15 files changed, 248 insertions, 181 deletions
diff --git a/arch/sparc64/Kconfig b/arch/sparc64/Kconfig
index 923a98959fa7..36b4b7ab9cfb 100644
--- a/arch/sparc64/Kconfig
+++ b/arch/sparc64/Kconfig
@@ -231,6 +231,7 @@ endmenu
 
 config NUMA
 	bool "NUMA support"
+	depends on SMP
 
 config NODES_SHIFT
 	int
diff --git a/arch/sparc64/kernel/irq.c b/arch/sparc64/kernel/irq.c
index ba43d85e8dde..7495bc774685 100644
--- a/arch/sparc64/kernel/irq.c
+++ b/arch/sparc64/kernel/irq.c
@@ -7,6 +7,7 @@
 
 #include <linux/module.h>
 #include <linux/sched.h>
+#include <linux/linkage.h>
 #include <linux/ptrace.h>
 #include <linux/errno.h>
 #include <linux/kernel_stat.h>
@@ -682,10 +683,32 @@ void ack_bad_irq(unsigned int virt_irq)
 	       ino, virt_irq);
 }
 
+void *hardirq_stack[NR_CPUS];
+void *softirq_stack[NR_CPUS];
+
+static __attribute__((always_inline)) void *set_hardirq_stack(void)
+{
+	void *orig_sp, *sp = hardirq_stack[smp_processor_id()];
+
+	__asm__ __volatile__("mov %%sp, %0" : "=r" (orig_sp));
+	if (orig_sp < sp ||
+	    orig_sp > (sp + THREAD_SIZE)) {
+		sp += THREAD_SIZE - 192 - STACK_BIAS;
+		__asm__ __volatile__("mov %0, %%sp" : : "r" (sp));
+	}
+
+	return orig_sp;
+}
+static __attribute__((always_inline)) void restore_hardirq_stack(void *orig_sp)
+{
+	__asm__ __volatile__("mov %0, %%sp" : : "r" (orig_sp));
+}
+
 void handler_irq(int irq, struct pt_regs *regs)
 {
 	unsigned long pstate, bucket_pa;
 	struct pt_regs *old_regs;
+	void *orig_sp;
 
 	clear_softint(1 << irq);
 
@@ -703,6 +726,8 @@ void handler_irq(int irq, struct pt_regs *regs)
 			       "i" (PSTATE_IE)
 			     : "memory");
 
+	orig_sp = set_hardirq_stack();
+
 	while (bucket_pa) {
 		struct irq_desc *desc;
 		unsigned long next_pa;
@@ -719,10 +744,38 @@ void handler_irq(int irq, struct pt_regs *regs)
 		bucket_pa = next_pa;
 	}
 
+	restore_hardirq_stack(orig_sp);
+
 	irq_exit();
 	set_irq_regs(old_regs);
 }
 
+void do_softirq(void)
+{
+	unsigned long flags;
+
+	if (in_interrupt())
+		return;
+
+	local_irq_save(flags);
+
+	if (local_softirq_pending()) {
+		void *orig_sp, *sp = softirq_stack[smp_processor_id()];
+
+		sp += THREAD_SIZE - 192 - STACK_BIAS;
+
+		__asm__ __volatile__("mov %%sp, %0\n\t"
+				     "mov %1, %%sp"
+				     : "=&r" (orig_sp)
+				     : "r" (sp));
+		__do_softirq();
+		__asm__ __volatile__("mov %0, %%sp"
+				     : : "r" (orig_sp));
+	}
+
+	local_irq_restore(flags);
+}
+
 #ifdef CONFIG_HOTPLUG_CPU
 void fixup_irqs(void)
 {
@@ -740,6 +793,8 @@ void fixup_irqs(void)
 		}
 		spin_unlock_irqrestore(&irq_desc[irq].lock, flags);
 	}
+
+	tick_ops->disable_irq();
 }
 #endif
 
@@ -812,7 +867,7 @@ static void kill_prom_timer(void)
 	: "g1", "g2");
 }
 
-void init_irqwork_curcpu(void)
+void notrace init_irqwork_curcpu(void)
 {
 	int cpu = hard_smp_processor_id();
 
@@ -843,7 +898,7 @@ static void __cpuinit register_one_mondo(unsigned long paddr, unsigned long type
 	}
 }
 
-void __cpuinit sun4v_register_mondo_queues(int this_cpu)
+void __cpuinit notrace sun4v_register_mondo_queues(int this_cpu)
 {
 	struct trap_per_cpu *tb = &trap_block[this_cpu];
 
