33 files changed, 305 insertions, 268 deletions

diff --git a/include/asm-x86_64/Kbuild b/include/asm-x86_64/Kbuild
index 1ee9b07f3fe6..ebd7117782a6 100644
--- a/include/asm-x86_64/Kbuild
+++ b/include/asm-x86_64/Kbuild
@@ -6,13 +6,11 @@ ALTARCHDEF := defined __i386__
 
 header-y += boot.h
 header-y += bootsetup.h
-header-y += cpufeature.h
 header-y += debugreg.h
 header-y += ldt.h
 header-y += msr.h
 header-y += prctl.h
 header-y += ptrace-abi.h
-header-y += setup.h
 header-y += sigcontext32.h
 header-y += ucontext.h
 header-y += vsyscall32.h
diff --git a/include/asm-x86_64/acpi.h b/include/asm-x86_64/acpi.h
index ed59aa4c6ff9..9d1916e59c04 100644
--- a/include/asm-x86_64/acpi.h
+++ b/include/asm-x86_64/acpi.h
@@ -163,6 +163,7 @@ extern u8 x86_acpiid_to_apicid[];
 #define ARCH_HAS_POWER_INIT 1
 
 extern int acpi_skip_timer_override;
+extern int acpi_use_timer_override;
 
 #endif /*__KERNEL__*/
 
diff --git a/include/asm-x86_64/alternative.h b/include/asm-x86_64/alternative.h
index a584826cc570..a6657b4f3e0e 100644
--- a/include/asm-x86_64/alternative.h
+++ b/include/asm-x86_64/alternative.h
@@ -4,6 +4,7 @@
 #ifdef __KERNEL__
 
 #include <linux/types.h>
+#include <linux/stddef.h>
 #include <asm/cpufeature.h>
 
 struct alt_instr {
@@ -133,4 +134,15 @@ static inline void alternatives_smp_switch(int smp) {}
 #define LOCK_PREFIX ""
 #endif
 
+struct paravirt_patch;
+#ifdef CONFIG_PARAVIRT
+void apply_paravirt(struct paravirt_patch *start, struct paravirt_patch *end);
+#else
+static inline void
+apply_paravirt(struct paravirt_patch *start, struct paravirt_patch *end)
+{}
+#define __start_parainstructions NULL
+#define __stop_parainstructions NULL
+#endif
+
 #endif /* _X86_64_ALTERNATIVE_H */
diff --git a/include/asm-x86_64/atomic.h b/include/asm-x86_64/atomic.h
index 007e88d6d43f..706ca4b60000 100644
--- a/include/asm-x86_64/atomic.h
+++ b/include/asm-x86_64/atomic.h
@@ -21,7 +21,7 @@
  * on us. We need to use _exactly_ the address the user gave us,
  * not some alias that contains the same information.
  */
-typedef struct { volatile int counter; } atomic_t;
+typedef struct { int counter; } atomic_t;
 
 #define ATOMIC_INIT(i)  { (i) }
 
@@ -189,9 +189,9 @@ static __inline__ int atomic_add_return(int i, atomic_t *v)
 {
         int __i = i;
         __asm__ __volatile__(
-                LOCK_PREFIX "xaddl %0, %1;"
-                :"=r"(i)
-                :"m"(v->counter), "0"(i));
+                LOCK_PREFIX "xaddl %0, %1"
+                :"+r" (i), "+m" (v->counter)
+                : : "memory");
         return i + __i;
 }
 
diff --git a/include/asm-x86_64/calgary.h b/include/asm-x86_64/calgary.h
index 6b93f5a3a5c8..7ee900645719 100644
--- a/include/asm-x86_64/calgary.h
+++ b/include/asm-x86_64/calgary.h
@@ -51,6 +51,8 @@ struct iommu_table {
 #define TCE_TABLE_SIZE_4M               6
 #define TCE_TABLE_SIZE_8M               7
 
+extern int use_calgary;
+
 #ifdef CONFIG_CALGARY_IOMMU
 extern int calgary_iommu_init(void);
 extern void detect_calgary(void);
diff --git a/include/asm-x86_64/checksum.h b/include/asm-x86_64/checksum.h
index 989469e8e0b7..419fe88a0342 100644
--- a/include/asm-x86_64/checksum.h
+++ b/include/asm-x86_64/checksum.h
@@ -19,15 +19,16 @@
  * the last step before putting a checksum into a packet.
  * Make sure not to mix with 64bit checksums.
  */
-static inline unsigned int csum_fold(unsigned int sum)
+static inline __sum16 csum_fold(__wsum sum)
 {
         __asm__(
                 "  addl %1,%0\n"
                 "  adcl $0xffff,%0"
                 : "=r" (sum)
-                : "r" (sum << 16), "0" (sum & 0xffff0000)
+                : "r" ((__force u32)sum << 16),
+                  "0" ((__force u32)sum & 0xffff0000)
         );
-        return (~sum) >> 16;
+        return (__force __sum16)(~(__force u32)sum >> 16);
 }
 
 /*
@@ -43,7 +44,7 @@ static inline unsigned int csum_fold(unsigned int sum)
  * iph: ipv4 header
  * ihl: length of header / 4
  */ 
-static inline unsigned short ip_fast_csum(unsigned char *iph, unsigned int ihl) 
+static inline __sum16 ip_fast_csum(const void *iph, unsigned int ihl)
 {
         unsigned int sum;
 
@@ -70,7 +71,7 @@ static inline unsigned short ip_fast_csum(unsigned char *iph, unsigned int ihl)
         : "=r" (sum), "=r" (iph), "=r" (ihl)
         : "1" (iph), "2" (ihl)
         : "memory");
-        return(sum);
+        return (__force __sum16)sum;
 }
 
 /** 
@@ -84,16 +85,17 @@ static inline unsigned short ip_fast_csum(unsigned char *iph, unsigned int ihl)
  * Returns the pseudo header checksum the input data. Result is 
  * 32bit unfolded.
  */
-static inline unsigned long 
-csum_tcpudp_nofold(unsigned saddr, unsigned daddr, unsigned short len,
-                   unsigned short proto, unsigned int sum) 
+static inline __wsum
+csum_tcpudp_nofold(__be32 saddr, __be32 daddr, unsigned short len,
+                   unsigned short proto, __wsum sum)
 {
         asm("  addl %1, %0\n"
             "  adcl %2, %0\n"
             "  adcl %3, %0\n"
             "  adcl $0, %0\n"
                 : "=r" (sum)
-            : "g" (daddr), "g" (saddr), "g" ((ntohs(len)<<16)+proto*256), "0" (sum));
+            : "g" (daddr), "g" (saddr),
+              "g" ((len + proto)<<8), "0" (sum));
     return sum;
 }
 
