Merge branch 'master' of master.kernel.org:/pub/scm/linux/kernel/git/davem/sparc-2.6

* 'master' of master.kernel.org:/pub/scm/linux/kernel/git/davem/sparc-2.6: (26 commits) [SPARC64]: Fix UP build. [SPARC64]: dr-cpu unconfigure support. [SERIAL]: Fix console write locking in sparc drivers. [SPARC64]: Give more accurate errors in dr_cpu_configure(). [SPARC64]: Clear cpu_{core,sibling}_map[] in smp_fill_in_sib_core_maps() [SPARC64]: Fix leak when DR added cpu does not bootup. [SPARC64]: Add ->set_affinity IRQ handlers. [SPARC64]: Process dr-cpu events in a kthread instead of workqueue. [SPARC64]: More sensible udelay implementation. [SPARC64]: SMP build fixes. [SPARC64]: mdesc.c needs linux/mm.h [SPARC64]: Fix build regressions added by dr-cpu changes. [SPARC64]: Unconditionally register vio_bus_type. [SPARC64]: Initial LDOM cpu hotplug support. [SPARC64]: Fix setting of variables in LDOM guest. [SPARC64]: Fix MD property lifetime bugs. [SPARC64]: Abstract out mdesc accesses for better MD update handling. [SPARC64]: Use more mearningful names for IRQ registry. [SPARC64]: Initial domain-services driver. [SPARC64]: Export powerd facilities for external entities. ...
2026-05-01 15:00:59 -07:00 · 2007-07-16 10:45:23 -07:00
parent b91cba52e9 d54bc2793e
commit 02b2318e07
45 changed files with 8619 additions and 547 deletions
@@ -108,6 +108,15 @@ config SECCOMP

 source kernel/Kconfig.hz

+config HOTPLUG_CPU
+	bool "Support for hot-pluggable CPUs"
+	depends on SMP
+	select HOTPLUG
+	---help---
+	  Say Y here to experiment with turning CPUs off and on.  CPUs
+	  can be controlled through /sys/devices/system/cpu/cpu#.
+	  Say N if you want to disable CPU hotplug.
+
 source "init/Kconfig"

 config SYSVIPC_COMPAT
@@ -305,6 +314,12 @@ config SUN_IO
 	bool
 	default y

+config SUN_LDOMS
+	bool "Sun Logical Domains support"
+	help
+	  Say Y here is you want to support virtual devices via
+	  Logical Domains.
+
 config PCI
 	bool "PCI support"
 	select ARCH_SUPPORTS_MSI
@@ -18,7 +18,7 @@ obj-$(CONFIG_STACKTRACE) += stacktrace.o
 obj-$(CONFIG_PCI)	 += ebus.o isa.o pci_common.o pci_iommu.o \
 			    pci_psycho.o pci_sabre.o pci_schizo.o \
 			    pci_sun4v.o pci_sun4v_asm.o pci_fire.o
-obj-$(CONFIG_SMP)	 += smp.o trampoline.o
+obj-$(CONFIG_SMP)	 += smp.o trampoline.o hvtramp.o
 obj-$(CONFIG_SPARC32_COMPAT) += sys32.o sys_sparc32.o signal32.o
 obj-$(CONFIG_BINFMT_ELF32) += binfmt_elf32.o
 obj-$(CONFIG_BINFMT_AOUT32) += binfmt_aout32.o
@@ -26,6 +26,7 @@ obj-$(CONFIG_MODULES) += module.o
 obj-$(CONFIG_US3_FREQ) += us3_cpufreq.o
 obj-$(CONFIG_US2E_FREQ) += us2e_cpufreq.o
 obj-$(CONFIG_KPROBES) += kprobes.o
+obj-$(CONFIG_SUN_LDOMS) += ldc.o vio.o viohs.o ds.o
 obj-$(CONFIG_AUDIT) += audit.o
 obj-$(CONFIG_AUDIT)$(CONFIG_SPARC32_COMPAT) += compat_audit.o
 obj-y += $(obj-yy)
@@ -0,0 +1,139 @@
+/* hvtramp.S: Hypervisor start-cpu trampoline code.
+ *
+ * Copyright (C) 2007 David S. Miller <davem@davemloft.net>
+ */
+
+#include <asm/thread_info.h>
+#include <asm/hypervisor.h>
+#include <asm/scratchpad.h>
+#include <asm/spitfire.h>
+#include <asm/hvtramp.h>
+#include <asm/pstate.h>
+#include <asm/ptrace.h>
+#include <asm/asi.h>
+
+	.text
+	.align		8
+	.globl		hv_cpu_startup, hv_cpu_startup_end
+
+	/* This code executes directly out of the hypervisor
+	 * with physical addressing (va==pa).  %o0 contains
+	 * our client argument which for Linux points to
+	 * a descriptor data structure which defines the
+	 * MMU entries we need to load up.
+	 *
+	 * After we set things up we enable the MMU and call
+	 * into the kernel.
+	 *
+	 * First setup basic privileged cpu state.
+	 */
+hv_cpu_startup:
+	wrpr		%g0, 0, %gl
+	wrpr		%g0, 15, %pil
+	wrpr		%g0, 0, %canrestore
+	wrpr		%g0, 0, %otherwin
+	wrpr		%g0, 6, %cansave
+	wrpr		%g0, 6, %cleanwin
+	wrpr		%g0, 0, %cwp
+	wrpr		%g0, 0, %wstate
+	wrpr		%g0, 0, %tl
+
+	sethi		%hi(sparc64_ttable_tl0), %g1
+	wrpr		%g1, %tba
+
+	mov		%o0, %l0
+
+	lduw		[%l0 + HVTRAMP_DESCR_CPU], %g1
+	mov		SCRATCHPAD_CPUID, %g2
+	stxa		%g1, [%g2] ASI_SCRATCHPAD
+
+	ldx		[%l0 + HVTRAMP_DESCR_FAULT_INFO_VA], %g2
+	stxa		%g2, [%g0] ASI_SCRATCHPAD
+
+	mov		0, %l1
+	lduw		[%l0 + HVTRAMP_DESCR_NUM_MAPPINGS], %l2
+	add		%l0, HVTRAMP_DESCR_MAPS, %l3
+
+1:	ldx		[%l3 + HVTRAMP_MAPPING_VADDR], %o0
+	clr		%o1
+	ldx		[%l3 + HVTRAMP_MAPPING_TTE], %o2
+	mov		HV_MMU_IMMU | HV_MMU_DMMU, %o3
+	mov		HV_FAST_MMU_MAP_PERM_ADDR, %o5
+	ta		HV_FAST_TRAP
+
+	brnz,pn		%o0, 80f
+	 nop
+
+	add		%l1, 1, %l1
+	cmp		%l1, %l2
+	blt,a,pt	%xcc, 1b
+	 add		%l3, HVTRAMP_MAPPING_SIZE, %l3
+
+	ldx		[%l0 + HVTRAMP_DESCR_FAULT_INFO_PA], %o0
+	mov		HV_FAST_MMU_FAULT_AREA_CONF, %o5
+	ta		HV_FAST_TRAP
+
+	brnz,pn		%o0, 80f
+	 nop
+
+	wrpr		%g0, (PSTATE_PRIV | PSTATE_PEF), %pstate
+
+	ldx		[%l0 + HVTRAMP_DESCR_THREAD_REG], %l6
+
+	mov		1, %o0
+	set		1f, %o1
+	mov		HV_FAST_MMU_ENABLE, %o5
+	ta		HV_FAST_TRAP
+
+	ba,pt		%xcc, 80f
+	 nop
+
+1:
+	wr		%g0, 0, %fprs
+	wr		%g0, ASI_P, %asi
+
+	mov		PRIMARY_CONTEXT, %g7
+	stxa		%g0, [%g7] ASI_MMU
+	membar		#Sync
+
+	mov		SECONDARY_CONTEXT, %g7
+	stxa		%g0, [%g7] ASI_MMU
+	membar		#Sync
+
+	mov		%l6, %g6
+	ldx		[%g6 + TI_TASK], %g4
+
+	mov		1, %g5
+	sllx		%g5, THREAD_SHIFT, %g5
+	sub		%g5, (STACKFRAME_SZ + STACK_BIAS), %g5
+	add		%g6, %g5, %sp
+	mov		0, %fp
+
+	call		init_irqwork_curcpu
+	 nop
+	call		hard_smp_processor_id
+	 nop
+
+	mov		%o0, %o1
+	mov		0, %o0
+	mov		0, %o2
+	call		sun4v_init_mondo_queues
+	 mov		1, %o3
+
+	call		init_cur_cpu_trap
+	 mov		%g6, %o0
+
+	wrpr		%g0, (PSTATE_PRIV | PSTATE_PEF | PSTATE_IE), %pstate
+
+	call		smp_callin
+	 nop
+	call		cpu_idle
+	 mov		0, %o0
+	call		cpu_panic
+	 nop
+
+80:	ba,pt		%xcc, 80b
+	 nop
+
+	.align		8
+hv_cpu_startup_end:
@@ -293,6 +293,11 @@ static void sun4u_irq_enable(unsigned int virt_irq)
 	}
 }

