89 files changed, 2944 insertions, 1426 deletions
diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig
index 96e0c2ebc388..03980cb04291 100644
--- a/arch/x86/Kconfig
+++ b/arch/x86/Kconfig
@@ -447,7 +447,6 @@ config PARAVIRT_DEBUG
 config MEMTEST
        bool "Memtest"
-        depends on X86_64
        help
          This option adds a kernel parameter 'memtest', which allows memtest
          to be set.
diff --git a/arch/x86/Kconfig.cpu b/arch/x86/Kconfig.cpu
index abff1b84ed5b..54b8c02c71e6 100644
--- a/arch/x86/Kconfig.cpu
+++ b/arch/x86/Kconfig.cpu
@@ -362,10 +362,6 @@ config X86_ALIGNMENT_16
        def_bool y
        depends on MWINCHIP3D || MWINCHIP2 || MWINCHIPC6 || MCYRIXIII || X86_ELAN || MK6 || M586MMX || M586TSC || M586 || M486 || MVIAC3_2 || MGEODEGX1
-config X86_GOOD_APIC
-        def_bool y
-        depends on MK7 || MPENTIUM4 || MPENTIUMM || MPENTIUMIII || MPENTIUMII || M686 || M586MMX || MK8 || MEFFICEON || MCORE2 || MVIAC7 || X86_64
 config X86_INTEL_USERCOPY
        def_bool y
        depends on MPENTIUM4 || MPENTIUMM || MPENTIUMIII || MPENTIUMII || M586MMX || X86_GENERIC || MK8 || MK7 || MEFFICEON || MCORE2
diff --git a/arch/x86/Kconfig.debug b/arch/x86/Kconfig.debug
index 51c821477951..85a87d2ac0c0 100644
--- a/arch/x86/Kconfig.debug
+++ b/arch/x86/Kconfig.debug
@@ -289,7 +289,6 @@ config CPA_DEBUG
 config OPTIMIZE_INLINING
        bool "Allow gcc to uninline functions marked 'inline'"
-        depends on BROKEN
        help
          This option determines if the kernel forces gcc to inline the functions
          developers have marked 'inline'. Doing so takes away freedom from gcc to
@@ -300,5 +299,7 @@ config OPTIMIZE_INLINING
          become the default in the future, until then this option is there to
          test gcc for this.
+          If unsure, say N.
 endmenu
diff --git a/arch/x86/boot/edd.c b/arch/x86/boot/edd.c
index 03399d64013b..d93cbc6464d0 100644
--- a/arch/x86/boot/edd.c
+++ b/arch/x86/boot/edd.c
@@ -167,9 +167,8 @@ void query_edd(void)
                 * Scan the BIOS-supported hard disks and query EDD
                 * information...
                 */
-                get_edd_info(devno, &ei);
+                if (!get_edd_info(devno, &ei)
+                    && boot_params.eddbuf_entries < EDDMAXNR) {
-                if (boot_params.eddbuf_entries < EDDMAXNR) {
                        memcpy(edp, &ei, sizeof ei);
                        edp++;
                        boot_params.eddbuf_entries++;
diff --git a/arch/x86/boot/pm.c b/arch/x86/boot/pm.c
index 328956fdb59e..85a1cd8a8ff8 100644
--- a/arch/x86/boot/pm.c
+++ b/arch/x86/boot/pm.c
@@ -98,12 +98,6 @@ static void reset_coprocessor(void)
 /*
 * Set up the GDT
 */
-#define GDT_ENTRY(flags, base, limit)           \
-        (((u64)(base & 0xff000000) << 32) |     \
-         ((u64)flags << 40) |                   \
-         ((u64)(limit & 0x00ff0000) << 32) |    \
-         ((u64)(base & 0x00ffffff) << 16) |     \
-         ((u64)(limit & 0x0000ffff)))
 struct gdt_ptr {
        u16 len;
diff --git a/arch/x86/ia32/ia32_signal.c b/arch/x86/ia32/ia32_signal.c
index cb3856a18c85..20af4c79579a 100644
--- a/arch/x86/ia32/ia32_signal.c
+++ b/arch/x86/ia32/ia32_signal.c
@@ -36,6 +36,11 @@
 #define _BLOCKABLE (~(sigmask(SIGKILL) | sigmask(SIGSTOP)))
+#define FIX_EFLAGS      (X86_EFLAGS_AC | X86_EFLAGS_OF | \
+                         X86_EFLAGS_DF | X86_EFLAGS_TF | X86_EFLAGS_SF | \
+                         X86_EFLAGS_ZF | X86_EFLAGS_AF | X86_EFLAGS_PF | \
+                         X86_EFLAGS_CF)
 asmlinkage int do_signal(struct pt_regs *regs, sigset_t *oldset);
 void signal_fault(struct pt_regs *regs, void __user *frame, char *where);
@@ -248,7 +253,7 @@ static int ia32_restore_sigcontext(struct pt_regs *regs,
        regs->ss |= 3;
        err |= __get_user(tmpflags, &sc->flags);
-        regs->flags = (regs->flags & ~0x40DD5) | (tmpflags & 0x40DD5);
+        regs->flags = (regs->flags & ~FIX_EFLAGS) | (tmpflags & FIX_EFLAGS);
        /* disable syscall checks */
        regs->orig_ax = -1;
@@ -515,7 +520,6 @@ int ia32_setup_rt_frame(int sig, struct k_sigaction *ka, siginfo_t *info,
                        compat_sigset_t *set, struct pt_regs *regs)
 {
        struct rt_sigframe __user *frame;
-        struct exec_domain *ed = current_thread_info()->exec_domain;
        void __user *restorer;
        int err = 0;
@@ -538,8 +542,7 @@ int ia32_setup_rt_frame(int sig, struct k_sigaction *ka, siginfo_t *info,
        if (!access_ok(VERIFY_WRITE, frame, sizeof(*frame)))
                goto give_sigsegv;
-        err |= __put_user((ed && ed->signal_invmap && sig < 32
+        err |= __put_user(sig, &frame->sig);
-                           ? ed->signal_invmap[sig] : sig), &frame->sig);
        err |= __put_user(ptr_to_compat(&frame->info), &frame->pinfo);
        err |= __put_user(ptr_to_compat(&frame->uc), &frame->puc);
        err |= copy_siginfo_to_user32(&frame->info, info);
diff --git a/arch/x86/ia32/ia32entry.S b/arch/x86/ia32/ia32entry.S
index 20371d0635e4..23d146ce676b 100644
--- a/arch/x86/ia32/ia32entry.S
+++ b/arch/x86/ia32/ia32entry.S
@@ -37,6 +37,11 @@
        movq    %rax,R8(%rsp)
        .endm
+        /*
+         * Reload arg registers from stack in case ptrace changed them.
+         * We don't reload %eax because syscall_trace_enter() returned
+         * the value it wants us to use in the table lookup.
+         */
        .macro LOAD_ARGS32 offset
        movl \offset(%rsp),%r11d
        movl \offset+8(%rsp),%r10d
@@ -46,7 +51,6 @@
        movl \offset+48(%rsp),%edx
        movl \offset+56(%rsp),%esi
        movl \offset+64(%rsp),%edi
-        movl \offset+72(%rsp),%eax
        .endm
        
        .macro CFI_STARTPROC32 simple
@@ -137,13 +141,12 @@ ENTRY(ia32_sysenter_target)
        .previous       
        GET_THREAD_INFO(%r10)
        orl    $TS_COMPAT,TI_status(%r10)
-        testl  $(_TIF_SYSCALL_TRACE|_TIF_SYSCALL_AUDIT|_TIF_SECCOMP), \
+        testl  $_TIF_WORK_SYSCALL_ENTRY,TI_flags(%r10)
-                 TI_flags(%r10)
        CFI_REMEMBER_STATE
        jnz  sysenter_tracesys
-sysenter_do_call:       
        cmpl    $(IA32_NR_syscalls-1),%eax
        ja      ia32_badsys
+sysenter_do_call:
        IA32_ARG_FIXUP 1
        call    *ia32_sys_call_table(,%rax,8)
        movq    %rax,RAX-ARGOFFSET(%rsp)
@@ -242,8 +245,7 @@ ENTRY(ia32_cstar_target)
        .previous       
        GET_THREAD_INFO(%r10)
        orl   $TS_COMPAT,TI_status(%r10)
-        testl $(_TIF_SYSCALL_TRACE|_TIF_SYSCALL_AUDIT|_TIF_SECCOMP), \
+        testl $_TIF_WORK_SYSCALL_ENTRY,TI_flags(%r10)
-                TI_flags(%r10)
        CFI_REMEMBER_STATE
        jnz   cstar_tracesys
 cstar_do_call:  
@@ -321,6 +323,7 @@ ENTRY(ia32_syscall)
        /*CFI_REL_OFFSET        rflags,EFLAGS-RIP*/
        /*CFI_REL_OFFSET        cs,CS-RIP*/
        CFI_REL_OFFSET  rip,RIP-RIP
+        PARAVIRT_ADJUST_EXCEPTION_FRAME
        SWAPGS
        /*
         * No need to follow this irqs on/off section: the syscall
@@ -336,8 +339,7 @@ ENTRY(ia32_syscall)
        SAVE_ARGS 0,0,1
        GET_THREAD_INFO(%r10)
        orl   $TS_COMPAT,TI_status(%r10)
-        testl $(_TIF_SYSCALL_TRACE|_TIF_SYSCALL_AUDIT|_TIF_SECCOMP), \
+        testl $_TIF_WORK_SYSCALL_ENTRY,TI_flags(%r10)
-                TI_flags(%r10)
        jnz ia32_tracesys
 ia32_do_syscall:        
        cmpl $(IA32_NR_syscalls-1),%eax
diff --git a/arch/x86/kernel/Makefile b/arch/x86/kernel/Makefile
index da140611bb57..b78a17b12810 100644
--- a/arch/x86/kernel/Makefile
+++ b/arch/x86/kernel/Makefile
@@ -102,6 +102,7 @@ obj-$(CONFIG_OLPC)		+= olpc.o
 # 64 bit specific files
 ifeq ($(CONFIG_X86_64),y)
        obj-y                           += genapic_64.o genapic_flat_64.o genx2apic_uv_x.o tlb_uv.o
+        obj-y                           += bios_uv.o
        obj-$(CONFIG_X86_PM_TIMER)      += pmtimer_64.o
        obj-$(CONFIG_AUDIT)             += audit_64.o
diff --git a/arch/x86/kernel/acpi/sleep.c b/arch/x86/kernel/acpi/sleep.c
index 868de3d5c39d..a3ddad18aaa3 100644
--- a/arch/x86/kernel/acpi/sleep.c
+++ b/arch/x86/kernel/acpi/sleep.c
@@ -9,6 +9,7 @@
 #include <linux/bootmem.h>
 #include <linux/dmi.h>
 #include <linux/cpumask.h>
+#include <asm/segment.h>
 #include "realmode/wakeup.h"
 #include "sleep.h"
@@ -23,15 +24,6 @@ static unsigned long acpi_realmode;
 static char temp_stack[10240];
 #endif
-/* XXX: this macro should move to asm-x86/segment.h and be shared with the
-   boot code... */
-#define GDT_ENTRY(flags, base, limit)           \
-        (((u64)(base & 0xff000000) << 32) |     \
-         ((u64)flags << 40) |                   \
-         ((u64)(limit & 0x00ff0000) << 32) |    \
-         ((u64)(base & 0x00ffffff) << 16) |     \
-         ((u64)(limit & 0x0000ffff)))
 /**
 * acpi_save_state_mem - save kernel state
 *
diff --git a/arch/x86/kernel/amd_iommu.c b/arch/x86/kernel/amd_iommu.c
index f2766d84c7a0..c25210e6ac88 100644
--- a/arch/x86/kernel/amd_iommu.c
+++ b/arch/x86/kernel/amd_iommu.c
@@ -23,7 +23,7 @@
 #include <linux/scatterlist.h>
 #include <linux/iommu-helper.h>
 #include <asm/proto.h>
-#include <asm/gart.h>
+#include <asm/iommu.h>
 #include <asm/amd_iommu_types.h>
 #include <asm/amd_iommu.h>
@@ -32,21 +32,37 @@
 #define to_pages(addr, size) \
         (round_up(((addr) & ~PAGE_MASK) + (size), PAGE_SIZE) >> PAGE_SHIFT)
+#define EXIT_LOOP_COUNT 10000000
 static DEFINE_RWLOCK(amd_iommu_devtable_lock);
-struct command {
+/*
+ * general struct to manage commands send to an IOMMU
+ */
+struct iommu_cmd {
        u32 data[4];
 };
 static int dma_ops_unity_map(struct dma_ops_domain *dma_dom,
                             struct unity_map_entry *e);
+/* returns !0 if the IOMMU is caching non-present entries in its TLB */
 static int iommu_has_npcache(struct amd_iommu *iommu)
 {
        return iommu->cap & IOMMU_CAP_NPCACHE;
 }
-static int __iommu_queue_command(struct amd_iommu *iommu, struct command *cmd)
+/****************************************************************************
+ *
+ * IOMMU command queuing functions
+ *
+ ****************************************************************************/
+/*
+ * Writes the command to the IOMMUs command buffer and informs the
+ * hardware about the new command. Must be called with iommu->lock held.
+ */
+static int __iommu_queue_command(struct amd_iommu *iommu, struct iommu_cmd *cmd)
 {
        u32 tail, head;
        u8 *target;
@@ -63,7 +79,11 @@ static int __iommu_queue_command(struct amd_iommu *iommu, struct command *cmd)
        return 0;
 }
-static int iommu_queue_command(struct amd_iommu *iommu, struct command *cmd)
+/*
+ * General queuing function for commands. Takes iommu->lock and calls
+ * __iommu_queue_command().
+ */
+static int iommu_queue_command(struct amd_iommu *iommu, struct iommu_cmd *cmd)
 {
        unsigned long flags;
        int ret;
@@ -75,16 +95,24 @@ static int iommu_queue_command(struct amd_iommu *iommu, struct command *cmd)
        return ret;
 }
+/*
+ * This function is called whenever we need to ensure that the IOMMU has
+ * completed execution of all commands we sent. It sends a
+ * COMPLETION_WAIT command and waits for it to finish. The IOMMU informs
+ * us about that by writing a value to a physical address we pass with
+ * the command.
+ */
 static int iommu_completion_wait(struct amd_iommu *iommu)
 {
        int ret;
-        struct command cmd;
+        struct iommu_cmd cmd;
        volatile u64 ready = 0;
        unsigned long ready_phys = virt_to_phys(&ready);
+        unsigned long i = 0;
        memset(&cmd, 0, sizeof(cmd));
        cmd.data[0] = LOW_U32(ready_phys) | CMD_COMPL_WAIT_STORE_MASK;
-        cmd.data[1] = HIGH_U32(ready_phys);
+        cmd.data[1] = upper_32_bits(ready_phys);
        cmd.data[2] = 1; /* value written to 'ready' */
        CMD_SET_TYPE(&cmd, CMD_COMPL_WAIT);
@@ -95,15 +123,23 @@ static int iommu_completion_wait(struct amd_iommu *iommu)
        if (ret)
                return ret;
-        while (!ready)
+        while (!ready && (i < EXIT_LOOP_COUNT)) {
+                ++i;
                cpu_relax();
+        }
+        if (unlikely((i == EXIT_LOOP_COUNT) && printk_ratelimit()))
+                printk(KERN_WARNING "AMD IOMMU: Completion wait loop failed\n");
        return 0;
 }
+/*
+ * Command send function for invalidating a device table entry
+ */
 static int iommu_queue_inv_dev_entry(struct amd_iommu *iommu, u16 devid)
 {
-        struct command cmd;
+        struct iommu_cmd cmd;
        BUG_ON(iommu == NULL);
@@ -116,20 +152,23 @@ static int iommu_queue_inv_dev_entry(struct amd_iommu *iommu, u16 devid)
        return iommu_queue_command(iommu, &cmd);
 }
+/*
+ * Generic command send function for invalidaing TLB entries
+ */
 static int iommu_queue_inv_iommu_pages(struct amd_iommu *iommu,
                u64 address, u16 domid, int pde, int s)
 {
-        struct command cmd;
+        struct iommu_cmd cmd;
        memset(&cmd, 0, sizeof(cmd));
        address &= PAGE_MASK;
        CMD_SET_TYPE(&cmd, CMD_INV_IOMMU_PAGES);
        cmd.data[1] |= domid;
        cmd.data[2] = LOW_U32(address);
-        cmd.data[3] = HIGH_U32(address);
+        cmd.data[3] = upper_32_bits(address);
-        if (s)
+        if (s) /* size bit - we flush more than one 4kb page */
                cmd.data[2] |= CMD_INV_IOMMU_PAGES_SIZE_MASK;
-        if (pde)
+        if (pde) /* PDE bit - we wan't flush everything not only the PTEs */
                cmd.data[2] |= CMD_INV_IOMMU_PAGES_PDE_MASK;
        iommu->need_sync = 1;
@@ -137,6 +176,11 @@ static int iommu_queue_inv_iommu_pages(struct amd_iommu *iommu,
        return iommu_queue_command(iommu, &cmd);
 }
+/*
+ * TLB invalidation function which is called from the mapping functions.
+ * It invalidates a single PTE if the range to flush is within a single
+ * page. Otherwise it flushes the whole TLB of the IOMMU.
+ */
 static int iommu_flush_pages(struct amd_iommu *iommu, u16 domid,
                u64 address, size_t size)
 {
@@ -159,6 +203,20 @@ static int iommu_flush_pages(struct amd_iommu *iommu, u16 domid,
        return 0;
 }
+/****************************************************************************
+ *
+ * The functions below are used the create the page table mappings for
+ * unity mapped regions.
+ *
+ ****************************************************************************/
+/*
+ * Generic mapping functions. It maps a physical address into a DMA
+ * address space. It allocates the page table pages if necessary.
+ * In the future it can be extended to a generic mapping function
+ * supporting all features of AMD IOMMU page tables like level skipping
+ * and full 64 bit address spaces.
+ */
 static int iommu_map(struct protection_domain *dom,
                     unsigned long bus_addr,
                     unsigned long phys_addr,
@@ -209,6 +267,10 @@ static int iommu_map(struct protection_domain *dom,
        return 0;
 }
+/*
+ * This function checks if a specific unity mapping entry is needed for
+ * this specific IOMMU.
+ */
 static int iommu_for_unity_map(struct amd_iommu *iommu,
                               struct unity_map_entry *entry)
 {
@@ -223,6 +285,12 @@ static int iommu_for_unity_map(struct amd_iommu *iommu,
        return 0;
 }
+/*
+ * Init the unity mappings for a specific IOMMU in the system
+ *
+ * Basically iterates over all unity mapping entries and applies them to
+ * the default domain DMA of that IOMMU if necessary.
+ */
 static int iommu_init_unity_mappings(struct amd_iommu *iommu)
 {
        struct unity_map_entry *entry;
@@ -239,6 +307,10 @@ static int iommu_init_unity_mappings(struct amd_iommu *iommu)
        return 0;
 }
+/*
+ * This function actually applies the mapping to the page table of the
+ * dma_ops domain.
+ */
 static int dma_ops_unity_map(struct dma_ops_domain *dma_dom,
                             struct unity_map_entry *e)
 {
@@ -261,6 +333,9 @@ static int dma_ops_unity_map(struct dma_ops_domain *dma_dom,
        return 0;
 }
+/*
+ * Inits the unity mappings required for a specific device
+ */
 static int init_unity_mappings_for_device(struct dma_ops_domain *dma_dom,
                                          u16 devid)
 {
@@ -278,12 +353,26 @@ static int init_unity_mappings_for_device(struct dma_ops_domain *dma_dom,
        return 0;
 }
+/****************************************************************************
+ *
+ * The next functions belong to the address allocator for the dma_ops
+ * interface functions. They work like the allocators in the other IOMMU
+ * drivers. Its basically a bitmap which marks the allocated pages in
+ * the aperture. Maybe it could be enhanced in the future to a more
+ * efficient allocator.
+ *
+ ****************************************************************************/
 static unsigned long dma_mask_to_pages(unsigned long mask)
 {
        return (mask >> PAGE_SHIFT) +
                (PAGE_ALIGN(mask & ~PAGE_MASK) >> PAGE_SHIFT);
 }
+/*
+ * The address allocator core function.
+ *
+ * called with domain->lock held
+ */
 static unsigned long dma_ops_alloc_addresses(struct device *dev,
                                             struct dma_ops_domain *dom,
                                             unsigned int pages)
@@ -317,6 +406,11 @@ static unsigned long dma_ops_alloc_addresses(struct device *dev,
        return address;
 }
+/*
+ * The address free function.
+ *
+ * called with domain->lock held
+ */
 static void dma_ops_free_addresses(struct dma_ops_domain *dom,
                                   unsigned long address,
                                   unsigned int pages)
@@ -325,6 +419,16 @@ static void dma_ops_free_addresses(struct dma_ops_domain *dom,
        iommu_area_free(dom->bitmap, address, pages);
 }
+/****************************************************************************
+ *
+ * The next functions belong to the domain allocation. A domain is
+ * allocated for every IOMMU as the default domain. If device isolation
+ * is enabled, every device get its own domain. The most important thing
+ * about domains is the page table mapping the DMA address space they
+ * contain.
+ *
+ ****************************************************************************/
 static u16 domain_id_alloc(void)
 {
        unsigned long flags;
@@ -342,6 +446,10 @@ static u16 domain_id_alloc(void)
        return id;
 }
+/*
+ * Used to reserve address ranges in the aperture (e.g. for exclusion
+ * ranges.
+ */
 static void dma_ops_reserve_addresses(struct dma_ops_domain *dom,
                                      unsigned long start_page,
                                      unsigned int pages)
@@ -382,6 +490,10 @@ static void dma_ops_free_pagetable(struct dma_ops_domain *dma_dom)
        free_page((unsigned long)p1);
 }
+/*
+ * Free a domain, only used if something went wrong in the
+ * allocation path and we need to free an already allocated page table
+ */
 static void dma_ops_domain_free(struct dma_ops_domain *dom)
 {
        if (!dom)
@@ -396,6 +508,11 @@ static void dma_ops_domain_free(struct dma_ops_domain *dom)
        kfree(dom);
 }
+/*
+ * Allocates a new protection domain usable for the dma_ops functions.
+ * It also intializes the page table and the address allocator data
+ * structures required for the dma_ops interface
+ */
 static struct dma_ops_domain *dma_ops_domain_alloc(struct amd_iommu *iommu,
                                                   unsigned order)
 {
@@ -436,6 +553,7 @@ static struct dma_ops_domain *dma_ops_domain_alloc(struct amd_iommu *iommu,
        dma_dom->bitmap[0] = 1;
        dma_dom->next_bit = 0;
+        /* Intialize the exclusion range if necessary */
        if (iommu->exclusion_start &&
            iommu->exclusion_start < dma_dom->aperture_size) {
                unsigned long startpage = iommu->exclusion_start >> PAGE_SHIFT;
@@ -444,6 +562,11 @@ static struct dma_ops_domain *dma_ops_domain_alloc(struct amd_iommu *iommu,
                dma_ops_reserve_addresses(dma_dom, startpage, pages);
        }
+        /*
+         * At the last step, build the page tables so we don't need to
+         * allocate page table pages in the dma_ops mapping/unmapping
+         * path.
+         */
        num_pte_pages = dma_dom->aperture_size / (PAGE_SIZE * 512);
        dma_dom->pte_pages = kzalloc(num_pte_pages * sizeof(void *),
                        GFP_KERNEL);
@@ -472,6 +595,10 @@ free_dma_dom:
        return NULL;
 }
+/*
+ * Find out the protection domain structure for a given PCI device. This
+ * will give us the pointer to the page table root for example.
+ */
 static struct protection_domain *domain_for_device(u16 devid)
 {
        struct protection_domain *dom;
@@ -484,6 +611,10 @@ static struct protection_domain *domain_for_device(u16 devid)
        return dom;
 }
+/*
+ * If a device is not yet associated with a domain, this function does
+ * assigns it visible for the hardware
+ */
 static void set_device_domain(struct amd_iommu *iommu,
                              struct protection_domain *domain,
                              u16 devid)
@@ -508,6 +639,19 @@ static void set_device_domain(struct amd_iommu *iommu,
        iommu->need_sync = 1;
 }
+/*****************************************************************************
+ *
+ * The next functions belong to the dma_ops mapping/unmapping code.
+ *
+ *****************************************************************************/
+/*
+ * In the dma_ops path we only have the struct device. This function
+ * finds the corresponding IOMMU, the protection domain and the
+ * requestor id for a given device.
+ * If the device is not yet associated with a domain this is also done
+ * in this function.
+ */
 static int get_device_resources(struct device *dev,
                                struct amd_iommu **iommu,
                                struct protection_domain **domain,
@@ -520,8 +664,9 @@ static int get_device_resources(struct device *dev,
        BUG_ON(!dev || dev->bus != &pci_bus_type || !dev->dma_mask);
        pcidev = to_pci_dev(dev);
-        _bdf = (pcidev->bus->number << 8) | pcidev->devfn;
+        _bdf = calc_devid(pcidev->bus->number, pcidev->devfn);
+        /* device not translated by any IOMMU in the system? */
        if (_bdf >= amd_iommu_last_bdf) {
                *iommu = NULL;
                *domain = NULL;
@@ -547,6 +692,10 @@ static int get_device_resources(struct device *dev,
        return 1;
 }
+/*
+ * This is the generic map function. It maps one 4kb page at paddr to
+ * the given address in the DMA address space for the domain.
+ */
 static dma_addr_t dma_ops_domain_map(struct amd_iommu *iommu,
                                     struct dma_ops_domain *dom,
                                     unsigned long address,
@@ -578,6 +727,9 @@ static dma_addr_t dma_ops_domain_map(struct amd_iommu *iommu,
        return (dma_addr_t)address;
 }
+/*
+ * The generic unmapping function for on page in the DMA address space.
+ */
 static void dma_ops_domain_unmap(struct amd_iommu *iommu,
                                 struct dma_ops_domain *dom,
                                 unsigned long address)
@@ -597,6 +749,12 @@ static void dma_ops_domain_unmap(struct amd_iommu *iommu,
        *pte = 0ULL;
 }
+/*
+ * This function contains common code for mapping of a physically
+ * contiguous memory region into DMA address space. It is uses by all
+ * mapping functions provided by this IOMMU driver.
+ * Must be called with the domain lock held.
+ */
 static dma_addr_t __map_single(struct device *dev,
                               struct amd_iommu *iommu,
                               struct dma_ops_domain *dma_dom,
@@ -628,6 +786,10 @@ out:
        return address;
 }
+/*
+ * Does the reverse of the __map_single function. Must be called with
+ * the domain lock held too
+ */
 static void __unmap_single(struct amd_iommu *iommu,
                           struct dma_ops_domain *dma_dom,
                           dma_addr_t dma_addr,
@@ -652,6 +814,9 @@ static void __unmap_single(struct amd_iommu *iommu,
        dma_ops_free_addresses(dma_dom, dma_addr, pages);
 }
+/*
+ * The exported map_single function for dma_ops.
+ */
 static dma_addr_t map_single(struct device *dev, phys_addr_t paddr,
                             size_t size, int dir)
 {
@@ -664,6 +829,7 @@ static dma_addr_t map_single(struct device *dev, phys_addr_t paddr,
        get_device_resources(dev, &iommu, &domain, &devid);
        if (iommu == NULL || domain == NULL)
+                /* device not handled by any AMD IOMMU */
                return (dma_addr_t)paddr;
        spin_lock_irqsave(&domain->lock, flags);
@@ -683,6 +849,9 @@ out:
        return addr;
 }
+/*
+ * The exported unmap_single function for dma_ops.
+ */
 static void unmap_single(struct device *dev, dma_addr_t dma_addr,
                         size_t size, int dir)
 {
@@ -692,6 +861,7 @@ static void unmap_single(struct device *dev, dma_addr_t dma_addr,
        u16 devid;
        if (!get_device_resources(dev, &iommu, &domain, &devid))
+                /* device not handled by any AMD IOMMU */
                return;
        spin_lock_irqsave(&domain->lock, flags);
@@ -706,6 +876,10 @@ static void unmap_single(struct device *dev, dma_addr_t dma_addr,
        spin_unlock_irqrestore(&domain->lock, flags);
 }
+/*
+ * This is a special map_sg function which is used if we should map a
+ * device which is not handled by an AMD IOMMU in the system.
+ */
 static int map_sg_no_iommu(struct device *dev, struct scatterlist *sglist,
                           int nelems, int dir)
 {
@@ -720,6 +894,10 @@ static int map_sg_no_iommu(struct device *dev, struct scatterlist *sglist,
        return nelems;
 }
+/*
+ * The exported map_sg function for dma_ops (handles scatter-gather
+ * lists).
+ */
 static int map_sg(struct device *dev, struct scatterlist *sglist,
                  int nelems, int dir)
 {
@@ -775,6 +953,10 @@ unmap:
        goto out;
 }
+/*
+ * The exported map_sg function for dma_ops (handles scatter-gather
+ * lists).
+ */
 static void unmap_sg(struct device *dev, struct scatterlist *sglist,
                     int nelems, int dir)
 {
@@ -804,6 +986,9 @@ static void unmap_sg(struct device *dev, struct scatterlist *sglist,
        spin_unlock_irqrestore(&domain->lock, flags);
 }
+/*
+ * The exported alloc_coherent function for dma_ops.
+ */
 static void *alloc_coherent(struct device *dev, size_t size,
                            dma_addr_t *dma_addr, gfp_t flag)
 {
@@ -851,6 +1036,11 @@ out:
        return virt_addr;
 }
+/*
+ * The exported free_coherent function for dma_ops.
+ * FIXME: fix the generic x86 DMA layer so that it actually calls that
+ *        function.
+ */
 static void free_coherent(struct device *dev, size_t size,
                          void *virt_addr, dma_addr_t dma_addr)
 {
@@ -879,6 +1069,8 @@ free_mem:
 }
 /*
+ * The function for pre-allocating protection domains.
+ *
 * If the driver core informs the DMA layer if a driver grabs a device
 * we don't need to preallocate the protection domains anymore.
 * For now we have to.
@@ -921,12 +1113,20 @@ static struct dma_mapping_ops amd_iommu_dma_ops = {
        .unmap_sg = unmap_sg,
 };
+/*
+ * The function which clues the AMD IOMMU driver into dma_ops.
+ */
 int __init amd_iommu_init_dma_ops(void)
 {
        struct amd_iommu *iommu;
        int order = amd_iommu_aperture_order;
        int ret;
+        /*
+         * first allocate a default protection domain for every IOMMU we
+         * found in the system. Devices not assigned to any other
+         * protection domain will be assigned to the default one.
+         */
        list_for_each_entry(iommu, &amd_iommu_list, list) {
                iommu->default_dom = dma_ops_domain_alloc(iommu, order);
                if (iommu->default_dom == NULL)
@@ -936,6 +1136,10 @@ int __init amd_iommu_init_dma_ops(void)
                        goto free_domains;
        }
+        /*
+         * If device isolation is enabled, pre-allocate the protection
+         * domains for each device.
+         */
        if (amd_iommu_isolate)
                prealloc_protection_domains();
@@ -947,6 +1151,7 @@ int __init amd_iommu_init_dma_ops(void)
        gart_iommu_aperture = 0;
 #endif
+        /* Make the driver finally visible to the drivers */
        dma_ops = &amd_iommu_dma_ops;
        return 0;
diff --git a/arch/x86/kernel/amd_iommu_init.c b/arch/x86/kernel/amd_iommu_init.c
index 2a13e430437d..c9d8ff2eb130 100644
--- a/arch/x86/kernel/amd_iommu_init.c
+++ b/arch/x86/kernel/amd_iommu_init.c
@@ -25,20 +25,13 @@
 #include <asm/pci-direct.h>
 #include <asm/amd_iommu_types.h>
 #include <asm/amd_iommu.h>
-#include <asm/gart.h>
+#include <asm/iommu.h>
 /*
 * definitions for the ACPI scanning code
 */
-#define UPDATE_LAST_BDF(x) do {\
-        if ((x) > amd_iommu_last_bdf) \
-                amd_iommu_last_bdf = (x); \
-        } while (0);
-#define DEVID(bus, devfn) (((bus) << 8) | (devfn))
 #define PCI_BUS(x) (((x) >> 8) & 0xff)
 #define IVRS_HEADER_LENGTH 48
-#define TBL_SIZE(x) (1 << (PAGE_SHIFT + get_order(amd_iommu_last_bdf * (x))))
 #define ACPI_IVHD_TYPE                  0x10
 #define ACPI_IVMD_TYPE_ALL              0x20
@@ -71,6 +64,17 @@
 #define ACPI_DEVFLAG_LINT1              0x80
 #define ACPI_DEVFLAG_ATSDIS             0x10000000
+/*
+ * ACPI table definitions
+ *
+ * These data structures are laid over the table to parse the important values
+ * out of it.
+ */
+/*
+ * structure describing one IOMMU in the ACPI table. Typically followed by one
+ * or more ivhd_entrys.
+ */
 struct ivhd_header {
        u8 type;
        u8 flags;
@@ -83,6 +87,10 @@ struct ivhd_header {
        u32 reserved;
 } __attribute__((packed));
+/*
+ * A device entry describing which devices a specific IOMMU translates and
+ * which requestor ids they use.
+ */
 struct ivhd_entry {
        u8 type;
        u16 devid;
@@ -90,6 +98,10 @@ struct ivhd_entry {
        u32 ext;
 } __attribute__((packed));
+/*
+ * An AMD IOMMU memory definition structure. It defines things like exclusion
+ * ranges for devices and regions that should be unity mapped.
+ */
 struct ivmd_header {
        u8 type;
        u8 flags;
@@ -103,22 +115,80 @@ struct ivmd_header {
 static int __initdata amd_iommu_detected;
-u16 amd_iommu_last_bdf;
+u16 amd_iommu_last_bdf;                 /* largest PCI device id we have
-struct list_head amd_iommu_unity_map;
+                                           to handle */
-unsigned amd_iommu_aperture_order = 26;
+LIST_HEAD(amd_iommu_unity_map);         /* a list of required unity mappings
-int amd_iommu_isolate;
+                                           we find in ACPI */
+unsigned amd_iommu_aperture_order = 26; /* size of aperture in power of 2 */
+int amd_iommu_isolate;                  /* if 1, device isolation is enabled */
+LIST_HEAD(amd_iommu_list);              /* list of all AMD IOMMUs in the
+                                           system */
-struct list_head amd_iommu_list;
+/*
+ * Pointer to the device table which is shared by all AMD IOMMUs
+ * it is indexed by the PCI device id or the HT unit id and contains
+ * information about the domain the device belongs to as well as the
+ * page table root pointer.
+ */
 struct dev_table_entry *amd_iommu_dev_table;
+/*
+ * The alias table is a driver specific data structure which contains the
+ * mappings of the PCI device ids to the actual requestor ids on the IOMMU.
+ * More than one device can share the same requestor id.
+ */
 u16 *amd_iommu_alias_table;
+/*
+ * The rlookup table is used to find the IOMMU which is responsible
+ * for a specific device. It is also indexed by the PCI device id.
+ */
 struct amd_iommu **amd_iommu_rlookup_table;
+/*
+ * The pd table (protection domain table) is used to find the protection domain
+ * data structure a device belongs to. Indexed with the PCI device id too.
+ */
 struct protection_domain **amd_iommu_pd_table;
+/*
+ * AMD IOMMU allows up to 2^16 differend protection domains. This is a bitmap
+ * to know which ones are already in use.
+ */
 unsigned long *amd_iommu_pd_alloc_bitmap;
-static u32 dev_table_size;
+static u32 dev_table_size;      /* size of the device table */
-static u32 alias_table_size;
+static u32 alias_table_size;    /* size of the alias table */
-static u32 rlookup_table_size;
+static u32 rlookup_table_size;  /* size if the rlookup table */
+static inline void update_last_devid(u16 devid)
+{
+        if (devid > amd_iommu_last_bdf)
+                amd_iommu_last_bdf = devid;
+}
+static inline unsigned long tbl_size(int entry_size)
+{
+        unsigned shift = PAGE_SHIFT +
+                         get_order(amd_iommu_last_bdf * entry_size);
+        return 1UL << shift;
+}
+/****************************************************************************
+ *
+ * AMD IOMMU MMIO register space handling functions
+ *
+ * These functions are used to program the IOMMU device registers in
+ * MMIO space required for that driver.
+ *
+ ****************************************************************************/
+/*
+ * This function set the exclusion range in the IOMMU. DMA accesses to the
+ * exclusion range are passed through untranslated
+ */
 static void __init iommu_set_exclusion_range(struct amd_iommu *iommu)
 {
        u64 start = iommu->exclusion_start & PAGE_MASK;
@@ -137,6 +207,7 @@ static void __init iommu_set_exclusion_range(struct amd_iommu *iommu)
                        &entry, sizeof(entry));
 }
+/* Programs the physical address of the device table into the IOMMU hardware */
 static void __init iommu_set_device_table(struct amd_iommu *iommu)
 {
        u32 entry;
@@ -149,6 +220,7 @@ static void __init iommu_set_device_table(struct amd_iommu *iommu)
                        &entry, sizeof(entry));
 }
+/* Generic functions to enable/disable certain features of the IOMMU. */
 static void __init iommu_feature_enable(struct amd_iommu *iommu, u8 bit)
 {
        u32 ctrl;
@@ -167,6 +239,7 @@ static void __init iommu_feature_disable(struct amd_iommu *iommu, u8 bit)
        writel(ctrl, iommu->mmio_base + MMIO_CONTROL_OFFSET);
 }
+/* Function to enable the hardware */
 void __init iommu_enable(struct amd_iommu *iommu)
 {
        printk(KERN_INFO "AMD IOMMU: Enabling IOMMU at ");
@@ -176,6 +249,10 @@ void __init iommu_enable(struct amd_iommu *iommu)
        iommu_feature_enable(iommu, CONTROL_IOMMU_EN);
 }
+/*
+ * mapping and unmapping functions for the IOMMU MMIO space. Each AMD IOMMU in
+ * the system has one.