@@ -109,9 +111,9 @@ csum_tcpudp_nofold(unsigned saddr, unsigned daddr, unsigned short len,
  * Returns the 16bit pseudo header checksum the input data already
  * complemented and ready to be filled in.
  */
-static inline unsigned short int 
-csum_tcpudp_magic(unsigned long saddr, unsigned long daddr,
-                  unsigned short len, unsigned short proto, unsigned int sum) 
+static inline __sum16
+csum_tcpudp_magic(__be32 saddr, __be32 daddr,
+                  unsigned short len, unsigned short proto, __wsum sum)
 {
         return csum_fold(csum_tcpudp_nofold(saddr,daddr,len,proto,sum));
 }
@@ -126,25 +128,25 @@ csum_tcpudp_magic(unsigned long saddr, unsigned long daddr,
  * Before filling it in it needs to be csum_fold()'ed.
  * buff should be aligned to a 64bit boundary if possible.
  */ 
-extern unsigned int csum_partial(const unsigned char *buff, unsigned len, unsigned int sum);
+extern __wsum csum_partial(const void *buff, int len, __wsum sum);
 
 #define  _HAVE_ARCH_COPY_AND_CSUM_FROM_USER 1
 #define HAVE_CSUM_COPY_USER 1
 
 
 /* Do not call this directly. Use the wrappers below */
-extern unsigned long csum_partial_copy_generic(const unsigned char *src, const unsigned char *dst,
-                                               unsigned len,
-                                               unsigned sum, 
+extern __wsum csum_partial_copy_generic(const void *src, const void *dst,
+                                               int len,
+                                               __wsum sum,
                                                int *src_err_ptr, int *dst_err_ptr);
 
 
-extern unsigned int csum_partial_copy_from_user(const unsigned char __user *src, unsigned char *dst,
-                                       int len, unsigned int isum, int *errp);
-extern unsigned int csum_partial_copy_to_user(const unsigned char *src, unsigned char __user *dst,
-                                      int len, unsigned int isum, int *errp);
-extern unsigned int csum_partial_copy_nocheck(const unsigned char *src, unsigned char *dst, int len,
-                                              unsigned int sum);
+extern __wsum csum_partial_copy_from_user(const void __user *src, void *dst,
+                                       int len, __wsum isum, int *errp);
+extern __wsum csum_partial_copy_to_user(const void *src, void __user *dst,
+                                      int len, __wsum isum, int *errp);
+extern __wsum csum_partial_copy_nocheck(const void *src, void *dst, int len,
+                                              __wsum sum);
 
 /* Old names. To be removed. */
 #define csum_and_copy_to_user csum_partial_copy_to_user
@@ -158,7 +160,7 @@ extern unsigned int csum_partial_copy_nocheck(const unsigned char *src, unsigned
  * Returns the 16bit folded/inverted checksum of the passed buffer.
  * Ready to fill in.
  */
-extern unsigned short ip_compute_csum(unsigned char * buff, int len);
+extern __sum16 ip_compute_csum(const void *buff, int len);
 
 /**
  * csum_ipv6_magic - Compute checksum of an IPv6 pseudo header.
@@ -176,9 +178,9 @@ extern unsigned short ip_compute_csum(unsigned char * buff, int len);
 struct in6_addr;
 
 #define _HAVE_ARCH_IPV6_CSUM 1
-extern unsigned short 
-csum_ipv6_magic(struct in6_addr *saddr, struct in6_addr *daddr,
-                __u32 len, unsigned short proto, unsigned int sum);
+extern __sum16
+csum_ipv6_magic(const struct in6_addr *saddr, const struct in6_addr *daddr,
+                __u32 len, unsigned short proto, __wsum sum);
 
 static inline unsigned add32_with_carry(unsigned a, unsigned b)
 {
diff --git a/include/asm-x86_64/cpufeature.h b/include/asm-x86_64/cpufeature.h
index ee792faaca01..0b3c686139f1 100644
--- a/include/asm-x86_64/cpufeature.h
+++ b/include/asm-x86_64/cpufeature.h
@@ -29,7 +29,7 @@
 #define X86_FEATURE_PSE36       (0*32+17) /* 36-bit PSEs */
 #define X86_FEATURE_PN          (0*32+18) /* Processor serial number */
 #define X86_FEATURE_CLFLSH      (0*32+19) /* Supports the CLFLUSH instruction */
-#define X86_FEATURE_DTES        (0*32+21) /* Debug Trace Store */
+#define X86_FEATURE_DS          (0*32+21) /* Debug Store */
 #define X86_FEATURE_ACPI        (0*32+22) /* ACPI via MSR */
 #define X86_FEATURE_MMX         (0*32+23) /* Multimedia Extensions */
 #define X86_FEATURE_FXSR        (0*32+24) /* FXSAVE and FXRSTOR instructions (fast save and restore */
@@ -68,6 +68,8 @@
 #define X86_FEATURE_FXSAVE_LEAK (3*32+7)  /* FIP/FOP/FDP leaks through FXSAVE */
 #define X86_FEATURE_UP          (3*32+8) /* SMP kernel running on UP */
 #define X86_FEATURE_ARCH_PERFMON (3*32+9) /* Intel Architectural PerfMon */
+#define X86_FEATURE_PEBS        (3*32+10) /* Precise-Event Based Sampling */
+#define X86_FEATURE_BTS         (3*32+11) /* Branch Trace Store */
 
 /* Intel-defined CPU features, CPUID level 0x00000001 (ecx), word 4 */
 #define X86_FEATURE_XMM3        (4*32+ 0) /* Streaming SIMD Extensions-3 */
@@ -112,5 +114,8 @@
 #define cpu_has_cyrix_arr      0
 #define cpu_has_centaur_mcr    0
 #define cpu_has_clflush        boot_cpu_has(X86_FEATURE_CLFLSH)
+#define cpu_has_ds             boot_cpu_has(X86_FEATURE_DS)
+#define cpu_has_pebs           boot_cpu_has(X86_FEATURE_PEBS)
+#define cpu_has_bts            boot_cpu_has(X86_FEATURE_BTS)
 