diff --git a/arch/sparc64/kernel/kstack.h b/arch/sparc64/kernel/kstack.h
new file mode 100644
index 000000000000..4248d969272f
--- /dev/null
+++ b/arch/sparc64/kernel/kstack.h
@@ -0,0 +1,60 @@
+#ifndef _KSTACK_H
+#define _KSTACK_H
+
+#include <linux/thread_info.h>
+#include <linux/sched.h>
+#include <asm/ptrace.h>
+#include <asm/irq.h>
+
+/* SP must be STACK_BIAS adjusted already.  */
+static inline bool kstack_valid(struct thread_info *tp, unsigned long sp)
+{
+	unsigned long base = (unsigned long) tp;
+
+	if (sp >= (base + sizeof(struct thread_info)) &&
+	    sp <= (base + THREAD_SIZE - sizeof(struct sparc_stackf)))
+		return true;
+
+	if (hardirq_stack[tp->cpu]) {
+		base = (unsigned long) hardirq_stack[tp->cpu];
+		if (sp >= base &&
+		    sp <= (base + THREAD_SIZE - sizeof(struct sparc_stackf)))
+			return true;
+		base = (unsigned long) softirq_stack[tp->cpu];
+		if (sp >= base &&
+		    sp <= (base + THREAD_SIZE - sizeof(struct sparc_stackf)))
+			return true;
+	}
+	return false;
+}
+
+/* Does "regs" point to a valid pt_regs trap frame?  */
+static inline bool kstack_is_trap_frame(struct thread_info *tp, struct pt_regs *regs)
+{
+	unsigned long base = (unsigned long) tp;
+	unsigned long addr = (unsigned long) regs;
+
+	if (addr >= base &&
+	    addr <= (base + THREAD_SIZE - sizeof(*regs)))
+		goto check_magic;
+
+	if (hardirq_stack[tp->cpu]) {
+		base = (unsigned long) hardirq_stack[tp->cpu];
+		if (addr >= base &&
+		    addr <= (base + THREAD_SIZE - sizeof(*regs)))
+			goto check_magic;
+		base = (unsigned long) softirq_stack[tp->cpu];
+		if (addr >= base &&
+		    addr <= (base + THREAD_SIZE - sizeof(*regs)))
+			goto check_magic;
+	}
+	return false;
+
+check_magic:
+	if ((regs->magic & ~0x1ff) == PT_REGS_MAGIC)
+		return true;
+	return false;
+
+}
+
+#endif /* _KSTACK_H */
diff --git a/arch/sparc64/kernel/of_device.c b/arch/sparc64/kernel/of_device.c
index f8b50cbf4bf7..100ebd527499 100644
--- a/arch/sparc64/kernel/of_device.c
+++ b/arch/sparc64/kernel/of_device.c
@@ -96,7 +96,7 @@ struct of_bus {
 				       int *addrc, int *sizec);
 	int		(*map)(u32 *addr, const u32 *range,
 			       int na, int ns, int pna);
-	unsigned int	(*get_flags)(const u32 *addr);
+	unsigned long	(*get_flags)(const u32 *addr, unsigned long);
 };
 
 /*
@@ -156,8 +156,10 @@ static int of_bus_default_map(u32 *addr, const u32 *range,
 	return 0;
 }
 
-static unsigned int of_bus_default_get_flags(const u32 *addr)
+static unsigned long of_bus_default_get_flags(const u32 *addr, unsigned long flags)
 {
+	if (flags)
+		return flags;
 	return IORESOURCE_MEM;
 }
 
@@ -167,7 +169,7 @@ static unsigned int of_bus_default_get_flags(const u32 *addr)
 
 static int of_bus_pci_match(struct device_node *np)
 {
-	if (!strcmp(np->type, "pci") || !strcmp(np->type, "pciex")) {
+	if (!strcmp(np->name, "pci")) {
 		const char *model = of_get_property(np, "model", NULL);
 
 		if (model && !strcmp(model, "SUNW,simba"))
@@ -198,7 +200,7 @@ static int of_bus_simba_match(struct device_node *np)
 	/* Treat PCI busses lacking ranges property just like
 	 * simba.
 	 */
-	if (!strcmp(np->type, "pci") || !strcmp(np->type, "pciex")) {
+	if (!strcmp(np->name, "pci")) {
 		if (!of_find_property(np, "ranges", NULL))
 			return 1;
 	}
@@ -249,17 +251,21 @@ static int of_bus_pci_map(u32 *addr, const u32 *range,
 	return 0;
 }
 
-static unsigned int of_bus_pci_get_flags(const u32 *addr)
+static unsigned long of_bus_pci_get_flags(const u32 *addr, unsigned long flags)
 {
-	unsigned int flags = 0;
 	u32 w = addr[0];
 
+	/* For PCI, we override whatever child busses may have used.  */
+	flags = 0;
 	switch((w >> 24) & 0x03) {
 	case 0x01:
 		flags |= IORESOURCE_IO;
+		break;
+
 	case 0x02: /* 32 bits */
 	case 0x03: /* 64 bits */
 		flags |= IORESOURCE_MEM;
+		break;
 	}
 	if (w & 0x40000000)
 		flags |= IORESOURCE_PREFETCH;
@@ -423,7 +429,7 @@ static int __init use_1to1_mapping(struct device_node *pp)
 	 * it lacks a ranges property, and this will include
 	 * cases like Simba.
 	 */
-	if (!strcmp(pp->type, "pci") || !strcmp(pp->type, "pciex"))
+	if (!strcmp(pp->name, "pci"))
 		return 0;
 
 	return 1;
@@ -478,10 +484,10 @@ static void __init build_device_resources(struct of_device *op,
 		int pna, pns;
 
 		size = of_read_addr(reg + na, ns);
-		flags = bus->get_flags(reg);
-
 		memcpy(addr, reg, na * 4);
 
+		flags = bus->get_flags(addr, 0);
+
 		if (use_1to1_mapping(pp)) {
 			result = of_read_addr(addr, na);
 			goto build_res;
@@ -506,6 +512,8 @@ static void __init build_device_resources(struct of_device *op,
 					       dna, dns, pna))
 				break;
 
