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
|
/*
* libev epoll fd activity backend
*
* Copyright (c) 2007 Marc Alexander Lehmann <libev@schmorp.de>
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions are
* met:
*
* * Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
*
* * Redistributions in binary form must reproduce the above
* copyright notice, this list of conditions and the following
* disclaimer in the documentation and/or other materials provided
* with the distribution.
*
* THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
* "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
* LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
* A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
* OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
* SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
* LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
* DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
* THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
* (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
* OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
*/
/*
* general notes about epoll:
*
* a) epoll silently removes fds from the fd set. as nothing tells us
* that an fd has been removed otherwise, we have to continually
* "rearm" fds that we suspect *might* have changed (same
* problem with kqueue, but much less costly there).
* b) the fact that ADD != MOD creates a lot of extra syscalls due to a)
* and seems not to have any advantage.
* c) the inability to handle fork or file descriptors (think dup)
* limits the applicability over poll, so this is not a generic
* poll replacement.
*
* lots of "weird code" and complication handling in this file is due
* to these design problems with epoll, as we try very hard to avoid
* epoll_ctl syscalls for common usage patterns.
*/
#include <sys/epoll.h>
static void
epoll_modify (EV_P_ int fd, int oev, int nev)
{
struct epoll_event ev;
/*
* we handle EPOLL_CTL_DEL by ignoring it here
* on the assumption that the fd is gone anyways
* if that is wrong, we have to handle the spurious
* event in epoll_poll.
*/
if (!nev)
return;
ev.data.u64 = fd; /* use u64 to fully initialise the struct, for nicer strace etc. */
ev.events = (nev & EV_READ ? EPOLLIN : 0)
| (nev & EV_WRITE ? EPOLLOUT : 0);
if (expect_true (!epoll_ctl (backend_fd, oev ? EPOLL_CTL_MOD : EPOLL_CTL_ADD, fd, &ev)))
return;
if (expect_true (errno == ENOENT))
{
/* on ENOENT the fd went away, so try to do the right thing */
if (!nev)
return;
if (!epoll_ctl (backend_fd, EPOLL_CTL_ADD, fd, &ev))
return;
}
else if (expect_true (errno == EEXIST))
{
/* on EEXIST we ignored a previous DEL */
if (!epoll_ctl (backend_fd, EPOLL_CTL_MOD, fd, &ev))
return;
}
fd_kill (EV_A_ fd);
}
static void
epoll_poll (EV_P_ ev_tstamp timeout)
{
int i;
int eventcnt = epoll_wait (backend_fd, epoll_events, epoll_eventmax, (int)ceil (timeout * 1000.));
if (expect_false (eventcnt < 0))
{
if (errno != EINTR)
syserr ("(libev) epoll_wait");
return;
}
for (i = 0; i < eventcnt; ++i)
{
struct epoll_event *ev = epoll_events + i;
int fd = ev->data.u64;
int got = (ev->events & (EPOLLOUT | EPOLLERR | EPOLLHUP) ? EV_WRITE : 0)
| (ev->events & (EPOLLIN | EPOLLERR | EPOLLHUP) ? EV_READ : 0);
int want = anfds [fd].events;
if (expect_false (got & ~want))
{
/* we received an event but are not interested in it, try mod or del */
ev->events = (want & EV_READ ? EPOLLIN : 0)
| (want & EV_WRITE ? EPOLLOUT : 0);
epoll_ctl (backend_fd, want ? EPOLL_CTL_MOD : EPOLL_CTL_DEL, fd, ev);
}
fd_event (EV_A_ fd, got);
}
/* if the receive array was full, increase its size */
if (expect_false (eventcnt == epoll_eventmax))
{
ev_free (epoll_events);
epoll_eventmax = array_nextsize (sizeof (struct epoll_event), epoll_eventmax, epoll_eventmax + 1);
epoll_events = (struct epoll_event *)ev_malloc (sizeof (struct epoll_event) * epoll_eventmax);
}
}
int inline_size
epoll_init (EV_P_ int flags)
{
backend_fd = epoll_create (256);
if (backend_fd < 0)
return 0;
fcntl (backend_fd, F_SETFD, FD_CLOEXEC);
backend_fudge = 0.; /* kernel sources seem to indicate this to be zero */
backend_modify = epoll_modify;
backend_poll = epoll_poll;
epoll_eventmax = 64; /* intiial number of events receivable per poll */
epoll_events = (struct epoll_event *)ev_malloc (sizeof (struct epoll_event) * epoll_eventmax);
return EVBACKEND_EPOLL;
}
void inline_size
epoll_destroy (EV_P)
{
ev_free (epoll_events);
}
void inline_size
epoll_fork (EV_P)
{
close (backend_fd);
while ((backend_fd = epoll_create (256)) < 0)
syserr ("(libev) epoll_create");
fcntl (backend_fd, F_SETFD, FD_CLOEXEC);
fd_rearm_all (EV_A);
}
|
