2 files changed, 203 insertions, 141 deletions

diff --git a/include/asm-generic/percpu.h b/include/asm-generic/percpu.h
index d17784ea37ff..0703aa75b5e8 100644
--- a/include/asm-generic/percpu.h
+++ b/include/asm-generic/percpu.h
@@ -56,17 +56,17 @@ extern unsigned long __per_cpu_offset[NR_CPUS];
 #define per_cpu(var, cpu) \
         (*SHIFT_PERCPU_PTR(&(var), per_cpu_offset(cpu)))
 
-#ifndef __this_cpu_ptr
-#define __this_cpu_ptr(ptr) SHIFT_PERCPU_PTR(ptr, __my_cpu_offset)
+#ifndef raw_cpu_ptr
+#define raw_cpu_ptr(ptr) SHIFT_PERCPU_PTR(ptr, __my_cpu_offset)
 #endif
 #ifdef CONFIG_DEBUG_PREEMPT
 #define this_cpu_ptr(ptr) SHIFT_PERCPU_PTR(ptr, my_cpu_offset)
 #else
-#define this_cpu_ptr(ptr) __this_cpu_ptr(ptr)
+#define this_cpu_ptr(ptr) raw_cpu_ptr(ptr)
 #endif
 
 #define __get_cpu_var(var) (*this_cpu_ptr(&(var)))
-#define __raw_get_cpu_var(var) (*__this_cpu_ptr(&(var)))
+#define __raw_get_cpu_var(var) (*raw_cpu_ptr(&(var)))
 
 #ifdef CONFIG_HAVE_SETUP_PER_CPU_AREA
 extern void setup_per_cpu_areas(void);
@@ -83,7 +83,7 @@ extern void setup_per_cpu_areas(void);
 #define __get_cpu_var(var)      (*VERIFY_PERCPU_PTR(&(var)))
 #define __raw_get_cpu_var(var)  (*VERIFY_PERCPU_PTR(&(var)))
 #define this_cpu_ptr(ptr)       per_cpu_ptr(ptr, 0)
-#define __this_cpu_ptr(ptr)     this_cpu_ptr(ptr)
+#define raw_cpu_ptr(ptr)        this_cpu_ptr(ptr)
 
 #endif  /* SMP */
 
@@ -122,4 +122,7 @@ extern void setup_per_cpu_areas(void);
 #define PER_CPU_DEF_ATTRIBUTES
 #endif
 
+/* Keep until we have removed all uses of __this_cpu_ptr */
+#define __this_cpu_ptr raw_cpu_ptr
+
 #endif /* _ASM_GENERIC_PERCPU_H_ */
diff --git a/include/linux/percpu.h b/include/linux/percpu.h
index e3817d2441b6..4e4d2afcc0c7 100644
--- a/include/linux/percpu.h
+++ b/include/linux/percpu.h
@@ -243,6 +243,8 @@ do {									\
 } while (0)
 
 /*
+ * this_cpu operations (C) 2008-2013 Christoph Lameter <cl@linux.com>
+ *
  * Optimized manipulation for memory allocated through the per cpu
  * allocator or for addresses of per cpu variables.
  *
@@ -296,7 +298,7 @@ do {									\
 do {                                                                    \
         unsigned long flags;                                            \
         raw_local_irq_save(flags);                                      \
-        *__this_cpu_ptr(&(pcp)) op val;                                 \
+        *raw_cpu_ptr(&(pcp)) op val;                                    \
         raw_local_irq_restore(flags);                                   \
 } while (0)
 