+static void sun4u_set_affinity(unsigned int virt_irq, cpumask_t mask)
+{
+	sun4u_irq_enable(virt_irq);
+}
+
 static void sun4u_irq_disable(unsigned int virt_irq)
 {
 	struct irq_handler_data *data = get_irq_chip_data(virt_irq);
@@ -309,6 +314,10 @@ static void sun4u_irq_disable(unsigned int virt_irq)
 static void sun4u_irq_end(unsigned int virt_irq)
 {
 	struct irq_handler_data *data = get_irq_chip_data(virt_irq);
+	struct irq_desc *desc = irq_desc + virt_irq;
+
+	if (unlikely(desc->status & (IRQ_DISABLED|IRQ_INPROGRESS)))
+		return;

 	if (likely(data))
 		upa_writeq(ICLR_IDLE, data->iclr);
@@ -340,6 +349,24 @@ static void sun4v_irq_enable(unsigned int virt_irq)
 	}
 }

+static void sun4v_set_affinity(unsigned int virt_irq, cpumask_t mask)
+{
+	struct ino_bucket *bucket = virt_irq_to_bucket(virt_irq);
+	unsigned int ino = bucket - &ivector_table[0];
+
+	if (likely(bucket)) {
+		unsigned long cpuid;
+		int err;
+
+		cpuid = irq_choose_cpu(virt_irq);
+
+		err = sun4v_intr_settarget(ino, cpuid);
+		if (err != HV_EOK)
+			printk("sun4v_intr_settarget(%x,%lu): err(%d)\n",
+			       ino, cpuid, err);
+	}
+}
+
 static void sun4v_irq_disable(unsigned int virt_irq)
 {
 	struct ino_bucket *bucket = virt_irq_to_bucket(virt_irq);
@@ -373,6 +400,10 @@ static void sun4v_irq_end(unsigned int virt_irq)
 {
 	struct ino_bucket *bucket = virt_irq_to_bucket(virt_irq);
 	unsigned int ino = bucket - &ivector_table[0];
+	struct irq_desc *desc = irq_desc + virt_irq;
+
+	if (unlikely(desc->status & (IRQ_DISABLED|IRQ_INPROGRESS)))
+		return;

 	if (likely(bucket)) {
 		int err;
@@ -418,6 +449,28 @@ static void sun4v_virq_enable(unsigned int virt_irq)
 	}
 }

+static void sun4v_virt_set_affinity(unsigned int virt_irq, cpumask_t mask)
+{
+	struct ino_bucket *bucket = virt_irq_to_bucket(virt_irq);
+	unsigned int ino = bucket - &ivector_table[0];
+
+	if (likely(bucket)) {
+		unsigned long cpuid, dev_handle, dev_ino;
+		int err;
+
+		cpuid = irq_choose_cpu(virt_irq);
+
+		dev_handle = ino & IMAP_IGN;
+		dev_ino = ino & IMAP_INO;
+
+		err = sun4v_vintr_set_target(dev_handle, dev_ino, cpuid);
+		if (err != HV_EOK)
+			printk("sun4v_vintr_set_target(%lx,%lx,%lu): "
+			       "err(%d)\n",
+			       dev_handle, dev_ino, cpuid, err);
+	}
+}
+
 static void sun4v_virq_disable(unsigned int virt_irq)
 {
 	struct ino_bucket *bucket = virt_irq_to_bucket(virt_irq);
@@ -443,6 +496,10 @@ static void sun4v_virq_end(unsigned int virt_irq)
 {
 	struct ino_bucket *bucket = virt_irq_to_bucket(virt_irq);
 	unsigned int ino = bucket - &ivector_table[0];
+	struct irq_desc *desc = irq_desc + virt_irq;
+
+	if (unlikely(desc->status & (IRQ_DISABLED|IRQ_INPROGRESS)))
+		return;

 	if (likely(bucket)) {
 		unsigned long dev_handle, dev_ino;
@@ -477,6 +534,7 @@ static struct irq_chip sun4u_irq = {
 	.enable		= sun4u_irq_enable,
 	.disable	= sun4u_irq_disable,
 	.end		= sun4u_irq_end,
+	.set_affinity	= sun4u_set_affinity,
 };

 static struct irq_chip sun4u_irq_ack = {
@@ -485,6 +543,7 @@ static struct irq_chip sun4u_irq_ack = {
 	.disable	= sun4u_irq_disable,
 	.ack		= run_pre_handler,
 	.end		= sun4u_irq_end,
+	.set_affinity	= sun4u_set_affinity,
 };

 static struct irq_chip sun4v_irq = {
@@ -492,6 +551,7 @@ static struct irq_chip sun4v_irq = {
 	.enable		= sun4v_irq_enable,
 	.disable	= sun4v_irq_disable,
 	.end		= sun4v_irq_end,
+	.set_affinity	= sun4v_set_affinity,
 };

 static struct irq_chip sun4v_irq_ack = {
@@ -500,6 +560,7 @@ static struct irq_chip sun4v_irq_ack = {
 	.disable	= sun4v_irq_disable,
 	.ack		= run_pre_handler,
 	.end		= sun4v_irq_end,
+	.set_affinity	= sun4v_set_affinity,
 };

 #ifdef CONFIG_PCI_MSI
@@ -511,6 +572,7 @@ static struct irq_chip sun4v_msi = {
 	.disable	= sun4v_msi_disable,
 	.ack		= run_pre_handler,
 	.end		= sun4v_irq_end,
+	.set_affinity	= sun4v_set_affinity,
 };
 #endif

