Merge branch 'x86-64'

* x86-64: (83 commits)
  [PATCH] x86_64: x86_64 stack usage debugging
  [PATCH] x86_64: (resend) x86_64 stack overflow debugging
  [PATCH] x86_64: msi_apic.c build fix
  [PATCH] x86_64: i386/x86-64 Add nmi watchdog support for new Intel CPUs
  [PATCH] x86_64: Avoid broadcasting NMI IPIs
  [PATCH] x86_64: fix apic error on bootup
  [PATCH] x86_64: enlarge window for stack growth
  [PATCH] x86_64: Minor string functions optimizations
  [PATCH] x86_64: Move export symbols to their C functions
  [PATCH] x86_64: Standardize i386/x86_64 handling of NMI_VECTOR
  [PATCH] x86_64: Fix modular pc speaker
  [PATCH] x86_64: remove sys32_ni_syscall()
  [PATCH] x86_64: Do not use -ffunction-sections for modules
  [PATCH] x86_64: Add cpu_relax to apic_wait_icr_idle
  [PATCH] x86_64: adjust kstack_depth_to_print default
  [PATCH] i386/x86-64: adjust /proc/interrupts column headings
  [PATCH] x86_64: Fix race in cpu_local_* on preemptible kernels
  [PATCH] x86_64: Fix fast check in safe_smp_processor_id
  [PATCH] x86_64: x86_64 setup.c - printing cmp related boottime information
  [PATCH] i386/x86-64/ia64: Move polling flag into thread_info_status
  ...

Manual resolve of trivial conflict in arch/i386/kernel/Makefile

23 files changed, 689 insertions, 140 deletions

diff --git a/arch/i386/Kconfig b/arch/i386/Kconfig
index 374fb50608a0..f3eaf22f273d 100644
--- a/arch/i386/Kconfig
+++ b/arch/i386/Kconfig
@@ -328,6 +328,15 @@ config X86_MCE_P4THERMAL
           Enabling this feature will cause a message to be printed when the P4
           enters thermal throttling.
 
+config VM86
+        default y
+        bool "Enable VM86 support" if EMBEDDED
+        help
+          This option is required by programs like DOSEMU to run 16-bit legacy
+          code on X86 processors. It also may be needed by software like
+          XFree86 to initialize some video cards via BIOS. Disabling this
+          option saves about 6k.
+
 config TOSHIBA
         tristate "Toshiba Laptop support"
         ---help---
@@ -1068,6 +1077,10 @@ config SCx200HR_TIMER
           processor goes idle (as is done by the scheduler).  The
           other workaround is idle=poll boot option.
 
+config K8_NB
+        def_bool y
+        depends on AGP_AMD64
+
 source "drivers/pcmcia/Kconfig"
 
 source "drivers/pci/hotplug/Kconfig"
diff --git a/arch/i386/boot/Makefile b/arch/i386/boot/Makefile
index 33e55476381b..e97946626064 100644
--- a/arch/i386/boot/Makefile
+++ b/arch/i386/boot/Makefile
@@ -109,8 +109,13 @@ fdimage288: $(BOOTIMAGE) $(obj)/mtools.conf
 isoimage: $(BOOTIMAGE)
         -rm -rf $(obj)/isoimage
         mkdir $(obj)/isoimage
-        cp `echo /usr/lib*/syslinux/isolinux.bin | awk '{ print $1; }'` \
-                $(obj)/isoimage
+        for i in lib lib64 share end ; do \
+                if [ -f /usr/$$i/syslinux/isolinux.bin ] ; then \
+                        cp /usr/$$i/syslinux/isolinux.bin $(obj)/isoimage ; \
+                        break ; \
+                fi ; \
+                if [ $$i = end ] ; then exit 1 ; fi ; \
+        done
         cp $(BOOTIMAGE) $(obj)/isoimage/linux
         echo '$(image_cmdline)' > $(obj)/isoimage/isolinux.cfg
         if [ -f '$(FDINITRD)' ] ; then \
diff --git a/arch/i386/boot/compressed/misc.c b/arch/i386/boot/compressed/misc.c
index f19f3a7492a5..b2ccd543410d 100644
--- a/arch/i386/boot/compressed/misc.c
+++ b/arch/i386/boot/compressed/misc.c
@@ -24,14 +24,6 @@
 
 #undef memset
 #undef memcpy
-
-/*
- * Why do we do this? Don't ask me..
- *
- * Incomprehensible are the ways of bootloaders.
- */
-static void* memset(void *, int, size_t);
-static void* memcpy(void *, __const void *, size_t);
 #define memzero(s, n)     memset ((s), 0, (n))
 
 typedef unsigned char  uch;
@@ -93,7 +85,7 @@ static unsigned char *real_mode; /* Pointer to real-mode data */
 #endif
 #define RM_SCREEN_INFO (*(struct screen_info *)(real_mode+0))
 
-extern char input_data[];
+extern unsigned char input_data[];
 extern int input_len;
 
 static long bytes_out = 0;
@@ -103,6 +95,9 @@ static unsigned long output_ptr = 0;
 static void *malloc(int size);
 static void free(void *where);
 
+static void *memset(void *s, int c, unsigned n);
+static void *memcpy(void *dest, const void *src, unsigned n);
+
 static void putstr(const char *);
 
 extern int end;
@@ -205,7 +200,7 @@ static void putstr(const char *s)
         outb_p(0xff & (pos >> 1), vidport+1);
 }
 
-static void* memset(void* s, int c, size_t n)
+static void* memset(void* s, int c, unsigned n)
 {
         int i;
         char *ss = (char*)s;
@@ -214,14 +209,13 @@ static void* memset(void* s, int c, size_t n)
         return s;
 }
 
-static void* memcpy(void* __dest, __const void* __src,
-                            size_t __n)
+static void* memcpy(void* dest, const void* src, unsigned n)
 {
         int i;
-        char *d = (char *)__dest, *s = (char *)__src;
+        char *d = (char *)dest, *s = (char *)src;
 
-        for (i=0;i<__n;i++) d[i] = s[i];
-        return __dest;
+        for (i=0;i<n;i++) d[i] = s[i];
+        return dest;
 }
 
 /* ===========================================================================
@@ -309,7 +303,7 @@ static void setup_normal_output_buffer(void)
 #else
         if ((RM_ALT_MEM_K > RM_EXT_MEM_K ? RM_ALT_MEM_K : RM_EXT_MEM_K) < 1024) error("Less than 2MB of memory");
 #endif
-        output_data = (char *)__PHYSICAL_START; /* Normally Points to 1M */
+        output_data = (unsigned char *)__PHYSICAL_START; /* Normally Points to 1M */
         free_mem_end_ptr = (long)real_mode;
 }
 
@@ -324,11 +318,9 @@ static void setup_output_buffer_if_we_run_high(struct moveparams *mv)
 #ifdef STANDARD_MEMORY_BIOS_CALL
         if (RM_EXT_MEM_K < (3*1024)) error("Less than 4MB of memory");
 #else
-        if ((RM_ALT_MEM_K > RM_EXT_MEM_K ? RM_ALT_MEM_K : RM_EXT_MEM_K) <
-                        (3*1024))
-                error("Less than 4MB of memory");
+        if ((RM_ALT_MEM_K > RM_EXT_MEM_K ? RM_ALT_MEM_K : RM_EXT_MEM_K) < (3*1024)) error("Less than 4MB of memory");
 #endif  
-        mv->low_buffer_start = output_data = (char *)LOW_BUFFER_START;
+        mv->low_buffer_start = output_data = (unsigned char *)LOW_BUFFER_START;
         low_buffer_end = ((unsigned int)real_mode > LOW_BUFFER_MAX
           ? LOW_BUFFER_MAX : (unsigned int)real_mode) & ~0xfff;
         low_buffer_size = low_buffer_end - LOW_BUFFER_START;
diff --git a/arch/i386/kernel/Makefile b/arch/i386/kernel/Makefile
index 0fac85df64f1..5e70c2fb273a 100644
--- a/arch/i386/kernel/Makefile
+++ b/arch/i386/kernel/Makefile
@@ -37,6 +37,7 @@ obj-$(CONFIG_DOUBLEFAULT) 	+= doublefault.o
 obj-$(CONFIG_VM86)              += vm86.o
 obj-$(CONFIG_EARLY_PRINTK)      += early_printk.o
 obj-$(CONFIG_HPET_TIMER)        += hpet.o
+obj-$(CONFIG_K8_NB)             += k8.o
 
 EXTRA_AFLAGS   := -traditional
 
@@ -76,3 +77,6 @@ SYSCFLAGS_vsyscall-syms.o = -r
 $(obj)/vsyscall-syms.o: $(src)/vsyscall.lds \
                         $(obj)/vsyscall-sysenter.o $(obj)/vsyscall-note.o FORCE
         $(call if_changed,syscall)
+
+k8-y                      += ../../x86_64/kernel/k8.o
+
diff --git a/arch/i386/kernel/alternative.c b/arch/i386/kernel/alternative.c
index 5cbd6f99fb2a..50eb0e03777e 100644
--- a/arch/i386/kernel/alternative.c
+++ b/arch/i386/kernel/alternative.c
@@ -4,27 +4,41 @@
 #include <asm/alternative.h>
 #include <asm/sections.h>
 
-#define DEBUG 0
-#if DEBUG
-# define DPRINTK(fmt, args...) printk(fmt, args)
-#else
-# define DPRINTK(fmt, args...)
-#endif
+static int no_replacement    = 0;
+static int smp_alt_once      = 0;
+static int debug_alternative = 0;
+
+static int __init noreplacement_setup(char *s)
+{
+        no_replacement = 1;
+        return 1;
+}
+static int __init bootonly(char *str)
+{
+        smp_alt_once = 1;
+        return 1;
+}
+static int __init debug_alt(char *str)
+{
+        debug_alternative = 1;
+        return 1;
+}
 
