
 		 rdev->mc.vram_width, rdev->mc.vram_is_ddr ? 'D' : 'S');
-	/* Initialize clocks */
-	r = radeon_clocks_init(rdev);
-	if (r) {
-		return r;
-	}
 	/* Initialize memory controller (also test AGP) */
 	r = radeon_mc_init(rdev);
 	if (r) {
@@ -626,6 +640,9 @@ int radeon_device_init(struct radeon_device *rdev,
 	if (!ret) {
 		DRM_INFO("radeon: kernel modesetting successfully initialized.\n");
 	}
+	if (radeon_testing) {
+		radeon_test_moves(rdev);
+	}
 	if (radeon_benchmarking) {
 		radeon_benchmark(rdev);
 	}
diff --git a/drivers/gpu/drm/radeon/radeon_display.c b/drivers/gpu/drm/radeon/radeon_display.c
index 3efcf1a526b..a8fa1bb84cf 100644
--- a/drivers/gpu/drm/radeon/radeon_display.c
+++ b/drivers/gpu/drm/radeon/radeon_display.c
@@ -187,6 +187,7 @@ static void radeon_crtc_init(struct drm_device *dev, int index)
 
 	drm_mode_crtc_set_gamma_size(&radeon_crtc->base, 256);
 	radeon_crtc->crtc_id = index;
+	rdev->mode_info.crtcs[index] = radeon_crtc;
 
 	radeon_crtc->mode_set.crtc = &radeon_crtc->base;
 	radeon_crtc->mode_set.connectors = (struct drm_connector **)(radeon_crtc + 1);
@@ -491,7 +492,11 @@ void radeon_compute_pll(struct radeon_pll *pll,
 					tmp += (uint64_t)pll->reference_freq * 1000 * frac_feedback_div;
 					current_freq = radeon_div(tmp, ref_div * post_div);
 
-					error = abs(current_freq - freq);
+					if (flags & RADEON_PLL_PREFER_CLOSEST_LOWER) {
+						error = freq - current_freq;
+						error = error < 0 ? 0xffffffff : error;
+					} else
+						error = abs(current_freq - freq);
 					vco_diff = abs(vco - best_vco);
 
 					if ((best_vco == 0 && error < best_error) ||
@@ -657,36 +662,51 @@ void radeon_modeset_fini(struct radeon_device *rdev)
 	}
 }
 
-void radeon_init_disp_bandwidth(struct drm_device *dev)
+bool radeon_crtc_scaling_mode_fixup(struct drm_crtc *crtc,
+				struct drm_display_mode *mode,
+				struct drm_display_mode *adjusted_mode)
 {
-	struct radeon_device *rdev = dev->dev_private;
-	struct drm_display_mode *modes[2];
-	int pixel_bytes[2];
-	struct drm_crtc *crtc;
-
-	pixel_bytes[0] = pixel_bytes[1] = 0;
-	modes[0] = modes[1] = NULL;
-
-	list_for_each_entry(crtc, &dev->mode_config.crtc_list, head) {
-		struct radeon_crtc *radeon_crtc = to_radeon_crtc(crtc);
+	struct drm_device *dev = crtc->dev;
+	struct drm_encoder *encoder;
+	struct radeon_crtc *radeon_crtc = to_radeon_crtc(crtc);
+	struct radeon_encoder *radeon_encoder;
+	bool first = true;
 
-		if (crtc->enabled && crtc->fb) {
-			modes[radeon_crtc->crtc_id] = &crtc->mode;
-			pixel_bytes[radeon_crtc->crtc_id] = crtc->fb->bits_per_pixel / 8;
+	list_for_each_entry(encoder, &dev->mode_config.encoder_list, head) {
+		radeon_encoder = to_radeon_encoder(encoder);
+		if (encoder->crtc != crtc)
+			continue;
+		if (first) {
+			radeon_crtc->rmx_type = radeon_encoder->rmx_type;
+			radeon_crtc->devices = radeon_encoder->devices;
+			memcpy(&radeon_crtc->native_mode,
+				&radeon_encoder->native_mode,
+				sizeof(struct radeon_native_mode));
+			first = false;
+		} else {
+			if (radeon_crtc->rmx_type != radeon_encoder->rmx_type) {
+				/* WARNING: Right now this can't happen but
+				 * in the future we need to check that scaling
+				 * are consistent accross different encoder
+				 * (ie all encoder can work with the same
+				 *  scaling).
+				 */
+				DRM_ERROR("Scaling not consistent accross encoder.\n");
+				return false;
+			}
 		}
 	}
-
-	if (ASIC_IS_AVIVO(rdev)) {
-		radeon_init_disp_bw_avivo(dev,
-					  modes[0],
-					  pixel_bytes[0],
-					  modes[1],
-					  pixel_bytes[1]);
+	if (radeon_crtc->rmx_type != RMX_OFF) {
+		fixed20_12 a, b;
+		a.full = rfixed_const(crtc->mode.vdisplay);
+		b.full = rfixed_const(radeon_crtc->native_mode.panel_xres);
+		radeon_crtc->vsc.full = rfixed_div(a, b);
+		a.full = rfixed_const(crtc->mode.hdisplay);
+		b.full = rfixed_const(radeon_crtc->native_mode.panel_yres);
+		radeon_crtc->hsc.full = rfixed_div(a, b);
 	} else {
-		radeon_init_disp_bw_legacy(dev,
-					   modes[0],
-					   pixel_bytes[0],
-					   modes[1],
-					   pixel_bytes[1]);
+		radeon_crtc->vsc.full = rfixed_const(1);
+		radeon_crtc->hsc.full = rfixed_const(1);
 	}
+	return true;
 }
diff --git a/drivers/gpu/drm/radeon/radeon_drv.c b/drivers/gpu/drm/radeon/radeon_drv.c
index 84ba69f4878..0bd5879a495 100644
--- a/drivers/gpu/drm/radeon/radeon_drv.c
+++ b/drivers/gpu/drm/radeon/radeon_drv.c
@@ -89,6 +89,7 @@ int radeon_agpmode = 0;
 int radeon_vram_limit = 0;
 int radeon_gart_size = 512; /* default gart size */
 int radeon_benchmarking = 0;
+int radeon_testing = 0;
 int radeon_connector_table = 0;
 #endif
 
@@ -117,6 +118,9 @@ module_param_named(gartsize, radeon_gart_size, int, 0600);
 MODULE_PARM_DESC(benchmark, "Run benchmark");
 module_param_named(benchmark, radeon_benchmarking, int, 0444);
 
+MODULE_PARM_DESC(test, "Run tests");
+module_param_named(test, radeon_testing, int, 0444);
+
 MODULE_PARM_DESC(connector_table, "Force connector table");
 module_param_named(connector_table, radeon_connector_table, int, 0444);
 #endif
@@ -314,6 +318,14 @@ static int __init radeon_init(void)
 	driver = &driver_old;
 	driver->num_ioctls = radeon_max_ioctl;
 #if defined(CONFIG_DRM_RADEON_KMS)
+#ifdef CONFIG_VGA_CONSOLE
+	if (vgacon_text_force() && radeon_modeset == -1) {
+		DRM_INFO("VGACON disable radeon kernel modesetting.\n");
+		driver = &driver_old;
+		driver->driver_features &= ~DRIVER_MODESET;
+		radeon_modeset = 0;
+	}
+#endif
 	/* if enabled by default */
 	if (radeon_modeset == -1) {
 		DRM_INFO("radeon default to kernel modesetting.\n");
@@ -325,17 +337,8 @@ static int __init radeon_init(void)
 		driver->driver_features |= DRIVER_MODESET;
 		driver->num_ioctls = radeon_max_kms_ioctl;
 	}
-
 	/* if the vga console setting is enabled still
 	 * let modprobe override it */
-#ifdef CONFIG_VGA_CONSOLE
-	if (vgacon_text_force() && radeon_modeset == -1) {
-		DRM_INFO("VGACON disable radeon kernel modesetting.\n");
-		driver = &driver_old;
-		driver->driver_features &= ~DRIVER_MODESET;
-		radeon_modeset = 0;
-	}
-#endif
 #endif
 	return drm_init(driver);
 }
diff --git a/drivers/gpu/drm/radeon/radeon_drv.h b/drivers/gpu/drm/radeon/radeon_drv.h
index 127d0456f62..3933f8216a3 100644
--- a/drivers/gpu/drm/radeon/radeon_drv.h
+++ b/drivers/gpu/drm/radeon/radeon_drv.h
@@ -143,6 +143,7 @@ enum radeon_family {
 	CHIP_RV635,
 	CHIP_RV670,
 	CHIP_RS780,
+	CHIP_RS880,
 	CHIP_RV770,
 	CHIP_RV730,
 	CHIP_RV710,
diff --git a/drivers/gpu/drm/radeon/radeon_encoders.c b/drivers/gpu/drm/radeon/radeon_encoders.c
index c8ef0d14ffa..0a92706eac1 100644
--- a/drivers/gpu/drm/radeon/radeon_encoders.c
+++ b/drivers/gpu/drm/radeon/radeon_encoders.c
@@ -154,7 +154,6 @@ void radeon_rmx_mode_fixup(struct drm_encoder *encoder,
 
 	if (mode->hdisplay < native_mode->panel_xres ||
 	    mode->vdisplay < native_mode->panel_yres) {
-		radeon_encoder->flags |= RADEON_USE_RMX;
 		if (ASIC_IS_AVIVO(rdev)) {
 			adjusted_mode->hdisplay = native_mode->panel_xres;
 			adjusted_mode->vdisplay = native_mode->panel_yres;
@@ -197,15 +196,13 @@ void radeon_rmx_mode_fixup(struct drm_encoder *encoder,
 	}
 }
 
+
 static bool radeon_atom_mode_fixup(struct drm_encoder *encoder,
 				   struct drm_display_mode *mode,
 				   struct drm_display_mode *adjusted_mode)
 {
-
 	struct radeon_encoder *radeon_encoder = to_radeon_encoder(encoder);
 
-	radeon_encoder->flags &= ~RADEON_USE_RMX;
-
 	drm_mode_set_crtcinfo(adjusted_mode, 0);
 
 	if (radeon_encoder->rmx_type != RMX_OFF)
@@ -808,234 +805,6 @@ atombios_dig_transmitter_setup(struct drm_encoder *encoder, int action)
 
 }
 
-static void atom_rv515_force_tv_scaler(struct radeon_device *rdev)
-{
-
-	WREG32(0x659C, 0x0);
-	WREG32(0x6594, 0x705);
-	WREG32(0x65A4, 0x10001);
-	WREG32(0x65D8, 0x0);
-	WREG32(0x65B0, 0x0);
-	WREG32(0x65C0, 0x0);
-	WREG32(0x65D4, 0x0);
-	WREG32(0x6578, 0x0);
-	WREG32(0x657C, 0x841880A8);
-	WREG32(0x6578, 0x1);
-	WREG32(0x657C, 0x84208680);
-	WREG32(0x6578, 0x2);
-	WREG32(0x657C, 0xBFF880B0);
-	WREG32(0x6578, 0x100);
-	WREG32(0x657C, 0x83D88088);
-	WREG32(0x6578, 0x101);
-	WREG32(0x657C, 0x84608680);
-	WREG32(0x6578, 0x102);
-	WREG32(0x657C, 0xBFF080D0);
-	WREG32(0x6578, 0x200);
-	WREG32(0x657C, 0x83988068);
-	WREG32(0x6578, 0x201);
-	WREG32(0x657C, 0x84A08680);
-	WREG32(0x6578, 0x202);
-	WREG32(0x657C, 0xBFF080F8);
-	WREG32(0x6578, 0x300);
-	WREG32(0x657C, 0x83588058);
-	WREG32(0x6578, 0x301);
-	WREG32(0x657C, 0x84E08660);
-	WREG32(0x6578, 0x302);
-	WREG32(0x657C, 0xBFF88120);
-	WREG32(0x6578, 0x400);
-	WREG32(0x657C, 0x83188040);
-	WREG32(0x6578, 0x401);
-	WREG32(0x657C, 0x85008660);
-	WREG32(0x6578, 0x402);
-	WREG32(0x657C, 0xBFF88150);
-	WREG32(0x6578, 0x500);
-	WREG32(0x657C, 0x82D88030);
-	WREG32(0x6578, 0x501);
-	WREG32(0x657C, 0x85408640);
-	WREG32(0x6578, 0x502);
-	WREG32(0x657C, 0xBFF88180);
-	WREG32(0x6578, 0x600);
-	WREG32(0x657C, 0x82A08018);
-	WREG32(0x6578, 0x601);
-	WREG32(0x657C, 0x85808620);
-	WREG32(0x6578, 0x602);
-	WREG32(0x657C, 0xBFF081B8);
-	WREG32(0x6578, 0x700);
-	WREG32(0x657C, 0x82608010);
-	WREG32(0x6578, 0x701);
-	WREG32(0x657C, 0x85A08600);
-	WREG32(0x6578, 0x702);
-	WREG32(0x657C, 0x800081F0);
-	WREG32(0x6578, 0x800);
-	WREG32(0x657C, 0x8228BFF8);
-	WREG32(0x6578, 0x801);
-	WREG32(0x657C, 0x85E085E0);
-	WREG32(0x6578, 0x802);
-	WREG32(0x657C, 0xBFF88228);
-	WREG32(0x6578, 0x10000);
-	WREG32(0x657C, 0x82A8BF00);
-	WREG32(0x6578, 0x10001);
-	WREG32(0x657C, 0x82A08CC0);
-	WREG32(0x6578, 0x10002);
-	WREG32(0x657C, 0x8008BEF8);
-	WREG32(0x6578, 0x10100);
-	WREG32(0x657C, 0x81F0BF28);
-	WREG32(0x6578, 0x10101);
-	WREG32(0x657C, 0x83608CA0);
-	WREG32(0x6578, 0x10102);
-	WREG32(0x657C, 0x8018BED0);
-	WREG32(0x6578, 0x10200);
-	WREG32(0x657C, 0x8148BF38);
-	WREG32(0x6578, 0x10201);
-	WREG32(0x657C, 0x84408C80);
-	WREG32(0x6578, 0x10202);
-	WREG32(0x657C, 0x8008BEB8);
-	WREG32(0x6578, 0x10300);
-	WREG32(0x657C, 0x80B0BF78);
-	WREG32(0x6578, 0x10301);
-	WREG32(0x657C, 0x85008C20);
-	WREG32(0x6578, 0x10302);
-	WREG32(0x657C, 0x8020BEA0);
-	WREG32(0x6578, 0x10400);
-	WREG32(0x657C, 0x8028BF90);
-	WREG32(0x6578, 0x10401);
-	WREG32(0x657C, 0x85E08BC0);
-	WREG32(0x6578, 0x10402);
-	WREG32(0x657C, 0x8018BE90);
-	WREG32(0x6578, 0x10500);
-	WREG32(0x657C, 0xBFB8BFB0);
-	WREG32(0x6578, 0x10501);
-	WREG32(0x657C, 0x86C08B40);
-	WREG32(0x6578, 0x10502);
-	WREG32(0x657C, 0x8010BE90);
-	WREG32(0x6578, 0x10600);
-	WREG32(0x657C, 0xBF58BFC8);
-	WREG32(0x6578, 0x10601);
-	WREG32(0x657C, 0x87A08AA0);
-	WREG32(0x6578, 0x10602);
-	WREG32(0x657C, 0x8010BE98);
-	WREG32(0x6578, 0x10700);
-	WREG32(0x657C, 0xBF10BFF0);
-	WREG32(0x6578, 0x10701);
-	WREG32(0x657C, 0x886089E0);
-	WREG32(0x6578, 0x10702);
-	WREG32(0x657C, 0x8018BEB0);
-	WREG32(0x6578, 0x10800);
-	WREG32(0x657C, 0xBED8BFE8);
-	WREG32(0x6578, 0x10801);
-	WREG32(0x657C, 0x89408940);
-	WREG32(0x6578, 0x10802);
-	WREG32(0x657C, 0xBFE8BED8);
-	WREG32(0x6578, 0x20000);
-	WREG32(0x657C, 0x80008000);
-	WREG32(0x6578, 0x20001);
-	WREG32(0x657C, 0x90008000);
-	WREG32(0x6578, 0x20002);
-	WREG32(0x657C, 0x80008000);
-	WREG32(0x6578, 0x20003);
-	WREG32(0x657C, 0x80008000);
-	WREG32(0x6578, 0x20100);
-	WREG32(0x657C, 0x80108000);
-	WREG32(0x6578, 0x20101);
-	WREG32(0x657C, 0x8FE0BF70);
-	WREG32(0x6578, 0x20102);
-	WREG32(0x657C, 0xBFE880C0);
-	WREG32(0x6578, 0x20103);
-	WREG32(0x657C, 0x80008000);
-	WREG32(0x6578, 0x20200);
-	WREG32(0x657C, 0x8018BFF8);
-	WREG32(0x6578, 0x20201);
-	WREG32(0x657C, 0x8F80BF08);
-	WREG32(0x6578, 0x20202);
-	WREG32(0x657C, 0xBFD081A0);
-	WREG32(0x6578, 0x20203);
-	WREG32(0x657C, 0xBFF88000);
-	WREG32(0x6578, 0x20300);
-	WREG32(0x657C, 0x80188000);
-	WREG32(0x6578, 0x20301);
-	WREG32(0x657C, 0x8EE0BEC0);
-	WREG32(0x6578, 0x20302);
-	WREG32(0x657C, 0xBFB082A0);
-	WREG32(0x6578, 0x20303);
-	WREG32(0x657C, 0x80008000);
-	WREG32(0x6578, 0x20400);
-	WREG32(0x657C, 0x80188000);
-	WREG32(0x6578, 0x20401);
-	WREG32(0x657C, 0x8E00BEA0);
-	WREG32(0x6578, 0x20402);
-	WREG32(0x657C, 0xBF8883C0);
-	WREG32(0x6578, 0x20403);
-	WREG32(0x657C, 0x80008000);
-	WREG32(0x6578, 0x20500);
-	WREG32(0x657C, 0x80188000);
-	WREG32(0x6578, 0x20501);
-	WREG32(0x657C, 0x8D00BE90);
-	WREG32(0x6578, 0x20502);
-	WREG32(0x657C, 0xBF588500);
-	WREG32(0x6578, 0x20503);
-	WREG32(0x657C, 0x80008008);
-	WREG32(0x6578, 0x20600);
-	WREG32(0x657C, 0x80188000);
-	WREG32(0x6578, 0x20601);
-	WREG32(0x657C, 0x8BC0BE98);
-	WREG32(0x6578, 0x20602);
-	WREG32(0x657C, 0xBF308660);
-	WREG32(0x6578, 0x20603);
-	WREG32(0x657C, 0x80008008);
-	WREG32(0x6578, 0x20700);
-	WREG32(0x657C, 0x80108000);
-	WREG32(0x6578, 0x20701);
-	WREG32(0x657C, 0x8A80BEB0);
-	WREG32(0x6578, 0x20702);
-	WREG32(0x657C, 0xBF0087C0);
-	WREG32(0x6578, 0x20703);
-	WREG32(0x657C, 0x80008008);
-	WREG32(0x6578, 0x20800);
-	WREG32(0x657C, 0x80108000);
-	WREG32(0x6578, 0x20801);
-	WREG32(0x657C, 0x8920BED0);
-	WREG32(0x6578, 0x20802);
-	WREG32(0x657C, 0xBED08920);
-	WREG32(0x6578, 0x20803);
-	WREG32(0x657C, 0x80008010);
-	WREG32(0x6578, 0x30000);
-	WREG32(0x657C, 0x90008000);
-	WREG32(0x6578, 0x30001);
-	WREG32(0x657C, 0x80008000);
-	WREG32(0x6578, 0x30100);
-	WREG32(0x657C, 0x8FE0BF90);
-	WREG32(0x6578, 0x30101);
-	WREG32(0x657C, 0xBFF880A0);
-	WREG32(0x6578, 0x30200);
-	WREG32(0x657C, 0x8F60BF40);
-	WREG32(0x6578, 0x30201);
-	WREG32(0x657C, 0xBFE88180);
-	WREG32(0x6578, 0x30300);
-	WREG32(0x657C, 0x8EC0BF00);
-	WREG32(0x6578, 0x30301);
-	WREG32(0x657C, 0xBFC88280);
-	WREG32(0x6578, 0x30400);
-	WREG32(0x657C, 0x8DE0BEE0);
-	WREG32(0x6578, 0x30401);
-	WREG32(0x657C, 0xBFA083A0);
-	WREG32(0x6578, 0x30500);
-	WREG32(0x657C, 0x8CE0BED0);
-	WREG32(0x6578, 0x30501);
-	WREG32(0x657C, 0xBF7884E0);
-	WREG32(0x6578, 0x30600);
-	WREG32(0x657C, 0x8BA0BED8);
-	WREG32(0x6578, 0x30601);
-	WREG32(0x657C, 0xBF508640);
-	WREG32(0x6578, 0x30700);
-	WREG32(0x657C, 0x8A60BEE8);
-	WREG32(0x6578, 0x30701);
-	WREG32(0x657C, 0xBF2087A0);
-	WREG32(0x6578, 0x30800);
-	WREG32(0x657C, 0x8900BF00);
-	WREG32(0x6578, 0x30801);
-	WREG32(0x657C, 0xBF008900);
-}
-
 static void
 atombios_yuv_setup(struct drm_encoder *encoder, bool enable)
 {
@@ -1074,129 +843,6 @@ atombios_yuv_setup(struct drm_encoder *encoder, bool enable)
 }
 
 static void
-atombios_overscan_setup(struct drm_encoder *encoder,
-			struct drm_display_mode *mode,
-			struct drm_display_mode *adjusted_mode)
-{
-	struct drm_device *dev = encoder->dev;
-	struct radeon_device *rdev = dev->dev_private;
-	struct radeon_encoder *radeon_encoder = to_radeon_encoder(encoder);
-	struct radeon_crtc *radeon_crtc = to_radeon_crtc(encoder->crtc);
-	SET_CRTC_OVERSCAN_PS_ALLOCATION args;
-	int index = GetIndexIntoMasterTable(COMMAND, SetCRTC_OverScan);
-
-	memset(&args, 0, sizeof(args));
-
-	args.usOverscanRight = 0;
-	args.usOverscanLeft = 0;
-	args.usOverscanBottom = 0;
-	args.usOverscanTop = 0;
-	args.ucCRTC = radeon_crtc->crtc_id;
-
-	if (radeon_encoder->flags & RADEON_USE_RMX) {
-		if (radeon_encoder->rmx_type == RMX_FULL) {
-			args.usOverscanRight = 0;
-			args.usOverscanLeft = 0;
-			args.usOverscanBottom = 0;
-			args.usOverscanTop = 0;
-		} else if (radeon_encoder->rmx_type == RMX_CENTER) {
-			args.usOverscanTop = (adjusted_mode->crtc_vdisplay - mode->crtc_vdisplay) / 2;
-			args.usOverscanBottom = (adjusted_mode->crtc_vdisplay - mode->crtc_vdisplay) / 2;
-			args.usOverscanLeft = (adjusted_mode->crtc_hdisplay - mode->crtc_hdisplay) / 2;
-			args.usOverscanRight = (adjusted_mode->crtc_hdisplay - mode->crtc_hdisplay) / 2;
-		} else if (radeon_encoder->rmx_type == RMX_ASPECT) {
-			int a1 = mode->crtc_vdisplay * adjusted_mode->crtc_hdisplay;
-			int a2 = adjusted_mode->crtc_vdisplay * mode->crtc_hdisplay;
-
-			if (a1 > a2) {
-				args.usOverscanLeft = (adjusted_mode->crtc_hdisplay - (a2 / mode->crtc_vdisplay)) / 2;
-				args.usOverscanRight = (adjusted_mode->crtc_hdisplay - (a2 / mode->crtc_vdisplay)) / 2;
-			} else if (a2 > a1) {
-				args.usOverscanLeft = (adjusted_mode->crtc_vdisplay - (a1 / mode->crtc_hdisplay)) / 2;
-				args.usOverscanRight = (adjusted_mode->crtc_vdisplay - (a1 / mode->crtc_hdisplay)) / 2;
-			}
-		}
-	}
-
-	atom_execute_table(rdev->mode_info.atom_context, index, (uint32_t *)&args);
-
-}
-
-static void
-atombios_scaler_setup(struct drm_encoder *encoder)
-{
-	struct drm_device *dev = encoder->dev;
-	struct radeon_device *rdev = dev->dev_private;
-	struct radeon_encoder *radeon_encoder = to_radeon_encoder(encoder);
-	struct radeon_crtc *radeon_crtc = to_radeon_crtc(encoder->crtc);
-	ENABLE_SCALER_PS_ALLOCATION args;
-	int index = GetIndexIntoMasterTable(COMMAND, EnableScaler);
-	/* fixme - fill in enc_priv for atom dac */
-	enum radeon_tv_std tv_std = TV_STD_NTSC;
-
-	if (!ASIC_IS_AVIVO(rdev) && radeon_crtc->crtc_id)
-		return;
-
-	memset(&args, 0, sizeof(args));
-
-	args.ucScaler = radeon_crtc->crtc_id;
-
-	if (radeon_encoder->devices & (ATOM_DEVICE_TV_SUPPORT)) {
-		switch (tv_std) {
-		case TV_STD_NTSC:
-		default:
-			args.ucTVStandard = ATOM_TV_NTSC;
-			break;
-		case TV_STD_PAL:
-			args.ucTVStandard = ATOM_TV_PAL;
-			break;
-		case TV_STD_PAL_M:
-			args.ucTVStandard = ATOM_TV_PALM;
-			break;
-		case TV_STD_PAL_60:
-			args.ucTVStandard = ATOM_TV_PAL60;
-			break;
-		case TV_STD_NTSC_J:
-			args.ucTVStandard = ATOM_TV_NTSCJ;
-			break;
-		case TV_STD_SCART_PAL:
-			args.ucTVStandard = ATOM_TV_PAL; /* ??? */
-			break;
-		case TV_STD_SECAM:
-			args.ucTVStandard = ATOM_TV_SECAM;
-			break;
-		case TV_STD_PAL_CN:
-			args.ucTVStandard = ATOM_TV_PALCN;
-			break;
-		}
-		args.ucEnable = SCALER_ENABLE_MULTITAP_MODE;
-	} else if (radeon_encoder->devices & (ATOM_DEVICE_CV_SUPPORT)) {
-		args.ucTVStandard = ATOM_TV_CV;
-		args.ucEnable = SCALER_ENABLE_MULTITAP_MODE;
-	} else if (radeon_encoder->flags & RADEON_USE_RMX) {
-		if (radeon_encoder->rmx_type == RMX_FULL)
-			args.ucEnable = ATOM_SCALER_EXPANSION;
-		else if (radeon_encoder->rmx_type == RMX_CENTER)
-			args.ucEnable = ATOM_SCALER_CENTER;
-		else if (radeon_encoder->rmx_type == RMX_ASPECT)
-			args.ucEnable = ATOM_SCALER_EXPANSION;
-	} else {
-		if (ASIC_IS_AVIVO(rdev))
-			args.ucEnable = ATOM_SCALER_DISABLE;
-		else
-			args.ucEnable = ATOM_SCALER_CENTER;
-	}
-
-	atom_execute_table(rdev->mode_info.atom_context, index, (uint32_t *)&args);
-
-	if (radeon_encoder->devices & (ATOM_DEVICE_CV_SUPPORT | ATOM_DEVICE_TV_SUPPORT)
-	    && rdev->family >= CHIP_RV515 && rdev->family <= CHIP_RV570) {
-		atom_rv515_force_tv_scaler(rdev);
-	}
-
-}
-
-static void
 radeon_atom_encoder_dpms(struct drm_encoder *encoder, int mode)
 {
 	struct drm_device *dev = encoder->dev;
@@ -1448,8 +1094,6 @@ radeon_atom_encoder_mode_set(struct drm_encoder *encoder,
 	radeon_encoder->pixel_clock = adjusted_mode->clock;
 
 	radeon_atombios_encoder_crtc_scratch_regs(encoder, radeon_crtc->crtc_id);
-	atombios_overscan_setup(encoder, mode, adjusted_mode);
-	atombios_scaler_setup(encoder);
 	atombios_set_encoder_crtc_source(encoder);
 
 	if (ASIC_IS_AVIVO(rdev)) {
@@ -1667,6 +1311,7 @@ radeon_add_atom_encoder(struct drm_device *dev, uint32_t encoder_id, uint32_t su
 
 	radeon_encoder->encoder_id = encoder_id;
 	radeon_encoder->devices = supported_device;
+	radeon_encoder->rmx_type = RMX_OFF;
 
 	switch (radeon_encoder->encoder_id) {
 	case ENCODER_OBJECT_ID_INTERNAL_LVDS:
diff --git a/drivers/gpu/drm/radeon/radeon_fb.c b/drivers/gpu/drm/radeon/radeon_fb.c
index 9e8f191eb64..3206c0ad7b6 100644
--- a/drivers/gpu/drm/radeon/radeon_fb.c
+++ b/drivers/gpu/drm/radeon/radeon_fb.c
@@ -101,9 +101,10 @@ static int radeonfb_setcolreg(unsigned regno,
 				break;
 			case 24:
 			case 32:
-				fb->pseudo_palette[regno] = ((red & 0xff00) << 8) |
-					(green & 0xff00) |
-					((blue  & 0xff00) >> 8);
+				fb->pseudo_palette[regno] =
+					(((red >> 8) & 0xff) << info->var.red.offset) |
+					(((green >> 8) & 0xff) << info->var.green.offset) |
+					(((blue >> 8) & 0xff) << info->var.blue.offset);
 				break;
 			}
 		}
@@ -154,6 +155,7 @@ static int radeonfb_check_var(struct fb_var_screeninfo *var,
 		var->transp.length = 0;
 		var->transp.offset = 0;
 		break;
+#ifdef __LITTLE_ENDIAN
 	case 15:
 		var->red.offset = 10;
 		var->green.offset = 5;
@@ -194,6 +196,28 @@ static int radeonfb_check_var(struct fb_var_screeninfo *var,
 		var->transp.length = 8;
 		var->transp.offset = 24;
 		break;
+#else
+	case 24:
+		var->red.offset = 8;
+		var->green.offset = 16;
+		var->blue.offset = 24;
+		var->red.length = 8;
+		var->green.length = 8;
+		var->blue.length = 8;
+		var->transp.length = 0;
+		var->transp.offset = 0;
+		break;
+	case 32:
+		var->red.offset = 8;
+		var->green.offset = 16;
+		var->blue.offset = 24;
+		var->red.length = 8;
+		var->green.length = 8;
+		var->blue.length = 8;
+		var->transp.length = 8;
+		var->transp.offset = 0;
+		break;
+#endif
 	default:
 		return -EINVAL;
 	}
@@ -447,10 +471,10 @@ static struct notifier_block paniced = {
 	.notifier_call = radeonfb_panic,
 };
 
-static int radeon_align_pitch(struct radeon_device *rdev, int width, int bpp)
+static int radeon_align_pitch(struct radeon_device *rdev, int width, int bpp, bool tiled)
 {
 	int aligned = width;
-	int align_large = (ASIC_IS_AVIVO(rdev));
+	int align_large = (ASIC_IS_AVIVO(rdev)) || tiled;
 	int pitch_mask = 0;
 
 	switch (bpp / 8) {
@@ -488,12 +512,13 @@ int radeonfb_create(struct radeon_device *rdev,
 	u64 fb_gpuaddr;
 	void *fbptr = NULL;
 	unsigned long tmp;
+	bool fb_tiled = false; /* useful for testing */
 
 	mode_cmd.width = surface_width;
 	mode_cmd.height = surface_height;
 	mode_cmd.bpp = 32;
 	/* need to align pitch with crtc limits */
-	mode_cmd.pitch = radeon_align_pitch(rdev, mode_cmd.width, mode_cmd.bpp) * ((mode_cmd.bpp + 1) / 8);
+	mode_cmd.pitch = radeon_align_pitch(rdev, mode_cmd.width, mode_cmd.bpp, fb_tiled) * ((mode_cmd.bpp + 1) / 8);
 	mode_cmd.depth = 24;
 
 	size = mode_cmd.pitch * mode_cmd.height;
@@ -511,6 +536,8 @@ int radeonfb_create(struct radeon_device *rdev,
 	}
 	robj = gobj->driver_private;
 
+	if (fb_tiled)
+		radeon_object_set_tiling_flags(robj, RADEON_TILING_MACRO|RADEON_TILING_SURFACE, mode_cmd.pitch);
 	mutex_lock(&rdev->ddev->struct_mutex);
 	fb = radeon_framebuffer_create(rdev->ddev, &mode_cmd, gobj);
 	if (fb == NULL) {
@@ -539,6 +566,9 @@ int radeonfb_create(struct radeon_device *rdev,
 	}
 	rfbdev = info->par;
 
+	if (fb_tiled)
+		radeon_object_check_tiling(robj, 0, 0);
+
 	ret = radeon_object_kmap(robj, &fbptr);
 	if (ret) {
 		goto out_unref;
@@ -572,6 +602,11 @@ int radeonfb_create(struct radeon_device *rdev,
 	info->var.width = -1;
 	info->var.xres = fb_width;
 	info->var.yres = fb_height;
+
+	/* setup aperture base/size for vesafb takeover */
+	info->aperture_base = rdev->ddev->mode_config.fb_base;
+	info->aperture_size = rdev->mc.real_vram_size;
+
 	info->fix.mmio_start = 0;
 	info->fix.mmio_len = 0;
 	info->pixmap.size = 64*1024;
@@ -600,6 +635,7 @@ int radeonfb_create(struct radeon_device *rdev,
 		info->var.transp.offset = 0;
 		info->var.transp.length = 0;
 		break;
+#ifdef __LITTLE_ENDIAN
 	case 15:
 		info->var.red.offset = 10;
 		info->var.green.offset = 5;
@@ -639,7 +675,29 @@ int radeonfb_create(struct radeon_device *rdev,
 		info->var.transp.offset = 24;
 		info->var.transp.length = 8;
 		break;
+#else
+	case 24:
+		info->var.red.offset = 8;
+		info->var.green.offset = 16;
+		info->var.blue.offset = 24;
+		info->var.red.length = 8;
+		info->var.green.length = 8;
+		info->var.blue.length = 8;
+		info->var.transp.offset = 0;
+		info->var.transp.length = 0;
+		break;
+	case 32:
+		info->var.red.offset = 8;
+		info->var.green.offset = 16;
+		info->var.blue.offset = 24;
+		info->var.red.length = 8;
+		info->var.green.length = 8;
+		info->var.blue.length = 8;
+		info->var.transp.offset = 0;
+		info->var.transp.length = 8;
+		break;
 	default:
+#endif
 		break;
 	}
 
diff --git a/drivers/gpu/drm/radeon/radeon_fence.c b/drivers/gpu/drm/radeon/radeon_fence.c
index 96afbf5ae2a..b4e48dd2e85 100644
--- a/drivers/gpu/drm/radeon/radeon_fence.c
+++ b/drivers/gpu/drm/radeon/radeon_fence.c
@@ -195,7 +195,7 @@ retry:
 		r = wait_event_interruptible_timeout(rdev->fence_drv.queue,
 				radeon_fence_signaled(fence), timeout);
 		if (unlikely(r == -ERESTARTSYS)) {
-			return -ERESTART;
+			return -EBUSY;
 		}
 	} else {
 		r = wait_event_timeout(rdev->fence_drv.queue,
diff --git a/drivers/gpu/drm/radeon/radeon_gart.c b/drivers/gpu/drm/radeon/radeon_gart.c
index d343a15316e..2977539880f 100644
--- a/drivers/gpu/drm/radeon/radeon_gart.c
+++ b/drivers/gpu/drm/radeon/radeon_gart.c
@@ -177,7 +177,7 @@ int radeon_gart_bind(struct radeon_device *rdev, unsigned offset,
 			return -ENOMEM;
 		}
 		rdev->gart.pages[p] = pagelist[i];
-		page_base = (uint32_t)rdev->gart.pages_addr[p];
+		page_base = rdev->gart.pages_addr[p];
 		for (j = 0; j < (PAGE_SIZE / 4096); j++, t++) {
 			radeon_gart_set_page(rdev, t, page_base);
 			page_base += 4096;
diff --git a/drivers/gpu/drm/radeon/radeon_gem.c b/drivers/gpu/drm/radeon/radeon_gem.c
index eb516034235..cded5180c75 100644
--- a/drivers/gpu/drm/radeon/radeon_gem.c
+++ b/drivers/gpu/drm/radeon/radeon_gem.c
@@ -157,9 +157,9 @@ int radeon_gem_info_ioctl(struct drm_device *dev, void *data,
 	struct radeon_device *rdev = dev->dev_private;
 	struct drm_radeon_gem_info *args = data;
 
-	args->vram_size = rdev->mc.vram_size;
+	args->vram_size = rdev->mc.real_vram_size;
 	/* FIXME: report somethings that makes sense */
-	args->vram_visible = rdev->mc.vram_size - (4 * 1024 * 1024);
+	args->vram_visible = rdev->mc.real_vram_size - (4 * 1024 * 1024);
 	args->gart_size = rdev->mc.gtt_size;
 	return 0;
 }
@@ -285,3 +285,44 @@ int radeon_gem_wait_idle_ioctl(struct drm_device *dev, void *data,
 	mutex_unlock(&dev->struct_mutex);
 	return r;
 }
+
+int radeon_gem_set_tiling_ioctl(struct drm_device *dev, void *data,
+				struct drm_file *filp)
+{
+	struct drm_radeon_gem_set_tiling *args = data;
+	struct drm_gem_object *gobj;
+	struct radeon_object *robj;
+	int r = 0;
+
+	DRM_DEBUG("%d \n", args->handle);
+	gobj = drm_gem_object_lookup(dev, filp, args->handle);
+	if (gobj == NULL)
+		return -EINVAL;
+	robj = gobj->driver_private;
+	radeon_object_set_tiling_flags(robj, args->tiling_flags, args->pitch);
+	mutex_lock(&dev->struct_mutex);
+	drm_gem_object_unreference(gobj);
+	mutex_unlock(&dev->struct_mutex);
+	return r;
+}
+
+int radeon_gem_get_tiling_ioctl(struct drm_device *dev, void *data,
+				struct drm_file *filp)
+{
+	struct drm_radeon_gem_get_tiling *args = data;
+	struct drm_gem_object *gobj;
+	struct radeon_object *robj;
+	int r = 0;
+
+	DRM_DEBUG("\n");
+	gobj = drm_gem_object_lookup(dev, filp, args->handle);
+	if (gobj == NULL)
+		return -EINVAL;
+	robj = gobj->driver_private;
+	radeon_object_get_tiling_flags(robj, &args->tiling_flags,
+				       &args->pitch);
+	mutex_lock(&dev->struct_mutex);
+	drm_gem_object_unreference(gobj);
+	mutex_unlock(&dev->struct_mutex);
+	return r;
+}
diff --git a/drivers/gpu/drm/radeon/radeon_kms.c b/drivers/gpu/drm/radeon/radeon_kms.c
index 4612a7c146d..3357110e30c 100644
--- a/drivers/gpu/drm/radeon/radeon_kms.c
+++ b/drivers/gpu/drm/radeon/radeon_kms.c
@@ -58,6 +58,8 @@ int radeon_driver_load_kms(struct drm_device *dev, unsigned long flags)
 	if (r) {
 		DRM_ERROR("Failed to initialize radeon, disabling IOCTL\n");
 		radeon_device_fini(rdev);
+		kfree(rdev);
+		dev->dev_private = NULL;
 		return r;
 	}
 	return 0;
@@ -291,5 +293,7 @@ struct drm_ioctl_desc radeon_ioctls_kms[] = {
 	DRM_IOCTL_DEF(DRM_RADEON_GEM_WAIT_IDLE, radeon_gem_wait_idle_ioctl, DRM_AUTH),
 	DRM_IOCTL_DEF(DRM_RADEON_CS, radeon_cs_ioctl, DRM_AUTH),
 	DRM_IOCTL_DEF(DRM_RADEON_INFO, radeon_info_ioctl, DRM_AUTH),
+	DRM_IOCTL_DEF(DRM_RADEON_GEM_SET_TILING, radeon_gem_set_tiling_ioctl, DRM_AUTH),
+	DRM_IOCTL_DEF(DRM_RADEON_GEM_GET_TILING, radeon_gem_get_tiling_ioctl, DRM_AUTH),
 };
 int radeon_max_kms_ioctl = DRM_ARRAY_SIZE(radeon_ioctls_kms);
diff --git a/drivers/gpu/drm/radeon/radeon_legacy_crtc.c b/drivers/gpu/drm/radeon/radeon_legacy_crtc.c
index 8086ecf7f03..7d06dc98a42 100644
--- a/drivers/gpu/drm/radeon/radeon_legacy_crtc.c
+++ b/drivers/gpu/drm/radeon/radeon_legacy_crtc.c
@@ -29,6 +29,171 @@
 #include "radeon_fixed.h"
 #include "radeon.h"
 
+static void radeon_legacy_rmx_mode_set(struct drm_crtc *crtc,
+				       struct drm_display_mode *mode,
+				       struct drm_display_mode *adjusted_mode)
+{
+	struct drm_device *dev = crtc->dev;
+	struct radeon_device *rdev = dev->dev_private;
+	struct radeon_crtc *radeon_crtc = to_radeon_crtc(crtc);
+	int xres = mode->hdisplay;
+	int yres = mode->vdisplay;
+	bool hscale = true, vscale = true;
+	int hsync_wid;
+	int vsync_wid;
+	int hsync_start;
+	int blank_width;
+	u32 scale, inc, crtc_more_cntl;
+	u32 fp_horz_stretch, fp_vert_stretch, fp_horz_vert_active;
+	u32 fp_h_sync_strt_wid, fp_crtc_h_total_disp;
+	u32 fp_v_sync_strt_wid, fp_crtc_v_total_disp;
+	struct radeon_native_mode *native_mode = &radeon_crtc->native_mode;
+
+	fp_vert_stretch = RREG32(RADEON_FP_VERT_STRETCH) &
+		(RADEON_VERT_STRETCH_RESERVED |
+		 RADEON_VERT_AUTO_RATIO_INC);
+	fp_horz_stretch = RREG32(RADEON_FP_HORZ_STRETCH) &
+		(RADEON_HORZ_FP_LOOP_STRETCH |
+		 RADEON_HORZ_AUTO_RATIO_INC);
+
+	crtc_more_cntl = 0;
+	if ((rdev->family == CHIP_RS100) ||
+	    (rdev->family == CHIP_RS200)) {
+		/* This is to workaround the asic bug for RMX, some versions
+		   of BIOS dosen't have this register initialized correctly. */
+		crtc_more_cntl |= RADEON_CRTC_H_CUTOFF_ACTIVE_EN;
+	}
+
+
+	fp_crtc_h_total_disp = ((((mode->crtc_htotal / 8) - 1) & 0x3ff)
+				| ((((mode->crtc_hdisplay / 8) - 1) & 0x1ff) << 16));
+
+	hsync_wid = (mode->crtc_hsync_end - mode->crtc_hsync_start) / 8;
+	if (!hsync_wid)
+		hsync_wid = 1;
+	hsync_start = mode->crtc_hsync_start - 8;
+
+	fp_h_sync_strt_wid = ((hsync_start & 0x1fff)
+			      | ((hsync_wid & 0x3f) << 16)
+			      | ((mode->flags & DRM_MODE_FLAG_NHSYNC)
+				 ? RADEON_CRTC_H_SYNC_POL
+				 : 0));
+
+	fp_crtc_v_total_disp = (((mode->crtc_vtotal - 1) & 0xffff)
+				| ((mode->crtc_vdisplay - 1) << 16));
+
+	vsync_wid = mode->crtc_vsync_end - mode->crtc_vsync_start;
+	if (!vsync_wid)
+		vsync_wid = 1;
+
+	fp_v_sync_strt_wid = (((mode->crtc_vsync_start - 1) & 0xfff)
+			      | ((vsync_wid & 0x1f) << 16)
+			      | ((mode->flags & DRM_MODE_FLAG_NVSYNC)
+				 ? RADEON_CRTC_V_SYNC_POL
+				 : 0));
+
+	fp_horz_vert_active = 0;
+
+	if (native_mode->panel_xres == 0 ||
+	    native_mode->panel_yres == 0) {
+		hscale = false;
+		vscale = false;
+	} else {
+		if (xres > native_mode->panel_xres)
+			xres = native_mode->panel_xres;
+		if (yres > native_mode->panel_yres)
+			yres = native_mode->panel_yres;
+
+		if (xres == native_mode->panel_xres)
+			hscale = false;
+		if (yres == native_mode->panel_yres)
+			vscale = false;
+	}
+
+	switch (radeon_crtc->rmx_type) {
+	case RMX_FULL:
+	case RMX_ASPECT:
+		if (!hscale)
+			fp_horz_stretch |= ((xres/8-1) << 16);
+		else {
+			inc = (fp_horz_stretch & RADEON_HORZ_AUTO_RATIO_INC) ? 1 : 0;
+			scale = ((xres + inc) * RADEON_HORZ_STRETCH_RATIO_MAX)
+				/ native_mode->panel_xres + 1;
+			fp_horz_stretch |= (((scale) & RADEON_HORZ_STRETCH_RATIO_MASK) |
+					RADEON_HORZ_STRETCH_BLEND |
+					RADEON_HORZ_STRETCH_ENABLE |
+					((native_mode->panel_xres/8-1) << 16));
+		}
+
+		if (!vscale)
+			fp_vert_stretch |= ((yres-1) << 12);
+		else {
+			inc = (fp_vert_stretch & RADEON_VERT_AUTO_RATIO_INC) ? 1 : 0;
+			scale = ((yres + inc) * RADEON_VERT_STRETCH_RATIO_MAX)
+				/ native_mode->panel_yres + 1;
+			fp_vert_stretch |= (((scale) & RADEON_VERT_STRETCH_RATIO_MASK) |
+					RADEON_VERT_STRETCH_ENABLE |
+					RADEON_VERT_STRETCH_BLEND |
+					((native_mode->panel_yres-1) << 12));
+		}
+		break;
+	case RMX_CENTER:
+		fp_horz_stretch |= ((xres/8-1) << 16);
+		fp_vert_stretch |= ((yres-1) << 12);
+
+		crtc_more_cntl |= (RADEON_CRTC_AUTO_HORZ_CENTER_EN |
+				RADEON_CRTC_AUTO_VERT_CENTER_EN);
+
+		blank_width = (mode->crtc_hblank_end - mode->crtc_hblank_start) / 8;
+		if (blank_width > 110)
+			blank_width = 110;
+
+		fp_crtc_h_total_disp = (((blank_width) & 0x3ff)
+				| ((((mode->crtc_hdisplay / 8) - 1) & 0x1ff) << 16));
+
+		hsync_wid = (mode->crtc_hsync_end - mode->crtc_hsync_start) / 8;
+		if (!hsync_wid)
+			hsync_wid = 1;
+
+		fp_h_sync_strt_wid = ((((mode->crtc_hsync_start - mode->crtc_hblank_start) / 8) & 0x1fff)
+				| ((hsync_wid & 0x3f) << 16)
+				| ((mode->flags & DRM_MODE_FLAG_NHSYNC)
+					? RADEON_CRTC_H_SYNC_POL
+					: 0));
+
+		fp_crtc_v_total_disp = (((mode->crtc_vblank_end - mode->crtc_vblank_start) & 0xffff)
+				| ((mode->crtc_vdisplay - 1) << 16));
+
+		vsync_wid = mode->crtc_vsync_end - mode->crtc_vsync_start;
+		if (!vsync_wid)
+			vsync_wid = 1;
+
+		fp_v_sync_strt_wid = ((((mode->crtc_vsync_start - mode->crtc_vblank_start) & 0xfff)
+					| ((vsync_wid & 0x1f) << 16)
+					| ((mode->flags & DRM_MODE_FLAG_NVSYNC)
+						? RADEON_CRTC_V_SYNC_POL
+						: 0)));
+
+		fp_horz_vert_active = (((native_mode->panel_yres) & 0xfff) |
+				(((native_mode->panel_xres / 8) & 0x1ff) << 16));
+		break;
+	case RMX_OFF:
+	default:
+		fp_horz_stretch |= ((xres/8-1) << 16);
+		fp_vert_stretch |= ((yres-1) << 12);
+		break;
+	}
+
+	WREG32(RADEON_FP_HORZ_STRETCH,      fp_horz_stretch);
+	WREG32(RADEON_FP_VERT_STRETCH,      fp_vert_stretch);
+	WREG32(RADEON_CRTC_MORE_CNTL,       crtc_more_cntl);
+	WREG32(RADEON_FP_HORZ_VERT_ACTIVE,  fp_horz_vert_active);
+	WREG32(RADEON_FP_H_SYNC_STRT_WID,   fp_h_sync_strt_wid);
+	WREG32(RADEON_FP_V_SYNC_STRT_WID,   fp_v_sync_strt_wid);
+	WREG32(RADEON_FP_CRTC_H_TOTAL_DISP, fp_crtc_h_total_disp);
+	WREG32(RADEON_FP_CRTC_V_TOTAL_DISP, fp_crtc_v_total_disp);
+}
+
 void radeon_restore_common_regs(struct drm_device *dev)
 {
 	/* don't need this yet */
@@ -235,6 +400,7 @@ int radeon_crtc_set_base(struct drm_crtc *crtc, int x, int y,
 	uint64_t base;
 	uint32_t crtc_offset, crtc_offset_cntl, crtc_tile_x0_y0 = 0;
 	uint32_t crtc_pitch, pitch_pixels;
+	uint32_t tiling_flags;
 
 	DRM_DEBUG("\n");
 
@@ -244,7 +410,12 @@ int radeon_crtc_set_base(struct drm_crtc *crtc, int x, int y,
 	if (radeon_gem_object_pin(obj, RADEON_GEM_DOMAIN_VRAM, &base)) {
 		return -EINVAL;
 	}
-	crtc_offset = (u32)base;
+	/* if scanout was in GTT this really wouldn't work */
+	/* crtc offset is from display base addr not FB location */
+	radeon_crtc->legacy_display_base_addr = rdev->mc.vram_location;
+
+	base -= radeon_crtc->legacy_display_base_addr;
+
 	crtc_offset_cntl = 0;
 
 	pitch_pixels = crtc->fb->pitch / (crtc->fb->bits_per_pixel / 8);
@@ -253,8 +424,12 @@ int radeon_crtc_set_base(struct drm_crtc *crtc, int x, int y,
 		       (crtc->fb->bits_per_pixel * 8));
 	crtc_pitch |= crtc_pitch << 16;
 
-	/* TODO tiling */
-	if (0) {
+	radeon_object_get_tiling_flags(obj->driver_private,
+				       &tiling_flags, NULL);
+	if (tiling_flags & RADEON_TILING_MICRO)
+		DRM_ERROR("trying to scanout microtiled buffer\n");
+
+	if (tiling_flags & RADEON_TILING_MACRO) {
 		if (ASIC_IS_R300(rdev))
 			crtc_offset_cntl |= (R300_CRTC_X_Y_MODE_EN |
 					     R300_CRTC_MICRO_TILE_BUFFER_DIS |
@@ -270,15 +445,13 @@ int radeon_crtc_set_base(struct drm_crtc *crtc, int x, int y,
 			crtc_offset_cntl &= ~RADEON_CRTC_TILE_EN;
 	}
 
-
-	/* TODO more tiling */
-	if (0) {
+	if (tiling_flags & RADEON_TILING_MACRO) {
 		if (ASIC_IS_R300(rdev)) {
 			crtc_tile_x0_y0 = x | (y << 16);
 			base &= ~0x7ff;
 		} else {
 			int byteshift = crtc->fb->bits_per_pixel >> 4;
-			int tile_addr = (((y >> 3) * crtc->fb->width + x) >> (8 - byteshift)) << 11;
+			int tile_addr = (((y >> 3) * pitch_pixels +  x) >> (8 - byteshift)) << 11;
 			base += tile_addr + ((x << byteshift) % 256) + ((y % 8) << 8);
 			crtc_offset_cntl |= (y % 16);
 		}
@@ -303,11 +476,9 @@ int radeon_crtc_set_base(struct drm_crtc *crtc, int x, int y,
 
 	base &= ~7;
 
-	/* update sarea TODO */
-
 	crtc_offset = (u32)base;
 
-	WREG32(RADEON_DISPLAY_BASE_ADDR + radeon_crtc->crtc_offset, rdev->mc.vram_location);
+	WREG32(RADEON_DISPLAY_BASE_ADDR + radeon_crtc->crtc_offset, radeon_crtc->legacy_display_base_addr);
 
 	if (ASIC_IS_R300(rdev)) {
 		if (radeon_crtc->crtc_id)
@@ -751,6 +922,8 @@ static bool radeon_crtc_mode_fixup(struct drm_crtc *crtc,
 				   struct drm_display_mode *mode,
 				   struct drm_display_mode *adjusted_mode)
 {
+	if (!radeon_crtc_scaling_mode_fixup(crtc, mode, adjusted_mode))
+		return false;
 	return true;
 }
 
@@ -759,16 +932,25 @@ static int radeon_crtc_mode_set(struct drm_crtc *crtc,
 				 struct drm_display_mode *adjusted_mode,
 				 int x, int y, struct drm_framebuffer *old_fb)
 {
-
-	DRM_DEBUG("\n");
+	struct radeon_crtc *radeon_crtc = to_radeon_crtc(crtc);
+	struct drm_device *dev = crtc->dev;
+	struct radeon_device *rdev = dev->dev_private;
 
 	/* TODO TV */
-
 	radeon_crtc_set_base(crtc, x, y, old_fb);
 	radeon_set_crtc_timing(crtc, adjusted_mode);
 	radeon_set_pll(crtc, adjusted_mode);
-	radeon_init_disp_bandwidth(crtc->dev);
-
+	radeon_bandwidth_update(rdev);
+	if (radeon_crtc->crtc_id == 0) {
+		radeon_legacy_rmx_mode_set(crtc, mode, adjusted_mode);
+	} else {
+		if (radeon_crtc->rmx_type != RMX_OFF) {
+			/* FIXME: only first crtc has rmx what should we
+			 * do ?
+			 */
+			DRM_ERROR("Mode need scaling but only first crtc can do that.\n");
+		}
+	}
 	return 0;
 }
 
@@ -799,478 +981,3 @@ void radeon_legacy_init_crtc(struct drm_device *dev,
 		radeon_crtc->crtc_offset = RADEON_CRTC2_H_TOTAL_DISP - RADEON_CRTC_H_TOTAL_DISP;
 	drm_crtc_helper_add(&radeon_crtc->base, &legacy_helper_funcs);
 }
-
-void radeon_init_disp_bw_legacy(struct drm_device *dev,
-				struct drm_display_mode *mode1,
-				uint32_t pixel_bytes1,
-				struct drm_display_mode *mode2,
-				uint32_t pixel_bytes2)
-{
-	struct radeon_device *rdev = dev->dev_private;
-	fixed20_12 trcd_ff, trp_ff, tras_ff, trbs_ff, tcas_ff;
-	fixed20_12 sclk_ff, mclk_ff, sclk_eff_ff, sclk_delay_ff;
-	fixed20_12 peak_disp_bw, mem_bw, pix_clk, pix_clk2, temp_ff, crit_point_ff;
-	uint32_t temp, data, mem_trcd, mem_trp, mem_tras;
-	fixed20_12 memtcas_ff[8] = {
-		fixed_init(1),
-		fixed_init(2),
-		fixed_init(3),
-		fixed_init(0),
-		fixed_init_half(1),
-		fixed_init_half(2),
-		fixed_init(0),
-	};
-	fixed20_12 memtcas_rs480_ff[8] = {
-		fixed_init(0),
-		fixed_init(1),
-		fixed_init(2),
-		fixed_init(3),
-		fixed_init(0),
-		fixed_init_half(1),
-		fixed_init_half(2),
-		fixed_init_half(3),
-	};
-	fixed20_12 memtcas2_ff[8] = {
-		fixed_init(0),
-		fixed_init(1),
-		fixed_init(2),
-		fixed_init(3),
-		fixed_init(4),
-		fixed_init(5),
-		fixed_init(6),
-		fixed_init(7),
-	};
-	fixed20_12 memtrbs[8] = {
-		fixed_init(1),
-		fixed_init_half(1),
-		fixed_init(2),
-		fixed_init_half(2),
-		fixed_init(3),
-		fixed_init_half(3),
-		fixed_init(4),
-		fixed_init_half(4)
-	};
-	fixed20_12 memtrbs_r4xx[8] = {
-		fixed_init(4),
-		fixed_init(5),
-		fixed_init(6),
-		fixed_init(7),
-		fixed_init(8),
-		fixed_init(9),
-		fixed_init(10),
-		fixed_init(11)
-	};
-	fixed20_12 min_mem_eff;
-	fixed20_12 mc_latency_sclk, mc_latency_mclk, k1;
-	fixed20_12 cur_latency_mclk, cur_latency_sclk;
-	fixed20_12 disp_latency, disp_latency_overhead, disp_drain_rate,
-		disp_drain_rate2, read_return_rate;
-	fixed20_12 time_disp1_drop_priority;
-	int c;
-	int cur_size = 16;       /* in octawords */
-	int critical_point = 0, critical_point2;
-/* 	uint32_t read_return_rate, time_disp1_drop_priority; */
-	int stop_req, max_stop_req;
-
-	min_mem_eff.full = rfixed_const_8(0);
-	/* get modes */
-	if ((rdev->disp_priority == 2) && ASIC_IS_R300(rdev)) {
-		uint32_t mc_init_misc_lat_timer = RREG32(R300_MC_INIT_MISC_LAT_TIMER);
-		mc_init_misc_lat_timer &= ~(R300_MC_DISP1R_INIT_LAT_MASK << R300_MC_DISP1R_INIT_LAT_SHIFT);
-		mc_init_misc_lat_timer &= ~(R300_MC_DISP0R_INIT_LAT_MASK << R300_MC_DISP0R_INIT_LAT_SHIFT);
-		/* check crtc enables */
-		if (mode2)
-			mc_init_misc_lat_timer |= (1 << R300_MC_DISP1R_INIT_LAT_SHIFT);
-		if (mode1)
-			mc_init_misc_lat_timer |= (1 << R300_MC_DISP0R_INIT_LAT_SHIFT);
-		WREG32(R300_MC_INIT_MISC_LAT_TIMER, mc_init_misc_lat_timer);
-	}
-
-	/*
-	 * determine is there is enough bw for current mode
-	 */
-	mclk_ff.full = rfixed_const(rdev->clock.default_mclk);
-	temp_ff.full = rfixed_const(100);
-	mclk_ff.full = rfixed_div(mclk_ff, temp_ff);
-	sclk_ff.full = rfixed_const(rdev->clock.default_sclk);
-	sclk_ff.full = rfixed_div(sclk_ff, temp_ff);
-
-	temp = (rdev->mc.vram_width / 8) * (rdev->mc.vram_is_ddr ? 2 : 1);
-	temp_ff.full = rfixed_const(temp);
-	mem_bw.full = rfixed_mul(mclk_ff, temp_ff);
-
-	pix_clk.full = 0;
-	pix_clk2.full = 0;
-	peak_disp_bw.full = 0;
-	if (mode1) {
-		temp_ff.full = rfixed_const(1000);
-		pix_clk.full = rfixed_const(mode1->clock); /* convert to fixed point */
-		pix_clk.full = rfixed_div(pix_clk, temp_ff);
-		temp_ff.full = rfixed_const(pixel_bytes1);
-		peak_disp_bw.full += rfixed_mul(pix_clk, temp_ff);
-	}
-	if (mode2) {
-		temp_ff.full = rfixed_const(1000);
-		pix_clk2.full = rfixed_const(mode2->clock); /* convert to fixed point */
-		pix_clk2.full = rfixed_div(pix_clk2, temp_ff);
-		temp_ff.full = rfixed_const(pixel_bytes2);
-		peak_disp_bw.full += rfixed_mul(pix_clk2, temp_ff);
-	}
-
-	mem_bw.full = rfixed_mul(mem_bw, min_mem_eff);
-	if (peak_disp_bw.full >= mem_bw.full) {
-		DRM_ERROR("You may not have enough display bandwidth for current mode\n"
-			  "If you have flickering problem, try to lower resolution, refresh rate, or color depth\n");
-	}
-
-	/*  Get values from the EXT_MEM_CNTL register...converting its contents. */
-	temp = RREG32(RADEON_MEM_TIMING_CNTL);
-	if ((rdev->family == CHIP_RV100) || (rdev->flags & RADEON_IS_IGP)) { /* RV100, M6, IGPs */
-		mem_trcd = ((temp >> 2) & 0x3) + 1;
-		mem_trp  = ((temp & 0x3)) + 1;
-		mem_tras = ((temp & 0x70) >> 4) + 1;
-	} else if (rdev->family == CHIP_R300 ||
-		   rdev->family == CHIP_R350) { /* r300, r350 */
-		mem_trcd = (temp & 0x7) + 1;
-		mem_trp = ((temp >> 8) & 0x7) + 1;
-		mem_tras = ((temp >> 11) & 0xf) + 4;
-	} else if (rdev->family == CHIP_RV350 ||
-		   rdev->family <= CHIP_RV380) {
-		/* rv3x0 */
-		mem_trcd = (temp & 0x7) + 3;
-		mem_trp = ((temp >> 8) & 0x7) + 3;
-		mem_tras = ((temp >> 11) & 0xf) + 6;
-	} else if (rdev->family == CHIP_R420 ||
-		   rdev->family == CHIP_R423 ||
-		   rdev->family == CHIP_RV410) {
-		/* r4xx */
-		mem_trcd = (temp & 0xf) + 3;
-		if (mem_trcd > 15)
-			mem_trcd = 15;
-		mem_trp = ((temp >> 8) & 0xf) + 3;
-		if (mem_trp > 15)
-			mem_trp = 15;
-		mem_tras = ((temp >> 12) & 0x1f) + 6;
-		if (mem_tras > 31)
-			mem_tras = 31;
-	} else { /* RV200, R200 */
-		mem_trcd = (temp & 0x7) + 1;
-		mem_trp = ((temp >> 8) & 0x7) + 1;
-		mem_tras = ((temp >> 12) & 0xf) + 4;
-	}
-	/* convert to FF */
-	trcd_ff.full = rfixed_const(mem_trcd);
-	trp_ff.full = rfixed_const(mem_trp);
-	tras_ff.full = rfixed_const(mem_tras);
-
-	/* Get values from the MEM_SDRAM_MODE_REG register...converting its */
-	temp = RREG32(RADEON_MEM_SDRAM_MODE_REG);
-	data = (temp & (7 << 20)) >> 20;
-	if ((rdev->family == CHIP_RV100) || rdev->flags & RADEON_IS_IGP) {
-		if (rdev->family == CHIP_RS480) /* don't think rs400 */
-			tcas_ff = memtcas_rs480_ff[data];
-		else
-			tcas_ff = memtcas_ff[data];
-	} else
-		tcas_ff = memtcas2_ff[data];
-
-	if (rdev->family == CHIP_RS400 ||
-	    rdev->family == CHIP_RS480) {
-		/* extra cas latency stored in bits 23-25 0-4 clocks */
-		data = (temp >> 23) & 0x7;
-		if (data < 5)
-			tcas_ff.full += rfixed_const(data);
-	}
-
-	if (ASIC_IS_R300(rdev) && !(rdev->flags & RADEON_IS_IGP)) {
-		/* on the R300, Tcas is included in Trbs.
-		 */
-		temp = RREG32(RADEON_MEM_CNTL);
-		data = (R300_MEM_NUM_CHANNELS_MASK & temp);
-		if (data == 1) {
-			if (R300_MEM_USE_CD_CH_ONLY & temp) {
-				temp = RREG32(R300_MC_IND_INDEX);
-				temp &= ~R300_MC_IND_ADDR_MASK;
-				temp |= R300_MC_READ_CNTL_CD_mcind;
-				WREG32(R300_MC_IND_INDEX, temp);
-				temp = RREG32(R300_MC_IND_DATA);
-				data = (R300_MEM_RBS_POSITION_C_MASK & temp);
-			} else {
-				temp = RREG32(R300_MC_READ_CNTL_AB);
-				data = (R300_MEM_RBS_POSITION_A_MASK & temp);
-			}
-		} else {
-			temp = RREG32(R300_MC_READ_CNTL_AB);
-			data = (R300_MEM_RBS_POSITION_A_MASK & temp);
-		}
-		if (rdev->family == CHIP_RV410 ||
-		    rdev->family == CHIP_R420 ||
-		    rdev->family == CHIP_R423)
-			trbs_ff = memtrbs_r4xx[data];
-		else
-			trbs_ff = memtrbs[data];
-		tcas_ff.full += trbs_ff.full;
-	}
-
-	sclk_eff_ff.full = sclk_ff.full;
-
-	if (rdev->flags & RADEON_IS_AGP) {
-		fixed20_12 agpmode_ff;
-		agpmode_ff.full = rfixed_const(radeon_agpmode);
-		temp_ff.full = rfixed_const_666(16);
-		sclk_eff_ff.full -= rfixed_mul(agpmode_ff, temp_ff);
-	}
-	/* TODO PCIE lanes may affect this - agpmode == 16?? */
-
-	if (ASIC_IS_R300(rdev)) {
-		sclk_delay_ff.full = rfixed_const(250);
-	} else {
-		if ((rdev->family == CHIP_RV100) ||
-		    rdev->flags & RADEON_IS_IGP) {
-			if (rdev->mc.vram_is_ddr)
-				sclk_delay_ff.full = rfixed_const(41);
-			else
-				sclk_delay_ff.full = rfixed_const(33);
-		} else {
-			if (rdev->mc.vram_width == 128)
-				sclk_delay_ff.full = rfixed_const(57);
-			else
-				sclk_delay_ff.full = rfixed_const(41);
-		}
-	}
-
-	mc_latency_sclk.full = rfixed_div(sclk_delay_ff, sclk_eff_ff);
-
-	if (rdev->mc.vram_is_ddr) {
-		if (rdev->mc.vram_width == 32) {
-			k1.full = rfixed_const(40);
-			c  = 3;
-		} else {
-			k1.full = rfixed_const(20);
-			c  = 1;
-		}
-	} else {
-		k1.full = rfixed_const(40);
-		c  = 3;
-	}
-
-	temp_ff.full = rfixed_const(2);
-	mc_latency_mclk.full = rfixed_mul(trcd_ff, temp_ff);
-	temp_ff.full = rfixed_const(c);
-	mc_latency_mclk.full += rfixed_mul(tcas_ff, temp_ff);
-	temp_ff.full = rfixed_const(4);
-	mc_latency_mclk.full += rfixed_mul(tras_ff, temp_ff);
-	mc_latency_mclk.full += rfixed_mul(trp_ff, temp_ff);
-	mc_latency_mclk.full += k1.full;
-
-	mc_latency_mclk.full = rfixed_div(mc_latency_mclk, mclk_ff);
-	mc_latency_mclk.full += rfixed_div(temp_ff, sclk_eff_ff);
-
-	/*
-	  HW cursor time assuming worst case of full size colour cursor.
-	*/
-	temp_ff.full = rfixed_const((2 * (cur_size - (rdev->mc.vram_is_ddr + 1))));
-	temp_ff.full += trcd_ff.full;
-	if (temp_ff.full < tras_ff.full)
-		temp_ff.full = tras_ff.full;
-	cur_latency_mclk.full = rfixed_div(temp_ff, mclk_ff);
-
-	temp_ff.full = rfixed_const(cur_size);
-	cur_latency_sclk.full = rfixed_div(temp_ff, sclk_eff_ff);
-	/*
-	  Find the total latency for the display data.
-	*/
-	disp_latency_overhead.full = rfixed_const(80);
-	disp_latency_overhead.full = rfixed_div(disp_latency_overhead, sclk_ff);
-	mc_latency_mclk.full += disp_latency_overhead.full + cur_latency_mclk.full;
-	mc_latency_sclk.full += disp_latency_overhead.full + cur_latency_sclk.full;
-
-	if (mc_latency_mclk.full > mc_latency_sclk.full)
-		disp_latency.full = mc_latency_mclk.full;
-	else
-		disp_latency.full = mc_latency_sclk.full;
-
-	/* setup Max GRPH_STOP_REQ default value */
-	if (ASIC_IS_RV100(rdev))
-		max_stop_req = 0x5c;
-	else
-		max_stop_req = 0x7c;
-
-	if (mode1) {
-		/*  CRTC1
-		    Set GRPH_BUFFER_CNTL register using h/w defined optimal values.
-		    GRPH_STOP_REQ <= MIN[ 0x7C, (CRTC_H_DISP + 1) * (bit depth) / 0x10 ]
-		*/
-		stop_req = mode1->hdisplay * pixel_bytes1 / 16;
-
-		if (stop_req > max_stop_req)
-			stop_req = max_stop_req;
-
-		/*
-		  Find the drain rate of the display buffer.
-		*/
-		temp_ff.full = rfixed_const((16/pixel_bytes1));
-		disp_drain_rate.full = rfixed_div(pix_clk, temp_ff);
-
-		/*
-		  Find the critical point of the display buffer.
-		*/
-		crit_point_ff.full = rfixed_mul(disp_drain_rate, disp_latency);
-		crit_point_ff.full += rfixed_const_half(0);
-
-		critical_point = rfixed_trunc(crit_point_ff);
-
-		if (rdev->disp_priority == 2) {
-			critical_point = 0;
-		}
-
-		/*
-		  The critical point should never be above max_stop_req-4.  Setting
-		  GRPH_CRITICAL_CNTL = 0 will thus force high priority all the time.
-		*/
-		if (max_stop_req - critical_point < 4)
-			critical_point = 0;
-
-		if (critical_point == 0 && mode2 && rdev->family == CHIP_R300) {
-			/* some R300 cards have problem with this set to 0, when CRTC2 is enabled.*/
-			critical_point = 0x10;
-		}
-
-		temp = RREG32(RADEON_GRPH_BUFFER_CNTL);
-		temp &= ~(RADEON_GRPH_STOP_REQ_MASK);
-		temp |= (stop_req << RADEON_GRPH_STOP_REQ_SHIFT);
-		temp &= ~(RADEON_GRPH_START_REQ_MASK);
-		if ((rdev->family == CHIP_R350) &&
-		    (stop_req > 0x15)) {
-			stop_req -= 0x10;
-		}
-		temp |= (stop_req << RADEON_GRPH_START_REQ_SHIFT);
-		temp |= RADEON_GRPH_BUFFER_SIZE;
-		temp &= ~(RADEON_GRPH_CRITICAL_CNTL   |
-			  RADEON_GRPH_CRITICAL_AT_SOF |
-			  RADEON_GRPH_STOP_CNTL);
-		/*
-		  Write the result into the register.
-		*/
-		WREG32(RADEON_GRPH_BUFFER_CNTL, ((temp & ~RADEON_GRPH_CRITICAL_POINT_MASK) |
-						       (critical_point << RADEON_GRPH_CRITICAL_POINT_SHIFT)));
-
-#if 0
-		if ((rdev->family == CHIP_RS400) ||
-		    (rdev->family == CHIP_RS480)) {
-			/* attempt to program RS400 disp regs correctly ??? */
-			temp = RREG32(RS400_DISP1_REG_CNTL);
-			temp &= ~(RS400_DISP1_START_REQ_LEVEL_MASK |
-				  RS400_DISP1_STOP_REQ_LEVEL_MASK);
-			WREG32(RS400_DISP1_REQ_CNTL1, (temp |
-						       (critical_point << RS400_DISP1_START_REQ_LEVEL_SHIFT) |
-						       (critical_point << RS400_DISP1_STOP_REQ_LEVEL_SHIFT)));
-			temp = RREG32(RS400_DMIF_MEM_CNTL1);
-			temp &= ~(RS400_DISP1_CRITICAL_POINT_START_MASK |
-				  RS400_DISP1_CRITICAL_POINT_STOP_MASK);
-			WREG32(RS400_DMIF_MEM_CNTL1, (temp |
-						      (critical_point << RS400_DISP1_CRITICAL_POINT_START_SHIFT) |
-						      (critical_point << RS400_DISP1_CRITICAL_POINT_STOP_SHIFT)));
-		}
-#endif
-
-		DRM_DEBUG("GRPH_BUFFER_CNTL from to %x\n",
-			  /* 	  (unsigned int)info->SavedReg->grph_buffer_cntl, */
-			  (unsigned int)RREG32(RADEON_GRPH_BUFFER_CNTL));
-	}
-
-	if (mode2) {
-		u32 grph2_cntl;
-		stop_req = mode2->hdisplay * pixel_bytes2 / 16;
-
-		if (stop_req > max_stop_req)
-			stop_req = max_stop_req;
-
-		/*
-		  Find the drain rate of the display buffer.
-		*/
-		temp_ff.full = rfixed_const((16/pixel_bytes2));
-		disp_drain_rate2.full = rfixed_div(pix_clk2, temp_ff);
-
-		grph2_cntl = RREG32(RADEON_GRPH2_BUFFER_CNTL);
-		grph2_cntl &= ~(RADEON_GRPH_STOP_REQ_MASK);
-		grph2_cntl |= (stop_req << RADEON_GRPH_STOP_REQ_SHIFT);
-		grph2_cntl &= ~(RADEON_GRPH_START_REQ_MASK);
-		if ((rdev->family == CHIP_R350) &&
-		    (stop_req > 0x15)) {
-			stop_req -= 0x10;
-		}
-		grph2_cntl |= (stop_req << RADEON_GRPH_START_REQ_SHIFT);
-		grph2_cntl |= RADEON_GRPH_BUFFER_SIZE;
-		grph2_cntl &= ~(RADEON_GRPH_CRITICAL_CNTL   |
-			  RADEON_GRPH_CRITICAL_AT_SOF |
-			  RADEON_GRPH_STOP_CNTL);
-
-		if ((rdev->family == CHIP_RS100) ||
-		    (rdev->family == CHIP_RS200))
-			critical_point2 = 0;
-		else {
-			temp = (rdev->mc.vram_width * rdev->mc.vram_is_ddr + 1)/128;
-			temp_ff.full = rfixed_const(temp);
-			temp_ff.full = rfixed_mul(mclk_ff, temp_ff);
-			if (sclk_ff.full < temp_ff.full)
-				temp_ff.full = sclk_ff.full;
-
-			read_return_rate.full = temp_ff.full;
-
-			if (mode1) {
-				temp_ff.full = read_return_rate.full - disp_drain_rate.full;
-				time_disp1_drop_priority.full = rfixed_div(crit_point_ff, temp_ff);
-			} else {
-				time_disp1_drop_priority.full = 0;
-			}
-			crit_point_ff.full = disp_latency.full + time_disp1_drop_priority.full + disp_latency.full;
-			crit_point_ff.full = rfixed_mul(crit_point_ff, disp_drain_rate2);
-			crit_point_ff.full += rfixed_const_half(0);
-
-			critical_point2 = rfixed_trunc(crit_point_ff);
-
-			if (rdev->disp_priority == 2) {
-				critical_point2 = 0;
-			}
-
-			if (max_stop_req - critical_point2 < 4)
-				critical_point2 = 0;
-
-		}
-
-		if (critical_point2 == 0 && rdev->family == CHIP_R300) {
-			/* some R300 cards have problem with this set to 0 */
-			critical_point2 = 0x10;
-		}
-
-		WREG32(RADEON_GRPH2_BUFFER_CNTL, ((grph2_cntl & ~RADEON_GRPH_CRITICAL_POINT_MASK) |
-						  (critical_point2 << RADEON_GRPH_CRITICAL_POINT_SHIFT)));
-
-		if ((rdev->family == CHIP_RS400) ||
-		    (rdev->family == CHIP_RS480)) {
-#if 0
-			/* attempt to program RS400 disp2 regs correctly ??? */
-			temp = RREG32(RS400_DISP2_REQ_CNTL1);
-			temp &= ~(RS400_DISP2_START_REQ_LEVEL_MASK |
-				  RS400_DISP2_STOP_REQ_LEVEL_MASK);
-			WREG32(RS400_DISP2_REQ_CNTL1, (temp |
-						       (critical_point2 << RS400_DISP1_START_REQ_LEVEL_SHIFT) |
-						       (critical_point2 << RS400_DISP1_STOP_REQ_LEVEL_SHIFT)));
-			temp = RREG32(RS400_DISP2_REQ_CNTL2);
-			temp &= ~(RS400_DISP2_CRITICAL_POINT_START_MASK |
-				  RS400_DISP2_CRITICAL_POINT_STOP_MASK);
-			WREG32(RS400_DISP2_REQ_CNTL2, (temp |
-						       (critical_point2 << RS400_DISP2_CRITICAL_POINT_START_SHIFT) |
-						       (critical_point2 << RS400_DISP2_CRITICAL_POINT_STOP_SHIFT)));
-#endif
-			WREG32(RS400_DISP2_REQ_CNTL1, 0x105DC1CC);
-			WREG32(RS400_DISP2_REQ_CNTL2, 0x2749D000);
-			WREG32(RS400_DMIF_MEM_CNTL1,  0x29CA71DC);
-			WREG32(RS400_DISP1_REQ_CNTL1, 0x28FBC3AC);
-		}
-
-		DRM_DEBUG("GRPH2_BUFFER_CNTL from to %x\n",
-			  (unsigned int)RREG32(RADEON_GRPH2_BUFFER_CNTL));
-	}
-}
diff --git a/drivers/gpu/drm/radeon/radeon_legacy_encoders.c b/drivers/gpu/drm/radeon/radeon_legacy_encoders.c
index 2c2f42de1d4..34d0f58eb94 100644
--- a/drivers/gpu/drm/radeon/radeon_legacy_encoders.c
+++ b/drivers/gpu/drm/radeon/radeon_legacy_encoders.c
@@ -30,170 +30,6 @@
 #include "atom.h"
 
 
-static void radeon_legacy_rmx_mode_set(struct drm_encoder *encoder,
-				       struct drm_display_mode *mode,
-				       struct drm_display_mode *adjusted_mode)
-{
-	struct drm_device *dev = encoder->dev;
-	struct radeon_device *rdev = dev->dev_private;
-	struct radeon_encoder *radeon_encoder = to_radeon_encoder(encoder);
-	int    xres = mode->hdisplay;
-	int    yres = mode->vdisplay;
-	bool   hscale = true, vscale = true;
-	int    hsync_wid;
-	int    vsync_wid;
-	int    hsync_start;
-	uint32_t scale, inc;
-	uint32_t fp_horz_stretch, fp_vert_stretch, crtc_more_cntl, fp_horz_vert_active;
-	uint32_t fp_h_sync_strt_wid, fp_v_sync_strt_wid, fp_crtc_h_total_disp, fp_crtc_v_total_disp;
-	struct radeon_native_mode *native_mode = &radeon_encoder->native_mode;
-
-	DRM_DEBUG("\n");
-
-	fp_vert_stretch = RREG32(RADEON_FP_VERT_STRETCH) &
-		(RADEON_VERT_STRETCH_RESERVED |
-		 RADEON_VERT_AUTO_RATIO_INC);
-	fp_horz_stretch = RREG32(RADEON_FP_HORZ_STRETCH) &
-		(RADEON_HORZ_FP_LOOP_STRETCH |
-		 RADEON_HORZ_AUTO_RATIO_INC);
-
-	crtc_more_cntl = 0;
-	if ((rdev->family == CHIP_RS100) ||
-	    (rdev->family == CHIP_RS200)) {
-		/* This is to workaround the asic bug for RMX, some versions
-		   of BIOS dosen't have this register initialized correctly. */
-		crtc_more_cntl |= RADEON_CRTC_H_CUTOFF_ACTIVE_EN;
-	}
-
-
-	fp_crtc_h_total_disp = ((((mode->crtc_htotal / 8) - 1) & 0x3ff)
-				| ((((mode->crtc_hdisplay / 8) - 1) & 0x1ff) << 16));
-
-	hsync_wid = (mode->crtc_hsync_end - mode->crtc_hsync_start) / 8;
-	if (!hsync_wid)
-		hsync_wid = 1;
-	hsync_start = mode->crtc_hsync_start - 8;
-
-	fp_h_sync_strt_wid = ((hsync_start & 0x1fff)
-			      | ((hsync_wid & 0x3f) << 16)
-			      | ((mode->flags & DRM_MODE_FLAG_NHSYNC)
-				 ? RADEON_CRTC_H_SYNC_POL
-				 : 0));
-
-	fp_crtc_v_total_disp = (((mode->crtc_vtotal - 1) & 0xffff)
-				| ((mode->crtc_vdisplay - 1) << 16));
-
-	vsync_wid = mode->crtc_vsync_end - mode->crtc_vsync_start;
-	if (!vsync_wid)
-		vsync_wid = 1;
-
-	fp_v_sync_strt_wid = (((mode->crtc_vsync_start - 1) & 0xfff)
-			      | ((vsync_wid & 0x1f) << 16)
-			      | ((mode->flags & DRM_MODE_FLAG_NVSYNC)
-				 ? RADEON_CRTC_V_SYNC_POL
-				 : 0));
-
-	fp_horz_vert_active = 0;
-
-	if (native_mode->panel_xres == 0 ||
-	    native_mode->panel_yres == 0) {
-		hscale = false;
-		vscale = false;
-	} else {
-		if (xres > native_mode->panel_xres)
-			xres = native_mode->panel_xres;
-		if (yres > native_mode->panel_yres)
-			yres = native_mode->panel_yres;
-
-		if (xres == native_mode->panel_xres)
-			hscale = false;
-		if (yres == native_mode->panel_yres)
-			vscale = false;
-	}
-
-	if (radeon_encoder->flags & RADEON_USE_RMX) {
-		if (radeon_encoder->rmx_type != RMX_CENTER) {
-			if (!hscale)
-				fp_horz_stretch |= ((xres/8-1) << 16);
-			else {
-				inc = (fp_horz_stretch & RADEON_HORZ_AUTO_RATIO_INC) ? 1 : 0;
-				scale = ((xres + inc) * RADEON_HORZ_STRETCH_RATIO_MAX)
-					/ native_mode->panel_xres + 1;
-				fp_horz_stretch |= (((scale) & RADEON_HORZ_STRETCH_RATIO_MASK) |
-						    RADEON_HORZ_STRETCH_BLEND |
-						    RADEON_HORZ_STRETCH_ENABLE |
-						    ((native_mode->panel_xres/8-1) << 16));
-			}
-
-			if (!vscale)
-				fp_vert_stretch |= ((yres-1) << 12);
-			else {
-				inc = (fp_vert_stretch & RADEON_VERT_AUTO_RATIO_INC) ? 1 : 0;
-				scale = ((yres + inc) * RADEON_VERT_STRETCH_RATIO_MAX)
-					/ native_mode->panel_yres + 1;
-				fp_vert_stretch |= (((scale) & RADEON_VERT_STRETCH_RATIO_MASK) |
-						    RADEON_VERT_STRETCH_ENABLE |
-						    RADEON_VERT_STRETCH_BLEND |
-						    ((native_mode->panel_yres-1) << 12));
-			}
-		} else if (radeon_encoder->rmx_type == RMX_CENTER) {
-			int    blank_width;
-
-			fp_horz_stretch |= ((xres/8-1) << 16);
-			fp_vert_stretch |= ((yres-1) << 12);
-
-			crtc_more_cntl |= (RADEON_CRTC_AUTO_HORZ_CENTER_EN |
-					   RADEON_CRTC_AUTO_VERT_CENTER_EN);
-
-			blank_width = (mode->crtc_hblank_end - mode->crtc_hblank_start) / 8;
-			if (blank_width > 110)
-				blank_width = 110;
-
-			fp_crtc_h_total_disp = (((blank_width) & 0x3ff)
-						| ((((mode->crtc_hdisplay / 8) - 1) & 0x1ff) << 16));
-
-			hsync_wid = (mode->crtc_hsync_end - mode->crtc_hsync_start) / 8;
-			if (!hsync_wid)
-				hsync_wid = 1;
-
-			fp_h_sync_strt_wid = ((((mode->crtc_hsync_start - mode->crtc_hblank_start) / 8) & 0x1fff)
-					      | ((hsync_wid & 0x3f) << 16)
-					      | ((mode->flags & DRM_MODE_FLAG_NHSYNC)
-						 ? RADEON_CRTC_H_SYNC_POL
-						 : 0));
-
-			fp_crtc_v_total_disp = (((mode->crtc_vblank_end - mode->crtc_vblank_start) & 0xffff)
-						| ((mode->crtc_vdisplay - 1) << 16));
-
-			vsync_wid = mode->crtc_vsync_end - mode->crtc_vsync_start;
-			if (!vsync_wid)
-				vsync_wid = 1;
-
-			fp_v_sync_strt_wid = ((((mode->crtc_vsync_start - mode->crtc_vblank_start) & 0xfff)
-					       | ((vsync_wid & 0x1f) << 16)
-					       | ((mode->flags & DRM_MODE_FLAG_NVSYNC)
-						  ? RADEON_CRTC_V_SYNC_POL
-						  : 0)));
-
-			fp_horz_vert_active = (((native_mode->panel_yres) & 0xfff) |
-					       (((native_mode->panel_xres / 8) & 0x1ff) << 16));
-		}
-	} else {
-		fp_horz_stretch |= ((xres/8-1) << 16);
-		fp_vert_stretch |= ((yres-1) << 12);
-	}
-
-	WREG32(RADEON_FP_HORZ_STRETCH,      fp_horz_stretch);
-	WREG32(RADEON_FP_VERT_STRETCH,      fp_vert_stretch);
-	WREG32(RADEON_CRTC_MORE_CNTL,       crtc_more_cntl);
-	WREG32(RADEON_FP_HORZ_VERT_ACTIVE,  fp_horz_vert_active);
-	WREG32(RADEON_FP_H_SYNC_STRT_WID,   fp_h_sync_strt_wid);
-	WREG32(RADEON_FP_V_SYNC_STRT_WID,   fp_v_sync_strt_wid);
-	WREG32(RADEON_FP_CRTC_H_TOTAL_DISP, fp_crtc_h_total_disp);
-	WREG32(RADEON_FP_CRTC_V_TOTAL_DISP, fp_crtc_v_total_disp);
-
-}
-
 static void radeon_legacy_lvds_dpms(struct drm_encoder *encoder, int mode)
 {
 	struct drm_device *dev = encoder->dev;
@@ -287,9 +123,6 @@ static void radeon_legacy_lvds_mode_set(struct drm_encoder *encoder,
 
 	DRM_DEBUG("\n");
 
-	if (radeon_crtc->crtc_id == 0)
-		radeon_legacy_rmx_mode_set(encoder, mode, adjusted_mode);
-
 	lvds_pll_cntl = RREG32(RADEON_LVDS_PLL_CNTL);
 	lvds_pll_cntl &= ~RADEON_LVDS_PLL_EN;
 
@@ -318,7 +151,7 @@ static void radeon_legacy_lvds_mode_set(struct drm_encoder *encoder,
 
 	if (radeon_crtc->crtc_id == 0) {
 		if (ASIC_IS_R300(rdev)) {
-			if (radeon_encoder->flags & RADEON_USE_RMX)
+			if (radeon_encoder->rmx_type != RMX_OFF)
 				lvds_pll_cntl |= R300_LVDS_SRC_SEL_RMX;
 		} else
 			lvds_gen_cntl &= ~RADEON_LVDS_SEL_CRTC2;
@@ -350,8 +183,6 @@ static bool radeon_legacy_lvds_mode_fixup(struct drm_encoder *encoder,
 
 	drm_mode_set_crtcinfo(adjusted_mode, 0);
 
-	radeon_encoder->flags &= ~RADEON_USE_RMX;
-
 	if (radeon_encoder->rmx_type != RMX_OFF)
 		radeon_rmx_mode_fixup(encoder, mode, adjusted_mode);
 
@@ -455,9 +286,6 @@ static void radeon_legacy_primary_dac_mode_set(struct drm_encoder *encoder,
 
 	DRM_DEBUG("\n");
 
-	if (radeon_crtc->crtc_id == 0)
-		radeon_legacy_rmx_mode_set(encoder, mode, adjusted_mode);
-
 	if (radeon_crtc->crtc_id == 0) {
 		if (rdev->family == CHIP_R200 || ASIC_IS_R300(rdev)) {
 			disp_output_cntl = RREG32(RADEON_DISP_OUTPUT_CNTL) &
@@ -653,9 +481,6 @@ static void radeon_legacy_tmds_int_mode_set(struct drm_encoder *encoder,
 
 	DRM_DEBUG("\n");
 
-	if (radeon_crtc->crtc_id == 0)
-		radeon_legacy_rmx_mode_set(encoder, mode, adjusted_mode);
-
 	tmp = tmds_pll_cntl = RREG32(RADEON_TMDS_PLL_CNTL);
 	tmp &= 0xfffff;
 	if (rdev->family == CHIP_RV280) {
@@ -711,7 +536,7 @@ static void radeon_legacy_tmds_int_mode_set(struct drm_encoder *encoder,
     if (radeon_crtc->crtc_id == 0) {
 	    if (ASIC_IS_R300(rdev) || rdev->family == CHIP_R200) {
 		    fp_gen_cntl &= ~R200_FP_SOURCE_SEL_MASK;
-		    if (radeon_encoder->flags & RADEON_USE_RMX)
+		    if (radeon_encoder->rmx_type != RMX_OFF)
 			    fp_gen_cntl |= R200_FP_SOURCE_SEL_RMX;
 		    else
 			    fp_gen_cntl |= R200_FP_SOURCE_SEL_CRTC1;
@@ -820,9 +645,6 @@ static void radeon_legacy_tmds_ext_mode_set(struct drm_encoder *encoder,
 
 	DRM_DEBUG("\n");
 
-	if (radeon_crtc->crtc_id == 0)
-		radeon_legacy_rmx_mode_set(encoder, mode, adjusted_mode);
-
 	if (rdev->is_atom_bios) {
 		radeon_encoder->pixel_clock = adjusted_mode->clock;
 		atombios_external_tmds_setup(encoder, ATOM_ENABLE);
@@ -856,7 +678,7 @@ static void radeon_legacy_tmds_ext_mode_set(struct drm_encoder *encoder,
 	if (radeon_crtc->crtc_id == 0) {
 		if ((rdev->family == CHIP_R200) || ASIC_IS_R300(rdev)) {
 			fp2_gen_cntl &= ~R200_FP2_SOURCE_SEL_MASK;
-			if (radeon_encoder->flags & RADEON_USE_RMX)
+			if (radeon_encoder->rmx_type != RMX_OFF)
 				fp2_gen_cntl |= R200_FP2_SOURCE_SEL_RMX;
 			else
 				fp2_gen_cntl |= R200_FP2_SOURCE_SEL_CRTC1;
@@ -1014,9 +836,6 @@ static void radeon_legacy_tv_dac_mode_set(struct drm_encoder *encoder,
 
 	DRM_DEBUG("\n");
 
-	if (radeon_crtc->crtc_id == 0)
-		radeon_legacy_rmx_mode_set(encoder, mode, adjusted_mode);
-
 	if (rdev->family != CHIP_R200) {
 		tv_dac_cntl = RREG32(RADEON_TV_DAC_CNTL);
 		if (rdev->family == CHIP_R420 ||
@@ -1243,6 +1062,7 @@ radeon_add_legacy_encoder(struct drm_device *dev, uint32_t encoder_id, uint32_t
 
 	radeon_encoder->encoder_id = encoder_id;
 	radeon_encoder->devices = supported_device;
+	radeon_encoder->rmx_type = RMX_OFF;
 
 	switch (radeon_encoder->encoder_id) {
 	case ENCODER_OBJECT_ID_INTERNAL_LVDS:
diff --git a/drivers/gpu/drm/radeon/radeon_mode.h b/drivers/gpu/drm/radeon/radeon_mode.h
index 9173b687462..3b09a1f2d8f 100644
--- a/drivers/gpu/drm/radeon/radeon_mode.h
+++ b/drivers/gpu/drm/radeon/radeon_mode.h
@@ -36,6 +36,9 @@
 #include <linux/i2c.h>
 #include <linux/i2c-id.h>
 #include <linux/i2c-algo-bit.h>
+#include "radeon_fixed.h"
+
+struct radeon_device;
 
 #define to_radeon_crtc(x) container_of(x, struct radeon_crtc, base)
 #define to_radeon_connector(x) container_of(x, struct radeon_connector, base)
@@ -124,6 +127,7 @@ struct radeon_tmds_pll {
 #define RADEON_PLL_PREFER_LOW_POST_DIV  (1 << 8)
 #define RADEON_PLL_PREFER_HIGH_POST_DIV (1 << 9)
 #define RADEON_PLL_USE_FRAC_FB_DIV      (1 << 10)
+#define RADEON_PLL_PREFER_CLOSEST_LOWER (1 << 11)
 
 struct radeon_pll {
 	uint16_t reference_freq;
@@ -170,6 +174,18 @@ struct radeon_mode_info {
 	struct atom_context *atom_context;
 	enum radeon_connector_table connector_table;
 	bool mode_config_initialized;
+	struct radeon_crtc *crtcs[2];
+};
+
+struct radeon_native_mode {
+	/* preferred mode */
+	uint32_t panel_xres, panel_yres;
+	uint32_t hoverplus, hsync_width;
+	uint32_t hblank;
+	uint32_t voverplus, vsync_width;
+	uint32_t vblank;
+	uint32_t dotclock;
+	uint32_t flags;
 };
 
 struct radeon_crtc {
@@ -185,19 +201,13 @@ struct radeon_crtc {
 	uint64_t cursor_addr;
 	int cursor_width;
 	int cursor_height;
-};
-
-#define RADEON_USE_RMX 1
-
-struct radeon_native_mode {
-	/* preferred mode */
-	uint32_t panel_xres, panel_yres;
-	uint32_t hoverplus, hsync_width;
-	uint32_t hblank;
-	uint32_t voverplus, vsync_width;
-	uint32_t vblank;
-	uint32_t dotclock;
-	uint32_t flags;
+	uint32_t legacy_display_base_addr;
+	uint32_t legacy_cursor_offset;
+	enum radeon_rmx_type rmx_type;
+	uint32_t devices;
+	fixed20_12 vsc;
+	fixed20_12 hsc;
+	struct radeon_native_mode native_mode;
 };
 
 struct radeon_encoder_primary_dac {
@@ -383,16 +393,9 @@ void radeon_enc_destroy(struct drm_encoder *encoder);
 void radeon_copy_fb(struct drm_device *dev, struct drm_gem_object *dst_obj);
 void radeon_combios_asic_init(struct drm_device *dev);
 extern int radeon_static_clocks_init(struct drm_device *dev);
-void radeon_init_disp_bw_legacy(struct drm_device *dev,
-				struct drm_display_mode *mode1,
-				uint32_t pixel_bytes1,
-				struct drm_display_mode *mode2,
-				uint32_t pixel_bytes2);
-void radeon_init_disp_bw_avivo(struct drm_device *dev,
-			       struct drm_display_mode *mode1,
-			       uint32_t pixel_bytes1,
-			       struct drm_display_mode *mode2,
-			       uint32_t pixel_bytes2);
-void radeon_init_disp_bandwidth(struct drm_device *dev);
+bool radeon_crtc_scaling_mode_fixup(struct drm_crtc *crtc,
+					struct drm_display_mode *mode,
+					struct drm_display_mode *adjusted_mode);
+void atom_rv515_force_tv_scaler(struct radeon_device *rdev);
 
 #endif
diff --git a/drivers/gpu/drm/radeon/radeon_object.c b/drivers/gpu/drm/radeon/radeon_object.c
index bac0d06c52a..e98cae3bf4a 100644
--- a/drivers/gpu/drm/radeon/radeon_object.c
+++ b/drivers/gpu/drm/radeon/radeon_object.c
@@ -44,6 +44,9 @@ struct radeon_object {
 	uint64_t			gpu_addr;
 	void				*kptr;
 	bool				is_iomem;
+	uint32_t			tiling_flags;
+	uint32_t			pitch;
+	int				surface_reg;
 };
 
 int radeon_ttm_init(struct radeon_device *rdev);
@@ -70,6 +73,7 @@ static void radeon_ttm_object_object_destroy(struct ttm_buffer_object *tobj)
 
 	robj = container_of(tobj, struct radeon_object, tobj);
 	list_del_init(&robj->list);
+	radeon_object_clear_surface_reg(robj);
 	kfree(robj);
 }
 
@@ -99,16 +103,16 @@ static inline uint32_t radeon_object_flags_from_domain(uint32_t domain)
 {
 	uint32_t flags = 0;
 	if (domain & RADEON_GEM_DOMAIN_VRAM) {
-		flags |= TTM_PL_FLAG_VRAM;
+		flags |= TTM_PL_FLAG_VRAM | TTM_PL_FLAG_WC | TTM_PL_FLAG_UNCACHED;
 	}
 	if (domain & RADEON_GEM_DOMAIN_GTT) {
-		flags |= TTM_PL_FLAG_TT;
+		flags |= TTM_PL_FLAG_TT | TTM_PL_MASK_CACHING;
 	}
 	if (domain & RADEON_GEM_DOMAIN_CPU) {
-		flags |= TTM_PL_FLAG_SYSTEM;
+		flags |= TTM_PL_FLAG_SYSTEM | TTM_PL_MASK_CACHING;
 	}
 	if (!flags) {
-		flags |= TTM_PL_FLAG_SYSTEM;
+		flags |= TTM_PL_FLAG_SYSTEM | TTM_PL_MASK_CACHING;
 	}
 	return flags;
 }
@@ -141,6 +145,7 @@ int radeon_object_create(struct radeon_device *rdev,
 	}
 	robj->rdev = rdev;
 	robj->gobj = gobj;
+	robj->surface_reg = -1;
 	INIT_LIST_HEAD(&robj->list);
 
 	flags = radeon_object_flags_from_domain(domain);
@@ -304,7 +309,7 @@ int radeon_object_wait(struct radeon_object *robj)
 	}
 	spin_lock(&robj->tobj.lock);
 	if (robj->tobj.sync_obj) {
-		r = ttm_bo_wait(&robj->tobj, true, false, false);
+		r = ttm_bo_wait(&robj->tobj, true, true, false);
 	}
 	spin_unlock(&robj->tobj.lock);
 	radeon_object_unreserve(robj);
@@ -403,7 +408,6 @@ int radeon_object_list_validate(struct list_head *head, void *fence)
 	struct radeon_object *robj;
 	struct radeon_fence *old_fence = NULL;
 	struct list_head *i;
-	uint32_t flags;
 	int r;
 
 	r = radeon_object_list_reserve(head);
@@ -414,27 +418,25 @@ int radeon_object_list_validate(struct list_head *head, void *fence)
 	list_for_each(i, head) {
 		lobj = list_entry(i, struct radeon_object_list, list);
 		robj = lobj->robj;
-		if (lobj->wdomain) {
-			flags = radeon_object_flags_from_domain(lobj->wdomain);
-			flags |= TTM_PL_FLAG_TT;
-		} else {
-			flags = radeon_object_flags_from_domain(lobj->rdomain);
-			flags |= TTM_PL_FLAG_TT;
-			flags |= TTM_PL_FLAG_VRAM;
-		}
 		if (!robj->pin_count) {
-			robj->tobj.proposed_placement = flags | TTM_PL_MASK_CACHING;
+			if (lobj->wdomain) {
+				robj->tobj.proposed_placement =
+					radeon_object_flags_from_domain(lobj->wdomain);
+			} else {
+				robj->tobj.proposed_placement =
+					radeon_object_flags_from_domain(lobj->rdomain);
+			}
 			r = ttm_buffer_object_validate(&robj->tobj,
 						       robj->tobj.proposed_placement,
 						       true, false);
 			if (unlikely(r)) {
-				radeon_object_list_unreserve(head);
 				DRM_ERROR("radeon: failed to validate.\n");
 				return r;
 			}
 			radeon_object_gpu_addr(robj);
 		}
 		lobj->gpu_offset = robj->gpu_addr;
+		lobj->tiling_flags = robj->tiling_flags;
 		if (fence) {
 			old_fence = (struct radeon_fence *)robj->tobj.sync_obj;
 			robj->tobj.sync_obj = radeon_fence_ref(fence);
@@ -479,3 +481,127 @@ unsigned long radeon_object_size(struct radeon_object *robj)
 {
 	return robj->tobj.num_pages << PAGE_SHIFT;
 }
+
+int radeon_object_get_surface_reg(struct radeon_object *robj)
+{
+	struct radeon_device *rdev = robj->rdev;
+	struct radeon_surface_reg *reg;
+	struct radeon_object *old_object;
+	int steal;
+	int i;
+
+	if (!robj->tiling_flags)
+		return 0;
+
+	if (robj->surface_reg >= 0) {
+		reg = &rdev->surface_regs[robj->surface_reg];
+		i = robj->surface_reg;
+		goto out;
+	}
+
+	steal = -1;
+	for (i = 0; i < RADEON_GEM_MAX_SURFACES; i++) {
+
+		reg = &rdev->surface_regs[i];
+		if (!reg->robj)
+			break;
+
+		old_object = reg->robj;
+		if (old_object->pin_count == 0)
+			steal = i;
+	}
+
+	/* if we are all out */
+	if (i == RADEON_GEM_MAX_SURFACES) {
+		if (steal == -1)
+			return -ENOMEM;
+		/* find someone with a surface reg and nuke their BO */
+		reg = &rdev->surface_regs[steal];
+		old_object = reg->robj;
+		/* blow away the mapping */
+		DRM_DEBUG("stealing surface reg %d from %p\n", steal, old_object);
+		ttm_bo_unmap_virtual(&old_object->tobj);
+		old_object->surface_reg = -1;
+		i = steal;
+	}
+
+	robj->surface_reg = i;
+	reg->robj = robj;
+
+out:
+	radeon_set_surface_reg(rdev, i, robj->tiling_flags, robj->pitch,
+			       robj->tobj.mem.mm_node->start << PAGE_SHIFT,
+			       robj->tobj.num_pages << PAGE_SHIFT);
+	return 0;
+}
+
+void radeon_object_clear_surface_reg(struct radeon_object *robj)
+{
+	struct radeon_device *rdev = robj->rdev;
+	struct radeon_surface_reg *reg;
+
+	if (robj->surface_reg == -1)
+		return;
+
+	reg = &rdev->surface_regs[robj->surface_reg];
+	radeon_clear_surface_reg(rdev, robj->surface_reg);
+
+	reg->robj = NULL;
+	robj->surface_reg = -1;
+}
+
+void radeon_object_set_tiling_flags(struct radeon_object *robj,
+				    uint32_t tiling_flags, uint32_t pitch)
+{
+	robj->tiling_flags = tiling_flags;
+	robj->pitch = pitch;
+}
+
+void radeon_object_get_tiling_flags(struct radeon_object *robj,
+				    uint32_t *tiling_flags,
+				    uint32_t *pitch)
+{
+	if (tiling_flags)
+		*tiling_flags = robj->tiling_flags;
+	if (pitch)
+		*pitch = robj->pitch;
+}
+
+int radeon_object_check_tiling(struct radeon_object *robj, bool has_moved,
+			       bool force_drop)
+{
+	if (!(robj->tiling_flags & RADEON_TILING_SURFACE))
+		return 0;
+
+	if (force_drop) {
+		radeon_object_clear_surface_reg(robj);
+		return 0;
+	}
+
+	if (robj->tobj.mem.mem_type != TTM_PL_VRAM) {
+		if (!has_moved)
+			return 0;
+
+		if (robj->surface_reg >= 0)
+			radeon_object_clear_surface_reg(robj);
+		return 0;
+	}
+
+	if ((robj->surface_reg >= 0) && !has_moved)
+		return 0;
+
+	return radeon_object_get_surface_reg(robj);
+}
+
+void radeon_bo_move_notify(struct ttm_buffer_object *bo,
+			  struct ttm_mem_reg *mem)
+{
+	struct radeon_object *robj = container_of(bo, struct radeon_object, tobj);
+	radeon_object_check_tiling(robj, 0, 1);
+}
+
+void radeon_bo_fault_reserve_notify(struct ttm_buffer_object *bo)
+{
+	struct radeon_object *robj = container_of(bo, struct radeon_object, tobj);
+	radeon_object_check_tiling(robj, 0, 0);
+}
diff --git a/drivers/gpu/drm/radeon/radeon_ring.c b/drivers/gpu/drm/radeon/radeon_ring.c
index a853261d188..60d159308b8 100644
--- a/drivers/gpu/drm/radeon/radeon_ring.c
+++ b/drivers/gpu/drm/radeon/radeon_ring.c
@@ -126,32 +126,19 @@ static void radeon_ib_align(struct radeon_device *rdev, struct radeon_ib *ib)
 	}
 }
 
-static void radeon_ib_cpu_flush(struct radeon_device *rdev,
-				struct radeon_ib *ib)
-{
-	unsigned long tmp;
-	unsigned i;
-
-	/* To force CPU cache flush ugly but seems reliable */
-	for (i = 0; i < ib->length_dw; i += (rdev->cp.align_mask + 1)) {
-		tmp = readl(&ib->ptr[i]);
-	}
-}
-
 int radeon_ib_schedule(struct radeon_device *rdev, struct radeon_ib *ib)
 {
 	int r = 0;
 
 	mutex_lock(&rdev->ib_pool.mutex);
 	radeon_ib_align(rdev, ib);
-	radeon_ib_cpu_flush(rdev, ib);
 	if (!ib->length_dw || !rdev->cp.ready) {
 		/* TODO: Nothings in the ib we should report. */
 		mutex_unlock(&rdev->ib_pool.mutex);
 		DRM_ERROR("radeon: couldn't schedule IB(%lu).\n", ib->idx);
 		return -EINVAL;
 	}
-	/* 64 dwords should be enought for fence too */
+	/* 64 dwords should be enough for fence too */
 	r = radeon_ring_lock(rdev, 64);
 	if (r) {
 		DRM_ERROR("radeon: scheduling IB failled (%d).\n", r);
diff --git a/drivers/gpu/drm/radeon/radeon_share.h b/drivers/gpu/drm/radeon/radeon_share.h
new file mode 100644
index 00000000000..63a773578f1
--- /dev/null
+++ b/drivers/gpu/drm/radeon/radeon_share.h
@@ -0,0 +1,39 @@
+/*
+ * Copyright 2008 Advanced Micro Devices, Inc.
+ * Copyright 2008 Red Hat Inc.
+ * Copyright 2009 Jerome Glisse.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: Dave Airlie
+ *          Alex Deucher
+ *          Jerome Glisse
+ */
+#ifndef __RADEON_SHARE_H__
+#define __RADEON_SHARE_H__
+
+void r100_vram_init_sizes(struct radeon_device *rdev);
+
+void rs690_line_buffer_adjust(struct radeon_device *rdev,
+			      struct drm_display_mode *mode1,
+			      struct drm_display_mode *mode2);
+
+void rv515_bandwidth_avivo_update(struct radeon_device *rdev);
+
+#endif
diff --git a/drivers/gpu/drm/radeon/radeon_test.c b/drivers/gpu/drm/radeon/radeon_test.c
new file mode 100644
index 00000000000..03c33cf4e14
--- /dev/null
+++ b/drivers/gpu/drm/radeon/radeon_test.c
@@ -0,0 +1,209 @@
+/*
+ * Copyright 2009 VMware, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: Michel Dänzer
+ */
+#include <drm/drmP.h>
+#include <drm/radeon_drm.h>
+#include "radeon_reg.h"
+#include "radeon.h"
+
+
+/* Test BO GTT->VRAM and VRAM->GTT GPU copies across the whole GTT aperture */
+void radeon_test_moves(struct radeon_device *rdev)
+{
+	struct radeon_object *vram_obj = NULL;
+	struct radeon_object **gtt_obj = NULL;
+	struct radeon_fence *fence = NULL;
+	uint64_t gtt_addr, vram_addr;
+	unsigned i, n, size;
+	int r;
+
+	size = 1024 * 1024;
+
+	/* Number of tests =
+	 * (Total GTT - IB pool - writeback page - ring buffer) / test size
+	 */
+	n = (rdev->mc.gtt_size - RADEON_IB_POOL_SIZE*64*1024 - 4096 -
+	     rdev->cp.ring_size) / size;
+
+	gtt_obj = kzalloc(n * sizeof(*gtt_obj), GFP_KERNEL);
+	if (!gtt_obj) {
+		DRM_ERROR("Failed to allocate %d pointers\n", n);
+		r = 1;
+		goto out_cleanup;
+	}
+
+	r = radeon_object_create(rdev, NULL, size, true, RADEON_GEM_DOMAIN_VRAM,
+				 false, &vram_obj);
+	if (r) {
+		DRM_ERROR("Failed to create VRAM object\n");
+		goto out_cleanup;
+	}
+
+	r = radeon_object_pin(vram_obj, RADEON_GEM_DOMAIN_VRAM, &vram_addr);
+	if (r) {
+		DRM_ERROR("Failed to pin VRAM object\n");
+		goto out_cleanup;
+	}
+
+	for (i = 0; i < n; i++) {
+		void *gtt_map, *vram_map;
+		void **gtt_start, **gtt_end;
+		void **vram_start, **vram_end;
+
+		r = radeon_object_create(rdev, NULL, size, true,
+					 RADEON_GEM_DOMAIN_GTT, false, gtt_obj + i);
+		if (r) {
+			DRM_ERROR("Failed to create GTT object %d\n", i);
+			goto out_cleanup;
+		}
+
+		r = radeon_object_pin(gtt_obj[i], RADEON_GEM_DOMAIN_GTT, &gtt_addr);
+		if (r) {
+			DRM_ERROR("Failed to pin GTT object %d\n", i);
+			goto out_cleanup;
+		}
+
+		r = radeon_object_kmap(gtt_obj[i], &gtt_map);
+		if (r) {
+			DRM_ERROR("Failed to map GTT object %d\n", i);
+			goto out_cleanup;
+		}
+
+		for (gtt_start = gtt_map, gtt_end = gtt_map + size;
+		     gtt_start < gtt_end;
+		     gtt_start++)
+			*gtt_start = gtt_start;
+
+		radeon_object_kunmap(gtt_obj[i]);
+
+		r = radeon_fence_create(rdev, &fence);
+		if (r) {
+			DRM_ERROR("Failed to create GTT->VRAM fence %d\n", i);
+			goto out_cleanup;
+		}
+
+		r = radeon_copy(rdev, gtt_addr, vram_addr, size / 4096, fence);
+		if (r) {
+			DRM_ERROR("Failed GTT->VRAM copy %d\n", i);
+			goto out_cleanup;
+		}
+
+		r = radeon_fence_wait(fence, false);
+		if (r) {
+			DRM_ERROR("Failed to wait for GTT->VRAM fence %d\n", i);
+			goto out_cleanup;
+		}
+
+		radeon_fence_unref(&fence);
+
+		r = radeon_object_kmap(vram_obj, &vram_map);
+		if (r) {
+			DRM_ERROR("Failed to map VRAM object after copy %d\n", i);
+			goto out_cleanup;
+		}
+
+		for (gtt_start = gtt_map, gtt_end = gtt_map + size,
+		     vram_start = vram_map, vram_end = vram_map + size;
+		     vram_start < vram_end;
+		     gtt_start++, vram_start++) {
+			if (*vram_start != gtt_start) {
+				DRM_ERROR("Incorrect GTT->VRAM copy %d: Got 0x%p, "
+					  "expected 0x%p (GTT map 0x%p-0x%p)\n",
+					  i, *vram_start, gtt_start, gtt_map,
+					  gtt_end);
+				radeon_object_kunmap(vram_obj);
+				goto out_cleanup;
+			}
+			*vram_start = vram_start;
+		}
+
+		radeon_object_kunmap(vram_obj);
+
+		r = radeon_fence_create(rdev, &fence);
+		if (r) {
+			DRM_ERROR("Failed to create VRAM->GTT fence %d\n", i);
+			goto out_cleanup;
+		}
+
+		r = radeon_copy(rdev, vram_addr, gtt_addr, size / 4096, fence);
+		if (r) {
+			DRM_ERROR("Failed VRAM->GTT copy %d\n", i);
+			goto out_cleanup;
+		}
+
+		r = radeon_fence_wait(fence, false);
+		if (r) {
+			DRM_ERROR("Failed to wait for VRAM->GTT fence %d\n", i);
+			goto out_cleanup;
+		}
+
+		radeon_fence_unref(&fence);
+
+		r = radeon_object_kmap(gtt_obj[i], &gtt_map);
+		if (r) {
+			DRM_ERROR("Failed to map GTT object after copy %d\n", i);
+			goto out_cleanup;
+		}
+
+		for (gtt_start = gtt_map, gtt_end = gtt_map + size,
+		     vram_start = vram_map, vram_end = vram_map + size;
+		     gtt_start < gtt_end;
+		     gtt_start++, vram_start++) {
+			if (*gtt_start != vram_start) {
+				DRM_ERROR("Incorrect VRAM->GTT copy %d: Got 0x%p, "
+					  "expected 0x%p (VRAM map 0x%p-0x%p)\n",
+					  i, *gtt_start, vram_start, vram_map,
+					  vram_end);
+				radeon_object_kunmap(gtt_obj[i]);
+				goto out_cleanup;
+			}
+		}
+
+		radeon_object_kunmap(gtt_obj[i]);
+
+		DRM_INFO("Tested GTT->VRAM and VRAM->GTT copy for GTT offset 0x%llx\n",
+			 gtt_addr - rdev->mc.gtt_location);
+	}
+
+out_cleanup:
+	if (vram_obj) {
+		radeon_object_unpin(vram_obj);
+		radeon_object_unref(&vram_obj);
+	}
+	if (gtt_obj) {
+		for (i = 0; i < n; i++) {
+			if (gtt_obj[i]) {
+				radeon_object_unpin(gtt_obj[i]);
+				radeon_object_unref(&gtt_obj[i]);
+			}
+		}
+		kfree(gtt_obj);
+	}
+	if (fence) {
+		radeon_fence_unref(&fence);
+	}
+	if (r) {
+		printk(KERN_WARNING "Error while testing BO move.\n");
+	}
+}
+
diff --git a/drivers/gpu/drm/radeon/radeon_ttm.c b/drivers/gpu/drm/radeon/radeon_ttm.c
index 1227a97f516..15c3531377e 100644
--- a/drivers/gpu/drm/radeon/radeon_ttm.c
+++ b/drivers/gpu/drm/radeon/radeon_ttm.c
@@ -355,23 +355,26 @@ static int radeon_bo_move(struct ttm_buffer_object *bo,
 	if (!rdev->cp.ready) {
 		/* use memcpy */
 		DRM_ERROR("CP is not ready use memcpy.\n");
-		return ttm_bo_move_memcpy(bo, evict, no_wait, new_mem);
+		goto memcpy;
 	}
 
 	if (old_mem->mem_type == TTM_PL_VRAM &&
 	    new_mem->mem_type == TTM_PL_SYSTEM) {
-		return radeon_move_vram_ram(bo, evict, interruptible,
+		r = radeon_move_vram_ram(bo, evict, interruptible,
 					    no_wait, new_mem);
 	} else if (old_mem->mem_type == TTM_PL_SYSTEM &&
 		   new_mem->mem_type == TTM_PL_VRAM) {
-		return radeon_move_ram_vram(bo, evict, interruptible,
+		r = radeon_move_ram_vram(bo, evict, interruptible,
 					    no_wait, new_mem);
 	} else {
 		r = radeon_move_blit(bo, evict, no_wait, new_mem, old_mem);
-		if (unlikely(r)) {
-			return r;
-		}
 	}
+
+	if (r) {
+memcpy:
+		r = ttm_bo_move_memcpy(bo, evict, no_wait, new_mem);
+	}
+
 	return r;
 }
 
@@ -429,6 +432,8 @@ static struct ttm_bo_driver radeon_bo_driver = {
 	.sync_obj_flush = &radeon_sync_obj_flush,
 	.sync_obj_unref = &radeon_sync_obj_unref,
 	.sync_obj_ref = &radeon_sync_obj_ref,
+	.move_notify = &radeon_bo_move_notify,
+	.fault_reserve_notify = &radeon_bo_fault_reserve_notify,
 };
 
 int radeon_ttm_init(struct radeon_device *rdev)
@@ -442,13 +447,14 @@ int radeon_ttm_init(struct radeon_device *rdev)
 	/* No others user of address space so set it to 0 */
 	r = ttm_bo_device_init(&rdev->mman.bdev,
 			       rdev->mman.mem_global_ref.object,
-			       &radeon_bo_driver, DRM_FILE_PAGE_OFFSET);
+			       &radeon_bo_driver, DRM_FILE_PAGE_OFFSET,
+			       rdev->need_dma32);
 	if (r) {
 		DRM_ERROR("failed initializing buffer object driver(%d).\n", r);
 		return r;
 	}
 	r = ttm_bo_init_mm(&rdev->mman.bdev, TTM_PL_VRAM, 0,
-			   ((rdev->mc.aper_size) >> PAGE_SHIFT));
+			   ((rdev->mc.real_vram_size) >> PAGE_SHIFT));
 	if (r) {
 		DRM_ERROR("Failed initializing VRAM heap.\n");
 		return r;
@@ -465,7 +471,7 @@ int radeon_ttm_init(struct radeon_device *rdev)
 		return r;
 	}
 	DRM_INFO("radeon: %uM of VRAM memory ready\n",
-		 rdev->mc.vram_size / (1024 * 1024));
+		 rdev->mc.real_vram_size / (1024 * 1024));
 	r = ttm_bo_init_mm(&rdev->mman.bdev, TTM_PL_TT, 0,
 			   ((rdev->mc.gtt_size) >> PAGE_SHIFT));
 	if (r) {
diff --git a/drivers/gpu/drm/radeon/rs400.c b/drivers/gpu/drm/radeon/rs400.c
index cc074b5a8f7..b29affd9c5d 100644
--- a/drivers/gpu/drm/radeon/rs400.c
+++ b/drivers/gpu/drm/radeon/rs400.c
@@ -29,6 +29,7 @@
 #include <drm/drmP.h>
 #include "radeon_reg.h"
 #include "radeon.h"
+#include "radeon_share.h"
 
 /* rs400,rs480 depends on : */
 void r100_hdp_reset(struct radeon_device *rdev);
@@ -164,7 +165,9 @@ int rs400_gart_enable(struct radeon_device *rdev)
 		WREG32(RADEON_BUS_CNTL, tmp);
 	}
 	/* Table should be in 32bits address space so ignore bits above. */
-	tmp = rdev->gart.table_addr & 0xfffff000;
+	tmp = (u32)rdev->gart.table_addr & 0xfffff000;
+	tmp |= (upper_32_bits(rdev->gart.table_addr) & 0xff) << 4;
+
 	WREG32_MC(RS480_GART_BASE, tmp);
 	/* TODO: more tweaking here */
 	WREG32_MC(RS480_GART_FEATURE_ID,
@@ -201,10 +204,17 @@ void rs400_gart_disable(struct radeon_device *rdev)
 
 int rs400_gart_set_page(struct radeon_device *rdev, int i, uint64_t addr)
 {
+	uint32_t entry;
+
 	if (i < 0 || i > rdev->gart.num_gpu_pages) {
 		return -EINVAL;
 	}
-	rdev->gart.table.ram.ptr[i] = cpu_to_le32(((uint32_t)addr) | 0xC);
+
+	entry = (lower_32_bits(addr) & PAGE_MASK) |
+		((upper_32_bits(addr) & 0xff) << 4) |
+		0xc;
+	entry = cpu_to_le32(entry);
+	rdev->gart.table.ram.ptr[i] = entry;
 	return 0;
 }
 
@@ -223,10 +233,9 @@ int rs400_mc_init(struct radeon_device *rdev)
 
 	rs400_gpu_init(rdev);
 	rs400_gart_disable(rdev);
-	rdev->mc.gtt_location = rdev->mc.vram_size;
+	rdev->mc.gtt_location = rdev->mc.mc_vram_size;
 	rdev->mc.gtt_location += (rdev->mc.gtt_size - 1);
 	rdev->mc.gtt_location &= ~(rdev->mc.gtt_size - 1);
-	rdev->mc.vram_location = 0xFFFFFFFFUL;
 	r = radeon_mc_setup(rdev);
 	if (r) {
 		return r;
@@ -238,7 +247,7 @@ int rs400_mc_init(struct radeon_device *rdev)
 		       "programming pipes. Bad things might happen.\n");
 	}
 
-	tmp = rdev->mc.vram_location + rdev->mc.vram_size - 1;
+	tmp = rdev->mc.vram_location + rdev->mc.mc_vram_size - 1;
 	tmp = REG_SET(RADEON_MC_FB_TOP, tmp >> 16);
 	tmp |= REG_SET(RADEON_MC_FB_START, rdev->mc.vram_location >> 16);
 	WREG32(RADEON_MC_FB_LOCATION, tmp);
@@ -284,21 +293,12 @@ void rs400_gpu_init(struct radeon_device *rdev)
  */
 void rs400_vram_info(struct radeon_device *rdev)
 {
-	uint32_t tom;
-
 	rs400_gart_adjust_size(rdev);
 	/* DDR for all card after R300 & IGP */
 	rdev->mc.vram_is_ddr = true;
 	rdev->mc.vram_width = 128;
 
-	/* read NB_TOM to get the amount of ram stolen for the GPU */
-	tom = RREG32(RADEON_NB_TOM);
-	rdev->mc.vram_size = (((tom >> 16) - (tom & 0xffff) + 1) << 16);
-	WREG32(RADEON_CONFIG_MEMSIZE, rdev->mc.vram_size);
-
-	/* Could aper size report 0 ? */
-	rdev->mc.aper_base = drm_get_resource_start(rdev->ddev, 0);
-	rdev->mc.aper_size = drm_get_resource_len(rdev->ddev, 0);
+	r100_vram_init_sizes(rdev);
 }
 
 
diff --git a/drivers/gpu/drm/radeon/rs600.c b/drivers/gpu/drm/radeon/rs600.c
index ab0c967553e..bbea6dee4a9 100644
--- a/drivers/gpu/drm/radeon/rs600.c
+++ b/drivers/gpu/drm/radeon/rs600.c
@@ -223,7 +223,7 @@ int rs600_mc_init(struct radeon_device *rdev)
 		printk(KERN_WARNING "Failed to wait MC idle while "
 		       "programming pipes. Bad things might happen.\n");
 	}
-	tmp = rdev->mc.vram_location + rdev->mc.vram_size - 1;
+	tmp = rdev->mc.vram_location + rdev->mc.mc_vram_size - 1;
 	tmp = REG_SET(RS600_MC_FB_TOP, tmp >> 16);
 	tmp |= REG_SET(RS600_MC_FB_START, rdev->mc.vram_location >> 16);
 	WREG32_MC(RS600_MC_FB_LOCATION, tmp);
@@ -301,6 +301,11 @@ void rs600_vram_info(struct radeon_device *rdev)
 	rdev->mc.vram_width = 128;
 }
 
+void rs600_bandwidth_update(struct radeon_device *rdev)
+{
+	/* FIXME: implement, should this be like rs690 ? */
+}
+
 
 /*
  * Indirect registers accessor
diff --git a/drivers/gpu/drm/radeon/rs690.c b/drivers/gpu/drm/radeon/rs690.c
index 79ba85042b5..839595b0072 100644
--- a/drivers/gpu/drm/radeon/rs690.c
+++ b/drivers/gpu/drm/radeon/rs690.c
@@ -28,6 +28,9 @@
 #include "drmP.h"
 #include "radeon_reg.h"
 #include "radeon.h"
+#include "rs690r.h"
+#include "atom.h"
+#include "atom-bits.h"
 
 /* rs690,rs740 depends on : */
 void r100_hdp_reset(struct radeon_device *rdev);
@@ -64,7 +67,7 @@ int rs690_mc_init(struct radeon_device *rdev)
 	rs400_gart_disable(rdev);
 