@@ -519,6 +581,7 @@ static struct irq_chip sun4v_virq = {
 	.enable		= sun4v_virq_enable,
 	.disable	= sun4v_virq_disable,
 	.end		= sun4v_virq_end,
+	.set_affinity	= sun4v_virt_set_affinity,
 };

 static struct irq_chip sun4v_virq_ack = {
@@ -527,6 +590,7 @@ static struct irq_chip sun4v_virq_ack = {
 	.disable	= sun4v_virq_disable,
 	.ack		= run_pre_handler,
 	.end		= sun4v_virq_end,
+	.set_affinity	= sun4v_virt_set_affinity,
 };

 void irq_install_pre_handler(int virt_irq,
@@ -739,6 +803,26 @@ void handler_irq(int irq, struct pt_regs *regs)
 	set_irq_regs(old_regs);
 }

+#ifdef CONFIG_HOTPLUG_CPU
+void fixup_irqs(void)
+{
+	unsigned int irq;
+
+	for (irq = 0; irq < NR_IRQS; irq++) {
+		unsigned long flags;
+
+		spin_lock_irqsave(&irq_desc[irq].lock, flags);
+		if (irq_desc[irq].action &&
+		    !(irq_desc[irq].status & IRQ_PER_CPU)) {
+			if (irq_desc[irq].chip->set_affinity)
+				irq_desc[irq].chip->set_affinity(irq,
+					irq_desc[irq].affinity);
+		}
+		spin_unlock_irqrestore(&irq_desc[irq].lock, flags);
+	}
+}
+#endif
+
 struct sun5_timer {
 	u64	count0;
 	u64	limit0;
@@ -1,7 +1,6 @@
-/* $Id: power.c,v 1.10 2001/12/11 01:57:16 davem Exp $
- * power.c: Power management driver.
+/* power.c: Power management driver.
 *
- * Copyright (C) 1999 David S. Miller (davem@redhat.com)
+ * Copyright (C) 1999, 2007 David S. Miller (davem@davemloft.net)
 */

 #include <linux/kernel.h>
@@ -19,6 +18,7 @@
 #include <asm/prom.h>
 #include <asm/of_device.h>
 #include <asm/io.h>
+#include <asm/power.h>
 #include <asm/sstate.h>

 #include <linux/unistd.h>
@@ -29,24 +29,26 @@
 */
 int scons_pwroff = 1; 

-#ifdef CONFIG_PCI
-#include <linux/pci.h>
 static void __iomem *power_reg;

 static DECLARE_WAIT_QUEUE_HEAD(powerd_wait);
 static int button_pressed;

-static irqreturn_t power_handler(int irq, void *dev_id)
+void wake_up_powerd(void)
 {
 	if (button_pressed == 0) {
 		button_pressed = 1;
 		wake_up(&powerd_wait);
 	}
+}
+
+static irqreturn_t power_handler(int irq, void *dev_id)
+{
+	wake_up_powerd();

 	/* FIXME: Check registers for status... */
 	return IRQ_HANDLED;
 }
-#endif /* CONFIG_PCI */

 extern void machine_halt(void);
 extern void machine_alt_power_off(void);
@@ -56,19 +58,18 @@ void machine_power_off(void)
 {
 	sstate_poweroff();
 	if (!serial_console || scons_pwroff) {
-#ifdef CONFIG_PCI
 		if (power_reg) {
 			/* Both register bits seem to have the
 			 * same effect, so until I figure out
 			 * what the difference is...
 			 */
 			writel(AUXIO_PCIO_CPWR_OFF | AUXIO_PCIO_SPWR_OFF, power_reg);
-		} else
-#endif /* CONFIG_PCI */
+		} else {
 			if (poweroff_method != NULL) {
 				poweroff_method();
 				/* not reached */
 			}
+		}
 	}
 	machine_halt();
 }
@@ -76,7 +77,6 @@ void machine_power_off(void)
 void (*pm_power_off)(void) = machine_power_off;
 EXPORT_SYMBOL(pm_power_off);

-#ifdef CONFIG_PCI
 static int powerd(void *__unused)
 {
 	static char *envp[] = { "HOME=/", "TERM=linux", "PATH=/sbin:/usr/sbin:/bin:/usr/bin", NULL };
@@ -86,7 +86,7 @@ static int powerd(void *__unused)
 	daemonize("powerd");

 	add_wait_queue(&powerd_wait, &wait);
-again:
+
 	for (;;) {
 		set_task_state(current, TASK_INTERRUPTIBLE);
 		if (button_pressed)
@@ -100,16 +100,28 @@ again:
 	/* Ok, down we go... */
 	button_pressed = 0;
 	if (kernel_execve("/sbin/shutdown", argv, envp) < 0) {
-		printk("powerd: shutdown execution failed\n");
-		add_wait_queue(&powerd_wait, &wait);
-		goto again;
+		printk(KERN_ERR "powerd: shutdown execution failed\n");
+		machine_power_off();
 	}
 	return 0;
 }

+int start_powerd(void)
+{
+	int err;
+
+	err = kernel_thread(powerd, NULL, CLONE_FS);
+	if (err < 0)
+		printk(KERN_ERR "power: Failed to start power daemon.\n");
+	else
+		printk(KERN_INFO "power: powerd running.\n");
+
+	return err;
+}
+
 static int __init has_button_interrupt(unsigned int irq, struct device_node *dp)
 {
-	if (irq == PCI_IRQ_NONE)
+	if (irq == 0xffffffff)
 		return 0;
 	if (!of_find_property(dp, "button", NULL))
 		return 0;
@@ -130,17 +142,14 @@ static int __devinit power_probe(struct of_device *op, const struct of_device_id
 	poweroff_method = machine_halt;  /* able to use the standard halt */

 	if (has_button_interrupt(irq, op->node)) {
-		if (kernel_thread(powerd, NULL, CLONE_FS) < 0) {
-			printk("Failed to start power daemon.\n");
+		if (start_powerd() < 0)
 			return 0;
-		}
-		printk("powerd running.\n");

 		if (request_irq(irq,
 				power_handler, 0, "power", NULL) < 0)
-			printk("power: Error, cannot register IRQ handler.\n");
+			printk(KERN_ERR "power: Cannot setup IRQ handler.\n");
 	} else {
-		printk("not using powerd.\n");
+		printk(KERN_INFO "power: Not using powerd.\n");
 	}

 	return 0;
@@ -164,4 +173,3 @@ void __init power_init(void)
 	of_register_driver(&power_driver, &of_bus_type);
 	return;
 }
-#endif /* CONFIG_PCI */
@@ -29,6 +29,7 @@
 #include <linux/compat.h>
 #include <linux/tick.h>
 #include <linux/init.h>
+#include <linux/cpu.h>

 #include <asm/oplib.h>
 #include <asm/uaccess.h>
@@ -49,7 +50,7 @@

 /* #define VERBOSE_SHOWREGS */

-static void sparc64_yield(void)
+static void sparc64_yield(int cpu)
 {
 	if (tlb_type != hypervisor)
 		return;
@@ -57,7 +58,7 @@ static void sparc64_yield(void)
 	clear_thread_flag(TIF_POLLING_NRFLAG);
 	smp_mb__after_clear_bit();

-	while (!need_resched()) {
+	while (!need_resched() && !cpu_is_offline(cpu)) {
 		unsigned long pstate;

 		/* Disable interrupts. */
@@ -68,7 +69,7 @@ static void sparc64_yield(void)
 			: "=&r" (pstate)
 			: "i" (PSTATE_IE));

-		if (!need_resched())
+		if (!need_resched() && !cpu_is_offline(cpu))
 			sun4v_cpu_yield();