+__setup("noreplacement", noreplacement_setup);
+__setup("smp-alt-boot", bootonly);
+__setup("debug-alternative", debug_alt);
+
+#define DPRINTK(fmt, args...) if (debug_alternative) \
+        printk(KERN_DEBUG fmt, args)
+
+#ifdef GENERIC_NOP1
 /* Use inline assembly to define this because the nops are defined
    as inline assembly strings in the include files and we cannot
    get them easily into strings. */
 asm("\t.data\nintelnops: "
         GENERIC_NOP1 GENERIC_NOP2 GENERIC_NOP3 GENERIC_NOP4 GENERIC_NOP5 GENERIC_NOP6
         GENERIC_NOP7 GENERIC_NOP8);
-asm("\t.data\nk8nops: "
-        K8_NOP1 K8_NOP2 K8_NOP3 K8_NOP4 K8_NOP5 K8_NOP6
-        K8_NOP7 K8_NOP8);
-asm("\t.data\nk7nops: "
-        K7_NOP1 K7_NOP2 K7_NOP3 K7_NOP4 K7_NOP5 K7_NOP6
-        K7_NOP7 K7_NOP8);
-
-extern unsigned char intelnops[], k8nops[], k7nops[];
+extern unsigned char intelnops[];
 static unsigned char *intel_nops[ASM_NOP_MAX+1] = {
         NULL,
         intelnops,
@@ -36,6 +50,13 @@ static unsigned char *intel_nops[ASM_NOP_MAX+1] = {
         intelnops + 1 + 2 + 3 + 4 + 5 + 6,
         intelnops + 1 + 2 + 3 + 4 + 5 + 6 + 7,
 };
+#endif
+
+#ifdef K8_NOP1
+asm("\t.data\nk8nops: "
+        K8_NOP1 K8_NOP2 K8_NOP3 K8_NOP4 K8_NOP5 K8_NOP6
+        K8_NOP7 K8_NOP8);
+extern unsigned char k8nops[];
 static unsigned char *k8_nops[ASM_NOP_MAX+1] = {
         NULL,
         k8nops,
@@ -47,6 +68,13 @@ static unsigned char *k8_nops[ASM_NOP_MAX+1] = {
         k8nops + 1 + 2 + 3 + 4 + 5 + 6,
         k8nops + 1 + 2 + 3 + 4 + 5 + 6 + 7,
 };
+#endif
+
+#ifdef K7_NOP1
+asm("\t.data\nk7nops: "
+        K7_NOP1 K7_NOP2 K7_NOP3 K7_NOP4 K7_NOP5 K7_NOP6
+        K7_NOP7 K7_NOP8);
+extern unsigned char k7nops[];
 static unsigned char *k7_nops[ASM_NOP_MAX+1] = {
         NULL,
         k7nops,
@@ -58,6 +86,18 @@ static unsigned char *k7_nops[ASM_NOP_MAX+1] = {
         k7nops + 1 + 2 + 3 + 4 + 5 + 6,
         k7nops + 1 + 2 + 3 + 4 + 5 + 6 + 7,
 };
+#endif
+
+#ifdef CONFIG_X86_64
+
+extern char __vsyscall_0;
+static inline unsigned char** find_nop_table(void)
+{
+        return k8_nops;
+}
+
+#else /* CONFIG_X86_64 */
+
 static struct nop {
         int cpuid;
         unsigned char **noptable;
@@ -67,14 +107,6 @@ static struct nop {
         { -1, NULL }
 };
 
-
-extern struct alt_instr __alt_instructions[], __alt_instructions_end[];
-extern struct alt_instr __smp_alt_instructions[], __smp_alt_instructions_end[];
-extern u8 *__smp_locks[], *__smp_locks_end[];
-
-extern u8 __smp_alt_begin[], __smp_alt_end[];
-
-
 static unsigned char** find_nop_table(void)
 {
         unsigned char **noptable = intel_nops;
@@ -89,6 +121,14 @@ static unsigned char** find_nop_table(void)
         return noptable;
 }
 
+#endif /* CONFIG_X86_64 */
+
+extern struct alt_instr __alt_instructions[], __alt_instructions_end[];
+extern struct alt_instr __smp_alt_instructions[], __smp_alt_instructions_end[];
+extern u8 *__smp_locks[], *__smp_locks_end[];
+
+extern u8 __smp_alt_begin[], __smp_alt_end[];
+
 /* Replace instructions with better alternatives for this CPU type.
    This runs before SMP is initialized to avoid SMP problems with
    self modifying code. This implies that assymetric systems where
@@ -99,6 +139,7 @@ void apply_alternatives(struct alt_instr *start, struct alt_instr *end)
 {
         unsigned char **noptable = find_nop_table();
         struct alt_instr *a;
+        u8 *instr;
         int diff, i, k;
 
         DPRINTK("%s: alt table %p -> %p\n", __FUNCTION__, start, end);
@@ -106,7 +147,16 @@ void apply_alternatives(struct alt_instr *start, struct alt_instr *end)
                 BUG_ON(a->replacementlen > a->instrlen);
                 if (!boot_cpu_has(a->cpuid))
                         continue;
-                memcpy(a->instr, a->replacement, a->replacementlen);
+                instr = a->instr;
+#ifdef CONFIG_X86_64
+                /* vsyscall code is not mapped yet. resolve it manually. */
+                if (instr >= (u8 *)VSYSCALL_START && instr < (u8*)VSYSCALL_END) {
+                        instr = __va(instr - (u8*)VSYSCALL_START + (u8*)__pa_symbol(&__vsyscall_0));
+                        DPRINTK("%s: vsyscall fixup: %p => %p\n",
+                                __FUNCTION__, a->instr, instr);
+                }
+#endif
+                memcpy(instr, a->replacement, a->replacementlen);
                 diff = a->instrlen - a->replacementlen;
                 /* Pad the rest with nops */
                 for (i = a->replacementlen; diff > 0; diff -= k, i += k) {
@@ -186,14 +236,6 @@ struct smp_alt_module {
 static LIST_HEAD(smp_alt_modules);
 static DEFINE_SPINLOCK(smp_alt);
 
-static int smp_alt_once = 0;
-static int __init bootonly(char *str)
-{
-        smp_alt_once = 1;
-        return 1;
-}
-__setup("smp-alt-boot", bootonly);
-
 void alternatives_smp_module_add(struct module *mod, char *name,
                                  void *locks, void *locks_end,
                                  void *text,  void *text_end)
@@ -201,6 +243,9 @@ void alternatives_smp_module_add(struct module *mod, char *name,
         struct smp_alt_module *smp;
         unsigned long flags;
 
+        if (no_replacement)
+                return;
+
         if (smp_alt_once) {
                 if (boot_cpu_has(X86_FEATURE_UP))
                         alternatives_smp_unlock(locks, locks_end,
@@ -235,7 +280,7 @@ void alternatives_smp_module_del(struct module *mod)
         struct smp_alt_module *item;
         unsigned long flags;
 
-        if (smp_alt_once)
+        if (no_replacement || smp_alt_once)
                 return;
 
         spin_lock_irqsave(&smp_alt, flags);
@@ -256,7 +301,7 @@ void alternatives_smp_switch(int smp)
         struct smp_alt_module *mod;
         unsigned long flags;
 
-        if (smp_alt_once)
+        if (no_replacement || smp_alt_once)
                 return;
         BUG_ON(!smp && (num_online_cpus() > 1));
 
@@ -285,6 +330,13 @@ void alternatives_smp_switch(int smp)
 
 void __init alternative_instructions(void)
 {
+        if (no_replacement) {
+                printk(KERN_INFO "(SMP-)alternatives turned off\n");
+                free_init_pages("SMP alternatives",
+                                (unsigned long)__smp_alt_begin,
+                                (unsigned long)__smp_alt_end);
+                return;
+        }
         apply_alternatives(__alt_instructions, __alt_instructions_end);
 
         /* switch to patch-once-at-boottime-only mode and free the
diff --git a/arch/i386/kernel/apic.c b/arch/i386/kernel/apic.c
index 5ab59c12335b..7ce09492fc0c 100644
--- a/arch/i386/kernel/apic.c
+++ b/arch/i386/kernel/apic.c
@@ -36,6 +36,7 @@
 #include <asm/arch_hooks.h>
 #include <asm/hpet.h>
 #include <asm/i8253.h>
+#include <asm/nmi.h>
 
 #include <mach_apic.h>
 #include <mach_apicdef.h>
@@ -156,7 +157,7 @@ void clear_local_APIC(void)
         maxlvt = get_maxlvt();
 
         /*
-         * Masking an LVT entry on a P6 can trigger a local APIC error
+         * Masking an LVT entry can trigger a local APIC error
          * if the vector is zero. Mask LVTERR first to prevent this.
          */
         if (maxlvt >= 3) {
@@ -1117,7 +1118,18 @@ void disable_APIC_timer(void)
                 unsigned long v;
 
                 v = apic_read(APIC_LVTT);
-                apic_write_around(APIC_LVTT, v | APIC_LVT_MASKED);
+                /*
+                 * When an illegal vector value (0-15) is written to an LVT
+                 * entry and delivery mode is Fixed, the APIC may signal an
+                 * illegal vector error, with out regard to whether the mask
+                 * bit is set or whether an interrupt is actually seen on input.
+                 *
+                 * Boot sequence might call this function when the LVTT has
+                 * '0' vector value. So make sure vector field is set to
+                 * valid value.
+                 */
+                v |= (APIC_LVT_MASKED | LOCAL_TIMER_VECTOR);
+                apic_write_around(APIC_LVTT, v);
         }
 }
 
diff --git a/arch/i386/kernel/apm.c b/arch/i386/kernel/apm.c
index 9e819eb68229..7c5729d1fd06 100644
--- a/arch/i386/kernel/apm.c
+++ b/arch/i386/kernel/apm.c
@@ -764,9 +764,9 @@ static int apm_do_idle(void)
         int     idled = 0;
         int     polling;
 
