File: socket-api-tests.c

package info (click to toggle)
mmlib 1.4.2-2.1
  • links: PTS, VCS
  • area: main
  • in suites: forky, sid, trixie
  • size: 1,716 kB
  • sloc: ansic: 18,071; makefile: 431; sh: 135; python: 63
file content (1225 lines) | stat: -rw-r--r-- 35,829 bytes parent folder | download
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
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
1001
1002
1003
1004
1005
1006
1007
1008
1009
1010
1011
1012
1013
1014
1015
1016
1017
1018
1019
1020
1021
1022
1023
1024
1025
1026
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
1040
1041
1042
1043
1044
1045
1046
1047
1048
1049
1050
1051
1052
1053
1054
1055
1056
1057
1058
1059
1060
1061
1062
1063
1064
1065
1066
1067
1068
1069
1070
1071
1072
1073
1074
1075
1076
1077
1078
1079
1080
1081
1082
1083
1084
1085
1086
1087
1088
1089
1090
1091
1092
1093
1094
1095
1096
1097
1098
1099
1100
1101
1102
1103
1104
1105
1106
1107
1108
1109
1110
1111
1112
1113
1114
1115
1116
1117
1118
1119
1120
1121
1122
1123
1124
1125
1126
1127
1128
1129
1130
1131
1132
1133
1134
1135
1136
1137
1138
1139
1140
1141
1142
1143
1144
1145
1146
1147
1148
1149
1150
1151
1152
1153
1154
1155
1156
1157
1158
1159
1160
1161
1162
1163
1164
1165
1166
1167
1168
1169
1170
1171
1172
1173
1174
1175
1176
1177
1178
1179
1180
1181
1182
1183
1184
1185
1186
1187
1188
1189
1190
1191
1192
1193
1194
1195
1196
1197
1198
1199
1200
1201
1202
1203
1204
1205
1206
1207
1208
1209
1210
1211
1212
1213
1214
1215
1216
1217
1218
1219
1220
1221
1222
1223
1224
1225
/*
   @mindmaze_header@
*/
#if HAVE_CONFIG_H
# include <config.h>
#endif

#include <check.h>
#include <stdarg.h>
#include <stdio.h>
#include <strings.h>

#include "api-testcases.h"
#include "mmerrno.h"
#include "mmlib.h"
#include "mmpredefs.h"
#include "mmsysio.h"
#include "socket-testlib.h"
#include "tests-child-proc.h"

#define PORT	32145
#define MULTIMSG_LEN	6
#define IOV_MAXLEN	8

static size_t socket_data_len[] = {
	4, 16, 64, 325, 512, 1000, 4096, 3000, 6951, 2412, 8192,
};

struct testsocket_config_case {
	int domain;
	int socktype;
};

static
struct testsocket_config_case test_cases[] = {
	{.domain = AF_INET, .socktype = SOCK_STREAM},
	{.domain = AF_INET, .socktype = SOCK_DGRAM},
	{.domain = AF_INET6, .socktype = SOCK_STREAM},
	{.domain = AF_INET6, .socktype = SOCK_DGRAM},
};
#define FIRST_IPV6_TEST_CASE	2


struct childproc {
	int wr_pipe_fd;
	int rd_pipe_fd;
	mm_pid_t pid;
	int sockfd;
	struct sockaddr_storage peer_addr;
	socklen_t peer_addr_len;
};


/**
 * spawn_childproc() - spawn a child test process with connected pipes
 * @child;      childproc structure to init
 * ...:         NULL terminated variable argument list to be supplied to child
 *              process spawn
 *
 * Execute the program tests-child-proc in a child process and connect it with
 * 2 pipe, one for reading and one for writing. The program will execute the
 * function "run_socket_client" with the argument passed in the variable
 * argument list (terminated by NULL). The data regarding child and pipes are
 * stored in @child.
 */
static
void spawn_childproc(struct childproc* child, ...)
{
	int i, pipe_fds_tochild[2], pipe_fds_fromchild[2];
	struct mm_remap_fd fdmap[2];
	va_list args;
	char *argv[16];
	char* arg;

	// Setup argument list array for passing to child creation function
	i = 0;
	argv[i++] = TESTS_CHILD_BIN;
	argv[i++] = "run_socket_client";
	va_start(args, child);
	do {
		arg = va_arg(args, char*);
		argv[i++] = arg;
	} while (arg);
	va_end(args);

	// Create pipe
	ck_assert(mm_pipe(pipe_fds_tochild) == 0);
	ck_assert(mm_pipe(pipe_fds_fromchild) == 0);

	child->wr_pipe_fd = pipe_fds_tochild[1];
	child->rd_pipe_fd = pipe_fds_fromchild[0];

	fdmap[0] = (struct mm_remap_fd){WR_PIPE_FD, pipe_fds_fromchild[1]};
	fdmap[1] = (struct mm_remap_fd){RD_PIPE_FD, pipe_fds_tochild[0]};

	// spawn child process
	ck_assert(mm_spawn(&child->pid, argv[0], 2, fdmap, 0, argv, NULL) == 0);

	// Close pipe end meant for child process
	mm_close(fdmap[0].parent_fd);
	mm_close(fdmap[1].parent_fd);
}


