/* * Mesa 3-D graphics library * Version: 6.5.3 * * Copyright (C) 2005-2007 Brian Paul All Rights Reserved. * * 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 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 * BRIAN PAUL 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. */ /** * \file slang_mem.c * * Memory manager for GLSL compiler. The general idea is to do all * allocations out of a large pool then just free the pool when done * compiling to avoid intricate malloc/free tracking and memory leaks. * * \author Brian Paul */ #include "main/context.h" #include "main/macros.h" #include "slang_mem.h" #define GRANULARITY 8 #define ROUND_UP(B) ( ((B) + (GRANULARITY - 1)) & ~(GRANULARITY - 1) ) /** If 1, use conventional malloc/free. Helpful for debugging */ #define USE_MALLOC_FREE 0 struct slang_mempool_ { GLuint Size, Used, Count, Largest; char *Data; struct slang_mempool_ *Next; }; slang_mempool * _slang_new_mempool(GLuint initialSize) { slang_mempool *pool = (slang_mempool *) _mesa_calloc(sizeof(slang_mempool)); if (pool) { pool->Data = (char *) _mesa_calloc(initialSize); /*printf("ALLOC MEMPOOL %d at %p\n", initialSize, pool->Data);*/ if (!pool->Data) { _mesa_free(pool); return NULL; } pool->Size = initialSize; pool->Used = 0; } return pool; } void _slang_delete_mempool(slang_mempool *pool) { GLuint total = 0; while (pool) { slang_mempool *next = pool->Next; /* printf("DELETE MEMPOOL %u / %u count=%u largest=%u\n", pool->Used, pool->Size, pool->Count, pool->Largest); */ total += pool->Used; _mesa_free(pool->Data); _mesa_free(pool); pool = next; } /*printf("TOTAL ALLOCATED: %u\n", total);*/ } #ifdef DEBUG static void check_zero(const char *addr, GLuint n) { GLuint i; for (i = 0; i < n; i++) { assert(addr[i]==0); } } #endif #ifdef DEBUG static GLboolean is_valid_address(const slang_mempool *pool, void *addr) { while (pool) { if ((char *) addr >= pool->Data && (char *) addr < pool->Data + pool->Used) return GL_TRUE; pool = pool->Next; } return GL_FALSE; } #endif /** * Alloc 'bytes' from shader mempool. */ void * _slang_alloc(GLuint bytes) { #if USE_MALLOC_FREE return _mesa_calloc(bytes); #else slang_mempool *pool; GET_CURRENT_CONTEXT(ctx); pool = (slang_mempool *) ctx->Shader.MemPool; if (bytes == 0) bytes = 1; while (pool) { if (pool->Used + bytes <= pool->Size) { /* found room */ void *addr = (void *) (pool->Data + pool->Used); #ifdef DEBUG check_zero((char*) addr, bytes); #endif pool->Used += ROUND_UP(bytes); pool->Largest = MAX2(pool->Largest, bytes); pool->Count++; /*printf("alloc %u Used %u\n", bytes, pool->Used);*/ return addr; } else if (pool->Next) { /* try next block */ pool = pool->Next; } else { /* alloc new pool */ const GLuint sz = MAX2(bytes, pool->Size); pool->Next = _slang_new_mempool(sz); if (!pool->Next) { /* we're _really_ out of memory */ return NULL; } else { pool = pool->Next; pool->Largest = bytes; pool->Count++; pool->Used = ROUND_UP(bytes); #ifdef DEBUG check_zero((char*) pool->Data, bytes); #endif return (void *) pool->Data; } } } return NULL; #endif } void * _slang_realloc(void *oldBuffer, GLuint oldSize, GLuint newSize) { #if USE_MALLOC_FREE return _mesa_realloc(oldBuffer, oldSize, newSize); #else GET_CURRENT_CONTEXT(ctx); slang_mempool *pool = (slang_mempool *) ctx->Shader.MemPool; (void) pool; if (newSize < oldSize) { return oldBuffer; } else { const GLuint copySize = (oldSize < newSize) ? oldSize : newSize; void *newBuffer = _slang_alloc(newSize); if (oldBuffer) ASSERT(is_valid_address(pool, oldBuffer)); if (newBuffer && oldBuffer && copySize > 0) _mesa_memcpy(newBuffer, oldBuffer, copySize); return newBuffer; } #endif } /** * Clone string, storing in current mempool. */ char * _slang_strdup(const char *s) { if (s) { size_t l = _mesa_strlen(s); char *s2 = (char *) _slang_alloc(l + 1); if (s2) _mesa_strcpy(s2, s); return s2; } else { return NULL; } } /** * Don't actually free memory, but mark it (for debugging). */ void _slang_free(void *addr) { #if USE_MALLOC_FREE _mesa_free(addr); #else if (addr) { GET_CURRENT_CONTEXT(ctx); slang_mempool *pool = (slang_mempool *) ctx->Shader.MemPool; (void) pool; ASSERT(is_valid_address(pool, addr)); } #endif } 'n73' href='#n73'>73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151 1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199 1200 1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 1230 1231 1232 1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 1251 1252 1253 1254 1255 1256 1257 1258 1259 1260 1261 1262 1263 1264 1265 1266 1267 1268 1269 1270 1271 1272 1273 1274 1275 1276 1277 1278 1279 1280 1281 1282 1283 1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 1299 1300 1301 1302 1303 1304 1305 1306 1307 1308 1309 1310 1311 1312 1313 1314 1315 1316 1317 1318 1319 1320 1321 1322 1323 1324 1325 1326 1327 1328 1329 1330 1331 1332 1333 1334 1335 1336 1337 1338 1339 1340 1341 1342 1343 1344 1345 1346 1347 1348 1349 1350 1351 1352 1353 1354 1355 1356 1357 1358 1359 1360 1361 1362 1363 1364 1365 1366 1367 1368 1369 1370 1371 1372 1373 1374 1375 1376 1377 1378 1379 1380 1381 1382 1383 1384 1385 1386 1387 1388 1389 1390 1391 1392 1393 1394 1395 1396 1397 1398 1399 1400 1401 1402 1403 1404 1405 1406 1407 1408 1409 1410 1411 1412 1413 1414 1415 1416 1417 1418 1419 1420 1421 1422 1423 1424 1425 1426 1427 1428 1429 1430 1431 1432 1433 1434 1435 1436 1437 1438 1439 1440 1441 1442 1443 1444 1445 1446 1447 1448 1449 1450 1451 1452 1453 1454 1455 1456 1457 1458 1459 1460 1461 1462 1463 1464 1465 1466 1467 1468 1469 1470 1471 1472 1473 1474 1475 1476 1477 1478 1479 1480 1481 1482 1483 1484 1485 1486 1487 1488 1489 1490 1491 1492 1493 1494 1495 1496 1497 1498 1499 1500 1501 1502 1503 1504 1505 1506 1507 1508 1509 1510 1511 1512 1513 1514 1515 1516 1517 1518 1519 1520 1521 1522 1523 1524 1525 1526 1527 1528 1529 1530 1531 1532 1533 1534 1535 1536 1537 1538 1539 1540 1541 1542 1543 1544 1545 1546 1547 1548 1549 1550 1551 1552 1553 1554 1555 1556 1557 1558 1559 1560 1561 1562 1563 1564 1565 1566 1567 1568 1569 1570 1571 1572 1573 1574 1575 1576 1577 1578 1579 1580 1581 1582 1583 1584 1585 1586 1587 1588 1589 1590 1591 1592 1593 1594 1595 1596 1597 1598 1599 1600 1601 1602 1603 1604 1605 1606 1607 1608 1609 1610 1611 1612 1613 1614 1615 1616 1617 1618 1619 1620 1621 1622 1623 1624 1625 1626 1627 1628 1629 1630 1631 1632 1633 1634 1635 1636 1637 1638 1639 1640 1641 1642 1643 1644 1645 1646 1647 1648 1649 1650 1651 1652 1653 1654 1655 1656 1657 1658 1659 1660 1661 1662 1663 1664 1665 1666 1667 1668 1669 1670 1671 1672 1673 1674 1675 1676 1677 1678 1679 1680 1681 1682 1683 1684 1685 1686 1687 1688 1689 1690 1691 1692 1693 1694 1695 1696 1697 1698 1699 1700 1701 1702 1703 1704 1705 1706 1707 1708 1709 1710 1711 1712 1713 1714 1715 1716 1717 1718 1719 1720 1721 1722 1723 1724 1725 1726 1727 1728 1729 1730 1731 1732 1733 1734 1735 1736 1737 1738 1739 1740 1741 1742 1743 1744 1745 1746 1747 1748 1749 1750 1751 1752 1753 1754 1755 1756 1757
/*
 * Mesa 3-D graphics library
 * Version:  5.1
 *
 * Copyright (C) 1999-2003  Brian Paul   All Rights Reserved.
 *
 * 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 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
 * BRIAN PAUL 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.
 */

/* Authors:
 *    David Bucciarelli
 *    Brian Paul
 *    Daryll Strauss
 *    Keith Whitwell
 *    Daniel Borca
 *    Hiroshi Morii
 */

/* fxdd.c - 3Dfx VooDoo Mesa device driver functions */


#ifdef HAVE_CONFIG_H
#include "conf.h"
#endif

#if defined(FX)

#include "image.h"
#include "mtypes.h"
#include "fxdrv.h"
#include "enums.h"
#include "extensions.h"
#include "macros.h"
#include "texstore.h"
#include "teximage.h"
#include "swrast/swrast.h"
#include "swrast/s_context.h"
#include "swrast_setup/swrast_setup.h"
#include "tnl/tnl.h"
#include "tnl/t_context.h"
#include "tnl/t_pipeline.h"
#include "array_cache/acache.h"



/* lookup table for scaling 4 bit colors up to 8 bits */
GLuint FX_rgb_scale_4[16] = {
   0,   17,  34,  51,  68,  85,  102, 119,
   136, 153, 170, 187, 204, 221, 238, 255
};

