i386: move xen

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>

17 files changed, 4198 insertions, 0 deletions

diff --git a/arch/x86/xen/Kconfig b/arch/x86/xen/Kconfig
new file mode 100644
index 00000000000..9df99e1885a
--- /dev/null
+++ b/arch/x86/xen/Kconfig
@@ -0,0 +1,11 @@
+#
+# This Kconfig describes xen options
+#
+
+config XEN
+	bool "Enable support for Xen hypervisor"
+	depends on PARAVIRT && X86_CMPXCHG && X86_TSC && !NEED_MULTIPLE_NODES
+	help
+	  This is the Linux Xen port.  Enabling this will allow the
+	  kernel to boot in a paravirtualized environment under the
+	  Xen hypervisor.
diff --git a/arch/x86/xen/Makefile b/arch/x86/xen/Makefile
new file mode 100644
index 00000000000..343df246bd3
--- /dev/null
+++ b/arch/x86/xen/Makefile
@@ -0,0 +1,4 @@
+obj-y		:= enlighten.o setup.o features.o multicalls.o mmu.o \
+			events.o time.o manage.o xen-asm.o
+
+obj-$(CONFIG_SMP)	+= smp.o
diff --git a/arch/x86/xen/enlighten.c b/arch/x86/xen/enlighten.c
new file mode 100644
index 00000000000..f01bfcd4bde
--- /dev/null
+++ b/arch/x86/xen/enlighten.c
@@ -0,0 +1,1146 @@
+/*
+ * Core of Xen paravirt_ops implementation.
+ *
+ * This file contains the xen_paravirt_ops structure itself, and the
+ * implementations for:
+ * - privileged instructions
+ * - interrupt flags
+ * - segment operations
+ * - booting and setup
+ *
+ * Jeremy Fitzhardinge <jeremy@xensource.com>, XenSource Inc, 2007
+ */
+
+#include <linux/kernel.h>
+#include <linux/init.h>
+#include <linux/smp.h>
+#include <linux/preempt.h>
+#include <linux/hardirq.h>
+#include <linux/percpu.h>
+#include <linux/delay.h>
+#include <linux/start_kernel.h>
+#include <linux/sched.h>
+#include <linux/bootmem.h>
+#include <linux/module.h>
+#include <linux/mm.h>
+#include <linux/page-flags.h>
+#include <linux/highmem.h>
+#include <linux/smp.h>
+
+#include <xen/interface/xen.h>
+#include <xen/interface/physdev.h>
+#include <xen/interface/vcpu.h>
+#include <xen/interface/sched.h>
+#include <xen/features.h>
+#include <xen/page.h>
+
+#include <asm/paravirt.h>
+#include <asm/page.h>
+#include <asm/xen/hypercall.h>
+#include <asm/xen/hypervisor.h>
+#include <asm/fixmap.h>
+#include <asm/processor.h>
+#include <asm/setup.h>
+#include <asm/desc.h>
+#include <asm/pgtable.h>
+#include <asm/tlbflush.h>
+#include <asm/reboot.h>
+
+#include "xen-ops.h"
+#include "mmu.h"
+#include "multicalls.h"
+
+EXPORT_SYMBOL_GPL(hypercall_page);
+
+DEFINE_PER_CPU(enum paravirt_lazy_mode, xen_lazy_mode);
+
+DEFINE_PER_CPU(struct vcpu_info *, xen_vcpu);
+DEFINE_PER_CPU(struct vcpu_info, xen_vcpu_info);
+DEFINE_PER_CPU(unsigned long, xen_cr3);
+
+struct start_info *xen_start_info;
+EXPORT_SYMBOL_GPL(xen_start_info);
+
+static /* __initdata */ struct shared_info dummy_shared_info;
+
+/*
+ * Point at some empty memory to start with. We map the real shared_info
+ * page as soon as fixmap is up and running.
+ */
+struct shared_info *HYPERVISOR_shared_info = (void *)&dummy_shared_info;
+
+/*
+ * Flag to determine whether vcpu info placement is available on all
+ * VCPUs.  We assume it is to start with, and then set it to zero on
+ * the first failure.  This is because it can succeed on some VCPUs
+ * and not others, since it can involve hypervisor memory allocation,
+ * or because the guest failed to guarantee all the appropriate
+ * constraints on all VCPUs (ie buffer can't cross a page boundary).
+ *
+ * Note that any particular CPU may be using a placed vcpu structure,
+ * but we can only optimise if the all are.
+ *
+ * 0: not available, 1: available
+ */
+static int have_vcpu_info_placement = 1;
+
+static void __init xen_vcpu_setup(int cpu)
+{
+	struct vcpu_register_vcpu_info info;
+	int err;
+	struct vcpu_info *vcpup;
+
+	per_cpu(xen_vcpu, cpu) = &HYPERVISOR_shared_info->vcpu_info[cpu];
+
+	if (!have_vcpu_info_placement)
+		return;		/* already tested, not available */
+
+	vcpup = &per_cpu(xen_vcpu_info, cpu);
+
+	info.mfn = virt_to_mfn(vcpup);
+	info.offset = offset_in_page(vcpup);
+
+	printk(KERN_DEBUG "trying to map vcpu_info %d at %p, mfn %x, offset %d\n",
+	       cpu, vcpup, info.mfn, info.offset);
+
+	/* Check to see if the hypervisor will put the vcpu_info
+	   structure where we want it, which allows direct access via
+	   a percpu-variable. */
+	err = HYPERVISOR_vcpu_op(VCPUOP_register_vcpu_info, cpu, &info);
+
+	if (err) {
+		printk(KERN_DEBUG "register_vcpu_info failed: err=%d\n", err);
+		have_vcpu_info_placement = 0;
+	} else {
+		/* This cpu is using the registered vcpu info, even if
+		   later ones fail to. */
+		per_cpu(xen_vcpu, cpu) = vcpup;
+
+		printk(KERN_DEBUG "cpu %d using vcpu_info at %p\n",
+		       cpu, vcpup);
+	}
+}
+
+static void __init xen_banner(void)
+{
+	printk(KERN_INFO "Booting paravirtualized kernel on %s\n",
+	       paravirt_ops.name);
+	printk(KERN_INFO "Hypervisor signature: %s\n", xen_start_info->magic);
+}
+
+static void xen_cpuid(unsigned int *eax, unsigned int *ebx,
+		      unsigned int *ecx, unsigned int *edx)
+{
+	unsigned maskedx = ~0;
+
+	/*
+	 * Mask out inconvenient features, to try and disable as many
+	 * unsupported kernel subsystems as possible.
+	 */
+	if (*eax == 1)
+		maskedx = ~((1 << X86_FEATURE_APIC) |  /* disable APIC */
+			    (1 << X86_FEATURE_ACPI) |  /* disable ACPI */
+			    (1 << X86_FEATURE_ACC));   /* thermal monitoring */
+
+	asm(XEN_EMULATE_PREFIX "cpuid"
+		: "=a" (*eax),
+		  "=b" (*ebx),
+		  "=c" (*ecx),
+		  "=d" (*edx)
+		: "0" (*eax), "2" (*ecx));
+	*edx &= maskedx;
+}
+
+static void xen_set_debugreg(int reg, unsigned long val)
+{
+	HYPERVISOR_set_debugreg(reg, val);
+}
+
+static unsigned long xen_get_debugreg(int reg)
+{
+	return HYPERVISOR_get_debugreg(reg);
+}
+
+static unsigned long xen_save_fl(void)
+{
+	struct vcpu_info *vcpu;
+	unsigned long flags;
+
+	vcpu = x86_read_percpu(xen_vcpu);
+
+	/* flag has opposite sense of mask */
+	flags = !vcpu->evtchn_upcall_mask;
+
+	/* convert to IF type flag
+	   -0 -> 0x00000000
+	   -1 -> 0xffffffff
+	*/
+	return (-flags) & X86_EFLAGS_IF;
+}
+
+static void xen_restore_fl(unsigned long flags)
+{
+	struct vcpu_info *vcpu;
+
+	/* convert from IF type flag */
+	flags = !(flags & X86_EFLAGS_IF);
+
+	/* There's a one instruction preempt window here.  We need to
+	   make sure we're don't switch CPUs between getting the vcpu
+	   pointer and updating the mask. */
+	preempt_disable();
+	vcpu = x86_read_percpu(xen_vcpu);
+	vcpu->evtchn_upcall_mask = flags;
+	preempt_enable_no_resched();
+
+	/* Doesn't matter if we get preempted here, because any
+	   pending event will get dealt with anyway. */
+
+	if (flags == 0) {
+		preempt_check_resched();
+		barrier(); /* unmask then check (avoid races) */
+		if (unlikely(vcpu->evtchn_upcall_pending))
+			force_evtchn_callback();
+	}
+}
+
+static void xen_irq_disable(void)
+{
+	/* There's a one instruction preempt window here.  We need to
+	   make sure we're don't switch CPUs between getting the vcpu
+	   pointer and updating the mask. */
+	preempt_disable();
+	x86_read_percpu(xen_vcpu)->evtchn_upcall_mask = 1;
+	preempt_enable_no_resched();
+}
+
+static void xen_irq_enable(void)
+{
+	struct vcpu_info *vcpu;
+
+	/* There's a one instruction preempt window here.  We need to
+	   make sure we're don't switch CPUs between getting the vcpu
+	   pointer and updating the mask. */
+	preempt_disable();
+	vcpu = x86_read_percpu(xen_vcpu);
+	vcpu->evtchn_upcall_mask = 0;
+	preempt_enable_no_resched();
+
+	/* Doesn't matter if we get preempted here, because any
+	   pending event will get dealt with anyway. */
+
+	barrier(); /* unmask then check (avoid races) */
+	if (unlikely(vcpu->evtchn_upcall_pending))
+		force_evtchn_callback();
+}
+
+static void xen_safe_halt(void)
+{
+	/* Blocking includes an implicit local_irq_enable(). */
+	if (HYPERVISOR_sched_op(SCHEDOP_block, 0) != 0)
+		BUG();
+}
+
+static void xen_halt(void)
+{
+	if (irqs_disabled())
+		HYPERVISOR_vcpu_op(VCPUOP_down, smp_processor_id(), NULL);
+	else
+		xen_safe_halt();
+}
+
+static void xen_set_lazy_mode(enum paravirt_lazy_mode mode)
+{
+	BUG_ON(preemptible());
+
+	switch (mode) {
+	case PARAVIRT_LAZY_NONE:
+		BUG_ON(x86_read_percpu(xen_lazy_mode) == PARAVIRT_LAZY_NONE);
+		break;
+
+	case PARAVIRT_LAZY_MMU:
+	case PARAVIRT_LAZY_CPU:
+		BUG_ON(x86_read_percpu(xen_lazy_mode) != PARAVIRT_LAZY_NONE);
+		break;
+
+	case PARAVIRT_LAZY_FLUSH:
+		/* flush if necessary, but don't change state */
+		if (x86_read_percpu(xen_lazy_mode) != PARAVIRT_LAZY_NONE)
+			xen_mc_flush();
+		return;
+	}
+
+	xen_mc_flush();
+	x86_write_percpu(xen_lazy_mode, mode);
+}
+
+static unsigned long xen_store_tr(void)
+{
+	return 0;
+}
+
+static void xen_set_ldt(const void *addr, unsigned entries)
+{
+	unsigned long linear_addr = (unsigned long)addr;
+	struct mmuext_op *op;
+	struct multicall_space mcs = xen_mc_entry(sizeof(*op));
+
+	op = mcs.args;
+	op->cmd = MMUEXT_SET_LDT;
+	if (linear_addr) {
+		/* ldt my be vmalloced, use arbitrary_virt_to_machine */
+		xmaddr_t maddr;
+		maddr = arbitrary_virt_to_machine((unsigned long)addr);
+		linear_addr = (unsigned long)maddr.maddr;
+	}
+	op->arg1.linear_addr = linear_addr;
+	op->arg2.nr_ents = entries;
+
+	MULTI_mmuext_op(mcs.mc, op, 1, NULL, DOMID_SELF);
+
+	xen_mc_issue(PARAVIRT_LAZY_CPU);
+}
+
+static void xen_load_gdt(const struct Xgt_desc_struct *dtr)
+{
+	unsigned long *frames;
+	unsigned long va = dtr->address;
+	unsigned int size = dtr->size + 1;
+	unsigned pages = (size + PAGE_SIZE - 1) / PAGE_SIZE;
+	int f;
+	struct multicall_space mcs;
+
+	/* A GDT can be up to 64k in size, which corresponds to 8192
+	   8-byte entries, or 16 4k pages.. */
+
+	BUG_ON(size > 65536);
+	BUG_ON(va & ~PAGE_MASK);
+
+	mcs = xen_mc_entry(sizeof(*frames) * pages);
+	frames = mcs.args;
+
+	for (f = 0; va < dtr->address + size; va += PAGE_SIZE, f++) {
+		frames[f] = virt_to_mfn(va);
+		make_lowmem_page_readonly((void *)va);
+	}
+
+	MULTI_set_gdt(mcs.mc, frames, size / sizeof(struct desc_struct));
+
+	xen_mc_issue(PARAVIRT_LAZY_CPU);
+}
+
+static void load_TLS_descriptor(struct thread_struct *t,
+				unsigned int cpu, unsigned int i)
+{
+	struct desc_struct *gdt = get_cpu_gdt_table(cpu);
+	xmaddr_t maddr = virt_to_machine(&gdt[GDT_ENTRY_TLS_MIN+i]);
+	struct multicall_space mc = __xen_mc_entry(0);
+
+	MULTI_update_descriptor(mc.mc, maddr.maddr, t->tls_array[i]);
+}
+
+static void xen_load_tls(struct thread_struct *t, unsigned int cpu)
+{
+	xen_mc_batch();
+
+	load_TLS_descriptor(t, cpu, 0);
+	load_TLS_descriptor(t, cpu, 1);
+	load_TLS_descriptor(t, cpu, 2);
+
+	xen_mc_issue(PARAVIRT_LAZY_CPU);
+
+	/*
+	 * XXX sleazy hack: If we're being called in a lazy-cpu zone,
+	 * it means we're in a context switch, and %gs has just been
+	 * saved.  This means we can zero it out to prevent faults on
+	 * exit from the hypervisor if the next process has no %gs.
+	 * Either way, it has been saved, and the new value will get
+	 * loaded properly.  This will go away as soon as Xen has been
+	 * modified to not save/restore %gs for normal hypercalls.
