Training courses

Kernel and Embedded Linux

Bootlin training courses

Embedded Linux, kernel,
Yocto Project, Buildroot, real-time,
graphics, boot time, debugging...

Bootlin logo

Elixir Cross Referencer

   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
*	$NetBSD: res_func.sa,v 1.5 2001/09/16 16:34:32 wiz Exp $

*	MOTOROLA MICROPROCESSOR & MEMORY TECHNOLOGY GROUP
*	M68000 Hi-Performance Microprocessor Division
*	M68040 Software Package 
*
*	M68040 Software Package Copyright (c) 1993, 1994 Motorola Inc.
*	All rights reserved.
*
*	THE SOFTWARE is provided on an "AS IS" basis and without warranty.
*	To the maximum extent permitted by applicable law,
*	MOTOROLA DISCLAIMS ALL WARRANTIES WHETHER EXPRESS OR IMPLIED,
*	INCLUDING IMPLIED WARRANTIES OF MERCHANTABILITY OR FITNESS FOR A
*	PARTICULAR PURPOSE and any warranty against infringement with
*	regard to the SOFTWARE (INCLUDING ANY MODIFIED VERSIONS THEREOF)
*	and any accompanying written materials. 
*
*	To the maximum extent permitted by applicable law,
*	IN NO EVENT SHALL MOTOROLA BE LIABLE FOR ANY DAMAGES WHATSOEVER
*	(INCLUDING WITHOUT LIMITATION, DAMAGES FOR LOSS OF BUSINESS
*	PROFITS, BUSINESS INTERRUPTION, LOSS OF BUSINESS INFORMATION, OR
*	OTHER PECUNIARY LOSS) ARISING OF THE USE OR INABILITY TO USE THE
*	SOFTWARE.  Motorola assumes no responsibility for the maintenance
*	and support of the SOFTWARE.  
*
*	You are hereby granted a copyright license to use, modify, and
*	distribute the SOFTWARE so long as this entire notice is retained
*	without alteration in any modified and/or redistributed versions,
*	and that such modified versions are clearly identified as such.
*	No licenses are granted by implication, estoppel or otherwise
*	under any patents or trademarks of Motorola, Inc.

*
*	res_func.sa 3.9 7/29/91
*
* Normalizes denormalized numbers if necessary and updates the
* stack frame.  The function is then restored back into the
* machine and the 040 completes the operation.  This routine
* is only used by the unsupported data type/format handler.
* (Exception vector 55).
*
* For packed move out (fmove.p fpm,<ea>) the operation is
* completed here; data is packed and moved to user memory. 
* The stack is restored to the 040 only in the case of a
* reportable exception in the conversion.
*

RES_FUNC    IDNT    2,1 Motorola 040 Floating Point Software Package

	section	8

	include	fpsp.h

sp_bnds:	dc.w	$3f81,$407e
		dc.w	$3f6a,$0000
dp_bnds:	dc.w	$3c01,$43fe
		dc.w	$3bcd,$0000

	xref	mem_write
	xref	bindec
	xref	get_fline
	xref	round
	xref	denorm
	xref	dest_ext
	xref	dest_dbl
	xref	dest_sgl
	xref	unf_sub
	xref	nrm_set
	xref	dnrm_lp
	xref	ovf_res
	xref	reg_dest
	xref	t_ovfl
	xref	t_unfl

	xdef	res_func
	xdef 	p_move

res_func:
	clr.b	DNRM_FLG(a6)
	clr.b	RES_FLG(a6)
	clr.b	CU_ONLY(a6)
	tst.b	DY_MO_FLG(a6)
	beq.b	monadic
dyadic:
	btst.b	#7,DTAG(a6)	;if dop = norm=000, zero=001,
*				;inf=010 or nan=011
	beq.b	monadic		;then branch
*				;else denorm
* HANDLE DESTINATION DENORM HERE
*				;set dtag to norm
*				;write the tag & fpte15 to the fstack
	lea.l	FPTEMP(a6),a0

	bclr.b	#sign_bit,LOCAL_EX(a0)
	sne	LOCAL_SGN(a0)

	bsr	nrm_set		;normalize number (exp will go negative)
	bclr.b	#sign_bit,LOCAL_EX(a0) ;get rid of false sign
	bfclr	LOCAL_SGN(a0){0:8}	;change back to IEEE ext format
	beq.b	dpos
	bset.b	#sign_bit,LOCAL_EX(a0)
dpos:
	bfclr	DTAG(a6){0:4}	;set tag to normalized, FPTE15 = 0
	bset.b	#4,DTAG(a6)	;set FPTE15
	or.b	#$0f,DNRM_FLG(a6)
monadic:
	lea.l	ETEMP(a6),a0
	btst.b	#direction_bit,CMDREG1B(a6)	;check direction
	bne.w	opclass3			;it is a mv out
*
* At this point, only oplcass 0 and 2 possible
*
	btst.b	#7,STAG(a6)	;if sop = norm=000, zero=001,
*				;inf=010 or nan=011
	bne.w	mon_dnrm	;else denorm
	tst.b	DY_MO_FLG(a6)	;all cases of dyadic instructions would
	bne.w	normal		;require normalization of denorm

* At this point:
*	monadic instructions:	fabs  = $18  fneg   = $1a  ftst   = $3a
*				fmove = $00  fsmove = $40  fdmove = $44
*				fsqrt = $05* fssqrt = $41  fdsqrt = $45
*				(*fsqrt reencoded to $05)
*
	move.w	CMDREG1B(a6),d0	;get command register
	andi.l	#$7f,d0			;strip to only command word
*
* At this point, fabs, fneg, fsmove, fdmove, ftst, fsqrt, fssqrt, and 
* fdsqrt are possible.
* For cases fabs, fneg, fsmove, and fdmove goto spos (do not normalize)
* For cases fsqrt, fssqrt, and fdsqrt goto nrm_src (do normalize)
*
	btst.l	#0,d0
	bne.w	normal			;weed out fsqrt instructions
*
* cu_norm handles fmove in instructions with normalized inputs.
* The routine round is used to correctly round the input for the
* destination precision and mode.
*
cu_norm:
	st	CU_ONLY(a6)		;set cu-only inst flag
	move.w	CMDREG1B(a6),d0
	andi.b	#$3b,d0		;isolate bits to select inst
	tst.b	d0
	beq.l	cu_nmove	;if zero, it is an fmove
	cmpi.b	#$18,d0
	beq.l	cu_nabs		;if $18, it is fabs
	cmpi.b	#$1a,d0
	beq.l	cu_nneg		;if $1a, it is fneg
*
* Inst is ftst.  Check the source operand and set the cc's accordingly.
* No write is done, so simply rts.
*
cu_ntst:
	move.w	LOCAL_EX(a0),d0
	bclr.l	#15,d0
	sne	LOCAL_SGN(a0)
	beq.b	cu_ntpo
	or.l	#neg_mask,USER_FPSR(a6) ;set N
cu_ntpo:
	cmpi.w	#$7fff,d0	;test for inf/nan
	bne.b	cu_ntcz
	tst.l	LOCAL_HI(a0)
	bne.b	cu_ntn
	tst.l	LOCAL_LO(a0)
	bne.b	cu_ntn
	or.l	#inf_mask,USER_FPSR(a6)
	rts
cu_ntn:
	or.l	#nan_mask,USER_FPSR(a6)
	move.l	ETEMP_EX(a6),FPTEMP_EX(a6)	;set up fptemp sign for 
*						;snan handler

