jfs_xtree.c 92.1 KB
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 2703 2704 2705 2706 2707 2708 2709 2710 2711 2712 2713 2714 2715 2716 2717 2718 2719 2720 2721 2722 2723 2724 2725 2726 2727 2728 2729 2730 2731 2732 2733 2734 2735 2736 2737 2738 2739 2740 2741 2742 2743 2744 2745 2746 2747 2748 2749 2750 2751 2752 2753 2754 2755 2756 2757 2758 2759 2760 2761 2762 2763 2764 2765 2766 2767 2768 2769 2770 2771 2772 2773 2774 2775 2776 2777 2778 2779 2780 2781 2782 2783 2784 2785 2786 2787 2788 2789 2790 2791 2792 2793 2794 2795 2796 2797 2798 2799 2800 2801 2802 2803 2804 2805 2806 2807 2808 2809 2810 2811 2812 2813 2814 2815 2816 2817 2818 2819 2820 2821 2822 2823 2824 2825 2826 2827 2828 2829 2830 2831 2832 2833 2834 2835 2836 2837 2838 2839 2840 2841 2842 2843 2844 2845 2846 2847 2848 2849 2850 2851 2852 2853 2854 2855 2856 2857 2858 2859 2860 2861 2862 2863 2864 2865 2866 2867 2868 2869 2870 2871 2872 2873 2874 2875 2876 2877 2878 2879 2880 2881 2882 2883 2884 2885 2886 2887 2888 2889 2890 2891 2892 2893 2894 2895 2896 2897 2898 2899 2900 2901 2902 2903 2904 2905 2906 2907 2908 2909 2910 2911 2912 2913 2914 2915 2916 2917 2918 2919 2920 2921 2922 2923 2924 2925 2926 2927 2928 2929 2930 2931 2932 2933 2934 2935 2936 2937 2938 2939 2940 2941 2942 2943 2944 2945 2946 2947 2948 2949 2950 2951 2952 2953 2954 2955 2956 2957 2958 2959 2960 2961 2962 2963 2964 2965 2966 2967 2968 2969 2970 2971 2972 2973 2974 2975 2976 2977 2978 2979 2980 2981 2982 2983 2984 2985 2986 2987 2988 2989 2990 2991 2992 2993 2994 2995 2996 2997 2998 2999 3000 3001 3002 3003 3004 3005 3006 3007 3008 3009 3010 3011 3012 3013 3014 3015 3016 3017 3018 3019 3020 3021 3022 3023 3024 3025 3026 3027 3028 3029 3030 3031 3032 3033 3034 3035 3036 3037 3038 3039 3040 3041 3042 3043 3044 3045 3046 3047 3048 3049 3050 3051 3052 3053 3054 3055 3056 3057 3058 3059 3060 3061 3062 3063 3064 3065 3066 3067 3068 3069 3070 3071 3072 3073 3074 3075 3076 3077 3078 3079 3080 3081 3082 3083 3084 3085 3086 3087 3088 3089 3090 3091 3092 3093 3094 3095 3096 3097 3098 3099 3100 3101 3102 3103 3104 3105 3106 3107 3108 3109 3110 3111 3112 3113 3114 3115 3116 3117 3118 3119 3120 3121 3122 3123 3124 3125 3126 3127 3128 3129 3130 3131 3132 3133 3134 3135 3136 3137 3138 3139 3140 3141 3142 3143 3144 3145 3146 3147 3148 3149 3150 3151 3152 3153 3154 3155 3156 3157 3158 3159 3160 3161 3162 3163 3164 3165 3166 3167 3168 3169 3170 3171 3172 3173 3174 3175 3176 3177 3178 3179 3180 3181 3182 3183 3184 3185 3186 3187 3188 3189 3190 3191 3192 3193 3194 3195 3196 3197 3198 3199 3200 3201 3202 3203 3204 3205 3206 3207 3208 3209 3210 3211 3212 3213 3214 3215 3216 3217 3218 3219 3220 3221 3222 3223 3224 3225 3226 3227 3228 3229 3230 3231 3232 3233 3234 3235 3236 3237 3238 3239 3240 3241 3242 3243 3244 3245 3246 3247 3248 3249 3250 3251 3252 3253 3254 3255 3256 3257 3258 3259 3260 3261 3262 3263 3264 3265 3266 3267 3268 3269 3270 3271 3272 3273 3274 3275 3276 3277 3278 3279 3280 3281 3282 3283 3284 3285 3286 3287 3288 3289 3290 3291 3292 3293 3294 3295 3296 3297 3298 3299 3300 3301 3302 3303 3304 3305 3306 3307 3308 3309 3310 3311 3312 3313 3314 3315 3316 3317 3318 3319 3320 3321 3322 3323 3324 3325 3326 3327 3328 3329 3330 3331 3332 3333 3334 3335 3336 3337 3338 3339 3340 3341 3342 3343 3344 3345 3346 3347 3348 3349 3350 3351 3352 3353 3354 3355 3356 3357 3358 3359 3360 3361 3362 3363 3364 3365 3366 3367 3368 3369 3370 3371 3372 3373 3374 3375 3376 3377 3378 3379 3380 3381 3382 3383 3384 3385 3386 3387 3388 3389 3390 3391 3392 3393 3394 3395 3396 3397 3398 3399 3400 3401 3402 3403 3404 3405 3406 3407 3408 3409 3410 3411 3412 3413 3414 3415 3416 3417 3418 3419 3420 3421 3422 3423 3424 3425 3426 3427 3428 3429 3430 3431 3432 3433 3434 3435 3436 3437 3438 3439 3440 3441 3442 3443 3444 3445 3446 3447 3448 3449 3450 3451 3452 3453 3454 3455 3456 3457 3458 3459 3460 3461 3462 3463 3464 3465 3466 3467 3468 3469 3470 3471 3472 3473 3474 3475 3476 3477 3478 3479 3480 3481 3482 3483 3484 3485 3486 3487 3488 3489 3490 3491 3492 3493 3494 3495 3496 3497 3498 3499 3500 3501 3502 3503 3504 3505 3506 3507 3508 3509 3510 3511 3512 3513 3514 3515 3516 3517 3518 3519 3520 3521 3522 3523 3524 3525 3526 3527 3528 3529 3530 3531 3532 3533 3534 3535 3536 3537 3538 3539 3540 3541 3542 3543 3544 3545 3546 3547 3548 3549 3550 3551 3552 3553 3554 3555 3556 3557 3558 3559 3560 3561 3562 3563 3564 3565 3566 3567 3568 3569 3570 3571 3572 3573 3574 3575 3576 3577 3578 3579 3580 3581 3582 3583 3584 3585 3586 3587 3588 3589 3590 3591 3592 3593 3594 3595 3596 3597 3598 3599 3600 3601 3602 3603 3604 3605 3606 3607 3608 3609 3610 3611 3612 3613 3614 3615 3616 3617 3618 3619 3620 3621 3622 3623 3624 3625 3626 3627 3628 3629 3630 3631 3632 3633 3634 3635 3636 3637 3638 3639 3640 3641 3642 3643 3644 3645 3646 3647 3648 3649 3650 3651 3652 3653 3654 3655 3656 3657 3658 3659 3660 3661 3662 3663 3664 3665 3666 3667 3668 3669 3670 3671 3672 3673 3674 3675 3676 3677 3678 3679 3680 3681 3682 3683 3684 3685 3686 3687 3688 3689 3690 3691 3692 3693 3694 3695 3696 3697 3698 3699 3700 3701 3702 3703 3704 3705 3706 3707 3708 3709 3710 3711 3712 3713 3714 3715 3716 3717 3718 3719 3720 3721 3722 3723 3724 3725 3726 3727 3728 3729 3730 3731 3732 3733 3734 3735 3736 3737 3738 3739 3740 3741 3742 3743 3744 3745 3746 3747 3748 3749 3750 3751 3752 3753 3754 3755 3756 3757 3758 3759 3760 3761 3762 3763 3764 3765 3766 3767 3768 3769 3770 3771 3772 3773 3774 3775 3776 3777 3778 3779 3780 3781 3782 3783 3784 3785 3786 3787 3788 3789 3790 3791 3792 3793 3794 3795 3796 3797 3798 3799 3800 3801 3802 3803 3804 3805 3806 3807 3808 3809 3810 3811 3812 3813 3814 3815 3816 3817 3818 3819 3820 3821 3822 3823 3824 3825 3826 3827 3828 3829 3830 3831 3832 3833 3834 3835 3836 3837 3838 3839 3840 3841 3842 3843 3844 3845 3846 3847 3848 3849 3850 3851 3852 3853 3854 3855 3856 3857 3858 3859 3860 3861 3862 3863 3864 3865 3866 3867 3868 3869 3870 3871 3872 3873 3874 3875 3876 3877 3878 3879 3880 3881 3882 3883 3884 3885 3886 3887 3888 3889 3890 3891 3892 3893 3894 3895 3896 3897 3898 3899 3900 3901 3902 3903
/*
 *   Copyright (C) International Business Machines Corp., 2000-2005
 *
 *   This program is free software;  you can redistribute it and/or modify
 *   it under the terms of the GNU General Public License as published by
 *   the Free Software Foundation; either version 2 of the License, or
 *   (at your option) any later version.
 *
 *   This program is distributed in the hope that it will be useful,
 *   but WITHOUT ANY WARRANTY;  without even the implied warranty of
 *   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See
 *   the GNU General Public License for more details.
 *
 *   You should have received a copy of the GNU General Public License
 *   along with this program;  if not, write to the Free Software
 *   Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
 */
/*
 *	jfs_xtree.c: extent allocation descriptor B+-tree manager
 */

#include <linux/fs.h>
#include <linux/module.h>
#include <linux/quotaops.h>
#include <linux/seq_file.h>
#include "jfs_incore.h"
#include "jfs_filsys.h"
#include "jfs_metapage.h"
#include "jfs_dmap.h"
#include "jfs_dinode.h"
#include "jfs_superblock.h"
#include "jfs_debug.h"

/*
 * xtree local flag
 */
#define XT_INSERT	0x00000001

/*
 *	xtree key/entry comparison: extent offset
 *
 * return:
 *	-1: k < start of extent
 *	 0: start_of_extent <= k <= end_of_extent
 *	 1: k > end_of_extent
 */
#define XT_CMP(CMP, K, X, OFFSET64)\
{\
	OFFSET64 = offsetXAD(X);\
	(CMP) = ((K) >= OFFSET64 + lengthXAD(X)) ? 1 :\
		((K) < OFFSET64) ? -1 : 0;\
}

/* write a xad entry */
#define XT_PUTENTRY(XAD, FLAG, OFF, LEN, ADDR)\
{\
	(XAD)->flag = (FLAG);\
	XADoffset((XAD), (OFF));\
	XADlength((XAD), (LEN));\
	XADaddress((XAD), (ADDR));\
}

#define XT_PAGE(IP, MP) BT_PAGE(IP, MP, xtpage_t, i_xtroot)

/* get page buffer for specified block address */
/* ToDo: Replace this ugly macro with a function */
#define XT_GETPAGE(IP, BN, MP, SIZE, P, RC)				\
do {									\
	BT_GETPAGE(IP, BN, MP, xtpage_t, SIZE, P, RC, i_xtroot);	\
	if (!(RC)) {							\
		if ((le16_to_cpu((P)->header.nextindex) < XTENTRYSTART) || \
		    (le16_to_cpu((P)->header.nextindex) >		\
		     le16_to_cpu((P)->header.maxentry)) ||		\
		    (le16_to_cpu((P)->header.maxentry) >		\
		     (((BN) == 0) ? XTROOTMAXSLOT : PSIZE >> L2XTSLOTSIZE))) { \
			jfs_error((IP)->i_sb,				\
				  "XT_GETPAGE: xtree page corrupt\n");	\
			BT_PUTPAGE(MP);					\
			MP = NULL;					\
			RC = -EIO;					\
		}							\
	}								\
} while (0)

/* for consistency */
#define XT_PUTPAGE(MP) BT_PUTPAGE(MP)

#define XT_GETSEARCH(IP, LEAF, BN, MP, P, INDEX) \
	BT_GETSEARCH(IP, LEAF, BN, MP, xtpage_t, P, INDEX, i_xtroot)
/* xtree entry parameter descriptor */
struct xtsplit {
	struct metapage *mp;
	s16 index;
	u8 flag;
	s64 off;
	s64 addr;
	int len;
	struct pxdlist *pxdlist;
};


/*
 *	statistics
 */
#ifdef CONFIG_JFS_STATISTICS
static struct {
	uint search;
	uint fastSearch;
	uint split;
} xtStat;
#endif


/*
 * forward references
 */
static int xtSearch(struct inode *ip, s64 xoff, s64 *next, int *cmpp,
		    struct btstack * btstack, int flag);

static int xtSplitUp(tid_t tid,
		     struct inode *ip,
		     struct xtsplit * split, struct btstack * btstack);

static int xtSplitPage(tid_t tid, struct inode *ip, struct xtsplit * split,
		       struct metapage ** rmpp, s64 * rbnp);

static int xtSplitRoot(tid_t tid, struct inode *ip,
		       struct xtsplit * split, struct metapage ** rmpp);

#ifdef _STILL_TO_PORT
static int xtDeleteUp(tid_t tid, struct inode *ip, struct metapage * fmp,
		      xtpage_t * fp, struct btstack * btstack);

static int xtSearchNode(struct inode *ip,
			xad_t * xad,
			int *cmpp, struct btstack * btstack, int flag);

static int xtRelink(tid_t tid, struct inode *ip, xtpage_t * fp);
#endif				/*  _STILL_TO_PORT */

/*
 *	xtLookup()
 *
 * function: map a single page into a physical extent;
 */
int xtLookup(struct inode *ip, s64 lstart,
	     s64 llen, int *pflag, s64 * paddr, s32 * plen, int no_check)
{
	int rc = 0;
	struct btstack btstack;
	int cmp;
	s64 bn;
	struct metapage *mp;
	xtpage_t *p;
	int index;
	xad_t *xad;
	s64 next, size, xoff, xend;
	int xlen;
	s64 xaddr;

	*paddr = 0;
	*plen = llen;

	if (!no_check) {
		/* is lookup offset beyond eof ? */
		size = ((u64) ip->i_size + (JFS_SBI(ip->i_sb)->bsize - 1)) >>
		    JFS_SBI(ip->i_sb)->l2bsize;
		if (lstart >= size)
			return 0;
	}

	/*
	 * search for the xad entry covering the logical extent
	 */
//search:
	if ((rc = xtSearch(ip, lstart, &next, &cmp, &btstack, 0))) {
		jfs_err("xtLookup: xtSearch returned %d", rc);
		return rc;
	}

	/*
	 *	compute the physical extent covering logical extent
	 *
	 * N.B. search may have failed (e.g., hole in sparse file),
	 * and returned the index of the next entry.
	 */
	/* retrieve search result */
	XT_GETSEARCH(ip, btstack.top, bn, mp, p, index);

	/* is xad found covering start of logical extent ?
	 * lstart is a page start address,
	 * i.e., lstart cannot start in a hole;
	 */
	if (cmp) {
		if (next)
			*plen = min(next - lstart, llen);
		goto out;
	}

	/*
	 * lxd covered by xad
	 */
	xad = &p->xad[index];
	xoff = offsetXAD(xad);
	xlen = lengthXAD(xad);
	xend = xoff + xlen;
	xaddr = addressXAD(xad);

