Merge branch 'linus' into x86/pat2

Conflicts:
	arch/x86/mm/init_64.c

59 files changed, 1997 insertions, 1007 deletions

diff --git a/arch/x86/kernel/acpi/boot.c b/arch/x86/kernel/acpi/boot.c
index bfd10fd211cd..7d40ef7b36e3 100644
--- a/arch/x86/kernel/acpi/boot.c
+++ b/arch/x86/kernel/acpi/boot.c
@@ -58,7 +58,6 @@ EXPORT_SYMBOL(acpi_disabled);
 #ifdef  CONFIG_X86_64
 
 #include <asm/proto.h>
-#include <asm/genapic.h>
 
 #else                           /* X86 */
 
@@ -97,8 +96,6 @@ static u64 acpi_lapic_addr __initdata = APIC_DEFAULT_PHYS_BASE;
 #warning ACPI uses CMPXCHG, i486 and later hardware
 #endif
 
-static int acpi_mcfg_64bit_base_addr __initdata = FALSE;
-
 /* --------------------------------------------------------------------------
                               Boot-time Configuration
    -------------------------------------------------------------------------- */
@@ -160,6 +157,8 @@ char *__init __acpi_map_table(unsigned long phys, unsigned long size)
 struct acpi_mcfg_allocation *pci_mmcfg_config;
 int pci_mmcfg_config_num;
 
+static int acpi_mcfg_64bit_base_addr __initdata = FALSE;
+
 static int __init acpi_mcfg_oem_check(struct acpi_table_mcfg *mcfg)
 {
         if (!strcmp(mcfg->header.oem_id, "SGI"))
@@ -1605,6 +1604,14 @@ static struct dmi_system_id __initdata acpi_dmi_table[] = {
          */
         {
          .callback = dmi_ignore_irq0_timer_override,
+         .ident = "HP nx6115 laptop",
+         .matches = {
+                     DMI_MATCH(DMI_SYS_VENDOR, "Hewlett-Packard"),
+                     DMI_MATCH(DMI_PRODUCT_NAME, "HP Compaq nx6115"),
+                     },
+         },
+        {
+         .callback = dmi_ignore_irq0_timer_override,
          .ident = "HP NX6125 laptop",
          .matches = {
                      DMI_MATCH(DMI_SYS_VENDOR, "Hewlett-Packard"),
@@ -1619,6 +1626,14 @@ static struct dmi_system_id __initdata acpi_dmi_table[] = {
                      DMI_MATCH(DMI_PRODUCT_NAME, "HP Compaq nx6325"),
                      },
          },
+        {
+         .callback = dmi_ignore_irq0_timer_override,
+         .ident = "HP 6715b laptop",
+         .matches = {
+                     DMI_MATCH(DMI_SYS_VENDOR, "Hewlett-Packard"),
+                     DMI_MATCH(DMI_PRODUCT_NAME, "HP Compaq 6715b"),
+                     },
+         },
         {}
 };
 
diff --git a/arch/x86/kernel/alternative.c b/arch/x86/kernel/alternative.c
index 65a0c1b48696..fb04e49776ba 100644
--- a/arch/x86/kernel/alternative.c
+++ b/arch/x86/kernel/alternative.c
@@ -231,25 +231,25 @@ static void alternatives_smp_lock(u8 **start, u8 **end, u8 *text, u8 *text_end)
                         continue;
                 if (*ptr > text_end)
                         continue;
-                text_poke(*ptr, ((unsigned char []){0xf0}), 1); /* add lock prefix */
+                /* turn DS segment override prefix into lock prefix */
+                text_poke(*ptr, ((unsigned char []){0xf0}), 1);
         };
 }
 
 static void alternatives_smp_unlock(u8 **start, u8 **end, u8 *text, u8 *text_end)
 {
         u8 **ptr;
-        char insn[1];
 
         if (noreplace_smp)
                 return;
 
-        add_nops(insn, 1);
         for (ptr = start; ptr < end; ptr++) {
                 if (*ptr < text)
                         continue;
                 if (*ptr > text_end)
                         continue;
-                text_poke(*ptr, insn, 1);
+                /* turn lock prefix into DS segment override prefix */
+                text_poke(*ptr, ((unsigned char []){0x3E}), 1);
         };
 }
 
diff --git a/arch/x86/kernel/amd_iommu.c b/arch/x86/kernel/amd_iommu.c
index 69b4d060b21c..042fdc27bc92 100644
--- a/arch/x86/kernel/amd_iommu.c
+++ b/arch/x86/kernel/amd_iommu.c
@@ -101,10 +101,10 @@ static int iommu_queue_command(struct amd_iommu *iommu, struct iommu_cmd *cmd)
  */
 static int iommu_completion_wait(struct amd_iommu *iommu)
 {
-        int ret, ready = 0;
+        int ret = 0, ready = 0;
         unsigned status = 0;
         struct iommu_cmd cmd;
-        unsigned long i = 0;
+        unsigned long flags, i = 0;
 
         memset(&cmd, 0, sizeof(cmd));
         cmd.data[0] = CMD_COMPL_WAIT_INT_MASK;
@@ -112,10 +112,12 @@ static int iommu_completion_wait(struct amd_iommu *iommu)
 
         iommu->need_sync = 0;
 
-        ret = iommu_queue_command(iommu, &cmd);
+        spin_lock_irqsave(&iommu->lock, flags);
+
+        ret = __iommu_queue_command(iommu, &cmd);
 
         if (ret)
-                return ret;
+                goto out;
 
         while (!ready && (i < EXIT_LOOP_COUNT)) {
                 ++i;
@@ -130,6 +132,8 @@ static int iommu_completion_wait(struct amd_iommu *iommu)
 
         if (unlikely((i == EXIT_LOOP_COUNT) && printk_ratelimit()))
                 printk(KERN_WARNING "AMD IOMMU: Completion wait loop failed\n");
+out:
+        spin_unlock_irqrestore(&iommu->lock, flags);
 
         return 0;
 }
@@ -140,6 +144,7 @@ static int iommu_completion_wait(struct amd_iommu *iommu)
 static int iommu_queue_inv_dev_entry(struct amd_iommu *iommu, u16 devid)
 {
         struct iommu_cmd cmd;
+        int ret;
 
         BUG_ON(iommu == NULL);
 
@@ -147,9 +152,11 @@ static int iommu_queue_inv_dev_entry(struct amd_iommu *iommu, u16 devid)
         CMD_SET_TYPE(&cmd, CMD_INV_DEV_ENTRY);
         cmd.data[0] = devid;
 
+        ret = iommu_queue_command(iommu, &cmd);
+
         iommu->need_sync = 1;
 
-        return iommu_queue_command(iommu, &cmd);
+        return ret;
 }
 
 /*
@@ -159,6 +166,7 @@ static int iommu_queue_inv_iommu_pages(struct amd_iommu *iommu,
                 u64 address, u16 domid, int pde, int s)
 {
         struct iommu_cmd cmd;
+        int ret;
 
         memset(&cmd, 0, sizeof(cmd));
         address &= PAGE_MASK;
@@ -171,9 +179,11 @@ static int iommu_queue_inv_iommu_pages(struct amd_iommu *iommu,
         if (pde) /* PDE bit - we wan't flush everything not only the PTEs */
                 cmd.data[2] |= CMD_INV_IOMMU_PAGES_PDE_MASK;
 
+        ret = iommu_queue_command(iommu, &cmd);
+
         iommu->need_sync = 1;
 
-        return iommu_queue_command(iommu, &cmd);
+        return ret;
 }
 
 /*
diff --git a/arch/x86/kernel/aperture_64.c b/arch/x86/kernel/aperture_64.c
index 44e21826db11..9a32b37ee2ee 100644
--- a/arch/x86/kernel/aperture_64.c
+++ b/arch/x86/kernel/aperture_64.c
@@ -455,11 +455,11 @@ out:
                    force_iommu ||
                    valid_agp ||
                    fallback_aper_force) {
-                printk(KERN_ERR
+                printk(KERN_INFO
                         "Your BIOS doesn't leave a aperture memory hole\n");
-                printk(KERN_ERR
+                printk(KERN_INFO
                         "Please enable the IOMMU option in the BIOS setup\n");
-                printk(KERN_ERR
+                printk(KERN_INFO
                         "This costs you %d MB of RAM\n",
                                 32 << fallback_aper_order);
 
diff --git a/arch/x86/kernel/apm_32.c b/arch/x86/kernel/apm_32.c
index 9ee24e6bc4b0..5145a6e72bbb 100644
--- a/arch/x86/kernel/apm_32.c
+++ b/arch/x86/kernel/apm_32.c
@@ -228,12 +228,12 @@
 #include <linux/suspend.h>
 #include <linux/kthread.h>
 #include <linux/jiffies.h>
-#include <linux/smp_lock.h>
 
 #include <asm/system.h>
 #include <asm/uaccess.h>
 #include <asm/desc.h>
 #include <asm/i8253.h>
+#include <asm/olpc.h>
 #include <asm/paravirt.h>
 #include <asm/reboot.h>
 
@@ -2217,7 +2217,7 @@ static int __init apm_init(void)
 
         dmi_check_system(apm_dmi_table);
 
-        if (apm_info.bios.version == 0 || paravirt_enabled()) {
+        if (apm_info.bios.version == 0 || paravirt_enabled() || machine_is_olpc()) {
                 printk(KERN_INFO "apm: BIOS not found.\n");
                 return -ENODEV;
         }
diff --git a/arch/x86/kernel/asm-offsets_64.c b/arch/x86/kernel/asm-offsets_64.c
index aa89387006fe..505543a75a56 100644
--- a/arch/x86/kernel/asm-offsets_64.c
+++ b/arch/x86/kernel/asm-offsets_64.c
@@ -22,7 +22,7 @@
 
 #define __NO_STUBS 1
 #undef __SYSCALL
-#undef _ASM_X86_64_UNISTD_H_
+#undef ASM_X86__UNISTD_64_H
 #define __SYSCALL(nr, sym) [nr] = 1,
 static char syscalls[] = {
 #include <asm/unistd.h>
diff --git a/arch/x86/kernel/bios_uv.c b/arch/x86/kernel/bios_uv.c
index c639bd55391c..fdd585f9c53d 100644
--- a/arch/x86/kernel/bios_uv.c
+++ b/arch/x86/kernel/bios_uv.c
@@ -25,11 +25,11 @@ 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;
+        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;
 }
diff --git a/arch/x86/kernel/cpu/common.c b/arch/x86/kernel/cpu/common.c
index 8aab8517642e..4e456bd955bb 100644
--- a/arch/x86/kernel/cpu/common.c
+++ b/arch/x86/kernel/cpu/common.c
@@ -344,31 +344,15 @@ static void __init early_cpu_detect(void)
 
 /*
  * The NOPL instruction is supposed to exist on all CPUs with
- * family >= 6, unfortunately, that's not true in practice because
+ * family >= 6; unfortunately, that's not true in practice because
  * of early VIA chips and (more importantly) broken virtualizers that
- * are not easy to detect.  Hence, probe for it based on first
- * principles.
+ * are not easy to detect.  In the latter case it doesn't even *fail*
+ * reliably, so probing for it doesn't even work.  Disable it completely
+ * unless we can find a reliable way to detect all the broken cases.
  */
 static void __cpuinit detect_nopl(struct cpuinfo_x86 *c)
 {
-        const u32 nopl_signature = 0x888c53b1; /* Random number */
-        u32 has_nopl = nopl_signature;
-
         clear_cpu_cap(c, X86_FEATURE_NOPL);
-        if (c->x86 >= 6) {
-                asm volatile("\n"
-                             "1:      .byte 0x0f,0x1f,0xc0\n" /* nopl %eax */
-                             "2:\n"
-                             "        .section .fixup,\"ax\"\n"
-                             "3:      xor %0,%0\n"
-                             "        jmp 2b\n"
-                             "        .previous\n"
-                             _ASM_EXTABLE(1b,3b)
-                             : "+a" (has_nopl));
-
-                if (has_nopl == nopl_signature)
-                        set_cpu_cap(c, X86_FEATURE_NOPL);
-        }
 }
 
 static void __cpuinit generic_identify(struct cpuinfo_x86 *c)
diff --git a/arch/x86/kernel/cpu/common_64.c b/arch/x86/kernel/cpu/common_64.c
index a11f5d4477cd..305b465889b0 100644
--- a/arch/x86/kernel/cpu/common_64.c
+++ b/arch/x86/kernel/cpu/common_64.c
@@ -430,6 +430,49 @@ static __init int setup_noclflush(char *arg)
 }
 __setup("noclflush", setup_noclflush);
 
+struct msr_range {
+        unsigned min;
+        unsigned max;
+};
+
+static struct msr_range msr_range_array[] __cpuinitdata = {
+        { 0x00000000, 0x00000418},
+        { 0xc0000000, 0xc000040b},
+        { 0xc0010000, 0xc0010142},
+        { 0xc0011000, 0xc001103b},
+};
+
+static void __cpuinit print_cpu_msr(void)
+{
+        unsigned index;
+        u64 val;
+        int i;
+        unsigned index_min, index_max;
+
+        for (i = 0; i < ARRAY_SIZE(msr_range_array); i++) {
+                index_min = msr_range_array[i].min;
+                index_max = msr_range_array[i].max;
+                for (index = index_min; index < index_max; index++) {
+                        if (rdmsrl_amd_safe(index, &val))
+                                continue;
+                        printk(KERN_INFO " MSR%08x: %016llx\n", index, val);
+                }
+        }
+}
+
+static int show_msr __cpuinitdata;
+static __init int setup_show_msr(char *arg)
+{
+        int num;
+
+        get_option(&arg, &num);
+
+        if (num > 0)
+                show_msr = num;
+        return 1;
+}
+__setup("show_msr=", setup_show_msr);
+
 void __cpuinit print_cpu_info(struct cpuinfo_x86 *c)
 {
         if (c->x86_model_id[0])
@@ -439,6 +482,14 @@ void __cpuinit print_cpu_info(struct cpuinfo_x86 *c)
                 printk(KERN_CONT " stepping %02x\n", c->x86_mask);
         else
                 printk(KERN_CONT "\n");
+
+#ifdef CONFIG_SMP
+        if (c->cpu_index < show_msr)
+                print_cpu_msr();
+#else
+        if (show_msr)
+                print_cpu_msr();
+#endif
 }
 
 static __init int setup_disablecpuid(char *arg)
diff --git a/arch/x86/kernel/cpu/cpufreq/p4-clockmod.c b/arch/x86/kernel/cpu/cpufreq/p4-clockmod.c
index f1685fb91fbd..b8e05ee4f736 100644
--- a/arch/x86/kernel/cpu/cpufreq/p4-clockmod.c
+++ b/arch/x86/kernel/cpu/cpufreq/p4-clockmod.c
@@ -171,7 +171,7 @@ static unsigned int cpufreq_p4_get_frequency(struct cpuinfo_x86 *c)
         }
 
         if (c->x86 != 0xF) {
-                printk(KERN_WARNING PFX "Unknown p4-clockmod-capable CPU. Please send an e-mail to <cpufreq@lists.linux.org.uk>\n");
+                printk(KERN_WARNING PFX "Unknown p4-clockmod-capable CPU. Please send an e-mail to <cpufreq@vger.kernel.org>\n");
                 return 0;
         }
 
diff --git a/arch/x86/kernel/cpu/cpufreq/speedstep-centrino.c b/arch/x86/kernel/cpu/cpufreq/speedstep-centrino.c
index 15e13c01cc36..3b5f06423e77 100644
--- a/arch/x86/kernel/cpu/cpufreq/speedstep-centrino.c
+++ b/arch/x86/kernel/cpu/cpufreq/speedstep-centrino.c
@@ -26,7 +26,7 @@
 #include <asm/cpufeature.h>
 
 #define PFX             "speedstep-centrino: "
-#define MAINTAINER      "cpufreq@lists.linux.org.uk"
+#define MAINTAINER      "cpufreq@vger.kernel.org"
 
 #define dprintk(msg...) \
         cpufreq_debug_printk(CPUFREQ_DEBUG_DRIVER, "speedstep-centrino", msg)
diff --git a/arch/x86/kernel/cpu/intel.c b/arch/x86/kernel/cpu/intel.c
index b75f2569b8f8..f113ef4595f6 100644
--- a/arch/x86/kernel/cpu/intel.c
+++ b/arch/x86/kernel/cpu/intel.c
@@ -222,10 +222,11 @@ static void __cpuinit init_intel(struct cpuinfo_x86 *c)
                         set_cpu_cap(c, X86_FEATURE_BTS);
                 if (!(l1 & (1<<12)))
                         set_cpu_cap(c, X86_FEATURE_PEBS);
+                ds_init_intel(c);
         }
 
         if (cpu_has_bts)
-                ds_init_intel(c);
+                ptrace_bts_init_intel(c);
 
         /*
          * See if we have a good local APIC by checking for buggy Pentia,
diff --git a/arch/x86/kernel/cpu/mtrr/generic.c b/arch/x86/kernel/cpu/mtrr/generic.c
index cb7d3b6a80eb..4e8d77f01eeb 100644
--- a/arch/x86/kernel/cpu/mtrr/generic.c
+++ b/arch/x86/kernel/cpu/mtrr/generic.c
@@ -401,12 +401,7 @@ static void generic_get_mtrr(unsigned int reg, unsigned long *base,
                 tmp |= ~((1<<(hi - 1)) - 1);
 
                 if (tmp != mask_lo) {
-                        static int once = 1;
-
-                        if (once) {
-                                printk(KERN_INFO "mtrr: your BIOS has set up an incorrect mask, fixing it up.\n");
-                                once = 0;
-                        }
+                        WARN_ONCE(1, KERN_INFO "mtrr: your BIOS has set up an incorrect mask, fixing it up.\n");
                         mask_lo = tmp;
                 }
         }
diff --git a/arch/x86/kernel/cpu/mtrr/if.c b/arch/x86/kernel/cpu/mtrr/if.c
index 84c480bb3715..4c4214690dd1 100644
--- a/arch/x86/kernel/cpu/mtrr/if.c
+++ b/arch/x86/kernel/cpu/mtrr/if.c
@@ -405,9 +405,9 @@ static int mtrr_seq_show(struct seq_file *seq, void *offset)
                         }
                         /* RED-PEN: base can be > 32bit */ 
                         len += seq_printf(seq, 
-                                   "reg%02i: base=0x%05lx000 (%4luMB), size=%4lu%cB: %s, count=%d\n",
+                                   "reg%02i: base=0x%06lx000 (%5luMB), size=%5lu%cB, count=%d: %s\n",
                              i, base, base >> (20 - PAGE_SHIFT), size, factor,
-                             mtrr_attrib_to_str(type), mtrr_usage_table[i]);
+                             mtrr_usage_table[i], mtrr_attrib_to_str(type));
                 }
         }
         return 0;
diff --git a/arch/x86/kernel/cpu/mtrr/main.c b/arch/x86/kernel/cpu/mtrr/main.c
index b117d7f8a564..c78c04821ea1 100644
--- a/arch/x86/kernel/cpu/mtrr/main.c
+++ b/arch/x86/kernel/cpu/mtrr/main.c
@@ -729,7 +729,7 @@ struct var_mtrr_range_state {
         mtrr_type type;
 };
 
-struct var_mtrr_range_state __initdata range_state[RANGE_NUM];
+static struct var_mtrr_range_state __initdata range_state[RANGE_NUM];
 static int __initdata debug_print;
 
 static int __init
@@ -759,7 +759,8 @@ x86_get_mtrr_mem_range(struct res_range *range, int nr_range,
         /* take out UC ranges */
         for (i = 0; i < num_var_ranges; i++) {
                 type = range_state[i].type;
-                if (type != MTRR_TYPE_UNCACHABLE)
+                if (type != MTRR_TYPE_UNCACHABLE &&
+                    type != MTRR_TYPE_WRPROT)
                         continue;
                 size = range_state[i].size_pfn;
                 if (!size)
@@ -834,7 +835,14 @@ static int __init enable_mtrr_cleanup_setup(char *str)
                 enable_mtrr_cleanup = 1;
         return 0;
 }
-early_param("enble_mtrr_cleanup", enable_mtrr_cleanup_setup);
+early_param("enable_mtrr_cleanup", enable_mtrr_cleanup_setup);
+
+static int __init mtrr_cleanup_debug_setup(char *str)
+{
+        debug_print = 1;
+        return 0;
+}
+early_param("mtrr_cleanup_debug", mtrr_cleanup_debug_setup);
 
 struct var_mtrr_state {
         unsigned long   range_startk;
@@ -898,6 +906,27 @@ set_var_mtrr_all(unsigned int address_bits)
         }
 }
 
+static unsigned long to_size_factor(unsigned long sizek, char *factorp)
+{
+        char factor;
+        unsigned long base = sizek;
+
+        if (base & ((1<<10) - 1)) {
+                /* not MB alignment */
+                factor = 'K';
+        } else if (base & ((1<<20) - 1)){
+                factor = 'M';
+                base >>= 10;
+        } else {
+                factor = 'G';
+                base >>= 20;
+        }
+
+        *factorp = factor;
+
+        return base;
+}
+
 static unsigned int __init
 range_to_mtrr(unsigned int reg, unsigned long range_startk,
               unsigned long range_sizek, unsigned char type)
@@ -919,13 +948,21 @@ range_to_mtrr(unsigned int reg, unsigned long range_startk,
                         align = max_align;
 
                 sizek = 1 << align;
-                if (debug_print)
+                if (debug_print) {
+                        char start_factor = 'K', size_factor = 'K';
+                        unsigned long start_base, size_base;
+
+                        start_base = to_size_factor(range_startk, &start_factor),
+                        size_base = to_size_factor(sizek, &size_factor),
+
                         printk(KERN_DEBUG "Setting variable MTRR %d, "
-                                "base: %ldMB, range: %ldMB, type %s\n",
-                                reg, range_startk >> 10, sizek >> 10,
+                                "base: %ld%cB, range: %ld%cB, type %s\n",
+                                reg, start_base, start_factor,
+                                size_base, size_factor,
                                 (type == MTRR_TYPE_UNCACHABLE)?"UC":
                                     ((type == MTRR_TYPE_WRBACK)?"WB":"Other")
                                 );
+                }
                 save_var_mtrr(reg++, range_startk, sizek, type);
                 range_startk += sizek;
                 range_sizek -= sizek;
@@ -970,6 +1007,8 @@ range_to_mtrr_with_hole(struct var_mtrr_state *state, unsigned long basek,
         /* try to append some small hole */
         range0_basek = state->range_startk;
         range0_sizek = ALIGN(state->range_sizek, chunk_sizek);
+
+        /* no increase */
         if (range0_sizek == state->range_sizek) {
                 if (debug_print)
                         printk(KERN_DEBUG "rangeX: %016lx - %016lx\n",
@@ -980,13 +1019,40 @@ range_to_mtrr_with_hole(struct var_mtrr_state *state, unsigned long basek,
                 return 0;
         }
 
-        range0_sizek -= chunk_sizek;
-        if (range0_sizek && sizek) {
-            while (range0_basek + range0_sizek > (basek + sizek)) {
-                range0_sizek -= chunk_sizek;
-                if (!range0_sizek)
-                        break;
-            }
+        /* only cut back, when it is not the last */
+        if (sizek) {
+                while (range0_basek + range0_sizek > (basek + sizek)) {
+                        if (range0_sizek >= chunk_sizek)
+                                range0_sizek -= chunk_sizek;
+                        else
+                                range0_sizek = 0;
+
+                        if (!range0_sizek)
+                                break;
+                }
+        }
+
+second_try:
+        range_basek = range0_basek + range0_sizek;
+
+        /* one hole in the middle */
+        if (range_basek > basek && range_basek <= (basek + sizek))
+                second_sizek = range_basek - basek;
+
+        if (range0_sizek > state->range_sizek) {
+
+                /* one hole in middle or at end */
+                hole_sizek = range0_sizek - state->range_sizek - second_sizek;
+
+                /* hole size should be less than half of range0 size */
+                if (hole_sizek >= (range0_sizek >> 1) &&
+                    range0_sizek >= chunk_sizek) {
+                        range0_sizek -= chunk_sizek;
+                        second_sizek = 0;
+                        hole_sizek = 0;
+
+                        goto second_try;
+                }
         }
 
         if (range0_sizek) {
@@ -996,50 +1062,28 @@ range_to_mtrr_with_hole(struct var_mtrr_state *state, unsigned long basek,
                                 (range0_basek + range0_sizek)<<10);
                 state->reg = range_to_mtrr(state->reg, range0_basek,
                                 range0_sizek, MTRR_TYPE_WRBACK);
-
-        }
-
-        range_basek = range0_basek + range0_sizek;
-        range_sizek = chunk_sizek;
-
-        if (range_basek + range_sizek > basek &&
-            range_basek + range_sizek <= (basek + sizek)) {
-                /* one hole */
-                second_basek = basek;
-                second_sizek = range_basek + range_sizek - basek;
         }
 
-        /* if last piece, only could one hole near end */
-        if ((second_basek || !basek) &&
-            range_sizek - (state->range_sizek - range0_sizek) - second_sizek <
-            (chunk_sizek >> 1)) {
-                /*
-                 * one hole in middle (second_sizek is 0) or at end
-                 * (second_sizek is 0 )
-                 */
-                hole_sizek = range_sizek - (state->range_sizek - range0_sizek)
-                                 - second_sizek;
-                hole_basek = range_basek + range_sizek - hole_sizek
-                                 - second_sizek;
-        } else {
-                /* fallback for big hole, or several holes */
+        if (range0_sizek < state->range_sizek) {
+                /* need to handle left over */
                 range_sizek = state->range_sizek - range0_sizek;
-                second_basek = 0;
-                second_sizek = 0;
+
+                if (debug_print)
+                        printk(KERN_DEBUG "range: %016lx - %016lx\n",
+                                 range_basek<<10,
+                                 (range_basek + range_sizek)<<10);
+                state->reg = range_to_mtrr(state->reg, range_basek,
+                                 range_sizek, MTRR_TYPE_WRBACK);
         }
 
-        if (debug_print)
-                printk(KERN_DEBUG "range: %016lx - %016lx\n", range_basek<<10,
-                         (range_basek + range_sizek)<<10);
-        state->reg = range_to_mtrr(state->reg, range_basek, range_sizek,
-                                         MTRR_TYPE_WRBACK);
         if (hole_sizek) {
+                hole_basek = range_basek - hole_sizek - second_sizek;
                 if (debug_print)
                         printk(KERN_DEBUG "hole: %016lx - %016lx\n",
-                                 hole_basek<<10, (hole_basek + hole_sizek)<<10);
-                state->reg = range_to_mtrr(state->reg, hole_basek, hole_sizek,
-                                                 MTRR_TYPE_UNCACHABLE);
-
+                                 hole_basek<<10,
+                                 (hole_basek + hole_sizek)<<10);
+                state->reg = range_to_mtrr(state->reg, hole_basek,
+                                 hole_sizek, MTRR_TYPE_UNCACHABLE);
         }
 
         return second_sizek;
@@ -1154,11 +1198,11 @@ struct mtrr_cleanup_result {
 };
 