 	/* Setup GPU memory space */
-	rdev->mc.gtt_location = rdev->mc.vram_size;
+	rdev->mc.gtt_location = rdev->mc.mc_vram_size;
 	rdev->mc.gtt_location += (rdev->mc.gtt_size - 1);
 	rdev->mc.gtt_location &= ~(rdev->mc.gtt_size - 1);
 	rdev->mc.vram_location = 0xFFFFFFFFUL;
@@ -79,7 +82,7 @@ int rs690_mc_init(struct radeon_device *rdev)
 		printk(KERN_WARNING "Failed to wait MC idle while "
 		       "programming pipes. Bad things might happen.\n");
 	}
-	tmp = rdev->mc.vram_location + rdev->mc.vram_size - 1;
+	tmp = rdev->mc.vram_location + rdev->mc.mc_vram_size - 1;
 	tmp = REG_SET(RS690_MC_FB_TOP, tmp >> 16);
 	tmp |= REG_SET(RS690_MC_FB_START, rdev->mc.vram_location >> 16);
 	WREG32_MC(RS690_MCCFG_FB_LOCATION, tmp);
@@ -138,9 +141,82 @@ void rs690_gpu_init(struct radeon_device *rdev)
 /*
  * VRAM info.
  */
+void rs690_pm_info(struct radeon_device *rdev)
+{
+	int index = GetIndexIntoMasterTable(DATA, IntegratedSystemInfo);
+	struct _ATOM_INTEGRATED_SYSTEM_INFO *info;
+	struct _ATOM_INTEGRATED_SYSTEM_INFO_V2 *info_v2;
+	void *ptr;
+	uint16_t data_offset;
+	uint8_t frev, crev;
+	fixed20_12 tmp;
+
+	atom_parse_data_header(rdev->mode_info.atom_context, index, NULL,
+			       &frev, &crev, &data_offset);
+	ptr = rdev->mode_info.atom_context->bios + data_offset;
+	info = (struct _ATOM_INTEGRATED_SYSTEM_INFO *)ptr;
+	info_v2 = (struct _ATOM_INTEGRATED_SYSTEM_INFO_V2 *)ptr;
+	/* Get various system informations from bios */
+	switch (crev) {
+	case 1:
+		tmp.full = rfixed_const(100);
+		rdev->pm.igp_sideport_mclk.full = rfixed_const(info->ulBootUpMemoryClock);
+		rdev->pm.igp_sideport_mclk.full = rfixed_div(rdev->pm.igp_sideport_mclk, tmp);
+		rdev->pm.igp_system_mclk.full = rfixed_const(le16_to_cpu(info->usK8MemoryClock));
+		rdev->pm.igp_ht_link_clk.full = rfixed_const(le16_to_cpu(info->usFSBClock));
+		rdev->pm.igp_ht_link_width.full = rfixed_const(info->ucHTLinkWidth);
+		break;
+	case 2:
+		tmp.full = rfixed_const(100);
+		rdev->pm.igp_sideport_mclk.full = rfixed_const(info_v2->ulBootUpSidePortClock);
+		rdev->pm.igp_sideport_mclk.full = rfixed_div(rdev->pm.igp_sideport_mclk, tmp);
+		rdev->pm.igp_system_mclk.full = rfixed_const(info_v2->ulBootUpUMAClock);
+		rdev->pm.igp_system_mclk.full = rfixed_div(rdev->pm.igp_system_mclk, tmp);
+		rdev->pm.igp_ht_link_clk.full = rfixed_const(info_v2->ulHTLinkFreq);
+		rdev->pm.igp_ht_link_clk.full = rfixed_div(rdev->pm.igp_ht_link_clk, tmp);
+		rdev->pm.igp_ht_link_width.full = rfixed_const(le16_to_cpu(info_v2->usMinHTLinkWidth));
+		break;
+	default:
+		tmp.full = rfixed_const(100);
+		/* We assume the slower possible clock ie worst case */
+		/* DDR 333Mhz */
+		rdev->pm.igp_sideport_mclk.full = rfixed_const(333);
+		/* FIXME: system clock ? */
+		rdev->pm.igp_system_mclk.full = rfixed_const(100);
+		rdev->pm.igp_system_mclk.full = rfixed_div(rdev->pm.igp_system_mclk, tmp);
+		rdev->pm.igp_ht_link_clk.full = rfixed_const(200);
+		rdev->pm.igp_ht_link_width.full = rfixed_const(8);
+		DRM_ERROR("No integrated system info for your GPU, using safe default\n");
+		break;
+	}
+	/* Compute various bandwidth */
+	/* k8_bandwidth = (memory_clk / 2) * 2 * 8 * 0.5 = memory_clk * 4  */
+	tmp.full = rfixed_const(4);
+	rdev->pm.k8_bandwidth.full = rfixed_mul(rdev->pm.igp_system_mclk, tmp);
+	/* ht_bandwidth = ht_clk * 2 * ht_width / 8 * 0.8
+	 *              = ht_clk * ht_width / 5
+	 */
+	tmp.full = rfixed_const(5);
+	rdev->pm.ht_bandwidth.full = rfixed_mul(rdev->pm.igp_ht_link_clk,
+						rdev->pm.igp_ht_link_width);
+	rdev->pm.ht_bandwidth.full = rfixed_div(rdev->pm.ht_bandwidth, tmp);
+	if (tmp.full < rdev->pm.max_bandwidth.full) {
+		/* HT link is a limiting factor */
+		rdev->pm.max_bandwidth.full = tmp.full;
+	}
+	/* sideport_bandwidth = (sideport_clk / 2) * 2 * 2 * 0.7
+	 *                    = (sideport_clk * 14) / 10
+	 */
+	tmp.full = rfixed_const(14);
+	rdev->pm.sideport_bandwidth.full = rfixed_mul(rdev->pm.igp_sideport_mclk, tmp);
+	tmp.full = rfixed_const(10);
+	rdev->pm.sideport_bandwidth.full = rfixed_div(rdev->pm.sideport_bandwidth, tmp);
+}
+
 void rs690_vram_info(struct radeon_device *rdev)
 {
 	uint32_t tmp;
+	fixed20_12 a;
 
 	rs400_gart_adjust_size(rdev);
 	/* DDR for all card after R300 & IGP */
@@ -152,12 +228,409 @@ void rs690_vram_info(struct radeon_device *rdev)
 	} else {
 		rdev->mc.vram_width = 64;
 	}
-	rdev->mc.vram_size = RREG32(RADEON_CONFIG_MEMSIZE);
+	rdev->mc.real_vram_size = RREG32(RADEON_CONFIG_MEMSIZE);
+	rdev->mc.mc_vram_size = rdev->mc.real_vram_size;
 
 	rdev->mc.aper_base = drm_get_resource_start(rdev->ddev, 0);
 	rdev->mc.aper_size = drm_get_resource_len(rdev->ddev, 0);
+	rs690_pm_info(rdev);
+	/* FIXME: we should enforce default clock in case GPU is not in
+	 * default setup
+	 */
+	a.full = rfixed_const(100);
+	rdev->pm.sclk.full = rfixed_const(rdev->clock.default_sclk);
+	rdev->pm.sclk.full = rfixed_div(rdev->pm.sclk, a);
+	a.full = rfixed_const(16);
+	/* core_bandwidth = sclk(Mhz) * 16 */
+	rdev->pm.core_bandwidth.full = rfixed_div(rdev->pm.sclk, a);
+}
+
+void rs690_line_buffer_adjust(struct radeon_device *rdev,
+			      struct drm_display_mode *mode1,
+			      struct drm_display_mode *mode2)
+{
+	u32 tmp;
+
+	/*
+	 * Line Buffer Setup
+	 * There is a single line buffer shared by both display controllers.
+	 * DC_LB_MEMORY_SPLIT controls how that line buffer is shared between
+	 * the display controllers.  The paritioning can either be done
+	 * manually or via one of four preset allocations specified in bits 1:0:
+	 *  0 - line buffer is divided in half and shared between crtc
+	 *  1 - D1 gets 3/4 of the line buffer, D2 gets 1/4
+	 *  2 - D1 gets the whole buffer
+	 *  3 - D1 gets 1/4 of the line buffer, D2 gets 3/4
+	 * Setting bit 2 of DC_LB_MEMORY_SPLIT controls switches to manual
+	 * allocation mode. In manual allocation mode, D1 always starts at 0,
+	 * D1 end/2 is specified in bits 14:4; D2 allocation follows D1.
+	 */
+	tmp = RREG32(DC_LB_MEMORY_SPLIT) & ~DC_LB_MEMORY_SPLIT_MASK;
+	tmp &= ~DC_LB_MEMORY_SPLIT_SHIFT_MODE;
+	/* auto */
+	if (mode1 && mode2) {
+		if (mode1->hdisplay > mode2->hdisplay) {
+			if (mode1->hdisplay > 2560)
+				tmp |= DC_LB_MEMORY_SPLIT_D1_3Q_D2_1Q;
+			else
+				tmp |= DC_LB_MEMORY_SPLIT_D1HALF_D2HALF;
+		} else if (mode2->hdisplay > mode1->hdisplay) {
+			if (mode2->hdisplay > 2560)
+				tmp |= DC_LB_MEMORY_SPLIT_D1_1Q_D2_3Q;
+			else
+				tmp |= DC_LB_MEMORY_SPLIT_D1HALF_D2HALF;
+		} else
+			tmp |= AVIVO_DC_LB_MEMORY_SPLIT_D1HALF_D2HALF;
+	} else if (mode1) {
+		tmp |= DC_LB_MEMORY_SPLIT_D1_ONLY;
+	} else if (mode2) {
+		tmp |= DC_LB_MEMORY_SPLIT_D1_1Q_D2_3Q;
+	}
+	WREG32(DC_LB_MEMORY_SPLIT, tmp);
 }
 
+struct rs690_watermark {
+	u32        lb_request_fifo_depth;
+	fixed20_12 num_line_pair;
+	fixed20_12 estimated_width;
+	fixed20_12 worst_case_latency;
+	fixed20_12 consumption_rate;
+	fixed20_12 active_time;
+	fixed20_12 dbpp;
+	fixed20_12 priority_mark_max;
+	fixed20_12 priority_mark;
+	fixed20_12 sclk;
+};
+
+void rs690_crtc_bandwidth_compute(struct radeon_device *rdev,
+				  struct radeon_crtc *crtc,
+				  struct rs690_watermark *wm)
+{
+	struct drm_display_mode *mode = &crtc->base.mode;
+	fixed20_12 a, b, c;
+	fixed20_12 pclk, request_fifo_depth, tolerable_latency, estimated_width;
+	fixed20_12 consumption_time, line_time, chunk_time, read_delay_latency;
+	/* FIXME: detect IGP with sideport memory, i don't think there is any
+	 * such product available
+	 */
+	bool sideport = false;
+
+	if (!crtc->base.enabled) {
+		/* FIXME: wouldn't it better to set priority mark to maximum */
+		wm->lb_request_fifo_depth = 4;
+		return;
+	}
+
+	if (crtc->vsc.full > rfixed_const(2))
+		wm->num_line_pair.full = rfixed_const(2);
+	else
+		wm->num_line_pair.full = rfixed_const(1);
+
+	b.full = rfixed_const(mode->crtc_hdisplay);
+	c.full = rfixed_const(256);
+	a.full = rfixed_mul(wm->num_line_pair, b);
+	request_fifo_depth.full = rfixed_div(a, c);
+	if (a.full < rfixed_const(4)) {
+		wm->lb_request_fifo_depth = 4;
+	} else {
+		wm->lb_request_fifo_depth = rfixed_trunc(request_fifo_depth);
+	}
+
+	/* Determine consumption rate
+	 *  pclk = pixel clock period(ns) = 1000 / (mode.clock / 1000)
+	 *  vtaps = number of vertical taps,
+	 *  vsc = vertical scaling ratio, defined as source/destination
+	 *  hsc = horizontal scaling ration, defined as source/destination
+	 */
+	a.full = rfixed_const(mode->clock);
+	b.full = rfixed_const(1000);
+	a.full = rfixed_div(a, b);
+	pclk.full = rfixed_div(b, a);
+	if (crtc->rmx_type != RMX_OFF) {
+		b.full = rfixed_const(2);
+		if (crtc->vsc.full > b.full)
+			b.full = crtc->vsc.full;
+		b.full = rfixed_mul(b, crtc->hsc);
+		c.full = rfixed_const(2);
+		b.full = rfixed_div(b, c);
+		consumption_time.full = rfixed_div(pclk, b);
+	} else {
+		consumption_time.full = pclk.full;
+	}
+	a.full = rfixed_const(1);
+	wm->consumption_rate.full = rfixed_div(a, consumption_time);
+
+
+	/* Determine line time
+	 *  LineTime = total time for one line of displayhtotal
+	 *  LineTime = total number of horizontal pixels
+	 *  pclk = pixel clock period(ns)
+	 */
+	a.full = rfixed_const(crtc->base.mode.crtc_htotal);
+	line_time.full = rfixed_mul(a, pclk);
+
+	/* Determine active time
+	 *  ActiveTime = time of active region of display within one line,
+	 *  hactive = total number of horizontal active pixels
+	 *  htotal = total number of horizontal pixels
+	 */
+	a.full = rfixed_const(crtc->base.mode.crtc_htotal);
+	b.full = rfixed_const(crtc->base.mode.crtc_hdisplay);
+	wm->active_time.full = rfixed_mul(line_time, b);
+	wm->active_time.full = rfixed_div(wm->active_time, a);
+
+	/* Maximun bandwidth is the minimun bandwidth of all component */
+	rdev->pm.max_bandwidth = rdev->pm.core_bandwidth;
+	if (sideport) {
+		if (rdev->pm.max_bandwidth.full > rdev->pm.sideport_bandwidth.full &&
+			rdev->pm.sideport_bandwidth.full)
+			rdev->pm.max_bandwidth = rdev->pm.sideport_bandwidth;
+		read_delay_latency.full = rfixed_const(370 * 800 * 1000);
+		read_delay_latency.full = rfixed_div(read_delay_latency,
+			rdev->pm.igp_sideport_mclk);
+	} else {
+		if (rdev->pm.max_bandwidth.full > rdev->pm.k8_bandwidth.full &&
+			rdev->pm.k8_bandwidth.full)
+			rdev->pm.max_bandwidth = rdev->pm.k8_bandwidth;
+		if (rdev->pm.max_bandwidth.full > rdev->pm.ht_bandwidth.full &&
+			rdev->pm.ht_bandwidth.full)
+			rdev->pm.max_bandwidth = rdev->pm.ht_bandwidth;
+		read_delay_latency.full = rfixed_const(5000);
+	}
+
+	/* sclk = system clocks(ns) = 1000 / max_bandwidth / 16 */
+	a.full = rfixed_const(16);
+	rdev->pm.sclk.full = rfixed_mul(rdev->pm.max_bandwidth, a);
+	a.full = rfixed_const(1000);
+	rdev->pm.sclk.full = rfixed_div(a, rdev->pm.sclk);
+	/* Determine chunk time
+	 * ChunkTime = the time it takes the DCP to send one chunk of data
+	 * to the LB which consists of pipeline delay and inter chunk gap
+	 * sclk = system clock(ns)
+	 */
+	a.full = rfixed_const(256 * 13);
+	chunk_time.full = rfixed_mul(rdev->pm.sclk, a);
+	a.full = rfixed_const(10);
+	chunk_time.full = rfixed_div(chunk_time, a);
+
+	/* Determine the worst case latency
+	 * NumLinePair = Number of line pairs to request(1=2 lines, 2=4 lines)
+	 * WorstCaseLatency = worst case time from urgent to when the MC starts
+	 *                    to return data
+	 * READ_DELAY_IDLE_MAX = constant of 1us
+	 * ChunkTime = time it takes the DCP to send one chunk of data to the LB
+	 *             which consists of pipeline delay and inter chunk gap
+	 */
+	if (rfixed_trunc(wm->num_line_pair) > 1) {
+		a.full = rfixed_const(3);
+		wm->worst_case_latency.full = rfixed_mul(a, chunk_time);
+		wm->worst_case_latency.full += read_delay_latency.full;
+	} else {
+		a.full = rfixed_const(2);
+		wm->worst_case_latency.full = rfixed_mul(a, chunk_time);
+		wm->worst_case_latency.full += read_delay_latency.full;
+	}
+
+	/* Determine the tolerable latency
+	 * TolerableLatency = Any given request has only 1 line time
+	 *                    for the data to be returned
+	 * LBRequestFifoDepth = Number of chunk requests the LB can
+	 *                      put into the request FIFO for a display
+	 *  LineTime = total time for one line of display
+	 *  ChunkTime = the time it takes the DCP to send one chunk
+	 *              of data to the LB which consists of
+	 *  pipeline delay and inter chunk gap
+	 */
+	if ((2+wm->lb_request_fifo_depth) >= rfixed_trunc(request_fifo_depth)) {
+		tolerable_latency.full = line_time.full;
+	} else {
+		tolerable_latency.full = rfixed_const(wm->lb_request_fifo_depth - 2);
+		tolerable_latency.full = request_fifo_depth.full - tolerable_latency.full;
+		tolerable_latency.full = rfixed_mul(tolerable_latency, chunk_time);
+		tolerable_latency.full = line_time.full - tolerable_latency.full;
+	}
+	/* We assume worst case 32bits (4 bytes) */
+	wm->dbpp.full = rfixed_const(4 * 8);
+
+	/* Determine the maximum priority mark
+	 *  width = viewport width in pixels
+	 */
+	a.full = rfixed_const(16);
+	wm->priority_mark_max.full = rfixed_const(crtc->base.mode.crtc_hdisplay);
+	wm->priority_mark_max.full = rfixed_div(wm->priority_mark_max, a);
+
+	/* Determine estimated width */
+	estimated_width.full = tolerable_latency.full - wm->worst_case_latency.full;
+	estimated_width.full = rfixed_div(estimated_width, consumption_time);
+	if (rfixed_trunc(estimated_width) > crtc->base.mode.crtc_hdisplay) {
+		wm->priority_mark.full = rfixed_const(10);
+	} else {
+		a.full = rfixed_const(16);
+		wm->priority_mark.full = rfixed_div(estimated_width, a);
+		wm->priority_mark.full = wm->priority_mark_max.full - wm->priority_mark.full;
+	}
+}
+
+void rs690_bandwidth_update(struct radeon_device *rdev)
+{
+	struct drm_display_mode *mode0 = NULL;
+	struct drm_display_mode *mode1 = NULL;
+	struct rs690_watermark wm0;
+	struct rs690_watermark wm1;
+	u32 tmp;
+	fixed20_12 priority_mark02, priority_mark12, fill_rate;
+	fixed20_12 a, b;
+
+	if (rdev->mode_info.crtcs[0]->base.enabled)
+		mode0 = &rdev->mode_info.crtcs[0]->base.mode;
+	if (rdev->mode_info.crtcs[1]->base.enabled)
+		mode1 = &rdev->mode_info.crtcs[1]->base.mode;
+	/*
+	 * Set display0/1 priority up in the memory controller for
+	 * modes if the user specifies HIGH for displaypriority
+	 * option.
+	 */
+	if (rdev->disp_priority == 2) {
+		tmp = RREG32_MC(MC_INIT_MISC_LAT_TIMER);
+		tmp &= ~MC_DISP1R_INIT_LAT_MASK;
+		tmp &= ~MC_DISP0R_INIT_LAT_MASK;
+		if (mode1)
+			tmp |= (1 << MC_DISP1R_INIT_LAT_SHIFT);
+		if (mode0)
+			tmp |= (1 << MC_DISP0R_INIT_LAT_SHIFT);
+		WREG32_MC(MC_INIT_MISC_LAT_TIMER, tmp);
+	}
+	rs690_line_buffer_adjust(rdev, mode0, mode1);
+
+	if ((rdev->family == CHIP_RS690) || (rdev->family == CHIP_RS740))
+		WREG32(DCP_CONTROL, 0);
+	if ((rdev->family == CHIP_RS780) || (rdev->family == CHIP_RS880))
+		WREG32(DCP_CONTROL, 2);
+
+	rs690_crtc_bandwidth_compute(rdev, rdev->mode_info.crtcs[0], &wm0);
+	rs690_crtc_bandwidth_compute(rdev, rdev->mode_info.crtcs[1], &wm1);
+
+	tmp = (wm0.lb_request_fifo_depth - 1);
+	tmp |= (wm1.lb_request_fifo_depth - 1) << 16;
+	WREG32(LB_MAX_REQ_OUTSTANDING, tmp);
+
+	if (mode0 && mode1) {
+		if (rfixed_trunc(wm0.dbpp) > 64)
+			a.full = rfixed_mul(wm0.dbpp, wm0.num_line_pair);
+		else
+			a.full = wm0.num_line_pair.full;
+		if (rfixed_trunc(wm1.dbpp) > 64)
+			b.full = rfixed_mul(wm1.dbpp, wm1.num_line_pair);
+		else
+			b.full = wm1.num_line_pair.full;
+		a.full += b.full;
+		fill_rate.full = rfixed_div(wm0.sclk, a);
+		if (wm0.consumption_rate.full > fill_rate.full) {
+			b.full = wm0.consumption_rate.full - fill_rate.full;
+			b.full = rfixed_mul(b, wm0.active_time);
+			a.full = rfixed_mul(wm0.worst_case_latency,
+						wm0.consumption_rate);
+			a.full = a.full + b.full;
+			b.full = rfixed_const(16 * 1000);
+			priority_mark02.full = rfixed_div(a, b);
+		} else {
+			a.full = rfixed_mul(wm0.worst_case_latency,
+						wm0.consumption_rate);
+			b.full = rfixed_const(16 * 1000);
+			priority_mark02.full = rfixed_div(a, b);
+		}
+		if (wm1.consumption_rate.full > fill_rate.full) {
+			b.full = wm1.consumption_rate.full - fill_rate.full;
+			b.full = rfixed_mul(b, wm1.active_time);
+			a.full = rfixed_mul(wm1.worst_case_latency,
+						wm1.consumption_rate);
+			a.full = a.full + b.full;
+			b.full = rfixed_const(16 * 1000);
+			priority_mark12.full = rfixed_div(a, b);
+		} else {
+			a.full = rfixed_mul(wm1.worst_case_latency,
+						wm1.consumption_rate);
+			b.full = rfixed_const(16 * 1000);
+			priority_mark12.full = rfixed_div(a, b);
+		}
+		if (wm0.priority_mark.full > priority_mark02.full)
+			priority_mark02.full = wm0.priority_mark.full;
+		if (rfixed_trunc(priority_mark02) < 0)
+			priority_mark02.full = 0;
+		if (wm0.priority_mark_max.full > priority_mark02.full)
+			priority_mark02.full = wm0.priority_mark_max.full;
+		if (wm1.priority_mark.full > priority_mark12.full)
+			priority_mark12.full = wm1.priority_mark.full;
+		if (rfixed_trunc(priority_mark12) < 0)
+			priority_mark12.full = 0;
+		if (wm1.priority_mark_max.full > priority_mark12.full)
+			priority_mark12.full = wm1.priority_mark_max.full;
+		WREG32(D1MODE_PRIORITY_A_CNT, rfixed_trunc(priority_mark02));
+		WREG32(D1MODE_PRIORITY_B_CNT, rfixed_trunc(priority_mark02));
+		WREG32(D2MODE_PRIORITY_A_CNT, rfixed_trunc(priority_mark12));
+		WREG32(D2MODE_PRIORITY_B_CNT, rfixed_trunc(priority_mark12));
+	} else if (mode0) {
+		if (rfixed_trunc(wm0.dbpp) > 64)
+			a.full = rfixed_mul(wm0.dbpp, wm0.num_line_pair);
+		else
+			a.full = wm0.num_line_pair.full;
+		fill_rate.full = rfixed_div(wm0.sclk, a);
+		if (wm0.consumption_rate.full > fill_rate.full) {
+			b.full = wm0.consumption_rate.full - fill_rate.full;
+			b.full = rfixed_mul(b, wm0.active_time);
+			a.full = rfixed_mul(wm0.worst_case_latency,
+						wm0.consumption_rate);
+			a.full = a.full + b.full;
+			b.full = rfixed_const(16 * 1000);
+			priority_mark02.full = rfixed_div(a, b);
+		} else {
+			a.full = rfixed_mul(wm0.worst_case_latency,
+						wm0.consumption_rate);
+			b.full = rfixed_const(16 * 1000);
+			priority_mark02.full = rfixed_div(a, b);
+		}
+		if (wm0.priority_mark.full > priority_mark02.full)
+			priority_mark02.full = wm0.priority_mark.full;
+		if (rfixed_trunc(priority_mark02) < 0)
+			priority_mark02.full = 0;
+		if (wm0.priority_mark_max.full > priority_mark02.full)
+			priority_mark02.full = wm0.priority_mark_max.full;
+		WREG32(D1MODE_PRIORITY_A_CNT, rfixed_trunc(priority_mark02));
+		WREG32(D1MODE_PRIORITY_B_CNT, rfixed_trunc(priority_mark02));
+		WREG32(D2MODE_PRIORITY_A_CNT, MODE_PRIORITY_OFF);
+		WREG32(D2MODE_PRIORITY_B_CNT, MODE_PRIORITY_OFF);
+	} else {
+		if (rfixed_trunc(wm1.dbpp) > 64)
+			a.full = rfixed_mul(wm1.dbpp, wm1.num_line_pair);
+		else
+			a.full = wm1.num_line_pair.full;
+		fill_rate.full = rfixed_div(wm1.sclk, a);
+		if (wm1.consumption_rate.full > fill_rate.full) {
+			b.full = wm1.consumption_rate.full - fill_rate.full;
+			b.full = rfixed_mul(b, wm1.active_time);
+			a.full = rfixed_mul(wm1.worst_case_latency,
+						wm1.consumption_rate);
+			a.full = a.full + b.full;
+			b.full = rfixed_const(16 * 1000);
+			priority_mark12.full = rfixed_div(a, b);
+		} else {
+			a.full = rfixed_mul(wm1.worst_case_latency,
+						wm1.consumption_rate);
+			b.full = rfixed_const(16 * 1000);
+			priority_mark12.full = rfixed_div(a, b);
+		}
+		if (wm1.priority_mark.full > priority_mark12.full)
+			priority_mark12.full = wm1.priority_mark.full;
+		if (rfixed_trunc(priority_mark12) < 0)
+			priority_mark12.full = 0;
+		if (wm1.priority_mark_max.full > priority_mark12.full)
+			priority_mark12.full = wm1.priority_mark_max.full;
+		WREG32(D1MODE_PRIORITY_A_CNT, MODE_PRIORITY_OFF);
+		WREG32(D1MODE_PRIORITY_B_CNT, MODE_PRIORITY_OFF);
+		WREG32(D2MODE_PRIORITY_A_CNT, rfixed_trunc(priority_mark12));
+		WREG32(D2MODE_PRIORITY_B_CNT, rfixed_trunc(priority_mark12));
+	}
+}
 
 /*
  * Indirect registers accessor
diff --git a/drivers/gpu/drm/radeon/rs690r.h b/drivers/gpu/drm/radeon/rs690r.h
new file mode 100644
index 00000000000..c0d9faa2175
--- /dev/null
+++ b/drivers/gpu/drm/radeon/rs690r.h
@@ -0,0 +1,99 @@
+/*
+ * Copyright 2008 Advanced Micro Devices, Inc.
+ * Copyright 2008 Red Hat Inc.
+ * Copyright 2009 Jerome Glisse.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: Dave Airlie
+ *          Alex Deucher
+ *          Jerome Glisse
+ */
+#ifndef RS690R_H
+#define RS690R_H
+
+/* RS690/RS740 registers */
+#define MC_INDEX			0x0078
+#	define MC_INDEX_MASK			0x1FF
+#	define MC_INDEX_WR_EN			(1 << 9)
+#	define MC_INDEX_WR_ACK			0x7F
+#define MC_DATA				0x007C
+#define HDP_FB_LOCATION			0x0134
+#define DC_LB_MEMORY_SPLIT		0x6520
+#define		DC_LB_MEMORY_SPLIT_MASK			0x00000003
+#define		DC_LB_MEMORY_SPLIT_SHIFT		0
+#define		DC_LB_MEMORY_SPLIT_D1HALF_D2HALF	0
+#define		DC_LB_MEMORY_SPLIT_D1_3Q_D2_1Q		1
+#define		DC_LB_MEMORY_SPLIT_D1_ONLY		2
+#define		DC_LB_MEMORY_SPLIT_D1_1Q_D2_3Q		3
+#define		DC_LB_MEMORY_SPLIT_SHIFT_MODE		(1 << 2)
+#define		DC_LB_DISP1_END_ADR_SHIFT		4
+#define		DC_LB_DISP1_END_ADR_MASK		0x00007FF0
+#define D1MODE_PRIORITY_A_CNT		0x6548
+#define		MODE_PRIORITY_MARK_MASK			0x00007FFF
+#define		MODE_PRIORITY_OFF			(1 << 16)
+#define		MODE_PRIORITY_ALWAYS_ON			(1 << 20)
+#define		MODE_PRIORITY_FORCE_MASK		(1 << 24)
+#define D1MODE_PRIORITY_B_CNT		0x654C
+#define LB_MAX_REQ_OUTSTANDING		0x6D58
+#define		LB_D1_MAX_REQ_OUTSTANDING_MASK		0x0000000F
+#define		LB_D1_MAX_REQ_OUTSTANDING_SHIFT		0
+#define		LB_D2_MAX_REQ_OUTSTANDING_MASK		0x000F0000
+#define		LB_D2_MAX_REQ_OUTSTANDING_SHIFT		16
+#define DCP_CONTROL			0x6C9C
+#define D2MODE_PRIORITY_A_CNT		0x6D48
+#define D2MODE_PRIORITY_B_CNT		0x6D4C
+
+/* MC indirect registers */
+#define MC_STATUS_IDLE				(1 << 0)
+#define MC_MISC_CNTL			0x18
+#define		DISABLE_GTW			(1 << 1)
+#define		GART_INDEX_REG_EN		(1 << 12)
+#define		BLOCK_GFX_D3_EN			(1 << 14)
+#define GART_FEATURE_ID			0x2B
+#define		HANG_EN				(1 << 11)
+#define		TLB_ENABLE			(1 << 18)
+#define		P2P_ENABLE			(1 << 19)
+#define		GTW_LAC_EN			(1 << 25)
+#define		LEVEL2_GART			(0 << 30)
+#define		LEVEL1_GART			(1 << 30)
+#define		PDC_EN				(1 << 31)
+#define GART_BASE			0x2C
+#define GART_CACHE_CNTRL		0x2E
+#	define GART_CACHE_INVALIDATE		(1 << 0)
+#define MC_STATUS			0x90
+#define MCCFG_FB_LOCATION		0x100
+#define		MC_FB_START_MASK		0x0000FFFF
+#define		MC_FB_START_SHIFT		0
+#define		MC_FB_TOP_MASK			0xFFFF0000
+#define		MC_FB_TOP_SHIFT			16
+#define MCCFG_AGP_LOCATION		0x101
+#define		MC_AGP_START_MASK		0x0000FFFF
+#define		MC_AGP_START_SHIFT		0
+#define		MC_AGP_TOP_MASK			0xFFFF0000
+#define		MC_AGP_TOP_SHIFT		16
+#define MCCFG_AGP_BASE			0x102
+#define MCCFG_AGP_BASE_2		0x103
+#define MC_INIT_MISC_LAT_TIMER		0x104
+#define		MC_DISP0R_INIT_LAT_SHIFT	8
+#define		MC_DISP0R_INIT_LAT_MASK		0x00000F00
+#define		MC_DISP1R_INIT_LAT_SHIFT	12
+#define		MC_DISP1R_INIT_LAT_MASK		0x0000F000
+
+#endif
diff --git a/drivers/gpu/drm/radeon/rv515.c b/drivers/gpu/drm/radeon/rv515.c
index ffea37b1b3e..fd8f3ca716e 100644
--- a/drivers/gpu/drm/radeon/rv515.c
+++ b/drivers/gpu/drm/radeon/rv515.c
@@ -27,8 +27,9 @@
  */
 #include <linux/seq_file.h>
 #include "drmP.h"
-#include "radeon_reg.h"
+#include "rv515r.h"
 #include "radeon.h"
+#include "radeon_share.h"
 
 /* rv515 depends on : */
 void r100_hdp_reset(struct radeon_device *rdev);
@@ -99,26 +100,26 @@ int rv515_mc_init(struct radeon_device *rdev)
 		       "programming pipes. Bad things might happen.\n");
 	}
 	/* Write VRAM size in case we are limiting it */
-	WREG32(RADEON_CONFIG_MEMSIZE, rdev->mc.vram_size);
-	tmp = REG_SET(RV515_MC_FB_START, rdev->mc.vram_location >> 16);
+	WREG32(RADEON_CONFIG_MEMSIZE, rdev->mc.real_vram_size);
+	tmp = REG_SET(MC_FB_START, rdev->mc.vram_location >> 16);
 	WREG32(0x134, tmp);
-	tmp = rdev->mc.vram_location + rdev->mc.vram_size - 1;
-	tmp = REG_SET(RV515_MC_FB_TOP, tmp >> 16);
-	tmp |= REG_SET(RV515_MC_FB_START, rdev->mc.vram_location >> 16);
-	WREG32_MC(RV515_MC_FB_LOCATION, tmp);
-	WREG32(RS690_HDP_FB_LOCATION, rdev->mc.vram_location >> 16);
+	tmp = rdev->mc.vram_location + rdev->mc.mc_vram_size - 1;
+	tmp = REG_SET(MC_FB_TOP, tmp >> 16);
+	tmp |= REG_SET(MC_FB_START, rdev->mc.vram_location >> 16);
+	WREG32_MC(MC_FB_LOCATION, tmp);
+	WREG32(HDP_FB_LOCATION, rdev->mc.vram_location >> 16);
 	WREG32(0x310, rdev->mc.vram_location);
 	if (rdev->flags & RADEON_IS_AGP) {
 		tmp = rdev->mc.gtt_location + rdev->mc.gtt_size - 1;
-		tmp = REG_SET(RV515_MC_AGP_TOP, tmp >> 16);
-		tmp |= REG_SET(RV515_MC_AGP_START, rdev->mc.gtt_location >> 16);
-		WREG32_MC(RV515_MC_AGP_LOCATION, tmp);
-		WREG32_MC(RV515_MC_AGP_BASE, rdev->mc.agp_base);
-		WREG32_MC(RV515_MC_AGP_BASE_2, 0);
+		tmp = REG_SET(MC_AGP_TOP, tmp >> 16);
+		tmp |= REG_SET(MC_AGP_START, rdev->mc.gtt_location >> 16);
+		WREG32_MC(MC_AGP_LOCATION, tmp);
+		WREG32_MC(MC_AGP_BASE, rdev->mc.agp_base);
+		WREG32_MC(MC_AGP_BASE_2, 0);
 	} else {
-		WREG32_MC(RV515_MC_AGP_LOCATION, 0x0FFFFFFF);
-		WREG32_MC(RV515_MC_AGP_BASE, 0);
-		WREG32_MC(RV515_MC_AGP_BASE_2, 0);
+		WREG32_MC(MC_AGP_LOCATION, 0x0FFFFFFF);
+		WREG32_MC(MC_AGP_BASE, 0);
+		WREG32_MC(MC_AGP_BASE_2, 0);
 	}
 	return 0;
 }
@@ -136,95 +137,67 @@ void rv515_mc_fini(struct radeon_device *rdev)
  */
 void rv515_ring_start(struct radeon_device *rdev)
 {
-	unsigned gb_tile_config;
 	int r;
 
-	/* Sub pixel 1/12 so we can have 4K rendering according to doc */
-	gb_tile_config = R300_ENABLE_TILING | R300_TILE_SIZE_16;
-	switch (rdev->num_gb_pipes) {
-	case 2:
-		gb_tile_config |= R300_PIPE_COUNT_R300;
-		break;
-	case 3:
-		gb_tile_config |= R300_PIPE_COUNT_R420_3P;
-		break;
-	case 4:
-		gb_tile_config |= R300_PIPE_COUNT_R420;
-		break;
-	case 1:
-	default:
-		gb_tile_config |= R300_PIPE_COUNT_RV350;
-		break;
-	}
-
 	r = radeon_ring_lock(rdev, 64);
 	if (r) {
 		return;
 	}
-	radeon_ring_write(rdev, PACKET0(RADEON_ISYNC_CNTL, 0));
-	radeon_ring_write(rdev,
-			  RADEON_ISYNC_ANY2D_IDLE3D |
-			  RADEON_ISYNC_ANY3D_IDLE2D |
-			  RADEON_ISYNC_WAIT_IDLEGUI |
-			  RADEON_ISYNC_CPSCRATCH_IDLEGUI);
-	radeon_ring_write(rdev, PACKET0(R300_GB_TILE_CONFIG, 0));
-	radeon_ring_write(rdev, gb_tile_config);
-	radeon_ring_write(rdev, PACKET0(RADEON_WAIT_UNTIL, 0));
+	radeon_ring_write(rdev, PACKET0(ISYNC_CNTL, 0));
 	radeon_ring_write(rdev,
-			  RADEON_WAIT_2D_IDLECLEAN |
-			  RADEON_WAIT_3D_IDLECLEAN);
+			  ISYNC_ANY2D_IDLE3D |
+			  ISYNC_ANY3D_IDLE2D |
+			  ISYNC_WAIT_IDLEGUI |
+			  ISYNC_CPSCRATCH_IDLEGUI);
+	radeon_ring_write(rdev, PACKET0(WAIT_UNTIL, 0));
+	radeon_ring_write(rdev, WAIT_2D_IDLECLEAN | WAIT_3D_IDLECLEAN);
 	radeon_ring_write(rdev, PACKET0(0x170C, 0));
 	radeon_ring_write(rdev, 1 << 31);
-	radeon_ring_write(rdev, PACKET0(R300_GB_SELECT, 0));
+	radeon_ring_write(rdev, PACKET0(GB_SELECT, 0));
 	radeon_ring_write(rdev, 0);
-	radeon_ring_write(rdev, PACKET0(R300_GB_ENABLE, 0));
+	radeon_ring_write(rdev, PACKET0(GB_ENABLE, 0));
 	radeon_ring_write(rdev, 0);
 	radeon_ring_write(rdev, PACKET0(0x42C8, 0));
 	radeon_ring_write(rdev, (1 << rdev->num_gb_pipes) - 1);
-	radeon_ring_write(rdev, PACKET0(R500_VAP_INDEX_OFFSET, 0));
+	radeon_ring_write(rdev, PACKET0(VAP_INDEX_OFFSET, 0));
 	radeon_ring_write(rdev, 0);
-	radeon_ring_write(rdev, PACKET0(R300_RB3D_DSTCACHE_CTLSTAT, 0));
-	radeon_ring_write(rdev, R300_RB3D_DC_FLUSH | R300_RB3D_DC_FREE);
-	radeon_ring_write(rdev, PACKET0(R300_RB3D_ZCACHE_CTLSTAT, 0));
-	radeon_ring_write(rdev, R300_ZC_FLUSH | R300_ZC_FREE);
-	radeon_ring_write(rdev, PACKET0(RADEON_WAIT_UNTIL, 0));
-	radeon_ring_write(rdev,
-			  RADEON_WAIT_2D_IDLECLEAN |
-			  RADEON_WAIT_3D_IDLECLEAN);
-	radeon_ring_write(rdev, PACKET0(R300_GB_AA_CONFIG, 0));
+	radeon_ring_write(rdev, PACKET0(RB3D_DSTCACHE_CTLSTAT, 0));
+	radeon_ring_write(rdev, RB3D_DC_FLUSH | RB3D_DC_FREE);
+	radeon_ring_write(rdev, PACKET0(ZB_ZCACHE_CTLSTAT, 0));
+	radeon_ring_write(rdev, ZC_FLUSH | ZC_FREE);
+	radeon_ring_write(rdev, PACKET0(WAIT_UNTIL, 0));
+	radeon_ring_write(rdev, WAIT_2D_IDLECLEAN | WAIT_3D_IDLECLEAN);
+	radeon_ring_write(rdev, PACKET0(GB_AA_CONFIG, 0));
 	radeon_ring_write(rdev, 0);
-	radeon_ring_write(rdev, PACKET0(R300_RB3D_DSTCACHE_CTLSTAT, 0));
-	radeon_ring_write(rdev, R300_RB3D_DC_FLUSH | R300_RB3D_DC_FREE);
-	radeon_ring_write(rdev, PACKET0(R300_RB3D_ZCACHE_CTLSTAT, 0));
-	radeon_ring_write(rdev, R300_ZC_FLUSH | R300_ZC_FREE);
-	radeon_ring_write(rdev, PACKET0(R300_GB_MSPOS0, 0));
-	radeon_ring_write(rdev,
-			  ((6 << R300_MS_X0_SHIFT) |
-			   (6 << R300_MS_Y0_SHIFT) |
-			   (6 << R300_MS_X1_SHIFT) |
-			   (6 << R300_MS_Y1_SHIFT) |
-			   (6 << R300_MS_X2_SHIFT) |
-			   (6 << R300_MS_Y2_SHIFT) |
-			   (6 << R300_MSBD0_Y_SHIFT) |
-			   (6 << R300_MSBD0_X_SHIFT)));
-	radeon_ring_write(rdev, PACKET0(R300_GB_MSPOS1, 0));
+	radeon_ring_write(rdev, PACKET0(RB3D_DSTCACHE_CTLSTAT, 0));
+	radeon_ring_write(rdev, RB3D_DC_FLUSH | RB3D_DC_FREE);
+	radeon_ring_write(rdev, PACKET0(ZB_ZCACHE_CTLSTAT, 0));
+	radeon_ring_write(rdev, ZC_FLUSH | ZC_FREE);
+	radeon_ring_write(rdev, PACKET0(GB_MSPOS0, 0));
 	radeon_ring_write(rdev,
-			  ((6 << R300_MS_X3_SHIFT) |
-			   (6 << R300_MS_Y3_SHIFT) |
-			   (6 << R300_MS_X4_SHIFT) |
-			   (6 << R300_MS_Y4_SHIFT) |
-			   (6 << R300_MS_X5_SHIFT) |
-			   (6 << R300_MS_Y5_SHIFT) |
-			   (6 << R300_MSBD1_SHIFT)));
-	radeon_ring_write(rdev, PACKET0(R300_GA_ENHANCE, 0));
-	radeon_ring_write(rdev, R300_GA_DEADLOCK_CNTL | R300_GA_FASTSYNC_CNTL);
-	radeon_ring_write(rdev, PACKET0(R300_GA_POLY_MODE, 0));
+			  ((6 << MS_X0_SHIFT) |
+			   (6 << MS_Y0_SHIFT) |
+			   (6 << MS_X1_SHIFT) |
+			   (6 << MS_Y1_SHIFT) |
+			   (6 << MS_X2_SHIFT) |
+			   (6 << MS_Y2_SHIFT) |
+			   (6 << MSBD0_Y_SHIFT) |
+			   (6 << MSBD0_X_SHIFT)));
+	radeon_ring_write(rdev, PACKET0(GB_MSPOS1, 0));
 	radeon_ring_write(rdev,
-			  R300_FRONT_PTYPE_TRIANGE | R300_BACK_PTYPE_TRIANGE);
-	radeon_ring_write(rdev, PACKET0(R300_GA_ROUND_MODE, 0));
-	radeon_ring_write(rdev,
-			  R300_GEOMETRY_ROUND_NEAREST |
-			  R300_COLOR_ROUND_NEAREST);
+			  ((6 << MS_X3_SHIFT) |
+			   (6 << MS_Y3_SHIFT) |
+			   (6 << MS_X4_SHIFT) |
+			   (6 << MS_Y4_SHIFT) |
+			   (6 << MS_X5_SHIFT) |
+			   (6 << MS_Y5_SHIFT) |
+			   (6 << MSBD1_SHIFT)));
+	radeon_ring_write(rdev, PACKET0(GA_ENHANCE, 0));
+	radeon_ring_write(rdev, GA_DEADLOCK_CNTL | GA_FASTSYNC_CNTL);
+	radeon_ring_write(rdev, PACKET0(GA_POLY_MODE, 0));
+	radeon_ring_write(rdev, FRONT_PTYPE_TRIANGE | BACK_PTYPE_TRIANGE);
+	radeon_ring_write(rdev, PACKET0(GA_ROUND_MODE, 0));
+	radeon_ring_write(rdev, GEOMETRY_ROUND_NEAREST | COLOR_ROUND_NEAREST);
 	radeon_ring_write(rdev, PACKET0(0x20C8, 0));
 	radeon_ring_write(rdev, 0);
 	radeon_ring_unlock_commit(rdev);
@@ -242,8 +215,8 @@ int rv515_mc_wait_for_idle(struct radeon_device *rdev)
 
 	for (i = 0; i < rdev->usec_timeout; i++) {
 		/* read MC_STATUS */
-		tmp = RREG32_MC(RV515_MC_STATUS);
-		if (tmp & RV515_MC_STATUS_IDLE) {
+		tmp = RREG32_MC(MC_STATUS);
+		if (tmp & MC_STATUS_IDLE) {
 			return 0;
 		}
 		DRM_UDELAY(1);
@@ -291,33 +264,33 @@ int rv515_ga_reset(struct radeon_device *rdev)
 	reinit_cp = rdev->cp.ready;
 	rdev->cp.ready = false;
 	for (i = 0; i < rdev->usec_timeout; i++) {
-		WREG32(RADEON_CP_CSQ_MODE, 0);
-		WREG32(RADEON_CP_CSQ_CNTL, 0);
-		WREG32(RADEON_RBBM_SOFT_RESET, 0x32005);
-		(void)RREG32(RADEON_RBBM_SOFT_RESET);
+		WREG32(CP_CSQ_MODE, 0);
+		WREG32(CP_CSQ_CNTL, 0);
+		WREG32(RBBM_SOFT_RESET, 0x32005);
+		(void)RREG32(RBBM_SOFT_RESET);
 		udelay(200);
-		WREG32(RADEON_RBBM_SOFT_RESET, 0);
+		WREG32(RBBM_SOFT_RESET, 0);
 		/* Wait to prevent race in RBBM_STATUS */
 		mdelay(1);
-		tmp = RREG32(RADEON_RBBM_STATUS);
+		tmp = RREG32(RBBM_STATUS);
 		if (tmp & ((1 << 20) | (1 << 26))) {
 			DRM_ERROR("VAP & CP still busy (RBBM_STATUS=0x%08X)\n", tmp);
 			/* GA still busy soft reset it */
 			WREG32(0x429C, 0x200);
-			WREG32(R300_VAP_PVS_STATE_FLUSH_REG, 0);
+			WREG32(VAP_PVS_STATE_FLUSH_REG, 0);
 			WREG32(0x43E0, 0);
 			WREG32(0x43E4, 0);
 			WREG32(0x24AC, 0);
 		}
 		/* Wait to prevent race in RBBM_STATUS */
 		mdelay(1);
-		tmp = RREG32(RADEON_RBBM_STATUS);
+		tmp = RREG32(RBBM_STATUS);
 		if (!(tmp & ((1 << 20) | (1 << 26)))) {
 			break;
 		}
 	}
 	for (i = 0; i < rdev->usec_timeout; i++) {
-		tmp = RREG32(RADEON_RBBM_STATUS);
+		tmp = RREG32(RBBM_STATUS);
 		if (!(tmp & ((1 << 20) | (1 << 26)))) {
 			DRM_INFO("GA reset succeed (RBBM_STATUS=0x%08X)\n",
 				 tmp);
@@ -331,7 +304,7 @@ int rv515_ga_reset(struct radeon_device *rdev)
 		}
 		DRM_UDELAY(1);
 	}
-	tmp = RREG32(RADEON_RBBM_STATUS);
+	tmp = RREG32(RBBM_STATUS);
 	DRM_ERROR("Failed to reset GA ! (RBBM_STATUS=0x%08X)\n", tmp);
 	return -1;
 }
@@ -341,7 +314,7 @@ int rv515_gpu_reset(struct radeon_device *rdev)
 	uint32_t status;
 
 	/* reset order likely matter */
-	status = RREG32(RADEON_RBBM_STATUS);
+	status = RREG32(RBBM_STATUS);
 	/* reset HDP */
 	r100_hdp_reset(rdev);
 	/* reset rb2d */
@@ -353,12 +326,12 @@ int rv515_gpu_reset(struct radeon_device *rdev)
 		rv515_ga_reset(rdev);
 	}
 	/* reset CP */
-	status = RREG32(RADEON_RBBM_STATUS);
+	status = RREG32(RBBM_STATUS);
 	if (status & (1 << 16)) {
 		r100_cp_reset(rdev);
 	}
 	/* Check if GPU is idle */
-	status = RREG32(RADEON_RBBM_STATUS);
+	status = RREG32(RBBM_STATUS);
 	if (status & (1 << 31)) {
 		DRM_ERROR("Failed to reset GPU (RBBM_STATUS=0x%08X)\n", status);
 		return -1;
@@ -377,8 +350,7 @@ static void rv515_vram_get_type(struct radeon_device *rdev)
 
 	rdev->mc.vram_width = 128;
 	rdev->mc.vram_is_ddr = true;
-	tmp = RREG32_MC(RV515_MC_CNTL);
-	tmp &= RV515_MEM_NUM_CHANNELS_MASK;
+	tmp = RREG32_MC(RV515_MC_CNTL) & MEM_NUM_CHANNELS_MASK;
 	switch (tmp) {
 	case 0:
 		rdev->mc.vram_width = 64;
@@ -394,11 +366,17 @@ static void rv515_vram_get_type(struct radeon_device *rdev)
 
 void rv515_vram_info(struct radeon_device *rdev)
 {
+	fixed20_12 a;
+
 	rv515_vram_get_type(rdev);
-	rdev->mc.vram_size = RREG32(RADEON_CONFIG_MEMSIZE);
 
-	rdev->mc.aper_base = drm_get_resource_start(rdev->ddev, 0);
-	rdev->mc.aper_size = drm_get_resource_len(rdev->ddev, 0);
+	r100_vram_init_sizes(rdev);
+	/* FIXME: we should enforce default clock in case GPU is not in
+	 * default setup
+	 */
+	a.full = rfixed_const(100);
+	rdev->pm.sclk.full = rfixed_const(rdev->clock.default_sclk);
+	rdev->pm.sclk.full = rfixed_div(rdev->pm.sclk, a);
 }
 
 
@@ -409,35 +387,35 @@ uint32_t rv515_mc_rreg(struct radeon_device *rdev, uint32_t reg)
 {
 	uint32_t r;
 
-	WREG32(R520_MC_IND_INDEX, 0x7f0000 | (reg & 0xffff));
-	r = RREG32(R520_MC_IND_DATA);
-	WREG32(R520_MC_IND_INDEX, 0);
+	WREG32(MC_IND_INDEX, 0x7f0000 | (reg & 0xffff));
+	r = RREG32(MC_IND_DATA);
+	WREG32(MC_IND_INDEX, 0);
 	return r;
 }
 
 void rv515_mc_wreg(struct radeon_device *rdev, uint32_t reg, uint32_t v)
 {
-	WREG32(R520_MC_IND_INDEX, 0xff0000 | ((reg) & 0xffff));
-	WREG32(R520_MC_IND_DATA, (v));
-	WREG32(R520_MC_IND_INDEX, 0);
+	WREG32(MC_IND_INDEX, 0xff0000 | ((reg) & 0xffff));
+	WREG32(MC_IND_DATA, (v));
+	WREG32(MC_IND_INDEX, 0);
 }
 
 uint32_t rv515_pcie_rreg(struct radeon_device *rdev, uint32_t reg)
 {
 	uint32_t r;
 
-	WREG32(RADEON_PCIE_INDEX, ((reg) & 0x7ff));
-	(void)RREG32(RADEON_PCIE_INDEX);
-	r = RREG32(RADEON_PCIE_DATA);
+	WREG32(PCIE_INDEX, ((reg) & 0x7ff));
+	(void)RREG32(PCIE_INDEX);
+	r = RREG32(PCIE_DATA);
 	return r;
 }
 
 void rv515_pcie_wreg(struct radeon_device *rdev, uint32_t reg, uint32_t v)
 {
-	WREG32(RADEON_PCIE_INDEX, ((reg) & 0x7ff));
-	(void)RREG32(RADEON_PCIE_INDEX);
-	WREG32(RADEON_PCIE_DATA, (v));
-	(void)RREG32(RADEON_PCIE_DATA);
+	WREG32(PCIE_INDEX, ((reg) & 0x7ff));
+	(void)RREG32(PCIE_INDEX);
+	WREG32(PCIE_DATA, (v));
+	(void)RREG32(PCIE_DATA);
 }
 
 
@@ -452,13 +430,13 @@ static int rv515_debugfs_pipes_info(struct seq_file *m, void *data)
 	struct radeon_device *rdev = dev->dev_private;
 	uint32_t tmp;
 
-	tmp = RREG32(R400_GB_PIPE_SELECT);
+	tmp = RREG32(GB_PIPE_SELECT);
 	seq_printf(m, "GB_PIPE_SELECT 0x%08x\n", tmp);
-	tmp = RREG32(R500_SU_REG_DEST);
+	tmp = RREG32(SU_REG_DEST);
 	seq_printf(m, "SU_REG_DEST 0x%08x\n", tmp);
-	tmp = RREG32(R300_GB_TILE_CONFIG);
+	tmp = RREG32(GB_TILE_CONFIG);
 	seq_printf(m, "GB_TILE_CONFIG 0x%08x\n", tmp);
-	tmp = RREG32(R300_DST_PIPE_CONFIG);
+	tmp = RREG32(DST_PIPE_CONFIG);
 	seq_printf(m, "DST_PIPE_CONFIG 0x%08x\n", tmp);
 	return 0;
 }
@@ -509,9 +487,9 @@ int rv515_debugfs_ga_info_init(struct radeon_device *rdev)
 /*
  * Asic initialization
  */
-static const unsigned r500_reg_safe_bm[159] = {
+static const unsigned r500_reg_safe_bm[219] = {
+	0xFFFFFFFF, 0xFFFFFFFF, 0xFFFFFFFF, 0xFFFFFFFF,
 	0xFFFFFFFF, 0xFFFFFFFF, 0xFFFFFFFF, 0xFFFFFFFF,
-	0xFFFFFFBF, 0xFFFFFFFF, 0xFFFFFFBF, 0xFFFFFFFF,
 	0xFFFFFFFF, 0xFFFFFFFF, 0xFFFFFFFF, 0xFFFFFFFF,
 	0xFFFFFFFF, 0xFFFFFFFF, 0xFFFFFFFF, 0xFFFFFFFF,
 	0xFFFFFFFF, 0xFFFFFFFF, 0xFFFFFFFF, 0xFFFFFFFF,
@@ -549,14 +527,575 @@ static const unsigned r500_reg_safe_bm[159] = {
 	0xFFFFFFFF, 0xFFFFFFFF, 0xFFFFFFFF, 0xFFFFFFFF,
 	0xFFFFFFFF, 0xFFFFFFFF, 0xFFFFFFFF, 0xFF80FFFF,
 	0x00000000, 0x00000000, 0x00000000, 0x00000000,
-	0x0003FC01, 0x3FFFFCF8, 0xFE800B19,
+	0x0003FC01, 0x3FFFFCF8, 0xFE800B19, 0xFFFFFFFF,
+	0xFFFFFFFF, 0xFFFFFFFF, 0xFFFFFFFF, 0xFFFFFFFF,
+	0xFFFFFFFF, 0xFFFFFFFF, 0xFFFFFFFF, 0xFFFFFFFF,
+	0xFFFFFFFF, 0xFFFFFFFF, 0xFFFFFFFF, 0xFFFFFFFF,
+	0xFFFFFFFF, 0xFFFFFFFF, 0xFFFFFFFF, 0xFFFFFFFF,
+	0xFFFFFFFF, 0xFFFFFFFF, 0xFFFFFFFF, 0xFFFFFFFF,
+	0xFFFFFFFF, 0xFFFFFFFF, 0xFFFFFFFF, 0xFFFFFFFF,
+	0xFFFFFFFF, 0xFFFFFFFF, 0xFFFFFFFF, 0xFFFFFFFF,
+	0xFFFFFFFF, 0xFFFFFFFF, 0xFFFFFFFF, 0xFFFFFFFF,
+	0xFFFFFFFF, 0xFFFFFFFF, 0xFFFFFFFF, 0xFFFFFFFF,
+	0xFFFFFFFF, 0xFFFFFFFF, 0xFFFFFFFF, 0xFFFFFFFF,
+	0xFFFFFFFF, 0xFFFFFFFF, 0xFFFFFFFF, 0xFFFFFFFF,
+	0xFFFFFFFF, 0xFFFFFFFF, 0xFFFFFFFF, 0xFFFFFFFF,
+	0xFFFFFFFF, 0xFFFFFFFF, 0xFFFFFFFF, 0xFFFFFFFF,
+	0xFFFFFFFF, 0xFFFFFFFF, 0xFFFFFFFF, 0xFFFFFFFF,
+	0xFFFFFFFF, 0xFFFFFFFF, 0xFFFFFFFF,
 };
 
-
-
 int rv515_init(struct radeon_device *rdev)
 {
 	rdev->config.r300.reg_safe_bm = r500_reg_safe_bm;
 	rdev->config.r300.reg_safe_bm_size = ARRAY_SIZE(r500_reg_safe_bm);
 	return 0;
 }
+
+void atom_rv515_force_tv_scaler(struct radeon_device *rdev)
+{
+
+	WREG32(0x659C, 0x0);
+	WREG32(0x6594, 0x705);
+	WREG32(0x65A4, 0x10001);
+	WREG32(0x65D8, 0x0);
+	WREG32(0x65B0, 0x0);
+	WREG32(0x65C0, 0x0);
+	WREG32(0x65D4, 0x0);
+	WREG32(0x6578, 0x0);
+	WREG32(0x657C, 0x841880A8);
+	WREG32(0x6578, 0x1);
+	WREG32(0x657C, 0x84208680);
+	WREG32(0x6578, 0x2);
+	WREG32(0x657C, 0xBFF880B0);
+	WREG32(0x6578, 0x100);
+	WREG32(0x657C, 0x83D88088);
+	WREG32(0x6578, 0x101);
+	WREG32(0x657C, 0x84608680);
+	WREG32(0x6578, 0x102);
+	WREG32(0x657C, 0xBFF080D0);
+	WREG32(0x6578, 0x200);
+	WREG32(0x657C, 0x83988068);
+	WREG32(0x6578, 0x201);
+	WREG32(0x657C, 0x84A08680);
+	WREG32(0x6578, 0x202);
+	WREG32(0x657C, 0xBFF080F8);
+	WREG32(0x6578, 0x300);
+	WREG32(0x657C, 0x83588058);
+	WREG32(0x6578, 0x301);
+	WREG32(0x657C, 0x84E08660);
+	WREG32(0x6578, 0x302);
+	WREG32(0x657C, 0xBFF88120);
+	WREG32(0x6578, 0x400);
+	WREG32(0x657C, 0x83188040);
+	WREG32(0x6578, 0x401);
+	WREG32(0x657C, 0x85008660);
+	WREG32(0x6578, 0x402);
+	WREG32(0x657C, 0xBFF88150);
+	WREG32(0x6578, 0x500);
+	WREG32(0x657C, 0x82D88030);
+	WREG32(0x6578, 0x501);
+	WREG32(0x657C, 0x85408640);
+	WREG32(0x6578, 0x502);
+	WREG32(0x657C, 0xBFF88180);
+	WREG32(0x6578, 0x600);
+	WREG32(0x657C, 0x82A08018);
+	WREG32(0x6578, 0x601);
+	WREG32(0x657C, 0x85808620);
+	WREG32(0x6578, 0x602);
+	WREG32(0x657C, 0xBFF081B8);
+	WREG32(0x6578, 0x700);
+	WREG32(0x657C, 0x82608010);
+	WREG32(0x6578, 0x701);
+	WREG32(0x657C, 0x85A08600);
+	WREG32(0x6578, 0x702);
+	WREG32(0x657C, 0x800081F0);
+	WREG32(0x6578, 0x800);
+	WREG32(0x657C, 0x8228BFF8);
+	WREG32(0x6578, 0x801);
+	WREG32(0x657C, 0x85E085E0);
+	WREG32(0x6578, 0x802);
+	WREG32(0x657C, 0xBFF88228);
+	WREG32(0x6578, 0x10000);
+	WREG32(0x657C, 0x82A8BF00);
+	WREG32(0x6578, 0x10001);
+	WREG32(0x657C, 0x82A08CC0);
+	WREG32(0x6578, 0x10002);
+	WREG32(0x657C, 0x8008BEF8);
+	WREG32(0x6578, 0x10100);
+	WREG32(0x657C, 0x81F0BF28);
+	WREG32(0x6578, 0x10101);
+	WREG32(0x657C, 0x83608CA0);
+	WREG32(0x6578, 0x10102);
+	WREG32(0x657C, 0x8018BED0);
+	WREG32(0x6578, 0x10200);
+	WREG32(0x657C, 0x8148BF38);
+	WREG32(0x6578, 0x10201);
+	WREG32(0x657C, 0x84408C80);
+	WREG32(0x6578, 0x10202);
+	WREG32(0x657C, 0x8008BEB8);
+	WREG32(0x6578, 0x10300);
+	WREG32(0x657C, 0x80B0BF78);
+	WREG32(0x6578, 0x10301);
+	WREG32(0x657C, 0x85008C20);
+	WREG32(0x6578, 0x10302);
+	WREG32(0x657C, 0x8020BEA0);
+	WREG32(0x6578, 0x10400);
+	WREG32(0x657C, 0x8028BF90);
+	WREG32(0x6578, 0x10401);
+	WREG32(0x657C, 0x85E08BC0);
+	WREG32(0x6578, 0x10402);
+	WREG32(0x657C, 0x8018BE90);
+	WREG32(0x6578, 0x10500);
+	WREG32(0x657C, 0xBFB8BFB0);
+	WREG32(0x6578, 0x10501);
+	WREG32(0x657C, 0x86C08B40);
+	WREG32(0x6578, 0x10502);
+	WREG32(0x657C, 0x8010BE90);
+	WREG32(0x6578, 0x10600);
+	WREG32(0x657C, 0xBF58BFC8);
+	WREG32(0x6578, 0x10601);
+	WREG32(0x657C, 0x87A08AA0);
+	WREG32(0x6578, 0x10602);
+	WREG32(0x657C, 0x8010BE98);
+	WREG32(0x6578, 0x10700);
+	WREG32(0x657C, 0xBF10BFF0);
+	WREG32(0x6578, 0x10701);
+	WREG32(0x657C, 0x886089E0);
+	WREG32(0x6578, 0x10702);
+	WREG32(0x657C, 0x8018BEB0);
+	WREG32(0x6578, 0x10800);
+	WREG32(0x657C, 0xBED8BFE8);
+	WREG32(0x6578, 0x10801);
+	WREG32(0x657C, 0x89408940);
+	WREG32(0x6578, 0x10802);
+	WREG32(0x657C, 0xBFE8BED8);
+	WREG32(0x6578, 0x20000);
+	WREG32(0x657C, 0x80008000);
+	WREG32(0x6578, 0x20001);
+	WREG32(0x657C, 0x90008000);
+	WREG32(0x6578, 0x20002);
+	WREG32(0x657C, 0x80008000);
+	WREG32(0x6578, 0x20003);
+	WREG32(0x657C, 0x80008000);
+	WREG32(0x6578, 0x20100);
+	WREG32(0x657C, 0x80108000);
+	WREG32(0x6578, 0x20101);
+	WREG32(0x657C, 0x8FE0BF70);
+	WREG32(0x6578, 0x20102);
+	WREG32(0x657C, 0xBFE880C0);
+	WREG32(0x6578, 0x20103);
+	WREG32(0x657C, 0x80008000);
+	WREG32(0x6578, 0x20200);
+	WREG32(0x657C, 0x8018BFF8);
+	WREG32(0x6578, 0x20201);
+	WREG32(0x657C, 0x8F80BF08);
+	WREG32(0x6578, 0x20202);
+	WREG32(0x657C, 0xBFD081A0);
+	WREG32(0x6578, 0x20203);
+	WREG32(0x657C, 0xBFF88000);
+	WREG32(0x6578, 0x20300);
+	WREG32(0x657C, 0x80188000);
+	WREG32(0x6578, 0x20301);
+	WREG32(0x657C, 0x8EE0BEC0);
+	WREG32(0x6578, 0x20302);
+	WREG32(0x657C, 0xBFB082A0);
+	WREG32(0x6578, 0x20303);
+	WREG32(0x657C, 0x80008000);
+	WREG32(0x6578, 0x20400);
+	WREG32(0x657C, 0x80188000);
+	WREG32(0x6578, 0x20401);
+	WREG32(0x657C, 0x8E00BEA0);
+	WREG32(0x6578, 0x20402);
+	WREG32(0x657C, 0xBF8883C0);
+	WREG32(0x6578, 0x20403);
+	WREG32(0x657C, 0x80008000);
+	WREG32(0x6578, 0x20500);
+	WREG32(0x657C, 0x80188000);
+	WREG32(0x6578, 0x20501);
+	WREG32(0x657C, 0x8D00BE90);
+	WREG32(0x6578, 0x20502);
+	WREG32(0x657C, 0xBF588500);
+	WREG32(0x6578, 0x20503);
+	WREG32(0x657C, 0x80008008);
+	WREG32(0x6578, 0x20600);
+	WREG32(0x657C, 0x80188000);
+	WREG32(0x6578, 0x20601);
+	WREG32(0x657C, 0x8BC0BE98);
+	WREG32(0x6578, 0x20602);
+	WREG32(0x657C, 0xBF308660);
+	WREG32(0x6578, 0x20603);
+	WREG32(0x657C, 0x80008008);
+	WREG32(0x6578, 0x20700);
+	WREG32(0x657C, 0x80108000);
+	WREG32(0x6578, 0x20701);
+	WREG32(0x657C, 0x8A80BEB0);
+	WREG32(0x6578, 0x20702);
+	WREG32(0x657C, 0xBF0087C0);
+	WREG32(0x6578, 0x20703);
+	WREG32(0x657C, 0x80008008);
+	WREG32(0x6578, 0x20800);
+	WREG32(0x657C, 0x80108000);
+	WREG32(0x6578, 0x20801);
+	WREG32(0x657C, 0x8920BED0);
+	WREG32(0x6578, 0x20802);
+	WREG32(0x657C, 0xBED08920);
+	WREG32(0x6578, 0x20803);
+	WREG32(0x657C, 0x80008010);
+	WREG32(0x6578, 0x30000);
+	WREG32(0x657C, 0x90008000);
+	WREG32(0x6578, 0x30001);
+	WREG32(0x657C, 0x80008000);
+	WREG32(0x6578, 0x30100);
+	WREG32(0x657C, 0x8FE0BF90);
+	WREG32(0x6578, 0x30101);
+	WREG32(0x657C, 0xBFF880A0);
+	WREG32(0x6578, 0x30200);
+	WREG32(0x657C, 0x8F60BF40);
+	WREG32(0x6578, 0x30201);
+	WREG32(0x657C, 0xBFE88180);
+	WREG32(0x6578, 0x30300);
+	WREG32(0x657C, 0x8EC0BF00);
+	WREG32(0x6578, 0x30301);
+	WREG32(0x657C, 0xBFC88280);
+	WREG32(0x6578, 0x30400);
+	WREG32(0x657C, 0x8DE0BEE0);
+	WREG32(0x6578, 0x30401);
+	WREG32(0x657C, 0xBFA083A0);
+	WREG32(0x6578, 0x30500);
+	WREG32(0x657C, 0x8CE0BED0);
+	WREG32(0x6578, 0x30501);
+	WREG32(0x657C, 0xBF7884E0);
+	WREG32(0x6578, 0x30600);
+	WREG32(0x657C, 0x8BA0BED8);
+	WREG32(0x6578, 0x30601);
+	WREG32(0x657C, 0xBF508640);
+	WREG32(0x6578, 0x30700);
+	WREG32(0x657C, 0x8A60BEE8);
+	WREG32(0x6578, 0x30701);
+	WREG32(0x657C, 0xBF2087A0);
+	WREG32(0x6578, 0x30800);
+	WREG32(0x657C, 0x8900BF00);
+	WREG32(0x6578, 0x30801);
+	WREG32(0x657C, 0xBF008900);
+}
+
+struct rv515_watermark {
+	u32        lb_request_fifo_depth;
+	fixed20_12 num_line_pair;
+	fixed20_12 estimated_width;
+	fixed20_12 worst_case_latency;
+	fixed20_12 consumption_rate;
+	fixed20_12 active_time;
+	fixed20_12 dbpp;
+	fixed20_12 priority_mark_max;
+	fixed20_12 priority_mark;
+	fixed20_12 sclk;
+};
+
+void rv515_crtc_bandwidth_compute(struct radeon_device *rdev,
+				  struct radeon_crtc *crtc,
+				  struct rv515_watermark *wm)
+{
+	struct drm_display_mode *mode = &crtc->base.mode;
+	fixed20_12 a, b, c;
+	fixed20_12 pclk, request_fifo_depth, tolerable_latency, estimated_width;
+	fixed20_12 consumption_time, line_time, chunk_time, read_delay_latency;
+
+	if (!crtc->base.enabled) {
+		/* FIXME: wouldn't it better to set priority mark to maximum */
+		wm->lb_request_fifo_depth = 4;
+		return;
+	}
+
+	if (crtc->vsc.full > rfixed_const(2))
+		wm->num_line_pair.full = rfixed_const(2);
+	else
+		wm->num_line_pair.full = rfixed_const(1);
+
+	b.full = rfixed_const(mode->crtc_hdisplay);
+	c.full = rfixed_const(256);
+	a.full = rfixed_mul(wm->num_line_pair, b);
+	request_fifo_depth.full = rfixed_div(a, c);
+	if (a.full < rfixed_const(4)) {
+		wm->lb_request_fifo_depth = 4;
+	} else {
+		wm->lb_request_fifo_depth = rfixed_trunc(request_fifo_depth);
+	}
+
+	/* Determine consumption rate
+	 *  pclk = pixel clock period(ns) = 1000 / (mode.clock / 1000)
+	 *  vtaps = number of vertical taps,
+	 *  vsc = vertical scaling ratio, defined as source/destination
+	 *  hsc = horizontal scaling ration, defined as source/destination
+	 */
+	a.full = rfixed_const(mode->clock);
+	b.full = rfixed_const(1000);
+	a.full = rfixed_div(a, b);
+	pclk.full = rfixed_div(b, a);
+	if (crtc->rmx_type != RMX_OFF) {
+		b.full = rfixed_const(2);
+		if (crtc->vsc.full > b.full)
+			b.full = crtc->vsc.full;
+		b.full = rfixed_mul(b, crtc->hsc);
+		c.full = rfixed_const(2);
+		b.full = rfixed_div(b, c);
+		consumption_time.full = rfixed_div(pclk, b);
+	} else {
+		consumption_time.full = pclk.full;
+	}
+	a.full = rfixed_const(1);
+	wm->consumption_rate.full = rfixed_div(a, consumption_time);
+
+
+	/* Determine line time
+	 *  LineTime = total time for one line of displayhtotal
+	 *  LineTime = total number of horizontal pixels
+	 *  pclk = pixel clock period(ns)
+	 */
+	a.full = rfixed_const(crtc->base.mode.crtc_htotal);
+	line_time.full = rfixed_mul(a, pclk);
+
+	/* Determine active time
+	 *  ActiveTime = time of active region of display within one line,
+	 *  hactive = total number of horizontal active pixels
+	 *  htotal = total number of horizontal pixels
+	 */
+	a.full = rfixed_const(crtc->base.mode.crtc_htotal);
+	b.full = rfixed_const(crtc->base.mode.crtc_hdisplay);
+	wm->active_time.full = rfixed_mul(line_time, b);
+	wm->active_time.full = rfixed_div(wm->active_time, a);
+
+	/* Determine chunk time
+	 * ChunkTime = the time it takes the DCP to send one chunk of data
+	 * to the LB which consists of pipeline delay and inter chunk gap
+	 * sclk = system clock(Mhz)
+	 */
+	a.full = rfixed_const(600 * 1000);
+	chunk_time.full = rfixed_div(a, rdev->pm.sclk);
+	read_delay_latency.full = rfixed_const(1000);
+
+	/* Determine the worst case latency
+	 * NumLinePair = Number of line pairs to request(1=2 lines, 2=4 lines)
+	 * WorstCaseLatency = worst case time from urgent to when the MC starts
+	 *                    to return data
+	 * READ_DELAY_IDLE_MAX = constant of 1us
+	 * ChunkTime = time it takes the DCP to send one chunk of data to the LB
+	 *             which consists of pipeline delay and inter chunk gap
+	 */
+	if (rfixed_trunc(wm->num_line_pair) > 1) {
+		a.full = rfixed_const(3);
+		wm->worst_case_latency.full = rfixed_mul(a, chunk_time);
+		wm->worst_case_latency.full += read_delay_latency.full;
+	} else {
+		wm->worst_case_latency.full = chunk_time.full + read_delay_latency.full;
+	}
+
+	/* Determine the tolerable latency
+	 * TolerableLatency = Any given request has only 1 line time
+	 *                    for the data to be returned
+	 * LBRequestFifoDepth = Number of chunk requests the LB can
+	 *                      put into the request FIFO for a display
+	 *  LineTime = total time for one line of display
+	 *  ChunkTime = the time it takes the DCP to send one chunk
+	 *              of data to the LB which consists of
+	 *  pipeline delay and inter chunk gap
+	 */
+	if ((2+wm->lb_request_fifo_depth) >= rfixed_trunc(request_fifo_depth)) {
+		tolerable_latency.full = line_time.full;
+	} else {
+		tolerable_latency.full = rfixed_const(wm->lb_request_fifo_depth - 2);
+		tolerable_latency.full = request_fifo_depth.full - tolerable_latency.full;
+		tolerable_latency.full = rfixed_mul(tolerable_latency, chunk_time);
+		tolerable_latency.full = line_time.full - tolerable_latency.full;
+	}
+	/* We assume worst case 32bits (4 bytes) */
+	wm->dbpp.full = rfixed_const(2 * 16);
+
+	/* Determine the maximum priority mark
+	 *  width = viewport width in pixels
+	 */
+	a.full = rfixed_const(16);
+	wm->priority_mark_max.full = rfixed_const(crtc->base.mode.crtc_hdisplay);
+	wm->priority_mark_max.full = rfixed_div(wm->priority_mark_max, a);
+
+	/* Determine estimated width */
+	estimated_width.full = tolerable_latency.full - wm->worst_case_latency.full;
+	estimated_width.full = rfixed_div(estimated_width, consumption_time);
+	if (rfixed_trunc(estimated_width) > crtc->base.mode.crtc_hdisplay) {
+		wm->priority_mark.full = rfixed_const(10);
+	} else {
+		a.full = rfixed_const(16);
+		wm->priority_mark.full = rfixed_div(estimated_width, a);
+		wm->priority_mark.full = wm->priority_mark_max.full - wm->priority_mark.full;
+	}
+}
+
+void rv515_bandwidth_avivo_update(struct radeon_device *rdev)
+{
+	struct drm_display_mode *mode0 = NULL;
+	struct drm_display_mode *mode1 = NULL;
+	struct rv515_watermark wm0;
+	struct rv515_watermark wm1;
+	u32 tmp;
+	fixed20_12 priority_mark02, priority_mark12, fill_rate;
+	fixed20_12 a, b;
+
+	if (rdev->mode_info.crtcs[0]->base.enabled)
+		mode0 = &rdev->mode_info.crtcs[0]->base.mode;
+	if (rdev->mode_info.crtcs[1]->base.enabled)
+		mode1 = &rdev->mode_info.crtcs[1]->base.mode;
+	rs690_line_buffer_adjust(rdev, mode0, mode1);
+
+	rv515_crtc_bandwidth_compute(rdev, rdev->mode_info.crtcs[0], &wm0);
+	rv515_crtc_bandwidth_compute(rdev, rdev->mode_info.crtcs[1], &wm1);
+
+	tmp = wm0.lb_request_fifo_depth;
+	tmp |= wm1.lb_request_fifo_depth << 16;
+	WREG32(LB_MAX_REQ_OUTSTANDING, tmp);
+
+	if (mode0 && mode1) {
+		if (rfixed_trunc(wm0.dbpp) > 64)
+			a.full = rfixed_div(wm0.dbpp, wm0.num_line_pair);
+		else
+			a.full = wm0.num_line_pair.full;
+		if (rfixed_trunc(wm1.dbpp) > 64)
+			b.full = rfixed_div(wm1.dbpp, wm1.num_line_pair);
+		else
+			b.full = wm1.num_line_pair.full;
+		a.full += b.full;
+		fill_rate.full = rfixed_div(wm0.sclk, a);
+		if (wm0.consumption_rate.full > fill_rate.full) {
+			b.full = wm0.consumption_rate.full - fill_rate.full;
+			b.full = rfixed_mul(b, wm0.active_time);
+			a.full = rfixed_const(16);
+			b.full = rfixed_div(b, a);
+			a.full = rfixed_mul(wm0.worst_case_latency,
+						wm0.consumption_rate);
+			priority_mark02.full = a.full + b.full;
+		} else {
+			a.full = rfixed_mul(wm0.worst_case_latency,
+						wm0.consumption_rate);
+			b.full = rfixed_const(16 * 1000);
+			priority_mark02.full = rfixed_div(a, b);
+		}
+		if (wm1.consumption_rate.full > fill_rate.full) {
+			b.full = wm1.consumption_rate.full - fill_rate.full;
+			b.full = rfixed_mul(b, wm1.active_time);
+			a.full = rfixed_const(16);
+			b.full = rfixed_div(b, a);
+			a.full = rfixed_mul(wm1.worst_case_latency,
+						wm1.consumption_rate);
+			priority_mark12.full = a.full + b.full;
+		} else {
+			a.full = rfixed_mul(wm1.worst_case_latency,
+						wm1.consumption_rate);
+			b.full = rfixed_const(16 * 1000);
+			priority_mark12.full = rfixed_div(a, b);
+		}
+		if (wm0.priority_mark.full > priority_mark02.full)
+			priority_mark02.full = wm0.priority_mark.full;
+		if (rfixed_trunc(priority_mark02) < 0)
+			priority_mark02.full = 0;
+		if (wm0.priority_mark_max.full > priority_mark02.full)
+			priority_mark02.full = wm0.priority_mark_max.full;
+		if (wm1.priority_mark.full > priority_mark12.full)
+			priority_mark12.full = wm1.priority_mark.full;
+		if (rfixed_trunc(priority_mark12) < 0)
+			priority_mark12.full = 0;
+		if (wm1.priority_mark_max.full > priority_mark12.full)
+			priority_mark12.full = wm1.priority_mark_max.full;
+		WREG32(D1MODE_PRIORITY_A_CNT, rfixed_trunc(priority_mark02));
+		WREG32(D1MODE_PRIORITY_B_CNT, rfixed_trunc(priority_mark02));
+		WREG32(D2MODE_PRIORITY_A_CNT, rfixed_trunc(priority_mark12));
+		WREG32(D2MODE_PRIORITY_B_CNT, rfixed_trunc(priority_mark12));
+	} else if (mode0) {
+		if (rfixed_trunc(wm0.dbpp) > 64)
+			a.full = rfixed_div(wm0.dbpp, wm0.num_line_pair);
+		else
+			a.full = wm0.num_line_pair.full;
+		fill_rate.full = rfixed_div(wm0.sclk, a);
+		if (wm0.consumption_rate.full > fill_rate.full) {
+			b.full = wm0.consumption_rate.full - fill_rate.full;
+			b.full = rfixed_mul(b, wm0.active_time);
+			a.full = rfixed_const(16);
+			b.full = rfixed_div(b, a);
+			a.full = rfixed_mul(wm0.worst_case_latency,
+						wm0.consumption_rate);
+			priority_mark02.full = a.full + b.full;
+		} else {
+			a.full = rfixed_mul(wm0.worst_case_latency,
+						wm0.consumption_rate);
+			b.full = rfixed_const(16);
+			priority_mark02.full = rfixed_div(a, b);
+		}
+		if (wm0.priority_mark.full > priority_mark02.full)
+			priority_mark02.full = wm0.priority_mark.full;
+		if (rfixed_trunc(priority_mark02) < 0)
+			priority_mark02.full = 0;
+		if (wm0.priority_mark_max.full > priority_mark02.full)
+			priority_mark02.full = wm0.priority_mark_max.full;
+		WREG32(D1MODE_PRIORITY_A_CNT, rfixed_trunc(priority_mark02));
+		WREG32(D1MODE_PRIORITY_B_CNT, rfixed_trunc(priority_mark02));
+		WREG32(D2MODE_PRIORITY_A_CNT, MODE_PRIORITY_OFF);
+		WREG32(D2MODE_PRIORITY_B_CNT, MODE_PRIORITY_OFF);
+	} else {
+		if (rfixed_trunc(wm1.dbpp) > 64)
+			a.full = rfixed_div(wm1.dbpp, wm1.num_line_pair);
+		else
+			a.full = wm1.num_line_pair.full;
+		fill_rate.full = rfixed_div(wm1.sclk, a);
+		if (wm1.consumption_rate.full > fill_rate.full) {
+			b.full = wm1.consumption_rate.full - fill_rate.full;
+			b.full = rfixed_mul(b, wm1.active_time);
+			a.full = rfixed_const(16);
+			b.full = rfixed_div(b, a);
+			a.full = rfixed_mul(wm1.worst_case_latency,
+						wm1.consumption_rate);
+			priority_mark12.full = a.full + b.full;
+		} else {
+			a.full = rfixed_mul(wm1.worst_case_latency,
+						wm1.consumption_rate);
+			b.full = rfixed_const(16 * 1000);
+			priority_mark12.full = rfixed_div(a, b);
+		}
+		if (wm1.priority_mark.full > priority_mark12.full)
+			priority_mark12.full = wm1.priority_mark.full;
+		if (rfixed_trunc(priority_mark12) < 0)
+			priority_mark12.full = 0;
+		if (wm1.priority_mark_max.full > priority_mark12.full)
+			priority_mark12.full = wm1.priority_mark_max.full;
+		WREG32(D1MODE_PRIORITY_A_CNT, MODE_PRIORITY_OFF);
+		WREG32(D1MODE_PRIORITY_B_CNT, MODE_PRIORITY_OFF);
+		WREG32(D2MODE_PRIORITY_A_CNT, rfixed_trunc(priority_mark12));
+		WREG32(D2MODE_PRIORITY_B_CNT, rfixed_trunc(priority_mark12));
+	}
+}
+
+void rv515_bandwidth_update(struct radeon_device *rdev)
+{
+	uint32_t tmp;
+	struct drm_display_mode *mode0 = NULL;
+	struct drm_display_mode *mode1 = NULL;
+
+	if (rdev->mode_info.crtcs[0]->base.enabled)
+		mode0 = &rdev->mode_info.crtcs[0]->base.mode;
+	if (rdev->mode_info.crtcs[1]->base.enabled)
+		mode1 = &rdev->mode_info.crtcs[1]->base.mode;
+	/*
+	 * Set display0/1 priority up in the memory controller for
+	 * modes if the user specifies HIGH for displaypriority
+	 * option.
+	 */
+	if (rdev->disp_priority == 2) {
+		tmp = RREG32_MC(MC_MISC_LAT_TIMER);
+		tmp &= ~MC_DISP1R_INIT_LAT_MASK;
+		tmp &= ~MC_DISP0R_INIT_LAT_MASK;
+		if (mode1)
+			tmp |= (1 << MC_DISP1R_INIT_LAT_SHIFT);
+		if (mode0)
+			tmp |= (1 << MC_DISP0R_INIT_LAT_SHIFT);
+		WREG32_MC(MC_MISC_LAT_TIMER, tmp);
+	}
+	rv515_bandwidth_avivo_update(rdev);
+}
diff --git a/drivers/gpu/drm/radeon/rv515r.h b/drivers/gpu/drm/radeon/rv515r.h
new file mode 100644
index 00000000000..f3cf8403990
--- /dev/null
+++ b/drivers/gpu/drm/radeon/rv515r.h
@@ -0,0 +1,170 @@
+/*
+ * Copyright 2008 Advanced Micro Devices, Inc.
+ * Copyright 2008 Red Hat Inc.
+ * Copyright 2009 Jerome Glisse.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: Dave Airlie
+ *          Alex Deucher
+ *          Jerome Glisse
+ */
+#ifndef RV515R_H
+#define RV515R_H
+
+/* RV515 registers */
+#define PCIE_INDEX			0x0030
+#define PCIE_DATA			0x0034
+#define	MC_IND_INDEX			0x0070
+#define		MC_IND_WR_EN				(1 << 24)
+#define	MC_IND_DATA			0x0074
+#define	RBBM_SOFT_RESET			0x00F0
+#define	CONFIG_MEMSIZE			0x00F8
+#define HDP_FB_LOCATION			0x0134
+#define	CP_CSQ_CNTL			0x0740
+#define	CP_CSQ_MODE			0x0744
+#define	CP_CSQ_ADDR			0x07F0
+#define	CP_CSQ_DATA			0x07F4
+#define	CP_CSQ_STAT			0x07F8
+#define	CP_CSQ2_STAT			0x07FC
+#define	RBBM_STATUS			0x0E40
+#define	DST_PIPE_CONFIG			0x170C
+#define	WAIT_UNTIL			0x1720
+#define		WAIT_2D_IDLE				(1 << 14)
+#define		WAIT_3D_IDLE				(1 << 15)
+#define		WAIT_2D_IDLECLEAN			(1 << 16)
+#define		WAIT_3D_IDLECLEAN			(1 << 17)
+#define	ISYNC_CNTL			0x1724
+#define		ISYNC_ANY2D_IDLE3D			(1 << 0)
+#define		ISYNC_ANY3D_IDLE2D			(1 << 1)
+#define		ISYNC_TRIG2D_IDLE3D			(1 << 2)
+#define		ISYNC_TRIG3D_IDLE2D			(1 << 3)
+#define		ISYNC_WAIT_IDLEGUI			(1 << 4)
+#define		ISYNC_CPSCRATCH_IDLEGUI			(1 << 5)
+#define	VAP_INDEX_OFFSET		0x208C
+#define	VAP_PVS_STATE_FLUSH_REG		0x2284
+#define	GB_ENABLE			0x4008
+#define	GB_MSPOS0			0x4010
+#define		MS_X0_SHIFT				0
+#define		MS_Y0_SHIFT				4
+#define		MS_X1_SHIFT				8
+#define		MS_Y1_SHIFT				12
+#define		MS_X2_SHIFT				16
+#define		MS_Y2_SHIFT				20
+#define		MSBD0_Y_SHIFT				24
+#define		MSBD0_X_SHIFT				28
+#define	GB_MSPOS1			0x4014
+#define		MS_X3_SHIFT				0
+#define		MS_Y3_SHIFT				4
+#define		MS_X4_SHIFT				8
+#define		MS_Y4_SHIFT				12
+#define		MS_X5_SHIFT				16
+#define		MS_Y5_SHIFT				20
+#define		MSBD1_SHIFT				24
+#define GB_TILE_CONFIG			0x4018
+#define		ENABLE_TILING				(1 << 0)
+#define		PIPE_COUNT_MASK				0x0000000E
+#define		PIPE_COUNT_SHIFT			1
+#define		TILE_SIZE_8				(0 << 4)
+#define		TILE_SIZE_16				(1 << 4)
+#define		TILE_SIZE_32				(2 << 4)
+#define		SUBPIXEL_1_12				(0 << 16)
+#define		SUBPIXEL_1_16				(1 << 16)
+#define	GB_SELECT			0x401C
+#define	GB_AA_CONFIG			0x4020
+#define	GB_PIPE_SELECT			0x402C
+#define	GA_ENHANCE			0x4274
+#define		GA_DEADLOCK_CNTL			(1 << 0)
+#define		GA_FASTSYNC_CNTL			(1 << 1)
+#define	GA_POLY_MODE			0x4288
+#define		FRONT_PTYPE_POINT			(0 << 4)
+#define		FRONT_PTYPE_LINE			(1 << 4)
+#define		FRONT_PTYPE_TRIANGE			(2 << 4)
+#define		BACK_PTYPE_POINT			(0 << 7)
+#define		BACK_PTYPE_LINE				(1 << 7)
+#define		BACK_PTYPE_TRIANGE			(2 << 7)
+#define	GA_ROUND_MODE			0x428C
+#define		GEOMETRY_ROUND_TRUNC			(0 << 0)
+#define		GEOMETRY_ROUND_NEAREST			(1 << 0)
+#define		COLOR_ROUND_TRUNC			(0 << 2)
+#define		COLOR_ROUND_NEAREST			(1 << 2)
+#define	SU_REG_DEST			0x42C8
+#define	RB3D_DSTCACHE_CTLSTAT		0x4E4C
+#define		RB3D_DC_FLUSH				(2 << 0)
+#define		RB3D_DC_FREE				(2 << 2)
+#define		RB3D_DC_FINISH				(1 << 4)
+#define ZB_ZCACHE_CTLSTAT		0x4F18
+#define		ZC_FLUSH				(1 << 0)
+#define		ZC_FREE					(1 << 1)
+#define DC_LB_MEMORY_SPLIT		0x6520
+#define		DC_LB_MEMORY_SPLIT_MASK			0x00000003
+#define		DC_LB_MEMORY_SPLIT_SHIFT		0
+#define		DC_LB_MEMORY_SPLIT_D1HALF_D2HALF	0
+#define		DC_LB_MEMORY_SPLIT_D1_3Q_D2_1Q		1
+#define		DC_LB_MEMORY_SPLIT_D1_ONLY		2
+#define		DC_LB_MEMORY_SPLIT_D1_1Q_D2_3Q		3
+#define		DC_LB_MEMORY_SPLIT_SHIFT_MODE		(1 << 2)
+#define		DC_LB_DISP1_END_ADR_SHIFT		4
+#define		DC_LB_DISP1_END_ADR_MASK		0x00007FF0
+#define D1MODE_PRIORITY_A_CNT		0x6548
+#define		MODE_PRIORITY_MARK_MASK			0x00007FFF
+#define		MODE_PRIORITY_OFF			(1 << 16)
+#define		MODE_PRIORITY_ALWAYS_ON			(1 << 20)
+#define		MODE_PRIORITY_FORCE_MASK		(1 << 24)
+#define D1MODE_PRIORITY_B_CNT		0x654C
+#define LB_MAX_REQ_OUTSTANDING		0x6D58
+#define		LB_D1_MAX_REQ_OUTSTANDING_MASK		0x0000000F
+#define		LB_D1_MAX_REQ_OUTSTANDING_SHIFT		0
+#define		LB_D2_MAX_REQ_OUTSTANDING_MASK		0x000F0000
+#define		LB_D2_MAX_REQ_OUTSTANDING_SHIFT		16
+#define D2MODE_PRIORITY_A_CNT		0x6D48
+#define D2MODE_PRIORITY_B_CNT		0x6D4C
+
+/* ix[MC] registers */
+#define MC_FB_LOCATION			0x01
+#define		MC_FB_START_MASK			0x0000FFFF
+#define		MC_FB_START_SHIFT			0
+#define		MC_FB_TOP_MASK				0xFFFF0000
+#define		MC_FB_TOP_SHIFT				16
+#define MC_AGP_LOCATION			0x02
+#define		MC_AGP_START_MASK			0x0000FFFF
+#define		MC_AGP_START_SHIFT			0
+#define		MC_AGP_TOP_MASK				0xFFFF0000
+#define		MC_AGP_TOP_SHIFT			16
+#define MC_AGP_BASE			0x03
+#define MC_AGP_BASE_2			0x04
+#define	MC_CNTL				0x5
+#define		MEM_NUM_CHANNELS_MASK			0x00000003
+#define	MC_STATUS			0x08
+#define		MC_STATUS_IDLE				(1 << 4)
+#define	MC_MISC_LAT_TIMER		0x09
+#define		MC_CPR_INIT_LAT_MASK			0x0000000F
+#define		MC_VF_INIT_LAT_MASK			0x000000F0
+#define		MC_DISP0R_INIT_LAT_MASK			0x00000F00
+#define		MC_DISP0R_INIT_LAT_SHIFT		8
+#define		MC_DISP1R_INIT_LAT_MASK			0x0000F000
+#define		MC_DISP1R_INIT_LAT_SHIFT		12
+#define		MC_FIXED_INIT_LAT_MASK			0x000F0000
+#define		MC_E2R_INIT_LAT_MASK			0x00F00000
+#define		SAME_PAGE_PRIO_MASK			0x0F000000
+#define		MC_GLOBW_INIT_LAT_MASK			0xF0000000
+
+
+#endif
+
diff --git a/drivers/gpu/drm/radeon/rv770.c b/drivers/gpu/drm/radeon/rv770.c
index da50cc51ede..21d8ffd5730 100644
--- a/drivers/gpu/drm/radeon/rv770.c
+++ b/drivers/gpu/drm/radeon/rv770.c
@@ -67,7 +67,7 @@ int rv770_mc_init(struct radeon_device *rdev)
 		       "programming pipes. Bad things might happen.\n");
 	}
 
-	tmp = rdev->mc.vram_location + rdev->mc.vram_size - 1;
+	tmp = rdev->mc.vram_location + rdev->mc.mc_vram_size - 1;
 	tmp = REG_SET(R700_MC_FB_TOP, tmp >> 24);
 	tmp |= REG_SET(R700_MC_FB_BASE, rdev->mc.vram_location >> 24);
 	WREG32(R700_MC_VM_FB_LOCATION, tmp);
diff --git a/drivers/gpu/drm/ttm/ttm_bo.c b/drivers/gpu/drm/ttm/ttm_bo.c
index c1c407f7cca..c2b0d710d10 100644
--- a/drivers/gpu/drm/ttm/ttm_bo.c
+++ b/drivers/gpu/drm/ttm/ttm_bo.c
@@ -43,7 +43,6 @@
 #define TTM_BO_HASH_ORDER 13
 
 static int ttm_bo_setup_vm(struct ttm_buffer_object *bo);
-static void ttm_bo_unmap_virtual(struct ttm_buffer_object *bo);
 static int ttm_bo_swapout(struct ttm_mem_shrink *shrink);
 
 static inline uint32_t ttm_bo_type_flags(unsigned type)
@@ -224,6 +223,9 @@ static int ttm_bo_add_ttm(struct ttm_buffer_object *bo, bool zero_alloc)
 	TTM_ASSERT_LOCKED(&bo->mutex);
 	bo->ttm = NULL;
 
+	if (bdev->need_dma32)
+		page_flags |= TTM_PAGE_FLAG_DMA32;
+
 	switch (bo->type) {
 	case ttm_bo_type_device:
 		if (zero_alloc)
@@ -304,6 +306,9 @@ static int ttm_bo_handle_move_mem(struct ttm_buffer_object *bo,
 
 	}
 
+	if (bdev->driver->move_notify)
+		bdev->driver->move_notify(bo, mem);
+
 	if (!(old_man->flags & TTM_MEMTYPE_FLAG_FIXED) &&
 	    !(new_man->flags & TTM_MEMTYPE_FLAG_FIXED))
 		ret = ttm_bo_move_ttm(bo, evict, no_wait, mem);
@@ -655,31 +660,52 @@ retry_pre_get:
 	return 0;
 }
 
+static uint32_t ttm_bo_select_caching(struct ttm_mem_type_manager *man,
+				      uint32_t cur_placement,
+				      uint32_t proposed_placement)
+{
+	uint32_t caching = proposed_placement & TTM_PL_MASK_CACHING;
+	uint32_t result = proposed_placement & ~TTM_PL_MASK_CACHING;
+
+	/**
+	 * Keep current caching if possible.
+	 */
+
+	if ((cur_placement & caching) != 0)
+		result |= (cur_placement & caching);
+	else if ((man->default_caching & caching) != 0)
+		result |= man->default_caching;
+	else if ((TTM_PL_FLAG_CACHED & caching) != 0)
+		result |= TTM_PL_FLAG_CACHED;
+	else if ((TTM_PL_FLAG_WC & caching) != 0)
+		result |= TTM_PL_FLAG_WC;
+	else if ((TTM_PL_FLAG_UNCACHED & caching) != 0)
+		result |= TTM_PL_FLAG_UNCACHED;
+
+	return result;
+}
+
+
 static bool ttm_bo_mt_compatible(struct ttm_mem_type_manager *man,
 				 bool disallow_fixed,
 				 uint32_t mem_type,
-				 uint32_t mask, uint32_t *res_mask)
+				 uint32_t proposed_placement,
+				 uint32_t *masked_placement)
 {
 	uint32_t cur_flags = ttm_bo_type_flags(mem_type);
 
 	if ((man->flags & TTM_MEMTYPE_FLAG_FIXED) && disallow_fixed)
 		return false;
 
-	if ((cur_flags & mask & TTM_PL_MASK_MEM) == 0)
+	if ((cur_flags & proposed_placement & TTM_PL_MASK_MEM) == 0)
 		return false;
 
-	if ((mask & man->available_caching) == 0)
+	if ((proposed_placement & man->available_caching) == 0)
 		return false;
-	if (mask & man->default_caching)
-		cur_flags |= man->default_caching;
-	else if (mask & TTM_PL_FLAG_CACHED)
-		cur_flags |= TTM_PL_FLAG_CACHED;
-	else if (mask & TTM_PL_FLAG_WC)
-		cur_flags |= TTM_PL_FLAG_WC;
-	else
-		cur_flags |= TTM_PL_FLAG_UNCACHED;
 
-	*res_mask = cur_flags;
+	cur_flags |= (proposed_placement & man->available_caching);
+
+	*masked_placement = cur_flags;
 	return true;
 }
 
@@ -723,6 +749,9 @@ int ttm_bo_mem_space(struct ttm_buffer_object *bo,
 		if (!type_ok)
 			continue;
 
+		cur_flags = ttm_bo_select_caching(man, bo->mem.placement,
+						  cur_flags);
+
 		if (mem_type == TTM_PL_SYSTEM)
 			break;
 
@@ -779,6 +808,9 @@ int ttm_bo_mem_space(struct ttm_buffer_object *bo,
 					  proposed_placement, &cur_flags))
 			continue;
 
+		cur_flags = ttm_bo_select_caching(man, bo->mem.placement,
+						  cur_flags);
+
 		ret = ttm_bo_mem_force_space(bdev, mem, mem_type,
 					     interruptible, no_wait);
 
@@ -1150,13 +1182,14 @@ static int ttm_bo_force_list_clean(struct ttm_bo_device *bdev,
 
 int ttm_bo_clean_mm(struct ttm_bo_device *bdev, unsigned mem_type)
 {
-	struct ttm_mem_type_manager *man = &bdev->man[mem_type];
+	struct ttm_mem_type_manager *man;
 	int ret = -EINVAL;
 
 	if (mem_type >= TTM_NUM_MEM_TYPES) {
 		printk(KERN_ERR TTM_PFX "Illegal memory type %d\n", mem_type);
 		return ret;
 	}
+	man = &bdev->man[mem_type];
 
 	if (!man->has_type) {
 		printk(KERN_ERR TTM_PFX "Trying to take down uninitialized "
@@ -1305,7 +1338,8 @@ EXPORT_SYMBOL(ttm_bo_device_release);
 
 int ttm_bo_device_init(struct ttm_bo_device *bdev,
 		       struct ttm_mem_global *mem_glob,
-		       struct ttm_bo_driver *driver, uint64_t file_page_offset)
+		       struct ttm_bo_driver *driver, uint64_t file_page_offset,
+		       bool need_dma32)
 {
 	int ret = -EINVAL;
 
@@ -1342,6 +1376,7 @@ int ttm_bo_device_init(struct ttm_bo_device *bdev,
 	INIT_LIST_HEAD(&bdev->ddestroy);
 	INIT_LIST_HEAD(&bdev->swap_lru);
 	bdev->dev_mapping = NULL;
+	bdev->need_dma32 = need_dma32;
 	ttm_mem_init_shrink(&bdev->shrink, ttm_bo_swapout);
 	ret = ttm_mem_register_shrink(mem_glob, &bdev->shrink);
 	if (unlikely(ret != 0)) {
@@ -1419,6 +1454,7 @@ void ttm_bo_unmap_virtual(struct ttm_buffer_object *bo)
 
 	unmap_mapping_range(bdev->dev_mapping, offset, holelen, 1);
 }
+EXPORT_SYMBOL(ttm_bo_unmap_virtual);
 
 static void ttm_bo_vm_insert_rb(struct ttm_buffer_object *bo)
 {
@@ -1540,6 +1576,10 @@ int ttm_bo_wait(struct ttm_buffer_object *bo,
 			driver->sync_obj_unref(&sync_obj);
 			driver->sync_obj_unref(&tmp_obj);
 			spin_lock(&bo->lock);
+		} else {
+			spin_unlock(&bo->lock);
+			driver->sync_obj_unref(&sync_obj);
+			spin_lock(&bo->lock);
 		}
 	}
 	return 0;
diff --git a/drivers/gpu/drm/ttm/ttm_bo_util.c b/drivers/gpu/drm/ttm/ttm_bo_util.c
index bdec583901e..ad4ada07c6c 100644
--- a/drivers/gpu/drm/ttm/ttm_bo_util.c
+++ b/drivers/gpu/drm/ttm/ttm_bo_util.c
@@ -136,7 +136,8 @@ static int ttm_copy_io_page(void *dst, void *src, unsigned long page)
 }
 
 static int ttm_copy_io_ttm_page(struct ttm_tt *ttm, void *src,
-				unsigned long page)
+				unsigned long page,
+				pgprot_t prot)
 {
 	struct page *d = ttm_tt_get_page(ttm, page);
 	void *dst;
@@ -145,17 +146,35 @@ static int ttm_copy_io_ttm_page(struct ttm_tt *ttm, void *src,
 		return -ENOMEM;
 
 	src = (void *)((unsigned long)src + (page << PAGE_SHIFT));
-	dst = kmap(d);
+
+#ifdef CONFIG_X86
+	dst = kmap_atomic_prot(d, KM_USER0, prot);
+#else
+	if (pgprot_val(prot) != pgprot_val(PAGE_KERNEL))
+		dst = vmap(&d, 1, 0, prot);
+	else
+		dst = kmap(d);
+#endif
 	if (!dst)
 		return -ENOMEM;
 
 	memcpy_fromio(dst, src, PAGE_SIZE);
-	kunmap(d);
+
+#ifdef CONFIG_X86
+	kunmap_atomic(dst, KM_USER0);
+#else
+	if (pgprot_val(prot) != pgprot_val(PAGE_KERNEL))
+		vunmap(dst);
+	else
+		kunmap(d);
+#endif
+
 	return 0;
 }
 
 static int ttm_copy_ttm_io_page(struct ttm_tt *ttm, void *dst,
-				unsigned long page)
+				unsigned long page,
+				pgprot_t prot)
 {
 	struct page *s = ttm_tt_get_page(ttm, page);
 	void *src;
@@ -164,12 +183,28 @@ static int ttm_copy_ttm_io_page(struct ttm_tt *ttm, void *dst,
 		return -ENOMEM;
 
 	dst = (void *)((unsigned long)dst + (page << PAGE_SHIFT));
-	src = kmap(s);
+#ifdef CONFIG_X86
+	src = kmap_atomic_prot(s, KM_USER0, prot);
+#else
+	if (pgprot_val(prot) != pgprot_val(PAGE_KERNEL))
+		src = vmap(&s, 1, 0, prot);
+	else
+		src = kmap(s);
+#endif
 	if (!src)
 		return -ENOMEM;
 
 	memcpy_toio(dst, src, PAGE_SIZE);
-	kunmap(s);
+
+#ifdef CONFIG_X86
+	kunmap_atomic(src, KM_USER0);
+#else
+	if (pgprot_val(prot) != pgprot_val(PAGE_KERNEL))
+		vunmap(src);
+	else
+		kunmap(s);
+#endif
+
 	return 0;
 }
 
@@ -214,11 +249,17 @@ int ttm_bo_move_memcpy(struct ttm_buffer_object *bo,
 
 	for (i = 0; i < new_mem->num_pages; ++i) {
 		page = i * dir + add;
-		if (old_iomap == NULL)
-			ret = ttm_copy_ttm_io_page(ttm, new_iomap, page);
-		else if (new_iomap == NULL)
-			ret = ttm_copy_io_ttm_page(ttm, old_iomap, page);
-		else
+		if (old_iomap == NULL) {
+			pgprot_t prot = ttm_io_prot(old_mem->placement,
+						    PAGE_KERNEL);
+			ret = ttm_copy_ttm_io_page(ttm, new_iomap, page,
+						   prot);
+		} else if (new_iomap == NULL) {
+			pgprot_t prot = ttm_io_prot(new_mem->placement,
+						    PAGE_KERNEL);
+			ret = ttm_copy_io_ttm_page(ttm, old_iomap, page,
+						   prot);
+		} else
 			ret = ttm_copy_io_page(new_iomap, old_iomap, page);
 		if (ret)
 			goto out1;
@@ -509,8 +550,8 @@ int ttm_bo_move_accel_cleanup(struct ttm_buffer_object *bo,
 	if (evict) {
 		ret = ttm_bo_wait(bo, false, false, false);
 		spin_unlock(&bo->lock);
-		driver->sync_obj_unref(&bo->sync_obj);
-
+		if (tmp_obj)
+			driver->sync_obj_unref(&tmp_obj);
 		if (ret)
 			return ret;
 
@@ -532,6 +573,8 @@ int ttm_bo_move_accel_cleanup(struct ttm_buffer_object *bo,
 
 		set_bit(TTM_BO_PRIV_FLAG_MOVING, &bo->priv_flags);
 		spin_unlock(&bo->lock);
+		if (tmp_obj)
+			driver->sync_obj_unref(&tmp_obj);
 
 		ret = ttm_buffer_object_transfer(bo, &ghost_obj);
 		if (ret)
diff --git a/drivers/gpu/drm/ttm/ttm_bo_vm.c b/drivers/gpu/drm/ttm/ttm_bo_vm.c
index fe949a12fe4..33de7637c0c 100644
--- a/drivers/gpu/drm/ttm/ttm_bo_vm.c
+++ b/drivers/gpu/drm/ttm/ttm_bo_vm.c
@@ -101,6 +101,9 @@ static int ttm_bo_vm_fault(struct vm_area_struct *vma, struct vm_fault *vmf)
 		return VM_FAULT_NOPAGE;
 	}
 
+	if (bdev->driver->fault_reserve_notify)
+		bdev->driver->fault_reserve_notify(bo);
+
 	/*
 	 * Wait for buffer data in transit, due to a pipelined
 	 * move.
diff --git a/drivers/gpu/drm/ttm/ttm_tt.c b/drivers/gpu/drm/ttm/ttm_tt.c
index 75dc8bd2459..b8b6c4a5f98 100644
--- a/drivers/gpu/drm/ttm/ttm_tt.c
+++ b/drivers/gpu/drm/ttm/ttm_tt.c
@@ -86,10 +86,16 @@ void ttm_tt_cache_flush(struct page *pages[], unsigned long num_pages)
 	unsigned long i;
 
 	for (i = 0; i < num_pages; ++i) {
-		if (pages[i]) {
-			unsigned long start = (unsigned long)page_address(pages[i]);
-			flush_dcache_range(start, start + PAGE_SIZE);
-		}
+		struct page *page = pages[i];
+		void *page_virtual;
+
+		if (unlikely(page == NULL))
+			continue;
+
+		page_virtual = kmap_atomic(page, KM_USER0);
+		flush_dcache_range((unsigned long) page_virtual,
+				   (unsigned long) page_virtual + PAGE_SIZE);
+		kunmap_atomic(page_virtual, KM_USER0);
 	}
 #else
 	if (on_each_cpu(ttm_tt_ipi_handler, NULL, 1) != 0)
@@ -131,10 +137,17 @@ static void ttm_tt_free_page_directory(struct ttm_tt *ttm)
 
 static struct page *ttm_tt_alloc_page(unsigned page_flags)
 {
+	gfp_t gfp_flags = GFP_USER;
+
 	if (page_flags & TTM_PAGE_FLAG_ZERO_ALLOC)
-		return alloc_page(GFP_HIGHUSER | __GFP_ZERO);
+		gfp_flags |= __GFP_ZERO;
+
+	if (page_flags & TTM_PAGE_FLAG_DMA32)
+		gfp_flags |= __GFP_DMA32;
+	else
+		gfp_flags |= __GFP_HIGHMEM;
 
-	return alloc_page(GFP_HIGHUSER);
+	return alloc_page(gfp_flags);
 }
 
 static void ttm_tt_free_user_pages(struct ttm_tt *ttm)
diff --git a/drivers/hwmon/asus_atk0110.c b/drivers/hwmon/asus_atk0110.c
index bff0103610c..fe4fa29c921 100644
--- a/drivers/hwmon/asus_atk0110.c
+++ b/drivers/hwmon/asus_atk0110.c
@@ -593,7 +593,11 @@ static int atk_add_sensor(struct atk_data *data, union acpi_object *obj)
 	sensor->data = data;
 	sensor->id = flags->integer.value;
 	sensor->limit1 = limit1->integer.value;
-	sensor->limit2 = limit2->integer.value;
+	if (data->old_interface)
+		sensor->limit2 = limit2->integer.value;
+	else
+		/* The upper limit is expressed as delta from lower limit */
+		sensor->limit2 = sensor->limit1 + limit2->integer.value;
 
 	snprintf(sensor->input_attr_name, ATTR_NAME_SIZE,
 			"%s%d_input", base_name, start + *num);
diff --git a/drivers/hwmon/smsc47m1.c b/drivers/hwmon/smsc47m1.c
index a92dbb97ee9..ba75bfcf14c 100644
--- a/drivers/hwmon/smsc47m1.c
+++ b/drivers/hwmon/smsc47m1.c
@@ -86,6 +86,7 @@ superio_exit(void)
 #define SUPERIO_REG_ACT		0x30
 #define SUPERIO_REG_BASE	0x60
 #define SUPERIO_REG_DEVID	0x20
+#define SUPERIO_REG_DEVREV	0x21
 
 /* Logical device registers */
 
@@ -429,6 +430,9 @@ static int __init smsc47m1_find(unsigned short *addr,
 	 * The LPC47M292 (device id 0x6B) is somewhat compatible, but it
 	 * supports a 3rd fan, and the pin configuration registers are
 	 * unfortunately different.
+	 * The LPC47M233 has the same device id (0x6B) but is not compatible.
+	 * We check the high bit of the device revision register to
+	 * differentiate them.
 	 */
 	switch (val) {
 	case 0x51:
@@ -448,6 +452,13 @@ static int __init smsc47m1_find(unsigned short *addr,
 		sio_data->type = smsc47m1;
 		break;
 	case 0x6B:
+		if (superio_inb(SUPERIO_REG_DEVREV) & 0x80) {
+			pr_debug(DRVNAME ": "
+				 "Found SMSC LPC47M233, unsupported\n");
+			superio_exit();
+			return -ENODEV;
+		}
+
 		pr_info(DRVNAME ": Found SMSC LPC47M292\n");
 		sio_data->type = smsc47m2;
 		break;
diff --git a/drivers/i2c/busses/i2c-omap.c b/drivers/i2c/busses/i2c-omap.c
index fdd83277c8a..d258b02aef4 100644
--- a/drivers/i2c/busses/i2c-omap.c
+++ b/drivers/i2c/busses/i2c-omap.c
@@ -672,9 +672,10 @@ omap_i2c_isr(int this_irq, void *dev_id)
 			break;
 		}
 
+		err = 0;
+complete:
 		omap_i2c_write_reg(dev, OMAP_I2C_STAT_REG, stat);
 
-		err = 0;
 		if (stat & OMAP_I2C_STAT_NACK) {
 			err |= OMAP_I2C_STAT_NACK;
 			omap_i2c_write_reg(dev, OMAP_I2C_CON_REG,
@@ -685,16 +686,19 @@ omap_i2c_isr(int this_irq, void *dev_id)
 			err |= OMAP_I2C_STAT_AL;
 		}
 		if (stat & (OMAP_I2C_STAT_ARDY | OMAP_I2C_STAT_NACK |
-					OMAP_I2C_STAT_AL))
+					OMAP_I2C_STAT_AL)) {
 			omap_i2c_complete_cmd(dev, err);
+			return IRQ_HANDLED;
+		}
 		if (stat & (OMAP_I2C_STAT_RRDY | OMAP_I2C_STAT_RDR)) {
 			u8 num_bytes = 1;
 			if (dev->fifo_size) {
 				if (stat & OMAP_I2C_STAT_RRDY)
 					num_bytes = dev->fifo_size;
-				else
-					num_bytes = omap_i2c_read_reg(dev,
-							OMAP_I2C_BUFSTAT_REG);
+				else    /* read RXSTAT on RDR interrupt */
+					num_bytes = (omap_i2c_read_reg(dev,
+							OMAP_I2C_BUFSTAT_REG)
+							>> 8) & 0x3F;
 			}
 			while (num_bytes) {
 				num_bytes--;
@@ -731,9 +735,10 @@ omap_i2c_isr(int this_irq, void *dev_id)
 			if (dev->fifo_size) {
 				if (stat & OMAP_I2C_STAT_XRDY)
 					num_bytes = dev->fifo_size;
-				else
+				else    /* read TXSTAT on XDR interrupt */
 					num_bytes = omap_i2c_read_reg(dev,
-							OMAP_I2C_BUFSTAT_REG);
+							OMAP_I2C_BUFSTAT_REG)
+							& 0x3F;
 			}
 			while (num_bytes) {
 				num_bytes--;
@@ -760,6 +765,27 @@ omap_i2c_isr(int this_irq, void *dev_id)
 							"data to send\n");
 					break;
 				}
+
+				/*
+				 * OMAP3430 Errata 1.153: When an XRDY/XDR
+				 * is hit, wait for XUDF before writing data
+				 * to DATA_REG. Otherwise some data bytes can
+				 * be lost while transferring them from the
+				 * memory to the I2C interface.
+				 */
+
+				if (cpu_is_omap34xx()) {
+						while (!(stat & OMAP_I2C_STAT_XUDF)) {
+							if (stat & (OMAP_I2C_STAT_NACK | OMAP_I2C_STAT_AL)) {
+								omap_i2c_ack_stat(dev, stat & (OMAP_I2C_STAT_XRDY | OMAP_I2C_STAT_XDR));
+								err |= OMAP_I2C_STAT_XUDF;
+								goto complete;
+							}
+							cpu_relax();
+							stat = omap_i2c_read_reg(dev, OMAP_I2C_STAT_REG);
+						}
+				}
+
 				omap_i2c_write_reg(dev, OMAP_I2C_DATA_REG, w);
 			}
 			omap_i2c_ack_stat(dev,
@@ -879,7 +905,7 @@ omap_i2c_probe(struct platform_device *pdev)
 	i2c_set_adapdata(adap, dev);
 	adap->owner = THIS_MODULE;
 	adap->class = I2C_CLASS_HWMON;
-	strncpy(adap->name, "OMAP I2C adapter", sizeof(adap->name));
+	strlcpy(adap->name, "OMAP I2C adapter", sizeof(adap->name));
 	adap->algo = &omap_i2c_algo;
 	adap->dev.parent = &pdev->dev;
 
diff --git a/drivers/i2c/busses/i2c-s3c2410.c b/drivers/i2c/busses/i2c-s3c2410.c
index 8f42a4536cd..20bb0ceb027 100644
--- a/drivers/i2c/busses/i2c-s3c2410.c
+++ b/drivers/i2c/busses/i2c-s3c2410.c
@@ -763,11 +763,6 @@ static int s3c24xx_i2c_init(struct s3c24xx_i2c *i2c)
 	dev_info(i2c->dev, "bus frequency set to %d KHz\n", freq);
 	dev_dbg(i2c->dev, "S3C2410_IICCON=0x%02lx\n", iicon);
 
-	/* check for s3c2440 i2c controller  */
-
-	if (s3c24xx_i2c_is2440(i2c))
-		writel(0x0, i2c->regs + S3C2440_IICLC);
-
 	return 0;
 }
 
diff --git a/drivers/i2c/busses/i2c-sh_mobile.c b/drivers/i2c/busses/i2c-sh_mobile.c
index 4f3d99cd169..820487d0d5c 100644
--- a/drivers/i2c/busses/i2c-sh_mobile.c
+++ b/drivers/i2c/busses/i2c-sh_mobile.c
@@ -637,7 +637,7 @@ static void __exit sh_mobile_i2c_adap_exit(void)
 	platform_driver_unregister(&sh_mobile_i2c_driver);
 }
 
-module_init(sh_mobile_i2c_adap_init);
+subsys_initcall(sh_mobile_i2c_adap_init);
 module_exit(sh_mobile_i2c_adap_exit);
 
 MODULE_DESCRIPTION("SuperH Mobile I2C Bus Controller driver");
diff --git a/drivers/i2c/chips/tsl2550.c b/drivers/i2c/chips/tsl2550.c
index 1a9cc135219..b96f3025e58 100644
--- a/drivers/i2c/chips/tsl2550.c
+++ b/drivers/i2c/chips/tsl2550.c
@@ -27,7 +27,7 @@
 #include <linux/delay.h>
 
 #define TSL2550_DRV_NAME	"tsl2550"
-#define DRIVER_VERSION		"1.1.1"
+#define DRIVER_VERSION		"1.1.2"
 
 /*
  * Defines
@@ -189,13 +189,16 @@ static int tsl2550_calculate_lux(u8 ch0, u8 ch1)
 	u8 r = 128;
 
 	/* Avoid division by 0 and count 1 cannot be greater than count 0 */
-	if (c0 && (c1 <= c0))
-		r = c1 * 128 / c0;
+	if (c1 <= c0)
+		if (c0) {
+			r = c1 * 128 / c0;
+
+			/* Calculate LUX */
+			lux = ((c0 - c1) * ratio_lut[r]) / 256;
+		} else
+			lux = 0;
 	else
-		return -1;
-
-	/* Calculate LUX */
-	lux = ((c0 - c1) * ratio_lut[r]) / 256;
+		return -EAGAIN;
 
 	/* LUX range check */
 	return lux > TSL2550_MAX_LUX ? TSL2550_MAX_LUX : lux;
diff --git a/drivers/input/keyboard/matrix_keypad.c b/drivers/input/keyboard/matrix_keypad.c
index e9b2e7cb05b..541b981ff07 100644
--- a/drivers/input/keyboard/matrix_keypad.c
+++ b/drivers/input/keyboard/matrix_keypad.c
@@ -27,6 +27,7 @@ struct matrix_keypad {
 	const struct matrix_keypad_platform_data *pdata;
 	struct input_dev *input_dev;
 	unsigned short *keycodes;
+	unsigned int row_shift;
 
 	uint32_t last_key_state[MATRIX_MAX_COLS];
 	struct delayed_work work;
@@ -136,7 +137,7 @@ static void matrix_keypad_scan(struct work_struct *work)
 			if ((bits_changed & (1 << row)) == 0)
 				continue;
 
-			code = (row << 4) + col;
+			code = MATRIX_SCAN_CODE(row, col, keypad->row_shift);
 			input_event(input_dev, EV_MSC, MSC_SCAN, code);
 			input_report_key(input_dev,
 					 keypad->keycodes[code],
@@ -317,6 +318,7 @@ static int __devinit matrix_keypad_probe(struct platform_device *pdev)
 	struct matrix_keypad *keypad;
 	struct input_dev *input_dev;
 	unsigned short *keycodes;
+	unsigned int row_shift;
 	int i;
 	int err;
 
@@ -332,14 +334,11 @@ static int __devinit matrix_keypad_probe(struct platform_device *pdev)
 		return -EINVAL;
 	}
 
-	if (!keymap_data->max_keymap_size) {
-		dev_err(&pdev->dev, "invalid keymap data supplied\n");
-		return -EINVAL;
-	}
+	row_shift = get_count_order(pdata->num_col_gpios);
 
 	keypad = kzalloc(sizeof(struct matrix_keypad), GFP_KERNEL);
-	keycodes = kzalloc(keymap_data->max_keymap_size *
-				sizeof(keypad->keycodes),
+	keycodes = kzalloc((pdata->num_row_gpios << row_shift) *
+				sizeof(*keycodes),
 			   GFP_KERNEL);
 	input_dev = input_allocate_device();
 	if (!keypad || !keycodes || !input_dev) {
@@ -350,6 +349,7 @@ static int __devinit matrix_keypad_probe(struct platform_device *pdev)
 	keypad->input_dev = input_dev;
 	keypad->pdata = pdata;
 	keypad->keycodes = keycodes;
+	keypad->row_shift = row_shift;
 	keypad->stopped = true;
 	INIT_DELAYED_WORK(&keypad->work, matrix_keypad_scan);
 	spin_lock_init(&keypad->lock);
@@ -363,7 +363,7 @@ static int __devinit matrix_keypad_probe(struct platform_device *pdev)
 
 	input_dev->keycode	= keycodes;
 	input_dev->keycodesize	= sizeof(*keycodes);
-	input_dev->keycodemax	= keymap_data->max_keymap_size;
+	input_dev->keycodemax	= pdata->num_row_gpios << keypad->row_shift;
 
 	for (i = 0; i < keymap_data->keymap_size; i++) {
 		unsigned int key = keymap_data->keymap[i];
@@ -371,7 +371,7 @@ static int __devinit matrix_keypad_probe(struct platform_device *pdev)
 		unsigned int col = KEY_COL(key);
 		unsigned short code = KEY_VAL(key);
 
-		keycodes[(row << 4) + col] = code;
+		keycodes[MATRIX_SCAN_CODE(row, col, row_shift)] = code;
 		__set_bit(code, input_dev->keybit);
 	}
 	__clear_bit(KEY_RESERVED, input_dev->keybit);
diff --git a/drivers/input/misc/wistron_btns.c b/drivers/input/misc/wistron_btns.c
index 26e17a9a22e..27ee976eb54 100644
--- a/drivers/input/misc/wistron_btns.c
+++ b/drivers/input/misc/wistron_btns.c
@@ -611,6 +611,20 @@ static struct key_entry keymap_wistron_generic[] __initdata = {
 	{ KE_END, 0 }
 };
 
+static struct key_entry keymap_prestigio[] __initdata = {
+	{ KE_KEY,  0x11, {KEY_PROG1} },
+	{ KE_KEY,  0x12, {KEY_PROG2} },
+	{ KE_WIFI,  0x30 },
+	{ KE_KEY,  0x22, {KEY_REWIND} },
+	{ KE_KEY,  0x23, {KEY_FORWARD} },
+	{ KE_KEY,  0x24, {KEY_PLAYPAUSE} },
+	{ KE_KEY,  0x25, {KEY_STOPCD} },
+	{ KE_KEY,  0x31, {KEY_MAIL} },
+	{ KE_KEY,  0x36, {KEY_WWW} },
+	{ KE_END,  0 }
+};
+
+
 /*
  * If your machine is not here (which is currently rather likely), please send
  * a list of buttons and their key codes (reported when loading this module
@@ -971,6 +985,8 @@ static int __init select_keymap(void)
 	if (keymap_name != NULL) {
 		if (strcmp (keymap_name, "1557/MS2141") == 0)
 			keymap = keymap_wistron_ms2141;
+		else if (strcmp (keymap_name, "prestigio") == 0)
+			keymap = keymap_prestigio;
 		else if (strcmp (keymap_name, "generic") == 0)
 			keymap = keymap_wistron_generic;
 		else {
diff --git a/drivers/input/serio/hp_sdc_mlc.c b/drivers/input/serio/hp_sdc_mlc.c
index b587e2d576a..820e51673b2 100644
--- a/drivers/input/serio/hp_sdc_mlc.c
+++ b/drivers/input/serio/hp_sdc_mlc.c
@@ -296,7 +296,7 @@ static void hp_sdc_mlc_out(hil_mlc *mlc)
 	priv->tseq[3] = 0;
 	if (mlc->opacket & HIL_CTRL_APE) {
 		priv->tseq[3] |= HP_SDC_LPC_APE_IPF;
-		down_trylock(&mlc->csem);
+		BUG_ON(down_trylock(&mlc->csem));
 	}
  enqueue:
 	hp_sdc_enqueue_transaction(&priv->trans);
diff --git a/drivers/input/serio/i8042-x86ia64io.h b/drivers/input/serio/i8042-x86ia64io.h
index 924e8ed7f2c..ae04d8a494e 100644
--- a/drivers/input/serio/i8042-x86ia64io.h
+++ b/drivers/input/serio/i8042-x86ia64io.h
@@ -78,6 +78,14 @@ static struct dmi_system_id __initdata i8042_dmi_noloop_table[] = {
 		},
 	},
 	{
+		.ident = "ASUS G1S",
+		.matches = {
+			DMI_MATCH(DMI_BOARD_VENDOR, "ASUSTeK Computer Inc."),
+			DMI_MATCH(DMI_BOARD_NAME, "G1S"),
+			DMI_MATCH(DMI_BOARD_VERSION, "1.0"),
+		},
+	},
+	{
 		/* AUX LOOP command does not raise AUX IRQ */
 		.ident = "ASUS P65UP5",
 		.matches = {
diff --git a/drivers/isdn/mISDN/l1oip_core.c b/drivers/isdn/mISDN/l1oip_core.c
index 990e6a7e667..7e5f30dbc0a 100644
--- a/drivers/isdn/mISDN/l1oip_core.c
+++ b/drivers/isdn/mISDN/l1oip_core.c
@@ -731,10 +731,10 @@ l1oip_socket_thread(void *data)
 	while (!signal_pending(current)) {
 		struct kvec iov = {
 			.iov_base = recvbuf,
-			.iov_len = sizeof(recvbuf),
+			.iov_len = recvbuf_size,
 		};
 		recvlen = kernel_recvmsg(socket, &msg, &iov, 1,
-					 sizeof(recvbuf), 0);
+					 recvbuf_size, 0);
 		if (recvlen > 0) {
 			l1oip_socket_parse(hc, &sin_rx, recvbuf, recvlen);
 		} else {
@@ -1480,7 +1480,7 @@ l1oip_init(void)
 		return -ENOMEM;
 
 	l1oip_cnt = 0;
-	while (type[l1oip_cnt] && l1oip_cnt < MAX_CARDS) {
+	while (l1oip_cnt < MAX_CARDS && type[l1oip_cnt]) {
 		switch (type[l1oip_cnt] & 0xff) {
 		case 1:
 			pri = 0;
diff --git a/drivers/lguest/core.c b/drivers/lguest/core.c
index a6974e9b8eb..1e2cb846b3c 100644
--- a/drivers/lguest/core.c
+++ b/drivers/lguest/core.c
@@ -1,6 +1,8 @@
-/*P:400 This contains run_guest() which actually calls into the Host<->Guest
+/*P:400
+ * This contains run_guest() which actually calls into the Host<->Guest
  * Switcher and analyzes the return, such as determining if the Guest wants the
- * Host to do something.  This file also contains useful helper routines. :*/
+ * Host to do something.  This file also contains useful helper routines.
+:*/
 #include <linux/module.h>
 #include <linux/stringify.h>
 #include <linux/stddef.h>
@@ -24,7 +26,8 @@ static struct page **switcher_page;
 /* This One Big lock protects all inter-guest data structures. */
 DEFINE_MUTEX(lguest_lock);
 
-/*H:010 We need to set up the Switcher at a high virtual address.  Remember the
+/*H:010
+ * We need to set up the Switcher at a high virtual address.  Remember the
  * Switcher is a few hundred bytes of assembler code which actually changes the
  * CPU to run the Guest, and then changes back to the Host when a trap or
  * interrupt happens.
@@ -33,7 +36,8 @@ DEFINE_MUTEX(lguest_lock);
  * Host since it will be running as the switchover occurs.
  *
  * Trying to map memory at a particular address is an unusual thing to do, so
- * it's not a simple one-liner. */
+ * it's not a simple one-liner.
+ */
 static __init int map_switcher(void)
 {
 	int i, err;
@@ -47,8 +51,10 @@ static __init int map_switcher(void)
 	 * easy.
 	 */
 
-	/* We allocate an array of struct page pointers.  map_vm_area() wants
-	 * this, rather than just an array of pages. */
+	/*
+	 * We allocate an array of struct page pointers.  map_vm_area() wants
+	 * this, rather than just an array of pages.
+	 */
 	switcher_page = kmalloc(sizeof(switcher_page[0])*TOTAL_SWITCHER_PAGES,
 				GFP_KERNEL);
 	if (!switcher_page) {
@@ -56,8 +62,10 @@ static __init int map_switcher(void)
 		goto out;
 	}
 