+ */
 static u8 * __init iommu_map_mmio_space(u64 address)
 {
        u8 *ret;
@@ -199,16 +276,33 @@ static void __init iommu_unmap_mmio_space(struct amd_iommu *iommu)
        release_mem_region(iommu->mmio_phys, MMIO_REGION_LENGTH);
 }
+/****************************************************************************
+ *
+ * The functions below belong to the first pass of AMD IOMMU ACPI table
+ * parsing. In this pass we try to find out the highest device id this
+ * code has to handle. Upon this information the size of the shared data
+ * structures is determined later.
+ *
+ ****************************************************************************/
+/*
+ * This function reads the last device id the IOMMU has to handle from the PCI
+ * capability header for this IOMMU
+ */
 static int __init find_last_devid_on_pci(int bus, int dev, int fn, int cap_ptr)
 {
        u32 cap;
        cap = read_pci_config(bus, dev, fn, cap_ptr+MMIO_RANGE_OFFSET);
-        UPDATE_LAST_BDF(DEVID(MMIO_GET_BUS(cap), MMIO_GET_LD(cap)));
+        update_last_devid(calc_devid(MMIO_GET_BUS(cap), MMIO_GET_LD(cap)));
        return 0;
 }
+/*
+ * After reading the highest device id from the IOMMU PCI capability header
+ * this function looks if there is a higher device id defined in the ACPI table
+ */
 static int __init find_last_devid_from_ivhd(struct ivhd_header *h)
 {
        u8 *p = (void *)h, *end = (void *)h;
@@ -229,7 +323,8 @@ static int __init find_last_devid_from_ivhd(struct ivhd_header *h)
                case IVHD_DEV_RANGE_END:
                case IVHD_DEV_ALIAS:
                case IVHD_DEV_EXT_SELECT:
-                        UPDATE_LAST_BDF(dev->devid);
+                        /* all the above subfield types refer to device ids */
+                        update_last_devid(dev->devid);
                        break;
                default:
                        break;
@@ -242,6 +337,11 @@ static int __init find_last_devid_from_ivhd(struct ivhd_header *h)
        return 0;
 }
+/*
+ * Iterate over all IVHD entries in the ACPI table and find the highest device
+ * id which we need to handle. This is the first of three functions which parse
+ * the ACPI table. So we check the checksum here.
+ */
 static int __init find_last_devid_acpi(struct acpi_table_header *table)
 {
        int i;
@@ -277,19 +377,31 @@ static int __init find_last_devid_acpi(struct acpi_table_header *table)
        return 0;
 }
+/****************************************************************************
+ *
+ * The following functions belong the the code path which parses the ACPI table
+ * the second time. In this ACPI parsing iteration we allocate IOMMU specific
+ * data structures, initialize the device/alias/rlookup table and also
+ * basically initialize the hardware.
+ *
+ ****************************************************************************/
+/*
+ * Allocates the command buffer. This buffer is per AMD IOMMU. We can
+ * write commands to that buffer later and the IOMMU will execute them
+ * asynchronously
+ */
 static u8 * __init alloc_command_buffer(struct amd_iommu *iommu)
 {
-        u8 *cmd_buf = (u8 *)__get_free_pages(GFP_KERNEL,
+        u8 *cmd_buf = (u8 *)__get_free_pages(GFP_KERNEL | __GFP_ZERO,
                        get_order(CMD_BUFFER_SIZE));
-        u64 entry = 0;
+        u64 entry;
        if (cmd_buf == NULL)
                return NULL;
        iommu->cmd_buf_size = CMD_BUFFER_SIZE;
-        memset(cmd_buf, 0, CMD_BUFFER_SIZE);
        entry = (u64)virt_to_phys(cmd_buf);
        entry |= MMIO_CMD_SIZE_512;
        memcpy_toio(iommu->mmio_base + MMIO_CMD_BUF_OFFSET,
@@ -302,11 +414,10 @@ static u8 * __init alloc_command_buffer(struct amd_iommu *iommu)
 static void __init free_command_buffer(struct amd_iommu *iommu)
 {
-        if (iommu->cmd_buf)
+        free_pages((unsigned long)iommu->cmd_buf, get_order(CMD_BUFFER_SIZE));
-                free_pages((unsigned long)iommu->cmd_buf,
-                                get_order(CMD_BUFFER_SIZE));
 }
+/* sets a specific bit in the device table entry. */
 static void set_dev_entry_bit(u16 devid, u8 bit)
 {
        int i = (bit >> 5) & 0x07;
@@ -315,7 +426,18 @@ static void set_dev_entry_bit(u16 devid, u8 bit)
        amd_iommu_dev_table[devid].data[i] |= (1 << _bit);
 }
-static void __init set_dev_entry_from_acpi(u16 devid, u32 flags, u32 ext_flags)
+/* Writes the specific IOMMU for a device into the rlookup table */
+static void __init set_iommu_for_device(struct amd_iommu *iommu, u16 devid)
+{
+        amd_iommu_rlookup_table[devid] = iommu;
+}
+/*
+ * This function takes the device specific flags read from the ACPI
+ * table and sets up the device table entry with that information
+ */
+static void __init set_dev_entry_from_acpi(struct amd_iommu *iommu,
+                                           u16 devid, u32 flags, u32 ext_flags)
 {
        if (flags & ACPI_DEVFLAG_INITPASS)
                set_dev_entry_bit(devid, DEV_ENTRY_INIT_PASS);
@@ -331,13 +453,14 @@ static void __init set_dev_entry_from_acpi(u16 devid, u32 flags, u32 ext_flags)
                set_dev_entry_bit(devid, DEV_ENTRY_LINT0_PASS);
        if (flags & ACPI_DEVFLAG_LINT1)
                set_dev_entry_bit(devid, DEV_ENTRY_LINT1_PASS);
-}
-static void __init set_iommu_for_device(struct amd_iommu *iommu, u16 devid)
+        set_iommu_for_device(iommu, devid);
-{
-        amd_iommu_rlookup_table[devid] = iommu;
 }
+/*
+ * Reads the device exclusion range from ACPI and initialize IOMMU with
+ * it
+ */
 static void __init set_device_exclusion_range(u16 devid, struct ivmd_header *m)
 {
        struct amd_iommu *iommu = amd_iommu_rlookup_table[devid];
@@ -346,12 +469,22 @@ static void __init set_device_exclusion_range(u16 devid, struct ivmd_header *m)
                return;
        if (iommu) {
+                /*
+                 * We only can configure exclusion ranges per IOMMU, not
+                 * per device. But we can enable the exclusion range per
+                 * device. This is done here
+                 */
                set_dev_entry_bit(m->devid, DEV_ENTRY_EX);
                iommu->exclusion_start = m->range_start;
                iommu->exclusion_length = m->range_length;
        }
 }
+/*
+ * This function reads some important data from the IOMMU PCI space and
+ * initializes the driver data structure with it. It reads the hardware
+ * capabilities and the first/last device entries
+ */
 static void __init init_iommu_from_pci(struct amd_iommu *iommu)
 {
        int bus = PCI_BUS(iommu->devid);
@@ -363,10 +496,16 @@ static void __init init_iommu_from_pci(struct amd_iommu *iommu)
        iommu->cap = read_pci_config(bus, dev, fn, cap_ptr+MMIO_CAP_HDR_OFFSET);
        range = read_pci_config(bus, dev, fn, cap_ptr+MMIO_RANGE_OFFSET);
-        iommu->first_device = DEVID(MMIO_GET_BUS(range), MMIO_GET_FD(range));
+        iommu->first_device = calc_devid(MMIO_GET_BUS(range),
-        iommu->last_device = DEVID(MMIO_GET_BUS(range), MMIO_GET_LD(range));
+                                         MMIO_GET_FD(range));
+        iommu->last_device = calc_devid(MMIO_GET_BUS(range),
+                                        MMIO_GET_LD(range));
 }
+/*
+ * Takes a pointer to an AMD IOMMU entry in the ACPI table and
+ * initializes the hardware and our data structures with it.
+ */
 static void __init init_iommu_from_acpi(struct amd_iommu *iommu,
                                        struct ivhd_header *h)
 {
@@ -374,7 +513,7 @@ static void __init init_iommu_from_acpi(struct amd_iommu *iommu,
        u8 *end = p, flags = 0;
        u16 dev_i, devid = 0, devid_start = 0, devid_to = 0;
        u32 ext_flags = 0;
-        bool alias = 0;
+        bool alias = false;
        struct ivhd_entry *e;
        /*
@@ -414,22 +553,23 @@ static void __init init_iommu_from_acpi(struct amd_iommu *iommu,
                case IVHD_DEV_ALL:
                        for (dev_i = iommu->first_device;
                                        dev_i <= iommu->last_device; ++dev_i)
-                                set_dev_entry_from_acpi(dev_i, e->flags, 0);
+                                set_dev_entry_from_acpi(iommu, dev_i,
+                                                        e->flags, 0);
                        break;
                case IVHD_DEV_SELECT:
                        devid = e->devid;
-                        set_dev_entry_from_acpi(devid, e->flags, 0);
+                        set_dev_entry_from_acpi(iommu, devid, e->flags, 0);
                        break;
                case IVHD_DEV_SELECT_RANGE_START:
                        devid_start = e->devid;
                        flags = e->flags;
                        ext_flags = 0;
-                        alias = 0;
+                        alias = false;
                        break;
                case IVHD_DEV_ALIAS:
                        devid = e->devid;
                        devid_to = e->ext >> 8;
-                        set_dev_entry_from_acpi(devid, e->flags, 0);
+                        set_dev_entry_from_acpi(iommu, devid, e->flags, 0);
                        amd_iommu_alias_table[devid] = devid_to;
                        break;
                case IVHD_DEV_ALIAS_RANGE:
@@ -437,24 +577,25 @@ static void __init init_iommu_from_acpi(struct amd_iommu *iommu,
                        flags = e->flags;
                        devid_to = e->ext >> 8;
                        ext_flags = 0;
-                        alias = 1;
+                        alias = true;
                        break;
                case IVHD_DEV_EXT_SELECT:
                        devid = e->devid;
-                        set_dev_entry_from_acpi(devid, e->flags, e->ext);
+                        set_dev_entry_from_acpi(iommu, devid, e->flags,
+                                                e->ext);
                        break;
                case IVHD_DEV_EXT_SELECT_RANGE:
                        devid_start = e->devid;
                        flags = e->flags;
                        ext_flags = e->ext;
-                        alias = 0;
+                        alias = false;
                        break;
                case IVHD_DEV_RANGE_END:
                        devid = e->devid;
                        for (dev_i = devid_start; dev_i <= devid; ++dev_i) {
                                if (alias)
                                        amd_iommu_alias_table[dev_i] = devid_to;
-                                set_dev_entry_from_acpi(
+                                set_dev_entry_from_acpi(iommu,
                                                amd_iommu_alias_table[dev_i],
                                                flags, ext_flags);
                        }
@@ -467,6 +608,7 @@ static void __init init_iommu_from_acpi(struct amd_iommu *iommu,
        }
 }
+/* Initializes the device->iommu mapping for the driver */
 static int __init init_iommu_devices(struct amd_iommu *iommu)
 {
        u16 i;
@@ -494,6 +636,11 @@ static void __init free_iommu_all(void)
        }
 }
+/*
+ * This function clues the initialization function for one IOMMU
+ * together and also allocates the command buffer and programs the
+ * hardware. It does NOT enable the IOMMU. This is done afterwards.
+ */
 static int __init init_iommu_one(struct amd_iommu *iommu, struct ivhd_header *h)
 {
        spin_lock_init(&iommu->lock);
@@ -521,6 +668,10 @@ static int __init init_iommu_one(struct amd_iommu *iommu, struct ivhd_header *h)
        return 0;
 }
+/*
+ * Iterates over all IOMMU entries in the ACPI table, allocates the
+ * IOMMU structure and initializes it with init_iommu_one()
+ */
 static int __init init_iommu_all(struct acpi_table_header *table)
 {
        u8 *p = (u8 *)table, *end = (u8 *)table;
@@ -528,8 +679,6 @@ static int __init init_iommu_all(struct acpi_table_header *table)
        struct amd_iommu *iommu;
        int ret;
-        INIT_LIST_HEAD(&amd_iommu_list);
        end += table->length;
        p += IVRS_HEADER_LENGTH;
@@ -555,6 +704,14 @@ static int __init init_iommu_all(struct acpi_table_header *table)
        return 0;
 }
+/****************************************************************************
+ *
+ * The next functions belong to the third pass of parsing the ACPI
+ * table. In this last pass the memory mapping requirements are
+ * gathered (like exclusion and unity mapping reanges).
+ *
+ ****************************************************************************/
 static void __init free_unity_maps(void)
 {
        struct unity_map_entry *entry, *next;
@@ -565,6 +722,7 @@ static void __init free_unity_maps(void)
        }
 }
+/* called when we find an exclusion range definition in ACPI */
 static int __init init_exclusion_range(struct ivmd_header *m)
 {
        int i;
@@ -588,6 +746,7 @@ static int __init init_exclusion_range(struct ivmd_header *m)
        return 0;
 }
+/* called for unity map ACPI definition */
 static int __init init_unity_map_range(struct ivmd_header *m)
 {
        struct unity_map_entry *e = 0;
@@ -619,13 +778,12 @@ static int __init init_unity_map_range(struct ivmd_header *m)
        return 0;
 }
+/* iterates over all memory definitions we find in the ACPI table */
 static int __init init_memory_definitions(struct acpi_table_header *table)
 {
        u8 *p = (u8 *)table, *end = (u8 *)table;
        struct ivmd_header *m;
-        INIT_LIST_HEAD(&amd_iommu_unity_map);
        end += table->length;
        p += IVRS_HEADER_LENGTH;
@@ -642,6 +800,10 @@ static int __init init_memory_definitions(struct acpi_table_header *table)
        return 0;
 }
+/*
+ * This function finally enables all IOMMUs found in the system after
+ * they have been initialized
+ */
 static void __init enable_iommus(void)
 {
        struct amd_iommu *iommu;
@@ -678,6 +840,34 @@ static struct sys_device device_amd_iommu = {
        .cls = &amd_iommu_sysdev_class,
 };
+/*
+ * This is the core init function for AMD IOMMU hardware in the system.
+ * This function is called from the generic x86 DMA layer initialization
+ * code.
+ *
+ * This function basically parses the ACPI table for AMD IOMMU (IVRS)
+ * three times:
+ *
+ *      1 pass) Find the highest PCI device id the driver has to handle.
+ *              Upon this information the size of the data structures is
+ *              determined that needs to be allocated.
+ *
+ *      2 pass) Initialize the data structures just allocated with the
+ *              information in the ACPI table about available AMD IOMMUs
+ *              in the system. It also maps the PCI devices in the
+ *              system to specific IOMMUs
+ *
+ *      3 pass) After the basic data structures are allocated and
+ *              initialized we update them with information about memory
+ *              remapping requirements parsed out of the ACPI table in
+ *              this last pass.
+ *
+ * After that the hardware is initialized and ready to go. In the last
+ * step we do some Linux specific things like registering the driver in
+ * the dma_ops interface and initializing the suspend/resume support
+ * functions. Finally it prints some information about AMD IOMMUs and
+ * the driver state and enables the hardware.
+ */
 int __init amd_iommu_init(void)
 {
        int i, ret = 0;
@@ -699,14 +889,14 @@ int __init amd_iommu_init(void)
        if (acpi_table_parse("IVRS", find_last_devid_acpi) != 0)
                return -ENODEV;
-        dev_table_size     = TBL_SIZE(DEV_TABLE_ENTRY_SIZE);
+        dev_table_size     = tbl_size(DEV_TABLE_ENTRY_SIZE);
-        alias_table_size   = TBL_SIZE(ALIAS_TABLE_ENTRY_SIZE);
+        alias_table_size   = tbl_size(ALIAS_TABLE_ENTRY_SIZE);
-        rlookup_table_size = TBL_SIZE(RLOOKUP_TABLE_ENTRY_SIZE);
+        rlookup_table_size = tbl_size(RLOOKUP_TABLE_ENTRY_SIZE);
        ret = -ENOMEM;
        /* Device table - directly used by all IOMMUs */
-        amd_iommu_dev_table = (void *)__get_free_pages(GFP_KERNEL,
+        amd_iommu_dev_table = (void *)__get_free_pages(GFP_KERNEL | __GFP_ZERO,
                                      get_order(dev_table_size));
        if (amd_iommu_dev_table == NULL)
                goto out;
@@ -730,27 +920,23 @@ int __init amd_iommu_init(void)
         * Protection Domain table - maps devices to protection domains
         * This table has the same size as the rlookup_table
         */
-        amd_iommu_pd_table = (void *)__get_free_pages(GFP_KERNEL,
+        amd_iommu_pd_table = (void *)__get_free_pages(GFP_KERNEL | __GFP_ZERO,
                                     get_order(rlookup_table_size));
        if (amd_iommu_pd_table == NULL)
                goto free;
-        amd_iommu_pd_alloc_bitmap = (void *)__get_free_pages(GFP_KERNEL,
+        amd_iommu_pd_alloc_bitmap = (void *)__get_free_pages(
+                                            GFP_KERNEL | __GFP_ZERO,
                                            get_order(MAX_DOMAIN_ID/8));
        if (amd_iommu_pd_alloc_bitmap == NULL)
                goto free;
        /*
-         * memory is allocated now; initialize the device table with all zeroes
+         * let all alias entries point to itself
-         * and let all alias entries point to itself
         */
-        memset(amd_iommu_dev_table, 0, dev_table_size);
        for (i = 0; i < amd_iommu_last_bdf; ++i)
                amd_iommu_alias_table[i] = i;
-        memset(amd_iommu_pd_table, 0, rlookup_table_size);
-        memset(amd_iommu_pd_alloc_bitmap, 0, MAX_DOMAIN_ID / 8);
        /*
         * never allocate domain 0 because its used as the non-allocated and
         * error value placeholder
@@ -795,24 +981,19 @@ out:
        return ret;
 free:
-        if (amd_iommu_pd_alloc_bitmap)
+        free_pages((unsigned long)amd_iommu_pd_alloc_bitmap, 1);
-                free_pages((unsigned long)amd_iommu_pd_alloc_bitmap, 1);
-        if (amd_iommu_pd_table)
+        free_pages((unsigned long)amd_iommu_pd_table,
-                free_pages((unsigned long)amd_iommu_pd_table,
+                   get_order(rlookup_table_size));
-                                get_order(rlookup_table_size));
-        if (amd_iommu_rlookup_table)
+        free_pages((unsigned long)amd_iommu_rlookup_table,
-                free_pages((unsigned long)amd_iommu_rlookup_table,
+                   get_order(rlookup_table_size));
-                                get_order(rlookup_table_size));
-        if (amd_iommu_alias_table)
+        free_pages((unsigned long)amd_iommu_alias_table,
-                free_pages((unsigned long)amd_iommu_alias_table,
+                   get_order(alias_table_size));
-                                get_order(alias_table_size));
-        if (amd_iommu_dev_table)
+        free_pages((unsigned long)amd_iommu_dev_table,
-                free_pages((unsigned long)amd_iommu_dev_table,
+                   get_order(dev_table_size));
-                                get_order(dev_table_size));
        free_iommu_all();
@@ -821,6 +1002,13 @@ free:
        goto out;
 }
+/****************************************************************************
+ *
+ * Early detect code. This code runs at IOMMU detection time in the DMA
+ * layer. It just looks if there is an IVRS ACPI table to detect AMD
+ * IOMMUs
+ *
+ ****************************************************************************/
 static int __init early_amd_iommu_detect(struct acpi_table_header *table)
 {
        return 0;
@@ -828,7 +1016,7 @@ static int __init early_amd_iommu_detect(struct acpi_table_header *table)
 void __init amd_iommu_detect(void)
 {
-        if (swiotlb || no_iommu || iommu_detected)
+        if (swiotlb || no_iommu || (iommu_detected && !gart_iommu_aperture))
                return;
        if (acpi_table_parse("IVRS", early_amd_iommu_detect) == 0) {
@@ -841,6 +1029,13 @@ void __init amd_iommu_detect(void)
        }
 }
+/****************************************************************************
+ *
+ * Parsing functions for the AMD IOMMU specific kernel command line
+ * options.
+ *
+ ****************************************************************************/
 static int __init parse_amd_iommu_options(char *str)
 {
        for (; *str; ++str) {
@@ -853,20 +1048,10 @@ static int __init parse_amd_iommu_options(char *str)
 static int __init parse_amd_iommu_size_options(char *str)
 {
-        for (; *str; ++str) {
+        unsigned order = PAGE_SHIFT + get_order(memparse(str, &str));
-                if (strcmp(str, "32M") == 0)
-                        amd_iommu_aperture_order = 25;
+        if ((order > 24) && (order < 31))
-                if (strcmp(str, "64M") == 0)
+                amd_iommu_aperture_order = order;
-                        amd_iommu_aperture_order = 26;
-                if (strcmp(str, "128M") == 0)
-                        amd_iommu_aperture_order = 27;
-                if (strcmp(str, "256M") == 0)
-                        amd_iommu_aperture_order = 28;
-                if (strcmp(str, "512M") == 0)
-                        amd_iommu_aperture_order = 29;
-                if (strcmp(str, "1G") == 0)
-                        amd_iommu_aperture_order = 30;
-        }
        return 1;
 }
diff --git a/arch/x86/kernel/aperture_64.c b/arch/x86/kernel/aperture_64.c
index 9f907806c1a5..44e21826db11 100644
--- a/arch/x86/kernel/aperture_64.c
+++ b/arch/x86/kernel/aperture_64.c
@@ -21,6 +21,7 @@
 #include <linux/suspend.h>
 #include <asm/e820.h>
 #include <asm/io.h>
+#include <asm/iommu.h>
 #include <asm/gart.h>
 #include <asm/pci-direct.h>
 #include <asm/dma.h>
diff --git a/arch/x86/kernel/apic_32.c b/arch/x86/kernel/apic_32.c
index a437d027f20b..d6c898358371 100644
--- a/arch/x86/kernel/apic_32.c
+++ b/arch/x86/kernel/apic_32.c
@@ -75,7 +75,7 @@ char system_vectors[NR_VECTORS] = { [0 ... NR_VECTORS-1] = SYS_VECTOR_FREE};
 /*
 * Debug level, exported for io_apic.c
 */
-int apic_verbosity;
+unsigned int apic_verbosity;
 int pic_mode;
@@ -177,7 +177,7 @@ void __cpuinit enable_NMI_through_LVT0(void)
        /* Level triggered for 82489DX */
        if (!lapic_is_integrated())
                v |= APIC_LVT_LEVEL_TRIGGER;
-        apic_write_around(APIC_LVT0, v);
+        apic_write(APIC_LVT0, v);
 }
 /**
@@ -212,9 +212,6 @@ int lapic_get_maxlvt(void)
 * this function twice on the boot CPU, once with a bogus timeout
 * value, second time for real. The other (noncalibrating) CPUs
 * call this function only once, with the real, calibrated value.
- *
- * We do reads before writes even if unnecessary, to get around the
- * P5 APIC double write bug.
 */
 static void __setup_APIC_LVTT(unsigned int clocks, int oneshot, int irqen)
 {
@@ -229,18 +226,18 @@ static void __setup_APIC_LVTT(unsigned int clocks, int oneshot, int irqen)
        if (!irqen)
                lvtt_value |= APIC_LVT_MASKED;
-        apic_write_around(APIC_LVTT, lvtt_value);
+        apic_write(APIC_LVTT, lvtt_value);
        /*
         * Divide PICLK by 16
         */
        tmp_value = apic_read(APIC_TDCR);
-        apic_write_around(APIC_TDCR, (tmp_value
+        apic_write(APIC_TDCR,
-                                & ~(APIC_TDR_DIV_1 | APIC_TDR_DIV_TMBASE))
+                   (tmp_value & ~(APIC_TDR_DIV_1 | APIC_TDR_DIV_TMBASE)) |
-                                | APIC_TDR_DIV_16);
+                   APIC_TDR_DIV_16);
        if (!oneshot)
-                apic_write_around(APIC_TMICT, clocks/APIC_DIVISOR);
+                apic_write(APIC_TMICT, clocks / APIC_DIVISOR);
 }
 /*
@@ -249,7 +246,7 @@ static void __setup_APIC_LVTT(unsigned int clocks, int oneshot, int irqen)
 static int lapic_next_event(unsigned long delta,
                            struct clock_event_device *evt)
 {
-        apic_write_around(APIC_TMICT, delta);
+        apic_write(APIC_TMICT, delta);
        return 0;
 }
@@ -278,7 +275,7 @@ static void lapic_timer_setup(enum clock_event_mode mode,
        case CLOCK_EVT_MODE_SHUTDOWN:
                v = apic_read(APIC_LVTT);
                v |= (APIC_LVT_MASKED | LOCAL_TIMER_VECTOR);
-                apic_write_around(APIC_LVTT, v);
+                apic_write(APIC_LVTT, v);
                break;
        case CLOCK_EVT_MODE_RESUME:
                /* Nothing to do here */
@@ -372,12 +369,7 @@ static void __init lapic_cal_handler(struct clock_event_device *dev)
        }
 }
-/*
+static int __init calibrate_APIC_clock(void)
- * Setup the boot APIC
- *
- * Calibrate and verify the result.
- */
-void __init setup_boot_APIC_clock(void)
 {
        struct clock_event_device *levt = &__get_cpu_var(lapic_events);
        const long pm_100ms = PMTMR_TICKS_PER_SEC/10;
@@ -387,24 +379,6 @@ void __init setup_boot_APIC_clock(void)
        long delta, deltapm;
        int pm_referenced = 0;
-        /*
-         * The local apic timer can be disabled via the kernel
-         * commandline or from the CPU detection code. Register the lapic
-         * timer as a dummy clock event source on SMP systems, so the
-         * broadcast mechanism is used. On UP systems simply ignore it.
-         */
-        if (local_apic_timer_disabled) {
-                /* No broadcast on UP ! */
-                if (num_possible_cpus() > 1) {
-                        lapic_clockevent.mult = 1;
-                        setup_APIC_timer();
-                }
-                return;
-        }
-        apic_printk(APIC_VERBOSE, "Using local APIC timer interrupts.\n"
-                    "calibrating APIC timer ...\n");
        local_irq_disable();
        /* Replace the global interrupt handler */
@@ -489,8 +463,6 @@ void __init setup_boot_APIC_clock(void)
                    calibration_result / (1000000 / HZ),
                    calibration_result % (1000000 / HZ));
-        local_apic_timer_verify_ok = 1;
        /*
         * Do a sanity check on the APIC calibration result
         */
@@ -498,12 +470,11 @@ void __init setup_boot_APIC_clock(void)
                local_irq_enable();
                printk(KERN_WARNING
                       "APIC frequency too slow, disabling apic timer\n");
-                /* No broadcast on UP ! */
+                return -1;
-                if (num_possible_cpus() > 1)
-                        setup_APIC_timer();
-                return;
        }
+        local_apic_timer_verify_ok = 1;
        /* We trust the pm timer based calibration */
        if (!pm_referenced) {
                apic_printk(APIC_VERBOSE, "... verify APIC timer\n");
@@ -543,22 +514,55 @@ void __init setup_boot_APIC_clock(void)
        if (!local_apic_timer_verify_ok) {
                printk(KERN_WARNING
                       "APIC timer disabled due to verification failure.\n");
+                        return -1;
+        }
+        return 0;
+}
+/*
+ * Setup the boot APIC
+ *
+ * Calibrate and verify the result.
+ */
+void __init setup_boot_APIC_clock(void)
+{
+        /*
+         * The local apic timer can be disabled via the kernel
+         * commandline or from the CPU detection code. Register the lapic
+         * timer as a dummy clock event source on SMP systems, so the
+         * broadcast mechanism is used. On UP systems simply ignore it.
+         */
+        if (local_apic_timer_disabled) {
                /* No broadcast on UP ! */
-                if (num_possible_cpus() == 1)
+                if (num_possible_cpus() > 1) {
-                        return;
+                        lapic_clockevent.mult = 1;
-        } else {
+                        setup_APIC_timer();
-                /*
+                }
-                 * If nmi_watchdog is set to IO_APIC, we need the
+                return;
-                 * PIT/HPET going.  Otherwise register lapic as a dummy
-                 * device.
-                 */
-                if (nmi_watchdog != NMI_IO_APIC)
-                        lapic_clockevent.features &= ~CLOCK_EVT_FEAT_DUMMY;
-                else
-                        printk(KERN_WARNING "APIC timer registered as dummy,"
-                                " due to nmi_watchdog=%d!\n", nmi_watchdog);
        }
+        apic_printk(APIC_VERBOSE, "Using local APIC timer interrupts.\n"
+                    "calibrating APIC timer ...\n");
+        if (calibrate_APIC_clock()) {
+                /* No broadcast on UP ! */
+                if (num_possible_cpus() > 1)
+                        setup_APIC_timer();
+                return;
+        }
+        /*
+         * If nmi_watchdog is set to IO_APIC, we need the
+         * PIT/HPET going.  Otherwise register lapic as a dummy
+         * device.
+         */
+        if (nmi_watchdog != NMI_IO_APIC)
+                lapic_clockevent.features &= ~CLOCK_EVT_FEAT_DUMMY;
+        else
+                printk(KERN_WARNING "APIC timer registered as dummy,"
+                        " due to nmi_watchdog=%d!\n", nmi_watchdog);
        /* Setup the lapic or request the broadcast */
        setup_APIC_timer();
 }
@@ -693,44 +697,44 @@ void clear_local_APIC(void)
         */
        if (maxlvt >= 3) {
                v = ERROR_APIC_VECTOR; /* any non-zero vector will do */
-                apic_write_around(APIC_LVTERR, v | APIC_LVT_MASKED);
+                apic_write(APIC_LVTERR, v | APIC_LVT_MASKED);
        }
        /*
         * Careful: we have to set masks only first to deassert
         * any level-triggered sources.
         */
        v = apic_read(APIC_LVTT);
-        apic_write_around(APIC_LVTT, v | APIC_LVT_MASKED);
+        apic_write(APIC_LVTT, v | APIC_LVT_MASKED);
        v = apic_read(APIC_LVT0);
-        apic_write_around(APIC_LVT0, v | APIC_LVT_MASKED);
+        apic_write(APIC_LVT0, v | APIC_LVT_MASKED);
        v = apic_read(APIC_LVT1);
-        apic_write_around(APIC_LVT1, v | APIC_LVT_MASKED);
+        apic_write(APIC_LVT1, v | APIC_LVT_MASKED);
        if (maxlvt >= 4) {
                v = apic_read(APIC_LVTPC);
-                apic_write_around(APIC_LVTPC, v | APIC_LVT_MASKED);
+                apic_write(APIC_LVTPC, v | APIC_LVT_MASKED);
        }
        /* lets not touch this if we didn't frob it */
 #ifdef CONFIG_X86_MCE_P4THERMAL
        if (maxlvt >= 5) {
                v = apic_read(APIC_LVTTHMR);
-                apic_write_around(APIC_LVTTHMR, v | APIC_LVT_MASKED);
+                apic_write(APIC_LVTTHMR, v | APIC_LVT_MASKED);
        }
 #endif
        /*
         * Clean APIC state for other OSs:
         */
-        apic_write_around(APIC_LVTT, APIC_LVT_MASKED);
+        apic_write(APIC_LVTT, APIC_LVT_MASKED);
-        apic_write_around(APIC_LVT0, APIC_LVT_MASKED);
+        apic_write(APIC_LVT0, APIC_LVT_MASKED);
-        apic_write_around(APIC_LVT1, APIC_LVT_MASKED);
+        apic_write(APIC_LVT1, APIC_LVT_MASKED);
        if (maxlvt >= 3)
-                apic_write_around(APIC_LVTERR, APIC_LVT_MASKED);
+                apic_write(APIC_LVTERR, APIC_LVT_MASKED);
        if (maxlvt >= 4)
-                apic_write_around(APIC_LVTPC, APIC_LVT_MASKED);
+                apic_write(APIC_LVTPC, APIC_LVT_MASKED);
 #ifdef CONFIG_X86_MCE_P4THERMAL
        if (maxlvt >= 5)
-                apic_write_around(APIC_LVTTHMR, APIC_LVT_MASKED);
+                apic_write(APIC_LVTTHMR, APIC_LVT_MASKED);
 #endif
        /* Integrated APIC (!82489DX) ? */
        if (lapic_is_integrated()) {
@@ -756,7 +760,7 @@ void disable_local_APIC(void)
         */
        value = apic_read(APIC_SPIV);
        value &= ~APIC_SPIV_APIC_ENABLED;
-        apic_write_around(APIC_SPIV, value);
+        apic_write(APIC_SPIV, value);
        /*
         * When LAPIC was disabled by the BIOS and enabled by the kernel,
@@ -865,8 +869,8 @@ void __init sync_Arb_IDs(void)
        apic_wait_icr_idle();
        apic_printk(APIC_DEBUG, "Synchronizing Arb IDs.\n");
-        apic_write_around(APIC_ICR, APIC_DEST_ALLINC | APIC_INT_LEVELTRIG
+        apic_write(APIC_ICR,
-                                | APIC_DM_INIT);
+                   APIC_DEST_ALLINC | APIC_INT_LEVELTRIG | APIC_DM_INIT);
 }
 /*
@@ -902,16 +906,16 @@ void __init init_bsp_APIC(void)
        else
                value |= APIC_SPIV_FOCUS_DISABLED;
        value |= SPURIOUS_APIC_VECTOR;
-        apic_write_around(APIC_SPIV, value);
+        apic_write(APIC_SPIV, value);
        /*
         * Set up the virtual wire mode.
         */
-        apic_write_around(APIC_LVT0, APIC_DM_EXTINT);
+        apic_write(APIC_LVT0, APIC_DM_EXTINT);
        value = APIC_DM_NMI;
        if (!lapic_is_integrated())             /* 82489DX */
                value |= APIC_LVT_LEVEL_TRIGGER;
-        apic_write_around(APIC_LVT1, value);
+        apic_write(APIC_LVT1, value);
 }
 static void __cpuinit lapic_setup_esr(void)
@@ -926,7 +930,7 @@ static void __cpuinit lapic_setup_esr(void)
                /* enables sending errors */
                value = ERROR_APIC_VECTOR;
-                apic_write_around(APIC_LVTERR, value);
+                apic_write(APIC_LVTERR, value);
                /*
                 * spec says clear errors after enabling vector.
                 */
@@ -989,7 +993,7 @@ void __cpuinit setup_local_APIC(void)
         */
        value = apic_read(APIC_TASKPRI);
        value &= ~APIC_TPRI_MASK;
-        apic_write_around(APIC_TASKPRI, value);
+        apic_write(APIC_TASKPRI, value);
        /*
         * After a crash, we no longer service the interrupts and a pending
@@ -1047,7 +1051,7 @@ void __cpuinit setup_local_APIC(void)
         * Set spurious IRQ vector
         */
        value |= SPURIOUS_APIC_VECTOR;
-        apic_write_around(APIC_SPIV, value);
+        apic_write(APIC_SPIV, value);
        /*
         * Set up LVT0, LVT1:
@@ -1069,7 +1073,7 @@ void __cpuinit setup_local_APIC(void)
                apic_printk(APIC_VERBOSE, "masked ExtINT on CPU#%d\n",
                                smp_processor_id());
        }
-        apic_write_around(APIC_LVT0, value);
+        apic_write(APIC_LVT0, value);
        /*
         * only the BP should see the LINT1 NMI signal, obviously.
