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
|
// SPDX-License-Identifier: GPL-2.0
#include <linux/kernel.h>
#include <linux/errno.h>
#include <linux/mm.h>
#include <linux/slab.h>
#include <linux/eventfd.h>
#include <linux/eventpoll.h>
#include <linux/io_uring.h>
#include <linux/io_uring_types.h>
#include "io-wq.h"
#include "eventfd.h"
struct io_ev_fd {
struct eventfd_ctx *cq_ev_fd;
unsigned int eventfd_async;
/* protected by ->completion_lock */
unsigned last_cq_tail;
refcount_t refs;
atomic_t ops;
struct rcu_head rcu;
};
enum {
IO_EVENTFD_OP_SIGNAL_BIT,
};
static void io_eventfd_free(struct rcu_head *rcu)
{
struct io_ev_fd *ev_fd = container_of(rcu, struct io_ev_fd, rcu);
eventfd_ctx_put(ev_fd->cq_ev_fd);
kfree(ev_fd);
}
static void io_eventfd_put(struct io_ev_fd *ev_fd)
{
if (refcount_dec_and_test(&ev_fd->refs))
call_rcu(&ev_fd->rcu, io_eventfd_free);
}
static void io_eventfd_do_signal(struct rcu_head *rcu)
{
struct io_ev_fd *ev_fd = container_of(rcu, struct io_ev_fd, rcu);
eventfd_signal_mask(ev_fd->cq_ev_fd, EPOLL_URING_WAKE);
io_eventfd_put(ev_fd);
}
static void io_eventfd_release(struct io_ev_fd *ev_fd, bool put_ref)
{
if (put_ref)
io_eventfd_put(ev_fd);
rcu_read_unlock();
}
/*
* Returns true if the caller should put the ev_fd reference, false if not.
*/
static bool __io_eventfd_signal(struct io_ev_fd *ev_fd)
{
if (eventfd_signal_allowed()) {
eventfd_signal_mask(ev_fd->cq_ev_fd, EPOLL_URING_WAKE);
return true;
}
if (!atomic_fetch_or(BIT(IO_EVENTFD_OP_SIGNAL_BIT), &ev_fd->ops)) {
call_rcu_hurry(&ev_fd->rcu, io_eventfd_do_signal);
return false;
}
return true;
}
/*
* Trigger if eventfd_async isn't set, or if it's set and the caller is
* an async worker. If ev_fd isn't valid, obviously return false.
*/
static bool io_eventfd_trigger(struct io_ev_fd *ev_fd)
{
if (ev_fd)
return !ev_fd->eventfd_async || io_wq_current_is_worker();
return false;
}
/*
* On success, returns with an ev_fd reference grabbed and the RCU read
* lock held.
*/
static struct io_ev_fd *io_eventfd_grab(struct io_ring_ctx *ctx)
{
struct io_ev_fd *ev_fd;
if (READ_ONCE(ctx->rings->cq_flags) & IORING_CQ_EVENTFD_DISABLED)
return NULL;
rcu_read_lock();
/*
* rcu_dereference ctx->io_ev_fd once and use it for both for checking
* and eventfd_signal
*/
ev_fd = rcu_dereference(ctx->io_ev_fd);
/*
* Check again if ev_fd exists in case an io_eventfd_unregister call
* completed between the NULL check of ctx->io_ev_fd at the start of
* the function and rcu_read_lock.
*/
if (io_eventfd_trigger(ev_fd) && refcount_inc_not_zero(&ev_fd->refs))
return ev_fd;
rcu_read_unlock();
return NULL;
}
void io_eventfd_signal(struct io_ring_ctx *ctx)
{
struct io_ev_fd *ev_fd;
ev_fd = io_eventfd_grab(ctx);
if (ev_fd)
io_eventfd_release(ev_fd, __io_eventfd_signal(ev_fd));
}
void io_eventfd_flush_signal(struct io_ring_ctx *ctx)
{
struct io_ev_fd *ev_fd;
ev_fd = io_eventfd_grab(ctx);
if (ev_fd) {
bool skip, put_ref = true;
/*
* Eventfd should only get triggered when at least one event
* has been posted. Some applications rely on the eventfd
* notification count only changing IFF a new CQE has been
* added to the CQ ring. There's no dependency on 1:1
* relationship between how many times this function is called
* (and hence the eventfd count) and number of CQEs posted to
* the CQ ring.
*/
spin_lock(&ctx->completion_lock);
skip = ctx->cached_cq_tail == ev_fd->last_cq_tail;
ev_fd->last_cq_tail = ctx->cached_cq_tail;
spin_unlock(&ctx->completion_lock);
if (!skip)
put_ref = __io_eventfd_signal(ev_fd);
io_eventfd_release(ev_fd, put_ref);
}
}
int io_eventfd_register(struct io_ring_ctx *ctx, void __user *arg,
unsigned int eventfd_async)
{
struct io_ev_fd *ev_fd;
__s32 __user *fds = arg;
int fd;
ev_fd = rcu_dereference_protected(ctx->io_ev_fd,
lockdep_is_held(&ctx->uring_lock));
if (ev_fd)
return -EBUSY;
if (copy_from_user(&fd, fds, sizeof(*fds)))
return -EFAULT;
ev_fd = kmalloc(sizeof(*ev_fd), GFP_KERNEL);
if (!ev_fd)
return -ENOMEM;
ev_fd->cq_ev_fd = eventfd_ctx_fdget(fd);
if (IS_ERR(ev_fd->cq_ev_fd)) {
int ret = PTR_ERR(ev_fd->cq_ev_fd);
kfree(ev_fd);
return ret;
}
spin_lock(&ctx->completion_lock);
ev_fd->last_cq_tail = ctx->cached_cq_tail;
spin_unlock(&ctx->completion_lock);
ev_fd->eventfd_async = eventfd_async;
ctx->has_evfd = true;
refcount_set(&ev_fd->refs, 1);
atomic_set(&ev_fd->ops, 0);
rcu_assign_pointer(ctx->io_ev_fd, ev_fd);
return 0;
}
int io_eventfd_unregister(struct io_ring_ctx *ctx)
{
struct io_ev_fd *ev_fd;
ev_fd = rcu_dereference_protected(ctx->io_ev_fd,
lockdep_is_held(&ctx->uring_lock));
if (ev_fd) {
ctx->has_evfd = false;
rcu_assign_pointer(ctx->io_ev_fd, NULL);
io_eventfd_put(ev_fd);
return 0;
}
return -ENXIO;
}
|
