Merge tag 'locking-core-2023-06-27' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip

Pull locking updates from Ingo Molnar:

 - Introduce cmpxchg128() -- aka. the demise of cmpxchg_double()

   The cmpxchg128() family of functions is basically & functionally the
   same as cmpxchg_double(), but with a saner interface.

   Instead of a 6-parameter horror that forced u128 - u64/u64-halves
   layout details on the interface and exposed users to complexity,
   fragility & bugs, use a natural 3-parameter interface with u128
   types.

 - Restructure the generated atomic headers, and add kerneldoc comments
   for all of the generic atomic{,64,_long}_t operations.

   The generated definitions are much cleaner now, and come with
   documentation.

 - Implement lock_set_cmp_fn() on lockdep, for defining an ordering when
   taking multiple locks of the same type.

   This gets rid of one use of lockdep_set_novalidate_class() in the
   bcache code.

 - Fix raw_cpu_generic_try_cmpxchg() bug due to an unintended variable
   shadowing generating garbage code on Clang on certain ARM builds.

* tag 'locking-core-2023-06-27' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: (43 commits)
  locking/atomic: scripts: fix ${atomic}_dec_if_positive() kerneldoc
  percpu: Fix self-assignment of __old in raw_cpu_generic_try_cmpxchg()
  locking/atomic: treewide: delete arch_atomic_*() kerneldoc
  locking/atomic: docs: Add atomic operations to the driver basic API documentation
  locking/atomic: scripts: generate kerneldoc comments
  docs: scripts: kernel-doc: accept bitwise negation like ~@var
  locking/atomic: scripts: simplify raw_atomic*() definitions
  locking/atomic: scripts: simplify raw_atomic_long*() definitions
  locking/atomic: scripts: split pfx/name/sfx/order
  locking/atomic: scripts: restructure fallback ifdeffery
  locking/atomic: scripts: build raw_atomic_long*() directly
  locking/atomic: treewide: use raw_atomic*_<op>()
  locking/atomic: scripts: add trivial raw_atomic*_<op>()
  locking/atomic: scripts: factor out order template generation
  locking/atomic: scripts: remove leftover "${mult}"
  locking/atomic: scripts: remove bogus order parameter
  locking/atomic: xtensa: add preprocessor symbols
  locking/atomic: x86: add preprocessor symbols
  locking/atomic: sparc: add preprocessor symbols
  locking/atomic: sh: add preprocessor symbols
  ...

53 files changed, 533 insertions, 930 deletions

diff --git a/arch/alpha/include/asm/atomic.h b/arch/alpha/include/asm/atomic.h
index f2861a43a61e..cbd9244571af 100644
--- a/arch/alpha/include/asm/atomic.h
+++ b/arch/alpha/include/asm/atomic.h
@@ -200,25 +200,6 @@ ATOMIC_OPS(xor, xor)
 #undef ATOMIC_OP_RETURN
 #undef ATOMIC_OP
 
-#define arch_atomic64_cmpxchg(v, old, new) \
-	(arch_cmpxchg(&((v)->counter), old, new))
-#define arch_atomic64_xchg(v, new) \
-	(arch_xchg(&((v)->counter), new))
-
-#define arch_atomic_cmpxchg(v, old, new) \
-	(arch_cmpxchg(&((v)->counter), old, new))
-#define arch_atomic_xchg(v, new) \
-	(arch_xchg(&((v)->counter), new))
-
-/**
- * arch_atomic_fetch_add_unless - add unless the number is a given value
- * @v: pointer of type atomic_t
- * @a: the amount to add to v...
- * @u: ...unless v is equal to u.
- *
- * Atomically adds @a to @v, so long as it was not @u.
- * Returns the old value of @v.
- */
 static __inline__ int arch_atomic_fetch_add_unless(atomic_t *v, int a, int u)
 {
 	int c, new, old;
@@ -242,15 +223,6 @@ static __inline__ int arch_atomic_fetch_add_unless(atomic_t *v, int a, int u)
 }
 #define arch_atomic_fetch_add_unless arch_atomic_fetch_add_unless
 