-        polling = test_thread_flag(TIF_POLLING_NRFLAG);
+        polling = !!(current_thread_info()->status & TS_POLLING);
         if (polling) {
-                clear_thread_flag(TIF_POLLING_NRFLAG);
+                current_thread_info()->status &= ~TS_POLLING;
                 smp_mb__after_clear_bit();
         }
         if (!need_resched()) {
@@ -774,7 +774,7 @@ static int apm_do_idle(void)
                 ret = apm_bios_call_simple(APM_FUNC_IDLE, 0, 0, &eax);
         }
         if (polling)
-                set_thread_flag(TIF_POLLING_NRFLAG);
+                current_thread_info()->status |= TS_POLLING;
 
         if (!idled)
                 return 0;
diff --git a/arch/i386/kernel/cpu/amd.c b/arch/i386/kernel/cpu/amd.c
index 786d1a57048b..fd0457c9c827 100644
--- a/arch/i386/kernel/cpu/amd.c
+++ b/arch/i386/kernel/cpu/amd.c
@@ -224,15 +224,17 @@ static void __init init_amd(struct cpuinfo_x86 *c)
 
 #ifdef CONFIG_X86_HT
         /*
-         * On a AMD dual core setup the lower bits of the APIC id
-         * distingush the cores.  Assumes number of cores is a power
-         * of two.
+         * On a AMD multi core setup the lower bits of the APIC id
+         * distingush the cores.
          */
         if (c->x86_max_cores > 1) {
                 int cpu = smp_processor_id();
-                unsigned bits = 0;
-                while ((1 << bits) < c->x86_max_cores)
-                        bits++;
+                unsigned bits = (cpuid_ecx(0x80000008) >> 12) & 0xf;
+
+                if (bits == 0) {
+                        while ((1 << bits) < c->x86_max_cores)
+                                bits++;
+                }
                 cpu_core_id[cpu] = phys_proc_id[cpu] & ((1<<bits)-1);
                 phys_proc_id[cpu] >>= bits;
                 printk(KERN_INFO "CPU %d(%d) -> Core %d\n",
@@ -240,6 +242,8 @@ static void __init init_amd(struct cpuinfo_x86 *c)
         }
 #endif
 
+        if (cpuid_eax(0x80000000) >= 0x80000006)
+                num_cache_leaves = 3;
 }
 
 static unsigned int amd_size_cache(struct cpuinfo_x86 * c, unsigned int size)
diff --git a/arch/i386/kernel/cpu/intel.c b/arch/i386/kernel/cpu/intel.c
index 5386b29bb5a5..10afc645c540 100644
--- a/arch/i386/kernel/cpu/intel.c
+++ b/arch/i386/kernel/cpu/intel.c
@@ -122,6 +122,12 @@ static void __cpuinit init_intel(struct cpuinfo_x86 *c)
 
         select_idle_routine(c);
         l2 = init_intel_cacheinfo(c);
+        if (c->cpuid_level > 9 ) {
+                unsigned eax = cpuid_eax(10);
+                /* Check for version and the number of counters */
+                if ((eax & 0xff) && (((eax>>8) & 0xff) > 1))
+                        set_bit(X86_FEATURE_ARCH_PERFMON, c->x86_capability);
+        }
 
         /* SEP CPUID bug: Pentium Pro reports SEP but doesn't have it until model 3 mask 3 */
         if ((c->x86<<8 | c->x86_model<<4 | c->x86_mask) < 0x633)
diff --git a/arch/i386/kernel/cpu/intel_cacheinfo.c b/arch/i386/kernel/cpu/intel_cacheinfo.c
index c8547a6fa7e6..6c37b4fd8ce2 100644
--- a/arch/i386/kernel/cpu/intel_cacheinfo.c
+++ b/arch/i386/kernel/cpu/intel_cacheinfo.c
@@ -4,6 +4,7 @@
  *      Changes:
  *      Venkatesh Pallipadi     : Adding cache identification through cpuid(4)
  *              Ashok Raj <ashok.raj@intel.com>: Work with CPU hotplug infrastructure.
+ *      Andi Kleen              : CPUID4 emulation on AMD.
  */
 
 #include <linux/init.h>
@@ -130,25 +131,111 @@ struct _cpuid4_info {
         cpumask_t shared_cpu_map;
 };
 
-static unsigned short                   num_cache_leaves;
+unsigned short                  num_cache_leaves;
+
+/* AMD doesn't have CPUID4. Emulate it here to report the same
+   information to the user.  This makes some assumptions about the machine:
+   No L3, L2 not shared, no SMT etc. that is currently true on AMD CPUs.
+
+   In theory the TLBs could be reported as fake type (they are in "dummy").
+   Maybe later */
+union l1_cache {
+        struct {
+                unsigned line_size : 8;
+                unsigned lines_per_tag : 8;
+                unsigned assoc : 8;
+                unsigned size_in_kb : 8;
+        };
+        unsigned val;
+};
+
+union l2_cache {
+        struct {
+                unsigned line_size : 8;
+                unsigned lines_per_tag : 4;
+                unsigned assoc : 4;
+                unsigned size_in_kb : 16;
+        };
+        unsigned val;
+};
+
+static unsigned short assocs[] = {
+        [1] = 1, [2] = 2, [4] = 4, [6] = 8,
+        [8] = 16,
+        [0xf] = 0xffff // ??
+        };
+static unsigned char levels[] = { 1, 1, 2 };
+static unsigned char types[] = { 1, 2, 3 };
+
+static void __cpuinit amd_cpuid4(int leaf, union _cpuid4_leaf_eax *eax,
+                       union _cpuid4_leaf_ebx *ebx,
+                       union _cpuid4_leaf_ecx *ecx)
+{
+        unsigned dummy;
+        unsigned line_size, lines_per_tag, assoc, size_in_kb;
+        union l1_cache l1i, l1d;
+        union l2_cache l2;
+
+        eax->full = 0;
+        ebx->full = 0;
+        ecx->full = 0;
+
+        cpuid(0x80000005, &dummy, &dummy, &l1d.val, &l1i.val);
+        cpuid(0x80000006, &dummy, &dummy, &l2.val, &dummy);
+
+        if (leaf > 2 || !l1d.val || !l1i.val || !l2.val)
+                return;
+
+        eax->split.is_self_initializing = 1;
+        eax->split.type = types[leaf];
+        eax->split.level = levels[leaf];
+        eax->split.num_threads_sharing = 0;
+        eax->split.num_cores_on_die = current_cpu_data.x86_max_cores - 1;
+
+        if (leaf <= 1) {
+                union l1_cache *l1 = leaf == 0 ? &l1d : &l1i;
+                assoc = l1->assoc;
+                line_size = l1->line_size;
+                lines_per_tag = l1->lines_per_tag;
+                size_in_kb = l1->size_in_kb;
+        } else {
+                assoc = l2.assoc;
+                line_size = l2.line_size;
+                lines_per_tag = l2.lines_per_tag;
+                /* cpu_data has errata corrections for K7 applied */
+                size_in_kb = current_cpu_data.x86_cache_size;
+        }
+
+        if (assoc == 0xf)
+                eax->split.is_fully_associative = 1;
+        ebx->split.coherency_line_size = line_size - 1;
+        ebx->split.ways_of_associativity = assocs[assoc] - 1;
+        ebx->split.physical_line_partition = lines_per_tag - 1;
+        ecx->split.number_of_sets = (size_in_kb * 1024) / line_size /
+                (ebx->split.ways_of_associativity + 1) - 1;
+}
 
 static int __cpuinit cpuid4_cache_lookup(int index, struct _cpuid4_info *this_leaf)
 {
-        unsigned int            eax, ebx, ecx, edx;
-        union _cpuid4_leaf_eax  cache_eax;
+        union _cpuid4_leaf_eax  eax;
+        union _cpuid4_leaf_ebx  ebx;
+        union _cpuid4_leaf_ecx  ecx;
+        unsigned                edx;
 
-        cpuid_count(4, index, &eax, &ebx, &ecx, &edx);
-        cache_eax.full = eax;
-        if (cache_eax.split.type == CACHE_TYPE_NULL)
+        if (boot_cpu_data.x86_vendor == X86_VENDOR_AMD)
+                amd_cpuid4(index, &eax, &ebx, &ecx);
+        else
+                cpuid_count(4, index, &eax.full, &ebx.full, &ecx.full,  &edx);
+        if (eax.split.type == CACHE_TYPE_NULL)
                 return -EIO; /* better error ? */
 
-        this_leaf->eax.full = eax;
-        this_leaf->ebx.full = ebx;
-        this_leaf->ecx.full = ecx;
-        this_leaf->size = (this_leaf->ecx.split.number_of_sets + 1) *
-                (this_leaf->ebx.split.coherency_line_size + 1) *
-                (this_leaf->ebx.split.physical_line_partition + 1) *
-                (this_leaf->ebx.split.ways_of_associativity + 1);
+        this_leaf->eax = eax;
+        this_leaf->ebx = ebx;
+        this_leaf->ecx = ecx;
+        this_leaf->size = (ecx.split.number_of_sets + 1) *
+                (ebx.split.coherency_line_size + 1) *
+                (ebx.split.physical_line_partition + 1) *
+                (ebx.split.ways_of_associativity + 1);
         return 0;
 }
 
diff --git a/arch/i386/kernel/crash.c b/arch/i386/kernel/crash.c
index 21dc1bbb8067..0c88d3ec8c18 100644
--- a/arch/i386/kernel/crash.c
+++ b/arch/i386/kernel/crash.c
@@ -120,14 +120,9 @@ static int crash_nmi_callback(struct pt_regs *regs, int cpu)
         return 1;
 }
 