-	/* Now we actually allocate the pages.  The Guest will see these pages,
-	 * so we make sure they're zeroed. */
+	/*
+	 * Now we actually allocate the pages.  The Guest will see these pages,
+	 * so we make sure they're zeroed.
+	 */
 	for (i = 0; i < TOTAL_SWITCHER_PAGES; i++) {
 		unsigned long addr = get_zeroed_page(GFP_KERNEL);
 		if (!addr) {
@@ -67,19 +75,23 @@ static __init int map_switcher(void)
 		switcher_page[i] = virt_to_page(addr);
 	}
 
-	/* First we check that the Switcher won't overlap the fixmap area at
+	/*
+	 * First we check that the Switcher won't overlap the fixmap area at
 	 * the top of memory.  It's currently nowhere near, but it could have
-	 * very strange effects if it ever happened. */
+	 * very strange effects if it ever happened.
+	 */
 	if (SWITCHER_ADDR + (TOTAL_SWITCHER_PAGES+1)*PAGE_SIZE > FIXADDR_START){
 		err = -ENOMEM;
 		printk("lguest: mapping switcher would thwack fixmap\n");
 		goto free_pages;
 	}
 
-	/* Now we reserve the "virtual memory area" we want: 0xFFC00000
+	/*
+	 * Now we reserve the "virtual memory area" we want: 0xFFC00000
 	 * (SWITCHER_ADDR).  We might not get it in theory, but in practice
 	 * it's worked so far.  The end address needs +1 because __get_vm_area
-	 * allocates an extra guard page, so we need space for that. */
+	 * allocates an extra guard page, so we need space for that.
+	 */
 	switcher_vma = __get_vm_area(TOTAL_SWITCHER_PAGES * PAGE_SIZE,
 				     VM_ALLOC, SWITCHER_ADDR, SWITCHER_ADDR
 				     + (TOTAL_SWITCHER_PAGES+1) * PAGE_SIZE);
@@ -89,11 +101,13 @@ static __init int map_switcher(void)
 		goto free_pages;
 	}
 
-	/* This code actually sets up the pages we've allocated to appear at
+	/*
+	 * This code actually sets up the pages we've allocated to appear at
 	 * SWITCHER_ADDR.  map_vm_area() takes the vma we allocated above, the
 	 * kind of pages we're mapping (kernel pages), and a pointer to our
 	 * array of struct pages.  It increments that pointer, but we don't
-	 * care. */
+	 * care.
+	 */
 	pagep = switcher_page;
 	err = map_vm_area(switcher_vma, PAGE_KERNEL_EXEC, &pagep);
 	if (err) {
@@ -101,8 +115,10 @@ static __init int map_switcher(void)
 		goto free_vma;
 	}
 
-	/* Now the Switcher is mapped at the right address, we can't fail!
-	 * Copy in the compiled-in Switcher code (from <arch>_switcher.S). */
+	/*
+	 * Now the Switcher is mapped at the right address, we can't fail!
+	 * Copy in the compiled-in Switcher code (from <arch>_switcher.S).
+	 */
 	memcpy(switcher_vma->addr, start_switcher_text,
 	       end_switcher_text - start_switcher_text);
 
@@ -124,8 +140,7 @@ out:
 }
 /*:*/
 
-/* Cleaning up the mapping when the module is unloaded is almost...
- * too easy. */
+/* Cleaning up the mapping when the module is unloaded is almost... too easy. */
 static void unmap_switcher(void)
 {
 	unsigned int i;
@@ -151,16 +166,19 @@ static void unmap_switcher(void)
  * But we can't trust the Guest: it might be trying to access the Launcher
  * code.  We have to check that the range is below the pfn_limit the Launcher
  * gave us.  We have to make sure that addr + len doesn't give us a false
- * positive by overflowing, too. */
+ * positive by overflowing, too.
+ */
 bool lguest_address_ok(const struct lguest *lg,
 		       unsigned long addr, unsigned long len)
 {
 	return (addr+len) / PAGE_SIZE < lg->pfn_limit && (addr+len >= addr);
 }
 
-/* This routine copies memory from the Guest.  Here we can see how useful the
+/*
+ * This routine copies memory from the Guest.  Here we can see how useful the
  * kill_lguest() routine we met in the Launcher can be: we return a random
- * value (all zeroes) instead of needing to return an error. */
+ * value (all zeroes) instead of needing to return an error.
+ */
 void __lgread(struct lg_cpu *cpu, void *b, unsigned long addr, unsigned bytes)
 {
 	if (!lguest_address_ok(cpu->lg, addr, bytes)
@@ -181,9 +199,11 @@ void __lgwrite(struct lg_cpu *cpu, unsigned long addr, const void *b,
 }
 /*:*/
 
-/*H:030 Let's jump straight to the the main loop which runs the Guest.
+/*H:030
+ * Let's jump straight to the the main loop which runs the Guest.
  * Remember, this is called by the Launcher reading /dev/lguest, and we keep
- * going around and around until something interesting happens. */
+ * going around and around until something interesting happens.
+ */
 int run_guest(struct lg_cpu *cpu, unsigned long __user *user)
 {
 	/* We stop running once the Guest is dead. */
@@ -195,10 +215,17 @@ int run_guest(struct lg_cpu *cpu, unsigned long __user *user)
 		if (cpu->hcall)
 			do_hypercalls(cpu);
 
-		/* It's possible the Guest did a NOTIFY hypercall to the
-		 * Launcher, in which case we return from the read() now. */
+		/*
+		 * It's possible the Guest did a NOTIFY hypercall to the
+		 * Launcher.
+		 */
 		if (cpu->pending_notify) {
+			/*
+			 * Does it just needs to write to a registered
+			 * eventfd (ie. the appropriate virtqueue thread)?
+			 */
 			if (!send_notify_to_eventfd(cpu)) {
+				/* OK, we tell the main Laucher. */
 				if (put_user(cpu->pending_notify, user))
 					return -EFAULT;
 				return sizeof(cpu->pending_notify);
@@ -209,29 +236,39 @@ int run_guest(struct lg_cpu *cpu, unsigned long __user *user)
 		if (signal_pending(current))
 			return -ERESTARTSYS;
 
-		/* Check if there are any interrupts which can be delivered now:
+		/*
+		 * Check if there are any interrupts which can be delivered now:
 		 * if so, this sets up the hander to be executed when we next
-		 * run the Guest. */
+		 * run the Guest.
+		 */
 		irq = interrupt_pending(cpu, &more);
 		if (irq < LGUEST_IRQS)
 			try_deliver_interrupt(cpu, irq, more);
 
-		/* All long-lived kernel loops need to check with this horrible
+		/*
+		 * All long-lived kernel loops need to check with this horrible
 		 * thing called the freezer.  If the Host is trying to suspend,
-		 * it stops us. */
+		 * it stops us.
+		 */
 		try_to_freeze();
 
-		/* Just make absolutely sure the Guest is still alive.  One of
-		 * those hypercalls could have been fatal, for example. */
+		/*
+		 * Just make absolutely sure the Guest is still alive.  One of
+		 * those hypercalls could have been fatal, for example.
+		 */
 		if (cpu->lg->dead)
 			break;
 
-		/* If the Guest asked to be stopped, we sleep.  The Guest's
-		 * clock timer will wake us. */
+		/*
+		 * If the Guest asked to be stopped, we sleep.  The Guest's
+		 * clock timer will wake us.
+		 */
 		if (cpu->halted) {
 			set_current_state(TASK_INTERRUPTIBLE);
-			/* Just before we sleep, make sure no interrupt snuck in
-			 * which we should be doing. */
+			/*
+			 * Just before we sleep, make sure no interrupt snuck in
+			 * which we should be doing.
+			 */
 			if (interrupt_pending(cpu, &more) < LGUEST_IRQS)
 				set_current_state(TASK_RUNNING);
 			else
@@ -239,8 +276,10 @@ int run_guest(struct lg_cpu *cpu, unsigned long __user *user)
 			continue;
 		}
 
-		/* OK, now we're ready to jump into the Guest.  First we put up
-		 * the "Do Not Disturb" sign: */
+		/*
+		 * OK, now we're ready to jump into the Guest.  First we put up
+		 * the "Do Not Disturb" sign:
+		 */
 		local_irq_disable();
 
 		/* Actually run the Guest until something happens. */
@@ -327,8 +366,10 @@ static void __exit fini(void)
 }
 /*:*/
 
-/* The Host side of lguest can be a module.  This is a nice way for people to
- * play with it.  */
+/*
+ * The Host side of lguest can be a module.  This is a nice way for people to
+ * play with it.
+ */
 module_init(init);
 module_exit(fini);
 MODULE_LICENSE("GPL");
diff --git a/drivers/lguest/hypercalls.c b/drivers/lguest/hypercalls.c
index c29ffa19cb7..83511eb0923 100644
--- a/drivers/lguest/hypercalls.c
+++ b/drivers/lguest/hypercalls.c
@@ -1,8 +1,10 @@
-/*P:500 Just as userspace programs request kernel operations through a system
+/*P:500
+ * Just as userspace programs request kernel operations through a system
  * call, the Guest requests Host operations through a "hypercall".  You might
  * notice this nomenclature doesn't really follow any logic, but the name has
  * been around for long enough that we're stuck with it.  As you'd expect, this
- * code is basically a one big switch statement. :*/
+ * code is basically a one big switch statement.
+:*/
 
 /*  Copyright (C) 2006 Rusty Russell IBM Corporation
 
@@ -28,30 +30,41 @@
 #include <asm/pgtable.h>
 #include "lg.h"
 
-/*H:120 This is the core hypercall routine: where the Guest gets what it wants.
- * Or gets killed.  Or, in the case of LHCALL_SHUTDOWN, both. */
+/*H:120
+ * This is the core hypercall routine: where the Guest gets what it wants.
+ * Or gets killed.  Or, in the case of LHCALL_SHUTDOWN, both.
+ */
 static void do_hcall(struct lg_cpu *cpu, struct hcall_args *args)
 {
 	switch (args->arg0) {
 	case LHCALL_FLUSH_ASYNC:
-		/* This call does nothing, except by breaking out of the Guest
-		 * it makes us process all the asynchronous hypercalls. */
+		/*
+		 * This call does nothing, except by breaking out of the Guest
+		 * it makes us process all the asynchronous hypercalls.
+		 */
 		break;
 	case LHCALL_SEND_INTERRUPTS:
-		/* This call does nothing too, but by breaking out of the Guest
-		 * it makes us process any pending interrupts. */
+		/*
+		 * This call does nothing too, but by breaking out of the Guest
+		 * it makes us process any pending interrupts.
+		 */
 		break;
 	case LHCALL_LGUEST_INIT:
-		/* You can't get here unless you're already initialized.  Don't
-		 * do that. */
+		/*
+		 * You can't get here unless you're already initialized.  Don't
+		 * do that.
+		 */
 		kill_guest(cpu, "already have lguest_data");
 		break;
 	case LHCALL_SHUTDOWN: {
-		/* Shutdown is such a trivial hypercall that we do it in four
-		 * lines right here. */
 		char msg[128];
-		/* If the lgread fails, it will call kill_guest() itself; the
-		 * kill_guest() with the message will be ignored. */
+		/*
+		 * Shutdown is such a trivial hypercall that we do it in five
+		 * lines right here.
+		 *
+		 * If the lgread fails, it will call kill_guest() itself; the
+		 * kill_guest() with the message will be ignored.
+		 */
 		__lgread(cpu, msg, args->arg1, sizeof(msg));
 		msg[sizeof(msg)-1] = '\0';
 		kill_guest(cpu, "CRASH: %s", msg);
@@ -60,16 +73,17 @@ static void do_hcall(struct lg_cpu *cpu, struct hcall_args *args)
 		break;
 	}
 	case LHCALL_FLUSH_TLB:
-		/* FLUSH_TLB comes in two flavors, depending on the
-		 * argument: */
+		/* FLUSH_TLB comes in two flavors, depending on the argument: */
 		if (args->arg1)
 			guest_pagetable_clear_all(cpu);
 		else
 			guest_pagetable_flush_user(cpu);
 		break;
 
-	/* All these calls simply pass the arguments through to the right
-	 * routines. */
+	/*
+	 * All these calls simply pass the arguments through to the right
+	 * routines.
+	 */
 	case LHCALL_NEW_PGTABLE:
 		guest_new_pagetable(cpu, args->arg1);
 		break;
@@ -112,15 +126,16 @@ static void do_hcall(struct lg_cpu *cpu, struct hcall_args *args)
 			kill_guest(cpu, "Bad hypercall %li\n", args->arg0);
 	}
 }
-/*:*/
 
-/*H:124 Asynchronous hypercalls are easy: we just look in the array in the
+/*H:124
+ * Asynchronous hypercalls are easy: we just look in the array in the
  * Guest's "struct lguest_data" to see if any new ones are marked "ready".
  *
  * We are careful to do these in order: obviously we respect the order the
  * Guest put them in the ring, but we also promise the Guest that they will
  * happen before any normal hypercall (which is why we check this before
- * checking for a normal hcall). */
+ * checking for a normal hcall).
+ */
 static void do_async_hcalls(struct lg_cpu *cpu)
 {
 	unsigned int i;
@@ -133,22 +148,28 @@ static void do_async_hcalls(struct lg_cpu *cpu)
 	/* We process "struct lguest_data"s hcalls[] ring once. */
 	for (i = 0; i < ARRAY_SIZE(st); i++) {
 		struct hcall_args args;
-		/* We remember where we were up to from last time.  This makes
+		/*
+		 * We remember where we were up to from last time.  This makes
 		 * sure that the hypercalls are done in the order the Guest
-		 * places them in the ring. */
+		 * places them in the ring.
+		 */
 		unsigned int n = cpu->next_hcall;
 
 		/* 0xFF means there's no call here (yet). */
 		if (st[n] == 0xFF)
 			break;
 
-		/* OK, we have hypercall.  Increment the "next_hcall" cursor,
-		 * and wrap back to 0 if we reach the end. */
+		/*
+		 * OK, we have hypercall.  Increment the "next_hcall" cursor,
+		 * and wrap back to 0 if we reach the end.
+		 */
 		if (++cpu->next_hcall == LHCALL_RING_SIZE)
 			cpu->next_hcall = 0;
 
-		/* Copy the hypercall arguments into a local copy of
-		 * the hcall_args struct. */
+		/*
+		 * Copy the hypercall arguments into a local copy of the
+		 * hcall_args struct.
+		 */
 		if (copy_from_user(&args, &cpu->lg->lguest_data->hcalls[n],
 				   sizeof(struct hcall_args))) {
 			kill_guest(cpu, "Fetching async hypercalls");
@@ -164,19 +185,25 @@ static void do_async_hcalls(struct lg_cpu *cpu)
 			break;
 		}
 
-		/* Stop doing hypercalls if they want to notify the Launcher:
-		 * it needs to service this first. */
+		/*
+		 * Stop doing hypercalls if they want to notify the Launcher:
+		 * it needs to service this first.
+		 */
 		if (cpu->pending_notify)
 			break;
 	}
 }
 
-/* Last of all, we look at what happens first of all.  The very first time the
- * Guest makes a hypercall, we end up here to set things up: */
+/*
+ * Last of all, we look at what happens first of all.  The very first time the
+ * Guest makes a hypercall, we end up here to set things up:
+ */
 static void initialize(struct lg_cpu *cpu)
 {
-	/* You can't do anything until you're initialized.  The Guest knows the
-	 * rules, so we're unforgiving here. */
+	/*
+	 * You can't do anything until you're initialized.  The Guest knows the
+	 * rules, so we're unforgiving here.
+	 */
 	if (cpu->hcall->arg0 != LHCALL_LGUEST_INIT) {
 		kill_guest(cpu, "hypercall %li before INIT", cpu->hcall->arg0);
 		return;
@@ -185,32 +212,44 @@ static void initialize(struct lg_cpu *cpu)
 	if (lguest_arch_init_hypercalls(cpu))
 		kill_guest(cpu, "bad guest page %p", cpu->lg->lguest_data);
 
-	/* The Guest tells us where we're not to deliver interrupts by putting
-	 * the range of addresses into "struct lguest_data". */
+	/*
+	 * The Guest tells us where we're not to deliver interrupts by putting
+	 * the range of addresses into "struct lguest_data".
+	 */
 	if (get_user(cpu->lg->noirq_start, &cpu->lg->lguest_data->noirq_start)
 	    || get_user(cpu->lg->noirq_end, &cpu->lg->lguest_data->noirq_end))
 		kill_guest(cpu, "bad guest page %p", cpu->lg->lguest_data);
 
-	/* We write the current time into the Guest's data page once so it can
-	 * set its clock. */
+	/*
+	 * We write the current time into the Guest's data page once so it can
+	 * set its clock.
+	 */
 	write_timestamp(cpu);
 
 	/* page_tables.c will also do some setup. */
 	page_table_guest_data_init(cpu);
 
-	/* This is the one case where the above accesses might have been the
+	/*
+	 * This is the one case where the above accesses might have been the
 	 * first write to a Guest page.  This may have caused a copy-on-write
 	 * fault, but the old page might be (read-only) in the Guest
-	 * pagetable. */
+	 * pagetable.
+	 */
 	guest_pagetable_clear_all(cpu);
 }
 /*:*/
 
-/*M:013 If a Guest reads from a page (so creates a mapping) that it has never
+/*M:013
+ * If a Guest reads from a page (so creates a mapping) that it has never
  * written to, and then the Launcher writes to it (ie. the output of a virtual
  * device), the Guest will still see the old page.  In practice, this never
  * happens: why would the Guest read a page which it has never written to?  But
- * a similar scenario might one day bite us, so it's worth mentioning. :*/
+ * a similar scenario might one day bite us, so it's worth mentioning.
+ *
+ * Note that if we used a shared anonymous mapping in the Launcher instead of
+ * mapping /dev/zero private, we wouldn't worry about cop-on-write.  And we
+ * need that to switch the Launcher to processes (away from threads) anyway.
+:*/
 
 /*H:100
  * Hypercalls
@@ -229,17 +268,22 @@ void do_hypercalls(struct lg_cpu *cpu)
 		return;
 	}
 
-	/* The Guest has initialized.
+	/*
+	 * The Guest has initialized.
 	 *
-	 * Look in the hypercall ring for the async hypercalls: */
+	 * Look in the hypercall ring for the async hypercalls:
+	 */
 	do_async_hcalls(cpu);
 
-	/* If we stopped reading the hypercall ring because the Guest did a
+	/*
+	 * If we stopped reading the hypercall ring because the Guest did a
 	 * NOTIFY to the Launcher, we want to return now.  Otherwise we do
-	 * the hypercall. */
+	 * the hypercall.
+	 */
 	if (!cpu->pending_notify) {
 		do_hcall(cpu, cpu->hcall);
-		/* Tricky point: we reset the hcall pointer to mark the
+		/*
+		 * Tricky point: we reset the hcall pointer to mark the
 		 * hypercall as "done".  We use the hcall pointer rather than
 		 * the trap number to indicate a hypercall is pending.
 		 * Normally it doesn't matter: the Guest will run again and
@@ -248,13 +292,16 @@ void do_hypercalls(struct lg_cpu *cpu)
 		 * However, if we are signalled or the Guest sends I/O to the
 		 * Launcher, the run_guest() loop will exit without running the
 		 * Guest.  When it comes back it would try to re-run the
-		 * hypercall.  Finding that bug sucked. */
+		 * hypercall.  Finding that bug sucked.
+		 */
 		cpu->hcall = NULL;
 	}
 }
 
-/* This routine supplies the Guest with time: it's used for wallclock time at
- * initial boot and as a rough time source if the TSC isn't available. */
+/*
+ * This routine supplies the Guest with time: it's used for wallclock time at
+ * initial boot and as a rough time source if the TSC isn't available.
+ */
 void write_timestamp(struct lg_cpu *cpu)
 {
 	struct timespec now;
diff --git a/drivers/lguest/interrupts_and_traps.c b/drivers/lguest/interrupts_and_traps.c
index 0e9067b0d50..18648180db0 100644
--- a/drivers/lguest/interrupts_and_traps.c
+++ b/drivers/lguest/interrupts_and_traps.c
@@ -1,4 +1,5 @@
-/*P:800 Interrupts (traps) are complicated enough to earn their own file.
+/*P:800
+ * Interrupts (traps) are complicated enough to earn their own file.
  * There are three classes of interrupts:
  *
  * 1) Real hardware interrupts which occur while we're running the Guest,
@@ -10,7 +11,8 @@
  * just like real hardware would deliver them.  Traps from the Guest can be set
  * up to go directly back into the Guest, but sometimes the Host wants to see
  * them first, so we also have a way of "reflecting" them into the Guest as if
- * they had been delivered to it directly. :*/
+ * they had been delivered to it directly.
+:*/
 #include <linux/uaccess.h>
 #include <linux/interrupt.h>
 #include <linux/module.h>
@@ -26,8 +28,10 @@ static unsigned long idt_address(u32 lo, u32 hi)
 	return (lo & 0x0000FFFF) | (hi & 0xFFFF0000);
 }
 
-/* The "type" of the interrupt handler is a 4 bit field: we only support a
- * couple of types. */
+/*
+ * The "type" of the interrupt handler is a 4 bit field: we only support a
+ * couple of types.
+ */
 static int idt_type(u32 lo, u32 hi)
 {
 	return (hi >> 8) & 0xF;
@@ -39,8 +43,10 @@ static bool idt_present(u32 lo, u32 hi)
 	return (hi & 0x8000);
 }
 
-/* We need a helper to "push" a value onto the Guest's stack, since that's a
- * big part of what delivering an interrupt does. */
+/*
+ * We need a helper to "push" a value onto the Guest's stack, since that's a
+ * big part of what delivering an interrupt does.
+ */
 static void push_guest_stack(struct lg_cpu *cpu, unsigned long *gstack, u32 val)
 {
 	/* Stack grows upwards: move stack then write value. */
@@ -48,7 +54,8 @@ static void push_guest_stack(struct lg_cpu *cpu, unsigned long *gstack, u32 val)
 	lgwrite(cpu, *gstack, u32, val);
 }
 
-/*H:210 The set_guest_interrupt() routine actually delivers the interrupt or
+/*H:210
+ * The set_guest_interrupt() routine actually delivers the interrupt or
  * trap.  The mechanics of delivering traps and interrupts to the Guest are the
  * same, except some traps have an "error code" which gets pushed onto the
  * stack as well: the caller tells us if this is one.
@@ -59,7 +66,8 @@ static void push_guest_stack(struct lg_cpu *cpu, unsigned long *gstack, u32 val)
  *
  * We set up the stack just like the CPU does for a real interrupt, so it's
  * identical for the Guest (and the standard "iret" instruction will undo
- * it). */
+ * it).
+ */
 static void set_guest_interrupt(struct lg_cpu *cpu, u32 lo, u32 hi,
 				bool has_err)
 {
@@ -67,20 +75,26 @@ static void set_guest_interrupt(struct lg_cpu *cpu, u32 lo, u32 hi,
 	u32 eflags, ss, irq_enable;
 	unsigned long virtstack;
 
-	/* There are two cases for interrupts: one where the Guest is already
+	/*
+	 * There are two cases for interrupts: one where the Guest is already
 	 * in the kernel, and a more complex one where the Guest is in
-	 * userspace.  We check the privilege level to find out. */
+	 * userspace.  We check the privilege level to find out.
+	 */
 	if ((cpu->regs->ss&0x3) != GUEST_PL) {
-		/* The Guest told us their kernel stack with the SET_STACK
-		 * hypercall: both the virtual address and the segment */
+		/*
+		 * The Guest told us their kernel stack with the SET_STACK
+		 * hypercall: both the virtual address and the segment.
+		 */
 		virtstack = cpu->esp1;
 		ss = cpu->ss1;
 
 		origstack = gstack = guest_pa(cpu, virtstack);
-		/* We push the old stack segment and pointer onto the new
+		/*
+		 * We push the old stack segment and pointer onto the new
 		 * stack: when the Guest does an "iret" back from the interrupt
 		 * handler the CPU will notice they're dropping privilege
-		 * levels and expect these here. */
+		 * levels and expect these here.
+		 */
 		push_guest_stack(cpu, &gstack, cpu->regs->ss);
 		push_guest_stack(cpu, &gstack, cpu->regs->esp);
 	} else {
@@ -91,18 +105,22 @@ static void set_guest_interrupt(struct lg_cpu *cpu, u32 lo, u32 hi,
 		origstack = gstack = guest_pa(cpu, virtstack);
 	}
 
-	/* Remember that we never let the Guest actually disable interrupts, so
+	/*
+	 * Remember that we never let the Guest actually disable interrupts, so
 	 * the "Interrupt Flag" bit is always set.  We copy that bit from the
 	 * Guest's "irq_enabled" field into the eflags word: we saw the Guest
-	 * copy it back in "lguest_iret". */
+	 * copy it back in "lguest_iret".
+	 */
 	eflags = cpu->regs->eflags;
 	if (get_user(irq_enable, &cpu->lg->lguest_data->irq_enabled) == 0
 	    && !(irq_enable & X86_EFLAGS_IF))
 		eflags &= ~X86_EFLAGS_IF;
 
-	/* An interrupt is expected to push three things on the stack: the old
+	/*
+	 * An interrupt is expected to push three things on the stack: the old
 	 * "eflags" word, the old code segment, and the old instruction
-	 * pointer. */
+	 * pointer.
+	 */
 	push_guest_stack(cpu, &gstack, eflags);
 	push_guest_stack(cpu, &gstack, cpu->regs->cs);
 	push_guest_stack(cpu, &gstack, cpu->regs->eip);
@@ -111,15 +129,19 @@ static void set_guest_interrupt(struct lg_cpu *cpu, u32 lo, u32 hi,
 	if (has_err)
 		push_guest_stack(cpu, &gstack, cpu->regs->errcode);
 
-	/* Now we've pushed all the old state, we change the stack, the code
-	 * segment and the address to execute. */
+	/*
+	 * Now we've pushed all the old state, we change the stack, the code
+	 * segment and the address to execute.
+	 */
 	cpu->regs->ss = ss;
 	cpu->regs->esp = virtstack + (gstack - origstack);
 	cpu->regs->cs = (__KERNEL_CS|GUEST_PL);
 	cpu->regs->eip = idt_address(lo, hi);
 
-	/* There are two kinds of interrupt handlers: 0xE is an "interrupt
-	 * gate" which expects interrupts to be disabled on entry. */
+	/*
+	 * There are two kinds of interrupt handlers: 0xE is an "interrupt
+	 * gate" which expects interrupts to be disabled on entry.
+	 */
 	if (idt_type(lo, hi) == 0xE)
 		if (put_user(0, &cpu->lg->lguest_data->irq_enabled))
 			kill_guest(cpu, "Disabling interrupts");
@@ -130,7 +152,8 @@ static void set_guest_interrupt(struct lg_cpu *cpu, u32 lo, u32 hi,
  *
  * interrupt_pending() returns the first pending interrupt which isn't blocked
  * by the Guest.  It is called before every entry to the Guest, and just before
- * we go to sleep when the Guest has halted itself. */
+ * we go to sleep when the Guest has halted itself.
+ */
 unsigned int interrupt_pending(struct lg_cpu *cpu, bool *more)
 {
 	unsigned int irq;
@@ -140,8 +163,10 @@ unsigned int interrupt_pending(struct lg_cpu *cpu, bool *more)
 	if (!cpu->lg->lguest_data)
 		return LGUEST_IRQS;
 
-	/* Take our "irqs_pending" array and remove any interrupts the Guest
-	 * wants blocked: the result ends up in "blk". */
+	/*
+	 * Take our "irqs_pending" array and remove any interrupts the Guest
+	 * wants blocked: the result ends up in "blk".
+	 */
 	if (copy_from_user(&blk, cpu->lg->lguest_data->blocked_interrupts,
 			   sizeof(blk)))
 		return LGUEST_IRQS;
@@ -154,16 +179,20 @@ unsigned int interrupt_pending(struct lg_cpu *cpu, bool *more)
 	return irq;
 }
 
-/* This actually diverts the Guest to running an interrupt handler, once an
- * interrupt has been identified by interrupt_pending(). */
+/*
+ * This actually diverts the Guest to running an interrupt handler, once an
+ * interrupt has been identified by interrupt_pending().
+ */
 void try_deliver_interrupt(struct lg_cpu *cpu, unsigned int irq, bool more)
 {
 	struct desc_struct *idt;
 
 	BUG_ON(irq >= LGUEST_IRQS);
 
-	/* They may be in the middle of an iret, where they asked us never to
-	 * deliver interrupts. */
+	/*
+	 * They may be in the middle of an iret, where they asked us never to
+	 * deliver interrupts.
+	 */
 	if (cpu->regs->eip >= cpu->lg->noirq_start &&
 	   (cpu->regs->eip < cpu->lg->noirq_end))
 		return;
@@ -187,29 +216,37 @@ void try_deliver_interrupt(struct lg_cpu *cpu, unsigned int irq, bool more)
 		}
 	}
 
-	/* Look at the IDT entry the Guest gave us for this interrupt.  The
+	/*
+	 * Look at the IDT entry the Guest gave us for this interrupt.  The
 	 * first 32 (FIRST_EXTERNAL_VECTOR) entries are for traps, so we skip
-	 * over them. */
+	 * over them.
+	 */
 	idt = &cpu->arch.idt[FIRST_EXTERNAL_VECTOR+irq];
 	/* If they don't have a handler (yet?), we just ignore it */
 	if (idt_present(idt->a, idt->b)) {
 		/* OK, mark it no longer pending and deliver it. */
 		clear_bit(irq, cpu->irqs_pending);
-		/* set_guest_interrupt() takes the interrupt descriptor and a
+		/*
+		 * set_guest_interrupt() takes the interrupt descriptor and a
 		 * flag to say whether this interrupt pushes an error code onto
-		 * the stack as well: virtual interrupts never do. */
+		 * the stack as well: virtual interrupts never do.
+		 */
 		set_guest_interrupt(cpu, idt->a, idt->b, false);
 	}
 
-	/* Every time we deliver an interrupt, we update the timestamp in the
+	/*
+	 * Every time we deliver an interrupt, we update the timestamp in the
 	 * Guest's lguest_data struct.  It would be better for the Guest if we
 	 * did this more often, but it can actually be quite slow: doing it
 	 * here is a compromise which means at least it gets updated every
-	 * timer interrupt. */
+	 * timer interrupt.
+	 */
 	write_timestamp(cpu);
 
-	/* If there are no other interrupts we want to deliver, clear
-	 * the pending flag. */
+	/*
+	 * If there are no other interrupts we want to deliver, clear
+	 * the pending flag.
+	 */
 	if (!more)
 		put_user(0, &cpu->lg->lguest_data->irq_pending);
 }
@@ -217,24 +254,29 @@ void try_deliver_interrupt(struct lg_cpu *cpu, unsigned int irq, bool more)
 /* And this is the routine when we want to set an interrupt for the Guest. */
 void set_interrupt(struct lg_cpu *cpu, unsigned int irq)
 {
-	/* Next time the Guest runs, the core code will see if it can deliver
-	 * this interrupt. */
+	/*
+	 * Next time the Guest runs, the core code will see if it can deliver
+	 * this interrupt.
+	 */
 	set_bit(irq, cpu->irqs_pending);
 
-	/* Make sure it sees it; it might be asleep (eg. halted), or
-	 * running the Guest right now, in which case kick_process()
-	 * will knock it out. */
+	/*
+	 * Make sure it sees it; it might be asleep (eg. halted), or running
+	 * the Guest right now, in which case kick_process() will knock it out.
+	 */
 	if (!wake_up_process(cpu->tsk))
 		kick_process(cpu->tsk);
 }
 /*:*/
 
-/* Linux uses trap 128 for system calls.  Plan9 uses 64, and Ron Minnich sent
+/*
+ * Linux uses trap 128 for system calls.  Plan9 uses 64, and Ron Minnich sent
  * me a patch, so we support that too.  It'd be a big step for lguest if half
  * the Plan 9 user base were to start using it.
  *
  * Actually now I think of it, it's possible that Ron *is* half the Plan 9
- * userbase.  Oh well. */
+ * userbase.  Oh well.
+ */
 static bool could_be_syscall(unsigned int num)
 {
 	/* Normal Linux SYSCALL_VECTOR or reserved vector? */
@@ -274,9 +316,11 @@ void free_interrupts(void)
 		clear_bit(syscall_vector, used_vectors);
 }
 
-/*H:220 Now we've got the routines to deliver interrupts, delivering traps like
+/*H:220
+ * Now we've got the routines to deliver interrupts, delivering traps like
  * page fault is easy.  The only trick is that Intel decided that some traps
- * should have error codes: */
+ * should have error codes:
+ */
 static bool has_err(unsigned int trap)
 {
 	return (trap == 8 || (trap >= 10 && trap <= 14) || trap == 17);
@@ -285,13 +329,17 @@ static bool has_err(unsigned int trap)
 /* deliver_trap() returns true if it could deliver the trap. */
 bool deliver_trap(struct lg_cpu *cpu, unsigned int num)
 {
-	/* Trap numbers are always 8 bit, but we set an impossible trap number
-	 * for traps inside the Switcher, so check that here. */
+	/*
+	 * Trap numbers are always 8 bit, but we set an impossible trap number
+	 * for traps inside the Switcher, so check that here.
+	 */
 	if (num >= ARRAY_SIZE(cpu->arch.idt))
 		return false;
 
-	/* Early on the Guest hasn't set the IDT entries (or maybe it put a
-	 * bogus one in): if we fail here, the Guest will be killed. */
+	/*
+	 * Early on the Guest hasn't set the IDT entries (or maybe it put a
+	 * bogus one in): if we fail here, the Guest will be killed.
+	 */
 	if (!idt_present(cpu->arch.idt[num].a, cpu->arch.idt[num].b))
 		return false;
 	set_guest_interrupt(cpu, cpu->arch.idt[num].a,
@@ -299,7 +347,8 @@ bool deliver_trap(struct lg_cpu *cpu, unsigned int num)
 	return true;
 }
 
-/*H:250 Here's the hard part: returning to the Host every time a trap happens
+/*H:250
+ * Here's the hard part: returning to the Host every time a trap happens
  * and then calling deliver_trap() and re-entering the Guest is slow.
  * Particularly because Guest userspace system calls are traps (usually trap
  * 128).
@@ -311,69 +360,87 @@ bool deliver_trap(struct lg_cpu *cpu, unsigned int num)
  * the other hypervisors would beat it up at lunchtime.
  *
  * This routine indicates if a particular trap number could be delivered
- * directly. */
+ * directly.
+ */
 static bool direct_trap(unsigned int num)
 {
-	/* Hardware interrupts don't go to the Guest at all (except system
-	 * call). */
+	/*
+	 * Hardware interrupts don't go to the Guest at all (except system
+	 * call).
+	 */
 	if (num >= FIRST_EXTERNAL_VECTOR && !could_be_syscall(num))
 		return false;
 
-	/* The Host needs to see page faults (for shadow paging and to save the
+	/*
+	 * The Host needs to see page faults (for shadow paging and to save the
 	 * fault address), general protection faults (in/out emulation) and
 	 * device not available (TS handling), invalid opcode fault (kvm hcall),
-	 * and of course, the hypercall trap. */
+	 * and of course, the hypercall trap.
+	 */
 	return num != 14 && num != 13 && num != 7 &&
 			num != 6 && num != LGUEST_TRAP_ENTRY;
 }
 /*:*/
 
-/*M:005 The Guest has the ability to turn its interrupt gates into trap gates,
+/*M:005
+ * The Guest has the ability to turn its interrupt gates into trap gates,
  * if it is careful.  The Host will let trap gates can go directly to the
  * Guest, but the Guest needs the interrupts atomically disabled for an
  * interrupt gate.  It can do this by pointing the trap gate at instructions
- * within noirq_start and noirq_end, where it can safely disable interrupts. */
+ * within noirq_start and noirq_end, where it can safely disable interrupts.
+ */
 
-/*M:006 The Guests do not use the sysenter (fast system call) instruction,
+/*M:006
+ * The Guests do not use the sysenter (fast system call) instruction,
  * because it's hardcoded to enter privilege level 0 and so can't go direct.
  * It's about twice as fast as the older "int 0x80" system call, so it might
  * still be worthwhile to handle it in the Switcher and lcall down to the
  * Guest.  The sysenter semantics are hairy tho: search for that keyword in
- * entry.S :*/
+ * entry.S
+:*/
 
-/*H:260 When we make traps go directly into the Guest, we need to make sure
+/*H:260
+ * When we make traps go directly into the Guest, we need to make sure
  * the kernel stack is valid (ie. mapped in the page tables).  Otherwise, the
  * CPU trying to deliver the trap will fault while trying to push the interrupt
  * words on the stack: this is called a double fault, and it forces us to kill
  * the Guest.
  *
- * Which is deeply unfair, because (literally!) it wasn't the Guests' fault. */
+ * Which is deeply unfair, because (literally!) it wasn't the Guests' fault.
+ */
 void pin_stack_pages(struct lg_cpu *cpu)
 {
 	unsigned int i;
 
-	/* Depending on the CONFIG_4KSTACKS option, the Guest can have one or
-	 * two pages of stack space. */
+	/*
+	 * Depending on the CONFIG_4KSTACKS option, the Guest can have one or
+	 * two pages of stack space.
+	 */
 	for (i = 0; i < cpu->lg->stack_pages; i++)
-		/* The stack grows *upwards*, so the address we're given is the
+		/*
+		 * The stack grows *upwards*, so the address we're given is the
 		 * start of the page after the kernel stack.  Subtract one to
 		 * get back onto the first stack page, and keep subtracting to
-		 * get to the rest of the stack pages. */
+		 * get to the rest of the stack pages.
+		 */
 		pin_page(cpu, cpu->esp1 - 1 - i * PAGE_SIZE);
 }
 
-/* Direct traps also mean that we need to know whenever the Guest wants to use
+/*
+ * Direct traps also mean that we need to know whenever the Guest wants to use
  * a different kernel stack, so we can change the IDT entries to use that
  * stack.  The IDT entries expect a virtual address, so unlike most addresses
  * the Guest gives us, the "esp" (stack pointer) value here is virtual, not
  * physical.
  *
  * In Linux each process has its own kernel stack, so this happens a lot: we
- * change stacks on each context switch. */
+ * change stacks on each context switch.
+ */
 void guest_set_stack(struct lg_cpu *cpu, u32 seg, u32 esp, unsigned int pages)
 {
-	/* You are not allowed have a stack segment with privilege level 0: bad
-	 * Guest! */
+	/*
+	 * You're not allowed a stack segment with privilege level 0: bad Guest!
+	 */
 	if ((seg & 0x3) != GUEST_PL)
 		kill_guest(cpu, "bad stack segment %i", seg);
 	/* We only expect one or two stack pages. */
@@ -387,11 +454,15 @@ void guest_set_stack(struct lg_cpu *cpu, u32 seg, u32 esp, unsigned int pages)
 	pin_stack_pages(cpu);
 }
 
-/* All this reference to mapping stacks leads us neatly into the other complex
- * part of the Host: page table handling. */
+/*
+ * All this reference to mapping stacks leads us neatly into the other complex
+ * part of the Host: page table handling.
+ */
 
-/*H:235 This is the routine which actually checks the Guest's IDT entry and
- * transfers it into the entry in "struct lguest": */
+/*H:235
+ * This is the routine which actually checks the Guest's IDT entry and
+ * transfers it into the entry in "struct lguest":
+ */
 static void set_trap(struct lg_cpu *cpu, struct desc_struct *trap,
 		     unsigned int num, u32 lo, u32 hi)
 {
@@ -407,30 +478,38 @@ static void set_trap(struct lg_cpu *cpu, struct desc_struct *trap,
 	if (type != 0xE && type != 0xF)
 		kill_guest(cpu, "bad IDT type %i", type);
 
-	/* We only copy the handler address, present bit, privilege level and
+	/*
+	 * We only copy the handler address, present bit, privilege level and
 	 * type.  The privilege level controls where the trap can be triggered
 	 * manually with an "int" instruction.  This is usually GUEST_PL,
-	 * except for system calls which userspace can use. */
+	 * except for system calls which userspace can use.
+	 */
 	trap->a = ((__KERNEL_CS|GUEST_PL)<<16) | (lo&0x0000FFFF);
 	trap->b = (hi&0xFFFFEF00);
 }
 
-/*H:230 While we're here, dealing with delivering traps and interrupts to the
+/*H:230
+ * While we're here, dealing with delivering traps and interrupts to the
  * Guest, we might as well complete the picture: how the Guest tells us where
  * it wants them to go.  This would be simple, except making traps fast
  * requires some tricks.
  *
  * We saw the Guest setting Interrupt Descriptor Table (IDT) entries with the
- * LHCALL_LOAD_IDT_ENTRY hypercall before: that comes here. */
+ * LHCALL_LOAD_IDT_ENTRY hypercall before: that comes here.
+ */
 void load_guest_idt_entry(struct lg_cpu *cpu, unsigned int num, u32 lo, u32 hi)
 {
-	/* Guest never handles: NMI, doublefault, spurious interrupt or
-	 * hypercall.  We ignore when it tries to set them. */
+	/*
+	 * Guest never handles: NMI, doublefault, spurious interrupt or
+	 * hypercall.  We ignore when it tries to set them.
+	 */
 	if (num == 2 || num == 8 || num == 15 || num == LGUEST_TRAP_ENTRY)
 		return;
 
-	/* Mark the IDT as changed: next time the Guest runs we'll know we have
-	 * to copy this again. */
+	/*
+	 * Mark the IDT as changed: next time the Guest runs we'll know we have
+	 * to copy this again.
+	 */
 	cpu->changed |= CHANGED_IDT;
 
 	/* Check that the Guest doesn't try to step outside the bounds. */
@@ -440,9 +519,11 @@ void load_guest_idt_entry(struct lg_cpu *cpu, unsigned int num, u32 lo, u32 hi)
 		set_trap(cpu, &cpu->arch.idt[num], num, lo, hi);
 }
 
-/* The default entry for each interrupt points into the Switcher routines which
+/*
+ * The default entry for each interrupt points into the Switcher routines which
  * simply return to the Host.  The run_guest() loop will then call
- * deliver_trap() to bounce it back into the Guest. */
+ * deliver_trap() to bounce it back into the Guest.
+ */
 static void default_idt_entry(struct desc_struct *idt,
 			      int trap,
 			      const unsigned long handler,
@@ -451,13 +532,17 @@ static void default_idt_entry(struct desc_struct *idt,
 	/* A present interrupt gate. */
 	u32 flags = 0x8e00;
 
-	/* Set the privilege level on the entry for the hypercall: this allows
-	 * the Guest to use the "int" instruction to trigger it. */
+	/*
+	 * Set the privilege level on the entry for the hypercall: this allows
+	 * the Guest to use the "int" instruction to trigger it.
+	 */
 	if (trap == LGUEST_TRAP_ENTRY)
 		flags |= (GUEST_PL << 13);
 	else if (base)
-		/* Copy priv. level from what Guest asked for.  This allows
-		 * debug (int 3) traps from Guest userspace, for example. */
+		/*
+		 * Copy privilege level from what Guest asked for.  This allows
+		 * debug (int 3) traps from Guest userspace, for example.
+		 */
 		flags |= (base->b & 0x6000);
 
 	/* Now pack it into the IDT entry in its weird format. */
@@ -475,16 +560,20 @@ void setup_default_idt_entries(struct lguest_ro_state *state,
 		default_idt_entry(&state->guest_idt[i], i, def[i], NULL);
 }
 
-/*H:240 We don't use the IDT entries in the "struct lguest" directly, instead
+/*H:240
+ * We don't use the IDT entries in the "struct lguest" directly, instead
  * we copy them into the IDT which we've set up for Guests on this CPU, just
- * before we run the Guest.  This routine does that copy. */
+ * before we run the Guest.  This routine does that copy.
+ */
 void copy_traps(const struct lg_cpu *cpu, struct desc_struct *idt,
 		const unsigned long *def)
 {
 	unsigned int i;
 
-	/* We can simply copy the direct traps, otherwise we use the default
-	 * ones in the Switcher: they will return to the Host. */
+	/*
+	 * We can simply copy the direct traps, otherwise we use the default
+	 * ones in the Switcher: they will return to the Host.
+	 */
 	for (i = 0; i < ARRAY_SIZE(cpu->arch.idt); i++) {
 		const struct desc_struct *gidt = &cpu->arch.idt[i];
 
@@ -492,14 +581,16 @@ void copy_traps(const struct lg_cpu *cpu, struct desc_struct *idt,
 		if (!direct_trap(i))
 			continue;
 
-		/* Only trap gates (type 15) can go direct to the Guest.
+		/*
+		 * Only trap gates (type 15) can go direct to the Guest.
 		 * Interrupt gates (type 14) disable interrupts as they are
 		 * entered, which we never let the Guest do.  Not present
 		 * entries (type 0x0) also can't go direct, of course.
 		 *
 		 * If it can't go direct, we still need to copy the priv. level:
 		 * they might want to give userspace access to a software
-		 * interrupt. */
+		 * interrupt.
+		 */
 		if (idt_type(gidt->a, gidt->b) == 0xF)
 			idt[i] = *gidt;
 		else
@@ -518,7 +609,8 @@ void copy_traps(const struct lg_cpu *cpu, struct desc_struct *idt,
  * the next timer interrupt (in nanoseconds).  We use the high-resolution timer
  * infrastructure to set a callback at that time.
  *
- * 0 means "turn off the clock". */
+ * 0 means "turn off the clock".
+ */
 void guest_set_clockevent(struct lg_cpu *cpu, unsigned long delta)
 {
 	ktime_t expires;
@@ -529,9 +621,11 @@ void guest_set_clockevent(struct lg_cpu *cpu, unsigned long delta)
 		return;
 	}
 
-	/* We use wallclock time here, so the Guest might not be running for
+	/*
+	 * We use wallclock time here, so the Guest might not be running for
 	 * all the time between now and the timer interrupt it asked for.  This
-	 * is almost always the right thing to do. */
+	 * is almost always the right thing to do.
+	 */
 	expires = ktime_add_ns(ktime_get_real(), delta);
 	hrtimer_start(&cpu->hrt, expires, HRTIMER_MODE_ABS);
 }
diff --git a/drivers/lguest/lg.h b/drivers/lguest/lg.h
index 01c59192379..bc28745d05a 100644
--- a/drivers/lguest/lg.h
+++ b/drivers/lguest/lg.h
@@ -16,15 +16,13 @@
 void free_pagetables(void);
 int init_pagetables(struct page **switcher_page, unsigned int pages);
 
-struct pgdir
-{
+struct pgdir {
 	unsigned long gpgdir;
 	pgd_t *pgdir;
 };
 
 /* We have two pages shared with guests, per cpu.  */
-struct lguest_pages
-{
+struct lguest_pages {
 	/* This is the stack page mapped rw in guest */
 	char spare[PAGE_SIZE - sizeof(struct lguest_regs)];
 	struct lguest_regs regs;
@@ -54,13 +52,13 @@ struct lg_cpu {
 
 	unsigned long pending_notify; /* pfn from LHCALL_NOTIFY */
 
-	/* At end of a page shared mapped over lguest_pages in guest.  */
+	/* At end of a page shared mapped over lguest_pages in guest. */
 	unsigned long regs_page;
 	struct lguest_regs *regs;
 
 	struct lguest_pages *last_pages;
 
-	int cpu_pgd; /* which pgd this cpu is currently using */
+	int cpu_pgd; /* Which pgd this cpu is currently using */
 
 	/* If a hypercall was asked for, this points to the arguments. */
 	struct hcall_args *hcall;
@@ -89,15 +87,17 @@ struct lg_eventfd_map {
 };
 
 /* The private info the thread maintains about the guest. */
-struct lguest
-{
+struct lguest {
 	struct lguest_data __user *lguest_data;
 	struct lg_cpu cpus[NR_CPUS];
 	unsigned int nr_cpus;
 
 	u32 pfn_limit;
-	/* This provides the offset to the base of guest-physical
-	 * memory in the Launcher. */
+
+	/*
+	 * This provides the offset to the base of guest-physical memory in the
+	 * Launcher.
+	 */
 	void __user *mem_base;
 	unsigned long kernel_address;
 
@@ -122,11 +122,13 @@ bool lguest_address_ok(const struct lguest *lg,
 void __lgread(struct lg_cpu *, void *, unsigned long, unsigned);
 void __lgwrite(struct lg_cpu *, unsigned long, const void *, unsigned);
 
-/*H:035 Using memory-copy operations like that is usually inconvient, so we
+/*H:035
+ * Using memory-copy operations like that is usually inconvient, so we
  * have the following helper macros which read and write a specific type (often
  * an unsigned long).
  *
- * This reads into a variable of the given type then returns that. */
+ * This reads into a variable of the given type then returns that.
+ */
 #define lgread(cpu, addr, type)						\
 	({ type _v; __lgread((cpu), &_v, (addr), sizeof(_v)); _v; })
 
@@ -140,9 +142,11 @@ void __lgwrite(struct lg_cpu *, unsigned long, const void *, unsigned);
 
 int run_guest(struct lg_cpu *cpu, unsigned long __user *user);
 
-/* Helper macros to obtain the first 12 or the last 20 bits, this is only the
+/*
+ * Helper macros to obtain the first 12 or the last 20 bits, this is only the
  * first step in the migration to the kernel types.  pte_pfn is already defined
- * in the kernel. */
+ * in the kernel.
+ */
 #define pgd_flags(x)	(pgd_val(x) & ~PAGE_MASK)
 #define pgd_pfn(x)	(pgd_val(x) >> PAGE_SHIFT)
 #define pmd_flags(x)    (pmd_val(x) & ~PAGE_MASK)
diff --git a/drivers/lguest/lguest_device.c b/drivers/lguest/lguest_device.c
index e082cdac88b..b6200bc39b5 100644
--- a/drivers/lguest/lguest_device.c
+++ b/drivers/lguest/lguest_device.c
@@ -1,10 +1,12 @@
-/*P:050 Lguest guests use a very simple method to describe devices.  It's a
+/*P:050
+ * Lguest guests use a very simple method to describe devices.  It's a
  * series of device descriptors contained just above the top of normal Guest
  * memory.
  *
  * We use the standard "virtio" device infrastructure, which provides us with a
  * console, a network and a block driver.  Each one expects some configuration
- * information and a "virtqueue" or two to send and receive data. :*/
+ * information and a "virtqueue" or two to send and receive data.
+:*/
 #include <linux/init.h>
 #include <linux/bootmem.h>
 #include <linux/lguest_launcher.h>
@@ -20,8 +22,10 @@
 /* The pointer to our (page) of device descriptions. */
 static void *lguest_devices;
 
-/* For Guests, device memory can be used as normal memory, so we cast away the
- * __iomem to quieten sparse. */
+/*
+ * For Guests, device memory can be used as normal memory, so we cast away the
+ * __iomem to quieten sparse.
+ */
 static inline void *lguest_map(unsigned long phys_addr, unsigned long pages)
 {
 	return (__force void *)ioremap_cache(phys_addr, PAGE_SIZE*pages);
@@ -32,8 +36,10 @@ static inline void lguest_unmap(void *addr)
 	iounmap((__force void __iomem *)addr);
 }
 
-/*D:100 Each lguest device is just a virtio device plus a pointer to its entry
- * in the lguest_devices page. */
+/*D:100
+ * Each lguest device is just a virtio device plus a pointer to its entry
+ * in the lguest_devices page.
+ */
 struct lguest_device {
 	struct virtio_device vdev;
 
@@ -41,9 +47,11 @@ struct lguest_device {
 	struct lguest_device_desc *desc;
 };
 
-/* Since the virtio infrastructure hands us a pointer to the virtio_device all
+/*
+ * Since the virtio infrastructure hands us a pointer to the virtio_device all
  * the time, it helps to have a curt macro to get a pointer to the struct
- * lguest_device it's enclosed in.  */
+ * lguest_device it's enclosed in.
+ */
 #define to_lgdev(vd) container_of(vd, struct lguest_device, vdev)
 
 /*D:130
@@ -55,7 +63,8 @@ struct lguest_device {
  * the driver will look at them during setup.
  *
  * A convenient routine to return the device's virtqueue config array:
- * immediately after the descriptor. */
+ * immediately after the descriptor.
+ */
 static struct lguest_vqconfig *lg_vq(const struct lguest_device_desc *desc)
 {
 	return (void *)(desc + 1);
@@ -98,10 +107,12 @@ static u32 lg_get_features(struct virtio_device *vdev)
 	return features;
 }
 
-/* The virtio core takes the features the Host offers, and copies the
- * ones supported by the driver into the vdev->features array.  Once
- * that's all sorted out, this routine is called so we can tell the
- * Host which features we understand and accept. */
+/*
+ * The virtio core takes the features the Host offers, and copies the ones
+ * supported by the driver into the vdev->features array.  Once that's all
+ * sorted out, this routine is called so we can tell the Host which features we
+ * understand and accept.
+ */
 static void lg_finalize_features(struct virtio_device *vdev)
 {
 	unsigned int i, bits;
@@ -112,10 +123,11 @@ static void lg_finalize_features(struct virtio_device *vdev)
 	/* Give virtio_ring a chance to accept features. */
 	vring_transport_features(vdev);
 
-	/* The vdev->feature array is a Linux bitmask: this isn't the
-	 * same as a the simple array of bits used by lguest devices
-	 * for features.  So we do this slow, manual conversion which is
-	 * completely general. */
+	/*
+	 * The vdev->feature array is a Linux bitmask: this isn't the same as a
+	 * the simple array of bits used by lguest devices for features.  So we
+	 * do this slow, manual conversion which is completely general.
+	 */
 	memset(out_features, 0, desc->feature_len);
 	bits = min_t(unsigned, desc->feature_len, sizeof(vdev->features)) * 8;
 	for (i = 0; i < bits; i++) {
@@ -146,15 +158,19 @@ static void lg_set(struct virtio_device *vdev, unsigned int offset,
 	memcpy(lg_config(desc) + offset, buf, len);
 }
 
-/* The operations to get and set the status word just access the status field
- * of the device descriptor. */
+/*
+ * The operations to get and set the status word just access the status field
+ * of the device descriptor.
+ */
 static u8 lg_get_status(struct virtio_device *vdev)
 {
 	return to_lgdev(vdev)->desc->status;
 }
 
-/* To notify on status updates, we (ab)use the NOTIFY hypercall, with the
- * descriptor address of the device.  A zero status means "reset". */
+/*
+ * To notify on status updates, we (ab)use the NOTIFY hypercall, with the
+ * descriptor address of the device.  A zero status means "reset".
+ */
 static void set_status(struct virtio_device *vdev, u8 status)
 {
 	unsigned long offset = (void *)to_lgdev(vdev)->desc - lguest_devices;
@@ -191,8 +207,7 @@ static void lg_reset(struct virtio_device *vdev)
  */
 
 /*D:140 This is the information we remember about each virtqueue. */
-struct lguest_vq_info
-{
+struct lguest_vq_info {
 	/* A copy of the information contained in the device config. */
 	struct lguest_vqconfig config;
 
@@ -200,13 +215,17 @@ struct lguest_vq_info
 	void *pages;
 };
 
-/* When the virtio_ring code wants to prod the Host, it calls us here and we
+/*
+ * When the virtio_ring code wants to prod the Host, it calls us here and we
  * make a hypercall.  We hand the physical address of the virtqueue so the Host
- * knows which virtqueue we're talking about. */
+ * knows which virtqueue we're talking about.
+ */
 static void lg_notify(struct virtqueue *vq)
 {
-	/* We store our virtqueue information in the "priv" pointer of the
-	 * virtqueue structure. */
+	/*
+	 * We store our virtqueue information in the "priv" pointer of the
+	 * virtqueue structure.
+	 */
 	struct lguest_vq_info *lvq = vq->priv;
 
 	kvm_hypercall1(LHCALL_NOTIFY, lvq->config.pfn << PAGE_SHIFT);
@@ -215,7 +234,8 @@ static void lg_notify(struct virtqueue *vq)
 /* An extern declaration inside a C file is bad form.  Don't do it. */
 extern void lguest_setup_irq(unsigned int irq);
 
-/* This routine finds the first virtqueue described in the configuration of
+/*
+ * This routine finds the Nth virtqueue described in the configuration of
  * this device and sets it up.
  *
  * This is kind of an ugly duckling.  It'd be nicer to have a standard
@@ -223,9 +243,7 @@ extern void lguest_setup_irq(unsigned int irq);
  * everyone wants to do it differently.  The KVM coders want the Guest to
  * allocate its own pages and tell the Host where they are, but for lguest it's
  * simpler for the Host to simply tell us where the pages are.
- *
- * So we provide drivers with a "find the Nth virtqueue and set it up"
- * function. */
+ */
 static struct virtqueue *lg_find_vq(struct virtio_device *vdev,
 				    unsigned index,
 				    void (*callback)(struct virtqueue *vq),
@@ -244,9 +262,11 @@ static struct virtqueue *lg_find_vq(struct virtio_device *vdev,
 	if (!lvq)
 		return ERR_PTR(-ENOMEM);
 
-	/* Make a copy of the "struct lguest_vqconfig" entry, which sits after
+	/*
+	 * Make a copy of the "struct lguest_vqconfig" entry, which sits after
 	 * the descriptor.  We need a copy because the config space might not
-	 * be aligned correctly. */
+	 * be aligned correctly.
+	 */
 	memcpy(&lvq->config, lg_vq(ldev->desc)+index, sizeof(lvq->config));
 
 	printk("Mapping virtqueue %i addr %lx\n", index,
@@ -261,8 +281,10 @@ static struct virtqueue *lg_find_vq(struct virtio_device *vdev,
 		goto free_lvq;
 	}
 
-	/* OK, tell virtio_ring.c to set up a virtqueue now we know its size
-	 * and we've got a pointer to its pages. */
+	/*
+	 * OK, tell virtio_ring.c to set up a virtqueue now we know its size
+	 * and we've got a pointer to its pages.
+	 */
 	vq = vring_new_virtqueue(lvq->config.num, LGUEST_VRING_ALIGN,
 				 vdev, lvq->pages, lg_notify, callback, name);
 	if (!vq) {
@@ -273,18 +295,23 @@ static struct virtqueue *lg_find_vq(struct virtio_device *vdev,
 	/* Make sure the interrupt is allocated. */
 	lguest_setup_irq(lvq->config.irq);
 
-	/* Tell the interrupt for this virtqueue to go to the virtio_ring
-	 * interrupt handler. */
-	/* FIXME: We used to have a flag for the Host to tell us we could use
+	/*
+	 * Tell the interrupt for this virtqueue to go to the virtio_ring
+	 * interrupt handler.
+	 *
+	 * FIXME: We used to have a flag for the Host to tell us we could use
 	 * the interrupt as a source of randomness: it'd be nice to have that
-	 * back.. */
+	 * back.
+	 */
 	err = request_irq(lvq->config.irq, vring_interrupt, IRQF_SHARED,
 			  dev_name(&vdev->dev), vq);
 	if (err)
 		goto destroy_vring;
 
-	/* Last of all we hook up our 'struct lguest_vq_info" to the
-	 * virtqueue's priv pointer. */
+	/*
+	 * Last of all we hook up our 'struct lguest_vq_info" to the
+	 * virtqueue's priv pointer.
+	 */
 	vq->priv = lvq;
 	return vq;
 
@@ -358,11 +385,14 @@ static struct virtio_config_ops lguest_config_ops = {
 	.del_vqs = lg_del_vqs,
 };
 
-/* The root device for the lguest virtio devices.  This makes them appear as
- * /sys/devices/lguest/0,1,2 not /sys/devices/0,1,2. */
+/*
+ * The root device for the lguest virtio devices.  This makes them appear as
+ * /sys/devices/lguest/0,1,2 not /sys/devices/0,1,2.
+ */
 static struct device *lguest_root;
 
-/*D:120 This is the core of the lguest bus: actually adding a new device.
+/*D:120
+ * This is the core of the lguest bus: actually adding a new device.
  * It's a separate function because it's neater that way, and because an
  * earlier version of the code supported hotplug and unplug.  They were removed
  * early on because they were never used.
@@ -371,14 +401,14 @@ static struct device *lguest_root;
  *
  * It's worth reading this carefully: we start with a pointer to the new device
  * descriptor in the "lguest_devices" page, and the offset into the device
- * descriptor page so we can uniquely identify it if things go badly wrong. */
+ * descriptor page so we can uniquely identify it if things go badly wrong.
+ */
 static void add_lguest_device(struct lguest_device_desc *d,
 			      unsigned int offset)
 {
 	struct lguest_device *ldev;
 
-	/* Start with zeroed memory; Linux's device layer seems to count on
-	 * it. */
+	/* Start with zeroed memory; Linux's device layer counts on it. */
 	ldev = kzalloc(sizeof(*ldev), GFP_KERNEL);
 	if (!ldev) {
 		printk(KERN_EMERG "Cannot allocate lguest dev %u type %u\n",
@@ -388,17 +418,25 @@ static void add_lguest_device(struct lguest_device_desc *d,
 
 	/* This devices' parent is the lguest/ dir. */
 	ldev->vdev.dev.parent = lguest_root;
-	/* We have a unique device index thanks to the dev_index counter. */
+	/*
+	 * The device type comes straight from the descriptor.  There's also a
+	 * device vendor field in the virtio_device struct, which we leave as
+	 * 0.
+	 */
 	ldev->vdev.id.device = d->type;
-	/* We have a simple set of routines for querying the device's
-	 * configuration information and setting its status. */
+	/*
+	 * We have a simple set of routines for querying the device's
+	 * configuration information and setting its status.
+	 */
 	ldev->vdev.config = &lguest_config_ops;
 	/* And we remember the device's descriptor for lguest_config_ops. */
 	ldev->desc = d;
 
-	/* register_virtio_device() sets up the generic fields for the struct
+	/*
+	 * register_virtio_device() sets up the generic fields for the struct
 	 * virtio_device and calls device_register().  This makes the bus
-	 * infrastructure look for a matching driver. */
+	 * infrastructure look for a matching driver.
+	 */
 	if (register_virtio_device(&ldev->vdev) != 0) {
 		printk(KERN_ERR "Failed to register lguest dev %u type %u\n",
 		       offset, d->type);
@@ -406,8 +444,10 @@ static void add_lguest_device(struct lguest_device_desc *d,
 	}
 }
 
-/*D:110 scan_devices() simply iterates through the device page.  The type 0 is
- * reserved to mean "end of devices". */
+/*D:110
+ * scan_devices() simply iterates through the device page.  The type 0 is
+ * reserved to mean "end of devices".
+ */
 static void scan_devices(void)
 {
 	unsigned int i;
@@ -426,7 +466,8 @@ static void scan_devices(void)
 	}
 }
 
-/*D:105 Fairly early in boot, lguest_devices_init() is called to set up the
+/*D:105
+ * Fairly early in boot, lguest_devices_init() is called to set up the
  * lguest device infrastructure.  We check that we are a Guest by checking
  * pv_info.name: there are other ways of checking, but this seems most
  * obvious to me.
@@ -437,7 +478,8 @@ static void scan_devices(void)
  * correct sysfs incantation).
  *
  * Finally we call scan_devices() which adds all the devices found in the
- * lguest_devices page. */
+ * lguest_devices page.
+ */
 static int __init lguest_devices_init(void)
 {
 	if (strcmp(pv_info.name, "lguest") != 0)
@@ -456,11 +498,13 @@ static int __init lguest_devices_init(void)
 /* We do this after core stuff, but before the drivers. */
 postcore_initcall(lguest_devices_init);
 
-/*D:150 At this point in the journey we used to now wade through the lguest
+/*D:150
+ * At this point in the journey we used to now wade through the lguest
  * devices themselves: net, block and console.  Since they're all now virtio
  * devices rather than lguest-specific, I've decided to ignore them.  Mostly,
  * they're kind of boring.  But this does mean you'll never experience the
  * thrill of reading the forbidden love scene buried deep in the block driver.
  *
  * "make Launcher" beckons, where we answer questions like "Where do Guests
- * come from?", and "What do you do when someone asks for optimization?". */
+ * come from?", and "What do you do when someone asks for optimization?".
+ */
diff --git a/drivers/lguest/lguest_user.c b/drivers/lguest/lguest_user.c
index 9f9a2953b38..b4d3f7ca554 100644
--- a/drivers/lguest/lguest_user.c
+++ b/drivers/lguest/lguest_user.c
@@ -1,8 +1,9 @@
 /*P:200 This contains all the /dev/lguest code, whereby the userspace launcher
  * controls and communicates with the Guest.  For example, the first write will
- * tell us the Guest's memory layout, pagetable, entry point and kernel address
- * offset.  A read will run the Guest until something happens, such as a signal
- * or the Guest doing a NOTIFY out to the Launcher. :*/
+ * tell us the Guest's memory layout and entry point.  A read will run the
+ * Guest until something happens, such as a signal or the Guest doing a NOTIFY
+ * out to the Launcher.
+:*/
 #include <linux/uaccess.h>
 #include <linux/miscdevice.h>
 #include <linux/fs.h>
@@ -11,14 +12,41 @@
 #include <linux/file.h>
 #include "lg.h"
 
+/*L:056
+ * Before we move on, let's jump ahead and look at what the kernel does when
+ * it needs to look up the eventfds.  That will complete our picture of how we
+ * use RCU.
+ *
+ * The notification value is in cpu->pending_notify: we return true if it went
+ * to an eventfd.
+ */
 bool send_notify_to_eventfd(struct lg_cpu *cpu)
 {
 	unsigned int i;
 	struct lg_eventfd_map *map;
 
-	/* lg->eventfds is RCU-protected */
+	/*
+	 * This "rcu_read_lock()" helps track when someone is still looking at
+	 * the (RCU-using) eventfds array.  It's not actually a lock at all;
+	 * indeed it's a noop in many configurations.  (You didn't expect me to
+	 * explain all the RCU secrets here, did you?)
+	 */
 	rcu_read_lock();
+	/*
+	 * rcu_dereference is the counter-side of rcu_assign_pointer(); it
+	 * makes sure we don't access the memory pointed to by
+	 * cpu->lg->eventfds before cpu->lg->eventfds is set.  Sounds crazy,
+	 * but Alpha allows this!  Paul McKenney points out that a really
+	 * aggressive compiler could have the same effect:
+	 *   http://lists.ozlabs.org/pipermail/lguest/2009-July/001560.html
+	 *
+	 * So play safe, use rcu_dereference to get the rcu-protected pointer:
+	 */
 	map = rcu_dereference(cpu->lg->eventfds);
+	/*
+	 * Simple array search: even if they add an eventfd while we do this,
+	 * we'll continue to use the old array and just won't see the new one.
+	 */
 	for (i = 0; i < map->num; i++) {
 		if (map->map[i].addr == cpu->pending_notify) {
 			eventfd_signal(map->map[i].event, 1);
@@ -26,19 +54,50 @@ bool send_notify_to_eventfd(struct lg_cpu *cpu)
 			break;
 		}
 	}
+	/* We're done with the rcu-protected variable cpu->lg->eventfds. */
 	rcu_read_unlock();
+
+	/* If we cleared the notification, it's because we found a match. */
 	return cpu->pending_notify == 0;
 }
 
+/*L:055
+ * One of the more tricksy tricks in the Linux Kernel is a technique called
+ * Read Copy Update.  Since one point of lguest is to teach lguest journeyers
+ * about kernel coding, I use it here.  (In case you're curious, other purposes
+ * include learning about virtualization and instilling a deep appreciation for
+ * simplicity and puppies).
+ *
+ * We keep a simple array which maps LHCALL_NOTIFY values to eventfds, but we
+ * add new eventfds without ever blocking readers from accessing the array.
+ * The current Launcher only does this during boot, so that never happens.  But
+ * Read Copy Update is cool, and adding a lock risks damaging even more puppies
+ * than this code does.
+ *
+ * We allocate a brand new one-larger array, copy the old one and add our new
+ * element.  Then we make the lg eventfd pointer point to the new array.
+ * That's the easy part: now we need to free the old one, but we need to make
+ * sure no slow CPU somewhere is still looking at it.  That's what
+ * synchronize_rcu does for us: waits until every CPU has indicated that it has
+ * moved on to know it's no longer using the old one.
+ *
+ * If that's unclear, see http://en.wikipedia.org/wiki/Read-copy-update.
+ */
 static int add_eventfd(struct lguest *lg, unsigned long addr, int fd)
 {
 	struct lg_eventfd_map *new, *old = lg->eventfds;
 
+	/*
+	 * We don't allow notifications on value 0 anyway (pending_notify of
+	 * 0 means "nothing pending").
+	 */
 	if (!addr)
 		return -EINVAL;
 
-	/* Replace the old array with the new one, carefully: others can
-	 * be accessing it at the same time */
+	/*
+	 * Replace the old array with the new one, carefully: others can
+	 * be accessing it at the same time.
+	 */
 	new = kmalloc(sizeof(*new) + sizeof(new->map[0]) * (old->num + 1),
 		      GFP_KERNEL);
 	if (!new)
@@ -52,22 +111,41 @@ static int add_eventfd(struct lguest *lg, unsigned long addr, int fd)
 	new->map[new->num].addr = addr;
 	new->map[new->num].event = eventfd_ctx_fdget(fd);
 	if (IS_ERR(new->map[new->num].event)) {
+		int err =  PTR_ERR(new->map[new->num].event);
 		kfree(new);
-		return PTR_ERR(new->map[new->num].event);
+		return err;
 	}
 	new->num++;
 
-	/* Now put new one in place. */
+	/*
+	 * Now put new one in place: rcu_assign_pointer() is a fancy way of
+	 * doing "lg->eventfds = new", but it uses memory barriers to make
+	 * absolutely sure that the contents of "new" written above is nailed
+	 * down before we actually do the assignment.
+	 *
+	 * We have to think about these kinds of things when we're operating on
+	 * live data without locks.
+	 */
 	rcu_assign_pointer(lg->eventfds, new);
 
-	/* We're not in a big hurry.  Wait until noone's looking at old
-	 * version, then delete it. */
+	/*
+	 * We're not in a big hurry.  Wait until noone's looking at old
+	 * version, then free it.
+	 */
 	synchronize_rcu();
 	kfree(old);
 
 	return 0;
 }
 
+/*L:052
+ * Receiving notifications from the Guest is usually done by attaching a
+ * particular LHCALL_NOTIFY value to an event filedescriptor.  The eventfd will
+ * become readable when the Guest does an LHCALL_NOTIFY with that value.
+ *
+ * This is really convenient for processing each virtqueue in a separate
+ * thread.
+ */
 static int attach_eventfd(struct lguest *lg, const unsigned long __user *input)
 {
 	unsigned long addr, fd;
@@ -79,15 +157,22 @@ static int attach_eventfd(struct lguest *lg, const unsigned long __user *input)
 	if (get_user(fd, input) != 0)
 		return -EFAULT;
 
+	/*
+	 * Just make sure two callers don't add eventfds at once.  We really
+	 * only need to lock against callers adding to the same Guest, so using
+	 * the Big Lguest Lock is overkill.  But this is setup, not a fast path.
+	 */
 	mutex_lock(&lguest_lock);
 	err = add_eventfd(lg, addr, fd);
 	mutex_unlock(&lguest_lock);
 
-	return 0;
+	return err;
 }
 
-/*L:050 Sending an interrupt is done by writing LHREQ_IRQ and an interrupt
- * number to /dev/lguest. */
+/*L:050
+ * Sending an interrupt is done by writing LHREQ_IRQ and an interrupt
+ * number to /dev/lguest.
+ */
 static int user_send_irq(struct lg_cpu *cpu, const unsigned long __user *input)
 {
 	unsigned long irq;
@@ -97,12 +182,18 @@ static int user_send_irq(struct lg_cpu *cpu, const unsigned long __user *input)
 	if (irq >= LGUEST_IRQS)
 		return -EINVAL;
 
+	/*
+	 * Next time the Guest runs, the core code will see if it can deliver
+	 * this interrupt.
+	 */
 	set_interrupt(cpu, irq);
 	return 0;
 }
 
-/*L:040 Once our Guest is initialized, the Launcher makes it run by reading
- * from /dev/lguest. */
+/*L:040
+ * Once our Guest is initialized, the Launcher makes it run by reading
+ * from /dev/lguest.
+ */
 static ssize_t read(struct file *file, char __user *user, size_t size,loff_t*o)
 {
 	struct lguest *lg = file->private_data;
@@ -138,8 +229,10 @@ static ssize_t read(struct file *file, char __user *user, size_t size,loff_t*o)
 		return len;
 	}
 
-	/* If we returned from read() last time because the Guest sent I/O,
-	 * clear the flag. */
+	/*
+	 * If we returned from read() last time because the Guest sent I/O,
+	 * clear the flag.
+	 */
 	if (cpu->pending_notify)
 		cpu->pending_notify = 0;
 
@@ -147,8 +240,10 @@ static ssize_t read(struct file *file, char __user *user, size_t size,loff_t*o)
 	return run_guest(cpu, (unsigned long __user *)user);
 }
 
-/*L:025 This actually initializes a CPU.  For the moment, a Guest is only
- * uniprocessor, so "id" is always 0. */
+/*L:025
+ * This actually initializes a CPU.  For the moment, a Guest is only
+ * uniprocessor, so "id" is always 0.
+ */
 static int lg_cpu_start(struct lg_cpu *cpu, unsigned id, unsigned long start_ip)
 {
 	/* We have a limited number the number of CPUs in the lguest struct. */
@@ -163,8 +258,10 @@ static int lg_cpu_start(struct lg_cpu *cpu, unsigned id, unsigned long start_ip)
 	/* Each CPU has a timer it can set. */
 	init_clockdev(cpu);
 
-	/* We need a complete page for the Guest registers: they are accessible
-	 * to the Guest and we can only grant it access to whole pages. */
+	/*
+	 * We need a complete page for the Guest registers: they are accessible
+	 * to the Guest and we can only grant it access to whole pages.
+	 */
 	cpu->regs_page = get_zeroed_page(GFP_KERNEL);
 	if (!cpu->regs_page)
 		return -ENOMEM;
@@ -172,29 +269,38 @@ static int lg_cpu_start(struct lg_cpu *cpu, unsigned id, unsigned long start_ip)
 	/* We actually put the registers at the bottom of the page. */
 	cpu->regs = (void *)cpu->regs_page + PAGE_SIZE - sizeof(*cpu->regs);
 
-	/* Now we initialize the Guest's registers, handing it the start
-	 * address. */
+	/*
+	 * Now we initialize the Guest's registers, handing it the start
+	 * address.
+	 */
 	lguest_arch_setup_regs(cpu, start_ip);
 
-	/* We keep a pointer to the Launcher task (ie. current task) for when
-	 * other Guests want to wake this one (eg. console input). */
+	/*
+	 * We keep a pointer to the Launcher task (ie. current task) for when
+	 * other Guests want to wake this one (eg. console input).
+	 */
 	cpu->tsk = current;
 
-	/* We need to keep a pointer to the Launcher's memory map, because if
+	/*
+	 * We need to keep a pointer to the Launcher's memory map, because if
 	 * the Launcher dies we need to clean it up.  If we don't keep a
-	 * reference, it is destroyed before close() is called. */
+	 * reference, it is destroyed before close() is called.
+	 */
 	cpu->mm = get_task_mm(cpu->tsk);
 
-	/* We remember which CPU's pages this Guest used last, for optimization
-	 * when the same Guest runs on the same CPU twice. */
+	/*
+	 * We remember which CPU's pages this Guest used last, for optimization
+	 * when the same Guest runs on the same CPU twice.
+	 */
 	cpu->last_pages = NULL;
 
 	/* No error == success. */
 	return 0;
 }
 
-/*L:020 The initialization write supplies 3 pointer sized (32 or 64 bit)
- * values (in addition to the LHREQ_INITIALIZE value).  These are:
+/*L:020
+ * The initialization write supplies 3 pointer sized (32 or 64 bit) values (in
+ * addition to the LHREQ_INITIALIZE value).  These are:
  *
  * base: The start of the Guest-physical memory inside the Launcher memory.
  *
@@ -206,14 +312,15 @@ static int lg_cpu_start(struct lg_cpu *cpu, unsigned id, unsigned long start_ip)
  */
 static int initialize(struct file *file, const unsigned long __user *input)
 {
-	/* "struct lguest" contains everything we (the Host) know about a
-	 * Guest. */
+	/* "struct lguest" contains all we (the Host) know about a Guest. */
 	struct lguest *lg;
 	int err;
 	unsigned long args[3];
 
-	/* We grab the Big Lguest lock, which protects against multiple
-	 * simultaneous initializations. */
+	/*
+	 * We grab the Big Lguest lock, which protects against multiple
+	 * simultaneous initializations.
+	 */
 	mutex_lock(&lguest_lock);
 	/* You can't initialize twice!  Close the device and start again... */
 	if (file->private_data) {
@@ -248,8 +355,10 @@ static int initialize(struct file *file, const unsigned long __user *input)
 	if (err)
 		goto free_eventfds;
 
-	/* Initialize the Guest's shadow page tables, using the toplevel
-	 * address the Launcher gave us.  This allocates memory, so can fail. */
+	/*
+	 * Initialize the Guest's shadow page tables, using the toplevel
+	 * address the Launcher gave us.  This allocates memory, so can fail.
+	 */
 	err = init_guest_pagetable(lg);
 	if (err)
 		goto free_regs;
@@ -274,20 +383,24 @@ unlock:
 	return err;
 }
 
-/*L:010 The first operation the Launcher does must be a write.  All writes
+/*L:010
+ * The first operation the Launcher does must be a write.  All writes
  * start with an unsigned long number: for the first write this must be
  * LHREQ_INITIALIZE to set up the Guest.  After that the Launcher can use
- * writes of other values to send interrupts.
+ * writes of other values to send interrupts or set up receipt of notifications.
  *
  * Note that we overload the "offset" in the /dev/lguest file to indicate what
- * CPU number we're dealing with.  Currently this is always 0, since we only
+ * CPU number we're dealing with.  Currently this is always 0 since we only
  * support uniprocessor Guests, but you can see the beginnings of SMP support
- * here. */
+ * here.
+ */
 static ssize_t write(struct file *file, const char __user *in,
 		     size_t size, loff_t *off)
 {
-	/* Once the Guest is initialized, we hold the "struct lguest" in the
-	 * file private data. */
+	/*
+	 * Once the Guest is initialized, we hold the "struct lguest" in the
+	 * file private data.
+	 */
 	struct lguest *lg = file->private_data;
 	const unsigned long __user *input = (const unsigned long __user *)in;
 	unsigned long req;
@@ -322,13 +435,15 @@ static ssize_t write(struct file *file, const char __user *in,
 	}
 }
 
-/*L:060 The final piece of interface code is the close() routine.  It reverses
+/*L:060
+ * The final piece of interface code is the close() routine.  It reverses
  * everything done in initialize().  This is usually called because the
  * Launcher exited.
  *
  * Note that the close routine returns 0 or a negative error number: it can't
  * really fail, but it can whine.  I blame Sun for this wart, and K&R C for
- * letting them do it. :*/
+ * letting them do it.
+:*/
 static int close(struct inode *inode, struct file *file)
 {
 	struct lguest *lg = file->private_data;
@@ -338,8 +453,10 @@ static int close(struct inode *inode, struct file *file)
 	if (!lg)
 		return 0;
 
-	/* We need the big lock, to protect from inter-guest I/O and other
-	 * Launchers initializing guests. */
+	/*
+	 * We need the big lock, to protect from inter-guest I/O and other
+	 * Launchers initializing guests.
+	 */
 	mutex_lock(&lguest_lock);
 
 	/* Free up the shadow page tables for the Guest. */
@@ -350,8 +467,10 @@ static int close(struct inode *inode, struct file *file)
 		hrtimer_cancel(&lg->cpus[i].hrt);
 		/* We can free up the register page we allocated. */
 		free_page(lg->cpus[i].regs_page);
-		/* Now all the memory cleanups are done, it's safe to release
-		 * the Launcher's memory management structure. */
+		/*
+		 * Now all the memory cleanups are done, it's safe to release
+		 * the Launcher's memory management structure.
+		 */
 		mmput(lg->cpus[i].mm);
 	}
 
@@ -360,8 +479,10 @@ static int close(struct inode *inode, struct file *file)
 		eventfd_ctx_put(lg->eventfds->map[i].event);
 	kfree(lg->eventfds);
 
-	/* If lg->dead doesn't contain an error code it will be NULL or a
-	 * kmalloc()ed string, either of which is ok to hand to kfree(). */
+	/*
+	 * If lg->dead doesn't contain an error code it will be NULL or a
+	 * kmalloc()ed string, either of which is ok to hand to kfree().
+	 */
 	if (!IS_ERR(lg->dead))
 		kfree(lg->dead);
 	/* Free the memory allocated to the lguest_struct */
@@ -385,7 +506,8 @@ static int close(struct inode *inode, struct file *file)
  *
  * We begin our understanding with the Host kernel interface which the Launcher
  * uses: reading and writing a character device called /dev/lguest.  All the
- * work happens in the read(), write() and close() routines: */
+ * work happens in the read(), write() and close() routines:
+ */
 static struct file_operations lguest_fops = {
 	.owner	 = THIS_MODULE,
 	.release = close,
@@ -393,8 +515,10 @@ static struct file_operations lguest_fops = {
 	.read	 = read,
 };
 
-/* This is a textbook example of a "misc" character device.  Populate a "struct
- * miscdevice" and register it with misc_register(). */
+/*
+ * This is a textbook example of a "misc" character device.  Populate a "struct
+ * miscdevice" and register it with misc_register().
+ */
 static struct miscdevice lguest_dev = {
 	.minor	= MISC_DYNAMIC_MINOR,
 	.name	= "lguest",
diff --git a/drivers/lguest/page_tables.c b/drivers/lguest/page_tables.c
index a6fe1abda24..a8d0aee3bc0 100644
--- a/drivers/lguest/page_tables.c
+++ b/drivers/lguest/page_tables.c
@@ -1,9 +1,11 @@
-/*P:700 The pagetable code, on the other hand, still shows the scars of
+/*P:700
+ * The pagetable code, on the other hand, still shows the scars of
  * previous encounters.  It's functional, and as neat as it can be in the
  * circumstances, but be wary, for these things are subtle and break easily.
  * The Guest provides a virtual to physical mapping, but we can neither trust
  * it nor use it: we verify and convert it here then point the CPU to the
- * converted Guest pages when running the Guest. :*/
+ * converted Guest pages when running the Guest.
+:*/
 
 /* Copyright (C) Rusty Russell IBM Corporation 2006.
  * GPL v2 and any later version */
@@ -17,18 +19,20 @@
 #include <asm/bootparam.h>
 #include "lg.h"
 
-/*M:008 We hold reference to pages, which prevents them from being swapped.
+/*M:008
+ * We hold reference to pages, which prevents them from being swapped.
  * It'd be nice to have a callback in the "struct mm_struct" when Linux wants
  * to swap out.  If we had this, and a shrinker callback to trim PTE pages, we
- * could probably consider launching Guests as non-root. :*/
+ * could probably consider launching Guests as non-root.
+:*/
 
 /*H:300
  * The Page Table Code
  *
- * We use two-level page tables for the Guest.  If you're not entirely
- * comfortable with virtual addresses, physical addresses and page tables then
- * I recommend you review arch/x86/lguest/boot.c's "Page Table Handling" (with
- * diagrams!).
+ * We use two-level page tables for the Guest, or three-level with PAE.  If
+ * you're not entirely comfortable with virtual addresses, physical addresses
+ * and page tables then I recommend you review arch/x86/lguest/boot.c's "Page
+ * Table Handling" (with diagrams!).
  *
  * The Guest keeps page tables, but we maintain the actual ones here: these are
  * called "shadow" page tables.  Which is a very Guest-centric name: these are
@@ -45,16 +49,18 @@
  *  (v) Flushing (throwing away) page tables,
  *  (vi) Mapping the Switcher when the Guest is about to run,
  *  (vii) Setting up the page tables initially.
- :*/
+:*/
 
-
-/* 1024 entries in a page table page maps 1024 pages: 4MB.  The Switcher is
- * conveniently placed at the top 4MB, so it uses a separate, complete PTE
- * page.  */
+/*
+ * The Switcher uses the complete top PTE page.  That's 1024 PTE entries (4MB)
+ * or 512 PTE entries with PAE (2MB).
+ */
 #define SWITCHER_PGD_INDEX (PTRS_PER_PGD - 1)
 
-/* For PAE we need the PMD index as well. We use the last 2MB, so we
- * will need the last pmd entry of the last pmd page.  */
+/*
+ * For PAE we need the PMD index as well. We use the last 2MB, so we
+ * will need the last pmd entry of the last pmd page.
+ */
 #ifdef CONFIG_X86_PAE
 #define SWITCHER_PMD_INDEX 	(PTRS_PER_PMD - 1)
 #define RESERVE_MEM 		2U
@@ -64,14 +70,18 @@
 #define CHECK_GPGD_MASK		_PAGE_TABLE
 #endif
 
-/* We actually need a separate PTE page for each CPU.  Remember that after the
+/*
+ * We actually need a separate PTE page for each CPU.  Remember that after the
  * Switcher code itself comes two pages for each CPU, and we don't want this
- * CPU's guest to see the pages of any other CPU. */
+ * CPU's guest to see the pages of any other CPU.
+ */
 static DEFINE_PER_CPU(pte_t *, switcher_pte_pages);
 #define switcher_pte_page(cpu) per_cpu(switcher_pte_pages, cpu)
 
-/*H:320 The page table code is curly enough to need helper functions to keep it
- * clear and clean.
+/*H:320
+ * The page table code is curly enough to need helper functions to keep it
+ * clear and clean.  The kernel itself provides many of them; one advantage
+ * of insisting that the Guest and Host use the same CONFIG_PAE setting.
  *
  * There are two functions which return pointers to the shadow (aka "real")
  * page tables.
@@ -79,7 +89,8 @@ static DEFINE_PER_CPU(pte_t *, switcher_pte_pages);
  * spgd_addr() takes the virtual address and returns a pointer to the top-level
  * page directory entry (PGD) for that address.  Since we keep track of several
  * page tables, the "i" argument tells us which one we're interested in (it's
- * usually the current one). */
+ * usually the current one).
+ */
 static pgd_t *spgd_addr(struct lg_cpu *cpu, u32 i, unsigned long vaddr)
 {
 	unsigned int index = pgd_index(vaddr);
@@ -96,9 +107,11 @@ static pgd_t *spgd_addr(struct lg_cpu *cpu, u32 i, unsigned long vaddr)
 }
 
 #ifdef CONFIG_X86_PAE
-/* This routine then takes the PGD entry given above, which contains the
+/*
+ * This routine then takes the PGD entry given above, which contains the
  * address of the PMD page.  It then returns a pointer to the PMD entry for the
- * given address. */
+ * given address.
+ */
 static pmd_t *spmd_addr(struct lg_cpu *cpu, pgd_t spgd, unsigned long vaddr)
 {
 	unsigned int index = pmd_index(vaddr);
@@ -119,9 +132,11 @@ static pmd_t *spmd_addr(struct lg_cpu *cpu, pgd_t spgd, unsigned long vaddr)
 }
 #endif
 
-/* This routine then takes the page directory entry returned above, which
+/*
+ * This routine then takes the page directory entry returned above, which
  * contains the address of the page table entry (PTE) page.  It then returns a
- * pointer to the PTE entry for the given address. */
+ * pointer to the PTE entry for the given address.
+ */
 static pte_t *spte_addr(struct lg_cpu *cpu, pgd_t spgd, unsigned long vaddr)
 {
 #ifdef CONFIG_X86_PAE
@@ -139,8 +154,10 @@ static pte_t *spte_addr(struct lg_cpu *cpu, pgd_t spgd, unsigned long vaddr)
 	return &page[pte_index(vaddr)];
 }
 
-/* These two functions just like the above two, except they access the Guest
- * page tables.  Hence they return a Guest address. */
+/*
+ * These functions are just like the above two, except they access the Guest
+ * page tables.  Hence they return a Guest address.
+ */
 static unsigned long gpgd_addr(struct lg_cpu *cpu, unsigned long vaddr)
 {
 	unsigned int index = vaddr >> (PGDIR_SHIFT);
@@ -148,6 +165,7 @@ static unsigned long gpgd_addr(struct lg_cpu *cpu, unsigned long vaddr)
 }
 
 #ifdef CONFIG_X86_PAE
+/* Follow the PGD to the PMD. */
 static unsigned long gpmd_addr(pgd_t gpgd, unsigned long vaddr)
 {
 	unsigned long gpage = pgd_pfn(gpgd) << PAGE_SHIFT;
@@ -155,6 +173,7 @@ static unsigned long gpmd_addr(pgd_t gpgd, unsigned long vaddr)
 	return gpage + pmd_index(vaddr) * sizeof(pmd_t);
 }
 
+/* Follow the PMD to the PTE. */
 static unsigned long gpte_addr(struct lg_cpu *cpu,
 			       pmd_t gpmd, unsigned long vaddr)
 {
@@ -164,6 +183,7 @@ static unsigned long gpte_addr(struct lg_cpu *cpu,
 	return gpage + pte_index(vaddr) * sizeof(pte_t);
 }
 #else
+/* Follow the PGD to the PTE (no mid-level for !PAE). */
 static unsigned long gpte_addr(struct lg_cpu *cpu,
 				pgd_t gpgd, unsigned long vaddr)
 {
@@ -175,17 +195,21 @@ static unsigned long gpte_addr(struct lg_cpu *cpu,
 #endif
 /*:*/
 
-/*M:014 get_pfn is slow: we could probably try to grab batches of pages here as
- * an optimization (ie. pre-faulting). :*/
+/*M:014
+ * get_pfn is slow: we could probably try to grab batches of pages here as
+ * an optimization (ie. pre-faulting).
+:*/
 
-/*H:350 This routine takes a page number given by the Guest and converts it to
+/*H:350
+ * This routine takes a page number given by the Guest and converts it to
  * an actual, physical page number.  It can fail for several reasons: the
  * virtual address might not be mapped by the Launcher, the write flag is set
  * and the page is read-only, or the write flag was set and the page was
  * shared so had to be copied, but we ran out of memory.
  *
  * This holds a reference to the page, so release_pte() is careful to put that
- * back. */
+ * back.
+ */
 static unsigned long get_pfn(unsigned long virtpfn, int write)
 {
 	struct page *page;
@@ -198,33 +222,41 @@ static unsigned long get_pfn(unsigned long virtpfn, int write)
 	return -1UL;
 }
 
-/*H:340 Converting a Guest page table entry to a shadow (ie. real) page table
+/*H:340
+ * Converting a Guest page table entry to a shadow (ie. real) page table
  * entry can be a little tricky.  The flags are (almost) the same, but the
  * Guest PTE contains a virtual page number: the CPU needs the real page
- * number. */
+ * number.
+ */
 static pte_t gpte_to_spte(struct lg_cpu *cpu, pte_t gpte, int write)
 {
 	unsigned long pfn, base, flags;
 
-	/* The Guest sets the global flag, because it thinks that it is using
+	/*
+	 * The Guest sets the global flag, because it thinks that it is using
 	 * PGE.  We only told it to use PGE so it would tell us whether it was
 	 * flushing a kernel mapping or a userspace mapping.  We don't actually
-	 * use the global bit, so throw it away. */
+	 * use the global bit, so throw it away.
+	 */
 	flags = (pte_flags(gpte) & ~_PAGE_GLOBAL);
 
 	/* The Guest's pages are offset inside the Launcher. */
 	base = (unsigned long)cpu->lg->mem_base / PAGE_SIZE;
 
-	/* We need a temporary "unsigned long" variable to hold the answer from
+	/*
+	 * We need a temporary "unsigned long" variable to hold the answer from
 	 * get_pfn(), because it returns 0xFFFFFFFF on failure, which wouldn't
 	 * fit in spte.pfn.  get_pfn() finds the real physical number of the
-	 * page, given the virtual number. */
+	 * page, given the virtual number.
+	 */
 	pfn = get_pfn(base + pte_pfn(gpte), write);
 	if (pfn == -1UL) {
 		kill_guest(cpu, "failed to get page %lu", pte_pfn(gpte));
-		/* When we destroy the Guest, we'll go through the shadow page
+		/*
+		 * When we destroy the Guest, we'll go through the shadow page
 		 * tables and release_pte() them.  Make sure we don't think
-		 * this one is valid! */
+		 * this one is valid!
+		 */
 		flags = 0;
 	}
 	/* Now we assemble our shadow PTE from the page number and flags. */
@@ -234,8 +266,10 @@ static pte_t gpte_to_spte(struct lg_cpu *cpu, pte_t gpte, int write)
 /*H:460 And to complete the chain, release_pte() looks like this: */
 static void release_pte(pte_t pte)
 {
-	/* Remember that get_user_pages_fast() took a reference to the page, in
-	 * get_pfn()?  We have to put it back now. */
+	/*
+	 * Remember that get_user_pages_fast() took a reference to the page, in
+	 * get_pfn()?  We have to put it back now.
+	 */
 	if (pte_flags(pte) & _PAGE_PRESENT)
 		put_page(pte_page(pte));
 }
@@ -273,7 +307,8 @@ static void check_gpmd(struct lg_cpu *cpu, pmd_t gpmd)
  * and return to the Guest without it knowing.
  *
  * If we fixed up the fault (ie. we mapped the address), this routine returns
- * true.  Otherwise, it was a real fault and we need to tell the Guest. */
+ * true.  Otherwise, it was a real fault and we need to tell the Guest.
+ */
 bool demand_page(struct lg_cpu *cpu, unsigned long vaddr, int errcode)
 {
 	pgd_t gpgd;
@@ -282,6 +317,7 @@ bool demand_page(struct lg_cpu *cpu, unsigned long vaddr, int errcode)
 	pte_t gpte;
 	pte_t *spte;
 
+	/* Mid level for PAE. */
 #ifdef CONFIG_X86_PAE
 	pmd_t *spmd;
 	pmd_t gpmd;
@@ -298,22 +334,26 @@ bool demand_page(struct lg_cpu *cpu, unsigned long vaddr, int errcode)
 	if (!(pgd_flags(*spgd) & _PAGE_PRESENT)) {
 		/* No shadow entry: allocate a new shadow PTE page. */
 		unsigned long ptepage = get_zeroed_page(GFP_KERNEL);
-		/* This is not really the Guest's fault, but killing it is
-		 * simple for this corner case. */
+		/*
+		 * This is not really the Guest's fault, but killing it is
+		 * simple for this corner case.
+		 */
 		if (!ptepage) {
 			kill_guest(cpu, "out of memory allocating pte page");
 			return false;
 		}
 		/* We check that the Guest pgd is OK. */
 		check_gpgd(cpu, gpgd);
-		/* And we copy the flags to the shadow PGD entry.  The page
-		 * number in the shadow PGD is the page we just allocated. */
+		/*
+		 * And we copy the flags to the shadow PGD entry.  The page
+		 * number in the shadow PGD is the page we just allocated.
+		 */
 		set_pgd(spgd, __pgd(__pa(ptepage) | pgd_flags(gpgd)));
 	}
 
 #ifdef CONFIG_X86_PAE
 	gpmd = lgread(cpu, gpmd_addr(gpgd, vaddr), pmd_t);
-	/* middle level not present?  We can't map it in. */
+	/* Middle level not present?  We can't map it in. */
 	if (!(pmd_flags(gpmd) & _PAGE_PRESENT))
 		return false;
 
@@ -324,8 +364,10 @@ bool demand_page(struct lg_cpu *cpu, unsigned long vaddr, int errcode)
 		/* No shadow entry: allocate a new shadow PTE page. */
 		unsigned long ptepage = get_zeroed_page(GFP_KERNEL);
 
-		/* This is not really the Guest's fault, but killing it is
-		* simple for this corner case. */
+		/*
+		 * This is not really the Guest's fault, but killing it is
+		 * simple for this corner case.
+		 */
 		if (!ptepage) {
 			kill_guest(cpu, "out of memory allocating pte page");
 			return false;
@@ -334,27 +376,37 @@ bool demand_page(struct lg_cpu *cpu, unsigned long vaddr, int errcode)
 		/* We check that the Guest pmd is OK. */
 		check_gpmd(cpu, gpmd);
 
-		/* And we copy the flags to the shadow PMD entry.  The page
-		 * number in the shadow PMD is the page we just allocated. */
+		/*
+		 * And we copy the flags to the shadow PMD entry.  The page
+		 * number in the shadow PMD is the page we just allocated.
+		 */
 		native_set_pmd(spmd, __pmd(__pa(ptepage) | pmd_flags(gpmd)));
 	}
 
-	/* OK, now we look at the lower level in the Guest page table: keep its
-	 * address, because we might update it later. */
+	/*
+	 * OK, now we look at the lower level in the Guest page table: keep its
+	 * address, because we might update it later.
+	 */
 	gpte_ptr = gpte_addr(cpu, gpmd, vaddr);
 #else
-	/* OK, now we look at the lower level in the Guest page table: keep its
-	 * address, because we might update it later. */
+	/*
+	 * OK, now we look at the lower level in the Guest page table: keep its
+	 * address, because we might update it later.
+	 */
 	gpte_ptr = gpte_addr(cpu, gpgd, vaddr);
 #endif
+
+	/* Read the actual PTE value. */
 	gpte = lgread(cpu, gpte_ptr, pte_t);
 
 	/* If this page isn't in the Guest page tables, we can't page it in. */
 	if (!(pte_flags(gpte) & _PAGE_PRESENT))
 		return false;
 
-	/* Check they're not trying to write to a page the Guest wants
-	 * read-only (bit 2 of errcode == write). */
+	/*
+	 * Check they're not trying to write to a page the Guest wants
+	 * read-only (bit 2 of errcode == write).
+	 */
 	if ((errcode & 2) && !(pte_flags(gpte) & _PAGE_RW))
 		return false;
 
@@ -362,8 +414,10 @@ bool demand_page(struct lg_cpu *cpu, unsigned long vaddr, int errcode)
 	if ((errcode & 4) && !(pte_flags(gpte) & _PAGE_USER))
 		return false;
 
-	/* Check that the Guest PTE flags are OK, and the page number is below
-	 * the pfn_limit (ie. not mapping the Launcher binary). */
+	/*
+	 * Check that the Guest PTE flags are OK, and the page number is below
+	 * the pfn_limit (ie. not mapping the Launcher binary).
+	 */
 	check_gpte(cpu, gpte);
 
 	/* Add the _PAGE_ACCESSED and (for a write) _PAGE_DIRTY flag */
@@ -373,29 +427,40 @@ bool demand_page(struct lg_cpu *cpu, unsigned long vaddr, int errcode)
 
 	/* Get the pointer to the shadow PTE entry we're going to set. */
 	spte = spte_addr(cpu, *spgd, vaddr);
-	/* If there was a valid shadow PTE entry here before, we release it.
-	 * This can happen with a write to a previously read-only entry. */
+
+	/*
+	 * If there was a valid shadow PTE entry here before, we release it.
+	 * This can happen with a write to a previously read-only entry.
+	 */
 	release_pte(*spte);
 
-	/* If this is a write, we insist that the Guest page is writable (the
-	 * final arg to gpte_to_spte()). */
+	/*
+	 * If this is a write, we insist that the Guest page is writable (the
+	 * final arg to gpte_to_spte()).
+	 */
 	if (pte_dirty(gpte))
 		*spte = gpte_to_spte(cpu, gpte, 1);
 	else
-		/* If this is a read, don't set the "writable" bit in the page
+		/*
+		 * If this is a read, don't set the "writable" bit in the page
 		 * table entry, even if the Guest says it's writable.  That way
 		 * we will come back here when a write does actually occur, so
-		 * we can update the Guest's _PAGE_DIRTY flag. */
+		 * we can update the Guest's _PAGE_DIRTY flag.
+		 */
 		native_set_pte(spte, gpte_to_spte(cpu, pte_wrprotect(gpte), 0));
 
-	/* Finally, we write the Guest PTE entry back: we've set the
-	 * _PAGE_ACCESSED and maybe the _PAGE_DIRTY flags. */
+	/*
+	 * Finally, we write the Guest PTE entry back: we've set the
+	 * _PAGE_ACCESSED and maybe the _PAGE_DIRTY flags.
+	 */
 	lgwrite(cpu, gpte_ptr, pte_t, gpte);
 
-	/* The fault is fixed, the page table is populated, the mapping
+	/*
+	 * The fault is fixed, the page table is populated, the mapping
 	 * manipulated, the result returned and the code complete.  A small
 	 * delay and a trace of alliteration are the only indications the Guest
-	 * has that a page fault occurred at all. */
+	 * has that a page fault occurred at all.
+	 */
 	return true;
 }
 
@@ -408,7 +473,8 @@ bool demand_page(struct lg_cpu *cpu, unsigned long vaddr, int errcode)
  * mapped, so it's overkill.
  *
  * This is a quick version which answers the question: is this virtual address
- * mapped by the shadow page tables, and is it writable? */
+ * mapped by the shadow page tables, and is it writable?
+ */
 static bool page_writable(struct lg_cpu *cpu, unsigned long vaddr)
 {
 	pgd_t *spgd;
@@ -428,21 +494,26 @@ static bool page_writable(struct lg_cpu *cpu, unsigned long vaddr)
 		return false;
 #endif
 
-	/* Check the flags on the pte entry itself: it must be present and
-	 * writable. */
+	/*
+	 * Check the flags on the pte entry itself: it must be present and
+	 * writable.
+	 */
 	flags = pte_flags(*(spte_addr(cpu, *spgd, vaddr)));
 
 	return (flags & (_PAGE_PRESENT|_PAGE_RW)) == (_PAGE_PRESENT|_PAGE_RW);
 }
 
-/* So, when pin_stack_pages() asks us to pin a page, we check if it's already
+/*
+ * So, when pin_stack_pages() asks us to pin a page, we check if it's already
  * in the page tables, and if not, we call demand_page() with error code 2
- * (meaning "write"). */
+ * (meaning "write").
+ */
 void pin_page(struct lg_cpu *cpu, unsigned long vaddr)
 {
 	if (!page_writable(cpu, vaddr) && !demand_page(cpu, vaddr, 2))
 		kill_guest(cpu, "bad stack page %#lx", vaddr);
 }
+/*:*/
 
 #ifdef CONFIG_X86_PAE
 static void release_pmd(pmd_t *spmd)
@@ -479,15 +550,21 @@ static void release_pgd(pgd_t *spgd)
 }
 
 #else /* !CONFIG_X86_PAE */
-/*H:450 If we chase down the release_pgd() code, it looks like this: */
+/*H:450
+ * If we chase down the release_pgd() code, the non-PAE version looks like
+ * this.  The PAE version is almost identical, but instead of calling
+ * release_pte it calls release_pmd(), which looks much like this.
+ */
 static void release_pgd(pgd_t *spgd)
 {
 	/* If the entry's not present, there's nothing to release. */
 	if (pgd_flags(*spgd) & _PAGE_PRESENT) {
 		unsigned int i;
-		/* Converting the pfn to find the actual PTE page is easy: turn
+		/*
+		 * Converting the pfn to find the actual PTE page is easy: turn
 		 * the page number into a physical address, then convert to a
-		 * virtual address (easy for kernel pages like this one). */
+		 * virtual address (easy for kernel pages like this one).
+		 */
 		pte_t *ptepage = __va(pgd_pfn(*spgd) << PAGE_SHIFT);
 		/* For each entry in the page, we might need to release it. */
 		for (i = 0; i < PTRS_PER_PTE; i++)
@@ -499,9 +576,12 @@ static void release_pgd(pgd_t *spgd)
 	}
 }
 #endif
-/*H:445 We saw flush_user_mappings() twice: once from the flush_user_mappings()
+
+/*H:445
+ * We saw flush_user_mappings() twice: once from the flush_user_mappings()
  * hypercall and once in new_pgdir() when we re-used a top-level pgdir page.
- * It simply releases every PTE page from 0 up to the Guest's kernel address. */
+ * It simply releases every PTE page from 0 up to the Guest's kernel address.
+ */
 static void flush_user_mappings(struct lguest *lg, int idx)
 {
 	unsigned int i;
@@ -510,10 +590,12 @@ static void flush_user_mappings(struct lguest *lg, int idx)
 		release_pgd(lg->pgdirs[idx].pgdir + i);
 }
 
-/*H:440 (v) Flushing (throwing away) page tables,
+/*H:440
+ * (v) Flushing (throwing away) page tables,
  *
  * The Guest has a hypercall to throw away the page tables: it's used when a
- * large number of mappings have been changed. */
+ * large number of mappings have been changed.
+ */
 void guest_pagetable_flush_user(struct lg_cpu *cpu)
 {
 	/* Drop the userspace part of the current page table. */
@@ -551,9 +633,11 @@ unsigned long guest_pa(struct lg_cpu *cpu, unsigned long vaddr)
 	return pte_pfn(gpte) * PAGE_SIZE | (vaddr & ~PAGE_MASK);
 }
 
-/* We keep several page tables.  This is a simple routine to find the page
+/*
+ * We keep several page tables.  This is a simple routine to find the page
  * table (if any) corresponding to this top-level address the Guest has given
- * us. */
+ * us.
+ */
 static unsigned int find_pgdir(struct lguest *lg, unsigned long pgtable)
 {
 	unsigned int i;
@@ -563,9 +647,11 @@ static unsigned int find_pgdir(struct lguest *lg, unsigned long pgtable)
 	return i;
 }
 
-/*H:435 And this is us, creating the new page directory.  If we really do
+/*H:435
+ * And this is us, creating the new page directory.  If we really do
  * allocate a new one (and so the kernel parts are not there), we set
- * blank_pgdir. */
+ * blank_pgdir.
+ */
 static unsigned int new_pgdir(struct lg_cpu *cpu,
 			      unsigned long gpgdir,
 			      int *blank_pgdir)
@@ -575,8 +661,10 @@ static unsigned int new_pgdir(struct lg_cpu *cpu,
 	pmd_t *pmd_table;
 #endif
 
-	/* We pick one entry at random to throw out.  Choosing the Least
-	 * Recently Used might be better, but this is easy. */
+	/*
+	 * We pick one entry at random to throw out.  Choosing the Least
+	 * Recently Used might be better, but this is easy.
+	 */
 	next = random32() % ARRAY_SIZE(cpu->lg->pgdirs);
 	/* If it's never been allocated at all before, try now. */
 	if (!cpu->lg->pgdirs[next].pgdir) {
@@ -587,8 +675,10 @@ static unsigned int new_pgdir(struct lg_cpu *cpu,
 			next = cpu->cpu_pgd;
 		else {
 #ifdef CONFIG_X86_PAE
-			/* In PAE mode, allocate a pmd page and populate the
-			 * last pgd entry. */
+			/*
+			 * In PAE mode, allocate a pmd page and populate the
+			 * last pgd entry.
+			 */
 			pmd_table = (pmd_t *)get_zeroed_page(GFP_KERNEL);
 			if (!pmd_table) {
 				free_page((long)cpu->lg->pgdirs[next].pgdir);
@@ -598,8 +688,10 @@ static unsigned int new_pgdir(struct lg_cpu *cpu,
 				set_pgd(cpu->lg->pgdirs[next].pgdir +
 					SWITCHER_PGD_INDEX,
 					__pgd(__pa(pmd_table) | _PAGE_PRESENT));
-				/* This is a blank page, so there are no kernel
-				 * mappings: caller must map the stack! */
+				/*
+				 * This is a blank page, so there are no kernel
+				 * mappings: caller must map the stack!
+				 */
 				*blank_pgdir = 1;
 			}
 #else
@@ -615,19 +707,23 @@ static unsigned int new_pgdir(struct lg_cpu *cpu,
 	return next;
 }
 
-/*H:430 (iv) Switching page tables
+/*H:430
+ * (iv) Switching page tables
  *
  * Now we've seen all the page table setting and manipulation, let's see
  * what happens when the Guest changes page tables (ie. changes the top-level
- * pgdir).  This occurs on almost every context switch. */
+ * pgdir).  This occurs on almost every context switch.
+ */
 void guest_new_pagetable(struct lg_cpu *cpu, unsigned long pgtable)
 {
 	int newpgdir, repin = 0;
 
 	/* Look to see if we have this one already. */
 	newpgdir = find_pgdir(cpu->lg, pgtable);
-	/* If not, we allocate or mug an existing one: if it's a fresh one,
-	 * repin gets set to 1. */
+	/*
+	 * If not, we allocate or mug an existing one: if it's a fresh one,
+	 * repin gets set to 1.
+	 */
 	if (newpgdir == ARRAY_SIZE(cpu->lg->pgdirs))
 		newpgdir = new_pgdir(cpu, pgtable, &repin);
 	/* Change the current pgd index to the new one. */
@@ -637,9 +733,11 @@ void guest_new_pagetable(struct lg_cpu *cpu, unsigned long pgtable)
 		pin_stack_pages(cpu);
 }
 
-/*H:470 Finally, a routine which throws away everything: all PGD entries in all
+/*H:470
+ * Finally, a routine which throws away everything: all PGD entries in all
  * the shadow page tables, including the Guest's kernel mappings.  This is used
- * when we destroy the Guest. */
+ * when we destroy the Guest.
+ */
 static void release_all_pagetables(struct lguest *lg)
 {
 	unsigned int i, j;
@@ -656,8 +754,10 @@ static void release_all_pagetables(struct lguest *lg)
 			spgd = lg->pgdirs[i].pgdir + SWITCHER_PGD_INDEX;
 			pmdpage = __va(pgd_pfn(*spgd) << PAGE_SHIFT);
 
-			/* And release the pmd entries of that pmd page,
-			 * except for the switcher pmd. */
+			/*
+			 * And release the pmd entries of that pmd page,
+			 * except for the switcher pmd.
+			 */
 			for (k = 0; k < SWITCHER_PMD_INDEX; k++)
 				release_pmd(&pmdpage[k]);
 #endif
@@ -667,10 +767,12 @@ static void release_all_pagetables(struct lguest *lg)
 		}
 }
 
-/* We also throw away everything when a Guest tells us it's changed a kernel
+/*
+ * We also throw away everything when a Guest tells us it's changed a kernel
  * mapping.  Since kernel mappings are in every page table, it's easiest to
  * throw them all away.  This traps the Guest in amber for a while as
- * everything faults back in, but it's rare. */
+ * everything faults back in, but it's rare.
+ */
 void guest_pagetable_clear_all(struct lg_cpu *cpu)
 {
 	release_all_pagetables(cpu->lg);
@@ -678,15 +780,19 @@ void guest_pagetable_clear_all(struct lg_cpu *cpu)
 	pin_stack_pages(cpu);
 }
 /*:*/
-/*M:009 Since we throw away all mappings when a kernel mapping changes, our
+
+/*M:009
+ * Since we throw away all mappings when a kernel mapping changes, our
  * performance sucks for guests using highmem.  In fact, a guest with
  * PAGE_OFFSET 0xc0000000 (the default) and more than about 700MB of RAM is
  * usually slower than a Guest with less memory.
  *
  * This, of course, cannot be fixed.  It would take some kind of... well, I
- * don't know, but the term "puissant code-fu" comes to mind. :*/
+ * don't know, but the term "puissant code-fu" comes to mind.
+:*/
 
-/*H:420 This is the routine which actually sets the page table entry for then
+/*H:420
+ * This is the routine which actually sets the page table entry for then
  * "idx"'th shadow page table.
  *
  * Normally, we can just throw out the old entry and replace it with 0: if they
@@ -715,31 +821,36 @@ static void do_set_pte(struct lg_cpu *cpu, int idx,
 		spmd = spmd_addr(cpu, *spgd, vaddr);
 		if (pmd_flags(*spmd) & _PAGE_PRESENT) {
 #endif
-			/* Otherwise, we start by releasing
-			 * the existing entry. */
+			/* Otherwise, start by releasing the existing entry. */
 			pte_t *spte = spte_addr(cpu, *spgd, vaddr);
 			release_pte(*spte);
 
-			/* If they're setting this entry as dirty or accessed,
-			 * we might as well put that entry they've given us
-			 * in now.  This shaves 10% off a
-			 * copy-on-write micro-benchmark. */
+			/*
+			 * If they're setting this entry as dirty or accessed,
+			 * we might as well put that entry they've given us in
+			 * now.  This shaves 10% off a copy-on-write
+			 * micro-benchmark.
+			 */
 			if (pte_flags(gpte) & (_PAGE_DIRTY | _PAGE_ACCESSED)) {
 				check_gpte(cpu, gpte);
 				native_set_pte(spte,
 						gpte_to_spte(cpu, gpte,
 						pte_flags(gpte) & _PAGE_DIRTY));
-			} else
-				/* Otherwise kill it and we can demand_page()
-				 * it in later. */
+			} else {
+				/*
+				 * Otherwise kill it and we can demand_page()
+				 * it in later.
+				 */
 				native_set_pte(spte, __pte(0));
+			}
 #ifdef CONFIG_X86_PAE
 		}
 #endif
 	}
 }
 
-/*H:410 Updating a PTE entry is a little trickier.
+/*H:410
+ * Updating a PTE entry is a little trickier.
  *
  * We keep track of several different page tables (the Guest uses one for each
  * process, so it makes sense to cache at least a few).  Each of these have
@@ -748,12 +859,15 @@ static void do_set_pte(struct lg_cpu *cpu, int idx,
  * all the page tables, not just the current one.  This is rare.
  *
  * The benefit is that when we have to track a new page table, we can keep all
- * the kernel mappings.  This speeds up context switch immensely. */
+ * the kernel mappings.  This speeds up context switch immensely.
+ */
 void guest_set_pte(struct lg_cpu *cpu,
 		   unsigned long gpgdir, unsigned long vaddr, pte_t gpte)
 {
-	/* Kernel mappings must be changed on all top levels.  Slow, but doesn't
-	 * happen often. */
+	/*
+	 * Kernel mappings must be changed on all top levels.  Slow, but doesn't
+	 * happen often.
+	 */
 	if (vaddr >= cpu->lg->kernel_address) {
 		unsigned int i;
 		for (i = 0; i < ARRAY_SIZE(cpu->lg->pgdirs); i++)
@@ -795,19 +909,25 @@ void guest_set_pgd(struct lguest *lg, unsigned long gpgdir, u32 idx)
 		/* ... throw it away. */
 		release_pgd(lg->pgdirs[pgdir].pgdir + idx);
 }
+
 #ifdef CONFIG_X86_PAE
+/* For setting a mid-level, we just throw everything away.  It's easy. */
 void guest_set_pmd(struct lguest *lg, unsigned long pmdp, u32 idx)
 {
 	guest_pagetable_clear_all(&lg->cpus[0]);
 }
 #endif
 
-/* Once we know how much memory we have we can construct simple identity
- * (which set virtual == physical) and linear mappings
- * which will get the Guest far enough into the boot to create its own.
+/*H:505
+ * To get through boot, we construct simple identity page mappings (which
+ * set virtual == physical) and linear mappings which will get the Guest far
+ * enough into the boot to create its own.  The linear mapping means we
+ * simplify the Guest boot, but it makes assumptions about their PAGE_OFFSET,
+ * as you'll see.
  *
  * We lay them out of the way, just below the initrd (which is why we need to
- * know its size here). */
+ * know its size here).
+ */
 static unsigned long setup_pagetables(struct lguest *lg,
 				      unsigned long mem,
 				      unsigned long initrd_size)
@@ -825,8 +945,10 @@ static unsigned long setup_pagetables(struct lguest *lg,
 	unsigned int phys_linear;
 #endif
 
-	/* We have mapped_pages frames to map, so we need
-	 * linear_pages page tables to map them. */
+	/*
+	 * We have mapped_pages frames to map, so we need linear_pages page
+	 * tables to map them.
+	 */
 	mapped_pages = mem / PAGE_SIZE;
 	linear_pages = (mapped_pages + PTRS_PER_PTE - 1) / PTRS_PER_PTE;
 
@@ -837,10 +959,16 @@ static unsigned long setup_pagetables(struct lguest *lg,
 	linear = (void *)pgdir - linear_pages * PAGE_SIZE;
 
 #ifdef CONFIG_X86_PAE
+	/*
+	 * And the single mid page goes below that.  We only use one, but
+	 * that's enough to map 1G, which definitely gets us through boot.
+	 */
 	pmds = (void *)linear - PAGE_SIZE;
 #endif
-	/* Linear mapping is easy: put every page's address into the
-	 * mapping in order. */
+	/*
+	 * Linear mapping is easy: put every page's address into the
+	 * mapping in order.
+	 */
 	for (i = 0; i < mapped_pages; i++) {
 		pte_t pte;
 		pte = pfn_pte(i, __pgprot(_PAGE_PRESENT|_PAGE_RW|_PAGE_USER));
@@ -848,11 +976,14 @@ static unsigned long setup_pagetables(struct lguest *lg,
 			return -EFAULT;
 	}
 
-	/* The top level points to the linear page table pages above.
-	 * We setup the identity and linear mappings here. */
 #ifdef CONFIG_X86_PAE
+	/*
+	 * Make the Guest PMD entries point to the corresponding place in the
+	 * linear mapping (up to one page worth of PMD).
+	 */
 	for (i = j = 0; i < mapped_pages && j < PTRS_PER_PMD;
 	     i += PTRS_PER_PTE, j++) {
+		/* FIXME: native_set_pmd is overkill here. */
 		native_set_pmd(&pmd, __pmd(((unsigned long)(linear + i)
 		- mem_base) | _PAGE_PRESENT | _PAGE_RW | _PAGE_USER));
 
@@ -860,18 +991,36 @@ static unsigned long setup_pagetables(struct lguest *lg,
 			return -EFAULT;
 	}
 
+	/* One PGD entry, pointing to that PMD page. */
 	set_pgd(&pgd, __pgd(((u32)pmds - mem_base) | _PAGE_PRESENT));
+	/* Copy it in as the first PGD entry (ie. addresses 0-1G). */
 	if (copy_to_user(&pgdir[0], &pgd, sizeof(pgd)) != 0)
 		return -EFAULT;
+	/*
+	 * And the third PGD entry (ie. addresses 3G-4G).
+	 *
+	 * FIXME: This assumes that PAGE_OFFSET for the Guest is 0xC0000000.
+	 */
 	if (copy_to_user(&pgdir[3], &pgd, sizeof(pgd)) != 0)
 		return -EFAULT;
 #else
+	/*
+	 * The top level points to the linear page table pages above.
+	 * We setup the identity and linear mappings here.
+	 */
 	phys_linear = (unsigned long)linear - mem_base;
 	for (i = 0; i < mapped_pages; i += PTRS_PER_PTE) {
 		pgd_t pgd;
+		/*
+		 * Create a PGD entry which points to the right part of the
+		 * linear PTE pages.
+		 */
 		pgd = __pgd((phys_linear + i * sizeof(pte_t)) |
 			    (_PAGE_PRESENT | _PAGE_RW | _PAGE_USER));
 
+		/*
+		 * Copy it into the PGD page at 0 and PAGE_OFFSET.
+		 */
 		if (copy_to_user(&pgdir[i / PTRS_PER_PTE], &pgd, sizeof(pgd))
 		    || copy_to_user(&pgdir[pgd_index(PAGE_OFFSET)
 					   + i / PTRS_PER_PTE],
@@ -880,15 +1029,19 @@ static unsigned long setup_pagetables(struct lguest *lg,
 	}
 #endif
 
-	/* We return the top level (guest-physical) address: remember where
-	 * this is. */
+	/*
+	 * We return the top level (guest-physical) address: we remember where
+	 * this is to write it into lguest_data when the Guest initializes.
+	 */
 	return (unsigned long)pgdir - mem_base;
 }
 
-/*H:500 (vii) Setting up the page tables initially.
+/*H:500
+ * (vii) Setting up the page tables initially.
  *
  * When a Guest is first created, the Launcher tells us where the toplevel of
- * its first page table is.  We set some things up here: */
+ * its first page table is.  We set some things up here:
+ */
 int init_guest_pagetable(struct lguest *lg)
 {
 	u64 mem;
@@ -898,21 +1051,27 @@ int init_guest_pagetable(struct lguest *lg)
 	pgd_t *pgd;
 	pmd_t *pmd_table;
 #endif
-	/* Get the Guest memory size and the ramdisk size from the boot header
-	 * located at lg->mem_base (Guest address 0). */
+	/*
+	 * Get the Guest memory size and the ramdisk size from the boot header
+	 * located at lg->mem_base (Guest address 0).
+	 */
 	if (copy_from_user(&mem, &boot->e820_map[0].size, sizeof(mem))
 	    || get_user(initrd_size, &boot->hdr.ramdisk_size))
 		return -EFAULT;
 
-	/* We start on the first shadow page table, and give it a blank PGD
-	 * page. */
+	/*
+	 * We start on the first shadow page table, and give it a blank PGD
+	 * page.
+	 */
 	lg->pgdirs[0].gpgdir = setup_pagetables(lg, mem, initrd_size);
 	if (IS_ERR_VALUE(lg->pgdirs[0].gpgdir))
 		return lg->pgdirs[0].gpgdir;
 	lg->pgdirs[0].pgdir = (pgd_t *)get_zeroed_page(GFP_KERNEL);
 	if (!lg->pgdirs[0].pgdir)
 		return -ENOMEM;
+
 #ifdef CONFIG_X86_PAE
+	/* For PAE, we also create the initial mid-level. */
 	pgd = lg->pgdirs[0].pgdir;
 	pmd_table = (pmd_t *) get_zeroed_page(GFP_KERNEL);
 	if (!pmd_table)
@@ -921,27 +1080,33 @@ int init_guest_pagetable(struct lguest *lg)
 	set_pgd(pgd + SWITCHER_PGD_INDEX,
 		__pgd(__pa(pmd_table) | _PAGE_PRESENT));
 #endif
+
+	/* This is the current page table. */
 	lg->cpus[0].cpu_pgd = 0;
 	return 0;
 }
 
-/* When the Guest calls LHCALL_LGUEST_INIT we do more setup. */
+/*H:508 When the Guest calls LHCALL_LGUEST_INIT we do more setup. */
 void page_table_guest_data_init(struct lg_cpu *cpu)
 {
 	/* We get the kernel address: above this is all kernel memory. */
 	if (get_user(cpu->lg->kernel_address,
 		&cpu->lg->lguest_data->kernel_address)
-		/* We tell the Guest that it can't use the top 2 or 4 MB
-		 * of virtual addresses used by the Switcher. */
+		/*
+		 * We tell the Guest that it can't use the top 2 or 4 MB
+		 * of virtual addresses used by the Switcher.
+		 */
 		|| put_user(RESERVE_MEM * 1024 * 1024,
 			&cpu->lg->lguest_data->reserve_mem)
 		|| put_user(cpu->lg->pgdirs[0].gpgdir,
 			&cpu->lg->lguest_data->pgdir))
 		kill_guest(cpu, "bad guest page %p", cpu->lg->lguest_data);
 
-	/* In flush_user_mappings() we loop from 0 to
+	/*
+	 * In flush_user_mappings() we loop from 0 to
 	 * "pgd_index(lg->kernel_address)".  This assumes it won't hit the
-	 * Switcher mappings, so check that now. */
+	 * Switcher mappings, so check that now.
+	 */
 #ifdef CONFIG_X86_PAE
 	if (pgd_index(cpu->lg->kernel_address) == SWITCHER_PGD_INDEX &&
 		pmd_index(cpu->lg->kernel_address) == SWITCHER_PMD_INDEX)
@@ -964,12 +1129,14 @@ void free_guest_pagetable(struct lguest *lg)
 		free_page((long)lg->pgdirs[i].pgdir);
 }
 
-/*H:480 (vi) Mapping the Switcher when the Guest is about to run.
+/*H:480
+ * (vi) Mapping the Switcher when the Guest is about to run.
  *
  * The Switcher and the two pages for this CPU need to be visible in the
  * Guest (and not the pages for other CPUs).  We have the appropriate PTE pages
  * for each CPU already set up, we just need to hook them in now we know which
- * Guest is about to run on this CPU. */
+ * Guest is about to run on this CPU.
+ */
 void map_switcher_in_guest(struct lg_cpu *cpu, struct lguest_pages *pages)
 {
 	pte_t *switcher_pte_page = __get_cpu_var(switcher_pte_pages);
@@ -980,30 +1147,38 @@ void map_switcher_in_guest(struct lg_cpu *cpu, struct lguest_pages *pages)
 	pmd_t switcher_pmd;
 	pmd_t *pmd_table;
 
+	/* FIXME: native_set_pmd is overkill here. */
 	native_set_pmd(&switcher_pmd, pfn_pmd(__pa(switcher_pte_page) >>
 		       PAGE_SHIFT, PAGE_KERNEL_EXEC));
 
+	/* Figure out where the pmd page is, by reading the PGD, and converting
+	 * it to a virtual address. */
 	pmd_table = __va(pgd_pfn(cpu->lg->
 			pgdirs[cpu->cpu_pgd].pgdir[SWITCHER_PGD_INDEX])
 								<< PAGE_SHIFT);
+	/* Now write it into the shadow page table. */
 	native_set_pmd(&pmd_table[SWITCHER_PMD_INDEX], switcher_pmd);
 #else
 	pgd_t switcher_pgd;
 
-	/* Make the last PGD entry for this Guest point to the Switcher's PTE
-	 * page for this CPU (with appropriate flags). */
+	/*
+	 * Make the last PGD entry for this Guest point to the Switcher's PTE
+	 * page for this CPU (with appropriate flags).
+	 */
 	switcher_pgd = __pgd(__pa(switcher_pte_page) | __PAGE_KERNEL_EXEC);
 
 	cpu->lg->pgdirs[cpu->cpu_pgd].pgdir[SWITCHER_PGD_INDEX] = switcher_pgd;
 
 #endif
-	/* We also change the Switcher PTE page.  When we're running the Guest,
+	/*
+	 * We also change the Switcher PTE page.  When we're running the Guest,
 	 * we want the Guest's "regs" page to appear where the first Switcher
 	 * page for this CPU is.  This is an optimization: when the Switcher
 	 * saves the Guest registers, it saves them into the first page of this
 	 * CPU's "struct lguest_pages": if we make sure the Guest's register
 	 * page is already mapped there, we don't have to copy them out
-	 * again. */
+	 * again.
+	 */
 	pfn = __pa(cpu->regs_page) >> PAGE_SHIFT;
 	native_set_pte(&regs_pte, pfn_pte(pfn, PAGE_KERNEL));
 	native_set_pte(&switcher_pte_page[pte_index((unsigned long)pages)],
@@ -1019,10 +1194,12 @@ static void free_switcher_pte_pages(void)
 		free_page((long)switcher_pte_page(i));
 }
 