-/**
- * arch_atomic64_fetch_add_unless - add unless the number is a given value
- * @v: pointer of type atomic64_t
- * @a: the amount to add to v...
- * @u: ...unless v is equal to u.
- *
- * Atomically adds @a to @v, so long as it was not @u.
- * Returns the old value of @v.
- */
 static __inline__ s64 arch_atomic64_fetch_add_unless(atomic64_t *v, s64 a, s64 u)
 {
 	s64 c, new, old;
@@ -274,13 +246,6 @@ static __inline__ s64 arch_atomic64_fetch_add_unless(atomic64_t *v, s64 a, s64 u
 }
 #define arch_atomic64_fetch_add_unless arch_atomic64_fetch_add_unless
 
-/*
- * arch_atomic64_dec_if_positive - decrement by 1 if old value positive
- * @v: pointer of type atomic_t
- *
- * The function returns the old value of *v minus 1, even if
- * the atomic variable, v, was not decremented.
- */
 static inline s64 arch_atomic64_dec_if_positive(atomic64_t *v)
 {
 	s64 old, tmp;
diff --git a/arch/arc/include/asm/atomic-spinlock.h b/arch/arc/include/asm/atomic-spinlock.h
index 2c830347bfb4..89d12a60f84c 100644
--- a/arch/arc/include/asm/atomic-spinlock.h
+++ b/arch/arc/include/asm/atomic-spinlock.h
@@ -81,6 +81,11 @@ static inline int arch_atomic_fetch_##op(int i, atomic_t *v)		\
 ATOMIC_OPS(add, +=, add)
 ATOMIC_OPS(sub, -=, sub)
 
+#define arch_atomic_fetch_add		arch_atomic_fetch_add
+#define arch_atomic_fetch_sub		arch_atomic_fetch_sub
+#define arch_atomic_add_return		arch_atomic_add_return
+#define arch_atomic_sub_return		arch_atomic_sub_return
+
 #undef ATOMIC_OPS
 #define ATOMIC_OPS(op, c_op, asm_op)					\
 	ATOMIC_OP(op, c_op, asm_op)					\
@@ -92,7 +97,11 @@ ATOMIC_OPS(or, |=, or)
 ATOMIC_OPS(xor, ^=, xor)
 
 #define arch_atomic_andnot		arch_atomic_andnot
+
+#define arch_atomic_fetch_and		arch_atomic_fetch_and
 #define arch_atomic_fetch_andnot	arch_atomic_fetch_andnot
+#define arch_atomic_fetch_or		arch_atomic_fetch_or
+#define arch_atomic_fetch_xor		arch_atomic_fetch_xor
 
 #undef ATOMIC_OPS
 #undef ATOMIC_FETCH_OP
diff --git a/arch/arc/include/asm/atomic.h b/arch/arc/include/asm/atomic.h
index 52ee51e1ff7c..592d7fffc223 100644
--- a/arch/arc/include/asm/atomic.h
+++ b/arch/arc/include/asm/atomic.h
@@ -22,30 +22,6 @@
 #include <asm/atomic-spinlock.h>
 #endif
 
-#define arch_atomic_cmpxchg(v, o, n)					\
-({									\
-	arch_cmpxchg(&((v)->counter), (o), (n));			\
-})
-
-#ifdef arch_cmpxchg_relaxed
-#define arch_atomic_cmpxchg_relaxed(v, o, n)				\
-({									\
-	arch_cmpxchg_relaxed(&((v)->counter), (o), (n));		\
-})
-#endif
-
-#define arch_atomic_xchg(v, n)						\
-({									\
-	arch_xchg(&((v)->counter), (n));				\
-})
-
-#ifdef arch_xchg_relaxed
-#define arch_atomic_xchg_relaxed(v, n)					\
-({									\
-	arch_xchg_relaxed(&((v)->counter), (n));			\
-})
-#endif
-
 /*
  * 64-bit atomics
  */
diff --git a/arch/arc/include/asm/atomic64-arcv2.h b/arch/arc/include/asm/atomic64-arcv2.h
index c5a8010fdc97..6b6db981967a 100644
--- a/arch/arc/include/asm/atomic64-arcv2.h
+++ b/arch/arc/include/asm/atomic64-arcv2.h
@@ -159,6 +159,7 @@ arch_atomic64_cmpxchg(atomic64_t *ptr, s64 expected, s64 new)
 
