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
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
|
/**
* Copyright (C) 2018-present MongoDB, Inc.
*
* This program is free software: you can redistribute it and/or modify
* it under the terms of the Server Side Public License, version 1,
* as published by MongoDB, Inc.
*
* This program is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* Server Side Public License for more details.
*
* You should have received a copy of the Server Side Public License
* along with this program. If not, see
* <http://www.mongodb.com/licensing/server-side-public-license>.
*
* As a special exception, the copyright holders give permission to link the
* code of portions of this program with the OpenSSL library under certain
* conditions as described in each individual source file and distribute
* linked combinations including the program with the OpenSSL library. You
* must comply with the Server Side Public License in all respects for
* all of the code used other than as permitted herein. If you modify file(s)
* with this exception, you may extend this exception to your version of the
* file(s), but you are not obligated to do so. If you do not wish to do so,
* delete this exception statement from your version. If you delete this
* exception statement from all source files in the program, then also delete
* it in the license file.
*/
#include "mongo/platform/basic.h"
#include "mongo/util/net/sock.h"
#ifndef _WIN32
#include <netdb.h>
#include <sys/socket.h>
#include <sys/types.h>
#endif
#include "mongo/db/server_options.h"
#include "mongo/stdx/thread.h"
#include "mongo/unittest/unittest.h"
#include "mongo/util/concurrency/notification.h"
#include "mongo/util/fail_point_service.h"
#include "mongo/util/net/socket_exception.h"
namespace {
using namespace mongo;
using std::shared_ptr;
typedef std::shared_ptr<Socket> SocketPtr;
typedef std::pair<SocketPtr, SocketPtr> SocketPair;
// On UNIX, make a connected pair of PF_LOCAL (aka PF_UNIX) sockets via the native 'socketpair'
// call. The 'type' parameter should be one of SOCK_STREAM, SOCK_DGRAM, SOCK_SEQPACKET, etc.
// For Win32, we don't have a native socketpair function, so we hack up a connected PF_INET
// pair on a random port.
SocketPair socketPair(const int type, const int protocol = 0);
#if defined(_WIN32)
namespace detail {
void awaitAccept(SOCKET* acceptSock, SOCKET listenSock, Notification<void>& notify) {
*acceptSock = INVALID_SOCKET;
const SOCKET result = ::accept(listenSock, nullptr, 0);
if (result != INVALID_SOCKET) {
*acceptSock = result;
}
notify.set();
}
void awaitConnect(SOCKET* connectSock, const struct addrinfo& where, Notification<void>& notify) {
*connectSock = INVALID_SOCKET;
SOCKET newSock = ::socket(where.ai_family, where.ai_socktype, where.ai_protocol);
if (newSock != INVALID_SOCKET) {
int result = ::connect(newSock, where.ai_addr, where.ai_addrlen);
if (result == 0) {
*connectSock = newSock;
}
}
notify.set();
}
} // namespace detail
SocketPair socketPair(const int type, const int protocol) {
const int domain = PF_INET;
// Create a listen socket and a connect socket.
const SOCKET listenSock = ::socket(domain, type, protocol);
if (listenSock == INVALID_SOCKET)
return SocketPair();
// Bind the listen socket on port zero, it will pick one for us, and start it listening
// for connections.
struct addrinfo hints, *res;
::memset(&hints, 0, sizeof(hints));
hints.ai_family = PF_INET;
hints.ai_socktype = type;
hints.ai_flags = AI_PASSIVE;
int result = ::getaddrinfo(nullptr, "0", &hints, &res);
if (result != 0) {
closesocket(listenSock);
return SocketPair();
}
result = ::bind(listenSock, res->ai_addr, res->ai_addrlen);
if (result != 0) {
closesocket(listenSock);
::freeaddrinfo(res);
return SocketPair();
}
// Read out the port to which we bound.