-/*H:520 Setting up the Switcher PTE page for given CPU is fairly easy, given
+/*H:520
+ * Setting up the Switcher PTE page for given CPU is fairly easy, given
  * the CPU number and the "struct page"s for the Switcher code itself.
  *
- * Currently the Switcher is less than a page long, so "pages" is always 1. */
+ * Currently the Switcher is less than a page long, so "pages" is always 1.
+ */
 static __init void populate_switcher_pte_page(unsigned int cpu,
 					      struct page *switcher_page[],
 					      unsigned int pages)
@@ -1043,13 +1220,16 @@ static __init void populate_switcher_pte_page(unsigned int cpu,
 	native_set_pte(&pte[i], pfn_pte(page_to_pfn(switcher_page[i]),
 			 __pgprot(_PAGE_PRESENT|_PAGE_ACCESSED|_PAGE_RW)));
 
-	/* The second page contains the "struct lguest_ro_state", and is
-	 * read-only. */
+	/*
+	 * The second page contains the "struct lguest_ro_state", and is
+	 * read-only.
+	 */
 	native_set_pte(&pte[i+1], pfn_pte(page_to_pfn(switcher_page[i+1]),
 			   __pgprot(_PAGE_PRESENT|_PAGE_ACCESSED)));
 }
 
-/* We've made it through the page table code.  Perhaps our tired brains are
+/*
+ * We've made it through the page table code.  Perhaps our tired brains are
  * still processing the details, or perhaps we're simply glad it's over.
  *
  * If nothing else, note that all this complexity in juggling shadow page tables
@@ -1058,10 +1238,13 @@ static __init void populate_switcher_pte_page(unsigned int cpu,
  * uses exotic direct Guest pagetable manipulation, and why both Intel and AMD
  * have implemented shadow page table support directly into hardware.
  *
- * There is just one file remaining in the Host. */
+ * There is just one file remaining in the Host.
+ */
 
-/*H:510 At boot or module load time, init_pagetables() allocates and populates
- * the Switcher PTE page for each CPU. */
+/*H:510
+ * At boot or module load time, init_pagetables() allocates and populates
+ * the Switcher PTE page for each CPU.
+ */
 __init int init_pagetables(struct page **switcher_page, unsigned int pages)
 {
 	unsigned int i;
diff --git a/drivers/lguest/segments.c b/drivers/lguest/segments.c
index 482ed5a1875..951c57b0a7e 100644
--- a/drivers/lguest/segments.c
+++ b/drivers/lguest/segments.c
@@ -1,4 +1,5 @@
-/*P:600 The x86 architecture has segments, which involve a table of descriptors
+/*P:600
+ * The x86 architecture has segments, which involve a table of descriptors
  * which can be used to do funky things with virtual address interpretation.
  * We originally used to use segments so the Guest couldn't alter the
  * Guest<->Host Switcher, and then we had to trim Guest segments, and restore
@@ -8,7 +9,8 @@
  *
  * In these modern times, the segment handling code consists of simple sanity
  * checks, and the worst you'll experience reading this code is butterfly-rash
- * from frolicking through its parklike serenity. :*/
+ * from frolicking through its parklike serenity.
+:*/
 #include "lg.h"
 
 /*H:600
@@ -41,10 +43,12 @@
  * begin.
  */
 
-/* There are several entries we don't let the Guest set.  The TSS entry is the
+/*
+ * There are several entries we don't let the Guest set.  The TSS entry is the
  * "Task State Segment" which controls all kinds of delicate things.  The
  * LGUEST_CS and LGUEST_DS entries are reserved for the Switcher, and the
- * the Guest can't be trusted to deal with double faults. */
+ * the Guest can't be trusted to deal with double faults.
+ */
 static bool ignored_gdt(unsigned int num)
 {
 	return (num == GDT_ENTRY_TSS
@@ -53,42 +57,52 @@ static bool ignored_gdt(unsigned int num)
 		|| num == GDT_ENTRY_DOUBLEFAULT_TSS);
 }
 
-/*H:630 Once the Guest gave us new GDT entries, we fix them up a little.  We
+/*H:630
+ * Once the Guest gave us new GDT entries, we fix them up a little.  We
  * don't care if they're invalid: the worst that can happen is a General
  * Protection Fault in the Switcher when it restores a Guest segment register
  * which tries to use that entry.  Then we kill the Guest for causing such a
- * mess: the message will be "unhandled trap 256". */
+ * mess: the message will be "unhandled trap 256".
+ */
 static void fixup_gdt_table(struct lg_cpu *cpu, unsigned start, unsigned end)
 {
 	unsigned int i;
 
 	for (i = start; i < end; i++) {
-		/* We never copy these ones to real GDT, so we don't care what
-		 * they say */
+		/*
+		 * We never copy these ones to real GDT, so we don't care what
+		 * they say
+		 */
 		if (ignored_gdt(i))
 			continue;
 
-		/* Segment descriptors contain a privilege level: the Guest is
+		/*
+		 * Segment descriptors contain a privilege level: the Guest is
 		 * sometimes careless and leaves this as 0, even though it's
-		 * running at privilege level 1.  If so, we fix it here. */
+		 * running at privilege level 1.  If so, we fix it here.
+		 */
 		if ((cpu->arch.gdt[i].b & 0x00006000) == 0)
 			cpu->arch.gdt[i].b |= (GUEST_PL << 13);
 
-		/* Each descriptor has an "accessed" bit.  If we don't set it
+		/*
+		 * Each descriptor has an "accessed" bit.  If we don't set it
 		 * now, the CPU will try to set it when the Guest first loads
 		 * that entry into a segment register.  But the GDT isn't
-		 * writable by the Guest, so bad things can happen. */
+		 * writable by the Guest, so bad things can happen.
+		 */
 		cpu->arch.gdt[i].b |= 0x00000100;
 	}
 }
 
-/*H:610 Like the IDT, we never simply use the GDT the Guest gives us.  We keep
+/*H:610
+ * Like the IDT, we never simply use the GDT the Guest gives us.  We keep
  * a GDT for each CPU, and copy across the Guest's entries each time we want to
  * run the Guest on that CPU.
  *
  * This routine is called at boot or modprobe time for each CPU to set up the
  * constant GDT entries: the ones which are the same no matter what Guest we're
- * running. */
+ * running.
+ */
 void setup_default_gdt_entries(struct lguest_ro_state *state)
 {
 	struct desc_struct *gdt = state->guest_gdt;
@@ -98,30 +112,37 @@ void setup_default_gdt_entries(struct lguest_ro_state *state)
 	gdt[GDT_ENTRY_LGUEST_CS] = FULL_EXEC_SEGMENT;
 	gdt[GDT_ENTRY_LGUEST_DS] = FULL_SEGMENT;
 
-	/* The TSS segment refers to the TSS entry for this particular CPU.
+	/*
+	 * The TSS segment refers to the TSS entry for this particular CPU.
 	 * Forgive the magic flags: the 0x8900 means the entry is Present, it's
 	 * privilege level 0 Available 386 TSS system segment, and the 0x67
-	 * means Saturn is eclipsed by Mercury in the twelfth house. */
+	 * means Saturn is eclipsed by Mercury in the twelfth house.
+	 */
 	gdt[GDT_ENTRY_TSS].a = 0x00000067 | (tss << 16);
 	gdt[GDT_ENTRY_TSS].b = 0x00008900 | (tss & 0xFF000000)
 		| ((tss >> 16) & 0x000000FF);
 }
 
-/* This routine sets up the initial Guest GDT for booting.  All entries start
- * as 0 (unusable). */
+/*
+ * This routine sets up the initial Guest GDT for booting.  All entries start
+ * as 0 (unusable).
+ */
 void setup_guest_gdt(struct lg_cpu *cpu)
 {
-	/* Start with full 0-4G segments... */
+	/*
+	 * Start with full 0-4G segments...except the Guest is allowed to use
+	 * them, so set the privilege level appropriately in the flags.
+	 */
 	cpu->arch.gdt[GDT_ENTRY_KERNEL_CS] = FULL_EXEC_SEGMENT;
 	cpu->arch.gdt[GDT_ENTRY_KERNEL_DS] = FULL_SEGMENT;
-	/* ...except the Guest is allowed to use them, so set the privilege
-	 * level appropriately in the flags. */
 	cpu->arch.gdt[GDT_ENTRY_KERNEL_CS].b |= (GUEST_PL << 13);
 	cpu->arch.gdt[GDT_ENTRY_KERNEL_DS].b |= (GUEST_PL << 13);
 }
 
-/*H:650 An optimization of copy_gdt(), for just the three "thead-local storage"
- * entries. */
+/*H:650
+ * An optimization of copy_gdt(), for just the three "thead-local storage"
+ * entries.
+ */
 void copy_gdt_tls(const struct lg_cpu *cpu, struct desc_struct *gdt)
 {
 	unsigned int i;
@@ -130,26 +151,34 @@ void copy_gdt_tls(const struct lg_cpu *cpu, struct desc_struct *gdt)
 		gdt[i] = cpu->arch.gdt[i];
 }
 
-/*H:640 When the Guest is run on a different CPU, or the GDT entries have
- * changed, copy_gdt() is called to copy the Guest's GDT entries across to this
- * CPU's GDT. */
+/*H:640
+ * When the Guest is run on a different CPU, or the GDT entries have changed,
+ * copy_gdt() is called to copy the Guest's GDT entries across to this CPU's
+ * GDT.
+ */
 void copy_gdt(const struct lg_cpu *cpu, struct desc_struct *gdt)
 {
 	unsigned int i;
 
-	/* The default entries from setup_default_gdt_entries() are not
-	 * replaced.  See ignored_gdt() above. */
+	/*
+	 * The default entries from setup_default_gdt_entries() are not
+	 * replaced.  See ignored_gdt() above.
+	 */
 	for (i = 0; i < GDT_ENTRIES; i++)
 		if (!ignored_gdt(i))
 			gdt[i] = cpu->arch.gdt[i];
 }
 
-/*H:620 This is where the Guest asks us to load a new GDT entry
- * (LHCALL_LOAD_GDT_ENTRY).  We tweak the entry and copy it in. */
+/*H:620
+ * This is where the Guest asks us to load a new GDT entry
+ * (LHCALL_LOAD_GDT_ENTRY).  We tweak the entry and copy it in.
+ */
 void load_guest_gdt_entry(struct lg_cpu *cpu, u32 num, u32 lo, u32 hi)
 {
-	/* We assume the Guest has the same number of GDT entries as the
-	 * Host, otherwise we'd have to dynamically allocate the Guest GDT. */
+	/*
+	 * We assume the Guest has the same number of GDT entries as the
+	 * Host, otherwise we'd have to dynamically allocate the Guest GDT.
+	 */
 	if (num >= ARRAY_SIZE(cpu->arch.gdt))
 		kill_guest(cpu, "too many gdt entries %i", num);
 
@@ -157,15 +186,19 @@ void load_guest_gdt_entry(struct lg_cpu *cpu, u32 num, u32 lo, u32 hi)
 	cpu->arch.gdt[num].a = lo;
 	cpu->arch.gdt[num].b = hi;
 	fixup_gdt_table(cpu, num, num+1);
-	/* Mark that the GDT changed so the core knows it has to copy it again,
-	 * even if the Guest is run on the same CPU. */
+	/*
+	 * Mark that the GDT changed so the core knows it has to copy it again,
+	 * even if the Guest is run on the same CPU.
+	 */
 	cpu->changed |= CHANGED_GDT;
 }
 
-/* This is the fast-track version for just changing the three TLS entries.
+/*
+ * This is the fast-track version for just changing the three TLS entries.
  * Remember that this happens on every context switch, so it's worth
  * optimizing.  But wouldn't it be neater to have a single hypercall to cover
- * both cases? */
+ * both cases?
+ */
 void guest_load_tls(struct lg_cpu *cpu, unsigned long gtls)
 {
 	struct desc_struct *tls = &cpu->arch.gdt[GDT_ENTRY_TLS_MIN];
@@ -175,7 +208,6 @@ void guest_load_tls(struct lg_cpu *cpu, unsigned long gtls)
 	/* Note that just the TLS entries have changed. */
 	cpu->changed |= CHANGED_GDT_TLS;
 }
-/*:*/
 
 /*H:660
  * With this, we have finished the Host.
diff --git a/drivers/lguest/x86/core.c b/drivers/lguest/x86/core.c
index eaf722fe309..6ae388849a3 100644
--- a/drivers/lguest/x86/core.c
+++ b/drivers/lguest/x86/core.c
@@ -17,13 +17,15 @@
  * along with this program; if not, write to the Free Software
  * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
  */
-/*P:450 This file contains the x86-specific lguest code.  It used to be all
+/*P:450
+ * This file contains the x86-specific lguest code.  It used to be all
  * mixed in with drivers/lguest/core.c but several foolhardy code slashers
  * wrestled most of the dependencies out to here in preparation for porting
  * lguest to other architectures (see what I mean by foolhardy?).
  *
  * This also contains a couple of non-obvious setup and teardown pieces which
- * were implemented after days of debugging pain. :*/
+ * were implemented after days of debugging pain.
+:*/
 #include <linux/kernel.h>
 #include <linux/start_kernel.h>
 #include <linux/string.h>
@@ -82,25 +84,33 @@ static DEFINE_PER_CPU(struct lg_cpu *, last_cpu);
  */
 static void copy_in_guest_info(struct lg_cpu *cpu, struct lguest_pages *pages)
 {
-	/* Copying all this data can be quite expensive.  We usually run the
+	/*
+	 * Copying all this data can be quite expensive.  We usually run the
 	 * same Guest we ran last time (and that Guest hasn't run anywhere else
 	 * meanwhile).  If that's not the case, we pretend everything in the
-	 * Guest has changed. */
+	 * Guest has changed.
+	 */
 	if (__get_cpu_var(last_cpu) != cpu || cpu->last_pages != pages) {
 		__get_cpu_var(last_cpu) = cpu;
 		cpu->last_pages = pages;
 		cpu->changed = CHANGED_ALL;
 	}
 
-	/* These copies are pretty cheap, so we do them unconditionally: */
-	/* Save the current Host top-level page directory. */
+	/*
+	 * These copies are pretty cheap, so we do them unconditionally: */
+	/* Save the current Host top-level page directory.
+	 */
 	pages->state.host_cr3 = __pa(current->mm->pgd);
-	/* Set up the Guest's page tables to see this CPU's pages (and no
-	 * other CPU's pages). */
+	/*
+	 * Set up the Guest's page tables to see this CPU's pages (and no
+	 * other CPU's pages).
+	 */
 	map_switcher_in_guest(cpu, pages);
-	/* Set up the two "TSS" members which tell the CPU what stack to use
+	/*
+	 * Set up the two "TSS" members which tell the CPU what stack to use
 	 * for traps which do directly into the Guest (ie. traps at privilege
-	 * level 1). */
+	 * level 1).
+	 */
 	pages->state.guest_tss.sp1 = cpu->esp1;
 	pages->state.guest_tss.ss1 = cpu->ss1;
 
@@ -125,97 +135,126 @@ static void run_guest_once(struct lg_cpu *cpu, struct lguest_pages *pages)
 	/* This is a dummy value we need for GCC's sake. */
 	unsigned int clobber;
 
-	/* Copy the guest-specific information into this CPU's "struct
-	 * lguest_pages". */
+	/*
+	 * Copy the guest-specific information into this CPU's "struct
+	 * lguest_pages".
+	 */
 	copy_in_guest_info(cpu, pages);
 
-	/* Set the trap number to 256 (impossible value).  If we fault while
+	/*
+	 * Set the trap number to 256 (impossible value).  If we fault while
 	 * switching to the Guest (bad segment registers or bug), this will
-	 * cause us to abort the Guest. */
+	 * cause us to abort the Guest.
+	 */
 	cpu->regs->trapnum = 256;
 
-	/* Now: we push the "eflags" register on the stack, then do an "lcall".
+	/*
+	 * Now: we push the "eflags" register on the stack, then do an "lcall".
 	 * This is how we change from using the kernel code segment to using
 	 * the dedicated lguest code segment, as well as jumping into the
 	 * Switcher.
 	 *
 	 * The lcall also pushes the old code segment (KERNEL_CS) onto the
 	 * stack, then the address of this call.  This stack layout happens to
-	 * exactly match the stack layout created by an interrupt... */
+	 * exactly match the stack layout created by an interrupt...
+	 */
 	asm volatile("pushf; lcall *lguest_entry"
-		     /* This is how we tell GCC that %eax ("a") and %ebx ("b")
-		      * are changed by this routine.  The "=" means output. */
+		     /*
+		      * This is how we tell GCC that %eax ("a") and %ebx ("b")
+		      * are changed by this routine.  The "=" means output.
+		      */
 		     : "=a"(clobber), "=b"(clobber)
-		     /* %eax contains the pages pointer.  ("0" refers to the
+		     /*
+		      * %eax contains the pages pointer.  ("0" refers to the
 		      * 0-th argument above, ie "a").  %ebx contains the
 		      * physical address of the Guest's top-level page
-		      * directory. */
+		      * directory.
+		      */
 		     : "0"(pages), "1"(__pa(cpu->lg->pgdirs[cpu->cpu_pgd].pgdir))
-		     /* We tell gcc that all these registers could change,
+		     /*
+		      * We tell gcc that all these registers could change,
 		      * which means we don't have to save and restore them in
-		      * the Switcher. */
+		      * the Switcher.
+		      */
 		     : "memory", "%edx", "%ecx", "%edi", "%esi");
 }
 /*:*/
 
-/*M:002 There are hooks in the scheduler which we can register to tell when we
+/*M:002
+ * There are hooks in the scheduler which we can register to tell when we
  * get kicked off the CPU (preempt_notifier_register()).  This would allow us
  * to lazily disable SYSENTER which would regain some performance, and should
  * also simplify copy_in_guest_info().  Note that we'd still need to restore
  * things when we exit to Launcher userspace, but that's fairly easy.
  *
- * We could also try using this hooks for PGE, but that might be too expensive.
+ * We could also try using these hooks for PGE, but that might be too expensive.
  *
- * The hooks were designed for KVM, but we can also put them to good use. :*/
+ * The hooks were designed for KVM, but we can also put them to good use.
+:*/
 
-/*H:040 This is the i386-specific code to setup and run the Guest.  Interrupts
- * are disabled: we own the CPU. */
+/*H:040
+ * This is the i386-specific code to setup and run the Guest.  Interrupts
+ * are disabled: we own the CPU.
+ */
 void lguest_arch_run_guest(struct lg_cpu *cpu)
 {
-	/* Remember the awfully-named TS bit?  If the Guest has asked to set it
+	/*
+	 * Remember the awfully-named TS bit?  If the Guest has asked to set it
 	 * we set it now, so we can trap and pass that trap to the Guest if it
-	 * uses the FPU. */
+	 * uses the FPU.
+	 */
 	if (cpu->ts)
 		unlazy_fpu(current);
 
-	/* SYSENTER is an optimized way of doing system calls.  We can't allow
+	/*
+	 * SYSENTER is an optimized way of doing system calls.  We can't allow
 	 * it because it always jumps to privilege level 0.  A normal Guest
 	 * won't try it because we don't advertise it in CPUID, but a malicious
 	 * Guest (or malicious Guest userspace program) could, so we tell the
-	 * CPU to disable it before running the Guest. */
+	 * CPU to disable it before running the Guest.
+	 */
 	if (boot_cpu_has(X86_FEATURE_SEP))
 		wrmsr(MSR_IA32_SYSENTER_CS, 0, 0);
 
-	/* Now we actually run the Guest.  It will return when something
+	/*
+	 * Now we actually run the Guest.  It will return when something
 	 * interesting happens, and we can examine its registers to see what it
-	 * was doing. */
+	 * was doing.
+	 */
 	run_guest_once(cpu, lguest_pages(raw_smp_processor_id()));
 
-	/* Note that the "regs" structure contains two extra entries which are
+	/*
+	 * Note that the "regs" structure contains two extra entries which are
 	 * not really registers: a trap number which says what interrupt or
 	 * trap made the switcher code come back, and an error code which some
-	 * traps set.  */
+	 * traps set.
+	 */
 
 	 /* Restore SYSENTER if it's supposed to be on. */
 	 if (boot_cpu_has(X86_FEATURE_SEP))
 		wrmsr(MSR_IA32_SYSENTER_CS, __KERNEL_CS, 0);
 
-	/* If the Guest page faulted, then the cr2 register will tell us the
+	/*
+	 * If the Guest page faulted, then the cr2 register will tell us the
 	 * bad virtual address.  We have to grab this now, because once we
 	 * re-enable interrupts an interrupt could fault and thus overwrite
-	 * cr2, or we could even move off to a different CPU. */
+	 * cr2, or we could even move off to a different CPU.
+	 */
 	if (cpu->regs->trapnum == 14)
 		cpu->arch.last_pagefault = read_cr2();
-	/* Similarly, if we took a trap because the Guest used the FPU,
+	/*
+	 * Similarly, if we took a trap because the Guest used the FPU,
 	 * we have to restore the FPU it expects to see.
 	 * math_state_restore() may sleep and we may even move off to
 	 * a different CPU. So all the critical stuff should be done
-	 * before this.  */
+	 * before this.
+	 */
 	else if (cpu->regs->trapnum == 7)
 		math_state_restore();
 }
 
-/*H:130 Now we've examined the hypercall code; our Guest can make requests.
+/*H:130
+ * Now we've examined the hypercall code; our Guest can make requests.
  * Our Guest is usually so well behaved; it never tries to do things it isn't
  * allowed to, and uses hypercalls instead.  Unfortunately, Linux's paravirtual
  * infrastructure isn't quite complete, because it doesn't contain replacements
@@ -225,26 +264,33 @@ void lguest_arch_run_guest(struct lg_cpu *cpu)
  *
  * When the Guest uses one of these instructions, we get a trap (General
  * Protection Fault) and come here.  We see if it's one of those troublesome
- * instructions and skip over it.  We return true if we did. */
+ * instructions and skip over it.  We return true if we did.
+ */
 static int emulate_insn(struct lg_cpu *cpu)
 {
 	u8 insn;
 	unsigned int insnlen = 0, in = 0, shift = 0;
-	/* The eip contains the *virtual* address of the Guest's instruction:
-	 * guest_pa just subtracts the Guest's page_offset. */
+	/*
+	 * The eip contains the *virtual* address of the Guest's instruction:
+	 * guest_pa just subtracts the Guest's page_offset.
+	 */
 	unsigned long physaddr = guest_pa(cpu, cpu->regs->eip);
 
-	/* This must be the Guest kernel trying to do something, not userspace!
+	/*
+	 * This must be the Guest kernel trying to do something, not userspace!
 	 * The bottom two bits of the CS segment register are the privilege
-	 * level. */
+	 * level.
+	 */
 	if ((cpu->regs->cs & 3) != GUEST_PL)
 		return 0;
 
 	/* Decoding x86 instructions is icky. */
 	insn = lgread(cpu, physaddr, u8);
 
-	/* 0x66 is an "operand prefix".  It means it's using the upper 16 bits
-	   of the eax register. */
+	/*
+	 * 0x66 is an "operand prefix".  It means it's using the upper 16 bits
+	 * of the eax register.
+	 */
 	if (insn == 0x66) {
 		shift = 16;
 		/* The instruction is 1 byte so far, read the next byte. */
@@ -252,8 +298,10 @@ static int emulate_insn(struct lg_cpu *cpu)
 		insn = lgread(cpu, physaddr + insnlen, u8);
 	}
 
-	/* We can ignore the lower bit for the moment and decode the 4 opcodes
-	 * we need to emulate. */
+	/*
+	 * We can ignore the lower bit for the moment and decode the 4 opcodes
+	 * we need to emulate.
+	 */
 	switch (insn & 0xFE) {
 	case 0xE4: /* in     <next byte>,%al */
 		insnlen += 2;
@@ -274,9 +322,11 @@ static int emulate_insn(struct lg_cpu *cpu)
 		return 0;
 	}
 
-	/* If it was an "IN" instruction, they expect the result to be read
+	/*
+	 * If it was an "IN" instruction, they expect the result to be read
 	 * into %eax, so we change %eax.  We always return all-ones, which
-	 * traditionally means "there's nothing there". */
+	 * traditionally means "there's nothing there".
+	 */
 	if (in) {
 		/* Lower bit tells is whether it's a 16 or 32 bit access */
 		if (insn & 0x1)
@@ -290,7 +340,8 @@ static int emulate_insn(struct lg_cpu *cpu)
 	return 1;
 }
 
-/* Our hypercalls mechanism used to be based on direct software interrupts.
+/*
+ * Our hypercalls mechanism used to be based on direct software interrupts.
  * After Anthony's "Refactor hypercall infrastructure" kvm patch, we decided to
  * change over to using kvm hypercalls.
  *
@@ -318,16 +369,20 @@ static int emulate_insn(struct lg_cpu *cpu)
  */
 static void rewrite_hypercall(struct lg_cpu *cpu)
 {
-	/* This are the opcodes we use to patch the Guest.  The opcode for "int
+	/*
+	 * This are the opcodes we use to patch the Guest.  The opcode for "int
 	 * $0x1f" is "0xcd 0x1f" but vmcall instruction is 3 bytes long, so we
-	 * complete the sequence with a NOP (0x90). */
+	 * complete the sequence with a NOP (0x90).
+	 */
 	u8 insn[3] = {0xcd, 0x1f, 0x90};
 
 	__lgwrite(cpu, guest_pa(cpu, cpu->regs->eip), insn, sizeof(insn));
-	/* The above write might have caused a copy of that page to be made
+	/*
+	 * The above write might have caused a copy of that page to be made
 	 * (if it was read-only).  We need to make sure the Guest has
 	 * up-to-date pagetables.  As this doesn't happen often, we can just
-	 * drop them all. */
+	 * drop them all.
+	 */
 	guest_pagetable_clear_all(cpu);
 }
 
@@ -335,9 +390,11 @@ static bool is_hypercall(struct lg_cpu *cpu)
 {
 	u8 insn[3];
 
-	/* This must be the Guest kernel trying to do something.
+	/*
+	 * This must be the Guest kernel trying to do something.
 	 * The bottom two bits of the CS segment register are the privilege
-	 * level. */
+	 * level.
+	 */
 	if ((cpu->regs->cs & 3) != GUEST_PL)
 		return false;
 
@@ -351,86 +408,105 @@ void lguest_arch_handle_trap(struct lg_cpu *cpu)
 {
 	switch (cpu->regs->trapnum) {
 	case 13: /* We've intercepted a General Protection Fault. */
-		/* Check if this was one of those annoying IN or OUT
+		/*
+		 * Check if this was one of those annoying IN or OUT
 		 * instructions which we need to emulate.  If so, we just go
-		 * back into the Guest after we've done it. */
+		 * back into the Guest after we've done it.
+		 */
 		if (cpu->regs->errcode == 0) {
 			if (emulate_insn(cpu))
 				return;
 		}
-		/* If KVM is active, the vmcall instruction triggers a
-		 * General Protection Fault.  Normally it triggers an
-		 * invalid opcode fault (6): */
+		/*
+		 * If KVM is active, the vmcall instruction triggers a General
+		 * Protection Fault.  Normally it triggers an invalid opcode
+		 * fault (6):
+		 */
 	case 6:
-		/* We need to check if ring == GUEST_PL and
-		 * faulting instruction == vmcall. */
+		/*
+		 * We need to check if ring == GUEST_PL and faulting
+		 * instruction == vmcall.
+		 */
 		if (is_hypercall(cpu)) {
 			rewrite_hypercall(cpu);
 			return;
 		}
 		break;
 	case 14: /* We've intercepted a Page Fault. */
-		/* The Guest accessed a virtual address that wasn't mapped.
+		/*
+		 * The Guest accessed a virtual address that wasn't mapped.
 		 * This happens a lot: we don't actually set up most of the page
 		 * tables for the Guest at all when we start: as it runs it asks
 		 * for more and more, and we set them up as required. In this
 		 * case, we don't even tell the Guest that the fault happened.
 		 *
 		 * The errcode tells whether this was a read or a write, and
-		 * whether kernel or userspace code. */
+		 * whether kernel or userspace code.
+		 */
 		if (demand_page(cpu, cpu->arch.last_pagefault,
 				cpu->regs->errcode))
 			return;
 
-		/* OK, it's really not there (or not OK): the Guest needs to
+		/*
+		 * OK, it's really not there (or not OK): the Guest needs to
 		 * know.  We write out the cr2 value so it knows where the
 		 * fault occurred.
 		 *
 		 * Note that if the Guest were really messed up, this could
 		 * happen before it's done the LHCALL_LGUEST_INIT hypercall, so
-		 * lg->lguest_data could be NULL */
+		 * lg->lguest_data could be NULL
+		 */
 		if (cpu->lg->lguest_data &&
 		    put_user(cpu->arch.last_pagefault,
 			     &cpu->lg->lguest_data->cr2))
 			kill_guest(cpu, "Writing cr2");
 		break;
 	case 7: /* We've intercepted a Device Not Available fault. */
-		/* If the Guest doesn't want to know, we already restored the
-		 * Floating Point Unit, so we just continue without telling
-		 * it. */
+		/*
+		 * If the Guest doesn't want to know, we already restored the
+		 * Floating Point Unit, so we just continue without telling it.
+		 */
 		if (!cpu->ts)
 			return;
 		break;
 	case 32 ... 255:
-		/* These values mean a real interrupt occurred, in which case
+		/*
+		 * These values mean a real interrupt occurred, in which case
 		 * the Host handler has already been run. We just do a
 		 * friendly check if another process should now be run, then
-		 * return to run the Guest again */
+		 * return to run the Guest again
+		 */
 		cond_resched();
 		return;
 	case LGUEST_TRAP_ENTRY:
-		/* Our 'struct hcall_args' maps directly over our regs: we set
-		 * up the pointer now to indicate a hypercall is pending. */
+		/*
+		 * Our 'struct hcall_args' maps directly over our regs: we set
+		 * up the pointer now to indicate a hypercall is pending.
+		 */
 		cpu->hcall = (struct hcall_args *)cpu->regs;
 		return;
 	}
 
 	/* We didn't handle the trap, so it needs to go to the Guest. */
 	if (!deliver_trap(cpu, cpu->regs->trapnum))
-		/* If the Guest doesn't have a handler (either it hasn't
+		/*
+		 * If the Guest doesn't have a handler (either it hasn't
 		 * registered any yet, or it's one of the faults we don't let
-		 * it handle), it dies with this cryptic error message. */
+		 * it handle), it dies with this cryptic error message.
+		 */
 		kill_guest(cpu, "unhandled trap %li at %#lx (%#lx)",
 			   cpu->regs->trapnum, cpu->regs->eip,
 			   cpu->regs->trapnum == 14 ? cpu->arch.last_pagefault
 			   : cpu->regs->errcode);
 }
 
-/* Now we can look at each of the routines this calls, in increasing order of
+/*
+ * Now we can look at each of the routines this calls, in increasing order of
  * complexity: do_hypercalls(), emulate_insn(), maybe_do_interrupt(),
  * deliver_trap() and demand_page().  After all those, we'll be ready to
  * examine the Switcher, and our philosophical understanding of the Host/Guest
- * duality will be complete. :*/
+ * duality will be complete.
+:*/
 static void adjust_pge(void *on)
 {
 	if (on)
@@ -439,13 +515,16 @@ static void adjust_pge(void *on)
 		write_cr4(read_cr4() & ~X86_CR4_PGE);
 }
 
-/*H:020 Now the Switcher is mapped and every thing else is ready, we need to do
- * some more i386-specific initialization. */
+/*H:020
+ * Now the Switcher is mapped and every thing else is ready, we need to do
+ * some more i386-specific initialization.
+ */
 void __init lguest_arch_host_init(void)
 {
 	int i;
 
-	/* Most of the i386/switcher.S doesn't care that it's been moved; on
+	/*
+	 * Most of the i386/switcher.S doesn't care that it's been moved; on
 	 * Intel, jumps are relative, and it doesn't access any references to
 	 * external code or data.
 	 *
@@ -453,7 +532,8 @@ void __init lguest_arch_host_init(void)
 	 * addresses are placed in a table (default_idt_entries), so we need to
 	 * update the table with the new addresses.  switcher_offset() is a
 	 * convenience function which returns the distance between the
-	 * compiled-in switcher code and the high-mapped copy we just made. */
+	 * compiled-in switcher code and the high-mapped copy we just made.
+	 */
 	for (i = 0; i < IDT_ENTRIES; i++)
 		default_idt_entries[i] += switcher_offset();
 
@@ -468,63 +548,81 @@ void __init lguest_arch_host_init(void)
 	for_each_possible_cpu(i) {
 		/* lguest_pages() returns this CPU's two pages. */
 		struct lguest_pages *pages = lguest_pages(i);
-		/* This is a convenience pointer to make the code fit one
-		 * statement to a line. */
+		/* This is a convenience pointer to make the code neater. */
 		struct lguest_ro_state *state = &pages->state;
 
-		/* The Global Descriptor Table: the Host has a different one
+		/*
+		 * The Global Descriptor Table: the Host has a different one
 		 * for each CPU.  We keep a descriptor for the GDT which says
 		 * where it is and how big it is (the size is actually the last
-		 * byte, not the size, hence the "-1"). */
+		 * byte, not the size, hence the "-1").
+		 */
 		state->host_gdt_desc.size = GDT_SIZE-1;
 		state->host_gdt_desc.address = (long)get_cpu_gdt_table(i);
 
-		/* All CPUs on the Host use the same Interrupt Descriptor
+		/*
+		 * All CPUs on the Host use the same Interrupt Descriptor
 		 * Table, so we just use store_idt(), which gets this CPU's IDT
-		 * descriptor. */
+		 * descriptor.
+		 */
 		store_idt(&state->host_idt_desc);
 
-		/* The descriptors for the Guest's GDT and IDT can be filled
+		/*
+		 * The descriptors for the Guest's GDT and IDT can be filled
 		 * out now, too.  We copy the GDT & IDT into ->guest_gdt and
-		 * ->guest_idt before actually running the Guest. */
+		 * ->guest_idt before actually running the Guest.
+		 */
 		state->guest_idt_desc.size = sizeof(state->guest_idt)-1;
 		state->guest_idt_desc.address = (long)&state->guest_idt;
 		state->guest_gdt_desc.size = sizeof(state->guest_gdt)-1;
 		state->guest_gdt_desc.address = (long)&state->guest_gdt;
 
-		/* We know where we want the stack to be when the Guest enters
+		/*
+		 * We know where we want the stack to be when the Guest enters
 		 * the Switcher: in pages->regs.  The stack grows upwards, so
-		 * we start it at the end of that structure. */
+		 * we start it at the end of that structure.
+		 */
 		state->guest_tss.sp0 = (long)(&pages->regs + 1);
-		/* And this is the GDT entry to use for the stack: we keep a
-		 * couple of special LGUEST entries. */
+		/*
+		 * And this is the GDT entry to use for the stack: we keep a
+		 * couple of special LGUEST entries.
+		 */
 		state->guest_tss.ss0 = LGUEST_DS;
 
-		/* x86 can have a finegrained bitmap which indicates what I/O
+		/*
+		 * x86 can have a finegrained bitmap which indicates what I/O
 		 * ports the process can use.  We set it to the end of our
-		 * structure, meaning "none". */
+		 * structure, meaning "none".
+		 */
 		state->guest_tss.io_bitmap_base = sizeof(state->guest_tss);
 
-		/* Some GDT entries are the same across all Guests, so we can
-		 * set them up now. */
+		/*
+		 * Some GDT entries are the same across all Guests, so we can
+		 * set them up now.
+		 */
 		setup_default_gdt_entries(state);
 		/* Most IDT entries are the same for all Guests, too.*/
 		setup_default_idt_entries(state, default_idt_entries);
 
-		/* The Host needs to be able to use the LGUEST segments on this
-		 * CPU, too, so put them in the Host GDT. */
+		/*
+		 * The Host needs to be able to use the LGUEST segments on this
+		 * CPU, too, so put them in the Host GDT.
+		 */
 		get_cpu_gdt_table(i)[GDT_ENTRY_LGUEST_CS] = FULL_EXEC_SEGMENT;
 		get_cpu_gdt_table(i)[GDT_ENTRY_LGUEST_DS] = FULL_SEGMENT;
 	}
 
-	/* In the Switcher, we want the %cs segment register to use the
+	/*
+	 * In the Switcher, we want the %cs segment register to use the
 	 * LGUEST_CS GDT entry: we've put that in the Host and Guest GDTs, so
 	 * it will be undisturbed when we switch.  To change %cs and jump we
-	 * need this structure to feed to Intel's "lcall" instruction. */
+	 * need this structure to feed to Intel's "lcall" instruction.
+	 */
 	lguest_entry.offset = (long)switch_to_guest + switcher_offset();
 	lguest_entry.segment = LGUEST_CS;
 
-	/* Finally, we need to turn off "Page Global Enable".  PGE is an
+	/*
+	 * Finally, we need to turn off "Page Global Enable".  PGE is an
 	 * optimization where page table entries are specially marked to show
 	 * they never change.  The Host kernel marks all the kernel pages this
 	 * way because it's always present, even when userspace is running.
@@ -534,16 +632,21 @@ void __init lguest_arch_host_init(void)
 	 * you'll get really weird bugs that you'll chase for two days.
 	 *
 	 * I used to turn PGE off every time we switched to the Guest and back
-	 * on when we return, but that slowed the Switcher down noticibly. */
+	 * on when we return, but that slowed the Switcher down noticibly.
+	 */
 
-	/* We don't need the complexity of CPUs coming and going while we're
-	 * doing this. */
+	/*
+	 * We don't need the complexity of CPUs coming and going while we're
+	 * doing this.
+	 */
 	get_online_cpus();
 	if (cpu_has_pge) { /* We have a broader idea of "global". */
 		/* Remember that this was originally set (for cleanup). */
 		cpu_had_pge = 1;
-		/* adjust_pge is a helper function which sets or unsets the PGE
-		 * bit on its CPU, depending on the argument (0 == unset). */
+		/*
+		 * adjust_pge is a helper function which sets or unsets the PGE
+		 * bit on its CPU, depending on the argument (0 == unset).
+		 */
 		on_each_cpu(adjust_pge, (void *)0, 1);
 		/* Turn off the feature in the global feature set. */
 		clear_cpu_cap(&boot_cpu_data, X86_FEATURE_PGE);
@@ -590,26 +693,32 @@ int lguest_arch_init_hypercalls(struct lg_cpu *cpu)
 {
 	u32 tsc_speed;
 
-	/* The pointer to the Guest's "struct lguest_data" is the only argument.
-	 * We check that address now. */
+	/*
+	 * The pointer to the Guest's "struct lguest_data" is the only argument.
+	 * We check that address now.
+	 */
 	if (!lguest_address_ok(cpu->lg, cpu->hcall->arg1,
 			       sizeof(*cpu->lg->lguest_data)))
 		return -EFAULT;
 
-	/* Having checked it, we simply set lg->lguest_data to point straight
+	/*
+	 * Having checked it, we simply set lg->lguest_data to point straight
 	 * into the Launcher's memory at the right place and then use
 	 * copy_to_user/from_user from now on, instead of lgread/write.  I put
 	 * this in to show that I'm not immune to writing stupid
-	 * optimizations. */
+	 * optimizations.
+	 */
 	cpu->lg->lguest_data = cpu->lg->mem_base + cpu->hcall->arg1;
 
-	/* We insist that the Time Stamp Counter exist and doesn't change with
+	/*
+	 * We insist that the Time Stamp Counter exist and doesn't change with
 	 * cpu frequency.  Some devious chip manufacturers decided that TSC
 	 * changes could be handled in software.  I decided that time going
 	 * backwards might be good for benchmarks, but it's bad for users.
 	 *
 	 * We also insist that the TSC be stable: the kernel detects unreliable
-	 * TSCs for its own purposes, and we use that here. */
+	 * TSCs for its own purposes, and we use that here.
+	 */
 	if (boot_cpu_has(X86_FEATURE_CONSTANT_TSC) && !check_tsc_unstable())
 		tsc_speed = tsc_khz;
 	else
@@ -625,38 +734,47 @@ int lguest_arch_init_hypercalls(struct lg_cpu *cpu)
 }
 /*:*/
 
-/*L:030 lguest_arch_setup_regs()
+/*L:030
+ * lguest_arch_setup_regs()
  *
  * Most of the Guest's registers are left alone: we used get_zeroed_page() to
- * allocate the structure, so they will be 0. */
+ * allocate the structure, so they will be 0.
+ */
 void lguest_arch_setup_regs(struct lg_cpu *cpu, unsigned long start)
 {
 	struct lguest_regs *regs = cpu->regs;
 
-	/* There are four "segment" registers which the Guest needs to boot:
+	/*
+	 * There are four "segment" registers which the Guest needs to boot:
 	 * The "code segment" register (cs) refers to the kernel code segment
 	 * __KERNEL_CS, and the "data", "extra" and "stack" segment registers
 	 * refer to the kernel data segment __KERNEL_DS.
 	 *
 	 * The privilege level is packed into the lower bits.  The Guest runs
-	 * at privilege level 1 (GUEST_PL).*/
+	 * at privilege level 1 (GUEST_PL).
+	 */
 	regs->ds = regs->es = regs->ss = __KERNEL_DS|GUEST_PL;
 	regs->cs = __KERNEL_CS|GUEST_PL;
 
-	/* The "eflags" register contains miscellaneous flags.  Bit 1 (0x002)
+	/*
+	 * The "eflags" register contains miscellaneous flags.  Bit 1 (0x002)
 	 * is supposed to always be "1".  Bit 9 (0x200) controls whether
 	 * interrupts are enabled.  We always leave interrupts enabled while
-	 * running the Guest. */
+	 * running the Guest.
+	 */
 	regs->eflags = X86_EFLAGS_IF | 0x2;
 
-	/* The "Extended Instruction Pointer" register says where the Guest is
-	 * running. */
+	/*
+	 * The "Extended Instruction Pointer" register says where the Guest is
+	 * running.
+	 */
 	regs->eip = start;
 
-	/* %esi points to our boot information, at physical address 0, so don't
-	 * touch it. */
+	/*
+	 * %esi points to our boot information, at physical address 0, so don't
+	 * touch it.
+	 */
 
-	/* There are a couple of GDT entries the Guest expects when first
-	 * booting. */
+	/* There are a couple of GDT entries the Guest expects at boot. */
 	setup_guest_gdt(cpu);
 }
diff --git a/drivers/lguest/x86/switcher_32.S b/drivers/lguest/x86/switcher_32.S
index 3fc15318a80..40634b0db9f 100644
--- a/drivers/lguest/x86/switcher_32.S
+++ b/drivers/lguest/x86/switcher_32.S
@@ -1,12 +1,15 @@
-/*P:900 This is the Switcher: code which sits at 0xFFC00000 astride both the
- * Host and Guest to do the low-level Guest<->Host switch.  It is as simple as
- * it can be made, but it's naturally very specific to x86.
+/*P:900
+ * This is the Switcher: code which sits at 0xFFC00000 (or 0xFFE00000) astride
+ * both the Host and Guest to do the low-level Guest<->Host switch.  It is as
+ * simple as it can be made, but it's naturally very specific to x86.
  *
  * You have now completed Preparation.  If this has whet your appetite; if you
  * are feeling invigorated and refreshed then the next, more challenging stage
- * can be found in "make Guest". :*/
+ * can be found in "make Guest".
+ :*/
 
-/*M:012 Lguest is meant to be simple: my rule of thumb is that 1% more LOC must
+/*M:012
+ * Lguest is meant to be simple: my rule of thumb is that 1% more LOC must
  * gain at least 1% more performance.  Since neither LOC nor performance can be
  * measured beforehand, it generally means implementing a feature then deciding
  * if it's worth it.  And once it's implemented, who can say no?
@@ -31,11 +34,14 @@
  * Host (which is actually really easy).
  *
  * Two questions remain.  Would the performance gain outweigh the complexity?
- * And who would write the verse documenting it? :*/
+ * And who would write the verse documenting it?
+:*/
 
-/*M:011 Lguest64 handles NMI.  This gave me NMI envy (until I looked at their
+/*M:011
+ * Lguest64 handles NMI.  This gave me NMI envy (until I looked at their
  * code).  It's worth doing though, since it would let us use oprofile in the
- * Host when a Guest is running. :*/
+ * Host when a Guest is running.
+:*/
 
 /*S:100
  * Welcome to the Switcher itself!
diff --git a/drivers/md/linear.c b/drivers/md/linear.c
index 5810fa906af..5fe39c2a3d2 100644
--- a/drivers/md/linear.c
+++ b/drivers/md/linear.c
@@ -220,6 +220,7 @@ static int linear_run (mddev_t *mddev)
 	mddev->queue->unplug_fn = linear_unplug;
 	mddev->queue->backing_dev_info.congested_fn = linear_congested;
 	mddev->queue->backing_dev_info.congested_data = mddev;
+	md_integrity_register(mddev);
 	return 0;
 }
 
@@ -256,6 +257,7 @@ static int linear_add(mddev_t *mddev, mdk_rdev_t *rdev)
 	rcu_assign_pointer(mddev->private, newconf);
 	md_set_array_sectors(mddev, linear_size(mddev, 0, 0));
 	set_capacity(mddev->gendisk, mddev->array_sectors);
+	revalidate_disk(mddev->gendisk);
 	call_rcu(&oldconf->rcu, free_conf);
 	return 0;
 }
diff --git a/drivers/md/md.c b/drivers/md/md.c
index d4351ff0849..5b98bea4ff9 100644
--- a/drivers/md/md.c
+++ b/drivers/md/md.c
@@ -1308,7 +1308,12 @@ static int super_1_validate(mddev_t *mddev, mdk_rdev_t *rdev)
 	}
 	if (mddev->level != LEVEL_MULTIPATH) {
 		int role;
-		role = le16_to_cpu(sb->dev_roles[rdev->desc_nr]);
+		if (rdev->desc_nr < 0 ||
+		    rdev->desc_nr >= le32_to_cpu(sb->max_dev)) {
+			role = 0xffff;
+			rdev->desc_nr = -1;
+		} else
+			role = le16_to_cpu(sb->dev_roles[rdev->desc_nr]);
 		switch(role) {
 		case 0xffff: /* spare */
 			break;
@@ -1394,8 +1399,14 @@ static void super_1_sync(mddev_t *mddev, mdk_rdev_t *rdev)
 		if (rdev2->desc_nr+1 > max_dev)
 			max_dev = rdev2->desc_nr+1;
 
-	if (max_dev > le32_to_cpu(sb->max_dev))
+	if (max_dev > le32_to_cpu(sb->max_dev)) {
+		int bmask;
 		sb->max_dev = cpu_to_le32(max_dev);
+		rdev->sb_size = max_dev * 2 + 256;
+		bmask = queue_logical_block_size(rdev->bdev->bd_disk->queue)-1;
+		if (rdev->sb_size & bmask)
+			rdev->sb_size = (rdev->sb_size | bmask) + 1;
+	}
 	for (i=0; i<max_dev;i++)
 		sb->dev_roles[i] = cpu_to_le16(0xfffe);
 	
@@ -1487,37 +1498,76 @@ static int match_mddev_units(mddev_t *mddev1, mddev_t *mddev2)
 
 static LIST_HEAD(pending_raid_disks);
 
-static void md_integrity_check(mdk_rdev_t *rdev, mddev_t *mddev)
+/*
+ * Try to register data integrity profile for an mddev
+ *
+ * This is called when an array is started and after a disk has been kicked
+ * from the array. It only succeeds if all working and active component devices
+ * are integrity capable with matching profiles.
+ */
+int md_integrity_register(mddev_t *mddev)
+{
+	mdk_rdev_t *rdev, *reference = NULL;
+
+	if (list_empty(&mddev->disks))
+		return 0; /* nothing to do */
+	if (blk_get_integrity(mddev->gendisk))
+		return 0; /* already registered */
+	list_for_each_entry(rdev, &mddev->disks, same_set) {
+		/* skip spares and non-functional disks */
+		if (test_bit(Faulty, &rdev->flags))
+			continue;
+		if (rdev->raid_disk < 0)
+			continue;
+		/*
+		 * If at least one rdev is not integrity capable, we can not
+		 * enable data integrity for the md device.
+		 */
+		if (!bdev_get_integrity(rdev->bdev))
+			return -EINVAL;
+		if (!reference) {
+			/* Use the first rdev as the reference */
+			reference = rdev;
+			continue;
+		}
+		/* does this rdev's profile match the reference profile? */
+		if (blk_integrity_compare(reference->bdev->bd_disk,
+				rdev->bdev->bd_disk) < 0)
+			return -EINVAL;
+	}
+	/*
+	 * All component devices are integrity capable and have matching
+	 * profiles, register the common profile for the md device.
+	 */
+	if (blk_integrity_register(mddev->gendisk,
+			bdev_get_integrity(reference->bdev)) != 0) {
+		printk(KERN_ERR "md: failed to register integrity for %s\n",
+			mdname(mddev));
+		return -EINVAL;
+	}
+	printk(KERN_NOTICE "md: data integrity on %s enabled\n",
+		mdname(mddev));
+	return 0;
+}
+EXPORT_SYMBOL(md_integrity_register);
+
+/* Disable data integrity if non-capable/non-matching disk is being added */
+void md_integrity_add_rdev(mdk_rdev_t *rdev, mddev_t *mddev)
 {
-	struct mdk_personality *pers = mddev->pers;
-	struct gendisk *disk = mddev->gendisk;
 	struct blk_integrity *bi_rdev = bdev_get_integrity(rdev->bdev);
-	struct blk_integrity *bi_mddev = blk_get_integrity(disk);
+	struct blk_integrity *bi_mddev = blk_get_integrity(mddev->gendisk);
 
-	/* Data integrity passthrough not supported on RAID 4, 5 and 6 */
-	if (pers && pers->level >= 4 && pers->level <= 6)
+	if (!bi_mddev) /* nothing to do */
 		return;
-
-	/* If rdev is integrity capable, register profile for mddev */
-	if (!bi_mddev && bi_rdev) {
-		if (blk_integrity_register(disk, bi_rdev))
-			printk(KERN_ERR "%s: %s Could not register integrity!\n",
-			       __func__, disk->disk_name);
-		else
-			printk(KERN_NOTICE "Enabling data integrity on %s\n",
-			       disk->disk_name);
+	if (rdev->raid_disk < 0) /* skip spares */
 		return;
-	}
-
-	/* Check that mddev and rdev have matching profiles */
-	if (blk_integrity_compare(disk, rdev->bdev->bd_disk) < 0) {
-		printk(KERN_ERR "%s: %s/%s integrity mismatch!\n", __func__,
-		       disk->disk_name, rdev->bdev->bd_disk->disk_name);
-		printk(KERN_NOTICE "Disabling data integrity on %s\n",
-		       disk->disk_name);
-		blk_integrity_unregister(disk);
-	}
+	if (bi_rdev && blk_integrity_compare(mddev->gendisk,
+					     rdev->bdev->bd_disk) >= 0)
+		return;
+	printk(KERN_NOTICE "disabling data integrity on %s\n", mdname(mddev));
+	blk_integrity_unregister(mddev->gendisk);
 }
+EXPORT_SYMBOL(md_integrity_add_rdev);
 
 static int bind_rdev_to_array(mdk_rdev_t * rdev, mddev_t * mddev)
 {
@@ -1591,7 +1641,6 @@ static int bind_rdev_to_array(mdk_rdev_t * rdev, mddev_t * mddev)
 	/* May as well allow recovery to be retried once */
 	mddev->recovery_disabled = 0;
 
-	md_integrity_check(rdev, mddev);
 	return 0;
 
  fail:
@@ -2657,6 +2706,7 @@ level_store(mddev_t *mddev, const char *buf, size_t len)
 	ssize_t rv = len;
 	struct mdk_personality *pers;
 	void *priv;
+	mdk_rdev_t *rdev;
 
 	if (mddev->pers == NULL) {
 		if (len == 0)
@@ -2736,6 +2786,12 @@ level_store(mddev_t *mddev, const char *buf, size_t len)
 	mddev_suspend(mddev);
 	mddev->pers->stop(mddev);
 	module_put(mddev->pers->owner);
+	/* Invalidate devices that are now superfluous */
+	list_for_each_entry(rdev, &mddev->disks, same_set)
+		if (rdev->raid_disk >= mddev->raid_disks) {
+			rdev->raid_disk = -1;
+			clear_bit(In_sync, &rdev->flags);
+		}
 	mddev->pers = pers;
 	mddev->private = priv;
 	strlcpy(mddev->clevel, pers->name, sizeof(mddev->clevel));
@@ -3685,17 +3741,8 @@ array_size_store(mddev_t *mddev, const char *buf, size_t len)
 
 	mddev->array_sectors = sectors;
 	set_capacity(mddev->gendisk, mddev->array_sectors);
-	if (mddev->pers) {
-		struct block_device *bdev = bdget_disk(mddev->gendisk, 0);
-
-		if (bdev) {
-			mutex_lock(&bdev->bd_inode->i_mutex);
-			i_size_write(bdev->bd_inode,
-				     (loff_t)mddev->array_sectors << 9);
-			mutex_unlock(&bdev->bd_inode->i_mutex);
-			bdput(bdev);
-		}
-	}
+	if (mddev->pers)
+		revalidate_disk(mddev->gendisk);
 
 	return len;
 }
@@ -4048,10 +4095,6 @@ static int do_md_run(mddev_t * mddev)
 	}
 	strlcpy(mddev->clevel, pers->name, sizeof(mddev->clevel));
 
-	if (pers->level >= 4 && pers->level <= 6)
-		/* Cannot support integrity (yet) */
-		blk_integrity_unregister(mddev->gendisk);
-
 	if (mddev->reshape_position != MaxSector &&
 	    pers->start_reshape == NULL) {
 		/* This personality cannot handle reshaping... */
@@ -4189,6 +4232,7 @@ static int do_md_run(mddev_t * mddev)
 	md_wakeup_thread(mddev->thread);
 	md_wakeup_thread(mddev->sync_thread); /* possibly kick off a reshape */
 
+	revalidate_disk(mddev->gendisk);
 	mddev->changed = 1;
 	md_new_event(mddev);
 	sysfs_notify_dirent(mddev->sysfs_state);
@@ -5087,18 +5131,8 @@ static int update_size(mddev_t *mddev, sector_t num_sectors)
 			return -ENOSPC;
 	}
 	rv = mddev->pers->resize(mddev, num_sectors);
-	if (!rv) {
-		struct block_device *bdev;
-
-		bdev = bdget_disk(mddev->gendisk, 0);
-		if (bdev) {
-			mutex_lock(&bdev->bd_inode->i_mutex);
-			i_size_write(bdev->bd_inode,
-				     (loff_t)mddev->array_sectors << 9);
-			mutex_unlock(&bdev->bd_inode->i_mutex);
-			bdput(bdev);
-		}
-	}
+	if (!rv)
+		revalidate_disk(mddev->gendisk);
 	return rv;
 }
 
diff --git a/drivers/md/md.h b/drivers/md/md.h
index 9430a110db9..78f03168baf 100644
--- a/drivers/md/md.h
+++ b/drivers/md/md.h
@@ -431,5 +431,7 @@ extern int md_allow_write(mddev_t *mddev);
 extern void md_wait_for_blocked_rdev(mdk_rdev_t *rdev, mddev_t *mddev);
 extern void md_set_array_sectors(mddev_t *mddev, sector_t array_sectors);
 extern int md_check_no_bitmap(mddev_t *mddev);
+extern int md_integrity_register(mddev_t *mddev);
+void md_integrity_add_rdev(mdk_rdev_t *rdev, mddev_t *mddev);
 
 #endif /* _MD_MD_H */
diff --git a/drivers/md/multipath.c b/drivers/md/multipath.c
index 237fe3fd235..7140909f666 100644
--- a/drivers/md/multipath.c
+++ b/drivers/md/multipath.c
@@ -313,6 +313,7 @@ static int multipath_add_disk(mddev_t *mddev, mdk_rdev_t *rdev)
 			set_bit(In_sync, &rdev->flags);
 			rcu_assign_pointer(p->rdev, rdev);
 			err = 0;
+			md_integrity_add_rdev(rdev, mddev);
 			break;
 		}
 
@@ -345,7 +346,9 @@ static int multipath_remove_disk(mddev_t *mddev, int number)
 			/* lost the race, try later */
 			err = -EBUSY;
 			p->rdev = rdev;
+			goto abort;
 		}
+		md_integrity_register(mddev);
 	}
 abort:
 
@@ -519,7 +522,7 @@ static int multipath_run (mddev_t *mddev)
 	mddev->queue->unplug_fn = multipath_unplug;
 	mddev->queue->backing_dev_info.congested_fn = multipath_congested;
 	mddev->queue->backing_dev_info.congested_data = mddev;
-
+	md_integrity_register(mddev);
 	return 0;
 
 out_free_conf:
diff --git a/drivers/md/raid0.c b/drivers/md/raid0.c
index 335f490dcad..898e2bdfee4 100644
--- a/drivers/md/raid0.c
+++ b/drivers/md/raid0.c
@@ -351,6 +351,7 @@ static int raid0_run(mddev_t *mddev)
 
 	blk_queue_merge_bvec(mddev->queue, raid0_mergeable_bvec);
 	dump_zones(mddev);
+	md_integrity_register(mddev);
 	return 0;
 }
 
diff --git a/drivers/md/raid1.c b/drivers/md/raid1.c
index 0569efba0c0..8726fd7ebce 100644
--- a/drivers/md/raid1.c
+++ b/drivers/md/raid1.c
@@ -1144,7 +1144,7 @@ static int raid1_add_disk(mddev_t *mddev, mdk_rdev_t *rdev)
 			rcu_assign_pointer(p->rdev, rdev);
 			break;
 		}
-
+	md_integrity_add_rdev(rdev, mddev);
 	print_conf(conf);
 	return err;
 }
@@ -1178,7 +1178,9 @@ static int raid1_remove_disk(mddev_t *mddev, int number)
 			/* lost the race, try later */
 			err = -EBUSY;
 			p->rdev = rdev;
+			goto abort;
 		}
+		md_integrity_register(mddev);
 	}
 abort:
 
@@ -2067,7 +2069,7 @@ static int run(mddev_t *mddev)
 	mddev->queue->unplug_fn = raid1_unplug;
 	mddev->queue->backing_dev_info.congested_fn = raid1_congested;
 	mddev->queue->backing_dev_info.congested_data = mddev;
-
+	md_integrity_register(mddev);
 	return 0;
 
 out_no_mem:
@@ -2132,6 +2134,7 @@ static int raid1_resize(mddev_t *mddev, sector_t sectors)
 		return -EINVAL;
 	set_capacity(mddev->gendisk, mddev->array_sectors);
 	mddev->changed = 1;
+	revalidate_disk(mddev->gendisk);
 	if (sectors > mddev->dev_sectors &&
 	    mddev->recovery_cp == MaxSector) {
 		mddev->recovery_cp = mddev->dev_sectors;
diff --git a/drivers/md/raid10.c b/drivers/md/raid10.c
index 7298a5e5a18..3d9020cf6f6 100644
--- a/drivers/md/raid10.c
+++ b/drivers/md/raid10.c
@@ -1170,6 +1170,7 @@ static int raid10_add_disk(mddev_t *mddev, mdk_rdev_t *rdev)
 			break;
 		}
 
+	md_integrity_add_rdev(rdev, mddev);
 	print_conf(conf);
 	return err;
 }
@@ -1203,7 +1204,9 @@ static int raid10_remove_disk(mddev_t *mddev, int number)
 			/* lost the race, try later */
 			err = -EBUSY;
 			p->rdev = rdev;
+			goto abort;
 		}
+		md_integrity_register(mddev);
 	}
 abort:
 
@@ -2225,6 +2228,7 @@ static int run(mddev_t *mddev)
 
 	if (conf->near_copies < mddev->raid_disks)
 		blk_queue_merge_bvec(mddev->queue, raid10_mergeable_bvec);
+	md_integrity_register(mddev);
 	return 0;
 
 out_free_conf:
diff --git a/drivers/md/raid5.c b/drivers/md/raid5.c
index 37835538b58..2b521ee67df 100644
--- a/drivers/md/raid5.c
+++ b/drivers/md/raid5.c
@@ -3999,6 +3999,9 @@ static inline sector_t sync_request(mddev_t *mddev, sector_t sector_nr, int *ski
 		return 0;
 	}
 
+	/* Allow raid5_quiesce to complete */
+	wait_event(conf->wait_for_overlap, conf->quiesce != 2);
+
 	if (test_bit(MD_RECOVERY_RESHAPE, &mddev->recovery))
 		return reshape_request(mddev, sector_nr, skipped);
 
@@ -4316,6 +4319,15 @@ raid5_size(mddev_t *mddev, sector_t sectors, int raid_disks)
 	return sectors * (raid_disks - conf->max_degraded);
 }
 
+static void free_conf(raid5_conf_t *conf)
+{
+	shrink_stripes(conf);
+	safe_put_page(conf->spare_page);
+	kfree(conf->disks);
+	kfree(conf->stripe_hashtbl);
+	kfree(conf);
+}
+
 static raid5_conf_t *setup_conf(mddev_t *mddev)
 {
 	raid5_conf_t *conf;
@@ -4447,11 +4459,7 @@ static raid5_conf_t *setup_conf(mddev_t *mddev)
 
  abort:
 	if (conf) {
-		shrink_stripes(conf);
-		safe_put_page(conf->spare_page);
-		kfree(conf->disks);
-		kfree(conf->stripe_hashtbl);
-		kfree(conf);
+		free_conf(conf);
 		return ERR_PTR(-EIO);
 	} else
 		return ERR_PTR(-ENOMEM);
@@ -4629,12 +4637,8 @@ abort:
 	md_unregister_thread(mddev->thread);
 	mddev->thread = NULL;
 	if (conf) {
-		shrink_stripes(conf);
 		print_raid5_conf(conf);
-		safe_put_page(conf->spare_page);
-		kfree(conf->disks);
-		kfree(conf->stripe_hashtbl);
-		kfree(conf);
+		free_conf(conf);
 	}
 	mddev->private = NULL;
 	printk(KERN_ALERT "raid5: failed to run raid set %s\n", mdname(mddev));
@@ -4649,13 +4653,10 @@ static int stop(mddev_t *mddev)
 
 	md_unregister_thread(mddev->thread);
 	mddev->thread = NULL;
-	shrink_stripes(conf);
-	kfree(conf->stripe_hashtbl);
 	mddev->queue->backing_dev_info.congested_fn = NULL;
 	blk_sync_queue(mddev->queue); /* the unplug fn references 'conf'*/
 	sysfs_remove_group(&mddev->kobj, &raid5_attrs_group);
-	kfree(conf->disks);
-	kfree(conf);
+	free_conf(conf);
 	mddev->private = NULL;
 	return 0;
 }
@@ -4857,6 +4858,7 @@ static int raid5_resize(mddev_t *mddev, sector_t sectors)
 		return -EINVAL;
 	set_capacity(mddev->gendisk, mddev->array_sectors);
 	mddev->changed = 1;
+	revalidate_disk(mddev->gendisk);
 	if (sectors > mddev->dev_sectors && mddev->recovery_cp == MaxSector) {
 		mddev->recovery_cp = mddev->dev_sectors;
 		set_bit(MD_RECOVERY_NEEDED, &mddev->recovery);
@@ -5002,7 +5004,7 @@ static int raid5_start_reshape(mddev_t *mddev)
 		spin_unlock_irqrestore(&conf->device_lock, flags);
 	}
 	mddev->raid_disks = conf->raid_disks;
-	mddev->reshape_position = 0;
+	mddev->reshape_position = conf->reshape_progress;
 	set_bit(MD_CHANGE_DEVS, &mddev->flags);
 
 	clear_bit(MD_RECOVERY_SYNC, &mddev->recovery);
@@ -5057,7 +5059,6 @@ static void end_reshape(raid5_conf_t *conf)
  */
 static void raid5_finish_reshape(mddev_t *mddev)
 {
-	struct block_device *bdev;
 	raid5_conf_t *conf = mddev->private;
 
 	if (!test_bit(MD_RECOVERY_INTR, &mddev->recovery)) {
@@ -5066,15 +5067,7 @@ static void raid5_finish_reshape(mddev_t *mddev)
 			md_set_array_sectors(mddev, raid5_size(mddev, 0, 0));
 			set_capacity(mddev->gendisk, mddev->array_sectors);
 			mddev->changed = 1;
-
-			bdev = bdget_disk(mddev->gendisk, 0);
-			if (bdev) {
-				mutex_lock(&bdev->bd_inode->i_mutex);
-				i_size_write(bdev->bd_inode,
-					     (loff_t)mddev->array_sectors << 9);
-				mutex_unlock(&bdev->bd_inode->i_mutex);
-				bdput(bdev);
-			}
+			revalidate_disk(mddev->gendisk);
 		} else {
 			int d;
 			mddev->degraded = conf->raid_disks;
@@ -5106,12 +5099,18 @@ static void raid5_quiesce(mddev_t *mddev, int state)
 
 	case 1: /* stop all writes */
 		spin_lock_irq(&conf->device_lock);
-		conf->quiesce = 1;
+		/* '2' tells resync/reshape to pause so that all
+		 * active stripes can drain
+		 */
+		conf->quiesce = 2;
 		wait_event_lock_irq(conf->wait_for_stripe,
 				    atomic_read(&conf->active_stripes) == 0 &&
 				    atomic_read(&conf->active_aligned_reads) == 0,
 				    conf->device_lock, /* nothing */);
+		conf->quiesce = 1;
 		spin_unlock_irq(&conf->device_lock);
+		/* allow reshape to continue */
+		wake_up(&conf->wait_for_overlap);
 		break;
 
 	case 0: /* re-enable writes */
diff --git a/drivers/mfd/twl4030-irq.c b/drivers/mfd/twl4030-irq.c
index bae61b22501..7d430835655 100644
--- a/drivers/mfd/twl4030-irq.c
+++ b/drivers/mfd/twl4030-irq.c
@@ -180,14 +180,9 @@ static struct completion irq_event;
 static int twl4030_irq_thread(void *data)
 {
 	long irq = (long)data;
-	struct irq_desc *desc = irq_to_desc(irq);
 	static unsigned i2c_errors;
 	static const unsigned max_i2c_errors = 100;
 
-	if (!desc) {
-		pr_err("twl4030: Invalid IRQ: %ld\n", irq);
-		return -EINVAL;
-	}
 
 	current->flags |= PF_NOFREEZE;
 
@@ -240,7 +235,7 @@ static int twl4030_irq_thread(void *data)
 		}
 		local_irq_enable();
 
-		desc->chip->unmask(irq);
+		enable_irq(irq);
 	}
 
 	return 0;
@@ -255,25 +250,13 @@ static int twl4030_irq_thread(void *data)
  * thread.  All we do here is acknowledge and mask the interrupt and wakeup
  * the kernel thread.
  */
-static void handle_twl4030_pih(unsigned int irq, struct irq_desc *desc)
+static irqreturn_t handle_twl4030_pih(int irq, void *devid)
 {
 	/* Acknowledge, clear *AND* mask the interrupt... */
-	desc->chip->ack(irq);
-	complete(&irq_event);
-}
-
-static struct task_struct *start_twl4030_irq_thread(long irq)
-{
-	struct task_struct *thread;
-
-	init_completion(&irq_event);
-	thread = kthread_run(twl4030_irq_thread, (void *)irq, "twl4030-irq");
-	if (!thread)
-		pr_err("twl4030: could not create irq %ld thread!\n", irq);
-
-	return thread;
+	disable_irq_nosync(irq);
+	complete(devid);
+	return IRQ_HANDLED;
 }
-
 /*----------------------------------------------------------------------*/
 
 /*
@@ -734,18 +717,28 @@ int twl_init_irq(int irq_num, unsigned irq_base, unsigned irq_end)
 	}
 
 	/* install an irq handler to demultiplex the TWL4030 interrupt */
-	task = start_twl4030_irq_thread(irq_num);
-	if (!task) {
-		pr_err("twl4030: irq thread FAIL\n");
-		status = -ESRCH;
-		goto fail;
-	}
 
-	set_irq_data(irq_num, task);
-	set_irq_chained_handler(irq_num, handle_twl4030_pih);
 
-	return status;
+	init_completion(&irq_event);
 
+	status = request_irq(irq_num, handle_twl4030_pih, IRQF_DISABLED,
+				"TWL4030-PIH", &irq_event);
+	if (status < 0) {
+		pr_err("twl4030: could not claim irq%d: %d\n", irq_num, status);
+		goto fail_rqirq;
+	}
+
+	task = kthread_run(twl4030_irq_thread, (void *)irq_num, "twl4030-irq");
+	if (IS_ERR(task)) {
+		pr_err("twl4030: could not create irq %d thread!\n", irq_num);
+		status = PTR_ERR(task);
+		goto fail_kthread;
+	}
+	return status;
+fail_kthread:
+	free_irq(irq_num, &irq_event);
+fail_rqirq:
+	/* clean up twl4030_sih_setup */
 fail:
 	for (i = irq_base; i < irq_end; i++)
 		set_irq_chip_and_handler(i, NULL, NULL);
diff --git a/drivers/misc/cb710/sgbuf2.c b/drivers/misc/cb710/sgbuf2.c
index d38a7acdb6e..d019746551f 100644
--- a/drivers/misc/cb710/sgbuf2.c
+++ b/drivers/misc/cb710/sgbuf2.c
@@ -114,7 +114,6 @@ static void sg_dwiter_write_slow(struct sg_mapping_iter *miter, uint32_t data)
 		if (!left)
 			return;
 		addr += len;
-		flush_kernel_dcache_page(miter->page);
 	} while (sg_dwiter_next(miter));
 }
 
@@ -142,9 +141,6 @@ void cb710_sg_dwiter_write_next_block(struct sg_mapping_iter *miter, uint32_t da
 			return;
 	} else
 		sg_dwiter_write_slow(miter, data);
-
-	if (miter->length == miter->consumed)
-		flush_kernel_dcache_page(miter->page);
 }
 EXPORT_SYMBOL_GPL(cb710_sg_dwiter_write_next_block);
 
diff --git a/drivers/misc/eeprom/at25.c b/drivers/misc/eeprom/at25.c
index b34cb5f79ee..2e535a0ccd5 100644
--- a/drivers/misc/eeprom/at25.c
+++ b/drivers/misc/eeprom/at25.c
@@ -173,6 +173,7 @@ at25_ee_write(struct at25_data *at25, const char *buf, loff_t off,
 		unsigned	segment;
 		unsigned	offset = (unsigned) off;
 		u8		*cp = bounce + 1;
+		int		sr;
 
 		*cp = AT25_WREN;
 		status = spi_write(at25->spi, cp, 1);
@@ -214,7 +215,6 @@ at25_ee_write(struct at25_data *at25, const char *buf, loff_t off,
 		timeout = jiffies + msecs_to_jiffies(EE_TIMEOUT);
 		retries = 0;
 		do {
-			int	sr;
 
 			sr = spi_w8r8(at25->spi, AT25_RDSR);
 			if (sr < 0 || (sr & AT25_SR_nRDY)) {
@@ -228,7 +228,7 @@ at25_ee_write(struct at25_data *at25, const char *buf, loff_t off,
 				break;
 		} while (retries++ < 3 || time_before_eq(jiffies, timeout));
 
-		if (time_after(jiffies, timeout)) {
+		if ((sr < 0) || (sr & AT25_SR_nRDY)) {
 			dev_err(&at25->spi->dev,
 				"write %d bytes offset %d, "
 				"timeout after %u msecs\n",
diff --git a/drivers/mmc/host/cb710-mmc.c b/drivers/mmc/host/cb710-mmc.c
index 11efefb1af5..4e72964a7b4 100644
--- a/drivers/mmc/host/cb710-mmc.c
+++ b/drivers/mmc/host/cb710-mmc.c
@@ -278,7 +278,7 @@ static int cb710_mmc_receive(struct cb710_slot *slot, struct mmc_data *data)
 	if (unlikely(data->blksz & 15 && (data->blocks != 1 || data->blksz != 8)))
 		return -EINVAL;
 
-	sg_miter_start(&miter, data->sg, data->sg_len, 0);
+	sg_miter_start(&miter, data->sg, data->sg_len, SG_MITER_TO_SG);
 
 	cb710_modify_port_8(slot, CB710_MMC_CONFIG2_PORT,
 		15, CB710_MMC_C2_READ_PIO_SIZE_MASK);
@@ -307,7 +307,7 @@ static int cb710_mmc_receive(struct cb710_slot *slot, struct mmc_data *data)
 			goto out;
 	}
 out:
-	cb710_sg_miter_stop_writing(&miter);
+	sg_miter_stop(&miter);
 	return err;
 }
 
@@ -322,7 +322,7 @@ static int cb710_mmc_send(struct cb710_slot *slot, struct mmc_data *data)
 	if (unlikely(data->blocks > 1 && data->blksz & 15))
 		return -EINVAL;
 
-	sg_miter_start(&miter, data->sg, data->sg_len, 0);
+	sg_miter_start(&miter, data->sg, data->sg_len, SG_MITER_FROM_SG);
 
 	cb710_modify_port_8(slot, CB710_MMC_CONFIG2_PORT,
 		0, CB710_MMC_C2_READ_PIO_SIZE_MASK);
diff --git a/drivers/mmc/host/imxmmc.c b/drivers/mmc/host/imxmmc.c
index e0be21a4a69..bf98d7cc928 100644
--- a/drivers/mmc/host/imxmmc.c
+++ b/drivers/mmc/host/imxmmc.c
@@ -652,7 +652,7 @@ static irqreturn_t imxmci_irq(int irq, void *devid)
 	set_bit(IMXMCI_PEND_STARTED_b, &host->pending_events);
 	tasklet_schedule(&host->tasklet);
 
-	return IRQ_RETVAL(handled);;
+	return IRQ_RETVAL(handled);
 }
 
 static void imxmci_tasklet_fnc(unsigned long data)
diff --git a/drivers/mmc/host/sdhci-of.c b/drivers/mmc/host/sdhci-of.c
index d79fa55c3b8..1e8aa590bb3 100644
--- a/drivers/mmc/host/sdhci-of.c
+++ b/drivers/mmc/host/sdhci-of.c
@@ -158,6 +158,13 @@ static unsigned int esdhc_get_max_clock(struct sdhci_host *host)
 	return of_host->clock;
 }
 
+static unsigned int esdhc_get_min_clock(struct sdhci_host *host)
+{
+	struct sdhci_of_host *of_host = sdhci_priv(host);
+
+	return of_host->clock / 256 / 16;
+}
+
 static unsigned int esdhc_get_timeout_clock(struct sdhci_host *host)
 {
 	struct sdhci_of_host *of_host = sdhci_priv(host);
@@ -184,6 +191,7 @@ static struct sdhci_of_data sdhci_esdhc = {
 		.set_clock = esdhc_set_clock,
 		.enable_dma = esdhc_enable_dma,
 		.get_max_clock = esdhc_get_max_clock,
+		.get_min_clock = esdhc_get_min_clock,
 		.get_timeout_clock = esdhc_get_timeout_clock,
 	},
 };
@@ -226,7 +234,7 @@ static int __devinit sdhci_of_probe(struct of_device *ofdev,
 		return -ENODEV;
 
 	host = sdhci_alloc_host(&ofdev->dev, sizeof(*of_host));
-	if (!host)
+	if (IS_ERR(host))
 		return -ENOMEM;
 
 	of_host = sdhci_priv(host);
diff --git a/drivers/mmc/host/sdhci.c b/drivers/mmc/host/sdhci.c
index 6779b4ecab1..fc96f8cb9c0 100644
--- a/drivers/mmc/host/sdhci.c
+++ b/drivers/mmc/host/sdhci.c
@@ -773,8 +773,14 @@ static void sdhci_prepare_data(struct sdhci_host *host, struct mmc_data *data)
 	}
 
 	if (!(host->flags & SDHCI_REQ_USE_DMA)) {
-		sg_miter_start(&host->sg_miter,
-			data->sg, data->sg_len, SG_MITER_ATOMIC);
+		int flags;
+
+		flags = SG_MITER_ATOMIC;
+		if (host->data->flags & MMC_DATA_READ)
+			flags |= SG_MITER_TO_SG;
+		else
+			flags |= SG_MITER_FROM_SG;
+		sg_miter_start(&host->sg_miter, data->sg, data->sg_len, flags);
 		host->blocks = data->blocks;
 	}
 
@@ -1766,7 +1772,10 @@ int sdhci_add_host(struct sdhci_host *host)
 	 * Set host parameters.
 	 */
 	mmc->ops = &sdhci_ops;
-	mmc->f_min = host->max_clk / 256;
+	if (host->ops->get_min_clock)
+		mmc->f_min = host->ops->get_min_clock(host);
+	else
+		mmc->f_min = host->max_clk / 256;
 	mmc->f_max = host->max_clk;
 	mmc->caps = MMC_CAP_SDIO_IRQ;
 
diff --git a/drivers/mmc/host/sdhci.h b/drivers/mmc/host/sdhci.h
index 831ddf7dcb4..c77e9ff3022 100644
--- a/drivers/mmc/host/sdhci.h
+++ b/drivers/mmc/host/sdhci.h
@@ -302,6 +302,7 @@ struct sdhci_ops {
 
 	int		(*enable_dma)(struct sdhci_host *host);
 	unsigned int	(*get_max_clock)(struct sdhci_host *host);
+	unsigned int	(*get_min_clock)(struct sdhci_host *host);
 	unsigned int	(*get_timeout_clock)(struct sdhci_host *host);
 };
 
diff --git a/drivers/mtd/maps/Kconfig b/drivers/mtd/maps/Kconfig
index 0b98654d8ee..7a58bd5522f 100644
--- a/drivers/mtd/maps/Kconfig
+++ b/drivers/mtd/maps/Kconfig
@@ -284,13 +284,6 @@ config MTD_L440GX
 
 	  BE VERY CAREFUL.
 
-config MTD_SBC8240
-	tristate "Flash device on SBC8240"
-	depends on MTD_JEDECPROBE && 8260
-	help
-          Flash access on the SBC8240 board from Wind River.  See
-          <http://www.windriver.com/products/sbc8240/>
-
 config MTD_TQM8XXL
 	tristate "CFI Flash device mapped on TQM8XXL"
 	depends on MTD_CFI && TQM8xxL
diff --git a/drivers/mtd/maps/Makefile b/drivers/mtd/maps/Makefile
index 8bae7f9850c..5beb0662d72 100644
--- a/drivers/mtd/maps/Makefile
+++ b/drivers/mtd/maps/Makefile
@@ -50,7 +50,6 @@ obj-$(CONFIG_MTD_UCLINUX)	+= uclinux.o
 obj-$(CONFIG_MTD_NETtel)	+= nettel.o
 obj-$(CONFIG_MTD_SCB2_FLASH)	+= scb2_flash.o
 obj-$(CONFIG_MTD_H720X)		+= h720x-flash.o
-obj-$(CONFIG_MTD_SBC8240)	+= sbc8240.o
 obj-$(CONFIG_MTD_IXP4XX)	+= ixp4xx.o
 obj-$(CONFIG_MTD_IXP2000)	+= ixp2000.o
 obj-$(CONFIG_MTD_WRSBC8260)	+= wr_sbc82xx_flash.o
diff --git a/drivers/mtd/maps/sbc8240.c b/drivers/mtd/maps/sbc8240.c
index d5374cdcb16..e69de29bb2d 100644
--- a/drivers/mtd/maps/sbc8240.c
+++ b/drivers/mtd/maps/sbc8240.c
@@ -1,250 +0,0 @@
-/*
- * Handle mapping of the flash memory access routines on the SBC8240 board.
- *
- * Carolyn Smith, Tektronix, Inc.
- *
- * This code is GPLed
- */
-
-/*
- * The SBC8240 has 2 flash banks.
- * Bank 0 is a 512 KiB AMD AM29F040B; 8 x 64 KiB sectors.
- * It contains the U-Boot code (7 sectors) and the environment (1 sector).
- * Bank 1 is 4 x 1 MiB AMD AM29LV800BT; 15 x 64 KiB sectors, 1 x 32 KiB sector,
- * 2 x 8 KiB sectors, 1 x 16 KiB sectors.
- * Both parts are JEDEC compatible.
- */
-
-#include <linux/module.h>
-#include <linux/types.h>
-#include <linux/kernel.h>
-#include <asm/io.h>
-
-#include <linux/mtd/mtd.h>
-#include <linux/mtd/map.h>
-#include <linux/mtd/cfi.h>
-
-#ifdef CONFIG_MTD_PARTITIONS
-#include <linux/mtd/partitions.h>
-#endif
-
-#define	DEBUG
-
-#ifdef	DEBUG
-# define debugk(fmt,args...)	printk(fmt ,##args)
-#else
-# define debugk(fmt,args...)
-#endif
-
-
-#define WINDOW_ADDR0	0xFFF00000		/* 512 KiB */
-#define WINDOW_SIZE0	0x00080000
-#define BUSWIDTH0	1
-
-#define WINDOW_ADDR1	0xFF000000		/* 4 MiB */
-#define WINDOW_SIZE1	0x00400000
-#define BUSWIDTH1	8
-
-#define MSG_PREFIX "sbc8240:"	/* prefix for our printk()'s */
-#define MTDID	   "sbc8240-%d"	/* for mtdparts= partitioning */
-
-
-static struct map_info sbc8240_map[2] = {
-	{
-		.name           = "sbc8240 Flash Bank #0",
-		.size           = WINDOW_SIZE0,
-		.bankwidth       = BUSWIDTH0,
-	},
-	{
-		.name           = "sbc8240 Flash Bank #1",
-		.size           = WINDOW_SIZE1,
-		.bankwidth       = BUSWIDTH1,
-	}
-};
-
-#define NUM_FLASH_BANKS	ARRAY_SIZE(sbc8240_map)
-
-/*
- * The following defines the partition layout of SBC8240 boards.
- *
- * See include/linux/mtd/partitions.h for definition of the
- * mtd_partition structure.
- *
- * The *_max_flash_size is the maximum possible mapped flash size
- * which is not necessarily the actual flash size. It must correspond
- * to the value specified in the mapping definition defined by the
- * "struct map_desc *_io_desc" for the corresponding machine.
- */
-
-#ifdef CONFIG_MTD_PARTITIONS
-
-static struct mtd_partition sbc8240_uboot_partitions [] = {
-	/* Bank 0 */
-	{
-		.name =	"U-boot",			/* U-Boot Firmware	*/
-		.offset =	0,
-		.size =	0x00070000,			/*  7 x 64 KiB sectors 	*/
-		.mask_flags = MTD_WRITEABLE,		/*  force read-only	*/
-	},
-	{
-		.name =	"environment",			/* U-Boot environment	*/
-		.offset =	0x00070000,
-		.size =	0x00010000,			/*  1 x 64 KiB sector	*/
-	},
-};
-
-static struct mtd_partition sbc8240_fs_partitions [] = {
-	{
-		.name =	"jffs",				/* JFFS  filesystem	*/
-		.offset =	0,
-		.size =	0x003C0000,			/*  4 * 15 * 64KiB	*/
-	},
-	{
-		.name =	"tmp32",
-		.offset =	0x003C0000,
-		.size =	0x00020000,			/*  4 * 32KiB		*/
-	},
-	{
-		.name =	"tmp8a",
-		.offset =	0x003E0000,
-		.size =	0x00008000,			/*  4 * 8KiB		*/
-	},
-	{
-		.name =	"tmp8b",
-		.offset =	0x003E8000,
-		.size =	0x00008000,			/*  4 * 8KiB		*/
-	},
-	{
-		.name =	"tmp16",
-		.offset =	0x003F0000,
-		.size =	0x00010000,			/*  4 * 16KiB		*/
-	}
-};
-
-/* trivial struct to describe partition information */
-struct mtd_part_def
-{
-	int nums;
-	unsigned char *type;
-	struct mtd_partition* mtd_part;
-};
-
-static struct mtd_info *sbc8240_mtd[NUM_FLASH_BANKS];
-static struct mtd_part_def sbc8240_part_banks[NUM_FLASH_BANKS];
-
-
-#endif	/* CONFIG_MTD_PARTITIONS */
-
-
-static int __init init_sbc8240_mtd (void)
-{
-	static struct _cjs {
-		u_long addr;
-		u_long size;
-	} pt[NUM_FLASH_BANKS] = {
-		{
-			.addr = WINDOW_ADDR0,
-			.size = WINDOW_SIZE0
-		},
-		{
-			.addr = WINDOW_ADDR1,
-			.size = WINDOW_SIZE1
-		},
-	};
-
-	int devicesfound = 0;
-	int i,j;
-
-	for (i = 0; i < NUM_FLASH_BANKS; i++) {
-		printk (KERN_NOTICE MSG_PREFIX
-			"Probing 0x%08lx at 0x%08lx\n", pt[i].size, pt[i].addr);
-
-		sbc8240_map[i].map_priv_1 =
-			(unsigned long) ioremap (pt[i].addr, pt[i].size);
-		if (!sbc8240_map[i].map_priv_1) {
-			printk (MSG_PREFIX "failed to ioremap\n");
-			for (j = 0; j < i; j++) {
-				iounmap((void *) sbc8240_map[j].map_priv_1);
-				sbc8240_map[j].map_priv_1 = 0;
-			}
-			return -EIO;
-		}
-		simple_map_init(&sbc8240_mtd[i]);
-
-		sbc8240_mtd[i] = do_map_probe("jedec_probe", &sbc8240_map[i]);
-
-		if (sbc8240_mtd[i]) {
-			sbc8240_mtd[i]->module = THIS_MODULE;
-			devicesfound++;
-		} else {
-			if (sbc8240_map[i].map_priv_1) {
-				iounmap((void *) sbc8240_map[i].map_priv_1);
-				sbc8240_map[i].map_priv_1 = 0;
-			}
-		}
-	}
-
-	if (!devicesfound) {
-		printk(KERN_NOTICE MSG_PREFIX
-		       "No suppported flash chips found!\n");
-		return -ENXIO;
-	}
-
-#ifdef CONFIG_MTD_PARTITIONS
-	sbc8240_part_banks[0].mtd_part   = sbc8240_uboot_partitions;
-	sbc8240_part_banks[0].type       = "static image";
-	sbc8240_part_banks[0].nums       = ARRAY_SIZE(sbc8240_uboot_partitions);
-	sbc8240_part_banks[1].mtd_part   = sbc8240_fs_partitions;
-	sbc8240_part_banks[1].type       = "static file system";
-	sbc8240_part_banks[1].nums       = ARRAY_SIZE(sbc8240_fs_partitions);
-
-	for (i = 0; i < NUM_FLASH_BANKS; i++) {
-
-		if (!sbc8240_mtd[i]) continue;
-		if (sbc8240_part_banks[i].nums == 0) {
-			printk (KERN_NOTICE MSG_PREFIX
-				"No partition info available, registering whole device\n");
-			add_mtd_device(sbc8240_mtd[i]);
-		} else {
-			printk (KERN_NOTICE MSG_PREFIX
-				"Using %s partition definition\n", sbc8240_part_banks[i].mtd_part->name);
-			add_mtd_partitions (sbc8240_mtd[i],
-					    sbc8240_part_banks[i].mtd_part,
-					    sbc8240_part_banks[i].nums);
-		}
-	}
-#else
-	printk(KERN_NOTICE MSG_PREFIX
-	       "Registering %d flash banks at once\n", devicesfound);
-
-	for (i = 0; i < devicesfound; i++) {
-		add_mtd_device(sbc8240_mtd[i]);
-	}
-#endif	/* CONFIG_MTD_PARTITIONS */
-
-	return devicesfound == 0 ? -ENXIO : 0;
-}
-
-static void __exit cleanup_sbc8240_mtd (void)
-{
-	int i;
-
-	for (i = 0; i < NUM_FLASH_BANKS; i++) {
-		if (sbc8240_mtd[i]) {
-			del_mtd_device (sbc8240_mtd[i]);
-			map_destroy (sbc8240_mtd[i]);
-		}
-		if (sbc8240_map[i].map_priv_1) {
-			iounmap ((void *) sbc8240_map[i].map_priv_1);
-			sbc8240_map[i].map_priv_1 = 0;
-		}
-	}
-}
-
-module_init (init_sbc8240_mtd);
-module_exit (cleanup_sbc8240_mtd);
-
-MODULE_LICENSE ("GPL");
-MODULE_AUTHOR ("Carolyn Smith <carolyn.smith@tektronix.com>");
-MODULE_DESCRIPTION ("MTD map driver for SBC8240 boards");
-
diff --git a/drivers/mtd/mtd_blkdevs.c b/drivers/mtd/mtd_blkdevs.c
index c3f62654b6d..7baba40c1ed 100644
--- a/drivers/mtd/mtd_blkdevs.c
+++ b/drivers/mtd/mtd_blkdevs.c
@@ -144,7 +144,7 @@ static int blktrans_open(struct block_device *bdev, fmode_t mode)
 	struct mtd_blktrans_ops *tr = dev->tr;
 	int ret = -ENODEV;
 
-	if (!try_module_get(dev->mtd->owner))
+	if (!get_mtd_device(NULL, dev->mtd->index))
 		goto out;
 
 	if (!try_module_get(tr->owner))
@@ -158,7 +158,7 @@ static int blktrans_open(struct block_device *bdev, fmode_t mode)
 	ret = 0;
 	if (tr->open && (ret = tr->open(dev))) {
 		dev->mtd->usecount--;
-		module_put(dev->mtd->owner);
+		put_mtd_device(dev->mtd);
 	out_tr:
 		module_put(tr->owner);
 	}
@@ -177,7 +177,7 @@ static int blktrans_release(struct gendisk *disk, fmode_t mode)
 
 	if (!ret) {
 		dev->mtd->usecount--;
-		module_put(dev->mtd->owner);
+		put_mtd_device(dev->mtd);
 		module_put(tr->owner);
 	}
 
diff --git a/drivers/mtd/mtdblock.c b/drivers/mtd/mtdblock.c
index 208c6faa035..77db5ce24d9 100644
--- a/drivers/mtd/mtdblock.c
+++ b/drivers/mtd/mtdblock.c
@@ -29,6 +29,8 @@ static struct mtdblk_dev {
 	enum { STATE_EMPTY, STATE_CLEAN, STATE_DIRTY } cache_state;
 } *mtdblks[MAX_MTD_DEVICES];
 
+static struct mutex mtdblks_lock;
+
 /*
  * Cache stuff...
  *
@@ -270,15 +272,19 @@ static int mtdblock_open(struct mtd_blktrans_dev *mbd)
 
 	DEBUG(MTD_DEBUG_LEVEL1,"mtdblock_open\n");
 
+	mutex_lock(&mtdblks_lock);
 	if (mtdblks[dev]) {
 		mtdblks[dev]->count++;
+		mutex_unlock(&mtdblks_lock);
 		return 0;
 	}
 
 	/* OK, it's not open. Create cache info for it */
 	mtdblk = kzalloc(sizeof(struct mtdblk_dev), GFP_KERNEL);
-	if (!mtdblk)
+	if (!mtdblk) {
+		mutex_unlock(&mtdblks_lock);
 		return -ENOMEM;
+	}
 
 	mtdblk->count = 1;
 	mtdblk->mtd = mtd;
@@ -291,6 +297,7 @@ static int mtdblock_open(struct mtd_blktrans_dev *mbd)
 	}
 
 	mtdblks[dev] = mtdblk;
+	mutex_unlock(&mtdblks_lock);
 
 	DEBUG(MTD_DEBUG_LEVEL1, "ok\n");
 
@@ -304,6 +311,8 @@ static int mtdblock_release(struct mtd_blktrans_dev *mbd)
 
    	DEBUG(MTD_DEBUG_LEVEL1, "mtdblock_release\n");
 
+	mutex_lock(&mtdblks_lock);
+
 	mutex_lock(&mtdblk->cache_mutex);
 	write_cached_data(mtdblk);
 	mutex_unlock(&mtdblk->cache_mutex);
@@ -316,6 +325,9 @@ static int mtdblock_release(struct mtd_blktrans_dev *mbd)
 		vfree(mtdblk->cache_data);
 		kfree(mtdblk);
 	}
+
+	mutex_unlock(&mtdblks_lock);
+
 	DEBUG(MTD_DEBUG_LEVEL1, "ok\n");
 
 	return 0;
@@ -376,6 +388,8 @@ static struct mtd_blktrans_ops mtdblock_tr = {
 
 static int __init init_mtdblock(void)
 {
+	mutex_init(&mtdblks_lock);
+
 	return register_mtd_blktrans(&mtdblock_tr);
 }
 
diff --git a/drivers/mtd/mtdcore.c b/drivers/mtd/mtdcore.c
index fac54a3fa3f..00ebf7af746 100644
--- a/drivers/mtd/mtdcore.c
+++ b/drivers/mtd/mtdcore.c
@@ -65,8 +65,8 @@ static void mtd_release(struct device *dev)
 static int mtd_cls_suspend(struct device *dev, pm_message_t state)
 {
 	struct mtd_info *mtd = dev_to_mtd(dev);
-	
-	if (mtd->suspend)
+
+	if (mtd && mtd->suspend)
 		return mtd->suspend(mtd);
 	else
 		return 0;
@@ -76,7 +76,7 @@ static int mtd_cls_resume(struct device *dev)
 {
 	struct mtd_info *mtd = dev_to_mtd(dev);
 	
-	if (mtd->resume)
+	if (mtd && mtd->resume)
 		mtd->resume(mtd);
 	return 0;
 }
@@ -298,6 +298,7 @@ int add_mtd_device(struct mtd_info *mtd)
 			mtd->dev.class = &mtd_class;
 			mtd->dev.devt = MTD_DEVT(i);
 			dev_set_name(&mtd->dev, "mtd%d", i);
+			dev_set_drvdata(&mtd->dev, mtd);
 			if (device_register(&mtd->dev) != 0) {
 				mtd_table[i] = NULL;
 				break;
diff --git a/drivers/mtd/onenand/omap2.c b/drivers/mtd/onenand/omap2.c
index 38d656b9b2e..0108ed42e87 100644
--- a/drivers/mtd/onenand/omap2.c
+++ b/drivers/mtd/onenand/omap2.c
@@ -266,7 +266,7 @@ static inline int omap2_onenand_bufferram_offset(struct mtd_info *mtd, int area)
 
 	if (ONENAND_CURRENT_BUFFERRAM(this)) {
 		if (area == ONENAND_DATARAM)
-			return mtd->writesize;
+			return this->writesize;
 		if (area == ONENAND_SPARERAM)
 			return mtd->oobsize;
 	}
@@ -770,6 +770,7 @@ static int __devexit omap2_onenand_remove(struct platform_device *pdev)
 	}
 	iounmap(c->onenand.base);
 	release_mem_region(c->phys_base, ONENAND_IO_SIZE);
+	gpmc_cs_free(c->gpmc_cs);
 	kfree(c);
 
 	return 0;
diff --git a/drivers/net/3c515.c b/drivers/net/3c515.c
index 3e00fa8ea65..4a7c32895be 100644
--- a/drivers/net/3c515.c
+++ b/drivers/net/3c515.c
@@ -832,7 +832,9 @@ static int corkscrew_open(struct net_device *dev)
 			skb_reserve(skb, 2);	/* Align IP on 16 byte boundaries */
 			vp->rx_ring[i].addr = isa_virt_to_bus(skb->data);
 		}
-		vp->rx_ring[i - 1].next = isa_virt_to_bus(&vp->rx_ring[0]);	/* Wrap the ring. */
+		if (i != 0)
+			vp->rx_ring[i - 1].next =
+				isa_virt_to_bus(&vp->rx_ring[0]);	/* Wrap the ring. */
 		outl(isa_virt_to_bus(&vp->rx_ring[0]), ioaddr + UpListPtr);
 	}
 	if (vp->full_bus_master_tx) {	/* Boomerang bus master Tx. */
diff --git a/drivers/net/3c59x.c b/drivers/net/3c59x.c
index c34aee91250..c2041685094 100644
--- a/drivers/net/3c59x.c
+++ b/drivers/net/3c59x.c
@@ -2721,13 +2721,15 @@ dump_tx_ring(struct net_device *dev)
 				   &vp->tx_ring[vp->dirty_tx % TX_RING_SIZE]);
 			issue_and_wait(dev, DownStall);
 			for (i = 0; i < TX_RING_SIZE; i++) {
-				pr_err("  %d: @%p  length %8.8x status %8.8x\n", i,
-					   &vp->tx_ring[i],
+				unsigned int length;
+
 #if DO_ZEROCOPY
-					   le32_to_cpu(vp->tx_ring[i].frag[0].length),
+				length = le32_to_cpu(vp->tx_ring[i].frag[0].length);
 #else
-					   le32_to_cpu(vp->tx_ring[i].length),
+				length = le32_to_cpu(vp->tx_ring[i].length);
 #endif
+				pr_err("  %d: @%p  length %8.8x status %8.8x\n",
+					   i, &vp->tx_ring[i], length,
 					   le32_to_cpu(vp->tx_ring[i].status));
 			}
 			if (!stalled)
diff --git a/drivers/net/eexpress.c b/drivers/net/eexpress.c
index 1686dca2874..1f016d66684 100644
--- a/drivers/net/eexpress.c
+++ b/drivers/net/eexpress.c
@@ -1474,13 +1474,13 @@ static void eexp_hw_init586(struct net_device *dev)
 	outw(0x0000, ioaddr + 0x800c);
 	outw(0x0000, ioaddr + 0x800e);
 
-	for (i = 0; i < (sizeof(start_code)); i+=32) {
+	for (i = 0; i < ARRAY_SIZE(start_code) * 2; i+=32) {
 		int j;
 		outw(i, ioaddr + SM_PTR);
-		for (j = 0; j < 16; j+=2)
+		for (j = 0; j < 16 && (i+j)/2 < ARRAY_SIZE(start_code); j+=2)
 			outw(start_code[(i+j)/2],
 			     ioaddr+0x4000+j);
-		for (j = 0; j < 16; j+=2)
+		for (j = 0; j < 16 && (i+j+16)/2 < ARRAY_SIZE(start_code); j+=2)
 			outw(start_code[(i+j+16)/2],
 			     ioaddr+0x8000+j);
 	}
diff --git a/drivers/net/ehea/ehea.h b/drivers/net/ehea/ehea.h
index 78952f8324e..fa311a95099 100644
--- a/drivers/net/ehea/ehea.h
+++ b/drivers/net/ehea/ehea.h
@@ -40,7 +40,7 @@
 #include <asm/io.h>
 
 #define DRV_NAME	"ehea"
-#define DRV_VERSION	"EHEA_0101"
+#define DRV_VERSION	"EHEA_0102"
 
 /* eHEA capability flags */
 #define DLPAR_PORT_ADD_REM 1
diff --git a/drivers/net/ehea/ehea_main.c b/drivers/net/ehea/ehea_main.c
index e8d46cc1bec..977c3d35827 100644
--- a/drivers/net/ehea/ehea_main.c
+++ b/drivers/net/ehea/ehea_main.c
@@ -1545,6 +1545,9 @@ static int ehea_clean_portres(struct ehea_port *port, struct ehea_port_res *pr)
 {
 	int ret, i;
 
+	if (pr->qp)
+		netif_napi_del(&pr->napi);
+
 	ret = ehea_destroy_qp(pr->qp);
 
 	if (!ret) {
diff --git a/drivers/net/gianfar_ethtool.c b/drivers/net/gianfar_ethtool.c
index dbf06e9313c..2234118eedb 100644
--- a/drivers/net/gianfar_ethtool.c
+++ b/drivers/net/gianfar_ethtool.c
@@ -366,9 +366,8 @@ static int gfar_scoalesce(struct net_device *dev, struct ethtool_coalesce *cvals
 		return -EINVAL;
 	}
 
-	priv->rxic = mk_ic_value(
-		gfar_usecs2ticks(priv, cvals->rx_coalesce_usecs),
-		cvals->rx_max_coalesced_frames);
+	priv->rxic = mk_ic_value(cvals->rx_max_coalesced_frames,
+		gfar_usecs2ticks(priv, cvals->rx_coalesce_usecs));
 
 	/* Set up tx coalescing */
 	if ((cvals->tx_coalesce_usecs == 0) ||
@@ -390,9 +389,8 @@ static int gfar_scoalesce(struct net_device *dev, struct ethtool_coalesce *cvals
 		return -EINVAL;
 	}
 
-	priv->txic = mk_ic_value(
-		gfar_usecs2ticks(priv, cvals->tx_coalesce_usecs),
-		cvals->tx_max_coalesced_frames);
+	priv->txic = mk_ic_value(cvals->tx_max_coalesced_frames,
+		gfar_usecs2ticks(priv, cvals->tx_coalesce_usecs));
 
 	gfar_write(&priv->regs->rxic, 0);
 	if (priv->rxcoalescing)
diff --git a/drivers/net/igbvf/vf.c b/drivers/net/igbvf/vf.c
index 2a4faf9ade6..a9a61efa964 100644
--- a/drivers/net/igbvf/vf.c
+++ b/drivers/net/igbvf/vf.c
@@ -274,6 +274,8 @@ static s32 e1000_set_vfta_vf(struct e1000_hw *hw, u16 vid, bool set)
 
 	err = mbx->ops.read_posted(hw, msgbuf, 2);
 
+	msgbuf[0] &= ~E1000_VT_MSGTYPE_CTS;
+
 	/* if nacked the vlan was rejected */
 	if (!err && (msgbuf[0] == (E1000_VF_SET_VLAN | E1000_VT_MSGTYPE_NACK)))
 		err = -E1000_ERR_MAC_INIT;
@@ -317,6 +319,8 @@ static void e1000_rar_set_vf(struct e1000_hw *hw, u8 * addr, u32 index)
 	if (!ret_val)
 		ret_val = mbx->ops.read_posted(hw, msgbuf, 3);
 
+	msgbuf[0] &= ~E1000_VT_MSGTYPE_CTS;
+
 	/* if nacked the address was rejected, use "perm_addr" */
 	if (!ret_val &&
 	    (msgbuf[0] == (E1000_VF_SET_MAC_ADDR | E1000_VT_MSGTYPE_NACK)))
diff --git a/drivers/net/ixgbe/ixgbe.h b/drivers/net/ixgbe/ixgbe.h
index 1b12c7ba275..e11d83d5852 100644
--- a/drivers/net/ixgbe/ixgbe.h
+++ b/drivers/net/ixgbe/ixgbe.h
@@ -96,6 +96,8 @@
 #define IXGBE_TX_FLAGS_VLAN_PRIO_MASK   0x0000e000
 #define IXGBE_TX_FLAGS_VLAN_SHIFT	16
 
+#define IXGBE_MAX_RSC_INT_RATE          162760
+
 /* wrapper around a pointer to a socket buffer,
  * so a DMA handle can be stored along with the buffer */
 struct ixgbe_tx_buffer {
diff --git a/drivers/net/ixgbe/ixgbe_82598.c b/drivers/net/ixgbe/ixgbe_82598.c
index b9923047ce1..522c03bc1da 100644
--- a/drivers/net/ixgbe/ixgbe_82598.c
+++ b/drivers/net/ixgbe/ixgbe_82598.c
@@ -50,6 +50,51 @@ static s32 ixgbe_read_i2c_eeprom_82598(struct ixgbe_hw *hw, u8 byte_offset,
                                        u8 *eeprom_data);
 
 /**
+ *  ixgbe_set_pcie_completion_timeout - set pci-e completion timeout
+ *  @hw: pointer to the HW structure
+ *
+ *  The defaults for 82598 should be in the range of 50us to 50ms,
+ *  however the hardware default for these parts is 500us to 1ms which is less
+ *  than the 10ms recommended by the pci-e spec.  To address this we need to
+ *  increase the value to either 10ms to 250ms for capability version 1 config,
+ *  or 16ms to 55ms for version 2.
+ **/
+void ixgbe_set_pcie_completion_timeout(struct ixgbe_hw *hw)
+{
+	struct ixgbe_adapter *adapter = hw->back;
+	u32 gcr = IXGBE_READ_REG(hw, IXGBE_GCR);
+	u16 pcie_devctl2;
+
+	/* only take action if timeout value is defaulted to 0 */
+	if (gcr & IXGBE_GCR_CMPL_TMOUT_MASK)
+		goto out;
+
+	/*
+	 * if capababilities version is type 1 we can write the
+	 * timeout of 10ms to 250ms through the GCR register
+	 */
+	if (!(gcr & IXGBE_GCR_CAP_VER2)) {
+		gcr |= IXGBE_GCR_CMPL_TMOUT_10ms;
+		goto out;
+	}
+
+	/*
+	 * for version 2 capabilities we need to write the config space
+	 * directly in order to set the completion timeout value for
+	 * 16ms to 55ms
+	 */
+	pci_read_config_word(adapter->pdev,
+	                     IXGBE_PCI_DEVICE_CONTROL2, &pcie_devctl2);
+	pcie_devctl2 |= IXGBE_PCI_DEVICE_CONTROL2_16ms;
+	pci_write_config_word(adapter->pdev,
+	                      IXGBE_PCI_DEVICE_CONTROL2, pcie_devctl2);
+out:
+	/* disable completion timeout resend */
+	gcr &= ~IXGBE_GCR_CMPL_TMOUT_RESEND;
+	IXGBE_WRITE_REG(hw, IXGBE_GCR, gcr);
+}
+
+/**
  *  ixgbe_get_pcie_msix_count_82598 - Gets MSI-X vector count
  *  @hw: pointer to hardware structure
  *
@@ -153,6 +198,26 @@ out:
 }
 
 /**
+ *  ixgbe_start_hw_82598 - Prepare hardware for Tx/Rx
+ *  @hw: pointer to hardware structure
+ *
+ *  Starts the hardware using the generic start_hw function.
+ *  Then set pcie completion timeout
+ **/
+s32 ixgbe_start_hw_82598(struct ixgbe_hw *hw)
+{
+	s32 ret_val = 0;
+
+	ret_val = ixgbe_start_hw_generic(hw);
+
+	/* set the completion timeout for interface */
+	if (ret_val == 0)
+		ixgbe_set_pcie_completion_timeout(hw);
+
+	return ret_val;
+}
+
+/**
  *  ixgbe_get_link_capabilities_82598 - Determines link capabilities
  *  @hw: pointer to hardware structure
  *  @speed: pointer to link speed
@@ -1085,7 +1150,7 @@ out:
 static struct ixgbe_mac_operations mac_ops_82598 = {
 	.init_hw		= &ixgbe_init_hw_generic,
 	.reset_hw		= &ixgbe_reset_hw_82598,
-	.start_hw		= &ixgbe_start_hw_generic,
+	.start_hw		= &ixgbe_start_hw_82598,
 	.clear_hw_cntrs		= &ixgbe_clear_hw_cntrs_generic,
 	.get_media_type		= &ixgbe_get_media_type_82598,
 	.get_supported_physical_layer = &ixgbe_get_supported_physical_layer_82598,
diff --git a/drivers/net/ixgbe/ixgbe_ethtool.c b/drivers/net/ixgbe/ixgbe_ethtool.c
index 2a978008fd6..79144e950a3 100644
--- a/drivers/net/ixgbe/ixgbe_ethtool.c
+++ b/drivers/net/ixgbe/ixgbe_ethtool.c
@@ -1975,7 +1975,10 @@ static int ixgbe_set_coalesce(struct net_device *netdev,
 		 * any other value means disable eitr, which is best
 		 * served by setting the interrupt rate very high
 		 */
-		adapter->eitr_param = IXGBE_MAX_INT_RATE;
+		if (adapter->flags2 & IXGBE_FLAG2_RSC_ENABLED)
+			adapter->eitr_param = IXGBE_MAX_RSC_INT_RATE;
+		else
+			adapter->eitr_param = IXGBE_MAX_INT_RATE;
 		adapter->itr_setting = 0;
 	}
 
@@ -1999,13 +2002,13 @@ static int ixgbe_set_flags(struct net_device *netdev, u32 data)
 
 	ethtool_op_set_flags(netdev, data);
 
-	if (!(adapter->flags & IXGBE_FLAG2_RSC_CAPABLE))
+	if (!(adapter->flags2 & IXGBE_FLAG2_RSC_CAPABLE))
 		return 0;
 
 	/* if state changes we need to update adapter->flags and reset */
 	if ((!!(data & ETH_FLAG_LRO)) != 
-	    (!!(adapter->flags & IXGBE_FLAG2_RSC_ENABLED))) {
-		adapter->flags ^= IXGBE_FLAG2_RSC_ENABLED;
+	    (!!(adapter->flags2 & IXGBE_FLAG2_RSC_ENABLED))) {
+		adapter->flags2 ^= IXGBE_FLAG2_RSC_ENABLED;
 		if (netif_running(netdev))
 			ixgbe_reinit_locked(adapter);
 		else
diff --git a/drivers/net/ixgbe/ixgbe_main.c b/drivers/net/ixgbe/ixgbe_main.c
index 200454f30f6..110c65ab5cb 100644
--- a/drivers/net/ixgbe/ixgbe_main.c
+++ b/drivers/net/ixgbe/ixgbe_main.c
@@ -780,7 +780,7 @@ static bool ixgbe_clean_rx_irq(struct ixgbe_q_vector *q_vector,
 		prefetch(next_rxd);
 		cleaned_count++;
 
-		if (adapter->flags & IXGBE_FLAG2_RSC_CAPABLE)
+		if (adapter->flags2 & IXGBE_FLAG2_RSC_CAPABLE)
 			rsc_count = ixgbe_get_rsc_count(rx_desc);
 
 		if (rsc_count) {
@@ -2036,7 +2036,7 @@ static void ixgbe_configure_rx(struct ixgbe_adapter *adapter)
 			IXGBE_WRITE_REG(hw, IXGBE_PSRTYPE(0), psrtype);
 		}
 	} else {
-		if (!(adapter->flags & IXGBE_FLAG2_RSC_ENABLED) &&
+		if (!(adapter->flags2 & IXGBE_FLAG2_RSC_ENABLED) &&
 		    (netdev->mtu <= ETH_DATA_LEN))
 			rx_buf_len = MAXIMUM_ETHERNET_VLAN_SIZE;
 		else
@@ -2165,7 +2165,7 @@ static void ixgbe_configure_rx(struct ixgbe_adapter *adapter)
 		IXGBE_WRITE_REG(hw, IXGBE_RDRXCTL, rdrxctl);
 	}
 
-	if (adapter->flags & IXGBE_FLAG2_RSC_ENABLED) {
+	if (adapter->flags2 & IXGBE_FLAG2_RSC_ENABLED) {
 		/* Enable 82599 HW-RSC */
 		for (i = 0; i < adapter->num_rx_queues; i++) {
 			j = adapter->rx_ring[i].reg_idx;
@@ -3812,8 +3812,8 @@ static int __devinit ixgbe_sw_init(struct ixgbe_adapter *adapter)
 		adapter->max_msix_q_vectors = MAX_MSIX_Q_VECTORS_82598;
 	} else if (hw->mac.type == ixgbe_mac_82599EB) {
 		adapter->max_msix_q_vectors = MAX_MSIX_Q_VECTORS_82599;
-		adapter->flags |= IXGBE_FLAG2_RSC_CAPABLE;
-		adapter->flags |= IXGBE_FLAG2_RSC_ENABLED;
+		adapter->flags2 |= IXGBE_FLAG2_RSC_CAPABLE;
+		adapter->flags2 |= IXGBE_FLAG2_RSC_ENABLED;
 		adapter->flags |= IXGBE_FLAG_FDIR_HASH_CAPABLE;
 		adapter->ring_feature[RING_F_FDIR].indices =
 		                                         IXGBE_MAX_FDIR_INDICES;
@@ -5360,12 +5360,19 @@ static int ixgbe_del_sanmac_netdev(struct net_device *dev)
 static void ixgbe_netpoll(struct net_device *netdev)
 {
 	struct ixgbe_adapter *adapter = netdev_priv(netdev);
+	int i;
 
-	disable_irq(adapter->pdev->irq);
 	adapter->flags |= IXGBE_FLAG_IN_NETPOLL;
-	ixgbe_intr(adapter->pdev->irq, netdev);
+	if (adapter->flags & IXGBE_FLAG_MSIX_ENABLED) {
+		int num_q_vectors = adapter->num_msix_vectors - NON_Q_VECTORS;
+		for (i = 0; i < num_q_vectors; i++) {
+			struct ixgbe_q_vector *q_vector = adapter->q_vector[i];
+			ixgbe_msix_clean_many(0, q_vector);
+		}
+	} else {
+		ixgbe_intr(adapter->pdev->irq, netdev);
+	}
 	adapter->flags &= ~IXGBE_FLAG_IN_NETPOLL;
-	enable_irq(adapter->pdev->irq);
 }
 #endif
 
@@ -5611,7 +5618,7 @@ static int __devinit ixgbe_probe(struct pci_dev *pdev,
 	if (pci_using_dac)
 		netdev->features |= NETIF_F_HIGHDMA;
 
-	if (adapter->flags & IXGBE_FLAG2_RSC_ENABLED)
+	if (adapter->flags2 & IXGBE_FLAG2_RSC_ENABLED)
 		netdev->features |= NETIF_F_LRO;
 
 	/* make sure the EEPROM is good */
diff --git a/drivers/net/ixgbe/ixgbe_type.h b/drivers/net/ixgbe/ixgbe_type.h
index fa87309dc08..be90eb4575f 100644
--- a/drivers/net/ixgbe/ixgbe_type.h
+++ b/drivers/net/ixgbe/ixgbe_type.h
@@ -718,6 +718,12 @@
 #define IXGBE_ECC_STATUS_82599  0x110E0
 #define IXGBE_BAR_CTRL_82599    0x110F4
 
+/* PCI Express Control */
+#define IXGBE_GCR_CMPL_TMOUT_MASK       0x0000F000
+#define IXGBE_GCR_CMPL_TMOUT_10ms       0x00001000
+#define IXGBE_GCR_CMPL_TMOUT_RESEND     0x00010000
+#define IXGBE_GCR_CAP_VER2              0x00040000
+
 /* Time Sync Registers */
 #define IXGBE_TSYNCRXCTL 0x05188 /* Rx Time Sync Control register - RW */
 #define IXGBE_TSYNCTXCTL 0x08C00 /* Tx Time Sync Control register - RW */
@@ -1521,6 +1527,7 @@
 
 /* PCI Bus Info */
 #define IXGBE_PCI_LINK_STATUS     0xB2
+#define IXGBE_PCI_DEVICE_CONTROL2 0xC8
 #define IXGBE_PCI_LINK_WIDTH      0x3F0
 #define IXGBE_PCI_LINK_WIDTH_1    0x10
 #define IXGBE_PCI_LINK_WIDTH_2    0x20
@@ -1531,6 +1538,7 @@
 #define IXGBE_PCI_LINK_SPEED_5000 0x2
 #define IXGBE_PCI_HEADER_TYPE_REGISTER  0x0E
 #define IXGBE_PCI_HEADER_TYPE_MULTIFUNC 0x80
+#define IXGBE_PCI_DEVICE_CONTROL2_16ms  0x0005
 
 /* Number of 100 microseconds we wait for PCI Express master disable */
 #define IXGBE_PCI_MASTER_DISABLE_TIMEOUT 800
diff --git a/drivers/net/mlx4/en_tx.c b/drivers/net/mlx4/en_tx.c
index 08c43f2ae72..5a88b3f5769 100644
--- a/drivers/net/mlx4/en_tx.c
+++ b/drivers/net/mlx4/en_tx.c
@@ -249,6 +249,7 @@ static u32 mlx4_en_free_tx_desc(struct mlx4_en_priv *priv,
 				pci_unmap_page(mdev->pdev,
 					(dma_addr_t) be64_to_cpu(data->addr),
 					 frag->size, PCI_DMA_TODEVICE);
+				++data;
 			}
 		}
 		/* Stamp the freed descriptor */
diff --git a/drivers/net/netxen/netxen_nic_main.c b/drivers/net/netxen/netxen_nic_main.c
index 637ac8b89ba..3cd8cfcf627 100644
--- a/drivers/net/netxen/netxen_nic_main.c
+++ b/drivers/net/netxen/netxen_nic_main.c
@@ -221,7 +221,7 @@ netxen_napi_disable(struct netxen_adapter *adapter)
 	}
 }
 
-static int nx_set_dma_mask(struct netxen_adapter *adapter, uint8_t revision_id)
+static int nx_set_dma_mask(struct netxen_adapter *adapter)
 {
 	struct pci_dev *pdev = adapter->pdev;
 	uint64_t mask, cmask;
@@ -229,19 +229,17 @@ static int nx_set_dma_mask(struct netxen_adapter *adapter, uint8_t revision_id)
 	adapter->pci_using_dac = 0;
 
 	mask = DMA_BIT_MASK(32);
-	/*
-	 * Consistent DMA mask is set to 32 bit because it cannot be set to
-	 * 35 bits. For P3 also leave it at 32 bits for now. Only the rings
-	 * come off this pool.
-	 */
 	cmask = DMA_BIT_MASK(32);
 
+	if (NX_IS_REVISION_P2(adapter->ahw.revision_id)) {
 #ifndef CONFIG_IA64
-	if (revision_id >= NX_P3_B0)
-		mask = DMA_BIT_MASK(39);
-	else if (revision_id == NX_P2_C1)
 		mask = DMA_BIT_MASK(35);
 #endif
+	} else {
+		mask = DMA_BIT_MASK(39);
+		cmask = mask;
+	}
+
 	if (pci_set_dma_mask(pdev, mask) == 0 &&
 		pci_set_consistent_dma_mask(pdev, cmask) == 0) {
 		adapter->pci_using_dac = 1;
@@ -256,7 +254,7 @@ static int
 nx_update_dma_mask(struct netxen_adapter *adapter)
 {
 	int change, shift, err;
-	uint64_t mask, old_mask;
+	uint64_t mask, old_mask, old_cmask;
 	struct pci_dev *pdev = adapter->pdev;
 
 	change = 0;
@@ -272,14 +270,29 @@ nx_update_dma_mask(struct netxen_adapter *adapter)
 
 	if (change) {
 		old_mask = pdev->dma_mask;
+		old_cmask = pdev->dev.coherent_dma_mask;
+
 		mask = (1ULL<<(32+shift)) - 1;
 
 		err = pci_set_dma_mask(pdev, mask);
 		if (err)
-			return pci_set_dma_mask(pdev, old_mask);
+			goto err_out;
+
+		if (NX_IS_REVISION_P3(adapter->ahw.revision_id)) {
+
+			err = pci_set_consistent_dma_mask(pdev, mask);
+			if (err)
+				goto err_out;
+		}
+		dev_info(&pdev->dev, "using %d-bit dma mask\n", 32+shift);
 	}
 
 	return 0;
+
+err_out:
+	pci_set_dma_mask(pdev, old_mask);
+	pci_set_consistent_dma_mask(pdev, old_cmask);
+	return err;
 }
 
 static void netxen_check_options(struct netxen_adapter *adapter)
@@ -1006,7 +1019,7 @@ netxen_nic_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
 	revision_id = pdev->revision;
 	adapter->ahw.revision_id = revision_id;
 
-	err = nx_set_dma_mask(adapter, revision_id);
+	err = nx_set_dma_mask(adapter);
 	if (err)
 		goto err_out_free_netdev;
 
diff --git a/drivers/net/pcnet32.c b/drivers/net/pcnet32.c
index 28368157dac..a646a445fda 100644
--- a/drivers/net/pcnet32.c
+++ b/drivers/net/pcnet32.c
@@ -1611,8 +1611,11 @@ pcnet32_probe1(unsigned long ioaddr, int shared, struct pci_dev *pdev)
 		if (pcnet32_dwio_read_csr(ioaddr, 0) == 4
 		    && pcnet32_dwio_check(ioaddr)) {
 			a = &pcnet32_dwio;
-		} else
+		} else {
+			if (pcnet32_debug & NETIF_MSG_PROBE)
+				printk(KERN_ERR PFX "No access methods\n");
 			goto err_release_region;
+		}
 	}
 
 	chip_version =
@@ -1719,7 +1722,9 @@ pcnet32_probe1(unsigned long ioaddr, int shared, struct pci_dev *pdev)
 		ret = -ENOMEM;
 		goto err_release_region;
 	}
-	SET_NETDEV_DEV(dev, &pdev->dev);
+
+	if (pdev)
+		SET_NETDEV_DEV(dev, &pdev->dev);
 
 	if (pcnet32_debug & NETIF_MSG_PROBE)
 		printk(KERN_INFO PFX "%s at %#3lx,", chipname, ioaddr);
@@ -1818,7 +1823,6 @@ pcnet32_probe1(unsigned long ioaddr, int shared, struct pci_dev *pdev)
 
 	spin_lock_init(&lp->lock);
 
-	SET_NETDEV_DEV(dev, &pdev->dev);
 	lp->name = chipname;
 	lp->shared_irq = shared;
 	lp->tx_ring_size = TX_RING_SIZE;	/* default tx ring size */
@@ -1852,12 +1856,6 @@ pcnet32_probe1(unsigned long ioaddr, int shared, struct pci_dev *pdev)
 	    ((cards_found >= MAX_UNITS) || full_duplex[cards_found]))
 		lp->options |= PCNET32_PORT_FD;
 
-	if (!a) {
-		if (pcnet32_debug & NETIF_MSG_PROBE)
-			printk(KERN_ERR PFX "No access methods\n");
-		ret = -ENODEV;
-		goto err_free_consistent;
-	}
 	lp->a = *a;
 
 	/* prior to register_netdev, dev->name is not yet correct */
@@ -1973,14 +1971,13 @@ pcnet32_probe1(unsigned long ioaddr, int shared, struct pci_dev *pdev)
 
 	return 0;
 
-      err_free_ring:
+err_free_ring:
 	pcnet32_free_ring(dev);
-      err_free_consistent:
 	pci_free_consistent(lp->pci_dev, sizeof(*lp->init_block),
 			    lp->init_block, lp->init_dma_addr);
-      err_free_netdev:
+err_free_netdev:
 	free_netdev(dev);
-      err_release_region:
+err_release_region:
 	release_region(ioaddr, PCNET32_TOTAL_SIZE);
 	return ret;
 }
@@ -2089,6 +2086,7 @@ static void pcnet32_free_ring(struct net_device *dev)
 static int pcnet32_open(struct net_device *dev)
 {
 	struct pcnet32_private *lp = netdev_priv(dev);
+	struct pci_dev *pdev = lp->pci_dev;
 	unsigned long ioaddr = dev->base_addr;
 	u16 val;
 	int i;
@@ -2149,9 +2147,9 @@ static int pcnet32_open(struct net_device *dev)
 	lp->a.write_csr(ioaddr, 124, val);
 
 	/* Allied Telesyn AT 2700/2701 FX are 100Mbit only and do not negotiate */
-	if (lp->pci_dev->subsystem_vendor == PCI_VENDOR_ID_AT &&
-	    (lp->pci_dev->subsystem_device == PCI_SUBDEVICE_ID_AT_2700FX ||
-	     lp->pci_dev->subsystem_device == PCI_SUBDEVICE_ID_AT_2701FX)) {
+	if (pdev && pdev->subsystem_vendor == PCI_VENDOR_ID_AT &&
+	    (pdev->subsystem_device == PCI_SUBDEVICE_ID_AT_2700FX ||
+	     pdev->subsystem_device == PCI_SUBDEVICE_ID_AT_2701FX)) {
 		if (lp->options & PCNET32_PORT_ASEL) {
 			lp->options = PCNET32_PORT_FD | PCNET32_PORT_100;
 			if (netif_msg_link(lp))
diff --git a/drivers/net/ppp_generic.c b/drivers/net/ppp_generic.c
index 639d11bc444..cd37d739ac7 100644
--- a/drivers/net/ppp_generic.c
+++ b/drivers/net/ppp_generic.c
@@ -1384,7 +1384,7 @@ static int ppp_mp_explode(struct ppp *ppp, struct sk_buff *skb)
 
 	/* create a	fragment for each channel */
 	bits = B;
-	while (nfree > 0 &&	len	> 0) {
+	while (len	> 0) {
 		list = list->next;
 		if (list ==	&ppp->channels)	{
 			i =	0;
@@ -1431,29 +1431,31 @@ static int ppp_mp_explode(struct ppp *ppp, struct sk_buff *skb)
 		*otherwise divide it according to the speed
 		*of the channel we are going to transmit on
 		*/
-		if (pch->speed == 0) {
-			flen = totlen/nfree	;
-			if (nbigger > 0) {
-				flen++;
-				nbigger--;
-			}
-		} else {
-			flen = (((totfree - nzero)*(totlen + hdrlen*totfree)) /
-				((totspeed*totfree)/pch->speed)) - hdrlen;
-			if (nbigger > 0) {
-				flen += ((totfree - nzero)*pch->speed)/totspeed;
-				nbigger -= ((totfree - nzero)*pch->speed)/
+		if (nfree > 0) {
+			if (pch->speed == 0) {
+				flen = totlen/nfree	;
+				if (nbigger > 0) {
+					flen++;
+					nbigger--;
+				}
+			} else {
+				flen = (((totfree - nzero)*(totlen + hdrlen*totfree)) /
+					((totspeed*totfree)/pch->speed)) - hdrlen;
+				if (nbigger > 0) {
+					flen += ((totfree - nzero)*pch->speed)/totspeed;
+					nbigger -= ((totfree - nzero)*pch->speed)/
 							totspeed;
+				}
 			}
+			nfree--;
 		}
-		nfree--;
 
 		/*
 		 *check	if we are on the last channel or
 		 *we exceded the lenght	of the data	to
 		 *fragment
 		 */
-		if ((nfree == 0) || (flen > len))
+		if ((nfree <= 0) || (flen > len))
 			flen = len;
 		/*
 		 *it is not worth to tx on slow channels:
@@ -1467,7 +1469,7 @@ static int ppp_mp_explode(struct ppp *ppp, struct sk_buff *skb)
 			continue;
 		}
 