/* lookup table for scaling 5 bit colors up to 8 bits */
GLuint FX_rgb_scale_5[32] = {
   0,   8,   16,  25,  33,  41,  49,  58,
   66,  74,  82,  90,  99,  107, 115, 123,
   132, 140, 148, 156, 165, 173, 181, 189,
   197, 206, 214, 222, 230, 239, 247, 255
};

/* lookup table for scaling 6 bit colors up to 8 bits */
GLuint FX_rgb_scale_6[64] = {
   0,   4,   8,   12,  16,  20,  24,  28,
   32,  36,  40,  45,  49,  53,  57,  61,
   65,  69,  73,  77,  81,  85,  89,  93,
   97,  101, 105, 109, 113, 117, 121, 125,
   130, 134, 138, 142, 146, 150, 154, 158,
   162, 166, 170, 174, 178, 182, 186, 190,
   194, 198, 202, 206, 210, 215, 219, 223,
   227, 231, 235, 239, 243, 247, 251, 255
};


/*
 * Disable color by masking out R, G, B, A
 */
static void fxDisableColor (fxMesaContext fxMesa)
{
 if (fxMesa->colDepth != 16) {
    /* 32bpp mode or 15bpp mode */
    fxMesa->Glide.grColorMaskExt(FXFALSE, FXFALSE, FXFALSE, FXFALSE);
 } else {
    /* 16 bpp mode */
    grColorMask(FXFALSE, FXFALSE);
 }
}


/**********************************************************************/
/*****                 Miscellaneous functions                    *****/
/**********************************************************************/

/* Return buffer size information */
static void
fxDDBufferSize(GLframebuffer *buffer, GLuint *width, GLuint *height)
{
   GET_CURRENT_CONTEXT(ctx);
   if (ctx && FX_CONTEXT(ctx)) {
      fxMesaContext fxMesa = FX_CONTEXT(ctx);

      if (TDFX_DEBUG & VERBOSE_DRIVER) {
         fprintf(stderr, "fxDDBufferSize(...)\n");
      }

      *width = fxMesa->width;
      *height = fxMesa->height;
   }
}


/* Implements glClearColor() */
static void
fxDDClearColor(GLcontext * ctx, const GLfloat color[4])
{
   fxMesaContext fxMesa = FX_CONTEXT(ctx);
   GLubyte col[4];

   if (TDFX_DEBUG & VERBOSE_DRIVER) {
      fprintf(stderr, "fxDDClearColor(%f, %f, %f, %f)\n",
	              color[0], color[1], color[2], color[3]);
   }

   CLAMPED_FLOAT_TO_UBYTE(col[0], color[0]);
   CLAMPED_FLOAT_TO_UBYTE(col[1], color[1]);
   CLAMPED_FLOAT_TO_UBYTE(col[2], color[2]);
   CLAMPED_FLOAT_TO_UBYTE(col[3], color[3]);

   fxMesa->clearC = FXCOLOR4(col);
   fxMesa->clearA = col[3];
}


/* Clear the color and/or depth buffers */
static void fxDDClear( GLcontext *ctx,
			GLbitfield mask, GLboolean all,
			GLint x, GLint y, GLint width, GLint height )
{
   fxMesaContext fxMesa = FX_CONTEXT(ctx);
   GLbitfield softwareMask = mask & (DD_ACCUM_BIT);
   const GLuint stencil_size = fxMesa->haveHwStencil ? ctx->Visual.stencilBits : 0;
   const FxU32 clearD = (FxU32) (((1 << ctx->Visual.depthBits) - 1) * ctx->Depth.Clear);
   const FxU8 clearS = (FxU8) (ctx->Stencil.Clear & 0xff);

   /* [dBorca] Hack alert:
    * if we set Mesa for 32bit depth, we'll get
    * clearD == 0
    * due to 32bit integer overflow!
    */

   if ( TDFX_DEBUG & MESA_VERBOSE ) {
      fprintf( stderr, "fxDDClear( %d, %d, %d, %d )\n",
	               (int) x, (int) y, (int) width, (int) height );
   }

   /* Need this check to respond to glScissor and clipping updates */
   /* should also take care of FX_NEW_COLOR_MASK, FX_NEW_STENCIL, depth? */
   if (fxMesa->new_state & FX_NEW_SCISSOR) {
      fxSetupScissor(ctx);
      fxMesa->new_state &= ~FX_NEW_SCISSOR;
   }

   /* we can't clear accum buffers */
   mask &= ~(DD_ACCUM_BIT);

   /*
    * As per GL spec, stencil masking should be obeyed when clearing
    */
   if (mask & DD_STENCIL_BIT) {
      if (!fxMesa->haveHwStencil || fxMesa->unitsState.stencilWriteMask != 0xff) {
         /* Napalm seems to have trouble with stencil write masks != 0xff */
         /* do stencil clear in software */
         softwareMask |= DD_STENCIL_BIT;
         mask &= ~(DD_STENCIL_BIT);
      }
   }

   /*
    * As per GL spec, color masking should be obeyed when clearing
    */
   if (ctx->Visual.greenBits != 8 && ctx->Visual.greenBits != 5) {
      /* can only do color masking if running in 24/32bpp on Napalm */
      if (ctx->Color.ColorMask[RCOMP] != ctx->Color.ColorMask[GCOMP] ||
          ctx->Color.ColorMask[GCOMP] != ctx->Color.ColorMask[BCOMP]) {
         softwareMask |= (mask & (DD_FRONT_LEFT_BIT | DD_BACK_LEFT_BIT));
         mask &= ~(DD_FRONT_LEFT_BIT | DD_BACK_LEFT_BIT);
      }
   }

   if (fxMesa->haveHwStencil) {
      /*
       * If we want to clear stencil, it must be enabled
       * in the HW, even if the stencil test is not enabled
       * in the OGL state.
       */
      BEGIN_BOARD_LOCK();
      if (mask & DD_STENCIL_BIT) {
	 fxMesa->Glide.grStencilMaskExt(0xff /*fxMesa->unitsState.stencilWriteMask*/);
	 /* set stencil ref value = desired clear value */
	 fxMesa->Glide.grStencilFuncExt(GR_CMP_ALWAYS, clearS, 0xff);
	 fxMesa->Glide.grStencilOpExt(GR_STENCILOP_REPLACE,
                                   GR_STENCILOP_REPLACE, GR_STENCILOP_REPLACE);
	 grEnable(GR_STENCIL_MODE_EXT);
      }
      else {
	 grDisable(GR_STENCIL_MODE_EXT);
      }
      END_BOARD_LOCK();
   }