 		/* Re-enable interrupts. */
@@ -86,15 +87,25 @@ static void sparc64_yield(void)
 /* The idle loop on sparc64. */
 void cpu_idle(void)
 {
+	int cpu = smp_processor_id();
+
 	set_thread_flag(TIF_POLLING_NRFLAG);

 	while(1) {
 		tick_nohz_stop_sched_tick();
-		while (!need_resched())
-			sparc64_yield();
+
+		while (!need_resched() && !cpu_is_offline(cpu))
+			sparc64_yield(cpu);
+
 		tick_nohz_restart_sched_tick();

 		preempt_enable_no_resched();
+
+#ifdef CONFIG_HOTPLUG_CPU
+		if (cpu_is_offline(cpu))
+			cpu_play_dead();
+#endif
+
 		schedule();
 		preempt_disable();
 	}
@@ -1808,7 +1808,7 @@ static void __init of_fill_in_cpu_data(void)

 #ifdef CONFIG_SMP
 		cpu_set(cpuid, cpu_present_map);
-		cpu_set(cpuid, phys_cpu_present_map);
+		cpu_set(cpuid, cpu_possible_map);
 #endif
 	}

@@ -442,7 +442,6 @@ static int show_cpuinfo(struct seq_file *m, void *__unused)
 		   "D$ parity tl1\t: %u\n"
 		   "I$ parity tl1\t: %u\n"
 #ifndef CONFIG_SMP
-		   "Cpu0Bogo\t: %lu.%02lu\n"
 		   "Cpu0ClkTck\t: %016lx\n"
 #endif
 		   ,
@@ -455,10 +454,8 @@ static int show_cpuinfo(struct seq_file *m, void *__unused)
 		   ncpus_probed,
 		   num_online_cpus(),
 		   dcache_parity_tl1_occurred,
-		   icache_parity_tl1_occurred
+		   icache_parity_tl1_occurred,
 #ifndef CONFIG_SMP
-		   , cpu_data(0).udelay_val/(500000/HZ),
-		   (cpu_data(0).udelay_val/(5000/HZ)) % 100,
 		   cpu_data(0).clock_tick
 #endif
 		);
@@ -1,6 +1,6 @@
 /* smp.c: Sparc64 SMP support.
 *
- * Copyright (C) 1997 David S. Miller (davem@caip.rutgers.edu)
+ * Copyright (C) 1997, 2007 David S. Miller (davem@davemloft.net)
 */

 #include <linux/module.h>
@@ -28,6 +28,8 @@
 #include <asm/tlbflush.h>
 #include <asm/mmu_context.h>
 #include <asm/cpudata.h>
+#include <asm/hvtramp.h>
+#include <asm/io.h>

 #include <asm/irq.h>
 #include <asm/irq_regs.h>
@@ -41,22 +43,26 @@
 #include <asm/sections.h>
 #include <asm/prom.h>
 #include <asm/mdesc.h>
+#include <asm/ldc.h>
+#include <asm/hypervisor.h>

 extern void calibrate_delay(void);

 int sparc64_multi_core __read_mostly;

-/* Please don't make this stuff initdata!!!  --DaveM */
-unsigned char boot_cpu_id;
-
+cpumask_t cpu_possible_map __read_mostly = CPU_MASK_NONE;
 cpumask_t cpu_online_map __read_mostly = CPU_MASK_NONE;
-cpumask_t phys_cpu_present_map __read_mostly = CPU_MASK_NONE;
 cpumask_t cpu_sibling_map[NR_CPUS] __read_mostly =
 	{ [0 ... NR_CPUS-1] = CPU_MASK_NONE };
 cpumask_t cpu_core_map[NR_CPUS] __read_mostly =
 	{ [0 ... NR_CPUS-1] = CPU_MASK_NONE };
+
+EXPORT_SYMBOL(cpu_possible_map);
+EXPORT_SYMBOL(cpu_online_map);
+EXPORT_SYMBOL(cpu_sibling_map);
+EXPORT_SYMBOL(cpu_core_map);
+
 static cpumask_t smp_commenced_mask;
-static cpumask_t cpu_callout_map;

 void smp_info(struct seq_file *m)
 {
@@ -73,18 +79,17 @@ void smp_bogo(struct seq_file *m)
 	
 	for_each_online_cpu(i)
 		seq_printf(m,
-			   "Cpu%dBogo\t: %lu.%02lu\n"
 			   "Cpu%dClkTck\t: %016lx\n",
-			   i, cpu_data(i).udelay_val / (500000/HZ),
-			   (cpu_data(i).udelay_val / (5000/HZ)) % 100,
 			   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;

-void __init smp_callin(void)
+void __devinit smp_callin(void)
 {
 	int cpuid = hard_smp_processor_id();

@@ -102,8 +107,6 @@ void __init smp_callin(void)

 	local_irq_enable();

-	calibrate_delay();
-	cpu_data(cpuid).udelay_val = loops_per_jiffy;
 	callin_flag = 1;
 	__asm__ __volatile__("membar #Sync\n\t"
 			     "flush  %%g6" : : : "memory");
@@ -120,7 +123,9 @@ void __init smp_callin(void)
 	while (!cpu_isset(cpuid, smp_commenced_mask))
 		rmb();

+	spin_lock(&call_lock);
 	cpu_set(cpuid, cpu_online_map);
+	spin_unlock(&call_lock);

 	/* idle thread is expected to have preempt disabled */
 	preempt_disable();
@@ -268,6 +273,67 @@ static void smp_synchronize_one_tick(int cpu)
 	spin_unlock_irqrestore(&itc_sync_lock, flags);
 }

+#if defined(CONFIG_SUN_LDOMS) && defined(CONFIG_HOTPLUG_CPU)
+/* XXX Put this in some common place. XXX */
+static unsigned long kimage_addr_to_ra(void *p)
+{
+	unsigned long val = (unsigned long) p;
+
+	return kern_base + (val - KERNBASE);
+}
+
+static void ldom_startcpu_cpuid(unsigned int cpu, unsigned long thread_reg)
+{
+	extern unsigned long sparc64_ttable_tl0;
+	extern unsigned long kern_locked_tte_data;
+	extern int bigkernel;
+	struct hvtramp_descr *hdesc;
+	unsigned long trampoline_ra;
+	struct trap_per_cpu *tb;
+	u64 tte_vaddr, tte_data;
+	unsigned long hv_err;
+
+	hdesc = kzalloc(sizeof(*hdesc), GFP_KERNEL);
+	if (!hdesc) {
+		printk(KERN_ERR "ldom_startcpu_cpuid: Cannot allocate "
+		       "hvtramp_descr.\n");
+		return;
+	}
+
+	hdesc->cpu = cpu;
+	hdesc->num_mappings = (bigkernel ? 2 : 1);
+
+	tb = &trap_block[cpu];
+	tb->hdesc = hdesc;
+
+	hdesc->fault_info_va = (unsigned long) &tb->fault_info;
+	hdesc->fault_info_pa = kimage_addr_to_ra(&tb->fault_info);
+
+	hdesc->thread_reg = thread_reg;
+
+	tte_vaddr = (unsigned long) KERNBASE;
+	tte_data = kern_locked_tte_data;
+
+	hdesc->maps[0].vaddr = tte_vaddr;
+	hdesc->maps[0].tte   = tte_data;
+	if (bigkernel) {
+		tte_vaddr += 0x400000;
+		tte_data  += 0x400000;
+		hdesc->maps[1].vaddr = tte_vaddr;
+		hdesc->maps[1].tte   = tte_data;
+	}
+
+	trampoline_ra = kimage_addr_to_ra(hv_cpu_startup);
+
+	hv_err = sun4v_cpu_start(cpu, trampoline_ra,
+				 kimage_addr_to_ra(&sparc64_ttable_tl0),
+				 __pa(hdesc));
+	if (hv_err)
+		printk(KERN_ERR "ldom_startcpu_cpuid: sun4v_cpu_start() "
+		       "gives error %lu\n", hv_err);
+}
+#endif
+
 extern void sun4v_init_mondo_queues(int use_bootmem, int cpu, int alloc, int load);

