Update from upstream with manual merge of Yasunori Goto's

changes to swiotlb.c made in commit 281dd25cdc0d6903929b79183816d151ea626341
since this file has been moved from arch/ia64/lib/swiotlb.c to
lib/swiotlb.c

Signed-off-by: Tony Luck <tony.luck@intel.com>

19 files changed, 161 insertions, 117 deletions

diff --git a/arch/ppc64/kernel/bpa_iommu.c b/arch/ppc64/kernel/bpa_iommu.c
index f33a7bccb0d..5f2460090e0 100644
--- a/arch/ppc64/kernel/bpa_iommu.c
+++ b/arch/ppc64/kernel/bpa_iommu.c
@@ -99,7 +99,11 @@ get_iost_entry(unsigned long iopt_base, unsigned long io_address, unsigned page_
 		break;
 
 	default: /* not a known compile time constant */
-		BUILD_BUG_ON(1);
+		{
+			/* BUILD_BUG_ON() is not usable here */
+			extern void __get_iost_entry_bad_page_size(void);
+			__get_iost_entry_bad_page_size();
+		}
 		break;
 	}
 
@@ -306,7 +310,7 @@ static void bpa_map_iommu(void)
 
 
 static void *bpa_alloc_coherent(struct device *hwdev, size_t size,
-			   dma_addr_t *dma_handle, unsigned int __nocast flag)
+			   dma_addr_t *dma_handle, gfp_t flag)
 {
 	void *ret;
 
diff --git a/arch/ppc64/kernel/dma.c b/arch/ppc64/kernel/dma.c
index 4da8e31b2b6..7c3419656cc 100644
--- a/arch/ppc64/kernel/dma.c
+++ b/arch/ppc64/kernel/dma.c
@@ -53,7 +53,7 @@ int dma_set_mask(struct device *dev, u64 dma_mask)
 EXPORT_SYMBOL(dma_set_mask);
 
 void *dma_alloc_coherent(struct device *dev, size_t size,
-		dma_addr_t *dma_handle, unsigned int __nocast flag)
+		dma_addr_t *dma_handle, gfp_t flag)
 {
 	struct dma_mapping_ops *dma_ops = get_dma_ops(dev);
 
diff --git a/arch/ppc64/kernel/head.S b/arch/ppc64/kernel/head.S
index 58c314738c9..72c61041151 100644
--- a/arch/ppc64/kernel/head.S
+++ b/arch/ppc64/kernel/head.S
@@ -1649,7 +1649,7 @@ _GLOBAL(__secondary_start)
 	ld	r3,0(r3)
 	lwz	r3,PLATFORM(r3)		/* r3 = platform flags		 */
 	andi.	r3,r3,PLATFORM_LPAR	/* Test if bit 0 is set (LPAR bit) */
-	bne	98f
+	beq	98f			/* branch if result is 0  */
 	mfspr	r3,PVR
 	srwi	r3,r3,16
 	cmpwi	r3,0x37			/* SStar  */
@@ -1813,7 +1813,7 @@ _STATIC(start_here_multiplatform)
 	ld	r3,0(r3)
 	lwz	r3,PLATFORM(r3)		/* r3 = platform flags */
 	andi.	r3,r3,PLATFORM_LPAR	/* Test if bit 0 is set (LPAR bit) */
-	bne	98f
+	beq	98f			/* branch if result is 0  */
 	mfspr	r3,PVR
 	srwi	r3,r3,16
 	cmpwi	r3,0x37			/* SStar */
@@ -1834,7 +1834,7 @@ _STATIC(start_here_multiplatform)
 	lwz	r3,PLATFORM(r3)		/* r3 = platform flags */
 	/* Test if bit 0 is set (LPAR bit) */
 	andi.	r3,r3,PLATFORM_LPAR
-	bne	98f
+	bne	98f			/* branch if result is !0  */
 	LOADADDR(r6,_SDR1)		/* Only if NOT LPAR */
 	sub	r6,r6,r26
 	ld	r6,0(r6)		/* get the value of _SDR1 */
diff --git a/arch/ppc64/kernel/iommu.c b/arch/ppc64/kernel/iommu.c
index 9032b6bfe03..4d9b4388918 100644
--- a/arch/ppc64/kernel/iommu.c
+++ b/arch/ppc64/kernel/iommu.c
@@ -519,7 +519,7 @@ void iommu_unmap_single(struct iommu_table *tbl, dma_addr_t dma_handle,
  * to the dma address (mapping) of the first page.
  */
 void *iommu_alloc_coherent(struct iommu_table *tbl, size_t size,
-		dma_addr_t *dma_handle, unsigned int __nocast flag)
+		dma_addr_t *dma_handle, gfp_t flag)
 {
 	void *ret = NULL;
 	dma_addr_t mapping;
diff --git a/arch/ppc64/kernel/kprobes.c b/arch/ppc64/kernel/kprobes.c
index 7e80d49c589..9c6facc24f7 100644
--- a/arch/ppc64/kernel/kprobes.c
+++ b/arch/ppc64/kernel/kprobes.c
@@ -59,9 +59,9 @@ int __kprobes arch_prepare_kprobe(struct kprobe *p)
 