-		mtu	= pch->chan->mtu + 2 - hdrlen;
+		mtu	= pch->chan->mtu - hdrlen;
 		if (mtu	< 4)
 			mtu	= 4;
 		if (flen > mtu)
diff --git a/drivers/net/pppoe.c b/drivers/net/pppoe.c
index f0031f1f97e..5f2090233d7 100644
--- a/drivers/net/pppoe.c
+++ b/drivers/net/pppoe.c
@@ -1063,6 +1063,7 @@ static void *pppoe_seq_next(struct seq_file *seq, void *v, loff_t *pos)
 	else {
 		int hash = hash_item(po->pppoe_pa.sid, po->pppoe_pa.remote);
 
+		po = NULL;
 		while (++hash < PPPOE_HASH_SIZE) {
 			po = pn->hash_table[hash];
 			if (po)
diff --git a/drivers/net/pppol2tp.c b/drivers/net/pppol2tp.c
index e7935d09c89..e0f9219a0ae 100644
--- a/drivers/net/pppol2tp.c
+++ b/drivers/net/pppol2tp.c
@@ -2680,6 +2680,7 @@ out_unregister_pppol2tp_proto:
 static void __exit pppol2tp_exit(void)
 {
 	unregister_pppox_proto(PX_PROTO_OL2TP);
+	unregister_pernet_gen_device(pppol2tp_net_id, &pppol2tp_net_ops);
 	proto_unregister(&pppol2tp_sk_proto);
 }
 
diff --git a/drivers/net/s6gmac.c b/drivers/net/s6gmac.c
index 5345e47b35a..4525cbe8dd6 100644
--- a/drivers/net/s6gmac.c
+++ b/drivers/net/s6gmac.c
@@ -793,7 +793,7 @@ static inline int s6gmac_phy_start(struct net_device *dev)
 	struct s6gmac *pd = netdev_priv(dev);
 	int i = 0;
 	struct phy_device *p = NULL;
-	while ((!(p = pd->mii.bus->phy_map[i])) && (i < PHY_MAX_ADDR))
+	while ((i < PHY_MAX_ADDR) && (!(p = pd->mii.bus->phy_map[i])))
 		i++;
 	p = phy_connect(dev, dev_name(&p->dev), &s6gmac_adjust_link, 0,
 			PHY_INTERFACE_MODE_RGMII);
diff --git a/drivers/net/sky2.c b/drivers/net/sky2.c
index 3550c5dcd93..0a551d8f5d9 100644
--- a/drivers/net/sky2.c
+++ b/drivers/net/sky2.c
@@ -1488,6 +1488,8 @@ static int sky2_up(struct net_device *dev)
 	sky2_set_vlan_mode(hw, port, sky2->vlgrp != NULL);
 #endif
 
+	sky2->restarting = 0;
+
 	err = sky2_rx_start(sky2);
 	if (err)
 		goto err_out;
@@ -1500,6 +1502,9 @@ static int sky2_up(struct net_device *dev)
 
 	sky2_set_multicast(dev);
 
+	/* wake queue incase we are restarting */
+	netif_wake_queue(dev);
+
 	if (netif_msg_ifup(sky2))
 		printk(KERN_INFO PFX "%s: enabling interface\n", dev->name);
 	return 0;
@@ -1533,6 +1538,8 @@ static inline int tx_dist(unsigned tail, unsigned head)
 /* Number of list elements available for next tx */
 static inline int tx_avail(const struct sky2_port *sky2)
 {
+	if (unlikely(sky2->restarting))
+		return 0;
 	return sky2->tx_pending - tx_dist(sky2->tx_cons, sky2->tx_prod);
 }
 
@@ -1818,6 +1825,10 @@ static int sky2_down(struct net_device *dev)
 	if (netif_msg_ifdown(sky2))
 		printk(KERN_INFO PFX "%s: disabling interface\n", dev->name);
 
+	/* explicitly shut off tx incase we're restarting */
+	sky2->restarting = 1;
+	netif_tx_disable(dev);
+
 	/* Force flow control off */
 	sky2_write8(hw, SK_REG(port, GMAC_CTRL), GMC_PAUSE_OFF);
 
@@ -2359,7 +2370,7 @@ static inline void sky2_tx_done(struct net_device *dev, u16 last)
 {
 	struct sky2_port *sky2 = netdev_priv(dev);
 
-	if (netif_running(dev)) {
+	if (likely(netif_running(dev) && !sky2->restarting)) {
 		netif_tx_lock(dev);
 		sky2_tx_complete(sky2, last);
 		netif_tx_unlock(dev);
@@ -4283,6 +4294,7 @@ static __devinit struct net_device *sky2_init_netdev(struct sky2_hw *hw,
 	spin_lock_init(&sky2->phy_lock);
 	sky2->tx_pending = TX_DEF_PENDING;
 	sky2->rx_pending = RX_DEF_PENDING;
+	sky2->restarting = 0;
 
 	hw->dev[port] = dev;
 
diff --git a/drivers/net/sky2.h b/drivers/net/sky2.h
index b5549c9e510..4486b066b43 100644
--- a/drivers/net/sky2.h
+++ b/drivers/net/sky2.h
@@ -2051,6 +2051,7 @@ struct sky2_port {
 	u8		     duplex;	/* DUPLEX_HALF, DUPLEX_FULL */
 	u8		     rx_csum;
 	u8		     wol;
+	u8		     restarting;
  	enum flow_control    flow_mode;
  	enum flow_control    flow_status;
 
diff --git a/drivers/net/tulip/de4x5.c b/drivers/net/tulip/de4x5.c
index eb72d2e9ab3..acfdccd4456 100644
--- a/drivers/net/tulip/de4x5.c
+++ b/drivers/net/tulip/de4x5.c
@@ -5059,7 +5059,7 @@ mii_get_phy(struct net_device *dev)
 	if ((id == 0) || (id == 65535)) continue;  /* Valid ID? */
 	for (j=0; j<limit; j++) {                  /* Search PHY table */
 	    if (id != phy_info[j].id) continue;    /* ID match? */
-	    for (k=0; lp->phy[k].id && (k < DE4X5_MAX_PHY); k++);
+	    for (k=0; k < DE4X5_MAX_PHY && lp->phy[k].id; k++);
 	    if (k < DE4X5_MAX_PHY) {
 		memcpy((char *)&lp->phy[k],
 		       (char *)&phy_info[j], sizeof(struct phy_table));
@@ -5072,7 +5072,7 @@ mii_get_phy(struct net_device *dev)
 	    break;
 	}
 	if ((j == limit) && (i < DE4X5_MAX_MII)) {
-	    for (k=0; lp->phy[k].id && (k < DE4X5_MAX_PHY); k++);
+	    for (k=0; k < DE4X5_MAX_PHY && lp->phy[k].id; k++);
 	    lp->phy[k].addr = i;
 	    lp->phy[k].id = id;
 	    lp->phy[k].spd.reg = GENERIC_REG;      /* ANLPA register         */
@@ -5091,7 +5091,7 @@ mii_get_phy(struct net_device *dev)
   purgatory:
     lp->active = 0;
     if (lp->phy[0].id) {                           /* Reset the PHY devices */
-	for (k=0; lp->phy[k].id && (k < DE4X5_MAX_PHY); k++) { /*For each PHY*/
+	for (k=0; k < DE4X5_MAX_PHY && lp->phy[k].id; k++) { /*For each PHY*/
 	    mii_wr(MII_CR_RST, MII_CR, lp->phy[k].addr, DE4X5_MII);
 	    while (mii_rd(MII_CR, lp->phy[k].addr, DE4X5_MII) & MII_CR_RST);
 
diff --git a/drivers/net/wireless/airo.c b/drivers/net/wireless/airo.c
index c70604f0329..8ce5e4cee16 100644
--- a/drivers/net/wireless/airo.c
+++ b/drivers/net/wireless/airo.c
@@ -5918,20 +5918,19 @@ static int airo_set_essid(struct net_device *dev,
 	readSsidRid(local, &SSID_rid);
 
 	/* Check if we asked for `any' */
-	if(dwrq->flags == 0) {
+	if (dwrq->flags == 0) {
 		/* Just send an empty SSID list */
 		memset(&SSID_rid, 0, sizeof(SSID_rid));
 	} else {
-		int	index = (dwrq->flags & IW_ENCODE_INDEX) - 1;
+		unsigned index = (dwrq->flags & IW_ENCODE_INDEX) - 1;
 
 		/* Check the size of the string */
-		if(dwrq->length > IW_ESSID_MAX_SIZE) {
+		if (dwrq->length > IW_ESSID_MAX_SIZE)
 			return -E2BIG ;
-		}
+
 		/* Check if index is valid */
-		if((index < 0) || (index >= 4)) {
+		if (index >= ARRAY_SIZE(SSID_rid.ssids))
 			return -EINVAL;
-		}
 
 		/* Set the SSID */
 		memset(SSID_rid.ssids[index].ssid, 0,
@@ -6819,7 +6818,7 @@ static int airo_set_txpow(struct net_device *dev,
 		return -EINVAL;
 	}
 	clear_bit (FLAG_RADIO_OFF, &local->flags);
-	for (i = 0; cap_rid.txPowerLevels[i] && (i < 8); i++)
+	for (i = 0; i < 8 && cap_rid.txPowerLevels[i]; i++)
 		if (v == cap_rid.txPowerLevels[i]) {
 			readConfigRid(local, 1);
 			local->config.txPower = v;
diff --git a/drivers/net/wireless/ath/ath9k/eeprom.c b/drivers/net/wireless/ath/ath9k/eeprom.c
index a2fda702b62..ce0e86c36a8 100644
--- a/drivers/net/wireless/ath/ath9k/eeprom.c
+++ b/drivers/net/wireless/ath/ath9k/eeprom.c
@@ -460,7 +460,7 @@ static int ath9k_hw_4k_check_eeprom(struct ath_hw *ah)
 		integer = swab32(eep->modalHeader.antCtrlCommon);
 		eep->modalHeader.antCtrlCommon = integer;
 
-		for (i = 0; i < AR5416_MAX_CHAINS; i++) {
+		for (i = 0; i < AR5416_EEP4K_MAX_CHAINS; i++) {
 			integer = swab32(eep->modalHeader.antCtrlChain[i]);
 			eep->modalHeader.antCtrlChain[i] = integer;
 		}
@@ -914,7 +914,7 @@ static void ath9k_hw_set_4k_power_per_rate_table(struct ath_hw *ah,
 			ctlMode, numCtlModes, isHt40CtlMode,
 			(pCtlMode[ctlMode] & EXT_ADDITIVE));
 
-		for (i = 0; (i < AR5416_NUM_CTLS) &&
+		for (i = 0; (i < AR5416_EEP4K_NUM_CTLS) &&
 				pEepData->ctlIndex[i]; i++) {
 			DPRINTF(ah->ah_sc, ATH_DBG_EEPROM,
 				"  LOOP-Ctlidx %d: cfgCtl 0x%2.2x "
diff --git a/drivers/net/wireless/iwlwifi/iwl-3945.h b/drivers/net/wireless/iwlwifi/iwl-3945.h
index fbb3a573463..2de6471d4be 100644
--- a/drivers/net/wireless/iwlwifi/iwl-3945.h
+++ b/drivers/net/wireless/iwlwifi/iwl-3945.h
@@ -112,7 +112,7 @@ enum iwl3945_antenna {
 #define IWL_TX_FIFO_NONE	7
 
 /* Minimum number of queues. MAX_NUM is defined in hw specific files */
-#define IWL_MIN_NUM_QUEUES	4
+#define IWL39_MIN_NUM_QUEUES	4
 
 #define IEEE80211_DATA_LEN              2304
 #define IEEE80211_4ADDR_LEN             30
diff --git a/drivers/net/wireless/iwlwifi/iwl-core.c b/drivers/net/wireless/iwlwifi/iwl-core.c
index 6ab07165ea2..18b135f510e 100644
--- a/drivers/net/wireless/iwlwifi/iwl-core.c
+++ b/drivers/net/wireless/iwlwifi/iwl-core.c
@@ -1332,6 +1332,9 @@ int iwl_setup_mac(struct iwl_priv *priv)
 
 	hw->wiphy->custom_regulatory = true;
 
+	/* Firmware does not support this */
+	hw->wiphy->disable_beacon_hints = true;
+
 	hw->wiphy->max_scan_ssids = PROBE_OPTION_MAX;
 	/* we create the 802.11 header and a zero-length SSID element */
 	hw->wiphy->max_scan_ie_len = IWL_MAX_PROBE_REQUEST - 24 - 2;
diff --git a/drivers/net/wireless/iwlwifi/iwl-debugfs.c b/drivers/net/wireless/iwlwifi/iwl-debugfs.c
index 11e08c06891..ca00cc8ad4c 100644
--- a/drivers/net/wireless/iwlwifi/iwl-debugfs.c
+++ b/drivers/net/wireless/iwlwifi/iwl-debugfs.c
@@ -308,18 +308,18 @@ static ssize_t iwl_dbgfs_nvm_read(struct file *file,
 		return -ENODATA;
 	}
 
+	ptr = priv->eeprom;
+	if (!ptr) {
+		IWL_ERR(priv, "Invalid EEPROM/OTP memory\n");
+		return -ENOMEM;
+	}
+
 	/* 4 characters for byte 0xYY */
 	buf = kzalloc(buf_size, GFP_KERNEL);
 	if (!buf) {
 		IWL_ERR(priv, "Can not allocate Buffer\n");
 		return -ENOMEM;
 	}
-
-	ptr = priv->eeprom;
-	if (!ptr) {
-		IWL_ERR(priv, "Invalid EEPROM/OTP memory\n");
-		return -ENOMEM;
-	}
 	pos += scnprintf(buf + pos, buf_size - pos, "NVM Type: %s\n",
 			(priv->nvm_device_type == NVM_DEVICE_TYPE_OTP)
 			? "OTP" : "EEPROM");
diff --git a/drivers/net/wireless/iwlwifi/iwl-dev.h b/drivers/net/wireless/iwlwifi/iwl-dev.h
index e2d620f0b6e..650e20af20f 100644
--- a/drivers/net/wireless/iwlwifi/iwl-dev.h
+++ b/drivers/net/wireless/iwlwifi/iwl-dev.h
@@ -258,8 +258,10 @@ struct iwl_channel_info {
 #define IWL_TX_FIFO_HCCA_2	6
 #define IWL_TX_FIFO_NONE	7
 
-/* Minimum number of queues. MAX_NUM is defined in hw specific files */
-#define IWL_MIN_NUM_QUEUES	4
+/* Minimum number of queues. MAX_NUM is defined in hw specific files.
+ * Set the minimum to accommodate the 4 standard TX queues, 1 command
+ * queue, 2 (unused) HCCA queues, and 4 HT queues (one for each AC) */
+#define IWL_MIN_NUM_QUEUES	10
 
 /* Power management (not Tx power) structures */
 
diff --git a/drivers/net/wireless/iwlwifi/iwl-sta.c b/drivers/net/wireless/iwlwifi/iwl-sta.c
index 2addf735b19..ffd5c61a755 100644
--- a/drivers/net/wireless/iwlwifi/iwl-sta.c
+++ b/drivers/net/wireless/iwlwifi/iwl-sta.c
@@ -566,6 +566,8 @@ int iwl_remove_default_wep_key(struct iwl_priv *priv,
 	unsigned long flags;
 
 	spin_lock_irqsave(&priv->sta_lock, flags);
+	IWL_DEBUG_WEP(priv, "Removing default WEP key: idx=%d\n",
+		      keyconf->keyidx);
 
 	if (!test_and_clear_bit(keyconf->keyidx, &priv->ucode_key_table))
 		IWL_ERR(priv, "index %d not used in uCode key table.\n",
@@ -573,6 +575,11 @@ int iwl_remove_default_wep_key(struct iwl_priv *priv,
 
 	priv->default_wep_key--;
 	memset(&priv->wep_keys[keyconf->keyidx], 0, sizeof(priv->wep_keys[0]));
+	if (iwl_is_rfkill(priv)) {
+		IWL_DEBUG_WEP(priv, "Not sending REPLY_WEPKEY command due to RFKILL.\n");
+		spin_unlock_irqrestore(&priv->sta_lock, flags);
+		return 0;
+	}
 	ret = iwl_send_static_wepkey_cmd(priv, 1);
 	IWL_DEBUG_WEP(priv, "Remove default WEP key: idx=%d ret=%d\n",
 		      keyconf->keyidx, ret);
@@ -853,6 +860,11 @@ int iwl_remove_dynamic_key(struct iwl_priv *priv,
 	priv->stations[sta_id].sta.sta.modify_mask = STA_MODIFY_KEY_MASK;
 	priv->stations[sta_id].sta.mode = STA_CONTROL_MODIFY_MSK;
 
+	if (iwl_is_rfkill(priv)) {
+		IWL_DEBUG_WEP(priv, "Not sending REPLY_ADD_STA command because RFKILL enabled. \n");
+		spin_unlock_irqrestore(&priv->sta_lock, flags);
+		return 0;
+	}
 	ret =  iwl_send_add_sta(priv, &priv->stations[sta_id].sta, CMD_ASYNC);
 	spin_unlock_irqrestore(&priv->sta_lock, flags);
 	return ret;
diff --git a/drivers/net/wireless/iwlwifi/iwl-tx.c b/drivers/net/wireless/iwlwifi/iwl-tx.c
index 9bbeec9427f..2e89040e63b 100644
--- a/drivers/net/wireless/iwlwifi/iwl-tx.c
+++ b/drivers/net/wireless/iwlwifi/iwl-tx.c
@@ -720,8 +720,6 @@ int iwl_tx_skb(struct iwl_priv *priv, struct sk_buff *skb)
 		goto drop_unlock;
 	}
 
-	spin_unlock_irqrestore(&priv->lock, flags);
-
 	hdr_len = ieee80211_hdrlen(fc);
 
 	/* Find (or create) index into station table for destination station */
@@ -729,7 +727,7 @@ int iwl_tx_skb(struct iwl_priv *priv, struct sk_buff *skb)
 	if (sta_id == IWL_INVALID_STATION) {
 		IWL_DEBUG_DROP(priv, "Dropping - INVALID STATION: %pM\n",
 			       hdr->addr1);
-		goto drop;
+		goto drop_unlock;
 	}
 
 	IWL_DEBUG_TX(priv, "station Id %d\n", sta_id);
@@ -750,14 +748,17 @@ int iwl_tx_skb(struct iwl_priv *priv, struct sk_buff *skb)
 			txq_id = priv->stations[sta_id].tid[tid].agg.txq_id;
 			swq_id = iwl_virtual_agg_queue_num(swq_id, txq_id);
 		}
-		priv->stations[sta_id].tid[tid].tfds_in_queue++;
 	}
 
 	txq = &priv->txq[txq_id];
 	q = &txq->q;
 	txq->swq_id = swq_id;
 
-	spin_lock_irqsave(&priv->lock, flags);
+	if (unlikely(iwl_queue_space(q) < q->high_mark))
+		goto drop_unlock;
+
+	if (ieee80211_is_data_qos(fc))
+		priv->stations[sta_id].tid[tid].tfds_in_queue++;
 
 	/* Set up driver data for this TFD */
 	memset(&(txq->txb[q->write_ptr]), 0, sizeof(struct iwl_tx_info));
@@ -902,7 +903,6 @@ int iwl_tx_skb(struct iwl_priv *priv, struct sk_buff *skb)
 
 drop_unlock:
 	spin_unlock_irqrestore(&priv->lock, flags);
-drop:
 	return -1;
 }
 EXPORT_SYMBOL(iwl_tx_skb);
@@ -1171,6 +1171,8 @@ int iwl_tx_agg_start(struct iwl_priv *priv, const u8 *ra, u16 tid, u16 *ssn)
 		IWL_ERR(priv, "Start AGG on invalid station\n");
 		return -ENXIO;
 	}
+	if (unlikely(tid >= MAX_TID_COUNT))
+		return -EINVAL;
 
 	if (priv->stations[sta_id].tid[tid].agg.state != IWL_AGG_OFF) {
 		IWL_ERR(priv, "Start AGG when state is not IWL_AGG_OFF !\n");
diff --git a/drivers/net/wireless/iwlwifi/iwl3945-base.c b/drivers/net/wireless/iwlwifi/iwl3945-base.c
index 956798f2c80..523843369ca 100644
--- a/drivers/net/wireless/iwlwifi/iwl3945-base.c
+++ b/drivers/net/wireless/iwlwifi/iwl3945-base.c
@@ -3968,6 +3968,9 @@ static int iwl3945_setup_mac(struct iwl_priv *priv)
 
 	hw->wiphy->custom_regulatory = true;
 
+	/* Firmware does not support this */
+	hw->wiphy->disable_beacon_hints = true;
+
 	hw->wiphy->max_scan_ssids = PROBE_OPTION_MAX_3945;
 	/* we create the 802.11 header and a zero-length SSID element */
 	hw->wiphy->max_scan_ie_len = IWL_MAX_PROBE_REQUEST - 24 - 2;
@@ -4018,10 +4021,10 @@ static int iwl3945_pci_probe(struct pci_dev *pdev, const struct pci_device_id *e
 	SET_IEEE80211_DEV(hw, &pdev->dev);
 
 	if ((iwl3945_mod_params.num_of_queues > IWL39_MAX_NUM_QUEUES) ||
-	     (iwl3945_mod_params.num_of_queues < IWL_MIN_NUM_QUEUES)) {
+	     (iwl3945_mod_params.num_of_queues < IWL39_MIN_NUM_QUEUES)) {
 		IWL_ERR(priv,
 			"invalid queues_num, should be between %d and %d\n",
-			IWL_MIN_NUM_QUEUES, IWL39_MAX_NUM_QUEUES);
+			IWL39_MIN_NUM_QUEUES, IWL39_MAX_NUM_QUEUES);
 		err = -EINVAL;
 		goto out_ieee80211_free_hw;
 	}
diff --git a/drivers/net/wireless/iwmc3200wifi/commands.c b/drivers/net/wireless/iwmc3200wifi/commands.c
index 834a7f544e5..e2334d12359 100644
--- a/drivers/net/wireless/iwmc3200wifi/commands.c
+++ b/drivers/net/wireless/iwmc3200wifi/commands.c
@@ -220,6 +220,7 @@ int iwm_store_rxiq_calib_result(struct iwm_priv *iwm)
 	eeprom_rxiq = iwm_eeprom_access(iwm, IWM_EEPROM_CALIB_RXIQ);
 	if (IS_ERR(eeprom_rxiq)) {
 		IWM_ERR(iwm, "Couldn't access EEPROM RX IQ entry\n");
+		kfree(rxiq);
 		return PTR_ERR(eeprom_rxiq);
 	}
 
diff --git a/drivers/net/wireless/iwmc3200wifi/netdev.c b/drivers/net/wireless/iwmc3200wifi/netdev.c
index aea5ccf24cc..bf294e41753 100644
--- a/drivers/net/wireless/iwmc3200wifi/netdev.c
+++ b/drivers/net/wireless/iwmc3200wifi/netdev.c
@@ -106,10 +106,8 @@ void *iwm_if_alloc(int sizeof_bus, struct device *dev,
 	int ret = 0;
 
 	wdev = iwm_wdev_alloc(sizeof_bus, dev);
-	if (!wdev) {
-		dev_err(dev, "no memory for wireless device instance\n");
-		return ERR_PTR(-ENOMEM);
-	}
+	if (IS_ERR(wdev))
+		return wdev;
 
 	iwm = wdev_to_iwm(wdev);
 	iwm->bus_ops = if_ops;
diff --git a/drivers/net/wireless/libertas/11d.c b/drivers/net/wireless/libertas/11d.c
index 9a5408e7d94..5c6968101f0 100644
--- a/drivers/net/wireless/libertas/11d.c
+++ b/drivers/net/wireless/libertas/11d.c
@@ -47,7 +47,7 @@ static u8 lbs_region_2_code(u8 *region)
 {
 	u8 i;
 
-	for (i = 0; region[i] && i < COUNTRY_CODE_LEN; i++)
+	for (i = 0; i < COUNTRY_CODE_LEN && region[i]; i++)
 		region[i] = toupper(region[i]);
 
 	for (i = 0; i < ARRAY_SIZE(region_code_mapping); i++) {
diff --git a/drivers/net/wireless/libertas/assoc.c b/drivers/net/wireless/libertas/assoc.c
index b9b37411903..d6997371c27 100644
--- a/drivers/net/wireless/libertas/assoc.c
+++ b/drivers/net/wireless/libertas/assoc.c
@@ -1,6 +1,7 @@
 /* Copyright (C) 2006, Red Hat, Inc. */
 
 #include <linux/types.h>
+#include <linux/kernel.h>
 #include <linux/etherdevice.h>
 #include <linux/ieee80211.h>
 #include <linux/if_arp.h>
@@ -43,21 +44,21 @@ static int get_common_rates(struct lbs_private *priv,
 	u16 *rates_size)
 {
 	u8 *card_rates = lbs_bg_rates;
-	size_t num_card_rates = sizeof(lbs_bg_rates);
 	int ret = 0, i, j;
-	u8 tmp[30];
+	u8 tmp[(ARRAY_SIZE(lbs_bg_rates) - 1) * (*rates_size - 1)];
 	size_t tmp_size = 0;
 
 	/* For each rate in card_rates that exists in rate1, copy to tmp */
-	for (i = 0; card_rates[i] && (i < num_card_rates); i++) {
-		for (j = 0; rates[j] && (j < *rates_size); j++) {
+	for (i = 0; i < ARRAY_SIZE(lbs_bg_rates) && card_rates[i]; i++) {
+		for (j = 0; j < *rates_size && rates[j]; j++) {
 			if (rates[j] == card_rates[i])
 				tmp[tmp_size++] = card_rates[i];
 		}
 	}
 
 	lbs_deb_hex(LBS_DEB_JOIN, "AP rates    ", rates, *rates_size);
-	lbs_deb_hex(LBS_DEB_JOIN, "card rates  ", card_rates, num_card_rates);
+	lbs_deb_hex(LBS_DEB_JOIN, "card rates  ", card_rates,
+			ARRAY_SIZE(lbs_bg_rates));
 	lbs_deb_hex(LBS_DEB_JOIN, "common rates", tmp, tmp_size);
 	lbs_deb_join("TX data rate 0x%02x\n", priv->cur_rate);
 
@@ -69,10 +70,7 @@ static int get_common_rates(struct lbs_private *priv,
 		lbs_pr_alert("Previously set fixed data rate %#x isn't "
 		       "compatible with the network.\n", priv->cur_rate);
 		ret = -1;
-		goto done;
 	}
-	ret = 0;
-
 done:
 	memset(rates, 0, *rates_size);
 	*rates_size = min_t(int, tmp_size, *rates_size);
@@ -322,7 +320,7 @@ static int lbs_associate(struct lbs_private *priv,
 	rates = (struct mrvl_ie_rates_param_set *) pos;
 	rates->header.type = cpu_to_le16(TLV_TYPE_RATES);
 	memcpy(&rates->rates, &bss->rates, MAX_RATES);
-	tmplen = MAX_RATES;
+	tmplen = min_t(u16, ARRAY_SIZE(rates->rates), MAX_RATES);
 	if (get_common_rates(priv, rates->rates, &tmplen)) {
 		ret = -1;
 		goto done;
@@ -598,7 +596,7 @@ static int lbs_adhoc_join(struct lbs_private *priv,
 
 	/* Copy Data rates from the rates recorded in scan response */
 	memset(cmd.bss.rates, 0, sizeof(cmd.bss.rates));
-	ratesize = min_t(u16, sizeof(cmd.bss.rates), MAX_RATES);
+	ratesize = min_t(u16, ARRAY_SIZE(cmd.bss.rates), MAX_RATES);
 	memcpy(cmd.bss.rates, bss->rates, ratesize);
 	if (get_common_rates(priv, cmd.bss.rates, &ratesize)) {
 		lbs_deb_join("ADHOC_JOIN: get_common_rates returned error.\n");
diff --git a/drivers/net/wireless/libertas/scan.c b/drivers/net/wireless/libertas/scan.c
index 601b5424967..6c95af3023c 100644
--- a/drivers/net/wireless/libertas/scan.c
+++ b/drivers/net/wireless/libertas/scan.c
@@ -5,6 +5,7 @@
   *  for sending scan commands to the firmware.
   */
 #include <linux/types.h>
+#include <linux/kernel.h>
 #include <linux/etherdevice.h>
 #include <linux/if_arp.h>
 #include <asm/unaligned.h>
@@ -876,7 +877,7 @@ static inline char *lbs_translate_scan(struct lbs_private *priv,
 	iwe.u.bitrate.disabled = 0;
 	iwe.u.bitrate.value = 0;
 
-	for (j = 0; bss->rates[j] && (j < sizeof(bss->rates)); j++) {
+	for (j = 0; j < ARRAY_SIZE(bss->rates) && bss->rates[j]; j++) {
 		/* Bit rate given in 500 kb/s units */
 		iwe.u.bitrate.value = bss->rates[j] * 500000;
 		current_val = iwe_stream_add_value(info, start, current_val,
diff --git a/drivers/net/wireless/zd1211rw/zd_mac.c b/drivers/net/wireless/zd1211rw/zd_mac.c
index 40b07b98822..3bd3c779fff 100644
--- a/drivers/net/wireless/zd1211rw/zd_mac.c
+++ b/drivers/net/wireless/zd1211rw/zd_mac.c
@@ -698,7 +698,7 @@ int zd_mac_rx(struct ieee80211_hw *hw, const u8 *buffer, unsigned int length)
 			&& !mac->pass_ctrl)
 		return 0;
 
-	fc = *(__le16 *)buffer;
+	fc = get_unaligned((__le16*)buffer);
 	need_padding = ieee80211_is_data_qos(fc) ^ ieee80211_has_a4(fc);
 
 	skb = dev_alloc_skb(length + (need_padding ? 2 : 0));
diff --git a/drivers/parisc/ccio-dma.c b/drivers/parisc/ccio-dma.c
index 0f0e0b919ef..a45b0c0d574 100644
--- a/drivers/parisc/ccio-dma.c
+++ b/drivers/parisc/ccio-dma.c
@@ -70,7 +70,6 @@
 #undef CCIO_COLLECT_STATS
 #endif
 
-#include <linux/proc_fs.h>
 #include <asm/runway.h>		/* for proc_runway_root */
 
 #ifdef DEBUG_CCIO_INIT
diff --git a/drivers/parisc/dino.c b/drivers/parisc/dino.c
index c590974e981..d69bde6a234 100644
--- a/drivers/parisc/dino.c
+++ b/drivers/parisc/dino.c
@@ -614,7 +614,7 @@ dino_fixup_bus(struct pci_bus *bus)
 			    dev_name(&bus->self->dev), i,
 			    bus->self->resource[i].start,
 			    bus->self->resource[i].end);
-			pci_assign_resource(bus->self, i);
+			WARN_ON(pci_assign_resource(bus->self, i));
 			DBG("DEBUG %s after assign %d [0x%lx,0x%lx]\n",
 			    dev_name(&bus->self->dev), i,
 			    bus->self->resource[i].start,
diff --git a/drivers/parisc/eisa_eeprom.c b/drivers/parisc/eisa_eeprom.c
index 685d94e69d4..8c0b26e9b98 100644
--- a/drivers/parisc/eisa_eeprom.c
+++ b/drivers/parisc/eisa_eeprom.c
@@ -55,7 +55,7 @@ static ssize_t eisa_eeprom_read(struct file * file,
 	ssize_t ret;
 	int i;
 	
-	if (*ppos >= HPEE_MAX_LENGTH)
+	if (*ppos < 0 || *ppos >= HPEE_MAX_LENGTH)
 		return 0;
 	
 	count = *ppos + count < HPEE_MAX_LENGTH ? count : HPEE_MAX_LENGTH - *ppos;
diff --git a/drivers/parisc/hppb.c b/drivers/parisc/hppb.c
index 13856415b43..815db175d42 100644
--- a/drivers/parisc/hppb.c
+++ b/drivers/parisc/hppb.c
@@ -62,7 +62,8 @@ static int hppb_probe(struct parisc_device *dev)
 		}
 		card = card->next;
 	}
-        printk(KERN_INFO "Found GeckoBoa at 0x%x\n", dev->hpa.start);
+	printk(KERN_INFO "Found GeckoBoa at 0x%llx\n",
+			(unsigned long long) dev->hpa.start);
 
 	card->hpa = dev->hpa.start;
 	card->mmio_region.name = "HP-PB Bus";
@@ -73,8 +74,10 @@ static int hppb_probe(struct parisc_device *dev)
 
 	status = ccio_request_resource(dev, &card->mmio_region);
 	if(status < 0) {
-		printk(KERN_ERR "%s: failed to claim HP-PB bus space (%08x, %08x)\n",
-			__FILE__, card->mmio_region.start, card->mmio_region.end);
+		printk(KERN_ERR "%s: failed to claim HP-PB "
+			"bus space (0x%08llx, 0x%08llx)\n",
+			__FILE__, (unsigned long long) card->mmio_region.start,
+			(unsigned long long) card->mmio_region.end);
 	}
 
         return 0;
diff --git a/drivers/parisc/lba_pci.c b/drivers/parisc/lba_pci.c
index ede614616f8..3aeb3279c92 100644
--- a/drivers/parisc/lba_pci.c
+++ b/drivers/parisc/lba_pci.c
@@ -992,7 +992,7 @@ lba_pat_resources(struct parisc_device *pa_dev, struct lba_device *lba_dev)
 		return;
 
 	io_pdc_cell = kzalloc(sizeof(pdc_pat_cell_mod_maddr_block_t), GFP_KERNEL);
-	if (!pa_pdc_cell) {
+	if (!io_pdc_cell) {
 		kfree(pa_pdc_cell);
 		return;
 	}
diff --git a/drivers/parisc/pdc_stable.c b/drivers/parisc/pdc_stable.c
index f9f9a5f1bbd..13a64bc081b 100644
--- a/drivers/parisc/pdc_stable.c
+++ b/drivers/parisc/pdc_stable.c
@@ -370,7 +370,7 @@ pdcspath_layer_read(struct pdcspath_entry *entry, char *buf)
 	if (!i)	/* entry is not ready */
 		return -ENODATA;
 	
-	for (i = 0; devpath->layers[i] && (likely(i < 6)); i++)
+	for (i = 0; i < 6 && devpath->layers[i]; i++)
 		out += sprintf(out, "%u ", devpath->layers[i]);
 
 	out += sprintf(out, "\n");
diff --git a/drivers/pci/intel-iommu.c b/drivers/pci/intel-iommu.c
index ebc9b8dca88..2314ad7ee5f 100644
--- a/drivers/pci/intel-iommu.c
+++ b/drivers/pci/intel-iommu.c
@@ -1505,7 +1505,6 @@ static int domain_context_mapping_one(struct dmar_domain *domain, int segment,
 			}
 
 			set_bit(num, iommu->domain_ids);
-			set_bit(iommu->seq_id, &domain->iommu_bmp);
 			iommu->domains[num] = domain;
 			id = num;
 		}
@@ -1648,6 +1647,14 @@ static int domain_context_mapped(struct pci_dev *pdev)
 					     tmp->devfn);
 }
 
+/* Returns a number of VTD pages, but aligned to MM page size */
+static inline unsigned long aligned_nrpages(unsigned long host_addr,
+					    size_t size)
+{
+	host_addr &= ~PAGE_MASK;
+	return PAGE_ALIGN(host_addr + size) >> VTD_PAGE_SHIFT;
+}
+
 static int __domain_mapping(struct dmar_domain *domain, unsigned long iov_pfn,
 			    struct scatterlist *sg, unsigned long phys_pfn,
 			    unsigned long nr_pages, int prot)
@@ -1675,7 +1682,7 @@ static int __domain_mapping(struct dmar_domain *domain, unsigned long iov_pfn,
 		uint64_t tmp;
 
 		if (!sg_res) {
-			sg_res = (sg->offset + sg->length + VTD_PAGE_SIZE - 1) >> VTD_PAGE_SHIFT;
+			sg_res = aligned_nrpages(sg->offset, sg->length);
 			sg->dma_address = ((dma_addr_t)iov_pfn << VTD_PAGE_SHIFT) + sg->offset;
 			sg->dma_length = sg->length;
 			pteval = page_to_phys(sg_page(sg)) | prot;
@@ -2415,14 +2422,6 @@ error:
 	return ret;
 }
 
-/* Returns a number of VTD pages, but aligned to MM page size */
-static inline unsigned long aligned_nrpages(unsigned long host_addr,
-					    size_t size)
-{
-	host_addr &= ~PAGE_MASK;
-	return PAGE_ALIGN(host_addr + size) >> VTD_PAGE_SHIFT;
-}
-
 /* This takes a number of _MM_ pages, not VTD pages */
 static struct iova *intel_alloc_iova(struct device *dev,
 				     struct dmar_domain *domain,
@@ -2551,6 +2550,7 @@ static dma_addr_t __intel_map_single(struct device *hwdev, phys_addr_t paddr,
 	int prot = 0;
 	int ret;
 	struct intel_iommu *iommu;
+	unsigned long paddr_pfn = paddr >> PAGE_SHIFT;
 
 	BUG_ON(dir == DMA_NONE);
 
@@ -2585,7 +2585,7 @@ static dma_addr_t __intel_map_single(struct device *hwdev, phys_addr_t paddr,
 	 * is not a big problem
 	 */
 	ret = domain_pfn_mapping(domain, mm_to_dma_pfn(iova->pfn_lo),
-				 paddr >> VTD_PAGE_SHIFT, size, prot);
+				 mm_to_dma_pfn(paddr_pfn), size, prot);
 	if (ret)
 		goto error;
 
@@ -2875,7 +2875,7 @@ static int intel_map_sg(struct device *hwdev, struct scatterlist *sglist, int ne
 
 	start_vpfn = mm_to_dma_pfn(iova->pfn_lo);
 
-	ret = domain_sg_mapping(domain, start_vpfn, sglist, mm_to_dma_pfn(size), prot);
+	ret = domain_sg_mapping(domain, start_vpfn, sglist, size, prot);
 	if (unlikely(ret)) {
 		/*  clear the page */
 		dma_pte_clear_range(domain, start_vpfn,
@@ -3408,6 +3408,7 @@ static int md_domain_init(struct dmar_domain *domain, int guest_width)
 
 	domain->iommu_count = 0;
 	domain->iommu_coherency = 0;
+	domain->iommu_snooping = 0;
 	domain->max_addr = 0;
 
 	/* always allocate the top pgd */
diff --git a/drivers/pci/setup-res.c b/drivers/pci/setup-res.c
index b711fb7181e..1898c7b4790 100644
--- a/drivers/pci/setup-res.c
+++ b/drivers/pci/setup-res.c
@@ -100,16 +100,16 @@ int pci_claim_resource(struct pci_dev *dev, int resource)
 {
 	struct resource *res = &dev->resource[resource];
 	struct resource *root;
-	char *dtype = resource < PCI_BRIDGE_RESOURCES ? "device" : "bridge";
 	int err;
 
 	root = pci_find_parent_resource(dev, res);
 
 	err = -EINVAL;
 	if (root != NULL)
-		err = insert_resource(root, res);
+		err = request_resource(root, res);
 
 	if (err) {
+		const char *dtype = resource < PCI_BRIDGE_RESOURCES ? "device" : "bridge";
 		dev_err(&dev->dev, "BAR %d: %s of %s %pR\n",
 			resource,
 			root ? "address space collision on" :
diff --git a/drivers/platform/x86/Kconfig b/drivers/platform/x86/Kconfig
index 46dad12f952..77c6097ced8 100644
--- a/drivers/platform/x86/Kconfig
+++ b/drivers/platform/x86/Kconfig
@@ -277,31 +277,6 @@ config THINKPAD_ACPI_UNSAFE_LEDS
 	  Say N here, unless you are building a kernel for your own
 	  use, and need to control the important firmware LEDs.
 
-config THINKPAD_ACPI_DOCK
-	bool "Legacy Docking Station Support"
-	depends on THINKPAD_ACPI
-	depends on ACPI_DOCK=n
-	default n
-	---help---
-	  Allows the thinkpad_acpi driver to handle docking station events.
-	  This support was made obsolete by the generic ACPI docking station
-	  support (CONFIG_ACPI_DOCK).  It will allow locking and removing the
-	  laptop from the docking station, but will not properly connect PCI
-	  devices.
-
-	  If you are not sure, say N here.
-
-config THINKPAD_ACPI_BAY
-	bool "Legacy Removable Bay Support"
-	depends on THINKPAD_ACPI
-	default y
-	---help---
-	  Allows the thinkpad_acpi driver to handle removable bays.  It will
-	  electrically disable the device in the bay, and also generate
-	  notifications when the bay lever is ejected or inserted.
-
-	  If you are not sure, say Y here.
-
 config THINKPAD_ACPI_VIDEO
 	bool "Video output control support"
 	depends on THINKPAD_ACPI
diff --git a/drivers/platform/x86/eeepc-laptop.c b/drivers/platform/x86/eeepc-laptop.c
index ec560f16d72..222ffb892f2 100644
--- a/drivers/platform/x86/eeepc-laptop.c
+++ b/drivers/platform/x86/eeepc-laptop.c
@@ -143,6 +143,7 @@ struct eeepc_hotk {
 	struct rfkill *bluetooth_rfkill;
 	struct rfkill *wwan3g_rfkill;
 	struct hotplug_slot *hotplug_slot;
+	struct work_struct hotplug_work;
 };
 
 /* The actual device the driver binds to */
@@ -660,7 +661,7 @@ static int eeepc_get_adapter_status(struct hotplug_slot *hotplug_slot,
 	return 0;
 }
 
-static void eeepc_rfkill_hotplug(void)
+static void eeepc_hotplug_work(struct work_struct *work)
 {
 	struct pci_dev *dev;
 	struct pci_bus *bus = pci_find_bus(0, 1);
@@ -701,7 +702,7 @@ static void eeepc_rfkill_notify(acpi_handle handle, u32 event, void *data)
 	if (event != ACPI_NOTIFY_BUS_CHECK)
 		return;
 
-	eeepc_rfkill_hotplug();
+	schedule_work(&ehotk->hotplug_work);
 }
 
 static void eeepc_hotk_notify(struct acpi_device *device, u32 event)
@@ -892,7 +893,7 @@ static int eeepc_hotk_resume(struct acpi_device *device)
 
 		rfkill_set_sw_state(ehotk->wlan_rfkill, wlan != 1);
 
-		eeepc_rfkill_hotplug();
+		schedule_work(&ehotk->hotplug_work);
 	}
 
 	if (ehotk->bluetooth_rfkill)
@@ -1093,6 +1094,8 @@ static int eeepc_rfkill_init(struct device *dev)
 {
 	int result = 0;
 
+	INIT_WORK(&ehotk->hotplug_work, eeepc_hotplug_work);
+
 	eeepc_register_rfkill_notifier("\\_SB.PCI0.P0P6");
 	eeepc_register_rfkill_notifier("\\_SB.PCI0.P0P7");
 
diff --git a/drivers/platform/x86/hp-wmi.c b/drivers/platform/x86/hp-wmi.c
index ca508564a18..a2ad53e1587 100644
--- a/drivers/platform/x86/hp-wmi.c
+++ b/drivers/platform/x86/hp-wmi.c
@@ -520,11 +520,13 @@ static int hp_wmi_resume_handler(struct platform_device *device)
 	 * the input layer will only actually pass it on if the state
 	 * changed.
 	 */
-
-	input_report_switch(hp_wmi_input_dev, SW_DOCK, hp_wmi_dock_state());
-	input_report_switch(hp_wmi_input_dev, SW_TABLET_MODE,
-			    hp_wmi_tablet_state());
-	input_sync(hp_wmi_input_dev);
+	if (hp_wmi_input_dev) {
+		input_report_switch(hp_wmi_input_dev, SW_DOCK,
+				    hp_wmi_dock_state());
+		input_report_switch(hp_wmi_input_dev, SW_TABLET_MODE,
+				    hp_wmi_tablet_state());
+		input_sync(hp_wmi_input_dev);
+	}
 
 	return 0;
 }
diff --git a/drivers/platform/x86/thinkpad_acpi.c b/drivers/platform/x86/thinkpad_acpi.c
index a463fd72c49..e8560085250 100644
--- a/drivers/platform/x86/thinkpad_acpi.c
+++ b/drivers/platform/x86/thinkpad_acpi.c
@@ -239,12 +239,6 @@ struct ibm_init_struct {
 };
 
 static struct {
-#ifdef CONFIG_THINKPAD_ACPI_BAY
-	u32 bay_status:1;
-	u32 bay_eject:1;
-	u32 bay_status2:1;
-	u32 bay_eject2:1;
-#endif
 	u32 bluetooth:1;
 	u32 hotkey:1;
 	u32 hotkey_mask:1;
@@ -589,18 +583,6 @@ static int acpi_ec_write(int i, u8 v)
 	return 1;
 }
 
-#if defined(CONFIG_THINKPAD_ACPI_DOCK) || defined(CONFIG_THINKPAD_ACPI_BAY)
-static int _sta(acpi_handle handle)
-{
-	int status;
-
-	if (!handle || !acpi_evalf(handle, &status, "_STA", "d"))
-		status = 0;
-
-	return status;
-}
-#endif
-
 static int issue_thinkpad_cmos_command(int cmos_cmd)
 {
 	if (!cmos_handle)
@@ -784,6 +766,8 @@ static int dispatch_procfs_write(struct file *file,
 
 	if (!ibm || !ibm->write)
 		return -EINVAL;
+	if (count > PAGE_SIZE - 2)
+		return -EINVAL;
 
 	kernbuf = kmalloc(count + 2, GFP_KERNEL);
 	if (!kernbuf)
@@ -4442,293 +4426,6 @@ static struct ibm_struct light_driver_data = {
 };
 
 /*************************************************************************
- * Dock subdriver
- */
-
-#ifdef CONFIG_THINKPAD_ACPI_DOCK
-
-static void dock_notify(struct ibm_struct *ibm, u32 event);
-static int dock_read(char *p);
-static int dock_write(char *buf);
-
-TPACPI_HANDLE(dock, root, "\\_SB.GDCK",	/* X30, X31, X40 */
-	   "\\_SB.PCI0.DOCK",	/* 600e/x,770e,770x,A2xm/p,T20-22,X20-21 */
-	   "\\_SB.PCI0.PCI1.DOCK",	/* all others */
-	   "\\_SB.PCI.ISA.SLCE",	/* 570 */
-    );				/* A21e,G4x,R30,R31,R32,R40,R40e,R50e */
-
-/* don't list other alternatives as we install a notify handler on the 570 */
-TPACPI_HANDLE(pci, root, "\\_SB.PCI");	/* 570 */
-
-static const struct acpi_device_id ibm_pci_device_ids[] = {
-	{PCI_ROOT_HID_STRING, 0},
-	{"", 0},
-};
-
-static struct tp_acpi_drv_struct ibm_dock_acpidriver[2] = {
-	{
-	 .notify = dock_notify,
-	 .handle = &dock_handle,
-	 .type = ACPI_SYSTEM_NOTIFY,
-	},
-	{
-	/* THIS ONE MUST NEVER BE USED FOR DRIVER AUTOLOADING.
-	 * We just use it to get notifications of dock hotplug
-	 * in very old thinkpads */
-	 .hid = ibm_pci_device_ids,
-	 .notify = dock_notify,
-	 .handle = &pci_handle,
-	 .type = ACPI_SYSTEM_NOTIFY,
-	},
-};
-
-static struct ibm_struct dock_driver_data[2] = {
-	{
-	 .name = "dock",
-	 .read = dock_read,
-	 .write = dock_write,
-	 .acpi = &ibm_dock_acpidriver[0],
-	},
-	{
-	 .name = "dock",
-	 .acpi = &ibm_dock_acpidriver[1],
-	},
-};
-
-#define dock_docked() (_sta(dock_handle) & 1)
-
-static int __init dock_init(struct ibm_init_struct *iibm)
-{
-	vdbg_printk(TPACPI_DBG_INIT, "initializing dock subdriver\n");
-
-	TPACPI_ACPIHANDLE_INIT(dock);
-
-	vdbg_printk(TPACPI_DBG_INIT, "dock is %s\n",
-		str_supported(dock_handle != NULL));
-
-	return (dock_handle)? 0 : 1;
-}
-
-static int __init dock_init2(struct ibm_init_struct *iibm)
-{
-	int dock2_needed;
-
-	vdbg_printk(TPACPI_DBG_INIT, "initializing dock subdriver part 2\n");
-
-	if (dock_driver_data[0].flags.acpi_driver_registered &&
-	    dock_driver_data[0].flags.acpi_notify_installed) {
-		TPACPI_ACPIHANDLE_INIT(pci);
-		dock2_needed = (pci_handle != NULL);
-		vdbg_printk(TPACPI_DBG_INIT,
-			    "dock PCI handler for the TP 570 is %s\n",
-			    str_supported(dock2_needed));
-	} else {
-		vdbg_printk(TPACPI_DBG_INIT,
-		"dock subdriver part 2 not required\n");
-		dock2_needed = 0;
-	}
-
-	return (dock2_needed)? 0 : 1;
-}
-
-static void dock_notify(struct ibm_struct *ibm, u32 event)
-{
-	int docked = dock_docked();
-	int pci = ibm->acpi->hid && ibm->acpi->device &&
-		acpi_match_device_ids(ibm->acpi->device, ibm_pci_device_ids);
-	int data;
-
-	if (event == 1 && !pci)	/* 570 */
-		data = 1;	/* button */
-	else if (event == 1 && pci)	/* 570 */
-		data = 3;	/* dock */
-	else if (event == 3 && docked)
-		data = 1;	/* button */
-	else if (event == 3 && !docked)
-		data = 2;	/* undock */
-	else if (event == 0 && docked)
-		data = 3;	/* dock */
-	else {
-		printk(TPACPI_ERR "unknown dock event %d, status %d\n",
-		       event, _sta(dock_handle));
-		data = 0;	/* unknown */
-	}
-	acpi_bus_generate_proc_event(ibm->acpi->device, event, data);
-	acpi_bus_generate_netlink_event(ibm->acpi->device->pnp.device_class,
-					  dev_name(&ibm->acpi->device->dev),
-					  event, data);
-}
-
-static int dock_read(char *p)
-{
-	int len = 0;
-	int docked = dock_docked();
-
-	if (!dock_handle)
-		len += sprintf(p + len, "status:\t\tnot supported\n");
-	else if (!docked)
-		len += sprintf(p + len, "status:\t\tundocked\n");
-	else {
-		len += sprintf(p + len, "status:\t\tdocked\n");
-		len += sprintf(p + len, "commands:\tdock, undock\n");
-	}
-
-	return len;
-}
-
-static int dock_write(char *buf)
-{
-	char *cmd;
-
-	if (!dock_docked())
-		return -ENODEV;
-
-	while ((cmd = next_cmd(&buf))) {
-		if (strlencmp(cmd, "undock") == 0) {
-			if (!acpi_evalf(dock_handle, NULL, "_DCK", "vd", 0) ||
-			    !acpi_evalf(dock_handle, NULL, "_EJ0", "vd", 1))
-				return -EIO;
-		} else if (strlencmp(cmd, "dock") == 0) {
-			if (!acpi_evalf(dock_handle, NULL, "_DCK", "vd", 1))
-				return -EIO;
-		} else
-			return -EINVAL;
-	}
-
-	return 0;
-}
-
-#endif /* CONFIG_THINKPAD_ACPI_DOCK */
-
-/*************************************************************************
- * Bay subdriver
- */
-
-#ifdef CONFIG_THINKPAD_ACPI_BAY
-
-TPACPI_HANDLE(bay, root, "\\_SB.PCI.IDE.SECN.MAST",	/* 570 */
-	   "\\_SB.PCI0.IDE0.IDES.IDSM",	/* 600e/x, 770e, 770x */
-	   "\\_SB.PCI0.SATA.SCND.MSTR",	/* T60, X60, Z60 */
-	   "\\_SB.PCI0.IDE0.SCND.MSTR",	/* all others */
-	   );				/* A21e, R30, R31 */
-TPACPI_HANDLE(bay_ej, bay, "_EJ3",	/* 600e/x, A2xm/p, A3x */
-	   "_EJ0",		/* all others */
-	   );			/* 570,A21e,G4x,R30,R31,R32,R40e,R50e */
-TPACPI_HANDLE(bay2, root, "\\_SB.PCI0.IDE0.PRIM.SLAV",	/* A3x, R32 */
-	   "\\_SB.PCI0.IDE0.IDEP.IDPS",	/* 600e/x, 770e, 770x */
-	   );				/* all others */
-TPACPI_HANDLE(bay2_ej, bay2, "_EJ3",	/* 600e/x, 770e, A3x */
-	   "_EJ0",			/* 770x */
-	   );				/* all others */
-
-static int __init bay_init(struct ibm_init_struct *iibm)
-{
-	vdbg_printk(TPACPI_DBG_INIT, "initializing bay subdriver\n");
-
-	TPACPI_ACPIHANDLE_INIT(bay);
-	if (bay_handle)
-		TPACPI_ACPIHANDLE_INIT(bay_ej);
-	TPACPI_ACPIHANDLE_INIT(bay2);
-	if (bay2_handle)
-		TPACPI_ACPIHANDLE_INIT(bay2_ej);
-
-	tp_features.bay_status = bay_handle &&
-		acpi_evalf(bay_handle, NULL, "_STA", "qv");
-	tp_features.bay_status2 = bay2_handle &&
-		acpi_evalf(bay2_handle, NULL, "_STA", "qv");
-
-	tp_features.bay_eject = bay_handle && bay_ej_handle &&
-		(strlencmp(bay_ej_path, "_EJ0") == 0 || experimental);
-	tp_features.bay_eject2 = bay2_handle && bay2_ej_handle &&
-		(strlencmp(bay2_ej_path, "_EJ0") == 0 || experimental);
-
-	vdbg_printk(TPACPI_DBG_INIT,
-		"bay 1: status %s, eject %s; bay 2: status %s, eject %s\n",
-		str_supported(tp_features.bay_status),
-		str_supported(tp_features.bay_eject),
-		str_supported(tp_features.bay_status2),
-		str_supported(tp_features.bay_eject2));
-
-	return (tp_features.bay_status || tp_features.bay_eject ||
-		tp_features.bay_status2 || tp_features.bay_eject2)? 0 : 1;
-}
-
-static void bay_notify(struct ibm_struct *ibm, u32 event)
-{
-	acpi_bus_generate_proc_event(ibm->acpi->device, event, 0);
-	acpi_bus_generate_netlink_event(ibm->acpi->device->pnp.device_class,
-					  dev_name(&ibm->acpi->device->dev),
-					  event, 0);
-}
-
-#define bay_occupied(b) (_sta(b##_handle) & 1)
-
-static int bay_read(char *p)
-{
-	int len = 0;
-	int occupied = bay_occupied(bay);
-	int occupied2 = bay_occupied(bay2);
-	int eject, eject2;
-
-	len += sprintf(p + len, "status:\t\t%s\n",
-		tp_features.bay_status ?
-			(occupied ? "occupied" : "unoccupied") :
-				"not supported");
-	if (tp_features.bay_status2)
-		len += sprintf(p + len, "status2:\t%s\n", occupied2 ?
-			       "occupied" : "unoccupied");
-
-	eject = tp_features.bay_eject && occupied;
-	eject2 = tp_features.bay_eject2 && occupied2;
-
-	if (eject && eject2)
-		len += sprintf(p + len, "commands:\teject, eject2\n");
-	else if (eject)
-		len += sprintf(p + len, "commands:\teject\n");
-	else if (eject2)
-		len += sprintf(p + len, "commands:\teject2\n");
-
-	return len;
-}
-
-static int bay_write(char *buf)
-{
-	char *cmd;
-
-	if (!tp_features.bay_eject && !tp_features.bay_eject2)
-		return -ENODEV;
-
-	while ((cmd = next_cmd(&buf))) {
-		if (tp_features.bay_eject && strlencmp(cmd, "eject") == 0) {
-			if (!acpi_evalf(bay_ej_handle, NULL, NULL, "vd", 1))
-				return -EIO;
-		} else if (tp_features.bay_eject2 &&
-			   strlencmp(cmd, "eject2") == 0) {
-			if (!acpi_evalf(bay2_ej_handle, NULL, NULL, "vd", 1))
-				return -EIO;
-		} else
-			return -EINVAL;
-	}
-
-	return 0;
-}
-
-static struct tp_acpi_drv_struct ibm_bay_acpidriver = {
-	.notify = bay_notify,
-	.handle = &bay_handle,
-	.type = ACPI_SYSTEM_NOTIFY,
-};
-
-static struct ibm_struct bay_driver_data = {
-	.name = "bay",
-	.read = bay_read,
-	.write = bay_write,
-	.acpi = &ibm_bay_acpidriver,
-};
-
-#endif /* CONFIG_THINKPAD_ACPI_BAY */
-
-/*************************************************************************
  * CMOS subdriver
  */
 
@@ -5945,14 +5642,48 @@ static struct backlight_ops ibm_backlight_data = {
 
 /* --------------------------------------------------------------------- */
 
+/*
+ * These are only useful for models that have only one possibility
+ * of GPU.  If the BIOS model handles both ATI and Intel, don't use
+ * these quirks.
+ */
+#define TPACPI_BRGHT_Q_NOEC	0x0001	/* Must NOT use EC HBRV */
+#define TPACPI_BRGHT_Q_EC	0x0002  /* Should or must use EC HBRV */
+#define TPACPI_BRGHT_Q_ASK	0x8000	/* Ask for user report */
+
+static const struct tpacpi_quirk brightness_quirk_table[] __initconst = {
+	/* Models with ATI GPUs known to require ECNVRAM mode */
+	TPACPI_Q_IBM('1', 'Y', TPACPI_BRGHT_Q_EC),	/* T43/p ATI */
+
+	/* Models with ATI GPUs (waiting confirmation) */
+	TPACPI_Q_IBM('1', 'R', TPACPI_BRGHT_Q_ASK|TPACPI_BRGHT_Q_EC),
+	TPACPI_Q_IBM('1', 'Q', TPACPI_BRGHT_Q_ASK|TPACPI_BRGHT_Q_EC),
+	TPACPI_Q_IBM('7', '6', TPACPI_BRGHT_Q_ASK|TPACPI_BRGHT_Q_EC),
+	TPACPI_Q_IBM('7', '8', TPACPI_BRGHT_Q_ASK|TPACPI_BRGHT_Q_EC),
+
+	/* Models with Intel Extreme Graphics 2 (waiting confirmation) */
+	TPACPI_Q_IBM('1', 'V', TPACPI_BRGHT_Q_ASK|TPACPI_BRGHT_Q_NOEC),
+	TPACPI_Q_IBM('1', 'W', TPACPI_BRGHT_Q_ASK|TPACPI_BRGHT_Q_NOEC),
+	TPACPI_Q_IBM('1', 'U', TPACPI_BRGHT_Q_ASK|TPACPI_BRGHT_Q_NOEC),
+
+	/* Models with Intel GMA900 */
+	TPACPI_Q_IBM('7', '0', TPACPI_BRGHT_Q_NOEC),	/* T43, R52 */
+	TPACPI_Q_IBM('7', '4', TPACPI_BRGHT_Q_NOEC),	/* X41 */
+	TPACPI_Q_IBM('7', '5', TPACPI_BRGHT_Q_NOEC),	/* X41 Tablet */
+};
+
 static int __init brightness_init(struct ibm_init_struct *iibm)
 {
 	int b;
+	unsigned long quirks;
 
 	vdbg_printk(TPACPI_DBG_INIT, "initializing brightness subdriver\n");
 
 	mutex_init(&brightness_mutex);
 
+	quirks = tpacpi_check_quirks(brightness_quirk_table,
+				ARRAY_SIZE(brightness_quirk_table));
+
 	/*
 	 * We always attempt to detect acpi support, so as to switch
 	 * Lenovo Vista BIOS to ACPI brightness mode even if we are not
@@ -6009,23 +5740,13 @@ static int __init brightness_init(struct ibm_init_struct *iibm)
 	/* TPACPI_BRGHT_MODE_AUTO not implemented yet, just use default */
 	if (brightness_mode == TPACPI_BRGHT_MODE_AUTO ||
 	    brightness_mode == TPACPI_BRGHT_MODE_MAX) {
-		if (thinkpad_id.vendor == PCI_VENDOR_ID_IBM) {
-			/*
-			 * IBM models that define HBRV probably have
-			 * EC-based backlight level control
-			 */
-			if (acpi_evalf(ec_handle, NULL, "HBRV", "qd"))
-				/* T40-T43, R50-R52, R50e, R51e, X31-X41 */
-				brightness_mode = TPACPI_BRGHT_MODE_ECNVRAM;
-			else
-				/* all other IBM ThinkPads */
-				brightness_mode = TPACPI_BRGHT_MODE_UCMS_STEP;
-		} else
-			/* All Lenovo ThinkPads */
+		if (quirks & TPACPI_BRGHT_Q_EC)
+			brightness_mode = TPACPI_BRGHT_MODE_ECNVRAM;
+		else
 			brightness_mode = TPACPI_BRGHT_MODE_UCMS_STEP;
 
 		dbg_printk(TPACPI_DBG_BRGHT,
-			   "selected brightness_mode=%d\n",
+			   "driver auto-selected brightness_mode=%d\n",
 			   brightness_mode);
 	}
 
@@ -6052,6 +5773,15 @@ static int __init brightness_init(struct ibm_init_struct *iibm)
 	vdbg_printk(TPACPI_DBG_INIT | TPACPI_DBG_BRGHT,
 			"brightness is supported\n");
 
+	if (quirks & TPACPI_BRGHT_Q_ASK) {
+		printk(TPACPI_NOTICE
+			"brightness: will use unverified default: "
+			"brightness_mode=%d\n", brightness_mode);
+		printk(TPACPI_NOTICE
+			"brightness: please report to %s whether it works well "
+			"or not on your ThinkPad\n", TPACPI_MAIL);
+	}
+
 	ibm_backlight_device->props.max_brightness =
 				(tp_features.bright_16levels)? 15 : 7;
 	ibm_backlight_device->props.brightness = b & TP_EC_BACKLIGHT_LVLMSK;
@@ -7854,22 +7584,6 @@ static struct ibm_init_struct ibms_init[] __initdata = {
 		.init = light_init,
 		.data = &light_driver_data,
 	},
-#ifdef CONFIG_THINKPAD_ACPI_DOCK
-	{
-		.init = dock_init,
-		.data = &dock_driver_data[0],
-	},
-	{
-		.init = dock_init2,
-		.data = &dock_driver_data[1],
-	},
-#endif
-#ifdef CONFIG_THINKPAD_ACPI_BAY
-	{
-		.init = bay_init,
-		.data = &bay_driver_data,
-	},
-#endif
 	{
 		.init = cmos_init,
 		.data = &cmos_driver_data,
@@ -7968,12 +7682,6 @@ TPACPI_PARAM(hotkey);
 TPACPI_PARAM(bluetooth);
 TPACPI_PARAM(video);
 TPACPI_PARAM(light);
-#ifdef CONFIG_THINKPAD_ACPI_DOCK
-TPACPI_PARAM(dock);
-#endif
-#ifdef CONFIG_THINKPAD_ACPI_BAY
-TPACPI_PARAM(bay);
-#endif /* CONFIG_THINKPAD_ACPI_BAY */
 TPACPI_PARAM(cmos);
 TPACPI_PARAM(led);
 TPACPI_PARAM(beep);
diff --git a/drivers/power/Kconfig b/drivers/power/Kconfig
index 7eda34838bf..bdbc4f73fcd 100644
--- a/drivers/power/Kconfig
+++ b/drivers/power/Kconfig
@@ -43,6 +43,13 @@ config BATTERY_DS2760
 	help
 	  Say Y here to enable support for batteries with ds2760 chip.
 
+config BATTERY_DS2782
+	tristate "DS2782 standalone gas-gauge"
+	depends on I2C
+	help
+	  Say Y here to enable support for the DS2782 standalone battery
+	  gas-gauge.
+
 config BATTERY_PMU
 	tristate "Apple PMU battery"
 	depends on PPC32 && ADB_PMU
diff --git a/drivers/power/Makefile b/drivers/power/Makefile
index daf3179689a..380d17c9ae2 100644
--- a/drivers/power/Makefile
+++ b/drivers/power/Makefile
@@ -19,6 +19,7 @@ obj-$(CONFIG_APM_POWER)		+= apm_power.o
 obj-$(CONFIG_WM8350_POWER)	+= wm8350_power.o
 
 obj-$(CONFIG_BATTERY_DS2760)	+= ds2760_battery.o
+obj-$(CONFIG_BATTERY_DS2782)	+= ds2782_battery.o
 obj-$(CONFIG_BATTERY_PMU)	+= pmu_battery.o
 obj-$(CONFIG_BATTERY_OLPC)	+= olpc_battery.o
 obj-$(CONFIG_BATTERY_TOSA)	+= tosa_battery.o
diff --git a/drivers/power/ds2782_battery.c b/drivers/power/ds2782_battery.c
new file mode 100644
index 00000000000..da14f374cb6
--- /dev/null
+++ b/drivers/power/ds2782_battery.c
@@ -0,0 +1,330 @@
+/*
+ * I2C client/driver for the Maxim/Dallas DS2782 Stand-Alone Fuel Gauge IC
+ *
+ * Copyright (C) 2009 Bluewater Systems Ltd
+ *
+ * Author: Ryan Mallon <ryan@bluewatersys.com>
+ *
+ * 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/kernel.h>
+#include <linux/module.h>
+#include <linux/types.h>
+#include <linux/errno.h>
+#include <linux/swab.h>
+#include <linux/i2c.h>
+#include <linux/idr.h>
+#include <linux/power_supply.h>
+
+#define DS2782_REG_RARC		0x06	/* Remaining active relative capacity */
+
+#define DS2782_REG_VOLT_MSB	0x0c
+#define DS2782_REG_TEMP_MSB	0x0a
+#define DS2782_REG_CURRENT_MSB	0x0e
+
+/* EEPROM Block */
+#define DS2782_REG_RSNSP	0x69	/* Sense resistor value */
+
+/* Current unit measurement in uA for a 1 milli-ohm sense resistor */
+#define DS2782_CURRENT_UNITS	1563
+
+#define to_ds2782_info(x) container_of(x, struct ds2782_info, battery)
+
+struct ds2782_info {
+	struct i2c_client	*client;
+	struct power_supply	battery;
+	int			id;
+};
+
+static DEFINE_IDR(battery_id);
+static DEFINE_MUTEX(battery_lock);
+
+static inline int ds2782_read_reg(struct ds2782_info *info, int reg, u8 *val)
+{
+	int ret;
+
+	ret = i2c_smbus_read_byte_data(info->client, reg);
+	if (ret < 0) {
+		dev_err(&info->client->dev, "register read failed\n");
+		return ret;
+	}
+
+	*val = ret;
+	return 0;
+}
+
+static inline int ds2782_read_reg16(struct ds2782_info *info, int reg_msb,
+				    s16 *val)
+{
+	int ret;
+
+	ret = swab16(i2c_smbus_read_word_data(info->client, reg_msb));
+	if (ret < 0) {
+		dev_err(&info->client->dev, "register read failed\n");
+		return ret;
+	}
+
+	*val = ret;
+	return 0;
+}
+
+static int ds2782_get_temp(struct ds2782_info *info, int *temp)
+{
+	s16 raw;
+	int err;
+
+	/*
+	 * Temperature is measured in units of 0.125 degrees celcius, the
+	 * power_supply class measures temperature in tenths of degrees
+	 * celsius. The temperature value is stored as a 10 bit number, plus
+	 * sign in the upper bits of a 16 bit register.
+	 */
+	err = ds2782_read_reg16(info, DS2782_REG_TEMP_MSB, &raw);
+	if (err)
+		return err;
+	*temp = ((raw / 32) * 125) / 100;
+	return 0;
+}
+
+static int ds2782_get_current(struct ds2782_info *info, int *current_uA)
+{
+	int sense_res;
+	int err;
+	u8 sense_res_raw;
+	s16 raw;
+
+	/*
+	 * The units of measurement for current are dependent on the value of
+	 * the sense resistor.
+	 */
+	err = ds2782_read_reg(info, DS2782_REG_RSNSP, &sense_res_raw);
+	if (err)
+		return err;
+	if (sense_res_raw == 0) {
+		dev_err(&info->client->dev, "sense resistor value is 0\n");
+		return -ENXIO;
+	}
+	sense_res = 1000 / sense_res_raw;
+
+	dev_dbg(&info->client->dev, "sense resistor = %d milli-ohms\n",
+		sense_res);
+	err = ds2782_read_reg16(info, DS2782_REG_CURRENT_MSB, &raw);
+	if (err)
+		return err;
+	*current_uA = raw * (DS2782_CURRENT_UNITS / sense_res);
+	return 0;
+}
+
+static int ds2782_get_voltage(struct ds2782_info *info, int *voltage_uA)
+{
+	s16 raw;
+	int err;
+
+	/*
+	 * Voltage is measured in units of 4.88mV. The voltage is stored as
+	 * a 10-bit number plus sign, in the upper bits of a 16-bit register
+	 */
+	err = ds2782_read_reg16(info, DS2782_REG_VOLT_MSB, &raw);
+	if (err)
+		return err;
+	*voltage_uA = (raw / 32) * 4800;
+	return 0;
+}
+
+static int ds2782_get_capacity(struct ds2782_info *info, int *capacity)
+{
+	int err;
+	u8 raw;
+
+	err = ds2782_read_reg(info, DS2782_REG_RARC, &raw);
+	if (err)
+		return err;
+	*capacity = raw;
+	return raw;
+}
+
+static int ds2782_get_status(struct ds2782_info *info, int *status)
+{
+	int err;
+	int current_uA;
+	int capacity;
+
+	err = ds2782_get_current(info, &current_uA);
+	if (err)
+		return err;
+
+	err = ds2782_get_capacity(info, &capacity);
+	if (err)
+		return err;
+
+	if (capacity == 100)
+		*status = POWER_SUPPLY_STATUS_FULL;
+	else if (current_uA == 0)
+		*status = POWER_SUPPLY_STATUS_NOT_CHARGING;
+	else if (current_uA < 0)
+		*status = POWER_SUPPLY_STATUS_DISCHARGING;
+	else
+		*status = POWER_SUPPLY_STATUS_CHARGING;
+
+	return 0;
+}
+
+static int ds2782_battery_get_property(struct power_supply *psy,
+				       enum power_supply_property prop,
+				       union power_supply_propval *val)
+{
+	struct ds2782_info *info = to_ds2782_info(psy);
+	int ret;
+
+	switch (prop) {
+	case POWER_SUPPLY_PROP_STATUS:
+		ret = ds2782_get_status(info, &val->intval);
+		break;
+
+	case POWER_SUPPLY_PROP_CAPACITY:
+		ret = ds2782_get_capacity(info, &val->intval);
+		break;
+
+	case POWER_SUPPLY_PROP_VOLTAGE_NOW:
+		ret = ds2782_get_voltage(info, &val->intval);
+		break;
+
+	case POWER_SUPPLY_PROP_CURRENT_NOW:
+		ret = ds2782_get_current(info, &val->intval);
+		break;
+
+	case POWER_SUPPLY_PROP_TEMP:
+		ret = ds2782_get_temp(info, &val->intval);
+		break;
+
+	default:
+		ret = -EINVAL;
+	}
+
+	return ret;
+}
+
+static enum power_supply_property ds2782_battery_props[] = {
+	POWER_SUPPLY_PROP_STATUS,
+	POWER_SUPPLY_PROP_CAPACITY,
+	POWER_SUPPLY_PROP_VOLTAGE_NOW,
+	POWER_SUPPLY_PROP_CURRENT_NOW,
+	POWER_SUPPLY_PROP_TEMP,
+};
+
+static void ds2782_power_supply_init(struct power_supply *battery)
+{
+	battery->type			= POWER_SUPPLY_TYPE_BATTERY;
+	battery->properties		= ds2782_battery_props;
+	battery->num_properties		= ARRAY_SIZE(ds2782_battery_props);
+	battery->get_property		= ds2782_battery_get_property;
+	battery->external_power_changed	= NULL;
+}
+
+static int ds2782_battery_remove(struct i2c_client *client)
+{
+	struct ds2782_info *info = i2c_get_clientdata(client);
+
+	power_supply_unregister(&info->battery);
+	kfree(info->battery.name);
+
+	mutex_lock(&battery_lock);
+	idr_remove(&battery_id, info->id);
+	mutex_unlock(&battery_lock);
+
+	i2c_set_clientdata(client, info);
+
+	kfree(info);
+	return 0;
+}
+
+static int ds2782_battery_probe(struct i2c_client *client,
+				const struct i2c_device_id *id)
+{
+	struct ds2782_info *info;
+	int ret;
+	int num;
+
+	/* Get an ID for this battery */
+	ret = idr_pre_get(&battery_id, GFP_KERNEL);
+	if (ret == 0) {
+		ret = -ENOMEM;
+		goto fail_id;
+	}
+
+	mutex_lock(&battery_lock);
+	ret = idr_get_new(&battery_id, client, &num);
+	mutex_unlock(&battery_lock);
+	if (ret < 0)
+		goto fail_id;
+
+	info = kzalloc(sizeof(*info), GFP_KERNEL);
+	if (!info) {
+		ret = -ENOMEM;
+		goto fail_info;
+	}
+
+	info->battery.name = kasprintf(GFP_KERNEL, "ds2782-%d", num);
+	if (!info->battery.name) {
+		ret = -ENOMEM;
+		goto fail_name;
+	}
+
+	i2c_set_clientdata(client, info);
+	info->client = client;
+	ds2782_power_supply_init(&info->battery);
+
+	ret = power_supply_register(&client->dev, &info->battery);
+	if (ret) {
+		dev_err(&client->dev, "failed to register battery\n");
+		goto fail_register;
+	}
+
+	return 0;
+
+fail_register:
+	kfree(info->battery.name);
+fail_name:
+	i2c_set_clientdata(client, info);
+	kfree(info);
+fail_info:
+	mutex_lock(&battery_lock);
+	idr_remove(&battery_id, num);
+	mutex_unlock(&battery_lock);
+fail_id:
+	return ret;
+}
+
+static const struct i2c_device_id ds2782_id[] = {
+	{"ds2782", 0},
+	{},
+};
+
+static struct i2c_driver ds2782_battery_driver = {
+	.driver 	= {
+		.name	= "ds2782-battery",
+	},
+	.probe		= ds2782_battery_probe,
+	.remove		= ds2782_battery_remove,
+	.id_table	= ds2782_id,
+};
+
+static int __init ds2782_init(void)
+{
+	return i2c_add_driver(&ds2782_battery_driver);
+}
+module_init(ds2782_init);
+
+static void __exit ds2782_exit(void)
+{
+	i2c_del_driver(&ds2782_battery_driver);
+}
+module_exit(ds2782_exit);
+
+MODULE_AUTHOR("Ryan Mallon <ryan@bluewatersys.com>");
+MODULE_DESCRIPTION("Maxim/Dallas DS2782 Stand-Alone Fuel Gauage IC driver");
+MODULE_LICENSE("GPL");
diff --git a/drivers/power/olpc_battery.c b/drivers/power/olpc_battery.c
index 5fbca2681ba..58e419299cd 100644
--- a/drivers/power/olpc_battery.c
+++ b/drivers/power/olpc_battery.c
@@ -8,6 +8,7 @@
  * published by the Free Software Foundation.
  */
 
+#include <linux/kernel.h>
 #include <linux/module.h>
 #include <linux/err.h>
 #include <linux/platform_device.h>
@@ -35,6 +36,7 @@
 #define BAT_STAT_AC		0x10
 #define BAT_STAT_CHARGING	0x20
 #define BAT_STAT_DISCHARGING	0x40
+#define BAT_STAT_TRICKLE	0x80
 
 #define BAT_ERR_INFOFAIL	0x02
 #define BAT_ERR_OVERVOLTAGE	0x04
@@ -89,7 +91,7 @@ static char bat_serial[17]; /* Ick */
 static int olpc_bat_get_status(union power_supply_propval *val, uint8_t ec_byte)
 {
 	if (olpc_platform_info.ecver > 0x44) {
-		if (ec_byte & BAT_STAT_CHARGING)
+		if (ec_byte & (BAT_STAT_CHARGING | BAT_STAT_TRICKLE))
 			val->intval = POWER_SUPPLY_STATUS_CHARGING;
 		else if (ec_byte & BAT_STAT_DISCHARGING)
 			val->intval = POWER_SUPPLY_STATUS_DISCHARGING;
@@ -219,7 +221,8 @@ static int olpc_bat_get_property(struct power_supply *psy,
 	   It doesn't matter though -- the EC will return the last-known
 	   information, and it's as if we just ran that _little_ bit faster
 	   and managed to read it out before the battery went away. */
-	if (!(ec_byte & BAT_STAT_PRESENT) && psp != POWER_SUPPLY_PROP_PRESENT)
+	if (!(ec_byte & (BAT_STAT_PRESENT | BAT_STAT_TRICKLE)) &&
+			psp != POWER_SUPPLY_PROP_PRESENT)
 		return -ENODEV;
 
 	switch (psp) {
@@ -229,7 +232,8 @@ static int olpc_bat_get_property(struct power_supply *psy,
 			return ret;
 		break;
 	case POWER_SUPPLY_PROP_PRESENT:
-		val->intval = !!(ec_byte & BAT_STAT_PRESENT);
+		val->intval = !!(ec_byte & (BAT_STAT_PRESENT |
+					    BAT_STAT_TRICKLE));
 		break;
 
 	case POWER_SUPPLY_PROP_HEALTH:
@@ -334,21 +338,21 @@ static ssize_t olpc_bat_eeprom_read(struct kobject *kobj,
 		struct bin_attribute *attr, char *buf, loff_t off, size_t count)
 {
 	uint8_t ec_byte;
-	int ret, end;
+	int ret;
+	int i;
 
 	if (off >= EEPROM_SIZE)
 		return 0;
 	if (off + count > EEPROM_SIZE)
 		count = EEPROM_SIZE - off;
 
-	end = EEPROM_START + off + count;
-	for (ec_byte = EEPROM_START + off; ec_byte < end; ec_byte++) {
-		ret = olpc_ec_cmd(EC_BAT_EEPROM, &ec_byte, 1,
-				&buf[ec_byte - EEPROM_START], 1);
+	for (i = 0; i < count; i++) {
+		ec_byte = EEPROM_START + off + i;
+		ret = olpc_ec_cmd(EC_BAT_EEPROM, &ec_byte, 1, &buf[i], 1);
 		if (ret) {
-			printk(KERN_ERR "olpc-battery:  EC command "
-					"EC_BAT_EEPROM @ 0x%x failed -"
-					" %d!\n", ec_byte, ret);
+			pr_err("olpc-battery: "
+			       "EC_BAT_EEPROM cmd @ 0x%x failed - %d!\n",
+			       ec_byte, ret);
 			return -EIO;
 		}
 	}
diff --git a/drivers/rtc/rtc-cmos.c b/drivers/rtc/rtc-cmos.c
index 23e10b6263d..f7a4701bf86 100644
--- a/drivers/rtc/rtc-cmos.c
+++ b/drivers/rtc/rtc-cmos.c
@@ -1174,23 +1174,34 @@ static struct platform_driver cmos_platform_driver = {
 	}
 };
 
+#ifdef CONFIG_PNP
+static bool pnp_driver_registered;
+#endif
+static bool platform_driver_registered;
+
 static int __init cmos_init(void)
 {
 	int retval = 0;
 
 #ifdef	CONFIG_PNP
-	pnp_register_driver(&cmos_pnp_driver);
+	retval = pnp_register_driver(&cmos_pnp_driver);
+	if (retval == 0)
+		pnp_driver_registered = true;
 #endif
 
-	if (!cmos_rtc.dev)
+	if (!cmos_rtc.dev) {
 		retval = platform_driver_probe(&cmos_platform_driver,
 					       cmos_platform_probe);
+		if (retval == 0)
+			platform_driver_registered = true;
+	}
 
 	if (retval == 0)
 		return 0;
 
 #ifdef	CONFIG_PNP
-	pnp_unregister_driver(&cmos_pnp_driver);
+	if (pnp_driver_registered)
+		pnp_unregister_driver(&cmos_pnp_driver);
 #endif
 	return retval;
 }
@@ -1199,9 +1210,11 @@ module_init(cmos_init);
 static void __exit cmos_exit(void)
 {
 #ifdef	CONFIG_PNP
-	pnp_unregister_driver(&cmos_pnp_driver);
+	if (pnp_driver_registered)
+		pnp_unregister_driver(&cmos_pnp_driver);
 #endif
-	platform_driver_unregister(&cmos_platform_driver);
+	if (platform_driver_registered)
+		platform_driver_unregister(&cmos_platform_driver);
 }
 module_exit(cmos_exit);
 
diff --git a/drivers/s390/scsi/zfcp_erp.c b/drivers/s390/scsi/zfcp_erp.c
index 8030e25152f..c75d6f35cb5 100644
--- a/drivers/s390/scsi/zfcp_erp.c
+++ b/drivers/s390/scsi/zfcp_erp.c
@@ -553,40 +553,35 @@ static void _zfcp_erp_unit_reopen_all(struct zfcp_port *port, int clear,
 		_zfcp_erp_unit_reopen(unit, clear, id, ref);
 }
 
-static void zfcp_erp_strategy_followup_actions(struct zfcp_erp_action *act)
+static void zfcp_erp_strategy_followup_failed(struct zfcp_erp_action *act)
 {
-	struct zfcp_adapter *adapter = act->adapter;
-	struct zfcp_port *port = act->port;
-	struct zfcp_unit *unit = act->unit;
-	u32 status = act->status;
-
-	/* initiate follow-up actions depending on success of finished action */
 	switch (act->action) {
-
 	case ZFCP_ERP_ACTION_REOPEN_ADAPTER:
-		if (status == ZFCP_ERP_SUCCEEDED)
-			_zfcp_erp_port_reopen_all(adapter, 0, "ersfa_1", NULL);
-		else
-			_zfcp_erp_adapter_reopen(adapter, 0, "ersfa_2", NULL);
+		_zfcp_erp_adapter_reopen(act->adapter, 0, "ersff_1", NULL);
 		break;
-
 	case ZFCP_ERP_ACTION_REOPEN_PORT_FORCED:
-		if (status == ZFCP_ERP_SUCCEEDED)
-			_zfcp_erp_port_reopen(port, 0, "ersfa_3", NULL);
-		else
-			_zfcp_erp_adapter_reopen(adapter, 0, "ersfa_4", NULL);
+		_zfcp_erp_port_forced_reopen(act->port, 0, "ersff_2", NULL);
 		break;
-
 	case ZFCP_ERP_ACTION_REOPEN_PORT:
-		if (status == ZFCP_ERP_SUCCEEDED)
-			_zfcp_erp_unit_reopen_all(port, 0, "ersfa_5", NULL);
-		else
-			_zfcp_erp_port_forced_reopen(port, 0, "ersfa_6", NULL);
+		_zfcp_erp_port_reopen(act->port, 0, "ersff_3", NULL);
 		break;
-
 	case ZFCP_ERP_ACTION_REOPEN_UNIT:
-		if (status != ZFCP_ERP_SUCCEEDED)
-			_zfcp_erp_port_reopen(unit->port, 0, "ersfa_7", NULL);
+		_zfcp_erp_unit_reopen(act->unit, 0, "ersff_4", NULL);
+		break;
+	}
+}
+
+static void zfcp_erp_strategy_followup_success(struct zfcp_erp_action *act)
+{
+	switch (act->action) {
+	case ZFCP_ERP_ACTION_REOPEN_ADAPTER:
+		_zfcp_erp_port_reopen_all(act->adapter, 0, "ersfs_1", NULL);
+		break;
+	case ZFCP_ERP_ACTION_REOPEN_PORT_FORCED:
+		_zfcp_erp_port_reopen(act->port, 0, "ersfs_2", NULL);
+		break;
+	case ZFCP_ERP_ACTION_REOPEN_PORT:
+		_zfcp_erp_unit_reopen_all(act->port, 0, "ersfs_3", NULL);
 		break;
 	}
 }
@@ -801,7 +796,7 @@ static int zfcp_erp_port_forced_strategy(struct zfcp_erp_action *erp_action)
 			return ZFCP_ERP_FAILED;
 
 	case ZFCP_ERP_STEP_PHYS_PORT_CLOSING:
-		if (status & ZFCP_STATUS_PORT_PHYS_OPEN)
+		if (!(status & ZFCP_STATUS_PORT_PHYS_OPEN))
 			return ZFCP_ERP_SUCCEEDED;
 	}
 	return ZFCP_ERP_FAILED;
@@ -853,11 +848,17 @@ void zfcp_erp_port_strategy_open_lookup(struct work_struct *work)
 					      gid_pn_work);
 
 	retval = zfcp_fc_ns_gid_pn(&port->erp_action);
-	if (retval == -ENOMEM)
-		zfcp_erp_notify(&port->erp_action, ZFCP_ERP_NOMEM);
-	port->erp_action.step = ZFCP_ERP_STEP_NAMESERVER_LOOKUP;
-	if (retval)
-		zfcp_erp_notify(&port->erp_action, ZFCP_ERP_FAILED);
+	if (!retval) {
+		port->erp_action.step = ZFCP_ERP_STEP_NAMESERVER_LOOKUP;
+		goto out;
+	}
+	if (retval == -ENOMEM) {
+		zfcp_erp_notify(&port->erp_action, ZFCP_STATUS_ERP_LOWMEM);
+		goto out;
+	}
+	/* all other error condtions */
+	zfcp_erp_notify(&port->erp_action, 0);
+out:
 	zfcp_port_put(port);
 }
 
@@ -1289,7 +1290,10 @@ static int zfcp_erp_strategy(struct zfcp_erp_action *erp_action)
 	retval = zfcp_erp_strategy_statechange(erp_action, retval);
 	if (retval == ZFCP_ERP_EXIT)
 		goto unlock;
-	zfcp_erp_strategy_followup_actions(erp_action);
+	if (retval == ZFCP_ERP_SUCCEEDED)
+		zfcp_erp_strategy_followup_success(erp_action);
+	if (retval == ZFCP_ERP_FAILED)
+		zfcp_erp_strategy_followup_failed(erp_action);
 
  unlock:
 	write_unlock(&adapter->erp_lock);
diff --git a/drivers/s390/scsi/zfcp_fc.c b/drivers/s390/scsi/zfcp_fc.c
index 2f0705d76b7..47daebfa7e5 100644
--- a/drivers/s390/scsi/zfcp_fc.c
+++ b/drivers/s390/scsi/zfcp_fc.c
@@ -79,11 +79,9 @@ static int zfcp_wka_port_get(struct zfcp_wka_port *wka_port)
 
 	mutex_unlock(&wka_port->mutex);
 
-	wait_event_timeout(
-		wka_port->completion_wq,
-		wka_port->status == ZFCP_WKA_PORT_ONLINE ||
-		wka_port->status == ZFCP_WKA_PORT_OFFLINE,
-		HZ >> 1);
+	wait_event(wka_port->completion_wq,
+		   wka_port->status == ZFCP_WKA_PORT_ONLINE ||
+		   wka_port->status == ZFCP_WKA_PORT_OFFLINE);
 
 	if (wka_port->status == ZFCP_WKA_PORT_ONLINE) {
 		atomic_inc(&wka_port->refcount);
diff --git a/drivers/s390/scsi/zfcp_fsf.c b/drivers/s390/scsi/zfcp_fsf.c
index c57658f3d34..47795fbf081 100644
--- a/drivers/s390/scsi/zfcp_fsf.c
+++ b/drivers/s390/scsi/zfcp_fsf.c
@@ -670,8 +670,11 @@ static int zfcp_fsf_req_sbal_get(struct zfcp_adapter *adapter)
 			       zfcp_fsf_sbal_check(adapter), 5 * HZ);
 	if (ret > 0)
 		return 0;
-	if (!ret)
+	if (!ret) {
 		atomic_inc(&adapter->qdio_outb_full);
+		/* assume hanging outbound queue, try queue recovery */
+		zfcp_erp_adapter_reopen(adapter, 0, "fsrsg_1", NULL);
+	}
 
 	spin_lock_bh(&adapter->req_q_lock);
 	return -EIO;
@@ -722,7 +725,7 @@ static struct zfcp_fsf_req *zfcp_fsf_req_create(struct zfcp_adapter *adapter,
 		req = zfcp_fsf_alloc_qtcb(pool);
 
 	if (unlikely(!req))
-		return ERR_PTR(-EIO);
+		return ERR_PTR(-ENOMEM);
 
 	if (adapter->req_no == 0)
 		adapter->req_no++;
@@ -1010,6 +1013,23 @@ skip_fsfstatus:
 		send_ct->handler(send_ct->handler_data);
 }
 
+static void zfcp_fsf_setup_ct_els_unchained(struct qdio_buffer_element *sbale,
+					    struct scatterlist *sg_req,
+					    struct scatterlist *sg_resp)
+{
+	sbale[0].flags |= SBAL_FLAGS0_TYPE_WRITE_READ;
+	sbale[2].addr   = sg_virt(sg_req);
+	sbale[2].length = sg_req->length;
+	sbale[3].addr   = sg_virt(sg_resp);
+	sbale[3].length = sg_resp->length;
+	sbale[3].flags |= SBAL_FLAGS_LAST_ENTRY;
+}
+
+static int zfcp_fsf_one_sbal(struct scatterlist *sg)
+{
+	return sg_is_last(sg) && sg->length <= PAGE_SIZE;
+}
+
 static int zfcp_fsf_setup_ct_els_sbals(struct zfcp_fsf_req *req,
 				       struct scatterlist *sg_req,
 				       struct scatterlist *sg_resp,
@@ -1020,30 +1040,30 @@ static int zfcp_fsf_setup_ct_els_sbals(struct zfcp_fsf_req *req,
 	int bytes;
 
 	if (!(feat & FSF_FEATURE_ELS_CT_CHAINED_SBALS)) {
-		if (sg_req->length > PAGE_SIZE || sg_resp->length > PAGE_SIZE ||
-		    !sg_is_last(sg_req) || !sg_is_last(sg_resp))
+		if (!zfcp_fsf_one_sbal(sg_req) || !zfcp_fsf_one_sbal(sg_resp))
 			return -EOPNOTSUPP;
 
-		sbale[0].flags |= SBAL_FLAGS0_TYPE_WRITE_READ;
-		sbale[2].addr   = sg_virt(sg_req);
-		sbale[2].length = sg_req->length;
-		sbale[3].addr   = sg_virt(sg_resp);
-		sbale[3].length = sg_resp->length;
-		sbale[3].flags |= SBAL_FLAGS_LAST_ENTRY;
+		zfcp_fsf_setup_ct_els_unchained(sbale, sg_req, sg_resp);
+		return 0;
+	}
+
+	/* use single, unchained SBAL if it can hold the request */
+	if (zfcp_fsf_one_sbal(sg_req) && zfcp_fsf_one_sbal(sg_resp)) {
+		zfcp_fsf_setup_ct_els_unchained(sbale, sg_req, sg_resp);
 		return 0;
 	}
 
 	bytes = zfcp_qdio_sbals_from_sg(req, SBAL_FLAGS0_TYPE_WRITE_READ,
 					sg_req, max_sbals);
 	if (bytes <= 0)
-		return -ENOMEM;
+		return -EIO;
 	req->qtcb->bottom.support.req_buf_length = bytes;
 	req->sbale_curr = ZFCP_LAST_SBALE_PER_SBAL;
 
 	bytes = zfcp_qdio_sbals_from_sg(req, SBAL_FLAGS0_TYPE_WRITE_READ,
 					sg_resp, max_sbals);
 	if (bytes <= 0)
-		return -ENOMEM;
+		return -EIO;
 	req->qtcb->bottom.support.resp_buf_length = bytes;
 
 	return 0;
@@ -1607,10 +1627,10 @@ static void zfcp_fsf_open_wka_port_handler(struct zfcp_fsf_req *req)
 	case FSF_ACCESS_DENIED:
 		wka_port->status = ZFCP_WKA_PORT_OFFLINE;
 		break;
-	case FSF_PORT_ALREADY_OPEN:
-		break;
 	case FSF_GOOD:
 		wka_port->handle = header->port_handle;
+		/* fall through */
+	case FSF_PORT_ALREADY_OPEN:
 		wka_port->status = ZFCP_WKA_PORT_ONLINE;
 	}
 out:
@@ -1731,15 +1751,16 @@ static void zfcp_fsf_close_physical_port_handler(struct zfcp_fsf_req *req)
 		zfcp_fsf_access_denied_port(req, port);
 		break;
 	case FSF_PORT_BOXED:
-		zfcp_erp_port_boxed(port, "fscpph2", req);
-		req->status |= ZFCP_STATUS_FSFREQ_ERROR |
-			       ZFCP_STATUS_FSFREQ_RETRY;
 		/* can't use generic zfcp_erp_modify_port_status because
 		 * ZFCP_STATUS_COMMON_OPEN must not be reset for the port */
 		atomic_clear_mask(ZFCP_STATUS_PORT_PHYS_OPEN, &port->status);
 		list_for_each_entry(unit, &port->unit_list_head, list)
 			atomic_clear_mask(ZFCP_STATUS_COMMON_OPEN,
 					  &unit->status);
+		zfcp_erp_port_boxed(port, "fscpph2", req);
+		req->status |= ZFCP_STATUS_FSFREQ_ERROR |
+			       ZFCP_STATUS_FSFREQ_RETRY;
+
 		break;
 	case FSF_ADAPTER_STATUS_AVAILABLE:
 		switch (header->fsf_status_qual.word[0]) {
@@ -2541,7 +2562,6 @@ struct zfcp_fsf_req *zfcp_fsf_control_file(struct zfcp_adapter *adapter,
 	bytes = zfcp_qdio_sbals_from_sg(req, direction, fsf_cfdc->sg,
 					FSF_MAX_SBALS_PER_REQ);
 	if (bytes != ZFCP_CFDC_MAX_SIZE) {
-		retval = -ENOMEM;
 		zfcp_fsf_req_free(req);
 		goto out;
 	}
diff --git a/drivers/s390/scsi/zfcp_scsi.c b/drivers/s390/scsi/zfcp_scsi.c
index 967ede73f4c..6925a178468 100644
--- a/drivers/s390/scsi/zfcp_scsi.c
+++ b/drivers/s390/scsi/zfcp_scsi.c
@@ -167,20 +167,21 @@ static int zfcp_scsi_eh_abort_handler(struct scsi_cmnd *scpnt)
 	struct zfcp_unit *unit = scpnt->device->hostdata;
 	struct zfcp_fsf_req *old_req, *abrt_req;
 	unsigned long flags;
-	unsigned long old_req_id = (unsigned long) scpnt->host_scribble;
+	unsigned long old_reqid = (unsigned long) scpnt->host_scribble;
 	int retval = SUCCESS;
 	int retry = 3;
+	char *dbf_tag;
 
 	/* avoid race condition between late normal completion and abort */
 	write_lock_irqsave(&adapter->abort_lock, flags);
 
 	spin_lock(&adapter->req_list_lock);
-	old_req = zfcp_reqlist_find(adapter, old_req_id);
+	old_req = zfcp_reqlist_find(adapter, old_reqid);
 	spin_unlock(&adapter->req_list_lock);
 	if (!old_req) {
 		write_unlock_irqrestore(&adapter->abort_lock, flags);
 		zfcp_scsi_dbf_event_abort("lte1", adapter, scpnt, NULL,
-					  old_req_id);
+					  old_reqid);
 		return FAILED; /* completion could be in progress */
 	}
 	old_req->data = NULL;
@@ -189,7 +190,7 @@ static int zfcp_scsi_eh_abort_handler(struct scsi_cmnd *scpnt)
 	write_unlock_irqrestore(&adapter->abort_lock, flags);
 
 	while (retry--) {
-		abrt_req = zfcp_fsf_abort_fcp_command(old_req_id, unit);
+		abrt_req = zfcp_fsf_abort_fcp_command(old_reqid, unit);
 		if (abrt_req)
 			break;
 
@@ -197,7 +198,7 @@ static int zfcp_scsi_eh_abort_handler(struct scsi_cmnd *scpnt)
 		if (!(atomic_read(&adapter->status) &
 		      ZFCP_STATUS_COMMON_RUNNING)) {
 			zfcp_scsi_dbf_event_abort("nres", adapter, scpnt, NULL,
-						  old_req_id);
+						  old_reqid);
 			return SUCCESS;
 		}
 	}
@@ -208,13 +209,14 @@ static int zfcp_scsi_eh_abort_handler(struct scsi_cmnd *scpnt)
 		   abrt_req->status & ZFCP_STATUS_FSFREQ_COMPLETED);
 
 	if (abrt_req->status & ZFCP_STATUS_FSFREQ_ABORTSUCCEEDED)
-		zfcp_scsi_dbf_event_abort("okay", adapter, scpnt, abrt_req, 0);
+		dbf_tag = "okay";
 	else if (abrt_req->status & ZFCP_STATUS_FSFREQ_ABORTNOTNEEDED)
-		zfcp_scsi_dbf_event_abort("lte2", adapter, scpnt, abrt_req, 0);
+		dbf_tag = "lte2";
 	else {
-		zfcp_scsi_dbf_event_abort("fail", adapter, scpnt, abrt_req, 0);
+		dbf_tag = "fail";
 		retval = FAILED;
 	}
+	zfcp_scsi_dbf_event_abort(dbf_tag, adapter, scpnt, abrt_req, old_reqid);
 	zfcp_fsf_req_free(abrt_req);
 	return retval;
 }
@@ -534,6 +536,9 @@ static void zfcp_scsi_rport_register(struct zfcp_port *port)
 	struct fc_rport_identifiers ids;
 	struct fc_rport *rport;
 
+	if (port->rport)
+		return;
+
 	ids.node_name = port->wwnn;
 	ids.port_name = port->wwpn;
 	ids.port_id = port->d_id;
@@ -557,8 +562,10 @@ static void zfcp_scsi_rport_block(struct zfcp_port *port)
 {
 	struct fc_rport *rport = port->rport;
 
-	if (rport)
+	if (rport) {
 		fc_remote_port_delete(rport);
+		port->rport = NULL;
+	}
 }
 
 void zfcp_scsi_schedule_rport_register(struct zfcp_port *port)
diff --git a/drivers/s390/scsi/zfcp_sysfs.c b/drivers/s390/scsi/zfcp_sysfs.c
index 3e51e64d110..0fe5cce818c 100644
--- a/drivers/s390/scsi/zfcp_sysfs.c
+++ b/drivers/s390/scsi/zfcp_sysfs.c
@@ -494,9 +494,14 @@ static ssize_t zfcp_sysfs_adapter_q_full_show(struct device *dev,
 	struct Scsi_Host *scsi_host = class_to_shost(dev);
 	struct zfcp_adapter *adapter =
 		(struct zfcp_adapter *) scsi_host->hostdata[0];
+	u64 util;
+
+	spin_lock_bh(&adapter->qdio_stat_lock);
+	util = adapter->req_q_util;
+	spin_unlock_bh(&adapter->qdio_stat_lock);
 
 	return sprintf(buf, "%d %llu\n", atomic_read(&adapter->qdio_outb_full),
-		       (unsigned long long)adapter->req_q_util);
+		       (unsigned long long)util);
 }
 static DEVICE_ATTR(queue_full, S_IRUGO, zfcp_sysfs_adapter_q_full_show, NULL);
 
diff --git a/drivers/scsi/libfc/fc_exch.c b/drivers/scsi/libfc/fc_exch.c
index 2bc22be5f84..145ab9ba55e 100644
--- a/drivers/scsi/libfc/fc_exch.c
+++ b/drivers/scsi/libfc/fc_exch.c
@@ -415,9 +415,9 @@ static void fc_exch_timeout(struct work_struct *work)
 	e_stat = ep->esb_stat;
 	if (e_stat & ESB_ST_COMPLETE) {
 		ep->esb_stat = e_stat & ~ESB_ST_REC_QUAL;
+		spin_unlock_bh(&ep->ex_lock);
 		if (e_stat & ESB_ST_REC_QUAL)
 			fc_exch_rrq(ep);
-		spin_unlock_bh(&ep->ex_lock);
 		goto done;
 	} else {
 		resp = ep->resp;
@@ -1624,14 +1624,14 @@ static void fc_exch_rrq(struct fc_exch *ep)
 	struct fc_lport *lp;
 	struct fc_els_rrq *rrq;
 	struct fc_frame *fp;
-	struct fc_seq *rrq_sp;
 	u32 did;
 
 	lp = ep->lp;
 
 	fp = fc_frame_alloc(lp, sizeof(*rrq));
 	if (!fp)
-		return;
+		goto retry;
+
 	rrq = fc_frame_payload_get(fp, sizeof(*rrq));
 	memset(rrq, 0, sizeof(*rrq));
 	rrq->rrq_cmd = ELS_RRQ;
@@ -1647,13 +1647,20 @@ static void fc_exch_rrq(struct fc_exch *ep)
 		       fc_host_port_id(lp->host), FC_TYPE_ELS,
 		       FC_FC_FIRST_SEQ | FC_FC_END_SEQ | FC_FC_SEQ_INIT, 0);
 
-	rrq_sp = fc_exch_seq_send(lp, fp, fc_exch_rrq_resp, NULL, ep,
-				  lp->e_d_tov);
-	if (!rrq_sp) {
-		ep->esb_stat |= ESB_ST_REC_QUAL;
-		fc_exch_timer_set_locked(ep, ep->r_a_tov);
+	if (fc_exch_seq_send(lp, fp, fc_exch_rrq_resp, NULL, ep, lp->e_d_tov))
+		return;
+
+retry:
+	spin_lock_bh(&ep->ex_lock);
+	if (ep->state & (FC_EX_RST_CLEANUP | FC_EX_DONE)) {
+		spin_unlock_bh(&ep->ex_lock);
+		/* drop hold for rec qual */
+		fc_exch_release(ep);
 		return;
 	}
+	ep->esb_stat |= ESB_ST_REC_QUAL;
+	fc_exch_timer_set_locked(ep, ep->r_a_tov);
+	spin_unlock_bh(&ep->ex_lock);
 }
 
 
diff --git a/drivers/scsi/libiscsi.c b/drivers/scsi/libiscsi.c
index 716cc344c5d..a751f6230c2 100644
--- a/drivers/scsi/libiscsi.c
+++ b/drivers/scsi/libiscsi.c
@@ -1974,10 +1974,10 @@ int iscsi_eh_abort(struct scsi_cmnd *sc)
 		 * good and have never sent us a successful tmf response
 		 * then sent more data for the cmd.
 		 */
-		spin_lock(&session->lock);
+		spin_lock_bh(&session->lock);
 		fail_scsi_task(task, DID_ABORT);
 		conn->tmf_state = TMF_INITIAL;
-		spin_unlock(&session->lock);
+		spin_unlock_bh(&session->lock);
 		iscsi_start_tx(conn);
 		goto success_unlocked;
 	case TMF_TIMEDOUT:
diff --git a/drivers/scsi/libsas/sas_expander.c b/drivers/scsi/libsas/sas_expander.c
index 54fa1e42dc4..b3381959acc 100644
--- a/drivers/scsi/libsas/sas_expander.c
+++ b/drivers/scsi/libsas/sas_expander.c
@@ -766,6 +766,7 @@ static int sas_ex_join_wide_port(struct domain_device *parent, int phy_id)
 		if (!memcmp(phy->attached_sas_addr, ephy->attached_sas_addr,
 			    SAS_ADDR_SIZE) && ephy->port) {
 			sas_port_add_phy(ephy->port, phy->phy);
+			phy->port = ephy->port;
 			phy->phy_state = PHY_DEVICE_DISCOVERED;
 			return 0;
 		}
@@ -945,11 +946,21 @@ static int sas_ex_discover_dev(struct domain_device *dev, int phy_id)
 			if (ex->ex_phy[i].phy_state == PHY_VACANT ||
 			    ex->ex_phy[i].phy_state == PHY_NOT_PRESENT)
 				continue;
-
+			/*
+			 * Due to races, the phy might not get added to the
+			 * wide port, so we add the phy to the wide port here.
+			 */
 			if (SAS_ADDR(ex->ex_phy[i].attached_sas_addr) ==
-			    SAS_ADDR(child->sas_addr))
+			    SAS_ADDR(child->sas_addr)) {
 				ex->ex_phy[i].phy_state= PHY_DEVICE_DISCOVERED;
+				res = sas_ex_join_wide_port(dev, i);
+				if (!res)
+					SAS_DPRINTK("Attaching ex phy%d to wide port %016llx\n",
+						    i, SAS_ADDR(ex->ex_phy[i].attached_sas_addr));
+
+			}
 		}
+		res = 0;
 	}
 
 	return res;
@@ -1598,7 +1609,7 @@ static int sas_get_phy_attached_sas_addr(struct domain_device *dev,
 }
 
 static int sas_find_bcast_phy(struct domain_device *dev, int *phy_id,
-			      int from_phy)
+			      int from_phy, bool update)
 {
 	struct expander_device *ex = &dev->ex_dev;
 	int res = 0;
@@ -1611,7 +1622,9 @@ static int sas_find_bcast_phy(struct domain_device *dev, int *phy_id,
 		if (res)
 			goto out;
 		else if (phy_change_count != ex->ex_phy[i].phy_change_count) {
-			ex->ex_phy[i].phy_change_count = phy_change_count;
+			if (update)
+				ex->ex_phy[i].phy_change_count =
+					phy_change_count;
 			*phy_id = i;
 			return 0;
 		}
@@ -1653,31 +1666,52 @@ out:
 	kfree(rg_req);
 	return res;
 }
+/**
+ * sas_find_bcast_dev -  find the device issue BROADCAST(CHANGE).
+ * @dev:domain device to be detect.
+ * @src_dev: the device which originated BROADCAST(CHANGE).
+ *
+ * Add self-configuration expander suport. Suppose two expander cascading,
+ * when the first level expander is self-configuring, hotplug the disks in
+ * second level expander, BROADCAST(CHANGE) will not only be originated
+ * in the second level expander, but also be originated in the first level
+ * expander (see SAS protocol SAS 2r-14, 7.11 for detail), it is to say,
+ * expander changed count in two level expanders will all increment at least
+ * once, but the phy which chang count has changed is the source device which
+ * we concerned.
+ */
 
 static int sas_find_bcast_dev(struct domain_device *dev,
 			      struct domain_device **src_dev)
 {
 	struct expander_device *ex = &dev->ex_dev;
 	int ex_change_count = -1;
+	int phy_id = -1;
 	int res;
+	struct domain_device *ch;
 
 	res = sas_get_ex_change_count(dev, &ex_change_count);
 	if (res)
 		goto out;
-	if (ex_change_count != -1 &&
-	    ex_change_count != ex->ex_change_count) {
-		*src_dev = dev;
-		ex->ex_change_count = ex_change_count;
-	} else {
-		struct domain_device *ch;
-
-		list_for_each_entry(ch, &ex->children, siblings) {
-			if (ch->dev_type == EDGE_DEV ||
-			    ch->dev_type == FANOUT_DEV) {
-				res = sas_find_bcast_dev(ch, src_dev);
-				if (src_dev)
-					return res;
-			}
+	if (ex_change_count != -1 && ex_change_count != ex->ex_change_count) {
+		/* Just detect if this expander phys phy change count changed,
+		* in order to determine if this expander originate BROADCAST,
+		* and do not update phy change count field in our structure.
+		*/
+		res = sas_find_bcast_phy(dev, &phy_id, 0, false);
+		if (phy_id != -1) {
+			*src_dev = dev;
+			ex->ex_change_count = ex_change_count;
+			SAS_DPRINTK("Expander phy change count has changed\n");
+			return res;
+		} else
+			SAS_DPRINTK("Expander phys DID NOT change\n");
+	}
+	list_for_each_entry(ch, &ex->children, siblings) {
+		if (ch->dev_type == EDGE_DEV || ch->dev_type == FANOUT_DEV) {
+			res = sas_find_bcast_dev(ch, src_dev);
+			if (src_dev)
+				return res;
 		}
 	}
 out:
@@ -1700,24 +1734,26 @@ static void sas_unregister_ex_tree(struct domain_device *dev)
 }
 
 static void sas_unregister_devs_sas_addr(struct domain_device *parent,
-					 int phy_id)
+					 int phy_id, bool last)
 {
 	struct expander_device *ex_dev = &parent->ex_dev;
 	struct ex_phy *phy = &ex_dev->ex_phy[phy_id];
 	struct domain_device *child, *n;
-
-	list_for_each_entry_safe(child, n, &ex_dev->children, siblings) {
-		if (SAS_ADDR(child->sas_addr) ==
-		    SAS_ADDR(phy->attached_sas_addr)) {
-			if (child->dev_type == EDGE_DEV ||
-			    child->dev_type == FANOUT_DEV)
-				sas_unregister_ex_tree(child);
-			else
-				sas_unregister_dev(child);
-			break;
+	if (last) {
+		list_for_each_entry_safe(child, n,
+			&ex_dev->children, siblings) {
+			if (SAS_ADDR(child->sas_addr) ==
+			    SAS_ADDR(phy->attached_sas_addr)) {
+				if (child->dev_type == EDGE_DEV ||
+				    child->dev_type == FANOUT_DEV)
+					sas_unregister_ex_tree(child);
+				else
+					sas_unregister_dev(child);
+				break;
+			}
 		}
+		sas_disable_routing(parent, phy->attached_sas_addr);
 	}
-	sas_disable_routing(parent, phy->attached_sas_addr);
 	memset(phy->attached_sas_addr, 0, SAS_ADDR_SIZE);
 	sas_port_delete_phy(phy->port, phy->phy);
 	if (phy->port->num_phys == 0)
@@ -1770,15 +1806,31 @@ static int sas_discover_new(struct domain_device *dev, int phy_id)
 {
 	struct ex_phy *ex_phy = &dev->ex_dev.ex_phy[phy_id];
 	struct domain_device *child;
-	int res;
+	bool found = false;
+	int res, i;
 
 	SAS_DPRINTK("ex %016llx phy%d new device attached\n",
 		    SAS_ADDR(dev->sas_addr), phy_id);
 	res = sas_ex_phy_discover(dev, phy_id);
 	if (res)
 		goto out;
+	/* to support the wide port inserted */
+	for (i = 0; i < dev->ex_dev.num_phys; i++) {
+		struct ex_phy *ex_phy_temp = &dev->ex_dev.ex_phy[i];
+		if (i == phy_id)
+			continue;
+		if (SAS_ADDR(ex_phy_temp->attached_sas_addr) ==
+		    SAS_ADDR(ex_phy->attached_sas_addr)) {
+			found = true;
+			break;
+		}
+	}
+	if (found) {
+		sas_ex_join_wide_port(dev, phy_id);
+		return 0;
+	}
 	res = sas_ex_discover_devices(dev, phy_id);
-	if (res)
+	if (!res)
 		goto out;
 	list_for_each_entry(child, &dev->ex_dev.children, siblings) {
 		if (SAS_ADDR(child->sas_addr) ==
@@ -1793,7 +1845,7 @@ out:
 	return res;
 }
 
-static int sas_rediscover_dev(struct domain_device *dev, int phy_id)
+static int sas_rediscover_dev(struct domain_device *dev, int phy_id, bool last)
 {
 	struct expander_device *ex = &dev->ex_dev;
 	struct ex_phy *phy = &ex->ex_phy[phy_id];
@@ -1804,11 +1856,11 @@ static int sas_rediscover_dev(struct domain_device *dev, int phy_id)
 	switch (res) {
 	case SMP_RESP_NO_PHY:
 		phy->phy_state = PHY_NOT_PRESENT;
-		sas_unregister_devs_sas_addr(dev, phy_id);
+		sas_unregister_devs_sas_addr(dev, phy_id, last);
 		goto out; break;
 	case SMP_RESP_PHY_VACANT:
 		phy->phy_state = PHY_VACANT;
-		sas_unregister_devs_sas_addr(dev, phy_id);
+		sas_unregister_devs_sas_addr(dev, phy_id, last);
 		goto out; break;
 	case SMP_RESP_FUNC_ACC:
 		break;
@@ -1816,7 +1868,7 @@ static int sas_rediscover_dev(struct domain_device *dev, int phy_id)
 
 	if (SAS_ADDR(attached_sas_addr) == 0) {
 		phy->phy_state = PHY_EMPTY;
-		sas_unregister_devs_sas_addr(dev, phy_id);
+		sas_unregister_devs_sas_addr(dev, phy_id, last);
 	} else if (SAS_ADDR(attached_sas_addr) ==
 		   SAS_ADDR(phy->attached_sas_addr)) {
 		SAS_DPRINTK("ex %016llx phy 0x%x broadcast flutter\n",
@@ -1828,12 +1880,27 @@ out:
 	return res;
 }
 
+/**
+ * sas_rediscover - revalidate the domain.
+ * @dev:domain device to be detect.
+ * @phy_id: the phy id will be detected.
+ *
+ * NOTE: this process _must_ quit (return) as soon as any connection
+ * errors are encountered.  Connection recovery is done elsewhere.
+ * Discover process only interrogates devices in order to discover the
+ * domain.For plugging out, we un-register the device only when it is
+ * the last phy in the port, for other phys in this port, we just delete it
+ * from the port.For inserting, we do discovery when it is the
+ * first phy,for other phys in this port, we add it to the port to
+ * forming the wide-port.
+ */
 static int sas_rediscover(struct domain_device *dev, const int phy_id)
 {
 	struct expander_device *ex = &dev->ex_dev;
 	struct ex_phy *changed_phy = &ex->ex_phy[phy_id];
 	int res = 0;
 	int i;
+	bool last = true;	/* is this the last phy of the port */
 
 	SAS_DPRINTK("ex %016llx phy%d originated BROADCAST(CHANGE)\n",
 		    SAS_ADDR(dev->sas_addr), phy_id);
@@ -1848,13 +1915,13 @@ static int sas_rediscover(struct domain_device *dev, const int phy_id)
 			    SAS_ADDR(changed_phy->attached_sas_addr)) {
 				SAS_DPRINTK("phy%d part of wide port with "
 					    "phy%d\n", phy_id, i);
-				goto out;
+				last = false;
+				break;
 			}
 		}
-		res = sas_rediscover_dev(dev, phy_id);
+		res = sas_rediscover_dev(dev, phy_id, last);
 	} else
 		res = sas_discover_new(dev, phy_id);
-out:
 	return res;
 }
 
@@ -1881,7 +1948,7 @@ int sas_ex_revalidate_domain(struct domain_device *port_dev)
 
 		do {
 			phy_id = -1;
-			res = sas_find_bcast_phy(dev, &phy_id, i);
+			res = sas_find_bcast_phy(dev, &phy_id, i, true);
 			if (phy_id == -1)
 				break;
 			res = sas_rediscover(dev, phy_id);
diff --git a/drivers/scsi/libsas/sas_port.c b/drivers/scsi/libsas/sas_port.c
index e6ac59c023f..fe8b74c706d 100644
--- a/drivers/scsi/libsas/sas_port.c
+++ b/drivers/scsi/libsas/sas_port.c
@@ -56,7 +56,7 @@ static void sas_form_port(struct asd_sas_phy *phy)
 		}
 	}
 
-	/* find a port */
+	/* see if the phy should be part of a wide port */
 	spin_lock_irqsave(&sas_ha->phy_port_lock, flags);
 	for (i = 0; i < sas_ha->num_phys; i++) {
 		port = sas_ha->sas_port[i];
@@ -69,12 +69,23 @@ static void sas_form_port(struct asd_sas_phy *phy)
 			SAS_DPRINTK("phy%d matched wide port%d\n", phy->id,
 				    port->id);
 			break;
-		} else if (*(u64 *) port->sas_addr == 0 && port->num_phys==0) {
-			memcpy(port->sas_addr, phy->sas_addr, SAS_ADDR_SIZE);
-			break;
 		}
 		spin_unlock(&port->phy_list_lock);
 	}
+	/* The phy does not match any existing port, create a new one */
+	if (i == sas_ha->num_phys) {
+		for (i = 0; i < sas_ha->num_phys; i++) {
+			port = sas_ha->sas_port[i];
+			spin_lock(&port->phy_list_lock);
+			if (*(u64 *)port->sas_addr == 0
+				&& port->num_phys == 0) {
+				memcpy(port->sas_addr, phy->sas_addr,
+					SAS_ADDR_SIZE);
+				break;
+			}
+			spin_unlock(&port->phy_list_lock);
+		}
+	}
 
 	if (i >= sas_ha->num_phys) {
 		printk(KERN_NOTICE "%s: couldn't find a free port, bug?\n",
diff --git a/drivers/scsi/qla4xxx/ql4_dbg.c b/drivers/scsi/qla4xxx/ql4_dbg.c
index fcc184cd066..cbceb0ebabf 100644
--- a/drivers/scsi/qla4xxx/ql4_dbg.c
+++ b/drivers/scsi/qla4xxx/ql4_dbg.c
@@ -15,19 +15,18 @@ void qla4xxx_dump_buffer(void *b, uint32_t size)
 	uint32_t cnt;
 	uint8_t *c = b;
 
-	printk(" 0   1	 2   3	 4   5	 6   7	 8   9	Ah  Bh	Ch  Dh	Eh  "
+	printk(" 0   1   2   3   4   5   6   7   8   9  Ah  Bh  Ch  Dh  Eh  "
 	       "Fh\n");
 	printk("------------------------------------------------------------"
 	       "--\n");
-	for (cnt = 0; cnt < size; cnt++, c++) {
-		printk(KERN_DEBUG "%02x", *c);
-		if (!(cnt % 16))
-			printk(KERN_DEBUG "\n");
+	for (cnt = 0; cnt < size; c++) {
+		printk(KERN_INFO "%02x", *c);
+		if (!(++cnt % 16))
+			printk(KERN_INFO "\n");
 
 		else
-			printk(KERN_DEBUG "  ");
+			printk(KERN_INFO "  ");
 	}
-	if (cnt % 16)
-		printk(KERN_DEBUG "\n");
+	printk(KERN_INFO "\n");
 }
 
diff --git a/drivers/scsi/qla4xxx/ql4_def.h b/drivers/scsi/qla4xxx/ql4_def.h
index b586f27c3bd..81b5f29254e 100644
--- a/drivers/scsi/qla4xxx/ql4_def.h
+++ b/drivers/scsi/qla4xxx/ql4_def.h
@@ -100,7 +100,6 @@
 #define MAX_SRBS		MAX_CMDS_TO_RISC
 #define MBOX_AEN_REG_COUNT	5
 #define MAX_INIT_RETRIES	5
-#define IOCB_HIWAT_CUSHION	16
 
 /*
  * Buffer sizes
@@ -184,6 +183,11 @@ struct srb {
 	uint16_t cc_stat;
 	u_long r_start;		/* Time we recieve a cmd from OS */
 	u_long u_start;		/* Time when we handed the cmd to F/W */
+
+	/* Used for extended sense / status continuation */
+	uint8_t *req_sense_ptr;
+	uint16_t req_sense_len;
+	uint16_t reserved2;
 };
 
 /*
@@ -302,7 +306,6 @@ struct scsi_qla_host {
 	uint32_t tot_ddbs;
 
 	uint16_t	iocb_cnt;
-	uint16_t	iocb_hiwat;
 
 	/* SRB cache. */
 #define SRB_MIN_REQ	128
@@ -436,6 +439,8 @@ struct scsi_qla_host {
 	/* Map ddb_list entry by FW ddb index */
 	struct ddb_entry *fw_ddb_index_map[MAX_DDB_ENTRIES];
 
+	/* Saved srb for status continuation entry processing */
+	struct srb *status_srb;
 };
 
 static inline int is_qla4010(struct scsi_qla_host *ha)
diff --git a/drivers/scsi/qla4xxx/ql4_fw.h b/drivers/scsi/qla4xxx/ql4_fw.h
index 1b667a70cff..9cd7a608df3 100644
--- a/drivers/scsi/qla4xxx/ql4_fw.h
+++ b/drivers/scsi/qla4xxx/ql4_fw.h
@@ -572,6 +572,7 @@ struct conn_event_log_entry {
  *************************************************************************/
 #define IOCB_MAX_CDB_LEN	    16	/* Bytes in a CBD */
 #define IOCB_MAX_SENSEDATA_LEN	    32	/* Bytes of sense data */
+#define IOCB_MAX_EXT_SENSEDATA_LEN  60  /* Bytes of extended sense data */
 
 /* IOCB header structure */
 struct qla4_header {
@@ -733,6 +734,12 @@ struct status_entry {
 
 };
 
+/* Status Continuation entry */
+struct status_cont_entry {
+       struct qla4_header hdr; /* 00-03 */
+       uint8_t ext_sense_data[IOCB_MAX_EXT_SENSEDATA_LEN]; /* 04-63 */
+};
+
 struct passthru0 {
 	struct qla4_header hdr;		       /* 00-03 */
 	uint32_t handle;	/* 04-07 */
diff --git a/drivers/scsi/qla4xxx/ql4_iocb.c b/drivers/scsi/qla4xxx/ql4_iocb.c
index 912a67494ad..e0c32159749 100644
--- a/drivers/scsi/qla4xxx/ql4_iocb.c
+++ b/drivers/scsi/qla4xxx/ql4_iocb.c
@@ -10,9 +10,42 @@
 #include "ql4_dbg.h"
 #include "ql4_inline.h"
 
-
 #include <scsi/scsi_tcq.h>
 
+static int
+qla4xxx_space_in_req_ring(struct scsi_qla_host *ha, uint16_t req_cnt)
+{
+	uint16_t cnt;
+
+	/* Calculate number of free request entries. */
+	if ((req_cnt + 2) >= ha->req_q_count) {
+		cnt = (uint16_t) le32_to_cpu(ha->shadow_regs->req_q_out);
+		if (ha->request_in < cnt)
+			ha->req_q_count = cnt - ha->request_in;
+		else
+			ha->req_q_count = REQUEST_QUEUE_DEPTH -
+						(ha->request_in - cnt);
+	}
+
+	/* Check if room for request in request ring. */
+	if ((req_cnt + 2) < ha->req_q_count)
+		return 1;
+	else
+		return 0;
+}
+
+static void qla4xxx_advance_req_ring_ptr(struct scsi_qla_host *ha)
+{
+	/* Advance request queue pointer */
+	if (ha->request_in == (REQUEST_QUEUE_DEPTH - 1)) {
+		ha->request_in = 0;
+		ha->request_ptr = ha->request_ring;
+	} else {
+		ha->request_in++;
+		ha->request_ptr++;
+	}
+}
+
 /**
  * qla4xxx_get_req_pkt - returns a valid entry in request queue.
  * @ha: Pointer to host adapter structure.
@@ -26,35 +59,18 @@
 static int qla4xxx_get_req_pkt(struct scsi_qla_host *ha,
 			       struct queue_entry **queue_entry)
 {
-	uint16_t request_in;
-	uint8_t status = QLA_SUCCESS;
-
-	*queue_entry = ha->request_ptr;
+	uint16_t req_cnt = 1;
 
-	/* get the latest request_in and request_out index */
-	request_in = ha->request_in;
-	ha->request_out = (uint16_t) le32_to_cpu(ha->shadow_regs->req_q_out);
-
-	/* Advance request queue pointer and check for queue full */
-	if (request_in == (REQUEST_QUEUE_DEPTH - 1)) {
-		request_in = 0;
-		ha->request_ptr = ha->request_ring;
-	} else {
-		request_in++;
-		ha->request_ptr++;
-	}
-
-	/* request queue is full, try again later */
-	if ((ha->iocb_cnt + 1) >= ha->iocb_hiwat) {
-		/* restore request pointer */
-		ha->request_ptr = *queue_entry;
-		status = QLA_ERROR;
-	} else {
-		ha->request_in = request_in;
+	if (qla4xxx_space_in_req_ring(ha, req_cnt)) {
+		*queue_entry = ha->request_ptr;
 		memset(*queue_entry, 0, sizeof(**queue_entry));
+
+		qla4xxx_advance_req_ring_ptr(ha);
+		ha->req_q_count -= req_cnt;
+		return QLA_SUCCESS;
 	}
 
-	return status;
+	return QLA_ERROR;
 }
 
 /**
@@ -100,21 +116,14 @@ exit_send_marker:
 	return status;
 }
 
-static struct continuation_t1_entry* qla4xxx_alloc_cont_entry(
-	struct scsi_qla_host *ha)
+static struct continuation_t1_entry *
+qla4xxx_alloc_cont_entry(struct scsi_qla_host *ha)
 {
 	struct continuation_t1_entry *cont_entry;
 
 	cont_entry = (struct continuation_t1_entry *)ha->request_ptr;
 
-	/* Advance request queue pointer */
-	if (ha->request_in == (REQUEST_QUEUE_DEPTH - 1)) {
-		ha->request_in = 0;
-		ha->request_ptr = ha->request_ring;
-	} else {
-		ha->request_in++;
-		ha->request_ptr++;
-	}
+	qla4xxx_advance_req_ring_ptr(ha);
 
 	/* Load packet defaults */
 	cont_entry->hdr.entryType = ET_CONTINUE;
@@ -197,13 +206,10 @@ int qla4xxx_send_command_to_isp(struct scsi_qla_host *ha, struct srb * srb)
 	struct scsi_cmnd *cmd = srb->cmd;
 	struct ddb_entry *ddb_entry;
 	struct command_t3_entry *cmd_entry;
-
 	int nseg;
 	uint16_t tot_dsds;
 	uint16_t req_cnt;
-
 	unsigned long flags;
-	uint16_t cnt;
 	uint32_t index;
 	char tag[2];
 
@@ -217,6 +223,19 @@ int qla4xxx_send_command_to_isp(struct scsi_qla_host *ha, struct srb * srb)
 
 	index = (uint32_t)cmd->request->tag;
 
