freedreno: fix staging resource size for arrays - mesa.git - Unnamed repository; edit this file 'description' to name the repository.

path: root/src/gallium/drivers/freedreno

diff options

author	Rob Clark <[email protected]>	2019-01-02 13:24:31 -0500
committer	Rob Clark <[email protected]>	2019-01-03 08:11:40 -0500
commit	e869481ef30d0c3576d402816fa341ea1e21456a (patch)
tree	bf89c8f47233c491975d02ba00b5735e172cae18 /src/gallium/drivers/freedreno
parent	67a7f6f24433beb5795bea5325700f50d845ed64 (diff)

freedreno: fix staging resource size for arrays

A 2d-array texture (for example), should get the # of array elements from box->depth, rather than depth0 which is minified. Fixes dEQP-GLES3.functional.shaders.texture_functions.texture.sampler2darray_bias_float_fragment with tiled textures. Reported-by: Kristian H. Kristensen <[email protected]> Signed-off-by: Rob Clark <[email protected]>

Diffstat (limited to 'src/gallium/drivers/freedreno')

-rw-r--r--

src/gallium/drivers/freedreno/freedreno_resource.c

1 files changed, 10 insertions, 2 deletions

181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151 1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199 1200 1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 1230 1231 1232 1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 1251 1252 1253 1254 1255 1256 1257 1258 1259 1260 1261 1262 1263 1264 1265 1266 1267 1268 1269 1270 1271 1272 1273 1274 1275 1276 1277 1278 1279 1280 1281 1282 1283 1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 1299 1300 1301 1302 1303 1304 1305 1306 1307 1308 1309 1310 1311 1312 1313 1314 1315 1316 1317 1318 1319 1320 1321 1322 1323 1324 1325 1326 1327 1328 1329 1330 1331 1332 1333 1334 1335 1336 1337 1338 1339 1340 1341 1342 1343 1344 1345 1346 1347 1348 1349 1350 1351 1352 1353 1354 1355 1356 1357 1358 1359 1360 1361 1362 1363 1364 1365 1366 1367 1368 1369 1370 1371 1372 1373 1374 1375 1376 1377 1378 1379 1380 1381 1382 1383 1384 1385 1386 1387 1388 1389 1390 1391 1392 1393 1394 1395 1396 1397 1398 1399 1400 1401 1402 1403 1404 1405 1406 1407 1408 1409 1410 1411 1412 1413 1414 1415 1416 1417 1418 1419 1420 1421 1422 1423 1424 1425 1426 1427 1428 1429 1430 1431 1432 1433 1434 1435 1436 1437 1438 1439 1440 1441 1442 1443 1444 1445 1446 1447 1448 1449 1450 1451 1452 1453 1454 1455 1456 1457 1458 1459 1460 1461 1462 1463 1464 1465 1466 1467 1468 1469 1470 1471 1472 1473 1474 1475 1476 1477 1478 1479 1480 1481 1482 1483 1484 1485 1486 1487 1488 1489 1490 1491 1492 1493 1494 1495 1496 1497 1498 1499 1500 1501 1502 1503 1504 1505 1506 1507 1508 1509 1510 1511 1512 1513 1514 1515 1516 1517 1518 1519 1520 1521 1522 1523 1524 1525 1526 1527 1528 1529 1530 1531 1532 1533 1534 1535 1536 1537 1538 1539 1540 1541 1542 1543 1544 1545 1546 1547 1548 1549 1550 1551 1552 1553 1554 1555 1556 1557 1558 1559 1560 1561 1562 1563 1564 1565 1566 1567 1568 1569 1570 1571 1572 1573 1574 1575 1576 1577 1578 1579 1580 1581 1582 1583 1584 1585 1586 1587 1588 1589 1590 1591 1592 1593 1594 1595 1596 1597 1598 1599 1600 1601 1602 1603 1604 1605 1606 1607 1608 1609 1610 1611 1612 1613 1614 1615 1616 1617 1618 1619 1620 1621 1622 1623 1624 1625 1626 1627 1628 1629 1630 1631 1632 1633 1634 1635 1636 1637 1638 1639 1640 1641 1642 1643 1644 1645 1646 1647 1648 1649 1650 1651 1652 1653 1654 1655 1656 1657 1658 1659 1660 1661 1662 1663 1664 1665 1666 1667 1668 1669 1670 1671 1672 1673 1674 1675 1676 1677 1678 1679 1680 1681 1682 1683 1684 1685 1686 1687 1688 1689 1690 1691 1692 1693 1694 1695 1696 1697 1698 1699 1700 1701 1702 1703 1704 1705 1706 1707 1708 1709 1710 1711 1712 1713 1714 1715 1716 1717 1718 1719 1720 1721 1722 1723 1724 1725 1726 1727 1728 1729 1730 1731 1732 1733 1734 1735 1736 1737 1738 1739 1740 1741 1742 1743 1744 1745 1746 1747 1748 1749 1750 1751 1752 1753 1754 1755 1756 1757 1758 1759 1760 1761 1762 1763 1764 1765 1766 1767 1768 1769 1770 1771 1772 1773 1774 1775 1776 1777 1778 1779 1780 1781 1782 1783 1784 1785 1786 1787 1788 1789 1790 1791 1792 1793 1794 1795 1796 1797 1798 1799 1800 1801 1802 1803 1804 1805 1806 1807 1808 1809 1810 1811 1812 1813 1814 1815 1816 1817 1818 1819 1820 1821 1822 1823 1824 1825 1826 1827 1828 1829 1830 1831 1832 1833 1834 1835 1836 1837 1838 1839 1840 1841 1842 1843 1844 1845 1846 1847 1848 1849 1850 1851 1852 1853 1854 1855 1856 1857 1858 1859 1860 1861 1862 1863 1864 1865 1866 1867 1868 1869 1870 1871 1872 1873 1874 1875 1876 1877 1878 1879 1880 1881 1882 1883 1884 1885 1886 1887 1888 1889 1890 1891 1892 1893 1894 1895 1896 1897 1898 1899 1900 1901 1902 1903 1904 1905 1906 1907 1908 1909 1910 1911 1912 1913 1914 1915 1916 1917 1918 1919 1920 1921 1922 1923 1924 1925 1926 1927 1928 1929 1930 1931 1932 1933 1934 1935 1936 1937 1938 1939 1940 1941 1942 1943 1944 1945 1946 1947 1948 1949 1950 1951 1952 1953 1954 1955 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 2034 2035 2036 2037 2038 2039 2040 2041 2042 2043 2044 2045 2046 2047 2048 2049 2050 2051 2052 2053 2054 2055 2056 2057 2058 2059 2060 2061 2062 2063 2064 2065 2066 2067 2068 2069 2070 2071 2072 2073 2074 2075 2076 2077 2078 2079 2080 2081 2082 2083 2084 2085 2086 2087 2088 2089 2090 2091 2092 2093 2094 2095 2096 2097 2098 2099 2100 2101 2102 2103 2104 2105 2106 2107 2108 2109 2110 2111 2112 2113 2114 2115 2116 2117 2118 2119 2120 2121 2122 2123 2124 2125 2126 2127 2128 2129 2130 2131 2132 2133 2134 2135 2136 2137 2138 2139 2140 2141 2142 2143 2144 2145 2146 2147 2148 2149 2150 2151 2152 2153 2154 2155 2156 2157 2158 2159 2160 2161 2162 2163 2164 2165 2166 2167 2168 2169 2170 2171 2172 2173 2174 2175 2176 2177 2178 2179 2180 2181 2182 2183 2184 2185 2186 2187 2188 2189 2190 2191 2192 2193 2194 2195 2196 2197 2198 2199 2200 2201 2202 2203 2204 2205 2206 2207 2208 2209 2210 2211 2212 2213 2214 2215 2216 2217 2218 2219 2220 2221 2222 2223 2224 2225 2226 2227 2228 2229 2230 2231 2232 2233 2234 2235 2236 2237 2238 2239 2240 2241 2242 2243 2244 2245 2246 2247 2248 2249 2250 2251 2252 2253 2254 2255 2256 2257 2258 2259 2260 2261 2262 2263 2264 2265 2266 2267 2268 2269 2270 2271 2272 2273 2274 2275 2276 2277 2278 2279 2280 2281 2282 2283 2284 2285 2286 2287 2288 2289 2290 2291 2292 2293 2294 2295 2296 2297 2298 2299 2300 2301 2302 2303 2304 2305 2306 2307 2308 2309 2310 2311 2312 2313 2314 2315 2316 2317 2318 2319 2320 2321 2322 2323 2324 2325 2326 2327 2328 2329 2330 2331 2332 2333 2334 2335 2336 2337 2338 2339 2340 2341 2342 2343 2344 2345 2346 2347 2348 2349 2350 2351 2352 2353 2354 2355 2356 2357 2358 2359 2360 2361 2362 2363 2364 2365 2366 2367 2368 2369 2370 2371 2372 2373 2374 2375 2376 2377 2378 2379 2380 2381 2382 2383 2384 2385 2386 2387 2388 2389 2390 2391 2392 2393 2394 2395 2396 2397 2398 2399 2400 2401 2402 2403 2404 2405 2406 2407 2408 2409 2410 2411 2412 2413 2414 2415 2416 2417 2418 2419 2420 2421 2422 2423 2424 2425 2426 2427 2428 2429 2430 2431 2432 2433 2434 2435 2436 2437 2438 2439 2440 2441 2442 2443 2444 2445 2446 2447 2448 2449 2450 2451 2452 2453 2454 2455 2456 2457 2458 2459 2460 2461 2462 2463 2464 2465 2466 2467 2468 2469 2470 2471 2472 2473 2474 2475 2476 2477 2478 2479 2480 2481 2482 2483 2484 2485 2486 2487 2488 2489 2490 2491 2492 2493 2494 2495 2496 2497 2498 2499 2500 2501 2502 2503 2504 2505 2506 2507 2508 2509 2510 2511 2512 2513 2514 2515 2516 2517 2518 2519 2520 2521 2522 2523 2524 2525 2526 2527 2528 2529 2530 2531 2532 2533 2534 2535 2536 2537 2538 2539 2540 2541 2542 2543 2544 2545 2546 2547 2548 2549 2550 2551 2552 2553 2554 2555 2556 2557 2558 2559 2560 2561 2562 2563 2564 2565 2566 2567 2568 2569 2570 2571 2572 2573 2574 2575 2576 2577 2578 2579 2580 2581 2582 2583 2584 2585 2586 2587 2588 2589 2590 2591 2592 2593 2594 2595 2596 2597 2598 2599 2600 2601 2602 2603 2604 2605 2606 2607 2608 2609 2610 2611 2612 2613 2614 2615 2616 2617 2618 2619 2620 2621 2622 2623 2624 2625 2626 2627 2628 2629 2630 2631 2632 2633 2634 2635 2636 2637 2638 2639 2640 2641 2642 2643 2644 2645 2646 2647 2648 2649 2650 2651 2652 2653 2654 2655 2656 2657 2658 2659 2660 2661 2662 2663 2664 2665 2666 2667 2668 2669 2670 2671 2672 2673 2674 2675 2676 2677 2678 2679 2680 2681 2682 2683 2684 2685 2686 2687 2688 2689 2690 2691 2692 2693 2694 2695 2696 2697 2698 2699 2700 2701 2702 2703 2704 2705 2706 2707 2708 2709 2710 2711 2712 2713 2714 2715 2716 2717 2718 2719 2720 2721 2722 2723 2724 2725 2726 2727 2728 2729 2730 2731 2732 2733 2734 2735 2736 2737 2738 2739 2740 2741 2742 2743 2744 2745 2746 2747 2748 2749 2750 2751 2752 2753 2754 2755 2756 2757 2758 2759 2760 2761 2762 2763 2764 2765 2766 2767 2768 2769 2770 2771 2772 2773 2774 2775 2776 2777 2778 2779 2780 2781 2782 2783 2784 2785 2786 2787 2788 2789 2790 2791 2792 2793 2794 2795 2796 2797 2798 2799 2800 2801 2802 2803 2804 2805 2806 2807 2808 2809 2810 2811 2812 2813 2814 2815 2816 2817 2818 2819 2820 2821 2822 2823 2824 2825 2826 2827 2828 2829 2830 2831 2832 2833 2834 2835 2836 2837 2838 2839 2840 2841 2842 2843 2844 2845 2846 2847 2848 2849 2850 2851 2852 2853 2854 2855 2856 2857 2858 2859 2860 2861 2862 2863 2864 2865 2866 2867 2868 2869 2870 2871 2872 2873 2874 2875 2876 2877 2878 2879 2880 2881 2882 2883 2884 2885 2886 2887 2888 2889 2890 2891 2892 2893 2894 2895 2896 2897 2898 2899 2900 2901 2902 2903 2904 2905 2906 2907 2908 2909 2910 2911 2912 2913 2914 2915 2916 2917 2918 2919 2920 2921 2922 2923 2924 2925 2926 2927 2928 2929 2930 2931 2932 2933 2934 2935 2936 2937 2938 2939 2940 2941 2942 2943 2944 2945 2946 2947 2948 2949 2950 2951 2952 2953 2954 2955 2956 2957 2958 2959 2960 2961 2962 2963 2964 2965 2966 2967 2968 2969 2970 2971 2972 2973 2974 2975 2976 2977 2978 2979 2980 2981 2982 2983 2984 2985 2986 2987 2988 2989 2990 2991 2992 2993 2994 2995 2996 2997 2998 2999 3000 3001 3002 3003 3004 3005 3006 3007 3008 3009 3010 3011 3012 3013 3014 3015 3016 3017 3018 3019 3020 3021 3022 3023 3024 3025 3026 3027 3028 3029 3030 3031 3032 3033 3034 3035 3036 3037 3038 3039 3040 3041 3042 3043 3044 3045 3046 3047 3048 3049 3050 3051 3052 3053 3054 3055 3056 3057 3058 3059 3060 3061 3062 3063 3064 3065 3066 3067 3068 3069 3070 3071 3072 3073 3074 3075 3076 3077 3078 3079 3080 3081 3082 3083 3084 3085 3086 3087 3088 3089 3090 3091 3092 3093 3094 3095 3096 3097 3098 3099 3100 3101 3102 3103 3104 3105 3106 3107 3108 3109 3110 3111 3112 3113 3114 3115 3116 3117 3118