-/*
- * By using the NMI code instead of a vector we just sneak thru the
- * word generator coming out with just what we want.  AND it does
- * not matter if clustered_apic_mode is set or not.
- */
 static void smp_send_nmi_allbutself(void)
 {
-        send_IPI_allbutself(APIC_DM_NMI);
+        send_IPI_allbutself(NMI_VECTOR);
 }
 
 static void nmi_shootdown_cpus(void)
diff --git a/arch/i386/kernel/entry.S b/arch/i386/kernel/entry.S
index cfc683f153b9..e6e4506e749a 100644
--- a/arch/i386/kernel/entry.S
+++ b/arch/i386/kernel/entry.S
@@ -48,6 +48,7 @@
 #include <asm/smp.h>
 #include <asm/page.h>
 #include <asm/desc.h>
+#include <asm/dwarf2.h>
 #include "irq_vectors.h"
 
 #define nr_syscalls ((syscall_table_size)/4)
@@ -85,31 +86,67 @@ VM_MASK		= 0x00020000
 #define SAVE_ALL \
         cld; \
         pushl %es; \
+        CFI_ADJUST_CFA_OFFSET 4;\
+        /*CFI_REL_OFFSET es, 0;*/\
         pushl %ds; \
+        CFI_ADJUST_CFA_OFFSET 4;\
+        /*CFI_REL_OFFSET ds, 0;*/\
         pushl %eax; \
+        CFI_ADJUST_CFA_OFFSET 4;\
+        CFI_REL_OFFSET eax, 0;\
         pushl %ebp; \
+        CFI_ADJUST_CFA_OFFSET 4;\
+        CFI_REL_OFFSET ebp, 0;\
         pushl %edi; \
+        CFI_ADJUST_CFA_OFFSET 4;\
+        CFI_REL_OFFSET edi, 0;\
         pushl %esi; \
+        CFI_ADJUST_CFA_OFFSET 4;\
+        CFI_REL_OFFSET esi, 0;\
         pushl %edx; \
+        CFI_ADJUST_CFA_OFFSET 4;\
+        CFI_REL_OFFSET edx, 0;\
         pushl %ecx; \
+        CFI_ADJUST_CFA_OFFSET 4;\
+        CFI_REL_OFFSET ecx, 0;\
         pushl %ebx; \
+        CFI_ADJUST_CFA_OFFSET 4;\
+        CFI_REL_OFFSET ebx, 0;\
         movl $(__USER_DS), %edx; \
         movl %edx, %ds; \
         movl %edx, %es;
 
 #define RESTORE_INT_REGS \
         popl %ebx;      \
+        CFI_ADJUST_CFA_OFFSET -4;\
+        CFI_RESTORE ebx;\
         popl %ecx;      \
+        CFI_ADJUST_CFA_OFFSET -4;\
+        CFI_RESTORE ecx;\
         popl %edx;      \
+        CFI_ADJUST_CFA_OFFSET -4;\
+        CFI_RESTORE edx;\
         popl %esi;      \
+        CFI_ADJUST_CFA_OFFSET -4;\
+        CFI_RESTORE esi;\
         popl %edi;      \
+        CFI_ADJUST_CFA_OFFSET -4;\
+        CFI_RESTORE edi;\
         popl %ebp;      \
-        popl %eax
+        CFI_ADJUST_CFA_OFFSET -4;\
+        CFI_RESTORE ebp;\
+        popl %eax;      \
+        CFI_ADJUST_CFA_OFFSET -4;\
+        CFI_RESTORE eax
 
 #define RESTORE_REGS    \
         RESTORE_INT_REGS; \
 1:      popl %ds;       \
+        CFI_ADJUST_CFA_OFFSET -4;\
+        /*CFI_RESTORE ds;*/\
 2:      popl %es;       \
+        CFI_ADJUST_CFA_OFFSET -4;\
+        /*CFI_RESTORE es;*/\
 .section .fixup,"ax";   \
 3:      movl $0,(%esp); \
         jmp 1b;         \
@@ -122,13 +159,43 @@ VM_MASK		= 0x00020000
         .long 2b,4b;    \
 .previous
 
+#define RING0_INT_FRAME \
+        CFI_STARTPROC simple;\
+        CFI_DEF_CFA esp, 3*4;\
+        /*CFI_OFFSET cs, -2*4;*/\
+        CFI_OFFSET eip, -3*4
+
+#define RING0_EC_FRAME \
+        CFI_STARTPROC simple;\
+        CFI_DEF_CFA esp, 4*4;\
+        /*CFI_OFFSET cs, -2*4;*/\
+        CFI_OFFSET eip, -3*4
+
+#define RING0_PTREGS_FRAME \
+        CFI_STARTPROC simple;\
+        CFI_DEF_CFA esp, OLDESP-EBX;\
+        /*CFI_OFFSET cs, CS-OLDESP;*/\
+        CFI_OFFSET eip, EIP-OLDESP;\
+        /*CFI_OFFSET es, ES-OLDESP;*/\
+        /*CFI_OFFSET ds, DS-OLDESP;*/\
+        CFI_OFFSET eax, EAX-OLDESP;\
+        CFI_OFFSET ebp, EBP-OLDESP;\
+        CFI_OFFSET edi, EDI-OLDESP;\
+        CFI_OFFSET esi, ESI-OLDESP;\
+        CFI_OFFSET edx, EDX-OLDESP;\
+        CFI_OFFSET ecx, ECX-OLDESP;\
+        CFI_OFFSET ebx, EBX-OLDESP
 
 ENTRY(ret_from_fork)
+        CFI_STARTPROC
         pushl %eax
+        CFI_ADJUST_CFA_OFFSET -4
         call schedule_tail
         GET_THREAD_INFO(%ebp)
         popl %eax
+        CFI_ADJUST_CFA_OFFSET -4
         jmp syscall_exit
+        CFI_ENDPROC
 
 /*
  * Return to user mode is not as complex as all this looks,
@@ -139,6 +206,7 @@ ENTRY(ret_from_fork)
 
         # userspace resumption stub bypassing syscall exit tracing
         ALIGN
+        RING0_PTREGS_FRAME
 ret_from_exception:
         preempt_stop
 ret_from_intr:
@@ -171,20 +239,33 @@ need_resched:
         call preempt_schedule_irq
         jmp need_resched
 #endif
+        CFI_ENDPROC
 
 /* SYSENTER_RETURN points to after the "sysenter" instruction in
    the vsyscall page.  See vsyscall-sysentry.S, which defines the symbol.  */
 
         # sysenter call handler stub
 ENTRY(sysenter_entry)
+        CFI_STARTPROC simple
+        CFI_DEF_CFA esp, 0
+        CFI_REGISTER esp, ebp
         movl TSS_sysenter_esp0(%esp),%esp
 sysenter_past_esp:
         sti
         pushl $(__USER_DS)
+        CFI_ADJUST_CFA_OFFSET 4
+        /*CFI_REL_OFFSET ss, 0*/
         pushl %ebp
+        CFI_ADJUST_CFA_OFFSET 4
+        CFI_REL_OFFSET esp, 0
         pushfl
+        CFI_ADJUST_CFA_OFFSET 4
         pushl $(__USER_CS)
+        CFI_ADJUST_CFA_OFFSET 4
+        /*CFI_REL_OFFSET cs, 0*/
         pushl $SYSENTER_RETURN
+        CFI_ADJUST_CFA_OFFSET 4
+        CFI_REL_OFFSET eip, 0
 
 /*
  * Load the potential sixth argument from user stack.
@@ -199,6 +280,7 @@ sysenter_past_esp:
 .previous
 
         pushl %eax
+        CFI_ADJUST_CFA_OFFSET 4
         SAVE_ALL
         GET_THREAD_INFO(%ebp)
 
@@ -219,11 +301,14 @@ sysenter_past_esp:
         xorl %ebp,%ebp
         sti
         sysexit
+        CFI_ENDPROC
 
 
         # system call handler stub
 ENTRY(system_call)
+        RING0_INT_FRAME                 # can't unwind into user space anyway
         pushl %eax                      # save orig_eax
+        CFI_ADJUST_CFA_OFFSET 4
         SAVE_ALL
         GET_THREAD_INFO(%ebp)
         testl $TF_MASK,EFLAGS(%esp)
@@ -256,10 +341,12 @@ restore_all:
         movb CS(%esp), %al
         andl $(VM_MASK | (4 << 8) | 3), %eax
         cmpl $((4 << 8) | 3), %eax
+        CFI_REMEMBER_STATE
         je ldt_ss                       # returning to user-space with LDT SS
 restore_nocheck:
         RESTORE_REGS
         addl $4, %esp
+        CFI_ADJUST_CFA_OFFSET -4
 1:      iret
 .section .fixup,"ax"
 iret_exc:
@@ -273,6 +360,7 @@ iret_exc:
         .long 1b,iret_exc
 .previous
 
