summaryrefslogtreecommitdiffhomepage
path: root/test/syscalls/linux/socket_ip_tcp_generic.cc
blob: 5b198f49d2d1c3bbd51567e89566de3d5c612484 (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
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
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
// Copyright 2018 The gVisor Authors.
//
// Licensed under the Apache License, Version 2.0 (the "License");
// you may not use this file except in compliance with the License.
// You may obtain a copy of the License at
//
//     http://www.apache.org/licenses/LICENSE-2.0
//
// Unless required by applicable law or agreed to in writing, software
// distributed under the License is distributed on an "AS IS" BASIS,
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
// See the License for the specific language governing permissions and
// limitations under the License.

#include "test/syscalls/linux/socket_ip_tcp_generic.h"

#include <netinet/in.h>
#include <netinet/tcp.h>
#include <poll.h>
#include <stdio.h>
#include <sys/ioctl.h>
#include <sys/socket.h>
#include <sys/types.h>
#include <sys/un.h>

#include "gtest/gtest.h"
#include "gtest/gtest.h"
#include "test/syscalls/linux/socket_test_util.h"
#include "test/util/test_util.h"

namespace gvisor {
namespace testing {

TEST_P(TCPSocketPairTest, TcpInfoSucceedes) {
  auto sockets = ASSERT_NO_ERRNO_AND_VALUE(NewSocketPair());

  struct tcp_info opt = {};
  socklen_t optLen = sizeof(opt);
  EXPECT_THAT(getsockopt(sockets->first_fd(), SOL_TCP, TCP_INFO, &opt, &optLen),
              SyscallSucceeds());
}

TEST_P(TCPSocketPairTest, ShortTcpInfoSucceedes) {
  auto sockets = ASSERT_NO_ERRNO_AND_VALUE(NewSocketPair());

  struct tcp_info opt = {};
  socklen_t optLen = 1;
  EXPECT_THAT(getsockopt(sockets->first_fd(), SOL_TCP, TCP_INFO, &opt, &optLen),
              SyscallSucceeds());
}

TEST_P(TCPSocketPairTest, ZeroTcpInfoSucceedes) {
  auto sockets = ASSERT_NO_ERRNO_AND_VALUE(NewSocketPair());

  struct tcp_info opt = {};
  socklen_t optLen = 0;
  EXPECT_THAT(getsockopt(sockets->first_fd(), SOL_TCP, TCP_INFO, &opt, &optLen),
              SyscallSucceeds());
}

// This test validates that an RST is sent instead of a FIN when data is
// unread on calls to close(2).
TEST_P(TCPSocketPairTest, RSTSentOnCloseWithUnreadData) {
  auto sockets = ASSERT_NO_ERRNO_AND_VALUE(NewSocketPair());

  char buf[10] = {};
  ASSERT_THAT(RetryEINTR(write)(sockets->first_fd(), buf, sizeof(buf)),
              SyscallSucceedsWithValue(sizeof(buf)));

  // Wait until t_ sees the data on its side but don't read it.
  struct pollfd poll_fd = {sockets->second_fd(), POLLIN | POLLHUP, 0};
  constexpr int kPollTimeoutMs = 20000;  // Wait up to 20 seconds for the data.
  ASSERT_THAT(RetryEINTR(poll)(&poll_fd, 1, kPollTimeoutMs),
              SyscallSucceedsWithValue(1));

  // Now close the connected without reading the data.
  ASSERT_THAT(close(sockets->release_second_fd()), SyscallSucceeds());

  // Wait for the other end to receive the RST (up to 20 seconds).
  struct pollfd poll_fd2 = {sockets->first_fd(), POLLIN | POLLHUP, 0};
  ASSERT_THAT(RetryEINTR(poll)(&poll_fd2, 1, kPollTimeoutMs),
              SyscallSucceedsWithValue(1));

