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 | /* * Copyright (c) 1999, 2000, 2003, 2005, 2008, 2012 Apple Inc. All rights reserved. * * @APPLE_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. 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_LICENSE_HEADER_END@ */ #include "internal.h" /* nano_malloc for 64bit ABI */ #if CONFIG_NANOZONE /********************* PROTOTYPES ***********************/ static void nano_statistics(nanozone_t *nanozone, malloc_statistics_t *stats); /********************* VERY LOW LEVEL UTILITIES ************************/ // msg prints after fmt, ... static MALLOC_ALWAYS_INLINE unsigned int nano_mag_index(const nanozone_t *nanozone) { if (os_likely(_os_cpu_number_override == -1)) { return (_os_cpu_number() >> hyper_shift) % nano_common_max_magazines; } return (_os_cpu_number_override >> hyper_shift) % nano_common_max_magazines; } #if NANO_PREALLOCATE_BAND_VM static boolean_t nano_preallocate_band_vm(void) { nano_blk_addr_t u; uintptr_t s, e; u.fields.nano_signature = NANOZONE_SIGNATURE; u.fields.nano_mag_index = 0; u.fields.nano_band = 0; u.fields.nano_slot = 0; u.fields.nano_offset = 0; s = u.addr; // start of first possible band u.fields.nano_mag_index = (1 << NANO_MAG_BITS) - 1; u.fields.nano_band = (1 << NANO_BAND_BITS) - 1; e = u.addr + BAND_SIZE; // end of last possible band return nano_common_allocate_vm_space(s, e - s); } #endif /* * We maintain separate free lists for each (quantized) size. The literature * calls this the "segregated policy". */ static boolean_t segregated_band_grow(nanozone_t *nanozone, nano_meta_admin_t pMeta, size_t slot_bytes, unsigned int mag_index) { nano_blk_addr_t u; // the compiler holds this in a register uintptr_t p, s; size_t watermark, hiwater; if (0 == pMeta->slot_current_base_addr) { // First encounter? u.fields.nano_signature = NANOZONE_SIGNATURE; u.fields.nano_mag_index = mag_index; u.fields.nano_band = 0; u.fields.nano_slot = (slot_bytes >> SHIFT_NANO_QUANTUM) - 1; u.fields.nano_offset = 0; p = u.addr; pMeta->slot_bytes = (unsigned int)slot_bytes; pMeta->slot_objects = SLOT_IN_BAND_SIZE / slot_bytes; } else { p = pMeta->slot_current_base_addr + BAND_SIZE; // Growing, so stride ahead by BAND_SIZE u.addr = (uint64_t)p; if (0 == u.fields.nano_band) { // Did the band index wrap? return FALSE; } assert(slot_bytes == pMeta->slot_bytes); } pMeta->slot_current_base_addr = p; mach_vm_address_t vm_addr = p & ~((uintptr_t)(BAND_SIZE - 1)); // Address of the (2MB) band covering this (128KB) slot if (nanozone->band_max_mapped_baseaddr[mag_index] < vm_addr) { #if !NANO_PREALLOCATE_BAND_VM // Obtain the next band to cover this slot kern_return_t kr = mach_vm_map(mach_task_self(), &vm_addr, BAND_SIZE, 0, VM_MAKE_TAG(VM_MEMORY_MALLOC_NANO), MEMORY_OBJECT_NULL, 0, FALSE, VM_PROT_DEFAULT, VM_PROT_ALL, VM_INHERIT_DEFAULT); void *q = (void *)vm_addr; if (kr || q != (void *)(p & ~((uintptr_t)(BAND_SIZE - 1)))) { // Must get exactly what we asked for if (!kr) { mach_vm_deallocate(mach_task_self(), vm_addr, BAND_SIZE); } return FALSE; } #endif nanozone->band_max_mapped_baseaddr[mag_index] = vm_addr; } // Randomize the starting allocation from this slot (introduces 11 to 14 bits of entropy) if (0 == pMeta->slot_objects_mapped) { // First encounter? pMeta->slot_objects_skipped = (malloc_entropy[1] % (SLOT_IN_BAND_SIZE / slot_bytes)); pMeta->slot_bump_addr = p + (pMeta->slot_objects_skipped * slot_bytes); } else { pMeta->slot_bump_addr = p; } pMeta->slot_limit_addr = p + (SLOT_IN_BAND_SIZE / slot_bytes) * slot_bytes; pMeta->slot_objects_mapped += (SLOT_IN_BAND_SIZE / slot_bytes); u.fields.nano_signature = NANOZONE_SIGNATURE; u.fields.nano_mag_index = mag_index; u.fields.nano_band = 0; u.fields.nano_slot = 0; u.fields.nano_offset = 0; s = u.addr; // Base for this core. // Set the high water mark for this CPU's entire magazine, if this resupply raised it. watermark = nanozone->core_mapped_size[mag_index]; hiwater = MAX(watermark, p - s + SLOT_IN_BAND_SIZE); nanozone->core_mapped_size[mag_index] = hiwater; return TRUE; } static inline unsigned long divrem(unsigned long a, unsigned int b, unsigned int *remainder) { // Encapsulating the modulo and division in an in-lined function convinces the compiler // to issue just a single divide instruction to obtain quotient and remainder. Go figure. *remainder = a % b; return a / b; } static MALLOC_INLINE void * segregated_next_block(nanozone_t *nanozone, nano_meta_admin_t pMeta, size_t slot_bytes, unsigned int mag_index) { while (1) { uintptr_t theLimit = pMeta->slot_limit_addr; // Capture the slot limit that bounds slot_bump_addr right now uintptr_t b = OSAtomicAdd64Barrier(slot_bytes, (volatile int64_t *)&(pMeta->slot_bump_addr)); b -= slot_bytes; // Atomic op returned addr of *next* free block. Subtract to get addr for *this* allocation. if (b < theLimit) { // Did we stay within the bound of the present slot allocation? return (void *)b; // Yep, so the slot_bump_addr this thread incremented is good to go } else { if (pMeta->slot_exhausted) { // exhausted all the bands availble for this slot? pMeta->slot_bump_addr = theLimit; return 0; // We're toast } else { // One thread will grow the heap, others will see its been grown and retry allocation _malloc_lock_lock(&nanozone->band_resupply_lock[mag_index]); // re-check state now that we've taken the lock if (pMeta->slot_exhausted) { _malloc_lock_unlock(&nanozone->band_resupply_lock[mag_index]); return 0; // Toast } else if (b < pMeta->slot_limit_addr) { _malloc_lock_unlock(&nanozone->band_resupply_lock[mag_index]); continue; // ... the slot was successfully grown by first-taker (not us). Now try again. } else if (segregated_band_grow(nanozone, pMeta, slot_bytes, mag_index)) { _malloc_lock_unlock(&nanozone->band_resupply_lock[mag_index]); continue; // ... the slot has been successfully grown by us. Now try again. } else { pMeta->slot_exhausted = TRUE; pMeta->slot_bump_addr = theLimit; _malloc_lock_unlock(&nanozone->band_resupply_lock[mag_index]); return 0; } } } } } static MALLOC_INLINE size_t segregated_size_to_fit(nanozone_t *nanozone, size_t size, size_t *pKey) { size_t k, slot_bytes; if (0 == size) { size = NANO_REGIME_QUANTA_SIZE; // Historical behavior } k = (size + NANO_REGIME_QUANTA_SIZE - 1) >> SHIFT_NANO_QUANTUM; // round up and shift for number of quanta slot_bytes = k << SHIFT_NANO_QUANTUM; // multiply by power of two quanta size *pKey = k - 1; // Zero-based! return slot_bytes; } static MALLOC_INLINE index_t offset_to_index(nanozone_t *nanozone, nano_meta_admin_t pMeta, uintptr_t offset) { unsigned int slot_bytes = pMeta->slot_bytes; unsigned int slot_objects = pMeta->slot_objects; // SLOT_IN_BAND_SIZE / slot_bytes; unsigned int rem; unsigned long quo = divrem(offset, BAND_SIZE, &rem); assert(0 == rem % slot_bytes || pMeta->slot_exhausted); return (index_t)((quo * slot_objects) + (rem / slot_bytes)); } static MALLOC_INLINE uintptr_t index_to_offset(nanozone_t *nanozone, nano_meta_admin_t pMeta, index_t i) { unsigned int slot_bytes = pMeta->slot_bytes; unsigned int slot_objects = pMeta->slot_objects; // SLOT_IN_BAND_SIZE / slot_bytes; unsigned int rem; unsigned long quo = divrem(i, slot_objects, &rem); return (quo * BAND_SIZE) + (rem * slot_bytes); } static kern_return_t segregated_in_use_enumerator(task_t task, void *context, unsigned type_mask, nanozone_t *nanozone, memory_reader_t reader, vm_range_recorder_t recorder) { unsigned int mag_index, slot_key; vm_range_t ptr_range; vm_range_t