summaryrefslogtreecommitdiffstatshomepage
path: root/scripts/src/netlist.lua
Commit message (Expand)AuthorAgeFilesLines
* -netlist: Added 82S123 256-bit (32x8) TTL bipolar PROM. [Ryan Holtz] MooglyGuy2016-12-241-0/+2
* -netlist: Added 82S126 4kbit (512x8) TTL bipolar PROM. [Ryan Holtz] therealmogminer@gmail.com2016-12-231-0/+2
* netlist: Added 4316 bilateral switch pack. Vas Crabb2016-12-231-0/+2
* -netlist: Added 74165, 74194, 7475, 7485, and DM9322 devices. [Ryan Holtz] therealmogminer@gmail.com2016-12-221-0/+10
* -netlist: Added shared RAM pointer, for use by netlist RAM devices which need... therealmogminer@gmail.com2016-12-211-0/+6
* -netlist: Added 74166 Parallel-Load 8-Bit Shift Register device. [Ryan Holtz] therealmogminer@gmail.com2016-12-211-0/+2
* -netlist: Added 2716 16 Kbit (2 Kbit x 8) UV EPROM. [Ryan Holtz] therealmogminer@gmail.com2016-12-211-0/+2
* -netlist: Added 74174 Hex D-Type Flip-Flop with Clear [Ryan Holtz] therealmogminer@gmail.com2016-12-211-0/+2
* netlist: Added 82S126 1K-bit bipolar PROM device. [Ryan Holtz] therealmogminer@gmail.com2016-12-211-0/+2
* netlist: Device additions: [Ryan Holtz] therealmogminer@gmail.com2016-12-211-0/+8
* Add Lua-cleaning ability to srcclean Vas Crabb2016-08-301-11/+11
* fix for Android build (nw) Miodrag Milanovic2016-06-281-0/+5
* Netlist changes couriersud2016-06-231-0/+3
* - More code cleanup. couriersud2016-06-071-1/+2
* - Minor cosmetic refactoring. Added a "uninitialised array" template to balr0g2016-06-071-1/+2
* Converted 7486 to macro module couriersud2016-05-051-2/+0
* Removed drivers/signal.h since netlist devices no longer use it. (nw) couriersud2016-05-051-1/+0
* Convert 7430 to macro module. couriersud2016-05-051-2/+0
* Convert 7425 to macro modul couriersud2016-05-051-2/+0
* Converted 7420 to macro modul couriersud2016-05-051-2/+0
* Convert 7432 to macro device. couriersud2016-05-041-2/+0
* Converted 7437 to macro device couriersud2016-05-041-2/+0
* Converted 7427 to macro device couriersud2016-05-041-2/+0
* Converted 7411 into a macro device. couriersud2016-05-041-2/+0
* Converted 7410 into macro device. couriersud2016-05-041-2/+0
* Implement 7402, 7408 as macro device. (nw) couriersud2016-05-031-4/+0
* Implement 7404 as macro device. (nw) couriersud2016-05-031-2/+0
* Implemented dynamic loading of precompiled solver code. [Couriersud] couriersud2016-04-231-0/+2
* Visual Studio 2015 netlist fix. (nw) dankan18902016-03-301-1/+2
* scrclean on lua scripts (nw) Miodrag Milanovic2016-03-291-3/+3
* Extend system library support (nw) Jeffrey Clark2016-03-281-6/+1
* Visual Studio compile fix. (nw) dankan18902016-03-281-1/+2
* netlist: couriersud2016-03-271-0/+1
* Netlist: code maintenance couriersud2016-03-231-2/+0
* Added addprojectflags function and added -Wsuggest-override for all that are ... Miodrag Milanovic2015-12-061-0/+2
* Added info header at the top of each .lua file (nw) Scott Stone2015-11-141-0/+8
* added more h files missing in lua scripts (nw) Miodrag Milanovic2015-11-121-4/+7
* Remove not longer needed ForceCpp flag (nw) Miodrag Milanovic2015-11-081-4/+0
* update build scripts for rename c - > cpp (nw) Miodrag Milanovic2015-11-081-63/+63
* typo fix (nw) Miodrag Milanovic2015-11-081-1/+1
* Remove last traces of printf from plib. Reorganized some code. (nw) couriersud2015-09-141-0/+3
* Moved netlist from emu to lib (nw) Miodrag Milanovic2015-09-121-141/+141
* Added simple stream classes to netlist code. (nw) couriersud2015-08-031-0/+2
* Move MC14584B to nlm_other.[ch] (nw) couriersud2015-07-271-0/+2
* Add macro library for opamp models. (nw) couriersud2015-07-161-0/+2
* Changed uppercase filenames to lowercase. (nw) couriersud2015-07-121-4/+4
* Created macro libraries for CD4XXX and 74XX chip families. Going couriersud2015-07-121-0/+4
* Added MM5837 Noise Generator used by Congo Bongo. [Couriersud] couriersud2015-07-041-0/+2
* no need for subdir for these projects, those are same for all mame builds (nw) Miodrag Milanovic2015-06-221-1/+0
* Move solver code into own folder. (nw) couriersud2015-06-191-6/+8
href='#n926'>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 3904 3905 3906 3907 3908 3909 3910 3911 3912 3913 3914 3915 3916 3917 3918 3919 3920 3921 3922 3923 3924 3925 3926 3927 3928 3929 3930 3931 3932 3933 3934 3935 3936 3937 3938 3939 3940 3941 3942 3943 3944 3945 3946 3947 3948 3949 3950 3951 3952 3953 3954 3955 3956 3957 3958 3959 3960 3961 3962 3963 3964 3965 3966 3967 3968 3969 3970 3971 3972 3973 3974 3975 3976 3977 3978 3979 3980 3981 3982 3983 3984 3985 3986 3987 3988 3989 3990 3991 3992 3993 3994 3995 3996 3997 3998 3999 4000 4001 4002 4003 4004 4005 4006 4007 4008 4009 4010 4011 4012 4013 4014 4015 4016 4017 4018 4019 4020 4021 4022 4023 4024 4025 4026 4027 4028 4029 4030 4031 4032 4033 4034 4035 4036 4037 4038 4039 4040 4041 4042 4043 4044 4045 4046 4047 4048 4049 4050 4051 4052 4053 4054 4055 4056 4057 4058 4059 4060 4061 4062 4063 4064 4065 4066 4067 4068 4069 4070 4071 4072 4073 4074 4075 4076 4077 4078 4079 4080 4081 4082 4083 4084 4085 4086 4087 4088 4089 4090 4091 4092 4093 4094 4095 4096 4097 4098 4099 4100 4101 4102 4103 4104 4105 4106 4107 4108 4109 4110 4111 4112 4113 4114 4115 4116 4117 4118 4119 4120 4121 4122 4123 4124 4125 4126 4127 4128 4129 4130 4131 4132 4133 4134 4135 4136 4137 4138 4139 4140 4141 4142 4143 4144 4145 4146 4147 4148 4149 4150 4151 4152 4153 4154 4155 4156 4157 4158 4159 4160 4161 4162 4163 4164 4165 4166 4167 4168 4169 4170 4171 4172 4173 4174 4175 4176 4177 4178 4179 4180 4181 4182 4183 4184 4185 4186 4187 4188 4189 4190 4191 4192 4193 4194 4195 4196 4197 4198 4199 4200 4201 4202 4203 4204 4205 4206 4207 4208 4209 4210 4211 4212 4213 4214 4215 4216 4217 4218 4219 4220 4221 4222 4223 4224 4225 4226 4227 4228 4229 4230 4231 4232 4233 4234 4235 4236 4237 4238 4239 4240 4241 4242 4243 4244 4245 4246 4247 4248 4249 4250 4251 4252 4253 4254 4255 4256 4257 4258 4259 4260 4261 4262 4263 4264 4265 4266 4267 4268 4269 4270 4271 4272 4273 4274 4275 4276 4277 4278 4279 4280 4281 4282 4283 4284 4285 4286 4287 4288 4289 4290 4291 4292 4293 4294 4295 4296 4297 4298 4299 4300 4301 4302 4303 4304 4305 4306 4307 4308 4309 4310 4311 4312 4313 4314 4315 4316 4317 4318 4319 4320 4321 4322 4323 4324 4325 4326 4327 4328 4329 4330 4331 4332 4333 4334 4335 4336 4337 4338 4339 4340 4341 4342 4343 4344 4345 4346 4347 4348 4349 4350 4351 4352 4353 4354 4355 4356 4357 4358 4359 4360 4361 4362 4363 4364 4365 4366 4367 4368 4369 4370 4371 4372 4373 4374 4375 4376 4377 4378 4379 4380 4381 4382 4383 4384 4385 4386 4387 4388 4389 4390 4391 4392 4393 4394 4395 4396 4397 4398 4399 4400 4401 4402 4403 4404 4405 4406 4407 4408 4409 4410 4411 4412 4413 4414 4415 4416 4417 4418 4419 4420 4421 4422 4423 4424 4425 4426 4427 4428 4429 4430 4431 4432 4433 4434 4435 4436 4437 4438 4439 4440 4441 4442 4443 4444 4445 4446 4447 4448 4449 4450 4451 4452 4453 4454 4455 4456 4457 4458 4459 4460 4461 4462 4463 4464 4465 4466 4467 4468 4469 4470 4471 4472 4473 4474 4475 4476 4477 4478 4479 4480 4481 4482 4483 4484 4485 4486 4487 4488 4489 4490 4491 4492 4493 4494 4495 4496 4497 4498 4499 4500 4501 4502 4503 4504 4505 4506 4507 4508 4509 4510 4511 4512 4513 4514 4515 4516 4517 4518 4519 4520 4521 4522 4523 4524 4525 4526 4527 4528 4529 4530 4531 4532 4533 4534 4535 4536 4537 4538 4539 4540 4541 4542 4543 4544 4545 4546 4547 4548 4549 4550 4551 4552 4553 4554 4555 4556 4557 4558 4559 4560 4561 4562 4563 4564 4565 4566 4567 4568 4569 4570 4571 4572 4573 4574 4575 4576 4577 4578 4579 4580 4581 4582 4583 4584 4585 4586 4587 4588 4589 4590 4591 4592 4593 4594 4595 4596 4597 4598 4599 4600 4601 4602 4603 4604 4605 4606 4607 4608 4609 4610 4611 4612 4613 4614 4615 4616 4617 4618 4619 4620 4621 4622 4623 4624 4625 4626 4627 4628 4629 4630 4631 4632 4633 4634 4635 4636 4637 4638 4639 4640 4641 4642 4643 4644 4645 4646 4647 4648 4649 4650 4651 4652 4653 4654 4655 4656 4657 4658 4659 4660 4661 4662 4663 4664 4665 4666 4667 4668 4669 4670 4671 4672 4673 4674 4675 4676 4677 4678 4679 4680 4681 4682 4683 4684 4685 4686 4687 4688 4689 4690 4691 4692 4693 4694 4695 4696 4697 4698 4699 4700 4701 4702 4703 4704 4705 4706 4707 4708 4709 4710 4711 4712 4713 4714 4715 4716 4717 4718 4719 4720 4721 4722 4723 4724 4725 4726 4727 4728 4729 4730 4731 4732 4733 4734 4735 4736 4737 4738 4739 4740 4741 4742 4743 4744 4745 4746 4747 4748 4749 4750 4751 4752 4753 4754 4755 4756 4757 4758 4759 4760 4761 4762 4763 4764 4765 4766 4767 4768 4769 4770 4771 4772 4773 4774 4775 4776 4777 4778 4779 4780 4781 4782 4783 4784 4785 4786 4787 4788 4789 4790 4791 4792 4793 4794 4795 4796 4797 4798 4799 4800 4801 4802 4803 4804 4805 4806 4807 4808 4809 4810 4811 4812 4813 4814 4815 4816 4817 4818 4819 4820 4821 4822 4823 4824 4825 4826 4827 4828 4829 4830 4831 4832 4833 4834 4835 4836 4837 4838 4839 4840 4841 4842 4843 4844 4845 4846 4847 4848 4849 4850 4851 4852 4853 4854 4855 4856 4857 4858 4859 4860 4861 4862 4863 4864 4865 4866 4867 4868 4869 4870 4871 4872 4873 4874 4875 4876 4877 4878 4879 4880 4881 4882 4883 4884 4885 4886 4887 4888 4889 4890 4891 4892 4893 4894 4895 4896 4897 4898 4899 4900 4901 4902 4903 4904 4905 4906 4907 4908 4909 4910 4911 4912 4913 4914 4915 4916 4917 4918 4919 4920 4921 4922 4923 4924 4925 4926 4927 4928 4929 4930 4931 4932 4933 4934 4935 4936 4937 4938 4939 4940 4941 4942 4943 4944 4945 4946 4947 4948 4949 4950 4951 4952 4953 4954 4955 4956 4957 4958 4959 4960 4961 4962 4963 4964 4965 4966 4967 4968 4969 4970 4971 4972 4973 4974 4975 4976 4977 4978 4979 4980 4981 4982 4983 4984 4985 4986 4987 4988 4989 4990 4991 4992 4993 4994 4995 4996 4997 4998 4999 5000 5001 5002 5003 5004 5005 5006 5007 5008 5009 5010 5011 5012 5013 5014 5015 5016 5017 5018 5019 5020 5021 5022 5023 5024 5025 5026 5027 5028 5029 5030 5031 5032 5033 5034 5035 5036 5037 5038 5039 5040 5041 5042 5043 5044 5045 5046 5047 5048 5049 5050 5051 5052 5053 5054 5055 5056 5057 5058 5059 5060 5061 5062 5063 5064 5065 5066 5067 5068 5069 5070 5071 5072 5073 5074 5075 5076 5077 5078 5079 5080 5081 5082 5083 5084 5085 5086 5087 5088 5089 5090 5091 5092 5093 5094 5095 5096 5097 5098 5099 5100 5101 5102 5103 5104 5105 5106 5107 5108 5109 5110 5111 5112 5113 5114 5115 5116 5117 5118 5119 5120 5121 5122 5123 5124 5125 5126 5127 5128 5129 5130 5131 5132 5133 5134 5135 5136 5137 5138 5139 5140 5141 5142 5143 5144 5145 5146 5147 5148 5149 5150 5151 5152 5153 5154 5155 5156 5157 5158 5159 5160 5161 5162 5163 5164 5165
// license:BSD-3-Clause
// copyright-holders:Aaron Giles
/***************************************************************************

    drcbex64.c

    64-bit x64 back-end for the universal machine language.

****************************************************************************

    Future improvements/changes:

    * Add support for FP registers

    * Optimize to avoid unnecessary reloads

    * Identify common pairs and optimize output

    * Convert SUB a,0,b to NEG

    * Optimize, e.g., and [r5],i0,$FF to use rbx as temporary register
        (avoid initial move) if i0 is not needed going forward

****************************************************************************

    -------------------------
    ABI/conventions (Windows)
    -------------------------

    Registers:
        RAX        - volatile, function return value
        RBX        - non-volatile
        RCX        - volatile, integer function parameter 1
        RDX        - volatile, integer function parameter 2
        RSI        - non-volatile
        RDI        - non-volatile
        RBP        - non-volatile
        R8         - volatile, integer function parameter 3
        R9         - volatile, integer function parameter 4
        R10        - volatile
        R11        - volatile, scratch immediate storage
        R12        - non-volatile
        R13        - non-volatile
        R14        - non-volatile
        R15        - non-volatile

        XMM0       - volatile, FP function parameter 1
        XMM1       - volatile, FP function parameter 2
        XMM2       - volatile, FP function parameter 3
        XMM3       - volatile, FP function parameter 4
        XMM4       - volatile
        XMM5       - volatile
        XMM6       - non-volatile
        XMM7       - non-volatile
        XMM8       - non-volatile
        XMM9       - non-volatile
        XMM10      - non-volatile
        XMM11      - non-volatile
        XMM12      - non-volatile
        XMM13      - non-volatile
        XMM14      - non-volatile
        XMM15      - non-volatile


    -----------------------------
    ABI/conventions (Linux/MacOS)
    -----------------------------

    Registers:
        RAX        - volatile, function return value
        RBX        - non-volatile
        RCX        - volatile, integer function parameter 4
        RDX        - volatile, integer function parameter 3
        RSI        - volatile, integer function parameter 2
        RDI        - volatile, integer function parameter 1
        RBP        - non-volatile
        R8         - volatile, integer function parameter 5
        R9         - volatile, integer function parameter 6
        R10        - volatile
        R11        - volatile, scratch immediate storage
        R12        - non-volatile
        R13        - non-volatile
        R14        - non-volatile
        R15        - non-volatile

        XMM0       - volatile, FP function parameter 1
        XMM1       - volatile, FP function parameter 2
        XMM2       - volatile, FP function parameter 3
        XMM3       - volatile, FP function parameter 4
        XMM4       - volatile
        XMM5       - volatile
        XMM6       - volatile
        XMM7       - volatile
        XMM8       - volatile
        XMM9       - volatile
        XMM10      - volatile
        XMM11      - volatile
        XMM12      - volatile
        XMM13      - volatile
        XMM14      - volatile
        XMM15      - volatile


    ---------------
    Execution model
    ---------------

    Registers (Windows):
        RAX        - scratch register
        RBX        - maps to I0
        RCX        - scratch register
        RDX        - scratch register
        RSI        - maps to I1
        RDI        - maps to I2
        RBP        - pointer to code cache
        R8         - scratch register
        R9         - scratch register
        R10        - scratch register
        R11        - scratch register
        R12        - maps to I3
        R13        - maps to I4
        R14        - maps to I5
        R15        - maps to I6

    Registers (Linux/MacOS):
        RAX        - scratch register
        RBX        - maps to I0
        RCX        - scratch register
        RDX        - scratch register
        RSI        - unused
        RDI        - unused
        RBP        - pointer to code cache
        R8         - scratch register
        R9         - scratch register
        R10        - scratch register
        R11        - scratch register
        R12        - maps to I1
        R13        - maps to I2
        R14        - maps to I3
        R15        - maps to I4

    Entry point:
        Assumes 1 parameter passed, which is the codeptr of the code
        to execute once the environment is set up.

    Exit point:
        Assumes exit value is in RAX.

    Entry stack:
        [rsp]      - return

    Runtime stack:
        [rsp]      - r9 home
        [rsp+8]    - r8 home
        [rsp+16]   - rdx home
        [rsp+24]   - rcx home
        [rsp+40]   - saved r15
        [rsp+48]   - saved r14
        [rsp+56]   - saved r13
        [rsp+64]   - saved r12
        [rsp+72]   - saved ebp
        [rsp+80]   - saved edi
        [rsp+88]   - saved esi
        [rsp+96]   - saved ebx
        [rsp+104]  - ret

***************************************************************************/

#include "emu.h"
#include "drcbex64.h"

#include "debugger.h"
#include "emuopts.h"

#include <cstddef>


// This is a trick to make it build on Android where the ARM SDK declares ::REG_Rn
// and the x64 SDK declares ::REG_Exx and ::REG_Rxx
namespace drc {

using namespace uml;

using namespace asmjit;
using namespace asmjit::x86;



//**************************************************************************
//  DEBUGGING
//**************************************************************************

#define LOG_HASHJMPS            (0)

#define USE_RCPSS_FOR_SINGLES   (0)
#define USE_RSQRTSS_FOR_SINGLES (0)
#define USE_RCPSS_FOR_DOUBLES   (0)
#define USE_RSQRTSS_FOR_DOUBLES (0)



//**************************************************************************
//  CONSTANTS
//**************************************************************************

const uint32_t PTYPE_M    = 1 << parameter::PTYPE_MEMORY;
const uint32_t PTYPE_I    = 1 << parameter::PTYPE_IMMEDIATE;
const uint32_t PTYPE_R    = 1 << parameter::PTYPE_INT_REGISTER;
const uint32_t PTYPE_F    = 1 << parameter::PTYPE_FLOAT_REGISTER;
//const uint32_t PTYPE_MI   = PTYPE_M | PTYPE_I;
//const uint32_t PTYPE_RI   = PTYPE_R | PTYPE_I;
const uint32_t PTYPE_MR   = PTYPE_M | PTYPE_R;
const uint32_t PTYPE_MRI  = PTYPE_M | PTYPE_R | PTYPE_I;
const uint32_t PTYPE_MF   = PTYPE_M | PTYPE_F;

#ifdef X64_WINDOWS_ABI

const Gp::Id REG_PARAM1    = Gp::kIdCx;
const Gp::Id REG_PARAM2    = Gp::kIdDx;
const Gp::Id REG_PARAM3    = Gp::kIdR8;
const Gp::Id REG_PARAM4    = Gp::kIdR9;

#else

const Gp::Id REG_PARAM1    = Gp::kIdDi;
const Gp::Id REG_PARAM2    = Gp::kIdSi;
const Gp::Id REG_PARAM3    = Gp::kIdDx;
const Gp::Id REG_PARAM4    = Gp::kIdCx;

#endif



//**************************************************************************
//  MACROS
//**************************************************************************

#define X86_CONDITION(condition)        (condition_map[condition - uml::COND_Z])
#define X86_NOT_CONDITION(condition)    negateCond(condition_map[condition - uml::COND_Z])

#define assert_no_condition(inst)       assert((inst).condition() == uml::COND_ALWAYS)
#define assert_any_condition(inst)      assert((inst).condition() == uml::COND_ALWAYS || ((inst).condition() >= uml::COND_Z && (inst).condition() < uml::COND_MAX))
#define assert_no_flags(inst)           assert((inst).flags() == 0)
#define assert_flags(inst, valid)       assert(((inst).flags() & ~(valid)) == 0)



//**************************************************************************
//  GLOBAL VARIABLES
//**************************************************************************

drcbe_x64::opcode_generate_func drcbe_x64::s_opcode_table[OP_MAX];

// size-to-mask table
//static const uint64_t size_to_mask[] = { 0, 0xff, 0xffff, 0, 0xffffffff, 0, 0, 0, 0xffffffffffffffffU };

// register mapping tables
static const Gp::Id int_register_map[REG_I_COUNT] =
{
#ifdef X64_WINDOWS_ABI
	Gp::kIdBx, Gp::kIdSi, Gp::kIdDi, Gp::kIdR12, Gp::kIdR13, Gp::kIdR14, Gp::kIdR15,
#else
	Gp::kIdBx, Gp::kIdR12, Gp::kIdR13, Gp::kIdR14, Gp::kIdR15
#endif
};

static uint32_t float_register_map[REG_F_COUNT] =
{
#ifdef X64_WINDOWS_ABI
	6, 7, 8, 9, 10, 11, 12, 13, 14, 15
#else
	// on AMD x64 ABI, XMM0-7 are FP function args.  since this code has no args, and we
	// save/restore them around CALLC, they should be safe for our use.
	0, 1, 2, 3, 4, 5, 6, 7
#endif
};

// condition mapping table
static const CondCode condition_map[uml::COND_MAX - uml::COND_Z] =
{
	CondCode::kZ,    // COND_Z = 0x80,    requires Z
	CondCode::kNZ,   // COND_NZ,          requires Z
	CondCode::kS,    // COND_S,           requires S
	CondCode::kNS,   // COND_NS,          requires S
	CondCode::kC,    // COND_C,           requires C
	CondCode::kNC,   // COND_NC,          requires C
	CondCode::kO,    // COND_V,           requires V
	CondCode::kNO,   // COND_NV,          requires V
	CondCode::kP,    // COND_U,           requires U
	CondCode::kNP,   // COND_NU,          requires U
	CondCode::kA,    // COND_A,           requires CZ
	CondCode::kBE,   // COND_BE,          requires CZ
	CondCode::kG,    // COND_G,           requires SVZ
	CondCode::kLE,   // COND_LE,          requires SVZ
	CondCode::kL,    // COND_L,           requires SV
	CondCode::kGE,   // COND_GE,          requires SV
};

#if 0
// rounding mode mapping table
static const uint8_t fprnd_map[4] =
{
	FPRND_CHOP,     // ROUND_TRUNC,   truncate
	FPRND_NEAR,     // ROUND_ROUND,   round
	FPRND_UP,       // ROUND_CEIL,    round up
	FPRND_DOWN      // ROUND_FLOOR    round down
};
#endif



//**************************************************************************
//  TABLES
//**************************************************************************

const drcbe_x64::opcode_table_entry drcbe_x64::s_opcode_table_source[] =
{
	// Compile-time opcodes
	{ uml::OP_HANDLE,  &drcbe_x64::op_handle },     // HANDLE  handle
	{ uml::OP_HASH,    &drcbe_x64::op_hash },       // HASH    mode,pc
	{ uml::OP_LABEL,   &drcbe_x64::op_label },      // LABEL   imm
	{ uml::OP_COMMENT, &drcbe_x64::op_comment },    // COMMENT string
	{ uml::OP_MAPVAR,  &drcbe_x64::op_mapvar },     // MAPVAR  mapvar,value

	// Control Flow Operations
	{ uml::OP_NOP,     &drcbe_x64::op_nop },        // NOP
	{ uml::OP_DEBUG,   &drcbe_x64::op_debug },      // DEBUG   pc
	{ uml::OP_EXIT,    &drcbe_x64::op_exit },       // EXIT    src1[,c]
	{ uml::OP_HASHJMP, &drcbe_x64::op_hashjmp },    // HASHJMP mode,pc,handle
	{ uml::OP_JMP,     &drcbe_x64::op_jmp },        // JMP     imm[,c]
	{ uml::OP_EXH,     &drcbe_x64::op_exh },        // EXH     handle,param[,c]
	{ uml::OP_CALLH,   &drcbe_x64::op_callh },      // CALLH   handle[,c]
	{ uml::OP_RET,     &drcbe_x64::op_ret },        // RET     [c]
	{ uml::OP_CALLC,   &drcbe_x64::op_callc },      // CALLC   func,ptr[,c]
	{ uml::OP_RECOVER, &drcbe_x64::op_recover },    // RECOVER dst,mapvar

	// Internal Register Operations
	{ uml::OP_SETFMOD, &drcbe_x64::op_setfmod },    // SETFMOD src
	{ uml::OP_GETFMOD, &drcbe_x64::op_getfmod },    // GETFMOD dst
	{ uml::OP_GETEXP,  &drcbe_x64::op_getexp },     // GETEXP  dst
	{ uml::OP_GETFLGS, &drcbe_x64::op_getflgs },    // GETFLGS dst[,f]
	{ uml::OP_SAVE,    &drcbe_x64::op_save },       // SAVE    dst
	{ uml::OP_RESTORE, &drcbe_x64::op_restore },    // RESTORE dst