	rts
cu_ntcz:
	tst.l	LOCAL_HI(a0)
	bne.l	cu_ntsx
	tst.l	LOCAL_LO(a0)
	bne.l	cu_ntsx
	or.l	#z_mask,USER_FPSR(a6)
cu_ntsx:
	rts
*
* Inst is fabs.  Execute the absolute value function on the input.
* Branch to the fmove code.  If the operand is NaN, do nothing.
*
cu_nabs:
	move.b	STAG(a6),d0
	btst.l	#5,d0			;test for NaN or zero
	bne	wr_etemp		;if either, simply write it
	bclr.b	#7,LOCAL_EX(a0)		;do abs
	bra.b	cu_nmove		;fmove code will finish
*
* Inst is fneg.  Execute the negate value function on the input.
* Fall though to the fmove code.  If the operand is NaN, do nothing.
*
cu_nneg:
	move.b	STAG(a6),d0
	btst.l	#5,d0			;test for NaN or zero
	bne	wr_etemp		;if either, simply write it
	bchg.b	#7,LOCAL_EX(a0)		;do neg
*
* Inst is fmove.  This code also handles all result writes.
* If bit 2 is set, round is forced to double.  If it is clear,
* and bit 6 is set, round is forced to single.  If both are clear,
* the round precision is found in the fpcr.  If the rounding precision
* is double or single, round the result before the write.
*
cu_nmove:
	move.b	STAG(a6),d0
	andi.b	#$e0,d0			;isolate stag bits
	bne	wr_etemp		;if not norm, simply write it
	btst.b	#2,CMDREG1B+1(a6)	;check for rd
	bne	cu_nmrd
	btst.b	#6,CMDREG1B+1(a6)	;check for rs
	bne	cu_nmrs
*
* The move or operation is not with forced precision.  Test for
* nan or inf as the input; if so, simply write it to FPn.  Use the
* FPCR_MODE byte to get rounding on norms and zeros.
*
cu_nmnr:
	bfextu	FPCR_MODE(a6){0:2},d0
	tst.b	d0			;check for extended
	beq	cu_wrexn		;if so, just write result
	cmpi.b	#1,d0			;check for single
	beq	cu_nmrs			;fall through to double
*
* The move is fdmove or round precision is double.
*
cu_nmrd:
	move.l	#2,d0			;set up the size for denorm
	move.w	LOCAL_EX(a0),d1		;compare exponent to double threshold
	and.w	#$7fff,d1	
	cmp.w	#$3c01,d1
	bls	cu_nunfl
	bfextu	FPCR_MODE(a6){2:2},d1	;get rmode
	or.l	#$00020000,d1		;or in rprec (double)
	clr.l	d0			;clear g,r,s for round
	bclr.b	#sign_bit,LOCAL_EX(a0)	;convert to internal format
	sne	LOCAL_SGN(a0)
	bsr.l	round
	bfclr	LOCAL_SGN(a0){0:8}
	beq.b	cu_nmrdc
	bset.b	#sign_bit,LOCAL_EX(a0)
cu_nmrdc:
	move.w	LOCAL_EX(a0),d1		;check for overflow
	and.w	#$7fff,d1
	cmp.w	#$43ff,d1
	bge	cu_novfl		;take care of overflow case
	bra.w	cu_wrexn
*
* The move is fsmove or round precision is single.
*
cu_nmrs:
	move.l	#1,d0
	move.w	LOCAL_EX(a0),d1
	and.w	#$7fff,d1
	cmp.w	#$3f81,d1
	bls	cu_nunfl
	bfextu	FPCR_MODE(a6){2:2},d1
	or.l	#$00010000,d1
	clr.l	d0
	bclr.b	#sign_bit,LOCAL_EX(a0)
	sne	LOCAL_SGN(a0)
	bsr.l	round
	bfclr	LOCAL_SGN(a0){0:8}
	beq.b	cu_nmrsc
	bset.b	#sign_bit,LOCAL_EX(a0)
cu_nmrsc:
	move.w	LOCAL_EX(a0),d1
	and.w	#$7FFF,d1
	cmp.w	#$407f,d1
	blt	cu_wrexn
*
* The operand is above precision boundaries.  Use t_ovfl to
* generate the correct value.
*
cu_novfl:
	bsr	t_ovfl
	bra	cu_wrexn
*
* The operand is below precision boundaries.  Use denorm to
* generate the correct value.
*
cu_nunfl:
	bclr.b	#sign_bit,LOCAL_EX(a0)
	sne	LOCAL_SGN(a0)
	bsr	denorm
	bfclr	LOCAL_SGN(a0){0:8}	;change back to IEEE ext format
	beq.b	cu_nucont
	bset.b	#sign_bit,LOCAL_EX(a0)
cu_nucont:
	bfextu	FPCR_MODE(a6){2:2},d1
	btst.b	#2,CMDREG1B+1(a6)	;check for rd
	bne	inst_d
	btst.b	#6,CMDREG1B+1(a6)	;check for rs
	bne	inst_s
	swap	d1
	move.b	FPCR_MODE(a6),d1
	lsr.b	#6,d1
	swap	d1
	bra	inst_sd
inst_d:
	or.l	#$00020000,d1
	bra	inst_sd
inst_s:
	or.l	#$00010000,d1
inst_sd:
	bclr.b	#sign_bit,LOCAL_EX(a0)
	sne	LOCAL_SGN(a0)
	bsr.l	round
	bfclr	LOCAL_SGN(a0){0:8}
	beq.b	cu_nuflp
	bset.b	#sign_bit,LOCAL_EX(a0)
cu_nuflp:
	btst.b	#inex2_bit,FPSR_EXCEPT(a6)
	beq.b	cu_nuninx
	or.l	#aunfl_mask,USER_FPSR(a6) ;if the round was inex, set AUNFL
cu_nuninx:
	tst.l	LOCAL_HI(a0)		;test for zero
	bne.b	cu_nunzro
	tst.l	LOCAL_LO(a0)
	bne.b	cu_nunzro
*
* The mantissa is zero from the denorm loop.  Check sign and rmode
* to see if rounding should have occurred which would leave the lsb.
*
	move.l	USER_FPCR(a6),d0
	andi.l	#$30,d0		;isolate rmode
	cmpi.l	#$20,d0
	blt.b	cu_nzro
	bne.b	cu_nrp
cu_nrm:
	tst.w	LOCAL_EX(a0)	;if positive, set lsb
	bge.b	cu_nzro
	btst.b	#7,FPCR_MODE(a6) ;check for double
	beq.b	cu_nincs
	bra.b	cu_nincd
cu_nrp:
	tst.w	LOCAL_EX(a0)	;if positive, set lsb
	blt.b	cu_nzro
	btst.b	#7,FPCR_MODE(a6) ;check for double
	beq.b	cu_nincs
cu_nincd:
	or.l	#$800,LOCAL_LO(a0) ;inc for double
	bra	cu_nunzro
cu_nincs:
	or.l	#$100,LOCAL_HI(a0) ;inc for single
	bra	cu_nunzro
cu_nzro:
	or.l	#z_mask,USER_FPSR(a6)
	move.b	STAG(a6),d0
	andi.b	#$e0,d0
	cmpi.b	#$40,d0		;check if input was tagged zero
	beq.b	cu_numv
cu_nunzro:
	or.l	#unfl_mask,USER_FPSR(a6) ;set unfl
cu_numv:
	move.l	(a0),ETEMP(a6)
	move.l	4(a0),ETEMP_HI(a6)
	move.l	8(a0),ETEMP_LO(a6)
*
* Write the result to memory, setting the fpsr cc bits.  NaN and Inf
* bypass cu_wrexn.
*
cu_wrexn:
	tst.w	LOCAL_EX(a0)		;test for zero
	beq.b	cu_wrzero
	cmp.w	#$8000,LOCAL_EX(a0)	;test for zero
	bne.b	cu_wreon
cu_wrzero:
	or.l	#z_mask,USER_FPSR(a6)	;set Z bit
cu_wreon:
	tst.w	LOCAL_EX(a0)
	bpl	wr_etemp
	or.l	#neg_mask,USER_FPSR(a6)
	bra	wr_etemp

*
* HANDLE SOURCE DENORM HERE
*
*				;clear denorm stag to norm
*				;write the new tag & ete15 to the fstack
mon_dnrm:
*
* At this point, check for the cases in which normalizing the 
* denorm produces incorrect results.
*
	tst.b	DY_MO_FLG(a6)	;all cases of dyadic instructions would
	bne.b	nrm_src		;require normalization of denorm

* At this point:
*	monadic instructions:	fabs  = $18  fneg   = $1a  ftst   = $3a
*				fmove = $00  fsmove = $40  fdmove = $44
*				fsqrt = $05* fssqrt = $41  fdsqrt = $45
*				(*fsqrt reencoded to $05)
*
	move.w	CMDREG1B(a6),d0	;get command register
	andi.l	#$7f,d0			;strip to only command word
*
* At this point, fabs, fneg, fsmove, fdmove, ftst, fsqrt, fssqrt, and 
* fdsqrt are possible.
* For cases fabs, fneg, fsmove, and fdmove goto spos (do not normalize)
* For cases fsqrt, fssqrt, and fdsqrt goto nrm_src (do normalize)
*
	btst.l	#0,d0
	bne.b	nrm_src		;weed out fsqrt instructions
	st	CU_ONLY(a6)	;set cu-only inst flag
	bra	cu_dnrm		;fmove, fabs, fneg, ftst 
*				;cases go to cu_dnrm
nrm_src:
	bclr.b	#sign_bit,LOCAL_EX(a0)
	sne	LOCAL_SGN(a0)
	bsr	nrm_set		;normalize number (exponent will go 
*				; negative)
	bclr.b	#sign_bit,LOCAL_EX(a0) ;get rid of false sign

	bfclr	LOCAL_SGN(a0){0:8}	;change back to IEEE ext format
	beq.b	spos
	bset.b	#sign_bit,LOCAL_EX(a0)
spos:
	bfclr	STAG(a6){0:4}	;set tag to normalized, FPTE15 = 0
	bset.b	#4,STAG(a6)	;set ETE15
	or.b	#$f0,DNRM_FLG(a6)
normal:
	tst.b	DNRM_FLG(a6)	;check if any of the ops were denorms
	bne	ck_wrap		;if so, check if it is a potential
*				;wrap-around case
fix_stk:
	move.b	#$fe,CU_SAVEPC(a6)
	bclr.b	#E1,E_BYTE(a6)

	clr.w	NMNEXC(a6)