 	return prev;
 }
+#define arch_atomic64_cmpxchg arch_atomic64_cmpxchg
 
 static inline s64 arch_atomic64_xchg(atomic64_t *ptr, s64 new)
 {
@@ -179,14 +180,7 @@ static inline s64 arch_atomic64_xchg(atomic64_t *ptr, s64 new)
 
 	return prev;
 }
-
-/**
- * arch_atomic64_dec_if_positive - decrement by 1 if old value positive
- * @v: pointer of type atomic64_t
- *
- * The function returns the old value of *v minus 1, even if
- * the atomic variable, v, was not decremented.
- */
+#define arch_atomic64_xchg arch_atomic64_xchg
 
 static inline s64 arch_atomic64_dec_if_positive(atomic64_t *v)
 {
@@ -212,15 +206,6 @@ static inline s64 arch_atomic64_dec_if_positive(atomic64_t *v)
 }
 #define arch_atomic64_dec_if_positive arch_atomic64_dec_if_positive
 
-/**
- * arch_atomic64_fetch_add_unless - add unless the number is a given value
- * @v: pointer of type atomic64_t
- * @a: the amount to add to v...
- * @u: ...unless v is equal to u.
- *
- * Atomically adds @a to @v, if it was not @u.
- * Returns the old value of @v
- */
 static inline s64 arch_atomic64_fetch_add_unless(atomic64_t *v, s64 a, s64 u)
 {
 	s64 old, temp;
diff --git a/arch/arm/include/asm/assembler.h b/arch/arm/include/asm/assembler.h
index 505a306e0271..aebe2c8f6a68 100644
--- a/arch/arm/include/asm/assembler.h
+++ b/arch/arm/include/asm/assembler.h
@@ -394,6 +394,23 @@ ALT_UP_B(.L0_\@)
 #endif
 	.endm
 
+/*
+ * Raw SMP data memory barrier
+ */
+	.macro	__smp_dmb mode
+#if __LINUX_ARM_ARCH__ >= 7
+	.ifeqs "\mode","arm"
+	dmb	ish
+	.else
+	W(dmb)	ish
+	.endif
+#elif __LINUX_ARM_ARCH__ == 6
+	mcr	p15, 0, r0, c7, c10, 5	@ dmb
+#else
+	.error "Incompatible SMP platform"
+#endif
+	.endm
+
 #if defined(CONFIG_CPU_V7M)
 	/*
 	 * setmode is used to assert to be in svc mode during boot. For v7-M
diff --git a/arch/arm/include/asm/atomic.h b/arch/arm/include/asm/atomic.h
index db8512d9a918..f0e3b01afa74 100644
--- a/arch/arm/include/asm/atomic.h
+++ b/arch/arm/include/asm/atomic.h
@@ -197,6 +197,16 @@ static inline int arch_atomic_fetch_##op(int i, atomic_t *v)		\
 	return val;							\
 }
 
+#define arch_atomic_add_return			arch_atomic_add_return
+#define arch_atomic_sub_return			arch_atomic_sub_return
+#define arch_atomic_fetch_add			arch_atomic_fetch_add
+#define arch_atomic_fetch_sub			arch_atomic_fetch_sub
+
+#define arch_atomic_fetch_and			arch_atomic_fetch_and
+#define arch_atomic_fetch_andnot		arch_atomic_fetch_andnot
+#define arch_atomic_fetch_or			arch_atomic_fetch_or
+#define arch_atomic_fetch_xor			arch_atomic_fetch_xor
+
 static inline int arch_atomic_cmpxchg(atomic_t *v, int old, int new)
 {
 	int ret;
@@ -210,8 +220,7 @@ static inline int arch_atomic_cmpxchg(atomic_t *v, int old, int new)
 
 	return ret;
 }
-
-#define arch_atomic_fetch_andnot		arch_atomic_fetch_andnot
+#define arch_atomic_cmpxchg arch_atomic_cmpxchg
 