 /*
- * gran_size: 1M, 2M, ..., 2G
- * chunk size: gran_size, ..., 4G
- * so we need (2+13)*6
+ * gran_size: 64K, 128K, 256K, 512K, 1M, 2M, ..., 2G
+ * chunk size: gran_size, ..., 2G
+ * so we need (1+16)*8
  */
-#define NUM_RESULT      90
+#define NUM_RESULT      136
 #define PSHIFT          (PAGE_SHIFT - 10)
 
 static struct mtrr_cleanup_result __initdata result[NUM_RESULT];
@@ -1168,13 +1212,14 @@ static unsigned long __initdata min_loss_pfn[RANGE_NUM];
 static int __init mtrr_cleanup(unsigned address_bits)
 {
         unsigned long extra_remove_base, extra_remove_size;
-        unsigned long i, base, size, def, dummy;
+        unsigned long base, size, def, dummy;
         mtrr_type type;
         int nr_range, nr_range_new;
         u64 chunk_size, gran_size;
         unsigned long range_sums, range_sums_new;
         int index_good;
         int num_reg_good;
+        int i;
 
         /* extra one for all 0 */
         int num[MTRR_NUM_TYPES + 1];
@@ -1204,6 +1249,8 @@ static int __init mtrr_cleanup(unsigned address_bits)
                         continue;
                 if (!size)
                         type = MTRR_NUM_TYPES;
+                if (type == MTRR_TYPE_WRPROT)
+                        type = MTRR_TYPE_UNCACHABLE;
                 num[type]++;
         }
 
@@ -1216,23 +1263,57 @@ static int __init mtrr_cleanup(unsigned address_bits)
                 num_var_ranges - num[MTRR_NUM_TYPES])
                 return 0;
 
+        /* print original var MTRRs at first, for debugging: */
+        printk(KERN_DEBUG "original variable MTRRs\n");
+        for (i = 0; i < num_var_ranges; i++) {
+                char start_factor = 'K', size_factor = 'K';
+                unsigned long start_base, size_base;
+
+                size_base = range_state[i].size_pfn << (PAGE_SHIFT - 10);
+                if (!size_base)
+                        continue;
+
+                size_base = to_size_factor(size_base, &size_factor),
+                start_base = range_state[i].base_pfn << (PAGE_SHIFT - 10);
+                start_base = to_size_factor(start_base, &start_factor),
+                type = range_state[i].type;
+
+                printk(KERN_DEBUG "reg %d, base: %ld%cB, range: %ld%cB, type %s\n",
+                        i, start_base, start_factor,
+                        size_base, size_factor,
+                        (type == MTRR_TYPE_UNCACHABLE) ? "UC" :
+                            ((type == MTRR_TYPE_WRPROT) ? "WP" :
+                             ((type == MTRR_TYPE_WRBACK) ? "WB" : "Other"))
+                        );
+        }
+
         memset(range, 0, sizeof(range));
         extra_remove_size = 0;
-        if (mtrr_tom2) {
-                extra_remove_base = 1 << (32 - PAGE_SHIFT);
+        extra_remove_base = 1 << (32 - PAGE_SHIFT);
+        if (mtrr_tom2)
                 extra_remove_size =
                         (mtrr_tom2 >> PAGE_SHIFT) - extra_remove_base;
-        }
         nr_range = x86_get_mtrr_mem_range(range, 0, extra_remove_base,
                                           extra_remove_size);
+        /*
+         * [0, 1M) should always be coverred by var mtrr with WB
+         * and fixed mtrrs should take effective before var mtrr for it
+         */
+        nr_range = add_range_with_merge(range, nr_range, 0,
+                                        (1ULL<<(20 - PAGE_SHIFT)) - 1);
+        /* sort the ranges */
+        sort(range, nr_range, sizeof(struct res_range), cmp_range, NULL);
+
         range_sums = sum_ranges(range, nr_range);
         printk(KERN_INFO "total RAM coverred: %ldM\n",
                range_sums >> (20 - PAGE_SHIFT));
 
         if (mtrr_chunk_size && mtrr_gran_size) {
                 int num_reg;
+                char gran_factor, chunk_factor, lose_factor;
+                unsigned long gran_base, chunk_base, lose_base;
 
-                debug_print = 1;
+                debug_print++;
                 /* convert ranges to var ranges state */
                 num_reg = x86_setup_var_mtrrs(range, nr_range, mtrr_chunk_size,
                                               mtrr_gran_size);
@@ -1256,34 +1337,48 @@ static int __init mtrr_cleanup(unsigned address_bits)
                         result[i].lose_cover_sizek =
                                 (range_sums - range_sums_new) << PSHIFT;
 
-                printk(KERN_INFO "%sgran_size: %ldM \tchunk_size: %ldM \t",
-                         result[i].bad?"*BAD*":" ", result[i].gran_sizek >> 10,
-                         result[i].chunk_sizek >> 10);
-                printk(KERN_CONT "num_reg: %d  \tlose cover RAM: %s%ldM \n",
+                gran_base = to_size_factor(result[i].gran_sizek, &gran_factor),
+                chunk_base = to_size_factor(result[i].chunk_sizek, &chunk_factor),
+                lose_base = to_size_factor(result[i].lose_cover_sizek, &lose_factor),
+                printk(KERN_INFO "%sgran_size: %ld%c \tchunk_size: %ld%c \t",
+                         result[i].bad?"*BAD*":" ",
+                         gran_base, gran_factor, chunk_base, chunk_factor);
+                printk(KERN_CONT "num_reg: %d  \tlose cover RAM: %s%ld%c\n",
                          result[i].num_reg, result[i].bad?"-":"",
-                         result[i].lose_cover_sizek >> 10);
+                         lose_base, lose_factor);
                 if (!result[i].bad) {
                         set_var_mtrr_all(address_bits);
                         return 1;
                 }
                 printk(KERN_INFO "invalid mtrr_gran_size or mtrr_chunk_size, "
                        "will find optimal one\n");
-                debug_print = 0;
+                debug_print--;
                 memset(result, 0, sizeof(result[0]));
         }
 
         i = 0;
         memset(min_loss_pfn, 0xff, sizeof(min_loss_pfn));
         memset(result, 0, sizeof(result));
-        for (gran_size = (1ULL<<20); gran_size < (1ULL<<32); gran_size <<= 1) {
-                for (chunk_size = gran_size; chunk_size < (1ULL<<33);
+        for (gran_size = (1ULL<<16); gran_size < (1ULL<<32); gran_size <<= 1) {
+                char gran_factor;
+                unsigned long gran_base;
+
+                if (debug_print)
+                        gran_base = to_size_factor(gran_size >> 10, &gran_factor);
+
+                for (chunk_size = gran_size; chunk_size < (1ULL<<32);
                      chunk_size <<= 1) {
                         int num_reg;
 
-                        if (debug_print)
-                                printk(KERN_INFO
-                               "\ngran_size: %lldM   chunk_size_size: %lldM\n",
-                                       gran_size >> 20, chunk_size >> 20);
+                        if (debug_print) {
+                                char chunk_factor;
+                                unsigned long chunk_base;
+
+                                chunk_base = to_size_factor(chunk_size>>10, &chunk_factor),
+                                printk(KERN_INFO "\n");
+                                printk(KERN_INFO "gran_size: %ld%c   chunk_size: %ld%c \n",
+                                       gran_base, gran_factor, chunk_base, chunk_factor);
+                        }
                         if (i >= NUM_RESULT)
                                 continue;
 
@@ -1326,12 +1421,18 @@ static int __init mtrr_cleanup(unsigned address_bits)
 
         /* print out all */
         for (i = 0; i < NUM_RESULT; i++) {
-                printk(KERN_INFO "%sgran_size: %ldM \tchunk_size: %ldM \t",
-                       result[i].bad?"*BAD* ":" ", result[i].gran_sizek >> 10,
-                       result[i].chunk_sizek >> 10);
-                printk(KERN_CONT "num_reg: %d \tlose RAM: %s%ldM\n",
-                       result[i].num_reg, result[i].bad?"-":"",
-                       result[i].lose_cover_sizek >> 10);
+                char gran_factor, chunk_factor, lose_factor;
+                unsigned long gran_base, chunk_base, lose_base;
+
+                gran_base = to_size_factor(result[i].gran_sizek, &gran_factor),
+                chunk_base = to_size_factor(result[i].chunk_sizek, &chunk_factor),
+                lose_base = to_size_factor(result[i].lose_cover_sizek, &lose_factor),
+                printk(KERN_INFO "%sgran_size: %ld%c \tchunk_size: %ld%c \t",
+                         result[i].bad?"*BAD*":" ",
+                         gran_base, gran_factor, chunk_base, chunk_factor);
+                printk(KERN_CONT "num_reg: %d  \tlose cover RAM: %s%ld%c\n",
+                         result[i].num_reg, result[i].bad?"-":"",
+                         lose_base, lose_factor);
         }
 
         /* try to find the optimal index */
@@ -1339,10 +1440,8 @@ static int __init mtrr_cleanup(unsigned address_bits)
                 nr_mtrr_spare_reg = num_var_ranges - 1;
         num_reg_good = -1;
         for (i = num_var_ranges - nr_mtrr_spare_reg; i > 0; i--) {
-                if (!min_loss_pfn[i]) {
+                if (!min_loss_pfn[i])
                         num_reg_good = i;
-                        break;
-                }
         }
 
         index_good = -1;
@@ -1358,21 +1457,26 @@ static int __init mtrr_cleanup(unsigned address_bits)
         }
 
         if (index_good != -1) {
+                char gran_factor, chunk_factor, lose_factor;
+                unsigned long gran_base, chunk_base, lose_base;
+
                 printk(KERN_INFO "Found optimal setting for mtrr clean up\n");
                 i = index_good;
-                printk(KERN_INFO "gran_size: %ldM \tchunk_size: %ldM \t",
-                                result[i].gran_sizek >> 10,
-                                result[i].chunk_sizek >> 10);
-                printk(KERN_CONT "num_reg: %d \tlose RAM: %ldM\n",
-                                result[i].num_reg,
-                                result[i].lose_cover_sizek >> 10);
+                gran_base = to_size_factor(result[i].gran_sizek, &gran_factor),
+                chunk_base = to_size_factor(result[i].chunk_sizek, &chunk_factor),
+                lose_base = to_size_factor(result[i].lose_cover_sizek, &lose_factor),
+                printk(KERN_INFO "gran_size: %ld%c \tchunk_size: %ld%c \t",
+                         gran_base, gran_factor, chunk_base, chunk_factor);
+                printk(KERN_CONT "num_reg: %d  \tlose RAM: %ld%c\n",
+                         result[i].num_reg, lose_base, lose_factor);
                 /* convert ranges to var ranges state */
                 chunk_size = result[i].chunk_sizek;
                 chunk_size <<= 10;
                 gran_size = result[i].gran_sizek;
                 gran_size <<= 10;
-                debug_print = 1;
+                debug_print++;
                 x86_setup_var_mtrrs(range, nr_range, chunk_size, gran_size);
+                debug_print--;
                 set_var_mtrr_all(address_bits);
                 return 1;
         }
diff --git a/arch/x86/kernel/cpu/perfctr-watchdog.c b/arch/x86/kernel/cpu/perfctr-watchdog.c
index 05cc22dbd4ff..6bff382094f5 100644
--- a/arch/x86/kernel/cpu/perfctr-watchdog.c
+++ b/arch/x86/kernel/cpu/perfctr-watchdog.c
@@ -295,13 +295,19 @@ static int setup_k7_watchdog(unsigned nmi_hz)
         /* setup the timer */
         wrmsr(evntsel_msr, evntsel, 0);
         write_watchdog_counter(perfctr_msr, "K7_PERFCTR0",nmi_hz);
-        apic_write(APIC_LVTPC, APIC_DM_NMI);
-        evntsel |= K7_EVNTSEL_ENABLE;
-        wrmsr(evntsel_msr, evntsel, 0);
 
+        /* initialize the wd struct before enabling */
         wd->perfctr_msr = perfctr_msr;
         wd->evntsel_msr = evntsel_msr;
         wd->cccr_msr = 0;  /* unused */
+
+        /* ok, everything is initialized, announce that we're set */
+        cpu_nmi_set_wd_enabled();
+
+        apic_write(APIC_LVTPC, APIC_DM_NMI);
+        evntsel |= K7_EVNTSEL_ENABLE;
+        wrmsr(evntsel_msr, evntsel, 0);
+
         return 1;
 }
 
@@ -379,13 +385,19 @@ static int setup_p6_watchdog(unsigned nmi_hz)
         wrmsr(evntsel_msr, evntsel, 0);
         nmi_hz = adjust_for_32bit_ctr(nmi_hz);
         write_watchdog_counter32(perfctr_msr, "P6_PERFCTR0",nmi_hz);
-        apic_write(APIC_LVTPC, APIC_DM_NMI);
-        evntsel |= P6_EVNTSEL0_ENABLE;
-        wrmsr(evntsel_msr, evntsel, 0);
 
+        /* initialize the wd struct before enabling */
         wd->perfctr_msr = perfctr_msr;
         wd->evntsel_msr = evntsel_msr;
         wd->cccr_msr = 0;  /* unused */
+
+        /* ok, everything is initialized, announce that we're set */
+        cpu_nmi_set_wd_enabled();
+
+        apic_write(APIC_LVTPC, APIC_DM_NMI);
+        evntsel |= P6_EVNTSEL0_ENABLE;
+        wrmsr(evntsel_msr, evntsel, 0);
+
         return 1;
 }
 
@@ -432,6 +444,27 @@ static const struct wd_ops p6_wd_ops = {
 #define P4_CCCR_ENABLE          (1 << 12)
 #define P4_CCCR_OVF             (1 << 31)
 
+#define P4_CONTROLS 18
+static unsigned int p4_controls[18] = {
+        MSR_P4_BPU_CCCR0,
+        MSR_P4_BPU_CCCR1,
+        MSR_P4_BPU_CCCR2,
+        MSR_P4_BPU_CCCR3,
+        MSR_P4_MS_CCCR0,
+        MSR_P4_MS_CCCR1,
+        MSR_P4_MS_CCCR2,
+        MSR_P4_MS_CCCR3,
+        MSR_P4_FLAME_CCCR0,
+        MSR_P4_FLAME_CCCR1,
+        MSR_P4_FLAME_CCCR2,
+        MSR_P4_FLAME_CCCR3,
+        MSR_P4_IQ_CCCR0,
+        MSR_P4_IQ_CCCR1,
+        MSR_P4_IQ_CCCR2,
+        MSR_P4_IQ_CCCR3,
+        MSR_P4_IQ_CCCR4,
+        MSR_P4_IQ_CCCR5,
+};
 /*
  * Set up IQ_COUNTER0 to behave like a clock, by having IQ_CCCR0 filter
  * CRU_ESCR0 (with any non-null event selector) through a complemented
@@ -473,6 +506,26 @@ static int setup_p4_watchdog(unsigned nmi_hz)
                 evntsel_msr = MSR_P4_CRU_ESCR0;
                 cccr_msr = MSR_P4_IQ_CCCR0;
                 cccr_val = P4_CCCR_OVF_PMI0 | P4_CCCR_ESCR_SELECT(4);
+
+                /*
+                 * If we're on the kdump kernel or other situation, we may
+                 * still have other performance counter registers set to
+                 * interrupt and they'll keep interrupting forever because
+                 * of the P4_CCCR_OVF quirk. So we need to ACK all the
+                 * pending interrupts and disable all the registers here,
+                 * before reenabling the NMI delivery. Refer to p4_rearm()
+                 * about the P4_CCCR_OVF quirk.
+                 */
+                if (reset_devices) {
+                        unsigned int low, high;
+                        int i;
+
+                        for (i = 0; i < P4_CONTROLS; i++) {
+                                rdmsr(p4_controls[i], low, high);
+                                low &= ~(P4_CCCR_ENABLE | P4_CCCR_OVF);
+                                wrmsr(p4_controls[i], low, high);
+                        }
+                }
         } else {
                 /* logical cpu 1 */
                 perfctr_msr = MSR_P4_IQ_PERFCTR1;
@@ -499,12 +552,17 @@ static int setup_p4_watchdog(unsigned nmi_hz)
         wrmsr(evntsel_msr, evntsel, 0);
         wrmsr(cccr_msr, cccr_val, 0);
         write_watchdog_counter(perfctr_msr, "P4_IQ_COUNTER0", nmi_hz);
-        apic_write(APIC_LVTPC, APIC_DM_NMI);
-        cccr_val |= P4_CCCR_ENABLE;
-        wrmsr(cccr_msr, cccr_val, 0);
+
         wd->perfctr_msr = perfctr_msr;
         wd->evntsel_msr = evntsel_msr;
         wd->cccr_msr = cccr_msr;
+
+        /* ok, everything is initialized, announce that we're set */
+        cpu_nmi_set_wd_enabled();
+
+        apic_write(APIC_LVTPC, APIC_DM_NMI);
+        cccr_val |= P4_CCCR_ENABLE;
+        wrmsr(cccr_msr, cccr_val, 0);
         return 1;
 }
 
@@ -620,13 +678,17 @@ static int setup_intel_arch_watchdog(unsigned nmi_hz)
         wrmsr(evntsel_msr, evntsel, 0);
         nmi_hz = adjust_for_32bit_ctr(nmi_hz);
         write_watchdog_counter32(perfctr_msr, "INTEL_ARCH_PERFCTR0", nmi_hz);
-        apic_write(APIC_LVTPC, APIC_DM_NMI);
-        evntsel |= ARCH_PERFMON_EVENTSEL0_ENABLE;
-        wrmsr(evntsel_msr, evntsel, 0);
 
         wd->perfctr_msr = perfctr_msr;
         wd->evntsel_msr = evntsel_msr;
         wd->cccr_msr = 0;  /* unused */
+
+        /* ok, everything is initialized, announce that we're set */
+        cpu_nmi_set_wd_enabled();
+
+        apic_write(APIC_LVTPC, APIC_DM_NMI);
+        evntsel |= ARCH_PERFMON_EVENTSEL0_ENABLE;
+        wrmsr(evntsel_msr, evntsel, 0);
         intel_arch_wd_ops.checkbit = 1ULL << (eax.split.bit_width - 1);
         return 1;
 }
diff --git a/arch/x86/kernel/cpuid.c b/arch/x86/kernel/cpuid.c
index 8e9cd6a8ec12..6a44d6465991 100644
--- a/arch/x86/kernel/cpuid.c
+++ b/arch/x86/kernel/cpuid.c
@@ -36,7 +36,6 @@
 #include <linux/smp_lock.h>
 #include <linux/major.h>
 #include <linux/fs.h>
-#include <linux/smp_lock.h>
 #include <linux/device.h>
 #include <linux/cpu.h>
 #include <linux/notifier.h>
diff --git a/arch/x86/kernel/crash_dump_64.c b/arch/x86/kernel/crash_dump_64.c
index 15e6c6bc4a46..e90a60ef10c2 100644
--- a/arch/x86/kernel/crash_dump_64.c
+++ b/arch/x86/kernel/crash_dump_64.c
@@ -7,9 +7,8 @@
 
 #include <linux/errno.h>
 #include <linux/crash_dump.h>
-
-#include <asm/uaccess.h>
-#include <asm/io.h>
+#include <linux/uaccess.h>
+#include <linux/io.h>
 
 /**
  * copy_oldmem_page - copy one page from "oldmem"
@@ -25,7 +24,7 @@
  * in the current kernel. We stitch up a pte, similar to kmap_atomic.
  */
 ssize_t copy_oldmem_page(unsigned long pfn, char *buf,
-                               size_t csize, unsigned long offset, int userbuf)
+                size_t csize, unsigned long offset, int userbuf)
 {
         void  *vaddr;
 
@@ -33,14 +32,16 @@ ssize_t copy_oldmem_page(unsigned long pfn, char *buf,
                 return 0;
 
         vaddr = ioremap(pfn << PAGE_SHIFT, PAGE_SIZE);
+        if (!vaddr)
+                return -ENOMEM;
 
         if (userbuf) {
-                if (copy_to_user(buf, (vaddr + offset), csize)) {
+                if (copy_to_user(buf, vaddr + offset, csize)) {
                         iounmap(vaddr);
                         return -EFAULT;
                 }
         } else
-        memcpy(buf, (vaddr + offset), csize);
+                memcpy(buf, vaddr + offset, csize);
 
         iounmap(vaddr);
         return csize;
diff --git a/arch/x86/kernel/ds.c b/arch/x86/kernel/ds.c
index 11c11b8ec48d..2b69994fd3a8 100644
--- a/arch/x86/kernel/ds.c
+++ b/arch/x86/kernel/ds.c
@@ -2,26 +2,49 @@
  * Debug Store support
  *
  * This provides a low-level interface to the hardware's Debug Store
- * feature that is used for last branch recording (LBR) and
+ * feature that is used for branch trace store (BTS) and
  * precise-event based sampling (PEBS).
  *
- * Different architectures use a different DS layout/pointer size.
- * The below functions therefore work on a void*.
+ * It manages:
+ * - per-thread and per-cpu allocation of BTS and PEBS
+ * - buffer memory allocation (optional)
+ * - buffer overflow handling
+ * - buffer access
  *
+ * It assumes:
+ * - get_task_struct on all parameter tasks
+ * - current is allowed to trace parameter tasks
  *
- * Since there is no user for PEBS, yet, only LBR (or branch
- * trace store, BTS) is supported.
  *
- *
- * Copyright (C) 2007 Intel Corporation.
- * Markus Metzger <markus.t.metzger@intel.com>, Dec 2007
+ * Copyright (C) 2007-2008 Intel Corporation.
+ * Markus Metzger <markus.t.metzger@intel.com>, 2007-2008
  */
 
+
+#ifdef CONFIG_X86_DS
+
 #include <asm/ds.h>
 
 #include <linux/errno.h>
 #include <linux/string.h>
 #include <linux/slab.h>
+#include <linux/sched.h>
+#include <linux/mm.h>
+
+
+/*
+ * The configuration for a particular DS hardware implementation.
+ */
+struct ds_configuration {
+        /* the size of the DS structure in bytes */
+        unsigned char  sizeof_ds;
+        /* the size of one pointer-typed field in the DS structure in bytes;
+           this covers the first 8 fields related to buffer management. */
+        unsigned char  sizeof_field;
+        /* the size of a BTS/PEBS record in bytes */
+        unsigned char  sizeof_rec[2];
+};
+static struct ds_configuration ds_cfg;
 
 
 /*
@@ -44,378 +67,747 @@
  *   (interrupt occurs when write pointer passes interrupt pointer)
  * - value to which counter is reset following counter overflow
  *
- * On later architectures, the last branch recording hardware uses
- * 64bit pointers even in 32bit mode.
- *
- *
- * Branch Trace Store (BTS) records store information about control
- * flow changes. They at least provide the following information:
- * - source linear address
- * - destination linear address
+ * Later architectures use 64bit pointers throughout, whereas earlier
+ * architectures use 32bit pointers in 32bit mode.
  *
- * Netburst supported a predicated bit that had been dropped in later
- * architectures. We do not suppor it.
  *
+ * We compute the base address for the first 8 fields based on:
+ * - the field size stored in the DS configuration
+ * - the relative field position
+ * - an offset giving the start of the respective region
  *
- * In order to abstract from the actual DS and BTS layout, we describe
- * the access to the relevant fields.
- * Thanks to Andi Kleen for proposing this design.
+ * This offset is further used to index various arrays holding
+ * information for BTS and PEBS at the respective index.
  *
- * The implementation, however, is not as general as it might seem. In
- * order to stay somewhat simple and efficient, we assume an
- * underlying unsigned type (mostly a pointer type) and we expect the
- * field to be at least as big as that type.
+ * On later 32bit processors, we only access the lower 32bit of the
+ * 64bit pointer fields. The upper halves will be zeroed out.
  */
 
-/*
- * A special from_ip address to indicate that the BTS record is an
- * info record that needs to be interpreted or skipped.
- */
-#define BTS_ESCAPE_ADDRESS (-1)
+enum ds_field {
+        ds_buffer_base = 0,
+        ds_index,
+        ds_absolute_maximum,
+        ds_interrupt_threshold,
+};
 
-/*
- * A field access descriptor
- */
-struct access_desc {
-        unsigned char offset;
-        unsigned char size;
+enum ds_qualifier {
+        ds_bts  = 0,
+        ds_pebs
 };
 
+static inline unsigned long ds_get(const unsigned char *base,
+                                   enum ds_qualifier qual, enum ds_field field)
+{
+        base += (ds_cfg.sizeof_field * (field + (4 * qual)));
+        return *(unsigned long *)base;
+}
+
+static inline void ds_set(unsigned char *base, enum ds_qualifier qual,
+                          enum ds_field field, unsigned long value)
+{
+        base += (ds_cfg.sizeof_field * (field + (4 * qual)));
+        (*(unsigned long *)base) = value;
+}
+
+
 /*
- * The configuration for a particular DS/BTS hardware implementation.
+ * Locking is done only for allocating BTS or PEBS resources and for
+ * guarding context and buffer memory allocation.
+ *
+ * Most functions require the current task to own the ds context part
+ * they are going to access. All the locking is done when validating
+ * access to the context.
  */
-struct ds_configuration {
-        /* the DS configuration */
-        unsigned char  sizeof_ds;
-        struct access_desc bts_buffer_base;
-        struct access_desc bts_index;
-        struct access_desc bts_absolute_maximum;
-        struct access_desc bts_interrupt_threshold;
-        /* the BTS configuration */
-        unsigned char  sizeof_bts;
-        struct access_desc from_ip;
-        struct access_desc to_ip;
-        /* BTS variants used to store additional information like
-           timestamps */
-        struct access_desc info_type;
-        struct access_desc info_data;
-        unsigned long debugctl_mask;
-};
+static spinlock_t ds_lock = __SPIN_LOCK_UNLOCKED(ds_lock);
 
 /*
- * The global configuration used by the below accessor functions
+ * Validate that the current task is allowed to access the BTS/PEBS
+ * buffer of the parameter task.
+ *
+ * Returns 0, if access is granted; -Eerrno, otherwise.
  */
-static struct ds_configuration ds_cfg;
+static inline int ds_validate_access(struct ds_context *context,
+                                     enum ds_qualifier qual)
+{
+        if (!context)
+                return -EPERM;
+
+        if (context->owner[qual] == current)
+                return 0;
+
+        return -EPERM;
+}
+
 