/*
 * Copyright © 2015 Intel Corporation
 *
 * Permission is hereby granted, free of charge, to any person obtaining a
 * copy of this software and associated documentation files (the "Software"),
 * to deal in the Software without restriction, including without limitation
 * the rights to use, copy, modify, merge, publish, distribute, sublicense,
 * and/or sell copies of the Software, and to permit persons to whom the
 * Software is furnished to do so, subject to the following conditions:
 *
 * The above copyright notice and this permission notice (including the next
 * paragraph) shall be included in all copies or substantial portions of the
 * Software.
 *
 * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
 * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
 * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
 * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
 * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
 * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
 * IN THE SOFTWARE.
 */

#ifndef ANV_PRIVATE_H
#define ANV_PRIVATE_H

#include <stdlib.h>
#include <stdio.h>
#include <stdbool.h>
#include <pthread.h>
#include <assert.h>
#include <stdint.h>
#include <i915_drm.h>

#ifdef HAVE_VALGRIND
#include <valgrind.h>
#include <memcheck.h>
#define VG(x) x
#define __gen_validate_value(x) VALGRIND_CHECK_MEM_IS_DEFINED(&(x), sizeof(x))
#else
#define VG(x)
#endif

#include "common/gen_clflush.h"
#include "dev/gen_device_info.h"
#include "blorp/blorp.h"
#include "compiler/brw_compiler.h"
#include "util/macros.h"
#include "util/list.h"
#include "util/u_atomic.h"
#include "util/u_vector.h"
#include "vk_alloc.h"
#include "vk_debug_report.h"

/* Pre-declarations needed for WSI entrypoints */
struct wl_surface;
struct wl_display;
typedef struct xcb_connection_t xcb_connection_t;
typedef uint32_t xcb_visualid_t;
typedef uint32_t xcb_window_t;

struct anv_buffer;
struct anv_buffer_view;
struct anv_image_view;
struct anv_instance;

struct gen_l3_config;

#include <vulkan/vulkan.h>
#include <vulkan/vulkan_intel.h>
#include <vulkan/vk_icd.h>
#include <vulkan/vk_android_native_buffer.h>

#include "anv_entrypoints.h"
#include "anv_extensions.h"
#include "isl/isl.h"

#include "common/gen_debug.h"
#include "common/intel_log.h"
#include "wsi_common.h"

/* Allowing different clear colors requires us to perform a depth resolve at
 * the end of certain render passes. This is because while slow clears store
 * the clear color in the HiZ buffer, fast clears (without a resolve) don't.
 * See the PRMs for examples describing when additional resolves would be
 * necessary. To enable fast clears without requiring extra resolves, we set
 * the clear value to a globally-defined one. We could allow different values
 * if the user doesn't expect coherent data during or after a render passes
 * (VK_ATTACHMENT_STORE_OP_DONT_CARE), but such users (aside from the CTS)
 * don't seem to exist yet. In almost all Vulkan applications tested thus far,
 * 1.0f seems to be the only value used. The only application that doesn't set
 * this value does so through the usage of an seemingly uninitialized clear
 * value.
 */
#define ANV_HZ_FC_VAL 1.0f

#define MAX_VBS         28
#define MAX_SETS         8
#define MAX_RTS          8
#define MAX_VIEWPORTS   16
#define MAX_SCISSORS    16
#define MAX_PUSH_CONSTANTS_SIZE 128
#define MAX_DYNAMIC_BUFFERS 16
#define MAX_IMAGES 8
#define MAX_PUSH_DESCRIPTORS 32 /* Minimum requirement */

/* The kernel relocation API has a limitation of a 32-bit delta value
 * applied to the address before it is written which, in spite of it being
 * unsigned, is treated as signed .  Because of the way that this maps to
 * the Vulkan API, we cannot handle an offset into a buffer that does not
 * fit into a signed 32 bits.  The only mechanism we have for dealing with
 * this at the moment is to limit all VkDeviceMemory objects to a maximum
 * of 2GB each.  The Vulkan spec allows us to do this:
 *
 *    "Some platforms may have a limit on the maximum size of a single
 *    allocation. For example, certain systems may fail to create
 *    allocations with a size greater than or equal to 4GB. Such a limit is
 *    implementation-dependent, and if such a failure occurs then the error
 *    VK_ERROR_OUT_OF_DEVICE_MEMORY should be returned."
 *
 * We don't use vk_error here because it's not an error so much as an
 * indication to the application that the allocation is too large.
 */
#define MAX_MEMORY_ALLOCATION_SIZE (1ull << 31)

#define ANV_SVGS_VB_INDEX    MAX_VBS
#define ANV_DRAWID_VB_INDEX (MAX_VBS + 1)

#define anv_printflike(a, b) __attribute__((__format__(__printf__, a, b)))

static inline uint32_t
align_down_npot_u32(uint32_t v, uint32_t a)
{
   return v - (v % a);
}

static inline uint32_t
align_u32(uint32_t v, uint32_t a)
{
   assert(a != 0 && a == (a & -a));
   return (v + a - 1) & ~(a - 1);
}

static inline uint64_t
align_u64(uint64_t v, uint64_t a)
{
   assert(a != 0 && a == (a & -a));
   return (v + a - 1) & ~(a - 1);
}

static inline int32_t
align_i32(int32_t v, int32_t a)
{
   assert(a != 0 && a == (a & -a));
   return (v + a - 1) & ~(a - 1);
}

/** Alignment must be a power of 2. */
static inline bool
anv_is_aligned(uintmax_t n, uintmax_t a)
{
   assert(a == (a & -a));
   return (n & (a - 1)) == 0;
}

static inline uint32_t
anv_minify(uint32_t n, uint32_t levels)
{
   if (unlikely(n == 0))
      return 0;
   else
      return MAX2(n >> levels, 1);
}

static inline float
anv_clamp_f(float f, float min, float max)
{
   assert(min < max);

   if (f > max)
      return max;
   else if (f < min)
      return min;
   else
      return f;
}

static inline bool
anv_clear_mask(uint32_t *inout_mask, uint32_t clear_mask)
{
   if (*inout_mask & clear_mask) {
      *inout_mask &= ~clear_mask;
      return true;
   } else {
      return false;
   }
}

static inline union isl_color_value
vk_to_isl_color(VkClearColorValue color)
{
   return (union isl_color_value) {
      .u32 = {
         color.uint32[0],
         color.uint32[1],
         color.uint32[2],
         color.uint32[3],
      },
   };
}

#define for_each_bit(b, dword)                          \
   for (uint32_t __dword = (dword);                     \
        (b) = __builtin_ffs(__dword) - 1, __dword;      \
        __dword &= ~(1 << (b)))

#define typed_memcpy(dest, src, count) ({ \
   STATIC_ASSERT(sizeof(*src) == sizeof(*dest)); \
   memcpy((dest), (src), (count) * sizeof(*(src))); \
})

/* Mapping from anv object to VkDebugReportObjectTypeEXT. New types need
 * to be added here in order to utilize mapping in debug/error/perf macros.
 */
#define REPORT_OBJECT_TYPE(o)                                                      \
   __builtin_choose_expr (                                                         \
   __builtin_types_compatible_p (__typeof (o), struct anv_instance*),              \
   VK_DEBUG_REPORT_OBJECT_TYPE_INSTANCE_EXT,                                       \
   __builtin_choose_expr (                                                         \
   __builtin_types_compatible_p (__typeof (o), struct anv_physical_device*),       \
   VK_DEBUG_REPORT_OBJECT_TYPE_PHYSICAL_DEVICE_EXT,                                \
   __builtin_choose_expr (                                                         \
   __builtin_types_compatible_p (__typeof (o), struct anv_device*),                \
   VK_DEBUG_REPORT_OBJECT_TYPE_DEVICE_EXT,                                         \
   __builtin_choose_expr (                                                         \
   __builtin_types_compatible_p (__typeof (o), const struct anv_device*),          \
   VK_DEBUG_REPORT_OBJECT_TYPE_DEVICE_EXT,                                         \
   __builtin_choose_expr (                                                         \
   __builtin_types_compatible_p (__typeof (o), struct anv_queue*),                 \
   VK_DEBUG_REPORT_OBJECT_TYPE_QUEUE_EXT,                                          \
   __builtin_choose_expr (                                                         \
   __builtin_types_compatible_p (__typeof (o), struct anv_semaphore*),             \
   VK_DEBUG_REPORT_OBJECT_TYPE_SEMAPHORE_EXT,                                      \
   __builtin_choose_expr (                                                         \
   __builtin_types_compatible_p (__typeof (o), struct anv_cmd_buffer*),            \
   VK_DEBUG_REPORT_OBJECT_TYPE_COMMAND_BUFFER_EXT,                                 \
   __builtin_choose_expr (                                                         \
   __builtin_types_compatible_p (__typeof (o), struct anv_fence*),                 \
   VK_DEBUG_REPORT_OBJECT_TYPE_FENCE_EXT,                                          \
   __builtin_choose_expr (                                                         \
   __builtin_types_compatible_p (__typeof (o), struct anv_device_memory*),         \
   VK_DEBUG_REPORT_OBJECT_TYPE_DEVICE_MEMORY_EXT,                                  \
   __builtin_choose_expr (                                                         \
   __builtin_types_compatible_p (__typeof (o), struct anv_buffer*),                \
   VK_DEBUG_REPORT_OBJECT_TYPE_BUFFER_EXT,                                         \
   __builtin_choose_expr (                                                         \
   __builtin_types_compatible_p (__typeof (o), struct anv_image*),                 \
   VK_DEBUG_REPORT_OBJECT_TYPE_IMAGE_EXT,                                          \
   __builtin_choose_expr (                                                         \
   __builtin_types_compatible_p (__typeof (o), const struct anv_image*),           \
   VK_DEBUG_REPORT_OBJECT_TYPE_IMAGE_EXT,                                          \
   __builtin_choose_expr (                                                         \
   __builtin_types_compatible_p (__typeof (o), struct anv_event*),                 \
   VK_DEBUG_REPORT_OBJECT_TYPE_EVENT_EXT,                                          \
   __builtin_choose_expr (                                                         \
   __builtin_types_compatible_p (__typeof (o), struct anv_query_pool*),            \
   VK_DEBUG_REPORT_OBJECT_TYPE_QUERY_POOL_EXT,                                     \
   __builtin_choose_expr (                                                         \
   __builtin_types_compatible_p (__typeof (o), struct anv_buffer_view*),           \
   VK_DEBUG_REPORT_OBJECT_TYPE_BUFFER_VIEW_EXT,                                    \
   __builtin_choose_expr (                                                         \
   __builtin_types_compatible_p (__typeof (o), struct anv_image_view*),            \
   VK_DEBUG_REPORT_OBJECT_TYPE_IMAGE_VIEW_EXT,                                     \
   __builtin_choose_expr (                                                         \
   __builtin_types_compatible_p (__typeof (o), struct anv_shader_module*),         \
   VK_DEBUG_REPORT_OBJECT_TYPE_SHADER_MODULE_EXT,                                  \
   __builtin_choose_expr (                                                         \
   __builtin_types_compatible_p (__typeof (o), struct anv_pipeline_cache*),        \
   VK_DEBUG_REPORT_OBJECT_TYPE_PIPELINE_CACHE_EXT,                                 \
   __builtin_choose_expr (                                                         \
   __builtin_types_compatible_p (__typeof (o), struct anv_pipeline_layout*),       \
   VK_DEBUG_REPORT_OBJECT_TYPE_PIPELINE_LAYOUT_EXT,                                \
   __builtin_choose_expr (                                                         \
   __builtin_types_compatible_p (__typeof (o), struct anv_render_pass*),           \
   VK_DEBUG_REPORT_OBJECT_TYPE_RENDER_PASS_EXT,                                    \
   __builtin_choose_expr (                                                         \
   __builtin_types_compatible_p (__typeof (o), struct anv_pipeline*),              \
   VK_DEBUG_REPORT_OBJECT_TYPE_PIPELINE_EXT,                                       \
   __builtin_choose_expr (                                                         \
   __builtin_types_compatible_p (__typeof (o), struct anv_descriptor_set_layout*), \
   VK_DEBUG_REPORT_OBJECT_TYPE_DESCRIPTOR_SET_LAYOUT_EXT,                          \
   __builtin_choose_expr (                                                         \
   __builtin_types_compatible_p (__typeof (o), struct anv_sampler*),               \
   VK_DEBUG_REPORT_OBJECT_TYPE_SAMPLER_EXT,                                        \
   __builtin_choose_expr (                                                         \
   __builtin_types_compatible_p (__typeof (o), struct anv_descriptor_pool*),       \
   VK_DEBUG_REPORT_OBJECT_TYPE_DESCRIPTOR_POOL_EXT,                                \
   __builtin_choose_expr (                                                         \
   __builtin_types_compatible_p (__typeof (o), struct anv_descriptor_set*),        \
   VK_DEBUG_REPORT_OBJECT_TYPE_DESCRIPTOR_SET_EXT,                                 \
   __builtin_choose_expr (                                                         \
   __builtin_types_compatible_p (__typeof (o), struct anv_framebuffer*),           \
   VK_DEBUG_REPORT_OBJECT_TYPE_FRAMEBUFFER_EXT,                                    \
   __builtin_choose_expr (                                                         \
   __builtin_types_compatible_p (__typeof (o), struct anv_cmd_pool*),              \
   VK_DEBUG_REPORT_OBJECT_TYPE_COMMAND_POOL_EXT,                                   \
   __builtin_choose_expr (                                                         \
   __builtin_types_compatible_p (__typeof (o), struct anv_surface*),               \
   VK_DEBUG_REPORT_OBJECT_TYPE_SURFACE_KHR_EXT,                                    \
   __builtin_choose_expr (                                                         \
   __builtin_types_compatible_p (__typeof (o), struct wsi_swapchain*),             \
   VK_DEBUG_REPORT_OBJECT_TYPE_SWAPCHAIN_KHR_EXT,                                  \
   __builtin_choose_expr (                                                         \
   __builtin_types_compatible_p (__typeof (o), struct vk_debug_callback*),         \
   VK_DEBUG_REPORT_OBJECT_TYPE_DEBUG_REPORT_CALLBACK_EXT_EXT,                      \
   __builtin_choose_expr (                                                         \
   __builtin_types_compatible_p (__typeof (o), void*),                             \
   VK_DEBUG_REPORT_OBJECT_TYPE_UNKNOWN_EXT,                                        \
   /* The void expression results in a compile-time error                          \
      when assigning the result to something.  */                                  \
   (void)0)))))))))))))))))))))))))))))))