 #endif /* __LINUX_ARM_ARCH__ */
 
@@ -240,8 +249,6 @@ ATOMIC_OPS(xor, ^=, eor)
 #undef ATOMIC_OP_RETURN
 #undef ATOMIC_OP
 
-#define arch_atomic_xchg(v, new) (arch_xchg(&((v)->counter), new))
-
 #ifndef CONFIG_GENERIC_ATOMIC64
 typedef struct {
 	s64 counter;
diff --git a/arch/arm/include/asm/sync_bitops.h b/arch/arm/include/asm/sync_bitops.h
index 6f5d627c44a3..f46b3c570f92 100644
--- a/arch/arm/include/asm/sync_bitops.h
+++ b/arch/arm/include/asm/sync_bitops.h
@@ -14,14 +14,35 @@
  * ops which are SMP safe even on a UP kernel.
  */
 
+/*
+ * Unordered
+ */
+
 #define sync_set_bit(nr, p)		_set_bit(nr, p)
 #define sync_clear_bit(nr, p)		_clear_bit(nr, p)
 #define sync_change_bit(nr, p)		_change_bit(nr, p)
-#define sync_test_and_set_bit(nr, p)	_test_and_set_bit(nr, p)
-#define sync_test_and_clear_bit(nr, p)	_test_and_clear_bit(nr, p)
-#define sync_test_and_change_bit(nr, p)	_test_and_change_bit(nr, p)
 #define sync_test_bit(nr, addr)		test_bit(nr, addr)
-#define arch_sync_cmpxchg		arch_cmpxchg
 
+/*
+ * Fully ordered
+ */
+
+int _sync_test_and_set_bit(int nr, volatile unsigned long * p);
+#define sync_test_and_set_bit(nr, p)	_sync_test_and_set_bit(nr, p)
+
+int _sync_test_and_clear_bit(int nr, volatile unsigned long * p);
+#define sync_test_and_clear_bit(nr, p)	_sync_test_and_clear_bit(nr, p)
+
+int _sync_test_and_change_bit(int nr, volatile unsigned long * p);
+#define sync_test_and_change_bit(nr, p)	_sync_test_and_change_bit(nr, p)
+
+#define arch_sync_cmpxchg(ptr, old, new)				\
+({									\
+	__typeof__(*(ptr)) __ret;					\
+	__smp_mb__before_atomic();					\
+	__ret = arch_cmpxchg_relaxed((ptr), (old), (new));		\
+	__smp_mb__after_atomic();					\
+	__ret;								\
+})
 
 #endif
diff --git a/arch/arm/lib/bitops.h b/arch/arm/lib/bitops.h
index 95bd35991288..f069d1b2318e 100644
--- a/arch/arm/lib/bitops.h
+++ b/arch/arm/lib/bitops.h
@@ -28,7 +28,7 @@ UNWIND(	.fnend		)
 ENDPROC(\name		)
 	.endm
 
-	.macro	testop, name, instr, store
+	.macro	__testop, name, instr, store, barrier
 ENTRY(	\name		)
 UNWIND(	.fnstart	)
 	ands	ip, r1, #3
@@ -38,7 +38,7 @@ UNWIND(	.fnstart	)
 	mov	r0, r0, lsr #5
 	add	r1, r1, r0, lsl #2	@ Get word offset
 	mov	r3, r2, lsl r3		@ create mask
-	smp_dmb
+	\barrier
 #if __LINUX_ARM_ARCH__ >= 7 && defined(CONFIG_SMP)
 	.arch_extension	mp
 	ALT_SMP(W(pldw)	[r1])
@@ -50,13 +50,21 @@ UNWIND(	.fnstart	)
 	strex	ip, r2, [r1]
 	cmp	ip, #0
 	bne	1b
-	smp_dmb
+	\barrier
 	cmp	r0, #0
 	movne	r0, #1
 2:	bx	lr
 UNWIND(	.fnend		)
 ENDPROC(\name		)
 	.endm
+
+	.macro	testop, name, instr, store
+	__testop \name, \instr, \store, smp_dmb
+	.endm
+
+	.macro	sync_testop, name, instr, store
+	__testop \name, \instr, \store, __smp_dmb
+	.endm
 #else
 	.macro	bitop, name, instr
 ENTRY(	\name		)
diff --git a/arch/arm/lib/testchangebit.S b/arch/arm/lib/testchangebit.S
index 4ebecc67e6e0..f13fe9bc2399 100644
--- a/arch/arm/lib/testchangebit.S
+++ b/arch/arm/lib/testchangebit.S
@@ -10,3 +10,7 @@
                 .text
 