+        CFI_RESTORE_STATE
 ldt_ss:
         larl OLDSS(%esp), %eax
         jnz restore_nocheck
@@ -285,11 +373,13 @@ ldt_ss:
          * CPUs, which we can try to work around to make
          * dosemu and wine happy. */
         subl $8, %esp           # reserve space for switch16 pointer
+        CFI_ADJUST_CFA_OFFSET 8
         cli
         movl %esp, %eax
         /* Set up the 16bit stack frame with switch32 pointer on top,
          * and a switch16 pointer on top of the current frame. */
         call setup_x86_bogus_stack
+        CFI_ADJUST_CFA_OFFSET -8        # frame has moved
         RESTORE_REGS
         lss 20+4(%esp), %esp    # switch to 16bit stack
 1:      iret
@@ -297,9 +387,11 @@ ldt_ss:
         .align 4
         .long 1b,iret_exc
 .previous
+        CFI_ENDPROC
 
         # perform work that needs to be done immediately before resumption
         ALIGN
+        RING0_PTREGS_FRAME              # can't unwind into user space anyway
 work_pending:
         testb $_TIF_NEED_RESCHED, %cl
         jz work_notifysig
@@ -329,8 +421,10 @@ work_notifysig:				# deal with pending signals and
 work_notifysig_v86:
 #ifdef CONFIG_VM86
         pushl %ecx                      # save ti_flags for do_notify_resume
+        CFI_ADJUST_CFA_OFFSET 4
         call save_v86_state             # %eax contains pt_regs pointer
         popl %ecx
+        CFI_ADJUST_CFA_OFFSET -4
         movl %eax, %esp
         xorl %edx, %edx
         call do_notify_resume
@@ -363,19 +457,21 @@ syscall_exit_work:
         movl $1, %edx
         call do_syscall_trace
         jmp resume_userspace
+        CFI_ENDPROC
 
-        ALIGN
+        RING0_INT_FRAME                 # can't unwind into user space anyway
 syscall_fault:
         pushl %eax                      # save orig_eax
+        CFI_ADJUST_CFA_OFFSET 4
         SAVE_ALL
         GET_THREAD_INFO(%ebp)
         movl $-EFAULT,EAX(%esp)
         jmp resume_userspace
 
-        ALIGN
 syscall_badsys:
         movl $-ENOSYS,EAX(%esp)
         jmp resume_userspace
+        CFI_ENDPROC
 
 #define FIXUP_ESPFIX_STACK \
         movl %esp, %eax; \
@@ -387,16 +483,21 @@ syscall_badsys:
         movl %eax, %esp;
 #define UNWIND_ESPFIX_STACK \
         pushl %eax; \
+        CFI_ADJUST_CFA_OFFSET 4; \
         movl %ss, %eax; \
         /* see if on 16bit stack */ \
         cmpw $__ESPFIX_SS, %ax; \
-        jne 28f; \
-        movl $__KERNEL_DS, %edx; \
-        movl %edx, %ds; \
-        movl %edx, %es; \
+        je 28f; \
+27:     popl %eax; \
+        CFI_ADJUST_CFA_OFFSET -4; \
+.section .fixup,"ax"; \
+28:     movl $__KERNEL_DS, %eax; \
+        movl %eax, %ds; \
+        movl %eax, %es; \
         /* switch to 32bit stack */ \
-        FIXUP_ESPFIX_STACK \
-28:     popl %eax;
+        FIXUP_ESPFIX_STACK; \
+        jmp 27b; \
+.previous
 
 /*
  * Build the entry stubs and pointer table with
@@ -408,9 +509,14 @@ ENTRY(interrupt)
 
 vector=0
 ENTRY(irq_entries_start)
+        RING0_INT_FRAME
 .rept NR_IRQS
         ALIGN
+ .if vector
+        CFI_ADJUST_CFA_OFFSET -4
+ .endif
 1:      pushl $vector-256
+        CFI_ADJUST_CFA_OFFSET 4
         jmp common_interrupt
 .data
         .long 1b
@@ -424,60 +530,99 @@ common_interrupt:
         movl %esp,%eax
         call do_IRQ
         jmp ret_from_intr
+        CFI_ENDPROC
 
 #define BUILD_INTERRUPT(name, nr)       \
 ENTRY(name)                             \
+        RING0_INT_FRAME;                \
         pushl $nr-256;                  \
-        SAVE_ALL                        \
+        CFI_ADJUST_CFA_OFFSET 4;        \
+        SAVE_ALL;                       \
         movl %esp,%eax;                 \
         call smp_/**/name;              \
-        jmp ret_from_intr;
+        jmp ret_from_intr;      \
+        CFI_ENDPROC
 
 /* The include is where all of the SMP etc. interrupts come from */
 #include "entry_arch.h"
 
 ENTRY(divide_error)
+        RING0_INT_FRAME
         pushl $0                        # no error code
+        CFI_ADJUST_CFA_OFFSET 4
         pushl $do_divide_error
+        CFI_ADJUST_CFA_OFFSET 4
         ALIGN
 error_code:
         pushl %ds
+        CFI_ADJUST_CFA_OFFSET 4
+        /*CFI_REL_OFFSET ds, 0*/
         pushl %eax
+        CFI_ADJUST_CFA_OFFSET 4
+        CFI_REL_OFFSET eax, 0
         xorl %eax, %eax
         pushl %ebp
+        CFI_ADJUST_CFA_OFFSET 4
+        CFI_REL_OFFSET ebp, 0
         pushl %edi
+        CFI_ADJUST_CFA_OFFSET 4
+        CFI_REL_OFFSET edi, 0
         pushl %esi
+        CFI_ADJUST_CFA_OFFSET 4
+        CFI_REL_OFFSET esi, 0
         pushl %edx
+        CFI_ADJUST_CFA_OFFSET 4
+        CFI_REL_OFFSET edx, 0
         decl %eax                       # eax = -1
         pushl %ecx
+        CFI_ADJUST_CFA_OFFSET 4
+        CFI_REL_OFFSET ecx, 0
         pushl %ebx
+        CFI_ADJUST_CFA_OFFSET 4
+        CFI_REL_OFFSET ebx, 0
         cld
         pushl %es
+        CFI_ADJUST_CFA_OFFSET 4
+        /*CFI_REL_OFFSET es, 0*/
         UNWIND_ESPFIX_STACK
         popl %ecx
+        CFI_ADJUST_CFA_OFFSET -4
+        /*CFI_REGISTER es, ecx*/
         movl ES(%esp), %edi             # get the function address
         movl ORIG_EAX(%esp), %edx       # get the error code
         movl %eax, ORIG_EAX(%esp)
         movl %ecx, ES(%esp)
+        /*CFI_REL_OFFSET es, ES*/
         movl $(__USER_DS), %ecx
         movl %ecx, %ds
         movl %ecx, %es
         movl %esp,%eax                  # pt_regs pointer
         call *%edi
         jmp ret_from_exception
+        CFI_ENDPROC
 
 ENTRY(coprocessor_error)
+        RING0_INT_FRAME
         pushl $0
+        CFI_ADJUST_CFA_OFFSET 4
         pushl $do_coprocessor_error
+        CFI_ADJUST_CFA_OFFSET 4
         jmp error_code
+        CFI_ENDPROC
 
 ENTRY(simd_coprocessor_error)
+        RING0_INT_FRAME
         pushl $0
+        CFI_ADJUST_CFA_OFFSET 4
         pushl $do_simd_coprocessor_error
+        CFI_ADJUST_CFA_OFFSET 4
         jmp error_code
+        CFI_ENDPROC
 
 ENTRY(device_not_available)
+        RING0_INT_FRAME
         pushl $-1                       # mark this as an int
+        CFI_ADJUST_CFA_OFFSET 4
         SAVE_ALL
         movl %cr0, %eax
         testl $0x4, %eax                # EM (math emulation bit)
@@ -487,9 +632,12 @@ ENTRY(device_not_available)
         jmp ret_from_exception
 device_not_available_emulate:
         pushl $0                        # temporary storage for ORIG_EIP
+        CFI_ADJUST_CFA_OFFSET 4
         call math_emulate
         addl $4, %esp
+        CFI_ADJUST_CFA_OFFSET -4
         jmp ret_from_exception
+        CFI_ENDPROC
 
 /*
  * Debug traps and NMI can happen at the one SYSENTER instruction
@@ -514,16 +662,19 @@ label:						\
         pushl $sysenter_past_esp
 
 KPROBE_ENTRY(debug)
+        RING0_INT_FRAME
         cmpl $sysenter_entry,(%esp)
         jne debug_stack_correct
         FIX_STACK(12, debug_stack_correct, debug_esp_fix_insn)
 debug_stack_correct:
         pushl $-1                       # mark this as an int
+        CFI_ADJUST_CFA_OFFSET 4
         SAVE_ALL
         xorl %edx,%edx                  # error code 0
         movl %esp,%eax                  # pt_regs pointer
         call do_debug
         jmp ret_from_exception
+        CFI_ENDPROC
         .previous .text
 /*
  * NMI is doubly nasty. It can happen _while_ we're handling
@@ -534,14 +685,18 @@ debug_stack_correct:
  * fault happened on the sysenter path.
  */
 ENTRY(nmi)
+        RING0_INT_FRAME
         pushl %eax
+        CFI_ADJUST_CFA_OFFSET 4
         movl %ss, %eax
         cmpw $__ESPFIX_SS, %ax
         popl %eax
+        CFI_ADJUST_CFA_OFFSET -4
         je nmi_16bit_stack
         cmpl $sysenter_entry,(%esp)
         je nmi_stack_fixup
         pushl %eax
+        CFI_ADJUST_CFA_OFFSET 4
         movl %esp,%eax
         /* Do not access memory above the end of our stack page,
          * it might not exist.
@@ -549,16 +704,19 @@ ENTRY(nmi)
         andl $(THREAD_SIZE-1),%eax
         cmpl $(THREAD_SIZE-20),%eax
         popl %eax
+        CFI_ADJUST_CFA_OFFSET -4
         jae nmi_stack_correct
         cmpl $sysenter_entry,12(%esp)
         je nmi_debug_stack_check
 nmi_stack_correct:
         pushl %eax
+        CFI_ADJUST_CFA_OFFSET 4
         SAVE_ALL
         xorl %edx,%edx          # zero error code
         movl %esp,%eax          # pt_regs pointer
         call do_nmi
         jmp restore_all
+        CFI_ENDPROC
 