   /*
    * This may be ugly, but it's needed in order to work around a number
    * of Glide bugs.
    */
   BEGIN_CLIP_LOOP();
   {
      /*
       * This could probably be done fancier but doing each possible case
       * explicitly is less error prone.
       */
      switch (mask & ~DD_STENCIL_BIT) {
      case DD_BACK_LEFT_BIT | DD_DEPTH_BIT:
	 /* back buffer & depth */
	 /* FX_grColorMaskv_NoLock(ctx, true4); */ /* work around Voodoo3 bug */
	 grDepthMask(FXTRUE);
	 grRenderBuffer(GR_BUFFER_BACKBUFFER);
	 if (stencil_size > 0) {
            fxMesa->Glide.grBufferClearExt(fxMesa->clearC,
                                           fxMesa->clearA,
                                           clearD, clearS);
         }
	 else
            grBufferClear(fxMesa->clearC,
                          fxMesa->clearA,
                          clearD);
	 if (!fxMesa->unitsState.depthTestEnabled) {
            grDepthMask(FXFALSE);
	 }
	 break;
      case DD_FRONT_LEFT_BIT | DD_DEPTH_BIT:
	 /* XXX it appears that the depth buffer isn't cleared when
	  * glRenderBuffer(GR_BUFFER_FRONTBUFFER) is set.
	  * This is a work-around/
	  */
	 /* clear depth */
	 grDepthMask(FXTRUE);
	 grRenderBuffer(GR_BUFFER_BACKBUFFER);
         fxDisableColor(fxMesa);
	 if (stencil_size > 0)
            fxMesa->Glide.grBufferClearExt(fxMesa->clearC,
                                           fxMesa->clearA,
                                           clearD, clearS);
	 else
            grBufferClear(fxMesa->clearC,
                          fxMesa->clearA,
                          clearD);
	 /* clear front */
	 fxSetupColorMask(ctx);
	 grRenderBuffer(GR_BUFFER_FRONTBUFFER);
	 if (stencil_size > 0)
            fxMesa->Glide.grBufferClearExt(fxMesa->clearC,
                                           fxMesa->clearA,
                                           clearD, clearS);
	 else
            grBufferClear(fxMesa->clearC,
                          fxMesa->clearA,
                          clearD);
	 if (!fxMesa->unitsState.depthTestEnabled) {
            grDepthMask(FXFALSE);
	 }
	 break;
      case DD_BACK_LEFT_BIT:
	 /* back buffer only */
	 grDepthMask(FXFALSE);
	 grRenderBuffer(GR_BUFFER_BACKBUFFER);
	 if (stencil_size > 0)
            fxMesa->Glide.grBufferClearExt(fxMesa->clearC,
                                           fxMesa->clearA,
                                           clearD, clearS);
	 else
            grBufferClear(fxMesa->clearC,
                          fxMesa->clearA,
                          clearD);
	 if (fxMesa->unitsState.depthTestEnabled) {
            grDepthMask(FXTRUE);
	 }
	 break;
      case DD_FRONT_LEFT_BIT:
	 /* front buffer only */
	 grDepthMask(FXFALSE);
	 grRenderBuffer(GR_BUFFER_FRONTBUFFER);
	 if (stencil_size > 0)
            fxMesa->Glide.grBufferClearExt(fxMesa->clearC,
                                           fxMesa->clearA,
                                           clearD, clearS);
	 else
            grBufferClear(fxMesa->clearC,
                          fxMesa->clearA,
                          clearD);
	 if (fxMesa->unitsState.depthTestEnabled) {
            grDepthMask(FXTRUE);
	 }
	 break;
      case DD_FRONT_LEFT_BIT | DD_BACK_LEFT_BIT:
	 /* front and back */
	 grDepthMask(FXFALSE);
	 grRenderBuffer(GR_BUFFER_BACKBUFFER);
	 if (stencil_size > 0)
            fxMesa->Glide.grBufferClearExt(fxMesa->clearC,
                                           fxMesa->clearA,
                                           clearD, clearS);
	 else
            grBufferClear(fxMesa->clearC,
                          fxMesa->clearA,
                          clearD);
	 grRenderBuffer(GR_BUFFER_FRONTBUFFER);
	 if (stencil_size > 0)
            fxMesa->Glide.grBufferClearExt(fxMesa->clearC,
                                           fxMesa->clearA,
                                           clearD, clearS);
	 else
            grBufferClear(fxMesa->clearC,
                          fxMesa->clearA,
                          clearD);
	 if (fxMesa->unitsState.depthTestEnabled) {
            grDepthMask(FXTRUE);
	 }
	 break;
      case DD_FRONT_LEFT_BIT | DD_BACK_LEFT_BIT | DD_DEPTH_BIT:
	 /* clear front */
	 grDepthMask(FXFALSE);
	 grRenderBuffer(GR_BUFFER_FRONTBUFFER);
	 if (stencil_size > 0)
            fxMesa->Glide.grBufferClearExt(fxMesa->clearC,
                                           fxMesa->clearA,
                                           clearD, clearS);
	 else
            grBufferClear(fxMesa->clearC,
                          fxMesa->clearA,
                          clearD);
	 /* clear back and depth */
	 grDepthMask(FXTRUE);
	 grRenderBuffer(GR_BUFFER_BACKBUFFER);
         if (stencil_size > 0)
            fxMesa->Glide.grBufferClearExt(fxMesa->clearC,
                                           fxMesa->clearA,
                                           clearD, clearS);
	 else
            grBufferClear(fxMesa->clearC,
                          fxMesa->clearA,
                          clearD);
	 if (!fxMesa->unitsState.depthTestEnabled) {
            grDepthMask(FXFALSE);
	 }
	 break;
      case DD_DEPTH_BIT:
	 /* just the depth buffer */
	 grRenderBuffer(GR_BUFFER_BACKBUFFER);
         fxDisableColor(fxMesa);
	 grDepthMask(FXTRUE);
	 if (stencil_size > 0)
            fxMesa->Glide.grBufferClearExt(fxMesa->clearC,
                                           fxMesa->clearA,
                                           clearD, clearS);
	 else
            grBufferClear(fxMesa->clearC,
                          fxMesa->clearA,
                          clearD);
	 fxSetupColorMask(ctx);
	 if (ctx->Color._DrawDestMask & FRONT_LEFT_BIT) {
            grRenderBuffer(GR_BUFFER_FRONTBUFFER);
         }
	 if (!fxMesa->unitsState.depthTestEnabled) {
	    grDepthMask(FXFALSE);
         }
	 break;
      default:
         /* clear no color buffers or depth buffer but might clear stencil */
	 if (stencil_size > 0 && (mask & DD_STENCIL_BIT)) {
            /* XXX need this RenderBuffer call to work around Glide bug */
            grRenderBuffer(GR_BUFFER_BACKBUFFER);
            grDepthMask(FXFALSE);
            fxDisableColor(fxMesa);
            fxMesa->Glide.grBufferClearExt(fxMesa->clearC,
                                           fxMesa->clearA,
                                           clearD, clearS);
            if (fxMesa->unitsState.depthTestEnabled) {
               grDepthMask(FXTRUE);
            }
            fxSetupColorMask(ctx);
            if (ctx->Color._DrawDestMask & FRONT_LEFT_BIT) {
               grRenderBuffer(GR_BUFFER_FRONTBUFFER);
            }
         }
      }
   }
   END_CLIP_LOOP();

   if (fxMesa->haveHwStencil && (mask & DD_STENCIL_BIT)) {
      /* We changed the stencil state above.  Signal that we need to
       * upload it again.
       */
      fxMesa->new_state |= FX_NEW_STENCIL;
   }

   if (softwareMask)
      _swrast_Clear( ctx, softwareMask, all, x, y, width, height );
}


/* Set the buffer used for drawing */
/* XXX support for separate read/draw buffers hasn't been tested */
static void
fxDDSetDrawBuffer(GLcontext * ctx, GLenum mode)
{
   fxMesaContext fxMesa = FX_CONTEXT(ctx);

   if (TDFX_DEBUG & VERBOSE_DRIVER) {
      fprintf(stderr, "fxDDSetDrawBuffer(%x)\n", (int)mode);
   }

   if (mode == GL_FRONT_LEFT) {
      fxMesa->currentFB = GR_BUFFER_FRONTBUFFER;
      grRenderBuffer(fxMesa->currentFB);
   }
   else if (mode == GL_BACK_LEFT) {
      fxMesa->currentFB = GR_BUFFER_BACKBUFFER;
      grRenderBuffer(fxMesa->currentFB);
   }
   else if (mode == GL_NONE) {
      fxDisableColor(fxMesa);
   }
   else {
      /* we'll need a software fallback */
      /* XXX not implemented */
   }

   /* update s/w fallback state */
   _swrast_DrawBuffer(ctx, mode);
}


static void
fxDDDrawBitmap2 (GLcontext *ctx, GLint px, GLint py,
		 GLsizei width, GLsizei height,
		 const struct gl_pixelstore_attrib *unpack,
		 const GLubyte *bitmap)
{
   fxMesaContext fxMesa = FX_CONTEXT(ctx);
   SWcontext *swrast = SWRAST_CONTEXT(ctx);
   GrLfbInfo_t info;
   GrLfbWriteMode_t mode;
   FxU16 color;
   const struct gl_pixelstore_attrib *finalUnpack;
   struct gl_pixelstore_attrib scissoredUnpack;

   /* check if there's any raster operations enabled which we can't handle */
   if ((swrast->_RasterMask & (ALPHATEST_BIT |
			      /*BLEND_BIT |*/   /* blending ok, through pixpipe */
			      DEPTH_BIT |       /* could be done with RGB:DEPTH */
			      FOG_BIT |         /* could be done with RGB:DEPTH */
			      LOGIC_OP_BIT |
			      /*CLIP_BIT |*/    /* clipping ok, below */
			      STENCIL_BIT |
			      /*MASKING_BIT |*/ /* masking ok, test follows */
			      ALPHABUF_BIT |    /* nope! see 565 span kludge */
			      MULTI_DRAW_BIT |
			      OCCLUSION_BIT |   /* nope! at least not yet */
			      TEXTURE_BIT |
			      FRAGPROG_BIT))
       ||
       ((swrast->_RasterMask & MASKING_BIT) /*&& (ctx->Visual.greenBits != 8)*/ && (ctx->Visual.greenBits != 5))
      ) {
      _swrast_Bitmap(ctx, px, py, width, height, unpack, bitmap);
      return;
   }

   /* make sure the pixelpipe is configured correctly */
   fxSetupFXUnits(ctx);

   if (ctx->Scissor.Enabled) {
      /* This is a bit tricky, but by carefully adjusting the px, py,
       * width, height, skipPixels and skipRows values we can do
       * scissoring without special code in the rendering loop.
       */

      /* we'll construct a new pixelstore struct */
      finalUnpack = &scissoredUnpack;
      scissoredUnpack = *unpack;
      if (scissoredUnpack.RowLength == 0)
	 scissoredUnpack.RowLength = width;

      /* clip left */
      if (px < ctx->Scissor.X) {
	 scissoredUnpack.SkipPixels += (ctx->Scissor.X - px);
	 width -= (ctx->Scissor.X - px);
	 px = ctx->Scissor.X;
      }
      /* clip right */
      if (px + width >= ctx->Scissor.X + ctx->Scissor.Width) {
	 width -= (px + width - (ctx->Scissor.X + ctx->Scissor.Width));
      }
      /* clip bottom */
      if (py < ctx->Scissor.Y) {
	 scissoredUnpack.SkipRows += (ctx->Scissor.Y - py);
	 height -= (ctx->Scissor.Y - py);
	 py = ctx->Scissor.Y;
      }
      /* clip top */
      if (py + height >= ctx->Scissor.Y + ctx->Scissor.Height) {
	 height -= (py + height - (ctx->Scissor.Y + ctx->Scissor.Height));
      }

      if (width <= 0 || height <= 0)
	 return;
   }
   else {
      finalUnpack = unpack;
   }

   /* compute pixel value */
   {
      GLint r = (GLint) (ctx->Current.RasterColor[RCOMP] * 255.0f);
      GLint g = (GLint) (ctx->Current.RasterColor[GCOMP] * 255.0f);
      GLint b = (GLint) (ctx->Current.RasterColor[BCOMP] * 255.0f);
      GLint a = (GLint) (ctx->Current.RasterColor[ACOMP] * 255.0f);
      if (fxMesa->colDepth == 15) {
         color = TDFXPACKCOLOR1555(b, g, r, a);
         mode = GR_LFBWRITEMODE_1555;
      } else {
         color = fxMesa->bgrOrder ? TDFXPACKCOLOR565(r, g, b) : TDFXPACKCOLOR565(b, g, r);
         mode = GR_LFBWRITEMODE_565;
      }
   }

   info.size = sizeof(info);
   if (!grLfbLock(GR_LFB_WRITE_ONLY,
		  fxMesa->currentFB,
		  mode,
		  GR_ORIGIN_UPPER_LEFT, FXTRUE, &info)) {
      _swrast_Bitmap(ctx, px, py, width, height, unpack, bitmap);
      return;
   }

   {
      const GLint winX = 0;
      const GLint winY = fxMesa->height - 1;
      /* The dest stride depends on the hardware and whether we're drawing
       * to the front or back buffer.  This compile-time test seems to do
       * the job for now.
       */
      const GLint dstStride = info.strideInBytes / 2;	/* stride in GLushorts */