 	/* insn must be on a special executable page on ppc64 */
 	if (!ret) {
-		up(&kprobe_mutex);
-		p->ainsn.insn = get_insn_slot();
 		down(&kprobe_mutex);
+		p->ainsn.insn = get_insn_slot();
+		up(&kprobe_mutex);
 		if (!p->ainsn.insn)
 			ret = -ENOMEM;
 	}
@@ -90,9 +90,9 @@ void __kprobes arch_disarm_kprobe(struct kprobe *p)
 
 void __kprobes arch_remove_kprobe(struct kprobe *p)
 {
-	up(&kprobe_mutex);
-	free_insn_slot(p->ainsn.insn);
 	down(&kprobe_mutex);
+	free_insn_slot(p->ainsn.insn);
+	up(&kprobe_mutex);
 }
 
 static inline void prepare_singlestep(struct kprobe *p, struct pt_regs *regs)
diff --git a/arch/ppc64/kernel/machine_kexec.c b/arch/ppc64/kernel/machine_kexec.c
index 4775f12a013..bf7cc4f8210 100644
--- a/arch/ppc64/kernel/machine_kexec.c
+++ b/arch/ppc64/kernel/machine_kexec.c
@@ -205,6 +205,7 @@ static void kexec_prepare_cpus(void)
 			continue;
 
 		while (paca[i].hw_cpu_id != -1) {
+			barrier();
 			if (!cpu_possible(i)) {
 				printk("kexec: cpu %d hw_cpu_id %d is not"
 						" possible, ignoring\n",
diff --git a/arch/ppc64/kernel/module.c b/arch/ppc64/kernel/module.c
index c683bf88e69..928b8581fcb 100644
--- a/arch/ppc64/kernel/module.c
+++ b/arch/ppc64/kernel/module.c
@@ -341,6 +341,19 @@ int apply_relocate_add(Elf64_Shdr *sechdrs,
 			*(unsigned long *)location = my_r2(sechdrs, me);
 			break;
 
+		case R_PPC64_TOC16:
+			/* Subtact TOC pointer */
+			value -= my_r2(sechdrs, me);
+			if (value + 0x8000 > 0xffff) {
+				printk("%s: bad TOC16 relocation (%lu)\n",
+				       me->name, value);
+				return -ENOEXEC;
+			}
+			*((uint16_t *) location)
+				= (*((uint16_t *) location) & ~0xffff)
+				| (value & 0xffff);
+			break;
+
 		case R_PPC64_TOC16_DS:
 			/* Subtact TOC pointer */
 			value -= my_r2(sechdrs, me);
diff --git a/arch/ppc64/kernel/of_device.c b/arch/ppc64/kernel/of_device.c
index da580812ddf..9f200f0f2ad 100644
--- a/arch/ppc64/kernel/of_device.c
+++ b/arch/ppc64/kernel/of_device.c
@@ -233,7 +233,9 @@ void of_device_unregister(struct of_device *ofdev)
 	device_unregister(&ofdev->dev);
 }
 
-struct of_device* of_platform_device_create(struct device_node *np, const char *bus_id)
+struct of_device* of_platform_device_create(struct device_node *np,
+					    const char *bus_id,
+					    struct device *parent)
 {
 	struct of_device *dev;
 
@@ -245,7 +247,7 @@ struct of_device* of_platform_device_create(struct device_node *np, const char *
 	dev->node = np;
 	dev->dma_mask = 0xffffffffUL;
 	dev->dev.dma_mask = &dev->dma_mask;
-	dev->dev.parent = NULL;
+	dev->dev.parent = parent;
 	dev->dev.bus = &of_platform_bus_type;
 	dev->dev.release = of_release_dev;
 
