aboutsummaryrefslogtreecommitdiff
path: root/block/nbd.c
blob: f6ff1c4fb472c054b5b6f2f874530ebdf2a25c5f (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
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
1226
1227
1228
1229
1230
1231
1232
1233
1234
1235
1236
1237
1238
1239
1240
1241
1242
1243
1244
1245
1246
1247
1248
1249
1250
1251
1252
1253
1254
1255
1256
1257
1258
1259
1260
1261
1262
1263
1264
1265
1266
1267
1268
1269
1270
1271
1272
1273
1274
1275
1276
1277
1278
1279
1280
1281
1282
1283
1284
1285
1286
1287
1288
1289
1290
1291
1292
1293
1294
1295
1296
1297
1298
1299
1300
1301
1302
1303
1304
1305
1306
1307
1308
1309
1310
1311
1312
1313
1314
1315
1316
1317
1318
1319
1320
1321
1322
1323
1324
1325
1326
1327
1328
1329
1330
1331
1332
1333
1334
1335
1336
1337
1338
1339
1340
1341
1342
1343
1344
1345
1346
1347
1348
1349
1350
1351
1352
1353
1354
1355
1356
1357
1358
1359
1360
1361
1362
1363
1364
1365
1366
1367
1368
1369
1370
1371
1372
1373
1374
1375
1376
1377
1378
1379
1380
1381
1382
1383
1384
1385
1386
1387
1388
1389
1390
1391
1392
1393
1394
1395
1396
1397
1398
1399
1400
1401
1402
1403
1404
1405
1406
1407
1408
1409
1410
1411
1412
1413
1414
1415
1416
1417
1418
1419
1420
1421
1422
1423
1424
1425
1426
1427
1428
1429
1430
1431
1432
1433
1434
1435
1436
1437
1438
1439
1440
1441
1442
1443
1444
1445
1446
1447
1448
1449
1450
1451
1452
1453
1454
1455
1456
1457
1458
1459
1460
1461
1462
1463
1464
1465
1466
1467
1468
1469
1470
1471
1472
1473
1474
1475
1476
1477
1478
1479
1480
1481
1482
1483
1484
1485
1486
1487
1488
1489
1490
1491
1492
1493
1494
1495
1496
1497
1498
1499
1500
1501
1502
1503
1504
1505
1506
1507
1508
1509
1510
1511
1512
1513
1514
1515
1516
1517
1518
1519
1520
1521
1522
1523
1524
1525
1526
1527
1528
1529
1530
1531
1532
1533
1534
1535
1536
1537
1538
1539
1540
1541
1542
1543
1544
1545
1546
1547
1548
1549
1550
1551
1552
1553
1554
1555
1556
1557
1558
1559
1560
1561
1562
1563
1564
1565
1566
1567
1568
1569
1570
1571
1572
1573
1574
1575
1576
1577
1578
1579
1580
1581
1582
1583
1584
1585
1586
1587
1588
1589
1590
1591
1592
1593
1594
1595
1596
1597
1598
1599
1600
1601
1602
1603
1604
1605
1606
1607
1608
1609
1610
1611
1612
1613
1614
1615
1616
1617
1618
1619
1620
1621
1622
1623
1624
1625
1626
1627
1628
1629
1630
1631
1632
1633
1634
1635
1636
1637
1638
1639
1640
1641
1642
1643
1644
1645
1646
1647
1648
1649
1650
1651
1652
1653
1654
1655
1656
1657
1658
1659
1660
1661
1662
1663
1664
1665
1666
1667
1668
1669
1670
1671
1672
1673
1674
1675
1676
1677
1678
1679
1680
1681
1682
1683
1684
1685
1686
1687
1688
1689
1690
1691
1692
1693
1694
1695
1696
1697
1698
1699
1700
1701
1702
1703
1704
1705
1706
1707
1708
1709
1710
1711
1712
1713
1714
1715
1716
1717
1718
1719
1720
1721
1722
1723
1724
1725
1726
1727
1728
1729
1730
1731
1732
1733
1734
1735
1736
1737
1738
1739
1740
1741
1742
1743
1744
1745
1746
1747
1748
1749
1750
1751
1752
1753
1754
1755
1756
1757
1758
1759
1760
1761
1762
1763
1764
1765
1766
1767
1768
1769
1770
1771
1772
1773
1774
1775
1776
1777
1778
1779
1780
1781
1782
1783
1784
1785
1786
1787
1788
1789
1790
1791
1792
1793
1794
1795
1796
1797
1798
1799
1800
1801
1802
1803
1804
1805
1806
1807
1808
1809
1810
1811
1812
1813
1814
1815
1816
1817
1818
1819
1820
1821
1822
1823
1824
1825
1826
1827
1828
1829
1830
1831
1832
1833
1834
1835
1836
1837
1838
1839
1840
1841
1842
1843
1844
1845
1846
1847
1848
1849
1850
1851
1852
1853
1854
1855
1856
1857
1858
1859
1860
1861
1862
1863
1864
1865
1866
1867
1868
1869
1870
1871
1872
1873
1874
1875
1876
1877
1878
1879
1880
1881
1882
1883
1884
1885
1886
1887
1888
1889
1890
1891
1892
1893
1894
1895
1896
1897
1898
1899
1900
1901
1902
1903
1904
1905
1906
1907
1908
1909
1910
1911
1912
1913
1914
1915
1916
1917
1918
1919
1920
1921
1922
1923
1924
1925
1926
1927
1928
1929
1930
1931
1932
1933
1934
1935
1936
1937
1938
1939
1940
1941
1942
1943
1944
1945
1946
1947
1948
1949
1950
1951
1952
1953
1954
1955
1956
1957
1958
1959
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
2030
2031
2032
2033
2034
2035
2036
2037
2038
2039
2040
2041
2042
2043
2044
2045
2046
2047
2048
2049
2050
2051
2052
2053
2054
2055
2056
2057
2058
2059
2060
2061
2062
2063
2064
2065
2066
2067
2068
2069
2070
2071
2072
2073
2074
2075
2076
2077
2078
2079
2080
2081
2082
2083
2084
2085
2086
2087
2088
2089
2090
2091
2092
2093
2094
2095
2096
2097
2098
2099
2100
2101
2102
2103
2104
2105
2106
2107
2108
2109
2110
2111
2112
2113
2114
2115
2116
2117
2118
2119
2120
2121
2122
2123
2124
2125
2126
2127
2128
2129
2130
2131
2132
2133
2134
2135
2136
2137
2138
2139
2140
2141
2142
2143
2144
2145
2146
2147
2148
2149
2150
2151
2152
2153
2154
2155
2156
2157
2158
2159
2160
2161
2162
2163
2164
2165
2166
2167
2168
2169
2170
2171
2172
2173
2174
2175
2176
2177
2178
2179
2180
2181
2182
2183
2184
2185
2186
2187
2188
2189
2190
2191
2192
2193
2194
2195
2196
2197
2198
2199
2200
2201
2202
2203
2204
2205
2206
2207
2208
2209
2210
2211
2212
2213
2214
2215
2216
2217
2218
2219
2220
2221
2222
2223
2224
2225
2226
2227
2228
2229
2230
2231
2232
2233
2234
2235
2236
2237
2238
2239
2240
2241
2242
2243
/*
 * QEMU Block driver for  NBD
 *
 * Copyright (c) 2019 Virtuozzo International GmbH.
 * Copyright (C) 2016 Red Hat, Inc.
 * Copyright (C) 2008 Bull S.A.S.
 *     Author: Laurent Vivier <Laurent.Vivier@bull.net>
 *
 * Some parts:
 *    Copyright (C) 2007 Anthony Liguori <anthony@codemonkey.ws>
 *
 * Permission is hereby granted, free of charge, to any person obtaining a copy
 * of this software and associated documentation files (the "Software"), to deal
 * in the Software without restriction, including without limitation the rights
 * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
 * copies of the Software, and to permit persons to whom the Software is
 * furnished to do so, subject to the following conditions:
 *
 * The above copyright notice and this permission notice shall be included in
 * all copies or substantial portions of the Software.
 *
 * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
 * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
 * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
 * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
 * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
 * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
 * THE SOFTWARE.
 */

#include "qemu/osdep.h"

#include "trace.h"
#include "qemu/uri.h"
#include "qemu/option.h"
#include "qemu/cutils.h"
#include "qemu/main-loop.h"
#include "qemu/atomic.h"

#include "qapi/qapi-visit-sockets.h"
#include "qapi/qmp/qstring.h"
#include "qapi/clone-visitor.h"

#include "block/qdict.h"
#include "block/nbd.h"
#include "block/block_int.h"
#include "block/coroutines.h"

#include "qemu/yank.h"

#define EN_OPTSTR ":exportname="
#define MAX_NBD_REQUESTS    16

#define HANDLE_TO_INDEX(bs, handle) ((handle) ^ (uint64_t)(intptr_t)(bs))
#define INDEX_TO_HANDLE(bs, index)  ((index)  ^ (uint64_t)(intptr_t)(bs))

typedef struct {
    Coroutine *coroutine;
    uint64_t offset;        /* original offset of the request */
    bool receiving;         /* waiting for connection_co? */
} NBDClientRequest;

typedef enum NBDClientState {
    NBD_CLIENT_CONNECTING_WAIT,
    NBD_CLIENT_CONNECTING_NOWAIT,
    NBD_CLIENT_CONNECTED,
    NBD_CLIENT_QUIT
} NBDClientState;

typedef struct BDRVNBDState {
    QIOChannel *ioc; /* The current I/O channel */
    NBDExportInfo info;

    CoMutex send_mutex;
    CoQueue free_sema;
    Coroutine *connection_co;
    Coroutine *teardown_co;
    QemuCoSleep reconnect_sleep;
    bool drained;
    bool wait_drained_end;
    int in_flight;
    NBDClientState state;
    bool wait_in_flight;

    QEMUTimer *reconnect_delay_timer;

    NBDClientRequest requests[MAX_NBD_REQUESTS];
    NBDReply reply;
    BlockDriverState *bs;