      GLint row;
      /* compute dest address of bottom-left pixel in bitmap */
      GLushort *dst = (GLushort *) info.lfbPtr
	 + (winY - py) * dstStride + (winX + px);

      for (row = 0; row < height; row++) {
	 const GLubyte *src =
	    (const GLubyte *) _mesa_image_address(finalUnpack,
						  bitmap, width, height,
						  GL_COLOR_INDEX, GL_BITMAP,
						  0, row, 0);
	 if (finalUnpack->LsbFirst) {
	    /* least significan bit first */
	    GLubyte mask = 1U << (finalUnpack->SkipPixels & 0x7);
	    GLint col;
	    for (col = 0; col < width; col++) {
	       if (*src & mask) {
		  dst[col] = color;
	       }
	       if (mask == 128U) {
		  src++;
		  mask = 1U;
	       }
	       else {
		  mask = mask << 1;
	       }
	    }
	    if (mask != 1)
	       src++;
	 }
	 else {
	    /* most significan bit first */
	    GLubyte mask = 128U >> (finalUnpack->SkipPixels & 0x7);
	    GLint col;
	    for (col = 0; col < width; col++) {
	       if (*src & mask) {
		  dst[col] = color;
	       }
	       if (mask == 1U) {
		  src++;
		  mask = 128U;
	       }
	       else {
		  mask = mask >> 1;
	       }
	    }
	    if (mask != 128)
	       src++;
	 }
	 dst -= dstStride;
      }
   }

   grLfbUnlock(GR_LFB_WRITE_ONLY, fxMesa->currentFB);
}

static void
fxDDDrawBitmap4 (GLcontext *ctx, GLint px, GLint py,
		 GLsizei width, GLsizei height,
		 const struct gl_pixelstore_attrib *unpack,
		 const GLubyte *bitmap)
{
   fxMesaContext fxMesa = FX_CONTEXT(ctx);
   SWcontext *swrast = SWRAST_CONTEXT(ctx);
   GrLfbInfo_t info;
   FxU32 color;
   const struct gl_pixelstore_attrib *finalUnpack;
   struct gl_pixelstore_attrib scissoredUnpack;

   /* check if there's any raster operations enabled which we can't handle */
   if ((swrast->_RasterMask & (ALPHATEST_BIT |
			      /*BLEND_BIT |*/   /* blending ok, through pixpipe */
			      DEPTH_BIT |       /* could be done with RGB:DEPTH */
			      FOG_BIT |         /* could be done with RGB:DEPTH */
			      LOGIC_OP_BIT |
			      /*CLIP_BIT |*/    /* clipping ok, below */
			      STENCIL_BIT |
			      /*MASKING_BIT |*/ /* masking ok, we're in 32bpp */
			      /*ALPHABUF_BIT |*//* alpha ok, we're in 32bpp */
			      MULTI_DRAW_BIT |
			      OCCLUSION_BIT |   /* nope! at least not yet */
			      TEXTURE_BIT |
			      FRAGPROG_BIT))
      ) {
      _swrast_Bitmap(ctx, px, py, width, height, unpack, bitmap);
      return;
   }

   /* make sure the pixelpipe is configured correctly */
   fxSetupFXUnits(ctx);

   if (ctx->Scissor.Enabled) {
      /* This is a bit tricky, but by carefully adjusting the px, py,
       * width, height, skipPixels and skipRows values we can do
       * scissoring without special code in the rendering loop.
       */

      /* we'll construct a new pixelstore struct */
      finalUnpack = &scissoredUnpack;
      scissoredUnpack = *unpack;
      if (scissoredUnpack.RowLength == 0)
	 scissoredUnpack.RowLength = width;

      /* clip left */
      if (px < ctx->Scissor.X) {
	 scissoredUnpack.SkipPixels += (ctx->Scissor.X - px);
	 width -= (ctx->Scissor.X - px);
	 px = ctx->Scissor.X;
      }
      /* clip right */
      if (px + width >= ctx->Scissor.X + ctx->Scissor.Width) {
	 width -= (px + width - (ctx->Scissor.X + ctx->Scissor.Width));
      }
      /* clip bottom */
      if (py < ctx->Scissor.Y) {
	 scissoredUnpack.SkipRows += (ctx->Scissor.Y - py);
	 height -= (ctx->Scissor.Y - py);
	 py = ctx->Scissor.Y;
      }
      /* clip top */
      if (py + height >= ctx->Scissor.Y + ctx->Scissor.Height) {
	 height -= (py + height - (ctx->Scissor.Y + ctx->Scissor.Height));
      }

      if (width <= 0 || height <= 0)
	 return;
   }
   else {
      finalUnpack = unpack;
   }

   /* compute pixel value */
   {
      GLint r = (GLint) (ctx->Current.RasterColor[RCOMP] * 255.0f);
      GLint g = (GLint) (ctx->Current.RasterColor[GCOMP] * 255.0f);
      GLint b = (GLint) (ctx->Current.RasterColor[BCOMP] * 255.0f);
      GLint a = (GLint) (ctx->Current.RasterColor[ACOMP] * 255.0f);
      color = TDFXPACKCOLOR8888(b, g, r, a);
   }

   info.size = sizeof(info);
   if (!grLfbLock(GR_LFB_WRITE_ONLY,
		  fxMesa->currentFB,
		  GR_LFBWRITEMODE_8888,
		  GR_ORIGIN_UPPER_LEFT, FXTRUE, &info)) {
      _swrast_Bitmap(ctx, px, py, width, height, unpack, bitmap);
      return;
   }

   {
      const GLint winX = 0;
      const GLint winY = fxMesa->height - 1;
      /* The dest stride depends on the hardware and whether we're drawing
       * to the front or back buffer.  This compile-time test seems to do
       * the job for now.
       */
      const GLint dstStride = info.strideInBytes / 4;	/* stride in GLuints */

      GLint row;
      /* compute dest address of bottom-left pixel in bitmap */
      GLuint *dst = (GLuint *) info.lfbPtr
	 + (winY - py) * dstStride + (winX + px);

      for (row = 0; row < height; row++) {
	 const GLubyte *src =
	    (const GLubyte *) _mesa_image_address(finalUnpack,
						  bitmap, width, height,
						  GL_COLOR_INDEX, GL_BITMAP,
						  0, row, 0);
	 if (finalUnpack->LsbFirst) {
	    /* least significan bit first */
	    GLubyte mask = 1U << (finalUnpack->SkipPixels & 0x7);
	    GLint col;
	    for (col = 0; col < width; col++) {
	       if (*src & mask) {
		  dst[col] = color;
	       }
	       if (mask == 128U) {
		  src++;
		  mask = 1U;
	       }
	       else {
		  mask = mask << 1;
	       }
	    }
	    if (mask != 1)
	       src++;
	 }
	 else {
	    /* most significan bit first */
	    GLubyte mask = 128U >> (finalUnpack->SkipPixels & 0x7);
	    GLint col;
	    for (col = 0; col < width; col++) {
	       if (*src & mask) {
		  dst[col] = color;
	       }
	       if (mask == 1U) {
		  src++;
		  mask = 128U;
	       }
	       else {
		  mask = mask >> 1;
	       }
	    }
	    if (mask != 128)
	       src++;
	 }
	 dst -= dstStride;
      }
   }

   grLfbUnlock(GR_LFB_WRITE_ONLY, fxMesa->currentFB);
}


static void
fxDDReadPixels565 (GLcontext * ctx,
		   GLint x, GLint y,
		   GLsizei width, GLsizei height,
		   GLenum format, GLenum type,
		   const struct gl_pixelstore_attrib *packing,
		   GLvoid *dstImage)
{
   if (ctx->_ImageTransferState/* & (IMAGE_SCALE_BIAS_BIT|IMAGE_MAP_COLOR_BIT)*/) {
      _swrast_ReadPixels(ctx, x, y, width, height, format, type,
			 packing, dstImage);
      return;
   }
   else {
      fxMesaContext fxMesa = FX_CONTEXT(ctx);
      GrLfbInfo_t info;

      BEGIN_BOARD_LOCK();
      info.size = sizeof(info);
      if (grLfbLock(GR_LFB_READ_ONLY,
		    fxMesa->currentFB,
		    GR_LFBWRITEMODE_ANY,
		    GR_ORIGIN_UPPER_LEFT, FXFALSE, &info)) {
	 const GLint winX = 0;
	 const GLint winY = fxMesa->height - 1;
	 const GLint srcStride = info.strideInBytes / 2;	/* stride in GLushorts */
	 const GLushort *src = (const GLushort *) info.lfbPtr
	    + (winY - y) * srcStride + (winX + x);
	 GLubyte *dst = (GLubyte *) _mesa_image_address(packing, dstImage,
							width, height, format,
							type, 0, 0, 0);
	 GLint dstStride =
	    _mesa_image_row_stride(packing, width, format, type);