+	 */
+	if (xen_get_lazy_mode() == PARAVIRT_LAZY_CPU)
+		loadsegment(gs, 0);
+}
+
+static void xen_write_ldt_entry(struct desc_struct *dt, int entrynum,
+				u32 low, u32 high)
+{
+	unsigned long lp = (unsigned long)&dt[entrynum];
+	xmaddr_t mach_lp = virt_to_machine(lp);
+	u64 entry = (u64)high << 32 | low;
+
+	preempt_disable();
+
+	xen_mc_flush();
+	if (HYPERVISOR_update_descriptor(mach_lp.maddr, entry))
+		BUG();
+
+	preempt_enable();
+}
+
+static int cvt_gate_to_trap(int vector, u32 low, u32 high,
+			    struct trap_info *info)
+{
+	u8 type, dpl;
+
+	type = (high >> 8) & 0x1f;
+	dpl = (high >> 13) & 3;
+
+	if (type != 0xf && type != 0xe)
+		return 0;
+
+	info->vector = vector;
+	info->address = (high & 0xffff0000) | (low & 0x0000ffff);
+	info->cs = low >> 16;
+	info->flags = dpl;
+	/* interrupt gates clear IF */
+	if (type == 0xe)
+		info->flags |= 4;
+
+	return 1;
+}
+
+/* Locations of each CPU's IDT */
+static DEFINE_PER_CPU(struct Xgt_desc_struct, idt_desc);
+
+/* Set an IDT entry.  If the entry is part of the current IDT, then
+   also update Xen. */
+static void xen_write_idt_entry(struct desc_struct *dt, int entrynum,
+				u32 low, u32 high)
+{
+	unsigned long p = (unsigned long)&dt[entrynum];
+	unsigned long start, end;
+
+	preempt_disable();
+
+	start = __get_cpu_var(idt_desc).address;
+	end = start + __get_cpu_var(idt_desc).size + 1;
+
+	xen_mc_flush();
+
+	write_dt_entry(dt, entrynum, low, high);
+
+	if (p >= start && (p + 8) <= end) {
+		struct trap_info info[2];
+
+		info[1].address = 0;
+
+		if (cvt_gate_to_trap(entrynum, low, high, &info[0]))
+			if (HYPERVISOR_set_trap_table(info))
+				BUG();
+	}
+
+	preempt_enable();
+}
+
+static void xen_convert_trap_info(const struct Xgt_desc_struct *desc,
+				  struct trap_info *traps)
+{
+	unsigned in, out, count;
+
+	count = (desc->size+1) / 8;
+	BUG_ON(count > 256);
+
+	for (in = out = 0; in < count; in++) {
+		const u32 *entry = (u32 *)(desc->address + in * 8);
+
+		if (cvt_gate_to_trap(in, entry[0], entry[1], &traps[out]))
+			out++;
+	}
+	traps[out].address = 0;
+}
+
+void xen_copy_trap_info(struct trap_info *traps)
+{
+	const struct Xgt_desc_struct *desc = &__get_cpu_var(idt_desc);
+
+	xen_convert_trap_info(desc, traps);
+}
+
+/* Load a new IDT into Xen.  In principle this can be per-CPU, so we
+   hold a spinlock to protect the static traps[] array (static because
+   it avoids allocation, and saves stack space). */
+static void xen_load_idt(const struct Xgt_desc_struct *desc)
+{
+	static DEFINE_SPINLOCK(lock);
+	static struct trap_info traps[257];
+
+	spin_lock(&lock);
+
+	__get_cpu_var(idt_desc) = *desc;
+
+	xen_convert_trap_info(desc, traps);
+
+	xen_mc_flush();
+	if (HYPERVISOR_set_trap_table(traps))
+		BUG();
+
+	spin_unlock(&lock);
+}
+
+/* Write a GDT descriptor entry.  Ignore LDT descriptors, since
+   they're handled differently. */
+static void xen_write_gdt_entry(struct desc_struct *dt, int entry,
+				u32 low, u32 high)
+{
+	preempt_disable();
+
+	switch ((high >> 8) & 0xff) {
+	case DESCTYPE_LDT:
+	case DESCTYPE_TSS:
+		/* ignore */
+		break;
+
+	default: {
+		xmaddr_t maddr = virt_to_machine(&dt[entry]);
+		u64 desc = (u64)high << 32 | low;
+
+		xen_mc_flush();
+		if (HYPERVISOR_update_descriptor(maddr.maddr, desc))
+			BUG();
+	}
+
+	}
+
+	preempt_enable();
+}
+
+static void xen_load_esp0(struct tss_struct *tss,
+			  struct thread_struct *thread)
+{
+	struct multicall_space mcs = xen_mc_entry(0);
+	MULTI_stack_switch(mcs.mc, __KERNEL_DS, thread->esp0);
+	xen_mc_issue(PARAVIRT_LAZY_CPU);
+}
+
+static void xen_set_iopl_mask(unsigned mask)
+{
+	struct physdev_set_iopl set_iopl;
+
+	/* Force the change at ring 0. */
+	set_iopl.iopl = (mask == 0) ? 1 : (mask >> 12) & 3;
+	HYPERVISOR_physdev_op(PHYSDEVOP_set_iopl, &set_iopl);
+}
+
+static void xen_io_delay(void)
+{
+}
+
+#ifdef CONFIG_X86_LOCAL_APIC
+static unsigned long xen_apic_read(unsigned long reg)
+{
+	return 0;
+}
+
+static void xen_apic_write(unsigned long reg, unsigned long val)
+{
+	/* Warn to see if there's any stray references */
+	WARN_ON(1);
+}
+#endif
+
+static void xen_flush_tlb(void)
+{
+	struct mmuext_op *op;
+	struct multicall_space mcs = xen_mc_entry(sizeof(*op));
+
+	op = mcs.args;
+	op->cmd = MMUEXT_TLB_FLUSH_LOCAL;
+	MULTI_mmuext_op(mcs.mc, op, 1, NULL, DOMID_SELF);
+
+	xen_mc_issue(PARAVIRT_LAZY_MMU);
+}
+
+static void xen_flush_tlb_single(unsigned long addr)
+{
+	struct mmuext_op *op;
+	struct multicall_space mcs = xen_mc_entry(sizeof(*op));
+
+	op = mcs.args;
+	op->cmd = MMUEXT_INVLPG_LOCAL;
+	op->arg1.linear_addr = addr & PAGE_MASK;
+	MULTI_mmuext_op(mcs.mc, op, 1, NULL, DOMID_SELF);
+
+	xen_mc_issue(PARAVIRT_LAZY_MMU);
+}
+
+static void xen_flush_tlb_others(const cpumask_t *cpus, struct mm_struct *mm,
+				 unsigned long va)
+{
+	struct {
+		struct mmuext_op op;
+		cpumask_t mask;
+	} *args;
+	cpumask_t cpumask = *cpus;
+	struct multicall_space mcs;
+
+	/*
+	 * A couple of (to be removed) sanity checks:
+	 *
+	 * - current CPU must not be in mask
+	 * - mask must exist :)
+	 */
+	BUG_ON(cpus_empty(cpumask));
+	BUG_ON(cpu_isset(smp_processor_id(), cpumask));
+	BUG_ON(!mm);
+
+	/* If a CPU which we ran on has gone down, OK. */
+	cpus_and(cpumask, cpumask, cpu_online_map);
+	if (cpus_empty(cpumask))
+		return;
+
+	mcs = xen_mc_entry(sizeof(*args));
+	args = mcs.args;
+	args->mask = cpumask;
+	args->op.arg2.vcpumask = &args->mask;
+
+	if (va == TLB_FLUSH_ALL) {
+		args->op.cmd = MMUEXT_TLB_FLUSH_MULTI;
+	} else {
+		args->op.cmd = MMUEXT_INVLPG_MULTI;
+		args->op.arg1.linear_addr = va;
+	}
+
+	MULTI_mmuext_op(mcs.mc, &args->op, 1, NULL, DOMID_SELF);
+
+	xen_mc_issue(PARAVIRT_LAZY_MMU);
+}
+
+static void xen_write_cr2(unsigned long cr2)
+{
+	x86_read_percpu(xen_vcpu)->arch.cr2 = cr2;
+}
+
+static unsigned long xen_read_cr2(void)
+{
+	return x86_read_percpu(xen_vcpu)->arch.cr2;
+}
+
+static unsigned long xen_read_cr2_direct(void)
+{
+	return x86_read_percpu(xen_vcpu_info.arch.cr2);
+}
+
+static void xen_write_cr4(unsigned long cr4)
+{
+	/* Just ignore cr4 changes; Xen doesn't allow us to do
+	   anything anyway. */
+}
+
+static unsigned long xen_read_cr3(void)
+{
+	return x86_read_percpu(xen_cr3);
+}
+
+static void xen_write_cr3(unsigned long cr3)
+{
+	BUG_ON(preemptible());
+
+	if (cr3 == x86_read_percpu(xen_cr3)) {
+		/* just a simple tlb flush */
+		xen_flush_tlb();
+		return;
+	}
+
+	x86_write_percpu(xen_cr3, cr3);
+
+
+	{
+		struct mmuext_op *op;
+		struct multicall_space mcs = xen_mc_entry(sizeof(*op));
+		unsigned long mfn = pfn_to_mfn(PFN_DOWN(cr3));
+
+		op = mcs.args;
+		op->cmd = MMUEXT_NEW_BASEPTR;
+		op->arg1.mfn = mfn;
+
+		MULTI_mmuext_op(mcs.mc, op, 1, NULL, DOMID_SELF);
+
+		xen_mc_issue(PARAVIRT_LAZY_CPU);
+	}
+}
+
+/* Early in boot, while setting up the initial pagetable, assume
+   everything is pinned. */
+static __init void xen_alloc_pt_init(struct mm_struct *mm, u32 pfn)
+{
+	BUG_ON(mem_map);	/* should only be used early */
+	make_lowmem_page_readonly(__va(PFN_PHYS(pfn)));
+}
+
+/* This needs to make sure the new pte page is pinned iff its being
+   attached to a pinned pagetable. */
+static void xen_alloc_pt(struct mm_struct *mm, u32 pfn)
+{
+	struct page *page = pfn_to_page(pfn);
+
+	if (PagePinned(virt_to_page(mm->pgd))) {
+		SetPagePinned(page);
+
+		if (!PageHighMem(page))
+			make_lowmem_page_readonly(__va(PFN_PHYS(pfn)));
+		else
+			/* make sure there are no stray mappings of
+			   this page */
+			kmap_flush_unused();
+	}
+}
+
+/* This should never happen until we're OK to use struct page */
+static void xen_release_pt(u32 pfn)
+{
+	struct page *page = pfn_to_page(pfn);
+
+	if (PagePinned(page)) {
+		if (!PageHighMem(page))
+			make_lowmem_page_readwrite(__va(PFN_PHYS(pfn)));
+	}
+}
+
+#ifdef CONFIG_HIGHPTE
+static void *xen_kmap_atomic_pte(struct page *page, enum km_type type)
+{
+	pgprot_t prot = PAGE_KERNEL;
+
+	if (PagePinned(page))
+		prot = PAGE_KERNEL_RO;
+
+	if (0 && PageHighMem(page))
+		printk("mapping highpte %lx type %d prot %s\n",
+		       page_to_pfn(page), type,
+		       (unsigned long)pgprot_val(prot) & _PAGE_RW ? "WRITE" : "READ");
+
+	return kmap_atomic_prot(page, type, prot);
+}
+#endif
+
+static __init pte_t mask_rw_pte(pte_t *ptep, pte_t pte)
+{
+	/* If there's an existing pte, then don't allow _PAGE_RW to be set */
+	if (pte_val_ma(*ptep) & _PAGE_PRESENT)
+		pte = __pte_ma(((pte_val_ma(*ptep) & _PAGE_RW) | ~_PAGE_RW) &
+			       pte_val_ma(pte));
+
+	return pte;
+}
+
+/* Init-time set_pte while constructing initial pagetables, which
+   doesn't allow RO pagetable pages to be remapped RW */
+static __init void xen_set_pte_init(pte_t *ptep, pte_t pte)
+{
+	pte = mask_rw_pte(ptep, pte);
+
+	xen_set_pte(ptep, pte);
+}
+
+static __init void xen_pagetable_setup_start(pgd_t *base)
+{
+	pgd_t *xen_pgd = (pgd_t *)xen_start_info->pt_base;
+
+	/* special set_pte for pagetable initialization */
+	paravirt_ops.set_pte = xen_set_pte_init;
+
+	init_mm.pgd = base;
+	/*
+	 * copy top-level of Xen-supplied pagetable into place.	 For
+	 * !PAE we can use this as-is, but for PAE it is a stand-in
+	 * while we copy the pmd pages.
+	 */
+	memcpy(base, xen_pgd, PTRS_PER_PGD * sizeof(pgd_t));
+
+	if (PTRS_PER_PMD > 1) {
+		int i;
+		/*
+		 * For PAE, need to allocate new pmds, rather than
+		 * share Xen's, since Xen doesn't like pmd's being
+		 * shared between address spaces.
+		 */
+		for (i = 0; i < PTRS_PER_PGD; i++) {
+			if (pgd_val_ma(xen_pgd[i]) & _PAGE_PRESENT) {
+				pmd_t *pmd = (pmd_t *)alloc_bootmem_low_pages(PAGE_SIZE);
+
+				memcpy(pmd, (void *)pgd_page_vaddr(xen_pgd[i]),
+				       PAGE_SIZE);
+
+				make_lowmem_page_readonly(pmd);
+
+				set_pgd(&base[i], __pgd(1 + __pa(pmd)));
+			} else
+				pgd_clear(&base[i]);
+		}
+	}
+
+	/* make sure zero_page is mapped RO so we can use it in pagetables */
+	make_lowmem_page_readonly(empty_zero_page);
+	make_lowmem_page_readonly(base);
+	/*
+	 * Switch to new pagetable.  This is done before
+	 * pagetable_init has done anything so that the new pages
+	 * added to the table can be prepared properly for Xen.
+	 */
+	xen_write_cr3(__pa(base));
+}
+
+static __init void xen_pagetable_setup_done(pgd_t *base)
+{
+	/* This will work as long as patching hasn't happened yet
+	   (which it hasn't) */
+	paravirt_ops.alloc_pt = xen_alloc_pt;
+	paravirt_ops.set_pte = xen_set_pte;
+
+	if (!xen_feature(XENFEAT_auto_translated_physmap)) {
+		/*
+		 * Create a mapping for the shared info page.
+		 * Should be set_fixmap(), but shared_info is a machine
+		 * address with no corresponding pseudo-phys address.
+		 */
+		set_pte_mfn(fix_to_virt(FIX_PARAVIRT_BOOTMAP),
+			    PFN_DOWN(xen_start_info->shared_info),
+			    PAGE_KERNEL);
+
+		HYPERVISOR_shared_info =
+			(struct shared_info *)fix_to_virt(FIX_PARAVIRT_BOOTMAP);
+
+	} else
+		HYPERVISOR_shared_info =
+			(struct shared_info *)__va(xen_start_info->shared_info);
+
+	/* Actually pin the pagetable down, but we can't set PG_pinned
+	   yet because the page structures don't exist yet. */
+	{
+		struct mmuext_op op;
+#ifdef CONFIG_X86_PAE
+		op.cmd = MMUEXT_PIN_L3_TABLE;
+#else
+		op.cmd = MMUEXT_PIN_L3_TABLE;
+#endif
+		op.arg1.mfn = pfn_to_mfn(PFN_DOWN(__pa(base)));
+		if (HYPERVISOR_mmuext_op(&op, 1, NULL, DOMID_SELF))
+			BUG();
+	}
+}
+
+/* This is called once we have the cpu_possible_map */
+void __init xen_setup_vcpu_info_placement(void)
+{
+	int cpu;
+
+	for_each_possible_cpu(cpu)
+		xen_vcpu_setup(cpu);
+
+	/* xen_vcpu_setup managed to place the vcpu_info within the
+	   percpu area for all cpus, so make use of it */
+	if (have_vcpu_info_placement) {
+		printk(KERN_INFO "Xen: using vcpu_info placement\n");
+
+		paravirt_ops.save_fl = xen_save_fl_direct;
+		paravirt_ops.restore_fl = xen_restore_fl_direct;
+		paravirt_ops.irq_disable = xen_irq_disable_direct;
+		paravirt_ops.irq_enable = xen_irq_enable_direct;
+		paravirt_ops.read_cr2 = xen_read_cr2_direct;
+		paravirt_ops.iret = xen_iret_direct;
+	}
+}
+
+static unsigned xen_patch(u8 type, u16 clobbers, void *insnbuf,
+			  unsigned long addr, unsigned len)
+{
+	char *start, *end, *reloc;
+	unsigned ret;
+
+	start = end = reloc = NULL;
+
+#define SITE(x)								\
+	case PARAVIRT_PATCH(x):						\
+	if (have_vcpu_info_placement) {					\
+		start = (char *)xen_##x##_direct;			\
+		end = xen_##x##_direct_end;				\
+		reloc = xen_##x##_direct_reloc;				\
+	}								\
+	goto patch_site
+
+	switch (type) {
+		SITE(irq_enable);
+		SITE(irq_disable);
+		SITE(save_fl);
+		SITE(restore_fl);
+#undef SITE
+
+	patch_site:
+		if (start == NULL || (end-start) > len)
+			goto default_patch;
+
+		ret = paravirt_patch_insns(insnbuf, len, start, end);
+
+		/* Note: because reloc is assigned from something that
+		   appears to be an array, gcc assumes it's non-null,
+		   but doesn't know its relationship with start and
+		   end. */
+		if (reloc > start && reloc < end) {
+			int reloc_off = reloc - start;
+			long *relocp = (long *)(insnbuf + reloc_off);
+			long delta = start - (char *)addr;
+
+			*relocp += delta;
+		}
+		break;
+
+	default_patch:
+	default:
+		ret = paravirt_patch_default(type, clobbers, insnbuf,
+					     addr, len);
+		break;
+	}
+
+	return ret;
+}
+
+static const struct paravirt_ops xen_paravirt_ops __initdata = {
+	.paravirt_enabled = 1,
+	.shared_kernel_pmd = 0,
+
+	.name = "Xen",
+	.banner = xen_banner,
+
+	.patch = xen_patch,
+
+	.memory_setup = xen_memory_setup,
+	.arch_setup = xen_arch_setup,
+	.init_IRQ = xen_init_IRQ,
+	.post_allocator_init = xen_mark_init_mm_pinned,
+
+	.time_init = xen_time_init,
+	.set_wallclock = xen_set_wallclock,
+	.get_wallclock = xen_get_wallclock,
+	.get_cpu_khz = xen_cpu_khz,
+	.sched_clock = xen_sched_clock,
+
+	.cpuid = xen_cpuid,
+
+	.set_debugreg = xen_set_debugreg,
+	.get_debugreg = xen_get_debugreg,
+
+	.clts = native_clts,
+
+	.read_cr0 = native_read_cr0,
+	.write_cr0 = native_write_cr0,
+
+	.read_cr2 = xen_read_cr2,
+	.write_cr2 = xen_write_cr2,
+
+	.read_cr3 = xen_read_cr3,
+	.write_cr3 = xen_write_cr3,
+
+	.read_cr4 = native_read_cr4,
+	.read_cr4_safe = native_read_cr4_safe,
+	.write_cr4 = xen_write_cr4,
+
+	.save_fl = xen_save_fl,
+	.restore_fl = xen_restore_fl,
+	.irq_disable = xen_irq_disable,
+	.irq_enable = xen_irq_enable,
+	.safe_halt = xen_safe_halt,
+	.halt = xen_halt,
+	.wbinvd = native_wbinvd,
+
+	.read_msr = native_read_msr_safe,
+	.write_msr = native_write_msr_safe,
+	.read_tsc = native_read_tsc,
+	.read_pmc = native_read_pmc,
+
+	.iret = (void *)&hypercall_page[__HYPERVISOR_iret],
+	.irq_enable_sysexit = NULL,  /* never called */
+
+	.load_tr_desc = paravirt_nop,
+	.set_ldt = xen_set_ldt,
+	.load_gdt = xen_load_gdt,
+	.load_idt = xen_load_idt,
+	.load_tls = xen_load_tls,
+
+	.store_gdt = native_store_gdt,
+	.store_idt = native_store_idt,
+	.store_tr = xen_store_tr,
+
+	.write_ldt_entry = xen_write_ldt_entry,
+	.write_gdt_entry = xen_write_gdt_entry,
+	.write_idt_entry = xen_write_idt_entry,
+	.load_esp0 = xen_load_esp0,
+
+	.set_iopl_mask = xen_set_iopl_mask,
+	.io_delay = xen_io_delay,
+
+#ifdef CONFIG_X86_LOCAL_APIC
+	.apic_write = xen_apic_write,
+	.apic_write_atomic = xen_apic_write,
+	.apic_read = xen_apic_read,
+	.setup_boot_clock = paravirt_nop,
+	.setup_secondary_clock = paravirt_nop,
+	.startup_ipi_hook = paravirt_nop,
+#endif
+
+	.flush_tlb_user = xen_flush_tlb,
+	.flush_tlb_kernel = xen_flush_tlb,
+	.flush_tlb_single = xen_flush_tlb_single,
+	.flush_tlb_others = xen_flush_tlb_others,
+
+	.pte_update = paravirt_nop,
+	.pte_update_defer = paravirt_nop,
+
+	.pagetable_setup_start = xen_pagetable_setup_start,
+	.pagetable_setup_done = xen_pagetable_setup_done,
+
+	.alloc_pt = xen_alloc_pt_init,
+	.release_pt = xen_release_pt,
+	.alloc_pd = paravirt_nop,
+	.alloc_pd_clone = paravirt_nop,
+	.release_pd = paravirt_nop,
+
+#ifdef CONFIG_HIGHPTE
+	.kmap_atomic_pte = xen_kmap_atomic_pte,
+#endif
+
+	.set_pte = NULL,	/* see xen_pagetable_setup_* */
+	.set_pte_at = xen_set_pte_at,
+	.set_pmd = xen_set_pmd,
+
+	.pte_val = xen_pte_val,
+	.pgd_val = xen_pgd_val,
+
+	.make_pte = xen_make_pte,
+	.make_pgd = xen_make_pgd,
+
+#ifdef CONFIG_X86_PAE
+	.set_pte_atomic = xen_set_pte_atomic,
+	.set_pte_present = xen_set_pte_at,
+	.set_pud = xen_set_pud,
+	.pte_clear = xen_pte_clear,
+	.pmd_clear = xen_pmd_clear,
+
+	.make_pmd = xen_make_pmd,
+	.pmd_val = xen_pmd_val,
+#endif	/* PAE */
+
+	.activate_mm = xen_activate_mm,
+	.dup_mmap = xen_dup_mmap,
+	.exit_mmap = xen_exit_mmap,
+
+	.set_lazy_mode = xen_set_lazy_mode,
+};
+
+#ifdef CONFIG_SMP
+static const struct smp_ops xen_smp_ops __initdata = {
+	.smp_prepare_boot_cpu = xen_smp_prepare_boot_cpu,
+	.smp_prepare_cpus = xen_smp_prepare_cpus,
+	.cpu_up = xen_cpu_up,
+	.smp_cpus_done = xen_smp_cpus_done,
+
+	.smp_send_stop = xen_smp_send_stop,
+	.smp_send_reschedule = xen_smp_send_reschedule,
+	.smp_call_function_mask = xen_smp_call_function_mask,
+};
+#endif	/* CONFIG_SMP */
+
+static void xen_reboot(int reason)
+{
+#ifdef CONFIG_SMP
+	smp_send_stop();
+#endif
+
+	if (HYPERVISOR_sched_op(SCHEDOP_shutdown, reason))
+		BUG();
+}
+
+static void xen_restart(char *msg)
+{
+	xen_reboot(SHUTDOWN_reboot);
+}
+
+static void xen_emergency_restart(void)
+{
+	xen_reboot(SHUTDOWN_reboot);
+}
+
+static void xen_machine_halt(void)
+{
+	xen_reboot(SHUTDOWN_poweroff);
+}
+
+static void xen_crash_shutdown(struct pt_regs *regs)
+{
+	xen_reboot(SHUTDOWN_crash);
+}
+
+static const struct machine_ops __initdata xen_machine_ops = {
+	.restart = xen_restart,
+	.halt = xen_machine_halt,
+	.power_off = xen_machine_halt,
+	.shutdown = xen_machine_halt,
+	.crash_shutdown = xen_crash_shutdown,
+	.emergency_restart = xen_emergency_restart,
+};
+
+
+/* First C function to be called on Xen boot */
+asmlinkage void __init xen_start_kernel(void)
+{
+	pgd_t *pgd;
+
+	if (!xen_start_info)
+		return;
+
+	BUG_ON(memcmp(xen_start_info->magic, "xen-3.0", 7) != 0);
+
+	/* Install Xen paravirt ops */
+	paravirt_ops = xen_paravirt_ops;
+	machine_ops = xen_machine_ops;
+
+#ifdef CONFIG_SMP
+	smp_ops = xen_smp_ops;
+#endif
+
+	xen_setup_features();
+
+	/* Get mfn list */
+	if (!xen_feature(XENFEAT_auto_translated_physmap))
+		phys_to_machine_mapping = (unsigned long *)xen_start_info->mfn_list;
+
+	pgd = (pgd_t *)xen_start_info->pt_base;
+
+	init_pg_tables_end = __pa(pgd) + xen_start_info->nr_pt_frames*PAGE_SIZE;
+
+	init_mm.pgd = pgd; /* use the Xen pagetables to start */
+
+	/* keep using Xen gdt for now; no urgent need to change it */
+
+	x86_write_percpu(xen_cr3, __pa(pgd));
+
+#ifdef CONFIG_SMP
+	/* Don't do the full vcpu_info placement stuff until we have a
+	   possible map. */
+	per_cpu(xen_vcpu, 0) = &HYPERVISOR_shared_info->vcpu_info[0];
+#else
+	/* May as well do it now, since there's no good time to call
+	   it later on UP. */
+	xen_setup_vcpu_info_placement();
+#endif
+
+	paravirt_ops.kernel_rpl = 1;
+	if (xen_feature(XENFEAT_supervisor_mode_kernel))
+		paravirt_ops.kernel_rpl = 0;
+
+	/* set the limit of our address space */
+	reserve_top_address(-HYPERVISOR_VIRT_START + 2 * PAGE_SIZE);
+
+	/* set up basic CPUID stuff */
+	cpu_detect(&new_cpu_data);
+	new_cpu_data.hard_math = 1;
+	new_cpu_data.x86_capability[0] = cpuid_edx(1);
+
+	/* Poke various useful things into boot_params */
+	LOADER_TYPE = (9 << 4) | 0;
+	INITRD_START = xen_start_info->mod_start ? __pa(xen_start_info->mod_start) : 0;
+	INITRD_SIZE = xen_start_info->mod_len;
+
+	/* Start the world */
+	start_kernel();
+}
diff --git a/arch/x86/xen/events.c b/arch/x86/xen/events.c
new file mode 100644
index 00000000000..da1b173547a
--- /dev/null
+++ b/arch/x86/xen/events.c
@@ -0,0 +1,591 @@
+/*
+ * Xen event channels
+ *
+ * Xen models interrupts with abstract event channels.  Because each
+ * domain gets 1024 event channels, but NR_IRQ is not that large, we
+ * must dynamically map irqs<->event channels.  The event channels
+ * interface with the rest of the kernel by defining a xen interrupt
+ * chip.  When an event is recieved, it is mapped to an irq and sent
+ * through the normal interrupt processing path.