 nmi_stack_fixup:
         FIX_STACK(12,nmi_stack_correct, 1)
@@ -574,94 +732,177 @@ nmi_debug_stack_check:
         jmp nmi_stack_correct
 
 nmi_16bit_stack:
+        RING0_INT_FRAME
         /* create the pointer to lss back */
         pushl %ss
+        CFI_ADJUST_CFA_OFFSET 4
         pushl %esp
+        CFI_ADJUST_CFA_OFFSET 4
         movzwl %sp, %esp
         addw $4, (%esp)
         /* copy the iret frame of 12 bytes */
         .rept 3
         pushl 16(%esp)
+        CFI_ADJUST_CFA_OFFSET 4
         .endr
         pushl %eax
+        CFI_ADJUST_CFA_OFFSET 4
         SAVE_ALL
         FIXUP_ESPFIX_STACK              # %eax == %esp
+        CFI_ADJUST_CFA_OFFSET -20       # the frame has now moved
         xorl %edx,%edx                  # zero error code
         call do_nmi
         RESTORE_REGS
         lss 12+4(%esp), %esp            # back to 16bit stack
 1:      iret
+        CFI_ENDPROC
 .section __ex_table,"a"
         .align 4
         .long 1b,iret_exc
 .previous
 
 KPROBE_ENTRY(int3)
+        RING0_INT_FRAME
         pushl $-1                       # mark this as an int
+        CFI_ADJUST_CFA_OFFSET 4
         SAVE_ALL
         xorl %edx,%edx          # zero error code
         movl %esp,%eax          # pt_regs pointer
         call do_int3
         jmp ret_from_exception
+        CFI_ENDPROC
         .previous .text
 
 ENTRY(overflow)
+        RING0_INT_FRAME
         pushl $0
+        CFI_ADJUST_CFA_OFFSET 4
         pushl $do_overflow
+        CFI_ADJUST_CFA_OFFSET 4
         jmp error_code
+        CFI_ENDPROC
 
 ENTRY(bounds)
+        RING0_INT_FRAME
         pushl $0
+        CFI_ADJUST_CFA_OFFSET 4
         pushl $do_bounds
+        CFI_ADJUST_CFA_OFFSET 4
         jmp error_code
+        CFI_ENDPROC
 
 ENTRY(invalid_op)
+        RING0_INT_FRAME
         pushl $0
+        CFI_ADJUST_CFA_OFFSET 4
         pushl $do_invalid_op
+        CFI_ADJUST_CFA_OFFSET 4
         jmp error_code
+        CFI_ENDPROC
 
 ENTRY(coprocessor_segment_overrun)
+        RING0_INT_FRAME
         pushl $0
+        CFI_ADJUST_CFA_OFFSET 4
         pushl $do_coprocessor_segment_overrun
+        CFI_ADJUST_CFA_OFFSET 4
         jmp error_code
+        CFI_ENDPROC
 
 ENTRY(invalid_TSS)
+        RING0_EC_FRAME
         pushl $do_invalid_TSS
+        CFI_ADJUST_CFA_OFFSET 4
         jmp error_code
+        CFI_ENDPROC
 
 ENTRY(segment_not_present)
+        RING0_EC_FRAME
         pushl $do_segment_not_present
+        CFI_ADJUST_CFA_OFFSET 4
         jmp error_code
+        CFI_ENDPROC
 
 ENTRY(stack_segment)
+        RING0_EC_FRAME
         pushl $do_stack_segment
+        CFI_ADJUST_CFA_OFFSET 4
         jmp error_code
+        CFI_ENDPROC
 
 KPROBE_ENTRY(general_protection)
+        RING0_EC_FRAME
         pushl $do_general_protection
+        CFI_ADJUST_CFA_OFFSET 4
         jmp error_code
+        CFI_ENDPROC
         .previous .text
 
 ENTRY(alignment_check)
+        RING0_EC_FRAME
         pushl $do_alignment_check
+        CFI_ADJUST_CFA_OFFSET 4
         jmp error_code
+        CFI_ENDPROC
 
 KPROBE_ENTRY(page_fault)
+        RING0_EC_FRAME
         pushl $do_page_fault
+        CFI_ADJUST_CFA_OFFSET 4
         jmp error_code
+        CFI_ENDPROC
         .previous .text
 
 #ifdef CONFIG_X86_MCE
 ENTRY(machine_check)
+        RING0_INT_FRAME
         pushl $0
+        CFI_ADJUST_CFA_OFFSET 4
         pushl machine_check_vector
+        CFI_ADJUST_CFA_OFFSET 4
         jmp error_code
+        CFI_ENDPROC
 #endif
 
 ENTRY(spurious_interrupt_bug)
+        RING0_INT_FRAME
         pushl $0
+        CFI_ADJUST_CFA_OFFSET 4
         pushl $do_spurious_interrupt_bug
+        CFI_ADJUST_CFA_OFFSET 4
         jmp error_code
+        CFI_ENDPROC
+
+#ifdef CONFIG_STACK_UNWIND
+ENTRY(arch_unwind_init_running)
+        CFI_STARTPROC
+        movl    4(%esp), %edx
+        movl    (%esp), %ecx
+        leal    4(%esp), %eax
+        movl    %ebx, EBX(%edx)
+        xorl    %ebx, %ebx
+        movl    %ebx, ECX(%edx)
+        movl    %ebx, EDX(%edx)
+        movl    %esi, ESI(%edx)
+        movl    %edi, EDI(%edx)
+        movl    %ebp, EBP(%edx)
+        movl    %ebx, EAX(%edx)
+        movl    $__USER_DS, DS(%edx)
+        movl    $__USER_DS, ES(%edx)
+        movl    %ebx, ORIG_EAX(%edx)
+        movl    %ecx, EIP(%edx)
+        movl    12(%esp), %ecx
+        movl    $__KERNEL_CS, CS(%edx)
+        movl    %ebx, EFLAGS(%edx)
+        movl    %eax, OLDESP(%edx)
+        movl    8(%esp), %eax
+        movl    %ecx, 8(%esp)
+        movl    EBX(%edx), %ebx
+        movl    $__KERNEL_DS, OLDSS(%edx)
+        jmpl    *%eax
+        CFI_ENDPROC
+ENDPROC(arch_unwind_init_running)
+#endif
 
 .section .rodata,"a"
 #include "syscall_table.S"
diff --git a/arch/i386/kernel/io_apic.c b/arch/i386/kernel/io_apic.c
index a62df3e764c5..72ae414e4d49 100644
--- a/arch/i386/kernel/io_apic.c
+++ b/arch/i386/kernel/io_apic.c
@@ -38,6 +38,7 @@
 #include <asm/desc.h>
 #include <asm/timer.h>
 #include <asm/i8259.h>
+#include <asm/nmi.h>
 
 #include <mach_apic.h>
 
@@ -50,6 +51,7 @@ atomic_t irq_mis_count;
 static struct { int pin, apic; } ioapic_i8259 = { -1, -1 };
 
 static DEFINE_SPINLOCK(ioapic_lock);
+static DEFINE_SPINLOCK(vector_lock);
 
 int timer_over_8254 __initdata = 1;
 
@@ -1161,10 +1163,17 @@ u8 irq_vector[NR_IRQ_VECTORS] __read_mostly = { FIRST_DEVICE_VECTOR , 0 };
 int assign_irq_vector(int irq)
 {
         static int current_vector = FIRST_DEVICE_VECTOR, offset = 0;
+        unsigned long flags;
+        int vector;
+
+        BUG_ON(irq != AUTO_ASSIGN && (unsigned)irq >= NR_IRQ_VECTORS);
 
-        BUG_ON(irq >= NR_IRQ_VECTORS);
-        if (irq != AUTO_ASSIGN && IO_APIC_VECTOR(irq) > 0)
+        spin_lock_irqsave(&vector_lock, flags);
+
+        if (irq != AUTO_ASSIGN && IO_APIC_VECTOR(irq) > 0) {
+                spin_unlock_irqrestore(&vector_lock, flags);
                 return IO_APIC_VECTOR(irq);
+        }
 next:
         current_vector += 8;
         if (current_vector == SYSCALL_VECTOR)
@@ -1172,16 +1181,21 @@ next:
 
         if (current_vector >= FIRST_SYSTEM_VECTOR) {
                 offset++;
-                if (!(offset%8))
+                if (!(offset%8)) {
+                        spin_unlock_irqrestore(&vector_lock, flags);
                         return -ENOSPC;
+                }
                 current_vector = FIRST_DEVICE_VECTOR + offset;
         }
 
-        vector_irq[current_vector] = irq;
+        vector = current_vector;
+        vector_irq[vector] = irq;
         if (irq != AUTO_ASSIGN)
-                IO_APIC_VECTOR(irq) = current_vector;
+                IO_APIC_VECTOR(irq) = vector;
 