/* Whenever we generate an error, pass it through this function. Useful for
 * debugging, where we can break on it. Only call at error site, not when
 * propagating errors. Might be useful to plug in a stack trace here.
 */

VkResult __vk_errorf(struct anv_instance *instance, const void *object,
                     VkDebugReportObjectTypeEXT type, VkResult error,
                     const char *file, int line, const char *format, ...);

#ifdef DEBUG
#define vk_error(error) __vk_errorf(NULL, NULL,\
                                    VK_DEBUG_REPORT_OBJECT_TYPE_UNKNOWN_EXT,\
                                    error, __FILE__, __LINE__, NULL)
#define vk_errorf(instance, obj, error, format, ...)\
    __vk_errorf(instance, obj, REPORT_OBJECT_TYPE(obj), error,\
                __FILE__, __LINE__, format, ## __VA_ARGS__)
#else
#define vk_error(error) error
#define vk_errorf(instance, obj, error, format, ...) error
#endif

/**
 * Warn on ignored extension structs.
 *
 * The Vulkan spec requires us to ignore unsupported or unknown structs in
 * a pNext chain.  In debug mode, emitting warnings for ignored structs may
 * help us discover structs that we should not have ignored.
 *
 *
 * From the Vulkan 1.0.38 spec:
 *
 *    Any component of the implementation (the loader, any enabled layers,
 *    and drivers) must skip over, without processing (other than reading the
 *    sType and pNext members) any chained structures with sType values not
 *    defined by extensions supported by that component.
 */
#define anv_debug_ignored_stype(sType) \
   intel_logd("%s: ignored VkStructureType %u\n", __func__, (sType))

void __anv_perf_warn(struct anv_instance *instance, const void *object,
                     VkDebugReportObjectTypeEXT type, const char *file,
                     int line, const char *format, ...)
   anv_printflike(6, 7);
void anv_loge(const char *format, ...) anv_printflike(1, 2);
void anv_loge_v(const char *format, va_list va);

/**
 * Print a FINISHME message, including its source location.
 */
#define anv_finishme(format, ...) \
   do { \
      static bool reported = false; \
      if (!reported) { \
         intel_logw("%s:%d: FINISHME: " format, __FILE__, __LINE__, \
                    ##__VA_ARGS__); \
         reported = true; \
      } \
   } while (0)

/**
 * Print a perf warning message.  Set INTEL_DEBUG=perf to see these.
 */
#define anv_perf_warn(instance, obj, format, ...) \
   do { \
      static bool reported = false; \
      if (!reported && unlikely(INTEL_DEBUG & DEBUG_PERF)) { \
         __anv_perf_warn(instance, obj, REPORT_OBJECT_TYPE(obj), __FILE__, __LINE__,\
                         format, ##__VA_ARGS__); \
         reported = true; \
      } \
   } while (0)

/* A non-fatal assert.  Useful for debugging. */
#ifdef DEBUG
#define anv_assert(x) ({ \
   if (unlikely(!(x))) \
      intel_loge("%s:%d ASSERT: %s", __FILE__, __LINE__, #x); \
})
#else
#define anv_assert(x)
#endif

/* A multi-pointer allocator
 *
 * When copying data structures from the user (such as a render pass), it's
 * common to need to allocate data for a bunch of different things.  Instead
 * of doing several allocations and having to handle all of the error checking
 * that entails, it can be easier to do a single allocation.  This struct
 * helps facilitate that.  The intended usage looks like this:
 *
 *    ANV_MULTIALLOC(ma)
 *    anv_multialloc_add(&ma, &main_ptr, 1);
 *    anv_multialloc_add(&ma, &substruct1, substruct1Count);
 *    anv_multialloc_add(&ma, &substruct2, substruct2Count);
 *
 *    if (!anv_multialloc_alloc(&ma, pAllocator, VK_ALLOCATION_SCOPE_FOO))
 *       return vk_error(VK_ERROR_OUT_OF_HOST_MEORY);
 */
struct anv_multialloc {
    size_t size;
    size_t align;

    uint32_t ptr_count;
    void **ptrs[8];
};

#define ANV_MULTIALLOC_INIT \
   ((struct anv_multialloc) { 0, })

#define ANV_MULTIALLOC(_name) \
   struct anv_multialloc _name = ANV_MULTIALLOC_INIT

__attribute__((always_inline))
static inline void
_anv_multialloc_add(struct anv_multialloc *ma,
                    void **ptr, size_t size, size_t align)
{
   size_t offset = align_u64(ma->size, align);
   ma->size = offset + size;
   ma->align = MAX2(ma->align, align);

   /* Store the offset in the pointer. */
   *ptr = (void *)(uintptr_t)offset;

   assert(ma->ptr_count < ARRAY_SIZE(ma->ptrs));
   ma->ptrs[ma->ptr_count++] = ptr;
}

#define anv_multialloc_add_size(_ma, _ptr, _size) \
   _anv_multialloc_add((_ma), (void **)(_ptr), (_size), __alignof__(**(_ptr)))

#define anv_multialloc_add(_ma, _ptr, _count) \
   anv_multialloc_add_size(_ma, _ptr, (_count) * sizeof(**(_ptr)));

__attribute__((always_inline))
static inline void *
anv_multialloc_alloc(struct anv_multialloc *ma,
                     const VkAllocationCallbacks *alloc,
                     VkSystemAllocationScope scope)
{
   void *ptr = vk_alloc(alloc, ma->size, ma->align, scope);
   if (!ptr)
      return NULL;

   /* Fill out each of the pointers with their final value.
    *
    *   for (uint32_t i = 0; i < ma->ptr_count; i++)
    *      *ma->ptrs[i] = ptr + (uintptr_t)*ma->ptrs[i];
    *
    * Unfortunately, even though ma->ptr_count is basically guaranteed to be a
    * constant, GCC is incapable of figuring this out and unrolling the loop
    * so we have to give it a little help.
    */
   STATIC_ASSERT(ARRAY_SIZE(ma->ptrs) == 8);
#define _ANV_MULTIALLOC_UPDATE_POINTER(_i) \
   if ((_i) < ma->ptr_count) \
      *ma->ptrs[_i] = ptr + (uintptr_t)*ma->ptrs[_i]
   _ANV_MULTIALLOC_UPDATE_POINTER(0);
   _ANV_MULTIALLOC_UPDATE_POINTER(1);
   _ANV_MULTIALLOC_UPDATE_POINTER(2);
   _ANV_MULTIALLOC_UPDATE_POINTER(3);
   _ANV_MULTIALLOC_UPDATE_POINTER(4);
   _ANV_MULTIALLOC_UPDATE_POINTER(5);
   _ANV_MULTIALLOC_UPDATE_POINTER(6);
   _ANV_MULTIALLOC_UPDATE_POINTER(7);
#undef _ANV_MULTIALLOC_UPDATE_POINTER

   return ptr;
}

__attribute__((always_inline))
static inline void *
anv_multialloc_alloc2(struct anv_multialloc *ma,
                      const VkAllocationCallbacks *parent_alloc,
                      const VkAllocationCallbacks *alloc,
                      VkSystemAllocationScope scope)
{
   return anv_multialloc_alloc(ma, alloc ? alloc : parent_alloc, scope);
}

struct anv_bo {
   uint32_t gem_handle;

   /* Index into the current validation list.  This is used by the
    * validation list building alrogithm to track which buffers are already
    * in the validation list so that we can ensure uniqueness.
    */
   uint32_t index;

   /* Last known offset.  This value is provided by the kernel when we
    * execbuf and is used as the presumed offset for the next bunch of
    * relocations.
    */
   uint64_t offset;

   uint64_t size;
   void *map;

   /** Flags to pass to the kernel through drm_i915_exec_object2::flags */
   uint32_t flags;
};

static inline void
anv_bo_init(struct anv_bo *bo, uint32_t gem_handle, uint64_t size)
{
   bo->gem_handle = gem_handle;
   bo->index = 0;
   bo->offset = -1;
   bo->size = size;
   bo->map = NULL;
   bo->flags = 0;
}

/* Represents a lock-free linked list of "free" things.  This is used by
 * both the block pool and the state pools.  Unfortunately, in order to
 * solve the ABA problem, we can't use a single uint32_t head.
 */
union anv_free_list {
   struct {
      int32_t offset;

      /* A simple count that is incremented every time the head changes. */
      uint32_t count;
   };
   uint64_t u64;
};

#define ANV_FREE_LIST_EMPTY ((union anv_free_list) { { 1, 0 } })

struct anv_block_state {
   union {
      struct {
         uint32_t next;
         uint32_t end;
      };
      uint64_t u64;
   };
};

struct anv_block_pool {
   struct anv_device *device;

   uint64_t bo_flags;

   struct anv_bo bo;

   /* The offset from the start of the bo to the "center" of the block
    * pool.  Pointers to allocated blocks are given by
    * bo.map + center_bo_offset + offsets.
    */
   uint32_t center_bo_offset;

   /* Current memory map of the block pool.  This pointer may or may not
    * point to the actual beginning of the block pool memory.  If
    * anv_block_pool_alloc_back has ever been called, then this pointer
    * will point to the "center" position of the buffer and all offsets
    * (negative or positive) given out by the block pool alloc functions
    * will be valid relative to this pointer.
    *
    * In particular, map == bo.map + center_offset
    */
   void *map;
   int fd;

   /**
    * Array of mmaps and gem handles owned by the block pool, reclaimed when
    * the block pool is destroyed.
    */
   struct u_vector mmap_cleanups;

   struct anv_block_state state;

   struct anv_block_state back_state;
};

/* Block pools are backed by a fixed-size 1GB memfd */
#define BLOCK_POOL_MEMFD_SIZE (1ul << 30)

/* The center of the block pool is also the middle of the memfd.  This may
 * change in the future if we decide differently for some reason.
 */
#define BLOCK_POOL_MEMFD_CENTER (BLOCK_POOL_MEMFD_SIZE / 2)

static inline uint32_t
anv_block_pool_size(struct anv_block_pool *pool)
{
   return pool->state.end + pool->back_state.end;
}

struct anv_state {
   int32_t offset;
   uint32_t alloc_size;
   void *map;
};

#define ANV_STATE_NULL ((struct anv_state) { .alloc_size = 0 })

struct anv_fixed_size_state_pool {
   union anv_free_list free_list;
   struct anv_block_state block;
};

#define ANV_MIN_STATE_SIZE_LOG2 6
#define ANV_MAX_STATE_SIZE_LOG2 20

#define ANV_STATE_BUCKETS (ANV_MAX_STATE_SIZE_LOG2 - ANV_MIN_STATE_SIZE_LOG2 + 1)

struct anv_state_pool {
   struct anv_block_pool block_pool;

   /* The size of blocks which will be allocated from the block pool */
   uint32_t block_size;

   /** Free list for "back" allocations */
   union anv_free_list back_alloc_free_list;

   struct anv_fixed_size_state_pool buckets[ANV_STATE_BUCKETS];
};

struct anv_state_stream_block;

struct anv_state_stream {
   struct anv_state_pool *state_pool;

   /* The size of blocks to allocate from the state pool */
   uint32_t block_size;

   /* Current block we're allocating from */
   struct anv_state block;

   /* Offset into the current block at which to allocate the next state */
   uint32_t next;

   /* List of all blocks allocated from this pool */
   struct anv_state_stream_block *block_list;
};

/* The block_pool functions exported for testing only.  The block pool should
 * only be used via a state pool (see below).
 */
VkResult anv_block_pool_init(struct anv_block_pool *pool,
                             struct anv_device *device,
                             uint32_t initial_size,
                             uint64_t bo_flags);
void anv_block_pool_finish(struct anv_block_pool *pool);
int32_t anv_block_pool_alloc(struct anv_block_pool *pool,
                             uint32_t block_size);
int32_t anv_block_pool_alloc_back(struct anv_block_pool *pool,
                                  uint32_t block_size);

VkResult anv_state_pool_init(struct anv_state_pool *pool,
                             struct anv_device *device,
                             uint32_t block_size,
                             uint64_t bo_flags);
void anv_state_pool_finish(struct anv_state_pool *pool);
struct anv_state anv_state_pool_alloc(struct anv_state_pool *pool,
                                      uint32_t state_size, uint32_t alignment);
struct anv_state anv_state_pool_alloc_back(struct anv_state_pool *pool);
void anv_state_pool_free(struct anv_state_pool *pool, struct anv_state state);
void anv_state_stream_init(struct anv_state_stream *stream,
                           struct anv_state_pool *state_pool,
                           uint32_t block_size);
void anv_state_stream_finish(struct anv_state_stream *stream);
struct anv_state anv_state_stream_alloc(struct anv_state_stream *stream,
                                        uint32_t size, uint32_t alignment);