  // A shutdown with unread data will cause a RST to be sent instead
  // of a FIN, per RFC 2525 section 2.17; this is also what Linux does.
  ASSERT_THAT(RetryEINTR(read)(sockets->first_fd(), buf, sizeof(buf)),
              SyscallFailsWithErrno(ECONNRESET));
}

// This test will validate that a RST will cause POLLHUP to trigger.
TEST_P(TCPSocketPairTest, RSTCausesPollHUP) {
  auto sockets = ASSERT_NO_ERRNO_AND_VALUE(NewSocketPair());

  char buf[10] = {};
  ASSERT_THAT(RetryEINTR(write)(sockets->first_fd(), buf, sizeof(buf)),
              SyscallSucceedsWithValue(sizeof(buf)));

  // Wait until second sees the data on its side but don't read it.
  struct pollfd poll_fd = {sockets->second_fd(), POLLIN, 0};
  constexpr int kPollTimeoutMs = 20000;  // Wait up to 20 seconds for the data.
  ASSERT_THAT(RetryEINTR(poll)(&poll_fd, 1, kPollTimeoutMs),
              SyscallSucceedsWithValue(1));
  EXPECT_EQ(poll_fd.revents & POLLIN, POLLIN);

  // Confirm we at least have one unread byte.
  int bytes_available = 0;
  ASSERT_THAT(
      RetryEINTR(ioctl)(sockets->second_fd(), FIONREAD, &bytes_available),
      SyscallSucceeds());
  EXPECT_GT(bytes_available, 0);

  // Now close the connected socket without reading the data from the second,
  // this will cause a RST and we should see that with POLLHUP.
  ASSERT_THAT(close(sockets->release_second_fd()), SyscallSucceeds());

  // Wait for the other end to receive the RST (up to 20 seconds).
  struct pollfd poll_fd3 = {sockets->first_fd(), POLLHUP, 0};
  ASSERT_THAT(RetryEINTR(poll)(&poll_fd3, 1, kPollTimeoutMs),
              SyscallSucceedsWithValue(1));
  ASSERT_NE(poll_fd.revents & (POLLHUP | POLLIN), 0);
}

// This test validates that even if a RST is sent the other end will not
// get an ECONNRESET until it's read all data.
TEST_P(TCPSocketPairTest, RSTSentOnCloseWithUnreadDataAllowsReadBuffered) {
  auto sockets = ASSERT_NO_ERRNO_AND_VALUE(NewSocketPair());

  char buf[10] = {};
  ASSERT_THAT(RetryEINTR(write)(sockets->first_fd(), buf, sizeof(buf)),
              SyscallSucceedsWithValue(sizeof(buf)));
  ASSERT_THAT(RetryEINTR(write)(sockets->second_fd(), buf, sizeof(buf)),
              SyscallSucceedsWithValue(sizeof(buf)));

  // Wait until second sees the data on its side but don't read it.
  struct pollfd poll_fd = {sockets->second_fd(), POLLIN, 0};
  constexpr int kPollTimeoutMs = 30000;  // Wait up to 30 seconds for the data.
  ASSERT_THAT(RetryEINTR(poll)(&poll_fd, 1, kPollTimeoutMs),
              SyscallSucceedsWithValue(1));

  // Wait until first sees the data on its side but don't read it.
  struct pollfd poll_fd2 = {sockets->first_fd(), POLLIN, 0};
  ASSERT_THAT(RetryEINTR(poll)(&poll_fd2, 1, kPollTimeoutMs),
              SyscallSucceedsWithValue(1));

  // Now close the connected socket without reading the data from the second.
  ASSERT_THAT(close(sockets->release_second_fd()), SyscallSucceeds());

  // Wait for the other end to receive the RST (up to 30 seconds).
  struct pollfd poll_fd3 = {sockets->first_fd(), POLLHUP, 0};
  ASSERT_THAT(RetryEINTR(poll)(&poll_fd3, 1, kPollTimeoutMs),
              SyscallSucceedsWithValue(1));