@@ -259,6 +261,7 @@ struct of_device* of_platform_device_create(struct device_node *np, const char *
 	return dev;
 }
 
+
 EXPORT_SYMBOL(of_match_device);
 EXPORT_SYMBOL(of_platform_bus_type);
 EXPORT_SYMBOL(of_register_driver);
diff --git a/arch/ppc64/kernel/pSeries_iommu.c b/arch/ppc64/kernel/pSeries_iommu.c
index f0fd7fbd653..d17f0108a03 100644
--- a/arch/ppc64/kernel/pSeries_iommu.c
+++ b/arch/ppc64/kernel/pSeries_iommu.c
@@ -265,8 +265,10 @@ static void iommu_table_setparms(struct pci_controller *phb,
 	tbl->it_offset = phb->dma_window_base_cur >> PAGE_SHIFT;
 	
 	/* Test if we are going over 2GB of DMA space */
-	if (phb->dma_window_base_cur + phb->dma_window_size > (1L << 31))
+	if (phb->dma_window_base_cur + phb->dma_window_size > 0x80000000ul) {
+		udbg_printf("PCI_DMA: Unexpected number of IOAs under this PHB.\n");
 		panic("PCI_DMA: Unexpected number of IOAs under this PHB.\n"); 
+	}
 	
 	phb->dma_window_base_cur += phb->dma_window_size;
 
@@ -310,92 +312,85 @@ static void iommu_table_setparms_lpar(struct pci_controller *phb,
 
 static void iommu_bus_setup_pSeries(struct pci_bus *bus)
 {
-	struct device_node *dn, *pdn;
-	struct pci_dn *pci;
+	struct device_node *dn;
 	struct iommu_table *tbl;
+	struct device_node *isa_dn, *isa_dn_orig;
+	struct device_node *tmp;
+	struct pci_dn *pci;
+	int children;
 
 	DBG("iommu_bus_setup_pSeries, bus %p, bus->self %p\n", bus, bus->self);
 
-	/* For each (root) bus, we carve up the available DMA space in 256MB
-	 * pieces. Since each piece is used by one (sub) bus/device, that would
-	 * give a maximum of 7 devices per PHB. In most cases, this is plenty.
-	 *
-	 * The exception is on Python PHBs (pre-POWER4). Here we don't have EADS
-	 * bridges below the PHB to allocate the sectioned tables to, so instead
-	 * we allocate a 1GB table at the PHB level.
+	dn = pci_bus_to_OF_node(bus);
+	pci = PCI_DN(dn);
+
+	if (bus->self) {
+		/* This is not a root bus, any setup will be done for the
+		 * device-side of the bridge in iommu_dev_setup_pSeries().
+		 */
+		return;
+	}
+
+	/* Check if the ISA bus on the system is under
+	 * this PHB.
 	 */
+	isa_dn = isa_dn_orig = of_find_node_by_type(NULL, "isa");
 
-	dn = pci_bus_to_OF_node(bus);
-	pci = dn->data;
-
-	if (!bus->self) {
-		/* Root bus */
-		if (is_python(dn)) {
-			unsigned int *iohole;
-
-			DBG("Python root bus %s\n", bus->name);
-
-			iohole = (unsigned int *)get_property(dn, "io-hole", 0);
-
-			if (iohole) {
-				/* On first bus we need to leave room for the
-				 * ISA address space. Just skip the first 256MB
-				 * alltogether. This leaves 768MB for the window.
-				 */
-				DBG("PHB has io-hole, reserving 256MB\n");
-				pci->phb->dma_window_size = 3 << 28;
-				pci->phb->dma_window_base_cur = 1 << 28;
-			} else {
-				/* 1GB window by default */
-				pci->phb->dma_window_size = 1 << 30;
-				pci->phb->dma_window_base_cur = 0;
-			}
-
-			tbl = kmalloc(sizeof(struct iommu_table), GFP_KERNEL);
-
-			iommu_table_setparms(pci->phb, dn, tbl);
-			pci->iommu_table = iommu_init_table(tbl);
-		} else {
-			/* Do a 128MB table at root. This is used for the IDE
-			 * controller on some SMP-mode POWER4 machines. It
-			 * doesn't hurt to allocate it on other machines
-			 * -- it'll just be unused since new tables are
-			 * allocated on the EADS level.
-			 *
-			 * Allocate at offset 128MB to avoid having to deal
-			 * with ISA holes; 128MB table for IDE is plenty.
-			 */
-			pci->phb->dma_window_size = 1 << 27;
-			pci->phb->dma_window_base_cur = 1 << 27;
-
-			tbl = kmalloc(sizeof(struct iommu_table), GFP_KERNEL);
-
-			iommu_table_setparms(pci->phb, dn, tbl);
-			pci->iommu_table = iommu_init_table(tbl);
-
-			/* All child buses have 256MB tables */
-			pci->phb->dma_window_size = 1 << 28;
-		}
-	} else {
-		pdn = pci_bus_to_OF_node(bus->parent);
+	while (isa_dn && isa_dn != dn)
+		isa_dn = isa_dn->parent;
+
+	if (isa_dn_orig)
+		of_node_put(isa_dn_orig);
 