/**
 * Implements a pool of re-usable BOs.  The interface is identical to that
 * of block_pool except that each block is its own BO.
 */
struct anv_bo_pool {
   struct anv_device *device;

   uint64_t bo_flags;

   void *free_list[16];
};

void anv_bo_pool_init(struct anv_bo_pool *pool, struct anv_device *device,
                      uint64_t bo_flags);
void anv_bo_pool_finish(struct anv_bo_pool *pool);
VkResult anv_bo_pool_alloc(struct anv_bo_pool *pool, struct anv_bo *bo,
                           uint32_t size);
void anv_bo_pool_free(struct anv_bo_pool *pool, const struct anv_bo *bo);

struct anv_scratch_bo {
   bool exists;
   struct anv_bo bo;
};

struct anv_scratch_pool {
   /* Indexed by Per-Thread Scratch Space number (the hardware value) and stage */
   struct anv_scratch_bo bos[16][MESA_SHADER_STAGES];
};

void anv_scratch_pool_init(struct anv_device *device,
                           struct anv_scratch_pool *pool);
void anv_scratch_pool_finish(struct anv_device *device,
                             struct anv_scratch_pool *pool);
struct anv_bo *anv_scratch_pool_alloc(struct anv_device *device,
                                      struct anv_scratch_pool *pool,
                                      gl_shader_stage stage,
                                      unsigned per_thread_scratch);

/** Implements a BO cache that ensures a 1-1 mapping of GEM BOs to anv_bos */
struct anv_bo_cache {
   struct hash_table *bo_map;
   pthread_mutex_t mutex;
};

VkResult anv_bo_cache_init(struct anv_bo_cache *cache);
void anv_bo_cache_finish(struct anv_bo_cache *cache);
VkResult anv_bo_cache_alloc(struct anv_device *device,
                            struct anv_bo_cache *cache,
                            uint64_t size, struct anv_bo **bo);
VkResult anv_bo_cache_import(struct anv_device *device,
                             struct anv_bo_cache *cache,
                             int fd, struct anv_bo **bo);
VkResult anv_bo_cache_export(struct anv_device *device,
                             struct anv_bo_cache *cache,
                             struct anv_bo *bo_in, int *fd_out);
void anv_bo_cache_release(struct anv_device *device,
                          struct anv_bo_cache *cache,
                          struct anv_bo *bo);

struct anv_memory_type {
   /* Standard bits passed on to the client */
   VkMemoryPropertyFlags   propertyFlags;
   uint32_t                heapIndex;

   /* Driver-internal book-keeping */
   VkBufferUsageFlags      valid_buffer_usage;
};

struct anv_memory_heap {
   /* Standard bits passed on to the client */
   VkDeviceSize      size;
   VkMemoryHeapFlags flags;

   /* Driver-internal book-keeping */
   bool              supports_48bit_addresses;
};

struct anv_physical_device {
    VK_LOADER_DATA                              _loader_data;

    struct anv_instance *                       instance;
    uint32_t                                    chipset_id;
    bool                                        no_hw;
    char                                        path[20];
    const char *                                name;
    struct gen_device_info                      info;
    /** Amount of "GPU memory" we want to advertise
     *
     * Clearly, this value is bogus since Intel is a UMA architecture.  On
     * gen7 platforms, we are limited by GTT size unless we want to implement
     * fine-grained tracking and GTT splitting.  On Broadwell and above we are
     * practically unlimited.  However, we will never report more than 3/4 of
     * the total system ram to try and avoid running out of RAM.
     */
    bool                                        supports_48bit_addresses;
    struct brw_compiler *                       compiler;
    struct isl_device                           isl_dev;
    int                                         cmd_parser_version;
    bool                                        has_exec_async;
    bool                                        has_exec_capture;
    bool                                        has_exec_fence;
    bool                                        has_syncobj;
    bool                                        has_syncobj_wait;
    bool                                        has_context_priority;

    struct anv_device_extension_table           supported_extensions;

    uint32_t                                    eu_total;
    uint32_t                                    subslice_total;

    struct {
      uint32_t                                  type_count;
      struct anv_memory_type                    types[VK_MAX_MEMORY_TYPES];
      uint32_t                                  heap_count;
      struct anv_memory_heap                    heaps[VK_MAX_MEMORY_HEAPS];
    } memory;

    uint8_t                                     pipeline_cache_uuid[VK_UUID_SIZE];
    uint8_t                                     driver_uuid[VK_UUID_SIZE];
    uint8_t                                     device_uuid[VK_UUID_SIZE];

    struct wsi_device                       wsi_device;
    int                                         local_fd;
};

struct anv_instance {
    VK_LOADER_DATA                              _loader_data;

    VkAllocationCallbacks                       alloc;

    uint32_t                                    apiVersion;
    struct anv_instance_extension_table         enabled_extensions;
    struct anv_dispatch_table                   dispatch;

    int                                         physicalDeviceCount;
    struct anv_physical_device                  physicalDevice;

    struct vk_debug_report_instance             debug_report_callbacks;
};

VkResult anv_init_wsi(struct anv_physical_device *physical_device);
void anv_finish_wsi(struct anv_physical_device *physical_device);

uint32_t anv_physical_device_api_version(struct anv_physical_device *dev);
bool anv_physical_device_extension_supported(struct anv_physical_device *dev,
                                             const char *name);

struct anv_queue {
    VK_LOADER_DATA                              _loader_data;

    struct anv_device *                         device;

    struct anv_state_pool *                     pool;
};

struct anv_pipeline_cache {
   struct anv_device *                          device;
   pthread_mutex_t                              mutex;

   struct hash_table *                          cache;
};

struct anv_pipeline_bind_map;

void anv_pipeline_cache_init(struct anv_pipeline_cache *cache,
                             struct anv_device *device,
                             bool cache_enabled);
void anv_pipeline_cache_finish(struct anv_pipeline_cache *cache);

struct anv_shader_bin *
anv_pipeline_cache_search(struct anv_pipeline_cache *cache,
                          const void *key, uint32_t key_size);
struct anv_shader_bin *
anv_pipeline_cache_upload_kernel(struct anv_pipeline_cache *cache,
                                 const void *key_data, uint32_t key_size,
                                 const void *kernel_data, uint32_t kernel_size,
                                 const struct brw_stage_prog_data *prog_data,
                                 uint32_t prog_data_size,
                                 const struct anv_pipeline_bind_map *bind_map);

struct anv_device {
    VK_LOADER_DATA                              _loader_data;

    VkAllocationCallbacks                       alloc;

    struct anv_instance *                       instance;
    uint32_t                                    chipset_id;
    bool                                        no_hw;
    struct gen_device_info                      info;
    struct isl_device                           isl_dev;
    int                                         context_id;
    int                                         fd;
    bool                                        can_chain_batches;
    bool                                        robust_buffer_access;
    struct anv_device_extension_table           enabled_extensions;
    struct anv_dispatch_table                   dispatch;

    struct anv_bo_pool                          batch_bo_pool;

    struct anv_bo_cache                         bo_cache;

    struct anv_state_pool                       dynamic_state_pool;
    struct anv_state_pool                       instruction_state_pool;
    struct anv_state_pool                       surface_state_pool;

    struct anv_bo                               workaround_bo;
    struct anv_bo                               trivial_batch_bo;

    struct anv_pipeline_cache                   blorp_shader_cache;
    struct blorp_context                        blorp;

    struct anv_state                            border_colors;

    struct anv_queue                            queue;

    struct anv_scratch_pool                     scratch_pool;

    uint32_t                                    default_mocs;

    pthread_mutex_t                             mutex;
    pthread_cond_t                              queue_submit;
    bool                                        lost;
};

static void inline
anv_state_flush(struct anv_device *device, struct anv_state state)
{
   if (device->info.has_llc)
      return;

   gen_flush_range(state.map, state.alloc_size);
}

void anv_device_init_blorp(struct anv_device *device);
void anv_device_finish_blorp(struct anv_device *device);

VkResult anv_device_execbuf(struct anv_device *device,
                            struct drm_i915_gem_execbuffer2 *execbuf,
                            struct anv_bo **execbuf_bos);
VkResult anv_device_query_status(struct anv_device *device);
VkResult anv_device_bo_busy(struct anv_device *device, struct anv_bo *bo);
VkResult anv_device_wait(struct anv_device *device, struct anv_bo *bo,
                         int64_t timeout);

void* anv_gem_mmap(struct anv_device *device,
                   uint32_t gem_handle, uint64_t offset, uint64_t size, uint32_t flags);
void anv_gem_munmap(void *p, uint64_t size);
uint32_t anv_gem_create(struct anv_device *device, uint64_t size);
void anv_gem_close(struct anv_device *device, uint32_t gem_handle);
uint32_t anv_gem_userptr(struct anv_device *device, void *mem, size_t size);
int anv_gem_busy(struct anv_device *device, uint32_t gem_handle);
int anv_gem_wait(struct anv_device *device, uint32_t gem_handle, int64_t *timeout_ns);
int anv_gem_execbuffer(struct anv_device *device,
                       struct drm_i915_gem_execbuffer2 *execbuf);
int anv_gem_set_tiling(struct anv_device *device, uint32_t gem_handle,
                       uint32_t stride, uint32_t tiling);
int anv_gem_create_context(struct anv_device *device);
bool anv_gem_has_context_priority(int fd);
int anv_gem_destroy_context(struct anv_device *device, int context);
int anv_gem_set_context_param(int fd, int context, uint32_t param,
                              uint64_t value);
int anv_gem_get_context_param(int fd, int context, uint32_t param,
                              uint64_t *value);
int anv_gem_get_param(int fd, uint32_t param);
int anv_gem_get_tiling(struct anv_device *device, uint32_t gem_handle);
bool anv_gem_get_bit6_swizzle(int fd, uint32_t tiling);
int anv_gem_get_aperture(int fd, uint64_t *size);
bool anv_gem_supports_48b_addresses(int fd);
int anv_gem_gpu_get_reset_stats(struct anv_device *device,
                                uint32_t *active, uint32_t *pending);
int anv_gem_handle_to_fd(struct anv_device *device, uint32_t gem_handle);
uint32_t anv_gem_fd_to_handle(struct anv_device *device, int fd);
int anv_gem_set_caching(struct anv_device *device, uint32_t gem_handle, uint32_t caching);
int anv_gem_set_domain(struct anv_device *device, uint32_t gem_handle,
                       uint32_t read_domains, uint32_t write_domain);
int anv_gem_sync_file_merge(struct anv_device *device, int fd1, int fd2);
uint32_t anv_gem_syncobj_create(struct anv_device *device, uint32_t flags);
void anv_gem_syncobj_destroy(struct anv_device *device, uint32_t handle);
int anv_gem_syncobj_handle_to_fd(struct anv_device *device, uint32_t handle);
uint32_t anv_gem_syncobj_fd_to_handle(struct anv_device *device, int fd);
int anv_gem_syncobj_export_sync_file(struct anv_device *device,
                                     uint32_t handle);
int anv_gem_syncobj_import_sync_file(struct anv_device *device,
                                     uint32_t handle, int fd);
void anv_gem_syncobj_reset(struct anv_device *device, uint32_t handle);
bool anv_gem_supports_syncobj_wait(int fd);
int anv_gem_syncobj_wait(struct anv_device *device,
                         uint32_t *handles, uint32_t num_handles,
                         int64_t abs_timeout_ns, bool wait_all);

VkResult anv_bo_init_new(struct anv_bo *bo, struct anv_device *device, uint64_t size);

struct anv_reloc_list {
   uint32_t                                     num_relocs;
   uint32_t                                     array_length;
   struct drm_i915_gem_relocation_entry *       relocs;
   struct anv_bo **                             reloc_bos;
};

VkResult anv_reloc_list_init(struct anv_reloc_list *list,
                             const VkAllocationCallbacks *alloc);
void anv_reloc_list_finish(struct anv_reloc_list *list,
                           const VkAllocationCallbacks *alloc);

VkResult anv_reloc_list_add(struct anv_reloc_list *list,
                            const VkAllocationCallbacks *alloc,
                            uint32_t offset, struct anv_bo *target_bo,
                            uint32_t delta);

struct anv_batch_bo {
   /* Link in the anv_cmd_buffer.owned_batch_bos list */
   struct list_head                             link;

   struct anv_bo                                bo;

   /* Bytes actually consumed in this batch BO */
   uint32_t                                     length;

   struct anv_reloc_list                        relocs;
};

struct anv_batch {
   const VkAllocationCallbacks *                alloc;

   void *                                       start;
   void *                                       end;
   void *                                       next;

   struct anv_reloc_list *                      relocs;

   /* This callback is called (with the associated user data) in the event
    * that the batch runs out of space.
    */
   VkResult (*extend_cb)(struct anv_batch *, void *);
   void *                                       user_data;

   /**
    * Current error status of the command buffer. Used to track inconsistent
    * or incomplete command buffer states that are the consequence of run-time
    * errors such as out of memory scenarios. We want to track this in the
    * batch because the command buffer object is not visible to some parts
    * of the driver.
    */
   VkResult                                     status;
};

void *anv_batch_emit_dwords(struct anv_batch *batch, int num_dwords);
void anv_batch_emit_batch(struct anv_batch *batch, struct anv_batch *other);
uint64_t anv_batch_emit_reloc(struct anv_batch *batch,
                              void *location, struct anv_bo *bo, uint32_t offset);
VkResult anv_device_submit_simple_batch(struct anv_device *device,
                                        struct anv_batch *batch);

static inline VkResult
anv_batch_set_error(struct anv_batch *batch, VkResult error)
{
   assert(error != VK_SUCCESS);
   if (batch->status == VK_SUCCESS)
      batch->status = error;
   return batch->status;
}

static inline bool
anv_batch_has_error(struct anv_batch *batch)
{
   return batch->status != VK_SUCCESS;
}

struct anv_address {
   struct anv_bo *bo;
   uint32_t offset;
};

static inline uint64_t
_anv_combine_address(struct anv_batch *batch, void *location,
                     const struct anv_address address, uint32_t delta)
{
   if (address.bo == NULL) {
      return address.offset + delta;
   } else {
      assert(batch->start <= location && location < batch->end);

      return anv_batch_emit_reloc(batch, location, address.bo, address.offset + delta);
   }
}

#define __gen_address_type struct anv_address
#define __gen_user_data struct anv_batch
#define __gen_combine_address _anv_combine_address