	/* initialize new pxd */
	*pflag = xad->flag;
	*paddr = xaddr + (lstart - xoff);
	/* a page must be fully covered by an xad */
	*plen = min(xend - lstart, llen);

      out:
	XT_PUTPAGE(mp);

	return rc;
}

/*
 *	xtSearch()
 *
 * function:	search for the xad entry covering specified offset.
 *
 * parameters:
 *	ip	- file object;
 *	xoff	- extent offset;
 *	nextp	- address of next extent (if any) for search miss
 *	cmpp	- comparison result:
 *	btstack - traverse stack;
 *	flag	- search process flag (XT_INSERT);
 *
 * returns:
 *	btstack contains (bn, index) of search path traversed to the entry.
 *	*cmpp is set to result of comparison with the entry returned.
 *	the page containing the entry is pinned at exit.
 */
static int xtSearch(struct inode *ip, s64 xoff,	s64 *nextp,
		    int *cmpp, struct btstack * btstack, int flag)
{
	struct jfs_inode_info *jfs_ip = JFS_IP(ip);
	int rc = 0;
	int cmp = 1;		/* init for empty page */
	s64 bn;			/* block number */
	struct metapage *mp;	/* page buffer */
	xtpage_t *p;		/* page */
	xad_t *xad;
	int base, index, lim, btindex;
	struct btframe *btsp;
	int nsplit = 0;		/* number of pages to split */
	s64 t64;
	s64 next = 0;

	INCREMENT(xtStat.search);

	BT_CLR(btstack);

	btstack->nsplit = 0;

	/*
	 *	search down tree from root:
	 *
	 * between two consecutive entries of <Ki, Pi> and <Kj, Pj> of
	 * internal page, child page Pi contains entry with k, Ki <= K < Kj.
	 *
	 * if entry with search key K is not found
	 * internal page search find the entry with largest key Ki
	 * less than K which point to the child page to search;
	 * leaf page search find the entry with smallest key Kj
	 * greater than K so that the returned index is the position of
	 * the entry to be shifted right for insertion of new entry.
	 * for empty tree, search key is greater than any key of the tree.
	 *
	 * by convention, root bn = 0.
	 */
	for (bn = 0;;) {
		/* get/pin the page to search */
		XT_GETPAGE(ip, bn, mp, PSIZE, p, rc);
		if (rc)
			return rc;

		/* try sequential access heuristics with the previous
		 * access entry in target leaf page:
		 * once search narrowed down into the target leaf,
		 * key must either match an entry in the leaf or
		 * key entry does not exist in the tree;
		 */
//fastSearch:
		if ((jfs_ip->btorder & BT_SEQUENTIAL) &&
		    (p->header.flag & BT_LEAF) &&
		    (index = jfs_ip->btindex) <
		    le16_to_cpu(p->header.nextindex)) {
			xad = &p->xad[index];
			t64 = offsetXAD(xad);
			if (xoff < t64 + lengthXAD(xad)) {
				if (xoff >= t64) {
					*cmpp = 0;
					goto out;
				}

				/* stop sequential access heuristics */
				goto binarySearch;
			} else {	/* (t64 + lengthXAD(xad)) <= xoff */

				/* try next sequential entry */
				index++;
				if (index <
				    le16_to_cpu(p->header.nextindex)) {
					xad++;
					t64 = offsetXAD(xad);
					if (xoff < t64 + lengthXAD(xad)) {
						if (xoff >= t64) {
							*cmpp = 0;
							goto out;
						}

						/* miss: key falls between
						 * previous and this entry
						 */
						*cmpp = 1;
						next = t64;
						goto out;
					}

					/* (xoff >= t64 + lengthXAD(xad));
					 * matching entry may be further out:
					 * stop heuristic search
					 */
					/* stop sequential access heuristics */
					goto binarySearch;
				}

				/* (index == p->header.nextindex);
				 * miss: key entry does not exist in
				 * the target leaf/tree
				 */
				*cmpp = 1;
				goto out;
			}

			/*
			 * if hit, return index of the entry found, and
			 * if miss, where new entry with search key is
			 * to be inserted;
			 */
		      out:
			/* compute number of pages to split */
			if (flag & XT_INSERT) {
				if (p->header.nextindex ==	/* little-endian */
				    p->header.maxentry)
					nsplit++;
				else
					nsplit = 0;
				btstack->nsplit = nsplit;
			}

			/* save search result */
			btsp = btstack->top;
			btsp->bn = bn;
			btsp->index = index;
			btsp->mp = mp;

			/* update sequential access heuristics */
			jfs_ip->btindex = index;

			if (nextp)
				*nextp = next;

			INCREMENT(xtStat.fastSearch);
			return 0;
		}

		/* well, ... full search now */
	      binarySearch:
		lim = le16_to_cpu(p->header.nextindex) - XTENTRYSTART;

		/*
		 * binary search with search key K on the current page
		 */
		for (base = XTENTRYSTART; lim; lim >>= 1) {
			index = base + (lim >> 1);

			XT_CMP(cmp, xoff, &p->xad[index], t64);
			if (cmp == 0) {
				/*
				 *	search hit
				 */
				/* search hit - leaf page:
				 * return the entry found
				 */
				if (p->header.flag & BT_LEAF) {
					*cmpp = cmp;

					/* compute number of pages to split */
					if (flag & XT_INSERT) {
						if (p->header.nextindex ==
						    p->header.maxentry)
							nsplit++;
						else
							nsplit = 0;
						btstack->nsplit = nsplit;
					}

					/* save search result */
					btsp = btstack->top;
					btsp->bn = bn;
					btsp->index = index;
					btsp->mp = mp;

					/* init sequential access heuristics */
					btindex = jfs_ip->btindex;
					if (index == btindex ||
					    index == btindex + 1)
						jfs_ip->btorder = BT_SEQUENTIAL;
					else
						jfs_ip->btorder = BT_RANDOM;
					jfs_ip->btindex = index;

					return 0;
				}
				/* search hit - internal page:
				 * descend/search its child page
				 */
				if (index < le16_to_cpu(p->header.nextindex)-1)
					next = offsetXAD(&p->xad[index + 1]);
				goto next;
			}

			if (cmp > 0) {
				base = index + 1;
				--lim;
			}
		}

		/*
		 *	search miss
		 *
		 * base is the smallest index with key (Kj) greater than
		 * search key (K) and may be zero or maxentry index.
		 */
		if (base < le16_to_cpu(p->header.nextindex))
			next = offsetXAD(&p->xad[base]);
		/*
		 * search miss - leaf page:
		 *
		 * return location of entry (base) where new entry with
		 * search key K is to be inserted.
		 */
		if (p->header.flag & BT_LEAF) {
			*cmpp = cmp;

			/* compute number of pages to split */
			if (flag & XT_INSERT) {
				if (p->header.nextindex ==
				    p->header.maxentry)
					nsplit++;
				else
					nsplit = 0;
				btstack->nsplit = nsplit;
			}

			/* save search result */
			btsp = btstack->top;
			btsp->bn = bn;
			btsp->index = base;
			btsp->mp = mp;

			/* init sequential access heuristics */
			btindex = jfs_ip->btindex;
			if (base == btindex || base == btindex + 1)
				jfs_ip->btorder = BT_SEQUENTIAL;
			else
				jfs_ip->btorder = BT_RANDOM;
			jfs_ip->btindex = base;

			if (nextp)
				*nextp = next;

			return 0;
		}

		/*
		 * search miss - non-leaf page:
		 *
		 * if base is non-zero, decrement base by one to get the parent
		 * entry of the child page to search.
		 */
		index = base ? base - 1 : base;

		/*
		 * go down to child page
		 */
	      next:
		/* update number of pages to split */
		if (p->header.nextindex == p->header.maxentry)
			nsplit++;
		else
			nsplit = 0;

		/* push (bn, index) of the parent page/entry */
		if (BT_STACK_FULL(btstack)) {
			jfs_error(ip->i_sb, "stack overrun!\n");
			XT_PUTPAGE(mp);
			return -EIO;
		}
		BT_PUSH(btstack, bn, index);

		/* get the child page block number */
		bn = addressXAD(&p->xad[index]);

		/* unpin the parent page */
		XT_PUTPAGE(mp);
	}
}

/*
 *	xtInsert()
 *
 * function:
 *
 * parameter:
 *	tid	- transaction id;
 *	ip	- file object;
 *	xflag	- extent flag (XAD_NOTRECORDED):
 *	xoff	- extent offset;
 *	xlen	- extent length;
 *	xaddrp	- extent address pointer (in/out):
 *		if (*xaddrp)
 *			caller allocated data extent at *xaddrp;
 *		else
 *			allocate data extent and return its xaddr;
 *	flag	-
 *
 * return:
 */
int xtInsert(tid_t tid,		/* transaction id */
	     struct inode *ip, int xflag, s64 xoff, s32 xlen, s64 * xaddrp,
	     int flag)
{
	int rc = 0;
	s64 xaddr, hint;
	struct metapage *mp;	/* meta-page buffer */
	xtpage_t *p;		/* base B+-tree index page */
	s64 bn;
	int index, nextindex;
	struct btstack btstack;	/* traverse stack */
	struct xtsplit split;	/* split information */
	xad_t *xad;
	int cmp;
	s64 next;
	struct tlock *tlck;
	struct xtlock *xtlck;

	jfs_info("xtInsert: nxoff:0x%lx nxlen:0x%x", (ulong) xoff, xlen);

	/*
	 *	search for the entry location at which to insert:
	 *
	 * xtFastSearch() and xtSearch() both returns (leaf page
	 * pinned, index at which to insert).
	 * n.b. xtSearch() may return index of maxentry of
	 * the full page.
	 */
	if ((rc = xtSearch(ip, xoff, &next, &cmp, &btstack, XT_INSERT)))
		return rc;

	/* retrieve search result */
	XT_GETSEARCH(ip, btstack.top, bn, mp, p, index);

	/* This test must follow XT_GETSEARCH since mp must be valid if
	 * we branch to out: */
	if ((cmp == 0) || (next && (xlen > next - xoff))) {
		rc = -EEXIST;
		goto out;
	}

	/*
	 * allocate data extent requested
	 *
	 * allocation hint: last xad
	 */
	if ((xaddr = *xaddrp) == 0) {
		if (index > XTENTRYSTART) {
			xad = &p->xad[index - 1];
			hint = addressXAD(xad) + lengthXAD(xad) - 1;
		} else
			hint = 0;
		if ((rc = dquot_alloc_block(ip, xlen)))
			goto out;
		if ((rc = dbAlloc(ip, hint, (s64) xlen, &xaddr))) {
			dquot_free_block(ip, xlen);
			goto out;
		}
	}

	/*
	 *	insert entry for new extent
	 */
	xflag |= XAD_NEW;

	/*
	 *	if the leaf page is full, split the page and
	 *	propagate up the router entry for the new page from split
	 *
	 * The xtSplitUp() will insert the entry and unpin the leaf page.
	 */
	nextindex = le16_to_cpu(p->header.nextindex);
	if (nextindex == le16_to_cpu(p->header.maxentry)) {
		split.mp = mp;
		split.index = index;
		split.flag = xflag;
		split.off = xoff;
		split.len = xlen;
		split.addr = xaddr;
		split.pxdlist = NULL;
		if ((rc = xtSplitUp(tid, ip, &split, &btstack))) {
			/* undo data extent allocation */
			if (*xaddrp == 0) {
				dbFree(ip, xaddr, (s64) xlen);
				dquot_free_block(ip, xlen);
			}
			return rc;
		}

		*xaddrp = xaddr;
		return 0;
	}

	/*
	 *	insert the new entry into the leaf page
	 */
	/*
	 * acquire a transaction lock on the leaf page;
	 *
	 * action: xad insertion/extension;
	 */
	BT_MARK_DIRTY(mp, ip);

	/* if insert into middle, shift right remaining entries. */
	if (index < nextindex)
		memmove(&p->xad[index + 1], &p->xad[index],
			(nextindex - index) * sizeof(xad_t));

	/* insert the new entry: mark the entry NEW */
	xad = &p->xad[index];
	XT_PUTENTRY(xad, xflag, xoff, xlen, xaddr);

	/* advance next available entry index */
	le16_add_cpu(&p->header.nextindex, 1);

	/* Don't log it if there are no links to the file */
	if (!test_cflag(COMMIT_Nolink, ip)) {
		tlck = txLock(tid, ip, mp, tlckXTREE | tlckGROW);
		xtlck = (struct xtlock *) & tlck->lock;
		xtlck->lwm.offset =
		    (xtlck->lwm.offset) ? min(index,
					      (int)xtlck->lwm.offset) : index;
		xtlck->lwm.length =
		    le16_to_cpu(p->header.nextindex) - xtlck->lwm.offset;
	}

	*xaddrp = xaddr;

      out:
	/* unpin the leaf page */
	XT_PUTPAGE(mp);

	return rc;
}


/*
 *	xtSplitUp()
 *
 * function:
 *	split full pages as propagating insertion up the tree
 *
 * parameter:
 *	tid	- transaction id;
 *	ip	- file object;
 *	split	- entry parameter descriptor;
 *	btstack - traverse stack from xtSearch()
 *
 * return:
 */
static int
xtSplitUp(tid_t tid,
	  struct inode *ip, struct xtsplit * split, struct btstack * btstack)
{
	int rc = 0;
	struct metapage *smp;
	xtpage_t *sp;		/* split page */
	struct metapage *rmp;
	s64 rbn;		/* new right page block number */
	struct metapage *rcmp;
	xtpage_t *rcp;		/* right child page */
	s64 rcbn;		/* right child page block number */
	int skip;		/* index of entry of insertion */
	int nextindex;		/* next available entry index of p */
	struct btframe *parent;	/* parent page entry on traverse stack */
	xad_t *xad;
	s64 xaddr;
	int xlen;
	int nsplit;		/* number of pages split */
	struct pxdlist pxdlist;
	pxd_t *pxd;
	struct tlock *tlck;
	struct xtlock *xtlck;

	smp = split->mp;
	sp = XT_PAGE(ip, smp);

	/* is inode xtree root extension/inline EA area free ? */
	if ((sp->header.flag & BT_ROOT) && (!S_ISDIR(ip->i_mode)) &&
	    (le16_to_cpu(sp->header.maxentry) < XTROOTMAXSLOT) &&
	    (JFS_IP(ip)->mode2 & INLINEEA)) {
		sp->header.maxentry = cpu_to_le16(XTROOTMAXSLOT);
		JFS_IP(ip)->mode2 &= ~INLINEEA;

		BT_MARK_DIRTY(smp, ip);
		/*
		 * acquire a transaction lock on the leaf page;
		 *
		 * action: xad insertion/extension;
		 */

		/* if insert into middle, shift right remaining entries. */
		skip = split->index;
		nextindex = le16_to_cpu(sp->header.nextindex);
		if (skip < nextindex)
			memmove(&sp->xad[skip + 1], &sp->xad[skip],
				(nextindex - skip) * sizeof(xad_t));

		/* insert the new entry: mark the entry NEW */
		xad = &sp->xad[skip];
		XT_PUTENTRY(xad, split->flag, split->off, split->len,
			    split->addr);

		/* advance next available entry index */
		le16_add_cpu(&sp->header.nextindex, 1);

		/* Don't log it if there are no links to the file */
		if (!test_cflag(COMMIT_Nolink, ip)) {
			tlck = txLock(tid, ip, smp, tlckXTREE | tlckGROW);
			xtlck = (struct xtlock *) & tlck->lock;
			xtlck->lwm.offset = (xtlck->lwm.offset) ?
			    min(skip, (int)xtlck->lwm.offset) : skip;
			xtlck->lwm.length =
			    le16_to_cpu(sp->header.nextindex) -
			    xtlck->lwm.offset;
		}

		return 0;
	}

	/*
	 * allocate new index blocks to cover index page split(s)
	 *
	 * allocation hint: ?
	 */
	if (split->pxdlist == NULL) {
		nsplit = btstack->nsplit;
		split->pxdlist = &pxdlist;
		pxdlist.maxnpxd = pxdlist.npxd = 0;
		pxd = &pxdlist.pxd[0];
		xlen = JFS_SBI(ip->i_sb)->nbperpage;
		for (; nsplit > 0; nsplit--, pxd++) {
			if ((rc = dbAlloc(ip, (s64) 0, (s64) xlen, &xaddr))
			    == 0) {
				PXDaddress(pxd, xaddr);
				PXDlength(pxd, xlen);

				pxdlist.maxnpxd++;

				continue;
			}

			/* undo allocation */

			XT_PUTPAGE(smp);
			return rc;
		}
	}

	/*
	 * Split leaf page <sp> into <sp> and a new right page <rp>.
	 *
	 * The split routines insert the new entry into the leaf page,
	 * and acquire txLock as appropriate.
	 * return <rp> pinned and its block number <rpbn>.
	 */
	rc = (sp->header.flag & BT_ROOT) ?
	    xtSplitRoot(tid, ip, split, &rmp) :
	    xtSplitPage(tid, ip, split, &rmp, &rbn);