-        return current_vector;
+        spin_unlock_irqrestore(&vector_lock, flags);
+
+        return vector;
 }
 
 static struct hw_interrupt_type ioapic_level_type;
@@ -1193,21 +1207,14 @@ static struct hw_interrupt_type ioapic_edge_type;
 
 static inline void ioapic_register_intr(int irq, int vector, unsigned long trigger)
 {
-        if (use_pci_vector() && !platform_legacy_irq(irq)) {
-                if ((trigger == IOAPIC_AUTO && IO_APIC_irq_trigger(irq)) ||
-                                trigger == IOAPIC_LEVEL)
-                        irq_desc[vector].handler = &ioapic_level_type;
-                else
-                        irq_desc[vector].handler = &ioapic_edge_type;
-                set_intr_gate(vector, interrupt[vector]);
-        } else  {
-                if ((trigger == IOAPIC_AUTO && IO_APIC_irq_trigger(irq)) ||
-                                trigger == IOAPIC_LEVEL)
-                        irq_desc[irq].handler = &ioapic_level_type;
-                else
-                        irq_desc[irq].handler = &ioapic_edge_type;
-                set_intr_gate(vector, interrupt[irq]);
-        }
+        unsigned idx = use_pci_vector() && !platform_legacy_irq(irq) ? vector : irq;
+
+        if ((trigger == IOAPIC_AUTO && IO_APIC_irq_trigger(irq)) ||
+                        trigger == IOAPIC_LEVEL)
+                irq_desc[idx].handler = &ioapic_level_type;
+        else
+                irq_desc[idx].handler = &ioapic_edge_type;
+        set_intr_gate(vector, interrupt[idx]);
 }
 
 static void __init setup_IO_APIC_irqs(void)
diff --git a/arch/i386/kernel/irq.c b/arch/i386/kernel/irq.c
index 49ce4c31b713..061533e0cb5e 100644
--- a/arch/i386/kernel/irq.c
+++ b/arch/i386/kernel/irq.c
@@ -227,7 +227,7 @@ int show_interrupts(struct seq_file *p, void *v)
         if (i == 0) {
                 seq_printf(p, "           ");
                 for_each_online_cpu(j)
-                        seq_printf(p, "CPU%d       ",j);
+                        seq_printf(p, "CPU%-8d",j);
                 seq_putc(p, '\n');
         }
 
diff --git a/arch/i386/kernel/nmi.c b/arch/i386/kernel/nmi.c
index d43b498ec745..a76e93146585 100644
--- a/arch/i386/kernel/nmi.c
+++ b/arch/i386/kernel/nmi.c
@@ -14,21 +14,17 @@
  */
 
 #include <linux/config.h>
-#include <linux/mm.h>
 #include <linux/delay.h>
-#include <linux/bootmem.h>
-#include <linux/smp_lock.h>
 #include <linux/interrupt.h>
-#include <linux/mc146818rtc.h>
-#include <linux/kernel_stat.h>
 #include <linux/module.h>
 #include <linux/nmi.h>
 #include <linux/sysdev.h>
 #include <linux/sysctl.h>
+#include <linux/percpu.h>
 
 #include <asm/smp.h>
-#include <asm/div64.h>
 #include <asm/nmi.h>
+#include <asm/intel_arch_perfmon.h>
 
 #include "mach_traps.h"
 
@@ -100,6 +96,9 @@ int nmi_active;
         (P4_CCCR_OVF_PMI0|P4_CCCR_THRESHOLD(15)|P4_CCCR_COMPLEMENT|     \
          P4_CCCR_COMPARE|P4_CCCR_REQUIRED|P4_CCCR_ESCR_SELECT(4)|P4_CCCR_ENABLE)
 
+#define ARCH_PERFMON_NMI_EVENT_SEL      ARCH_PERFMON_UNHALTED_CORE_CYCLES_SEL
+#define ARCH_PERFMON_NMI_EVENT_UMASK    ARCH_PERFMON_UNHALTED_CORE_CYCLES_UMASK
+
 #ifdef CONFIG_SMP
 /* The performance counters used by NMI_LOCAL_APIC don't trigger when
  * the CPU is idle. To make sure the NMI watchdog really ticks on all
@@ -212,6 +211,8 @@ static int __init setup_nmi_watchdog(char *str)
 
 __setup("nmi_watchdog=", setup_nmi_watchdog);
 
+static void disable_intel_arch_watchdog(void);
+
 static void disable_lapic_nmi_watchdog(void)
 {
         if (nmi_active <= 0)
@@ -221,6 +222,10 @@ static void disable_lapic_nmi_watchdog(void)
                 wrmsr(MSR_K7_EVNTSEL0, 0, 0);
                 break;
         case X86_VENDOR_INTEL:
+                if (cpu_has(&boot_cpu_data, X86_FEATURE_ARCH_PERFMON)) {
+                        disable_intel_arch_watchdog();
+                        break;
+                }
                 switch (boot_cpu_data.x86) {
                 case 6:
                         if (boot_cpu_data.x86_model > 0xd)
@@ -449,6 +454,53 @@ static int setup_p4_watchdog(void)
         return 1;
 }
 
+static void disable_intel_arch_watchdog(void)
+{
+        unsigned ebx;
+
+        /*
+         * Check whether the Architectural PerfMon supports
+         * Unhalted Core Cycles Event or not.
+         * NOTE: Corresponding bit = 0 in ebp indicates event present.
+         */
+        ebx = cpuid_ebx(10);
+        if (!(ebx & ARCH_PERFMON_UNHALTED_CORE_CYCLES_PRESENT))
+                wrmsr(MSR_ARCH_PERFMON_EVENTSEL0, 0, 0);
+}
+
+static int setup_intel_arch_watchdog(void)
+{
+        unsigned int evntsel;
+        unsigned ebx;
+
+        /*
+         * Check whether the Architectural PerfMon supports
+         * Unhalted Core Cycles Event or not.
+         * NOTE: Corresponding bit = 0 in ebp indicates event present.
+         */
+        ebx = cpuid_ebx(10);
+        if ((ebx & ARCH_PERFMON_UNHALTED_CORE_CYCLES_PRESENT))
+                return 0;
+
+        nmi_perfctr_msr = MSR_ARCH_PERFMON_PERFCTR0;
+
+        clear_msr_range(MSR_ARCH_PERFMON_EVENTSEL0, 2);
+        clear_msr_range(MSR_ARCH_PERFMON_PERFCTR0, 2);
+
+        evntsel = ARCH_PERFMON_EVENTSEL_INT
+                | ARCH_PERFMON_EVENTSEL_OS
+                | ARCH_PERFMON_EVENTSEL_USR
+                | ARCH_PERFMON_NMI_EVENT_SEL
+                | ARCH_PERFMON_NMI_EVENT_UMASK;
+
+        wrmsr(MSR_ARCH_PERFMON_EVENTSEL0, evntsel, 0);
+        write_watchdog_counter("INTEL_ARCH_PERFCTR0");
+        apic_write(APIC_LVTPC, APIC_DM_NMI);
+        evntsel |= ARCH_PERFMON_EVENTSEL0_ENABLE;
+        wrmsr(MSR_ARCH_PERFMON_EVENTSEL0, evntsel, 0);
+        return 1;
+}
+
 void setup_apic_nmi_watchdog (void)
 {
         switch (boot_cpu_data.x86_vendor) {
@@ -458,6 +510,11 @@ void setup_apic_nmi_watchdog (void)
                 setup_k7_watchdog();
                 break;
         case X86_VENDOR_INTEL:
+                if (cpu_has(&boot_cpu_data, X86_FEATURE_ARCH_PERFMON)) {
+                        if (!setup_intel_arch_watchdog())
+                                return;
+                        break;
+                }
                 switch (boot_cpu_data.x86) {
                 case 6:
                         if (boot_cpu_data.x86_model > 0xd)
@@ -561,7 +618,8 @@ void nmi_watchdog_tick (struct pt_regs * regs)
                         wrmsr(MSR_P4_IQ_CCCR0, nmi_p4_cccr_val, 0);
                         apic_write(APIC_LVTPC, APIC_DM_NMI);
                 }
-                else if (nmi_perfctr_msr == MSR_P6_PERFCTR0) {
+                else if (nmi_perfctr_msr == MSR_P6_PERFCTR0 ||
+                         nmi_perfctr_msr == MSR_ARCH_PERFMON_PERFCTR0) {
                         /* Only P6 based Pentium M need to re-unmask
                          * the apic vector but it doesn't hurt
                          * other P6 variant */
diff --git a/arch/i386/kernel/process.c b/arch/i386/kernel/process.c
index 6259afea46d1..6946b06e2784 100644
--- a/arch/i386/kernel/process.c
+++ b/arch/i386/kernel/process.c
@@ -102,7 +102,7 @@ void default_idle(void)
         local_irq_enable();
 
         if (!hlt_counter && boot_cpu_data.hlt_works_ok) {
-                clear_thread_flag(TIF_POLLING_NRFLAG);
+                current_thread_info()->status &= ~TS_POLLING;
                 smp_mb__after_clear_bit();
                 while (!need_resched()) {
                         local_irq_disable();
@@ -111,7 +111,7 @@ void default_idle(void)
                         else
                                 local_irq_enable();
                 }
-                set_thread_flag(TIF_POLLING_NRFLAG);
+                current_thread_info()->status |= TS_POLLING;
         } else {
                 while (!need_resched())
                         cpu_relax();
@@ -174,7 +174,7 @@ void cpu_idle(void)
 {
         int cpu = smp_processor_id();
 
-        set_thread_flag(TIF_POLLING_NRFLAG);
+        current_thread_info()->status |= TS_POLLING;
 
         /* endless idle loop with no priority at all */
         while (1) {
@@ -312,7 +312,7 @@ void show_regs(struct pt_regs * regs)
         cr3 = read_cr3();
         cr4 = read_cr4_safe();
         printk("CR0: %08lx CR2: %08lx CR3: %08lx CR4: %08lx\n", cr0, cr2, cr3, cr4);
-        show_trace(NULL, &regs->esp);
+        show_trace(NULL, regs, &regs->esp);
 }
 