/* Wrapper macros needed to work around preprocessor argument issues.  In
 * particular, arguments don't get pre-evaluated if they are concatenated.
 * This means that, if you pass GENX(3DSTATE_PS) into the emit macro, the
 * GENX macro won't get evaluated if the emit macro contains "cmd ## foo".
 * We can work around this easily enough with these helpers.
 */
#define __anv_cmd_length(cmd) cmd ## _length
#define __anv_cmd_length_bias(cmd) cmd ## _length_bias
#define __anv_cmd_header(cmd) cmd ## _header
#define __anv_cmd_pack(cmd) cmd ## _pack
#define __anv_reg_num(reg) reg ## _num

#define anv_pack_struct(dst, struc, ...) do {                              \
      struct struc __template = {                                          \
         __VA_ARGS__                                                       \
      };                                                                   \
      __anv_cmd_pack(struc)(NULL, dst, &__template);                       \
      VG(VALGRIND_CHECK_MEM_IS_DEFINED(dst, __anv_cmd_length(struc) * 4)); \
   } while (0)

#define anv_batch_emitn(batch, n, cmd, ...) ({             \
      void *__dst = anv_batch_emit_dwords(batch, n);       \
      if (__dst) {                                         \
         struct cmd __template = {                         \
            __anv_cmd_header(cmd),                         \
           .DWordLength = n - __anv_cmd_length_bias(cmd),  \
            __VA_ARGS__                                    \
         };                                                \
         __anv_cmd_pack(cmd)(batch, __dst, &__template);   \
      }                                                    \
      __dst;                                               \
   })

#define anv_batch_emit_merge(batch, dwords0, dwords1)                   \
   do {                                                                 \
      uint32_t *dw;                                                     \
                                                                        \
      STATIC_ASSERT(ARRAY_SIZE(dwords0) == ARRAY_SIZE(dwords1));        \
      dw = anv_batch_emit_dwords((batch), ARRAY_SIZE(dwords0));         \
      if (!dw)                                                          \
         break;                                                         \
      for (uint32_t i = 0; i < ARRAY_SIZE(dwords0); i++)                \
         dw[i] = (dwords0)[i] | (dwords1)[i];                           \
      VG(VALGRIND_CHECK_MEM_IS_DEFINED(dw, ARRAY_SIZE(dwords0) * 4));\
   } while (0)

#define anv_batch_emit(batch, cmd, name)                            \
   for (struct cmd name = { __anv_cmd_header(cmd) },                    \
        *_dst = anv_batch_emit_dwords(batch, __anv_cmd_length(cmd));    \
        __builtin_expect(_dst != NULL, 1);                              \
        ({ __anv_cmd_pack(cmd)(batch, _dst, &name);                     \
           VG(VALGRIND_CHECK_MEM_IS_DEFINED(_dst, __anv_cmd_length(cmd) * 4)); \
           _dst = NULL;                                                 \
         }))

#define GEN7_MOCS (struct GEN7_MEMORY_OBJECT_CONTROL_STATE) {  \
   .GraphicsDataTypeGFDT                        = 0,           \
   .LLCCacheabilityControlLLCCC                 = 0,           \
   .L3CacheabilityControlL3CC                   = 1,           \
}

#define GEN75_MOCS (struct GEN75_MEMORY_OBJECT_CONTROL_STATE) {  \
   .LLCeLLCCacheabilityControlLLCCC             = 0,           \
   .L3CacheabilityControlL3CC                   = 1,           \
}

#define GEN8_MOCS (struct GEN8_MEMORY_OBJECT_CONTROL_STATE) {  \
      .MemoryTypeLLCeLLCCacheabilityControl = WB,              \
      .TargetCache = L3DefertoPATforLLCeLLCselection,          \
      .AgeforQUADLRU = 0                                       \
   }

/* Skylake: MOCS is now an index into an array of 62 different caching
 * configurations programmed by the kernel.
 */

#define GEN9_MOCS (struct GEN9_MEMORY_OBJECT_CONTROL_STATE) {  \
      /* TC=LLC/eLLC, LeCC=WB, LRUM=3, L3CC=WB */              \
      .IndextoMOCSTables                           = 2         \
   }

#define GEN9_MOCS_PTE {                                 \
      /* TC=LLC/eLLC, LeCC=WB, LRUM=3, L3CC=WB */       \
      .IndextoMOCSTables                           = 1  \
   }

/* Cannonlake MOCS defines are duplicates of Skylake MOCS defines. */
#define GEN10_MOCS (struct GEN10_MEMORY_OBJECT_CONTROL_STATE) {  \
      /* TC=LLC/eLLC, LeCC=WB, LRUM=3, L3CC=WB */              \
      .IndextoMOCSTables                           = 2         \
   }

#define GEN10_MOCS_PTE {                                 \
      /* TC=LLC/eLLC, LeCC=WB, LRUM=3, L3CC=WB */       \
      .IndextoMOCSTables                           = 1  \
   }

/* Ice Lake MOCS defines are duplicates of Skylake MOCS defines. */
#define GEN11_MOCS (struct GEN11_MEMORY_OBJECT_CONTROL_STATE) {  \
      /* TC=LLC/eLLC, LeCC=WB, LRUM=3, L3CC=WB */              \
      .IndextoMOCSTables                           = 2         \
   }

#define GEN11_MOCS_PTE {                                 \
      /* TC=LLC/eLLC, LeCC=WB, LRUM=3, L3CC=WB */       \
      .IndextoMOCSTables                           = 1  \
   }

struct anv_device_memory {
   struct anv_bo *                              bo;
   struct anv_memory_type *                     type;
   VkDeviceSize                                 map_size;
   void *                                       map;
};

/**
 * Header for Vertex URB Entry (VUE)
 */
struct anv_vue_header {
   uint32_t Reserved;
   uint32_t RTAIndex; /* RenderTargetArrayIndex */
   uint32_t ViewportIndex;
   float PointWidth;
};

struct anv_descriptor_set_binding_layout {
#ifndef NDEBUG
   /* The type of the descriptors in this binding */
   VkDescriptorType type;
#endif

   /* Number of array elements in this binding */
   uint16_t array_size;

   /* Index into the flattend descriptor set */
   uint16_t descriptor_index;

   /* Index into the dynamic state array for a dynamic buffer */
   int16_t dynamic_offset_index;

   /* Index into the descriptor set buffer views */
   int16_t buffer_index;

   struct {
      /* Index into the binding table for the associated surface */
      int16_t surface_index;

      /* Index into the sampler table for the associated sampler */
      int16_t sampler_index;

      /* Index into the image table for the associated image */
      int16_t image_index;
   } stage[MESA_SHADER_STAGES];

   /* Immutable samplers (or NULL if no immutable samplers) */
   struct anv_sampler **immutable_samplers;
};

struct anv_descriptor_set_layout {
   /* Descriptor set layouts can be destroyed at almost any time */
   uint32_t ref_cnt;

   /* Number of bindings in this descriptor set */
   uint16_t binding_count;

   /* Total size of the descriptor set with room for all array entries */
   uint16_t size;

   /* Shader stages affected by this descriptor set */
   uint16_t shader_stages;

   /* Number of buffers in this descriptor set */
   uint16_t buffer_count;

   /* Number of dynamic offsets used by this descriptor set */
   uint16_t dynamic_offset_count;

   /* Bindings in this descriptor set */
   struct anv_descriptor_set_binding_layout binding[0];
};

static inline void
anv_descriptor_set_layout_ref(struct anv_descriptor_set_layout *layout)
{
   assert(layout && layout->ref_cnt >= 1);
   p_atomic_inc(&layout->ref_cnt);
}

static inline void
anv_descriptor_set_layout_unref(struct anv_device *device,
                                struct anv_descriptor_set_layout *layout)
{
   assert(layout && layout->ref_cnt >= 1);
   if (p_atomic_dec_zero(&layout->ref_cnt))
      vk_free(&device->alloc, layout);
}

struct anv_descriptor {
   VkDescriptorType type;

   union {
      struct {
         VkImageLayout layout;
         struct anv_image_view *image_view;
         struct anv_sampler *sampler;
      };

      struct {
         struct anv_buffer *buffer;
         uint64_t offset;
         uint64_t range;
      };

      struct anv_buffer_view *buffer_view;
   };
};

struct anv_descriptor_set {
   struct anv_descriptor_set_layout *layout;
   uint32_t size;
   uint32_t buffer_count;
   struct anv_buffer_view *buffer_views;
   struct anv_descriptor descriptors[0];
};

struct anv_buffer_view {
   enum isl_format format; /**< VkBufferViewCreateInfo::format */
   struct anv_bo *bo;
   uint32_t offset; /**< Offset into bo. */
   uint64_t range; /**< VkBufferViewCreateInfo::range */

   struct anv_state surface_state;
   struct anv_state storage_surface_state;
   struct anv_state writeonly_storage_surface_state;

   struct brw_image_param storage_image_param;
};

struct anv_push_descriptor_set {
   struct anv_descriptor_set set;

   /* Put this field right behind anv_descriptor_set so it fills up the
    * descriptors[0] field. */
   struct anv_descriptor descriptors[MAX_PUSH_DESCRIPTORS];
   struct anv_buffer_view buffer_views[MAX_PUSH_DESCRIPTORS];
};

struct anv_descriptor_pool {
   uint32_t size;
   uint32_t next;
   uint32_t free_list;

   struct anv_state_stream surface_state_stream;
   void *surface_state_free_list;

   char data[0];
};

enum anv_descriptor_template_entry_type {
   ANV_DESCRIPTOR_TEMPLATE_ENTRY_TYPE_IMAGE,
   ANV_DESCRIPTOR_TEMPLATE_ENTRY_TYPE_BUFFER,
   ANV_DESCRIPTOR_TEMPLATE_ENTRY_TYPE_BUFFER_VIEW
};

struct anv_descriptor_template_entry {
   /* The type of descriptor in this entry */
   VkDescriptorType type;

   /* Binding in the descriptor set */
   uint32_t binding;

   /* Offset at which to write into the descriptor set binding */
   uint32_t array_element;

   /* Number of elements to write into the descriptor set binding */
   uint32_t array_count;

   /* Offset into the user provided data */
   size_t offset;

   /* Stride between elements into the user provided data */
   size_t stride;
};

struct anv_descriptor_update_template {
    VkPipelineBindPoint bind_point;

   /* The descriptor set this template corresponds to. This value is only
    * valid if the template was created with the templateType
    * VK_DESCRIPTOR_UPDATE_TEMPLATE_TYPE_DESCRIPTOR_SET_KHR.
    */
   uint8_t set;

   /* Number of entries in this template */
   uint32_t entry_count;

   /* Entries of the template */
   struct anv_descriptor_template_entry entries[0];
};

size_t
anv_descriptor_set_binding_layout_get_hw_size(const struct anv_descriptor_set_binding_layout *binding);

size_t
anv_descriptor_set_layout_size(const struct anv_descriptor_set_layout *layout);

void
anv_descriptor_set_write_image_view(struct anv_descriptor_set *set,
                                    const struct gen_device_info * const devinfo,
                                    const VkDescriptorImageInfo * const info,
                                    VkDescriptorType type,
                                    uint32_t binding,
                                    uint32_t element);

void
anv_descriptor_set_write_buffer_view(struct anv_descriptor_set *set,
                                     VkDescriptorType type,
                                     struct anv_buffer_view *buffer_view,
                                     uint32_t binding,
                                     uint32_t element);

void
anv_descriptor_set_write_buffer(struct anv_descriptor_set *set,
                                struct anv_device *device,
                                struct anv_state_stream *alloc_stream,
                                VkDescriptorType type,
                                struct anv_buffer *buffer,
                                uint32_t binding,
                                uint32_t element,
                                VkDeviceSize offset,
                                VkDeviceSize range);

void
anv_descriptor_set_write_template(struct anv_descriptor_set *set,
                                  struct anv_device *device,
                                  struct anv_state_stream *alloc_stream,
                                  const struct anv_descriptor_update_template *template,
                                  const void *data);

VkResult
anv_descriptor_set_create(struct anv_device *device,
                          struct anv_descriptor_pool *pool,
                          struct anv_descriptor_set_layout *layout,
                          struct anv_descriptor_set **out_set);

void
anv_descriptor_set_destroy(struct anv_device *device,
                           struct anv_descriptor_pool *pool,
                           struct anv_descriptor_set *set);

#define ANV_DESCRIPTOR_SET_COLOR_ATTACHMENTS UINT8_MAX

struct anv_pipeline_binding {
   /* The descriptor set this surface corresponds to.  The special value of
    * ANV_DESCRIPTOR_SET_COLOR_ATTACHMENTS indicates that the offset refers
    * to a color attachment and not a regular descriptor.
    */
   uint8_t set;

   /* Binding in the descriptor set */
   uint32_t binding;

   /* Index in the binding */
   uint32_t index;

   /* Plane in the binding index */
   uint8_t plane;

   /* Input attachment index (relative to the subpass) */
   uint8_t input_attachment_index;

   /* For a storage image, whether it is write-only */
   bool write_only;
};

struct anv_pipeline_layout {
   struct {
      struct anv_descriptor_set_layout *layout;
      uint32_t dynamic_offset_start;
   } set[MAX_SETS];

   uint32_t num_sets;

   struct {
      bool has_dynamic_offsets;
   } stage[MESA_SHADER_STAGES];

   unsigned char sha1[20];
};

struct anv_buffer {
   struct anv_device *                          device;
   VkDeviceSize                                 size;

   VkBufferUsageFlags                           usage;