 #endif /* __ASM_X8664_CPUFEATURE_H */
diff --git a/include/asm-x86_64/delay.h b/include/asm-x86_64/delay.h
index 65f64acc5319..c2669f1f5529 100644
--- a/include/asm-x86_64/delay.h
+++ b/include/asm-x86_64/delay.h
@@ -7,18 +7,21 @@
  * Delay routines calling functions in arch/x86_64/lib/delay.c
  */
  
+/* Undefined functions to get compile-time errors */
 extern void __bad_udelay(void);
 extern void __bad_ndelay(void);
 
 extern void __udelay(unsigned long usecs);
-extern void __ndelay(unsigned long usecs);
+extern void __ndelay(unsigned long nsecs);
 extern void __const_udelay(unsigned long usecs);
 extern void __delay(unsigned long loops);
 
+/* 0x10c7 is 2**32 / 1000000 (rounded up) */
 #define udelay(n) (__builtin_constant_p(n) ? \
-        ((n) > 20000 ? __bad_udelay() : __const_udelay((n) * 0x10c6ul)) : \
+        ((n) > 20000 ? __bad_udelay() : __const_udelay((n) * 0x10c7ul)) : \
         __udelay(n))
 
+/* 0x5 is 2**32 / 1000000000 (rounded up) */
 #define ndelay(n) (__builtin_constant_p(n) ? \
        ((n) > 20000 ? __bad_ndelay() : __const_udelay((n) * 5ul)) : \
        __ndelay(n))
diff --git a/include/asm-x86_64/desc.h b/include/asm-x86_64/desc.h
index eb7723a46790..913d6ac00033 100644
--- a/include/asm-x86_64/desc.h
+++ b/include/asm-x86_64/desc.h
@@ -9,64 +9,13 @@
 
 #include <linux/string.h>
 #include <linux/smp.h>
+#include <asm/desc_defs.h>
 
 #include <asm/segment.h>
 #include <asm/mmu.h>
 
-// 8 byte segment descriptor
-struct desc_struct { 
-        u16 limit0;
-        u16 base0;
-        unsigned base1 : 8, type : 4, s : 1, dpl : 2, p : 1;
-        unsigned limit : 4, avl : 1, l : 1, d : 1, g : 1, base2 : 8;
-} __attribute__((packed)); 
-
-struct n_desc_struct { 
-        unsigned int a,b;
-};      
-
 extern struct desc_struct cpu_gdt_table[GDT_ENTRIES];
 
-enum { 
-        GATE_INTERRUPT = 0xE, 
-        GATE_TRAP = 0xF,        
-        GATE_CALL = 0xC,
-};      
-
-// 16byte gate
-struct gate_struct {          
-        u16 offset_low;
-        u16 segment; 
-        unsigned ist : 3, zero0 : 5, type : 5, dpl : 2, p : 1;
-        u16 offset_middle;
-        u32 offset_high;
-        u32 zero1; 
-} __attribute__((packed));
-
-#define PTR_LOW(x) ((unsigned long)(x) & 0xFFFF) 
-#define PTR_MIDDLE(x) (((unsigned long)(x) >> 16) & 0xFFFF)
-#define PTR_HIGH(x) ((unsigned long)(x) >> 32)
-
-enum { 
-        DESC_TSS = 0x9,
-        DESC_LDT = 0x2,
-}; 
-
-// LDT or TSS descriptor in the GDT. 16 bytes.
-struct ldttss_desc { 
-        u16 limit0;
-        u16 base0;
-        unsigned base1 : 8, type : 5, dpl : 2, p : 1;
-        unsigned limit1 : 4, zero0 : 3, g : 1, base2 : 8;
-        u32 base3;
-        u32 zero1; 
-} __attribute__((packed)); 
-
-struct desc_ptr {
-        unsigned short size;
-        unsigned long address;
-} __attribute__((packed)) ;
-
 #define load_TR_desc() asm volatile("ltr %w0"::"r" (GDT_ENTRY_TSS*8))
 #define load_LDT_desc() asm volatile("lldt %w0"::"r" (GDT_ENTRY_LDT*8))
 #define clear_LDT()  asm volatile("lldt %w0"::"r" (0))
diff --git a/include/asm-x86_64/desc_defs.h b/include/asm-x86_64/desc_defs.h
new file mode 100644
index 000000000000..089004070099
--- /dev/null
+++ b/include/asm-x86_64/desc_defs.h
@@ -0,0 +1,69 @@
+/* Written 2000 by Andi Kleen */
+#ifndef __ARCH_DESC_DEFS_H
+#define __ARCH_DESC_DEFS_H
+
+/*
+ * Segment descriptor structure definitions, usable from both x86_64 and i386
+ * archs.
+ */
+
+#ifndef __ASSEMBLY__
+
+#include <linux/types.h>
+
+// 8 byte segment descriptor
+struct desc_struct {
+        u16 limit0;
+        u16 base0;
+        unsigned base1 : 8, type : 4, s : 1, dpl : 2, p : 1;
+        unsigned limit : 4, avl : 1, l : 1, d : 1, g : 1, base2 : 8;
+} __attribute__((packed));
+
+struct n_desc_struct {
+        unsigned int a,b;
+};
+
+enum {
+        GATE_INTERRUPT = 0xE,
+        GATE_TRAP = 0xF,
+        GATE_CALL = 0xC,
+};
+
+// 16byte gate
+struct gate_struct {
+        u16 offset_low;
+        u16 segment;
+        unsigned ist : 3, zero0 : 5, type : 5, dpl : 2, p : 1;
+        u16 offset_middle;
+        u32 offset_high;
+        u32 zero1;
+} __attribute__((packed));
+
+#define PTR_LOW(x) ((unsigned long)(x) & 0xFFFF)
+#define PTR_MIDDLE(x) (((unsigned long)(x) >> 16) & 0xFFFF)
+#define PTR_HIGH(x) ((unsigned long)(x) >> 32)
+
+enum {
+        DESC_TSS = 0x9,
+        DESC_LDT = 0x2,
+};
+
+// LDT or TSS descriptor in the GDT. 16 bytes.
+struct ldttss_desc {
+        u16 limit0;
+        u16 base0;
+        unsigned base1 : 8, type : 5, dpl : 2, p : 1;
+        unsigned limit1 : 4, zero0 : 3, g : 1, base2 : 8;
+        u32 base3;
+        u32 zero1;
+} __attribute__((packed));
+
+struct desc_ptr {
+        unsigned short size;
+        unsigned long address;
+} __attribute__((packed)) ;
+
+
+#endif /* !__ASSEMBLY__ */
+
+#endif
diff --git a/include/asm-x86_64/device.h b/include/asm-x86_64/device.h
new file mode 100644
index 000000000000..3afa03f33a36
--- /dev/null
+++ b/include/asm-x86_64/device.h
@@ -0,0 +1,15 @@
+/*
+ * Arch specific extensions to struct device
+ *
+ * This file is released under the GPLv2
+ */
+#ifndef _ASM_X86_64_DEVICE_H
+#define _ASM_X86_64_DEVICE_H
+
+struct dev_archdata {
+#ifdef CONFIG_ACPI
+        void    *acpi_handle;
+#endif
+};
+
+#endif /* _ASM_X86_64_DEVICE_H */
diff --git a/include/asm-x86_64/dma-mapping.h b/include/asm-x86_64/dma-mapping.h
index 10174b110a5c..be9ec6890723 100644
--- a/include/asm-x86_64/dma-mapping.h
+++ b/include/asm-x86_64/dma-mapping.h
@@ -180,12 +180,13 @@ static inline int dma_get_cache_alignment(void)
         return boot_cpu_data.x86_clflush_size;
 }
 