 extern unsigned long sparc64_cpu_startup;
@@ -280,6 +346,7 @@ static struct thread_info *cpu_new_thread = NULL;

 static int __devinit smp_boot_one_cpu(unsigned int cpu)
 {
+	struct trap_per_cpu *tb = &trap_block[cpu];
 	unsigned long entry =
 		(unsigned long)(&sparc64_cpu_startup);
 	unsigned long cookie =
@@ -290,20 +357,25 @@ static int __devinit smp_boot_one_cpu(unsigned int cpu)
 	p = fork_idle(cpu);
 	callin_flag = 0;
 	cpu_new_thread = task_thread_info(p);
-	cpu_set(cpu, cpu_callout_map);

 	if (tlb_type == hypervisor) {
 		/* Alloc the mondo queues, cpu will load them.  */
 		sun4v_init_mondo_queues(0, cpu, 1, 0);

-		prom_startcpu_cpuid(cpu, entry, cookie);
+#if defined(CONFIG_SUN_LDOMS) && defined(CONFIG_HOTPLUG_CPU)
+		if (ldom_domaining_enabled)
+			ldom_startcpu_cpuid(cpu,
+					    (unsigned long) cpu_new_thread);
+		else
+#endif
+			prom_startcpu_cpuid(cpu, entry, cookie);
 	} else {
 		struct device_node *dp = of_find_node_by_cpuid(cpu);

 		prom_startcpu(dp->node, entry, cookie);
 	}

-	for (timeout = 0; timeout < 5000000; timeout++) {
+	for (timeout = 0; timeout < 50000; timeout++) {
 		if (callin_flag)
 			break;
 		udelay(100);
@@ -313,11 +385,15 @@ static int __devinit smp_boot_one_cpu(unsigned int cpu)
 		ret = 0;
 	} else {
 		printk("Processor %d is stuck.\n", cpu);
-		cpu_clear(cpu, cpu_callout_map);
 		ret = -ENODEV;
 	}
 	cpu_new_thread = NULL;

+	if (tb->hdesc) {
+		kfree(tb->hdesc);
+		tb->hdesc = NULL;
+	}
+
 	return ret;
 }

@@ -720,7 +796,6 @@ struct call_data_struct {
 	int wait;
 };

-static __cacheline_aligned_in_smp DEFINE_SPINLOCK(call_lock);
 static struct call_data_struct *call_data;

 extern unsigned long xcall_call_function;
@@ -1152,34 +1227,14 @@ void smp_penguin_jailcell(int irq, struct pt_regs *regs)
 	preempt_enable();
 }

-void __init smp_tick_init(void)
-{
-	boot_cpu_id = hard_smp_processor_id();
-}
-
 /* /proc/profile writes can call this, don't __init it please. */
 int setup_profiling_timer(unsigned int multiplier)
 {
 	return -EINVAL;
 }

-/* Constrain the number of cpus to max_cpus.  */
 void __init smp_prepare_cpus(unsigned int max_cpus)
 {
-	int i;
-
-	if (num_possible_cpus() > max_cpus) {
-		for_each_possible_cpu(i) {
-			if (i != boot_cpu_id) {
-				cpu_clear(i, phys_cpu_present_map);
-				cpu_clear(i, cpu_present_map);
-				if (num_possible_cpus() <= max_cpus)
-					break;
-			}
-		}
-	}
-
-	cpu_data(boot_cpu_id).udelay_val = loops_per_jiffy;
 }

 void __devinit smp_prepare_boot_cpu(void)
@@ -1190,30 +1245,32 @@ void __devinit smp_fill_in_sib_core_maps(void)
 {
 	unsigned int i;

-	for_each_possible_cpu(i) {
+	for_each_present_cpu(i) {
 		unsigned int j;

+		cpus_clear(cpu_core_map[i]);
 		if (cpu_data(i).core_id == 0) {
 			cpu_set(i, cpu_core_map[i]);
 			continue;
 		}

-		for_each_possible_cpu(j) {
+		for_each_present_cpu(j) {
 			if (cpu_data(i).core_id ==
 			    cpu_data(j).core_id)
 				cpu_set(j, cpu_core_map[i]);
 		}
 	}

-	for_each_possible_cpu(i) {
+	for_each_present_cpu(i) {
 		unsigned int j;

+		cpus_clear(cpu_sibling_map[i]);
 		if (cpu_data(i).proc_id == -1) {
 			cpu_set(i, cpu_sibling_map[i]);
 			continue;
 		}

-		for_each_possible_cpu(j) {
+		for_each_present_cpu(j) {
 			if (cpu_data(i).proc_id ==
 			    cpu_data(j).proc_id)
 				cpu_set(j, cpu_sibling_map[i]);
@@ -1242,18 +1299,112 @@ int __cpuinit __cpu_up(unsigned int cpu)
 	return ret;
 }

-void __init smp_cpus_done(unsigned int max_cpus)
+#ifdef CONFIG_HOTPLUG_CPU
+void cpu_play_dead(void)
 {
-	unsigned long bogosum = 0;
+	int cpu = smp_processor_id();
+	unsigned long pstate;
+
+	idle_task_exit();
+
+	if (tlb_type == hypervisor) {
+		struct trap_per_cpu *tb = &trap_block[cpu];
+
+		sun4v_cpu_qconf(HV_CPU_QUEUE_CPU_MONDO,
+				tb->cpu_mondo_pa, 0);
+		sun4v_cpu_qconf(HV_CPU_QUEUE_DEVICE_MONDO,
+				tb->dev_mondo_pa, 0);
+		sun4v_cpu_qconf(HV_CPU_QUEUE_RES_ERROR,
+				tb->resum_mondo_pa, 0);
+		sun4v_cpu_qconf(HV_CPU_QUEUE_NONRES_ERROR,
+				tb->nonresum_mondo_pa, 0);
+	}
+
+	cpu_clear(cpu, smp_commenced_mask);
+	membar_safe("#Sync");
+
+	local_irq_disable();
+
+	__asm__ __volatile__(
+		"rdpr	%%pstate, %0\n\t"
+		"wrpr	%0, %1, %%pstate"
+		: "=r" (pstate)
+		: "i" (PSTATE_IE));
+
+	while (1)
+		barrier();
+}
+
+int __cpu_disable(void)
+{
+	int cpu = smp_processor_id();
+	cpuinfo_sparc *c;
 	int i;