	// Integer Operations
	{ uml::OP_LOAD,    &drcbe_x64::op_load },       // LOAD    dst,base,index,size
	{ uml::OP_LOADS,   &drcbe_x64::op_loads },      // LOADS   dst,base,index,size
	{ uml::OP_STORE,   &drcbe_x64::op_store },      // STORE   base,index,src,size
	{ uml::OP_READ,    &drcbe_x64::op_read },       // READ    dst,src1,spacesize
	{ uml::OP_READM,   &drcbe_x64::op_readm },      // READM   dst,src1,mask,spacesize
	{ uml::OP_WRITE,   &drcbe_x64::op_write },      // WRITE   dst,src1,spacesize
	{ uml::OP_WRITEM,  &drcbe_x64::op_writem },     // WRITEM  dst,src1,spacesize
	{ uml::OP_CARRY,   &drcbe_x64::op_carry },      // CARRY   src,bitnum
	{ uml::OP_SET,     &drcbe_x64::op_set },        // SET     dst,c
	{ uml::OP_MOV,     &drcbe_x64::op_mov },        // MOV     dst,src[,c]
	{ uml::OP_SEXT,    &drcbe_x64::op_sext },       // SEXT    dst,src
	{ uml::OP_ROLAND,  &drcbe_x64::op_roland },     // ROLAND  dst,src1,src2,src3
	{ uml::OP_ROLINS,  &drcbe_x64::op_rolins },     // ROLINS  dst,src1,src2,src3
	{ uml::OP_ADD,     &drcbe_x64::op_add },        // ADD     dst,src1,src2[,f]
	{ uml::OP_ADDC,    &drcbe_x64::op_addc },       // ADDC    dst,src1,src2[,f]
	{ uml::OP_SUB,     &drcbe_x64::op_sub },        // SUB     dst,src1,src2[,f]
	{ uml::OP_SUBB,    &drcbe_x64::op_subc },       // SUBB    dst,src1,src2[,f]
	{ uml::OP_CMP,     &drcbe_x64::op_cmp },        // CMP     src1,src2[,f]
	{ uml::OP_MULU,    &drcbe_x64::op_mulu },       // MULU    dst,edst,src1,src2[,f]
	{ uml::OP_MULS,    &drcbe_x64::op_muls },       // MULS    dst,edst,src1,src2[,f]
	{ uml::OP_DIVU,    &drcbe_x64::op_divu },       // DIVU    dst,edst,src1,src2[,f]
	{ uml::OP_DIVS,    &drcbe_x64::op_divs },       // DIVS    dst,edst,src1,src2[,f]
	{ uml::OP_AND,     &drcbe_x64::op_and },        // AND     dst,src1,src2[,f]
	{ uml::OP_TEST,    &drcbe_x64::op_test },       // TEST    src1,src2[,f]
	{ uml::OP_OR,      &drcbe_x64::op_or },         // OR      dst,src1,src2[,f]
	{ uml::OP_XOR,     &drcbe_x64::op_xor },        // XOR     dst,src1,src2[,f]
	{ uml::OP_LZCNT,   &drcbe_x64::op_lzcnt },      // LZCNT   dst,src[,f]
	{ uml::OP_TZCNT,   &drcbe_x64::op_tzcnt },      // TZCNT   dst,src[,f]
	{ uml::OP_BSWAP,   &drcbe_x64::op_bswap },      // BSWAP   dst,src
	{ uml::OP_SHL,     &drcbe_x64::op_shift<Inst::kIdShl> },        // SHL     dst,src,count[,f]
	{ uml::OP_SHR,     &drcbe_x64::op_shift<Inst::kIdShr> },        // SHR     dst,src,count[,f]
	{ uml::OP_SAR,     &drcbe_x64::op_shift<Inst::kIdSar> },        // SAR     dst,src,count[,f]
	{ uml::OP_ROL,     &drcbe_x64::op_shift<Inst::kIdRol> },        // ROL     dst,src,count[,f]
	{ uml::OP_ROLC,    &drcbe_x64::op_shift<Inst::kIdRcl> },        // ROLC    dst,src,count[,f]
	{ uml::OP_ROR,     &drcbe_x64::op_shift<Inst::kIdRor> },        // ROR     dst,src,count[,f]
	{ uml::OP_RORC,    &drcbe_x64::op_shift<Inst::kIdRcr> },        // RORC    dst,src,count[,f]

	// Floating Point Operations
	{ uml::OP_FLOAD,   &drcbe_x64::op_fload },      // FLOAD   dst,base,index
	{ uml::OP_FSTORE,  &drcbe_x64::op_fstore },     // FSTORE  base,index,src
	{ uml::OP_FREAD,   &drcbe_x64::op_fread },      // FREAD   dst,space,src1
	{ uml::OP_FWRITE,  &drcbe_x64::op_fwrite },     // FWRITE  space,dst,src1
	{ uml::OP_FMOV,    &drcbe_x64::op_fmov },       // FMOV    dst,src1[,c]
	{ uml::OP_FTOINT,  &drcbe_x64::op_ftoint },     // FTOINT  dst,src1,size,round
	{ uml::OP_FFRINT,  &drcbe_x64::op_ffrint },     // FFRINT  dst,src1,size
	{ uml::OP_FFRFLT,  &drcbe_x64::op_ffrflt },     // FFRFLT  dst,src1,size
	{ uml::OP_FRNDS,   &drcbe_x64::op_frnds },      // FRNDS   dst,src1
	{ uml::OP_FADD,    &drcbe_x64::op_fadd },       // FADD    dst,src1,src2
	{ uml::OP_FSUB,    &drcbe_x64::op_fsub },       // FSUB    dst,src1,src2
	{ uml::OP_FCMP,    &drcbe_x64::op_fcmp },       // FCMP    src1,src2
	{ uml::OP_FMUL,    &drcbe_x64::op_fmul },       // FMUL    dst,src1,src2
	{ uml::OP_FDIV,    &drcbe_x64::op_fdiv },       // FDIV    dst,src1,src2
	{ uml::OP_FNEG,    &drcbe_x64::op_fneg },       // FNEG    dst,src1
	{ uml::OP_FABS,    &drcbe_x64::op_fabs },       // FABS    dst,src1
	{ uml::OP_FSQRT,   &drcbe_x64::op_fsqrt },      // FSQRT   dst,src1
	{ uml::OP_FRECIP,  &drcbe_x64::op_frecip },     // FRECIP  dst,src1
	{ uml::OP_FRSQRT,  &drcbe_x64::op_frsqrt },     // FRSQRT  dst,src1
	{ uml::OP_FCOPYI,  &drcbe_x64::op_fcopyi },     // FCOPYI  dst,src
	{ uml::OP_ICOPYF,  &drcbe_x64::op_icopyf }      // ICOPYF  dst,src
};

class ThrowableErrorHandler : public ErrorHandler
{
public:
	void handleError(Error err, const char *message, BaseEmitter *origin) override
	{
		throw emu_fatalerror("asmjit error %d: %s", err, message);
	}
};


//**************************************************************************
//  INLINE FUNCTIONS
//**************************************************************************

//-------------------------------------------------
//  param_normalize - convert a full parameter
//  into a reduced set
//-------------------------------------------------

drcbe_x64::be_parameter::be_parameter(drcbe_x64 &drcbe, const parameter &param, uint32_t allowed)
{
	int regnum;

	switch (param.type())
	{
		// immediates pass through
		case parameter::PTYPE_IMMEDIATE:
			assert(allowed & PTYPE_I);
			*this = param.immediate();
			break;

		// memory passes through
		case parameter::PTYPE_MEMORY:
			assert(allowed & PTYPE_M);
			*this = make_memory(param.memory());
			break;

		// if a register maps to a register, keep it as a register; otherwise map it to memory
		case parameter::PTYPE_INT_REGISTER:
			assert(allowed & PTYPE_R);
			assert(allowed & PTYPE_M);
			regnum = int_register_map[param.ireg() - REG_I0];
			if (regnum != 0)
				*this = make_ireg(regnum);
			else
				*this = make_memory(&drcbe.m_state.r[param.ireg() - REG_I0]);
			break;

		// if a register maps to a register, keep it as a register; otherwise map it to memory
		case parameter::PTYPE_FLOAT_REGISTER:
			assert(allowed & PTYPE_F);
			assert(allowed & PTYPE_M);
			regnum = float_register_map[param.freg() - REG_F0];
			if (regnum != 0)
				*this = make_freg(regnum);
			else
				*this = make_memory(&drcbe.m_state.f[param.freg() - REG_F0]);
			break;

		// everything else is unexpected
		default:
			fatalerror("Unexpected parameter type\n");
	}
}


//-------------------------------------------------
//  select_register - select a register to use,
//  avoiding conflicts with the optional
//  checkparam
//-------------------------------------------------

inline Gp drcbe_x64::be_parameter::select_register(Gp defreg) const
{
	if (m_type == PTYPE_INT_REGISTER)
		return Gp(defreg, m_value);
	return defreg;
}

inline Xmm drcbe_x64::be_parameter::select_register(Xmm defreg) const
{
	if (m_type == PTYPE_FLOAT_REGISTER)
		return Xmm(m_value);
	return defreg;
}

Gp drcbe_x64::be_parameter::select_register(Gp defreg, be_parameter const &checkparam) const
{
	if (*this == checkparam)
		return defreg;
	return select_register(defreg);
}

Gp drcbe_x64::be_parameter::select_register(Gp defreg, be_parameter const &checkparam, be_parameter const &checkparam2) const
{
	if (*this == checkparam || *this == checkparam2)
		return defreg;
	return select_register(defreg);
}

Xmm drcbe_x64::be_parameter::select_register(Xmm defreg, be_parameter const &checkparam) const
{
	if (*this == checkparam)
		return defreg;
	return select_register(defreg);
}

//-------------------------------------------------
//  select_register - select a register to use,
//  avoiding conflicts with the optional
//  checkparam
//-------------------------------------------------

inline void drcbe_x64::normalize_commutative(be_parameter &inner, be_parameter &outer)
{
	// if the inner parameter is a memory operand, push it to the outer
	if (inner.is_memory())
	{
		be_parameter temp = inner;
		inner = outer;
		outer = temp;
	}

	// if the inner parameter is an immediate, push it to the outer
	if (inner.is_immediate())
	{
		be_parameter temp = inner;
		inner = outer;
		outer = temp;
	}
}


//-------------------------------------------------
//  offset_from_rbp - return the verified offset
//  from rbp
//-------------------------------------------------

inline int32_t drcbe_x64::offset_from_rbp(const void *ptr) const
{
	const int64_t delta = reinterpret_cast<const uint8_t *>(ptr) - m_rbpvalue;
	if (int32_t(delta) != delta)
		throw emu_fatalerror("drcbe_x64::offset_from_rbp: delta out of range");
	return int32_t(delta);
}


//-------------------------------------------------
//  get_base_register_and_offset - determine right
//  base register and offset to access the given
//  target address
//-------------------------------------------------

inline Gp drcbe_x64::get_base_register_and_offset(Assembler &a, void *target, Gp const &reg, int32_t &offset)
{
	const int64_t delta = reinterpret_cast<uint8_t *>(target) - m_rbpvalue;
	if (short_immediate(delta))
	{
		offset = delta;
		return rbp;
	}
	else
	{
		offset = 0;
		mov_r64_imm(a, reg, uintptr_t(target));                                         // mov   reg,target
		return reg;
	}
}


//-------------------------------------------------
//  smart_call_r64 - generate a call either
//  directly or via a call through pointer
//-------------------------------------------------

inline void drcbe_x64::smart_call_r64(Assembler &a, x86code *target, Gp const &reg)
{
	const int64_t delta = target - (x86code *)(a.code()->baseAddress() + a.offset() + 5);
	if (short_immediate(delta))
		a.call(imm(target));                                                            // call  target
	else
	{
		mov_r64_imm(a, reg, uintptr_t(target));                                         // mov   reg,target
		a.call(reg);                                                                    // call  reg
	}
}


//-------------------------------------------------
//  smart_call_m64 - generate a call either
//  directly or via a call through pointer
//-------------------------------------------------

inline void drcbe_x64::smart_call_m64(Assembler &a, x86code **target)
{
	const int64_t delta = *target - (x86code *)(a.code()->baseAddress() + a.offset() + 5);
	if (short_immediate(delta))
		a.call(imm(*target));                                                           // call  *target
	else
		a.call(MABS(target));                                                           // call  [target]
}



//**************************************************************************
//  BACKEND CALLBACKS
//**************************************************************************

//-------------------------------------------------
//  drcbe_x64 - constructor
//-------------------------------------------------

drcbe_x64::drcbe_x64(drcuml_state &drcuml, device_t &device, drc_cache &cache, uint32_t flags, int modes, int addrbits, int ignorebits)
	: drcbe_interface(drcuml, cache, device)
	, m_hash(cache, modes, addrbits, ignorebits)
	, m_map(cache, 0xaaaaaaaa5555)
	, m_log(nullptr)
	, m_log_asmjit(nullptr)
	, m_absmask32((uint32_t *)cache.alloc_near(16*2 + 15))
	, m_absmask64(nullptr)
	, m_rbpvalue(cache.near() + 0x80)
	, m_entry(nullptr)
	, m_exit(nullptr)
	, m_nocode(nullptr)
	, m_near(*(near_state *)cache.alloc_near(sizeof(m_near)))
{
	// build up necessary arrays
	static const uint32_t sse_control[4] =
	{
		0xff80,     // ROUND_TRUNC
		0x9f80,     // ROUND_ROUND
		0xdf80,     // ROUND_CEIL
		0xbf80      // ROUND_FLOOR
	};
	memcpy(m_near.ssecontrol, sse_control, sizeof(m_near.ssecontrol));
	m_near.single1 = 1.0f;
	m_near.double1 = 1.0;

	// create absolute value masks that are aligned to SSE boundaries
	m_absmask32 = (uint32_t *)(((uintptr_t)m_absmask32 + 15) & ~15);
	m_absmask32[0] = m_absmask32[1] = m_absmask32[2] = m_absmask32[3] = 0x7fffffff;
	m_absmask64 = (uint64_t *)&m_absmask32[4];
	m_absmask64[0] = m_absmask64[1] = 0x7fffffffffffffffU;

	// get pointers to C functions we need to call
	using debugger_hook_func = void (*)(device_debug *, offs_t);
	static const debugger_hook_func debugger_inst_hook = [] (device_debug *dbg, offs_t pc) { dbg->instruction_hook(pc); }; // TODO: kill trampoline if possible
	m_near.debug_cpu_instruction_hook = (x86code *)debugger_inst_hook;
	if (LOG_HASHJMPS)
	{
		m_near.debug_log_hashjmp = (x86code *)debug_log_hashjmp;
		m_near.debug_log_hashjmp_fail = (x86code *)debug_log_hashjmp_fail;
	}
	m_near.drcmap_get_value = (x86code *)&drc_map_variables::static_get_value;

	// build the flags map
	for (int entry = 0; entry < std::size(m_near.flagsmap); entry++)
	{
		uint8_t flags = 0;
		if (entry & 0x001) flags |= FLAG_C;
		if (entry & 0x004) flags |= FLAG_U;
		if (entry & 0x040) flags |= FLAG_Z;
		if (entry & 0x080) flags |= FLAG_S;
		if (entry & 0x800) flags |= FLAG_V;
		m_near.flagsmap[entry] = flags;
	}
	for (int entry = 0; entry < std::size(m_near.flagsunmap); entry++)
	{
		uint64_t flags = 0;
		if (entry & FLAG_C) flags |= 0x001;
		if (entry & FLAG_U) flags |= 0x004;
		if (entry & FLAG_Z) flags |= 0x040;
		if (entry & FLAG_S) flags |= 0x080;
		if (entry & FLAG_V) flags |= 0x800;
		m_near.flagsunmap[entry] = flags;
	}

	// resolve the actual addresses of the address space handlers
	auto const resolve_accessor =
			[] (resolved_handler &handler, address_space &space, auto accessor)
			{
				if (MAME_DELEGATE_USE_TYPE == MAME_DELEGATE_TYPE_ITANIUM)
				{
					struct { uintptr_t ptr; ptrdiff_t adj; } equiv;
					assert(sizeof(accessor) == sizeof(equiv));
					*reinterpret_cast<decltype(accessor) *>(&equiv) = accessor;
					handler.obj = uintptr_t(reinterpret_cast<u8 *>(&space) + equiv.adj);
					if (BIT(equiv.ptr, 0))
					{
						auto const vptr = *reinterpret_cast<u8 const *const *>(handler.obj) + equiv.ptr - 1;
						handler.func = *reinterpret_cast<x86code *const *>(vptr);
					}
					else
					{
						handler.func = reinterpret_cast<x86code *>(equiv.ptr);
					}
				}
				else if (MAME_DELEGATE_USE_TYPE == MAME_DELEGATE_TYPE_MSVC)
				{
					// interpret the pointer to member function ignoring the virtual inheritance variant
					struct single { uintptr_t ptr; };
					struct multi { uintptr_t ptr; int adj; };
					struct { uintptr_t ptr; int adj; int vadj; int vindex; } unknown;
					assert(sizeof(accessor) <= sizeof(unknown));
					*reinterpret_cast<decltype(accessor) *>(&unknown) = accessor;
					handler.func = reinterpret_cast<x86code *>(unknown.ptr);
					handler.obj = uintptr_t(&space);
					if ((sizeof(unknown) == sizeof(accessor)) && unknown.vindex)
					{
						handler.obj += unknown.vadj;
						auto const vptr = *reinterpret_cast<std::uint8_t const *const *>(handler.obj);
						handler.obj += *reinterpret_cast<int const *>(vptr + unknown.vindex);
					}
					if (sizeof(single) < sizeof(accessor))
						handler.obj += unknown.adj;

					// walk past thunks
					while (true)
					{
						if (0xe9 == handler.func[0])
						{
							// absolute jump with 32-bit displacement
							handler.func += 5 + *reinterpret_cast<s32 const *>(handler.func + 1);
						}
						else if ((0x48 == handler.func[0]) && (0x8b == handler.func[1]) && (0x01 == handler.func[2]) && (0xff == handler.func[3]) && ((0x60 == handler.func[4]) || (0xa0 == handler.func[4])))
						{
							// virtual function call thunk
							auto const vptr = *reinterpret_cast<std::uint8_t const *const *>(handler.obj);
							if (0x60 == handler.func[4])
								handler.func = *reinterpret_cast<x86code *const *>(vptr + *reinterpret_cast<s8 const *>(handler.func + 5));
							else
								handler.func = *reinterpret_cast<x86code *const *>(vptr + *reinterpret_cast<s32 const *>(handler.func + 5));
						}
						else
						{
							// not something we can easily bypass
							break;
						}
					}
				}
			};
	m_resolved_accessors.resize(m_space.size());
	for (int space = 0; m_space.size() > space; ++space)
	{
		if (m_space[space])
		{
			resolve_accessor(m_resolved_accessors[space].read_byte,         *m_space[space], static_cast<u8  (address_space::*)(offs_t)     >(&address_space::read_byte));
			resolve_accessor(m_resolved_accessors[space].read_word,         *m_space[space], static_cast<u16 (address_space::*)(offs_t)     >(&address_space::read_word));
			resolve_accessor(m_resolved_accessors[space].read_word_masked,  *m_space[space], static_cast<u16 (address_space::*)(offs_t, u16)>(&address_space::read_word));
			resolve_accessor(m_resolved_accessors[space].read_dword,        *m_space[space], static_cast<u32 (address_space::*)(offs_t)     >(&address_space::read_dword));
			resolve_accessor(m_resolved_accessors[space].read_dword_masked, *m_space[space], static_cast<u32 (address_space::*)(offs_t, u32)>(&address_space::read_dword));
			resolve_accessor(m_resolved_accessors[space].read_qword,        *m_space[space], static_cast<u64 (address_space::*)(offs_t)     >(&address_space::read_qword));
			resolve_accessor(m_resolved_accessors[space].read_qword_masked, *m_space[space], static_cast<u64 (address_space::*)(offs_t, u64)>(&address_space::read_qword));

			resolve_accessor(m_resolved_accessors[space].write_byte,         *m_space[space], static_cast<void (address_space::*)(offs_t, u8)      >(&address_space::write_byte));
			resolve_accessor(m_resolved_accessors[space].write_word,         *m_space[space], static_cast<void (address_space::*)(offs_t, u16)     >(&address_space::write_word));
			resolve_accessor(m_resolved_accessors[space].write_word_masked,  *m_space[space], static_cast<void (address_space::*)(offs_t, u16, u16)>(&address_space::write_word));
			resolve_accessor(m_resolved_accessors[space].write_dword,        *m_space[space], static_cast<void (address_space::*)(offs_t, u32)     >(&address_space::write_dword));
			resolve_accessor(m_resolved_accessors[space].write_dword_masked, *m_space[space], static_cast<void (address_space::*)(offs_t, u32, u32)>(&address_space::write_dword));
			resolve_accessor(m_resolved_accessors[space].write_qword,        *m_space[space], static_cast<void (address_space::*)(offs_t, u64)     >(&address_space::write_qword));
			resolve_accessor(m_resolved_accessors[space].write_qword_masked, *m_space[space], static_cast<void (address_space::*)(offs_t, u64, u64)>(&address_space::write_qword));
		}
	}

	// build the opcode table (static but it doesn't hurt to regenerate it)
	for (auto & elem : s_opcode_table_source)
		s_opcode_table[elem.opcode] = elem.func;

	// create the log
	if (device.machine().options().drc_log_native())
	{
		std::string filename = std::string("drcbex64_").append(device.shortname()).append(".asm");
		m_log = x86log_create_context(filename.c_str());
		m_log_asmjit = fopen(std::string("drcbex64_asmjit_").append(device.shortname()).append(".asm").c_str(), "w");
	}
}


//-------------------------------------------------
//  ~drcbe_x64 - destructor
//-------------------------------------------------

drcbe_x64::~drcbe_x64()
{
	// free the log context
	if (m_log != nullptr)
		x86log_free_context(m_log);

	if (m_log_asmjit)
		fclose(m_log_asmjit);
}

size_t drcbe_x64::emit(CodeHolder &ch)
{
	Error err;

	// the following three calls aren't currently required, but may be if
	// other asmjist features are used in future
	if (false)
	{
		err = ch.flatten();
		if (err)
			throw emu_fatalerror("asmjit::CodeHolder::flatten() error %d", err);

		err = ch.resolveUnresolvedLinks();
		if (err)
			throw emu_fatalerror("asmjit::CodeHolder::resolveUnresolvedLinks() error %d", err);

		err = ch.relocateToBase(ch.baseAddress());
		if (err)
			throw emu_fatalerror("asmjit::CodeHolder::relocateToBase() error %d", err);
	}

	size_t const alignment = ch.baseAddress() - uint64_t(m_cache.top());
	size_t const code_size = ch.codeSize();

	// test if enough room remains in drc cache
	drccodeptr *cachetop = m_cache.begin_codegen(alignment + code_size);
	if (cachetop == nullptr)
		return 0;

	err = ch.copyFlattenedData(drccodeptr(ch.baseAddress()), code_size, CopySectionFlags::kPadTargetBuffer);
	if (err)
		throw emu_fatalerror("asmjit::CodeHolder::copyFlattenedData() error %d", err);

	// update the drc cache and end codegen
	*cachetop += alignment + code_size;
	m_cache.end_codegen();

	return code_size;
}

//-------------------------------------------------
//  reset - reset back-end specific state
//-------------------------------------------------

void drcbe_x64::reset()
{
	// output a note to the log
	if (m_log != nullptr)
		x86log_printf(m_log, "%s", "\n\n===========\nCACHE RESET\n===========\n\n");

	// generate a little bit of glue code to set up the environment
	x86code *dst = (x86code *)m_cache.top();

	CodeHolder ch;
	ch.init(Environment::host(), uint64_t(dst));

	FileLogger logger(m_log_asmjit);
	if (logger.file())
	{
		logger.setFlags(FormatFlags::kHexOffsets | FormatFlags::kHexImms | FormatFlags::kMachineCode);
		logger.setIndentation(FormatIndentationGroup::kCode, 4);
		ch.setLogger(&logger);
	}

	Assembler a(&ch);
	if (logger.file())
		a.addDiagnosticOptions(DiagnosticOptions::kValidateIntermediate);

	// generate an entry point
	m_entry = (x86_entry_point_func)dst;
	a.bind(a.newNamedLabel("entry_point"));

	FuncDetail entry_point;
	entry_point.init(FuncSignatureT<uint32_t, uint8_t *, x86code *>(CallConvId::kHost), Environment::host());

	FuncFrame frame;
	frame.init(entry_point);
	frame.addDirtyRegs(rbx, rbp, rsi, rdi, r12, r13, r14, r15);
	FuncArgsAssignment args(&entry_point);
	args.assignAll(rbp);
	args.updateFuncFrame(frame);
	frame.finalize();

	a.emitProlog(frame);
	a.emitArgsAssignment(frame, args);

	a.sub(rsp, 32);
	a.mov(MABS(&m_near.hashstacksave), rsp);
	a.sub(rsp, 8);
	a.mov(MABS(&m_near.stacksave), rsp);
	a.stmxcsr(MABS(&m_near.ssemode));
	a.jmp(Gpq(REG_PARAM2));

	// generate an exit point
	m_exit = dst + a.offset();
	a.bind(a.newNamedLabel("exit_point"));
	a.ldmxcsr(MABS(&m_near.ssemode));
	a.mov(rsp, MABS(&m_near.hashstacksave));
	a.add(rsp, 32);
	a.emitEpilog(frame);

	// generate a no code point
	m_nocode = dst + a.offset();
	a.bind(a.newNamedLabel("nocode_point"));
	a.ret();

	// emit the generated code
	size_t bytes = emit(ch);

	if (m_log != nullptr)
	{
		x86log_disasm_code_range(m_log, "entry_point", dst, m_exit);
		x86log_disasm_code_range(m_log, "exit_point", m_exit, m_nocode);
		x86log_disasm_code_range(m_log, "nocode_point", m_nocode, dst + bytes);
	}

	// reset our hash tables
	m_hash.reset();
	m_hash.set_default_codeptr(m_nocode);
}


//-------------------------------------------------
//  execute - execute a block of code referenced
//  by the given handle
//-------------------------------------------------

int drcbe_x64::execute(code_handle &entry)
{
	// call our entry point which will jump to the destination
	m_cache.codegen_complete();
	return (*m_entry)(m_rbpvalue, (x86code *)entry.codeptr());
}


//-------------------------------------------------
//  drcbex64_generate - generate code
//-------------------------------------------------

void drcbe_x64::generate(drcuml_block &block, const instruction *instlist, uint32_t numinst)
{
	// tell all of our utility objects that a block is beginning
	m_hash.block_begin(block, instlist, numinst);
	m_map.block_begin(block);

	// compute the base by aligning the cache top to a cache line (assumed to be 64 bytes)
	x86code *dst = (x86code *)(uint64_t(m_cache.top() + 63) & ~63);

	CodeHolder ch;
	ch.init(Environment::host(), uint64_t(dst));
	ThrowableErrorHandler e;
	ch.setErrorHandler(&e);

	FileLogger logger(m_log_asmjit);
	if (logger.file())
	{
		logger.setFlags(FormatFlags::kHexOffsets | FormatFlags::kHexImms | FormatFlags::kMachineCode);
		logger.setIndentation(FormatIndentationGroup::kCode, 4);
		ch.setLogger(&logger);
	}

