Loading...
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 2244 2245 2246 2247 2248 2249 2250 2251 2252 2253 2254 2255 2256 2257 2258 2259 2260 2261 2262 2263 2264 2265 2266 2267 2268 2269 2270 2271 2272 2273 2274 2275 2276 2277 2278 2279 2280 2281 2282 2283 2284 2285 2286 2287 2288 2289 2290 2291 2292 2293 2294 2295 2296 2297 2298 2299 2300 2301 2302 2303 2304 2305 2306 2307 2308 2309 2310 2311 2312 2313 2314 2315 2316 2317 2318 2319 2320 2321 2322 2323 2324 2325 2326 2327 2328 2329 2330 2331 2332 2333 2334 2335 2336 2337 2338 2339 2340 2341 2342 2343 2344 2345 2346 2347 2348 2349 2350 2351 2352 2353 2354 2355 2356 2357 2358 2359 2360 2361 2362 2363 2364 2365 2366 2367 2368 2369 2370 2371 2372 2373 2374 2375 2376 2377 2378 2379 2380 2381 2382 2383 2384 2385 2386 2387 2388 2389 2390 2391 2392 2393 2394 2395 2396 2397 2398 2399 2400 2401 2402 2403 2404 2405 2406 2407 2408 2409 2410 2411 2412 2413 2414 2415 2416 2417 2418 2419 2420 2421 2422 2423 2424 2425 2426 2427 2428 2429 2430 2431 2432 2433 2434 2435 2436 2437 2438 2439 2440 2441 2442 2443 2444 2445 2446 2447 2448 2449 2450 2451 2452 2453 2454 2455 2456 2457 2458 2459 2460 2461 2462 2463 2464 2465 2466 2467 2468 2469 2470 2471 2472 2473 2474 2475 2476 2477 2478 2479 2480 2481 2482 2483 2484 2485 2486 2487 2488 2489 2490 2491 2492 2493 2494 2495 2496 2497 2498 2499 2500 2501 2502 2503 2504 2505 2506 2507 2508 2509 2510 2511 2512 2513 2514 2515 2516 2517 2518 2519 2520 2521 2522 2523 2524 2525 2526 2527 2528 2529 2530 2531 2532 2533 2534 2535 2536 2537 2538 2539 2540 2541 2542 2543 2544 2545 2546 2547 2548 2549 2550 2551 2552 2553 2554 2555 2556 2557 2558 2559 2560 2561 2562 2563 2564 2565 2566 2567 2568 2569 2570 2571 2572 2573 2574 2575 2576 2577 2578 2579 2580 2581 2582 2583 2584 2585 2586 2587 2588 2589 2590 2591 2592 2593 2594 2595 2596 2597 2598 2599 2600 2601 2602 2603 2604 2605 2606 2607 2608 2609 2610 2611 2612 2613 2614 2615 2616 2617 2618 2619 2620 2621 2622 2623 2624 2625 2626 2627 2628 2629 2630 2631 2632 2633 2634 2635 2636 2637 2638 2639 2640 2641 2642 2643 2644 2645 2646 2647 2648 2649 2650 2651 2652 2653 2654 2655 2656 2657 2658 2659 2660 2661 2662 2663 2664 2665 2666 2667 2668 2669 2670 2671 2672 2673 2674 2675 2676 2677 2678 2679 2680 2681 2682 2683 2684 2685 2686 2687 2688 2689 2690 2691 2692 2693 2694 2695 2696 2697 2698 2699 2700 2701 2702 | /* * Copyright (c) 2000-2017 Apple Inc. All rights reserved. * * @APPLE_OSREFERENCE_LICENSE_HEADER_START@ * * This file contains Original Code and/or Modifications of Original Code * as defined in and that are subject to the Apple Public Source License * Version 2.0 (the 'License'). You may not use this file except in * compliance with the License. The rights granted to you under the License * may not be used to create, or enable the creation or redistribution of, * unlawful or unlicensed copies of an Apple operating system, or to * circumvent, violate, or enable the circumvention or violation of, any * terms of an Apple operating system software license agreement. * * Please obtain a copy of the License at * http://www.opensource.apple.com/apsl/ and read it before using this file. * * The Original Code and all software distributed under the License are * distributed on an 'AS IS' basis, WITHOUT WARRANTY OF ANY KIND, EITHER * EXPRESS OR IMPLIED, AND APPLE HEREBY DISCLAIMS ALL SUCH WARRANTIES, * INCLUDING WITHOUT LIMITATION, ANY WARRANTIES OF MERCHANTABILITY, * FITNESS FOR A PARTICULAR PURPOSE, QUIET ENJOYMENT OR NON-INFRINGEMENT. * Please see the License for the specific language governing rights and * limitations under the License. * * @APPLE_OSREFERENCE_LICENSE_HEADER_END@ */ /* * Copyright (c) 1982, 1986, 1988, 1990, 1993, 1995 * The Regents of the University of California. All rights reserved. * * Redistribution and use in source and binary forms, with or without * modification, are permitted provided that the following conditions * are met: * 1. Redistributions of source code must retain the above copyright * notice, this list of conditions and the following disclaimer. * 2. Redistributions in binary form must reproduce the above copyright * notice, this list of conditions and the following disclaimer in the * documentation and/or other materials provided with the distribution. * 3. All advertising materials mentioning features or use of this software * must display the following acknowledgement: * This product includes software developed by the University of * California, Berkeley and its contributors. * 4. Neither the name of the University nor the names of its contributors * may be used to endorse or promote products derived from this software * without specific prior written permission. * * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF * SUCH DAMAGE. * * @(#)tcp_timer.c 8.2 (Berkeley) 5/24/95 * $FreeBSD: src/sys/netinet/tcp_timer.c,v 1.34.2.11 2001/08/22 00:59:12 silby Exp $ */ #include <sys/param.h> #include <sys/systm.h> #include <sys/kernel.h> #include <sys/mbuf.h> #include <sys/sysctl.h> #include <sys/socket.h> #include <sys/socketvar.h> #include <sys/protosw.h> #include <sys/domain.h> #include <sys/mcache.h> #include <sys/queue.h> #include <kern/locks.h> #include <kern/cpu_number.h> /* before tcp_seq.h, for tcp_random18() */ #include <mach/boolean.h> #include <net/route.h> #include <net/if_var.h> #include <net/ntstat.h> #include <netinet/in.h> #include <netinet/in_systm.h> #include <netinet/in_pcb.h> #if INET6 #include <netinet6/in6_pcb.h> #endif #include <netinet/ip_var.h> #include <netinet/tcp.h> #include <netinet/tcp_cache.h> #include <netinet/tcp_fsm.h> #include <netinet/tcp_seq.h> #include <netinet/tcp_timer.h> #include <netinet/tcp_var.h> #include <netinet/tcp_cc.h> #if INET6 #include <netinet6/tcp6_var.h> #endif #include <netinet/tcpip.h> #if TCPDEBUG #include <netinet/tcp_debug.h> #endif #include <sys/kdebug.h> #include <mach/sdt.h> #include <netinet/mptcp_var.h> /* Max number of times a stretch ack can be delayed on a connection */ #define TCP_STRETCHACK_DELAY_THRESHOLD 5 /* * If the host processor has been sleeping for too long, this is the threshold * used to avoid sending stale retransmissions. */ #define TCP_SLEEP_TOO_LONG (10 * 60 * 1000) /* 10 minutes in ms */ /* tcp timer list */ struct tcptimerlist tcp_timer_list; /* List of pcbs in timewait state, protected by tcbinfo's ipi_lock */ struct tcptailq tcp_tw_tailq; static int sysctl_msec_to_ticks SYSCTL_HANDLER_ARGS { #pragma unused(arg2) int error, s, tt; tt = *(int *)arg1; s = tt * 1000 / TCP_RETRANSHZ;; error = sysctl_handle_int(oidp, &s, 0, req); if (error || !req->newptr) return (error); tt = s * TCP_RETRANSHZ / 1000; if (tt < 1) return (EINVAL); *(int *)arg1 = tt; SYSCTL_SKMEM_UPDATE_AT_OFFSET(arg2, *(int*)arg1); return (0); } #if SYSCTL_SKMEM int tcp_keepinit = TCPTV_KEEP_INIT; SYSCTL_PROC(_net_inet_tcp, TCPCTL_KEEPINIT, keepinit, CTLTYPE_INT | CTLFLAG_RW | CTLFLAG_LOCKED, &tcp_keepinit, offsetof(skmem_sysctl, tcp.keepinit), sysctl_msec_to_ticks, "I", ""); int tcp_keepidle = TCPTV_KEEP_IDLE; SYSCTL_PROC(_net_inet_tcp, TCPCTL_KEEPIDLE, keepidle, CTLTYPE_INT | CTLFLAG_RW | CTLFLAG_LOCKED, &tcp_keepidle, offsetof(skmem_sysctl, tcp.keepidle), sysctl_msec_to_ticks, "I", ""); int tcp_keepintvl = TCPTV_KEEPINTVL; SYSCTL_PROC(_net_inet_tcp, TCPCTL_KEEPINTVL, keepintvl, CTLTYPE_INT | CTLFLAG_RW | CTLFLAG_LOCKED, &tcp_keepintvl, offsetof(skmem_sysctl, tcp.keepintvl), sysctl_msec_to_ticks, "I", ""); SYSCTL_SKMEM_TCP_INT(OID_AUTO, keepcnt, CTLTYPE_INT | CTLFLAG_RW | CTLFLAG_LOCKED, int, tcp_keepcnt, TCPTV_KEEPCNT, "number of times to repeat keepalive"); int tcp_msl = TCPTV_MSL; SYSCTL_PROC(_net_inet_tcp, OID_AUTO, msl, CTLTYPE_INT | CTLFLAG_RW | CTLFLAG_LOCKED, &tcp_msl, offsetof(skmem_sysctl, tcp.msl), sysctl_msec_to_ticks, "I", "Maximum segment lifetime"); #else /* SYSCTL_SKMEM */ int tcp_keepinit; SYSCTL_PROC(_net_inet_tcp, TCPCTL_KEEPINIT, keepinit, CTLTYPE_INT | CTLFLAG_RW | CTLFLAG_LOCKED, &tcp_keepinit, 0, sysctl_msec_to_ticks, "I", ""); int tcp_keepidle; SYSCTL_PROC(_net_inet_tcp, TCPCTL_KEEPIDLE, keepidle, CTLTYPE_INT | CTLFLAG_RW | CTLFLAG_LOCKED, &tcp_keepidle, 0, sysctl_msec_to_ticks, "I", ""); int tcp_keepintvl; SYSCTL_PROC(_net_inet_tcp, TCPCTL_KEEPINTVL, keepintvl, CTLTYPE_INT | CTLFLAG_RW | CTLFLAG_LOCKED, &tcp_keepintvl, 0, sysctl_msec_to_ticks, "I", ""); int tcp_keepcnt; SYSCTL_INT(_net_inet_tcp, OID_AUTO, keepcnt, CTLTYPE_INT | CTLFLAG_RW | CTLFLAG_LOCKED, &tcp_keepcnt, 0, "number of times to repeat keepalive"); int tcp_msl; SYSCTL_PROC(_net_inet_tcp, OID_AUTO, msl, CTLTYPE_INT | CTLFLAG_RW | CTLFLAG_LOCKED, &tcp_msl, 0, sysctl_msec_to_ticks, "I", "Maximum segment lifetime"); #endif /* SYSCTL_SKMEM */ /* * Avoid DoS via TCP Robustness in Persist Condition * (see http://www.ietf.org/id/draft-ananth-tcpm-persist-02.txt) * by allowing a system wide maximum persistence timeout value when in * Zero Window Probe mode. * * Expressed in milliseconds to be consistent without timeout related * values, the TCP socket option is in seconds. */ #if SYSCTL_SKMEM u_int32_t tcp_max_persist_timeout = 0; SYSCTL_PROC(_net_inet_tcp, OID_AUTO, max_persist_timeout, CTLTYPE_INT | CTLFLAG_RW | CTLFLAG_LOCKED, &tcp_max_persist_timeout, offsetof(skmem_sysctl, tcp.max_persist_timeout), sysctl_msec_to_ticks, "I", "Maximum persistence timeout for ZWP"); #else /* SYSCTL_SKMEM */ u_int32_t tcp_max_persist_timeout = 0; SYSCTL_PROC(_net_inet_tcp, OID_AUTO, max_persist_timeout, CTLTYPE_INT | CTLFLAG_RW | CTLFLAG_LOCKED, &tcp_max_persist_timeout, 0, sysctl_msec_to_ticks, "I", "Maximum persistence timeout for ZWP"); #endif /* SYSCTL_SKMEM */ SYSCTL_SKMEM_TCP_INT(OID_AUTO, always_keepalive, CTLFLAG_RW | CTLFLAG_LOCKED, static int, always_keepalive, 0, "Assume SO_KEEPALIVE on all TCP connections"); /* * This parameter determines how long the timer list will stay in fast or * quick mode even though all connections are idle. In this state, the * timer will run more frequently anticipating new data. */ SYSCTL_SKMEM_TCP_INT(OID_AUTO, timer_fastmode_idlemax, CTLFLAG_RW | CTLFLAG_LOCKED, int, timer_fastmode_idlemax, TCP_FASTMODE_IDLERUN_MAX, "Maximum idle generations in fast mode"); /* * See tcp_syn_backoff[] for interval values between SYN retransmits; * the value set below defines the number of retransmits, before we * disable the timestamp and window scaling options during subsequent * SYN retransmits. Setting it to 0 disables the dropping off of those * two options. */ SYSCTL_SKMEM_TCP_INT(OID_AUTO, broken_peer_syn_rexmit_thres, CTLFLAG_RW | CTLFLAG_LOCKED, static int, tcp_broken_peer_syn_rxmit_thres, 10, "Number of retransmitted SYNs before disabling RFC 1323 " "options on local connections"); static int tcp_timer_advanced = 0; SYSCTL_INT(_net_inet_tcp, OID_AUTO, tcp_timer_advanced, CTLFLAG_RD | CTLFLAG_LOCKED, &tcp_timer_advanced, 0, "Number of times one of the timers was advanced"); static int tcp_resched_timerlist = 0; SYSCTL_INT(_net_inet_tcp, OID_AUTO, tcp_resched_timerlist, CTLFLAG_RD | CTLFLAG_LOCKED, &tcp_resched_timerlist, 0, "Number of times timer list was rescheduled as part of processing a packet"); SYSCTL_SKMEM_TCP_INT(OID_AUTO, pmtud_blackhole_detection, CTLFLAG_RW | CTLFLAG_LOCKED, int, tcp_pmtud_black_hole_detect, 1, "Path MTU Discovery Black Hole Detection"); SYSCTL_SKMEM_TCP_INT(OID_AUTO, pmtud_blackhole_mss, CTLFLAG_RW | CTLFLAG_LOCKED, int, tcp_pmtud_black_hole_mss, 1200, "Path MTU Discovery Black Hole Detection lowered MSS"); static u_int32_t tcp_mss_rec_medium = 1200; static u_int32_t tcp_mss_rec_low = 512; #define TCP_REPORT_STATS_INTERVAL 43200 /* 12 hours, in seconds */ int tcp_report_stats_interval = TCP_REPORT_STATS_INTERVAL; /* performed garbage collection of "used" sockets */ static boolean_t tcp_gc_done = FALSE; /* max idle probes */ int tcp_maxpersistidle = TCPTV_KEEP_IDLE; /* * TCP delack timer is set to 100 ms. Since the processing of timer list * in fast mode will happen no faster than 100 ms, the delayed ack timer * will fire some where between 100 and 200 ms. */ int tcp_delack = TCP_RETRANSHZ / 10; #if MPTCP /* * MP_JOIN retransmission of 3rd ACK will be every 500 msecs without backoff */ int tcp_jack_rxmt = TCP_RETRANSHZ / 2; #endif /* MPTCP */ static boolean_t tcp_itimer_done = FALSE; static void tcp_remove_timer(struct tcpcb *tp); static void tcp_sched_timerlist(uint32_t offset); static u_int32_t tcp_run_conn_timer(struct tcpcb *tp, u_int16_t *mode, u_int16_t probe_if_index); static void tcp_sched_timers(struct tcpcb *tp); static inline void tcp_set_lotimer_index(struct tcpcb *); __private_extern__ void tcp_remove_from_time_wait(struct inpcb *inp); static inline void tcp_update_mss_core(struct tcpcb *tp, struct ifnet *ifp); __private_extern__ void tcp_report_stats(void); static u_int64_t tcp_last_report_time; /* * Structure to store previously reported stats so that we can send * incremental changes in each report interval. */ struct tcp_last_report_stats { u_int32_t tcps_connattempt; u_int32_t tcps_accepts; u_int32_t tcps_ecn_client_setup; u_int32_t tcps_ecn_server_setup; u_int32_t tcps_ecn_client_success; u_int32_t tcps_ecn_server_success; u_int32_t tcps_ecn_not_supported; u_int32_t tcps_ecn_lost_syn; u_int32_t tcps_ecn_lost_synack; u_int32_t tcps_ecn_recv_ce; u_int32_t tcps_ecn_recv_ece; u_int32_t tcps_ecn_sent_ece; u_int32_t tcps_ecn_conn_recv_ce; u_int32_t tcps_ecn_conn_recv_ece; u_int32_t tcps_ecn_conn_plnoce; u_int32_t tcps_ecn_conn_pl_ce; u_int32_t tcps_ecn_conn_nopl_ce; u_int32_t tcps_ecn_fallback_synloss; u_int32_t tcps_ecn_fallback_reorder; u_int32_t tcps_ecn_fallback_ce; /* TFO-related statistics */ u_int32_t tcps_tfo_syn_data_rcv; u_int32_t tcps_tfo_cookie_req_rcv; u_int32_t tcps_tfo_cookie_sent; u_int32_t tcps_tfo_cookie_invalid; u_int32_t tcps_tfo_cookie_req; u_int32_t tcps_tfo_cookie_rcv; u_int32_t tcps_tfo_syn_data_sent; u_int32_t tcps_tfo_syn_data_acked; u_int32_t tcps_tfo_syn_loss; u_int32_t tcps_tfo_blackhole; u_int32_t tcps_tfo_cookie_wrong; u_int32_t tcps_tfo_no_cookie_rcv; u_int32_t tcps_tfo_heuristics_disable; u_int32_t tcps_tfo_sndblackhole; /* MPTCP-related statistics */ u_int32_t tcps_mptcp_handover_attempt; u_int32_t tcps_mptcp_interactive_attempt; u_int32_t tcps_mptcp_aggregate_attempt; u_int32_t tcps_mptcp_fp_handover_attempt; u_int32_t tcps_mptcp_fp_interactive_attempt; u_int32_t tcps_mptcp_fp_aggregate_attempt; u_int32_t tcps_mptcp_heuristic_fallback; u_int32_t tcps_mptcp_fp_heuristic_fallback; u_int32_t tcps_mptcp_handover_success_wifi; u_int32_t tcps_mptcp_handover_success_cell; u_int32_t tcps_mptcp_interactive_success; u_int32_t tcps_mptcp_aggregate_success; u_int32_t tcps_mptcp_fp_handover_success_wifi; u_int32_t tcps_mptcp_fp_handover_success_cell; u_int32_t tcps_mptcp_fp_interactive_success; u_int32_t tcps_mptcp_fp_aggregate_success; u_int32_t tcps_mptcp_handover_cell_from_wifi; u_int32_t tcps_mptcp_handover_wifi_from_cell; u_int32_t tcps_mptcp_interactive_cell_from_wifi; u_int64_t tcps_mptcp_handover_cell_bytes; u_int64_t tcps_mptcp_interactive_cell_bytes; u_int64_t tcps_mptcp_aggregate_cell_bytes; u_int64_t tcps_mptcp_handover_all_bytes; u_int64_t tcps_mptcp_interactive_all_bytes; u_int64_t tcps_mptcp_aggregate_all_bytes; u_int32_t tcps_mptcp_back_to_wifi; u_int32_t tcps_mptcp_wifi_proxy; u_int32_t tcps_mptcp_cell_proxy; u_int32_t tcps_mptcp_triggered_cell; }; /* Returns true if the timer is on the timer list */ #define TIMER_IS_ON_LIST(tp) ((tp)->t_flags & TF_TIMER_ONLIST) /* Run the TCP timerlist atleast once every hour */ #define TCP_TIMERLIST_MAX_OFFSET (60 * 60 * TCP_RETRANSHZ) static void add_to_time_wait_locked(struct tcpcb *tp, uint32_t delay); static boolean_t tcp_garbage_collect(struct inpcb *, int); #define TIMERENTRY_TO_TP(te) ((struct tcpcb *)((uintptr_t)te - offsetof(struct tcpcb, tentry.le.le_next))) #define VERIFY_NEXT_LINK(elm,field) do { \ if (LIST_NEXT((elm),field) != NULL && \ LIST_NEXT((elm),field)->field.le_prev != \ &((elm)->field.le_next)) \ panic("Bad link elm %p next->prev != elm", (elm)); \ } while(0) #define VERIFY_PREV_LINK(elm,field) do { \ if (*(elm)->field.le_prev != (elm)) \ panic("Bad link elm %p prev->next != elm", (elm)); \ } while(0) #define TCP_SET_TIMER_MODE(mode, i) do { \ if (IS_TIMER_HZ_10MS(i)) \ (mode) |= TCP_TIMERLIST_10MS_MODE; \ else if (IS_TIMER_HZ_100MS(i)) \ (mode) |= TCP_TIMERLIST_100MS_MODE; \ else \ (mode) |= TCP_TIMERLIST_500MS_MODE; \ } while(0) #if (DEVELOPMENT || DEBUG) SYSCTL_UINT(_net_inet_tcp, OID_AUTO, mss_rec_medium, CTLFLAG_RW | CTLFLAG_LOCKED, &tcp_mss_rec_medium, 0, "Medium MSS based on recommendation in link status report"); SYSCTL_UINT(_net_inet_tcp, OID_AUTO, mss_rec_low, CTLFLAG_RW | CTLFLAG_LOCKED, &tcp_mss_rec_low, 0, "Low MSS based on recommendation in link status report"); static int32_t tcp_change_mss_recommended = 0; static int sysctl_change_mss_recommended SYSCTL_HANDLER_ARGS { #pragma unused(oidp, arg1, arg2) int i, err = 0, changed = 0; struct ifnet *ifp; struct if_link_status ifsr; struct if_cellular_status_v1 *new_cell_sr; err = sysctl_io_number(req, tcp_change_mss_recommended, sizeof (int32_t), &i, &changed); if (changed) { ifnet_head_lock_shared(); TAILQ_FOREACH(ifp, &ifnet_head, if_link) { if (IFNET_IS_CELLULAR(ifp)) { bzero(&ifsr, sizeof (ifsr)); new_cell_sr = &ifsr.ifsr_u.ifsr_cell.if_cell_u.if_status_v1; ifsr.ifsr_version = IF_CELLULAR_STATUS_REPORT_CURRENT_VERSION; ifsr.ifsr_len = sizeof(*new_cell_sr); /* Set MSS recommended */ new_cell_sr->valid_bitmask |= IF_CELL_UL_MSS_RECOMMENDED_VALID; new_cell_sr->mss_recommended = i; err = ifnet_link_status_report(ifp, new_cell_sr, sizeof (new_cell_sr)); if (err == 0) { tcp_change_mss_recommended = i; } else { break; } } } ifnet_head_done(); } return (err); } SYSCTL_PROC(_net_inet_tcp, OID_AUTO, change_mss_recommended, CTLTYPE_INT | CTLFLAG_RW | CTLFLAG_LOCKED, &tcp_change_mss_recommended, 0, sysctl_change_mss_recommended, "IU", "Change MSS recommended"); SYSCTL_INT(_net_inet_tcp, OID_AUTO, report_stats_interval, CTLFLAG_RW | CTLFLAG_LOCKED, &tcp_report_stats_interval, 0, "Report stats interval"); #endif /* (DEVELOPMENT || DEBUG) */ /* * Macro to compare two timers. If there is a reset of the sign bit, * it is safe to assume that the timer has wrapped around. By doing * signed comparision, we take care of wrap around such that the value * with the sign bit reset is actually ahead of the other. */ inline int32_t timer_diff(uint32_t t1, uint32_t toff1, uint32_t t2, uint32_t toff2) { return (int32_t)((t1 + toff1) - (t2 + toff2)); }; /* * Add to tcp timewait list, delay is given in milliseconds. */ static void add_to_time_wait_locked(struct tcpcb *tp, uint32_t delay) { struct inpcbinfo *pcbinfo = &tcbinfo; struct inpcb *inp = tp->t_inpcb; uint32_t timer; /* pcb list should be locked when we get here */ LCK_RW_ASSERT(pcbinfo->ipi_lock, LCK_RW_ASSERT_EXCLUSIVE); /* We may get here multiple times, so check */ if (!