@@ -1080,7 +1084,7 @@ void __cpuinit setup_local_APIC(void)
                value = APIC_DM_NMI | APIC_LVT_MASKED;
        if (!integrated)                /* 82489DX */
                value |= APIC_LVT_LEVEL_TRIGGER;
-        apic_write_around(APIC_LVT1, value);
+        apic_write(APIC_LVT1, value);
 }
 void __cpuinit end_local_APIC_setup(void)
@@ -1091,7 +1095,7 @@ void __cpuinit end_local_APIC_setup(void)
        /* Disable the local apic timer */
        value = apic_read(APIC_LVTT);
        value |= (APIC_LVT_MASKED | LOCAL_TIMER_VECTOR);
-        apic_write_around(APIC_LVTT, value);
+        apic_write(APIC_LVTT, value);
        setup_apic_nmi_watchdog(NULL);
        apic_pm_activate();
@@ -1214,9 +1218,6 @@ int apic_version[MAX_APICS];
 int __init APIC_init_uniprocessor(void)
 {
-        if (disable_apic)
-                clear_cpu_cap(&boot_cpu_data, X86_FEATURE_APIC);
        if (!smp_found_config && !cpu_has_apic)
                return -1;
@@ -1419,7 +1420,7 @@ void disconnect_bsp_APIC(int virt_wire_setup)
                value &= ~APIC_VECTOR_MASK;
                value |= APIC_SPIV_APIC_ENABLED;
                value |= 0xf;
-                apic_write_around(APIC_SPIV, value);
+                apic_write(APIC_SPIV, value);
                if (!virt_wire_setup) {
                        /*
@@ -1432,10 +1433,10 @@ void disconnect_bsp_APIC(int virt_wire_setup)
                                APIC_LVT_LEVEL_TRIGGER | APIC_LVT_MASKED);
                        value |= APIC_LVT_REMOTE_IRR | APIC_SEND_PENDING;
                        value = SET_APIC_DELIVERY_MODE(value, APIC_MODE_EXTINT);
-                        apic_write_around(APIC_LVT0, value);
+                        apic_write(APIC_LVT0, value);
                } else {
                        /* Disable LVT0 */
-                        apic_write_around(APIC_LVT0, APIC_LVT_MASKED);
+                        apic_write(APIC_LVT0, APIC_LVT_MASKED);
                }
                /*
@@ -1449,7 +1450,7 @@ void disconnect_bsp_APIC(int virt_wire_setup)
                        APIC_LVT_LEVEL_TRIGGER | APIC_LVT_MASKED);
                value |= APIC_LVT_REMOTE_IRR | APIC_SEND_PENDING;
                value = SET_APIC_DELIVERY_MODE(value, APIC_MODE_NMI);
-                apic_write_around(APIC_LVT1, value);
+                apic_write(APIC_LVT1, value);
        }
 }
@@ -1700,7 +1701,7 @@ early_param("lapic", parse_lapic);
 static int __init parse_nolapic(char *arg)
 {
        disable_apic = 1;
-        clear_cpu_cap(&boot_cpu_data, X86_FEATURE_APIC);
+        setup_clear_cpu_cap(X86_FEATURE_APIC);
        return 0;
 }
 early_param("nolapic", parse_nolapic);
diff --git a/arch/x86/kernel/apic_64.c b/arch/x86/kernel/apic_64.c
index 1e3d32e27c14..7f1f030da7ee 100644
--- a/arch/x86/kernel/apic_64.c
+++ b/arch/x86/kernel/apic_64.c
@@ -54,7 +54,7 @@ EXPORT_SYMBOL_GPL(local_apic_timer_c2_ok);
 /*
 * Debug level, exported for io_apic.c
 */
-int apic_verbosity;
+unsigned int apic_verbosity;
 /* Have we found an MP table */
 int smp_found_config;
@@ -314,7 +314,7 @@ static void setup_APIC_timer(void)
 #define TICK_COUNT 100000000
-static void __init calibrate_APIC_clock(void)
+static int __init calibrate_APIC_clock(void)
 {
        unsigned apic, apic_start;
        unsigned long tsc, tsc_start;
@@ -368,6 +368,17 @@ static void __init calibrate_APIC_clock(void)
                clockevent_delta2ns(0xF, &lapic_clockevent);
        calibration_result = result / HZ;
+        /*
+         * Do a sanity check on the APIC calibration result
+         */
+        if (calibration_result < (1000000 / HZ)) {
+                printk(KERN_WARNING
+                        "APIC frequency too slow, disabling apic timer\n");
+                return -1;
+        }
+        return 0;
 }
 /*
@@ -394,14 +405,7 @@ void __init setup_boot_APIC_clock(void)
        }
        printk(KERN_INFO "Using local APIC timer interrupts.\n");
-        calibrate_APIC_clock();
+        if (calibrate_APIC_clock()) {
-        /*
-         * Do a sanity check on the APIC calibration result
-         */
-        if (calibration_result < (1000000 / HZ)) {
-                printk(KERN_WARNING
-                       "APIC frequency too slow, disabling apic timer\n");
                /* No broadcast on UP ! */
                if (num_possible_cpus() > 1)
                        setup_APIC_timer();
@@ -1337,7 +1341,7 @@ early_param("apic", apic_set_verbosity);
 static __init int setup_disableapic(char *str)
 {
        disable_apic = 1;
-        clear_cpu_cap(&boot_cpu_data, X86_FEATURE_APIC);
+        setup_clear_cpu_cap(X86_FEATURE_APIC);
        return 0;
 }
 early_param("disableapic", setup_disableapic);
diff --git a/arch/x86/kernel/asm-offsets_64.c b/arch/x86/kernel/asm-offsets_64.c
index bacf5deeec2d..aa89387006fe 100644
--- a/arch/x86/kernel/asm-offsets_64.c
+++ b/arch/x86/kernel/asm-offsets_64.c
@@ -18,6 +18,8 @@
 #include <asm/ia32.h>
 #include <asm/bootparam.h>
+#include <xen/interface/xen.h>
 #define __NO_STUBS 1
 #undef __SYSCALL
 #undef _ASM_X86_64_UNISTD_H_
@@ -131,5 +133,14 @@ int main(void)
        OFFSET(BP_loadflags, boot_params, hdr.loadflags);
        OFFSET(BP_hardware_subarch, boot_params, hdr.hardware_subarch);
        OFFSET(BP_version, boot_params, hdr.version);
+        BLANK();
+        DEFINE(PAGE_SIZE_asm, PAGE_SIZE);
+#ifdef CONFIG_XEN
+        BLANK();
+        OFFSET(XEN_vcpu_info_mask, vcpu_info, evtchn_upcall_mask);
+        OFFSET(XEN_vcpu_info_pending, vcpu_info, evtchn_upcall_pending);
+#undef ENTRY
+#endif
        return 0;
 }
diff --git a/arch/x86/kernel/bios_uv.c b/arch/x86/kernel/bios_uv.c
new file mode 100644
index 000000000000..c639bd55391c
--- /dev/null
+++ b/arch/x86/kernel/bios_uv.c
@@ -0,0 +1,48 @@
+/*
+ * BIOS run time interface routines.
+ *
+ *  Copyright (c) 2008 Silicon Graphics, Inc.  All Rights Reserved.
+ *
+ *  This program is free software; you can redistribute it and/or modify
+ *  it under the terms of the GNU General Public License as published by
+ *  the Free Software Foundation; either version 2 of the License, or
+ *  (at your option) any later version.
+ *
+ *  This program is distributed in the hope that it will be useful,
+ *  but WITHOUT ANY WARRANTY; without even the implied warranty of
+ *  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ *  GNU General Public License for more details.
+ *
+ *  You should have received a copy of the GNU General Public License
+ *  along with this program; if not, write to the Free Software
+ *  Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307 USA
+ */
+#include <asm/uv/bios.h>
+const char *
+x86_bios_strerror(long status)
+{
+        const char *str;
+        switch (status) {
+        case  0: str = "Call completed without error"; break;
+        case -1: str = "Not implemented"; break;
+        case -2: str = "Invalid argument"; break;
+        case -3: str = "Call completed with error"; break;
+        default: str = "Unknown BIOS status code"; break;
+        }
+        return str;
+}
+long
+x86_bios_freq_base(unsigned long which, unsigned long *ticks_per_second,
+                   unsigned long *drift_info)
+{
+        struct uv_bios_retval isrv;
+        BIOS_CALL(isrv, BIOS_FREQ_BASE, which, 0, 0, 0, 0, 0, 0);
+        *ticks_per_second = isrv.v0;
+        *drift_info = isrv.v1;
+        return isrv.status;
+}
+EXPORT_SYMBOL_GPL(x86_bios_freq_base);
diff --git a/arch/x86/kernel/cpu/amd.c b/arch/x86/kernel/cpu/amd.c
index 81a07ca65d44..cae9cabc3031 100644
--- a/arch/x86/kernel/cpu/amd.c
+++ b/arch/x86/kernel/cpu/amd.c
@@ -24,8 +24,6 @@
 extern void vide(void);
 __asm__(".align 4\nvide: ret");
-int force_mwait __cpuinitdata;
 static void __cpuinit early_init_amd(struct cpuinfo_x86 *c)
 {
        if (cpuid_eax(0x80000000) >= 0x80000007) {
diff --git a/arch/x86/kernel/cpu/amd_64.c b/arch/x86/kernel/cpu/amd_64.c
index 7c36fb8a28d4..d1692b2a41ff 100644
--- a/arch/x86/kernel/cpu/amd_64.c
+++ b/arch/x86/kernel/cpu/amd_64.c
@@ -115,6 +115,8 @@ static void __cpuinit early_init_amd(struct cpuinfo_x86 *c)
        /* c->x86_power is 8000_0007 edx. Bit 8 is constant TSC */
        if (c->x86_power & (1<<8))
                set_cpu_cap(c, X86_FEATURE_CONSTANT_TSC);
+        set_cpu_cap(c, X86_FEATURE_SYSCALL32);
 }
 static void __cpuinit init_amd(struct cpuinfo_x86 *c)
diff --git a/arch/x86/kernel/cpu/bugs.c b/arch/x86/kernel/cpu/bugs.c
index 1b1c56bb338f..c9b58a806e85 100644
--- a/arch/x86/kernel/cpu/bugs.c
+++ b/arch/x86/kernel/cpu/bugs.c
@@ -131,13 +131,7 @@ static void __init check_popad(void)
 *   (for due to lack of "invlpg" and working WP on a i386)
 * - In order to run on anything without a TSC, we need to be
 *   compiled for a i486.
- * - In order to support the local APIC on a buggy Pentium machine,
+ */
- *   we need to be compiled with CONFIG_X86_GOOD_APIC disabled,
- *   which happens implicitly if compiled for a Pentium or lower
- *   (unless an advanced selection of CPU features is used) as an
- *   otherwise config implies a properly working local APIC without
- *   the need to do extra reads from the APIC.
-*/
 static void __init check_config(void)
 {
@@ -151,21 +145,6 @@ static void __init check_config(void)
        if (boot_cpu_data.x86 == 3)
                panic("Kernel requires i486+ for 'invlpg' and other features");
 #endif
-/*
- * If we were told we had a good local APIC, check for buggy Pentia,
- * i.e. all B steppings and the C2 stepping of P54C when using their
- * integrated APIC (see 11AP erratum in "Pentium Processor
- * Specification Update").
- */
-#if defined(CONFIG_X86_LOCAL_APIC) && defined(CONFIG_X86_GOOD_APIC)
-        if (boot_cpu_data.x86_vendor == X86_VENDOR_INTEL
-            && cpu_has_apic
-            && boot_cpu_data.x86 == 5
-            && boot_cpu_data.x86_model == 2
-            && (boot_cpu_data.x86_mask < 6 || boot_cpu_data.x86_mask == 11))
-                panic("Kernel compiled for PMMX+, assumes a local APIC without the read-before-write bug!");
-#endif
 }
diff --git a/arch/x86/kernel/cpu/common_64.c b/arch/x86/kernel/cpu/common_64.c
index 7b8cc72feb40..dd6e3f15017e 100644
--- a/arch/x86/kernel/cpu/common_64.c
+++ b/arch/x86/kernel/cpu/common_64.c
@@ -7,15 +7,13 @@
 #include <linux/module.h>
 #include <linux/kgdb.h>
 #include <linux/topology.h>
-#include <linux/string.h>
 #include <linux/delay.h>
 #include <linux/smp.h>
-#include <linux/module.h>
 #include <linux/percpu.h>
-#include <asm/processor.h>
 #include <asm/i387.h>
 #include <asm/msr.h>
 #include <asm/io.h>
+#include <asm/linkage.h>
 #include <asm/mmu_context.h>
 #include <asm/mtrr.h>
 #include <asm/mce.h>
@@ -305,7 +303,6 @@ static void __cpuinit early_identify_cpu(struct cpuinfo_x86 *c)
                        c->x86_capability[2] = cpuid_edx(0x80860001);
        }
-        c->extended_cpuid_level = cpuid_eax(0x80000000);
        if (c->extended_cpuid_level >= 0x80000007)
                c->x86_power = cpuid_edx(0x80000007);
@@ -316,18 +313,11 @@ static void __cpuinit early_identify_cpu(struct cpuinfo_x86 *c)
                c->x86_phys_bits = eax & 0xff;
        }
-        /* Assume all 64-bit CPUs support 32-bit syscall */
-        set_cpu_cap(c, X86_FEATURE_SYSCALL32);
        if (c->x86_vendor != X86_VENDOR_UNKNOWN &&
            cpu_devs[c->x86_vendor]->c_early_init)
                cpu_devs[c->x86_vendor]->c_early_init(c);
        validate_pat_support(c);
-        /* early_param could clear that, but recall get it set again */
-        if (disable_apic)
-                clear_cpu_cap(c, X86_FEATURE_APIC);
 }
 /*
@@ -517,8 +507,7 @@ void pda_init(int cpu)
 }
 char boot_exception_stacks[(N_EXCEPTION_STACKS - 1) * EXCEPTION_STKSZ +
-                           DEBUG_STKSZ]
+                           DEBUG_STKSZ] __page_aligned_bss;
-__attribute__((section(".bss.page_aligned")));
 extern asmlinkage void ignore_sysret(void);
diff --git a/arch/x86/kernel/cpu/intel.c b/arch/x86/kernel/cpu/intel.c
index 70609efdf1da..b75f2569b8f8 100644
--- a/arch/x86/kernel/cpu/intel.c
+++ b/arch/x86/kernel/cpu/intel.c
@@ -227,6 +227,16 @@ static void __cpuinit init_intel(struct cpuinfo_x86 *c)
        if (cpu_has_bts)
                ds_init_intel(c);
+        /*
+         * See if we have a good local APIC by checking for buggy Pentia,
+         * i.e. all B steppings and the C2 stepping of P54C when using their
+         * integrated APIC (see 11AP erratum in "Pentium Processor
+         * Specification Update").
+         */
+        if (cpu_has_apic && (c->x86<<8 | c->x86_model<<4) == 0x520 &&
+            (c->x86_mask < 0x6 || c->x86_mask == 0xb))
+                set_cpu_cap(c, X86_FEATURE_11AP);
 #ifdef CONFIG_X86_NUMAQ
        numaq_tsc_disable();
 #endif
diff --git a/arch/x86/kernel/cpu/mcheck/p4.c b/arch/x86/kernel/cpu/mcheck/p4.c
index eef001ad3bde..9b60fce09f75 100644
--- a/arch/x86/kernel/cpu/mcheck/p4.c
+++ b/arch/x86/kernel/cpu/mcheck/p4.c
@@ -102,7 +102,7 @@ static void intel_init_thermal(struct cpuinfo_x86 *c)
        /* The temperature transition interrupt handler setup */
        h = THERMAL_APIC_VECTOR;                /* our delivery vector */
        h |= (APIC_DM_FIXED | APIC_LVT_MASKED); /* we'll mask till we're ready */
-        apic_write_around(APIC_LVTTHMR, h);
+        apic_write(APIC_LVTTHMR, h);
        rdmsr(MSR_IA32_THERM_INTERRUPT, l, h);
        wrmsr(MSR_IA32_THERM_INTERRUPT, l | 0x03 , h);
@@ -114,7 +114,7 @@ static void intel_init_thermal(struct cpuinfo_x86 *c)
        wrmsr(MSR_IA32_MISC_ENABLE, l | (1<<3), h);
        l = apic_read(APIC_LVTTHMR);
-        apic_write_around(APIC_LVTTHMR, l & ~APIC_LVT_MASKED);
+        apic_write(APIC_LVTTHMR, l & ~APIC_LVT_MASKED);
        printk(KERN_INFO "CPU%d: Thermal monitoring enabled\n", cpu);
        /* enable thermal throttle processing */
diff --git a/arch/x86/kernel/e820.c b/arch/x86/kernel/e820.c
index 28c29180b380..9af89078f7bb 100644
--- a/arch/x86/kernel/e820.c
+++ b/arch/x86/kernel/e820.c
@@ -877,7 +877,8 @@ void __init early_res_to_bootmem(u64 start, u64 end)
        for (i = 0; i < MAX_EARLY_RES && early_res[i].end; i++)
                count++;
-        printk(KERN_INFO "(%d early reservations) ==> bootmem\n", count);
+        printk(KERN_INFO "(%d early reservations) ==> bootmem [%010llx - %010llx]\n",
+                         count, start, end);
        for (i = 0; i < count; i++) {
                struct early_res *r = &early_res[i];
                printk(KERN_INFO "  #%d [%010llx - %010llx] %16s", i,
@@ -1298,11 +1299,6 @@ void __init e820_reserve_resources(void)
        }
 }
-/*
- * Non-standard memory setup can be specified via this quirk:
- */
-char * (*arch_memory_setup_quirk)(void);
 char *__init default_machine_specific_memory_setup(void)
 {
        char *who = "BIOS-e820";
@@ -1343,8 +1339,8 @@ char *__init default_machine_specific_memory_setup(void)
 char *__init __attribute__((weak)) machine_specific_memory_setup(void)
 {
-        if (arch_memory_setup_quirk) {
+        if (x86_quirks->arch_memory_setup) {
-                char *who = arch_memory_setup_quirk();
+                char *who = x86_quirks->arch_memory_setup();
                if (who)
                        return who;
@@ -1367,24 +1363,3 @@ void __init setup_memory_map(void)
        printk(KERN_INFO "BIOS-provided physical RAM map:\n");
        e820_print_map(who);
 }
-#ifdef CONFIG_X86_64
-int __init arch_get_ram_range(int slot, u64 *addr, u64 *size)
-{
-        int i;
-        if (slot < 0 || slot >= e820.nr_map)
-                return -1;
-        for (i = slot; i < e820.nr_map; i++) {
-                if (e820.map[i].type != E820_RAM)
-                        continue;
-                break;
-        }
-        if (i == e820.nr_map || e820.map[i].addr > (max_pfn << PAGE_SHIFT))
-                return -1;
-        *addr = e820.map[i].addr;
-        *size = min_t(u64, e820.map[i].size + e820.map[i].addr,
-                max_pfn << PAGE_SHIFT) - *addr;
-        return i + 1;
-}
-#endif
diff --git a/arch/x86/kernel/early-quirks.c b/arch/x86/kernel/early-quirks.c
index a0e11c0cc872..4353cf5e6fac 100644
--- a/arch/x86/kernel/early-quirks.c
+++ b/arch/x86/kernel/early-quirks.c
@@ -16,10 +16,7 @@
 #include <asm/dma.h>
 #include <asm/io_apic.h>
 #include <asm/apic.h>
+#include <asm/iommu.h>
-#ifdef CONFIG_GART_IOMMU
-#include <asm/gart.h>
-#endif
 static void __init fix_hypertransport_config(int num, int slot, int func)
 {
diff --git a/arch/x86/kernel/entry_32.S b/arch/x86/kernel/entry_32.S
index 6bc07f0f1202..cdfd94cc6b14 100644
--- a/arch/x86/kernel/entry_32.S
+++ b/arch/x86/kernel/entry_32.S
@@ -332,7 +332,7 @@ sysenter_past_esp:
        GET_THREAD_INFO(%ebp)
        /* Note, _TIF_SECCOMP is bit number 8, and so it needs testw and not testb */
-        testw $(_TIF_SYSCALL_EMU|_TIF_SYSCALL_TRACE|_TIF_SECCOMP|_TIF_SYSCALL_AUDIT),TI_flags(%ebp)
+        testw $_TIF_WORK_SYSCALL_ENTRY,TI_flags(%ebp)
        jnz syscall_trace_entry
        cmpl $(nr_syscalls), %eax
        jae syscall_badsys
@@ -370,7 +370,7 @@ ENTRY(system_call)
        GET_THREAD_INFO(%ebp)
                                        # system call tracing in operation / emulation
        /* Note, _TIF_SECCOMP is bit number 8, and so it needs testw and not testb */
-        testw $(_TIF_SYSCALL_EMU|_TIF_SYSCALL_TRACE|_TIF_SECCOMP|_TIF_SYSCALL_AUDIT),TI_flags(%ebp)
+        testw $_TIF_WORK_SYSCALL_ENTRY,TI_flags(%ebp)
        jnz syscall_trace_entry
        cmpl $(nr_syscalls), %eax
        jae syscall_badsys
@@ -383,10 +383,6 @@ syscall_exit:
                                        # setting need_resched or sigpending
                                        # between sampling and the iret
        TRACE_IRQS_OFF
-        testl $X86_EFLAGS_TF,PT_EFLAGS(%esp)    # If tracing set singlestep flag on exit
-        jz no_singlestep
-        orl $_TIF_SINGLESTEP,TI_flags(%ebp)
-no_singlestep:
        movl TI_flags(%ebp), %ecx
        testw $_TIF_ALLWORK_MASK, %cx   # current->work
        jne syscall_exit_work
@@ -514,12 +510,8 @@ END(work_pending)
 syscall_trace_entry:
        movl $-ENOSYS,PT_EAX(%esp)
        movl %esp, %eax
-        xorl %edx,%edx
+        call syscall_trace_enter
-        call do_syscall_trace
+        /* What it returned is what we'll actually use.  */
-        cmpl $0, %eax
-        jne resume_userspace            # ret != 0 -> running under PTRACE_SYSEMU,
-                                        # so must skip actual syscall
-        movl PT_ORIG_EAX(%esp), %eax
        cmpl $(nr_syscalls), %eax
        jnae syscall_call
        jmp syscall_exit
@@ -528,14 +520,13 @@ END(syscall_trace_entry)
        # perform syscall exit tracing
        ALIGN
 syscall_exit_work:
-        testb $(_TIF_SYSCALL_TRACE|_TIF_SYSCALL_AUDIT|_TIF_SINGLESTEP), %cl
+        testb $_TIF_WORK_SYSCALL_EXIT, %cl
        jz work_pending
        TRACE_IRQS_ON
-        ENABLE_INTERRUPTS(CLBR_ANY)     # could let do_syscall_trace() call
+        ENABLE_INTERRUPTS(CLBR_ANY)     # could let syscall_trace_leave() call
                                        # schedule() instead
        movl %esp, %eax
-        movl $1, %edx
+        call syscall_trace_leave
-        call do_syscall_trace
        jmp resume_userspace
 END(syscall_exit_work)
        CFI_ENDPROC
@@ -1024,6 +1015,7 @@ ENDPROC(kernel_thread_helper)
 ENTRY(xen_sysenter_target)
        RING0_INT_FRAME
        addl $5*4, %esp         /* remove xen-provided frame */
+        CFI_ADJUST_CFA_OFFSET -5*4
        jmp sysenter_past_esp
        CFI_ENDPROC
diff --git a/arch/x86/kernel/entry_64.S b/arch/x86/kernel/entry_64.S
index ae63e584c340..8410e26f4183 100644
--- a/arch/x86/kernel/entry_64.S
+++ b/arch/x86/kernel/entry_64.S
@@ -349,8 +349,7 @@ ENTRY(system_call_after_swapgs)
        movq  %rcx,RIP-ARGOFFSET(%rsp)
        CFI_REL_OFFSET rip,RIP-ARGOFFSET
        GET_THREAD_INFO(%rcx)
-        testl $(_TIF_SYSCALL_TRACE|_TIF_SYSCALL_AUDIT|_TIF_SECCOMP), \
+        testl $_TIF_WORK_SYSCALL_ENTRY,TI_flags(%rcx)
-                TI_flags(%rcx)
        jnz tracesys
        cmpq $__NR_syscall_max,%rax
        ja badsys
@@ -430,7 +429,12 @@ tracesys:
        FIXUP_TOP_OF_STACK %rdi
        movq %rsp,%rdi
        call syscall_trace_enter
-        LOAD_ARGS ARGOFFSET  /* reload args from stack in case ptrace changed it */
+        /*
+         * Reload arg registers from stack in case ptrace changed them.
+         * We don't reload %rax because syscall_trace_enter() returned
+         * the value it wants us to use in the table lookup.
+         */
+        LOAD_ARGS ARGOFFSET, 1
        RESTORE_REST
        cmpq $__NR_syscall_max,%rax
        ja   int_ret_from_sys_call      /* RAX(%rsp) set to -ENOSYS above */
@@ -483,7 +487,7 @@ int_very_careful:
        ENABLE_INTERRUPTS(CLBR_NONE)
        SAVE_REST
        /* Check for syscall exit trace */      
-        testl $(_TIF_SYSCALL_TRACE|_TIF_SYSCALL_AUDIT|_TIF_SINGLESTEP),%edx
+        testl $_TIF_WORK_SYSCALL_EXIT,%edx
        jz int_signal
        pushq %rdi
        CFI_ADJUST_CFA_OFFSET 8
@@ -491,7 +495,7 @@ int_very_careful:
        call syscall_trace_leave
        popq %rdi
        CFI_ADJUST_CFA_OFFSET -8
-        andl $~(_TIF_SYSCALL_TRACE|_TIF_SYSCALL_AUDIT|_TIF_SINGLESTEP),%edi
+        andl $~(_TIF_WORK_SYSCALL_EXIT|_TIF_SYSCALL_EMU),%edi
        jmp int_restore_rest
        
 int_signal:
@@ -1189,6 +1193,7 @@ END(device_not_available)
        /* runs on exception stack */
 KPROBE_ENTRY(debug)
        INTR_FRAME
+        PARAVIRT_ADJUST_EXCEPTION_FRAME
        pushq $0
        CFI_ADJUST_CFA_OFFSET 8         
        paranoidentry do_debug, DEBUG_STACK
@@ -1198,6 +1203,7 @@ KPROBE_END(debug)
        /* runs on exception stack */   
 KPROBE_ENTRY(nmi)
        INTR_FRAME
+        PARAVIRT_ADJUST_EXCEPTION_FRAME
        pushq $-1
        CFI_ADJUST_CFA_OFFSET 8
        paranoidentry do_nmi, 0, 0
@@ -1211,6 +1217,7 @@ KPROBE_END(nmi)
 KPROBE_ENTRY(int3)
        INTR_FRAME
+        PARAVIRT_ADJUST_EXCEPTION_FRAME
        pushq $0
        CFI_ADJUST_CFA_OFFSET 8
        paranoidentry do_int3, DEBUG_STACK
@@ -1237,6 +1244,7 @@ END(coprocessor_segment_overrun)
        /* runs on exception stack */
 ENTRY(double_fault)
        XCPT_FRAME
+        PARAVIRT_ADJUST_EXCEPTION_FRAME
        paranoidentry do_double_fault
        jmp paranoid_exit1
        CFI_ENDPROC
@@ -1253,6 +1261,7 @@ END(segment_not_present)
        /* runs on exception stack */
 ENTRY(stack_segment)
        XCPT_FRAME
+        PARAVIRT_ADJUST_EXCEPTION_FRAME
        paranoidentry do_stack_segment
        jmp paranoid_exit1
        CFI_ENDPROC
@@ -1278,6 +1287,7 @@ END(spurious_interrupt_bug)
        /* runs on exception stack */
 ENTRY(machine_check)
        INTR_FRAME
+        PARAVIRT_ADJUST_EXCEPTION_FRAME
        pushq $0
        CFI_ADJUST_CFA_OFFSET 8 
        paranoidentry do_machine_check
@@ -1312,3 +1322,103 @@ KPROBE_ENTRY(ignore_sysret)
        sysret
        CFI_ENDPROC
 ENDPROC(ignore_sysret)
+#ifdef CONFIG_XEN
+ENTRY(xen_hypervisor_callback)
+        zeroentry xen_do_hypervisor_callback
+END(xen_hypervisor_callback)
+/*
+# A note on the "critical region" in our callback handler.
+# We want to avoid stacking callback handlers due to events occurring
+# during handling of the last event. To do this, we keep events disabled
+# until we've done all processing. HOWEVER, we must enable events before
+# popping the stack frame (can't be done atomically) and so it would still
+# be possible to get enough handler activations to overflow the stack.
+# Although unlikely, bugs of that kind are hard to track down, so we'd
+# like to avoid the possibility.
+# So, on entry to the handler we detect whether we interrupted an
+# existing activation in its critical region -- if so, we pop the current
+# activation and restart the handler using the previous one.
+*/
+ENTRY(xen_do_hypervisor_callback)   # do_hypervisor_callback(struct *pt_regs)
+        CFI_STARTPROC
+/* Since we don't modify %rdi, evtchn_do_upall(struct *pt_regs) will
+   see the correct pointer to the pt_regs */
+        movq %rdi, %rsp            # we don't return, adjust the stack frame
+        CFI_ENDPROC
+        CFI_DEFAULT_STACK
+11:     incl %gs:pda_irqcount
+        movq %rsp,%rbp
+        CFI_DEF_CFA_REGISTER rbp
+        cmovzq %gs:pda_irqstackptr,%rsp
+        pushq %rbp                      # backlink for old unwinder
+        call xen_evtchn_do_upcall
+        popq %rsp
+        CFI_DEF_CFA_REGISTER rsp
+        decl %gs:pda_irqcount
+        jmp  error_exit
+        CFI_ENDPROC
+END(do_hypervisor_callback)
+/*
+# Hypervisor uses this for application faults while it executes.
+# We get here for two reasons:
+#  1. Fault while reloading DS, ES, FS or GS
+#  2. Fault while executing IRET
+# Category 1 we do not need to fix up as Xen has already reloaded all segment
+# registers that could be reloaded and zeroed the others.
+# Category 2 we fix up by killing the current process. We cannot use the
+# normal Linux return path in this case because if we use the IRET hypercall
+# to pop the stack frame we end up in an infinite loop of failsafe callbacks.
+# We distinguish between categories by comparing each saved segment register
+# with its current contents: any discrepancy means we in category 1.
+*/
+ENTRY(xen_failsafe_callback)
+        framesz = (RIP-0x30)    /* workaround buggy gas */
+        _frame framesz
+        CFI_REL_OFFSET rcx, 0
+        CFI_REL_OFFSET r11, 8
+        movw %ds,%cx
+        cmpw %cx,0x10(%rsp)
+        CFI_REMEMBER_STATE
+        jne 1f
+        movw %es,%cx
+        cmpw %cx,0x18(%rsp)
+        jne 1f
+        movw %fs,%cx
+        cmpw %cx,0x20(%rsp)
+        jne 1f
+        movw %gs,%cx
+        cmpw %cx,0x28(%rsp)
+        jne 1f
+        /* All segments match their saved values => Category 2 (Bad IRET). */
+        movq (%rsp),%rcx
+        CFI_RESTORE rcx
+        movq 8(%rsp),%r11
+        CFI_RESTORE r11
+        addq $0x30,%rsp
+        CFI_ADJUST_CFA_OFFSET -0x30
+        pushq $0
+        CFI_ADJUST_CFA_OFFSET 8
+        pushq %r11
+        CFI_ADJUST_CFA_OFFSET 8
+        pushq %rcx
+        CFI_ADJUST_CFA_OFFSET 8
+        jmp general_protection
+        CFI_RESTORE_STATE
+1:      /* Segment mismatch => Category 1 (Bad segment). Retry the IRET. */
+        movq (%rsp),%rcx
+        CFI_RESTORE rcx
+        movq 8(%rsp),%r11
+        CFI_RESTORE r11
+        addq $0x30,%rsp
+        CFI_ADJUST_CFA_OFFSET -0x30
+        pushq $0
+        CFI_ADJUST_CFA_OFFSET 8
+        SAVE_ALL
+        jmp error_exit
+        CFI_ENDPROC
+END(xen_failsafe_callback)
+#endif /* CONFIG_XEN */
diff --git a/arch/x86/kernel/genx2apic_uv_x.c b/arch/x86/kernel/genx2apic_uv_x.c
index 711f11c30b06..3c3929340692 100644
--- a/arch/x86/kernel/genx2apic_uv_x.c
+++ b/arch/x86/kernel/genx2apic_uv_x.c
@@ -24,6 +24,7 @@
 #include <asm/pgtable.h>
 #include <asm/uv/uv_mmrs.h>
 #include <asm/uv/uv_hub.h>
+#include <asm/uv/bios.h>
 DEFINE_PER_CPU(struct uv_hub_info_s, __uv_hub_info);
 EXPORT_PER_CPU_SYMBOL_GPL(__uv_hub_info);
@@ -40,6 +41,9 @@ EXPORT_SYMBOL_GPL(uv_cpu_to_blade);
 short uv_possible_blades;
 EXPORT_SYMBOL_GPL(uv_possible_blades);
+unsigned long sn_rtc_cycles_per_second;
+EXPORT_SYMBOL(sn_rtc_cycles_per_second);
 /* Start with all IRQs pointing to boot CPU.  IRQ balancing will shift them. */
 static cpumask_t uv_target_cpus(void)
@@ -272,6 +276,23 @@ static __init void map_mmioh_high(int max_pnode)
                map_high("MMIOH", mmioh.s.base, shift, map_uc);
 }
+static __init void uv_rtc_init(void)
+{
+        long status, ticks_per_sec, drift;
+        status =
+            x86_bios_freq_base(BIOS_FREQ_BASE_REALTIME_CLOCK, &ticks_per_sec,
+                                        &drift);
+        if (status != 0 || ticks_per_sec < 100000) {
+                printk(KERN_WARNING
+                        "unable to determine platform RTC clock frequency, "
+                        "guessing.\n");
+                /* BIOS gives wrong value for clock freq. so guess */
+                sn_rtc_cycles_per_second = 1000000000000UL / 30000UL;
+        } else
+                sn_rtc_cycles_per_second = ticks_per_sec;
+}
 static __init void uv_system_init(void)
 {
        union uvh_si_addr_map_config_u m_n_config;
@@ -326,6 +347,8 @@ static __init void uv_system_init(void)
        gnode_upper = (((unsigned long)node_id.s.node_id) &
                       ~((1 << n_val) - 1)) << m_val;
+        uv_rtc_init();
        for_each_present_cpu(cpu) {
                nid = cpu_to_node(cpu);
                pnode = uv_apicid_to_pnode(per_cpu(x86_cpu_to_apicid, cpu));
diff --git a/arch/x86/kernel/head64.c b/arch/x86/kernel/head64.c
index c97819829146..1b318e903bf6 100644
--- a/arch/x86/kernel/head64.c
+++ b/arch/x86/kernel/head64.c
@@ -39,6 +39,13 @@ static struct x8664_pda *__cpu_pda[NR_CPUS] __initdata;
 static struct x8664_pda *__cpu_pda[NR_CPUS] __read_mostly;
 #endif
+void __init x86_64_init_pda(void)
+{
+        _cpu_pda = __cpu_pda;
+        cpu_pda(0) = &_boot_cpu_pda;
+        pda_init(0);
+}
 static void __init zap_identity_mappings(void)
 {
        pgd_t *pgd = pgd_offset_k(0UL);
@@ -102,9 +109,7 @@ void __init x86_64_start_kernel(char * real_mode_data)
        early_printk("Kernel alive\n");
-        _cpu_pda = __cpu_pda;
+        x86_64_init_pda();
-        cpu_pda(0) = &_boot_cpu_pda;
-        pda_init(0);
        early_printk("Kernel really alive\n");
diff --git a/arch/x86/kernel/head_64.S b/arch/x86/kernel/head_64.S
index b07ac7b217cb..db3280afe886 100644
--- a/arch/x86/kernel/head_64.S
+++ b/arch/x86/kernel/head_64.S
@@ -407,6 +407,7 @@ ENTRY(phys_base)
        /* This must match the first entry in level2_kernel_pgt */
        .quad   0x0000000000000000
+#include "../../x86/xen/xen-head.S"
        
        .section .bss, "aw", @nobits
        .align L1_CACHE_BYTES
diff --git a/arch/x86/kernel/io_apic_32.c b/arch/x86/kernel/io_apic_32.c
index 558abf4c796a..de9aa0e3a9c5 100644
--- a/arch/x86/kernel/io_apic_32.c
+++ b/arch/x86/kernel/io_apic_32.c
@@ -756,7 +756,7 @@ void send_IPI_self(int vector)
        /*
         * Send the IPI. The write to APIC_ICR fires this off.
         */
-        apic_write_around(APIC_ICR, cfg);
+        apic_write(APIC_ICR, cfg);
 }
 #endif /* !CONFIG_SMP */
@@ -2030,7 +2030,7 @@ static void mask_lapic_irq(unsigned int irq)
        unsigned long v;
        v = apic_read(APIC_LVT0);
-        apic_write_around(APIC_LVT0, v | APIC_LVT_MASKED);
+        apic_write(APIC_LVT0, v | APIC_LVT_MASKED);
 }
 static void unmask_lapic_irq(unsigned int irq)
@@ -2038,7 +2038,7 @@ static void unmask_lapic_irq(unsigned int irq)
        unsigned long v;
        v = apic_read(APIC_LVT0);
-        apic_write_around(APIC_LVT0, v & ~APIC_LVT_MASKED);
+        apic_write(APIC_LVT0, v & ~APIC_LVT_MASKED);
 }
 static struct irq_chip lapic_chip __read_mostly = {
@@ -2168,7 +2168,7 @@ static inline void __init check_timer(void)
         * The AEOI mode will finish them in the 8259A
         * automatically.