+ *
+ * There are four kinds of events which can be mapped to an event
+ * channel:
+ *
+ * 1. Inter-domain notifications.  This includes all the virtual
+ *    device events, since they're driven by front-ends in another domain
+ *    (typically dom0).
+ * 2. VIRQs, typically used for timers.  These are per-cpu events.
+ * 3. IPIs.
+ * 4. Hardware interrupts. Not supported at present.
+ *
+ * Jeremy Fitzhardinge <jeremy@xensource.com>, XenSource Inc, 2007
+ */
+
+#include <linux/linkage.h>
+#include <linux/interrupt.h>
+#include <linux/irq.h>
+#include <linux/module.h>
+#include <linux/string.h>
+
+#include <asm/ptrace.h>
+#include <asm/irq.h>
+#include <asm/sync_bitops.h>
+#include <asm/xen/hypercall.h>
+#include <asm/xen/hypervisor.h>
+
+#include <xen/events.h>
+#include <xen/interface/xen.h>
+#include <xen/interface/event_channel.h>
+
+#include "xen-ops.h"
+
+/*
+ * This lock protects updates to the following mapping and reference-count
+ * arrays. The lock does not need to be acquired to read the mapping tables.
+ */
+static DEFINE_SPINLOCK(irq_mapping_update_lock);
+
+/* IRQ <-> VIRQ mapping. */
+static DEFINE_PER_CPU(int, virq_to_irq[NR_VIRQS]) = {[0 ... NR_VIRQS-1] = -1};
+
+/* IRQ <-> IPI mapping */
+static DEFINE_PER_CPU(int, ipi_to_irq[XEN_NR_IPIS]) = {[0 ... XEN_NR_IPIS-1] = -1};
+
+/* Packed IRQ information: binding type, sub-type index, and event channel. */
+struct packed_irq
+{
+	unsigned short evtchn;
+	unsigned char index;
+	unsigned char type;
+};
+
+static struct packed_irq irq_info[NR_IRQS];
+
+/* Binding types. */
+enum {
+	IRQT_UNBOUND,
+	IRQT_PIRQ,
+	IRQT_VIRQ,
+	IRQT_IPI,
+	IRQT_EVTCHN
+};
+
+/* Convenient shorthand for packed representation of an unbound IRQ. */
+#define IRQ_UNBOUND	mk_irq_info(IRQT_UNBOUND, 0, 0)
+
+static int evtchn_to_irq[NR_EVENT_CHANNELS] = {
+	[0 ... NR_EVENT_CHANNELS-1] = -1
+};
+static unsigned long cpu_evtchn_mask[NR_CPUS][NR_EVENT_CHANNELS/BITS_PER_LONG];
+static u8 cpu_evtchn[NR_EVENT_CHANNELS];
+
+/* Reference counts for bindings to IRQs. */
+static int irq_bindcount[NR_IRQS];
+
+/* Xen will never allocate port zero for any purpose. */
+#define VALID_EVTCHN(chn)	((chn) != 0)
+
+/*
+ * Force a proper event-channel callback from Xen after clearing the
+ * callback mask. We do this in a very simple manner, by making a call
+ * down into Xen. The pending flag will be checked by Xen on return.
+ */
+void force_evtchn_callback(void)
+{
+	(void)HYPERVISOR_xen_version(0, NULL);
+}
+EXPORT_SYMBOL_GPL(force_evtchn_callback);
+
+static struct irq_chip xen_dynamic_chip;
+
+/* Constructor for packed IRQ information. */
+static inline struct packed_irq mk_irq_info(u32 type, u32 index, u32 evtchn)
+{
+	return (struct packed_irq) { evtchn, index, type };
+}
+
+/*
+ * Accessors for packed IRQ information.
+ */
+static inline unsigned int evtchn_from_irq(int irq)
+{
+	return irq_info[irq].evtchn;
+}
+
+static inline unsigned int index_from_irq(int irq)
+{
+	return irq_info[irq].index;
+}
+
+static inline unsigned int type_from_irq(int irq)
+{
+	return irq_info[irq].type;
+}
+
+static inline unsigned long active_evtchns(unsigned int cpu,
+					   struct shared_info *sh,
+					   unsigned int idx)
+{
+	return (sh->evtchn_pending[idx] &
+		cpu_evtchn_mask[cpu][idx] &
+		~sh->evtchn_mask[idx]);
+}
+
+static void bind_evtchn_to_cpu(unsigned int chn, unsigned int cpu)
+{
+	int irq = evtchn_to_irq[chn];
+
+	BUG_ON(irq == -1);
+#ifdef CONFIG_SMP
+	irq_desc[irq].affinity = cpumask_of_cpu(cpu);
+#endif
+
+	__clear_bit(chn, cpu_evtchn_mask[cpu_evtchn[chn]]);
+	__set_bit(chn, cpu_evtchn_mask[cpu]);
+
+	cpu_evtchn[chn] = cpu;
+}
+
+static void init_evtchn_cpu_bindings(void)
+{
+#ifdef CONFIG_SMP
+	int i;
+	/* By default all event channels notify CPU#0. */
+	for (i = 0; i < NR_IRQS; i++)
+		irq_desc[i].affinity = cpumask_of_cpu(0);
+#endif
+
+	memset(cpu_evtchn, 0, sizeof(cpu_evtchn));
+	memset(cpu_evtchn_mask[0], ~0, sizeof(cpu_evtchn_mask[0]));
+}
+
+static inline unsigned int cpu_from_evtchn(unsigned int evtchn)
+{
+	return cpu_evtchn[evtchn];
+}
+
+static inline void clear_evtchn(int port)
+{
+	struct shared_info *s = HYPERVISOR_shared_info;
+	sync_clear_bit(port, &s->evtchn_pending[0]);
+}
+
+static inline void set_evtchn(int port)
+{
+	struct shared_info *s = HYPERVISOR_shared_info;
+	sync_set_bit(port, &s->evtchn_pending[0]);
+}
+
+
+/**
+ * notify_remote_via_irq - send event to remote end of event channel via irq
+ * @irq: irq of event channel to send event to
+ *
+ * Unlike notify_remote_via_evtchn(), this is safe to use across
+ * save/restore. Notifications on a broken connection are silently
+ * dropped.
+ */
+void notify_remote_via_irq(int irq)
+{
+	int evtchn = evtchn_from_irq(irq);
+
+	if (VALID_EVTCHN(evtchn))
+		notify_remote_via_evtchn(evtchn);
+}
+EXPORT_SYMBOL_GPL(notify_remote_via_irq);
+
+static void mask_evtchn(int port)
+{
+	struct shared_info *s = HYPERVISOR_shared_info;
+	sync_set_bit(port, &s->evtchn_mask[0]);
+}
+
+static void unmask_evtchn(int port)
+{
+	struct shared_info *s = HYPERVISOR_shared_info;
+	unsigned int cpu = get_cpu();
+
+	BUG_ON(!irqs_disabled());
+
+	/* Slow path (hypercall) if this is a non-local port. */
+	if (unlikely(cpu != cpu_from_evtchn(port))) {
+		struct evtchn_unmask unmask = { .port = port };
+		(void)HYPERVISOR_event_channel_op(EVTCHNOP_unmask, &unmask);
+	} else {
+		struct vcpu_info *vcpu_info = __get_cpu_var(xen_vcpu);
+
+		sync_clear_bit(port, &s->evtchn_mask[0]);
+
+		/*
+		 * The following is basically the equivalent of
+		 * 'hw_resend_irq'. Just like a real IO-APIC we 'lose
+		 * the interrupt edge' if the channel is masked.
+		 */
+		if (sync_test_bit(port, &s->evtchn_pending[0]) &&
+		    !sync_test_and_set_bit(port / BITS_PER_LONG,
+					   &vcpu_info->evtchn_pending_sel))
+			vcpu_info->evtchn_upcall_pending = 1;
+	}
+
+	put_cpu();
+}
+
+static int find_unbound_irq(void)
+{
+	int irq;
+
+	/* Only allocate from dynirq range */
+	for (irq = 0; irq < NR_IRQS; irq++)
+		if (irq_bindcount[irq] == 0)
+			break;
+
+	if (irq == NR_IRQS)
+		panic("No available IRQ to bind to: increase NR_IRQS!\n");
+
+	return irq;
+}
+
+int bind_evtchn_to_irq(unsigned int evtchn)
+{
+	int irq;
+
+	spin_lock(&irq_mapping_update_lock);
+
+	irq = evtchn_to_irq[evtchn];
+
+	if (irq == -1) {
+		irq = find_unbound_irq();
+
+		dynamic_irq_init(irq);
+		set_irq_chip_and_handler_name(irq, &xen_dynamic_chip,
+					      handle_level_irq, "event");
+
+		evtchn_to_irq[evtchn] = irq;
+		irq_info[irq] = mk_irq_info(IRQT_EVTCHN, 0, evtchn);
+	}
+
+	irq_bindcount[irq]++;
+
+	spin_unlock(&irq_mapping_update_lock);
+
+	return irq;
+}
+EXPORT_SYMBOL_GPL(bind_evtchn_to_irq);
+
+static int bind_ipi_to_irq(unsigned int ipi, unsigned int cpu)
+{
+	struct evtchn_bind_ipi bind_ipi;
+	int evtchn, irq;
+
+	spin_lock(&irq_mapping_update_lock);
+
+	irq = per_cpu(ipi_to_irq, cpu)[ipi];
+	if (irq == -1) {
+		irq = find_unbound_irq();
+		if (irq < 0)
+			goto out;
+
+		dynamic_irq_init(irq);
+		set_irq_chip_and_handler_name(irq, &xen_dynamic_chip,
+					      handle_level_irq, "ipi");
+
+		bind_ipi.vcpu = cpu;
+		if (HYPERVISOR_event_channel_op(EVTCHNOP_bind_ipi,
+						&bind_ipi) != 0)
+			BUG();
+		evtchn = bind_ipi.port;
+
+		evtchn_to_irq[evtchn] = irq;
+		irq_info[irq] = mk_irq_info(IRQT_IPI, ipi, evtchn);
+
+		per_cpu(ipi_to_irq, cpu)[ipi] = irq;
+
+		bind_evtchn_to_cpu(evtchn, cpu);
+	}
+
+	irq_bindcount[irq]++;
+
+ out:
+	spin_unlock(&irq_mapping_update_lock);
+	return irq;
+}
+
+
+static int bind_virq_to_irq(unsigned int virq, unsigned int cpu)
+{
+	struct evtchn_bind_virq bind_virq;
+	int evtchn, irq;
+
+	spin_lock(&irq_mapping_update_lock);
+
+	irq = per_cpu(virq_to_irq, cpu)[virq];
+
+	if (irq == -1) {
+		bind_virq.virq = virq;
+		bind_virq.vcpu = cpu;
+		if (HYPERVISOR_event_channel_op(EVTCHNOP_bind_virq,
+						&bind_virq) != 0)
+			BUG();
+		evtchn = bind_virq.port;
+
+		irq = find_unbound_irq();
+
+		dynamic_irq_init(irq);
+		set_irq_chip_and_handler_name(irq, &xen_dynamic_chip,
+					      handle_level_irq, "virq");
+
+		evtchn_to_irq[evtchn] = irq;
+		irq_info[irq] = mk_irq_info(IRQT_VIRQ, virq, evtchn);
+
+		per_cpu(virq_to_irq, cpu)[virq] = irq;
+
+		bind_evtchn_to_cpu(evtchn, cpu);
+	}
+
+	irq_bindcount[irq]++;
+
+	spin_unlock(&irq_mapping_update_lock);
+
+	return irq;
+}
+
+static void unbind_from_irq(unsigned int irq)
+{
+	struct evtchn_close close;
+	int evtchn = evtchn_from_irq(irq);
+
+	spin_lock(&irq_mapping_update_lock);
+
+	if (VALID_EVTCHN(evtchn) && (--irq_bindcount[irq] == 0)) {
+		close.port = evtchn;
+		if (HYPERVISOR_event_channel_op(EVTCHNOP_close, &close) != 0)
+			BUG();
+
+		switch (type_from_irq(irq)) {
+		case IRQT_VIRQ:
+			per_cpu(virq_to_irq, cpu_from_evtchn(evtchn))
+				[index_from_irq(irq)] = -1;
+			break;
+		default:
+			break;
+		}
+
+		/* Closed ports are implicitly re-bound to VCPU0. */
+		bind_evtchn_to_cpu(evtchn, 0);
+
+		evtchn_to_irq[evtchn] = -1;
+		irq_info[irq] = IRQ_UNBOUND;
+
+		dynamic_irq_init(irq);
+	}
+
+	spin_unlock(&irq_mapping_update_lock);
+}
+
+int bind_evtchn_to_irqhandler(unsigned int evtchn,
+			      irqreturn_t (*handler)(int, void *),
+			      unsigned long irqflags,
+			      const char *devname, void *dev_id)
+{
+	unsigned int irq;
+	int retval;
+
+	irq = bind_evtchn_to_irq(evtchn);
+	retval = request_irq(irq, handler, irqflags, devname, dev_id);
+	if (retval != 0) {
+		unbind_from_irq(irq);
+		return retval;
+	}
+
+	return irq;
+}
+EXPORT_SYMBOL_GPL(bind_evtchn_to_irqhandler);
+
+int bind_virq_to_irqhandler(unsigned int virq, unsigned int cpu,
+			    irqreturn_t (*handler)(int, void *),
+			    unsigned long irqflags, const char *devname, void *dev_id)
+{
+	unsigned int irq;
+	int retval;
+
+	irq = bind_virq_to_irq(virq, cpu);
+	retval = request_irq(irq, handler, irqflags, devname, dev_id);
+	if (retval != 0) {
+		unbind_from_irq(irq);
+		return retval;
+	}
+
+	return irq;
+}
+EXPORT_SYMBOL_GPL(bind_virq_to_irqhandler);
+
+int bind_ipi_to_irqhandler(enum ipi_vector ipi,
+			   unsigned int cpu,
+			   irq_handler_t handler,
+			   unsigned long irqflags,
+			   const char *devname,
+			   void *dev_id)
+{
+	int irq, retval;
+
+	irq = bind_ipi_to_irq(ipi, cpu);
+	if (irq < 0)
+		return irq;
+
+	retval = request_irq(irq, handler, irqflags, devname, dev_id);
+	if (retval != 0) {
+		unbind_from_irq(irq);
+		return retval;
+	}
+
+	return irq;
+}
+
+void unbind_from_irqhandler(unsigned int irq, void *dev_id)
+{
+	free_irq(irq, dev_id);
+	unbind_from_irq(irq);
+}
+EXPORT_SYMBOL_GPL(unbind_from_irqhandler);
+
+void xen_send_IPI_one(unsigned int cpu, enum ipi_vector vector)
+{
+	int irq = per_cpu(ipi_to_irq, cpu)[vector];
+	BUG_ON(irq < 0);
+	notify_remote_via_irq(irq);
+}
+
+
+/*
+ * Search the CPUs pending events bitmasks.  For each one found, map
+ * the event number to an irq, and feed it into do_IRQ() for
+ * handling.