 /*
- * Accessor functions for some DS and BTS fields using the above
- * global ptrace_bts_cfg.
+ * We either support (system-wide) per-cpu or per-thread allocation.
+ * We distinguish the two based on the task_struct pointer, where a
+ * NULL pointer indicates per-cpu allocation for the current cpu.
+ *
+ * Allocations are use-counted. As soon as resources are allocated,
+ * further allocations must be of the same type (per-cpu or
+ * per-thread). We model this by counting allocations (i.e. the number
+ * of tracers of a certain type) for one type negatively:
+ *   =0  no tracers
+ *   >0  number of per-thread tracers
+ *   <0  number of per-cpu tracers
+ *
+ * The below functions to get and put tracers and to check the
+ * allocation type require the ds_lock to be held by the caller.
+ *
+ * Tracers essentially gives the number of ds contexts for a certain
+ * type of allocation.
  */
-static inline unsigned long get_bts_buffer_base(char *base)
+static long tracers;
+
+static inline void get_tracer(struct task_struct *task)
 {
-        return *(unsigned long *)(base + ds_cfg.bts_buffer_base.offset);
+        tracers += (task ? 1 : -1);
 }
-static inline void set_bts_buffer_base(char *base, unsigned long value)
+
+static inline void put_tracer(struct task_struct *task)
 {
-        (*(unsigned long *)(base + ds_cfg.bts_buffer_base.offset)) = value;
+        tracers -= (task ? 1 : -1);
 }
-static inline unsigned long get_bts_index(char *base)
+
+static inline int check_tracer(struct task_struct *task)
 {
-        return *(unsigned long *)(base + ds_cfg.bts_index.offset);
+        return (task ? (tracers >= 0) : (tracers <= 0));
 }
-static inline void set_bts_index(char *base, unsigned long value)
+
+
+/*
+ * The DS context is either attached to a thread or to a cpu:
+ * - in the former case, the thread_struct contains a pointer to the
+ *   attached context.
+ * - in the latter case, we use a static array of per-cpu context
+ *   pointers.
+ *
+ * Contexts are use-counted. They are allocated on first access and
+ * deallocated when the last user puts the context.
+ *
+ * We distinguish between an allocating and a non-allocating get of a
+ * context:
+ * - the allocating get is used for requesting BTS/PEBS resources. It
+ *   requires the caller to hold the global ds_lock.
+ * - the non-allocating get is used for all other cases. A
+ *   non-existing context indicates an error. It acquires and releases
+ *   the ds_lock itself for obtaining the context.
+ *
+ * A context and its DS configuration are allocated and deallocated
+ * together. A context always has a DS configuration of the
+ * appropriate size.
+ */
+static DEFINE_PER_CPU(struct ds_context *, system_context);
+
+#define this_system_context per_cpu(system_context, smp_processor_id())
+
+/*
+ * Returns the pointer to the parameter task's context or to the
+ * system-wide context, if task is NULL.
+ *
+ * Increases the use count of the returned context, if not NULL.
+ */
+static inline struct ds_context *ds_get_context(struct task_struct *task)
 {
-        (*(unsigned long *)(base + ds_cfg.bts_index.offset)) = value;
+        struct ds_context *context;
+
+        spin_lock(&ds_lock);
+
+        context = (task ? task->thread.ds_ctx : this_system_context);
+        if (context)
+                context->count++;
+
+        spin_unlock(&ds_lock);
+
+        return context;
 }
-static inline unsigned long get_bts_absolute_maximum(char *base)
+
+/*
+ * Same as ds_get_context, but allocates the context and it's DS
+ * structure, if necessary; returns NULL; if out of memory.
+ *
+ * pre: requires ds_lock to be held
+ */
+static inline struct ds_context *ds_alloc_context(struct task_struct *task)
 {
-        return *(unsigned long *)(base + ds_cfg.bts_absolute_maximum.offset);
+        struct ds_context **p_context =
+                (task ? &task->thread.ds_ctx : &this_system_context);
+        struct ds_context *context = *p_context;
+
+        if (!context) {
+                context = kzalloc(sizeof(*context), GFP_KERNEL);
+
+                if (!context)
+                        return NULL;
+
+                context->ds = kzalloc(ds_cfg.sizeof_ds, GFP_KERNEL);
+                if (!context->ds) {
+                        kfree(context);
+                        return NULL;
+                }
+
+                *p_context = context;
+
+                context->this = p_context;
+                context->task = task;
+
+                if (task)
+                        set_tsk_thread_flag(task, TIF_DS_AREA_MSR);
+
+                if (!task || (task == current))
+                        wrmsr(MSR_IA32_DS_AREA, (unsigned long)context->ds, 0);
+
+                get_tracer(task);
+        }
+
+        context->count++;
+
+        return context;
 }
-static inline void set_bts_absolute_maximum(char *base, unsigned long value)
+
+/*
+ * Decreases the use count of the parameter context, if not NULL.
+ * Deallocates the context, if the use count reaches zero.
+ */
+static inline void ds_put_context(struct ds_context *context)
 {
-        (*(unsigned long *)(base + ds_cfg.bts_absolute_maximum.offset)) = value;
+        if (!context)
+                return;
+
+        spin_lock(&ds_lock);
+
+        if (--context->count)
+                goto out;
+
+        *(context->this) = NULL;
+
+        if (context->task)
+                clear_tsk_thread_flag(context->task, TIF_DS_AREA_MSR);
+
+        if (!context->task || (context->task == current))
+                wrmsrl(MSR_IA32_DS_AREA, 0);
+
+        put_tracer(context->task);
+
+        /* free any leftover buffers from tracers that did not
+         * deallocate them properly. */
+        kfree(context->buffer[ds_bts]);
+        kfree(context->buffer[ds_pebs]);
+        kfree(context->ds);
+        kfree(context);
+ out:
+        spin_unlock(&ds_lock);
 }
-static inline unsigned long get_bts_interrupt_threshold(char *base)
+
+
+/*
+ * Handle a buffer overflow
+ *
+ * task: the task whose buffers are overflowing;
+ *       NULL for a buffer overflow on the current cpu
+ * context: the ds context
+ * qual: the buffer type
+ */
+static void ds_overflow(struct task_struct *task, struct ds_context *context,
+                        enum ds_qualifier qual)
 {
-        return *(unsigned long *)(base + ds_cfg.bts_interrupt_threshold.offset);
+        if (!context)
+                return;
+
+        if (context->callback[qual])
+                (*context->callback[qual])(task);
+
+        /* todo: do some more overflow handling */
 }
-static inline void set_bts_interrupt_threshold(char *base, unsigned long value)
+
+
+/*
+ * Allocate a non-pageable buffer of the parameter size.
+ * Checks the memory and the locked memory rlimit.
+ *
+ * Returns the buffer, if successful;
+ *         NULL, if out of memory or rlimit exceeded.
+ *
+ * size: the requested buffer size in bytes
+ * pages (out): if not NULL, contains the number of pages reserved
+ */
+static inline void *ds_allocate_buffer(size_t size, unsigned int *pages)
 {
-        (*(unsigned long *)(base + ds_cfg.bts_interrupt_threshold.offset)) = value;
+        unsigned long rlim, vm, pgsz;
+        void *buffer;
+
+        pgsz = PAGE_ALIGN(size) >> PAGE_SHIFT;
+
+        rlim = current->signal->rlim[RLIMIT_AS].rlim_cur >> PAGE_SHIFT;
+        vm   = current->mm->total_vm  + pgsz;
+        if (rlim < vm)
+                return NULL;
+
+        rlim = current->signal->rlim[RLIMIT_MEMLOCK].rlim_cur >> PAGE_SHIFT;
+        vm   = current->mm->locked_vm  + pgsz;
+        if (rlim < vm)
+                return NULL;
+
+        buffer = kzalloc(size, GFP_KERNEL);
+        if (!buffer)
+                return NULL;
+
+        current->mm->total_vm  += pgsz;
+        current->mm->locked_vm += pgsz;
+
+        if (pages)
+                *pages = pgsz;
+
+        return buffer;
 }
-static inline unsigned long get_from_ip(char *base)
+
+static int ds_request(struct task_struct *task, void *base, size_t size,
+                      ds_ovfl_callback_t ovfl, enum ds_qualifier qual)
 {
-        return *(unsigned long *)(base + ds_cfg.from_ip.offset);
+        struct ds_context *context;
+        unsigned long buffer, adj;
+        const unsigned long alignment = (1 << 3);
+        int error = 0;
+
+        if (!ds_cfg.sizeof_ds)
+                return -EOPNOTSUPP;
+
+        /* we require some space to do alignment adjustments below */
+        if (size < (alignment + ds_cfg.sizeof_rec[qual]))
+                return -EINVAL;
+
+        /* buffer overflow notification is not yet implemented */
+        if (ovfl)
+                return -EOPNOTSUPP;
+
+
+        spin_lock(&ds_lock);
+
+        if (!check_tracer(task))
+                return -EPERM;
+
+        error = -ENOMEM;
+        context = ds_alloc_context(task);
+        if (!context)
+                goto out_unlock;
+
+        error = -EALREADY;
+        if (context->owner[qual] == current)
+                goto out_unlock;
+        error = -EPERM;
+        if (context->owner[qual] != NULL)
+                goto out_unlock;
+        context->owner[qual] = current;
+
+        spin_unlock(&ds_lock);
+
+
+        error = -ENOMEM;
+        if (!base) {
+                base = ds_allocate_buffer(size, &context->pages[qual]);
+                if (!base)
+                        goto out_release;
+
+                context->buffer[qual]   = base;
+        }
+        error = 0;
+
+        context->callback[qual] = ovfl;
+
+        /* adjust the buffer address and size to meet alignment
+         * constraints:
+         * - buffer is double-word aligned
+         * - size is multiple of record size
+         *
+         * We checked the size at the very beginning; we have enough
+         * space to do the adjustment.
+         */
+        buffer = (unsigned long)base;
+
+        adj = ALIGN(buffer, alignment) - buffer;
+        buffer += adj;
+        size   -= adj;
+
+        size /= ds_cfg.sizeof_rec[qual];
+        size *= ds_cfg.sizeof_rec[qual];
+
+        ds_set(context->ds, qual, ds_buffer_base, buffer);
+        ds_set(context->ds, qual, ds_index, buffer);
+        ds_set(context->ds, qual, ds_absolute_maximum, buffer + size);
+
+        if (ovfl) {
+                /* todo: select a suitable interrupt threshold */
+        } else
+                ds_set(context->ds, qual,
+                       ds_interrupt_threshold, buffer + size + 1);
+
+        /* we keep the context until ds_release */
+        return error;
+
+ out_release:
+        context->owner[qual] = NULL;
+        ds_put_context(context);
+        return error;
+
+ out_unlock:
+        spin_unlock(&ds_lock);
+        ds_put_context(context);
+        return error;
 }
-static inline void set_from_ip(char *base, unsigned long value)
+
+int ds_request_bts(struct task_struct *task, void *base, size_t size,
+                   ds_ovfl_callback_t ovfl)
 {
-        (*(unsigned long *)(base + ds_cfg.from_ip.offset)) = value;
+        return ds_request(task, base, size, ovfl, ds_bts);
 }
-static inline unsigned long get_to_ip(char *base)
+
+int ds_request_pebs(struct task_struct *task, void *base, size_t size,
+                    ds_ovfl_callback_t ovfl)
 {
-        return *(unsigned long *)(base + ds_cfg.to_ip.offset);
+        return ds_request(task, base, size, ovfl, ds_pebs);
 }
-static inline void set_to_ip(char *base, unsigned long value)
+
+static int ds_release(struct task_struct *task, enum ds_qualifier qual)
 {
-        (*(unsigned long *)(base + ds_cfg.to_ip.offset)) = value;
+        struct ds_context *context;
+        int error;
+
+        context = ds_get_context(task);
+        error = ds_validate_access(context, qual);
+        if (error < 0)
+                goto out;
+
+        kfree(context->buffer[qual]);
+        context->buffer[qual] = NULL;
+
+        current->mm->total_vm  -= context->pages[qual];
+        current->mm->locked_vm -= context->pages[qual];
+        context->pages[qual] = 0;
+        context->owner[qual] = NULL;
+
+        /*
+         * we put the context twice:
+         *   once for the ds_get_context
+         *   once for the corresponding ds_request
+         */
+        ds_put_context(context);
+ out:
+        ds_put_context(context);
+        return error;
 }
-static inline unsigned char get_info_type(char *base)
+
+int ds_release_bts(struct task_struct *task)
 {
-        return *(unsigned char *)(base + ds_cfg.info_type.offset);
+        return ds_release(task, ds_bts);
 }
-static inline void set_info_type(char *base, unsigned char value)
+
+int ds_release_pebs(struct task_struct *task)
 {
-        (*(unsigned char *)(base + ds_cfg.info_type.offset)) = value;
+        return ds_release(task, ds_pebs);
 }
-static inline unsigned long get_info_data(char *base)
+
+static int ds_get_index(struct task_struct *task, size_t *pos,
+                        enum ds_qualifier qual)
 {
-        return *(unsigned long *)(base + ds_cfg.info_data.offset);
+        struct ds_context *context;
+        unsigned long base, index;
+        int error;
+
+        context = ds_get_context(task);
+        error = ds_validate_access(context, qual);
+        if (error < 0)
+                goto out;
+
+        base  = ds_get(context->ds, qual, ds_buffer_base);
+        index = ds_get(context->ds, qual, ds_index);
+
+        error = ((index - base) / ds_cfg.sizeof_rec[qual]);
+        if (pos)
+                *pos = error;
+ out:
+        ds_put_context(context);
+        return error;
 }
-static inline void set_info_data(char *base, unsigned long value)
+
+int ds_get_bts_index(struct task_struct *task, size_t *pos)
 {
-        (*(unsigned long *)(base + ds_cfg.info_data.offset)) = value;
+        return ds_get_index(task, pos, ds_bts);
 }
 
+int ds_get_pebs_index(struct task_struct *task, size_t *pos)
+{
+        return ds_get_index(task, pos, ds_pebs);
+}
 
-int ds_allocate(void **dsp, size_t bts_size_in_bytes)
+static int ds_get_end(struct task_struct *task, size_t *pos,
+                      enum ds_qualifier qual)
 {
-        size_t bts_size_in_records;
-        unsigned long bts;
-        void *ds;
+        struct ds_context *context;
+        unsigned long base, end;
+        int error;
+
+        context = ds_get_context(task);
+        error = ds_validate_access(context, qual);
+        if (error < 0)
+                goto out;
+
+        base = ds_get(context->ds, qual, ds_buffer_base);
+        end  = ds_get(context->ds, qual, ds_absolute_maximum);
+
+        error = ((end - base) / ds_cfg.sizeof_rec[qual]);
+        if (pos)
+                *pos = error;
+ out:
+        ds_put_context(context);
+        return error;
+}
 
-        if (!ds_cfg.sizeof_ds || !ds_cfg.sizeof_bts)
-                return -EOPNOTSUPP;
+int ds_get_bts_end(struct task_struct *task, size_t *pos)
+{
+        return ds_get_end(task, pos, ds_bts);
+}
 
-        if (bts_size_in_bytes < 0)
-                return -EINVAL;
+int ds_get_pebs_end(struct task_struct *task, size_t *pos)
+{
+        return ds_get_end(task, pos, ds_pebs);
+}
 
-        bts_size_in_records =
-                bts_size_in_bytes / ds_cfg.sizeof_bts;
-        bts_size_in_bytes =
-                bts_size_in_records * ds_cfg.sizeof_bts;
+static int ds_access(struct task_struct *task, size_t index,
+                     const void **record, enum ds_qualifier qual)
+{
+        struct ds_context *context;
+        unsigned long base, idx;
+        int error;
 
-        if (bts_size_in_bytes <= 0)
+        if (!record)
                 return -EINVAL;
 
-        bts = (unsigned long)kzalloc(bts_size_in_bytes, GFP_KERNEL);
-
-        if (!bts)
-                return -ENOMEM;
+        context = ds_get_context(task);
+        error = ds_validate_access(context, qual);
+        if (error < 0)
+                goto out;
 
-        ds = kzalloc(ds_cfg.sizeof_ds, GFP_KERNEL);
+        base = ds_get(context->ds, qual, ds_buffer_base);
+        idx = base + (index * ds_cfg.sizeof_rec[qual]);
 
-        if (!ds) {
-                kfree((void *)bts);
-                return -ENOMEM;
-        }
-
-        set_bts_buffer_base(ds, bts);
-        set_bts_index(ds, bts);
-        set_bts_absolute_maximum(ds, bts + bts_size_in_bytes);
-        set_bts_interrupt_threshold(ds, bts + bts_size_in_bytes + 1);
+        error = -EINVAL;
+        if (idx > ds_get(context->ds, qual, ds_absolute_maximum))
+                goto out;
 
-        *dsp = ds;
-        return 0;
+        *record = (const void *)idx;
+        error = ds_cfg.sizeof_rec[qual];
+ out:
+        ds_put_context(context);
+        return error;
 }
 
-int ds_free(void **dsp)
+int ds_access_bts(struct task_struct *task, size_t index, const void **record)
 {
-        if (*dsp) {
-                kfree((void *)get_bts_buffer_base(*dsp));
-                kfree(*dsp);
-                *dsp = NULL;
-        }
-        return 0;
+        return ds_access(task, index, record, ds_bts);
 }
 
-int ds_get_bts_size(void *ds)
+int ds_access_pebs(struct task_struct *task, size_t index, const void **record)
 {
-        int size_in_bytes;
-
-        if (!ds_cfg.sizeof_ds || !ds_cfg.sizeof_bts)
-                return -EOPNOTSUPP;
-
-        if (!ds)
-                return 0;
-
-        size_in_bytes =
-                get_bts_absolute_maximum(ds) -
-                get_bts_buffer_base(ds);
-        return size_in_bytes;
+        return ds_access(task, index, record, ds_pebs);
 }
 
-int ds_get_bts_end(void *ds)
+static int ds_write(struct task_struct *task, const void *record, size_t size,
+                    enum ds_qualifier qual, int force)
 {
-        int size_in_bytes = ds_get_bts_size(ds);
-
-        if (size_in_bytes <= 0)
-                return size_in_bytes;
+        struct ds_context *context;
+        int error;
 
-        return size_in_bytes / ds_cfg.sizeof_bts;
-}
+        if (!record)
+                return -EINVAL;
 
-int ds_get_bts_index(void *ds)
-{
-        int index_offset_in_bytes;
+        error = -EPERM;
+        context = ds_get_context(task);
+        if (!context)
+                goto out;
 
-        if (!ds_cfg.sizeof_ds || !ds_cfg.sizeof_bts)
-                return -EOPNOTSUPP;
+        if (!force) {
+                error = ds_validate_access(context, qual);
+                if (error < 0)
+                        goto out;
+        }
 
-        index_offset_in_bytes =
-                get_bts_index(ds) -
-                get_bts_buffer_base(ds);
+        error = 0;
+        while (size) {
+                unsigned long base, index, end, write_end, int_th;
+                unsigned long write_size, adj_write_size;
+
+                /*
+                 * write as much as possible without producing an
+                 * overflow interrupt.
+                 *
+                 * interrupt_threshold must either be
+                 * - bigger than absolute_maximum or
+                 * - point to a record between buffer_base and absolute_maximum
+                 *
+                 * index points to a valid record.
+                 */
+                base   = ds_get(context->ds, qual, ds_buffer_base);
+                index  = ds_get(context->ds, qual, ds_index);
+                end    = ds_get(context->ds, qual, ds_absolute_maximum);
+                int_th = ds_get(context->ds, qual, ds_interrupt_threshold);
+
+                write_end = min(end, int_th);
+
+                /* if we are already beyond the interrupt threshold,
+                 * we fill the entire buffer */
+                if (write_end <= index)
+                        write_end = end;
+
+                if (write_end <= index)
+                        goto out;
+
+                write_size = min((unsigned long) size, write_end - index);
+                memcpy((void *)index, record, write_size);
+
+                record = (const char *)record + write_size;
+                size  -= write_size;
+                error += write_size;
+
+                adj_write_size = write_size / ds_cfg.sizeof_rec[qual];
+                adj_write_size *= ds_cfg.sizeof_rec[qual];
+
+                /* zero out trailing bytes */
+                memset((char *)index + write_size, 0,
+                       adj_write_size - write_size);
+                index += adj_write_size;
+
+                if (index >= end)
+                        index = base;
+                ds_set(context->ds, qual, ds_index, index);
+
+                if (index >= int_th)
+                        ds_overflow(task, context, qual);
+        }
 
-        return index_offset_in_bytes / ds_cfg.sizeof_bts;
+ out:
+        ds_put_context(context);
+        return error;
 }
 
-int ds_set_overflow(void *ds, int method)
+int ds_write_bts(struct task_struct *task, const void *record, size_t size)
 {
-        switch (method) {
-        case DS_O_SIGNAL:
-                return -EOPNOTSUPP;
-        case DS_O_WRAP:
-                return 0;
-        default:
-                return -EINVAL;
-        }
+        return ds_write(task, record, size, ds_bts, /* force = */ 0);
 }
 
-int ds_get_overflow(void *ds)
+int ds_write_pebs(struct task_struct *task, const void *record, size_t size)
 {
-        return DS_O_WRAP;
+        return ds_write(task, record, size, ds_pebs, /* force = */ 0);
 }
 
-int ds_clear(void *ds)
+int ds_unchecked_write_bts(struct task_struct *task,
+                           const void *record, size_t size)
 {
-        int bts_size = ds_get_bts_size(ds);
-        unsigned long bts_base;
-
-        if (bts_size <= 0)
-                return bts_size;
-
-        bts_base = get_bts_buffer_base(ds);
-        memset((void *)bts_base, 0, bts_size);
-
-        set_bts_index(ds, bts_base);
-        return 0;
+        return ds_write(task, record, size, ds_bts, /* force = */ 1);
 }
 
-int ds_read_bts(void *ds, int index, struct bts_struct *out)
+int ds_unchecked_write_pebs(struct task_struct *task,
+                            const void *record, size_t size)
 {
-        void *bts;
+        return ds_write(task, record, size, ds_pebs, /* force = */ 1);
+}
 
-        if (!ds_cfg.sizeof_ds || !ds_cfg.sizeof_bts)
-                return -EOPNOTSUPP;
+static int ds_reset_or_clear(struct task_struct *task,
+                             enum ds_qualifier qual, int clear)
+{
+        struct ds_context *context;
+        unsigned long base, end;
+        int error;
 
-        if (index < 0)
-                return -EINVAL;
+        context = ds_get_context(task);
+        error = ds_validate_access(context, qual);
+        if (error < 0)
+                goto out;
 
-        if (index >= ds_get_bts_size(ds))
-                return -EINVAL;
+        base = ds_get(context->ds, qual, ds_buffer_base);
+        end  = ds_get(context->ds, qual, ds_absolute_maximum);
 
-        bts = (void *)(get_bts_buffer_base(ds) + (index * ds_cfg.sizeof_bts));
+        if (clear)
+                memset((void *)base, 0, end - base);
 
-        memset(out, 0, sizeof(*out));
-        if (get_from_ip(bts) == BTS_ESCAPE_ADDRESS) {
-                out->qualifier       = get_info_type(bts);
-                out->variant.jiffies = get_info_data(bts);
-        } else {
-                out->qualifier = BTS_BRANCH;
-                out->variant.lbr.from_ip = get_from_ip(bts);
-                out->variant.lbr.to_ip   = get_to_ip(bts);
-        }
+        ds_set(context->ds, qual, ds_index, base);
 
-        return sizeof(*out);;
+        error = 0;
+ out:
+        ds_put_context(context);
+        return error;
 }
 
-int ds_write_bts(void *ds, const struct bts_struct *in)
+int ds_reset_bts(struct task_struct *task)
 {
-        unsigned long bts;
-
-        if (!ds_cfg.sizeof_ds || !ds_cfg.sizeof_bts)
-                return -EOPNOTSUPP;
-
-        if (ds_get_bts_size(ds) <= 0)
-                return -ENXIO;
+        return ds_reset_or_clear(task, ds_bts, /* clear = */ 0);
+}
 
-        bts = get_bts_index(ds);
+int ds_reset_pebs(struct task_struct *task)
+{
+        return ds_reset_or_clear(task, ds_pebs, /* clear = */ 0);
+}
 
-        memset((void *)bts, 0, ds_cfg.sizeof_bts);
-        switch (in->qualifier) {
-        case BTS_INVALID:
-                break;
+int ds_clear_bts(struct task_struct *task)
+{
+        return ds_reset_or_clear(task, ds_bts, /* clear = */ 1);
+}
 
-        case BTS_BRANCH:
-                set_from_ip((void *)bts, in->variant.lbr.from_ip);
-                set_to_ip((void *)bts, in->variant.lbr.to_ip);
-                break;
+int ds_clear_pebs(struct task_struct *task)
+{
+        return ds_reset_or_clear(task, ds_pebs, /* clear = */ 1);
+}
 
-        case BTS_TASK_ARRIVES:
-        case BTS_TASK_DEPARTS:
-                set_from_ip((void *)bts, BTS_ESCAPE_ADDRESS);
-                set_info_type((void *)bts, in->qualifier);
-                set_info_data((void *)bts, in->variant.jiffies);
-                break;
+int ds_get_pebs_reset(struct task_struct *task, u64 *value)
+{
+        struct ds_context *context;
+        int error;
 
-        default:
+        if (!value)
                 return -EINVAL;
-        }
 
-        bts = bts + ds_cfg.sizeof_bts;
-        if (bts >= get_bts_absolute_maximum(ds))
-                bts = get_bts_buffer_base(ds);
-        set_bts_index(ds, bts);
+        context = ds_get_context(task);
+        error = ds_validate_access(context, ds_pebs);
+        if (error < 0)
+                goto out;
 
-        return ds_cfg.sizeof_bts;
+        *value = *(u64 *)(context->ds + (ds_cfg.sizeof_field * 8));
+
+        error = 0;
+ out:
+        ds_put_context(context);
+        return error;
 }
 
-unsigned long ds_debugctl_mask(void)
+int ds_set_pebs_reset(struct task_struct *task, u64 value)
 {
-        return ds_cfg.debugctl_mask;
-}
+        struct ds_context *context;
+        int error;
 
-#ifdef __i386__
-static const struct ds_configuration ds_cfg_netburst = {
-        .sizeof_ds = 9 * 4,
-        .bts_buffer_base = { 0, 4 },
-        .bts_index = { 4, 4 },
-        .bts_absolute_maximum = { 8, 4 },
-        .bts_interrupt_threshold = { 12, 4 },
-        .sizeof_bts = 3 * 4,
-        .from_ip = { 0, 4 },
-        .to_ip = { 4, 4 },
-        .info_type = { 4, 1 },
-        .info_data = { 8, 4 },
-        .debugctl_mask = (1<<2)|(1<<3)
-};
+        context = ds_get_context(task);
+        error = ds_validate_access(context, ds_pebs);
+        if (error < 0)
+                goto out;
 