sockaddr_in bindAddr;
::socklen_t len = sizeof(bindAddr);
::memset(&bindAddr, 0, sizeof(bindAddr));
result = ::getsockname(listenSock, reinterpret_cast<struct sockaddr*>(&bindAddr), &len);
if (result != 0) {
closesocket(listenSock);
::freeaddrinfo(res);
return SocketPair();
}
result = ::listen(listenSock, 1);
if (result != 0) {
closesocket(listenSock);
::freeaddrinfo(res);
return SocketPair();
}
struct addrinfo connectHints, *connectRes;
::memset(&connectHints, 0, sizeof(connectHints));
connectHints.ai_family = PF_INET;
connectHints.ai_socktype = type;
std::stringstream portStream;
portStream << ntohs(bindAddr.sin_port);
result = ::getaddrinfo(nullptr, portStream.str().c_str(), &connectHints, &connectRes);
if (result != 0) {
closesocket(listenSock);
::freeaddrinfo(res);
return SocketPair();
}
// I'd prefer to avoid trying to do this non-blocking on Windows. Just spin up some
// threads to do the connect and acccept.
Notification<void> accepted;
SOCKET acceptSock = INVALID_SOCKET;
stdx::thread acceptor([&] { detail::awaitAccept(&acceptSock, listenSock, accepted); });
Notification<void> connected;
SOCKET connectSock = INVALID_SOCKET;
stdx::thread connector([&] { detail::awaitConnect(&connectSock, *connectRes, connected); });
connected.get();
connector.join();
if (connectSock == INVALID_SOCKET) {
closesocket(listenSock);
::freeaddrinfo(res);
::freeaddrinfo(connectRes);
closesocket(acceptSock);
closesocket(connectSock);
return SocketPair();
}
accepted.get();
acceptor.join();
if (acceptSock == INVALID_SOCKET) {
closesocket(listenSock);
::freeaddrinfo(res);
::freeaddrinfo(connectRes);
closesocket(acceptSock);
closesocket(connectSock);
return SocketPair();
}
closesocket(listenSock);
::freeaddrinfo(res);
::freeaddrinfo(connectRes);
SocketPtr first(new Socket(static_cast<int>(acceptSock), SockAddr()));
SocketPtr second(new Socket(static_cast<int>(connectSock), SockAddr()));
return SocketPair(first, second);
}
#else
// We can just use ::socketpair and wrap up the result in a Socket.
SocketPair socketPair(const int type, const int protocol) {
// PF_LOCAL is the POSIX name for Unix domain sockets, while PF_UNIX
// is the name that BSD used. We use the BSD name because it is more
// widely supported (e.g. Solaris 10).
const int domain = PF_UNIX;
int socks[2];
const int result = ::socketpair(domain, type, protocol, socks);
if (result == 0) {
return SocketPair(SocketPtr(new Socket(socks[0], SockAddr())),
SocketPtr(new Socket(socks[1], SockAddr())));
}
return SocketPair();
}
#endif
// This should match the name of the fail point declared in sock.cpp.
const char kSocketFailPointName[] = "throwSockExcep";
class SocketFailPointTest : public unittest::Test {
public:
SocketFailPointTest()
: _failPoint(getGlobalFailPointRegistry()->getFailPoint(kSocketFailPointName)),
_sockets(socketPair(SOCK_STREAM)) {
ASSERT_TRUE(_failPoint != nullptr);
ASSERT_TRUE(_sockets.first);
ASSERT_TRUE(_sockets.second);
}
~SocketFailPointTest() {}
bool trySend() {
char byte = 'x';
_sockets.first->send(&byte, sizeof(byte), "SocketFailPointTest::trySend");
return true;
}
bool trySendVector() {
std::vector<std::pair<char*, int>> data;
char byte = 'x';
data.push_back(std::make_pair(&byte, sizeof(byte)));
_sockets.first->send(data, "SocketFailPointTest::trySendVector");
return true;
}
bool tryRecv() {
char byte;
_sockets.second->recv(&byte, sizeof(byte));
return true;
}
// You must queue at least one byte on the send socket before calling this function.