	Assembler a(&ch);
	if (logger.file())
		a.addDiagnosticOptions(DiagnosticOptions::kValidateIntermediate);

	// generate code
	std::string blockname;
	for (int inum = 0; inum < numinst; inum++)
	{
		const instruction &inst = instlist[inum];
		assert(inst.opcode() < std::size(s_opcode_table));

		// must remain in scope until output
		std::string dasm;

		// add a comment
		if (m_log != nullptr)
		{
			dasm = inst.disasm(&m_drcuml);

			x86log_add_comment(m_log, dst + a.offset(), "%s", dasm.c_str());
			a.setInlineComment(dasm.c_str());
		}

		// extract a blockname
		if (blockname.empty())
		{
			if (inst.opcode() == OP_HANDLE)
				blockname = inst.param(0).handle().string();
			else if (inst.opcode() == OP_HASH)
				blockname = string_format("Code: mode=%d PC=%08X", (uint32_t)inst.param(0).immediate(), (offs_t)inst.param(1).immediate());
		}

		// generate code
		(this->*s_opcode_table[inst.opcode()])(a, inst);
	}

	// emit the generated code
	size_t const bytes = emit(ch);
	if (!bytes)
		block.abort();

	// log it
	if (m_log != nullptr)
		x86log_disasm_code_range(m_log, (blockname.empty()) ? "Unknown block" : blockname.c_str(), dst, dst + bytes);

	// tell all of our utility objects that the block is finished
	m_hash.block_end(block);
	m_map.block_end(block);
}


//-------------------------------------------------
//  hash_exists - return true if the given mode/pc
//  exists in the hash table
//-------------------------------------------------

bool drcbe_x64::hash_exists(uint32_t mode, uint32_t pc)
{
	return m_hash.code_exists(mode, pc);
}


//-------------------------------------------------
//  get_info - return information about the
//  back-end implementation
//-------------------------------------------------

void drcbe_x64::get_info(drcbe_info &info)
{
	for (info.direct_iregs = 0; info.direct_iregs < REG_I_COUNT; info.direct_iregs++)
		if (int_register_map[info.direct_iregs] == 0)
			break;
	for (info.direct_fregs = 0; info.direct_fregs < REG_F_COUNT; info.direct_fregs++)
		if (float_register_map[info.direct_fregs] == 0)
			break;
}

void drcbe_x64::alu_op_param(Assembler &a, Inst::Id const opcode, Operand const &dst, be_parameter const &param, std::function<bool(Assembler &a, Operand const &dst, be_parameter const &src)> optimize)
{
	bool const is64 = dst.size() == 8;

	if (param.is_immediate())
	{
		if (!optimize(a, dst, param))
		{
			if (is64 && !short_immediate(param.immediate()))
			{
				// use scratch register for 64-bit immediate
				a.mov(r11, param.immediate());                                          // mov   r11,param
				a.emit(opcode, dst, r11);                                               // op    dst,r11
			}
			else
				a.emit(opcode, dst, param.immediate());                                 // op    dst,param
		}
	}
	else if (param.is_memory())
	{
		if (dst.isMem())
		{
			// use temporary register for memory,memory
			Gp const tmp = is64 ? param.select_register(rax) : param.select_register(eax);

			a.mov(tmp, MABS(param.memory()));                                           // mov   tmp,param
			a.emit(opcode, dst, tmp);                                                   // op    [dst],tmp
		}
		else if (opcode != Inst::kIdTest)
			// most instructions are register,memory
			a.emit(opcode, dst, MABS(param.memory()));                                  // op    dst,[param]
		else
			// test instruction requires memory,register
			a.emit(opcode, MABS(param.memory()), dst);                                  // op    [param],dst
	}
	else if (param.is_int_register())
	{
		Gp const src = Gp::fromTypeAndId(is64 ? RegType::kX86_Gpq : RegType::kX86_Gpd, param.ireg());

		a.emit(opcode, dst, src);                                                       // op    dst,param
	}
}

void drcbe_x64::shift_op_param(Assembler &a, Inst::Id const opcode, Operand const &dst, be_parameter const &param)
{
	Operand shift = cl;
	if (param.is_immediate())
		shift = imm(param.immediate());
	else
		mov_reg_param(a, ecx, param);

	a.emit(opcode, dst, shift);
}

void drcbe_x64::mov_reg_param(Assembler &a, Gp const &reg, be_parameter const &param, bool const keepflags)
{
	if (param.is_immediate())
	{
		if (param.immediate() == 0 && !keepflags)
			a.xor_(reg.r32(), reg.r32());                                               // xor   reg,reg
		else if (reg.isGpq())
			mov_r64_imm(a, reg, param.immediate());
		else
			a.mov(reg, param.immediate());                                              // mov   reg,param
	}
	else if (param.is_memory())
		a.mov(reg, MABS(param.memory()));                                               // mov   reg,[param]
	else if (param.is_int_register())
	{
		if (reg.id() != param.ireg())
			a.mov(reg, Gp(reg, param.ireg()));                                          // mov   reg,param
	}
}

void drcbe_x64::mov_param_reg(Assembler &a, be_parameter const &param, Gp const &reg)
{
	assert(!param.is_immediate());

	if (param.is_memory())
		a.mov(MABS(param.memory()), reg);                                               // mov   [param],reg
	else if (param.is_int_register())
	{
		if (reg.id() != param.ireg())
			a.mov(Gp(reg, param.ireg()), reg);                                          // mov   param,reg
	}
}

void drcbe_x64::mov_mem_param(Assembler &a, Mem const &mem, be_parameter const &param)
{
	bool const is64 = mem.size() == 8;

	if (param.is_immediate())
	{
		if (is64 && !short_immediate(param.immediate()))
		{
			a.mov(rax, param.immediate());                                              // mov   tmp,param
			a.mov(mem, rax);                                                            // mov   [mem],tmp
		}
		else
			a.mov(mem, param.immediate());                                              // mov   [mem],param
	}
	else if (param.is_memory())
	{
		Gp const tmp = Gp::fromTypeAndId(is64 ? RegType::kX86_Gpq : RegType::kX86_Gpd, Gp::kIdAx);

		a.mov(tmp, MABS(param.memory()));                                               // mov   tmp,[param]
		a.mov(mem, tmp);                                                                // mov   [mem],tmp
	}
	else if (param.is_int_register())
	{
		Gp const src = Gp::fromTypeAndId(is64 ? RegType::kX86_Gpq : RegType::kX86_Gpd, param.ireg());

		a.mov(mem, src);                                                                // mov   [mem],param
	}
}

void drcbe_x64::movsx_r64_p32(Assembler &a, Gp const &reg, be_parameter const &param)
{
	if (param.is_immediate())
	{
		if (param.immediate() == 0)
			a.xor_(reg.r32(), reg.r32());                                               // xor   reg,reg
		else if ((int32_t)param.immediate() >= 0)
			a.mov(reg.r32(), param.immediate());                                        // mov   reg,param
		else
			mov_r64_imm(a, reg, int32_t(param.immediate()));                            // mov   reg,param
	}
	else if (param.is_memory())
		a.movsxd(reg, MABS(param.memory()));                                            // movsxd reg,[param]
	else if (param.is_int_register())
		a.movsxd(reg, Gpd(param.ireg()));                                               // movsxd reg,param
}

void drcbe_x64::mov_r64_imm(Assembler &a, Gp const &reg, uint64_t const imm)
{
	if (u32(imm) == imm)
		a.mov(reg.r32(), imm);
	else if (s32(imm) == imm)
		a.mov(reg.r64(), s32(imm));
	else
		a.mov(reg.r64(), imm);
}


/***************************************************************************
    EMITTERS FOR FLOATING POINT OPERATIONS WITH PARAMETERS
***************************************************************************/

//-------------------------------------------------
//  movss_r128_p32 - move a 32-bit parameter
//  into a register
//-------------------------------------------------

void drcbe_x64::movss_r128_p32(Assembler &a, Xmm const &reg, be_parameter const &param)
{
	assert(!param.is_immediate());
	if (param.is_memory())
		a.movss(reg, MABS(param.memory(), 4));                                          // movss reg,[param]
	else if (param.is_float_register())
	{
		if (reg.id() != param.freg())
			a.movss(reg, Xmm(param.freg()));                                            // movss reg,param
	}
}


//-------------------------------------------------
//  movss_p32_r128 - move a register into a
//  32-bit parameter
//-------------------------------------------------

void drcbe_x64::movss_p32_r128(Assembler &a, be_parameter const &param, Xmm const &reg)
{
	assert(!param.is_immediate());
	if (param.is_memory())
		a.movss(MABS(param.memory(), 4), reg);                                          // movss [param],reg
	else if (param.is_float_register())
	{
		if (reg.id() != param.freg())
			a.movss(Xmm(param.freg()), reg);                                            // movss param,reg
	}
}


//-------------------------------------------------
//  movsd_r128_p64 - move a 64-bit parameter
//  into a register
//-------------------------------------------------

void drcbe_x64::movsd_r128_p64(Assembler &a, Xmm const &reg, be_parameter const &param)
{
	assert(!param.is_immediate());
	if (param.is_memory())
		a.movsd(reg, MABS(param.memory(), 8));                                          // movsd reg,[param]
	else if (param.is_float_register())
	{
		if (reg.id() != param.freg())
			a.movsd(reg, Xmm(param.freg()));                                            // movsd reg,param
	}
}


//-------------------------------------------------
//  movsd_p64_r128 - move a register into a
//  64-bit parameter
//-------------------------------------------------

void drcbe_x64::movsd_p64_r128(Assembler &a, be_parameter const &param, Xmm const &reg)
{
	assert(!param.is_immediate());
	if (param.is_memory())
		a.movsd(MABS(param.memory(), 8), reg);                                          // movsd [param],reg
	else if (param.is_float_register())
	{
		if (reg.id() != param.freg())
			a.movsd(Xmm(param.freg()), reg);                                            // movsd param,reg
	}
}


//**************************************************************************
//  DEBUG HELPERS
//**************************************************************************

//-------------------------------------------------
//  debug_log_hashjmp - callback to handle
//  logging of hashjmps
//-------------------------------------------------

void drcbe_x64::debug_log_hashjmp(offs_t pc, int mode)
{
	printf("mode=%d PC=%08X\n", mode, pc);
}


//-------------------------------------------------
//  debug_log_hashjmp - callback to handle
//  logging of hashjmps
//-------------------------------------------------

void drcbe_x64::debug_log_hashjmp_fail()
{
	printf("  (FAIL)\n");
}


/***************************************************************************
    COMPILE-TIME OPCODES
***************************************************************************/

//-------------------------------------------------
//  op_handle - process a HANDLE opcode
//-------------------------------------------------

void drcbe_x64::op_handle(Assembler &a, const instruction &inst)
{
	assert_no_condition(inst);
	assert_no_flags(inst);
	assert(inst.numparams() == 1);
	assert(inst.param(0).is_code_handle());

	// make a label for documentation
	Label handle = a.newNamedLabel(inst.param(0).handle().string());
	a.bind(handle);

	// emit a jump around the stack adjust in case code falls through here
	Label skip = a.newLabel();
	a.short_().jmp(skip);                                                               // jmp   skip

	// register the current pointer for the handle
	inst.param(0).handle().set_codeptr(drccodeptr(a.code()->baseAddress() + a.offset()));

	// by default, the handle points to prolog code that moves the stack pointer
	a.lea(rsp, ptr(rsp, -40));                                                          // lea   rsp,[rsp-40]
	a.bind(skip);                                                                   // skip:
}


//-------------------------------------------------
//  op_hash - process a HASH opcode
//-------------------------------------------------

void drcbe_x64::op_hash(Assembler &a, const instruction &inst)
{
	assert_no_condition(inst);
	assert_no_flags(inst);
	assert(inst.numparams() == 2);
	assert(inst.param(0).is_immediate());
	assert(inst.param(1).is_immediate());

	// register the current pointer for the mode/PC
	m_hash.set_codeptr(inst.param(0).immediate(), inst.param(1).immediate(), drccodeptr(a.code()->baseAddress() + a.offset()));
}


//-------------------------------------------------
//  op_label - process a LABEL opcode
//-------------------------------------------------

void drcbe_x64::op_label(Assembler &a, const instruction &inst)
{
	assert_no_condition(inst);
	assert_no_flags(inst);
	assert(inst.numparams() == 1);
	assert(inst.param(0).is_code_label());

	std::string labelName = util::string_format("PC$%x", inst.param(0).label());
	Label label = a.labelByName(labelName.c_str());
	if (!label.isValid())
		label = a.newNamedLabel(labelName.c_str());

	// register the current pointer for the label
	a.bind(label);
}


//-------------------------------------------------
//  op_comment - process a COMMENT opcode
//-------------------------------------------------

void drcbe_x64::op_comment(Assembler &a, const instruction &inst)
{
	assert_no_condition(inst);
	assert_no_flags(inst);
	assert(inst.numparams() == 1);
	assert(inst.param(0).is_string());

	// do nothing
}


//-------------------------------------------------
//  op_mapvar - process a MAPVAR opcode
//-------------------------------------------------

void drcbe_x64::op_mapvar(Assembler &a, const instruction &inst)
{
	assert_no_condition(inst);
	assert_no_flags(inst);
	assert(inst.numparams() == 2);
	assert(inst.param(0).is_mapvar());
	assert(inst.param(1).is_immediate());

	// set the value of the specified mapvar
	m_map.set_value(drccodeptr(a.code()->baseAddress() + a.offset()), inst.param(0).mapvar(), inst.param(1).immediate());
}



/***************************************************************************
    CONTROL FLOW OPCODES
***************************************************************************/

//-------------------------------------------------
//  op_nop - process a NOP opcode
//-------------------------------------------------

void drcbe_x64::op_nop(Assembler &a, const instruction &inst)
{
	// nothing
}


//-------------------------------------------------
//  op_debug - process a DEBUG opcode
//-------------------------------------------------

void drcbe_x64::op_debug(Assembler &a, const instruction &inst)
{
	// validate instruction
	assert(inst.size() == 4);
	assert_no_condition(inst);
	assert_no_flags(inst);

	if ((m_device.machine().debug_flags & DEBUG_FLAG_ENABLED) != 0)
	{
		// normalize parameters
		be_parameter pcp(*this, inst.param(0), PTYPE_MRI);

		// test and branch
		mov_r64_imm(a, rax, (uintptr_t)&m_device.machine().debug_flags);                // mov   rax,&debug_flags
		a.test(dword_ptr(rax), DEBUG_FLAG_CALL_HOOK);                                   // test  [debug_flags],DEBUG_FLAG_CALL_HOOK
		Label skip = a.newLabel();
		a.short_().jz(skip);

		// push the parameter
		mov_r64_imm(a, Gpq(REG_PARAM1), (uintptr_t)m_device.debug());                   // mov   param1,device.debug
		mov_reg_param(a, Gpd(REG_PARAM2), pcp);                                         // mov   param2,pcp
		smart_call_m64(a, &m_near.debug_cpu_instruction_hook);                          // call  debug_cpu_instruction_hook

		a.bind(skip);
	}
}


//-------------------------------------------------
//  op_exit - process an EXIT opcode
//-------------------------------------------------

void drcbe_x64::op_exit(Assembler &a, const instruction &inst)
{
	// validate instruction
	assert(inst.size() == 4);
	assert_any_condition(inst);
	assert_no_flags(inst);

	// normalize parameters
	be_parameter retp(*this, inst.param(0), PTYPE_MRI);

	// load the parameter into EAX
	mov_reg_param(a, eax, retp);                                                        // mov   eax,retp
	if (inst.condition() == uml::COND_ALWAYS)
		a.jmp(imm(m_exit));                                                             // jmp   exit
	else
		a.j(X86_CONDITION(inst.condition()), imm(m_exit));                              // jcc   exit
}


//-------------------------------------------------
//  op_hashjmp - process a HASHJMP opcode
//-------------------------------------------------

void drcbe_x64::op_hashjmp(Assembler &a, const instruction &inst)
{
	// validate instruction
	assert(inst.size() == 4);
	assert_no_condition(inst);
	assert_no_flags(inst);

	// normalize parameters
	be_parameter modep(*this, inst.param(0), PTYPE_MRI);
	be_parameter pcp(*this, inst.param(1), PTYPE_MRI);
	const parameter &exp = inst.param(2);
	assert(exp.is_code_handle());

	if (LOG_HASHJMPS)
	{
		mov_reg_param(a, Gpd(REG_PARAM1), pcp);
		mov_reg_param(a, Gpd(REG_PARAM2), modep);
		smart_call_m64(a, &m_near.debug_log_hashjmp);
	}

	// load the stack base one word early so we end up at the right spot after our call below
	a.mov(rsp, MABS(&m_near.hashstacksave));                                            // mov   rsp,[hashstacksave]

	// fixed mode cases
	if (modep.is_immediate() && m_hash.is_mode_populated(modep.immediate()))
	{
		// a straight immediate jump is direct, though we need the PC in EAX in case of failure
		if (pcp.is_immediate())
		{
			uint32_t l1val = (pcp.immediate() >> m_hash.l1shift()) & m_hash.l1mask();
			uint32_t l2val = (pcp.immediate() >> m_hash.l2shift()) & m_hash.l2mask();
			a.call(MABS(&m_hash.base()[modep.immediate()][l1val][l2val]));              // call  hash[modep][l1val][l2val]
		}

		// a fixed mode but variable PC
		else
		{
			mov_reg_param(a, eax, pcp);                                                 // mov   eax,pcp
			a.mov(edx, eax);                                                            // mov   edx,eax
			a.shr(edx, m_hash.l1shift());                                               // shr   edx,l1shift
			a.and_(eax, m_hash.l2mask() << m_hash.l2shift());                           // and  eax,l2mask << l2shift
			a.mov(rdx, ptr(rbp, rdx, 3, offset_from_rbp(&m_hash.base()[modep.immediate()][0])));
																						// mov   rdx,hash[modep+edx*8]
			a.call(ptr(rdx, rax, 3 - m_hash.l2shift()));                                // call  [rdx+rax*shift]
		}
	}
	else
	{
		// variable mode
		Gp modereg = modep.select_register(ecx);
		mov_reg_param(a, modereg, modep);                                               // mov   modereg,modep
		a.mov(rcx, ptr(rbp, modereg, 3, offset_from_rbp(m_hash.base())));               // mov   rcx,hash[modereg*8]

		// fixed PC
		if (pcp.is_immediate())
		{
			uint32_t l1val = (pcp.immediate() >> m_hash.l1shift()) & m_hash.l1mask();
			uint32_t l2val = (pcp.immediate() >> m_hash.l2shift()) & m_hash.l2mask();
			a.mov(rdx, ptr(rcx, l1val * 8));                                            // mov   rdx,[rcx+l1val*8]
			a.call(ptr(rdx, l2val * 8));                                                // call  [l2val*8]
		}

		// variable PC
		else
		{
			mov_reg_param(a, eax, pcp);                                                 // mov   eax,pcp
			a.mov(edx, eax);                                                            // mov   edx,eax
			a.shr(edx, m_hash.l1shift());                                               // shr   edx,l1shift
			a.mov(rdx, ptr(rcx, rdx, 3));                                               // mov   rdx,[rcx+rdx*8]
			a.and_(eax, m_hash.l2mask() << m_hash.l2shift());                           // and   eax,l2mask << l2shift
			a.call(ptr(rdx, rax, 3 - m_hash.l2shift()));                                // call  [rdx+rax*shift]
		}
	}

	// in all cases, if there is no code, we return here to generate the exception
	if (LOG_HASHJMPS)
		smart_call_m64(a, &m_near.debug_log_hashjmp_fail);

	mov_mem_param(a, MABS(&m_state.exp, 4), pcp);                                       // mov   [exp],param
	a.sub(rsp, 8);                                                                      // sub   rsp,8
	a.call(MABS(exp.handle().codeptr_addr()));                                          // call  [exp]
}


//-------------------------------------------------
//  op_jmp - process a JMP opcode
//-------------------------------------------------

void drcbe_x64::op_jmp(Assembler &a, const instruction &inst)
{
	// validate instruction
	assert(inst.size() == 4);
	assert_any_condition(inst);
	assert_no_flags(inst);

	// normalize parameters
	const parameter &labelp = inst.param(0);
	assert(labelp.is_code_label());

	std::string labelName = util::string_format("PC$%x", labelp.label());
	Label jmptarget = a.labelByName(labelName.c_str());
	if (!jmptarget.isValid())
		jmptarget = a.newNamedLabel(labelName.c_str());

	if (inst.condition() == uml::COND_ALWAYS)
		a.jmp(jmptarget);                                                               // jmp   target
	else
		a.j(X86_CONDITION(inst.condition()), jmptarget);                                // jcc   target
}


//-------------------------------------------------
//  op_exh - process an EXH opcode
//-------------------------------------------------

void drcbe_x64::op_exh(Assembler &a, const instruction &inst)
{
	// validate instruction
	assert(inst.size() == 4);
	assert_any_condition(inst);
	assert_no_flags(inst);

	// normalize parameters
	const parameter &handp = inst.param(0);
	assert(handp.is_code_handle());
	be_parameter exp(*this, inst.param(1), PTYPE_MRI);

	// look up the handle target
	drccodeptr *targetptr = handp.handle().codeptr_addr();

	// perform the exception processing
	Label no_exception = a.newLabel();
	if (inst.condition() != uml::COND_ALWAYS)
		a.short_().j(X86_NOT_CONDITION(inst.condition()), no_exception);                // jcc   no_exception
	mov_mem_param(a, MABS(&m_state.exp, 4), exp);                                       // mov   [exp],exp
	if (*targetptr != nullptr)
		a.call(imm(*targetptr));                                                        // call  *targetptr
	else
		a.call(MABS(targetptr));                                                        // call  [targetptr]
	if (inst.condition() != uml::COND_ALWAYS)
		a.bind(no_exception);
}


//-------------------------------------------------
//  op_callh - process a CALLH opcode
//-------------------------------------------------

void drcbe_x64::op_callh(Assembler &a, const instruction &inst)
{
	// validate instruction
	assert(inst.size() == 4);
	assert_any_condition(inst);
	assert_no_flags(inst);

	// normalize parameters
	const parameter &handp = inst.param(0);
	assert(handp.is_code_handle());

	// look up the handle target
	drccodeptr *targetptr = handp.handle().codeptr_addr();

	// skip if conditional
	Label skip = a.newLabel();
	if (inst.condition() != uml::COND_ALWAYS)
		a.short_().j(X86_NOT_CONDITION(inst.condition()), skip);                        // jcc   skip

	// jump through the handle; directly if a normal jump
	if (*targetptr != nullptr)
		a.call(imm(*targetptr));                                                        // call  *targetptr
	else
		a.call(MABS(targetptr));                                                        // call  [targetptr]

	// resolve the conditional link
	if (inst.condition() != uml::COND_ALWAYS)
		a.bind(skip);                                                               // skip:
}


//-------------------------------------------------
//  op_ret - process a RET opcode
//-------------------------------------------------

void drcbe_x64::op_ret(Assembler &a, const instruction &inst)
{
	// validate instruction
	assert(inst.size() == 4);
	assert_any_condition(inst);
	assert_no_flags(inst);
	assert(inst.numparams() == 0);

	// skip if conditional
	Label skip = a.newLabel();
	if (inst.condition() != uml::COND_ALWAYS)
		a.short_().j(X86_NOT_CONDITION(inst.condition()), skip);                        // jcc   skip

	// return
	a.lea(rsp, ptr(rsp, 40));                                                           // lea   rsp,[rsp+40]
	a.ret();                                                                            // ret

	// resolve the conditional link
	if (inst.condition() != uml::COND_ALWAYS)
		a.bind(skip);                                                               // skip:
}


//-------------------------------------------------
//  op_callc - process a CALLC opcode
//-------------------------------------------------

void drcbe_x64::op_callc(Assembler &a, const instruction &inst)
{
	// validate instruction
	assert(inst.size() == 4);
	assert_any_condition(inst);
	assert_no_flags(inst);

	// normalize parameters
	const parameter &funcp = inst.param(0);
	assert(funcp.is_c_function());
	be_parameter paramp(*this, inst.param(1), PTYPE_M);

	// skip if conditional
	Label skip = a.newLabel();
	if (inst.condition() != uml::COND_ALWAYS)
		a.short_().j(X86_NOT_CONDITION(inst.condition()), skip);                        // jcc   skip

	// perform the call
	mov_r64_imm(a, Gpq(REG_PARAM1), (uintptr_t)paramp.memory());                        // mov   param1,paramp
	smart_call_r64(a, (x86code *)(uintptr_t)funcp.cfunc(), rax);                        // call  funcp

	// resolve the conditional link
	if (inst.condition() != uml::COND_ALWAYS)
		a.bind(skip);                                                               // skip:
}


//-------------------------------------------------
//  op_recover - process a RECOVER opcode
//-------------------------------------------------

void drcbe_x64::op_recover(Assembler &a, const instruction &inst)
{
	// validate instruction
	assert(inst.size() == 4);
	assert_no_condition(inst);
	assert_no_flags(inst);

	// normalize parameters
	be_parameter dstp(*this, inst.param(0), PTYPE_MR);

	// call the recovery code
	a.mov(rax, MABS(&m_near.stacksave));                                                // mov   rax,stacksave
	a.mov(rax, ptr(rax, -8));                                                           // mov   rax,[rax-8]
	mov_r64_imm(a, Gpq(REG_PARAM1), (uintptr_t)&m_map);                                 // mov   param1,m_map
	a.lea(Gpq(REG_PARAM2), ptr(rax, -1));                                               // lea   param2,[rax-1]
	mov_r64_imm(a, Gpq(REG_PARAM3), inst.param(1).mapvar());                            // mov   param3,param[1].value
	smart_call_m64(a, &m_near.drcmap_get_value);                                        // call  drcmap_get_value
	mov_param_reg(a, dstp, eax);                                                        // mov   dstp,eax
}



/***************************************************************************
    INTERNAL REGISTER OPCODES
***************************************************************************/

//-------------------------------------------------
//  op_setfmod - process a SETFMOD opcode
//-------------------------------------------------

void drcbe_x64::op_setfmod(Assembler &a, const instruction &inst)
{
	// validate instruction
	assert(inst.size() == 4);
	assert_no_condition(inst);
	assert_no_flags(inst);

	// normalize parameters
	be_parameter srcp(*this, inst.param(0), PTYPE_MRI);

	// immediate case
	if (srcp.is_immediate())
	{
		int value = srcp.immediate() & 3;
		a.mov(MABS(&m_state.fmod), value);                                              // mov   [fmod],srcp
		a.ldmxcsr(MABS(&m_near.ssecontrol[value]));                                     // ldmxcsr fp_control[srcp]
	}

	// register/memory case
	else
	{
		mov_reg_param(a, eax, srcp);                                                    // mov   eax,srcp
		a.and_(eax, 3);                                                                 // and   eax,3
		a.mov(MABS(&m_state.fmod), al);                                                 // mov   [fmod],al
		a.ldmxcsr(ptr(rbp, rax, 2, offset_from_rbp(&m_near.ssecontrol[0])));            // ldmxcsr fp_control[eax]
	}
}