+ *
+ * Xen uses a two-level bitmap to speed searching.  The first level is
+ * a bitset of words which contain pending event bits.  The second
+ * level is a bitset of pending events themselves.
+ */
+fastcall void xen_evtchn_do_upcall(struct pt_regs *regs)
+{
+	int cpu = get_cpu();
+	struct shared_info *s = HYPERVISOR_shared_info;
+	struct vcpu_info *vcpu_info = __get_cpu_var(xen_vcpu);
+	unsigned long pending_words;
+
+	vcpu_info->evtchn_upcall_pending = 0;
+
+	/* NB. No need for a barrier here -- XCHG is a barrier on x86. */
+	pending_words = xchg(&vcpu_info->evtchn_pending_sel, 0);
+	while (pending_words != 0) {
+		unsigned long pending_bits;
+		int word_idx = __ffs(pending_words);
+		pending_words &= ~(1UL << word_idx);
+
+		while ((pending_bits = active_evtchns(cpu, s, word_idx)) != 0) {
+			int bit_idx = __ffs(pending_bits);
+			int port = (word_idx * BITS_PER_LONG) + bit_idx;
+			int irq = evtchn_to_irq[port];
+
+			if (irq != -1) {
+				regs->orig_eax = ~irq;
+				do_IRQ(regs);
+			}
+		}
+	}
+
+	put_cpu();
+}
+
+/* Rebind an evtchn so that it gets delivered to a specific cpu */
+static void rebind_irq_to_cpu(unsigned irq, unsigned tcpu)
+{
+	struct evtchn_bind_vcpu bind_vcpu;
+	int evtchn = evtchn_from_irq(irq);
+
+	if (!VALID_EVTCHN(evtchn))
+		return;
+
+	/* Send future instances of this interrupt to other vcpu. */
+	bind_vcpu.port = evtchn;
+	bind_vcpu.vcpu = tcpu;
+
+	/*
+	 * If this fails, it usually just indicates that we're dealing with a
+	 * virq or IPI channel, which don't actually need to be rebound. Ignore
+	 * it, but don't do the xenlinux-level rebind in that case.
+	 */
+	if (HYPERVISOR_event_channel_op(EVTCHNOP_bind_vcpu, &bind_vcpu) >= 0)
+		bind_evtchn_to_cpu(evtchn, tcpu);
+}
+
+
+static void set_affinity_irq(unsigned irq, cpumask_t dest)
+{
+	unsigned tcpu = first_cpu(dest);
+	rebind_irq_to_cpu(irq, tcpu);
+}
+
+static void enable_dynirq(unsigned int irq)
+{
+	int evtchn = evtchn_from_irq(irq);
+
+	if (VALID_EVTCHN(evtchn))
+		unmask_evtchn(evtchn);
+}
+
+static void disable_dynirq(unsigned int irq)
+{
+	int evtchn = evtchn_from_irq(irq);
+
+	if (VALID_EVTCHN(evtchn))
+		mask_evtchn(evtchn);
+}
+
+static void ack_dynirq(unsigned int irq)
+{
+	int evtchn = evtchn_from_irq(irq);
+
+	move_native_irq(irq);
+
+	if (VALID_EVTCHN(evtchn))
+		clear_evtchn(evtchn);
+}
+
+static int retrigger_dynirq(unsigned int irq)
+{
+	int evtchn = evtchn_from_irq(irq);
+	int ret = 0;
+
+	if (VALID_EVTCHN(evtchn)) {
+		set_evtchn(evtchn);
+		ret = 1;
+	}
+
+	return ret;
+}
+
+static struct irq_chip xen_dynamic_chip __read_mostly = {
+	.name		= "xen-dyn",
+	.mask		= disable_dynirq,
+	.unmask		= enable_dynirq,
+	.ack		= ack_dynirq,
+	.set_affinity	= set_affinity_irq,
+	.retrigger	= retrigger_dynirq,
+};
+
+void __init xen_init_IRQ(void)
+{
+	int i;
+
+	init_evtchn_cpu_bindings();
+
+	/* No event channels are 'live' right now. */
+	for (i = 0; i < NR_EVENT_CHANNELS; i++)
+		mask_evtchn(i);
+
+	/* Dynamic IRQ space is currently unbound. Zero the refcnts. */
+	for (i = 0; i < NR_IRQS; i++)
+		irq_bindcount[i] = 0;
+
+	irq_ctx_init(smp_processor_id());
+}
diff --git a/arch/x86/xen/features.c b/arch/x86/xen/features.c
new file mode 100644
index 00000000000..0707714e40d
--- /dev/null
+++ b/arch/x86/xen/features.c
@@ -0,0 +1,29 @@
+/******************************************************************************
+ * features.c
+ *
+ * Xen feature flags.
+ *
+ * Copyright (c) 2006, Ian Campbell, XenSource Inc.
+ */
+#include <linux/types.h>
+#include <linux/cache.h>
+#include <linux/module.h>
+#include <asm/xen/hypervisor.h>
+#include <xen/features.h>
+
+u8 xen_features[XENFEAT_NR_SUBMAPS * 32] __read_mostly;
+EXPORT_SYMBOL_GPL(xen_features);
+
+void xen_setup_features(void)
+{
+	struct xen_feature_info fi;
+	int i, j;
+
+	for (i = 0; i < XENFEAT_NR_SUBMAPS; i++) {
+		fi.submap_idx = i;
+		if (HYPERVISOR_xen_version(XENVER_get_features, &fi) < 0)
+			break;
+		for (j = 0; j < 32; j++)
+			xen_features[i * 32 + j] = !!(fi.submap & 1<<j);
+	}
+}
diff --git a/arch/x86/xen/manage.c b/arch/x86/xen/manage.c
new file mode 100644
index 00000000000..aa7af9e6abc
--- /dev/null
+++ b/arch/x86/xen/manage.c
@@ -0,0 +1,143 @@
+/*
+ * Handle extern requests for shutdown, reboot and sysrq
+ */
+#include <linux/kernel.h>
+#include <linux/err.h>
+#include <linux/reboot.h>
+#include <linux/sysrq.h>
+
+#include <xen/xenbus.h>
+
+#define SHUTDOWN_INVALID  -1
+#define SHUTDOWN_POWEROFF  0
+#define SHUTDOWN_SUSPEND   2
+/* Code 3 is SHUTDOWN_CRASH, which we don't use because the domain can only
+ * report a crash, not be instructed to crash!
+ * HALT is the same as POWEROFF, as far as we're concerned.  The tools use
+ * the distinction when we return the reason code to them.
+ */
+#define SHUTDOWN_HALT      4
+
+/* Ignore multiple shutdown requests. */
+static int shutting_down = SHUTDOWN_INVALID;
+
+static void shutdown_handler(struct xenbus_watch *watch,
+			     const char **vec, unsigned int len)
+{
+	char *str;
+	struct xenbus_transaction xbt;
+	int err;
+
+	if (shutting_down != SHUTDOWN_INVALID)
+		return;
+
+ again:
+	err = xenbus_transaction_start(&xbt);
+	if (err)
+		return;
+
+	str = (char *)xenbus_read(xbt, "control", "shutdown", NULL);
+	/* Ignore read errors and empty reads. */
+	if (XENBUS_IS_ERR_READ(str)) {
+		xenbus_transaction_end(xbt, 1);
+		return;
+	}
+
+	xenbus_write(xbt, "control", "shutdown", "");
+
+	err = xenbus_transaction_end(xbt, 0);
+	if (err == -EAGAIN) {
+		kfree(str);
+		goto again;
+	}
+
+	if (strcmp(str, "poweroff") == 0 ||
+	    strcmp(str, "halt") == 0)
+		orderly_poweroff(false);
+	else if (strcmp(str, "reboot") == 0)
+		ctrl_alt_del();
+	else {
+		printk(KERN_INFO "Ignoring shutdown request: %s\n", str);
+		shutting_down = SHUTDOWN_INVALID;
+	}
+
+	kfree(str);
+}
+
+static void sysrq_handler(struct xenbus_watch *watch, const char **vec,
+			  unsigned int len)
+{
+	char sysrq_key = '\0';
+	struct xenbus_transaction xbt;
+	int err;
+
+ again:
+	err = xenbus_transaction_start(&xbt);
+	if (err)
+		return;
+	if (!xenbus_scanf(xbt, "control", "sysrq", "%c", &sysrq_key)) {
+		printk(KERN_ERR "Unable to read sysrq code in "
+		       "control/sysrq\n");
+		xenbus_transaction_end(xbt, 1);
+		return;
+	}
+
+	if (sysrq_key != '\0')
+		xenbus_printf(xbt, "control", "sysrq", "%c", '\0');
+
+	err = xenbus_transaction_end(xbt, 0);
+	if (err == -EAGAIN)
+		goto again;
+
+	if (sysrq_key != '\0')
+		handle_sysrq(sysrq_key, NULL);
+}
+
+static struct xenbus_watch shutdown_watch = {
+	.node = "control/shutdown",
+	.callback = shutdown_handler
+};
+
+static struct xenbus_watch sysrq_watch = {
+	.node = "control/sysrq",
+	.callback = sysrq_handler
+};
+
+static int setup_shutdown_watcher(void)
+{
+	int err;
+
+	err = register_xenbus_watch(&shutdown_watch);
+	if (err) {
+		printk(KERN_ERR "Failed to set shutdown watcher\n");
+		return err;
+	}
+
+	err = register_xenbus_watch(&sysrq_watch);
+	if (err) {
+		printk(KERN_ERR "Failed to set sysrq watcher\n");
+		return err;
+	}
+
+	return 0;
+}
+
+static int shutdown_event(struct notifier_block *notifier,
+			  unsigned long event,
+			  void *data)
+{
+	setup_shutdown_watcher();
+	return NOTIFY_DONE;
+}
+
+static int __init setup_shutdown_event(void)
+{
+	static struct notifier_block xenstore_notifier = {
+		.notifier_call = shutdown_event
+	};
+	register_xenstore_notifier(&xenstore_notifier);
+
+	return 0;
+}
+
+subsys_initcall(setup_shutdown_event);
diff --git a/arch/x86/xen/mmu.c b/arch/x86/xen/mmu.c
new file mode 100644
index 00000000000..874db0cd1d2
--- /dev/null
+++ b/arch/x86/xen/mmu.c
@@ -0,0 +1,567 @@
+/*
+ * Xen mmu operations
+ *
+ * This file contains the various mmu fetch and update operations.
+ * The most important job they must perform is the mapping between the
+ * domain's pfn and the overall machine mfns.
+ *
+ * Xen allows guests to directly update the pagetable, in a controlled
+ * fashion.  In other words, the guest modifies the same pagetable
+ * that the CPU actually uses, which eliminates the overhead of having
+ * a separate shadow pagetable.
+ *
+ * In order to allow this, it falls on the guest domain to map its
+ * notion of a "physical" pfn - which is just a domain-local linear
+ * address - into a real "machine address" which the CPU's MMU can
+ * use.
+ *
+ * A pgd_t/pmd_t/pte_t will typically contain an mfn, and so can be
+ * inserted directly into the pagetable.  When creating a new
+ * pte/pmd/pgd, it converts the passed pfn into an mfn.  Conversely,
+ * when reading the content back with __(pgd|pmd|pte)_val, it converts
+ * the mfn back into a pfn.
+ *
+ * The other constraint is that all pages which make up a pagetable
+ * must be mapped read-only in the guest.  This prevents uncontrolled
+ * guest updates to the pagetable.  Xen strictly enforces this, and
+ * will disallow any pagetable update which will end up mapping a
+ * pagetable page RW, and will disallow using any writable page as a
+ * pagetable.
+ *
+ * Naively, when loading %cr3 with the base of a new pagetable, Xen
+ * would need to validate the whole pagetable before going on.
+ * Naturally, this is quite slow.  The solution is to "pin" a
+ * pagetable, which enforces all the constraints on the pagetable even
+ * when it is not actively in use.  This menas that Xen can be assured
+ * that it is still valid when you do load it into %cr3, and doesn't
+ * need to revalidate it.
+ *
+ * Jeremy Fitzhardinge <jeremy@xensource.com>, XenSource Inc, 2007
+ */
+#include <linux/sched.h>
+#include <linux/highmem.h>
+#include <linux/bug.h>
+#include <linux/sched.h>
+
+#include <asm/pgtable.h>
+#include <asm/tlbflush.h>
+#include <asm/mmu_context.h>
+#include <asm/paravirt.h>
+
+#include <asm/xen/hypercall.h>
+#include <asm/xen/hypervisor.h>
+
+#include <xen/page.h>
+#include <xen/interface/xen.h>
+
+#include "multicalls.h"
+#include "mmu.h"
+
+xmaddr_t arbitrary_virt_to_machine(unsigned long address)
+{
+	pte_t *pte = lookup_address(address);
+	unsigned offset = address & PAGE_MASK;
+
+	BUG_ON(pte == NULL);
+
+	return XMADDR((pte_mfn(*pte) << PAGE_SHIFT) + offset);
+}
+
+void make_lowmem_page_readonly(void *vaddr)
+{
+	pte_t *pte, ptev;
+	unsigned long address = (unsigned long)vaddr;
+
+	pte = lookup_address(address);
+	BUG_ON(pte == NULL);
+
+	ptev = pte_wrprotect(*pte);
+
+	if (HYPERVISOR_update_va_mapping(address, ptev, 0))
+		BUG();
+}
+
+void make_lowmem_page_readwrite(void *vaddr)
+{
+	pte_t *pte, ptev;
+	unsigned long address = (unsigned long)vaddr;
+
+	pte = lookup_address(address);
+	BUG_ON(pte == NULL);
+
+	ptev = pte_mkwrite(*pte);
+
+	if (HYPERVISOR_update_va_mapping(address, ptev, 0))
+		BUG();
+}
+
+
+void xen_set_pmd(pmd_t *ptr, pmd_t val)
+{
+	struct multicall_space mcs;
+	struct mmu_update *u;
+
+	preempt_disable();
+
+	mcs = xen_mc_entry(sizeof(*u));
+	u = mcs.args;
+	u->ptr = virt_to_machine(ptr).maddr;
+	u->val = pmd_val_ma(val);
+	MULTI_mmu_update(mcs.mc, u, 1, NULL, DOMID_SELF);
+
+	xen_mc_issue(PARAVIRT_LAZY_MMU);
+
+	preempt_enable();
+}
+
+/*
+ * Associate a virtual page frame with a given physical page frame
+ * and protection flags for that frame.
+ */
+void set_pte_mfn(unsigned long vaddr, unsigned long mfn, pgprot_t flags)
+{
+	pgd_t *pgd;
+	pud_t *pud;
+	pmd_t *pmd;
+	pte_t *pte;
+
+	pgd = swapper_pg_dir + pgd_index(vaddr);
+	if (pgd_none(*pgd)) {
+		BUG();
+		return;
+	}
+	pud = pud_offset(pgd, vaddr);
+	if (pud_none(*pud)) {
+		BUG();
+		return;
+	}
+	pmd = pmd_offset(pud, vaddr);
+	if (pmd_none(*pmd)) {
+		BUG();
+		return;
+	}
+	pte = pte_offset_kernel(pmd, vaddr);
+	/* <mfn,flags> stored as-is, to permit clearing entries */
+	xen_set_pte(pte, mfn_pte(mfn, flags));
+
+	/*
+	 * It's enough to flush this one mapping.