size_t countRecvable(size_t max) {
std::vector<char> buf(max);
// This isn't great, because we don't have a guarantee that multiple sends will be
// captured in one recv. However, sock doesn't let us pass flags into recv, so we
// can't make this non blocking, and therefore can't risk another call.
return _sockets.second->unsafe_recv(&buf[0], max);
}
FailPoint* const _failPoint;
const SocketPair _sockets;
};
class ScopedFailPointEnabler {
public:
ScopedFailPointEnabler(FailPoint& fp) : _fp(fp) {
_fp.setMode(FailPoint::alwaysOn);
}
~ScopedFailPointEnabler() {
_fp.setMode(FailPoint::off);
}
private:
FailPoint& _fp;
};
TEST_F(SocketFailPointTest, TestSend) {
ASSERT_TRUE(trySend());
ASSERT_TRUE(tryRecv());
{
const ScopedFailPointEnabler enabled(*_failPoint);
auto expectedEx =
makeSocketError(SocketErrorKind::SEND_ERROR, _sockets.first->remoteString());
ASSERT_THROWS_WHAT(trySend(), NetworkException, expectedEx.reason());
}
// Channel should be working again
ASSERT_TRUE(trySend());
ASSERT_TRUE(tryRecv());
}
TEST_F(SocketFailPointTest, TestSendVector) {
ASSERT_TRUE(trySendVector());
ASSERT_TRUE(tryRecv());
{
const ScopedFailPointEnabler enabled(*_failPoint);
ASSERT_THROWS(trySendVector(), NetworkException);
}
ASSERT_TRUE(trySendVector());
ASSERT_TRUE(tryRecv());
}
TEST_F(SocketFailPointTest, TestRecv) {
ASSERT_TRUE(trySend()); // data for recv
ASSERT_TRUE(tryRecv());
{
ASSERT_TRUE(trySend()); // data for recv
const ScopedFailPointEnabler enabled(*_failPoint);
auto expectedEx =
makeSocketError(SocketErrorKind::RECV_ERROR, _sockets.first->remoteString());
ASSERT_THROWS_WHAT(tryRecv(), NetworkException, expectedEx.reason());
}
ASSERT_TRUE(trySend()); // data for recv
ASSERT_TRUE(tryRecv());
}
TEST_F(SocketFailPointTest, TestFailedSendsDontSend) {
ASSERT_TRUE(trySend());
ASSERT_TRUE(tryRecv());
{
ASSERT_TRUE(trySend()); // queue 1 byte
const ScopedFailPointEnabler enabled(*_failPoint);
// Fail to queue another byte
ASSERT_THROWS(trySend(), NetworkException);
}
// Failed byte should not have been transmitted.
ASSERT_EQUALS(size_t(1), countRecvable(2));
}
// Ensure that calling send doesn't actually enqueue data to the socket
TEST_F(SocketFailPointTest, TestFailedVectorSendsDontSend) {
ASSERT_TRUE(trySend());
ASSERT_TRUE(tryRecv());
{
ASSERT_TRUE(trySend()); // queue 1 byte
const ScopedFailPointEnabler enabled(*_failPoint);
// Fail to queue another byte
ASSERT_THROWS(trySendVector(), NetworkException);
}
// Failed byte should not have been transmitted.
ASSERT_EQUALS(size_t(1), countRecvable(2));
}
TEST_F(SocketFailPointTest, TestFailedRecvsDontRecv) {
ASSERT_TRUE(trySend());
ASSERT_TRUE(tryRecv());
{
ASSERT_TRUE(trySend());
const ScopedFailPointEnabler enabled(*_failPoint);
// Fail to recv that byte
ASSERT_THROWS(tryRecv(), NetworkException);
}
// Failed byte should still be queued to recv.
ASSERT_EQUALS(size_t(1), countRecvable(1));
// Channel should be working again
ASSERT_TRUE(trySend());
ASSERT_TRUE(tryRecv());
}
} // namespace
|