-static const struct ds_configuration ds_cfg_pentium_m = {
-        .sizeof_ds = 9 * 4,
-        .bts_buffer_base = { 0, 4 },
-        .bts_index = { 4, 4 },
-        .bts_absolute_maximum = { 8, 4 },
-        .bts_interrupt_threshold = { 12, 4 },
-        .sizeof_bts = 3 * 4,
-        .from_ip = { 0, 4 },
-        .to_ip = { 4, 4 },
-        .info_type = { 4, 1 },
-        .info_data = { 8, 4 },
-        .debugctl_mask = (1<<6)|(1<<7)
+        *(u64 *)(context->ds + (ds_cfg.sizeof_field * 8)) = value;
+
+        error = 0;
+ out:
+        ds_put_context(context);
+        return error;
+}
+
+static const struct ds_configuration ds_cfg_var = {
+        .sizeof_ds    = sizeof(long) * 12,
+        .sizeof_field = sizeof(long),
+        .sizeof_rec[ds_bts]   = sizeof(long) * 3,
+        .sizeof_rec[ds_pebs]  = sizeof(long) * 10
 };
-#endif /* _i386_ */
-
-static const struct ds_configuration ds_cfg_core2 = {
-        .sizeof_ds = 9 * 8,
-        .bts_buffer_base = { 0, 8 },
-        .bts_index = { 8, 8 },
-        .bts_absolute_maximum = { 16, 8 },
-        .bts_interrupt_threshold = { 24, 8 },
-        .sizeof_bts = 3 * 8,
-        .from_ip = { 0, 8 },
-        .to_ip = { 8, 8 },
-        .info_type = { 8, 1 },
-        .info_data = { 16, 8 },
-        .debugctl_mask = (1<<6)|(1<<7)|(1<<9)
+static const struct ds_configuration ds_cfg_64 = {
+        .sizeof_ds    = 8 * 12,
+        .sizeof_field = 8,
+        .sizeof_rec[ds_bts]   = 8 * 3,
+        .sizeof_rec[ds_pebs]  = 8 * 10
 };
 
 static inline void
@@ -429,14 +821,13 @@ void __cpuinit ds_init_intel(struct cpuinfo_x86 *c)
         switch (c->x86) {
         case 0x6:
                 switch (c->x86_model) {
-#ifdef __i386__
                 case 0xD:
                 case 0xE: /* Pentium M */
-                        ds_configure(&ds_cfg_pentium_m);
+                        ds_configure(&ds_cfg_var);
                         break;
-#endif /* _i386_ */
                 case 0xF: /* Core2 */
-                        ds_configure(&ds_cfg_core2);
+                case 0x1C: /* Atom */
+                        ds_configure(&ds_cfg_64);
                         break;
                 default:
                         /* sorry, don't know about them */
@@ -445,13 +836,11 @@ void __cpuinit ds_init_intel(struct cpuinfo_x86 *c)
                 break;
         case 0xF:
                 switch (c->x86_model) {
-#ifdef __i386__
                 case 0x0:
                 case 0x1:
                 case 0x2: /* Netburst */
-                        ds_configure(&ds_cfg_netburst);
+                        ds_configure(&ds_cfg_var);
                         break;
-#endif /* _i386_ */
                 default:
                         /* sorry, don't know about them */
                         break;
@@ -462,3 +851,14 @@ void __cpuinit ds_init_intel(struct cpuinfo_x86 *c)
                 break;
         }
 }
+
+void ds_free(struct ds_context *context)
+{
+        /* This is called when the task owning the parameter context
+         * is dying. There should not be any user of that context left
+         * to disturb us, anymore. */
+        unsigned long leftovers = context->count;
+        while (leftovers--)
+                ds_put_context(context);
+}
+#endif /* CONFIG_X86_DS */
diff --git a/arch/x86/kernel/efi.c b/arch/x86/kernel/efi.c
index 06cc8d4254b1..945a31cdd81f 100644
--- a/arch/x86/kernel/efi.c
+++ b/arch/x86/kernel/efi.c
@@ -414,9 +414,11 @@ void __init efi_init(void)
         if (memmap.map == NULL)
                 printk(KERN_ERR "Could not map the EFI memory map!\n");
         memmap.map_end = memmap.map + (memmap.nr_map * memmap.desc_size);
+
         if (memmap.desc_size != sizeof(efi_memory_desc_t))
-                printk(KERN_WARNING "Kernel-defined memdesc"
-                       "doesn't match the one from EFI!\n");
+                printk(KERN_WARNING
+                  "Kernel-defined memdesc doesn't match the one from EFI!\n");
+
         if (add_efi_memmap)
                 do_add_efi_memmap();
 
diff --git a/arch/x86/kernel/entry_64.S b/arch/x86/kernel/entry_64.S
index 89434d439605..cf3a0b2d0059 100644
--- a/arch/x86/kernel/entry_64.S
+++ b/arch/x86/kernel/entry_64.S
@@ -275,9 +275,9 @@ ENTRY(native_usergs_sysret64)
 ENTRY(ret_from_fork)
         CFI_DEFAULT_STACK
         push kernel_eflags(%rip)
-        CFI_ADJUST_CFA_OFFSET 4
+        CFI_ADJUST_CFA_OFFSET 8
         popf                            # reset kernel eflags
-        CFI_ADJUST_CFA_OFFSET -4
+        CFI_ADJUST_CFA_OFFSET -8
         call schedule_tail
         GET_THREAD_INFO(%rcx)
         testl $(_TIF_SYSCALL_TRACE|_TIF_SYSCALL_AUDIT),TI_flags(%rcx)
diff --git a/arch/x86/kernel/head64.c b/arch/x86/kernel/head64.c
index 9bfc4d72fb2e..d16084f90649 100644
--- a/arch/x86/kernel/head64.c
+++ b/arch/x86/kernel/head64.c
@@ -108,12 +108,11 @@ void __init x86_64_start_kernel(char * real_mode_data)
         }
         load_idt((const struct desc_ptr *)&idt_descr);
 
-        early_printk("Kernel alive\n");
+        if (console_loglevel == 10)
+                early_printk("Kernel alive\n");
 
         x86_64_init_pda();
 
-        early_printk("Kernel really alive\n");
-
         x86_64_start_reservations(real_mode_data);
 }
 
diff --git a/arch/x86/kernel/ioport.c b/arch/x86/kernel/ioport.c
index 50e5e4a31c85..191914302744 100644
--- a/arch/x86/kernel/ioport.c
+++ b/arch/x86/kernel/ioport.c
@@ -14,6 +14,7 @@
 #include <linux/slab.h>
 #include <linux/thread_info.h>
 #include <linux/syscalls.h>
+#include <asm/syscalls.h>
 
 /* Set EXTENT bits starting at BASE in BITMAP to value TURN_ON. */
 static void set_bitmap(unsigned long *bitmap, unsigned int base,
diff --git a/arch/x86/kernel/ipi.c b/arch/x86/kernel/ipi.c
index 3f7537b669d3..f1c688e46f35 100644
--- a/arch/x86/kernel/ipi.c
+++ b/arch/x86/kernel/ipi.c
@@ -20,6 +20,8 @@
 
 #ifdef CONFIG_X86_32
 #include <mach_apic.h>
+#include <mach_ipi.h>
+
 /*
  * the following functions deal with sending IPIs between CPUs.
  *
@@ -147,7 +149,6 @@ void send_IPI_mask_sequence(cpumask_t mask, int vector)
 }
 
 /* must come after the send_IPI functions above for inlining */
-#include <mach_ipi.h>
 static int convert_apicid_to_cpu(int apic_id)
 {
         int i;
diff --git a/arch/x86/kernel/irq_32.c b/arch/x86/kernel/irq_32.c
index 1cf8c1fcc088..b71e02d42f4f 100644
--- a/arch/x86/kernel/irq_32.c
+++ b/arch/x86/kernel/irq_32.c
@@ -325,7 +325,7 @@ skip:
                 for_each_online_cpu(j)
                         seq_printf(p, "%10u ",
                                 per_cpu(irq_stat,j).irq_call_count);
-                seq_printf(p, "  function call interrupts\n");
+                seq_printf(p, "  Function call interrupts\n");
                 seq_printf(p, "TLB: ");
                 for_each_online_cpu(j)
                         seq_printf(p, "%10u ",
diff --git a/arch/x86/kernel/irq_64.c b/arch/x86/kernel/irq_64.c
index 1f78b238d8d2..f065fe9071b9 100644
--- a/arch/x86/kernel/irq_64.c
+++ b/arch/x86/kernel/irq_64.c
@@ -129,7 +129,7 @@ skip:
                 seq_printf(p, "CAL: ");
                 for_each_online_cpu(j)
                         seq_printf(p, "%10u ", cpu_pda(j)->irq_call_count);
-                seq_printf(p, "  function call interrupts\n");
+                seq_printf(p, "  Function call interrupts\n");
                 seq_printf(p, "TLB: ");
                 for_each_online_cpu(j)
                         seq_printf(p, "%10u ", cpu_pda(j)->irq_tlb_count);
diff --git a/arch/x86/kernel/kdebugfs.c b/arch/x86/kernel/kdebugfs.c
index f2d43bc75514..ff7d3b0124f1 100644
--- a/arch/x86/kernel/kdebugfs.c
+++ b/arch/x86/kernel/kdebugfs.c
@@ -139,6 +139,7 @@ static int __init create_setup_data_nodes(struct dentry *parent)
                 if (PageHighMem(pg)) {
                         data = ioremap_cache(pa_data, sizeof(*data));
                         if (!data) {
+                                kfree(node);
                                 error = -ENXIO;
                                 goto err_dir;
                         }
diff --git a/arch/x86/kernel/kgdb.c b/arch/x86/kernel/kgdb.c
index f47f0eb886b8..10435a120d22 100644
--- a/arch/x86/kernel/kgdb.c
+++ b/arch/x86/kernel/kgdb.c
@@ -69,6 +69,9 @@ static int gdb_x86vector = -1;
  */
 void pt_regs_to_gdb_regs(unsigned long *gdb_regs, struct pt_regs *regs)
 {
+#ifndef CONFIG_X86_32
+        u32 *gdb_regs32 = (u32 *)gdb_regs;
+#endif
         gdb_regs[GDB_AX]        = regs->ax;
         gdb_regs[GDB_BX]        = regs->bx;
         gdb_regs[GDB_CX]        = regs->cx;
@@ -76,9 +79,9 @@ void pt_regs_to_gdb_regs(unsigned long *gdb_regs, struct pt_regs *regs)
         gdb_regs[GDB_SI]        = regs->si;
         gdb_regs[GDB_DI]        = regs->di;
         gdb_regs[GDB_BP]        = regs->bp;
-        gdb_regs[GDB_PS]        = regs->flags;
         gdb_regs[GDB_PC]        = regs->ip;
 #ifdef CONFIG_X86_32
+        gdb_regs[GDB_PS]        = regs->flags;
         gdb_regs[GDB_DS]        = regs->ds;
         gdb_regs[GDB_ES]        = regs->es;
         gdb_regs[GDB_CS]        = regs->cs;
@@ -94,6 +97,9 @@ void pt_regs_to_gdb_regs(unsigned long *gdb_regs, struct pt_regs *regs)
         gdb_regs[GDB_R13]       = regs->r13;
         gdb_regs[GDB_R14]       = regs->r14;
         gdb_regs[GDB_R15]       = regs->r15;
+        gdb_regs32[GDB_PS]      = regs->flags;
+        gdb_regs32[GDB_CS]      = regs->cs;
+        gdb_regs32[GDB_SS]      = regs->ss;
 #endif
         gdb_regs[GDB_SP]        = regs->sp;
 }
@@ -112,6 +118,9 @@ void pt_regs_to_gdb_regs(unsigned long *gdb_regs, struct pt_regs *regs)
  */
 void sleeping_thread_to_gdb_regs(unsigned long *gdb_regs, struct task_struct *p)
 {
+#ifndef CONFIG_X86_32
+        u32 *gdb_regs32 = (u32 *)gdb_regs;
+#endif
         gdb_regs[GDB_AX]        = 0;
         gdb_regs[GDB_BX]        = 0;
         gdb_regs[GDB_CX]        = 0;
@@ -129,8 +138,10 @@ void sleeping_thread_to_gdb_regs(unsigned long *gdb_regs, struct task_struct *p)
         gdb_regs[GDB_FS]        = 0xFFFF;
         gdb_regs[GDB_GS]        = 0xFFFF;
 #else
-        gdb_regs[GDB_PS]        = *(unsigned long *)(p->thread.sp + 8);
-        gdb_regs[GDB_PC]        = 0;
+        gdb_regs32[GDB_PS]      = *(unsigned long *)(p->thread.sp + 8);
+        gdb_regs32[GDB_CS]      = __KERNEL_CS;
+        gdb_regs32[GDB_SS]      = __KERNEL_DS;
+        gdb_regs[GDB_PC]        = p->thread.ip;
         gdb_regs[GDB_R8]        = 0;
         gdb_regs[GDB_R9]        = 0;
         gdb_regs[GDB_R10]       = 0;
@@ -153,6 +164,9 @@ void sleeping_thread_to_gdb_regs(unsigned long *gdb_regs, struct task_struct *p)
  */
 void gdb_regs_to_pt_regs(unsigned long *gdb_regs, struct pt_regs *regs)
 {
+#ifndef CONFIG_X86_32
+        u32 *gdb_regs32 = (u32 *)gdb_regs;
+#endif
         regs->ax                = gdb_regs[GDB_AX];
         regs->bx                = gdb_regs[GDB_BX];
         regs->cx                = gdb_regs[GDB_CX];
@@ -160,9 +174,9 @@ void gdb_regs_to_pt_regs(unsigned long *gdb_regs, struct pt_regs *regs)
         regs->si                = gdb_regs[GDB_SI];
         regs->di                = gdb_regs[GDB_DI];
         regs->bp                = gdb_regs[GDB_BP];
-        regs->flags             = gdb_regs[GDB_PS];
         regs->ip                = gdb_regs[GDB_PC];
 #ifdef CONFIG_X86_32
+        regs->flags             = gdb_regs[GDB_PS];
         regs->ds                = gdb_regs[GDB_DS];
         regs->es                = gdb_regs[GDB_ES];
         regs->cs                = gdb_regs[GDB_CS];
@@ -175,6 +189,9 @@ void gdb_regs_to_pt_regs(unsigned long *gdb_regs, struct pt_regs *regs)
         regs->r13               = gdb_regs[GDB_R13];
         regs->r14               = gdb_regs[GDB_R14];
         regs->r15               = gdb_regs[GDB_R15];
+        regs->flags             = gdb_regs32[GDB_PS];
+        regs->cs                = gdb_regs32[GDB_CS];
+        regs->ss                = gdb_regs32[GDB_SS];
 #endif
 }
 
@@ -378,10 +395,8 @@ int kgdb_arch_handle_exception(int e_vector, int signo, int err_code,
                 if (remcomInBuffer[0] == 's') {
                         linux_regs->flags |= X86_EFLAGS_TF;
                         kgdb_single_step = 1;
-                        if (kgdb_contthread) {
-                                atomic_set(&kgdb_cpu_doing_single_step,
-                                           raw_smp_processor_id());
-                        }
+                        atomic_set(&kgdb_cpu_doing_single_step,
+                                   raw_smp_processor_id());
                 }
 
                 get_debugreg(dr6, 6);
@@ -440,12 +455,7 @@ static int __kgdb_notify(struct die_args *args, unsigned long cmd)
                 return NOTIFY_DONE;
 
         case DIE_NMI_IPI:
-                if (atomic_read(&kgdb_active) != -1) {
-                        /* KGDB CPU roundup */
-                        kgdb_nmicallback(raw_smp_processor_id(), regs);
-                        was_in_debug_nmi[raw_smp_processor_id()] = 1;
-                        touch_nmi_watchdog();
-                }
+                /* Just ignore, we will handle the roundup on DIE_NMI. */
                 return NOTIFY_DONE;
 
         case DIE_NMIUNKNOWN:
@@ -466,9 +476,15 @@ static int __kgdb_notify(struct die_args *args, unsigned long cmd)
 
         case DIE_DEBUG:
                 if (atomic_read(&kgdb_cpu_doing_single_step) ==
-                        raw_smp_processor_id() &&
-                        user_mode(regs))
-                        return single_step_cont(regs, args);
+                    raw_smp_processor_id()) {
+                        if (user_mode(regs))
+                                return single_step_cont(regs, args);
+                        break;
+                } else if (test_thread_flag(TIF_SINGLESTEP))
+                        /* This means a user thread is single stepping
+                         * a system call which should be ignored
+                         */
+                        return NOTIFY_DONE;
                 /* fall through */
         default:
                 if (user_mode(regs))
diff --git a/arch/x86/kernel/kvm.c b/arch/x86/kernel/kvm.c
index 8b7a3cf37d2b..478bca986eca 100644
--- a/arch/x86/kernel/kvm.c
+++ b/arch/x86/kernel/kvm.c
@@ -178,7 +178,7 @@ static void kvm_flush_tlb(void)
         kvm_deferred_mmu_op(&ftlb, sizeof ftlb);
 }
 
-static void kvm_release_pt(u32 pfn)
+static void kvm_release_pt(unsigned long pfn)
 {
         struct kvm_mmu_op_release_pt rpt = {
                 .header.op = KVM_MMU_OP_RELEASE_PT,
diff --git a/arch/x86/kernel/ldt.c b/arch/x86/kernel/ldt.c
index b68e21f06f4f..0ed5f939b905 100644
--- a/arch/x86/kernel/ldt.c
+++ b/arch/x86/kernel/ldt.c
@@ -18,6 +18,7 @@
 #include <asm/ldt.h>
 #include <asm/desc.h>
 #include <asm/mmu_context.h>
+#include <asm/syscalls.h>
 
 #ifdef CONFIG_SMP
 static void flush_ldt(void *current_mm)
diff --git a/arch/x86/kernel/nmi.c b/arch/x86/kernel/nmi.c
index abb78a2cc4ad..2c97f07f1c2c 100644
--- a/arch/x86/kernel/nmi.c
+++ b/arch/x86/kernel/nmi.c
@@ -299,6 +299,15 @@ void acpi_nmi_disable(void)
                 on_each_cpu(__acpi_nmi_disable, NULL, 1);
 }
 
+/*
+ * This function is called as soon the LAPIC NMI watchdog driver has everything
+ * in place and it's ready to check if the NMIs belong to the NMI watchdog
+ */
+void cpu_nmi_set_wd_enabled(void)
+{
+        __get_cpu_var(wd_enabled) = 1;
+}
+
 void setup_apic_nmi_watchdog(void *unused)
 {
         if (__get_cpu_var(wd_enabled))
@@ -311,8 +320,6 @@ void setup_apic_nmi_watchdog(void *unused)
 
         switch (nmi_watchdog) {
         case NMI_LOCAL_APIC:
-                 /* enable it before to avoid race with handler */
-                __get_cpu_var(wd_enabled) = 1;
                 if (lapic_watchdog_init(nmi_hz) < 0) {
                         __get_cpu_var(wd_enabled) = 0;
                         return;
diff --git a/arch/x86/kernel/olpc.c b/arch/x86/kernel/olpc.c
index 3e6672274807..7a13fac63a1f 100644
--- a/arch/x86/kernel/olpc.c
+++ b/arch/x86/kernel/olpc.c
@@ -190,12 +190,12 @@ EXPORT_SYMBOL_GPL(olpc_ec_cmd);
 static void __init platform_detect(void)
 {
         size_t propsize;
-        u32 rev;
+        __be32 rev;
 
         if (ofw("getprop", 4, 1, NULL, "board-revision-int", &rev, 4,
                         &propsize) || propsize != 4) {
                 printk(KERN_ERR "ofw: getprop call failed!\n");
-                rev = 0;
+                rev = cpu_to_be32(0);
         }
         olpc_platform_info.boardrev = be32_to_cpu(rev);
 }
@@ -203,7 +203,7 @@ static void __init platform_detect(void)
 static void __init platform_detect(void)
 {
         /* stopgap until OFW support is added to the kernel */
-        olpc_platform_info.boardrev = be32_to_cpu(0xc2);
+        olpc_platform_info.boardrev = 0xc2;
 }
 #endif
 
diff --git a/arch/x86/kernel/paravirt.c b/arch/x86/kernel/paravirt.c
index 300da17e61cb..e2f43768723a 100644
--- a/arch/x86/kernel/paravirt.c
+++ b/arch/x86/kernel/paravirt.c
@@ -330,6 +330,7 @@ struct pv_cpu_ops pv_cpu_ops = {
 #endif
         .wbinvd = native_wbinvd,
         .read_msr = native_read_msr_safe,
+        .read_msr_amd = native_read_msr_amd_safe,
         .write_msr = native_write_msr_safe,
         .read_tsc = native_read_tsc,
         .read_pmc = native_read_pmc,
diff --git a/arch/x86/kernel/paravirt_patch_32.c b/arch/x86/kernel/paravirt_patch_32.c
index 58262218781b..9fe644f4861d 100644
--- a/arch/x86/kernel/paravirt_patch_32.c
+++ b/arch/x86/kernel/paravirt_patch_32.c
@@ -23,7 +23,7 @@ unsigned native_patch(u8 type, u16 clobbers, void *ibuf,
                         start = start_##ops##_##x;              \
                         end = end_##ops##_##x;                  \
                         goto patch_site
-        switch(type) {
+        switch (type) {
                 PATCH_SITE(pv_irq_ops, irq_disable);
                 PATCH_SITE(pv_irq_ops, irq_enable);
                 PATCH_SITE(pv_irq_ops, restore_fl);
diff --git a/arch/x86/kernel/pci-dma.c b/arch/x86/kernel/pci-dma.c
index 87d4d6964ec2..f704cb51ff82 100644
--- a/arch/x86/kernel/pci-dma.c
+++ b/arch/x86/kernel/pci-dma.c
@@ -82,7 +82,7 @@ void __init dma32_reserve_bootmem(void)
          * using 512M as goal
          */
         align = 64ULL<<20;
-        size = round_up(dma32_bootmem_size, align);
+        size = roundup(dma32_bootmem_size, align);
         dma32_bootmem_ptr = __alloc_bootmem_nopanic(size, align,
                                  512ULL<<20);
         if (dma32_bootmem_ptr)
diff --git a/arch/x86/kernel/pci-gart_64.c b/arch/x86/kernel/pci-gart_64.c
index 49285f8fd4d5..1a895a582534 100644
--- a/arch/x86/kernel/pci-gart_64.c
+++ b/arch/x86/kernel/pci-gart_64.c
@@ -82,7 +82,8 @@ AGPEXTERN __u32 *agp_gatt_table;
 static unsigned long next_bit;  /* protected by iommu_bitmap_lock */
 static int need_flush;          /* global flush state. set for each gart wrap */
 
-static unsigned long alloc_iommu(struct device *dev, int size)
+static unsigned long alloc_iommu(struct device *dev, int size,
+                                 unsigned long align_mask)
 {
         unsigned long offset, flags;
         unsigned long boundary_size;
@@ -90,16 +91,17 @@ static unsigned long alloc_iommu(struct device *dev, int size)
 
         base_index = ALIGN(iommu_bus_base & dma_get_seg_boundary(dev),
                            PAGE_SIZE) >> PAGE_SHIFT;
-        boundary_size = ALIGN(dma_get_seg_boundary(dev) + 1,
+        boundary_size = ALIGN((unsigned long long)dma_get_seg_boundary(dev) + 1,
                               PAGE_SIZE) >> PAGE_SHIFT;
 
         spin_lock_irqsave(&iommu_bitmap_lock, flags);
         offset = iommu_area_alloc(iommu_gart_bitmap, iommu_pages, next_bit,
-                                  size, base_index, boundary_size, 0);
+                                  size, base_index, boundary_size, align_mask);
         if (offset == -1) {
                 need_flush = 1;
                 offset = iommu_area_alloc(iommu_gart_bitmap, iommu_pages, 0,
-                                          size, base_index, boundary_size, 0);
+                                          size, base_index, boundary_size,
+                                          align_mask);
         }
         if (offset != -1) {
                 next_bit = offset+size;
@@ -236,10 +238,10 @@ nonforced_iommu(struct device *dev, unsigned long addr, size_t size)
  * Caller needs to check if the iommu is needed and flush.
  */
 static dma_addr_t dma_map_area(struct device *dev, dma_addr_t phys_mem,
-                                size_t size, int dir)
+                                size_t size, int dir, unsigned long align_mask)
 {
         unsigned long npages = iommu_num_pages(phys_mem, size);
-        unsigned long iommu_page = alloc_iommu(dev, npages);
+        unsigned long iommu_page = alloc_iommu(dev, npages, align_mask);
         int i;
 
         if (iommu_page == -1) {
@@ -262,7 +264,11 @@ static dma_addr_t dma_map_area(struct device *dev, dma_addr_t phys_mem,
 static dma_addr_t
 gart_map_simple(struct device *dev, phys_addr_t paddr, size_t size, int dir)
 {
-        dma_addr_t map = dma_map_area(dev, paddr, size, dir);
+        dma_addr_t map;
+        unsigned long align_mask;
+
+        align_mask = (1UL << get_order(size)) - 1;
+        map = dma_map_area(dev, paddr, size, dir, align_mask);
 
         flush_gart();
 
@@ -281,7 +287,8 @@ gart_map_single(struct device *dev, phys_addr_t paddr, size_t size, int dir)
         if (!need_iommu(dev, paddr, size))
                 return paddr;
 
-        bus = gart_map_simple(dev, paddr, size, dir);
+        bus = dma_map_area(dev, paddr, size, dir, 0);
+        flush_gart();
 
         return bus;
 }
@@ -340,7 +347,7 @@ static int dma_map_sg_nonforce(struct device *dev, struct scatterlist *sg,
                 unsigned long addr = sg_phys(s);
 
                 if (nonforced_iommu(dev, addr, s->length)) {
-                        addr = dma_map_area(dev, addr, s->length, dir);
+                        addr = dma_map_area(dev, addr, s->length, dir, 0);
                         if (addr == bad_dma_address) {
                                 if (i > 0)
                                         gart_unmap_sg(dev, sg, i, dir);
@@ -362,7 +369,7 @@ static int __dma_map_cont(struct device *dev, struct scatterlist *start,
                           int nelems, struct scatterlist *sout,
                           unsigned long pages)
 {
-        unsigned long iommu_start = alloc_iommu(dev, pages);
+        unsigned long iommu_start = alloc_iommu(dev, pages, 0);
         unsigned long iommu_page = iommu_start;
         struct scatterlist *s;
         int i;
@@ -626,7 +633,6 @@ static __init int init_k8_gatt(struct agp_kern_info *info)
         struct pci_dev *dev;
         void *gatt;
         int i, error;
-        unsigned long start_pfn, end_pfn;
 
         printk(KERN_INFO "PCI-DMA: Disabling AGP.\n");
         aper_size = aper_base = info->aper_size = 0;
@@ -672,12 +678,6 @@ static __init int init_k8_gatt(struct agp_kern_info *info)
         printk(KERN_INFO "PCI-DMA: aperture base @ %x size %u KB\n",
                aper_base, aper_size>>10);
 
-        /* need to map that range */
-        end_pfn = (aper_base>>PAGE_SHIFT) + (aper_size>>PAGE_SHIFT);
-        if (end_pfn > max_low_pfn_mapped) {
-                start_pfn = (aper_base>>PAGE_SHIFT);
-                init_memory_mapping(start_pfn<<PAGE_SHIFT, end_pfn<<PAGE_SHIFT);
-        }
         return 0;
 
  nommu:
@@ -727,7 +727,8 @@ void __init gart_iommu_init(void)
 {
         struct agp_kern_info info;
         unsigned long iommu_start;
-        unsigned long aper_size;
+        unsigned long aper_base, aper_size;
+        unsigned long start_pfn, end_pfn;
         unsigned long scratch;
         long i;
 
@@ -765,8 +766,16 @@ void __init gart_iommu_init(void)
                 return;
         }
 