-		if (!bus->parent->self && !is_python(pdn)) {
-			struct iommu_table *tbl;
-			/* First child and not python means this is the EADS
-			 * level. Allocate new table for this slot with 256MB
-			 * window.
-			 */
+	/* Count number of direct PCI children of the PHB.
+	 * All PCI device nodes have class-code property, so it's
+	 * an easy way to find them.
+	 */
+	for (children = 0, tmp = dn->child; tmp; tmp = tmp->sibling)
+		if (get_property(tmp, "class-code", NULL))
+			children++;
 
-			tbl = kmalloc(sizeof(struct iommu_table), GFP_KERNEL);
+	DBG("Children: %d\n", children);
 
-			iommu_table_setparms(pci->phb, dn, tbl);
+	/* Calculate amount of DMA window per slot. Each window must be
+	 * a power of two (due to pci_alloc_consistent requirements).
+	 *
+	 * Keep 256MB aside for PHBs with ISA.
+	 */
 
-			pci->iommu_table = iommu_init_table(tbl);
-		} else {
-			/* Lower than first child or under python, use parent table */
-			pci->iommu_table = PCI_DN(pdn)->iommu_table;
-		}
+	if (!isa_dn) {
+		/* No ISA/IDE - just set window size and return */
+		pci->phb->dma_window_size = 0x80000000ul; /* To be divided */
+
+		while (pci->phb->dma_window_size * children > 0x80000000ul)
+			pci->phb->dma_window_size >>= 1;
+		DBG("No ISA/IDE, window size is 0x%lx\n",
+			pci->phb->dma_window_size);
+		pci->phb->dma_window_base_cur = 0;
+
+		return;
 	}
+
+	/* If we have ISA, then we probably have an IDE
+	 * controller too. Allocate a 128MB table but
+	 * skip the first 128MB to avoid stepping on ISA
+	 * space.
+	 */
+	pci->phb->dma_window_size = 0x8000000ul;
+	pci->phb->dma_window_base_cur = 0x8000000ul;
+
+	tbl = kmalloc(sizeof(struct iommu_table), GFP_KERNEL);
+
+	iommu_table_setparms(pci->phb, dn, tbl);
+	pci->iommu_table = iommu_init_table(tbl);
+
+	/* Divide the rest (1.75GB) among the children */
+	pci->phb->dma_window_size = 0x80000000ul;
+	while (pci->phb->dma_window_size * children > 0x70000000ul)
+		pci->phb->dma_window_size >>= 1;
+
+	DBG("ISA/IDE, window size is 0x%lx\n", pci->phb->dma_window_size);
+
 }
 
 
@@ -446,21 +441,36 @@ static void iommu_bus_setup_pSeriesLP(struct pci_bus *bus)
 static void iommu_dev_setup_pSeries(struct pci_dev *dev)
 {
 	struct device_node *dn, *mydn;
+	struct iommu_table *tbl;
+
+	DBG("iommu_dev_setup_pSeries, dev %p (%s)\n", dev, pci_name(dev));
 
-	DBG("iommu_dev_setup_pSeries, dev %p (%s)\n", dev, dev->pretty_name);
-	/* Now copy the iommu_table ptr from the bus device down to the
-	 * pci device_node.  This means get_iommu_table() won't need to search
-	 * up the device tree to find it.
-	 */
 	mydn = dn = pci_device_to_OF_node(dev);
 