  // Since we also have data buffered we should be able to read it before
  // the syscall will fail with ECONNRESET.
  ASSERT_THAT(RetryEINTR(read)(sockets->first_fd(), buf, sizeof(buf)),
              SyscallSucceedsWithValue(sizeof(buf)));

  // A shutdown with unread data will cause a RST to be sent instead
  // of a FIN, per RFC 2525 section 2.17; this is also what Linux does.
  ASSERT_THAT(RetryEINTR(read)(sockets->first_fd(), buf, sizeof(buf)),
              SyscallFailsWithErrno(ECONNRESET));
}

// This test will verify that a clean shutdown (FIN) is preformed when there
// is unread data but only the write side is closed.
TEST_P(TCPSocketPairTest, FINSentOnShutdownWrWithUnreadData) {
  auto sockets = ASSERT_NO_ERRNO_AND_VALUE(NewSocketPair());

  char buf[10] = {};
  ASSERT_THAT(RetryEINTR(write)(sockets->first_fd(), buf, sizeof(buf)),
              SyscallSucceedsWithValue(sizeof(buf)));

  // Wait until t_ sees the data on its side but don't read it.
  struct pollfd poll_fd = {sockets->second_fd(), POLLIN | POLLHUP, 0};
  constexpr int kPollTimeoutMs = 20000;  // Wait up to 20 seconds for the data.
  ASSERT_THAT(RetryEINTR(poll)(&poll_fd, 1, kPollTimeoutMs),
              SyscallSucceedsWithValue(1));

  // Now shutdown the write end leaving the read end open.
  ASSERT_THAT(shutdown(sockets->second_fd(), SHUT_WR), SyscallSucceeds());

  // Wait for the other end to receive the FIN (up to 20 seconds).
  struct pollfd poll_fd2 = {sockets->first_fd(), POLLIN | POLLHUP, 0};
  ASSERT_THAT(RetryEINTR(poll)(&poll_fd2, 1, kPollTimeoutMs),
              SyscallSucceedsWithValue(1));

  // Since we didn't shutdown the read end this will be a clean close.
  ASSERT_THAT(RetryEINTR(read)(sockets->first_fd(), buf, sizeof(buf)),
              SyscallSucceedsWithValue(0));
}

// This test will verify that when data is received by a socket, even if it's
// not read SHUT_RD will not cause any packets to be generated.
TEST_P(TCPSocketPairTest, ShutdownRdShouldCauseNoPacketsWithUnreadData) {
  auto sockets = ASSERT_NO_ERRNO_AND_VALUE(NewSocketPair());

  char buf[10] = {};
  ASSERT_THAT(RetryEINTR(write)(sockets->first_fd(), buf, sizeof(buf)),
              SyscallSucceedsWithValue(sizeof(buf)));

  // Wait until t_ sees the data on its side but don't read it.
  struct pollfd poll_fd = {sockets->second_fd(), POLLIN | POLLHUP, 0};
  constexpr int kPollTimeoutMs = 20000;  // Wait up to 20 seconds for the data.
  ASSERT_THAT(RetryEINTR(poll)(&poll_fd, 1, kPollTimeoutMs),
              SyscallSucceedsWithValue(1));

  // Now shutdown the read end, this will generate no packets to the other end.
  ASSERT_THAT(shutdown(sockets->second_fd(), SHUT_RD), SyscallSucceeds());

  // We should not receive any events on the other side of the socket.
  struct pollfd poll_fd2 = {sockets->first_fd(), POLLIN | POLLHUP, 0};
  constexpr int kPollNoResponseTimeoutMs = 3000;
  ASSERT_THAT(RetryEINTR(poll)(&poll_fd2, 1, kPollNoResponseTimeoutMs),
              SyscallSucceedsWithValue(0));  // Timeout.
}