         */
-        apic_write_around(APIC_LVT0, APIC_LVT_MASKED | APIC_DM_EXTINT);
+        apic_write(APIC_LVT0, APIC_LVT_MASKED | APIC_DM_EXTINT);
        init_8259A(1);
        timer_ack = (nmi_watchdog == NMI_IO_APIC && !APIC_INTEGRATED(ver));
@@ -2177,8 +2177,9 @@ static inline void __init check_timer(void)
        pin2  = ioapic_i8259.pin;
        apic2 = ioapic_i8259.apic;
-        printk(KERN_INFO "..TIMER: vector=0x%02X apic1=%d pin1=%d apic2=%d pin2=%d\n",
+        apic_printk(APIC_QUIET, KERN_INFO "..TIMER: vector=0x%02X "
-                vector, apic1, pin1, apic2, pin2);
+                    "apic1=%d pin1=%d apic2=%d pin2=%d\n",
+                    vector, apic1, pin1, apic2, pin2);
        /*
         * Some BIOS writers are clueless and report the ExtINTA
@@ -2216,12 +2217,13 @@ static inline void __init check_timer(void)
                }
                clear_IO_APIC_pin(apic1, pin1);
                if (!no_pin1)
-                        printk(KERN_ERR "..MP-BIOS bug: "
+                        apic_printk(APIC_QUIET, KERN_ERR "..MP-BIOS bug: "
-                               "8254 timer not connected to IO-APIC\n");
+                                    "8254 timer not connected to IO-APIC\n");
-                printk(KERN_INFO "...trying to set up timer (IRQ0) "
+                apic_printk(APIC_QUIET, KERN_INFO "...trying to set up timer "
-                       "through the 8259A ... ");
+                            "(IRQ0) through the 8259A ...\n");
-                printk("\n..... (found pin %d) ...", pin2);
+                apic_printk(APIC_QUIET, KERN_INFO
+                            "..... (found apic %d pin %d) ...\n", apic2, pin2);
                /*
                 * legacy devices should be connected to IO APIC #0
                 */
@@ -2230,7 +2232,7 @@ static inline void __init check_timer(void)
                unmask_IO_APIC_irq(0);
                enable_8259A_irq(0);
                if (timer_irq_works()) {
-                        printk("works.\n");
+                        apic_printk(APIC_QUIET, KERN_INFO "....... works.\n");
                        timer_through_8259 = 1;
                        if (nmi_watchdog == NMI_IO_APIC) {
                                disable_8259A_irq(0);
@@ -2244,44 +2246,47 @@ static inline void __init check_timer(void)
                 */
                disable_8259A_irq(0);
                clear_IO_APIC_pin(apic2, pin2);
-                printk(" failed.\n");
+                apic_printk(APIC_QUIET, KERN_INFO "....... failed.\n");
        }
        if (nmi_watchdog == NMI_IO_APIC) {
-                printk(KERN_WARNING "timer doesn't work through the IO-APIC - disabling NMI Watchdog!\n");
+                apic_printk(APIC_QUIET, KERN_WARNING "timer doesn't work "
+                            "through the IO-APIC - disabling NMI Watchdog!\n");
                nmi_watchdog = NMI_NONE;
        }
        timer_ack = 0;
-        printk(KERN_INFO "...trying to set up timer as Virtual Wire IRQ...");
+        apic_printk(APIC_QUIET, KERN_INFO
+                    "...trying to set up timer as Virtual Wire IRQ...\n");
        lapic_register_intr(0, vector);
-        apic_write_around(APIC_LVT0, APIC_DM_FIXED | vector);   /* Fixed mode */
+        apic_write(APIC_LVT0, APIC_DM_FIXED | vector);  /* Fixed mode */
        enable_8259A_irq(0);
        if (timer_irq_works()) {
-                printk(" works.\n");
+                apic_printk(APIC_QUIET, KERN_INFO "..... works.\n");
                goto out;
        }
        disable_8259A_irq(0);
-        apic_write_around(APIC_LVT0, APIC_LVT_MASKED | APIC_DM_FIXED | vector);
+        apic_write(APIC_LVT0, APIC_LVT_MASKED | APIC_DM_FIXED | vector);
-        printk(" failed.\n");
+        apic_printk(APIC_QUIET, KERN_INFO "..... failed.\n");
-        printk(KERN_INFO "...trying to set up timer as ExtINT IRQ...");
+        apic_printk(APIC_QUIET, KERN_INFO
+                    "...trying to set up timer as ExtINT IRQ...\n");
        init_8259A(0);
        make_8259A_irq(0);
-        apic_write_around(APIC_LVT0, APIC_DM_EXTINT);
+        apic_write(APIC_LVT0, APIC_DM_EXTINT);
        unlock_ExtINT_logic();
        if (timer_irq_works()) {
-                printk(" works.\n");
+                apic_printk(APIC_QUIET, KERN_INFO "..... works.\n");
                goto out;
        }
-        printk(" failed :(.\n");
+        apic_printk(APIC_QUIET, KERN_INFO "..... failed :(.\n");
        panic("IO-APIC + timer doesn't work!  Boot with apic=debug and send a "
-                "report.  Then try booting with the 'noapic' option");
+                "report.  Then try booting with the 'noapic' option.\n");
 out:
        local_irq_restore(flags);
 }
diff --git a/arch/x86/kernel/io_apic_64.c b/arch/x86/kernel/io_apic_64.c
index 6510cde36b35..64a46affd858 100644
--- a/arch/x86/kernel/io_apic_64.c
+++ b/arch/x86/kernel/io_apic_64.c
@@ -45,6 +45,7 @@
 #include <asm/proto.h>
 #include <asm/acpi.h>
 #include <asm/dma.h>
+#include <asm/i8259.h>
 #include <asm/nmi.h>
 #include <asm/msidef.h>
 #include <asm/hypertransport.h>
@@ -1696,8 +1697,9 @@ static inline void __init check_timer(void)
        pin2  = ioapic_i8259.pin;
        apic2 = ioapic_i8259.apic;
-        apic_printk(APIC_VERBOSE,KERN_INFO "..TIMER: vector=0x%02X apic1=%d pin1=%d apic2=%d pin2=%d\n",
+        apic_printk(APIC_QUIET, KERN_INFO "..TIMER: vector=0x%02X "
-                cfg->vector, apic1, pin1, apic2, pin2);
+                    "apic1=%d pin1=%d apic2=%d pin2=%d\n",
+                    cfg->vector, apic1, pin1, apic2, pin2);
        /*
         * Some BIOS writers are clueless and report the ExtINTA
@@ -1735,14 +1737,13 @@ static inline void __init check_timer(void)
                }
                clear_IO_APIC_pin(apic1, pin1);
                if (!no_pin1)
-                        apic_printk(APIC_QUIET,KERN_ERR "..MP-BIOS bug: "
+                        apic_printk(APIC_QUIET, KERN_ERR "..MP-BIOS bug: "
                                    "8254 timer not connected to IO-APIC\n");
-                apic_printk(APIC_VERBOSE,KERN_INFO
+                apic_printk(APIC_QUIET, KERN_INFO "...trying to set up timer "
-                        "...trying to set up timer (IRQ0) "
+                            "(IRQ0) through the 8259A ...\n");
-                        "through the 8259A ... ");
+                apic_printk(APIC_QUIET, KERN_INFO
-                apic_printk(APIC_VERBOSE,"\n..... (found apic %d pin %d) ...",
+                            "..... (found apic %d pin %d) ...\n", apic2, pin2);
-                        apic2, pin2);
                /*
                 * legacy devices should be connected to IO APIC #0
                 */
@@ -1751,7 +1752,7 @@ static inline void __init check_timer(void)
                unmask_IO_APIC_irq(0);
                enable_8259A_irq(0);
                if (timer_irq_works()) {
-                        apic_printk(APIC_VERBOSE," works.\n");
+                        apic_printk(APIC_QUIET, KERN_INFO "....... works.\n");
                        timer_through_8259 = 1;
                        if (nmi_watchdog == NMI_IO_APIC) {
                                disable_8259A_irq(0);
@@ -1765,29 +1766,32 @@ static inline void __init check_timer(void)
                 */
                disable_8259A_irq(0);
                clear_IO_APIC_pin(apic2, pin2);
-                apic_printk(APIC_VERBOSE," failed.\n");
+                apic_printk(APIC_QUIET, KERN_INFO "....... failed.\n");
        }
        if (nmi_watchdog == NMI_IO_APIC) {
-                printk(KERN_WARNING "timer doesn't work through the IO-APIC - disabling NMI Watchdog!\n");
+                apic_printk(APIC_QUIET, KERN_WARNING "timer doesn't work "
+                            "through the IO-APIC - disabling NMI Watchdog!\n");
                nmi_watchdog = NMI_NONE;
        }
-        apic_printk(APIC_VERBOSE, KERN_INFO "...trying to set up timer as Virtual Wire IRQ...");
+        apic_printk(APIC_QUIET, KERN_INFO
+                    "...trying to set up timer as Virtual Wire IRQ...\n");
        lapic_register_intr(0);
        apic_write(APIC_LVT0, APIC_DM_FIXED | cfg->vector);     /* Fixed mode */
        enable_8259A_irq(0);
        if (timer_irq_works()) {
-                apic_printk(APIC_VERBOSE," works.\n");
+                apic_printk(APIC_QUIET, KERN_INFO "..... works.\n");
                goto out;
        }
        disable_8259A_irq(0);
        apic_write(APIC_LVT0, APIC_LVT_MASKED | APIC_DM_FIXED | cfg->vector);
-        apic_printk(APIC_VERBOSE," failed.\n");
+        apic_printk(APIC_QUIET, KERN_INFO "..... failed.\n");
-        apic_printk(APIC_VERBOSE, KERN_INFO "...trying to set up timer as ExtINT IRQ...");
+        apic_printk(APIC_QUIET, KERN_INFO
+                    "...trying to set up timer as ExtINT IRQ...\n");
        init_8259A(0);
        make_8259A_irq(0);
@@ -1796,11 +1800,12 @@ static inline void __init check_timer(void)
        unlock_ExtINT_logic();
        if (timer_irq_works()) {
-                apic_printk(APIC_VERBOSE," works.\n");
+                apic_printk(APIC_QUIET, KERN_INFO "..... works.\n");
                goto out;
        }
-        apic_printk(APIC_VERBOSE," failed :(.\n");
+        apic_printk(APIC_QUIET, KERN_INFO "..... failed :(.\n");
-        panic("IO-APIC + timer doesn't work! Try using the 'noapic' kernel parameter\n");
+        panic("IO-APIC + timer doesn't work!  Boot with apic=debug and send a "
+                "report.  Then try booting with the 'noapic' option.\n");
 out:
        local_irq_restore(flags);
 }
diff --git a/arch/x86/kernel/io_delay.c b/arch/x86/kernel/io_delay.c
index 5921e5f0a640..1c3a66a67f83 100644
--- a/arch/x86/kernel/io_delay.c
+++ b/arch/x86/kernel/io_delay.c
@@ -103,6 +103,9 @@ void __init io_delay_init(void)
 static int __init io_delay_param(char *s)
 {
+        if (!s)
+                return -EINVAL;
        if (!strcmp(s, "0x80"))
                io_delay_type = CONFIG_IO_DELAY_TYPE_0X80;
        else if (!strcmp(s, "0xed"))
diff --git a/arch/x86/kernel/ipi.c b/arch/x86/kernel/ipi.c
index 9d98cda39ad9..3f7537b669d3 100644
--- a/arch/x86/kernel/ipi.c
+++ b/arch/x86/kernel/ipi.c
@@ -70,7 +70,7 @@ void __send_IPI_shortcut(unsigned int shortcut, int vector)
        /*
         * Send the IPI. The write to APIC_ICR fires this off.
         */
-        apic_write_around(APIC_ICR, cfg);
+        apic_write(APIC_ICR, cfg);
 }
 void send_IPI_self(int vector)
@@ -98,7 +98,7 @@ static inline void __send_IPI_dest_field(unsigned long mask, int vector)
         * prepare target chip field
         */
        cfg = __prepare_ICR2(mask);
-        apic_write_around(APIC_ICR2, cfg);
+        apic_write(APIC_ICR2, cfg);
        /*
         * program the ICR
@@ -108,7 +108,7 @@ static inline void __send_IPI_dest_field(unsigned long mask, int vector)
        /*
         * Send the IPI. The write to APIC_ICR fires this off.
         */
-        apic_write_around(APIC_ICR, cfg);
+        apic_write(APIC_ICR, cfg);
 }
 /*
diff --git a/arch/x86/kernel/irq_32.c b/arch/x86/kernel/irq_32.c
index 47a6f6f12478..1cf8c1fcc088 100644
--- a/arch/x86/kernel/irq_32.c
+++ b/arch/x86/kernel/irq_32.c
@@ -83,11 +83,8 @@ union irq_ctx {
 static union irq_ctx *hardirq_ctx[NR_CPUS] __read_mostly;
 static union irq_ctx *softirq_ctx[NR_CPUS] __read_mostly;
-static char softirq_stack[NR_CPUS * THREAD_SIZE]
+static char softirq_stack[NR_CPUS * THREAD_SIZE] __page_aligned_bss;
-                __attribute__((__section__(".bss.page_aligned")));
+static char hardirq_stack[NR_CPUS * THREAD_SIZE] __page_aligned_bss;
-static char hardirq_stack[NR_CPUS * THREAD_SIZE]
-                __attribute__((__section__(".bss.page_aligned")));
 static void call_on_stack(void *func, void *stack)
 {
diff --git a/arch/x86/kernel/kdebugfs.c b/arch/x86/kernel/kdebugfs.c
index c03205991718..f2d43bc75514 100644
--- a/arch/x86/kernel/kdebugfs.c
+++ b/arch/x86/kernel/kdebugfs.c
@@ -12,9 +12,13 @@
 #include <linux/init.h>
 #include <linux/io.h>
 #include <linux/mm.h>
+#include <linux/module.h>
 #include <asm/setup.h>
+struct dentry *arch_debugfs_dir;
+EXPORT_SYMBOL(arch_debugfs_dir);
 #ifdef CONFIG_DEBUG_BOOT_PARAMS
 struct setup_data_node {
        u64 paddr;
@@ -209,6 +213,10 @@ static int __init arch_kdebugfs_init(void)
 {
        int error = 0;
+        arch_debugfs_dir = debugfs_create_dir("x86", NULL);
+        if (!arch_debugfs_dir)
+                return -ENOMEM;
 #ifdef CONFIG_DEBUG_BOOT_PARAMS
        error = boot_params_kdebugfs_init();
 #endif
diff --git a/arch/x86/kernel/kprobes.c b/arch/x86/kernel/kprobes.c
index b8c6743a13da..43c019f85f0d 100644
--- a/arch/x86/kernel/kprobes.c
+++ b/arch/x86/kernel/kprobes.c
@@ -860,7 +860,6 @@ static int __kprobes post_kprobe_handler(struct pt_regs *regs)
        resume_execution(cur, regs, kcb);
        regs->flags |= kcb->kprobe_saved_flags;
-        trace_hardirqs_fixup_flags(regs->flags);
        if ((kcb->kprobe_status != KPROBE_REENTER) && cur->post_handler) {
                kcb->kprobe_status = KPROBE_HIT_SSDONE;
diff --git a/arch/x86/kernel/module_64.c b/arch/x86/kernel/module_64.c
index a888e67f5874..0e867676b5a5 100644
--- a/arch/x86/kernel/module_64.c
+++ b/arch/x86/kernel/module_64.c
@@ -150,7 +150,8 @@ int module_finalize(const Elf_Ehdr *hdr,
                    const Elf_Shdr *sechdrs,
                    struct module *me)
 {
-        const Elf_Shdr *s, *text = NULL, *alt = NULL, *locks = NULL;
+        const Elf_Shdr *s, *text = NULL, *alt = NULL, *locks = NULL,
+                *para = NULL;
        char *secstrings = (void *)hdr + sechdrs[hdr->e_shstrndx].sh_offset;
        for (s = sechdrs; s < sechdrs + hdr->e_shnum; s++) {
@@ -160,6 +161,8 @@ int module_finalize(const Elf_Ehdr *hdr,
                        alt = s;
                if (!strcmp(".smp_locks", secstrings + s->sh_name))
                        locks= s;
+                if (!strcmp(".parainstructions", secstrings + s->sh_name))
+                        para = s;
        }
        if (alt) {
@@ -175,6 +178,11 @@ int module_finalize(const Elf_Ehdr *hdr,
                                            tseg, tseg + text->sh_size);
        }
+        if (para) {
+                void *pseg = (void *)para->sh_addr;
+                apply_paravirt(pseg, pseg + para->sh_size);
+        }
        return module_bug_finalize(hdr, sechdrs, me);
 }
diff --git a/arch/x86/kernel/mpparse.c b/arch/x86/kernel/mpparse.c
index 3b25e49380c6..6ae005ccaed8 100644
--- a/arch/x86/kernel/mpparse.c
+++ b/arch/x86/kernel/mpparse.c
@@ -27,6 +27,7 @@
 #include <asm/bios_ebda.h>
 #include <asm/e820.h>
 #include <asm/trampoline.h>
+#include <asm/setup.h>
 #include <mach_apic.h>
 #ifdef CONFIG_X86_32
@@ -48,76 +49,6 @@ static int __init mpf_checksum(unsigned char *mp, int len)
        return sum & 0xFF;
 }
-#ifdef CONFIG_X86_NUMAQ
-int found_numaq;
-/*
- * Have to match translation table entries to main table entries by counter
- * hence the mpc_record variable .... can't see a less disgusting way of
- * doing this ....
- */
-struct mpc_config_translation {
-        unsigned char mpc_type;
-        unsigned char trans_len;
-        unsigned char trans_type;
-        unsigned char trans_quad;
-        unsigned char trans_global;
-        unsigned char trans_local;
-        unsigned short trans_reserved;
-};
-static int mpc_record;
-static struct mpc_config_translation *translation_table[MAX_MPC_ENTRY]
-    __cpuinitdata;
-static inline int generate_logical_apicid(int quad, int phys_apicid)
-{
-        return (quad << 4) + (phys_apicid ? phys_apicid << 1 : 1);
-}
-static inline int mpc_apic_id(struct mpc_config_processor *m,
-                        struct mpc_config_translation *translation_record)
-{
-        int quad = translation_record->trans_quad;
-        int logical_apicid = generate_logical_apicid(quad, m->mpc_apicid);
-        printk(KERN_DEBUG "Processor #%d %u:%u APIC version %d (quad %d, apic %d)\n",
-               m->mpc_apicid,
-               (m->mpc_cpufeature & CPU_FAMILY_MASK) >> 8,
-               (m->mpc_cpufeature & CPU_MODEL_MASK) >> 4,
-               m->mpc_apicver, quad, logical_apicid);
-        return logical_apicid;
-}
-int mp_bus_id_to_node[MAX_MP_BUSSES];
-int mp_bus_id_to_local[MAX_MP_BUSSES];
-static void mpc_oem_bus_info(struct mpc_config_bus *m, char *name,
-        struct mpc_config_translation *translation)
-{
-        int quad = translation->trans_quad;
-        int local = translation->trans_local;
-        mp_bus_id_to_node[m->mpc_busid] = quad;
-        mp_bus_id_to_local[m->mpc_busid] = local;
-        printk(KERN_INFO "Bus #%d is %s (node %d)\n",
-               m->mpc_busid, name, quad);
-}
-int quad_local_to_mp_bus_id [NR_CPUS/4][4];
-static void mpc_oem_pci_bus(struct mpc_config_bus *m,
-        struct mpc_config_translation *translation)
-{
-        int quad = translation->trans_quad;
-        int local = translation->trans_local;
-        quad_local_to_mp_bus_id[quad][local] = m->mpc_busid;
-}
-#endif
 static void __cpuinit MP_processor_info(struct mpc_config_processor *m)
 {
        int apicid;
@@ -127,14 +58,12 @@ static void __cpuinit MP_processor_info(struct mpc_config_processor *m)
                disabled_cpus++;
                return;
        }
-#ifdef CONFIG_X86_NUMAQ
-        if (found_numaq)
+        if (x86_quirks->mpc_apic_id)
-                apicid = mpc_apic_id(m, translation_table[mpc_record]);
+                apicid = x86_quirks->mpc_apic_id(m);
        else
                apicid = m->mpc_apicid;
-#else
-        apicid = m->mpc_apicid;
-#endif
        if (m->mpc_cpuflag & CPU_BOOTPROCESSOR) {
                bootup_cpu = " (Bootup-CPU)";
                boot_cpu_physical_apicid = m->mpc_apicid;
@@ -151,12 +80,10 @@ static void __init MP_bus_info(struct mpc_config_bus *m)
        memcpy(str, m->mpc_bustype, 6);
        str[6] = 0;
-#ifdef CONFIG_X86_NUMAQ
+        if (x86_quirks->mpc_oem_bus_info)
-        if (found_numaq)
+                x86_quirks->mpc_oem_bus_info(m, str);
-                mpc_oem_bus_info(m, str, translation_table[mpc_record]);
+        else
-#else
+                printk(KERN_INFO "Bus #%d is %s\n", m->mpc_busid, str);
-        printk(KERN_INFO "Bus #%d is %s\n", m->mpc_busid, str);
-#endif
 #if MAX_MP_BUSSES < 256
        if (m->mpc_busid >= MAX_MP_BUSSES) {
@@ -173,10 +100,9 @@ static void __init MP_bus_info(struct mpc_config_bus *m)
                mp_bus_id_to_type[m->mpc_busid] = MP_BUS_ISA;
 #endif
        } else if (strncmp(str, BUSTYPE_PCI, sizeof(BUSTYPE_PCI) - 1) == 0) {
-#ifdef CONFIG_X86_NUMAQ
+                if (x86_quirks->mpc_oem_pci_bus)
-                if (found_numaq)
+                        x86_quirks->mpc_oem_pci_bus(m);
-                        mpc_oem_pci_bus(m, translation_table[mpc_record]);
-#endif
                clear_bit(m->mpc_busid, mp_bus_not_pci);
 #if defined(CONFIG_EISA) || defined (CONFIG_MCA)
                mp_bus_id_to_type[m->mpc_busid] = MP_BUS_PCI;
@@ -316,83 +242,6 @@ static void __init MP_lintsrc_info(struct mpc_config_lintsrc *m)
                m->mpc_srcbusirq, m->mpc_destapic, m->mpc_destapiclint);
 }
-#ifdef CONFIG_X86_NUMAQ
-static void __init MP_translation_info(struct mpc_config_translation *m)
-{
-        printk(KERN_INFO
-               "Translation: record %d, type %d, quad %d, global %d, local %d\n",
-               mpc_record, m->trans_type, m->trans_quad, m->trans_global,
-               m->trans_local);
-        if (mpc_record >= MAX_MPC_ENTRY)
-                printk(KERN_ERR "MAX_MPC_ENTRY exceeded!\n");
-        else
-                translation_table[mpc_record] = m;      /* stash this for later */
-        if (m->trans_quad < MAX_NUMNODES && !node_online(m->trans_quad))
-                node_set_online(m->trans_quad);
-}
-/*
- * Read/parse the MPC oem tables
- */
-static void __init smp_read_mpc_oem(struct mp_config_oemtable *oemtable,
-                                    unsigned short oemsize)
-{
-        int count = sizeof(*oemtable);  /* the header size */
-        unsigned char *oemptr = ((unsigned char *)oemtable) + count;
-        mpc_record = 0;
-        printk(KERN_INFO "Found an OEM MPC table at %8p - parsing it ... \n",
-               oemtable);
-        if (memcmp(oemtable->oem_signature, MPC_OEM_SIGNATURE, 4)) {
-                printk(KERN_WARNING
-                       "SMP mpc oemtable: bad signature [%c%c%c%c]!\n",
-                       oemtable->oem_signature[0], oemtable->oem_signature[1],
-                       oemtable->oem_signature[2], oemtable->oem_signature[3]);
-                return;
-        }
-        if (mpf_checksum((unsigned char *)oemtable, oemtable->oem_length)) {
-                printk(KERN_WARNING "SMP oem mptable: checksum error!\n");
-                return;
-        }
-        while (count < oemtable->oem_length) {
-                switch (*oemptr) {
-                case MP_TRANSLATION:
-                        {
-                                struct mpc_config_translation *m =
-                                    (struct mpc_config_translation *)oemptr;
-                                MP_translation_info(m);
-                                oemptr += sizeof(*m);
-                                count += sizeof(*m);
-                                ++mpc_record;
-                                break;
-                        }
-                default:
-                        {
-                                printk(KERN_WARNING
-                                       "Unrecognised OEM table entry type! - %d\n",
-                                       (int)*oemptr);
-                                return;
-                        }
-                }
-        }
-}
-void numaq_mps_oem_check(struct mp_config_table *mpc, char *oem,
-                                 char *productid)
-{
-        if (strncmp(oem, "IBM NUMA", 8))
-                printk("Warning!  Not a NUMA-Q system!\n");
-        else
-                found_numaq = 1;
-        if (mpc->mpc_oemptr)
-                smp_read_mpc_oem((struct mp_config_oemtable *)mpc->mpc_oemptr,
-                                 mpc->mpc_oemsize);
-}
-#endif /* CONFIG_X86_NUMAQ */
 /*
 * Read/parse the MPC
 */
@@ -457,7 +306,6 @@ static int __init smp_read_mpc(struct mp_config_table *mpc, unsigned early)
        } else
                mps_oem_check(mpc, oem, str);
 #endif
        /* save the local APIC address, it might be non-default */
        if (!acpi_lapic)
                mp_lapic_addr = mpc->mpc_lapic;
@@ -465,12 +313,17 @@ static int __init smp_read_mpc(struct mp_config_table *mpc, unsigned early)
        if (early)
                return 1;
+        if (mpc->mpc_oemptr && x86_quirks->smp_read_mpc_oem) {
+                struct mp_config_oemtable *oem_table = (struct mp_config_oemtable *)(unsigned long)mpc->mpc_oemptr;
+                x86_quirks->smp_read_mpc_oem(oem_table, mpc->mpc_oemsize);
+        }
        /*
         *      Now process the configuration blocks.
         */
-#ifdef CONFIG_X86_NUMAQ
+        if (x86_quirks->mpc_record)
-        mpc_record = 0;
+                *x86_quirks->mpc_record = 0;
-#endif
        while (count < mpc->mpc_length) {
                switch (*mpt) {
                case MP_PROCESSOR:
@@ -536,9 +389,8 @@ static int __init smp_read_mpc(struct mp_config_table *mpc, unsigned early)
                        count = mpc->mpc_length;
                        break;
                }
-#ifdef CONFIG_X86_NUMAQ
+                if (x86_quirks->mpc_record)
-                ++mpc_record;
+                        (*x86_quirks->mpc_record)++;
-#endif
        }
 #ifdef CONFIG_X86_GENERICARCH
@@ -726,20 +578,14 @@ static inline void __init construct_default_ISA_mptable(int mpc_default_type)
 static struct intel_mp_floating *mpf_found;
 /*
- * Machine specific quirk for finding the SMP config before other setup
- * activities destroy the table:
- */
-int (*mach_get_smp_config_quirk)(unsigned int early);
-/*
 * Scan the memory blocks for an SMP configuration block.
 */
 static void __init __get_smp_config(unsigned int early)
 {
        struct intel_mp_floating *mpf = mpf_found;
-        if (mach_get_smp_config_quirk) {
+        if (x86_quirks->mach_get_smp_config) {
-                if (mach_get_smp_config_quirk(early))
+                if (x86_quirks->mach_get_smp_config(early))
                        return;
        }
        if (acpi_lapic && early)
@@ -899,14 +745,12 @@ static int __init smp_scan_config(unsigned long base, unsigned long length,
        return 0;
 }
-int (*mach_find_smp_config_quirk)(unsigned int reserve);
 static void __init __find_smp_config(unsigned int reserve)
 {
        unsigned int address;
-        if (mach_find_smp_config_quirk) {
+        if (x86_quirks->mach_find_smp_config) {
-                if (mach_find_smp_config_quirk(reserve))
+                if (x86_quirks->mach_find_smp_config(reserve))
                        return;
        }
        /*
diff --git a/arch/x86/kernel/nmi.c b/arch/x86/kernel/nmi.c
index ec024b3baad0..ac6d51222e7d 100644
--- a/arch/x86/kernel/nmi.c
+++ b/arch/x86/kernel/nmi.c
@@ -263,7 +263,7 @@ late_initcall(init_lapic_nmi_sysfs);
 static void __acpi_nmi_enable(void *__unused)
 {
-        apic_write_around(APIC_LVT0, APIC_DM_NMI);
+        apic_write(APIC_LVT0, APIC_DM_NMI);
 }
 /*
@@ -277,7 +277,7 @@ void acpi_nmi_enable(void)
 static void __acpi_nmi_disable(void *__unused)
 {
-        apic_write_around(APIC_LVT0, APIC_DM_NMI | APIC_LVT_MASKED);
+        apic_write(APIC_LVT0, APIC_DM_NMI | APIC_LVT_MASKED);
 }
 /*
@@ -448,6 +448,13 @@ nmi_watchdog_tick(struct pt_regs *regs, unsigned reason)
 #ifdef CONFIG_SYSCTL
+static int __init setup_unknown_nmi_panic(char *str)
+{
+        unknown_nmi_panic = 1;
+        return 1;
+}
+__setup("unknown_nmi_panic", setup_unknown_nmi_panic);
 static int unknown_nmi_panic_callback(struct pt_regs *regs, int cpu)
 {
        unsigned char reason = get_nmi_reason();
diff --git a/arch/x86/kernel/numaq_32.c b/arch/x86/kernel/numaq_32.c
index a23e8233b9ac..b8c45610b20a 100644
--- a/arch/x86/kernel/numaq_32.c
+++ b/arch/x86/kernel/numaq_32.c
@@ -33,6 +33,7 @@
 #include <asm/processor.h>
 #include <asm/mpspec.h>
 #include <asm/e820.h>
+#include <asm/setup.h>
 #define MB_TO_PAGES(addr) ((addr) << (20 - PAGE_SHIFT))
@@ -71,6 +72,188 @@ static void __init smp_dump_qct(void)
        }
 }
+void __init numaq_tsc_disable(void)
+{
+        if (!found_numaq)
+                return;
+        if (num_online_nodes() > 1) {
+                printk(KERN_DEBUG "NUMAQ: disabling TSC\n");
+                setup_clear_cpu_cap(X86_FEATURE_TSC);
+        }
+}
+static int __init numaq_pre_time_init(void)
+{
+        numaq_tsc_disable();
+        return 0;
+}
+int found_numaq;
+/*
+ * Have to match translation table entries to main table entries by counter
+ * hence the mpc_record variable .... can't see a less disgusting way of
+ * doing this ....
+ */
+struct mpc_config_translation {
+        unsigned char mpc_type;
+        unsigned char trans_len;
+        unsigned char trans_type;
+        unsigned char trans_quad;
+        unsigned char trans_global;
+        unsigned char trans_local;
+        unsigned short trans_reserved;
+};
+/* x86_quirks member */
+static int mpc_record;
+static struct mpc_config_translation *translation_table[MAX_MPC_ENTRY]
+    __cpuinitdata;
+static inline int generate_logical_apicid(int quad, int phys_apicid)
+{
+        return (quad << 4) + (phys_apicid ? phys_apicid << 1 : 1);
+}
+/* x86_quirks member */
+static int mpc_apic_id(struct mpc_config_processor *m)
+{
+        int quad = translation_table[mpc_record]->trans_quad;
+        int logical_apicid = generate_logical_apicid(quad, m->mpc_apicid);
+        printk(KERN_DEBUG "Processor #%d %u:%u APIC version %d (quad %d, apic %d)\n",
+               m->mpc_apicid,
+               (m->mpc_cpufeature & CPU_FAMILY_MASK) >> 8,
+               (m->mpc_cpufeature & CPU_MODEL_MASK) >> 4,
+               m->mpc_apicver, quad, logical_apicid);
+        return logical_apicid;
+}
+int mp_bus_id_to_node[MAX_MP_BUSSES];
+int mp_bus_id_to_local[MAX_MP_BUSSES];
+/* x86_quirks member */
+static void mpc_oem_bus_info(struct mpc_config_bus *m, char *name)
+{
+        int quad = translation_table[mpc_record]->trans_quad;
+        int local = translation_table[mpc_record]->trans_local;
+        mp_bus_id_to_node[m->mpc_busid] = quad;
+        mp_bus_id_to_local[m->mpc_busid] = local;
+        printk(KERN_INFO "Bus #%d is %s (node %d)\n",
+               m->mpc_busid, name, quad);
+}
+int quad_local_to_mp_bus_id [NR_CPUS/4][4];
+/* x86_quirks member */
+static void mpc_oem_pci_bus(struct mpc_config_bus *m)
+{
+        int quad = translation_table[mpc_record]->trans_quad;
+        int local = translation_table[mpc_record]->trans_local;
+        quad_local_to_mp_bus_id[quad][local] = m->mpc_busid;
+}
+static void __init MP_translation_info(struct mpc_config_translation *m)
+{
+        printk(KERN_INFO
+               "Translation: record %d, type %d, quad %d, global %d, local %d\n",
+               mpc_record, m->trans_type, m->trans_quad, m->trans_global,
+               m->trans_local);
+        if (mpc_record >= MAX_MPC_ENTRY)
+                printk(KERN_ERR "MAX_MPC_ENTRY exceeded!\n");
+        else
+                translation_table[mpc_record] = m;      /* stash this for later */
+        if (m->trans_quad < MAX_NUMNODES && !node_online(m->trans_quad))
+                node_set_online(m->trans_quad);
+}
+static int __init mpf_checksum(unsigned char *mp, int len)
+{
+        int sum = 0;
+        while (len--)
+                sum += *mp++;
+        return sum & 0xFF;
+}
+/*
+ * Read/parse the MPC oem tables
+ */
+static void __init smp_read_mpc_oem(struct mp_config_oemtable *oemtable,
+                                    unsigned short oemsize)
+{
+        int count = sizeof(*oemtable);  /* the header size */
+        unsigned char *oemptr = ((unsigned char *)oemtable) + count;
+        mpc_record = 0;
+        printk(KERN_INFO "Found an OEM MPC table at %8p - parsing it ... \n",
+               oemtable);
+        if (memcmp(oemtable->oem_signature, MPC_OEM_SIGNATURE, 4)) {
+                printk(KERN_WARNING
+                       "SMP mpc oemtable: bad signature [%c%c%c%c]!\n",
+                       oemtable->oem_signature[0], oemtable->oem_signature[1],
+                       oemtable->oem_signature[2], oemtable->oem_signature[3]);
+                return;
+        }
+        if (mpf_checksum((unsigned char *)oemtable, oemtable->oem_length)) {
+                printk(KERN_WARNING "SMP oem mptable: checksum error!\n");
+                return;
+        }
+        while (count < oemtable->oem_length) {
+                switch (*oemptr) {
+                case MP_TRANSLATION:
+                        {
+                                struct mpc_config_translation *m =
+                                    (struct mpc_config_translation *)oemptr;
+                                MP_translation_info(m);
+                                oemptr += sizeof(*m);
+                                count += sizeof(*m);
+                                ++mpc_record;
+                                break;
+                        }
+                default:
+                        {
+                                printk(KERN_WARNING
+                                       "Unrecognised OEM table entry type! - %d\n",
+                                       (int)*oemptr);
+                                return;
+                        }
+                }
+        }
+}
+static struct x86_quirks numaq_x86_quirks __initdata = {
+        .arch_pre_time_init     = numaq_pre_time_init,
+        .arch_time_init         = NULL,
+        .arch_pre_intr_init     = NULL,
+        .arch_memory_setup      = NULL,
+        .arch_intr_init         = NULL,
+        .arch_trap_init         = NULL,
+        .mach_get_smp_config    = NULL,
+        .mach_find_smp_config   = NULL,
+        .mpc_record             = &mpc_record,
+        .mpc_apic_id            = mpc_apic_id,
+        .mpc_oem_bus_info       = mpc_oem_bus_info,
+        .mpc_oem_pci_bus        = mpc_oem_pci_bus,
+        .smp_read_mpc_oem       = smp_read_mpc_oem,
+};
+void numaq_mps_oem_check(struct mp_config_table *mpc, char *oem,
+                                 char *productid)
+{
+        if (strncmp(oem, "IBM NUMA", 8))
+                printk("Warning!  Not a NUMA-Q system!\n");
+        else
+                found_numaq = 1;
+}
 static __init void early_check_numaq(void)
 {
        /*
@@ -82,6 +265,9 @@ static __init void early_check_numaq(void)
         */
        if (smp_found_config)
                early_get_smp_config();
+        if (found_numaq)
+                x86_quirks = &numaq_x86_quirks;
 }
 int __init get_memcfg_numaq(void)
@@ -92,14 +278,3 @@ int __init get_memcfg_numaq(void)
        smp_dump_qct();
        return 1;
 }
-void __init numaq_tsc_disable(void)
-{
-        if (!found_numaq)
-                return;
-        if (num_online_nodes() > 1) {
-                printk(KERN_DEBUG "NUMAQ: disabling TSC\n");
-                setup_clear_cpu_cap(X86_FEATURE_TSC);
-        }
-}
diff --git a/arch/x86/kernel/paravirt.c b/arch/x86/kernel/paravirt.c
index e0f571d58c19..b4564d089b43 100644
--- a/arch/x86/kernel/paravirt.c
+++ b/arch/x86/kernel/paravirt.c
@@ -29,6 +29,7 @@
 #include <asm/desc.h>
 #include <asm/setup.h>
 #include <asm/arch_hooks.h>
+#include <asm/pgtable.h>
 #include <asm/time.h>
 #include <asm/pgalloc.h>
 #include <asm/irq.h>
@@ -361,7 +362,6 @@ struct pv_cpu_ops pv_cpu_ops = {
 struct pv_apic_ops pv_apic_ops = {
 #ifdef CONFIG_X86_LOCAL_APIC
        .apic_write = native_apic_write,
-        .apic_write_atomic = native_apic_write_atomic,
        .apic_read = native_apic_read,
        .setup_boot_clock = setup_boot_APIC_clock,
        .setup_secondary_clock = setup_secondary_APIC_clock,
@@ -373,6 +373,9 @@ struct pv_mmu_ops pv_mmu_ops = {
 #ifndef CONFIG_X86_64
        .pagetable_setup_start = native_pagetable_setup_start,
        .pagetable_setup_done = native_pagetable_setup_done,
+#else
+        .pagetable_setup_start = paravirt_nop,
+        .pagetable_setup_done = paravirt_nop,
 #endif
        .read_cr2 = native_read_cr2,
diff --git a/arch/x86/kernel/pci-calgary_64.c b/arch/x86/kernel/pci-calgary_64.c
index 6959b5c45df4..151f2d171f7c 100644
--- a/arch/x86/kernel/pci-calgary_64.c
+++ b/arch/x86/kernel/pci-calgary_64.c
@@ -36,7 +36,7 @@
 #include <linux/delay.h>
 #include <linux/scatterlist.h>
 #include <linux/iommu-helper.h>
-#include <asm/gart.h>
+#include <asm/iommu.h>
 #include <asm/calgary.h>
 #include <asm/tce.h>
 #include <asm/pci-direct.h>
diff --git a/arch/x86/kernel/pci-dma.c b/arch/x86/kernel/pci-dma.c
index 8467ec2320f1..a4213c00dffc 100644
--- a/arch/x86/kernel/pci-dma.c
+++ b/arch/x86/kernel/pci-dma.c
@@ -5,12 +5,11 @@
 #include <asm/proto.h>
 #include <asm/dma.h>
-#include <asm/gart.h>
+#include <asm/iommu.h>
 #include <asm/calgary.h>
 #include <asm/amd_iommu.h>
-int forbid_dac __read_mostly;
+static int forbid_dac __read_mostly;
-EXPORT_SYMBOL(forbid_dac);
 const struct dma_mapping_ops *dma_ops;
 EXPORT_SYMBOL(dma_ops);
@@ -114,21 +113,15 @@ void __init pci_iommu_alloc(void)
         * The order of these functions is important for
         * fall-back/fail-over reasons
         */
-#ifdef CONFIG_GART_IOMMU
        gart_iommu_hole_init();
-#endif
-#ifdef CONFIG_CALGARY_IOMMU
        detect_calgary();
-#endif
        detect_intel_iommu();
        amd_iommu_detect();
-#ifdef CONFIG_SWIOTLB
        pci_swiotlb_init();
-#endif
 }
 #endif
@@ -184,9 +177,7 @@ static __init int iommu_setup(char *p)
                        swiotlb = 1;
 #endif
-#ifdef CONFIG_GART_IOMMU
                gart_parse_options(p);
-#endif
 #ifdef CONFIG_CALGARY_IOMMU
                if (!strncmp(p, "calgary", 7))
@@ -500,17 +491,13 @@ EXPORT_SYMBOL(dma_free_coherent);
 static int __init pci_iommu_init(void)
 {
-#ifdef CONFIG_CALGARY_IOMMU
        calgary_iommu_init();
-#endif
        intel_iommu_init();
        amd_iommu_init();
-#ifdef CONFIG_GART_IOMMU
        gart_iommu_init();
-#endif
        no_iommu_init();
        return 0;
diff --git a/arch/x86/kernel/pci-gart_64.c b/arch/x86/kernel/pci-gart_64.c
index c3fe78406d18..be60961f8695 100644
--- a/arch/x86/kernel/pci-gart_64.c
+++ b/arch/x86/kernel/pci-gart_64.c
@@ -32,6 +32,7 @@
 #include <asm/mtrr.h>
 #include <asm/pgtable.h>
 #include <asm/proto.h>
+#include <asm/iommu.h>
 #include <asm/gart.h>
 #include <asm/cacheflush.h>
 #include <asm/swiotlb.h>
diff --git a/arch/x86/kernel/pci-nommu.c b/arch/x86/kernel/pci-nommu.c
index aec43d56f49c..792b9179eff3 100644
--- a/arch/x86/kernel/pci-nommu.c
+++ b/arch/x86/kernel/pci-nommu.c
@@ -7,7 +7,7 @@
 #include <linux/dma-mapping.h>
 #include <linux/scatterlist.h>
-#include <asm/gart.h>
+#include <asm/iommu.h>
 #include <asm/processor.h>
 #include <asm/dma.h>
diff --git a/arch/x86/kernel/pci-swiotlb_64.c b/arch/x86/kernel/pci-swiotlb_64.c
index 82299cd1d04d..20df839b9c20 100644
--- a/arch/x86/kernel/pci-swiotlb_64.c
+++ b/arch/x86/kernel/pci-swiotlb_64.c
@@ -5,7 +5,7 @@
 #include <linux/module.h>
 #include <linux/dma-mapping.h>
-#include <asm/gart.h>
+#include <asm/iommu.h>
 #include <asm/swiotlb.h>
 #include <asm/dma.h>
diff --git a/arch/x86/kernel/process.c b/arch/x86/kernel/process.c
index 4d629c62f4f8..7fc4d5b0a6a0 100644
--- a/arch/x86/kernel/process.c
+++ b/arch/x86/kernel/process.c
@@ -15,6 +15,7 @@ unsigned long idle_nomwait;
 EXPORT_SYMBOL(idle_nomwait);
 struct kmem_cache *task_xstate_cachep;
+static int force_mwait __cpuinitdata;
 int arch_dup_task_struct(struct task_struct *dst, struct task_struct *src)
 {
@@ -199,6 +200,7 @@ static void poll_idle(void)
 *
 * idle=mwait overrides this decision and forces the usage of mwait.
