summaryrefslogtreecommitdiffstats
path: root/libswscale/swscale.c
blob: f45a4dfa4670a86b21b8ba4a5bd67d4577f9c7cf (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
1001
1002
1003
1004
1005
1006
1007
1008
1009
1010
1011
1012
1013
1014
1015
1016
1017
1018
1019
1020
1021
1022
1023
1024
1025
1026
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
1040
1041
1042
1043
1044
1045
1046
1047
1048
1049
1050
1051
1052
1053
1054
1055
1056
1057
1058
1059
1060
1061
1062
1063
1064
1065
1066
1067
1068
1069
1070
1071
1072
1073
1074
1075
1076
1077
1078
1079
1080
1081
1082
1083
1084
1085
1086
1087
1088
1089
1090
1091
1092
1093
1094
1095
1096
1097
1098
1099
1100
1101
1102
1103
1104
1105
1106
1107
1108
1109
1110
1111
1112
1113
1114
1115
1116
1117
1118
1119
1120
1121
1122
1123
1124
1125
1126
1127
1128
1129
1130
1131
1132
1133
1134
1135
1136
1137
1138
1139
1140
1141
1142
1143
1144
1145
1146
1147
1148
1149
1150
1151
1152
1153
1154
1155
1156
1157
1158
1159
1160
1161
1162
1163
1164
1165
1166
1167
1168
1169
1170
1171
1172
1173
1174
1175
1176
1177
1178
1179
1180
1181
1182
1183
1184
1185
1186
1187
1188
1189
1190
1191
1192
1193
1194
1195
1196
1197
1198
1199
1200
1201
1202
1203
1204
1205
1206
1207
1208
1209
1210
1211
1212
1213
1214
1215
1216
1217
1218
1219
1220
1221
1222
1223
1224
1225
1226
1227
1228
1229
1230
1231
1232
1233
1234
1235
1236
1237
1238
1239
1240
1241
1242
1243
1244
1245
1246
1247
1248
1249
1250
1251
1252
1253
1254
1255
1256
1257
1258
1259
1260
1261
1262
1263
1264
1265
1266
1267
1268
1269
1270
1271
1272
1273
1274
1275
1276
1277
1278
1279
1280
1281
1282
1283
1284
1285
1286
1287
1288
1289
1290
1291
1292
1293
1294
1295
1296
1297
1298
1299
1300
1301
1302
1303
1304
1305
1306
1307
1308
1309
1310
1311
1312
1313
1314
1315
1316
1317
1318
1319
1320
1321
1322
1323
1324
1325
1326
1327
1328
1329
1330
1331
1332
1333
1334
1335
1336
1337
1338
1339
1340
1341
1342
1343
1344
1345
1346
1347
1348
1349
1350
1351
1352
1353
1354
1355
1356
1357
1358
1359
1360
1361
1362
1363
1364
1365
1366
1367
1368
1369
1370
1371
1372
1373
1374
1375
1376
1377
1378
1379
1380
1381
1382
1383
1384
1385
1386
1387
1388
1389
1390
1391
1392
1393
1394
1395
1396
1397
1398
1399
1400
1401
1402
1403
1404
1405
1406
1407
1408
1409
1410
1411
1412
1413
1414
1415
1416
1417
1418
1419
1420
1421
1422
1423
1424
1425
1426
1427
1428
1429
1430
1431
1432
1433
1434
1435
1436
1437
1438
1439
1440
1441
1442
1443
1444
1445
1446
1447
1448
1449
1450
1451
1452
1453
1454
1455
1456
1457
1458
1459
1460
1461
1462
1463
1464
1465
1466
1467
1468
1469
1470
1471
1472
1473
1474
1475
1476
1477
1478
1479
1480
1481
1482
1483
1484
1485
1486
1487
1488
1489
1490
1491
1492
1493
1494
1495
1496
1497
1498
1499
1500
1501
1502
1503
1504
1505
1506
1507
1508
1509
1510
1511
1512
1513
1514
1515
1516
1517
1518
1519
1520
1521
1522
1523
1524
1525
1526
1527
1528
1529
1530
1531
1532
1533
1534
1535
1536
1537
1538
1539
1540
1541
1542
1543
1544
1545
1546
1547
1548
1549
1550
1551
1552
1553
1554
1555
1556
1557
1558
1559
1560
1561
1562
1563
1564
1565
1566
1567
1568
1569
1570
1571
1572
1573
1574
1575
1576
1577
1578
1579
1580
1581
1582
1583
1584
1585
1586
1587
1588
1589
1590
1591
1592
1593
1594
1595
1596
1597
1598
1599
1600
1601
1602
1603
1604
1605
1606
1607
1608
1609
1610
1611
1612
1613
1614
1615
1616
1617
1618
1619
1620
1621
1622
1623
1624
1625
1626
1627
1628
1629
1630
1631
1632
1633
1634
1635
1636
1637
1638
1639
1640
1641
1642
1643
1644
1645
1646
1647
1648
1649
1650
1651
1652
1653
1654
1655
1656
1657
1658
1659
1660
1661
1662
1663
1664
1665
1666
1667
1668
1669
1670
1671
1672
1673
1674
1675
1676
1677
1678
1679
1680
1681
1682
1683
1684
1685
1686
1687
1688
1689
1690
1691
1692
1693
1694
1695
1696
1697
1698
1699
1700
1701
1702
1703
1704
1705
1706
1707
1708
1709
1710
1711
1712
1713
1714
1715
1716
1717
1718
1719
1720
1721
1722
1723
1724
1725
1726
1727
1728
1729
1730
1731
1732
1733
1734
1735
1736
1737
1738
1739
1740
1741
1742
1743
1744
1745
1746
1747
1748
1749
1750
1751
1752
1753
1754
1755
1756
1757
1758
1759
1760
1761
1762
1763
1764
1765
1766
1767
1768
1769
1770
1771
1772
1773
1774
1775
1776
1777
1778
1779
1780
1781
1782
1783
1784
1785
1786
1787
1788
1789
1790
1791
1792
1793
1794
1795
1796
1797
1798
1799
1800
1801
1802
1803
1804
1805
1806
1807
1808
1809
1810
1811
1812
1813
1814
1815
1816
1817
1818
1819
1820
1821
1822
1823
1824
1825
1826
1827
1828
1829
1830
1831
1832
1833
1834
1835
1836
1837
1838
1839
1840
1841
1842
1843
1844
1845
1846
1847
1848
1849
1850
1851
1852
1853
1854
1855
1856
1857
1858
1859
1860
1861
1862
1863
1864
1865
1866
1867
1868
1869
1870
1871
1872
1873
1874
1875
1876
1877
1878
1879
1880
1881
1882
1883
1884
1885
1886
1887
1888
1889
1890
1891
1892
1893
1894
1895
1896
1897
1898
1899
1900
1901
1902
1903
1904
1905
1906
1907
1908
1909
1910
1911
1912
1913
1914
1915
1916
1917
1918
1919
1920
1921
1922
1923
1924
1925
1926
1927
1928
1929
1930
1931
1932
1933
1934
1935
1936
1937
1938
1939
1940
1941
1942
1943
1944
1945
1946
1947
1948
1949
1950
1951
1952
1953
1954
1955
1956
1957
1958
1959
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
2030
2031
2032
2033
2034
2035
2036
2037
2038
2039
2040
2041
2042
2043
2044
2045
2046
2047
2048
2049
2050
2051
2052
2053
2054
2055
2056
2057
2058
2059
2060
2061
2062
2063
2064
2065
2066
2067
2068
2069
2070
2071
2072
2073
2074
2075
2076
2077
2078
2079
2080
2081
2082
2083
2084
2085
2086
2087
2088
2089
2090
2091
2092
2093
2094
2095
2096
2097
2098
2099
2100
2101
2102
2103
2104
2105
2106
2107
2108
2109
2110
2111
2112
2113
2114
2115
2116
2117
2118
2119
2120
2121
2122
2123
2124
2125
2126
2127
2128
2129
2130
2131
2132
2133
2134
2135
2136
2137
2138
2139
2140
2141
2142
2143
2144
2145
2146
2147
2148
2149
2150
2151
2152
2153
2154
2155
2156
2157
2158
2159
2160
2161
2162
2163
2164
2165
2166
2167
2168
2169
2170
2171
2172
2173
2174
2175
2176
2177
2178
2179
2180
2181
2182
2183
2184
2185
2186
2187
2188
2189
2190
2191
2192
2193
2194
2195
2196
2197
2198
2199
2200
2201
2202
2203
2204
2205
2206
2207
2208
2209
2210
2211
2212
2213
2214
2215
2216
2217
2218
2219
2220
2221
2222
2223
2224
2225
2226
2227
2228
2229
2230
2231
2232
2233
2234
2235
2236
2237
2238
2239
2240
2241
2242
2243
2244
2245
2246
2247
2248
2249
2250
2251
2252
2253
2254
2255
2256
2257
2258
2259
2260
2261
2262
2263
2264
2265
2266
2267
2268
2269
2270
2271
2272
2273
2274
2275
2276
2277
2278
2279
2280
2281
2282
2283
2284
2285
2286
2287
2288
2289
2290
2291
2292
2293
2294
2295
2296
2297
2298
2299
2300
2301
2302
2303
2304
2305
2306
2307
2308
2309
2310
2311
2312
2313
2314
2315
2316
2317
2318
2319
2320
2321
2322
2323
2324
2325
2326
2327
2328
2329
2330
2331
2332
2333
2334
2335
2336
2337
2338
2339
2340
2341
2342
2343
2344
2345
2346
2347
2348
2349
2350
2351
2352
2353
2354
2355
2356
2357
2358
2359
2360
2361
2362
2363
2364
2365
2366
2367
2368
2369
2370
2371
2372
2373
2374
2375
2376
2377
2378
2379
2380
2381
2382
2383
2384
2385
2386
2387
2388
2389
2390
2391
2392
2393
2394
2395
2396
2397
2398
2399
2400
2401
2402
2403
2404
2405
2406
2407
2408
2409
2410
2411
2412
2413
2414
2415
2416
2417
2418
2419
2420
2421
2422
2423
2424
2425
2426
2427
2428
2429
2430
2431
2432
2433
2434
2435
2436
2437
2438
2439
2440
2441
2442
2443
2444
2445
2446
2447
2448
2449
2450
2451
2452
2453
2454
2455
2456
2457
2458
2459
2460
2461
2462
2463
2464
2465
2466
2467
2468
2469
2470
2471
2472
2473
2474
2475
2476
2477
2478
2479
2480
2481
2482
2483
2484
2485
2486
2487
2488
2489
2490
2491
2492
2493
2494
2495
2496
2497
2498
2499
2500
2501
2502
2503
2504
2505
2506
2507
2508
2509
2510
2511
2512
2513
2514
2515
2516
2517
2518
2519
2520
2521
2522
2523
2524
2525
2526
2527
2528
2529
2530
2531
2532
2533
2534
2535
2536
2537
2538
2539
2540
2541
2542
2543
2544
2545
2546
2547
2548
2549
2550
2551
2552
2553
2554
2555
2556
2557
2558
2559
2560
2561
2562
2563
2564
2565
2566
2567
2568
2569
2570
2571
2572
2573
2574
2575
2576
2577
2578
2579
2580
2581
2582
2583
2584
2585
2586
2587
2588
2589
2590
2591
2592
2593
2594
2595
2596
2597
2598
2599
2600
2601
2602
2603
2604
2605
2606
2607
2608
2609
2610
2611
2612
2613
2614
2615
2616
2617
2618
2619
2620
2621
2622
2623
2624
2625
2626
2627
2628
2629
2630
2631
2632
2633
2634
2635
2636
2637
2638
2639
2640
2641
2642
2643
2644
2645
2646
2647
2648
2649
2650
2651
2652
2653
2654
2655
2656
2657
2658
2659
2660
2661
2662
2663
2664
2665
2666
2667
2668
2669
2670
2671
2672
2673
2674
2675
2676
2677
2678
2679
2680
2681
2682
2683
2684
2685
2686
2687
2688
2689
2690
2691
2692
2693
2694
2695
2696
2697
2698
2699
2700
2701
2702
2703
2704
2705
2706
2707
2708
2709
2710
2711
2712
2713
2714
2715
2716
2717
2718
2719
2720
2721
2722
2723
/*
    Copyright (C) 2001-2003 Michael Niedermayer <michaelni@gmx.at>

    This program is free software; you can redistribute it and/or modify
    it under the terms of the GNU General Public License as published by
    the Free Software Foundation; either version 2 of the License, or
    (at your option) any later version.

    This program is distributed in the hope that it will be useful,
    but WITHOUT ANY WARRANTY; without even the implied warranty of
    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
    GNU General Public License for more details.

    You should have received a copy of the GNU General Public License
    along with this program; if not, write to the Free Software
    Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
*/

/*
  supported Input formats: YV12, I420/IYUV, YUY2, UYVY, BGR32, BGR24, BGR16, BGR15, RGB32, RGB24, Y8/Y800, YVU9/IF09
  supported output formats: YV12, I420/IYUV, YUY2, UYVY, {BGR,RGB}{1,4,8,15,16,24,32}, Y8/Y800, YVU9/IF09
  {BGR,RGB}{1,4,8,15,16} support dithering
  
  unscaled special converters (YV12=I420=IYUV, Y800=Y8)
  YV12 -> {BGR,RGB}{1,4,8,15,16,24,32}
  x -> x
  YUV9 -> YV12
  YUV9/YV12 -> Y800
  Y800 -> YUV9/YV12
  BGR24 -> BGR32 & RGB24 -> RGB32
  BGR32 -> BGR24 & RGB32 -> RGB24
  BGR15 -> BGR16
*/

/* 
tested special converters (most are tested actually but i didnt write it down ...)
 YV12 -> BGR16
 YV12 -> YV12
 BGR15 -> BGR16
 BGR16 -> BGR16
 YVU9 -> YV12

untested special converters
  YV12/I420 -> BGR15/BGR24/BGR32 (its the yuv2rgb stuff, so it should be ok)
  YV12/I420 -> YV12/I420
  YUY2/BGR15/BGR24/BGR32/RGB24/RGB32 -> same format
  BGR24 -> BGR32 & RGB24 -> RGB32
  BGR32 -> BGR24 & RGB32 -> RGB24
  BGR24 -> YV12
*/

#include <inttypes.h>
#include <string.h>
#include <math.h>
#include <stdio.h>
#include <unistd.h>
#include "config.h"
#include <assert.h>
#ifdef HAVE_MALLOC_H
#include <malloc.h>
#else
#include <stdlib.h>
#endif
#ifdef HAVE_SYS_MMAN_H
#include <sys/mman.h>
#if defined(MAP_ANON) && !defined(MAP_ANONYMOUS)
#define MAP_ANONYMOUS MAP_ANON
#endif
#endif
#include "swscale.h"
#include "swscale_internal.h"
#include "x86_cpu.h"
#include "bswap.h"
#include "img_format.h"
#include "rgb2rgb.h"
#ifdef USE_FASTMEMCPY
#include "libvo/fastmemcpy.h"
#endif

#undef MOVNTQ
#undef PAVGB

//#undef HAVE_MMX2
//#define HAVE_3DNOW
//#undef HAVE_MMX
//#undef ARCH_X86
//#define WORDS_BIGENDIAN
#define DITHER1XBPP