(inp->inp_flags2 & INP2_TIMEWAIT)) { pcbinfo->ipi_twcount++; inp->inp_flags2 |= INP2_TIMEWAIT; /* Remove from global inp list */ LIST_REMOVE(inp, inp_list); } else { TAILQ_REMOVE(&tcp_tw_tailq, tp, t_twentry); } /* Compute the time at which this socket can be closed */ timer = tcp_now + delay; /* We will use the TCPT_2MSL timer for tracking this delay */ if (TIMER_IS_ON_LIST(tp)) tcp_remove_timer(tp); tp->t_timer[TCPT_2MSL] = timer; TAILQ_INSERT_TAIL(&tcp_tw_tailq, tp, t_twentry); } void add_to_time_wait(struct tcpcb *tp, uint32_t delay) { struct inpcbinfo *pcbinfo = &tcbinfo; if (tp->t_inpcb->inp_socket->so_options & SO_NOWAKEFROMSLEEP) socket_post_kev_msg_closed(tp->t_inpcb->inp_socket); /* 19182803: Notify nstat that connection is closing before waiting. */ nstat_pcb_detach(tp->t_inpcb); if (!lck_rw_try_lock_exclusive(pcbinfo->ipi_lock)) { socket_unlock(tp->t_inpcb->inp_socket, 0); lck_rw_lock_exclusive(pcbinfo->ipi_lock); socket_lock(tp->t_inpcb->inp_socket, 0); } add_to_time_wait_locked(tp, delay); lck_rw_done(pcbinfo->ipi_lock); inpcb_gc_sched(pcbinfo, INPCB_TIMER_LAZY); } /* If this is on time wait queue, remove it. */ void tcp_remove_from_time_wait(struct inpcb *inp) { struct tcpcb *tp = intotcpcb(inp); if (inp->inp_flags2 & INP2_TIMEWAIT) TAILQ_REMOVE(&tcp_tw_tailq, tp, t_twentry); } static boolean_t tcp_garbage_collect(struct inpcb *inp, int istimewait) { boolean_t active = FALSE; struct socket *so, *mp_so = NULL; struct tcpcb *tp; so = inp->inp_socket; tp = intotcpcb(inp); if (so->so_flags & SOF_MP_SUBFLOW) { mp_so = mptetoso(tptomptp(tp)->mpt_mpte); if (!socket_try_lock(mp_so)) { mp_so = NULL; active = TRUE; goto out; } mp_so->so_usecount++; } /* * Skip if still in use or busy; it would have been more efficient * if we were to test so_usecount against 0, but this isn't possible * due to the current implementation of tcp_dropdropablreq() where * overflow sockets that are eligible for garbage collection have * their usecounts set to 1. */ if (!lck_mtx_try_lock_spin(&inp->inpcb_mtx)) { active = TRUE; goto out; } /* Check again under the lock */ if (so->so_usecount > 1) { if (inp->inp_wantcnt == WNT_STOPUSING) active = TRUE; lck_mtx_unlock(&inp->inpcb_mtx); goto out; } if (istimewait && TSTMP_GEQ(tcp_now, tp->t_timer[TCPT_2MSL]) && tp->t_state != TCPS_CLOSED) { /* Become a regular mutex */ lck_mtx_convert_spin(&inp->inpcb_mtx); tcp_close(tp); } /* * Overflowed socket dropped from the listening queue? Do this * only if we are called to clean up the time wait slots, since * tcp_dropdropablreq() considers a socket to have been fully * dropped after add_to_time_wait() is finished. * Also handle the case of connections getting closed by the peer * while in the queue as seen with rdar://6422317 * */ if (so->so_usecount == 1 && ((istimewait && (so->so_flags & SOF_OVERFLOW)) || ((tp != NULL) && (tp->t_state == TCPS_CLOSED) && (so->so_head != NULL) && ((so->so_state & (SS_INCOMP|SS_CANTSENDMORE|SS_CANTRCVMORE)) == (SS_INCOMP|SS_CANTSENDMORE|SS_CANTRCVMORE))))) { if (inp->inp_state != INPCB_STATE_DEAD) { /* Become a regular mutex */ lck_mtx_convert_spin(&inp->inpcb_mtx); #if INET6 if (SOCK_CHECK_DOM(so, PF_INET6)) in6_pcbdetach(inp); else #endif /* INET6 */ in_pcbdetach(inp); } VERIFY(so->so_usecount > 0); so->so_usecount--; if (inp->inp_wantcnt == WNT_STOPUSING) active = TRUE; lck_mtx_unlock(&inp->inpcb_mtx); goto out; } else if (inp->inp_wantcnt != WNT_STOPUSING) { lck_mtx_unlock(&inp->inpcb_mtx); active = FALSE; goto out; } /* * We get here because the PCB is no longer searchable * (WNT_STOPUSING); detach (if needed) and dispose if it is dead * (usecount is 0). This covers all cases, including overflow * sockets and those that are considered as "embryonic", * i.e. created by sonewconn() in TCP input path, and have * not yet been committed. For the former, we reduce the usecount * to 0 as done by the code above. For the latter, the usecount * would have reduced to 0 as part calling soabort() when the * socket is dropped at the end of tcp_input(). */ if (so->so_usecount == 0) { DTRACE_TCP4(state__change, void, NULL, struct inpcb *, inp, struct tcpcb *, tp, int32_t, TCPS_CLOSED); /* Become a regular mutex */ lck_mtx_convert_spin(&inp->inpcb_mtx); /* * If this tp still happens to be on the timer list, * take it out */ if (TIMER_IS_ON_LIST(tp)) { tcp_remove_timer(tp); } if (inp->inp_state != INPCB_STATE_DEAD) { #if INET6 if (SOCK_CHECK_DOM(so, PF_INET6)) in6_pcbdetach(inp); else #endif /* INET6 */ in_pcbdetach(inp); } if (mp_so) { mptcp_subflow_del(tptomptp(tp)->mpt_mpte, tp->t_mpsub); /* so is now unlinked from mp_so - let's drop the lock */ socket_unlock(mp_so, 1); mp_so = NULL; } in_pcbdispose(inp); active = FALSE; goto out; } lck_mtx_unlock(&inp->inpcb_mtx); active = TRUE; out: if (mp_so) socket_unlock(mp_so, 1); return (active); } /* * TCP garbage collector callback (inpcb_timer_func_t). * * Returns the number of pcbs that will need to be gc-ed soon, * returnining > 0 will keep timer active. */ void tcp_gc(struct inpcbinfo *ipi) { struct inpcb *inp, *nxt; struct tcpcb *tw_tp, *tw_ntp; #if TCPDEBUG int ostate; #endif #if KDEBUG static int tws_checked = 0; #endif KERNEL_DEBUG(DBG_FNC_TCP_SLOW | DBG_FUNC_START, 0, 0, 0, 0, 0); /* * Update tcp_now here as it may get used while * processing the slow timer. */ calculate_tcp_clock(); /* * Garbage collect socket/tcpcb: We need to acquire the list lock * exclusively to do this */ if (lck_rw_try_lock_exclusive(ipi->ipi_lock) == FALSE) { /* don't sweat it this time; cleanup was done last time */ if (tcp_gc_done == TRUE) { tcp_gc_done = FALSE; KERNEL_DEBUG(DBG_FNC_TCP_SLOW | DBG_FUNC_END, tws_checked, cur_tw_slot, 0, 0, 0); /* Lock upgrade failed, give up this round */ atomic_add_32(&ipi->ipi_gc_req.intimer_fast, 1); return; } /* Upgrade failed, lost lock now take it again exclusive */ lck_rw_lock_exclusive(ipi->ipi_lock); } tcp_gc_done = TRUE; LIST_FOREACH_SAFE(inp, &tcb, inp_list, nxt) { if (tcp_garbage_collect(inp, 0)) atomic_add_32(&ipi->ipi_gc_req.intimer_fast, 1); } /* Now cleanup the time wait ones */ TAILQ_FOREACH_SAFE(tw_tp, &tcp_tw_tailq, t_twentry, tw_ntp) { /* * We check the timestamp here without holding the * socket lock for better performance. If there are * any pcbs in time-wait, the timer will get rescheduled. * Hence some error in this check can be tolerated. * * Sometimes a socket on time-wait queue can be closed if * 2MSL timer expired but the application still has a * usecount on it. */ if (tw_tp->t_state == TCPS_CLOSED || TSTMP_GEQ(tcp_now, tw_tp->t_timer[TCPT_2MSL])) { if (tcp_garbage_collect(tw_tp->t_inpcb, 1)) atomic_add_32(&ipi->ipi_gc_req.intimer_lazy, 1); } } /* take into account pcbs that are still in time_wait_slots */ atomic_add_32(&ipi->ipi_gc_req.intimer_lazy, ipi->ipi_twcount); lck_rw_done(ipi->ipi_lock); /* Clean up the socache while we are here */ if (so_cache_timer()) atomic_add_32(&ipi->ipi_gc_req.intimer_lazy, 1); KERNEL_DEBUG(DBG_FNC_TCP_SLOW | DBG_FUNC_END, tws_checked, cur_tw_slot, 0, 0, 0); return; } /* * Cancel all timers for TCP tp. */ void tcp_canceltimers(struct tcpcb *tp) { int i; tcp_remove_timer(tp); for (i = 0; i < TCPT_NTIMERS; i++) tp->t_timer[i] = 0; tp->tentry.timer_start = tcp_now; tp->tentry.index = TCPT_NONE; } int tcp_syn_backoff[TCP_MAXRXTSHIFT + 1] = { 1, 1, 1, 1, 1, 2, 4, 8, 16, 32, 64, 64, 64 }; int tcp_backoff[TCP_MAXRXTSHIFT + 1] = { 1, 2, 4, 8, 16, 32, 64, 64, 64, 64, 64, 64, 64 }; static int tcp_totbackoff = 511; /* sum of tcp_backoff[] */ void tcp_rexmt_save_state(struct tcpcb *tp) { u_int32_t fsize; if (TSTMP_SUPPORTED(tp)) { /* * Since timestamps are supported on the connection, * we can do recovery as described in rfc 4015. */ fsize = tp->snd_max - tp->snd_una; tp->snd_ssthresh_prev = max(fsize, tp->snd_ssthresh); tp->snd_recover_prev = tp->snd_recover; } else { /* * Timestamp option is not supported on this connection. * Record ssthresh and cwnd so they can * be recovered if this turns out to be a "bad" retransmit. * A retransmit is considered "bad" if an ACK for this * segment is received within RTT/2 interval; the assumption * here is that the ACK was already in flight. See * "On Estimating End-to-End Network Path Properties" by * Allman and Paxson for more details. */ tp->snd_cwnd_prev = tp->snd_cwnd; tp->snd_ssthresh_prev = tp->snd_ssthresh; tp->snd_recover_prev = tp->snd_recover; if (IN_FASTRECOVERY(tp)) tp->t_flags |= TF_WASFRECOVERY; else tp->t_flags &= ~TF_WASFRECOVERY; } tp->t_srtt_prev = (tp->t_srtt >> TCP_RTT_SHIFT) + 2; tp->t_rttvar_prev = (tp->t_rttvar >> TCP_RTTVAR_SHIFT); tp->t_flagsext &= ~(TF_RECOMPUTE_RTT); } /* * Revert to the older segment size if there is an indication that PMTU * blackhole detection was not needed. */ void tcp_pmtud_revert_segment_size(struct tcpcb *tp) { int32_t optlen; VERIFY(tp->t_pmtud_saved_maxopd > 0); tp->t_flags |= TF_PMTUD; tp->t_flags &= ~TF_BLACKHOLE; optlen = tp->t_maxopd - tp->t_maxseg; tp->t_maxopd = tp->t_pmtud_saved_maxopd; tp->t_maxseg = tp->t_maxopd - optlen; /* * Reset the slow-start flight size as it * may depend on the new MSS */ if (CC_ALGO(tp)->cwnd_init != NULL) CC_ALGO(tp)->cwnd_init(tp); tp->t_pmtud_start_ts = 0; tcpstat.tcps_pmtudbh_reverted++; /* change MSS according to recommendation, if there was one */ tcp_update_mss_locked(tp->t_inpcb->inp_socket, NULL); } /* * TCP timer processing. */ struct tcpcb * tcp_timers(struct tcpcb *tp, int timer) { int32_t rexmt, optlen = 0, idle_time = 0; struct socket *so; struct tcptemp *t_template; #if TCPDEBUG int ostate; #endif #if INET6 int isipv6 = (tp->t_inpcb->inp_vflag & INP_IPV4) == 0; #endif /* INET6 */ u_int64_t accsleep_ms; u_int32_t last_sleep_ms = 0; so = tp->t_inpcb->inp_socket; idle_time = tcp_now - tp->t_rcvtime; switch (timer) { /* * 2 MSL timeout in shutdown went off. If we're closed but * still waiting for peer to close and connection has been idle * too long, or if 2MSL time is up from TIME_WAIT or