+	/*
+	 * Check to see if adapter is online before placing request on
+	 * request queue.  If a reset occurs and a request is in the queue,
+	 * the firmware will still attempt to process the request, retrieving
+	 * garbage for pointers.
+	 */
+	if (!test_bit(AF_ONLINE, &ha->flags)) {
+		DEBUG2(printk("scsi%ld: %s: Adapter OFFLINE! "
+			      "Do not issue command.\n",
+			      ha->host_no, __func__));
+		goto queuing_error;
+	}
+
 	/* Calculate the number of request entries needed. */
 	nseg = scsi_dma_map(cmd);
 	if (nseg < 0)
@@ -224,17 +243,7 @@ int qla4xxx_send_command_to_isp(struct scsi_qla_host *ha, struct srb * srb)
 	tot_dsds = nseg;
 
 	req_cnt = qla4xxx_calc_request_entries(tot_dsds);
-
-	if (ha->req_q_count < (req_cnt + 2)) {
-		cnt = (uint16_t) le32_to_cpu(ha->shadow_regs->req_q_out);
-		if (ha->request_in < cnt)
-			ha->req_q_count = cnt - ha->request_in;
-		else
-			ha->req_q_count = REQUEST_QUEUE_DEPTH -
-				(ha->request_in - cnt);
-	}
-
-	if (ha->req_q_count < (req_cnt + 2))
+	if (!qla4xxx_space_in_req_ring(ha, req_cnt))
 		goto queuing_error;
 
 	/* total iocbs active */
@@ -286,32 +295,10 @@ int qla4xxx_send_command_to_isp(struct scsi_qla_host *ha, struct srb * srb)
 			break;
 		}
 
-
-	/* Advance request queue pointer */
-	ha->request_in++;
-	if (ha->request_in == REQUEST_QUEUE_DEPTH) {
-		ha->request_in = 0;
-		ha->request_ptr = ha->request_ring;
-	} else
-		ha->request_ptr++;
-
-
+	qla4xxx_advance_req_ring_ptr(ha);
 	qla4xxx_build_scsi_iocbs(srb, cmd_entry, tot_dsds);
 	wmb();
 
-	/*
-	 * Check to see if adapter is online before placing request on
-	 * request queue.  If a reset occurs and a request is in the queue,
-	 * the firmware will still attempt to process the request, retrieving
-	 * garbage for pointers.
-	 */
-	if (!test_bit(AF_ONLINE, &ha->flags)) {
-		DEBUG2(printk("scsi%ld: %s: Adapter OFFLINE! "
-			      "Do not issue command.\n",
-			      ha->host_no, __func__));
-		goto queuing_error;
-	}
-
 	srb->cmd->host_scribble = (unsigned char *)srb;
 
 	/* update counters */
diff --git a/drivers/scsi/qla4xxx/ql4_isr.c b/drivers/scsi/qla4xxx/ql4_isr.c
index 799120fcb9b..8025ee16588 100644
--- a/drivers/scsi/qla4xxx/ql4_isr.c
+++ b/drivers/scsi/qla4xxx/ql4_isr.c
@@ -11,6 +11,98 @@
 #include "ql4_inline.h"
 
 /**
+ * qla4xxx_copy_sense - copy sense data	into cmd sense buffer
+ * @ha: Pointer to host adapter structure.
+ * @sts_entry: Pointer to status entry structure.
+ * @srb: Pointer to srb structure.
+ **/
+static void qla4xxx_copy_sense(struct scsi_qla_host *ha,
+                               struct status_entry *sts_entry,
+                               struct srb *srb)
+{
+	struct scsi_cmnd *cmd = srb->cmd;
+	uint16_t sense_len;
+
+	memset(cmd->sense_buffer, 0, SCSI_SENSE_BUFFERSIZE);
+	sense_len = le16_to_cpu(sts_entry->senseDataByteCnt);
+	if (sense_len == 0)
+		return;
+
+	/* Save total available sense length,
+	 * not to exceed cmd's sense buffer size */
+	sense_len = min_t(uint16_t, sense_len, SCSI_SENSE_BUFFERSIZE);
+	srb->req_sense_ptr = cmd->sense_buffer;
+	srb->req_sense_len = sense_len;
+
+	/* Copy sense from sts_entry pkt */
+	sense_len = min_t(uint16_t, sense_len, IOCB_MAX_SENSEDATA_LEN);
+	memcpy(cmd->sense_buffer, sts_entry->senseData, sense_len);
+
+	DEBUG2(printk(KERN_INFO "scsi%ld:%d:%d:%d: %s: sense key = %x, "
+		"ASL= %02x, ASC/ASCQ = %02x/%02x\n", ha->host_no,
+		cmd->device->channel, cmd->device->id,
+		cmd->device->lun, __func__,
+		sts_entry->senseData[2] & 0x0f,
+		sts_entry->senseData[7],
+		sts_entry->senseData[12],
+		sts_entry->senseData[13]));
+
+	DEBUG5(qla4xxx_dump_buffer(cmd->sense_buffer, sense_len));
+	srb->flags |= SRB_GOT_SENSE;
+
+	/* Update srb, in case a sts_cont pkt follows */
+	srb->req_sense_ptr += sense_len;
+	srb->req_sense_len -= sense_len;
+	if (srb->req_sense_len != 0)
+		ha->status_srb = srb;
+	else
+		ha->status_srb = NULL;
+}
+
+/**
+ * qla4xxx_status_cont_entry - Process a Status Continuations entry.
+ * @ha: SCSI driver HA context
+ * @sts_cont: Entry pointer
+ *
+ * Extended sense data.
+ */
+static void
+qla4xxx_status_cont_entry(struct scsi_qla_host *ha,
+			  struct status_cont_entry *sts_cont)
+{
+	struct srb *srb = ha->status_srb;
+	struct scsi_cmnd *cmd;
+	uint8_t sense_len;
+
+	if (srb == NULL)
+		return;
+
+	cmd = srb->cmd;
+	if (cmd == NULL) {
+		DEBUG2(printk(KERN_INFO "scsi%ld: %s: Cmd already returned "
+			"back to OS srb=%p srb->state:%d\n", ha->host_no,
+			__func__, srb, srb->state));
+		ha->status_srb = NULL;
+		return;
+	}
+
+	/* Copy sense data. */
+	sense_len = min_t(uint16_t, srb->req_sense_len,
+			  IOCB_MAX_EXT_SENSEDATA_LEN);
+	memcpy(srb->req_sense_ptr, sts_cont->ext_sense_data, sense_len);
+	DEBUG5(qla4xxx_dump_buffer(srb->req_sense_ptr, sense_len));
+
+	srb->req_sense_ptr += sense_len;
+	srb->req_sense_len -= sense_len;
+
+	/* Place command on done queue. */
+	if (srb->req_sense_len == 0) {
+		qla4xxx_srb_compl(ha, srb);
+		ha->status_srb = NULL;
+	}
+}
+
+/**
  * qla4xxx_status_entry - processes status IOCBs
  * @ha: Pointer to host adapter structure.
  * @sts_entry: Pointer to status entry structure.
@@ -23,7 +115,6 @@ static void qla4xxx_status_entry(struct scsi_qla_host *ha,
 	struct srb *srb;
 	struct ddb_entry *ddb_entry;
 	uint32_t residual;
-	uint16_t sensebytecnt;
 
 	srb = qla4xxx_del_from_active_array(ha, le32_to_cpu(sts_entry->handle));
 	if (!srb) {
@@ -92,24 +183,7 @@ static void qla4xxx_status_entry(struct scsi_qla_host *ha,
 			break;
 
 		/* Copy Sense Data into sense buffer. */
-		memset(cmd->sense_buffer, 0, SCSI_SENSE_BUFFERSIZE);
-
-		sensebytecnt = le16_to_cpu(sts_entry->senseDataByteCnt);
-		if (sensebytecnt == 0)
-			break;
-
-		memcpy(cmd->sense_buffer, sts_entry->senseData,
-		       min_t(uint16_t, sensebytecnt, SCSI_SENSE_BUFFERSIZE));
-
-		DEBUG2(printk("scsi%ld:%d:%d:%d: %s: sense key = %x, "
-			      "ASC/ASCQ = %02x/%02x\n", ha->host_no,
-			      cmd->device->channel, cmd->device->id,
-			      cmd->device->lun, __func__,
-			      sts_entry->senseData[2] & 0x0f,
-			      sts_entry->senseData[12],
-			      sts_entry->senseData[13]));
-
-		srb->flags |= SRB_GOT_SENSE;
+		qla4xxx_copy_sense(ha, sts_entry, srb);
 		break;
 
 	case SCS_INCOMPLETE:
@@ -176,23 +250,7 @@ static void qla4xxx_status_entry(struct scsi_qla_host *ha,
 				break;
 
 			/* Copy Sense Data into sense buffer. */
-			memset(cmd->sense_buffer, 0, SCSI_SENSE_BUFFERSIZE);
-
-			sensebytecnt =
-				le16_to_cpu(sts_entry->senseDataByteCnt);
-			if (sensebytecnt == 0)
-				break;
-
-			memcpy(cmd->sense_buffer, sts_entry->senseData,
-			       min_t(uint16_t, sensebytecnt, SCSI_SENSE_BUFFERSIZE));
-
-			DEBUG2(printk("scsi%ld:%d:%d:%d: %s: sense key = %x, "
-				      "ASC/ASCQ = %02x/%02x\n", ha->host_no,
-				      cmd->device->channel, cmd->device->id,
-				      cmd->device->lun, __func__,
-				      sts_entry->senseData[2] & 0x0f,
-				      sts_entry->senseData[12],
-				      sts_entry->senseData[13]));
+			qla4xxx_copy_sense(ha, sts_entry, srb);
 		} else {
 			/*
 			 * If RISC reports underrun and target does not
@@ -268,9 +326,10 @@ static void qla4xxx_status_entry(struct scsi_qla_host *ha,
 
 status_entry_exit:
 
-	/* complete the request */
+	/* complete the request, if not waiting for status_continuation pkt */
 	srb->cc_stat = sts_entry->completionStatus;
-	qla4xxx_srb_compl(ha, srb);
+	if (ha->status_srb == NULL)
+		qla4xxx_srb_compl(ha, srb);
 }
 
 /**
@@ -305,10 +364,7 @@ static void qla4xxx_process_response_queue(struct scsi_qla_host * ha)
 		/* process entry */
 		switch (sts_entry->hdr.entryType) {
 		case ET_STATUS:
-			/*
-			 * Common status - Single completion posted in single
-			 * IOSB.
-			 */
+			/* Common status */
 			qla4xxx_status_entry(ha, sts_entry);
 			break;
 
@@ -316,9 +372,8 @@ static void qla4xxx_process_response_queue(struct scsi_qla_host * ha)
 			break;
 
 		case ET_STATUS_CONTINUATION:
-			/* Just throw away the status continuation entries */
-			DEBUG2(printk("scsi%ld: %s: Status Continuation entry "
-				      "- ignoring\n", ha->host_no, __func__));
+			qla4xxx_status_cont_entry(ha,
+				(struct status_cont_entry *) sts_entry);
 			break;
 
 		case ET_COMMAND:
diff --git a/drivers/scsi/qla4xxx/ql4_mbx.c b/drivers/scsi/qla4xxx/ql4_mbx.c
index 051b0f5e8c8..09d6d4b76f3 100644
--- a/drivers/scsi/qla4xxx/ql4_mbx.c
+++ b/drivers/scsi/qla4xxx/ql4_mbx.c
@@ -385,16 +385,6 @@ int qla4xxx_get_firmware_status(struct scsi_qla_host * ha)
 			      mbox_sts[0]));
 		return QLA_ERROR;
 	}
-
-	/* High-water mark of IOCBs */
-	ha->iocb_hiwat = mbox_sts[2];
-	if (ha->iocb_hiwat > IOCB_HIWAT_CUSHION)
-		ha->iocb_hiwat -= IOCB_HIWAT_CUSHION;
-	else
-		dev_info(&ha->pdev->dev, "WARNING!!!  You have less than %d "
-			   "firmware IOCBs available (%d).\n",
-			   IOCB_HIWAT_CUSHION, ha->iocb_hiwat);
-
 	return QLA_SUCCESS;
 }
 
diff --git a/drivers/scsi/qla4xxx/ql4_os.c b/drivers/scsi/qla4xxx/ql4_os.c
index ec9da6ce848..40e3cafb3a9 100644
--- a/drivers/scsi/qla4xxx/ql4_os.c
+++ b/drivers/scsi/qla4xxx/ql4_os.c
@@ -66,6 +66,7 @@ static int qla4xxx_sess_get_param(struct iscsi_cls_session *sess,
 static int qla4xxx_host_get_param(struct Scsi_Host *shost,
 				  enum iscsi_host_param param, char *buf);
 static void qla4xxx_recovery_timedout(struct iscsi_cls_session *session);
+static enum blk_eh_timer_return qla4xxx_eh_cmd_timed_out(struct scsi_cmnd *sc);
 
 /*
  * SCSI host template entry points
@@ -89,6 +90,7 @@ static struct scsi_host_template qla4xxx_driver_template = {
 	.eh_device_reset_handler = qla4xxx_eh_device_reset,
 	.eh_target_reset_handler = qla4xxx_eh_target_reset,
 	.eh_host_reset_handler	= qla4xxx_eh_host_reset,
+	.eh_timed_out		= qla4xxx_eh_cmd_timed_out,
 
 	.slave_configure	= qla4xxx_slave_configure,
 	.slave_alloc		= qla4xxx_slave_alloc,
@@ -124,6 +126,21 @@ static struct iscsi_transport qla4xxx_iscsi_transport = {
 
 static struct scsi_transport_template *qla4xxx_scsi_transport;
 
+static enum blk_eh_timer_return qla4xxx_eh_cmd_timed_out(struct scsi_cmnd *sc)
+{
+	struct iscsi_cls_session *session;
+	struct ddb_entry *ddb_entry;
+
+	session = starget_to_session(scsi_target(sc->device));
+	ddb_entry = session->dd_data;
+
+	/* if we are not logged in then the LLD is going to clean up the cmd */
+	if (atomic_read(&ddb_entry->state) != DDB_STATE_ONLINE)
+		return BLK_EH_RESET_TIMER;
+	else
+		return BLK_EH_NOT_HANDLED;
+}
+
 static void qla4xxx_recovery_timedout(struct iscsi_cls_session *session)
 {
 	struct ddb_entry *ddb_entry = session->dd_data;
@@ -904,18 +921,17 @@ static int qla4xxx_recover_adapter(struct scsi_qla_host *ha,
 	/* Flush any pending ddb changed AENs */
 	qla4xxx_process_aen(ha, FLUSH_DDB_CHANGED_AENS);
 
+	qla4xxx_flush_active_srbs(ha);
+
 	/* Reset the firmware.	If successful, function
 	 * returns with ISP interrupts enabled.
 	 */
-	if (status == QLA_SUCCESS) {
-		DEBUG2(printk("scsi%ld: %s - Performing soft reset..\n",
-			      ha->host_no, __func__));
-		qla4xxx_flush_active_srbs(ha);
-		if (ql4xxx_lock_drvr_wait(ha) == QLA_SUCCESS)
-			status = qla4xxx_soft_reset(ha);
-		else
-			status = QLA_ERROR;
-	}
+	DEBUG2(printk("scsi%ld: %s - Performing soft reset..\n",
+		      ha->host_no, __func__));
+	if (ql4xxx_lock_drvr_wait(ha) == QLA_SUCCESS)
+		status = qla4xxx_soft_reset(ha);
+	else
+		status = QLA_ERROR;
 
 	/* Flush any pending ddb changed AENs */
 	qla4xxx_process_aen(ha, FLUSH_DDB_CHANGED_AENS);
@@ -1527,11 +1543,9 @@ static int qla4xxx_eh_device_reset(struct scsi_cmnd *cmd)
 {
 	struct scsi_qla_host *ha = to_qla_host(cmd->device->host);
 	struct ddb_entry *ddb_entry = cmd->device->hostdata;
-	struct srb *sp;
 	int ret = FAILED, stat;
 
-	sp = (struct srb *) cmd->SCp.ptr;
-	if (!sp || !ddb_entry)
+	if (!ddb_entry)
 		return ret;
 
 	dev_info(&ha->pdev->dev,
@@ -1644,7 +1658,7 @@ static int qla4xxx_eh_host_reset(struct scsi_cmnd *cmd)
 	ha = (struct scsi_qla_host *) cmd->device->host->hostdata;
 
 	dev_info(&ha->pdev->dev,
-		   "scsi(%ld:%d:%d:%d): ADAPTER RESET ISSUED.\n", ha->host_no,
+		   "scsi(%ld:%d:%d:%d): HOST RESET ISSUED.\n", ha->host_no,
 		   cmd->device->channel, cmd->device->id, cmd->device->lun);
 
 	if (qla4xxx_wait_for_hba_online(ha) != QLA_SUCCESS) {
diff --git a/drivers/scsi/qla4xxx/ql4_version.h b/drivers/scsi/qla4xxx/ql4_version.h
index ab984cb89ce..6980cb279c8 100644
--- a/drivers/scsi/qla4xxx/ql4_version.h
+++ b/drivers/scsi/qla4xxx/ql4_version.h
@@ -5,5 +5,5 @@
  * See LICENSE.qla4xxx for copyright and licensing details.
  */
 
-#define QLA4XXX_DRIVER_VERSION	"5.01.00-k8"
+#define QLA4XXX_DRIVER_VERSION	"5.01.00-k9"
 
diff --git a/drivers/scsi/scsi_transport_iscsi.c b/drivers/scsi/scsi_transport_iscsi.c
index 783e33c65eb..b47240ca4b1 100644
--- a/drivers/scsi/scsi_transport_iscsi.c
+++ b/drivers/scsi/scsi_transport_iscsi.c
@@ -990,7 +990,7 @@ int iscsi_offload_mesg(struct Scsi_Host *shost,
 	struct iscsi_uevent *ev;
 	int len = NLMSG_SPACE(sizeof(*ev) + data_size);
 
-	skb = alloc_skb(len, GFP_NOIO);
+	skb = alloc_skb(len, GFP_ATOMIC);
 	if (!skb) {
 		printk(KERN_ERR "can not deliver iscsi offload message:OOM\n");
 		return -ENOMEM;
@@ -1012,7 +1012,7 @@ int iscsi_offload_mesg(struct Scsi_Host *shost,
 
 	memcpy((char *)ev + sizeof(*ev), data, data_size);
 
-	return iscsi_multicast_skb(skb, ISCSI_NL_GRP_UIP, GFP_NOIO);
+	return iscsi_multicast_skb(skb, ISCSI_NL_GRP_UIP, GFP_ATOMIC);
 }
 EXPORT_SYMBOL_GPL(iscsi_offload_mesg);
 
diff --git a/drivers/scsi/sd.c b/drivers/scsi/sd.c
index 5616cd780ff..b7b9fec67a9 100644
--- a/drivers/scsi/sd.c
+++ b/drivers/scsi/sd.c
@@ -1840,6 +1840,18 @@ static void sd_read_block_characteristics(struct scsi_disk *sdkp)
 	kfree(buffer);
 }
 
+static int sd_try_extended_inquiry(struct scsi_device *sdp)
+{
+	/*
+	 * Although VPD inquiries can go to SCSI-2 type devices,
+	 * some USB ones crash on receiving them, and the pages
+	 * we currently ask for are for SPC-3 and beyond
+	 */
+	if (sdp->scsi_level > SCSI_SPC_2)
+		return 1;
+	return 0;
+}
+
 /**
  *	sd_revalidate_disk - called the first time a new disk is seen,
  *	performs disk spin up, read_capacity, etc.
@@ -1877,8 +1889,12 @@ static int sd_revalidate_disk(struct gendisk *disk)
 	 */
 	if (sdkp->media_present) {
 		sd_read_capacity(sdkp, buffer);
-		sd_read_block_limits(sdkp);
-		sd_read_block_characteristics(sdkp);
+
+		if (sd_try_extended_inquiry(sdp)) {
+			sd_read_block_limits(sdkp);
+			sd_read_block_characteristics(sdkp);
+		}
+
 		sd_read_write_protect_flag(sdkp, buffer);
 		sd_read_cache_type(sdkp, buffer);
 		sd_read_app_tag_own(sdkp, buffer);
diff --git a/drivers/serial/atmel_serial.c b/drivers/serial/atmel_serial.c
index 338b15c0a54..607d43a3104 100644
--- a/drivers/serial/atmel_serial.c
+++ b/drivers/serial/atmel_serial.c
@@ -1551,6 +1551,7 @@ static int __devinit atmel_serial_probe(struct platform_device *pdev)
 	if (ret)
 		goto err_add_port;
 
+#ifdef CONFIG_SERIAL_ATMEL_CONSOLE
 	if (atmel_is_console_port(&port->uart)
 			&& ATMEL_CONSOLE_DEVICE->flags & CON_ENABLED) {
 		/*
@@ -1559,6 +1560,7 @@ static int __devinit atmel_serial_probe(struct platform_device *pdev)
 		 */
 		clk_disable(port->clk);
 	}
+#endif
 
 	device_init_wakeup(&pdev->dev, 1);
 	platform_set_drvdata(pdev, port);
diff --git a/drivers/serial/cpm_uart/cpm_uart_cpm2.c b/drivers/serial/cpm_uart/cpm_uart_cpm2.c
index 141c0a3333a..a9802e76b5f 100644
--- a/drivers/serial/cpm_uart/cpm_uart_cpm2.c
+++ b/drivers/serial/cpm_uart/cpm_uart_cpm2.c
@@ -132,7 +132,7 @@ int cpm_uart_allocbuf(struct uart_cpm_port *pinfo, unsigned int is_con)
 	memsz = L1_CACHE_ALIGN(pinfo->rx_nrfifos * pinfo->rx_fifosize) +
 	    L1_CACHE_ALIGN(pinfo->tx_nrfifos * pinfo->tx_fifosize);
 	if (is_con) {
-		mem_addr = alloc_bootmem(memsz);
+		mem_addr = kzalloc(memsz, GFP_NOWAIT);
 		dma_addr = virt_to_bus(mem_addr);
 	}
 	else
diff --git a/drivers/serial/s3c2400.c b/drivers/serial/s3c2400.c
index fb00ed5296e..fed1a9a1ffb 100644
--- a/drivers/serial/s3c2400.c
+++ b/drivers/serial/s3c2400.c
@@ -76,7 +76,7 @@ static int s3c2400_serial_probe(struct platform_device *dev)
 	return s3c24xx_serial_probe(dev, &s3c2400_uart_inf);
 }
 
-static struct platform_driver s3c2400_serial_drv = {
+static struct platform_driver s3c2400_serial_driver = {
 	.probe		= s3c2400_serial_probe,
 	.remove		= __devexit_p(s3c24xx_serial_remove),
 	.driver		= {
@@ -85,16 +85,16 @@ static struct platform_driver s3c2400_serial_drv = {
 	},
 };
 
-s3c24xx_console_init(&s3c2400_serial_drv, &s3c2400_uart_inf);
+s3c24xx_console_init(&s3c2400_serial_driver, &s3c2400_uart_inf);
 
 static inline int s3c2400_serial_init(void)
 {
-	return s3c24xx_serial_init(&s3c2400_serial_drv, &s3c2400_uart_inf);
+	return s3c24xx_serial_init(&s3c2400_serial_driver, &s3c2400_uart_inf);
 }
 
 static inline void s3c2400_serial_exit(void)
 {
-	platform_driver_unregister(&s3c2400_serial_drv);
+	platform_driver_unregister(&s3c2400_serial_driver);
 }
 
 module_init(s3c2400_serial_init);
diff --git a/drivers/serial/s3c2410.c b/drivers/serial/s3c2410.c
index b5d7cbcba2a..c99f0821cae 100644
--- a/drivers/serial/s3c2410.c
+++ b/drivers/serial/s3c2410.c
@@ -88,7 +88,7 @@ static int s3c2410_serial_probe(struct platform_device *dev)
 	return s3c24xx_serial_probe(dev, &s3c2410_uart_inf);
 }
 
-static struct platform_driver s3c2410_serial_drv = {
+static struct platform_driver s3c2410_serial_driver = {
 	.probe		= s3c2410_serial_probe,
 	.remove		= __devexit_p(s3c24xx_serial_remove),
 	.driver		= {
@@ -97,16 +97,16 @@ static struct platform_driver s3c2410_serial_drv = {
 	},
 };
 
-s3c24xx_console_init(&s3c2410_serial_drv, &s3c2410_uart_inf);
+s3c24xx_console_init(&s3c2410_serial_driver, &s3c2410_uart_inf);
 
 static int __init s3c2410_serial_init(void)
 {
-	return s3c24xx_serial_init(&s3c2410_serial_drv, &s3c2410_uart_inf);
+	return s3c24xx_serial_init(&s3c2410_serial_driver, &s3c2410_uart_inf);
 }
 
 static void __exit s3c2410_serial_exit(void)
 {
-	platform_driver_unregister(&s3c2410_serial_drv);
+	platform_driver_unregister(&s3c2410_serial_driver);
 }
 
 module_init(s3c2410_serial_init);
diff --git a/drivers/serial/s3c2412.c b/drivers/serial/s3c2412.c
index 11dcb90bdfe..6e057d8809d 100644
--- a/drivers/serial/s3c2412.c
+++ b/drivers/serial/s3c2412.c
@@ -121,7 +121,7 @@ static int s3c2412_serial_probe(struct platform_device *dev)
 	return s3c24xx_serial_probe(dev, &s3c2412_uart_inf);
 }
 
-static struct platform_driver s3c2412_serial_drv = {
+static struct platform_driver s3c2412_serial_driver = {
 	.probe		= s3c2412_serial_probe,
 	.remove		= __devexit_p(s3c24xx_serial_remove),
 	.driver		= {
@@ -130,16 +130,16 @@ static struct platform_driver s3c2412_serial_drv = {
 	},
 };
 
-s3c24xx_console_init(&s3c2412_serial_drv, &s3c2412_uart_inf);
+s3c24xx_console_init(&s3c2412_serial_driver, &s3c2412_uart_inf);
 
 static inline int s3c2412_serial_init(void)
 {
-	return s3c24xx_serial_init(&s3c2412_serial_drv, &s3c2412_uart_inf);
+	return s3c24xx_serial_init(&s3c2412_serial_driver, &s3c2412_uart_inf);
 }
 
 static inline void s3c2412_serial_exit(void)
 {
-	platform_driver_unregister(&s3c2412_serial_drv);
+	platform_driver_unregister(&s3c2412_serial_driver);
 }
 
 module_init(s3c2412_serial_init);
diff --git a/drivers/serial/s3c2440.c b/drivers/serial/s3c2440.c
index 06c5b0cc47a..69ff5d340f0 100644
--- a/drivers/serial/s3c2440.c
+++ b/drivers/serial/s3c2440.c
@@ -151,7 +151,7 @@ static int s3c2440_serial_probe(struct platform_device *dev)
 	return s3c24xx_serial_probe(dev, &s3c2440_uart_inf);
 }
 
-static struct platform_driver s3c2440_serial_drv = {
+static struct platform_driver s3c2440_serial_driver = {
 	.probe		= s3c2440_serial_probe,
 	.remove		= __devexit_p(s3c24xx_serial_remove),
 	.driver		= {
@@ -160,16 +160,16 @@ static struct platform_driver s3c2440_serial_drv = {
 	},
 };
 
-s3c24xx_console_init(&s3c2440_serial_drv, &s3c2440_uart_inf);
+s3c24xx_console_init(&s3c2440_serial_driver, &s3c2440_uart_inf);
 
 static int __init s3c2440_serial_init(void)
 {
-	return s3c24xx_serial_init(&s3c2440_serial_drv, &s3c2440_uart_inf);
+	return s3c24xx_serial_init(&s3c2440_serial_driver, &s3c2440_uart_inf);
 }
 
 static void __exit s3c2440_serial_exit(void)
 {
-	platform_driver_unregister(&s3c2440_serial_drv);
+	platform_driver_unregister(&s3c2440_serial_driver);
 }
 
 module_init(s3c2440_serial_init);
diff --git a/drivers/serial/s3c24a0.c b/drivers/serial/s3c24a0.c
index 786a067d62a..26c49e18bdd 100644
--- a/drivers/serial/s3c24a0.c
+++ b/drivers/serial/s3c24a0.c
@@ -92,7 +92,7 @@ static int s3c24a0_serial_probe(struct platform_device *dev)
 	return s3c24xx_serial_probe(dev, &s3c24a0_uart_inf);
 }
 
-static struct platform_driver s3c24a0_serial_drv = {
+static struct platform_driver s3c24a0_serial_driver = {
 	.probe		= s3c24a0_serial_probe,
 	.remove		= __devexit_p(s3c24xx_serial_remove),
 	.driver		= {
@@ -101,16 +101,16 @@ static struct platform_driver s3c24a0_serial_drv = {
 	},
 };
 
-s3c24xx_console_init(&s3c24a0_serial_drv, &s3c24a0_uart_inf);
+s3c24xx_console_init(&s3c24a0_serial_driver, &s3c24a0_uart_inf);
 
 static int __init s3c24a0_serial_init(void)
 {
-	return s3c24xx_serial_init(&s3c24a0_serial_drv, &s3c24a0_uart_inf);
+	return s3c24xx_serial_init(&s3c24a0_serial_driver, &s3c24a0_uart_inf);
 }
 
 static void __exit s3c24a0_serial_exit(void)
 {
-	platform_driver_unregister(&s3c24a0_serial_drv);
+	platform_driver_unregister(&s3c24a0_serial_driver);
 }
 
 module_init(s3c24a0_serial_init);
diff --git a/drivers/serial/s3c6400.c b/drivers/serial/s3c6400.c
index 48f1a3781f0..4be92ab5005 100644
--- a/drivers/serial/s3c6400.c
+++ b/drivers/serial/s3c6400.c
@@ -122,7 +122,7 @@ static int s3c6400_serial_probe(struct platform_device *dev)
 	return s3c24xx_serial_probe(dev, &s3c6400_uart_inf);
 }
 
-static struct platform_driver s3c6400_serial_drv = {
+static struct platform_driver s3c6400_serial_driver = {
 	.probe		= s3c6400_serial_probe,
 	.remove		= __devexit_p(s3c24xx_serial_remove),
 	.driver		= {
@@ -131,16 +131,16 @@ static struct platform_driver s3c6400_serial_drv = {
 	},
 };
 
-s3c24xx_console_init(&s3c6400_serial_drv, &s3c6400_uart_inf);
+s3c24xx_console_init(&s3c6400_serial_driver, &s3c6400_uart_inf);
 
 static int __init s3c6400_serial_init(void)
 {
-	return s3c24xx_serial_init(&s3c6400_serial_drv, &s3c6400_uart_inf);
+	return s3c24xx_serial_init(&s3c6400_serial_driver, &s3c6400_uart_inf);
 }
 
 static void __exit s3c6400_serial_exit(void)
 {
-	platform_driver_unregister(&s3c6400_serial_drv);
+	platform_driver_unregister(&s3c6400_serial_driver);
 }
 
 module_init(s3c6400_serial_init);
diff --git a/drivers/serial/serial_ks8695.c b/drivers/serial/serial_ks8695.c
index 998e89dc5aa..e0665630e4d 100644
--- a/drivers/serial/serial_ks8695.c
+++ b/drivers/serial/serial_ks8695.c
@@ -549,7 +549,7 @@ static struct uart_port ks8695uart_ports[SERIAL_KS8695_NR] = {
 		.mapbase	= KS8695_UART_VA,
 		.iotype		= SERIAL_IO_MEM,
 		.irq		= KS8695_IRQ_UART_TX,
-		.uartclk	= CLOCK_TICK_RATE * 16,
+		.uartclk	= KS8695_CLOCK_RATE * 16,
 		.fifosize	= 16,
 		.ops		= &ks8695uart_pops,
 		.flags		= ASYNC_BOOT_AUTOCONF,
diff --git a/drivers/spi/omap2_mcspi.c b/drivers/spi/omap2_mcspi.c
index eee4b6e0af2..9b80ad36dbb 100644
--- a/drivers/spi/omap2_mcspi.c
+++ b/drivers/spi/omap2_mcspi.c
@@ -59,6 +59,8 @@
 
 /* per-register bitmasks: */
 
+#define OMAP2_MCSPI_SYSCONFIG_SMARTIDLE	(2 << 3)
+#define OMAP2_MCSPI_SYSCONFIG_ENAWAKEUP	(1 << 2)
 #define OMAP2_MCSPI_SYSCONFIG_AUTOIDLE	(1 << 0)
 #define OMAP2_MCSPI_SYSCONFIG_SOFTRESET	(1 << 1)
 
@@ -90,6 +92,7 @@
 
 #define OMAP2_MCSPI_CHCTRL_EN		(1 << 0)
 
+#define OMAP2_MCSPI_WAKEUPENABLE_WKEN	(1 << 0)
 
 /* We have 2 DMA channels per CS, one for RX and one for TX */
 struct omap2_mcspi_dma {
@@ -269,7 +272,7 @@ omap2_mcspi_txrx_dma(struct spi_device *spi, struct spi_transfer *xfer)
 
 	if (rx != NULL) {
 		omap_set_dma_transfer_params(mcspi_dma->dma_rx_channel,
-				data_type, element_count, 1,
+				data_type, element_count - 1, 1,
 				OMAP_DMA_SYNC_ELEMENT,
 				mcspi_dma->dma_rx_sync_dev, 1);
 
@@ -300,6 +303,25 @@ omap2_mcspi_txrx_dma(struct spi_device *spi, struct spi_transfer *xfer)
 	if (rx != NULL) {
 		wait_for_completion(&mcspi_dma->dma_rx_completion);
 		dma_unmap_single(NULL, xfer->rx_dma, count, DMA_FROM_DEVICE);
+		omap2_mcspi_set_enable(spi, 0);
+		if (likely(mcspi_read_cs_reg(spi, OMAP2_MCSPI_CHSTAT0)
+				& OMAP2_MCSPI_CHSTAT_RXS)) {
+			u32 w;
+
+			w = mcspi_read_cs_reg(spi, OMAP2_MCSPI_RX0);
+			if (word_len <= 8)
+				((u8 *)xfer->rx_buf)[element_count - 1] = w;
+			else if (word_len <= 16)
+				((u16 *)xfer->rx_buf)[element_count - 1] = w;
+			else /* word_len <= 32 */
+				((u32 *)xfer->rx_buf)[element_count - 1] = w;
+		} else {
+			dev_err(&spi->dev, "DMA RX last word empty");
+			count -= (word_len <= 8)  ? 1 :
+				 (word_len <= 16) ? 2 :
+			       /* word_len <= 32 */ 4;
+		}
+		omap2_mcspi_set_enable(spi, 1);
 	}
 	return count;
 }
@@ -873,8 +895,12 @@ static int __init omap2_mcspi_reset(struct omap2_mcspi *mcspi)
 	} while (!(tmp & OMAP2_MCSPI_SYSSTATUS_RESETDONE));
 
 	mcspi_write_reg(master, OMAP2_MCSPI_SYSCONFIG,
-			/* (3 << 8) | (2 << 3) | */
-			OMAP2_MCSPI_SYSCONFIG_AUTOIDLE);
+			OMAP2_MCSPI_SYSCONFIG_AUTOIDLE |
+			OMAP2_MCSPI_SYSCONFIG_ENAWAKEUP |
+			OMAP2_MCSPI_SYSCONFIG_SMARTIDLE);
+
+	mcspi_write_reg(master, OMAP2_MCSPI_WAKEUPENABLE,
+			OMAP2_MCSPI_WAKEUPENABLE_WKEN);
 
 	omap2_mcspi_set_master_mode(master);
 
diff --git a/drivers/staging/Kconfig b/drivers/staging/Kconfig
index 348bf61a8fe..975ecddbce3 100644
--- a/drivers/staging/Kconfig
+++ b/drivers/staging/Kconfig
@@ -103,8 +103,6 @@ source "drivers/staging/pohmelfs/Kconfig"
 
 source "drivers/staging/stlc45xx/Kconfig"
 
-source "drivers/staging/uc2322/Kconfig"
-
 source "drivers/staging/b3dfg/Kconfig"
 
 source "drivers/staging/phison/Kconfig"
diff --git a/drivers/staging/Makefile b/drivers/staging/Makefile
index 8d61d7b4deb..2241ae1b21e 100644
--- a/drivers/staging/Makefile
+++ b/drivers/staging/Makefile
@@ -34,7 +34,6 @@ obj-$(CONFIG_ANDROID)		+= android/
 obj-$(CONFIG_DST)		+= dst/
 obj-$(CONFIG_POHMELFS)		+= pohmelfs/
 obj-$(CONFIG_STLC45XX)		+= stlc45xx/
-obj-$(CONFIG_USB_SERIAL_ATEN2011)	+= uc2322/
 obj-$(CONFIG_B3DFG)		+= b3dfg/
 obj-$(CONFIG_IDE_PHISON)	+= phison/
 obj-$(CONFIG_PLAN9AUTH)		+= p9auth/
diff --git a/drivers/staging/android/lowmemorykiller.c b/drivers/staging/android/lowmemorykiller.c
index fe72240f5a9..f934393f395 100644
--- a/drivers/staging/android/lowmemorykiller.c
+++ b/drivers/staging/android/lowmemorykiller.c
@@ -96,19 +96,21 @@ static int lowmem_shrink(int nr_to_scan, gfp_t gfp_mask)
 
 	read_lock(&tasklist_lock);
 	for_each_process(p) {
+		struct mm_struct *mm;
 		int oom_adj;
 
 		task_lock(p);
-		if (!p->mm) {
+		mm = p->mm;
+		if (!mm) {
 			task_unlock(p);
 			continue;
 		}
-		oom_adj = p->oomkilladj;
+		oom_adj = mm->oom_adj;
 		if (oom_adj < min_adj) {
 			task_unlock(p);
 			continue;
 		}
-		tasksize = get_mm_rss(p->mm);
+		tasksize = get_mm_rss(mm);
 		task_unlock(p);
 		if (tasksize <= 0)
 			continue;
diff --git a/drivers/staging/b3dfg/Kconfig b/drivers/staging/b3dfg/Kconfig
index 524231047de..9e6573cf97d 100644
--- a/drivers/staging/b3dfg/Kconfig
+++ b/drivers/staging/b3dfg/Kconfig
@@ -1,5 +1,6 @@
 config B3DFG
        tristate "Brontes 3d Frame Framegrabber"
+       depends on PCI
        default n
        ---help---
          This driver provides support for the Brontes 3d Framegrabber
diff --git a/drivers/staging/heci/Kconfig b/drivers/staging/heci/Kconfig
index ae8d588d3a2..c7206f8bcd9 100644
--- a/drivers/staging/heci/Kconfig
+++ b/drivers/staging/heci/Kconfig
@@ -1,5 +1,6 @@
 config HECI
 	tristate "Intel Management Engine Interface (MEI) Support"
+	depends on PCI
 	---help---
 	  The Intel Management Engine Interface (Intel MEI) driver allows
 	  applications to access the Active Management Technology
diff --git a/drivers/staging/rspiusb/rspiusb.c b/drivers/staging/rspiusb/rspiusb.c
index 2f8155c1968..04e2f92c0f6 100644
--- a/drivers/staging/rspiusb/rspiusb.c
+++ b/drivers/staging/rspiusb/rspiusb.c
@@ -716,6 +716,8 @@ static int MapUserBuffer(struct ioctl_struct *io, struct device_extension *pdx)
 		pdx->PixelUrb[frameInfo][i]->transfer_flags =
 		    URB_NO_TRANSFER_DMA_MAP | URB_NO_INTERRUPT;
 	}
+	if (i == 0)
+		return -EINVAL;
 	/* only interrupt when last URB completes */
 	pdx->PixelUrb[frameInfo][--i]->transfer_flags &= ~URB_NO_INTERRUPT;
 	pdx->pendedPixelUrbs[frameInfo] =
diff --git a/drivers/staging/rt2860/rt_linux.h b/drivers/staging/rt2860/rt_linux.h
index 85175c18243..25b53ac3f82 100644
--- a/drivers/staging/rt2860/rt_linux.h
+++ b/drivers/staging/rt2860/rt_linux.h
@@ -43,9 +43,6 @@
 #include "rtmp_type.h"
 #include <linux/module.h>
 #include <linux/kernel.h>
-#if !defined(RT2860) && !defined(RT30xx)
-#include <linux/kthread.h>
-#endif
 
 #include <linux/spinlock.h>
 #include <linux/init.h>
@@ -166,9 +163,7 @@ typedef int (*HARD_START_XMIT_FUNC)(struct sk_buff *skb, struct net_device *net_
 
 #ifndef RT30xx
 typedef	struct pid *	THREAD_PID;
-#ifdef RT2860
 #define	THREAD_PID_INIT_VALUE	NULL
-#endif
 #define	GET_PID(_v)	find_get_pid(_v)
 #define	GET_PID_NUMBER(_v)	pid_nr(_v)
 #define CHECK_PID_LEGALITY(_pid)	if (pid_nr(_pid) >= 0)
@@ -188,12 +183,12 @@ struct os_cookie {
 	dma_addr_t		  		pAd_pa;
 #endif
 #ifdef RT2870
-	struct usb_device	*pUsb_Dev;
+	struct usb_device		*pUsb_Dev;
 
 #ifndef RT30xx
-	struct task_struct	*MLMEThr_task;
-	struct task_struct	*RTUSBCmdThr_task;
-	struct task_struct	*TimerQThr_task;
+	THREAD_PID				MLMEThr_pid;
+	THREAD_PID				RTUSBCmdThr_pid;
+	THREAD_PID				TimerQThr_pid;
 #endif
 #ifdef RT30xx
 	struct pid	*MLMEThr_pid;
diff --git a/drivers/staging/rt2870/2870_main_dev.c b/drivers/staging/rt2870/2870_main_dev.c
index dd01c64fbf6..a4e8696ca39 100644
--- a/drivers/staging/rt2870/2870_main_dev.c
+++ b/drivers/staging/rt2870/2870_main_dev.c
@@ -235,7 +235,7 @@ INT MlmeThread(
 	DBGPRINT(RT_DEBUG_TRACE,( "<---%s\n",__func__));
 
 #ifndef RT30xx
-	pObj->MLMEThr_task = NULL;
+	pObj->MLMEThr_pid = THREAD_PID_INIT_VALUE;
 #endif
 #ifdef RT30xx
 	pObj->MLMEThr_pid = NULL;
@@ -348,7 +348,7 @@ INT RTUSBCmdThread(
 	DBGPRINT(RT_DEBUG_TRACE,( "<---RTUSBCmdThread\n"));
 
 #ifndef RT30xx
-	pObj->RTUSBCmdThr_task = NULL;
+	pObj->RTUSBCmdThr_pid = THREAD_PID_INIT_VALUE;
 #endif
 #ifdef RT30xx
 	pObj->RTUSBCmdThr_pid = NULL;
@@ -447,7 +447,7 @@ INT TimerQThread(
 	DBGPRINT(RT_DEBUG_TRACE,( "<---%s\n",__func__));
 
 #ifndef RT30xx
-	pObj->TimerQThr_task = NULL;
+	pObj->TimerQThr_pid = THREAD_PID_INIT_VALUE;
 #endif
 #ifdef RT30xx
 	pObj->TimerQThr_pid = NULL;
@@ -883,46 +883,69 @@ VOID RT28xxThreadTerminate(
 
 	// Terminate Threads
 #ifndef RT30xx
-	BUG_ON(pObj->TimerQThr_task == NULL);
-	CHECK_PID_LEGALITY(task_pid(pObj->TimerQThr_task))
+	CHECK_PID_LEGALITY(pObj->TimerQThr_pid)
 	{
 		POS_COOKIE pObj = (POS_COOKIE)pAd->OS_Cookie;
 
-		printk(KERN_DEBUG "Terminate the TimerQThr pid=%d!\n",
-			pid_nr(task_pid(pObj->TimerQThr_task)));
+		printk("Terminate the TimerQThr_pid=%d!\n", GET_PID_NUMBER(pObj->TimerQThr_pid));
 		mb();
 		pAd->TimerFunc_kill = 1;
 		mb();
-		kthread_stop(pObj->TimerQThr_task);
-		pObj->TimerQThr_task = NULL;
+		ret = KILL_THREAD_PID(pObj->TimerQThr_pid, SIGTERM, 1);
+		if (ret)
+		{
+			printk(KERN_WARNING "%s: unable to stop TimerQThread, pid=%d, ret=%d!\n",
+					pAd->net_dev->name, GET_PID_NUMBER(pObj->TimerQThr_pid), ret);
+		}
+		else
+		{
+			wait_for_completion(&pAd->TimerQComplete);
+			pObj->TimerQThr_pid = THREAD_PID_INIT_VALUE;
+		}
 	}
 
-	BUG_ON(pObj->MLMEThr_task == NULL);
-	CHECK_PID_LEGALITY(task_pid(pObj->MLMEThr_task))
+	CHECK_PID_LEGALITY(pObj->MLMEThr_pid)
 	{
-		printk(KERN_DEBUG "Terminate the MLMEThr pid=%d!\n",
-			pid_nr(task_pid(pObj->MLMEThr_task)));
+		printk("Terminate the MLMEThr_pid=%d!\n", GET_PID_NUMBER(pObj->MLMEThr_pid));
 		mb();
 		pAd->mlme_kill = 1;
 		//RT28XX_MLME_HANDLER(pAd);
 		mb();
-		kthread_stop(pObj->MLMEThr_task);
-		pObj->MLMEThr_task = NULL;
+		ret = KILL_THREAD_PID(pObj->MLMEThr_pid, SIGTERM, 1);
+		if (ret)
+		{
+			printk (KERN_WARNING "%s: unable to Mlme thread, pid=%d, ret=%d!\n",
+					pAd->net_dev->name, GET_PID_NUMBER(pObj->MLMEThr_pid), ret);
+		}
+		else
+		{
+			//wait_for_completion (&pAd->notify);
+			wait_for_completion (&pAd->mlmeComplete);
+			pObj->MLMEThr_pid = THREAD_PID_INIT_VALUE;
+		}
 	}
 
-	BUG_ON(pObj->RTUSBCmdThr_task == NULL);
-	CHECK_PID_LEGALITY(task_pid(pObj->RTUSBCmdThr_task))
+	CHECK_PID_LEGALITY(pObj->RTUSBCmdThr_pid)
 	{
-		printk(KERN_DEBUG "Terminate the RTUSBCmdThr pid=%d!\n",
-			pid_nr(task_pid(pObj->RTUSBCmdThr_task)));
+		printk("Terminate the RTUSBCmdThr_pid=%d!\n", GET_PID_NUMBER(pObj->RTUSBCmdThr_pid));
 		mb();
 		NdisAcquireSpinLock(&pAd->CmdQLock);
 		pAd->CmdQ.CmdQState = RT2870_THREAD_STOPED;
 		NdisReleaseSpinLock(&pAd->CmdQLock);
 		mb();
 		//RTUSBCMDUp(pAd);
-		kthread_stop(pObj->RTUSBCmdThr_task);
-		pObj->RTUSBCmdThr_task = NULL;
+		ret = KILL_THREAD_PID(pObj->RTUSBCmdThr_pid, SIGTERM, 1);
+		if (ret)
+		{
+			printk(KERN_WARNING "%s: unable to RTUSBCmd thread, pid=%d, ret=%d!\n",
+					pAd->net_dev->name, GET_PID_NUMBER(pObj->RTUSBCmdThr_pid), ret);
+		}
+		else
+		{
+			//wait_for_completion (&pAd->notify);
+			wait_for_completion (&pAd->CmdQComplete);
+			pObj->RTUSBCmdThr_pid = THREAD_PID_INIT_VALUE;
+		}
 	}
 #endif
 #ifdef RT30xx
@@ -1045,7 +1068,7 @@ BOOLEAN RT28XXChipsetCheck(
 			dev_p->descriptor.idProduct == rtusb_usb_id[i].idProduct)
 		{
 #ifndef RT30xx
-			printk(KERN_DEBUG "rt2870: idVendor = 0x%x, idProduct = 0x%x\n",
+			printk("rt2870: idVendor = 0x%x, idProduct = 0x%x\n",
 #endif
 #ifdef RT30xx
 			printk("rt2870: idVendor = 0x%x, idProduct = 0x%x\n",
diff --git a/drivers/staging/rt2870/common/2870_rtmp_init.c b/drivers/staging/rt2870/common/2870_rtmp_init.c
index 0f4c8af97e4..80909e9ab5a 100644
--- a/drivers/staging/rt2870/common/2870_rtmp_init.c
+++ b/drivers/staging/rt2870/common/2870_rtmp_init.c
@@ -700,8 +700,8 @@ NDIS_STATUS AdapterBlockAllocateMemory(
 	usb_dev = pObj->pUsb_Dev;
 
 #ifndef RT30xx
-	pObj->MLMEThr_task		= NULL;
-	pObj->RTUSBCmdThr_task	= NULL;
+	pObj->MLMEThr_pid		= THREAD_PID_INIT_VALUE;
+	pObj->RTUSBCmdThr_pid	= THREAD_PID_INIT_VALUE;
 #endif
 #ifdef RT30xx
 	pObj->MLMEThr_pid	= NULL;
@@ -743,7 +743,7 @@ NDIS_STATUS	 CreateThreads(
 	PRTMP_ADAPTER pAd = net_dev->ml_priv;
 	POS_COOKIE pObj = (POS_COOKIE) pAd->OS_Cookie;
 #ifndef RT30xx
-	struct task_struct *tsk;
+	pid_t pid_number = -1;
 #endif
 #ifdef RT30xx
 	pid_t pid_number;
@@ -762,10 +762,10 @@ NDIS_STATUS	 CreateThreads(
 
 	// Creat MLME Thread
 #ifndef RT30xx
-	pObj->MLMEThr_task = NULL;
-	tsk = kthread_run(MlmeThread, pAd, "%s", pAd->net_dev->name);
-
-	if (IS_ERR(tsk)) {
+	pObj->MLMEThr_pid= THREAD_PID_INIT_VALUE;
+	pid_number = kernel_thread(MlmeThread, pAd, CLONE_VM);
+	if (pid_number < 0)
+	{
 #endif
 #ifdef RT30xx
 	pObj->MLMEThr_pid = NULL;
@@ -778,7 +778,7 @@ NDIS_STATUS	 CreateThreads(
 	}
 
 #ifndef RT30xx
-	pObj->MLMEThr_task = tsk;
+	pObj->MLMEThr_pid = GET_PID(pid_number);
 #endif
 #ifdef RT30xx
 	pObj->MLMEThr_pid = find_get_pid(pid_number);
@@ -788,10 +788,9 @@ NDIS_STATUS	 CreateThreads(
 
 	// Creat Command Thread
 #ifndef RT30xx
-	pObj->RTUSBCmdThr_task = NULL;
-	tsk = kthread_run(RTUSBCmdThread, pAd, "%s", pAd->net_dev->name);
-
-	if (IS_ERR(tsk) < 0)
+	pObj->RTUSBCmdThr_pid= THREAD_PID_INIT_VALUE;
+	pid_number = kernel_thread(RTUSBCmdThread, pAd, CLONE_VM);
+	if (pid_number < 0)
 #endif
 #ifdef RT30xx
 	pObj->RTUSBCmdThr_pid = NULL;
@@ -804,7 +803,7 @@ NDIS_STATUS	 CreateThreads(
 	}
 
 #ifndef RT30xx
-	pObj->RTUSBCmdThr_task = tsk;
+	pObj->RTUSBCmdThr_pid = GET_PID(pid_number);
 #endif
 #ifdef RT30xx
 	pObj->RTUSBCmdThr_pid = find_get_pid(pid_number);
@@ -812,9 +811,9 @@ NDIS_STATUS	 CreateThreads(
 	wait_for_completion(&(pAd->CmdQComplete));
 
 #ifndef RT30xx
-	pObj->TimerQThr_task = NULL;
-	tsk = kthread_run(TimerQThread, pAd, "%s", pAd->net_dev->name);
-	if (IS_ERR(tsk) < 0)
+	pObj->TimerQThr_pid= THREAD_PID_INIT_VALUE;
+	pid_number = kernel_thread(TimerQThread, pAd, CLONE_VM);
+	if (pid_number < 0)
 #endif
 #ifdef RT30xx
 	pObj->TimerQThr_pid = NULL;
@@ -826,7 +825,7 @@ NDIS_STATUS	 CreateThreads(
 		return NDIS_STATUS_FAILURE;
 	}
 #ifndef RT30xx
-	pObj->TimerQThr_task = tsk;
+	pObj->TimerQThr_pid = GET_PID(pid_number);
 #endif
 #ifdef RT30xx
 	pObj->TimerQThr_pid = find_get_pid(pid_number);
diff --git a/drivers/staging/rt2870/common/rtusb_io.c b/drivers/staging/rt2870/common/rtusb_io.c
index fd1b0c18f2a..704b5c2d509 100644
--- a/drivers/staging/rt2870/common/rtusb_io.c
+++ b/drivers/staging/rt2870/common/rtusb_io.c
@@ -984,8 +984,7 @@ NDIS_STATUS	RTUSBEnqueueCmdFromNdis(
 	POS_COOKIE pObj = (POS_COOKIE) pAd->OS_Cookie;
 
 #ifndef RT30xx
-	BUG_ON(pObj->RTUSBCmdThr_task == NULL);
-	CHECK_PID_LEGALITY(task_pid(pObj->RTUSBCmdThr_task))
+	CHECK_PID_LEGALITY(pObj->RTUSBCmdThr_pid)
 #endif
 #ifdef RT30xx
 	if (pObj->RTUSBCmdThr_pid < 0)
diff --git a/drivers/staging/rt2870/rt2870.h b/drivers/staging/rt2870/rt2870.h
index 29e3b53e52a..2b8872b2fd9 100644
--- a/drivers/staging/rt2870/rt2870.h
+++ b/drivers/staging/rt2870/rt2870.h
@@ -79,6 +79,7 @@
 {	\
 	{USB_DEVICE(0x148F,0x2770)}, /* Ralink */		\
 	{USB_DEVICE(0x1737,0x0071)}, /* Linksys WUSB600N */	\
+	{USB_DEVICE(0x1737,0x0070)}, /* Linksys */	\
 	{USB_DEVICE(0x148F,0x2870)}, /* Ralink */		\
 	{USB_DEVICE(0x148F,0x3070)}, /* Ralink */		\
 	{USB_DEVICE(0x0B05,0x1731)}, /* Asus */			\
@@ -93,12 +94,14 @@
 	{USB_DEVICE(0x14B2,0x3C06)}, /* Conceptronic */		\
 	{USB_DEVICE(0x14B2,0x3C28)}, /* Conceptronic */		\
 	{USB_DEVICE(0x2019,0xED06)}, /* Planex Communications, Inc. */		\
+	{USB_DEVICE(0x2019,0xED14)}, /* Planex Communications, Inc. */		\
 	{USB_DEVICE(0x2019,0xAB25)}, /* Planex Communications, Inc. RT3070 */		\
 	{USB_DEVICE(0x07D1,0x3C09)}, /* D-Link */		\
 	{USB_DEVICE(0x07D1,0x3C11)}, /* D-Link */		\
 	{USB_DEVICE(0x14B2,0x3C07)}, /* AL */			\
 	{USB_DEVICE(0x14B2,0x3C12)}, /* AL */           \
 	{USB_DEVICE(0x050D,0x8053)}, /* Belkin */		\
+	{USB_DEVICE(0x050D,0x815C)}, /* Belkin */		\
 	{USB_DEVICE(0x14B2,0x3C23)}, /* Airlink */		\
 	{USB_DEVICE(0x14B2,0x3C27)}, /* Airlink */		\
 	{USB_DEVICE(0x07AA,0x002F)}, /* Corega */		\
@@ -587,16 +590,14 @@ VOID RTUSBBulkRxComplete(purbb_t pUrb, struct pt_regs *pt_regs);
 #define RTUSBMlmeUp(pAd)	        \
 {								    \
 	POS_COOKIE pObj = (POS_COOKIE) pAd->OS_Cookie;	\
-	BUG_ON(pObj->MLMEThr_task == NULL);		    \
-	CHECK_PID_LEGALITY(task_pid(pObj->MLMEThr_task))		    \
+	CHECK_PID_LEGALITY(pObj->MLMEThr_pid)		    \
         up(&(pAd->mlme_semaphore)); \
 }
 
 #define RTUSBCMDUp(pAd)	                \
 {									    \
 	POS_COOKIE pObj = (POS_COOKIE) pAd->OS_Cookie;	\
-	BUG_ON(pObj->RTUSBCmdThr_task == NULL);	    \
-	CHECK_PID_LEGALITY(task_pid(pObj->RTUSBCmdThr_task))	    \
+	CHECK_PID_LEGALITY(pObj->RTUSBCmdThr_pid)	    \
 	    up(&(pAd->RTUSBCmd_semaphore)); \
 }
 #endif
diff --git a/drivers/staging/rtl8192su/ieee80211.h b/drivers/staging/rtl8192su/ieee80211.h
index 0edb09a536f..ea973931803 100644
--- a/drivers/staging/rtl8192su/ieee80211.h
+++ b/drivers/staging/rtl8192su/ieee80211.h
@@ -2645,7 +2645,7 @@ extern int ieee80211_encrypt_fragment(
 	struct sk_buff *frag,
 	int hdr_len);
 
-extern int ieee80211_xmit(struct sk_buff *skb,
+extern int rtl8192_ieee80211_xmit(struct sk_buff *skb,
 			  struct net_device *dev);
 extern void ieee80211_txb_free(struct ieee80211_txb *);
 
diff --git a/drivers/staging/rtl8192su/ieee80211/ieee80211.h b/drivers/staging/rtl8192su/ieee80211/ieee80211.h
index 720bfcbfadc..5e3a2cbed2b 100644
--- a/drivers/staging/rtl8192su/ieee80211/ieee80211.h
+++ b/drivers/staging/rtl8192su/ieee80211/ieee80211.h
@@ -2645,7 +2645,7 @@ extern int ieee80211_encrypt_fragment(
 	struct sk_buff *frag,
 	int hdr_len);
 
-extern int ieee80211_xmit(struct sk_buff *skb,
+extern int rtl8192_ieee80211_xmit(struct sk_buff *skb,
 			  struct net_device *dev);
 extern void ieee80211_txb_free(struct ieee80211_txb *);
 
diff --git a/drivers/staging/rtl8192su/ieee80211/ieee80211_tx.c b/drivers/staging/rtl8192su/ieee80211/ieee80211_tx.c
index 7294572b990..cba12b84be5 100644
--- a/drivers/staging/rtl8192su/ieee80211/ieee80211_tx.c
+++ b/drivers/staging/rtl8192su/ieee80211/ieee80211_tx.c
@@ -618,7 +618,7 @@ void ieee80211_query_seqnum(struct ieee80211_device*ieee, struct sk_buff* skb, u
 	}
 }
 
-int ieee80211_xmit(struct sk_buff *skb, struct net_device *dev)
+int rtl8192_ieee80211_xmit(struct sk_buff *skb, struct net_device *dev)
 {
 #if (LINUX_VERSION_CODE >= KERNEL_VERSION(2,5,0))
 	struct ieee80211_device *ieee = netdev_priv(dev);
@@ -943,5 +943,6 @@ int ieee80211_xmit(struct sk_buff *skb, struct net_device *dev)
 	return 1;
 
 }
+EXPORT_SYMBOL(rtl8192_ieee80211_xmit);
 
 EXPORT_SYMBOL(ieee80211_txb_free);
diff --git a/drivers/staging/rtl8192su/r8192U_core.c b/drivers/staging/rtl8192su/r8192U_core.c
index 4ab250743e8..70f81a8f129 100644
--- a/drivers/staging/rtl8192su/r8192U_core.c
+++ b/drivers/staging/rtl8192su/r8192U_core.c
@@ -12142,7 +12142,7 @@ static const struct net_device_ops rtl8192_netdev_ops = {
 	.ndo_set_mac_address	= r8192_set_mac_adr,
 	.ndo_validate_addr	= eth_validate_addr,
 	.ndo_change_mtu		= eth_change_mtu,
-	.ndo_start_xmit		= ieee80211_xmit,
+	.ndo_start_xmit		= rtl8192_ieee80211_xmit,
 };
 
 #if LINUX_VERSION_CODE > KERNEL_VERSION(2,5,0)
diff --git a/drivers/staging/serqt_usb2/serqt_usb2.c b/drivers/staging/serqt_usb2/serqt_usb2.c
index a9bd4106beb..0fdf8c6dc64 100644
--- a/drivers/staging/serqt_usb2/serqt_usb2.c
+++ b/drivers/staging/serqt_usb2/serqt_usb2.c
@@ -360,18 +360,18 @@ static void qt_read_bulk_callback(struct urb *urb)
 	if (port_paranoia_check(port, __func__) != 0) {
 		dbg("%s - port_paranoia_check, exiting\n", __func__);
 		qt_port->ReadBulkStopped = 1;
-		return;
+		goto exit;
 	}
 
 	if (!serial) {
 		dbg("%s - bad serial pointer, exiting\n", __func__);
-		return;
+		goto exit;
 	}
 	if (qt_port->closePending == 1) {
 		/* Were closing , stop reading */
 		dbg("%s - (qt_port->closepending == 1\n", __func__);
 		qt_port->ReadBulkStopped = 1;
-		return;
+		goto exit;
 	}
 
 	/*
@@ -381,7 +381,7 @@ static void qt_read_bulk_callback(struct urb *urb)
 	 */
 	if (qt_port->RxHolding == 1) {
 		qt_port->ReadBulkStopped = 1;
-		return;
+		goto exit;
 	}
 
 	if (urb->status) {
@@ -389,7 +389,7 @@ static void qt_read_bulk_callback(struct urb *urb)
 
 		dbg("%s - nonzero read bulk status received: %d\n",
 		    __func__, urb->status);
-		return;
+		goto exit;
 	}
 
 	if (tty && RxCount) {
@@ -463,6 +463,8 @@ static void qt_read_bulk_callback(struct urb *urb)
 	}
 
 	schedule_work(&port->work);
+exit:
+	tty_kref_put(tty);
 }
 
 /*
@@ -736,6 +738,11 @@ static int qt_startup(struct usb_serial *serial)
 		if (!qt_port) {
 			dbg("%s: kmalloc for quatech_port (%d) failed!.",
 			    __func__, i);
+			for(--i; i >= 0; i--) {
+				port = serial->port[i];
+				kfree(usb_get_serial_port_data(port));
+				usb_set_serial_port_data(port, NULL);
+			}
 			return -ENOMEM;
 		}
 		spin_lock_init(&qt_port->lock);
@@ -1041,7 +1048,7 @@ static void qt_block_until_empty(struct tty_struct *tty,
 	}
 }
 
-static void qt_close( struct usb_serial_port *port)
+static void qt_close(struct usb_serial_port *port)
 {
 	struct usb_serial *serial = port->serial;
 	struct quatech_port *qt_port;
@@ -1068,6 +1075,7 @@ static void qt_close( struct usb_serial_port *port)
 	/* wait up to for transmitter to empty */
 	if (serial->dev)
 		qt_block_until_empty(tty, qt_port);
+	tty_kref_put(tty);
 
 	/* Close uart channel */
 	status = qt_close_channel(serial, index);
diff --git a/drivers/staging/uc2322/Kconfig b/drivers/staging/uc2322/Kconfig
deleted file mode 100644
index 2e0c6e79df2..00000000000
--- a/drivers/staging/uc2322/Kconfig
+++ /dev/null
@@ -1,10 +0,0 @@
-config USB_SERIAL_ATEN2011
-	tristate "ATEN 2011 USB to serial device support"
-	depends on USB_SERIAL
-	default N
-	---help---
-	  Say Y here if you want to use a ATEN 2011 dual port USB to serial
-	  adapter.
-
-	  To compile this driver as a module, choose M here: the module will be
-	  called aten2011.
diff --git a/drivers/staging/uc2322/Makefile b/drivers/staging/uc2322/Makefile
deleted file mode 100644
index 49c18d6e579..00000000000
--- a/drivers/staging/uc2322/Makefile
+++ /dev/null
@@ -1 +0,0 @@
-obj-$(CONFIG_USB_SERIAL_ATEN2011)	+= aten2011.o
diff --git a/drivers/staging/uc2322/TODO b/drivers/staging/uc2322/TODO
deleted file mode 100644
index c189a64c418..00000000000
--- a/drivers/staging/uc2322/TODO
+++ /dev/null
@@ -1,7 +0,0 @@
-TODO:
-	- checkpatch.pl cleanups
-	- remove dead and useless code (auditing the tty ioctls to
-	  verify that they really are correct and needed.)
-
-Please send any patches to Greg Kroah-Hartman <greg@kroah.com> and
-Russell Lang <gsview@ghostgum.com.au>.
diff --git a/drivers/staging/uc2322/aten2011.c b/drivers/staging/uc2322/aten2011.c
deleted file mode 100644
index 39d0926d1a9..00000000000
--- a/drivers/staging/uc2322/aten2011.c
+++ /dev/null
@@ -1,2430 +0,0 @@
-/*
- * Aten 2011 USB serial driver for 4 port devices
- *
- * Copyright (C) 2000 Inside Out Networks
- * Copyright (C) 2001-2002, 2009 Greg Kroah-Hartman <greg@kroah.com>
- * Copyright (C) 2009 Novell Inc.
- *
- * This program is free software; you can redistribute it and/or modify
- * it under the terms of the GNU General Public License as published by
- * the Free Software Foundation; either version 2 of the License, or
- * (at your option) any later version.
- *
- */
-
-#include <linux/kernel.h>
-#include <linux/errno.h>
-#include <linux/init.h>
-#include <linux/slab.h>
-#include <linux/tty.h>
-#include <linux/tty_driver.h>
-#include <linux/tty_flip.h>
-#include <linux/module.h>
-#include <linux/serial.h>
-#include <linux/uaccess.h>
-#include <linux/usb.h>
-#include <linux/usb/serial.h>
-
-
-#define ZLP_REG1		0x3A	/* Zero_Flag_Reg1 58 */
-#define ZLP_REG2		0x3B	/* Zero_Flag_Reg2 59 */
-#define ZLP_REG3		0x3C	/* Zero_Flag_Reg3 60 */
-#define ZLP_REG4		0x3D	/* Zero_Flag_Reg4 61 */
-#define ZLP_REG5		0x3E	/* Zero_Flag_Reg5 62 */
-
-/* Interrupt Rotinue Defines	*/
-#define SERIAL_IIR_RLS		0x06
-#define SERIAL_IIR_RDA		0x04
-#define SERIAL_IIR_CTI		0x0c
-#define SERIAL_IIR_THR		0x02
-#define SERIAL_IIR_MS		0x00
-
-/* Emulation of the bit mask on the LINE STATUS REGISTER.  */
-#define SERIAL_LSR_DR		0x0001
-#define SERIAL_LSR_OE		0x0002
-#define SERIAL_LSR_PE		0x0004
-#define SERIAL_LSR_FE		0x0008
-#define SERIAL_LSR_BI		0x0010
-#define SERIAL_LSR_THRE		0x0020
-#define SERIAL_LSR_TEMT		0x0040
-#define SERIAL_LSR_FIFOERR	0x0080
-
-/* MSR bit defines(place holders) */
-#define ATEN_MSR_DELTA_CTS	0x10
-#define ATEN_MSR_DELTA_DSR	0x20
-#define ATEN_MSR_DELTA_RI	0x40
-#define ATEN_MSR_DELTA_CD	0x80
-
-/* Serial Port register Address */
-#define RECEIVE_BUFFER_REGISTER		((__u16)(0x00))
-#define TRANSMIT_HOLDING_REGISTER	((__u16)(0x00))
-#define INTERRUPT_ENABLE_REGISTER	((__u16)(0x01))
-#define INTERRUPT_IDENT_REGISTER	((__u16)(0x02))
-#define FIFO_CONTROL_REGISTER		((__u16)(0x02))
-#define LINE_CONTROL_REGISTER		((__u16)(0x03))
-#define MODEM_CONTROL_REGISTER		((__u16)(0x04))
-#define LINE_STATUS_REGISTER		((__u16)(0x05))
-#define MODEM_STATUS_REGISTER		((__u16)(0x06))
-#define SCRATCH_PAD_REGISTER		((__u16)(0x07))
-#define DIVISOR_LATCH_LSB		((__u16)(0x00))
-#define DIVISOR_LATCH_MSB		((__u16)(0x01))
-
-#define SP1_REGISTER			((__u16)(0x00))
-#define CONTROL1_REGISTER		((__u16)(0x01))
-#define CLK_MULTI_REGISTER		((__u16)(0x02))
-#define CLK_START_VALUE_REGISTER	((__u16)(0x03))
-#define DCR1_REGISTER			((__u16)(0x04))
-#define GPIO_REGISTER			((__u16)(0x07))
-
-#define SERIAL_LCR_DLAB			((__u16)(0x0080))
-
-/*
- * URB POOL related defines
- */
-#define NUM_URBS			16	/* URB Count */
-#define URB_TRANSFER_BUFFER_SIZE	32	/* URB Size  */
-
-#define USB_VENDOR_ID_ATENINTL		0x0557
-#define ATENINTL_DEVICE_ID_2011		0x2011
-#define ATENINTL_DEVICE_ID_7820		0x7820
-
-static struct usb_device_id id_table[] = {
-	{ USB_DEVICE(USB_VENDOR_ID_ATENINTL, ATENINTL_DEVICE_ID_2011) },
-	{ USB_DEVICE(USB_VENDOR_ID_ATENINTL, ATENINTL_DEVICE_ID_7820) },
-	{ } /* terminating entry */
-};
-MODULE_DEVICE_TABLE(usb, id_table);
-
-/* This structure holds all of the local port information */
-struct ATENINTL_port {
-	int		port_num;          /*Actual port number in the device(1,2,etc)*/
-	__u8		bulk_out_endpoint;   	/* the bulk out endpoint handle */
-	unsigned char	*bulk_out_buffer;     	/* buffer used for the bulk out endpoint */
-	struct urb	*write_urb;	     	/* write URB for this port */
-	__u8		bulk_in_endpoint;	/* the bulk in endpoint handle */
-	unsigned char	*bulk_in_buffer;	/* the buffer we use for the bulk in endpoint */
-	struct urb	*read_urb;	     	/* read URB for this port */
-	__u8		shadowLCR;		/* last LCR value received */
-	__u8		shadowMCR;		/* last MCR value received */
-	char		open;
-	char		chaseResponsePending;
-	wait_queue_head_t	wait_chase;		/* for handling sleeping while waiting for chase to finish */
-	wait_queue_head_t	wait_command;		/* for handling sleeping while waiting for command to finish */
-	struct async_icount	icount;
-	struct usb_serial_port	*port;			/* loop back to the owner of this object */
-	/*Offsets*/
-	__u8		SpRegOffset;
-	__u8		ControlRegOffset;
-	__u8		DcrRegOffset;
-	/* for processing control URBS in interrupt context */
-	struct urb 	*control_urb;
-	char		*ctrl_buf;
-	int		MsrLsr;
-
-	struct urb	*write_urb_pool[NUM_URBS];
-	/* we pass a pointer to this as the arguement sent to cypress_set_termios old_termios */
-	struct ktermios	tmp_termios;        /* stores the old termios settings */
-	spinlock_t 	lock;                   /* private lock */
-};
-
-/* This structure holds all of the individual serial device information */
-struct ATENINTL_serial {
-	__u8		interrupt_in_endpoint;		/* the interrupt endpoint handle */
-	unsigned char	*interrupt_in_buffer;		/* the buffer we use for the interrupt endpoint */
-	struct urb	*interrupt_read_urb;	/* our interrupt urb */
-	__u8		bulk_in_endpoint;	/* the bulk in endpoint handle */
-	unsigned char	*bulk_in_buffer;		/* the buffer we use for the bulk in endpoint */
-	struct urb 	*read_urb;		/* our bulk read urb */
-	__u8		bulk_out_endpoint;	/* the bulk out endpoint handle */
-	struct usb_serial	*serial;	/* loop back to the owner of this object */
-	int	ATEN2011_spectrum_2or4ports; 	/* this says the number of ports in the device */
-	/* Indicates about the no.of opened ports of an individual USB-serial adapater. */
-	unsigned int	NoOfOpenPorts;
-	/* a flag for Status endpoint polling */
-	unsigned char	status_polling_started;
-};
-
-static void ATEN2011_set_termios(struct tty_struct *tty,
-				 struct usb_serial_port *port,
-				 struct ktermios *old_termios);
-static void ATEN2011_change_port_settings(struct tty_struct *tty,
-					  struct ATENINTL_port *ATEN2011_port,
-					  struct ktermios *old_termios);
-
-/*************************************
- * Bit definitions for each register *
- *************************************/
-#define LCR_BITS_5		0x00	/* 5 bits/char */
-#define LCR_BITS_6		0x01	/* 6 bits/char */
-#define LCR_BITS_7		0x02	/* 7 bits/char */
-#define LCR_BITS_8		0x03	/* 8 bits/char */
-#define LCR_BITS_MASK		0x03	/* Mask for bits/char field */
-
-#define LCR_STOP_1		0x00	/* 1 stop bit */
-#define LCR_STOP_1_5		0x04	/* 1.5 stop bits (if 5   bits/char) */
-#define LCR_STOP_2		0x04	/* 2 stop bits   (if 6-8 bits/char) */
-#define LCR_STOP_MASK		0x04	/* Mask for stop bits field */
-
-#define LCR_PAR_NONE		0x00	/* No parity */
-#define LCR_PAR_ODD		0x08	/* Odd parity */
-#define LCR_PAR_EVEN		0x18	/* Even parity */
-#define LCR_PAR_MARK		0x28	/* Force parity bit to 1 */
-#define LCR_PAR_SPACE		0x38	/* Force parity bit to 0 */
-#define LCR_PAR_MASK		0x38	/* Mask for parity field */
-
-#define LCR_SET_BREAK		0x40	/* Set Break condition */
-#define LCR_DL_ENABLE		0x80	/* Enable access to divisor latch */
-
-#define MCR_DTR			0x01	/* Assert DTR */
-#define MCR_RTS			0x02	/* Assert RTS */
-#define MCR_OUT1		0x04	/* Loopback only: Sets state of RI */
-#define MCR_MASTER_IE		0x08	/* Enable interrupt outputs */
-#define MCR_LOOPBACK		0x10	/* Set internal (digital) loopback mode */
-#define MCR_XON_ANY		0x20	/* Enable any char to exit XOFF mode */
-
-#define ATEN2011_MSR_CTS	0x10	/* Current state of CTS */
-#define ATEN2011_MSR_DSR	0x20	/* Current state of DSR */
-#define ATEN2011_MSR_RI		0x40	/* Current state of RI */
-#define ATEN2011_MSR_CD		0x80	/* Current state of CD */
-
-
-static int debug;
-
-/*
- * Version Information
- */
-#define DRIVER_VERSION "2.0"
-#define DRIVER_DESC "ATENINTL 2011 USB Serial Adapter"
-
-/*
- * Defines used for sending commands to port
- */
-
-#define ATEN_WDR_TIMEOUT	(50)	/* default urb timeout */
-
-/* Requests */
-#define ATEN_RD_RTYPE		0xC0
-#define ATEN_WR_RTYPE		0x40
-#define ATEN_RDREQ		0x0D
-#define ATEN_WRREQ		0x0E
-#define ATEN_CTRL_TIMEOUT	500
-#define VENDOR_READ_LENGTH	(0x01)
-
-/* set to 1 for RS485 mode and 0 for RS232 mode */
-/* FIXME make this somehow dynamic and not build time specific */
-static int RS485mode;
-
-static int set_reg_sync(struct usb_serial_port *port, __u16 reg, __u16 val)
-{
-	struct usb_device *dev = port->serial->dev;
-	val = val & 0x00ff;
-
-	dbg("%s: is %x, value %x", __func__, reg, val);
-
-	return usb_control_msg(dev, usb_sndctrlpipe(dev, 0), ATEN_WRREQ,
-			       ATEN_WR_RTYPE, val, reg, NULL, 0,
-			       ATEN_WDR_TIMEOUT);
-}
-
-static int get_reg_sync(struct usb_serial_port *port, __u16 reg, __u16 *val)
-{
-	struct usb_device *dev = port->serial->dev;
-	int ret;
-
-	ret = usb_control_msg(dev, usb_rcvctrlpipe(dev, 0), ATEN_RDREQ,
-			      ATEN_RD_RTYPE, 0, reg, val, VENDOR_READ_LENGTH,
-			      ATEN_WDR_TIMEOUT);
-	dbg("%s: offset is %x, return val %x", __func__, reg, *val);
-	*val = (*val) & 0x00ff;
-	return ret;
-}
-
-static int set_uart_reg(struct usb_serial_port *port, __u16 reg, __u16 val)
-{
-	struct usb_device *dev = port->serial->dev;
-	struct ATENINTL_serial *a_serial;
-	__u16 minor;
-
-	a_serial = usb_get_serial_data(port->serial);
-	minor = port->serial->minor;
-	if (minor == SERIAL_TTY_NO_MINOR)
-		minor = 0;
-	val = val & 0x00ff;
-
-	/*
-	 * For the UART control registers,
-	 * the application number need to be Or'ed
-	 */
-	if (a_serial->ATEN2011_spectrum_2or4ports == 4)
-		val |= (((__u16)port->number - minor) + 1) << 8;
-	else {
-		if (((__u16) port->number - minor) == 0)
-			val |= (((__u16)port->number - minor) + 1) << 8;
-		else
-			val |= (((__u16)port->number - minor) + 2) << 8;
-	}
-	dbg("%s: application number is %x", __func__, val);
-
-	return usb_control_msg(dev, usb_sndctrlpipe(dev, 0), ATEN_WRREQ,
-			       ATEN_WR_RTYPE, val, reg, NULL, 0,
-			       ATEN_WDR_TIMEOUT);
-}
-
-static int get_uart_reg(struct usb_serial_port *port, __u16 reg, __u16 *val)
-{
-	struct usb_device *dev = port->serial->dev;
-	int ret = 0;
-	__u16 wval;
-	struct ATENINTL_serial *a_serial;
-	__u16 minor = port->serial->minor;
-
-	a_serial = usb_get_serial_data(port->serial);
-	if (minor == SERIAL_TTY_NO_MINOR)
-		minor = 0;
-
-	/* wval is same as application number */
-	if (a_serial->ATEN2011_spectrum_2or4ports == 4)
-		wval = (((__u16)port->number - minor) + 1) << 8;
-	else {
-		if (((__u16) port->number - minor) == 0)
-			wval = (((__u16) port->number - minor) + 1) << 8;
-		else
-			wval = (((__u16) port->number - minor) + 2) << 8;
-	}
-	dbg("%s: application number is %x", __func__, wval);
-	ret = usb_control_msg(dev, usb_rcvctrlpipe(dev, 0), ATEN_RDREQ,
-			      ATEN_RD_RTYPE, wval, reg, val, VENDOR_READ_LENGTH,
-			      ATEN_WDR_TIMEOUT);
-	*val = (*val) & 0x00ff;
-	return ret;
-}
-
-static int handle_newMsr(struct ATENINTL_port *port, __u8 newMsr)
-{
-	struct ATENINTL_port *ATEN2011_port;
-	struct async_icount *icount;
-	ATEN2011_port = port;
-	icount = &ATEN2011_port->icount;
-	if (newMsr &
-	    (ATEN_MSR_DELTA_CTS | ATEN_MSR_DELTA_DSR | ATEN_MSR_DELTA_RI |
-	     ATEN_MSR_DELTA_CD)) {
-		icount = &ATEN2011_port->icount;
-
-		/* update input line counters */
-		if (newMsr & ATEN_MSR_DELTA_CTS)
-			icount->cts++;
-		if (newMsr & ATEN_MSR_DELTA_DSR)
-			icount->dsr++;
-		if (newMsr & ATEN_MSR_DELTA_CD)
-			icount->dcd++;
-		if (newMsr & ATEN_MSR_DELTA_RI)
-			icount->rng++;
-	}
-
-	return 0;
-}
-
-static int handle_newLsr(struct ATENINTL_port *port, __u8 newLsr)
-{
-	struct async_icount *icount;
-
-	dbg("%s - %02x", __func__, newLsr);
-
-	if (newLsr & SERIAL_LSR_BI) {
-		/*
-		 * Parity and Framing errors only count if they occur exclusive
-		 * of a break being received.
-		 */
-		newLsr &= (__u8) (SERIAL_LSR_OE | SERIAL_LSR_BI);
-	}
-
-	/* update input line counters */
-	icount = &port->icount;
-	if (newLsr & SERIAL_LSR_BI)
-		icount->brk++;
-	if (newLsr & SERIAL_LSR_OE)
-		icount->overrun++;
-	if (newLsr & SERIAL_LSR_PE)
-		icount->parity++;
-	if (newLsr & SERIAL_LSR_FE)
-		icount->frame++;
-
-	return 0;
-}
-
-static void ATEN2011_control_callback(struct urb *urb)
-{
-	unsigned char *data;
-	struct ATENINTL_port *ATEN2011_port;
-	__u8 regval = 0x0;
-
-	switch (urb->status) {
-	case 0:
-		/* success */
-		break;
-	case -ECONNRESET:
-	case -ENOENT:
-	case -ESHUTDOWN:
-		/* this urb is terminated, clean up */
-		dbg("%s - urb shutting down with status: %d", __func__,
-		    urb->status);
-		return;
-	default:
-		dbg("%s - nonzero urb status received: %d", __func__,
-		    urb->status);
-		goto exit;
-	}
-
-	ATEN2011_port = (struct ATENINTL_port *)urb->context;
-
-	dbg("%s urb buffer size is %d", __func__, urb->actual_length);
-	dbg("%s ATEN2011_port->MsrLsr is %d port %d", __func__,
-		ATEN2011_port->MsrLsr, ATEN2011_port->port_num);
-	data = urb->transfer_buffer;
-	regval = (__u8) data[0];
-	dbg("%s data is %x", __func__, regval);
-	if (ATEN2011_port->MsrLsr == 0)
-		handle_newMsr(ATEN2011_port, regval);
-	else if (ATEN2011_port->MsrLsr == 1)
-		handle_newLsr(ATEN2011_port, regval);
-
-exit:
-	return;
-}
-
-static int ATEN2011_get_reg(struct ATENINTL_port *ATEN, __u16 Wval, __u16 reg,
-			    __u16 *val)
-{
-	struct usb_device *dev = ATEN->port->serial->dev;
-	struct usb_ctrlrequest *dr = NULL;
-	unsigned char *buffer = NULL;
-	int ret = 0;
-	buffer = (__u8 *) ATEN->ctrl_buf;
-
-	dr = (void *)(buffer + 2);
-	dr->bRequestType = ATEN_RD_RTYPE;
-	dr->bRequest = ATEN_RDREQ;
-	dr->wValue = cpu_to_le16(Wval);
-	dr->wIndex = cpu_to_le16(reg);
-	dr->wLength = cpu_to_le16(2);
-
-	usb_fill_control_urb(ATEN->control_urb, dev, usb_rcvctrlpipe(dev, 0),
-			     (unsigned char *)dr, buffer, 2,
-			     ATEN2011_control_callback, ATEN);
-	ATEN->control_urb->transfer_buffer_length = 2;
-	ret = usb_submit_urb(ATEN->control_urb, GFP_ATOMIC);
-	return ret;
-}
-
-static void ATEN2011_interrupt_callback(struct urb *urb)
-{
-	int result;
-	int length;
-	struct ATENINTL_port *ATEN2011_port;
-	struct ATENINTL_serial *ATEN2011_serial;
-	struct usb_serial *serial;
-	__u16 Data;
-	unsigned char *data;
-	__u8 sp[5], st;
-	int i;
-	__u16 wval;
-	int minor;
-
-	dbg("%s", " : Entering");
-
-	ATEN2011_serial = (struct ATENINTL_serial *)urb->context;
-
-	switch (urb->status) {
-	case 0:
-		/* success */
-		break;
-	case -ECONNRESET:
-	case -ENOENT:
-	case -ESHUTDOWN:
-		/* this urb is terminated, clean up */
-		dbg("%s - urb shutting down with status: %d", __func__,
-		    urb->status);
-		return;
-	default:
-		dbg("%s - nonzero urb status received: %d", __func__,
-		    urb->status);
-		goto exit;
-	}
-	length = urb->actual_length;
-	data = urb->transfer_buffer;
-
-	serial = ATEN2011_serial->serial;
-
-	/* ATENINTL get 5 bytes
-	 * Byte 1 IIR Port 1 (port.number is 0)
-	 * Byte 2 IIR Port 2 (port.number is 1)
-	 * Byte 3 IIR Port 3 (port.number is 2)
-	 * Byte 4 IIR Port 4 (port.number is 3)
-	 * Byte 5 FIFO status for both */
-
-	if (length && length > 5) {
-		dbg("%s", "Wrong data !!!");
-		return;
-	}
-
-	/* MATRIX */
-	if (ATEN2011_serial->ATEN2011_spectrum_2or4ports == 4) {
-		sp[0] = (__u8) data[0];
-		sp[1] = (__u8) data[1];
-		sp[2] = (__u8) data[2];
-		sp[3] = (__u8) data[3];
-		st = (__u8) data[4];
-	} else {
-		sp[0] = (__u8) data[0];
-		sp[1] = (__u8) data[2];
-		/* sp[2]=(__u8)data[2]; */
-		/* sp[3]=(__u8)data[3]; */
-		st = (__u8) data[4];
-
-	}
-	for (i = 0; i < serial->num_ports; i++) {
-		ATEN2011_port = usb_get_serial_port_data(serial->port[i]);
-		minor = serial->minor;
-		if (minor == SERIAL_TTY_NO_MINOR)
-			minor = 0;
-		if ((ATEN2011_serial->ATEN2011_spectrum_2or4ports == 2)
-		    && (i != 0))
-			wval =
-			    (((__u16) serial->port[i]->number -
-			      (__u16) (minor)) + 2) << 8;
-		else
-			wval =
-			    (((__u16) serial->port[i]->number -
-			      (__u16) (minor)) + 1) << 8;
-		if (ATEN2011_port->open != 0) {
-			if (sp[i] & 0x01) {
-				dbg("SP%d No Interrupt !!!", i);
-			} else {
-				switch (sp[i] & 0x0f) {
-				case SERIAL_IIR_RLS:
-					dbg("Serial Port %d: Receiver status error or address bit detected in 9-bit mode", i);
-					ATEN2011_port->MsrLsr = 1;
-					ATEN2011_get_reg(ATEN2011_port, wval,
-							 LINE_STATUS_REGISTER,
-							 &Data);
-					break;
-				case SERIAL_IIR_MS:
-					dbg("Serial Port %d: Modem status change", i);
-					ATEN2011_port->MsrLsr = 0;
-					ATEN2011_get_reg(ATEN2011_port, wval,
-							 MODEM_STATUS_REGISTER,
-							 &Data);
-					break;
-				}
-			}
-		}
-
-	}
-exit:
-	if (ATEN2011_serial->status_polling_started == 0)
-		return;
-
-	result = usb_submit_urb(urb, GFP_ATOMIC);
-	if (result) {
-		dev_err(&urb->dev->dev,
-			"%s - Error %d submitting interrupt urb\n",
-			__func__, result);
-	}
-
-	return;
-}
-
-static void ATEN2011_bulk_in_callback(struct urb *urb)
-{
-	int status;
-	unsigned char *data;
-	struct usb_serial *serial;
-	struct usb_serial_port *port;
-	struct ATENINTL_serial *ATEN2011_serial;
-	struct ATENINTL_port *ATEN2011_port;
-	struct tty_struct *tty;
-
-	if (urb->status) {
-		dbg("nonzero read bulk status received: %d", urb->status);
-		return;
-	}
-
-	ATEN2011_port = (struct ATENINTL_port *)urb->context;
-
-	port = (struct usb_serial_port *)ATEN2011_port->port;
-	serial = port->serial;
-
-	dbg("%s", "Entering...");
-
-	data = urb->transfer_buffer;
-	ATEN2011_serial = usb_get_serial_data(serial);
-
-	if (urb->actual_length) {
-		tty = tty_port_tty_get(&ATEN2011_port->port->port);
-		if (tty) {
-			tty_buffer_request_room(tty, urb->actual_length);
-			tty_insert_flip_string(tty, data, urb->actual_length);
-			tty_flip_buffer_push(tty);
-			tty_kref_put(tty);
-		}
-
-		ATEN2011_port->icount.rx += urb->actual_length;
-		dbg("ATEN2011_port->icount.rx is %d:",
-			ATEN2011_port->icount.rx);
-	}
-
-	if (!ATEN2011_port->read_urb) {
-		dbg("%s", "URB KILLED !!!");
-		return;
-	}
-
-	if (ATEN2011_port->read_urb->status != -EINPROGRESS) {
-		ATEN2011_port->read_urb->dev = serial->dev;
-
-		status = usb_submit_urb(ATEN2011_port->read_urb, GFP_ATOMIC);
-		if (status)
-			dbg("usb_submit_urb(read bulk) failed, status = %d", status);
-	}
-}
-
-static void ATEN2011_bulk_out_data_callback(struct urb *urb)
-{
-	struct ATENINTL_port *ATEN2011_port;
-	struct tty_struct *tty;
-
-	if (urb->status) {
-		dbg("nonzero write bulk status received:%d", urb->status);
-		return;
-	}
-
-	ATEN2011_port = (struct ATENINTL_port *)urb->context;
-
-	dbg("%s", "Entering .........");
-
-	tty = tty_port_tty_get(&ATEN2011_port->port->port);
-
-	if (tty && ATEN2011_port->open)
-		/* tell the tty driver that something has changed */
-		tty_wakeup(tty);
-
-	/* schedule_work(&ATEN2011_port->port->work); */
-	tty_kref_put(tty);
-
-}
-
-#ifdef ATENSerialProbe
-static int ATEN2011_serial_probe(struct usb_serial *serial,
-				 const struct usb_device_id *id)
-{
-
-	/*need to implement the mode_reg reading and updating\
-	   structures usb_serial_ device_type\
-	   (i.e num_ports, num_bulkin,bulkout etc) */
-	/* Also we can update the changes  attach */
-	return 1;
-}
-#endif
-
-static int ATEN2011_open(struct tty_struct *tty, struct usb_serial_port *port,
-			 struct file *filp)
-{
-	int response;
-	int j;
-	struct usb_serial *serial;
-	struct urb *urb;
-	__u16 Data;
-	int status;
-	struct ATENINTL_serial *ATEN2011_serial;
-	struct ATENINTL_port *ATEN2011_port;
-	struct ktermios tmp_termios;
-	int minor;
-
-	serial = port->serial;
-
-	ATEN2011_port = usb_get_serial_port_data(port);
-
-	if (ATEN2011_port == NULL)
-		return -ENODEV;
-
-	ATEN2011_serial = usb_get_serial_data(serial);
-	if (ATEN2011_serial == NULL)
-		return -ENODEV;
-
-	/* increment the number of opened ports counter here */
-	ATEN2011_serial->NoOfOpenPorts++;
-
-	usb_clear_halt(serial->dev, port->write_urb->pipe);
-	usb_clear_halt(serial->dev, port->read_urb->pipe);
-
-	/* Initialising the write urb pool */
-	for (j = 0; j < NUM_URBS; ++j) {
-		urb = usb_alloc_urb(0, GFP_ATOMIC);
-		ATEN2011_port->write_urb_pool[j] = urb;
-
-		if (urb == NULL) {
-			err("No more urbs???");
-			continue;
-		}
-
-		urb->transfer_buffer = NULL;
-		urb->transfer_buffer =
-		    kmalloc(URB_TRANSFER_BUFFER_SIZE, GFP_KERNEL);
-		if (!urb->transfer_buffer) {
-			err("%s-out of memory for urb buffers.", __func__);
-			continue;
-		}
-	}
-
-/*****************************************************************************
- * Initialize ATEN2011 -- Write Init values to corresponding Registers
- *
- * Register Index
- * 1 : IER
- * 2 : FCR
- * 3 : LCR
- * 4 : MCR
- *
- * 0x08 : SP1/2 Control Reg
- *****************************************************************************/
-
-/* NEED to check the fallowing Block */
-
-	Data = 0x0;
-	status = get_reg_sync(port, ATEN2011_port->SpRegOffset, &Data);
-	if (status < 0) {
-		dbg("Reading Spreg failed");
-		return -1;
-	}
-	Data |= 0x80;
-	status = set_reg_sync(port, ATEN2011_port->SpRegOffset, Data);
-	if (status < 0) {
-		dbg("writing Spreg failed");
-		return -1;
-	}
-
-	Data &= ~0x80;
-	status = set_reg_sync(port, ATEN2011_port->SpRegOffset, Data);
-	if (status < 0) {
-		dbg("writing Spreg failed");
-		return -1;
-	}
-
-/* End of block to be checked */
-/**************************CHECK***************************/
-
-	if (RS485mode == 0)
-		Data = 0xC0;
-	else
-		Data = 0x00;
-	status = set_uart_reg(port, SCRATCH_PAD_REGISTER, Data);
-	if (status < 0) {
-		dbg("Writing SCRATCH_PAD_REGISTER failed status-0x%x", status);
-		return -1;
-	} else
-		dbg("SCRATCH_PAD_REGISTER Writing success status%d", status);
-
-/**************************CHECK***************************/
-
-	Data = 0x0;
-	status = get_reg_sync(port, ATEN2011_port->ControlRegOffset, &Data);
-	if (status < 0) {
-		dbg("Reading Controlreg failed");
-		return -1;
-	}
-	Data |= 0x08;		/* Driver done bit */
-	Data |= 0x20;		/* rx_disable */
-	status = 0;
-	status =
-	    set_reg_sync(port, ATEN2011_port->ControlRegOffset, Data);
-	if (status < 0) {
-		dbg("writing Controlreg failed");
-		return -1;
-	}
-	/*
-	 * do register settings here
-	 * Set all regs to the device default values.
-	 * First Disable all interrupts.
-	 */
-
-	Data = 0x00;
-	status = set_uart_reg(port, INTERRUPT_ENABLE_REGISTER, Data);
-	if (status < 0) {
-		dbg("disableing interrupts failed");
-		return -1;
-	}
-	/* Set FIFO_CONTROL_REGISTER to the default value */
-	Data = 0x00;
-	status = set_uart_reg(port, FIFO_CONTROL_REGISTER, Data);
-	if (status < 0) {
-		dbg("Writing FIFO_CONTROL_REGISTER  failed");
-		return -1;
-	}
-
-	Data = 0xcf;		/* chk */
-	status = set_uart_reg(port, FIFO_CONTROL_REGISTER, Data);
-	if (status < 0) {
-		dbg("Writing FIFO_CONTROL_REGISTER  failed");
-		return -1;
-	}
-
-	Data = 0x03;		/* LCR_BITS_8 */
-	status = set_uart_reg(port, LINE_CONTROL_REGISTER, Data);
-	ATEN2011_port->shadowLCR = Data;
-
-	Data = 0x0b;		/* MCR_DTR|MCR_RTS|MCR_MASTER_IE */
-	status = set_uart_reg(port, MODEM_CONTROL_REGISTER, Data);
-	ATEN2011_port->shadowMCR = Data;
-
-#ifdef Check
-	Data = 0x00;
-	status = get_uart_reg(port, LINE_CONTROL_REGISTER, &Data);
-	ATEN2011_port->shadowLCR = Data;
-
-	Data |= SERIAL_LCR_DLAB;	/* data latch enable in LCR 0x80 */
-	status = set_uart_reg(port, LINE_CONTROL_REGISTER, Data);
-
-	Data = 0x0c;
-	status = set_uart_reg(port, DIVISOR_LATCH_LSB, Data);
-
-	Data = 0x0;
-	status = set_uart_reg(port, DIVISOR_LATCH_MSB, Data);
-
-	Data = 0x00;
-	status = get_uart_reg(port, LINE_CONTROL_REGISTER, &Data);
-
-/*      Data = ATEN2011_port->shadowLCR; */	/* data latch disable */
-	Data = Data & ~SERIAL_LCR_DLAB;
-	status = set_uart_reg(port, LINE_CONTROL_REGISTER, Data);
-	ATEN2011_port->shadowLCR = Data;
-#endif
-	/* clearing Bulkin and Bulkout Fifo */
-	Data = 0x0;
-	status = get_reg_sync(port, ATEN2011_port->SpRegOffset, &Data);
-
-	Data = Data | 0x0c;
-	status = set_reg_sync(port, ATEN2011_port->SpRegOffset, Data);
-
-	Data = Data & ~0x0c;
-	status = set_reg_sync(port, ATEN2011_port->SpRegOffset, Data);
-	/* Finally enable all interrupts */
-	Data = 0x0;
-	Data = 0x0c;
-	status = set_uart_reg(port, INTERRUPT_ENABLE_REGISTER, Data);
-
-	/* clearing rx_disable */
-	Data = 0x0;
-	status = get_reg_sync(port, ATEN2011_port->ControlRegOffset, &Data);
-	Data = Data & ~0x20;
-	status = set_reg_sync(port, ATEN2011_port->ControlRegOffset, Data);
-
-	/* rx_negate */
-	Data = 0x0;
-	status = get_reg_sync(port, ATEN2011_port->ControlRegOffset, &Data);
-	Data = Data | 0x10;
-	status = 0;
-	status = set_reg_sync(port, ATEN2011_port->ControlRegOffset, Data);
-
-	/*
-	 * Check to see if we've set up our endpoint info yet
-	 * (can't set it up in ATEN2011_startup as the structures
-	 * were not set up at that time.)
-	 */
-	if (ATEN2011_serial->NoOfOpenPorts == 1) {
-		/* start the status polling here */
-		ATEN2011_serial->status_polling_started = 1;
-		/* If not yet set, Set here */
-		ATEN2011_serial->interrupt_in_buffer =
-		    serial->port[0]->interrupt_in_buffer;
-		ATEN2011_serial->interrupt_in_endpoint =
-		    serial->port[0]->interrupt_in_endpointAddress;
-		ATEN2011_serial->interrupt_read_urb =
-		    serial->port[0]->interrupt_in_urb;
-
-		/* set up interrupt urb */
-		usb_fill_int_urb(ATEN2011_serial->interrupt_read_urb,
-				 serial->dev,
-				 usb_rcvintpipe(serial->dev,
-						ATEN2011_serial->
-						interrupt_in_endpoint),
-				 ATEN2011_serial->interrupt_in_buffer,
-				 ATEN2011_serial->interrupt_read_urb->
-				 transfer_buffer_length,
-				 ATEN2011_interrupt_callback, ATEN2011_serial,
-				 ATEN2011_serial->interrupt_read_urb->interval);
-
-		/* start interrupt read for ATEN2011               *
-		 * will continue as long as ATEN2011 is connected  */
-
-		response =
-		    usb_submit_urb(ATEN2011_serial->interrupt_read_urb,
-				   GFP_KERNEL);
-		if (response) {
-			dbg("%s - Error %d submitting interrupt urb",
-				__func__, response);
-		}
-
-	}
-
-	/*
-	 * See if we've set up our endpoint info yet
-	 * (can't set it up in ATEN2011_startup as the
-	 * structures were not set up at that time.)
-	 */
-
-	dbg("port number is %d", port->number);
-	dbg("serial number is %d", port->serial->minor);
-	dbg("Bulkin endpoint is %d", port->bulk_in_endpointAddress);
-	dbg("BulkOut endpoint is %d", port->bulk_out_endpointAddress);
-	dbg("Interrupt endpoint is %d",
-		port->interrupt_in_endpointAddress);
-	dbg("port's number in the device is %d", ATEN2011_port->port_num);
-	ATEN2011_port->bulk_in_buffer = port->bulk_in_buffer;
-	ATEN2011_port->bulk_in_endpoint = port->bulk_in_endpointAddress;
-	ATEN2011_port->read_urb = port->read_urb;
-	ATEN2011_port->bulk_out_endpoint = port->bulk_out_endpointAddress;
-
-	minor = port->serial->minor;
-	if (minor == SERIAL_TTY_NO_MINOR)
-		minor = 0;
-
-	/* set up our bulk in urb */
-	if ((ATEN2011_serial->ATEN2011_spectrum_2or4ports == 2)
-	    && (((__u16) port->number - (__u16) (minor)) != 0)) {
-		usb_fill_bulk_urb(ATEN2011_port->read_urb, serial->dev,
-				  usb_rcvbulkpipe(serial->dev,
-						  (port->
-						   bulk_in_endpointAddress +
-						   2)), port->bulk_in_buffer,
-				  ATEN2011_port->read_urb->
-				  transfer_buffer_length,
-				  ATEN2011_bulk_in_callback, ATEN2011_port);
-	} else
-		usb_fill_bulk_urb(ATEN2011_port->read_urb,
-				  serial->dev,
-				  usb_rcvbulkpipe(serial->dev,
-						  port->
-						  bulk_in_endpointAddress),
-				  port->bulk_in_buffer,
-				  ATEN2011_port->read_urb->
-				  transfer_buffer_length,
-				  ATEN2011_bulk_in_callback, ATEN2011_port);
-
-	dbg("ATEN2011_open: bulkin endpoint is %d",
-		port->bulk_in_endpointAddress);
-	response = usb_submit_urb(ATEN2011_port->read_urb, GFP_KERNEL);
-	if (response) {
-		err("%s - Error %d submitting control urb", __func__,
-		    response);
-	}
-
-	/* initialize our wait queues */
-	init_waitqueue_head(&ATEN2011_port->wait_chase);
-	init_waitqueue_head(&ATEN2011_port->wait_command);
-
-	/* initialize our icount structure */
-	memset(&(ATEN2011_port->icount), 0x00, sizeof(ATEN2011_port->icount));
-
-	/* initialize our port settings */
-	ATEN2011_port->shadowMCR = MCR_MASTER_IE;	/* Must set to enable ints! */
-	ATEN2011_port->chaseResponsePending = 0;
-	/* send a open port command */
-	ATEN2011_port->open = 1;
-	/* ATEN2011_change_port_settings(ATEN2011_port,old_termios); */
-	/* Setup termios */
-	ATEN2011_set_termios(tty, port, &tmp_termios);
-	ATEN2011_port->icount.tx = 0;
-	ATEN2011_port->icount.rx = 0;
-
-	dbg("usb_serial serial:%x       ATEN2011_port:%x\nATEN2011_serial:%x      usb_serial_port port:%x",
-	     (unsigned int)serial, (unsigned int)ATEN2011_port,
-	     (unsigned int)ATEN2011_serial, (unsigned int)port);
-
-	return 0;
-
-}
-
-static int ATEN2011_chars_in_buffer(struct tty_struct *tty)
-{
-	struct usb_serial_port *port = tty->driver_data;
-	int i;
-	int chars = 0;
-	struct ATENINTL_port *ATEN2011_port;
-
-	/* dbg("%s"," ATEN2011_chars_in_buffer:entering ..........."); */
-
-	ATEN2011_port = usb_get_serial_port_data(port);
-	if (ATEN2011_port == NULL) {
-		dbg("%s", "ATEN2011_break:leaving ...........");
-		return -1;
-	}
-
-	for (i = 0; i < NUM_URBS; ++i)
-		if (ATEN2011_port->write_urb_pool[i]->status == -EINPROGRESS)
-			chars += URB_TRANSFER_BUFFER_SIZE;
-
-	dbg("%s - returns %d", __func__, chars);
-	return chars;
-
-}
-
-static void ATEN2011_block_until_tx_empty(struct tty_struct *tty,
-					  struct ATENINTL_port *ATEN2011_port)
-{
-	int timeout = HZ / 10;
-	int wait = 30;
-	int count;
-
-	while (1) {
-		count = ATEN2011_chars_in_buffer(tty);
-
-		/* Check for Buffer status */
-		if (count <= 0)
-			return;
-
-		/* Block the thread for a while */
-		interruptible_sleep_on_timeout(&ATEN2011_port->wait_chase,
-					       timeout);
-
-		/* No activity.. count down section */
-		wait--;
-		if (wait == 0) {
-			dbg("%s - TIMEOUT", __func__);
-			return;
-		} else {
-			/* Reset timout value back to seconds */
-			wait = 30;
-		}
-	}
-}
-
-static void ATEN2011_close(struct tty_struct *tty, struct usb_serial_port *port,
-			   struct file *filp)
-{
-	struct usb_serial *serial;
-	struct ATENINTL_serial *ATEN2011_serial;
-	struct ATENINTL_port *ATEN2011_port;
-	int no_urbs;
-	__u16 Data;
-
-	dbg("%s", "ATEN2011_close:entering...");
-	serial = port->serial;
-
-	/* take the Adpater and port's private data */
-	ATEN2011_serial = usb_get_serial_data(serial);
-	ATEN2011_port = usb_get_serial_port_data(port);
-	if ((ATEN2011_serial == NULL) || (ATEN2011_port == NULL))
-		return;
-
-	if (serial->dev) {
-		/* flush and block(wait) until tx is empty */
-		ATEN2011_block_until_tx_empty(tty, ATEN2011_port);
-	}
-	/* kill the ports URB's */
-	for (no_urbs = 0; no_urbs < NUM_URBS; no_urbs++)
-		usb_kill_urb(ATEN2011_port->write_urb_pool[no_urbs]);
-	/* Freeing Write URBs */
-	for (no_urbs = 0; no_urbs < NUM_URBS; ++no_urbs) {
-		kfree(ATEN2011_port->write_urb_pool[no_urbs]->transfer_buffer);
-		usb_free_urb(ATEN2011_port->write_urb_pool[no_urbs]);
-	}
-	/* While closing port, shutdown all bulk read, write  *
-	 * and interrupt read if they exists                  */
-	if (serial->dev) {
-		if (ATEN2011_port->write_urb) {
-			dbg("%s", "Shutdown bulk write");
-			usb_kill_urb(ATEN2011_port->write_urb);
-		}
-		if (ATEN2011_port->read_urb) {
-			dbg("%s", "Shutdown bulk read");
-			usb_kill_urb(ATEN2011_port->read_urb);
-		}
-		if ((&ATEN2011_port->control_urb)) {
-			dbg("%s", "Shutdown control read");
-			/* usb_kill_urb (ATEN2011_port->control_urb); */
-
-		}
-	}
-	/* if(ATEN2011_port->ctrl_buf != NULL) */
-		/* kfree(ATEN2011_port->ctrl_buf); */
-	/* decrement the no.of open ports counter of an individual USB-serial adapter. */
-	ATEN2011_serial->NoOfOpenPorts--;
-	dbg("NoOfOpenPorts in close%d:in port%d",
-		ATEN2011_serial->NoOfOpenPorts, port->number);
-	if (ATEN2011_serial->NoOfOpenPorts == 0) {
-		/* stop the stus polling here */
-		ATEN2011_serial->status_polling_started = 0;
-		if (ATEN2011_serial->interrupt_read_urb) {
-			dbg("%s", "Shutdown interrupt_read_urb");
-			/* ATEN2011_serial->interrupt_in_buffer=NULL; */
-			/* usb_kill_urb (ATEN2011_serial->interrupt_read_urb); */
-		}
-	}
-	if (ATEN2011_port->write_urb) {
-		/* if this urb had a transfer buffer already (old tx) free it */
-		kfree(ATEN2011_port->write_urb->transfer_buffer);
-		usb_free_urb(ATEN2011_port->write_urb);
-	}
-
-	/* clear the MCR & IER */
-	Data = 0x00;
-	set_uart_reg(port, MODEM_CONTROL_REGISTER, Data);
-	Data = 0x00;
-	set_uart_reg(port, INTERRUPT_ENABLE_REGISTER, Data);
-
-	ATEN2011_port->open = 0;
-	dbg("%s", "Leaving ............");
-
-}
-
-static void ATEN2011_block_until_chase_response(struct tty_struct *tty,
-						struct ATENINTL_port
-						*ATEN2011_port)
-{
-	int timeout = 1 * HZ;
-	int wait = 10;
-	int count;
-
-	while (1) {
-		count = ATEN2011_chars_in_buffer(tty);
-
-		/* Check for Buffer status */
-		if (count <= 0) {
-			ATEN2011_port->chaseResponsePending = 0;
-			return;
-		}
-
-		/* Block the thread for a while */
-		interruptible_sleep_on_timeout(&ATEN2011_port->wait_chase,
-					       timeout);
-		/* No activity.. count down section */
-		wait--;
-		if (wait == 0) {
-			dbg("%s - TIMEOUT", __func__);
-			return;
-		} else {
-			/* Reset timout value back to seconds */
-			wait = 10;
-		}
-	}
-
-}
-
-static void ATEN2011_break(struct tty_struct *tty, int break_state)
-{
-	struct usb_serial_port *port = tty->driver_data;
-	unsigned char data;
-	struct usb_serial *serial;
-	struct ATENINTL_serial *ATEN2011_serial;
-	struct ATENINTL_port *ATEN2011_port;
-
-	dbg("%s", "Entering ...........");
-	dbg("ATEN2011_break: Start");
-
-	serial = port->serial;
-
-	ATEN2011_serial = usb_get_serial_data(serial);
-	ATEN2011_port = usb_get_serial_port_data(port);
-
-	if ((ATEN2011_serial == NULL) || (ATEN2011_port == NULL))
-		return;
-
-	/* flush and chase */
-	ATEN2011_port->chaseResponsePending = 1;
-
-	if (serial->dev) {
-		/* flush and block until tx is empty */
-		ATEN2011_block_until_chase_response(tty, ATEN2011_port);
-	}
-
-	if (break_state == -1)
-		data = ATEN2011_port->shadowLCR | LCR_SET_BREAK;
-	else
-		data = ATEN2011_port->shadowLCR & ~LCR_SET_BREAK;
-
-	ATEN2011_port->shadowLCR = data;
-	dbg("ATEN2011_break ATEN2011_port->shadowLCR is %x",
-		ATEN2011_port->shadowLCR);
-	set_uart_reg(port, LINE_CONTROL_REGISTER, ATEN2011_port->shadowLCR);
-
-	return;
-}
-
-static int ATEN2011_write_room(struct tty_struct *tty)
-{
-	struct usb_serial_port *port = tty->driver_data;
-	int i;
-	int room = 0;
-	struct ATENINTL_port *ATEN2011_port;
-
-	ATEN2011_port = usb_get_serial_port_data(port);
-	if (ATEN2011_port == NULL) {
-		dbg("%s", "ATEN2011_break:leaving ...........");
-		return -1;
-	}
-
-	for (i = 0; i < NUM_URBS; ++i)
-		if (ATEN2011_port->write_urb_pool[i]->status != -EINPROGRESS)
-			room += URB_TRANSFER_BUFFER_SIZE;
-
-	dbg("%s - returns %d", __func__, room);
-	return room;
-
-}
-
-static int ATEN2011_write(struct tty_struct *tty, struct usb_serial_port *port,
-			  const unsigned char *data, int count)
-{
-	int status;
-	int i;
-	int bytes_sent = 0;
-	int transfer_size;
-	int minor;
-
-	struct ATENINTL_port *ATEN2011_port;
-	struct usb_serial *serial;
-	struct ATENINTL_serial *ATEN2011_serial;
-	struct urb *urb;
-	const unsigned char *current_position = data;
-	unsigned char *data1;
-	dbg("%s", "entering ...........");
-
-	serial = port->serial;
-
-	ATEN2011_port = usb_get_serial_port_data(port);
-	if (ATEN2011_port == NULL) {
-		dbg("%s", "ATEN2011_port is NULL");
-		return -1;
-	}
-
-	ATEN2011_serial = usb_get_serial_data(serial);
-	if (ATEN2011_serial == NULL) {
-		dbg("%s", "ATEN2011_serial is NULL");
-		return -1;
-	}
-
-	/* try to find a free urb in the list */
-	urb = NULL;
-
-	for (i = 0; i < NUM_URBS; ++i) {
-		if (ATEN2011_port->write_urb_pool[i]->status != -EINPROGRESS) {
-			urb = ATEN2011_port->write_urb_pool[i];
-			dbg("URB:%d", i);
-			break;
-		}
-	}
-
-	if (urb == NULL) {
-		dbg("%s - no more free urbs", __func__);
-		goto exit;
-	}
-
-	if (urb->transfer_buffer == NULL) {
-		urb->transfer_buffer =
-		    kmalloc(URB_TRANSFER_BUFFER_SIZE, GFP_KERNEL);
-
-		if (urb->transfer_buffer == NULL) {
-			err("%s no more kernel memory...", __func__);
-			goto exit;
-		}
-	}
-	transfer_size = min(count, URB_TRANSFER_BUFFER_SIZE);
-
-	memcpy(urb->transfer_buffer, current_position, transfer_size);
-	/* usb_serial_debug_data (__FILE__, __func__, transfer_size, urb->transfer_buffer); */
-
-	/* fill urb with data and submit  */
-	minor = port->serial->minor;
-	if (minor == SERIAL_TTY_NO_MINOR)
-		minor = 0;
-	if ((ATEN2011_serial->ATEN2011_spectrum_2or4ports == 2)
-	    && (((__u16) port->number - (__u16) (minor)) != 0)) {
-		usb_fill_bulk_urb(urb, ATEN2011_serial->serial->dev,
-				  usb_sndbulkpipe(ATEN2011_serial->serial->dev,
-						  (port->
-						   bulk_out_endpointAddress) +
-						  2), urb->transfer_buffer,
-				  transfer_size,
-				  ATEN2011_bulk_out_data_callback,
-				  ATEN2011_port);
-	} else
-
-		usb_fill_bulk_urb(urb,
-				  ATEN2011_serial->serial->dev,
-				  usb_sndbulkpipe(ATEN2011_serial->serial->dev,
-						  port->
-						  bulk_out_endpointAddress),
-				  urb->transfer_buffer, transfer_size,
-				  ATEN2011_bulk_out_data_callback,
-				  ATEN2011_port);
-
-	data1 = urb->transfer_buffer;
-	dbg("bulkout endpoint is %d", port->bulk_out_endpointAddress);
-	/* for(i=0;i < urb->actual_length;i++) */
-		/* dbg("Data is %c ",data1[i]); */
-
-	/* send it down the pipe */
-	status = usb_submit_urb(urb, GFP_ATOMIC);
-
-	if (status) {
-		err("%s - usb_submit_urb(write bulk) failed with status = %d",
-		    __func__, status);
-		bytes_sent = status;
-		goto exit;
-	}
-	bytes_sent = transfer_size;
-	ATEN2011_port->icount.tx += transfer_size;
-	dbg("ATEN2011_port->icount.tx is %d:", ATEN2011_port->icount.tx);
-
-exit:
-	return bytes_sent;
-}
-
-static void ATEN2011_throttle(struct tty_struct *tty)
-{
-	struct usb_serial_port *port = tty->driver_data;
-	struct ATENINTL_port *ATEN2011_port;
-	int status;
-
-	dbg("- port %d", port->number);
-
-	ATEN2011_port = usb_get_serial_port_data(port);
-
-	if (ATEN2011_port == NULL)
-		return;
-
-	if (!ATEN2011_port->open) {
-		dbg("%s", "port not opened");
-		return;
-	}
-
-	dbg("%s", "Entering .......... ");
-
-	if (!tty) {
-		dbg("%s - no tty available", __func__);
-		return;
-	}
-
-	/* if we are implementing XON/XOFF, send the stop character */
-	if (I_IXOFF(tty)) {
-		unsigned char stop_char = STOP_CHAR(tty);
-		status = ATEN2011_write(tty, port, &stop_char, 1);
-		if (status <= 0)
-			return;
-	}
-
-	/* if we are implementing RTS/CTS, toggle that line */
-	if (tty->termios->c_cflag & CRTSCTS) {
-		ATEN2011_port->shadowMCR &= ~MCR_RTS;
-		status = set_uart_reg(port, MODEM_CONTROL_REGISTER,
-				      ATEN2011_port->shadowMCR);
-		if (status < 0)
-			return;
-	}
-
-	return;
-}
-
-static void ATEN2011_unthrottle(struct tty_struct *tty)
-{
-	struct usb_serial_port *port = tty->driver_data;
-	int status;
-	struct ATENINTL_port *ATEN2011_port = usb_get_serial_port_data(port);
-
-	if (ATEN2011_port == NULL)
-		return;
-
-	if (!ATEN2011_port->open) {
-		dbg("%s - port not opened", __func__);
-		return;
-	}
-
-	dbg("%s", "Entering .......... ");
-
-	if (!tty) {
-		dbg("%s - no tty available", __func__);
-		return;
-	}
-
-	/* if we are implementing XON/XOFF, send the start character */
-	if (I_IXOFF(tty)) {
-		unsigned char start_char = START_CHAR(tty);
-		status = ATEN2011_write(tty, port, &start_char, 1);
-		if (status <= 0)
-			return;
-	}
-
-	/* if we are implementing RTS/CTS, toggle that line */
-	if (tty->termios->c_cflag & CRTSCTS) {
-		ATEN2011_port->shadowMCR |= MCR_RTS;
-		status = set_uart_reg(port, MODEM_CONTROL_REGISTER,
-				      ATEN2011_port->shadowMCR);
-		if (status < 0)
-			return;
-	}
-
-	return;
-}
-
-static int ATEN2011_tiocmget(struct tty_struct *tty, struct file *file)
-{
-	struct usb_serial_port *port = tty->driver_data;
-	struct ATENINTL_port *ATEN2011_port;
-	unsigned int result;
-	__u16 msr;
-	__u16 mcr;
-	/* unsigned int mcr; */
-	int status = 0;
-	ATEN2011_port = usb_get_serial_port_data(port);
-
-	dbg("%s - port %d", __func__, port->number);
-
-	if (ATEN2011_port == NULL)
-		return -ENODEV;
-
-	status = get_uart_reg(port, MODEM_STATUS_REGISTER, &msr);
-	status = get_uart_reg(port, MODEM_CONTROL_REGISTER, &mcr);
-	/* mcr = ATEN2011_port->shadowMCR; */
-	/* COMMENT2: the Fallowing three line are commented for updating only MSR values */
-	result = ((mcr & MCR_DTR) ? TIOCM_DTR : 0)
-	    | ((mcr & MCR_RTS) ? TIOCM_RTS : 0)
-	    | ((mcr & MCR_LOOPBACK) ? TIOCM_LOOP : 0)
-	    | ((msr & ATEN2011_MSR_CTS) ? TIOCM_CTS : 0)
-	    | ((msr & ATEN2011_MSR_CD) ? TIOCM_CAR : 0)
-	    | ((msr & ATEN2011_MSR_RI) ? TIOCM_RI : 0)
-	    | ((msr & ATEN2011_MSR_DSR) ? TIOCM_DSR : 0);
-
-	dbg("%s - 0x%04X", __func__, result);
-
-	return result;
-}
-
-static int ATEN2011_tiocmset(struct tty_struct *tty, struct file *file,
-			     unsigned int set, unsigned int clear)
-{
-	struct usb_serial_port *port = tty->driver_data;
-	struct ATENINTL_port *ATEN2011_port;
-	unsigned int mcr;
-	unsigned int status;
-
-	dbg("%s - port %d", __func__, port->number);
-
-	ATEN2011_port = usb_get_serial_port_data(port);
-
-	if (ATEN2011_port == NULL)
-		return -ENODEV;
-
-	mcr = ATEN2011_port->shadowMCR;
-	if (clear & TIOCM_RTS)
-		mcr &= ~MCR_RTS;
-	if (clear & TIOCM_DTR)
-		mcr &= ~MCR_DTR;
-	if (clear & TIOCM_LOOP)
-		mcr &= ~MCR_LOOPBACK;
-
-	if (set & TIOCM_RTS)
-		mcr |= MCR_RTS;
-	if (set & TIOCM_DTR)
-		mcr |= MCR_DTR;
-	if (set & TIOCM_LOOP)
-		mcr |= MCR_LOOPBACK;
-
-	ATEN2011_port->shadowMCR = mcr;
-
-	status = set_uart_reg(port, MODEM_CONTROL_REGISTER, mcr);
-	if (status < 0) {
-		dbg("setting MODEM_CONTROL_REGISTER Failed");
-		return -1;
-	}
-
-	return 0;
-}
-
-static void ATEN2011_set_termios(struct tty_struct *tty,
-				 struct usb_serial_port *port,
-				 struct ktermios *old_termios)
-{
-	int status;
-	unsigned int cflag;
-	struct usb_serial *serial;
-	struct ATENINTL_port *ATEN2011_port;
-
-	dbg("ATEN2011_set_termios: START");
-
-	serial = port->serial;
-
-	ATEN2011_port = usb_get_serial_port_data(port);
-
-	if (ATEN2011_port == NULL)
-		return;
-
-	if (!ATEN2011_port->open) {
-		dbg("%s - port not opened", __func__);
-		return;
-	}
-
-	dbg("%s", "setting termios - ");
-
-	cflag = tty->termios->c_cflag;
-
-	dbg("%s - cflag %08x iflag %08x", __func__,
-	    tty->termios->c_cflag, RELEVANT_IFLAG(tty->termios->c_iflag));
-
-	if (old_termios) {
-		dbg("%s - old clfag %08x old iflag %08x", __func__,
-		    old_termios->c_cflag, RELEVANT_IFLAG(old_termios->c_iflag));
-	}
-
-	dbg("%s - port %d", __func__, port->number);
-
-	/* change the port settings to the new ones specified */
-
-	ATEN2011_change_port_settings(tty, ATEN2011_port, old_termios);
-
-	if (!ATEN2011_port->read_urb) {
-		dbg("%s", "URB KILLED !!!!!");
-		return;
-	}
-
-	if (ATEN2011_port->read_urb->status != -EINPROGRESS) {
-		ATEN2011_port->read_urb->dev = serial->dev;
-		status = usb_submit_urb(ATEN2011_port->read_urb, GFP_ATOMIC);
-		if (status) {
-			dbg
-			    (" usb_submit_urb(read bulk) failed, status = %d",
-			     status);
-		}
-	}
-	return;
-}
-
-static int get_lsr_info(struct tty_struct *tty,
-			struct ATENINTL_port *ATEN2011_port,
-			unsigned int __user *value)
-{
-	int count;
-	unsigned int result = 0;
-
-	count = ATEN2011_chars_in_buffer(tty);
-	if (count == 0) {
-		dbg("%s -- Empty", __func__);
-		result = TIOCSER_TEMT;
-	}
-
-	if (copy_to_user(value, &result, sizeof(int)))
-		return -EFAULT;
-	return 0;
-}
-
-static int get_number_bytes_avail(struct tty_struct *tty,
-				  struct ATENINTL_port *ATEN2011_port,
-				  unsigned int __user *value)
-{
-	unsigned int result = 0;
-
-	if (!tty)
-		return -ENOIOCTLCMD;
-
-	result = tty->read_cnt;
-
-	dbg("%s(%d) = %d", __func__, ATEN2011_port->port->number, result);
-	if (copy_to_user(value, &result, sizeof(int)))
-		return -EFAULT;
-
-	return -ENOIOCTLCMD;
-}
-
-static int set_modem_info(struct ATENINTL_port *ATEN2011_port, unsigned int cmd,
-			  unsigned int __user *value)
-{
-	unsigned int mcr;
-	unsigned int arg;
-	__u16 Data;
-	int status;
-	struct usb_serial_port *port;
-
-	if (ATEN2011_port == NULL)
-		return -1;
-
-	port = (struct usb_serial_port *)ATEN2011_port->port;
-
-	mcr = ATEN2011_port->shadowMCR;
-
-	if (copy_from_user(&arg, value, sizeof(int)))
-		return -EFAULT;
-
-	switch (cmd) {
-	case TIOCMBIS:
-		if (arg & TIOCM_RTS)
-			mcr |= MCR_RTS;
-		if (arg & TIOCM_DTR)
-			mcr |= MCR_RTS;
-		if (arg & TIOCM_LOOP)
-			mcr |= MCR_LOOPBACK;
-		break;
-
-	case TIOCMBIC:
-		if (arg & TIOCM_RTS)
-			mcr &= ~MCR_RTS;
-		if (arg & TIOCM_DTR)
-			mcr &= ~MCR_RTS;
-		if (arg & TIOCM_LOOP)
-			mcr &= ~MCR_LOOPBACK;
-		break;
-
-	case TIOCMSET:
-		/* turn off the RTS and DTR and LOOPBACK
-		 * and then only turn on what was asked to */
-		mcr &= ~(MCR_RTS | MCR_DTR | MCR_LOOPBACK);
-		mcr |= ((arg & TIOCM_RTS) ? MCR_RTS : 0);
-		mcr |= ((arg & TIOCM_DTR) ? MCR_DTR : 0);
-		mcr |= ((arg & TIOCM_LOOP) ? MCR_LOOPBACK : 0);
-		break;
-	}
-
-	ATEN2011_port->shadowMCR = mcr;
-
-	Data = ATEN2011_port->shadowMCR;
-	status = set_uart_reg(port, MODEM_CONTROL_REGISTER, Data);
-	if (status < 0) {
-		dbg("setting MODEM_CONTROL_REGISTER Failed");
-		return -1;
-	}
-
-	return 0;
-}
-
-static int get_modem_info(struct ATENINTL_port *ATEN2011_port,
-			  unsigned int __user *value)
-{
-	unsigned int result = 0;
-	__u16 msr;
-	unsigned int mcr = ATEN2011_port->shadowMCR;
-	int status;
-
-	status = get_uart_reg(ATEN2011_port->port, MODEM_STATUS_REGISTER, &msr);
-	result = ((mcr & MCR_DTR) ? TIOCM_DTR : 0)	/* 0x002 */
-	    |((mcr & MCR_RTS) ? TIOCM_RTS : 0)	/* 0x004 */
-	    |((msr & ATEN2011_MSR_CTS) ? TIOCM_CTS : 0)	/* 0x020 */
-	    |((msr & ATEN2011_MSR_CD) ? TIOCM_CAR : 0)	/* 0x040 */
-	    |((msr & ATEN2011_MSR_RI) ? TIOCM_RI : 0)	/* 0x080 */
-	    |((msr & ATEN2011_MSR_DSR) ? TIOCM_DSR : 0);	/* 0x100 */
-
-	dbg("%s -- %x", __func__, result);
-
-	if (copy_to_user(value, &result, sizeof(int)))
-		return -EFAULT;
-	return 0;
-}
-
-static int get_serial_info(struct ATENINTL_port *ATEN2011_port,
-			   struct serial_struct __user *retinfo)
-{
-	struct serial_struct tmp;
-
-	if (ATEN2011_port == NULL)
-		return -1;
-
-	if (!retinfo)
-		return -EFAULT;
-
-	memset(&tmp, 0, sizeof(tmp));
-
-	tmp.type = PORT_16550A;
-	tmp.line = ATEN2011_port->port->serial->minor;
-	if (tmp.line == SERIAL_TTY_NO_MINOR)
-		tmp.line = 0;
-	tmp.port = ATEN2011_port->port->number;
-	tmp.irq = 0;
-	tmp.flags = ASYNC_SKIP_TEST | ASYNC_AUTO_IRQ;
-	tmp.xmit_fifo_size = NUM_URBS * URB_TRANSFER_BUFFER_SIZE;
-	tmp.baud_base = 9600;
-	tmp.close_delay = 5 * HZ;
-	tmp.closing_wait = 30 * HZ;
-
-	if (copy_to_user(retinfo, &tmp, sizeof(*retinfo)))
-		return -EFAULT;
-	return 0;
-}
-
-static int ATEN2011_ioctl(struct tty_struct *tty, struct file *file,
-			  unsigned int cmd, unsigned long arg)
-{
-	struct usb_serial_port *port = tty->driver_data;
-	struct ATENINTL_port *ATEN2011_port;
-	struct async_icount cnow;
-	struct async_icount cprev;
-	struct serial_icounter_struct icount;
-	int ATENret = 0;
-	unsigned int __user *user_arg = (unsigned int __user *)arg;
-
-	ATEN2011_port = usb_get_serial_port_data(port);
-
-	if (ATEN2011_port == NULL)
-		return -1;
-
-	dbg("%s - port %d, cmd = 0x%x", __func__, port->number, cmd);
-
-	switch (cmd) {
-		/* return number of bytes available */
-
-	case TIOCINQ:
-		dbg("%s (%d) TIOCINQ", __func__, port->number);
-		return get_number_bytes_avail(tty, ATEN2011_port, user_arg);
-		break;
-
-	case TIOCOUTQ:
-		dbg("%s (%d) TIOCOUTQ", __func__, port->number);
-		return put_user(ATEN2011_chars_in_buffer(tty), user_arg);
-		break;
-
-	case TIOCSERGETLSR:
-		dbg("%s (%d) TIOCSERGETLSR", __func__, port->number);
-		return get_lsr_info(tty, ATEN2011_port, user_arg);
-		return 0;
-
-	case TIOCMBIS:
-	case TIOCMBIC:
-	case TIOCMSET:
-		dbg("%s (%d) TIOCMSET/TIOCMBIC/TIOCMSET", __func__,
-		    port->number);
-		ATENret = set_modem_info(ATEN2011_port, cmd, user_arg);
-		return ATENret;
-
-	case TIOCMGET:
-		dbg("%s (%d) TIOCMGET", __func__, port->number);
-		return get_modem_info(ATEN2011_port, user_arg);
-
-	case TIOCGSERIAL:
-		dbg("%s (%d) TIOCGSERIAL", __func__, port->number);
-		return get_serial_info(ATEN2011_port,
-				       (struct serial_struct __user *)arg);
-
-	case TIOCSSERIAL:
-		dbg("%s (%d) TIOCSSERIAL", __func__, port->number);
-		break;
-
-	case TIOCMIWAIT:
-		dbg("%s (%d) TIOCMIWAIT", __func__, port->number);
-		cprev = ATEN2011_port->icount;
-		while (1) {
-			/* see if a signal did it */
-			if (signal_pending(current))
-				return -ERESTARTSYS;
-			cnow = ATEN2011_port->icount;
-			if (cnow.rng == cprev.rng && cnow.dsr == cprev.dsr &&
-			    cnow.dcd == cprev.dcd && cnow.cts == cprev.cts)
-				return -EIO;	/* no change => error */
-			if (((arg & TIOCM_RNG) && (cnow.rng != cprev.rng)) ||
-			    ((arg & TIOCM_DSR) && (cnow.dsr != cprev.dsr)) ||
-			    ((arg & TIOCM_CD) && (cnow.dcd != cprev.dcd)) ||
-			    ((arg & TIOCM_CTS) && (cnow.cts != cprev.cts))) {
-				return 0;
-			}
-			cprev = cnow;
-		}
-		/* NOTREACHED */
-		break;
-
-	case TIOCGICOUNT:
-		cnow = ATEN2011_port->icount;
-		icount.cts = cnow.cts;
-		icount.dsr = cnow.dsr;
-		icount.rng = cnow.rng;
-		icount.dcd = cnow.dcd;
-		icount.rx = cnow.rx;
-		icount.tx = cnow.tx;
-		icount.frame = cnow.frame;
-		icount.overrun = cnow.overrun;
-		icount.parity = cnow.parity;
-		icount.brk = cnow.brk;
-		icount.buf_overrun = cnow.buf_overrun;
-
-		dbg("%s (%d) TIOCGICOUNT RX=%d, TX=%d", __func__,
-		    port->number, icount.rx, icount.tx);
-		if (copy_to_user((void __user *)arg, &icount, sizeof(icount)))
-			return -EFAULT;
-		return 0;
-
-	default:
-		break;
-	}
-
-	return -ENOIOCTLCMD;
-}
-
-static int ATEN2011_calc_baud_rate_divisor(int baudRate, int *divisor,
-					   __u16 *clk_sel_val)
-{
-	dbg("%s - %d", __func__, baudRate);
-
-	if (baudRate <= 115200) {
-		*divisor = 115200 / baudRate;
-		*clk_sel_val = 0x0;
-	}
-	if ((baudRate > 115200) && (baudRate <= 230400)) {
-		*divisor = 230400 / baudRate;
-		*clk_sel_val = 0x10;
-	} else if ((baudRate > 230400) && (baudRate <= 403200)) {
-		*divisor = 403200 / baudRate;
-		*clk_sel_val = 0x20;
-	} else if ((baudRate > 403200) && (baudRate <= 460800)) {
-		*divisor = 460800 / baudRate;
-		*clk_sel_val = 0x30;
-	} else if ((baudRate > 460800) && (baudRate <= 806400)) {
-		*divisor = 806400 / baudRate;
-		*clk_sel_val = 0x40;
-	} else if ((baudRate > 806400) && (baudRate <= 921600)) {
-		*divisor = 921600 / baudRate;
-		*clk_sel_val = 0x50;
-	} else if ((baudRate > 921600) && (baudRate <= 1572864)) {
-		*divisor = 1572864 / baudRate;
-		*clk_sel_val = 0x60;
-	} else if ((baudRate > 1572864) && (baudRate <= 3145728)) {
-		*divisor = 3145728 / baudRate;
-		*clk_sel_val = 0x70;
-	}
-	return 0;
-}
-
-static int ATEN2011_send_cmd_write_baud_rate(struct ATENINTL_port
-					     *ATEN2011_port, int baudRate)
-{
-	int divisor = 0;
-	int status;
-	__u16 Data;
-	unsigned char number;
-	__u16 clk_sel_val;
-	struct usb_serial_port *port;
-	int minor;
-
-	if (ATEN2011_port == NULL)
-		return -1;
-
-	port = (struct usb_serial_port *)ATEN2011_port->port;
-
-	dbg("%s", "Entering .......... ");
-
-	minor = ATEN2011_port->port->serial->minor;
-	if (minor == SERIAL_TTY_NO_MINOR)
-		minor = 0;
-	number = ATEN2011_port->port->number - minor;
-
-	dbg("%s - port = %d, baud = %d", __func__,
-	    ATEN2011_port->port->number, baudRate);
-	/* reset clk_uart_sel in spregOffset */
-	if (baudRate > 115200) {
-#ifdef HW_flow_control
-		/*
-		 * NOTE: need to see the pther register to modify
-		 * setting h/w flow control bit to 1;
-		 */
-		/* Data = ATEN2011_port->shadowMCR; */
-		Data = 0x2b;
-		ATEN2011_port->shadowMCR = Data;
-		status = set_uart_reg(port, MODEM_CONTROL_REGISTER, Data);
-		if (status < 0) {
-			dbg("Writing spreg failed in set_serial_baud");
-			return -1;
-		}
-#endif
-
-	} else {
-#ifdef HW_flow_control
-		/* setting h/w flow control bit to 0; */
-		/* Data = ATEN2011_port->shadowMCR; */
-		Data = 0xb;
-		ATEN2011_port->shadowMCR = Data;
-		status = set_uart_reg(port, MODEM_CONTROL_REGISTER, Data);
-		if (status < 0) {
-			dbg("Writing spreg failed in set_serial_baud");
-			return -1;
-		}
-#endif
-
-	}
-
-	if (1)			/* baudRate <= 115200) */ {
-		clk_sel_val = 0x0;
-		Data = 0x0;
-		status =
-		    ATEN2011_calc_baud_rate_divisor(baudRate, &divisor,
-						    &clk_sel_val);
-		status = get_reg_sync(port, ATEN2011_port->SpRegOffset, &Data);
-		if (status < 0) {
-			dbg("reading spreg failed in set_serial_baud");
-			return -1;
-		}
-		Data = (Data & 0x8f) | clk_sel_val;
-		status = set_reg_sync(port, ATEN2011_port->SpRegOffset, Data);
-		if (status < 0) {
-			dbg("Writing spreg failed in set_serial_baud");
-			return -1;
-		}
-		/* Calculate the Divisor */
-
-		if (status) {
-			err("%s - bad baud rate", __func__);
-			dbg("%s", "bad baud rate");
-			return status;
-		}
-		/* Enable access to divisor latch */
-		Data = ATEN2011_port->shadowLCR | SERIAL_LCR_DLAB;
-		ATEN2011_port->shadowLCR = Data;
-		set_uart_reg(port, LINE_CONTROL_REGISTER, Data);
-
-		/* Write the divisor */
-		Data = (unsigned char)(divisor & 0xff);
-		dbg("set_serial_baud Value to write DLL is %x", Data);
-		set_uart_reg(port, DIVISOR_LATCH_LSB, Data);
-
-		Data = (unsigned char)((divisor & 0xff00) >> 8);
-		dbg("set_serial_baud Value to write DLM is %x", Data);
-		set_uart_reg(port, DIVISOR_LATCH_MSB, Data);
-
-		/* Disable access to divisor latch */
-		Data = ATEN2011_port->shadowLCR & ~SERIAL_LCR_DLAB;
-		ATEN2011_port->shadowLCR = Data;
-		set_uart_reg(port, LINE_CONTROL_REGISTER, Data);
-
-	}
-
-	return status;
-}
-
-static void ATEN2011_change_port_settings(struct tty_struct *tty,
-					  struct ATENINTL_port *ATEN2011_port,
-					  struct ktermios *old_termios)
-{
-	int baud;
-	unsigned cflag;
-	unsigned iflag;
-	__u8 lData;
-	__u8 lParity;
-	__u8 lStop;
-	int status;
-	__u16 Data;
-	struct usb_serial_port *port;
-	struct usb_serial *serial;
-
-	if (ATEN2011_port == NULL)
-		return;
-
-	port = (struct usb_serial_port *)ATEN2011_port->port;
-
-	serial = port->serial;
-
-	dbg("%s - port %d", __func__, ATEN2011_port->port->number);
-
-	if (!ATEN2011_port->open) {
-		dbg("%s - port not opened", __func__);
-		return;
-	}
-
-	if ((!tty) || (!tty->termios)) {
-		dbg("%s - no tty structures", __func__);
-		return;
-	}
-
-	dbg("%s", "Entering .......... ");
-
-	lData = LCR_BITS_8;
-	lStop = LCR_STOP_1;
-	lParity = LCR_PAR_NONE;
-
-	cflag = tty->termios->c_cflag;
-	iflag = tty->termios->c_iflag;
-
-	/* Change the number of bits */
-
-	/* COMMENT1: the below Line"if(cflag & CSIZE)" is added for the errors we get for serial loop data test i.e serial_loopback.pl -v */
-	/* if(cflag & CSIZE) */
-	{
-		switch (cflag & CSIZE) {
-		case CS5:
-			lData = LCR_BITS_5;
-			break;
-
-		case CS6:
-			lData = LCR_BITS_6;
-			break;
-
-		case CS7:
-			lData = LCR_BITS_7;
-			break;
-		default:
-		case CS8:
-			lData = LCR_BITS_8;
-			break;
-		}
-	}
-	/* Change the Parity bit */
-	if (cflag & PARENB) {
-		if (cflag & PARODD) {
-			lParity = LCR_PAR_ODD;
-			dbg("%s - parity = odd", __func__);
-		} else {
-			lParity = LCR_PAR_EVEN;
-			dbg("%s - parity = even", __func__);
-		}
-
-	} else {
-		dbg("%s - parity = none", __func__);
-	}
-
-	if (cflag & CMSPAR)
-		lParity = lParity | 0x20;
-
-	/* Change the Stop bit */
-	if (cflag & CSTOPB) {
-		lStop = LCR_STOP_2;
-		dbg("%s - stop bits = 2", __func__);
-	} else {
-		lStop = LCR_STOP_1;
-		dbg("%s - stop bits = 1", __func__);
-	}
-
-	/* Update the LCR with the correct value */
-	ATEN2011_port->shadowLCR &=
-	    ~(LCR_BITS_MASK | LCR_STOP_MASK | LCR_PAR_MASK);
-	ATEN2011_port->shadowLCR |= (lData | lParity | lStop);
-
-	dbg
-	    ("ATEN2011_change_port_settings ATEN2011_port->shadowLCR is %x",
-	     ATEN2011_port->shadowLCR);
-	/* Disable Interrupts */
-	Data = 0x00;
-	set_uart_reg(port, INTERRUPT_ENABLE_REGISTER, Data);
-
-	Data = 0x00;
-	set_uart_reg(port, FIFO_CONTROL_REGISTER, Data);
-
-	Data = 0xcf;
-	set_uart_reg(port, FIFO_CONTROL_REGISTER, Data);
-
-	/* Send the updated LCR value to the ATEN2011 */
-	Data = ATEN2011_port->shadowLCR;
-
-	set_uart_reg(port, LINE_CONTROL_REGISTER, Data);
-
-	Data = 0x00b;
-	ATEN2011_port->shadowMCR = Data;
-	set_uart_reg(port, MODEM_CONTROL_REGISTER, Data);
-	Data = 0x00b;
-	set_uart_reg(port, MODEM_CONTROL_REGISTER, Data);
-
-	/* set up the MCR register and send it to the ATEN2011 */
-
-	ATEN2011_port->shadowMCR = MCR_MASTER_IE;
-	if (cflag & CBAUD)
-		ATEN2011_port->shadowMCR |= (MCR_DTR | MCR_RTS);
-
-	if (cflag & CRTSCTS)
-		ATEN2011_port->shadowMCR |= (MCR_XON_ANY);
-	else
-		ATEN2011_port->shadowMCR &= ~(MCR_XON_ANY);
-
-	Data = ATEN2011_port->shadowMCR;
-	set_uart_reg(port, MODEM_CONTROL_REGISTER, Data);
-
-	/* Determine divisor based on baud rate */
-	baud = tty_get_baud_rate(tty);
-
-	if (!baud) {
-		/* pick a default, any default... */
-		dbg("%s", "Picked default baud...");
-		baud = 9600;
-	}
-
-	dbg("%s - baud rate = %d", __func__, baud);
-	status = ATEN2011_send_cmd_write_baud_rate(ATEN2011_port, baud);
-
-	/* Enable Interrupts */
-	Data = 0x0c;
-	set_uart_reg(port, INTERRUPT_ENABLE_REGISTER, Data);
-
-	if (ATEN2011_port->read_urb->status != -EINPROGRESS) {
-		ATEN2011_port->read_urb->dev = serial->dev;
-
-		status = usb_submit_urb(ATEN2011_port->read_urb, GFP_ATOMIC);
-
-		if (status) {
-			dbg
-			    (" usb_submit_urb(read bulk) failed, status = %d",
-			     status);
-		}
-	}
-	dbg
-	    ("ATEN2011_change_port_settings ATEN2011_port->shadowLCR is End %x",
-	     ATEN2011_port->shadowLCR);
-
-	return;
-}
-
-static int ATEN2011_calc_num_ports(struct usb_serial *serial)
-{
-
-	__u16 Data = 0x00;
-	int ret = 0;
-	int ATEN2011_2or4ports;
-	ret = usb_control_msg(serial->dev, usb_rcvctrlpipe(serial->dev, 0),
-			      ATEN_RDREQ, ATEN_RD_RTYPE, 0, GPIO_REGISTER,
-			      &Data, VENDOR_READ_LENGTH, ATEN_WDR_TIMEOUT);
-
-/* ghostgum: here is where the problem appears to bet */
-/* Which of the following are needed? */
-/* Greg used the serial->type->num_ports=2 */
-/* But the code in the ATEN2011_open relies on serial->num_ports=2 */
-	if ((Data & 0x01) == 0) {
-		ATEN2011_2or4ports = 2;
-		serial->type->num_ports = 2;
-		serial->num_ports = 2;
-	}
-	/* else if(serial->interface->cur_altsetting->desc.bNumEndpoints == 9) */
-	else {
-		ATEN2011_2or4ports = 4;
-		serial->type->num_ports = 4;
-		serial->num_ports = 4;
-
-	}
-
-	return ATEN2011_2or4ports;
-}
-
-static int ATEN2011_startup(struct usb_serial *serial)
-{
-	struct ATENINTL_serial *ATEN2011_serial;
-	struct ATENINTL_port *ATEN2011_port;
-	struct usb_device *dev;
-	int i, status;
-	int minor;
-
-	__u16 Data;
-	dbg("%s", " ATEN2011_startup :entering..........");
-
-	if (!serial) {
-		dbg("%s", "Invalid Handler");
-		return -1;
-	}
-
-	dev = serial->dev;
-
-	dbg("%s", "Entering...");
-
-	/* create our private serial structure */
-	ATEN2011_serial = kzalloc(sizeof(struct ATENINTL_serial), GFP_KERNEL);
-	if (ATEN2011_serial == NULL) {
-		err("%s - Out of memory", __func__);
-		return -ENOMEM;
-	}
-
-	/* resetting the private structure field values to zero */
-	memset(ATEN2011_serial, 0, sizeof(struct ATENINTL_serial));
-
-	ATEN2011_serial->serial = serial;
-	/* initilize status polling flag to 0 */
-	ATEN2011_serial->status_polling_started = 0;
-
-	usb_set_serial_data(serial, ATEN2011_serial);
-	ATEN2011_serial->ATEN2011_spectrum_2or4ports =
-	    ATEN2011_calc_num_ports(serial);
-	/* we set up the pointers to the endpoints in the ATEN2011_open *
-	 * function, as the structures aren't created yet.             */
-
-	/* set up port private structures */
-	for (i = 0; i < serial->num_ports; ++i) {
-		ATEN2011_port =
-		    kmalloc(sizeof(struct ATENINTL_port), GFP_KERNEL);
-		if (ATEN2011_port == NULL) {
-			err("%s - Out of memory", __func__);
-			usb_set_serial_data(serial, NULL);
-			kfree(ATEN2011_serial);
-			return -ENOMEM;
-		}
-		memset(ATEN2011_port, 0, sizeof(struct ATENINTL_port));
-
-		/*
-		 * Initialize all port interrupt end point to port 0
-		 * int endpoint. Our device has only one interrupt end point
-		 * comman to all port
-		 */
-		/* serial->port[i]->interrupt_in_endpointAddress = serial->port[0]->interrupt_in_endpointAddress; */
-
-		ATEN2011_port->port = serial->port[i];
-		usb_set_serial_port_data(serial->port[i], ATEN2011_port);
-
-		minor = serial->port[i]->serial->minor;
-		if (minor == SERIAL_TTY_NO_MINOR)
-			minor = 0;
-		ATEN2011_port->port_num =
-		    ((serial->port[i]->number - minor) + 1);
-
-		if (ATEN2011_port->port_num == 1) {
-			ATEN2011_port->SpRegOffset = 0x0;
-			ATEN2011_port->ControlRegOffset = 0x1;
-			ATEN2011_port->DcrRegOffset = 0x4;
-		} else if ((ATEN2011_port->port_num == 2)
-			   && (ATEN2011_serial->ATEN2011_spectrum_2or4ports ==
-			       4)) {
-			ATEN2011_port->SpRegOffset = 0x8;
-			ATEN2011_port->ControlRegOffset = 0x9;
-			ATEN2011_port->DcrRegOffset = 0x16;
-		} else if ((ATEN2011_port->port_num == 2)
-			   && (ATEN2011_serial->ATEN2011_spectrum_2or4ports ==
-			       2)) {
-			ATEN2011_port->SpRegOffset = 0xa;
-			ATEN2011_port->ControlRegOffset = 0xb;
-			ATEN2011_port->DcrRegOffset = 0x19;
-		} else if ((ATEN2011_port->port_num == 3)
-			   && (ATEN2011_serial->ATEN2011_spectrum_2or4ports ==
-			       4)) {
-			ATEN2011_port->SpRegOffset = 0xa;
-			ATEN2011_port->ControlRegOffset = 0xb;
-			ATEN2011_port->DcrRegOffset = 0x19;
-		} else if ((ATEN2011_port->port_num == 4)
-			   && (ATEN2011_serial->ATEN2011_spectrum_2or4ports ==
-			       4)) {
-			ATEN2011_port->SpRegOffset = 0xc;
-			ATEN2011_port->ControlRegOffset = 0xd;
-			ATEN2011_port->DcrRegOffset = 0x1c;
-		}
-
-		usb_set_serial_port_data(serial->port[i], ATEN2011_port);
-
-		/* enable rx_disable bit in control register */
-
-		status = get_reg_sync(serial->port[i],
-				      ATEN2011_port->ControlRegOffset, &Data);
-		if (status < 0) {
-			dbg("Reading ControlReg failed status-0x%x",
-				status);
-			break;
-		} else
-			dbg
-			    ("ControlReg Reading success val is %x, status%d",
-			     Data, status);
-		Data |= 0x08;	/* setting driver done bit */
-		Data |= 0x04;	/* sp1_bit to have cts change reflect in modem status reg */
-
-		/* Data |= 0x20; */	/* rx_disable bit */
-		status = set_reg_sync(serial->port[i],
-				      ATEN2011_port->ControlRegOffset, Data);
-		if (status < 0) {
-			dbg
-			    ("Writing ControlReg failed(rx_disable) status-0x%x",
-			     status);
-			break;
-		} else
-			dbg
-			    ("ControlReg Writing success(rx_disable) status%d",
-			     status);
-
-		/*
-		 * Write default values in DCR (i.e 0x01 in DCR0, 0x05 in DCR2
-		 * and 0x24 in DCR3
-		 */
-		Data = 0x01;
-		status = set_reg_sync(serial->port[i],
-				      (__u16)(ATEN2011_port->DcrRegOffset + 0),
-				      Data);
-		if (status < 0) {
-			dbg("Writing DCR0 failed status-0x%x", status);
-			break;
-		} else
-			dbg("DCR0 Writing success status%d", status);
-
-		Data = 0x05;
-		status = set_reg_sync(serial->port[i],
-				      (__u16)(ATEN2011_port->DcrRegOffset + 1),
-				      Data);
-		if (status < 0) {
-			dbg("Writing DCR1 failed status-0x%x", status);
-			break;
-		} else
-			dbg("DCR1 Writing success status%d", status);
-
-		Data = 0x24;
-		status = set_reg_sync(serial->port[i],
-				      (__u16)(ATEN2011_port->DcrRegOffset + 2),
-				      Data);
-		if (status < 0) {
-			dbg("Writing DCR2 failed status-0x%x", status);
-			break;
-		} else
-			dbg("DCR2 Writing success status%d", status);
-
-		/* write values in clkstart0x0 and clkmulti 0x20 */
-		Data = 0x0;
-		status = set_reg_sync(serial->port[i], CLK_START_VALUE_REGISTER,
-				      Data);
-		if (status < 0) {
-			dbg
-			    ("Writing CLK_START_VALUE_REGISTER failed status-0x%x",
-			     status);
-			break;
-		} else
-			dbg
-			    ("CLK_START_VALUE_REGISTER Writing success status%d",
-			     status);
-
-		Data = 0x20;
-		status = set_reg_sync(serial->port[i], CLK_MULTI_REGISTER,
-				      Data);
-		if (status < 0) {
-			dbg
-			    ("Writing CLK_MULTI_REGISTER failed status-0x%x",
-			     status);
-			break;
-		} else
-			dbg("CLK_MULTI_REGISTER Writing success status%d",
-				status);
-
-		/* Zero Length flag register */
-		if ((ATEN2011_port->port_num != 1)
-		    && (ATEN2011_serial->ATEN2011_spectrum_2or4ports == 2)) {
-
-			Data = 0xff;
-			status = set_reg_sync(serial->port[i],
-					      (__u16)(ZLP_REG1 + ((__u16)ATEN2011_port->port_num)),
-					      Data);
-			dbg("ZLIP offset%x",
-				(__u16) (ZLP_REG1 +
-					 ((__u16) ATEN2011_port->port_num)));
-			if (status < 0) {
-				dbg
-				    ("Writing ZLP_REG%d failed status-0x%x",
-				     i + 2, status);
-				break;
-			} else
-				dbg("ZLP_REG%d Writing success status%d",
-					i + 2, status);
-		} else {
-			Data = 0xff;
-			status = set_reg_sync(serial->port[i],
-					      (__u16)(ZLP_REG1 + ((__u16)ATEN2011_port->port_num) - 0x1),
-					      Data);
-			dbg("ZLIP offset%x",
-				(__u16) (ZLP_REG1 +
-					 ((__u16) ATEN2011_port->port_num) -
-					 0x1));
-			if (status < 0) {
-				dbg
-				    ("Writing ZLP_REG%d failed status-0x%x",
-				     i + 1, status);
-				break;
-			} else
-				dbg("ZLP_REG%d Writing success status%d",
-					i + 1, status);
-
-		}
-		ATEN2011_port->control_urb = usb_alloc_urb(0, GFP_ATOMIC);
-		ATEN2011_port->ctrl_buf = kmalloc(16, GFP_KERNEL);
-
-	}
-
-	/* Zero Length flag enable */
-	Data = 0x0f;
-	status = set_reg_sync(serial->port[0], ZLP_REG5, Data);
-	if (status < 0) {
-		dbg("Writing ZLP_REG5 failed status-0x%x", status);
-		return -1;
-	} else
-		dbg("ZLP_REG5 Writing success status%d", status);
-
-	/* setting configuration feature to one */
-	usb_control_msg(serial->dev, usb_sndctrlpipe(serial->dev, 0),
-			(__u8) 0x03, 0x00, 0x01, 0x00, NULL, 0x00, 5 * HZ);
-	return 0;
-}
-
-static void ATEN2011_release(struct usb_serial *serial)
-{
-	int i;
-	struct ATENINTL_port *ATEN2011_port;
-
-	/* check for the ports to be closed,close the ports and disconnect */
-
-	/* free private structure allocated for serial port  *
-	 * stop reads and writes on all ports                */
-
-	for (i = 0; i < serial->num_ports; ++i) {
-		ATEN2011_port = usb_get_serial_port_data(serial->port[i]);
-		kfree(ATEN2011_port->ctrl_buf);
-		usb_kill_urb(ATEN2011_port->control_urb);
-		kfree(ATEN2011_port);
-		usb_set_serial_port_data(serial->port[i], NULL);
-	}
-
-	/* free private structure allocated for serial device */
-
-	kfree(usb_get_serial_data(serial));
-	usb_set_serial_data(serial, NULL);
-}
-
-static struct usb_serial_driver aten_serial_driver = {
-	.driver = {
-		.owner =	THIS_MODULE,
-		.name =		"aten2011",
-		},
-	.description =		DRIVER_DESC,
-	.id_table =		id_table,
-	.open =			ATEN2011_open,
-	.close =		ATEN2011_close,
-	.write =		ATEN2011_write,
-	.write_room =		ATEN2011_write_room,
-	.chars_in_buffer =	ATEN2011_chars_in_buffer,
-	.throttle =		ATEN2011_throttle,
-	.unthrottle =		ATEN2011_unthrottle,
-	.calc_num_ports =	ATEN2011_calc_num_ports,
-
-	.ioctl =		ATEN2011_ioctl,
-	.set_termios =		ATEN2011_set_termios,
-	.break_ctl =		ATEN2011_break,
-	.tiocmget =		ATEN2011_tiocmget,
-	.tiocmset =		ATEN2011_tiocmset,
-	.attach =		ATEN2011_startup,
-	.release =		ATEN2011_release,
-	.read_bulk_callback =	ATEN2011_bulk_in_callback,
-	.read_int_callback =	ATEN2011_interrupt_callback,
-};
-
-static struct usb_driver aten_driver = {
-	.name =		"aten2011",
-	.probe =	usb_serial_probe,
-	.disconnect =	usb_serial_disconnect,
-	.id_table =	id_table,
-};
-
-static int __init aten_init(void)
-{
-	int retval;
-
-	/* Register with the usb serial */
-	retval = usb_serial_register(&aten_serial_driver);
-	if (retval)
-		return retval;
-
-	printk(KERN_INFO KBUILD_MODNAME ":"
-	       DRIVER_DESC " " DRIVER_VERSION "\n");
-
-	/* Register with the usb */
-	retval = usb_register(&aten_driver);
-	if (retval)
-		usb_serial_deregister(&aten_serial_driver);
-
-	return retval;
-}
-
-static void __exit aten_exit(void)
-{
-	usb_deregister(&aten_driver);
-	usb_serial_deregister(&aten_serial_driver);
-}
-
-module_init(aten_init);
-module_exit(aten_exit);
-
-/* Module information */
-MODULE_DESCRIPTION(DRIVER_DESC);
-MODULE_LICENSE("GPL");
-
-MODULE_PARM_DESC(debug, "Debug enabled or not");
diff --git a/drivers/staging/udlfb/udlfb.c b/drivers/staging/udlfb/udlfb.c
index 0ab9d15f343..f5416af1e90 100644
--- a/drivers/staging/udlfb/udlfb.c
+++ b/drivers/staging/udlfb/udlfb.c
@@ -21,6 +21,7 @@
 #include <linux/mm.h>
 #include <linux/fb.h>
 #include <linux/mutex.h>
+#include <linux/vmalloc.h>
 