/**
 * clean_childproc() - stop and clean child
 * @child:      pointer to childproc holding the data relative to child
 *
 * Call this function when you want to stop the child process. The order of the
 * connection of connection stop will make the child process to stop normally.
 *
 * Typically called in the teardown function of the tests
 */
static
void clean_childproc(struct childproc* child)
{
	mm_close(child->wr_pipe_fd);
	mm_close(child->sockfd);
	mm_close(child->rd_pipe_fd);

	child->wr_pipe_fd = -1;
	child->sockfd = -1;
	child->rd_pipe_fd = -1;

	if (child->pid)
		mm_wait_process(child->pid, NULL);

	child->pid = 0;
}


/**
 * childproc_order_read() - instruct child to read from its socket end and return data to pipe
 * @child:      pointer to childproc holding the data relative to child
 * @len:        length of the data to try to read from socket
 * @data:       socket_data structure to use to get the data that child will return
 */
static
void childproc_order_read(struct childproc* child, size_t len, struct socket_data* data)
{
	ssize_t rsz;
	struct testsocket_order order = {
		.cmd = READ_SOCKET_DATA,
		.opt_len = len,
	};

	// Instruct (over pipe) child to read data from its socket endpoint
	rsz = mm_write(child->wr_pipe_fd, &order, sizeof(order));
	ck_assert(rsz == (ssize_t)sizeof(order));

	// The child should have returned what it has read from socket onto the
	// pipe, so get it
	ck_assert(pipe_read_socketdata(child->rd_pipe_fd, data) == 0);
}


/**
 * childproc_order_read() - instruct child to write data on its socket end
 * @child:      pointer to childproc holding the data relative to child
 * @data:       socket_data structure containing data child has to send to socket
 */
static
void childproc_order_write(struct childproc* child, struct socket_data* data)
{
	ssize_t rsz;
	struct testsocket_order order = {
		.cmd = WRITE_SOCKET_DATA,
	};

	// Instruct (over pipe) child to write data on its socket endpoint
	rsz = mm_write(child->wr_pipe_fd, &order, sizeof(order));
	ck_assert(rsz == (ssize_t)sizeof(order));

	// Send over the pipe the data that child must send on the socket
	ck_assert(pipe_write_socketdata(child->wr_pipe_fd, data) == 0);
}


/**
 * create_connected_socket_and_child() - setup child and socket to test
 * @child:      pointer to childproc holding the data relative to child
 * @domain:     communications domain in which a socket is to be created (AF_INET or AF_INET6)
 * @socktype:   type of socket to create (SOCK_STREAM or SOCK_DGRAM)
 *
 * This function creates a server socket according to @domain and @socktype,
 * spawn a child process executing tests-child-proc with the function
 * "run_socket_client" which will try to connect the server socket. This
 * process will then:
 *
 * - if TCP: accept the connection and return the connected socket
 * - if UDP: read the first datagram that client is supposed to send in order
 *           to bind the server socket to it.
 *
 * This means that in any case, this function return a connected socket to
 * client (no matter it is SOCK_DGRAM or SOCK_STREAM).
 *
 * This internal data use to communicate with child process is stored in
 * @child.
 *
 * Return: file descriptor of connected socket to child
 */
static
int create_connected_socket_and_child(struct childproc* child,
                                      int domain, int socktype)
{
	int fd;
	int timeout = 500;
	char domain_str[16], socktype_str[16];

	// Create listening socket
	fd = create_server_socket(domain, socktype, PORT);
	ck_assert(fd != -1);

	// Make server socket listening in case of TCP
	if (socktype == SOCK_STREAM)
		ck_assert(mm_listen(fd, 10) == 0);

	// Spawn child for the test
	sprintf(domain_str, "%i", domain);
	sprintf(socktype_str, "%i", socktype);
	spawn_childproc(child,
	                domain_str, socktype_str, "localhost", MM_STRINGIFY(PORT), NULL);

	child->peer_addr_len = sizeof(child->peer_addr);
	if (socktype == SOCK_DGRAM) {
		struct msghdr msg = {
			.msg_name = &child->peer_addr,
			.msg_namelen = child->peer_addr_len,
		};

		// Get first packet from client (when UDP, the child process
		// send a first packet with empty payload). And use it to know
		// the address and connect to the client socket
		ck_assert(mm_recvmsg(fd, &msg, 0) == 0);
		ck_assert(mm_connect(fd, msg.msg_name, msg.msg_namelen) == 0);
		child->peer_addr_len = msg.msg_namelen;

		child->sockfd = fd;
	} else {
		// Accept incoming TCP connection and close listening socket
		child->sockfd = mm_accept(fd, (struct sockaddr*)&child->peer_addr,
		                              &child->peer_addr_len);
		mm_close(fd);
		ck_assert(child->sockfd != -1);
	}

	ck_assert(mm_setsockopt(child->sockfd, SOL_SOCKET, SO_RCVTIMEO, &timeout, sizeof(timeout)) == 0);
	return child->sockfd;
}


/**
 * gen_random_buffer() - fill buffer with random data
 * @buffer:     pointer to buffer to fill
 * @len:        size of buffer
 */
static
void gen_random_buffer(void* buffer, size_t len)
{
	int* int_buf = buffer;
	int rand_val;

	while (len >= sizeof(int)) {
		*(int_buf++) = rand();
		len -= sizeof(int);
	}

	rand_val = rand();
	memcpy(int_buf, &rand_val, len);
}


/**
 * gen_socket_data() - generate random a piece of specific size
 * @data:       socket data to initialize
 * @len:        length of the data
 */
static
void gen_socket_data(struct socket_data* data, size_t len)
{
	ck_assert(len <= sizeof(data->buf));
	data->len = len;
	gen_random_buffer(data->buf, len);
}


/**
 * gen_random_int() - draw a random int from uniform distribution of [0,max]
 * @max:        maximum value that the function can generate
 *
 * Return: random int between 0 and @max
 */
static
int gen_random_int(int max)
{
	return (max+1)*((double)rand()/RAND_MAX);
}