	 if (format == GL_RGB && type == GL_UNSIGNED_BYTE) {
	    /* convert 5R6G5B into 8R8G8B */
	    GLint row, col;
	    const GLint halfWidth = width >> 1;
	    const GLint extraPixel = (width & 1);
	    for (row = 0; row < height; row++) {
	       GLubyte *d = dst;
	       for (col = 0; col < halfWidth; col++) {
		  const GLuint pixel = ((const GLuint *) src)[col];
                  *d++ = FX_rgb_scale_5[(pixel >> 11) & 0x1f];
                  *d++ = FX_rgb_scale_6[(pixel >> 5)  & 0x3f];
                  *d++ = FX_rgb_scale_5[ pixel        & 0x1f];
                  *d++ = FX_rgb_scale_5[(pixel >> 27) & 0x1f];
                  *d++ = FX_rgb_scale_6[(pixel >> 21) & 0x3f];
                  *d++ = FX_rgb_scale_5[(pixel >> 16) & 0x1f];
	       }
	       if (extraPixel) {
		  GLushort pixel = src[width - 1];
                  *d++ = FX_rgb_scale_5[(pixel >> 11) & 0x1f];
                  *d++ = FX_rgb_scale_6[(pixel >> 5)  & 0x3f];
                  *d++ = FX_rgb_scale_5[ pixel        & 0x1f];
	       }
	       dst += dstStride;
	       src -= srcStride;
	    }
	 }
	 else if (format == GL_RGBA && type == GL_UNSIGNED_BYTE) {
	    /* convert 5R6G5B into 8R8G8B8A */
	    GLint row, col;
	    const GLint halfWidth = width >> 1;
	    const GLint extraPixel = (width & 1);
	    for (row = 0; row < height; row++) {
	       GLubyte *d = dst;
	       for (col = 0; col < halfWidth; col++) {
		  const GLuint pixel = ((const GLuint *) src)[col];
                  *d++ = FX_rgb_scale_5[(pixel >> 11) & 0x1f];
                  *d++ = FX_rgb_scale_6[(pixel >> 5)  & 0x3f];
                  *d++ = FX_rgb_scale_5[ pixel        & 0x1f];
		  *d++ = 255;
                  *d++ = FX_rgb_scale_5[(pixel >> 27) & 0x1f];
                  *d++ = FX_rgb_scale_6[(pixel >> 21) & 0x3f];
                  *d++ = FX_rgb_scale_5[(pixel >> 16) & 0x1f];
		  *d++ = 255;
	       }
	       if (extraPixel) {
		  const GLushort pixel = src[width - 1];
                  *d++ = FX_rgb_scale_5[(pixel >> 11) & 0x1f];
                  *d++ = FX_rgb_scale_6[(pixel >> 5)  & 0x3f];
                  *d++ = FX_rgb_scale_5[ pixel        & 0x1f];
		  *d++ = 255;
	       }
	       dst += dstStride;
	       src -= srcStride;
	    }
	 }
	 else if (format == GL_RGB && type == GL_UNSIGNED_SHORT_5_6_5) {
	    /* directly memcpy 5R6G5B pixels into client's buffer */
	    const GLint widthInBytes = width * 2;
	    GLint row;
	    for (row = 0; row < height; row++) {
	       MEMCPY(dst, src, widthInBytes);
	       dst += dstStride;
	       src -= srcStride;
	    }
	 }
	 else {
	    grLfbUnlock(GR_LFB_READ_ONLY, fxMesa->currentFB);
	    END_BOARD_LOCK();
	    _swrast_ReadPixels(ctx, x, y, width, height, format, type,
			       packing, dstImage);
	    return;
	 }

	 grLfbUnlock(GR_LFB_READ_ONLY, fxMesa->currentFB);
      }
      END_BOARD_LOCK();
   }
}

static void
fxDDReadPixels555 (GLcontext * ctx,
		   GLint x, GLint y,
		   GLsizei width, GLsizei height,
		   GLenum format, GLenum type,
		   const struct gl_pixelstore_attrib *packing,
		   GLvoid *dstImage)
{
   if (ctx->_ImageTransferState/* & (IMAGE_SCALE_BIAS_BIT|IMAGE_MAP_COLOR_BIT)*/) {
      _swrast_ReadPixels(ctx, x, y, width, height, format, type,
			 packing, dstImage);
      return;
   }
   else {
      fxMesaContext fxMesa = FX_CONTEXT(ctx);
      GrLfbInfo_t info;

      BEGIN_BOARD_LOCK();
      info.size = sizeof(info);
      if (grLfbLock(GR_LFB_READ_ONLY,
		    fxMesa->currentFB,
		    GR_LFBWRITEMODE_ANY,
		    GR_ORIGIN_UPPER_LEFT, FXFALSE, &info)) {
	 const GLint winX = 0;
	 const GLint winY = fxMesa->height - 1;
	 const GLint srcStride = info.strideInBytes / 2;	/* stride in GLushorts */
	 const GLushort *src = (const GLushort *) info.lfbPtr
	    + (winY - y) * srcStride + (winX + x);
	 GLubyte *dst = (GLubyte *) _mesa_image_address(packing, dstImage,
							width, height, format,
							type, 0, 0, 0);
	 GLint dstStride =
	    _mesa_image_row_stride(packing, width, format, type);

	 if (format == GL_RGB && type == GL_UNSIGNED_BYTE) {
	    /* convert 5R5G5B into 8R8G8B */
	    GLint row, col;
	    const GLint halfWidth = width >> 1;
	    const GLint extraPixel = (width & 1);
	    for (row = 0; row < height; row++) {
	       GLubyte *d = dst;
	       for (col = 0; col < halfWidth; col++) {
		  const GLuint pixel = ((const GLuint *) src)[col];
                  *d++ = FX_rgb_scale_5[(pixel >> 10) & 0x1f];
                  *d++ = FX_rgb_scale_5[(pixel >> 5)  & 0x1f];
                  *d++ = FX_rgb_scale_5[ pixel        & 0x1f];
                  *d++ = FX_rgb_scale_5[(pixel >> 26) & 0x1f];
                  *d++ = FX_rgb_scale_5[(pixel >> 21) & 0x1f];
                  *d++ = FX_rgb_scale_5[(pixel >> 16) & 0x1f];
	       }
	       if (extraPixel) {
		  GLushort pixel = src[width - 1];
                  *d++ = FX_rgb_scale_5[(pixel >> 10) & 0x1f];
                  *d++ = FX_rgb_scale_5[(pixel >> 5)  & 0x1f];
                  *d++ = FX_rgb_scale_5[ pixel        & 0x1f];
	       }
	       dst += dstStride;
	       src -= srcStride;
	    }
	 }
	 else if (format == GL_RGBA && type == GL_UNSIGNED_BYTE) {
	    /* convert 5R6G5B into 8R8G8B8A */
	    GLint row, col;
	    const GLint halfWidth = width >> 1;
	    const GLint extraPixel = (width & 1);
	    for (row = 0; row < height; row++) {
	       GLubyte *d = dst;
	       for (col = 0; col < halfWidth; col++) {
		  const GLuint pixel = ((const GLuint *) src)[col];
                  *d++ = FX_rgb_scale_5[(pixel >> 10) & 0x1f];
                  *d++ = FX_rgb_scale_5[(pixel >>  5) & 0x1f];
                  *d++ = FX_rgb_scale_5[ pixel        & 0x1f];
		  *d++ =  (pixel & 0x8000) ? 255 : 0;
                  *d++ = FX_rgb_scale_5[(pixel >> 26) & 0x1f];
                  *d++ = FX_rgb_scale_5[(pixel >> 21) & 0x1f];
                  *d++ = FX_rgb_scale_5[(pixel >> 16) & 0x1f];
		  *d++ =  (pixel & 0x80000000) ? 255 : 0;
	       }
	       if (extraPixel) {
		  const GLushort pixel = src[width - 1];
                  *d++ = FX_rgb_scale_5[(pixel >> 10) & 0x1f];
                  *d++ = FX_rgb_scale_5[(pixel >>  5) & 0x1f];
                  *d++ = FX_rgb_scale_5[ pixel        & 0x1f];
		  *d++ =  (pixel & 0x8000) ? 255 : 0;
	       }
	       dst += dstStride;
	       src -= srcStride;
	    }
	 }
	 else if (format == GL_RGB && type == GL_UNSIGNED_SHORT_5_6_5) {
	    /* directly memcpy 5R5G5B pixels into client's buffer */
	    const GLint widthInBytes = width * 2;
	    GLint row;
	    for (row = 0; row < height; row++) {
	       MEMCPY(dst, src, widthInBytes);
	       dst += dstStride;
	       src -= srcStride;
	    }
	 }
	 else {
	    grLfbUnlock(GR_LFB_READ_ONLY, fxMesa->currentFB);
	    END_BOARD_LOCK();
	    _swrast_ReadPixels(ctx, x, y, width, height, format, type,
			       packing, dstImage);
	    return;
	 }

	 grLfbUnlock(GR_LFB_READ_ONLY, fxMesa->currentFB);
      }
      END_BOARD_LOCK();
   }
}

static void
fxDDReadPixels8888 (GLcontext * ctx,
		    GLint x, GLint y,
		    GLsizei width, GLsizei height,
		    GLenum format, GLenum type,
		    const struct gl_pixelstore_attrib *packing,
		    GLvoid *dstImage)
{
   if (ctx->_ImageTransferState/* & (IMAGE_SCALE_BIAS_BIT|IMAGE_MAP_COLOR_BIT)*/) {
      _swrast_ReadPixels(ctx, x, y, width, height, format, type,
			 packing, dstImage);
      return;
   }
   else {
      fxMesaContext fxMesa = FX_CONTEXT(ctx);
      GrLfbInfo_t info;