+	/* If we're the direct child of a root bus, then we need to allocate
+	 * an iommu table ourselves. The bus setup code should have setup
+	 * the window sizes already.
+	 */
+	if (!dev->bus->self) {
+		DBG(" --> first child, no bridge. Allocating iommu table.\n");
+		tbl = kmalloc(sizeof(struct iommu_table), GFP_KERNEL);
+		iommu_table_setparms(PCI_DN(dn)->phb, dn, tbl);
+		PCI_DN(mydn)->iommu_table = iommu_init_table(tbl);
+
+		return;
+	}
+
+	/* If this device is further down the bus tree, search upwards until
+	 * an already allocated iommu table is found and use that.
+	 */
+
 	while (dn && dn->data && PCI_DN(dn)->iommu_table == NULL)
 		dn = dn->parent;
 
 	if (dn && dn->data) {
 		PCI_DN(mydn)->iommu_table = PCI_DN(dn)->iommu_table;
 	} else {
-		DBG("iommu_dev_setup_pSeries, dev %p (%s) has no iommu table\n", dev, dev->pretty_name);
+		DBG("iommu_dev_setup_pSeries, dev %p (%s) has no iommu table\n", dev, pci_name(dev));
 	}
 }
 
@@ -494,7 +504,7 @@ static void iommu_dev_setup_pSeriesLP(struct pci_dev *dev)
 	int *dma_window = NULL;
 	struct pci_dn *pci;
 
-	DBG("iommu_dev_setup_pSeriesLP, dev %p (%s)\n", dev, dev->pretty_name);
+	DBG("iommu_dev_setup_pSeriesLP, dev %p (%s)\n", dev, pci_name(dev));
 
 	/* dev setup for LPAR is a little tricky, since the device tree might
 	 * contain the dma-window properties per-device and not neccesarily
@@ -516,9 +526,8 @@ static void iommu_dev_setup_pSeriesLP(struct pci_dev *dev)
 	 * slots on POWER4 machines.
 	 */
 	if (dma_window == NULL || pdn->parent == NULL) {
-		/* Fall back to regular (non-LPAR) dev setup */
-		DBG("No dma window for device, falling back to regular setup\n");
-		iommu_dev_setup_pSeries(dev);
+		DBG("No dma window for device, linking to parent\n");
+		PCI_DN(dn)->iommu_table = PCI_DN(pdn)->iommu_table;
 		return;
 	} else {
 		DBG("Found DMA window, allocating table\n");
diff --git a/arch/ppc64/kernel/pSeries_pci.c b/arch/ppc64/kernel/pSeries_pci.c
index 1f5f141fb7a..928f8febdb3 100644
--- a/arch/ppc64/kernel/pSeries_pci.c
+++ b/arch/ppc64/kernel/pSeries_pci.c
@@ -32,7 +32,7 @@
 
 #include "pci.h"
 
-static int __initdata s7a_workaround = -1;
+static int __devinitdata s7a_workaround = -1;
 
 #if 0
 void pcibios_name_device(struct pci_dev *dev)
@@ -60,7 +60,7 @@ void pcibios_name_device(struct pci_dev *dev)
 DECLARE_PCI_FIXUP_HEADER(PCI_ANY_ID, PCI_ANY_ID, pcibios_name_device);
 #endif
 
-static void __init check_s7a(void)
+static void __devinit check_s7a(void)
 {
 	struct device_node *root;
 	char *model;
diff --git a/arch/ppc64/kernel/pci.c b/arch/ppc64/kernel/pci.c
index 861138ad092..ff4be1da69d 100644
--- a/arch/ppc64/kernel/pci.c
+++ b/arch/ppc64/kernel/pci.c
@@ -246,11 +246,14 @@ static unsigned int pci_parse_of_flags(u32 addr0)
 	unsigned int flags = 0;
 
 	if (addr0 & 0x02000000) {
-		flags |= IORESOURCE_MEM;
+		flags = IORESOURCE_MEM | PCI_BASE_ADDRESS_SPACE_MEMORY;
+		flags |= (addr0 >> 22) & PCI_BASE_ADDRESS_MEM_TYPE_64;
+		flags |= (addr0 >> 28) & PCI_BASE_ADDRESS_MEM_TYPE_1M;
 		if (addr0 & 0x40000000)
-			flags |= IORESOURCE_PREFETCH;
+			flags |= IORESOURCE_PREFETCH
+				 | PCI_BASE_ADDRESS_MEM_PREFETCH;
 	} else if (addr0 & 0x01000000)
-		flags |= IORESOURCE_IO;
+		flags = IORESOURCE_IO | PCI_BASE_ADDRESS_SPACE_IO;
 	return flags;
 }
 