/**
 * gen_random_msg_iov() - generate random split of scatter/gather buffers of msg
 * @num_iov_max:        maximum number of scatter/gather buffers
 * @msg:                msghdr structure whose iov should be configured
 * @buf:                buffer that will back the scatter/gather buffers
 * @buflen:             size of @buf
 *
 * This function configures the msg_iov split of @msg with a random lookup
 * partition of a buffer @buf of size @buflen. The partition is random, nothing
 * prevents that the elements in @msg->msg_iov will not overlap.
 *
 * It is however guaranteed that:
 * - the element in @msg->msg_iov will always point inside @buf
 * - the combined size of @msg->msg_iov elements will be less or equal to @buflen
 */
static
void gen_random_msg_iov(int num_iov_max, struct msghdr* msg, void* buf, size_t buflen)
{
	int i, offset;
	size_t len, rem_sz;
	char* cbuf = buf;

	rem_sz = buflen;
	for (i = 0; i < num_iov_max; i++) {
		offset = gen_random_int(buflen-1);
		len = gen_random_int(buflen-offset);
		if (len > rem_sz)
			len = rem_sz;

		msg->msg_iov[i].iov_base = cbuf + offset;
		msg->msg_iov[i].iov_len = len;

		rem_sz -= len;
		if (rem_sz == 0) {
			i += 1;
			break;
		}
	}

	msg->msg_iovlen = i;
}


/**
 * copy_msg_data_to_buffer() - copy scatter/gather buffers content to flat buffer
 * @msg:        msghdr whose scatter/gather buffers must be copied from
 * @buffer:     flat buffer to which the data must be copied
 */
static
void copy_msg_data_to_buffer(const struct msghdr* msg, void* buffer)
{
	int i;
	size_t iovlen;
	char* dstbuf;
	int num_iov = msg->msg_iovlen;

	dstbuf = buffer;
	for (i = 0; i < num_iov; i++) {
		iovlen = msg->msg_iov[i].iov_len;
		memcpy(dstbuf, msg->msg_iov[i].iov_base, iovlen);
		dstbuf += iovlen;
	}
}


/**
 * copy_buffer_to_msg_data() - copy flat buffer to scatter/gather buffers
 * @buffer:     flat buffer from which which the data must be copied
 * @msg:        msghdr whose scatter/gather buffers must receive the data
 */
static
void copy_buffer_to_msg_data(const void* buffer, const struct msghdr* msg)
{
	int i;
	size_t iovlen;
	const char* srcbuf;

	srcbuf = buffer;
	for (i = 0; i < (int)msg->msg_iovlen; i++) {
		iovlen = msg->msg_iov[i].iov_len;
		memcpy(msg->msg_iov[i].iov_base, srcbuf, iovlen);
		srcbuf += iovlen;
	}
}


/**
 * cmp_msg_dat() compare content of scatter/gather buffers of 2 msghdr
 * msg1:        first message to compare
 * msg2:        second message to compare
 *
 * Return: 0 if the data contained in scatter/gather buffers of both message
 * match, non-zero otherwise.
 */
static
int cmp_msg_data(const struct msghdr* msg1, const struct msghdr* msg2)
{
	int i, r;
	size_t iovlen;
	struct iovec *iov1 = msg1->msg_iov;
	struct iovec *iov2 = msg2->msg_iov;
	int num_iov = msg1->msg_iovlen;

	if (msg1->msg_iovlen != msg2->msg_iovlen)
		return -1;

	for (i = 0; i < num_iov; i++) {
		iovlen = iov1[i].iov_len;
		if (iovlen != iov2[i].iov_len)
			return -1;

		r = memcmp(iov1[i].iov_base, iov2[i].iov_base, iovlen);
		if (r != 0)
			return r;
	}

	return 0;
}


/**
 * get_iov_size() - compute the total size held scatter/gather buffers
 * @msg:        msghdr whose scatter/gather buffers must be inspected
 *
 * Return: summed size of @msg->msg_iov[*].iov_len.
 */
static
size_t get_iov_size(const struct msghdr* msg)
{
	size_t sz;
	int i;
	int num_iov = msg->msg_iovlen;

	sz = 0;
	for (i = 0; i < num_iov; i++)
		sz += msg->msg_iov[i].iov_len;

	return sz;
}


/**
 * copy_msg_iov() - copy the split of scatter/gather buffer of a message
 * @src:        source message
 * @dst:	message to which the split must be copied
 *
 * This function copy the split of @scr into @dst. The buffers element are
 * still backed by the buffer used in @src. Use retarget_msg_iov() to change
 * this.
 */
static
void copy_msg_iov(const struct msghdr* src, struct msghdr* dst)
{
	memcpy(dst->msg_iov, src->msg_iov,
	       src->msg_iovlen*sizeof(src->msg_iov[0]));

	dst->msg_iovlen = src->msg_iovlen;
}


/**
 * retarget_msg_iov() - change the backing flat buffer of the scatter/gather buffers
 * @msg:        message whose iov structures must be changed
 * @prev:       pointer to the previous backing buffer
 * @new:        pointer to the new backing buffer
 *
 * This function assumes that the scatter/gather buffers of the message
 * point inside the same buffer. While this is not the case in general, it
 * is true in the context of those current tests.
 *
 * This takes the @new and @prev argument to compute the offset to apply
 * to the element @msg->msg_iov[*].iov_base so that the new element now
 * point inside the buffer @new. (this assumes that the allocated size of
 * @new and @prev are consistent).
 */
static
void retarget_msg_iov(struct msghdr* msg, const void* prev, const void* new)
{
	int i;
	char* base_ptr;
	int num_iov = msg->msg_iovlen;
	intptr_t offset = (intptr_t)new - (intptr_t)prev;

	for (i = 0; i < num_iov; i++) {
		base_ptr = msg->msg_iov[i].iov_base;
		msg->msg_iov[i].iov_base = base_ptr + offset;
	}
}