    /* Connection parameters */
    uint32_t reconnect_delay;
    SocketAddress *saddr;
    char *export, *tlscredsid;
    QCryptoTLSCreds *tlscreds;
    const char *hostname;
    char *x_dirty_bitmap;
    bool alloc_depth;

    NBDClientConnection *conn;
} BDRVNBDState;

static void nbd_yank(void *opaque);

static void nbd_clear_bdrvstate(BlockDriverState *bs)
{
    BDRVNBDState *s = (BDRVNBDState *)bs->opaque;

    nbd_client_connection_release(s->conn);
    s->conn = NULL;

    yank_unregister_instance(BLOCKDEV_YANK_INSTANCE(bs->node_name));

    object_unref(OBJECT(s->tlscreds));
    qapi_free_SocketAddress(s->saddr);
    s->saddr = NULL;
    g_free(s->export);
    s->export = NULL;
    g_free(s->tlscredsid);
    s->tlscredsid = NULL;
    g_free(s->x_dirty_bitmap);
    s->x_dirty_bitmap = NULL;
}

static bool nbd_client_connected(BDRVNBDState *s)
{
    return qatomic_load_acquire(&s->state) == NBD_CLIENT_CONNECTED;
}

static void nbd_channel_error(BDRVNBDState *s, int ret)
{
    if (ret == -EIO) {
        if (nbd_client_connected(s)) {
            s->state = s->reconnect_delay ? NBD_CLIENT_CONNECTING_WAIT :
                                            NBD_CLIENT_CONNECTING_NOWAIT;
        }
    } else {
        if (nbd_client_connected(s)) {
            qio_channel_shutdown(s->ioc, QIO_CHANNEL_SHUTDOWN_BOTH, NULL);
        }
        s->state = NBD_CLIENT_QUIT;
    }
}

static void nbd_recv_coroutines_wake_all(BDRVNBDState *s)
{
    int i;

    for (i = 0; i < MAX_NBD_REQUESTS; i++) {
        NBDClientRequest *req = &s->requests[i];

        if (req->coroutine && req->receiving) {
            req->receiving = false;
            aio_co_wake(req->coroutine);
        }
    }
}

static void reconnect_delay_timer_del(BDRVNBDState *s)
{
    if (s->reconnect_delay_timer) {
        timer_free(s->reconnect_delay_timer);
        s->reconnect_delay_timer = NULL;
    }
}

static void reconnect_delay_timer_cb(void *opaque)
{
    BDRVNBDState *s = opaque;

    if (qatomic_load_acquire(&s->state) == NBD_CLIENT_CONNECTING_WAIT) {
        s->state = NBD_CLIENT_CONNECTING_NOWAIT;
        while (qemu_co_enter_next(&s->free_sema, NULL)) {
            /* Resume all queued requests */
        }
    }

    reconnect_delay_timer_del(s);
}

static void reconnect_delay_timer_init(BDRVNBDState *s, uint64_t expire_time_ns)
{
    if (qatomic_load_acquire(&s->state) != NBD_CLIENT_CONNECTING_WAIT) {
        return;
    }

    assert(!s->reconnect_delay_timer);
    s->reconnect_delay_timer = aio_timer_new(bdrv_get_aio_context(s->bs),
                                             QEMU_CLOCK_REALTIME,
                                             SCALE_NS,
                                             reconnect_delay_timer_cb, s);
    timer_mod(s->reconnect_delay_timer, expire_time_ns);
}

static void nbd_client_detach_aio_context(BlockDriverState *bs)
{
    BDRVNBDState *s = (BDRVNBDState *)bs->opaque;

    /* Timer is deleted in nbd_client_co_drain_begin() */
    assert(!s->reconnect_delay_timer);
    /*
     * If reconnect is in progress we may have no ->ioc.  It will be
     * re-instantiated in the proper aio context once the connection is
     * reestablished.
     */
    if (s->ioc) {
        qio_channel_detach_aio_context(QIO_CHANNEL(s->ioc));
    }
}

static void nbd_client_attach_aio_context_bh(void *opaque)
{
    BlockDriverState *bs = opaque;
    BDRVNBDState *s = (BDRVNBDState *)bs->opaque;

    if (s->connection_co) {
        /*
         * The node is still drained, so we know the coroutine has yielded in
         * nbd_read_eof(), the only place where bs->in_flight can reach 0, or
         * it is entered for the first time. Both places are safe for entering
         * the coroutine.
         */
        qemu_aio_coroutine_enter(bs->aio_context, s->connection_co);
    }
    bdrv_dec_in_flight(bs);
}

static void nbd_client_attach_aio_context(BlockDriverState *bs,
                                          AioContext *new_context)
{
    BDRVNBDState *s = (BDRVNBDState *)bs->opaque;

    /*
     * s->connection_co is either yielded from nbd_receive_reply or from
     * nbd_co_reconnect_loop()
     */
    if (nbd_client_connected(s)) {
        qio_channel_attach_aio_context(QIO_CHANNEL(s->ioc), new_context);
    }

    bdrv_inc_in_flight(bs);

    /*
     * Need to wait here for the BH to run because the BH must run while the
     * node is still drained.
     */
    aio_wait_bh_oneshot(new_context, nbd_client_attach_aio_context_bh, bs);
}

static void coroutine_fn nbd_client_co_drain_begin(BlockDriverState *bs)
{
    BDRVNBDState *s = (BDRVNBDState *)bs->opaque;

    s->drained = true;
    qemu_co_sleep_wake(&s->reconnect_sleep);

    nbd_co_establish_connection_cancel(s->conn);

    reconnect_delay_timer_del(s);

    if (qatomic_load_acquire(&s->state) == NBD_CLIENT_CONNECTING_WAIT) {
        s->state = NBD_CLIENT_CONNECTING_NOWAIT;
        qemu_co_queue_restart_all(&s->free_sema);
    }
}

static void coroutine_fn nbd_client_co_drain_end(BlockDriverState *bs)
{
    BDRVNBDState *s = (BDRVNBDState *)bs->opaque;

    s->drained = false;
    if (s->wait_drained_end) {
        s->wait_drained_end = false;
        aio_co_wake(s->connection_co);
    }
}


static void nbd_teardown_connection(BlockDriverState *bs)
{
    BDRVNBDState *s = (BDRVNBDState *)bs->opaque;

    if (s->ioc) {
        /* finish any pending coroutines */
        qio_channel_shutdown(s->ioc, QIO_CHANNEL_SHUTDOWN_BOTH, NULL);
    }

    s->state = NBD_CLIENT_QUIT;
    if (s->connection_co) {
        qemu_co_sleep_wake(&s->reconnect_sleep);
        nbd_co_establish_connection_cancel(s->conn);
    }
    if (qemu_in_coroutine()) {
        s->teardown_co = qemu_coroutine_self();
        /* connection_co resumes us when it terminates */
        qemu_coroutine_yield();
        s->teardown_co = NULL;
    } else {
        BDRV_POLL_WHILE(bs, s->connection_co);
    }
    assert(!s->connection_co);
}

static bool nbd_client_connecting(BDRVNBDState *s)
{
    NBDClientState state = qatomic_load_acquire(&s->state);
    return state == NBD_CLIENT_CONNECTING_WAIT ||
        state == NBD_CLIENT_CONNECTING_NOWAIT;
}

static bool nbd_client_connecting_wait(BDRVNBDState *s)
{
    return qatomic_load_acquire(&s->state) == NBD_CLIENT_CONNECTING_WAIT;
}

/*
 * Update @bs with information learned during a completed negotiation process.
 * Return failure if the server's advertised options are incompatible with the
 * client's needs.
 */
static int nbd_handle_updated_info(BlockDriverState *bs, Error **errp)
{
    BDRVNBDState *s = (BDRVNBDState *)bs->opaque;
    int ret;

    if (s->x_dirty_bitmap) {
        if (!s->info.base_allocation) {
            error_setg(errp, "requested x-dirty-bitmap %s not found",
                       s->x_dirty_bitmap);
            return -EINVAL;
        }
        if (strcmp(s->x_dirty_bitmap, "qemu:allocation-depth") == 0) {
            s->alloc_depth = true;
        }
    }

    if (s->info.flags & NBD_FLAG_READ_ONLY) {
        ret = bdrv_apply_auto_read_only(bs, "NBD export is read-only", errp);
        if (ret < 0) {
            return ret;
        }
    }

    if (s->info.flags & NBD_FLAG_SEND_FUA) {
        bs->supported_write_flags = BDRV_REQ_FUA;
        bs->supported_zero_flags |= BDRV_REQ_FUA;
    }

    if (s->info.flags & NBD_FLAG_SEND_WRITE_ZEROES) {
        bs->supported_zero_flags |= BDRV_REQ_MAY_UNMAP;
        if (s->info.flags & NBD_FLAG_SEND_FAST_ZERO) {
            bs->supported_zero_flags |= BDRV_REQ_NO_FALLBACK;
        }
    }

    trace_nbd_client_handshake_success(s->export);

    return 0;
}

int coroutine_fn nbd_co_do_establish_connection(BlockDriverState *bs,
                                                Error **errp)
{
    BDRVNBDState *s = (BDRVNBDState *)bs->opaque;
    int ret;

    assert(!s->ioc);

    s->ioc = nbd_co_establish_connection(s->conn, &s->info, true, errp);
    if (!s->ioc) {
        return -ECONNREFUSED;
    }

    yank_register_function(BLOCKDEV_YANK_INSTANCE(s->bs->node_name), nbd_yank,
                           bs);

    ret = nbd_handle_updated_info(s->bs, NULL);
    if (ret < 0) {
        /*
         * We have connected, but must fail for other reasons.
         * Send NBD_CMD_DISC as a courtesy to the server.
         */
        NBDRequest request = { .type = NBD_CMD_DISC };

        nbd_send_request(s->ioc, &request);

        yank_unregister_function(BLOCKDEV_YANK_INSTANCE(s->bs->node_name),
                                 nbd_yank, bs);
        object_unref(OBJECT(s->ioc));
        s->ioc = NULL;

        return ret;
    }

    qio_channel_set_blocking(s->ioc, false, NULL);
    qio_channel_attach_aio_context(s->ioc, bdrv_get_aio_context(bs));