FIN_WAIT_2, * delete connection control block. * Otherwise, (this case shouldn't happen) check again in a bit * we keep the socket in the main list in that case. */ case TCPT_2MSL: tcp_free_sackholes(tp); if (tp->t_state != TCPS_TIME_WAIT && tp->t_state != TCPS_FIN_WAIT_2 && ((idle_time > 0) && (idle_time < TCP_CONN_MAXIDLE(tp)))) { tp->t_timer[TCPT_2MSL] = OFFSET_FROM_START(tp, (u_int32_t)TCP_CONN_KEEPINTVL(tp)); } else { tp = tcp_close(tp); return(tp); } break; /* * Retransmission timer went off. Message has not * been acked within retransmit interval. Back off * to a longer retransmit interval and retransmit one segment. */ case TCPT_REXMT: absolutetime_to_nanoseconds(mach_absolutetime_asleep, &accsleep_ms); accsleep_ms = accsleep_ms / 1000000UL; if (accsleep_ms > tp->t_accsleep_ms) last_sleep_ms = accsleep_ms - tp->t_accsleep_ms; /* * Drop a connection in the retransmit timer * 1. If we have retransmitted more than TCP_MAXRXTSHIFT * times * 2. If the time spent in this retransmission episode is * more than the time limit set with TCP_RXT_CONNDROPTIME * socket option * 3. If TCP_RXT_FINDROP socket option was set and * we have already retransmitted the FIN 3 times without * receiving an ack */ if (++tp->t_rxtshift > TCP_MAXRXTSHIFT || (tp->t_rxt_conndroptime > 0 && tp->t_rxtstart > 0 && (tcp_now - tp->t_rxtstart) >= tp->t_rxt_conndroptime) || ((tp->t_flagsext & TF_RXTFINDROP) != 0 && (tp->t_flags & TF_SENTFIN) != 0 && tp->t_rxtshift >= 4) || (tp->t_rxtshift > 4 && last_sleep_ms >= TCP_SLEEP_TOO_LONG)) { if (tp->t_state == TCPS_ESTABLISHED && tp->t_rxt_minimum_timeout > 0) { /* * Avoid dropping a connection if minimum * timeout is set and that time did not * pass. We will retry sending * retransmissions at the maximum interval */ if (TSTMP_LT(tcp_now, (tp->t_rxtstart + tp->t_rxt_minimum_timeout))) { tp->t_rxtshift = TCP_MAXRXTSHIFT - 1; goto retransmit_packet; } } if ((tp->t_flagsext & TF_RXTFINDROP) != 0) { tcpstat.tcps_rxtfindrop++; } else if (last_sleep_ms >= TCP_SLEEP_TOO_LONG) { tcpstat.tcps_drop_after_sleep++; } else { tcpstat.tcps_timeoutdrop++; } if (tp->t_rxtshift >= TCP_MAXRXTSHIFT) { if (TCP_ECN_ENABLED(tp)) { INP_INC_IFNET_STAT(tp->t_inpcb, ecn_on.rxmit_drop); } else { INP_INC_IFNET_STAT(tp->t_inpcb, ecn_off.rxmit_drop); } } tp->t_rxtshift = TCP_MAXRXTSHIFT; postevent(so, 0, EV_TIMEOUT); soevent(so, (SO_FILT_HINT_LOCKED|SO_FILT_HINT_TIMEOUT)); if (TCP_ECN_ENABLED(tp) && tp->t_state == TCPS_ESTABLISHED) tcp_heuristic_ecn_droprxmt(tp); tp = tcp_drop(tp, tp->t_softerror ? tp->t_softerror : ETIMEDOUT); break; } retransmit_packet: tcpstat.tcps_rexmttimeo++; tp->t_accsleep_ms = accsleep_ms; if (tp->t_rxtshift == 1 && tp->t_state == TCPS_ESTABLISHED) { /* Set the time at which retransmission started. */ tp->t_rxtstart = tcp_now; /* * if this is the first retransmit timeout, save * the state so that we can recover if the timeout * is spurious. */ tcp_rexmt_save_state(tp); } #if MPTCP if ((tp->t_rxtshift >= mptcp_fail_thresh) && (tp->t_state == TCPS_ESTABLISHED) && (tp->t_mpflags & TMPF_MPTCP_TRUE)) mptcp_act_on_txfail(so); if (so->so_flags & SOF_MP_SUBFLOW) { struct mptses *mpte = tptomptp(tp)->mpt_mpte; mptcp_check_subflows_and_add(mpte); } #endif /* MPTCP */ if (tp->t_adaptive_wtimo > 0 && tp->t_rxtshift > tp->t_adaptive_wtimo && TCPS_HAVEESTABLISHED(tp->t_state)) { /* Send an event to the application */ soevent(so, (SO_FILT_HINT_LOCKED| SO_FILT_HINT_ADAPTIVE_WTIMO)); } /* * If this is a retransmit timeout after PTO, the PTO * was not effective */ if (tp->t_flagsext & TF_SENT_TLPROBE) { tp->t_flagsext &= ~(TF_SENT_TLPROBE); tcpstat.tcps_rto_after_pto++; } if (tp->t_flagsext & TF_DELAY_RECOVERY) { /* * Retransmit timer fired before entering recovery * on a connection with packet re-ordering. This * suggests that the reordering metrics computed * are not accurate. */ tp->t_reorderwin = 0; tp->t_timer[TCPT_DELAYFR] = 0; tp->t_flagsext &= ~(TF_DELAY_RECOVERY); } if (tp->t_state == TCPS_SYN_RECEIVED) tcp_disable_tfo(tp); if (!(tp->t_tfo_flags & TFO_F_HEURISTIC_DONE) && (tp->t_tfo_stats & TFO_S_SYN_DATA_SENT) && !(tp->t_tfo_flags & TFO_F_NO_SNDPROBING) && ((tp->t_state != TCPS_SYN_SENT && tp->t_rxtshift > 1) || tp->t_rxtshift > 4)) { /* * For regular retransmissions, a first one is being * done for tail-loss probe. * Thus, if rxtshift > 1, this means we have sent the segment * a total of 3 times. * * If we are in SYN-SENT state, then there is no tail-loss * probe thus we have to let rxtshift go up to 3. */ tcp_heuristic_tfo_middlebox(tp); so->so_error = ENODATA; sorwakeup(so); sowwakeup(so); tp->t_tfo_stats |= TFO_S_SEND_BLACKHOLE; tcpstat.tcps_tfo_sndblackhole++; } if (!(tp->t_tfo_flags & TFO_F_HEURISTIC_DONE) && (tp->t_tfo_stats & TFO_S_SYN_DATA_ACKED) && tp->t_rxtshift > 3) { if (TSTMP_GT(tp->t_sndtime - 10 * TCP_RETRANSHZ, tp->t_rcvtime)) { tcp_heuristic_tfo_middlebox(tp); so->so_error = ENODATA; sorwakeup(so); sowwakeup(so); } } if (tp->t_state == TCPS_SYN_SENT) { rexmt = TCP_REXMTVAL(tp) * tcp_syn_backoff[tp->t_rxtshift]; tp->t_stat.synrxtshift = tp->t_rxtshift; /* When retransmitting, disable TFO */ if (tfo_enabled(tp) && (!(so->so_flags1 & SOF1_DATA_AUTHENTICATED) || (tp->t_flagsext & TF_FASTOPEN_HEUR))) { tp->t_flagsext &= ~TF_FASTOPEN; tp->t_tfo_flags |= TFO_F_SYN_LOSS; } } else { rexmt = TCP_REXMTVAL(tp) * tcp_backoff[tp->t_rxtshift]; } TCPT_RANGESET(tp->t_rxtcur, rexmt, tp->t_rttmin, TCPTV_REXMTMAX, TCP_ADD_REXMTSLOP(tp)); tp->t_timer[TCPT_REXMT] = OFFSET_FROM_START(tp, tp->t_rxtcur); if (INP_WAIT_FOR_IF_FEEDBACK(tp->t_inpcb)) goto fc_output; tcp_free_sackholes(tp); /* * Check for potential Path MTU Discovery Black Hole */ if (tcp_pmtud_black_hole_detect && !(tp->t_flagsext & TF_NOBLACKHOLE_DETECTION) && (tp->t_state == TCPS_ESTABLISHED)) { if ((tp->t_flags & TF_PMTUD) && ((tp->t_flags & TF_MAXSEGSNT) || tp->t_pmtud_lastseg_size > tcp_pmtud_black_hole_mss) && tp->t_rxtshift == 2) { /* * Enter Path MTU Black-hole Detection mechanism: * - Disable Path MTU Discovery (IP "DF" bit). * - Reduce MTU to lower value than what we * negotiated with the peer. */ /* Disable Path MTU Discovery for now */ tp->t_flags &= ~TF_PMTUD; /* Record that we may have found a black hole */ tp->t_flags |= TF_BLACKHOLE; optlen = tp->t_maxopd - tp->t_maxseg; /* Keep track of previous MSS */ tp->t_pmtud_saved_maxopd = tp->t_maxopd; tp->t_pmtud_start_ts = tcp_now; if (tp->t_pmtud_start_ts == 0) tp->t_pmtud_start_ts++; /* Reduce the MSS to intermediary value */ if (tp->t_maxopd > tcp_pmtud_black_hole_mss) { tp->t_maxopd = tcp_pmtud_black_hole_mss; } else { tp->t_maxopd = /* use the default MSS */ #if INET6 isipv6 ? tcp_v6mssdflt : #endif /* INET6 */ tcp_mssdflt; } tp->t_maxseg = tp->t_maxopd - optlen; /* * Reset the slow-start flight size * as it may depend on the new MSS */ if (CC_ALGO(tp)->cwnd_init != NULL) CC_ALGO(tp)->cwnd_init(tp); tp->snd_cwnd = tp->t_maxseg; } /* * If further retransmissions are still * unsuccessful with a lowered MTU, maybe this * isn't a Black Hole and we restore the previous * MSS and blackhole detection flags. */ else { if ((tp->t_flags & TF_BLACKHOLE) && (tp->t_rxtshift > 4)) { tcp_pmtud_revert_segment_size(tp); tp->snd_cwnd = tp->t_maxseg; } } } /* * Disable rfc1323 and rfc1644 if we haven't got any * response to our SYN (after we reach the threshold) * to work-around some broken terminal servers (most of * which have hopefully been retired) that have bad VJ * header compression code which trashes TCP segments * containing unknown-to-them TCP options. * Do this only on non-local connections. */ if (tp->t_state == TCPS_SYN_SENT && tp->t_rxtshift == tcp_broken_peer_syn_rxmit_thres) tp->t_flags &= ~(TF_REQ_SCALE|TF_REQ_TSTMP|TF_REQ_CC); /* * If losing, let the lower level know and try for * a better route. Also, if we backed off this far, * our srtt estimate is probably bogus. Clobber it * so we'll take the next rtt measurement as our srtt; * move the current srtt into rttvar to keep the current * retransmit times until then. */ if (tp->t_rxtshift > TCP_MAXRXTSHIFT / 4) { #if INET6 if (isipv6) in6_losing(tp->t_inpcb); else #endif /* INET6 */ in_losing(tp->t_inpcb); tp->t_rttvar += (tp->t_srtt >> TCP_RTT_SHIFT); tp->t_srtt = 0; } tp->snd_nxt = tp->snd_una; /* * Note: We overload snd_recover to function also as the * snd_last variable described in RFC 2582 */ tp->snd_recover = tp->snd_max; /* * Force a segment to be sent. */ tp->t_flags |= TF_ACKNOW; /* If timing a segment in this window, stop the timer */ tp->t_rtttime = 0; if (!IN_FASTRECOVERY(tp) && tp->t_rxtshift == 1) tcpstat.tcps_tailloss_rto++; /* * RFC 5681 says: when a TCP sender detects segment loss * using retransmit timer and the given segment has already * been retransmitted by way of the retransmission timer at * least once, the value of ssthresh is held constant */ if (tp->t_rxtshift == 1 && CC_ALGO(tp)->after_timeout != NULL) { CC_ALGO(tp)->after_timeout(tp); /* * CWR notifications are to be sent on new data * right after Fast Retransmits and ECE * notification receipts. */ if (TCP_ECN_ENABLED(tp)) tp->ecn_flags |= TE_SENDCWR; } EXIT_FASTRECOVERY(tp); /* Exit cwnd non validated phase */ tp->t_flagsext &= ~TF_CWND_NONVALIDATED; fc_output: tcp_ccdbg_trace(tp, NULL, TCP_CC_REXMT_TIMEOUT); (void) tcp_output(tp); break; /* * Persistance timer into zero window. * Force a byte to be output, if possible. */ case TCPT_PERSIST: tcpstat.tcps_persisttimeo++; /* * Hack: if the peer is dead/unreachable, we do not * time out if the window is closed. After a full * backoff, drop the connection if the idle time * (no responses to probes) reaches the