//-------------------------------------------------
//  op_getfmod - process a GETFMOD opcode
//-------------------------------------------------

void drcbe_x64::op_getfmod(Assembler &a, const instruction &inst)
{
	// validate instruction
	assert(inst.size() == 4);
	assert_no_condition(inst);
	assert_no_flags(inst);

	// normalize parameters
	be_parameter dstp(*this, inst.param(0), PTYPE_MR);

	Mem fmod = MABS(&m_state.fmod);
	fmod.setSize(1);

	// fetch the current mode and store to the destination
	if (dstp.is_int_register())
		a.movzx(Gpd(dstp.ireg()), fmod);                                                // movzx reg,[fmod]
	else
	{
		a.movzx(eax, fmod);                                                             // movzx eax,[fmod]
		a.mov(MABS(dstp.memory()), eax);                                                // mov   [dstp],eax
	}
}


//-------------------------------------------------
//  op_getexp - process a GETEXP opcode
//-------------------------------------------------

void drcbe_x64::op_getexp(Assembler &a, const instruction &inst)
{
	// validate instruction
	assert(inst.size() == 4);
	assert_no_condition(inst);
	assert_no_flags(inst);

	// normalize parameters
	be_parameter dstp(*this, inst.param(0), PTYPE_MR);

	// fetch the exception parameter and store to the destination
	if (dstp.is_int_register())
		a.mov(Gpd(dstp.ireg()), MABS(&m_state.exp));                                    // mov   reg,[exp]
	else
	{
		a.mov(eax, MABS(&m_state.exp));                                                 // mov   eax,[exp]
		a.mov(MABS(dstp.memory()), eax);                                                // mov   [dstp],eax
	}
}


//-------------------------------------------------
//  op_getflgs - process a GETFLGS opcode
//-------------------------------------------------

void drcbe_x64::op_getflgs(Assembler &a, const instruction &inst)
{
	// validate instruction
	assert(inst.size() == 4);
	assert_no_condition(inst);
	assert_no_flags(inst);

	// normalize parameters
	be_parameter dstp(*this, inst.param(0), PTYPE_MR);
	be_parameter maskp(*this, inst.param(1), PTYPE_I);

	// pick a target register for the general case
	Gp dstreg = dstp.select_register(eax);

	// compute mask for flags
	uint32_t flagmask = 0;
	if (maskp.immediate() & FLAG_C) flagmask |= 0x001;
	if (maskp.immediate() & FLAG_V) flagmask |= 0x800;
	if (maskp.immediate() & FLAG_Z) flagmask |= 0x040;
	if (maskp.immediate() & FLAG_S) flagmask |= 0x080;
	if (maskp.immediate() & FLAG_U) flagmask |= 0x004;

	switch (maskp.immediate())
	{
		// single flags only
		case FLAG_C:
			a.setc(al);                                                                 // setc   al
			a.movzx(dstreg, al);                                                        // movzx  dstreg,al
			break;

		case FLAG_V:
			a.seto(al);                                                                 // seto   al
			a.movzx(dstreg, al);                                                        // movzx  dstreg,al
			a.shl(dstreg, 1);                                                           // shl    dstreg,1
			break;

		case FLAG_Z:
			a.setz(al);                                                                 // setz   al
			a.movzx(dstreg, al);                                                        // movzx  dstreg,al
			a.shl(dstreg, 2);                                                           // shl    dstreg,2
			break;

		case FLAG_S:
			a.sets(al);                                                                 // sets   al
			a.movzx(dstreg, al);                                                        // movzx  dstreg,al
			a.shl(dstreg, 3);                                                           // shl    dstreg,3
			break;

		case FLAG_U:
			a.setp(al);                                                                 // setp   al
			a.movzx(dstreg, al);                                                        // movzx  dstreg,al
			a.shl(dstreg, 4);                                                           // shl    dstreg,4
			break;

		// carry plus another flag
		case FLAG_C | FLAG_V:
			a.setc(al);                                                                 // setc   al
			a.seto(cl);                                                                 // seto   cl
			a.movzx(eax, al);                                                           // movzx  eax,al
			a.movzx(ecx, cl);                                                           // movzx  ecx,cl
			a.lea(dstreg, ptr(eax, ecx, 1));                                            // lea    dstreg,[eax+ecx*2]
			break;

		case FLAG_C | FLAG_Z:
			a.setc(al);                                                                 // setc   al
			a.setz(cl);                                                                 // setz   cl
			a.movzx(eax, al);                                                           // movzx  eax,al
			a.movzx(ecx, cl);                                                           // movzx  ecx,cl
			a.lea(dstreg, ptr(eax, ecx, 2));                                            // lea    dstreg,[eax+ecx*4]
			break;

		case FLAG_C | FLAG_S:
			a.setc(al);                                                                 // setc   al
			a.sets(cl);                                                                 // sets   cl
			a.movzx(eax, al);                                                           // movzx  eax,al
			a.movzx(ecx, cl);                                                           // movzx  ecx,cl
			a.lea(dstreg, ptr(eax, ecx, 3));                                            // lea    dstreg,[eax+ecx*8]
			break;

		// overflow plus another flag
		case FLAG_V | FLAG_Z:
			a.seto(al);                                                                 // seto   al
			a.setz(cl);                                                                 // setz   cl
			a.movzx(eax, al);                                                           // movzx  eax,al
			a.movzx(ecx, cl);                                                           // movzx  ecx,cl
			a.lea(dstreg, ptr(eax, ecx, 1));                                            // lea    dstreg,[eax+ecx*2]
			a.shl(dstreg, 1);                                                           // shl    dstreg,1
			break;

		case FLAG_V | FLAG_S:
			a.seto(al);                                                                 // seto   al
			a.sets(cl);                                                                 // sets   cl
			a.movzx(eax, al);                                                           // movzx  eax,al
			a.movzx(ecx, cl);                                                           // movzx  ecx,cl
			a.lea(dstreg, ptr(eax, ecx, 2));                                            // lea    dstreg,[eax+ecx*4]
			a.shl(dstreg, 1);                                                           // shl    dstreg,1
			break;

		// zero plus another flag
		case FLAG_Z | FLAG_S:
			a.setz(al);                                                                 // setz   al
			a.sets(cl);                                                                 // sets   cl
			a.movzx(eax, al);                                                           // movzx  eax,al
			a.movzx(ecx, cl);                                                           // movzx  ecx,cl
			a.lea(dstreg, ptr(eax, ecx, 1));                                            // lea    dstreg,[eax+ecx*2]
			a.shl(dstreg, 2);                                                           // shl    dstreg,2
			break;

		// default cases
		default:
			a.pushfq();                                                                 // pushf
			a.pop(eax);                                                                 // pop    eax
			a.and_(eax, flagmask);                                                      // and    eax,flagmask
			a.movzx(dstreg, byte_ptr(rbp, rax, 0, offset_from_rbp(&m_near.flagsmap[0]))); // movzx  dstreg,[flags_map]
			break;
	}

	// 32-bit form
	if (inst.size() == 4)
		mov_param_reg(a, dstp, dstreg);                                                 // mov   dstp,dstreg

	// 64-bit form
	else if (inst.size() == 8)
		mov_param_reg(a, dstp, dstreg.r64());                                           // mov   dstp,dstreg
}


//-------------------------------------------------
//  op_save - process a SAVE opcode
//-------------------------------------------------

void drcbe_x64::op_save(Assembler &a, const instruction &inst)
{
	// validate instruction
	assert(inst.size() == 4);
	assert_no_condition(inst);
	assert_no_flags(inst);

	// normalize parameters
	be_parameter dstp(*this, inst.param(0), PTYPE_M);

	// copy live state to the destination
	mov_r64_imm(a, rcx, (uintptr_t)dstp.memory());                                      // mov    rcx,dstp

	// copy flags
	a.pushfq();                                                                         // pushf
	a.pop(rax);                                                                         // pop    rax
	a.and_(eax, 0x8c5);                                                                 // and    eax,0x8c5
	a.mov(al, ptr(rbp, rax, 0, offset_from_rbp(&m_near.flagsmap[0])));                  // mov    al,[flags_map]
	a.mov(ptr(rcx, offsetof(drcuml_machine_state, flags)), al);                         // mov    state->flags,al

	// copy fmod and exp
	a.mov(al, MABS(&m_state.fmod));                                                     // mov    al,[fmod]
	a.mov(ptr(rcx, offsetof(drcuml_machine_state, fmod)), al);                          // mov    state->fmod,al
	a.mov(eax, MABS(&m_state.exp));                                                     // mov    eax,[exp]
	a.mov(ptr(rcx, offsetof(drcuml_machine_state, exp)), eax);                          // mov    state->exp,eax

	// copy integer registers
	int regoffs = offsetof(drcuml_machine_state, r);
	for (int regnum = 0; regnum < std::size(m_state.r); regnum++)
	{
		if (int_register_map[regnum] != 0)
			a.mov(ptr(rcx, regoffs + 8 * regnum), Gpq(regnum));
		else
		{
			a.mov(rax, MABS(&m_state.r[regnum].d));
			a.mov(ptr(rcx, regoffs + 8 * regnum), rax);
		}
	}

	// copy FP registers
	regoffs = offsetof(drcuml_machine_state, f);
	for (int regnum = 0; regnum < std::size(m_state.f); regnum++)
	{
		if (float_register_map[regnum] != 0)
			a.movsd(ptr(rcx, regoffs + 8 * regnum), Xmm(regnum));
		else
		{
			a.mov(rax, MABS(&m_state.f[regnum].d));
			a.mov(ptr(rcx, regoffs + 8 * regnum), rax);
		}
	}
}


//-------------------------------------------------
//  op_restore - process a RESTORE opcode
//-------------------------------------------------

void drcbe_x64::op_restore(Assembler &a, const instruction &inst)
{
	// validate instruction
	assert(inst.size() == 4);
	assert_no_condition(inst);

	// normalize parameters
	be_parameter srcp(*this, inst.param(0), PTYPE_M);

	// copy live state from the destination
	mov_r64_imm(a, rcx, (uintptr_t)srcp.memory());                                      // mov    rcx,dstp

	// copy integer registers
	int regoffs = offsetof(drcuml_machine_state, r);
	for (int regnum = 0; regnum < std::size(m_state.r); regnum++)
	{
		if (int_register_map[regnum] != 0)
			a.mov(Gpq(regnum), ptr(rcx, regoffs + 8 * regnum));
		else
		{
			a.mov(rax, ptr(rcx, regoffs + 8 * regnum));
			a.mov(MABS(&m_state.r[regnum].d), rax);
		}
	}

	// copy FP registers
	regoffs = offsetof(drcuml_machine_state, f);
	for (int regnum = 0; regnum < std::size(m_state.f); regnum++)
	{
		if (float_register_map[regnum] != 0)
			a.movsd(Xmm(regnum), ptr(rcx, regoffs + 8 * regnum));
		else
		{
			a.mov(rax, ptr(rcx, regoffs + 8 * regnum));
			a.mov(MABS(&m_state.f[regnum].d), rax);
		}
	}

	Mem fmod = MABS(&m_state.fmod);
	fmod.setSize(1);

	// copy fmod and exp
	a.movzx(eax, byte_ptr(rcx, offsetof(drcuml_machine_state, fmod)));                  // movzx  eax,state->fmod
	a.and_(eax, 3);                                                                     // and    eax,3
	a.mov(MABS(&m_state.fmod), al);                                                     // mov    [fmod],al
	a.ldmxcsr(ptr(rbp, rax, 2, offset_from_rbp(&m_near.ssecontrol[0])));                // ldmxcsr fp_control[eax]
	a.mov(eax, ptr(rcx, offsetof(drcuml_machine_state, exp)));                          // mov    eax,state->exp
	a.mov(MABS(&m_state.exp), eax);                                                     // mov    [exp],eax

	// copy flags
	a.movzx(eax, byte_ptr(rcx, offsetof(drcuml_machine_state, flags)));                 // movzx  eax,state->flags
	a.push(ptr(rbp, rax, 3, offset_from_rbp(&m_near.flagsunmap[0])));                   // push   flags_unmap[eax*8]
	a.popfq();                                                                          // popf
}



/***************************************************************************
    INTEGER OPERATIONS
***************************************************************************/

//-------------------------------------------------
//  op_load - process a LOAD opcode
//-------------------------------------------------

void drcbe_x64::op_load(Assembler &a, const instruction &inst)
{
	// validate instruction
	assert(inst.size() == 4 || inst.size() == 8);
	assert_no_condition(inst);
	assert_no_flags(inst);

	// normalize parameters
	be_parameter dstp(*this, inst.param(0), PTYPE_MR);
	be_parameter basep(*this, inst.param(1), PTYPE_M);
	be_parameter indp(*this, inst.param(2), PTYPE_MRI);
	const parameter &scalesizep = inst.param(3);
	assert(scalesizep.is_size_scale());
	int size = scalesizep.size();

	// determine the pointer base
	int32_t baseoffs;
	Gp basereg = get_base_register_and_offset(a, basep.memory(), rdx, baseoffs);

	// pick a target register for the general case
	Gp dstreg = dstp.select_register(eax);

	// immediate index
	if (indp.is_immediate())
	{
		s32 const offset = baseoffs + (s32(indp.immediate()) << scalesizep.scale());

		if (size == SIZE_BYTE)
			a.movzx(dstreg, byte_ptr(basereg, offset));                                 // movzx dstreg,[basep + scale*indp]
		else if (size == SIZE_WORD)
			a.movzx(dstreg, word_ptr(basereg, offset));                                 // movzx dstreg,[basep + scale*indp]
		else if (size == SIZE_DWORD)
			a.mov(dstreg, ptr(basereg, offset));                                        // mov   dstreg,[basep + scale*indp]
		else if (size == SIZE_QWORD)
			a.mov(dstreg.r64(), ptr(basereg, offset));                                  // mov   dstreg,[basep + scale*indp]
	}

	// other index
	else
	{
		Gp indreg = indp.select_register(rcx);
		movsx_r64_p32(a, indreg, indp);
		if (size == SIZE_BYTE)
			a.movzx(dstreg, byte_ptr(basereg, indreg, scalesizep.scale(), baseoffs));   // movzx dstreg,[basep + scale*indp]
		else if (size == SIZE_WORD)
			a.movzx(dstreg, word_ptr(basereg, indreg, scalesizep.scale(), baseoffs));   // movzx dstreg,[basep + scale*indp]
		else if (size == SIZE_DWORD)
			a.mov(dstreg, ptr(basereg, indreg, scalesizep.scale(), baseoffs));          // mov   dstreg,[basep + scale*indp]
		else if (size == SIZE_QWORD)
			a.mov(dstreg.r64(), ptr(basereg, indreg, scalesizep.scale(), baseoffs));    // mov   dstreg,[basep + scale*indp]
	}

	// store result
	if (inst.size() == 4)
		mov_param_reg(a, dstp, dstreg);                                                 // mov   dstp,dstreg
	else
		mov_param_reg(a, dstp, dstreg.r64());                                           // mov   dstp,dstreg
}


//-------------------------------------------------
//  op_loads - process a LOADS opcode
//-------------------------------------------------

void drcbe_x64::op_loads(Assembler &a, const instruction &inst)
{
	// validate instruction
	assert(inst.size() == 4 || inst.size() == 8);
	assert_no_condition(inst);
	assert_no_flags(inst);

	// normalize parameters
	be_parameter dstp(*this, inst.param(0), PTYPE_MR);
	be_parameter basep(*this, inst.param(1), PTYPE_M);
	be_parameter indp(*this, inst.param(2), PTYPE_MRI);
	const parameter &scalesizep = inst.param(3);
	assert(scalesizep.is_size_scale());
	int size = scalesizep.size();

	// determine the pointer base
	int32_t baseoffs;
	Gp basereg = get_base_register_and_offset(a, basep.memory(), rdx, baseoffs);

	// pick a target register for the general case
	Gp dstreg = (inst.size() == 4) ? dstp.select_register(eax) : dstp.select_register(rax);

	// immediate index
	if (indp.is_immediate())
	{
		s32 const offset = baseoffs + (s32(indp.immediate()) << scalesizep.scale());

		if (size == SIZE_BYTE)
			a.movsx(dstreg, byte_ptr(basereg, offset));                                 // movsx dstreg,[basep + scale*indp]
		else if (size == SIZE_WORD)
			a.movsx(dstreg, word_ptr(basereg, offset));                                 // movsx dstreg,[basep + scale*indp]
		else if (size == SIZE_DWORD && inst.size() == 4)
			a.mov(dstreg, ptr(basereg, offset));                                        // mov   dstreg,[basep + scale*indp]
		else if (size == SIZE_DWORD)
			a.movsxd(dstreg, ptr(basereg, offset));                                     // movsxd dstreg,[basep + scale*indp]
		else if (size == SIZE_QWORD)
			a.mov(dstreg, ptr(basereg, offset));                                        // mov   dstreg,[basep + scale*indp]
	}

	// other index
	else
	{
		Gp indreg = indp.select_register(rcx);
		movsx_r64_p32(a, indreg, indp);
		if (size == SIZE_BYTE)
			a.movsx(dstreg, byte_ptr(basereg, indreg, scalesizep.scale(), baseoffs));   // movsx dstreg,[basep + scale*indp]
		else if (size == SIZE_WORD)
			a.movsx(dstreg, word_ptr(basereg, indreg, scalesizep.scale(), baseoffs));   // movsx dstreg,[basep + scale*indp]
		else if (size == SIZE_DWORD && inst.size() == 4)
			a.mov(dstreg, ptr(basereg, indreg, scalesizep.scale(), baseoffs));          // mov   dstreg,[basep + scale*indp]
		else if (size == SIZE_DWORD)
			a.movsxd(dstreg, ptr(basereg, indreg, scalesizep.scale(), baseoffs));       // movsxd dstreg,[basep + scale*indp]
		else if (size == SIZE_QWORD)
			a.mov(dstreg, ptr(basereg, indreg, scalesizep.scale(), baseoffs));          // mov   dstreg,[basep + scale*indp]
	}

	// store result
	mov_param_reg(a, dstp, dstreg);                                                     // mov   dstp,dstreg
}


//-------------------------------------------------
//  op_store - process a STORE opcode
//-------------------------------------------------

void drcbe_x64::op_store(Assembler &a, const instruction &inst)
{
	// validate instruction
	assert(inst.size() == 4 || inst.size() == 8);
	assert_no_condition(inst);
	assert_no_flags(inst);

	// normalize parameters
	be_parameter basep(*this, inst.param(0), PTYPE_M);
	be_parameter indp(*this, inst.param(1), PTYPE_MRI);
	be_parameter srcp(*this, inst.param(2), PTYPE_MRI);
	const parameter &scalesizep = inst.param(3);
	int size = scalesizep.size();

	// determine the pointer base
	int32_t baseoffs;
	Gp basereg = get_base_register_and_offset(a, basep.memory(), rdx, baseoffs);

	// pick a source register for the general case
	Gp srcreg = srcp.select_register(rax);

	// degenerate case: constant index
	if (indp.is_immediate())
	{
		s32 const offset = baseoffs + (s32(indp.immediate()) << scalesizep.scale());

		// immediate source
		if (srcp.is_immediate())
		{
			if (size == SIZE_QWORD)
			{
				if (short_immediate(srcp.immediate()))
					a.mov(qword_ptr(basereg, offset), s32(srcp.immediate()));           // mov   [basep + scale*indp],srcp
				else
				{
					a.mov(ptr(basereg, offset + 0), u32(srcp.immediate() >>  0));       // mov   [basep + scale*indp],srcp
					a.mov(ptr(basereg, offset + 4), u32(srcp.immediate() >> 32));       // mov   [basep + scale*indp + 4],srcp >> 32
				}
			}
			else
				a.mov(ptr(basereg, offset, 1 << size), srcp.immediate());               // mov   [basep + scale*indp],srcp
		}

		// variable source
		else
		{
			if (size != SIZE_QWORD)
				mov_reg_param(a, srcreg.r32(), srcp);                                   // mov   srcreg,srcp
			else
				mov_reg_param(a, srcreg.r64(), srcp);                                   // mov   srcreg,srcp
			if (size == SIZE_BYTE)
				a.mov(ptr(basereg, offset), srcreg.r8());                               // mov   [basep + scale*indp],srcreg
			else if (size == SIZE_WORD)
				a.mov(ptr(basereg, offset), srcreg.r16());                              // mov   [basep + scale*indp],srcreg
			else if (size == SIZE_DWORD)
				a.mov(ptr(basereg, offset), srcreg.r32());                              // mov   [basep + scale*indp],srcreg
			else if (size == SIZE_QWORD)
				a.mov(ptr(basereg, offset), srcreg.r64());                              // mov   [basep + scale*indp],srcreg
		}
	}

	// normal case: variable index
	else
	{
		Gp indreg = indp.select_register(rcx);
		movsx_r64_p32(a, indreg, indp);                                                 // mov   indreg,indp

		// immediate source
		if (srcp.is_immediate())
		{
			if (size == SIZE_QWORD)
			{
				if (short_immediate(srcp.immediate()))
					a.mov(qword_ptr(basereg, indreg, scalesizep.scale(), baseoffs), s32(srcp.immediate()));     // mov   [basep + scale*indp],srcp
				else
				{
					a.mov(ptr(basereg, indreg, scalesizep.scale(), baseoffs + 0), u32(srcp.immediate() >>  0)); // mov   [basep + scale*ecx],srcp
					a.mov(ptr(basereg, indreg, scalesizep.scale(), baseoffs + 4), u32(srcp.immediate() >> 32)); // mov   [basep + scale*ecx + 4],srcp >> 32
				}
			}
			else
				a.mov(ptr(basereg, indreg, scalesizep.scale(), baseoffs, 1 << size), srcp.immediate());         // mov   [basep + scale*ecx],srcp
		}

		// variable source
		else
		{
			if (size != SIZE_QWORD)
				mov_reg_param(a, srcreg.r32(), srcp);                                   // mov   srcreg,srcp
			else
				mov_reg_param(a, srcreg.r64(), srcp);                                   // mov   edx:srcreg,srcp
			if (size == SIZE_BYTE)
				a.mov(ptr(basereg, indreg, scalesizep.scale(), baseoffs), srcreg.r8()); // mov   [basep + scale*ecx],srcreg
			else if (size == SIZE_WORD)
				a.mov(ptr(basereg, indreg, scalesizep.scale(), baseoffs), srcreg.r16());// mov   [basep + scale*ecx],srcreg
			else if (size == SIZE_DWORD)
				a.mov(ptr(basereg, indreg, scalesizep.scale(), baseoffs), srcreg.r32());// mov   [basep + scale*ecx],srcreg
			else if (size == SIZE_QWORD)
				a.mov(ptr(basereg, indreg, scalesizep.scale(), baseoffs), srcreg.r64());// mov   [basep + scale*ecx],srcreg
		}
	}
}


//-------------------------------------------------
//  op_read - process a READ opcode
//-------------------------------------------------

void drcbe_x64::op_read(Assembler &a, const instruction &inst)
{
	// validate instruction
	assert(inst.size() == 4 || inst.size() == 8);
	assert_no_condition(inst);
	assert_no_flags(inst);

	// normalize parameters
	be_parameter dstp(*this, inst.param(0), PTYPE_MR);
	be_parameter addrp(*this, inst.param(1), PTYPE_MRI);
	const parameter &spacesizep = inst.param(2);
	assert(spacesizep.is_size_space());

	// pick a target register for the general case
	Gp dstreg = dstp.select_register(eax);

	// set up a call to the read handler
	auto &trampolines = m_accessors[spacesizep.space()];
	auto &resolved = m_resolved_accessors[spacesizep.space()];
	mov_reg_param(a, Gpd(REG_PARAM2), addrp);                                           // mov    param2,addrp
	if (spacesizep.size() == SIZE_BYTE)
	{
		if (resolved.read_byte.func)
		{
			mov_r64_imm(a, Gpq(REG_PARAM1), resolved.read_byte.obj);                    // mov    param1,space
			smart_call_r64(a, resolved.read_byte.func, rax);                            // call   read_byte
		}
		else
		{
			mov_r64_imm(a, Gpq(REG_PARAM1), (uintptr_t)m_space[spacesizep.space()]);    // mov    param1,space
			smart_call_m64(a, (x86code **)&trampolines.read_byte);                      // call   read_byte
		}
		a.movzx(dstreg, al);                                                            // movzx  dstreg,al
	}
	else if (spacesizep.size() == SIZE_WORD)
	{
		if (resolved.read_word.func)
		{
			mov_r64_imm(a, Gpq(REG_PARAM1), resolved.read_word.obj);                    // mov    param1,space
			smart_call_r64(a, resolved.read_word.func, rax);                            // call   read_word
		}
		else
		{
			mov_r64_imm(a, Gpq(REG_PARAM1), (uintptr_t)m_space[spacesizep.space()]);    // mov    param1,space
			smart_call_m64(a, (x86code **)&trampolines.read_word);                      // call   read_word
		}
		a.movzx(dstreg, ax);                                                            // movzx  dstreg,ax
	}
	else if (spacesizep.size() == SIZE_DWORD)
	{
		if (resolved.read_dword.func)
		{
			mov_r64_imm(a, Gpq(REG_PARAM1), resolved.read_dword.obj);                   // mov    param1,space
			smart_call_r64(a, resolved.read_dword.func, rax);                           // call   read_dword
		}
		else
		{
			mov_r64_imm(a, Gpq(REG_PARAM1), (uintptr_t)m_space[spacesizep.space()]);    // mov    param1,space
			smart_call_m64(a, (x86code **)&trampolines.read_dword);                     // call   read_dword
		}
		if (dstreg != eax || inst.size() == 8)
			a.mov(dstreg, eax);                                                         // mov    dstreg,eax
	}
	else if (spacesizep.size() == SIZE_QWORD)
	{
		if (resolved.read_qword.func)
		{
			mov_r64_imm(a, Gpq(REG_PARAM1), resolved.read_qword.obj);                   // mov    param1,space
			smart_call_r64(a, resolved.read_qword.func, rax);                           // call   read_qword
		}
		else
		{
			mov_r64_imm(a, Gpq(REG_PARAM1), (uintptr_t)m_space[spacesizep.space()]);    // mov    param1,space
			smart_call_m64(a, (x86code **)&trampolines.read_qword);                     // call   read_qword
		}
		if (dstreg != eax)
			a.mov(dstreg.r64(), rax);                                                   // mov    dstreg,rax
	}

	// store result
	if (inst.size() == 4)
		mov_param_reg(a, dstp, dstreg);                                                 // mov   dstp,dstreg
	else
		mov_param_reg(a, dstp, dstreg.r64());                                           // mov   dstp,dstreg
}


//-------------------------------------------------
//  op_readm - process a READM opcode
//-------------------------------------------------

void drcbe_x64::op_readm(Assembler &a, const instruction &inst)
{
	// validate instruction
	assert(inst.size() == 4 || inst.size() == 8);
	assert_no_condition(inst);
	assert_no_flags(inst);