 #include "udlfb.h"
 
diff --git a/drivers/usb/class/cdc-acm.c b/drivers/usb/class/cdc-acm.c
index e1f89416ef8..2bfc41ece0e 100644
--- a/drivers/usb/class/cdc-acm.c
+++ b/drivers/usb/class/cdc-acm.c
@@ -387,7 +387,6 @@ static void acm_rx_tasklet(unsigned long _acm)
 	struct acm_ru *rcv;
 	unsigned long flags;
 	unsigned char throttled;
-	struct usb_host_endpoint *ep;
 
 	dbg("Entering acm_rx_tasklet");
 
@@ -463,14 +462,12 @@ urbs:
 
 		rcv->buffer = buf;
 
-		ep = (usb_pipein(acm->rx_endpoint) ? acm->dev->ep_in : acm->dev->ep_out)
-				[usb_pipeendpoint(acm->rx_endpoint)];
-		if (usb_endpoint_xfer_int(&ep->desc))
+		if (acm->is_int_ep)
 			usb_fill_int_urb(rcv->urb, acm->dev,
 					 acm->rx_endpoint,
 					 buf->base,
 					 acm->readsize,
-					 acm_read_bulk, rcv, ep->desc.bInterval);
+					 acm_read_bulk, rcv, acm->bInterval);
 		else
 			usb_fill_bulk_urb(rcv->urb, acm->dev,
 					  acm->rx_endpoint,
@@ -1183,6 +1180,9 @@ made_compressed_probe:
 	spin_lock_init(&acm->read_lock);
 	mutex_init(&acm->mutex);
 	acm->rx_endpoint = usb_rcvbulkpipe(usb_dev, epread->bEndpointAddress);
+	acm->is_int_ep = usb_endpoint_xfer_int(epread);
+	if (acm->is_int_ep)
+		acm->bInterval = epread->bInterval;
 	tty_port_init(&acm->port);
 	acm->port.ops = &acm_port_ops;
 
diff --git a/drivers/usb/class/cdc-acm.h b/drivers/usb/class/cdc-acm.h
index 1602324808b..c4a0ee8ffcc 100644
--- a/drivers/usb/class/cdc-acm.h
+++ b/drivers/usb/class/cdc-acm.h
@@ -126,6 +126,8 @@ struct acm {
 	unsigned int ctrl_caps;				/* control capabilities from the class specific header */
 	unsigned int susp_count;			/* number of suspended interfaces */
 	int combined_interfaces:1;			/* control and data collapsed */
+	int is_int_ep:1;				/* interrupt endpoints contrary to spec used */
+	u8 bInterval;
 	struct acm_wb *delayed_wb;			/* write queued for a device about to be woken */
 };
 
diff --git a/drivers/usb/core/config.c b/drivers/usb/core/config.c
index 24dfb33f90c..a16c538d013 100644
--- a/drivers/usb/core/config.c
+++ b/drivers/usb/core/config.c
@@ -80,38 +80,18 @@ static int usb_parse_ss_endpoint_companion(struct device *ddev, int cfgno,
 	int max_tx;
 	int i;
 
-	/* Allocate space for the SS endpoint companion descriptor */
-	ep->ss_ep_comp = kzalloc(sizeof(struct usb_host_ss_ep_comp),
-			GFP_KERNEL);
-	if (!ep->ss_ep_comp)
-		return -ENOMEM;
 	desc = (struct usb_ss_ep_comp_descriptor *) buffer;
 	if (desc->bDescriptorType != USB_DT_SS_ENDPOINT_COMP) {
 		dev_warn(ddev, "No SuperSpeed endpoint companion for config %d "
 				" interface %d altsetting %d ep %d: "
 				"using minimum values\n",
 				cfgno, inum, asnum, ep->desc.bEndpointAddress);
-		ep->ss_ep_comp->desc.bLength = USB_DT_SS_EP_COMP_SIZE;
-		ep->ss_ep_comp->desc.bDescriptorType = USB_DT_SS_ENDPOINT_COMP;
-		ep->ss_ep_comp->desc.bMaxBurst = 0;
-		/*
-		 * Leave bmAttributes as zero, which will mean no streams for
-		 * bulk, and isoc won't support multiple bursts of packets.
-		 * With bursts of only one packet, and a Mult of 1, the max
-		 * amount of data moved per endpoint service interval is one
-		 * packet.
-		 */
-		if (usb_endpoint_xfer_isoc(&ep->desc) ||
-				usb_endpoint_xfer_int(&ep->desc))
-			ep->ss_ep_comp->desc.wBytesPerInterval =
-				ep->desc.wMaxPacketSize;
 		/*
 		 * The next descriptor is for an Endpoint or Interface,
 		 * no extra descriptors to copy into the companion structure,
 		 * and we didn't eat up any of the buffer.
 		 */
-		retval = 0;
-		goto valid;
+		return 0;
 	}
 	memcpy(&ep->ss_ep_comp->desc, desc, USB_DT_SS_EP_COMP_SIZE);
 	desc = &ep->ss_ep_comp->desc;
@@ -320,6 +300,28 @@ static int usb_parse_endpoint(struct device *ddev, int cfgno, int inum,
 		buffer += i;
 		size -= i;
 
+		/* Allocate space for the SS endpoint companion descriptor */
+		endpoint->ss_ep_comp = kzalloc(sizeof(struct usb_host_ss_ep_comp),
+				GFP_KERNEL);
+		if (!endpoint->ss_ep_comp)
+			return -ENOMEM;
+
+		/* Fill in some default values (may be overwritten later) */
+		endpoint->ss_ep_comp->desc.bLength = USB_DT_SS_EP_COMP_SIZE;
+		endpoint->ss_ep_comp->desc.bDescriptorType = USB_DT_SS_ENDPOINT_COMP;
+		endpoint->ss_ep_comp->desc.bMaxBurst = 0;
+		/*
+		 * Leave bmAttributes as zero, which will mean no streams for
+		 * bulk, and isoc won't support multiple bursts of packets.
+		 * With bursts of only one packet, and a Mult of 1, the max
+		 * amount of data moved per endpoint service interval is one
+		 * packet.
+		 */
+		if (usb_endpoint_xfer_isoc(&endpoint->desc) ||
+				usb_endpoint_xfer_int(&endpoint->desc))
+			endpoint->ss_ep_comp->desc.wBytesPerInterval =
+				endpoint->desc.wMaxPacketSize;
+
 		if (size > 0) {
 			retval = usb_parse_ss_endpoint_companion(ddev, cfgno,
 					inum, asnum, endpoint, num_ep, buffer,
@@ -329,6 +331,10 @@ static int usb_parse_endpoint(struct device *ddev, int cfgno, int inum,
 				retval = buffer - buffer0;
 			}
 		} else {
+			dev_warn(ddev, "config %d interface %d altsetting %d "
+				"endpoint 0x%X has no "
+				"SuperSpeed companion descriptor\n",
+				cfgno, inum, asnum, d->bEndpointAddress);
 			retval = buffer - buffer0;
 		}
 	} else {
diff --git a/drivers/usb/core/devio.c b/drivers/usb/core/devio.c
index 38b8bce782d..4247eccf858 100644
--- a/drivers/usb/core/devio.c
+++ b/drivers/usb/core/devio.c
@@ -595,7 +595,7 @@ static int usbdev_open(struct inode *inode, struct file *file)
 	if (!ps)
 		goto out;
 
-	ret = -ENOENT;
+	ret = -ENODEV;
 
 	/* usbdev device-node */
 	if (imajor(inode) == USB_DEVICE_MAJOR)
@@ -1321,7 +1321,8 @@ static int get_urb32(struct usbdevfs_urb *kurb,
 		     struct usbdevfs_urb32 __user *uurb)
 {
 	__u32  uptr;
-	if (get_user(kurb->type, &uurb->type) ||
+	if (!access_ok(VERIFY_READ, uurb, sizeof(*uurb)) ||
+	    __get_user(kurb->type, &uurb->type) ||
 	    __get_user(kurb->endpoint, &uurb->endpoint) ||
 	    __get_user(kurb->status, &uurb->status) ||
 	    __get_user(kurb->flags, &uurb->flags) ||
@@ -1536,8 +1537,9 @@ static int proc_ioctl_compat(struct dev_state *ps, compat_uptr_t arg)
 	u32 udata;
 
 	uioc = compat_ptr((long)arg);
-	if (get_user(ctrl.ifno, &uioc->ifno) ||
-	    get_user(ctrl.ioctl_code, &uioc->ioctl_code) ||
+	if (!access_ok(VERIFY_READ, uioc, sizeof(*uioc)) ||
+	    __get_user(ctrl.ifno, &uioc->ifno) ||
+	    __get_user(ctrl.ioctl_code, &uioc->ioctl_code) ||
 	    __get_user(udata, &uioc->data))
 		return -EFAULT;
 	ctrl.data = compat_ptr(udata);
diff --git a/drivers/usb/host/ehci-hcd.c b/drivers/usb/host/ehci-hcd.c
index 7d03549c333..11c627ce602 100644
--- a/drivers/usb/host/ehci-hcd.c
+++ b/drivers/usb/host/ehci-hcd.c
@@ -903,7 +903,8 @@ static int ehci_urb_dequeue(struct usb_hcd *hcd, struct urb *urb, int status)
 			/* already started */
 			break;
 		case QH_STATE_IDLE:
-			WARN_ON(1);
+			/* QH might be waiting for a Clear-TT-Buffer */
+			qh_completions(ehci, qh);
 			break;
 		}
 		break;
diff --git a/drivers/usb/host/ehci-orion.c b/drivers/usb/host/ehci-orion.c
index dc2ac613a9d..1d283e1b2b8 100644
--- a/drivers/usb/host/ehci-orion.c
+++ b/drivers/usb/host/ehci-orion.c
@@ -105,6 +105,7 @@ static int ehci_orion_setup(struct usb_hcd *hcd)
 	struct ehci_hcd *ehci = hcd_to_ehci(hcd);
 	int retval;
 
+	ehci_reset(ehci);
 	retval = ehci_halt(ehci);
 	if (retval)
 		return retval;
@@ -118,7 +119,6 @@ static int ehci_orion_setup(struct usb_hcd *hcd)
 
 	hcd->has_tt = 1;
 
-	ehci_reset(ehci);
 	ehci_port_power(ehci, 0);
 
 	return retval;
diff --git a/drivers/usb/host/ehci-q.c b/drivers/usb/host/ehci-q.c
index 9a1384747f3..7673554fa64 100644
--- a/drivers/usb/host/ehci-q.c
+++ b/drivers/usb/host/ehci-q.c
@@ -375,12 +375,11 @@ qh_completions (struct ehci_hcd *ehci, struct ehci_qh *qh)
 				 */
 				if ((token & QTD_STS_XACT) &&
 						QTD_CERR(token) == 0 &&
-						--qh->xacterrs > 0 &&
+						++qh->xacterrs < QH_XACTERR_MAX &&
 						!urb->unlinked) {
 					ehci_dbg(ehci,
 	"detected XactErr len %zu/%zu retry %d\n",
-	qtd->length - QTD_LENGTH(token), qtd->length,
-	QH_XACTERR_MAX - qh->xacterrs);
+	qtd->length - QTD_LENGTH(token), qtd->length, qh->xacterrs);
 
 					/* reset the token in the qtd and the
 					 * qh overlay (which still contains
@@ -494,7 +493,7 @@ halt:
 		last = qtd;
 
 		/* reinit the xacterr counter for the next qtd */
-		qh->xacterrs = QH_XACTERR_MAX;
+		qh->xacterrs = 0;
 	}
 
 	/* last urb's completion might still need calling */
@@ -940,7 +939,8 @@ static void qh_link_async (struct ehci_hcd *ehci, struct ehci_qh *qh)
 	head->qh_next.qh = qh;
 	head->hw_next = dma;
 
-	qh->xacterrs = QH_XACTERR_MAX;
+	qh_get(qh);
+	qh->xacterrs = 0;
 	qh->qh_state = QH_STATE_LINKED;
 	/* qtd completions reported later by interrupt */
 }
@@ -1080,7 +1080,7 @@ submit_async (
 	 * the HC and TT handle it when the TT has a buffer ready.
 	 */
 	if (likely (qh->qh_state == QH_STATE_IDLE))
-		qh_link_async (ehci, qh_get (qh));
+		qh_link_async(ehci, qh);
  done:
 	spin_unlock_irqrestore (&ehci->lock, flags);
 	if (unlikely (qh == NULL))
@@ -1115,8 +1115,6 @@ static void end_unlink_async (struct ehci_hcd *ehci)
 			&& HC_IS_RUNNING (ehci_to_hcd(ehci)->state))
 		qh_link_async (ehci, qh);
 	else {
-		qh_put (qh);		// refcount from async list
-
 		/* it's not free to turn the async schedule on/off; leave it
 		 * active but idle for a while once it empties.
 		 */
@@ -1124,6 +1122,7 @@ static void end_unlink_async (struct ehci_hcd *ehci)
 				&& ehci->async->qh_next.qh == NULL)
 			timer_action (ehci, TIMER_ASYNC_OFF);
 	}
+	qh_put(qh);			/* refcount from async list */
 
 	if (next) {
 		ehci->reclaim = NULL;
diff --git a/drivers/usb/host/ehci-sched.c b/drivers/usb/host/ehci-sched.c
index 74f7f83b29a..edd61ee9032 100644
--- a/drivers/usb/host/ehci-sched.c
+++ b/drivers/usb/host/ehci-sched.c
@@ -542,6 +542,7 @@ static int qh_link_periodic (struct ehci_hcd *ehci, struct ehci_qh *qh)
 		}
 	}
 	qh->qh_state = QH_STATE_LINKED;
+	qh->xacterrs = 0;
 	qh_get (qh);
 
 	/* update per-qh bandwidth for usbfs */
diff --git a/drivers/usb/host/ohci-omap.c b/drivers/usb/host/ohci-omap.c
index f3aaba35e91..83cbecd2a1e 100644
--- a/drivers/usb/host/ohci-omap.c
+++ b/drivers/usb/host/ohci-omap.c
@@ -282,6 +282,7 @@ static int ohci_omap_init(struct usb_hcd *hcd)
 static void ohci_omap_stop(struct usb_hcd *hcd)
 {
 	dev_dbg(hcd->self.controller, "stopping USB Controller\n");
+	ohci_stop(hcd);
 	omap_ohci_clock_power(0);
 }
 
diff --git a/drivers/usb/host/xhci-dbg.c b/drivers/usb/host/xhci-dbg.c
index 2501c571f85..705e3432415 100644
--- a/drivers/usb/host/xhci-dbg.c
+++ b/drivers/usb/host/xhci-dbg.c
@@ -173,6 +173,7 @@ void xhci_print_ir_set(struct xhci_hcd *xhci, struct xhci_intr_reg *ir_set, int
 {
 	void *addr;
 	u32 temp;
+	u64 temp_64;
 
 	addr = &ir_set->irq_pending;
 	temp = xhci_readl(xhci, addr);
@@ -200,25 +201,15 @@ void xhci_print_ir_set(struct xhci_hcd *xhci, struct xhci_intr_reg *ir_set, int
 		xhci_dbg(xhci, "  WARN: %p: ir_set.rsvd = 0x%x\n",
 				addr, (unsigned int)temp);
 
-	addr = &ir_set->erst_base[0];
-	temp = xhci_readl(xhci, addr);
-	xhci_dbg(xhci, "  %p: ir_set.erst_base[0] = 0x%x\n",
-			addr, (unsigned int) temp);
-
-	addr = &ir_set->erst_base[1];
-	temp = xhci_readl(xhci, addr);
-	xhci_dbg(xhci, "  %p: ir_set.erst_base[1] = 0x%x\n",
-			addr, (unsigned int) temp);
+	addr = &ir_set->erst_base;
+	temp_64 = xhci_read_64(xhci, addr);
+	xhci_dbg(xhci, "  %p: ir_set.erst_base = @%08llx\n",
+			addr, temp_64);
 
-	addr = &ir_set->erst_dequeue[0];
-	temp = xhci_readl(xhci, addr);
-	xhci_dbg(xhci, "  %p: ir_set.erst_dequeue[0] = 0x%x\n",
-			addr, (unsigned int) temp);
-
-	addr = &ir_set->erst_dequeue[1];
-	temp = xhci_readl(xhci, addr);
-	xhci_dbg(xhci, "  %p: ir_set.erst_dequeue[1] = 0x%x\n",
-			addr, (unsigned int) temp);
+	addr = &ir_set->erst_dequeue;
+	temp_64 = xhci_read_64(xhci, addr);
+	xhci_dbg(xhci, "  %p: ir_set.erst_dequeue = @%08llx\n",
+			addr, temp_64);
 }
 
 void xhci_print_run_regs(struct xhci_hcd *xhci)
@@ -268,8 +259,7 @@ void xhci_debug_trb(struct xhci_hcd *xhci, union xhci_trb *trb)
 		xhci_dbg(xhci, "Link TRB:\n");
 		xhci_print_trb_offsets(xhci, trb);
 
-		address = trb->link.segment_ptr[0] +
-			(((u64) trb->link.segment_ptr[1]) << 32);
+		address = trb->link.segment_ptr;
 		xhci_dbg(xhci, "Next ring segment DMA address = 0x%llx\n", address);
 
 		xhci_dbg(xhci, "Interrupter target = 0x%x\n",
@@ -282,8 +272,7 @@ void xhci_debug_trb(struct xhci_hcd *xhci, union xhci_trb *trb)
 				(unsigned int) (trb->link.control & TRB_NO_SNOOP));
 		break;
 	case TRB_TYPE(TRB_TRANSFER):
-		address = trb->trans_event.buffer[0] +
-			(((u64) trb->trans_event.buffer[1]) << 32);
+		address = trb->trans_event.buffer;
 		/*
 		 * FIXME: look at flags to figure out if it's an address or if
 		 * the data is directly in the buffer field.
@@ -291,8 +280,7 @@ void xhci_debug_trb(struct xhci_hcd *xhci, union xhci_trb *trb)
 		xhci_dbg(xhci, "DMA address or buffer contents= %llu\n", address);
 		break;
 	case TRB_TYPE(TRB_COMPLETION):
-		address = trb->event_cmd.cmd_trb[0] +
-			(((u64) trb->event_cmd.cmd_trb[1]) << 32);
+		address = trb->event_cmd.cmd_trb;
 		xhci_dbg(xhci, "Command TRB pointer = %llu\n", address);
 		xhci_dbg(xhci, "Completion status = %u\n",
 				(unsigned int) GET_COMP_CODE(trb->event_cmd.status));
@@ -328,8 +316,8 @@ void xhci_debug_segment(struct xhci_hcd *xhci, struct xhci_segment *seg)
 	for (i = 0; i < TRBS_PER_SEGMENT; ++i) {
 		trb = &seg->trbs[i];
 		xhci_dbg(xhci, "@%08x %08x %08x %08x %08x\n", addr,
-				(unsigned int) trb->link.segment_ptr[0],
-				(unsigned int) trb->link.segment_ptr[1],
+				lower_32_bits(trb->link.segment_ptr),
+				upper_32_bits(trb->link.segment_ptr),
 				(unsigned int) trb->link.intr_target,
 				(unsigned int) trb->link.control);
 		addr += sizeof(*trb);
@@ -386,8 +374,8 @@ void xhci_dbg_erst(struct xhci_hcd *xhci, struct xhci_erst *erst)
 		entry = &erst->entries[i];
 		xhci_dbg(xhci, "@%08x %08x %08x %08x %08x\n",
 				(unsigned int) addr,
-				(unsigned int) entry->seg_addr[0],
-				(unsigned int) entry->seg_addr[1],
+				lower_32_bits(entry->seg_addr),
+				upper_32_bits(entry->seg_addr),
 				(unsigned int) entry->seg_size,
 				(unsigned int) entry->rsvd);
 		addr += sizeof(*entry);
@@ -396,90 +384,147 @@ void xhci_dbg_erst(struct xhci_hcd *xhci, struct xhci_erst *erst)
 
 void xhci_dbg_cmd_ptrs(struct xhci_hcd *xhci)
 {
-	u32 val;
+	u64 val;
 
-	val = xhci_readl(xhci, &xhci->op_regs->cmd_ring[0]);
-	xhci_dbg(xhci, "// xHC command ring deq ptr low bits + flags = 0x%x\n", val);
-	val = xhci_readl(xhci, &xhci->op_regs->cmd_ring[1]);
-	xhci_dbg(xhci, "// xHC command ring deq ptr high bits = 0x%x\n", val);
+	val = xhci_read_64(xhci, &xhci->op_regs->cmd_ring);
+	xhci_dbg(xhci, "// xHC command ring deq ptr low bits + flags = @%08x\n",
+			lower_32_bits(val));
+	xhci_dbg(xhci, "// xHC command ring deq ptr high bits = @%08x\n",
+			upper_32_bits(val));
 }
 
-void xhci_dbg_ctx(struct xhci_hcd *xhci, struct xhci_device_control *ctx, dma_addr_t dma, unsigned int last_ep)
+/* Print the last 32 bytes for 64-byte contexts */
+static void dbg_rsvd64(struct xhci_hcd *xhci, u64 *ctx, dma_addr_t dma)
+{
+	int i;
+	for (i = 0; i < 4; ++i) {
+		xhci_dbg(xhci, "@%p (virt) @%08llx "
+			 "(dma) %#08llx - rsvd64[%d]\n",
+			 &ctx[4 + i], (unsigned long long)dma,
+			 ctx[4 + i], i);
+		dma += 8;
+	}
+}
+
+void xhci_dbg_slot_ctx(struct xhci_hcd *xhci, struct xhci_container_ctx *ctx)
 {
-	int i, j;
-	int last_ep_ctx = 31;
 	/* Fields are 32 bits wide, DMA addresses are in bytes */
 	int field_size = 32 / 8;
+	int i;
 
-	xhci_dbg(xhci, "@%p (virt) @%08llx (dma) %#08x - drop flags\n",
-			&ctx->drop_flags, (unsigned long long)dma,
-			ctx->drop_flags);
-	dma += field_size;
-	xhci_dbg(xhci, "@%p (virt) @%08llx (dma) %#08x - add flags\n",
-			&ctx->add_flags, (unsigned long long)dma,
-			ctx->add_flags);
-	dma += field_size;
-	for (i = 0; i > 6; ++i) {
-		xhci_dbg(xhci, "@%p (virt) @%08llx (dma) %#08x - rsvd[%d]\n",
-				&ctx->rsvd[i], (unsigned long long)dma,
-				ctx->rsvd[i], i);
-		dma += field_size;
-	}
+	struct xhci_slot_ctx *slot_ctx = xhci_get_slot_ctx(xhci, ctx);
+	dma_addr_t dma = ctx->dma + ((unsigned long)slot_ctx - (unsigned long)ctx);
+	int csz = HCC_64BYTE_CONTEXT(xhci->hcc_params);
 
 	xhci_dbg(xhci, "Slot Context:\n");
 	xhci_dbg(xhci, "@%p (virt) @%08llx (dma) %#08x - dev_info\n",
-			&ctx->slot.dev_info,
-			(unsigned long long)dma, ctx->slot.dev_info);
+			&slot_ctx->dev_info,
+			(unsigned long long)dma, slot_ctx->dev_info);
 	dma += field_size;
 	xhci_dbg(xhci, "@%p (virt) @%08llx (dma) %#08x - dev_info2\n",
-			&ctx->slot.dev_info2,
-			(unsigned long long)dma, ctx->slot.dev_info2);
+			&slot_ctx->dev_info2,
+			(unsigned long long)dma, slot_ctx->dev_info2);
 	dma += field_size;
 	xhci_dbg(xhci, "@%p (virt) @%08llx (dma) %#08x - tt_info\n",
-			&ctx->slot.tt_info,
-			(unsigned long long)dma, ctx->slot.tt_info);
+			&slot_ctx->tt_info,
+			(unsigned long long)dma, slot_ctx->tt_info);
 	dma += field_size;
 	xhci_dbg(xhci, "@%p (virt) @%08llx (dma) %#08x - dev_state\n",
-			&ctx->slot.dev_state,
-			(unsigned long long)dma, ctx->slot.dev_state);
+			&slot_ctx->dev_state,
+			(unsigned long long)dma, slot_ctx->dev_state);
 	dma += field_size;
-	for (i = 0; i > 4; ++i) {
+	for (i = 0; i < 4; ++i) {
 		xhci_dbg(xhci, "@%p (virt) @%08llx (dma) %#08x - rsvd[%d]\n",
-				&ctx->slot.reserved[i], (unsigned long long)dma,
-				ctx->slot.reserved[i], i);
+				&slot_ctx->reserved[i], (unsigned long long)dma,
+				slot_ctx->reserved[i], i);
 		dma += field_size;
 	}
 
+	if (csz)
+		dbg_rsvd64(xhci, (u64 *)slot_ctx, dma);
+}
+
+void xhci_dbg_ep_ctx(struct xhci_hcd *xhci,
+		     struct xhci_container_ctx *ctx,
+		     unsigned int last_ep)
+{
+	int i, j;
+	int last_ep_ctx = 31;
+	/* Fields are 32 bits wide, DMA addresses are in bytes */
+	int field_size = 32 / 8;
+	int csz = HCC_64BYTE_CONTEXT(xhci->hcc_params);
+
 	if (last_ep < 31)
 		last_ep_ctx = last_ep + 1;
 	for (i = 0; i < last_ep_ctx; ++i) {
+		struct xhci_ep_ctx *ep_ctx = xhci_get_ep_ctx(xhci, ctx, i);
+		dma_addr_t dma = ctx->dma +
+			((unsigned long)ep_ctx - (unsigned long)ctx);
+
 		xhci_dbg(xhci, "Endpoint %02d Context:\n", i);
 		xhci_dbg(xhci, "@%p (virt) @%08llx (dma) %#08x - ep_info\n",
-				&ctx->ep[i].ep_info,
-				(unsigned long long)dma, ctx->ep[i].ep_info);
+				&ep_ctx->ep_info,
+				(unsigned long long)dma, ep_ctx->ep_info);
 		dma += field_size;
 		xhci_dbg(xhci, "@%p (virt) @%08llx (dma) %#08x - ep_info2\n",
-				&ctx->ep[i].ep_info2,
-				(unsigned long long)dma, ctx->ep[i].ep_info2);
-		dma += field_size;
-		xhci_dbg(xhci, "@%p (virt) @%08llx (dma) %#08x - deq[0]\n",
-				&ctx->ep[i].deq[0],
-				(unsigned long long)dma, ctx->ep[i].deq[0]);
-		dma += field_size;
-		xhci_dbg(xhci, "@%p (virt) @%08llx (dma) %#08x - deq[1]\n",
-				&ctx->ep[i].deq[1],
-				(unsigned long long)dma, ctx->ep[i].deq[1]);
+				&ep_ctx->ep_info2,
+				(unsigned long long)dma, ep_ctx->ep_info2);
 		dma += field_size;
+		xhci_dbg(xhci, "@%p (virt) @%08llx (dma) %#08llx - deq\n",
+				&ep_ctx->deq,
+				(unsigned long long)dma, ep_ctx->deq);
+		dma += 2*field_size;
 		xhci_dbg(xhci, "@%p (virt) @%08llx (dma) %#08x - tx_info\n",
-				&ctx->ep[i].tx_info,
-				(unsigned long long)dma, ctx->ep[i].tx_info);
+				&ep_ctx->tx_info,
+				(unsigned long long)dma, ep_ctx->tx_info);
 		dma += field_size;
 		for (j = 0; j < 3; ++j) {
 			xhci_dbg(xhci, "@%p (virt) @%08llx (dma) %#08x - rsvd[%d]\n",
-					&ctx->ep[i].reserved[j],
+					&ep_ctx->reserved[j],
 					(unsigned long long)dma,
-					ctx->ep[i].reserved[j], j);
+					ep_ctx->reserved[j], j);
+			dma += field_size;
+		}
+
+		if (csz)
+			dbg_rsvd64(xhci, (u64 *)ep_ctx, dma);
+	}
+}
+
+void xhci_dbg_ctx(struct xhci_hcd *xhci,
+		  struct xhci_container_ctx *ctx,
+		  unsigned int last_ep)
+{
+	int i;
+	/* Fields are 32 bits wide, DMA addresses are in bytes */
+	int field_size = 32 / 8;
+	struct xhci_slot_ctx *slot_ctx;
+	dma_addr_t dma = ctx->dma;
+	int csz = HCC_64BYTE_CONTEXT(xhci->hcc_params);
+
+	if (ctx->type == XHCI_CTX_TYPE_INPUT) {
+		struct xhci_input_control_ctx *ctrl_ctx =
+			xhci_get_input_control_ctx(xhci, ctx);
+		xhci_dbg(xhci, "@%p (virt) @%08llx (dma) %#08x - drop flags\n",
+			 &ctrl_ctx->drop_flags, (unsigned long long)dma,
+			 ctrl_ctx->drop_flags);
+		dma += field_size;
+		xhci_dbg(xhci, "@%p (virt) @%08llx (dma) %#08x - add flags\n",
+			 &ctrl_ctx->add_flags, (unsigned long long)dma,
+			 ctrl_ctx->add_flags);
+		dma += field_size;
+		for (i = 0; i < 6; ++i) {
+			xhci_dbg(xhci, "@%p (virt) @%08llx (dma) %#08x - rsvd2[%d]\n",
+				 &ctrl_ctx->rsvd2[i], (unsigned long long)dma,
+				 ctrl_ctx->rsvd2[i], i);
 			dma += field_size;
 		}
+
+		if (csz)
+			dbg_rsvd64(xhci, (u64 *)ctrl_ctx, dma);
 	}
+
+	slot_ctx = xhci_get_slot_ctx(xhci, ctx);
+	xhci_dbg_slot_ctx(xhci, ctx);
+	xhci_dbg_ep_ctx(xhci, ctx, last_ep);
 }
diff --git a/drivers/usb/host/xhci-hcd.c b/drivers/usb/host/xhci-hcd.c
index dba3e07ccd0..816c39caca1 100644
--- a/drivers/usb/host/xhci-hcd.c
+++ b/drivers/usb/host/xhci-hcd.c
@@ -103,7 +103,10 @@ int xhci_reset(struct xhci_hcd *xhci)
 	u32 state;
 
 	state = xhci_readl(xhci, &xhci->op_regs->status);
-	BUG_ON((state & STS_HALT) == 0);
+	if ((state & STS_HALT) == 0) {
+		xhci_warn(xhci, "Host controller not halted, aborting reset.\n");
+		return 0;
+	}
 
 	xhci_dbg(xhci, "// Reset the HC\n");
 	command = xhci_readl(xhci, &xhci->op_regs->command);
@@ -226,6 +229,7 @@ int xhci_init(struct usb_hcd *hcd)
 static void xhci_work(struct xhci_hcd *xhci)
 {
 	u32 temp;
+	u64 temp_64;
 
 	/*
 	 * Clear the op reg interrupt status first,
@@ -248,9 +252,9 @@ static void xhci_work(struct xhci_hcd *xhci)
 	/* FIXME this should be a delayed service routine that clears the EHB */
 	xhci_handle_event(xhci);
 
-	/* Clear the event handler busy flag; the event ring should be empty. */
-	temp = xhci_readl(xhci, &xhci->ir_set->erst_dequeue[0]);
-	xhci_writel(xhci, temp & ~ERST_EHB, &xhci->ir_set->erst_dequeue[0]);
+	/* Clear the event handler busy flag (RW1C); the event ring should be empty. */
+	temp_64 = xhci_read_64(xhci, &xhci->ir_set->erst_dequeue);
+	xhci_write_64(xhci, temp_64 | ERST_EHB, &xhci->ir_set->erst_dequeue);
 	/* Flush posted writes -- FIXME is this necessary? */
 	xhci_readl(xhci, &xhci->ir_set->irq_pending);
 }
@@ -266,19 +270,34 @@ irqreturn_t xhci_irq(struct usb_hcd *hcd)
 {
 	struct xhci_hcd *xhci = hcd_to_xhci(hcd);
 	u32 temp, temp2;
+	union xhci_trb *trb;
 
 	spin_lock(&xhci->lock);
+	trb = xhci->event_ring->dequeue;
 	/* Check if the xHC generated the interrupt, or the irq is shared */
 	temp = xhci_readl(xhci, &xhci->op_regs->status);
 	temp2 = xhci_readl(xhci, &xhci->ir_set->irq_pending);
+	if (temp == 0xffffffff && temp2 == 0xffffffff)
+		goto hw_died;
+
 	if (!(temp & STS_EINT) && !ER_IRQ_PENDING(temp2)) {
 		spin_unlock(&xhci->lock);
 		return IRQ_NONE;
 	}
+	xhci_dbg(xhci, "op reg status = %08x\n", temp);
+	xhci_dbg(xhci, "ir set irq_pending = %08x\n", temp2);
+	xhci_dbg(xhci, "Event ring dequeue ptr:\n");
+	xhci_dbg(xhci, "@%llx %08x %08x %08x %08x\n",
+			(unsigned long long)xhci_trb_virt_to_dma(xhci->event_ring->deq_seg, trb),
+			lower_32_bits(trb->link.segment_ptr),
+			upper_32_bits(trb->link.segment_ptr),
+			(unsigned int) trb->link.intr_target,
+			(unsigned int) trb->link.control);
 
 	if (temp & STS_FATAL) {
 		xhci_warn(xhci, "WARNING: Host System Error\n");
 		xhci_halt(xhci);
+hw_died:
 		xhci_to_hcd(xhci)->state = HC_STATE_HALT;
 		spin_unlock(&xhci->lock);
 		return -ESHUTDOWN;
@@ -295,6 +314,7 @@ void xhci_event_ring_work(unsigned long arg)
 {
 	unsigned long flags;
 	int temp;
+	u64 temp_64;
 	struct xhci_hcd *xhci = (struct xhci_hcd *) arg;
 	int i, j;
 
@@ -311,9 +331,9 @@ void xhci_event_ring_work(unsigned long arg)
 	xhci_dbg(xhci, "Event ring:\n");
 	xhci_debug_segment(xhci, xhci->event_ring->deq_seg);
 	xhci_dbg_ring_ptrs(xhci, xhci->event_ring);
-	temp = xhci_readl(xhci, &xhci->ir_set->erst_dequeue[0]);
-	temp &= ERST_PTR_MASK;
-	xhci_dbg(xhci, "ERST deq = 0x%x\n", temp);
+	temp_64 = xhci_read_64(xhci, &xhci->ir_set->erst_dequeue);
+	temp_64 &= ~ERST_PTR_MASK;
+	xhci_dbg(xhci, "ERST deq = 64'h%0lx\n", (long unsigned int) temp_64);
 	xhci_dbg(xhci, "Command ring:\n");
 	xhci_debug_segment(xhci, xhci->cmd_ring->deq_seg);
 	xhci_dbg_ring_ptrs(xhci, xhci->cmd_ring);
@@ -356,6 +376,7 @@ void xhci_event_ring_work(unsigned long arg)
 int xhci_run(struct usb_hcd *hcd)
 {
 	u32 temp;
+	u64 temp_64;
 	struct xhci_hcd *xhci = hcd_to_xhci(hcd);
 	void (*doorbell)(struct xhci_hcd *) = NULL;
 
@@ -382,6 +403,20 @@ int xhci_run(struct usb_hcd *hcd)
 	add_timer(&xhci->event_ring_timer);
 #endif
 
+	xhci_dbg(xhci, "Command ring memory map follows:\n");
+	xhci_debug_ring(xhci, xhci->cmd_ring);
+	xhci_dbg_ring_ptrs(xhci, xhci->cmd_ring);
+	xhci_dbg_cmd_ptrs(xhci);
+
+	xhci_dbg(xhci, "ERST memory map follows:\n");
+	xhci_dbg_erst(xhci, &xhci->erst);
+	xhci_dbg(xhci, "Event ring:\n");
+	xhci_debug_ring(xhci, xhci->event_ring);
+	xhci_dbg_ring_ptrs(xhci, xhci->event_ring);
+	temp_64 = xhci_read_64(xhci, &xhci->ir_set->erst_dequeue);
+	temp_64 &= ~ERST_PTR_MASK;
+	xhci_dbg(xhci, "ERST deq = 64'h%0lx\n", (long unsigned int) temp_64);
+
 	xhci_dbg(xhci, "// Set the interrupt modulation register\n");
 	temp = xhci_readl(xhci, &xhci->ir_set->irq_control);
 	temp &= ~ER_IRQ_INTERVAL_MASK;
@@ -406,22 +441,6 @@ int xhci_run(struct usb_hcd *hcd)
 	if (NUM_TEST_NOOPS > 0)
 		doorbell = xhci_setup_one_noop(xhci);
 
-	xhci_dbg(xhci, "Command ring memory map follows:\n");
-	xhci_debug_ring(xhci, xhci->cmd_ring);
-	xhci_dbg_ring_ptrs(xhci, xhci->cmd_ring);
-	xhci_dbg_cmd_ptrs(xhci);
-
-	xhci_dbg(xhci, "ERST memory map follows:\n");
-	xhci_dbg_erst(xhci, &xhci->erst);
-	xhci_dbg(xhci, "Event ring:\n");
-	xhci_debug_ring(xhci, xhci->event_ring);
-	xhci_dbg_ring_ptrs(xhci, xhci->event_ring);
-	temp = xhci_readl(xhci, &xhci->ir_set->erst_dequeue[0]);
-	temp &= ERST_PTR_MASK;
-	xhci_dbg(xhci, "ERST deq = 0x%x\n", temp);
-	temp = xhci_readl(xhci, &xhci->ir_set->erst_dequeue[1]);
-	xhci_dbg(xhci, "ERST deq upper = 0x%x\n", temp);
-
 	temp = xhci_readl(xhci, &xhci->op_regs->command);
 	temp |= (CMD_RUN);
 	xhci_dbg(xhci, "// Turn on HC, cmd = 0x%x.\n",
@@ -601,10 +620,13 @@ int xhci_urb_enqueue(struct usb_hcd *hcd, struct urb *urb, gfp_t mem_flags)
 		goto exit;
 	}
 	if (usb_endpoint_xfer_control(&urb->ep->desc))
-		ret = xhci_queue_ctrl_tx(xhci, mem_flags, urb,
+		/* We have a spinlock and interrupts disabled, so we must pass
+		 * atomic context to this function, which may allocate memory.
+		 */
+		ret = xhci_queue_ctrl_tx(xhci, GFP_ATOMIC, urb,
 				slot_id, ep_index);
 	else if (usb_endpoint_xfer_bulk(&urb->ep->desc))
-		ret = xhci_queue_bulk_tx(xhci, mem_flags, urb,
+		ret = xhci_queue_bulk_tx(xhci, GFP_ATOMIC, urb,
 				slot_id, ep_index);
 	else
 		ret = -EINVAL;
@@ -661,8 +683,12 @@ int xhci_urb_dequeue(struct usb_hcd *hcd, struct urb *urb, int status)
 		goto done;
 
 	xhci_dbg(xhci, "Cancel URB %p\n", urb);
+	xhci_dbg(xhci, "Event ring:\n");
+	xhci_debug_ring(xhci, xhci->event_ring);
 	ep_index = xhci_get_endpoint_index(&urb->ep->desc);
 	ep_ring = xhci->devs[urb->dev->slot_id]->ep_rings[ep_index];
+	xhci_dbg(xhci, "Endpoint ring:\n");
+	xhci_debug_ring(xhci, ep_ring);
 	td = (struct xhci_td *) urb->hcpriv;
 
 	ep_ring->cancels_pending++;
@@ -696,7 +722,9 @@ int xhci_drop_endpoint(struct usb_hcd *hcd, struct usb_device *udev,
 		struct usb_host_endpoint *ep)
 {
 	struct xhci_hcd *xhci;
-	struct xhci_device_control *in_ctx;
+	struct xhci_container_ctx *in_ctx, *out_ctx;
+	struct xhci_input_control_ctx *ctrl_ctx;
+	struct xhci_slot_ctx *slot_ctx;
 	unsigned int last_ctx;
 	unsigned int ep_index;
 	struct xhci_ep_ctx *ep_ctx;
@@ -724,31 +752,34 @@ int xhci_drop_endpoint(struct usb_hcd *hcd, struct usb_device *udev,
 	}
 
 	in_ctx = xhci->devs[udev->slot_id]->in_ctx;
+	out_ctx = xhci->devs[udev->slot_id]->out_ctx;
+	ctrl_ctx = xhci_get_input_control_ctx(xhci, in_ctx);
 	ep_index = xhci_get_endpoint_index(&ep->desc);
-	ep_ctx = &xhci->devs[udev->slot_id]->out_ctx->ep[ep_index];
+	ep_ctx = xhci_get_ep_ctx(xhci, out_ctx, ep_index);
 	/* If the HC already knows the endpoint is disabled,
 	 * or the HCD has noted it is disabled, ignore this request
 	 */
 	if ((ep_ctx->ep_info & EP_STATE_MASK) == EP_STATE_DISABLED ||
-			in_ctx->drop_flags & xhci_get_endpoint_flag(&ep->desc)) {
+			ctrl_ctx->drop_flags & xhci_get_endpoint_flag(&ep->desc)) {
 		xhci_warn(xhci, "xHCI %s called with disabled ep %p\n",
 				__func__, ep);
 		return 0;
 	}
 
-	in_ctx->drop_flags |= drop_flag;
-	new_drop_flags = in_ctx->drop_flags;
+	ctrl_ctx->drop_flags |= drop_flag;
+	new_drop_flags = ctrl_ctx->drop_flags;
 
-	in_ctx->add_flags = ~drop_flag;
-	new_add_flags = in_ctx->add_flags;
+	ctrl_ctx->add_flags = ~drop_flag;
+	new_add_flags = ctrl_ctx->add_flags;
 
-	last_ctx = xhci_last_valid_endpoint(in_ctx->add_flags);
+	last_ctx = xhci_last_valid_endpoint(ctrl_ctx->add_flags);
+	slot_ctx = xhci_get_slot_ctx(xhci, in_ctx);
 	/* Update the last valid endpoint context, if we deleted the last one */
-	if ((in_ctx->slot.dev_info & LAST_CTX_MASK) > LAST_CTX(last_ctx)) {
-		in_ctx->slot.dev_info &= ~LAST_CTX_MASK;
-		in_ctx->slot.dev_info |= LAST_CTX(last_ctx);
+	if ((slot_ctx->dev_info & LAST_CTX_MASK) > LAST_CTX(last_ctx)) {
+		slot_ctx->dev_info &= ~LAST_CTX_MASK;
+		slot_ctx->dev_info |= LAST_CTX(last_ctx);
 	}
-	new_slot_info = in_ctx->slot.dev_info;
+	new_slot_info = slot_ctx->dev_info;
 
 	xhci_endpoint_zero(xhci, xhci->devs[udev->slot_id], ep);
 
@@ -778,17 +809,22 @@ int xhci_add_endpoint(struct usb_hcd *hcd, struct usb_device *udev,
 		struct usb_host_endpoint *ep)
 {
 	struct xhci_hcd *xhci;
-	struct xhci_device_control *in_ctx;
+	struct xhci_container_ctx *in_ctx, *out_ctx;
 	unsigned int ep_index;
 	struct xhci_ep_ctx *ep_ctx;
+	struct xhci_slot_ctx *slot_ctx;
+	struct xhci_input_control_ctx *ctrl_ctx;
 	u32 added_ctxs;
 	unsigned int last_ctx;
 	u32 new_add_flags, new_drop_flags, new_slot_info;
 	int ret = 0;
 
 	ret = xhci_check_args(hcd, udev, ep, 1, __func__);
-	if (ret <= 0)
+	if (ret <= 0) {
+		/* So we won't queue a reset ep command for a root hub */
+		ep->hcpriv = NULL;
 		return ret;
+	}
 	xhci = hcd_to_xhci(hcd);
 
 	added_ctxs = xhci_get_endpoint_flag(&ep->desc);
@@ -810,12 +846,14 @@ int xhci_add_endpoint(struct usb_hcd *hcd, struct usb_device *udev,
 	}
 
 	in_ctx = xhci->devs[udev->slot_id]->in_ctx;
+	out_ctx = xhci->devs[udev->slot_id]->out_ctx;
+	ctrl_ctx = xhci_get_input_control_ctx(xhci, in_ctx);
 	ep_index = xhci_get_endpoint_index(&ep->desc);
-	ep_ctx = &xhci->devs[udev->slot_id]->out_ctx->ep[ep_index];
+	ep_ctx = xhci_get_ep_ctx(xhci, out_ctx, ep_index);
 	/* If the HCD has already noted the endpoint is enabled,
 	 * ignore this request.
 	 */
-	if (in_ctx->add_flags & xhci_get_endpoint_flag(&ep->desc)) {
+	if (ctrl_ctx->add_flags & xhci_get_endpoint_flag(&ep->desc)) {
 		xhci_warn(xhci, "xHCI %s called with enabled ep %p\n",
 				__func__, ep);
 		return 0;
@@ -833,8 +871,8 @@ int xhci_add_endpoint(struct usb_hcd *hcd, struct usb_device *udev,
 		return -ENOMEM;
 	}
 
-	in_ctx->add_flags |= added_ctxs;
-	new_add_flags = in_ctx->add_flags;
+	ctrl_ctx->add_flags |= added_ctxs;
+	new_add_flags = ctrl_ctx->add_flags;
 
 	/* If xhci_endpoint_disable() was called for this endpoint, but the
 	 * xHC hasn't been notified yet through the check_bandwidth() call,
@@ -842,14 +880,18 @@ int xhci_add_endpoint(struct usb_hcd *hcd, struct usb_device *udev,
 	 * descriptors.  We must drop and re-add this endpoint, so we leave the
 	 * drop flags alone.
 	 */
-	new_drop_flags = in_ctx->drop_flags;
+	new_drop_flags = ctrl_ctx->drop_flags;
 
+	slot_ctx = xhci_get_slot_ctx(xhci, in_ctx);
 	/* Update the last valid endpoint context, if we just added one past */
-	if ((in_ctx->slot.dev_info & LAST_CTX_MASK) < LAST_CTX(last_ctx)) {
-		in_ctx->slot.dev_info &= ~LAST_CTX_MASK;
-		in_ctx->slot.dev_info |= LAST_CTX(last_ctx);
+	if ((slot_ctx->dev_info & LAST_CTX_MASK) < LAST_CTX(last_ctx)) {
+		slot_ctx->dev_info &= ~LAST_CTX_MASK;
+		slot_ctx->dev_info |= LAST_CTX(last_ctx);
 	}
-	new_slot_info = in_ctx->slot.dev_info;
+	new_slot_info = slot_ctx->dev_info;
+
+	/* Store the usb_device pointer for later use */
+	ep->hcpriv = udev;
 
 	xhci_dbg(xhci, "add ep 0x%x, slot id %d, new drop flags = %#x, new add flags = %#x, new slot info = %#x\n",
 			(unsigned int) ep->desc.bEndpointAddress,
@@ -860,9 +902,11 @@ int xhci_add_endpoint(struct usb_hcd *hcd, struct usb_device *udev,
 	return 0;
 }
 
-static void xhci_zero_in_ctx(struct xhci_virt_device *virt_dev)
+static void xhci_zero_in_ctx(struct xhci_hcd *xhci, struct xhci_virt_device *virt_dev)
 {
+	struct xhci_input_control_ctx *ctrl_ctx;
 	struct xhci_ep_ctx *ep_ctx;
+	struct xhci_slot_ctx *slot_ctx;
 	int i;
 
 	/* When a device's add flag and drop flag are zero, any subsequent
@@ -870,17 +914,18 @@ static void xhci_zero_in_ctx(struct xhci_virt_device *virt_dev)
 	 * untouched.  Make sure we don't leave any old state in the input
 	 * endpoint contexts.
 	 */
-	virt_dev->in_ctx->drop_flags = 0;
-	virt_dev->in_ctx->add_flags = 0;
-	virt_dev->in_ctx->slot.dev_info &= ~LAST_CTX_MASK;
+	ctrl_ctx = xhci_get_input_control_ctx(xhci, virt_dev->in_ctx);
+	ctrl_ctx->drop_flags = 0;
+	ctrl_ctx->add_flags = 0;
+	slot_ctx = xhci_get_slot_ctx(xhci, virt_dev->in_ctx);
+	slot_ctx->dev_info &= ~LAST_CTX_MASK;
 	/* Endpoint 0 is always valid */
-	virt_dev->in_ctx->slot.dev_info |= LAST_CTX(1);
+	slot_ctx->dev_info |= LAST_CTX(1);
 	for (i = 1; i < 31; ++i) {
-		ep_ctx = &virt_dev->in_ctx->ep[i];
+		ep_ctx = xhci_get_ep_ctx(xhci, virt_dev->in_ctx, i);
 		ep_ctx->ep_info = 0;
 		ep_ctx->ep_info2 = 0;
-		ep_ctx->deq[0] = 0;
-		ep_ctx->deq[1] = 0;
+		ep_ctx->deq = 0;
 		ep_ctx->tx_info = 0;
 	}
 }
@@ -903,6 +948,8 @@ int xhci_check_bandwidth(struct usb_hcd *hcd, struct usb_device *udev)
 	unsigned long flags;
 	struct xhci_hcd *xhci;
 	struct xhci_virt_device	*virt_dev;
+	struct xhci_input_control_ctx *ctrl_ctx;
+	struct xhci_slot_ctx *slot_ctx;
 
 	ret = xhci_check_args(hcd, udev, NULL, 0, __func__);
 	if (ret <= 0)
@@ -918,16 +965,18 @@ int xhci_check_bandwidth(struct usb_hcd *hcd, struct usb_device *udev)
 	virt_dev = xhci->devs[udev->slot_id];
 
 	/* See section 4.6.6 - A0 = 1; A1 = D0 = D1 = 0 */
-	virt_dev->in_ctx->add_flags |= SLOT_FLAG;
-	virt_dev->in_ctx->add_flags &= ~EP0_FLAG;
-	virt_dev->in_ctx->drop_flags &= ~SLOT_FLAG;
-	virt_dev->in_ctx->drop_flags &= ~EP0_FLAG;
+	ctrl_ctx = xhci_get_input_control_ctx(xhci, virt_dev->in_ctx);
+	ctrl_ctx->add_flags |= SLOT_FLAG;
+	ctrl_ctx->add_flags &= ~EP0_FLAG;
+	ctrl_ctx->drop_flags &= ~SLOT_FLAG;
+	ctrl_ctx->drop_flags &= ~EP0_FLAG;
 	xhci_dbg(xhci, "New Input Control Context:\n");
-	xhci_dbg_ctx(xhci, virt_dev->in_ctx, virt_dev->in_ctx_dma,
-			LAST_CTX_TO_EP_NUM(virt_dev->in_ctx->slot.dev_info));
+	slot_ctx = xhci_get_slot_ctx(xhci, virt_dev->in_ctx);
+	xhci_dbg_ctx(xhci, virt_dev->in_ctx,
+			LAST_CTX_TO_EP_NUM(slot_ctx->dev_info));
 
 	spin_lock_irqsave(&xhci->lock, flags);
-	ret = xhci_queue_configure_endpoint(xhci, virt_dev->in_ctx_dma,
+	ret = xhci_queue_configure_endpoint(xhci, virt_dev->in_ctx->dma,
 			udev->slot_id);
 	if (ret < 0) {
 		spin_unlock_irqrestore(&xhci->lock, flags);
@@ -982,10 +1031,10 @@ int xhci_check_bandwidth(struct usb_hcd *hcd, struct usb_device *udev)
 	}
 
 	xhci_dbg(xhci, "Output context after successful config ep cmd:\n");
-	xhci_dbg_ctx(xhci, virt_dev->out_ctx, virt_dev->out_ctx_dma,
-			LAST_CTX_TO_EP_NUM(virt_dev->in_ctx->slot.dev_info));
+	xhci_dbg_ctx(xhci, virt_dev->out_ctx,
+			LAST_CTX_TO_EP_NUM(slot_ctx->dev_info));
 
-	xhci_zero_in_ctx(virt_dev);
+	xhci_zero_in_ctx(xhci, virt_dev);
 	/* Free any old rings */
 	for (i = 1; i < 31; ++i) {
 		if (virt_dev->new_ep_rings[i]) {
@@ -1023,7 +1072,67 @@ void xhci_reset_bandwidth(struct usb_hcd *hcd, struct usb_device *udev)
 			virt_dev->new_ep_rings[i] = NULL;
 		}
 	}
-	xhci_zero_in_ctx(virt_dev);
+	xhci_zero_in_ctx(xhci, virt_dev);
+}
+
+/* Deal with stalled endpoints.  The core should have sent the control message
+ * to clear the halt condition.  However, we need to make the xHCI hardware
+ * reset its sequence number, since a device will expect a sequence number of
+ * zero after the halt condition is cleared.
+ * Context: in_interrupt
+ */
+void xhci_endpoint_reset(struct usb_hcd *hcd,
+		struct usb_host_endpoint *ep)
+{
+	struct xhci_hcd *xhci;
+	struct usb_device *udev;
+	unsigned int ep_index;
+	unsigned long flags;
+	int ret;
+	struct xhci_dequeue_state deq_state;
+	struct xhci_ring *ep_ring;
+
+	xhci = hcd_to_xhci(hcd);
+	udev = (struct usb_device *) ep->hcpriv;
+	/* Called with a root hub endpoint (or an endpoint that wasn't added
+	 * with xhci_add_endpoint()
+	 */
+	if (!ep->hcpriv)
+		return;
+	ep_index = xhci_get_endpoint_index(&ep->desc);
+	ep_ring = xhci->devs[udev->slot_id]->ep_rings[ep_index];
+	if (!ep_ring->stopped_td) {
+		xhci_dbg(xhci, "Endpoint 0x%x not halted, refusing to reset.\n",
+				ep->desc.bEndpointAddress);
+		return;
+	}
+
+	xhci_dbg(xhci, "Queueing reset endpoint command\n");
+	spin_lock_irqsave(&xhci->lock, flags);
+	ret = xhci_queue_reset_ep(xhci, udev->slot_id, ep_index);
+	/*
+	 * Can't change the ring dequeue pointer until it's transitioned to the
+	 * stopped state, which is only upon a successful reset endpoint
+	 * command.  Better hope that last command worked!
+	 */
+	if (!ret) {
+		xhci_dbg(xhci, "Cleaning up stalled endpoint ring\n");
+		/* We need to move the HW's dequeue pointer past this TD,
+		 * or it will attempt to resend it on the next doorbell ring.
+		 */
+		xhci_find_new_dequeue_state(xhci, udev->slot_id,
+				ep_index, ep_ring->stopped_td, &deq_state);
+		xhci_dbg(xhci, "Queueing new dequeue state\n");
+		xhci_queue_new_dequeue_state(xhci, ep_ring,
+				udev->slot_id,
+				ep_index, &deq_state);
+		kfree(ep_ring->stopped_td);
+		xhci_ring_cmd_db(xhci);
+	}
+	spin_unlock_irqrestore(&xhci->lock, flags);
+
+	if (ret)
+		xhci_warn(xhci, "FIXME allocate a new ring segment\n");
 }
 
 /*
@@ -1120,7 +1229,9 @@ int xhci_address_device(struct usb_hcd *hcd, struct usb_device *udev)
 	struct xhci_virt_device *virt_dev;
 	int ret = 0;
 	struct xhci_hcd *xhci = hcd_to_xhci(hcd);
-	u32 temp;
+	struct xhci_slot_ctx *slot_ctx;
+	struct xhci_input_control_ctx *ctrl_ctx;
+	u64 temp_64;
 
 	if (!udev->slot_id) {
 		xhci_dbg(xhci, "Bad Slot ID %d\n", udev->slot_id);
@@ -1133,10 +1244,12 @@ int xhci_address_device(struct usb_hcd *hcd, struct usb_device *udev)
 	if (!udev->config)
 		xhci_setup_addressable_virt_dev(xhci, udev);
 	/* Otherwise, assume the core has the device configured how it wants */
+	xhci_dbg(xhci, "Slot ID %d Input Context:\n", udev->slot_id);
+	xhci_dbg_ctx(xhci, virt_dev->in_ctx, 2);
 
 	spin_lock_irqsave(&xhci->lock, flags);
-	ret = xhci_queue_address_device(xhci, virt_dev->in_ctx_dma,
-			udev->slot_id);
+	ret = xhci_queue_address_device(xhci, virt_dev->in_ctx->dma,
+					udev->slot_id);
 	if (ret) {
 		spin_unlock_irqrestore(&xhci->lock, flags);
 		xhci_dbg(xhci, "FIXME: allocate a command ring segment\n");
@@ -1176,41 +1289,37 @@ int xhci_address_device(struct usb_hcd *hcd, struct usb_device *udev)
 	default:
 		xhci_err(xhci, "ERROR: unexpected command completion "
 				"code 0x%x.\n", virt_dev->cmd_status);
+		xhci_dbg(xhci, "Slot ID %d Output Context:\n", udev->slot_id);
+		xhci_dbg_ctx(xhci, virt_dev->out_ctx, 2);
 		ret = -EINVAL;
 		break;
 	}
 	if (ret) {
 		return ret;
 	}
-	temp = xhci_readl(xhci, &xhci->op_regs->dcbaa_ptr[0]);
-	xhci_dbg(xhci, "Op regs DCBAA ptr[0] = %#08x\n", temp);
-	temp = xhci_readl(xhci, &xhci->op_regs->dcbaa_ptr[1]);
-	xhci_dbg(xhci, "Op regs DCBAA ptr[1] = %#08x\n", temp);
-	xhci_dbg(xhci, "Slot ID %d dcbaa entry[0] @%p = %#08x\n",
-			udev->slot_id,
-			&xhci->dcbaa->dev_context_ptrs[2*udev->slot_id],
-			xhci->dcbaa->dev_context_ptrs[2*udev->slot_id]);
-	xhci_dbg(xhci, "Slot ID %d dcbaa entry[1] @%p = %#08x\n",
+	temp_64 = xhci_read_64(xhci, &xhci->op_regs->dcbaa_ptr);
+	xhci_dbg(xhci, "Op regs DCBAA ptr = %#016llx\n", temp_64);
+	xhci_dbg(xhci, "Slot ID %d dcbaa entry @%p = %#016llx\n",
 			udev->slot_id,
-			&xhci->dcbaa->dev_context_ptrs[2*udev->slot_id+1],
-			xhci->dcbaa->dev_context_ptrs[2*udev->slot_id+1]);
+			&xhci->dcbaa->dev_context_ptrs[udev->slot_id],
+			(unsigned long long)
+				xhci->dcbaa->dev_context_ptrs[udev->slot_id]);
 	xhci_dbg(xhci, "Output Context DMA address = %#08llx\n",
-			(unsigned long long)virt_dev->out_ctx_dma);
+			(unsigned long long)virt_dev->out_ctx->dma);
 	xhci_dbg(xhci, "Slot ID %d Input Context:\n", udev->slot_id);
-	xhci_dbg_ctx(xhci, virt_dev->in_ctx, virt_dev->in_ctx_dma, 2);
+	xhci_dbg_ctx(xhci, virt_dev->in_ctx, 2);
 	xhci_dbg(xhci, "Slot ID %d Output Context:\n", udev->slot_id);
-	xhci_dbg_ctx(xhci, virt_dev->out_ctx, virt_dev->out_ctx_dma, 2);
+	xhci_dbg_ctx(xhci, virt_dev->out_ctx, 2);
 	/*
 	 * USB core uses address 1 for the roothubs, so we add one to the
 	 * address given back to us by the HC.
 	 */
-	udev->devnum = (virt_dev->out_ctx->slot.dev_state & DEV_ADDR_MASK) + 1;
+	slot_ctx = xhci_get_slot_ctx(xhci, virt_dev->out_ctx);
+	udev->devnum = (slot_ctx->dev_state & DEV_ADDR_MASK) + 1;
 	/* Zero the input context control for later use */
-	virt_dev->in_ctx->add_flags = 0;
-	virt_dev->in_ctx->drop_flags = 0;
-	/* Mirror flags in the output context for future ep enable/disable */
-	virt_dev->out_ctx->add_flags = SLOT_FLAG | EP0_FLAG;
-	virt_dev->out_ctx->drop_flags = 0;
+	ctrl_ctx = xhci_get_input_control_ctx(xhci, virt_dev->in_ctx);
+	ctrl_ctx->add_flags = 0;
+	ctrl_ctx->drop_flags = 0;
 
 	xhci_dbg(xhci, "Device address = %d\n", udev->devnum);
 	/* XXX Meh, not sure if anyone else but choose_address uses this. */
@@ -1252,7 +1361,6 @@ static int __init xhci_hcd_init(void)
 	/* xhci_device_control has eight fields, and also
 	 * embeds one xhci_slot_ctx and 31 xhci_ep_ctx
 	 */
-	BUILD_BUG_ON(sizeof(struct xhci_device_control) != (8+8+8*31)*32/8);
 	BUILD_BUG_ON(sizeof(struct xhci_stream_ctx) != 4*32/8);
 	BUILD_BUG_ON(sizeof(union xhci_trb) != 4*32/8);
 	BUILD_BUG_ON(sizeof(struct xhci_erst_entry) != 4*32/8);
diff --git a/drivers/usb/host/xhci-mem.c b/drivers/usb/host/xhci-mem.c
index c8a72de1c50..e6b9a1c6002 100644
--- a/drivers/usb/host/xhci-mem.c
+++ b/drivers/usb/host/xhci-mem.c
@@ -88,7 +88,7 @@ static void xhci_link_segments(struct xhci_hcd *xhci, struct xhci_segment *prev,
 		return;
 	prev->next = next;
 	if (link_trbs) {
-		prev->trbs[TRBS_PER_SEGMENT-1].link.segment_ptr[0] = next->dma;
+		prev->trbs[TRBS_PER_SEGMENT-1].link.segment_ptr = next->dma;
 
 		/* Set the last TRB in the segment to have a TRB type ID of Link TRB */
 		val = prev->trbs[TRBS_PER_SEGMENT-1].link.control;
@@ -189,6 +189,63 @@ fail:
 	return 0;
 }
 
+#define CTX_SIZE(_hcc) (HCC_64BYTE_CONTEXT(_hcc) ? 64 : 32)
+
+struct xhci_container_ctx *xhci_alloc_container_ctx(struct xhci_hcd *xhci,
+						    int type, gfp_t flags)
+{
+	struct xhci_container_ctx *ctx = kzalloc(sizeof(*ctx), flags);
+	if (!ctx)
+		return NULL;
+
+	BUG_ON((type != XHCI_CTX_TYPE_DEVICE) && (type != XHCI_CTX_TYPE_INPUT));
+	ctx->type = type;
+	ctx->size = HCC_64BYTE_CONTEXT(xhci->hcc_params) ? 2048 : 1024;
+	if (type == XHCI_CTX_TYPE_INPUT)
+		ctx->size += CTX_SIZE(xhci->hcc_params);
+
+	ctx->bytes = dma_pool_alloc(xhci->device_pool, flags, &ctx->dma);
+	memset(ctx->bytes, 0, ctx->size);
+	return ctx;
+}
+
+void xhci_free_container_ctx(struct xhci_hcd *xhci,
+			     struct xhci_container_ctx *ctx)
+{
+	dma_pool_free(xhci->device_pool, ctx->bytes, ctx->dma);
+	kfree(ctx);
+}
+
+struct xhci_input_control_ctx *xhci_get_input_control_ctx(struct xhci_hcd *xhci,
+					      struct xhci_container_ctx *ctx)
+{
+	BUG_ON(ctx->type != XHCI_CTX_TYPE_INPUT);
+	return (struct xhci_input_control_ctx *)ctx->bytes;
+}
+
+struct xhci_slot_ctx *xhci_get_slot_ctx(struct xhci_hcd *xhci,
+					struct xhci_container_ctx *ctx)
+{
+	if (ctx->type == XHCI_CTX_TYPE_DEVICE)
+		return (struct xhci_slot_ctx *)ctx->bytes;
+
+	return (struct xhci_slot_ctx *)
+		(ctx->bytes + CTX_SIZE(xhci->hcc_params));
+}
+
+struct xhci_ep_ctx *xhci_get_ep_ctx(struct xhci_hcd *xhci,
+				    struct xhci_container_ctx *ctx,
+				    unsigned int ep_index)
+{
+	/* increment ep index by offset of start of ep ctx array */
+	ep_index++;
+	if (ctx->type == XHCI_CTX_TYPE_INPUT)
+		ep_index++;
+
+	return (struct xhci_ep_ctx *)
+		(ctx->bytes + (ep_index * CTX_SIZE(xhci->hcc_params)));
+}
+
 /* All the xhci_tds in the ring's TD list should be freed at this point */
 void xhci_free_virt_device(struct xhci_hcd *xhci, int slot_id)
 {
@@ -200,8 +257,7 @@ void xhci_free_virt_device(struct xhci_hcd *xhci, int slot_id)
 		return;
 
 	dev = xhci->devs[slot_id];
-	xhci->dcbaa->dev_context_ptrs[2*slot_id] = 0;
-	xhci->dcbaa->dev_context_ptrs[2*slot_id + 1] = 0;
+	xhci->dcbaa->dev_context_ptrs[slot_id] = 0;
 	if (!dev)
 		return;
 
@@ -210,11 +266,10 @@ void xhci_free_virt_device(struct xhci_hcd *xhci, int slot_id)
 			xhci_ring_free(xhci, dev->ep_rings[i]);
 
 	if (dev->in_ctx)
-		dma_pool_free(xhci->device_pool,
-				dev->in_ctx, dev->in_ctx_dma);
+		xhci_free_container_ctx(xhci, dev->in_ctx);
 	if (dev->out_ctx)
-		dma_pool_free(xhci->device_pool,
-				dev->out_ctx, dev->out_ctx_dma);
+		xhci_free_container_ctx(xhci, dev->out_ctx);
+
 	kfree(xhci->devs[slot_id]);
 	xhci->devs[slot_id] = 0;
 }
@@ -222,7 +277,6 @@ void xhci_free_virt_device(struct xhci_hcd *xhci, int slot_id)
 int xhci_alloc_virt_device(struct xhci_hcd *xhci, int slot_id,
 		struct usb_device *udev, gfp_t flags)
 {
-	dma_addr_t	dma;
 	struct xhci_virt_device *dev;
 
 	/* Slot ID 0 is reserved */
@@ -236,23 +290,21 @@ int xhci_alloc_virt_device(struct xhci_hcd *xhci, int slot_id,
 		return 0;
 	dev = xhci->devs[slot_id];
 
-	/* Allocate the (output) device context that will be used in the HC */
-	dev->out_ctx = dma_pool_alloc(xhci->device_pool, flags, &dma);
+	/* Allocate the (output) device context that will be used in the HC. */
+	dev->out_ctx = xhci_alloc_container_ctx(xhci, XHCI_CTX_TYPE_DEVICE, flags);
 	if (!dev->out_ctx)
 		goto fail;
-	dev->out_ctx_dma = dma;
+
 	xhci_dbg(xhci, "Slot %d output ctx = 0x%llx (dma)\n", slot_id,
-			(unsigned long long)dma);
-	memset(dev->out_ctx, 0, sizeof(*dev->out_ctx));
+			(unsigned long long)dev->out_ctx->dma);
 
 	/* Allocate the (input) device context for address device command */
-	dev->in_ctx = dma_pool_alloc(xhci->device_pool, flags, &dma);
+	dev->in_ctx = xhci_alloc_container_ctx(xhci, XHCI_CTX_TYPE_INPUT, flags);
 	if (!dev->in_ctx)
 		goto fail;
-	dev->in_ctx_dma = dma;
+
 	xhci_dbg(xhci, "Slot %d input ctx = 0x%llx (dma)\n", slot_id,
-			(unsigned long long)dma);
-	memset(dev->in_ctx, 0, sizeof(*dev->in_ctx));
+			(unsigned long long)dev->in_ctx->dma);
 
 	/* Allocate endpoint 0 ring */
 	dev->ep_rings[0] = xhci_ring_alloc(xhci, 1, true, flags);
@@ -261,17 +313,12 @@ int xhci_alloc_virt_device(struct xhci_hcd *xhci, int slot_id,
 
 	init_completion(&dev->cmd_completion);
 
-	/*
-	 * Point to output device context in dcbaa; skip the output control
-	 * context, which is eight 32 bit fields (or 32 bytes long)
-	 */
-	xhci->dcbaa->dev_context_ptrs[2*slot_id] =
-		(u32) dev->out_ctx_dma + (32);
+	/* Point to output device context in dcbaa. */
+	xhci->dcbaa->dev_context_ptrs[slot_id] = dev->out_ctx->dma;
 	xhci_dbg(xhci, "Set slot id %d dcbaa entry %p to 0x%llx\n",
 			slot_id,
-			&xhci->dcbaa->dev_context_ptrs[2*slot_id],
-			(unsigned long long)dev->out_ctx_dma);
-	xhci->dcbaa->dev_context_ptrs[2*slot_id + 1] = 0;
+			&xhci->dcbaa->dev_context_ptrs[slot_id],
+			(unsigned long long) xhci->dcbaa->dev_context_ptrs[slot_id]);
 
 	return 1;
 fail:
@@ -285,6 +332,8 @@ int xhci_setup_addressable_virt_dev(struct xhci_hcd *xhci, struct usb_device *ud
 	struct xhci_virt_device *dev;
 	struct xhci_ep_ctx	*ep0_ctx;
 	struct usb_device	*top_dev;
+	struct xhci_slot_ctx    *slot_ctx;
+	struct xhci_input_control_ctx *ctrl_ctx;
 
 	dev = xhci->devs[udev->slot_id];
 	/* Slot ID 0 is reserved */
@@ -293,27 +342,29 @@ int xhci_setup_addressable_virt_dev(struct xhci_hcd *xhci, struct usb_device *ud
 				udev->slot_id);
 		return -EINVAL;
 	}
-	ep0_ctx = &dev->in_ctx->ep[0];
+	ep0_ctx = xhci_get_ep_ctx(xhci, dev->in_ctx, 0);
+	ctrl_ctx = xhci_get_input_control_ctx(xhci, dev->in_ctx);
+	slot_ctx = xhci_get_slot_ctx(xhci, dev->in_ctx);
 
 	/* 2) New slot context and endpoint 0 context are valid*/
-	dev->in_ctx->add_flags = SLOT_FLAG | EP0_FLAG;
+	ctrl_ctx->add_flags = SLOT_FLAG | EP0_FLAG;
 
 	/* 3) Only the control endpoint is valid - one endpoint context */
-	dev->in_ctx->slot.dev_info |= LAST_CTX(1);
+	slot_ctx->dev_info |= LAST_CTX(1);
 
 	switch (udev->speed) {
 	case USB_SPEED_SUPER:
-		dev->in_ctx->slot.dev_info |= (u32) udev->route;
-		dev->in_ctx->slot.dev_info |= (u32) SLOT_SPEED_SS;
+		slot_ctx->dev_info |= (u32) udev->route;
+		slot_ctx->dev_info |= (u32) SLOT_SPEED_SS;
 		break;
 	case USB_SPEED_HIGH:
-		dev->in_ctx->slot.dev_info |= (u32) SLOT_SPEED_HS;
+		slot_ctx->dev_info |= (u32) SLOT_SPEED_HS;
 		break;
 	case USB_SPEED_FULL:
-		dev->in_ctx->slot.dev_info |= (u32) SLOT_SPEED_FS;
+		slot_ctx->dev_info |= (u32) SLOT_SPEED_FS;
 		break;
 	case USB_SPEED_LOW:
-		dev->in_ctx->slot.dev_info |= (u32) SLOT_SPEED_LS;
+		slot_ctx->dev_info |= (u32) SLOT_SPEED_LS;
 		break;
 	case USB_SPEED_VARIABLE:
 		xhci_dbg(xhci, "FIXME xHCI doesn't support wireless speeds\n");
@@ -327,7 +378,7 @@ int xhci_setup_addressable_virt_dev(struct xhci_hcd *xhci, struct usb_device *ud
 	for (top_dev = udev; top_dev->parent && top_dev->parent->parent;
 			top_dev = top_dev->parent)
 		/* Found device below root hub */;
-	dev->in_ctx->slot.dev_info2 |= (u32) ROOT_HUB_PORT(top_dev->portnum);
+	slot_ctx->dev_info2 |= (u32) ROOT_HUB_PORT(top_dev->portnum);
 	xhci_dbg(xhci, "Set root hub portnum to %d\n", top_dev->portnum);
 
 	/* Is this a LS/FS device under a HS hub? */
@@ -337,8 +388,8 @@ int xhci_setup_addressable_virt_dev(struct xhci_hcd *xhci, struct usb_device *ud
 	 */
 	if ((udev->speed == USB_SPEED_LOW || udev->speed == USB_SPEED_FULL) &&
 			udev->tt) {
-		dev->in_ctx->slot.tt_info = udev->tt->hub->slot_id;
-		dev->in_ctx->slot.tt_info |= udev->ttport << 8;
+		slot_ctx->tt_info = udev->tt->hub->slot_id;
+		slot_ctx->tt_info |= udev->ttport << 8;
 	}
 	xhci_dbg(xhci, "udev->tt = %p\n", udev->tt);
 	xhci_dbg(xhci, "udev->ttport = 0x%x\n", udev->ttport);
@@ -360,10 +411,9 @@ int xhci_setup_addressable_virt_dev(struct xhci_hcd *xhci, struct usb_device *ud
 	ep0_ctx->ep_info2 |= MAX_BURST(0);
 	ep0_ctx->ep_info2 |= ERROR_COUNT(3);
 
-	ep0_ctx->deq[0] =
+	ep0_ctx->deq =
 		dev->ep_rings[0]->first_seg->dma;
-	ep0_ctx->deq[0] |= dev->ep_rings[0]->cycle_state;
-	ep0_ctx->deq[1] = 0;
+	ep0_ctx->deq |= dev->ep_rings[0]->cycle_state;
 
 	/* Steps 7 and 8 were done in xhci_alloc_virt_device() */
 
@@ -470,25 +520,26 @@ int xhci_endpoint_init(struct xhci_hcd *xhci,
 	unsigned int max_burst;
 
 	ep_index = xhci_get_endpoint_index(&ep->desc);
-	ep_ctx = &virt_dev->in_ctx->ep[ep_index];
+	ep_ctx = xhci_get_ep_ctx(xhci, virt_dev->in_ctx, ep_index);
 
 	/* Set up the endpoint ring */
 	virt_dev->new_ep_rings[ep_index] = xhci_ring_alloc(xhci, 1, true, mem_flags);
 	if (!virt_dev->new_ep_rings[ep_index])
 		return -ENOMEM;
 	ep_ring = virt_dev->new_ep_rings[ep_index];
-	ep_ctx->deq[0] = ep_ring->first_seg->dma | ep_ring->cycle_state;
-	ep_ctx->deq[1] = 0;
+	ep_ctx->deq = ep_ring->first_seg->dma | ep_ring->cycle_state;
 
 	ep_ctx->ep_info = xhci_get_endpoint_interval(udev, ep);
 
 	/* FIXME dig Mult and streams info out of ep companion desc */
 
-	/* Allow 3 retries for everything but isoc */
+	/* Allow 3 retries for everything but isoc;
+	 * error count = 0 means infinite retries.
+	 */
 	if (!usb_endpoint_xfer_isoc(&ep->desc))
 		ep_ctx->ep_info2 = ERROR_COUNT(3);
 	else
-		ep_ctx->ep_info2 = ERROR_COUNT(0);
+		ep_ctx->ep_info2 = ERROR_COUNT(1);
 
 	ep_ctx->ep_info2 |= xhci_get_endpoint_type(udev, ep);
 
@@ -498,7 +549,12 @@ int xhci_endpoint_init(struct xhci_hcd *xhci,
 		max_packet = ep->desc.wMaxPacketSize;
 		ep_ctx->ep_info2 |= MAX_PACKET(max_packet);
 		/* dig out max burst from ep companion desc */
-		max_packet = ep->ss_ep_comp->desc.bMaxBurst;
+		if (!ep->ss_ep_comp) {
+			xhci_warn(xhci, "WARN no SS endpoint companion descriptor.\n");
+			max_packet = 0;
+		} else {
+			max_packet = ep->ss_ep_comp->desc.bMaxBurst;
+		}
 		ep_ctx->ep_info2 |= MAX_BURST(max_packet);
 		break;
 	case USB_SPEED_HIGH:
@@ -531,18 +587,114 @@ void xhci_endpoint_zero(struct xhci_hcd *xhci,
 	struct xhci_ep_ctx *ep_ctx;
 
 	ep_index = xhci_get_endpoint_index(&ep->desc);
-	ep_ctx = &virt_dev->in_ctx->ep[ep_index];
+	ep_ctx = xhci_get_ep_ctx(xhci, virt_dev->in_ctx, ep_index);
 
 	ep_ctx->ep_info = 0;
 	ep_ctx->ep_info2 = 0;
-	ep_ctx->deq[0] = 0;
-	ep_ctx->deq[1] = 0;
+	ep_ctx->deq = 0;
 	ep_ctx->tx_info = 0;
 	/* Don't free the endpoint ring until the set interface or configuration
 	 * request succeeds.
 	 */
 }
 
+/* Set up the scratchpad buffer array and scratchpad buffers, if needed. */
+static int scratchpad_alloc(struct xhci_hcd *xhci, gfp_t flags)
+{
+	int i;
+	struct device *dev = xhci_to_hcd(xhci)->self.controller;
+	int num_sp = HCS_MAX_SCRATCHPAD(xhci->hcs_params2);
+
+	xhci_dbg(xhci, "Allocating %d scratchpad buffers\n", num_sp);
+
+	if (!num_sp)
+		return 0;
+
+	xhci->scratchpad = kzalloc(sizeof(*xhci->scratchpad), flags);
+	if (!xhci->scratchpad)
+		goto fail_sp;
+
+	xhci->scratchpad->sp_array =
+		pci_alloc_consistent(to_pci_dev(dev),
+				     num_sp * sizeof(u64),
+				     &xhci->scratchpad->sp_dma);
+	if (!xhci->scratchpad->sp_array)
+		goto fail_sp2;
+
+	xhci->scratchpad->sp_buffers = kzalloc(sizeof(void *) * num_sp, flags);
+	if (!xhci->scratchpad->sp_buffers)
+		goto fail_sp3;
+
+	xhci->scratchpad->sp_dma_buffers =
+		kzalloc(sizeof(dma_addr_t) * num_sp, flags);
+
+	if (!xhci->scratchpad->sp_dma_buffers)
+		goto fail_sp4;
+
+	xhci->dcbaa->dev_context_ptrs[0] = xhci->scratchpad->sp_dma;
+	for (i = 0; i < num_sp; i++) {
+		dma_addr_t dma;
+		void *buf = pci_alloc_consistent(to_pci_dev(dev),
+						 xhci->page_size, &dma);
+		if (!buf)
+			goto fail_sp5;
+
+		xhci->scratchpad->sp_array[i] = dma;
+		xhci->scratchpad->sp_buffers[i] = buf;
+		xhci->scratchpad->sp_dma_buffers[i] = dma;
+	}
+
+	return 0;
+
+ fail_sp5:
+	for (i = i - 1; i >= 0; i--) {
+		pci_free_consistent(to_pci_dev(dev), xhci->page_size,
+				    xhci->scratchpad->sp_buffers[i],
+				    xhci->scratchpad->sp_dma_buffers[i]);
+	}
+	kfree(xhci->scratchpad->sp_dma_buffers);
+
+ fail_sp4:
+	kfree(xhci->scratchpad->sp_buffers);
+
+ fail_sp3:
+	pci_free_consistent(to_pci_dev(dev), num_sp * sizeof(u64),
+			    xhci->scratchpad->sp_array,
+			    xhci->scratchpad->sp_dma);
+
+ fail_sp2:
+	kfree(xhci->scratchpad);
+	xhci->scratchpad = NULL;
+
+ fail_sp:
+	return -ENOMEM;
+}
+
+static void scratchpad_free(struct xhci_hcd *xhci)
+{
+	int num_sp;
+	int i;
+	struct pci_dev	*pdev = to_pci_dev(xhci_to_hcd(xhci)->self.controller);
+
+	if (!xhci->scratchpad)
+		return;
+
+	num_sp = HCS_MAX_SCRATCHPAD(xhci->hcs_params2);
+
+	for (i = 0; i < num_sp; i++) {
+		pci_free_consistent(pdev, xhci->page_size,
+				    xhci->scratchpad->sp_buffers[i],
+				    xhci->scratchpad->sp_dma_buffers[i]);
+	}
+	kfree(xhci->scratchpad->sp_dma_buffers);
+	kfree(xhci->scratchpad->sp_buffers);
+	pci_free_consistent(pdev, num_sp * sizeof(u64),
+			    xhci->scratchpad->sp_array,
+			    xhci->scratchpad->sp_dma);
+	kfree(xhci->scratchpad);
+	xhci->scratchpad = NULL;
+}
+
 void xhci_mem_cleanup(struct xhci_hcd *xhci)
 {
 	struct pci_dev	*pdev = to_pci_dev(xhci_to_hcd(xhci)->self.controller);
@@ -551,10 +703,8 @@ void xhci_mem_cleanup(struct xhci_hcd *xhci)
 
 	/* Free the Event Ring Segment Table and the actual Event Ring */
 	xhci_writel(xhci, 0, &xhci->ir_set->erst_size);
-	xhci_writel(xhci, 0, &xhci->ir_set->erst_base[0]);
-	xhci_writel(xhci, 0, &xhci->ir_set->erst_base[1]);
-	xhci_writel(xhci, 0, &xhci->ir_set->erst_dequeue[0]);
-	xhci_writel(xhci, 0, &xhci->ir_set->erst_dequeue[1]);
+	xhci_write_64(xhci, 0, &xhci->ir_set->erst_base);
+	xhci_write_64(xhci, 0, &xhci->ir_set->erst_dequeue);
 	size = sizeof(struct xhci_erst_entry)*(xhci->erst.num_entries);
 	if (xhci->erst.entries)
 		pci_free_consistent(pdev, size,
@@ -566,8 +716,7 @@ void xhci_mem_cleanup(struct xhci_hcd *xhci)
 	xhci->event_ring = NULL;
 	xhci_dbg(xhci, "Freed event ring\n");
 
-	xhci_writel(xhci, 0, &xhci->op_regs->cmd_ring[0]);
-	xhci_writel(xhci, 0, &xhci->op_regs->cmd_ring[1]);
+	xhci_write_64(xhci, 0, &xhci->op_regs->cmd_ring);
 	if (xhci->cmd_ring)
 		xhci_ring_free(xhci, xhci->cmd_ring);
 	xhci->cmd_ring = NULL;
@@ -586,8 +735,7 @@ void xhci_mem_cleanup(struct xhci_hcd *xhci)
 	xhci->device_pool = NULL;
 	xhci_dbg(xhci, "Freed device context pool\n");
 
-	xhci_writel(xhci, 0, &xhci->op_regs->dcbaa_ptr[0]);
-	xhci_writel(xhci, 0, &xhci->op_regs->dcbaa_ptr[1]);
+	xhci_write_64(xhci, 0, &xhci->op_regs->dcbaa_ptr);
 	if (xhci->dcbaa)
 		pci_free_consistent(pdev, sizeof(*xhci->dcbaa),
 				xhci->dcbaa, xhci->dcbaa->dma);
@@ -595,6 +743,7 @@ void xhci_mem_cleanup(struct xhci_hcd *xhci)
 
 	xhci->page_size = 0;
 	xhci->page_shift = 0;
+	scratchpad_free(xhci);
 }
 
 int xhci_mem_init(struct xhci_hcd *xhci, gfp_t flags)
@@ -602,6 +751,7 @@ int xhci_mem_init(struct xhci_hcd *xhci, gfp_t flags)
 	dma_addr_t	dma;
 	struct device	*dev = xhci_to_hcd(xhci)->self.controller;
 	unsigned int	val, val2;
+	u64		val_64;
 	struct xhci_segment	*seg;
 	u32 page_size;
 	int i;
@@ -647,8 +797,7 @@ int xhci_mem_init(struct xhci_hcd *xhci, gfp_t flags)
 	xhci->dcbaa->dma = dma;
 	xhci_dbg(xhci, "// Device context base array address = 0x%llx (DMA), %p (virt)\n",
 			(unsigned long long)xhci->dcbaa->dma, xhci->dcbaa);
-	xhci_writel(xhci, dma, &xhci->op_regs->dcbaa_ptr[0]);
-	xhci_writel(xhci, (u32) 0, &xhci->op_regs->dcbaa_ptr[1]);
+	xhci_write_64(xhci, dma, &xhci->op_regs->dcbaa_ptr);
 
 	/*
 	 * Initialize the ring segment pool.  The ring must be a contiguous
@@ -658,11 +807,10 @@ int xhci_mem_init(struct xhci_hcd *xhci, gfp_t flags)
 	 */
 	xhci->segment_pool = dma_pool_create("xHCI ring segments", dev,
 			SEGMENT_SIZE, 64, xhci->page_size);
+
 	/* See Table 46 and Note on Figure 55 */
-	/* FIXME support 64-byte contexts */
 	xhci->device_pool = dma_pool_create("xHCI input/output contexts", dev,
-			sizeof(struct xhci_device_control),
-			64, xhci->page_size);
+			2112, 64, xhci->page_size);
 	if (!xhci->segment_pool || !xhci->device_pool)
 		goto fail;
 
@@ -675,14 +823,12 @@ int xhci_mem_init(struct xhci_hcd *xhci, gfp_t flags)
 			(unsigned long long)xhci->cmd_ring->first_seg->dma);
 
 	/* Set the address in the Command Ring Control register */
-	val = xhci_readl(xhci, &xhci->op_regs->cmd_ring[0]);
-	val = (val & ~CMD_RING_ADDR_MASK) |
-		(xhci->cmd_ring->first_seg->dma & CMD_RING_ADDR_MASK) |
+	val_64 = xhci_read_64(xhci, &xhci->op_regs->cmd_ring);
+	val_64 = (val_64 & (u64) CMD_RING_RSVD_BITS) |
+		(xhci->cmd_ring->first_seg->dma & (u64) ~CMD_RING_RSVD_BITS) |
 		xhci->cmd_ring->cycle_state;
-	xhci_dbg(xhci, "// Setting command ring address low bits to 0x%x\n", val);
-	xhci_writel(xhci, val, &xhci->op_regs->cmd_ring[0]);
-	xhci_dbg(xhci, "// Setting command ring address high bits to 0x0\n");
-	xhci_writel(xhci, (u32) 0, &xhci->op_regs->cmd_ring[1]);
+	xhci_dbg(xhci, "// Setting command ring address to 0x%x\n", val);
+	xhci_write_64(xhci, val_64, &xhci->op_regs->cmd_ring);
 	xhci_dbg_cmd_ptrs(xhci);
 
 	val = xhci_readl(xhci, &xhci->cap_regs->db_off);
@@ -722,8 +868,7 @@ int xhci_mem_init(struct xhci_hcd *xhci, gfp_t flags)
 	/* set ring base address and size for each segment table entry */
 	for (val = 0, seg = xhci->event_ring->first_seg; val < ERST_NUM_SEGS; val++) {
 		struct xhci_erst_entry *entry = &xhci->erst.entries[val];
-		entry->seg_addr[0] = seg->dma;
-		entry->seg_addr[1] = 0;
+		entry->seg_addr = seg->dma;
 		entry->seg_size = TRBS_PER_SEGMENT;
 		entry->rsvd = 0;
 		seg = seg->next;
@@ -741,11 +886,10 @@ int xhci_mem_init(struct xhci_hcd *xhci, gfp_t flags)
 	/* set the segment table base address */
 	xhci_dbg(xhci, "// Set ERST base address for ir_set 0 = 0x%llx\n",
 			(unsigned long long)xhci->erst.erst_dma_addr);
-	val = xhci_readl(xhci, &xhci->ir_set->erst_base[0]);
-	val &= ERST_PTR_MASK;
-	val |= (xhci->erst.erst_dma_addr & ~ERST_PTR_MASK);
-	xhci_writel(xhci, val, &xhci->ir_set->erst_base[0]);
-	xhci_writel(xhci, 0, &xhci->ir_set->erst_base[1]);
+	val_64 = xhci_read_64(xhci, &xhci->ir_set->erst_base);
+	val_64 &= ERST_PTR_MASK;
+	val_64 |= (xhci->erst.erst_dma_addr & (u64) ~ERST_PTR_MASK);
+	xhci_write_64(xhci, val_64, &xhci->ir_set->erst_base);
 
 	/* Set the event ring dequeue address */
 	xhci_set_hc_event_deq(xhci);
@@ -761,7 +905,11 @@ int xhci_mem_init(struct xhci_hcd *xhci, gfp_t flags)
 	for (i = 0; i < MAX_HC_SLOTS; ++i)
 		xhci->devs[i] = 0;
 
+	if (scratchpad_alloc(xhci, flags))
+		goto fail;
+
 	return 0;
+
 fail:
 	xhci_warn(xhci, "Couldn't initialize memory\n");
 	xhci_mem_cleanup(xhci);
diff --git a/drivers/usb/host/xhci-pci.c b/drivers/usb/host/xhci-pci.c
index 1462709e26c..592fe7e623f 100644
--- a/drivers/usb/host/xhci-pci.c
+++ b/drivers/usb/host/xhci-pci.c
@@ -117,6 +117,7 @@ static const struct hc_driver xhci_pci_hc_driver = {
 	.free_dev =		xhci_free_dev,
 	.add_endpoint =		xhci_add_endpoint,
 	.drop_endpoint =	xhci_drop_endpoint,
+	.endpoint_reset =	xhci_endpoint_reset,
 	.check_bandwidth =	xhci_check_bandwidth,
 	.reset_bandwidth =	xhci_reset_bandwidth,
 	.address_device =	xhci_address_device,
diff --git a/drivers/usb/host/xhci-ring.c b/drivers/usb/host/xhci-ring.c
index 02d81985c45..aa88a067148 100644
--- a/drivers/usb/host/xhci-ring.c
+++ b/drivers/usb/host/xhci-ring.c
@@ -135,6 +135,7 @@ static void next_trb(struct xhci_hcd *xhci,
 static void inc_deq(struct xhci_hcd *xhci, struct xhci_ring *ring, bool consumer)
 {
 	union xhci_trb *next = ++(ring->dequeue);
+	unsigned long long addr;
 
 	ring->deq_updates++;
 	/* Update the dequeue pointer further if that was a link TRB or we're at
@@ -152,6 +153,13 @@ static void inc_deq(struct xhci_hcd *xhci, struct xhci_ring *ring, bool consumer
 		ring->dequeue = ring->deq_seg->trbs;
 		next = ring->dequeue;
 	}
+	addr = (unsigned long long) xhci_trb_virt_to_dma(ring->deq_seg, ring->dequeue);
+	if (ring == xhci->event_ring)
+		xhci_dbg(xhci, "Event ring deq = 0x%llx (DMA)\n", addr);
+	else if (ring == xhci->cmd_ring)
+		xhci_dbg(xhci, "Command ring deq = 0x%llx (DMA)\n", addr);
+	else
+		xhci_dbg(xhci, "Ring deq = 0x%llx (DMA)\n", addr);
 }
 
 /*
@@ -171,6 +179,7 @@ static void inc_enq(struct xhci_hcd *xhci, struct xhci_ring *ring, bool consumer
 {
 	u32 chain;
 	union xhci_trb *next;
+	unsigned long long addr;
 
 	chain = ring->enqueue->generic.field[3] & TRB_CHAIN;
 	next = ++(ring->enqueue);
@@ -204,6 +213,13 @@ static void inc_enq(struct xhci_hcd *xhci, struct xhci_ring *ring, bool consumer
 		ring->enqueue = ring->enq_seg->trbs;
 		next = ring->enqueue;
 	}
+	addr = (unsigned long long) xhci_trb_virt_to_dma(ring->enq_seg, ring->enqueue);
+	if (ring == xhci->event_ring)
+		xhci_dbg(xhci, "Event ring enq = 0x%llx (DMA)\n", addr);
+	else if (ring == xhci->cmd_ring)
+		xhci_dbg(xhci, "Command ring enq = 0x%llx (DMA)\n", addr);
+	else
+		xhci_dbg(xhci, "Ring enq = 0x%llx (DMA)\n", addr);
 }
 
 /*
@@ -237,7 +253,7 @@ static int room_on_ring(struct xhci_hcd *xhci, struct xhci_ring *ring,
 
 void xhci_set_hc_event_deq(struct xhci_hcd *xhci)
 {
-	u32 temp;
+	u64 temp;
 	dma_addr_t deq;
 
 	deq = xhci_trb_virt_to_dma(xhci->event_ring->deq_seg,
@@ -246,13 +262,15 @@ void xhci_set_hc_event_deq(struct xhci_hcd *xhci)
 		xhci_warn(xhci, "WARN something wrong with SW event ring "
 				"dequeue ptr.\n");
 	/* Update HC event ring dequeue pointer */
-	temp = xhci_readl(xhci, &xhci->ir_set->erst_dequeue[0]);
+	temp = xhci_read_64(xhci, &xhci->ir_set->erst_dequeue);
 	temp &= ERST_PTR_MASK;
-	if (!in_interrupt())
-		xhci_dbg(xhci, "// Write event ring dequeue pointer\n");
-	xhci_writel(xhci, 0, &xhci->ir_set->erst_dequeue[1]);
-	xhci_writel(xhci, (deq & ~ERST_PTR_MASK) | temp,
-			&xhci->ir_set->erst_dequeue[0]);
+	/* Don't clear the EHB bit (which is RW1C) because
+	 * there might be more events to service.
+	 */
+	temp &= ~ERST_EHB;
+	xhci_dbg(xhci, "// Write event ring dequeue pointer, preserving EHB bit\n");
+	xhci_write_64(xhci, ((u64) deq & (u64) ~ERST_PTR_MASK) | temp,
+			&xhci->ir_set->erst_dequeue);
 }
 
 /* Ring the host controller doorbell after placing a command on the ring */
@@ -279,7 +297,8 @@ static void ring_ep_doorbell(struct xhci_hcd *xhci,
 	/* Don't ring the doorbell for this endpoint if there are pending
 	 * cancellations because the we don't want to interrupt processing.
 	 */
-	if (!ep_ring->cancels_pending && !(ep_ring->state & SET_DEQ_PENDING)) {
+	if (!ep_ring->cancels_pending && !(ep_ring->state & SET_DEQ_PENDING)
+			&& !(ep_ring->state & EP_HALTED)) {
 		field = xhci_readl(xhci, db_addr) & DB_MASK;
 		xhci_writel(xhci, field | EPI_TO_DB(ep_index), db_addr);
 		/* Flush PCI posted writes - FIXME Matthew Wilcox says this
@@ -316,12 +335,6 @@ static struct xhci_segment *find_trb_seg(
 	return cur_seg;
 }
 
-struct dequeue_state {
-	struct xhci_segment *new_deq_seg;
-	union xhci_trb *new_deq_ptr;
-	int new_cycle_state;
-};
-
 /*
  * Move the xHC's endpoint ring dequeue pointer past cur_td.
  * Record the new state of the xHC's endpoint ring dequeue segment,
@@ -336,24 +349,30 @@ struct dequeue_state {
  *  - Finally we move the dequeue state one TRB further, toggling the cycle bit
  *    if we've moved it past a link TRB with the toggle cycle bit set.
  */
-static void find_new_dequeue_state(struct xhci_hcd *xhci,
+void xhci_find_new_dequeue_state(struct xhci_hcd *xhci,
 		unsigned int slot_id, unsigned int ep_index,
-		struct xhci_td *cur_td, struct dequeue_state *state)
+		struct xhci_td *cur_td, struct xhci_dequeue_state *state)
 {
 	struct xhci_virt_device *dev = xhci->devs[slot_id];
 	struct xhci_ring *ep_ring = dev->ep_rings[ep_index];
 	struct xhci_generic_trb *trb;
+	struct xhci_ep_ctx *ep_ctx;
+	dma_addr_t addr;
 
 	state->new_cycle_state = 0;
+	xhci_dbg(xhci, "Finding segment containing stopped TRB.\n");
 	state->new_deq_seg = find_trb_seg(cur_td->start_seg,
 			ep_ring->stopped_trb,
 			&state->new_cycle_state);
 	if (!state->new_deq_seg)
 		BUG();
 	/* Dig out the cycle state saved by the xHC during the stop ep cmd */
-	state->new_cycle_state = 0x1 & dev->out_ctx->ep[ep_index].deq[0];
+	xhci_dbg(xhci, "Finding endpoint context\n");
+	ep_ctx = xhci_get_ep_ctx(xhci, dev->out_ctx, ep_index);
+	state->new_cycle_state = 0x1 & ep_ctx->deq;
 
 	state->new_deq_ptr = cur_td->last_trb;
+	xhci_dbg(xhci, "Finding segment containing last TRB in TD.\n");
 	state->new_deq_seg = find_trb_seg(state->new_deq_seg,
 			state->new_deq_ptr,
 			&state->new_cycle_state);
@@ -367,6 +386,12 @@ static void find_new_dequeue_state(struct xhci_hcd *xhci,
 	next_trb(xhci, ep_ring, &state->new_deq_seg, &state->new_deq_ptr);
 