diff --git a/arch/ppc64/kernel/pci_direct_iommu.c b/arch/ppc64/kernel/pci_direct_iommu.c
index b8f7f58824f..54055c81017 100644
--- a/arch/ppc64/kernel/pci_direct_iommu.c
+++ b/arch/ppc64/kernel/pci_direct_iommu.c
@@ -31,7 +31,7 @@
 #include "pci.h"
 
 static void *pci_direct_alloc_coherent(struct device *hwdev, size_t size,
-				   dma_addr_t *dma_handle, unsigned int __nocast flag)
+				   dma_addr_t *dma_handle, gfp_t flag)
 {
 	void *ret;
 
diff --git a/arch/ppc64/kernel/pci_iommu.c b/arch/ppc64/kernel/pci_iommu.c
index 14647e09c9c..d9e33b7d420 100644
--- a/arch/ppc64/kernel/pci_iommu.c
+++ b/arch/ppc64/kernel/pci_iommu.c
@@ -76,7 +76,7 @@ static inline struct iommu_table *devnode_table(struct device *dev)
  * to the dma address (mapping) of the first page.
  */
 static void *pci_iommu_alloc_coherent(struct device *hwdev, size_t size,
-			   dma_addr_t *dma_handle, unsigned int __nocast flag)
+			   dma_addr_t *dma_handle, gfp_t flag)
 {
 	return iommu_alloc_coherent(devnode_table(hwdev), size, dma_handle,
 			flag);
diff --git a/arch/ppc64/kernel/pmac_setup.c b/arch/ppc64/kernel/pmac_setup.c
index 325426c7bed..fa8121d53b8 100644
--- a/arch/ppc64/kernel/pmac_setup.c
+++ b/arch/ppc64/kernel/pmac_setup.c
@@ -115,7 +115,7 @@ static void __pmac pmac_show_cpuinfo(struct seq_file *m)
 	
 	/* find motherboard type */
 	seq_printf(m, "machine\t\t: ");
-	np = find_devices("device-tree");
+	np = of_find_node_by_path("/");
 	if (np != NULL) {
 		pp = (char *) get_property(np, "model", NULL);
 		if (pp != NULL)
@@ -133,6 +133,7 @@ static void __pmac pmac_show_cpuinfo(struct seq_file *m)
 			}
 			seq_printf(m, "\n");
 		}
+		of_node_put(np);
 	} else
 		seq_printf(m, "PowerMac\n");
 
@@ -434,15 +435,23 @@ static int pmac_check_legacy_ioport(unsigned int baseport)
 
 static int __init pmac_declare_of_platform_devices(void)
 {
-	struct device_node *np;
+	struct device_node *np, *npp;
 
-	np = find_devices("u3");
-	if (np) {
-		for (np = np->child; np != NULL; np = np->sibling)
+	npp = of_find_node_by_name(NULL, "u3");
+	if (npp) {
+		for (np = NULL; (np = of_get_next_child(npp, np)) != NULL;) {
 			if (strncmp(np->name, "i2c", 3) == 0) {
-				of_platform_device_create(np, "u3-i2c");
+				of_platform_device_create(np, "u3-i2c", NULL);
+				of_node_put(np);
 				break;
 			}
+		}
+		of_node_put(npp);
+	}
+        npp = of_find_node_by_type(NULL, "smu");
+        if (npp) {
+		of_platform_device_create(npp, "smu", NULL);
+		of_node_put(npp);
 	}
 
 	return 0;
diff --git a/arch/ppc64/kernel/pmac_time.c b/arch/ppc64/kernel/pmac_time.c
index 3059edb09cc..41bbb8c5969 100644
--- a/arch/ppc64/kernel/pmac_time.c
+++ b/arch/ppc64/kernel/pmac_time.c
@@ -84,7 +84,7 @@ void __pmac pmac_get_rtc_time(struct rtc_time *tm)
 