	XT_PUTPAGE(smp);

	if (rc)
		return -EIO;
	/*
	 * propagate up the router entry for the leaf page just split
	 *
	 * insert a router entry for the new page into the parent page,
	 * propagate the insert/split up the tree by walking back the stack
	 * of (bn of parent page, index of child page entry in parent page)
	 * that were traversed during the search for the page that split.
	 *
	 * the propagation of insert/split up the tree stops if the root
	 * splits or the page inserted into doesn't have to split to hold
	 * the new entry.
	 *
	 * the parent entry for the split page remains the same, and
	 * a new entry is inserted at its right with the first key and
	 * block number of the new right page.
	 *
	 * There are a maximum of 3 pages pinned at any time:
	 * right child, left parent and right parent (when the parent splits)
	 * to keep the child page pinned while working on the parent.
	 * make sure that all pins are released at exit.
	 */
	while ((parent = BT_POP(btstack)) != NULL) {
		/* parent page specified by stack frame <parent> */

		/* keep current child pages <rcp> pinned */
		rcmp = rmp;
		rcbn = rbn;
		rcp = XT_PAGE(ip, rcmp);

		/*
		 * insert router entry in parent for new right child page <rp>
		 */
		/* get/pin the parent page <sp> */
		XT_GETPAGE(ip, parent->bn, smp, PSIZE, sp, rc);
		if (rc) {
			XT_PUTPAGE(rcmp);
			return rc;
		}

		/*
		 * The new key entry goes ONE AFTER the index of parent entry,
		 * because the split was to the right.
		 */
		skip = parent->index + 1;

		/*
		 * split or shift right remaining entries of the parent page
		 */
		nextindex = le16_to_cpu(sp->header.nextindex);
		/*
		 * parent page is full - split the parent page
		 */
		if (nextindex == le16_to_cpu(sp->header.maxentry)) {
			/* init for parent page split */
			split->mp = smp;
			split->index = skip;	/* index at insert */
			split->flag = XAD_NEW;
			split->off = offsetXAD(&rcp->xad[XTENTRYSTART]);
			split->len = JFS_SBI(ip->i_sb)->nbperpage;
			split->addr = rcbn;

			/* unpin previous right child page */
			XT_PUTPAGE(rcmp);

			/* The split routines insert the new entry,
			 * and acquire txLock as appropriate.
			 * return <rp> pinned and its block number <rpbn>.
			 */
			rc = (sp->header.flag & BT_ROOT) ?
			    xtSplitRoot(tid, ip, split, &rmp) :
			    xtSplitPage(tid, ip, split, &rmp, &rbn);
			if (rc) {
				XT_PUTPAGE(smp);
				return rc;
			}

			XT_PUTPAGE(smp);
			/* keep new child page <rp> pinned */
		}
		/*
		 * parent page is not full - insert in parent page
		 */
		else {
			/*
			 * insert router entry in parent for the right child
			 * page from the first entry of the right child page:
			 */
			/*
			 * acquire a transaction lock on the parent page;
			 *
			 * action: router xad insertion;
			 */
			BT_MARK_DIRTY(smp, ip);

			/*
			 * if insert into middle, shift right remaining entries
			 */
			if (skip < nextindex)
				memmove(&sp->xad[skip + 1], &sp->xad[skip],
					(nextindex -
					 skip) << L2XTSLOTSIZE);

			/* insert the router entry */
			xad = &sp->xad[skip];
			XT_PUTENTRY(xad, XAD_NEW,
				    offsetXAD(&rcp->xad[XTENTRYSTART]),
				    JFS_SBI(ip->i_sb)->nbperpage, rcbn);

			/* advance next available entry index. */
			le16_add_cpu(&sp->header.nextindex, 1);

			/* Don't log it if there are no links to the file */
			if (!test_cflag(COMMIT_Nolink, ip)) {
				tlck = txLock(tid, ip, smp,
					      tlckXTREE | tlckGROW);
				xtlck = (struct xtlock *) & tlck->lock;
				xtlck->lwm.offset = (xtlck->lwm.offset) ?
				    min(skip, (int)xtlck->lwm.offset) : skip;
				xtlck->lwm.length =
				    le16_to_cpu(sp->header.nextindex) -
				    xtlck->lwm.offset;
			}

			/* unpin parent page */
			XT_PUTPAGE(smp);

			/* exit propagate up */
			break;
		}
	}

	/* unpin current right page */
	XT_PUTPAGE(rmp);

	return 0;
}


/*
 *	xtSplitPage()
 *
 * function:
 *	split a full non-root page into
 *	original/split/left page and new right page
 *	i.e., the original/split page remains as left page.
 *
 * parameter:
 *	int		tid,
 *	struct inode	*ip,
 *	struct xtsplit	*split,
 *	struct metapage	**rmpp,
 *	u64		*rbnp,
 *
 * return:
 *	Pointer to page in which to insert or NULL on error.
 */
static int
xtSplitPage(tid_t tid, struct inode *ip,
	    struct xtsplit * split, struct metapage ** rmpp, s64 * rbnp)
{
	int rc = 0;
	struct metapage *smp;
	xtpage_t *sp;
	struct metapage *rmp;
	xtpage_t *rp;		/* new right page allocated */
	s64 rbn;		/* new right page block number */
	struct metapage *mp;
	xtpage_t *p;
	s64 nextbn;
	int skip, maxentry, middle, righthalf, n;
	xad_t *xad;
	struct pxdlist *pxdlist;
	pxd_t *pxd;
	struct tlock *tlck;
	struct xtlock *sxtlck = NULL, *rxtlck = NULL;
	int quota_allocation = 0;

	smp = split->mp;
	sp = XT_PAGE(ip, smp);

	INCREMENT(xtStat.split);

	pxdlist = split->pxdlist;
	pxd = &pxdlist->pxd[pxdlist->npxd];
	pxdlist->npxd++;
	rbn = addressPXD(pxd);

	/* Allocate blocks to quota. */
	rc = dquot_alloc_block(ip, lengthPXD(pxd));
	if (rc)
		goto clean_up;

	quota_allocation += lengthPXD(pxd);

	/*
	 * allocate the new right page for the split
	 */
	rmp = get_metapage(ip, rbn, PSIZE, 1);
	if (rmp == NULL) {
		rc = -EIO;
		goto clean_up;
	}

	jfs_info("xtSplitPage: ip:0x%p smp:0x%p rmp:0x%p", ip, smp, rmp);

	BT_MARK_DIRTY(rmp, ip);
	/*
	 * action: new page;
	 */

	rp = (xtpage_t *) rmp->data;
	rp->header.self = *pxd;
	rp->header.flag = sp->header.flag & BT_TYPE;
	rp->header.maxentry = sp->header.maxentry;	/* little-endian */
	rp->header.nextindex = cpu_to_le16(XTENTRYSTART);

	BT_MARK_DIRTY(smp, ip);
	/* Don't log it if there are no links to the file */
	if (!test_cflag(COMMIT_Nolink, ip)) {
		/*
		 * acquire a transaction lock on the new right page;
		 */
		tlck = txLock(tid, ip, rmp, tlckXTREE | tlckNEW);
		rxtlck = (struct xtlock *) & tlck->lock;
		rxtlck->lwm.offset = XTENTRYSTART;
		/*
		 * acquire a transaction lock on the split page
		 */
		tlck = txLock(tid, ip, smp, tlckXTREE | tlckGROW);
		sxtlck = (struct xtlock *) & tlck->lock;
	}

	/*
	 * initialize/update sibling pointers of <sp> and <rp>
	 */
	nextbn = le64_to_cpu(sp->header.next);
	rp->header.next = cpu_to_le64(nextbn);
	rp->header.prev = cpu_to_le64(addressPXD(&sp->header.self));
	sp->header.next = cpu_to_le64(rbn);

	skip = split->index;

	/*
	 *	sequential append at tail (after last entry of last page)
	 *
	 * if splitting the last page on a level because of appending
	 * a entry to it (skip is maxentry), it's likely that the access is
	 * sequential. adding an empty page on the side of the level is less
	 * work and can push the fill factor much higher than normal.
	 * if we're wrong it's no big deal -  we will do the split the right
	 * way next time.
	 * (it may look like it's equally easy to do a similar hack for
	 * reverse sorted data, that is, split the tree left, but it's not.
	 * Be my guest.)
	 */
	if (nextbn == 0 && skip == le16_to_cpu(sp->header.maxentry)) {
		/*
		 * acquire a transaction lock on the new/right page;
		 *
		 * action: xad insertion;
		 */
		/* insert entry at the first entry of the new right page */
		xad = &rp->xad[XTENTRYSTART];
		XT_PUTENTRY(xad, split->flag, split->off, split->len,
			    split->addr);

		rp->header.nextindex = cpu_to_le16(XTENTRYSTART + 1);

		if (!test_cflag(COMMIT_Nolink, ip)) {
			/* rxtlck->lwm.offset = XTENTRYSTART; */
			rxtlck->lwm.length = 1;
		}

		*rmpp = rmp;
		*rbnp = rbn;

		jfs_info("xtSplitPage: sp:0x%p rp:0x%p", sp, rp);
		return 0;
	}

	/*
	 *	non-sequential insert (at possibly middle page)
	 */

	/*
	 * update previous pointer of old next/right page of <sp>
	 */
	if (nextbn != 0) {
		XT_GETPAGE(ip, nextbn, mp, PSIZE, p, rc);
		if (rc) {
			XT_PUTPAGE(rmp);
			goto clean_up;
		}

		BT_MARK_DIRTY(mp, ip);
		/*
		 * acquire a transaction lock on the next page;
		 *
		 * action:sibling pointer update;
		 */
		if (!test_cflag(COMMIT_Nolink, ip))
			tlck = txLock(tid, ip, mp, tlckXTREE | tlckRELINK);

		p->header.prev = cpu_to_le64(rbn);

		/* sibling page may have been updated previously, or
		 * it may be updated later;
		 */

		XT_PUTPAGE(mp);
	}

	/*
	 * split the data between the split and new/right pages
	 */
	maxentry = le16_to_cpu(sp->header.maxentry);
	middle = maxentry >> 1;
	righthalf = maxentry - middle;

	/*
	 * skip index in old split/left page - insert into left page:
	 */
	if (skip <= middle) {
		/* move right half of split page to the new right page */
		memmove(&rp->xad[XTENTRYSTART], &sp->xad[middle],
			righthalf << L2XTSLOTSIZE);

		/* shift right tail of left half to make room for new entry */
		if (skip < middle)
			memmove(&sp->xad[skip + 1], &sp->xad[skip],
				(middle - skip) << L2XTSLOTSIZE);

		/* insert new entry */
		xad = &sp->xad[skip];
		XT_PUTENTRY(xad, split->flag, split->off, split->len,
			    split->addr);

		/* update page header */
		sp->header.nextindex = cpu_to_le16(middle + 1);
		if (!test_cflag(COMMIT_Nolink, ip)) {
			sxtlck->lwm.offset = (sxtlck->lwm.offset) ?
			    min(skip, (int)sxtlck->lwm.offset) : skip;
		}

		rp->header.nextindex =
		    cpu_to_le16(XTENTRYSTART + righthalf);
	}
	/*
	 * skip index in new right page - insert into right page:
	 */
	else {
		/* move left head of right half to right page */
		n = skip - middle;
		memmove(&rp->xad[XTENTRYSTART], &sp->xad[middle],
			n << L2XTSLOTSIZE);

		/* insert new entry */
		n += XTENTRYSTART;
		xad = &rp->xad[n];
		XT_PUTENTRY(xad, split->flag, split->off, split->len,
			    split->addr);

		/* move right tail of right half to right page */
		if (skip < maxentry)
			memmove(&rp->xad[n + 1], &sp->xad[skip],
				(maxentry - skip) << L2XTSLOTSIZE);

		/* update page header */
		sp->header.nextindex = cpu_to_le16(middle);
		if (!test_cflag(COMMIT_Nolink, ip)) {
			sxtlck->lwm.offset = (sxtlck->lwm.offset) ?
			    min(middle, (int)sxtlck->lwm.offset) : middle;
		}

		rp->header.nextindex = cpu_to_le16(XTENTRYSTART +
						   righthalf + 1);
	}

	if (!test_cflag(COMMIT_Nolink, ip)) {
		sxtlck->lwm.length = le16_to_cpu(sp->header.nextindex) -
		    sxtlck->lwm.offset;

		/* rxtlck->lwm.offset = XTENTRYSTART; */
		rxtlck->lwm.length = le16_to_cpu(rp->header.nextindex) -
		    XTENTRYSTART;
	}

	*rmpp = rmp;
	*rbnp = rbn;

	jfs_info("xtSplitPage: sp:0x%p rp:0x%p", sp, rp);
	return rc;

      clean_up:

	/* Rollback quota allocation. */
	if (quota_allocation)
		dquot_free_block(ip, quota_allocation);

	return (rc);
}


/*
 *	xtSplitRoot()
 *
 * function:
 *	split the full root page into original/root/split page and new
 *	right page
 *	i.e., root remains fixed in tree anchor (inode) and the root is
 *	copied to a single new right child page since root page <<
 *	non-root page, and the split root page contains a single entry
 *	for the new right child page.
 *
 * parameter:
 *	int		tid,
 *	struct inode	*ip,
 *	struct xtsplit	*split,
 *	struct metapage	**rmpp)
 *
 * return:
 *	Pointer to page in which to insert or NULL on error.
 */
static int
xtSplitRoot(tid_t tid,
	    struct inode *ip, struct xtsplit * split, struct metapage ** rmpp)
{
	xtpage_t *sp;
	struct metapage *rmp;
	xtpage_t *rp;
	s64 rbn;
	int skip, nextindex;
	xad_t *xad;
	pxd_t *pxd;
	struct pxdlist *pxdlist;
	struct tlock *tlck;
	struct xtlock *xtlck;
	int rc;

	sp = &JFS_IP(ip)->i_xtroot;

	INCREMENT(xtStat.split);

	/*
	 *	allocate a single (right) child page
	 */
	pxdlist = split->pxdlist;
	pxd = &pxdlist->pxd[pxdlist->npxd];
	pxdlist->npxd++;
	rbn = addressPXD(pxd);
	rmp = get_metapage(ip, rbn, PSIZE, 1);
	if (rmp == NULL)
		return -EIO;

	/* Allocate blocks to quota. */
	rc = dquot_alloc_block(ip, lengthPXD(pxd));
	if (rc) {
		release_metapage(rmp);
		return rc;
	}

	jfs_info("xtSplitRoot: ip:0x%p rmp:0x%p", ip, rmp);

	/*
	 * acquire a transaction lock on the new right page;
	 *
	 * action: new page;
	 */
	BT_MARK_DIRTY(rmp, ip);

	rp = (xtpage_t *) rmp->data;
	rp->header.flag =
	    (sp->header.flag & BT_LEAF) ? BT_LEAF : BT_INTERNAL;
	rp->header.self = *pxd;
	rp->header.nextindex = cpu_to_le16(XTENTRYSTART);
	rp->header.maxentry = cpu_to_le16(PSIZE >> L2XTSLOTSIZE);

	/* initialize sibling pointers */
	rp->header.next = 0;
	rp->header.prev = 0;

	/*
	 * copy the in-line root page into new right page extent
	 */
	nextindex = le16_to_cpu(sp->header.maxentry);
	memmove(&rp->xad[XTENTRYSTART], &sp->xad[XTENTRYSTART],
		(nextindex - XTENTRYSTART) << L2XTSLOTSIZE);

	/*
	 * insert the new entry into the new right/child page
	 * (skip index in the new right page will not change)
	 */
	skip = split->index;
	/* if insert into middle, shift right remaining entries */
	if (skip != nextindex)
		memmove(&rp->xad[skip + 1], &rp->xad[skip],
			(nextindex - skip) * sizeof(xad_t));

	xad = &rp->xad[skip];
	XT_PUTENTRY(xad, split->flag, split->off, split->len, split->addr);