-#define dma_is_consistent(h) 1
+#define dma_is_consistent(d, h) 1
 
 extern int dma_set_mask(struct device *dev, u64 mask);
 
 static inline void
-dma_cache_sync(void *vaddr, size_t size, enum dma_data_direction dir)
+dma_cache_sync(struct device *dev, void *vaddr, size_t size,
+        enum dma_data_direction dir)
 {
         flush_write_buffers();
 }
diff --git a/include/asm-x86_64/elf.h b/include/asm-x86_64/elf.h
index a406fcb1e924..6d24ea7c4d9d 100644
--- a/include/asm-x86_64/elf.h
+++ b/include/asm-x86_64/elf.h
@@ -45,7 +45,6 @@ typedef struct user_i387_struct elf_fpregset_t;
 
 #ifdef __KERNEL__
 #include <asm/processor.h>
-#include <asm/compat.h>
 
 /*
  * This is used to ensure we don't load something for the wrong architecture.
diff --git a/include/asm-x86_64/futex.h b/include/asm-x86_64/futex.h
index 9804bf07b092..5cdfb08013c3 100644
--- a/include/asm-x86_64/futex.h
+++ b/include/asm-x86_64/futex.h
@@ -55,7 +55,7 @@ futex_atomic_op_inuser (int encoded_op, int __user *uaddr)
         if (! access_ok (VERIFY_WRITE, uaddr, sizeof(int)))
                 return -EFAULT;
 
-        inc_preempt_count();
+        pagefault_disable();
 
         switch (op) {
         case FUTEX_OP_SET:
@@ -78,7 +78,7 @@ futex_atomic_op_inuser (int encoded_op, int __user *uaddr)
                 ret = -ENOSYS;
         }
 
-        dec_preempt_count();
+        pagefault_enable();
 
         if (!ret) {
                 switch (cmp) {
diff --git a/include/asm-x86_64/genapic.h b/include/asm-x86_64/genapic.h
index a0e9a4b93484..b80f4bb5f273 100644
--- a/include/asm-x86_64/genapic.h
+++ b/include/asm-x86_64/genapic.h
@@ -30,6 +30,6 @@ struct genapic {
 };
 
 
-extern struct genapic *genapic;
+extern struct genapic *genapic, *genapic_force, apic_flat;
 
 #endif
diff --git a/include/asm-x86_64/io_apic.h b/include/asm-x86_64/io_apic.h
index 171ec2dc8c04..561ecbfd4cb5 100644
--- a/include/asm-x86_64/io_apic.h
+++ b/include/asm-x86_64/io_apic.h
@@ -12,10 +12,6 @@
 
 #define APIC_MISMATCH_DEBUG
 
-#define IO_APIC_BASE(idx) \
-                ((volatile int *)(__fix_to_virt(FIX_IO_APIC_BASE_0 + idx) \
-                + (mp_ioapics[idx].mpc_apicaddr & ~PAGE_MASK)))
-
 /*
  * The structure of the IO-APIC:
  */
@@ -119,36 +115,6 @@ extern struct mpc_config_intsrc mp_irqs[MAX_IRQ_SOURCES];
 /* non-0 if default (table-less) MP configuration */
 extern int mpc_default_type;
 
-static inline unsigned int io_apic_read(unsigned int apic, unsigned int reg)
-{
-        *IO_APIC_BASE(apic) = reg;
-        return *(IO_APIC_BASE(apic)+4);
-}
-
-static inline void io_apic_write(unsigned int apic, unsigned int reg, unsigned int value)
-{
-        *IO_APIC_BASE(apic) = reg;
-        *(IO_APIC_BASE(apic)+4) = value;
-}
-
-/*
- * Re-write a value: to be used for read-modify-write
- * cycles where the read already set up the index register.
- */
-static inline void io_apic_modify(unsigned int apic, unsigned int value)
-{
-        *(IO_APIC_BASE(apic)+4) = value;
-}
-
-/*
- * Synchronize the IO-APIC and the CPU by doing
- * a dummy read from the IO-APIC
- */
-static inline void io_apic_sync(unsigned int apic)
-{
-        (void) *(IO_APIC_BASE(apic)+4);
-}
-
 /* 1 if "noapic" boot option passed */
 extern int skip_ioapic_setup;
 
