Commit a7bafa79 authored by Mark Rutland's avatar Mark Rutland Committed by Peter Zijlstra
Browse files

locking/atomic: hexagon: remove redundant arch_atomic_cmpxchg 



Hexagon's implementation of arch_atomic_cmpxchg() is identical to its
implementation of arch_cmpxchg(). Have it define arch_atomic_cmpxchg()
in terms of arch_cmpxchg(), matching what it does for arch_atomic_xchg()
and arch_xchg().

At the same time, remove the kerneldoc comments for hexagon's
arch_atomic_xchg() and arch_atomic_cmpxchg(). The arch_atomic_*()
namespace is shared by all architectures and the API should be
documented centrally, and the comments aren't all that helpful as-is.

There should be no functional change as a result of this patch.

Signed-off-by: default avatarMark Rutland <mark.rutland@arm.com>
Signed-off-by: default avatarPeter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: default avatarKees Cook <keescook@chromium.org>
Link: https://lore.kernel.org/r/20230605070124.3741859-4-mark.rutland@arm.com
parent 14d72d4b

arch/hexagon/include/asm/atomic.h

+4 −42
Original line number Diff line number Diff line
@@ -36,49 +36,11 @@ static inline void arch_atomic_set(atomic_t *v, int new) 
 */

#define arch_atomic_read(v)		READ_ONCE((v)->counter)



/**

 * arch_atomic_xchg - atomic

 * @v: pointer to memory to change

 * @new: new value (technically passed in a register -- see xchg)

 */

#define arch_atomic_xchg(v, new)	(arch_xchg(&((v)->counter), (new)))





/**

 * arch_atomic_cmpxchg - atomic compare-and-exchange values

 * @v: pointer to value to change

 * @old:  desired old value to match

 * @new:  new value to put in

 *

 * Parameters are then pointer, value-in-register, value-in-register,

 * and the output is the old value.

 *

 * Apparently this is complicated for archs that don't support

 * the memw_locked like we do (or it's broken or whatever).

 *

 * Kind of the lynchpin of the rest of the generically defined routines.

 * Remember V2 had that bug with dotnew predicate set by memw_locked.

 *

 * "old" is "expected" old val, __oldval is actual old value

 */

static inline int arch_atomic_cmpxchg(atomic_t *v, int old, int new)

{

	int __oldval;

#define arch_atomic_xchg(v, new)					\

	(arch_xchg(&((v)->counter), (new)))



	asm volatile(

		"1:	%0 = memw_locked(%1);\n"

		"	{ P0 = cmp.eq(%0,%2);\n"

		"	  if (!P0.new) jump:nt 2f; }\n"

		"	memw_locked(%1,P0) = %3;\n"

		"	if (!P0) jump 1b;\n"

		"2:\n"

		: "=&r" (__oldval)

		: "r" (&v->counter), "r" (old), "r" (new)

		: "memory", "p0"

	);



	return __oldval;

}

#define arch_atomic_cmpxchg(v, old, new)				\

	(arch_cmpxchg(&((v)->counter), (old), (new)))



#define ATOMIC_OP(op)							\

static inline void arch_atomic_##op(int i, atomic_t *v)			\