    /* successfully connected */
    s->state = NBD_CLIENT_CONNECTED;
    qemu_co_queue_restart_all(&s->free_sema);

    return 0;
}

static coroutine_fn void nbd_reconnect_attempt(BDRVNBDState *s)
{
    if (!nbd_client_connecting(s)) {
        return;
    }

    /* Wait for completion of all in-flight requests */

    qemu_co_mutex_lock(&s->send_mutex);

    while (s->in_flight > 0) {
        qemu_co_mutex_unlock(&s->send_mutex);
        nbd_recv_coroutines_wake_all(s);
        s->wait_in_flight = true;
        qemu_coroutine_yield();
        s->wait_in_flight = false;
        qemu_co_mutex_lock(&s->send_mutex);
    }

    qemu_co_mutex_unlock(&s->send_mutex);

    if (!nbd_client_connecting(s)) {
        return;
    }

    /*
     * Now we are sure that nobody is accessing the channel, and no one will
     * try until we set the state to CONNECTED.
     */

    /* Finalize previous connection if any */
    if (s->ioc) {
        qio_channel_detach_aio_context(QIO_CHANNEL(s->ioc));
        yank_unregister_function(BLOCKDEV_YANK_INSTANCE(s->bs->node_name),
                                 nbd_yank, s->bs);
        object_unref(OBJECT(s->ioc));
        s->ioc = NULL;
    }

    nbd_co_do_establish_connection(s->bs, NULL);
}

static coroutine_fn void nbd_co_reconnect_loop(BDRVNBDState *s)
{
    uint64_t timeout = 1 * NANOSECONDS_PER_SECOND;
    uint64_t max_timeout = 16 * NANOSECONDS_PER_SECOND;

    if (qatomic_load_acquire(&s->state) == NBD_CLIENT_CONNECTING_WAIT) {
        reconnect_delay_timer_init(s, qemu_clock_get_ns(QEMU_CLOCK_REALTIME) +
                                   s->reconnect_delay * NANOSECONDS_PER_SECOND);
    }

    nbd_reconnect_attempt(s);

    while (nbd_client_connecting(s)) {
        if (s->drained) {
            bdrv_dec_in_flight(s->bs);
            s->wait_drained_end = true;
            while (s->drained) {
                /*
                 * We may be entered once from nbd_client_attach_aio_context_bh
                 * and then from nbd_client_co_drain_end. So here is a loop.
                 */
                qemu_coroutine_yield();
            }
            bdrv_inc_in_flight(s->bs);
        } else {
            qemu_co_sleep_ns_wakeable(&s->reconnect_sleep,
                                      QEMU_CLOCK_REALTIME, timeout);
            if (s->drained) {
                continue;
            }
            if (timeout < max_timeout) {
                timeout *= 2;
            }
        }

        nbd_reconnect_attempt(s);
    }

    reconnect_delay_timer_del(s);
}

static coroutine_fn void nbd_connection_entry(void *opaque)
{
    BDRVNBDState *s = opaque;
    uint64_t i;
    int ret = 0;
    Error *local_err = NULL;

    while (qatomic_load_acquire(&s->state) != NBD_CLIENT_QUIT) {
        /*
         * The NBD client can only really be considered idle when it has
         * yielded from qio_channel_readv_all_eof(), waiting for data. This is
         * the point where the additional scheduled coroutine entry happens
         * after nbd_client_attach_aio_context().
         *
         * Therefore we keep an additional in_flight reference all the time and
         * only drop it temporarily here.
         */

        if (nbd_client_connecting(s)) {
            nbd_co_reconnect_loop(s);
        }

        if (!nbd_client_connected(s)) {
            continue;
        }

        assert(s->reply.handle == 0);
        ret = nbd_receive_reply(s->bs, s->ioc, &s->reply, &local_err);

        if (local_err) {
            trace_nbd_read_reply_entry_fail(ret, error_get_pretty(local_err));
            error_free(local_err);
            local_err = NULL;
        }
        if (ret <= 0) {
            nbd_channel_error(s, ret ? ret : -EIO);
            continue;
        }

        /*
         * There's no need for a mutex on the receive side, because the
         * handler acts as a synchronization point and ensures that only
         * one coroutine is called until the reply finishes.
         */
        i = HANDLE_TO_INDEX(s, s->reply.handle);
        if (i >= MAX_NBD_REQUESTS ||
            !s->requests[i].coroutine ||
            !s->requests[i].receiving ||
            (nbd_reply_is_structured(&s->reply) && !s->info.structured_reply))
        {
            nbd_channel_error(s, -EINVAL);
            continue;
        }

        /*
         * We're woken up again by the request itself.  Note that there
         * is no race between yielding and reentering connection_co.  This
         * is because:
         *
         * - if the request runs on the same AioContext, it is only
         *   entered after we yield
         *
         * - if the request runs on a different AioContext, reentering
         *   connection_co happens through a bottom half, which can only
         *   run after we yield.
         */
        s->requests[i].receiving = false;
        aio_co_wake(s->requests[i].coroutine);
        qemu_coroutine_yield();
    }

    qemu_co_queue_restart_all(&s->free_sema);
    nbd_recv_coroutines_wake_all(s);
    bdrv_dec_in_flight(s->bs);

    s->connection_co = NULL;
    if (s->ioc) {
        qio_channel_detach_aio_context(QIO_CHANNEL(s->ioc));
        yank_unregister_function(BLOCKDEV_YANK_INSTANCE(s->bs->node_name),
                                 nbd_yank, s->bs);
        object_unref(OBJECT(s->ioc));
        s->ioc = NULL;
    }

    if (s->teardown_co) {
        aio_co_wake(s->teardown_co);
    }
    aio_wait_kick();
}

static int nbd_co_send_request(BlockDriverState *bs,
                               NBDRequest *request,
                               QEMUIOVector *qiov)
{
    BDRVNBDState *s = (BDRVNBDState *)bs->opaque;
    int rc, i = -1;

    qemu_co_mutex_lock(&s->send_mutex);
    while (s->in_flight == MAX_NBD_REQUESTS || nbd_client_connecting_wait(s)) {
        qemu_co_queue_wait(&s->free_sema, &s->send_mutex);
    }

    if (!nbd_client_connected(s)) {
        rc = -EIO;
        goto err;
    }

    s->in_flight++;

    for (i = 0; i < MAX_NBD_REQUESTS; i++) {
        if (s->requests[i].coroutine == NULL) {
            break;
        }
    }

    g_assert(qemu_in_coroutine());
    assert(i < MAX_NBD_REQUESTS);

    s->requests[i].coroutine = qemu_coroutine_self();
    s->requests[i].offset = request->from;
    s->requests[i].receiving = false;

    request->handle = INDEX_TO_HANDLE(s, i);

    assert(s->ioc);

    if (qiov) {
        qio_channel_set_cork(s->ioc, true);
        rc = nbd_send_request(s->ioc, request);
        if (nbd_client_connected(s) && rc >= 0) {
            if (qio_channel_writev_all(s->ioc, qiov->iov, qiov->niov,
                                       NULL) < 0) {
                rc = -EIO;
            }
        } else if (rc >= 0) {
            rc = -EIO;
        }
        qio_channel_set_cork(s->ioc, false);
    } else {
        rc = nbd_send_request(s->ioc, request);
    }

err:
    if (rc < 0) {
        nbd_channel_error(s, rc);
        if (i != -1) {
            s->requests[i].coroutine = NULL;
            s->in_flight--;
        }
        if (s->in_flight == 0 && s->wait_in_flight) {
            aio_co_wake(s->connection_co);
        } else {
            qemu_co_queue_next(&s->free_sema);
        }
    }
    qemu_co_mutex_unlock(&s->send_mutex);
    return rc;
}

static inline uint16_t payload_advance16(uint8_t **payload)
{
    *payload += 2;
    return lduw_be_p(*payload - 2);
}

static inline uint32_t payload_advance32(uint8_t **payload)
{
    *payload += 4;
    return ldl_be_p(*payload - 4);
}

static inline uint64_t payload_advance64(uint8_t **payload)
{
    *payload += 8;
    return ldq_be_p(*payload - 8);
}

static int nbd_parse_offset_hole_payload(BDRVNBDState *s,
                                         NBDStructuredReplyChunk *chunk,
                                         uint8_t *payload, uint64_t orig_offset,
                                         QEMUIOVector *qiov, Error **errp)
{
    uint64_t offset;
    uint32_t hole_size;

    if (chunk->length != sizeof(offset) + sizeof(hole_size)) {
        error_setg(errp, "Protocol error: invalid payload for "
                         "NBD_REPLY_TYPE_OFFSET_HOLE");
        return -EINVAL;
    }

    offset = payload_advance64(&payload);
    hole_size = payload_advance32(&payload);

    if (!hole_size || offset < orig_offset || hole_size > qiov->size ||
        offset > orig_offset + qiov->size - hole_size) {
        error_setg(errp, "Protocol error: server sent chunk exceeding requested"
                         " region");
        return -EINVAL;
    }
    if (s->info.min_block &&
        !QEMU_IS_ALIGNED(hole_size, s->info.min_block)) {
        trace_nbd_structured_read_compliance("hole");
    }

    qemu_iovec_memset(qiov, offset - orig_offset, 0, hole_size);

    return 0;
}

/*
 * nbd_parse_blockstatus_payload
 * Based on our request, we expect only one extent in reply, for the
 * base:allocation context.
 */
static int nbd_parse_blockstatus_payload(BDRVNBDState *s,
                                         NBDStructuredReplyChunk *chunk,
                                         uint8_t *payload, uint64_t orig_length,
                                         NBDExtent *extent, Error **errp)
{
    uint32_t context_id;

    /* The server succeeded, so it must have sent [at least] one extent */
    if (chunk->length < sizeof(context_id) + sizeof(*extent)) {
        error_setg(errp, "Protocol error: invalid payload for "
                         "NBD_REPLY_TYPE_BLOCK_STATUS");
        return -EINVAL;
    }

    context_id = payload_advance32(&payload);
    if (s->info.context_id != context_id) {
        error_setg(errp, "Protocol error: unexpected context id %d for "
                         "NBD_REPLY_TYPE_BLOCK_STATUS, when negotiated context "
                         "id is %d", context_id,
                         s->info.context_id);
        return -EINVAL;
    }

    extent->length = payload_advance32(&payload);
    extent->flags = payload_advance32(&payload);

    if (extent->length == 0) {
        error_setg(errp, "Protocol error: server sent status chunk with "
                   "zero length");
        return -EINVAL;
    }