 testop	_test_and_change_bit, eor, str
+
+#if __LINUX_ARM_ARCH__ >= 6
+sync_testop	_sync_test_and_change_bit, eor, str
+#endif
diff --git a/arch/arm/lib/testclearbit.S b/arch/arm/lib/testclearbit.S
index 009afa0f5b4a..4d2c5ca620eb 100644
--- a/arch/arm/lib/testclearbit.S
+++ b/arch/arm/lib/testclearbit.S
@@ -10,3 +10,7 @@
                 .text
 
 testop	_test_and_clear_bit, bicne, strne
+
+#if __LINUX_ARM_ARCH__ >= 6
+sync_testop	_sync_test_and_clear_bit, bicne, strne
+#endif
diff --git a/arch/arm/lib/testsetbit.S b/arch/arm/lib/testsetbit.S
index f3192e55acc8..649dbab65d8d 100644
--- a/arch/arm/lib/testsetbit.S
+++ b/arch/arm/lib/testsetbit.S
@@ -10,3 +10,7 @@
                 .text
 
 testop	_test_and_set_bit, orreq, streq
+
+#if __LINUX_ARM_ARCH__ >= 6
+sync_testop	_sync_test_and_set_bit, orreq, streq
+#endif
diff --git a/arch/arm64/include/asm/atomic.h b/arch/arm64/include/asm/atomic.h
index c9979273d389..400d279e0f8d 100644
--- a/arch/arm64/include/asm/atomic.h
+++ b/arch/arm64/include/asm/atomic.h
@@ -142,24 +142,6 @@ static __always_inline long arch_atomic64_dec_if_positive(atomic64_t *v)
 #define arch_atomic_fetch_xor_release		arch_atomic_fetch_xor_release
 #define arch_atomic_fetch_xor			arch_atomic_fetch_xor
 
-#define arch_atomic_xchg_relaxed(v, new) \
-	arch_xchg_relaxed(&((v)->counter), (new))
-#define arch_atomic_xchg_acquire(v, new) \
-	arch_xchg_acquire(&((v)->counter), (new))
-#define arch_atomic_xchg_release(v, new) \
-	arch_xchg_release(&((v)->counter), (new))
-#define arch_atomic_xchg(v, new) \
-	arch_xchg(&((v)->counter), (new))
-
-#define arch_atomic_cmpxchg_relaxed(v, old, new) \
-	arch_cmpxchg_relaxed(&((v)->counter), (old), (new))
-#define arch_atomic_cmpxchg_acquire(v, old, new) \
-	arch_cmpxchg_acquire(&((v)->counter), (old), (new))
-#define arch_atomic_cmpxchg_release(v, old, new) \
-	arch_cmpxchg_release(&((v)->counter), (old), (new))
-#define arch_atomic_cmpxchg(v, old, new) \
-	arch_cmpxchg(&((v)->counter), (old), (new))
-
 #define arch_atomic_andnot			arch_atomic_andnot
 
 /*
@@ -209,16 +191,6 @@ static __always_inline long arch_atomic64_dec_if_positive(atomic64_t *v)
 #define arch_atomic64_fetch_xor_release		arch_atomic64_fetch_xor_release
 #define arch_atomic64_fetch_xor			arch_atomic64_fetch_xor
 