// This test will verify that a socket which has unread data will still allow
// the data to be read after shutting down the read side, and once there is no
// unread data left, then read will return an EOF.
TEST_P(TCPSocketPairTest, ShutdownRdAllowsReadOfReceivedDataBeforeEOF) {
  auto sockets = ASSERT_NO_ERRNO_AND_VALUE(NewSocketPair());

  char buf[10] = {};
  ASSERT_THAT(RetryEINTR(write)(sockets->first_fd(), buf, sizeof(buf)),
              SyscallSucceedsWithValue(sizeof(buf)));

  // Wait until t_ sees the data on its side but don't read it.
  struct pollfd poll_fd = {sockets->second_fd(), POLLIN | POLLHUP, 0};
  constexpr int kPollTimeoutMs = 20000;  // Wait up to 20 seconds for the data.
  ASSERT_THAT(RetryEINTR(poll)(&poll_fd, 1, kPollTimeoutMs),
              SyscallSucceedsWithValue(1));

  // Now shutdown the read end.
  ASSERT_THAT(shutdown(sockets->second_fd(), SHUT_RD), SyscallSucceeds());

  // Even though we did a SHUT_RD on the read end we can still read the data.
  ASSERT_THAT(RetryEINTR(read)(sockets->second_fd(), buf, sizeof(buf)),
              SyscallSucceedsWithValue(sizeof(buf)));

  // After reading all of the data, reading the closed read end returns EOF.
  ASSERT_THAT(RetryEINTR(poll)(&poll_fd, 1, kPollTimeoutMs),
              SyscallSucceedsWithValue(1));
  ASSERT_THAT(RetryEINTR(read)(sockets->second_fd(), buf, sizeof(buf)),
              SyscallSucceedsWithValue(0));
}

TEST_P(TCPSocketPairTest, ClosedReadNonBlockingSocket) {
  auto sockets = ASSERT_NO_ERRNO_AND_VALUE(NewSocketPair());

  // Set the read end to O_NONBLOCK.
  int opts = 0;
  ASSERT_THAT(opts = fcntl(sockets->second_fd(), F_GETFL), SyscallSucceeds());
  ASSERT_THAT(fcntl(sockets->second_fd(), F_SETFL, opts | O_NONBLOCK),
              SyscallSucceeds());

  char buf[10] = {};
  ASSERT_THAT(RetryEINTR(send)(sockets->first_fd(), buf, sizeof(buf), 0),
              SyscallSucceedsWithValue(sizeof(buf)));

  // Wait until second_fd sees the data and then recv it.
  struct pollfd poll_fd = {sockets->second_fd(), POLLIN, 0};
  constexpr int kPollTimeoutMs = 2000;  // Wait up to 2 seconds for the data.
  ASSERT_THAT(RetryEINTR(poll)(&poll_fd, 1, kPollTimeoutMs),
              SyscallSucceedsWithValue(1));

  ASSERT_THAT(RetryEINTR(recv)(sockets->second_fd(), buf, sizeof(buf), 0),
              SyscallSucceedsWithValue(sizeof(buf)));

  // Now shutdown the write end leaving the read end open.
  ASSERT_THAT(close(sockets->release_first_fd()), SyscallSucceeds());

  // Wait for close notification and recv again.
  struct pollfd poll_fd2 = {sockets->second_fd(), POLLIN, 0};
  ASSERT_THAT(RetryEINTR(poll)(&poll_fd2, 1, kPollTimeoutMs),
              SyscallSucceedsWithValue(1));

  ASSERT_THAT(RetryEINTR(recv)(sockets->second_fd(), buf, sizeof(buf), 0),
              SyscallSucceedsWithValue(0));
}

TEST_P(TCPSocketPairTest,
       ShutdownRdUnreadDataShouldCauseNoPacketsUnlessClosed) {
  auto sockets = ASSERT_NO_ERRNO_AND_VALUE(NewSocketPair());

  char buf[10] = {};
  ASSERT_THAT(RetryEINTR(write)(sockets->first_fd(), buf, sizeof(buf)),
              SyscallSucceedsWithValue(sizeof(buf)));