@@ -381,8 +383,8 @@ do {									\
         typeof(pcp) ret__;                                              \
         unsigned long flags;                                            \
         raw_local_irq_save(flags);                                      \
-        __this_cpu_add(pcp, val);                                       \
-        ret__ = __this_cpu_read(pcp);                                   \
+        raw_cpu_add(pcp, val);                                  \
+        ret__ = raw_cpu_read(pcp);                                      \
         raw_local_irq_restore(flags);                                   \
         ret__;                                                          \
 })
@@ -411,8 +413,8 @@ do {									\
 ({      typeof(pcp) ret__;                                              \
         unsigned long flags;                                            \
         raw_local_irq_save(flags);                                      \
-        ret__ = __this_cpu_read(pcp);                                   \
-        __this_cpu_write(pcp, nval);                                    \
+        ret__ = raw_cpu_read(pcp);                                      \
+        raw_cpu_write(pcp, nval);                                       \
         raw_local_irq_restore(flags);                                   \
         ret__;                                                          \
 })
@@ -439,9 +441,9 @@ do {									\
         typeof(pcp) ret__;                                              \
         unsigned long flags;                                            \
         raw_local_irq_save(flags);                                      \
-        ret__ = __this_cpu_read(pcp);                                   \
+        ret__ = raw_cpu_read(pcp);                                      \
         if (ret__ == (oval))                                            \
-                __this_cpu_write(pcp, nval);                            \
+                raw_cpu_write(pcp, nval);                               \
         raw_local_irq_restore(flags);                                   \
         ret__;                                                          \
 })
@@ -476,7 +478,7 @@ do {									\
         int ret__;                                                      \
         unsigned long flags;                                            \
         raw_local_irq_save(flags);                                      \
-        ret__ = __this_cpu_generic_cmpxchg_double(pcp1, pcp2,           \
+        ret__ = raw_cpu_generic_cmpxchg_double(pcp1, pcp2,              \
                         oval1, oval2, nval1, nval2);                    \
         raw_local_irq_restore(flags);                                   \
         ret__;                                                          \
@@ -504,12 +506,8 @@ do {									\
 #endif
 
 /*
- * Generic percpu operations for context that are safe from preemption/interrupts.
- * Either we do not care about races or the caller has the
- * responsibility of handling preemption/interrupt issues. Arch code can still
- * override these instructions since the arch per cpu code may be more
- * efficient and may actually get race freeness for free (that is the
- * case for x86 for example).
+ * Generic percpu operations for contexts where we do not want to do
+ * any checks for preemptiosn.
  *
  * If there is no other protection through preempt disable and/or
  * disabling interupts then one of these RMW operations can show unexpected
@@ -517,211 +515,272 @@ do {									\
  * or an interrupt occurred and the same percpu variable was modified from
  * the interrupt context.
  */
-#ifndef __this_cpu_read
-# ifndef __this_cpu_read_1
-#  define __this_cpu_read_1(pcp)        (*__this_cpu_ptr(&(pcp)))
+#ifndef raw_cpu_read
+# ifndef raw_cpu_read_1
+#  define raw_cpu_read_1(pcp)   (*raw_cpu_ptr(&(pcp)))
 # endif
-# ifndef __this_cpu_read_2
-#  define __this_cpu_read_2(pcp)        (*__this_cpu_ptr(&(pcp)))
+# ifndef raw_cpu_read_2
+#  define raw_cpu_read_2(pcp)   (*raw_cpu_ptr(&(pcp)))
 # endif
-# ifndef __this_cpu_read_4
-#  define __this_cpu_read_4(pcp)        (*__this_cpu_ptr(&(pcp)))
+# ifndef raw_cpu_read_4
+#  define raw_cpu_read_4(pcp)   (*raw_cpu_ptr(&(pcp)))
 # endif
-# ifndef __this_cpu_read_8
-#  define __this_cpu_read_8(pcp)        (*__this_cpu_ptr(&(pcp)))
+# ifndef raw_cpu_read_8
+#  define raw_cpu_read_8(pcp)   (*raw_cpu_ptr(&(pcp)))
 # endif
-# define __this_cpu_read(pcp)   __pcpu_size_call_return(__this_cpu_read_, (pcp))
+# define raw_cpu_read(pcp)      __pcpu_size_call_return(raw_cpu_read_, (pcp))
 #endif
 
-#define __this_cpu_generic_to_op(pcp, val, op)                          \
+#define raw_cpu_generic_to_op(pcp, val, op)                             \
 do {                                                                    \
-        *__this_cpu_ptr(&(pcp)) op val;                                 \
+        *raw_cpu_ptr(&(pcp)) op val;                                    \
 } while (0)
 
-#ifndef __this_cpu_write
-# ifndef __this_cpu_write_1
-#  define __this_cpu_write_1(pcp, val)  __this_cpu_generic_to_op((pcp), (val), =)
+
+#ifndef raw_cpu_write
+# ifndef raw_cpu_write_1