	// normalize parameters
	be_parameter dstp(*this, inst.param(0), PTYPE_MR);
	be_parameter addrp(*this, inst.param(1), PTYPE_MRI);
	be_parameter maskp(*this, inst.param(2), PTYPE_MRI);
	const parameter &spacesizep = inst.param(3);
	assert(spacesizep.is_size_space());

	// pick a target register for the general case
	Gp dstreg = dstp.select_register(eax);

	// set up a call to the read handler
	auto &trampolines = m_accessors[spacesizep.space()];
	auto &resolved = m_resolved_accessors[spacesizep.space()];
	mov_reg_param(a, Gpd(REG_PARAM2), addrp);                                           // mov    param2,addrp
	if (spacesizep.size() != SIZE_QWORD)
		mov_reg_param(a, Gpd(REG_PARAM3), maskp);                                       // mov    param3,maskp
	else
		mov_reg_param(a, Gpq(REG_PARAM3), maskp);                                       // mov    param3,maskp
	if (spacesizep.size() == SIZE_WORD)
	{
		if (resolved.read_word_masked.func)
		{
			mov_r64_imm(a, Gpq(REG_PARAM1), resolved.read_word_masked.obj);             // mov    param1,space
			smart_call_r64(a, resolved.read_word_masked.func, rax);                     // call   read_byte_masked
		}
		else
		{
			mov_r64_imm(a, Gpq(REG_PARAM1), (uintptr_t)m_space[spacesizep.space()]);    // mov    param1,space
			smart_call_m64(a, (x86code **)&trampolines.read_word_masked);               // call   read_word_masked
		}
		a.movzx(dstreg, ax);                                                            // movzx  dstreg,ax
	}
	else if (spacesizep.size() == SIZE_DWORD)
	{
		if (resolved.read_dword_masked.func)
		{
			mov_r64_imm(a, Gpq(REG_PARAM1), resolved.read_dword_masked.obj);            // mov    param1,space
			smart_call_r64(a, resolved.read_dword_masked.func, rax);                    // call   read_dword_masked
		}
		else
		{
			mov_r64_imm(a, Gpq(REG_PARAM1), (uintptr_t)m_space[spacesizep.space()]);    // mov    param1,space
			smart_call_m64(a, (x86code **)&trampolines.read_dword_masked);              // call   read_word_masked
		}
		if (dstreg != eax || inst.size() == 8)
			a.mov(dstreg, eax);                                                         // mov    dstreg,eax
	}
	else if (spacesizep.size() == SIZE_QWORD)
	{
		if (resolved.read_qword_masked.func)
		{
			mov_r64_imm(a, Gpq(REG_PARAM1), resolved.read_qword_masked.obj);            // mov    param1,space
			smart_call_r64(a, resolved.read_qword_masked.func, rax);                    // call   read_qword_masked
		}
		else
		{
			mov_r64_imm(a, Gpq(REG_PARAM1), (uintptr_t)m_space[spacesizep.space()]);    // mov    param1,space
			smart_call_m64(a, (x86code **)&trampolines.read_qword_masked);              // call   read_word_masked
		}
		if (dstreg != eax)
			a.mov(dstreg.r64(), rax);                                                   // mov    dstreg,rax
	}

	// store result
	if (inst.size() == 4)
		mov_param_reg(a, dstp, dstreg);                                                 // mov   dstp,dstreg
	else
		mov_param_reg(a, dstp, dstreg.r64());                                           // mov   dstp,dstreg
}


//-------------------------------------------------
//  op_write - process a WRITE opcode
//-------------------------------------------------

void drcbe_x64::op_write(Assembler &a, const instruction &inst)
{
	// validate instruction
	assert(inst.size() == 4 || inst.size() == 8);
	assert_no_condition(inst);
	assert_no_flags(inst);

	// normalize parameters
	be_parameter addrp(*this, inst.param(0), PTYPE_MRI);
	be_parameter srcp(*this, inst.param(1), PTYPE_MRI);
	const parameter &spacesizep = inst.param(2);
	assert(spacesizep.is_size_space());

	// set up a call to the write handler
	auto &trampolines = m_accessors[spacesizep.space()];
	auto &resolved = m_resolved_accessors[spacesizep.space()];
	mov_reg_param(a, Gpd(REG_PARAM2), addrp);                                           // mov    param2,addrp
	if (spacesizep.size() != SIZE_QWORD)
		mov_reg_param(a, Gpd(REG_PARAM3), srcp);                                        // mov    param3,srcp
	else
		mov_reg_param(a, Gpq(REG_PARAM3), srcp);                                        // mov    param3,srcp
	if (spacesizep.size() == SIZE_BYTE)
	{
		if (resolved.write_byte.func)
		{
			mov_r64_imm(a, Gpq(REG_PARAM1), resolved.write_byte.obj);                   // mov    param1,space
			smart_call_r64(a, resolved.write_byte.func, rax);                           // call   write_byte
		}
		else
		{
			mov_r64_imm(a, Gpq(REG_PARAM1), (uintptr_t)m_space[spacesizep.space()]);    // mov    param1,space
			smart_call_m64(a, (x86code **)&trampolines.write_byte);                     // call   write_byte
		}
	}
	else if (spacesizep.size() == SIZE_WORD)
	{
		if (resolved.write_word.func)
		{
			mov_r64_imm(a, Gpq(REG_PARAM1), resolved.write_word.obj);                   // mov    param1,space
			smart_call_r64(a, resolved.write_word.func, rax);                           // call   write_word
		}
		else
		{
			mov_r64_imm(a, Gpq(REG_PARAM1), (uintptr_t)m_space[spacesizep.space()]);    // mov    param1,space
			smart_call_m64(a, (x86code **)&trampolines.write_word);                     // call   write_word
		}
	}
	else if (spacesizep.size() == SIZE_DWORD)
	{
		if (resolved.write_dword.func)
		{
			mov_r64_imm(a, Gpq(REG_PARAM1), resolved.write_dword.obj);                  // mov    param1,space
			smart_call_r64(a, resolved.write_dword.func, rax);                          // call   write_dword
		}
		else
		{
			mov_r64_imm(a, Gpq(REG_PARAM1), (uintptr_t)m_space[spacesizep.space()]);    // mov    param1,space
			smart_call_m64(a, (x86code **)&trampolines.write_dword);                    // call   write_dword
		}
	}
	else if (spacesizep.size() == SIZE_QWORD)
	{
		if (resolved.write_qword.func)
		{
			mov_r64_imm(a, Gpq(REG_PARAM1), resolved.write_qword.obj);                  // mov    param1,space
			smart_call_r64(a, resolved.write_qword.func, rax);                          // call   write_qword
		}
		else
		{
			mov_r64_imm(a, Gpq(REG_PARAM1), (uintptr_t)m_space[spacesizep.space()]);    // mov    param1,space
			smart_call_m64(a, (x86code **)&trampolines.write_qword);                    // call   write_qword
		}
	}
}


//-------------------------------------------------
//  op_writem - process a WRITEM opcode
//-------------------------------------------------

void drcbe_x64::op_writem(Assembler &a, const instruction &inst)
{
	// validate instruction
	assert(inst.size() == 4 || inst.size() == 8);
	assert_no_condition(inst);
	assert_no_flags(inst);

	// normalize parameters
	be_parameter addrp(*this, inst.param(0), PTYPE_MRI);
	be_parameter srcp(*this, inst.param(1), PTYPE_MRI);
	be_parameter maskp(*this, inst.param(2), PTYPE_MRI);
	const parameter &spacesizep = inst.param(3);
	assert(spacesizep.is_size_space());

	// set up a call to the write handler
	auto &trampolines = m_accessors[spacesizep.space()];
	auto &resolved = m_resolved_accessors[spacesizep.space()];
	mov_reg_param(a, Gpd(REG_PARAM2), addrp);                                           // mov    param2,addrp
	if (spacesizep.size() != SIZE_QWORD)
	{
		mov_reg_param(a, Gpd(REG_PARAM3), srcp);                                        // mov    param3,srcp
		mov_reg_param(a, Gpd(REG_PARAM4), maskp);                                       // mov    param4,maskp
	}
	else
	{
		mov_reg_param(a, Gpq(REG_PARAM3), srcp);                                        // mov    param3,srcp
		mov_reg_param(a, Gpq(REG_PARAM4), maskp);                                       // mov    param4,maskp
	}
	if (spacesizep.size() == SIZE_WORD)
	{
		if (resolved.write_word.func)
		{
			mov_r64_imm(a, Gpq(REG_PARAM1), resolved.write_word_masked.obj);            // mov    param1,space
			smart_call_r64(a, resolved.write_word_masked.func, rax);                    // call   write_word_masked
		}
		else
		{
			mov_r64_imm(a, Gpq(REG_PARAM1), (uintptr_t)m_space[spacesizep.space()]);    // mov    param1,space
			smart_call_m64(a, (x86code **)&trampolines.write_word_masked);              // call   write_word_masked
		}
	}
	else if (spacesizep.size() == SIZE_DWORD)
	{
		if (resolved.write_word.func)
		{
			mov_r64_imm(a, Gpq(REG_PARAM1), resolved.write_dword_masked.obj);           // mov    param1,space
			smart_call_r64(a, resolved.write_dword_masked.func, rax);                   // call   write_dword_masked
		}
		else
		{
			mov_r64_imm(a, Gpq(REG_PARAM1), (uintptr_t)m_space[spacesizep.space()]);    // mov    param1,space
			smart_call_m64(a, (x86code **)&trampolines.write_dword_masked);             // call   write_dword_masked
		}
	}
	else if (spacesizep.size() == SIZE_QWORD)
	{
		if (resolved.write_word.func)
		{
			mov_r64_imm(a, Gpq(REG_PARAM1), resolved.write_qword_masked.obj);           // mov    param1,space
			smart_call_r64(a, resolved.write_qword_masked.func, rax);                   // call   write_qword_masked
		}
		else
		{
			mov_r64_imm(a, Gpq(REG_PARAM1), (uintptr_t)m_space[spacesizep.space()]);    // mov    param1,space
			smart_call_m64(a, (x86code **)&trampolines.write_qword_masked);             // call   write_qword_masked
		}
	}
}


//-------------------------------------------------
//  op_carry - process a CARRY opcode
//-------------------------------------------------

void drcbe_x64::op_carry(Assembler &a, const instruction &inst)
{
	// validate instruction
	assert(inst.size() == 4 || inst.size() == 8);
	assert_no_condition(inst);
	assert_flags(inst, FLAG_C);

	// normalize parameters
	be_parameter srcp(*this, inst.param(0), PTYPE_MRI);
	be_parameter bitp(*this, inst.param(1), PTYPE_MRI);

	Gp const src = Gp::fromTypeAndId((inst.size() == 4) ? RegType::kX86_Gpd : RegType::kX86_Gpq, srcp.ireg());

	// degenerate case: source is immediate
	if (srcp.is_immediate() && bitp.is_immediate())
	{
		if (srcp.immediate() & ((uint64_t)1 << bitp.immediate()))
			a.stc();
		else
			a.clc();
	}

	// load non-immediate bit numbers into a register
	if (!bitp.is_immediate())
	{
		mov_reg_param(a, ecx, bitp);
		a.and_(ecx, inst.size() * 8 - 1);
	}

	if (bitp.is_immediate())
	{
		if (srcp.is_memory())
			a.bt(MABS(srcp.memory(), inst.size()), bitp.immediate());                   // bt     [srcp],bitp
		else if (srcp.is_int_register())
			a.bt(src, bitp.immediate());                                                // bt     srcp,bitp
	}
	else
	{
		if (srcp.is_memory())
			a.bt(MABS(srcp.memory(), inst.size()), ecx);                                // bt     [srcp],ecx
		else if (srcp.is_int_register())
			a.bt(src, ecx);                                                             // bt     srcp,ecx
	}
}


//-------------------------------------------------
//  op_set - process a SET opcode
//-------------------------------------------------

void drcbe_x64::op_set(Assembler &a, const instruction &inst)
{
	// validate instruction
	assert(inst.size() == 4 || inst.size() == 8);
	assert_any_condition(inst);
	assert_no_flags(inst);

	// normalize parameters
	be_parameter dstp(*this, inst.param(0), PTYPE_MR);

	// pick a target register for the general case
	Gp dstreg = (inst.size() == 4) ? dstp.select_register(eax) : dstp.select_register(rax);

	// set to AL
	a.set(X86_CONDITION(inst.condition()), al);                                         // setcc  al
	a.movzx(dstreg.r32(), al);                                                          // movzx  dstreg,al
	mov_param_reg(a, dstp, dstreg);                                                     // mov   dstp,dstreg
}


//-------------------------------------------------
//  op_mov - process a MOV opcode
//-------------------------------------------------

void drcbe_x64::op_mov(Assembler &a, const instruction &inst)
{
	// validate instruction
	assert(inst.size() == 4 || inst.size() == 8);
	assert_any_condition(inst);
	assert_no_flags(inst);

	// normalize parameters
	be_parameter dstp(*this, inst.param(0), PTYPE_MR);
	be_parameter srcp(*this, inst.param(1), PTYPE_MRI);

	// add a conditional branch unless a conditional move is possible
	Label skip = a.newLabel();
	if (inst.condition() != uml::COND_ALWAYS && !(dstp.is_int_register() && !srcp.is_immediate()))
		a.short_().j(X86_NOT_CONDITION(inst.condition()), skip);                        // jcc   skip

	// register to memory
	if (dstp.is_memory() && srcp.is_int_register())
	{
		Gp const src = Gp::fromTypeAndId((inst.size() == 4) ? RegType::kX86_Gpd : RegType::kX86_Gpq, srcp.ireg());

		a.mov(MABS(dstp.memory()), src);                                                // mov   [dstp],srcp
	}
	// immediate to memory
	else if (dstp.is_memory() && srcp.is_immediate() && short_immediate(srcp.immediate()))
		a.mov(MABS(dstp.memory(), inst.size()), s32(srcp.immediate()));                 // mov   [dstp],srcp

	// conditional memory to register
	else if (inst.condition() != 0 && dstp.is_int_register() && srcp.is_memory())
	{
		Gp const dst = Gp::fromTypeAndId((inst.size() == 4) ? RegType::kX86_Gpd : RegType::kX86_Gpq, dstp.ireg());
		a.cmov(X86_CONDITION(inst.condition()), dst, MABS(srcp.memory()));              // cmovcc dstp,[srcp]
	}

	// conditional register to register
	else if (inst.condition() != 0 && dstp.is_int_register() && srcp.is_int_register())
	{
		Gp const src = Gp::fromTypeAndId((inst.size() == 4) ? RegType::kX86_Gpd : RegType::kX86_Gpq, srcp.ireg());
		Gp const dst = Gp::fromTypeAndId((inst.size() == 4) ? RegType::kX86_Gpd : RegType::kX86_Gpq, dstp.ireg());
		a.cmov(X86_CONDITION(inst.condition()), dst, src);                              // cmovcc dstp,srcp
	}

	// general case
	else
	{
		Gp dstreg = (inst.size() == 4) ? dstp.select_register(eax) : dstp.select_register(rax);

		mov_reg_param(a, dstreg, srcp, true);                                           // mov   dstreg,srcp
		mov_param_reg(a, dstp, dstreg);                                                 // mov   dstp,dstreg
	}

	// resolve the jump
	if (inst.condition() != uml::COND_ALWAYS && !(dstp.is_int_register() && !srcp.is_immediate()))
		a.bind(skip);
}


//-------------------------------------------------
//  op_sext - process a SEXT opcode
//-------------------------------------------------

void drcbe_x64::op_sext(Assembler &a, const instruction &inst)
{
	// validate instruction
	assert(inst.size() == 4 || inst.size() == 8);
	assert_no_condition(inst);
	assert_flags(inst, FLAG_S | FLAG_Z);

	// normalize parameters
	be_parameter dstp(*this, inst.param(0), PTYPE_MR);
	be_parameter srcp(*this, inst.param(1), PTYPE_MRI);
	const parameter &sizep = inst.param(2);
	assert(sizep.is_size());

	Gp dstreg = dstp.select_register(rax);

	// 32-bit form
	if (inst.size() == 4)
	{
		dstreg = dstreg.r32();

		// general case
		if (srcp.is_memory())
		{
			if (sizep.size() == SIZE_BYTE)
				a.movsx(dstreg, MABS(srcp.memory(), 1));                                // movsx dstreg,[srcp]
			else if (sizep.size() == SIZE_WORD)
				a.movsx(dstreg, MABS(srcp.memory(), 2));                                // movsx dstreg,[srcp]
			else if (sizep.size() == SIZE_DWORD)
				a.mov(dstreg, MABS(srcp.memory()));                                     // mov   dstreg,[srcp]
		}
		else if (srcp.is_int_register())
		{
			if (sizep.size() == SIZE_BYTE)
				a.movsx(dstreg, GpbLo(srcp.ireg()));                                    // movsx dstreg,srcp
			else if (sizep.size() == SIZE_WORD)
				a.movsx(dstreg, Gpw(srcp.ireg()));                                      // movsx dstreg,srcp
			else if (sizep.size() == SIZE_DWORD)
				a.mov(dstreg, Gpd(srcp.ireg()));                                        // mov   dstreg,srcp
		}
		mov_param_reg(a, dstp, dstreg);                                                 // mov   dstp,dstreg
	}

	// 64-bit form
	else if (inst.size() == 8)
	{
		// general case
		if (srcp.is_memory())
		{
			if (sizep.size() == SIZE_BYTE)
				a.movsx(dstreg, MABS(srcp.memory(), 1));                                // movsx dstreg,[srcp]
			else if (sizep.size() == SIZE_WORD)
				a.movsx(dstreg, MABS(srcp.memory(), 2));                                // movsx dstreg,[srcp]
			else if (sizep.size() == SIZE_DWORD)
				a.movsxd(dstreg, MABS(srcp.memory(), 4));                               // movsxd dstreg,[srcp]
			else if (sizep.size() == SIZE_QWORD)
				a.mov(dstreg, MABS(srcp.memory()));                                     // mov   dstreg,[srcp]
		}
		else if (srcp.is_int_register())
		{
			if (sizep.size() == SIZE_BYTE)
				a.movsx(dstreg, GpbLo(srcp.ireg()));                                    // movsx dstreg,srcp
			else if (sizep.size() == SIZE_WORD)
				a.movsx(dstreg, Gpw(srcp.ireg()));                                      // movsx dstreg,srcp
			else if (sizep.size() == SIZE_DWORD)
				a.movsxd(dstreg, Gpd(srcp.ireg()));                                     // movsxd dstreg,srcp
			else if (sizep.size() == SIZE_QWORD)
				a.mov(dstreg, Gpq(srcp.ireg()));                                        // mov   dstreg,srcp
		}
		mov_param_reg(a, dstp, dstreg);                                                 // mov   dstp,dstreg
	}

	if (inst.flags() != 0)
		a.test(dstreg, dstreg);                                                         // test  dstreg,dstreg
}


//-------------------------------------------------
//  op_roland - process an ROLAND opcode
//-------------------------------------------------

void drcbe_x64::op_roland(Assembler &a, const instruction &inst)
{
	// validate instruction
	assert(inst.size() == 4 || inst.size() == 8);
	assert_no_condition(inst);
	assert_flags(inst, FLAG_S | FLAG_Z);

	// normalize parameters
	be_parameter dstp(*this, inst.param(0), PTYPE_MR);
	be_parameter srcp(*this, inst.param(1), PTYPE_MRI);
	be_parameter shiftp(*this, inst.param(2), PTYPE_MRI);
	be_parameter maskp(*this, inst.param(3), PTYPE_MRI);

	// pick a target register
	Gp dstreg = (inst.size() == 4) ? dstp.select_register(eax, shiftp, maskp) : dstp.select_register(rax, shiftp, maskp);

	mov_reg_param(a, dstreg, srcp);                                                     // mov   dstreg,srcp
	if (!shiftp.is_immediate_value(0))
		shift_op_param(a, Inst::kIdRol, dstreg, shiftp);                                // rol   dstreg,shiftp
	alu_op_param(a, Inst::kIdAnd, dstreg, maskp,                                        // and   dstreg,maskp
		[inst](Assembler &a, Operand const &dst, be_parameter const &src)
		{
			// optimize all-zero and all-one cases
			if (!inst.flags() && !src.immediate())
			{
				a.xor_(dst.as<Gpd>(), dst.as<Gpd>());
				return true;
			}
			else if (!inst.flags() && ones(src.immediate(), inst.size()))
				return true;

			return false;
		});
	mov_param_reg(a, dstp, dstreg);                                                     // mov   dstp,dstreg
}


//-------------------------------------------------
//  op_rolins - process an ROLINS opcode
//-------------------------------------------------

void drcbe_x64::op_rolins(Assembler &a, const instruction &inst)
{
	// validate instruction
	assert(inst.size() == 4 || inst.size() == 8);
	assert_no_condition(inst);
	assert_flags(inst, FLAG_S | FLAG_Z);

	// normalize parameters
	be_parameter dstp(*this, inst.param(0), PTYPE_MR);
	be_parameter srcp(*this, inst.param(1), PTYPE_MRI);
	be_parameter shiftp(*this, inst.param(2), PTYPE_MRI);
	be_parameter maskp(*this, inst.param(3), PTYPE_MRI);

	// 32-bit form
	if (inst.size() == 4)
	{
		// pick a target register
		Gp dstreg = dstp.select_register(ecx, shiftp, maskp);

		mov_reg_param(a, eax, srcp);                                                    // mov   eax,srcp
		if (!shiftp.is_immediate_value(0))
			shift_op_param(a, Inst::kIdRol, eax, shiftp);                               // rol   eax,shiftp
		mov_reg_param(a, dstreg, dstp);                                                 // mov   dstreg,dstp
		if (maskp.is_immediate())
		{
			a.and_(eax, maskp.immediate());                                             // and   eax,maskp
			a.and_(dstreg, ~maskp.immediate());                                         // and   dstreg,~maskp
		}
		else
		{
			mov_reg_param(a, edx, maskp);                                               // mov   edx,maskp
			a.and_(eax, edx);                                                           // and   eax,edx
			a.not_(edx);                                                                // not   edx
			a.and_(dstreg, edx);                                                        // and   dstreg,edx
		}
		a.or_(dstreg, eax);                                                             // or    dstreg,eax
		mov_param_reg(a, dstp, dstreg);                                                 // mov   dstp,dstreg
	}

	// 64-bit form
	else if (inst.size() == 8)
	{
		// pick a target register
		Gp dstreg = dstp.select_register(rcx, shiftp, maskp);

		mov_reg_param(a, rax, srcp);                                                    // mov   rax,srcp
		mov_reg_param(a, rdx, maskp);                                                   // mov   rdx,maskp
		if (!shiftp.is_immediate_value(0))
			shift_op_param(a, Inst::kIdRol, rax, shiftp);                               // rol   rax,shiftp
		mov_reg_param(a, dstreg, dstp);                                                 // mov   dstreg,dstp
		a.and_(rax, rdx);                                                               // and   eax,rdx
		a.not_(rdx);                                                                    // not   rdx
		a.and_(dstreg, rdx);                                                            // and   dstreg,rdx
		a.or_(dstreg, rax);                                                             // or    dstreg,rax
		mov_param_reg(a, dstp, dstreg);                                                 // mov   dstp,dstreg
	}
}


//-------------------------------------------------
//  op_add - process a ADD opcode
//-------------------------------------------------

void drcbe_x64::op_add(Assembler &a, const instruction &inst)
{
	// validate instruction
	assert(inst.size() == 4 || inst.size() == 8);
	assert_no_condition(inst);
	assert_flags(inst, FLAG_C | FLAG_V | FLAG_Z | FLAG_S);

	// normalize parameters
	be_parameter dstp(*this, inst.param(0), PTYPE_MR);
	be_parameter src1p(*this, inst.param(1), PTYPE_MRI);
	be_parameter src2p(*this, inst.param(2), PTYPE_MRI);
	normalize_commutative(src1p, src2p);

	// dstp == src1p in memory
	if (dstp.is_memory() && dstp == src1p)
		alu_op_param(a, Inst::kIdAdd, MABS(dstp.memory(), inst.size()), src2p,          // add   [dstp],src2p
			[inst](Assembler &a, Operand const &dst, be_parameter const &src)
			{
				// optimize zero case
				return (!inst.flags() && !src.immediate());
			});

	// dstp == src2p in memory
	else if (dstp.is_memory() && dstp == src2p)
		alu_op_param(a, Inst::kIdAdd, MABS(dstp.memory(), inst.size()), src1p,          // add   [dstp],src1p
			[inst](Assembler &a, Operand const &dst, be_parameter const &src)
			{
				// optimize zero case
				return (!inst.flags() && !src.immediate());
			});

	// reg = reg + imm
	else if (dstp.is_int_register() && src1p.is_int_register() && src2p.is_immediate() && short_immediate(src2p.immediate()) && !inst.flags())
	{
		Gp const dst = Gp::fromTypeAndId((inst.size() == 4) ? RegType::kX86_Gpd : RegType::kX86_Gpq, dstp.ireg());
		Gp const src1 = Gp::fromTypeAndId((inst.size() == 4) ? RegType::kX86_Gpd : RegType::kX86_Gpq, src1p.ireg());

		a.lea(dst, ptr(src1, src2p.immediate()));                                       // lea   dstp,[src1p+src2p]
	}

	// reg = reg + reg
	else if (dstp.is_int_register() && src1p.is_int_register() && src2p.is_int_register() && !inst.flags())
	{
		Gp const dst = Gp::fromTypeAndId((inst.size() == 4) ? RegType::kX86_Gpd : RegType::kX86_Gpq, dstp.ireg());
		Gp const src1 = Gp::fromTypeAndId((inst.size() == 4) ? RegType::kX86_Gpd : RegType::kX86_Gpq, src1p.ireg());
		Gp const src2 = Gp::fromTypeAndId((inst.size() == 4) ? RegType::kX86_Gpd : RegType::kX86_Gpq, src2p.ireg());

		a.lea(dst, ptr(src1, src2));                                                    // lea   dstp,[src1p+src2p]
	}

	// general case
	else
	{
		// pick a target register for the general case
		Gp dstreg = (inst.size() == 4) ? dstp.select_register(eax, src2p) : dstp.select_register(rax, src2p);

		mov_reg_param(a, dstreg, src1p);                                                // mov   dstreg,src1p
		alu_op_param(a, Inst::kIdAdd, dstreg, src2p,                                    // add   dstreg,src2p
			[inst](Assembler &a, Operand const &dst, be_parameter const &src)
			{
				// optimize zero case
				return (!inst.flags() && !src.immediate());
			});
		mov_param_reg(a, dstp, dstreg);                                                 // mov   dstp,dstreg
	}
}


//-------------------------------------------------
//  op_addc - process a ADDC opcode
//-------------------------------------------------

void drcbe_x64::op_addc(Assembler &a, const instruction &inst)
{
	// validate instruction
	assert(inst.size() == 4 || inst.size() == 8);
	assert_no_condition(inst);
	assert_flags(inst, FLAG_C | FLAG_V | FLAG_Z | FLAG_S);

