aboutsummaryrefslogtreecommitdiff
path: root/arch/ia64/include/asm/spinlock.h
blob: 54ff557d474e49a2376cefe8a78e646fda36cc78 (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
#ifndef _ASM_IA64_SPINLOCK_H
#define _ASM_IA64_SPINLOCK_H

/*
 * Copyright (C) 1998-2003 Hewlett-Packard Co
 *	David Mosberger-Tang <davidm@hpl.hp.com>
 * Copyright (C) 1999 Walt Drummond <drummond@valinux.com>
 *
 * This file is used for SMP configurations only.
 */

#include <linux/compiler.h>
#include <linux/kernel.h>
#include <linux/bitops.h>

#include <linux/atomic.h>
#include <asm/intrinsics.h>

#define arch_spin_lock_init(x)			((x)->lock = 0)

/*
 * Ticket locks are conceptually two parts, one indicating the current head of
 * the queue, and the other indicating the current tail. The lock is acquired
 * by atomically noting the tail and incrementing it by one (thus adding
 * ourself to the queue and noting our position), then waiting until the head
 * becomes equal to the the initial value of the tail.
 * The pad bits in the middle are used to prevent the next_ticket number
 * overflowing into the now_serving number.
 *
 *   31             17  16    15  14                    0
 *  +----------------------------------------------------+
 *  |  now_serving     | padding |   next_ticket         |
 *  +----------------------------------------------------+
 */

#define TICKET_SHIFT	17
#define TICKET_BITS	15
#define	TICKET_MASK	((1 << TICKET_BITS) - 1)

static __always_inline void __ticket_spin_lock(arch_spinlock_t *lock)
{
	int	*p = (int *)&lock->lock, ticket, serve;

	ticket = ia64_fetchadd(1, p, acq);

	if (!(((ticket >> TICKET_SHIFT) ^ ticket) & TICKET_MASK))
		return;

	ia64_invala();

	for (;;) {
		asm volatile ("ld4.c.nc %0=[%1]" : "=r"(serve) : "r"(p) : "memory");

		if (!(((serve >> TICKET_SHIFT) ^ ticket) & TICKET_MASK))
			return;
		cpu_relax();
	}
}

static __always_inline int __ticket_spin_trylock(arch_spinlock_t *lock)
{
	int tmp = ACCESS_ONCE(lock->lock);

	if (!(((tmp >> TICKET_SHIFT) ^ tmp) & TICKET_MASK))
		return ia64_cmpxchg(acq, &lock->lock, tmp, tmp + 1, sizeof (tmp)) == tmp;
	return 0;
}

static __always_inline void __ticket_spin_unlock(arch_spinlock_t *lock)
{
	unsigned short	*p = (unsigned short *)&lock->lock + 1, tmp;

	asm volatile ("ld2.bias %0=[%1]" : "=r"(tmp) : "r"(p));
	ACCESS_ONCE(*p) = (tmp + 2) & ~1;
}

static __always_inline void __ticket_spin_unlock_wait(arch_spinlock_t *lock)
{
	int	*p = (int *)&lock->lock, ticket;

	ia64_invala();

	for (;;) {
		asm volatile ("ld4.c.nc %0=[%1]" : "=r"(ticket) : "r"(p) : "memory");
		if (!(((ticket >> TICKET_SHIFT) ^ ticket) & TICKET_MASK))
			return;
		cpu_relax();
	}
}

static inline int __ticket_spin_is_locked(arch_spinlock_t *lock)
{
	long tmp = ACCESS_ONCE(lock->lock);

	return !!(((tmp >> TICKET_SHIFT) ^ tmp) & TICKET_MASK);
}

static inline int __ticket_spin_is_contended(arch_spinlock_t *lock)
{
	long tmp = ACCESS_ONCE(lock->lock);

	return ((tmp - (tmp >> TICKET_SHIFT)) & TICKET_MASK) > 1;
}

static inline int arch_spin_is_locked(arch_spinlock_t *lock)
{
	return __ticket_spin_is_locked(lock);
}

static inline int arch_spin_is_contended(arch_spinlock_t *lock)
{
	return __ticket_spin_is_contended(lock);
}
#define arch_spin_is_contended	arch_spin_is_contended

static __always_inline void arch_spin_lock(arch_spinlock_t *lock)
{
	__ticket_spin_lock(lock);
}

static __always_inline int arch_spin_trylock(arch_spinlock_t *lock)
{
	return __ticket_spin_trylock(lock);
}

static __always_inline void arch_spin_unlock(arch_spinlock_t *lock)
{
	__ticket_spin_unlock(lock);
}

static __always_inline void arch_spin_lock_flags(arch_spinlock_t *lock,
						  unsigned long flags)
{
	arch_spin_lock(lock);
}

static inline void arch_spin_unlock_wait(arch_spinlock_t *lock)
{
	__ticket_spin_unlock_wait(lock);
}

#define arch_read_can_lock(rw)		(*(volatile int *)(rw) >= 0)
#define arch_write_can_lock(rw)	(*(volatile int *)(rw) == 0)