 	/* Don't update the ring cycle state for the producer (us). */
+	xhci_dbg(xhci, "New dequeue segment = %p (virtual)\n",
+			state->new_deq_seg);
+	addr = xhci_trb_virt_to_dma(state->new_deq_seg, state->new_deq_ptr);
+	xhci_dbg(xhci, "New dequeue pointer = 0x%llx (DMA)\n",
+			(unsigned long long) addr);
+	xhci_dbg(xhci, "Setting dequeue pointer in internal ring state.\n");
 	ep_ring->dequeue = state->new_deq_ptr;
 	ep_ring->deq_seg = state->new_deq_seg;
 }
@@ -416,6 +441,30 @@ static int queue_set_tr_deq(struct xhci_hcd *xhci, int slot_id,
 		unsigned int ep_index, struct xhci_segment *deq_seg,
 		union xhci_trb *deq_ptr, u32 cycle_state);
 
+void xhci_queue_new_dequeue_state(struct xhci_hcd *xhci,
+		struct xhci_ring *ep_ring, unsigned int slot_id,
+		unsigned int ep_index, struct xhci_dequeue_state *deq_state)
+{
+	xhci_dbg(xhci, "Set TR Deq Ptr cmd, new deq seg = %p (0x%llx dma), "
+			"new deq ptr = %p (0x%llx dma), new cycle = %u\n",
+			deq_state->new_deq_seg,
+			(unsigned long long)deq_state->new_deq_seg->dma,
+			deq_state->new_deq_ptr,
+			(unsigned long long)xhci_trb_virt_to_dma(deq_state->new_deq_seg, deq_state->new_deq_ptr),
+			deq_state->new_cycle_state);
+	queue_set_tr_deq(xhci, slot_id, ep_index,
+			deq_state->new_deq_seg,
+			deq_state->new_deq_ptr,
+			(u32) deq_state->new_cycle_state);
+	/* Stop the TD queueing code from ringing the doorbell until
+	 * this command completes.  The HC won't set the dequeue pointer
+	 * if the ring is running, and ringing the doorbell starts the
+	 * ring running.
+	 */
+	ep_ring->state |= SET_DEQ_PENDING;
+	xhci_ring_cmd_db(xhci);
+}
+
 /*
  * When we get a command completion for a Stop Endpoint Command, we need to
  * unlink any cancelled TDs from the ring.  There are two ways to do that:
@@ -436,7 +485,7 @@ static void handle_stopped_endpoint(struct xhci_hcd *xhci,
 	struct xhci_td *cur_td = 0;
 	struct xhci_td *last_unlinked_td;
 
-	struct dequeue_state deq_state;
+	struct xhci_dequeue_state deq_state;
 #ifdef CONFIG_USB_HCD_STAT
 	ktime_t stop_time = ktime_get();
 #endif
@@ -464,7 +513,7 @@ static void handle_stopped_endpoint(struct xhci_hcd *xhci,
 		 * move the xHC endpoint ring dequeue pointer past this TD.
 		 */
 		if (cur_td == ep_ring->stopped_td)
-			find_new_dequeue_state(xhci, slot_id, ep_index, cur_td,
+			xhci_find_new_dequeue_state(xhci, slot_id, ep_index, cur_td,
 					&deq_state);
 		else
 			td_to_noop(xhci, ep_ring, cur_td);
@@ -480,24 +529,8 @@ static void handle_stopped_endpoint(struct xhci_hcd *xhci,
 
 	/* If necessary, queue a Set Transfer Ring Dequeue Pointer command */
 	if (deq_state.new_deq_ptr && deq_state.new_deq_seg) {
-		xhci_dbg(xhci, "Set TR Deq Ptr cmd, new deq seg = %p (0x%llx dma), "
-				"new deq ptr = %p (0x%llx dma), new cycle = %u\n",
-				deq_state.new_deq_seg,
-				(unsigned long long)deq_state.new_deq_seg->dma,
-				deq_state.new_deq_ptr,
-				(unsigned long long)xhci_trb_virt_to_dma(deq_state.new_deq_seg, deq_state.new_deq_ptr),
-				deq_state.new_cycle_state);
-		queue_set_tr_deq(xhci, slot_id, ep_index,
-				deq_state.new_deq_seg,
-				deq_state.new_deq_ptr,
-				(u32) deq_state.new_cycle_state);
-		/* Stop the TD queueing code from ringing the doorbell until
-		 * this command completes.  The HC won't set the dequeue pointer
-		 * if the ring is running, and ringing the doorbell starts the
-		 * ring running.
-		 */
-		ep_ring->state |= SET_DEQ_PENDING;
-		xhci_ring_cmd_db(xhci);
+		xhci_queue_new_dequeue_state(xhci, ep_ring,
+				slot_id, ep_index, &deq_state);
 	} else {
 		/* Otherwise just ring the doorbell to restart the ring */
 		ring_ep_doorbell(xhci, slot_id, ep_index);
@@ -551,11 +584,15 @@ static void handle_set_deq_completion(struct xhci_hcd *xhci,
 	unsigned int ep_index;
 	struct xhci_ring *ep_ring;
 	struct xhci_virt_device *dev;
+	struct xhci_ep_ctx *ep_ctx;
+	struct xhci_slot_ctx *slot_ctx;
 
 	slot_id = TRB_TO_SLOT_ID(trb->generic.field[3]);
 	ep_index = TRB_TO_EP_INDEX(trb->generic.field[3]);
 	dev = xhci->devs[slot_id];
 	ep_ring = dev->ep_rings[ep_index];
+	ep_ctx = xhci_get_ep_ctx(xhci, dev->out_ctx, ep_index);
+	slot_ctx = xhci_get_slot_ctx(xhci, dev->out_ctx);
 
 	if (GET_COMP_CODE(event->status) != COMP_SUCCESS) {
 		unsigned int ep_state;
@@ -569,9 +606,9 @@ static void handle_set_deq_completion(struct xhci_hcd *xhci,
 		case COMP_CTX_STATE:
 			xhci_warn(xhci, "WARN Set TR Deq Ptr cmd failed due "
 					"to incorrect slot or ep state.\n");
-			ep_state = dev->out_ctx->ep[ep_index].ep_info;
+			ep_state = ep_ctx->ep_info;
 			ep_state &= EP_STATE_MASK;
-			slot_state = dev->out_ctx->slot.dev_state;
+			slot_state = slot_ctx->dev_state;
 			slot_state = GET_SLOT_STATE(slot_state);
 			xhci_dbg(xhci, "Slot state = %u, EP state = %u\n",
 					slot_state, ep_state);
@@ -593,16 +630,33 @@ static void handle_set_deq_completion(struct xhci_hcd *xhci,
 		 * cancelling URBs, which might not be an error...
 		 */
 	} else {
-		xhci_dbg(xhci, "Successful Set TR Deq Ptr cmd, deq[0] = 0x%x, "
-				"deq[1] = 0x%x.\n",
-				dev->out_ctx->ep[ep_index].deq[0],
-				dev->out_ctx->ep[ep_index].deq[1]);
+		xhci_dbg(xhci, "Successful Set TR Deq Ptr cmd, deq = @%08llx\n",
+				ep_ctx->deq);
 	}
 
 	ep_ring->state &= ~SET_DEQ_PENDING;
 	ring_ep_doorbell(xhci, slot_id, ep_index);
 }
 
+static void handle_reset_ep_completion(struct xhci_hcd *xhci,
+		struct xhci_event_cmd *event,
+		union xhci_trb *trb)
+{
+	int slot_id;
+	unsigned int ep_index;
+
+	slot_id = TRB_TO_SLOT_ID(trb->generic.field[3]);
+	ep_index = TRB_TO_EP_INDEX(trb->generic.field[3]);
+	/* This command will only fail if the endpoint wasn't halted,
+	 * but we don't care.
+	 */
+	xhci_dbg(xhci, "Ignoring reset ep completion code of %u\n",
+			(unsigned int) GET_COMP_CODE(event->status));
+
+	/* Clear our internal halted state and restart the ring */
+	xhci->devs[slot_id]->ep_rings[ep_index]->state &= ~EP_HALTED;
+	ring_ep_doorbell(xhci, slot_id, ep_index);
+}
 
 static void handle_cmd_completion(struct xhci_hcd *xhci,
 		struct xhci_event_cmd *event)
@@ -611,7 +665,7 @@ static void handle_cmd_completion(struct xhci_hcd *xhci,
 	u64 cmd_dma;
 	dma_addr_t cmd_dequeue_dma;
 
-	cmd_dma = (((u64) event->cmd_trb[1]) << 32) + event->cmd_trb[0];
+	cmd_dma = event->cmd_trb;
 	cmd_dequeue_dma = xhci_trb_virt_to_dma(xhci->cmd_ring->deq_seg,
 			xhci->cmd_ring->dequeue);
 	/* Is the command ring deq ptr out of sync with the deq seg ptr? */
@@ -653,6 +707,9 @@ static void handle_cmd_completion(struct xhci_hcd *xhci,
 	case TRB_TYPE(TRB_CMD_NOOP):
 		++xhci->noops_handled;
 		break;
+	case TRB_TYPE(TRB_RESET_EP):
+		handle_reset_ep_completion(xhci, event, xhci->cmd_ring->dequeue);
+		break;
 	default:
 		/* Skip over unknown commands on the event ring */
 		xhci->error_bitmask |= 1 << 6;
@@ -756,7 +813,9 @@ static int handle_tx_event(struct xhci_hcd *xhci,
 	union xhci_trb *event_trb;
 	struct urb *urb = 0;
 	int status = -EINPROGRESS;
+	struct xhci_ep_ctx *ep_ctx;
 
+	xhci_dbg(xhci, "In %s\n", __func__);
 	xdev = xhci->devs[TRB_TO_SLOT_ID(event->flags)];
 	if (!xdev) {
 		xhci_err(xhci, "ERROR Transfer event pointed to bad slot\n");
@@ -765,17 +824,17 @@ static int handle_tx_event(struct xhci_hcd *xhci,
 
 	/* Endpoint ID is 1 based, our index is zero based */
 	ep_index = TRB_TO_EP_ID(event->flags) - 1;
+	xhci_dbg(xhci, "%s - ep index = %d\n", __func__, ep_index);
 	ep_ring = xdev->ep_rings[ep_index];
-	if (!ep_ring || (xdev->out_ctx->ep[ep_index].ep_info & EP_STATE_MASK) == EP_STATE_DISABLED) {
+	ep_ctx = xhci_get_ep_ctx(xhci, xdev->out_ctx, ep_index);
+	if (!ep_ring || (ep_ctx->ep_info & EP_STATE_MASK) == EP_STATE_DISABLED) {
 		xhci_err(xhci, "ERROR Transfer event pointed to disabled endpoint\n");
 		return -ENODEV;
 	}
 
-	event_dma = event->buffer[0];
-	if (event->buffer[1] != 0)
-		xhci_warn(xhci, "WARN ignoring upper 32-bits of 64-bit TRB dma address\n");
-
+	event_dma = event->buffer;
 	/* This TRB should be in the TD at the head of this ring's TD list */
+	xhci_dbg(xhci, "%s - checking for list empty\n", __func__);
 	if (list_empty(&ep_ring->td_list)) {
 		xhci_warn(xhci, "WARN Event TRB for slot %d ep %d with no TDs queued?\n",
 				TRB_TO_SLOT_ID(event->flags), ep_index);
@@ -785,11 +844,14 @@ static int handle_tx_event(struct xhci_hcd *xhci,
 		urb = NULL;
 		goto cleanup;
 	}
+	xhci_dbg(xhci, "%s - getting list entry\n", __func__);
 	td = list_entry(ep_ring->td_list.next, struct xhci_td, td_list);
 
 	/* Is this a TRB in the currently executing TD? */
+	xhci_dbg(xhci, "%s - looking for TD\n", __func__);
 	event_seg = trb_in_td(ep_ring->deq_seg, ep_ring->dequeue,
 			td->last_trb, event_dma);
+	xhci_dbg(xhci, "%s - found event_seg = %p\n", __func__, event_seg);
 	if (!event_seg) {
 		/* HC is busted, give up! */
 		xhci_err(xhci, "ERROR Transfer event TRB DMA ptr not part of current TD\n");
@@ -798,10 +860,10 @@ static int handle_tx_event(struct xhci_hcd *xhci,
 	event_trb = &event_seg->trbs[(event_dma - event_seg->dma) / sizeof(*event_trb)];
 	xhci_dbg(xhci, "Event TRB with TRB type ID %u\n",
 			(unsigned int) (event->flags & TRB_TYPE_BITMASK)>>10);
-	xhci_dbg(xhci, "Offset 0x00 (buffer[0]) = 0x%x\n",
-			(unsigned int) event->buffer[0]);
-	xhci_dbg(xhci, "Offset 0x04 (buffer[0]) = 0x%x\n",
-			(unsigned int) event->buffer[1]);
+	xhci_dbg(xhci, "Offset 0x00 (buffer lo) = 0x%x\n",
+			lower_32_bits(event->buffer));
+	xhci_dbg(xhci, "Offset 0x04 (buffer hi) = 0x%x\n",
+			upper_32_bits(event->buffer));
 	xhci_dbg(xhci, "Offset 0x08 (transfer length) = 0x%x\n",
 			(unsigned int) event->transfer_len);
 	xhci_dbg(xhci, "Offset 0x0C (flags) = 0x%x\n",
@@ -823,6 +885,7 @@ static int handle_tx_event(struct xhci_hcd *xhci,
 		break;
 	case COMP_STALL:
 		xhci_warn(xhci, "WARN: Stalled endpoint\n");
+		ep_ring->state |= EP_HALTED;
 		status = -EPIPE;
 		break;
 	case COMP_TRB_ERR:
@@ -833,6 +896,10 @@ static int handle_tx_event(struct xhci_hcd *xhci,
 		xhci_warn(xhci, "WARN: transfer error on endpoint\n");
 		status = -EPROTO;
 		break;
+	case COMP_BABBLE:
+		xhci_warn(xhci, "WARN: babble error on endpoint\n");
+		status = -EOVERFLOW;
+		break;
 	case COMP_DB_ERR:
 		xhci_warn(xhci, "WARN: HC couldn't access mem fast enough\n");
 		status = -ENOSR;
@@ -874,15 +941,26 @@ static int handle_tx_event(struct xhci_hcd *xhci,
 		if (event_trb != ep_ring->dequeue) {
 			/* The event was for the status stage */
 			if (event_trb == td->last_trb) {
-				td->urb->actual_length =
-					td->urb->transfer_buffer_length;
+				if (td->urb->actual_length != 0) {
+					/* Don't overwrite a previously set error code */
+					if (status == -EINPROGRESS || status == 0)
+						/* Did we already see a short data stage? */
+						status = -EREMOTEIO;
+				} else {
+					td->urb->actual_length =
+						td->urb->transfer_buffer_length;
+				}
 			} else {
 			/* Maybe the event was for the data stage? */
-				if (GET_COMP_CODE(event->transfer_len) != COMP_STOP_INVAL)
+				if (GET_COMP_CODE(event->transfer_len) != COMP_STOP_INVAL) {
 					/* We didn't stop on a link TRB in the middle */
 					td->urb->actual_length =
 						td->urb->transfer_buffer_length -
 						TRB_LEN(event->transfer_len);
+					xhci_dbg(xhci, "Waiting for status stage event\n");
+					urb = NULL;
+					goto cleanup;
+				}
 			}
 		}
 	} else {
@@ -929,16 +1007,20 @@ static int handle_tx_event(struct xhci_hcd *xhci,
 							TRB_LEN(event->transfer_len));
 					td->urb->actual_length = 0;
 				}
-				if (td->urb->transfer_flags & URB_SHORT_NOT_OK)
-					status = -EREMOTEIO;
-				else
-					status = 0;
+				/* Don't overwrite a previously set error code */
+				if (status == -EINPROGRESS) {
+					if (td->urb->transfer_flags & URB_SHORT_NOT_OK)
+						status = -EREMOTEIO;
+					else
+						status = 0;
+				}
 			} else {
 				td->urb->actual_length = td->urb->transfer_buffer_length;
 				/* Ignore a short packet completion if the
 				 * untransferred length was zero.
 				 */
-				status = 0;
+				if (status == -EREMOTEIO)
+					status = 0;
 			}
 		} else {
 			/* Slow path - walk the list, starting from the dequeue
@@ -965,19 +1047,30 @@ static int handle_tx_event(struct xhci_hcd *xhci,
 					TRB_LEN(event->transfer_len);
 		}
 	}
-	/* The Endpoint Stop Command completion will take care of
-	 * any stopped TDs.  A stopped TD may be restarted, so don't update the
-	 * ring dequeue pointer or take this TD off any lists yet.
-	 */
 	if (GET_COMP_CODE(event->transfer_len) == COMP_STOP_INVAL ||
 			GET_COMP_CODE(event->transfer_len) == COMP_STOP) {
+		/* The Endpoint Stop Command completion will take care of any
+		 * stopped TDs.  A stopped TD may be restarted, so don't update
+		 * the ring dequeue pointer or take this TD off any lists yet.
+		 */
 		ep_ring->stopped_td = td;
 		ep_ring->stopped_trb = event_trb;
 	} else {
-		/* Update ring dequeue pointer */
-		while (ep_ring->dequeue != td->last_trb)
+		if (GET_COMP_CODE(event->transfer_len) == COMP_STALL) {
+			/* The transfer is completed from the driver's
+			 * perspective, but we need to issue a set dequeue
+			 * command for this stalled endpoint to move the dequeue
+			 * pointer past the TD.  We can't do that here because
+			 * the halt condition must be cleared first.
+			 */
+			ep_ring->stopped_td = td;
+			ep_ring->stopped_trb = event_trb;
+		} else {
+			/* Update ring dequeue pointer */
+			while (ep_ring->dequeue != td->last_trb)
+				inc_deq(xhci, ep_ring, false);
 			inc_deq(xhci, ep_ring, false);
-		inc_deq(xhci, ep_ring, false);
+		}
 
 		/* Clean up the endpoint's TD list */
 		urb = td->urb;
@@ -987,7 +1080,10 @@ static int handle_tx_event(struct xhci_hcd *xhci,
 			list_del(&td->cancelled_td_list);
 			ep_ring->cancels_pending--;
 		}
-		kfree(td);
+		/* Leave the TD around for the reset endpoint function to use */
+		if (GET_COMP_CODE(event->transfer_len) != COMP_STALL) {
+			kfree(td);
+		}
 		urb->hcpriv = NULL;
 	}
 cleanup:
@@ -997,6 +1093,8 @@ cleanup:
 	/* FIXME for multi-TD URBs (who have buffers bigger than 64MB) */
 	if (urb) {
 		usb_hcd_unlink_urb_from_ep(xhci_to_hcd(xhci), urb);
+		xhci_dbg(xhci, "Giveback URB %p, len = %d, status = %d\n",
+				urb, td->urb->actual_length, status);
 		spin_unlock(&xhci->lock);
 		usb_hcd_giveback_urb(xhci_to_hcd(xhci), urb, status);
 		spin_lock(&xhci->lock);
@@ -1014,6 +1112,7 @@ void xhci_handle_event(struct xhci_hcd *xhci)
 	int update_ptrs = 1;
 	int ret;
 
+	xhci_dbg(xhci, "In %s\n", __func__);
 	if (!xhci->event_ring || !xhci->event_ring->dequeue) {
 		xhci->error_bitmask |= 1 << 1;
 		return;
@@ -1026,18 +1125,25 @@ void xhci_handle_event(struct xhci_hcd *xhci)
 		xhci->error_bitmask |= 1 << 2;
 		return;
 	}
+	xhci_dbg(xhci, "%s - OS owns TRB\n", __func__);
 
 	/* FIXME: Handle more event types. */
 	switch ((event->event_cmd.flags & TRB_TYPE_BITMASK)) {
 	case TRB_TYPE(TRB_COMPLETION):
+		xhci_dbg(xhci, "%s - calling handle_cmd_completion\n", __func__);
 		handle_cmd_completion(xhci, &event->event_cmd);
+		xhci_dbg(xhci, "%s - returned from handle_cmd_completion\n", __func__);
 		break;
 	case TRB_TYPE(TRB_PORT_STATUS):
+		xhci_dbg(xhci, "%s - calling handle_port_status\n", __func__);
 		handle_port_status(xhci, event);
+		xhci_dbg(xhci, "%s - returned from handle_port_status\n", __func__);
 		update_ptrs = 0;
 		break;
 	case TRB_TYPE(TRB_TRANSFER):
+		xhci_dbg(xhci, "%s - calling handle_tx_event\n", __func__);
 		ret = handle_tx_event(xhci, &event->trans_event);
+		xhci_dbg(xhci, "%s - returned from handle_tx_event\n", __func__);
 		if (ret < 0)
 			xhci->error_bitmask |= 1 << 9;
 		else
@@ -1093,13 +1199,13 @@ static int prepare_ring(struct xhci_hcd *xhci, struct xhci_ring *ep_ring,
 		 */
 		xhci_warn(xhci, "WARN urb submitted to disabled ep\n");
 		return -ENOENT;
-	case EP_STATE_HALTED:
 	case EP_STATE_ERROR:
-		xhci_warn(xhci, "WARN waiting for halt or error on ep "
-				"to be cleared\n");
+		xhci_warn(xhci, "WARN waiting for error on ep to be cleared\n");
 		/* FIXME event handling code for error needs to clear it */
 		/* XXX not sure if this should be -ENOENT or not */
 		return -EINVAL;
+	case EP_STATE_HALTED:
+		xhci_dbg(xhci, "WARN halted endpoint, queueing URB anyway.\n");
 	case EP_STATE_STOPPED:
 	case EP_STATE_RUNNING:
 		break;
@@ -1128,9 +1234,9 @@ static int prepare_transfer(struct xhci_hcd *xhci,
 		gfp_t mem_flags)
 {
 	int ret;
-
+	struct xhci_ep_ctx *ep_ctx = xhci_get_ep_ctx(xhci, xdev->out_ctx, ep_index);
 	ret = prepare_ring(xhci, xdev->ep_rings[ep_index],
-			xdev->out_ctx->ep[ep_index].ep_info & EP_STATE_MASK,
+			ep_ctx->ep_info & EP_STATE_MASK,
 			num_trbs, mem_flags);
 	if (ret)
 		return ret;
@@ -1285,6 +1391,7 @@ static int queue_bulk_sg_tx(struct xhci_hcd *xhci, gfp_t mem_flags,
 	/* Queue the first TRB, even if it's zero-length */
 	do {
 		u32 field = 0;
+		u32 length_field = 0;
 
 		/* Don't change the cycle bit of the first TRB until later */
 		if (first_trb)
@@ -1314,10 +1421,13 @@ static int queue_bulk_sg_tx(struct xhci_hcd *xhci, gfp_t mem_flags,
 					(unsigned int) (addr + TRB_MAX_BUFF_SIZE) & ~(TRB_MAX_BUFF_SIZE - 1),
 					(unsigned int) addr + trb_buff_len);
 		}
+		length_field = TRB_LEN(trb_buff_len) |
+			TD_REMAINDER(urb->transfer_buffer_length - running_total) |
+			TRB_INTR_TARGET(0);
 		queue_trb(xhci, ep_ring, false,
-				(u32) addr,
-				(u32) ((u64) addr >> 32),
-				TRB_LEN(trb_buff_len) | TRB_INTR_TARGET(0),
+				lower_32_bits(addr),
+				upper_32_bits(addr),
+				length_field,
 				/* We always want to know if the TRB was short,
 				 * or we won't get an event when it completes.
 				 * (Unless we use event data TRBs, which are a
@@ -1365,7 +1475,7 @@ int xhci_queue_bulk_tx(struct xhci_hcd *xhci, gfp_t mem_flags,
 	struct xhci_generic_trb *start_trb;
 	bool first_trb;
 	int start_cycle;
-	u32 field;
+	u32 field, length_field;
 
 	int running_total, trb_buff_len, ret;
 	u64 addr;
@@ -1443,10 +1553,13 @@ int xhci_queue_bulk_tx(struct xhci_hcd *xhci, gfp_t mem_flags,
 			td->last_trb = ep_ring->enqueue;
 			field |= TRB_IOC;
 		}
+		length_field = TRB_LEN(trb_buff_len) |
+			TD_REMAINDER(urb->transfer_buffer_length - running_total) |
+			TRB_INTR_TARGET(0);
 		queue_trb(xhci, ep_ring, false,
-				(u32) addr,
-				(u32) ((u64) addr >> 32),
-				TRB_LEN(trb_buff_len) | TRB_INTR_TARGET(0),
+				lower_32_bits(addr),
+				upper_32_bits(addr),
+				length_field,
 				/* We always want to know if the TRB was short,
 				 * or we won't get an event when it completes.
 				 * (Unless we use event data TRBs, which are a
@@ -1478,7 +1591,7 @@ int xhci_queue_ctrl_tx(struct xhci_hcd *xhci, gfp_t mem_flags,
 	struct usb_ctrlrequest *setup;
 	struct xhci_generic_trb *start_trb;
 	int start_cycle;
-	u32 field;
+	u32 field, length_field;
 	struct xhci_td *td;
 
 	ep_ring = xhci->devs[slot_id]->ep_rings[ep_index];
@@ -1528,13 +1641,16 @@ int xhci_queue_ctrl_tx(struct xhci_hcd *xhci, gfp_t mem_flags,
 
 	/* If there's data, queue data TRBs */
 	field = 0;
+	length_field = TRB_LEN(urb->transfer_buffer_length) |
+		TD_REMAINDER(urb->transfer_buffer_length) |
+		TRB_INTR_TARGET(0);
 	if (urb->transfer_buffer_length > 0) {
 		if (setup->bRequestType & USB_DIR_IN)
 			field |= TRB_DIR_IN;
 		queue_trb(xhci, ep_ring, false,
 				lower_32_bits(urb->transfer_dma),
 				upper_32_bits(urb->transfer_dma),
-				TRB_LEN(urb->transfer_buffer_length) | TRB_INTR_TARGET(0),
+				length_field,
 				/* Event on short tx */
 				field | TRB_ISP | TRB_TYPE(TRB_DATA) | ep_ring->cycle_state);
 	}
@@ -1603,7 +1719,8 @@ int xhci_queue_slot_control(struct xhci_hcd *xhci, u32 trb_type, u32 slot_id)
 int xhci_queue_address_device(struct xhci_hcd *xhci, dma_addr_t in_ctx_ptr,
 		u32 slot_id)
 {
-	return queue_command(xhci, in_ctx_ptr, 0, 0,
+	return queue_command(xhci, lower_32_bits(in_ctx_ptr),
+			upper_32_bits(in_ctx_ptr), 0,
 			TRB_TYPE(TRB_ADDR_DEV) | SLOT_ID_FOR_TRB(slot_id));
 }
 
@@ -1611,7 +1728,8 @@ int xhci_queue_address_device(struct xhci_hcd *xhci, dma_addr_t in_ctx_ptr,
 int xhci_queue_configure_endpoint(struct xhci_hcd *xhci, dma_addr_t in_ctx_ptr,
 		u32 slot_id)
 {
-	return queue_command(xhci, in_ctx_ptr, 0, 0,
+	return queue_command(xhci, lower_32_bits(in_ctx_ptr),
+			upper_32_bits(in_ctx_ptr), 0,
 			TRB_TYPE(TRB_CONFIG_EP) | SLOT_ID_FOR_TRB(slot_id));
 }
 
@@ -1639,10 +1757,23 @@ static int queue_set_tr_deq(struct xhci_hcd *xhci, int slot_id,
 	u32 type = TRB_TYPE(TRB_SET_DEQ);
 
 	addr = xhci_trb_virt_to_dma(deq_seg, deq_ptr);
-	if (addr == 0)
+	if (addr == 0) {
 		xhci_warn(xhci, "WARN Cannot submit Set TR Deq Ptr\n");
 		xhci_warn(xhci, "WARN deq seg = %p, deq pt = %p\n",
 				deq_seg, deq_ptr);
-	return queue_command(xhci, (u32) addr | cycle_state, 0, 0,
+		return 0;
+	}
+	return queue_command(xhci, lower_32_bits(addr) | cycle_state,
+			upper_32_bits(addr), 0,
 			trb_slot_id | trb_ep_index | type);
 }
+
+int xhci_queue_reset_ep(struct xhci_hcd *xhci, int slot_id,
+		unsigned int ep_index)
+{
+	u32 trb_slot_id = SLOT_ID_FOR_TRB(slot_id);
+	u32 trb_ep_index = EP_ID_FOR_TRB(ep_index);
+	u32 type = TRB_TYPE(TRB_RESET_EP);
+
+	return queue_command(xhci, 0, 0, 0, trb_slot_id | trb_ep_index | type);
+}
diff --git a/drivers/usb/host/xhci.h b/drivers/usb/host/xhci.h
index 8936eeb5588..d31d32206ba 100644
--- a/drivers/usb/host/xhci.h
+++ b/drivers/usb/host/xhci.h
@@ -25,6 +25,7 @@
 
 #include <linux/usb.h>
 #include <linux/timer.h>
+#include <linux/kernel.h>
 
 #include "../core/hcd.h"
 /* Code sharing between pci-quirks and xhci hcd */
@@ -42,14 +43,6 @@
  * xHCI register interface.
  * This corresponds to the eXtensible Host Controller Interface (xHCI)
  * Revision 0.95 specification
- *
- * Registers should always be accessed with double word or quad word accesses.
- *
- * Some xHCI implementations may support 64-bit address pointers.  Registers
- * with 64-bit address pointers should be written to with dword accesses by
- * writing the low dword first (ptr[0]), then the high dword (ptr[1]) second.
- * xHCI implementations that do not support 64-bit address pointers will ignore
- * the high dword, and write order is irrelevant.
  */
 
 /**
@@ -96,6 +89,7 @@ struct xhci_cap_regs {
 #define HCS_ERST_MAX(p)		(((p) >> 4) & 0xf)
 /* bit 26 Scratchpad restore - for save/restore HW state - not used yet */
 /* bits 27:31 number of Scratchpad buffers SW must allocate for the HW */
+#define HCS_MAX_SCRATCHPAD(p)   (((p) >> 27) & 0x1f)
 
 /* HCSPARAMS3 - hcs_params3 - bitmasks */
 /* bits 0:7, Max U1 to U0 latency for the roothub ports */
@@ -166,10 +160,10 @@ struct xhci_op_regs {
 	u32	reserved1;
 	u32	reserved2;
 	u32	dev_notification;
-	u32	cmd_ring[2];
+	u64	cmd_ring;
 	/* rsvd: offset 0x20-2F */
 	u32	reserved3[4];
-	u32	dcbaa_ptr[2];
+	u64	dcbaa_ptr;
 	u32	config_reg;
 	/* rsvd: offset 0x3C-3FF */
 	u32	reserved4[241];
@@ -254,7 +248,7 @@ struct xhci_op_regs {
 #define CMD_RING_RUNNING	(1 << 3)
 /* bits 4:5 reserved and should be preserved */
 /* Command Ring pointer - bit mask for the lower 32 bits. */
-#define CMD_RING_ADDR_MASK	(0xffffffc0)
+#define CMD_RING_RSVD_BITS	(0x3f)
 
 /* CONFIG - Configure Register - config_reg bitmasks */
 /* bits 0:7 - maximum number of device slots enabled (NumSlotsEn) */
@@ -382,8 +376,8 @@ struct xhci_intr_reg {
 	u32	irq_control;
 	u32	erst_size;
 	u32	rsvd;
-	u32	erst_base[2];
-	u32	erst_dequeue[2];
+	u64	erst_base;
+	u64	erst_dequeue;
 };
 
 /* irq_pending bitmasks */
@@ -453,6 +447,27 @@ struct xhci_doorbell_array {
 
 
 /**
+ * struct xhci_container_ctx
+ * @type: Type of context.  Used to calculated offsets to contained contexts.
+ * @size: Size of the context data
+ * @bytes: The raw context data given to HW
+ * @dma: dma address of the bytes
+ *
+ * Represents either a Device or Input context.  Holds a pointer to the raw
+ * memory used for the context (bytes) and dma address of it (dma).
+ */
+struct xhci_container_ctx {
+	unsigned type;
+#define XHCI_CTX_TYPE_DEVICE  0x1
+#define XHCI_CTX_TYPE_INPUT   0x2
+
+	int size;
+
+	u8 *bytes;
+	dma_addr_t dma;
+};
+
+/**
  * struct xhci_slot_ctx
  * @dev_info:	Route string, device speed, hub info, and last valid endpoint
  * @dev_info2:	Max exit latency for device number, root hub port number
@@ -538,7 +553,7 @@ struct xhci_slot_ctx {
 struct xhci_ep_ctx {
 	u32	ep_info;
 	u32	ep_info2;
-	u32	deq[2];
+	u64	deq;
 	u32	tx_info;
 	/* offset 0x14 - 0x1f reserved for HC internal use */
 	u32	reserved[3];
@@ -589,18 +604,16 @@ struct xhci_ep_ctx {
 
 
 /**
- * struct xhci_device_control
- * Input/Output context; see section 6.2.5.
+ * struct xhci_input_control_context
+ * Input control context; see section 6.2.5.
  *
  * @drop_context:	set the bit of the endpoint context you want to disable
  * @add_context:	set the bit of the endpoint context you want to enable
  */
-struct xhci_device_control {
+struct xhci_input_control_ctx {
 	u32	drop_flags;
 	u32	add_flags;
-	u32	rsvd[6];
-	struct xhci_slot_ctx	slot;
-	struct xhci_ep_ctx	ep[31];
+	u32	rsvd2[6];
 };
 
 /* drop context bitmasks */
@@ -608,7 +621,6 @@ struct xhci_device_control {
 /* add context bitmasks */
 #define	ADD_EP(x)	(0x1 << x)
 
-
 struct xhci_virt_device {
 	/*
 	 * Commands to the hardware are passed an "input context" that
@@ -618,11 +630,10 @@ struct xhci_virt_device {
 	 * track of input and output contexts separately because
 	 * these commands might fail and we don't trust the hardware.
 	 */
-	struct xhci_device_control	*out_ctx;
-	dma_addr_t			out_ctx_dma;
+	struct xhci_container_ctx       *out_ctx;
 	/* Used for addressing devices and configuration changes */
-	struct xhci_device_control	*in_ctx;
-	dma_addr_t			in_ctx_dma;
+	struct xhci_container_ctx       *in_ctx;
+
 	/* FIXME when stream support is added */
 	struct xhci_ring		*ep_rings[31];
 	/* Temporary storage in case the configure endpoint command fails and we
@@ -641,7 +652,7 @@ struct xhci_virt_device {
  */
 struct xhci_device_context_array {
 	/* 64-bit device addresses; we only write 32-bit addresses */
-	u32			dev_context_ptrs[2*MAX_HC_SLOTS];
+	u64			dev_context_ptrs[MAX_HC_SLOTS];
 	/* private xHCD pointers */
 	dma_addr_t	dma;
 };
@@ -654,7 +665,7 @@ struct xhci_device_context_array {
 
 struct xhci_stream_ctx {
 	/* 64-bit stream ring address, cycle state, and stream type */
-	u32	stream_ring[2];
+	u64	stream_ring;
 	/* offset 0x14 - 0x1f reserved for HC internal use */
 	u32	reserved[2];
 };
@@ -662,7 +673,7 @@ struct xhci_stream_ctx {
 
 struct xhci_transfer_event {
 	/* 64-bit buffer address, or immediate data */
-	u32	buffer[2];
+	u64	buffer;
 	u32	transfer_len;
 	/* This field is interpreted differently based on the type of TRB */
 	u32	flags;
@@ -744,7 +755,7 @@ struct xhci_transfer_event {
 
 struct xhci_link_trb {
 	/* 64-bit segment pointer*/
-	u32 segment_ptr[2];
+	u64 segment_ptr;
 	u32 intr_target;
 	u32 control;
 };
@@ -755,7 +766,7 @@ struct xhci_link_trb {
 /* Command completion event TRB */
 struct xhci_event_cmd {
 	/* Pointer to command TRB, or the value passed by the event data trb */
-	u32 cmd_trb[2];
+	u64 cmd_trb;
 	u32 status;
 	u32 flags;
 };
@@ -848,8 +859,8 @@ union xhci_trb {
 #define TRB_CONFIG_EP		12
 /* Evaluate Context Command */
 #define TRB_EVAL_CONTEXT	13
-/* Reset Transfer Ring Command */
-#define TRB_RESET_RING		14
+/* Reset Endpoint Command */
+#define TRB_RESET_EP		14
 /* Stop Transfer Ring Command */
 #define TRB_STOP_RING		15
 /* Set Transfer Ring Dequeue Pointer Command */
@@ -929,6 +940,7 @@ struct xhci_ring {
 	unsigned int		cancels_pending;
 	unsigned int		state;
 #define SET_DEQ_PENDING		(1 << 0)
+#define EP_HALTED		(1 << 1)
 	/* The TRB that was last reported in a stopped endpoint ring */
 	union xhci_trb		*stopped_trb;
 	struct xhci_td		*stopped_td;
@@ -940,9 +952,15 @@ struct xhci_ring {
 	u32			cycle_state;
 };
 
+struct xhci_dequeue_state {
+	struct xhci_segment *new_deq_seg;
+	union xhci_trb *new_deq_ptr;
+	int new_cycle_state;
+};
+
 struct xhci_erst_entry {
 	/* 64-bit event ring segment address */
-	u32	seg_addr[2];
+	u64	seg_addr;
 	u32	seg_size;
 	/* Set to zero */
 	u32	rsvd;
@@ -957,6 +975,13 @@ struct xhci_erst {
 	unsigned int		erst_size;
 };
 
+struct xhci_scratchpad {
+	u64 *sp_array;
+	dma_addr_t sp_dma;
+	void **sp_buffers;
+	dma_addr_t *sp_dma_buffers;
+};
+
 /*
  * Each segment table entry is 4*32bits long.  1K seems like an ok size:
  * (1K bytes * 8bytes/bit) / (4*32 bits) = 64 segment entries in the table,
@@ -1011,6 +1036,9 @@ struct xhci_hcd {
 	struct xhci_ring	*cmd_ring;
 	struct xhci_ring	*event_ring;
 	struct xhci_erst	erst;
+	/* Scratchpad */
+	struct xhci_scratchpad  *scratchpad;
+
 	/* slot enabling and address device helpers */
 	struct completion	addr_dev;
 	int slot_id;
@@ -1071,13 +1099,43 @@ static inline unsigned int xhci_readl(const struct xhci_hcd *xhci,
 static inline void xhci_writel(struct xhci_hcd *xhci,
 		const unsigned int val, __u32 __iomem *regs)
 {
-	if (!in_interrupt())
-		xhci_dbg(xhci,
-			 "`MEM_WRITE_DWORD(3'b000, 32'h%p, 32'h%0x, 4'hf);\n",
-			 regs, val);
+	xhci_dbg(xhci,
+			"`MEM_WRITE_DWORD(3'b000, 32'h%p, 32'h%0x, 4'hf);\n",
+			regs, val);
 	writel(val, regs);
 }
 
+/*
+ * Registers should always be accessed with double word or quad word accesses.
+ *
+ * Some xHCI implementations may support 64-bit address pointers.  Registers
+ * with 64-bit address pointers should be written to with dword accesses by
+ * writing the low dword first (ptr[0]), then the high dword (ptr[1]) second.
+ * xHCI implementations that do not support 64-bit address pointers will ignore
+ * the high dword, and write order is irrelevant.
+ */
+static inline u64 xhci_read_64(const struct xhci_hcd *xhci,
+		__u64 __iomem *regs)
+{
+	__u32 __iomem *ptr = (__u32 __iomem *) regs;
+	u64 val_lo = readl(ptr);
+	u64 val_hi = readl(ptr + 1);
+	return val_lo + (val_hi << 32);
+}
+static inline void xhci_write_64(struct xhci_hcd *xhci,
+		const u64 val, __u64 __iomem *regs)
+{
+	__u32 __iomem *ptr = (__u32 __iomem *) regs;
+	u32 val_lo = lower_32_bits(val);
+	u32 val_hi = upper_32_bits(val);
+
+	xhci_dbg(xhci,
+			"`MEM_WRITE_DWORD(3'b000, 64'h%p, 64'h%0lx, 4'hf);\n",
+			regs, (long unsigned int) val);
+	writel(val_lo, ptr);
+	writel(val_hi, ptr + 1);
+}
+
 /* xHCI debugging */
 void xhci_print_ir_set(struct xhci_hcd *xhci, struct xhci_intr_reg *ir_set, int set_num);
 void xhci_print_registers(struct xhci_hcd *xhci);
@@ -1090,7 +1148,7 @@ void xhci_debug_ring(struct xhci_hcd *xhci, struct xhci_ring *ring);
 void xhci_dbg_erst(struct xhci_hcd *xhci, struct xhci_erst *erst);
 void xhci_dbg_cmd_ptrs(struct xhci_hcd *xhci);
 void xhci_dbg_ring_ptrs(struct xhci_hcd *xhci, struct xhci_ring *ring);
-void xhci_dbg_ctx(struct xhci_hcd *xhci, struct xhci_device_control *ctx, dma_addr_t dma, unsigned int last_ep);
+void xhci_dbg_ctx(struct xhci_hcd *xhci, struct xhci_container_ctx *ctx, unsigned int last_ep);
 
 /* xHCI memory managment */
 void xhci_mem_cleanup(struct xhci_hcd *xhci);
@@ -1128,6 +1186,7 @@ int xhci_urb_enqueue(struct usb_hcd *hcd, struct urb *urb, gfp_t mem_flags);
 int xhci_urb_dequeue(struct usb_hcd *hcd, struct urb *urb, int status);
 int xhci_add_endpoint(struct usb_hcd *hcd, struct usb_device *udev, struct usb_host_endpoint *ep);
 int xhci_drop_endpoint(struct usb_hcd *hcd, struct usb_device *udev, struct usb_host_endpoint *ep);
+void xhci_endpoint_reset(struct usb_hcd *hcd, struct usb_host_endpoint *ep);
 int xhci_check_bandwidth(struct usb_hcd *hcd, struct usb_device *udev);
 void xhci_reset_bandwidth(struct usb_hcd *hcd, struct usb_device *udev);
 
@@ -1148,10 +1207,23 @@ int xhci_queue_bulk_tx(struct xhci_hcd *xhci, gfp_t mem_flags, struct urb *urb,
 		int slot_id, unsigned int ep_index);
 int xhci_queue_configure_endpoint(struct xhci_hcd *xhci, dma_addr_t in_ctx_ptr,
 		u32 slot_id);
+int xhci_queue_reset_ep(struct xhci_hcd *xhci, int slot_id,
+		unsigned int ep_index);
+void xhci_find_new_dequeue_state(struct xhci_hcd *xhci,
+		unsigned int slot_id, unsigned int ep_index,
+		struct xhci_td *cur_td, struct xhci_dequeue_state *state);
+void xhci_queue_new_dequeue_state(struct xhci_hcd *xhci,
+		struct xhci_ring *ep_ring, unsigned int slot_id,
+		unsigned int ep_index, struct xhci_dequeue_state *deq_state);
 
 /* xHCI roothub code */
 int xhci_hub_control(struct usb_hcd *hcd, u16 typeReq, u16 wValue, u16 wIndex,
 		char *buf, u16 wLength);
 int xhci_hub_status_data(struct usb_hcd *hcd, char *buf);
 
+/* xHCI contexts */
+struct xhci_input_control_ctx *xhci_get_input_control_ctx(struct xhci_hcd *xhci, struct xhci_container_ctx *ctx);
+struct xhci_slot_ctx *xhci_get_slot_ctx(struct xhci_hcd *xhci, struct xhci_container_ctx *ctx);
+struct xhci_ep_ctx *xhci_get_ep_ctx(struct xhci_hcd *xhci, struct xhci_container_ctx *ctx, unsigned int ep_index);
+
 #endif /* __LINUX_XHCI_HCD_H */
diff --git a/drivers/usb/misc/Kconfig b/drivers/usb/misc/Kconfig
index a68d91a11be..abe3aa67ed0 100644
--- a/drivers/usb/misc/Kconfig
+++ b/drivers/usb/misc/Kconfig
@@ -220,7 +220,7 @@ config USB_IOWARRIOR
 
 config USB_TEST
 	tristate "USB testing driver"
-	depends on USB && USB_DEVICEFS
+	depends on USB
 	help
 	  This driver is for testing host controller software.  It is used
 	  with specialized device firmware for regression and stress testing,
diff --git a/drivers/usb/musb/Kconfig b/drivers/usb/musb/Kconfig
index 70073b157f0..803adcb5ac1 100644
--- a/drivers/usb/musb/Kconfig
+++ b/drivers/usb/musb/Kconfig
@@ -12,6 +12,7 @@ config USB_MUSB_HDRC
 	depends on !SUPERH
 	select NOP_USB_XCEIV if ARCH_DAVINCI
 	select TWL4030_USB if MACH_OMAP_3430SDP
+	select NOP_USB_XCEIV if MACH_OMAP3EVM
 	select USB_OTG_UTILS
 	tristate 'Inventra Highspeed Dual Role Controller (TI, ADI, ...)'
 	help
diff --git a/drivers/usb/musb/musb_core.c b/drivers/usb/musb/musb_core.c
index 554a414f65d..c7c1ca0494c 100644
--- a/drivers/usb/musb/musb_core.c
+++ b/drivers/usb/musb/musb_core.c
@@ -1326,7 +1326,6 @@ static int __init musb_core_init(u16 musb_type, struct musb *musb)
 	int		i;
 
 	/* log core options (read using indexed model) */
-	musb_ep_select(mbase, 0);
 	reg = musb_read_configdata(mbase);
 
 	strcpy(aInfo, (reg & MUSB_CONFIGDATA_UTMIDW) ? "UTMI-16" : "UTMI-8");
@@ -1990,7 +1989,7 @@ bad_config:
 	if (status < 0)
 		goto fail2;
 
-#ifdef CONFIG_USB_OTG
+#ifdef CONFIG_USB_MUSB_OTG
 	setup_timer(&musb->otg_timer, musb_otg_timer_func, (unsigned long) musb);
 #endif
 
diff --git a/drivers/usb/musb/musb_gadget_ep0.c b/drivers/usb/musb/musb_gadget_ep0.c
index 40ed50ecedf..7a6778675ad 100644
--- a/drivers/usb/musb/musb_gadget_ep0.c
+++ b/drivers/usb/musb/musb_gadget_ep0.c
@@ -407,7 +407,7 @@ stall:
 					csr |= MUSB_RXCSR_P_SENDSTALL
 						| MUSB_RXCSR_FLUSHFIFO
 						| MUSB_RXCSR_CLRDATATOG
-						| MUSB_TXCSR_P_WZC_BITS;
+						| MUSB_RXCSR_P_WZC_BITS;
 					musb_writew(regs, MUSB_RXCSR,
 							csr);
 				}
diff --git a/drivers/usb/musb/musb_regs.h b/drivers/usb/musb/musb_regs.h
index de3b2f18db4..fbfd3fd9ce1 100644
--- a/drivers/usb/musb/musb_regs.h
+++ b/drivers/usb/musb/musb_regs.h
@@ -323,6 +323,7 @@ static inline void  musb_write_rxfifoadd(void __iomem *mbase, u16 c_off)
 
 static inline u8 musb_read_configdata(void __iomem *mbase)
 {
+	musb_writeb(mbase, MUSB_INDEX, 0);
 	return musb_readb(mbase, 0x10 + MUSB_CONFIGDATA);
 }
 
diff --git a/drivers/usb/serial/cp210x.c b/drivers/usb/serial/cp210x.c
index e9a40b820fd..985cbcf48bd 100644
--- a/drivers/usb/serial/cp210x.c
+++ b/drivers/usb/serial/cp210x.c
@@ -80,6 +80,7 @@ static struct usb_device_id id_table [] = {
 	{ USB_DEVICE(0x10C4, 0x80F6) }, /* Suunto sports instrument */
 	{ USB_DEVICE(0x10C4, 0x8115) }, /* Arygon NFC/Mifare Reader */
 	{ USB_DEVICE(0x10C4, 0x813D) }, /* Burnside Telecom Deskmobile */
+	{ USB_DEVICE(0x10C4, 0x813F) }, /* Tams Master Easy Control */
 	{ USB_DEVICE(0x10C4, 0x814A) }, /* West Mountain Radio RIGblaster P&P */
 	{ USB_DEVICE(0x10C4, 0x814B) }, /* West Mountain Radio RIGtalk */
 	{ USB_DEVICE(0x10C4, 0x815E) }, /* Helicomm IP-Link 1220-DVM */
@@ -96,7 +97,9 @@ static struct usb_device_id id_table [] = {
 	{ USB_DEVICE(0x10c4, 0x8293) }, /* Telegesys ETRX2USB */
 	{ USB_DEVICE(0x10C4, 0x82F9) }, /* Procyon AVS */
 	{ USB_DEVICE(0x10C4, 0x8341) }, /* Siemens MC35PU GPRS Modem */
+	{ USB_DEVICE(0x10C4, 0x8382) }, /* Cygnal Integrated Products, Inc. */
 	{ USB_DEVICE(0x10C4, 0x83A8) }, /* Amber Wireless AMB2560 */
+	{ USB_DEVICE(0x10C4, 0x8411) }, /* Kyocera GPS Module */
 	{ USB_DEVICE(0x10C4, 0x846E) }, /* BEI USB Sensor Interface (VCP) */
 	{ USB_DEVICE(0x10C4, 0xEA60) }, /* Silicon Labs factory default */
 	{ USB_DEVICE(0x10C4, 0xEA61) }, /* Silicon Labs factory default */
diff --git a/drivers/usb/serial/ftdi_sio.c b/drivers/usb/serial/ftdi_sio.c
index 60c64cc5be2..8fec5d4455c 100644
--- a/drivers/usb/serial/ftdi_sio.c
+++ b/drivers/usb/serial/ftdi_sio.c
@@ -698,6 +698,10 @@ static struct usb_device_id id_table_combined [] = {
 	{ USB_DEVICE(MARVELL_VID, MARVELL_SHEEVAPLUG_PID),
 		.driver_info = (kernel_ulong_t)&ftdi_jtag_quirk },
 	{ USB_DEVICE(LARSENBRUSGAARD_VID, LB_ALTITRACK_PID) },
+	{ USB_DEVICE(GN_OTOMETRICS_VID, AURICAL_USB_PID) },
+	{ USB_DEVICE(BAYER_VID, BAYER_CONTOUR_CABLE_PID) },
+	{ USB_DEVICE(FTDI_VID, MARVELL_OPENRD_PID),
+		.driver_info = (kernel_ulong_t)&ftdi_jtag_quirk },
 	{ },					/* Optional parameter entry */
 	{ }					/* Terminating entry */
 };
diff --git a/drivers/usb/serial/ftdi_sio.h b/drivers/usb/serial/ftdi_sio.h
index c9fbd741509..8c92b88166a 100644
--- a/drivers/usb/serial/ftdi_sio.h
+++ b/drivers/usb/serial/ftdi_sio.h
@@ -947,6 +947,27 @@
 #define FTDI_TURTELIZER_PID	0xBDC8 /* JTAG/RS-232 adapter by egnite GmBH */
 
 /*
+ * GN Otometrics (http://www.otometrics.com)
+ * Submitted by Ville Sundberg.
+ */
+#define GN_OTOMETRICS_VID	0x0c33	/* Vendor ID */
+#define AURICAL_USB_PID		0x0010	/* Aurical USB Audiometer */
+
+/*
+ * Bayer Ascensia Contour blood glucose meter USB-converter cable.
+ * http://winglucofacts.com/cables/
+ */
+#define BAYER_VID                      0x1A79
+#define BAYER_CONTOUR_CABLE_PID        0x6001
+
+/*
+ * Marvell OpenRD Base, Client
+ * http://www.open-rd.org
+ * OpenRD Base, Client use VID 0x0403
+ */
+#define MARVELL_OPENRD_PID	0x9e90
+
+/*
  *   BmRequestType:  1100 0000b
  *   bRequest:       FTDI_E2_READ
  *   wValue:         0
diff --git a/drivers/usb/serial/mos7840.c b/drivers/usb/serial/mos7840.c
index c31940a307f..270009afdf7 100644
--- a/drivers/usb/serial/mos7840.c
+++ b/drivers/usb/serial/mos7840.c
@@ -124,10 +124,13 @@
 #define BANDB_DEVICE_ID_USOPTL4_4       0xAC44
 #define BANDB_DEVICE_ID_USOPTL4_2       0xAC42
 
-/* This driver also supports the ATEN UC2324 device since it is mos7840 based
- *  - if I knew the device id it would also support the ATEN UC2322 */
+/* This driver also supports
+ * ATEN UC2324 device using Moschip MCS7840
+ * ATEN UC2322 device using Moschip MCS7820
+ */
 #define USB_VENDOR_ID_ATENINTL		0x0557
 #define ATENINTL_DEVICE_ID_UC2324	0x2011
+#define ATENINTL_DEVICE_ID_UC2322	0x7820
 
 /* Interrupt Routine Defines    */
 
@@ -177,6 +180,7 @@ static struct usb_device_id moschip_port_id_table[] = {
 	{USB_DEVICE(USB_VENDOR_ID_BANDB, BANDB_DEVICE_ID_USOPTL4_4)},
 	{USB_DEVICE(USB_VENDOR_ID_BANDB, BANDB_DEVICE_ID_USOPTL4_2)},
 	{USB_DEVICE(USB_VENDOR_ID_ATENINTL, ATENINTL_DEVICE_ID_UC2324)},
+	{USB_DEVICE(USB_VENDOR_ID_ATENINTL, ATENINTL_DEVICE_ID_UC2322)},
 	{}			/* terminating entry */
 };
 
@@ -186,6 +190,7 @@ static __devinitdata struct usb_device_id moschip_id_table_combined[] = {
 	{USB_DEVICE(USB_VENDOR_ID_BANDB, BANDB_DEVICE_ID_USOPTL4_4)},
 	{USB_DEVICE(USB_VENDOR_ID_BANDB, BANDB_DEVICE_ID_USOPTL4_2)},
 	{USB_DEVICE(USB_VENDOR_ID_ATENINTL, ATENINTL_DEVICE_ID_UC2324)},
+	{USB_DEVICE(USB_VENDOR_ID_ATENINTL, ATENINTL_DEVICE_ID_UC2322)},
 	{}			/* terminating entry */
 };
 
diff --git a/drivers/usb/serial/option.c b/drivers/usb/serial/option.c
index 98262dd552b..c784ddbe7b6 100644
--- a/drivers/usb/serial/option.c
+++ b/drivers/usb/serial/option.c
@@ -66,8 +66,10 @@ static int  option_tiocmget(struct tty_struct *tty, struct file *file);
 static int  option_tiocmset(struct tty_struct *tty, struct file *file,
 				unsigned int set, unsigned int clear);
 static int  option_send_setup(struct usb_serial_port *port);
+#ifdef CONFIG_PM
 static int  option_suspend(struct usb_serial *serial, pm_message_t message);
 static int  option_resume(struct usb_serial *serial);
+#endif
 
 /* Vendor and product IDs */
 #define OPTION_VENDOR_ID			0x0AF0
@@ -205,6 +207,7 @@ static int  option_resume(struct usb_serial *serial);
 #define NOVATELWIRELESS_PRODUCT_MC727		0x4100
 #define NOVATELWIRELESS_PRODUCT_MC950D		0x4400
 #define NOVATELWIRELESS_PRODUCT_U727		0x5010
+#define NOVATELWIRELESS_PRODUCT_MC727_NEW	0x5100
 #define NOVATELWIRELESS_PRODUCT_MC760		0x6000
 #define NOVATELWIRELESS_PRODUCT_OVMC760		0x6002
 
@@ -259,11 +262,6 @@ static int  option_resume(struct usb_serial *serial);
 #define AXESSTEL_VENDOR_ID			0x1726
 #define AXESSTEL_PRODUCT_MV110H			0x1000
 
-#define ONDA_VENDOR_ID				0x19d2
-#define ONDA_PRODUCT_MSA501HS			0x0001
-#define ONDA_PRODUCT_ET502HS			0x0002
-#define ONDA_PRODUCT_MT503HS			0x2000
-
 #define BANDRICH_VENDOR_ID			0x1A8D
 #define BANDRICH_PRODUCT_C100_1			0x1002
 #define BANDRICH_PRODUCT_C100_2			0x1003
@@ -301,6 +299,7 @@ static int  option_resume(struct usb_serial *serial);
 #define ZTE_PRODUCT_MF628			0x0015
 #define ZTE_PRODUCT_MF626			0x0031
 #define ZTE_PRODUCT_CDMA_TECH			0xfffe
+#define ZTE_PRODUCT_AC8710			0xfff1
 
 #define BENQ_VENDOR_ID				0x04a5
 #define BENQ_PRODUCT_H10			0x4068
@@ -322,6 +321,11 @@ static int  option_resume(struct usb_serial *serial);
 #define ALINK_VENDOR_ID				0x1e0e
 #define ALINK_PRODUCT_3GU			0x9200
 
+/* ALCATEL PRODUCTS */
+#define ALCATEL_VENDOR_ID			0x1bbb
+#define ALCATEL_PRODUCT_X060S			0x0000
+
+
 static struct usb_device_id option_ids[] = {
 	{ USB_DEVICE(OPTION_VENDOR_ID, OPTION_PRODUCT_COLT) },
 	{ USB_DEVICE(OPTION_VENDOR_ID, OPTION_PRODUCT_RICOLA) },
@@ -438,6 +442,7 @@ static struct usb_device_id option_ids[] = {
 	{ USB_DEVICE(NOVATELWIRELESS_VENDOR_ID, NOVATELWIRELESS_PRODUCT_EU870D) }, /* Novatel EU850D/EU860D/EU870D */
 	{ USB_DEVICE(NOVATELWIRELESS_VENDOR_ID, NOVATELWIRELESS_PRODUCT_MC950D) }, /* Novatel MC930D/MC950D */
 	{ USB_DEVICE(NOVATELWIRELESS_VENDOR_ID, NOVATELWIRELESS_PRODUCT_MC727) }, /* Novatel MC727/U727/USB727 */
+	{ USB_DEVICE(NOVATELWIRELESS_VENDOR_ID, NOVATELWIRELESS_PRODUCT_MC727_NEW) }, /* Novatel MC727/U727/USB727 refresh */
 	{ USB_DEVICE(NOVATELWIRELESS_VENDOR_ID, NOVATELWIRELESS_PRODUCT_U727) }, /* Novatel MC727/U727/USB727 */
 	{ USB_DEVICE(NOVATELWIRELESS_VENDOR_ID, NOVATELWIRELESS_PRODUCT_MC760) }, /* Novatel MC760/U760/USB760 */
 	{ USB_DEVICE(NOVATELWIRELESS_VENDOR_ID, NOVATELWIRELESS_PRODUCT_OVMC760) }, /* Novatel Ovation MC760 */
@@ -474,42 +479,6 @@ static struct usb_device_id option_ids[] = {
 	{ USB_DEVICE(ANYDATA_VENDOR_ID, ANYDATA_PRODUCT_ADU_500A) },
 	{ USB_DEVICE(ANYDATA_VENDOR_ID, ANYDATA_PRODUCT_ADU_620UW) },
 	{ USB_DEVICE(AXESSTEL_VENDOR_ID, AXESSTEL_PRODUCT_MV110H) },
-	{ USB_DEVICE(ONDA_VENDOR_ID, ONDA_PRODUCT_MSA501HS) },
-	{ USB_DEVICE(ONDA_VENDOR_ID, ONDA_PRODUCT_ET502HS) },
-	{ USB_DEVICE(ONDA_VENDOR_ID, 0x0003) },
-	{ USB_DEVICE(ONDA_VENDOR_ID, 0x0004) },
-	{ USB_DEVICE(ONDA_VENDOR_ID, 0x0005) },
-	{ USB_DEVICE(ONDA_VENDOR_ID, 0x0006) },
-	{ USB_DEVICE(ONDA_VENDOR_ID, 0x0007) },
-	{ USB_DEVICE(ONDA_VENDOR_ID, 0x0008) },
-	{ USB_DEVICE(ONDA_VENDOR_ID, 0x0009) },
-	{ USB_DEVICE(ONDA_VENDOR_ID, 0x000a) },
-	{ USB_DEVICE(ONDA_VENDOR_ID, 0x000b) },
-	{ USB_DEVICE(ONDA_VENDOR_ID, 0x000c) },
-	{ USB_DEVICE(ONDA_VENDOR_ID, 0x000d) },
-	{ USB_DEVICE(ONDA_VENDOR_ID, 0x000e) },
-	{ USB_DEVICE(ONDA_VENDOR_ID, 0x000f) },
-	{ USB_DEVICE(ONDA_VENDOR_ID, 0x0010) },
-	{ USB_DEVICE(ONDA_VENDOR_ID, 0x0011) },
-	{ USB_DEVICE(ONDA_VENDOR_ID, 0x0012) },
-	{ USB_DEVICE(ONDA_VENDOR_ID, 0x0013) },
-	{ USB_DEVICE(ONDA_VENDOR_ID, 0x0014) },
-	{ USB_DEVICE(ONDA_VENDOR_ID, 0x0015) },
-	{ USB_DEVICE(ONDA_VENDOR_ID, 0x0016) },
-	{ USB_DEVICE(ONDA_VENDOR_ID, 0x0017) },
-	{ USB_DEVICE(ONDA_VENDOR_ID, 0x0018) },
-	{ USB_DEVICE(ONDA_VENDOR_ID, 0x0019) },
-	{ USB_DEVICE(ONDA_VENDOR_ID, 0x0020) },
-	{ USB_DEVICE(ONDA_VENDOR_ID, 0x0021) },
-	{ USB_DEVICE(ONDA_VENDOR_ID, 0x0022) },
-	{ USB_DEVICE(ONDA_VENDOR_ID, 0x0023) },
-	{ USB_DEVICE(ONDA_VENDOR_ID, 0x0024) },
-	{ USB_DEVICE(ONDA_VENDOR_ID, 0x0025) },
-	{ USB_DEVICE(ONDA_VENDOR_ID, 0x0026) },
-	{ USB_DEVICE(ONDA_VENDOR_ID, 0x0027) },
-	{ USB_DEVICE(ONDA_VENDOR_ID, 0x0028) },
-	{ USB_DEVICE(ONDA_VENDOR_ID, 0x0029) },
-	{ USB_DEVICE(ONDA_VENDOR_ID, ONDA_PRODUCT_MT503HS) },
 	{ USB_DEVICE(YISO_VENDOR_ID, YISO_PRODUCT_U893) },
 	{ USB_DEVICE(BANDRICH_VENDOR_ID, BANDRICH_PRODUCT_C100_1) },
 	{ USB_DEVICE(BANDRICH_VENDOR_ID, BANDRICH_PRODUCT_C100_2) },
@@ -534,10 +503,75 @@ static struct usb_device_id option_ids[] = {
 	{ USB_DEVICE(QUALCOMM_VENDOR_ID, 0x6613)}, /* Onda H600/ZTE MF330 */
 	{ USB_DEVICE(MAXON_VENDOR_ID, 0x6280) }, /* BP3-USB & BP3-EXT HSDPA */
 	{ USB_DEVICE(TELIT_VENDOR_ID, TELIT_PRODUCT_UC864E) },
-	{ USB_DEVICE(ZTE_VENDOR_ID, ZTE_PRODUCT_MF622) },
-	{ USB_DEVICE(ZTE_VENDOR_ID, ZTE_PRODUCT_MF626) },
-	{ USB_DEVICE(ZTE_VENDOR_ID, ZTE_PRODUCT_MF628) },
-	{ USB_DEVICE(ZTE_VENDOR_ID, ZTE_PRODUCT_CDMA_TECH) },
+	{ USB_DEVICE_AND_INTERFACE_INFO(ZTE_VENDOR_ID, ZTE_PRODUCT_MF622, 0xff, 0xff, 0xff) }, /* ZTE WCDMA products */
+	{ USB_DEVICE_AND_INTERFACE_INFO(ZTE_VENDOR_ID, 0x0002, 0xff, 0xff, 0xff) },
+	{ USB_DEVICE_AND_INTERFACE_INFO(ZTE_VENDOR_ID, 0x0003, 0xff, 0xff, 0xff) },
+	{ USB_DEVICE_AND_INTERFACE_INFO(ZTE_VENDOR_ID, 0x0004, 0xff, 0xff, 0xff) },
+	{ USB_DEVICE_AND_INTERFACE_INFO(ZTE_VENDOR_ID, 0x0005, 0xff, 0xff, 0xff) },
+	{ USB_DEVICE_AND_INTERFACE_INFO(ZTE_VENDOR_ID, 0x0006, 0xff, 0xff, 0xff) },
+	{ USB_DEVICE_AND_INTERFACE_INFO(ZTE_VENDOR_ID, 0x0007, 0xff, 0xff, 0xff) },
+	{ USB_DEVICE_AND_INTERFACE_INFO(ZTE_VENDOR_ID, 0x0008, 0xff, 0xff, 0xff) },
+	{ USB_DEVICE_AND_INTERFACE_INFO(ZTE_VENDOR_ID, 0x0009, 0xff, 0xff, 0xff) },
+	{ USB_DEVICE_AND_INTERFACE_INFO(ZTE_VENDOR_ID, 0x000a, 0xff, 0xff, 0xff) },
+	{ USB_DEVICE_AND_INTERFACE_INFO(ZTE_VENDOR_ID, 0x000b, 0xff, 0xff, 0xff) },
+	{ USB_DEVICE_AND_INTERFACE_INFO(ZTE_VENDOR_ID, 0x000c, 0xff, 0xff, 0xff) },
+	{ USB_DEVICE_AND_INTERFACE_INFO(ZTE_VENDOR_ID, 0x000d, 0xff, 0xff, 0xff) },
+	{ USB_DEVICE_AND_INTERFACE_INFO(ZTE_VENDOR_ID, 0x000e, 0xff, 0xff, 0xff) },
+	{ USB_DEVICE_AND_INTERFACE_INFO(ZTE_VENDOR_ID, 0x000f, 0xff, 0xff, 0xff) },
+	{ USB_DEVICE_AND_INTERFACE_INFO(ZTE_VENDOR_ID, 0x0010, 0xff, 0xff, 0xff) },
+	{ USB_DEVICE_AND_INTERFACE_INFO(ZTE_VENDOR_ID, 0x0011, 0xff, 0xff, 0xff) },
+	{ USB_DEVICE_AND_INTERFACE_INFO(ZTE_VENDOR_ID, 0x0012, 0xff, 0xff, 0xff) },
+	{ USB_DEVICE_AND_INTERFACE_INFO(ZTE_VENDOR_ID, 0x0013, 0xff, 0xff, 0xff) },
+	{ USB_DEVICE_AND_INTERFACE_INFO(ZTE_VENDOR_ID, ZTE_PRODUCT_MF628, 0xff, 0xff, 0xff) },
+	{ USB_DEVICE_AND_INTERFACE_INFO(ZTE_VENDOR_ID, 0x0016, 0xff, 0xff, 0xff) },
+	{ USB_DEVICE_AND_INTERFACE_INFO(ZTE_VENDOR_ID, 0x0017, 0xff, 0xff, 0xff) },
+	{ USB_DEVICE_AND_INTERFACE_INFO(ZTE_VENDOR_ID, 0x0018, 0xff, 0xff, 0xff) },
+	{ USB_DEVICE_AND_INTERFACE_INFO(ZTE_VENDOR_ID, 0x0019, 0xff, 0xff, 0xff) },
+	{ USB_DEVICE_AND_INTERFACE_INFO(ZTE_VENDOR_ID, 0x0020, 0xff, 0xff, 0xff) },
+	{ USB_DEVICE_AND_INTERFACE_INFO(ZTE_VENDOR_ID, 0x0021, 0xff, 0xff, 0xff) },
+	{ USB_DEVICE_AND_INTERFACE_INFO(ZTE_VENDOR_ID, 0x0022, 0xff, 0xff, 0xff) },
+	{ USB_DEVICE_AND_INTERFACE_INFO(ZTE_VENDOR_ID, 0x0023, 0xff, 0xff, 0xff) },
+	{ USB_DEVICE_AND_INTERFACE_INFO(ZTE_VENDOR_ID, 0x0024, 0xff, 0xff, 0xff) },
+	{ USB_DEVICE_AND_INTERFACE_INFO(ZTE_VENDOR_ID, 0x0025, 0xff, 0xff, 0xff) },
+	{ USB_DEVICE_AND_INTERFACE_INFO(ZTE_VENDOR_ID, 0x0026, 0xff, 0xff, 0xff) },
+	{ USB_DEVICE_AND_INTERFACE_INFO(ZTE_VENDOR_ID, 0x0028, 0xff, 0xff, 0xff) },
+	{ USB_DEVICE_AND_INTERFACE_INFO(ZTE_VENDOR_ID, 0x0029, 0xff, 0xff, 0xff) },
+	{ USB_DEVICE_AND_INTERFACE_INFO(ZTE_VENDOR_ID, 0x0030, 0xff, 0xff, 0xff) },
+	{ USB_DEVICE_AND_INTERFACE_INFO(ZTE_VENDOR_ID, ZTE_PRODUCT_MF626, 0xff, 0xff, 0xff) },
+	{ USB_DEVICE_AND_INTERFACE_INFO(ZTE_VENDOR_ID, 0x0032, 0xff, 0xff, 0xff) },
+	{ USB_DEVICE_AND_INTERFACE_INFO(ZTE_VENDOR_ID, 0x0033, 0xff, 0xff, 0xff) },
+	{ USB_DEVICE_AND_INTERFACE_INFO(ZTE_VENDOR_ID, 0x0037, 0xff, 0xff, 0xff) },
+	{ USB_DEVICE_AND_INTERFACE_INFO(ZTE_VENDOR_ID, 0x0039, 0xff, 0xff, 0xff) },
+	{ USB_DEVICE_AND_INTERFACE_INFO(ZTE_VENDOR_ID, 0x0042, 0xff, 0xff, 0xff) },
+	{ USB_DEVICE_AND_INTERFACE_INFO(ZTE_VENDOR_ID, 0x0043, 0xff, 0xff, 0xff) },
+	{ USB_DEVICE_AND_INTERFACE_INFO(ZTE_VENDOR_ID, 0x0048, 0xff, 0xff, 0xff) },
+	{ USB_DEVICE_AND_INTERFACE_INFO(ZTE_VENDOR_ID, 0x0049, 0xff, 0xff, 0xff) },
+	{ USB_DEVICE_AND_INTERFACE_INFO(ZTE_VENDOR_ID, 0x0051, 0xff, 0xff, 0xff) },
+	{ USB_DEVICE_AND_INTERFACE_INFO(ZTE_VENDOR_ID, 0x0052, 0xff, 0xff, 0xff) },
+	{ USB_DEVICE_AND_INTERFACE_INFO(ZTE_VENDOR_ID, 0x0054, 0xff, 0xff, 0xff) },
+	{ USB_DEVICE_AND_INTERFACE_INFO(ZTE_VENDOR_ID, 0x0055, 0xff, 0xff, 0xff) },
+	{ USB_DEVICE_AND_INTERFACE_INFO(ZTE_VENDOR_ID, 0x0057, 0xff, 0xff, 0xff) },
+	{ USB_DEVICE_AND_INTERFACE_INFO(ZTE_VENDOR_ID, 0x0058, 0xff, 0xff, 0xff) },
+	{ USB_DEVICE_AND_INTERFACE_INFO(ZTE_VENDOR_ID, 0x0061, 0xff, 0xff, 0xff) },
+	{ USB_DEVICE_AND_INTERFACE_INFO(ZTE_VENDOR_ID, 0x0062, 0xff, 0xff, 0xff) },
+	{ USB_DEVICE_AND_INTERFACE_INFO(ZTE_VENDOR_ID, 0x0063, 0xff, 0xff, 0xff) },
+	{ USB_DEVICE_AND_INTERFACE_INFO(ZTE_VENDOR_ID, 0x0064, 0xff, 0xff, 0xff) },
+	{ USB_DEVICE_AND_INTERFACE_INFO(ZTE_VENDOR_ID, 0x0066, 0xff, 0xff, 0xff) },
+	{ USB_DEVICE_AND_INTERFACE_INFO(ZTE_VENDOR_ID, 0x0069, 0xff, 0xff, 0xff) },
+	{ USB_DEVICE_AND_INTERFACE_INFO(ZTE_VENDOR_ID, 0x0076, 0xff, 0xff, 0xff) },
+	{ USB_DEVICE_AND_INTERFACE_INFO(ZTE_VENDOR_ID, 0x0078, 0xff, 0xff, 0xff) },
+	{ USB_DEVICE_AND_INTERFACE_INFO(ZTE_VENDOR_ID, 0x0082, 0xff, 0xff, 0xff) },
+	{ USB_DEVICE_AND_INTERFACE_INFO(ZTE_VENDOR_ID, 0x0086, 0xff, 0xff, 0xff) },
+	{ USB_DEVICE_AND_INTERFACE_INFO(ZTE_VENDOR_ID, 0x2002, 0xff, 0xff, 0xff) },
+	{ USB_DEVICE_AND_INTERFACE_INFO(ZTE_VENDOR_ID, 0x2003, 0xff, 0xff, 0xff) },
+	{ USB_DEVICE_AND_INTERFACE_INFO(ZTE_VENDOR_ID, 0x0014, 0xff, 0xff, 0xff) }, /* ZTE CDMA products */
+	{ USB_DEVICE_AND_INTERFACE_INFO(ZTE_VENDOR_ID, 0x0027, 0xff, 0xff, 0xff) },
+	{ USB_DEVICE_AND_INTERFACE_INFO(ZTE_VENDOR_ID, 0x0059, 0xff, 0xff, 0xff) },
+	{ USB_DEVICE_AND_INTERFACE_INFO(ZTE_VENDOR_ID, 0x0060, 0xff, 0xff, 0xff) },
+	{ USB_DEVICE_AND_INTERFACE_INFO(ZTE_VENDOR_ID, 0x0070, 0xff, 0xff, 0xff) },
+	{ USB_DEVICE_AND_INTERFACE_INFO(ZTE_VENDOR_ID, 0x0073, 0xff, 0xff, 0xff) },
+	{ USB_DEVICE_AND_INTERFACE_INFO(ZTE_VENDOR_ID, ZTE_PRODUCT_CDMA_TECH, 0xff, 0xff, 0xff) },
+	{ USB_DEVICE_AND_INTERFACE_INFO(ZTE_VENDOR_ID, ZTE_PRODUCT_AC8710, 0xff, 0xff, 0xff) },
 	{ USB_DEVICE(BENQ_VENDOR_ID, BENQ_PRODUCT_H10) },
 	{ USB_DEVICE(DLINK_VENDOR_ID, DLINK_PRODUCT_DWM_652) },
 	{ USB_DEVICE(QISDA_VENDOR_ID, QISDA_PRODUCT_H21_4512) },
@@ -547,6 +581,7 @@ static struct usb_device_id option_ids[] = {
 	{ USB_DEVICE(TOSHIBA_VENDOR_ID, TOSHIBA_PRODUCT_HSDPA_MINICARD ) }, /* Toshiba 3G HSDPA == Novatel Expedite EU870D MiniCard */
 	{ USB_DEVICE(ALINK_VENDOR_ID, 0x9000) },
 	{ USB_DEVICE_AND_INTERFACE_INFO(ALINK_VENDOR_ID, ALINK_PRODUCT_3GU, 0xff, 0xff, 0xff) },
+	{ USB_DEVICE(ALCATEL_VENDOR_ID, ALCATEL_PRODUCT_X060S) },
 	{ } /* Terminating entry */
 };
 MODULE_DEVICE_TABLE(usb, option_ids);
@@ -555,8 +590,10 @@ static struct usb_driver option_driver = {
 	.name       = "option",
 	.probe      = usb_serial_probe,
 	.disconnect = usb_serial_disconnect,
+#ifdef CONFIG_PM
 	.suspend    = usb_serial_suspend,
 	.resume     = usb_serial_resume,
+#endif
 	.id_table   = option_ids,
 	.no_dynamic_id = 	1,
 };
@@ -588,8 +625,10 @@ static struct usb_serial_driver option_1port_device = {
 	.disconnect        = option_disconnect,
 	.release           = option_release,
 	.read_int_callback = option_instat_callback,
+#ifdef CONFIG_PM
 	.suspend           = option_suspend,
 	.resume            = option_resume,
+#endif
 };
 
 static int debug;
@@ -831,7 +870,6 @@ static void option_instat_callback(struct urb *urb)
 	int status = urb->status;
 	struct usb_serial_port *port =  urb->context;
 	struct option_port_private *portdata = usb_get_serial_port_data(port);
-	struct usb_serial *serial = port->serial;
 
 	dbg("%s", __func__);
 	dbg("%s: urb %p port %p has data %p", __func__, urb, port, portdata);
@@ -927,7 +965,6 @@ static int option_open(struct tty_struct *tty,
 			struct usb_serial_port *port, struct file *filp)
 {
 	struct option_port_private *portdata;
-	struct usb_serial *serial = port->serial;
 	int i, err;
 	struct urb *urb;
 
@@ -1187,6 +1224,7 @@ static void option_release(struct usb_serial *serial)
 	}
 }
 
+#ifdef CONFIG_PM
 static int option_suspend(struct usb_serial *serial, pm_message_t message)
 {
 	dbg("%s entered", __func__);
@@ -1245,6 +1283,7 @@ static int option_resume(struct usb_serial *serial)
 	}
 	return 0;
 }
+#endif
 
 MODULE_AUTHOR(DRIVER_AUTHOR);
 MODULE_DESCRIPTION(DRIVER_DESC);
diff --git a/drivers/usb/serial/pl2303.c b/drivers/usb/serial/pl2303.c
index 7d15bfa7c2d..3e86815b270 100644
--- a/drivers/usb/serial/pl2303.c
+++ b/drivers/usb/serial/pl2303.c
@@ -95,6 +95,7 @@ static struct usb_device_id id_table [] = {
 	{ USB_DEVICE(SUPERIAL_VENDOR_ID, SUPERIAL_PRODUCT_ID) },
 	{ USB_DEVICE(HP_VENDOR_ID, HP_LD220_PRODUCT_ID) },
 	{ USB_DEVICE(CRESSI_VENDOR_ID, CRESSI_EDY_PRODUCT_ID) },
+	{ USB_DEVICE(SONY_VENDOR_ID, SONY_QN3USB_PRODUCT_ID) },
 	{ }					/* Terminating entry */
 };
 
diff --git a/drivers/usb/serial/pl2303.h b/drivers/usb/serial/pl2303.h
index 12aac7d2462..ee9505e1dd9 100644
--- a/drivers/usb/serial/pl2303.h
+++ b/drivers/usb/serial/pl2303.h
@@ -126,3 +126,7 @@
 /* Cressi Edy (diving computer) PC interface */
 #define CRESSI_VENDOR_ID	0x04b8
 #define CRESSI_EDY_PRODUCT_ID	0x0521
+
+/* Sony, USB data cable for CMD-Jxx mobile phones */
+#define SONY_VENDOR_ID		0x054c
+#define SONY_QN3USB_PRODUCT_ID	0x0437
diff --git a/drivers/usb/storage/transport.c b/drivers/usb/storage/transport.c
index fcb32021721..e20dc525d17 100644
--- a/drivers/usb/storage/transport.c
+++ b/drivers/usb/storage/transport.c
@@ -961,7 +961,7 @@ int usb_stor_Bulk_max_lun(struct us_data *us)
 				 US_BULK_GET_MAX_LUN, 
 				 USB_DIR_IN | USB_TYPE_CLASS | 
 				 USB_RECIP_INTERFACE,
-				 0, us->ifnum, us->iobuf, 1, HZ);
+				 0, us->ifnum, us->iobuf, 1, 10*HZ);
 
 	US_DEBUGP("GetMaxLUN command result is %d, data is %d\n", 
 		  result, us->iobuf[0]);
diff --git a/drivers/usb/storage/unusual_devs.h b/drivers/usb/storage/unusual_devs.h
index 1b9c5dd0fb2..7477d411959 100644
--- a/drivers/usb/storage/unusual_devs.h
+++ b/drivers/usb/storage/unusual_devs.h
@@ -838,6 +838,13 @@ UNUSUAL_DEV( 0x066f, 0x8000, 0x0001, 0x0001,
 		US_SC_DEVICE, US_PR_DEVICE, NULL,
 		US_FL_FIX_CAPACITY ),
 
+/* Reported by Rogerio Brito <rbrito@ime.usp.br> */
+UNUSUAL_DEV( 0x067b, 0x2317, 0x0001, 0x001,
+		"Prolific Technology, Inc.",
+		"Mass Storage Device",
+		US_SC_DEVICE, US_PR_DEVICE, NULL,
+		US_FL_NOT_LOCKABLE ),
+
 /* Reported by Richard -=[]=- <micro_flyer@hotmail.com> */
 /* Change to bcdDeviceMin (0x0100 to 0x0001) reported by
  * Thomas Bartosik <tbartdev@gmx-topmail.de> */
diff --git a/drivers/video/backlight/jornada720_bl.c b/drivers/video/backlight/jornada720_bl.c
index c3ebb6b41ce..7aed2565c1b 100644
--- a/drivers/video/backlight/jornada720_bl.c
+++ b/drivers/video/backlight/jornada720_bl.c
@@ -72,7 +72,7 @@ static int jornada_bl_update_status(struct backlight_device *bd)
 		if (jornada_ssp_byte(SETBRIGHTNESS) != TXDUMMY) {
 			printk(KERN_INFO "bl : failed to set brightness\n");
 			ret = -ETIMEDOUT;
-			goto out
+			goto out;
 		}
 
 		/* at this point we expect that the mcu has accepted
diff --git a/drivers/video/console/fbcon.c b/drivers/video/console/fbcon.c
index 471a9a60376..3a44695b9c0 100644
--- a/drivers/video/console/fbcon.c
+++ b/drivers/video/console/fbcon.c
@@ -1082,7 +1082,6 @@ static void fbcon_init(struct vc_data *vc, int init)
 	new_rows = FBCON_SWAP(ops->rotate, info->var.yres, info->var.xres);
 	new_cols /= vc->vc_font.width;
 	new_rows /= vc->vc_font.height;
-	vc_resize(vc, new_cols, new_rows);
 
 	/*
 	 * We must always set the mode. The mode of the previous console
@@ -1111,10 +1110,11 @@ static void fbcon_init(struct vc_data *vc, int init)
 	 *  vc_{cols,rows}, but we must not set those if we are only
 	 *  resizing the console.
 	 */
-	if (!init) {
+	if (init) {
 		vc->vc_cols = new_cols;
 		vc->vc_rows = new_rows;
-	}
+	} else
+		vc_resize(vc, new_cols, new_rows);
 
 	if (logo)
 		fbcon_prepare_logo(vc, info, cols, rows, new_cols, new_rows);
diff --git a/drivers/video/console/fbcon_rotate.h b/drivers/video/console/fbcon_rotate.h
index 75be5ce53dc..e233444cda6 100644
--- a/drivers/video/console/fbcon_rotate.h
+++ b/drivers/video/console/fbcon_rotate.h
@@ -45,7 +45,7 @@ static inline void rotate_ud(const char *in, char *out, u32 width, u32 height)
 	width = (width + 7) & ~7;
 
 	for (i = 0; i < height; i++) {
-		for (j = 0; j < width; j++) {
+		for (j = 0; j < width - shift; j++) {
 			if (pattern_test_bit(j, i, width, in))
 				pattern_set_bit(width - (1 + j + shift),
 						height - (1 + i),
diff --git a/drivers/video/console/sticore.c b/drivers/video/console/sticore.c
index ef7870f5ea0..857b3668b3b 100644
--- a/drivers/video/console/sticore.c
+++ b/drivers/video/console/sticore.c
@@ -957,9 +957,14 @@ static int __devinit sticore_pci_init(struct pci_dev *pd,
 #ifdef CONFIG_PCI
 	unsigned long fb_base, rom_base;
 	unsigned int fb_len, rom_len;
+	int err;
 	struct sti_struct *sti;
 	
-	pci_enable_device(pd);
+	err = pci_enable_device(pd);
+	if (err < 0) {
+		dev_err(&pd->dev, "Cannot enable PCI device\n");
+		return err;
+	}
 
 	fb_base = pci_resource_start(pd, 0);
 	fb_len = pci_resource_len(pd, 0);
@@ -1048,7 +1053,7 @@ static void __devinit sti_init_roms(void)
 
 	/* Register drivers for native & PCI cards */
 	register_parisc_driver(&pa_sti_driver);
-	pci_register_driver(&pci_sti_driver);
+	WARN_ON(pci_register_driver(&pci_sti_driver));
 
 	/* if we didn't find the given default sti, take the first one */
 	if (!default_sti)
diff --git a/drivers/video/mx3fb.c b/drivers/video/mx3fb.c
index f8778cde218..054ef29be47 100644
--- a/drivers/video/mx3fb.c
+++ b/drivers/video/mx3fb.c
@@ -669,7 +669,8 @@ static uint32_t bpp_to_pixfmt(int bpp)
 }
 
 static int mx3fb_blank(int blank, struct fb_info *fbi);
-static int mx3fb_map_video_memory(struct fb_info *fbi, unsigned int mem_len);
+static int mx3fb_map_video_memory(struct fb_info *fbi, unsigned int mem_len,
+				  bool lock);
 static int mx3fb_unmap_video_memory(struct fb_info *fbi);
 
 /**
@@ -711,12 +712,7 @@ static void mx3fb_dma_done(void *arg)
 	complete(&mx3_fbi->flip_cmpl);
 }
 
-/**
- * mx3fb_set_par() - set framebuffer parameters and change the operating mode.
- * @fbi:	framebuffer information pointer.
- * @return:	0 on success or negative error code on failure.
- */
-static int mx3fb_set_par(struct fb_info *fbi)
+static int __set_par(struct fb_info *fbi, bool lock)
 {
 	u32 mem_len;
 	struct ipu_di_signal_cfg sig_cfg;
@@ -727,10 +723,6 @@ static int mx3fb_set_par(struct fb_info *fbi)
 	struct idmac_video_param *video = &ichan->params.video;
 	struct scatterlist *sg = mx3_fbi->sg;
 
-	dev_dbg(mx3fb->dev, "%s [%c]\n", __func__, list_empty(&ichan->queue) ? '-' : '+');
-
-	mutex_lock(&mx3_fbi->mutex);
-
 	/* Total cleanup */
 	if (mx3_fbi->txd)
 		sdc_disable_channel(mx3_fbi);
@@ -742,10 +734,8 @@ static int mx3fb_set_par(struct fb_info *fbi)
 		if (fbi->fix.smem_start)
 			mx3fb_unmap_video_memory(fbi);
 
-		if (mx3fb_map_video_memory(fbi, mem_len) < 0) {
-			mutex_unlock(&mx3_fbi->mutex);
+		if (mx3fb_map_video_memory(fbi, mem_len, lock) < 0)
 			return -ENOMEM;
-		}
 	}
 
 	sg_init_table(&sg[0], 1);
@@ -791,7 +781,6 @@ static int mx3fb_set_par(struct fb_info *fbi)
 				   fbi->var.vsync_len,
 				   fbi->var.lower_margin +
 				   fbi->var.vsync_len, sig_cfg) != 0) {
-			mutex_unlock(&mx3_fbi->mutex);
 			dev_err(fbi->device,
 				"mx3fb: Error initializing panel.\n");
 			return -EINVAL;
@@ -810,9 +799,30 @@ static int mx3fb_set_par(struct fb_info *fbi)
 	if (mx3_fbi->blank == FB_BLANK_UNBLANK)
 		sdc_enable_channel(mx3_fbi);
 
+	return 0;
+}
+
+/**
+ * mx3fb_set_par() - set framebuffer parameters and change the operating mode.
+ * @fbi:	framebuffer information pointer.
+ * @return:	0 on success or negative error code on failure.
+ */
+static int mx3fb_set_par(struct fb_info *fbi)
+{
+	struct mx3fb_info *mx3_fbi = fbi->par;
+	struct mx3fb_data *mx3fb = mx3_fbi->mx3fb;
+	struct idmac_channel *ichan = mx3_fbi->idmac_channel;
+	int ret;
+
+	dev_dbg(mx3fb->dev, "%s [%c]\n", __func__, list_empty(&ichan->queue) ? '-' : '+');
+
+	mutex_lock(&mx3_fbi->mutex);
+
+	ret = __set_par(fbi, true);
+
 	mutex_unlock(&mx3_fbi->mutex);
 
-	return 0;
+	return ret;
 }
 
 /**
@@ -966,21 +976,11 @@ static int mx3fb_setcolreg(unsigned int regno, unsigned int red,
 	return ret;
 }
 
-/**
- * mx3fb_blank() - blank the display.
- */
-static int mx3fb_blank(int blank, struct fb_info *fbi)
+static void __blank(int blank, struct fb_info *fbi)
 {
 	struct mx3fb_info *mx3_fbi = fbi->par;
 	struct mx3fb_data *mx3fb = mx3_fbi->mx3fb;
 
-	dev_dbg(fbi->device, "%s, blank = %d, base %p, len %u\n", __func__,
-		blank, fbi->screen_base, fbi->fix.smem_len);
-
-	if (mx3_fbi->blank == blank)
-		return 0;
-
-	mutex_lock(&mx3_fbi->mutex);
 	mx3_fbi->blank = blank;
 
 	switch (blank) {
@@ -999,6 +999,23 @@ static int mx3fb_blank(int blank, struct fb_info *fbi)
 		sdc_set_brightness(mx3fb, mx3fb->backlight_level);
 		break;
 	}
+}
+
+/**
+ * mx3fb_blank() - blank the display.
+ */
+static int mx3fb_blank(int blank, struct fb_info *fbi)
+{
+	struct mx3fb_info *mx3_fbi = fbi->par;
+
+	dev_dbg(fbi->device, "%s, blank = %d, base %p, len %u\n", __func__,
+		blank, fbi->screen_base, fbi->fix.smem_len);
+
+	if (mx3_fbi->blank == blank)
+		return 0;
+
+	mutex_lock(&mx3_fbi->mutex);
+	__blank(blank, fbi);
 	mutex_unlock(&mx3_fbi->mutex);
 
 	return 0;
@@ -1198,6 +1215,7 @@ static int mx3fb_resume(struct platform_device *pdev)
  * mx3fb_map_video_memory() - allocates the DRAM memory for the frame buffer.
  * @fbi:	framebuffer information pointer
  * @mem_len:	length of mapped memory
+ * @lock:	do not lock during initialisation
  * @return:	Error code indicating success or failure
  *
  * This buffer is remapped into a non-cached, non-buffered, memory region to
@@ -1205,7 +1223,8 @@ static int mx3fb_resume(struct platform_device *pdev)
  * area is remapped, all virtual memory access to the video memory should occur
  * at the new region.
  */
-static int mx3fb_map_video_memory(struct fb_info *fbi, unsigned int mem_len)
+static int mx3fb_map_video_memory(struct fb_info *fbi, unsigned int mem_len,
+				  bool lock)
 {
 	int retval = 0;
 	dma_addr_t addr;
@@ -1221,10 +1240,12 @@ static int mx3fb_map_video_memory(struct fb_info *fbi, unsigned int mem_len)
 		goto err0;
 	}
 
-	mutex_lock(&fbi->mm_lock);
+	if (lock)
+		mutex_lock(&fbi->mm_lock);
 	fbi->fix.smem_start = addr;
 	fbi->fix.smem_len = mem_len;
-	mutex_unlock(&fbi->mm_lock);
+	if (lock)
+		mutex_unlock(&fbi->mm_lock);
 
 	dev_dbg(fbi->device, "allocated fb @ p=0x%08x, v=0x%p, size=%d.\n",
 		(uint32_t) fbi->fix.smem_start, fbi->screen_base, fbi->fix.smem_len);
@@ -1365,6 +1386,11 @@ static int init_fb_chan(struct mx3fb_data *mx3fb, struct idmac_channel *ichan)
 	init_completion(&mx3fbi->flip_cmpl);
 	disable_irq(ichan->eof_irq);
 	dev_dbg(mx3fb->dev, "disabling irq %d\n", ichan->eof_irq);
+	ret = __set_par(fbi, false);
+	if (ret < 0)
+		goto esetpar;
+
+	__blank(FB_BLANK_UNBLANK, fbi);
 
 	dev_info(dev, "registered, using mode %s\n", fb_mode);
 
diff --git a/drivers/video/s3c-fb.c b/drivers/video/s3c-fb.c
index bb63c07e13d..5a72083dc67 100644
--- a/drivers/video/s3c-fb.c
+++ b/drivers/video/s3c-fb.c
@@ -964,7 +964,7 @@ static int __devexit s3c_fb_remove(struct platform_device *pdev)
 	struct s3c_fb *sfb = platform_get_drvdata(pdev);
 	int win;
 
-	for (win = 0; win <= S3C_FB_MAX_WIN; win++)
+	for (win = 0; win < S3C_FB_MAX_WIN; win++)
 		if (sfb->windows[win])
 			s3c_fb_release_win(sfb, sfb->windows[win]);
 
@@ -988,7 +988,7 @@ static int s3c_fb_suspend(struct platform_device *pdev, pm_message_t state)
 	struct s3c_fb_win *win;
 	int win_no;
 
-	for (win_no = S3C_FB_MAX_WIN; win_no >= 0; win_no--) {
+	for (win_no = S3C_FB_MAX_WIN - 1; win_no >= 0; win_no--) {
 		win = sfb->windows[win_no];
 		if (!win)
 			continue;
diff --git a/drivers/video/via/hw.c b/drivers/video/via/hw.c
index fcd53ceb88f..c8960003f47 100644
--- a/drivers/video/via/hw.c
+++ b/drivers/video/via/hw.c
@@ -2407,14 +2407,14 @@ int viafb_setmode(int vmode_index, int hor_res, int ver_res, int video_bpp,
 			viafb_dvi_set_mode(viafb_get_mode_index
 				     (viaparinfo->tmds_setting_info->h_active,
 				      viaparinfo->tmds_setting_info->
-				      v_active, 1),
+				      v_active),
 				     video_bpp1, viaparinfo->
 				     tmds_setting_info->iga_path);
 		} else {
 			viafb_dvi_set_mode(viafb_get_mode_index
 				     (viaparinfo->tmds_setting_info->h_active,
 				      viaparinfo->
-				      tmds_setting_info->v_active, 0),
+				      tmds_setting_info->v_active),
 				     video_bpp, viaparinfo->
 				     tmds_setting_info->iga_path);
 		}
diff --git a/drivers/video/via/lcd.c b/drivers/video/via/lcd.c
index 6c7290a6a44..78c6b338794 100644
--- a/drivers/video/via/lcd.c
+++ b/drivers/video/via/lcd.c
@@ -580,10 +580,7 @@ static void load_lcd_k400_patch_tbl(int set_hres, int set_vres,
 	int reg_num = 0;
 	struct io_reg *lcd_patch_reg = NULL;
 
-	if (viaparinfo->lvds_setting_info->iga_path == IGA2)
-		vmode_index = viafb_get_mode_index(set_hres, set_vres, 1);
-	else
-		vmode_index = viafb_get_mode_index(set_hres, set_vres, 0);
+	vmode_index = viafb_get_mode_index(set_hres, set_vres);
 	switch (panel_id) {
 		/* LCD 800x600 */
 	case LCD_PANEL_ID1_800X600:
@@ -761,10 +758,7 @@ static void load_lcd_p880_patch_tbl(int set_hres, int set_vres,
 	int reg_num = 0;
 	struct io_reg *lcd_patch_reg = NULL;
 
-	if (viaparinfo->lvds_setting_info->iga_path == IGA2)
-		vmode_index = viafb_get_mode_index(set_hres, set_vres, 1);
-	else
-		vmode_index = viafb_get_mode_index(set_hres, set_vres, 0);
+	vmode_index = viafb_get_mode_index(set_hres, set_vres);
 
 	switch (panel_id) {
 	case LCD_PANEL_ID5_1400X1050:
@@ -832,10 +826,7 @@ static void load_lcd_patch_regs(int set_hres, int set_vres,
 {
 	int vmode_index;
 
-	if (viaparinfo->lvds_setting_info->iga_path == IGA2)
-		vmode_index = viafb_get_mode_index(set_hres, set_vres, 1);
-	else
-		vmode_index = viafb_get_mode_index(set_hres, set_vres, 0);
+	vmode_index = viafb_get_mode_index(set_hres, set_vres);
 
 	viafb_unlock_crt();
 
diff --git a/drivers/video/via/viafbdev.c b/drivers/video/via/viafbdev.c
index a0fec298216..72833f3334b 100644
--- a/drivers/video/via/viafbdev.c
+++ b/drivers/video/via/viafbdev.c
@@ -32,7 +32,6 @@ static u32 pseudo_pal[17];
 /* video mode */
 static char *viafb_mode = "640x480";
 static char *viafb_mode1 = "640x480";
-static int viafb_resMode = VIA_RES_640X480;
 
 /* Added for specifying active devices.*/
 char *viafb_active_dev = "";
@@ -56,47 +55,47 @@ static void viafb_get_video_device(u32 *video_dev_info);
 
 /* Mode information */
 static const struct viafb_modeinfo viafb_modentry[] = {
-	{480, 640, VIA_RES_480X640, "480x640"},
-	{640, 480, VIA_RES_640X480, "640x480"},
-	{800, 480, VIA_RES_800X480, "800x480"},
-	{800, 600, VIA_RES_800X600, "800x600"},
-	{1024, 768, VIA_RES_1024X768, "1024x768"},
-	{1152, 864, VIA_RES_1152X864, "1152x864"},
-	{1280, 1024, VIA_RES_1280X1024, "1280x1024"},
-	{1600, 1200, VIA_RES_1600X1200, "1600x1200"},
-	{1440, 1050, VIA_RES_1440X1050, "1440x1050"},
-	{1280, 768, VIA_RES_1280X768, "1280x768"},
-	{1280, 800, VIA_RES_1280X800, "1280x800"},
-	{1280, 960, VIA_RES_1280X960, "1280x960"},
-	{1920, 1440, VIA_RES_1920X1440, "1920x1440"},
-	{848, 480, VIA_RES_848X480, "848x480"},
-	{1400, 1050, VIA_RES_1400X1050, "1400x1050"},
-	{720, 480, VIA_RES_720X480, "720x480"},
-	{720, 576, VIA_RES_720X576, "720x576"},
-	{1024, 512, VIA_RES_1024X512, "1024x512"},
-	{1024, 576, VIA_RES_1024X576, "1024x576"},
-	{1024, 600, VIA_RES_1024X600, "1024x600"},
-	{1280, 720, VIA_RES_1280X720, "1280x720"},
-	{1920, 1080, VIA_RES_1920X1080, "1920x1080"},
-	{1366, 768, VIA_RES_1368X768, "1368x768"},
-	{1680, 1050, VIA_RES_1680X1050, "1680x1050"},
-	{960, 600, VIA_RES_960X600, "960x600"},
-	{1000, 600, VIA_RES_1000X600, "1000x600"},
-	{1024, 576, VIA_RES_1024X576, "1024x576"},
-	{1024, 600, VIA_RES_1024X600, "1024x600"},
-	{1088, 612, VIA_RES_1088X612, "1088x612"},
-	{1152, 720, VIA_RES_1152X720, "1152x720"},
-	{1200, 720, VIA_RES_1200X720, "1200x720"},
-	{1280, 600, VIA_RES_1280X600, "1280x600"},
-	{1360, 768, VIA_RES_1360X768, "1360x768"},
-	{1440, 900, VIA_RES_1440X900, "1440x900"},
-	{1600, 900, VIA_RES_1600X900, "1600x900"},
-	{1600, 1024, VIA_RES_1600X1024, "1600x1024"},
-	{1792, 1344, VIA_RES_1792X1344, "1792x1344"},
-	{1856, 1392, VIA_RES_1856X1392, "1856x1392"},
-	{1920, 1200, VIA_RES_1920X1200, "1920x1200"},
-	{2048, 1536, VIA_RES_2048X1536, "2048x1536"},
-	{0, 0, VIA_RES_INVALID, "640x480"}
+	{480, 640, VIA_RES_480X640},
+	{640, 480, VIA_RES_640X480},
+	{800, 480, VIA_RES_800X480},
+	{800, 600, VIA_RES_800X600},
+	{1024, 768, VIA_RES_1024X768},
+	{1152, 864, VIA_RES_1152X864},
+	{1280, 1024, VIA_RES_1280X1024},
+	{1600, 1200, VIA_RES_1600X1200},
+	{1440, 1050, VIA_RES_1440X1050},
+	{1280, 768, VIA_RES_1280X768,},
+	{1280, 800, VIA_RES_1280X800},
+	{1280, 960, VIA_RES_1280X960},
+	{1920, 1440, VIA_RES_1920X1440},
+	{848, 480, VIA_RES_848X480},
+	{1400, 1050, VIA_RES_1400X1050},
+	{720, 480, VIA_RES_720X480},
+	{720, 576, VIA_RES_720X576},
+	{1024, 512, VIA_RES_1024X512},
+	{1024, 576, VIA_RES_1024X576},
+	{1024, 600, VIA_RES_1024X600},
+	{1280, 720, VIA_RES_1280X720},
+	{1920, 1080, VIA_RES_1920X1080},
+	{1366, 768, VIA_RES_1368X768},
+	{1680, 1050, VIA_RES_1680X1050},
+	{960, 600, VIA_RES_960X600},
+	{1000, 600, VIA_RES_1000X600},
+	{1024, 576, VIA_RES_1024X576},
+	{1024, 600, VIA_RES_1024X600},
+	{1088, 612, VIA_RES_1088X612},
+	{1152, 720, VIA_RES_1152X720},
+	{1200, 720, VIA_RES_1200X720},
+	{1280, 600, VIA_RES_1280X600},
+	{1360, 768, VIA_RES_1360X768},
+	{1440, 900, VIA_RES_1440X900},
+	{1600, 900, VIA_RES_1600X900},
+	{1600, 1024, VIA_RES_1600X1024},
+	{1792, 1344, VIA_RES_1792X1344},
+	{1856, 1392, VIA_RES_1856X1392},
+	{1920, 1200, VIA_RES_1920X1200},
+	{2048, 1536, VIA_RES_2048X1536},
+	{0, 0, VIA_RES_INVALID}
 };
 
 static struct fb_ops viafb_ops;
@@ -177,7 +176,7 @@ static int viafb_check_var(struct fb_var_screeninfo *var,
 	if (var->vmode & FB_VMODE_INTERLACED || var->vmode & FB_VMODE_DOUBLE)
 		return -EINVAL;
 
-	vmode_index = viafb_get_mode_index(var->xres, var->yres, 0);
+	vmode_index = viafb_get_mode_index(var->xres, var->yres);
 	if (vmode_index == VIA_RES_INVALID) {
 		DEBUG_MSG(KERN_INFO
 			  "viafb: Mode %dx%dx%d not supported!!\n",
@@ -233,14 +232,14 @@ static int viafb_set_par(struct fb_info *info)
 	viafb_update_device_setting(info->var.xres, info->var.yres,
 			      info->var.bits_per_pixel, viafb_refresh, 0);
 
-	vmode_index = viafb_get_mode_index(info->var.xres, info->var.yres, 0);
+	vmode_index = viafb_get_mode_index(info->var.xres, info->var.yres);
 
 	if (viafb_SAMM_ON == 1) {
 		DEBUG_MSG(KERN_INFO
 		"viafb_second_xres = %d, viafb_second_yres = %d, bpp = %d\n",
 			  viafb_second_xres, viafb_second_yres, viafb_bpp1);
 		vmode_index1 = viafb_get_mode_index(viafb_second_xres,
-			viafb_second_yres, 1);
+			viafb_second_yres);
 		DEBUG_MSG(KERN_INFO "->viafb_SAMM_ON: index=%d\n",
 			vmode_index1);
 
@@ -1262,7 +1261,7 @@ static int viafb_sync(struct fb_info *info)
 	return 0;
 }
 
-int viafb_get_mode_index(int hres, int vres, int flag)
+int viafb_get_mode_index(int hres, int vres)
 {
 	u32 i;
 	DEBUG_MSG(KERN_INFO "viafb_get_mode_index!\n");
@@ -1272,13 +1271,7 @@ int viafb_get_mode_index(int hres, int vres, int flag)
 			viafb_modentry[i].yres == vres)
 			break;
 
-	viafb_resMode = viafb_modentry[i].mode_index;
-	if (flag)
-		viafb_mode1 = viafb_modentry[i].mode_res;
-	else
-		viafb_mode = viafb_modentry[i].mode_res;
-
-	return viafb_resMode;
+	return viafb_modentry[i].mode_index;
 }
 
 static void check_available_device_to_enable(int device_id)
@@ -2199,7 +2192,7 @@ static int __devinit via_pci_probe(void)
 	strict_strtoul(tmpc, 0, &default_xres);
 	strict_strtoul(tmpm, 0, &default_yres);
 
-	vmode_index = viafb_get_mode_index(default_xres, default_yres, 0);
+	vmode_index = viafb_get_mode_index(default_xres, default_yres);
 	DEBUG_MSG(KERN_INFO "0->index=%d\n", vmode_index);
 
 	if (viafb_SAMM_ON == 1) {
diff --git a/drivers/video/via/viafbdev.h b/drivers/video/via/viafbdev.h
index a4158e87287..227b000feb3 100644
--- a/drivers/video/via/viafbdev.h
+++ b/drivers/video/via/viafbdev.h
@@ -81,7 +81,6 @@ struct viafb_modeinfo {
 	u32 xres;
 	u32 yres;
 	int mode_index;
-	char *mode_res;
 };
 extern unsigned int viafb_second_virtual_yres;
 extern unsigned int viafb_second_virtual_xres;
@@ -102,7 +101,7 @@ extern int strict_strtoul(const char *cp, unsigned int base,
 void viafb_memory_pitch_patch(struct fb_info *info);
 void viafb_fill_var_timing_info(struct fb_var_screeninfo *var, int refresh,
 			  int mode_index);
-int viafb_get_mode_index(int hres, int vres, int flag);
+int viafb_get_mode_index(int hres, int vres);
 u8 viafb_gpio_i2c_read_lvds(struct lvds_setting_information
 	*plvds_setting_info, struct lvds_chip_information
 	*plvds_chip_info, u8 index);
diff --git a/drivers/virtio/virtio_pci.c b/drivers/virtio/virtio_pci.c
index bcec78ffc76..248e00ec4dc 100644
--- a/drivers/virtio/virtio_pci.c
+++ b/drivers/virtio/virtio_pci.c
@@ -52,8 +52,10 @@ struct virtio_pci_device
 	char (*msix_names)[256];
 	/* Number of available vectors */
 	unsigned msix_vectors;
-	/* Vectors allocated */
+	/* Vectors allocated, excluding per-vq vectors if any */
 	unsigned msix_used_vectors;
+	/* Whether we have vector per vq */
+	bool per_vq_vectors;
 };
 
 /* Constants for MSI-X */
@@ -258,7 +260,6 @@ static void vp_free_vectors(struct virtio_device *vdev)
 
 	for (i = 0; i < vp_dev->msix_used_vectors; ++i)
 		free_irq(vp_dev->msix_entries[i].vector, vp_dev);
-	vp_dev->msix_used_vectors = 0;
 
 	if (vp_dev->msix_enabled) {
 		/* Disable the vector used for configuration */
@@ -267,80 +268,77 @@ static void vp_free_vectors(struct virtio_device *vdev)
 		/* Flush the write out to device */
 		ioread16(vp_dev->ioaddr + VIRTIO_MSI_CONFIG_VECTOR);
 
-		vp_dev->msix_enabled = 0;
 		pci_disable_msix(vp_dev->pci_dev);
+		vp_dev->msix_enabled = 0;
+		vp_dev->msix_vectors = 0;
 	}
-}
 
-static int vp_enable_msix(struct pci_dev *dev, struct msix_entry *entries,
-			  int *options, int noptions)
-{
-	int i;
-	for (i = 0; i < noptions; ++i)
-		if (!pci_enable_msix(dev, entries, options[i]))
-			return options[i];
-	return -EBUSY;
+	vp_dev->msix_used_vectors = 0;
+	kfree(vp_dev->msix_names);
+	vp_dev->msix_names = NULL;
+	kfree(vp_dev->msix_entries);
+	vp_dev->msix_entries = NULL;
 }
 
-static int vp_request_vectors(struct virtio_device *vdev, unsigned max_vqs)
+static int vp_request_vectors(struct virtio_device *vdev, int nvectors,
+			      bool per_vq_vectors)
 {
 	struct virtio_pci_device *vp_dev = to_vp_device(vdev);
 	const char *name = dev_name(&vp_dev->vdev.dev);
 	unsigned i, v;
 	int err = -ENOMEM;
-	/* We want at most one vector per queue and one for config changes.
-	 * Fallback to separate vectors for config and a shared for queues.
-	 * Finally fall back to regular interrupts. */
-	int options[] = { max_vqs + 1, 2 };
-	int nvectors = max(options[0], options[1]);
+
+	if (!nvectors) {
+		/* Can't allocate MSI-X vectors, use regular interrupt */
+		vp_dev->msix_vectors = 0;
+		err = request_irq(vp_dev->pci_dev->irq, vp_interrupt,
+				  IRQF_SHARED, name, vp_dev);
+		if (err)
+			return err;
+		vp_dev->intx_enabled = 1;
+		return 0;
+	}
 
 	vp_dev->msix_entries = kmalloc(nvectors * sizeof *vp_dev->msix_entries,
 				       GFP_KERNEL);
 	if (!vp_dev->msix_entries)
-		goto error_entries;
+		goto error;
 	vp_dev->msix_names = kmalloc(nvectors * sizeof *vp_dev->msix_names,
 				     GFP_KERNEL);
 	if (!vp_dev->msix_names)
-		goto error_names;
+		goto error;
 
 	for (i = 0; i < nvectors; ++i)
 		vp_dev->msix_entries[i].entry = i;
 
-	err = vp_enable_msix(vp_dev->pci_dev, vp_dev->msix_entries,
-			     options, ARRAY_SIZE(options));
-	if (err < 0) {
-		/* Can't allocate enough MSI-X vectors, use regular interrupt */
-		vp_dev->msix_vectors = 0;
-		err = request_irq(vp_dev->pci_dev->irq, vp_interrupt,
-				  IRQF_SHARED, name, vp_dev);
-		if (err)
-			goto error_irq;
-		vp_dev->intx_enabled = 1;
-	} else {
-		vp_dev->msix_vectors = err;
-		vp_dev->msix_enabled = 1;
-
-		/* Set the vector used for configuration */
-		v = vp_dev->msix_used_vectors;
-		snprintf(vp_dev->msix_names[v], sizeof *vp_dev->msix_names,
-			 "%s-config", name);
-		err = request_irq(vp_dev->msix_entries[v].vector,
-				  vp_config_changed, 0, vp_dev->msix_names[v],
-				  vp_dev);
-		if (err)
-			goto error_irq;
-		++vp_dev->msix_used_vectors;
-
-		iowrite16(v, vp_dev->ioaddr + VIRTIO_MSI_CONFIG_VECTOR);
-		/* Verify we had enough resources to assign the vector */
-		v = ioread16(vp_dev->ioaddr + VIRTIO_MSI_CONFIG_VECTOR);
-		if (v == VIRTIO_MSI_NO_VECTOR) {
-			err = -EBUSY;
-			goto error_irq;
-		}
+	err = pci_enable_msix(vp_dev->pci_dev, vp_dev->msix_entries, nvectors);
+	if (err > 0)
+		err = -ENOSPC;
+	if (err)
+		goto error;
+	vp_dev->msix_vectors = nvectors;
+	vp_dev->msix_enabled = 1;
+
+	/* Set the vector used for configuration */
+	v = vp_dev->msix_used_vectors;
+	snprintf(vp_dev->msix_names[v], sizeof *vp_dev->msix_names,
+		 "%s-config", name);
+	err = request_irq(vp_dev->msix_entries[v].vector,
+			  vp_config_changed, 0, vp_dev->msix_names[v],
+			  vp_dev);
+	if (err)
+		goto error;
+	++vp_dev->msix_used_vectors;
+
+	iowrite16(v, vp_dev->ioaddr + VIRTIO_MSI_CONFIG_VECTOR);
+	/* Verify we had enough resources to assign the vector */
+	v = ioread16(vp_dev->ioaddr + VIRTIO_MSI_CONFIG_VECTOR);
+	if (v == VIRTIO_MSI_NO_VECTOR) {
+		err = -EBUSY;
+		goto error;
 	}
 
-	if (vp_dev->msix_vectors && vp_dev->msix_vectors != max_vqs + 1) {
+	if (!per_vq_vectors) {
 		/* Shared vector for all VQs */
 		v = vp_dev->msix_used_vectors;
 		snprintf(vp_dev->msix_names[v], sizeof *vp_dev->msix_names,
@@ -349,28 +347,25 @@ static int vp_request_vectors(struct virtio_device *vdev, unsigned max_vqs)
 				  vp_vring_interrupt, 0, vp_dev->msix_names[v],
 				  vp_dev);
 		if (err)
-			goto error_irq;
+			goto error;
 		++vp_dev->msix_used_vectors;
 	}
 	return 0;
-error_irq:
+error:
 	vp_free_vectors(vdev);
-	kfree(vp_dev->msix_names);
-error_names:
-	kfree(vp_dev->msix_entries);
-error_entries:
 	return err;
 }
 
 static struct virtqueue *vp_find_vq(struct virtio_device *vdev, unsigned index,
 				    void (*callback)(struct virtqueue *vq),
-				    const char *name)
+				    const char *name,
+				    u16 vector)
 {
 	struct virtio_pci_device *vp_dev = to_vp_device(vdev);
 	struct virtio_pci_vq_info *info;
 	struct virtqueue *vq;
 	unsigned long flags, size;
-	u16 num, vector;
+	u16 num;
 	int err;
 
 	/* Select the queue we're interested in */
@@ -389,7 +384,7 @@ static struct virtqueue *vp_find_vq(struct virtio_device *vdev, unsigned index,
 
 	info->queue_index = index;
 	info->num = num;
-	info->vector = VIRTIO_MSI_NO_VECTOR;
+	info->vector = vector;
 
 	size = PAGE_ALIGN(vring_size(num, VIRTIO_PCI_VRING_ALIGN));
 	info->queue = alloc_pages_exact(size, GFP_KERNEL|__GFP_ZERO);
@@ -413,22 +408,7 @@ static struct virtqueue *vp_find_vq(struct virtio_device *vdev, unsigned index,
 	vq->priv = info;
 	info->vq = vq;
 
-	/* allocate per-vq vector if available and necessary */
-	if (callback && vp_dev->msix_used_vectors < vp_dev->msix_vectors) {
-		vector = vp_dev->msix_used_vectors;
-		snprintf(vp_dev->msix_names[vector], sizeof *vp_dev->msix_names,
-			 "%s-%s", dev_name(&vp_dev->vdev.dev), name);
-		err = request_irq(vp_dev->msix_entries[vector].vector,
-				  vring_interrupt, 0,
-				  vp_dev->msix_names[vector], vq);
-		if (err)
-			goto out_request_irq;
-		info->vector = vector;
-		++vp_dev->msix_used_vectors;
-	} else
-		vector = VP_MSIX_VQ_VECTOR;
-
-	 if (callback && vp_dev->msix_enabled) {
+	 if (vector != VIRTIO_MSI_NO_VECTOR) {
 		iowrite16(vector, vp_dev->ioaddr + VIRTIO_MSI_QUEUE_VECTOR);
 		vector = ioread16(vp_dev->ioaddr + VIRTIO_MSI_QUEUE_VECTOR);
 		if (vector == VIRTIO_MSI_NO_VECTOR) {
@@ -444,11 +424,6 @@ static struct virtqueue *vp_find_vq(struct virtio_device *vdev, unsigned index,
 	return vq;
 
 out_assign:
-	if (info->vector != VIRTIO_MSI_NO_VECTOR) {
-		free_irq(vp_dev->msix_entries[info->vector].vector, vq);
-		--vp_dev->msix_used_vectors;
-	}
-out_request_irq:
 	vring_del_virtqueue(vq);
 out_activate_queue:
 	iowrite32(0, vp_dev->ioaddr + VIRTIO_PCI_QUEUE_PFN);
@@ -462,12 +437,13 @@ static void vp_del_vq(struct virtqueue *vq)
 {
 	struct virtio_pci_device *vp_dev = to_vp_device(vq->vdev);
 	struct virtio_pci_vq_info *info = vq->priv;
-	unsigned long size;
+	unsigned long flags, size;
 
-	iowrite16(info->queue_index, vp_dev->ioaddr + VIRTIO_PCI_QUEUE_SEL);
+	spin_lock_irqsave(&vp_dev->lock, flags);
+	list_del(&info->node);
+	spin_unlock_irqrestore(&vp_dev->lock, flags);
 
-	if (info->vector != VIRTIO_MSI_NO_VECTOR)
-		free_irq(vp_dev->msix_entries[info->vector].vector, vq);
+	iowrite16(info->queue_index, vp_dev->ioaddr + VIRTIO_PCI_QUEUE_SEL);
 
 	if (vp_dev->msix_enabled) {
 		iowrite16(VIRTIO_MSI_NO_VECTOR,
@@ -489,36 +465,62 @@ static void vp_del_vq(struct virtqueue *vq)
 /* the config->del_vqs() implementation */
 static void vp_del_vqs(struct virtio_device *vdev)
 {
+	struct virtio_pci_device *vp_dev = to_vp_device(vdev);
 	struct virtqueue *vq, *n;
+	struct virtio_pci_vq_info *info;
 
-	list_for_each_entry_safe(vq, n, &vdev->vqs, list)
+	list_for_each_entry_safe(vq, n, &vdev->vqs, list) {
+		info = vq->priv;
+		if (vp_dev->per_vq_vectors)
+			free_irq(vp_dev->msix_entries[info->vector].vector, vq);
 		vp_del_vq(vq);
+	}
+	vp_dev->per_vq_vectors = false;
 
 	vp_free_vectors(vdev);
 }
 
-/* the config->find_vqs() implementation */
-static int vp_find_vqs(struct virtio_device *vdev, unsigned nvqs,
-		       struct virtqueue *vqs[],
-		       vq_callback_t *callbacks[],
-		       const char *names[])
+static int vp_try_to_find_vqs(struct virtio_device *vdev, unsigned nvqs,
+			      struct virtqueue *vqs[],
+			      vq_callback_t *callbacks[],
+			      const char *names[],
+			      int nvectors,
+			      bool per_vq_vectors)
 {
-	int vectors = 0;
-	int i, err;
-
-	/* How many vectors would we like? */
-	for (i = 0; i < nvqs; ++i)
-		if (callbacks[i])
-			++vectors;
+	struct virtio_pci_device *vp_dev = to_vp_device(vdev);
+	u16 vector;
+	int i, err, allocated_vectors;
 
-	err = vp_request_vectors(vdev, vectors);
+	err = vp_request_vectors(vdev, nvectors, per_vq_vectors);
 	if (err)
 		goto error_request;
 
+	vp_dev->per_vq_vectors = per_vq_vectors;
+	allocated_vectors = vp_dev->msix_used_vectors;
 	for (i = 0; i < nvqs; ++i) {
-		vqs[i] = vp_find_vq(vdev, i, callbacks[i], names[i]);
-		if (IS_ERR(vqs[i]))
+		if (!callbacks[i] || !vp_dev->msix_enabled)
+			vector = VIRTIO_MSI_NO_VECTOR;
+		else if (vp_dev->per_vq_vectors)
+			vector = allocated_vectors++;
+		else
+			vector = VP_MSIX_VQ_VECTOR;
+		vqs[i] = vp_find_vq(vdev, i, callbacks[i], names[i], vector);
+		if (IS_ERR(vqs[i])) {
+			err = PTR_ERR(vqs[i]);
 			goto error_find;
+		}
+		/* allocate per-vq irq if available and necessary */
+		if (vp_dev->per_vq_vectors && vector != VIRTIO_MSI_NO_VECTOR) {
+			snprintf(vp_dev->msix_names[vector], sizeof *vp_dev->msix_names,
+				 "%s-%s", dev_name(&vp_dev->vdev.dev), names[i]);
+			err = request_irq(vp_dev->msix_entries[vector].vector,
+					  vring_interrupt, 0,
+					  vp_dev->msix_names[vector], vqs[i]);
+			if (err) {
+				vp_del_vq(vqs[i]);
+				goto error_find;
+			}
+		}
 	}
 	return 0;
 
@@ -526,7 +528,37 @@ error_find:
 	vp_del_vqs(vdev);
 
 error_request:
-	return PTR_ERR(vqs[i]);
+	return err;
+}
+
+/* the config->find_vqs() implementation */
+static int vp_find_vqs(struct virtio_device *vdev, unsigned nvqs,
+		       struct virtqueue *vqs[],
+		       vq_callback_t *callbacks[],
+		       const char *names[])
+{
+	int vectors = 0;
+	int i, uninitialized_var(err);
+
+	/* How many vectors would we like? */
+	for (i = 0; i < nvqs; ++i)
+		if (callbacks[i])
+			++vectors;
+
+	/* We want at most one vector per queue and one for config changes. */
+	err = vp_try_to_find_vqs(vdev, nvqs, vqs, callbacks, names,
+				 vectors + 1, true);
+	if (!err)
+		return 0;
+	/* Fallback to separate vectors for config and a shared for queues. */
+	err = vp_try_to_find_vqs(vdev, nvqs, vqs, callbacks, names,
+				 2, false);
+	if (!err)
+		return 0;
+	/* Finally fall back to regular interrupts. */
+	err = vp_try_to_find_vqs(vdev, nvqs, vqs, callbacks, names,
+				 0, false);
+	return err;
 }
 
 static struct virtio_config_ops virtio_pci_config_ops = {
diff --git a/drivers/w1/masters/omap_hdq.c b/drivers/w1/masters/omap_hdq.c
index a7e3b706b9d..0d92969404c 100644
--- a/drivers/w1/masters/omap_hdq.c
+++ b/drivers/w1/masters/omap_hdq.c
@@ -687,6 +687,7 @@ static int omap_hdq_remove(struct platform_device *pdev)
 
 	if (hdq_data->hdq_usecount) {
 		dev_dbg(&pdev->dev, "removed when use count is not zero\n");
+		mutex_unlock(&hdq_data->hdq_mutex);
 		return -EBUSY;
 	}
 
diff --git a/drivers/watchdog/coh901327_wdt.c b/drivers/watchdog/coh901327_wdt.c
index fecb307d28e..aec7cefdef2 100644
--- a/drivers/watchdog/coh901327_wdt.c
+++ b/drivers/watchdog/coh901327_wdt.c
@@ -18,6 +18,7 @@
 #include <linux/bitops.h>
 #include <linux/uaccess.h>
 #include <linux/clk.h>
+#include <linux/delay.h>
 
 #define DRV_NAME "WDOG COH 901 327"
 
@@ -92,6 +93,8 @@ static struct clk *clk;
 static void coh901327_enable(u16 timeout)
 {
 	u16 val;
+	unsigned long freq;
+	unsigned long delay_ns;
 
 	clk_enable(clk);
 	/* Restart timer if it is disabled */
@@ -102,6 +105,14 @@ static void coh901327_enable(u16 timeout)
 	/* Acknowledge any pending interrupt so it doesn't just fire off */
 	writew(U300_WDOG_IER_WILL_BARK_IRQ_ACK_ENABLE,
 	       virtbase + U300_WDOG_IER);
+	/*
+	 * The interrupt is cleared in the 32 kHz clock domain.
+	 * Wait 3 32 kHz cycles for it to take effect
+	 */
+	freq = clk_get_rate(clk);
+	delay_ns = (1000000000 + freq - 1) / freq; /* Freq to ns and round up */
+	delay_ns = 3 * delay_ns; /* Wait 3 cycles */
+	ndelay(delay_ns);
 	/* Enable the watchdog interrupt */
 	writew(U300_WDOG_IMR_WILL_BARK_IRQ_ENABLE, virtbase + U300_WDOG_IMR);
 	/* Activate the watchdog timer */
diff --git a/drivers/watchdog/ks8695_wdt.c b/drivers/watchdog/ks8695_wdt.c
index 00b03eb43bf..e1c82769b08 100644
--- a/drivers/watchdog/ks8695_wdt.c
+++ b/drivers/watchdog/ks8695_wdt.c
@@ -66,7 +66,7 @@ static inline void ks8695_wdt_stop(void)
 static inline void ks8695_wdt_start(void)
 {
 	unsigned long tmcon;
-	unsigned long tval = wdt_time * CLOCK_TICK_RATE;
+	unsigned long tval = wdt_time * KS8695_CLOCK_RATE;
 
 	spin_lock(&ks8695_lock);
 	/* disable timer0 */
@@ -103,7 +103,7 @@ static inline void ks8695_wdt_reload(void)
 static int ks8695_wdt_settimeout(int new_time)
 {
 	/*
-	 * All counting occurs at SLOW_CLOCK / 128 = 0.256 Hz
+	 * All counting occurs at KS8695_CLOCK_RATE / 128 = 0.256 Hz
 	 *
 	 * Since WDV is a 16-bit counter, the maximum period is
 	 * 65536 / 0.256 = 256 seconds.