 */
+static int __cpuinitdata force_mwait;
 #define MWAIT_INFO                      0x05
 #define MWAIT_ECX_EXTENDED_INFO         0x01
@@ -326,6 +328,9 @@ void __cpuinit select_idle_routine(const struct cpuinfo_x86 *c)
 static int __init idle_setup(char *str)
 {
+        if (!str)
+                return -EINVAL;
        if (!strcmp(str, "poll")) {
                printk("using polling idle threads.\n");
                pm_idle = poll_idle;
diff --git a/arch/x86/kernel/process_64.c b/arch/x86/kernel/process_64.c
index a8e53626ac9a..e8a8e1b99817 100644
--- a/arch/x86/kernel/process_64.c
+++ b/arch/x86/kernel/process_64.c
@@ -537,8 +537,8 @@ static inline void __switch_to_xtra(struct task_struct *prev_p,
 struct task_struct *
 __switch_to(struct task_struct *prev_p, struct task_struct *next_p)
 {
-        struct thread_struct *prev = &prev_p->thread,
+        struct thread_struct *prev = &prev_p->thread;
-                                 *next = &next_p->thread;
+        struct thread_struct *next = &next_p->thread;
        int cpu = smp_processor_id();
        struct tss_struct *tss = &per_cpu(init_tss, cpu);
        unsigned fsindex, gsindex;
@@ -586,35 +586,34 @@ __switch_to(struct task_struct *prev_p, struct task_struct *next_p)
        /* 
         * Switch FS and GS.
+         *
+         * Segment register != 0 always requires a reload.  Also
+         * reload when it has changed.  When prev process used 64bit
+         * base always reload to avoid an information leak.
         */
-        { 
+        if (unlikely(fsindex | next->fsindex | prev->fs)) {
-                /* segment register != 0 always requires a reload. 
+                loadsegment(fs, next->fsindex);
-                   also reload when it has changed. 
+                /* 
-                   when prev process used 64bit base always reload
+                 * Check if the user used a selector != 0; if yes
-                   to avoid an information leak. */
+                 *  clear 64bit base, since overloaded base is always
-                if (unlikely(fsindex | next->fsindex | prev->fs)) {
+                 *  mapped to the Null selector
-                        loadsegment(fs, next->fsindex);
+                 */
-                        /* check if the user used a selector != 0
+                if (fsindex)
-                         * if yes clear 64bit base, since overloaded base
-                         * is always mapped to the Null selector
-                         */
-                        if (fsindex)
                        prev->fs = 0;                           
-                }
+        }
-                /* when next process has a 64bit base use it */
+        /* when next process has a 64bit base use it */
-                if (next->fs) 
+        if (next->fs)
-                        wrmsrl(MSR_FS_BASE, next->fs); 
+                wrmsrl(MSR_FS_BASE, next->fs);
-                prev->fsindex = fsindex;
+        prev->fsindex = fsindex;
-                if (unlikely(gsindex | next->gsindex | prev->gs)) {
+        if (unlikely(gsindex | next->gsindex | prev->gs)) {
-                        load_gs_index(next->gsindex);
+                load_gs_index(next->gsindex);
-                        if (gsindex)
+                if (gsindex)
                        prev->gs = 0;                           
-                }
-                if (next->gs)
-                        wrmsrl(MSR_KERNEL_GS_BASE, next->gs); 
-                prev->gsindex = gsindex;
        }
+        if (next->gs)
+                wrmsrl(MSR_KERNEL_GS_BASE, next->gs);
+        prev->gsindex = gsindex;
        /* Must be after DS reload */
        unlazy_fpu(prev_p);
@@ -627,7 +626,8 @@ __switch_to(struct task_struct *prev_p, struct task_struct *next_p)
        write_pda(pcurrent, next_p); 
        write_pda(kernelstack,
-        (unsigned long)task_stack_page(next_p) + THREAD_SIZE - PDA_STACKOFFSET);
+                  (unsigned long)task_stack_page(next_p) +
+                  THREAD_SIZE - PDA_STACKOFFSET);
 #ifdef CONFIG_CC_STACKPROTECTOR
        write_pda(stack_canary, next_p->stack_canary);
        /*
diff --git a/arch/x86/kernel/ptrace.c b/arch/x86/kernel/ptrace.c
index 77040b6070e1..e37dccce85db 100644
--- a/arch/x86/kernel/ptrace.c
+++ b/arch/x86/kernel/ptrace.c
@@ -1357,8 +1357,6 @@ const struct user_regset_view *task_user_regset_view(struct task_struct *task)
 #endif
 }
-#ifdef CONFIG_X86_32
 void send_sigtrap(struct task_struct *tsk, struct pt_regs *regs, int error_code)
 {
        struct siginfo info;
@@ -1377,89 +1375,10 @@ void send_sigtrap(struct task_struct *tsk, struct pt_regs *regs, int error_code)
        force_sig_info(SIGTRAP, &info, tsk);
 }
-/* notification of system call entry/exit
- * - triggered by current->work.syscall_trace
- */
-int do_syscall_trace(struct pt_regs *regs, int entryexit)
-{
-        int is_sysemu = test_thread_flag(TIF_SYSCALL_EMU);
-        /*
-         * With TIF_SYSCALL_EMU set we want to ignore TIF_SINGLESTEP for syscall
-         * interception
-         */
-        int is_singlestep = !is_sysemu && test_thread_flag(TIF_SINGLESTEP);
-        int ret = 0;
-        /* do the secure computing check first */
-        if (!entryexit)
-                secure_computing(regs->orig_ax);
-        if (unlikely(current->audit_context)) {
-                if (entryexit)
-                        audit_syscall_exit(AUDITSC_RESULT(regs->ax),
-                                                regs->ax);
-                /* Debug traps, when using PTRACE_SINGLESTEP, must be sent only
-                 * on the syscall exit path. Normally, when TIF_SYSCALL_AUDIT is
-                 * not used, entry.S will call us only on syscall exit, not
-                 * entry; so when TIF_SYSCALL_AUDIT is used we must avoid
-                 * calling send_sigtrap() on syscall entry.
-                 *
-                 * Note that when PTRACE_SYSEMU_SINGLESTEP is used,
-                 * is_singlestep is false, despite his name, so we will still do
-                 * the correct thing.
-                 */
-                else if (is_singlestep)
-                        goto out;
-        }
-        if (!(current->ptrace & PT_PTRACED))
-                goto out;
-        /* If a process stops on the 1st tracepoint with SYSCALL_TRACE
-         * and then is resumed with SYSEMU_SINGLESTEP, it will come in
-         * here. We have to check this and return */
-        if (is_sysemu && entryexit)
-                return 0;
-        /* Fake a debug trap */
-        if (is_singlestep)
-                send_sigtrap(current, regs, 0);
-        if (!test_thread_flag(TIF_SYSCALL_TRACE) && !is_sysemu)
-                goto out;
-        /* the 0x80 provides a way for the tracing parent to distinguish
-           between a syscall stop and SIGTRAP delivery */
-        /* Note that the debugger could change the result of test_thread_flag!*/
-        ptrace_notify(SIGTRAP | ((current->ptrace & PT_TRACESYSGOOD) ? 0x80:0));
-        /*
-         * this isn't the same as continuing with a signal, but it will do
-         * for normal use.  strace only continues with a signal if the
-         * stopping signal is not SIGTRAP.  -brl
-         */
-        if (current->exit_code) {
-                send_sig(current->exit_code, current, 1);
-                current->exit_code = 0;
-        }
-        ret = is_sysemu;
-out:
-        if (unlikely(current->audit_context) && !entryexit)
-                audit_syscall_entry(AUDIT_ARCH_I386, regs->orig_ax,
-                                    regs->bx, regs->cx, regs->dx, regs->si);
-        if (ret == 0)
-                return 0;
-        regs->orig_ax = -1; /* force skip of syscall restarting */
-        if (unlikely(current->audit_context))
-                audit_syscall_exit(AUDITSC_RESULT(regs->ax), regs->ax);
-        return 1;
-}
-#else  /* CONFIG_X86_64 */
 static void syscall_trace(struct pt_regs *regs)
 {
+        if (!(current->ptrace & PT_PTRACED))
+                return;
 #if 0
        printk("trace %s ip %lx sp %lx ax %d origrax %d caller %lx tiflags %x ptrace %x\n",
@@ -1481,39 +1400,81 @@ static void syscall_trace(struct pt_regs *regs)
        }
 }
-asmlinkage void syscall_trace_enter(struct pt_regs *regs)
+#ifdef CONFIG_X86_32
+# define IS_IA32        1
+#elif defined CONFIG_IA32_EMULATION
+# define IS_IA32        test_thread_flag(TIF_IA32)
+#else
+# define IS_IA32        0
+#endif
+/*
+ * We must return the syscall number to actually look up in the table.
+ * This can be -1L to skip running any syscall at all.
+ */
+asmregparm long syscall_trace_enter(struct pt_regs *regs)
 {
+        long ret = 0;
+        /*
+         * If we stepped into a sysenter/syscall insn, it trapped in
+         * kernel mode; do_debug() cleared TF and set TIF_SINGLESTEP.
+         * If user-mode had set TF itself, then it's still clear from
+         * do_debug() and we need to set it again to restore the user
+         * state.  If we entered on the slow path, TF was already set.
+         */
+        if (test_thread_flag(TIF_SINGLESTEP))
+                regs->flags |= X86_EFLAGS_TF;
        /* do the secure computing check first */
        secure_computing(regs->orig_ax);
-        if (test_thread_flag(TIF_SYSCALL_TRACE)
+        if (unlikely(test_thread_flag(TIF_SYSCALL_EMU)))
-            && (current->ptrace & PT_PTRACED))
+                ret = -1L;
+        if (ret || test_thread_flag(TIF_SYSCALL_TRACE))
                syscall_trace(regs);
        if (unlikely(current->audit_context)) {
-                if (test_thread_flag(TIF_IA32)) {
+                if (IS_IA32)
                        audit_syscall_entry(AUDIT_ARCH_I386,
                                            regs->orig_ax,
                                            regs->bx, regs->cx,
                                            regs->dx, regs->si);
-                } else {
+#ifdef CONFIG_X86_64
+                else
                        audit_syscall_entry(AUDIT_ARCH_X86_64,
                                            regs->orig_ax,
                                            regs->di, regs->si,
                                            regs->dx, regs->r10);
-                }
+#endif
        }
+        return ret ?: regs->orig_ax;
 }
-asmlinkage void syscall_trace_leave(struct pt_regs *regs)
+asmregparm void syscall_trace_leave(struct pt_regs *regs)
 {
        if (unlikely(current->audit_context))
                audit_syscall_exit(AUDITSC_RESULT(regs->ax), regs->ax);
-        if ((test_thread_flag(TIF_SYSCALL_TRACE)
+        if (test_thread_flag(TIF_SYSCALL_TRACE))
-             || test_thread_flag(TIF_SINGLESTEP))
-            && (current->ptrace & PT_PTRACED))
                syscall_trace(regs);
-}
-#endif  /* CONFIG_X86_32 */
+        /*
+         * If TIF_SYSCALL_EMU is set, we only get here because of
+         * TIF_SINGLESTEP (i.e. this is PTRACE_SYSEMU_SINGLESTEP).
+         * We already reported this syscall instruction in
+         * syscall_trace_enter(), so don't do any more now.
+         */
+        if (unlikely(test_thread_flag(TIF_SYSCALL_EMU)))
+                return;
+        /*
+         * If we are single-stepping, synthesize a trap to follow the
+         * system call instruction.
+         */
+        if (test_thread_flag(TIF_SINGLESTEP) &&
+            (current->ptrace & PT_PTRACED))
+                send_sigtrap(current, regs, 0);
+}
diff --git a/arch/x86/kernel/reboot.c b/arch/x86/kernel/reboot.c
index f8a62160e151..9dcf39c02972 100644
--- a/arch/x86/kernel/reboot.c
+++ b/arch/x86/kernel/reboot.c
@@ -177,6 +177,14 @@ static struct dmi_system_id __initdata reboot_dmi_table[] = {
                        DMI_MATCH(DMI_PRODUCT_NAME, "PowerEdge 2400"),
                },
        },
+        {       /* Handle problems with rebooting on Dell T5400's */
+                .callback = set_bios_reboot,
+                .ident = "Dell Precision T5400",
+                .matches = {
+                        DMI_MATCH(DMI_SYS_VENDOR, "Dell Inc."),
+                        DMI_MATCH(DMI_PRODUCT_NAME, "Precision WorkStation T5400"),
+                },
+        },
        {       /* Handle problems with rebooting on HP laptops */
                .callback = set_bios_reboot,
                .ident = "HP Compaq Laptop",
diff --git a/arch/x86/kernel/setup.c b/arch/x86/kernel/setup.c
index 531b55b8e81a..ec952aa5394a 100644
--- a/arch/x86/kernel/setup.c
+++ b/arch/x86/kernel/setup.c
@@ -57,12 +57,8 @@
 #include <linux/slab.h>
 #include <linux/user.h>
 #include <linux/delay.h>
-#include <linux/highmem.h>
 #include <linux/kallsyms.h>
-#include <linux/edd.h>
-#include <linux/iscsi_ibft.h>
-#include <linux/kexec.h>
 #include <linux/cpufreq.h>
 #include <linux/dma-mapping.h>
 #include <linux/ctype.h>
@@ -96,7 +92,7 @@
 #include <asm/smp.h>
 #include <asm/desc.h>
 #include <asm/dma.h>
-#include <asm/gart.h>
+#include <asm/iommu.h>
 #include <asm/mmu_context.h>
 #include <asm/proto.h>
@@ -104,7 +100,6 @@
 #include <asm/paravirt.h>
 #include <asm/percpu.h>
-#include <asm/sections.h>
 #include <asm/topology.h>
 #include <asm/apicdef.h>
 #ifdef CONFIG_X86_64
@@ -579,6 +574,10 @@ static int __init setup_elfcorehdr(char *arg)
 early_param("elfcorehdr", setup_elfcorehdr);
 #endif
+static struct x86_quirks default_x86_quirks __initdata;
+struct x86_quirks *x86_quirks __initdata = &default_x86_quirks;
 /*
 * Determine if we were loaded by an EFI loader.  If so, then we have also been
 * passed the efi memmap, systab, etc., so we should use these data structures
@@ -824,7 +823,10 @@ void __init setup_arch(char **cmdline_p)
        vmi_init();
 #endif
+        paravirt_pagetable_setup_start(swapper_pg_dir);
        paging_init();
+        paravirt_pagetable_setup_done(swapper_pg_dir);
+        paravirt_post_allocator_init();
 #ifdef CONFIG_X86_64
        map_vsyscall();
@@ -854,14 +856,6 @@ void __init setup_arch(char **cmdline_p)
        init_cpu_to_node();
 #endif
-#ifdef CONFIG_X86_NUMAQ
-        /*
-         * need to check online nodes num, call it
-         * here before time_init/tsc_init
-         */
-        numaq_tsc_disable();
-#endif
        init_apic_mappings();
        ioapic_init_mappings();
diff --git a/arch/x86/kernel/signal_32.c b/arch/x86/kernel/signal_32.c
index d92373630963..07faaa5109cb 100644
--- a/arch/x86/kernel/signal_32.c
+++ b/arch/x86/kernel/signal_32.c
@@ -212,7 +212,7 @@ asmlinkage unsigned long sys_sigreturn(unsigned long __unused)
 badframe:
        if (show_unhandled_signals && printk_ratelimit()) {
-                printk(KERN_INFO "%s%s[%d] bad frame in sigreturn frame:"
+                printk("%s%s[%d] bad frame in sigreturn frame:"
                        "%p ip:%lx sp:%lx oeax:%lx",
                    task_pid_nr(current) > 1 ? KERN_INFO : KERN_EMERG,
                    current->comm, task_pid_nr(current), frame, regs->ip,
@@ -657,12 +657,6 @@ static void do_signal(struct pt_regs *regs)
 void
 do_notify_resume(struct pt_regs *regs, void *unused, __u32 thread_info_flags)
 {
-        /* Pending single-step? */
-        if (thread_info_flags & _TIF_SINGLESTEP) {
-                regs->flags |= X86_EFLAGS_TF;
-                clear_thread_flag(TIF_SINGLESTEP);
-        }
        /* deal with pending signal delivery */
        if (thread_info_flags & _TIF_SIGPENDING)
                do_signal(regs);
diff --git a/arch/x86/kernel/signal_64.c b/arch/x86/kernel/signal_64.c
index e53b267662e7..bf87684474f1 100644
--- a/arch/x86/kernel/signal_64.c
+++ b/arch/x86/kernel/signal_64.c
@@ -487,12 +487,6 @@ static void do_signal(struct pt_regs *regs)
 void do_notify_resume(struct pt_regs *regs, void *unused,
                      __u32 thread_info_flags)
 {
-        /* Pending single-step? */
-        if (thread_info_flags & _TIF_SINGLESTEP) {
-                regs->flags |= X86_EFLAGS_TF;
-                clear_thread_flag(TIF_SINGLESTEP);
-        }
 #ifdef CONFIG_X86_MCE
        /* notify userspace of pending MCEs */
        if (thread_info_flags & _TIF_MCE_NOTIFY)
diff --git a/arch/x86/kernel/smpboot.c b/arch/x86/kernel/smpboot.c
index 27456574f070..27640196eb7c 100644
--- a/arch/x86/kernel/smpboot.c
+++ b/arch/x86/kernel/smpboot.c
@@ -546,8 +546,8 @@ static inline void __inquire_remote_apic(int apicid)
                        printk(KERN_CONT
                               "a previous APIC delivery may have failed\n");
-                apic_write_around(APIC_ICR2, SET_APIC_DEST_FIELD(apicid));
+                apic_write(APIC_ICR2, SET_APIC_DEST_FIELD(apicid));
-                apic_write_around(APIC_ICR, APIC_DM_REMRD | regs[i]);
+                apic_write(APIC_ICR, APIC_DM_REMRD | regs[i]);
                timeout = 0;
                do {
@@ -579,11 +579,11 @@ wakeup_secondary_cpu(int logical_apicid, unsigned long start_eip)
        int maxlvt;
        /* Target chip */
-        apic_write_around(APIC_ICR2, SET_APIC_DEST_FIELD(logical_apicid));
+        apic_write(APIC_ICR2, SET_APIC_DEST_FIELD(logical_apicid));
        /* Boot on the stack */
        /* Kick the second */
-        apic_write_around(APIC_ICR, APIC_DM_NMI | APIC_DEST_LOGICAL);
+        apic_write(APIC_ICR, APIC_DM_NMI | APIC_DEST_LOGICAL);
        Dprintk("Waiting for send to finish...\n");
        send_status = safe_apic_wait_icr_idle();
@@ -592,14 +592,9 @@ wakeup_secondary_cpu(int logical_apicid, unsigned long start_eip)
         * Give the other CPU some time to accept the IPI.
         */
        udelay(200);
-        /*
-         * Due to the Pentium erratum 3AP.
-         */
        maxlvt = lapic_get_maxlvt();
-        if (maxlvt > 3) {
+        if (maxlvt > 3)                 /* Due to the Pentium erratum 3AP.  */
-                apic_read_around(APIC_SPIV);
                apic_write(APIC_ESR, 0);
-        }
        accept_status = (apic_read(APIC_ESR) & 0xEF);
        Dprintk("NMI sent.\n");
@@ -625,12 +620,14 @@ wakeup_secondary_cpu(int phys_apicid, unsigned long start_eip)
                return send_status;
        }
+        maxlvt = lapic_get_maxlvt();
        /*
         * Be paranoid about clearing APIC errors.
         */
        if (APIC_INTEGRATED(apic_version[phys_apicid])) {
-                apic_read_around(APIC_SPIV);
+                if (maxlvt > 3)         /* Due to the Pentium erratum 3AP.  */
-                apic_write(APIC_ESR, 0);
+                        apic_write(APIC_ESR, 0);
                apic_read(APIC_ESR);
        }
@@ -639,13 +636,13 @@ wakeup_secondary_cpu(int phys_apicid, unsigned long start_eip)
        /*
         * Turn INIT on target chip
         */
-        apic_write_around(APIC_ICR2, SET_APIC_DEST_FIELD(phys_apicid));
+        apic_write(APIC_ICR2, SET_APIC_DEST_FIELD(phys_apicid));
        /*
         * Send IPI
         */
-        apic_write_around(APIC_ICR, APIC_INT_LEVELTRIG | APIC_INT_ASSERT
+        apic_write(APIC_ICR,
-                                | APIC_DM_INIT);
+                   APIC_INT_LEVELTRIG | APIC_INT_ASSERT | APIC_DM_INIT);
        Dprintk("Waiting for send to finish...\n");
        send_status = safe_apic_wait_icr_idle();
@@ -655,10 +652,10 @@ wakeup_secondary_cpu(int phys_apicid, unsigned long start_eip)
        Dprintk("Deasserting INIT.\n");
        /* Target chip */
-        apic_write_around(APIC_ICR2, SET_APIC_DEST_FIELD(phys_apicid));
+        apic_write(APIC_ICR2, SET_APIC_DEST_FIELD(phys_apicid));
        /* Send IPI */
-        apic_write_around(APIC_ICR, APIC_INT_LEVELTRIG | APIC_DM_INIT);
+        apic_write(APIC_ICR, APIC_INT_LEVELTRIG | APIC_DM_INIT);
        Dprintk("Waiting for send to finish...\n");
        send_status = safe_apic_wait_icr_idle();
@@ -689,12 +686,10 @@ wakeup_secondary_cpu(int phys_apicid, unsigned long start_eip)
         */
        Dprintk("#startup loops: %d.\n", num_starts);
-        maxlvt = lapic_get_maxlvt();
        for (j = 1; j <= num_starts; j++) {
                Dprintk("Sending STARTUP #%d.\n", j);
-                apic_read_around(APIC_SPIV);
+                if (maxlvt > 3)         /* Due to the Pentium erratum 3AP.  */
-                apic_write(APIC_ESR, 0);
+                        apic_write(APIC_ESR, 0);
                apic_read(APIC_ESR);
                Dprintk("After apic_write.\n");
@@ -703,12 +698,11 @@ wakeup_secondary_cpu(int phys_apicid, unsigned long start_eip)
                 */
                /* Target chip */
-                apic_write_around(APIC_ICR2, SET_APIC_DEST_FIELD(phys_apicid));
+                apic_write(APIC_ICR2, SET_APIC_DEST_FIELD(phys_apicid));
                /* Boot on the stack */
                /* Kick the second */
-                apic_write_around(APIC_ICR, APIC_DM_STARTUP
+                apic_write(APIC_ICR, APIC_DM_STARTUP | (start_eip >> 12));
-                                        | (start_eip >> 12));
                /*
                 * Give the other CPU some time to accept the IPI.
@@ -724,13 +718,8 @@ wakeup_secondary_cpu(int phys_apicid, unsigned long start_eip)
                 * Give the other CPU some time to accept the IPI.
                 */
                udelay(200);
-                /*
+                if (maxlvt > 3)         /* Due to the Pentium erratum 3AP.  */
-                 * Due to the Pentium erratum 3AP.
-                 */
-                if (maxlvt > 3) {
-                        apic_read_around(APIC_SPIV);
                        apic_write(APIC_ESR, 0);
-                }
                accept_status = (apic_read(APIC_ESR) & 0xEF);
                if (send_status || accept_status)
                        break;
@@ -768,7 +757,7 @@ static void __cpuinit do_fork_idle(struct work_struct *work)
 *
 * Must be called after the _cpu_pda pointer table is initialized.
 */
-static int __cpuinit get_local_pda(int cpu)
+int __cpuinit get_local_pda(int cpu)
 {
        struct x8664_pda *oldpda, *newpda;
        unsigned long size = sizeof(struct x8664_pda);
@@ -1390,7 +1379,8 @@ static int __init parse_maxcpus(char *arg)
 {
        extern unsigned int maxcpus;
-        maxcpus = simple_strtoul(arg, NULL, 0);
+        if (arg)
+                maxcpus = simple_strtoul(arg, NULL, 0);
        return 0;
 }
 early_param("maxcpus", parse_maxcpus);
diff --git a/arch/x86/kernel/step.c b/arch/x86/kernel/step.c
index 92c20fee6781..e8b9863ef8c4 100644
--- a/arch/x86/kernel/step.c
+++ b/arch/x86/kernel/step.c
@@ -105,6 +105,20 @@ static int is_setting_trap_flag(struct task_struct *child, struct pt_regs *regs)
 static int enable_single_step(struct task_struct *child)
 {
        struct pt_regs *regs = task_pt_regs(child);
+        unsigned long oflags;
+        /*
+         * If we stepped into a sysenter/syscall insn, it trapped in
+         * kernel mode; do_debug() cleared TF and set TIF_SINGLESTEP.
+         * If user-mode had set TF itself, then it's still clear from
+         * do_debug() and we need to set it again to restore the user
+         * state so we don't wrongly set TIF_FORCED_TF below.
+         * If enable_single_step() was used last and that is what
+         * set TIF_SINGLESTEP, then both TF and TIF_FORCED_TF are
+         * already set and our bookkeeping is fine.
+         */
+        if (unlikely(test_tsk_thread_flag(child, TIF_SINGLESTEP)))
+                regs->flags |= X86_EFLAGS_TF;
        /*
         * Always set TIF_SINGLESTEP - this guarantees that
@@ -113,11 +127,7 @@ static int enable_single_step(struct task_struct *child)
         */
        set_tsk_thread_flag(child, TIF_SINGLESTEP);
-        /*
+        oflags = regs->flags;
-         * If TF was already set, don't do anything else
-         */
-        if (regs->flags & X86_EFLAGS_TF)
-                return 0;
        /* Set TF on the kernel stack.. */
        regs->flags |= X86_EFLAGS_TF;
@@ -126,9 +136,22 @@ static int enable_single_step(struct task_struct *child)
         * ..but if TF is changed by the instruction we will trace,
         * don't mark it as being "us" that set it, so that we
         * won't clear it by hand later.
+         *
+         * Note that if we don't actually execute the popf because
+         * of a signal arriving right now or suchlike, we will lose
+         * track of the fact that it really was "us" that set it.
         */
-        if (is_setting_trap_flag(child, regs))
+        if (is_setting_trap_flag(child, regs)) {
+                clear_tsk_thread_flag(child, TIF_FORCED_TF);
                return 0;
+        }
+        /*
+         * If TF was already set, check whether it was us who set it.
+         * If not, we should never attempt a block step.
+         */
+        if (oflags & X86_EFLAGS_TF)
+                return test_tsk_thread_flag(child, TIF_FORCED_TF);
        set_tsk_thread_flag(child, TIF_FORCED_TF);
diff --git a/arch/x86/kernel/time_32.c b/arch/x86/kernel/time_32.c
index 059ca6ee59b4..ffe3c664afc0 100644
--- a/arch/x86/kernel/time_32.c
+++ b/arch/x86/kernel/time_32.c
@@ -129,6 +129,7 @@ void __init hpet_time_init(void)
 */
 void __init time_init(void)
 {
+        pre_time_init_hook();
        tsc_init();
        late_time_init = choose_time_init();
 }
diff --git a/arch/x86/kernel/traps_32.c b/arch/x86/kernel/traps_32.c
index 8a768973c4f0..03df8e45e5a1 100644
--- a/arch/x86/kernel/traps_32.c
+++ b/arch/x86/kernel/traps_32.c
@@ -58,6 +58,7 @@
 #include <asm/nmi.h>
 #include <asm/smp.h>
 #include <asm/io.h>
+#include <asm/traps.h>
 #include "mach_traps.h"
@@ -77,26 +78,6 @@ char ignore_fpu_irq;
 gate_desc idt_table[256]
        __attribute__((__section__(".data.idt"))) = { { { { 0, 0 } } }, };
-asmlinkage void divide_error(void);
-asmlinkage void debug(void);
-asmlinkage void nmi(void);
-asmlinkage void int3(void);
-asmlinkage void overflow(void);
-asmlinkage void bounds(void);
-asmlinkage void invalid_op(void);
-asmlinkage void device_not_available(void);
-asmlinkage void coprocessor_segment_overrun(void);
-asmlinkage void invalid_TSS(void);
-asmlinkage void segment_not_present(void);
-asmlinkage void stack_segment(void);
-asmlinkage void general_protection(void);
-asmlinkage void page_fault(void);
-asmlinkage void coprocessor_error(void);
-asmlinkage void simd_coprocessor_error(void);
-asmlinkage void alignment_check(void);
-asmlinkage void spurious_interrupt_bug(void);
-asmlinkage void machine_check(void);
 int panic_on_unrecovered_nmi;
 int kstack_depth_to_print = 24;
 static unsigned int code_bytes = 64;
@@ -256,7 +237,7 @@ static const struct stacktrace_ops print_trace_ops = {
 static void
 show_trace_log_lvl(struct task_struct *task, struct pt_regs *regs,
-                   unsigned long *stack, unsigned long bp, char *log_lvl)
+                unsigned long *stack, unsigned long bp, char *log_lvl)
 {
        dump_trace(task, regs, stack, bp, &print_trace_ops, log_lvl);
        printk("%s =======================\n", log_lvl);
@@ -383,6 +364,54 @@ int is_valid_bugaddr(unsigned long ip)
        return ud2 == 0x0b0f;
 }
+static raw_spinlock_t die_lock = __RAW_SPIN_LOCK_UNLOCKED;
+static int die_owner = -1;
+static unsigned int die_nest_count;
+unsigned __kprobes long oops_begin(void)
+{
+        unsigned long flags;
+        oops_enter();
+        if (die_owner != raw_smp_processor_id()) {
+                console_verbose();
+                raw_local_irq_save(flags);
+                __raw_spin_lock(&die_lock);
+                die_owner = smp_processor_id();
+                die_nest_count = 0;
+                bust_spinlocks(1);
+        } else {
+                raw_local_irq_save(flags);
+        }
+        die_nest_count++;
+        return flags;
+}
+void __kprobes oops_end(unsigned long flags, struct pt_regs *regs, int signr)
+{
+        bust_spinlocks(0);
+        die_owner = -1;
+        add_taint(TAINT_DIE);
+        __raw_spin_unlock(&die_lock);
+        raw_local_irq_restore(flags);
+        if (!regs)
+                return;
+        if (kexec_should_crash(current))
+                crash_kexec(regs);
+        if (in_interrupt())
+                panic("Fatal exception in interrupt");
+        if (panic_on_oops)
+                panic("Fatal exception");
+        oops_exit();
+        do_exit(signr);
+}
 int __kprobes __die(const char *str, struct pt_regs *regs, long err)
 {
        unsigned short ss;
@@ -423,31 +452,9 @@ int __kprobes __die(const char *str, struct pt_regs *regs, long err)
 */
 void die(const char *str, struct pt_regs *regs, long err)
 {
-        static struct {
+        unsigned long flags = oops_begin();
-                raw_spinlock_t lock;
-                u32 lock_owner;
-                int lock_owner_depth;
-        } die = {
-                .lock =                 __RAW_SPIN_LOCK_UNLOCKED,
-                .lock_owner =           -1,
-                .lock_owner_depth =     0
-        };
-        unsigned long flags;
-        oops_enter();
-        if (die.lock_owner != raw_smp_processor_id()) {
-                console_verbose();
-                raw_local_irq_save(flags);
-                __raw_spin_lock(&die.lock);
-                die.lock_owner = smp_processor_id();
-                die.lock_owner_depth = 0;
-                bust_spinlocks(1);
-        } else {
-                raw_local_irq_save(flags);
-        }
-        if (++die.lock_owner_depth < 3) {
+        if (die_nest_count < 3) {
                report_bug(regs->ip, regs);
                if (__die(str, regs, err))
@@ -456,26 +463,7 @@ void die(const char *str, struct pt_regs *regs, long err)
                printk(KERN_EMERG "Recursive die() failure, output suppressed\n");
        }
-        bust_spinlocks(0);
+        oops_end(flags, regs, SIGSEGV);
-        die.lock_owner = -1;
-        add_taint(TAINT_DIE);
-        __raw_spin_unlock(&die.lock);
-        raw_local_irq_restore(flags);
-        if (!regs)
-                return;
-        if (kexec_should_crash(current))
-                crash_kexec(regs);
-        if (in_interrupt())
-                panic("Fatal exception in interrupt");
-        if (panic_on_oops)
-                panic("Fatal exception");
-        oops_exit();
-        do_exit(SIGSEGV);
 }
 static inline void
diff --git a/arch/x86/kernel/traps_64.c b/arch/x86/kernel/traps_64.c
index 2696a6837782..3f18d73f420c 100644
--- a/arch/x86/kernel/traps_64.c
+++ b/arch/x86/kernel/traps_64.c
@@ -51,30 +51,10 @@
 #include <asm/pgalloc.h>
 #include <asm/proto.h>
 #include <asm/pda.h>
+#include <asm/traps.h>
 #include <mach_traps.h>
-asmlinkage void divide_error(void);
-asmlinkage void debug(void);
-asmlinkage void nmi(void);
-asmlinkage void int3(void);
-asmlinkage void overflow(void);
-asmlinkage void bounds(void);
-asmlinkage void invalid_op(void);
-asmlinkage void device_not_available(void);
-asmlinkage void double_fault(void);
-asmlinkage void coprocessor_segment_overrun(void);
-asmlinkage void invalid_TSS(void);
-asmlinkage void segment_not_present(void);
-asmlinkage void stack_segment(void);
-asmlinkage void general_protection(void);
-asmlinkage void page_fault(void);
-asmlinkage void coprocessor_error(void);
-asmlinkage void simd_coprocessor_error(void);
-asmlinkage void alignment_check(void);
-asmlinkage void spurious_interrupt_bug(void);
-asmlinkage void machine_check(void);
 int panic_on_unrecovered_nmi;
 int kstack_depth_to_print = 12;
 static unsigned int code_bytes = 64;
@@ -355,17 +335,24 @@ static const struct stacktrace_ops print_trace_ops = {
        .address = print_trace_address,
 };
-void show_trace(struct task_struct *task, struct pt_regs *regs,
+static void
-                unsigned long *stack, unsigned long bp)
+show_trace_log_lvl(struct task_struct *task, struct pt_regs *regs,
+                unsigned long *stack, unsigned long bp, char *log_lvl)
 {
        printk("\nCall Trace:\n");
-        dump_trace(task, regs, stack, bp, &print_trace_ops, NULL);
+        dump_trace(task, regs, stack, bp, &print_trace_ops, log_lvl);
        printk("\n");
 }
+void show_trace(struct task_struct *task, struct pt_regs *regs,
+                unsigned long *stack, unsigned long bp)
+{
+        show_trace_log_lvl(task, regs, stack, bp, "");
+}
 static void
-_show_stack(struct task_struct *task, struct pt_regs *regs,
+show_stack_log_lvl(struct task_struct *task, struct pt_regs *regs,
-                unsigned long *sp, unsigned long bp)
+                unsigned long *sp, unsigned long bp, char *log_lvl)
 {
        unsigned long *stack;
        int i;
@@ -399,12 +386,12 @@ _show_stack(struct task_struct *task, struct pt_regs *regs,
                printk(" %016lx", *stack++);
                touch_nmi_watchdog();
        }
-        show_trace(task, regs, sp, bp);
+        show_trace_log_lvl(task, regs, sp, bp, log_lvl);
 }
 void show_stack(struct task_struct *task, unsigned long *sp)
 {
-        _show_stack(task, NULL, sp, 0);
+        show_stack_log_lvl(task, NULL, sp, 0, "");
 }
 /*
@@ -454,7 +441,8 @@ void show_registers(struct pt_regs *regs)
                u8 *ip;
                printk("Stack: ");
-                _show_stack(NULL, regs, (unsigned long *)sp, regs->bp);
+                show_stack_log_lvl(NULL, regs, (unsigned long *)sp,
+                                regs->bp, "");
                printk("\n");
                printk(KERN_EMERG "Code: ");
@@ -518,7 +506,7 @@ unsigned __kprobes long oops_begin(void)
 }
 void __kprobes oops_end(unsigned long flags, struct pt_regs *regs, int signr)
-{ 
+{
        die_owner = -1;
        bust_spinlocks(0);
        die_nest_count--;
diff --git a/arch/x86/kernel/visws_quirks.c b/arch/x86/kernel/visws_quirks.c
index e94bdb6add1d..41e01b145c48 100644
--- a/arch/x86/kernel/visws_quirks.c
+++ b/arch/x86/kernel/visws_quirks.c
@@ -73,7 +73,7 @@ int is_visws_box(void)
        return visws_board_type >= 0;
 }
-static int __init visws_time_init_quirk(void)
+static int __init visws_time_init(void)
 {
        printk(KERN_INFO "Starting Cobalt Timer system clock\n");
@@ -93,7 +93,7 @@ static int __init visws_time_init_quirk(void)
        return 0;
 }
-static int __init visws_pre_intr_init_quirk(void)
+static int __init visws_pre_intr_init(void)
 {
        init_VISWS_APIC_irqs();
@@ -114,7 +114,7 @@ EXPORT_SYMBOL(sgivwfb_mem_size);
 long long mem_size __initdata = 0;
-static char * __init visws_memory_setup_quirk(void)
+static char * __init visws_memory_setup(void)
 {
        long long gfx_mem_size = 8 * MB;
@@ -176,7 +176,7 @@ static void visws_machine_power_off(void)
        outl(PIIX_SPECIAL_STOP, 0xCFC);
 }
-static int __init visws_get_smp_config_quirk(unsigned int early)
+static int __init visws_get_smp_config(unsigned int early)
 {
        /*
         * Prevent MP-table parsing by the generic code:
@@ -192,7 +192,7 @@ extern unsigned int __cpuinitdata maxcpus;
 * No problem for Linux.