	st.b	RES_FLG(a6)	;indicate that a restore is needed
	rts

*
* cu_dnrm handles all cu-only instructions (fmove, fabs, fneg, and
* ftst) completly in software without an frestore to the 040. 
*
cu_dnrm:
	st.b	CU_ONLY(a6)
	move.w	CMDREG1B(a6),d0
	andi.b	#$3b,d0		;isolate bits to select inst
	tst.b	d0
	beq.l	cu_dmove	;if zero, it is an fmove
	cmpi.b	#$18,d0
	beq.l	cu_dabs		;if $18, it is fabs
	cmpi.b	#$1a,d0
	beq.l	cu_dneg		;if $1a, it is fneg
*
* Inst is ftst.  Check the source operand and set the cc's accordingly.
* No write is done, so simply rts.
*
cu_dtst:
	move.w	LOCAL_EX(a0),d0
	bclr.l	#15,d0
	sne	LOCAL_SGN(a0)
	beq.b	cu_dtpo
	or.l	#neg_mask,USER_FPSR(a6) ;set N
cu_dtpo:
	cmpi.w	#$7fff,d0	;test for inf/nan
	bne.b	cu_dtcz
	tst.l	LOCAL_HI(a0)
	bne.b	cu_dtn
	tst.l	LOCAL_LO(a0)
	bne.b	cu_dtn
	or.l	#inf_mask,USER_FPSR(a6)
	rts
cu_dtn:
	or.l	#nan_mask,USER_FPSR(a6)
	move.l	ETEMP_EX(a6),FPTEMP_EX(a6)	;set up fptemp sign for 
*						;snan handler
	rts
cu_dtcz:
	tst.l	LOCAL_HI(a0)
	bne.l	cu_dtsx
	tst.l	LOCAL_LO(a0)
	bne.l	cu_dtsx
	or.l	#z_mask,USER_FPSR(a6)
cu_dtsx:
	rts
*
* Inst is fabs.  Execute the absolute value function on the input.
* Branch to the fmove code.
*
cu_dabs:
	bclr.b	#7,LOCAL_EX(a0)		;do abs
	bra.b	cu_dmove		;fmove code will finish
*
* Inst is fneg.  Execute the negate value function on the input.
* Fall though to the fmove code.
*
cu_dneg:
	bchg.b	#7,LOCAL_EX(a0)		;do neg
*
* Inst is fmove.  This code also handles all result writes.
* If bit 2 is set, round is forced to double.  If it is clear,
* and bit 6 is set, round is forced to single.  If both are clear,
* the round precision is found in the fpcr.  If the rounding precision
* is double or single, the result is zero, and the mode is checked
* to determine if the lsb of the result should be set.
*
cu_dmove:
	btst.b	#2,CMDREG1B+1(a6)	;check for rd
	bne	cu_dmrd
	btst.b	#6,CMDREG1B+1(a6)	;check for rs
	bne	cu_dmrs
*
* The move or operation is not with forced precision.  Use the
* FPCR_MODE byte to get rounding.
*
cu_dmnr:
	bfextu	FPCR_MODE(a6){0:2},d0
	tst.b	d0			;check for extended
	beq	cu_wrexd		;if so, just write result
	cmpi.b	#1,d0			;check for single
	beq	cu_dmrs			;fall through to double
*
* The move is fdmove or round precision is double.  Result is zero.
* Check rmode for rp or rm and set lsb accordingly.
*
cu_dmrd:
	bfextu	FPCR_MODE(a6){2:2},d1	;get rmode
	tst.w	LOCAL_EX(a0)		;check sign
	blt.b	cu_dmdn
	cmpi.b	#3,d1			;check for rp
	bne	cu_dpd			;load double pos zero
	bra	cu_dpdr			;load double pos zero w/lsb
cu_dmdn:
	cmpi.b	#2,d1			;check for rm
	bne	cu_dnd			;load double neg zero
	bra	cu_dndr			;load double neg zero w/lsb
*
* The move is fsmove or round precision is single.  Result is zero.
* Check for rp or rm and set lsb accordingly.
*
cu_dmrs:
	bfextu	FPCR_MODE(a6){2:2},d1	;get rmode
	tst.w	LOCAL_EX(a0)		;check sign
	blt.b	cu_dmsn
	cmpi.b	#3,d1			;check for rp
	bne	cu_spd			;load single pos zero
	bra	cu_spdr			;load single pos zero w/lsb
cu_dmsn:
	cmpi.b	#2,d1			;check for rm
	bne	cu_snd			;load single neg zero
	bra	cu_sndr			;load single neg zero w/lsb
*
* The precision is extended, so the result in etemp is correct.
* Simply set unfl (not inex2 or aunfl) and write the result to 
* the correct fp register.
cu_wrexd:
	or.l	#unfl_mask,USER_FPSR(a6)
	tst.w	LOCAL_EX(a0)
	beq	wr_etemp
	or.l	#neg_mask,USER_FPSR(a6)
	bra	wr_etemp
*
* These routines write +/- zero in double format.  The routines
* cu_dpdr and cu_dndr set the double lsb.
*
cu_dpd:
	move.l	#$3c010000,LOCAL_EX(a0)	;force pos double zero
	clr.l	LOCAL_HI(a0)
	clr.l	LOCAL_LO(a0)
	or.l	#z_mask,USER_FPSR(a6)
	or.l	#unfinx_mask,USER_FPSR(a6)
	bra	wr_etemp
cu_dpdr:
	move.l	#$3c010000,LOCAL_EX(a0)	;force pos double zero
	clr.l	LOCAL_HI(a0)
	move.l	#$800,LOCAL_LO(a0)	;with lsb set
	or.l	#unfinx_mask,USER_FPSR(a6)
	bra	wr_etemp
cu_dnd:
	move.l	#$bc010000,LOCAL_EX(a0)	;force pos double zero
	clr.l	LOCAL_HI(a0)
	clr.l	LOCAL_LO(a0)
	or.l	#z_mask,USER_FPSR(a6)
	or.l	#neg_mask,USER_FPSR(a6)
	or.l	#unfinx_mask,USER_FPSR(a6)
	bra	wr_etemp
cu_dndr:
	move.l	#$bc010000,LOCAL_EX(a0)	;force pos double zero
	clr.l	LOCAL_HI(a0)
	move.l	#$800,LOCAL_LO(a0)	;with lsb set
	or.l	#neg_mask,USER_FPSR(a6)
	or.l	#unfinx_mask,USER_FPSR(a6)
	bra	wr_etemp
*
* These routines write +/- zero in single format.  The routines
* cu_dpdr and cu_dndr set the single lsb.
*
cu_spd:
	move.l	#$3f810000,LOCAL_EX(a0)	;force pos single zero
	clr.l	LOCAL_HI(a0)
	clr.l	LOCAL_LO(a0)
	or.l	#z_mask,USER_FPSR(a6)
	or.l	#unfinx_mask,USER_FPSR(a6)
	bra	wr_etemp
cu_spdr:
	move.l	#$3f810000,LOCAL_EX(a0)	;force pos single zero
	move.l	#$100,LOCAL_HI(a0)	;with lsb set
	clr.l	LOCAL_LO(a0)
	or.l	#unfinx_mask,USER_FPSR(a6)
	bra	wr_etemp
cu_snd:
	move.l	#$bf810000,LOCAL_EX(a0)	;force pos single zero
	clr.l	LOCAL_HI(a0)
	clr.l	LOCAL_LO(a0)
	or.l	#z_mask,USER_FPSR(a6)
	or.l	#neg_mask,USER_FPSR(a6)
	or.l	#unfinx_mask,USER_FPSR(a6)
	bra	wr_etemp
cu_sndr:
	move.l	#$bf810000,LOCAL_EX(a0)	;force pos single zero
	move.l	#$100,LOCAL_HI(a0)	;with lsb set
	clr.l	LOCAL_LO(a0)
	or.l	#neg_mask,USER_FPSR(a6)
	or.l	#unfinx_mask,USER_FPSR(a6)
	bra	wr_etemp
	
*
* This code checks for 16-bit overflow conditions on dyadic
* operations which are not restorable into the floating-point
* unit and must be completed in software.  Basically, this
* condition exists with a very large norm and a denorm.  One
* of the operands must be denormalized to enter this code.
*
* Flags used:
*	DY_MO_FLG contains 0 for monadic op, $ff for dyadic
*	DNRM_FLG contains $00 for neither op denormalized
*	                  $0f for the destination op denormalized
*	                  $f0 for the source op denormalized
*	                  $ff for both ops denormalzed
*
* The wrap-around condition occurs for add, sub, div, and cmp
* when 
*
*	abs(dest_exp - src_exp) >= $8000
*
* and for mul when
*
*	(dest_exp + src_exp) < $0
*
* we must process the operation here if this case is true.
*
* The rts following the frcfpn routine is the exit from res_func
* for this condition.  The restore flag (RES_FLG) is left clear.
* No frestore is done unless an exception is to be reported.
*
* For fadd: 
*	if(sign_of(dest) != sign_of(src))
*		replace exponent of src with $3fff (keep sign)
*		use fpu to perform dest+new_src (user's rmode and X)
*		clr sticky
*	else
*		set sticky
*	call round with user's precision and mode
*	move result to fpn and wbtemp
*
* For fsub:
*	if(sign_of(dest) == sign_of(src))
*		replace exponent of src with $3fff (keep sign)
*		use fpu to perform dest+new_src (user's rmode and X)
*		clr sticky
*	else
*		set sticky
*	call round with user's precision and mode
*	move result to fpn and wbtemp
*
* For fdiv/fsgldiv:
*	if(both operands are denorm)
*		restore_to_fpu;
*	if(dest is norm)
*		force_ovf;
*	else(dest is denorm)
*		force_unf:
*
* For fcmp:
*	if(dest is norm)
*		N = sign_of(dest);
*	else(dest is denorm)
*		N = sign_of(src);
*
* For fmul:
*	if(both operands are denorm)
*		force_unf;
*	if((dest_exp + src_exp) < 0)
*		force_unf:
*	else
*		restore_to_fpu;
*
* local equates:
addcode	equ	$22
subcode	equ	$28
mulcode	equ	$23
divcode	equ	$20
cmpcode	equ	$38
ck_wrap:
	tst.b	DY_MO_FLG(a6)	;check for fsqrt
	beq	fix_stk		;if zero, it is fsqrt
	move.w	CMDREG1B(a6),d0
	andi.w	#$3b,d0		;strip to command bits
	cmpi.w	#addcode,d0
	beq	wrap_add
	cmpi.w	#subcode,d0
	beq	wrap_sub
	cmpi.w	#mulcode,d0
	beq	wrap_mul
	cmpi.w	#cmpcode,d0
	beq	wrap_cmp
*
* Inst is fdiv.  
*
wrap_div:
	cmp.b	#$ff,DNRM_FLG(a6) ;if both ops denorm, 
	beq	fix_stk		 ;restore to fpu
*
* One of the ops is denormalized.  Test for wrap condition
* and force the result.
*
	cmp.b	#$0f,DNRM_FLG(a6) ;check for dest denorm
	bne.b	div_srcd
div_destd:
	bsr.l	ckinf_ns
	bne	fix_stk
	bfextu	ETEMP_EX(a6){1:15},d0	;get src exp (always pos)
	bfexts	FPTEMP_EX(a6){1:15},d1	;get dest exp (always neg)
	sub.l	d1,d0			;subtract dest from src
	cmp.l	#$7fff,d0
	blt	fix_stk			;if less, not wrap case
	clr.b	WBTEMP_SGN(a6)
	move.w	ETEMP_EX(a6),d0		;find the sign of the result
	move.w	FPTEMP_EX(a6),d1
	eor.w	d1,d0
	andi.w	#$8000,d0
	beq	force_unf
	st.b	WBTEMP_SGN(a6)
	bra	force_unf

ckinf_ns:
	move.b	STAG(a6),d0		;check source tag for inf or nan
	bra	ck_in_com
ckinf_nd:
	move.b	DTAG(a6),d0		;check destination tag for inf or nan
ck_in_com:	
	andi.b	#$60,d0			;isolate tag bits
	cmp.b	#$40,d0			;is it inf?
	beq	nan_or_inf		;not wrap case
	cmp.b	#$60,d0			;is it nan?
	beq	nan_or_inf		;yes, not wrap case?
	cmp.b	#$20,d0			;is it a zero?
	beq	nan_or_inf		;yes
	clr.l	d0
	rts				;then it is either a zero of norm,
*					;check wrap case
nan_or_inf:
	moveq.l	#-1,d0
	rts



div_srcd:
	bsr.l	ckinf_nd
	bne	fix_stk
	bfextu	FPTEMP_EX(a6){1:15},d0	;get dest exp (always pos)
	bfexts	ETEMP_EX(a6){1:15},d1	;get src exp (always neg)
	sub.l	d1,d0			;subtract src from dest
	cmp.l	#$8000,d0
	blt	fix_stk			;if less, not wrap case
	clr.b	WBTEMP_SGN(a6)
	move.w	ETEMP_EX(a6),d0		;find the sign of the result
	move.w	FPTEMP_EX(a6),d1
	eor.w	d1,d0
	andi.w	#$8000,d0
	beq.b	force_ovf
	st.b	WBTEMP_SGN(a6)
*
* This code handles the case of the instruction resulting in 
* an overflow condition.
*
force_ovf:
	bclr.b	#E1,E_BYTE(a6)
	or.l	#ovfl_inx_mask,USER_FPSR(a6)
	clr.w	NMNEXC(a6)
	lea.l	WBTEMP(a6),a0		;point a0 to memory location
	move.w	CMDREG1B(a6),d0
	btst.l	#6,d0			;test for forced precision
	beq.b	frcovf_fpcr
	btst.l	#2,d0			;check for double
	bne.b	frcovf_dbl
	move.l	#$1,d0			;inst is forced single
	bra.b	frcovf_rnd
frcovf_dbl:
	move.l	#$2,d0			;inst is forced double
	bra.b	frcovf_rnd
frcovf_fpcr:
	bfextu	FPCR_MODE(a6){0:2},d0	;inst not forced - use fpcr prec
frcovf_rnd:

* The 881/882 does not set inex2 for the following case, so the 
* line is commented out to be compatible with 881/882
*	tst.b	d0
*	beq.b	frcovf_x
*	or.l	#inex2_mask,USER_FPSR(a6) ;if prec is s or d, set inex2

*frcovf_x:
	bsr.l	ovf_res			;get correct result based on
*					;round precision/mode.  This 
*					;sets FPSR_CC correctly
*					;returns in external format
	bfclr	WBTEMP_SGN(a6){0:8}
	beq	frcfpn
	bset.b	#sign_bit,WBTEMP_EX(a6)
	bra	frcfpn
*
* Inst is fadd.
*
wrap_add:
	cmp.b	#$ff,DNRM_FLG(a6) ;if both ops denorm, 
	beq	fix_stk		 ;restore to fpu
*
* One of the ops is denormalized.  Test for wrap condition
* and complete the instruction.
*
	cmp.b	#$0f,DNRM_FLG(a6) ;check for dest denorm
	bne.b	add_srcd
add_destd:
	bsr.l	ckinf_ns
	bne	fix_stk
	bfextu	ETEMP_EX(a6){1:15},d0	;get src exp (always pos)
	bfexts	FPTEMP_EX(a6){1:15},d1	;get dest exp (always neg)
	sub.l	d1,d0			;subtract dest from src
	cmp.l	#$8000,d0
	blt	fix_stk			;if less, not wrap case
	bra	add_wrap
add_srcd:
	bsr.l	ckinf_nd
	bne	fix_stk
	bfextu	FPTEMP_EX(a6){1:15},d0	;get dest exp (always pos)
	bfexts	ETEMP_EX(a6){1:15},d1	;get src exp (always neg)
	sub.l	d1,d0			;subtract src from dest
	cmp.l	#$8000,d0
	blt	fix_stk			;if less, not wrap case
*
* Check the signs of the operands.  If they are unlike, the fpu
* can be used to add the norm and 1.0 with the sign of the
* denorm and it will correctly generate the result in extended
* precision.  We can then call round with no sticky and the result
* will be correct for the user's rounding mode and precision.  If
* the signs are the same, we call round with the sticky bit set
* and the result will be correctfor the user's rounding mode and
* precision.
*
add_wrap:
	move.w	ETEMP_EX(a6),d0
	move.w	FPTEMP_EX(a6),d1
	eor.w	d1,d0
	andi.w	#$8000,d0
	beq	add_same
*
* The signs are unlike.
*
	cmp.b	#$0f,DNRM_FLG(a6) ;is dest the denorm?
	bne.b	add_u_srcd
	move.w	FPTEMP_EX(a6),d0
	andi.w	#$8000,d0
	or.w	#$3fff,d0	;force the exponent to +/- 1
	move.w	d0,FPTEMP_EX(a6) ;in the denorm
	move.l	USER_FPCR(a6),d0
	andi.l	#$30,d0
	fmove.l	d0,fpcr		;set up users rmode and X
	fmove.x	ETEMP(a6),fp0
	fadd.x	FPTEMP(a6),fp0
	lea.l	WBTEMP(a6),a0	;point a0 to wbtemp in frame
	fmove.l	fpsr,d1
	or.l	d1,USER_FPSR(a6) ;capture cc's and inex from fadd
	fmove.x	fp0,WBTEMP(a6)	;write result to memory
	lsr.l	#4,d0		;put rmode in lower 2 bits
	move.l	USER_FPCR(a6),d1
	andi.l	#$c0,d1
	lsr.l	#6,d1		;put precision in upper word
	swap	d1
	or.l	d0,d1		;set up for round call
	clr.l	d0		;force sticky to zero
	bclr.b	#sign_bit,WBTEMP_EX(a6)
	sne	WBTEMP_SGN(a6)
	bsr.l	round		;round result to users rmode & prec
	bfclr	WBTEMP_SGN(a6){0:8}	;convert back to IEEE ext format
	beq	frcfpnr
	bset.b	#sign_bit,WBTEMP_EX(a6)
	bra	frcfpnr
add_u_srcd:
	move.w	ETEMP_EX(a6),d0
	andi.w	#$8000,d0
	or.w	#$3fff,d0	;force the exponent to +/- 1
	move.w	d0,ETEMP_EX(a6) ;in the denorm
	move.l	USER_FPCR(a6),d0
	andi.l	#$30,d0
	fmove.l	d0,fpcr		;set up users rmode and X
	fmove.x	ETEMP(a6),fp0
	fadd.x	FPTEMP(a6),fp0
	fmove.l	fpsr,d1
	or.l	d1,USER_FPSR(a6) ;capture cc's and inex from fadd
	lea.l	WBTEMP(a6),a0	;point a0 to wbtemp in frame
	fmove.x	fp0,WBTEMP(a6)	;write result to memory
	lsr.l	#4,d0		;put rmode in lower 2 bits
	move.l	USER_FPCR(a6),d1
	andi.l	#$c0,d1
	lsr.l	#6,d1		;put precision in upper word
	swap	d1
	or.l	d0,d1		;set up for round call
	clr.l	d0		;force sticky to zero
	bclr.b	#sign_bit,WBTEMP_EX(a6)
	sne	WBTEMP_SGN(a6)	;use internal format for round
	bsr.l	round		;round result to users rmode & prec
	bfclr	WBTEMP_SGN(a6){0:8}	;convert back to IEEE ext format
	beq	frcfpnr
	bset.b	#sign_bit,WBTEMP_EX(a6)
	bra	frcfpnr
*
* Signs are alike:
*
add_same:
	cmp.b	#$0f,DNRM_FLG(a6) ;is dest the denorm?
	bne.b	add_s_srcd
add_s_destd:
	lea.l	ETEMP(a6),a0
	move.l	USER_FPCR(a6),d0
	andi.l	#$30,d0
	lsr.l	#4,d0		;put rmode in lower 2 bits
	move.l	USER_FPCR(a6),d1
	andi.l	#$c0,d1
	lsr.l	#6,d1		;put precision in upper word
	swap	d1
	or.l	d0,d1		;set up for round call
	move.l	#$20000000,d0	;set sticky for round
	bclr.b	#sign_bit,ETEMP_EX(a6)
	sne	ETEMP_SGN(a6)
	bsr.l	round		;round result to users rmode & prec
	bfclr	ETEMP_SGN(a6){0:8}	;convert back to IEEE ext format
	beq.b	add_s_dclr
	bset.b	#sign_bit,ETEMP_EX(a6)
add_s_dclr:
	lea.l	WBTEMP(a6),a0
	move.l	ETEMP(a6),(a0)	;write result to wbtemp
	move.l	ETEMP_HI(a6),4(a0)
	move.l	ETEMP_LO(a6),8(a0)
	tst.w	ETEMP_EX(a6)
	bgt	add_ckovf
	or.l	#neg_mask,USER_FPSR(a6)
	bra	add_ckovf
add_s_srcd:
	lea.l	FPTEMP(a6),a0
	move.l	USER_FPCR(a6),d0
	andi.l	#$30,d0
	lsr.l	#4,d0		;put rmode in lower 2 bits
	move.l	USER_FPCR(a6),d1
	andi.l	#$c0,d1
	lsr.l	#6,d1		;put precision in upper word
	swap	d1
	or.l	d0,d1		;set up for round call
	move.l	#$20000000,d0	;set sticky for round
	bclr.b	#sign_bit,FPTEMP_EX(a6)
	sne	FPTEMP_SGN(a6)
	bsr.l	round		;round result to users rmode & prec
	bfclr	FPTEMP_SGN(a6){0:8}	;convert back to IEEE ext format
	beq.b	add_s_sclr
	bset.b	#sign_bit,FPTEMP_EX(a6)
add_s_sclr:
	lea.l	WBTEMP(a6),a0
	move.l	FPTEMP(a6),(a0)	;write result to wbtemp
	move.l	FPTEMP_HI(a6),4(a0)
	move.l	FPTEMP_LO(a6),8(a0)
	tst.w	FPTEMP_EX(a6)
	bgt	add_ckovf
	or.l	#neg_mask,USER_FPSR(a6)
add_ckovf:
	move.w	WBTEMP_EX(a6),d0
	andi.w	#$7fff,d0
	cmpi.w	#$7fff,d0
	bne	frcfpnr
*
* The result has overflowed to $7fff exponent.  Set I, ovfl,
* and aovfl, and clr the mantissa (incorrectly set by the
* round routine.)
*
	or.l	#inf_mask+ovfl_inx_mask,USER_FPSR(a6)	
	clr.l	4(a0)
	bra	frcfpnr
*
* Inst is fsub.
*
wrap_sub:
	cmp.b	#$ff,DNRM_FLG(a6) ;if both ops denorm, 
	beq	fix_stk		 ;restore to fpu
*
* One of the ops is denormalized.  Test for wrap condition
* and complete the instruction.
*
	cmp.b	#$0f,DNRM_FLG(a6) ;check for dest denorm
	bne.b	sub_srcd
sub_destd:
	bsr.l	ckinf_ns
	bne	fix_stk
	bfextu	ETEMP_EX(a6){1:15},d0	;get src exp (always pos)
	bfexts	FPTEMP_EX(a6){1:15},d1	;get dest exp (always neg)
	sub.l	d1,d0			;subtract src from dest
	cmp.l	#$8000,d0
	blt	fix_stk			;if less, not wrap case
	bra	sub_wrap
sub_srcd:
	bsr.l	ckinf_nd
	bne	fix_stk
	bfextu	FPTEMP_EX(a6){1:15},d0	;get dest exp (always pos)
	bfexts	ETEMP_EX(a6){1:15},d1	;get src exp (always neg)
