---
 arch/x86/Makefile_64        |    1 +
 arch/x86/kernel/entry_64.S  |    4 ++--
 include/asm-x86/percpu_64.h |   32 +++++++++++++++++++-------------
 3 files changed, 22 insertions(+), 15 deletions(-)

Index: linux-2.6/arch/x86/Makefile_64
===================================================================
--- linux-2.6.orig/arch/x86/Makefile_64	2007-10-29 18:23:24.340823073 -0700
+++ linux-2.6/arch/x86/Makefile_64	2007-10-29 18:23:47.749072234 -0700
@@ -61,6 +61,7 @@ cflags-y += $(call cc-option,-mno-sse -m
 # this works around some issues with generating unwind tables in older gccs
 # newer gccs do it by default
 cflags-y += -maccumulate-outgoing-args
+cflags-y += -mtls-direct-seg-other
 
 # do binutils support CFI?
 cflags-y += $(call as-instr,.cfi_startproc\n.cfi_rel_offset rsp${comma}0\n.cfi_endproc,-DCONFIG_AS_CFI=1,)
Index: linux-2.6/arch/x86/kernel/entry_64.S
===================================================================
--- linux-2.6.orig/arch/x86/kernel/entry_64.S	2007-10-29 17:56:59.111573397 -0700
+++ linux-2.6/arch/x86/kernel/entry_64.S	2007-10-29 17:58:19.380322743 -0700
@@ -741,11 +741,11 @@ END(spurious_interrupt)
 	movq ORIG_RAX(%rsp),%rsi
 	movq $-1,ORIG_RAX(%rsp)
 	.if \ist
-	subq	$EXCEPTION_STKSZ, per_cpu__init_tss + TSS_ist + (\ist - 1) * 8(%rbp)
+//	subq	$EXCEPTION_STKSZ, per_cpu__init_tss + TSS_ist + (\ist - 1) * 8(%rbp)
 	.endif
 	call \sym
 	.if \ist
-	addq	$EXCEPTION_STKSZ, per_cpu__init_tss + TSS_ist + (\ist - 1) * 8(%rbp)
+//	addq	$EXCEPTION_STKSZ, per_cpu__init_tss + TSS_ist + (\ist - 1) * 8(%rbp)
 	.endif
 	cli
 	.if \irqtrace
Index: linux-2.6/include/asm-x86/percpu_64.h
===================================================================
--- linux-2.6.orig/include/asm-x86/percpu_64.h	2007-10-29 17:44:24.635822796 -0700
+++ linux-2.6/include/asm-x86/percpu_64.h	2007-10-29 17:59:51.440722104 -0700
@@ -12,29 +12,24 @@
 #include <asm/pda.h>
 
 #define __per_cpu_offset(cpu) (cpu_pda(cpu)->data_offset)
-#define __my_cpu_offset() read_pda(data_offset)
+#define __my_cpu_offset() 0
 
 #define per_cpu_offset(x) (__per_cpu_offset(x))
 
 /* Separate out the type, so (int[3], foo) works. */
 #define DEFINE_PER_CPU(type, name) \
-    __attribute__((__section__(".data.percpu"))) __typeof__(type) per_cpu__##name
+    __attribute__((__section__(".data.percpu"))) __typeof__(type) __thread per_cpu__##name
 
 #define DEFINE_PER_CPU_SHARED_ALIGNED(type, name)		\
     __attribute__((__section__(".data.percpu.shared_aligned"))) \
-    __typeof__(type) per_cpu__##name				\
-    ____cacheline_internodealigned_in_smp
+    __typeof__(type) __thread per_cpu__##name
 
 /* var is in discarded region: offset to particular copy we want */
 #define per_cpu(var, cpu) (*({				\
 	extern int simple_identifier_##var(void);	\
 	RELOC_HIDE(&per_cpu__##var, __per_cpu_offset(cpu)); }))
-#define __get_cpu_var(var) (*({				\
-	extern int simple_identifier_##var(void);	\
-	RELOC_HIDE(&per_cpu__##var, __my_cpu_offset()); }))
-#define __raw_get_cpu_var(var) (*({			\
-	extern int simple_identifier_##var(void);	\
-	RELOC_HIDE(&per_cpu__##var, __my_cpu_offset()); }))
+#define __get_cpu_var(var) per_cpu__##var
+#define __raw_get_cpu_var(var) per_cpu__##var
 
 /* A macro to avoid #include hell... */
 #define percpu_modcopy(pcpudst, src, size)			\
@@ -60,9 +55,20 @@ extern void setup_per_cpu_areas(void);
 
 #endif	/* SMP */
 
-#define DECLARE_PER_CPU(type, name) extern __typeof__(type) per_cpu__##name
+#define DECLARE_PER_CPU(type, name) extern __typeof__(type) __thread per_cpu__##name
+
+#define __EXPORT_PER_CPUSYMBOL(sym, sec)                        \
+        extern typeof(sym) __thread sym;                        \
+        __CRC_SYMBOL(sym, sec)                                  \
+        static const char __kstrtab_##sym[]                     \
+        __attribute__((section("__ksymtab_strings")))           \
+        = MODULE_SYMBOL_PREFIX #sym;                            \
+        static const struct kernel_symbol __ksymtab_##sym       \
+        __attribute_used__                                      \
+        __attribute__((section("__ksymtab" sec), unused))       \
+        = { /* FIXME: addr not working */ 0, __kstrtab_##sym }
 
-#define EXPORT_PER_CPU_SYMBOL(var) EXPORT_SYMBOL(per_cpu__##var)
-#define EXPORT_PER_CPU_SYMBOL_GPL(var) EXPORT_SYMBOL_GPL(per_cpu__##var)
+#define EXPORT_PER_CPU_SYMBOL(var) __EXPORT_PER_CPUSYMBOL(per_cpu__##var, "")
+#define EXPORT_PER_CPU_SYMBOL_GPL(var) EXPORT_PER_CPUSYMBOL_GPL(per_cpu__##var, "_gpl")
 
 #endif /* _ASM_X8664_PERCPU_H_ */