diff --git a/include/asm-x86_64/msr.h b/include/asm-x86_64/msr.h
index 37e194169fac..952783d35c7b 100644
--- a/include/asm-x86_64/msr.h
+++ b/include/asm-x86_64/msr.h
@@ -169,8 +169,8 @@ static inline unsigned int cpuid_edx(unsigned int op)
 #define MSR_LSTAR 0xc0000082            /* long mode SYSCALL target */
 #define MSR_CSTAR 0xc0000083            /* compatibility mode SYSCALL target */
 #define MSR_SYSCALL_MASK 0xc0000084     /* EFLAGS mask for syscall */
-#define MSR_FS_BASE 0xc0000100          /* 64bit GS base */
-#define MSR_GS_BASE 0xc0000101          /* 64bit FS base */
+#define MSR_FS_BASE 0xc0000100          /* 64bit FS base */
+#define MSR_GS_BASE 0xc0000101          /* 64bit GS base */
 #define MSR_KERNEL_GS_BASE  0xc0000102  /* SwapGS GS shadow (or USER_GS from kernel) */ 
 /* EFER bits: */ 
 #define _EFER_SCE 0  /* SYSCALL/SYSRET */
@@ -210,6 +210,10 @@ static inline unsigned int cpuid_edx(unsigned int op)
 #define MSR_IA32_LASTINTFROMIP     0x1dd
 #define MSR_IA32_LASTINTTOIP       0x1de
 
+#define MSR_IA32_PEBS_ENABLE            0x3f1
+#define MSR_IA32_DS_AREA                0x600
+#define MSR_IA32_PERF_CAPABILITIES      0x345
+
 #define MSR_MTRRfix64K_00000    0x250
 #define MSR_MTRRfix16K_80000    0x258
 #define MSR_MTRRfix16K_A0000    0x259
@@ -407,4 +411,13 @@ static inline unsigned int cpuid_edx(unsigned int op)
 #define MSR_P4_U2L_ESCR0                0x3b0
 #define MSR_P4_U2L_ESCR1                0x3b1
 
+/* Intel Core-based CPU performance counters */
+#define MSR_CORE_PERF_FIXED_CTR0        0x309
+#define MSR_CORE_PERF_FIXED_CTR1        0x30a
+#define MSR_CORE_PERF_FIXED_CTR2        0x30b
+#define MSR_CORE_PERF_FIXED_CTR_CTRL    0x38d
+#define MSR_CORE_PERF_GLOBAL_STATUS     0x38e
+#define MSR_CORE_PERF_GLOBAL_CTRL       0x38f
+#define MSR_CORE_PERF_GLOBAL_OVF_CTRL   0x390
+
 #endif
diff --git a/include/asm-x86_64/nmi.h b/include/asm-x86_64/nmi.h
index f367d4014b42..72375e7d32a8 100644
--- a/include/asm-x86_64/nmi.h
+++ b/include/asm-x86_64/nmi.h
@@ -77,4 +77,7 @@ extern int proc_nmi_enabled(struct ctl_table *, int , struct file *,
 
 extern int unknown_nmi_panic;
 
+void __trigger_all_cpu_backtrace(void);
+#define trigger_all_cpu_backtrace() __trigger_all_cpu_backtrace()
+
 #endif /* ASM_NMI_H */
diff --git a/include/asm-x86_64/pci-direct.h b/include/asm-x86_64/pci-direct.h
index eba9cb471df3..6823fa4f1afa 100644
--- a/include/asm-x86_64/pci-direct.h
+++ b/include/asm-x86_64/pci-direct.h
@@ -10,6 +10,7 @@ extern u32 read_pci_config(u8 bus, u8 slot, u8 func, u8 offset);
 extern u8 read_pci_config_byte(u8 bus, u8 slot, u8 func, u8 offset);
 extern u16 read_pci_config_16(u8 bus, u8 slot, u8 func, u8 offset);
 extern void write_pci_config(u8 bus, u8 slot, u8 func, u8 offset, u32 val);
+extern void write_pci_config_byte(u8 bus, u8 slot, u8 func, u8 offset, u8 val);
 
 extern int early_pci_allowed(void);
 
diff --git a/include/asm-x86_64/pda.h b/include/asm-x86_64/pda.h
index 14996d962bac..5642634843c4 100644
--- a/include/asm-x86_64/pda.h
+++ b/include/asm-x86_64/pda.h
@@ -109,6 +109,15 @@ extern struct x8664_pda _proxy_pda;
 #define sub_pda(field,val) pda_to_op("sub",field,val)
 #define or_pda(field,val) pda_to_op("or",field,val)
 
+/* This is not atomic against other CPUs -- CPU preemption needs to be off */
+#define test_and_clear_bit_pda(bit,field) ({            \
+        int old__;                                              \
+        asm volatile("btr %2,%%gs:%c3\n\tsbbl %0,%0"            \
+            : "=r" (old__), "+m" (_proxy_pda.field)             \
+            : "dIr" (bit), "i" (pda_offset(field)) : "memory"); \
+        old__;                                                  \
+})
+
 #endif
 
 #define PDA_STACKOFFSET (5*8)
diff --git a/include/asm-x86_64/pgtable.h b/include/asm-x86_64/pgtable.h
index 0555c1c4d8fa..59901c690a0d 100644
--- a/include/asm-x86_64/pgtable.h
+++ b/include/asm-x86_64/pgtable.h
@@ -221,20 +221,19 @@ static inline pte_t ptep_get_and_clear_full(struct mm_struct *mm, unsigned long
 #define __S110  PAGE_SHARED_EXEC
 #define __S111  PAGE_SHARED_EXEC
 
-static inline unsigned long pgd_bad(pgd_t pgd) 
-{ 
-       unsigned long val = pgd_val(pgd);
-       val &= ~PTE_MASK; 
-       val &= ~(_PAGE_USER | _PAGE_DIRTY); 
-       return val & ~(_PAGE_PRESENT | _PAGE_RW | _PAGE_ACCESSED);      
-} 
+static inline unsigned long pgd_bad(pgd_t pgd)
+{
+        return pgd_val(pgd) & ~(PTE_MASK | _KERNPG_TABLE | _PAGE_USER);
+}
 
 static inline unsigned long pud_bad(pud_t pud)
 {
-       unsigned long val = pud_val(pud);
-       val &= ~PTE_MASK;
-       val &= ~(_PAGE_USER | _PAGE_DIRTY);
-       return val & ~(_PAGE_PRESENT | _PAGE_RW | _PAGE_ACCESSED);
+        return pud_val(pud) & ~(PTE_MASK | _KERNPG_TABLE | _PAGE_USER);
+}
+
+static inline unsigned long pmd_bad(pmd_t pmd)
+{
+        return pmd_val(pmd) & ~(PTE_MASK | _KERNPG_TABLE | _PAGE_USER);
 }
 
