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
|
/*-
* Copyright (c) 1993 The Regents of the University of California.
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
* 1. Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* 2. Redistributions in binary form must reproduce the above copyright
* notice, this list of conditions and the following disclaimer in the
* documentation and/or other materials provided with the distribution.
* 3. All advertising materials mentioning features or use of this software
* must display the following acknowledgement:
* This product includes software developed by the University of
* California, Berkeley and its contributors.
* 4. Neither the name of the University nor the names of its contributors
* may be used to endorse or promote products derived from this software
* without specific prior written permission.
*
* THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
* ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
* ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
* FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
* DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
* OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
* HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
* LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
* OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
* SUCH DAMAGE.
*
* $FreeBSD$
*/
#include "opt_npx.h"
#include <machine/asmacros.h>
#include <machine/cputypes.h>
#include <machine/pmap.h>
#include <machine/specialreg.h>
#include "assym.s"
#define IDXSHIFT 10
.data
.globl bcopy_vector
bcopy_vector:
.long generic_bcopy
.globl bzero
bzero:
.long generic_bzero
.globl copyin_vector
copyin_vector:
.long generic_copyin
.globl copyout_vector
copyout_vector:
.long generic_copyout
.globl ovbcopy_vector
ovbcopy_vector:
.long generic_bcopy
#if defined(I586_CPU) && defined(DEV_NPX)
kernel_fpu_lock:
.byte 0xfe
.space 3
#endif
.text
/*
* bcopy family
* void bzero(void *buf, u_int len)
*/
ENTRY(generic_bzero)
pushl %edi
movl 8(%esp),%edi
movl 12(%esp),%ecx
xorl %eax,%eax
shrl $2,%ecx
cld
rep
stosl
movl 12(%esp),%ecx
andl $3,%ecx
rep
stosb
popl %edi
ret
#ifdef I486_CPU
ENTRY(i486_bzero)
movl 4(%esp),%edx
movl 8(%esp),%ecx
xorl %eax,%eax
/*
* do 64 byte chunks first
*
* XXX this is probably over-unrolled at least for DX2's
*/
2:
cmpl $64,%ecx
jb 3f
movl %eax,(%edx)
movl %eax,4(%edx)
movl %eax,8(%edx)
movl %eax,12(%edx)
movl %eax,16(%edx)
movl %eax,20(%edx)
movl %eax,24(%edx)
movl %eax,28(%edx)
movl %eax,32(%edx)
movl %eax,36(%edx)
movl %eax,40(%edx)
movl %eax,44(%edx)
movl %eax,48(%edx)
movl %eax,52(%edx)
movl %eax,56(%edx)
movl %eax,60(%edx)
addl $64,%edx
subl $64,%ecx
jnz 2b
ret
/*
* do 16 byte chunks
*/
SUPERALIGN_TEXT
3:
cmpl $16,%ecx
jb 4f
movl %eax,(%edx)
movl %eax,4(%edx)
movl %eax,8(%edx)
movl %eax,12(%edx)
addl $16,%edx
subl $16,%ecx
jnz 3b
ret
/*
* do 4 byte chunks
*/
SUPERALIGN_TEXT
4:
cmpl $4,%ecx
jb 5f
movl %eax,(%edx)
addl $4,%edx
subl $4,%ecx
jnz 4b
ret
/*
* do 1 byte chunks
* a jump table seems to be faster than a loop or more range reductions
*
* XXX need a const section for non-text
*/
.data
jtab:
.long do0
.long do1
.long do2
.long do3
.text
SUPERALIGN_TEXT
5:
jmp *jtab(,%ecx,4)
SUPERALIGN_TEXT
do3:
movw %ax,(%edx)
movb %al,2(%edx)
ret
SUPERALIGN_TEXT
do2:
movw %ax,(%edx)
ret
SUPERALIGN_TEXT
do1:
movb %al,(%edx)
ret
SUPERALIGN_TEXT
do0:
ret
#endif
#if defined(I586_CPU) && defined(DEV_NPX)
ENTRY(i586_bzero)
movl 4(%esp),%edx
movl 8(%esp),%ecx
/*
* The FPU register method is twice as fast as the integer register
* method unless the target is in the L1 cache and we pre-allocate a
* cache line for it (then the integer register method is 4-5 times
* faster). However, we never pre-allocate cache lines, since that
* would make the integer method 25% or more slower for the common
* case when the target isn't in either the L1 cache or the L2 cache.