	sub.l	d1,d0			;subtract dest from src
	cmp.l	#$8000,d0
	blt	fix_stk			;if less, not wrap case
*
* Check the signs of the operands.  If they are alike, the fpu
* can be used to subtract from the norm 1.0 with the sign of the
* denorm and it will correctly generate the result in extended
* precision.  We can then call round with no sticky and the result
* will be correct for the user's rounding mode and precision.  If
* the signs are unlike, we call round with the sticky bit set
* and the result will be correctfor the user's rounding mode and
* precision.
*
sub_wrap:
	move.w	ETEMP_EX(a6),d0
	move.w	FPTEMP_EX(a6),d1
	eor.w	d1,d0
	andi.w	#$8000,d0
	bne	sub_diff
*
* The signs are alike.
*
	cmp.b	#$0f,DNRM_FLG(a6) ;is dest the denorm?
	bne.b	sub_u_srcd
	move.w	FPTEMP_EX(a6),d0
	andi.w	#$8000,d0
	or.w	#$3fff,d0	;force the exponent to +/- 1
	move.w	d0,FPTEMP_EX(a6) ;in the denorm
	move.l	USER_FPCR(a6),d0
	andi.l	#$30,d0
	fmove.l	d0,fpcr		;set up users rmode and X
	fmove.x	FPTEMP(a6),fp0
	fsub.x	ETEMP(a6),fp0
	fmove.l	fpsr,d1
	or.l	d1,USER_FPSR(a6) ;capture cc's and inex from fadd
	lea.l	WBTEMP(a6),a0	;point a0 to wbtemp in frame
	fmove.x	fp0,WBTEMP(a6)	;write result to memory
	lsr.l	#4,d0		;put rmode in lower 2 bits
	move.l	USER_FPCR(a6),d1
	andi.l	#$c0,d1
	lsr.l	#6,d1		;put precision in upper word
	swap	d1
	or.l	d0,d1		;set up for round call
	clr.l	d0		;force sticky to zero
	bclr.b	#sign_bit,WBTEMP_EX(a6)
	sne	WBTEMP_SGN(a6)
	bsr.l	round		;round result to users rmode & prec
	bfclr	WBTEMP_SGN(a6){0:8}	;convert back to IEEE ext format
	beq	frcfpnr
	bset.b	#sign_bit,WBTEMP_EX(a6)
	bra	frcfpnr
sub_u_srcd:
	move.w	ETEMP_EX(a6),d0
	andi.w	#$8000,d0
	or.w	#$3fff,d0	;force the exponent to +/- 1
	move.w	d0,ETEMP_EX(a6) ;in the denorm
	move.l	USER_FPCR(a6),d0
	andi.l	#$30,d0
	fmove.l	d0,fpcr		;set up users rmode and X
	fmove.x	FPTEMP(a6),fp0
	fsub.x	ETEMP(a6),fp0
	fmove.l	fpsr,d1
	or.l	d1,USER_FPSR(a6) ;capture cc's and inex from fadd
	lea.l	WBTEMP(a6),a0	;point a0 to wbtemp in frame
	fmove.x	fp0,WBTEMP(a6)	;write result to memory
	lsr.l	#4,d0		;put rmode in lower 2 bits
	move.l	USER_FPCR(a6),d1
	andi.l	#$c0,d1
	lsr.l	#6,d1		;put precision in upper word
	swap	d1
	or.l	d0,d1		;set up for round call
	clr.l	d0		;force sticky to zero
	bclr.b	#sign_bit,WBTEMP_EX(a6)
	sne	WBTEMP_SGN(a6)
	bsr.l	round		;round result to users rmode & prec
	bfclr	WBTEMP_SGN(a6){0:8}	;convert back to IEEE ext format
	beq	frcfpnr
	bset.b	#sign_bit,WBTEMP_EX(a6)
	bra	frcfpnr
*
* Signs are unlike:
*
sub_diff:
	cmp.b	#$0f,DNRM_FLG(a6) ;is dest the denorm?
	bne.b	sub_s_srcd
sub_s_destd:
	lea.l	ETEMP(a6),a0
	move.l	USER_FPCR(a6),d0
	andi.l	#$30,d0
	lsr.l	#4,d0		;put rmode in lower 2 bits
	move.l	USER_FPCR(a6),d1
	andi.l	#$c0,d1
	lsr.l	#6,d1		;put precision in upper word
	swap	d1
	or.l	d0,d1		;set up for round call
	move.l	#$20000000,d0	;set sticky for round
*
* Since the dest is the denorm, the sign is the opposite of the
* norm sign.
*
	eori.w	#$8000,ETEMP_EX(a6)	;flip sign on result
	tst.w	ETEMP_EX(a6)
	bgt.b	sub_s_dwr
	or.l	#neg_mask,USER_FPSR(a6)
sub_s_dwr:
	bclr.b	#sign_bit,ETEMP_EX(a6)
	sne	ETEMP_SGN(a6)
	bsr.l	round		;round result to users rmode & prec
	bfclr	ETEMP_SGN(a6){0:8}	;convert back to IEEE ext format
	beq.b	sub_s_dclr
	bset.b	#sign_bit,ETEMP_EX(a6)
sub_s_dclr:
	lea.l	WBTEMP(a6),a0
	move.l	ETEMP(a6),(a0)	;write result to wbtemp
	move.l	ETEMP_HI(a6),4(a0)
	move.l	ETEMP_LO(a6),8(a0)
	bra	sub_ckovf
sub_s_srcd:
	lea.l	FPTEMP(a6),a0
	move.l	USER_FPCR(a6),d0
	andi.l	#$30,d0
	lsr.l	#4,d0		;put rmode in lower 2 bits
	move.l	USER_FPCR(a6),d1
	andi.l	#$c0,d1
	lsr.l	#6,d1		;put precision in upper word
	swap	d1
	or.l	d0,d1		;set up for round call
	move.l	#$20000000,d0	;set sticky for round
	bclr.b	#sign_bit,FPTEMP_EX(a6)
	sne	FPTEMP_SGN(a6)
	bsr.l	round		;round result to users rmode & prec
	bfclr	FPTEMP_SGN(a6){0:8}	;convert back to IEEE ext format
	beq.b	sub_s_sclr
	bset.b	#sign_bit,FPTEMP_EX(a6)
sub_s_sclr:
	lea.l	WBTEMP(a6),a0
	move.l	FPTEMP(a6),(a0)	;write result to wbtemp
	move.l	FPTEMP_HI(a6),4(a0)
	move.l	FPTEMP_LO(a6),8(a0)
	tst.w	FPTEMP_EX(a6)
	bgt	sub_ckovf
	or.l	#neg_mask,USER_FPSR(a6)
sub_ckovf:
	move.w	WBTEMP_EX(a6),d0
	andi.w	#$7fff,d0
	cmpi.w	#$7fff,d0
	bne	frcfpnr
*
* The result has overflowed to $7fff exponent.  Set I, ovfl,
* and aovfl, and clr the mantissa (incorrectly set by the
* round routine.)
*
	or.l	#inf_mask+ovfl_inx_mask,USER_FPSR(a6)	
	clr.l	4(a0)
	bra	frcfpnr
*
* Inst is fcmp.
*
wrap_cmp:
	cmp.b	#$ff,DNRM_FLG(a6) ;if both ops denorm, 
	beq	fix_stk		 ;restore to fpu
*
* One of the ops is denormalized.  Test for wrap condition
* and complete the instruction.
*
	cmp.b	#$0f,DNRM_FLG(a6) ;check for dest denorm
	bne.b	cmp_srcd
cmp_destd:
	bsr.l	ckinf_ns
	bne	fix_stk
	bfextu	ETEMP_EX(a6){1:15},d0	;get src exp (always pos)
	bfexts	FPTEMP_EX(a6){1:15},d1	;get dest exp (always neg)
	sub.l	d1,d0			;subtract dest from src
	cmp.l	#$8000,d0
	blt	fix_stk			;if less, not wrap case
	tst.w	ETEMP_EX(a6)		;set N to ~sign_of(src)
	bge	cmp_setn
	rts
cmp_srcd:
	bsr.l	ckinf_nd
	bne	fix_stk
	bfextu	FPTEMP_EX(a6){1:15},d0	;get dest exp (always pos)
	bfexts	ETEMP_EX(a6){1:15},d1	;get src exp (always neg)
	sub.l	d1,d0			;subtract src from dest
	cmp.l	#$8000,d0
	blt	fix_stk			;if less, not wrap case
	tst.w	FPTEMP_EX(a6)		;set N to sign_of(dest)
	blt	cmp_setn
	rts
cmp_setn:
	or.l	#neg_mask,USER_FPSR(a6)
	rts

*
* Inst is fmul.
*
wrap_mul:
	cmp.b	#$ff,DNRM_FLG(a6) ;if both ops denorm, 
	beq	force_unf	;force an underflow (really!)
*
* One of the ops is denormalized.  Test for wrap condition
* and complete the instruction.
*
	cmp.b	#$0f,DNRM_FLG(a6) ;check for dest denorm
	bne.b	mul_srcd
mul_destd:
	bsr.l	ckinf_ns
	bne	fix_stk
	bfextu	ETEMP_EX(a6){1:15},d0	;get src exp (always pos)
	bfexts	FPTEMP_EX(a6){1:15},d1	;get dest exp (always neg)
	add.l	d1,d0			;subtract dest from src
	bgt	fix_stk
	bra	force_unf
mul_srcd:
	bsr.l	ckinf_nd
	bne	fix_stk
	bfextu	FPTEMP_EX(a6){1:15},d0	;get dest exp (always pos)
	bfexts	ETEMP_EX(a6){1:15},d1	;get src exp (always neg)
	add.l	d1,d0			;subtract src from dest
	bgt	fix_stk
	