  // Wait until t_ sees the data on its side but don't read it.
  struct pollfd poll_fd = {sockets->second_fd(), POLLIN | POLLHUP, 0};
  constexpr int kPollTimeoutMs = 20000;  // Wait up to 20 seconds for the data.
  ASSERT_THAT(RetryEINTR(poll)(&poll_fd, 1, kPollTimeoutMs),
              SyscallSucceedsWithValue(1));

  // Now shutdown the read end, this will generate no packets to the other end.
  ASSERT_THAT(shutdown(sockets->second_fd(), SHUT_RD), SyscallSucceeds());

  // We should not receive any events on the other side of the socket.
  struct pollfd poll_fd2 = {sockets->first_fd(), POLLIN | POLLHUP, 0};
  constexpr int kPollNoResponseTimeoutMs = 3000;
  ASSERT_THAT(RetryEINTR(poll)(&poll_fd2, 1, kPollNoResponseTimeoutMs),
              SyscallSucceedsWithValue(0));  // Timeout.

  // Now since we've fully closed the connection it will generate a RST.
  ASSERT_THAT(close(sockets->release_second_fd()), SyscallSucceeds());
  ASSERT_THAT(RetryEINTR(poll)(&poll_fd2, 1, kPollTimeoutMs),
              SyscallSucceedsWithValue(1));  // The other end has closed.

  // A shutdown with unread data will cause a RST to be sent instead
  // of a FIN, per RFC 2525 section 2.17; this is also what Linux does.
  ASSERT_THAT(RetryEINTR(read)(sockets->first_fd(), buf, sizeof(buf)),
              SyscallFailsWithErrno(ECONNRESET));
}

TEST_P(TCPSocketPairTest, TCPCorkDefault) {
  auto sockets = ASSERT_NO_ERRNO_AND_VALUE(NewSocketPair());

  int get = -1;
  socklen_t get_len = sizeof(get);
  EXPECT_THAT(
      getsockopt(sockets->first_fd(), IPPROTO_TCP, TCP_CORK, &get, &get_len),
      SyscallSucceedsWithValue(0));
  EXPECT_EQ(get_len, sizeof(get));
  EXPECT_EQ(get, kSockOptOff);
}

TEST_P(TCPSocketPairTest, SetTCPCork) {
  auto sockets = ASSERT_NO_ERRNO_AND_VALUE(NewSocketPair());

  ASSERT_THAT(setsockopt(sockets->first_fd(), IPPROTO_TCP, TCP_CORK,
                         &kSockOptOn, sizeof(kSockOptOn)),
              SyscallSucceeds());

  int get = -1;
  socklen_t get_len = sizeof(get);
  EXPECT_THAT(
      getsockopt(sockets->first_fd(), IPPROTO_TCP, TCP_CORK, &get, &get_len),
      SyscallSucceedsWithValue(0));
  EXPECT_EQ(get_len, sizeof(get));
  EXPECT_EQ(get, kSockOptOn);

  ASSERT_THAT(setsockopt(sockets->first_fd(), IPPROTO_TCP, TCP_CORK,
                         &kSockOptOff, sizeof(kSockOptOff)),
              SyscallSucceeds());

  EXPECT_THAT(
      getsockopt(sockets->first_fd(), IPPROTO_TCP, TCP_CORK, &get, &get_len),
      SyscallSucceedsWithValue(0));
  EXPECT_EQ(get_len, sizeof(get));
  EXPECT_EQ(get, kSockOptOff);
}

TEST_P(TCPSocketPairTest, TCPCork) {
  auto sockets = ASSERT_NO_ERRNO_AND_VALUE(NewSocketPair());

  EXPECT_THAT(setsockopt(sockets->first_fd(), IPPROTO_TCP, TCP_CORK,
                         &kSockOptOn, sizeof(kSockOptOn)),
              SyscallSucceeds());

  constexpr char kData[] = "abc";
  ASSERT_THAT(WriteFd(sockets->first_fd(), kData, sizeof(kData)),
              SyscallSucceedsWithValue(sizeof(kData)));