buffer[MAX_RECORDER_BUFFER]; kern_return_t err; unsigned count = 0; for (mag_index = 0; mag_index < nano_common_max_magazines; mag_index++) { uintptr_t clone_magazine; // magazine base for ourselves nano_blk_addr_t p; // slot base for remote uintptr_t clone_slot_base; // slot base for ourselves (tracks with "p") // Establish p as base address for slot 0 in remote p.fields.nano_signature = NANOZONE_SIGNATURE; p.fields.nano_mag_index = mag_index; p.fields.nano_band = 0; p.fields.nano_slot = 0; p.fields.nano_offset = 0; if (type_mask & MALLOC_PTR_IN_USE_RANGE_TYPE) { mach_vm_address_t vm_addr; mach_vm_size_t alloc_size = nanozone->core_mapped_size[mag_index]; int alloc_flags = VM_FLAGS_ANYWHERE | VM_MAKE_TAG(VM_MEMORY_MALLOC); vm_addr = vm_page_size; kern_return_t kr = mach_vm_allocate(mach_task_self(), &vm_addr, alloc_size, alloc_flags); if (kr) { return kr; } clone_magazine = (uintptr_t)vm_addr; clone_slot_base = clone_magazine; // base for slot 0 in this local magazine } else { clone_slot_base = clone_magazine = 0; // and won't be used in this loop } for (slot_key = 0; slot_key < SLOT_KEY_LIMIT; p.addr += SLOT_IN_BAND_SIZE, // Advance to next slot base for remote clone_slot_base += SLOT_IN_BAND_SIZE, // Advance to next slot base for ourselves slot_key++) { nano_meta_admin_t pMeta = &(nanozone->meta_data[mag_index][slot_key]); size_t slot_objects_mapped = pMeta->slot_objects_mapped; // capture this volatile count if (0 == slot_objects_mapped) { // Nothing allocated in this magazine for this slot? continue; } if (type_mask & MALLOC_ADMIN_REGION_RANGE_TYPE) { /* do NOTHING as there is no distinct admin region */ } if (type_mask & (MALLOC_PTR_REGION_RANGE_TYPE | MALLOC_ADMIN_REGION_RANGE_TYPE)) { nano_blk_addr_t q = p; uintptr_t skip_adj = index_to_offset(nanozone, pMeta, (index_t)pMeta->slot_objects_skipped); while (q.addr < pMeta->slot_limit_addr) { ptr_range.address = q.addr + skip_adj; ptr_range.size = SLOT_IN_BAND_SIZE - skip_adj; skip_adj = 0; recorder(task, context, MALLOC_PTR_REGION_RANGE_TYPE, &ptr_range, 1); q.addr += BAND_SIZE; } } if (type_mask & MALLOC_PTR_IN_USE_RANGE_TYPE) { nano_blk_addr_t q = p; uintptr_t slot_band, clone_slot_band_base = clone_slot_base; uintptr_t skip_adj = index_to_offset(nanozone, pMeta, (index_t)pMeta->slot_objects_skipped); // Copy the bitarray_t denoting madvise()'d pages (if any) into *this* task's address space bitarray_t madv_page_bitarray; int log_page_count; if (pMeta->slot_madvised_pages) { log_page_count = pMeta->slot_madvised_log_page_count; err = reader(task, (vm_address_t)(pMeta->slot_madvised_pages), bitarray_size(log_page_count), (void **)&madv_page_bitarray); if (err) { return err; } } else { madv_page_bitarray = NULL; log_page_count = 0; } while (q.addr < pMeta->slot_limit_addr) { // read slot in each remote band. Lands in some random location. Do not read // parts of the slot that are in madvised pages. if (!madv_page_bitarray) { // Nothing madvised yet - read everything in one go. size_t len = MIN(pMeta->slot_bump_addr - q.addr, SLOT_IN_BAND_SIZE) - skip_adj; err = reader(task, (vm_address_t)(q.addr + skip_adj), len, (void **)&slot_band); if (err) { return err; } // Place the data just read in the correct position relative to the local magazine. memcpy((void *)(clone_slot_band_base + skip_adj), (void *)slot_band, len); } else { // We madvised at least one page. Read only the pages that // have not been madvised. If bitarray_t had operations // like "get next bit set after a given bit" and "find // next unset bit after a given bit", we could do this more // efficiently but given that it doesn't, we have to walk // through each page individually. In practice this is not // much of an issue because this code is only used by // sampling tools and the additional time required is not // really noticeable. size_t len = MIN(pMeta->slot_bump_addr - q.addr, SLOT_IN_BAND_SIZE) - skip_adj; vm_address_t start_addr = (vm_address_t)(q.addr + skip_adj); vm_address_t end_addr = (vm_address_t)(start_addr + len); void *target_addr = (void *)(clone_slot_band_base + skip_adj); for (vm_address_t addr = start_addr; addr < end_addr;) { vm_address_t next_page_addr = trunc_page_kernel(addr + vm_kernel_page_size); size_t read_size = MIN(len, next_page_addr - addr); boolean_t madvised = false; nano_blk_addr_t r; r.addr = addr; index_t pgnum = ((((unsigned)r.fields.nano_band) << NANO_OFFSET_BITS) | ((unsigned)r.fields.nano_offset)) >> vm_kernel_page_shift; unsigned int log_page_count = pMeta->slot_madvised_log_page_count; madvised = (pgnum < (1 << log_page_count)) && bitarray_get(madv_page_bitarray, log_page_count, pgnum); if (!madvised) { // This is not an madvised page - grab the data. err = reader(task, addr, read_size, (void **)&slot_band); if (err) { return err; } // Place the data just read in the correct position relative to the local magazine. memcpy(target_addr, (void *)slot_band, read_size); } else { // This is an madvised page - there should be nothing in here that's // on the freelist, so just write garbage to the target memory. memset(target_addr, (char)0xee, read_size); } addr = next_page_addr; target_addr += read_size; len -= read_size; } } // Simultaneously advance pointers in remote and ourselves to the next band. q.addr += BAND_SIZE; clone_slot_band_base += BAND_SIZE; skip_adj = 0; } // Walk the slot free list and populate a bitarray_t int log_size = 64 - __builtin_clzl(slot_objects_mapped); bitarray_t slot_bitarray = bitarray_create(log_size); if (!slot_bitarray) { return errno; } chained_block_t t; unsigned stoploss = (unsigned)slot_objects_mapped; while ((t = OSAtomicDequeue( &(pMeta->slot_LIFO), offsetof(struct chained_block_s, next) + (clone_slot_base - p.addr)))) { if (0 == stoploss) { malloc_report(ASL_LEVEL_ERR, "Free list walk in segregated_in_use_enumerator exceeded object count.\n"); break; } stoploss--; uintptr_t offset = ((uintptr_t)t - p.addr); // offset from beginning of slot, task-independent index_t block_index = offset_to_index(nanozone, pMeta, offset); if (block_index < slot_objects_mapped) { bitarray_set(slot_bitarray, log_size, block_index); } } // N.B. pMeta->slot_LIFO in *this* task is now drained (remote free list has *not* been disturbed) // Enumerate all the block indices issued to date, and report those not on the free list index_t i; for (i = (index_t)pMeta->slot_objects_skipped; i < slot_objects_mapped; ++i) { uintptr_t block_offset = index_to_offset(nanozone, pMeta, i); if (p.addr + block_offset >= pMeta->slot_bump_addr) { break; } // blocks falling on madvise()'d pages are free! So not enumerated. if (madv_page_bitarray) { nano_blk_addr_t q; index_t pgnum, pgnum_end; q.addr = p.addr + block_offset; pgnum = ((((unsigned)q.fields.nano_band) << NANO_OFFSET_BITS) | ((unsigned)q.fields.nano_offset)) >> vm_kernel_page_shift; q.addr += pMeta->slot_bytes - 1; pgnum_end = ((((unsigned)q.fields.nano_band) << NANO_OFFSET_BITS) | ((unsigned)q.fields.nano_offset)) >> vm_kernel_page_shift; if (pgnum < (1 << log_page_count)) { // bounds check for bitarray_get()'s that follow if (bitarray_get(madv_page_bitarray, log_page_count, pgnum) || bitarray_get(madv_page_bitarray, log_page_count, pgnum_end)) { continue; } } } if (!bitarray_get(slot_bitarray, log_size, i)) { buffer[count].address = p.addr + block_offset; buffer[count].size = (slot_key + 1) << SHIFT_NANO_QUANTUM; count++; if (count >= MAX_RECORDER_BUFFER) { recorder(task, context, MALLOC_PTR_IN_USE_RANGE_TYPE, buffer, count); count = 0; } } } if (count) { recorder(task, context, MALLOC_PTR_IN_USE_RANGE_TYPE, buffer, count); count = 0; } free(slot_bitarray); } } if (clone_magazine) { mach_vm_address_t vm_addr = clone_magazine; mach_vm_size_t alloc_size = nanozone->core_mapped_size[mag_index]; mach_vm_deallocate(mach_task_self(), vm_addr, alloc_size); } } return 0; } /****************** nanozone methods **********************/ /* * These methods are called with "ptr" known to possess the nano signature (from * which we can additionally infer "ptr" is not NULL), and with "size" bounded to * the extent of the nano allocation regime -- (0, 256]. */ static MALLOC_INLINE MALLOC_UNUSED boolean_t _nano_block_inuse_p(nanozone_t *nanozone, const void *ptr) { nano_blk_addr_t p; // happily, the compiler holds this in a register nano_meta_admin_t pMeta; chained_block_t head = NULL, tail = NULL, t; boolean_t inuse = TRUE; p.addr = (uint64_t)ptr; // place ptr on the dissecting table pMeta = &(nanozone->meta_data[p.fields.nano_mag_index][p.fields.nano_slot]); // pop elements off the free list all the while looking for ptr. unsigned stoploss = (unsigned)pMeta->slot_objects_mapped; while ((t = OSAtomicDequeue(&(pMeta->slot_LIFO), offsetof(struct chained_block_s, next)))) { if (0 == stoploss) { malloc_zone_error(nanozone->debug_flags, true, "Free list walk for slot %p in _nano_block_inuse_p exceeded object count.\n", (void *)&(pMeta->slot_LIFO)); } stoploss--; if (NULL == head) { head = t; } else { tail->next = t; } tail = t; if (ptr == t) { inuse = FALSE; break; } } if (tail) { tail->next = NULL; } // push the free list extracted above back onto the LIFO, all at once if (head) { OSAtomicEnqueue(&(pMeta->slot_LIFO), head, (uintptr_t)tail - (uintptr_t)head + offsetof(struct chained_block_s, next)); } return inuse; } static MALLOC_INLINE size_t __nano_vet_and_size_inner(nanozone_t *nanozone, const void *ptr, boolean_t inner) { // Extracts the size of the block in bytes. Checks for a plausible ptr. nano_blk_addr_t p; // the compiler holds this in a register nano_meta_admin_t pMeta; p.addr = (uint64_t)ptr; // Begin the dissection of ptr if (NANOZONE_SIGNATURE != p.fields.nano_signature) { return 0; } if (nano_common_max_magazines <= p.fields.nano_mag_index) { return 0; } if (!inner && p.fields.nano_offset & NANO_QUANTA_MASK) { // stray low-order bits? return 0; } pMeta = &(nanozone->meta_data[p.fields.nano_mag_index][p.fields.nano_slot]); if ((void *)(pMeta->slot_bump_addr) <= ptr) { return 0; // Beyond what's ever been allocated! } if (!inner && ((p.fields.nano_offset % pMeta->slot_bytes) != 0)) { return 0; // Not an exact multiple of the block size for this slot } return pMeta->slot_bytes; } static MALLOC_INLINE size_t __nano_vet_and_size(nanozone_t *nanozone, const void *ptr) { return __nano_vet_and_size_inner(nanozone, ptr, false); } static MALLOC_ALWAYS_INLINE boolean_t _nano_block_has_canary_value(nanozone_t *nanozone, const void *ptr) { return (((chained_block_t)ptr)->double_free_guard ^ nanozone->cookie) == (uintptr_t)ptr; } static MALLOC_ALWAYS_INLINE void _nano_block_set_canary_value(nanozone_t *nanozone, const void *ptr) { ((chained_block_t)ptr)->double_free_guard = ((uintptr_t)ptr) ^ nanozone->cookie; } static MALLOC_INLINE size_t _nano_vet_and_size_of_live(nanozone_t *nanozone, const void *ptr) { size_t size = __nano_vet_and_size(nanozone, ptr); if (0 == size) { // ptr fails sanity check? return 0; } // We have the invariant: If ptr is on a free list, then ptr->double_free_guard is the canary. // So if ptr->double_free_guard is NOT the canary, then ptr is not on a free list, hence is live. if (!_nano_block_has_canary_value(nanozone, ptr)) { return size; // Common case: not on a free list, hence live. Return its size. } else { // confirm that ptr is live despite ptr->double_free_guard having the canary value if (_nano_block_inuse_p(nanozone, ptr)) { return size; // live block that exhibits canary } else { return 0; // ptr wasn't live after all (likely a double free) } } } static MALLOC_INLINE size_t _nano_vet_and_size_of_free(nanozone_t *nanozone, const void *ptr) { size_t size = __nano_vet_and_size(nanozone, ptr); if (0 == size) { // ptr fails sanity check? return 0; } // ptr was just dequed from a free list, so ptr->double_free_guard must have the canary value. if (_nano_block_has_canary_value(nanozone, ptr)) { return size; // return the size of this well formed free block. } else { return 0; // Broken invariant: If ptr is on a free list, then ptr->double_free_guard is the canary. (likely use after free) } } static void * _nano_malloc_check_clear(nanozone_t *nanozone, size_t size, boolean_t cleared_requested) { MALLOC_TRACE(TRACE_nano_malloc, (uintptr_t)nanozone, size, cleared_requested, 0); void *ptr; size_t slot_key; size_t slot_bytes = segregated_size_to_fit(nanozone, size, &slot_key); // Note slot_key is set here mag_index_t mag_index = nano_mag_index(nanozone); nano_meta_admin_t pMeta = &(nanozone->meta_data[mag_index][slot_key]); ptr = OSAtomicDequeue(&(pMeta->slot_LIFO), offsetof(struct chained_block_s, next)); if (ptr) { unsigned debug_flags = nanozone->debug_flags; #if NANO_FREE_DEQUEUE_DILIGENCE size_t gotSize; nano_blk_addr_t p; // the compiler holds this in a register p.addr = (uint64_t)ptr; // Begin the dissection of ptr if (NANOZONE_SIGNATURE != p.fields.nano_signature) { malloc_zone_error(debug_flags, true, "Invalid signature for pointer %p dequeued from free list\n", ptr); } if (mag_index != p.fields.nano_mag_index) { malloc_zone_error(debug_flags, true, "Mismatched magazine for pointer %p dequeued from free list\n", ptr); } gotSize = _nano_vet_and_size_of_free(nanozone, ptr); if (0 == gotSize) { malloc_zone_error(debug_flags, true, "Invalid pointer %p dequeued from free list\n", ptr); } if (gotSize != slot_bytes) { malloc_zone_error(debug_flags, true, "Mismatched size for pointer %p dequeued from free list\n", ptr); } if (!_nano_block_has_canary_value(nanozone, ptr)) { malloc_zone_error(debug_flags, true, "Heap corruption detected, free list canary is damaged for %p\n" "*** Incorrect guard value: %lu\n", ptr, ((chained_block_t)ptr)->double_free_guard); } #if defined(DEBUG) void *next = (void *)(((chained_block_t)ptr)->next); if (next) { p.addr = (uint64_t)next; // Begin the dissection of next if (NANOZONE_SIGNATURE != p.fields.nano_signature) { malloc_zone_error(debug_flags, true, "Invalid next signature for pointer %p dequeued from free " "list, next = %p\n", ptr, "next"); } if (mag_index != p.fields.nano_mag_index) { malloc_zone_error(debug_flags, true, "Mismatched next magazine for pointer %p dequeued from " "free list, next = %p\n", ptr, next); } gotSize = _nano_vet_and_size_of_free(nanozone, next); if (0 == gotSize) { malloc_zone_error(debug_flags, true, "Invalid next for pointer %p dequeued from free list, " "next = %p\n", ptr, next); } if (gotSize != slot_bytes) { malloc_zone_error(debug_flags, true, "Mismatched next size for pointer %p dequeued from free " "list, next = %p\n", ptr, next); } } #endif /* DEBUG */ #endif /* NANO_FREE_DEQUEUE_DILIGENCE */ ((chained_block_t)ptr)->double_free_guard = 0; ((chained_block_t)ptr)->next = NULL; // clear out next pointer to protect free list } else { ptr = segregated_next_block(nanozone, pMeta, slot_bytes, mag_index); } if (cleared_requested && ptr) { memset(ptr, 0, slot_bytes); // TODO: Needs a memory barrier after memset to ensure zeroes land first? } return ptr; } static void * _nano_malloc_check_scribble(nanozone_t *nanozone, size_t size) { void *ptr = _nano_malloc_check_clear(nanozone, size, 0); /* * Scribble on allocated memory when requested. */ if ((nanozone->debug_flags & MALLOC_DO_SCRIBBLE) && ptr && size) { memset(ptr, SCRIBBLE_BYTE, _nano_vet_and_size_of_live(nanozone, ptr)); } return ptr; } static MALLOC_INLINE size_t _nano_size(nanozone_t *nanozone, const void *ptr) { return _nano_vet_and_size_of_live(nanozone, ptr); } static MALLOC_INLINE size_t _nano_good_size(nanozone_t *nanozone, size_t size) { return (size <= NANO_REGIME_QUANTA_SIZE) ? NANO_REGIME_QUANTA_SIZE : (((size + NANO_REGIME_QUANTA_SIZE - 1) >> SHIFT_NANO_QUANTUM) << SHIFT_NANO_QUANTUM); } static MALLOC_INLINE void _nano_free_trusted_size_check_scribble(nanozone_t *nanozone, void *ptr, size_t trusted_size, boolean_t do_scribble) MALLOC_ALWAYS_INLINE; static MALLOC_INLINE void _nano_free_trusted_size_check_scribble(nanozone_t *nanozone, void *ptr, size_t trusted_size, boolean_t do_scribble) { if (trusted_size) { nano_blk_addr_t p; // happily, the compiler holds this in a register nano_meta_admin_t pMeta; if (do_scribble) { (void)memset(ptr, SCRABBLE_BYTE, trusted_size); } _nano_block_set_canary_value(nanozone, ptr); p.addr = (uint64_t)ptr; // place ptr on the dissecting table pMeta = &(nanozone->meta_data[p.fields.nano_mag_index][p.fields.nano_slot]); OSAtomicEnqueue(&(pMeta->slot_LIFO), ptr, offsetof(struct chained_block_s, next)); } else { malloc_zone_error(nanozone->debug_flags, true, "Freeing unallocated pointer %p\n", ptr); } } static MALLOC_INLINE void _nano_free_check_scribble(nanozone_t *nanozone, void *ptr, boolean_t do_scribble) MALLOC_ALWAYS_INLINE; static MALLOC_INLINE void _nano_free_check_scribble(nanozone_t *nanozone, void *ptr, boolean_t do_scribble) { _nano_free_trusted_size_check_scribble(nanozone, ptr, _nano_vet_and_size_of_live(nanozone, ptr), do_scribble); } static MALLOC_INLINE void * _nano_realloc(nanozone_t *nanozone, void *ptr, size_t new_size) { size_t old_size, new_good_size, valid_size; void *new_ptr; if (FALSE && NULL == ptr) { // ptr has our_signature so can't be NULL, but if it were Posix sez ... // If ptr is a null pointer, realloc() shall be equivalent to malloc() for the specified size. return _nano_malloc_check_scribble(nanozone, new_size); } else if (0 == new_size) { // If size is 0 and ptr is not a null pointer, the object pointed to is freed. _nano_free_check_scribble(nanozone, ptr, (nanozone->debug_flags & MALLOC_DO_SCRIBBLE)); // If size is 0, either a null pointer or a unique pointer that can be successfully passed // to free() shall be returned. return _nano_malloc_check_scribble(nanozone, 1); } old_size = _nano_vet_and_size_of_live(nanozone, ptr); if (!old_size) { malloc_zone_error(nanozone->debug_flags, true, "pointer %p being reallocated was not allocated\n", ptr); return NULL; } new_good_size = _nano_good_size(nanozone, new_size); if (new_good_size > old_size) { /* Must grow. FALL THROUGH to alloc/copy/free. */ } else if (new_good_size <= (old_size >> 1)) { /* Serious shrinkage (more than half). FALL THROUGH to alloc/copy/free. */ } else { /* Let's hang on to what we got. */ if (nanozone->debug_flags & MALLOC_DO_SCRIBBLE) { memset(ptr + new_size, SCRIBBLE_BYTE, old_size - new_size); } return ptr; } /* * Allocate a new buffer and copy. */ new_ptr = _nano_malloc_check_scribble(nanozone, new_good_size); if (new_ptr == NULL) { return NULL; } valid_size = MIN(old_size, new_good_size); memcpy(new_ptr, ptr, valid_size); _nano_free_check_scribble(nanozone, ptr, (nanozone->debug_flags & MALLOC_DO_SCRIBBLE)); return new_ptr; } static MALLOC_INLINE void _nano_destroy(nanozone_t *nanozone) { /* Now destroy the separate nanozone region */ nano_common_deallocate_pages((void *)nanozone, NANOZONE_PAGED_SIZE, nanozone->debug_flags); } /****************** nanozone dispatch **********************/ static void * nano_malloc(nanozone_t *nanozone, size_t size) { if (size <= NANO_MAX_SIZE) { void *p = _nano_malloc_check_clear(nanozone, size, 0); if (p) { return p; } else { /* FALLTHROUGH to helper zone */ } } malloc_zone_t *zone = (malloc_zone_t *)(nanozone->helper_zone); return zone->malloc(zone, size); } static void * nano_forked_malloc(nanozone_t *nanozone, size_t size) { malloc_zone_t *zone = (malloc_zone_t *)(nanozone->helper_zone); return zone->malloc(zone, size); } static void * nano_malloc_scribble(nanozone_t *nanozone, size_t size) { if (size <= NANO_MAX_SIZE) { void *ptr = _nano_malloc_check_clear(nanozone, size, 0); if (ptr) { /* * Scribble on allocated memory. */ if (size) { memset(ptr, SCRIBBLE_BYTE, _nano_vet_and_size_of_live(nanozone, ptr)); } return ptr; } else { /* FALLTHROUGH to helper zone */ } } malloc_zone_t *zone = (malloc_zone_t *)(nanozone->helper_zone); return zone->malloc(zone, size); } static void * nano_calloc(nanozone_t *nanozone, size_t num_items, size_t size) { size_t total_bytes; if (calloc_get_size(num_items, size, 0, &total_bytes)) { return NULL; } if (total_bytes <= NANO_MAX_SIZE) { void *p = _nano_malloc_check_clear(nanozone, total_bytes, 1); if (p) { return p; } else { /* FALLTHROUGH to helper zone */ } } malloc_zone_t *zone = (malloc_zone_t *)(nanozone->helper_zone); return zone->calloc(zone, 1, total_bytes); } static void * nano_forked_calloc(nanozone_t *nanozone, size_t num_items, size_t size) { malloc_zone_t *zone = (malloc_zone_t *)(nanozone->helper_zone); return zone->calloc(zone, num_items, size); } static void * nano_valloc(nanozone_t *nanozone, size_t size) { malloc_zone_t *zone = (malloc_zone_t *)(nanozone->helper_zone); return zone->valloc(zone, size); } static MALLOC_INLINE void __nano_free_definite_size(nanozone_t *nanozone, void *ptr, size_t size, boolean_t do_scribble) MALLOC_ALWAYS_INLINE; static MALLOC_INLINE void __nano_free_definite_size(nanozone_t *nanozone, void *ptr, size_t size, boolean_t do_scribble) { nano_blk_addr_t p; // happily, the compiler holds this in a register p.addr = (uint64_t)ptr; // place ptr on the dissecting table if (NANOZONE_SIGNATURE == p.fields.nano_signature) { if (size == ((p.fields.nano_slot + 1) << SHIFT_NANO_QUANTUM)) { // "Trust but verify." _nano_free_trusted_size_check_scribble(nanozone, ptr, size, do_scribble); return; } else { malloc_zone_error(nanozone->debug_flags, true, "Freeing pointer %p whose size was misdeclared\n", ptr); } } else { malloc_zone_t *zone = (malloc_zone_t *)(nanozone->helper_zone); zone->free_definite_size(zone, ptr, size); return; } /* NOTREACHED */ } static void nano_free_definite_size(nanozone_t *nanozone, void *ptr, size_t size) { __nano_free_definite_size(nanozone, ptr, size, 0); } static void nano_free_definite_size_scribble(nanozone_t *nanozone, void *ptr, size_t size) { __nano_free_definite_size(nanozone, ptr, size, 1); } static MALLOC_INLINE void __nano_free(nanozone_t *nanozone, void *ptr, boolean_t do_scribble) MALLOC_ALWAYS_INLINE; static MALLOC_INLINE void __nano_free(nanozone_t *nanozone, void *ptr, boolean_t do_scribble) { MALLOC_TRACE(TRACE_nano_free, (uintptr_t)nanozone, (uintptr_t)ptr, do_scribble, 0); if (!ptr) { return; // Protect against malloc_zone_free() passing NULL. } // <rdar://problem/26481467> exhausting a slot may result in a pointer with // the nanozone prefix being given to nano_free via malloc_zone_free. Calling // vet_and_size here, instead of in _nano_free_check_scribble means we can // early-out into the helper_zone if it turns out nano does not own this ptr. size_t sz = _nano_vet_and_size_of_live(nanozone, ptr); if (sz) { _nano_free_trusted_size_check_scribble(nanozone, ptr, sz, do_scribble); return; } else { malloc_zone_t *zone = (malloc_zone_t *)(nanozone->helper_zone); zone->free(zone, ptr); return; } /* NOTREACHED */ } static void nano_free(nanozone_t *nanozone, void *ptr) { __nano_free(nanozone, ptr, 0); } static void nano_forked_free(nanozone_t *nanozone, void *ptr) { if (!ptr) { return; // Protect against malloc_zone_free() passing NULL. } // <rdar://problem/26481467> exhausting a slot may result in a pointer with // the nanozone prefix being given to nano_free via malloc_zone_free. Calling // vet_and_size here, instead of in _nano_free_check_scribble means we can // early-out into the helper_zone if it turns out nano does not own this ptr. size_t sz = _nano_vet_and_size_of_live(nanozone, ptr); if (sz) { /* NOTHING. Drop it on the floor as nanozone metadata could be fouled by fork. */ return; } else { malloc_zone_t *zone = (malloc_zone_t *)(nanozone->helper_zone); zone->free(zone, ptr); return; } /* NOTREACHED */ } static void nano_forked_free_definite_size(nanozone_t *nanozone, void *ptr, size_t size) { nano_forked_free(nanozone, ptr); } static void nano_free_scribble(nanozone_t *nanozone, void *ptr) { __nano_free(nanozone, ptr, 1); } static size_t nano_size(nanozone_t *nanozone, const void *ptr) { nano_blk_addr_t p; // happily, the compiler holds this in a register p.addr = (uint64_t)ptr; // place ptr on the dissecting table if (NANOZONE_SIGNATURE == p.fields.nano_signature) { // Our signature? return _nano_size(nanozone, ptr); } else { malloc_zone_t *zone = (malloc_zone_t *)(nanozone->helper_zone); return zone->size(zone, ptr); // Not nano. Try other sizes. } /* NOTREACHED */ } static void * nano_realloc(nanozone_t *nanozone, void *ptr, size_t new_size) { // could occur through malloc_zone_realloc() path if (!ptr) { // If ptr is a null pointer, realloc() shall be equivalent to malloc() for the specified size. return nano_malloc(nanozone, new_size); } size_t old_size = _nano_vet_and_size_of_live(nanozone, ptr); if (!old_size) { // not-nano pointer, hand down to helper zone malloc_zone_t *zone = (malloc_zone_t *)(nanozone->helper_zone); return zone->realloc(zone, ptr, new_size); } else { if (new_size <= NANO_MAX_SIZE) { // nano to nano? void *q = _nano_realloc(nanozone, ptr, new_size); if (q) { return q; } else { // nano exhausted /* FALLTHROUGH to helper zone copying case */ } } malloc_zone_t *zone = (malloc_zone_t *)(nanozone->helper_zone); void *new_ptr = zone->malloc(zone, new_size); if (new_ptr) { size_t valid_size = MIN(old_size, new_size); memcpy(new_ptr, ptr, valid_size); _nano_free_check_scribble(nanozone, ptr, (nanozone->debug_flags & MALLOC_DO_SCRIBBLE)); return new_ptr; } else { /* Original ptr is left intact */ return NULL; } /* NOTREACHED */ } /* NOTREACHED */ } static void * nano_forked_realloc(nanozone_t *nanozone, void *ptr, size_t new_size) { // could occur through malloc_zone_realloc() path if (!ptr) { // If ptr is a null pointer, realloc() shall be equivalent to malloc() for the specified size. return nano_forked_malloc(nanozone, new_size); } size_t old_size = _nano_vet_and_size_of_live(nanozone, ptr); if (!old_size) { // not-nano pointer, hand down to helper zone malloc_zone_t *zone = (malloc_zone_t *)(nanozone->helper_zone); return zone->realloc(zone, ptr, new_size); } else { if (0 == new_size) { // If size is 0 and ptr is not a null pointer, the object pointed to is freed. // However as nanozone metadata could be fouled by fork, we'll intentionally leak it. // If size is 0, either a null pointer or a unique pointer that can be successfully passed // to free() shall be returned. return nano_forked_malloc(nanozone, 1); } malloc_zone_t *zone = (malloc_zone_t *)(nanozone->helper_zone); void *new_ptr = zone->malloc(zone, new_size); if (new_ptr) { size_t valid_size = MIN(old_size, new_size); memcpy(new_ptr, ptr, valid_size); /* Original pointer is intentionally leaked as nanozone metadata could be fouled by fork. */ return new_ptr; } else { /* Original ptr is left intact */ return NULL; } /* NOTREACHED */ } /* NOTREACHED */ } static void nano_destroy(nanozone_t *nanozone) { malloc_zone_t *zone = (malloc_zone_t *)(nanozone->helper_zone); zone->destroy(zone); _nano_destroy(nanozone); } static unsigned nano_batch_malloc(nanozone_t *nanozone, size_t size, void **results, unsigned count) { unsigned found = 0; if (size <= NANO_MAX_SIZE) { while (found < count) { void *ptr = _nano_malloc_check_clear(nanozone, size, 0); if (!ptr) { break; } *results++ = ptr; found++; } if (found == count) { return found; } else { /* FALLTHROUGH to mop-up in the helper zone */ } } malloc_zone_t *zone = (malloc_zone_t *)(nanozone->helper_zone); return found + zone->batch_malloc(zone, size, results, count - found); } static unsigned nano_forked_batch_malloc(nanozone_t *nanozone, size_t size, void **results, unsigned count) { malloc_zone_t *zone = (malloc_zone_t *)(nanozone->helper_zone); return zone->batch_malloc(zone, size, results, count); } static void nano_batch_free(nanozone_t *nanozone, void **to_be_freed, unsigned count) { void *ptr; // frees all the pointers in to_be_freed // note that to_be_freed may be overwritten during the process if (!count) { return; } while (count--) { ptr = to_be_freed[count]; if (ptr) { nano_free(nanozone, ptr); } } } static void nano_forked_batch_free(nanozone_t *nanozone, void **to_be_freed, unsigned count) { void *ptr; // frees all the pointers in to_be_freed // note that to_be_freed may be overwritten during the process if (!count) { return; } while (count--) { ptr = to_be_freed[count]; if (ptr) { nano_forked_free(nanozone, ptr); } } } static void * nano_memalign(nanozone_t *nanozone, size_t alignment, size_t size) { malloc_zone_t *zone = (malloc_zone_t *)(nanozone->helper_zone); return zone->memalign(zone, alignment, size); } static boolean_t nano_claimed_address(nanozone_t *nanozone, void *ptr) { nano_blk_addr_t p; p.addr = (uint64_t)ptr; if (NANOZONE_SIGNATURE != p.fields.nano_signature) { // Not a nano address - let the helper zone handle it. malloc_zone_t *helper_zone = nanozone->helper_zone; return malloc_zone_claimed_address(helper_zone, ptr); } return __nano_vet_and_size_inner(nanozone, ptr, true) != 0; } static boolean_t nano_forked_claimed_address(struct _malloc_zone_t *zone, void *ptr) { // This does not operate after fork - default to true to avoid // false negatives. return true; } static size_t nano_try_madvise(nanozone_t *nanozone, size_t goal) { unsigned int mag_index, slot_key; size_t bytes_toward_goal = 0; for (mag_index = 0; mag_index < nano_common_max_magazines; mag_index++) { nano_blk_addr_t p; // Establish p as base address for band 0, slot 0, offset 0 p.fields.nano_signature = NANOZONE_SIGNATURE; p.fields.nano_mag_index = mag_index; p.fields.nano_band = 0; p.fields.nano_slot = 0; p.fields.nano_offset = 0; for (slot_key = 0; slot_key < SLOT_KEY_LIMIT; p.addr += SLOT_IN_BAND_SIZE, // Advance to next slot base slot_key++) { // malloc_report(ASL_LEVEL_WARNING,"nano_try_madvise examining slot base %p\n", p.addr); nano_meta_admin_t pMeta = &(nanozone->meta_data[mag_index][slot_key]); uintptr_t slot_bump_addr = pMeta->slot_bump_addr; // capture this volatile pointer size_t slot_objects_mapped = pMeta->slot_objects_mapped; // capture this volatile count if (0 == slot_objects_mapped) { // Nothing allocated in this magazine for this slot? continue; } else { // Walk the slot free list and populate a bitarray_t int log_size = 64 - __builtin_clzl(slot_objects_mapped); bitarray_t slot_bitarray = bitarray_create(log_size); unsigned int slot_bytes = pMeta->slot_bytes; int log_page_count = 64 - __builtin_clzl((slot_objects_mapped * slot_bytes) / vm_kernel_page_size); log_page_count = 1 + MAX(0, log_page_count); bitarray_t page_bitarray = bitarray_create(log_page_count); // malloc_report(ASL_LEVEL_WARNING,"slot_bitarray: %db page_bitarray: %db\n", bitarray_size(log_size), // bitarray_size(log_page_count)); if (!slot_bitarray) { malloc_report(ASL_LEVEL_ERR, "bitarray_create(%d) in nano_try_madvise returned errno=%d.