 */
-static void __init MP_processor_info (struct mpc_config_processor *m)
+static void __init MP_processor_info(struct mpc_config_processor *m)
 {
        int ver, logical_apicid;
        physid_mask_t apic_cpus;
@@ -232,7 +232,7 @@ static void __init MP_processor_info (struct mpc_config_processor *m)
        apic_version[m->mpc_apicid] = ver;
 }
-int __init visws_find_smp_config_quirk(unsigned int reserve)
+static int __init visws_find_smp_config(unsigned int reserve)
 {
        struct mpc_config_processor *mp = phys_to_virt(CO_CPU_TAB_PHYS);
        unsigned short ncpus = readw(phys_to_virt(CO_CPU_NUM_PHYS));
@@ -258,7 +258,17 @@ int __init visws_find_smp_config_quirk(unsigned int reserve)
        return 1;
 }
-extern int visws_trap_init_quirk(void);
+static int visws_trap_init(void);
+static struct x86_quirks visws_x86_quirks __initdata = {
+        .arch_time_init         = visws_time_init,
+        .arch_pre_intr_init     = visws_pre_intr_init,
+        .arch_memory_setup      = visws_memory_setup,
+        .arch_intr_init         = NULL,
+        .arch_trap_init         = visws_trap_init,
+        .mach_get_smp_config    = visws_get_smp_config,
+        .mach_find_smp_config   = visws_find_smp_config,
+};
 void __init visws_early_detect(void)
 {
@@ -272,16 +282,10 @@ void __init visws_early_detect(void)
        /*
         * Install special quirks for timer, interrupt and memory setup:
-         */
-        arch_time_init_quirk            = visws_time_init_quirk;
-        arch_pre_intr_init_quirk        = visws_pre_intr_init_quirk;
-        arch_memory_setup_quirk         = visws_memory_setup_quirk;
-        /*
         * Fall back to generic behavior for traps:
+         * Override generic MP-table parsing:
         */
-        arch_intr_init_quirk            = NULL;
+        x86_quirks = &visws_x86_quirks;
-        arch_trap_init_quirk            = visws_trap_init_quirk;
        /*
         * Install reboot quirks:
@@ -294,12 +298,6 @@ void __init visws_early_detect(void)
         */
        no_broadcast = 0;
-        /*
-         * Override generic MP-table parsing:
-         */
-        mach_get_smp_config_quirk       = visws_get_smp_config_quirk;
-        mach_find_smp_config_quirk      = visws_find_smp_config_quirk;
 #ifdef CONFIG_X86_IO_APIC
        /*
         * Turn off IO-APIC detection and initialization:
@@ -426,7 +424,7 @@ static __init void cobalt_init(void)
                co_apic_read(CO_APIC_ID));
 }
-int __init visws_trap_init_quirk(void)
+static int __init visws_trap_init(void)
 {
        lithium_init();
        cobalt_init();
diff --git a/arch/x86/kernel/vmi_32.c b/arch/x86/kernel/vmi_32.c
index b15346092b7b..0a1b1a9d922d 100644
--- a/arch/x86/kernel/vmi_32.c
+++ b/arch/x86/kernel/vmi_32.c
@@ -906,7 +906,6 @@ static inline int __init activate_vmi(void)
 #ifdef CONFIG_X86_LOCAL_APIC
        para_fill(pv_apic_ops.apic_read, APICRead);
        para_fill(pv_apic_ops.apic_write, APICWrite);
-        para_fill(pv_apic_ops.apic_write_atomic, APICWrite);
 #endif
        /*
diff --git a/arch/x86/lguest/boot.c b/arch/x86/lguest/boot.c
index 50dad44fb542..0313a5eec412 100644
--- a/arch/x86/lguest/boot.c
+++ b/arch/x86/lguest/boot.c
@@ -991,7 +991,6 @@ __init void lguest_init(void)
 #ifdef CONFIG_X86_LOCAL_APIC
        /* apic read/write intercepts */
        pv_apic_ops.apic_write = lguest_apic_write;
-        pv_apic_ops.apic_write_atomic = lguest_apic_write;
        pv_apic_ops.apic_read = lguest_apic_read;
 #endif
diff --git a/arch/x86/mach-default/setup.c b/arch/x86/mach-default/setup.c
index 48278fa7d3de..3d317836be9e 100644
--- a/arch/x86/mach-default/setup.c
+++ b/arch/x86/mach-default/setup.c
@@ -10,14 +10,6 @@
 #include <asm/e820.h>
 #include <asm/setup.h>
-/*
- * Any quirks to be performed to initialize timers/irqs/etc?
- */
-int (*arch_time_init_quirk)(void);
-int (*arch_pre_intr_init_quirk)(void);
-int (*arch_intr_init_quirk)(void);
-int (*arch_trap_init_quirk)(void);
 #ifdef CONFIG_HOTPLUG_CPU
 #define DEFAULT_SEND_IPI        (1)
 #else
@@ -37,8 +29,8 @@ int no_broadcast=DEFAULT_SEND_IPI;
 **/
 void __init pre_intr_init_hook(void)
 {
-        if (arch_pre_intr_init_quirk) {
+        if (x86_quirks->arch_pre_intr_init) {
-                if (arch_pre_intr_init_quirk())
+                if (x86_quirks->arch_pre_intr_init())
                        return;
        }
        init_ISA_irqs();
@@ -64,8 +56,8 @@ static struct irqaction irq2 = {
 **/
 void __init intr_init_hook(void)
 {
-        if (arch_intr_init_quirk) {
+        if (x86_quirks->arch_intr_init) {
-                if (arch_intr_init_quirk())
+                if (x86_quirks->arch_intr_init())
                        return;
        }
 #ifdef CONFIG_X86_LOCAL_APIC
@@ -97,8 +89,8 @@ void __init pre_setup_arch_hook(void)
 **/
 void __init trap_init_hook(void)
 {
-        if (arch_trap_init_quirk) {
+        if (x86_quirks->arch_trap_init) {
-                if (arch_trap_init_quirk())
+                if (x86_quirks->arch_trap_init())
                        return;
        }
 }
@@ -111,6 +103,16 @@ static struct irqaction irq0  = {
 };
 /**
+ * pre_time_init_hook - do any specific initialisations before.
+ *
+ **/
+void __init pre_time_init_hook(void)
+{
+        if (x86_quirks->arch_pre_time_init)
+                x86_quirks->arch_pre_time_init();
+}
+/**
 * time_init_hook - do any specific initialisations for the system timer.
 *
 * Description:
@@ -119,13 +121,13 @@ static struct irqaction irq0  = {
 **/
 void __init time_init_hook(void)
 {
-        if (arch_time_init_quirk) {
+        if (x86_quirks->arch_time_init) {
                /*
                 * A nonzero return code does not mean failure, it means
                 * that the architecture quirk does not want any
                 * generic (timer) setup to be performed after this:
                 */
-                if (arch_time_init_quirk())
+                if (x86_quirks->arch_time_init())
                        return;
        }
diff --git a/arch/x86/mm/Makefile b/arch/x86/mm/Makefile
index 9873716e9f76..1fbb844c3d7a 100644
--- a/arch/x86/mm/Makefile
+++ b/arch/x86/mm/Makefile
@@ -21,3 +21,4 @@ obj-$(CONFIG_K8_NUMA)		+= k8topology_64.o
 endif
 obj-$(CONFIG_ACPI_NUMA)         += srat_$(BITS).o
+obj-$(CONFIG_MEMTEST)           += memtest.o
diff --git a/arch/x86/mm/init_32.c b/arch/x86/mm/init_32.c
index 9689a5138e64..d37f29376b0c 100644
--- a/arch/x86/mm/init_32.c
+++ b/arch/x86/mm/init_32.c
@@ -844,6 +844,9 @@ unsigned long __init_refok init_memory_mapping(unsigned long start,
                reserve_early(table_start << PAGE_SHIFT,
                                 table_end << PAGE_SHIFT, "PGTABLE");
+        if (!after_init_bootmem)
+                early_memtest(start, end);
        return end >> PAGE_SHIFT;
 }
@@ -868,8 +871,6 @@ void __init paging_init(void)
         */
        sparse_init();
        zone_sizes_init();
-        paravirt_post_allocator_init();
 }
 /*
diff --git a/arch/x86/mm/init_64.c b/arch/x86/mm/init_64.c
index 306049edd553..ec37121f6709 100644
--- a/arch/x86/mm/init_64.c
+++ b/arch/x86/mm/init_64.c
@@ -517,118 +517,6 @@ static void __init init_gbpages(void)
                direct_gbpages = 0;
 }
-#ifdef CONFIG_MEMTEST
-static void __init memtest(unsigned long start_phys, unsigned long size,
-                                 unsigned pattern)
-{
-        unsigned long i;
-        unsigned long *start;
-        unsigned long start_bad;
-        unsigned long last_bad;
-        unsigned long val;
-        unsigned long start_phys_aligned;
-        unsigned long count;
-        unsigned long incr;
-        switch (pattern) {
-        case 0:
-                val = 0UL;
-                break;
-        case 1:
-                val = -1UL;
-                break;
-        case 2:
-                val = 0x5555555555555555UL;
-                break;
-        case 3:
-                val = 0xaaaaaaaaaaaaaaaaUL;
-                break;
-        default:
-                return;
-        }
-        incr = sizeof(unsigned long);
-        start_phys_aligned = ALIGN(start_phys, incr);
-        count = (size - (start_phys_aligned - start_phys))/incr;
-        start = __va(start_phys_aligned);
-        start_bad = 0;
-        last_bad = 0;
-        for (i = 0; i < count; i++)
-                start[i] = val;
-        for (i = 0; i < count; i++, start++, start_phys_aligned += incr) {
-                if (*start != val) {
-                        if (start_phys_aligned == last_bad + incr) {
-                                last_bad += incr;
-                        } else {
-                                if (start_bad) {
-                                        printk(KERN_CONT "\n  %016lx bad mem addr %016lx - %016lx reserved",
-                                                val, start_bad, last_bad + incr);
-                                        reserve_early(start_bad, last_bad - start_bad, "BAD RAM");
-                                }
-                                start_bad = last_bad = start_phys_aligned;
-                        }
-                }
-        }
-        if (start_bad) {
-                printk(KERN_CONT "\n  %016lx bad mem addr %016lx - %016lx reserved",
-                        val, start_bad, last_bad + incr);
-                reserve_early(start_bad, last_bad - start_bad, "BAD RAM");
-        }
-}
-/* default is disabled */
-static int memtest_pattern __initdata;
-static int __init parse_memtest(char *arg)
-{
-        if (arg)
-                memtest_pattern = simple_strtoul(arg, NULL, 0);
-        return 0;
-}
-early_param("memtest", parse_memtest);
-static void __init early_memtest(unsigned long start, unsigned long end)
-{
-        u64 t_start, t_size;
-        unsigned pattern;
-        if (!memtest_pattern)
-                return;
-        printk(KERN_INFO "early_memtest: pattern num %d", memtest_pattern);
-        for (pattern = 0; pattern < memtest_pattern; pattern++) {
-                t_start = start;
-                t_size = 0;
-                while (t_start < end) {
-                        t_start = find_e820_area_size(t_start, &t_size, 1);
-                        /* done ? */
-                        if (t_start >= end)
-                                break;
-                        if (t_start + t_size > end)
-                                t_size = end - t_start;
-                        printk(KERN_CONT "\n  %016llx - %016llx pattern %d",
-                                (unsigned long long)t_start,
-                                (unsigned long long)t_start + t_size, pattern);
-                        memtest(t_start, t_size, pattern);
-                        t_start += t_size;
-                }
-        }
-        printk(KERN_CONT "\n");
-}
-#else
-static void __init early_memtest(unsigned long start, unsigned long end)
-{
-}
-#endif
 static unsigned long __init kernel_physical_mapping_init(unsigned long start,
                                                unsigned long end,
                                                unsigned long page_size_mask)
diff --git a/arch/x86/mm/memtest.c b/arch/x86/mm/memtest.c
new file mode 100644
index 000000000000..672e17f8262a
--- /dev/null
+++ b/arch/x86/mm/memtest.c
@@ -0,0 +1,123 @@
+#include <linux/kernel.h>
+#include <linux/errno.h>
+#include <linux/string.h>
+#include <linux/types.h>
+#include <linux/mm.h>
+#include <linux/smp.h>
+#include <linux/init.h>
+#include <linux/pfn.h>
+#include <asm/e820.h>
+static void __init memtest(unsigned long start_phys, unsigned long size,
+                                 unsigned pattern)
+{
+        unsigned long i;
+        unsigned long *start;
+        unsigned long start_bad;
+        unsigned long last_bad;
+        unsigned long val;
+        unsigned long start_phys_aligned;
+        unsigned long count;
+        unsigned long incr;
+        switch (pattern) {
+        case 0:
+                val = 0UL;
+                break;
+        case 1:
+                val = -1UL;
+                break;
+        case 2:
+#ifdef CONFIG_X86_64
+                val = 0x5555555555555555UL;
+#else
+                val = 0x55555555UL;
+#endif
+                break;
+        case 3:
+#ifdef CONFIG_X86_64
+                val = 0xaaaaaaaaaaaaaaaaUL;
+#else
+                val = 0xaaaaaaaaUL;
+#endif
+                break;
+        default:
+                return;
+        }
+        incr = sizeof(unsigned long);
+        start_phys_aligned = ALIGN(start_phys, incr);
+        count = (size - (start_phys_aligned - start_phys))/incr;
+        start = __va(start_phys_aligned);
+        start_bad = 0;
+        last_bad = 0;
+        for (i = 0; i < count; i++)
+                start[i] = val;
+        for (i = 0; i < count; i++, start++, start_phys_aligned += incr) {
+                if (*start != val) {
+                        if (start_phys_aligned == last_bad + incr) {
+                                last_bad += incr;
+                        } else {
+                                if (start_bad) {
+                                        printk(KERN_CONT "\n  %010lx bad mem addr %010lx - %010lx reserved",
+                                                val, start_bad, last_bad + incr);
+                                        reserve_early(start_bad, last_bad - start_bad, "BAD RAM");
+                                }
+                                start_bad = last_bad = start_phys_aligned;
+                        }
+                }
+        }
+        if (start_bad) {
+                printk(KERN_CONT "\n  %016lx bad mem addr %010lx - %010lx reserved",
+                        val, start_bad, last_bad + incr);
+                reserve_early(start_bad, last_bad - start_bad, "BAD RAM");
+        }
+}
+/* default is disabled */
+static int memtest_pattern __initdata;
+static int __init parse_memtest(char *arg)
+{
+        if (arg)
+                memtest_pattern = simple_strtoul(arg, NULL, 0);
+        return 0;
+}
+early_param("memtest", parse_memtest);
+void __init early_memtest(unsigned long start, unsigned long end)
+{
+        u64 t_start, t_size;
+        unsigned pattern;
+        if (!memtest_pattern)
+                return;
+        printk(KERN_INFO "early_memtest: pattern num %d", memtest_pattern);
+        for (pattern = 0; pattern < memtest_pattern; pattern++) {
+                t_start = start;
+                t_size = 0;
+                while (t_start < end) {
+                        t_start = find_e820_area_size(t_start, &t_size, 1);
+                        /* done ? */
+                        if (t_start >= end)
+                                break;
+                        if (t_start + t_size > end)
+                                t_size = end - t_start;
+                        printk(KERN_CONT "\n  %010llx - %010llx pattern %d",
+                                (unsigned long long)t_start,
+                                (unsigned long long)t_start + t_size, pattern);
+                        memtest(t_start, t_size, pattern);
+                        t_start += t_size;
+                }
+        }
+        printk(KERN_CONT "\n");
+}
diff --git a/arch/x86/mm/pat.c b/arch/x86/mm/pat.c
index 6bb597f4d701..2fe30916d4b6 100644
--- a/arch/x86/mm/pat.c
+++ b/arch/x86/mm/pat.c
@@ -12,6 +12,8 @@
 #include <linux/gfp.h>
 #include <linux/fs.h>
 #include <linux/bootmem.h>
+#include <linux/debugfs.h>
+#include <linux/seq_file.h>
 #include <asm/msr.h>
 #include <asm/tlbflush.h>
@@ -489,3 +491,89 @@ void unmap_devmem(unsigned long pfn, unsigned long size, pgprot_t vma_prot)
        free_memtype(addr, addr + size);
 }
+#if defined(CONFIG_DEBUG_FS)
+/* get Nth element of the linked list */
+static struct memtype *memtype_get_idx(loff_t pos)
+{
+        struct memtype *list_node, *print_entry;
+        int i = 1;
+        print_entry  = kmalloc(sizeof(struct memtype), GFP_KERNEL);
+        if (!print_entry)
+                return NULL;
+        spin_lock(&memtype_lock);
+        list_for_each_entry(list_node, &memtype_list, nd) {
+                if (pos == i) {
+                        *print_entry = *list_node;
+                        spin_unlock(&memtype_lock);
+                        return print_entry;
+                }
+                ++i;
+        }
+        spin_unlock(&memtype_lock);
+        kfree(print_entry);
+        return NULL;
+}
+static void *memtype_seq_start(struct seq_file *seq, loff_t *pos)
+{
+        if (*pos == 0) {
+                ++*pos;
+                seq_printf(seq, "PAT memtype list:\n");
+        }
+        return memtype_get_idx(*pos);
+}
+static void *memtype_seq_next(struct seq_file *seq, void *v, loff_t *pos)
+{
+        ++*pos;
+        return memtype_get_idx(*pos);
+}
+static void memtype_seq_stop(struct seq_file *seq, void *v)
+{
+}
+static int memtype_seq_show(struct seq_file *seq, void *v)
+{
+        struct memtype *print_entry = (struct memtype *)v;
+        seq_printf(seq, "%s @ 0x%Lx-0x%Lx\n", cattr_name(print_entry->type),
+                        print_entry->start, print_entry->end);
+        kfree(print_entry);
+        return 0;
+}
+static struct seq_operations memtype_seq_ops = {
+        .start = memtype_seq_start,
+        .next  = memtype_seq_next,
+        .stop  = memtype_seq_stop,
+        .show  = memtype_seq_show,
+};
+static int memtype_seq_open(struct inode *inode, struct file *file)
+{
+        return seq_open(file, &memtype_seq_ops);
+}
+static const struct file_operations memtype_fops = {
+        .open    = memtype_seq_open,
+        .read    = seq_read,
+        .llseek  = seq_lseek,
+        .release = seq_release,
+};
+static int __init pat_memtype_list_init(void)
+{
+        debugfs_create_file("pat_memtype_list", S_IRUSR, arch_debugfs_dir,
+                                NULL, &memtype_fops);
+        return 0;
+}
+late_initcall(pat_memtype_list_init);
+#endif /* CONFIG_DEBUG_FS */
diff --git a/arch/x86/pci/Makefile b/arch/x86/pci/Makefile
index e515e8db842a..d49202e740ea 100644
--- a/arch/x86/pci/Makefile
+++ b/arch/x86/pci/Makefile
@@ -5,13 +5,13 @@ obj-$(CONFIG_PCI_MMCONFIG)	+= mmconfig_$(BITS).o direct.o mmconfig-shared.o
 obj-$(CONFIG_PCI_DIRECT)        += direct.o
 obj-$(CONFIG_PCI_OLPC)          += olpc.o
-pci-y                           := fixup.o
+obj-y                           += fixup.o
-pci-$(CONFIG_ACPI)              += acpi.o
+obj-$(CONFIG_ACPI)              += acpi.o
-pci-y                           += legacy.o irq.o
+obj-y                           += legacy.o irq.o
-pci-$(CONFIG_X86_VISWS)         += visws.o
+obj-$(CONFIG_X86_VISWS)         += visws.o
-pci-$(CONFIG_X86_NUMAQ)         += numa.o
+obj-$(CONFIG_X86_NUMAQ)         += numaq_32.o
-obj-y                           += $(pci-y) common.o early.o
+obj-y                           += common.o early.o
 obj-y                           += amd_bus.o
diff --git a/arch/x86/pci/legacy.c b/arch/x86/pci/legacy.c
index 132876cc6fca..ec9ce35e44d6 100644
--- a/arch/x86/pci/legacy.c
+++ b/arch/x86/pci/legacy.c
@@ -57,14 +57,17 @@ static int __init pci_legacy_init(void)
 int __init pci_subsys_init(void)
 {
+#ifdef CONFIG_X86_NUMAQ
+        pci_numaq_init();
+#endif
 #ifdef CONFIG_ACPI
        pci_acpi_init();
 #endif
+#ifdef CONFIG_X86_VISWS
+        pci_visws_init();
+#endif
        pci_legacy_init();
        pcibios_irq_init();
-#ifdef CONFIG_X86_NUMAQ
-        pci_numa_init();
-#endif
        pcibios_init();
        return 0;
diff --git a/arch/x86/pci/numa.c b/arch/x86/pci/numaq_32.c
index 8b5ca1966731..f4b16dc11dad 100644
--- a/arch/x86/pci/numa.c
+++ b/arch/x86/pci/numaq_32.c
@@ -1,5 +1,5 @@
 /*
- * numa.c - Low-level PCI access for NUMA-Q machines
+ * numaq_32.c - Low-level PCI access for NUMA-Q machines
 */
 #include <linux/pci.h>
@@ -151,7 +151,7 @@ static void __devinit pci_fixup_i450nx(struct pci_dev *d)
 }
 DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_82451NX, pci_fixup_i450nx);
-int __init pci_numa_init(void)
+int __init pci_numaq_init(void)
 {
        int quad;
diff --git a/arch/x86/pci/pci.h b/arch/x86/pci/pci.h
index 3e25deb821ac..15b9cf6be729 100644
--- a/arch/x86/pci/pci.h
+++ b/arch/x86/pci/pci.h
@@ -108,7 +108,8 @@ extern void __init dmi_check_skip_isa_align(void);
 /* some common used subsys_initcalls */
 extern int __init pci_acpi_init(void);
 extern int __init pcibios_irq_init(void);
-extern int __init pci_numa_init(void);
+extern int __init pci_visws_init(void);
+extern int __init pci_numaq_init(void);
 extern int __init pcibios_init(void);
 /* pci-mmconfig.c */
diff --git a/arch/x86/pci/visws.c b/arch/x86/pci/visws.c
index 1a7bed492bb1..42f4cb19faca 100644
--- a/arch/x86/pci/visws.c
+++ b/arch/x86/pci/visws.c
@@ -86,8 +86,14 @@ void __init pcibios_update_irq(struct pci_dev *dev, int irq)
        pci_write_config_byte(dev, PCI_INTERRUPT_LINE, irq);
 }
-static int __init pci_visws_init(void)
+int __init pci_visws_init(void)
 {
+        if (!is_visws_box())
+                return -1;
+        pcibios_enable_irq = &pci_visws_enable_irq;
+        pcibios_disable_irq = &pci_visws_disable_irq;
        /* The VISWS supports configuration access type 1 only */
        pci_probe = (pci_probe | PCI_PROBE_CONF1) &
                    ~(PCI_PROBE_BIOS | PCI_PROBE_CONF2);
@@ -105,18 +111,3 @@ static int __init pci_visws_init(void)
        pcibios_resource_survey();
        return 0;
 }
-static __init int pci_subsys_init(void)
-{
-        if (!is_visws_box())
-                return -1;
-        pcibios_enable_irq = &pci_visws_enable_irq;
-        pcibios_disable_irq = &pci_visws_disable_irq;
-        pci_visws_init();
-        pcibios_init();
-        return 0;
-}
-subsys_initcall(pci_subsys_init);
diff --git a/arch/x86/vdso/Makefile b/arch/x86/vdso/Makefile
index b7ad9f89d21f..4d6ef0a336d6 100644
--- a/arch/x86/vdso/Makefile
+++ b/arch/x86/vdso/Makefile
@@ -62,7 +62,7 @@ $(obj)/%-syms.lds: $(obj)/%.so.dbg FORCE
 # Build multiple 32-bit vDSO images to choose from at boot time.
 #
 obj-$(VDSO32-y)                 += vdso32-syms.lds
-vdso32.so-$(CONFIG_X86_32)      += int80
+vdso32.so-$(VDSO32-y)           += int80
 vdso32.so-$(CONFIG_COMPAT)      += syscall
 vdso32.so-$(VDSO32-y)           += sysenter
diff --git a/arch/x86/vdso/vdso32-setup.c b/arch/x86/vdso/vdso32-setup.c
index 0bce5429a515..513f330c5832 100644
--- a/arch/x86/vdso/vdso32-setup.c
+++ b/arch/x86/vdso/vdso32-setup.c
@@ -193,17 +193,12 @@ static __init void relocate_vdso(Elf32_Ehdr *ehdr)
        }
 }
-/*
- * These symbols are defined by vdso32.S to mark the bounds
- * of the ELF DSO images included therein.
- */
-extern const char vdso32_default_start, vdso32_default_end;
-extern const char vdso32_sysenter_start, vdso32_sysenter_end;
 static struct page *vdso32_pages[1];
 #ifdef CONFIG_X86_64
 #define vdso32_sysenter()       (boot_cpu_has(X86_FEATURE_SYSENTER32))
+#define vdso32_syscall()        (boot_cpu_has(X86_FEATURE_SYSCALL32))
 /* May not be __init: called during resume */
 void syscall32_cpu_init(void)
@@ -226,6 +221,7 @@ static inline void map_compat_vdso(int map)
 #else  /* CONFIG_X86_32 */
 #define vdso32_sysenter()       (boot_cpu_has(X86_FEATURE_SEP))
+#define vdso32_syscall()        (0)
 void enable_sep_cpu(void)
 {
@@ -296,12 +292,15 @@ int __init sysenter_setup(void)
        gate_vma_init();
 #endif
-        if (!vdso32_sysenter()) {
+        if (vdso32_syscall()) {
-                vsyscall = &vdso32_default_start;
+                vsyscall = &vdso32_syscall_start;
-                vsyscall_len = &vdso32_default_end - &vdso32_default_start;
+                vsyscall_len = &vdso32_syscall_end - &vdso32_syscall_start;
-        } else {
+        } else if (vdso32_sysenter()){
                vsyscall = &vdso32_sysenter_start;
                vsyscall_len = &vdso32_sysenter_end - &vdso32_sysenter_start;
+        } else {
+                vsyscall = &vdso32_int80_start;
+                vsyscall_len = &vdso32_int80_end - &vdso32_int80_start;
        }
        memcpy(syscall_page, vsyscall, vsyscall_len);
diff --git a/arch/x86/vdso/vdso32.S b/arch/x86/vdso/vdso32.S
index 1e36f72cab86..2ce5f82c333b 100644
--- a/arch/x86/vdso/vdso32.S
+++ b/arch/x86/vdso/vdso32.S
@@ -2,14 +2,17 @@
 __INITDATA
-        .globl vdso32_default_start, vdso32_default_end
+        .globl vdso32_int80_start, vdso32_int80_end
-vdso32_default_start:
+vdso32_int80_start:
-#ifdef CONFIG_X86_32
        .incbin "arch/x86/vdso/vdso32-int80.so"
-#else
+vdso32_int80_end:
+        .globl vdso32_syscall_start, vdso32_syscall_end
+vdso32_syscall_start:
+#ifdef CONFIG_COMPAT
        .incbin "arch/x86/vdso/vdso32-syscall.so"
 #endif
-vdso32_default_end:
+vdso32_syscall_end:
        .globl vdso32_sysenter_start, vdso32_sysenter_end
 vdso32_sysenter_start:
diff --git a/arch/x86/vdso/vma.c b/arch/x86/vdso/vma.c
index 19a6cfaf5db9..257ba4a10abf 100644
--- a/arch/x86/vdso/vma.c
+++ b/arch/x86/vdso/vma.c
@@ -21,7 +21,8 @@ unsigned int __read_mostly vdso_enabled = 1;
 extern char vdso_start[], vdso_end[];
 extern unsigned short vdso_sync_cpuid;
-struct page **vdso_pages;
+static struct page **vdso_pages;
+static unsigned vdso_size;
 static inline void *var_ref(void *p, char *name)
 {
@@ -38,6 +39,7 @@ static int __init init_vdso_vars(void)
        int i;
        char *vbase;
+        vdso_size = npages << PAGE_SHIFT;
        vdso_pages = kmalloc(sizeof(struct page *) * npages, GFP_KERNEL);
        if (!vdso_pages)
                goto oom;
@@ -101,20 +103,19 @@ int arch_setup_additional_pages(struct linux_binprm *bprm, int exstack)
        struct mm_struct *mm = current->mm;
        unsigned long addr;
        int ret;
-        unsigned len = round_up(vdso_end - vdso_start, PAGE_SIZE);
        if (!vdso_enabled)
                return 0;
        down_write(&mm->mmap_sem);
-        addr = vdso_addr(mm->start_stack, len);
+        addr = vdso_addr(mm->start_stack, vdso_size);
-        addr = get_unmapped_area(NULL, addr, len, 0, 0);
+        addr = get_unmapped_area(NULL, addr, vdso_size, 0, 0);
        if (IS_ERR_VALUE(addr)) {
                ret = addr;
                goto up_fail;
        }
-        ret = install_special_mapping(mm, addr, len,
+        ret = install_special_mapping(mm, addr, vdso_size,
                                      VM_READ|VM_EXEC|
                                      VM_MAYREAD|VM_MAYWRITE|VM_MAYEXEC|
                                      VM_ALWAYSDUMP,
diff --git a/arch/x86/xen/Kconfig b/arch/x86/xen/Kconfig
index c2cc99580871..3815e425f470 100644
--- a/arch/x86/xen/Kconfig
+++ b/arch/x86/xen/Kconfig
@@ -6,8 +6,8 @@ config XEN
        bool "Xen guest support"
        select PARAVIRT
        select PARAVIRT_CLOCK
-        depends on X86_32
+        depends on X86_64 || (X86_32 && X86_PAE && !(X86_VISWS || X86_VOYAGER))
-        depends on X86_CMPXCHG && X86_TSC && X86_PAE && !(X86_VISWS || X86_VOYAGER)
+        depends on X86_CMPXCHG && X86_TSC
        help
          This is the Linux Xen port.  Enabling this will allow the
          kernel to boot in a paravirtualized environment under the
@@ -15,10 +15,16 @@ config XEN
 config XEN_MAX_DOMAIN_MEMORY
       int "Maximum allowed size of a domain in gigabytes"
-       default 8
+       default 8 if X86_32
+       default 32 if X86_64
       depends on XEN
       help
         The pseudo-physical to machine address array is sized
         according to the maximum possible memory size of a Xen
         domain.  This array uses 1 page per gigabyte, so there's no
-         need to be too stingy here.
-\ No newline at end of file
+         need to be too stingy here.
+config XEN_SAVE_RESTORE
+       bool
+       depends on PM
+       default y
+\ No newline at end of file
diff --git a/arch/x86/xen/Makefile b/arch/x86/xen/Makefile
index 2ba2d1649131..59c1e539aed2 100644
--- a/arch/x86/xen/Makefile
+++ b/arch/x86/xen/Makefile
@@ -1,4 +1,4 @@
 obj-y           := enlighten.o setup.o multicalls.o mmu.o \
-                        time.o xen-asm.o grant-table.o suspend.o
+                        time.o xen-asm_$(BITS).o grant-table.o suspend.o
 obj-$(CONFIG_SMP)       += smp.o
diff --git a/arch/x86/xen/enlighten.c b/arch/x86/xen/enlighten.c
index bb508456ef52..194bbd6e3241 100644
--- a/arch/x86/xen/enlighten.c
+++ b/arch/x86/xen/enlighten.c
@@ -33,6 +33,7 @@
 #include <xen/interface/sched.h>
 #include <xen/features.h>
 #include <xen/page.h>
+#include <xen/hvc-console.h>
 #include <asm/paravirt.h>
 #include <asm/page.h>
@@ -40,12 +41,12 @@
 #include <asm/xen/hypervisor.h>
 #include <asm/fixmap.h>
 #include <asm/processor.h>
+#include <asm/msr-index.h>
 #include <asm/setup.h>
 #include <asm/desc.h>
 #include <asm/pgtable.h>
 #include <asm/tlbflush.h>
 #include <asm/reboot.h>
-#include <asm/pgalloc.h>
 #include "xen-ops.h"
 #include "mmu.h"
@@ -57,6 +58,18 @@ DEFINE_PER_CPU(struct vcpu_info *, xen_vcpu);
 DEFINE_PER_CPU(struct vcpu_info, xen_vcpu_info);
 /*
+ * Identity map, in addition to plain kernel map.  This needs to be
+ * large enough to allocate page table pages to allocate the rest.
+ * Each page can map 2MB.
+ */
+static pte_t level1_ident_pgt[PTRS_PER_PTE * 4] __page_aligned_bss;
+#ifdef CONFIG_X86_64
+/* l3 pud for userspace vsyscall mapping */
+static pud_t level3_user_vsyscall[PTRS_PER_PUD] __page_aligned_bss;
+#endif /* CONFIG_X86_64 */
+/*
 * Note about cr3 (pagetable base) values:
 *
 * xen_cr3 contains the current logical cr3 value; it contains the
@@ -167,10 +180,14 @@ void xen_vcpu_restore(void)
 static void __init xen_banner(void)
 {
+        unsigned version = HYPERVISOR_xen_version(XENVER_version, NULL);
+        struct xen_extraversion extra;
+        HYPERVISOR_xen_version(XENVER_extraversion, &extra);
        printk(KERN_INFO "Booting paravirtualized kernel on %s\n",
               pv_info.name);
-        printk(KERN_INFO "Hypervisor signature: %s%s\n",
+        printk(KERN_INFO "Xen version: %d.%d%s%s\n",
-               xen_start_info->magic,
+               version >> 16, version & 0xffff, extra.extraversion,
               xen_feature(XENFEAT_mmu_pt_update_preserve_ad) ? " (preserve-AD)" : "");
 }
@@ -363,14 +380,6 @@ static void load_TLS_descriptor(struct thread_struct *t,
 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
@@ -379,10 +388,39 @@ static void xen_load_tls(struct thread_struct *t, unsigned int cpu)
         * 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.
+         *
+         * On x86_64, this hack is not used for %gs, because gs points
+         * to KERNEL_GS_BASE (and uses it for PDA references), so we
+         * must not zero %gs on x86_64
+         *
+         * For x86_64, we need to zero %fs, otherwise we may get an
+         * exception between the new %fs descriptor being loaded and
+         * %fs being effectively cleared at __switch_to().