  ASSERT_NO_FATAL_FAILURE(RecvNoData(sockets->second_fd()));

  EXPECT_THAT(setsockopt(sockets->first_fd(), IPPROTO_TCP, TCP_CORK,
                         &kSockOptOff, sizeof(kSockOptOff)),
              SyscallSucceeds());

  // Create a receive buffer larger than kData.
  char buf[(sizeof(kData) + 1) * 2] = {};
  ASSERT_THAT(RetryEINTR(recv)(sockets->second_fd(), buf, sizeof(buf), 0),
              SyscallSucceedsWithValue(sizeof(kData)));
  EXPECT_EQ(absl::string_view(kData, sizeof(kData)),
            absl::string_view(buf, sizeof(kData)));
}

TEST_P(TCPSocketPairTest, TCPQuickAckDefault) {
  auto sockets = ASSERT_NO_ERRNO_AND_VALUE(NewSocketPair());

  int get = -1;
  socklen_t get_len = sizeof(get);
  EXPECT_THAT(getsockopt(sockets->first_fd(), IPPROTO_TCP, TCP_QUICKACK, &get,
                         &get_len),
              SyscallSucceedsWithValue(0));
  EXPECT_EQ(get_len, sizeof(get));
  EXPECT_EQ(get, kSockOptOn);
}

TEST_P(TCPSocketPairTest, SetTCPQuickAck) {
  auto sockets = ASSERT_NO_ERRNO_AND_VALUE(NewSocketPair());

  ASSERT_THAT(setsockopt(sockets->first_fd(), IPPROTO_TCP, TCP_QUICKACK,
                         &kSockOptOff, sizeof(kSockOptOff)),
              SyscallSucceeds());

  int get = -1;
  socklen_t get_len = sizeof(get);
  EXPECT_THAT(getsockopt(sockets->first_fd(), IPPROTO_TCP, TCP_QUICKACK, &get,
                         &get_len),
              SyscallSucceedsWithValue(0));
  EXPECT_EQ(get_len, sizeof(get));
  EXPECT_EQ(get, kSockOptOff);

  ASSERT_THAT(setsockopt(sockets->first_fd(), IPPROTO_TCP, TCP_QUICKACK,
                         &kSockOptOn, sizeof(kSockOptOn)),
              SyscallSucceeds());

  EXPECT_THAT(getsockopt(sockets->first_fd(), IPPROTO_TCP, TCP_QUICKACK, &get,
                         &get_len),
              SyscallSucceedsWithValue(0));
  EXPECT_EQ(get_len, sizeof(get));
  EXPECT_EQ(get, kSockOptOn);
}

TEST_P(TCPSocketPairTest, SoKeepaliveDefault) {
  auto sockets = ASSERT_NO_ERRNO_AND_VALUE(NewSocketPair());

  int get = -1;
  socklen_t get_len = sizeof(get);
  EXPECT_THAT(
      getsockopt(sockets->first_fd(), SOL_SOCKET, SO_KEEPALIVE, &get, &get_len),
      SyscallSucceedsWithValue(0));
  EXPECT_EQ(get_len, sizeof(get));
  EXPECT_EQ(get, kSockOptOff);
}

TEST_P(TCPSocketPairTest, SetSoKeepalive) {
  auto sockets = ASSERT_NO_ERRNO_AND_VALUE(NewSocketPair());

  ASSERT_THAT(setsockopt(sockets->first_fd(), SOL_SOCKET, SO_KEEPALIVE,
                         &kSockOptOn, sizeof(kSockOptOn)),
              SyscallSucceeds());

  int get = -1;
  socklen_t get_len = sizeof(get);
  EXPECT_THAT(
      getsockopt(sockets->first_fd(), SOL_SOCKET, SO_KEEPALIVE, &get, &get_len),
      SyscallSucceedsWithValue(0));
  EXPECT_EQ(get_len, sizeof(get));
  EXPECT_EQ(get, kSockOptOn);

