Commit b2c0ab17 authored by Anton Blanchard's avatar Anton Blanchard Committed by Paul Mackerras
Browse files

[PATCH] ppc64: speedup cmpxchg 



cmpxchg has the following code:

__typeof__(*(ptr)) _o_ = (o);
__typeof__(*(ptr)) _n_ = (n);

Unfortunately it makes gcc 4.0 store and load the variables to the stack.
Eg in atomic_dec_and_test we get:

  stw     r10,112(r1)
  stw     r9,116(r1)
  lwz     r9,112(r1)
  lwz     r0,116(r1)

x86 is just casting the values so do that instead. Also change __xchg*
and __cmpxchg* to take unsigned values, removing a few sign extensions.

Signed-off-by: default avatarAnton Blanchard <anton@samba.org>
Signed-off-by: default avatarPaul Mackerras <paulus@samba.org>
parent 4721e221

include/asm-ppc64/system.h

+8 −11
Original line number Diff line number Diff line
@@ -158,7 +158,7 @@ static inline int __is_processor(unsigned long pv) 
 * is more like most of the other architectures.

 */

static __inline__ unsigned long

__xchg_u32(volatile int *m, unsigned long val)

__xchg_u32(volatile unsigned int *m, unsigned long val)

{

	unsigned long dummy;
@@ -200,7 +200,7 @@ __xchg_u64(volatile long *m, unsigned long val) 
extern void __xchg_called_with_bad_pointer(void);



static __inline__ unsigned long

__xchg(volatile void *ptr, unsigned long x, int size)

__xchg(volatile void *ptr, unsigned long x, unsigned int size)

{

	switch (size) {

	case 4:
@@ -223,7 +223,7 @@ __xchg(volatile void *ptr, unsigned long x, int size) 
#define __HAVE_ARCH_CMPXCHG	1



static __inline__ unsigned long

__cmpxchg_u32(volatile int *p, int old, int new)

__cmpxchg_u32(volatile unsigned int *p, unsigned long old, unsigned long new)

{

	unsigned int prev;
@@ -271,7 +271,8 @@ __cmpxchg_u64(volatile long *p, unsigned long old, unsigned long new) 
extern void __cmpxchg_called_with_bad_pointer(void);



static __inline__ unsigned long

__cmpxchg(volatile void *ptr, unsigned long old, unsigned long new, int size)

__cmpxchg(volatile void *ptr, unsigned long old, unsigned long new,

	  unsigned int size)

{

	switch (size) {

	case 4:
@@ -284,12 +285,8 @@ __cmpxchg(volatile void *ptr, unsigned long old, unsigned long new, int size) 
}



#define cmpxchg(ptr,o,n)\

  ({									 \

     __typeof__(*(ptr)) _o_ = (o);					 \

     __typeof__(*(ptr)) _n_ = (n);					 \

     (__typeof__(*(ptr))) __cmpxchg((ptr), (unsigned long)_o_,		 \

				    (unsigned long)_n_, sizeof(*(ptr))); \

  })

	((__typeof__(*(ptr)))__cmpxchg((ptr),(unsigned long)(o),\

	(unsigned long)(n),sizeof(*(ptr))))



/*

 * We handle most unaligned accesses in hardware. On the other hand

CRA Git | Maintained and supported by SUSTech CRA and CCSE