+        /* need to map that range */
+        aper_size = info.aper_size << 20;
+        aper_base = info.aper_base;
+        end_pfn = (aper_base>>PAGE_SHIFT) + (aper_size>>PAGE_SHIFT);
+        if (end_pfn > max_low_pfn_mapped) {
+                start_pfn = (aper_base>>PAGE_SHIFT);
+                init_memory_mapping(start_pfn<<PAGE_SHIFT, end_pfn<<PAGE_SHIFT);
+        }
+
         printk(KERN_INFO "PCI-DMA: using GART IOMMU.\n");
-        aper_size = info.aper_size * 1024 * 1024;
         iommu_size = check_iommu_size(info.aper_base, aper_size);
         iommu_pages = iommu_size >> PAGE_SHIFT;
 
diff --git a/arch/x86/kernel/pcspeaker.c b/arch/x86/kernel/pcspeaker.c
index bc1f2d3ea277..a311ffcaad16 100644
--- a/arch/x86/kernel/pcspeaker.c
+++ b/arch/x86/kernel/pcspeaker.c
@@ -1,20 +1,13 @@
 #include <linux/platform_device.h>
-#include <linux/errno.h>
+#include <linux/err.h>
 #include <linux/init.h>
 
 static __init int add_pcspkr(void)
 {
         struct platform_device *pd;
-        int ret;
 
-        pd = platform_device_alloc("pcspkr", -1);
-        if (!pd)
-                return -ENOMEM;
+        pd = platform_device_register_simple("pcspkr", -1, NULL, 0);
 
-        ret = platform_device_add(pd);
-        if (ret)
-                platform_device_put(pd);
-
-        return ret;
+        return IS_ERR(pd) ? PTR_ERR(pd) : 0;
 }
 device_initcall(add_pcspkr);
diff --git a/arch/x86/kernel/process.c b/arch/x86/kernel/process.c
index 7fc4d5b0a6a0..ec7a2ba9bce8 100644
--- a/arch/x86/kernel/process.c
+++ b/arch/x86/kernel/process.c
@@ -185,7 +185,8 @@ static void mwait_idle(void)
 static void poll_idle(void)
 {
         local_irq_enable();
-        cpu_relax();
+        while (!need_resched())
+                cpu_relax();
 }
 
 /*
@@ -246,6 +247,14 @@ static int __cpuinit check_c1e_idle(const struct cpuinfo_x86 *c)
         return 1;
 }
 
+static cpumask_t c1e_mask = CPU_MASK_NONE;
+static int c1e_detected;
+
+void c1e_remove_cpu(int cpu)
+{
+        cpu_clear(cpu, c1e_mask);
+}
+
 /*
  * C1E aware idle routine. We check for C1E active in the interrupt
  * pending message MSR. If we detect C1E, then we handle it the same
@@ -253,9 +262,6 @@ static int __cpuinit check_c1e_idle(const struct cpuinfo_x86 *c)
  */
 static void c1e_idle(void)
 {
-        static cpumask_t c1e_mask = CPU_MASK_NONE;
-        static int c1e_detected;
-
         if (need_resched())
                 return;
 
@@ -265,8 +271,10 @@ static void c1e_idle(void)
                 rdmsr(MSR_K8_INT_PENDING_MSG, lo, hi);
                 if (lo & K8_INTP_C1E_ACTIVE_MASK) {
                         c1e_detected = 1;
-                        mark_tsc_unstable("TSC halt in C1E");
-                        printk(KERN_INFO "System has C1E enabled\n");
+                        if (!boot_cpu_has(X86_FEATURE_CONSTANT_TSC))
+                                mark_tsc_unstable("TSC halt in AMD C1E");
+                        printk(KERN_INFO "System has AMD C1E enabled\n");
+                        set_cpu_cap(&boot_cpu_data, X86_FEATURE_AMDC1E);
                 }
         }
 
diff --git a/arch/x86/kernel/process_32.c b/arch/x86/kernel/process_32.c
index 3b7a1ddcc0bc..205188db9626 100644
--- a/arch/x86/kernel/process_32.c
+++ b/arch/x86/kernel/process_32.c
@@ -37,6 +37,7 @@
 #include <linux/tick.h>
 #include <linux/percpu.h>
 #include <linux/prctl.h>
+#include <linux/dmi.h>
 
 #include <asm/uaccess.h>
 #include <asm/pgtable.h>
@@ -55,6 +56,9 @@
 #include <asm/tlbflush.h>
 #include <asm/cpu.h>
 #include <asm/kdebug.h>
+#include <asm/idle.h>
+#include <asm/syscalls.h>
+#include <asm/smp.h>
 
 asmlinkage void ret_from_fork(void) __asm__("ret_from_fork");
 
@@ -88,6 +92,7 @@ static void cpu_exit_clear(void)
         cpu_clear(cpu, cpu_callin_map);
 
         numa_remove_cpu(cpu);
+        c1e_remove_cpu(cpu);
 }
 
 /* We don't actually take CPU down, just spin without interrupts. */
@@ -159,6 +164,7 @@ void __show_registers(struct pt_regs *regs, int all)
         unsigned long d0, d1, d2, d3, d6, d7;
         unsigned long sp;
         unsigned short ss, gs;
+        const char *board;
 
         if (user_mode_vm(regs)) {
                 sp = regs->sp;
@@ -171,11 +177,15 @@ void __show_registers(struct pt_regs *regs, int all)
         }
 
         printk("\n");
-        printk("Pid: %d, comm: %s %s (%s %.*s)\n",
+
+        board = dmi_get_system_info(DMI_PRODUCT_NAME);
+        if (!board)
+                board = "";
+        printk("Pid: %d, comm: %s %s (%s %.*s) %s\n",
                         task_pid_nr(current), current->comm,
                         print_tainted(), init_utsname()->release,
                         (int)strcspn(init_utsname()->version, " "),
-                        init_utsname()->version);
+                        init_utsname()->version, board);
 
         printk("EIP: %04x:[<%08lx>] EFLAGS: %08lx CPU: %d\n",
                         (u16)regs->cs, regs->ip, regs->flags,
@@ -275,6 +285,14 @@ void exit_thread(void)
                 tss->x86_tss.io_bitmap_base = INVALID_IO_BITMAP_OFFSET;
                 put_cpu();
         }
+#ifdef CONFIG_X86_DS
+        /* Free any DS contexts that have not been properly released. */
+        if (unlikely(current->thread.ds_ctx)) {
+                /* we clear debugctl to make sure DS is not used. */
+                update_debugctlmsr(0);
+                ds_free(current->thread.ds_ctx);
+        }
+#endif /* CONFIG_X86_DS */
 }
 
 void flush_thread(void)
@@ -436,6 +454,35 @@ int set_tsc_mode(unsigned int val)
         return 0;
 }
 
+#ifdef CONFIG_X86_DS
+static int update_debugctl(struct thread_struct *prev,
+                        struct thread_struct *next, unsigned long debugctl)
+{
+        unsigned long ds_prev = 0;
+        unsigned long ds_next = 0;
+
+        if (prev->ds_ctx)
+                ds_prev = (unsigned long)prev->ds_ctx->ds;
+        if (next->ds_ctx)
+                ds_next = (unsigned long)next->ds_ctx->ds;
+
+        if (ds_next != ds_prev) {
+                /* we clear debugctl to make sure DS
+                 * is not in use when we change it */
+                debugctl = 0;
+                update_debugctlmsr(0);
+                wrmsr(MSR_IA32_DS_AREA, ds_next, 0);
+        }
+        return debugctl;
+}
+#else
+static int update_debugctl(struct thread_struct *prev,
+                        struct thread_struct *next, unsigned long debugctl)
+{
+        return debugctl;
+}
+#endif /* CONFIG_X86_DS */
+
 static noinline void
 __switch_to_xtra(struct task_struct *prev_p, struct task_struct *next_p,
                  struct tss_struct *tss)
@@ -446,14 +493,7 @@ __switch_to_xtra(struct task_struct *prev_p, struct task_struct *next_p,
         prev = &prev_p->thread;
         next = &next_p->thread;
 
-        debugctl = prev->debugctlmsr;
-        if (next->ds_area_msr != prev->ds_area_msr) {
-                /* we clear debugctl to make sure DS
-                 * is not in use when we change it */
-                debugctl = 0;
-                update_debugctlmsr(0);
-                wrmsr(MSR_IA32_DS_AREA, next->ds_area_msr, 0);
-        }
+        debugctl = update_debugctl(prev, next, prev->debugctlmsr);
 
         if (next->debugctlmsr != debugctl)
                 update_debugctlmsr(next->debugctlmsr);
@@ -477,13 +517,13 @@ __switch_to_xtra(struct task_struct *prev_p, struct task_struct *next_p,
                         hard_enable_TSC();
         }
 
-#ifdef X86_BTS
+#ifdef CONFIG_X86_PTRACE_BTS
         if (test_tsk_thread_flag(prev_p, TIF_BTS_TRACE_TS))
                 ptrace_bts_take_timestamp(prev_p, BTS_TASK_DEPARTS);
 
         if (test_tsk_thread_flag(next_p, TIF_BTS_TRACE_TS))
                 ptrace_bts_take_timestamp(next_p, BTS_TASK_ARRIVES);
-#endif
+#endif /* CONFIG_X86_PTRACE_BTS */
 
 
         if (!test_tsk_thread_flag(next_p, TIF_IO_BITMAP)) {
diff --git a/arch/x86/kernel/process_64.c b/arch/x86/kernel/process_64.c
index 71553b664e2a..2a8ccb9238b4 100644
--- a/arch/x86/kernel/process_64.c
+++ b/arch/x86/kernel/process_64.c
@@ -37,11 +37,11 @@
 #include <linux/kdebug.h>
 #include <linux/tick.h>
 #include <linux/prctl.h>
+#include <linux/uaccess.h>
+#include <linux/io.h>
 
-#include <asm/uaccess.h>
 #include <asm/pgtable.h>
 #include <asm/system.h>
-#include <asm/io.h>
 #include <asm/processor.h>
 #include <asm/i387.h>
 #include <asm/mmu_context.h>
@@ -51,6 +51,7 @@
 #include <asm/proto.h>
 #include <asm/ia32.h>
 #include <asm/idle.h>
+#include <asm/syscalls.h>
 
 asmlinkage extern void ret_from_fork(void);
 
@@ -88,11 +89,13 @@ void exit_idle(void)
 #ifdef CONFIG_HOTPLUG_CPU
 DECLARE_PER_CPU(int, cpu_state);
 
-#include <asm/nmi.h>
+#include <linux/nmi.h>
 /* We halt the CPU with physical CPU hotplug */
 static inline void play_dead(void)
 {
         idle_task_exit();
+        c1e_remove_cpu(raw_smp_processor_id());
+
         mb();
         /* Ack it */
         __get_cpu_var(cpu_state) = CPU_DEAD;
@@ -151,7 +154,7 @@ void cpu_idle(void)
 }
 
 /* Prints also some state that isn't saved in the pt_regs */
-void __show_regs(struct pt_regs * regs)
+void __show_regs(struct pt_regs *regs)
 {
         unsigned long cr0 = 0L, cr2 = 0L, cr3 = 0L, cr4 = 0L, fs, gs, shadowgs;
         unsigned long d0, d1, d2, d3, d6, d7;
@@ -160,59 +163,61 @@ void __show_regs(struct pt_regs * regs)
 
         printk("\n");
         print_modules();
-        printk("Pid: %d, comm: %.20s %s %s %.*s\n",
+        printk(KERN_INFO "Pid: %d, comm: %.20s %s %s %.*s\n",
                 current->pid, current->comm, print_tainted(),
                 init_utsname()->release,
                 (int)strcspn(init_utsname()->version, " "),
                 init_utsname()->version);
-        printk("RIP: %04lx:[<%016lx>] ", regs->cs & 0xffff, regs->ip);
+        printk(KERN_INFO "RIP: %04lx:[<%016lx>] ", regs->cs & 0xffff, regs->ip);
         printk_address(regs->ip, 1);
-        printk("RSP: %04lx:%016lx  EFLAGS: %08lx\n", regs->ss, regs->sp,
-                regs->flags);
-        printk("RAX: %016lx RBX: %016lx RCX: %016lx\n",
+        printk(KERN_INFO "RSP: %04lx:%016lx  EFLAGS: %08lx\n", regs->ss,
+                        regs->sp, regs->flags);
+        printk(KERN_INFO "RAX: %016lx RBX: %016lx RCX: %016lx\n",
                regs->ax, regs->bx, regs->cx);
-        printk("RDX: %016lx RSI: %016lx RDI: %016lx\n",
+        printk(KERN_INFO "RDX: %016lx RSI: %016lx RDI: %016lx\n",
                regs->dx, regs->si, regs->di);
-        printk("RBP: %016lx R08: %016lx R09: %016lx\n",
+        printk(KERN_INFO "RBP: %016lx R08: %016lx R09: %016lx\n",
                regs->bp, regs->r8, regs->r9);
-        printk("R10: %016lx R11: %016lx R12: %016lx\n",
-               regs->r10, regs->r11, regs->r12); 
-        printk("R13: %016lx R14: %016lx R15: %016lx\n",
-               regs->r13, regs->r14, regs->r15); 
-
-        asm("movl %%ds,%0" : "=r" (ds)); 
-        asm("movl %%cs,%0" : "=r" (cs)); 
-        asm("movl %%es,%0" : "=r" (es)); 
+        printk(KERN_INFO "R10: %016lx R11: %016lx R12: %016lx\n",
+               regs->r10, regs->r11, regs->r12);
+        printk(KERN_INFO "R13: %016lx R14: %016lx R15: %016lx\n",
+               regs->r13, regs->r14, regs->r15);
+
+        asm("movl %%ds,%0" : "=r" (ds));
+        asm("movl %%cs,%0" : "=r" (cs));
+        asm("movl %%es,%0" : "=r" (es));
         asm("movl %%fs,%0" : "=r" (fsindex));
         asm("movl %%gs,%0" : "=r" (gsindex));
 
         rdmsrl(MSR_FS_BASE, fs);
-        rdmsrl(MSR_GS_BASE, gs); 
-        rdmsrl(MSR_KERNEL_GS_BASE, shadowgs); 
+        rdmsrl(MSR_GS_BASE, gs);
+        rdmsrl(MSR_KERNEL_GS_BASE, shadowgs);
 
         cr0 = read_cr0();
         cr2 = read_cr2();
         cr3 = read_cr3();
         cr4 = read_cr4();
 
-        printk("FS:  %016lx(%04x) GS:%016lx(%04x) knlGS:%016lx\n", 
-               fs,fsindex,gs,gsindex,shadowgs); 
-        printk("CS:  %04x DS: %04x ES: %04x CR0: %016lx\n", cs, ds, es, cr0); 
-        printk("CR2: %016lx CR3: %016lx CR4: %016lx\n", cr2, cr3, cr4);
+        printk(KERN_INFO "FS:  %016lx(%04x) GS:%016lx(%04x) knlGS:%016lx\n",
+               fs, fsindex, gs, gsindex, shadowgs);
+        printk(KERN_INFO "CS:  %04x DS: %04x ES: %04x CR0: %016lx\n", cs, ds,
+                        es, cr0);
+        printk(KERN_INFO "CR2: %016lx CR3: %016lx CR4: %016lx\n", cr2, cr3,
+                        cr4);
 
         get_debugreg(d0, 0);
         get_debugreg(d1, 1);
         get_debugreg(d2, 2);
-        printk("DR0: %016lx DR1: %016lx DR2: %016lx\n", d0, d1, d2);
+        printk(KERN_INFO "DR0: %016lx DR1: %016lx DR2: %016lx\n", d0, d1, d2);
         get_debugreg(d3, 3);
         get_debugreg(d6, 6);
         get_debugreg(d7, 7);
-        printk("DR3: %016lx DR6: %016lx DR7: %016lx\n", d3, d6, d7);
+        printk(KERN_INFO "DR3: %016lx DR6: %016lx DR7: %016lx\n", d3, d6, d7);
 }
 
 void show_regs(struct pt_regs *regs)
 {
-        printk("CPU %d:", smp_processor_id());
+        printk(KERN_INFO "CPU %d:", smp_processor_id());
         __show_regs(regs);
         show_trace(NULL, regs, (void *)(regs + 1), regs->bp);
 }
@@ -238,6 +243,14 @@ void exit_thread(void)
                 t->io_bitmap_max = 0;
                 put_cpu();
         }
+#ifdef CONFIG_X86_DS
+        /* Free any DS contexts that have not been properly released. */
+        if (unlikely(t->ds_ctx)) {
+                /* we clear debugctl to make sure DS is not used. */
+                update_debugctlmsr(0);
+                ds_free(t->ds_ctx);
+        }
+#endif /* CONFIG_X86_DS */
 }
 
 void flush_thread(void)
@@ -313,10 +326,10 @@ void prepare_to_copy(struct task_struct *tsk)
 
 int copy_thread(int nr, unsigned long clone_flags, unsigned long sp,
                 unsigned long unused,
-        struct task_struct * p, struct pt_regs * regs)
+        struct task_struct *p, struct pt_regs *regs)
 {
         int err;
-        struct pt_regs * childregs;
+        struct pt_regs *childregs;
         struct task_struct *me = current;
 
         childregs = ((struct pt_regs *)
@@ -361,10 +374,10 @@ int copy_thread(int nr, unsigned long clone_flags, unsigned long sp,
                 if (test_thread_flag(TIF_IA32))
                         err = do_set_thread_area(p, -1,
                                 (struct user_desc __user *)childregs->si, 0);
-                else                    
-#endif   
-                        err = do_arch_prctl(p, ARCH_SET_FS, childregs->r8); 
-                if (err) 
+                else
+#endif
+                        err = do_arch_prctl(p, ARCH_SET_FS, childregs->r8);
+                if (err)
                         goto out;
         }
         err = 0;
@@ -471,13 +484,27 @@ static inline void __switch_to_xtra(struct task_struct *prev_p,
         next = &next_p->thread;
 
         debugctl = prev->debugctlmsr;
-        if (next->ds_area_msr != prev->ds_area_msr) {
-                /* we clear debugctl to make sure DS
-                 * is not in use when we change it */
-                debugctl = 0;
-                update_debugctlmsr(0);
-                wrmsrl(MSR_IA32_DS_AREA, next->ds_area_msr);
+
+#ifdef CONFIG_X86_DS
+        {
+                unsigned long ds_prev = 0, ds_next = 0;
+
+                if (prev->ds_ctx)
+                        ds_prev = (unsigned long)prev->ds_ctx->ds;
+                if (next->ds_ctx)
+                        ds_next = (unsigned long)next->ds_ctx->ds;
+
+                if (ds_next != ds_prev) {
+                        /*
+                         * We clear debugctl to make sure DS
+                         * is not in use when we change it:
+                         */
+                        debugctl = 0;
+                        update_debugctlmsr(0);
+                        wrmsrl(MSR_IA32_DS_AREA, ds_next);
+                }
         }
+#endif /* CONFIG_X86_DS */
 
         if (next->debugctlmsr != debugctl)
                 update_debugctlmsr(next->debugctlmsr);
@@ -515,13 +542,13 @@ static inline void __switch_to_xtra(struct task_struct *prev_p,
                 memset(tss->io_bitmap, 0xff, prev->io_bitmap_max);
         }
 
-#ifdef X86_BTS
+#ifdef CONFIG_X86_PTRACE_BTS
         if (test_tsk_thread_flag(prev_p, TIF_BTS_TRACE_TS))
                 ptrace_bts_take_timestamp(prev_p, BTS_TASK_DEPARTS);
 
         if (test_tsk_thread_flag(next_p, TIF_BTS_TRACE_TS))
                 ptrace_bts_take_timestamp(next_p, BTS_TASK_ARRIVES);
-#endif
+#endif /* CONFIG_X86_PTRACE_BTS */
 }
 
 /*
@@ -543,7 +570,7 @@ __switch_to(struct task_struct *prev_p, struct task_struct *next_p)
         unsigned fsindex, gsindex;
 
         /* we're going to use this soon, after a few expensive things */
-        if (next_p->fpu_counter>5)
+        if (next_p->fpu_counter > 5)
                 prefetch(next->xstate);
 
         /*
@@ -551,13 +578,13 @@ __switch_to(struct task_struct *prev_p, struct task_struct *next_p)
          */
         load_sp0(tss, next);
 
-        /* 
+        /*
          * Switch DS and ES.
          * This won't pick up thread selector changes, but I guess that is ok.
          */
         savesegment(es, prev->es);
         if (unlikely(next->es | prev->es))
-                loadsegment(es, next->es); 
+                loadsegment(es, next->es);
 
         savesegment(ds, prev->ds);
         if (unlikely(next->ds | prev->ds))
@@ -583,7 +610,7 @@ __switch_to(struct task_struct *prev_p, struct task_struct *next_p)
          */
         arch_leave_lazy_cpu_mode();
 
-        /* 
+        /*
          * Switch FS and GS.
          *
          * Segment register != 0 always requires a reload.  Also
@@ -592,13 +619,13 @@ __switch_to(struct task_struct *prev_p, struct task_struct *next_p)
          */
         if (unlikely(fsindex | next->fsindex | prev->fs)) {
                 loadsegment(fs, next->fsindex);
-                /* 
+                /*
                  * Check if the user used a selector != 0; if yes
                  *  clear 64bit base, since overloaded base is always
                  *  mapped to the Null selector
                  */
                 if (fsindex)
-                        prev->fs = 0;                           
+                        prev->fs = 0;
         }
         /* when next process has a 64bit base use it */
         if (next->fs)
@@ -608,7 +635,7 @@ __switch_to(struct task_struct *prev_p, struct task_struct *next_p)
         if (unlikely(gsindex | next->gsindex | prev->gs)) {
                 load_gs_index(next->gsindex);
                 if (gsindex)
-                        prev->gs = 0;                           
+                        prev->gs = 0;
         }
         if (next->gs)
                 wrmsrl(MSR_KERNEL_GS_BASE, next->gs);
@@ -617,12 +644,12 @@ __switch_to(struct task_struct *prev_p, struct task_struct *next_p)
         /* Must be after DS reload */
         unlazy_fpu(prev_p);
 
-        /* 
+        /*
          * Switch the PDA and FPU contexts.
          */
         prev->usersp = read_pda(oldrsp);
         write_pda(oldrsp, next->usersp);
-        write_pda(pcurrent, next_p); 
+        write_pda(pcurrent, next_p);
 
         write_pda(kernelstack,
                   (unsigned long)task_stack_page(next_p) +
@@ -663,7 +690,7 @@ long sys_execve(char __user *name, char __user * __user *argv,
                 char __user * __user *envp, struct pt_regs *regs)
 {
         long error;
-        char * filename;
+        char *filename;
 
         filename = getname(name);
         error = PTR_ERR(filename);
@@ -721,55 +748,55 @@ asmlinkage long sys_vfork(struct pt_regs *regs)
 unsigned long get_wchan(struct task_struct *p)
 {
         unsigned long stack;
-        u64 fp,ip;
+        u64 fp, ip;
         int count = 0;
 
-        if (!p || p == current || p->state==TASK_RUNNING)
-                return 0; 
+        if (!p || p == current || p->state == TASK_RUNNING)
+                return 0;
         stack = (unsigned long)task_stack_page(p);
         if (p->thread.sp < stack || p->thread.sp > stack+THREAD_SIZE)
                 return 0;
         fp = *(u64 *)(p->thread.sp);
-        do { 
+        do {
                 if (fp < (unsigned long)stack ||
                     fp > (unsigned long)stack+THREAD_SIZE)
-                        return 0; 
+                        return 0;
                 ip = *(u64 *)(fp+8);
                 if (!in_sched_functions(ip))
                         return ip;
-                fp = *(u64 *)fp; 
-        } while (count++ < 16); 
+                fp = *(u64 *)fp;
+        } while (count++ < 16);
         return 0;
 }
 
 long do_arch_prctl(struct task_struct *task, int code, unsigned long addr)
-{ 
-        int ret = 0; 
+{
+        int ret = 0;
         int doit = task == current;
         int cpu;
 
-        switch (code) { 
+        switch (code) {
         case ARCH_SET_GS:
                 if (addr >= TASK_SIZE_OF(task))
-                        return -EPERM; 
+                        return -EPERM;
                 cpu = get_cpu();
-                /* handle small bases via the GDT because that's faster to 
+                /* handle small bases via the GDT because that's faster to
                    switch. */
-                if (addr <= 0xffffffff) {  
-                        set_32bit_tls(task, GS_TLS, addr); 
-                        if (doit) { 
+                if (addr <= 0xffffffff) {
+                        set_32bit_tls(task, GS_TLS, addr);
+                        if (doit) {
                                 load_TLS(&task->thread, cpu);
-                                load_gs_index(GS_TLS_SEL); 
+                                load_gs_index(GS_TLS_SEL);
                         }
-                        task->thread.gsindex = GS_TLS_SEL; 
+                        task->thread.gsindex = GS_TLS_SEL;
                         task->thread.gs = 0;
-                } else { 
+                } else {
                         task->thread.gsindex = 0;
                         task->thread.gs = addr;
                         if (doit) {
                                 load_gs_index(0);
                                 ret = checking_wrmsrl(MSR_KERNEL_GS_BASE, addr);
-                        } 
+                        }
                 }
                 put_cpu();
                 break;
@@ -823,8 +850,7 @@ long do_arch_prctl(struct task_struct *task, int code, unsigned long addr)
                                 rdmsrl(MSR_KERNEL_GS_BASE, base);
                         else
                                 base = task->thread.gs;
-                }
-                else
+                } else
                         base = task->thread.gs;
                 ret = put_user(base, (unsigned long __user *)addr);
                 break;
diff --git a/arch/x86/kernel/ptrace.c b/arch/x86/kernel/ptrace.c
index e37dccce85db..e375b658efc3 100644
--- a/arch/x86/kernel/ptrace.c
+++ b/arch/x86/kernel/ptrace.c
@@ -14,6 +14,7 @@
 #include <linux/errno.h>
 #include <linux/ptrace.h>
 #include <linux/regset.h>
+#include <linux/tracehook.h>
 #include <linux/user.h>
 #include <linux/elf.h>
 #include <linux/security.h>
@@ -69,7 +70,7 @@ static inline bool invalid_selector(u16 value)
 
 #define FLAG_MASK               FLAG_MASK_32
 
-static long *pt_regs_access(struct pt_regs *regs, unsigned long regno)
+static unsigned long *pt_regs_access(struct pt_regs *regs, unsigned long regno)
 {
         BUILD_BUG_ON(offsetof(struct pt_regs, bx) != 0);
         regno >>= 2;
@@ -554,45 +555,115 @@ static int ptrace_set_debugreg(struct task_struct *child,
         return 0;
 }
 
-#ifdef X86_BTS
+#ifdef CONFIG_X86_PTRACE_BTS
+/*
+ * The configuration for a particular BTS hardware implementation.
+ */
+struct bts_configuration {
+        /* the size of a BTS record in bytes; at most BTS_MAX_RECORD_SIZE */
+        unsigned char  sizeof_bts;
+        /* the size of a field in the BTS record in bytes */
+        unsigned char  sizeof_field;
+        /* a bitmask to enable/disable BTS in DEBUGCTL MSR */
+        unsigned long debugctl_mask;
+};
+static struct bts_configuration bts_cfg;
+
+#define BTS_MAX_RECORD_SIZE (8 * 3)
+
+
+/*
+ * Branch Trace Store (BTS) uses the following format. Different
+ * architectures vary in the size of those fields.
+ * - source linear address
+ * - destination linear address
+ * - flags
+ *
+ * Later architectures use 64bit pointers throughout, whereas earlier
+ * architectures use 32bit pointers in 32bit mode.
+ *
+ * We compute the base address for the first 8 fields based on:
+ * - the field size stored in the DS configuration
+ * - the relative field position
+ *
+ * In order to store additional information in the BTS buffer, we use
+ * a special source address to indicate that the record requires
+ * special interpretation.
+ *
+ * Netburst indicated via a bit in the flags field whether the branch
+ * was predicted; this is ignored.
+ */
+
+enum bts_field {
+        bts_from = 0,
+        bts_to,
+        bts_flags,
+
+        bts_escape = (unsigned long)-1,
+        bts_qual = bts_to,
+        bts_jiffies = bts_flags
+};
+
+static inline unsigned long bts_get(const char *base, enum bts_field field)
+{
+        base += (bts_cfg.sizeof_field * field);
+        return *(unsigned long *)base;
+}
 