#define FAST_BGR2YV12 // use 7 bit coeffs instead of 15bit

#define RET 0xC3 //near return opcode for X86

#ifdef MP_DEBUG
#define ASSERT(x) assert(x);
#else
#define ASSERT(x) ;
#endif

#ifdef M_PI
#define PI M_PI
#else
#define PI 3.14159265358979323846
#endif

//FIXME replace this with something faster
#define isPlanarYUV(x) ((x)==IMGFMT_YV12 || (x)==IMGFMT_YVU9 \
			|| (x)==IMGFMT_NV12 || (x)==IMGFMT_NV21 \
			|| (x)==IMGFMT_444P || (x)==IMGFMT_422P || (x)==IMGFMT_411P)
#define isYUV(x)       ((x)==IMGFMT_UYVY || (x)==IMGFMT_YUY2 || isPlanarYUV(x))
#define isGray(x)      ((x)==IMGFMT_Y800)
#define isRGB(x)       (((x)&IMGFMT_RGB_MASK)==IMGFMT_RGB)
#define isBGR(x)       (((x)&IMGFMT_BGR_MASK)==IMGFMT_BGR)
#define isSupportedIn(x)  ((x)==IMGFMT_YV12 || (x)==IMGFMT_YUY2 || (x)==IMGFMT_UYVY\
			|| (x)==IMGFMT_BGR32|| (x)==IMGFMT_BGR24|| (x)==IMGFMT_BGR16|| (x)==IMGFMT_BGR15\
			|| (x)==IMGFMT_RGB32|| (x)==IMGFMT_RGB24\
			|| (x)==IMGFMT_Y800 || (x)==IMGFMT_YVU9\
			|| (x)==IMGFMT_444P || (x)==IMGFMT_422P || (x)==IMGFMT_411P)
#define isSupportedOut(x) ((x)==IMGFMT_YV12 || (x)==IMGFMT_YUY2 || (x)==IMGFMT_UYVY\
			|| (x)==IMGFMT_444P || (x)==IMGFMT_422P || (x)==IMGFMT_411P\
			|| isRGB(x) || isBGR(x)\
			|| (x)==IMGFMT_NV12 || (x)==IMGFMT_NV21\
			|| (x)==IMGFMT_Y800 || (x)==IMGFMT_YVU9)
#define isPacked(x)    ((x)==IMGFMT_YUY2 || (x)==IMGFMT_UYVY ||isRGB(x) || isBGR(x))

#define RGB2YUV_SHIFT 16
#define BY ((int)( 0.098*(1<<RGB2YUV_SHIFT)+0.5))
#define BV ((int)(-0.071*(1<<RGB2YUV_SHIFT)+0.5))
#define BU ((int)( 0.439*(1<<RGB2YUV_SHIFT)+0.5))
#define GY ((int)( 0.504*(1<<RGB2YUV_SHIFT)+0.5))
#define GV ((int)(-0.368*(1<<RGB2YUV_SHIFT)+0.5))
#define GU ((int)(-0.291*(1<<RGB2YUV_SHIFT)+0.5))
#define RY ((int)( 0.257*(1<<RGB2YUV_SHIFT)+0.5))
#define RV ((int)( 0.439*(1<<RGB2YUV_SHIFT)+0.5))
#define RU ((int)(-0.148*(1<<RGB2YUV_SHIFT)+0.5))

extern const int32_t Inverse_Table_6_9[8][4];

/*
NOTES
Special versions: fast Y 1:1 scaling (no interpolation in y direction)

TODO
more intelligent missalignment avoidance for the horizontal scaler
write special vertical cubic upscale version
Optimize C code (yv12 / minmax)
add support for packed pixel yuv input & output
add support for Y8 output
optimize bgr24 & bgr32
add BGR4 output support
write special BGR->BGR scaler
*/

#define MIN(a,b) ((a) > (b) ? (b) : (a))
#define MAX(a,b) ((a) < (b) ? (b) : (a))

#if defined(ARCH_X86) || defined(ARCH_X86_64)
static uint64_t attribute_used __attribute__((aligned(8))) bF8=       0xF8F8F8F8F8F8F8F8LL;
static uint64_t attribute_used __attribute__((aligned(8))) bFC=       0xFCFCFCFCFCFCFCFCLL;
static uint64_t __attribute__((aligned(8))) w10=       0x0010001000100010LL;
static uint64_t attribute_used __attribute__((aligned(8))) w02=       0x0002000200020002LL;
static uint64_t attribute_used __attribute__((aligned(8))) bm00001111=0x00000000FFFFFFFFLL;
static uint64_t attribute_used __attribute__((aligned(8))) bm00000111=0x0000000000FFFFFFLL;
static uint64_t attribute_used __attribute__((aligned(8))) bm11111000=0xFFFFFFFFFF000000LL;
static uint64_t attribute_used __attribute__((aligned(8))) bm01010101=0x00FF00FF00FF00FFLL;

static volatile uint64_t attribute_used __attribute__((aligned(8))) b5Dither;
static volatile uint64_t attribute_used __attribute__((aligned(8))) g5Dither;
static volatile uint64_t attribute_used __attribute__((aligned(8))) g6Dither;
static volatile uint64_t attribute_used __attribute__((aligned(8))) r5Dither;

static uint64_t __attribute__((aligned(8))) dither4[2]={
	0x0103010301030103LL,
	0x0200020002000200LL,};

static uint64_t __attribute__((aligned(8))) dither8[2]={
	0x0602060206020602LL,
	0x0004000400040004LL,};

static uint64_t __attribute__((aligned(8))) b16Mask=   0x001F001F001F001FLL;
static uint64_t attribute_used __attribute__((aligned(8))) g16Mask=   0x07E007E007E007E0LL;
static uint64_t attribute_used __attribute__((aligned(8))) r16Mask=   0xF800F800F800F800LL;
static uint64_t __attribute__((aligned(8))) b15Mask=   0x001F001F001F001FLL;
static uint64_t attribute_used __attribute__((aligned(8))) g15Mask=   0x03E003E003E003E0LL;
static uint64_t attribute_used __attribute__((aligned(8))) r15Mask=   0x7C007C007C007C00LL;

static uint64_t attribute_used __attribute__((aligned(8))) M24A=   0x00FF0000FF0000FFLL;
static uint64_t attribute_used __attribute__((aligned(8))) M24B=   0xFF0000FF0000FF00LL;
static uint64_t attribute_used __attribute__((aligned(8))) M24C=   0x0000FF0000FF0000LL;

#ifdef FAST_BGR2YV12
static const uint64_t bgr2YCoeff  attribute_used __attribute__((aligned(8))) = 0x000000210041000DULL;
static const uint64_t bgr2UCoeff  attribute_used __attribute__((aligned(8))) = 0x0000FFEEFFDC0038ULL;
static const uint64_t bgr2VCoeff  attribute_used __attribute__((aligned(8))) = 0x00000038FFD2FFF8ULL;
#else
static const uint64_t bgr2YCoeff  attribute_used __attribute__((aligned(8))) = 0x000020E540830C8BULL;
static const uint64_t bgr2UCoeff  attribute_used __attribute__((aligned(8))) = 0x0000ED0FDAC23831ULL;
static const uint64_t bgr2VCoeff  attribute_used __attribute__((aligned(8))) = 0x00003831D0E6F6EAULL;
#endif
static const uint64_t bgr2YOffset attribute_used __attribute__((aligned(8))) = 0x1010101010101010ULL;
static const uint64_t bgr2UVOffset attribute_used __attribute__((aligned(8)))= 0x8080808080808080ULL;
static const uint64_t w1111       attribute_used __attribute__((aligned(8))) = 0x0001000100010001ULL;
#endif

// clipping helper table for C implementations:
static unsigned char clip_table[768];

static SwsVector *sws_getConvVec(SwsVector *a, SwsVector *b);
		  
extern const uint8_t dither_2x2_4[2][8];
extern const uint8_t dither_2x2_8[2][8];
extern const uint8_t dither_8x8_32[8][8];
extern const uint8_t dither_8x8_73[8][8];
extern const uint8_t dither_8x8_220[8][8];

char *sws_format_name(int format)
{
    static char fmt_name[64];
    char *res;
    static int buffer;

    res = fmt_name + buffer * 32;
    buffer = 1 - buffer;
    snprintf(res, 32, "0x%x (%c%c%c%c)", format,
		    format >> 24, (format >> 16) & 0xFF,
		    (format >> 8) & 0xFF,
		    format & 0xFF);

    return res;
}

#if defined(ARCH_X86) || defined(ARCH_X86_64)
void in_asm_used_var_warning_killer()
{
 volatile int i= bF8+bFC+w10+
 bm00001111+bm00000111+bm11111000+b16Mask+g16Mask+r16Mask+b15Mask+g15Mask+r15Mask+
 M24A+M24B+M24C+w02 + b5Dither+g5Dither+r5Dither+g6Dither+dither4[0]+dither8[0]+bm01010101;
 if(i) i=0;
}
#endif

static inline void yuv2yuvXinC(int16_t *lumFilter, int16_t **lumSrc, int lumFilterSize,
				    int16_t *chrFilter, int16_t **chrSrc, int chrFilterSize,
				    uint8_t *dest, uint8_t *uDest, uint8_t *vDest, int dstW, int chrDstW)
{
	//FIXME Optimize (just quickly writen not opti..)
	int i;
	for(i=0; i<dstW; i++)
	{
		int val=1<<18;
		int j;
		for(j=0; j<lumFilterSize; j++)
			val += lumSrc[j][i] * lumFilter[j];

		dest[i]= MIN(MAX(val>>19, 0), 255);
	}

	if(uDest != NULL)
		for(i=0; i<chrDstW; i++)
		{
			int u=1<<18;
			int v=1<<18;
			int j;
			for(j=0; j<chrFilterSize; j++)
			{
				u += chrSrc[j][i] * chrFilter[j];
				v += chrSrc[j][i + 2048] * chrFilter[j];
			}

			uDest[i]= MIN(MAX(u>>19, 0), 255);
			vDest[i]= MIN(MAX(v>>19, 0), 255);
		}
}

static inline void yuv2nv12XinC(int16_t *lumFilter, int16_t **lumSrc, int lumFilterSize,
				int16_t *chrFilter, int16_t **chrSrc, int chrFilterSize,
				uint8_t *dest, uint8_t *uDest, int dstW, int chrDstW, int dstFormat)
{
	//FIXME Optimize (just quickly writen not opti..)
	int i;
	for(i=0; i<dstW; i++)
	{
		int val=1<<18;
		int j;
		for(j=0; j<lumFilterSize; j++)
			val += lumSrc[j][i] * lumFilter[j];

		dest[i]= MIN(MAX(val>>19, 0), 255);
	}

	if(uDest == NULL)
		return;

	if(dstFormat == IMGFMT_NV12)
		for(i=0; i<chrDstW; i++)
		{
			int u=1<<18;
			int v=1<<18;
			int j;
			for(j=0; j<chrFilterSize; j++)
			{
				u += chrSrc[j][i] * chrFilter[j];
				v += chrSrc[j][i + 2048] * chrFilter[j];
			}

			uDest[2*i]= MIN(MAX(u>>19, 0), 255);
			uDest[2*i+1]= MIN(MAX(v>>19, 0), 255);
		}
	else
		for(i=0; i<chrDstW; i++)
		{
			int u=1<<18;
			int v=1<<18;
			int j;
			for(j=0; j<chrFilterSize; j++)
			{
				u += chrSrc[j][i] * chrFilter[j];
				v += chrSrc[j][i + 2048] * chrFilter[j];
			}

			uDest[2*i]= MIN(MAX(v>>19, 0), 255);
			uDest[2*i+1]= MIN(MAX(u>>19, 0), 255);
		}
}

#define YSCALE_YUV_2_PACKEDX_C(type) \
		for(i=0; i<(dstW>>1); i++){\
			int j;\
			int Y1=1<<18;\
			int Y2=1<<18;\
			int U=1<<18;\
			int V=1<<18;\
			type *r, *b, *g;\
			const int i2= 2*i;\
			\
			for(j=0; j<lumFilterSize; j++)\
			{\
				Y1 += lumSrc[j][i2] * lumFilter[j];\
				Y2 += lumSrc[j][i2+1] * lumFilter[j];\
			}\
			for(j=0; j<chrFilterSize; j++)\
			{\
				U += chrSrc[j][i] * chrFilter[j];\
				V += chrSrc[j][i+2048] * chrFilter[j];\
			}\
			Y1>>=19;\
			Y2>>=19;\
			U >>=19;\
			V >>=19;\
			if((Y1|Y2|U|V)&256)\
			{\
				if(Y1>255)   Y1=255;\
				else if(Y1<0)Y1=0;\
				if(Y2>255)   Y2=255;\
				else if(Y2<0)Y2=0;\
				if(U>255)    U=255;\
				else if(U<0) U=0;\
				if(V>255)    V=255;\
				else if(V<0) V=0;\
			}
                        
#define YSCALE_YUV_2_RGBX_C(type) \
			YSCALE_YUV_2_PACKEDX_C(type)\
			r = c->table_rV[V];\
			g = c->table_gU[U] + c->table_gV[V];\
			b = c->table_bU[U];\

#define YSCALE_YUV_2_PACKED2_C \
		for(i=0; i<(dstW>>1); i++){\
			const int i2= 2*i;\
			int Y1= (buf0[i2  ]*yalpha1+buf1[i2  ]*yalpha)>>19;\
			int Y2= (buf0[i2+1]*yalpha1+buf1[i2+1]*yalpha)>>19;\
			int U= (uvbuf0[i     ]*uvalpha1+uvbuf1[i     ]*uvalpha)>>19;\
			int V= (uvbuf0[i+2048]*uvalpha1+uvbuf1[i+2048]*uvalpha)>>19;\

#define YSCALE_YUV_2_RGB2_C(type) \
			YSCALE_YUV_2_PACKED2_C\
			type *r, *b, *g;\
			r = c->table_rV[V];\
			g = c->table_gU[U] + c->table_gV[V];\
			b = c->table_bU[U];\

#define YSCALE_YUV_2_PACKED1_C \
		for(i=0; i<(dstW>>1); i++){\
			const int i2= 2*i;\
			int Y1= buf0[i2  ]>>7;\
			int Y2= buf0[i2+1]>>7;\
			int U= (uvbuf1[i     ])>>7;\
			int V= (uvbuf1[i+2048])>>7;\

#define YSCALE_YUV_2_RGB1_C(type) \
			YSCALE_YUV_2_PACKED1_C\
			type *r, *b, *g;\
			r = c->table_rV[V];\
			g = c->table_gU[U] + c->table_gV[V];\
			b = c->table_bU[U];\

#define YSCALE_YUV_2_PACKED1B_C \
		for(i=0; i<(dstW>>1); i++){\
			const int i2= 2*i;\
			int Y1= buf0[i2  ]>>7;\
			int Y2= buf0[i2+1]>>7;\
			int U= (uvbuf0[i     ] + uvbuf1[i     ])>>8;\
			int V= (uvbuf0[i+2048] + uvbuf1[i+2048])>>8;\

#define YSCALE_YUV_2_RGB1B_C(type) \
			YSCALE_YUV_2_PACKED1B_C\
			type *r, *b, *g;\
			r = c->table_rV[V];\
			g = c->table_gU[U] + c->table_gV[V];\
			b = c->table_bU[U];\