      BEGIN_BOARD_LOCK();
      info.size = sizeof(info);
      if (grLfbLock(GR_LFB_READ_ONLY,
		    fxMesa->currentFB,
		    GR_LFBWRITEMODE_ANY,
		    GR_ORIGIN_UPPER_LEFT, FXFALSE, &info)) {
	 const GLint winX = 0;
	 const GLint winY = fxMesa->height - 1;
	 const GLint srcStride = info.strideInBytes / 4;	/* stride in GLuints */
	 const GLuint *src = (const GLuint *) info.lfbPtr
	    + (winY - y) * srcStride + (winX + x);
	 GLubyte *dst = (GLubyte *) _mesa_image_address(packing, dstImage,
							width, height, format,
							type, 0, 0, 0);
	 GLint dstStride =
	    _mesa_image_row_stride(packing, width, format, type);

	 if (format == GL_RGB && type == GL_UNSIGNED_BYTE) {
	    /* convert 8A8R8G8B into 8R8G8B */
	    GLint row, col;
	    for (row = 0; row < height; row++) {
	       GLubyte *d = dst;
	       for (col = 0; col < width; col++) {
		  const GLuint pixel = ((const GLuint *) src)[col];
                  *d++ = pixel >> 16;
                  *d++ = pixel >> 8;
                  *d++ = pixel;
	       }
	       dst += dstStride;
	       src -= srcStride;
	    }
	 }
	 else if (format == GL_RGBA && type == GL_UNSIGNED_BYTE) {
	    /* 8A8R8G8B pixels into client's buffer */
	    GLint row, col;
	    for (row = 0; row < height; row++) {
	       GLubyte *d = dst;
	       for (col = 0; col < width; col++) {
		  const GLuint pixel = ((const GLuint *) src)[col];
                  *d++ = pixel >> 16;
                  *d++ = pixel >> 8;
                  *d++ = pixel;
                  *d++ = pixel >> 24;
	       }
	       dst += dstStride;
	       src -= srcStride;
	    }
	 }
	 else if (format == GL_RGB && type == GL_UNSIGNED_SHORT_5_6_5) {
	    /* convert 8A8R8G8B into 5R6G5B */
	    GLint row, col;
	    for (row = 0; row < height; row++) {
	       GLushort *d = (GLushort *)dst;
	       for (col = 0; col < width; col++) {
		  const GLuint pixel = ((const GLuint *) src)[col];
                  *d++ = (((pixel >> 16) & 0xf8) << 8) |
                         (((pixel >>  8) & 0xfc) << 3) |
                          ((pixel        & 0xf8) >> 3);
	       }
	       dst += dstStride;
	       src -= srcStride;
	    }
	 }
	 else {
	    grLfbUnlock(GR_LFB_READ_ONLY, fxMesa->currentFB);
	    END_BOARD_LOCK();
	    _swrast_ReadPixels(ctx, x, y, width, height, format, type,
			       packing, dstImage);
	    return;
	 }

	 grLfbUnlock(GR_LFB_READ_ONLY, fxMesa->currentFB);
      }
      END_BOARD_LOCK();
   }
}


/* [dBorca] Hack alert:
 * not finished!!!
 * revise fallback tests and fix scissor; implement new formats
 * also write its siblings: 565 and 1555
 */
void
fxDDDrawPixels8888 (GLcontext * ctx, GLint x, GLint y,
                    GLsizei width, GLsizei height,
                    GLenum format, GLenum type,
                    const struct gl_pixelstore_attrib *unpack,
                    const GLvoid * pixels)
{
   fxMesaContext fxMesa = FX_CONTEXT(ctx);
   GrLfbInfo_t info;

   if (ctx->Pixel.ZoomX != 1.0F ||
       ctx->Pixel.ZoomY != 1.0F ||
       (ctx->_ImageTransferState & (IMAGE_SCALE_BIAS_BIT|
				    IMAGE_MAP_COLOR_BIT)) ||
       ctx->Color.AlphaEnabled ||
       ctx->Depth.Test ||
       ctx->Fog.Enabled ||
       ctx->Scissor.Enabled ||
       ctx->Stencil.Enabled ||
       !ctx->Color.ColorMask[0] ||
       !ctx->Color.ColorMask[1] ||
       !ctx->Color.ColorMask[2] ||
       !ctx->Color.ColorMask[3] ||
       ctx->Color.ColorLogicOpEnabled ||
       ctx->Texture._EnabledUnits ||
       ctx->Depth.OcclusionTest ||
       fxMesa->fallback)
   {
      _swrast_DrawPixels( ctx, x, y, width, height, format, type, 
			  unpack, pixels );
      return; 
   }

   /* lock early to make sure cliprects are right */
   BEGIN_BOARD_LOCK();

   /* make sure the pixelpipe is configured correctly */
   fxSetupFXUnits(ctx);

   /* look for clipmasks, giveup if region obscured */
#if 0
   if (ctx->Color.DrawBuffer == GL_FRONT) {
      if (!inClipRects_Region(fxMesa, scrX, scrY, width, height)) {
         END_BOARD_LOCK(fxMesa);
         _swrast_DrawPixels(ctx, x, y, width, height, format, type, unpack, pixels);
         return;
      }
   }
#endif

   info.size = sizeof(info);
   if (!grLfbLock(GR_LFB_WRITE_ONLY,
                  fxMesa->currentFB,
                  GR_LFBWRITEMODE_8888,
                  GR_ORIGIN_UPPER_LEFT, FXTRUE, &info)) {
      _swrast_DrawPixels(ctx, x, y, width, height, format, type, unpack, pixels);
      return;
   }

   {
      const GLint winX = 0;
      const GLint winY = fxMesa->height - 1;

      const GLint dstStride = info.strideInBytes / 4;	/* stride in GLuints */
      GLuint *dst = (GLuint *) info.lfbPtr + (winY - y) * dstStride + (winX + x);
      const GLubyte *src = (GLubyte *)_mesa_image_address(unpack, pixels,
							width, height, format,
							type, 0, 0, 0);
      const GLint srcStride = _mesa_image_row_stride(unpack, width, format, type);

      if (format == GL_RGBA && type == GL_UNSIGNED_BYTE) {
         /* directly memcpy 8A8R8G8B pixels to screen */
         const GLint widthInBytes = width * 4;
         GLint row;
         for (row = 0; row < height; row++) {
             MEMCPY(dst, src, widthInBytes);
             dst -= dstStride;
             src += srcStride;
         }
      }
      else {
         grLfbUnlock(GR_LFB_WRITE_ONLY, fxMesa->currentFB);
         END_BOARD_LOCK();
         _swrast_DrawPixels(ctx, x, y, width, height, format, type, unpack, pixels);
         return;
      }

   }

   grLfbUnlock(GR_LFB_WRITE_ONLY, fxMesa->currentFB);
   END_BOARD_LOCK();
}


static void
fxDDFinish(GLcontext * ctx)
{
   grFlush();
}





/* KW: Put the word Mesa in the render string because quakeworld
 * checks for this rather than doing a glGet(GL_MAX_TEXTURE_SIZE).
 * Why?
 */
static const GLubyte *
fxDDGetString(GLcontext * ctx, GLenum name)
{
 fxMesaContext fxMesa = FX_CONTEXT(ctx);

 switch (name) {
        case GL_RENDERER:
             return (GLubyte *)fxMesa->rendererString;
#if __WIN32__ /* hack to advertise vanilla extension names */
        case GL_EXTENSIONS:
             if (ctx->Extensions.String == NULL) {
                GLubyte *ext = _mesa_make_extension_string(ctx);
                if (ext != NULL) {
                   ctx->Extensions.String = _mesa_malloc(strlen((char *)ext) + 256);
                   if (ctx->Extensions.String != NULL) {
                      strcpy((char *)ctx->Extensions.String, (char *)ext);
                      /* put any additional extension names here */
#if 0
                      strcat((char *)ctx->Extensions.String, " 3DFX_set_global_palette");
#endif
#if __WIN32__
                      strcat((char *)ctx->Extensions.String, " WGL_3DFX_gamma_control");
                      strcat((char *)ctx->Extensions.String, " WGL_EXT_swap_control");
                      strcat((char *)ctx->Extensions.String, " WGL_EXT_extensions_string WGL_ARB_extensions_string");
#endif
                      /* put any additional extension names here */
                      _mesa_free(ext);
                   } else {
                      ctx->Extensions.String = ext;
                   }
                }
             }
             return ctx->Extensions.String;
#endif
        default:
             return NULL;
 }
}

static const struct tnl_pipeline_stage *fx_pipeline[] = {
   &_tnl_vertex_transform_stage,	/* TODO: Add the fastpath here */
   &_tnl_normal_transform_stage,
   &_tnl_lighting_stage,
   /*&_tnl_fog_coordinate_stage,*/	/* TODO: Omit fog stage ZZZ ZZZ ZZZ */
   &_tnl_texgen_stage,
   &_tnl_texture_transform_stage,
   &_tnl_point_attenuation_stage,
   &_tnl_render_stage,
   0,
};




int
fxDDInitFxMesaContext(fxMesaContext fxMesa)
{
   int i;
   GLcontext *ctx = fxMesa->glCtx;

   FX_setupGrVertexLayout();

   fxMesa->color = 0xffffffff;
   fxMesa->clearC = 0;
   fxMesa->clearA = 0;

   fxMesa->stats.swapBuffer = 0;
   fxMesa->stats.reqTexUpload = 0;
   fxMesa->stats.texUpload = 0;
   fxMesa->stats.memTexUpload = 0;

   fxMesa->tmuSrc = FX_TMU_NONE;
   fxMesa->lastUnitsMode = FX_UM_NONE;
   fxTMInit(fxMesa);

