* [PATCH v2 2/3] powerpc/atomics: Use immediate operand when possible
2021-04-13 16:36 [PATCH v2 1/3] powerpc/bitops: Use immediate operand when possible Christophe Leroy
@ 2021-04-13 16:36 ` Christophe Leroy
2021-04-13 16:36 ` [PATCH v2 3/3] powerpc/atomics: Remove atomic_inc()/atomic_dec() and friends Christophe Leroy
1 sibling, 0 replies; 3+ messages in thread
From: Christophe Leroy @ 2021-04-13 16:36 UTC (permalink / raw)
To: Benjamin Herrenschmidt, Paul Mackerras, Michael Ellerman
Cc: linux-kernel, linuxppc-dev
Today we get the following code generation for atomic operations:
c001bb2c: 39 20 00 01 li r9,1
c001bb30: 7d 40 18 28 lwarx r10,0,r3
c001bb34: 7d 09 50 50 subf r8,r9,r10
c001bb38: 7d 00 19 2d stwcx. r8,0,r3
c001c7a8: 39 40 00 01 li r10,1
c001c7ac: 7d 00 18 28 lwarx r8,0,r3
c001c7b0: 7c ea 42 14 add r7,r10,r8
c001c7b4: 7c e0 19 2d stwcx. r7,0,r3
By allowing GCC to choose between immediate or regular operation,
we get:
c001bb2c: 7d 20 18 28 lwarx r9,0,r3
c001bb30: 39 49 ff ff addi r10,r9,-1
c001bb34: 7d 40 19 2d stwcx. r10,0,r3
--
c001c7a4: 7d 40 18 28 lwarx r10,0,r3
c001c7a8: 39 0a 00 01 addi r8,r10,1
c001c7ac: 7d 00 19 2d stwcx. r8,0,r3
For "and", the dot form has to be used because "andi" doesn't exist.
For logical operations we use unsigned 16 bits immediate.
For arithmetic operations we use signed 16 bits immediate.
On pmac32_defconfig, it reduces the text by approx another 8 kbytes.
Signed-off-by: Christophe Leroy <christophe.leroy@csgroup.eu>
Acked-by: Segher Boessenkool <segher@kernel.crashing.org>
---
v2: Use "addc/addic"
---
arch/powerpc/include/asm/atomic.h | 56 +++++++++++++++----------------
1 file changed, 28 insertions(+), 28 deletions(-)
diff --git a/arch/powerpc/include/asm/atomic.h b/arch/powerpc/include/asm/atomic.h
index 61c6e8b200e8..eb1bdf14f67c 100644
--- a/arch/powerpc/include/asm/atomic.h
+++ b/arch/powerpc/include/asm/atomic.h
@@ -37,62 +37,62 @@ static __inline__ void atomic_set(atomic_t *v, int i)
__asm__ __volatile__("stw%U0%X0 %1,%0" : "=m"UPD_CONSTR(v->counter) : "r"(i));
}
-#define ATOMIC_OP(op, asm_op) \
+#define ATOMIC_OP(op, asm_op, suffix, sign, ...) \
static __inline__ void atomic_##op(int a, atomic_t *v) \
{ \
int t; \
\
__asm__ __volatile__( \
"1: lwarx %0,0,%3 # atomic_" #op "\n" \
- #asm_op " %0,%2,%0\n" \
+ #asm_op "%I2" suffix " %0,%0,%2\n" \
" stwcx. %0,0,%3 \n" \
" bne- 1b\n" \
: "=&r" (t), "+m" (v->counter) \
- : "r" (a), "r" (&v->counter) \
- : "cc"); \
+ : "r"#sign (a), "r" (&v->counter) \
+ : "cc", ##__VA_ARGS__); \
} \
-#define ATOMIC_OP_RETURN_RELAXED(op, asm_op) \
+#define ATOMIC_OP_RETURN_RELAXED(op, asm_op, suffix, sign, ...) \
static inline int atomic_##op##_return_relaxed(int a, atomic_t *v) \
{ \
int t; \
\
__asm__ __volatile__( \
"1: lwarx %0,0,%3 # atomic_" #op "_return_relaxed\n" \
- #asm_op " %0,%2,%0\n" \
+ #asm_op "%I2" suffix " %0,%0,%2\n" \
" stwcx. %0,0,%3\n" \
" bne- 1b\n" \
: "=&r" (t), "+m" (v->counter) \
- : "r" (a), "r" (&v->counter) \
- : "cc"); \
+ : "r"#sign (a), "r" (&v->counter) \
+ : "cc", ##__VA_ARGS__); \
\
return t; \
}
-#define ATOMIC_FETCH_OP_RELAXED(op, asm_op) \
+#define ATOMIC_FETCH_OP_RELAXED(op, asm_op, suffix, sign, ...) \
static inline int atomic_fetch_##op##_relaxed(int a, atomic_t *v) \
{ \
int res, t; \
\
__asm__ __volatile__( \
"1: lwarx %0,0,%4 # atomic_fetch_" #op "_relaxed\n" \
- #asm_op " %1,%3,%0\n" \
+ #asm_op "%I3" suffix " %1,%0,%3\n" \
" stwcx. %1,0,%4\n" \
" bne- 1b\n" \
: "=&r" (res), "=&r" (t), "+m" (v->counter) \
- : "r" (a), "r" (&v->counter) \
- : "cc"); \
+ : "r"#sign (a), "r" (&v->counter) \
+ : "cc", ##__VA_ARGS__); \
\
return res; \
}
-#define ATOMIC_OPS(op, asm_op) \
- ATOMIC_OP(op, asm_op) \
- ATOMIC_OP_RETURN_RELAXED(op, asm_op) \
- ATOMIC_FETCH_OP_RELAXED(op, asm_op)
+#define ATOMIC_OPS(op, asm_op, suffix, sign, ...) \
+ ATOMIC_OP(op, asm_op, suffix, sign, ##__VA_ARGS__) \
+ ATOMIC_OP_RETURN_RELAXED(op, asm_op, suffix, sign, ##__VA_ARGS__)\
+ ATOMIC_FETCH_OP_RELAXED(op, asm_op, suffix, sign, ##__VA_ARGS__)
-ATOMIC_OPS(add, add)
-ATOMIC_OPS(sub, subf)
+ATOMIC_OPS(add, add, "c", I, "xer")
+ATOMIC_OPS(sub, sub, "c", I, "xer")
#define atomic_add_return_relaxed atomic_add_return_relaxed
#define atomic_sub_return_relaxed atomic_sub_return_relaxed
@@ -101,13 +101,13 @@ ATOMIC_OPS(sub, subf)
#define atomic_fetch_sub_relaxed atomic_fetch_sub_relaxed
#undef ATOMIC_OPS
-#define ATOMIC_OPS(op, asm_op) \
- ATOMIC_OP(op, asm_op) \
- ATOMIC_FETCH_OP_RELAXED(op, asm_op)
+#define ATOMIC_OPS(op, asm_op, suffix, sign) \
+ ATOMIC_OP(op, asm_op, suffix, sign) \
+ ATOMIC_FETCH_OP_RELAXED(op, asm_op, suffix, sign)
-ATOMIC_OPS(and, and)
-ATOMIC_OPS(or, or)
-ATOMIC_OPS(xor, xor)
+ATOMIC_OPS(and, and, ".", K)
+ATOMIC_OPS(or, or, "", K)
+ATOMIC_OPS(xor, xor, "", K)
#define atomic_fetch_and_relaxed atomic_fetch_and_relaxed
#define atomic_fetch_or_relaxed atomic_fetch_or_relaxed
@@ -238,15 +238,15 @@ static __inline__ int atomic_fetch_add_unless(atomic_t *v, int a, int u)
"1: lwarx %0,0,%1 # atomic_fetch_add_unless\n\
cmpw 0,%0,%3 \n\
beq 2f \n\
- add %0,%2,%0 \n"
+ add%I2c %0,%0,%2 \n"
" stwcx. %0,0,%1 \n\
bne- 1b \n"
PPC_ATOMIC_EXIT_BARRIER
-" subf %0,%2,%0 \n\
+" sub%I2c %0,%0,%2 \n\
2:"
: "=&r" (t)
- : "r" (&v->counter), "r" (a), "r" (u)
- : "cc", "memory");
+ : "r" (&v->counter), "rI" (a), "r" (u)
+ : "cc", "memory", "xer");
return t;
}
--
2.25.0
^ permalink raw reply related [flat|nested] 3+ messages in thread
* [PATCH v2 3/3] powerpc/atomics: Remove atomic_inc()/atomic_dec() and friends
2021-04-13 16:36 [PATCH v2 1/3] powerpc/bitops: Use immediate operand when possible Christophe Leroy
2021-04-13 16:36 ` [PATCH v2 2/3] powerpc/atomics: " Christophe Leroy
@ 2021-04-13 16:36 ` Christophe Leroy
1 sibling, 0 replies; 3+ messages in thread
From: Christophe Leroy @ 2021-04-13 16:36 UTC (permalink / raw)
To: Benjamin Herrenschmidt, Paul Mackerras, Michael Ellerman
Cc: linux-kernel, linuxppc-dev
Now that atomic_add() and atomic_sub() handle immediate operands,
atomic_inc() and atomic_dec() have no added value compared to the
generic fallback which calls atomic_add(1) and atomic_sub(1).
Also remove atomic_inc_not_zero() which fallsback to
atomic_add_unless() which itself fallsback to
atomic_fetch_add_unless() which now handles immediate operands.
Signed-off-by: Christophe Leroy <christophe.leroy@csgroup.eu>
---
v2: New
---
arch/powerpc/include/asm/atomic.h | 95 -------------------------------
1 file changed, 95 deletions(-)
diff --git a/arch/powerpc/include/asm/atomic.h b/arch/powerpc/include/asm/atomic.h
index eb1bdf14f67c..00ba5d9e837b 100644
--- a/arch/powerpc/include/asm/atomic.h
+++ b/arch/powerpc/include/asm/atomic.h
@@ -118,71 +118,6 @@ ATOMIC_OPS(xor, xor, "", K)
#undef ATOMIC_OP_RETURN_RELAXED
#undef ATOMIC_OP
-static __inline__ void atomic_inc(atomic_t *v)
-{
- int t;
-
- __asm__ __volatile__(
-"1: lwarx %0,0,%2 # atomic_inc\n\
- addic %0,%0,1\n"
-" stwcx. %0,0,%2 \n\
- bne- 1b"
- : "=&r" (t), "+m" (v->counter)
- : "r" (&v->counter)
- : "cc", "xer");
-}
-#define atomic_inc atomic_inc
-
-static __inline__ int atomic_inc_return_relaxed(atomic_t *v)
-{
- int t;
-
- __asm__ __volatile__(
-"1: lwarx %0,0,%2 # atomic_inc_return_relaxed\n"
-" addic %0,%0,1\n"
-" stwcx. %0,0,%2\n"
-" bne- 1b"
- : "=&r" (t), "+m" (v->counter)
- : "r" (&v->counter)
- : "cc", "xer");
-
- return t;
-}
-
-static __inline__ void atomic_dec(atomic_t *v)
-{
- int t;
-
- __asm__ __volatile__(
-"1: lwarx %0,0,%2 # atomic_dec\n\
- addic %0,%0,-1\n"
-" stwcx. %0,0,%2\n\
- bne- 1b"
- : "=&r" (t), "+m" (v->counter)
- : "r" (&v->counter)
- : "cc", "xer");
-}
-#define atomic_dec atomic_dec
-
-static __inline__ int atomic_dec_return_relaxed(atomic_t *v)
-{
- int t;
-
- __asm__ __volatile__(
-"1: lwarx %0,0,%2 # atomic_dec_return_relaxed\n"
-" addic %0,%0,-1\n"
-" stwcx. %0,0,%2\n"
-" bne- 1b"
- : "=&r" (t), "+m" (v->counter)
- : "r" (&v->counter)
- : "cc", "xer");
-
- return t;
-}
-
-#define atomic_inc_return_relaxed atomic_inc_return_relaxed
-#define atomic_dec_return_relaxed atomic_dec_return_relaxed
-
#define atomic_cmpxchg(v, o, n) (cmpxchg(&((v)->counter), (o), (n)))
#define atomic_cmpxchg_relaxed(v, o, n) \
cmpxchg_relaxed(&((v)->counter), (o), (n))
@@ -252,36 +187,6 @@ static __inline__ int atomic_fetch_add_unless(atomic_t *v, int a, int u)
}
#define atomic_fetch_add_unless atomic_fetch_add_unless
-/**
- * atomic_inc_not_zero - increment unless the number is zero
- * @v: pointer of type atomic_t
- *
- * Atomically increments @v by 1, so long as @v is non-zero.
- * Returns non-zero if @v was non-zero, and zero otherwise.
- */
-static __inline__ int atomic_inc_not_zero(atomic_t *v)
-{
- int t1, t2;
-
- __asm__ __volatile__ (
- PPC_ATOMIC_ENTRY_BARRIER
-"1: lwarx %0,0,%2 # atomic_inc_not_zero\n\
- cmpwi 0,%0,0\n\
- beq- 2f\n\
- addic %1,%0,1\n"
-" stwcx. %1,0,%2\n\
- bne- 1b\n"
- PPC_ATOMIC_EXIT_BARRIER
- "\n\
-2:"
- : "=&r" (t1), "=&r" (t2)
- : "r" (&v->counter)
- : "cc", "xer", "memory");
-
- return t1;
-}
-#define atomic_inc_not_zero(v) atomic_inc_not_zero((v))
-
/*
* Atomically test *v and decrement if it is greater than 0.
* The function returns the old value of *v minus 1, even if
--
2.25.0
^ permalink raw reply related [flat|nested] 3+ messages in thread