---
 include/asm-x86/pda.h       |   11 ++---------
 include/asm-x86/percpu_64.h |   38 +++++++++++++++++++++++++++++++++++++-
 2 files changed, 39 insertions(+), 10 deletions(-)

Index: linux-2.6/include/asm-x86/pda.h
===================================================================
--- linux-2.6.orig/include/asm-x86/pda.h	2007-11-19 17:08:40.767874989 -0800
+++ linux-2.6/include/asm-x86/pda.h	2007-11-19 17:08:40.871625076 -0800
@@ -53,15 +53,8 @@ extern struct x8664_pda _proxy_pda;
 #define add_pda(field,val) CPU_ADD(per_cpu_var(pda).field, val)
 #define sub_pda(field,val) CPU_ADD(per_cpu_var(pda).field, val)
 #define or_pda(field,val) CPU_OR(per_cpu_var(pda).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__;							\
-})
+#define test_and_clear_bit_pda(bit, field) \
+		CPU_TEST_AND_CLEAR(per_cpu_var(pda).field, bit)
 
 #endif
 
Index: linux-2.6/include/asm-x86/percpu_64.h
===================================================================
--- linux-2.6.orig/include/asm-x86/percpu_64.h	2007-11-19 17:08:40.767874989 -0800
+++ linux-2.6/include/asm-x86/percpu_64.h	2007-11-19 17:08:40.871625076 -0800
@@ -304,6 +304,8 @@ static inline unsigned long __cmpxchg_lo
 	((__typeof__(obj))__cmpxchg_local_gs(&(obj),(unsigned long)(o),\
 					(unsigned long)(n),sizeof(obj)))
 
+/* x86 only functions */
+
 static inline void __cpu_or_gs(volatile void *ptr,
 				long data, int size)
 {
@@ -331,6 +333,40 @@ static inline void __cpu_or_gs(volatile 
 #define cpu_or_gs(obj, value)\
 	__cpu_or_gs(&(obj), (unsigned long)value, sizeof(obj))
 
+
+static inline int __cpu_test_and_clear_gs(volatile void *ptr,
+				int bit, int size)
+{
+	int result;
+
+	switch (size) {
+	case 1:
+		__asm__ ("btr %2, %%gs:%3\n\tsbbl %0,%0"
+			: "r=" (result), "m"(*__xp(ptr))
+			: "dIr" (bit), "m"(*__xp(ptr)) : "memory");
+		return result;
+	case 2:
+		__asm__ ("btr %2, %%gs:%3\n\tsbbl %0,%0"
+			: "r=" (result), "m"(*__xp(ptr))
+			: "dIr" (bit), "m"(*__xp(ptr)) : "memory");
+		return result;
+	case 4:
+		__asm__ ("btr %2, %%gs:%3\n\tsbbl %0,%0"
+			: "r=" (result), "m"(*__xp(ptr))
+			: "dIr" (bit), "m"(*__xp(ptr)) : "memory");
+		return result;
+	case 8:
+		__asm__ ("btr %2, %%gs:%3\n\tsbbl %0,%0"
+			: "r=" (result), "m"(*__xp(ptr))
+			: "dIr" (bit), "m"(*__xp(ptr)) : "memory");
+		return result;
+	}
+	BUG();
+}
+
+#define cpu_test_and_clear_gs(obj, value)\
+	__cpu_test_and_clear_gs(&(obj), (unsigned long)value, sizeof(obj))
+
 #define CPU_READ(obj)		cpu_read_gs(obj)
 #define CPU_WRITE(obj,val)	cpu_write_gs(obj, val)
 #define CPU_ADD(obj,val)	cpu_add_gs(obj, val)
@@ -338,7 +374,7 @@ static inline void __cpu_or_gs(volatile 
 #define CPU_INC(obj)		cpu_inc_gs(obj)
 #define CPU_DEC(obj)		cpu_dec_gs(obj)
 #define CPU_OR(obj, val)	cpu_or_gs(obj, val)
-
+#define CPU_TEST_AND_CLEAR(obj, bit)	cpu_test_and_clear_gs(obj, bit)
 #define CPU_XCHG(obj,val)	cpu_xchg_gs(obj, val)
 #define CPU_CMPXCHG(obj, old, new) cmpxchg_local_gs(obj, old, new)