   /* FX units setup */

   fxMesa->unitsState.alphaTestEnabled = GL_FALSE;
   fxMesa->unitsState.alphaTestFunc = GL_ALWAYS;
   fxMesa->unitsState.alphaTestRefValue = 0.0;

   fxMesa->unitsState.blendEnabled = GL_FALSE;
   fxMesa->unitsState.blendSrcFuncRGB = GR_BLEND_ONE;
   fxMesa->unitsState.blendDstFuncRGB = GR_BLEND_ZERO;
   fxMesa->unitsState.blendSrcFuncAlpha = GR_BLEND_ONE;
   fxMesa->unitsState.blendDstFuncAlpha = GR_BLEND_ZERO;
   fxMesa->unitsState.blendEq = GR_BLEND_OP_ADD;

   fxMesa->unitsState.depthTestEnabled = GL_FALSE;
   fxMesa->unitsState.depthMask = GL_TRUE;
   fxMesa->unitsState.depthTestFunc = GL_LESS;
   fxMesa->unitsState.depthBias = 0;

   fxMesa->unitsState.stencilWriteMask = 0xff;

   if (fxMesa->colDepth != 16) {
      /* 32bpp mode or 15bpp mode */
      fxMesa->Glide.grColorMaskExt(FXTRUE, FXTRUE, FXTRUE, fxMesa->haveHwAlpha);
   } else {
      /* 16 bpp mode */
      grColorMask(FXTRUE, fxMesa->haveHwAlpha);
   }

   fxMesa->currentFB = fxMesa->haveDoubleBuffer ? GR_BUFFER_BACKBUFFER : GR_BUFFER_FRONTBUFFER;
   grRenderBuffer(fxMesa->currentFB);

   fxMesa->state = MALLOC(FX_grGetInteger(GR_GLIDE_STATE_SIZE));
   fxMesa->fogTable = (GrFog_t *) MALLOC(FX_grGetInteger(GR_FOG_TABLE_ENTRIES) *
			     sizeof(GrFog_t));

   if (!fxMesa->state || !fxMesa->fogTable) {
      if (fxMesa->state)
	 FREE(fxMesa->state);
      if (fxMesa->fogTable)
	 FREE(fxMesa->fogTable);
      return 0;
   }

   if (fxMesa->haveZBuffer)
      grDepthBufferMode(GR_DEPTHBUFFER_ZBUFFER);

   if (!fxMesa->bgrOrder) {
      grLfbWriteColorFormat(GR_COLORFORMAT_ABGR);
   }

   fxMesa->textureAlign = FX_grGetInteger(GR_TEXTURE_ALIGN);
   /* [koolsmoky] */
   {
    char *env;
    int textureLevels = 0;
    int textureSize = FX_grGetInteger(GR_MAX_TEXTURE_SIZE);
    do {
        textureLevels++;
    } while ((textureSize >>= 0x1) & 0x7ff);
    fxMesa->textureMaxLod = textureLevels - 1;
    ctx->Const.MaxTextureLevels = textureLevels;
    if ((env = getenv("MESA_FX_MAXLOD")) != NULL) {
       int maxLevels = atoi(env) + 1;
       if ((maxLevels <= MAX_TEXTURE_LEVELS) && (maxLevels > textureLevels)) {
          ctx->Const.MaxTextureLevels = maxLevels;
       }
    }
   }
   ctx->Const.MaxTextureCoordUnits =
   ctx->Const.MaxTextureImageUnits = fxMesa->haveTwoTMUs ? 2 : 1;
   ctx->Const.MaxTextureUnits = MAX2(ctx->Const.MaxTextureImageUnits, ctx->Const.MaxTextureCoordUnits);
   
   fxMesa->new_state = _NEW_ALL;
   if (!fxMesa->haveHwStencil) {
      /* don't touch stencil if there is none */
      fxMesa->new_state &= ~FX_NEW_STENCIL;
   }

   /* Initialize the software rasterizer and helper modules.
    */
   _swrast_CreateContext(ctx);
   _ac_CreateContext(ctx);
   _tnl_CreateContext(ctx);
   _swsetup_CreateContext(ctx);

   /* Install customized pipeline */
   _tnl_destroy_pipeline(ctx);
   _tnl_install_pipeline(ctx, fx_pipeline);

   fxAllocVB(ctx);

   fxSetupDDPointers(ctx);
   fxDDInitTriFuncs(ctx);

   /* Tell the software rasterizer to use pixel fog always.
    */
   _swrast_allow_vertex_fog(ctx, GL_FALSE);
   _swrast_allow_pixel_fog(ctx, GL_TRUE);

   /* Tell tnl not to calculate or use vertex fog factors.  (Needed to
    * tell render stage not to clip fog coords).
    */
/*     _tnl_calculate_vertex_fog( ctx, GL_FALSE ); */

   fxDDInitExtensions(ctx);

#if 0
   /* [dBorca] Hack alert:
    * do we want dither? It just looks bad...
    */
   grEnable(GR_ALLOW_MIPMAP_DITHER);
   grTexNccTable(GR_NCCTABLE_NCC0); /* set this once... no multipass */
#endif
   grGlideGetState((GrState *) fxMesa->state);

   return 1;
}

/* Undo the above.
 */
void
fxDDDestroyFxMesaContext(fxMesaContext fxMesa)
{
   _swsetup_DestroyContext(fxMesa->glCtx);
   _tnl_DestroyContext(fxMesa->glCtx);
   _ac_DestroyContext(fxMesa->glCtx);
   _swrast_DestroyContext(fxMesa->glCtx);

   if (fxMesa->state)
      FREE(fxMesa->state);
   if (fxMesa->fogTable)
      FREE(fxMesa->fogTable);
   fxTMClose(fxMesa);
   fxFreeVB(fxMesa->glCtx);
}




void
fxDDInitExtensions(GLcontext * ctx)
{
   fxMesaContext fxMesa = FX_CONTEXT(ctx);

#if 1 /* multipass ColorSum stage */
   _mesa_enable_extension(ctx, "GL_EXT_secondary_color");
#endif

   _mesa_enable_extension(ctx, "GL_EXT_point_parameters");
   _mesa_enable_extension(ctx, "GL_EXT_paletted_texture");
   _mesa_enable_extension(ctx, "GL_EXT_texture_lod_bias");
   _mesa_enable_extension(ctx, "GL_EXT_shared_texture_palette");
   _mesa_enable_extension(ctx, "GL_EXT_blend_func_separate");

   if (fxMesa->haveTwoTMUs) {
      _mesa_enable_extension(ctx, "GL_EXT_texture_env_add");
      _mesa_enable_extension(ctx, "GL_ARB_multitexture");
   }

   if (fxMesa->haveHwStencil) {
      _mesa_enable_extension( ctx, "GL_EXT_stencil_wrap" );
   }

   /* [dBorca] Hack alert:
    * True texture compression can be done only on Napalm.
    * We will advertise, however, generic texture compression
    * on all Voodoo cards; the Mesa logic allows us to eventually
    * fallback to uncompressed. This will fix those dumb applications
    * which refuse to run w/o texture compression! We actually _can_
    * do texture compression for pre-Napalm cores, through NCC. But
    * NCC poses many issues:
    * 1) NCC w/o DITHER_ERR has poor quality and NCC w/ DITHER_ERR is
    *    damn slow!
    * 2) NCC compression cannot be used with multitexturing, because
    *    the decompression tables are not per TMU anymore (bear in mind
    *    that earlier Voodoos could handle 2 NCC tables for each TMU --
    *    just look for POINTCAST_PALETTE). As a last resort, we could
    *    fake NCC multitexturing through multipass rendering, but...
    *    ohwell, it's not worth the effort...
    *    This stand true for multitexturing palletized textures.
    * 3) since NCC is not an OpenGL standard (as opposed to FXT1), we
    *    would need to plug deeper into the core... First, we would need to
    *    bind NCC to GL_COMPRESSED_RGB[A]. Then, we would need to trick
    *    Mesa into reporting our texture as compressed. Last, we would need
    *    to stash the NCC decompression table into the mipmap data and adjust
    *    CompressedSize accordingly!
    */
   if (fxMesa->HaveTexus2) {
      _mesa_enable_extension(ctx, "GL_ARB_texture_compression");

      if (fxMesa->type >= GR_SSTTYPE_Voodoo4) {
         _mesa_enable_extension(ctx, "GL_3DFX_texture_compression_FXT1");
         _mesa_enable_extension(ctx, "GL_EXT_texture_compression_s3tc");
         _mesa_enable_extension(ctx, "GL_S3_s3tc");
      }
   }

   if (fxMesa->HaveCmbExt) {
      _mesa_enable_extension(ctx, "GL_EXT_texture_env_combine");
   }

   if (fxMesa->HavePixExt) {
      _mesa_enable_extension(ctx, "GL_EXT_blend_subtract");
   }

   if (fxMesa->HaveMirExt) {
      _mesa_enable_extension(ctx, "GL_ARB_texture_mirrored_repeat");
   }
}


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

/* Check if the hardware supports the current context
 *
 * Performs similar work to fxDDChooseRenderState() - should be merged.
 */
GLuint
fx_check_IsInHardware(GLcontext * ctx)
{
   fxMesaContext fxMesa = FX_CONTEXT(ctx);

   if (ctx->RenderMode != GL_RENDER) {
      return FX_FALLBACK_RENDER_MODE;
   }

   if (ctx->Stencil.Enabled && !fxMesa->haveHwStencil) {
      return FX_FALLBACK_STENCIL;
   }

   if (ctx->Color._DrawDestMask != FRONT_LEFT_BIT && ctx->Color._DrawDestMask != BACK_LEFT_BIT) {
      return FX_FALLBACK_DRAW_BUFFER;
   }

   if (ctx->Color.BlendEnabled) {
      if (ctx->Color.BlendEquation != GL_FUNC_ADD_EXT) {
         if (!fxMesa->HavePixExt ||
             ((ctx->Color.BlendEquation != GL_FUNC_SUBTRACT_EXT) &&
              (ctx->Color.BlendEquation != GL_FUNC_REVERSE_SUBTRACT_EXT))) {
            return FX_FALLBACK_BLEND;
         }
      }
   }

   if (ctx->Color.ColorLogicOpEnabled && (ctx->Color.LogicOp != GL_COPY)) {
      return FX_FALLBACK_LOGICOP;
   }

#if 0 /* multipass ColorSum stage */
   if (ctx->Light.Model.ColorControl == GL_SEPARATE_SPECULAR_COLOR) {
      return FX_FALLBACK_SPECULAR;
   }
#endif

   if ((fxMesa->colDepth == 16) &&
       ((ctx->Color.ColorMask[RCOMP] != ctx->Color.ColorMask[GCOMP]) ||
        (ctx->Color.ColorMask[GCOMP] != ctx->Color.ColorMask[BCOMP]))) {
      return FX_FALLBACK_COLORMASK;
   }