	/* update page header */
	rp->header.nextindex = cpu_to_le16(nextindex + 1);

	if (!test_cflag(COMMIT_Nolink, ip)) {
		tlck = txLock(tid, ip, rmp, tlckXTREE | tlckNEW);
		xtlck = (struct xtlock *) & tlck->lock;
		xtlck->lwm.offset = XTENTRYSTART;
		xtlck->lwm.length = le16_to_cpu(rp->header.nextindex) -
		    XTENTRYSTART;
	}

	/*
	 *	reset the root
	 *
	 * init root with the single entry for the new right page
	 * set the 1st entry offset to 0, which force the left-most key
	 * at any level of the tree to be less than any search key.
	 */
	/*
	 * acquire a transaction lock on the root page (in-memory inode);
	 *
	 * action: root split;
	 */
	BT_MARK_DIRTY(split->mp, ip);

	xad = &sp->xad[XTENTRYSTART];
	XT_PUTENTRY(xad, XAD_NEW, 0, JFS_SBI(ip->i_sb)->nbperpage, rbn);

	/* update page header of root */
	sp->header.flag &= ~BT_LEAF;
	sp->header.flag |= BT_INTERNAL;

	sp->header.nextindex = cpu_to_le16(XTENTRYSTART + 1);

	if (!test_cflag(COMMIT_Nolink, ip)) {
		tlck = txLock(tid, ip, split->mp, tlckXTREE | tlckGROW);
		xtlck = (struct xtlock *) & tlck->lock;
		xtlck->lwm.offset = XTENTRYSTART;
		xtlck->lwm.length = 1;
	}

	*rmpp = rmp;

	jfs_info("xtSplitRoot: sp:0x%p rp:0x%p", sp, rp);
	return 0;
}


/*
 *	xtExtend()
 *
 * function: extend in-place;
 *
 * note: existing extent may or may not have been committed.
 * caller is responsible for pager buffer cache update, and
 * working block allocation map update;
 * update pmap: alloc whole extended extent;
 */
int xtExtend(tid_t tid,		/* transaction id */
	     struct inode *ip, s64 xoff,	/* delta extent offset */
	     s32 xlen,		/* delta extent length */
	     int flag)
{
	int rc = 0;
	int cmp;
	struct metapage *mp;	/* meta-page buffer */
	xtpage_t *p;		/* base B+-tree index page */
	s64 bn;
	int index, nextindex, len;
	struct btstack btstack;	/* traverse stack */
	struct xtsplit split;	/* split information */
	xad_t *xad;
	s64 xaddr;
	struct tlock *tlck;
	struct xtlock *xtlck = NULL;

	jfs_info("xtExtend: nxoff:0x%lx nxlen:0x%x", (ulong) xoff, xlen);

	/* there must exist extent to be extended */
	if ((rc = xtSearch(ip, xoff - 1, NULL, &cmp, &btstack, XT_INSERT)))
		return rc;

	/* retrieve search result */
	XT_GETSEARCH(ip, btstack.top, bn, mp, p, index);

	if (cmp != 0) {
		XT_PUTPAGE(mp);
		jfs_error(ip->i_sb, "xtSearch did not find extent\n");
		return -EIO;
	}

	/* extension must be contiguous */
	xad = &p->xad[index];
	if ((offsetXAD(xad) + lengthXAD(xad)) != xoff) {
		XT_PUTPAGE(mp);
		jfs_error(ip->i_sb, "extension is not contiguous\n");
		return -EIO;
	}

	/*
	 * acquire a transaction lock on the leaf page;
	 *
	 * action: xad insertion/extension;
	 */
	BT_MARK_DIRTY(mp, ip);
	if (!test_cflag(COMMIT_Nolink, ip)) {
		tlck = txLock(tid, ip, mp, tlckXTREE | tlckGROW);
		xtlck = (struct xtlock *) & tlck->lock;
	}

	/* extend will overflow extent ? */
	xlen = lengthXAD(xad) + xlen;
	if ((len = xlen - MAXXLEN) <= 0)
		goto extendOld;

	/*
	 *	extent overflow: insert entry for new extent
	 */
//insertNew:
	xoff = offsetXAD(xad) + MAXXLEN;
	xaddr = addressXAD(xad) + MAXXLEN;
	nextindex = le16_to_cpu(p->header.nextindex);

	/*
	 *	if the leaf page is full, insert the new entry and
	 *	propagate up the router entry for the new page from split
	 *
	 * The xtSplitUp() will insert the entry and unpin the leaf page.
	 */
	if (nextindex == le16_to_cpu(p->header.maxentry)) {
		/* xtSpliUp() unpins leaf pages */
		split.mp = mp;
		split.index = index + 1;
		split.flag = XAD_NEW;
		split.off = xoff;	/* split offset */
		split.len = len;
		split.addr = xaddr;
		split.pxdlist = NULL;
		if ((rc = xtSplitUp(tid, ip, &split, &btstack)))
			return rc;

		/* get back old page */
		XT_GETPAGE(ip, bn, mp, PSIZE, p, rc);
		if (rc)
			return rc;
		/*
		 * if leaf root has been split, original root has been
		 * copied to new child page, i.e., original entry now
		 * resides on the new child page;
		 */
		if (p->header.flag & BT_INTERNAL) {
			ASSERT(p->header.nextindex ==
			       cpu_to_le16(XTENTRYSTART + 1));
			xad = &p->xad[XTENTRYSTART];
			bn = addressXAD(xad);
			XT_PUTPAGE(mp);

			/* get new child page */
			XT_GETPAGE(ip, bn, mp, PSIZE, p, rc);
			if (rc)
				return rc;

			BT_MARK_DIRTY(mp, ip);
			if (!test_cflag(COMMIT_Nolink, ip)) {
				tlck = txLock(tid, ip, mp, tlckXTREE|tlckGROW);
				xtlck = (struct xtlock *) & tlck->lock;
			}
		}
	}
	/*
	 *	insert the new entry into the leaf page
	 */
	else {
		/* insert the new entry: mark the entry NEW */
		xad = &p->xad[index + 1];
		XT_PUTENTRY(xad, XAD_NEW, xoff, len, xaddr);

		/* advance next available entry index */
		le16_add_cpu(&p->header.nextindex, 1);
	}

	/* get back old entry */
	xad = &p->xad[index];
	xlen = MAXXLEN;

	/*
	 * extend old extent
	 */
      extendOld:
	XADlength(xad, xlen);
	if (!(xad->flag & XAD_NEW))
		xad->flag |= XAD_EXTENDED;

	if (!test_cflag(COMMIT_Nolink, ip)) {
		xtlck->lwm.offset =
		    (xtlck->lwm.offset) ? min(index,
					      (int)xtlck->lwm.offset) : index;
		xtlck->lwm.length =
		    le16_to_cpu(p->header.nextindex) - xtlck->lwm.offset;
	}

	/* unpin the leaf page */
	XT_PUTPAGE(mp);

	return rc;
}

#ifdef _NOTYET
/*
 *	xtTailgate()
 *
 * function: split existing 'tail' extent
 *	(split offset >= start offset of tail extent), and
 *	relocate and extend the split tail half;
 *
 * note: existing extent may or may not have been committed.
 * caller is responsible for pager buffer cache update, and
 * working block allocation map update;
 * update pmap: free old split tail extent, alloc new extent;
 */
int xtTailgate(tid_t tid,		/* transaction id */
	       struct inode *ip, s64 xoff,	/* split/new extent offset */
	       s32 xlen,	/* new extent length */
	       s64 xaddr,	/* new extent address */
	       int flag)
{
	int rc = 0;
	int cmp;
	struct metapage *mp;	/* meta-page buffer */
	xtpage_t *p;		/* base B+-tree index page */
	s64 bn;
	int index, nextindex, llen, rlen;
	struct btstack btstack;	/* traverse stack */
	struct xtsplit split;	/* split information */
	xad_t *xad;
	struct tlock *tlck;
	struct xtlock *xtlck = 0;
	struct tlock *mtlck;
	struct maplock *pxdlock;

/*
printf("xtTailgate: nxoff:0x%lx nxlen:0x%x nxaddr:0x%lx\n",
	(ulong)xoff, xlen, (ulong)xaddr);
*/

	/* there must exist extent to be tailgated */
	if ((rc = xtSearch(ip, xoff, NULL, &cmp, &btstack, XT_INSERT)))
		return rc;

	/* retrieve search result */
	XT_GETSEARCH(ip, btstack.top, bn, mp, p, index);

	if (cmp != 0) {
		XT_PUTPAGE(mp);
		jfs_error(ip->i_sb, "couldn't find extent\n");
		return -EIO;
	}

	/* entry found must be last entry */
	nextindex = le16_to_cpu(p->header.nextindex);
	if (index != nextindex - 1) {
		XT_PUTPAGE(mp);
		jfs_error(ip->i_sb, "the entry found is not the last entry\n");
		return -EIO;
	}

	BT_MARK_DIRTY(mp, ip);
	/*
	 * acquire tlock of the leaf page containing original entry
	 */
	if (!test_cflag(COMMIT_Nolink, ip)) {
		tlck = txLock(tid, ip, mp, tlckXTREE | tlckGROW);
		xtlck = (struct xtlock *) & tlck->lock;
	}

	/* completely replace extent ? */
	xad = &p->xad[index];
/*
printf("xtTailgate: xoff:0x%lx xlen:0x%x xaddr:0x%lx\n",
	(ulong)offsetXAD(xad), lengthXAD(xad), (ulong)addressXAD(xad));
*/
	if ((llen = xoff - offsetXAD(xad)) == 0)
		goto updateOld;

	/*
	 *	partially replace extent: insert entry for new extent
	 */
//insertNew:
	/*
	 *	if the leaf page is full, insert the new entry and
	 *	propagate up the router entry for the new page from split
	 *
	 * The xtSplitUp() will insert the entry and unpin the leaf page.
	 */
	if (nextindex == le16_to_cpu(p->header.maxentry)) {
		/* xtSpliUp() unpins leaf pages */
		split.mp = mp;
		split.index = index + 1;
		split.flag = XAD_NEW;
		split.off = xoff;	/* split offset */
		split.len = xlen;
		split.addr = xaddr;
		split.pxdlist = NULL;
		if ((rc = xtSplitUp(tid, ip, &split, &btstack)))
			return rc;

		/* get back old page */
		XT_GETPAGE(ip, bn, mp, PSIZE, p, rc);
		if (rc)
			return rc;
		/*
		 * if leaf root has been split, original root has been
		 * copied to new child page, i.e., original entry now
		 * resides on the new child page;
		 */
		if (p->header.flag & BT_INTERNAL) {
			ASSERT(p->header.nextindex ==
			       cpu_to_le16(XTENTRYSTART + 1));
			xad = &p->xad[XTENTRYSTART];
			bn = addressXAD(xad);
			XT_PUTPAGE(mp);

			/* get new child page */
			XT_GETPAGE(ip, bn, mp, PSIZE, p, rc);
			if (rc)
				return rc;

			BT_MARK_DIRTY(mp, ip);
			if (!test_cflag(COMMIT_Nolink, ip)) {
				tlck = txLock(tid, ip, mp, tlckXTREE|tlckGROW);
				xtlck = (struct xtlock *) & tlck->lock;
			}
		}
	}
	/*
	 *	insert the new entry into the leaf page
	 */
	else {
		/* insert the new entry: mark the entry NEW */
		xad = &p->xad[index + 1];
		XT_PUTENTRY(xad, XAD_NEW, xoff, xlen, xaddr);

		/* advance next available entry index */
		le16_add_cpu(&p->header.nextindex, 1);
	}

	/* get back old XAD */
	xad = &p->xad[index];

	/*
	 * truncate/relocate old extent at split offset
	 */
      updateOld:
	/* update dmap for old/committed/truncated extent */
	rlen = lengthXAD(xad) - llen;
	if (!(xad->flag & XAD_NEW)) {
		/* free from PWMAP at commit */
		if (!test_cflag(COMMIT_Nolink, ip)) {
			mtlck = txMaplock(tid, ip, tlckMAP);
			pxdlock = (struct maplock *) & mtlck->lock;
			pxdlock->flag = mlckFREEPXD;
			PXDaddress(&pxdlock->pxd, addressXAD(xad) + llen);
			PXDlength(&pxdlock->pxd, rlen);
			pxdlock->index = 1;
		}
	} else
		/* free from WMAP */
		dbFree(ip, addressXAD(xad) + llen, (s64) rlen);

	if (llen)
		/* truncate */
		XADlength(xad, llen);
	else
		/* replace */
		XT_PUTENTRY(xad, XAD_NEW, xoff, xlen, xaddr);

	if (!test_cflag(COMMIT_Nolink, ip)) {
		xtlck->lwm.offset = (xtlck->lwm.offset) ?
		    min(index, (int)xtlck->lwm.offset) : index;
		xtlck->lwm.length = le16_to_cpu(p->header.nextindex) -
		    xtlck->lwm.offset;
	}

	/* unpin the leaf page */
	XT_PUTPAGE(mp);

	return rc;
}
#endif /* _NOTYET */

/*
 *	xtUpdate()
 *
 * function: update XAD;
 *
 *	update extent for allocated_but_not_recorded or
 *	compressed extent;
 *
 * parameter:
 *	nxad	- new XAD;
 *		logical extent of the specified XAD must be completely
 *		contained by an existing XAD;
 */
int xtUpdate(tid_t tid, struct inode *ip, xad_t * nxad)
{				/* new XAD */
	int rc = 0;
	int cmp;
	struct metapage *mp;	/* meta-page buffer */
	xtpage_t *p;		/* base B+-tree index page */
	s64 bn;
	int index0, index, newindex, nextindex;
	struct btstack btstack;	/* traverse stack */
	struct xtsplit split;	/* split information */
	xad_t *xad, *lxad, *rxad;
	int xflag;
	s64 nxoff, xoff;
	int nxlen, xlen, lxlen, rxlen;
	s64 nxaddr, xaddr;
	struct tlock *tlck;
	struct xtlock *xtlck = NULL;
	int newpage = 0;

	/* there must exist extent to be tailgated */
	nxoff = offsetXAD(nxad);
	nxlen = lengthXAD(nxad);
	nxaddr = addressXAD(nxad);

	if ((rc = xtSearch(ip, nxoff, NULL, &cmp, &btstack, XT_INSERT)))
		return rc;

	/* retrieve search result */
	XT_GETSEARCH(ip, btstack.top, bn, mp, p, index0);

	if (cmp != 0) {
		XT_PUTPAGE(mp);
		jfs_error(ip->i_sb, "Could not find extent\n");
		return -EIO;
	}

	BT_MARK_DIRTY(mp, ip);
	/*
	 * acquire tlock of the leaf page containing original entry
	 */
	if (!test_cflag(COMMIT_Nolink, ip)) {
		tlck = txLock(tid, ip, mp, tlckXTREE | tlckGROW);
		xtlck = (struct xtlock *) & tlck->lock;
	}

	xad = &p->xad[index0];
	xflag = xad->flag;
	xoff = offsetXAD(xad);
	xlen = lengthXAD(xad);
	xaddr = addressXAD(xad);

	/* nXAD must be completely contained within XAD */
	if ((xoff > nxoff) ||
	    (nxoff + nxlen > xoff + xlen)) {
		XT_PUTPAGE(mp);
		jfs_error(ip->i_sb,
			  "nXAD in not completely contained within XAD\n");
		return -EIO;
	}

	index = index0;
	newindex = index + 1;
	nextindex = le16_to_cpu(p->header.nextindex);

#ifdef  _JFS_WIP_NOCOALESCE
	if (xoff < nxoff)
		goto updateRight;

	/*
	 * replace XAD with nXAD
	 */
      replace:			/* (nxoff == xoff) */
	if (nxlen == xlen) {
		/* replace XAD with nXAD:recorded */
		*xad = *nxad;
		xad->flag = xflag & ~XAD_NOTRECORDED;

		goto out;
	} else			/* (nxlen < xlen) */
		goto updateLeft;
#endif				/* _JFS_WIP_NOCOALESCE */