         */
-        if (paravirt_get_lazy_mode() == PARAVIRT_LAZY_CPU)
+        if (paravirt_get_lazy_mode() == PARAVIRT_LAZY_CPU) {
+#ifdef CONFIG_X86_32
                loadsegment(gs, 0);
+#else
+                loadsegment(fs, 0);
+#endif
+        }
+        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);
+}
+#ifdef CONFIG_X86_64
+static void xen_load_gs_index(unsigned int idx)
+{
+        if (HYPERVISOR_set_segment_base(SEGBASE_GS_USER_SEL, idx))
+                BUG();
 }
+#endif
 static void xen_write_ldt_entry(struct desc_struct *dt, int entrynum,
                                const void *ptr)
@@ -400,23 +438,18 @@ static void xen_write_ldt_entry(struct desc_struct *dt, int entrynum,
        preempt_enable();
 }
-static int cvt_gate_to_trap(int vector, u32 low, u32 high,
+static int cvt_gate_to_trap(int vector, const gate_desc *val,
                            struct trap_info *info)
 {
-        u8 type, dpl;
+        if (val->type != 0xf && val->type != 0xe)
-        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->address = gate_offset(*val);
-        info->cs = low >> 16;
+        info->cs = gate_segment(*val);
-        info->flags = dpl;
+        info->flags = val->dpl;
        /* interrupt gates clear IF */
-        if (type == 0xe)
+        if (val->type == 0xe)
                info->flags |= 4;
        return 1;
@@ -443,11 +476,10 @@ static void xen_write_idt_entry(gate_desc *dt, int entrynum, const gate_desc *g)
        if (p >= start && (p + 8) <= end) {
                struct trap_info info[2];
-                u32 *desc = (u32 *)g;
                info[1].address = 0;
-                if (cvt_gate_to_trap(entrynum, desc[0], desc[1], &info[0]))
+                if (cvt_gate_to_trap(entrynum, g, &info[0]))
                        if (HYPERVISOR_set_trap_table(info))
                                BUG();
        }
@@ -460,13 +492,13 @@ static void xen_convert_trap_info(const struct desc_ptr *desc,
 {
        unsigned in, out, count;
-        count = (desc->size+1) / 8;
+        count = (desc->size+1) / sizeof(gate_desc);
        BUG_ON(count > 256);
        for (in = out = 0; in < count; in++) {
-                const u32 *entry = (u32 *)(desc->address + in * 8);
+                gate_desc *entry = (gate_desc*)(desc->address) + in;
-                if (cvt_gate_to_trap(in, entry[0], entry[1], &traps[out]))
+                if (cvt_gate_to_trap(in, entry, &traps[out]))
                        out++;
        }
        traps[out].address = 0;
@@ -695,33 +727,89 @@ static void set_current_cr3(void *v)
        x86_write_percpu(xen_current_cr3, (unsigned long)v);
 }
-static void xen_write_cr3(unsigned long cr3)
+static void __xen_write_cr3(bool kernel, unsigned long cr3)
 {
        struct mmuext_op *op;
        struct multicall_space mcs;
-        unsigned long mfn = pfn_to_mfn(PFN_DOWN(cr3));
+        unsigned long mfn;
-        BUG_ON(preemptible());
+        if (cr3)
+                mfn = pfn_to_mfn(PFN_DOWN(cr3));
+        else
+                mfn = 0;
-        mcs = xen_mc_entry(sizeof(*op));  /* disables interrupts */
+        WARN_ON(mfn == 0 && kernel);
-        /* Update while interrupts are disabled, so its atomic with
+        mcs = __xen_mc_entry(sizeof(*op));
-           respect to ipis */
-        x86_write_percpu(xen_cr3, cr3);
        op = mcs.args;
-        op->cmd = MMUEXT_NEW_BASEPTR;
+        op->cmd = kernel ? MMUEXT_NEW_BASEPTR : MMUEXT_NEW_USER_BASEPTR;
        op->arg1.mfn = mfn;
        MULTI_mmuext_op(mcs.mc, op, 1, NULL, DOMID_SELF);
-        /* Update xen_update_cr3 once the batch has actually
+        if (kernel) {
-           been submitted. */
+                x86_write_percpu(xen_cr3, cr3);
-        xen_mc_callback(set_current_cr3, (void *)cr3);
+                /* Update xen_current_cr3 once the batch has actually
+                   been submitted. */
+                xen_mc_callback(set_current_cr3, (void *)cr3);
+        }
+}
+static void xen_write_cr3(unsigned long cr3)
+{
+        BUG_ON(preemptible());
+        xen_mc_batch();  /* disables interrupts */
+        /* Update while interrupts are disabled, so its atomic with
+           respect to ipis */
+        x86_write_percpu(xen_cr3, cr3);
+        __xen_write_cr3(true, cr3);
+#ifdef CONFIG_X86_64
+        {
+                pgd_t *user_pgd = xen_get_user_pgd(__va(cr3));
+                if (user_pgd)
+                        __xen_write_cr3(false, __pa(user_pgd));
+                else
+                        __xen_write_cr3(false, 0);
+        }
+#endif
        xen_mc_issue(PARAVIRT_LAZY_CPU);  /* interrupts restored */
 }
+static int xen_write_msr_safe(unsigned int msr, unsigned low, unsigned high)
+{
+        int ret;
+        ret = 0;
+        switch(msr) {
+#ifdef CONFIG_X86_64
+                unsigned which;
+                u64 base;
+        case MSR_FS_BASE:               which = SEGBASE_FS; goto set;
+        case MSR_KERNEL_GS_BASE:        which = SEGBASE_GS_USER; goto set;
+        case MSR_GS_BASE:               which = SEGBASE_GS_KERNEL; goto set;
+        set:
+                base = ((u64)high << 32) | low;
+                if (HYPERVISOR_set_segment_base(which, base) != 0)
+                        ret = -EFAULT;
+                break;
+#endif
+        default:
+                ret = native_write_msr_safe(msr, low, high);
+        }
+        return ret;
+}
 /* Early in boot, while setting up the initial pagetable, assume
   everything is pinned. */
 static __init void xen_alloc_pte_init(struct mm_struct *mm, u32 pfn)
@@ -778,6 +866,48 @@ static void xen_alloc_pmd(struct mm_struct *mm, u32 pfn)
        xen_alloc_ptpage(mm, pfn, PT_PMD);
 }
+static int xen_pgd_alloc(struct mm_struct *mm)
+{
+        pgd_t *pgd = mm->pgd;
+        int ret = 0;
+        BUG_ON(PagePinned(virt_to_page(pgd)));
+#ifdef CONFIG_X86_64
+        {
+                struct page *page = virt_to_page(pgd);
+                pgd_t *user_pgd;
+                BUG_ON(page->private != 0);
+                ret = -ENOMEM;
+                user_pgd = (pgd_t *)__get_free_page(GFP_KERNEL | __GFP_ZERO);
+                page->private = (unsigned long)user_pgd;
+                if (user_pgd != NULL) {
+                        user_pgd[pgd_index(VSYSCALL_START)] =
+                                __pgd(__pa(level3_user_vsyscall) | _PAGE_TABLE);
+                        ret = 0;
+                }
+                BUG_ON(PagePinned(virt_to_page(xen_get_user_pgd(pgd))));
+        }
+#endif
+        return ret;
+}
+static void xen_pgd_free(struct mm_struct *mm, pgd_t *pgd)
+{
+#ifdef CONFIG_X86_64
+        pgd_t *user_pgd = xen_get_user_pgd(pgd);
+        if (user_pgd)
+                free_page((unsigned long)user_pgd);
+#endif
+}
 /* This should never happen until we're OK to use struct page */
 static void xen_release_ptpage(u32 pfn, unsigned level)
 {
@@ -803,6 +933,18 @@ static void xen_release_pmd(u32 pfn)
        xen_release_ptpage(pfn, PT_PMD);
 }
+#if PAGETABLE_LEVELS == 4
+static void xen_alloc_pud(struct mm_struct *mm, u32 pfn)
+{
+        xen_alloc_ptpage(mm, pfn, PT_PUD);
+}
+static void xen_release_pud(u32 pfn)
+{
+        xen_release_ptpage(pfn, PT_PUD);
+}
+#endif
 #ifdef CONFIG_HIGHPTE
 static void *xen_kmap_atomic_pte(struct page *page, enum km_type type)
 {
@@ -841,68 +983,16 @@ static __init void xen_set_pte_init(pte_t *ptep, pte_t pte)
 static __init void xen_pagetable_setup_start(pgd_t *base)
 {
-        pgd_t *xen_pgd = (pgd_t *)xen_start_info->pt_base;
-        int i;
-        /* special set_pte for pagetable initialization */
-        pv_mmu_ops.set_pte = xen_set_pte_init;
-        init_mm.pgd = base;
-        /*
-         * copy top-level of Xen-supplied pagetable into place.  This
-         * is a stand-in while we copy the pmd pages.
-         */
-        memcpy(base, xen_pgd, PTRS_PER_PGD * sizeof(pgd_t));
-        /*
-         * 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));
-        /* Unpin initial Xen pagetable */
-        pin_pagetable_pfn(MMUEXT_UNPIN_TABLE,
-                          PFN_DOWN(__pa(xen_start_info->pt_base)));
 }
 void xen_setup_shared_info(void)
 {
        if (!xen_feature(XENFEAT_auto_translated_physmap)) {
-                unsigned long addr = fix_to_virt(FIX_PARAVIRT_BOOTMAP);
+                set_fixmap(FIX_PARAVIRT_BOOTMAP,
+                           xen_start_info->shared_info);
-                /*
-                 * Create a mapping for the shared info page.
+                HYPERVISOR_shared_info =
-                 * Should be set_fixmap(), but shared_info is a machine
+                        (struct shared_info *)fix_to_virt(FIX_PARAVIRT_BOOTMAP);
-                 * address with no corresponding pseudo-phys address.
-                 */
-                set_pte_mfn(addr,
-                            PFN_DOWN(xen_start_info->shared_info),
-                            PAGE_KERNEL);
-                HYPERVISOR_shared_info = (struct shared_info *)addr;
        } else
                HYPERVISOR_shared_info =
                        (struct shared_info *)__va(xen_start_info->shared_info);
@@ -917,26 +1007,32 @@ void xen_setup_shared_info(void)
 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) */
-        pv_mmu_ops.alloc_pte = xen_alloc_pte;
-        pv_mmu_ops.alloc_pmd = xen_alloc_pmd;
-        pv_mmu_ops.release_pte = xen_release_pte;
-        pv_mmu_ops.release_pmd = xen_release_pmd;
-        pv_mmu_ops.set_pte = xen_set_pte;
        xen_setup_shared_info();
-        /* Actually pin the pagetable down, but we can't set PG_pinned
-           yet because the page structures don't exist yet. */
-        pin_pagetable_pfn(MMUEXT_PIN_L3_TABLE, PFN_DOWN(__pa(base)));
 }
 static __init void xen_post_allocator_init(void)
 {
+        pv_mmu_ops.set_pte = xen_set_pte;
        pv_mmu_ops.set_pmd = xen_set_pmd;
        pv_mmu_ops.set_pud = xen_set_pud;
+#if PAGETABLE_LEVELS == 4
+        pv_mmu_ops.set_pgd = xen_set_pgd;
+#endif
+        /* This will work as long as patching hasn't happened yet
+           (which it hasn't) */
+        pv_mmu_ops.alloc_pte = xen_alloc_pte;
+        pv_mmu_ops.alloc_pmd = xen_alloc_pmd;
+        pv_mmu_ops.release_pte = xen_release_pte;
+        pv_mmu_ops.release_pmd = xen_release_pmd;
+#if PAGETABLE_LEVELS == 4
+        pv_mmu_ops.alloc_pud = xen_alloc_pud;
+        pv_mmu_ops.release_pud = xen_release_pud;
+#endif
+#ifdef CONFIG_X86_64
+        SetPagePinned(virt_to_page(level3_user_vsyscall));
+#endif
        xen_mark_init_mm_pinned();
 }
@@ -950,6 +1046,7 @@ void xen_setup_vcpu_info_placement(void)
        /* xen_vcpu_setup managed to place the vcpu_info within the
           percpu area for all cpus, so make use of it */
+#ifdef CONFIG_X86_32
        if (have_vcpu_info_placement) {
                printk(KERN_INFO "Xen: using vcpu_info placement\n");
@@ -959,6 +1056,7 @@ void xen_setup_vcpu_info_placement(void)
                pv_irq_ops.irq_enable = xen_irq_enable_direct;
                pv_mmu_ops.read_cr2 = xen_read_cr2_direct;
        }
+#endif
 }
 static unsigned xen_patch(u8 type, u16 clobbers, void *insnbuf,
@@ -979,10 +1077,12 @@ static unsigned xen_patch(u8 type, u16 clobbers, void *insnbuf,
        goto patch_site
        switch (type) {
+#ifdef CONFIG_X86_32
                SITE(pv_irq_ops, irq_enable);
                SITE(pv_irq_ops, irq_disable);
                SITE(pv_irq_ops, save_fl);
                SITE(pv_irq_ops, restore_fl);
+#endif /* CONFIG_X86_32 */
 #undef SITE
        patch_site:
@@ -1025,8 +1125,15 @@ static void xen_set_fixmap(unsigned idx, unsigned long phys, pgprot_t prot)
 #ifdef CONFIG_X86_F00F_BUG
        case FIX_F00F_IDT:
 #endif
+#ifdef CONFIG_X86_32
        case FIX_WP_TEST:
        case FIX_VDSO:
+# ifdef CONFIG_HIGHMEM
+        case FIX_KMAP_BEGIN ... FIX_KMAP_END:
+# endif
+#else
+        case VSYSCALL_LAST_PAGE ... VSYSCALL_FIRST_PAGE:
+#endif
 #ifdef CONFIG_X86_LOCAL_APIC
        case FIX_APIC_BASE:     /* maps dummy local APIC */
 #endif
@@ -1039,6 +1146,15 @@ static void xen_set_fixmap(unsigned idx, unsigned long phys, pgprot_t prot)
        }
        __native_set_fixmap(idx, pte);
+#ifdef CONFIG_X86_64
+        /* Replicate changes to map the vsyscall page into the user
+           pagetable vsyscall mapping. */
+        if (idx >= VSYSCALL_LAST_PAGE && idx <= VSYSCALL_FIRST_PAGE) {
+                unsigned long vaddr = __fix_to_virt(idx);
+                set_pte_vaddr_pud(level3_user_vsyscall, vaddr, pte);
+        }
+#endif
 }
 static const struct pv_info xen_info __initdata = {
@@ -1084,18 +1200,25 @@ static const struct pv_cpu_ops xen_cpu_ops __initdata = {
        .wbinvd = native_wbinvd,
        .read_msr = native_read_msr_safe,
-        .write_msr = native_write_msr_safe,
+        .write_msr = xen_write_msr_safe,
        .read_tsc = native_read_tsc,
        .read_pmc = native_read_pmc,
        .iret = xen_iret,
        .irq_enable_sysexit = xen_sysexit,
+#ifdef CONFIG_X86_64
+        .usergs_sysret32 = xen_sysret32,
+        .usergs_sysret64 = xen_sysret64,
+#endif
        .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,
+#ifdef CONFIG_X86_64
+        .load_gs_index = xen_load_gs_index,
+#endif
        .store_gdt = native_store_gdt,
        .store_idt = native_store_idt,
@@ -1109,14 +1232,34 @@ static const struct pv_cpu_ops xen_cpu_ops __initdata = {
        .set_iopl_mask = xen_set_iopl_mask,
        .io_delay = xen_io_delay,
+        /* Xen takes care of %gs when switching to usermode for us */
+        .swapgs = paravirt_nop,
        .lazy_mode = {
                .enter = paravirt_enter_lazy_cpu,
                .leave = xen_leave_lazy,
        },
 };
+static void __init __xen_init_IRQ(void)
+{
+#ifdef CONFIG_X86_64
+        int i;
+        /* Create identity vector->irq map */
+        for(i = 0; i < NR_VECTORS; i++) {
+                int cpu;
+                for_each_possible_cpu(cpu)
+                        per_cpu(vector_irq, cpu)[i] = i;
+        }
+#endif  /* CONFIG_X86_64 */
+        xen_init_IRQ();
+}
 static const struct pv_irq_ops xen_irq_ops __initdata = {
-        .init_IRQ = xen_init_IRQ,
+        .init_IRQ = __xen_init_IRQ,
        .save_fl = xen_save_fl,
        .restore_fl = xen_restore_fl,
        .irq_disable = xen_irq_disable,
@@ -1124,14 +1267,13 @@ static const struct pv_irq_ops xen_irq_ops __initdata = {
        .safe_halt = xen_safe_halt,
        .halt = xen_halt,
 #ifdef CONFIG_X86_64
-        .adjust_exception_frame = paravirt_nop,
+        .adjust_exception_frame = xen_adjust_exception_frame,
 #endif
 };
 static const struct pv_apic_ops xen_apic_ops __initdata = {
 #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,
@@ -1157,8 +1299,8 @@ static const struct pv_mmu_ops xen_mmu_ops __initdata = {
        .pte_update = paravirt_nop,
        .pte_update_defer = paravirt_nop,
-        .pgd_alloc = __paravirt_pgd_alloc,
+        .pgd_alloc = xen_pgd_alloc,
-        .pgd_free = paravirt_nop,
+        .pgd_free = xen_pgd_free,
        .alloc_pte = xen_alloc_pte_init,
        .release_pte = xen_release_pte_init,
@@ -1170,7 +1312,11 @@ static const struct pv_mmu_ops xen_mmu_ops __initdata = {
        .kmap_atomic_pte = xen_kmap_atomic_pte,
 #endif
-        .set_pte = NULL,        /* see xen_pagetable_setup_* */
+#ifdef CONFIG_X86_64
+        .set_pte = xen_set_pte,
+#else
+        .set_pte = xen_set_pte_init,
+#endif
        .set_pte_at = xen_set_pte_at,
        .set_pmd = xen_set_pmd_hyper,
@@ -1184,15 +1330,26 @@ static const struct pv_mmu_ops xen_mmu_ops __initdata = {
        .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_hyper,
        .pte_clear = xen_pte_clear,
        .pmd_clear = xen_pmd_clear,
+#endif  /* CONFIG_X86_PAE */
+        .set_pud = xen_set_pud_hyper,
        .make_pmd = xen_make_pmd,
        .pmd_val = xen_pmd_val,
+#if PAGETABLE_LEVELS == 4
+        .pud_val = xen_pud_val,
+        .make_pud = xen_make_pud,
+        .set_pgd = xen_set_pgd_hyper,
+        .alloc_pud = xen_alloc_pte_init,
+        .release_pud = xen_release_pte_init,
+#endif  /* PAGETABLE_LEVELS == 4 */
        .activate_mm = xen_activate_mm,
        .dup_mmap = xen_dup_mmap,
        .exit_mmap = xen_exit_mmap,
@@ -1205,21 +1362,6 @@ static const struct pv_mmu_ops xen_mmu_ops __initdata = {
        .set_fixmap = xen_set_fixmap,
 };
-#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,
-        .send_call_func_ipi = xen_smp_send_call_function_ipi,
-        .send_call_func_single_ipi = xen_smp_send_call_function_single_ipi,
-};
-#endif  /* CONFIG_SMP */
 static void xen_reboot(int reason)
 {
        struct sched_shutdown r = { .reason = reason };
@@ -1264,6 +1406,7 @@ static const struct machine_ops __initdata xen_machine_ops = {
 static void __init xen_reserve_top(void)
 {
+#ifdef CONFIG_X86_32
        unsigned long top = HYPERVISOR_VIRT_START;
        struct xen_platform_parameters pp;
@@ -1271,8 +1414,248 @@ static void __init xen_reserve_top(void)
                top = pp.virt_start;
        reserve_top_address(-top + 2 * PAGE_SIZE);
+#endif  /* CONFIG_X86_32 */
+}
+/*
+ * Like __va(), but returns address in the kernel mapping (which is
+ * all we have until the physical memory mapping has been set up.
+ */
+static void *__ka(phys_addr_t paddr)
+{
+#ifdef CONFIG_X86_64
+        return (void *)(paddr + __START_KERNEL_map);
+#else
+        return __va(paddr);
+#endif
 }
+/* Convert a machine address to physical address */
+static unsigned long m2p(phys_addr_t maddr)
+{
+        phys_addr_t paddr;
+        maddr &= PTE_MASK;
+        paddr = mfn_to_pfn(maddr >> PAGE_SHIFT) << PAGE_SHIFT;
+        return paddr;
+}
+/* Convert a machine address to kernel virtual */
+static void *m2v(phys_addr_t maddr)
+{
+        return __ka(m2p(maddr));
+}
+#ifdef CONFIG_X86_64
+static void walk(pgd_t *pgd, unsigned long addr)
+{
+        unsigned l4idx = pgd_index(addr);
+        unsigned l3idx = pud_index(addr);
+        unsigned l2idx = pmd_index(addr);
+        unsigned l1idx = pte_index(addr);
+        pgd_t l4;
+        pud_t l3;
+        pmd_t l2;
+        pte_t l1;
+        xen_raw_printk("walk %p, %lx -> %d %d %d %d\n",
+                       pgd, addr, l4idx, l3idx, l2idx, l1idx);
+        l4 = pgd[l4idx];
+        xen_raw_printk("  l4: %016lx\n", l4.pgd);
+        xen_raw_printk("      %016lx\n", pgd_val(l4));
+        l3 = ((pud_t *)(m2v(l4.pgd)))[l3idx];
+        xen_raw_printk("  l3: %016lx\n", l3.pud);
+        xen_raw_printk("      %016lx\n", pud_val(l3));
+        l2 = ((pmd_t *)(m2v(l3.pud)))[l2idx];
+        xen_raw_printk("  l2: %016lx\n", l2.pmd);
+        xen_raw_printk("      %016lx\n", pmd_val(l2));
+        l1 = ((pte_t *)(m2v(l2.pmd)))[l1idx];
+        xen_raw_printk("  l1: %016lx\n", l1.pte);
+        xen_raw_printk("      %016lx\n", pte_val(l1));
+}
+#endif
+static void set_page_prot(void *addr, pgprot_t prot)
+{
+        unsigned long pfn = __pa(addr) >> PAGE_SHIFT;
+        pte_t pte = pfn_pte(pfn, prot);
+        xen_raw_printk("addr=%p pfn=%lx mfn=%lx prot=%016llx pte=%016llx\n",
+                       addr, pfn, get_phys_to_machine(pfn),
+                       pgprot_val(prot), pte.pte);
+        if (HYPERVISOR_update_va_mapping((unsigned long)addr, pte, 0))
+                BUG();
+}
+static __init void xen_map_identity_early(pmd_t *pmd, unsigned long max_pfn)
+{
+        unsigned pmdidx, pteidx;
+        unsigned ident_pte;
+        unsigned long pfn;
+        ident_pte = 0;
+        pfn = 0;
+        for(pmdidx = 0; pmdidx < PTRS_PER_PMD && pfn < max_pfn; pmdidx++) {
+                pte_t *pte_page;
+                /* Reuse or allocate a page of ptes */
+                if (pmd_present(pmd[pmdidx]))
+                        pte_page = m2v(pmd[pmdidx].pmd);
+                else {
+                        /* Check for free pte pages */
+                        if (ident_pte == ARRAY_SIZE(level1_ident_pgt))
+                                break;
+                        pte_page = &level1_ident_pgt[ident_pte];
+                        ident_pte += PTRS_PER_PTE;
+                        pmd[pmdidx] = __pmd(__pa(pte_page) | _PAGE_TABLE);
+                }
+                /* Install mappings */
+                for(pteidx = 0; pteidx < PTRS_PER_PTE; pteidx++, pfn++) {
+                        pte_t pte;
+                        if (pfn > max_pfn_mapped)
+                                max_pfn_mapped = pfn;
+                        if (!pte_none(pte_page[pteidx]))
+                                continue;
+                        pte = pfn_pte(pfn, PAGE_KERNEL_EXEC);
+                        pte_page[pteidx] = pte;
+                }
+        }
+        for(pteidx = 0; pteidx < ident_pte; pteidx += PTRS_PER_PTE)
+                set_page_prot(&level1_ident_pgt[pteidx], PAGE_KERNEL_RO);
+        set_page_prot(pmd, PAGE_KERNEL_RO);
+}
+#ifdef CONFIG_X86_64
+static void convert_pfn_mfn(void *v)
+{
+        pte_t *pte = v;
+        int i;
+        /* All levels are converted the same way, so just treat them
+           as ptes. */
+        for(i = 0; i < PTRS_PER_PTE; i++)
+                pte[i] = xen_make_pte(pte[i].pte);
+}
+/*
+ * Set up the inital kernel pagetable.
+ *
+ * We can construct this by grafting the Xen provided pagetable into
+ * head_64.S's preconstructed pagetables.  We copy the Xen L2's into
+ * level2_ident_pgt, level2_kernel_pgt and level2_fixmap_pgt.  This
+ * means that only the kernel has a physical mapping to start with -
+ * but that's enough to get __va working.  We need to fill in the rest
+ * of the physical mapping once some sort of allocator has been set
+ * up.
+ */
+static __init pgd_t *xen_setup_kernel_pagetable(pgd_t *pgd, unsigned long max_pfn)
+{
+        pud_t *l3;
+        pmd_t *l2;
+        /* Zap identity mapping */
+        init_level4_pgt[0] = __pgd(0);
+        /* Pre-constructed entries are in pfn, so convert to mfn */
+        convert_pfn_mfn(init_level4_pgt);
+        convert_pfn_mfn(level3_ident_pgt);
+        convert_pfn_mfn(level3_kernel_pgt);
+        l3 = m2v(pgd[pgd_index(__START_KERNEL_map)].pgd);
+        l2 = m2v(l3[pud_index(__START_KERNEL_map)].pud);
+        memcpy(level2_ident_pgt, l2, sizeof(pmd_t) * PTRS_PER_PMD);
+        memcpy(level2_kernel_pgt, l2, sizeof(pmd_t) * PTRS_PER_PMD);
+        l3 = m2v(pgd[pgd_index(__START_KERNEL_map + PMD_SIZE)].pgd);
+        l2 = m2v(l3[pud_index(__START_KERNEL_map + PMD_SIZE)].pud);
+        memcpy(level2_fixmap_pgt, l2, sizeof(pmd_t) * PTRS_PER_PMD);
+        /* Set up identity map */
+        xen_map_identity_early(level2_ident_pgt, max_pfn);
+        /* Make pagetable pieces RO */
+        set_page_prot(init_level4_pgt, PAGE_KERNEL_RO);
+        set_page_prot(level3_ident_pgt, PAGE_KERNEL_RO);
+        set_page_prot(level3_kernel_pgt, PAGE_KERNEL_RO);
+        set_page_prot(level3_user_vsyscall, PAGE_KERNEL_RO);
+        set_page_prot(level2_kernel_pgt, PAGE_KERNEL_RO);
+        set_page_prot(level2_fixmap_pgt, PAGE_KERNEL_RO);
+        /* Pin down new L4 */
+        pin_pagetable_pfn(MMUEXT_PIN_L4_TABLE,
+                          PFN_DOWN(__pa_symbol(init_level4_pgt)));
+        /* Unpin Xen-provided one */
+        pin_pagetable_pfn(MMUEXT_UNPIN_TABLE, PFN_DOWN(__pa(pgd)));
+        /* Switch over */
+        pgd = init_level4_pgt;
+        /*
+         * At this stage there can be no user pgd, and no page
+         * structure to attach it to, so make sure we just set kernel
+         * pgd.
+         */
+        xen_mc_batch();
+        __xen_write_cr3(true, __pa(pgd));
+        xen_mc_issue(PARAVIRT_LAZY_CPU);
+        reserve_early(__pa(xen_start_info->pt_base),
+                      __pa(xen_start_info->pt_base +
+                           xen_start_info->nr_pt_frames * PAGE_SIZE),
+                      "XEN PAGETABLES");
+        return pgd;
+}
+#else   /* !CONFIG_X86_64 */
+static pmd_t level2_kernel_pgt[PTRS_PER_PMD] __page_aligned_bss;
+static __init pgd_t *xen_setup_kernel_pagetable(pgd_t *pgd, unsigned long max_pfn)
+{
+        pmd_t *kernel_pmd;
+        init_pg_tables_start = __pa(pgd);
+        init_pg_tables_end = __pa(pgd) + xen_start_info->nr_pt_frames*PAGE_SIZE;
+        max_pfn_mapped = PFN_DOWN(init_pg_tables_end + 512*1024);
+        kernel_pmd = m2v(pgd[KERNEL_PGD_BOUNDARY].pgd);
+        memcpy(level2_kernel_pgt, kernel_pmd, sizeof(pmd_t) * PTRS_PER_PMD);
+        xen_map_identity_early(level2_kernel_pgt, max_pfn);
+        memcpy(swapper_pg_dir, pgd, sizeof(pgd_t) * PTRS_PER_PGD);
+        set_pgd(&swapper_pg_dir[KERNEL_PGD_BOUNDARY],
+                        __pgd(__pa(level2_kernel_pgt) | _PAGE_PRESENT));
+        set_page_prot(level2_kernel_pgt, PAGE_KERNEL_RO);
+        set_page_prot(swapper_pg_dir, PAGE_KERNEL_RO);
+        set_page_prot(empty_zero_page, PAGE_KERNEL_RO);
+        pin_pagetable_pfn(MMUEXT_UNPIN_TABLE, PFN_DOWN(__pa(pgd)));
+        xen_write_cr3(__pa(swapper_pg_dir));
+        pin_pagetable_pfn(MMUEXT_PIN_L3_TABLE, PFN_DOWN(__pa(swapper_pg_dir)));
+        return swapper_pg_dir;
+}
+#endif  /* CONFIG_X86_64 */
 /* First C function to be called on Xen boot */
 asmlinkage void __init xen_start_kernel(void)
 {
@@ -1301,53 +1684,56 @@ asmlinkage void __init xen_start_kernel(void)
        machine_ops = xen_machine_ops;
-#ifdef CONFIG_SMP
+#ifdef CONFIG_X86_64
-        smp_ops = xen_smp_ops;
+        /* Disable until direct per-cpu data access. */
+        have_vcpu_info_placement = 0;
+        x86_64_init_pda();
 #endif
+        xen_smp_init();
        /* Get mfn list */
        if (!xen_feature(XENFEAT_auto_translated_physmap))
                xen_build_dynamic_phys_to_machine();
        pgd = (pgd_t *)xen_start_info->pt_base;
-        init_pg_tables_start = __pa(pgd);
+        /* Prevent unwanted bits from being set in PTEs. */
-        init_pg_tables_end = __pa(pgd) + xen_start_info->nr_pt_frames*PAGE_SIZE;
+        __supported_pte_mask &= ~_PAGE_GLOBAL;
-        max_pfn_mapped = (init_pg_tables_end + 512*1024) >> PAGE_SHIFT;
+        if (!is_initial_xendomain())
+                __supported_pte_mask &= ~(_PAGE_PWT | _PAGE_PCD);
-        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));
-        x86_write_percpu(xen_current_cr3, __pa(pgd));
        /* Don't do the full vcpu_info placement stuff until we have a
           possible map and a non-dummy shared_info. */
        per_cpu(xen_vcpu, 0) = &HYPERVISOR_shared_info->vcpu_info[0];
+        xen_raw_console_write("mapping kernel into physical memory\n");
+        pgd = xen_setup_kernel_pagetable(pgd, xen_start_info->nr_pages);
+        init_mm.pgd = pgd;
+        /* keep using Xen gdt for now; no urgent need to change it */
        pv_info.kernel_rpl = 1;
        if (xen_feature(XENFEAT_supervisor_mode_kernel))
                pv_info.kernel_rpl = 0;
-        /* Prevent unwanted bits from being set in PTEs. */
-        __supported_pte_mask &= ~_PAGE_GLOBAL;
-        if (!is_initial_xendomain())
-                __supported_pte_mask &= ~(_PAGE_PWT | _PAGE_PCD);
        /* set the limit of our address space */
        xen_reserve_top();
+#ifdef CONFIG_X86_32
        /* 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);
+#endif
        /* Poke various useful things into boot_params */
        boot_params.hdr.type_of_loader = (9 << 4) | 0;
        boot_params.hdr.ramdisk_image = xen_start_info->mod_start
                ? __pa(xen_start_info->mod_start) : 0;
        boot_params.hdr.ramdisk_size = xen_start_info->mod_len;
+        boot_params.hdr.cmd_line_ptr = __pa(xen_start_info->cmd_line);
        if (!is_initial_xendomain()) {
                add_preferred_console("xenboot", 0, NULL);
@@ -1355,6 +1741,21 @@ asmlinkage void __init xen_start_kernel(void)
                add_preferred_console("hvc", 0, NULL);
        }
+        xen_raw_console_write("about to get started...\n");
+#if 0
+        xen_raw_printk("&boot_params=%p __pa(&boot_params)=%lx __va(__pa(&boot_params))=%lx\n",
+                       &boot_params, __pa_symbol(&boot_params),
+                       __va(__pa_symbol(&boot_params)));
+        walk(pgd, &boot_params);
+        walk(pgd, __va(__pa(&boot_params)));
+#endif
        /* Start the world */
+#ifdef CONFIG_X86_32
        i386_start_kernel();
+#else
+        x86_64_start_reservations((char *)__pa_symbol(&boot_params));
+#endif
 }
diff --git a/arch/x86/xen/mmu.c b/arch/x86/xen/mmu.c
index ff0aa74afaa1..a44d56e38bd1 100644
--- a/arch/x86/xen/mmu.c
+++ b/arch/x86/xen/mmu.c
@@ -44,8 +44,10 @@
 #include <asm/pgtable.h>
 #include <asm/tlbflush.h>
+#include <asm/fixmap.h>
 #include <asm/mmu_context.h>
 #include <asm/paravirt.h>
+#include <asm/linkage.h>
 #include <asm/xen/hypercall.h>
 #include <asm/xen/hypervisor.h>
@@ -56,26 +58,29 @@
 #include "multicalls.h"
 #include "mmu.h"
+/*
+ * Just beyond the highest usermode address.  STACK_TOP_MAX has a
+ * redzone above it, so round it up to a PGD boundary.
+ */
+#define USER_LIMIT      ((STACK_TOP_MAX + PGDIR_SIZE - 1) & PGDIR_MASK)
 #define P2M_ENTRIES_PER_PAGE    (PAGE_SIZE / sizeof(unsigned long))
 #define TOP_ENTRIES             (MAX_DOMAIN_PAGES / P2M_ENTRIES_PER_PAGE)
 /* Placeholder for holes in the address space */
-static unsigned long p2m_missing[P2M_ENTRIES_PER_PAGE]
+static unsigned long p2m_missing[P2M_ENTRIES_PER_PAGE] __page_aligned_data =
-        __attribute__((section(".data.page_aligned"))) =
                { [ 0 ... P2M_ENTRIES_PER_PAGE-1 ] = ~0UL };
 /* Array of pointers to pages containing p2m entries */
-static unsigned long *p2m_top[TOP_ENTRIES]
+static unsigned long *p2m_top[TOP_ENTRIES] __page_aligned_data =
-        __attribute__((section(".data.page_aligned"))) =
                { [ 0 ... TOP_ENTRIES - 1] = &p2m_missing[0] };
 /* Arrays of p2m arrays expressed in mfns used for save/restore */
-static unsigned long p2m_top_mfn[TOP_ENTRIES]
+static unsigned long p2m_top_mfn[TOP_ENTRIES] __page_aligned_bss;
-        __attribute__((section(".bss.page_aligned")));
-static unsigned long p2m_top_mfn_list[
+static unsigned long p2m_top_mfn_list[TOP_ENTRIES / P2M_ENTRIES_PER_PAGE]
-                        PAGE_ALIGN(TOP_ENTRIES / P2M_ENTRIES_PER_PAGE)]
+        __page_aligned_bss;
-        __attribute__((section(".bss.page_aligned")));
 static inline unsigned p2m_top_index(unsigned long pfn)
 {
@@ -181,15 +186,16 @@ void set_phys_to_machine(unsigned long pfn, unsigned long mfn)
        p2m_top[topidx][idx] = mfn;
 }
-xmaddr_t arbitrary_virt_to_machine(unsigned long address)
+xmaddr_t arbitrary_virt_to_machine(void *vaddr)
 {
+        unsigned long address = (unsigned long)vaddr;
        unsigned int level;
        pte_t *pte = lookup_address(address, &level);
        unsigned offset = address & ~PAGE_MASK;
        BUG_ON(pte == NULL);
-        return XMADDR((pte_mfn(*pte) << PAGE_SHIFT) + offset);
+        return XMADDR(((phys_addr_t)pte_mfn(*pte) << PAGE_SHIFT) + offset);
 }
 void make_lowmem_page_readonly(void *vaddr)
@@ -256,7 +262,8 @@ void xen_set_pmd_hyper(pmd_t *ptr, pmd_t val)
        xen_mc_batch();
-        u.ptr = virt_to_machine(ptr).maddr;
+        /* ptr may be ioremapped for 64-bit pagetable setup */
+        u.ptr = arbitrary_virt_to_machine(ptr).maddr;
        u.val = pmd_val_ma(val);
        extend_mmu_update(&u);
@@ -283,35 +290,7 @@ void xen_set_pmd(pmd_t *ptr, pmd_t val)
 */
 void set_pte_mfn(unsigned long vaddr, unsigned long mfn, pgprot_t flags)
 {
-        pgd_t *pgd;
+        set_pte_vaddr(vaddr, mfn_pte(mfn, flags));
-        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,
@@ -418,7 +397,8 @@ void xen_set_pud_hyper(pud_t *ptr, pud_t val)
        xen_mc_batch();
-        u.ptr = virt_to_machine(ptr).maddr;
+        /* ptr may be ioremapped for 64-bit pagetable setup */
+        u.ptr = arbitrary_virt_to_machine(ptr).maddr;
        u.val = pud_val_ma(val);
        extend_mmu_update(&u);
@@ -441,14 +421,19 @@ void xen_set_pud(pud_t *ptr, pud_t val)
 void xen_set_pte(pte_t *ptep, pte_t pte)
 {
+#ifdef CONFIG_X86_PAE
        ptep->pte_high = pte.pte_high;
        smp_wmb();
        ptep->pte_low = pte.pte_low;
+#else
+        *ptep = pte;
+#endif
 }
+#ifdef CONFIG_X86_PAE
 void xen_set_pte_atomic(pte_t *ptep, pte_t pte)
 {
-        set_64bit((u64 *)ptep, pte_val_ma(pte));
+        set_64bit((u64 *)ptep, native_pte_val(pte));
 }
 void xen_pte_clear(struct mm_struct *mm, unsigned long addr, pte_t *ptep)
@@ -462,6 +447,7 @@ void xen_pmd_clear(pmd_t *pmdp)
 {
        set_pmd(pmdp, __pmd(0));
 }
+#endif  /* CONFIG_X86_PAE */
 pmd_t xen_make_pmd(pmdval_t pmd)
 {
@@ -469,78 +455,189 @@ pmd_t xen_make_pmd(pmdval_t pmd)
        return native_make_pmd(pmd);
 }
+#if PAGETABLE_LEVELS == 4
+pudval_t xen_pud_val(pud_t pud)
+{
+        return pte_mfn_to_pfn(pud.pud);
+}
+pud_t xen_make_pud(pudval_t pud)
+{
+        pud = pte_pfn_to_mfn(pud);
+        return native_make_pud(pud);
+}
+pgd_t *xen_get_user_pgd(pgd_t *pgd)
+{
+        pgd_t *pgd_page = (pgd_t *)(((unsigned long)pgd) & PAGE_MASK);
+        unsigned offset = pgd - pgd_page;
+        pgd_t *user_ptr = NULL;
+        if (offset < pgd_index(USER_LIMIT)) {
+                struct page *page = virt_to_page(pgd_page);
+                user_ptr = (pgd_t *)page->private;
+                if (user_ptr)
+                        user_ptr += offset;
+        }
+        return user_ptr;
+}
+static void __xen_set_pgd_hyper(pgd_t *ptr, pgd_t val)
+{
+        struct mmu_update u;
+        u.ptr = virt_to_machine(ptr).maddr;
+        u.val = pgd_val_ma(val);
+        extend_mmu_update(&u);
+}
+/*
+ * Raw hypercall-based set_pgd, intended for in early boot before
+ * there's a page structure.  This implies:
+ *  1. The only existing pagetable is the kernel's
+ *  2. It is always pinned
+ *  3. It has no user pagetable attached to it
+ */
+void __init xen_set_pgd_hyper(pgd_t *ptr, pgd_t val)
+{
+        preempt_disable();
+        xen_mc_batch();
+        __xen_set_pgd_hyper(ptr, val);
+        xen_mc_issue(PARAVIRT_LAZY_MMU);
+        preempt_enable();
+}
+void xen_set_pgd(pgd_t *ptr, pgd_t val)
+{
+        pgd_t *user_ptr = xen_get_user_pgd(ptr);
+        /* If page is not pinned, we can just update the entry
+           directly */
+        if (!page_pinned(ptr)) {
+                *ptr = val;
+                if (user_ptr) {
+                        WARN_ON(page_pinned(user_ptr));
+                        *user_ptr = val;
+                }
+                return;
+        }
+        /* If it's pinned, then we can at least batch the kernel and
+           user updates together. */
+        xen_mc_batch();
+        __xen_set_pgd_hyper(ptr, val);
+        if (user_ptr)
+                __xen_set_pgd_hyper(user_ptr, val);
+        xen_mc_issue(PARAVIRT_LAZY_MMU);
+}
+#endif  /* PAGETABLE_LEVELS == 4 */
 /*
-  (Yet another) pagetable walker.  This one is intended for pinning a
+ * (Yet another) pagetable walker.  This one is intended for pinning a
-  pagetable.  This means that it walks a pagetable and calls the
+ * pagetable.  This means that it walks a pagetable and calls the
-  callback function on each page it finds making up the page table,
+ * callback function on each page it finds making up the page table,
-  at every level.  It walks the entire pagetable, but it only bothers
+ * at every level.  It walks the entire pagetable, but it only bothers
-  pinning pte pages which are below pte_limit.  In the normal case
+ * pinning pte pages which are below limit.  In the normal case this
-  this will be TASK_SIZE, but at boot we need to pin up to
+ * will be STACK_TOP_MAX, but at boot we need to pin up to
-  FIXADDR_TOP.  But the important bit is that we don't pin beyond
+ * FIXADDR_TOP.