+	 * (PGE mappings get flushed as well)
+	 */
+	__flush_tlb_one(vaddr);
+}
+
+void xen_set_pte_at(struct mm_struct *mm, unsigned long addr,
+		    pte_t *ptep, pte_t pteval)
+{
+	if (mm == current->mm || mm == &init_mm) {
+		if (xen_get_lazy_mode() == PARAVIRT_LAZY_MMU) {
+			struct multicall_space mcs;
+			mcs = xen_mc_entry(0);
+
+			MULTI_update_va_mapping(mcs.mc, addr, pteval, 0);
+			xen_mc_issue(PARAVIRT_LAZY_MMU);
+			return;
+		} else
+			if (HYPERVISOR_update_va_mapping(addr, pteval, 0) == 0)
+				return;
+	}
+	xen_set_pte(ptep, pteval);
+}
+
+#ifdef CONFIG_X86_PAE
+void xen_set_pud(pud_t *ptr, pud_t val)
+{
+	struct multicall_space mcs;
+	struct mmu_update *u;
+
+	preempt_disable();
+
+	mcs = xen_mc_entry(sizeof(*u));
+	u = mcs.args;
+	u->ptr = virt_to_machine(ptr).maddr;
+	u->val = pud_val_ma(val);
+	MULTI_mmu_update(mcs.mc, u, 1, NULL, DOMID_SELF);
+
+	xen_mc_issue(PARAVIRT_LAZY_MMU);
+
+	preempt_enable();
+}
+
+void xen_set_pte(pte_t *ptep, pte_t pte)
+{
+	ptep->pte_high = pte.pte_high;
+	smp_wmb();
+	ptep->pte_low = pte.pte_low;
+}
+
+void xen_set_pte_atomic(pte_t *ptep, pte_t pte)
+{
+	set_64bit((u64 *)ptep, pte_val_ma(pte));
+}
+
+void xen_pte_clear(struct mm_struct *mm, unsigned long addr, pte_t *ptep)
+{
+	ptep->pte_low = 0;
+	smp_wmb();		/* make sure low gets written first */
+	ptep->pte_high = 0;
+}
+
+void xen_pmd_clear(pmd_t *pmdp)
+{
+	xen_set_pmd(pmdp, __pmd(0));
+}
+
+unsigned long long xen_pte_val(pte_t pte)
+{
+	unsigned long long ret = 0;
+
+	if (pte.pte_low) {
+		ret = ((unsigned long long)pte.pte_high << 32) | pte.pte_low;
+		ret = machine_to_phys(XMADDR(ret)).paddr | 1;
+	}
+
+	return ret;
+}
+
+unsigned long long xen_pmd_val(pmd_t pmd)
+{
+	unsigned long long ret = pmd.pmd;
+	if (ret)
+		ret = machine_to_phys(XMADDR(ret)).paddr | 1;
+	return ret;
+}
+
+unsigned long long xen_pgd_val(pgd_t pgd)
+{
+	unsigned long long ret = pgd.pgd;
+	if (ret)
+		ret = machine_to_phys(XMADDR(ret)).paddr | 1;
+	return ret;
+}
+
+pte_t xen_make_pte(unsigned long long pte)
+{
+	if (pte & 1)
+		pte = phys_to_machine(XPADDR(pte)).maddr;
+
+	return (pte_t){ pte, pte >> 32 };
+}
+
+pmd_t xen_make_pmd(unsigned long long pmd)
+{
+	if (pmd & 1)
+		pmd = phys_to_machine(XPADDR(pmd)).maddr;
+
+	return (pmd_t){ pmd };
+}
+
+pgd_t xen_make_pgd(unsigned long long pgd)
+{
+	if (pgd & _PAGE_PRESENT)
+		pgd = phys_to_machine(XPADDR(pgd)).maddr;
+
+	return (pgd_t){ pgd };
+}
+#else  /* !PAE */
+void xen_set_pte(pte_t *ptep, pte_t pte)
+{
+	*ptep = pte;
+}
+
+unsigned long xen_pte_val(pte_t pte)
+{
+	unsigned long ret = pte.pte_low;
+
+	if (ret & _PAGE_PRESENT)
+		ret = machine_to_phys(XMADDR(ret)).paddr;
+
+	return ret;
+}
+
+unsigned long xen_pgd_val(pgd_t pgd)
+{
+	unsigned long ret = pgd.pgd;
+	if (ret)
+		ret = machine_to_phys(XMADDR(ret)).paddr | 1;
+	return ret;
+}
+
+pte_t xen_make_pte(unsigned long pte)
+{
+	if (pte & _PAGE_PRESENT)
+		pte = phys_to_machine(XPADDR(pte)).maddr;
+
+	return (pte_t){ pte };
+}
+
+pgd_t xen_make_pgd(unsigned long pgd)
+{
+	if (pgd & _PAGE_PRESENT)
+		pgd = phys_to_machine(XPADDR(pgd)).maddr;
+
+	return (pgd_t){ pgd };
+}
+#endif	/* CONFIG_X86_PAE */
+
+
+
+/*
+  (Yet another) pagetable walker.  This one is intended for pinning a
+  pagetable.  This means that it walks a pagetable and calls the
+  callback function on each page it finds making up the page table,
+  at every level.  It walks the entire pagetable, but it only bothers
+  pinning pte pages which are below pte_limit.  In the normal case
+  this will be TASK_SIZE, but at boot we need to pin up to
+  FIXADDR_TOP.  But the important bit is that we don't pin beyond
+  there, because then we start getting into Xen's ptes.
+*/
+static int pgd_walk(pgd_t *pgd_base, int (*func)(struct page *, unsigned),
+		    unsigned long limit)
+{
+	pgd_t *pgd = pgd_base;
+	int flush = 0;
+	unsigned long addr = 0;
+	unsigned long pgd_next;
+
+	BUG_ON(limit > FIXADDR_TOP);
+
+	if (xen_feature(XENFEAT_auto_translated_physmap))
+		return 0;
+
+	for (; addr != FIXADDR_TOP; pgd++, addr = pgd_next) {
+		pud_t *pud;
+		unsigned long pud_limit, pud_next;
+
+		pgd_next = pud_limit = pgd_addr_end(addr, FIXADDR_TOP);
+
+		if (!pgd_val(*pgd))
+			continue;
+
+		pud = pud_offset(pgd, 0);
+
+		if (PTRS_PER_PUD > 1) /* not folded */
+			flush |= (*func)(virt_to_page(pud), 0);
+
+		for (; addr != pud_limit; pud++, addr = pud_next) {
+			pmd_t *pmd;
+			unsigned long pmd_limit;
+
+			pud_next = pud_addr_end(addr, pud_limit);
+
+			if (pud_next < limit)
+				pmd_limit = pud_next;
+			else
+				pmd_limit = limit;
+
+			if (pud_none(*pud))
+				continue;
+
+			pmd = pmd_offset(pud, 0);
+
+			if (PTRS_PER_PMD > 1) /* not folded */
+				flush |= (*func)(virt_to_page(pmd), 0);
+
+			for (; addr != pmd_limit; pmd++) {
+				addr += (PAGE_SIZE * PTRS_PER_PTE);
+				if ((pmd_limit-1) < (addr-1)) {
+					addr = pmd_limit;
+					break;
+				}
+
+				if (pmd_none(*pmd))
+					continue;
+
+				flush |= (*func)(pmd_page(*pmd), 0);
+			}
+		}
+	}
+
+	flush |= (*func)(virt_to_page(pgd_base), UVMF_TLB_FLUSH);
+
+	return flush;
+}
+
+static int pin_page(struct page *page, unsigned flags)
+{
+	unsigned pgfl = test_and_set_bit(PG_pinned, &page->flags);
+	int flush;
+
+	if (pgfl)
+		flush = 0;		/* already pinned */
+	else if (PageHighMem(page))
+		/* kmaps need flushing if we found an unpinned
+		   highpage */
+		flush = 1;
+	else {
+		void *pt = lowmem_page_address(page);
+		unsigned long pfn = page_to_pfn(page);
+		struct multicall_space mcs = __xen_mc_entry(0);
+
+		flush = 0;
+
+		MULTI_update_va_mapping(mcs.mc, (unsigned long)pt,
+					pfn_pte(pfn, PAGE_KERNEL_RO),
+					flags);
+	}
+
+	return flush;
+}
+
+/* This is called just after a mm has been created, but it has not
+   been used yet.  We need to make sure that its pagetable is all
+   read-only, and can be pinned. */
+void xen_pgd_pin(pgd_t *pgd)
+{
+	struct multicall_space mcs;
+	struct mmuext_op *op;
+
+	xen_mc_batch();
+
+	if (pgd_walk(pgd, pin_page, TASK_SIZE)) {
+		/* re-enable interrupts for kmap_flush_unused */
+		xen_mc_issue(0);
+		kmap_flush_unused();
+		xen_mc_batch();
+	}
+
+	mcs = __xen_mc_entry(sizeof(*op));
+	op = mcs.args;
+
+#ifdef CONFIG_X86_PAE
+	op->cmd = MMUEXT_PIN_L3_TABLE;
+#else
+	op->cmd = MMUEXT_PIN_L2_TABLE;
+#endif
+	op->arg1.mfn = pfn_to_mfn(PFN_DOWN(__pa(pgd)));
+	MULTI_mmuext_op(mcs.mc, op, 1, NULL, DOMID_SELF);
+
+	xen_mc_issue(0);
+}
+
+/* The init_mm pagetable is really pinned as soon as its created, but
+   that's before we have page structures to store the bits.  So do all
+   the book-keeping now. */
+static __init int mark_pinned(struct page *page, unsigned flags)
+{
+	SetPagePinned(page);
+	return 0;
+}
+
+void __init xen_mark_init_mm_pinned(void)
+{
+	pgd_walk(init_mm.pgd, mark_pinned, FIXADDR_TOP);
+}
+
+static int unpin_page(struct page *page, unsigned flags)
+{
+	unsigned pgfl = test_and_clear_bit(PG_pinned, &page->flags);
+
+	if (pgfl && !PageHighMem(page)) {
+		void *pt = lowmem_page_address(page);
+		unsigned long pfn = page_to_pfn(page);
+		struct multicall_space mcs = __xen_mc_entry(0);
+
+		MULTI_update_va_mapping(mcs.mc, (unsigned long)pt,
+					pfn_pte(pfn, PAGE_KERNEL),
+					flags);
+	}
+
+	return 0;		/* never need to flush on unpin */
+}
+
+/* Release a pagetables pages back as normal RW */
+static void xen_pgd_unpin(pgd_t *pgd)
+{
+	struct mmuext_op *op;
+	struct multicall_space mcs;
+
+	xen_mc_batch();
+
+	mcs = __xen_mc_entry(sizeof(*op));
+
+	op = mcs.args;
+	op->cmd = MMUEXT_UNPIN_TABLE;
+	op->arg1.mfn = pfn_to_mfn(PFN_DOWN(__pa(pgd)));
+
+	MULTI_mmuext_op(mcs.mc, op, 1, NULL, DOMID_SELF);
+
+	pgd_walk(pgd, unpin_page, TASK_SIZE);
+
+	xen_mc_issue(0);
+}
+
+void xen_activate_mm(struct mm_struct *prev, struct mm_struct *next)
+{
+	spin_lock(&next->page_table_lock);
+	xen_pgd_pin(next->pgd);
+	spin_unlock(&next->page_table_lock);
+}
+
+void xen_dup_mmap(struct mm_struct *oldmm, struct mm_struct *mm)
+{
+	spin_lock(&mm->page_table_lock);
+	xen_pgd_pin(mm->pgd);
+	spin_unlock(&mm->page_table_lock);
+}
+
+
+#ifdef CONFIG_SMP
+/* Another cpu may still have their %cr3 pointing at the pagetable, so
+   we need to repoint it somewhere else before we can unpin it. */
+static void drop_other_mm_ref(void *info)
+{
+	struct mm_struct *mm = info;
+
+	if (__get_cpu_var(cpu_tlbstate).active_mm == mm)
+		leave_mm(smp_processor_id());
+}
+
+static void drop_mm_ref(struct mm_struct *mm)
+{
+	if (current->active_mm == mm) {
+		if (current->mm == mm)
+			load_cr3(swapper_pg_dir);
+		else
+			leave_mm(smp_processor_id());
+	}
+
+	if (!cpus_empty(mm->cpu_vm_mask))
+		xen_smp_call_function_mask(mm->cpu_vm_mask, drop_other_mm_ref,
+					   mm, 1);
+}
+#else
+static void drop_mm_ref(struct mm_struct *mm)
+{
+	if (current->active_mm == mm)
+		load_cr3(swapper_pg_dir);
+}
+#endif
+
+/*
+ * While a process runs, Xen pins its pagetables, which means that the
+ * hypervisor forces it to be read-only, and it controls all updates
+ * to it.  This means that all pagetable updates have to go via the
+ * hypervisor, which is moderately expensive.
+ *
+ * Since we're pulling the pagetable down, we switch to use init_mm,
+ * unpin old process pagetable and mark it all read-write, which
+ * allows further operations on it to be simple memory accesses.
+ *
+ * The only subtle point is that another CPU may be still using the
+ * pagetable because of lazy tlb flushing.  This means we need need to
+ * switch all CPUs off this pagetable before we can unpin it.
+ */
+void xen_exit_mmap(struct mm_struct *mm)
+{
+	get_cpu();		/* make sure we don't move around */
+	drop_mm_ref(mm);
+	put_cpu();
+
+	spin_lock(&mm->page_table_lock);
+
+	/* pgd may not be pinned in the error exit path of execve */
+	if (PagePinned(virt_to_page(mm->pgd)))
+		xen_pgd_unpin(mm->pgd);
+	spin_unlock(&mm->page_table_lock);
+}
diff --git a/arch/x86/xen/mmu.h b/arch/x86/xen/mmu.h
new file mode 100644
index 00000000000..c9ff27f3ac3
--- /dev/null
+++ b/arch/x86/xen/mmu.h
@@ -0,0 +1,60 @@
+#ifndef _XEN_MMU_H
+
+#include <linux/linkage.h>
+#include <asm/page.h>
+
+/*
+ * Page-directory addresses above 4GB do not fit into architectural %cr3.
+ * When accessing %cr3, or equivalent field in vcpu_guest_context, guests
+ * must use the following accessor macros to pack/unpack valid MFNs.
+ *
+ * Note that Xen is using the fact that the pagetable base is always
+ * page-aligned, and putting the 12 MSB of the address into the 12 LSB
+ * of cr3.
+ */
+#define xen_pfn_to_cr3(pfn) (((unsigned)(pfn) << 12) | ((unsigned)(pfn) >> 20))
+#define xen_cr3_to_pfn(cr3) (((unsigned)(cr3) >> 12) | ((unsigned)(cr3) << 20))
+
+
+void set_pte_mfn(unsigned long vaddr, unsigned long pfn, pgprot_t flags);
+
+void xen_set_pte(pte_t *ptep, pte_t pteval);
+void xen_set_pte_at(struct mm_struct *mm, unsigned long addr,
+		    pte_t *ptep, pte_t pteval);
+void xen_set_pmd(pmd_t *pmdp, pmd_t pmdval);
+
+void xen_activate_mm(struct mm_struct *prev, struct mm_struct *next);
+void xen_dup_mmap(struct mm_struct *oldmm, struct mm_struct *mm);
+void xen_exit_mmap(struct mm_struct *mm);
+
+void xen_pgd_pin(pgd_t *pgd);
+//void xen_pgd_unpin(pgd_t *pgd);
+
+#ifdef CONFIG_X86_PAE
+unsigned long long xen_pte_val(pte_t);
+unsigned long long xen_pmd_val(pmd_t);
+unsigned long long xen_pgd_val(pgd_t);
+
+pte_t xen_make_pte(unsigned long long);
+pmd_t xen_make_pmd(unsigned long long);
+pgd_t xen_make_pgd(unsigned long long);
+
+void xen_set_pte_at(struct mm_struct *mm, unsigned long addr,
+		    pte_t *ptep, pte_t pteval);
+void xen_set_pte_atomic(pte_t *ptep, pte_t pte);
+void xen_set_pud(pud_t *ptr, pud_t val);
+void xen_pte_clear(struct mm_struct *mm, unsigned long addr, pte_t *ptep);
+void xen_pmd_clear(pmd_t *pmdp);
+
+
+#else
+unsigned long xen_pte_val(pte_t);
+unsigned long xen_pmd_val(pmd_t);
+unsigned long xen_pgd_val(pgd_t);
+
+pte_t xen_make_pte(unsigned long);
+pmd_t xen_make_pmd(unsigned long);
+pgd_t xen_make_pgd(unsigned long);
+#endif
+
+#endif	/* _XEN_MMU_H */
diff --git a/arch/x86/xen/multicalls.c b/arch/x86/xen/multicalls.c
new file mode 100644
index 00000000000..c837e8e463d
--- /dev/null
+++ b/arch/x86/xen/multicalls.c
@@ -0,0 +1,90 @@
+/*
+ * Xen hypercall batching.
+ *
+ * Xen allows multiple hypercalls to be issued at once, using the
+ * multicall interface.  This allows the cost of trapping into the
+ * hypervisor to be amortized over several calls.
+ *
+ * This file implements a simple interface for multicalls.  There's a
+ * per-cpu buffer of outstanding multicalls.  When you want to queue a
+ * multicall for issuing, you can allocate a multicall slot for the
+ * call and its arguments, along with storage for space which is
+ * pointed to by the arguments (for passing pointers to structures,
+ * etc).  When the multicall is actually issued, all the space for the
+ * commands and allocated memory is freed for reuse.
+ *
+ * Multicalls are flushed whenever any of the buffers get full, or
+ * when explicitly requested.  There's no way to get per-multicall
+ * return results back.  It will BUG if any of the multicalls fail.