/* #ifdef _JFS_WIP_COALESCE */
	if (xoff < nxoff)
		goto coalesceRight;

	/*
	 * coalesce with left XAD
	 */
//coalesceLeft: /* (xoff == nxoff) */
	/* is XAD first entry of page ? */
	if (index == XTENTRYSTART)
		goto replace;

	/* is nXAD logically and physically contiguous with lXAD ? */
	lxad = &p->xad[index - 1];
	lxlen = lengthXAD(lxad);
	if (!(lxad->flag & XAD_NOTRECORDED) &&
	    (nxoff == offsetXAD(lxad) + lxlen) &&
	    (nxaddr == addressXAD(lxad) + lxlen) &&
	    (lxlen + nxlen < MAXXLEN)) {
		/* extend right lXAD */
		index0 = index - 1;
		XADlength(lxad, lxlen + nxlen);

		/* If we just merged two extents together, need to make sure the
		 * right extent gets logged.  If the left one is marked XAD_NEW,
		 * then we know it will be logged.  Otherwise, mark as
		 * XAD_EXTENDED
		 */
		if (!(lxad->flag & XAD_NEW))
			lxad->flag |= XAD_EXTENDED;

		if (xlen > nxlen) {
			/* truncate XAD */
			XADoffset(xad, xoff + nxlen);
			XADlength(xad, xlen - nxlen);
			XADaddress(xad, xaddr + nxlen);
			goto out;
		} else {	/* (xlen == nxlen) */

			/* remove XAD */
			if (index < nextindex - 1)
				memmove(&p->xad[index], &p->xad[index + 1],
					(nextindex - index -
					 1) << L2XTSLOTSIZE);

			p->header.nextindex =
			    cpu_to_le16(le16_to_cpu(p->header.nextindex) -
					1);

			index = index0;
			newindex = index + 1;
			nextindex = le16_to_cpu(p->header.nextindex);
			xoff = nxoff = offsetXAD(lxad);
			xlen = nxlen = lxlen + nxlen;
			xaddr = nxaddr = addressXAD(lxad);
			goto coalesceRight;
		}
	}

	/*
	 * replace XAD with nXAD
	 */
      replace:			/* (nxoff == xoff) */
	if (nxlen == xlen) {
		/* replace XAD with nXAD:recorded */
		*xad = *nxad;
		xad->flag = xflag & ~XAD_NOTRECORDED;

		goto coalesceRight;
	} else			/* (nxlen < xlen) */
		goto updateLeft;

	/*
	 * coalesce with right XAD
	 */
      coalesceRight:		/* (xoff <= nxoff) */
	/* is XAD last entry of page ? */
	if (newindex == nextindex) {
		if (xoff == nxoff)
			goto out;
		goto updateRight;
	}

	/* is nXAD logically and physically contiguous with rXAD ? */
	rxad = &p->xad[index + 1];
	rxlen = lengthXAD(rxad);
	if (!(rxad->flag & XAD_NOTRECORDED) &&
	    (nxoff + nxlen == offsetXAD(rxad)) &&
	    (nxaddr + nxlen == addressXAD(rxad)) &&
	    (rxlen + nxlen < MAXXLEN)) {
		/* extend left rXAD */
		XADoffset(rxad, nxoff);
		XADlength(rxad, rxlen + nxlen);
		XADaddress(rxad, nxaddr);

		/* If we just merged two extents together, need to make sure
		 * the left extent gets logged.  If the right one is marked
		 * XAD_NEW, then we know it will be logged.  Otherwise, mark as
		 * XAD_EXTENDED
		 */
		if (!(rxad->flag & XAD_NEW))
			rxad->flag |= XAD_EXTENDED;

		if (xlen > nxlen)
			/* truncate XAD */
			XADlength(xad, xlen - nxlen);
		else {		/* (xlen == nxlen) */

			/* remove XAD */
			memmove(&p->xad[index], &p->xad[index + 1],
				(nextindex - index - 1) << L2XTSLOTSIZE);

			p->header.nextindex =
			    cpu_to_le16(le16_to_cpu(p->header.nextindex) -
					1);
		}

		goto out;
	} else if (xoff == nxoff)
		goto out;

	if (xoff >= nxoff) {
		XT_PUTPAGE(mp);
		jfs_error(ip->i_sb, "xoff >= nxoff\n");
		return -EIO;
	}
/* #endif _JFS_WIP_COALESCE */

	/*
	 * split XAD into (lXAD, nXAD):
	 *
	 *          |---nXAD--->
	 * --|----------XAD----------|--
	 *   |-lXAD-|
	 */
      updateRight:		/* (xoff < nxoff) */
	/* truncate old XAD as lXAD:not_recorded */
	xad = &p->xad[index];
	XADlength(xad, nxoff - xoff);

	/* insert nXAD:recorded */
	if (nextindex == le16_to_cpu(p->header.maxentry)) {

		/* xtSpliUp() unpins leaf pages */
		split.mp = mp;
		split.index = newindex;
		split.flag = xflag & ~XAD_NOTRECORDED;
		split.off = nxoff;
		split.len = nxlen;
		split.addr = nxaddr;
		split.pxdlist = NULL;
		if ((rc = xtSplitUp(tid, ip, &split, &btstack)))
			return rc;

		/* get back old page */
		XT_GETPAGE(ip, bn, mp, PSIZE, p, rc);
		if (rc)
			return rc;
		/*
		 * if leaf root has been split, original root has been
		 * copied to new child page, i.e., original entry now
		 * resides on the new child page;
		 */
		if (p->header.flag & BT_INTERNAL) {
			ASSERT(p->header.nextindex ==
			       cpu_to_le16(XTENTRYSTART + 1));
			xad = &p->xad[XTENTRYSTART];
			bn = addressXAD(xad);
			XT_PUTPAGE(mp);

			/* get new child page */
			XT_GETPAGE(ip, bn, mp, PSIZE, p, rc);
			if (rc)
				return rc;

			BT_MARK_DIRTY(mp, ip);
			if (!test_cflag(COMMIT_Nolink, ip)) {
				tlck = txLock(tid, ip, mp, tlckXTREE|tlckGROW);
				xtlck = (struct xtlock *) & tlck->lock;
			}
		} else {
			/* is nXAD on new page ? */
			if (newindex >
			    (le16_to_cpu(p->header.maxentry) >> 1)) {
				newindex =
				    newindex -
				    le16_to_cpu(p->header.nextindex) +
				    XTENTRYSTART;
				newpage = 1;
			}
		}
	} else {
		/* if insert into middle, shift right remaining entries */
		if (newindex < nextindex)
			memmove(&p->xad[newindex + 1], &p->xad[newindex],
				(nextindex - newindex) << L2XTSLOTSIZE);

		/* insert the entry */
		xad = &p->xad[newindex];
		*xad = *nxad;
		xad->flag = xflag & ~XAD_NOTRECORDED;

		/* advance next available entry index. */
		p->header.nextindex =
		    cpu_to_le16(le16_to_cpu(p->header.nextindex) + 1);
	}

	/*
	 * does nXAD force 3-way split ?
	 *
	 *          |---nXAD--->|
	 * --|----------XAD-------------|--
	 *   |-lXAD-|           |-rXAD -|
	 */
	if (nxoff + nxlen == xoff + xlen)
		goto out;

	/* reorient nXAD as XAD for further split XAD into (nXAD, rXAD) */
	if (newpage) {
		/* close out old page */
		if (!test_cflag(COMMIT_Nolink, ip)) {
			xtlck->lwm.offset = (xtlck->lwm.offset) ?
			    min(index0, (int)xtlck->lwm.offset) : index0;
			xtlck->lwm.length =
			    le16_to_cpu(p->header.nextindex) -
			    xtlck->lwm.offset;
		}

		bn = le64_to_cpu(p->header.next);
		XT_PUTPAGE(mp);

		/* get new right page */
		XT_GETPAGE(ip, bn, mp, PSIZE, p, rc);
		if (rc)
			return rc;

		BT_MARK_DIRTY(mp, ip);
		if (!test_cflag(COMMIT_Nolink, ip)) {
			tlck = txLock(tid, ip, mp, tlckXTREE | tlckGROW);
			xtlck = (struct xtlock *) & tlck->lock;
		}

		index0 = index = newindex;
	} else
		index++;

	newindex = index + 1;
	nextindex = le16_to_cpu(p->header.nextindex);
	xlen = xlen - (nxoff - xoff);
	xoff = nxoff;
	xaddr = nxaddr;

	/* recompute split pages */
	if (nextindex == le16_to_cpu(p->header.maxentry)) {
		XT_PUTPAGE(mp);

		if ((rc = xtSearch(ip, nxoff, NULL, &cmp, &btstack, XT_INSERT)))
			return rc;

		/* retrieve search result */
		XT_GETSEARCH(ip, btstack.top, bn, mp, p, index0);

		if (cmp != 0) {
			XT_PUTPAGE(mp);
			jfs_error(ip->i_sb, "xtSearch failed\n");
			return -EIO;
		}

		if (index0 != index) {
			XT_PUTPAGE(mp);
			jfs_error(ip->i_sb, "unexpected value of index\n");
			return -EIO;
		}
	}

	/*
	 * split XAD into (nXAD, rXAD)
	 *
	 *          ---nXAD---|
	 * --|----------XAD----------|--
	 *                    |-rXAD-|
	 */
      updateLeft:		/* (nxoff == xoff) && (nxlen < xlen) */
	/* update old XAD with nXAD:recorded */
	xad = &p->xad[index];
	*xad = *nxad;
	xad->flag = xflag & ~XAD_NOTRECORDED;

	/* insert rXAD:not_recorded */
	xoff = xoff + nxlen;
	xlen = xlen - nxlen;
	xaddr = xaddr + nxlen;
	if (nextindex == le16_to_cpu(p->header.maxentry)) {
/*
printf("xtUpdate.updateLeft.split p:0x%p\n", p);
*/
		/* xtSpliUp() unpins leaf pages */
		split.mp = mp;
		split.index = newindex;
		split.flag = xflag;
		split.off = xoff;
		split.len = xlen;
		split.addr = xaddr;
		split.pxdlist = NULL;
		if ((rc = xtSplitUp(tid, ip, &split, &btstack)))
			return rc;

		/* get back old page */
		XT_GETPAGE(ip, bn, mp, PSIZE, p, rc);
		if (rc)
			return rc;

		/*
		 * if leaf root has been split, original root has been
		 * copied to new child page, i.e., original entry now
		 * resides on the new child page;
		 */
		if (p->header.flag & BT_INTERNAL) {
			ASSERT(p->header.nextindex ==
			       cpu_to_le16(XTENTRYSTART + 1));
			xad = &p->xad[XTENTRYSTART];
			bn = addressXAD(xad);
			XT_PUTPAGE(mp);

			/* get new child page */
			XT_GETPAGE(ip, bn, mp, PSIZE, p, rc);
			if (rc)
				return rc;

			BT_MARK_DIRTY(mp, ip);
			if (!test_cflag(COMMIT_Nolink, ip)) {
				tlck = txLock(tid, ip, mp, tlckXTREE|tlckGROW);
				xtlck = (struct xtlock *) & tlck->lock;
			}
		}
	} else {
		/* if insert into middle, shift right remaining entries */
		if (newindex < nextindex)
			memmove(&p->xad[newindex + 1], &p->xad[newindex],
				(nextindex - newindex) << L2XTSLOTSIZE);

		/* insert the entry */
		xad = &p->xad[newindex];
		XT_PUTENTRY(xad, xflag, xoff, xlen, xaddr);

		/* advance next available entry index. */
		p->header.nextindex =
		    cpu_to_le16(le16_to_cpu(p->header.nextindex) + 1);
	}

      out:
	if (!test_cflag(COMMIT_Nolink, ip)) {
		xtlck->lwm.offset = (xtlck->lwm.offset) ?
		    min(index0, (int)xtlck->lwm.offset) : index0;
		xtlck->lwm.length = le16_to_cpu(p->header.nextindex) -
		    xtlck->lwm.offset;
	}

	/* unpin the leaf page */
	XT_PUTPAGE(mp);

	return rc;
}


/*
 *	xtAppend()
 *
 * function: grow in append mode from contiguous region specified ;
 *
 * parameter:
 *	tid		- transaction id;
 *	ip		- file object;
 *	xflag		- extent flag:
 *	xoff		- extent offset;
 *	maxblocks	- max extent length;
 *	xlen		- extent length (in/out);
 *	xaddrp		- extent address pointer (in/out):
 *	flag		-
 *
 * return:
 */
int xtAppend(tid_t tid,		/* transaction id */
	     struct inode *ip, int xflag, s64 xoff, s32 maxblocks,
	     s32 * xlenp,	/* (in/out) */
	     s64 * xaddrp,	/* (in/out) */
	     int flag)
{
	int rc = 0;
	struct metapage *mp;	/* meta-page buffer */
	xtpage_t *p;		/* base B+-tree index page */
	s64 bn, xaddr;
	int index, nextindex;
	struct btstack btstack;	/* traverse stack */
	struct xtsplit split;	/* split information */
	xad_t *xad;
	int cmp;
	struct tlock *tlck;
	struct xtlock *xtlck;
	int nsplit, nblocks, xlen;
	struct pxdlist pxdlist;
	pxd_t *pxd;
	s64 next;

	xaddr = *xaddrp;
	xlen = *xlenp;
	jfs_info("xtAppend: xoff:0x%lx maxblocks:%d xlen:%d xaddr:0x%lx",
		 (ulong) xoff, maxblocks, xlen, (ulong) xaddr);

	/*
	 *	search for the entry location at which to insert:
	 *
	 * xtFastSearch() and xtSearch() both returns (leaf page
	 * pinned, index at which to insert).
	 * n.b. xtSearch() may return index of maxentry of
	 * the full page.
	 */
	if ((rc = xtSearch(ip, xoff, &next, &cmp, &btstack, XT_INSERT)))
		return rc;

	/* retrieve search result */
	XT_GETSEARCH(ip, btstack.top, bn, mp, p, index);

	if (cmp == 0) {
		rc = -EEXIST;
		goto out;
	}

	if (next)
		xlen = min(xlen, (int)(next - xoff));
//insert:
	/*
	 *	insert entry for new extent
	 */
	xflag |= XAD_NEW;

	/*
	 *	if the leaf page is full, split the page and
	 *	propagate up the router entry for the new page from split
	 *
	 * The xtSplitUp() will insert the entry and unpin the leaf page.
	 */
	nextindex = le16_to_cpu(p->header.nextindex);
	if (nextindex < le16_to_cpu(p->header.maxentry))
		goto insertLeaf;

	/*
	 * allocate new index blocks to cover index page split(s)
	 */
	nsplit = btstack.nsplit;
	split.pxdlist = &pxdlist;
	pxdlist.maxnpxd = pxdlist.npxd = 0;
	pxd = &pxdlist.pxd[0];
	nblocks = JFS_SBI(ip->i_sb)->nbperpage;
	for (; nsplit > 0; nsplit--, pxd++, xaddr += nblocks, maxblocks -= nblocks) {
		if ((rc = dbAllocBottomUp(ip, xaddr, (s64) nblocks)) == 0) {
			PXDaddress(pxd, xaddr);
			PXDlength(pxd, nblocks);

			pxdlist.maxnpxd++;

			continue;
		}

		/* undo allocation */

		goto out;
	}

	xlen = min(xlen, maxblocks);

	/*
	 * allocate data extent requested
	 */
	if ((rc = dbAllocBottomUp(ip, xaddr, (s64) xlen)))
		goto out;

	split.mp = mp;
	split.index = index;
	split.flag = xflag;
	split.off = xoff;
	split.len = xlen;
	split.addr = xaddr;
	if ((rc = xtSplitUp(tid, ip, &split, &btstack))) {
		/* undo data extent allocation */
		dbFree(ip, *xaddrp, (s64) * xlenp);

		return rc;
	}

	*xaddrp = xaddr;
	*xlenp = xlen;
	return 0;