    /*
     * A server sending unaligned block status is in violation of the
     * protocol, but as qemu-nbd 3.1 is such a server (at least for
     * POSIX files that are not a multiple of 512 bytes, since qemu
     * rounds files up to 512-byte multiples but lseek(SEEK_HOLE)
     * still sees an implicit hole beyond the real EOF), it's nicer to
     * work around the misbehaving server. If the request included
     * more than the final unaligned block, truncate it back to an
     * aligned result; if the request was only the final block, round
     * up to the full block and change the status to fully-allocated
     * (always a safe status, even if it loses information).
     */
    if (s->info.min_block && !QEMU_IS_ALIGNED(extent->length,
                                                   s->info.min_block)) {
        trace_nbd_parse_blockstatus_compliance("extent length is unaligned");
        if (extent->length > s->info.min_block) {
            extent->length = QEMU_ALIGN_DOWN(extent->length,
                                             s->info.min_block);
        } else {
            extent->length = s->info.min_block;
            extent->flags = 0;
        }
    }

    /*
     * We used NBD_CMD_FLAG_REQ_ONE, so the server should not have
     * sent us any more than one extent, nor should it have included
     * status beyond our request in that extent. However, it's easy
     * enough to ignore the server's noncompliance without killing the
     * connection; just ignore trailing extents, and clamp things to
     * the length of our request.
     */
    if (chunk->length > sizeof(context_id) + sizeof(*extent)) {
        trace_nbd_parse_blockstatus_compliance("more than one extent");
    }
    if (extent->length > orig_length) {
        extent->length = orig_length;
        trace_nbd_parse_blockstatus_compliance("extent length too large");
    }

    /*
     * HACK: if we are using x-dirty-bitmaps to access
     * qemu:allocation-depth, treat all depths > 2 the same as 2,
     * since nbd_client_co_block_status is only expecting the low two
     * bits to be set.
     */
    if (s->alloc_depth && extent->flags > 2) {
        extent->flags = 2;
    }

    return 0;
}

/*
 * nbd_parse_error_payload
 * on success @errp contains message describing nbd error reply
 */
static int nbd_parse_error_payload(NBDStructuredReplyChunk *chunk,
                                   uint8_t *payload, int *request_ret,
                                   Error **errp)
{
    uint32_t error;
    uint16_t message_size;

    assert(chunk->type & (1 << 15));

    if (chunk->length < sizeof(error) + sizeof(message_size)) {
        error_setg(errp,
                   "Protocol error: invalid payload for structured error");
        return -EINVAL;
    }

    error = nbd_errno_to_system_errno(payload_advance32(&payload));
    if (error == 0) {
        error_setg(errp, "Protocol error: server sent structured error chunk "
                         "with error = 0");
        return -EINVAL;
    }

    *request_ret = -error;
    message_size = payload_advance16(&payload);

    if (message_size > chunk->length - sizeof(error) - sizeof(message_size)) {
        error_setg(errp, "Protocol error: server sent structured error chunk "
                         "with incorrect message size");
        return -EINVAL;
    }

    /* TODO: Add a trace point to mention the server complaint */

    /* TODO handle ERROR_OFFSET */

    return 0;
}

static int nbd_co_receive_offset_data_payload(BDRVNBDState *s,
                                              uint64_t orig_offset,
                                              QEMUIOVector *qiov, Error **errp)
{
    QEMUIOVector sub_qiov;
    uint64_t offset;
    size_t data_size;
    int ret;
    NBDStructuredReplyChunk *chunk = &s->reply.structured;

    assert(nbd_reply_is_structured(&s->reply));

    /* The NBD spec requires at least one byte of payload */
    if (chunk->length <= sizeof(offset)) {
        error_setg(errp, "Protocol error: invalid payload for "
                         "NBD_REPLY_TYPE_OFFSET_DATA");
        return -EINVAL;
    }

    if (nbd_read64(s->ioc, &offset, "OFFSET_DATA offset", errp) < 0) {
        return -EIO;
    }

    data_size = chunk->length - sizeof(offset);
    assert(data_size);
    if (offset < orig_offset || data_size > qiov->size ||
        offset > orig_offset + qiov->size - data_size) {
        error_setg(errp, "Protocol error: server sent chunk exceeding requested"
                         " region");
        return -EINVAL;
    }
    if (s->info.min_block && !QEMU_IS_ALIGNED(data_size, s->info.min_block)) {
        trace_nbd_structured_read_compliance("data");
    }

    qemu_iovec_init(&sub_qiov, qiov->niov);
    qemu_iovec_concat(&sub_qiov, qiov, offset - orig_offset, data_size);
    ret = qio_channel_readv_all(s->ioc, sub_qiov.iov, sub_qiov.niov, errp);
    qemu_iovec_destroy(&sub_qiov);

    return ret < 0 ? -EIO : 0;
}

#define NBD_MAX_MALLOC_PAYLOAD 1000
static coroutine_fn int nbd_co_receive_structured_payload(
        BDRVNBDState *s, void **payload, Error **errp)
{
    int ret;
    uint32_t len;

    assert(nbd_reply_is_structured(&s->reply));

    len = s->reply.structured.length;

    if (len == 0) {
        return 0;
    }

    if (payload == NULL) {
        error_setg(errp, "Unexpected structured payload");
        return -EINVAL;
    }

    if (len > NBD_MAX_MALLOC_PAYLOAD) {
        error_setg(errp, "Payload too large");
        return -EINVAL;
    }

    *payload = g_new(char, len);
    ret = nbd_read(s->ioc, *payload, len, "structured payload", errp);
    if (ret < 0) {
        g_free(*payload);
        *payload = NULL;
        return ret;
    }

    return 0;
}

/*
 * nbd_co_do_receive_one_chunk
 * for simple reply:
 *   set request_ret to received reply error
 *   if qiov is not NULL: read payload to @qiov
 * for structured reply chunk:
 *   if error chunk: read payload, set @request_ret, do not set @payload
 *   else if offset_data chunk: read payload data to @qiov, do not set @payload
 *   else: read payload to @payload
 *
 * If function fails, @errp contains corresponding error message, and the
 * connection with the server is suspect.  If it returns 0, then the
 * transaction succeeded (although @request_ret may be a negative errno
 * corresponding to the server's error reply), and errp is unchanged.
 */
static coroutine_fn int nbd_co_do_receive_one_chunk(
        BDRVNBDState *s, uint64_t handle, bool only_structured,
        int *request_ret, QEMUIOVector *qiov, void **payload, Error **errp)
{
    int ret;
    int i = HANDLE_TO_INDEX(s, handle);
    void *local_payload = NULL;
    NBDStructuredReplyChunk *chunk;

    if (payload) {
        *payload = NULL;
    }
    *request_ret = 0;

    /* Wait until we're woken up by nbd_connection_entry.  */
    s->requests[i].receiving = true;
    qemu_coroutine_yield();
    assert(!s->requests[i].receiving);
    if (!nbd_client_connected(s)) {
        error_setg(errp, "Connection closed");
        return -EIO;
    }
    assert(s->ioc);

    assert(s->reply.handle == handle);

    if (nbd_reply_is_simple(&s->reply)) {
        if (only_structured) {
            error_setg(errp, "Protocol error: simple reply when structured "
                             "reply chunk was expected");
            return -EINVAL;
        }

        *request_ret = -nbd_errno_to_system_errno(s->reply.simple.error);
        if (*request_ret < 0 || !qiov) {
            return 0;
        }

        return qio_channel_readv_all(s->ioc, qiov->iov, qiov->niov,
                                     errp) < 0 ? -EIO : 0;
    }

    /* handle structured reply chunk */
    assert(s->info.structured_reply);
    chunk = &s->reply.structured;

    if (chunk->type == NBD_REPLY_TYPE_NONE) {
        if (!(chunk->flags & NBD_REPLY_FLAG_DONE)) {
            error_setg(errp, "Protocol error: NBD_REPLY_TYPE_NONE chunk without"
                       " NBD_REPLY_FLAG_DONE flag set");
            return -EINVAL;
        }
        if (chunk->length) {
            error_setg(errp, "Protocol error: NBD_REPLY_TYPE_NONE chunk with"
                       " nonzero length");
            return -EINVAL;
        }
        return 0;
    }

    if (chunk->type == NBD_REPLY_TYPE_OFFSET_DATA) {
        if (!qiov) {
            error_setg(errp, "Unexpected NBD_REPLY_TYPE_OFFSET_DATA chunk");
            return -EINVAL;
        }

        return nbd_co_receive_offset_data_payload(s, s->requests[i].offset,
                                                  qiov, errp);
    }

    if (nbd_reply_type_is_error(chunk->type)) {
        payload = &local_payload;
    }

    ret = nbd_co_receive_structured_payload(s, payload, errp);
    if (ret < 0) {
        return ret;
    }

    if (nbd_reply_type_is_error(chunk->type)) {
        ret = nbd_parse_error_payload(chunk, local_payload, request_ret, errp);
        g_free(local_payload);
        return ret;
    }

    return 0;
}

/*
 * nbd_co_receive_one_chunk
 * Read reply, wake up connection_co and set s->quit if needed.
 * Return value is a fatal error code or normal nbd reply error code
 */
static coroutine_fn int nbd_co_receive_one_chunk(
        BDRVNBDState *s, uint64_t handle, bool only_structured,
        int *request_ret, QEMUIOVector *qiov, NBDReply *reply, void **payload,
        Error **errp)
{
    int ret = nbd_co_do_receive_one_chunk(s, handle, only_structured,
                                          request_ret, qiov, payload, errp);

    if (ret < 0) {
        memset(reply, 0, sizeof(*reply));
        nbd_channel_error(s, ret);
    } else {
        /* For assert at loop start in nbd_connection_entry */
        *reply = s->reply;
    }
    s->reply.handle = 0;

    if (s->connection_co && !s->wait_in_flight) {
        /*
         * We must check s->wait_in_flight, because we may entered by
         * nbd_recv_coroutines_wake_all(), in this case we should not
         * wake connection_co here, it will woken by last request.
         */
        aio_co_wake(s->connection_co);
    }

    return ret;
}

typedef struct NBDReplyChunkIter {
    int ret;
    int request_ret;
    Error *err;
    bool done, only_structured;
} NBDReplyChunkIter;

static void nbd_iter_channel_error(NBDReplyChunkIter *iter,
                                   int ret, Error **local_err)
{
    assert(local_err && *local_err);
    assert(ret < 0);

    if (!iter->ret) {
        iter->ret = ret;
        error_propagate(&iter->err, *local_err);
    } else {
        error_free(*local_err);
    }