   /* Set when bound */
   struct anv_bo *                              bo;
   VkDeviceSize                                 offset;
};

static inline uint64_t
anv_buffer_get_range(struct anv_buffer *buffer, uint64_t offset, uint64_t range)
{
   assert(offset <= buffer->size);
   if (range == VK_WHOLE_SIZE) {
      return buffer->size - offset;
   } else {
      assert(range <= buffer->size);
      return range;
   }
}

enum anv_cmd_dirty_bits {
   ANV_CMD_DIRTY_DYNAMIC_VIEWPORT                  = 1 << 0, /* VK_DYNAMIC_STATE_VIEWPORT */
   ANV_CMD_DIRTY_DYNAMIC_SCISSOR                   = 1 << 1, /* VK_DYNAMIC_STATE_SCISSOR */
   ANV_CMD_DIRTY_DYNAMIC_LINE_WIDTH                = 1 << 2, /* VK_DYNAMIC_STATE_LINE_WIDTH */
   ANV_CMD_DIRTY_DYNAMIC_DEPTH_BIAS                = 1 << 3, /* VK_DYNAMIC_STATE_DEPTH_BIAS */
   ANV_CMD_DIRTY_DYNAMIC_BLEND_CONSTANTS           = 1 << 4, /* VK_DYNAMIC_STATE_BLEND_CONSTANTS */
   ANV_CMD_DIRTY_DYNAMIC_DEPTH_BOUNDS              = 1 << 5, /* VK_DYNAMIC_STATE_DEPTH_BOUNDS */
   ANV_CMD_DIRTY_DYNAMIC_STENCIL_COMPARE_MASK      = 1 << 6, /* VK_DYNAMIC_STATE_STENCIL_COMPARE_MASK */
   ANV_CMD_DIRTY_DYNAMIC_STENCIL_WRITE_MASK        = 1 << 7, /* VK_DYNAMIC_STATE_STENCIL_WRITE_MASK */
   ANV_CMD_DIRTY_DYNAMIC_STENCIL_REFERENCE         = 1 << 8, /* VK_DYNAMIC_STATE_STENCIL_REFERENCE */
   ANV_CMD_DIRTY_DYNAMIC_ALL                       = (1 << 9) - 1,
   ANV_CMD_DIRTY_PIPELINE                          = 1 << 9,
   ANV_CMD_DIRTY_INDEX_BUFFER                      = 1 << 10,
   ANV_CMD_DIRTY_RENDER_TARGETS                    = 1 << 11,
};
typedef uint32_t anv_cmd_dirty_mask_t;

enum anv_pipe_bits {
   ANV_PIPE_DEPTH_CACHE_FLUSH_BIT            = (1 << 0),
   ANV_PIPE_STALL_AT_SCOREBOARD_BIT          = (1 << 1),
   ANV_PIPE_STATE_CACHE_INVALIDATE_BIT       = (1 << 2),
   ANV_PIPE_CONSTANT_CACHE_INVALIDATE_BIT    = (1 << 3),
   ANV_PIPE_VF_CACHE_INVALIDATE_BIT          = (1 << 4),
   ANV_PIPE_DATA_CACHE_FLUSH_BIT             = (1 << 5),
   ANV_PIPE_TEXTURE_CACHE_INVALIDATE_BIT     = (1 << 10),
   ANV_PIPE_INSTRUCTION_CACHE_INVALIDATE_BIT = (1 << 11),
   ANV_PIPE_RENDER_TARGET_CACHE_FLUSH_BIT    = (1 << 12),
   ANV_PIPE_DEPTH_STALL_BIT                  = (1 << 13),
   ANV_PIPE_CS_STALL_BIT                     = (1 << 20),

   /* This bit does not exist directly in PIPE_CONTROL.  Instead it means that
    * a flush has happened but not a CS stall.  The next time we do any sort
    * of invalidation we need to insert a CS stall at that time.  Otherwise,
    * we would have to CS stall on every flush which could be bad.
    */
   ANV_PIPE_NEEDS_CS_STALL_BIT               = (1 << 21),
};

#define ANV_PIPE_FLUSH_BITS ( \
   ANV_PIPE_DEPTH_CACHE_FLUSH_BIT | \
   ANV_PIPE_DATA_CACHE_FLUSH_BIT | \
   ANV_PIPE_RENDER_TARGET_CACHE_FLUSH_BIT)

#define ANV_PIPE_STALL_BITS ( \
   ANV_PIPE_STALL_AT_SCOREBOARD_BIT | \
   ANV_PIPE_DEPTH_STALL_BIT | \
   ANV_PIPE_CS_STALL_BIT)

#define ANV_PIPE_INVALIDATE_BITS ( \
   ANV_PIPE_STATE_CACHE_INVALIDATE_BIT | \
   ANV_PIPE_CONSTANT_CACHE_INVALIDATE_BIT | \
   ANV_PIPE_VF_CACHE_INVALIDATE_BIT | \
   ANV_PIPE_DATA_CACHE_FLUSH_BIT | \
   ANV_PIPE_TEXTURE_CACHE_INVALIDATE_BIT | \
   ANV_PIPE_INSTRUCTION_CACHE_INVALIDATE_BIT)

static inline enum anv_pipe_bits
anv_pipe_flush_bits_for_access_flags(VkAccessFlags flags)
{
   enum anv_pipe_bits pipe_bits = 0;

   unsigned b;
   for_each_bit(b, flags) {
      switch ((VkAccessFlagBits)(1 << b)) {
      case VK_ACCESS_SHADER_WRITE_BIT:
         pipe_bits |= ANV_PIPE_DATA_CACHE_FLUSH_BIT;
         break;
      case VK_ACCESS_COLOR_ATTACHMENT_WRITE_BIT:
         pipe_bits |= ANV_PIPE_RENDER_TARGET_CACHE_FLUSH_BIT;
         break;
      case VK_ACCESS_DEPTH_STENCIL_ATTACHMENT_WRITE_BIT:
         pipe_bits |= ANV_PIPE_DEPTH_CACHE_FLUSH_BIT;
         break;
      case VK_ACCESS_TRANSFER_WRITE_BIT:
         pipe_bits |= ANV_PIPE_RENDER_TARGET_CACHE_FLUSH_BIT;
         pipe_bits |= ANV_PIPE_DEPTH_CACHE_FLUSH_BIT;
         break;
      default:
         break; /* Nothing to do */
      }
   }

   return pipe_bits;
}

static inline enum anv_pipe_bits
anv_pipe_invalidate_bits_for_access_flags(VkAccessFlags flags)
{
   enum anv_pipe_bits pipe_bits = 0;

   unsigned b;
   for_each_bit(b, flags) {
      switch ((VkAccessFlagBits)(1 << b)) {
      case VK_ACCESS_INDIRECT_COMMAND_READ_BIT:
      case VK_ACCESS_INDEX_READ_BIT:
      case VK_ACCESS_VERTEX_ATTRIBUTE_READ_BIT:
         pipe_bits |= ANV_PIPE_VF_CACHE_INVALIDATE_BIT;
         break;
      case VK_ACCESS_UNIFORM_READ_BIT:
         pipe_bits |= ANV_PIPE_CONSTANT_CACHE_INVALIDATE_BIT;
         pipe_bits |= ANV_PIPE_TEXTURE_CACHE_INVALIDATE_BIT;
         break;
      case VK_ACCESS_SHADER_READ_BIT:
      case VK_ACCESS_INPUT_ATTACHMENT_READ_BIT:
      case VK_ACCESS_TRANSFER_READ_BIT:
         pipe_bits |= ANV_PIPE_TEXTURE_CACHE_INVALIDATE_BIT;
         break;
      default:
         break; /* Nothing to do */
      }
   }

   return pipe_bits;
}

#define VK_IMAGE_ASPECT_ANY_COLOR_BIT_ANV (         \
   VK_IMAGE_ASPECT_COLOR_BIT | \
   VK_IMAGE_ASPECT_PLANE_0_BIT | \
   VK_IMAGE_ASPECT_PLANE_1_BIT | \
   VK_IMAGE_ASPECT_PLANE_2_BIT)
#define VK_IMAGE_ASPECT_PLANES_BITS_ANV ( \
   VK_IMAGE_ASPECT_PLANE_0_BIT | \
   VK_IMAGE_ASPECT_PLANE_1_BIT | \
   VK_IMAGE_ASPECT_PLANE_2_BIT)

struct anv_vertex_binding {
   struct anv_buffer *                          buffer;
   VkDeviceSize                                 offset;
};

#define ANV_PARAM_PUSH(offset)         ((1 << 16) | (uint32_t)(offset))
#define ANV_PARAM_PUSH_OFFSET(param)   ((param) & 0xffff)

struct anv_push_constants {
   /* Current allocated size of this push constants data structure.
    * Because a decent chunk of it may not be used (images on SKL, for
    * instance), we won't actually allocate the entire structure up-front.
    */
   uint32_t size;

   /* Push constant data provided by the client through vkPushConstants */
   uint8_t client_data[MAX_PUSH_CONSTANTS_SIZE];

   /* Used for vkCmdDispatchBase */
   uint32_t base_work_group_id[3];

   /* Image data for image_load_store on pre-SKL */
   struct brw_image_param images[MAX_IMAGES];
};

struct anv_dynamic_state {
   struct {
      uint32_t                                  count;
      VkViewport                                viewports[MAX_VIEWPORTS];
   } viewport;

   struct {
      uint32_t                                  count;
      VkRect2D                                  scissors[MAX_SCISSORS];
   } scissor;

   float                                        line_width;

   struct {
      float                                     bias;
      float                                     clamp;
      float                                     slope;
   } depth_bias;

   float                                        blend_constants[4];

   struct {
      float                                     min;
      float                                     max;
   } depth_bounds;

   struct {
      uint32_t                                  front;
      uint32_t                                  back;
   } stencil_compare_mask;

   struct {
      uint32_t                                  front;
      uint32_t                                  back;
   } stencil_write_mask;

   struct {
      uint32_t                                  front;
      uint32_t                                  back;
   } stencil_reference;
};

extern const struct anv_dynamic_state default_dynamic_state;

void anv_dynamic_state_copy(struct anv_dynamic_state *dest,
                            const struct anv_dynamic_state *src,
                            uint32_t copy_mask);

struct anv_surface_state {
   struct anv_state state;
   /** Address of the surface referred to by this state
    *
    * This address is relative to the start of the BO.
    */
   uint64_t address;
   /* Address of the aux surface, if any
    *
    * This field is 0 if and only if no aux surface exists.
    *
    * This address is relative to the start of the BO.  On gen7, the bottom 12
    * bits of this address include extra aux information.
    */
   uint64_t aux_address;
};

/**
 * Attachment state when recording a renderpass instance.
 *
 * The clear value is valid only if there exists a pending clear.
 */
struct anv_attachment_state {
   enum isl_aux_usage                           aux_usage;
   enum isl_aux_usage                           input_aux_usage;
   struct anv_surface_state                     color;
   struct anv_surface_state                     input;

   VkImageLayout                                current_layout;
   VkImageAspectFlags                           pending_clear_aspects;
   VkImageAspectFlags                           pending_load_aspects;
   bool                                         fast_clear;
   VkClearValue                                 clear_value;
   bool                                         clear_color_is_zero_one;
   bool                                         clear_color_is_zero;
};

/** State tracking for particular pipeline bind point
 *
 * This struct is the base struct for anv_cmd_graphics_state and
 * anv_cmd_compute_state.  These are used to track state which is bound to a
 * particular type of pipeline.  Generic state that applies per-stage such as
 * binding table offsets and push constants is tracked generically with a
 * per-stage array in anv_cmd_state.
 */
struct anv_cmd_pipeline_state {
   struct anv_pipeline *pipeline;
   struct anv_pipeline_layout *layout;

   struct anv_descriptor_set *descriptors[MAX_SETS];
   uint32_t dynamic_offsets[MAX_DYNAMIC_BUFFERS];

   struct anv_push_descriptor_set *push_descriptors[MAX_SETS];
};

/** State tracking for graphics pipeline
 *
 * This has anv_cmd_pipeline_state as a base struct to track things which get
 * bound to a graphics pipeline.  Along with general pipeline bind point state
 * which is in the anv_cmd_pipeline_state base struct, it also contains other
 * state which is graphics-specific.
 */
struct anv_cmd_graphics_state {
   struct anv_cmd_pipeline_state base;

   anv_cmd_dirty_mask_t dirty;
   uint32_t vb_dirty;

   struct anv_dynamic_state dynamic;

   struct {
      struct anv_buffer *index_buffer;
      uint32_t index_type; /**< 3DSTATE_INDEX_BUFFER.IndexFormat */
      uint32_t index_offset;
   } gen7;
};

/** State tracking for compute pipeline
 *
 * This has anv_cmd_pipeline_state as a base struct to track things which get
 * bound to a compute pipeline.  Along with general pipeline bind point state
 * which is in the anv_cmd_pipeline_state base struct, it also contains other
 * state which is compute-specific.
 */
struct anv_cmd_compute_state {
   struct anv_cmd_pipeline_state base;

   bool pipeline_dirty;

   struct anv_address num_workgroups;
};

/** State required while building cmd buffer */
struct anv_cmd_state {
   /* PIPELINE_SELECT.PipelineSelection */
   uint32_t                                     current_pipeline;
   const struct gen_l3_config *                 current_l3_config;

   struct anv_cmd_graphics_state                gfx;
   struct anv_cmd_compute_state                 compute;

   enum anv_pipe_bits                           pending_pipe_bits;
   VkShaderStageFlags                           descriptors_dirty;
   VkShaderStageFlags                           push_constants_dirty;

   struct anv_framebuffer *                     framebuffer;
   struct anv_render_pass *                     pass;
   struct anv_subpass *                         subpass;
   VkRect2D                                     render_area;
   uint32_t                                     restart_index;
   struct anv_vertex_binding                    vertex_bindings[MAX_VBS];
   VkShaderStageFlags                           push_constant_stages;
   struct anv_push_constants *                  push_constants[MESA_SHADER_STAGES];
   struct anv_state                             binding_tables[MESA_SHADER_STAGES];
   struct anv_state                             samplers[MESA_SHADER_STAGES];

   /**
    * Whether or not the gen8 PMA fix is enabled.  We ensure that, at the top
    * of any command buffer it is disabled by disabling it in EndCommandBuffer
    * and before invoking the secondary in ExecuteCommands.
    */
   bool                                         pma_fix_enabled;