	/*
	 *	insert the new entry into the leaf page
	 */
      insertLeaf:
	/*
	 * allocate data extent requested
	 */
	if ((rc = dbAllocBottomUp(ip, xaddr, (s64) xlen)))
		goto out;

	BT_MARK_DIRTY(mp, ip);
	/*
	 * acquire a transaction lock on the leaf page;
	 *
	 * action: xad insertion/extension;
	 */
	tlck = txLock(tid, ip, mp, tlckXTREE | tlckGROW);
	xtlck = (struct xtlock *) & tlck->lock;

	/* insert the new entry: mark the entry NEW */
	xad = &p->xad[index];
	XT_PUTENTRY(xad, xflag, xoff, xlen, xaddr);

	/* advance next available entry index */
	le16_add_cpu(&p->header.nextindex, 1);

	xtlck->lwm.offset =
	    (xtlck->lwm.offset) ? min(index,(int) xtlck->lwm.offset) : index;
	xtlck->lwm.length = le16_to_cpu(p->header.nextindex) -
	    xtlck->lwm.offset;

	*xaddrp = xaddr;
	*xlenp = xlen;

      out:
	/* unpin the leaf page */
	XT_PUTPAGE(mp);

	return rc;
}
#ifdef _STILL_TO_PORT

/* - TBD for defragmentaion/reorganization -
 *
 *	xtDelete()
 *
 * function:
 *	delete the entry with the specified key.
 *
 *	N.B.: whole extent of the entry is assumed to be deleted.
 *
 * parameter:
 *
 * return:
 *	ENOENT: if the entry is not found.
 *
 * exception:
 */
int xtDelete(tid_t tid, struct inode *ip, s64 xoff, s32 xlen, int flag)
{
	int rc = 0;
	struct btstack btstack;
	int cmp;
	s64 bn;
	struct metapage *mp;
	xtpage_t *p;
	int index, nextindex;
	struct tlock *tlck;
	struct xtlock *xtlck;

	/*
	 * find the matching entry; xtSearch() pins the page
	 */
	if ((rc = xtSearch(ip, xoff, NULL, &cmp, &btstack, 0)))
		return rc;

	XT_GETSEARCH(ip, btstack.top, bn, mp, p, index);
	if (cmp) {
		/* unpin the leaf page */
		XT_PUTPAGE(mp);
		return -ENOENT;
	}

	/*
	 * delete the entry from the leaf page
	 */
	nextindex = le16_to_cpu(p->header.nextindex);
	le16_add_cpu(&p->header.nextindex, -1);

	/*
	 * if the leaf page bocome empty, free the page
	 */
	if (p->header.nextindex == cpu_to_le16(XTENTRYSTART))
		return (xtDeleteUp(tid, ip, mp, p, &btstack));

	BT_MARK_DIRTY(mp, ip);
	/*
	 * acquire a transaction lock on the leaf page;
	 *
	 * action:xad deletion;
	 */
	tlck = txLock(tid, ip, mp, tlckXTREE);
	xtlck = (struct xtlock *) & tlck->lock;
	xtlck->lwm.offset =
	    (xtlck->lwm.offset) ? min(index, xtlck->lwm.offset) : index;

	/* if delete from middle, shift left/compact the remaining entries */
	if (index < nextindex - 1)
		memmove(&p->xad[index], &p->xad[index + 1],
			(nextindex - index - 1) * sizeof(xad_t));

	XT_PUTPAGE(mp);

	return 0;
}


/* - TBD for defragmentaion/reorganization -
 *
 *	xtDeleteUp()
 *
 * function:
 *	free empty pages as propagating deletion up the tree
 *
 * parameter:
 *
 * return:
 */
static int
xtDeleteUp(tid_t tid, struct inode *ip,
	   struct metapage * fmp, xtpage_t * fp, struct btstack * btstack)
{
	int rc = 0;
	struct metapage *mp;
	xtpage_t *p;
	int index, nextindex;
	s64 xaddr;
	int xlen;
	struct btframe *parent;
	struct tlock *tlck;
	struct xtlock *xtlck;

	/*
	 * keep root leaf page which has become empty
	 */
	if (fp->header.flag & BT_ROOT) {
		/* keep the root page */
		fp->header.flag &= ~BT_INTERNAL;
		fp->header.flag |= BT_LEAF;
		fp->header.nextindex = cpu_to_le16(XTENTRYSTART);

		/* XT_PUTPAGE(fmp); */

		return 0;
	}

	/*
	 * free non-root leaf page
	 */
	if ((rc = xtRelink(tid, ip, fp))) {
		XT_PUTPAGE(fmp);
		return rc;
	}

	xaddr = addressPXD(&fp->header.self);
	xlen = lengthPXD(&fp->header.self);
	/* free the page extent */
	dbFree(ip, xaddr, (s64) xlen);

	/* free the buffer page */
	discard_metapage(fmp);

	/*
	 * propagate page deletion up the index tree
	 *
	 * If the delete from the parent page makes it empty,
	 * continue all the way up the tree.
	 * stop if the root page is reached (which is never deleted) or
	 * if the entry deletion does not empty the page.
	 */
	while ((parent = BT_POP(btstack)) != NULL) {
		/* get/pin the parent page <sp> */
		XT_GETPAGE(ip, parent->bn, mp, PSIZE, p, rc);
		if (rc)
			return rc;

		index = parent->index;

		/* delete the entry for the freed child page from parent.
		 */
		nextindex = le16_to_cpu(p->header.nextindex);

		/*
		 * the parent has the single entry being deleted:
		 * free the parent page which has become empty.
		 */
		if (nextindex == 1) {
			if (p->header.flag & BT_ROOT) {
				/* keep the root page */
				p->header.flag &= ~BT_INTERNAL;
				p->header.flag |= BT_LEAF;
				p->header.nextindex =
				    cpu_to_le16(XTENTRYSTART);

				/* XT_PUTPAGE(mp); */

				break;
			} else {
				/* free the parent page */
				if ((rc = xtRelink(tid, ip, p)))
					return rc;

				xaddr = addressPXD(&p->header.self);
				/* free the page extent */
				dbFree(ip, xaddr,
				       (s64) JFS_SBI(ip->i_sb)->nbperpage);

				/* unpin/free the buffer page */
				discard_metapage(mp);

				/* propagate up */
				continue;
			}
		}
		/*
		 * the parent has other entries remaining:
		 * delete the router entry from the parent page.
		 */
		else {
			BT_MARK_DIRTY(mp, ip);
			/*
			 * acquire a transaction lock on the leaf page;
			 *
			 * action:xad deletion;
			 */
			tlck = txLock(tid, ip, mp, tlckXTREE);
			xtlck = (struct xtlock *) & tlck->lock;
			xtlck->lwm.offset =
			    (xtlck->lwm.offset) ? min(index,
						      xtlck->lwm.
						      offset) : index;

			/* if delete from middle,
			 * shift left/compact the remaining entries in the page
			 */
			if (index < nextindex - 1)
				memmove(&p->xad[index], &p->xad[index + 1],
					(nextindex - index -
					 1) << L2XTSLOTSIZE);

			le16_add_cpu(&p->header.nextindex, -1);
			jfs_info("xtDeleteUp(entry): 0x%lx[%d]",
				 (ulong) parent->bn, index);
		}

		/* unpin the parent page */
		XT_PUTPAGE(mp);

		/* exit propagation up */
		break;
	}

	return 0;
}


/*
 * NAME:	xtRelocate()
 *
 * FUNCTION:	relocate xtpage or data extent of regular file;
 *		This function is mainly used by defragfs utility.
 *
 * NOTE:	This routine does not have the logic to handle
 *		uncommitted allocated extent. The caller should call
 *		txCommit() to commit all the allocation before call
 *		this routine.
 */
int
xtRelocate(tid_t tid, struct inode * ip, xad_t * oxad,	/* old XAD */
	   s64 nxaddr,		/* new xaddr */
	   int xtype)
{				/* extent type: XTPAGE or DATAEXT */
	int rc = 0;
	struct tblock *tblk;
	struct tlock *tlck;
	struct xtlock *xtlck;
	struct metapage *mp, *pmp, *lmp, *rmp;	/* meta-page buffer */
	xtpage_t *p, *pp, *rp, *lp;	/* base B+-tree index page */
	xad_t *xad;
	pxd_t *pxd;
	s64 xoff, xsize;
	int xlen;
	s64 oxaddr, sxaddr, dxaddr, nextbn, prevbn;
	cbuf_t *cp;
	s64 offset, nbytes, nbrd, pno;
	int nb, npages, nblks;
	s64 bn;
	int cmp;
	int index;
	struct pxd_lock *pxdlock;
	struct btstack btstack;	/* traverse stack */

	xtype = xtype & EXTENT_TYPE;

	xoff = offsetXAD(oxad);
	oxaddr = addressXAD(oxad);
	xlen = lengthXAD(oxad);

	/* validate extent offset */
	offset = xoff << JFS_SBI(ip->i_sb)->l2bsize;
	if (offset >= ip->i_size)
		return -ESTALE;	/* stale extent */

	jfs_info("xtRelocate: xtype:%d xoff:0x%lx xlen:0x%x xaddr:0x%lx:0x%lx",
		 xtype, (ulong) xoff, xlen, (ulong) oxaddr, (ulong) nxaddr);

	/*
	 *	1. get and validate the parent xtpage/xad entry
	 *	covering the source extent to be relocated;
	 */
	if (xtype == DATAEXT) {
		/* search in leaf entry */
		rc = xtSearch(ip, xoff, NULL, &cmp, &btstack, 0);
		if (rc)
			return rc;

		/* retrieve search result */
		XT_GETSEARCH(ip, btstack.top, bn, pmp, pp, index);

		if (cmp) {
			XT_PUTPAGE(pmp);
			return -ESTALE;
		}

		/* validate for exact match with a single entry */
		xad = &pp->xad[index];
		if (addressXAD(xad) != oxaddr || lengthXAD(xad) != xlen) {
			XT_PUTPAGE(pmp);
			return -ESTALE;
		}
	} else {		/* (xtype == XTPAGE) */

		/* search in internal entry */
		rc = xtSearchNode(ip, oxad, &cmp, &btstack, 0);
		if (rc)
			return rc;

		/* retrieve search result */
		XT_GETSEARCH(ip, btstack.top, bn, pmp, pp, index);

		if (cmp) {
			XT_PUTPAGE(pmp);
			return -ESTALE;
		}

		/* xtSearchNode() validated for exact match with a single entry
		 */
		xad = &pp->xad[index];
	}
	jfs_info("xtRelocate: parent xad entry validated.");

	/*
	 *	2. relocate the extent
	 */
	if (xtype == DATAEXT) {
		/* if the extent is allocated-but-not-recorded
		 * there is no real data to be moved in this extent,
		 */
		if (xad->flag & XAD_NOTRECORDED)
			goto out;
		else
			/* release xtpage for cmRead()/xtLookup() */
			XT_PUTPAGE(pmp);

		/*
		 *	cmRelocate()
		 *
		 * copy target data pages to be relocated;
		 *
		 * data extent must start at page boundary and
		 * multiple of page size (except the last data extent);
		 * read in each page of the source data extent into cbuf,
		 * update the cbuf extent descriptor of the page to be
		 * homeward bound to new dst data extent
		 * copy the data from the old extent to new extent.
		 * copy is essential for compressed files to avoid problems
		 * that can arise if there was a change in compression
		 * algorithms.
		 * it is a good strategy because it may disrupt cache
		 * policy to keep the pages in memory afterwards.
		 */
		offset = xoff << JFS_SBI(ip->i_sb)->l2bsize;
		assert((offset & CM_OFFSET) == 0);
		nbytes = xlen << JFS_SBI(ip->i_sb)->l2bsize;
		pno = offset >> CM_L2BSIZE;
		npages = (nbytes + (CM_BSIZE - 1)) >> CM_L2BSIZE;
/*
		npages = ((offset + nbytes - 1) >> CM_L2BSIZE) -
			  (offset >> CM_L2BSIZE) + 1;
*/
		sxaddr = oxaddr;
		dxaddr = nxaddr;

		/* process the request one cache buffer at a time */
		for (nbrd = 0; nbrd < nbytes; nbrd += nb,
		     offset += nb, pno++, npages--) {
			/* compute page size */
			nb = min(nbytes - nbrd, CM_BSIZE);

			/* get the cache buffer of the page */
			if (rc = cmRead(ip, offset, npages, &cp))
				break;

			assert(addressPXD(&cp->cm_pxd) == sxaddr);
			assert(!cp->cm_modified);

			/* bind buffer with the new extent address */
			nblks = nb >> JFS_IP(ip->i_sb)->l2bsize;
			cmSetXD(ip, cp, pno, dxaddr, nblks);

			/* release the cbuf, mark it as modified */
			cmPut(cp, true);

			dxaddr += nblks;
			sxaddr += nblks;
		}

		/* get back parent page */
		if ((rc = xtSearch(ip, xoff, NULL, &cmp, &btstack, 0)))
			return rc;

		XT_GETSEARCH(ip, btstack.top, bn, pmp, pp, index);
		jfs_info("xtRelocate: target data extent relocated.");
	} else {		/* (xtype == XTPAGE) */

		/*
		 * read in the target xtpage from the source extent;
		 */
		XT_GETPAGE(ip, oxaddr, mp, PSIZE, p, rc);
		if (rc) {
			XT_PUTPAGE(pmp);
			return rc;
		}

		/*
		 * read in sibling pages if any to update sibling pointers;
		 */
		rmp = NULL;
		if (p->header.next) {
			nextbn = le64_to_cpu(p->header.next);
			XT_GETPAGE(ip, nextbn, rmp, PSIZE, rp, rc);
			if (rc) {
				XT_PUTPAGE(pmp);
				XT_PUTPAGE(mp);
				return (rc);
			}
		}

		lmp = NULL;
		if (p->header.prev) {
			prevbn = le64_to_cpu(p->header.prev);
			XT_GETPAGE(ip, prevbn, lmp, PSIZE, lp, rc);
			if (rc) {
				XT_PUTPAGE(pmp);
				XT_PUTPAGE(mp);
				if (rmp)
					XT_PUTPAGE(rmp);
				return (rc);
			}
		}

		/* at this point, all xtpages to be updated are in memory */

		/*
		 * update sibling pointers of sibling xtpages if any;
		 */
		if (lmp) {
			BT_MARK_DIRTY(lmp, ip);
			tlck = txLock(tid, ip, lmp, tlckXTREE | tlckRELINK);
			lp->header.next = cpu_to_le64(nxaddr);
			XT_PUTPAGE(lmp);
		}

		if (rmp) {
			BT_MARK_DIRTY(rmp, ip);
			tlck = txLock(tid, ip, rmp, tlckXTREE | tlckRELINK);
			rp->header.prev = cpu_to_le64(nxaddr);
			XT_PUTPAGE(rmp);
		}

		/*
		 * update the target xtpage to be relocated
		 *
		 * update the self address of the target page
		 * and write to destination extent;
		 * redo image covers the whole xtpage since it is new page
		 * to the destination extent;
		 * update of bmap for the free of source extent
		 * of the target xtpage itself:
		 * update of bmap for the allocation of destination extent
		 * of the target xtpage itself:
		 * update of bmap for the extents covered by xad entries in
		 * the target xtpage is not necessary since they are not
		 * updated;
		 * if not committed before this relocation,
		 * target page may contain XAD_NEW entries which must
		 * be scanned for bmap update (logredo() always
		 * scan xtpage REDOPAGE image for bmap update);
		 * if committed before this relocation (tlckRELOCATE),
		 * scan may be skipped by commit() and logredo();
		 */
		BT_MARK_DIRTY(mp, ip);
		/* tlckNEW init xtlck->lwm.offset = XTENTRYSTART; */
		tlck = txLock(tid, ip, mp, tlckXTREE | tlckNEW);
		xtlck = (struct xtlock *) & tlck->lock;

		/* update the self address in the xtpage header */
		pxd = &p->header.self;
		PXDaddress(pxd, nxaddr);

