forked from torvalds/linux
-
Notifications
You must be signed in to change notification settings - Fork 1
/
closure.c
297 lines (238 loc) · 6.75 KB
/
closure.c
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
292
293
294
295
296
297
// SPDX-License-Identifier: GPL-2.0
/*
* Asynchronous refcounty things
*
* Copyright 2010, 2011 Kent Overstreet <[email protected]>
* Copyright 2012 Google, Inc.
*/
#include <linux/closure.h>
#include <linux/debugfs.h>
#include <linux/export.h>
#include <linux/rcupdate.h>
#include <linux/seq_file.h>
#include <linux/sched/debug.h>
static inline void closure_put_after_sub_checks(int flags)
{
int r = flags & CLOSURE_REMAINING_MASK;
if (WARN(flags & CLOSURE_GUARD_MASK,
"closure has guard bits set: %x (%u)",
flags & CLOSURE_GUARD_MASK, (unsigned) __fls(r)))
r &= ~CLOSURE_GUARD_MASK;
WARN(!r && (flags & ~CLOSURE_DESTRUCTOR),
"closure ref hit 0 with incorrect flags set: %x (%u)",
flags & ~CLOSURE_DESTRUCTOR, (unsigned) __fls(flags));
}
static inline void closure_put_after_sub(struct closure *cl, int flags)
{
closure_put_after_sub_checks(flags);
if (!(flags & CLOSURE_REMAINING_MASK)) {
smp_acquire__after_ctrl_dep();
cl->closure_get_happened = false;
if (cl->fn && !(flags & CLOSURE_DESTRUCTOR)) {
atomic_set(&cl->remaining,
CLOSURE_REMAINING_INITIALIZER);
closure_queue(cl);
} else {
struct closure *parent = cl->parent;
closure_fn *destructor = cl->fn;
closure_debug_destroy(cl);
if (destructor)
destructor(&cl->work);
if (parent)
closure_put(parent);
}
}
}
/* For clearing flags with the same atomic op as a put */
void closure_sub(struct closure *cl, int v)
{
closure_put_after_sub(cl, atomic_sub_return_release(v, &cl->remaining));
}
EXPORT_SYMBOL(closure_sub);
/*
* closure_put - decrement a closure's refcount
*/
void closure_put(struct closure *cl)
{
closure_put_after_sub(cl, atomic_dec_return_release(&cl->remaining));
}
EXPORT_SYMBOL(closure_put);
/*
* closure_wake_up - wake up all closures on a wait list, without memory barrier
*/
void __closure_wake_up(struct closure_waitlist *wait_list)
{
struct llist_node *list;
struct closure *cl, *t;
struct llist_node *reverse = NULL;
list = llist_del_all(&wait_list->list);
/* We first reverse the list to preserve FIFO ordering and fairness */
reverse = llist_reverse_order(list);
/* Then do the wakeups */
llist_for_each_entry_safe(cl, t, reverse, list) {
closure_set_waiting(cl, 0);
closure_sub(cl, CLOSURE_WAITING + 1);
}
}
EXPORT_SYMBOL(__closure_wake_up);
/**
* closure_wait - add a closure to a waitlist
* @waitlist: will own a ref on @cl, which will be released when
* closure_wake_up() is called on @waitlist.
* @cl: closure pointer.
*
*/
bool closure_wait(struct closure_waitlist *waitlist, struct closure *cl)
{
if (atomic_read(&cl->remaining) & CLOSURE_WAITING)
return false;
cl->closure_get_happened = true;
closure_set_waiting(cl, _RET_IP_);
atomic_add(CLOSURE_WAITING + 1, &cl->remaining);
llist_add(&cl->list, &waitlist->list);
return true;
}
EXPORT_SYMBOL(closure_wait);
struct closure_syncer {
struct task_struct *task;
int done;
};
static CLOSURE_CALLBACK(closure_sync_fn)
{
struct closure *cl = container_of(ws, struct closure, work);
struct closure_syncer *s = cl->s;
struct task_struct *p;
rcu_read_lock();
p = READ_ONCE(s->task);
s->done = 1;
wake_up_process(p);
rcu_read_unlock();
}
void __sched __closure_sync(struct closure *cl)
{
struct closure_syncer s = { .task = current };
cl->s = &s;
continue_at(cl, closure_sync_fn, NULL);
while (1) {
set_current_state(TASK_UNINTERRUPTIBLE);
if (s.done)
break;
schedule();
}
__set_current_state(TASK_RUNNING);
}
EXPORT_SYMBOL(__closure_sync);
/*
* closure_return_sync - finish running a closure, synchronously (i.e. waiting
* for outstanding get()s to finish) and returning once closure refcount is 0.