* Thus we normally use the FPU register method unless the overhead
* would be too large.
*/
cmpl $256,%ecx /* empirical; clts, fninit, smsw cost a lot */
jb intreg_i586_bzero
/*
* The FPU registers may belong to an application or to fastmove()
* or to another invocation of bcopy() or ourself in a higher level
* interrupt or trap handler. Preserving the registers is
* complicated since we avoid it if possible at all levels. We
* want to localize the complications even when that increases them.
* Here the extra work involves preserving CR0_TS in TS.
* `fpcurthread != NULL' is supposed to be the condition that all the
* FPU resources belong to an application, but fpcurthread and CR0_TS
* aren't set atomically enough for this condition to work in
* interrupt handlers.
*
* Case 1: FPU registers belong to the application: we must preserve
* the registers if we use them, so we only use the FPU register
* method if the target size is large enough to amortize the extra
* overhead for preserving them. CR0_TS must be preserved although
* it is very likely to end up as set.
*
* Case 2: FPU registers belong to fastmove(): fastmove() currently
* makes the registers look like they belong to an application so
* that cpu_switch() and savectx() don't have to know about it, so
* this case reduces to case 1.
*
* Case 3: FPU registers belong to the kernel: don't use the FPU
* register method. This case is unlikely, and supporting it would
* be more complicated and might take too much stack.
*
* Case 4: FPU registers don't belong to anyone: the FPU registers
* don't need to be preserved, so we always use the FPU register
* method. CR0_TS must be preserved although it is very likely to
* always end up as clear.
*/
cmpl $0,PCPU(FPCURTHREAD)
je i586_bz1
/*
* XXX don't use the FPU for cases 1 and 2, since preemptive
* scheduling of ithreads broke these cases. Note that we can
* no longer get here from an interrupt handler, since the
* context sitch to the interrupt handler will have saved the
* FPU state.
*/
jmp intreg_i586_bzero
cmpl $256+184,%ecx /* empirical; not quite 2*108 more */
jb intreg_i586_bzero
sarb $1,kernel_fpu_lock
jc intreg_i586_bzero
smsw %ax
clts
subl $108,%esp
fnsave 0(%esp)
jmp i586_bz2
i586_bz1:
sarb $1,kernel_fpu_lock
jc intreg_i586_bzero
smsw %ax
clts
fninit /* XXX should avoid needing this */
i586_bz2:
fldz
/*
* Align to an 8 byte boundary (misalignment in the main loop would
* cost a factor of >= 2). Avoid jumps (at little cost if it is
* already aligned) by always zeroing 8 bytes and using the part up
* to the _next_ alignment position.
*/
fstl 0(%edx)
addl %edx,%ecx /* part of %ecx -= new_%edx - %edx */
addl $8,%edx
andl $~7,%edx
subl %edx,%ecx
/*
* Similarly align `len' to a multiple of 8.
*/
fstl -8(%edx,%ecx)
decl %ecx
andl $~7,%ecx
/*
* This wouldn't be any faster if it were unrolled, since the loop
* control instructions are much faster than the fstl and/or done
* in parallel with it so their overhead is insignificant.
*/
fpureg_i586_bzero_loop:
fstl 0(%edx)
addl $8,%edx
subl $8,%ecx
cmpl $8,%ecx
jae fpureg_i586_bzero_loop
cmpl $0,PCPU(FPCURTHREAD)
je i586_bz3
/* XXX check that the condition for cases 1-2 stayed false. */
i586_bzero_oops:
int $3
jmp i586_bzero_oops
frstor 0(%esp)
addl $108,%esp
lmsw %ax
movb $0xfe,kernel_fpu_lock
ret
i586_bz3:
fstp %st(0)
lmsw %ax
movb $0xfe,kernel_fpu_lock
ret
intreg_i586_bzero:
/*
* `rep stos' seems to be the best method in practice for small
* counts. Fancy methods usually take too long to start up due
* to cache and BTB misses.