		/* linelock for the after image of the whole page */
		xtlck->lwm.length =
		    le16_to_cpu(p->header.nextindex) - xtlck->lwm.offset;

		/* update the buffer extent descriptor of target xtpage */
		xsize = xlen << JFS_SBI(ip->i_sb)->l2bsize;
		bmSetXD(mp, nxaddr, xsize);

		/* unpin the target page to new homeward bound */
		XT_PUTPAGE(mp);
		jfs_info("xtRelocate: target xtpage relocated.");
	}

	/*
	 *	3. acquire maplock for the source extent to be freed;
	 *
	 * acquire a maplock saving the src relocated extent address;
	 * to free of the extent at commit time;
	 */
      out:
	/* if DATAEXT relocation, write a LOG_UPDATEMAP record for
	 * free PXD of the source data extent (logredo() will update
	 * bmap for free of source data extent), and update bmap for
	 * free of the source data extent;
	 */
	if (xtype == DATAEXT)
		tlck = txMaplock(tid, ip, tlckMAP);
	/* if XTPAGE relocation, write a LOG_NOREDOPAGE record
	 * for the source xtpage (logredo() will init NoRedoPage
	 * filter and will also update bmap for free of the source
	 * xtpage), and update bmap for free of the source xtpage;
	 * N.B. We use tlckMAP instead of tlkcXTREE because there
	 *      is no buffer associated with this lock since the buffer
	 *      has been redirected to the target location.
	 */
	else			/* (xtype == XTPAGE) */
		tlck = txMaplock(tid, ip, tlckMAP | tlckRELOCATE);

	pxdlock = (struct pxd_lock *) & tlck->lock;
	pxdlock->flag = mlckFREEPXD;
	PXDaddress(&pxdlock->pxd, oxaddr);
	PXDlength(&pxdlock->pxd, xlen);
	pxdlock->index = 1;

	/*
	 *	4. update the parent xad entry for relocation;
	 *
	 * acquire tlck for the parent entry with XAD_NEW as entry
	 * update which will write LOG_REDOPAGE and update bmap for
	 * allocation of XAD_NEW destination extent;
	 */
	jfs_info("xtRelocate: update parent xad entry.");
	BT_MARK_DIRTY(pmp, ip);
	tlck = txLock(tid, ip, pmp, tlckXTREE | tlckGROW);
	xtlck = (struct xtlock *) & tlck->lock;

	/* update the XAD with the new destination extent; */
	xad = &pp->xad[index];
	xad->flag |= XAD_NEW;
	XADaddress(xad, nxaddr);

	xtlck->lwm.offset = min(index, xtlck->lwm.offset);
	xtlck->lwm.length = le16_to_cpu(pp->header.nextindex) -
	    xtlck->lwm.offset;

	/* unpin the parent xtpage */
	XT_PUTPAGE(pmp);

	return rc;
}


/*
 *	xtSearchNode()
 *
 * function:	search for the internal xad entry covering specified extent.
 *		This function is mainly used by defragfs utility.
 *
 * parameters:
 *	ip	- file object;
 *	xad	- extent to find;
 *	cmpp	- comparison result:
 *	btstack - traverse stack;
 *	flag	- search process flag;
 *
 * returns:
 *	btstack contains (bn, index) of search path traversed to the entry.
 *	*cmpp is set to result of comparison with the entry returned.
 *	the page containing the entry is pinned at exit.
 */
static int xtSearchNode(struct inode *ip, xad_t * xad,	/* required XAD entry */
			int *cmpp, struct btstack * btstack, int flag)
{
	int rc = 0;
	s64 xoff, xaddr;
	int xlen;
	int cmp = 1;		/* init for empty page */
	s64 bn;			/* block number */
	struct metapage *mp;	/* meta-page buffer */
	xtpage_t *p;		/* page */
	int base, index, lim;
	struct btframe *btsp;
	s64 t64;

	BT_CLR(btstack);

	xoff = offsetXAD(xad);
	xlen = lengthXAD(xad);
	xaddr = addressXAD(xad);

	/*
	 *	search down tree from root:
	 *
	 * between two consecutive entries of <Ki, Pi> and <Kj, Pj> of
	 * internal page, child page Pi contains entry with k, Ki <= K < Kj.
	 *
	 * if entry with search key K is not found
	 * internal page search find the entry with largest key Ki
	 * less than K which point to the child page to search;
	 * leaf page search find the entry with smallest key Kj
	 * greater than K so that the returned index is the position of
	 * the entry to be shifted right for insertion of new entry.
	 * for empty tree, search key is greater than any key of the tree.
	 *
	 * by convention, root bn = 0.
	 */
	for (bn = 0;;) {
		/* get/pin the page to search */
		XT_GETPAGE(ip, bn, mp, PSIZE, p, rc);
		if (rc)
			return rc;
		if (p->header.flag & BT_LEAF) {
			XT_PUTPAGE(mp);
			return -ESTALE;
		}

		lim = le16_to_cpu(p->header.nextindex) - XTENTRYSTART;

		/*
		 * binary search with search key K on the current page
		 */
		for (base = XTENTRYSTART; lim; lim >>= 1) {
			index = base + (lim >> 1);

			XT_CMP(cmp, xoff, &p->xad[index], t64);
			if (cmp == 0) {
				/*
				 *	search hit
				 *
				 * verify for exact match;
				 */
				if (xaddr == addressXAD(&p->xad[index]) &&
				    xoff == offsetXAD(&p->xad[index])) {
					*cmpp = cmp;

					/* save search result */
					btsp = btstack->top;
					btsp->bn = bn;
					btsp->index = index;
					btsp->mp = mp;

					return 0;
				}

				/* descend/search its child page */
				goto next;
			}

			if (cmp > 0) {
				base = index + 1;
				--lim;
			}
		}

		/*
		 *	search miss - non-leaf page:
		 *
		 * base is the smallest index with key (Kj) greater than
		 * search key (K) and may be zero or maxentry index.
		 * if base is non-zero, decrement base by one to get the parent
		 * entry of the child page to search.
		 */
		index = base ? base - 1 : base;

		/*
		 * go down to child page
		 */
	      next:
		/* get the child page block number */
		bn = addressXAD(&p->xad[index]);

		/* unpin the parent page */
		XT_PUTPAGE(mp);
	}
}


/*
 *	xtRelink()
 *
 * function:
 *	link around a freed page.
 *
 * Parameter:
 *	int		tid,
 *	struct inode	*ip,
 *	xtpage_t	*p)
 *
 * returns:
 */
static int xtRelink(tid_t tid, struct inode *ip, xtpage_t * p)
{
	int rc = 0;
	struct metapage *mp;
	s64 nextbn, prevbn;
	struct tlock *tlck;

	nextbn = le64_to_cpu(p->header.next);
	prevbn = le64_to_cpu(p->header.prev);

	/* update prev pointer of the next page */
	if (nextbn != 0) {
		XT_GETPAGE(ip, nextbn, mp, PSIZE, p, rc);
		if (rc)
			return rc;

		/*
		 * acquire a transaction lock on the page;
		 *
		 * action: update prev pointer;
		 */
		BT_MARK_DIRTY(mp, ip);
		tlck = txLock(tid, ip, mp, tlckXTREE | tlckRELINK);

		/* the page may already have been tlock'd */

		p->header.prev = cpu_to_le64(prevbn);

		XT_PUTPAGE(mp);
	}

	/* update next pointer of the previous page */
	if (prevbn != 0) {
		XT_GETPAGE(ip, prevbn, mp, PSIZE, p, rc);
		if (rc)
			return rc;

		/*
		 * acquire a transaction lock on the page;
		 *
		 * action: update next pointer;
		 */
		BT_MARK_DIRTY(mp, ip);
		tlck = txLock(tid, ip, mp, tlckXTREE | tlckRELINK);

		/* the page may already have been tlock'd */

		p->header.next = le64_to_cpu(nextbn);

		XT_PUTPAGE(mp);
	}

	return 0;
}
#endif				/*  _STILL_TO_PORT */


/*
 *	xtInitRoot()
 *
 * initialize file root (inline in inode)
 */
void xtInitRoot(tid_t tid, struct inode *ip)
{
	xtpage_t *p;

	/*
	 * acquire a transaction lock on the root
	 *
	 * action:
	 */
	txLock(tid, ip, (struct metapage *) &JFS_IP(ip)->bxflag,
		      tlckXTREE | tlckNEW);
	p = &JFS_IP(ip)->i_xtroot;

	p->header.flag = DXD_INDEX | BT_ROOT | BT_LEAF;
	p->header.nextindex = cpu_to_le16(XTENTRYSTART);

	if (S_ISDIR(ip->i_mode))
		p->header.maxentry = cpu_to_le16(XTROOTINITSLOT_DIR);
	else {
		p->header.maxentry = cpu_to_le16(XTROOTINITSLOT);
		ip->i_size = 0;
	}


	return;
}


/*
 * We can run into a deadlock truncating a file with a large number of
 * xtree pages (large fragmented file).  A robust fix would entail a
 * reservation system where we would reserve a number of metadata pages
 * and tlocks which we would be guaranteed without a deadlock.  Without
 * this, a partial fix is to limit number of metadata pages we will lock
 * in a single transaction.  Currently we will truncate the file so that
 * no more than 50 leaf pages will be locked.  The caller of xtTruncate
 * will be responsible for ensuring that the current transaction gets
 * committed, and that subsequent transactions are created to truncate
 * the file further if needed.
 */
#define MAX_TRUNCATE_LEAVES 50

/*
 *	xtTruncate()
 *
 * function:
 *	traverse for truncation logging backward bottom up;
 *	terminate at the last extent entry at the current subtree
 *	root page covering new down size.
 *	truncation may occur within the last extent entry.
 *
 * parameter:
 *	int		tid,
 *	struct inode	*ip,
 *	s64		newsize,
 *	int		type)	{PWMAP, PMAP, WMAP; DELETE, TRUNCATE}
 *
 * return:
 *
 * note:
 *	PWMAP:
 *	 1. truncate (non-COMMIT_NOLINK file)
 *	    by jfs_truncate() or jfs_open(O_TRUNC):
 *	    xtree is updated;
 *	 2. truncate index table of directory when last entry removed
 *	map update via tlock at commit time;
 *	PMAP:
 *	 Call xtTruncate_pmap instead
 *	WMAP:
 *	 1. remove (free zero link count) on last reference release
 *	    (pmap has been freed at commit zero link count);
 *	 2. truncate (COMMIT_NOLINK file, i.e., tmp file):
 *	    xtree is updated;
 *	 map update directly at truncation time;
 *
 *	if (DELETE)
 *		no LOG_NOREDOPAGE is required (NOREDOFILE is sufficient);
 *	else if (TRUNCATE)
 *		must write LOG_NOREDOPAGE for deleted index page;
 *
 * pages may already have been tlocked by anonymous transactions
 * during file growth (i.e., write) before truncation;
 *
 * except last truncated entry, deleted entries remains as is
 * in the page (nextindex is updated) for other use
 * (e.g., log/update allocation map): this avoid copying the page
 * info but delay free of pages;
 *
 */
s64 xtTruncate(tid_t tid, struct inode *ip, s64 newsize, int flag)
{
	int rc = 0;
	s64 teof;
	struct metapage *mp;
	xtpage_t *p;
	s64 bn;
	int index, nextindex;
	xad_t *xad;
	s64 xoff, xaddr;
	int xlen, len, freexlen;
	struct btstack btstack;
	struct btframe *parent;
	struct tblock *tblk = NULL;
	struct tlock *tlck = NULL;
	struct xtlock *xtlck = NULL;
	struct xdlistlock xadlock;	/* maplock for COMMIT_WMAP */
	struct pxd_lock *pxdlock;		/* maplock for COMMIT_WMAP */
	s64 nfreed;
	int freed, log;
	int locked_leaves = 0;

	/* save object truncation type */
	if (tid) {
		tblk = tid_to_tblock(tid);
		tblk->xflag |= flag;
	}

	nfreed = 0;

	flag &= COMMIT_MAP;
	assert(flag != COMMIT_PMAP);

	if (flag == COMMIT_PWMAP)
		log = 1;
	else {
		log = 0;
		xadlock.flag = mlckFREEXADLIST;
		xadlock.index = 1;
	}

	/*
	 * if the newsize is not an integral number of pages,
	 * the file between newsize and next page boundary will
	 * be cleared.
	 * if truncating into a file hole, it will cause
	 * a full block to be allocated for the logical block.
	 */

	/*
	 * release page blocks of truncated region <teof, eof>
	 *
	 * free the data blocks from the leaf index blocks.
	 * delete the parent index entries corresponding to
	 * the freed child data/index blocks.
	 * free the index blocks themselves which aren't needed
	 * in new sized file.
	 *
	 * index blocks are updated only if the blocks are to be
	 * retained in the new sized file.
	 * if type is PMAP, the data and index pages are NOT
	 * freed, and the data and index blocks are NOT freed
	 * from working map.
	 * (this will allow continued access of data/index of
	 * temporary file (zerolink count file truncated to zero-length)).
	 */
	teof = (newsize + (JFS_SBI(ip->i_sb)->bsize - 1)) >>
	    JFS_SBI(ip->i_sb)->l2bsize;

	/* clear stack */
	BT_CLR(&btstack);

	/*
	 * start with root
	 *
	 * root resides in the inode
	 */
	bn = 0;

	/*
	 * first access of each page:
	 */
      getPage:
	XT_GETPAGE(ip, bn, mp, PSIZE, p, rc);
	if (rc)
		return rc;

	/* process entries backward from last index */
	index = le16_to_cpu(p->header.nextindex) - 1;


	/* Since this is the rightmost page at this level, and we may have
	 * already freed a page that was formerly to the right, let's make
	 * sure that the next pointer is zero.
	 */
	if (p->header.next) {
		if (log)
			/*
			 * Make sure this change to the header is logged.
			 * If we really truncate this leaf, the flag
			 * will be changed to tlckTRUNCATE
			 */
			tlck = txLock(tid, ip, mp, tlckXTREE|tlckGROW);
		BT_MARK_DIRTY(mp, ip);
		p->header.next = 0;
	}

	if (p->header.flag & BT_INTERNAL)
		goto getChild;

	/*
	 *	leaf page
	 */
	freed = 0;

	/* does region covered by leaf page precede Teof ? */
	xad = &p->xad[index];
	xoff = offsetXAD(xad);
	xlen = lengthXAD(xad);
	if (teof >= xoff + xlen) {
		XT_PUTPAGE(mp);
		goto getParent;
	}

	/* (re)acquire tlock of the leaf page */
	if (log) {
		if (++locked_leaves > MAX_TRUNCATE_LEAVES) {
			/*
			 * We need to limit the size of the transaction
			 * to avoid exhausting pagecache & tlocks
			 */
			XT_PUTPAGE(mp);
			newsize = (xoff + xlen) << JFS_SBI(ip->i_sb)->l2bsize;
			goto getParent;
		}
		tlck = txLock(tid, ip, mp, tlckXTREE);
		tlck->type = tlckXTREE | tlckTRUNCATE;
		xtlck = (struct xtlock *) & tlck->lock;
		xtlck->hwm.offset = le16_to_cpu(p->header.nextindex) - 1;
	}
	BT_MARK_DIRTY(mp, ip);

	/*
	 * scan backward leaf page entries
	 */
	for (; index >= XTENTRYSTART; index--) {
		xad = &p->xad[index];
		xoff = offsetXAD(xad);
		xlen = lengthXAD(xad);
		xaddr = addressXAD(xad);

		/*
		 * The "data" for a directory is indexed by the block
		 * device's address space.  This metadata must be invalidated
		 * here
		 */
		if (S_ISDIR(ip->i_mode) && (teof == 0))
			invalidate_xad_metapages(ip, *xad);
		/*
		 * entry beyond eof: continue scan of current page
		 *          xad
		 * ---|---=======------->
		 *   eof
		 */
		if (teof < xoff) {
			nfreed += xlen;
			continue;
		}