+			flags = pbus->get_flags(addr, flags);
+
 			dna = pna;
 			dns = pns;
 			dbus = pbus;
@@ -706,8 +714,7 @@ static unsigned int __init build_one_device_irq(struct of_device *op,
 				break;
 			}
 		} else {
-			if (!strcmp(pp->type, "pci") ||
-			    !strcmp(pp->type, "pciex")) {
+			if (!strcmp(pp->name, "pci")) {
 				unsigned int this_orig_irq = irq;
 
 				irq = pci_irq_swizzle(dp, pp, irq);
diff --git a/arch/sparc64/kernel/pci.c b/arch/sparc64/kernel/pci.c
index 55096195458f..80dad76f8b81 100644
--- a/arch/sparc64/kernel/pci.c
+++ b/arch/sparc64/kernel/pci.c
@@ -425,7 +425,7 @@ struct pci_dev *of_create_pci_dev(struct pci_pbm_info *pbm,
 	dev->current_state = 4;		/* unknown power state */
 	dev->error_state = pci_channel_io_normal;
 
-	if (!strcmp(type, "pci") || !strcmp(type, "pciex")) {
+	if (!strcmp(node->name, "pci")) {
 		/* a PCI-PCI bridge */
 		dev->hdr_type = PCI_HEADER_TYPE_BRIDGE;
 		dev->rom_base_reg = PCI_ROM_ADDRESS1;
diff --git a/arch/sparc64/kernel/pci_psycho.c b/arch/sparc64/kernel/pci_psycho.c
index ef5fe29202c2..f85b6bebb0be 100644
--- a/arch/sparc64/kernel/pci_psycho.c
+++ b/arch/sparc64/kernel/pci_psycho.c
@@ -575,7 +575,7 @@ static irqreturn_t psycho_pcierr_intr_other(struct pci_pbm_info *pbm, int is_pbm
 {
 	unsigned long csr_reg, csr, csr_error_bits;
 	irqreturn_t ret = IRQ_NONE;
-	u16 stat;
+	u16 stat, *addr;
 
 	if (is_pbm_a) {
 		csr_reg = pbm->controller_regs + PSYCHO_PCIA_CTRL;
@@ -597,7 +597,9 @@ static irqreturn_t psycho_pcierr_intr_other(struct pci_pbm_info *pbm, int is_pbm
 			printk("%s: PCI SERR signal asserted.\n", pbm->name);
 		ret = IRQ_HANDLED;
 	}
-	pci_read_config_word(pbm->pci_bus->self, PCI_STATUS, &stat);
+	addr = psycho_pci_config_mkaddr(pbm, pbm->pci_first_busno,
+					0, PCI_STATUS);
+	pci_config_read16(addr, &stat);
 	if (stat & (PCI_STATUS_PARITY |
 		    PCI_STATUS_SIG_TARGET_ABORT |
 		    PCI_STATUS_REC_TARGET_ABORT |
@@ -605,7 +607,7 @@ static irqreturn_t psycho_pcierr_intr_other(struct pci_pbm_info *pbm, int is_pbm
 		    PCI_STATUS_SIG_SYSTEM_ERROR)) {
 		printk("%s: PCI bus error, PCI_STATUS[%04x]\n",
 		       pbm->name, stat);
-		pci_write_config_word(pbm->pci_bus->self, PCI_STATUS, 0xffff);
+		pci_config_write16(addr, 0xffff);
 		ret = IRQ_HANDLED;
 	}
 	return ret;
@@ -744,16 +746,16 @@ static void psycho_register_error_handlers(struct pci_pbm_info *pbm)
 	 * the second will just error out since we do not pass in
 	 * IRQF_SHARED.
 	 */
-	err = request_irq(op->irqs[1], psycho_ue_intr, 0,
+	err = request_irq(op->irqs[1], psycho_ue_intr, IRQF_SHARED,
 			  "PSYCHO_UE", pbm);
-	err = request_irq(op->irqs[2], psycho_ce_intr, 0,
+	err = request_irq(op->irqs[2], psycho_ce_intr, IRQF_SHARED,
 			  "PSYCHO_CE", pbm);
 
 	/* This one, however, ought not to fail.  We can just warn
 	 * about it since the system can still operate properly even
 	 * if this fails.
 	 */
-	err = request_irq(op->irqs[0], psycho_pcierr_intr, 0,
+	err = request_irq(op->irqs[0], psycho_pcierr_intr, IRQF_SHARED,
 			  "PSYCHO_PCIERR", pbm);
 	if (err)
 		printk(KERN_WARNING "%s: Could not register PCIERR, "
diff --git a/arch/sparc64/kernel/process.c b/arch/sparc64/kernel/process.c
index 7f5debdc5fed..15f4178592e7 100644
--- a/arch/sparc64/kernel/process.c
+++ b/arch/sparc64/kernel/process.c
@@ -52,6 +52,8 @@
 #include <asm/irq_regs.h>
 #include <asm/smp.h>
 
+#include "kstack.h"
+
 static void sparc64_yield(int cpu)
 {
 	if (tlb_type != hypervisor)
@@ -235,19 +237,6 @@ void show_regs(struct pt_regs *regs)
 struct global_reg_snapshot global_reg_snapshot[NR_CPUS];
 static DEFINE_SPINLOCK(global_reg_snapshot_lock);
 