    *local_err = NULL;
}

static void nbd_iter_request_error(NBDReplyChunkIter *iter, int ret)
{
    assert(ret < 0);

    if (!iter->request_ret) {
        iter->request_ret = ret;
    }
}

/*
 * NBD_FOREACH_REPLY_CHUNK
 * The pointer stored in @payload requires g_free() to free it.
 */
#define NBD_FOREACH_REPLY_CHUNK(s, iter, handle, structured, \
                                qiov, reply, payload) \
    for (iter = (NBDReplyChunkIter) { .only_structured = structured }; \
         nbd_reply_chunk_iter_receive(s, &iter, handle, qiov, reply, payload);)

/*
 * nbd_reply_chunk_iter_receive
 * The pointer stored in @payload requires g_free() to free it.
 */
static bool nbd_reply_chunk_iter_receive(BDRVNBDState *s,
                                         NBDReplyChunkIter *iter,
                                         uint64_t handle,
                                         QEMUIOVector *qiov, NBDReply *reply,
                                         void **payload)
{
    int ret, request_ret;
    NBDReply local_reply;
    NBDStructuredReplyChunk *chunk;
    Error *local_err = NULL;
    if (!nbd_client_connected(s)) {
        error_setg(&local_err, "Connection closed");
        nbd_iter_channel_error(iter, -EIO, &local_err);
        goto break_loop;
    }

    if (iter->done) {
        /* Previous iteration was last. */
        goto break_loop;
    }

    if (reply == NULL) {
        reply = &local_reply;
    }

    ret = nbd_co_receive_one_chunk(s, handle, iter->only_structured,
                                   &request_ret, qiov, reply, payload,
                                   &local_err);
    if (ret < 0) {
        nbd_iter_channel_error(iter, ret, &local_err);
    } else if (request_ret < 0) {
        nbd_iter_request_error(iter, request_ret);
    }

    /* Do not execute the body of NBD_FOREACH_REPLY_CHUNK for simple reply. */
    if (nbd_reply_is_simple(reply) || !nbd_client_connected(s)) {
        goto break_loop;
    }

    chunk = &reply->structured;
    iter->only_structured = true;

    if (chunk->type == NBD_REPLY_TYPE_NONE) {
        /* NBD_REPLY_FLAG_DONE is already checked in nbd_co_receive_one_chunk */
        assert(chunk->flags & NBD_REPLY_FLAG_DONE);
        goto break_loop;
    }

    if (chunk->flags & NBD_REPLY_FLAG_DONE) {
        /* This iteration is last. */
        iter->done = true;
    }

    /* Execute the loop body */
    return true;

break_loop:
    s->requests[HANDLE_TO_INDEX(s, handle)].coroutine = NULL;

    qemu_co_mutex_lock(&s->send_mutex);
    s->in_flight--;
    if (s->in_flight == 0 && s->wait_in_flight) {
        aio_co_wake(s->connection_co);
    } else {
        qemu_co_queue_next(&s->free_sema);
    }
    qemu_co_mutex_unlock(&s->send_mutex);

    return false;
}

static int nbd_co_receive_return_code(BDRVNBDState *s, uint64_t handle,
                                      int *request_ret, Error **errp)
{
    NBDReplyChunkIter iter;

    NBD_FOREACH_REPLY_CHUNK(s, iter, handle, false, NULL, NULL, NULL) {
        /* nbd_reply_chunk_iter_receive does all the work */
    }

    error_propagate(errp, iter.err);
    *request_ret = iter.request_ret;
    return iter.ret;
}

static int nbd_co_receive_cmdread_reply(BDRVNBDState *s, uint64_t handle,
                                        uint64_t offset, QEMUIOVector *qiov,
                                        int *request_ret, Error **errp)
{
    NBDReplyChunkIter iter;
    NBDReply reply;
    void *payload = NULL;
    Error *local_err = NULL;

    NBD_FOREACH_REPLY_CHUNK(s, iter, handle, s->info.structured_reply,
                            qiov, &reply, &payload)
    {
        int ret;
        NBDStructuredReplyChunk *chunk = &reply.structured;

        assert(nbd_reply_is_structured(&reply));

        switch (chunk->type) {
        case NBD_REPLY_TYPE_OFFSET_DATA:
            /*
             * special cased in nbd_co_receive_one_chunk, data is already
             * in qiov
             */
            break;
        case NBD_REPLY_TYPE_OFFSET_HOLE:
            ret = nbd_parse_offset_hole_payload(s, &reply.structured, payload,
                                                offset, qiov, &local_err);
            if (ret < 0) {
                nbd_channel_error(s, ret);
                nbd_iter_channel_error(&iter, ret, &local_err);
            }
            break;
        default:
            if (!nbd_reply_type_is_error(chunk->type)) {
                /* not allowed reply type */
                nbd_channel_error(s, -EINVAL);
                error_setg(&local_err,
                           "Unexpected reply type: %d (%s) for CMD_READ",
                           chunk->type, nbd_reply_type_lookup(chunk->type));
                nbd_iter_channel_error(&iter, -EINVAL, &local_err);
            }
        }

        g_free(payload);
        payload = NULL;
    }

    error_propagate(errp, iter.err);
    *request_ret = iter.request_ret;
    return iter.ret;
}

static int nbd_co_receive_blockstatus_reply(BDRVNBDState *s,
                                            uint64_t handle, uint64_t length,
                                            NBDExtent *extent,
                                            int *request_ret, Error **errp)
{
    NBDReplyChunkIter iter;
    NBDReply reply;
    void *payload = NULL;
    Error *local_err = NULL;
    bool received = false;

    assert(!extent->length);
    NBD_FOREACH_REPLY_CHUNK(s, iter, handle, false, NULL, &reply, &payload) {
        int ret;
        NBDStructuredReplyChunk *chunk = &reply.structured;

        assert(nbd_reply_is_structured(&reply));

        switch (chunk->type) {
        case NBD_REPLY_TYPE_BLOCK_STATUS:
            if (received) {
                nbd_channel_error(s, -EINVAL);
                error_setg(&local_err, "Several BLOCK_STATUS chunks in reply");
                nbd_iter_channel_error(&iter, -EINVAL, &local_err);
            }
            received = true;

            ret = nbd_parse_blockstatus_payload(s, &reply.structured,
                                                payload, length, extent,
                                                &local_err);
            if (ret < 0) {
                nbd_channel_error(s, ret);
                nbd_iter_channel_error(&iter, ret, &local_err);
            }
            break;
        default:
            if (!nbd_reply_type_is_error(chunk->type)) {
                nbd_channel_error(s, -EINVAL);
                error_setg(&local_err,
                           "Unexpected reply type: %d (%s) "
                           "for CMD_BLOCK_STATUS",
                           chunk->type, nbd_reply_type_lookup(chunk->type));
                nbd_iter_channel_error(&iter, -EINVAL, &local_err);
            }
        }

        g_free(payload);
        payload = NULL;
    }

    if (!extent->length && !iter.request_ret) {
        error_setg(&local_err, "Server did not reply with any status extents");
        nbd_iter_channel_error(&iter, -EIO, &local_err);
    }

    error_propagate(errp, iter.err);
    *request_ret = iter.request_ret;
    return iter.ret;
}

static int nbd_co_request(BlockDriverState *bs, NBDRequest *request,
                          QEMUIOVector *write_qiov)
{
    int ret, request_ret;
    Error *local_err = NULL;
    BDRVNBDState *s = (BDRVNBDState *)bs->opaque;

    assert(request->type != NBD_CMD_READ);
    if (write_qiov) {
        assert(request->type == NBD_CMD_WRITE);
        assert(request->len == iov_size(write_qiov->iov, write_qiov->niov));
    } else {
        assert(request->type != NBD_CMD_WRITE);
    }

    do {
        ret = nbd_co_send_request(bs, request, write_qiov);
        if (ret < 0) {
            continue;
        }

        ret = nbd_co_receive_return_code(s, request->handle,
                                         &request_ret, &local_err);
        if (local_err) {
            trace_nbd_co_request_fail(request->from, request->len,
                                      request->handle, request->flags,
                                      request->type,
                                      nbd_cmd_lookup(request->type),
                                      ret, error_get_pretty(local_err));
            error_free(local_err);
            local_err = NULL;
        }
    } while (ret < 0 && nbd_client_connecting_wait(s));

    return ret ? ret : request_ret;
}

static int nbd_client_co_preadv(BlockDriverState *bs, uint64_t offset,
                                uint64_t bytes, QEMUIOVector *qiov, int flags)
{
    int ret, request_ret;
    Error *local_err = NULL;
    BDRVNBDState *s = (BDRVNBDState *)bs->opaque;
    NBDRequest request = {
        .type = NBD_CMD_READ,
        .from = offset,
        .len = bytes,
    };

    assert(bytes <= NBD_MAX_BUFFER_SIZE);
    assert(!flags);

    if (!bytes) {
        return 0;
    }
    /*
     * Work around the fact that the block layer doesn't do
     * byte-accurate sizing yet - if the read exceeds the server's
     * advertised size because the block layer rounded size up, then
     * truncate the request to the server and tail-pad with zero.
     */
    if (offset >= s->info.size) {
        assert(bytes < BDRV_SECTOR_SIZE);
        qemu_iovec_memset(qiov, 0, 0, bytes);
        return 0;
    }
    if (offset + bytes > s->info.size) {
        uint64_t slop = offset + bytes - s->info.size;