	// normalize parameters
	be_parameter dstp(*this, inst.param(0), PTYPE_MR);
	be_parameter src1p(*this, inst.param(1), PTYPE_MRI);
	be_parameter src2p(*this, inst.param(2), PTYPE_MRI);
	normalize_commutative(src1p, src2p);

	// dstp == src1p in memory
	if (dstp.is_memory() && dstp == src1p)
		alu_op_param(a, Inst::kIdAdc, MABS(dstp.memory(), inst.size()), src2p);         // adc   [dstp],src2p

	// general case
	else
	{
		// pick a target register for the general case
		Gp dstreg = (inst.size() == 4) ? dstp.select_register(eax, src2p) : dstp.select_register(rax, src2p);

		mov_reg_param(a, dstreg, src1p, true);                                          // mov   dstreg,src1p
		alu_op_param(a, Inst::kIdAdc, dstreg, src2p);                                   // adc   dstreg,src2p
		mov_param_reg(a, dstp, dstreg);                                                 // mov   dstp,dstreg
	}
}


//-------------------------------------------------
//  op_sub - process a SUB opcode
//-------------------------------------------------

void drcbe_x64::op_sub(Assembler &a, const instruction &inst)
{
	// validate instruction
	assert(inst.size() == 4 || inst.size() == 8);
	assert_no_condition(inst);
	assert_flags(inst, FLAG_C | FLAG_V | FLAG_Z | FLAG_S);

	// normalize parameters
	be_parameter dstp(*this, inst.param(0), PTYPE_MR);
	be_parameter src1p(*this, inst.param(1), PTYPE_MRI);
	be_parameter src2p(*this, inst.param(2), PTYPE_MRI);

	// dstp == src1p in memory
	if (dstp.is_memory() && dstp == src1p)
		alu_op_param(a, Inst::kIdSub, MABS(dstp.memory(), inst.size()), src2p,          // sub   [dstp],src2p
			[inst](Assembler &a, Operand const &dst, be_parameter const &src)
			{
				// optimize zero case
				return (!inst.flags() && !src.immediate());
			});

	// reg = reg - imm
	else if (dstp.is_int_register() && src1p.is_int_register() && src2p.is_immediate() && short_immediate(src2p.immediate()) && !inst.flags())
	{
		Gp const dst = Gp::fromTypeAndId((inst.size() == 4) ? RegType::kX86_Gpd : RegType::kX86_Gpq, dstp.ireg());
		Gp const src1 = Gp::fromTypeAndId((inst.size() == 4) ? RegType::kX86_Gpd : RegType::kX86_Gpq, src1p.ireg());

		a.lea(dst, ptr(src1, -src2p.immediate()));                                      // lea   dstp,[src1p-src2p]
	}

	// general case
	else
	{
		// pick a target register for the general case
		Gp dstreg = (inst.size() == 4) ? dstp.select_register(eax, src2p) : dstp.select_register(rax, src2p);

		mov_reg_param(a, dstreg, src1p);                                                // mov   dstreg,src1p
		alu_op_param(a, Inst::kIdSub, dstreg, src2p,                                    // sub   dstreg,src2p
			[inst](Assembler &a, Operand const &dst, be_parameter const &src)
			{
				// optimize zero case
				return (!inst.flags() && !src.immediate());
			});
		mov_param_reg(a, dstp, dstreg);                                                 // mov   dstp,dstreg
	}
}


//-------------------------------------------------
//  op_subc - process a SUBC opcode
//-------------------------------------------------

void drcbe_x64::op_subc(Assembler &a, const instruction &inst)
{
	// validate instruction
	assert(inst.size() == 4 || inst.size() == 8);
	assert_no_condition(inst);
	assert_flags(inst, FLAG_C | FLAG_V | FLAG_Z | FLAG_S);

	// normalize parameters
	be_parameter dstp(*this, inst.param(0), PTYPE_MR);
	be_parameter src1p(*this, inst.param(1), PTYPE_MRI);
	be_parameter src2p(*this, inst.param(2), PTYPE_MRI);

	// dstp == src1p in memory
	if (dstp.is_memory() && dstp == src1p)
		alu_op_param(a, Inst::kIdSbb, MABS(dstp.memory(), inst.size()), src2p);         // sbb   [dstp],src2p

	// general case
	else
	{
		// pick a target register for the general case
		Gp dstreg = (inst.size() == 4) ? dstp.select_register(eax, src2p) : dstp.select_register(rax, src2p);

		mov_reg_param(a, dstreg, src1p, true);                                          // mov   dstreg,src1p
		alu_op_param(a, Inst::kIdSbb, dstreg, src2p);                                   // sbb   dstreg,src2p
		mov_param_reg(a, dstp, dstreg);                                                 // mov   dstp,dstreg
	}
}


//-------------------------------------------------
//  op_cmp - process a CMP opcode
//-------------------------------------------------

void drcbe_x64::op_cmp(Assembler &a, const instruction &inst)
{
	// validate instruction
	assert(inst.size() == 4 || inst.size() == 8);
	assert_no_condition(inst);
	assert_flags(inst, FLAG_C | FLAG_V | FLAG_Z | FLAG_S);

	// normalize parameters
	be_parameter src1p(*this, inst.param(0), PTYPE_MRI);
	be_parameter src2p(*this, inst.param(1), PTYPE_MRI);

	// memory versus anything
	if (src1p.is_memory())
		alu_op_param(a, Inst::kIdCmp, MABS(src1p.memory(), inst.size()), src2p);        // cmp   [dstp],src2p

	// general case
	else
	{
		// pick a target register for the general case
		Gp src1reg = (inst.size() == 4) ? src1p.select_register(eax) : src1p.select_register(rax);

		if (src1p.is_immediate())
		{
			if (inst.size() == 4)
				a.mov(src1reg, src1p.immediate());                                      // mov   src1reg,imm
			else
				mov_r64_imm(a, src1reg, src1p.immediate());                             // mov   src1reg,imm
		}
		alu_op_param(a, Inst::kIdCmp, src1reg, src2p);                                  // cmp   src1reg,src2p
	}
}


//-------------------------------------------------
//  op_mulu - process a MULU opcode
//-------------------------------------------------

void drcbe_x64::op_mulu(Assembler &a, const instruction &inst)
{
	uint8_t zsflags = inst.flags() & (FLAG_Z | FLAG_S);
	uint8_t vflag = inst.flags() & FLAG_V;

	// validate instruction
	assert(inst.size() == 4 || inst.size() == 8);
	assert_no_condition(inst);
	assert_flags(inst, FLAG_V | FLAG_Z | FLAG_S);

	// normalize parameters
	be_parameter dstp(*this, inst.param(0), PTYPE_MR);
	be_parameter edstp(*this, inst.param(1), PTYPE_MR);
	be_parameter src1p(*this, inst.param(2), PTYPE_MRI);
	be_parameter src2p(*this, inst.param(3), PTYPE_MRI);
	normalize_commutative(src1p, src2p);
	bool compute_hi = (dstp != edstp);

	// 32-bit form
	if (inst.size() == 4)
	{
		// general case
		mov_reg_param(a, eax, src1p);                                                   // mov   eax,src1p
		if (src2p.is_memory())
			a.mul(MABS(src2p.memory(), 4));                                             // mul   [src2p]
		else if (src2p.is_int_register())
			a.mul(Gpd(src2p.ireg()));                                                   // mul   src2p
		else if (src2p.is_immediate())
		{
			a.mov(edx, src2p.immediate());                                              // mov   edx,src2p
			a.mul(edx);                                                                 // mul   edx
		}
		mov_param_reg(a, dstp, eax);                                                    // mov   dstp,eax
		if (compute_hi)
			mov_param_reg(a, edstp, edx);                                               // mov   edstp,edx

		// compute flags
		if (inst.flags() != 0)
		{
			if (zsflags != 0)
			{
				if (vflag)
					a.pushfq();                                                         // pushf
				if (compute_hi)
				{
					if (zsflags == FLAG_Z)
						a.or_(edx, eax);                                                // or    edx,eax
					else if (zsflags == FLAG_S)
						a.test(edx, edx);                                               // test  edx,edx
					else
					{
						a.movzx(ecx, ax);                                               // movzx ecx,ax
						a.shr(eax, 16);                                                 // shr   eax,16
						a.or_(edx, ecx);                                                // or    edx,ecx
						a.or_(edx, eax);                                                // or    edx,eax
					}
				}
				else
					a.test(eax, eax);                                                   // test  eax,eax

				// we rely on the fact that OF is cleared by all logical operations above
				if (vflag)
				{
					a.pushfq();                                                         // pushf
					a.pop(rax);                                                         // pop   rax
					a.and_(qword_ptr(rsp), ~0x84);                                      // and   [rsp],~0x84
					a.or_(ptr(rsp), rax);                                               // or    [rsp],rax
					a.popfq();                                                          // popf
				}
			}
		}
	}

	// 64-bit form
	else if (inst.size() == 8)
	{
		// general case
		mov_reg_param(a, rax, src1p);                                                   // mov   rax,src1p
		if (src2p.is_memory())
			a.mul(MABS(src2p.memory(), 8));                                             // mul   [src2p]
		else if (src2p.is_int_register())
			a.mul(Gpq(src2p.ireg()));                                                   // mul   src2p
		else if (src2p.is_immediate())
		{
			mov_r64_imm(a, rdx, src2p.immediate());                                     // mov   rdx,src2p
			a.mul(rdx);                                                                 // mul   rdx
		}
		mov_param_reg(a, dstp, rax);                                                    // mov   dstp,rax
		if (compute_hi)
			mov_param_reg(a, edstp, rdx);                                               // mov   edstp,rdx

		// compute flags
		if (inst.flags() != 0)
		{
			if (zsflags != 0)
			{
				if (vflag)
					a.pushfq();                                                         // pushf
				if (compute_hi)
				{
					if (zsflags == FLAG_Z)
						a.or_(rdx, rax);                                                // or    rdx,rax
					else if (zsflags == FLAG_S)
						a.test(rdx, rdx);                                               // test  rdx,rdx
					else
					{
						a.mov(ecx, eax);                                                // mov   ecx,eax
						a.shr(rax, 32);                                                 // shr   rax,32
						a.or_(rdx, rcx);                                                // or    rdx,rcx
						a.or_(rdx, rax);                                                // or    rdx,rax
					}
				}
				else
					a.test(rax, rax);                                                   // test  rax,rax

				// we rely on the fact that OF is cleared by all logical operations above
				if (vflag)
				{
					a.pushfq();                                                         // pushf
					a.pop(rax);                                                         // pop   rax
					a.and_(qword_ptr(rsp), ~0x84);                                      // and   [rsp],~0x84
					a.or_(ptr(rsp), rax);                                               // or    [rsp],rax
					a.popfq();                                                          // popf
				}
			}
		}
	}
}


//-------------------------------------------------
//  op_muls - process a MULS opcode
//-------------------------------------------------

void drcbe_x64::op_muls(Assembler &a, const instruction &inst)
{
	uint8_t zsflags = inst.flags() & (FLAG_Z | FLAG_S);
	uint8_t vflag =  inst.flags() & FLAG_V;

	// validate instruction
	assert(inst.size() == 4 || inst.size() == 8);
	assert_no_condition(inst);
	assert_flags(inst, FLAG_V | FLAG_Z | FLAG_S);

	// normalize parameters
	be_parameter dstp(*this, inst.param(0), PTYPE_MR);
	be_parameter edstp(*this, inst.param(1), PTYPE_MR);
	be_parameter src1p(*this, inst.param(2), PTYPE_MRI);
	be_parameter src2p(*this, inst.param(3), PTYPE_MRI);
	normalize_commutative(src1p, src2p);
	bool compute_hi = (dstp != edstp);

	// 32-bit form
	if (inst.size() == 4)
	{
		// 32-bit destination with memory/immediate or register/immediate
		if (!compute_hi && !src1p.is_immediate() && src2p.is_immediate())
		{
			Gp dstreg = dstp.is_int_register() ? Gpd(dstp.ireg()) : eax;
			if (src1p.is_memory())
				a.imul(dstreg, MABS(src1p.memory()), src2p.immediate());                // imul  dstreg,[src1p],src2p
			else if (src1p.is_int_register())
				a.imul(dstreg, Gpd(src1p.ireg()), src2p.immediate());                   // imul  dstreg,src1p,src2p
			mov_param_reg(a, dstp, dstreg);                                             // mov   dstp,eax
		}

		// 32-bit destination, general case
		else if (!compute_hi)
		{
			Gp dstreg = dstp.is_int_register() ? Gpd(dstp.ireg()) : eax;
			mov_reg_param(a, dstreg, src1p);                                            // mov   dstreg,src1p
			if (src2p.is_memory())
				a.imul(dstreg, MABS(src2p.memory()));                                   // imul  dstreg,[src2p]
			else if (src2p.is_int_register())
				a.imul(dstreg, Gpd(src2p.ireg()));                                      // imul  dstreg,src2p
			mov_param_reg(a, dstp, dstreg);                                             // mov   dstp,dstreg
		}

		// 64-bit destination, general case
		else
		{
			mov_reg_param(a, eax, src1p);                                               // mov   eax,src1p
			if (src2p.is_memory())
				a.imul(MABS(src2p.memory(), 4));                                        // imul  [src2p]
			else if (src2p.is_int_register())
				a.imul(Gpd(src2p.ireg()));                                              // imul  src2p
			else if (src2p.is_immediate())
			{
				a.mov(edx, src2p.immediate());                                          // mov   edx,src2p
				a.imul(edx);                                                            // imul  edx
			}
			mov_param_reg(a, dstp, eax);                                                // mov   dstp,eax
			mov_param_reg(a, edstp, edx);                                               // mov   edstp,edx
		}

		// compute flags
		if (inst.flags() != 0)
		{
			if (zsflags != 0)
			{
				if (vflag)
					a.pushfq();                                                         // pushf
				if (compute_hi)
				{
					if (zsflags == FLAG_Z)
						a.or_(edx, eax);                                                // or    edx,eax
					else if (zsflags == FLAG_S)
						a.test(edx, edx);                                               // test  edx,edx
					else
					{
						a.movzx(ecx, ax);                                               // movzx ecx,ax
						a.shr(eax, 16);                                                 // shr   eax,16
						a.or_(edx, ecx);                                                // or    edx,ecx
						a.or_(edx, eax);                                                // or    edx,eax
					}
				}
				else
					a.test(eax, eax);                                                   // test  eax,eax

				// we rely on the fact that OF is cleared by all logical operations above
				if (vflag)
				{
					a.pushfq();                                                         // pushf
					a.pop(rax);                                                         // pop   rax
					a.and_(qword_ptr(rsp), ~0x84);                                      // and   [rsp],~0x84
					a.or_(ptr(rsp), rax);                                               // or    [rsp],rax
					a.popfq();                                                          // popf
				}
			}
		}
	}

	// 64-bit form
	else if (inst.size() == 8)
	{
		// 64-bit destination with memory/immediate or register/immediate
		if (!compute_hi && !src1p.is_immediate() && src2p.is_immediate() && short_immediate(src2p.immediate()))
		{
			Gp dstreg = dstp.is_int_register() ? Gpq(dstp.ireg()) : rax;
			if (src1p.is_memory())
				a.imul(dstreg, MABS(src1p.memory()), src2p.immediate());                // imul  dstreg,[src1p],src2p
			else if (src1p.is_int_register())
				a.imul(dstreg, Gpq(src1p.ireg()), src2p.immediate());                   // imul  rax,src1p,src2p
			mov_param_reg(a, dstp, dstreg);                                             // mov   dstp,rax
		}

		// 64-bit destination, general case
		else if (!compute_hi)
		{
			Gp dstreg = dstp.is_int_register() ? Gpq(dstp.ireg()) : rax;
			mov_reg_param(a, dstreg, src1p);                                            // mov   dstreg,src1p
			if (src2p.is_memory())
				a.imul(dstreg, MABS(src2p.memory()));                                   // imul  dstreg,[src2p]
			else if (src2p.is_int_register())
				a.imul(dstreg, Gpq(src2p.ireg()));                                      // imul  dstreg,src2p
			mov_param_reg(a, dstp, dstreg);                                             // mov   dstp,dstreg
		}

		// 128-bit destination, general case
		else
		{
			mov_reg_param(a, rax, src1p);                                               // mov   rax,src1p
			if (src2p.is_memory())
				a.imul(MABS(src2p.memory(), 8));                                        // imul  [src2p]
			else if (src2p.is_int_register())
				a.imul(Gpq(src2p.ireg()));                                              // imul  src2p
			else if (src2p.is_immediate())
			{
				mov_r64_imm(a, rdx, src2p.immediate());                                 // mov   rdx,src2p
				a.imul(rdx);                                                            // imul  rdx
			}
			mov_param_reg(a, dstp, rax);                                                // mov   dstp,rax
			mov_param_reg(a, edstp, rdx);                                               // mov   edstp,rdx
		}

		// compute flags
		if (inst.flags() != 0)
		{
			if (zsflags != 0)
			{
				if (vflag)
					a.pushfq();                                                         // pushf
				if (compute_hi)
				{
					if (zsflags == FLAG_Z)
						a.or_(rdx, rax);                                                // or    rdx,rax
					else if (zsflags == FLAG_S)
						a.test(rdx, rdx);                                               // test  rdx,rdx
					else
					{
						a.mov(ecx, eax);                                                // mov   ecx,eax
						a.shr(rax, 32);                                                 // shr   rax,32
						a.or_(rdx, rcx);                                                // or    rdx,rcx
						a.or_(rdx, rax);                                                // or    rdx,rax
					}
				}
				else
					a.test(rax, rax);                                                   // test  rax,rax

				// we rely on the fact that OF is cleared by all logical operations above
				if (vflag)
				{
					a.pushfq();                                                         // pushf
					a.pop(rax);                                                         // pop   rax
					a.and_(qword_ptr(rsp), ~0x84);                                      // and   [rsp],~0x84
					a.or_(ptr(rsp), rax);                                               // or    [rsp],rax
					a.popfq();                                                          // popf
				}
			}
		}
	}
}


//-------------------------------------------------
//  op_divu - process a DIVU opcode
//-------------------------------------------------

void drcbe_x64::op_divu(Assembler &a, const instruction &inst)
{
	// validate instruction
	assert(inst.size() == 4 || inst.size() == 8);
	assert_no_condition(inst);
	assert_flags(inst, FLAG_V | FLAG_Z | FLAG_S);

	// normalize parameters
	be_parameter dstp(*this, inst.param(0), PTYPE_MR);
	be_parameter edstp(*this, inst.param(1), PTYPE_MR);
	be_parameter src1p(*this, inst.param(2), PTYPE_MRI);
	be_parameter src2p(*this, inst.param(3), PTYPE_MRI);
	bool compute_rem = (dstp != edstp);

	Label skip = a.newLabel();

	// 32-bit form
	if (inst.size() == 4)
	{
		// general case
		mov_reg_param(a, ecx, src2p);                                                   // mov   ecx,src2p
		if (inst.flags() != 0)
		{
			a.mov(eax, 0xa0000000);                                                     // mov   eax,0xa0000000
			a.add(eax, eax);                                                            // add   eax,eax
		}
		a.short_().jecxz(skip);                                                         // jecxz skip
		mov_reg_param(a, eax, src1p);                                                   // mov   eax,src1p
		a.xor_(edx, edx);                                                               // xor   edx,edx
		a.div(ecx);                                                                     // div   ecx
		mov_param_reg(a, dstp, eax);                                                    // mov   dstp,eax
		if (compute_rem)
			mov_param_reg(a, edstp, edx);                                               // mov   edstp,edx
		if (inst.flags() != 0)
			a.test(eax, eax);                                                           // test  eax,eax
	}

	// 64-bit form
	else if (inst.size() == 8)
	{
		// general case
		mov_reg_param(a, rcx, src2p);                                                   // mov   rcx,src2p
		if (inst.flags() != 0)
		{
			a.mov(eax, 0xa0000000);                                                     // mov   eax,0xa0000000
			a.add(eax, eax);                                                            // add   eax,eax
		}
		a.short_().jecxz(skip);                                                         // jrcxz skip
		mov_reg_param(a, rax, src1p);                                                   // mov   rax,src1p
		a.xor_(edx, edx);                                                               // xor   edx,edx
		a.div(rcx);                                                                     // div   rcx
		mov_param_reg(a, dstp, rax);                                                    // mov   dstp,rax
		if (compute_rem)
			mov_param_reg(a, edstp, rdx);                                               // mov   edstp,rdx
		if (inst.flags() != 0)
			a.test(rax, rax);                                                           // test  eax,eax
	}

	a.bind(skip);                                                                   // skip:
}


//-------------------------------------------------
//  op_divs - process a DIVS opcode
//-------------------------------------------------

void drcbe_x64::op_divs(Assembler &a, const instruction &inst)
{
	// validate instruction
	assert(inst.size() == 4 || inst.size() == 8);
	assert_no_condition(inst);
	assert_flags(inst, FLAG_V | FLAG_Z | FLAG_S);

	// normalize parameters
	be_parameter dstp(*this, inst.param(0), PTYPE_MR);
	be_parameter edstp(*this, inst.param(1), PTYPE_MR);
	be_parameter src1p(*this, inst.param(2), PTYPE_MRI);
	be_parameter src2p(*this, inst.param(3), PTYPE_MRI);
	bool compute_rem = (dstp != edstp);

	Label skip = a.newLabel();

	// 32-bit form
	if (inst.size() == 4)
	{
		// general case
		mov_reg_param(a, ecx, src2p);                                                   // mov   ecx,src2p
		if (inst.flags() != 0)
		{
			a.mov(eax, 0xa0000000);                                                     // mov   eax,0xa0000000
			a.add(eax, eax);                                                            // add   eax,eax
		}
		a.short_().jecxz(skip);                                                         // jecxz skip
		mov_reg_param(a, eax, src1p);                                                   // mov   eax,src1p
		a.cdq();                                                                        // cdq
		a.idiv(ecx);                                                                    // idiv  ecx
		mov_param_reg(a, dstp, eax);                                                    // mov   dstp,eax
		if (compute_rem)
			mov_param_reg(a, edstp, edx);                                               // mov   edstp,edx
		if (inst.flags() != 0)
			a.test(eax, eax);                                                           // test  eax,eax
	}

	// 64-bit form
	else if (inst.size() == 8)
	{
		// general case
		mov_reg_param(a, rcx, src2p);                                                   // mov   rcx,src2p
		if (inst.flags() != 0)
		{
			a.mov(eax, 0xa0000000);                                                     // mov   eax,0xa0000000
			a.add(eax, eax);                                                            // add   eax,eax
		}
		a.short_().jecxz(skip);                                                         // jrcxz skip
		mov_reg_param(a, rax, src1p);                                                   // mov   rax,src1p
		a.cqo();                                                                        // cqo
		a.idiv(rcx);                                                                    // idiv  rcx
		mov_param_reg(a, dstp, rax);                                                    // mov   dstp,rax
		if (compute_rem)
			mov_param_reg(a, edstp, rdx);                                               // mov   edstp,rdx
		if (inst.flags() != 0)
			a.test(rax, rax);                                                           // test  eax,eax
	}

	a.bind(skip);                                                                   // skip:
}


//-------------------------------------------------
//  op_and - process a AND opcode
//-------------------------------------------------

void drcbe_x64::op_and(Assembler &a, const instruction &inst)
{
	// validate instruction
	assert(inst.size() == 4 || inst.size() == 8);
	assert_no_condition(inst);
	assert_flags(inst, FLAG_Z | FLAG_S);

	// normalize parameters
	be_parameter dstp(*this, inst.param(0), PTYPE_MR);
	be_parameter src1p(*this, inst.param(1), PTYPE_MRI);
	be_parameter src2p(*this, inst.param(2), PTYPE_MRI);
	normalize_commutative(src1p, src2p);

	// pick a target register
	Gp dstreg = (inst.size() == 4) ? dstp.select_register(eax, src2p) : dstp.select_register(rax, src2p);

	// dstp == src1p in memory
	if (dstp.is_memory() && dstp == src1p)
		alu_op_param(a, Inst::kIdAnd, MABS(dstp.memory(), inst.size()), src2p,          // and   [dstp],src2p
			[inst](Assembler &a, Operand const &dst, be_parameter const &src)
			{
				// optimize all-zero and all-one cases
				if (!inst.flags() && !src.immediate())
				{
					a.mov(dst.as<Mem>(), imm(0));
					return true;
				}
				else if (!inst.flags() && ones(src.immediate(), inst.size()))
					return true;

				return false;
			});

	// dstp == src2p in memory
	else if (dstp.is_memory() && dstp == src2p)
		alu_op_param(a, Inst::kIdAnd, MABS(dstp.memory(), inst.size()), src1p,          // and   [dstp],src1p
			[inst](Assembler &a, Operand const &dst, be_parameter const &src)
			{
				// optimize all-zero and all-one cases
				if (!inst.flags() && !src.immediate())
				{
					a.mov(dst.as<Mem>(), imm(0));
					return true;
				}
				else if (!inst.flags() && ones(src.immediate(), inst.size()))
					return true;

				return false;
			});

	// immediate 0xff
	else if (src2p.is_immediate_value(0xff) && !inst.flags())
	{
		if (src1p.is_int_register())
			a.movzx(dstreg, GpbLo(src1p.ireg()));                                       // movzx dstreg,src1p
		else if (src1p.is_memory())
			a.movzx(dstreg, MABS(src1p.memory(), 1));                                   // movzx dstreg,[src1p]
		mov_param_reg(a, dstp, dstreg);                                                 // mov   dstp,dstreg
	}

	// immediate 0xffff
	else if (src2p.is_immediate_value(0xffff) && !inst.flags())
	{
		if (src1p.is_int_register())
			a.movzx(dstreg, Gpw(src1p.ireg()));                                         // movzx dstreg,src1p
		else if (src1p.is_memory())
			a.movzx(dstreg, MABS(src1p.memory(), 2));                                   // movzx dstreg,[src1p]
		mov_param_reg(a, dstp, dstreg);                                                 // mov   dstp,dstreg
	}

	// immediate 0xffffffff
	else if (src2p.is_immediate_value(0xffffffff) && !inst.flags() && inst.size() == 8)
	{
		if (dstp.is_int_register() && src1p == dstp)
			a.mov(dstreg.r32(), dstreg.r32());                                          // mov   dstreg,dstreg
		else
		{
			mov_reg_param(a, dstreg.r32(), src1p);                                      // mov   dstreg,src1p
			mov_param_reg(a, dstp, dstreg);                                             // mov   dstp,dstreg
		}
	}