		/*
		 * (xoff <= teof): last entry to be deleted from page;
		 * If other entries remain in page: keep and update the page.
		 */

		/*
		 * eof == entry_start: delete the entry
		 *           xad
		 * -------|=======------->
		 *       eof
		 *
		 */
		if (teof == xoff) {
			nfreed += xlen;

			if (index == XTENTRYSTART)
				break;

			nextindex = index;
		}
		/*
		 * eof within the entry: truncate the entry.
		 *          xad
		 * -------===|===------->
		 *          eof
		 */
		else if (teof < xoff + xlen) {
			/* update truncated entry */
			len = teof - xoff;
			freexlen = xlen - len;
			XADlength(xad, len);

			/* save pxd of truncated extent in tlck */
			xaddr += len;
			if (log) {	/* COMMIT_PWMAP */
				xtlck->lwm.offset = (xtlck->lwm.offset) ?
				    min(index, (int)xtlck->lwm.offset) : index;
				xtlck->lwm.length = index + 1 -
				    xtlck->lwm.offset;
				xtlck->twm.offset = index;
				pxdlock = (struct pxd_lock *) & xtlck->pxdlock;
				pxdlock->flag = mlckFREEPXD;
				PXDaddress(&pxdlock->pxd, xaddr);
				PXDlength(&pxdlock->pxd, freexlen);
			}
			/* free truncated extent */
			else {	/* COMMIT_WMAP */

				pxdlock = (struct pxd_lock *) & xadlock;
				pxdlock->flag = mlckFREEPXD;
				PXDaddress(&pxdlock->pxd, xaddr);
				PXDlength(&pxdlock->pxd, freexlen);
				txFreeMap(ip, pxdlock, NULL, COMMIT_WMAP);

				/* reset map lock */
				xadlock.flag = mlckFREEXADLIST;
			}

			/* current entry is new last entry; */
			nextindex = index + 1;

			nfreed += freexlen;
		}
		/*
		 * eof beyond the entry:
		 *          xad
		 * -------=======---|--->
		 *                 eof
		 */
		else {		/* (xoff + xlen < teof) */

			nextindex = index + 1;
		}

		if (nextindex < le16_to_cpu(p->header.nextindex)) {
			if (!log) {	/* COMMIT_WAMP */
				xadlock.xdlist = &p->xad[nextindex];
				xadlock.count =
				    le16_to_cpu(p->header.nextindex) -
				    nextindex;
				txFreeMap(ip, (struct maplock *) & xadlock,
					  NULL, COMMIT_WMAP);
			}
			p->header.nextindex = cpu_to_le16(nextindex);
		}

		XT_PUTPAGE(mp);

		/* assert(freed == 0); */
		goto getParent;
	}			/* end scan of leaf page entries */

	freed = 1;

	/*
	 * leaf page become empty: free the page if type != PMAP
	 */
	if (log) {		/* COMMIT_PWMAP */
		/* txCommit() with tlckFREE:
		 * free data extents covered by leaf [XTENTRYSTART:hwm);
		 * invalidate leaf if COMMIT_PWMAP;
		 * if (TRUNCATE), will write LOG_NOREDOPAGE;
		 */
		tlck->type = tlckXTREE | tlckFREE;
	} else {		/* COMMIT_WAMP */

		/* free data extents covered by leaf */
		xadlock.xdlist = &p->xad[XTENTRYSTART];
		xadlock.count =
		    le16_to_cpu(p->header.nextindex) - XTENTRYSTART;
		txFreeMap(ip, (struct maplock *) & xadlock, NULL, COMMIT_WMAP);
	}

	if (p->header.flag & BT_ROOT) {
		p->header.flag &= ~BT_INTERNAL;
		p->header.flag |= BT_LEAF;
		p->header.nextindex = cpu_to_le16(XTENTRYSTART);

		XT_PUTPAGE(mp);	/* debug */
		goto out;
	} else {
		if (log) {	/* COMMIT_PWMAP */
			/* page will be invalidated at tx completion
			 */
			XT_PUTPAGE(mp);
		} else {	/* COMMIT_WMAP */

			if (mp->lid)
				lid_to_tlock(mp->lid)->flag |= tlckFREELOCK;

			/* invalidate empty leaf page */
			discard_metapage(mp);
		}
	}

	/*
	 * the leaf page become empty: delete the parent entry
	 * for the leaf page if the parent page is to be kept
	 * in the new sized file.
	 */

	/*
	 * go back up to the parent page
	 */
      getParent:
	/* pop/restore parent entry for the current child page */
	if ((parent = BT_POP(&btstack)) == NULL)
		/* current page must have been root */
		goto out;

	/* get back the parent page */
	bn = parent->bn;
	XT_GETPAGE(ip, bn, mp, PSIZE, p, rc);
	if (rc)
		return rc;

	index = parent->index;

	/*
	 * child page was not empty:
	 */
	if (freed == 0) {
		/* has any entry deleted from parent ? */
		if (index < le16_to_cpu(p->header.nextindex) - 1) {
			/* (re)acquire tlock on the parent page */
			if (log) {	/* COMMIT_PWMAP */
				/* txCommit() with tlckTRUNCATE:
				 * free child extents covered by parent [);
				 */
				tlck = txLock(tid, ip, mp, tlckXTREE);
				xtlck = (struct xtlock *) & tlck->lock;
				if (!(tlck->type & tlckTRUNCATE)) {
					xtlck->hwm.offset =
					    le16_to_cpu(p->header.
							nextindex) - 1;
					tlck->type =
					    tlckXTREE | tlckTRUNCATE;
				}
			} else {	/* COMMIT_WMAP */

				/* free child extents covered by parent */
				xadlock.xdlist = &p->xad[index + 1];
				xadlock.count =
				    le16_to_cpu(p->header.nextindex) -
				    index - 1;
				txFreeMap(ip, (struct maplock *) & xadlock,
					  NULL, COMMIT_WMAP);
			}
			BT_MARK_DIRTY(mp, ip);

			p->header.nextindex = cpu_to_le16(index + 1);
		}
		XT_PUTPAGE(mp);
		goto getParent;
	}

	/*
	 * child page was empty:
	 */
	nfreed += lengthXAD(&p->xad[index]);

	/*
	 * During working map update, child page's tlock must be handled
	 * before parent's.  This is because the parent's tlock will cause
	 * the child's disk space to be marked available in the wmap, so
	 * it's important that the child page be released by that time.
	 *
	 * ToDo:  tlocks should be on doubly-linked list, so we can
	 * quickly remove it and add it to the end.
	 */

	/*
	 * Move parent page's tlock to the end of the tid's tlock list
	 */
	if (log && mp->lid && (tblk->last != mp->lid) &&
	    lid_to_tlock(mp->lid)->tid) {
		lid_t lid = mp->lid;
		struct tlock *prev;

		tlck = lid_to_tlock(lid);

		if (tblk->next == lid)
			tblk->next = tlck->next;
		else {
			for (prev = lid_to_tlock(tblk->next);
			     prev->next != lid;
			     prev = lid_to_tlock(prev->next)) {
				assert(prev->next);
			}
			prev->next = tlck->next;
		}
		lid_to_tlock(tblk->last)->next = lid;
		tlck->next = 0;
		tblk->last = lid;
	}

	/*
	 * parent page become empty: free the page
	 */
	if (index == XTENTRYSTART) {
		if (log) {	/* COMMIT_PWMAP */
			/* txCommit() with tlckFREE:
			 * free child extents covered by parent;
			 * invalidate parent if COMMIT_PWMAP;
			 */
			tlck = txLock(tid, ip, mp, tlckXTREE);
			xtlck = (struct xtlock *) & tlck->lock;
			xtlck->hwm.offset =
			    le16_to_cpu(p->header.nextindex) - 1;
			tlck->type = tlckXTREE | tlckFREE;
		} else {	/* COMMIT_WMAP */

			/* free child extents covered by parent */
			xadlock.xdlist = &p->xad[XTENTRYSTART];
			xadlock.count =
			    le16_to_cpu(p->header.nextindex) -
			    XTENTRYSTART;
			txFreeMap(ip, (struct maplock *) & xadlock, NULL,
				  COMMIT_WMAP);
		}
		BT_MARK_DIRTY(mp, ip);

		if (p->header.flag & BT_ROOT) {
			p->header.flag &= ~BT_INTERNAL;
			p->header.flag |= BT_LEAF;
			p->header.nextindex = cpu_to_le16(XTENTRYSTART);
			if (le16_to_cpu(p->header.maxentry) == XTROOTMAXSLOT) {
				/*
				 * Shrink root down to allow inline
				 * EA (otherwise fsck complains)
				 */
				p->header.maxentry =
				    cpu_to_le16(XTROOTINITSLOT);
				JFS_IP(ip)->mode2 |= INLINEEA;
			}

			XT_PUTPAGE(mp);	/* debug */
			goto out;
		} else {
			if (log) {	/* COMMIT_PWMAP */
				/* page will be invalidated at tx completion
				 */
				XT_PUTPAGE(mp);
			} else {	/* COMMIT_WMAP */

				if (mp->lid)
					lid_to_tlock(mp->lid)->flag |=
						tlckFREELOCK;

				/* invalidate parent page */
				discard_metapage(mp);
			}

			/* parent has become empty and freed:
			 * go back up to its parent page
			 */
			/* freed = 1; */
			goto getParent;
		}
	}
	/*
	 * parent page still has entries for front region;
	 */
	else {
		/* try truncate region covered by preceding entry
		 * (process backward)
		 */
		index--;

		/* go back down to the child page corresponding
		 * to the entry
		 */
		goto getChild;
	}

	/*
	 *	internal page: go down to child page of current entry
	 */
      getChild:
	/* save current parent entry for the child page */
	if (BT_STACK_FULL(&btstack)) {
		jfs_error(ip->i_sb, "stack overrun!\n");
		XT_PUTPAGE(mp);
		return -EIO;
	}
	BT_PUSH(&btstack, bn, index);

	/* get child page */
	xad = &p->xad[index];
	bn = addressXAD(xad);

	/*
	 * first access of each internal entry:
	 */
	/* release parent page */
	XT_PUTPAGE(mp);

	/* process the child page */
	goto getPage;

      out:
	/*
	 * update file resource stat
	 */
	/* set size
	 */
	if (S_ISDIR(ip->i_mode) && !newsize)
		ip->i_size = 1;	/* fsck hates zero-length directories */
	else
		ip->i_size = newsize;

	/* update quota allocation to reflect freed blocks */
	dquot_free_block(ip, nfreed);

	/*
	 * free tlock of invalidated pages
	 */
	if (flag == COMMIT_WMAP)
		txFreelock(ip);

	return newsize;
}


/*
 *	xtTruncate_pmap()
 *
 * function:
 *	Perform truncate to zero length for deleted file, leaving the
 *	the xtree and working map untouched.  This allows the file to
 *	be accessed via open file handles, while the delete of the file
 *	is committed to disk.
 *
 * parameter:
 *	tid_t		tid,
 *	struct inode	*ip,
 *	s64		committed_size)
 *
 * return: new committed size
 *
 * note:
 *
 *	To avoid deadlock by holding too many transaction locks, the
 *	truncation may be broken up into multiple transactions.
 *	The committed_size keeps track of part of the file has been
 *	freed from the pmaps.
 */
s64 xtTruncate_pmap(tid_t tid, struct inode *ip, s64 committed_size)
{
	s64 bn;
	struct btstack btstack;
	int cmp;
	int index;
	int locked_leaves = 0;
	struct metapage *mp;
	xtpage_t *p;
	struct btframe *parent;
	int rc;
	struct tblock *tblk;
	struct tlock *tlck = NULL;
	xad_t *xad;
	int xlen;
	s64 xoff;
	struct xtlock *xtlck = NULL;

	/* save object truncation type */
	tblk = tid_to_tblock(tid);
	tblk->xflag |= COMMIT_PMAP;

	/* clear stack */
	BT_CLR(&btstack);

	if (committed_size) {
		xoff = (committed_size >> JFS_SBI(ip->i_sb)->l2bsize) - 1;
		rc = xtSearch(ip, xoff, NULL, &cmp, &btstack, 0);
		if (rc)
			return rc;

		XT_GETSEARCH(ip, btstack.top, bn, mp, p, index);

		if (cmp != 0) {
			XT_PUTPAGE(mp);
			jfs_error(ip->i_sb, "did not find extent\n");
			return -EIO;
		}
	} else {
		/*
		 * start with root
		 *
		 * root resides in the inode
		 */
		bn = 0;

		/*
		 * first access of each page:
		 */
      getPage:
		XT_GETPAGE(ip, bn, mp, PSIZE, p, rc);
		if (rc)
			return rc;

		/* process entries backward from last index */
		index = le16_to_cpu(p->header.nextindex) - 1;

		if (p->header.flag & BT_INTERNAL)
			goto getChild;
	}

	/*
	 *	leaf page
	 */

	if (++locked_leaves > MAX_TRUNCATE_LEAVES) {
		/*
		 * We need to limit the size of the transaction
		 * to avoid exhausting pagecache & tlocks
		 */
		xad = &p->xad[index];
		xoff = offsetXAD(xad);
		xlen = lengthXAD(xad);
		XT_PUTPAGE(mp);
		return (xoff + xlen) << JFS_SBI(ip->i_sb)->l2bsize;
	}
	tlck = txLock(tid, ip, mp, tlckXTREE);
	tlck->type = tlckXTREE | tlckFREE;
	xtlck = (struct xtlock *) & tlck->lock;
	xtlck->hwm.offset = index;


	XT_PUTPAGE(mp);

	/*
	 * go back up to the parent page
	 */
      getParent:
	/* pop/restore parent entry for the current child page */
	if ((parent = BT_POP(&btstack)) == NULL)
		/* current page must have been root */
		goto out;

	/* get back the parent page */
	bn = parent->bn;
	XT_GETPAGE(ip, bn, mp, PSIZE, p, rc);
	if (rc)
		return rc;

	index = parent->index;

	/*
	 * parent page become empty: free the page
	 */
	if (index == XTENTRYSTART) {
		/* txCommit() with tlckFREE:
		 * free child extents covered by parent;
		 * invalidate parent if COMMIT_PWMAP;
		 */
		tlck = txLock(tid, ip, mp, tlckXTREE);
		xtlck = (struct xtlock *) & tlck->lock;
		xtlck->hwm.offset = le16_to_cpu(p->header.nextindex) - 1;
		tlck->type = tlckXTREE | tlckFREE;

		XT_PUTPAGE(mp);

		if (p->header.flag & BT_ROOT) {

			goto out;
		} else {
			goto getParent;
		}
	}
	/*
	 * parent page still has entries for front region;
	 */
	else
		index--;
	/*
	 *	internal page: go down to child page of current entry
	 */
      getChild:
	/* save current parent entry for the child page */
	if (BT_STACK_FULL(&btstack)) {
		jfs_error(ip->i_sb, "stack overrun!\n");
		XT_PUTPAGE(mp);
		return -EIO;
	}
	BT_PUSH(&btstack, bn, index);

	/* get child page */
	xad = &p->xad[index];
	bn = addressXAD(xad);

	/*
	 * first access of each internal entry:
	 */
	/* release parent page */
	XT_PUTPAGE(mp);

	/* process the child page */
	goto getPage;

      out:

	return 0;
}

#ifdef CONFIG_JFS_STATISTICS
static int jfs_xtstat_proc_show(struct seq_file *m, void *v)
{
	seq_printf(m,
		       "JFS Xtree statistics\n"
		       "====================\n"
		       "searches = %d\n"
		       "fast searches = %d\n"
		       "splits = %d\n",
		       xtStat.search,
		       xtStat.fastSearch,
		       xtStat.split);
	return 0;
}

static int jfs_xtstat_proc_open(struct inode *inode, struct file *file)
{
	return single_open(file, jfs_xtstat_proc_show, NULL);
}

const struct file_operations jfs_xtstat_proc_fops = {
	.owner		= THIS_MODULE,
	.open		= jfs_xtstat_proc_open,
	.read		= seq_read,
	.llseek		= seq_lseek,
	.release	= single_release,
};
#endif