+ *
+ * Jeremy Fitzhardinge <jeremy@xensource.com>, XenSource Inc, 2007
+ */
+#include <linux/percpu.h>
+#include <linux/hardirq.h>
+
+#include <asm/xen/hypercall.h>
+
+#include "multicalls.h"
+
+#define MC_BATCH	32
+#define MC_ARGS		(MC_BATCH * 16 / sizeof(u64))
+
+struct mc_buffer {
+	struct multicall_entry entries[MC_BATCH];
+	u64 args[MC_ARGS];
+	unsigned mcidx, argidx;
+};
+
+static DEFINE_PER_CPU(struct mc_buffer, mc_buffer);
+DEFINE_PER_CPU(unsigned long, xen_mc_irq_flags);
+
+void xen_mc_flush(void)
+{
+	struct mc_buffer *b = &__get_cpu_var(mc_buffer);
+	int ret = 0;
+	unsigned long flags;
+
+	BUG_ON(preemptible());
+
+	/* Disable interrupts in case someone comes in and queues
+	   something in the middle */
+	local_irq_save(flags);
+
+	if (b->mcidx) {
+		int i;
+
+		if (HYPERVISOR_multicall(b->entries, b->mcidx) != 0)
+			BUG();
+		for (i = 0; i < b->mcidx; i++)
+			if (b->entries[i].result < 0)
+				ret++;
+		b->mcidx = 0;
+		b->argidx = 0;
+	} else
+		BUG_ON(b->argidx != 0);
+
+	local_irq_restore(flags);
+
+	BUG_ON(ret);
+}
+
+struct multicall_space __xen_mc_entry(size_t args)
+{
+	struct mc_buffer *b = &__get_cpu_var(mc_buffer);
+	struct multicall_space ret;
+	unsigned argspace = (args + sizeof(u64) - 1) / sizeof(u64);
+
+	BUG_ON(preemptible());
+	BUG_ON(argspace > MC_ARGS);
+
+	if (b->mcidx == MC_BATCH ||
+	    (b->argidx + argspace) > MC_ARGS)
+		xen_mc_flush();
+
+	ret.mc = &b->entries[b->mcidx];
+	b->mcidx++;
+	ret.args = &b->args[b->argidx];
+	b->argidx += argspace;
+
+	return ret;
+}
diff --git a/arch/x86/xen/multicalls.h b/arch/x86/xen/multicalls.h
new file mode 100644
index 00000000000..e6f7530b156
--- /dev/null
+++ b/arch/x86/xen/multicalls.h
@@ -0,0 +1,45 @@
+#ifndef _XEN_MULTICALLS_H
+#define _XEN_MULTICALLS_H
+
+#include "xen-ops.h"
+
+/* Multicalls */
+struct multicall_space
+{
+	struct multicall_entry *mc;
+	void *args;
+};
+
+/* Allocate room for a multicall and its args */
+struct multicall_space __xen_mc_entry(size_t args);
+
+DECLARE_PER_CPU(unsigned long, xen_mc_irq_flags);
+
+/* Call to start a batch of multiple __xen_mc_entry()s.  Must be
+   paired with xen_mc_issue() */
+static inline void xen_mc_batch(void)
+{
+	/* need to disable interrupts until this entry is complete */
+	local_irq_save(__get_cpu_var(xen_mc_irq_flags));
+}
+
+static inline struct multicall_space xen_mc_entry(size_t args)
+{
+	xen_mc_batch();
+	return __xen_mc_entry(args);
+}
+
+/* Flush all pending multicalls */
+void xen_mc_flush(void);
+
+/* Issue a multicall if we're not in a lazy mode */
+static inline void xen_mc_issue(unsigned mode)
+{
+	if ((xen_get_lazy_mode() & mode) == 0)
+		xen_mc_flush();
+
+	/* restore flags saved in xen_mc_batch */
+	local_irq_restore(x86_read_percpu(xen_mc_irq_flags));
+}
+
+#endif /* _XEN_MULTICALLS_H */
diff --git a/arch/x86/xen/setup.c b/arch/x86/xen/setup.c
new file mode 100644
index 00000000000..f84e7722664
--- /dev/null
+++ b/arch/x86/xen/setup.c
@@ -0,0 +1,111 @@
+/*
+ * Machine specific setup for xen
+ *
+ * Jeremy Fitzhardinge <jeremy@xensource.com>, XenSource Inc, 2007
+ */
+
+#include <linux/module.h>
+#include <linux/sched.h>
+#include <linux/mm.h>
+#include <linux/pm.h>
+
+#include <asm/elf.h>
+#include <asm/e820.h>
+#include <asm/setup.h>
+#include <asm/xen/hypervisor.h>
+#include <asm/xen/hypercall.h>
+
+#include <xen/interface/physdev.h>
+#include <xen/features.h>
+
+#include "xen-ops.h"
+#include "vdso.h"
+
+/* These are code, but not functions.  Defined in entry.S */
+extern const char xen_hypervisor_callback[];
+extern const char xen_failsafe_callback[];
+
+unsigned long *phys_to_machine_mapping;
+EXPORT_SYMBOL(phys_to_machine_mapping);
+
+/**
+ * machine_specific_memory_setup - Hook for machine specific memory setup.
+ **/
+
+char * __init xen_memory_setup(void)
+{
+	unsigned long max_pfn = xen_start_info->nr_pages;
+
+	e820.nr_map = 0;
+	add_memory_region(0, PFN_PHYS(max_pfn), E820_RAM);
+
+	return "Xen";
+}
+
+static void xen_idle(void)
+{
+	local_irq_disable();
+
+	if (need_resched())
+		local_irq_enable();
+	else {
+		current_thread_info()->status &= ~TS_POLLING;
+		smp_mb__after_clear_bit();
+		safe_halt();
+		current_thread_info()->status |= TS_POLLING;
+	}
+}
+
+/*
+ * Set the bit indicating "nosegneg" library variants should be used.
+ */
+static void fiddle_vdso(void)
+{
+	extern u32 VDSO_NOTE_MASK; /* See ../kernel/vsyscall-note.S.  */
+	extern char vsyscall_int80_start;
+	u32 *mask = (u32 *) ((unsigned long) &VDSO_NOTE_MASK - VDSO_PRELINK +
+			     &vsyscall_int80_start);
+	*mask |= 1 << VDSO_NOTE_NONEGSEG_BIT;
+}
+
+void __init xen_arch_setup(void)
+{
+	struct physdev_set_iopl set_iopl;
+	int rc;
+
+	HYPERVISOR_vm_assist(VMASST_CMD_enable, VMASST_TYPE_4gb_segments);
+	HYPERVISOR_vm_assist(VMASST_CMD_enable, VMASST_TYPE_writable_pagetables);
+
+	if (!xen_feature(XENFEAT_auto_translated_physmap))
+		HYPERVISOR_vm_assist(VMASST_CMD_enable, VMASST_TYPE_pae_extended_cr3);
+
+	HYPERVISOR_set_callbacks(__KERNEL_CS, (unsigned long)xen_hypervisor_callback,
+				 __KERNEL_CS, (unsigned long)xen_failsafe_callback);
+
+	set_iopl.iopl = 1;
+	rc = HYPERVISOR_physdev_op(PHYSDEVOP_set_iopl, &set_iopl);
+	if (rc != 0)
+		printk(KERN_INFO "physdev_op failed %d\n", rc);
+
+#ifdef CONFIG_ACPI
+	if (!(xen_start_info->flags & SIF_INITDOMAIN)) {
+		printk(KERN_INFO "ACPI in unprivileged domain disabled\n");
+		disable_acpi();
+	}
+#endif
+
+	memcpy(boot_command_line, xen_start_info->cmd_line,
+	       MAX_GUEST_CMDLINE > COMMAND_LINE_SIZE ?
+	       COMMAND_LINE_SIZE : MAX_GUEST_CMDLINE);
+
+	pm_idle = xen_idle;
+
+#ifdef CONFIG_SMP
+	/* fill cpus_possible with all available cpus */
+	xen_fill_possible_map();
+#endif
+
+	paravirt_disable_iospace();
+
+	fiddle_vdso();
+}
diff --git a/arch/x86/xen/smp.c b/arch/x86/xen/smp.c
new file mode 100644
index 00000000000..557b8e24706
--- /dev/null
+++ b/arch/x86/xen/smp.c
@@ -0,0 +1,404 @@
+/*
+ * Xen SMP support
+ *
+ * This file implements the Xen versions of smp_ops.  SMP under Xen is
+ * very straightforward.  Bringing a CPU up is simply a matter of
+ * loading its initial context and setting it running.
+ *
+ * IPIs are handled through the Xen event mechanism.
+ *
+ * Because virtual CPUs can be scheduled onto any real CPU, there's no
+ * useful topology information for the kernel to make use of.  As a
+ * result, all CPUs are treated as if they're single-core and
+ * single-threaded.
+ *
+ * This does not handle HOTPLUG_CPU yet.
+ */
+#include <linux/sched.h>
+#include <linux/err.h>
+#include <linux/smp.h>
+
+#include <asm/paravirt.h>
+#include <asm/desc.h>
+#include <asm/pgtable.h>
+#include <asm/cpu.h>
+
+#include <xen/interface/xen.h>
+#include <xen/interface/vcpu.h>
+
+#include <asm/xen/interface.h>
+#include <asm/xen/hypercall.h>
+
+#include <xen/page.h>
+#include <xen/events.h>
+
+#include "xen-ops.h"
+#include "mmu.h"
+
+static cpumask_t cpu_initialized_map;
+static DEFINE_PER_CPU(int, resched_irq);
+static DEFINE_PER_CPU(int, callfunc_irq);
+
+/*
+ * Structure and data for smp_call_function(). This is designed to minimise
+ * static memory requirements. It also looks cleaner.
+ */
+static DEFINE_SPINLOCK(call_lock);
+
+struct call_data_struct {
+	void (*func) (void *info);
+	void *info;
+	atomic_t started;
+	atomic_t finished;
+	int wait;
+};
+
+static irqreturn_t xen_call_function_interrupt(int irq, void *dev_id);
+
+static struct call_data_struct *call_data;
+
+/*
+ * Reschedule call back. Nothing to do,
+ * all the work is done automatically when
+ * we return from the interrupt.
+ */
+static irqreturn_t xen_reschedule_interrupt(int irq, void *dev_id)
+{
+	return IRQ_HANDLED;
+}
+
+static __cpuinit void cpu_bringup_and_idle(void)
+{
+	int cpu = smp_processor_id();
+
+	cpu_init();
+
+	preempt_disable();
+	per_cpu(cpu_state, cpu) = CPU_ONLINE;
+
+	xen_setup_cpu_clockevents();
+
+	/* We can take interrupts now: we're officially "up". */
+	local_irq_enable();
+
+	wmb();			/* make sure everything is out */
+	cpu_idle();
+}
+
+static int xen_smp_intr_init(unsigned int cpu)
+{
+	int rc;
+	const char *resched_name, *callfunc_name;
+
+	per_cpu(resched_irq, cpu) = per_cpu(callfunc_irq, cpu) = -1;
+
+	resched_name = kasprintf(GFP_KERNEL, "resched%d", cpu);
+	rc = bind_ipi_to_irqhandler(XEN_RESCHEDULE_VECTOR,
+				    cpu,
+				    xen_reschedule_interrupt,
+				    IRQF_DISABLED|IRQF_PERCPU|IRQF_NOBALANCING,
+				    resched_name,
+				    NULL);
+	if (rc < 0)
+		goto fail;
+	per_cpu(resched_irq, cpu) = rc;
+
+	callfunc_name = kasprintf(GFP_KERNEL, "callfunc%d", cpu);
+	rc = bind_ipi_to_irqhandler(XEN_CALL_FUNCTION_VECTOR,
+				    cpu,
+				    xen_call_function_interrupt,
+				    IRQF_DISABLED|IRQF_PERCPU|IRQF_NOBALANCING,
+				    callfunc_name,
+				    NULL);
+	if (rc < 0)
+		goto fail;
+	per_cpu(callfunc_irq, cpu) = rc;
+
+	return 0;
+
+ fail:
+	if (per_cpu(resched_irq, cpu) >= 0)
+		unbind_from_irqhandler(per_cpu(resched_irq, cpu), NULL);
+	if (per_cpu(callfunc_irq, cpu) >= 0)
+		unbind_from_irqhandler(per_cpu(callfunc_irq, cpu), NULL);
+	return rc;
+}
+
+void __init xen_fill_possible_map(void)
+{
+	int i, rc;
+
+	for (i = 0; i < NR_CPUS; i++) {
+		rc = HYPERVISOR_vcpu_op(VCPUOP_is_up, i, NULL);
+		if (rc >= 0)
+			cpu_set(i, cpu_possible_map);
+	}
+}
+
+void __init xen_smp_prepare_boot_cpu(void)
+{
+	int cpu;
+
+	BUG_ON(smp_processor_id() != 0);
+	native_smp_prepare_boot_cpu();
+
+	/* We've switched to the "real" per-cpu gdt, so make sure the
+	   old memory can be recycled */
+	make_lowmem_page_readwrite(&per_cpu__gdt_page);
+
+	for (cpu = 0; cpu < NR_CPUS; cpu++) {
+		cpus_clear(cpu_sibling_map[cpu]);
+		cpus_clear(cpu_core_map[cpu]);
+	}
+
+	xen_setup_vcpu_info_placement();
+}
+
+void __init xen_smp_prepare_cpus(unsigned int max_cpus)
+{
+	unsigned cpu;
+
+	for (cpu = 0; cpu < NR_CPUS; cpu++) {
+		cpus_clear(cpu_sibling_map[cpu]);
+		cpus_clear(cpu_core_map[cpu]);
+	}
+
+	smp_store_cpu_info(0);
+	set_cpu_sibling_map(0);
+
+	if (xen_smp_intr_init(0))
+		BUG();
+
+	cpu_initialized_map = cpumask_of_cpu(0);
+
+	/* Restrict the possible_map according to max_cpus. */
+	while ((num_possible_cpus() > 1) && (num_possible_cpus() > max_cpus)) {
+		for (cpu = NR_CPUS-1; !cpu_isset(cpu, cpu_possible_map); cpu--)
+			continue;
+		cpu_clear(cpu, cpu_possible_map);
+	}
+
+	for_each_possible_cpu (cpu) {
+		struct task_struct *idle;
+
+		if (cpu == 0)
+			continue;
+
+		idle = fork_idle(cpu);
+		if (IS_ERR(idle))
+			panic("failed fork for CPU %d", cpu);
+
+		cpu_set(cpu, cpu_present_map);
+	}
+
+	//init_xenbus_allowed_cpumask();
+}
+
+static __cpuinit int
+cpu_initialize_context(unsigned int cpu, struct task_struct *idle)
+{
+	struct vcpu_guest_context *ctxt;
+	struct gdt_page *gdt = &per_cpu(gdt_page, cpu);
+
+	if (cpu_test_and_set(cpu, cpu_initialized_map))
+		return 0;
+
+	ctxt = kzalloc(sizeof(*ctxt), GFP_KERNEL);
+	if (ctxt == NULL)
+		return -ENOMEM;
+
+	ctxt->flags = VGCF_IN_KERNEL;
+	ctxt->user_regs.ds = __USER_DS;
+	ctxt->user_regs.es = __USER_DS;
+	ctxt->user_regs.fs = __KERNEL_PERCPU;
+	ctxt->user_regs.gs = 0;
+	ctxt->user_regs.ss = __KERNEL_DS;
+	ctxt->user_regs.eip = (unsigned long)cpu_bringup_and_idle;
+	ctxt->user_regs.eflags = 0x1000; /* IOPL_RING1 */
+
+	memset(&ctxt->fpu_ctxt, 0, sizeof(ctxt->fpu_ctxt));
+
+	xen_copy_trap_info(ctxt->trap_ctxt);
+
+	ctxt->ldt_ents = 0;
+
+	BUG_ON((unsigned long)gdt->gdt & ~PAGE_MASK);
+	make_lowmem_page_readonly(gdt->gdt);
+
+	ctxt->gdt_frames[0] = virt_to_mfn(gdt->gdt);
+	ctxt->gdt_ents      = ARRAY_SIZE(gdt->gdt);
+
+	ctxt->user_regs.cs = __KERNEL_CS;
+	ctxt->user_regs.esp = idle->thread.esp0 - sizeof(struct pt_regs);
+
+	ctxt->kernel_ss = __KERNEL_DS;
+	ctxt->kernel_sp = idle->thread.esp0;
+
+	ctxt->event_callback_cs     = __KERNEL_CS;
+	ctxt->event_callback_eip    = (unsigned long)xen_hypervisor_callback;
+	ctxt->failsafe_callback_cs  = __KERNEL_CS;
+	ctxt->failsafe_callback_eip = (unsigned long)xen_failsafe_callback;
+
+	per_cpu(xen_cr3, cpu) = __pa(swapper_pg_dir);
+	ctxt->ctrlreg[3] = xen_pfn_to_cr3(virt_to_mfn(swapper_pg_dir));
+
+	if (HYPERVISOR_vcpu_op(VCPUOP_initialise, cpu, ctxt))
+		BUG();
+
+	kfree(ctxt);
+	return 0;
+}
+
+int __cpuinit xen_cpu_up(unsigned int cpu)
+{
+	struct task_struct *idle = idle_task(cpu);
+	int rc;
+
+#if 0
+	rc = cpu_up_check(cpu);
+	if (rc)
+		return rc;
+#endif
+
+	init_gdt(cpu);
+	per_cpu(current_task, cpu) = idle;
+	irq_ctx_init(cpu);
+	xen_setup_timer(cpu);
+
+	/* make sure interrupts start blocked */
+	per_cpu(xen_vcpu, cpu)->evtchn_upcall_mask = 1;
+
+	rc = cpu_initialize_context(cpu, idle);
+	if (rc)
+		return rc;
+
+	if (num_online_cpus() == 1)
+		alternatives_smp_switch(1);
+
+	rc = xen_smp_intr_init(cpu);
+	if (rc)
+		return rc;
+
+	smp_store_cpu_info(cpu);
+	set_cpu_sibling_map(cpu);
+	/* This must be done before setting cpu_online_map */
+	wmb();
+
+	cpu_set(cpu, cpu_online_map);
+
+	rc = HYPERVISOR_vcpu_op(VCPUOP_up, cpu, NULL);
+	BUG_ON(rc);
+
+	return 0;
+}
+
+void xen_smp_cpus_done(unsigned int max_cpus)
+{
+}
+
+static void stop_self(void *v)
+{
+	int cpu = smp_processor_id();
+
+	/* make sure we're not pinning something down */
+	load_cr3(swapper_pg_dir);
+	/* should set up a minimal gdt */
+
+	HYPERVISOR_vcpu_op(VCPUOP_down, cpu, NULL);
+	BUG();
+}
+
+void xen_smp_send_stop(void)
+{
+	smp_call_function(stop_self, NULL, 0, 0);
+}
+
+void xen_smp_send_reschedule(int cpu)
+{
+	xen_send_IPI_one(cpu, XEN_RESCHEDULE_VECTOR);
+}
+
+
+static void xen_send_IPI_mask(cpumask_t mask, enum ipi_vector vector)
+{
+	unsigned cpu;
+
+	cpus_and(mask, mask, cpu_online_map);
+
+	for_each_cpu_mask(cpu, mask)
+		xen_send_IPI_one(cpu, vector);
+}
+
+static irqreturn_t xen_call_function_interrupt(int irq, void *dev_id)
+{
+	void (*func) (void *info) = call_data->func;
+	void *info = call_data->info;
+	int wait = call_data->wait;
+
+	/*
+	 * Notify initiating CPU that I've grabbed the data and am
+	 * about to execute the function
+	 */
+	mb();
+	atomic_inc(&call_data->started);
+	/*
+	 * At this point the info structure may be out of scope unless wait==1
+	 */
+	irq_enter();
+	(*func)(info);
+	irq_exit();
+
+	if (wait) {
+		mb();		/* commit everything before setting finished */
+		atomic_inc(&call_data->finished);
+	}
+
+	return IRQ_HANDLED;
+}
+
+int xen_smp_call_function_mask(cpumask_t mask, void (*func)(void *),
+			       void *info, int wait)
+{
+	struct call_data_struct data;
+	int cpus;
+
+	/* Holding any lock stops cpus from going down. */
+	spin_lock(&call_lock);
+
+	cpu_clear(smp_processor_id(), mask);
+
+	cpus = cpus_weight(mask);
+	if (!cpus) {
+		spin_unlock(&call_lock);
+		return 0;
+	}
+
+	/* Can deadlock when called with interrupts disabled */
+	WARN_ON(irqs_disabled());
+
+	data.func = func;
+	data.info = info;
+	atomic_set(&data.started, 0);
+	data.wait = wait;
+	if (wait)
+		atomic_set(&data.finished, 0);
+
+	call_data = &data;
+	mb();			/* write everything before IPI */
+
+	/* Send a message to other CPUs and wait for them to respond */
+	xen_send_IPI_mask(mask, XEN_CALL_FUNCTION_VECTOR);
+
+	/* Make sure other vcpus get a chance to run.