-static bool kstack_valid(struct thread_info *tp, struct reg_window *rw)
-{
-	unsigned long thread_base, fp;
-
-	thread_base = (unsigned long) tp;
-	fp = (unsigned long) rw;
-
-	if (fp < (thread_base + sizeof(struct thread_info)) ||
-	    fp >= (thread_base + THREAD_SIZE))
-		return false;
-	return true;
-}
-
 static void __global_reg_self(struct thread_info *tp, struct pt_regs *regs,
 			      int this_cpu)
 {
@@ -264,11 +253,11 @@ static void __global_reg_self(struct thread_info *tp, struct pt_regs *regs,
 
 		rw = (struct reg_window *)
 			(regs->u_regs[UREG_FP] + STACK_BIAS);
-		if (kstack_valid(tp, rw)) {
+		if (kstack_valid(tp, (unsigned long) rw)) {
 			global_reg_snapshot[this_cpu].i7 = rw->ins[7];
 			rw = (struct reg_window *)
 				(rw->ins[6] + STACK_BIAS);
-			if (kstack_valid(tp, rw))
+			if (kstack_valid(tp, (unsigned long) rw))
 				global_reg_snapshot[this_cpu].rpc = rw->ins[7];
 		}
 	} else {
@@ -828,7 +817,7 @@ out:
 unsigned long get_wchan(struct task_struct *task)
 {
 	unsigned long pc, fp, bias = 0;
-	unsigned long thread_info_base;
+	struct thread_info *tp;
 	struct reg_window *rw;
         unsigned long ret = 0;
 	int count = 0; 
@@ -837,14 +826,12 @@ unsigned long get_wchan(struct task_struct *task)
             task->state == TASK_RUNNING)
 		goto out;
 
-	thread_info_base = (unsigned long) task_stack_page(task);
+	tp = task_thread_info(task);
 	bias = STACK_BIAS;
 	fp = task_thread_info(task)->ksp + bias;
 
 	do {
-		/* Bogus frame pointer? */
-		if (fp < (thread_info_base + sizeof(struct thread_info)) ||
-		    fp >= (thread_info_base + THREAD_SIZE))
+		if (!kstack_valid(tp, fp))
 			break;
 		rw = (struct reg_window *) fp;
 		pc = rw->ins[7];
diff --git a/arch/sparc64/kernel/prom.c b/arch/sparc64/kernel/prom.c
index 3c048ac4e638..7151513f156e 100644
--- a/arch/sparc64/kernel/prom.c
+++ b/arch/sparc64/kernel/prom.c
@@ -156,55 +156,11 @@ static unsigned long psycho_pcislot_imap_offset(unsigned long ino)
 		return PSYCHO_IMAP_B_SLOT0 + (slot * 8);
 }
 
-#define PSYCHO_IMAP_SCSI	0x1000UL
-#define PSYCHO_IMAP_ETH		0x1008UL
-#define PSYCHO_IMAP_BPP		0x1010UL
-#define PSYCHO_IMAP_AU_REC	0x1018UL
-#define PSYCHO_IMAP_AU_PLAY	0x1020UL
-#define PSYCHO_IMAP_PFAIL	0x1028UL
-#define PSYCHO_IMAP_KMS		0x1030UL
-#define PSYCHO_IMAP_FLPY	0x1038UL
-#define PSYCHO_IMAP_SHW		0x1040UL
-#define PSYCHO_IMAP_KBD		0x1048UL
-#define PSYCHO_IMAP_MS		0x1050UL
-#define PSYCHO_IMAP_SER		0x1058UL
-#define PSYCHO_IMAP_TIM0	0x1060UL
-#define PSYCHO_IMAP_TIM1	0x1068UL
-#define PSYCHO_IMAP_UE		0x1070UL
-#define PSYCHO_IMAP_CE		0x1078UL
-#define PSYCHO_IMAP_A_ERR	0x1080UL
-#define PSYCHO_IMAP_B_ERR	0x1088UL
-#define PSYCHO_IMAP_PMGMT	0x1090UL
-#define PSYCHO_IMAP_GFX		0x1098UL
-#define PSYCHO_IMAP_EUPA	0x10a0UL
-
-static unsigned long __psycho_onboard_imap_off[] = {
-/*0x20*/	PSYCHO_IMAP_SCSI,
-/*0x21*/	PSYCHO_IMAP_ETH,
-/*0x22*/	PSYCHO_IMAP_BPP,
-/*0x23*/	PSYCHO_IMAP_AU_REC,
-/*0x24*/	PSYCHO_IMAP_AU_PLAY,
-/*0x25*/	PSYCHO_IMAP_PFAIL,
-/*0x26*/	PSYCHO_IMAP_KMS,
-/*0x27*/	PSYCHO_IMAP_FLPY,
-/*0x28*/	PSYCHO_IMAP_SHW,
-/*0x29*/	PSYCHO_IMAP_KBD,
-/*0x2a*/	PSYCHO_IMAP_MS,
-/*0x2b*/	PSYCHO_IMAP_SER,
-/*0x2c*/	PSYCHO_IMAP_TIM0,
-/*0x2d*/	PSYCHO_IMAP_TIM1,
-/*0x2e*/	PSYCHO_IMAP_UE,
-/*0x2f*/	PSYCHO_IMAP_CE,
-/*0x30*/	PSYCHO_IMAP_A_ERR,
-/*0x31*/	PSYCHO_IMAP_B_ERR,
-/*0x32*/	PSYCHO_IMAP_PMGMT,
-/*0x33*/	PSYCHO_IMAP_GFX,
-/*0x34*/	PSYCHO_IMAP_EUPA,
-};
+#define PSYCHO_OBIO_IMAP_BASE	0x1000UL
+
 #define PSYCHO_ONBOARD_IRQ_BASE		0x20
-#define PSYCHO_ONBOARD_IRQ_LAST		0x34
 #define psycho_onboard_imap_offset(__ino) \
-	__psycho_onboard_imap_off[(__ino) - PSYCHO_ONBOARD_IRQ_BASE]
+	(PSYCHO_OBIO_IMAP_BASE + (((__ino) & 0x1f) << 3))
 