-	for_each_online_cpu(i)
-		bogosum += cpu_data(i).udelay_val;
-	printk("Total of %ld processors activated "
-	       "(%lu.%02lu BogoMIPS).\n",
-	       (long) num_online_cpus(),
-	       bogosum/(500000/HZ),
-	       (bogosum/(5000/HZ))%100);
+	for_each_cpu_mask(i, cpu_core_map[cpu])
+		cpu_clear(cpu, cpu_core_map[i]);
+	cpus_clear(cpu_core_map[cpu]);
+
+	for_each_cpu_mask(i, cpu_sibling_map[cpu])
+		cpu_clear(cpu, cpu_sibling_map[i]);
+	cpus_clear(cpu_sibling_map[cpu]);
+
+	c = &cpu_data(cpu);
+
+	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.  */
+	fixup_irqs();
+
+	local_irq_enable();
+	mdelay(1);
+	local_irq_disable();
+
+	return 0;
+}
+
+void __cpu_die(unsigned int cpu)
+{
+	int i;
+
+	for (i = 0; i < 100; i++) {
+		smp_rmb();
+		if (!cpu_isset(cpu, smp_commenced_mask))
+			break;
+		msleep(100);
+	}
+	if (cpu_isset(cpu, smp_commenced_mask)) {
+		printk(KERN_ERR "CPU %u didn't die...\n", cpu);
+	} else {
+#if defined(CONFIG_SUN_LDOMS)
+		unsigned long hv_err;
+		int limit = 100;
+
+		do {
+			hv_err = sun4v_cpu_stop(cpu);
+			if (hv_err == HV_EOK) {
+				cpu_clear(cpu, cpu_present_map);
+				break;
+			}
+		} while (--limit > 0);
+		if (limit <= 0) {
+			printk(KERN_ERR "sun4v_cpu_stop() fails err=%lu\n",
+			       hv_err);
+		}
+#endif
+	}
+}
+#endif
+
+void __init smp_cpus_done(unsigned int max_cpus)
+{
 }

 void smp_send_reschedule(int cpu)
@@ -1,7 +1,6 @@
-/* $Id: sparc64_ksyms.c,v 1.121 2002/02/09 19:49:31 davem Exp $
- * arch/sparc64/kernel/sparc64_ksyms.c: Sparc64 specific ksyms support.
+/* arch/sparc64/kernel/sparc64_ksyms.c: Sparc64 specific ksyms support.
 *
- * Copyright (C) 1996 David S. Miller (davem@caip.rutgers.edu)
+ * Copyright (C) 1996, 2007 David S. Miller (davem@davemloft.net)
 * Copyright (C) 1996 Eddie C. Dost (ecd@skynet.be)
 * Copyright (C) 1999 Jakub Jelinek (jj@ultra.linux.cz)
 */
@@ -28,7 +27,6 @@
 #include <net/compat.h>

 #include <asm/oplib.h>
-#include <asm/delay.h>
 #include <asm/system.h>
 #include <asm/auxio.h>
 #include <asm/pgtable.h>
@@ -124,10 +122,6 @@ EXPORT_SYMBOL(__write_lock);
 EXPORT_SYMBOL(__write_unlock);
 EXPORT_SYMBOL(__write_trylock);

-/* CPU online map and active count.  */
-EXPORT_SYMBOL(cpu_online_map);
-EXPORT_SYMBOL(phys_cpu_present_map);
-
 EXPORT_SYMBOL(smp_call_function);
 #endif /* CONFIG_SMP */

@@ -330,12 +324,6 @@ EXPORT_SYMBOL(memset);
 EXPORT_SYMBOL(memmove);
 EXPORT_SYMBOL(strncmp);

-/* Delay routines. */
-EXPORT_SYMBOL(__udelay);
-EXPORT_SYMBOL(__ndelay);
-EXPORT_SYMBOL(__const_udelay);
-EXPORT_SYMBOL(__delay);
-
 void VISenter(void);
 /* RAID code needs this */
 EXPORT_SYMBOL(VISenter);
@@ -193,7 +193,6 @@ static ssize_t show_##NAME(struct sys_device *dev, char *buf) \
 }

 SHOW_CPUDATA_ULONG_NAME(clock_tick, clock_tick);
-SHOW_CPUDATA_ULONG_NAME(udelay_val, udelay_val);
 SHOW_CPUDATA_UINT_NAME(l1_dcache_size, dcache_size);
 SHOW_CPUDATA_UINT_NAME(l1_dcache_line_size, dcache_line_size);
 SHOW_CPUDATA_UINT_NAME(l1_icache_size, icache_size);
@@ -203,7 +202,6 @@ SHOW_CPUDATA_UINT_NAME(l2_cache_line_size, ecache_line_size);

 static struct sysdev_attribute cpu_core_attrs[] = {
 	_SYSDEV_ATTR(clock_tick,          0444, show_clock_tick, NULL),
-	_SYSDEV_ATTR(udelay_val,          0444, show_udelay_val, NULL),
 	_SYSDEV_ATTR(l1_dcache_size,      0444, show_l1_dcache_size, NULL),
 	_SYSDEV_ATTR(l1_dcache_line_size, 0444, show_l1_dcache_line_size, NULL),
 	_SYSDEV_ATTR(l1_icache_size,      0444, show_l1_icache_size, NULL),
@@ -849,9 +849,6 @@ static unsigned long sparc64_init_timers(void)
 {
 	struct device_node *dp;
 	unsigned long clock;
-#ifdef CONFIG_SMP
-	extern void smp_tick_init(void);
-#endif

 	dp = of_find_node_by_path("/");
 	if (tlb_type == spitfire) {
@@ -874,10 +871,6 @@ static unsigned long sparc64_init_timers(void)
 		clock = of_getintprop_default(dp, "stick-frequency", 0);
 	}

-#ifdef CONFIG_SMP
-	smp_tick_init();
-#endif
-
 	return clock;
 }

@@ -1038,10 +1031,31 @@ static void __init setup_clockevent_multiplier(unsigned long hz)
 	sparc64_clockevent.mult = mult;
 }

+static unsigned long tb_ticks_per_usec __read_mostly;
+
+void __delay(unsigned long loops)
+{
+	unsigned long bclock, now;
+
+	bclock = tick_ops->get_tick();
+	do {
+		now = tick_ops->get_tick();
+	} while ((now-bclock) < loops);
+}
+EXPORT_SYMBOL(__delay);
+
+void udelay(unsigned long usecs)
+{
+	__delay(tb_ticks_per_usec * usecs);
+}
+EXPORT_SYMBOL(udelay);
+
 void __init time_init(void)
 {
 	unsigned long clock = sparc64_init_timers();

+	tb_ticks_per_usec = clock / USEC_PER_SEC;
+
 	timer_ticks_per_nsec_quotient =
 		clocksource_hz2mult(clock, SPARC64_NSEC_PER_CYC_SHIFT);