/**
 * setup_recvmsg_test_iteration() - prepare messages for recvmsg and request write from child
 * @child:      pointer to childproc holding the data relative to child
 * @size_in_iov: size held by scatter/buffer in messages
 * @msg_ref:    reference message
 * @buff_ref:   flat buffer backing @msg_ref
 * @msg_test:   test message
 * @buff_rest:  flat buffer backing @msg_test
 *
 * setup 2 message using the same split for their scatter/gather buffers but
 * backed by 2 different flat buffers. Generate random data that is copied
 * into @msg_ref (simulating read from net interface into @msg_ref) and
 * order the child to write the same random data onto the socket.
 */
static
void setup_recvmsg_test_iteration(struct childproc* child, size_t size_in_iov,
                                  struct msghdr* msg_ref, void* buff_ref,
                                  struct msghdr* msg_test, void* buff_test)
{
	struct socket_data data;

	memset(buff_ref, 0, DGRAM_MAXSIZE);
	memset(buff_test, 0, DGRAM_MAXSIZE);

	// Generate 2 msg with identical iov split, but backed by 2 different
	// buffers (ie, iov element have same size and same offset but point to
	// 2 different buffers)
	gen_random_msg_iov(IOV_MAXLEN, msg_ref, buff_ref, size_in_iov);
	copy_msg_iov(msg_ref, msg_test);
	retarget_msg_iov(msg_test, buff_ref, buff_test);

	// generate random data that child will feed to socket
	gen_socket_data(&data, get_iov_size(msg_ref));
	copy_buffer_to_msg_data(data.buf, msg_ref);

	// instruct child to read data from pipe and issue a send call to
	// socket with it
	childproc_order_write(child, &data);
}


/**************************************************************************
 *                                                                        *
 *                       socket tests implementation                      *
 *                                                                        *
 **************************************************************************/
static struct childproc child = {
	.wr_pipe_fd = -1,
	.rd_pipe_fd = -1,
	.pid = 0,
	.sockfd = -1,
};


START_TEST(recv_on_localhost)
{
	int sockfd, i, flags;
	size_t req_sz;
	ssize_t rsz;
	struct socket_data data;
	char buffer[DGRAM_MAXSIZE];
	int domain = test_cases[_i].domain;
	int socktype = test_cases[_i].socktype;

	flags = (socktype == SOCK_STREAM) ? MSG_WAITALL : 0;

	// Create connected socket and child process (the created child and
	// socket are cleaned up in teardown)
	sockfd = create_connected_socket_and_child(&child, domain, socktype);

	// Send a piece of data to pipe and check that child has sent this
	// data on socket
	for (i = 0; i < MM_NELEM(socket_data_len); i++) {
		gen_socket_data(&data, socket_data_len[i]);
		childproc_order_write(&child, &data);

		// If not SOCK_STREAM, message boundaries must be kept. So in
		// such a case, received size must equal equal to what we have
		// sent, ie if we request more, we must received what has been
		// sent into the pipe
		req_sz = (socktype == SOCK_STREAM) ? data.len : sizeof(buffer);
		rsz = mm_recv(sockfd, buffer, req_sz, flags);

		ck_assert_int_eq(rsz, data.len);
		ck_assert(memcmp(data.buf, buffer, data.len) == 0);
	}
}
END_TEST


START_TEST(send_on_localhost)
{
	int sockfd, i;
	ssize_t rsz;
	struct socket_data data;
	char buffer[DGRAM_MAXSIZE];
	int domain = test_cases[_i].domain;
	int socktype = test_cases[_i].socktype;
	size_t buf_sz;

	// Create connected socket and child process (the created child and
	// socket are cleaned up in teardown)
	sockfd = create_connected_socket_and_child(&child, domain, socktype);

	// Send a piece of data to pipe and check that child has sent this
	// data on socket
	for (i = 0; i < MM_NELEM(socket_data_len); i++) {
		buf_sz = socket_data_len[i];
		gen_random_buffer(buffer, buf_sz);
		rsz = mm_send(sockfd, buffer, buf_sz, 0);
		ck_assert(rsz >= 0);
		childproc_order_read(&child, sizeof(data.buf), &data);

		ck_assert_int_eq(rsz, data.len);
		ck_assert(memcmp(data.buf, buffer, data.len) == 0);
	}
}
END_TEST


START_TEST(read_on_localhost)
{
	int sockfd, i, k;
	ssize_t rsz, rsz_tmp;
	size_t req_sz;
	struct socket_data data;
	char buffer[DGRAM_MAXSIZE];
	int domain = test_cases[_i].domain;
	int socktype = test_cases[_i].socktype;

	// Create connected socket and child process (the created child and
	// socket are cleaned up in teardown)
	sockfd = create_connected_socket_and_child(&child, domain, socktype);

	// Send a piece of data to pipe and check that child has sent this
	// data on socket
	for (i = 0; i < MM_NELEM(socket_data_len); i++) {
		gen_socket_data(&data, socket_data_len[i]);
		childproc_order_write(&child, &data);

		if (socktype == SOCK_STREAM) {
			// If SOCK_STREAM (TCP), the mm_read() call is allowed
			// to read less than requested
			req_sz = data.len;
			rsz = 0;
			for (k = 0; (k < 100) && (rsz < (ssize_t)req_sz); k++) {
				rsz_tmp = mm_read(sockfd, buffer+rsz, req_sz-rsz);
				if (rsz_tmp < 0)
					break;

				rsz += rsz_tmp;
			}
		} else {
			// If not SOCK_STREAM, message boundaries must be kept.
			// So in such a case, received size must equal to what
			// we have sent, ie if we request more, we must
			// received what has been sent into the pipe
			rsz = mm_read(sockfd, buffer, sizeof(buffer));
		}
		ck_assert_int_eq(rsz, data.len);
		ck_assert(memcmp(data.buf, buffer, data.len) == 0);
	}
}
END_TEST