-#define arch_atomic64_xchg_relaxed		arch_atomic_xchg_relaxed
-#define arch_atomic64_xchg_acquire		arch_atomic_xchg_acquire
-#define arch_atomic64_xchg_release		arch_atomic_xchg_release
-#define arch_atomic64_xchg			arch_atomic_xchg
-
-#define arch_atomic64_cmpxchg_relaxed		arch_atomic_cmpxchg_relaxed
-#define arch_atomic64_cmpxchg_acquire		arch_atomic_cmpxchg_acquire
-#define arch_atomic64_cmpxchg_release		arch_atomic_cmpxchg_release
-#define arch_atomic64_cmpxchg			arch_atomic_cmpxchg
-
 #define arch_atomic64_andnot			arch_atomic64_andnot
 
 #define arch_atomic64_dec_if_positive		arch_atomic64_dec_if_positive
diff --git a/arch/arm64/include/asm/atomic_ll_sc.h b/arch/arm64/include/asm/atomic_ll_sc.h
index cbb3d961123b..89d2ba272359 100644
--- a/arch/arm64/include/asm/atomic_ll_sc.h
+++ b/arch/arm64/include/asm/atomic_ll_sc.h
@@ -294,38 +294,46 @@ __CMPXCHG_CASE( ,  ,  mb_, 64, dmb ish,  , l, "memory", L)
 
 #undef __CMPXCHG_CASE
 
-#define __CMPXCHG_DBL(name, mb, rel, cl)				\
-static __always_inline long						\
-__ll_sc__cmpxchg_double##name(unsigned long old1,			\
-				      unsigned long old2,		\
-				      unsigned long new1,		\
-				      unsigned long new2,		\
-				      volatile void *ptr)		\
+union __u128_halves {
+	u128 full;
+	struct {
+		u64 low, high;
+	};
+};
+
+#define __CMPXCHG128(name, mb, rel, cl...)                             \
+static __always_inline u128						\
+__ll_sc__cmpxchg128##name(volatile u128 *ptr, u128 old, u128 new)	\
 {									\
-	unsigned long tmp, ret;						\
+	union __u128_halves r, o = { .full = (old) },			\
+			       n = { .full = (new) };			\
+       unsigned int tmp;                                               \
 									\
-	asm volatile("// __cmpxchg_double" #name "\n"			\
-	"	prfm	pstl1strm, %2\n"				\
-	"1:	ldxp	%0, %1, %2\n"					\
-	"	eor	%0, %0, %3\n"					\
-	"	eor	%1, %1, %4\n"					\
-	"	orr	%1, %0, %1\n"					\
-	"	cbnz	%1, 2f\n"					\
-	"	st" #rel "xp	%w0, %5, %6, %2\n"			\
-	"	cbnz	%w0, 1b\n"					\
+	asm volatile("// __cmpxchg128" #name "\n"			\
+       "       prfm    pstl1strm, %[v]\n"                              \
+       "1:     ldxp    %[rl], %[rh], %[v]\n"                           \
+       "       cmp     %[rl], %[ol]\n"                                 \
+       "       ccmp    %[rh], %[oh], 0, eq\n"                          \
+       "       b.ne    2f\n"                                           \
+       "       st" #rel "xp    %w[tmp], %[nl], %[nh], %[v]\n"          \
+       "       cbnz    %w[tmp], 1b\n"                                  \
 	"	" #mb "\n"						\
 	"2:"								\
-	: "=&r" (tmp), "=&r" (ret), "+Q" (*(__uint128_t *)ptr)		\
-	: "r" (old1), "r" (old2), "r" (new1), "r" (new2)		\
-	: cl);								\
+       : [v] "+Q" (*(u128 *)ptr),                                      \
+         [rl] "=&r" (r.low), [rh] "=&r" (r.high),                      \
+         [tmp] "=&r" (tmp)                                             \
+       : [ol] "r" (o.low), [oh] "r" (o.high),                          \
+         [nl] "r" (n.low), [nh] "r" (n.high)                           \
+       : "cc", ##cl);                                                  \
 									\
-	return ret;							\
+	return r.full;							\
 }
 