	// general case
	else
	{
		mov_reg_param(a, dstreg, src1p);                                                // mov   dstreg,src1p
		alu_op_param(a, Inst::kIdAnd, dstreg, src2p,                                    // and   dstreg,src2p
			[inst](Assembler &a, Operand const &dst, be_parameter const &src)
			{
				// optimize all-zero and all-one cases
				if (!inst.flags() && !src.immediate())
				{
					a.xor_(dst.as<Gpd>(), dst.as<Gpd>());
					return true;
				}
				else if (!inst.flags() && ones(src.immediate(), inst.size()))
					return true;

				return false;
			});
		mov_param_reg(a, dstp, dstreg);                                                 // mov   dstp,dstreg
	}
}


//-------------------------------------------------
//  op_test - process a TEST opcode
//-------------------------------------------------

void drcbe_x64::op_test(Assembler &a, const instruction &inst)
{
	// validate instruction
	assert(inst.size() == 4 || inst.size() == 8);
	assert_no_condition(inst);
	assert_flags(inst, FLAG_Z | FLAG_S);

	// normalize parameters
	be_parameter src1p(*this, inst.param(0), PTYPE_MRI);
	be_parameter src2p(*this, inst.param(1), PTYPE_MRI);
	normalize_commutative(src1p, src2p);

	// src1p in memory
	if (src1p.is_memory())
		alu_op_param(a, Inst::kIdTest, MABS(src1p.memory(), inst.size()), src2p);       // test  [src1p],src2p

	// general case
	else
	{
		// pick a target register for the general case
		Gp src1reg = (inst.size() == 4) ? src1p.select_register(eax) : src1p.select_register(rax);

		mov_reg_param(a, src1reg, src1p);                                               // mov   src1reg,src1p
		alu_op_param(a, Inst::kIdTest, src1reg, src2p);                                 // test  src1reg,src2p
	}
}


//-------------------------------------------------
//  op_or - process a OR opcode
//-------------------------------------------------

void drcbe_x64::op_or(Assembler &a, const instruction &inst)
{
	// validate instruction
	assert(inst.size() == 4 || inst.size() == 8);
	assert_no_condition(inst);
	assert_flags(inst, FLAG_Z | FLAG_S);

	// normalize parameters
	be_parameter dstp(*this, inst.param(0), PTYPE_MR);
	be_parameter src1p(*this, inst.param(1), PTYPE_MRI);
	be_parameter src2p(*this, inst.param(2), PTYPE_MRI);
	normalize_commutative(src1p, src2p);

	// dstp == src1p in memory
	if (dstp.is_memory() && dstp == src1p)
		alu_op_param(a, Inst::kIdOr, MABS(dstp.memory(), inst.size()), src2p,           // or    [dstp],src2p
			[inst](Assembler &a, Operand const &dst, be_parameter const &src)
			{
				// optimize all-zero and all-one cases
				if (!inst.flags() && ones(src.immediate(), inst.size()))
				{
					a.mov(dst.as<Mem>(), imm(-1));
					return true;
				}
				else if (!inst.flags() && !src.immediate())
					return true;

				return false;
			});

	// dstp == src2p in memory
	else if (dstp.is_memory() && dstp == src2p)
		alu_op_param(a, Inst::kIdOr, MABS(dstp.memory(), inst.size()), src1p,           // or    [dstp],src1p
			[inst](Assembler &a, Operand const &dst, be_parameter const &src)
			{
				// optimize all-zero and all-one cases
				if (!inst.flags() && ones(src.immediate(), inst.size()))
				{
					a.mov(dst.as<Mem>(), imm(-1));
					return true;
				}
				else if (!inst.flags() && !src.immediate())
					return true;

				return false;
			});

	// general case
	else
	{
		// pick a target register for the general case
		Gp dstreg = (inst.size() == 4) ? dstp.select_register(eax, src2p) : dstp.select_register(rax, src2p);

		mov_reg_param(a, dstreg, src1p);                                                // mov   dstreg,src1p
		alu_op_param(a, Inst::kIdOr, dstreg, src2p,                                     // or    dstreg,src2p
			[inst](Assembler &a, Operand const &dst, be_parameter const &src)
			{
				// optimize all-zero and all-one cases
				if (!inst.flags() && ones(src.immediate(), inst.size()))
				{
					a.mov(dst.as<Gp>(), imm(-1));
					return true;
				}
				else if (!inst.flags() && !src.immediate())
					return true;

				return false;
			});
		mov_param_reg(a, dstp, dstreg);                                                 // mov   dstp,dstreg
	}
}


//-------------------------------------------------
//  op_xor - process a XOR opcode
//-------------------------------------------------

void drcbe_x64::op_xor(Assembler &a, const instruction &inst)
{
	// validate instruction
	assert(inst.size() == 4 || inst.size() == 8);
	assert_no_condition(inst);
	assert_flags(inst, FLAG_Z | FLAG_S);

	// normalize parameters
	be_parameter dstp(*this, inst.param(0), PTYPE_MR);
	be_parameter src1p(*this, inst.param(1), PTYPE_MRI);
	be_parameter src2p(*this, inst.param(2), PTYPE_MRI);
	normalize_commutative(src1p, src2p);

	// dstp == src1p in memory
	if (dstp.is_memory() && dstp == src1p)
		alu_op_param(a, Inst::kIdXor, MABS(dstp.memory(), inst.size()), src2p,          // xor   [dstp],src2p
			[inst](Assembler &a, Operand const &dst, be_parameter const &src)
			{
				// optimize all-zero and all-one cases
				if (!inst.flags() && ones(src.immediate(), inst.size()))
				{
					a.not_(dst.as<Mem>());
					return true;
				}
				else if (!inst.flags() && !src.immediate())
					return true;

				return false;
			});

	// dstp == src2p in memory
	else if (dstp.is_memory() && dstp == src2p)
		alu_op_param(a, Inst::kIdXor, MABS(dstp.memory(), inst.size()), src1p,          // xor   [dstp],src1p
			[inst](Assembler &a, Operand const &dst, be_parameter const &src)
			{
				// optimize all-zero and all-one cases
				if (!inst.flags() && ones(src.immediate(), inst.size()))
				{
					a.not_(dst.as<Mem>());
					return true;
				}
				else if (!inst.flags() && !src.immediate())
					return true;

				return false;
			});

	// dstp == src1p register
	else if (dstp.is_int_register() && dstp == src1p)
	{
		Gp const dst = Gp::fromTypeAndId((inst.size() == 4) ? RegType::kX86_Gpd : RegType::kX86_Gpq, dstp.ireg());

		alu_op_param(a, Inst::kIdXor, dst, src2p,                                       // xor   dstp,src2p
			[inst](Assembler &a, Operand const &dst, be_parameter const &src)
			{
				// optimize all-zero and all-one cases
				if (!inst.flags() && ones(src.immediate(), inst.size()))
				{
					a.not_(dst.as<Gp>());
					return true;
				}
				else if (!inst.flags() && !src.immediate())
					return true;

				return false;
			});
	}
	// general case
	else
	{
		// pick a target register for the general case
		Gp dstreg = (inst.size() == 4) ? dstp.select_register(eax, src2p) : dstp.select_register(rax, src2p);

		mov_reg_param(a, dstreg, src1p);                                                // mov   dstreg,src1p
		alu_op_param(a, Inst::kIdXor, dstreg, src2p,                                    // xor   dstreg,src2p
			[inst](Assembler &a, Operand const &dst, be_parameter const &src)
			{
				// optimize all-zero and all-one cases
				if (!inst.flags() && ones(src.immediate(), inst.size()))
				{
					a.not_(dst.as<Gp>());
					return true;
				}
				else if (!inst.flags() && !src.immediate())
					return true;

				return false;
			});
		mov_param_reg(a, dstp, dstreg);                                                 // mov   dstp,dstreg
	}
}


//-------------------------------------------------
//  op_lzcnt - process a LZCNT opcode
//-------------------------------------------------

void drcbe_x64::op_lzcnt(Assembler &a, const instruction &inst)
{
	// validate instruction
	assert(inst.size() == 4 || inst.size() == 8);
	assert_no_condition(inst);
	assert_flags(inst, FLAG_Z | FLAG_S);

	// normalize parameters
	be_parameter dstp(*this, inst.param(0), PTYPE_MR);
	be_parameter srcp(*this, inst.param(1), PTYPE_MRI);

	// 32-bit form
	if (inst.size() == 4)
	{
		// pick a target register
		Gp dstreg = dstp.select_register(eax);

		mov_reg_param(a, dstreg, srcp);                                                 // mov   dstreg,src1p
		a.mov(ecx, 32 ^ 31);                                                            // mov   ecx,32 ^ 31
		a.bsr(dstreg, dstreg);                                                          // bsr   dstreg,dstreg
		a.cmovz(dstreg, ecx);                                                           // cmovz dstreg,ecx
		a.xor_(dstreg, 31);                                                             // xor   dstreg,31
		mov_param_reg(a, dstp, dstreg);                                                 // mov   dstp,dstreg
	}

	// 64-bit form
	else if (inst.size() == 8)
	{
		// pick a target register
		Gp dstreg = dstp.select_register(rax);

		mov_reg_param(a, dstreg, srcp);                                                 // mov   dstreg,src1p
		a.mov(ecx, 64 ^ 63);                                                            // mov   ecx,64 ^ 63
		a.bsr(dstreg, dstreg);                                                          // bsr   dstreg,dstreg
		a.cmovz(dstreg, rcx);                                                           // cmovz dstreg,rcx
		a.xor_(dstreg, 63);                                                             // xor   dstreg,63
		mov_param_reg(a, dstp, dstreg);                                                 // mov   dstp,dstreg
	}
}


//-------------------------------------------------
//  op_tzcnt - process a TZCNT opcode
//-------------------------------------------------

void drcbe_x64::op_tzcnt(Assembler &a, const instruction &inst)
{
	// validate instruction
	assert(inst.size() == 4 || inst.size() == 8);
	assert_no_condition(inst);
	assert_flags(inst, FLAG_Z | FLAG_S);

	// normalize parameters
	be_parameter dstp(*this, inst.param(0), PTYPE_MR);
	be_parameter srcp(*this, inst.param(1), PTYPE_MRI);

	// 32-bit form
	if (inst.size() == 4)
	{
		Gp dstreg = dstp.select_register(eax);

		mov_reg_param(a, dstreg, srcp);                                                 // mov   dstreg,srcp
		a.mov(ecx, 32);                                                                 // mov   ecx,32
		a.bsf(dstreg, dstreg);                                                          // bsf   dstreg,dstreg
		a.cmovz(dstreg, ecx);                                                           // cmovz dstreg,ecx
		mov_param_reg(a, dstp, dstreg);                                                 // mov   dstp,dstreg
	}

	// 64-bit form
	else if (inst.size() == 8)
	{
		Gp dstreg = dstp.select_register(rax);

		mov_reg_param(a, dstreg, srcp);                                                 // mov   dstreg,srcp
		a.mov(ecx, 64);                                                                 // mov   ecx,64
		a.bsf(dstreg, dstreg);                                                          // bsf   dstreg,dstreg
		a.cmovz(dstreg, rcx);                                                           // cmovz dstreg,rcx
		mov_param_reg(a, dstp, dstreg);                                                 // mov   dstp,dstreg
	}
}


//-------------------------------------------------
//  op_bswap - process a BSWAP opcode
//-------------------------------------------------

void drcbe_x64::op_bswap(Assembler &a, const instruction &inst)
{
	// validate instruction
	assert(inst.size() == 4 || inst.size() == 8);
	assert_no_condition(inst);
	assert_flags(inst, FLAG_Z | FLAG_S);

	// normalize parameters
	be_parameter dstp(*this, inst.param(0), PTYPE_MR);
	be_parameter srcp(*this, inst.param(1), PTYPE_MRI);

	// pick a target register
	Gp dstreg = (inst.size() == 4) ? dstp.select_register(eax) : dstp.select_register(rax);

	mov_reg_param(a, dstreg, srcp);                                                     // mov   dstreg,src1p
	a.bswap(dstreg);                                                                    // bswap dstreg
	if (inst.flags() != 0)
		a.test(dstreg, dstreg);                                                         // test  dstreg,dstreg
	mov_param_reg(a, dstp, dstreg);                                                     // mov   dstp,dstreg
}

template <Inst::Id Opcode> void drcbe_x64::op_shift(Assembler &a, const uml::instruction &inst)
{
	// validate instruction
	assert(inst.size() == 4 || inst.size() == 8);
	assert_no_condition(inst);
	assert_flags(inst, FLAG_C | FLAG_Z | FLAG_S);

	// normalize parameters
	be_parameter dstp(*this, inst.param(0), PTYPE_MR);
	be_parameter src1p(*this, inst.param(1), PTYPE_MRI);
	be_parameter src2p(*this, inst.param(2), PTYPE_MRI);

	const bool carry = (Opcode == Inst::kIdRcl) || (Opcode == Inst::kIdRcr);

	// optimize immediate zero case
	if (carry || inst.flags() || !src2p.is_immediate_value(0))
	{
		// dstp == src1p in memory
		if (dstp.is_memory() && dstp == src1p)
			shift_op_param(a, Opcode, MABS(dstp.memory(), inst.size()), src2p);         // op   [dstp],src2p

		// general case
		else
		{
			// pick a target register
			Gp dstreg = (inst.size() == 4) ? dstp.select_register(eax, src2p) : dstp.select_register(rax, src2p);

			if (carry)
				mov_reg_param(a, dstreg, src1p, true);                                  // mov   dstreg,src1p
			else
				mov_reg_param(a, dstreg, src1p);                                        // mov   dstreg,src1p
			shift_op_param(a, Opcode, dstreg, src2p);                                   // op    dstreg,src2p
			mov_param_reg(a, dstp, dstreg);                                             // mov   dstp,dstreg
		}
	}
}


/***************************************************************************
    FLOATING POINT OPERATIONS
***************************************************************************/

//-------------------------------------------------
//  op_fload - process a FLOAD opcode
//-------------------------------------------------

void drcbe_x64::op_fload(Assembler &a, const instruction &inst)
{
	// validate instruction
	assert(inst.size() == 4 || inst.size() == 8);
	assert_no_condition(inst);
	assert_no_flags(inst);

	// normalize parameters
	be_parameter dstp(*this, inst.param(0), PTYPE_MF);
	be_parameter basep(*this, inst.param(1), PTYPE_M);
	be_parameter indp(*this, inst.param(2), PTYPE_MRI);

	// pick a target register for the general case
	Xmm dstreg = dstp.select_register(xmm0);

	// determine the pointer base
	int32_t baseoffs;
	Gp basereg = get_base_register_and_offset(a, basep.memory(), rdx, baseoffs);

	// 32-bit form
	if (inst.size() == 4)
	{
		if (indp.is_immediate())
			a.movss(dstreg, ptr(basereg, baseoffs + 4*indp.immediate()));               // movss  dstreg,[basep + 4*indp]
		else
		{
			Gp indreg = indp.select_register(ecx);
			mov_reg_param(a, indreg, indp);                                             // mov    indreg,indp
			a.movss(dstreg, ptr(basereg, indreg, 2, baseoffs));                         // movss  dstreg,[basep + 4*indp]
		}
		movss_p32_r128(a, dstp, dstreg);                                                // movss  dstp,dstreg
	}

	// 64-bit form
	else if (inst.size() == 8)
	{
		if (indp.is_immediate())
			a.movsd(dstreg, ptr(basereg, baseoffs + 8*indp.immediate()));               // movsd  dstreg,[basep + 8*indp]
		else
		{
			Gp indreg = indp.select_register(ecx);
			mov_reg_param(a, indreg, indp);                                             // mov    indreg,indp
			a.movsd(dstreg, ptr(basereg, indreg, 3, baseoffs));                         // movsd  dstreg,[basep + 8*indp]
		}
		movsd_p64_r128(a, dstp, dstreg);                                                // movsd  dstp,dstreg
	}
}


//-------------------------------------------------
//  op_fstore - process a FSTORE opcode
//-------------------------------------------------

void drcbe_x64::op_fstore(Assembler &a, const instruction &inst)
{
	// validate instruction
	assert(inst.size() == 4 || inst.size() == 8);
	assert_no_condition(inst);
	assert_no_flags(inst);

	// normalize parameters
	be_parameter basep(*this, inst.param(0), PTYPE_M);
	be_parameter indp(*this, inst.param(1), PTYPE_MRI);
	be_parameter srcp(*this, inst.param(2), PTYPE_MF);

	// pick a target register for the general case
	Xmm srcreg = srcp.select_register(xmm0);

	// determine the pointer base
	int32_t baseoffs;
	Gp basereg = get_base_register_and_offset(a, basep.memory(), rdx, baseoffs);

	// 32-bit form
	if (inst.size() == 4)
	{
		movss_r128_p32(a, srcreg, srcp);                                                // movss  srcreg,srcp
		if (indp.is_immediate())
			a.movss(ptr(basereg, baseoffs + 4*indp.immediate()), srcreg);               // movss  [basep + 4*indp],srcreg
		else
		{
			Gp indreg = indp.select_register(ecx);
			mov_reg_param(a, indreg, indp);                                             // mov    indreg,indp
			a.movss(ptr(basereg, indreg, 2, baseoffs), srcreg);                         // movss  [basep + 4*indp],srcreg
		}
	}

	// 64-bit form
	else if (inst.size() == 8)
	{
		movsd_r128_p64(a, srcreg, srcp);                                                // movsd  srcreg,srcp
		if (indp.is_immediate())
			a.movsd(ptr(basereg, baseoffs + 8*indp.immediate()), srcreg);               // movsd  [basep + 8*indp],srcreg
		else
		{
			Gp indreg = indp.select_register(ecx);
			mov_reg_param(a, indreg, indp);                                             // mov    indreg,indp
			a.movsd(ptr(basereg, indreg, 3, baseoffs), srcreg);                         // movsd  [basep + 8*indp],srcreg
		}
	}
}


//-------------------------------------------------
//  op_fread - process a FREAD opcode
//-------------------------------------------------

void drcbe_x64::op_fread(Assembler &a, const instruction &inst)
{
	// validate instruction
	assert(inst.size() == 4 || inst.size() == 8);
	assert_no_condition(inst);
	assert_no_flags(inst);

	// normalize parameters
	be_parameter dstp(*this, inst.param(0), PTYPE_MF);
	be_parameter addrp(*this, inst.param(1), PTYPE_MRI);
	const parameter &spacep = inst.param(2);
	assert(spacep.is_size_space());
	assert((1 << spacep.size()) == inst.size());

	// set up a call to the read dword/qword handler
	mov_r64_imm(a, Gpq(REG_PARAM1), (uintptr_t)m_space[spacep.space()]);                // mov    param1,space
	mov_reg_param(a, Gpd(REG_PARAM2), addrp);                                           // mov    param2,addrp
	if (inst.size() == 4)
		smart_call_m64(a, (x86code **)&m_accessors[spacep.space()].read_dword);         // call   read_dword
	else if (inst.size() == 8)
		smart_call_m64(a, (x86code **)&m_accessors[spacep.space()].read_qword);         // call   read_qword

	// store result
	if (inst.size() == 4)
	{
		if (dstp.is_memory())
			a.mov(MABS(dstp.memory()), eax);                                            // mov   [dstp],eax
		else if (dstp.is_float_register())
			a.movd(Xmm(dstp.freg()), eax);                                              // movd  dstp,eax
	}
	else if (inst.size() == 8)
	{
		if (dstp.is_memory())
			a.mov(MABS(dstp.memory()), rax);                                            // mov   [dstp],rax
		else if (dstp.is_float_register())
			a.movq(Xmm(dstp.freg()), rax);                                              // movq  dstp,rax
	}
}


//-------------------------------------------------
//  op_fwrite - process a FWRITE opcode
//-------------------------------------------------

void drcbe_x64::op_fwrite(Assembler &a, const instruction &inst)
{
	// validate instruction
	assert(inst.size() == 4 || inst.size() == 8);
	assert_no_condition(inst);
	assert_no_flags(inst);

	// normalize parameters
	be_parameter addrp(*this, inst.param(0), PTYPE_MRI);
	be_parameter srcp(*this, inst.param(1), PTYPE_MF);
	const parameter &spacep = inst.param(2);
	assert(spacep.is_size_space());
	assert((1 << spacep.size()) == inst.size());

	// general case
	mov_r64_imm(a, Gpq(REG_PARAM1), (uintptr_t)m_space[spacep.space()]);                // mov    param1,space
	mov_reg_param(a, Gpd(REG_PARAM2), addrp);                                           // mov    param21,addrp

	// 32-bit form
	if (inst.size() == 4)
	{
		if (srcp.is_memory())
			a.mov(Gpd(REG_PARAM3), MABS(srcp.memory()));                                // mov    param3,[srcp]
		else if (srcp.is_float_register())
			a.movd(Gpd(REG_PARAM3), Xmm(srcp.freg()));                                  // movd   param3,srcp
		smart_call_m64(a, (x86code **)&m_accessors[spacep.space()].write_dword);        // call   write_dword
	}

	// 64-bit form
	else if (inst.size() == 8)
	{
		if (srcp.is_memory())
			a.mov(Gpq(REG_PARAM3), MABS(srcp.memory()));                                // mov    param3,[srcp]
		else if (srcp.is_float_register())
			a.movq(Gpq(REG_PARAM3), Xmm(srcp.freg()));                                  // movq   param3,srcp
		smart_call_m64(a, (x86code **)&m_accessors[spacep.space()].write_qword);        // call   write_qword
	}
}


//-------------------------------------------------
//  op_fmov - process a FMOV opcode
//-------------------------------------------------

void drcbe_x64::op_fmov(Assembler &a, const instruction &inst)
{
	// validate instruction
	assert(inst.size() == 4 || inst.size() == 8);
	assert_any_condition(inst);
	assert_no_flags(inst);

	// normalize parameters
	be_parameter dstp(*this, inst.param(0), PTYPE_MF);
	be_parameter srcp(*this, inst.param(1), PTYPE_MF);

	// pick a target register for the general case
	Xmm dstreg = dstp.select_register(xmm0);

	// always start with a jmp
	Label skip = a.newLabel();
	if (inst.condition() != uml::COND_ALWAYS)
		a.short_().j(X86_NOT_CONDITION(inst.condition()), skip);                        // jcc   skip

	// 32-bit form
	if (inst.size() == 4)
	{
		if (srcp.is_float_register())
		{
			movss_p32_r128(a, dstp, Xmm(srcp.freg()));                                  // movss dstp,srcp
		}
		else
		{
			movss_r128_p32(a, dstreg, srcp);                                            // movss dstreg,srcp
			movss_p32_r128(a, dstp, dstreg);                                            // movss dstp,dstreg
		}
	}

	// 64-bit form
	else if (inst.size() == 8)
	{
		if (srcp.is_float_register())
		{
			movsd_p64_r128(a, dstp, Xmm(srcp.freg()));                                  // movsd dstp,srcp
		}
		else
		{
			movsd_r128_p64(a, dstreg, srcp);                                            // movsd dstreg,srcp
			movsd_p64_r128(a, dstp, dstreg);                                            // movsd dstp,dstreg
		}
	}

	// resolve the jump
	if (inst.condition() != uml::COND_ALWAYS)
		a.bind(skip);                                                               // skip:
}


//-------------------------------------------------
//  op_ftoint - process a FTOINT opcode
//-------------------------------------------------

void drcbe_x64::op_ftoint(Assembler &a, const instruction &inst)
{
	// validate instruction
	assert(inst.size() == 4 || inst.size() == 8);
	assert_no_condition(inst);
	assert_no_flags(inst);

	// normalize parameters
	be_parameter dstp(*this, inst.param(0), PTYPE_MR);
	be_parameter srcp(*this, inst.param(1), PTYPE_MF);
	const parameter &sizep = inst.param(2);
	assert(sizep.is_size());
	const parameter &roundp = inst.param(3);
	assert(roundp.is_rounding());

	// pick a target register for the general case
	Gp dstreg = (sizep.size() == SIZE_DWORD) ? dstp.select_register(eax) : dstp.select_register(rax);

	// set rounding mode if necessary
	if (roundp.rounding() != ROUND_DEFAULT && roundp.rounding() != ROUND_TRUNC)
	{
		a.stmxcsr(MABS(&m_near.ssemodesave));                                           // stmxcsr [ssemodesave]
		a.ldmxcsr(MABS(&m_near.ssecontrol[roundp.rounding()]));                         // ldmxcsr fpcontrol[mode]
	}

	// 32-bit form
	if (inst.size() == 4)
	{
		if (srcp.is_memory())
		{
			if (roundp.rounding() != ROUND_TRUNC)
				a.cvtss2si(dstreg, MABS(srcp.memory()));                                // cvtss2si dstreg,[srcp]
			else
				a.cvttss2si(dstreg, MABS(srcp.memory()));                               // cvttss2si dstreg,[srcp]
		}
		else if (srcp.is_float_register())
		{
			if (roundp.rounding() != ROUND_TRUNC)
				a.cvtss2si(dstreg, Xmm(srcp.freg()));                                   // cvtss2si dstreg,srcp
			else
				a.cvttss2si(dstreg, Xmm(srcp.freg()));                                  // cvttss2si dstreg,srcp
		}
	}

	// 64-bit form
	else if (inst.size() == 8)
	{
		if (srcp.is_memory())
		{
			if (roundp.rounding() != ROUND_TRUNC)
				a.cvtsd2si(dstreg, MABS(srcp.memory()));                                // cvtsd2si dstreg,[srcp]
			else
				a.cvttsd2si(dstreg, MABS(srcp.memory()));                               // cvttsd2si dstreg,[srcp]
		}
		else if (srcp.is_float_register())
		{
			if (roundp.rounding() != ROUND_TRUNC)
				a.cvtsd2si(dstreg, Xmm(srcp.freg()));                                   // cvtsd2si dstreg,srcp
			else
				a.cvttsd2si(dstreg, Xmm(srcp.freg()));                                  // cvttsd2si dstreg,srcp
		}
	}

	mov_param_reg(a, dstp, dstreg);                                                     // mov   dstp,dstreg

	// restore rounding mode
	if (roundp.rounding() != ROUND_DEFAULT && roundp.rounding() != ROUND_TRUNC)
		a.ldmxcsr(MABS(&m_near.ssemodesave));                                           // ldmxcsr [ssemodesave]
}


//-------------------------------------------------
//  op_ffrint - process a FFRINT opcode
//-------------------------------------------------

void drcbe_x64::op_ffrint(Assembler &a, const instruction &inst)
{
	// validate instruction
	assert(inst.size() == 4 || inst.size() == 8);
	assert_no_condition(inst);
	assert_no_flags(inst);

	// normalize parameters
	be_parameter dstp(*this, inst.param(0), PTYPE_MF);
	be_parameter srcp(*this, inst.param(1), PTYPE_MRI);
	const parameter &sizep = inst.param(2);
	assert(sizep.is_size());

	// pick a target register for the general case
	Xmm dstreg = dstp.select_register(xmm0);

	// 32-bit form
	if (inst.size() == 4)
	{
		// 32-bit integer source
		if (sizep.size() == SIZE_DWORD)
		{
			if (srcp.is_memory())
				a.cvtsi2ss(dstreg, MABS(srcp.memory(), 4));                             // cvtsi2ss dstreg,[srcp]
			else
			{
				Gp srcreg = srcp.select_register(eax);
				mov_reg_param(a, srcreg, srcp);                                         // mov      srcreg,srcp
				a.cvtsi2ss(dstreg, srcreg);                                             // cvtsi2ss dstreg,srcreg
			}
		}