*
* Unlike closure_sync() this doesn't reinit the ref to 1; subsequent
* closure_get_not_zero() calls waill fail.
*/
void __sched closure_return_sync(struct closure *cl)
{
struct closure_syncer s = { .task = current };
cl->s = &s;
set_closure_fn(cl, closure_sync_fn, NULL);
unsigned flags = atomic_sub_return_release(1 + CLOSURE_RUNNING - CLOSURE_DESTRUCTOR,
&cl->remaining);
closure_put_after_sub_checks(flags);
if (unlikely(flags & CLOSURE_REMAINING_MASK)) {
while (1) {
set_current_state(TASK_UNINTERRUPTIBLE);
if (s.done)
break;
schedule();
}
__set_current_state(TASK_RUNNING);
}
if (cl->parent)
closure_put(cl->parent);
}
EXPORT_SYMBOL(closure_return_sync);
int __sched __closure_sync_timeout(struct closure *cl, unsigned long timeout)
{
struct closure_syncer s = { .task = current };
int ret = 0;
cl->s = &s;
continue_at(cl, closure_sync_fn, NULL);
while (1) {
set_current_state(TASK_UNINTERRUPTIBLE);
if (s.done)
break;
if (!timeout) {
/*
* Carefully undo the continue_at() - but only if it
* hasn't completed, i.e. the final closure_put() hasn't
* happened yet:
*/
unsigned old, new, v = atomic_read(&cl->remaining);
do {
old = v;
if (!old || (old & CLOSURE_RUNNING))
goto success;
new = old + CLOSURE_REMAINING_INITIALIZER;
} while ((v = atomic_cmpxchg(&cl->remaining, old, new)) != old);
ret = -ETIME;
}
timeout = schedule_timeout(timeout);
}
success:
__set_current_state(TASK_RUNNING);
return ret;
}
EXPORT_SYMBOL(__closure_sync_timeout);
#ifdef CONFIG_DEBUG_CLOSURES
static LIST_HEAD(closure_list);
static DEFINE_SPINLOCK(closure_list_lock);
void closure_debug_create(struct closure *cl)
{
unsigned long flags;
BUG_ON(cl->magic == CLOSURE_MAGIC_ALIVE);
cl->magic = CLOSURE_MAGIC_ALIVE;
spin_lock_irqsave(&closure_list_lock, flags);
list_add(&cl->all, &closure_list);
spin_unlock_irqrestore(&closure_list_lock, flags);
}
EXPORT_SYMBOL(closure_debug_create);
void closure_debug_destroy(struct closure *cl)
{
unsigned long flags;
if (cl->magic == CLOSURE_MAGIC_STACK)
return;
BUG_ON(cl->magic != CLOSURE_MAGIC_ALIVE);
cl->magic = CLOSURE_MAGIC_DEAD;
spin_lock_irqsave(&closure_list_lock, flags);
list_del(&cl->all);
spin_unlock_irqrestore(&closure_list_lock, flags);
}
EXPORT_SYMBOL(closure_debug_destroy);
static int debug_show(struct seq_file *f, void *data)
{
struct closure *cl;
spin_lock_irq(&closure_list_lock);
list_for_each_entry(cl, &closure_list, all) {
int r = atomic_read(&cl->remaining);
seq_printf(f, "%p: %pS -> %pS p %p r %i ",
cl, (void *) cl->ip, cl->fn, cl->parent,
r & CLOSURE_REMAINING_MASK);
seq_printf(f, "%s%s\n",
test_bit(WORK_STRUCT_PENDING_BIT,
work_data_bits(&cl->work)) ? "Q" : "",
r & CLOSURE_RUNNING ? "R" : "");
if (r & CLOSURE_WAITING)
seq_printf(f, " W %pS\n",
(void *) cl->waiting_on);
seq_putc(f, '\n');
}
spin_unlock_irq(&closure_list_lock);
return 0;
}
DEFINE_SHOW_ATTRIBUTE(debug);
static int __init closure_debug_init(void)
{
debugfs_create_file("closures", 0400, NULL, NULL, &debug_fops);
return 0;
}
late_initcall(closure_debug_init)
#endif