  ASSERT_THAT(setsockopt(sockets->first_fd(), SOL_SOCKET, SO_KEEPALIVE,
                         &kSockOptOff, sizeof(kSockOptOff)),
              SyscallSucceeds());

  EXPECT_THAT(
      getsockopt(sockets->first_fd(), SOL_SOCKET, SO_KEEPALIVE, &get, &get_len),
      SyscallSucceedsWithValue(0));
  EXPECT_EQ(get_len, sizeof(get));
  EXPECT_EQ(get, kSockOptOff);
}

TEST_P(TCPSocketPairTest, TCPKeepidleDefault) {
  auto sockets = ASSERT_NO_ERRNO_AND_VALUE(NewSocketPair());

  int get = -1;
  socklen_t get_len = sizeof(get);
  EXPECT_THAT(getsockopt(sockets->first_fd(), IPPROTO_TCP, TCP_KEEPIDLE, &get,
                         &get_len),
              SyscallSucceedsWithValue(0));
  EXPECT_EQ(get_len, sizeof(get));
  EXPECT_EQ(get, 2 * 60 * 60);  // 2 hours.
}

TEST_P(TCPSocketPairTest, TCPKeepintvlDefault) {
  auto sockets = ASSERT_NO_ERRNO_AND_VALUE(NewSocketPair());

  int get = -1;
  socklen_t get_len = sizeof(get);
  EXPECT_THAT(getsockopt(sockets->first_fd(), IPPROTO_TCP, TCP_KEEPINTVL, &get,
                         &get_len),
              SyscallSucceedsWithValue(0));
  EXPECT_EQ(get_len, sizeof(get));
  EXPECT_EQ(get, 75);  // 75 seconds.
}

TEST_P(TCPSocketPairTest, SetTCPKeepidleZero) {
  auto sockets = ASSERT_NO_ERRNO_AND_VALUE(NewSocketPair());

  constexpr int kZero = 0;
  EXPECT_THAT(setsockopt(sockets->first_fd(), IPPROTO_TCP, TCP_KEEPIDLE, &kZero,
                         sizeof(kZero)),
              SyscallFailsWithErrno(EINVAL));
}

TEST_P(TCPSocketPairTest, SetTCPKeepintvlZero) {
  auto sockets = ASSERT_NO_ERRNO_AND_VALUE(NewSocketPair());

  constexpr int kZero = 0;
  EXPECT_THAT(setsockopt(sockets->first_fd(), IPPROTO_TCP, TCP_KEEPINTVL,
                         &kZero, sizeof(kZero)),
              SyscallFailsWithErrno(EINVAL));
}

// Copied from include/net/tcp.h.
constexpr int MAX_TCP_KEEPIDLE = 32767;
constexpr int MAX_TCP_KEEPINTVL = 32767;

TEST_P(TCPSocketPairTest, SetTCPKeepidleAboveMax) {
  auto sockets = ASSERT_NO_ERRNO_AND_VALUE(NewSocketPair());

  constexpr int kAboveMax = MAX_TCP_KEEPIDLE + 1;
  EXPECT_THAT(setsockopt(sockets->first_fd(), IPPROTO_TCP, TCP_KEEPIDLE,
                         &kAboveMax, sizeof(kAboveMax)),
              SyscallFailsWithErrno(EINVAL));
}

TEST_P(TCPSocketPairTest, SetTCPKeepintvlAboveMax) {
  auto sockets = ASSERT_NO_ERRNO_AND_VALUE(NewSocketPair());

  constexpr int kAboveMax = MAX_TCP_KEEPINTVL + 1;
  EXPECT_THAT(setsockopt(sockets->first_fd(), IPPROTO_TCP, TCP_KEEPINTVL,
                         &kAboveMax, sizeof(kAboveMax)),
              SyscallFailsWithErrno(EINVAL));
}

TEST_P(TCPSocketPairTest, SetTCPKeepidleToMax) {
  auto sockets = ASSERT_NO_ERRNO_AND_VALUE(NewSocketPair());