-static int ptrace_bts_get_size(struct task_struct *child)
+static inline void bts_set(char *base, enum bts_field field, unsigned long val)
 {
-        if (!child->thread.ds_area_msr)
-                return -ENXIO;
+        base += (bts_cfg.sizeof_field * field);;
+        (*(unsigned long *)base) = val;
+}
 
-        return ds_get_bts_index((void *)child->thread.ds_area_msr);
+/*
+ * Translate a BTS record from the raw format into the bts_struct format
+ *
+ * out (out): bts_struct interpretation
+ * raw: raw BTS record
+ */
+static void ptrace_bts_translate_record(struct bts_struct *out, const void *raw)
+{
+        memset(out, 0, sizeof(*out));
+        if (bts_get(raw, bts_from) == bts_escape) {
+                out->qualifier       = bts_get(raw, bts_qual);
+                out->variant.jiffies = bts_get(raw, bts_jiffies);
+        } else {
+                out->qualifier = BTS_BRANCH;
+                out->variant.lbr.from_ip = bts_get(raw, bts_from);
+                out->variant.lbr.to_ip   = bts_get(raw, bts_to);
+        }
 }
 
-static int ptrace_bts_read_record(struct task_struct *child,
-                                  long index,
+static int ptrace_bts_read_record(struct task_struct *child, size_t index,
                                   struct bts_struct __user *out)
 {
         struct bts_struct ret;
-        int retval;
-        int bts_end;
-        int bts_index;
-
-        if (!child->thread.ds_area_msr)
-                return -ENXIO;
+        const void *bts_record;
+        size_t bts_index, bts_end;
+        int error;
 
-        if (index < 0)
-                return -EINVAL;
+        error = ds_get_bts_end(child, &bts_end);
+        if (error < 0)
+                return error;
 
-        bts_end = ds_get_bts_end((void *)child->thread.ds_area_msr);
         if (bts_end <= index)
                 return -EINVAL;
 
+        error = ds_get_bts_index(child, &bts_index);
+        if (error < 0)
+                return error;
+
         /* translate the ptrace bts index into the ds bts index */
-        bts_index = ds_get_bts_index((void *)child->thread.ds_area_msr);
-        bts_index -= (index + 1);
-        if (bts_index < 0)
-                bts_index += bts_end;
+        bts_index += bts_end - (index + 1);
+        if (bts_end <= bts_index)
+                bts_index -= bts_end;
 
-        retval = ds_read_bts((void *)child->thread.ds_area_msr,
-                             bts_index, &ret);
-        if (retval < 0)
-                return retval;
+        error = ds_access_bts(child, bts_index, &bts_record);
+        if (error < 0)
+                return error;
+
+        ptrace_bts_translate_record(&ret, bts_record);
 
         if (copy_to_user(out, &ret, sizeof(ret)))
                 return -EFAULT;
@@ -600,101 +671,106 @@ static int ptrace_bts_read_record(struct task_struct *child,
         return sizeof(ret);
 }
 
-static int ptrace_bts_clear(struct task_struct *child)
-{
-        if (!child->thread.ds_area_msr)
-                return -ENXIO;
-
-        return ds_clear((void *)child->thread.ds_area_msr);
-}
-
 static int ptrace_bts_drain(struct task_struct *child,
                             long size,
                             struct bts_struct __user *out)
 {
-        int end, i;
-        void *ds = (void *)child->thread.ds_area_msr;
-
-        if (!ds)
-                return -ENXIO;
+        struct bts_struct ret;
+        const unsigned char *raw;
+        size_t end, i;
+        int error;
 
-        end = ds_get_bts_index(ds);
-        if (end <= 0)
-                return end;
+        error = ds_get_bts_index(child, &end);
+        if (error < 0)
+                return error;
 
         if (size < (end * sizeof(struct bts_struct)))
                 return -EIO;
 
-        for (i = 0; i < end; i++, out++) {
-                struct bts_struct ret;
-                int retval;
+        error = ds_access_bts(child, 0, (const void **)&raw);
+        if (error < 0)
+                return error;
 
-                retval = ds_read_bts(ds, i, &ret);
-                if (retval < 0)
-                        return retval;
+        for (i = 0; i < end; i++, out++, raw += bts_cfg.sizeof_bts) {
+                ptrace_bts_translate_record(&ret, raw);
 
                 if (copy_to_user(out, &ret, sizeof(ret)))
                         return -EFAULT;
         }
 
-        ds_clear(ds);
+        error = ds_clear_bts(child);
+        if (error < 0)
+                return error;
 
         return end;
 }
 
+static void ptrace_bts_ovfl(struct task_struct *child)
+{
+        send_sig(child->thread.bts_ovfl_signal, child, 0);
+}
+
 static int ptrace_bts_config(struct task_struct *child,
                              long cfg_size,
                              const struct ptrace_bts_config __user *ucfg)
 {
         struct ptrace_bts_config cfg;
-        int bts_size, ret = 0;
-        void *ds;
+        int error = 0;
+
+        error = -EOPNOTSUPP;
+        if (!bts_cfg.sizeof_bts)
+                goto errout;
 
+        error = -EIO;
         if (cfg_size < sizeof(cfg))
-                return -EIO;
+                goto errout;
 
+        error = -EFAULT;
         if (copy_from_user(&cfg, ucfg, sizeof(cfg)))
-                return -EFAULT;
+                goto errout;
 
-        if ((int)cfg.size < 0)
-                return -EINVAL;
+        error = -EINVAL;
+        if ((cfg.flags & PTRACE_BTS_O_SIGNAL) &&
+            !(cfg.flags & PTRACE_BTS_O_ALLOC))
+                goto errout;
 
-        bts_size = 0;
-        ds = (void *)child->thread.ds_area_msr;
-        if (ds) {
-                bts_size = ds_get_bts_size(ds);
-                if (bts_size < 0)
-                        return bts_size;
-        }
-        cfg.size = PAGE_ALIGN(cfg.size);
+        if (cfg.flags & PTRACE_BTS_O_ALLOC) {
+                ds_ovfl_callback_t ovfl = NULL;
+                unsigned int sig = 0;
+
+                /* we ignore the error in case we were not tracing child */
+                (void)ds_release_bts(child);
 
-        if (bts_size != cfg.size) {
-                ret = ptrace_bts_realloc(child, cfg.size,
-                                         cfg.flags & PTRACE_BTS_O_CUT_SIZE);
-                if (ret < 0)
+                if (cfg.flags & PTRACE_BTS_O_SIGNAL) {
+                        if (!cfg.signal)
+                                goto errout;
+
+                        sig  = cfg.signal;
+                        ovfl = ptrace_bts_ovfl;
+                }
+
+                error = ds_request_bts(child, /* base = */ NULL, cfg.size, ovfl);
+                if (error < 0)
                         goto errout;
 
-                ds = (void *)child->thread.ds_area_msr;
+                child->thread.bts_ovfl_signal = sig;
         }
 
-        if (cfg.flags & PTRACE_BTS_O_SIGNAL)
-                ret = ds_set_overflow(ds, DS_O_SIGNAL);
-        else
-                ret = ds_set_overflow(ds, DS_O_WRAP);
-        if (ret < 0)
+        error = -EINVAL;
+        if (!child->thread.ds_ctx && cfg.flags)
                 goto errout;
 
         if (cfg.flags & PTRACE_BTS_O_TRACE)
-                child->thread.debugctlmsr |= ds_debugctl_mask();
+                child->thread.debugctlmsr |= bts_cfg.debugctl_mask;
         else
-                child->thread.debugctlmsr &= ~ds_debugctl_mask();
+                child->thread.debugctlmsr &= ~bts_cfg.debugctl_mask;
 
         if (cfg.flags & PTRACE_BTS_O_SCHED)
                 set_tsk_thread_flag(child, TIF_BTS_TRACE_TS);
         else
                 clear_tsk_thread_flag(child, TIF_BTS_TRACE_TS);
 
-        ret = sizeof(cfg);
+        error = sizeof(cfg);
 
 out:
         if (child->thread.debugctlmsr)
@@ -702,10 +778,10 @@ out:
         else
                 clear_tsk_thread_flag(child, TIF_DEBUGCTLMSR);
 
-        return ret;
+        return error;
 
 errout:
-        child->thread.debugctlmsr &= ~ds_debugctl_mask();
+        child->thread.debugctlmsr &= ~bts_cfg.debugctl_mask;
         clear_tsk_thread_flag(child, TIF_BTS_TRACE_TS);
         goto out;
 }
@@ -714,29 +790,40 @@ static int ptrace_bts_status(struct task_struct *child,
                              long cfg_size,
                              struct ptrace_bts_config __user *ucfg)
 {
-        void *ds = (void *)child->thread.ds_area_msr;
         struct ptrace_bts_config cfg;
+        size_t end;
+        const void *base, *max;
+        int error;
 
         if (cfg_size < sizeof(cfg))
                 return -EIO;
 
-        memset(&cfg, 0, sizeof(cfg));
+        error = ds_get_bts_end(child, &end);
+        if (error < 0)
+                return error;
 
-        if (ds) {
-                cfg.size = ds_get_bts_size(ds);
+        error = ds_access_bts(child, /* index = */ 0, &base);
+        if (error < 0)
+                return error;
 
-                if (ds_get_overflow(ds) == DS_O_SIGNAL)
-                        cfg.flags |= PTRACE_BTS_O_SIGNAL;
+        error = ds_access_bts(child, /* index = */ end, &max);
+        if (error < 0)
+                return error;
 
-                if (test_tsk_thread_flag(child, TIF_DEBUGCTLMSR) &&
-                    child->thread.debugctlmsr & ds_debugctl_mask())
-                        cfg.flags |= PTRACE_BTS_O_TRACE;
+        memset(&cfg, 0, sizeof(cfg));
+        cfg.size = (max - base);
+        cfg.signal = child->thread.bts_ovfl_signal;
+        cfg.bts_size = sizeof(struct bts_struct);
 
-                if (test_tsk_thread_flag(child, TIF_BTS_TRACE_TS))
-                        cfg.flags |= PTRACE_BTS_O_SCHED;
-        }
+        if (cfg.signal)
+                cfg.flags |= PTRACE_BTS_O_SIGNAL;
 
-        cfg.bts_size = sizeof(struct bts_struct);
+        if (test_tsk_thread_flag(child, TIF_DEBUGCTLMSR) &&
+            child->thread.debugctlmsr & bts_cfg.debugctl_mask)
+                cfg.flags |= PTRACE_BTS_O_TRACE;
+
+        if (test_tsk_thread_flag(child, TIF_BTS_TRACE_TS))
+                cfg.flags |= PTRACE_BTS_O_SCHED;
 
         if (copy_to_user(ucfg, &cfg, sizeof(cfg)))
                 return -EFAULT;
@@ -744,89 +831,38 @@ static int ptrace_bts_status(struct task_struct *child,
         return sizeof(cfg);
 }
 
-
 static int ptrace_bts_write_record(struct task_struct *child,
                                    const struct bts_struct *in)
 {
-        int retval;
+        unsigned char bts_record[BTS_MAX_RECORD_SIZE];
 
-        if (!child->thread.ds_area_msr)
-                return -ENXIO;
+        BUG_ON(BTS_MAX_RECORD_SIZE < bts_cfg.sizeof_bts);
 
-        retval = ds_write_bts((void *)child->thread.ds_area_msr, in);
-        if (retval)
-                return retval;
+        memset(bts_record, 0, bts_cfg.sizeof_bts);
+        switch (in->qualifier) {
+        case BTS_INVALID:
+                break;
 
-        return sizeof(*in);
-}
+        case BTS_BRANCH:
+                bts_set(bts_record, bts_from, in->variant.lbr.from_ip);
+                bts_set(bts_record, bts_to,   in->variant.lbr.to_ip);
+                break;
 
-static int ptrace_bts_realloc(struct task_struct *child,
-                              int size, int reduce_size)
-{
-        unsigned long rlim, vm;
-        int ret, old_size;
+        case BTS_TASK_ARRIVES:
+        case BTS_TASK_DEPARTS:
+                bts_set(bts_record, bts_from,    bts_escape);
+                bts_set(bts_record, bts_qual,    in->qualifier);
+                bts_set(bts_record, bts_jiffies, in->variant.jiffies);
+                break;
 
-        if (size < 0)
+        default:
                 return -EINVAL;
-
-        old_size = ds_get_bts_size((void *)child->thread.ds_area_msr);
-        if (old_size < 0)
-                return old_size;
-
-        ret = ds_free((void **)&child->thread.ds_area_msr);
-        if (ret < 0)
-                goto out;
-
-        size >>= PAGE_SHIFT;
-        old_size >>= PAGE_SHIFT;
-
-        current->mm->total_vm  -= old_size;
-        current->mm->locked_vm -= old_size;
-
-        if (size == 0)
-                goto out;
-
-        rlim = current->signal->rlim[RLIMIT_AS].rlim_cur >> PAGE_SHIFT;
-        vm = current->mm->total_vm  + size;
-        if (rlim < vm) {
-                ret = -ENOMEM;
-
-                if (!reduce_size)
-                        goto out;
-
-                size = rlim - current->mm->total_vm;
-                if (size <= 0)
-                        goto out;
-        }
-
-        rlim = current->signal->rlim[RLIMIT_MEMLOCK].rlim_cur >> PAGE_SHIFT;
-        vm = current->mm->locked_vm  + size;
-        if (rlim < vm) {
-                ret = -ENOMEM;
-
-                if (!reduce_size)
-                        goto out;
-
-                size = rlim - current->mm->locked_vm;
-                if (size <= 0)
-                        goto out;
         }
 
-        ret = ds_allocate((void **)&child->thread.ds_area_msr,
-                          size << PAGE_SHIFT);
-        if (ret < 0)
-                goto out;
-
-        current->mm->total_vm  += size;
-        current->mm->locked_vm += size;
-
-out:
-        if (child->thread.ds_area_msr)
-                set_tsk_thread_flag(child, TIF_DS_AREA_MSR);
-        else
-                clear_tsk_thread_flag(child, TIF_DS_AREA_MSR);
-
-        return ret;
+        /* The writing task will be the switched-to task on a context
+         * switch. It needs to write into the switched-from task's BTS
+         * buffer. */
+        return ds_unchecked_write_bts(child, bts_record, bts_cfg.sizeof_bts);
 }
 
 void ptrace_bts_take_timestamp(struct task_struct *tsk,
@@ -839,7 +875,66 @@ void ptrace_bts_take_timestamp(struct task_struct *tsk,
 
         ptrace_bts_write_record(tsk, &rec);
 }
-#endif /* X86_BTS */
+
+static const struct bts_configuration bts_cfg_netburst = {
+        .sizeof_bts    = sizeof(long) * 3,
+        .sizeof_field  = sizeof(long),
+        .debugctl_mask = (1<<2)|(1<<3)|(1<<5)
+};
+
+static const struct bts_configuration bts_cfg_pentium_m = {
+        .sizeof_bts    = sizeof(long) * 3,
+        .sizeof_field  = sizeof(long),
+        .debugctl_mask = (1<<6)|(1<<7)
+};
+
+static const struct bts_configuration bts_cfg_core2 = {
+        .sizeof_bts    = 8 * 3,
+        .sizeof_field  = 8,
+        .debugctl_mask = (1<<6)|(1<<7)|(1<<9)
+};
+
+static inline void bts_configure(const struct bts_configuration *cfg)
+{
+        bts_cfg = *cfg;
+}
+
+void __cpuinit ptrace_bts_init_intel(struct cpuinfo_x86 *c)
+{
+        switch (c->x86) {
+        case 0x6:
+                switch (c->x86_model) {
+                case 0xD:
+                case 0xE: /* Pentium M */
+                        bts_configure(&bts_cfg_pentium_m);
+                        break;
+                case 0xF: /* Core2 */
+        case 0x1C: /* Atom */
+                        bts_configure(&bts_cfg_core2);
+                        break;
+                default:
+                        /* sorry, don't know about them */
+                        break;
+                }
+                break;
+        case 0xF:
+                switch (c->x86_model) {
+                case 0x0:
+                case 0x1:
+                case 0x2: /* Netburst */
+                        bts_configure(&bts_cfg_netburst);
+                        break;
+                default:
+                        /* sorry, don't know about them */
+                        break;
+                }
+                break;
+        default:
+                /* sorry, don't know about them */
+                break;
+        }
+}
+#endif /* CONFIG_X86_PTRACE_BTS */
 
 /*
  * Called by kernel/ptrace.c when detaching..
@@ -852,15 +947,15 @@ void ptrace_disable(struct task_struct *child)
 #ifdef TIF_SYSCALL_EMU
         clear_tsk_thread_flag(child, TIF_SYSCALL_EMU);
 #endif
-        if (child->thread.ds_area_msr) {
-#ifdef X86_BTS
-                ptrace_bts_realloc(child, 0, 0);
-#endif
-                child->thread.debugctlmsr &= ~ds_debugctl_mask();
-                if (!child->thread.debugctlmsr)
-                        clear_tsk_thread_flag(child, TIF_DEBUGCTLMSR);
-                clear_tsk_thread_flag(child, TIF_BTS_TRACE_TS);
-        }
+#ifdef CONFIG_X86_PTRACE_BTS
+        (void)ds_release_bts(child);
+
+        child->thread.debugctlmsr &= ~bts_cfg.debugctl_mask;
+        if (!child->thread.debugctlmsr)
+                clear_tsk_thread_flag(child, TIF_DEBUGCTLMSR);
+
+        clear_tsk_thread_flag(child, TIF_BTS_TRACE_TS);
+#endif /* CONFIG_X86_PTRACE_BTS */
 }
 
 #if defined CONFIG_X86_32 || defined CONFIG_IA32_EMULATION
@@ -980,7 +1075,7 @@ long arch_ptrace(struct task_struct *child, long request, long addr, long data)
         /*
          * These bits need more cooking - not enabled yet:
          */
-#ifdef X86_BTS
+#ifdef CONFIG_X86_PTRACE_BTS
         case PTRACE_BTS_CONFIG:
                 ret = ptrace_bts_config
                         (child, data, (struct ptrace_bts_config __user *)addr);
@@ -992,7 +1087,7 @@ long arch_ptrace(struct task_struct *child, long request, long addr, long data)
                 break;
 
         case PTRACE_BTS_SIZE:
-                ret = ptrace_bts_get_size(child);
+                ret = ds_get_bts_index(child, /* pos = */ NULL);
                 break;
 
         case PTRACE_BTS_GET:
@@ -1001,14 +1096,14 @@ long arch_ptrace(struct task_struct *child, long request, long addr, long data)
                 break;
 
         case PTRACE_BTS_CLEAR:
-                ret = ptrace_bts_clear(child);
+                ret = ds_clear_bts(child);
                 break;
 
         case PTRACE_BTS_DRAIN:
                 ret = ptrace_bts_drain
                         (child, data, (struct bts_struct __user *) addr);
                 break;
-#endif
+#endif /* CONFIG_X86_PTRACE_BTS */
 
         default:
                 ret = ptrace_request(child, request, addr, data);
@@ -1375,30 +1470,6 @@ void send_sigtrap(struct task_struct *tsk, struct pt_regs *regs, int error_code)
         force_sig_info(SIGTRAP, &info, tsk);
 }
 
-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",
-               current->comm,
-               regs->ip, regs->sp, regs->ax, regs->orig_ax, __builtin_return_address(0),
-               current_thread_info()->flags, current->ptrace);
-#endif
-
-        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;
-        }
-}
 
 #ifdef CONFIG_X86_32
 # define IS_IA32        1
@@ -1432,8 +1503,9 @@ asmregparm long syscall_trace_enter(struct pt_regs *regs)
         if (unlikely(test_thread_flag(TIF_SYSCALL_EMU)))
                 ret = -1L;
 
-        if (ret || test_thread_flag(TIF_SYSCALL_TRACE))
-                syscall_trace(regs);
+        if ((ret || test_thread_flag(TIF_SYSCALL_TRACE)) &&
+            tracehook_report_syscall_entry(regs))
+                ret = -1L;
 
         if (unlikely(current->audit_context)) {
                 if (IS_IA32)
@@ -1459,7 +1531,7 @@ asmregparm void syscall_trace_leave(struct pt_regs *regs)
                 audit_syscall_exit(AUDITSC_RESULT(regs->ax), regs->ax);
 
         if (test_thread_flag(TIF_SYSCALL_TRACE))
-                syscall_trace(regs);
+                tracehook_report_syscall_exit(regs, 0);
 
         /*
          * If TIF_SYSCALL_EMU is set, we only get here because of
@@ -1475,6 +1547,6 @@ asmregparm void syscall_trace_leave(struct pt_regs *regs)
          * system call instruction.
          */
         if (test_thread_flag(TIF_SINGLESTEP) &&
-            (current->ptrace & PT_PTRACED))
+            tracehook_consider_fatal_signal(current, SIGTRAP, SIG_DFL))
                 send_sigtrap(current, regs, 0);
 }
diff --git a/arch/x86/kernel/reboot.c b/arch/x86/kernel/reboot.c
index 724adfc63cb9..f4c93f1cfc19 100644
--- a/arch/x86/kernel/reboot.c
+++ b/arch/x86/kernel/reboot.c
@@ -29,7 +29,11 @@ EXPORT_SYMBOL(pm_power_off);
 
 static const struct desc_ptr no_idt = {};
 static int reboot_mode;
-enum reboot_type reboot_type = BOOT_KBD;
+/*
+ * Keyboard reset and triple fault may result in INIT, not RESET, which
+ * doesn't work when we're in vmx root mode.  Try ACPI first.
+ */
+enum reboot_type reboot_type = BOOT_ACPI;
 int reboot_force;
 
 #if defined(CONFIG_X86_32) && defined(CONFIG_SMP)
diff --git a/arch/x86/kernel/setup.c b/arch/x86/kernel/setup.c
index 362d4e7f2d38..141efab52400 100644
--- a/arch/x86/kernel/setup.c
+++ b/arch/x86/kernel/setup.c
@@ -223,6 +223,9 @@ unsigned long saved_video_mode;
 #define RAMDISK_LOAD_FLAG               0x4000
 
 static char __initdata command_line[COMMAND_LINE_SIZE];
+#ifdef CONFIG_CMDLINE_BOOL
+static char __initdata builtin_cmdline[COMMAND_LINE_SIZE] = CONFIG_CMDLINE;
+#endif
 
 #if defined(CONFIG_EDD) || defined(CONFIG_EDD_MODULE)
 struct edd edd;
@@ -665,11 +668,28 @@ void __init setup_arch(char **cmdline_p)
         bss_resource.start = virt_to_phys(&__bss_start);
         bss_resource.end = virt_to_phys(&__bss_stop)-1;
 
+#ifdef CONFIG_CMDLINE_BOOL
+#ifdef CONFIG_CMDLINE_OVERRIDE
+        strlcpy(boot_command_line, builtin_cmdline, COMMAND_LINE_SIZE);
+#else
+        if (builtin_cmdline[0]) {
+                /* append boot loader cmdline to builtin */
+                strlcat(builtin_cmdline, " ", COMMAND_LINE_SIZE);
+                strlcat(builtin_cmdline, boot_command_line, COMMAND_LINE_SIZE);
+                strlcpy(boot_command_line, builtin_cmdline, COMMAND_LINE_SIZE);
+        }
+#endif
+#endif
+
         strlcpy(command_line, boot_command_line, COMMAND_LINE_SIZE);
         *cmdline_p = command_line;
 
         parse_early_param();
 
+#ifdef CONFIG_X86_64
+        check_efer();
+#endif
+
 #if defined(CONFIG_VMI) && defined(CONFIG_X86_32)
         /*
          * Must be before kernel pagetables are setup
@@ -738,7 +758,6 @@ void __init setup_arch(char **cmdline_p)
 #else
         num_physpages = max_pfn;
 
-        check_efer();
 
         /* How many end-of-memory variables you have, grandma! */
         /* need this before calling reserve_initrd */
diff --git a/arch/x86/kernel/setup_percpu.c b/arch/x86/kernel/setup_percpu.c
index 76e305e064f9..0e67f72d9316 100644
--- a/arch/x86/kernel/setup_percpu.c
+++ b/arch/x86/kernel/setup_percpu.c
@@ -162,9 +162,16 @@ void __init setup_per_cpu_areas(void)
                         printk(KERN_INFO
                                "cpu %d has no node %d or node-local memory\n",
                                 cpu, node);
+                        if (ptr)
+                                printk(KERN_DEBUG "per cpu data for cpu%d at %016lx\n",
+                                         cpu, __pa(ptr));
                 }
-                else
+                else {
                         ptr = alloc_bootmem_pages_node(NODE_DATA(node), size);
+                        if (ptr)
+                                printk(KERN_DEBUG "per cpu data for cpu%d on node%d at %016lx\n",
+                                         cpu, node, __pa(ptr));
+                }
 #endif
                 per_cpu_offset(cpu) = ptr - __per_cpu_start;
                 memcpy(ptr, __per_cpu_start, __per_cpu_end - __per_cpu_start);
diff --git a/arch/x86/kernel/sigframe.h b/arch/x86/kernel/sigframe.h
index 72bbb519d2dc..8b4956e800ac 100644
--- a/arch/x86/kernel/sigframe.h
+++ b/arch/x86/kernel/sigframe.h
@@ -24,4 +24,9 @@ struct rt_sigframe {
         struct ucontext uc;
         struct siginfo info;
 };
+
+int ia32_setup_rt_frame(int sig, struct k_sigaction *ka, siginfo_t *info,
+                sigset_t *set, struct pt_regs *regs);
+int ia32_setup_frame(int sig, struct k_sigaction *ka,
+                sigset_t *set, struct pt_regs *regs);
 #endif
diff --git a/arch/x86/kernel/signal_32.c b/arch/x86/kernel/signal_32.c
index 6fb5bcdd8933..2a2435d3037d 100644
--- a/arch/x86/kernel/signal_32.c
+++ b/arch/x86/kernel/signal_32.c
@@ -17,6 +17,7 @@
 #include <linux/errno.h>
 #include <linux/sched.h>
 #include <linux/wait.h>
+#include <linux/tracehook.h>
 #include <linux/elf.h>
 #include <linux/smp.h>
 #include <linux/mm.h>
@@ -26,6 +27,7 @@
 #include <asm/uaccess.h>
 #include <asm/i387.h>
 #include <asm/vdso.h>
+#include <asm/syscalls.h>
 
 #include "sigframe.h"
 