 #define PSYCHO_ICLR_A_SLOT0	0x1400UL
 #define PSYCHO_ICLR_SCSI	0x1800UL
@@ -228,10 +184,6 @@ static unsigned int psycho_irq_build(struct device_node *dp,
 		imap_off = psycho_pcislot_imap_offset(ino);
 	} else {
 		/* Onboard device */
-		if (ino > PSYCHO_ONBOARD_IRQ_LAST) {
-			prom_printf("psycho_irq_build: Wacky INO [%x]\n", ino);
-			prom_halt();
-		}
 		imap_off = psycho_onboard_imap_offset(ino);
 	}
 
@@ -318,23 +270,6 @@ static void sabre_wsync_handler(unsigned int ino, void *_arg1, void *_arg2)
 
 #define SABRE_IMAP_A_SLOT0	0x0c00UL
 #define SABRE_IMAP_B_SLOT0	0x0c20UL
-#define SABRE_IMAP_SCSI		0x1000UL
-#define SABRE_IMAP_ETH		0x1008UL
-#define SABRE_IMAP_BPP		0x1010UL
-#define SABRE_IMAP_AU_REC	0x1018UL
-#define SABRE_IMAP_AU_PLAY	0x1020UL
-#define SABRE_IMAP_PFAIL	0x1028UL
-#define SABRE_IMAP_KMS		0x1030UL
-#define SABRE_IMAP_FLPY		0x1038UL
-#define SABRE_IMAP_SHW		0x1040UL
-#define SABRE_IMAP_KBD		0x1048UL
-#define SABRE_IMAP_MS		0x1050UL
-#define SABRE_IMAP_SER		0x1058UL
-#define SABRE_IMAP_UE		0x1070UL
-#define SABRE_IMAP_CE		0x1078UL
-#define SABRE_IMAP_PCIERR	0x1080UL
-#define SABRE_IMAP_GFX		0x1098UL
-#define SABRE_IMAP_EUPA		0x10a0UL
 #define SABRE_ICLR_A_SLOT0	0x1400UL
 #define SABRE_ICLR_B_SLOT0	0x1480UL
 #define SABRE_ICLR_SCSI		0x1800UL
@@ -364,33 +299,10 @@ static unsigned long sabre_pcislot_imap_offset(unsigned long ino)
 		return SABRE_IMAP_B_SLOT0 + (slot * 8);
 }
 
-static unsigned long __sabre_onboard_imap_off[] = {
-/*0x20*/	SABRE_IMAP_SCSI,
-/*0x21*/	SABRE_IMAP_ETH,
-/*0x22*/	SABRE_IMAP_BPP,
-/*0x23*/	SABRE_IMAP_AU_REC,
-/*0x24*/	SABRE_IMAP_AU_PLAY,
-/*0x25*/	SABRE_IMAP_PFAIL,
-/*0x26*/	SABRE_IMAP_KMS,
-/*0x27*/	SABRE_IMAP_FLPY,
-/*0x28*/	SABRE_IMAP_SHW,
-/*0x29*/	SABRE_IMAP_KBD,
-/*0x2a*/	SABRE_IMAP_MS,
-/*0x2b*/	SABRE_IMAP_SER,
-/*0x2c*/	0 /* reserved */,
-/*0x2d*/	0 /* reserved */,
-/*0x2e*/	SABRE_IMAP_UE,
-/*0x2f*/	SABRE_IMAP_CE,
-/*0x30*/	SABRE_IMAP_PCIERR,
-/*0x31*/	0 /* reserved */,
-/*0x32*/	0 /* reserved */,
-/*0x33*/	SABRE_IMAP_GFX,
-/*0x34*/	SABRE_IMAP_EUPA,
-};
-#define SABRE_ONBOARD_IRQ_BASE		0x20
-#define SABRE_ONBOARD_IRQ_LAST		0x30
+#define SABRE_OBIO_IMAP_BASE	0x1000UL
+#define SABRE_ONBOARD_IRQ_BASE	0x20
 #define sabre_onboard_imap_offset(__ino) \
-	__sabre_onboard_imap_off[(__ino) - SABRE_ONBOARD_IRQ_BASE]
+	(SABRE_OBIO_IMAP_BASE + (((__ino) & 0x1f) << 3))
 