  EXPECT_THAT(setsockopt(sockets->first_fd(), IPPROTO_TCP, TCP_KEEPIDLE,
                         &MAX_TCP_KEEPIDLE, sizeof(MAX_TCP_KEEPIDLE)),
              SyscallSucceedsWithValue(0));

  int get = -1;
  socklen_t get_len = sizeof(get);
  EXPECT_THAT(getsockopt(sockets->first_fd(), IPPROTO_TCP, TCP_KEEPIDLE, &get,
                         &get_len),
              SyscallSucceedsWithValue(0));
  EXPECT_EQ(get_len, sizeof(get));
  EXPECT_EQ(get, MAX_TCP_KEEPIDLE);
}

TEST_P(TCPSocketPairTest, SetTCPKeepintvlToMax) {
  auto sockets = ASSERT_NO_ERRNO_AND_VALUE(NewSocketPair());

  EXPECT_THAT(setsockopt(sockets->first_fd(), IPPROTO_TCP, TCP_KEEPINTVL,
                         &MAX_TCP_KEEPINTVL, sizeof(MAX_TCP_KEEPINTVL)),
              SyscallSucceedsWithValue(0));

  int get = -1;
  socklen_t get_len = sizeof(get);
  EXPECT_THAT(getsockopt(sockets->first_fd(), IPPROTO_TCP, TCP_KEEPINTVL, &get,
                         &get_len),
              SyscallSucceedsWithValue(0));
  EXPECT_EQ(get_len, sizeof(get));
  EXPECT_EQ(get, MAX_TCP_KEEPINTVL);
}

TEST_P(TCPSocketPairTest, SetOOBInline) {
  auto sockets = ASSERT_NO_ERRNO_AND_VALUE(NewSocketPair());

  EXPECT_THAT(setsockopt(sockets->first_fd(), SOL_SOCKET, SO_OOBINLINE,
                         &kSockOptOn, sizeof(kSockOptOn)),
              SyscallSucceeds());

  int get = -1;
  socklen_t get_len = sizeof(get);
  EXPECT_THAT(
      getsockopt(sockets->first_fd(), SOL_SOCKET, SO_OOBINLINE, &get, &get_len),
      SyscallSucceedsWithValue(0));
  EXPECT_EQ(get_len, sizeof(get));
  EXPECT_EQ(get, kSockOptOn);
}

TEST_P(TCPSocketPairTest, MsgTruncMsgPeek) {
  auto sockets = ASSERT_NO_ERRNO_AND_VALUE(NewSocketPair());

  char sent_data[512];
  RandomizeBuffer(sent_data, sizeof(sent_data));
  ASSERT_THAT(
      RetryEINTR(send)(sockets->first_fd(), sent_data, sizeof(sent_data), 0),
      SyscallSucceedsWithValue(sizeof(sent_data)));

  // Read half of the data with MSG_TRUNC | MSG_PEEK. This way there will still
  // be some data left to read in the next step even if the data gets consumed.
  char received_data1[sizeof(sent_data) / 2] = {};
  ASSERT_THAT(RetryEINTR(recv)(sockets->second_fd(), received_data1,
                               sizeof(received_data1), MSG_TRUNC | MSG_PEEK),
              SyscallSucceedsWithValue(sizeof(received_data1)));

  // Check that we didn't get anything.
  char zeros[sizeof(received_data1)] = {};
  EXPECT_EQ(0, memcmp(zeros, received_data1, sizeof(received_data1)));

  // Check that all of the data is still there.
  char received_data2[sizeof(sent_data)] = {};
  ASSERT_THAT(RetryEINTR(recv)(sockets->second_fd(), received_data2,
                               sizeof(received_data2), 0),
              SyscallSucceedsWithValue(sizeof(sent_data)));

  EXPECT_EQ(0, memcmp(received_data2, sent_data, sizeof(sent_data)));
}

}  // namespace testing
}  // namespace gvisor