@@ -558,8 +560,6 @@ handle_signal(unsigned long sig, siginfo_t *info, struct k_sigaction *ka,
          * handler too.
          */
         regs->flags &= ~X86_EFLAGS_TF;
-        if (test_thread_flag(TIF_SINGLESTEP))
-                ptrace_notify(SIGTRAP);
 
         spin_lock_irq(&current->sighand->siglock);
         sigorsets(&current->blocked, &current->blocked, &ka->sa.sa_mask);
@@ -568,6 +568,9 @@ handle_signal(unsigned long sig, siginfo_t *info, struct k_sigaction *ka,
         recalc_sigpending();
         spin_unlock_irq(&current->sighand->siglock);
 
+        tracehook_signal_handler(sig, info, ka, regs,
+                                 test_thread_flag(TIF_SINGLESTEP));
+
         return 0;
 }
 
@@ -661,5 +664,10 @@ do_notify_resume(struct pt_regs *regs, void *unused, __u32 thread_info_flags)
         if (thread_info_flags & _TIF_SIGPENDING)
                 do_signal(regs);
 
+        if (thread_info_flags & _TIF_NOTIFY_RESUME) {
+                clear_thread_flag(TIF_NOTIFY_RESUME);
+                tracehook_notify_resume(regs);
+        }
+
         clear_thread_flag(TIF_IRET);
 }
diff --git a/arch/x86/kernel/signal_64.c b/arch/x86/kernel/signal_64.c
index ca316b5b742c..694aa888bb19 100644
--- a/arch/x86/kernel/signal_64.c
+++ b/arch/x86/kernel/signal_64.c
@@ -15,17 +15,21 @@
 #include <linux/errno.h>
 #include <linux/wait.h>
 #include <linux/ptrace.h>
+#include <linux/tracehook.h>
 #include <linux/unistd.h>
 #include <linux/stddef.h>
 #include <linux/personality.h>
 #include <linux/compiler.h>
+#include <linux/uaccess.h>
+
 #include <asm/processor.h>
 #include <asm/ucontext.h>
-#include <asm/uaccess.h>
 #include <asm/i387.h>
 #include <asm/proto.h>
 #include <asm/ia32_unistd.h>
 #include <asm/mce.h>
+#include <asm/syscall.h>
+#include <asm/syscalls.h>
 #include "sigframe.h"
 
 #define _BLOCKABLE (~(sigmask(SIGKILL) | sigmask(SIGSTOP)))
@@ -41,11 +45,6 @@
 # define FIX_EFLAGS     __FIX_EFLAGS
 #endif
 
-int ia32_setup_rt_frame(int sig, struct k_sigaction *ka, siginfo_t *info,
-               sigset_t *set, struct pt_regs * regs); 
-int ia32_setup_frame(int sig, struct k_sigaction *ka,
-            sigset_t *set, struct pt_regs * regs); 
-
 asmlinkage long
 sys_sigaltstack(const stack_t __user *uss, stack_t __user *uoss,
                 struct pt_regs *regs)
@@ -128,7 +127,7 @@ restore_sigcontext(struct pt_regs *regs, struct sigcontext __user *sc,
         /* Always make any pending restarted system calls return -EINTR */
         current_thread_info()->restart_block.fn = do_no_restart_syscall;
 
-#define COPY(x)         err |= __get_user(regs->x, &sc->x)
+#define COPY(x)         (err |= __get_user(regs->x, &sc->x))
 
         COPY(di); COPY(si); COPY(bp); COPY(sp); COPY(bx);
         COPY(dx); COPY(cx); COPY(ip);
@@ -158,7 +157,7 @@ restore_sigcontext(struct pt_regs *regs, struct sigcontext __user *sc,
         }
 
         {
-                struct _fpstate __user * buf;
+                struct _fpstate __user *buf;
                 err |= __get_user(buf, &sc->fpstate);
 
                 if (buf) {
@@ -198,7 +197,7 @@ asmlinkage long sys_rt_sigreturn(struct pt_regs *regs)
         current->blocked = set;
         recalc_sigpending();
         spin_unlock_irq(&current->sighand->siglock);
-        
+
         if (restore_sigcontext(regs, &frame->uc.uc_mcontext, &ax))
                 goto badframe;
 
@@ -208,16 +207,17 @@ asmlinkage long sys_rt_sigreturn(struct pt_regs *regs)
         return ax;
 
 badframe:
-        signal_fault(regs,frame,"sigreturn");
+        signal_fault(regs, frame, "sigreturn");
         return 0;
-}       
+}
 
 /*
  * Set up a signal frame.
  */
 
 static inline int
-setup_sigcontext(struct sigcontext __user *sc, struct pt_regs *regs, unsigned long mask, struct task_struct *me)
+setup_sigcontext(struct sigcontext __user *sc, struct pt_regs *regs,
+                unsigned long mask, struct task_struct *me)
 {
         int err = 0;
 
@@ -273,35 +273,35 @@ get_stack(struct k_sigaction *ka, struct pt_regs *regs, unsigned long size)
 }
 
 static int setup_rt_frame(int sig, struct k_sigaction *ka, siginfo_t *info,
-                           sigset_t *set, struct pt_regs * regs)
+                           sigset_t *set, struct pt_regs *regs)
 {
         struct rt_sigframe __user *frame;
-        struct _fpstate __user *fp = NULL; 
+        struct _fpstate __user *fp = NULL;
         int err = 0;
         struct task_struct *me = current;
 
         if (used_math()) {
-                fp = get_stack(ka, regs, sizeof(struct _fpstate)); 
+                fp = get_stack(ka, regs, sizeof(struct _fpstate));
                 frame = (void __user *)round_down(
                         (unsigned long)fp - sizeof(struct rt_sigframe), 16) - 8;
 
                 if (!access_ok(VERIFY_WRITE, fp, sizeof(struct _fpstate)))
                         goto give_sigsegv;
 
-                if (save_i387(fp) < 0) 
-                        err |= -1; 
+                if (save_i387(fp) < 0)
+                        err |= -1;
         } else
                 frame = get_stack(ka, regs, sizeof(struct rt_sigframe)) - 8;
 
         if (!access_ok(VERIFY_WRITE, frame, sizeof(*frame)))
                 goto give_sigsegv;
 
-        if (ka->sa.sa_flags & SA_SIGINFO) { 
+        if (ka->sa.sa_flags & SA_SIGINFO) {
                 err |= copy_siginfo_to_user(&frame->info, info);
                 if (err)
                         goto give_sigsegv;
         }
-                
+
         /* Create the ucontext.  */
         err |= __put_user(0, &frame->uc.uc_flags);
         err |= __put_user(0, &frame->uc.uc_link);
@@ -311,9 +311,9 @@ static int setup_rt_frame(int sig, struct k_sigaction *ka, siginfo_t *info,
         err |= __put_user(me->sas_ss_size, &frame->uc.uc_stack.ss_size);
         err |= setup_sigcontext(&frame->uc.uc_mcontext, regs, set->sig[0], me);
         err |= __put_user(fp, &frame->uc.uc_mcontext.fpstate);
-        if (sizeof(*set) == 16) { 
+        if (sizeof(*set) == 16) {
                 __put_user(set->sig[0], &frame->uc.uc_sigmask.sig[0]);
-                __put_user(set->sig[1], &frame->uc.uc_sigmask.sig[1]); 
+                __put_user(set->sig[1], &frame->uc.uc_sigmask.sig[1]);
         } else
                 err |= __copy_to_user(&frame->uc.uc_sigmask, set, sizeof(*set));
 
@@ -324,7 +324,7 @@ static int setup_rt_frame(int sig, struct k_sigaction *ka, siginfo_t *info,
                 err |= __put_user(ka->sa.sa_restorer, &frame->pretcode);
         } else {
                 /* could use a vstub here */
-                goto give_sigsegv; 
+                goto give_sigsegv;
         }
 
         if (err)
@@ -332,7 +332,7 @@ static int setup_rt_frame(int sig, struct k_sigaction *ka, siginfo_t *info,
 
         /* Set up registers for signal handler */
         regs->di = sig;
-        /* In case the signal handler was declared without prototypes */ 
+        /* In case the signal handler was declared without prototypes */
         regs->ax = 0;
 
         /* This also works for non SA_SIGINFO handlers because they expect the
@@ -355,37 +355,8 @@ give_sigsegv:
 }
 
 /*
- * Return -1L or the syscall number that @regs is executing.
- */
-static long current_syscall(struct pt_regs *regs)
-{
-        /*
-         * We always sign-extend a -1 value being set here,
-         * so this is always either -1L or a syscall number.
-         */
-        return regs->orig_ax;
-}
-
-/*
- * Return a value that is -EFOO if the system call in @regs->orig_ax
- * returned an error.  This only works for @regs from @current.
- */
-static long current_syscall_ret(struct pt_regs *regs)
-{
-#ifdef CONFIG_IA32_EMULATION
-        if (test_thread_flag(TIF_IA32))
-                /*
-                 * Sign-extend the value so (int)-EFOO becomes (long)-EFOO
-                 * and will match correctly in comparisons.
-                 */
-                return (int) regs->ax;
-#endif
-        return regs->ax;
-}
-
-/*
  * OK, we're invoking a handler
- */     
+ */
 
 static int
 handle_signal(unsigned long sig, siginfo_t *info, struct k_sigaction *ka,
@@ -394,9 +365,9 @@ handle_signal(unsigned long sig, siginfo_t *info, struct k_sigaction *ka,
         int ret;
 
         /* Are we from a system call? */
-        if (current_syscall(regs) >= 0) {
+        if (syscall_get_nr(current, regs) >= 0) {
                 /* If so, check system call restarting.. */
-                switch (current_syscall_ret(regs)) {
+                switch (syscall_get_error(current, regs)) {
                 case -ERESTART_RESTARTBLOCK:
                 case -ERESTARTNOHAND:
                         regs->ax = -EINTR;
@@ -429,7 +400,7 @@ handle_signal(unsigned long sig, siginfo_t *info, struct k_sigaction *ka,
                         ret = ia32_setup_rt_frame(sig, ka, info, oldset, regs);
                 else
                         ret = ia32_setup_frame(sig, ka, oldset, regs);
-        } else 
+        } else
 #endif
         ret = setup_rt_frame(sig, ka, info, oldset, regs);
 
@@ -453,15 +424,16 @@ handle_signal(unsigned long sig, siginfo_t *info, struct k_sigaction *ka,
                  * handler too.
                  */
                 regs->flags &= ~X86_EFLAGS_TF;
-                if (test_thread_flag(TIF_SINGLESTEP))
-                        ptrace_notify(SIGTRAP);
 
                 spin_lock_irq(&current->sighand->siglock);
-                sigorsets(&current->blocked,&current->blocked,&ka->sa.sa_mask);
+                sigorsets(&current->blocked, &current->blocked, &ka->sa.sa_mask);
                 if (!(ka->sa.sa_flags & SA_NODEFER))
-                        sigaddset(&current->blocked,sig);
+                        sigaddset(&current->blocked, sig);
                 recalc_sigpending();
                 spin_unlock_irq(&current->sighand->siglock);
+
+                tracehook_signal_handler(sig, info, ka, regs,
+                                         test_thread_flag(TIF_SINGLESTEP));
         }
 
         return ret;
@@ -518,9 +490,9 @@ static void do_signal(struct pt_regs *regs)
         }
 
         /* Did we come from a system call? */
-        if (current_syscall(regs) >= 0) {
+        if (syscall_get_nr(current, regs) >= 0) {
                 /* Restart the system call - no handlers present */
-                switch (current_syscall_ret(regs)) {
+                switch (syscall_get_error(current, regs)) {
                 case -ERESTARTNOHAND:
                 case -ERESTARTSYS:
                 case -ERESTARTNOINTR:
@@ -558,17 +530,23 @@ void do_notify_resume(struct pt_regs *regs, void *unused,
         /* deal with pending signal delivery */
         if (thread_info_flags & _TIF_SIGPENDING)
                 do_signal(regs);
+
+        if (thread_info_flags & _TIF_NOTIFY_RESUME) {
+                clear_thread_flag(TIF_NOTIFY_RESUME);
+                tracehook_notify_resume(regs);
+        }
 }
 
 void signal_fault(struct pt_regs *regs, void __user *frame, char *where)
-{ 
-        struct task_struct *me = current; 
+{
+        struct task_struct *me = current;
         if (show_unhandled_signals && printk_ratelimit()) {
                 printk("%s[%d] bad frame in %s frame:%p ip:%lx sp:%lx orax:%lx",
-               me->comm,me->pid,where,frame,regs->ip,regs->sp,regs->orig_ax);
+               me->comm, me->pid, where, frame, regs->ip,
+                   regs->sp, regs->orig_ax);
                 print_vma_addr(" in ", regs->ip);
                 printk("\n");
         }
 
-        force_sig(SIGSEGV, me); 
-} 
+        force_sig(SIGSEGV, me);
+}
diff --git a/arch/x86/kernel/smpboot.c b/arch/x86/kernel/smpboot.c
index 7985c5b3f916..45531e3ba194 100644
--- a/arch/x86/kernel/smpboot.c
+++ b/arch/x86/kernel/smpboot.c
@@ -88,7 +88,7 @@ static DEFINE_PER_CPU(struct task_struct *, idle_thread_array);
 #define get_idle_for_cpu(x)      (per_cpu(idle_thread_array, x))
 #define set_idle_for_cpu(x, p)   (per_cpu(idle_thread_array, x) = (p))
 #else
-struct task_struct *idle_thread_array[NR_CPUS] __cpuinitdata ;
+static struct task_struct *idle_thread_array[NR_CPUS] __cpuinitdata ;
 #define get_idle_for_cpu(x)      (idle_thread_array[(x)])
 #define set_idle_for_cpu(x, p)   (idle_thread_array[(x)] = (p))
 #endif
@@ -129,7 +129,7 @@ static int boot_cpu_logical_apicid;
 static cpumask_t cpu_sibling_setup_map;
 
 /* Set if we find a B stepping CPU */
-int __cpuinitdata smp_b_stepping;
+static int __cpuinitdata smp_b_stepping;
 
 #if defined(CONFIG_NUMA) && defined(CONFIG_X86_32)
 
@@ -1313,16 +1313,13 @@ __init void prefill_possible_map(void)
         if (!num_processors)
                 num_processors = 1;
 
-#ifdef CONFIG_HOTPLUG_CPU
         if (additional_cpus == -1) {
                 if (disabled_cpus > 0)
                         additional_cpus = disabled_cpus;
                 else
                         additional_cpus = 0;
         }
-#else
-        additional_cpus = 0;
-#endif
+
         possible = num_processors + additional_cpus;
         if (possible > NR_CPUS)
                 possible = NR_CPUS;
diff --git a/arch/x86/kernel/sys_i386_32.c b/arch/x86/kernel/sys_i386_32.c
index 7066cb855a60..1884a8d12bfa 100644
--- a/arch/x86/kernel/sys_i386_32.c
+++ b/arch/x86/kernel/sys_i386_32.c
@@ -22,6 +22,8 @@
 #include <linux/uaccess.h>
 #include <linux/unistd.h>
 
+#include <asm/syscalls.h>
+
 asmlinkage long sys_mmap2(unsigned long addr, unsigned long len,
                           unsigned long prot, unsigned long flags,
                           unsigned long fd, unsigned long pgoff)
diff --git a/arch/x86/kernel/sys_x86_64.c b/arch/x86/kernel/sys_x86_64.c
index 3b360ef33817..6bc211accf08 100644
--- a/arch/x86/kernel/sys_x86_64.c
+++ b/arch/x86/kernel/sys_x86_64.c
@@ -13,15 +13,17 @@
 #include <linux/utsname.h>
 #include <linux/personality.h>
 #include <linux/random.h>
+#include <linux/uaccess.h>
 
-#include <asm/uaccess.h>
 #include <asm/ia32.h>
+#include <asm/syscalls.h>
 
-asmlinkage long sys_mmap(unsigned long addr, unsigned long len, unsigned long prot, unsigned long flags,
-        unsigned long fd, unsigned long off)
+asmlinkage long sys_mmap(unsigned long addr, unsigned long len,
+                unsigned long prot, unsigned long flags,
+                unsigned long fd, unsigned long off)
 {
         long error;
-        struct file * file;
+        struct file *file;
 
         error = -EINVAL;
         if (off & ~PAGE_MASK)
@@ -56,9 +58,9 @@ static void find_start_end(unsigned long flags, unsigned long *begin,
                    unmapped base down for this case. This can give
                    conflicts with the heap, but we assume that glibc
                    malloc knows how to fall back to mmap. Give it 1GB
-                   of playground for now. -AK */ 
-                *begin = 0x40000000; 
-                *end = 0x80000000;              
+                   of playground for now. -AK */
+                *begin = 0x40000000;
+                *end = 0x80000000;
                 if (current->flags & PF_RANDOMIZE) {
                         new_begin = randomize_range(*begin, *begin + 0x02000000, 0);
                         if (new_begin)
@@ -66,9 +68,9 @@ static void find_start_end(unsigned long flags, unsigned long *begin,
                 }
         } else {
                 *begin = TASK_UNMAPPED_BASE;
-                *end = TASK_SIZE; 
+                *end = TASK_SIZE;
         }
-} 
+}
 
 unsigned long
 arch_get_unmapped_area(struct file *filp, unsigned long addr,
@@ -78,11 +80,11 @@ arch_get_unmapped_area(struct file *filp, unsigned long addr,
         struct vm_area_struct *vma;
         unsigned long start_addr;
         unsigned long begin, end;
-        
+
         if (flags & MAP_FIXED)
                 return addr;
 
-        find_start_end(flags, &begin, &end); 
+        find_start_end(flags, &begin, &end);
 
         if (len > end)
                 return -ENOMEM;
@@ -96,12 +98,12 @@ arch_get_unmapped_area(struct file *filp, unsigned long addr,
         }
         if (((flags & MAP_32BIT) || test_thread_flag(TIF_IA32))
             && len <= mm->cached_hole_size) {
-                mm->cached_hole_size = 0;
+                mm->cached_hole_size = 0;
                 mm->free_area_cache = begin;
         }
         addr = mm->free_area_cache;
-        if (addr < begin) 
-                addr = begin; 
+        if (addr < begin)
+                addr = begin;
         start_addr = addr;
 
 full_search:
@@ -127,7 +129,7 @@ full_search:
                         return addr;
                 }
                 if (addr + mm->cached_hole_size < vma->vm_start)
-                        mm->cached_hole_size = vma->vm_start - addr;
+                        mm->cached_hole_size = vma->vm_start - addr;
 
                 addr = vma->vm_end;
         }
@@ -177,7 +179,7 @@ arch_get_unmapped_area_topdown(struct file *filp, const unsigned long addr0,
                 vma = find_vma(mm, addr-len);
                 if (!vma || addr <= vma->vm_start)
                         /* remember the address as a hint for next time */
-                        return (mm->free_area_cache = addr-len);
+                        return mm->free_area_cache = addr-len;
         }
 
         if (mm->mmap_base < len)
@@ -194,7 +196,7 @@ arch_get_unmapped_area_topdown(struct file *filp, const unsigned long addr0,
                 vma = find_vma(mm, addr);
                 if (!vma || addr+len <= vma->vm_start)
                         /* remember the address as a hint for next time */
-                        return (mm->free_area_cache = addr);
+                        return mm->free_area_cache = addr;
 
                 /* remember the largest hole we saw so far */
                 if (addr + mm->cached_hole_size < vma->vm_start)
@@ -224,13 +226,13 @@ bottomup:
 }
 
 
-asmlinkage long sys_uname(struct new_utsname __user * name)
+asmlinkage long sys_uname(struct new_utsname __user *name)
 {
         int err;
         down_read(&uts_sem);
-        err = copy_to_user(name, utsname(), sizeof (*name));
+        err = copy_to_user(name, utsname(), sizeof(*name));
         up_read(&uts_sem);
-        if (personality(current->personality) == PER_LINUX32) 
-                err |= copy_to_user(&name->machine, "i686", 5);                 
+        if (personality(current->personality) == PER_LINUX32)
+                err |= copy_to_user(&name->machine, "i686", 5);
         return err ? -EFAULT : 0;
 }
diff --git a/arch/x86/kernel/syscall_64.c b/arch/x86/kernel/syscall_64.c
index 170d43c17487..3d1be4f0fac5 100644
--- a/arch/x86/kernel/syscall_64.c
+++ b/arch/x86/kernel/syscall_64.c
@@ -8,12 +8,12 @@
 #define __NO_STUBS
 
 #define __SYSCALL(nr, sym) extern asmlinkage void sym(void) ;
-#undef _ASM_X86_64_UNISTD_H_
+#undef ASM_X86__UNISTD_64_H
 #include <asm/unistd_64.h>
 
 #undef __SYSCALL
 #define __SYSCALL(nr, sym) [nr] = sym,
-#undef _ASM_X86_64_UNISTD_H_
+#undef ASM_X86__UNISTD_64_H
 
 typedef void (*sys_call_ptr_t)(void);
 
diff --git a/arch/x86/kernel/time_32.c b/arch/x86/kernel/time_32.c
index ffe3c664afc0..bbecf8b6bf96 100644
--- a/arch/x86/kernel/time_32.c
+++ b/arch/x86/kernel/time_32.c
@@ -36,6 +36,7 @@
 #include <asm/arch_hooks.h>
 #include <asm/hpet.h>
 #include <asm/time.h>
+#include <asm/timer.h>
 
 #include "do_timer.h"
 
diff --git a/arch/x86/kernel/tls.c b/arch/x86/kernel/tls.c
index ab6bf375a307..6bb7b8579e70 100644
--- a/arch/x86/kernel/tls.c
+++ b/arch/x86/kernel/tls.c
@@ -10,6 +10,7 @@
 #include <asm/ldt.h>
 #include <asm/processor.h>
 #include <asm/proto.h>
+#include <asm/syscalls.h>
 
 #include "tls.h"
 
diff --git a/arch/x86/kernel/traps_64.c b/arch/x86/kernel/traps_64.c
index 513caaca7115..7a31f104bef9 100644
--- a/arch/x86/kernel/traps_64.c
+++ b/arch/x86/kernel/traps_64.c
@@ -32,6 +32,8 @@
 #include <linux/bug.h>
 #include <linux/nmi.h>
 #include <linux/mm.h>
+#include <linux/smp.h>
+#include <linux/io.h>
 
 #if defined(CONFIG_EDAC)
 #include <linux/edac.h>
@@ -45,9 +47,6 @@
 #include <asm/unwind.h>
 #include <asm/desc.h>
 #include <asm/i387.h>
-#include <asm/nmi.h>
-#include <asm/smp.h>
-#include <asm/io.h>
 #include <asm/pgalloc.h>
 #include <asm/proto.h>
 #include <asm/pda.h>
@@ -85,7 +84,8 @@ static inline void preempt_conditional_cli(struct pt_regs *regs)
 
 void printk_address(unsigned long address, int reliable)
 {
-        printk(" [<%016lx>] %s%pS\n", address, reliable ? "": "? ", (void *) address);
+        printk(" [<%016lx>] %s%pS\n",
+                        address, reliable ?     "" : "? ", (void *) address);
 }
 
 static unsigned long *in_exception_stack(unsigned cpu, unsigned long stack,
@@ -98,7 +98,8 @@ static unsigned long *in_exception_stack(unsigned cpu, unsigned long stack,
                 [STACKFAULT_STACK - 1] = "#SS",
                 [MCE_STACK - 1] = "#MC",
 #if DEBUG_STKSZ > EXCEPTION_STKSZ
-                [N_EXCEPTION_STACKS ... N_EXCEPTION_STACKS + DEBUG_STKSZ / EXCEPTION_STKSZ - 2] = "#DB[?]"
+                [N_EXCEPTION_STACKS ...
+                        N_EXCEPTION_STACKS + DEBUG_STKSZ / EXCEPTION_STKSZ - 2] = "#DB[?]"
 #endif
         };
         unsigned k;
@@ -163,7 +164,7 @@ static unsigned long *in_exception_stack(unsigned cpu, unsigned long stack,
 }
 
 /*
- * x86-64 can have up to three kernel stacks: 
+ * x86-64 can have up to three kernel stacks:
  * process stack
  * interrupt stack
  * severe exception (double fault, nmi, stack fault, debug, mce) hardware stack
@@ -219,7 +220,7 @@ void dump_trace(struct task_struct *task, struct pt_regs *regs,
                 const struct stacktrace_ops *ops, void *data)
 {
         const unsigned cpu = get_cpu();
-        unsigned long *irqstack_end = (unsigned long*)cpu_pda(cpu)->irqstackptr;
+        unsigned long *irqstack_end = (unsigned long *)cpu_pda(cpu)->irqstackptr;
         unsigned used = 0;
         struct thread_info *tinfo;
 
@@ -237,7 +238,7 @@ void dump_trace(struct task_struct *task, struct pt_regs *regs,
         if (!bp) {
                 if (task == current) {
                         /* Grab bp right from our regs */
-                        asm("movq %%rbp, %0" : "=r" (bp) :);
+                        asm("movq %%rbp, %0" : "=r" (bp) : );
                 } else {
                         /* bp is the last reg pushed by switch_to */
                         bp = *(unsigned long *) task->thread.sp;
@@ -339,9 +340,8 @@ static void
 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");
+        printk("Call Trace:\n");
         dump_trace(task, regs, stack, bp, &print_trace_ops, log_lvl);
-        printk("\n");
 }
 
 void show_trace(struct task_struct *task, struct pt_regs *regs,
@@ -357,11 +357,15 @@ show_stack_log_lvl(struct task_struct *task, struct pt_regs *regs,
         unsigned long *stack;
         int i;
         const int cpu = smp_processor_id();
-        unsigned long *irqstack_end = (unsigned long *) (cpu_pda(cpu)->irqstackptr);
-        unsigned long *irqstack = (unsigned long *) (cpu_pda(cpu)->irqstackptr - IRQSTACKSIZE);
+        unsigned long *irqstack_end =
+                (unsigned long *) (cpu_pda(cpu)->irqstackptr);
+        unsigned long *irqstack =
+                (unsigned long *) (cpu_pda(cpu)->irqstackptr - IRQSTACKSIZE);
 
-        // debugging aid: "show_stack(NULL, NULL);" prints the
-        // back trace for this cpu.
+        /*
+         * debugging aid: "show_stack(NULL, NULL);" prints the
+         * back trace for this cpu.
+         */
 
         if (sp == NULL) {
                 if (task)
@@ -386,6 +390,7 @@ show_stack_log_lvl(struct task_struct *task, struct pt_regs *regs,
                 printk(" %016lx", *stack++);
                 touch_nmi_watchdog();
         }
+        printk("\n");
         show_trace_log_lvl(task, regs, sp, bp, log_lvl);
 }
 
@@ -404,7 +409,7 @@ void dump_stack(void)
 
 #ifdef CONFIG_FRAME_POINTER
         if (!bp)
-                asm("movq %%rbp, %0" : "=r" (bp):);
+                asm("movq %%rbp, %0" : "=r" (bp) : );
 #endif
 
         printk("Pid: %d, comm: %.20s %s %s %.*s\n",
@@ -414,7 +419,6 @@ void dump_stack(void)
                 init_utsname()->version);
         show_trace(NULL, NULL, &stack, bp);
 }
-
 EXPORT_SYMBOL(dump_stack);
 
 void show_registers(struct pt_regs *regs)
@@ -443,7 +447,6 @@ void show_registers(struct pt_regs *regs)
                 printk("Stack: ");
                 show_stack_log_lvl(NULL, regs, (unsigned long *)sp,
                                 regs->bp, "");
-                printk("\n");
 
                 printk(KERN_EMERG "Code: ");
 