#define YSCALE_YUV_2_ANYRGB_C(func, func2)\
	switch(c->dstFormat)\
	{\
	case IMGFMT_BGR32:\
	case IMGFMT_RGB32:\
		func(uint32_t)\
			((uint32_t*)dest)[i2+0]= r[Y1] + g[Y1] + b[Y1];\
			((uint32_t*)dest)[i2+1]= r[Y2] + g[Y2] + b[Y2];\
		}		\
		break;\
	case IMGFMT_RGB24:\
		func(uint8_t)\
			((uint8_t*)dest)[0]= r[Y1];\
			((uint8_t*)dest)[1]= g[Y1];\
			((uint8_t*)dest)[2]= b[Y1];\
			((uint8_t*)dest)[3]= r[Y2];\
			((uint8_t*)dest)[4]= g[Y2];\
			((uint8_t*)dest)[5]= b[Y2];\
			dest+=6;\
		}\
		break;\
	case IMGFMT_BGR24:\
		func(uint8_t)\
			((uint8_t*)dest)[0]= b[Y1];\
			((uint8_t*)dest)[1]= g[Y1];\
			((uint8_t*)dest)[2]= r[Y1];\
			((uint8_t*)dest)[3]= b[Y2];\
			((uint8_t*)dest)[4]= g[Y2];\
			((uint8_t*)dest)[5]= r[Y2];\
			dest+=6;\
		}\
		break;\
	case IMGFMT_RGB16:\
	case IMGFMT_BGR16:\
		{\
			const int dr1= dither_2x2_8[y&1    ][0];\
			const int dg1= dither_2x2_4[y&1    ][0];\
			const int db1= dither_2x2_8[(y&1)^1][0];\
			const int dr2= dither_2x2_8[y&1    ][1];\
			const int dg2= dither_2x2_4[y&1    ][1];\
			const int db2= dither_2x2_8[(y&1)^1][1];\
			func(uint16_t)\
				((uint16_t*)dest)[i2+0]= r[Y1+dr1] + g[Y1+dg1] + b[Y1+db1];\
				((uint16_t*)dest)[i2+1]= r[Y2+dr2] + g[Y2+dg2] + b[Y2+db2];\
			}\
		}\
		break;\
	case IMGFMT_RGB15:\
	case IMGFMT_BGR15:\
		{\
			const int dr1= dither_2x2_8[y&1    ][0];\
			const int dg1= dither_2x2_8[y&1    ][1];\
			const int db1= dither_2x2_8[(y&1)^1][0];\
			const int dr2= dither_2x2_8[y&1    ][1];\
			const int dg2= dither_2x2_8[y&1    ][0];\
			const int db2= dither_2x2_8[(y&1)^1][1];\
			func(uint16_t)\
				((uint16_t*)dest)[i2+0]= r[Y1+dr1] + g[Y1+dg1] + b[Y1+db1];\
				((uint16_t*)dest)[i2+1]= r[Y2+dr2] + g[Y2+dg2] + b[Y2+db2];\
			}\
		}\
		break;\
	case IMGFMT_RGB8:\
	case IMGFMT_BGR8:\
		{\
			const uint8_t * const d64= dither_8x8_73[y&7];\
			const uint8_t * const d32= dither_8x8_32[y&7];\
			func(uint8_t)\
				((uint8_t*)dest)[i2+0]= r[Y1+d32[(i2+0)&7]] + g[Y1+d32[(i2+0)&7]] + b[Y1+d64[(i2+0)&7]];\
				((uint8_t*)dest)[i2+1]= r[Y2+d32[(i2+1)&7]] + g[Y2+d32[(i2+1)&7]] + b[Y2+d64[(i2+1)&7]];\
			}\
		}\
		break;\
	case IMGFMT_RGB4:\
	case IMGFMT_BGR4:\
		{\
			const uint8_t * const d64= dither_8x8_73 [y&7];\
			const uint8_t * const d128=dither_8x8_220[y&7];\
			func(uint8_t)\
				((uint8_t*)dest)[i]= r[Y1+d128[(i2+0)&7]] + g[Y1+d64[(i2+0)&7]] + b[Y1+d128[(i2+0)&7]]\
				                 + ((r[Y2+d128[(i2+1)&7]] + g[Y2+d64[(i2+1)&7]] + b[Y2+d128[(i2+1)&7]])<<4);\
			}\
		}\
		break;\
	case IMGFMT_RG4B:\
	case IMGFMT_BG4B:\
		{\
			const uint8_t * const d64= dither_8x8_73 [y&7];\
			const uint8_t * const d128=dither_8x8_220[y&7];\
			func(uint8_t)\
				((uint8_t*)dest)[i2+0]= r[Y1+d128[(i2+0)&7]] + g[Y1+d64[(i2+0)&7]] + b[Y1+d128[(i2+0)&7]];\
				((uint8_t*)dest)[i2+1]= r[Y2+d128[(i2+1)&7]] + g[Y2+d64[(i2+1)&7]] + b[Y2+d128[(i2+1)&7]];\
			}\
		}\
		break;\
	case IMGFMT_RGB1:\
	case IMGFMT_BGR1:\
		{\
			const uint8_t * const d128=dither_8x8_220[y&7];\
			uint8_t *g= c->table_gU[128] + c->table_gV[128];\
			for(i=0; i<dstW-7; i+=8){\
				int acc;\
				acc =       g[((buf0[i  ]*yalpha1+buf1[i  ]*yalpha)>>19) + d128[0]];\
				acc+= acc + g[((buf0[i+1]*yalpha1+buf1[i+1]*yalpha)>>19) + d128[1]];\
				acc+= acc + g[((buf0[i+2]*yalpha1+buf1[i+2]*yalpha)>>19) + d128[2]];\
				acc+= acc + g[((buf0[i+3]*yalpha1+buf1[i+3]*yalpha)>>19) + d128[3]];\
				acc+= acc + g[((buf0[i+4]*yalpha1+buf1[i+4]*yalpha)>>19) + d128[4]];\
				acc+= acc + g[((buf0[i+5]*yalpha1+buf1[i+5]*yalpha)>>19) + d128[5]];\
				acc+= acc + g[((buf0[i+6]*yalpha1+buf1[i+6]*yalpha)>>19) + d128[6]];\
				acc+= acc + g[((buf0[i+7]*yalpha1+buf1[i+7]*yalpha)>>19) + d128[7]];\
				((uint8_t*)dest)[0]= acc;\
				dest++;\
			}\
\
/*\
((uint8_t*)dest)-= dstW>>4;\
{\
			int acc=0;\
			int left=0;\
			static int top[1024];\
			static int last_new[1024][1024];\
			static int last_in3[1024][1024];\
			static int drift[1024][1024];\
			int topLeft=0;\
			int shift=0;\
			int count=0;\
			const uint8_t * const d128=dither_8x8_220[y&7];\
			int error_new=0;\
			int error_in3=0;\
			int f=0;\
			\
			for(i=dstW>>1; i<dstW; i++){\
				int in= ((buf0[i  ]*yalpha1+buf1[i  ]*yalpha)>>19);\
				int in2 = (76309 * (in - 16) + 32768) >> 16;\
				int in3 = (in2 < 0) ? 0 : ((in2 > 255) ? 255 : in2);\
				int old= (left*7 + topLeft + top[i]*5 + top[i+1]*3)/20 + in3\
					+ (last_new[y][i] - in3)*f/256;\
				int new= old> 128 ? 255 : 0;\
\
				error_new+= ABS(last_new[y][i] - new);\
				error_in3+= ABS(last_in3[y][i] - in3);\
				f= error_new - error_in3*4;\
				if(f<0) f=0;\
				if(f>256) f=256;\
\
				topLeft= top[i];\
				left= top[i]= old - new;\
				last_new[y][i]= new;\
				last_in3[y][i]= in3;\
\
				acc+= acc + (new&1);\
				if((i&7)==6){\
					((uint8_t*)dest)[0]= acc;\
					((uint8_t*)dest)++;\
				}\
			}\
}\
*/\
		}\
		break;\
	case IMGFMT_YUY2:\
		func2\
			((uint8_t*)dest)[2*i2+0]= Y1;\
			((uint8_t*)dest)[2*i2+1]= U;\
			((uint8_t*)dest)[2*i2+2]= Y2;\
			((uint8_t*)dest)[2*i2+3]= V;\
		}		\
		break;\
	case IMGFMT_UYVY:\
		func2\
			((uint8_t*)dest)[2*i2+0]= U;\
			((uint8_t*)dest)[2*i2+1]= Y1;\
			((uint8_t*)dest)[2*i2+2]= V;\
			((uint8_t*)dest)[2*i2+3]= Y2;\
		}		\
		break;\
	}\


static inline void yuv2packedXinC(SwsContext *c, int16_t *lumFilter, int16_t **lumSrc, int lumFilterSize,
				    int16_t *chrFilter, int16_t **chrSrc, int chrFilterSize,
				    uint8_t *dest, int dstW, int y)
{
	int i;
	switch(c->dstFormat)
	{
	case IMGFMT_RGB32:
	case IMGFMT_BGR32:
		YSCALE_YUV_2_RGBX_C(uint32_t)
			((uint32_t*)dest)[i2+0]= r[Y1] + g[Y1] + b[Y1];
			((uint32_t*)dest)[i2+1]= r[Y2] + g[Y2] + b[Y2];
		}
		break;
	case IMGFMT_RGB24:
		YSCALE_YUV_2_RGBX_C(uint8_t)
			((uint8_t*)dest)[0]= r[Y1];
			((uint8_t*)dest)[1]= g[Y1];
			((uint8_t*)dest)[2]= b[Y1];
			((uint8_t*)dest)[3]= r[Y2];
			((uint8_t*)dest)[4]= g[Y2];
			((uint8_t*)dest)[5]= b[Y2];
			dest+=6;
		}
		break;
	case IMGFMT_BGR24:
		YSCALE_YUV_2_RGBX_C(uint8_t)
			((uint8_t*)dest)[0]= b[Y1];
			((uint8_t*)dest)[1]= g[Y1];
			((uint8_t*)dest)[2]= r[Y1];
			((uint8_t*)dest)[3]= b[Y2];
			((uint8_t*)dest)[4]= g[Y2];
			((uint8_t*)dest)[5]= r[Y2];
			dest+=6;
		}
		break;
	case IMGFMT_RGB16:
	case IMGFMT_BGR16:
		{
			const int dr1= dither_2x2_8[y&1    ][0];
			const int dg1= dither_2x2_4[y&1    ][0];
			const int db1= dither_2x2_8[(y&1)^1][0];
			const int dr2= dither_2x2_8[y&1    ][1];
			const int dg2= dither_2x2_4[y&1    ][1];
			const int db2= dither_2x2_8[(y&1)^1][1];
			YSCALE_YUV_2_RGBX_C(uint16_t)
				((uint16_t*)dest)[i2+0]= r[Y1+dr1] + g[Y1+dg1] + b[Y1+db1];
				((uint16_t*)dest)[i2+1]= r[Y2+dr2] + g[Y2+dg2] + b[Y2+db2];
			}
		}
		break;
	case IMGFMT_RGB15:
	case IMGFMT_BGR15:
		{
			const int dr1= dither_2x2_8[y&1    ][0];
			const int dg1= dither_2x2_8[y&1    ][1];
			const int db1= dither_2x2_8[(y&1)^1][0];
			const int dr2= dither_2x2_8[y&1    ][1];
			const int dg2= dither_2x2_8[y&1    ][0];
			const int db2= dither_2x2_8[(y&1)^1][1];
			YSCALE_YUV_2_RGBX_C(uint16_t)
				((uint16_t*)dest)[i2+0]= r[Y1+dr1] + g[Y1+dg1] + b[Y1+db1];
				((uint16_t*)dest)[i2+1]= r[Y2+dr2] + g[Y2+dg2] + b[Y2+db2];
			}
		}
		break;
	case IMGFMT_RGB8:
	case IMGFMT_BGR8:
		{
			const uint8_t * const d64= dither_8x8_73[y&7];
			const uint8_t * const d32= dither_8x8_32[y&7];
			YSCALE_YUV_2_RGBX_C(uint8_t)
				((uint8_t*)dest)[i2+0]= r[Y1+d32[(i2+0)&7]] + g[Y1+d32[(i2+0)&7]] + b[Y1+d64[(i2+0)&7]];
				((uint8_t*)dest)[i2+1]= r[Y2+d32[(i2+1)&7]] + g[Y2+d32[(i2+1)&7]] + b[Y2+d64[(i2+1)&7]];
			}
		}
		break;
	case IMGFMT_RGB4:
	case IMGFMT_BGR4:
		{
			const uint8_t * const d64= dither_8x8_73 [y&7];
			const uint8_t * const d128=dither_8x8_220[y&7];
			YSCALE_YUV_2_RGBX_C(uint8_t)
				((uint8_t*)dest)[i]= r[Y1+d128[(i2+0)&7]] + g[Y1+d64[(i2+0)&7]] + b[Y1+d128[(i2+0)&7]]
				                  +((r[Y2+d128[(i2+1)&7]] + g[Y2+d64[(i2+1)&7]] + b[Y2+d128[(i2+1)&7]])<<4);
			}
		}
		break;
	case IMGFMT_RG4B:
	case IMGFMT_BG4B:
		{
			const uint8_t * const d64= dither_8x8_73 [y&7];
			const uint8_t * const d128=dither_8x8_220[y&7];
			YSCALE_YUV_2_RGBX_C(uint8_t)
				((uint8_t*)dest)[i2+0]= r[Y1+d128[(i2+0)&7]] + g[Y1+d64[(i2+0)&7]] + b[Y1+d128[(i2+0)&7]];
				((uint8_t*)dest)[i2+1]= r[Y2+d128[(i2+1)&7]] + g[Y2+d64[(i2+1)&7]] + b[Y2+d128[(i2+1)&7]];
			}
		}
		break;
	case IMGFMT_RGB1:
	case IMGFMT_BGR1:
		{
			const uint8_t * const d128=dither_8x8_220[y&7];
			uint8_t *g= c->table_gU[128] + c->table_gV[128];
			int acc=0;
			for(i=0; i<dstW-1; i+=2){
				int j;
				int Y1=1<<18;
				int Y2=1<<18;

				for(j=0; j<lumFilterSize; j++)
				{
					Y1 += lumSrc[j][i] * lumFilter[j];
					Y2 += lumSrc[j][i+1] * lumFilter[j];
				}
				Y1>>=19;
				Y2>>=19;
				if((Y1|Y2)&256)
				{
					if(Y1>255)   Y1=255;
					else if(Y1<0)Y1=0;
					if(Y2>255)   Y2=255;
					else if(Y2<0)Y2=0;
				}
				acc+= acc + g[Y1+d128[(i+0)&7]];
				acc+= acc + g[Y2+d128[(i+1)&7]];
				if((i&7)==6){
					((uint8_t*)dest)[0]= acc;
					dest++;
				}
			}
		}
		break;
	case IMGFMT_YUY2:
		YSCALE_YUV_2_PACKEDX_C(void)
			((uint8_t*)dest)[2*i2+0]= Y1;
			((uint8_t*)dest)[2*i2+1]= U;
			((uint8_t*)dest)[2*i2+2]= Y2;
			((uint8_t*)dest)[2*i2+3]= V;
		}
                break;
	case IMGFMT_UYVY:
		YSCALE_YUV_2_PACKEDX_C(void)
			((uint8_t*)dest)[2*i2+0]= U;
			((uint8_t*)dest)[2*i2+1]= Y1;
			((uint8_t*)dest)[2*i2+2]= V;
			((uint8_t*)dest)[2*i2+3]= Y2;
		}
                break;
	}
}


//Note: we have C, X86, MMX, MMX2, 3DNOW version therse no 3DNOW+MMX2 one
//Plain C versions
#if !defined (HAVE_MMX) || defined (RUNTIME_CPUDETECT)
#define COMPILE_C
#endif

#ifdef ARCH_POWERPC
#if defined (HAVE_ALTIVEC) || defined (RUNTIME_CPUDETECT)
#define COMPILE_ALTIVEC
#endif //HAVE_ALTIVEC
#endif //ARCH_POWERPC

#if defined(ARCH_X86) || defined(ARCH_X86_64)

#if (defined (HAVE_MMX) && !defined (HAVE_3DNOW) && !defined (HAVE_MMX2)) || defined (RUNTIME_CPUDETECT)
#define COMPILE_MMX
#endif

#if defined (HAVE_MMX2) || defined (RUNTIME_CPUDETECT)
#define COMPILE_MMX2
#endif

#if (defined (HAVE_3DNOW) && !defined (HAVE_MMX2)) || defined (RUNTIME_CPUDETECT)
#define COMPILE_3DNOW
#endif
#endif //ARCH_X86 || ARCH_X86_64