#ifdef ASM_SUPPORTED

static __always_inline void
arch_read_lock_flags(arch_rwlock_t *lock, unsigned long flags)
{
	__asm__ __volatile__ (
		"tbit.nz p6, p0 = %1,%2\n"
		"br.few 3f\n"
		"1:\n"
		"fetchadd4.rel r2 = [%0], -1;;\n"
		"(p6) ssm psr.i\n"
		"2:\n"
		"hint @pause\n"
		"ld4 r2 = [%0];;\n"
		"cmp4.lt p7,p0 = r2, r0\n"
		"(p7) br.cond.spnt.few 2b\n"
		"(p6) rsm psr.i\n"
		";;\n"
		"3:\n"
		"fetchadd4.acq r2 = [%0], 1;;\n"
		"cmp4.lt p7,p0 = r2, r0\n"
		"(p7) br.cond.spnt.few 1b\n"
		: : "r"(lock), "r"(flags), "i"(IA64_PSR_I_BIT)
		: "p6", "p7", "r2", "memory");
}

#define arch_read_lock(lock) arch_read_lock_flags(lock, 0)

#else /* !ASM_SUPPORTED */

#define arch_read_lock_flags(rw, flags) arch_read_lock(rw)

#define arch_read_lock(rw)								\
do {											\
	arch_rwlock_t *__read_lock_ptr = (rw);						\
											\
	while (unlikely(ia64_fetchadd(1, (int *) __read_lock_ptr, acq) < 0)) {		\
		ia64_fetchadd(-1, (int *) __read_lock_ptr, rel);			\
		while (*(volatile int *)__read_lock_ptr < 0)				\
			cpu_relax();							\
	}										\
} while (0)

#endif /* !ASM_SUPPORTED */

#define arch_read_unlock(rw)					\
do {								\
	arch_rwlock_t *__read_lock_ptr = (rw);			\
	ia64_fetchadd(-1, (int *) __read_lock_ptr, rel);	\
} while (0)

#ifdef ASM_SUPPORTED

static __always_inline void
arch_write_lock_flags(arch_rwlock_t *lock, unsigned long flags)
{
	__asm__ __volatile__ (
		"tbit.nz p6, p0 = %1, %2\n"
		"mov ar.ccv = r0\n"
		"dep r29 = -1, r0, 31, 1\n"
		"br.few 3f;;\n"
		"1:\n"
		"(p6) ssm psr.i\n"
		"2:\n"
		"hint @pause\n"
		"ld4 r2 = [%0];;\n"
		"cmp4.eq p0,p7 = r0, r2\n"
		"(p7) br.cond.spnt.few 2b\n"
		"(p6) rsm psr.i\n"
		";;\n"
		"3:\n"
		"cmpxchg4.acq r2 = [%0], r29, ar.ccv;;\n"
		"cmp4.eq p0,p7 = r0, r2\n"
		"(p7) br.cond.spnt.few 1b;;\n"
		: : "r"(lock), "r"(flags), "i"(IA64_PSR_I_BIT)
		: "ar.ccv", "p6", "p7", "r2", "r29", "memory");
}

#define arch_write_lock(rw) arch_write_lock_flags(rw, 0)

#define arch_write_trylock(rw)							\
({										\
	register long result;							\
										\
	__asm__ __volatile__ (							\
		"mov ar.ccv = r0\n"						\
		"dep r29 = -1, r0, 31, 1;;\n"					\
		"cmpxchg4.acq %0 = [%1], r29, ar.ccv\n"				\
		: "=r"(result) : "r"(rw) : "ar.ccv", "r29", "memory");		\
	(result == 0);								\
})

static inline void arch_write_unlock(arch_rwlock_t *x)
{
	u8 *y = (u8 *)x;
	barrier();
	asm volatile ("st1.rel.nta [%0] = r0\n\t" :: "r"(y+3) : "memory" );
}

#else /* !ASM_SUPPORTED */

#define arch_write_lock_flags(l, flags) arch_write_lock(l)

#define arch_write_lock(l)								\
({											\
	__u64 ia64_val, ia64_set_val = ia64_dep_mi(-1, 0, 31, 1);			\
	__u32 *ia64_write_lock_ptr = (__u32 *) (l);					\
	do {										\
		while (*ia64_write_lock_ptr)						\
			ia64_barrier();							\
		ia64_val = ia64_cmpxchg4_acq(ia64_write_lock_ptr, ia64_set_val, 0);	\
	} while (ia64_val);								\
})

#define arch_write_trylock(rw)						\
({									\
	__u64 ia64_val;							\
	__u64 ia64_set_val = ia64_dep_mi(-1, 0, 31,1);			\
	ia64_val = ia64_cmpxchg4_acq((__u32 *)(rw), ia64_set_val, 0);	\
	(ia64_val == 0);						\
})

static inline void arch_write_unlock(arch_rwlock_t *x)
{
	barrier();
	x->write_lock = 0;
}

#endif /* !ASM_SUPPORTED */

static inline int arch_read_trylock(arch_rwlock_t *x)
{
	union {
		arch_rwlock_t lock;
		__u32 word;
	} old, new;
	old.lock = new.lock = *x;
	old.lock.write_lock = new.lock.write_lock = 0;
	++new.lock.read_counter;
	return (u32)ia64_cmpxchg4_acq((__u32 *)(x), new.word, old.word) == old.word;
}

#define arch_spin_relax(lock)	cpu_relax()
#define arch_read_relax(lock)	cpu_relax()
#define arch_write_relax(lock)	cpu_relax()

#endif /*  _ASM_IA64_SPINLOCK_H */