   /**
    * Whether or not we know for certain that HiZ is enabled for the current
    * subpass.  If, for whatever reason, we are unsure as to whether HiZ is
    * enabled or not, this will be false.
    */
   bool                                         hiz_enabled;

   /**
    * Array length is anv_cmd_state::pass::attachment_count. Array content is
    * valid only when recording a render pass instance.
    */
   struct anv_attachment_state *                attachments;

   /**
    * Surface states for color render targets.  These are stored in a single
    * flat array.  For depth-stencil attachments, the surface state is simply
    * left blank.
    */
   struct anv_state                             render_pass_states;

   /**
    * A null surface state of the right size to match the framebuffer.  This
    * is one of the states in render_pass_states.
    */
   struct anv_state                             null_surface_state;
};

struct anv_cmd_pool {
   VkAllocationCallbacks                        alloc;
   struct list_head                             cmd_buffers;
};

#define ANV_CMD_BUFFER_BATCH_SIZE 8192

enum anv_cmd_buffer_exec_mode {
   ANV_CMD_BUFFER_EXEC_MODE_PRIMARY,
   ANV_CMD_BUFFER_EXEC_MODE_EMIT,
   ANV_CMD_BUFFER_EXEC_MODE_GROW_AND_EMIT,
   ANV_CMD_BUFFER_EXEC_MODE_CHAIN,
   ANV_CMD_BUFFER_EXEC_MODE_COPY_AND_CHAIN,
};

struct anv_cmd_buffer {
   VK_LOADER_DATA                               _loader_data;

   struct anv_device *                          device;

   struct anv_cmd_pool *                        pool;
   struct list_head                             pool_link;

   struct anv_batch                             batch;

   /* Fields required for the actual chain of anv_batch_bo's.
    *
    * These fields are initialized by anv_cmd_buffer_init_batch_bo_chain().
    */
   struct list_head                             batch_bos;
   enum anv_cmd_buffer_exec_mode                exec_mode;

   /* A vector of anv_batch_bo pointers for every batch or surface buffer
    * referenced by this command buffer
    *
    * initialized by anv_cmd_buffer_init_batch_bo_chain()
    */
   struct u_vector                            seen_bbos;

   /* A vector of int32_t's for every block of binding tables.
    *
    * initialized by anv_cmd_buffer_init_batch_bo_chain()
    */
   struct u_vector                              bt_block_states;
   uint32_t                                     bt_next;

   struct anv_reloc_list                        surface_relocs;
   /** Last seen surface state block pool center bo offset */
   uint32_t                                     last_ss_pool_center;

   /* Serial for tracking buffer completion */
   uint32_t                                     serial;

   /* Stream objects for storing temporary data */
   struct anv_state_stream                      surface_state_stream;
   struct anv_state_stream                      dynamic_state_stream;

   VkCommandBufferUsageFlags                    usage_flags;
   VkCommandBufferLevel                         level;

   struct anv_cmd_state                         state;
};

VkResult anv_cmd_buffer_init_batch_bo_chain(struct anv_cmd_buffer *cmd_buffer);
void anv_cmd_buffer_fini_batch_bo_chain(struct anv_cmd_buffer *cmd_buffer);
void anv_cmd_buffer_reset_batch_bo_chain(struct anv_cmd_buffer *cmd_buffer);
void anv_cmd_buffer_end_batch_buffer(struct anv_cmd_buffer *cmd_buffer);
void anv_cmd_buffer_add_secondary(struct anv_cmd_buffer *primary,
                                  struct anv_cmd_buffer *secondary);
void anv_cmd_buffer_prepare_execbuf(struct anv_cmd_buffer *cmd_buffer);
VkResult anv_cmd_buffer_execbuf(struct anv_device *device,
                                struct anv_cmd_buffer *cmd_buffer,
                                const VkSemaphore *in_semaphores,
                                uint32_t num_in_semaphores,
                                const VkSemaphore *out_semaphores,
                                uint32_t num_out_semaphores,
                                VkFence fence);

VkResult anv_cmd_buffer_reset(struct anv_cmd_buffer *cmd_buffer);

VkResult
anv_cmd_buffer_ensure_push_constants_size(struct anv_cmd_buffer *cmd_buffer,
                                          gl_shader_stage stage, uint32_t size);
#define anv_cmd_buffer_ensure_push_constant_field(cmd_buffer, stage, field) \
   anv_cmd_buffer_ensure_push_constants_size(cmd_buffer, stage, \
      (offsetof(struct anv_push_constants, field) + \
       sizeof(cmd_buffer->state.push_constants[0]->field)))

struct anv_state anv_cmd_buffer_emit_dynamic(struct anv_cmd_buffer *cmd_buffer,
                                             const void *data, uint32_t size, uint32_t alignment);
struct anv_state anv_cmd_buffer_merge_dynamic(struct anv_cmd_buffer *cmd_buffer,
                                              uint32_t *a, uint32_t *b,
                                              uint32_t dwords, uint32_t alignment);

struct anv_address
anv_cmd_buffer_surface_base_address(struct anv_cmd_buffer *cmd_buffer);
struct anv_state
anv_cmd_buffer_alloc_binding_table(struct anv_cmd_buffer *cmd_buffer,
                                   uint32_t entries, uint32_t *state_offset);
struct anv_state
anv_cmd_buffer_alloc_surface_state(struct anv_cmd_buffer *cmd_buffer);
struct anv_state
anv_cmd_buffer_alloc_dynamic_state(struct anv_cmd_buffer *cmd_buffer,
                                   uint32_t size, uint32_t alignment);

VkResult
anv_cmd_buffer_new_binding_table_block(struct anv_cmd_buffer *cmd_buffer);

void gen8_cmd_buffer_emit_viewport(struct anv_cmd_buffer *cmd_buffer);
void gen8_cmd_buffer_emit_depth_viewport(struct anv_cmd_buffer *cmd_buffer,
                                         bool depth_clamp_enable);
void gen7_cmd_buffer_emit_scissor(struct anv_cmd_buffer *cmd_buffer);

void anv_cmd_buffer_setup_attachments(struct anv_cmd_buffer *cmd_buffer,
                                      struct anv_render_pass *pass,
                                      struct anv_framebuffer *framebuffer,
                                      const VkClearValue *clear_values);

void anv_cmd_buffer_emit_state_base_address(struct anv_cmd_buffer *cmd_buffer);

struct anv_state
anv_cmd_buffer_push_constants(struct anv_cmd_buffer *cmd_buffer,
                              gl_shader_stage stage);
struct anv_state
anv_cmd_buffer_cs_push_constants(struct anv_cmd_buffer *cmd_buffer);

void anv_cmd_buffer_resolve_subpass(struct anv_cmd_buffer *cmd_buffer);

const struct anv_image_view *
anv_cmd_buffer_get_depth_stencil_view(const struct anv_cmd_buffer *cmd_buffer);

VkResult
anv_cmd_buffer_alloc_blorp_binding_table(struct anv_cmd_buffer *cmd_buffer,
                                         uint32_t num_entries,
                                         uint32_t *state_offset,
                                         struct anv_state *bt_state);

void anv_cmd_buffer_dump(struct anv_cmd_buffer *cmd_buffer);

enum anv_fence_type {
   ANV_FENCE_TYPE_NONE = 0,
   ANV_FENCE_TYPE_BO,
   ANV_FENCE_TYPE_SYNCOBJ,
};

enum anv_bo_fence_state {
   /** Indicates that this is a new (or newly reset fence) */
   ANV_BO_FENCE_STATE_RESET,

   /** Indicates that this fence has been submitted to the GPU but is still
    * (as far as we know) in use by the GPU.
    */
   ANV_BO_FENCE_STATE_SUBMITTED,

   ANV_BO_FENCE_STATE_SIGNALED,
};

struct anv_fence_impl {
   enum anv_fence_type type;

   union {
      /** Fence implementation for BO fences
       *
       * These fences use a BO and a set of CPU-tracked state flags.  The BO
       * is added to the object list of the last execbuf call in a QueueSubmit
       * and is marked EXEC_WRITE.  The state flags track when the BO has been
       * submitted to the kernel.  We need to do this because Vulkan lets you
       * wait on a fence that has not yet been submitted and I915_GEM_BUSY
       * will say it's idle in this case.
       */
      struct {
         struct anv_bo bo;
         enum anv_bo_fence_state state;
      } bo;

      /** DRM syncobj handle for syncobj-based fences */
      uint32_t syncobj;
   };
};

struct anv_fence {
   /* Permanent fence state.  Every fence has some form of permanent state
    * (type != ANV_SEMAPHORE_TYPE_NONE).  This may be a BO to fence on (for
    * cross-process fences) or it could just be a dummy for use internally.
    */
   struct anv_fence_impl permanent;

   /* Temporary fence state.  A fence *may* have temporary state.  That state
    * is added to the fence by an import operation and is reset back to
    * ANV_SEMAPHORE_TYPE_NONE when the fence is reset.  A fence with temporary
    * state cannot be signaled because the fence must already be signaled
    * before the temporary state can be exported from the fence in the other
    * process and imported here.
    */
   struct anv_fence_impl temporary;
};

struct anv_event {
   uint64_t                                     semaphore;
   struct anv_state                             state;
};

enum anv_semaphore_type {
   ANV_SEMAPHORE_TYPE_NONE = 0,
   ANV_SEMAPHORE_TYPE_DUMMY,
   ANV_SEMAPHORE_TYPE_BO,
   ANV_SEMAPHORE_TYPE_SYNC_FILE,
   ANV_SEMAPHORE_TYPE_DRM_SYNCOBJ,
};

struct anv_semaphore_impl {
   enum anv_semaphore_type type;

   union {
      /* A BO representing this semaphore when type == ANV_SEMAPHORE_TYPE_BO.
       * This BO will be added to the object list on any execbuf2 calls for
       * which this semaphore is used as a wait or signal fence.  When used as
       * a signal fence, the EXEC_OBJECT_WRITE flag will be set.
       */
      struct anv_bo *bo;

      /* The sync file descriptor when type == ANV_SEMAPHORE_TYPE_SYNC_FILE.
       * If the semaphore is in the unsignaled state due to either just being
       * created or because it has been used for a wait, fd will be -1.
       */
      int fd;

      /* Sync object handle when type == ANV_SEMAPHORE_TYPE_DRM_SYNCOBJ.
       * Unlike GEM BOs, DRM sync objects aren't deduplicated by the kernel on
       * import so we don't need to bother with a userspace cache.
       */
      uint32_t syncobj;
   };
};

struct anv_semaphore {
   /* Permanent semaphore state.  Every semaphore has some form of permanent
    * state (type != ANV_SEMAPHORE_TYPE_NONE).  This may be a BO to fence on
    * (for cross-process semaphores0 or it could just be a dummy for use
    * internally.
    */
   struct anv_semaphore_impl permanent;

   /* Temporary semaphore state.  A semaphore *may* have temporary state.
    * That state is added to the semaphore by an import operation and is reset
    * back to ANV_SEMAPHORE_TYPE_NONE when the semaphore is waited on.  A
    * semaphore with temporary state cannot be signaled because the semaphore
    * must already be signaled before the temporary state can be exported from
    * the semaphore in the other process and imported here.
    */
   struct anv_semaphore_impl temporary;
};

void anv_semaphore_reset_temporary(struct anv_device *device,
                                   struct anv_semaphore *semaphore);

struct anv_shader_module {
   unsigned char                                sha1[20];
   uint32_t                                     size;
   char                                         data[0];
};

static inline gl_shader_stage
vk_to_mesa_shader_stage(VkShaderStageFlagBits vk_stage)
{
   assert(__builtin_popcount(vk_stage) == 1);
   return ffs(vk_stage) - 1;
}

static inline VkShaderStageFlagBits
mesa_to_vk_shader_stage(gl_shader_stage mesa_stage)
{
   return (1 << mesa_stage);
}

#define ANV_STAGE_MASK ((1 << MESA_SHADER_STAGES) - 1)

#define anv_foreach_stage(stage, stage_bits)                         \
   for (gl_shader_stage stage,                                       \
        __tmp = (gl_shader_stage)((stage_bits) & ANV_STAGE_MASK);    \
        stage = __builtin_ffs(__tmp) - 1, __tmp;                     \
        __tmp &= ~(1 << (stage)))

struct anv_pipeline_bind_map {
   uint32_t surface_count;
   uint32_t sampler_count;
   uint32_t image_count;

   struct anv_pipeline_binding *                surface_to_descriptor;
   struct anv_pipeline_binding *                sampler_to_descriptor;
};

struct anv_shader_bin_key {
   uint32_t size;
   uint8_t data[0];
};

struct anv_shader_bin {
   uint32_t ref_cnt;

   const struct anv_shader_bin_key *key;

   struct anv_state kernel;
   uint32_t kernel_size;

   const struct brw_stage_prog_data *prog_data;
   uint32_t prog_data_size;

   struct anv_pipeline_bind_map bind_map;
};

struct anv_shader_bin *
anv_shader_bin_create(struct anv_device *device,
                      const void *key, uint32_t key_size,
                      const void *kernel, uint32_t kernel_size,
                      const struct brw_stage_prog_data *prog_data,
                      uint32_t prog_data_size, const void *prog_data_param,
                      const struct anv_pipeline_bind_map *bind_map);

void
anv_shader_bin_destroy(struct anv_device *device, struct anv_shader_bin *shader);

static inline void
anv_shader_bin_ref(struct anv_shader_bin *shader)
{
   assert(shader && shader->ref_cnt >= 1);
   p_atomic_inc(&shader->ref_cnt);
}

static inline void
anv_shader_bin_unref(struct anv_device *device, struct anv_shader_bin *shader)
{
   assert(shader && shader->ref_cnt >= 1);
   if (p_atomic_dec_zero(&shader->ref_cnt))
      anv_shader_bin_destroy(device, shader);
}

struct anv_pipeline {
   struct anv_device *                          device;
   struct anv_batch                             batch;
   uint32_t                                     batch_data[512];
   struct anv_reloc_list                        batch_relocs;
   uint32_t                                     dynamic_state_mask;
   struct anv_dynamic_state                     dynamic_state;

   struct anv_subpass *                         subpass;

   bool                                         needs_data_cache;

   struct anv_shader_bin *                      shaders[MESA_SHADER_STAGES];

   struct {
      const struct gen_l3_config *              l3_config;
      uint32_t                                  total_size;
   } urb;