#undef HAVE_MMX
#undef HAVE_MMX2
#undef HAVE_3DNOW

#ifdef COMPILE_C
#undef HAVE_MMX
#undef HAVE_MMX2
#undef HAVE_3DNOW
#undef HAVE_ALTIVEC
#define RENAME(a) a ## _C
#include "swscale_template.c"
#endif

#ifdef ARCH_POWERPC
#ifdef COMPILE_ALTIVEC
#undef RENAME
#define HAVE_ALTIVEC
#define RENAME(a) a ## _altivec
#include "swscale_template.c"
#endif
#endif //ARCH_POWERPC

#if defined(ARCH_X86) || defined(ARCH_X86_64)

//X86 versions
/*
#undef RENAME
#undef HAVE_MMX
#undef HAVE_MMX2
#undef HAVE_3DNOW
#define ARCH_X86
#define RENAME(a) a ## _X86
#include "swscale_template.c"
*/
//MMX versions
#ifdef COMPILE_MMX
#undef RENAME
#define HAVE_MMX
#undef HAVE_MMX2
#undef HAVE_3DNOW
#define RENAME(a) a ## _MMX
#include "swscale_template.c"
#endif

//MMX2 versions
#ifdef COMPILE_MMX2
#undef RENAME
#define HAVE_MMX
#define HAVE_MMX2
#undef HAVE_3DNOW
#define RENAME(a) a ## _MMX2
#include "swscale_template.c"
#endif

//3DNOW versions
#ifdef COMPILE_3DNOW
#undef RENAME
#define HAVE_MMX
#undef HAVE_MMX2
#define HAVE_3DNOW
#define RENAME(a) a ## _3DNow
#include "swscale_template.c"
#endif

#endif //ARCH_X86 || ARCH_X86_64

// minor note: the HAVE_xyz is messed up after that line so don't use it

static double getSplineCoeff(double a, double b, double c, double d, double dist)
{
//	printf("%f %f %f %f %f\n", a,b,c,d,dist);
	if(dist<=1.0) 	return ((d*dist + c)*dist + b)*dist +a;
	else		return getSplineCoeff(	0.0, 
						 b+ 2.0*c + 3.0*d,
						        c + 3.0*d,
						-b- 3.0*c - 6.0*d,
						dist-1.0);
}

static inline void initFilter(int16_t **outFilter, int16_t **filterPos, int *outFilterSize, int xInc,
			      int srcW, int dstW, int filterAlign, int one, int flags,
			      SwsVector *srcFilter, SwsVector *dstFilter, double param[2])
{
	int i;
	int filterSize;
	int filter2Size;
	int minFilterSize;
	double *filter=NULL;
	double *filter2=NULL;
#if defined(ARCH_X86) || defined(ARCH_X86_64)
	if(flags & SWS_CPU_CAPS_MMX)
		asm volatile("emms\n\t"::: "memory"); //FIXME this shouldnt be required but it IS (even for non mmx versions)
#endif

	// Note the +1 is for the MMXscaler which reads over the end
	*filterPos = av_malloc((dstW+1)*sizeof(int16_t));

	if(ABS(xInc - 0x10000) <10) // unscaled
	{
		int i;
		filterSize= 1;
		filter= av_malloc(dstW*sizeof(double)*filterSize);
		for(i=0; i<dstW*filterSize; i++) filter[i]=0;

		for(i=0; i<dstW; i++)
		{
			filter[i*filterSize]=1;
			(*filterPos)[i]=i;
		}

	}
	else if(flags&SWS_POINT) // lame looking point sampling mode
	{
		int i;
		int xDstInSrc;
		filterSize= 1;
		filter= av_malloc(dstW*sizeof(double)*filterSize);
		
		xDstInSrc= xInc/2 - 0x8000;
		for(i=0; i<dstW; i++)
		{
			int xx= (xDstInSrc - ((filterSize-1)<<15) + (1<<15))>>16;

			(*filterPos)[i]= xx;
			filter[i]= 1.0;
			xDstInSrc+= xInc;
		}
	}
	else if((xInc <= (1<<16) && (flags&SWS_AREA)) || (flags&SWS_FAST_BILINEAR)) // bilinear upscale
	{
		int i;
		int xDstInSrc;
		if     (flags&SWS_BICUBIC) filterSize= 4;
		else if(flags&SWS_X      ) filterSize= 4;
		else			   filterSize= 2; // SWS_BILINEAR / SWS_AREA 
		filter= av_malloc(dstW*sizeof(double)*filterSize);

		xDstInSrc= xInc/2 - 0x8000;
		for(i=0; i<dstW; i++)
		{
			int xx= (xDstInSrc - ((filterSize-1)<<15) + (1<<15))>>16;
			int j;

			(*filterPos)[i]= xx;
				//Bilinear upscale / linear interpolate / Area averaging
				for(j=0; j<filterSize; j++)
				{
					double d= ABS((xx<<16) - xDstInSrc)/(double)(1<<16);
					double coeff= 1.0 - d;
					if(coeff<0) coeff=0;
					filter[i*filterSize + j]= coeff;
					xx++;
				}
			xDstInSrc+= xInc;
		}
	}
	else
	{
		double xDstInSrc;
		double sizeFactor, filterSizeInSrc;
		const double xInc1= (double)xInc / (double)(1<<16);

		if     (flags&SWS_BICUBIC)	sizeFactor= 4.0;
		else if(flags&SWS_X)		sizeFactor= 8.0;
		else if(flags&SWS_AREA)		sizeFactor= 1.0; //downscale only, for upscale it is bilinear
		else if(flags&SWS_GAUSS)	sizeFactor= 8.0;   // infinite ;)
		else if(flags&SWS_LANCZOS)	sizeFactor= param[0] != SWS_PARAM_DEFAULT ? 2.0*param[0] : 6.0;
		else if(flags&SWS_SINC)		sizeFactor= 20.0; // infinite ;)
		else if(flags&SWS_SPLINE)	sizeFactor= 20.0;  // infinite ;)
		else if(flags&SWS_BILINEAR)	sizeFactor= 2.0;
		else {
			sizeFactor= 0.0; //GCC warning killer
			ASSERT(0)
		}
		
		if(xInc1 <= 1.0)	filterSizeInSrc= sizeFactor; // upscale
		else			filterSizeInSrc= sizeFactor*srcW / (double)dstW;

		filterSize= (int)ceil(1 + filterSizeInSrc); // will be reduced later if possible
		if(filterSize > srcW-2) filterSize=srcW-2;

		filter= av_malloc(dstW*sizeof(double)*filterSize);

		xDstInSrc= xInc1 / 2.0 - 0.5;
		for(i=0; i<dstW; i++)
		{
			int xx= (int)(xDstInSrc - (filterSize-1)*0.5 + 0.5);
			int j;
			(*filterPos)[i]= xx;
			for(j=0; j<filterSize; j++)
			{
				double d= ABS(xx - xDstInSrc)/filterSizeInSrc*sizeFactor;
				double coeff;
				if(flags & SWS_BICUBIC)
				{
					double B= param[0] != SWS_PARAM_DEFAULT ? param[0] : 0.0;
					double C= param[1] != SWS_PARAM_DEFAULT ? param[1] : 0.6;

					if(d<1.0) 
						coeff = (12-9*B-6*C)*d*d*d + (-18+12*B+6*C)*d*d + 6-2*B;
					else if(d<2.0)
						coeff = (-B-6*C)*d*d*d + (6*B+30*C)*d*d + (-12*B-48*C)*d +8*B+24*C;
					else
						coeff=0.0;
				}
/*				else if(flags & SWS_X)
				{
					double p= param ? param*0.01 : 0.3;
					coeff = d ? sin(d*PI)/(d*PI) : 1.0;
					coeff*= pow(2.0, - p*d*d);
				}*/
				else if(flags & SWS_X)
				{
					double A= param[0] != SWS_PARAM_DEFAULT ? param[0] : 1.0;
					
					if(d<1.0)
						coeff = cos(d*PI);
					else
						coeff=-1.0;
					if(coeff<0.0) 	coeff= -pow(-coeff, A);
					else		coeff=  pow( coeff, A);
					coeff= coeff*0.5 + 0.5;
				}
				else if(flags & SWS_AREA)
				{
					double srcPixelSize= 1.0/xInc1;
					if(d + srcPixelSize/2 < 0.5) coeff= 1.0;
					else if(d - srcPixelSize/2 < 0.5) coeff= (0.5-d)/srcPixelSize + 0.5;
					else coeff=0.0;
				}
				else if(flags & SWS_GAUSS)
				{
					double p= param[0] != SWS_PARAM_DEFAULT ? param[0] : 3.0;
					coeff = pow(2.0, - p*d*d);
				}
				else if(flags & SWS_SINC)
				{
					coeff = d ? sin(d*PI)/(d*PI) : 1.0;
				}
				else if(flags & SWS_LANCZOS)
				{
					double p= param[0] != SWS_PARAM_DEFAULT ? param[0] : 3.0; 
					coeff = d ? sin(d*PI)*sin(d*PI/p)/(d*d*PI*PI/p) : 1.0;
					if(d>p) coeff=0;
				}
				else if(flags & SWS_BILINEAR)
				{
					coeff= 1.0 - d;
					if(coeff<0) coeff=0;
				}
				else if(flags & SWS_SPLINE)
				{
					double p=-2.196152422706632;
					coeff = getSplineCoeff(1.0, 0.0, p, -p-1.0, d);
				}
				else {
					coeff= 0.0; //GCC warning killer
					ASSERT(0)
				}

				filter[i*filterSize + j]= coeff;
				xx++;
			}
			xDstInSrc+= xInc1;
		}
	}

	/* apply src & dst Filter to filter -> filter2
	   av_free(filter);
	*/
	ASSERT(filterSize>0)
	filter2Size= filterSize;
	if(srcFilter) filter2Size+= srcFilter->length - 1;
	if(dstFilter) filter2Size+= dstFilter->length - 1;
	ASSERT(filter2Size>0)
	filter2= av_malloc(filter2Size*dstW*sizeof(double));

	for(i=0; i<dstW; i++)
	{
		int j;
		SwsVector scaleFilter;
		SwsVector *outVec;

		scaleFilter.coeff= filter + i*filterSize;
		scaleFilter.length= filterSize;

		if(srcFilter) outVec= sws_getConvVec(srcFilter, &scaleFilter);
		else	      outVec= &scaleFilter;

		ASSERT(outVec->length == filter2Size)
		//FIXME dstFilter

		for(j=0; j<outVec->length; j++)
		{
			filter2[i*filter2Size + j]= outVec->coeff[j];
		}

		(*filterPos)[i]+= (filterSize-1)/2 - (filter2Size-1)/2;

		if(outVec != &scaleFilter) sws_freeVec(outVec);
	}
	av_free(filter); filter=NULL;

	/* try to reduce the filter-size (step1 find size and shift left) */
	// Assume its near normalized (*0.5 or *2.0 is ok but * 0.001 is not)
	minFilterSize= 0;
	for(i=dstW-1; i>=0; i--)
	{
		int min= filter2Size;
		int j;
		double cutOff=0.0;

		/* get rid off near zero elements on the left by shifting left */
		for(j=0; j<filter2Size; j++)
		{
			int k;
			cutOff += ABS(filter2[i*filter2Size]);

			if(cutOff > SWS_MAX_REDUCE_CUTOFF) break;

			/* preserve Monotonicity because the core can't handle the filter otherwise */
			if(i<dstW-1 && (*filterPos)[i] >= (*filterPos)[i+1]) break;

			// Move filter coeffs left
			for(k=1; k<filter2Size; k++)
				filter2[i*filter2Size + k - 1]= filter2[i*filter2Size + k];
			filter2[i*filter2Size + k - 1]= 0.0;
			(*filterPos)[i]++;
		}

		cutOff=0.0;
		/* count near zeros on the right */
		for(j=filter2Size-1; j>0; j--)
		{
			cutOff += ABS(filter2[i*filter2Size + j]);

			if(cutOff > SWS_MAX_REDUCE_CUTOFF) break;
			min--;
		}

		if(min>minFilterSize) minFilterSize= min;
	}

        if (flags & SWS_CPU_CAPS_ALTIVEC) {
          // we can handle the special case 4,
          // so we don't want to go to the full 8
          if (minFilterSize < 5)
            filterAlign = 4;

          // we really don't want to waste our time
          // doing useless computation, so fall-back on
          // the scalar C code for very small filter.
          // vectorizing is worth it only if you have
          // decent-sized vector.
          if (minFilterSize < 3)
            filterAlign = 1;
        }

	ASSERT(minFilterSize > 0)
	filterSize= (minFilterSize +(filterAlign-1)) & (~(filterAlign-1));
	ASSERT(filterSize > 0)
	filter= av_malloc(filterSize*dstW*sizeof(double));
	*outFilterSize= filterSize;

	if(flags&SWS_PRINT_INFO)
		MSG_V("SwScaler: reducing / aligning filtersize %d -> %d\n", filter2Size, filterSize);
	/* try to reduce the filter-size (step2 reduce it) */
	for(i=0; i<dstW; i++)
	{
		int j;

		for(j=0; j<filterSize; j++)
		{
			if(j>=filter2Size) filter[i*filterSize + j]= 0.0;
			else		   filter[i*filterSize + j]= filter2[i*filter2Size + j];
		}
	}
	av_free(filter2); filter2=NULL;
	

	//FIXME try to align filterpos if possible

	//fix borders
	for(i=0; i<dstW; i++)
	{
		int j;
		if((*filterPos)[i] < 0)
		{
			// Move filter coeffs left to compensate for filterPos
			for(j=1; j<filterSize; j++)
			{
				int left= MAX(j + (*filterPos)[i], 0);
				filter[i*filterSize + left] += filter[i*filterSize + j];
				filter[i*filterSize + j]=0;
			}
			(*filterPos)[i]= 0;
		}

		if((*filterPos)[i] + filterSize > srcW)
		{
			int shift= (*filterPos)[i] + filterSize - srcW;
			// Move filter coeffs right to compensate for filterPos
			for(j=filterSize-2; j>=0; j--)
			{
				int right= MIN(j + shift, filterSize-1);
				filter[i*filterSize +right] += filter[i*filterSize +j];
				filter[i*filterSize +j]=0;
			}
			(*filterPos)[i]= srcW - filterSize;
		}
	}

	// Note the +1 is for the MMXscaler which reads over the end
	/* align at 16 for AltiVec (needed by hScale_altivec_real) */
	*outFilter= av_malloc(*outFilterSize*(dstW+1)*sizeof(int16_t));
	memset(*outFilter, 0, *outFilterSize*(dstW+1)*sizeof(int16_t));

	/* Normalize & Store in outFilter */
	for(i=0; i<dstW; i++)
	{
		int j;
		double error=0;
		double sum=0;
		double scale= one;

		for(j=0; j<filterSize; j++)
		{
			sum+= filter[i*filterSize + j];
		}
		scale/= sum;
		for(j=0; j<*outFilterSize; j++)
		{
			double v= filter[i*filterSize + j]*scale + error;
			int intV= floor(v + 0.5);
			(*outFilter)[i*(*outFilterSize) + j]= intV;
			error = v - intV;
		}
	}
	
	(*filterPos)[dstW]= (*filterPos)[dstW-1]; // the MMX scaler will read over the end
	for(i=0; i<*outFilterSize; i++)
	{
		int j= dstW*(*outFilterSize);
		(*outFilter)[j + i]= (*outFilter)[j + i - (*outFilterSize)];
	}

	av_free(filter);
}

#if defined(ARCH_X86) || defined(ARCH_X86_64)
static void initMMX2HScaler(int dstW, int xInc, uint8_t *funnyCode, int16_t *filter, int32_t *filterPos, int numSplits)
{
	uint8_t *fragmentA;
	long imm8OfPShufW1A;
	long imm8OfPShufW2A;
	long fragmentLengthA;
	uint8_t *fragmentB;
	long imm8OfPShufW1B;
	long imm8OfPShufW2B;
	long fragmentLengthB;
	int fragmentPos;

	int xpos, i;