*
* This code handles the case of the instruction resulting in 
* an underflow condition.
*
force_unf:
	bclr.b	#E1,E_BYTE(a6)
	or.l	#unfinx_mask,USER_FPSR(a6)
	clr.w	NMNEXC(a6)
	clr.b	WBTEMP_SGN(a6)
	move.w	ETEMP_EX(a6),d0		;find the sign of the result
	move.w	FPTEMP_EX(a6),d1
	eor.w	d1,d0
	andi.w	#$8000,d0
	beq.b	frcunfcont
	st.b	WBTEMP_SGN(a6)
frcunfcont:
	lea	WBTEMP(a6),a0		;point a0 to memory location
	move.w	CMDREG1B(a6),d0
	btst.l	#6,d0			;test for forced precision
	beq.b	frcunf_fpcr
	btst.l	#2,d0			;check for double
	bne.b	frcunf_dbl
	move.l	#$1,d0			;inst is forced single
	bra.b	frcunf_rnd
frcunf_dbl:
	move.l	#$2,d0			;inst is forced double
	bra.b	frcunf_rnd
frcunf_fpcr:
	bfextu	FPCR_MODE(a6){0:2},d0	;inst not forced - use fpcr prec
frcunf_rnd:
	bsr.l	unf_sub			;get correct result based on
*					;round precision/mode.  This 
*					;sets FPSR_CC correctly
	bfclr	WBTEMP_SGN(a6){0:8}	;convert back to IEEE ext format
	beq.b	frcfpn
	bset.b	#sign_bit,WBTEMP_EX(a6)
	bra	frcfpn

*
* Write the result to the user's fpn.  All results must be HUGE to be
* written; otherwise the results would have overflowed or underflowed.
* If the rounding precision is single or double, the ovf_res routine
* is needed to correctly supply the max value.
*
frcfpnr:
	move.w	CMDREG1B(a6),d0
	btst.l	#6,d0			;test for forced precision
	beq.b	frcfpn_fpcr
	btst.l	#2,d0			;check for double
	bne.b	frcfpn_dbl
	move.l	#$1,d0			;inst is forced single
	bra.b	frcfpn_rnd
frcfpn_dbl:
	move.l	#$2,d0			;inst is forced double
	bra.b	frcfpn_rnd
frcfpn_fpcr:
	bfextu	FPCR_MODE(a6){0:2},d0	;inst not forced - use fpcr prec
	tst.b	d0
	beq.b	frcfpn			;if extended, write what you got
frcfpn_rnd:
	bclr.b	#sign_bit,WBTEMP_EX(a6)
	sne	WBTEMP_SGN(a6)
	bsr.l	ovf_res			;get correct result based on
*					;round precision/mode.  This 
*					;sets FPSR_CC correctly
	bfclr	WBTEMP_SGN(a6){0:8}	;convert back to IEEE ext format
	beq.b	frcfpn_clr
	bset.b	#sign_bit,WBTEMP_EX(a6)
frcfpn_clr:
	or.l	#ovfinx_mask,USER_FPSR(a6)
* 
* Perform the write.
*
frcfpn:
	bfextu	CMDREG1B(a6){6:3},d0	;extract fp destination register
	cmpi.b	#3,d0
	ble.b	frc0123			;check if dest is fp0-fp3
	move.l	#7,d1
	sub.l	d0,d1
	clr.l	d0
	bset.l	d1,d0
	fmovem.x WBTEMP(a6),d0
	rts
frc0123:
	tst.b	d0
	beq.b	frc0_dst
	cmpi.b	#1,d0
	beq.b	frc1_dst 
	cmpi.b	#2,d0
	beq.b	frc2_dst 
frc3_dst:
	move.l	WBTEMP_EX(a6),USER_FP3(a6)
	move.l	WBTEMP_HI(a6),USER_FP3+4(a6)
	move.l	WBTEMP_LO(a6),USER_FP3+8(a6)
	rts
frc2_dst:
	move.l	WBTEMP_EX(a6),USER_FP2(a6)
	move.l	WBTEMP_HI(a6),USER_FP2+4(a6)
	move.l	WBTEMP_LO(a6),USER_FP2+8(a6)
	rts
frc1_dst:
	move.l	WBTEMP_EX(a6),USER_FP1(a6)
	move.l	WBTEMP_HI(a6),USER_FP1+4(a6)
	move.l	WBTEMP_LO(a6),USER_FP1+8(a6)
	rts
frc0_dst:
	move.l	WBTEMP_EX(a6),USER_FP0(a6)
	move.l	WBTEMP_HI(a6),USER_FP0+4(a6)
	move.l	WBTEMP_LO(a6),USER_FP0+8(a6)
	rts

*
* Write etemp to fpn.
* A check is made on enabled and signalled snan exceptions,
* and the destination is not overwritten if this condition exists.
* This code is designed to make fmoveins of unsupported data types
* faster.
*
wr_etemp:
	btst.b	#snan_bit,FPSR_EXCEPT(a6)	;if snan is set, and
	beq.b	fmoveinc		;enabled, force restore
	btst.b	#snan_bit,FPCR_ENABLE(a6) ;and don't overwrite
	beq.b	fmoveinc		;the dest
	move.l	ETEMP_EX(a6),FPTEMP_EX(a6)	;set up fptemp sign for 
*						;snan handler
	tst.b	ETEMP(a6)		;check for negative
	blt.b	snan_neg
	rts
snan_neg:
	or.l	#neg_bit,USER_FPSR(a6)	;snan is negative; set N
	rts
fmoveinc:
	clr.w	NMNEXC(a6)
	bclr.b	#E1,E_BYTE(a6)
	move.b	STAG(a6),d0		;check if stag is inf
	andi.b	#$e0,d0
	cmpi.b	#$40,d0
	bne.b	fminc_cnan
	or.l	#inf_mask,USER_FPSR(a6) ;if inf, nothing yet has set I
	tst.w	LOCAL_EX(a0)		;check sign
	bge.b	fminc_con
	or.l	#neg_mask,USER_FPSR(a6)
	bra	fminc_con
fminc_cnan:
	cmpi.b	#$60,d0			;check if stag is NaN
	bne.b	fminc_czero
	or.l	#nan_mask,USER_FPSR(a6) ;if nan, nothing yet has set NaN
	move.l	ETEMP_EX(a6),FPTEMP_EX(a6)	;set up fptemp sign for 
*						;snan handler
	tst.w	LOCAL_EX(a0)		;check sign
	bge.b	fminc_con
	or.l	#neg_mask,USER_FPSR(a6)
	bra	fminc_con
fminc_czero:
	cmpi.b	#$20,d0			;check if zero
	bne.b	fminc_con
	or.l	#z_mask,USER_FPSR(a6)	;if zero, set Z
	tst.w	LOCAL_EX(a0)		;check sign
	bge.b	fminc_con
	or.l	#neg_mask,USER_FPSR(a6)
fminc_con:
	bfextu	CMDREG1B(a6){6:3},d0	;extract fp destination register
	cmpi.b	#3,d0
	ble.b	fp0123			;check if dest is fp0-fp3
	move.l	#7,d1
	sub.l	d0,d1
	clr.l	d0
	bset.l	d1,d0
	fmovem.x ETEMP(a6),d0
	rts

fp0123:
	tst.b	d0
	beq.b	fp0_dst
	cmpi.b	#1,d0
	beq.b	fp1_dst 
	cmpi.b	#2,d0
	beq.b	fp2_dst 
fp3_dst:
	move.l	ETEMP_EX(a6),USER_FP3(a6)
	move.l	ETEMP_HI(a6),USER_FP3+4(a6)
	move.l	ETEMP_LO(a6),USER_FP3+8(a6)
	rts
fp2_dst:
	move.l	ETEMP_EX(a6),USER_FP2(a6)
	move.l	ETEMP_HI(a6),USER_FP2+4(a6)
	move.l	ETEMP_LO(a6),USER_FP2+8(a6)
	rts
fp1_dst:
	move.l	ETEMP_EX(a6),USER_FP1(a6)
	move.l	ETEMP_HI(a6),USER_FP1+4(a6)
	move.l	ETEMP_LO(a6),USER_FP1+8(a6)
	rts
fp0_dst:
	move.l	ETEMP_EX(a6),USER_FP0(a6)
	move.l	ETEMP_HI(a6),USER_FP0+4(a6)
	move.l	ETEMP_LO(a6),USER_FP0+8(a6)
	rts

opclass3:
	st.b	CU_ONLY(a6)
	move.w	CMDREG1B(a6),d0	;check if packed moveout
	andi.w	#$0c00,d0	;isolate last 2 bits of size field
	cmpi.w	#$0c00,d0	;if size is 011 or 111, it is packed
	beq.w	pack_out	;else it is norm or denorm
	bra.w	mv_out

	
*
*	MOVE OUT
*

mv_tbl:
	dc.l	li
	dc.l 	sgp
	dc.l 	xp
	dc.l 	mvout_end	;should never be taken
	dc.l 	wi
	dc.l 	dp
	dc.l 	bi
	dc.l 	mvout_end	;should never be taken
mv_out:
	bfextu	CMDREG1B(a6){3:3},d1	;put source specifier in d1
	lea.l	mv_tbl,a0
	move.l	(a0,d1*4),a0
	jmp	(a0)