\n", log_size, errno); free(page_bitarray); return bytes_toward_goal; } if (!page_bitarray) { malloc_report(ASL_LEVEL_ERR, "bitarray_create(%d) in nano_try_madvise returned errno=%d.\n", log_page_count, errno); free(slot_bitarray); return bytes_toward_goal; } chained_block_t head = NULL, tail = NULL, t; unsigned stoploss = (unsigned)slot_objects_mapped; while ((t = OSAtomicDequeue(&(pMeta->slot_LIFO), offsetof(struct chained_block_s, next)))) { if (0 == stoploss) { malloc_report(ASL_LEVEL_ERR, "Free list walk in nano_try_madvise exceeded object count.\n"); break; } stoploss--; uintptr_t offset = ((uintptr_t)t - p.addr); // offset from beginning of slot index_t block_index = offset_to_index(nanozone, pMeta, offset); // build a simple linked list of the free blocks we're able to obtain if (NULL == head) { head = t; } else { tail->next = t; } tail = t; // take note in a bitarray_t of each free block we're able to obtain (allows fast lookup below) if (block_index < slot_objects_mapped) { bitarray_set(slot_bitarray, log_size, block_index); } } if (tail) { tail->next = NULL; } if (NULL == head) { free(slot_bitarray); free(page_bitarray); continue; } index_t i; nano_blk_addr_t q; size_t pgnum; for (i = (index_t)pMeta->slot_objects_skipped; i < slot_objects_mapped; ++i) { uintptr_t block_offset = index_to_offset(nanozone, pMeta, i); if (p.addr + block_offset >= slot_bump_addr) { break; } if (!bitarray_get(slot_bitarray, log_size, i)) { // is block i allocated or already on an madvise'd page? // Mark the page(s) it resides on as live q.addr = p.addr + block_offset; pgnum = ((((unsigned)q.fields.nano_band) << NANO_OFFSET_BITS) | ((unsigned)q.fields.nano_offset)) >> vm_kernel_page_shift; bitarray_set(page_bitarray, log_page_count, (index_t)pgnum); q.addr += slot_bytes - 1; pgnum = ((((unsigned)q.fields.nano_band) << NANO_OFFSET_BITS) | ((unsigned)q.fields.nano_offset)) >> vm_kernel_page_shift; bitarray_set(page_bitarray, log_page_count, (index_t)pgnum); } } free(slot_bitarray); q.addr = p.addr + index_to_offset(nanozone, pMeta, (index_t)pMeta->slot_objects_skipped); index_t pgstart = ((((unsigned)q.fields.nano_band) << NANO_OFFSET_BITS) | ((unsigned)q.fields.nano_offset)) >> vm_kernel_page_shift; q.addr = slot_bump_addr - slot_bytes; pgnum = ((((unsigned)q.fields.nano_band) << NANO_OFFSET_BITS) | ((unsigned)q.fields.nano_offset)) >> vm_kernel_page_shift; // malloc_report(ASL_LEVEL_WARNING,"Examining %d pages. Slot base %p.\n", pgnum - pgstart + 1, p.addr); if (pMeta->slot_madvised_pages) { if (pMeta->slot_madvised_log_page_count < log_page_count) { bitarray_t new_madvised_pages = bitarray_create(log_page_count); index_t index; while (bitarray_zap_first_set(pMeta->slot_madvised_pages, pMeta->slot_madvised_log_page_count, &index)) { bitarray_set(new_madvised_pages, log_page_count, index); } free(pMeta->slot_madvised_pages); pMeta->slot_madvised_pages = new_madvised_pages; pMeta->slot_madvised_log_page_count = log_page_count; } } else { pMeta->slot_madvised_pages = bitarray_create(log_page_count); pMeta->slot_madvised_log_page_count = log_page_count; } bitarray_t will_madvise_pages = bitarray_create(log_page_count); int num_advised = 0; for (i = pgstart; i < pgnum; ++i) { if ((i < (1 << log_page_count)) && // bounds check for the bitarray_get()'s that follow. !bitarray_get(pMeta->slot_madvised_pages, log_page_count, i) && // already madvise'd? !bitarray_get(page_bitarray, log_page_count, i)) // no live allocations? { num_advised++; bitarray_set(will_madvise_pages, log_page_count, i); } } free(page_bitarray); if (num_advised) { chained_block_t new_head = NULL, new_tail = NULL; // malloc_report(ASL_LEVEL_WARNING,"Constructing residual free list starting at %p num_advised %d\n", head, // num_advised); t = head; while (t) { q.addr = (uintptr_t)t; index_t pgnum_start = ((((unsigned)q.fields.nano_band) << NANO_OFFSET_BITS) | ((unsigned)q.fields.nano_offset)) >> vm_kernel_page_shift; q.addr += slot_bytes - 1; index_t pgnum_end = ((((unsigned)q.fields.nano_band) << NANO_OFFSET_BITS) | ((unsigned)q.fields.nano_offset)) >> vm_kernel_page_shift; // bounds check for the bitarray_get()'s that follow. If the pgnum is beyond the // capacity of the will_madvise_pages just restore the block to the free list. if (pgnum_start >= (1 << log_page_count)) { if (NULL == new_head) { new_head = t; } else { new_tail->next = t; } new_tail = t; } // If the block nowhere lies on an madvise()'d page restore it to the slot free list. else if (!bitarray_get(will_madvise_pages, log_page_count, pgnum_start) && !bitarray_get(will_madvise_pages, log_page_count, pgnum_end)) { if (NULL == new_head) { new_head = t; } else { new_tail->next = t; } new_tail = t; } t = t->next; } if (new_tail) { new_tail->next = NULL; } // push the free list extracted above back onto the LIFO, all at once if (new_head) { OSAtomicEnqueue(&(pMeta->slot_LIFO), new_head, (uintptr_t)new_tail - (uintptr_t)new_head + offsetof(struct chained_block_s, next)); } } else { // malloc_report(ASL_LEVEL_WARNING,"Reinstating free list since no pages were madvised (%d).\n", num_advised); if (head) { OSAtomicEnqueue(&(pMeta->slot_LIFO), head, (uintptr_t)tail - (uintptr_t)head + offsetof(struct chained_block_s, next)); } } for (i = pgstart; i < pgnum; ++i) { if ((i < (1 << log_page_count)) && bitarray_get(will_madvise_pages, log_page_count, i)) { q = p; q.fields.nano_band = (i << vm_kernel_page_shift) >> NANO_OFFSET_BITS; q.fields.nano_offset = (i << vm_kernel_page_shift) & ((1 << NANO_OFFSET_BITS) - 1); // malloc_report(ASL_LEVEL_WARNING,"Entire page non-live: %d. Slot base %p, madvising %p\n", i, p.addr, // q.addr); if (nanozone->debug_flags & MALLOC_DO_SCRIBBLE) { memset((void *)q.addr, SCRUBBLE_BYTE, vm_kernel_page_size); } if (-1 == madvise((void *)q.addr, vm_kernel_page_size, MADV_FREE_REUSABLE)) { /* -1 return: VM map entry change makes this unfit for reuse. Something evil lurks. */ #if DEBUG_MADVISE nanozone_error(nanozone, 0, "madvise(..., MADV_FREE_REUSABLE) failed", (void *)cwq.addrpgLo, "length=%d\n", vm_page_size); #endif } else { bytes_toward_goal += vm_kernel_page_size; bitarray_set(pMeta->slot_madvised_pages, log_page_count, i); } } } free(will_madvise_pages); if (!bitarray_first_set(pMeta->slot_madvised_pages, log_page_count)) { free(pMeta->slot_madvised_pages); pMeta->slot_madvised_pages = NULL; pMeta->slot_madvised_log_page_count = 0; } if (goal && bytes_toward_goal >= goal) { return bytes_toward_goal; } } } } return bytes_toward_goal; } static size_t nano_pressure_relief(nanozone_t *nanozone, size_t goal) { MAGMALLOC_PRESSURERELIEFBEGIN((void *)nanozone, nanozone->basic_zone.zone_name, (int)goal); MALLOC_TRACE(TRACE_nano_memory_pressure | DBG_FUNC_START, (uint64_t)nanozone, goal, 0, 0); size_t total = nano_try_madvise(nanozone, goal); MAGMALLOC_PRESSURERELIEFEND((void *)nanozone, nanozone->basic_zone.zone_name, (int)goal, (int)total); MALLOC_TRACE(TRACE_nano_memory_pressure | DBG_FUNC_END, (uint64_t)nanozone, goal, total, 0); return total; } /**************** introspection methods *********************/ static kern_return_t nano_ptr_in_use_enumerator(task_t task, void *context, unsigned type_mask, vm_address_t zone_address, memory_reader_t reader, vm_range_recorder_t recorder) { nanozone_t *nanozone; kern_return_t err; struct nanozone_s zone_copy; if (!reader) { reader = nano_common_default_reader; } err = reader(task, zone_address, sizeof(nanozone_t), (void **)&nanozone); if (err) { return err; } memcpy(&zone_copy, nanozone, sizeof(zone_copy)); err = segregated_in_use_enumerator(task, context, type_mask, &zone_copy, reader, recorder); return err; } static size_t nano_good_size(nanozone_t *nanozone, size_t size) { if (size <= NANO_MAX_SIZE) { return _nano_common_good_size(size); } else { malloc_zone_t *zone = (malloc_zone_t *)(nanozone->helper_zone); return zone->introspect->good_size(zone, size); } } // TODO sanity checks static unsigned nanozone_check_counter = 0; static unsigned nanozone_check_start = 0; static unsigned nanozone_check_modulo = 1; static boolean_t nano_check_all(nanozone_t *nanozone, const char *function) { return 1; } static boolean_t nanozone_check(nanozone_t *nanozone) { if ((++nanozone_check_counter % 10000) == 0) { malloc_report(ASL_LEVEL_NOTICE, "at nanozone_check counter=%d\n", nanozone_check_counter); } if (nanozone_check_counter < nanozone_check_start) { return 1; } if (nanozone_check_counter % nanozone_check_modulo) { return 1; } return nano_check_all(nanozone, ""); } static unsigned count_free(nanozone_t *nanozone, nano_meta_admin_t pMeta) { chained_block_t head = NULL, tail = NULL, t; unsigned count = 0; unsigned stoploss = (unsigned)pMeta->slot_objects_mapped; while ((t = OSAtomicDequeue(&(pMeta->slot_LIFO), offsetof(struct chained_block_s, next)))) { if (0 == stoploss) { malloc_zone_error(nanozone->debug_flags, true, "Free list walk in count_free exceeded object count.\n", (void *)&(pMeta->slot_LIFO), NULL); } stoploss--; if (NULL == head) { head = t; } else { tail->next = t; } tail = t; count++; } if (tail) { tail->next = NULL; } // push the free list extracted above back onto the LIFO, all at once if (head) { OSAtomicEnqueue(&(pMeta->slot_LIFO), head, (uintptr_t)tail - (uintptr_t)head + offsetof(struct chained_block_s, next)); } return count; } static void nano_print(nanozone_t *nanozone, boolean_t verbose) { unsigned int mag_index, slot_key; malloc_statistics_t stats; nano_statistics(nanozone, &stats); malloc_report(MALLOC_REPORT_NOLOG | MALLOC_REPORT_NOPREFIX, "Nanozone %p: inUse=%d(%lluKB) touched=%lluKB allocated=%lluMB\n", nanozone, stats.blocks_in_use, (uint64_t)stats.size_in_use >> 10, (uint64_t)stats.max_size_in_use >> 10, (uint64_t)stats.size_allocated >> 20); for (mag_index = 0; mag_index < nano_common_max_magazines; mag_index++) { nano_blk_addr_t p; // Establish p as base address for band 0, slot 0, offset 0 p.fields.nano_signature = NANOZONE_SIGNATURE; p.fields.nano_mag_index = mag_index; p.fields.nano_band = 0; p.fields.nano_slot = 0; p.fields.nano_offset = 0; for (slot_key = 0; slot_key < SLOT_KEY_LIMIT; p.addr += SLOT_IN_BAND_SIZE, // Advance to next slot base slot_key++) { nano_meta_admin_t pMeta = &(nanozone->meta_data[mag_index][slot_key]); uintptr_t slot_bump_addr = pMeta->slot_bump_addr; // capture this volatile pointer size_t slot_objects_mapped = pMeta->slot_objects_mapped; // capture this volatile count if (0 == slot_objects_mapped) { // Nothing allocated in this magazine for this slot? malloc_report(MALLOC_REPORT_NOLOG | MALLOC_REPORT_NOPREFIX, "Magazine %2d(%3d) Unrealized\n", mag_index, (slot_key + 1) << SHIFT_NANO_QUANTUM); continue; } uintptr_t offset = (0 == slot_bump_addr ? 0 : slot_bump_addr - p.addr); unsigned blocks_touched = offset_to_index(nanozone, pMeta, offset) - (unsigned)pMeta->slot_objects_skipped; unsigned blocks_now_free = count_free(nanozone, pMeta); unsigned blocks_in_use = blocks_touched - blocks_now_free; size_t size_hiwater = ((slot_key + 1) << SHIFT_NANO_QUANTUM) * blocks_touched; size_t size_in_use = ((slot_key + 1) << SHIFT_NANO_QUANTUM) * blocks_in_use; size_t size_allocated = ((offset / BAND_SIZE) + 1) * SLOT_IN_BAND_SIZE; malloc_report(MALLOC_REPORT_NOLOG | MALLOC_REPORT_NOPREFIX, "Magazine %2d(%3d) [%p, %3lluKB] \t Allocations in use=%4d \t Bytes in use=%llub \t Untouched=%lluKB\n", mag_index, (slot_key + 1) << SHIFT_NANO_QUANTUM, (void *)p.addr, (uint64_t)(size_allocated >> 10), blocks_in_use, (uint64_t)size_in_use, (uint64_t)((size_allocated - size_hiwater) >> 10)); if (!verbose) { continue; } else { // Walk the slot free list and populate a bitarray_t int log_size = 64 - __builtin_clzl(slot_objects_mapped); bitarray_t slot_bitarray = bitarray_create(log_size); if (!slot_bitarray) { malloc_report(ASL_LEVEL_ERR, "bitarray_create(%d) in nano_print returned errno=%d.\n", log_size, errno); return; } chained_block_t head = NULL, tail = NULL, t; unsigned stoploss = (unsigned)slot_objects_mapped; while ((t = OSAtomicDequeue(&(pMeta->slot_LIFO), offsetof(struct chained_block_s, next)))) { if (0 == stoploss) { malloc_report(ASL_LEVEL_ERR, "Free list walk in nano_print exceeded object count.\n"); break; } stoploss--; uintptr_t offset = ((uintptr_t)t - p.addr); // offset from beginning of slot index_t block_index = offset_to_index(nanozone, pMeta, offset); if (NULL == head) { head = t; } else { tail->next = t; } tail = t; if (block_index < slot_objects_mapped) { bitarray_set(slot_bitarray, log_size, block_index); } } if (tail) { tail->next = NULL; } index_t i; for (i = 0; i < slot_objects_mapped; ++i) { nano_blk_addr_t q; size_t pgnum; uintptr_t block_offset = index_to_offset(nanozone, pMeta, i); if (p.addr + block_offset >= slot_bump_addr) { break; } q.addr = p.addr + block_offset; pgnum = ((((unsigned)q.fields.nano_band) << NANO_OFFSET_BITS) | ((unsigned)q.fields.nano_offset)) >> vm_kernel_page_shift; if (i < pMeta->slot_objects_skipped) { malloc_report(MALLOC_REPORT_NOLOG | MALLOC_REPORT_NOPREFIX, "_"); } else if (bitarray_get(slot_bitarray, log_size, i)) { malloc_report(MALLOC_REPORT_NOLOG | MALLOC_REPORT_NOPREFIX, "F"); } else if (pMeta->slot_madvised_pages && (pgnum < (1 << pMeta->slot_madvised_log_page_count)) && bitarray_get(pMeta->slot_madvised_pages, pMeta->slot_madvised_log_page_count, (index_t)pgnum)) { malloc_report(MALLOC_REPORT_NOLOG | MALLOC_REPORT_NOPREFIX, "M"); } else { malloc_report(MALLOC_REPORT_NOLOG | MALLOC_REPORT_NOPREFIX, "."); } } malloc_report(MALLOC_REPORT_NOLOG | MALLOC_REPORT_NOPREFIX, "\n"); free(slot_bitarray); // push the free list extracted above back onto the LIFO, all at once if (head) { OSAtomicEnqueue( &(pMeta->slot_LIFO), head, (uintptr_t)tail - (uintptr_t)head + offsetof(struct chained_block_s, next)); } } } } return; } static void nano_log(malloc_zone_t *zone, void *log_address) { } static void nano_force_lock(nanozone_t *nanozone) { int i; for (i = 0; i < nano_common_max_magazines; ++i) { _malloc_lock_lock(&nanozone->band_resupply_lock[i]); } } static void nano_force_unlock(nanozone_t *nanozone) { int i; for (i = 0; i < nano_common_max_magazines; ++i) { _malloc_lock_unlock(&nanozone->band_resupply_lock[i]); } } static void nano_reinit_lock(nanozone_t *nanozone) { int i; for (i = 0; i < nano_common_max_magazines; ++i) { _malloc_lock_init(&nanozone->band_resupply_lock[i]); } } static void nano_statistics(nanozone_t *nanozone, malloc_statistics_t *stats) { int i, j; bzero(stats, sizeof(*stats)); for (i = 0; i < nano_common_max_magazines; ++i) { nano_blk_addr_t p; // Establish p as base address for slot 0 in this CPU magazine p.fields.nano_signature = NANOZONE_SIGNATURE; p.fields.nano_mag_index = i; p.fields.nano_band = 0; p.fields.nano_slot = 0; p.fields.nano_offset = 0; for (j = 0; j < NANO_SLOT_SIZE; p.addr += SLOT_IN_BAND_SIZE, // Advance to next slot base ++j) { nano_meta_admin_t pMeta = &nanozone->meta_data[i][j]; uintptr_t offset = pMeta->slot_bump_addr - p.addr; if (0 == pMeta->slot_current_base_addr) { // Nothing allocated in this magazine for this slot? continue; } else { unsigned blocks_touched = offset_to_index(nanozone, pMeta, offset) - (unsigned)pMeta->slot_objects_skipped; unsigned blocks_now_free = count_free(nanozone, pMeta); unsigned blocks_in_use = blocks_touched - blocks_now_free; size_t size_hiwater = ((j + 1) << SHIFT_NANO_QUANTUM) * blocks_touched; size_t size_in_use = ((j + 1) << SHIFT_NANO_QUANTUM) * blocks_in_use; size_t size_allocated = ((offset / BAND_SIZE) + 1) * SLOT_IN_BAND_SIZE; stats->blocks_in_use += blocks_in_use; stats->max_size_in_use += size_hiwater; stats->size_in_use += size_in_use; stats->size_allocated += size_allocated; } } } } static boolean_t _nano_locked(nanozone_t *nanozone) { int i; for (i = 0; i < nano_common_max_magazines; ++i) { if (_malloc_lock_trylock(&nanozone->band_resupply_lock[i])) { _malloc_lock_unlock(&nanozone->band_resupply_lock[i]); return TRUE; } } return FALSE; } static boolean_t nano_locked(nanozone_t *nanozone) { malloc_zone_t *zone = (malloc_zone_t *)(nanozone->helper_zone); return _nano_locked(nanozone) || zone->introspect->zone_locked(zone); } static const struct malloc_introspection_t nano_introspect = { (void *)nano_ptr_in_use_enumerator, (void *)nano_good_size, (void *)nanozone_check, (void *)nano_print, nano_log, (void *)nano_force_lock, (void *)nano_force_unlock, (void *)nano_statistics, (void *)nano_locked, NULL, NULL, NULL, NULL, /* Zone enumeration version 7 and forward. */ (void *)nano_reinit_lock, // reinit_lock version 9 and foward }; // marked as const to spare the DATA section void nano_forked_zone(nanozone_t *nanozone) { /* * Hobble the nano zone in the child of a fork prior to an exec since * the state of the zone can be made inconsistent by a parent thread while the * fork is underway. * All new allocations will be referred to the helper zone (which is more stable.) * All free()'s of existing nano objects will be leaked. */ mprotect(nanozone, sizeof(nanozone->basic_zone), PROT_READ | PROT_WRITE); nanozone->basic_zone.size = (void *)nano_size; /* Unchanged. */ nanozone->basic_zone.malloc = (void *)nano_forked_malloc; nanozone->basic_zone.calloc = (void *)nano_forked_calloc; nanozone->basic_zone.valloc = (void *)nano_valloc; /* Unchanged, already always obtained from helper zone. */ nanozone->basic_zone.free = (void *)nano_forked_free; nanozone->basic_zone.realloc = (void *)nano_forked_realloc; nanozone->basic_zone.destroy = (void *)nano_destroy; /* Unchanged. */ nanozone->basic_zone.batch_malloc = (void *)nano_forked_batch_malloc; nanozone->basic_zone.batch_free = (void *)nano_forked_batch_free; nanozone->basic_zone.introspect = (struct malloc_introspection_t *)&nano_introspect; /* Unchanged. */ nanozone->basic_zone.memalign = (void *)nano_memalign; /* Unchanged. */ nanozone->basic_zone.free_definite_size = (void *)nano_forked_free_definite_size; nanozone->basic_zone.claimed_address = nano_forked_claimed_address; mprotect(nanozone, sizeof(nanozone->basic_zone), PROT_READ); } malloc_zone_t * nano_create_zone(malloc_zone_t *helper_zone, unsigned debug_flags) { nanozone_t *nanozone; int i, j; /* Note: It is important that nano_create_zone resets _malloc_engaged_nano * if it is unable to enable the nanozone (and chooses not to abort). As * several functions rely on _malloc_engaged_nano to determine if they * should manipulate the nanozone, and these should not run if we failed * to create the zone. */ MALLOC_ASSERT(_malloc_engaged_nano == NANO_V1); /* get memory for the zone. */ nanozone = nano_common_allocate_based_pages(NANOZONE_PAGED_SIZE, 0, 0, VM_MEMORY_MALLOC, 0); if (!nanozone) { _malloc_engaged_nano = NANO_NONE; return NULL; } /* set up the basic_zone portion of the nanozone structure */ nanozone->basic_zone.version = 10; nanozone->basic_zone.size = (void *)nano_size; nanozone->basic_zone.malloc = (debug_flags & MALLOC_DO_SCRIBBLE) ? (void *)nano_malloc_scribble : (void *)nano_malloc; nanozone->basic_zone.calloc = (void *)nano_calloc; nanozone->basic_zone.valloc = (void *)nano_valloc; nanozone->basic_zone.free = (debug_flags & MALLOC_DO_SCRIBBLE) ? (void *)nano_free_scribble : (void *)nano_free; nanozone->basic_zone.realloc = (void *)nano_realloc; nanozone->basic_zone.destroy = (void *)nano_destroy; nanozone->basic_zone.batch_malloc = (void *)nano_batch_malloc; nanozone->basic_zone.batch_free = (void *)nano_batch_free; nanozone->basic_zone.introspect = (struct malloc_introspection_t *)&nano_introspect; nanozone->basic_zone.memalign = (void *)nano_memalign; nanozone->basic_zone.free_definite_size = (debug_flags & MALLOC_DO_SCRIBBLE) ? (void *)nano_free_definite_size_scribble : (void *)nano_free_definite_size; nanozone->basic_zone.pressure_relief = (void *)nano_pressure_relief; nanozone->basic_zone.claimed_address = (void *)nano_claimed_address; nanozone->basic_zone.reserved1 = 0; /* Set to zero once and for all as required by CFAllocator. */ nanozone->basic_zone.reserved2 = 0; /* Set to zero once and for all as required by CFAllocator. */ mprotect(nanozone, sizeof(nanozone->basic_zone), PROT_READ); /* Prevent overwriting the function pointers in basic_zone. */ /* Nano zone does not support MALLOC_ADD_GUARD_PAGES. */ if (debug_flags & MALLOC_ADD_GUARD_PAGES) { malloc_report(ASL_LEVEL_INFO, "nano zone does not support guard pages\n"); debug_flags &= ~MALLOC_ADD_GUARD_PAGES; } /* set up the remainder of the nanozone structure */ nanozone->debug_flags = debug_flags; if (phys_ncpus > sizeof(nanozone->core_mapped_size) / sizeof(nanozone->core_mapped_size[0])) { MALLOC_REPORT_FATAL_ERROR(phys_ncpus, "nanozone abandoned because NCPUS > max magazines.\n"); } /* Initialize slot queue heads and resupply locks. */ OSQueueHead q0 = OS_ATOMIC_QUEUE_INIT; for (i = 0; i < nano_common_max_magazines; ++i) { _malloc_lock_init(&nanozone->band_resupply_lock[i]); for (j = 0; j < NANO_SLOT_SIZE; ++j) { nanozone->meta_data[i][j].slot_LIFO = q0; } } /* Initialize the security token. */ nanozone->cookie = (uintptr_t)malloc_entropy[0] & 0x0000ffffffff0000ULL; // scramble central 32bits with this cookie nanozone->helper_zone = helper_zone; return (malloc_zone_t *)nanozone; } void nano_init(const char *envp[], const char *apple[], const char *bootargs MALLOC_UNUSED) { #if NANO_PREALLOCATE_BAND_VM // Unconditionally preallocate the VA space set aside for nano malloc to // reserve it in all configurations. rdar://problem/33392283 boolean_t preallocated = nano_preallocate_band_vm(); if (!preallocated) { malloc_report(ASL_LEVEL_NOTICE, "nano zone abandoned due to inability to preallocate reserved vm space.\n"); _malloc_engaged_nano = NANO_NONE; } #endif } // Second phase of initialization, called during _malloc_initialize(), after // environment variables have been read and processed. void nano_configure() { // Nothing to do. } #endif // CONFIG_NANOZONE /* vim: set noet:ts=4:sw=4:cindent: */ |