 #define pte_none(x)     (!pte_val(x))
@@ -347,7 +346,6 @@ static inline int pmd_large(pmd_t pte) {
 #define pmd_none(x)     (!pmd_val(x))
 #define pmd_present(x)  (pmd_val(x) & _PAGE_PRESENT)
 #define pmd_clear(xp)   do { set_pmd(xp, __pmd(0)); } while (0)
-#define pmd_bad(x)      ((pmd_val(x) & (~PTE_MASK & ~_PAGE_USER)) != _KERNPG_TABLE )
 #define pfn_pmd(nr,prot) (__pmd(((nr) << PAGE_SHIFT) | pgprot_val(prot)))
 #define pmd_pfn(x)  ((pmd_val(x) & __PHYSICAL_MASK) >> PAGE_SHIFT)
 
diff --git a/include/asm-x86_64/processor.h b/include/asm-x86_64/processor.h
index cef17e0f828c..76552d72804c 100644
--- a/include/asm-x86_64/processor.h
+++ b/include/asm-x86_64/processor.h
@@ -475,6 +475,14 @@ static inline void __mwait(unsigned long eax, unsigned long ecx)
                 : :"a" (eax), "c" (ecx));
 }
 
+static inline void __sti_mwait(unsigned long eax, unsigned long ecx)
+{
+        /* "mwait %eax,%ecx;" */
+        asm volatile(
+                "sti; .byte 0x0f,0x01,0xc9;"
+                : :"a" (eax), "c" (ecx));
+}
+
 extern void mwait_idle_with_hints(unsigned long eax, unsigned long ecx);
 
 #define stack_current() \
diff --git a/include/asm-x86_64/proto.h b/include/asm-x86_64/proto.h
index e72cfcdf5344..6d324b838972 100644
--- a/include/asm-x86_64/proto.h
+++ b/include/asm-x86_64/proto.h
@@ -61,7 +61,6 @@ extern void numa_initmem_init(unsigned long start_pfn, unsigned long end_pfn);
 extern unsigned long numa_free_all_bootmem(void);
 
 extern void reserve_bootmem_generic(unsigned long phys, unsigned len);
-extern void free_bootmem_generic(unsigned long phys, unsigned len);
 
 extern void load_gs_index(unsigned gs);
 
@@ -88,6 +87,7 @@ extern void syscall32_cpu_init(void);
 extern void setup_node_bootmem(int nodeid, unsigned long start, unsigned long end);
 
 extern void early_quirks(void);
+extern void quirk_intel_irqbalance(void);
 extern void check_efer(void);
 
 extern int unhandled_signal(struct task_struct *tsk, int sig);
diff --git a/include/asm-x86_64/rio.h b/include/asm-x86_64/rio.h
new file mode 100644
index 000000000000..c7350f6d2015
--- /dev/null
+++ b/include/asm-x86_64/rio.h
@@ -0,0 +1,74 @@
+/*
+ * Derived from include/asm-i386/mach-summit/mach_mpparse.h
+ *          and include/asm-i386/mach-default/bios_ebda.h
+ *
+ * Author: Laurent Vivier <Laurent.Vivier@bull.net>
+ */
+
+#ifndef __ASM_RIO_H
+#define __ASM_RIO_H
+
+#define RIO_TABLE_VERSION       3
+
+struct rio_table_hdr {
+        u8 version;      /* Version number of this data structure  */
+        u8 num_scal_dev; /* # of Scalability devices               */
+        u8 num_rio_dev;  /* # of RIO I/O devices                   */
+} __attribute__((packed));
+
+struct scal_detail {
+        u8 node_id;      /* Scalability Node ID                    */
+        u32 CBAR;        /* Address of 1MB register space          */
+        u8 port0node;    /* Node ID port connected to: 0xFF=None   */
+        u8 port0port;    /* Port num port connected to: 0,1,2, or  */
+                         /* 0xFF=None                              */
+        u8 port1node;    /* Node ID port connected to: 0xFF = None */
+        u8 port1port;    /* Port num port connected to: 0,1,2, or  */
+                         /* 0xFF=None                              */
+        u8 port2node;    /* Node ID port connected to: 0xFF = None */
+        u8 port2port;    /* Port num port connected to: 0,1,2, or  */
+                         /* 0xFF=None                              */
+        u8 chassis_num;  /* 1 based Chassis number (1 = boot node) */
+} __attribute__((packed));
+
+struct rio_detail {
+        u8 node_id;      /* RIO Node ID                            */
+        u32 BBAR;        /* Address of 1MB register space          */
+        u8 type;         /* Type of device                         */
+        u8 owner_id;     /* Node ID of Hurricane that owns this    */
+                         /* node                                   */
+        u8 port0node;    /* Node ID port connected to: 0xFF=None   */
+        u8 port0port;    /* Port num port connected to: 0,1,2, or  */
+                         /* 0xFF=None                              */
+        u8 port1node;    /* Node ID port connected to: 0xFF=None   */
+        u8 port1port;    /* Port num port connected to: 0,1,2, or  */
+                         /* 0xFF=None                              */
+        u8 first_slot;   /* Lowest slot number below this Calgary  */
+        u8 status;       /* Bit 0 = 1 : the XAPIC is used          */
+                         /*       = 0 : the XAPIC is not used, ie: */
+                         /*            ints fwded to another XAPIC */
+                         /*           Bits1:7 Reserved             */
+        u8 WP_index;     /* instance index - lower ones have       */
+                         /*     lower slot numbers/PCI bus numbers */
+        u8 chassis_num;  /* 1 based Chassis number                 */
+} __attribute__((packed));
+
+enum {
+        HURR_SCALABILTY = 0,  /* Hurricane Scalability info */
+        HURR_RIOIB      = 2,  /* Hurricane RIOIB info       */
+        COMPAT_CALGARY  = 4,  /* Compatibility Calgary      */
+        ALT_CALGARY     = 5,  /* Second Planar Calgary      */
+};
+
+/*
+ * there is a real-mode segmented pointer pointing to the
+ * 4K EBDA area at 0x40E.
+ */
+static inline unsigned long get_bios_ebda(void)
+{
+        unsigned long address = *(unsigned short *)phys_to_virt(0x40EUL);
+        address <<= 4;
+        return address;
+}
+
+#endif /* __ASM_RIO_H */
diff --git a/include/asm-x86_64/smp.h b/include/asm-x86_64/smp.h
index d6b7c057edba..e17b9ec42e98 100644
--- a/include/asm-x86_64/smp.h
+++ b/include/asm-x86_64/smp.h
@@ -82,11 +82,6 @@ extern u8 x86_cpu_to_apicid[NR_CPUS];	/* physical ID */
 extern u8 x86_cpu_to_log_apicid[NR_CPUS];
 extern u8 bios_cpu_apicid[];
 