*
* This exit is for move-out to memory.  The aunfl bit is 
* set if the result is inex and unfl is signalled.
*
mvout_end:
	btst.b	#inex2_bit,FPSR_EXCEPT(a6)
	beq.b	no_aufl
	btst.b	#unfl_bit,FPSR_EXCEPT(a6)
	beq.b	no_aufl
	bset.b	#aunfl_bit,FPSR_AEXCEPT(a6)
no_aufl:
	clr.w	NMNEXC(a6)
	bclr.b	#E1,E_BYTE(a6)
	fmove.l	#0,FPSR			;clear any cc bits from res_func
*
* Return ETEMP to extended format from internal extended format so
* that gen_except will have a correctly signed value for ovfl/unfl
* handlers.
*
	bfclr	ETEMP_SGN(a6){0:8}
	beq.b	mvout_con
	bset.b	#sign_bit,ETEMP_EX(a6)
mvout_con:
	rts
*
* This exit is for move-out to int register.  The aunfl bit is 
* not set in any case for this move.
*
mvouti_end:
	clr.w	NMNEXC(a6)
	bclr.b	#E1,E_BYTE(a6)
	fmove.l	#0,FPSR			;clear any cc bits from res_func
*
* Return ETEMP to extended format from internal extended format so
* that gen_except will have a correctly signed value for ovfl/unfl
* handlers.
*
	bfclr	ETEMP_SGN(a6){0:8}
	beq.b	mvouti_con
	bset.b	#sign_bit,ETEMP_EX(a6)
mvouti_con:
	rts
*
* li is used to handle a long integer source specifier
*

li:
	moveq.l	#4,d0		;set byte count

	btst.b	#7,STAG(a6)	;check for extended denorm
	bne.w	int_dnrm	;if so, branch

	fmovem.x ETEMP(a6),fp0
	fcmp.d	#:41dfffffffc00000,fp0
* 41dfffffffc00000 in dbl prec = 401d0000fffffffe00000000 in ext prec
	fbge.w	lo_plrg	
	fcmp.d	#:c1e0000000000000,fp0
* c1e0000000000000 in dbl prec = c01e00008000000000000000 in ext prec
	fble.w	lo_nlrg
*
* at this point, the answer is between the largest pos and neg values
*
	move.l	USER_FPCR(a6),d1	;use user's rounding mode
	andi.l	#$30,d1
	fmove.l	d1,fpcr
	fmove.l	fp0,L_SCR1(a6)	;let the 040 perform conversion
	fmove.l fpsr,d1
	or.l	d1,USER_FPSR(a6)	;capture inex2/ainex if set
	bra.w	int_wrt


lo_plrg:
	move.l	#$7fffffff,L_SCR1(a6)	;answer is largest positive int
	fbeq.w	int_wrt			;exact answer
	fcmp.d	#:41dfffffffe00000,fp0
* 41dfffffffe00000 in dbl prec = 401d0000ffffffff00000000 in ext prec
	fbge.w	int_operr		;set operr
	bra.w	int_inx			;set inexact

lo_nlrg:
	move.l	#$80000000,L_SCR1(a6)
	fbeq.w	int_wrt			;exact answer
	fcmp.d	#:c1e0000000100000,fp0
* c1e0000000100000 in dbl prec = c01e00008000000080000000 in ext prec
	fblt.w	int_operr		;set operr
	bra.w	int_inx			;set inexact

*
* wi is used to handle a word integer source specifier
*

wi:
	moveq.l	#2,d0		;set byte count

	btst.b	#7,STAG(a6)	;check for extended denorm
	bne.w	int_dnrm	;branch if so

	fmovem.x ETEMP(a6),fp0
	fcmp.s	#:46fffe00,fp0
* 46fffe00 in sgl prec = 400d0000fffe000000000000 in ext prec
	fbge.w	wo_plrg	
	fcmp.s	#:c7000000,fp0
* c7000000 in sgl prec = c00e00008000000000000000 in ext prec
	fble.w	wo_nlrg

*
* at this point, the answer is between the largest pos and neg values
*
	move.l	USER_FPCR(a6),d1	;use user's rounding mode
	andi.l	#$30,d1
	fmove.l	d1,fpcr
	fmove.w	fp0,L_SCR1(a6)	;let the 040 perform conversion
	fmove.l fpsr,d1
	or.l	d1,USER_FPSR(a6)	;capture inex2/ainex if set
	bra.w	int_wrt

wo_plrg:
	move.w	#$7fff,L_SCR1(a6)	;answer is largest positive int
	fbeq.w	int_wrt			;exact answer
	fcmp.s	#:46ffff00,fp0
* 46ffff00 in sgl prec = 400d0000ffff000000000000 in ext prec
	fbge.w	int_operr		;set operr
	bra.w	int_inx			;set inexact

wo_nlrg:
	move.w	#$8000,L_SCR1(a6)
	fbeq.w	int_wrt			;exact answer
	fcmp.s	#:c7000080,fp0
* c7000080 in sgl prec = c00e00008000800000000000 in ext prec
	fblt.w	int_operr		;set operr
	bra.w	int_inx			;set inexact

*
* bi is used to handle a byte integer source specifier
*

bi:
	moveq.l	#1,d0		;set byte count

	btst.b	#7,STAG(a6)	;check for extended denorm
	bne.w	int_dnrm	;branch if so

	fmovem.x ETEMP(a6),fp0
	fcmp.s	#:42fe0000,fp0
* 42fe0000 in sgl prec = 40050000fe00000000000000 in ext prec
	fbge.w	by_plrg	
	fcmp.s	#:c3000000,fp0
* c3000000 in sgl prec = c00600008000000000000000 in ext prec
	fble.w	by_nlrg

*
* at this point, the answer is between the largest pos and neg values
*
	move.l	USER_FPCR(a6),d1	;use user's rounding mode
	andi.l	#$30,d1
	fmove.l	d1,fpcr
	fmove.b	fp0,L_SCR1(a6)	;let the 040 perform conversion
	fmove.l fpsr,d1
	or.l	d1,USER_FPSR(a6)	;capture inex2/ainex if set
	bra.w	int_wrt

by_plrg:
	move.b	#$7f,L_SCR1(a6)		;answer is largest positive int
	fbeq.w	int_wrt			;exact answer
	fcmp.s	#:42ff0000,fp0
* 42ff0000 in sgl prec = 40050000ff00000000000000 in ext prec
	fbge.w	int_operr		;set operr
	bra.w	int_inx			;set inexact

by_nlrg:
	move.b	#$80,L_SCR1(a6)
	fbeq.w	int_wrt			;exact answer
	fcmp.s	#:c3008000,fp0
* c3008000 in sgl prec = c00600008080000000000000 in ext prec
	fblt.w	int_operr		;set operr
	bra.w	int_inx			;set inexact

*
* Common integer routines
*
* int_drnrm---account for possible nonzero result for round up with positive
* operand and round down for negative answer.  In the first case (result = 1)
* byte-width (store in d0) of result must be honored.  In the second case,
* -1 in L_SCR1(a6) will cover all contingencies (FMOVE.B/W/L out).

int_dnrm:
	clr.l	L_SCR1(a6)	; initialize result to 0
	bfextu	FPCR_MODE(a6){2:2},d1	; d1 is the rounding mode
	cmp.b	#2,d1		
	bmi.b	int_inx		; if RN or RZ, done
	bne.b	int_rp		; if RP, continue below
	tst.w	ETEMP(a6)	; RM: store -1 in L_SCR1 if src is negative
	bpl.b	int_inx		; otherwise result is 0
	move.l	#-1,L_SCR1(a6)
	bra.b	int_inx
int_rp:
	tst.w	ETEMP(a6)	; RP: store +1 of proper width in L_SCR1 if
*				; source is greater than 0
	bmi.b	int_inx		; otherwise, result is 0
	lea	L_SCR1(a6),a1	; a1 is address of L_SCR1
	adda.l	d0,a1		; offset by destination width -1
	suba.l	#1,a1		
	bset.b	#0,(a1)		; set low bit at a1 address
int_inx:
	ori.l	#inx2a_mask,USER_FPSR(a6)
	bra.b	int_wrt
int_operr:
	fmovem.x fp0,FPTEMP(a6)	;FPTEMP must contain the extended
*				;precision source that needs to be
*				;converted to integer this is required
*				;if the operr exception is enabled.
*				;set operr/aiop (no inex2 on int ovfl)

	ori.l	#opaop_mask,USER_FPSR(a6)
*				;fall through to perform int_wrt
int_wrt: 
	move.l	EXC_EA(a6),a1	;load destination address
	tst.l	a1		;check to see if it is a dest register
	beq.b	wrt_dn		;write data register 
	lea	L_SCR1(a6),a0	;point to supervisor source address
	bsr.l	mem_write
	bra.w	mvouti_end

wrt_dn:
	move.l	d0,-(sp)	;d0 currently contains the size to write
	bsr.l	get_fline	;get_fline returns Dn in d0
	andi.w	#$7,d0		;isolate register
	move.l	(sp)+,d1	;get size
	cmpi.l	#4,d1		;most frequent case
	beq.b	sz_long
	cmpi.l	#2,d1
	bne.b	sz_con
	or.l	#8,d0		;add 'word' size to register#
	bra.b	sz_con
sz_long:
	or.l	#$10,d0		;add 'long' size to register#
sz_con:
	move.l	d0,d1		;reg_dest expects size:reg in d1
	bsr.l	reg_dest	;load proper data register
	bra.w	mvouti_end 
xp:
	lea	ETEMP(a6),a0
	bclr.b	#sign_bit,LOCAL_EX(a0)
	sne	LOCAL_SGN(a0)
	btst.b	#7,STAG(a6)	;check for extended denorm
	bne.w	xdnrm
	clr.l	d0
	bra.b	do_fp		;do normal case
sgp:
	lea	ETEMP(a6),a0
	bclr.b	#sign_bit,LOCAL_EX(a0)
	sne	LOCAL_SGN(a0)
	btst.b	#7,STAG(a6)	;check for extended denorm
	bne.w	sp_catas	;branch if so
	move.w	LOCAL_EX(a0),d0
	lea	sp_bnds,a1
	cmp.w	(a1),d0
	blt.w	sp_under
	cmp.w	2(a1),d0
	bgt.w	sp_over
	move.l	#1,d0		;set destination format to single
	bra.b	do_fp		;do normal case
dp:
	lea	ETEMP(a6),a0
	bclr.b	#sign_bit,LOCAL_EX(a0)
	sne	LOCAL_SGN(a0)

	btst.b	#7,STAG(a6)	;check for extended denorm
	bne.w	dp_catas	;branch if so

	move.w	LOCAL_EX(a0),d0
	lea	dp_bnds,a1

	cmp.w	(a1),d0
	blt.w	dp_under
	cmp.w	2(a1),d0
	bgt.w	dp_over
	
	move.l	#2,d0		;set destination format to double
*				;fall through to do_fp
*
do_fp:
	bfextu	FPCR_MODE(a6){2:2},d1	;rnd mode in d1
	swap	d0			;rnd prec in upper word
	add.l	d0,d1			;d1 has PREC/MODE info
	
	clr.l	d0			;clear g,r,s 

	bsr.l	round			;round 

	move.l	a0,a1
	move.l	EXC_EA(a6),a0

	bfextu	CMDREG1B(a6){3:3},d1	;extract destination format
*					;at this point only the dest
*					;formats sgl, dbl, ext are
*					;possible
	cmp.b	#2,d1
	bgt.b	ddbl			;double=5, extended=2, single=1
	bne.b	dsgl
*					;fall through to dext
dext:
	bsr.l	dest_ext
	bra.w	mvout_end
dsgl:
	bsr.l	dest_sgl
	bra.w	mvout_end
ddbl:
	bsr.l	dest_dbl
	bra.w	mvout_end