-  there, because then we start getting into Xen's ptes.
+ *
-*/
+ * For 32-bit the important bit is that we don't pin beyond there,
-static int pgd_walk(pgd_t *pgd_base, int (*func)(struct page *, enum pt_level),
+ * because then we start getting into Xen's ptes.
+ *
+ * For 64-bit, we must skip the Xen hole in the middle of the address
+ * space, just after the big x86-64 virtual hole.
+ */
+static int pgd_walk(pgd_t *pgd, int (*func)(struct page *, enum pt_level),
                    unsigned long limit)
 {
-        pgd_t *pgd = pgd_base;
        int flush = 0;
-        unsigned long addr = 0;
+        unsigned hole_low, hole_high;
-        unsigned long pgd_next;
+        unsigned pgdidx_limit, pudidx_limit, pmdidx_limit;
+        unsigned pgdidx, pudidx, pmdidx;
-        BUG_ON(limit > FIXADDR_TOP);
+        /* The limit is the last byte to be touched */
+        limit--;
+        BUG_ON(limit >= FIXADDR_TOP);
        if (xen_feature(XENFEAT_auto_translated_physmap))
                return 0;
-        for (; addr != FIXADDR_TOP; pgd++, addr = pgd_next) {
+        /*
+         * 64-bit has a great big hole in the middle of the address
+         * space, which contains the Xen mappings.  On 32-bit these
+         * will end up making a zero-sized hole and so is a no-op.
+         */
+        hole_low = pgd_index(USER_LIMIT);
+        hole_high = pgd_index(PAGE_OFFSET);
+        pgdidx_limit = pgd_index(limit);
+#if PTRS_PER_PUD > 1
+        pudidx_limit = pud_index(limit);
+#else
+        pudidx_limit = 0;
+#endif
+#if PTRS_PER_PMD > 1
+        pmdidx_limit = pmd_index(limit);
+#else
+        pmdidx_limit = 0;
+#endif
+        flush |= (*func)(virt_to_page(pgd), PT_PGD);
+        for (pgdidx = 0; pgdidx <= pgdidx_limit; pgdidx++) {
                pud_t *pud;
-                unsigned long pud_limit, pud_next;
-                pgd_next = pud_limit = pgd_addr_end(addr, FIXADDR_TOP);
+                if (pgdidx >= hole_low && pgdidx < hole_high)
+                        continue;
-                if (!pgd_val(*pgd))
+                if (!pgd_val(pgd[pgdidx]))
                        continue;
-                pud = pud_offset(pgd, 0);
+                pud = pud_offset(&pgd[pgdidx], 0);
                if (PTRS_PER_PUD > 1) /* not folded */
                        flush |= (*func)(virt_to_page(pud), PT_PUD);
-                for (; addr != pud_limit; pud++, addr = pud_next) {
+                for (pudidx = 0; pudidx < PTRS_PER_PUD; pudidx++) {
                        pmd_t *pmd;
-                        unsigned long pmd_limit;
-                        pud_next = pud_addr_end(addr, pud_limit);
+                        if (pgdidx == pgdidx_limit &&
+                            pudidx > pudidx_limit)
-                        if (pud_next < limit)
+                                goto out;
-                                pmd_limit = pud_next;
-                        else
-                                pmd_limit = limit;
-                        if (pud_none(*pud))
+                        if (pud_none(pud[pudidx]))
                                continue;
-                        pmd = pmd_offset(pud, 0);
+                        pmd = pmd_offset(&pud[pudidx], 0);
                        if (PTRS_PER_PMD > 1) /* not folded */
                                flush |= (*func)(virt_to_page(pmd), PT_PMD);
-                        for (; addr != pmd_limit; pmd++) {
+                        for (pmdidx = 0; pmdidx < PTRS_PER_PMD; pmdidx++) {
-                                addr += (PAGE_SIZE * PTRS_PER_PTE);
+                                struct page *pte;
-                                if ((pmd_limit-1) < (addr-1)) {
-                                        addr = pmd_limit;
+                                if (pgdidx == pgdidx_limit &&
-                                        break;
+                                    pudidx == pudidx_limit &&
-                                }
+                                    pmdidx > pmdidx_limit)
+                                        goto out;
-                                if (pmd_none(*pmd))
+                                if (pmd_none(pmd[pmdidx]))
                                        continue;
-                                flush |= (*func)(pmd_page(*pmd), PT_PTE);
+                                pte = pmd_page(pmd[pmdidx]);
+                                flush |= (*func)(pte, PT_PTE);
                        }
                }
        }
+out:
-        flush |= (*func)(virt_to_page(pgd_base), PT_PGD);
        return flush;
 }
@@ -622,14 +719,31 @@ void xen_pgd_pin(pgd_t *pgd)
 {
        xen_mc_batch();
-        if (pgd_walk(pgd, pin_page, TASK_SIZE)) {
+        if (pgd_walk(pgd, pin_page, USER_LIMIT)) {
                /* re-enable interrupts for kmap_flush_unused */
                xen_mc_issue(0);
                kmap_flush_unused();
                xen_mc_batch();
        }
+#ifdef CONFIG_X86_64
+        {
+                pgd_t *user_pgd = xen_get_user_pgd(pgd);
+                xen_do_pin(MMUEXT_PIN_L4_TABLE, PFN_DOWN(__pa(pgd)));
+                if (user_pgd) {
+                        pin_page(virt_to_page(user_pgd), PT_PGD);
+                        xen_do_pin(MMUEXT_PIN_L4_TABLE, PFN_DOWN(__pa(user_pgd)));
+                }
+        }
+#else /* CONFIG_X86_32 */
+#ifdef CONFIG_X86_PAE
+        /* Need to make sure unshared kernel PMD is pinnable */
+        pin_page(virt_to_page(pgd_page(pgd[pgd_index(TASK_SIZE)])), PT_PMD);
+#endif
        xen_do_pin(MMUEXT_PIN_L3_TABLE, PFN_DOWN(__pa(pgd)));
+#endif /* CONFIG_X86_64 */
        xen_mc_issue(0);
 }
@@ -656,9 +770,11 @@ void xen_mm_pin_all(void)
        spin_unlock_irqrestore(&pgd_lock, flags);
 }
-/* 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 init_mm pagetable is really pinned as soon as its created, but
-   the book-keeping now. */
+ * 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, enum pt_level level)
 {
        SetPagePinned(page);
@@ -708,7 +824,23 @@ static void xen_pgd_unpin(pgd_t *pgd)
        xen_do_pin(MMUEXT_UNPIN_TABLE, PFN_DOWN(__pa(pgd)));
-        pgd_walk(pgd, unpin_page, TASK_SIZE);
+#ifdef CONFIG_X86_64
+        {
+                pgd_t *user_pgd = xen_get_user_pgd(pgd);
+                if (user_pgd) {
+                        xen_do_pin(MMUEXT_UNPIN_TABLE, PFN_DOWN(__pa(user_pgd)));
+                        unpin_page(virt_to_page(user_pgd), PT_PGD);
+                }
+        }
+#endif
+#ifdef CONFIG_X86_PAE
+        /* Need to make sure unshared kernel PMD is unpinned */
+        pin_page(virt_to_page(pgd_page(pgd[pgd_index(TASK_SIZE)])), PT_PMD);
+#endif
+        pgd_walk(pgd, unpin_page, USER_LIMIT);
        xen_mc_issue(0);
 }
@@ -727,7 +859,6 @@ void xen_mm_unpin_all(void)
        list_for_each_entry(page, &pgd_list, lru) {
                if (PageSavePinned(page)) {
                        BUG_ON(!PagePinned(page));
-                        printk("unpinning pinned %p\n", page_address(page));
                        xen_pgd_unpin((pgd_t *)page_address(page));
                        ClearPageSavePinned(page);
                }
@@ -757,8 +888,15 @@ void xen_dup_mmap(struct mm_struct *oldmm, struct mm_struct *mm)
 static void drop_other_mm_ref(void *info)
 {
        struct mm_struct *mm = info;
+        struct mm_struct *active_mm;
+#ifdef CONFIG_X86_64
+        active_mm = read_pda(active_mm);
+#else
+        active_mm = __get_cpu_var(cpu_tlbstate).active_mm;
+#endif
-        if (__get_cpu_var(cpu_tlbstate).active_mm == mm)
+        if (active_mm == mm)
                leave_mm(smp_processor_id());
        /* If this cpu still has a stale cr3 reference, then make sure
diff --git a/arch/x86/xen/mmu.h b/arch/x86/xen/mmu.h
index 297bf9f5b8bc..0f59bd03f9e3 100644
--- a/arch/x86/xen/mmu.h
+++ b/arch/x86/xen/mmu.h
@@ -10,18 +10,6 @@ enum pt_level {
        PT_PTE
 };
-/*
- * 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);
@@ -44,13 +32,26 @@ pgd_t xen_make_pgd(pgdval_t);
 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);
+#ifdef CONFIG_X86_PAE
 void xen_set_pte_atomic(pte_t *ptep, pte_t pte);
+void xen_pte_clear(struct mm_struct *mm, unsigned long addr, pte_t *ptep);
+void xen_pmd_clear(pmd_t *pmdp);
+#endif  /* CONFIG_X86_PAE */
 void xen_set_pmd(pmd_t *pmdp, pmd_t pmdval);
 void xen_set_pud(pud_t *ptr, pud_t val);
 void xen_set_pmd_hyper(pmd_t *pmdp, pmd_t pmdval);
 void xen_set_pud_hyper(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);
+#if PAGETABLE_LEVELS == 4
+pudval_t xen_pud_val(pud_t pud);
+pud_t xen_make_pud(pudval_t pudval);
+void xen_set_pgd(pgd_t *pgdp, pgd_t pgd);
+void xen_set_pgd_hyper(pgd_t *pgdp, pgd_t pgd);
+#endif
+pgd_t *xen_get_user_pgd(pgd_t *pgd);
 pte_t xen_ptep_modify_prot_start(struct mm_struct *mm, unsigned long addr, pte_t *ptep);
 void  xen_ptep_modify_prot_commit(struct mm_struct *mm, unsigned long addr,
diff --git a/arch/x86/xen/multicalls.c b/arch/x86/xen/multicalls.c
index 3c63c4da7ed1..9efd1c6c9776 100644
--- a/arch/x86/xen/multicalls.c
+++ b/arch/x86/xen/multicalls.c
@@ -76,6 +76,7 @@ void xen_mc_flush(void)
                if (ret) {
                        printk(KERN_ERR "%d multicall(s) failed: cpu %d\n",
                               ret, smp_processor_id());
+                        dump_stack();
                        for (i = 0; i < b->mcidx; i++) {
                                printk("  call %2d/%d: op=%lu arg=[%lx] result=%ld\n",
                                       i+1, b->mcidx,
diff --git a/arch/x86/xen/setup.c b/arch/x86/xen/setup.c
index e0a39595bde3..b6acc3a0af46 100644
--- a/arch/x86/xen/setup.c
+++ b/arch/x86/xen/setup.c
@@ -83,30 +83,72 @@ static void xen_idle(void)
 /*
 * Set the bit indicating "nosegneg" library variants should be used.
+ * We only need to bother in pure 32-bit mode; compat 32-bit processes
+ * can have un-truncated segments, so wrapping around is allowed.
 */
 static void __init fiddle_vdso(void)
 {
-        extern const char vdso32_default_start;
+#ifdef CONFIG_X86_32
-        u32 *mask = VDSO32_SYMBOL(&vdso32_default_start, NOTE_MASK);
+        u32 *mask;
+        mask = VDSO32_SYMBOL(&vdso32_int80_start, NOTE_MASK);
        *mask |= 1 << VDSO_NOTE_NONEGSEG_BIT;
+        mask = VDSO32_SYMBOL(&vdso32_sysenter_start, NOTE_MASK);
+        *mask |= 1 << VDSO_NOTE_NONEGSEG_BIT;
+#endif
 }
-void xen_enable_sysenter(void)
+static __cpuinit int register_callback(unsigned type, const void *func)
 {
-        int cpu = smp_processor_id();
+        struct callback_register callback = {
-        extern void xen_sysenter_target(void);
+                .type = type,
-        /* Mask events on entry, even though they get enabled immediately */
+                .address = XEN_CALLBACK(__KERNEL_CS, func),
-        static struct callback_register sysenter = {
-                .type = CALLBACKTYPE_sysenter,
-                .address = { __KERNEL_CS, (unsigned long)xen_sysenter_target },
                .flags = CALLBACKF_mask_events,
        };
-        if (!boot_cpu_has(X86_FEATURE_SEP) ||
+        return HYPERVISOR_callback_op(CALLBACKOP_register, &callback);
-            HYPERVISOR_callback_op(CALLBACKOP_register, &sysenter) != 0) {
+}
-                clear_cpu_cap(&cpu_data(cpu), X86_FEATURE_SEP);
-                clear_cpu_cap(&boot_cpu_data, X86_FEATURE_SEP);
+void __cpuinit xen_enable_sysenter(void)
+{
+        extern void xen_sysenter_target(void);
+        int ret;
+        unsigned sysenter_feature;
+#ifdef CONFIG_X86_32
+        sysenter_feature = X86_FEATURE_SEP;
+#else
+        sysenter_feature = X86_FEATURE_SYSENTER32;
+#endif
+        if (!boot_cpu_has(sysenter_feature))
+                return;
+        ret = register_callback(CALLBACKTYPE_sysenter, xen_sysenter_target);
+        if(ret != 0)
+                setup_clear_cpu_cap(sysenter_feature);
+}
+void __cpuinit xen_enable_syscall(void)
+{
+#ifdef CONFIG_X86_64
+        int ret;
+        extern void xen_syscall_target(void);
+        extern void xen_syscall32_target(void);
+        ret = register_callback(CALLBACKTYPE_syscall, xen_syscall_target);
+        if (ret != 0) {
+                printk(KERN_ERR "Failed to set syscall callback: %d\n", ret);
+                /* Pretty fatal; 64-bit userspace has no other
+                   mechanism for syscalls. */
        }
+        if (boot_cpu_has(X86_FEATURE_SYSCALL32)) {
+                ret = register_callback(CALLBACKTYPE_syscall32,
+                                        xen_syscall32_target);
+                if (ret != 0)
+                        setup_clear_cpu_cap(X86_FEATURE_SYSCALL32);
+        }
+#endif /* CONFIG_X86_64 */
 }
 void __init xen_arch_setup(void)
@@ -120,10 +162,12 @@ void __init xen_arch_setup(void)
        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,
+        if (register_callback(CALLBACKTYPE_event, xen_hypervisor_callback) ||
-                                 __KERNEL_CS, (unsigned long)xen_failsafe_callback);
+            register_callback(CALLBACKTYPE_failsafe, xen_failsafe_callback))
+                BUG();
        xen_enable_sysenter();
+        xen_enable_syscall();
        set_iopl.iopl = 1;
        rc = HYPERVISOR_physdev_op(PHYSDEVOP_set_iopl, &set_iopl);
@@ -143,11 +187,6 @@ void __init xen_arch_setup(void)
        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
index 233156f39b7f..f702199312a5 100644
--- a/arch/x86/xen/smp.c
+++ b/arch/x86/xen/smp.c
@@ -66,13 +66,22 @@ static __cpuinit void cpu_bringup_and_idle(void)
        int cpu = smp_processor_id();
        cpu_init();
+        preempt_disable();
        xen_enable_sysenter();
+        xen_enable_syscall();
-        preempt_disable();
+        cpu = smp_processor_id();
-        per_cpu(cpu_state, cpu) = CPU_ONLINE;
+        smp_store_cpu_info(cpu);
+        cpu_data(cpu).x86_max_cores = 1;
+        set_cpu_sibling_map(cpu);
        xen_setup_cpu_clockevents();
+        cpu_set(cpu, cpu_online_map);
+        x86_write_percpu(cpu_state, CPU_ONLINE);
+        wmb();
        /* We can take interrupts now: we're officially "up". */
        local_irq_enable();
@@ -141,56 +150,37 @@ static int xen_smp_intr_init(unsigned int cpu)
        return rc;
 }
-void __init xen_fill_possible_map(void)
+static 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)
+                if (rc >= 0) {
+                        num_processors++;
                        cpu_set(i, cpu_possible_map);
+                }
        }
 }
-void __init xen_smp_prepare_boot_cpu(void)
+static 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);
+        make_lowmem_page_readwrite(&per_cpu_var(gdt_page));
-        for_each_possible_cpu(cpu) {
-                cpus_clear(per_cpu(cpu_sibling_map, cpu));
-                /*
-                 * cpu_core_map lives in a per cpu area that is cleared
-                 * when the per cpu array is allocated.
-                 *
-                 * cpus_clear(per_cpu(cpu_core_map, cpu));
-                 */
-        }
        xen_setup_vcpu_info_placement();
 }
-void __init xen_smp_prepare_cpus(unsigned int max_cpus)
+static void __init xen_smp_prepare_cpus(unsigned int max_cpus)
 {
        unsigned cpu;
-        for_each_possible_cpu(cpu) {
-                cpus_clear(per_cpu(cpu_sibling_map, cpu));
-                /*
-                 * cpu_core_ map will be zeroed when the per
-                 * cpu area is allocated.
-                 *
-                 * cpus_clear(per_cpu(cpu_core_map, cpu));
-                 */
-        }
        smp_store_cpu_info(0);
+        cpu_data(0).x86_max_cores = 1;
        set_cpu_sibling_map(0);
        if (xen_smp_intr_init(0))
@@ -225,7 +215,7 @@ 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);
+        struct desc_struct *gdt;
        if (cpu_test_and_set(cpu, xen_cpu_initialized_map))
                return 0;
@@ -234,12 +224,15 @@ cpu_initialize_context(unsigned int cpu, struct task_struct *idle)
        if (ctxt == NULL)
                return -ENOMEM;
+        gdt = get_cpu_gdt_table(cpu);
        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;
+#ifdef CONFIG_X86_32
+        ctxt->user_regs.fs = __KERNEL_PERCPU;
+#endif
        ctxt->user_regs.eip = (unsigned long)cpu_bringup_and_idle;
        ctxt->user_regs.eflags = 0x1000; /* IOPL_RING1 */
@@ -249,11 +242,11 @@ cpu_initialize_context(unsigned int cpu, struct task_struct *idle)
        ctxt->ldt_ents = 0;
-        BUG_ON((unsigned long)gdt->gdt & ~PAGE_MASK);
+        BUG_ON((unsigned long)gdt & ~PAGE_MASK);
-        make_lowmem_page_readonly(gdt->gdt);
+        make_lowmem_page_readonly(gdt);
-        ctxt->gdt_frames[0] = virt_to_mfn(gdt->gdt);
+        ctxt->gdt_frames[0] = virt_to_mfn(gdt);
-        ctxt->gdt_ents      = ARRAY_SIZE(gdt->gdt);
+        ctxt->gdt_ents      = GDT_ENTRIES;
        ctxt->user_regs.cs = __KERNEL_CS;
        ctxt->user_regs.esp = idle->thread.sp0 - sizeof(struct pt_regs);
@@ -261,9 +254,11 @@ cpu_initialize_context(unsigned int cpu, struct task_struct *idle)
        ctxt->kernel_ss = __KERNEL_DS;
        ctxt->kernel_sp = idle->thread.sp0;
+#ifdef CONFIG_X86_32
        ctxt->event_callback_cs     = __KERNEL_CS;
-        ctxt->event_callback_eip    = (unsigned long)xen_hypervisor_callback;
        ctxt->failsafe_callback_cs  = __KERNEL_CS;
+#endif
+        ctxt->event_callback_eip    = (unsigned long)xen_hypervisor_callback;
        ctxt->failsafe_callback_eip = (unsigned long)xen_failsafe_callback;
        per_cpu(xen_cr3, cpu) = __pa(swapper_pg_dir);
@@ -276,7 +271,7 @@ cpu_initialize_context(unsigned int cpu, struct task_struct *idle)
        return 0;
 }
-int __cpuinit xen_cpu_up(unsigned int cpu)
+static int __cpuinit xen_cpu_up(unsigned int cpu)
 {
        struct task_struct *idle = idle_task(cpu);
        int rc;
@@ -287,11 +282,28 @@ int __cpuinit xen_cpu_up(unsigned int cpu)
                return rc;
 #endif
+#ifdef CONFIG_X86_64
+        /* Allocate node local memory for AP pdas */
+        WARN_ON(cpu == 0);
+        if (cpu > 0) {
+                rc = get_local_pda(cpu);
+                if (rc)
+                        return rc;
+        }
+#endif
+#ifdef CONFIG_X86_32
        init_gdt(cpu);
        per_cpu(current_task, cpu) = idle;
        irq_ctx_init(cpu);
+#else
+        cpu_pda(cpu)->pcurrent = idle;
+        clear_tsk_thread_flag(idle, TIF_FORK);
+#endif
        xen_setup_timer(cpu);
+        per_cpu(cpu_state, cpu) = CPU_UP_PREPARE;
        /* make sure interrupts start blocked */
        per_cpu(xen_vcpu, cpu)->evtchn_upcall_mask = 1;
@@ -306,20 +318,18 @@ int __cpuinit xen_cpu_up(unsigned int 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);
+        while(per_cpu(cpu_state, cpu) != CPU_ONLINE) {
+                HYPERVISOR_sched_op(SCHEDOP_yield, 0);
+                barrier();
+        }
        return 0;
 }
-void xen_smp_cpus_done(unsigned int max_cpus)
+static void xen_smp_cpus_done(unsigned int max_cpus)
 {
 }
@@ -335,12 +345,12 @@ static void stop_self(void *v)
        BUG();
 }
-void xen_smp_send_stop(void)
+static void xen_smp_send_stop(void)
 {
        smp_call_function(stop_self, NULL, 0);
 }
-void xen_smp_send_reschedule(int cpu)
+static void xen_smp_send_reschedule(int cpu)
 {
        xen_send_IPI_one(cpu, XEN_RESCHEDULE_VECTOR);
 }
@@ -355,7 +365,7 @@ static void xen_send_IPI_mask(cpumask_t mask, enum ipi_vector vector)
                xen_send_IPI_one(cpu, vector);
 }
-void xen_smp_send_call_function_ipi(cpumask_t mask)
+static void xen_smp_send_call_function_ipi(cpumask_t mask)
 {
        int cpu;
@@ -370,7 +380,7 @@ void xen_smp_send_call_function_ipi(cpumask_t mask)
        }
 }
-void xen_smp_send_call_function_single_ipi(int cpu)
+static void xen_smp_send_call_function_single_ipi(int cpu)
 {
        xen_send_IPI_mask(cpumask_of_cpu(cpu), XEN_CALL_FUNCTION_SINGLE_VECTOR);
 }
@@ -379,7 +389,11 @@ static irqreturn_t xen_call_function_interrupt(int irq, void *dev_id)
 {
        irq_enter();
        generic_smp_call_function_interrupt();
+#ifdef CONFIG_X86_32
        __get_cpu_var(irq_stat).irq_call_count++;
+#else
+        add_pda(irq_call_count, 1);
+#endif
        irq_exit();
        return IRQ_HANDLED;
@@ -389,8 +403,31 @@ static irqreturn_t xen_call_function_single_interrupt(int irq, void *dev_id)
 {
        irq_enter();
        generic_smp_call_function_single_interrupt();
+#ifdef CONFIG_X86_32
        __get_cpu_var(irq_stat).irq_call_count++;
+#else
+        add_pda(irq_call_count, 1);
+#endif
        irq_exit();
        return IRQ_HANDLED;
 }
+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,
+        .send_call_func_ipi = xen_smp_send_call_function_ipi,
+        .send_call_func_single_ipi = xen_smp_send_call_function_single_ipi,
+};
+void __init xen_smp_init(void)
+{
+        smp_ops = xen_smp_ops;
+        xen_fill_possible_map();
+}
diff --git a/arch/x86/xen/suspend.c b/arch/x86/xen/suspend.c
index 251669a932d4..2a234db5949b 100644
--- a/arch/x86/xen/suspend.c
+++ b/arch/x86/xen/suspend.c
@@ -38,8 +38,11 @@ void xen_post_suspend(int suspend_cancelled)
                xen_cpu_initialized_map = cpu_online_map;
 #endif
                xen_vcpu_restore();
-                xen_timer_resume();
        }
 }
+void xen_arch_resume(void)
+{
+        /* nothing */
+}
diff --git a/arch/x86/xen/xen-asm.S b/arch/x86/xen/xen-asm_32.S
index 2497a30f41de..2497a30f41de 100644
--- a/arch/x86/xen/xen-asm.S
+++ b/arch/x86/xen/xen-asm_32.S
diff --git a/arch/x86/xen/xen-asm_64.S b/arch/x86/xen/xen-asm_64.S
new file mode 100644
index 000000000000..4038cbfe3331
--- /dev/null
+++ b/arch/x86/xen/xen-asm_64.S
@@ -0,0 +1,271 @@
+/*
+        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/processor-flags.h>
+#include <asm/errno.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
+#if 0
+#include <asm/percpu.h>
+/*
+        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)
+        /* Unmask events */
+        movb $0, 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. */
+        /* Test for pending */
+        testb $0xff, PER_CPU_VAR(xen_vcpu_info, XEN_vcpu_info_pending)
+        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)
+/*
+        Force an event check by making a hypercall,
+        but preserve regs before making the call.
+ */
+check_events:
+        push %rax
+        push %rcx
+        push %rdx
+        push %rsi
+        push %rdi
+        push %r8
+        push %r9
+        push %r10
+        push %r11
+        call force_evtchn_callback
+        pop %r11
+        pop %r10
+        pop %r9
+        pop %r8
+        pop %rdi
+        pop %rsi
+        pop %rdx
+        pop %rcx
+        pop %rax
+        ret
+#endif
+ENTRY(xen_adjust_exception_frame)
+        mov 8+0(%rsp),%rcx
+        mov 8+8(%rsp),%r11
+        ret $16
+hypercall_iret = hypercall_page + __HYPERVISOR_iret * 32
+/*
+        Xen64 iret frame:
+        ss
+        rsp
+        rflags
+        cs
+        rip             <-- standard iret frame
+        flags
+        rcx             }
+        r11             }<-- pushed by hypercall page
+rsp ->  rax             }
+ */
+ENTRY(xen_iret)
+        pushq $0
+1:      jmp hypercall_iret
+ENDPATCH(xen_iret)
+RELOC(xen_iret, 1b+1)
+/*
+        sysexit is not used for 64-bit processes, so it's
+        only ever used to return to 32-bit compat userspace.
+ */
+ENTRY(xen_sysexit)
+        pushq $__USER32_DS
+        pushq %rcx
+        pushq $X86_EFLAGS_IF
+        pushq $__USER32_CS
+        pushq %rdx
+        pushq $VGCF_in_syscall
+1:      jmp hypercall_iret
+ENDPATCH(xen_sysexit)
+RELOC(xen_sysexit, 1b+1)
+ENTRY(xen_sysret64)
+        /* We're already on the usermode stack at this point, but still
+           with the kernel gs, so we can easily switch back */
+        movq %rsp, %gs:pda_oldrsp
+        movq %gs:pda_kernelstack,%rsp
+        pushq $__USER_DS
+        pushq %gs:pda_oldrsp
+        pushq %r11
+        pushq $__USER_CS
+        pushq %rcx
+        pushq $VGCF_in_syscall
+1:      jmp hypercall_iret
+ENDPATCH(xen_sysret64)
+RELOC(xen_sysret64, 1b+1)
+ENTRY(xen_sysret32)
+        /* We're already on the usermode stack at this point, but still
+           with the kernel gs, so we can easily switch back */
+        movq %rsp, %gs:pda_oldrsp
+        movq %gs:pda_kernelstack, %rsp
+        pushq $__USER32_DS
+        pushq %gs:pda_oldrsp
+        pushq %r11
+        pushq $__USER32_CS
+        pushq %rcx
+        pushq $VGCF_in_syscall
+1:      jmp hypercall_iret
+ENDPATCH(xen_sysret32)
+RELOC(xen_sysret32, 1b+1)
+/*
+        Xen handles syscall callbacks much like ordinary exceptions,
+        which means we have:
+         - kernel gs
+         - kernel rsp
+         - an iret-like stack frame on the stack (including rcx and r11):
+                ss
+                rsp
+                rflags
+                cs
+                rip
+                r11
+        rsp->   rcx
+        In all the entrypoints, we undo all that to make it look
+        like a CPU-generated syscall/sysenter and jump to the normal
+        entrypoint.
+ */
+.macro undo_xen_syscall
+        mov 0*8(%rsp),%rcx
+        mov 1*8(%rsp),%r11
+        mov 5*8(%rsp),%rsp
+.endm
+/* Normal 64-bit system call target */
+ENTRY(xen_syscall_target)
+        undo_xen_syscall
+        jmp system_call_after_swapgs
+ENDPROC(xen_syscall_target)
+#ifdef CONFIG_IA32_EMULATION
+/* 32-bit compat syscall target */
+ENTRY(xen_syscall32_target)
+        undo_xen_syscall
+        jmp ia32_cstar_target
+ENDPROC(xen_syscall32_target)
+/* 32-bit compat sysenter target */
+ENTRY(xen_sysenter_target)
+        undo_xen_syscall
+        jmp ia32_sysenter_target
+ENDPROC(xen_sysenter_target)
+#else /* !CONFIG_IA32_EMULATION */
+ENTRY(xen_syscall32_target)
+ENTRY(xen_sysenter_target)
+        lea 16(%rsp), %rsp      /* strip %rcx,%r11 */
+        mov $-ENOSYS, %rax
+        pushq $VGCF_in_syscall
+        jmp hypercall_iret
+ENDPROC(xen_syscall32_target)
+ENDPROC(xen_sysenter_target)
+#endif  /* CONFIG_IA32_EMULATION */
diff --git a/arch/x86/xen/xen-head.S b/arch/x86/xen/xen-head.S
index 7c0cf6320a0a..63d49a523ed3 100644
--- a/arch/x86/xen/xen-head.S
+++ b/arch/x86/xen/xen-head.S
@@ -5,15 +5,24 @@
 #include <linux/elfnote.h>
 #include <linux/init.h>
 #include <asm/boot.h>
+#include <asm/asm.h>
+#include <asm/page.h>
 #include <xen/interface/elfnote.h>
 #include <asm/xen/interface.h>
        __INIT
 ENTRY(startup_xen)
-        movl %esi,xen_start_info
        cld
-        movl $(init_thread_union+THREAD_SIZE),%esp
+#ifdef CONFIG_X86_32
+        mov %esi,xen_start_info
+        mov $init_thread_union+THREAD_SIZE,%esp
+#else
+        mov %rsi,xen_start_info
+        mov $init_thread_union+THREAD_SIZE,%rsp
+#endif
        jmp xen_start_kernel
        __FINIT
@@ -21,21 +30,26 @@ ENTRY(startup_xen)
 .pushsection .text
        .align PAGE_SIZE_asm
 ENTRY(hypercall_page)
-        .skip 0x1000
+        .skip PAGE_SIZE_asm
 .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)
+#ifdef CONFIG_X86_32
-        ELFNOTE(Xen, XEN_ELFNOTE_ENTRY,          .long  startup_xen)
+        ELFNOTE(Xen, XEN_ELFNOTE_VIRT_BASE,      _ASM_PTR __PAGE_OFFSET)
-        ELFNOTE(Xen, XEN_ELFNOTE_HYPERCALL_PAGE, .long  hypercall_page)
+#else
+        ELFNOTE(Xen, XEN_ELFNOTE_VIRT_BASE,      _ASM_PTR __START_KERNEL_map)
+#endif
+        ELFNOTE(Xen, XEN_ELFNOTE_ENTRY,          _ASM_PTR startup_xen)
+        ELFNOTE(Xen, XEN_ELFNOTE_HYPERCALL_PAGE, _ASM_PTR hypercall_page)
        ELFNOTE(Xen, XEN_ELFNOTE_FEATURES,       .asciz "!writable_page_tables|pae_pgdir_above_4gb")
        ELFNOTE(Xen, XEN_ELFNOTE_PAE_MODE,       .asciz "yes")
        ELFNOTE(Xen, XEN_ELFNOTE_LOADER,         .asciz "generic")
        ELFNOTE(Xen, XEN_ELFNOTE_L1_MFN_VALID,
                .quad _PAGE_PRESENT; .quad _PAGE_PRESENT)
        ELFNOTE(Xen, XEN_ELFNOTE_SUSPEND_CANCEL, .long 1)
-        ELFNOTE(Xen, XEN_ELFNOTE_HV_START_LOW,   .long __HYPERVISOR_VIRT_START)
+        ELFNOTE(Xen, XEN_ELFNOTE_HV_START_LOW,   _ASM_PTR __HYPERVISOR_VIRT_START)
+        ELFNOTE(Xen, XEN_ELFNOTE_PADDR_OFFSET,   _ASM_PTR 0)
 #endif /*CONFIG_XEN */
diff --git a/arch/x86/xen/xen-ops.h b/arch/x86/xen/xen-ops.h
index 6f4b1045c1c2..dd3c23152a2e 100644
--- a/arch/x86/xen/xen-ops.h
+++ b/arch/x86/xen/xen-ops.h
@@ -26,6 +26,7 @@ char * __init xen_memory_setup(void);
 void __init xen_arch_setup(void);
 void __init xen_init_IRQ(void);
 void xen_enable_sysenter(void);
+void xen_enable_syscall(void);
 void xen_vcpu_restore(void);
 void __init xen_build_dynamic_phys_to_machine(void);
@@ -37,7 +38,6 @@ 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_timer_resume(void);
 irqreturn_t xen_debug_interrupt(int irq, void *dev_id);
@@ -45,20 +45,15 @@ bool xen_vcpu_stolen(int vcpu);
 void xen_mark_init_mm_pinned(void);
-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);
+#ifdef CONFIG_SMP
-void xen_smp_send_reschedule(int cpu);
+void xen_smp_init(void);
-void xen_smp_send_call_function_ipi(cpumask_t mask);
-void xen_smp_send_call_function_single_ipi(int cpu);
 extern cpumask_t xen_cpu_initialized_map;
+#else
+static inline void xen_smp_init(void) {}
+#endif
 /* Declare an asm function, along with symbols needed to make it
@@ -73,7 +68,11 @@ 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);
+/* These are not functions, and cannot be called normally */
 void xen_iret(void);
 void xen_sysexit(void);
+void xen_sysret32(void);
+void xen_sysret64(void);
+void xen_adjust_exception_frame(void);
 #endif /* XEN_OPS_H */