-__CMPXCHG_DBL(   ,        ,  ,         )
-__CMPXCHG_DBL(_mb, dmb ish, l, "memory")
+__CMPXCHG128(   ,        ,  )
+__CMPXCHG128(_mb, dmb ish, l, "memory")
+
+#undef __CMPXCHG128
 
-#undef __CMPXCHG_DBL
 #undef K
 
 #endif	/* __ASM_ATOMIC_LL_SC_H */
diff --git a/arch/arm64/include/asm/atomic_lse.h b/arch/arm64/include/asm/atomic_lse.h
index 319958b95cfd..87f568a94e55 100644
--- a/arch/arm64/include/asm/atomic_lse.h
+++ b/arch/arm64/include/asm/atomic_lse.h
@@ -281,40 +281,35 @@ __CMPXCHG_CASE(x,  ,  mb_, 64, al, "memory")
 
 #undef __CMPXCHG_CASE
 
-#define __CMPXCHG_DBL(name, mb, cl...)					\
-static __always_inline long						\
-__lse__cmpxchg_double##name(unsigned long old1,				\
-					 unsigned long old2,		\
-					 unsigned long new1,		\
-					 unsigned long new2,		\
-					 volatile void *ptr)		\
+#define __CMPXCHG128(name, mb, cl...)					\
+static __always_inline u128						\
+__lse__cmpxchg128##name(volatile u128 *ptr, u128 old, u128 new)		\
 {									\
-	unsigned long oldval1 = old1;					\
-	unsigned long oldval2 = old2;					\
-	register unsigned long x0 asm ("x0") = old1;			\
-	register unsigned long x1 asm ("x1") = old2;			\
-	register unsigned long x2 asm ("x2") = new1;			\
-	register unsigned long x3 asm ("x3") = new2;			\
+	union __u128_halves r, o = { .full = (old) },			\
+			       n = { .full = (new) };			\
+	register unsigned long x0 asm ("x0") = o.low;			\
+	register unsigned long x1 asm ("x1") = o.high;			\
+	register unsigned long x2 asm ("x2") = n.low;			\
+	register unsigned long x3 asm ("x3") = n.high;			\
 	register unsigned long x4 asm ("x4") = (unsigned long)ptr;	\
 									\
 	asm volatile(							\
 	__LSE_PREAMBLE							\
 	"	casp" #mb "\t%[old1], %[old2], %[new1], %[new2], %[v]\n"\
-	"	eor	%[old1], %[old1], %[oldval1]\n"			\
-	"	eor	%[old2], %[old2], %[oldval2]\n"			\
-	"	orr	%[old1], %[old1], %[old2]"			\
 	: [old1] "+&r" (x0), [old2] "+&r" (x1),				\
-	  [v] "+Q" (*(__uint128_t *)ptr)				\
+	  [v] "+Q" (*(u128 *)ptr)					\
 	: [new1] "r" (x2), [new2] "r" (x3), [ptr] "r" (x4),		\
-	  [oldval1] "r" (oldval1), [oldval2] "r" (oldval2)		\
+	  [oldval1] "r" (o.low), [oldval2] "r" (o.high)			\
 	: cl);								\
 									\
-	return x0;							\
+	r.low = x0; r.high = x1;					\
+									\
+	return r.full;							\
 }
 
-__CMPXCHG_DBL(   ,   )
-__CMPXCHG_DBL(_mb, al, "memory")
+__CMPXCHG128(   ,   )
+__CMPXCHG128(_mb, al, "memory")
 
-#undef __CMPXCHG_DBL
+#undef __CMPXCHG128
 
 #endif	/* __ASM_ATOMIC_LSE_H */
diff --git a/arch/arm64/include/asm/cmpxchg.h b/arch/arm64/include/asm/cmpxchg.h
index c6bc5d8ec3ca..d7a540736741 100644
--- a/arch/arm64/include/asm/cmpxchg.h
+++ b/arch/arm64/include/asm/cmpxchg.h
@@ -130,21 +130,18 @@ __CMPXCHG_CASE(mb_, 64)
 
 #undef __CMPXCHG_CASE