	// create an optimized horizontal scaling routine

	//code fragment

	asm volatile(
		"jmp 9f				\n\t"
	// Begin
		"0:				\n\t"
		"movq (%%"REG_d", %%"REG_a"), %%mm3\n\t" 
		"movd (%%"REG_c", %%"REG_S"), %%mm0\n\t" 
		"movd 1(%%"REG_c", %%"REG_S"), %%mm1\n\t"
		"punpcklbw %%mm7, %%mm1		\n\t"
		"punpcklbw %%mm7, %%mm0		\n\t"
		"pshufw $0xFF, %%mm1, %%mm1	\n\t"
		"1:				\n\t"
		"pshufw $0xFF, %%mm0, %%mm0	\n\t"
		"2:				\n\t"
		"psubw %%mm1, %%mm0		\n\t"
		"movl 8(%%"REG_b", %%"REG_a"), %%esi\n\t"
		"pmullw %%mm3, %%mm0		\n\t"
		"psllw $7, %%mm1		\n\t"
		"paddw %%mm1, %%mm0		\n\t"

		"movq %%mm0, (%%"REG_D", %%"REG_a")\n\t"

		"add $8, %%"REG_a"		\n\t"
	// End
		"9:				\n\t"
//		"int $3\n\t"
		"lea 0b, %0			\n\t"
		"lea 1b, %1			\n\t"
		"lea 2b, %2			\n\t"
		"dec %1				\n\t"
		"dec %2				\n\t"
		"sub %0, %1			\n\t"
		"sub %0, %2			\n\t"
		"lea 9b, %3			\n\t"
		"sub %0, %3			\n\t"


		:"=r" (fragmentA), "=r" (imm8OfPShufW1A), "=r" (imm8OfPShufW2A),
		"=r" (fragmentLengthA)
	);

	asm volatile(
		"jmp 9f				\n\t"
	// Begin
		"0:				\n\t"
		"movq (%%"REG_d", %%"REG_a"), %%mm3\n\t" 
		"movd (%%"REG_c", %%"REG_S"), %%mm0\n\t" 
		"punpcklbw %%mm7, %%mm0		\n\t"
		"pshufw $0xFF, %%mm0, %%mm1	\n\t"
		"1:				\n\t"
		"pshufw $0xFF, %%mm0, %%mm0	\n\t"
		"2:				\n\t"
		"psubw %%mm1, %%mm0		\n\t"
		"movl 8(%%"REG_b", %%"REG_a"), %%esi\n\t"
		"pmullw %%mm3, %%mm0		\n\t"
		"psllw $7, %%mm1		\n\t"
		"paddw %%mm1, %%mm0		\n\t"

		"movq %%mm0, (%%"REG_D", %%"REG_a")\n\t"

		"add $8, %%"REG_a"		\n\t"
	// End
		"9:				\n\t"
//		"int $3\n\t"
		"lea 0b, %0			\n\t"
		"lea 1b, %1			\n\t"
		"lea 2b, %2			\n\t"
		"dec %1				\n\t"
		"dec %2				\n\t"
		"sub %0, %1			\n\t"
		"sub %0, %2			\n\t"
		"lea 9b, %3			\n\t"
		"sub %0, %3			\n\t"


		:"=r" (fragmentB), "=r" (imm8OfPShufW1B), "=r" (imm8OfPShufW2B),
		"=r" (fragmentLengthB)
	);

	xpos= 0; //lumXInc/2 - 0x8000; // difference between pixel centers
	fragmentPos=0;
	
	for(i=0; i<dstW/numSplits; i++)
	{
		int xx=xpos>>16;

		if((i&3) == 0)
		{
			int a=0;
			int b=((xpos+xInc)>>16) - xx;
			int c=((xpos+xInc*2)>>16) - xx;
			int d=((xpos+xInc*3)>>16) - xx;

			filter[i  ] = (( xpos         & 0xFFFF) ^ 0xFFFF)>>9;
			filter[i+1] = (((xpos+xInc  ) & 0xFFFF) ^ 0xFFFF)>>9;
			filter[i+2] = (((xpos+xInc*2) & 0xFFFF) ^ 0xFFFF)>>9;
			filter[i+3] = (((xpos+xInc*3) & 0xFFFF) ^ 0xFFFF)>>9;
			filterPos[i/2]= xx;

			if(d+1<4)
			{
				int maxShift= 3-(d+1);
				int shift=0;

				memcpy(funnyCode + fragmentPos, fragmentB, fragmentLengthB);

				funnyCode[fragmentPos + imm8OfPShufW1B]=
					(a+1) | ((b+1)<<2) | ((c+1)<<4) | ((d+1)<<6);
				funnyCode[fragmentPos + imm8OfPShufW2B]=
					a | (b<<2) | (c<<4) | (d<<6);

				if(i+3>=dstW) shift=maxShift; //avoid overread
				else if((filterPos[i/2]&3) <= maxShift) shift=filterPos[i/2]&3; //Align

				if(shift && i>=shift)
				{
					funnyCode[fragmentPos + imm8OfPShufW1B]+= 0x55*shift;
					funnyCode[fragmentPos + imm8OfPShufW2B]+= 0x55*shift;
					filterPos[i/2]-=shift;
				}

				fragmentPos+= fragmentLengthB;
			}
			else
			{
				int maxShift= 3-d;
				int shift=0;

				memcpy(funnyCode + fragmentPos, fragmentA, fragmentLengthA);

				funnyCode[fragmentPos + imm8OfPShufW1A]=
				funnyCode[fragmentPos + imm8OfPShufW2A]=
					a | (b<<2) | (c<<4) | (d<<6);

				if(i+4>=dstW) shift=maxShift; //avoid overread
				else if((filterPos[i/2]&3) <= maxShift) shift=filterPos[i/2]&3; //partial align

				if(shift && i>=shift)
				{
					funnyCode[fragmentPos + imm8OfPShufW1A]+= 0x55*shift;
					funnyCode[fragmentPos + imm8OfPShufW2A]+= 0x55*shift;
					filterPos[i/2]-=shift;
				}

				fragmentPos+= fragmentLengthA;
			}

			funnyCode[fragmentPos]= RET;
		}
		xpos+=xInc;
	}
	filterPos[i/2]= xpos>>16; // needed to jump to the next part
}
#endif // ARCH_X86 || ARCH_X86_64

static void globalInit(void){
    // generating tables:
    int i;
    for(i=0; i<768; i++){
	int c= MIN(MAX(i-256, 0), 255);
	clip_table[i]=c;
    }
}

static SwsFunc getSwsFunc(int flags){
    
#ifdef RUNTIME_CPUDETECT
#if defined(ARCH_X86) || defined(ARCH_X86_64)
	// ordered per speed fasterst first
	if(flags & SWS_CPU_CAPS_MMX2)
		return swScale_MMX2;
	else if(flags & SWS_CPU_CAPS_3DNOW)
		return swScale_3DNow;
	else if(flags & SWS_CPU_CAPS_MMX)
		return swScale_MMX;
	else
		return swScale_C;

#else
#ifdef ARCH_POWERPC
	if(flags & SWS_CPU_CAPS_ALTIVEC)
	  return swScale_altivec;
	else
	  return swScale_C;
#endif
	return swScale_C;
#endif
#else //RUNTIME_CPUDETECT
#ifdef HAVE_MMX2
	return swScale_MMX2;
#elif defined (HAVE_3DNOW)
	return swScale_3DNow;
#elif defined (HAVE_MMX)
	return swScale_MMX;
#elif defined (HAVE_ALTIVEC)
	return swScale_altivec;
#else
	return swScale_C;
#endif
#endif //!RUNTIME_CPUDETECT
}

static int PlanarToNV12Wrapper(SwsContext *c, uint8_t* src[], int srcStride[], int srcSliceY,
             int srcSliceH, uint8_t* dstParam[], int dstStride[]){
	uint8_t *dst=dstParam[0] + dstStride[0]*srcSliceY;
	/* Copy Y plane */
	if(dstStride[0]==srcStride[0] && srcStride[0] > 0)
		memcpy(dst, src[0], srcSliceH*dstStride[0]);
	else
	{
		int i;
		uint8_t *srcPtr= src[0];
		uint8_t *dstPtr= dst;
		for(i=0; i<srcSliceH; i++)
		{
			memcpy(dstPtr, srcPtr, c->srcW);
			srcPtr+= srcStride[0];
			dstPtr+= dstStride[0];
		}
	}
	dst = dstParam[1] + dstStride[1]*srcSliceY/2;
	if (c->dstFormat == IMGFMT_NV12)
		interleaveBytes( src[1],src[2],dst,c->srcW/2,srcSliceH/2,srcStride[1],srcStride[2],dstStride[0] );
	else
		interleaveBytes( src[2],src[1],dst,c->srcW/2,srcSliceH/2,srcStride[2],srcStride[1],dstStride[0] );

	return srcSliceH;
}

static int PlanarToYuy2Wrapper(SwsContext *c, uint8_t* src[], int srcStride[], int srcSliceY,
             int srcSliceH, uint8_t* dstParam[], int dstStride[]){
	uint8_t *dst=dstParam[0] + dstStride[0]*srcSliceY;

	yv12toyuy2( src[0],src[1],src[2],dst,c->srcW,srcSliceH,srcStride[0],srcStride[1],dstStride[0] );

	return srcSliceH;
}

static int PlanarToUyvyWrapper(SwsContext *c, uint8_t* src[], int srcStride[], int srcSliceY,
             int srcSliceH, uint8_t* dstParam[], int dstStride[]){
	uint8_t *dst=dstParam[0] + dstStride[0]*srcSliceY;

	yv12touyvy( src[0],src[1],src[2],dst,c->srcW,srcSliceH,srcStride[0],srcStride[1],dstStride[0] );

	return srcSliceH;
}

/* {RGB,BGR}{15,16,24,32} -> {RGB,BGR}{15,16,24,32} */
static int rgb2rgbWrapper(SwsContext *c, uint8_t* src[], int srcStride[], int srcSliceY,
			   int srcSliceH, uint8_t* dst[], int dstStride[]){
	const int srcFormat= c->srcFormat;
	const int dstFormat= c->dstFormat;
	const int srcBpp= ((srcFormat&0xFF) + 7)>>3;
	const int dstBpp= ((dstFormat&0xFF) + 7)>>3;
	const int srcId= (srcFormat&0xFF)>>2; // 1:0, 4:1, 8:2, 15:3, 16:4, 24:6, 32:8 
	const int dstId= (dstFormat&0xFF)>>2;
	void (*conv)(const uint8_t *src, uint8_t *dst, long src_size)=NULL;

	/* BGR -> BGR */
	if(   (isBGR(srcFormat) && isBGR(dstFormat))
	   || (isRGB(srcFormat) && isRGB(dstFormat))){
		switch(srcId | (dstId<<4)){
		case 0x34: conv= rgb16to15; break;
		case 0x36: conv= rgb24to15; break;
		case 0x38: conv= rgb32to15; break;
		case 0x43: conv= rgb15to16; break;
		case 0x46: conv= rgb24to16; break;
		case 0x48: conv= rgb32to16; break;
		case 0x63: conv= rgb15to24; break;
		case 0x64: conv= rgb16to24; break;
		case 0x68: conv= rgb32to24; break;
		case 0x83: conv= rgb15to32; break;
		case 0x84: conv= rgb16to32; break;
		case 0x86: conv= rgb24to32; break;
		default: MSG_ERR("swScaler: internal error %s -> %s converter\n", 
				 sws_format_name(srcFormat), sws_format_name(dstFormat)); break;
		}
	}else if(   (isBGR(srcFormat) && isRGB(dstFormat))
		 || (isRGB(srcFormat) && isBGR(dstFormat))){
		switch(srcId | (dstId<<4)){
		case 0x33: conv= rgb15tobgr15; break;
		case 0x34: conv= rgb16tobgr15; break;
		case 0x36: conv= rgb24tobgr15; break;
		case 0x38: conv= rgb32tobgr15; break;
		case 0x43: conv= rgb15tobgr16; break;
		case 0x44: conv= rgb16tobgr16; break;
		case 0x46: conv= rgb24tobgr16; break;
		case 0x48: conv= rgb32tobgr16; break;
		case 0x63: conv= rgb15tobgr24; break;
		case 0x64: conv= rgb16tobgr24; break;
		case 0x66: conv= rgb24tobgr24; break;
		case 0x68: conv= rgb32tobgr24; break;
		case 0x83: conv= rgb15tobgr32; break;
		case 0x84: conv= rgb16tobgr32; break;
		case 0x86: conv= rgb24tobgr32; break;
		case 0x88: conv= rgb32tobgr32; break;
		default: MSG_ERR("swScaler: internal error %s -> %s converter\n", 
				 sws_format_name(srcFormat), sws_format_name(dstFormat)); break;
		}
	}else{
		MSG_ERR("swScaler: internal error %s -> %s converter\n", 
			 sws_format_name(srcFormat), sws_format_name(dstFormat));
	}

	if(dstStride[0]*srcBpp == srcStride[0]*dstBpp)
		conv(src[0], dst[0] + dstStride[0]*srcSliceY, srcSliceH*srcStride[0]);
	else
	{
		int i;
		uint8_t *srcPtr= src[0];
		uint8_t *dstPtr= dst[0] + dstStride[0]*srcSliceY;

		for(i=0; i<srcSliceH; i++)
		{
			conv(srcPtr, dstPtr, c->srcW*srcBpp);
			srcPtr+= srcStride[0];
			dstPtr+= dstStride[0];
		}
	}     
	return srcSliceH;
}

static int bgr24toyv12Wrapper(SwsContext *c, uint8_t* src[], int srcStride[], int srcSliceY,
             int srcSliceH, uint8_t* dst[], int dstStride[]){

	rgb24toyv12(
		src[0], 
		dst[0]+ srcSliceY    *dstStride[0], 
		dst[1]+(srcSliceY>>1)*dstStride[1], 
		dst[2]+(srcSliceY>>1)*dstStride[2],
		c->srcW, srcSliceH, 
		dstStride[0], dstStride[1], srcStride[0]);
	return srcSliceH;
}

static int yvu9toyv12Wrapper(SwsContext *c, uint8_t* src[], int srcStride[], int srcSliceY,
             int srcSliceH, uint8_t* dst[], int dstStride[]){
	int i;

	/* copy Y */
	if(srcStride[0]==dstStride[0] && srcStride[0] > 0) 
		memcpy(dst[0]+ srcSliceY*dstStride[0], src[0], srcStride[0]*srcSliceH);
	else{
		uint8_t *srcPtr= src[0];
		uint8_t *dstPtr= dst[0] + dstStride[0]*srcSliceY;

		for(i=0; i<srcSliceH; i++)
		{
			memcpy(dstPtr, srcPtr, c->srcW);
			srcPtr+= srcStride[0];
			dstPtr+= dstStride[0];
		}
	}

	if(c->dstFormat==IMGFMT_YV12){
		planar2x(src[1], dst[1], c->chrSrcW, c->chrSrcH, srcStride[1], dstStride[1]);
		planar2x(src[2], dst[2], c->chrSrcW, c->chrSrcH, srcStride[2], dstStride[2]);
	}else{
		planar2x(src[1], dst[2], c->chrSrcW, c->chrSrcH, srcStride[1], dstStride[2]);
		planar2x(src[2], dst[1], c->chrSrcW, c->chrSrcH, srcStride[2], dstStride[1]);
	}
	return srcSliceH;
}