+	   XXX too severe?  Maybe we should check the other CPU's states? */
+	HYPERVISOR_sched_op(SCHEDOP_yield, 0);
+
+	/* Wait for response */
+	while (atomic_read(&data.started) != cpus ||
+	       (wait && atomic_read(&data.finished) != cpus))
+		cpu_relax();
+
+	spin_unlock(&call_lock);
+
+	return 0;
+}
diff --git a/arch/x86/xen/time.c b/arch/x86/xen/time.c
new file mode 100644
index 00000000000..dfd6db69ead
--- /dev/null
+++ b/arch/x86/xen/time.c
@@ -0,0 +1,593 @@
+/*
+ * Xen time implementation.
+ *
+ * This is implemented in terms of a clocksource driver which uses
+ * the hypervisor clock as a nanosecond timebase, and a clockevent
+ * driver which uses the hypervisor's timer mechanism.
+ *
+ * Jeremy Fitzhardinge <jeremy@xensource.com>, XenSource Inc, 2007
+ */
+#include <linux/kernel.h>
+#include <linux/interrupt.h>
+#include <linux/clocksource.h>
+#include <linux/clockchips.h>
+#include <linux/kernel_stat.h>
+
+#include <asm/xen/hypervisor.h>
+#include <asm/xen/hypercall.h>
+
+#include <xen/events.h>
+#include <xen/interface/xen.h>
+#include <xen/interface/vcpu.h>
+
+#include "xen-ops.h"
+
+#define XEN_SHIFT 22
+
+/* Xen may fire a timer up to this many ns early */
+#define TIMER_SLOP	100000
+#define NS_PER_TICK	(1000000000LL / HZ)
+
+static cycle_t xen_clocksource_read(void);
+
+/* These are perodically updated in shared_info, and then copied here. */
+struct shadow_time_info {
+	u64 tsc_timestamp;     /* TSC at last update of time vals.  */
+	u64 system_timestamp;  /* Time, in nanosecs, since boot.    */
+	u32 tsc_to_nsec_mul;
+	int tsc_shift;
+	u32 version;
+};
+
+static DEFINE_PER_CPU(struct shadow_time_info, shadow_time);
+
+/* runstate info updated by Xen */
+static DEFINE_PER_CPU(struct vcpu_runstate_info, runstate);
+
+/* snapshots of runstate info */
+static DEFINE_PER_CPU(struct vcpu_runstate_info, runstate_snapshot);
+
+/* unused ns of stolen and blocked time */
+static DEFINE_PER_CPU(u64, residual_stolen);
+static DEFINE_PER_CPU(u64, residual_blocked);
+
+/* return an consistent snapshot of 64-bit time/counter value */
+static u64 get64(const u64 *p)
+{
+	u64 ret;
+
+	if (BITS_PER_LONG < 64) {
+		u32 *p32 = (u32 *)p;
+		u32 h, l;
+
+		/*
+		 * Read high then low, and then make sure high is
+		 * still the same; this will only loop if low wraps
+		 * and carries into high.
+		 * XXX some clean way to make this endian-proof?
+		 */
+		do {
+			h = p32[1];
+			barrier();
+			l = p32[0];
+			barrier();
+		} while (p32[1] != h);
+
+		ret = (((u64)h) << 32) | l;
+	} else
+		ret = *p;
+
+	return ret;
+}
+
+/*
+ * Runstate accounting
+ */
+static void get_runstate_snapshot(struct vcpu_runstate_info *res)
+{
+	u64 state_time;
+	struct vcpu_runstate_info *state;
+
+	BUG_ON(preemptible());
+
+	state = &__get_cpu_var(runstate);
+
+	/*
+	 * The runstate info is always updated by the hypervisor on
+	 * the current CPU, so there's no need to use anything
+	 * stronger than a compiler barrier when fetching it.
+	 */
+	do {
+		state_time = get64(&state->state_entry_time);
+		barrier();
+		*res = *state;
+		barrier();
+	} while (get64(&state->state_entry_time) != state_time);
+}
+
+static void setup_runstate_info(int cpu)
+{
+	struct vcpu_register_runstate_memory_area area;
+
+	area.addr.v = &per_cpu(runstate, cpu);
+
+	if (HYPERVISOR_vcpu_op(VCPUOP_register_runstate_memory_area,
+			       cpu, &area))
+		BUG();
+}
+
+static void do_stolen_accounting(void)
+{
+	struct vcpu_runstate_info state;
+	struct vcpu_runstate_info *snap;
+	s64 blocked, runnable, offline, stolen;
+	cputime_t ticks;
+
+	get_runstate_snapshot(&state);
+
+	WARN_ON(state.state != RUNSTATE_running);
+
+	snap = &__get_cpu_var(runstate_snapshot);
+
+	/* work out how much time the VCPU has not been runn*ing*  */
+	blocked = state.time[RUNSTATE_blocked] - snap->time[RUNSTATE_blocked];
+	runnable = state.time[RUNSTATE_runnable] - snap->time[RUNSTATE_runnable];
+	offline = state.time[RUNSTATE_offline] - snap->time[RUNSTATE_offline];
+
+	*snap = state;
+
+	/* Add the appropriate number of ticks of stolen time,
+	   including any left-overs from last time.  Passing NULL to
+	   account_steal_time accounts the time as stolen. */
+	stolen = runnable + offline + __get_cpu_var(residual_stolen);
+
+	if (stolen < 0)
+		stolen = 0;
+
+	ticks = 0;
+	while (stolen >= NS_PER_TICK) {
+		ticks++;
+		stolen -= NS_PER_TICK;
+	}
+	__get_cpu_var(residual_stolen) = stolen;
+	account_steal_time(NULL, ticks);
+
+	/* Add the appropriate number of ticks of blocked time,
+	   including any left-overs from last time.  Passing idle to
+	   account_steal_time accounts the time as idle/wait. */
+	blocked += __get_cpu_var(residual_blocked);
+
+	if (blocked < 0)
+		blocked = 0;
+
+	ticks = 0;
+	while (blocked >= NS_PER_TICK) {
+		ticks++;
+		blocked -= NS_PER_TICK;
+	}
+	__get_cpu_var(residual_blocked) = blocked;
+	account_steal_time(idle_task(smp_processor_id()), ticks);
+}
+
+/*
+ * Xen sched_clock implementation.  Returns the number of unstolen
+ * nanoseconds, which is nanoseconds the VCPU spent in RUNNING+BLOCKED
+ * states.
+ */
+unsigned long long xen_sched_clock(void)
+{
+	struct vcpu_runstate_info state;
+	cycle_t now;
+	u64 ret;
+	s64 offset;
+
+	/*
+	 * Ideally sched_clock should be called on a per-cpu basis
+	 * anyway, so preempt should already be disabled, but that's
+	 * not current practice at the moment.
+	 */
+	preempt_disable();
+
+	now = xen_clocksource_read();
+
+	get_runstate_snapshot(&state);
+
+	WARN_ON(state.state != RUNSTATE_running);
+
+	offset = now - state.state_entry_time;
+	if (offset < 0)
+		offset = 0;
+
+	ret = state.time[RUNSTATE_blocked] +
+		state.time[RUNSTATE_running] +
+		offset;
+
+	preempt_enable();
+
+	return ret;
+}
+
+
+/* Get the CPU speed from Xen */
+unsigned long xen_cpu_khz(void)
+{
+	u64 cpu_khz = 1000000ULL << 32;
+	const struct vcpu_time_info *info =
+		&HYPERVISOR_shared_info->vcpu_info[0].time;
+
+	do_div(cpu_khz, info->tsc_to_system_mul);
+	if (info->tsc_shift < 0)
+		cpu_khz <<= -info->tsc_shift;
+	else
+		cpu_khz >>= info->tsc_shift;
+
+	return cpu_khz;
+}
+
+/*
+ * Reads a consistent set of time-base values from Xen, into a shadow data
+ * area.
+ */
+static unsigned get_time_values_from_xen(void)
+{
+	struct vcpu_time_info   *src;
+	struct shadow_time_info *dst;
+
+	/* src is shared memory with the hypervisor, so we need to
+	   make sure we get a consistent snapshot, even in the face of
+	   being preempted. */
+	src = &__get_cpu_var(xen_vcpu)->time;
+	dst = &__get_cpu_var(shadow_time);
+
+	do {
+		dst->version = src->version;
+		rmb();		/* fetch version before data */
+		dst->tsc_timestamp     = src->tsc_timestamp;
+		dst->system_timestamp  = src->system_time;
+		dst->tsc_to_nsec_mul   = src->tsc_to_system_mul;
+		dst->tsc_shift         = src->tsc_shift;
+		rmb();		/* test version after fetching data */
+	} while ((src->version & 1) | (dst->version ^ src->version));
+
+	return dst->version;
+}
+
+/*
+ * Scale a 64-bit delta by scaling and multiplying by a 32-bit fraction,
+ * yielding a 64-bit result.
+ */
+static inline u64 scale_delta(u64 delta, u32 mul_frac, int shift)
+{
+	u64 product;
+#ifdef __i386__
+	u32 tmp1, tmp2;
+#endif
+
+	if (shift < 0)
+		delta >>= -shift;
+	else
+		delta <<= shift;
+
+#ifdef __i386__
+	__asm__ (
+		"mul  %5       ; "
+		"mov  %4,%%eax ; "
+		"mov  %%edx,%4 ; "
+		"mul  %5       ; "
+		"xor  %5,%5    ; "
+		"add  %4,%%eax ; "
+		"adc  %5,%%edx ; "
+		: "=A" (product), "=r" (tmp1), "=r" (tmp2)
+		: "a" ((u32)delta), "1" ((u32)(delta >> 32)), "2" (mul_frac) );
+#elif __x86_64__
+	__asm__ (
+		"mul %%rdx ; shrd $32,%%rdx,%%rax"
+		: "=a" (product) : "0" (delta), "d" ((u64)mul_frac) );
+#else
+#error implement me!
+#endif
+
+	return product;
+}
+
+static u64 get_nsec_offset(struct shadow_time_info *shadow)
+{
+	u64 now, delta;
+	now = native_read_tsc();
+	delta = now - shadow->tsc_timestamp;
+	return scale_delta(delta, shadow->tsc_to_nsec_mul, shadow->tsc_shift);
+}
+
+static cycle_t xen_clocksource_read(void)
+{
+	struct shadow_time_info *shadow = &get_cpu_var(shadow_time);
+	cycle_t ret;
+	unsigned version;
+
+	do {
+		version = get_time_values_from_xen();
+		barrier();
+		ret = shadow->system_timestamp + get_nsec_offset(shadow);
+		barrier();
+	} while (version != __get_cpu_var(xen_vcpu)->time.version);
+
+	put_cpu_var(shadow_time);
+
+	return ret;
+}
+
+static void xen_read_wallclock(struct timespec *ts)
+{
+	const struct shared_info *s = HYPERVISOR_shared_info;
+	u32 version;
+	u64 delta;
+	struct timespec now;
+
+	/* get wallclock at system boot */
+	do {
+		version = s->wc_version;
+		rmb();		/* fetch version before time */
+		now.tv_sec  = s->wc_sec;
+		now.tv_nsec = s->wc_nsec;
+		rmb();		/* fetch time before checking version */
+	} while ((s->wc_version & 1) | (version ^ s->wc_version));
+
+	delta = xen_clocksource_read();	/* time since system boot */
+	delta += now.tv_sec * (u64)NSEC_PER_SEC + now.tv_nsec;
+
+	now.tv_nsec = do_div(delta, NSEC_PER_SEC);
+	now.tv_sec = delta;
+
+	set_normalized_timespec(ts, now.tv_sec, now.tv_nsec);
+}
+
+unsigned long xen_get_wallclock(void)
+{
+	struct timespec ts;
+
+	xen_read_wallclock(&ts);
+
+	return ts.tv_sec;
+}
+
+int xen_set_wallclock(unsigned long now)
+{
+	/* do nothing for domU */
+	return -1;
+}
+
+static struct clocksource xen_clocksource __read_mostly = {
+	.name = "xen",
+	.rating = 400,
+	.read = xen_clocksource_read,
+	.mask = ~0,
+	.mult = 1<<XEN_SHIFT,		/* time directly in nanoseconds */
+	.shift = XEN_SHIFT,
+	.flags = CLOCK_SOURCE_IS_CONTINUOUS,
+};
+
+/*
+   Xen clockevent implementation
+
+   Xen has two clockevent implementations:
+
+   The old timer_op one works with all released versions of Xen prior
+   to version 3.0.4.  This version of the hypervisor provides a
+   single-shot timer with nanosecond resolution.  However, sharing the
+   same event channel is a 100Hz tick which is delivered while the
+   vcpu is running.  We don't care about or use this tick, but it will
+   cause the core time code to think the timer fired too soon, and
+   will end up resetting it each time.  It could be filtered, but
+   doing so has complications when the ktime clocksource is not yet
+   the xen clocksource (ie, at boot time).
+
+   The new vcpu_op-based timer interface allows the tick timer period
+   to be changed or turned off.  The tick timer is not useful as a
+   periodic timer because events are only delivered to running vcpus.
+   The one-shot timer can report when a timeout is in the past, so
+   set_next_event is capable of returning -ETIME when appropriate.
+   This interface is used when available.
+*/
+
+
+/*
+  Get a hypervisor absolute time.  In theory we could maintain an
+  offset between the kernel's time and the hypervisor's time, and
+  apply that to a kernel's absolute timeout.  Unfortunately the
+  hypervisor and kernel times can drift even if the kernel is using
+  the Xen clocksource, because ntp can warp the kernel's clocksource.
+*/
+static s64 get_abs_timeout(unsigned long delta)
+{
+	return xen_clocksource_read() + delta;
+}
+
+static void xen_timerop_set_mode(enum clock_event_mode mode,
+				 struct clock_event_device *evt)
+{
+	switch (mode) {
+	case CLOCK_EVT_MODE_PERIODIC:
+		/* unsupported */
+		WARN_ON(1);
+		break;
+
+	case CLOCK_EVT_MODE_ONESHOT:
+	case CLOCK_EVT_MODE_RESUME:
+		break;
+
+	case CLOCK_EVT_MODE_UNUSED:
+	case CLOCK_EVT_MODE_SHUTDOWN:
+		HYPERVISOR_set_timer_op(0);  /* cancel timeout */
+		break;
+	}
+}
+
+static int xen_timerop_set_next_event(unsigned long delta,
+				      struct clock_event_device *evt)
+{
+	WARN_ON(evt->mode != CLOCK_EVT_MODE_ONESHOT);
+
+	if (HYPERVISOR_set_timer_op(get_abs_timeout(delta)) < 0)
+		BUG();
+
+	/* We may have missed the deadline, but there's no real way of
+	   knowing for sure.  If the event was in the past, then we'll
+	   get an immediate interrupt. */
+
+	return 0;
+}
+
+static const struct clock_event_device xen_timerop_clockevent = {
+	.name = "xen",
+	.features = CLOCK_EVT_FEAT_ONESHOT,
+
+	.max_delta_ns = 0xffffffff,
+	.min_delta_ns = TIMER_SLOP,
+
+	.mult = 1,
+	.shift = 0,
+	.rating = 500,
+
+	.set_mode = xen_timerop_set_mode,
+	.set_next_event = xen_timerop_set_next_event,
+};
+
+
+
+static void xen_vcpuop_set_mode(enum clock_event_mode mode,
+				struct clock_event_device *evt)
+{
+	int cpu = smp_processor_id();
+
+	switch (mode) {
+	case CLOCK_EVT_MODE_PERIODIC:
+		WARN_ON(1);	/* unsupported */
+		break;
+
+	case CLOCK_EVT_MODE_ONESHOT:
+		if (HYPERVISOR_vcpu_op(VCPUOP_stop_periodic_timer, cpu, NULL))
+			BUG();
+		break;
+
+	case CLOCK_EVT_MODE_UNUSED:
+	case CLOCK_EVT_MODE_SHUTDOWN:
+		if (HYPERVISOR_vcpu_op(VCPUOP_stop_singleshot_timer, cpu, NULL) ||
+		    HYPERVISOR_vcpu_op(VCPUOP_stop_periodic_timer, cpu, NULL))
+			BUG();
+		break;
+	case CLOCK_EVT_MODE_RESUME:
+		break;
+	}
+}
+
+static int xen_vcpuop_set_next_event(unsigned long delta,
+				     struct clock_event_device *evt)
+{
+	int cpu = smp_processor_id();
+	struct vcpu_set_singleshot_timer single;
+	int ret;
+
+	WARN_ON(evt->mode != CLOCK_EVT_MODE_ONESHOT);
+
+	single.timeout_abs_ns = get_abs_timeout(delta);
+	single.flags = VCPU_SSHOTTMR_future;
+
+	ret = HYPERVISOR_vcpu_op(VCPUOP_set_singleshot_timer, cpu, &single);
+
+	BUG_ON(ret != 0 && ret != -ETIME);
+
+	return ret;
+}
+
+static const struct clock_event_device xen_vcpuop_clockevent = {
+	.name = "xen",
+	.features = CLOCK_EVT_FEAT_ONESHOT,
+
+	.max_delta_ns = 0xffffffff,
+	.min_delta_ns = TIMER_SLOP,
+
+	.mult = 1,
+	.shift = 0,
+	.rating = 500,
+
+	.set_mode = xen_vcpuop_set_mode,
+	.set_next_event = xen_vcpuop_set_next_event,
+};
+
+static const struct clock_event_device *xen_clockevent =
+	&xen_timerop_clockevent;
+static DEFINE_PER_CPU(struct clock_event_device, xen_clock_events);
+
+static irqreturn_t xen_timer_interrupt(int irq, void *dev_id)
+{
+	struct clock_event_device *evt = &__get_cpu_var(xen_clock_events);
+	irqreturn_t ret;
+
+	ret = IRQ_NONE;
+	if (evt->event_handler) {
+		evt->event_handler(evt);
+		ret = IRQ_HANDLED;
+	}
+
+	do_stolen_accounting();
+
+	return ret;
+}
+
+void xen_setup_timer(int cpu)
+{
+	const char *name;
+	struct clock_event_device *evt;
+	int irq;
+
+	printk(KERN_INFO "installing Xen timer for CPU %d\n", cpu);
+
+	name = kasprintf(GFP_KERNEL, "timer%d", cpu);
+	if (!name)
+		name = "<timer kasprintf failed>";
+
+	irq = bind_virq_to_irqhandler(VIRQ_TIMER, cpu, xen_timer_interrupt,
+				      IRQF_DISABLED|IRQF_PERCPU|IRQF_NOBALANCING,
+				      name, NULL);
+
+	evt = &per_cpu(xen_clock_events, cpu);
+	memcpy(evt, xen_clockevent, sizeof(*evt));
+
+	evt->cpumask = cpumask_of_cpu(cpu);
+	evt->irq = irq;
+
+	setup_runstate_info(cpu);
+}
+
+void xen_setup_cpu_clockevents(void)
+{
+	BUG_ON(preemptible());
+
+	clockevents_register_device(&__get_cpu_var(xen_clock_events));
+}
+
+__init void xen_time_init(void)
+{
+	int cpu = smp_processor_id();
+
+	get_time_values_from_xen();
+
+	clocksource_register(&xen_clocksource);
+
+	if (HYPERVISOR_vcpu_op(VCPUOP_stop_periodic_timer, cpu, NULL) == 0) {
+		/* Successfully turned off 100Hz tick, so we have the
+		   vcpuop-based timer interface */
+		printk(KERN_DEBUG "Xen: using vcpuop timer interface\n");
+		xen_clockevent = &xen_vcpuop_clockevent;
+	}
+
+	/* Set initial system time with full resolution */
+	xen_read_wallclock(&xtime);
+	set_normalized_timespec(&wall_to_monotonic,
+				-xtime.tv_sec, -xtime.tv_nsec);
+
+	tsc_disable = 0;
+
+	xen_setup_timer(cpu);
+	xen_setup_cpu_clockevents();
+}
diff --git a/arch/x86/xen/vdso.h b/arch/x86/xen/vdso.h
new file mode 100644
index 00000000000..861fedfe523
--- /dev/null
+++ b/arch/x86/xen/vdso.h
@@ -0,0 +1,4 @@
+/* Bit used for the pseudo-hwcap for non-negative segments.  We use
+   bit 1 to avoid bugs in some versions of glibc when bit 0 is
+   used; the choice is otherwise arbitrary. */
+#define VDSO_NOTE_NONEGSEG_BIT	1
diff --git a/arch/x86/xen/xen-asm.S b/arch/x86/xen/xen-asm.S
new file mode 100644
index 00000000000..1a43b60c0c6
--- /dev/null
+++ b/arch/x86/xen/xen-asm.S
@@ -0,0 +1,291 @@
+/*
+	Asm versions of Xen pv-ops, suitable for either direct use or inlining.