START_TEST(write_on_localhost)
{
	int sockfd, i;
	ssize_t rsz;
	struct socket_data data;
	char buffer[DGRAM_MAXSIZE];
	int domain = test_cases[_i].domain;
	int socktype = test_cases[_i].socktype;
	size_t buf_sz;

	// Create connected socket and child process (the created child and
	// socket are cleaned up in teardown)
	sockfd = create_connected_socket_and_child(&child, domain, socktype);

	// Send a piece of data to pipe and check that child has sent this
	// data on socket
	for (i = 0; i < MM_NELEM(socket_data_len); i++) {
		buf_sz = socket_data_len[i];
		gen_random_buffer(buffer, buf_sz);
		rsz = mm_write(sockfd, buffer, buf_sz);
		ck_assert_int_eq(rsz, buf_sz);
		childproc_order_read(&child, sizeof(data.buf), &data);

		ck_assert_int_eq(rsz, data.len);
		ck_assert(memcmp(data.buf, buffer, data.len) == 0);
	}
}
END_TEST


START_TEST(recvmsg_on_localhost)
{
	int sockfd, i, flags;
	ssize_t rsz;
	char buffer_test[DGRAM_MAXSIZE], buffer_ref[DGRAM_MAXSIZE];
	int domain = test_cases[_i].domain;
	int socktype = test_cases[_i].socktype;
	struct iovec iov_test[IOV_MAXLEN], iov_ref[IOV_MAXLEN];
	struct msghdr msg_test = {.msg_iov = iov_test};
	struct msghdr msg_ref = {.msg_iov = iov_ref};

	flags = (socktype == SOCK_STREAM) ? MSG_WAITALL : 0;

	// Create connected socket and child process (the created child and
	// socket are cleaned up in teardown)
	sockfd = create_connected_socket_and_child(&child, domain, socktype);

	// Send a piece of data to pipe and check that child has sent this
	// data on socket
	for (i = 0; i < MM_NELEM(socket_data_len); i++) {
		setup_recvmsg_test_iteration(&child, socket_data_len[i],
		                             &msg_ref, buffer_ref,
		                             &msg_test, buffer_test);

		rsz = mm_recvmsg(sockfd, &msg_test, flags);

		// Check data receing from mm_recvmsg match expected results
		ck_assert_int_eq(rsz, get_iov_size(&msg_ref));
		ck_assert(cmp_msg_data(&msg_ref, &msg_test) == 0);
		ck_assert(memcmp(buffer_ref, buffer_test, sizeof(buffer_test)) == 0);
	}
}
END_TEST


START_TEST(sendmsg_on_localhost)
{
	int sockfd, i;
	ssize_t rsz;
	struct socket_data data;
	char data_src[DGRAM_MAXSIZE], buffer_ref[DGRAM_MAXSIZE];
	int domain = test_cases[_i].domain;
	int socktype = test_cases[_i].socktype;
	struct iovec iov[IOV_MAXLEN];
	struct msghdr msg = {.msg_iov = iov};

	// Create connected socket and child process (the created child and
	// socket are cleaned up in teardown)
	sockfd = create_connected_socket_and_child(&child, domain, socktype);

	// Send a piece of data to socket and check that child has received
	// it properly (check this by inspecting what child send over pipe
	// which should match what we sent on socket)
	for (i = 0; i < MM_NELEM(socket_data_len); i++) {
		// Generate random data and iov (within size limits)
		gen_random_msg_iov(IOV_MAXLEN, &msg, data_src, socket_data_len[i]);
		gen_random_buffer(data_src, DGRAM_MAXSIZE);

		// order child to read data from socket obtained through
		// mm_sendmsg
		rsz = mm_sendmsg(sockfd, &msg, 0);
		ck_assert(rsz >= 0);
		childproc_order_read(&child, sizeof(data.buf), &data);

		// Simulate what does sendmsg do
		copy_msg_data_to_buffer(&msg, buffer_ref);

		// Check data received from mm_recvmsg match expected results
		ck_assert_int_eq(rsz, get_iov_size(&msg));
		ck_assert_int_eq(rsz, data.len);
		ck_assert(memcmp(data.buf, buffer_ref, data.len) == 0);
	}
}
END_TEST


START_TEST(recv_multimsg_on_localhost)
{
	int sockfd, i, j, num_msg, flags;
	size_t len;
	char fbuffer_test[MULTIMSG_LEN*DGRAM_MAXSIZE];
	char fbuffer_ref[MULTIMSG_LEN*DGRAM_MAXSIZE];
	char *buff_test, *buff_ref;
	struct iovec iov[MULTIMSG_LEN*IOV_MAXLEN];
	struct iovec iov_ref[MULTIMSG_LEN*IOV_MAXLEN];
	struct msghdr *msg_test, *msg_ref;
	struct mm_sock_multimsg msgvec_ref[MULTIMSG_LEN] = {{.datalen = 0}};
	struct mm_sock_multimsg msgvec_test[MULTIMSG_LEN] = {{.datalen = 0}};
	int domain = test_cases[_i].domain;
	int socktype = test_cases[_i].socktype;

	flags = (socktype == SOCK_STREAM) ? MSG_WAITALL : 0;