 #define sabre_iclr_offset(ino)					      \
 	((ino & 0x20) ? (SABRE_ICLR_SCSI + (((ino) & 0x1f) << 3)) :  \
@@ -453,10 +365,6 @@ static unsigned int sabre_irq_build(struct device_node *dp,
 		imap_off = sabre_pcislot_imap_offset(ino);
 	} else {
 		/* onboard device */
-		if (ino > SABRE_ONBOARD_IRQ_LAST) {
-			prom_printf("sabre_irq_build: Wacky INO [%x]\n", ino);
-			prom_halt();
-		}
 		imap_off = sabre_onboard_imap_offset(ino);
 	}
 
diff --git a/arch/sparc64/kernel/ptrace.c b/arch/sparc64/kernel/ptrace.c
index bd578cc4856d..10306e476e38 100644
--- a/arch/sparc64/kernel/ptrace.c
+++ b/arch/sparc64/kernel/ptrace.c
@@ -443,7 +443,7 @@ static const struct user_regset sparc64_regsets[] = {
 	 */
 	[REGSET_GENERAL] = {
 		.core_note_type = NT_PRSTATUS,
-		.n = 36 * sizeof(u64),
+		.n = 36,
 		.size = sizeof(u64), .align = sizeof(u64),
 		.get = genregs64_get, .set = genregs64_set
 	},
@@ -455,7 +455,7 @@ static const struct user_regset sparc64_regsets[] = {
 	 */
 	[REGSET_FP] = {
 		.core_note_type = NT_PRFPREG,
-		.n = 35 * sizeof(u64),
+		.n = 35,
 		.size = sizeof(u64), .align = sizeof(u64),
 		.get = fpregs64_get, .set = fpregs64_set
 	},
@@ -801,7 +801,7 @@ static const struct user_regset sparc32_regsets[] = {
 	 */
 	[REGSET_GENERAL] = {
 		.core_note_type = NT_PRSTATUS,
-		.n = 38 * sizeof(u32),
+		.n = 38,
 		.size = sizeof(u32), .align = sizeof(u32),
 		.get = genregs32_get, .set = genregs32_set
 	},
@@ -817,7 +817,7 @@ static const struct user_regset sparc32_regsets[] = {
 	 */
 	[REGSET_FP] = {
 		.core_note_type = NT_PRFPREG,
-		.n = 99 * sizeof(u32),
+		.n = 99,
 		.size = sizeof(u32), .align = sizeof(u32),
 		.get = fpregs32_get, .set = fpregs32_set
 	},
diff --git a/arch/sparc64/kernel/smp.c b/arch/sparc64/kernel/smp.c
index 27b81775a4de..2be166c544ca 100644
--- a/arch/sparc64/kernel/smp.c
+++ b/arch/sparc64/kernel/smp.c
@@ -80,8 +80,6 @@ void smp_bogo(struct seq_file *m)
 			   i, cpu_data(i).clock_tick);
 }
 
-static __cacheline_aligned_in_smp DEFINE_SPINLOCK(call_lock);
-
 extern void setup_sparc64_timer(void);
 
 static volatile unsigned long callin_flag = 0;
@@ -120,9 +118,9 @@ void __cpuinit smp_callin(void)
 	while (!cpu_isset(cpuid, smp_commenced_mask))
 		rmb();
 
-	spin_lock(&call_lock);
+	ipi_call_lock();
 	cpu_set(cpuid, cpu_online_map);
-	spin_unlock(&call_lock);
+	ipi_call_unlock();
 
 	/* idle thread is expected to have preempt disabled */
 	preempt_disable();
@@ -858,9 +856,7 @@ void smp_tsb_sync(struct mm_struct *mm)
 extern unsigned long xcall_flush_tlb_mm;
 extern unsigned long xcall_flush_tlb_pending;
 extern unsigned long xcall_flush_tlb_kernel_range;
-#ifdef CONFIG_MAGIC_SYSRQ
 extern unsigned long xcall_fetch_glob_regs;
-#endif
 extern unsigned long xcall_receive_signal;
 extern unsigned long xcall_new_mmu_context_version;
 #ifdef CONFIG_KGDB
@@ -1005,12 +1001,10 @@ void kgdb_roundup_cpus(unsigned long flags)
 }
 #endif
 
-#ifdef CONFIG_MAGIC_SYSRQ
 void smp_fetch_global_regs(void)
 {
 	smp_cross_call(&xcall_fetch_glob_regs, 0, 0, 0);
 }
-#endif
 
 /* We know that the window frames of the user have been flushed
  * to the stack before we get here because all callers of us
@@ -1309,10 +1303,6 @@ int __cpu_disable(void)
 	c->core_id = 0;
 	c->proc_id = -1;
 
-	spin_lock(&call_lock);
-	cpu_clear(cpu, cpu_online_map);
-	spin_unlock(&call_lock);
-
 	smp_wmb();
 
 	/* Make sure no interrupts point to this cpu.  */
@@ -1322,6 +1312,10 @@ int __cpu_disable(void)
 	mdelay(1);
 	local_irq_disable();
 
+	ipi_call_lock();
+	cpu_clear(cpu, cpu_online_map);
+	ipi_call_unlock();
+
 	return 0;
 }
 
diff --git a/arch/sparc64/kernel/stacktrace.c b/arch/sparc64/kernel/stacktrace.c
index e9d7f0660f2e..4e21d4a57d3b 100644
--- a/arch/sparc64/kernel/stacktrace.c
+++ b/arch/sparc64/kernel/stacktrace.c
@@ -5,10 +5,12 @@
 #include <asm/ptrace.h>
 #include <asm/stacktrace.h>
 