+	The inline versions are the same as the direct-use versions, with the
+	pre- and post-amble chopped off.
+
+	This code is encoded for size rather than absolute efficiency,
+	with a view to being able to inline as much as possible.
+
+	We only bother with direct forms (ie, vcpu in pda) of the operations
+	here; the indirect forms are better handled in C, since they're
+	generally too large to inline anyway.
+ */
+
+#include <linux/linkage.h>
+
+#include <asm/asm-offsets.h>
+#include <asm/thread_info.h>
+#include <asm/percpu.h>
+#include <asm/processor-flags.h>
+#include <asm/segment.h>
+
+#include <xen/interface/xen.h>
+
+#define RELOC(x, v)	.globl x##_reloc; x##_reloc=v
+#define ENDPATCH(x)	.globl x##_end; x##_end=.
+
+/* Pseudo-flag used for virtual NMI, which we don't implement yet */
+#define XEN_EFLAGS_NMI	0x80000000
+
+/*
+	Enable events.  This clears the event mask and tests the pending
+	event status with one and operation.  If there are pending
+	events, then enter the hypervisor to get them handled.
+ */
+ENTRY(xen_irq_enable_direct)
+	/* Clear mask and test pending */
+	andw $0x00ff, PER_CPU_VAR(xen_vcpu_info)+XEN_vcpu_info_pending
+	/* Preempt here doesn't matter because that will deal with
+	   any pending interrupts.  The pending check may end up being
+	   run on the wrong CPU, but that doesn't hurt. */
+	jz 1f
+2:	call check_events
+1:
+ENDPATCH(xen_irq_enable_direct)
+	ret
+	ENDPROC(xen_irq_enable_direct)
+	RELOC(xen_irq_enable_direct, 2b+1)
+
+
+/*
+	Disabling events is simply a matter of making the event mask
+	non-zero.
+ */
+ENTRY(xen_irq_disable_direct)
+	movb $1, PER_CPU_VAR(xen_vcpu_info)+XEN_vcpu_info_mask
+ENDPATCH(xen_irq_disable_direct)
+	ret
+	ENDPROC(xen_irq_disable_direct)
+	RELOC(xen_irq_disable_direct, 0)
+
+/*
+	(xen_)save_fl is used to get the current interrupt enable status.
+	Callers expect the status to be in X86_EFLAGS_IF, and other bits
+	may be set in the return value.  We take advantage of this by
+	making sure that X86_EFLAGS_IF has the right value (and other bits
+	in that byte are 0), but other bits in the return value are
+	undefined.  We need to toggle the state of the bit, because
+	Xen and x86 use opposite senses (mask vs enable).
+ */
+ENTRY(xen_save_fl_direct)
+	testb $0xff, PER_CPU_VAR(xen_vcpu_info)+XEN_vcpu_info_mask
+	setz %ah
+	addb %ah,%ah
+ENDPATCH(xen_save_fl_direct)
+	ret
+	ENDPROC(xen_save_fl_direct)
+	RELOC(xen_save_fl_direct, 0)
+
+
+/*
+	In principle the caller should be passing us a value return
+	from xen_save_fl_direct, but for robustness sake we test only
+	the X86_EFLAGS_IF flag rather than the whole byte. After
+	setting the interrupt mask state, it checks for unmasked
+	pending events and enters the hypervisor to get them delivered
+	if so.
+ */
+ENTRY(xen_restore_fl_direct)
+	testb $X86_EFLAGS_IF>>8, %ah
+	setz PER_CPU_VAR(xen_vcpu_info)+XEN_vcpu_info_mask
+	/* Preempt here doesn't matter because that will deal with
+	   any pending interrupts.  The pending check may end up being
+	   run on the wrong CPU, but that doesn't hurt. */
+
+	/* check for unmasked and pending */
+	cmpw $0x0001, PER_CPU_VAR(xen_vcpu_info)+XEN_vcpu_info_pending
+	jz 1f
+2:	call check_events
+1:
+ENDPATCH(xen_restore_fl_direct)
+	ret
+	ENDPROC(xen_restore_fl_direct)
+	RELOC(xen_restore_fl_direct, 2b+1)
+
+/*
+	This is run where a normal iret would be run, with the same stack setup:
+	      8: eflags
+	      4: cs
+	esp-> 0: eip
+
+	This attempts to make sure that any pending events are dealt
+	with on return to usermode, but there is a small window in
+	which an event can happen just before entering usermode.  If
+	the nested interrupt ends up setting one of the TIF_WORK_MASK
+	pending work flags, they will not be tested again before
+	returning to usermode. This means that a process can end up
+	with pending work, which will be unprocessed until the process
+	enters and leaves the kernel again, which could be an
+	unbounded amount of time.  This means that a pending signal or
+	reschedule event could be indefinitely delayed.
+
+	The fix is to notice a nested interrupt in the critical
+	window, and if one occurs, then fold the nested interrupt into
+	the current interrupt stack frame, and re-process it
+	iteratively rather than recursively.  This means that it will
+	exit via the normal path, and all pending work will be dealt
+	with appropriately.
+
+	Because the nested interrupt handler needs to deal with the
+	current stack state in whatever form its in, we keep things
+	simple by only using a single register which is pushed/popped
+	on the stack.
+
+	Non-direct iret could be done in the same way, but it would
+	require an annoying amount of code duplication.  We'll assume
+	that direct mode will be the common case once the hypervisor
+	support becomes commonplace.
+ */
+ENTRY(xen_iret_direct)
+	/* test eflags for special cases */
+	testl $(X86_EFLAGS_VM | XEN_EFLAGS_NMI), 8(%esp)
+	jnz hyper_iret
+
+	push %eax
+	ESP_OFFSET=4	# bytes pushed onto stack
+
+	/* Store vcpu_info pointer for easy access.  Do it this
+	   way to avoid having to reload %fs */
+#ifdef CONFIG_SMP
+	GET_THREAD_INFO(%eax)
+	movl TI_cpu(%eax),%eax
+	movl __per_cpu_offset(,%eax,4),%eax
+	lea per_cpu__xen_vcpu_info(%eax),%eax
+#else
+	movl $per_cpu__xen_vcpu_info, %eax
+#endif
+
+	/* check IF state we're restoring */
+	testb $X86_EFLAGS_IF>>8, 8+1+ESP_OFFSET(%esp)
+
+	/* Maybe enable events.  Once this happens we could get a
+	   recursive event, so the critical region starts immediately
+	   afterwards.  However, if that happens we don't end up
+	   resuming the code, so we don't have to be worried about
+	   being preempted to another CPU. */
+	setz XEN_vcpu_info_mask(%eax)
+xen_iret_start_crit:
+
+	/* check for unmasked and pending */
+	cmpw $0x0001, XEN_vcpu_info_pending(%eax)
+
+	/* If there's something pending, mask events again so we
+	   can jump back into xen_hypervisor_callback */
+	sete XEN_vcpu_info_mask(%eax)
+
+	popl %eax
+
+	/* From this point on the registers are restored and the stack
+	   updated, so we don't need to worry about it if we're preempted */
+iret_restore_end:
+
+	/* Jump to hypervisor_callback after fixing up the stack.
+	   Events are masked, so jumping out of the critical
+	   region is OK. */
+	je xen_hypervisor_callback
+
+	iret
+xen_iret_end_crit:
+
+hyper_iret:
+	/* put this out of line since its very rarely used */
+	jmp hypercall_page + __HYPERVISOR_iret * 32
+
+	.globl xen_iret_start_crit, xen_iret_end_crit
+
+/*
+   This is called by xen_hypervisor_callback in entry.S when it sees
+   that the EIP at the time of interrupt was between xen_iret_start_crit
+   and xen_iret_end_crit.  We're passed the EIP in %eax so we can do
+   a more refined determination of what to do.
+
+   The stack format at this point is:
+	----------------
+	 ss		: (ss/esp may be present if we came from usermode)
+	 esp		:
+	 eflags		}  outer exception info
+	 cs		}
+	 eip		}
+	---------------- <- edi (copy dest)
+	 eax		:  outer eax if it hasn't been restored
+	----------------
+	 eflags		}  nested exception info
+	 cs		}   (no ss/esp because we're nested
+	 eip		}    from the same ring)
+	 orig_eax	}<- esi (copy src)
+	 - - - - - - - -
+	 fs		}
+	 es		}
+	 ds		}  SAVE_ALL state
+	 eax		}
+	  :		:
+	 ebx		}
+	----------------
+	 return addr	 <- esp
+	----------------
+
+   In order to deliver the nested exception properly, we need to shift
+   everything from the return addr up to the error code so it
+   sits just under the outer exception info.  This means that when we
+   handle the exception, we do it in the context of the outer exception
+   rather than starting a new one.
+
+   The only caveat is that if the outer eax hasn't been
+   restored yet (ie, it's still on stack), we need to insert
+   its value into the SAVE_ALL state before going on, since
+   it's usermode state which we eventually need to restore.
+ */
+ENTRY(xen_iret_crit_fixup)
+	/* offsets +4 for return address */
+
+	/*
+	   Paranoia: Make sure we're really coming from userspace.
+	   One could imagine a case where userspace jumps into the
+	   critical range address, but just before the CPU delivers a GP,
+	   it decides to deliver an interrupt instead.  Unlikely?
+	   Definitely.  Easy to avoid?  Yes.  The Intel documents
+	   explicitly say that the reported EIP for a bad jump is the
+	   jump instruction itself, not the destination, but some virtual
+	   environments get this wrong.
+	 */
+	movl PT_CS+4(%esp), %ecx
+	andl $SEGMENT_RPL_MASK, %ecx
+	cmpl $USER_RPL, %ecx
+	je 2f
+
+	lea PT_ORIG_EAX+4(%esp), %esi
+	lea PT_EFLAGS+4(%esp), %edi
+
+	/* If eip is before iret_restore_end then stack
+	   hasn't been restored yet. */
+	cmp $iret_restore_end, %eax
+	jae 1f
+
+	movl 0+4(%edi),%eax		/* copy EAX */
+	movl %eax, PT_EAX+4(%esp)
+
+	lea ESP_OFFSET(%edi),%edi	/* move dest up over saved regs */
+
+	/* set up the copy */
+1:	std
+	mov $(PT_EIP+4) / 4, %ecx	/* copy ret+saved regs up to orig_eax */
+	rep movsl
+	cld
+
+	lea 4(%edi),%esp		/* point esp to new frame */
+2:	ret
+
+
+/*
+	Force an event check by making a hypercall,
+	but preserve regs before making the call.
+ */
+check_events:
+	push %eax
+	push %ecx
+	push %edx
+	call force_evtchn_callback
+	pop %edx
+	pop %ecx
+	pop %eax
+	ret
diff --git a/arch/x86/xen/xen-head.S b/arch/x86/xen/xen-head.S
new file mode 100644
index 00000000000..f8d6937db2e
--- /dev/null
+++ b/arch/x86/xen/xen-head.S
@@ -0,0 +1,38 @@
+/* Xen-specific pieces of head.S, intended to be included in the right
+	place in head.S */
+
+#ifdef CONFIG_XEN
+
+#include <linux/elfnote.h>
+#include <asm/boot.h>
+#include <xen/interface/elfnote.h>
+
+.pushsection .init.text
+ENTRY(startup_xen)
+	movl %esi,xen_start_info
+	cld
+	movl $(init_thread_union+THREAD_SIZE),%esp
+	jmp xen_start_kernel
+.popsection
+
+.pushsection .bss.page_aligned
+	.align PAGE_SIZE_asm
+ENTRY(hypercall_page)
+	.skip 0x1000
+.popsection
+
+	ELFNOTE(Xen, XEN_ELFNOTE_GUEST_OS,       .asciz "linux")
+	ELFNOTE(Xen, XEN_ELFNOTE_GUEST_VERSION,  .asciz "2.6")
+	ELFNOTE(Xen, XEN_ELFNOTE_XEN_VERSION,    .asciz "xen-3.0")
+	ELFNOTE(Xen, XEN_ELFNOTE_VIRT_BASE,      .long  __PAGE_OFFSET)
+	ELFNOTE(Xen, XEN_ELFNOTE_ENTRY,          .long  startup_xen)
+	ELFNOTE(Xen, XEN_ELFNOTE_HYPERCALL_PAGE, .long  hypercall_page)
+	ELFNOTE(Xen, XEN_ELFNOTE_FEATURES,       .asciz "!writable_page_tables|pae_pgdir_above_4gb")
+#ifdef CONFIG_X86_PAE
+	ELFNOTE(Xen, XEN_ELFNOTE_PAE_MODE,       .asciz "yes")
+#else
+	ELFNOTE(Xen, XEN_ELFNOTE_PAE_MODE,       .asciz "no")
+#endif
+	ELFNOTE(Xen, XEN_ELFNOTE_LOADER,         .asciz "generic")
+
+#endif /*CONFIG_XEN */
diff --git a/arch/x86/xen/xen-ops.h b/arch/x86/xen/xen-ops.h
new file mode 100644
index 00000000000..b9aaea45f07
--- /dev/null
+++ b/arch/x86/xen/xen-ops.h
@@ -0,0 +1,71 @@
+#ifndef XEN_OPS_H
+#define XEN_OPS_H
+
+#include <linux/init.h>
+
+/* These are code, but not functions.  Defined in entry.S */
+extern const char xen_hypervisor_callback[];
+extern const char xen_failsafe_callback[];
+
+void xen_copy_trap_info(struct trap_info *traps);
+
+DECLARE_PER_CPU(struct vcpu_info *, xen_vcpu);
+DECLARE_PER_CPU(unsigned long, xen_cr3);
+
+extern struct start_info *xen_start_info;
+extern struct shared_info *HYPERVISOR_shared_info;
+
+char * __init xen_memory_setup(void);
+void __init xen_arch_setup(void);
+void __init xen_init_IRQ(void);
+
+void xen_setup_timer(int cpu);
+void xen_setup_cpu_clockevents(void);
+unsigned long xen_cpu_khz(void);
+void __init xen_time_init(void);
+unsigned long xen_get_wallclock(void);
+int xen_set_wallclock(unsigned long time);
+unsigned long long xen_sched_clock(void);
+
+void xen_mark_init_mm_pinned(void);
+
+DECLARE_PER_CPU(enum paravirt_lazy_mode, xen_lazy_mode);
+
+static inline unsigned xen_get_lazy_mode(void)
+{
+	return x86_read_percpu(xen_lazy_mode);
+}
+
+void __init xen_fill_possible_map(void);
+
+void __init xen_setup_vcpu_info_placement(void);
+void xen_smp_prepare_boot_cpu(void);
+void xen_smp_prepare_cpus(unsigned int max_cpus);
+int xen_cpu_up(unsigned int cpu);
+void xen_smp_cpus_done(unsigned int max_cpus);
+
+void xen_smp_send_stop(void);
+void xen_smp_send_reschedule(int cpu);
+int xen_smp_call_function (void (*func) (void *info), void *info, int nonatomic,
+			   int wait);
+int xen_smp_call_function_single(int cpu, void (*func) (void *info), void *info,
+				 int nonatomic, int wait);
+
+int xen_smp_call_function_mask(cpumask_t mask, void (*func)(void *),
+			       void *info, int wait);
+
+
+/* Declare an asm function, along with symbols needed to make it
+   inlineable */
+#define DECL_ASM(ret, name, ...)		\
+	ret name(__VA_ARGS__);			\
+	extern char name##_end[];		\
+	extern char name##_reloc[]		\
+
+DECL_ASM(void, xen_irq_enable_direct, void);
+DECL_ASM(void, xen_irq_disable_direct, void);
+DECL_ASM(unsigned long, xen_save_fl_direct, void);
+DECL_ASM(void, xen_restore_fl_direct, unsigned long);
+
+void xen_iret_direct(void);
+#endif /* XEN_OPS_H */