maximum * backoff that we would use if retransmitting. * * Drop the connection if we reached the maximum allowed time for * Zero Window Probes without a non-zero update from the peer. * See rdar://5805356 */ if ((tp->t_rxtshift == TCP_MAXRXTSHIFT && (idle_time >= tcp_maxpersistidle || idle_time >= TCP_REXMTVAL(tp) * tcp_totbackoff)) || ((tp->t_persist_stop != 0) && TSTMP_LEQ(tp->t_persist_stop, tcp_now))) { tcpstat.tcps_persistdrop++; postevent(so, 0, EV_TIMEOUT); soevent(so, (SO_FILT_HINT_LOCKED|SO_FILT_HINT_TIMEOUT)); tp = tcp_drop(tp, ETIMEDOUT); break; } tcp_setpersist(tp); tp->t_flagsext |= TF_FORCE; (void) tcp_output(tp); tp->t_flagsext &= ~TF_FORCE; break; /* * Keep-alive timer went off; send something * or drop connection if idle for too long. */ case TCPT_KEEP: tcpstat.tcps_keeptimeo++; #if MPTCP /* * Regular TCP connections do not send keepalives after closing * MPTCP must not also, after sending Data FINs. */ struct mptcb *mp_tp = tptomptp(tp); if ((tp->t_mpflags & TMPF_MPTCP_TRUE) && (tp->t_state > TCPS_ESTABLISHED)) { goto dropit; } else if (mp_tp != NULL) { if ((mptcp_ok_to_keepalive(mp_tp) == 0)) goto dropit; } #endif /* MPTCP */ if (tp->t_state < TCPS_ESTABLISHED) goto dropit; if ((always_keepalive || (tp->t_inpcb->inp_socket->so_options & SO_KEEPALIVE) || (tp->t_flagsext & TF_DETECT_READSTALL) || (tp->t_tfo_probe_state == TFO_PROBE_PROBING)) && (tp->t_state <= TCPS_CLOSING || tp->t_state == TCPS_FIN_WAIT_2)) { if (idle_time >= TCP_CONN_KEEPIDLE(tp) + TCP_CONN_MAXIDLE(tp)) goto dropit; /* * Send a packet designed to force a response * if the peer is up and reachable: * either an ACK if the connection is still alive, * or an RST if the peer has closed the connection * due to timeout or reboot. * Using sequence number tp->snd_una-1 * causes the transmitted zero-length segment * to lie outside the receive window; * by the protocol spec, this requires the * correspondent TCP to respond. */ tcpstat.tcps_keepprobe++; t_template = tcp_maketemplate(tp); if (t_template) { struct inpcb *inp = tp->t_inpcb; struct tcp_respond_args tra; bzero(&tra, sizeof(tra)); tra.nocell = INP_NO_CELLULAR(inp); tra.noexpensive = INP_NO_EXPENSIVE(inp); tra.awdl_unrestricted = INP_AWDL_UNRESTRICTED(inp); tra.intcoproc_allowed = INP_INTCOPROC_ALLOWED(inp); if (tp->t_inpcb->inp_flags & INP_BOUND_IF) tra.ifscope = tp->t_inpcb->inp_boundifp->if_index; else tra.ifscope = IFSCOPE_NONE; tcp_respond(tp, t_template->tt_ipgen, &t_template->tt_t, (struct mbuf *)NULL, tp->rcv_nxt, tp->snd_una - 1, 0, &tra); (void) m_free(dtom(t_template)); if (tp->t_flagsext & TF_DETECT_READSTALL) tp->t_rtimo_probes++; } tp->t_timer[TCPT_KEEP] = OFFSET_FROM_START(tp, TCP_CONN_KEEPINTVL(tp)); } else { tp->t_timer[TCPT_KEEP] = OFFSET_FROM_START(tp, TCP_CONN_KEEPIDLE(tp)); } if (tp->t_flagsext & TF_DETECT_READSTALL) { struct ifnet *outifp = tp->t_inpcb->inp_last_outifp; bool reenable_probe = false; /* * The keep alive packets sent to detect a read * stall did not get a response from the * peer. Generate more keep-alives to confirm this. * If the number of probes sent reaches the limit, * generate an event. */ if (tp->t_adaptive_rtimo > 0) { if (tp->t_rtimo_probes > tp->t_adaptive_rtimo) { /* Generate an event */ soevent(so, (SO_FILT_HINT_LOCKED | SO_FILT_HINT_ADAPTIVE_RTIMO)); tcp_keepalive_reset(tp); } else { reenable_probe = true; } } else if (outifp != NULL && (outifp->if_eflags & IFEF_PROBE_CONNECTIVITY) && tp->t_rtimo_probes <= TCP_CONNECTIVITY_PROBES_MAX) { reenable_probe = true; } else { tp->t_flagsext &= ~TF_DETECT_READSTALL; } if (reenable_probe) { int ind = min(tp->t_rtimo_probes, TCP_MAXRXTSHIFT); tp->t_timer[TCPT_KEEP] = OFFSET_FROM_START( tp, tcp_backoff[ind] * TCP_REXMTVAL(tp)); } } if (tp->t_tfo_probe_state == TFO_PROBE_PROBING) { int ind; tp->t_tfo_probes++; ind = min(tp->t_tfo_probes, TCP_MAXRXTSHIFT); /* * We take the minimum among the time set by true * keepalive (see above) and the backoff'd RTO. That * way we backoff in case of packet-loss but will never * timeout slower than regular keepalive due to the * backing off. */ tp->t_timer[TCPT_KEEP] = min(OFFSET_FROM_START( tp, tcp_backoff[ind] * TCP_REXMTVAL(tp)), tp->t_timer[TCPT_KEEP]); } else if (!(tp->t_tfo_flags & TFO_F_HEURISTIC_DONE) && tp->t_tfo_probe_state == TFO_PROBE_WAIT_DATA) { /* Still no data! Let's assume a TFO-error and err out... */ tcp_heuristic_tfo_middlebox(tp); so->so_error = ENODATA; sorwakeup(so); tp->t_tfo_stats |= TFO_S_RECV_BLACKHOLE; tcpstat.tcps_tfo_blackhole++; } break; case TCPT_DELACK: if (tcp_delack_enabled && (tp->t_flags & TF_DELACK)) { tp->t_flags &= ~TF_DELACK; tp->t_timer[TCPT_DELACK] = 0; tp->t_flags |= TF_ACKNOW; /* * If delayed ack timer fired while stretching * acks, count the number of times the streaming * detection was not correct. If this exceeds a * threshold, disable strech ack on this * connection * * Also, go back to acking every other packet. */ if ((tp->t_flags & TF_STRETCHACK)) { if (tp->t_unacksegs > 1 && tp->t_unacksegs < maxseg_unacked) tp->t_stretchack_delayed++; if (tp->t_stretchack_delayed > TCP_STRETCHACK_DELAY_THRESHOLD) { tp->t_flagsext |= TF_DISABLE_STRETCHACK; /* * Note the time at which stretch * ack was disabled automatically */ tp->rcv_nostrack_ts = tcp_now; tcpstat.tcps_nostretchack++; tp->t_stretchack_delayed = 0; tp->rcv_nostrack_pkts = 0; } tcp_reset_stretch_ack(tp); } /* * If we are measuring inter packet arrival jitter * for throttling a connection, this delayed ack * might be the reason for accumulating some * jitter. So let's restart the measurement. */ CLEAR_IAJ_STATE(tp); tcpstat.tcps_delack++; (void) tcp_output(tp); } break; #if MPTCP case TCPT_JACK_RXMT: if ((tp->t_state == TCPS_ESTABLISHED) && (tp->t_mpflags & TMPF_PREESTABLISHED) && (tp->t_mpflags & TMPF_JOINED_FLOW)) { if (++tp->t_mprxtshift > TCP_MAXRXTSHIFT) { tcpstat.tcps_timeoutdrop++; postevent(so, 0, EV_TIMEOUT); soevent(so, (SO_FILT_HINT_LOCKED| SO_FILT_HINT_TIMEOUT)); tp = tcp_drop(tp, tp->t_softerror ? tp->t_softerror : ETIMEDOUT); break; } tcpstat.tcps_join_rxmts++; tp->t_mpflags |= TMPF_SND_JACK; tp->t_flags |= TF_ACKNOW; /* * No backoff is implemented for simplicity for this * corner case. */ (void) tcp_output(tp); } break; #endif /* MPTCP */ case TCPT_PTO: { int32_t snd_len; tp->t_flagsext &= ~(TF_SENT_TLPROBE); /* * Check if the connection is in the right state to * send a probe */ if (tp->t_state != TCPS_ESTABLISHED || (tp->t_rxtshift > 0 && !(tp->t_flagsext & TF_PROBING)) || tp->snd_max == tp->snd_una || !SACK_ENABLED(tp) || !TAILQ_EMPTY(&tp->snd_holes) || IN_FASTRECOVERY(tp)) break; /* * If there is no new data to send or if the * connection is limited by receive window then * retransmit the last segment, otherwise send * new data. */ snd_len = min(so->so_snd.sb_cc, tp->snd_wnd) - (tp->snd_max - tp->snd_una); if (snd_len > 0) { tp->snd_nxt = tp->snd_max; } else { snd_len = min((tp->snd_max - tp->snd_una), tp->t_maxseg); tp->snd_nxt = tp->snd_max - snd_len; } tcpstat.tcps_pto++; if (tp->t_flagsext & TF_PROBING) tcpstat.tcps_probe_if++; /* If timing a segment in this window, stop the timer */ tp->t_rtttime = 0; /* Note that tail loss probe is being sent */ tp->t_flagsext |= TF_SENT_TLPROBE; tp->t_tlpstart = tcp_now; tp->snd_cwnd += tp->t_maxseg; /* * When tail-loss-probe fires, we reset the RTO timer, because * a probe just got sent, so we are good to push out the timer. * * Set to 0 to ensure that tcp_output() will reschedule it */ tp->t_timer[TCPT_REXMT] = 0; (void )tcp_output(tp); tp->snd_cwnd -= tp->t_maxseg; tp->t_tlphighrxt = tp->snd_nxt; break; } case TCPT_DELAYFR: tp->t_flagsext &= ~TF_DELAY_RECOVERY; /* * Don't do anything if one of the following is true: * - the connection is already in recovery * - sequence until snd_recover has been acknowledged. * - retransmit timeout has fired */ if (IN_FASTRECOVERY(tp) || SEQ_GEQ(tp->snd_una, tp->snd_recover) || tp->t_rxtshift > 0) break; VERIFY(SACK_ENABLED(tp)); tcp_rexmt_save_state(tp); if (CC_ALGO(tp)->pre_fr != NULL) { CC_ALGO(tp)->pre_fr(tp); if (TCP_ECN_ENABLED(tp)) tp->ecn_flags |= TE_SENDCWR; } ENTER_FASTRECOVERY(tp); tp->t_timer[TCPT_REXMT] = 0; tcpstat.tcps_sack_recovery_episode++; tp->t_sack_recovery_episode++; tp->sack_newdata = tp->snd_nxt; tp->snd_cwnd = tp->t_maxseg; tcp_ccdbg_trace(tp, NULL, TCP_CC_ENTER_FASTRECOVERY); (void) tcp_output(tp); break; dropit: tcpstat.tcps_keepdrops++; postevent(so, 0, EV_TIMEOUT); soevent(so, (SO_FILT_HINT_LOCKED|SO_FILT_HINT_TIMEOUT)); tp = tcp_drop(tp, ETIMEDOUT); break; } #if TCPDEBUG if (tp->t_inpcb->inp_socket->so_options & SO_DEBUG) tcp_trace(TA_USER, ostate, tp, (void *)0, (struct tcphdr *)0, PRU_SLOWTIMO); #endif return (tp); } /* Remove a timer entry from timer list */ void tcp_remove_timer(struct tcpcb *tp) { struct tcptimerlist *listp = &tcp_timer_list; socket_lock_assert_owned(tp->t_inpcb->inp_socket); if (!(TIMER_IS_ON_LIST(tp))) { return; } lck_mtx_lock(listp->mtx); /* Check if pcb is on timer list again after acquiring the lock */ if (!