	// Create connected socket and child process (the created child and
	// socket are cleaned up in teardown)
	sockfd = create_connected_socket_and_child(&child, domain, socktype);

	// Send a piece of data to pipe and check that child has sent this
	// data on socket
	for (i = 0; i < MM_NELEM(socket_data_len); i++) {
		// Setup multiple msg (ref and test) and issue write order to
		// child so that we can retrieve all those msg data in one call
		// to mm_recv_multimsg()
		for (j = 0; j < MULTIMSG_LEN; j++) {
			msg_ref = &msgvec_ref[j].msg;
			msg_test = &msgvec_test[j].msg;
			buff_ref = fbuffer_ref + j*DGRAM_MAXSIZE;
			buff_test = fbuffer_test + j*DGRAM_MAXSIZE;

			msg_ref->msg_iov = iov_ref + j*IOV_MAXLEN;
			msg_test->msg_iov = iov + j*IOV_MAXLEN;

			setup_recvmsg_test_iteration(&child, socket_data_len[i],
                                                     msg_ref, buff_ref,
			                             msg_test, buff_test);
		}

		// Read all message data
		num_msg = mm_recv_multimsg(sockfd, MULTIMSG_LEN, msgvec_test, flags, NULL);
		ck_assert_int_eq(num_msg, MULTIMSG_LEN);

		// Check data receing from each individual msg match expected results
		for (j = 0; j < MULTIMSG_LEN; j++) {
			msg_ref = &msgvec_ref[j].msg;
			msg_test = &msgvec_test[j].msg;
			len = msgvec_test[j].datalen;
			ck_assert_int_eq(len, get_iov_size(msg_ref));
			ck_assert(cmp_msg_data(msg_ref, msg_test) == 0);
		}
		ck_assert(memcmp(fbuffer_ref, fbuffer_test, sizeof(fbuffer_test)) == 0);
	}
}
END_TEST


START_TEST(send_multimsg_on_localhost)
{
	int sockfd, i, j, num_msg;
	ssize_t len;
	struct socket_data data;
	char fdata_src[MULTIMSG_LEN*DGRAM_MAXSIZE], buff_ref[DGRAM_MAXSIZE];
	char* data_src;
	int domain = test_cases[_i].domain;
	int socktype = test_cases[_i].socktype;
	struct iovec iov[DGRAM_MAXSIZE * IOV_MAXLEN];
	struct msghdr* msg;
	struct mm_sock_multimsg msgvec[MULTIMSG_LEN] = {{.datalen = 0}};

	// Create connected socket and child process (the created child and
	// socket are cleaned up in teardown)
	sockfd = create_connected_socket_and_child(&child, domain, socktype);

	// Send a piece of data to pipe and check that child has sent this
	// data on socket
	for (i = 0; i < MM_NELEM(socket_data_len); i++) {
		for (j = 0; j < MULTIMSG_LEN; j++) {
			msg = &msgvec[j].msg;
			data_src = fdata_src + j*DGRAM_MAXSIZE;
			msg->msg_iov = iov + j*IOV_MAXLEN;

			// Generate random data and iov (within size limits)
			gen_random_msg_iov(IOV_MAXLEN, msg, data_src, socket_data_len[i]);
			gen_random_buffer(data_src, DGRAM_MAXSIZE);
		}

		num_msg = mm_send_multimsg(sockfd, MULTIMSG_LEN, msgvec, 0);
		ck_assert_int_eq(num_msg, MULTIMSG_LEN);

		for (j = 0; j < MULTIMSG_LEN; j++) {
			msg = &msgvec[j].msg;
			len = msgvec[j].datalen;
			ck_assert_int_eq(len, get_iov_size(msg));

			// Simulate what does sendmsg do
			childproc_order_read(&child, len, &data);
			copy_msg_data_to_buffer(msg, buff_ref);

			// Check data receing from mm_recvmsg match expected results
			ck_assert_int_eq(len, data.len);
			ck_assert(memcmp(data.buf, buff_ref, data.len) == 0);
		}
	}
}
END_TEST

START_TEST(test_poll_all_negative)
{
	struct mm_pollfd pollfds = {
		.fd = -1,
		.events = POLLIN | POLLOUT,
	};

	ck_assert(mm_poll(&pollfds, 1, 100) == 0);
}
END_TEST

START_TEST(test_poll_simple)
{
	struct mm_pollfd pollfds = {
		.fd = create_server_socket(AF_INET, SOCK_DGRAM, PORT),
		.events = POLLIN | POLLOUT,
	};

	ck_assert(pollfds.fd > 0);
	ck_assert(mm_poll(&pollfds, 1, 100) == 1);
	ck_assert((pollfds.revents & POLLOUT) != 0);  // can write
	mm_close(pollfds.fd);
}
END_TEST

/* test that if the fd is negative then the corresponding events field is
 * ignored and the revents field returns zero */
START_TEST(test_poll_ignore_negative_socket)
{
	struct mm_pollfd pollfds[] = {
		{
			.fd = -1,
			.events = POLLIN | POLLOUT,
		},
		{
			.fd = create_server_socket(AF_INET, SOCK_DGRAM, PORT),
			.events = POLLIN | POLLOUT,
		},
	};
	ck_assert(mm_poll(pollfds, 2, 100) == 1);  // only one valid fd
	ck_assert((pollfds[0].revents) == 0);  // empty event
	ck_assert((pollfds[1].revents & POLLOUT) != 0);  // can write
	mm_close(pollfds[0].fd);
	mm_close(pollfds[1].fd);
}
END_TEST