/**
 * bring pointers in YUV order instead of YVU
 */
static inline void sws_orderYUV(int format, uint8_t * sortedP[], int sortedStride[], uint8_t * p[], int stride[]){
	if(format == IMGFMT_YV12 || format == IMGFMT_YVU9
           || format == IMGFMT_444P || format == IMGFMT_422P || format == IMGFMT_411P){
		sortedP[0]= p[0];
		sortedP[1]= p[2];
		sortedP[2]= p[1];
		sortedStride[0]= stride[0];
		sortedStride[1]= stride[2];
		sortedStride[2]= stride[1];
	}
	else if(isPacked(format) || isGray(format) || format == IMGFMT_Y8)
	{
		sortedP[0]= p[0];
		sortedP[1]= 
		sortedP[2]= NULL;
		sortedStride[0]= stride[0];
		sortedStride[1]= 
		sortedStride[2]= 0;
	}
	else if(format == IMGFMT_I420 || format == IMGFMT_IYUV)
	{
		sortedP[0]= p[0];
		sortedP[1]= p[1];
		sortedP[2]= p[2];
		sortedStride[0]= stride[0];
		sortedStride[1]= stride[1];
		sortedStride[2]= stride[2];
	}
	else if(format == IMGFMT_NV12 || format == IMGFMT_NV21)
	{
		sortedP[0]= p[0];
		sortedP[1]= p[1];
		sortedP[2]= NULL;
		sortedStride[0]= stride[0];
		sortedStride[1]= stride[1];
		sortedStride[2]= 0;
	}else{
		MSG_ERR("internal error in orderYUV\n");
	}
}

/* unscaled copy like stuff (assumes nearly identical formats) */
static int simpleCopy(SwsContext *c, uint8_t* src[], int srcStride[], int srcSliceY,
             int srcSliceH, uint8_t* dst[], int dstStride[]){

	if(isPacked(c->srcFormat))
	{
		if(dstStride[0]==srcStride[0] && srcStride[0] > 0)
			memcpy(dst[0] + dstStride[0]*srcSliceY, src[0], srcSliceH*dstStride[0]);
		else
		{
			int i;
			uint8_t *srcPtr= src[0];
			uint8_t *dstPtr= dst[0] + dstStride[0]*srcSliceY;
			int length=0;

			/* universal length finder */
			while(length+c->srcW <= ABS(dstStride[0]) 
			   && length+c->srcW <= ABS(srcStride[0])) length+= c->srcW;
			ASSERT(length!=0);

			for(i=0; i<srcSliceH; i++)
			{
				memcpy(dstPtr, srcPtr, length);
				srcPtr+= srcStride[0];
				dstPtr+= dstStride[0];
			}
		}
	}
	else 
	{ /* Planar YUV or gray */
		int plane;
		for(plane=0; plane<3; plane++)
		{
			int length= plane==0 ? c->srcW  : -((-c->srcW  )>>c->chrDstHSubSample);
			int y=      plane==0 ? srcSliceY: -((-srcSliceY)>>c->chrDstVSubSample);
			int height= plane==0 ? srcSliceH: -((-srcSliceH)>>c->chrDstVSubSample);

			if((isGray(c->srcFormat) || isGray(c->dstFormat)) && plane>0)
			{
				if(!isGray(c->dstFormat))
					memset(dst[plane], 128, dstStride[plane]*height);
			}
			else
			{
				if(dstStride[plane]==srcStride[plane] && srcStride[plane] > 0)
					memcpy(dst[plane] + dstStride[plane]*y, src[plane], height*dstStride[plane]);
				else
				{
					int i;
					uint8_t *srcPtr= src[plane];
					uint8_t *dstPtr= dst[plane] + dstStride[plane]*y;
					for(i=0; i<height; i++)
					{
						memcpy(dstPtr, srcPtr, length);
						srcPtr+= srcStride[plane];
						dstPtr+= dstStride[plane];
					}
				}
			}
		}
	}
	return srcSliceH;
}

static int remove_dup_fourcc(int fourcc)
{
	switch(fourcc)
	{
	    case IMGFMT_I420:
	    case IMGFMT_IYUV: return IMGFMT_YV12;
	    case IMGFMT_Y8  : return IMGFMT_Y800;
	    case IMGFMT_IF09: return IMGFMT_YVU9;
	    default: return fourcc;
	}
}

static void getSubSampleFactors(int *h, int *v, int format){
	switch(format){
	case IMGFMT_UYVY:
	case IMGFMT_YUY2:
		*h=1;
		*v=0;
		break;
	case IMGFMT_YV12:
	case IMGFMT_Y800: //FIXME remove after different subsamplings are fully implemented
	case IMGFMT_NV12:
	case IMGFMT_NV21:
		*h=1;
		*v=1;
		break;
	case IMGFMT_YVU9:
		*h=2;
		*v=2;
		break;
	case IMGFMT_444P:
		*h=0;
		*v=0;
		break;
	case IMGFMT_422P:
		*h=1;
		*v=0;
		break;
	case IMGFMT_411P:
		*h=2;
		*v=0;
		break;
	default:
		*h=0;
		*v=0;
		break;
	}
}

static uint16_t roundToInt16(int64_t f){
	int r= (f + (1<<15))>>16;
	     if(r<-0x7FFF) return 0x8000;
	else if(r> 0x7FFF) return 0x7FFF;
	else               return r;
}

/**
 * @param inv_table the yuv2rgb coeffs, normally Inverse_Table_6_9[x]
 * @param fullRange if 1 then the luma range is 0..255 if 0 its 16..235
 * @return -1 if not supported
 */
int sws_setColorspaceDetails(SwsContext *c, const int inv_table[4], int srcRange, const int table[4], int dstRange, int brightness, int contrast, int saturation){
	int64_t crv =  inv_table[0];
	int64_t cbu =  inv_table[1];
	int64_t cgu = -inv_table[2];
	int64_t cgv = -inv_table[3];
	int64_t cy  = 1<<16;
	int64_t oy  = 0;

	if(isYUV(c->dstFormat) || isGray(c->dstFormat)) return -1;
	memcpy(c->srcColorspaceTable, inv_table, sizeof(int)*4);
	memcpy(c->dstColorspaceTable,     table, sizeof(int)*4);

	c->brightness= brightness;
	c->contrast  = contrast;
	c->saturation= saturation;
	c->srcRange  = srcRange;
	c->dstRange  = dstRange;

	c->uOffset=   0x0400040004000400LL;
	c->vOffset=   0x0400040004000400LL;

	if(!srcRange){
		cy= (cy*255) / 219;
		oy= 16<<16;
	}

	cy = (cy *contrast             )>>16;
	crv= (crv*contrast * saturation)>>32;
	cbu= (cbu*contrast * saturation)>>32;
	cgu= (cgu*contrast * saturation)>>32;
	cgv= (cgv*contrast * saturation)>>32;

	oy -= 256*brightness;

	c->yCoeff=    roundToInt16(cy *8192) * 0x0001000100010001ULL;
	c->vrCoeff=   roundToInt16(crv*8192) * 0x0001000100010001ULL;
	c->ubCoeff=   roundToInt16(cbu*8192) * 0x0001000100010001ULL;
	c->vgCoeff=   roundToInt16(cgv*8192) * 0x0001000100010001ULL;
	c->ugCoeff=   roundToInt16(cgu*8192) * 0x0001000100010001ULL;
	c->yOffset=   roundToInt16(oy *   8) * 0x0001000100010001ULL;

	yuv2rgb_c_init_tables(c, inv_table, srcRange, brightness, contrast, saturation);
	//FIXME factorize

#ifdef COMPILE_ALTIVEC
	if (c->flags & SWS_CPU_CAPS_ALTIVEC)
	    yuv2rgb_altivec_init_tables (c, inv_table, brightness, contrast, saturation);
#endif	
	return 0;
}

/**
 * @return -1 if not supported
 */
int sws_getColorspaceDetails(SwsContext *c, int **inv_table, int *srcRange, int **table, int *dstRange, int *brightness, int *contrast, int *saturation){
	if(isYUV(c->dstFormat) || isGray(c->dstFormat)) return -1;

	*inv_table = c->srcColorspaceTable;
	*table     = c->dstColorspaceTable;
	*srcRange  = c->srcRange;
	*dstRange  = c->dstRange;
	*brightness= c->brightness;
	*contrast  = c->contrast;
	*saturation= c->saturation;
	
	return 0;	
}

SwsContext *sws_getContext(int srcW, int srcH, int origSrcFormat, int dstW, int dstH, int origDstFormat, int flags,
                         SwsFilter *srcFilter, SwsFilter *dstFilter, double *param){

	SwsContext *c;
	int i;
	int usesVFilter, usesHFilter;
	int unscaled, needsDither;
	int srcFormat, dstFormat;
	SwsFilter dummyFilter= {NULL, NULL, NULL, NULL};
#if defined(ARCH_X86) || defined(ARCH_X86_64)
	if(flags & SWS_CPU_CAPS_MMX)
		asm volatile("emms\n\t"::: "memory");
#endif

#ifndef RUNTIME_CPUDETECT //ensure that the flags match the compiled variant if cpudetect is off
	flags &= ~(SWS_CPU_CAPS_MMX|SWS_CPU_CAPS_MMX2|SWS_CPU_CAPS_3DNOW|SWS_CPU_CAPS_ALTIVEC);
#ifdef HAVE_MMX2
	flags |= SWS_CPU_CAPS_MMX|SWS_CPU_CAPS_MMX2;
#elif defined (HAVE_3DNOW)
	flags |= SWS_CPU_CAPS_MMX|SWS_CPU_CAPS_3DNOW;
#elif defined (HAVE_MMX)
	flags |= SWS_CPU_CAPS_MMX;
#elif defined (HAVE_ALTIVEC)
	flags |= SWS_CPU_CAPS_ALTIVEC;
#endif
#endif
	if(clip_table[512] != 255) globalInit();
	if(rgb15to16 == NULL) sws_rgb2rgb_init(flags);

	/* avoid duplicate Formats, so we don't need to check to much */
	srcFormat = remove_dup_fourcc(origSrcFormat);
	dstFormat = remove_dup_fourcc(origDstFormat);

	unscaled = (srcW == dstW && srcH == dstH);
	needsDither= (isBGR(dstFormat) || isRGB(dstFormat)) 
		     && (dstFormat&0xFF)<24
		     && ((dstFormat&0xFF)<(srcFormat&0xFF) || (!(isRGB(srcFormat) || isBGR(srcFormat))));

	if(!isSupportedIn(srcFormat)) 
	{
		MSG_ERR("swScaler: %s is not supported as input format\n", sws_format_name(srcFormat));
		return NULL;
	}
	if(!isSupportedOut(dstFormat))
	{
		MSG_ERR("swScaler: %s is not supported as output format\n", sws_format_name(dstFormat));
		return NULL;
	}

	/* sanity check */
	if(srcW<4 || srcH<1 || dstW<8 || dstH<1) //FIXME check if these are enough and try to lowwer them after fixing the relevant parts of the code
	{
		 MSG_ERR("swScaler: %dx%d -> %dx%d is invalid scaling dimension\n", 
			srcW, srcH, dstW, dstH);
		return NULL;
	}

	if(!dstFilter) dstFilter= &dummyFilter;
	if(!srcFilter) srcFilter= &dummyFilter;

	c= av_malloc(sizeof(SwsContext));
	memset(c, 0, sizeof(SwsContext));

	c->srcW= srcW;
	c->srcH= srcH;
	c->dstW= dstW;
	c->dstH= dstH;
	c->lumXInc= ((srcW<<16) + (dstW>>1))/dstW;
	c->lumYInc= ((srcH<<16) + (dstH>>1))/dstH;
	c->flags= flags;
	c->dstFormat= dstFormat;
	c->srcFormat= srcFormat;
	c->origDstFormat= origDstFormat;
	c->origSrcFormat= origSrcFormat;
        c->vRounder= 4* 0x0001000100010001ULL;

	usesHFilter= usesVFilter= 0;
	if(dstFilter->lumV!=NULL && dstFilter->lumV->length>1) usesVFilter=1;
	if(dstFilter->lumH!=NULL && dstFilter->lumH->length>1) usesHFilter=1;
	if(dstFilter->chrV!=NULL && dstFilter->chrV->length>1) usesVFilter=1;
	if(dstFilter->chrH!=NULL && dstFilter->chrH->length>1) usesHFilter=1;
	if(srcFilter->lumV!=NULL && srcFilter->lumV->length>1) usesVFilter=1;
	if(srcFilter->lumH!=NULL && srcFilter->lumH->length>1) usesHFilter=1;
	if(srcFilter->chrV!=NULL && srcFilter->chrV->length>1) usesVFilter=1;
	if(srcFilter->chrH!=NULL && srcFilter->chrH->length>1) usesHFilter=1;

	getSubSampleFactors(&c->chrSrcHSubSample, &c->chrSrcVSubSample, srcFormat);
	getSubSampleFactors(&c->chrDstHSubSample, &c->chrDstVSubSample, dstFormat);

	// reuse chroma for 2 pixles rgb/bgr unless user wants full chroma interpolation
	if((isBGR(dstFormat) || isRGB(dstFormat)) && !(flags&SWS_FULL_CHR_H_INT)) c->chrDstHSubSample=1;

	// drop some chroma lines if the user wants it
	c->vChrDrop= (flags&SWS_SRC_V_CHR_DROP_MASK)>>SWS_SRC_V_CHR_DROP_SHIFT;
	c->chrSrcVSubSample+= c->vChrDrop;

	// drop every 2. pixel for chroma calculation unless user wants full chroma
	if((isBGR(srcFormat) || isRGB(srcFormat)) && !(flags&SWS_FULL_CHR_H_INP)) 
		c->chrSrcHSubSample=1;

	if(param){
		c->param[0] = param[0];
		c->param[1] = param[1];
	}else{
		c->param[0] =
		c->param[1] = SWS_PARAM_DEFAULT;
	}

	c->chrIntHSubSample= c->chrDstHSubSample;
	c->chrIntVSubSample= c->chrSrcVSubSample;

	// note the -((-x)>>y) is so that we allways round toward +inf
	c->chrSrcW= -((-srcW) >> c->chrSrcHSubSample);
	c->chrSrcH= -((-srcH) >> c->chrSrcVSubSample);
	c->chrDstW= -((-dstW) >> c->chrDstHSubSample);
	c->chrDstH= -((-dstH) >> c->chrDstVSubSample);

	sws_setColorspaceDetails(c, Inverse_Table_6_9[SWS_CS_DEFAULT], 0, Inverse_Table_6_9[SWS_CS_DEFAULT] /* FIXME*/, 0, 0, 1<<16, 1<<16); 

	/* unscaled special Cases */
	if(unscaled && !usesHFilter && !usesVFilter)
	{
		/* yv12_to_nv12 */
		if(srcFormat == IMGFMT_YV12 && (dstFormat == IMGFMT_NV12 || dstFormat == IMGFMT_NV21))
		{
			c->swScale= PlanarToNV12Wrapper;
		}
		/* yuv2bgr */
		if((srcFormat==IMGFMT_YV12 || srcFormat==IMGFMT_422P) && (isBGR(dstFormat) || isRGB(dstFormat)))
		{
			c->swScale= yuv2rgb_get_func_ptr(c);
		}
		
		if( srcFormat==IMGFMT_YVU9 && dstFormat==IMGFMT_YV12 )
		{
			c->swScale= yvu9toyv12Wrapper;
		}

		/* bgr24toYV12 */
		if(srcFormat==IMGFMT_BGR24 && dstFormat==IMGFMT_YV12)
			c->swScale= bgr24toyv12Wrapper;
		
		/* rgb/bgr -> rgb/bgr (no dither needed forms) */
		if(   (isBGR(srcFormat) || isRGB(srcFormat))
		   && (isBGR(dstFormat) || isRGB(dstFormat)) 
		   && !needsDither)
			c->swScale= rgb2rgbWrapper;

		/* LQ converters if -sws 0 or -sws 4*/
		if(c->flags&(SWS_FAST_BILINEAR|SWS_POINT)){
			/* rgb/bgr -> rgb/bgr (dither needed forms) */
			if(  (isBGR(srcFormat) || isRGB(srcFormat))
			  && (isBGR(dstFormat) || isRGB(dstFormat)) 
			  && needsDither)
				c->swScale= rgb2rgbWrapper;

			/* yv12_to_yuy2 */
			if(srcFormat == IMGFMT_YV12 && 
			    (dstFormat == IMGFMT_YUY2 || dstFormat == IMGFMT_UYVY))
			{
				if (dstFormat == IMGFMT_YUY2)
				    c->swScale= PlanarToYuy2Wrapper;
				else
				    c->swScale= PlanarToUyvyWrapper;
			}
		}

#ifdef COMPILE_ALTIVEC
		if ((c->flags & SWS_CPU_CAPS_ALTIVEC) &&
		    ((srcFormat == IMGFMT_YV12 && 
		      (dstFormat == IMGFMT_YUY2 || dstFormat == IMGFMT_UYVY)))) {
		  // unscaled YV12 -> packed YUV, we want speed
		  if (dstFormat == IMGFMT_YUY2)
		    c->swScale= yv12toyuy2_unscaled_altivec;
		  else
		    c->swScale= yv12touyvy_unscaled_altivec;
		}
#endif

		/* simple copy */
		if(   srcFormat == dstFormat
		   || (isPlanarYUV(srcFormat) && isGray(dstFormat))
		   || (isPlanarYUV(dstFormat) && isGray(srcFormat))
		  )
		{
			c->swScale= simpleCopy;
		}

		if(c->swScale){
			if(flags&SWS_PRINT_INFO)
				MSG_INFO("SwScaler: using unscaled %s -> %s special converter\n", 
					sws_format_name(srcFormat), sws_format_name(dstFormat));
			return c;
		}
	}

	if(flags & SWS_CPU_CAPS_MMX2)
	{
		c->canMMX2BeUsed= (dstW >=srcW && (dstW&31)==0 && (srcW&15)==0) ? 1 : 0;
		if(!c->canMMX2BeUsed && dstW >=srcW && (srcW&15)==0 && (flags&SWS_FAST_BILINEAR))
		{
			if(flags&SWS_PRINT_INFO)
				MSG_INFO("SwScaler: output Width is not a multiple of 32 -> no MMX2 scaler\n");
		}
		if(usesHFilter) c->canMMX2BeUsed=0;
	}
	else
		c->canMMX2BeUsed=0;

	c->chrXInc= ((c->chrSrcW<<16) + (c->chrDstW>>1))/c->chrDstW;
	c->chrYInc= ((c->chrSrcH<<16) + (c->chrDstH>>1))/c->chrDstH;

	// match pixel 0 of the src to pixel 0 of dst and match pixel n-2 of src to pixel n-2 of dst
	// but only for the FAST_BILINEAR mode otherwise do correct scaling
	// n-2 is the last chrominance sample available
	// this is not perfect, but noone shuld notice the difference, the more correct variant
	// would be like the vertical one, but that would require some special code for the
	// first and last pixel
	if(flags&SWS_FAST_BILINEAR)
	{
		if(c->canMMX2BeUsed)
		{
			c->lumXInc+= 20;
			c->chrXInc+= 20;
		}
		//we don't use the x86asm scaler if mmx is available
		else if(flags & SWS_CPU_CAPS_MMX)
		{
			c->lumXInc = ((srcW-2)<<16)/(dstW-2) - 20;
			c->chrXInc = ((c->chrSrcW-2)<<16)/(c->chrDstW-2) - 20;
		}
	}

	/* precalculate horizontal scaler filter coefficients */
	{
		const int filterAlign=
		  (flags & SWS_CPU_CAPS_MMX) ? 4 :
		  (flags & SWS_CPU_CAPS_ALTIVEC) ? 8 :
		  1;

		initFilter(&c->hLumFilter, &c->hLumFilterPos, &c->hLumFilterSize, c->lumXInc,
				 srcW      ,       dstW, filterAlign, 1<<14,
				 (flags&SWS_BICUBLIN) ? (flags|SWS_BICUBIC)  : flags,
				 srcFilter->lumH, dstFilter->lumH, c->param);
		initFilter(&c->hChrFilter, &c->hChrFilterPos, &c->hChrFilterSize, c->chrXInc,
				 c->chrSrcW, c->chrDstW, filterAlign, 1<<14,
				 (flags&SWS_BICUBLIN) ? (flags|SWS_BILINEAR) : flags,
				 srcFilter->chrH, dstFilter->chrH, c->param);

#if defined(ARCH_X86) || defined(ARCH_X86_64)
// can't downscale !!!
		if(c->canMMX2BeUsed && (flags & SWS_FAST_BILINEAR))
		{
#define MAX_FUNNY_CODE_SIZE 10000
#ifdef MAP_ANONYMOUS
			c->funnyYCode = (uint8_t*)mmap(NULL, MAX_FUNNY_CODE_SIZE, PROT_EXEC | PROT_READ | PROT_WRITE, MAP_PRIVATE | MAP_ANONYMOUS, 0, 0);
			c->funnyUVCode = (uint8_t*)mmap(NULL, MAX_FUNNY_CODE_SIZE, PROT_EXEC | PROT_READ | PROT_WRITE, MAP_PRIVATE | MAP_ANONYMOUS, 0, 0);
#else
			c->funnyYCode = av_malloc(MAX_FUNNY_CODE_SIZE);
			c->funnyUVCode = av_malloc(MAX_FUNNY_CODE_SIZE);
#endif

			c->lumMmx2Filter   = av_malloc((dstW        /8+8)*sizeof(int16_t));
			c->chrMmx2Filter   = av_malloc((c->chrDstW  /4+8)*sizeof(int16_t));
			c->lumMmx2FilterPos= av_malloc((dstW      /2/8+8)*sizeof(int32_t));
			c->chrMmx2FilterPos= av_malloc((c->chrDstW/2/4+8)*sizeof(int32_t));

			initMMX2HScaler(      dstW, c->lumXInc, c->funnyYCode , c->lumMmx2Filter, c->lumMmx2FilterPos, 8);
			initMMX2HScaler(c->chrDstW, c->chrXInc, c->funnyUVCode, c->chrMmx2Filter, c->chrMmx2FilterPos, 4);
		}
#endif
	} // Init Horizontal stuff



	/* precalculate vertical scaler filter coefficients */
	{
		const int filterAlign=
		  (flags & SWS_CPU_CAPS_ALTIVEC) ? 8 :
		  1;

		initFilter(&c->vLumFilter, &c->vLumFilterPos, &c->vLumFilterSize, c->lumYInc,
				srcH      ,        dstH, filterAlign, (1<<12)-4,
				(flags&SWS_BICUBLIN) ? (flags|SWS_BICUBIC)  : flags,
				srcFilter->lumV, dstFilter->lumV, c->param);
		initFilter(&c->vChrFilter, &c->vChrFilterPos, &c->vChrFilterSize, c->chrYInc,
				c->chrSrcH, c->chrDstH, filterAlign, (1<<12)-4,
				(flags&SWS_BICUBLIN) ? (flags|SWS_BILINEAR) : flags,
				srcFilter->chrV, dstFilter->chrV, c->param);

#ifdef HAVE_ALTIVEC
		c->vYCoeffsBank = av_malloc(sizeof (vector signed short)*c->vLumFilterSize*c->dstH);
		c->vCCoeffsBank = av_malloc(sizeof (vector signed short)*c->vChrFilterSize*c->chrDstH);

		for (i=0;i<c->vLumFilterSize*c->dstH;i++) {
                  int j;
		  short *p = (short *)&c->vYCoeffsBank[i];
		  for (j=0;j<8;j++)
		    p[j] = c->vLumFilter[i];
		}

		for (i=0;i<c->vChrFilterSize*c->chrDstH;i++) {
                  int j;
		  short *p = (short *)&c->vCCoeffsBank[i];
		  for (j=0;j<8;j++)
		    p[j] = c->vChrFilter[i];
		}
#endif
	}

	// Calculate Buffer Sizes so that they won't run out while handling these damn slices
	c->vLumBufSize= c->vLumFilterSize;
	c->vChrBufSize= c->vChrFilterSize;
	for(i=0; i<dstH; i++)
	{
		int chrI= i*c->chrDstH / dstH;
		int nextSlice= MAX(c->vLumFilterPos[i   ] + c->vLumFilterSize - 1,
				 ((c->vChrFilterPos[chrI] + c->vChrFilterSize - 1)<<c->chrSrcVSubSample));

		nextSlice>>= c->chrSrcVSubSample;
		nextSlice<<= c->chrSrcVSubSample;
		if(c->vLumFilterPos[i   ] + c->vLumBufSize < nextSlice)
			c->vLumBufSize= nextSlice - c->vLumFilterPos[i   ];
		if(c->vChrFilterPos[chrI] + c->vChrBufSize < (nextSlice>>c->chrSrcVSubSample))
			c->vChrBufSize= (nextSlice>>c->chrSrcVSubSample) - c->vChrFilterPos[chrI];
	}

	// allocate pixbufs (we use dynamic allocation because otherwise we would need to
	c->lumPixBuf= av_malloc(c->vLumBufSize*2*sizeof(int16_t*));
	c->chrPixBuf= av_malloc(c->vChrBufSize*2*sizeof(int16_t*));
	//Note we need at least one pixel more at the end because of the mmx code (just in case someone wanna replace the 4000/8000)
	/* align at 16 bytes for AltiVec */
	for(i=0; i<c->vLumBufSize; i++)
		c->lumPixBuf[i]= c->lumPixBuf[i+c->vLumBufSize]= av_malloc(4000);
	for(i=0; i<c->vChrBufSize; i++)
		c->chrPixBuf[i]= c->chrPixBuf[i+c->vChrBufSize]= av_malloc(8000);

	//try to avoid drawing green stuff between the right end and the stride end
	for(i=0; i<c->vLumBufSize; i++) memset(c->lumPixBuf[i], 0, 4000);
	for(i=0; i<c->vChrBufSize; i++) memset(c->chrPixBuf[i], 64, 8000);

	ASSERT(c->chrDstH <= dstH)

	if(flags&SWS_PRINT_INFO)
	{
#ifdef DITHER1XBPP
		char *dither= " dithered";
#else
		char *dither= "";
#endif
		if(flags&SWS_FAST_BILINEAR)
			MSG_INFO("\nSwScaler: FAST_BILINEAR scaler, ");
		else if(flags&SWS_BILINEAR)
			MSG_INFO("\nSwScaler: BILINEAR scaler, ");
		else if(flags&SWS_BICUBIC)
			MSG_INFO("\nSwScaler: BICUBIC scaler, ");
		else if(flags&SWS_X)
			MSG_INFO("\nSwScaler: Experimental scaler, ");
		else if(flags&SWS_POINT)
			MSG_INFO("\nSwScaler: Nearest Neighbor / POINT scaler, ");
		else if(flags&SWS_AREA)
			MSG_INFO("\nSwScaler: Area Averageing scaler, ");
		else if(flags&SWS_BICUBLIN)
			MSG_INFO("\nSwScaler: luma BICUBIC / chroma BILINEAR scaler, ");
		else if(flags&SWS_GAUSS)
			MSG_INFO("\nSwScaler: Gaussian scaler, ");
		else if(flags&SWS_SINC)
			MSG_INFO("\nSwScaler: Sinc scaler, ");
		else if(flags&SWS_LANCZOS)
			MSG_INFO("\nSwScaler: Lanczos scaler, ");
		else if(flags&SWS_SPLINE)
			MSG_INFO("\nSwScaler: Bicubic spline scaler, ");
		else
			MSG_INFO("\nSwScaler: ehh flags invalid?! ");

		if(dstFormat==IMGFMT_BGR15 || dstFormat==IMGFMT_BGR16)
			MSG_INFO("from %s to%s %s ", 
				sws_format_name(srcFormat), dither, sws_format_name(dstFormat));
		else
			MSG_INFO("from %s to %s ", 
				sws_format_name(srcFormat), sws_format_name(dstFormat));

		if(flags & SWS_CPU_CAPS_MMX2)
			MSG_INFO("using MMX2\n");
		else if(flags & SWS_CPU_CAPS_3DNOW)
			MSG_INFO("using 3DNOW\n");
		else if(flags & SWS_CPU_CAPS_MMX)
			MSG_INFO("using MMX\n");
		else if(flags & SWS_CPU_CAPS_ALTIVEC)
			MSG_INFO("using AltiVec\n");
		else 
			MSG_INFO("using C\n");
	}

	if(flags & SWS_PRINT_INFO)
	{
		if(flags & SWS_CPU_CAPS_MMX)
		{
			if(c->canMMX2BeUsed && (flags&SWS_FAST_BILINEAR))
				MSG_V("SwScaler: using FAST_BILINEAR MMX2 scaler for horizontal scaling\n");
			else
			{
				if(c->hLumFilterSize==4)
					MSG_V("SwScaler: using 4-tap MMX scaler for horizontal luminance scaling\n");
				else if(c->hLumFilterSize==8)
					MSG_V("SwScaler: using 8-tap MMX scaler for horizontal luminance scaling\n");
				else
					MSG_V("SwScaler: using n-tap MMX scaler for horizontal luminance scaling\n");

				if(c->hChrFilterSize==4)
					MSG_V("SwScaler: using 4-tap MMX scaler for horizontal chrominance scaling\n");
				else if(c->hChrFilterSize==8)
					MSG_V("SwScaler: using 8-tap MMX scaler for horizontal chrominance scaling\n");
				else
					MSG_V("SwScaler: using n-tap MMX scaler for horizontal chrominance scaling\n");
			}
		}
		else
		{
#if defined(ARCH_X86) || defined(ARCH_X86_64)
			MSG_V("SwScaler: using X86-Asm scaler for horizontal scaling\n");
#else
			if(flags & SWS_FAST_BILINEAR)
				MSG_V("SwScaler: using FAST_BILINEAR C scaler for horizontal scaling\n");
			else
				MSG_V("SwScaler: using C scaler for horizontal scaling\n");
#endif
		}
		if(isPlanarYUV(dstFormat))
		{
			if(c->vLumFilterSize==1)
				MSG_V("SwScaler: using 1-tap %s \"scaler\" for vertical scaling (YV12 like)\n", (flags & SWS_CPU_CAPS_MMX) ? "MMX" : "C");
			else
				MSG_V("SwScaler: using n-tap %s scaler for vertical scaling (YV12 like)\n", (flags & SWS_CPU_CAPS_MMX) ? "MMX" : "C");
		}
		else
		{
			if(c->vLumFilterSize==1 && c->vChrFilterSize==2)
				MSG_V("SwScaler: using 1-tap %s \"scaler\" for vertical luminance scaling (BGR)\n"
				       "SwScaler:       2-tap scaler for vertical chrominance scaling (BGR)\n",(flags & SWS_CPU_CAPS_MMX) ? "MMX" : "C");
			else if(c->vLumFilterSize==2 && c->vChrFilterSize==2)
				MSG_V("SwScaler: using 2-tap linear %s scaler for vertical scaling (BGR)\n", (flags & SWS_CPU_CAPS_MMX) ? "MMX" : "C");
			else
				MSG_V("SwScaler: using n-tap %s scaler for vertical scaling (BGR)\n", (flags & SWS_CPU_CAPS_MMX) ? "MMX" : "C");
		}

		if(dstFormat==IMGFMT_BGR24)
			MSG_V("SwScaler: using %s YV12->BGR24 Converter\n",
				(flags & SWS_CPU_CAPS_MMX2) ? "MMX2" : ((flags & SWS_CPU_CAPS_MMX) ? "MMX" : "C"));
		else if(dstFormat==IMGFMT_BGR32)
			MSG_V("SwScaler: using %s YV12->BGR32 Converter\n", (flags & SWS_CPU_CAPS_MMX) ? "MMX" : "C");
		else if(dstFormat==IMGFMT_BGR16)
			MSG_V("SwScaler: using %s YV12->BGR16 Converter\n", (flags & SWS_CPU_CAPS_MMX) ? "MMX" : "C");
		else if(dstFormat==IMGFMT_BGR15)
			MSG_V("SwScaler: using %s YV12->BGR15 Converter\n", (flags & SWS_CPU_CAPS_MMX) ? "MMX" : "C");