 #ifdef CONFIG_PMAC_SMU
 	case SYS_CTRLER_SMU:
-		smu_get_rtc_time(tm);
+		smu_get_rtc_time(tm, 1);
 		break;
 #endif /* CONFIG_PMAC_SMU */
 	default:
@@ -128,7 +128,7 @@ int __pmac pmac_set_rtc_time(struct rtc_time *tm)
 
 #ifdef CONFIG_PMAC_SMU
 	case SYS_CTRLER_SMU:
-		return smu_set_rtc_time(tm);
+		return smu_set_rtc_time(tm, 1);
 #endif /* CONFIG_PMAC_SMU */
 	default:
 		return -ENODEV;
diff --git a/arch/ppc64/kernel/prom_init.c b/arch/ppc64/kernel/prom_init.c
index 9979919cdf9..f252670874a 100644
--- a/arch/ppc64/kernel/prom_init.c
+++ b/arch/ppc64/kernel/prom_init.c
@@ -1711,6 +1711,7 @@ static void __init flatten_device_tree(void)
 	unsigned long offset = reloc_offset();
 	unsigned long mem_start, mem_end, room;
 	struct boot_param_header *hdr;
+	struct prom_t *_prom = PTRRELOC(&prom);
 	char *namep;
 	u64 *rsvmap;
 
@@ -1765,6 +1766,7 @@ static void __init flatten_device_tree(void)
 	RELOC(dt_struct_end) = PAGE_ALIGN(mem_start);
 
 	/* Finish header */
+	hdr->boot_cpuid_phys = _prom->cpu;
 	hdr->magic = OF_DT_HEADER;
 	hdr->totalsize = RELOC(dt_struct_end) - RELOC(dt_header_start);
 	hdr->off_dt_struct = RELOC(dt_struct_start) - RELOC(dt_header_start);
@@ -1854,7 +1856,6 @@ static void __init prom_find_boot_cpu(void)
 
 	cpu_pkg = call_prom("instance-to-package", 1, 1, prom_cpu);
 
-	prom_setprop(cpu_pkg, "linux,boot-cpu", NULL, 0);
 	prom_getprop(cpu_pkg, "reg", &getprop_rval, sizeof(getprop_rval));
 	_prom->cpu = getprop_rval;
 
diff --git a/arch/ppc64/kernel/ptrace.c b/arch/ppc64/kernel/ptrace.c
index 85ed3188a91..b1c044ca575 100644
--- a/arch/ppc64/kernel/ptrace.c
+++ b/arch/ppc64/kernel/ptrace.c
@@ -219,6 +219,7 @@ int sys_ptrace(long request, long pid, long addr, long data)
 
 	case PTRACE_SET_DEBUGREG:
 		ret = ptrace_set_debugreg(child, addr, data);
+		break;
 
 	case PTRACE_DETACH:
 		ret = ptrace_detach(child, data);
diff --git a/arch/ppc64/kernel/vdso32/gettimeofday.S b/arch/ppc64/kernel/vdso32/gettimeofday.S
index 07f1c1c650c..e243c1d24af 100644
--- a/arch/ppc64/kernel/vdso32/gettimeofday.S
+++ b/arch/ppc64/kernel/vdso32/gettimeofday.S
@@ -109,7 +109,7 @@ __do_get_xsec:
 	lwz	r6,(CFG_TB_TO_XS+4)(r9)
 	mulhwu	r4,r7,r5
 	mulhwu	r6,r7,r6
-	mullw	r6,r7,r5
+	mullw	r0,r7,r5
 	addc	r6,r6,r0
 
 	/* At this point, we have the scaled xsec value in r4 + XER:CA
diff --git a/arch/ppc64/kernel/vio.c b/arch/ppc64/kernel/vio.c
index c90e1dd875c..0e555b7a658 100644
--- a/arch/ppc64/kernel/vio.c
+++ b/arch/ppc64/kernel/vio.c
@@ -218,7 +218,7 @@ static void vio_unmap_sg(struct device *dev, struct scatterlist *sglist,
 }
 
 static void *vio_alloc_coherent(struct device *dev, size_t size,
-			   dma_addr_t *dma_handle, unsigned int __nocast flag)
+			   dma_addr_t *dma_handle, gfp_t flag)
 {
 	return iommu_alloc_coherent(to_vio_dev(dev)->iommu_table, size,
 			dma_handle, flag);