-static inline unsigned int cpu_mask_to_apicid(cpumask_t cpumask)
-{
-        return cpus_addr(cpumask)[0];
-}
-
 static inline int cpu_present_to_apicid(int mps_cpu)
 {
         if (mps_cpu < NR_CPUS)
@@ -118,13 +113,6 @@ static __inline int logical_smp_processor_id(void)
 #define cpu_physical_id(cpu)            x86_cpu_to_apicid[cpu]
 #else
 #define cpu_physical_id(cpu)            boot_cpu_id
-static inline int smp_call_function_single(int cpuid, void (*func) (void *info),
-                                void *info, int retry, int wait)
-{
-        /* Disable interrupts here? */
-        func(info);
-        return 0;
-}
 #endif /* !CONFIG_SMP */
 #endif
 
diff --git a/include/asm-x86_64/spinlock.h b/include/asm-x86_64/spinlock.h
index 05ef097ba55b..88bf981e73cf 100644
--- a/include/asm-x86_64/spinlock.h
+++ b/include/asm-x86_64/spinlock.h
@@ -36,7 +36,34 @@ static inline void __raw_spin_lock(raw_spinlock_t *lock)
                 "2:\t" : "=m" (lock->slock) : : "memory");
 }
 
-#define __raw_spin_lock_flags(lock, flags) __raw_spin_lock(lock)
+/*
+ * Same as __raw_spin_lock, but reenable interrupts during spinning.
+ */
+#ifndef CONFIG_PROVE_LOCKING
+static inline void __raw_spin_lock_flags(raw_spinlock_t *lock, unsigned long flags)
+{
+        asm volatile(
+                "\n1:\t"
+                LOCK_PREFIX " ; decl %0\n\t"
+                "jns 5f\n"
+                "testl $0x200, %1\n\t"  /* interrupts were disabled? */
+                "jz 4f\n\t"
+                "sti\n"
+                "3:\t"
+                "rep;nop\n\t"
+                "cmpl $0, %0\n\t"
+                "jle 3b\n\t"
+                "cli\n\t"
+                "jmp 1b\n"
+                "4:\t"
+                "rep;nop\n\t"
+                "cmpl $0, %0\n\t"
+                "jg 1b\n\t"
+                "jmp 4b\n"
+                "5:\n\t"
+                : "+m" (lock->slock) : "r" ((unsigned)flags) : "memory");
+}
+#endif
 
 static inline int __raw_spin_trylock(raw_spinlock_t *lock)
 {
diff --git a/include/asm-x86_64/spinlock_types.h b/include/asm-x86_64/spinlock_types.h
index 59efe849f351..4da9345c1500 100644
--- a/include/asm-x86_64/spinlock_types.h
+++ b/include/asm-x86_64/spinlock_types.h
@@ -6,13 +6,13 @@
 #endif
 
 typedef struct {
-        volatile unsigned int slock;
+        unsigned int slock;
 } raw_spinlock_t;
 
 #define __RAW_SPIN_LOCK_UNLOCKED        { 1 }
 
 typedef struct {
-        volatile unsigned int lock;
+        unsigned int lock;
 } raw_rwlock_t;
 
 #define __RAW_RW_LOCK_UNLOCKED          { RW_LOCK_BIAS }
diff --git a/include/asm-x86_64/stacktrace.h b/include/asm-x86_64/stacktrace.h
index 5eb9799bef76..6f0b54594307 100644
--- a/include/asm-x86_64/stacktrace.h
+++ b/include/asm-x86_64/stacktrace.h
@@ -1,6 +1,8 @@
 #ifndef _ASM_STACKTRACE_H
 #define _ASM_STACKTRACE_H 1
 
+extern int kstack_depth_to_print;
+
 /* Generic stack tracer with callbacks */
 
 struct stacktrace_ops {
diff --git a/include/asm-x86_64/types.h b/include/asm-x86_64/types.h
index c86c2e6793e2..2d4491aae281 100644
--- a/include/asm-x86_64/types.h
+++ b/include/asm-x86_64/types.h
@@ -48,9 +48,6 @@ typedef unsigned long long u64;
 typedef u64 dma64_addr_t;
 typedef u64 dma_addr_t;
 
-typedef u64 sector_t;
-#define HAVE_SECTOR_T
-
 #endif /* __ASSEMBLY__ */
 
 #endif /* __KERNEL__ */
diff --git a/include/asm-x86_64/uaccess.h b/include/asm-x86_64/uaccess.h
index 19f99178fe83..d5dbc87274f8 100644
--- a/include/asm-x86_64/uaccess.h
+++ b/include/asm-x86_64/uaccess.h
@@ -6,7 +6,6 @@
  */
 #include <linux/compiler.h>
 #include <linux/errno.h>
-#include <linux/sched.h>
 #include <linux/prefetch.h>
 #include <asm/page.h>
 
diff --git a/include/asm-x86_64/unistd.h b/include/asm-x86_64/unistd.h
index 777288eb7e75..c5f596e71faa 100644
--- a/include/asm-x86_64/unistd.h
+++ b/include/asm-x86_64/unistd.h
@@ -622,25 +622,7 @@ __SYSCALL(__NR_move_pages, sys_move_pages)
 