   /* Unsupported texture/multitexture cases */

   if (fxMesa->haveTwoTMUs) {
      /* we can only do 2D textures */
      if (ctx->Texture.Unit[0]._ReallyEnabled & ~TEXTURE_2D_BIT)
	 return FX_FALLBACK_TEXTURE_1D_3D;
      if (ctx->Texture.Unit[1]._ReallyEnabled & ~TEXTURE_2D_BIT)
	 return FX_FALLBACK_TEXTURE_1D_3D;

      if (ctx->Texture.Unit[0]._ReallyEnabled & TEXTURE_2D_BIT) {
         if (fxMesa->type < GR_SSTTYPE_Voodoo2)
	 if (ctx->Texture.Unit[0].EnvMode == GL_BLEND &&
	     (ctx->Texture.Unit[1]._ReallyEnabled & TEXTURE_2D_BIT ||
	      ctx->Texture.Unit[0].EnvColor[0] != 0 ||
	      ctx->Texture.Unit[0].EnvColor[1] != 0 ||
	      ctx->Texture.Unit[0].EnvColor[2] != 0 ||
	      ctx->Texture.Unit[0].EnvColor[3] != 1)) {
	    return FX_FALLBACK_TEXTURE_ENV;
	 }
	 if (ctx->Texture.Unit[0]._Current->Image[0]->Border > 0)
	    return FX_FALLBACK_TEXTURE_BORDER;
      }

      if (ctx->Texture.Unit[1]._ReallyEnabled & TEXTURE_2D_BIT) {
         if (fxMesa->type < GR_SSTTYPE_Voodoo2)
	 if (ctx->Texture.Unit[1].EnvMode == GL_BLEND)
	    return FX_FALLBACK_TEXTURE_ENV;
	 if (ctx->Texture.Unit[1]._Current->Image[0]->Border > 0)
	    return FX_FALLBACK_TEXTURE_BORDER;
      }

      if (TDFX_DEBUG & (VERBOSE_DRIVER | VERBOSE_TEXTURE))
	 fprintf(stderr, "fx_check_IsInHardware: envmode is %s/%s\n",
		 _mesa_lookup_enum_by_nr(ctx->Texture.Unit[0].EnvMode),
		 _mesa_lookup_enum_by_nr(ctx->Texture.Unit[1].EnvMode));

      /* KW: This was wrong (I think) and I changed it... which doesn't mean
       * it is now correct...
       * BP: The old condition just seemed to test if both texture units
       * were enabled.  That's easy!
       */
      if (ctx->Texture._EnabledUnits == 0x3) {
	 /* Can't use multipass to blend a multitextured triangle - fall
	  * back to software.
	  */
	 if (!fxMesa->haveTwoTMUs && ctx->Color.BlendEnabled) {
	    return FX_FALLBACK_TEXTURE_MULTI;
	 }

	 if ((ctx->Texture.Unit[0].EnvMode != ctx->Texture.Unit[1].EnvMode) &&
	     (ctx->Texture.Unit[0].EnvMode != GL_MODULATE) &&
	     (ctx->Texture.Unit[0].EnvMode != GL_REPLACE)) {	/* q2, seems ok... */
	    if (TDFX_DEBUG & VERBOSE_DRIVER)
	       fprintf(stderr, "fx_check_IsInHardware: unsupported multitex env mode\n");
	    return FX_FALLBACK_TEXTURE_MULTI;
	 }
      }
   }
   else {
      /* we have just one texture unit */
      if (ctx->Texture._EnabledUnits > 0x1) {
	 return FX_FALLBACK_TEXTURE_MULTI;
      }

      if (fxMesa->type < GR_SSTTYPE_Voodoo2)
      if ((ctx->Texture.Unit[0]._ReallyEnabled & TEXTURE_2D_BIT) &&
	  (ctx->Texture.Unit[0].EnvMode == GL_BLEND)) {
	 return FX_FALLBACK_TEXTURE_ENV;
      }
   }

   return 0;
}



static void
fxDDUpdateDDPointers(GLcontext * ctx, GLuint new_state)
{
   /*   TNLcontext *tnl = TNL_CONTEXT(ctx);*/
   fxMesaContext fxMesa = FX_CONTEXT(ctx);

   if (TDFX_DEBUG & VERBOSE_DRIVER) {
      fprintf(stderr, "fxDDUpdateDDPointers(%08x)\n", new_state);
   }

   _swrast_InvalidateState(ctx, new_state);
   _ac_InvalidateState(ctx, new_state);
   _tnl_InvalidateState(ctx, new_state);
   _swsetup_InvalidateState(ctx, new_state);

   fxMesa->new_gl_state |= new_state;
}




void
fxSetupDDPointers(GLcontext * ctx)
{
   fxMesaContext fxMesa = FX_CONTEXT(ctx);
   TNLcontext *tnl = TNL_CONTEXT(ctx);

   if (TDFX_DEBUG & VERBOSE_DRIVER) {
      fprintf(stderr, "fxSetupDDPointers()\n");
   }

   ctx->Driver.UpdateState = fxDDUpdateDDPointers;
   ctx->Driver.GetString = fxDDGetString;
   ctx->Driver.ClearIndex = NULL;
   ctx->Driver.ClearColor = fxDDClearColor;
   ctx->Driver.Clear = fxDDClear;
   ctx->Driver.DrawBuffer = fxDDSetDrawBuffer;
   ctx->Driver.GetBufferSize = fxDDBufferSize;
   ctx->Driver.Accum = _swrast_Accum;
   ctx->Driver.CopyPixels = _swrast_CopyPixels;
   ctx->Driver.DrawPixels = _swrast_DrawPixels;
   switch (fxMesa->colDepth) {
          case 15:
               ctx->Driver.ReadPixels = fxDDReadPixels555;
               ctx->Driver.Bitmap = fxDDDrawBitmap2;
               break;
          case 16:
               ctx->Driver.ReadPixels = fxDDReadPixels565;
               ctx->Driver.Bitmap = fxDDDrawBitmap2;
               break;
          case 32:
               ctx->Driver.DrawPixels = fxDDDrawPixels8888;
               ctx->Driver.ReadPixels = fxDDReadPixels8888;
               ctx->Driver.Bitmap = fxDDDrawBitmap4;
               break;
   }
   ctx->Driver.ResizeBuffers = _swrast_alloc_buffers;
   ctx->Driver.Finish = fxDDFinish;
   ctx->Driver.Flush = NULL;
   ctx->Driver.ChooseTextureFormat = fxDDChooseTextureFormat;
   ctx->Driver.TexImage1D = _mesa_store_teximage1d;
   ctx->Driver.TexImage2D = fxDDTexImage2D;
   ctx->Driver.TexImage3D = _mesa_store_teximage3d;
   ctx->Driver.TexSubImage1D = _mesa_store_texsubimage1d;
   ctx->Driver.TexSubImage2D = fxDDTexSubImage2D;
   ctx->Driver.TexSubImage3D = _mesa_store_texsubimage3d;
   ctx->Driver.CompressedTexImage1D = _mesa_store_compressed_teximage1d;
   ctx->Driver.CompressedTexImage2D = fxDDCompressedTexImage2D;
   ctx->Driver.CompressedTexImage3D = _mesa_store_compressed_teximage3d;
   ctx->Driver.CompressedTexSubImage1D = _mesa_store_compressed_texsubimage1d;
   ctx->Driver.CompressedTexSubImage2D = fxDDCompressedTexSubImage2D;
   ctx->Driver.CompressedTexSubImage3D = _mesa_store_compressed_texsubimage3d;
   ctx->Driver.IsCompressedFormat = fxDDIsCompressedFormat;
   ctx->Driver.CompressedTextureSize = fxDDCompressedTextureSize;
   ctx->Driver.CopyTexImage1D = _swrast_copy_teximage1d;
   ctx->Driver.CopyTexImage2D = _swrast_copy_teximage2d;
   ctx->Driver.CopyTexSubImage1D = _swrast_copy_texsubimage1d;
   ctx->Driver.CopyTexSubImage2D = _swrast_copy_texsubimage2d;
   ctx->Driver.CopyTexSubImage3D = _swrast_copy_texsubimage3d;
   ctx->Driver.TestProxyTexImage = _mesa_test_proxy_teximage;
   ctx->Driver.CopyColorTable = _swrast_CopyColorTable;
   ctx->Driver.CopyColorSubTable = _swrast_CopyColorSubTable;
   ctx->Driver.CopyConvolutionFilter1D = _swrast_CopyConvolutionFilter1D;
   ctx->Driver.CopyConvolutionFilter2D = _swrast_CopyConvolutionFilter2D;
   ctx->Driver.TexEnv = fxDDTexEnv;
   ctx->Driver.TexParameter = fxDDTexParam;
   ctx->Driver.BindTexture = fxDDTexBind;
   ctx->Driver.DeleteTexture = fxDDTexDel;
   ctx->Driver.IsTextureResident = fxDDIsTextureResident;
   ctx->Driver.UpdateTexturePalette = fxDDTexPalette;
   ctx->Driver.AlphaFunc = fxDDAlphaFunc;
   ctx->Driver.BlendFunc = fxDDBlendFunc;
   ctx->Driver.BlendFuncSeparate = fxDDBlendFuncSeparate;
   ctx->Driver.BlendEquation = fxDDBlendEquation;
   ctx->Driver.DepthFunc = fxDDDepthFunc;
   ctx->Driver.DepthMask = fxDDDepthMask;
   ctx->Driver.ColorMask = fxDDColorMask;
   ctx->Driver.Fogfv = fxDDFogfv;
   ctx->Driver.Scissor = fxDDScissor;
   ctx->Driver.FrontFace = fxDDFrontFace;
   ctx->Driver.CullFace = fxDDCullFace;
   ctx->Driver.ShadeModel = fxDDShadeModel;
   ctx->Driver.Enable = fxDDEnable;
   if (fxMesa->haveHwStencil) {
      ctx->Driver.StencilFunc	= fxDDStencilFunc;
      ctx->Driver.StencilMask	= fxDDStencilMask;
      ctx->Driver.StencilOp	= fxDDStencilOp;
   }

   fxSetupDDSpanPointers(ctx);
   fxDDUpdateDDPointers(ctx, ~0);
}


#else


/*
 * Need this to provide at least one external definition.
 */

extern int gl_fx_dummy_function_dd(void);
int
gl_fx_dummy_function_dd(void)
{
   return 0;
}

#endif /* FX */