		MSG_V("SwScaler: %dx%d -> %dx%d\n", srcW, srcH, dstW, dstH);
	}
	if(flags & SWS_PRINT_INFO)
	{
		MSG_DBG2("SwScaler:Lum srcW=%d srcH=%d dstW=%d dstH=%d xInc=%d yInc=%d\n",
			c->srcW, c->srcH, c->dstW, c->dstH, c->lumXInc, c->lumYInc);
		MSG_DBG2("SwScaler:Chr srcW=%d srcH=%d dstW=%d dstH=%d xInc=%d yInc=%d\n",
			c->chrSrcW, c->chrSrcH, c->chrDstW, c->chrDstH, c->chrXInc, c->chrYInc);
	}

	c->swScale= getSwsFunc(flags);
	return c;
}

/**
 * swscale warper, so we don't need to export the SwsContext.
 * assumes planar YUV to be in YUV order instead of YVU
 */
int sws_scale_ordered(SwsContext *c, uint8_t* src[], int srcStride[], int srcSliceY,
                           int srcSliceH, uint8_t* dst[], int dstStride[]){
	if (c->sliceDir == 0 && srcSliceY != 0 && srcSliceY + srcSliceH != c->srcH) {
	    MSG_ERR("swScaler: slices start in the middle!\n");
	    return 0;
	}
	if (c->sliceDir == 0) {
	    if (srcSliceY == 0) c->sliceDir = 1; else c->sliceDir = -1;
	}

	// copy strides, so they can safely be modified
	if (c->sliceDir == 1) {
	    // slices go from top to bottom
	    int srcStride2[3]= {srcStride[0], srcStride[1], srcStride[2]};
	    int dstStride2[3]= {dstStride[0], dstStride[1], dstStride[2]};
	    return c->swScale(c, src, srcStride2, srcSliceY, srcSliceH, dst, dstStride2);
	} else {
	    // slices go from bottom to top => we flip the image internally
	    uint8_t* src2[3]= {src[0] + (srcSliceH-1)*srcStride[0],
			       src[1] + ((srcSliceH>>c->chrSrcVSubSample)-1)*srcStride[1],
			       src[2] + ((srcSliceH>>c->chrSrcVSubSample)-1)*srcStride[2]
	    };
	    uint8_t* dst2[3]= {dst[0] + (c->dstH-1)*dstStride[0],
			       dst[1] + ((c->dstH>>c->chrDstVSubSample)-1)*dstStride[1],
			       dst[2] + ((c->dstH>>c->chrDstVSubSample)-1)*dstStride[2]};
	    int srcStride2[3]= {-srcStride[0], -srcStride[1], -srcStride[2]};
	    int dstStride2[3]= {-dstStride[0], -dstStride[1], -dstStride[2]};
	    
	    return c->swScale(c, src2, srcStride2, c->srcH-srcSliceY-srcSliceH, srcSliceH, dst2, dstStride2);
	}
}

/**
 * swscale warper, so we don't need to export the SwsContext
 */
int sws_scale(SwsContext *c, uint8_t* srcParam[], int srcStrideParam[], int srcSliceY,
                           int srcSliceH, uint8_t* dstParam[], int dstStrideParam[]){
	int srcStride[3];
	int dstStride[3];
	uint8_t *src[3];
	uint8_t *dst[3];
	sws_orderYUV(c->origSrcFormat, src, srcStride, srcParam, srcStrideParam);
	sws_orderYUV(c->origDstFormat, dst, dstStride, dstParam, dstStrideParam);
//printf("sws: slice %d %d\n", srcSliceY, srcSliceH);

	return c->swScale(c, src, srcStride, srcSliceY, srcSliceH, dst, dstStride);
}

SwsFilter *sws_getDefaultFilter(float lumaGBlur, float chromaGBlur, 
				float lumaSharpen, float chromaSharpen,
				float chromaHShift, float chromaVShift,
				int verbose)
{
	SwsFilter *filter= av_malloc(sizeof(SwsFilter));

	if(lumaGBlur!=0.0){
		filter->lumH= sws_getGaussianVec(lumaGBlur, 3.0);
		filter->lumV= sws_getGaussianVec(lumaGBlur, 3.0);
	}else{
		filter->lumH= sws_getIdentityVec();
		filter->lumV= sws_getIdentityVec();
	}

	if(chromaGBlur!=0.0){
		filter->chrH= sws_getGaussianVec(chromaGBlur, 3.0);
		filter->chrV= sws_getGaussianVec(chromaGBlur, 3.0);
	}else{
		filter->chrH= sws_getIdentityVec();
		filter->chrV= sws_getIdentityVec();
	}

	if(chromaSharpen!=0.0){
		SwsVector *id= sws_getIdentityVec();
                sws_scaleVec(filter->chrH, -chromaSharpen);
                sws_scaleVec(filter->chrV, -chromaSharpen);
		sws_addVec(filter->chrH, id);
		sws_addVec(filter->chrV, id);
		sws_freeVec(id);
	}

	if(lumaSharpen!=0.0){
		SwsVector *id= sws_getIdentityVec();
                sws_scaleVec(filter->lumH, -lumaSharpen);
                sws_scaleVec(filter->lumV, -lumaSharpen);
		sws_addVec(filter->lumH, id);
		sws_addVec(filter->lumV, id);
		sws_freeVec(id);
	}

	if(chromaHShift != 0.0)
		sws_shiftVec(filter->chrH, (int)(chromaHShift+0.5));

	if(chromaVShift != 0.0)
		sws_shiftVec(filter->chrV, (int)(chromaVShift+0.5));

	sws_normalizeVec(filter->chrH, 1.0);
	sws_normalizeVec(filter->chrV, 1.0);
	sws_normalizeVec(filter->lumH, 1.0);
	sws_normalizeVec(filter->lumV, 1.0);

	if(verbose) sws_printVec(filter->chrH);
	if(verbose) sws_printVec(filter->lumH);

        return filter;
}

/**
 * returns a normalized gaussian curve used to filter stuff
 * quality=3 is high quality, lowwer is lowwer quality
 */
SwsVector *sws_getGaussianVec(double variance, double quality){
	const int length= (int)(variance*quality + 0.5) | 1;
	int i;
	double *coeff= av_malloc(length*sizeof(double));
	double middle= (length-1)*0.5;
	SwsVector *vec= av_malloc(sizeof(SwsVector));

	vec->coeff= coeff;
	vec->length= length;

	for(i=0; i<length; i++)
	{
		double dist= i-middle;
		coeff[i]= exp( -dist*dist/(2*variance*variance) ) / sqrt(2*variance*PI);
	}

	sws_normalizeVec(vec, 1.0);

	return vec;
}

SwsVector *sws_getConstVec(double c, int length){
	int i;
	double *coeff= av_malloc(length*sizeof(double));
	SwsVector *vec= av_malloc(sizeof(SwsVector));

	vec->coeff= coeff;
	vec->length= length;

	for(i=0; i<length; i++)
		coeff[i]= c;

	return vec;
}


SwsVector *sws_getIdentityVec(void){
        return sws_getConstVec(1.0, 1);
}

double sws_dcVec(SwsVector *a){
	int i;
        double sum=0;

	for(i=0; i<a->length; i++)
		sum+= a->coeff[i];

        return sum;
}

void sws_scaleVec(SwsVector *a, double scalar){
	int i;

	for(i=0; i<a->length; i++)
		a->coeff[i]*= scalar;
}

void sws_normalizeVec(SwsVector *a, double height){
        sws_scaleVec(a, height/sws_dcVec(a));
}

static SwsVector *sws_getConvVec(SwsVector *a, SwsVector *b){
	int length= a->length + b->length - 1;
	double *coeff= av_malloc(length*sizeof(double));
	int i, j;
	SwsVector *vec= av_malloc(sizeof(SwsVector));

	vec->coeff= coeff;
	vec->length= length;

	for(i=0; i<length; i++) coeff[i]= 0.0;

	for(i=0; i<a->length; i++)
	{
		for(j=0; j<b->length; j++)
		{
			coeff[i+j]+= a->coeff[i]*b->coeff[j];
		}
	}

	return vec;
}

static SwsVector *sws_sumVec(SwsVector *a, SwsVector *b){
	int length= MAX(a->length, b->length);
	double *coeff= av_malloc(length*sizeof(double));
	int i;
	SwsVector *vec= av_malloc(sizeof(SwsVector));

	vec->coeff= coeff;
	vec->length= length;

	for(i=0; i<length; i++) coeff[i]= 0.0;

	for(i=0; i<a->length; i++) coeff[i + (length-1)/2 - (a->length-1)/2]+= a->coeff[i];
	for(i=0; i<b->length; i++) coeff[i + (length-1)/2 - (b->length-1)/2]+= b->coeff[i];

	return vec;
}

static SwsVector *sws_diffVec(SwsVector *a, SwsVector *b){
	int length= MAX(a->length, b->length);
	double *coeff= av_malloc(length*sizeof(double));
	int i;
	SwsVector *vec= av_malloc(sizeof(SwsVector));

	vec->coeff= coeff;
	vec->length= length;

	for(i=0; i<length; i++) coeff[i]= 0.0;

	for(i=0; i<a->length; i++) coeff[i + (length-1)/2 - (a->length-1)/2]+= a->coeff[i];
	for(i=0; i<b->length; i++) coeff[i + (length-1)/2 - (b->length-1)/2]-= b->coeff[i];

	return vec;
}

/* shift left / or right if "shift" is negative */
static SwsVector *sws_getShiftedVec(SwsVector *a, int shift){
	int length= a->length + ABS(shift)*2;
	double *coeff= av_malloc(length*sizeof(double));
	int i;
	SwsVector *vec= av_malloc(sizeof(SwsVector));

	vec->coeff= coeff;
	vec->length= length;

	for(i=0; i<length; i++) coeff[i]= 0.0;

	for(i=0; i<a->length; i++)
	{
		coeff[i + (length-1)/2 - (a->length-1)/2 - shift]= a->coeff[i];
	}

	return vec;
}

void sws_shiftVec(SwsVector *a, int shift){
	SwsVector *shifted= sws_getShiftedVec(a, shift);
	av_free(a->coeff);
	a->coeff= shifted->coeff;
	a->length= shifted->length;
	av_free(shifted);
}

void sws_addVec(SwsVector *a, SwsVector *b){
	SwsVector *sum= sws_sumVec(a, b);
	av_free(a->coeff);
	a->coeff= sum->coeff;
	a->length= sum->length;
	av_free(sum);
}

void sws_subVec(SwsVector *a, SwsVector *b){
	SwsVector *diff= sws_diffVec(a, b);
	av_free(a->coeff);
	a->coeff= diff->coeff;
	a->length= diff->length;
	av_free(diff);
}

void sws_convVec(SwsVector *a, SwsVector *b){
	SwsVector *conv= sws_getConvVec(a, b);
	av_free(a->coeff);  
	a->coeff= conv->coeff;
	a->length= conv->length;
	av_free(conv);
}

SwsVector *sws_cloneVec(SwsVector *a){
	double *coeff= av_malloc(a->length*sizeof(double));
	int i;
	SwsVector *vec= av_malloc(sizeof(SwsVector));

	vec->coeff= coeff;
	vec->length= a->length;

	for(i=0; i<a->length; i++) coeff[i]= a->coeff[i];

	return vec;
}

void sws_printVec(SwsVector *a){
	int i;
	double max=0;
	double min=0;
	double range;

	for(i=0; i<a->length; i++)
		if(a->coeff[i]>max) max= a->coeff[i];

	for(i=0; i<a->length; i++)
		if(a->coeff[i]<min) min= a->coeff[i];

	range= max - min;

	for(i=0; i<a->length; i++)
	{
		int x= (int)((a->coeff[i]-min)*60.0/range +0.5);
		MSG_DBG2("%1.3f ", a->coeff[i]);
		for(;x>0; x--) MSG_DBG2(" ");
		MSG_DBG2("|\n");
	}
}

void sws_freeVec(SwsVector *a){
	if(!a) return;
	av_free(a->coeff);
	a->coeff=NULL;
	a->length=0;
	av_free(a);
}

void sws_freeFilter(SwsFilter *filter){
	if(!filter) return;

	if(filter->lumH) sws_freeVec(filter->lumH);
	if(filter->lumV) sws_freeVec(filter->lumV);
	if(filter->chrH) sws_freeVec(filter->chrH);
	if(filter->chrV) sws_freeVec(filter->chrV);
	av_free(filter);
}


void sws_freeContext(SwsContext *c){
	int i;
	if(!c) return;

	if(c->lumPixBuf)
	{
		for(i=0; i<c->vLumBufSize; i++)
		{
			av_free(c->lumPixBuf[i]);
			c->lumPixBuf[i]=NULL;
		}
		av_free(c->lumPixBuf);
		c->lumPixBuf=NULL;
	}

	if(c->chrPixBuf)
	{
		for(i=0; i<c->vChrBufSize; i++)
		{
			av_free(c->chrPixBuf[i]);
			c->chrPixBuf[i]=NULL;
		}
		av_free(c->chrPixBuf);
		c->chrPixBuf=NULL;
	}

	av_free(c->vLumFilter);
	c->vLumFilter = NULL;
	av_free(c->vChrFilter);
	c->vChrFilter = NULL;
	av_free(c->hLumFilter);
	c->hLumFilter = NULL;
	av_free(c->hChrFilter);
	c->hChrFilter = NULL;
#ifdef HAVE_ALTIVEC
	av_free(c->vYCoeffsBank);
	c->vYCoeffsBank = NULL;
	av_free(c->vCCoeffsBank);
	c->vCCoeffsBank = NULL;
#endif

	av_free(c->vLumFilterPos);
	c->vLumFilterPos = NULL;
	av_free(c->vChrFilterPos);
	c->vChrFilterPos = NULL;
	av_free(c->hLumFilterPos);
	c->hLumFilterPos = NULL;
	av_free(c->hChrFilterPos);
	c->hChrFilterPos = NULL;

#if defined(ARCH_X86) || defined(ARCH_X86_64)
#ifdef MAP_ANONYMOUS
	if(c->funnyYCode) munmap(c->funnyYCode, MAX_FUNNY_CODE_SIZE);
	if(c->funnyUVCode) munmap(c->funnyUVCode, MAX_FUNNY_CODE_SIZE);
#else
	av_free(c->funnyYCode);
	av_free(c->funnyUVCode);
#endif
	c->funnyYCode=NULL;
	c->funnyUVCode=NULL;
#endif

	av_free(c->lumMmx2Filter);
	c->lumMmx2Filter=NULL;
	av_free(c->chrMmx2Filter);
	c->chrMmx2Filter=NULL;
	av_free(c->lumMmx2FilterPos);
	c->lumMmx2FilterPos=NULL;
	av_free(c->chrMmx2FilterPos);
	c->chrMmx2FilterPos=NULL;
	av_free(c->yuvTable);
	c->yuvTable=NULL;

	av_free(c);
}