START_TEST(test_getsockname)
{
	int rv, fd;
	char service[16];
	struct sockaddr_in addr;
	socklen_t addrlen = sizeof(addr);
	struct addrinfo *res = NULL;
	struct addrinfo hints = {.ai_family = AF_INET, .ai_socktype = SOCK_DGRAM,};

	snprintf(service, sizeof(service), "%i", PORT);
	rv = mm_getaddrinfo(NULL, service, &hints, &res);
	ck_assert(rv == 0 && res != NULL);

	fd = create_server_socket(AF_INET, SOCK_DGRAM, PORT);
	ck_assert(fd > 0);

	rv = mm_getsockname(fd,(struct sockaddr *) &addr, &addrlen);
	ck_assert(rv == 0);

	ck_assert(addr.sin_family == AF_INET);
	ck_assert(ntohs(addr.sin_port) == PORT);
	ck_assert(addr.sin_addr.s_addr == ((struct sockaddr_in *)res->ai_addr)->sin_addr.s_addr);
	mm_freeaddrinfo(res);

	mm_close(fd);
}
END_TEST


START_TEST(test_getpeername)
{
	int sockfd;
	int domain = test_cases[_i].domain;
	int socktype = test_cases[_i].socktype;
	struct sockaddr_storage addr;
	socklen_t addrlen =  sizeof(addr);

	// Create connected socket and child process (the created child and
	// socket are cleaned up in teardown)
	sockfd = create_connected_socket_and_child(&child, domain, socktype);

	ck_assert(mm_getpeername(sockfd, (struct sockaddr*)&addr, &addrlen) != -1);
	ck_assert_int_eq(addrlen, child.peer_addr_len);
	ck_assert(memcmp(&addr, &child.peer_addr, addrlen) == 0);
}
END_TEST


static
void assert_addrinfo(struct addrinfo* rp, int exp_socktype, short exp_port)
{
	struct sockaddr_in* addrin;
	struct sockaddr_in6* addrin6;

	ck_assert(rp->ai_socktype == exp_socktype);
	if (rp->ai_family == AF_INET) {
		addrin = (struct sockaddr_in*)rp->ai_addr;
		ck_assert_int_eq(htons(addrin->sin_port), exp_port);
	} else if (rp->ai_family == AF_INET6) {
		addrin6 = (struct sockaddr_in6*)rp->ai_addr;
		ck_assert_int_eq(htons(addrin6->sin6_port), exp_port);
	} else {
		ck_abort_msg("Not matching address family");
	}
}


START_TEST(getaddrinfo_valid)
{
	struct addrinfo *rp, *res = NULL;
	struct addrinfo hints = {.ai_family = AF_UNSPEC};

	ck_assert(mm_getaddrinfo("localhost", "ssh", &hints, &res) == 0);
	for (rp = res; rp != NULL; rp = rp->ai_next)
		assert_addrinfo(rp, SOCK_STREAM, 22);
	mm_freeaddrinfo(res);

	hints.ai_flags = AI_NUMERICSERV;
	hints.ai_socktype = SOCK_DGRAM;
	ck_assert(mm_getaddrinfo("localhost", "42", &hints, &res) == 0);
	for (rp = res; rp != NULL; rp = rp->ai_next)
		assert_addrinfo(rp, SOCK_DGRAM, 42);
	mm_freeaddrinfo(res);
}
END_TEST


START_TEST(getaddrinfo_error)
{
	struct addrinfo *res = NULL;
	struct addrinfo hints = {.ai_family = AF_INET};

	ck_assert(mm_getaddrinfo("notanhost.localdomain", "ssh", &hints, &res) == -1);
	ck_assert_int_eq(mm_get_lasterror_number(), MM_ENONAME);

	ck_assert(mm_getaddrinfo("localhost", "joke", &hints, &res) == -1);
	ck_assert_int_eq(mm_get_lasterror_number(), MM_ENOTFOUND);

	hints.ai_socktype = SOCK_DGRAM;
	ck_assert(mm_getaddrinfo("localhost", "ssh", &hints, &res) == -1);
	ck_assert_int_eq(mm_get_lasterror_number(), MM_ENOTFOUND);
	hints.ai_socktype = 0;

	ck_assert(mm_getaddrinfo(NULL, NULL, &hints, &res) == -1);
	ck_assert_int_eq(mm_get_lasterror_number(), EINVAL);

	hints.ai_flags = AI_CANONNAME;
	ck_assert(mm_getaddrinfo(NULL, "ssh", &hints, &res) == -1);
	ck_assert_int_eq(mm_get_lasterror_number(), EINVAL);
	hints.ai_flags = 0;

	hints.ai_flags = AI_NUMERICSERV;
	ck_assert(mm_getaddrinfo("localhost", "123joke", &hints, &res) == -1);
	ck_assert_int_eq(mm_get_lasterror_number(), EINVAL);
	hints.ai_flags = 0;

	hints.ai_socktype = SOCK_STREAM;
	hints.ai_protocol = IPPROTO_UDP;
	ck_assert(mm_getaddrinfo("localhost", "ssh", &hints, &res) == -1);
	ck_assert_int_eq(mm_get_lasterror_number(), EPROTOTYPE);
	hints.ai_socktype = 0;
	hints.ai_protocol = 0;
}
END_TEST