+#include "kstack.h"
+
 void save_stack_trace(struct stack_trace *trace)
 {
-	unsigned long ksp, fp, thread_base;
 	struct thread_info *tp = task_thread_info(current);
+	unsigned long ksp, fp;
 
 	stack_trace_flush();
 
@@ -18,23 +20,18 @@ void save_stack_trace(struct stack_trace *trace)
 	);
 
 	fp = ksp + STACK_BIAS;
-	thread_base = (unsigned long) tp;
 	do {
 		struct sparc_stackf *sf;
 		struct pt_regs *regs;
 		unsigned long pc;
 
-		/* Bogus frame pointer? */
-		if (fp < (thread_base + sizeof(struct thread_info)) ||
-		    fp > (thread_base + THREAD_SIZE - sizeof(struct sparc_stackf)))
+		if (!kstack_valid(tp, fp))
 			break;
 
 		sf = (struct sparc_stackf *) fp;
 		regs = (struct pt_regs *) (sf + 1);
 
-		if (((unsigned long)regs <=
-		     (thread_base + THREAD_SIZE - sizeof(*regs))) &&
-		    (regs->magic & ~0x1ff) == PT_REGS_MAGIC) {
+		if (kstack_is_trap_frame(tp, regs)) {
 			if (!(regs->tstate & TSTATE_PRIV))
 				break;
 			pc = regs->tpc;
diff --git a/arch/sparc64/kernel/traps.c b/arch/sparc64/kernel/traps.c
index 404e8561e2d0..c824df13f589 100644
--- a/arch/sparc64/kernel/traps.c
+++ b/arch/sparc64/kernel/traps.c
@@ -10,6 +10,7 @@
 
 #include <linux/module.h>
 #include <linux/sched.h>
+#include <linux/linkage.h>
 #include <linux/kernel.h>
 #include <linux/signal.h>
 #include <linux/smp.h>
@@ -39,6 +40,7 @@
 #include <asm/prom.h>
 
 #include "entry.h"
+#include "kstack.h"
 
 /* When an irrecoverable trap occurs at tl > 0, the trap entry
  * code logs the trap state registers at every level in the trap
@@ -2115,14 +2117,12 @@ void show_stack(struct task_struct *tsk, unsigned long *_ksp)
 		struct pt_regs *regs;
 		unsigned long pc;
 
-		/* Bogus frame pointer? */
-		if (fp < (thread_base + sizeof(struct thread_info)) ||
-		    fp >= (thread_base + THREAD_SIZE))
+		if (!kstack_valid(tp, fp))
 			break;
 		sf = (struct sparc_stackf *) fp;
 		regs = (struct pt_regs *) (sf + 1);
 
-		if ((regs->magic & ~0x1ff) == PT_REGS_MAGIC) {
+		if (kstack_is_trap_frame(tp, regs)) {
 			if (!(regs->tstate & TSTATE_PRIV))
 				break;
 			pc = regs->tpc;
@@ -2454,7 +2454,7 @@ struct trap_per_cpu trap_block[NR_CPUS];
 /* This can get invoked before sched_init() so play it super safe
  * and use hard_smp_processor_id().
  */
-void init_cur_cpu_trap(struct thread_info *t)
+void notrace init_cur_cpu_trap(struct thread_info *t)
 {
 	int cpu = hard_smp_processor_id();
 	struct trap_per_cpu *p = &trap_block[cpu];
diff --git a/arch/sparc64/lib/mcount.S b/arch/sparc64/lib/mcount.S
index 7735a7a60533..fad90ddb3a28 100644
--- a/arch/sparc64/lib/mcount.S
+++ b/arch/sparc64/lib/mcount.S
@@ -48,12 +48,45 @@ mcount:
 	sub		%g3, STACK_BIAS, %g3
 	cmp		%sp, %g3
 	bg,pt		%xcc, 1f
-	 sethi		%hi(panicstring), %g3
+	 nop
+	lduh		[%g6 + TI_CPU], %g1
+	sethi		%hi(hardirq_stack), %g3
+	or		%g3, %lo(hardirq_stack), %g3
+	sllx		%g1, 3, %g1
+	ldx		[%g3 + %g1], %g7
+	sub		%g7, STACK_BIAS, %g7
+	cmp		%sp, %g7
+	bleu,pt		%xcc, 2f
+	 sethi		%hi(THREAD_SIZE), %g3
+	add		%g7, %g3, %g7
+	cmp		%sp, %g7
+	blu,pn		%xcc, 1f
+2:	 sethi		%hi(softirq_stack), %g3
+	or		%g3, %lo(softirq_stack), %g3
+	ldx		[%g3 + %g1], %g7
+	cmp		%sp, %g7
+	bleu,pt		%xcc, 2f
+	 sethi		%hi(THREAD_SIZE), %g3
+	add		%g7, %g3, %g7
+	cmp		%sp, %g7
+	blu,pn		%xcc, 1f
+	 nop
+	/* If we are already on ovstack, don't hop onto it
+	 * again, we are already trying to output the stack overflow
+	 * message.
+	 */
 	sethi		%hi(ovstack), %g7		! cant move to panic stack fast enough
 	 or		%g7, %lo(ovstack), %g7
-	add		%g7, OVSTACKSIZE, %g7
+	add		%g7, OVSTACKSIZE, %g3
+	sub		%g3, STACK_BIAS + 192, %g3
 	sub		%g7, STACK_BIAS, %g7
-	mov		%g7, %sp
+	cmp		%sp, %g7
+	blu,pn		%xcc, 2f
+	 cmp		%sp, %g3
+	bleu,pn		%xcc, 1f
+	 nop
+2:	mov		%g3, %sp
+	sethi		%hi(panicstring), %g3
 	call		prom_printf
 	 or		%g3, %lo(panicstring), %o0
 	call		prom_halt
diff --git a/arch/sparc64/mm/init.c b/arch/sparc64/mm/init.c
index 4e821b3ecb03..a41df7bef035 100644
--- a/arch/sparc64/mm/init.c
+++ b/arch/sparc64/mm/init.c
@@ -49,6 +49,7 @@
 #include <asm/sstate.h>
 #include <asm/mdesc.h>
 #include <asm/cpudata.h>
+#include <asm/irq.h>
 