@@ -0,0 +1,395 @@
+/* vio.c: Virtual I/O channel devices probing infrastructure.
+ *
+ *    Copyright (c) 2003-2005 IBM Corp.
+ *     Dave Engebretsen engebret@us.ibm.com
+ *     Santiago Leon santil@us.ibm.com
+ *     Hollis Blanchard <hollisb@us.ibm.com>
+ *     Stephen Rothwell
+ *
+ * Adapted to sparc64 by David S. Miller davem@davemloft.net
+ */
+
+#include <linux/kernel.h>
+#include <linux/irq.h>
+#include <linux/init.h>
+
+#include <asm/mdesc.h>
+#include <asm/vio.h>
+
+static inline int find_in_proplist(const char *list, const char *match,
+				   int len)
+{
+	while (len > 0) {
+		int l;
+
+		if (!strcmp(list, match))
+			return 1;
+		l = strlen(list) + 1;
+		list += l;
+		len -= l;
+	}
+	return 0;
+}
+
+static const struct vio_device_id *vio_match_device(
+	const struct vio_device_id *matches,
+	const struct vio_dev *dev)
+{
+	const char *type, *compat;
+	int len;
+
+	type = dev->type;
+	compat = dev->compat;
+	len = dev->compat_len;
+
+	while (matches->type[0] || matches->compat[0]) {
+		int match = 1;
+		if (matches->type[0])
+			match &= !strcmp(matches->type, type);
+
+		if (matches->compat[0]) {
+			match &= len &&
+				find_in_proplist(compat, matches->compat, len);
+		}
+		if (match)
+			return matches;
+		matches++;
+	}
+	return NULL;
+}
+
+static int vio_bus_match(struct device *dev, struct device_driver *drv)
+{
+	struct vio_dev *vio_dev = to_vio_dev(dev);
+	struct vio_driver *vio_drv = to_vio_driver(drv);
+	const struct vio_device_id *matches = vio_drv->id_table;
+
+	if (!matches)
+		return 0;
+
+	return vio_match_device(matches, vio_dev) != NULL;
+}
+
+static int vio_device_probe(struct device *dev)
+{
+	struct vio_dev *vdev = to_vio_dev(dev);
+	struct vio_driver *drv = to_vio_driver(dev->driver);
+	const struct vio_device_id *id;
+	int error = -ENODEV;
+
+	if (drv->probe) {
+		id = vio_match_device(drv->id_table, vdev);
+		if (id)
+			error = drv->probe(vdev, id);
+	}
+
+	return error;
+}
+
+static int vio_device_remove(struct device *dev)
+{
+	struct vio_dev *vdev = to_vio_dev(dev);
+	struct vio_driver *drv = to_vio_driver(dev->driver);
+
+	if (drv->remove)
+		return drv->remove(vdev);
+
+	return 1;
+}
+
+static ssize_t devspec_show(struct device *dev,
+		struct device_attribute *attr, char *buf)
+{
+	struct vio_dev *vdev = to_vio_dev(dev);
+	const char *str = "none";
+
+	if (!strcmp(vdev->type, "network"))
+		str = "vnet";
+	else if (!strcmp(vdev->type, "block"))
+		str = "vdisk";
+
+	return sprintf(buf, "%s\n", str);
+}
+
+static ssize_t type_show(struct device *dev,
+		struct device_attribute *attr, char *buf)
+{
+	struct vio_dev *vdev = to_vio_dev(dev);
+	return sprintf(buf, "%s\n", vdev->type);
+}
+
+static struct device_attribute vio_dev_attrs[] = {
+	__ATTR_RO(devspec),
+	__ATTR_RO(type),
+	__ATTR_NULL
+};
+
+static struct bus_type vio_bus_type = {
+	.name		= "vio",
+	.dev_attrs	= vio_dev_attrs,
+	.match		= vio_bus_match,
+	.probe		= vio_device_probe,
+	.remove		= vio_device_remove,
+};
+
+int vio_register_driver(struct vio_driver *viodrv)
+{
+	viodrv->driver.bus = &vio_bus_type;
+
+	return driver_register(&viodrv->driver);
+}
+EXPORT_SYMBOL(vio_register_driver);
+
+void vio_unregister_driver(struct vio_driver *viodrv)
+{
+	driver_unregister(&viodrv->driver);
+}
+EXPORT_SYMBOL(vio_unregister_driver);
+
+static void __devinit vio_dev_release(struct device *dev)
+{
+	kfree(to_vio_dev(dev));
+}
+
+static ssize_t
+show_pciobppath_attr(struct device *dev, struct device_attribute *attr,
+		     char *buf)
+{
+	struct vio_dev *vdev;
+	struct device_node *dp;
+
+	vdev = to_vio_dev(dev);
+	dp = vdev->dp;
+
+	return snprintf (buf, PAGE_SIZE, "%s\n", dp->full_name);
+}
+
+static DEVICE_ATTR(obppath, S_IRUSR | S_IRGRP | S_IROTH,
+		   show_pciobppath_attr, NULL);
+
+struct device_node *cdev_node;
+
+static struct vio_dev *root_vdev;
+static u64 cdev_cfg_handle;
+
+static void vio_fill_channel_info(struct mdesc_handle *hp, u64 mp,
+				  struct vio_dev *vdev)
+{
+	u64 a;
+
+	mdesc_for_each_arc(a, hp, mp, MDESC_ARC_TYPE_FWD) {
+		const u64 *chan_id;
+		const u64 *irq;
+		u64 target;
+
+		target = mdesc_arc_target(hp, a);
+
+		irq = mdesc_get_property(hp, target, "tx-ino", NULL);
+		if (irq)
+			vdev->tx_irq = sun4v_build_virq(cdev_cfg_handle, *irq);
+
+		irq = mdesc_get_property(hp, target, "rx-ino", NULL);
+		if (irq)
+			vdev->rx_irq = sun4v_build_virq(cdev_cfg_handle, *irq);
+
+		chan_id = mdesc_get_property(hp, target, "id", NULL);
+		if (chan_id)
+			vdev->channel_id = *chan_id;
+	}
+}
+
+static struct vio_dev *vio_create_one(struct mdesc_handle *hp, u64 mp,
+				      struct device *parent)
+{
+	const char *type, *compat;
+	struct device_node *dp;
+	struct vio_dev *vdev;
+	int err, tlen, clen;
+
+	type = mdesc_get_property(hp, mp, "device-type", &tlen);
+	if (!type) {
+		type = mdesc_get_property(hp, mp, "name", &tlen);
+		if (!type) {
+			type = mdesc_node_name(hp, mp);
+			tlen = strlen(type) + 1;
+		}
+	}
+	if (tlen > VIO_MAX_TYPE_LEN) {
+		printk(KERN_ERR "VIO: Type string [%s] is too long.\n",
+		       type);
+		return NULL;
+	}
+
+	compat = mdesc_get_property(hp, mp, "device-type", &clen);
+	if (!compat) {
+		clen = 0;
+	} else if (clen > VIO_MAX_COMPAT_LEN) {
+		printk(KERN_ERR "VIO: Compat len %d for [%s] is too long.\n",
+		       clen, type);
+		return NULL;
+	}
+
+	vdev = kzalloc(sizeof(*vdev), GFP_KERNEL);
+	if (!vdev) {
+		printk(KERN_ERR "VIO: Could not allocate vio_dev\n");
+		return NULL;
+	}
+
+	vdev->mp = mp;
+	memcpy(vdev->type, type, tlen);
+	if (compat)
+		memcpy(vdev->compat, compat, clen);
+	else
+		memset(vdev->compat, 0, sizeof(vdev->compat));
+	vdev->compat_len = clen;
+
+	vdev->channel_id = ~0UL;
+	vdev->tx_irq = ~0;
+	vdev->rx_irq = ~0;
+
+	vio_fill_channel_info(hp, mp, vdev);
+
+	snprintf(vdev->dev.bus_id, BUS_ID_SIZE, "%lx", mp);
+	vdev->dev.parent = parent;
+	vdev->dev.bus = &vio_bus_type;
+	vdev->dev.release = vio_dev_release;
+
+	if (parent == NULL) {
+		dp = cdev_node;
+	} else if (to_vio_dev(parent) == root_vdev) {
+		dp = of_get_next_child(cdev_node, NULL);
+		while (dp) {
+			if (!strcmp(dp->type, type))
+				break;
+
+			dp = of_get_next_child(cdev_node, dp);
+		}
+	} else {
+		dp = to_vio_dev(parent)->dp;
+	}
+	vdev->dp = dp;
+
+	err = device_register(&vdev->dev);
+	if (err) {
+		printk(KERN_ERR "VIO: Could not register device %s, err=%d\n",
+		       vdev->dev.bus_id, err);
+		kfree(vdev);
+		return NULL;
+	}
+	if (vdev->dp)
+		err = sysfs_create_file(&vdev->dev.kobj,
+					&dev_attr_obppath.attr);
+
+	return vdev;
+}
+
+static void walk_tree(struct mdesc_handle *hp, u64 n, struct vio_dev *parent)
+{
+	u64 a;
+
+	mdesc_for_each_arc(a, hp, n, MDESC_ARC_TYPE_FWD) {
+		struct vio_dev *vdev;
+		u64 target;
+
+		target = mdesc_arc_target(hp, a);
+		vdev = vio_create_one(hp, target, &parent->dev);
+		if (vdev)
+			walk_tree(hp, target, vdev);
+	}
+}
+
+static void create_devices(struct mdesc_handle *hp, u64 root)
+{
+	u64 mp;
+
+	root_vdev = vio_create_one(hp, root, NULL);
+	if (!root_vdev) {
+		printk(KERN_ERR "VIO: Coult not create root device.\n");
+		return;
+	}
+
+	walk_tree(hp, root, root_vdev);
+
+	/* Domain services is odd as it doesn't sit underneath the
+	 * channel-devices node, so we plug it in manually.
+	 */
+	mp = mdesc_node_by_name(hp, MDESC_NODE_NULL, "domain-services");
+	if (mp != MDESC_NODE_NULL) {
+		struct vio_dev *parent = vio_create_one(hp, mp,
+							&root_vdev->dev);
+
+		if (parent)
+			walk_tree(hp, mp, parent);
+	}
+}
+
+const char *channel_devices_node = "channel-devices";
+const char *channel_devices_compat = "SUNW,sun4v-channel-devices";
+const char *cfg_handle_prop = "cfg-handle";
+
+static int __init vio_init(void)
+{
+	struct mdesc_handle *hp;
+	const char *compat;
+	const u64 *cfg_handle;
+	int err, len;
+	u64 root;
+
+	err = bus_register(&vio_bus_type);
+	if (err) {
+		printk(KERN_ERR "VIO: Could not register bus type err=%d\n",
+		       err);
+		return err;
+	}
+
+	hp = mdesc_grab();
+	if (!hp)
+		return 0;
+
+	root = mdesc_node_by_name(hp, MDESC_NODE_NULL, channel_devices_node);
+	if (root == MDESC_NODE_NULL) {
+		printk(KERN_INFO "VIO: No channel-devices MDESC node.\n");
+		mdesc_release(hp);
+		return 0;
+	}
+
+	cdev_node = of_find_node_by_name(NULL, "channel-devices");
+	err = -ENODEV;
+	if (!cdev_node) {
+		printk(KERN_INFO "VIO: No channel-devices OBP node.\n");
+		goto out_release;
+	}
+
+	compat = mdesc_get_property(hp, root, "compatible", &len);
+	if (!compat) {
+		printk(KERN_ERR "VIO: Channel devices lacks compatible "
+		       "property\n");
+		goto out_release;
+	}
+	if (!find_in_proplist(compat, channel_devices_compat, len)) {
+		printk(KERN_ERR "VIO: Channel devices node lacks (%s) "
+		       "compat entry.\n", channel_devices_compat);
+		goto out_release;
+	}
+
+	cfg_handle = mdesc_get_property(hp, root, cfg_handle_prop, NULL);
+	if (!cfg_handle) {
+		printk(KERN_ERR "VIO: Channel devices lacks %s property\n",
+		       cfg_handle_prop);
+		goto out_release;
+	}
+
+	cdev_cfg_handle = *cfg_handle;
+
+	create_devices(hp, root);
+
+	mdesc_release(hp);
+
+	return 0;
+
+out_release:
+	mdesc_release(hp);
+	return err;
+}
+
+postcore_initcall(vio_init);
@@ -14,6 +14,6 @@ lib-y := PeeCeeI.o copy_page.o clear_page.o strlen.o strncmp.o \
 	 NGmemcpy.o NGcopy_from_user.o NGcopy_to_user.o NGpatch.o \
 	 NGpage.o NGbzero.o \
 	 copy_in_user.o user_fixup.o memmove.o \
-	 mcount.o ipcsum.o rwsem.o xor.o delay.o
+	 mcount.o ipcsum.o rwsem.o xor.o