 /*
diff --git a/arch/i386/kernel/smp.c b/arch/i386/kernel/smp.c
index d134e9643a58..c10789d7a9d3 100644
--- a/arch/i386/kernel/smp.c
+++ b/arch/i386/kernel/smp.c
@@ -114,7 +114,17 @@ DEFINE_PER_CPU(struct tlb_state, cpu_tlbstate) ____cacheline_aligned = { &init_m
 
 static inline int __prepare_ICR (unsigned int shortcut, int vector)
 {
-        return APIC_DM_FIXED | shortcut | vector | APIC_DEST_LOGICAL;
+        unsigned int icr = shortcut | APIC_DEST_LOGICAL;
+
+        switch (vector) {
+        default:
+                icr |= APIC_DM_FIXED | vector;
+                break;
+        case NMI_VECTOR:
+                icr |= APIC_DM_NMI;
+                break;
+        }
+        return icr;
 }
 
 static inline int __prepare_ICR2 (unsigned int mask)
diff --git a/arch/i386/kernel/smpboot.c b/arch/i386/kernel/smpboot.c
index bd0ca5c9f053..bce5470ecb42 100644
--- a/arch/i386/kernel/smpboot.c
+++ b/arch/i386/kernel/smpboot.c
@@ -52,6 +52,7 @@
 #include <asm/tlbflush.h>
 #include <asm/desc.h>
 #include <asm/arch_hooks.h>
+#include <asm/nmi.h>
 
 #include <mach_apic.h>
 #include <mach_wakecpu.h>
diff --git a/arch/i386/kernel/traps.c b/arch/i386/kernel/traps.c
index dcc14477af1f..78464097470a 100644
--- a/arch/i386/kernel/traps.c
+++ b/arch/i386/kernel/traps.c
@@ -28,6 +28,7 @@
 #include <linux/utsname.h>
 #include <linux/kprobes.h>
 #include <linux/kexec.h>
+#include <linux/unwind.h>
 
 #ifdef CONFIG_EISA
 #include <linux/ioport.h>
@@ -47,7 +48,7 @@
 #include <asm/desc.h>
 #include <asm/i387.h>
 #include <asm/nmi.h>
-
+#include <asm/unwind.h>
 #include <asm/smp.h>
 #include <asm/arch_hooks.h>
 #include <asm/kdebug.h>
@@ -92,6 +93,7 @@ asmlinkage void spurious_interrupt_bug(void);
 asmlinkage void machine_check(void);
 
 static int kstack_depth_to_print = 24;
+static int call_trace = 1;
 ATOMIC_NOTIFIER_HEAD(i386die_chain);
 
 int register_die_notifier(struct notifier_block *nb)
@@ -170,7 +172,23 @@ static inline unsigned long print_context_stack(struct thread_info *tinfo,
         return ebp;
 }
 
-static void show_trace_log_lvl(struct task_struct *task,
+static asmlinkage int show_trace_unwind(struct unwind_frame_info *info, void *log_lvl)
+{
+        int n = 0;
+        int printed = 0; /* nr of entries already printed on current line */
+
+        while (unwind(info) == 0 && UNW_PC(info)) {
+                ++n;
+                printed = print_addr_and_symbol(UNW_PC(info), log_lvl, printed);
+                if (arch_unw_user_mode(info))
+                        break;
+        }
+        if (printed)
+                printk("\n");
+        return n;
+}
+
+static void show_trace_log_lvl(struct task_struct *task, struct pt_regs *regs,
                                unsigned long *stack, char *log_lvl)
 {
         unsigned long ebp;
@@ -178,6 +196,26 @@ static void show_trace_log_lvl(struct task_struct *task,
         if (!task)
                 task = current;
 
+        if (call_trace >= 0) {
+                int unw_ret = 0;
+                struct unwind_frame_info info;
+
+                if (regs) {
+                        if (unwind_init_frame_info(&info, task, regs) == 0)
+                                unw_ret = show_trace_unwind(&info, log_lvl);
+                } else if (task == current)
+                        unw_ret = unwind_init_running(&info, show_trace_unwind, log_lvl);
+                else {
+                        if (unwind_init_blocked(&info, task) == 0)
+                                unw_ret = show_trace_unwind(&info, log_lvl);
+                }
+                if (unw_ret > 0) {
+                        if (call_trace > 0)
+                                return;
+                        printk("%sLegacy call trace:\n", log_lvl);
+                }
+        }
+
         if (task == current) {
                 /* Grab ebp right from our regs */
                 asm ("movl %%ebp, %0" : "=r" (ebp) : );
@@ -198,13 +236,13 @@ static void show_trace_log_lvl(struct task_struct *task,
         }
 }
 
-void show_trace(struct task_struct *task, unsigned long * stack)
+void show_trace(struct task_struct *task, struct pt_regs *regs, unsigned long * stack)
 {
-        show_trace_log_lvl(task, stack, "");
+        show_trace_log_lvl(task, regs, stack, "");
 }
 
-static void show_stack_log_lvl(struct task_struct *task, unsigned long *esp,
-                               char *log_lvl)
+static void show_stack_log_lvl(struct task_struct *task, struct pt_regs *regs,
+                               unsigned long *esp, char *log_lvl)
 {
         unsigned long *stack;
         int i;
@@ -225,13 +263,13 @@ static void show_stack_log_lvl(struct task_struct *task, unsigned long *esp,
                 printk("%08lx ", *stack++);
         }
         printk("\n%sCall Trace:\n", log_lvl);
-        show_trace_log_lvl(task, esp, log_lvl);
+        show_trace_log_lvl(task, regs, esp, log_lvl);
 }
 
 void show_stack(struct task_struct *task, unsigned long *esp)
 {
         printk("       ");
-        show_stack_log_lvl(task, esp, "");
+        show_stack_log_lvl(task, NULL, esp, "");
 }
 
 /*
@@ -241,7 +279,7 @@ void dump_stack(void)
 {
         unsigned long stack;
 
-        show_trace(current, &stack);
+        show_trace(current, NULL, &stack);
 }
 
 EXPORT_SYMBOL(dump_stack);
@@ -285,7 +323,7 @@ void show_registers(struct pt_regs *regs)
                 u8 __user *eip;
 
                 printk("\n" KERN_EMERG "Stack: ");
-                show_stack_log_lvl(NULL, (unsigned long *)esp, KERN_EMERG);
+                show_stack_log_lvl(NULL, regs, (unsigned long *)esp, KERN_EMERG);
 
                 printk(KERN_EMERG "Code: ");
 
@@ -1215,3 +1253,15 @@ static int __init kstack_setup(char *s)
         return 1;
 }
 __setup("kstack=", kstack_setup);
+
+static int __init call_trace_setup(char *s)
+{
+        if (strcmp(s, "old") == 0)
+                call_trace = -1;
+        else if (strcmp(s, "both") == 0)
+                call_trace = 0;
+        else if (strcmp(s, "new") == 0)
+                call_trace = 1;
+        return 1;
+}
+__setup("call_trace=", call_trace_setup);
diff --git a/arch/i386/kernel/vmlinux.lds.S b/arch/i386/kernel/vmlinux.lds.S
index 7512f39c9f25..2d4f1386e2b1 100644
--- a/arch/i386/kernel/vmlinux.lds.S
+++ b/arch/i386/kernel/vmlinux.lds.S
@@ -71,6 +71,15 @@ SECTIONS
   .data.read_mostly : AT(ADDR(.data.read_mostly) - LOAD_OFFSET) { *(.data.read_mostly) }
   _edata = .;                   /* End of data section */
 
+#ifdef CONFIG_STACK_UNWIND
+  . = ALIGN(4);
+  .eh_frame : AT(ADDR(.eh_frame) - LOAD_OFFSET) {
+        __start_unwind = .;
+        *(.eh_frame)
+        __end_unwind = .;
+  }
+#endif
+
   . = ALIGN(THREAD_SIZE);       /* init_task */
   .data.init_task : AT(ADDR(.data.init_task) - LOAD_OFFSET) {
         *(.data.init_task)
diff --git a/arch/i386/oprofile/op_model_athlon.c b/arch/i386/oprofile/op_model_athlon.c
index 3ad9a72a5036..693bdea4a52b 100644
--- a/arch/i386/oprofile/op_model_athlon.c
+++ b/arch/i386/oprofile/op_model_athlon.c
@@ -13,6 +13,7 @@
 #include <linux/oprofile.h>
 #include <asm/ptrace.h>
 #include <asm/msr.h>
+#include <asm/nmi.h>
  
 #include "op_x86_model.h"
 #include "op_counter.h"
diff --git a/arch/i386/oprofile/op_model_p4.c b/arch/i386/oprofile/op_model_p4.c
index ac8a066035c2..7c61d357b82b 100644
--- a/arch/i386/oprofile/op_model_p4.c
+++ b/arch/i386/oprofile/op_model_p4.c
@@ -14,6 +14,7 @@
 #include <asm/ptrace.h>
 #include <asm/fixmap.h>
 #include <asm/apic.h>
+#include <asm/nmi.h>
 
 #include "op_x86_model.h"
 #include "op_counter.h"
diff --git a/arch/i386/oprofile/op_model_ppro.c b/arch/i386/oprofile/op_model_ppro.c
index d719015fc044..5c3ab4b027ad 100644
--- a/arch/i386/oprofile/op_model_ppro.c
+++ b/arch/i386/oprofile/op_model_ppro.c
@@ -14,6 +14,7 @@
 #include <asm/ptrace.h>
 #include <asm/msr.h>
 #include <asm/apic.h>
+#include <asm/nmi.h>
  
 #include "op_x86_model.h"
 #include "op_counter.h"