(TIMER_IS_ON_LIST(tp))) { lck_mtx_unlock(listp->mtx); return; } if (listp->next_te != NULL && listp->next_te == &tp->tentry) listp->next_te = LIST_NEXT(&tp->tentry, le); LIST_REMOVE(&tp->tentry, le); tp->t_flags &= ~(TF_TIMER_ONLIST); listp->entries--; tp->tentry.le.le_next = NULL; tp->tentry.le.le_prev = NULL; lck_mtx_unlock(listp->mtx); } /* * Function to check if the timerlist needs to be rescheduled to run * the timer entry correctly. Basically, this is to check if we can avoid * taking the list lock. */ static boolean_t need_to_resched_timerlist(u_int32_t runtime, u_int16_t mode) { struct tcptimerlist *listp = &tcp_timer_list; int32_t diff; /* * If the list is being processed then the state of the list is * in flux. In this case always acquire the lock and set the state * correctly. */ if (listp->running) return (TRUE); if (!listp->scheduled) return (TRUE); diff = timer_diff(listp->runtime, 0, runtime, 0); if (diff <= 0) { /* The list is going to run before this timer */ return (FALSE); } else { if (mode & TCP_TIMERLIST_10MS_MODE) { if (diff <= TCP_TIMER_10MS_QUANTUM) return (FALSE); } else if (mode & TCP_TIMERLIST_100MS_MODE) { if (diff <= TCP_TIMER_100MS_QUANTUM) return (FALSE); } else { if (diff <= TCP_TIMER_500MS_QUANTUM) return (FALSE); } } return (TRUE); } void tcp_sched_timerlist(uint32_t offset) { uint64_t deadline = 0; struct tcptimerlist *listp = &tcp_timer_list; LCK_MTX_ASSERT(listp->mtx, LCK_MTX_ASSERT_OWNED); offset = min(offset, TCP_TIMERLIST_MAX_OFFSET); listp->runtime = tcp_now + offset; listp->schedtime = tcp_now; if (listp->runtime == 0) { listp->runtime++; offset++; } clock_interval_to_deadline(offset, USEC_PER_SEC, &deadline); thread_call_enter_delayed(listp->call, deadline); listp->scheduled = TRUE; } /* * Function to run the timers for a connection. * * Returns the offset of next timer to be run for this connection which * can be used to reschedule the timerlist. * * te_mode is an out parameter that indicates the modes of active * timers for this connection. */ u_int32_t tcp_run_conn_timer(struct tcpcb *tp, u_int16_t *te_mode, u_int16_t probe_if_index) { struct socket *so; u_int16_t i = 0, index = TCPT_NONE, lo_index = TCPT_NONE; u_int32_t timer_val, offset = 0, lo_timer = 0; int32_t diff; boolean_t needtorun[TCPT_NTIMERS]; int count = 0; VERIFY(tp != NULL); bzero(needtorun, sizeof(needtorun)); *te_mode = 0; socket_lock(tp->t_inpcb->inp_socket, 1); so = tp->t_inpcb->inp_socket; /* Release the want count on inp */ if (in_pcb_checkstate(tp->t_inpcb, WNT_RELEASE, 1) == WNT_STOPUSING) { if (TIMER_IS_ON_LIST(tp)) { tcp_remove_timer(tp); } /* Looks like the TCP connection got closed while we * were waiting for the lock.. Done */ goto done; } /* * If this connection is over an interface that needs to * be probed, send probe packets to reinitiate communication. */ if (probe_if_index > 0 && tp->t_inpcb->inp_last_outifp != NULL && tp->t_inpcb->inp_last_outifp->if_index == probe_if_index) { tp->t_flagsext |= TF_PROBING; tcp_timers(tp, TCPT_PTO); tp->t_timer[TCPT_PTO] = 0; tp->t_flagsext &= ~TF_PROBING; } /* * Since the timer thread needs to wait for tcp lock, it may race * with another thread that can cancel or reschedule the timer * that is about to run. Check if we need to run anything. */ if ((index = tp->tentry.index) == TCPT_NONE) goto done; timer_val = tp->t_timer[index]; diff = timer_diff(tp->tentry.runtime, 0, tcp_now, 0); if (diff > 0) { if (tp->tentry.index != TCPT_NONE) { offset = diff; *(te_mode) = tp->tentry.mode; } goto done; } tp->t_timer[index] = 0; if (timer_val > 0) { tp = tcp_timers(tp, index); if (tp == NULL) goto done; } /* * Check if there are any other timers that need to be run. * While doing it, adjust the timer values wrt tcp_now. */ tp->tentry.mode = 0; for (i = 0; i < TCPT_NTIMERS; ++i) { if (tp->t_timer[i] != 0) { diff = timer_diff(tp->tentry.timer_start, tp->t_timer[i], tcp_now, 0); if (diff <= 0) { needtorun[i] = TRUE; count++; } else { tp->t_timer[i] = diff; needtorun[i] = FALSE; if (lo_timer == 0 || diff < lo_timer) { lo_timer = diff; lo_index = i; } TCP_SET_TIMER_MODE(tp->tentry.mode, i); } } } tp->tentry.timer_start = tcp_now; tp->tentry.index = lo_index; VERIFY(tp->tentry.index == TCPT_NONE || tp->tentry.mode > 0); if (tp->tentry.index != TCPT_NONE) { tp->tentry.runtime = tp->tentry.timer_start + tp->t_timer[tp->tentry.index]; if (tp->tentry.runtime == 0) tp->tentry.runtime++; } if (count > 0) { /* run any other timers outstanding at this time. */ for (i = 0; i < TCPT_NTIMERS; ++i) { if (needtorun[i]) { tp->t_timer[i] = 0; tp = tcp_timers(tp, i); if (tp == NULL) { offset = 0; *(te_mode) = 0; goto done; } } } tcp_set_lotimer_index(tp); } if (tp->tentry.index < TCPT_NONE) { offset = tp->t_timer[tp->tentry.index]; *(te_mode) = tp->tentry.mode; } done: if (tp != NULL && tp->tentry.index == TCPT_NONE) { tcp_remove_timer(tp); offset = 0; } socket_unlock(so, 1); return(offset); } void tcp_run_timerlist(void * arg1, void * arg2) { #pragma unused(arg1, arg2) struct tcptimerentry *te, *next_te; struct tcptimerlist *listp = &tcp_timer_list; struct tcpcb *tp; uint32_t next_timer = 0; /* offset of the next timer on the list */ u_int16_t te_mode = 0; /* modes of all active timers in a tcpcb */ u_int16_t list_mode = 0; /* cumulative of modes of all tcpcbs */ uint32_t active_count = 0; calculate_tcp_clock(); lck_mtx_lock(listp->mtx); int32_t drift = tcp_now - listp->runtime; if (drift <= 1) { tcpstat.tcps_timer_drift_le_1_ms++; } else if (drift <= 10) { tcpstat.tcps_timer_drift_le_10_ms++; } else if (drift <= 20) { tcpstat.tcps_timer_drift_le_20_ms++; } else if (drift <= 50) { tcpstat.tcps_timer_drift_le_50_ms++; } else if (drift <= 100) { tcpstat.tcps_timer_drift_le_100_ms++; } else if (drift <= 200) { tcpstat.tcps_timer_drift_le_200_ms++; } else if (drift <= 500) { tcpstat.tcps_timer_drift_le_500_ms++; } else if (drift <= 1000) { tcpstat.tcps_timer_drift_le_1000_ms++; } else { tcpstat.tcps_timer_drift_gt_1000_ms++; } listp->running = TRUE; LIST_FOREACH_SAFE(te, &listp->lhead, le, next_te) { uint32_t offset = 0; uint32_t runtime = te->runtime; if (te->index < TCPT_NONE && TSTMP_GT(runtime, tcp_now)) { offset = timer_diff(runtime, 0, tcp_now, 0); if (next_timer == 0 || offset < next_timer) { next_timer = offset; } list_mode |= te->mode; continue; } tp = TIMERENTRY_TO_TP(te); /* * Acquire an inp wantcnt on the inpcb so that the socket * won't get detached even if tcp_close is called */ if (in_pcb_checkstate(tp->t_inpcb, WNT_ACQUIRE, 0) == WNT_STOPUSING) { /* * Some how this pcb went into dead state while * on the timer list, just take it off the list. * Since the timer list entry pointers are * protected by the timer list lock, we can * do it here without the socket lock. */ if (TIMER_IS_ON_LIST(tp)) { tp->t_flags &= ~(TF_TIMER_ONLIST); LIST_REMOVE(&tp->tentry, le); listp->entries--; tp->tentry.le.le_next = NULL; tp->tentry.le.le_prev = NULL; } continue; } active_count++; /* * Store the next timerentry pointer before releasing the * list lock. If that entry has to be removed when we * release the lock, this pointer will be updated to the * element after that. */ listp->next_te = next_te; VERIFY_NEXT_LINK(&tp->tentry, le); VERIFY_PREV_LINK(&tp->tentry, le); lck_mtx_unlock(listp->mtx); offset = tcp_run_conn_timer(tp, &te_mode, listp->probe_if_index); lck_mtx_lock(listp->mtx); next_te = listp->next_te; listp->next_te = NULL; if (offset > 0 && te_mode != 0) { list_mode |= te_mode; if (next_timer == 0 || offset < next_timer) next_timer = offset; } } if (!LIST_EMPTY(&listp->lhead)) { u_int16_t next_mode = 0; if ((list_mode & TCP_TIMERLIST_10MS_MODE) || (listp->pref_mode & TCP_TIMERLIST_10MS_MODE)) next_mode = TCP_TIMERLIST_10MS_MODE; else if ((list_mode & TCP_TIMERLIST_100MS_MODE) || (listp->pref_mode & TCP_TIMERLIST_100MS_MODE)) next_mode = TCP_TIMERLIST_100MS_MODE; else next_mode = TCP_TIMERLIST_500MS_MODE; if (next_mode != TCP_TIMERLIST_500MS_MODE) { listp->idleruns = 0; } else { /* * the next required mode is slow mode, but if * the last one was a faster mode and we did not * have enough idle runs, repeat the last mode. * * We try to keep the timer list in fast mode for * some idle time in expectation of new data. */ if (listp->mode != next_mode && listp->idleruns < timer_fastmode_idlemax) { listp->idleruns++; next_mode = listp->mode; next_timer = TCP_TIMER_100MS_QUANTUM; } else { listp->idleruns = 0; } } listp->mode = next_mode; if (listp->pref_offset != 0) next_timer = min(listp->pref_offset, next_timer); if (listp->mode == TCP_TIMERLIST_500MS_MODE) next_timer = max(next_timer, TCP_TIMER_500MS_QUANTUM); tcp_sched_timerlist(next_timer); } else { /* * No need to reschedule this timer, but always run * periodically at a much higher granularity. */ tcp_sched_timerlist(TCP_TIMERLIST_MAX_OFFSET); } listp->running = FALSE; listp->pref_mode = 0; listp->pref_offset = 0; listp->probe_if_index = 0; lck_mtx_unlock(listp->mtx); } /* * Function to check if the timerlist needs to be rescheduled to run this * connection's timers correctly. */ void tcp_sched_timers(struct tcpcb *tp) { struct tcptimerentry *te = &tp->tentry; u_int16_t index = te->index; u_int16_t mode = te->mode; struct tcptimerlist *listp = &tcp_timer_list; int32_t offset = 0; boolean_t list_locked = FALSE; if (tp->t_inpcb->inp_state == INPCB_STATE_DEAD) { /* Just return without adding the dead pcb to the list */ if (TIMER_IS_ON_LIST(tp)) { tcp_remove_timer(tp); } return; } if (index == TCPT_NONE) { /* Nothing to run */ tcp_remove_timer(tp); return; } /* * compute the offset at which the next timer for this connection * has to run. */ offset = timer_diff(te->runtime, 0, tcp_now, 0); if (offset <= 0) { offset = 1; tcp_timer_advanced++; } if (!TIMER_IS_ON_LIST(tp)) { if (!list_locked) { lck_mtx_lock(listp->mtx); list_locked = TRUE; } if (!TIMER_IS_ON_LIST(tp)) { LIST_INSERT_HEAD(&listp->lhead, te, le); tp->t_flags |= TF_TIMER_ONLIST; listp->entries++; if (listp->entries > listp->maxentries) listp->maxentries = listp->entries; /* if the list is not scheduled, just schedule it */ if (!listp->scheduled) goto schedule; } } /* * Timer entry is currently on the list, check if the list needs * to be rescheduled. */ if (need_to_resched_timerlist(te->runtime, mode)) { tcp_resched_timerlist++; if (!list_locked) { lck_mtx_lock(listp->mtx); list_locked = TRUE; } VERIFY_NEXT_LINK(te, le); VERIFY_PREV_LINK(te, le); if (listp->running) { listp->pref_mode |= mode; if (listp->pref_offset == 0 || offset < listp->pref_offset) { listp->pref_offset = offset; } } else { /* * The list could have got rescheduled while * this thread was waiting for the lock */ if (listp->scheduled) { int32_t diff; diff = timer_diff(listp->runtime, 0, tcp_now, offset); if (diff <= 0) goto done; else goto schedule; } else { goto schedule; } } } goto done; schedule: /* * Since a connection with timers is getting scheduled, the timer * list moves from idle to active state and that is why idlegen is * reset */ if (mode & TCP_TIMERLIST_10MS_MODE) { listp->mode = TCP_TIMERLIST_10MS_MODE; listp->idleruns = 0; offset = min(offset, TCP_TIMER_10MS_QUANTUM); } else if (mode & TCP_TIMERLIST_100MS_MODE) { if (listp->mode > TCP_TIMERLIST_100MS_MODE) listp->mode = TCP_TIMERLIST_100MS_MODE; listp->idleruns = 0; offset = min(offset, TCP_TIMER_100MS_QUANTUM); } tcp_sched_timerlist(offset); done: if (list_locked) lck_mtx_unlock(listp->mtx); return; } static inline void tcp_set_lotimer_index(struct tcpcb *tp) { uint16_t i, lo_index = TCPT_NONE, mode = 0; uint32_t lo_timer = 0; for (i = 0; i < TCPT_NTIMERS; ++i) { if (tp->t_timer[i] != 0) { TCP_SET_TIMER_MODE(mode, i); if (lo_timer == 0 || tp->t_timer[i] < lo_timer) { lo_timer = tp->t_timer[i]; lo_index = i; } } } tp->tentry.index = lo_index; tp->tentry.mode = mode; VERIFY(tp->tentry.index == TCPT_NONE || tp->tentry.mode > 0); if (tp->tentry.index != TCPT_NONE) { tp->tentry.runtime = tp->tentry.timer_start + tp->t_timer[tp->tentry.index]; if (tp->tentry.runtime == 0) tp->tentry.runtime++; } } void tcp_check_timer_state(struct tcpcb *tp) { socket_lock_assert_owned(tp->t_inpcb->inp_socket); if (tp->t_inpcb->inp_flags2 & INP2_TIMEWAIT) return; tcp_set_lotimer_index(tp); tcp_sched_timers(tp); return; } static inline void tcp_cumulative_stat(u_int32_t cur, u_int32_t *prev, u_int32_t *dest) { /* handle wrap around */ int32_t diff = (int32_t) (cur - *prev); if (diff > 0) *dest = diff; else *dest = 0; *prev = cur; return; } static inline void tcp_cumulative_stat64(u_int64_t cur, u_int64_t *prev, u_int64_t *dest) { /* handle wrap around */ int64_t diff = (int64_t) (cur - *prev); if (diff > 0) *dest = diff; else *dest = 0; *prev = cur; return; } __private_extern__ void tcp_report_stats(void) { struct nstat_sysinfo_data data; struct sockaddr_in dst; struct sockaddr_in6 dst6; struct rtentry *rt = NULL; static struct tcp_last_report_stats prev; u_int64_t var, uptime; #define stat data.u.tcp_stats if (((uptime = net_uptime()) - tcp_last_report_time) < tcp_report_stats_interval) return; tcp_last_report_time = uptime; bzero(&data, sizeof(data)); data.flags = NSTAT_SYSINFO_TCP_STATS; bzero(&dst, sizeof(dst)); dst.sin_len = sizeof(dst); dst.sin_family = AF_INET; /* ipv4 avg rtt */ lck_mtx_lock(rnh_lock); rt = rt_lookup(TRUE, (struct sockaddr *)&dst, NULL, rt_tables[AF_INET], IFSCOPE_NONE); lck_mtx_unlock(rnh_lock); if (rt != NULL) { RT_LOCK(rt); if (rt_primary_default(rt, rt_key(rt)) && rt->rt_stats != NULL) { stat.ipv4_avgrtt = rt->rt_stats->nstat_avg_rtt; } RT_UNLOCK(rt); rtfree(rt); rt = NULL; } /* ipv6 avg rtt */ bzero(&dst6, sizeof(dst6)); dst6.sin6_len = sizeof(dst6); dst6.sin6_family = AF_INET6; lck_mtx_lock(rnh_lock); rt = rt_lookup(TRUE,(struct sockaddr *)&dst6, NULL, rt_tables[AF_INET6], IFSCOPE_NONE); lck_mtx_unlock(rnh_lock); if (rt != NULL) { RT_LOCK(rt); if (rt_primary_default(rt, rt_key(rt)) && rt->rt_stats != NULL) { stat.ipv6_avgrtt = rt->rt_stats->nstat_avg_rtt; } RT_UNLOCK(rt); rtfree(rt); rt = NULL; } /* send packet loss rate, shift by 10 for precision */ if (tcpstat.tcps_sndpack > 0 && tcpstat.tcps_sndrexmitpack > 0) { var = tcpstat.tcps_sndrexmitpack << 10; stat.send_plr = (var * 100) / tcpstat.tcps_sndpack; } /* recv packet loss rate, shift by 10 for precision */ if (tcpstat.tcps_rcvpack > 0 && tcpstat.tcps_recovered_pkts > 0) { var = tcpstat.tcps_recovered_pkts << 10; stat.recv_plr = (var * 100) / tcpstat.tcps_rcvpack; } /* RTO after tail loss, shift by 10 for precision */ if (tcpstat.tcps_sndrexmitpack > 0 && tcpstat.tcps_tailloss_rto > 0) { var = tcpstat.tcps_tailloss_rto << 10; stat.send_tlrto_rate = (var * 100) / tcpstat.tcps_sndrexmitpack; } /* packet reordering */ if (tcpstat.tcps_sndpack > 0 && tcpstat.tcps_reordered_pkts > 0) { var = tcpstat.tcps_reordered_pkts << 10; stat.send_reorder_rate = (var * 100) / tcpstat.tcps_sndpack; } if (tcp_ecn_outbound == 1) stat.ecn_client_enabled = 1; if (tcp_ecn_inbound == 1) stat.ecn_server_enabled = 1; tcp_cumulative_stat(tcpstat.tcps_connattempt, &prev.tcps_connattempt, &stat.connection_attempts); tcp_cumulative_stat(tcpstat.tcps_accepts, &prev.tcps_accepts, &stat.connection_accepts); tcp_cumulative_stat(tcpstat.tcps_ecn_client_setup, &prev.tcps_ecn_client_setup, &stat.ecn_client_setup); tcp_cumulative_stat(tcpstat.tcps_ecn_server_setup, &prev.tcps_ecn_server_setup, &stat.ecn_server_setup); tcp_cumulative_stat(tcpstat.tcps_ecn_client_success, &prev.tcps_ecn_client_success, &stat.ecn_client_success); tcp_cumulative_stat(tcpstat.tcps_ecn_server_success, &prev.tcps_ecn_server_success, &stat.ecn_server_success); tcp_cumulative_stat(tcpstat.tcps_ecn_not_supported, &prev.tcps_ecn_not_supported, &stat.ecn_not_supported); tcp_cumulative_stat(tcpstat.tcps_ecn_lost_syn, &prev.tcps_ecn_lost_syn, &stat.ecn_lost_syn); tcp_cumulative_stat(tcpstat.tcps_ecn_lost_synack, &prev.tcps_ecn_lost_synack, &stat.ecn_lost_synack); tcp_cumulative_stat(tcpstat.tcps_ecn_recv_ce, &prev.tcps_ecn_recv_ce, &stat.ecn_recv_ce); tcp_cumulative_stat(tcpstat.tcps_ecn_recv_ece, &prev.tcps_ecn_recv_ece, &stat.ecn_recv_ece); tcp_cumulative_stat(tcpstat.tcps_ecn_recv_ece, &prev.tcps_ecn_recv_ece, &stat.ecn_recv_ece); tcp_cumulative_stat(tcpstat.tcps_ecn_sent_ece, &prev.tcps_ecn_sent_ece, &stat.ecn_sent_ece); tcp_cumulative_stat(tcpstat.tcps_ecn_sent_ece, &prev.tcps_ecn_sent_ece, &stat.ecn_sent_ece); tcp_cumulative_stat(tcpstat.tcps_ecn_conn_recv_ce, &prev.tcps_ecn_conn_recv_ce, &stat.ecn_conn_recv_ce); tcp_cumulative_stat(tcpstat.tcps_ecn_conn_recv_ece, &prev.tcps_ecn_conn_recv_ece, &stat.ecn_conn_recv_ece); tcp_cumulative_stat(tcpstat.tcps_ecn_conn_plnoce, &prev.tcps_ecn_conn_plnoce, &stat.ecn_conn_plnoce); tcp_cumulative_stat(tcpstat.tcps_ecn_conn_pl_ce, &prev.tcps_ecn_conn_pl_ce, &stat.ecn_conn_pl_ce); tcp_cumulative_stat(tcpstat.tcps_ecn_conn_nopl_ce, &prev.tcps_ecn_conn_nopl_ce, &stat.ecn_conn_nopl_ce); tcp_cumulative_stat(tcpstat.tcps_ecn_fallback_synloss, &prev.tcps_ecn_fallback_synloss, &stat.ecn_fallback_synloss); tcp_cumulative_stat(tcpstat.tcps_ecn_fallback_reorder, &prev.tcps_ecn_fallback_reorder, &stat.ecn_fallback_reorder); tcp_cumulative_stat(tcpstat.tcps_ecn_fallback_ce, &prev.tcps_ecn_fallback_ce, &stat.ecn_fallback_ce); tcp_cumulative_stat(tcpstat.tcps_tfo_syn_data_rcv, &prev.tcps_tfo_syn_data_rcv, &stat.tfo_syn_data_rcv); tcp_cumulative_stat(tcpstat.tcps_tfo_cookie_req_rcv, &prev.tcps_tfo_cookie_req_rcv, &stat.tfo_cookie_req_rcv); tcp_cumulative_stat(tcpstat.tcps_tfo_cookie_sent, &prev.tcps_tfo_cookie_sent, &stat.tfo_cookie_sent); tcp_cumulative_stat(tcpstat.tcps_tfo_cookie_invalid, &prev.tcps_tfo_cookie_invalid, &stat.tfo_cookie_invalid); tcp_cumulative_stat(tcpstat.tcps_tfo_cookie_req, &prev.tcps_tfo_cookie_req, &stat.tfo_cookie_req); tcp_cumulative_stat(tcpstat.tcps_tfo_cookie_rcv, &prev.tcps_tfo_cookie_rcv, &stat.tfo_cookie_rcv); tcp_cumulative_stat(tcpstat.tcps_tfo_syn_data_sent, &prev.tcps_tfo_syn_data_sent, &stat.tfo_syn_data_sent); tcp_cumulative_stat(tcpstat.tcps_tfo_syn_data_acked, &prev.tcps_tfo_syn_data_acked, &stat.tfo_syn_data_acked); tcp_cumulative_stat(tcpstat.tcps_tfo_syn_loss, &prev.tcps_tfo_syn_loss, &stat.tfo_syn_loss); tcp_cumulative_stat(tcpstat.tcps_tfo_blackhole, &prev.tcps_tfo_blackhole, &stat.tfo_blackhole); tcp_cumulative_stat(tcpstat.tcps_tfo_cookie_wrong, &prev.tcps_tfo_cookie_wrong, &stat.tfo_cookie_wrong); tcp_cumulative_stat(tcpstat.tcps_tfo_no_cookie_rcv, &prev.tcps_tfo_no_cookie_rcv, &stat.tfo_no_cookie_rcv); tcp_cumulative_stat(tcpstat.tcps_tfo_heuristics_disable, &prev.tcps_tfo_heuristics_disable, &stat.tfo_heuristics_disable); tcp_cumulative_stat(tcpstat.tcps_tfo_sndblackhole, &prev.tcps_tfo_sndblackhole, &stat.tfo_sndblackhole); tcp_cumulative_stat(tcpstat.tcps_mptcp_handover_attempt, &prev.tcps_mptcp_handover_attempt , &stat.mptcp_handover_attempt); tcp_cumulative_stat(tcpstat.tcps_mptcp_interactive_attempt, &prev.tcps_mptcp_interactive_attempt , &stat.mptcp_interactive_attempt); tcp_cumulative_stat(tcpstat.tcps_mptcp_aggregate_attempt, &prev.tcps_mptcp_aggregate_attempt , &stat.mptcp_aggregate_attempt); tcp_cumulative_stat(tcpstat.tcps_mptcp_fp_handover_attempt, &prev.tcps_mptcp_fp_handover_attempt , &stat.mptcp_fp_handover_attempt); tcp_cumulative_stat(tcpstat.tcps_mptcp_fp_interactive_attempt, &prev.tcps_mptcp_fp_interactive_attempt , &stat.mptcp_fp_interactive_attempt); tcp_cumulative_stat(tcpstat.tcps_mptcp_fp_aggregate_attempt, &prev.tcps_mptcp_fp_aggregate_attempt , &stat.mptcp_fp_aggregate_attempt); tcp_cumulative_stat(tcpstat.tcps_mptcp_heuristic_fallback, &prev.tcps_mptcp_heuristic_fallback , &stat.mptcp_heuristic_fallback); tcp_cumulative_stat(tcpstat.tcps_mptcp_fp_heuristic_fallback, &prev.tcps_mptcp_fp_heuristic_fallback , &stat.mptcp_fp_heuristic_fallback); tcp_cumulative_stat(tcpstat.tcps_mptcp_handover_success_wifi, &prev.tcps_mptcp_handover_success_wifi , &stat.mptcp_handover_success_wifi); tcp_cumulative_stat(tcpstat.tcps_mptcp_handover_success_cell, &prev.tcps_mptcp_handover_success_cell , &stat.mptcp_handover_success_cell); tcp_cumulative_stat(tcpstat.tcps_mptcp_interactive_success, &prev.tcps_mptcp_interactive_success , &stat.mptcp_interactive_success); tcp_cumulative_stat(tcpstat.tcps_mptcp_aggregate_success, &prev.tcps_mptcp_aggregate_success , &stat.mptcp_aggregate_success); tcp_cumulative_stat(tcpstat.tcps_mptcp_fp_handover_success_wifi, &prev.tcps_mptcp_fp_handover_success_wifi , &stat.mptcp_fp_handover_success_wifi); tcp_cumulative_stat(tcpstat.tcps_mptcp_fp_handover_success_cell, &prev.tcps_mptcp_fp_handover_success_cell , &stat.mptcp_fp_handover_success_cell); tcp_cumulative_stat(tcpstat.tcps_mptcp_fp_interactive_success, &prev.tcps_mptcp_fp_interactive_success , &stat.mptcp_fp_interactive_success); tcp_cumulative_stat(tcpstat.tcps_mptcp_fp_aggregate_success, &prev.tcps_mptcp_fp_aggregate_success , &stat.mptcp_fp_aggregate_success); tcp_cumulative_stat(tcpstat.tcps_mptcp_handover_cell_from_wifi, &prev.tcps_mptcp_handover_cell_from_wifi , &stat.mptcp_handover_cell_from_wifi); tcp_cumulative_stat(tcpstat.tcps_mptcp_handover_wifi_from_cell, &prev.tcps_mptcp_handover_wifi_from_cell , &stat.mptcp_handover_wifi_from_cell); tcp_cumulative_stat(tcpstat.tcps_mptcp_interactive_cell_from_wifi, &prev.tcps_mptcp_interactive_cell_from_wifi , &stat.mptcp_interactive_cell_from_wifi); tcp_cumulative_stat64(tcpstat.tcps_mptcp_handover_cell_bytes, &prev.tcps_mptcp_handover_cell_bytes , &stat.mptcp_handover_cell_bytes); tcp_cumulative_stat64(tcpstat.tcps_mptcp_interactive_cell_bytes, &prev.tcps_mptcp_interactive_cell_bytes , &stat.mptcp_interactive_cell_bytes); tcp_cumulative_stat64(tcpstat.tcps_mptcp_aggregate_cell_bytes, &prev.tcps_mptcp_aggregate_cell_bytes , &stat.mptcp_aggregate_cell_bytes); tcp_cumulative_stat64(tcpstat.tcps_mptcp_handover_all_bytes, &prev.tcps_mptcp_handover_all_bytes , &stat.mptcp_handover_all_bytes); tcp_cumulative_stat64(tcpstat.tcps_mptcp_interactive_all_bytes, &prev.tcps_mptcp_interactive_all_bytes , &stat.mptcp_interactive_all_bytes); tcp_cumulative_stat64(tcpstat.tcps_mptcp_aggregate_all_bytes, &prev.tcps_mptcp_aggregate_all_bytes , &stat.mptcp_aggregate_all_bytes); tcp_cumulative_stat(tcpstat.tcps_mptcp_back_to_wifi, &prev.tcps_mptcp_back_to_wifi , &stat.mptcp_back_to_wifi); tcp_cumulative_stat(tcpstat.tcps_mptcp_wifi_proxy, &prev.tcps_mptcp_wifi_proxy , &stat.mptcp_wifi_proxy); tcp_cumulative_stat(tcpstat.tcps_mptcp_cell_proxy, &prev.tcps_mptcp_cell_proxy , &stat.mptcp_cell_proxy); tcp_cumulative_stat(tcpstat.tcps_mptcp_triggered_cell, &prev.tcps_mptcp_triggered_cell, &stat.mptcp_triggered_cell); nstat_sysinfo_send_data(&data); #undef stat } void tcp_interface_send_probe(u_int16_t probe_if_index) { int32_t offset = 0; struct tcptimerlist *listp = &tcp_timer_list; /* Make sure TCP clock is up to date */ calculate_tcp_clock(); lck_mtx_lock(listp->mtx); if (listp->probe_if_index > 0) { tcpstat.tcps_probe_if_conflict++; goto done; } listp->probe_if_index = probe_if_index; if (listp->running) goto done; /* * Reschedule the timerlist to run within the next 10ms, which is * the fastest that we can do. */ offset = TCP_TIMER_10MS_QUANTUM; if (listp->scheduled) { int32_t diff; diff = timer_diff(listp->runtime, 0, tcp_now, offset); if (diff <= 0) { /* The timer will fire sooner than what's needed */ goto done; } } listp->mode = TCP_TIMERLIST_10MS_MODE; listp->idleruns = 0; tcp_sched_timerlist(offset); done: lck_mtx_unlock(listp->mtx); return; } /* * Enable read probes on this connection, if: * - it is in established state * - doesn't have any data outstanding * - the outgoing ifp matches * - we have not already sent any read probes */ static void tcp_enable_read_probe(struct tcpcb *tp, struct ifnet *ifp) { if (tp->t_state == TCPS_ESTABLISHED && tp->snd_max == tp->snd_una && tp->t_inpcb->inp_last_outifp == ifp && !(tp->t_flagsext & TF_DETECT_READSTALL) && tp->t_rtimo_probes == 0) { tp->t_flagsext |= TF_DETECT_READSTALL; tp->t_rtimo_probes = 0; tp->t_timer[TCPT_KEEP] = OFFSET_FROM_START(tp, TCP_TIMER_10MS_QUANTUM); if (tp->tentry.index == TCPT_NONE) { tp->tentry.index = TCPT_KEEP; tp->tentry.runtime = tcp_now + TCP_TIMER_10MS_QUANTUM; } else { int32_t diff = 0; /* Reset runtime to be in next 10ms */ diff = timer_diff(tp->tentry.runtime, 0, tcp_now, TCP_TIMER_10MS_QUANTUM); if (diff > 0) { tp->tentry.index = TCPT_KEEP; tp->tentry.runtime = tcp_now + TCP_TIMER_10MS_QUANTUM; if (tp->tentry.runtime == 0) tp->tentry.runtime++; } } } } /* * Disable read probe and reset the keep alive timer */ static void tcp_disable_read_probe(struct tcpcb *tp) { if (tp->t_adaptive_rtimo == 0 && ((tp->t_flagsext & TF_DETECT_READSTALL) || tp->t_rtimo_probes > 0)) { tcp_keepalive_reset(tp); if (tp->t_mpsub) mptcp_reset_keepalive(tp); } } /* * Reschedule the tcp timerlist in the next 10ms to re-enable read/write * probes on connections going over a particular interface. */ void tcp_probe_connectivity(struct ifnet *ifp, u_int32_t enable) { int32_t offset; struct tcptimerlist *listp = &tcp_timer_list; struct inpcbinfo *pcbinfo = &tcbinfo; struct inpcb *inp, *nxt; if (ifp == NULL) return; /* update clock */ calculate_tcp_clock(); /* * Enable keep alive timer on all connections that are * active/established on this interface. */ lck_rw_lock_shared(pcbinfo->ipi_lock); LIST_FOREACH_SAFE(inp, pcbinfo->ipi_listhead, inp_list, nxt) { struct tcpcb *tp = NULL; if (in_pcb_checkstate(inp, WNT_ACQUIRE, 0) == WNT_STOPUSING) continue; /* Acquire lock to look at the state of the connection */ socket_lock(inp->inp_socket, 1); /* Release the want count */ if (inp->inp_ppcb == NULL || (in_pcb_checkstate(inp, WNT_RELEASE, 1) == WNT_STOPUSING)) { socket_unlock(inp->inp_socket, 1); continue; } tp = intotcpcb(inp); if (enable) tcp_enable_read_probe(tp, ifp); else tcp_disable_read_probe(tp); socket_unlock(inp->inp_socket, 1); } lck_rw_done(pcbinfo->ipi_lock); lck_mtx_lock(listp->mtx); if (listp->running) { listp->pref_mode |= TCP_TIMERLIST_10MS_MODE; goto done; } /* Reschedule within the next 10ms */ offset = TCP_TIMER_10MS_QUANTUM; if (listp->scheduled) { int32_t diff; diff = timer_diff(listp->runtime, 0, tcp_now, offset); if (diff <= 0) { /* The timer will fire sooner than what's needed */ goto done; } } listp->mode = TCP_TIMERLIST_10MS_MODE; listp->idleruns = 0; tcp_sched_timerlist(offset); done: lck_mtx_unlock(listp->mtx); return; } inline void tcp_update_mss_core(struct tcpcb *tp, struct ifnet *ifp) { struct if_cellular_status_v1 *ifsr; u_int32_t optlen; ifsr = &ifp->if_link_status->ifsr_u.ifsr_cell.if_cell_u.if_status_v1; if (ifsr->valid_bitmask & IF_CELL_UL_MSS_RECOMMENDED_VALID) { optlen = tp->t_maxopd - tp->t_maxseg; if (ifsr->mss_recommended == IF_CELL_UL_MSS_RECOMMENDED_NONE && tp->t_cached_maxopd > 0 && tp->t_maxopd < tp->t_cached_maxopd) { tp->t_maxopd = tp->t_cached_maxopd; tcpstat.tcps_mss_to_default++; } else if (ifsr->mss_recommended == IF_CELL_UL_MSS_RECOMMENDED_MEDIUM && tp->t_maxopd > tcp_mss_rec_medium) { tp->t_cached_maxopd = tp->t_maxopd; tp->t_maxopd = tcp_mss_rec_medium; tcpstat.tcps_mss_to_medium++; } else if (ifsr->mss_recommended == IF_CELL_UL_MSS_RECOMMENDED_LOW && tp->t_maxopd > tcp_mss_rec_low) { tp->t_cached_maxopd = tp->t_maxopd; tp->t_maxopd = tcp_mss_rec_low; tcpstat.tcps_mss_to_low++; } tp->t_maxseg = tp->t_maxopd - optlen; /* * clear the cached value if it is same as the current */ if (tp->t_maxopd == tp->t_cached_maxopd) tp->t_cached_maxopd = 0; } } void tcp_update_mss_locked(struct socket *so, struct ifnet *ifp) { struct inpcb *inp = sotoinpcb(so); struct tcpcb *tp = intotcpcb(inp); if (ifp == NULL && (ifp = inp->inp_last_outifp) == NULL) return; if (!IFNET_IS_CELLULAR(ifp)) { /* * This optimization is implemented for cellular * networks only */ return; } if ( tp->t_state <= TCPS_CLOSE_WAIT) { /* * If the connection is currently doing or has done PMTU * blackhole detection, do not change the MSS */ if (tp->t_flags & TF_BLACKHOLE) return; if (ifp->if_link_status == NULL) return; tcp_update_mss_core(tp, ifp); } } void tcp_itimer(struct inpcbinfo *ipi) { struct inpcb *inp, *nxt; if (lck_rw_try_lock_exclusive(ipi->ipi_lock) == FALSE) { if (tcp_itimer_done == TRUE) { tcp_itimer_done = FALSE; atomic_add_32(&ipi->ipi_timer_req.intimer_fast, 1); return; } /* Upgrade failed, lost lock now take it again exclusive */ lck_rw_lock_exclusive(ipi->ipi_lock); } tcp_itimer_done = TRUE; LIST_FOREACH_SAFE(inp, &tcb, inp_list, nxt) { struct socket *so; struct ifnet *ifp; if (inp->inp_ppcb == NULL || in_pcb_checkstate(inp, WNT_ACQUIRE, 0) == WNT_STOPUSING) continue; so = inp->inp_socket; ifp = inp->inp_last_outifp; socket_lock(so, 1); if (in_pcb_checkstate(inp, WNT_RELEASE, 1) == WNT_STOPUSING) { socket_unlock(so, 1); continue; } so_check_extended_bk_idle_time(so); if (ipi->ipi_flags & INPCBINFO_UPDATE_MSS) { tcp_update_mss_locked(so, NULL); } socket_unlock(so, 1); /* * Defunct all system-initiated background sockets if the * socket is using the cellular interface and the interface * has its LQM set to abort. */ if ((ipi->ipi_flags & INPCBINFO_HANDLE_LQM_ABORT) && IS_SO_TC_BACKGROUNDSYSTEM(so->so_traffic_class) && ifp != NULL && IFNET_IS_CELLULAR(ifp) && (ifp->if_interface_state.valid_bitmask & IF_INTERFACE_STATE_LQM_STATE_VALID) && ifp->if_interface_state.lqm_state == IFNET_LQM_THRESH_ABORT) { socket_defunct(current_proc(), so, SHUTDOWN_SOCKET_LEVEL_DISCONNECT_ALL); } } ipi->ipi_flags &= ~(INPCBINFO_UPDATE_MSS | INPCBINFO_HANDLE_LQM_ABORT); lck_rw_done(ipi->ipi_lock); } |