 obj-y += iomap.o
@@ -1,46 +0,0 @@
-/* delay.c: Delay loops for sparc64
- *
- * Copyright (C) 2004, 2006 David S. Miller <davem@davemloft.net>
- *
- * Based heavily upon x86 variant which is:
- *	Copyright (C) 1993 Linus Torvalds
- *	Copyright (C) 1997 Martin Mares <mj@atrey.karlin.mff.cuni.cz>
- */
-
-#include <linux/delay.h>
-#include <asm/timer.h>
-
-void __delay(unsigned long loops)
-{
-	unsigned long bclock, now;
-	
-	bclock = tick_ops->get_tick();
-	do {
-		now = tick_ops->get_tick();
-	} while ((now-bclock) < loops);
-}
-
-/* We used to multiply by HZ after shifting down by 32 bits
- * but that runs into problems for higher values of HZ and
- * slow cpus.
- */
-void __const_udelay(unsigned long n)
-{
-	n *= 4;
-
-	n *= (cpu_data(raw_smp_processor_id()).udelay_val * (HZ/4));
-	n >>= 32;
-
-	__delay(n + 1);
-}
-
-void __udelay(unsigned long n)
-{
-	__const_udelay(n * 0x10c7UL);
-}
-
-
-void __ndelay(unsigned long n)
-{
-	__const_udelay(n * 0x5UL);
-}
@@ -14,6 +14,7 @@
 #include <asm/openprom.h>
 #include <asm/oplib.h>
 #include <asm/system.h>
+#include <asm/ldc.h>

 int prom_service_exists(const char *service_name)
 {
@@ -37,6 +38,10 @@ void prom_sun4v_guest_soft_state(void)
 /* Reset and reboot the machine with the command 'bcommand'. */
 void prom_reboot(const char *bcommand)
 {
+#ifdef CONFIG_SUN_LDOMS
+	if (ldom_domaining_enabled)
+		ldom_reboot(bcommand);
+#endif
 	p1275_cmd("boot", P1275_ARG(0, P1275_ARG_IN_STRING) |
 		  P1275_INOUT(1, 0), bcommand);
 }
@@ -91,6 +96,10 @@ void prom_cmdline(void)
 */
 void prom_halt(void)
 {
+#ifdef CONFIG_SUN_LDOMS
+	if (ldom_domaining_enabled)
+		ldom_power_off();
+#endif
 again:
 	p1275_cmd("exit", P1275_INOUT(0, 0));
 	goto again; /* PROM is out to get me -DaveM */
@@ -98,6 +107,10 @@ again:

 void prom_halt_power_off(void)
 {
+#ifdef CONFIG_SUN_LDOMS
+	if (ldom_domaining_enabled)
+		ldom_power_off();
+#endif
 	p1275_cmd("SUNW,power-off", P1275_INOUT(0, 0));

 	/* if nothing else helps, we just halt */
--- a/Show More
+++ b/Show More