        assert(slop < BDRV_SECTOR_SIZE);
        qemu_iovec_memset(qiov, bytes - slop, 0, slop);
        request.len -= slop;
    }

    do {
        ret = nbd_co_send_request(bs, &request, NULL);
        if (ret < 0) {
            continue;
        }

        ret = nbd_co_receive_cmdread_reply(s, request.handle, offset, qiov,
                                           &request_ret, &local_err);
        if (local_err) {
            trace_nbd_co_request_fail(request.from, request.len, request.handle,
                                      request.flags, request.type,
                                      nbd_cmd_lookup(request.type),
                                      ret, error_get_pretty(local_err));
            error_free(local_err);
            local_err = NULL;
        }
    } while (ret < 0 && nbd_client_connecting_wait(s));

    return ret ? ret : request_ret;
}

static int nbd_client_co_pwritev(BlockDriverState *bs, uint64_t offset,
                                 uint64_t bytes, QEMUIOVector *qiov, int flags)
{
    BDRVNBDState *s = (BDRVNBDState *)bs->opaque;
    NBDRequest request = {
        .type = NBD_CMD_WRITE,
        .from = offset,
        .len = bytes,
    };

    assert(!(s->info.flags & NBD_FLAG_READ_ONLY));
    if (flags & BDRV_REQ_FUA) {
        assert(s->info.flags & NBD_FLAG_SEND_FUA);
        request.flags |= NBD_CMD_FLAG_FUA;
    }

    assert(bytes <= NBD_MAX_BUFFER_SIZE);

    if (!bytes) {
        return 0;
    }
    return nbd_co_request(bs, &request, qiov);
}

static int nbd_client_co_pwrite_zeroes(BlockDriverState *bs, int64_t offset,
                                       int bytes, BdrvRequestFlags flags)
{
    BDRVNBDState *s = (BDRVNBDState *)bs->opaque;
    NBDRequest request = {
        .type = NBD_CMD_WRITE_ZEROES,
        .from = offset,
        .len = bytes,
    };

    assert(!(s->info.flags & NBD_FLAG_READ_ONLY));
    if (!(s->info.flags & NBD_FLAG_SEND_WRITE_ZEROES)) {
        return -ENOTSUP;
    }

    if (flags & BDRV_REQ_FUA) {
        assert(s->info.flags & NBD_FLAG_SEND_FUA);
        request.flags |= NBD_CMD_FLAG_FUA;
    }
    if (!(flags & BDRV_REQ_MAY_UNMAP)) {
        request.flags |= NBD_CMD_FLAG_NO_HOLE;
    }
    if (flags & BDRV_REQ_NO_FALLBACK) {
        assert(s->info.flags & NBD_FLAG_SEND_FAST_ZERO);
        request.flags |= NBD_CMD_FLAG_FAST_ZERO;
    }

    if (!bytes) {
        return 0;
    }
    return nbd_co_request(bs, &request, NULL);
}

static int nbd_client_co_flush(BlockDriverState *bs)
{
    BDRVNBDState *s = (BDRVNBDState *)bs->opaque;
    NBDRequest request = { .type = NBD_CMD_FLUSH };

    if (!(s->info.flags & NBD_FLAG_SEND_FLUSH)) {
        return 0;
    }

    request.from = 0;
    request.len = 0;

    return nbd_co_request(bs, &request, NULL);
}

static int nbd_client_co_pdiscard(BlockDriverState *bs, int64_t offset,
                                  int bytes)
{
    BDRVNBDState *s = (BDRVNBDState *)bs->opaque;
    NBDRequest request = {
        .type = NBD_CMD_TRIM,
        .from = offset,
        .len = bytes,
    };

    assert(!(s->info.flags & NBD_FLAG_READ_ONLY));
    if (!(s->info.flags & NBD_FLAG_SEND_TRIM) || !bytes) {
        return 0;
    }

    return nbd_co_request(bs, &request, NULL);
}

static int coroutine_fn nbd_client_co_block_status(
        BlockDriverState *bs, bool want_zero, int64_t offset, int64_t bytes,
        int64_t *pnum, int64_t *map, BlockDriverState **file)
{
    int ret, request_ret;
    NBDExtent extent = { 0 };
    BDRVNBDState *s = (BDRVNBDState *)bs->opaque;
    Error *local_err = NULL;

    NBDRequest request = {
        .type = NBD_CMD_BLOCK_STATUS,
        .from = offset,
        .len = MIN(QEMU_ALIGN_DOWN(INT_MAX, bs->bl.request_alignment),
                   MIN(bytes, s->info.size - offset)),
        .flags = NBD_CMD_FLAG_REQ_ONE,
    };

    if (!s->info.base_allocation) {
        *pnum = bytes;
        *map = offset;
        *file = bs;
        return BDRV_BLOCK_DATA | BDRV_BLOCK_OFFSET_VALID;
    }

    /*
     * Work around the fact that the block layer doesn't do
     * byte-accurate sizing yet - if the status request exceeds the
     * server's advertised size because the block layer rounded size
     * up, we truncated the request to the server (above), or are
     * called on just the hole.
     */
    if (offset >= s->info.size) {
        *pnum = bytes;
        assert(bytes < BDRV_SECTOR_SIZE);
        /* Intentionally don't report offset_valid for the hole */
        return BDRV_BLOCK_ZERO;
    }

    if (s->info.min_block) {
        assert(QEMU_IS_ALIGNED(request.len, s->info.min_block));
    }
    do {
        ret = nbd_co_send_request(bs, &request, NULL);
        if (ret < 0) {
            continue;
        }

        ret = nbd_co_receive_blockstatus_reply(s, request.handle, bytes,
                                               &extent, &request_ret,
                                               &local_err);
        if (local_err) {
            trace_nbd_co_request_fail(request.from, request.len, request.handle,
                                      request.flags, request.type,
                                      nbd_cmd_lookup(request.type),
                                      ret, error_get_pretty(local_err));
            error_free(local_err);
            local_err = NULL;
        }
    } while (ret < 0 && nbd_client_connecting_wait(s));

    if (ret < 0 || request_ret < 0) {
        return ret ? ret : request_ret;
    }

    assert(extent.length);
    *pnum = extent.length;
    *map = offset;
    *file = bs;
    return (extent.flags & NBD_STATE_HOLE ? 0 : BDRV_BLOCK_DATA) |
        (extent.flags & NBD_STATE_ZERO ? BDRV_BLOCK_ZERO : 0) |
        BDRV_BLOCK_OFFSET_VALID;
}

static int nbd_client_reopen_prepare(BDRVReopenState *state,
                                     BlockReopenQueue *queue, Error **errp)
{
    BDRVNBDState *s = (BDRVNBDState *)state->bs->opaque;

    if ((state->flags & BDRV_O_RDWR) && (s->info.flags & NBD_FLAG_READ_ONLY)) {
        error_setg(errp, "Can't reopen read-only NBD mount as read/write");
        return -EACCES;
    }
    return 0;
}

static void nbd_yank(void *opaque)
{
    BlockDriverState *bs = opaque;
    BDRVNBDState *s = (BDRVNBDState *)bs->opaque;

    qatomic_store_release(&s->state, NBD_CLIENT_QUIT);
    qio_channel_shutdown(QIO_CHANNEL(s->ioc), QIO_CHANNEL_SHUTDOWN_BOTH, NULL);
}

static void nbd_client_close(BlockDriverState *bs)
{
    BDRVNBDState *s = (BDRVNBDState *)bs->opaque;
    NBDRequest request = { .type = NBD_CMD_DISC };

    if (s->ioc) {
        nbd_send_request(s->ioc, &request);
    }

    nbd_teardown_connection(bs);
}


/*
 * Parse nbd_open options
 */

static int nbd_parse_uri(const char *filename, QDict *options)
{
    URI *uri;
    const char *p;
    QueryParams *qp = NULL;
    int ret = 0;
    bool is_unix;

    uri = uri_parse(filename);
    if (!uri) {
        return -EINVAL;
    }

    /* transport */
    if (!g_strcmp0(uri->scheme, "nbd")) {
        is_unix = false;
    } else if (!g_strcmp0(uri->scheme, "nbd+tcp")) {
        is_unix = false;
    } else if (!g_strcmp0(uri->scheme, "nbd+unix")) {
        is_unix = true;
    } else {
        ret = -EINVAL;
        goto out;
    }

    p = uri->path ? uri->path : "";
    if (p[0] == '/') {
        p++;
    }
    if (p[0]) {
        qdict_put_str(options, "export", p);
    }

    qp = query_params_parse(uri->query);
    if (qp->n > 1 || (is_unix && !qp->n) || (!is_unix && qp->n)) {
        ret = -EINVAL;
        goto out;
    }

    if (is_unix) {
        /* nbd+unix:///export?socket=path */
        if (uri->server || uri->port || strcmp(qp->p[0].name, "socket")) {
            ret = -EINVAL;
            goto out;
        }
        qdict_put_str(options, "server.type", "unix");
        qdict_put_str(options, "server.path", qp->p[0].value);
    } else {
        QString *host;
        char *port_str;

        /* nbd[+tcp]://host[:port]/export */
        if (!uri->server) {
            ret = -EINVAL;
            goto out;
        }