@@ -493,7 +496,7 @@ unsigned __kprobes long oops_begin(void)
         raw_local_irq_save(flags);
         cpu = smp_processor_id();
         if (!__raw_spin_trylock(&die_lock)) {
-                if (cpu == die_owner) 
+                if (cpu == die_owner)
                         /* nested oops. should stop eventually */;
                 else
                         __raw_spin_lock(&die_lock);
@@ -638,7 +641,7 @@ kernel_trap:
 }
 
 #define DO_ERROR(trapnr, signr, str, name) \
-asmlinkage void do_##name(struct pt_regs * regs, long error_code)       \
+asmlinkage void do_##name(struct pt_regs *regs, long error_code)        \
 {                                                                       \
         if (notify_die(DIE_TRAP, str, regs, error_code, trapnr, signr)  \
                                                         == NOTIFY_STOP) \
@@ -648,7 +651,7 @@ asmlinkage void do_##name(struct pt_regs * regs, long error_code)	\
 }
 
 #define DO_ERROR_INFO(trapnr, signr, str, name, sicode, siaddr)         \
-asmlinkage void do_##name(struct pt_regs * regs, long error_code)       \
+asmlinkage void do_##name(struct pt_regs *regs, long error_code)        \
 {                                                                       \
         siginfo_t info;                                                 \
         info.si_signo = signr;                                          \
@@ -683,7 +686,7 @@ asmlinkage void do_stack_segment(struct pt_regs *regs, long error_code)
         preempt_conditional_cli(regs);
 }
 
-asmlinkage void do_double_fault(struct pt_regs * regs, long error_code)
+asmlinkage void do_double_fault(struct pt_regs *regs, long error_code)
 {
         static const char str[] = "double fault";
         struct task_struct *tsk = current;
@@ -778,9 +781,10 @@ io_check_error(unsigned char reason, struct pt_regs *regs)
 }
 
 static notrace __kprobes void
-unknown_nmi_error(unsigned char reason, struct pt_regs * regs)
+unknown_nmi_error(unsigned char reason, struct pt_regs *regs)
 {
-        if (notify_die(DIE_NMIUNKNOWN, "nmi", regs, reason, 2, SIGINT) == NOTIFY_STOP)
+        if (notify_die(DIE_NMIUNKNOWN, "nmi", regs, reason, 2, SIGINT) ==
+                        NOTIFY_STOP)
                 return;
         printk(KERN_EMERG "Uhhuh. NMI received for unknown reason %02x.\n",
                 reason);
@@ -882,7 +886,7 @@ asmlinkage __kprobes struct pt_regs *sync_regs(struct pt_regs *eregs)
         else if (user_mode(eregs))
                 regs = task_pt_regs(current);
         /* Exception from kernel and interrupts are enabled. Move to
-           kernel process stack. */
+           kernel process stack. */
         else if (eregs->flags & X86_EFLAGS_IF)
                 regs = (struct pt_regs *)(eregs->sp -= sizeof(struct pt_regs));
         if (eregs != regs)
@@ -891,7 +895,7 @@ asmlinkage __kprobes struct pt_regs *sync_regs(struct pt_regs *eregs)
 }
 
 /* runs on IST stack. */
-asmlinkage void __kprobes do_debug(struct pt_regs * regs,
+asmlinkage void __kprobes do_debug(struct pt_regs *regs,
                                    unsigned long error_code)
 {
         struct task_struct *tsk = current;
@@ -1035,7 +1039,7 @@ asmlinkage void do_coprocessor_error(struct pt_regs *regs)
 
 asmlinkage void bad_intr(void)
 {
-        printk("bad interrupt"); 
+        printk("bad interrupt");
 }
 
 asmlinkage void do_simd_coprocessor_error(struct pt_regs *regs)
@@ -1047,7 +1051,7 @@ asmlinkage void do_simd_coprocessor_error(struct pt_regs *regs)
 
         conditional_sti(regs);
         if (!user_mode(regs) &&
-                kernel_math_error(regs, "kernel simd math error", 19))
+                        kernel_math_error(regs, "kernel simd math error", 19))
                 return;
 
         /*
@@ -1092,7 +1096,7 @@ asmlinkage void do_simd_coprocessor_error(struct pt_regs *regs)
         force_sig_info(SIGFPE, &info, task);
 }
 
-asmlinkage void do_spurious_interrupt_bug(struct pt_regs * regs)
+asmlinkage void do_spurious_interrupt_bug(struct pt_regs *regs)
 {
 }
 
@@ -1149,8 +1153,10 @@ void __init trap_init(void)
         set_intr_gate(0, &divide_error);
         set_intr_gate_ist(1, &debug, DEBUG_STACK);
         set_intr_gate_ist(2, &nmi, NMI_STACK);
-        set_system_gate_ist(3, &int3, DEBUG_STACK); /* int3 can be called from all */
-        set_system_gate(4, &overflow); /* int4 can be called from all */
+        /* int3 can be called from all */
+        set_system_gate_ist(3, &int3, DEBUG_STACK);
+        /* int4 can be called from all */
+        set_system_gate(4, &overflow);
         set_intr_gate(5, &bounds);
         set_intr_gate(6, &invalid_op);
         set_intr_gate(7, &device_not_available);
diff --git a/arch/x86/kernel/tsc.c b/arch/x86/kernel/tsc.c
index 8f98e9de1b82..161bb850fc47 100644
--- a/arch/x86/kernel/tsc.c
+++ b/arch/x86/kernel/tsc.c
@@ -104,7 +104,7 @@ __setup("notsc", notsc_setup);
 /*
  * Read TSC and the reference counters. Take care of SMI disturbance
  */
-static u64 tsc_read_refs(u64 *pm, u64 *hpet)
+static u64 tsc_read_refs(u64 *p, int hpet)
 {
         u64 t1, t2;
         int i;
@@ -112,9 +112,9 @@ static u64 tsc_read_refs(u64 *pm, u64 *hpet)
         for (i = 0; i < MAX_RETRIES; i++) {
                 t1 = get_cycles();
                 if (hpet)
-                        *hpet = hpet_readl(HPET_COUNTER) & 0xFFFFFFFF;
+                        *p = hpet_readl(HPET_COUNTER) & 0xFFFFFFFF;
                 else
-                        *pm = acpi_pm_read_early();
+                        *p = acpi_pm_read_early();
                 t2 = get_cycles();
                 if ((t2 - t1) < SMI_TRESHOLD)
                         return t2;
@@ -123,13 +123,59 @@ static u64 tsc_read_refs(u64 *pm, u64 *hpet)
 }
 
 /*
+ * Calculate the TSC frequency from HPET reference
+ */
+static unsigned long calc_hpet_ref(u64 deltatsc, u64 hpet1, u64 hpet2)
+{
+        u64 tmp;
+
+        if (hpet2 < hpet1)
+                hpet2 += 0x100000000ULL;
+        hpet2 -= hpet1;
+        tmp = ((u64)hpet2 * hpet_readl(HPET_PERIOD));
+        do_div(tmp, 1000000);
+        do_div(deltatsc, tmp);
+
+        return (unsigned long) deltatsc;
+}
+
+/*
+ * Calculate the TSC frequency from PMTimer reference
+ */
+static unsigned long calc_pmtimer_ref(u64 deltatsc, u64 pm1, u64 pm2)
+{
+        u64 tmp;
+
+        if (!pm1 && !pm2)
+                return ULONG_MAX;
+
+        if (pm2 < pm1)
+                pm2 += (u64)ACPI_PM_OVRRUN;
+        pm2 -= pm1;
+        tmp = pm2 * 1000000000LL;
+        do_div(tmp, PMTMR_TICKS_PER_SEC);
+        do_div(deltatsc, tmp);
+
+        return (unsigned long) deltatsc;
+}
+
+#define CAL_MS          10
+#define CAL_LATCH       (CLOCK_TICK_RATE / (1000 / CAL_MS))
+#define CAL_PIT_LOOPS   1000
+
+#define CAL2_MS         50
+#define CAL2_LATCH      (CLOCK_TICK_RATE / (1000 / CAL2_MS))
+#define CAL2_PIT_LOOPS  5000
+
+
+/*
  * Try to calibrate the TSC against the Programmable
  * Interrupt Timer and return the frequency of the TSC
  * in kHz.
  *
  * Return ULONG_MAX on failure to calibrate.
  */
-static unsigned long pit_calibrate_tsc(void)
+static unsigned long pit_calibrate_tsc(u32 latch, unsigned long ms, int loopmin)
 {
         u64 tsc, t1, t2, delta;
         unsigned long tscmin, tscmax;
@@ -144,8 +190,8 @@ static unsigned long pit_calibrate_tsc(void)
          * (LSB then MSB) to begin countdown.
          */
         outb(0xb0, 0x43);
-        outb((CLOCK_TICK_RATE / (1000 / 50)) & 0xff, 0x42);
-        outb((CLOCK_TICK_RATE / (1000 / 50)) >> 8, 0x42);
+        outb(latch & 0xff, 0x42);
+        outb(latch >> 8, 0x42);
 
         tsc = t1 = t2 = get_cycles();
 
@@ -166,31 +212,154 @@ static unsigned long pit_calibrate_tsc(void)
         /*
          * Sanity checks:
          *
-         * If we were not able to read the PIT more than 5000
+         * If we were not able to read the PIT more than loopmin
          * times, then we have been hit by a massive SMI
          *
          * If the maximum is 10 times larger than the minimum,
          * then we got hit by an SMI as well.
          */
-        if (pitcnt < 5000 || tscmax > 10 * tscmin)
+        if (pitcnt < loopmin || tscmax > 10 * tscmin)
                 return ULONG_MAX;
 
         /* Calculate the PIT value */
         delta = t2 - t1;
-        do_div(delta, 50);
+        do_div(delta, ms);
         return delta;
 }
 
+/*
+ * This reads the current MSB of the PIT counter, and
+ * checks if we are running on sufficiently fast and
+ * non-virtualized hardware.
+ *
+ * Our expectations are:
+ *
+ *  - the PIT is running at roughly 1.19MHz
+ *
+ *  - each IO is going to take about 1us on real hardware,
+ *    but we allow it to be much faster (by a factor of 10) or
+ *    _slightly_ slower (ie we allow up to a 2us read+counter
+ *    update - anything else implies a unacceptably slow CPU
+ *    or PIT for the fast calibration to work.
+ *
+ *  - with 256 PIT ticks to read the value, we have 214us to
+ *    see the same MSB (and overhead like doing a single TSC
+ *    read per MSB value etc).
+ *
+ *  - We're doing 2 reads per loop (LSB, MSB), and we expect
+ *    them each to take about a microsecond on real hardware.
+ *    So we expect a count value of around 100. But we'll be
+ *    generous, and accept anything over 50.
+ *
+ *  - if the PIT is stuck, and we see *many* more reads, we
+ *    return early (and the next caller of pit_expect_msb()
+ *    then consider it a failure when they don't see the
+ *    next expected value).
+ *
+ * These expectations mean that we know that we have seen the
+ * transition from one expected value to another with a fairly
+ * high accuracy, and we didn't miss any events. We can thus
+ * use the TSC value at the transitions to calculate a pretty
+ * good value for the TSC frequencty.
+ */
+static inline int pit_expect_msb(unsigned char val)
+{
+        int count = 0;
+
+        for (count = 0; count < 50000; count++) {
+                /* Ignore LSB */
+                inb(0x42);
+                if (inb(0x42) != val)
+                        break;
+        }
+        return count > 50;
+}
+
+/*
+ * How many MSB values do we want to see? We aim for a
+ * 15ms calibration, which assuming a 2us counter read
+ * error should give us roughly 150 ppm precision for
+ * the calibration.
+ */
+#define QUICK_PIT_MS 15
+#define QUICK_PIT_ITERATIONS (QUICK_PIT_MS * PIT_TICK_RATE / 1000 / 256)
+
+static unsigned long quick_pit_calibrate(void)
+{
+        /* Set the Gate high, disable speaker */
+        outb((inb(0x61) & ~0x02) | 0x01, 0x61);
+
+        /*
+         * Counter 2, mode 0 (one-shot), binary count
+         *
+         * NOTE! Mode 2 decrements by two (and then the
+         * output is flipped each time, giving the same
+         * final output frequency as a decrement-by-one),
+         * so mode 0 is much better when looking at the
+         * individual counts.
+         */
+        outb(0xb0, 0x43);
+
+        /* Start at 0xffff */
+        outb(0xff, 0x42);
+        outb(0xff, 0x42);
+
+        if (pit_expect_msb(0xff)) {
+                int i;
+                u64 t1, t2, delta;
+                unsigned char expect = 0xfe;
+
+                t1 = get_cycles();
+                for (i = 0; i < QUICK_PIT_ITERATIONS; i++, expect--) {
+                        if (!pit_expect_msb(expect))
+                                goto failed;
+                }
+                t2 = get_cycles();
+
+                /*
+                 * Make sure we can rely on the second TSC timestamp:
+                 */
+                if (!pit_expect_msb(expect))
+                        goto failed;
+
+                /*
+                 * Ok, if we get here, then we've seen the
+                 * MSB of the PIT decrement QUICK_PIT_ITERATIONS
+                 * times, and each MSB had many hits, so we never
+                 * had any sudden jumps.
+                 *
+                 * As a result, we can depend on there not being
+                 * any odd delays anywhere, and the TSC reads are
+                 * reliable.
+                 *
+                 * kHz = ticks / time-in-seconds / 1000;
+                 * kHz = (t2 - t1) / (QPI * 256 / PIT_TICK_RATE) / 1000
+                 * kHz = ((t2 - t1) * PIT_TICK_RATE) / (QPI * 256 * 1000)
+                 */
+                delta = (t2 - t1)*PIT_TICK_RATE;
+                do_div(delta, QUICK_PIT_ITERATIONS*256*1000);
+                printk("Fast TSC calibration using PIT\n");
+                return delta;
+        }
+failed:
+        return 0;
+}
 
 /**
  * native_calibrate_tsc - calibrate the tsc on boot
  */
 unsigned long native_calibrate_tsc(void)
 {
-        u64 tsc1, tsc2, delta, pm1, pm2, hpet1, hpet2;
+        u64 tsc1, tsc2, delta, ref1, ref2;
         unsigned long tsc_pit_min = ULONG_MAX, tsc_ref_min = ULONG_MAX;
-        unsigned long flags;
-        int hpet = is_hpet_enabled(), i;
+        unsigned long flags, latch, ms, fast_calibrate;
+        int hpet = is_hpet_enabled(), i, loopmin;
+
+        local_irq_save(flags);
+        fast_calibrate = quick_pit_calibrate();
+        local_irq_restore(flags);
+        if (fast_calibrate)
+                return fast_calibrate;
 
         /*
          * Run 5 calibration loops to get the lowest frequency value
@@ -216,7 +385,13 @@ unsigned long native_calibrate_tsc(void)
          * calibration delay loop as we have to wait for a certain
          * amount of time anyway.
          */
-        for (i = 0; i < 5; i++) {
+
+        /* Preset PIT loop values */
+        latch = CAL_LATCH;
+        ms = CAL_MS;
+        loopmin = CAL_PIT_LOOPS;
+
+        for (i = 0; i < 3; i++) {
                 unsigned long tsc_pit_khz;
 
                 /*
@@ -226,16 +401,16 @@ unsigned long native_calibrate_tsc(void)
                  * read the end value.
                  */
                 local_irq_save(flags);
-                tsc1 = tsc_read_refs(&pm1, hpet ? &hpet1 : NULL);
-                tsc_pit_khz = pit_calibrate_tsc();
-                tsc2 = tsc_read_refs(&pm2, hpet ? &hpet2 : NULL);
+                tsc1 = tsc_read_refs(&ref1, hpet);
+                tsc_pit_khz = pit_calibrate_tsc(latch, ms, loopmin);
+                tsc2 = tsc_read_refs(&ref2, hpet);
                 local_irq_restore(flags);
 
                 /* Pick the lowest PIT TSC calibration so far */
                 tsc_pit_min = min(tsc_pit_min, tsc_pit_khz);
 
                 /* hpet or pmtimer available ? */
-                if (!hpet && !pm1 && !pm2)
+                if (!hpet && !ref1 && !ref2)
                         continue;
 
                 /* Check, whether the sampling was disturbed by an SMI */
@@ -243,23 +418,41 @@ unsigned long native_calibrate_tsc(void)
                         continue;
 
                 tsc2 = (tsc2 - tsc1) * 1000000LL;
+                if (hpet)
+                        tsc2 = calc_hpet_ref(tsc2, ref1, ref2);
+                else
+                        tsc2 = calc_pmtimer_ref(tsc2, ref1, ref2);
 
-                if (hpet) {
-                        if (hpet2 < hpet1)
-                                hpet2 += 0x100000000ULL;
-                        hpet2 -= hpet1;
-                        tsc1 = ((u64)hpet2 * hpet_readl(HPET_PERIOD));
-                        do_div(tsc1, 1000000);
-                } else {
-                        if (pm2 < pm1)
-                                pm2 += (u64)ACPI_PM_OVRRUN;
-                        pm2 -= pm1;
-                        tsc1 = pm2 * 1000000000LL;
-                        do_div(tsc1, PMTMR_TICKS_PER_SEC);
+                tsc_ref_min = min(tsc_ref_min, (unsigned long) tsc2);
+
+                /* Check the reference deviation */
+                delta = ((u64) tsc_pit_min) * 100;
+                do_div(delta, tsc_ref_min);
+
+                /*
+                 * If both calibration results are inside a 10% window
+                 * then we can be sure, that the calibration
+                 * succeeded. We break out of the loop right away. We
+                 * use the reference value, as it is more precise.
+                 */
+                if (delta >= 90 && delta <= 110) {
+                        printk(KERN_INFO
+                               "TSC: PIT calibration matches %s. %d loops\n",
+                               hpet ? "HPET" : "PMTIMER", i + 1);
+                        return tsc_ref_min;
                 }
 
-                do_div(tsc2, tsc1);
-                tsc_ref_min = min(tsc_ref_min, (unsigned long) tsc2);
+                /*
+                 * Check whether PIT failed more than once. This
+                 * happens in virtualized environments. We need to
+                 * give the virtual PC a slightly longer timeframe for
+                 * the HPET/PMTIMER to make the result precise.
+                 */
+                if (i == 1 && tsc_pit_min == ULONG_MAX) {
+                        latch = CAL2_LATCH;
+                        ms = CAL2_MS;
+                        loopmin = CAL2_PIT_LOOPS;
+                }
         }
 
         /*
@@ -270,7 +463,7 @@ unsigned long native_calibrate_tsc(void)
                 printk(KERN_WARNING "TSC: Unable to calibrate against PIT\n");
 
                 /* We don't have an alternative source, disable TSC */
-                if (!hpet && !pm1 && !pm2) {
+                if (!hpet && !ref1 && !ref2) {
                         printk("TSC: No reference (HPET/PMTIMER) available\n");
                         return 0;
                 }
@@ -278,7 +471,7 @@ unsigned long native_calibrate_tsc(void)
                 /* The alternative source failed as well, disable TSC */
                 if (tsc_ref_min == ULONG_MAX) {
                         printk(KERN_WARNING "TSC: HPET/PMTIMER calibration "
-                               "failed due to SMI disturbance.\n");
+                               "failed.\n");
                         return 0;
                 }
 
@@ -290,44 +483,25 @@ unsigned long native_calibrate_tsc(void)
         }
 
         /* We don't have an alternative source, use the PIT calibration value */
-        if (!hpet && !pm1 && !pm2) {
+        if (!hpet && !ref1 && !ref2) {
                 printk(KERN_INFO "TSC: Using PIT calibration value\n");
                 return tsc_pit_min;
         }
 
         /* The alternative source failed, use the PIT calibration value */
         if (tsc_ref_min == ULONG_MAX) {
-                printk(KERN_WARNING "TSC: HPET/PMTIMER calibration failed due "
-                       "to SMI disturbance. Using PIT calibration\n");
+                printk(KERN_WARNING "TSC: HPET/PMTIMER calibration failed. "
+                       "Using PIT calibration\n");
                 return tsc_pit_min;
         }
 
-        /* Check the reference deviation */
-        delta = ((u64) tsc_pit_min) * 100;
-        do_div(delta, tsc_ref_min);
-
-        /*
-         * If both calibration results are inside a 5% window, the we
-         * use the lower frequency of those as it is probably the
-         * closest estimate.
-         */
-        if (delta >= 95 && delta <= 105) {
-                printk(KERN_INFO "TSC: PIT calibration confirmed by %s.\n",
-                       hpet ? "HPET" : "PMTIMER");
-                printk(KERN_INFO "TSC: using %s calibration value\n",
-                       tsc_pit_min <= tsc_ref_min ? "PIT" :
-                       hpet ? "HPET" : "PMTIMER");
-                return tsc_pit_min <= tsc_ref_min ? tsc_pit_min : tsc_ref_min;
-        }
-
-        printk(KERN_WARNING "TSC: PIT calibration deviates from %s: %lu %lu.\n",
-               hpet ? "HPET" : "PMTIMER", tsc_pit_min, tsc_ref_min);
-
         /*
          * The calibration values differ too much. In doubt, we use
          * the PIT value as we know that there are PMTIMERs around
-         * running at double speed.
+         * running at double speed. At least we let the user know:
          */
+        printk(KERN_WARNING "TSC: PIT calibration deviates from %s: %lu %lu.\n",
+               hpet ? "HPET" : "PMTIMER", tsc_pit_min, tsc_ref_min);
         printk(KERN_INFO "TSC: Using PIT calibration value\n");
         return tsc_pit_min;
 }
diff --git a/arch/x86/kernel/visws_quirks.c b/arch/x86/kernel/visws_quirks.c
index 594ef47f0a63..61a97e616f70 100644
--- a/arch/x86/kernel/visws_quirks.c
+++ b/arch/x86/kernel/visws_quirks.c
@@ -25,45 +25,31 @@
 #include <asm/visws/cobalt.h>
 #include <asm/visws/piix4.h>
 #include <asm/arch_hooks.h>
+#include <asm/io_apic.h>
 #include <asm/fixmap.h>
 #include <asm/reboot.h>
 #include <asm/setup.h>
 #include <asm/e820.h>
-#include <asm/smp.h>
 #include <asm/io.h>
 
 #include <mach_ipi.h>
 
 #include "mach_apic.h"
 
-#include <linux/init.h>
-#include <linux/smp.h>
-
 #include <linux/kernel_stat.h>
-#include <linux/interrupt.h>
-#include <linux/init.h>
 
-#include <asm/io.h>
-#include <asm/apic.h>
 #include <asm/i8259.h>
 #include <asm/irq_vectors.h>
-#include <asm/visws/cobalt.h>
 #include <asm/visws/lithium.h>
-#include <asm/visws/piix4.h>
 
 #include <linux/sched.h>
 #include <linux/kernel.h>
-#include <linux/init.h>
 #include <linux/pci.h>
 #include <linux/pci_ids.h>
 
 extern int no_broadcast;
 
-#include <asm/io.h>
 #include <asm/apic.h>
-#include <asm/arch_hooks.h>
-#include <asm/visws/cobalt.h>
-#include <asm/visws/lithium.h>
 
 char visws_board_type   = -1;
 char visws_board_rev    = -1;
diff --git a/arch/x86/kernel/vm86_32.c b/arch/x86/kernel/vm86_32.c
index 38f566fa27d2..4eeb5cf9720d 100644
--- a/arch/x86/kernel/vm86_32.c
+++ b/arch/x86/kernel/vm86_32.c
@@ -46,6 +46,7 @@
 #include <asm/io.h>
 #include <asm/tlbflush.h>
 #include <asm/irq.h>
+#include <asm/syscalls.h>
 
 /*
  * Known problems:
diff --git a/arch/x86/kernel/vmi_32.c b/arch/x86/kernel/vmi_32.c
index 6ca515d6db54..8c9ad02af5a2 100644
--- a/arch/x86/kernel/vmi_32.c
+++ b/arch/x86/kernel/vmi_32.c
@@ -235,7 +235,7 @@ static void vmi_write_ldt_entry(struct desc_struct *dt, int entry,
                                 const void *desc)
 {
         u32 *ldt_entry = (u32 *)desc;
-        vmi_ops.write_idt_entry(dt, entry, ldt_entry[0], ldt_entry[1]);
+        vmi_ops.write_ldt_entry(dt, entry, ldt_entry[0], ldt_entry[1]);
 }
 
 static void vmi_load_sp0(struct tss_struct *tss,
@@ -393,13 +393,13 @@ static void *vmi_kmap_atomic_pte(struct page *page, enum km_type type)
 }
 #endif
 
-static void vmi_allocate_pte(struct mm_struct *mm, u32 pfn)
+static void vmi_allocate_pte(struct mm_struct *mm, unsigned long pfn)
 {
         vmi_set_page_type(pfn, VMI_PAGE_L1);
         vmi_ops.allocate_page(pfn, VMI_PAGE_L1, 0, 0, 0);
 }
 
-static void vmi_allocate_pmd(struct mm_struct *mm, u32 pfn)
+static void vmi_allocate_pmd(struct mm_struct *mm, unsigned long pfn)
 {
         /*
          * This call comes in very early, before mem_map is setup.
@@ -410,20 +410,20 @@ static void vmi_allocate_pmd(struct mm_struct *mm, u32 pfn)
         vmi_ops.allocate_page(pfn, VMI_PAGE_L2, 0, 0, 0);
 }
 
-static void vmi_allocate_pmd_clone(u32 pfn, u32 clonepfn, u32 start, u32 count)
+static void vmi_allocate_pmd_clone(unsigned long pfn, unsigned long clonepfn, unsigned long start, unsigned long count)
 {
         vmi_set_page_type(pfn, VMI_PAGE_L2 | VMI_PAGE_CLONE);
         vmi_check_page_type(clonepfn, VMI_PAGE_L2);
         vmi_ops.allocate_page(pfn, VMI_PAGE_L2 | VMI_PAGE_CLONE, clonepfn, start, count);
 }
 
-static void vmi_release_pte(u32 pfn)
+static void vmi_release_pte(unsigned long pfn)
 {
         vmi_ops.release_page(pfn, VMI_PAGE_L1);
         vmi_set_page_type(pfn, VMI_PAGE_NORMAL);
 }
 
-static void vmi_release_pmd(u32 pfn)
+static void vmi_release_pmd(unsigned long pfn)
 {
         vmi_ops.release_page(pfn, VMI_PAGE_L2);
         vmi_set_page_type(pfn, VMI_PAGE_NORMAL);
diff --git a/arch/x86/kernel/vsmp_64.c b/arch/x86/kernel/vsmp_64.c
index 0c029e8959c7..7766d36983fc 100644
--- a/arch/x86/kernel/vsmp_64.c
+++ b/arch/x86/kernel/vsmp_64.c
@@ -61,7 +61,7 @@ static void vsmp_irq_enable(void)
         native_restore_fl((flags | X86_EFLAGS_IF) & (~X86_EFLAGS_AC));
 }
 
-static unsigned __init vsmp_patch(u8 type, u16 clobbers, void *ibuf,
+static unsigned __init_or_module vsmp_patch(u8 type, u16 clobbers, void *ibuf,
                                   unsigned long addr, unsigned len)
 {
         switch (type) {