   VkShaderStageFlags                           active_stages;
   struct anv_state                             blend_state;

   uint32_t                                     vb_used;
   uint32_t                                     binding_stride[MAX_VBS];
   bool                                         instancing_enable[MAX_VBS];
   bool                                         primitive_restart;
   uint32_t                                     topology;

   uint32_t                                     cs_right_mask;

   bool                                         writes_depth;
   bool                                         depth_test_enable;
   bool                                         writes_stencil;
   bool                                         stencil_test_enable;
   bool                                         depth_clamp_enable;
   bool                                         sample_shading_enable;
   bool                                         kill_pixel;

   struct {
      uint32_t                                  sf[7];
      uint32_t                                  depth_stencil_state[3];
   } gen7;

   struct {
      uint32_t                                  sf[4];
      uint32_t                                  raster[5];
      uint32_t                                  wm_depth_stencil[3];
   } gen8;

   struct {
      uint32_t                                  wm_depth_stencil[4];
   } gen9;

   uint32_t                                     interface_descriptor_data[8];
};

static inline bool
anv_pipeline_has_stage(const struct anv_pipeline *pipeline,
                       gl_shader_stage stage)
{
   return (pipeline->active_stages & mesa_to_vk_shader_stage(stage)) != 0;
}

#define ANV_DECL_GET_PROG_DATA_FUNC(prefix, stage)                   \
static inline const struct brw_##prefix##_prog_data *                \
get_##prefix##_prog_data(const struct anv_pipeline *pipeline)        \
{                                                                    \
   if (anv_pipeline_has_stage(pipeline, stage)) {                    \
      return (const struct brw_##prefix##_prog_data *)               \
             pipeline->shaders[stage]->prog_data;                    \
   } else {                                                          \
      return NULL;                                                   \
   }                                                                 \
}

ANV_DECL_GET_PROG_DATA_FUNC(vs, MESA_SHADER_VERTEX)
ANV_DECL_GET_PROG_DATA_FUNC(tcs, MESA_SHADER_TESS_CTRL)
ANV_DECL_GET_PROG_DATA_FUNC(tes, MESA_SHADER_TESS_EVAL)
ANV_DECL_GET_PROG_DATA_FUNC(gs, MESA_SHADER_GEOMETRY)
ANV_DECL_GET_PROG_DATA_FUNC(wm, MESA_SHADER_FRAGMENT)
ANV_DECL_GET_PROG_DATA_FUNC(cs, MESA_SHADER_COMPUTE)

static inline const struct brw_vue_prog_data *
anv_pipeline_get_last_vue_prog_data(const struct anv_pipeline *pipeline)
{
   if (anv_pipeline_has_stage(pipeline, MESA_SHADER_GEOMETRY))
      return &get_gs_prog_data(pipeline)->base;
   else if (anv_pipeline_has_stage(pipeline, MESA_SHADER_TESS_EVAL))
      return &get_tes_prog_data(pipeline)->base;
   else
      return &get_vs_prog_data(pipeline)->base;
}

VkResult
anv_pipeline_init(struct anv_pipeline *pipeline, struct anv_device *device,
                  struct anv_pipeline_cache *cache,
                  const VkGraphicsPipelineCreateInfo *pCreateInfo,
                  const VkAllocationCallbacks *alloc);

VkResult
anv_pipeline_compile_cs(struct anv_pipeline *pipeline,
                        struct anv_pipeline_cache *cache,
                        const VkComputePipelineCreateInfo *info,
                        struct anv_shader_module *module,
                        const char *entrypoint,
                        const VkSpecializationInfo *spec_info);

struct anv_format_plane {
   enum isl_format isl_format:16;
   struct isl_swizzle swizzle;

   /* Whether this plane contains chroma channels */
   bool has_chroma;

   /* For downscaling of YUV planes */
   uint8_t denominator_scales[2];

   /* How to map sampled ycbcr planes to a single 4 component element. */
   struct isl_swizzle ycbcr_swizzle;
};


struct anv_format {
   struct anv_format_plane planes[3];
   uint8_t n_planes;
   bool can_ycbcr;
};

static inline uint32_t
anv_image_aspect_to_plane(VkImageAspectFlags image_aspects,
                          VkImageAspectFlags aspect_mask)
{
   switch (aspect_mask) {
   case VK_IMAGE_ASPECT_COLOR_BIT:
   case VK_IMAGE_ASPECT_DEPTH_BIT:
   case VK_IMAGE_ASPECT_PLANE_0_BIT:
      return 0;
   case VK_IMAGE_ASPECT_STENCIL_BIT:
      if ((image_aspects & VK_IMAGE_ASPECT_DEPTH_BIT) == 0)
         return 0;
      /* Fall-through */
   case VK_IMAGE_ASPECT_PLANE_1_BIT:
      return 1;
   case VK_IMAGE_ASPECT_PLANE_2_BIT:
      return 2;
   default:
      /* Purposefully assert with depth/stencil aspects. */
      unreachable("invalid image aspect");
   }
}

static inline uint32_t
anv_image_aspect_get_planes(VkImageAspectFlags aspect_mask)
{
   uint32_t planes = 0;

   if (aspect_mask & (VK_IMAGE_ASPECT_COLOR_BIT |
                      VK_IMAGE_ASPECT_DEPTH_BIT |
                      VK_IMAGE_ASPECT_STENCIL_BIT |
                      VK_IMAGE_ASPECT_PLANE_0_BIT))
      planes++;
   if (aspect_mask & VK_IMAGE_ASPECT_PLANE_1_BIT)
      planes++;
   if (aspect_mask & VK_IMAGE_ASPECT_PLANE_2_BIT)
      planes++;

   return planes;
}

static inline VkImageAspectFlags
anv_plane_to_aspect(VkImageAspectFlags image_aspects,
                    uint32_t plane)
{
   if (image_aspects & VK_IMAGE_ASPECT_ANY_COLOR_BIT_ANV) {
      if (_mesa_bitcount(image_aspects) > 1)
         return VK_IMAGE_ASPECT_PLANE_0_BIT << plane;
      return VK_IMAGE_ASPECT_COLOR_BIT;
   }
   if (image_aspects & VK_IMAGE_ASPECT_DEPTH_BIT)
      return VK_IMAGE_ASPECT_DEPTH_BIT << plane;
   assert(image_aspects == VK_IMAGE_ASPECT_STENCIL_BIT);
   return VK_IMAGE_ASPECT_STENCIL_BIT;
}

#define anv_foreach_image_aspect_bit(b, image, aspects) \
   for_each_bit(b, anv_image_expand_aspects(image, aspects))

const struct anv_format *
anv_get_format(VkFormat format);

static inline uint32_t
anv_get_format_planes(VkFormat vk_format)
{
   const struct anv_format *format = anv_get_format(vk_format);

   return format != NULL ? format->n_planes : 0;
}

struct anv_format_plane
anv_get_format_plane(const struct gen_device_info *devinfo, VkFormat vk_format,
                     VkImageAspectFlagBits aspect, VkImageTiling tiling);

static inline enum isl_format
anv_get_isl_format(const struct gen_device_info *devinfo, VkFormat vk_format,
                   VkImageAspectFlags aspect, VkImageTiling tiling)
{
   return anv_get_format_plane(devinfo, vk_format, aspect, tiling).isl_format;
}

static inline struct isl_swizzle
anv_swizzle_for_render(struct isl_swizzle swizzle)
{
   /* Sometimes the swizzle will have alpha map to one.  We do this to fake
    * RGB as RGBA for texturing
    */
   assert(swizzle.a == ISL_CHANNEL_SELECT_ONE ||
          swizzle.a == ISL_CHANNEL_SELECT_ALPHA);

   /* But it doesn't matter what we render to that channel */
   swizzle.a = ISL_CHANNEL_SELECT_ALPHA;

   return swizzle;
}

void
anv_pipeline_setup_l3_config(struct anv_pipeline *pipeline, bool needs_slm);

/**
 * Subsurface of an anv_image.
 */
struct anv_surface {
   /** Valid only if isl_surf::size > 0. */
   struct isl_surf isl;

   /**
    * Offset from VkImage's base address, as bound by vkBindImageMemory().
    */
   uint32_t offset;
};

struct anv_image {
   VkImageType type;
   /* The original VkFormat provided by the client.  This may not match any
    * of the actual surface formats.
    */
   VkFormat vk_format;
   const struct anv_format *format;

   VkImageAspectFlags aspects;
   VkExtent3D extent;
   uint32_t levels;
   uint32_t array_size;
   uint32_t samples; /**< VkImageCreateInfo::samples */
   uint32_t n_planes;
   VkImageUsageFlags usage; /**< Superset of VkImageCreateInfo::usage. */
   VkImageTiling tiling; /** VkImageCreateInfo::tiling */

   /** True if this is needs to be bound to an appropriately tiled BO.
    *
    * When not using modifiers, consumers such as X11, Wayland, and KMS need
    * the tiling passed via I915_GEM_SET_TILING.  When exporting these buffers
    * we require a dedicated allocation so that we can know to allocate a
    * tiled buffer.
    */
   bool needs_set_tiling;

   /**
    * Must be DRM_FORMAT_MOD_INVALID unless tiling is
    * VK_IMAGE_TILING_DRM_FORMAT_MODIFIER_EXT.
    */
   uint64_t drm_format_mod;

   VkDeviceSize size;
   uint32_t alignment;

   /* Whether the image is made of several underlying buffer objects rather a
    * single one with different offsets.
    */
   bool disjoint;

   /**
    * Image subsurfaces
    *
    * For each foo, anv_image::planes[x].surface is valid if and only if
    * anv_image::aspects has a x aspect. Refer to anv_image_aspect_to_plane()
    * to figure the number associated with a given aspect.
    *
    * The hardware requires that the depth buffer and stencil buffer be
    * separate surfaces.  From Vulkan's perspective, though, depth and stencil
    * reside in the same VkImage.  To satisfy both the hardware and Vulkan, we
    * allocate the depth and stencil buffers as separate surfaces in the same
    * bo.
    *
    * Memory layout :
    *
    * -----------------------
    * |     surface0        |   /|\
    * -----------------------    |
    * |   shadow surface0   |    |
    * -----------------------    | Plane 0
    * |    aux surface0     |    |
    * -----------------------    |
    * | fast clear colors0  |   \|/
    * -----------------------
    * |     surface1        |   /|\
    * -----------------------    |
    * |   shadow surface1   |    |
    * -----------------------    | Plane 1
    * |    aux surface1     |    |
    * -----------------------    |
    * | fast clear colors1  |   \|/
    * -----------------------
    * |        ...          |
    * |                     |
    * -----------------------
    */
   struct {
      /**
       * Offset of the entire plane (whenever the image is disjoint this is
       * set to 0).
       */
      uint32_t offset;

      VkDeviceSize size;
      uint32_t alignment;

      struct anv_surface surface;

      /**
       * A surface which shadows the main surface and may have different
       * tiling. This is used for sampling using a tiling that isn't supported
       * for other operations.
       */
      struct anv_surface shadow_surface;

      /**
       * For color images, this is the aux usage for this image when not used
       * as a color attachment.
       *
       * For depth/stencil images, this is set to ISL_AUX_USAGE_HIZ if the
       * image has a HiZ buffer.
       */
      enum isl_aux_usage aux_usage;

      struct anv_surface aux_surface;

      /**
       * Offset of the fast clear state (used to compute the
       * fast_clear_state_offset of the following planes).
       */
      uint32_t fast_clear_state_offset;

      /**
       * BO associated with this plane, set when bound.
       */
      struct anv_bo *bo;
      VkDeviceSize bo_offset;

      /**
       * When destroying the image, also free the bo.
       * */
      bool bo_is_owned;
   } planes[3];
};

/* The ordering of this enum is important */
enum anv_fast_clear_type {
   /** Image does not have/support any fast-clear blocks */
   ANV_FAST_CLEAR_NONE = 0,
   /** Image has/supports fast-clear but only to the default value */
   ANV_FAST_CLEAR_DEFAULT_VALUE = 1,
   /** Image has/supports fast-clear with an arbitrary fast-clear value */
   ANV_FAST_CLEAR_ANY = 2,
};

/* Returns the number of auxiliary buffer levels attached to an image. */
static inline uint8_t
anv_image_aux_levels(const struct anv_image * const image,
                     VkImageAspectFlagBits aspect)
{
   uint32_t plane = anv_image_aspect_to_plane(image->aspects, aspect);
   return image->planes[plane].aux_surface.isl.size > 0 ?
          image->planes[plane].aux_surface.isl.levels : 0;
}

/* Returns the number of auxiliary buffer layers attached to an image. */
static inline uint32_t
anv_image_aux_layers(const struct anv_image * const image,
                     VkImageAspectFlagBits aspect,
                     const uint8_t miplevel)
{
   assert(image);

   /* The miplevel must exist in the main buffer. */
   assert(miplevel < image->levels);

   if (miplevel >= anv_image_aux_levels(image, aspect)) {
      /* There are no layers with auxiliary data because the miplevel has no
       * auxiliary data.
       */
      return 0;
   } else {
      uint32_t plane = anv_image_aspect_to_plane(image->aspects, aspect);
      return MAX2(image->planes[plane].aux_surface.isl.logical_level0_px.array_len,
                  image->planes[plane].aux_surface.isl.logical_level0_px.depth >> miplevel);
   }
}

static inline struct anv_address
anv_image_get_clear_color_addr(const struct anv_device *device,
                               const struct anv_image *image,
                               VkImageAspectFlagBits aspect)
{
   assert(image->aspects & VK_IMAGE_ASPECT_ANY_COLOR_BIT_ANV);


context:
space:
mode: