linux_dsm_epyc7002/arch/s390/include/asm/atomic_ops.h
Martin Schwidefsky eb3b7b848f s390/rwlock: introduce rwlock wait queueing
Like the common queued rwlock code the s390 implementation uses the
queued spinlock code on a spinlock_t embedded in the rwlock_t to achieve
the queueing. The encoding of the rwlock_t differs though, the counter
field in the rwlock_t is split into two parts. The upper two bytes hold
the write bit and the write wait counter, the lower two bytes hold the
read counter.

The arch_read_lock operation works exactly like the common qrwlock but
the enqueue operation for a writer follows a diffent logic. After the
failed inline try to get the rwlock in write, the writer first increases
the write wait counter, acquires the wait spin_lock for the queueing,
and then loops until there are no readers and the write bit is zero.
Without the write wait counter a CPU that just released the rwlock
could immediately reacquire the lock in the inline code, bypassing all
outstanding read and write waiters. For s390 this would cause massive
imbalances in favour of writers in case of a contended rwlock.

Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
2017-09-28 07:29:44 +02:00

143 lines
4.0 KiB
C

/*
* Low level function for atomic operations
*
* Copyright IBM Corp. 1999, 2016
*/
#ifndef __ARCH_S390_ATOMIC_OPS__
#define __ARCH_S390_ATOMIC_OPS__
#ifdef CONFIG_HAVE_MARCH_Z196_FEATURES
#define __ATOMIC_OP(op_name, op_type, op_string, op_barrier) \
static inline op_type op_name(op_type val, op_type *ptr) \
{ \
op_type old; \
\
asm volatile( \
op_string " %[old],%[val],%[ptr]\n" \
op_barrier \
: [old] "=d" (old), [ptr] "+Q" (*ptr) \
: [val] "d" (val) : "cc", "memory"); \
return old; \
} \
#define __ATOMIC_OPS(op_name, op_type, op_string) \
__ATOMIC_OP(op_name, op_type, op_string, "\n") \
__ATOMIC_OP(op_name##_barrier, op_type, op_string, "bcr 14,0\n")
__ATOMIC_OPS(__atomic_add, int, "laa")
__ATOMIC_OPS(__atomic_and, int, "lan")
__ATOMIC_OPS(__atomic_or, int, "lao")
__ATOMIC_OPS(__atomic_xor, int, "lax")
__ATOMIC_OPS(__atomic64_add, long, "laag")
__ATOMIC_OPS(__atomic64_and, long, "lang")
__ATOMIC_OPS(__atomic64_or, long, "laog")
__ATOMIC_OPS(__atomic64_xor, long, "laxg")
#undef __ATOMIC_OPS
#undef __ATOMIC_OP
#define __ATOMIC_CONST_OP(op_name, op_type, op_string, op_barrier) \
static inline void op_name(op_type val, op_type *ptr) \
{ \
asm volatile( \
op_string " %[ptr],%[val]\n" \
op_barrier \
: [ptr] "+Q" (*ptr) : [val] "i" (val) : "cc", "memory");\
}
#define __ATOMIC_CONST_OPS(op_name, op_type, op_string) \
__ATOMIC_CONST_OP(op_name, op_type, op_string, "\n") \
__ATOMIC_CONST_OP(op_name##_barrier, op_type, op_string, "bcr 14,0\n")
__ATOMIC_CONST_OPS(__atomic_add_const, int, "asi")
__ATOMIC_CONST_OPS(__atomic64_add_const, long, "agsi")
#undef __ATOMIC_CONST_OPS
#undef __ATOMIC_CONST_OP
#else /* CONFIG_HAVE_MARCH_Z196_FEATURES */
#define __ATOMIC_OP(op_name, op_string) \
static inline int op_name(int val, int *ptr) \
{ \
int old, new; \
\
asm volatile( \
"0: lr %[new],%[old]\n" \
op_string " %[new],%[val]\n" \
" cs %[old],%[new],%[ptr]\n" \
" jl 0b" \
: [old] "=d" (old), [new] "=&d" (new), [ptr] "+Q" (*ptr)\
: [val] "d" (val), "0" (*ptr) : "cc", "memory"); \
return old; \
}
#define __ATOMIC_OPS(op_name, op_string) \
__ATOMIC_OP(op_name, op_string) \
__ATOMIC_OP(op_name##_barrier, op_string)
__ATOMIC_OPS(__atomic_add, "ar")
__ATOMIC_OPS(__atomic_and, "nr")
__ATOMIC_OPS(__atomic_or, "or")
__ATOMIC_OPS(__atomic_xor, "xr")
#undef __ATOMIC_OPS
#define __ATOMIC64_OP(op_name, op_string) \
static inline long op_name(long val, long *ptr) \
{ \
long old, new; \
\
asm volatile( \
"0: lgr %[new],%[old]\n" \
op_string " %[new],%[val]\n" \
" csg %[old],%[new],%[ptr]\n" \
" jl 0b" \
: [old] "=d" (old), [new] "=&d" (new), [ptr] "+Q" (*ptr)\
: [val] "d" (val), "0" (*ptr) : "cc", "memory"); \
return old; \
}
#define __ATOMIC64_OPS(op_name, op_string) \
__ATOMIC64_OP(op_name, op_string) \
__ATOMIC64_OP(op_name##_barrier, op_string)
__ATOMIC64_OPS(__atomic64_add, "agr")
__ATOMIC64_OPS(__atomic64_and, "ngr")
__ATOMIC64_OPS(__atomic64_or, "ogr")
__ATOMIC64_OPS(__atomic64_xor, "xgr")
#undef __ATOMIC64_OPS
#define __atomic_add_const(val, ptr) __atomic_add(val, ptr)
#define __atomic_add_const_barrier(val, ptr) __atomic_add(val, ptr)
#define __atomic64_add_const(val, ptr) __atomic64_add(val, ptr)
#define __atomic64_add_const_barrier(val, ptr) __atomic64_add(val, ptr)
#endif /* CONFIG_HAVE_MARCH_Z196_FEATURES */
static inline int __atomic_cmpxchg(int *ptr, int old, int new)
{
return __sync_val_compare_and_swap(ptr, old, new);
}
static inline int __atomic_cmpxchg_bool(int *ptr, int old, int new)
{
return __sync_bool_compare_and_swap(ptr, old, new);
}
static inline long __atomic64_cmpxchg(long *ptr, long old, long new)
{
return __sync_val_compare_and_swap(ptr, old, new);
}
static inline long __atomic64_cmpxchg_bool(long *ptr, long old, long new)
{
return __sync_bool_compare_and_swap(ptr, old, new);
}
#endif /* __ARCH_S390_ATOMIC_OPS__ */