*
* Handle possible denorm or catastrophic underflow cases here
*
xdnrm:
	bsr.w	set_xop		;initialize WBTEMP
	bset.b	#wbtemp15_bit,WB_BYTE(a6) ;set wbtemp15

	move.l	a0,a1
	move.l	EXC_EA(a6),a0	;a0 has the destination pointer
	bsr.l	dest_ext	;store to memory
	bset.b	#unfl_bit,FPSR_EXCEPT(a6)
	bra.w	mvout_end
	
sp_under:
	bset.b	#etemp15_bit,STAG(a6)

	cmp.w	4(a1),d0
	blt.b	sp_catas	;catastrophic underflow case	

	move.l	#1,d0		;load in round precision
	move.l	#sgl_thresh,d1	;load in single denorm threshold
	bsr.l	dpspdnrm	;expects d1 to have the proper
*				;denorm threshold
	bsr.l	dest_sgl	;stores value to destination
	bset.b	#unfl_bit,FPSR_EXCEPT(a6)
	bra.w	mvout_end	;exit

dp_under:
	bset.b	#etemp15_bit,STAG(a6)

	cmp.w	4(a1),d0
	blt.b	dp_catas	;catastrophic underflow case
		
	move.l	#dbl_thresh,d1	;load in double precision threshold
	move.l	#2,d0		
	bsr.l	dpspdnrm	;expects d1 to have proper
*				;denorm threshold
*				;expects d0 to have round precision
	bsr.l	dest_dbl	;store value to destination
	bset.b	#unfl_bit,FPSR_EXCEPT(a6)
	bra.w	mvout_end	;exit

*
* Handle catastrophic underflow cases here
*
sp_catas:
* Temp fix for z bit set in unf_sub
	move.l	USER_FPSR(a6),-(a7)

	move.l	#1,d0		;set round precision to sgl

	bsr.l	unf_sub		;a0 points to result

	move.l	(a7)+,USER_FPSR(a6)

	move.l	#1,d0
	sub.w	d0,LOCAL_EX(a0) ;account for difference between
*				;denorm/norm bias

	move.l	a0,a1		;a1 has the operand input
	move.l	EXC_EA(a6),a0	;a0 has the destination pointer
	
	bsr.l	dest_sgl	;store the result
	ori.l	#unfinx_mask,USER_FPSR(a6)
	bra.w	mvout_end
	
dp_catas:
* Temp fix for z bit set in unf_sub
	move.l	USER_FPSR(a6),-(a7)

	move.l	#2,d0		;set round precision to dbl
	bsr.l	unf_sub		;a0 points to result

	move.l	(a7)+,USER_FPSR(a6)

	move.l	#1,d0
	sub.w	d0,LOCAL_EX(a0) ;account for difference between 
*				;denorm/norm bias

	move.l	a0,a1		;a1 has the operand input
	move.l	EXC_EA(a6),a0	;a0 has the destination pointer
	
	bsr.l	dest_dbl	;store the result
	ori.l	#unfinx_mask,USER_FPSR(a6)
	bra.w	mvout_end

*
* Handle catastrophic overflow cases here
*
sp_over:
* Temp fix for z bit set in unf_sub
	move.l	USER_FPSR(a6),-(a7)

	move.l	#1,d0
	lea.l	FP_SCR1(a6),a0	;use FP_SCR1 for creating result
	move.l	ETEMP_EX(a6),(a0)
	move.l	ETEMP_HI(a6),4(a0)
	move.l	ETEMP_LO(a6),8(a0)
	bsr.l	ovf_res

	move.l	(a7)+,USER_FPSR(a6)

	move.l	a0,a1
	move.l	EXC_EA(a6),a0
	bsr.l	dest_sgl
	or.l	#ovfinx_mask,USER_FPSR(a6)
	bra.w	mvout_end

dp_over:
* Temp fix for z bit set in ovf_res
	move.l	USER_FPSR(a6),-(a7)

	move.l	#2,d0
	lea.l	FP_SCR1(a6),a0	;use FP_SCR1 for creating result
	move.l	ETEMP_EX(a6),(a0)
	move.l	ETEMP_HI(a6),4(a0)
	move.l	ETEMP_LO(a6),8(a0)
	bsr.l	ovf_res

	move.l	(a7)+,USER_FPSR(a6)

	move.l	a0,a1
	move.l	EXC_EA(a6),a0
	bsr.l	dest_dbl
	or.l	#ovfinx_mask,USER_FPSR(a6)
	bra.w	mvout_end

*
* 	DPSPDNRM
*
* This subroutine takes an extended normalized number and denormalizes
* it to the given round precision. This subroutine also decrements
* the input operand's exponent by 1 to account for the fact that
* dest_sgl or dest_dbl expects a normalized number's bias.
*
* Input: a0  points to a normalized number in internal extended format
*	 d0  is the round precision (=1 for sgl; =2 for dbl)
*	 d1  is the single precision or double precision
*	     denorm threshold
*
* Output: (In the format for dest_sgl or dest_dbl)
*	 a0   points to the destination
*   	 a1   points to the operand
*
* Exceptions: Reports inexact 2 exception by setting USER_FPSR bits
*
dpspdnrm:
	move.l	d0,-(a7)	;save round precision
	clr.l	d0		;clear initial g,r,s
	bsr.l	dnrm_lp		;careful with d0, it's needed by round

	bfextu	FPCR_MODE(a6){2:2},d1 ;get rounding mode
	swap	d1
	move.w	2(a7),d1	;set rounding precision 
	swap	d1		;at this point d1 has PREC/MODE info
	bsr.l	round		;round result, sets the inex bit in
*				;USER_FPSR if needed

	move.w	#1,d0
	sub.w	d0,LOCAL_EX(a0) ;account for difference in denorm
*				;vs norm bias

	move.l	a0,a1		;a1 has the operand input
	move.l	EXC_EA(a6),a0	;a0 has the destination pointer
	addq.l	#4,a7		;pop stack
	rts
*
* SET_XOP initialized WBTEMP with the value pointed to by a0
* input: a0 points to input operand in the internal extended format
*
set_xop:
	move.l	LOCAL_EX(a0),WBTEMP_EX(a6)
	move.l	LOCAL_HI(a0),WBTEMP_HI(a6)
	move.l	LOCAL_LO(a0),WBTEMP_LO(a6)
	bfclr	WBTEMP_SGN(a6){0:8}
	beq.b	sxop
	bset.b	#sign_bit,WBTEMP_EX(a6)
sxop:
	bfclr	STAG(a6){5:4}	;clear wbtm66,wbtm1,wbtm0,sbit
	rts
*
*	P_MOVE
*
p_movet:
	dc.l	p_move
	dc.l	p_movez
	dc.l	p_movei
	dc.l	p_moven
	dc.l	p_move
p_regd:
	dc.l	p_dyd0
	dc.l	p_dyd1
	dc.l	p_dyd2
	dc.l	p_dyd3
	dc.l	p_dyd4
	dc.l	p_dyd5
	dc.l	p_dyd6
	dc.l	p_dyd7

pack_out:
 	lea.l	p_movet,a0	;load jmp table address
	move.w	STAG(a6),d0	;get source tag
	bfextu	d0{16:3},d0	;isolate source bits
	move.l	(a0,d0.w*4),a0	;load a0 with routine label for tag
	jmp	(a0)		;go to the routine

p_write:
	move.l	#$0c,d0 	;get byte count
	move.l	EXC_EA(a6),a1	;get the destination address
	bsr 	mem_write	;write the user's destination
	clr.b	CU_SAVEPC(a6) ;set the cu save pc to all 0's

*
* Also note that the dtag must be set to norm here - this is because 
* the 040 uses the dtag to execute the correct microcode.
*
        bfclr    DTAG(a6){0:3}  ;set dtag to norm

	rts

* Notes on handling of special case (zero, inf, and nan) inputs:
*	1. Operr is not signalled if the k-factor is greater than 18.
*	2. Per the manual, status bits are not set.
*

p_move:
	move.w	CMDREG1B(a6),d0
	btst.l	#kfact_bit,d0	;test for dynamic k-factor
	beq.b	statick		;if clear, k-factor is static
dynamick:
	bfextu	d0{25:3},d0	;isolate register for dynamic k-factor
	lea	p_regd,a0
	move.l	(a0,d0*4),a0
	jmp	(a0)
statick:
	andi.w	#$007f,d0	;get k-factor
	bfexts	d0{25:7},d0	;sign extend d0 for bindec
	lea.l	ETEMP(a6),a0	;a0 will point to the packed decimal
	bsr.l	bindec		;perform the convert; data at a6
	lea.l	FP_SCR1(a6),a0	;load a0 with result address
	bra.l	p_write
p_movez:
	lea.l	ETEMP(a6),a0	;a0 will point to the packed decimal
	clr.w	2(a0)		;clear lower word of exp
	clr.l	4(a0)		;load second lword of ZERO
	clr.l	8(a0)		;load third lword of ZERO
	bra.w	p_write		;go write results
p_movei:
	fmove.l	#0,FPSR		;clear aiop
	lea.l	ETEMP(a6),a0	;a0 will point to the packed decimal
	clr.w	2(a0)		;clear lower word of exp
	bra.w	p_write		;go write the result
p_moven:
	lea.l	ETEMP(a6),a0	;a0 will point to the packed decimal
	clr.w	2(a0)		;clear lower word of exp
	bra.w	p_write		;go write the result

*
* Routines to read the dynamic k-factor from Dn.
*
p_dyd0:
	move.l	USER_D0(a6),d0
	bra.b	statick
p_dyd1:
	move.l	USER_D1(a6),d0
	bra.b	statick
p_dyd2:
	move.l	d2,d0
	bra.b	statick
p_dyd3:
	move.l	d3,d0
	bra.b	statick
p_dyd4:
	move.l	d4,d0
	bra.b	statick
p_dyd5:
	move.l	d5,d0
	bra.b	statick
p_dyd6:
	move.l	d6,d0
	bra.w	statick
p_dyd7:
	move.l	d7,d0
	bra.w	statick

	end