 #define __NR_syscall_max __NR_move_pages
 
-#ifdef __KERNEL__
-#include <linux/err.h>
-#endif
-
 #ifndef __NO_STUBS
-
-/* user-visible error numbers are in the range -1 - -MAX_ERRNO */
-
-#define __syscall_clobber "r11","rcx","memory" 
-
-#define __syscall_return(type, res) \
-do { \
-        if ((unsigned long)(res) >= (unsigned long)(-MAX_ERRNO)) { \
-                errno = -(res); \
-                res = -1; \
-        } \
-        return (type) (res); \
-} while (0)
-
 #define __ARCH_WANT_OLD_READDIR
 #define __ARCH_WANT_OLD_STAT
 #define __ARCH_WANT_SYS_ALARM
@@ -664,87 +646,6 @@ do { \
 #define __ARCH_WANT_SYS_TIME
 #define __ARCH_WANT_COMPAT_SYS_TIME
 
-#define __syscall "syscall"
-
-#define _syscall0(type,name) \
-type name(void) \
-{ \
-long __res; \
-__asm__ volatile (__syscall \
-        : "=a" (__res) \
-        : "0" (__NR_##name) : __syscall_clobber ); \
-__syscall_return(type,__res); \
-}
-
-#define _syscall1(type,name,type1,arg1) \
-type name(type1 arg1) \
-{ \
-long __res; \
-__asm__ volatile (__syscall \
-        : "=a" (__res) \
-        : "0" (__NR_##name),"D" ((long)(arg1)) : __syscall_clobber ); \
-__syscall_return(type,__res); \
-}
-
-#define _syscall2(type,name,type1,arg1,type2,arg2) \
-type name(type1 arg1,type2 arg2) \
-{ \
-long __res; \
-__asm__ volatile (__syscall \
-        : "=a" (__res) \
-        : "0" (__NR_##name),"D" ((long)(arg1)),"S" ((long)(arg2)) : __syscall_clobber ); \
-__syscall_return(type,__res); \
-}
-
-#define _syscall3(type,name,type1,arg1,type2,arg2,type3,arg3) \
-type name(type1 arg1,type2 arg2,type3 arg3) \
-{ \
-long __res; \
-__asm__ volatile (__syscall \
-        : "=a" (__res) \
-        : "0" (__NR_##name),"D" ((long)(arg1)),"S" ((long)(arg2)), \
-                  "d" ((long)(arg3)) : __syscall_clobber); \
-__syscall_return(type,__res); \
-}
-
-#define _syscall4(type,name,type1,arg1,type2,arg2,type3,arg3,type4,arg4) \
-type name (type1 arg1, type2 arg2, type3 arg3, type4 arg4) \
-{ \
-long __res; \
-__asm__ volatile ("movq %5,%%r10 ;" __syscall \
-        : "=a" (__res) \
-        : "0" (__NR_##name),"D" ((long)(arg1)),"S" ((long)(arg2)), \
-          "d" ((long)(arg3)),"g" ((long)(arg4)) : __syscall_clobber,"r10" ); \
-__syscall_return(type,__res); \
-} 
-
-#define _syscall5(type,name,type1,arg1,type2,arg2,type3,arg3,type4,arg4, \
-          type5,arg5) \
-type name (type1 arg1,type2 arg2,type3 arg3,type4 arg4,type5 arg5) \
-{ \
-long __res; \
-__asm__ volatile ("movq %5,%%r10 ; movq %6,%%r8 ; " __syscall \
-        : "=a" (__res) \
-        : "0" (__NR_##name),"D" ((long)(arg1)),"S" ((long)(arg2)), \
-          "d" ((long)(arg3)),"g" ((long)(arg4)),"g" ((long)(arg5)) : \
-        __syscall_clobber,"r8","r10" ); \
-__syscall_return(type,__res); \
-}
-
-#define _syscall6(type,name,type1,arg1,type2,arg2,type3,arg3,type4,arg4, \
-          type5,arg5,type6,arg6) \
-type name (type1 arg1,type2 arg2,type3 arg3,type4 arg4,type5 arg5,type6 arg6) \
-{ \
-long __res; \
-__asm__ volatile ("movq %5,%%r10 ; movq %6,%%r8 ; movq %7,%%r9 ; " __syscall \
-        : "=a" (__res) \
-        : "0" (__NR_##name),"D" ((long)(arg1)),"S" ((long)(arg2)), \
-          "d" ((long)(arg3)), "g" ((long)(arg4)), "g" ((long)(arg5)), \
-          "g" ((long)(arg6)) : \
-        __syscall_clobber,"r8","r10","r9" ); \
-__syscall_return(type,__res); \
-}
-
 #ifdef __KERNEL__
 #ifndef __ASSEMBLY__
 
diff --git a/include/asm-x86_64/unwind.h b/include/asm-x86_64/unwind.h
index 2e7ff10fd775..2f6349e48717 100644
--- a/include/asm-x86_64/unwind.h
+++ b/include/asm-x86_64/unwind.h
@@ -87,14 +87,10 @@ extern int arch_unwind_init_running(struct unwind_frame_info *,
 
 static inline int arch_unw_user_mode(const struct unwind_frame_info *info)
 {
-#if 0 /* This can only work when selector register saves/restores
-         are properly annotated (and tracked in UNW_REGISTER_INFO). */
-        return user_mode(&info->regs);
-#else
-        return (long)info->regs.rip >= 0
+        return user_mode(&info->regs)
+               || (long)info->regs.rip >= 0
                || (info->regs.rip >= VSYSCALL_START && info->regs.rip < VSYSCALL_END)
                || (long)info->regs.rsp >= 0;
-#endif
 }
 
 #else
diff --git a/include/asm-x86_64/vsyscall.h b/include/asm-x86_64/vsyscall.h
index fd452fc2c037..05cb8dd200de 100644
--- a/include/asm-x86_64/vsyscall.h
+++ b/include/asm-x86_64/vsyscall.h
@@ -10,6 +10,7 @@ enum vsyscall_num {
 #define VSYSCALL_START (-10UL << 20)
 #define VSYSCALL_SIZE 1024
 #define VSYSCALL_END (-2UL << 20)
+#define VSYSCALL_MAPPED_PAGES 1
 #define VSYSCALL_ADDR(vsyscall_nr) (VSYSCALL_START+VSYSCALL_SIZE*(vsyscall_nr))
 
 #ifdef __KERNEL__
@@ -59,8 +60,6 @@ extern seqlock_t xtime_lock;
 
 extern int sysctl_vsyscall;
 
-extern void vsyscall_set_cpu(int cpu);
-
 #define ARCH_HAVE_XTIME_LOCK 1
 
 #endif /* __KERNEL__ */