		// 64-bit integer source
		else
		{
			if (srcp.is_memory())
				a.cvtsi2ss(dstreg, MABS(srcp.memory(), 8));                             // cvtsi2ss dstreg,[srcp]
			else
			{
				Gp srcreg = srcp.select_register(rax);
				mov_reg_param(a, srcreg, srcp);                                         // mov      srcreg,srcp
				a.cvtsi2ss(dstreg, srcreg);                                             // cvtsi2ss dstreg,srcreg
			}
		}
		movss_p32_r128(a, dstp, dstreg);                                                // movss    dstp,dstreg
	}

	// 64-bit form
	else if (inst.size() == 8)
	{
		// 32-bit integer source
		if (sizep.size() == SIZE_DWORD)
		{
			if (srcp.is_memory())
				a.cvtsi2sd(dstreg, MABS(srcp.memory(), 4));                             // cvtsi2sd dstreg,[srcp]
			else
			{
				Gp srcreg = srcp.select_register(eax);
				mov_reg_param(a, srcreg, srcp);                                         // mov      srcreg,srcp
				a.cvtsi2sd(dstreg, srcreg);                                             // cvtsi2sd dstreg,srcreg
			}
		}

		// 64-bit integer source
		else
		{
			if (srcp.is_memory())
				a.cvtsi2sd(dstreg, MABS(srcp.memory(), 8));                             // cvtsi2sd dstreg,[srcp]
			else
			{
				Gp srcreg = srcp.select_register(rax);
				mov_reg_param(a, srcreg, srcp);                                         // mov      srcreg,srcp
				a.cvtsi2sd(dstreg, srcreg);                                             // cvtsi2sd dstreg,srcreg
			}
		}
		movsd_p64_r128(a, dstp, dstreg);                                                // movsd    dstp,dstreg
	}
}


//-------------------------------------------------
//  op_ffrflt - process a FFRFLT opcode
//-------------------------------------------------

void drcbe_x64::op_ffrflt(Assembler &a, const instruction &inst)
{
	// validate instruction
	assert(inst.size() == 4 || inst.size() == 8);
	assert_no_condition(inst);
	assert_no_flags(inst);

	// normalize parameters
	be_parameter dstp(*this, inst.param(0), PTYPE_MF);
	be_parameter srcp(*this, inst.param(1), PTYPE_MF);
	const parameter &sizep = inst.param(2);
	assert(sizep.is_size());

	// pick a target register for the general case
	Xmm dstreg = dstp.select_register(xmm0);

	// single-to-double
	if (inst.size() == 8 && sizep.size() == SIZE_DWORD)
	{
		if (srcp.is_memory())
			a.cvtss2sd(dstreg, MABS(srcp.memory()));                                    // cvtss2sd dstreg,[srcp]
		else if (srcp.is_float_register())
			a.cvtss2sd(dstreg, Xmm(srcp.freg()));                                       // cvtss2sd dstreg,srcp
		movsd_p64_r128(a, dstp, dstreg);                                                // movsd    dstp,dstreg
	}

	// double-to-single
	else if (inst.size() == 4 && sizep.size() == SIZE_QWORD)
	{
		if (srcp.is_memory())
			a.cvtsd2ss(dstreg, MABS(srcp.memory()));                                    // cvtsd2ss dstreg,[srcp]
		else if (srcp.is_float_register())
			a.cvtsd2ss(dstreg, Xmm(srcp.freg()));                                       // cvtsd2ss dstreg,srcp
		movss_p32_r128(a, dstp, dstreg);                                                // movss    dstp,dstreg
	}
}


//-------------------------------------------------
//  op_frnds - process a FRNDS opcode
//-------------------------------------------------

void drcbe_x64::op_frnds(Assembler &a, const instruction &inst)
{
	// validate instruction
	assert(inst.size() == 8);
	assert_no_condition(inst);
	assert_no_flags(inst);

	// normalize parameters
	be_parameter dstp(*this, inst.param(0), PTYPE_MF);
	be_parameter srcp(*this, inst.param(1), PTYPE_MF);

	// pick a target register for the general case
	Xmm dstreg = dstp.select_register(xmm0);

	// 64-bit form
	if (srcp.is_memory())
		a.cvtsd2ss(dstreg, MABS(srcp.memory(), 8));                                     // cvtsd2ss dstreg,[srcp]
	else if (srcp.is_float_register())
		a.cvtsd2ss(dstreg, Xmm(srcp.freg()));                                           // cvtsd2ss dstreg,srcp
	a.cvtss2sd(dstreg, dstreg);                                                         // cvtss2sd dstreg,dstreg
	movsd_p64_r128(a, dstp, dstreg);                                                    // movsd    dstp,dstreg
}


//-------------------------------------------------
//  op_fadd - process a FADD opcode
//-------------------------------------------------

void drcbe_x64::op_fadd(Assembler &a, const instruction &inst)
{
	// validate instruction
	assert(inst.size() == 4 || inst.size() == 8);
	assert_no_condition(inst);
	assert_no_flags(inst);

	// normalize parameters
	be_parameter dstp(*this, inst.param(0), PTYPE_MF);
	be_parameter src1p(*this, inst.param(1), PTYPE_MF);
	be_parameter src2p(*this, inst.param(2), PTYPE_MF);
	normalize_commutative(src1p, src2p);

	// pick a target register for the general case
	Xmm dstreg = dstp.select_register(xmm0, src2p);

	// 32-bit form
	if (inst.size() == 4)
	{
		movss_r128_p32(a, dstreg, src1p);                                               // movss dstreg,src1p
		if (src2p.is_memory())
			a.addss(dstreg, MABS(src2p.memory()));                                      // addss dstreg,[src2p]
		else if (src2p.is_float_register())
			a.addss(dstreg, Xmm(src2p.freg()));                                         // addss dstreg,src2p
		movss_p32_r128(a, dstp, dstreg);                                                // movss dstp,dstreg
	}

	// 64-bit form
	else if (inst.size() == 8)
	{
		movsd_r128_p64(a, dstreg, src1p);                                               // movsd dstreg,src1p
		if (src2p.is_memory())
			a.addsd(dstreg, MABS(src2p.memory()));                                      // addsd dstreg,[src2p]
		else if (src2p.is_float_register())
			a.addsd(dstreg, Xmm(src2p.freg()));                                         // addsd dstreg,src2p
		movsd_p64_r128(a, dstp, dstreg);                                                // movsd dstp,dstreg
	}
}


//-------------------------------------------------
//  op_fsub - process a FSUB opcode
//-------------------------------------------------

void drcbe_x64::op_fsub(Assembler &a, const instruction &inst)
{
	// validate instruction
	assert(inst.size() == 4 || inst.size() == 8);
	assert_no_condition(inst);
	assert_no_flags(inst);

	// normalize parameters
	be_parameter dstp(*this, inst.param(0), PTYPE_MF);
	be_parameter src1p(*this, inst.param(1), PTYPE_MF);
	be_parameter src2p(*this, inst.param(2), PTYPE_MF);

	// pick a target register for the general case
	Xmm dstreg = dstp.select_register(xmm0, src2p);

	// 32-bit form
	if (inst.size() == 4)
	{
		movss_r128_p32(a, dstreg, src1p);                                               // movss dstreg,src1p
		if (src2p.is_memory())
			a.subss(dstreg, MABS(src2p.memory()));                                      // subss dstreg,[src2p]
		else if (src2p.is_float_register())
			a.subss(dstreg, Xmm(src2p.freg()));                                         // subss dstreg,src2p
		movss_p32_r128(a, dstp, dstreg);                                                // movss dstp,dstreg
	}

	// 64-bit form
	else if (inst.size() == 8)
	{
		movsd_r128_p64(a, dstreg, src1p);                                               // movsd dstreg,src1p
		if (src2p.is_memory())
			a.subsd(dstreg, MABS(src2p.memory()));                                      // subsd dstreg,[src2p]
		else if (src2p.is_float_register())
			a.subsd(dstreg, Xmm(src2p.freg()));                                         // subsd dstreg,src2p
		movsd_p64_r128(a, dstp, dstreg);                                                // movsd dstp,dstreg
	}
}


//-------------------------------------------------
//  op_fcmp - process a FCMP opcode
//-------------------------------------------------

void drcbe_x64::op_fcmp(Assembler &a, const instruction &inst)
{
	// validate instruction
	assert(inst.size() == 4 || inst.size() == 8);
	assert_no_condition(inst);
	assert_flags(inst, FLAG_C | FLAG_Z | FLAG_U);

	// normalize parameters
	be_parameter src1p(*this, inst.param(0), PTYPE_MF);
	be_parameter src2p(*this, inst.param(1), PTYPE_MF);

	// pick a target register for the general case
	Xmm src1reg = src1p.select_register(xmm0);

	// 32-bit form
	if (inst.size() == 4)
	{
		movss_r128_p32(a, src1reg, src1p);                                              // movss src1reg,src1p
		if (src2p.is_memory())
			a.comiss(src1reg, MABS(src2p.memory()));                                    // comiss src1reg,[src2p]
		else if (src2p.is_float_register())
			a.comiss(src1reg, Xmm(src2p.freg()));                                       // comiss src1reg,src2p
	}

	// 64-bit form
	else if (inst.size() == 8)
	{
		movsd_r128_p64(a, src1reg, src1p);                                              // movsd src1reg,src1p
		if (src2p.is_memory())
			a.comisd(src1reg, MABS(src2p.memory()));                                    // comisd src1reg,[src2p]
		else if (src2p.is_float_register())
			a.comisd(src1reg, Xmm(src2p.freg()));                                       // comisd src1reg,src2p
	}
}


//-------------------------------------------------
//  op_fmul - process a FMUL opcode
//-------------------------------------------------

void drcbe_x64::op_fmul(Assembler &a, const instruction &inst)
{
	// validate instruction
	assert(inst.size() == 4 || inst.size() == 8);
	assert_no_condition(inst);
	assert_no_flags(inst);

	// normalize parameters
	be_parameter dstp(*this, inst.param(0), PTYPE_MF);
	be_parameter src1p(*this, inst.param(1), PTYPE_MF);
	be_parameter src2p(*this, inst.param(2), PTYPE_MF);
	normalize_commutative(src1p, src2p);

	// pick a target register for the general case
	Xmm dstreg = dstp.select_register(xmm0, src2p);

	// 32-bit form
	if (inst.size() == 4)
	{
		movss_r128_p32(a, dstreg, src1p);                                               // movss dstreg,src1p
		if (src2p.is_memory())
			a.mulss(dstreg, MABS(src2p.memory()));                                      // mulss dstreg,[src2p]
		else if (src2p.is_float_register())
			a.mulss(dstreg, Xmm(src2p.freg()));                                         // mulss dstreg,src2p
		movss_p32_r128(a, dstp, dstreg);                                                // movss dstp,dstreg
	}

	// 64-bit form
	else if (inst.size() == 8)
	{
		movsd_r128_p64(a, dstreg, src1p);                                               // movsd dstreg,src1p
		if (src2p.is_memory())
			a.mulsd(dstreg, MABS(src2p.memory()));                                      // mulsd dstreg,[src2p]
		else if (src2p.is_float_register())
			a.mulsd(dstreg, Xmm(src2p.freg()));                                         // mulsd dstreg,src2p
		movsd_p64_r128(a, dstp, dstreg);                                                // movsd dstp,dstreg
	}
}


//-------------------------------------------------
//  op_fdiv - process a FDIV opcode
//-------------------------------------------------

void drcbe_x64::op_fdiv(Assembler &a, const instruction &inst)
{
	// validate instruction
	assert(inst.size() == 4 || inst.size() == 8);
	assert_no_condition(inst);
	assert_no_flags(inst);

	// normalize parameters
	be_parameter dstp(*this, inst.param(0), PTYPE_MF);
	be_parameter src1p(*this, inst.param(1), PTYPE_MF);
	be_parameter src2p(*this, inst.param(2), PTYPE_MF);

	// pick a target register for the general case
	Xmm dstreg = dstp.select_register(xmm0, src2p);

	// 32-bit form
	if (inst.size() == 4)
	{
		movss_r128_p32(a, dstreg, src1p);                                               // movss dstreg,src1p
		if (src2p.is_memory())
			a.divss(dstreg, MABS(src2p.memory()));                                      // divss dstreg,[src2p]
		else if (src2p.is_float_register())
			a.divss(dstreg, Xmm(src2p.freg()));                                         // divss dstreg,src2p
		movss_p32_r128(a, dstp, dstreg);                                                // movss dstp,dstreg
	}

	// 64-bit form
	else if (inst.size() == 8)
	{
		movsd_r128_p64(a, dstreg, src1p);                                               // movsd dstreg,src1p
		if (src2p.is_memory())
			a.divsd(dstreg, MABS(src2p.memory()));                                      // divsd dstreg,[src2p]
		else if (src2p.is_float_register())
			a.divsd(dstreg, Xmm(src2p.freg()));                                         // divsd dstreg,src2p
		movsd_p64_r128(a, dstp, dstreg);                                                // movsd dstp,dstreg
	}
}


//-------------------------------------------------
//  op_fneg - process a FNEG opcode
//-------------------------------------------------

void drcbe_x64::op_fneg(Assembler &a, const instruction &inst)
{
	// validate instruction
	assert(inst.size() == 4 || inst.size() == 8);
	assert_no_condition(inst);
	assert_no_flags(inst);

	// normalize parameters
	be_parameter dstp(*this, inst.param(0), PTYPE_MF);
	be_parameter srcp(*this, inst.param(1), PTYPE_MF);

	// pick a target register for the general case
	Xmm dstreg = dstp.select_register(xmm0, srcp);

	// 32-bit form
	if (inst.size() == 4)
	{
		a.xorps(dstreg, dstreg);                                                        // xorps dstreg,dstreg
		if (srcp.is_memory())
			a.subss(dstreg, MABS(srcp.memory()));                                       // subss dstreg,[srcp]
		else if (srcp.is_float_register())
			a.subss(dstreg, Xmm(srcp.freg()));                                          // subss dstreg,srcp
		movss_p32_r128(a, dstp, dstreg);                                                // movss dstp,dstreg
	}

	// 64-bit form
	else if (inst.size() == 8)
	{
		a.xorpd(dstreg, dstreg);                                                        // xorpd dstreg,dstreg
		if (srcp.is_memory())
			a.subsd(dstreg, MABS(srcp.memory()));                                       // subsd dstreg,[srcp]
		else if (srcp.is_float_register())
			a.subsd(dstreg, Xmm(srcp.freg()));                                          // subsd dstreg,srcp
		movsd_p64_r128(a, dstp, dstreg);                                                // movsd dstp,dstreg
	}
}


//-------------------------------------------------
//  op_fabs - process a FABS opcode
//-------------------------------------------------

void drcbe_x64::op_fabs(Assembler &a, const instruction &inst)
{
	// validate instruction
	assert(inst.size() == 4 || inst.size() == 8);
	assert_no_condition(inst);
	assert_no_flags(inst);

	// normalize parameters
	be_parameter dstp(*this, inst.param(0), PTYPE_MF);
	be_parameter srcp(*this, inst.param(1), PTYPE_MF);

	// pick a target register for the general case
	Xmm dstreg = dstp.select_register(xmm0, srcp);

	// 32-bit form
	if (inst.size() == 4)
	{
		movss_r128_p32(a, dstreg, srcp);                                                // movss dstreg,srcp
		a.andps(dstreg, MABS(m_absmask32));                                             // andps dstreg,[absmask32]
		movss_p32_r128(a, dstp, dstreg);                                                // movss dstp,dstreg
	}

	// 64-bit form
	else if (inst.size() == 8)
	{
		movsd_r128_p64(a, dstreg, srcp);                                                // movsd dstreg,srcp
		a.andpd(dstreg, MABS(m_absmask64));                                             // andpd dstreg,[absmask64]
		movsd_p64_r128(a, dstp, dstreg);                                                // movsd dstp,dstreg
	}
}


//-------------------------------------------------
//  op_fsqrt - process a FSQRT opcode
//-------------------------------------------------

void drcbe_x64::op_fsqrt(Assembler &a, const instruction &inst)
{
	// validate instruction
	assert(inst.size() == 4 || inst.size() == 8);
	assert_no_condition(inst);
	assert_no_flags(inst);

	// normalize parameters
	be_parameter dstp(*this, inst.param(0), PTYPE_MF);
	be_parameter srcp(*this, inst.param(1), PTYPE_MF);

	// pick a target register for the general case
	Xmm dstreg = dstp.select_register(xmm0);

	// 32-bit form
	if (inst.size() == 4)
	{
		if (srcp.is_memory())
			a.sqrtss(dstreg, MABS(srcp.memory()));                                      // sqrtss dstreg,[srcp]
		else if (srcp.is_float_register())
			a.sqrtss(dstreg, Xmm(srcp.freg()));                                         // sqrtss dstreg,srcp
		movss_p32_r128(a, dstp, dstreg);                                                // movss dstp,dstreg
	}

	// 64-bit form
	else if (inst.size() == 8)
	{
		if (srcp.is_memory())
			a.sqrtsd(dstreg, MABS(srcp.memory()));                                      // sqrtsd dstreg,[srcp]
		else if (srcp.is_float_register())
			a.sqrtsd(dstreg, Xmm(srcp.freg()));                                         // sqrtsd dstreg,srcp
		movsd_p64_r128(a, dstp, dstreg);                                                // movsd dstp,dstreg
	}
}


//-------------------------------------------------
//  op_frecip - process a FRECIP opcode
//-------------------------------------------------

void drcbe_x64::op_frecip(Assembler &a, const instruction &inst)
{
	// validate instruction
	assert(inst.size() == 4 || inst.size() == 8);
	assert_no_condition(inst);
	assert_no_flags(inst);

	// normalize parameters
	be_parameter dstp(*this, inst.param(0), PTYPE_MF);
	be_parameter srcp(*this, inst.param(1), PTYPE_MF);

	// pick a target register for the general case
	Xmm dstreg = dstp.select_register(xmm0);

	// 32-bit form
	if (inst.size() == 4)
	{
		if (USE_RCPSS_FOR_SINGLES)
		{
			if (srcp.is_memory())
				a.rcpss(dstreg, MABS(srcp.memory()));                                   // rcpss dstreg,[srcp]
			else if (srcp.is_float_register())
				a.rcpss(dstreg, Xmm(srcp.freg()));                                      // rcpss dstreg,srcp
			movss_p32_r128(a, dstp, dstreg);                                            // movss dstp,dstreg
		}
		else
		{
			a.movss(xmm1, MABS(&m_near.single1));                                       // movss xmm1,1.0
			if (srcp.is_memory())
				a.divss(xmm1, MABS(srcp.memory()));                                     // divss xmm1,[srcp]
			else if (srcp.is_float_register())
				a.divss(xmm1, Xmm(srcp.freg()));                                        // divss xmm1,srcp
			movss_p32_r128(a, dstp, xmm1);                                              // movss dstp,xmm1
		}
	}

	// 64-bit form
	else if (inst.size() == 8)
	{
		if (USE_RCPSS_FOR_DOUBLES)
		{
			if (srcp.is_memory())
				a.cvtsd2ss(dstreg, MABS(srcp.memory()));                                // cvtsd2ss dstreg,[srcp]
			else if (srcp.is_float_register())
				a.cvtsd2ss(dstreg, Xmm(srcp.freg()));                                   // cvtsd2ss dstreg,srcp
			a.rcpss(dstreg, dstreg);                                                    // rcpss dstreg,dstreg
			a.cvtss2sd(dstreg, dstreg);                                                 // cvtss2sd dstreg,dstreg
			movsd_p64_r128(a, dstp, dstreg);                                            // movsd dstp,dstreg
		}
		else
		{
			a.movsd(xmm1, MABS(&m_near.double1));                                       // movsd xmm1,1.0
			if (srcp.is_memory())
				a.divsd(xmm1, MABS(srcp.memory()));                                     // divsd xmm1,[srcp]
			else if (srcp.is_float_register())
				a.divsd(xmm1, Xmm(srcp.freg()));                                        // divsd xmm1,srcp
			movsd_p64_r128(a, dstp, xmm1);                                              // movsd dstp,xmm1
		}
	}
}


//-------------------------------------------------
//  op_frsqrt - process a FRSQRT opcode
//-------------------------------------------------

void drcbe_x64::op_frsqrt(Assembler &a, const instruction &inst)
{
	// validate instruction
	assert(inst.size() == 4 || inst.size() == 8);
	assert_no_condition(inst);
	assert_no_flags(inst);

	// normalize parameters
	be_parameter dstp(*this, inst.param(0), PTYPE_MF);
	be_parameter srcp(*this, inst.param(1), PTYPE_MF);

	// pick a target register for the general case
	Xmm dstreg = dstp.select_register(xmm0);

	// 32-bit form
	if (inst.size() == 4)
	{
		if (USE_RSQRTSS_FOR_SINGLES)
		{
			if (srcp.is_memory())
				a.rsqrtss(dstreg, MABS(srcp.memory()));                                 // rsqrtss dstreg,[srcp]
			else if (srcp.is_float_register())
				a.rsqrtss(dstreg, Xmm(srcp.freg()));                                    // rsqrtss dstreg,srcp
		}
		else
		{
			if (srcp.is_memory())
				a.sqrtss(xmm1, MABS(srcp.memory()));                                    // sqrtss xmm1,[srcp]
			else if (srcp.is_float_register())
				a.sqrtss(xmm1, Xmm(srcp.freg()));                                       // sqrtss xmm1,srcp
			a.movss(dstreg, MABS(&m_near.single1));                                     // movss dstreg,1.0
			a.divss(dstreg, xmm1);                                                      // divss dstreg,xmm1
		}
		movss_p32_r128(a, dstp, dstreg);                                                // movss dstp,dstreg
	}

	// 64-bit form
	else if (inst.size() == 8)
	{
		if (USE_RSQRTSS_FOR_DOUBLES)
		{
			if (srcp.is_memory())
				a.cvtsd2ss(dstreg, MABS(srcp.memory()));                                // cvtsd2ss dstreg,[srcp]
			else if (srcp.is_float_register())
				a.cvtsd2ss(dstreg, Xmm(srcp.freg()));                                   // cvtsd2ss dstreg,srcp
			a.rsqrtss(dstreg, dstreg);                                                  // rsqrtss dstreg,dstreg
			a.cvtss2sd(dstreg, dstreg);                                                 // cvtss2sd dstreg,dstreg
		}
		else
		{
			if (srcp.is_memory())
				a.sqrtsd(xmm1, MABS(srcp.memory()));                                    // sqrtsd xmm1,[srcp]
			else if (srcp.is_float_register())
				a.sqrtsd(xmm1, Xmm(srcp.freg()));                                       // sqrtsd xmm1,srcp
			a.movsd(dstreg, MABS(&m_near.double1));                                     // movsd dstreg,1.0
			a.divsd(dstreg, xmm1);                                                      // divsd dstreg,xmm1
		}
		movsd_p64_r128(a, dstp, dstreg);                                                // movsd dstp,dstreg
	}
}


//-------------------------------------------------
//  op_fcopyi - process a FCOPYI opcode
//-------------------------------------------------

void drcbe_x64::op_fcopyi(Assembler &a, const instruction &inst)
{
	// validate instruction
	assert(inst.size() == 4 || inst.size() == 8);
	assert_no_condition(inst);
	assert_no_flags(inst);

	// normalize parameters
	be_parameter dstp(*this, inst.param(0), PTYPE_MF);
	be_parameter srcp(*this, inst.param(1), PTYPE_MR);

	// pick a target register for the general case
	Xmm dstreg = dstp.select_register(xmm0);

	// 32-bit form
	if (inst.size() == 4)
	{
		if (srcp.is_memory())
		{
			a.movd(dstreg, MABS(srcp.memory()));                                        // movd     dstreg,[srcp]
			movss_p32_r128(a, dstp, dstreg);                                            // movss    dstp,dstreg
		}
		else
		{
			if (dstp.is_memory())
			{
				mov_param_reg(a, dstp, Gpd(srcp.ireg()));                               // mov      dstp,srcp
			}
			else
			{
				a.movd(dstreg, Gpd(srcp.ireg()));                                       // movd     dstreg,srcp
				movss_p32_r128(a, dstp, dstreg);                                        // movss    dstp,dstreg
			}
		}

	}

	// 64-bit form
	else if (inst.size() == 8)
	{
		if (srcp.is_memory())
		{
			a.movq(dstreg, MABS(srcp.memory()));                                        // movq     dstreg,[srcp]
			movsd_p64_r128(a, dstp, dstreg);                                            // movsd    dstp,dstreg
		}
		else
		{
			if (dstp.is_memory())
			{
				mov_param_reg(a, dstp, Gpq(srcp.ireg()));                               // mov      dstp,srcp
			}
			else
			{
				a.movq(dstreg, Gpq(srcp.ireg()));                                       // movq     dstreg,srcp
				movsd_p64_r128(a, dstp, dstreg);                                        // movsd    dstp,dstreg
			}
		}

	}
}


//-------------------------------------------------
//  op_icopyf - process a ICOPYF opcode
//-------------------------------------------------

void drcbe_x64::op_icopyf(Assembler &a, const instruction &inst)
{
	// validate instruction
	assert(inst.size() == 4 || inst.size() == 8);
	assert_no_condition(inst);
	assert_no_flags(inst);

	// normalize parameters
	be_parameter dstp(*this, inst.param(0), PTYPE_MR);
	be_parameter srcp(*this, inst.param(1), PTYPE_MF);

	// 32-bit form
	if (inst.size() == 4)
	{
		if (srcp.is_memory())
		{
			Gp dstreg = dstp.select_register(eax);
			a.mov(dstreg, MABS(srcp.memory()));                                         // mov      dstreg,[srcp]
			mov_param_reg(a, dstp, dstreg);                                             // mov      dstp,dstreg
		}
		else
		{
			if (dstp.is_memory())
			{
				a.movd(MABS(dstp.memory()), Xmm(srcp.freg()));                          // movd     dstp,srcp
			}
			else
			{
				a.movd(Gpd(dstp.ireg()), Xmm(srcp.freg()));                             // movd     dstp,srcp
			}
		}
	}

	// 64-bit form
	else if (inst.size() == 8)
	{
		if (srcp.is_memory())
		{
			Gp dstreg = dstp.select_register(rax);
			a.mov(dstreg, MABS(srcp.memory()));                                         // mov      dstreg,[srcp]
			mov_param_reg(a, dstp, dstreg);                                             // mov      dstp,dstreg
		}
		else
		{
			if (dstp.is_memory())
			{
				a.movq(MABS(dstp.memory()), Xmm(srcp.freg()));                          // movq     dstp,srcp
			}
			else
			{
				a.movq(Gpq(dstp.ireg()), Xmm(srcp.freg()));                             // movq     dstp,srcp
			}
		}
	}
}

} // namespace drc