/**************************************************************************
 *                                                                        *
 *                           socket helper tests                          *
 *                                                                        *
 **************************************************************************/
static int num_fd_to_close = 0;
static int fds_to_close[8];

static
void clean_helper_test_data(void)
{
	while (num_fd_to_close)
		mm_close(fds_to_close[--num_fd_to_close]);
}


static
void add_fd_to_close(int fd)
{
	ck_assert(num_fd_to_close != MM_NELEM(fds_to_close));
	fds_to_close[num_fd_to_close++] = fd;
}


static
int get_socktype(int fd)
{
	int socktype = -1;
	socklen_t len = sizeof(socktype);

	if (mm_getsockopt(fd, SOL_SOCKET, SO_TYPE, &socktype, &len)
	   || len != sizeof(socktype))
		return -1;

	return socktype;
}


static
int get_peer_port(int fd)
{
	struct sockaddr_in6 addr = {.sin6_port = 0};
	socklen_t len = sizeof(addr);

	if (mm_getpeername(fd, (struct sockaddr*)&addr, &len)
	   || len != sizeof(addr))
		return -1;

	return ntohs(addr.sin6_port);
}


static const struct {
	const char* uri;
	int exp_socktype;
	int exp_port;
} sockclient_cases[] = {
#if _WIN32
	{"msnp://localhost", SOCK_STREAM, 1863},
#else
	{"socks://localhost", SOCK_STREAM, 1080},
#endif
	{"ntp://localhost", SOCK_DGRAM, 123},
	{"tcp://localhost:" MM_STRINGIFY(PORT), SOCK_STREAM, PORT},
	{"udp://localhost:" MM_STRINGIFY(PORT), SOCK_DGRAM, PORT},
};


START_TEST(create_sockclient)
{
	int fd, socktype, port;
	int exp_socktype = sockclient_cases[_i].exp_socktype;
	int exp_port = sockclient_cases[_i].exp_port;
	const char* uri = sockclient_cases[_i].uri;

	// Try to create listening server socket (stream). If the address
	// is already bound, the port is already opened out of the process:
	// this is fine for us, we can use this one.
	if (exp_socktype == SOCK_STREAM) {
		fd = create_server_socket(AF_INET6, SOCK_STREAM, exp_port);
		if (fd != -1) {
			ck_assert(mm_listen(fd, 1) == 0);
			add_fd_to_close(fd);
		} else if (mm_get_lasterror_number() != EADDRINUSE) {
			ck_abort_msg("Cannot create server socket");
		}
	}

	fd = mm_create_sockclient(uri);
	if (fd == -1) {
		fprintf(stderr, "Failed to create server socker with uri: %s, port %d\n"
		        "If you do not support this protocol, or if this port is closed "
		        "you can ignore this failure\n",
		        uri, exp_port);
	}
	ck_assert(fd != -1);

	socktype = get_socktype(fd);
	port = get_peer_port(fd);
	mm_close(fd);

	ck_assert_int_eq(socktype, exp_socktype);
	ck_assert_int_eq(port, exp_port);
}
END_TEST


START_TEST(create_invalid_sockclient)
{
	ck_assert(mm_create_sockclient(NULL) == -1);
	ck_assert_int_eq(mm_get_lasterror_number(), EINVAL);

	ck_assert(mm_create_sockclient("localhost") == -1);
	ck_assert_int_eq(mm_get_lasterror_number(), EINVAL);

	ck_assert(mm_create_sockclient("dummy://localhost") == -1);
	ck_assert_int_eq(mm_get_lasterror_number(), MM_ENOTFOUND);

	ck_assert(mm_create_sockclient("ssh://localhost:10") == -1);
	ck_assert_int_eq(mm_get_lasterror_number(), ECONNREFUSED);

	ck_assert(mm_create_sockclient("tcp://localhost") == -1);
	ck_assert_int_eq(mm_get_lasterror_number(), EINVAL);
}
END_TEST


/**************************************************************************
 *                                                                        *
 *                          Test suite setup                              *
 *                                                                        *
 **************************************************************************/

static
void socket_test_teardown(void)
{
	int flags = mm_error_set_flags(MM_ERROR_SET, MM_ERROR_IGNORE);
	clean_childproc(&child);
	clean_helper_test_data();
	mm_error_set_flags(flags, MM_ERROR_IGNORE);
}


LOCAL_SYMBOL
TCase* create_socket_tcase(void)
{
	int num_cases;

	TCase *tc = tcase_create("socket");

	tcase_add_checked_fixture(tc, NULL, socket_test_teardown);

	num_cases = MM_NELEM(test_cases);
	if (!strcmp(mm_getenv("MMLIB_DISABLE_IPV6_TESTS", "no"), "yes")) {
		fputs("Disable IPv6 socket tests\n", stderr);
		num_cases = FIRST_IPV6_TEST_CASE;
	}

	tcase_add_loop_test(tc, recv_on_localhost, 0, num_cases);
	tcase_add_loop_test(tc, send_on_localhost, 0, num_cases);
	tcase_add_loop_test(tc, read_on_localhost, 0, num_cases);
	tcase_add_loop_test(tc, write_on_localhost, 0, num_cases);
	tcase_add_loop_test(tc, recvmsg_on_localhost, 0, num_cases);
	tcase_add_loop_test(tc, sendmsg_on_localhost, 0, num_cases);
	tcase_add_loop_test(tc, recv_multimsg_on_localhost, 0, num_cases);
	tcase_add_loop_test(tc, send_multimsg_on_localhost, 0, num_cases);

	tcase_add_test(tc, test_poll_all_negative);
	tcase_add_test(tc, test_poll_simple);
	tcase_add_test(tc, test_poll_ignore_negative_socket);
	tcase_add_test(tc, test_getsockname);
	tcase_add_loop_test(tc, test_getpeername, 0, num_cases);
	tcase_add_test(tc, getaddrinfo_valid);
	tcase_add_test(tc, getaddrinfo_error);

	tcase_add_loop_test(tc, create_sockclient,
	                    0, MM_NELEM(sockclient_cases));
	tcase_add_test(tc, create_invalid_sockclient);

	return tc;
}