        /* strip braces from literal IPv6 address */
        if (uri->server[0] == '[') {
            host = qstring_from_substr(uri->server, 1,
                                       strlen(uri->server) - 1);
        } else {
            host = qstring_from_str(uri->server);
        }

        qdict_put_str(options, "server.type", "inet");
        qdict_put(options, "server.host", host);

        port_str = g_strdup_printf("%d", uri->port ?: NBD_DEFAULT_PORT);
        qdict_put_str(options, "server.port", port_str);
        g_free(port_str);
    }

out:
    if (qp) {
        query_params_free(qp);
    }
    uri_free(uri);
    return ret;
}

static bool nbd_has_filename_options_conflict(QDict *options, Error **errp)
{
    const QDictEntry *e;

    for (e = qdict_first(options); e; e = qdict_next(options, e)) {
        if (!strcmp(e->key, "host") ||
            !strcmp(e->key, "port") ||
            !strcmp(e->key, "path") ||
            !strcmp(e->key, "export") ||
            strstart(e->key, "server.", NULL))
        {
            error_setg(errp, "Option '%s' cannot be used with a file name",
                       e->key);
            return true;
        }
    }

    return false;
}

static void nbd_parse_filename(const char *filename, QDict *options,
                               Error **errp)
{
    g_autofree char *file = NULL;
    char *export_name;
    const char *host_spec;
    const char *unixpath;

    if (nbd_has_filename_options_conflict(options, errp)) {
        return;
    }

    if (strstr(filename, "://")) {
        int ret = nbd_parse_uri(filename, options);
        if (ret < 0) {
            error_setg(errp, "No valid URL specified");
        }
        return;
    }

    file = g_strdup(filename);

    export_name = strstr(file, EN_OPTSTR);
    if (export_name) {
        if (export_name[strlen(EN_OPTSTR)] == 0) {
            return;
        }
        export_name[0] = 0; /* truncate 'file' */
        export_name += strlen(EN_OPTSTR);

        qdict_put_str(options, "export", export_name);
    }

    /* extract the host_spec - fail if it's not nbd:... */
    if (!strstart(file, "nbd:", &host_spec)) {
        error_setg(errp, "File name string for NBD must start with 'nbd:'");
        return;
    }

    if (!*host_spec) {
        return;
    }

    /* are we a UNIX or TCP socket? */
    if (strstart(host_spec, "unix:", &unixpath)) {
        qdict_put_str(options, "server.type", "unix");
        qdict_put_str(options, "server.path", unixpath);
    } else {
        InetSocketAddress *addr = g_new(InetSocketAddress, 1);

        if (inet_parse(addr, host_spec, errp)) {
            goto out_inet;
        }

        qdict_put_str(options, "server.type", "inet");
        qdict_put_str(options, "server.host", addr->host);
        qdict_put_str(options, "server.port", addr->port);
    out_inet:
        qapi_free_InetSocketAddress(addr);
    }
}

static bool nbd_process_legacy_socket_options(QDict *output_options,
                                              QemuOpts *legacy_opts,
                                              Error **errp)
{
    const char *path = qemu_opt_get(legacy_opts, "path");
    const char *host = qemu_opt_get(legacy_opts, "host");
    const char *port = qemu_opt_get(legacy_opts, "port");
    const QDictEntry *e;

    if (!path && !host && !port) {
        return true;
    }

    for (e = qdict_first(output_options); e; e = qdict_next(output_options, e))
    {
        if (strstart(e->key, "server.", NULL)) {
            error_setg(errp, "Cannot use 'server' and path/host/port at the "
                       "same time");
            return false;
        }
    }

    if (path && host) {
        error_setg(errp, "path and host may not be used at the same time");
        return false;
    } else if (path) {
        if (port) {
            error_setg(errp, "port may not be used without host");
            return false;
        }

        qdict_put_str(output_options, "server.type", "unix");
        qdict_put_str(output_options, "server.path", path);
    } else if (host) {
        qdict_put_str(output_options, "server.type", "inet");
        qdict_put_str(output_options, "server.host", host);
        qdict_put_str(output_options, "server.port",
                      port ?: stringify(NBD_DEFAULT_PORT));
    }

    return true;
}

static SocketAddress *nbd_config(BDRVNBDState *s, QDict *options,
                                 Error **errp)
{
    SocketAddress *saddr = NULL;
    QDict *addr = NULL;
    Visitor *iv = NULL;

    qdict_extract_subqdict(options, &addr, "server.");
    if (!qdict_size(addr)) {
        error_setg(errp, "NBD server address missing");
        goto done;
    }

    iv = qobject_input_visitor_new_flat_confused(addr, errp);
    if (!iv) {
        goto done;
    }

    if (!visit_type_SocketAddress(iv, NULL, &saddr, errp)) {
        goto done;
    }

    if (socket_address_parse_named_fd(saddr, errp) < 0) {
        qapi_free_SocketAddress(saddr);
        saddr = NULL;
        goto done;
    }

done:
    qobject_unref(addr);
    visit_free(iv);
    return saddr;
}

static QCryptoTLSCreds *nbd_get_tls_creds(const char *id, Error **errp)
{
    Object *obj;
    QCryptoTLSCreds *creds;

    obj = object_resolve_path_component(
        object_get_objects_root(), id);
    if (!obj) {
        error_setg(errp, "No TLS credentials with id '%s'",
                   id);
        return NULL;
    }
    creds = (QCryptoTLSCreds *)
        object_dynamic_cast(obj, TYPE_QCRYPTO_TLS_CREDS);
    if (!creds) {
        error_setg(errp, "Object with id '%s' is not TLS credentials",
                   id);
        return NULL;
    }

    if (!qcrypto_tls_creds_check_endpoint(creds,
                                          QCRYPTO_TLS_CREDS_ENDPOINT_CLIENT,
                                          errp)) {
        return NULL;
    }
    object_ref(obj);
    return creds;
}


static QemuOptsList nbd_runtime_opts = {
    .name = "nbd",
    .head = QTAILQ_HEAD_INITIALIZER(nbd_runtime_opts.head),
    .desc = {
        {
            .name = "host",
            .type = QEMU_OPT_STRING,
            .help = "TCP host to connect to",
        },
        {
            .name = "port",
            .type = QEMU_OPT_STRING,
            .help = "TCP port to connect to",
        },
        {
            .name = "path",
            .type = QEMU_OPT_STRING,
            .help = "Unix socket path to connect to",
        },
        {
            .name = "export",
            .type = QEMU_OPT_STRING,
            .help = "Name of the NBD export to open",
        },
        {
            .name = "tls-creds",
            .type = QEMU_OPT_STRING,
            .help = "ID of the TLS credentials to use",
        },
        {
            .name = "x-dirty-bitmap",
            .type = QEMU_OPT_STRING,
            .help = "experimental: expose named dirty bitmap in place of "
                    "block status",
        },
        {
            .name = "reconnect-delay",
            .type = QEMU_OPT_NUMBER,
            .help = "On an unexpected disconnect, the nbd client tries to "
                    "connect again until succeeding or encountering a serious "
                    "error.  During the first @reconnect-delay seconds, all "
                    "requests are paused and will be rerun on a successful "
                    "reconnect. After that time, any delayed requests and all "
                    "future requests before a successful reconnect will "
                    "immediately fail. Default 0",
        },
        { /* end of list */ }
    },
};

static int nbd_process_options(BlockDriverState *bs, QDict *options,
                               Error **errp)
{
    BDRVNBDState *s = bs->opaque;
    QemuOpts *opts;
    int ret = -EINVAL;

    opts = qemu_opts_create(&nbd_runtime_opts, NULL, 0, &error_abort);
    if (!qemu_opts_absorb_qdict(opts, options, errp)) {
        goto error;
    }

    /* Translate @host, @port, and @path to a SocketAddress */
    if (!nbd_process_legacy_socket_options(options, opts, errp)) {
        goto error;
    }

    /* Pop the config into our state object. Exit if invalid. */
    s->saddr = nbd_config(s, options, errp);
    if (!s->saddr) {
        goto error;
    }

    s->export = g_strdup(qemu_opt_get(opts, "export"));
    if (s->export && strlen(s->export) > NBD_MAX_STRING_SIZE) {
        error_setg(errp, "export name too long to send to server");
        goto error;
    }

    s->tlscredsid = g_strdup(qemu_opt_get(opts, "tls-creds"));
    if (s->tlscredsid) {
        s->tlscreds = nbd_get_tls_creds(s->tlscredsid, errp);
        if (!s->tlscreds) {
            goto error;
        }

        /* TODO SOCKET_ADDRESS_KIND_FD where fd has AF_INET or AF_INET6 */
        if (s->saddr->type != SOCKET_ADDRESS_TYPE_INET) {
            error_setg(errp, "TLS only supported over IP sockets");
            goto error;
        }
        s->hostname = s->saddr->u.inet.host;
    }

    s->x_dirty_bitmap = g_strdup(qemu_opt_get(opts, "x-dirty-bitmap"));
    if (s->x_dirty_bitmap && strlen(s->x_dirty_bitmap) > NBD_MAX_STRING_SIZE) {
        error_setg(errp, "x-dirty-bitmap query too long to send to server");
        goto error;
    }

    s->reconnect_delay = qemu_opt_get_number(opts, "reconnect-delay", 0);

    ret = 0;

 error:
    qemu_opts_del(opts);
    return ret;
}

static int nbd_open(BlockDriverState *bs, QDict *options, int flags,
                    Error **errp)
{
    int ret;
    BDRVNBDState *s = (BDRVNBDState *)bs->opaque;

    s->bs = bs;
    qemu_co_mutex_init(&s->send_mutex);
    qemu_co_queue_init(&s->free_sema);

    if (!yank_register_instance(BLOCKDEV_YANK_INSTANCE(bs->node_name), errp)) {
        return -EEXIST;
    }

    ret = nbd_process_options(bs, options, errp);
    if (ret < 0) {
        goto fail;
    }

    s->conn = nbd_client_connection_new(s->saddr, true, s->export,
                                        s->x_dirty_bitmap, s->tlscreds);

    /* TODO: Configurable retry-until-timeout behaviour. */
    ret = nbd_do_establish_connection(bs, errp);
    if (ret < 0) {
        goto fail;
    }

    s->connection_co = qemu_coroutine_create(nbd_connection_entry, s);
    bdrv_inc_in_flight(bs);
    aio_co_schedule(bdrv_get_aio_context(bs), s->connection_co);

    return 0;

fail:
    nbd_clear_bdrvstate(bs);
    return ret;
}

static int nbd_co_flush(BlockDriverState *bs)
{
    return nbd_client_co_flush(bs);
}

static void nbd_refresh_limits(BlockDriverState *bs, Error **errp)
{
    BDRVNBDState *s = (BDRVNBDState *)bs->opaque;
    uint32_t min = s->info.min_block;
    uint32_t max = MIN_NON_ZERO(NBD_MAX_BUFFER_SIZE, s->info.max_block);