 #define MAX_PHYS_ADDRESS	(1UL << 42UL)
 #define KPTE_BITMAP_CHUNK_SZ	(256UL * 1024UL * 1024UL)
@@ -795,6 +796,9 @@ static unsigned long nid_range(unsigned long start, unsigned long end,
 		start += PAGE_SIZE;
 	}
 
+	if (start > end)
+		start = end;
+
 	return start;
 }
 #else
@@ -1722,8 +1726,7 @@ void __init paging_init(void)
 
 	find_ramdisk(phys_base);
 
-	if (cmdline_memory_size)
-		lmb_enforce_memory_limit(phys_base + cmdline_memory_size);
+	lmb_enforce_memory_limit(cmdline_memory_size);
 
 	lmb_analyze();
 	lmb_dump_all();
@@ -1771,6 +1774,16 @@ void __init paging_init(void)
 	if (tlb_type == hypervisor)
 		sun4v_mdesc_init();
 
+	/* Once the OF device tree and MDESC have been setup, we know
+	 * the list of possible cpus.  Therefore we can allocate the
+	 * IRQ stacks.
+	 */
+	for_each_possible_cpu(i) {
+		/* XXX Use node local allocations... XXX */
+		softirq_stack[i] = __va(lmb_alloc(THREAD_SIZE, THREAD_SIZE));
+		hardirq_stack[i] = __va(lmb_alloc(THREAD_SIZE, THREAD_SIZE));
+	}
+
 	/* Setup bootmem... */
 	last_valid_pfn = end_pfn = bootmem_init(phys_base);
 
@@ -1830,7 +1843,7 @@ static int pavail_rescan_ents __initdata;
  * memory list again, and make sure it provides at least as much
  * memory as 'pavail' does.
  */
-static void setup_valid_addr_bitmap_from_pavail(void)
+static void __init setup_valid_addr_bitmap_from_pavail(void)
 {
 	int i;
 
@@ -1950,6 +1963,15 @@ void __init mem_init(void)
 void free_initmem(void)
 {
 	unsigned long addr, initend;
+	int do_free = 1;
+
+	/* If the physical memory maps were trimmed by kernel command
+	 * line options, don't even try freeing this initmem stuff up.
+	 * The kernel image could have been in the trimmed out region
+	 * and if so the freeing below will free invalid page structs.
+	 */
+	if (cmdline_memory_size)
+		do_free = 0;
 
 	/*
 	 * The init section is aligned to 8k in vmlinux.lds. Page align for >8k pagesizes.
@@ -1964,13 +1986,16 @@ void free_initmem(void)
 			((unsigned long) __va(kern_base)) -
 			((unsigned long) KERNBASE));
 		memset((void *)addr, POISON_FREE_INITMEM, PAGE_SIZE);
-		p = virt_to_page(page);
 
-		ClearPageReserved(p);
-		init_page_count(p);
-		__free_page(p);
-		num_physpages++;
-		totalram_pages++;
+		if (do_free) {
+			p = virt_to_page(page);
+
+			ClearPageReserved(p);
+			init_page_count(p);
+			__free_page(p);
+			num_physpages++;
+			totalram_pages++;
+		}
 	}
 }
 
diff --git a/arch/sparc64/mm/ultra.S b/arch/sparc64/mm/ultra.S
index ff1dc44d363e..86773e89dc1b 100644
--- a/arch/sparc64/mm/ultra.S
+++ b/arch/sparc64/mm/ultra.S
@@ -480,7 +480,6 @@ xcall_sync_tick:
 	b		rtrap_xcall
 	 ldx		[%sp + PTREGS_OFF + PT_V9_TSTATE], %l1
 
-#ifdef CONFIG_MAGIC_SYSRQ
 	.globl		xcall_fetch_glob_regs
 xcall_fetch_glob_regs:
 	sethi		%hi(global_reg_snapshot), %g1
@@ -511,7 +510,6 @@ xcall_fetch_glob_regs:
 	membar		#StoreStore
 	stx		%g3, [%g1 + GR_SNAP_THREAD]
 	retry
-#endif /* CONFIG_MAGIC_SYSRQ */
 
 #ifdef DCACHE_ALIASING_POSSIBLE
 	.align		32