    /*
     * If the server did not advertise an alignment:
     * - a size that is not sector-aligned implies that an alignment
     *   of 1 can be used to access those tail bytes
     * - advertisement of block status requires an alignment of 1, so
     *   that we don't violate block layer constraints that block
     *   status is always aligned (as we can't control whether the
     *   server will report sub-sector extents, such as a hole at EOF
     *   on an unaligned POSIX file)
     * - otherwise, assume the server is so old that we are safer avoiding
     *   sub-sector requests
     */
    if (!min) {
        min = (!QEMU_IS_ALIGNED(s->info.size, BDRV_SECTOR_SIZE) ||
               s->info.base_allocation) ? 1 : BDRV_SECTOR_SIZE;
    }

    bs->bl.request_alignment = min;
    bs->bl.max_pdiscard = QEMU_ALIGN_DOWN(INT_MAX, min);
    bs->bl.max_pwrite_zeroes = max;
    bs->bl.max_transfer = max;

    if (s->info.opt_block &&
        s->info.opt_block > bs->bl.opt_transfer) {
        bs->bl.opt_transfer = s->info.opt_block;
    }
}

static void nbd_close(BlockDriverState *bs)
{
    nbd_client_close(bs);
    nbd_clear_bdrvstate(bs);
}

/*
 * NBD cannot truncate, but if the caller asks to truncate to the same size, or
 * to a smaller size with exact=false, there is no reason to fail the
 * operation.
 *
 * Preallocation mode is ignored since it does not seems useful to fail when
 * we never change anything.
 */
static int coroutine_fn nbd_co_truncate(BlockDriverState *bs, int64_t offset,
                                        bool exact, PreallocMode prealloc,
                                        BdrvRequestFlags flags, Error **errp)
{
    BDRVNBDState *s = bs->opaque;

    if (offset != s->info.size && exact) {
        error_setg(errp, "Cannot resize NBD nodes");
        return -ENOTSUP;
    }

    if (offset > s->info.size) {
        error_setg(errp, "Cannot grow NBD nodes");
        return -EINVAL;
    }

    return 0;
}

static int64_t nbd_getlength(BlockDriverState *bs)
{
    BDRVNBDState *s = bs->opaque;

    return s->info.size;
}

static void nbd_refresh_filename(BlockDriverState *bs)
{
    BDRVNBDState *s = bs->opaque;
    const char *host = NULL, *port = NULL, *path = NULL;
    size_t len = 0;

    if (s->saddr->type == SOCKET_ADDRESS_TYPE_INET) {
        const InetSocketAddress *inet = &s->saddr->u.inet;
        if (!inet->has_ipv4 && !inet->has_ipv6 && !inet->has_to) {
            host = inet->host;
            port = inet->port;
        }
    } else if (s->saddr->type == SOCKET_ADDRESS_TYPE_UNIX) {
        path = s->saddr->u.q_unix.path;
    } /* else can't represent as pseudo-filename */

    if (path && s->export) {
        len = snprintf(bs->exact_filename, sizeof(bs->exact_filename),
                       "nbd+unix:///%s?socket=%s", s->export, path);
    } else if (path && !s->export) {
        len = snprintf(bs->exact_filename, sizeof(bs->exact_filename),
                       "nbd+unix://?socket=%s", path);
    } else if (host && s->export) {
        len = snprintf(bs->exact_filename, sizeof(bs->exact_filename),
                       "nbd://%s:%s/%s", host, port, s->export);
    } else if (host && !s->export) {
        len = snprintf(bs->exact_filename, sizeof(bs->exact_filename),
                       "nbd://%s:%s", host, port);
    }
    if (len >= sizeof(bs->exact_filename)) {
        /* Name is too long to represent exactly, so leave it empty. */
        bs->exact_filename[0] = '\0';
    }
}

static char *nbd_dirname(BlockDriverState *bs, Error **errp)
{
    /* The generic bdrv_dirname() implementation is able to work out some
     * directory name for NBD nodes, but that would be wrong. So far there is no
     * specification for how "export paths" would work, so NBD does not have
     * directory names. */
    error_setg(errp, "Cannot generate a base directory for NBD nodes");
    return NULL;
}

static const char *const nbd_strong_runtime_opts[] = {
    "path",
    "host",
    "port",
    "export",
    "tls-creds",
    "server.",

    NULL
};

static void nbd_cancel_in_flight(BlockDriverState *bs)
{
    BDRVNBDState *s = (BDRVNBDState *)bs->opaque;

    reconnect_delay_timer_del(s);

    if (s->state == NBD_CLIENT_CONNECTING_WAIT) {
        s->state = NBD_CLIENT_CONNECTING_NOWAIT;
        qemu_co_queue_restart_all(&s->free_sema);
    }
}

static BlockDriver bdrv_nbd = {
    .format_name                = "nbd",
    .protocol_name              = "nbd",
    .instance_size              = sizeof(BDRVNBDState),
    .bdrv_parse_filename        = nbd_parse_filename,
    .bdrv_co_create_opts        = bdrv_co_create_opts_simple,
    .create_opts                = &bdrv_create_opts_simple,
    .bdrv_file_open             = nbd_open,
    .bdrv_reopen_prepare        = nbd_client_reopen_prepare,
    .bdrv_co_preadv             = nbd_client_co_preadv,
    .bdrv_co_pwritev            = nbd_client_co_pwritev,
    .bdrv_co_pwrite_zeroes      = nbd_client_co_pwrite_zeroes,
    .bdrv_close                 = nbd_close,
    .bdrv_co_flush_to_os        = nbd_co_flush,
    .bdrv_co_pdiscard           = nbd_client_co_pdiscard,
    .bdrv_refresh_limits        = nbd_refresh_limits,
    .bdrv_co_truncate           = nbd_co_truncate,
    .bdrv_getlength             = nbd_getlength,
    .bdrv_detach_aio_context    = nbd_client_detach_aio_context,
    .bdrv_attach_aio_context    = nbd_client_attach_aio_context,
    .bdrv_co_drain_begin        = nbd_client_co_drain_begin,
    .bdrv_co_drain_end          = nbd_client_co_drain_end,
    .bdrv_refresh_filename      = nbd_refresh_filename,
    .bdrv_co_block_status       = nbd_client_co_block_status,
    .bdrv_dirname               = nbd_dirname,
    .strong_runtime_opts        = nbd_strong_runtime_opts,
    .bdrv_cancel_in_flight      = nbd_cancel_in_flight,
};

static BlockDriver bdrv_nbd_tcp = {
    .format_name                = "nbd",
    .protocol_name              = "nbd+tcp",
    .instance_size              = sizeof(BDRVNBDState),
    .bdrv_parse_filename        = nbd_parse_filename,
    .bdrv_co_create_opts        = bdrv_co_create_opts_simple,
    .create_opts                = &bdrv_create_opts_simple,
    .bdrv_file_open             = nbd_open,
    .bdrv_reopen_prepare        = nbd_client_reopen_prepare,
    .bdrv_co_preadv             = nbd_client_co_preadv,
    .bdrv_co_pwritev            = nbd_client_co_pwritev,
    .bdrv_co_pwrite_zeroes      = nbd_client_co_pwrite_zeroes,
    .bdrv_close                 = nbd_close,
    .bdrv_co_flush_to_os        = nbd_co_flush,
    .bdrv_co_pdiscard           = nbd_client_co_pdiscard,
    .bdrv_refresh_limits        = nbd_refresh_limits,
    .bdrv_co_truncate           = nbd_co_truncate,
    .bdrv_getlength             = nbd_getlength,
    .bdrv_detach_aio_context    = nbd_client_detach_aio_context,
    .bdrv_attach_aio_context    = nbd_client_attach_aio_context,
    .bdrv_co_drain_begin        = nbd_client_co_drain_begin,
    .bdrv_co_drain_end          = nbd_client_co_drain_end,
    .bdrv_refresh_filename      = nbd_refresh_filename,
    .bdrv_co_block_status       = nbd_client_co_block_status,
    .bdrv_dirname               = nbd_dirname,
    .strong_runtime_opts        = nbd_strong_runtime_opts,
    .bdrv_cancel_in_flight      = nbd_cancel_in_flight,
};

static BlockDriver bdrv_nbd_unix = {
    .format_name                = "nbd",
    .protocol_name              = "nbd+unix",
    .instance_size              = sizeof(BDRVNBDState),
    .bdrv_parse_filename        = nbd_parse_filename,
    .bdrv_co_create_opts        = bdrv_co_create_opts_simple,
    .create_opts                = &bdrv_create_opts_simple,
    .bdrv_file_open             = nbd_open,
    .bdrv_reopen_prepare        = nbd_client_reopen_prepare,
    .bdrv_co_preadv             = nbd_client_co_preadv,
    .bdrv_co_pwritev            = nbd_client_co_pwritev,
    .bdrv_co_pwrite_zeroes      = nbd_client_co_pwrite_zeroes,
    .bdrv_close                 = nbd_close,
    .bdrv_co_flush_to_os        = nbd_co_flush,
    .bdrv_co_pdiscard           = nbd_client_co_pdiscard,
    .bdrv_refresh_limits        = nbd_refresh_limits,
    .bdrv_co_truncate           = nbd_co_truncate,
    .bdrv_getlength             = nbd_getlength,
    .bdrv_detach_aio_context    = nbd_client_detach_aio_context,
    .bdrv_attach_aio_context    = nbd_client_attach_aio_context,
    .bdrv_co_drain_begin        = nbd_client_co_drain_begin,
    .bdrv_co_drain_end          = nbd_client_co_drain_end,
    .bdrv_refresh_filename      = nbd_refresh_filename,
    .bdrv_co_block_status       = nbd_client_co_block_status,
    .bdrv_dirname               = nbd_dirname,
    .strong_runtime_opts        = nbd_strong_runtime_opts,
    .bdrv_cancel_in_flight      = nbd_cancel_in_flight,
};

static void bdrv_nbd_init(void)
{
    bdrv_register(&bdrv_nbd);
    bdrv_register(&bdrv_nbd_tcp);
    bdrv_register(&bdrv_nbd_unix);
}

block_init(bdrv_nbd_init);