*/
pushl %edi
movl %edx,%edi
xorl %eax,%eax
shrl $2,%ecx
cld
rep
stosl
movl 12(%esp),%ecx
andl $3,%ecx
jne 1f
popl %edi
ret
1:
rep
stosb
popl %edi
ret
#endif /* I586_CPU && defined(DEV_NPX) */
ENTRY(i686_pagezero)
pushl %edi
pushl %ebx
movl 12(%esp), %edi
movl $1024, %ecx
cld
ALIGN_TEXT
1:
xorl %eax, %eax
repe
scasl
jnz 2f
popl %ebx
popl %edi
ret
ALIGN_TEXT
2:
incl %ecx
subl $4, %edi
movl %ecx, %edx
cmpl $16, %ecx
jge 3f
movl %edi, %ebx
andl $0x3f, %ebx
shrl %ebx
shrl %ebx
movl $16, %ecx
subl %ebx, %ecx
3:
subl %ecx, %edx
rep
stosl
movl %edx, %ecx
testl %edx, %edx
jnz 1b
popl %ebx
popl %edi
ret
/* fillw(pat, base, cnt) */
ENTRY(fillw)
pushl %edi
movl 8(%esp),%eax
movl 12(%esp),%edi
movl 16(%esp),%ecx
cld
rep
stosw
popl %edi
ret
ENTRY(bcopyb)
pushl %esi
pushl %edi
movl 12(%esp),%esi
movl 16(%esp),%edi
movl 20(%esp),%ecx
movl %edi,%eax
subl %esi,%eax
cmpl %ecx,%eax /* overlapping && src < dst? */
jb 1f
cld /* nope, copy forwards */
rep
movsb
popl %edi
popl %esi
ret
ALIGN_TEXT
1:
addl %ecx,%edi /* copy backwards. */
addl %ecx,%esi
decl %edi
decl %esi
std
rep
movsb
popl %edi
popl %esi
cld
ret
ENTRY(bcopy)
MEXITCOUNT
jmp *bcopy_vector
ENTRY(ovbcopy)
MEXITCOUNT
jmp *ovbcopy_vector
/*
* generic_bcopy(src, dst, cnt)
* ws@tools.de (Wolfgang Solfrank, TooLs GmbH) +49-228-985800
*/
ENTRY(generic_bcopy)
pushl %esi
pushl %edi
movl 12(%esp),%esi
movl 16(%esp),%edi
movl 20(%esp),%ecx
movl %edi,%eax
subl %esi,%eax
cmpl %ecx,%eax /* overlapping && src < dst? */
jb 1f
shrl $2,%ecx /* copy by 32-bit words */
cld /* nope, copy forwards */
rep
movsl
movl 20(%esp),%ecx
andl $3,%ecx /* any bytes left? */
rep
movsb
popl %edi
popl %esi
ret
ALIGN_TEXT
1:
addl %ecx,%edi /* copy backwards */
addl %ecx,%esi
decl %edi
decl %esi
andl $3,%ecx /* any fractional bytes? */
std
rep
movsb
movl 20(%esp),%ecx /* copy remainder by 32-bit words */
shrl $2,%ecx
subl $3,%esi
subl $3,%edi
rep
movsl
popl %edi
popl %esi
cld
ret
#if defined(I586_CPU) && defined(DEV_NPX)
ENTRY(i586_bcopy)
pushl %esi
pushl %edi
movl 12(%esp),%esi
movl 16(%esp),%edi
movl 20(%esp),%ecx
movl %edi,%eax
subl %esi,%eax
cmpl %ecx,%eax /* overlapping && src < dst? */
jb 1f
cmpl $1024,%ecx
jb small_i586_bcopy
sarb $1,kernel_fpu_lock
jc small_i586_bcopy
cmpl $0,PCPU(FPCURTHREAD)
je i586_bc1
/* XXX turn off handling of cases 1-2, as above. */
movb $0xfe,kernel_fpu_lock
jmp small_i586_bcopy
smsw %dx
clts
subl $108,%esp
fnsave 0(%esp)
jmp 4f
i586_bc1:
smsw %dx
clts
fninit /* XXX should avoid needing this */
ALIGN_TEXT
4:
pushl %ecx
#define DCACHE_SIZE 8192
cmpl $(DCACHE_SIZE-512)/2,%ecx
jbe 2f
movl $(DCACHE_SIZE-512)/2,%ecx
2:
subl %ecx,0(%esp)
cmpl $256,%ecx
jb 5f /* XXX should prefetch if %ecx >= 32 */
pushl %esi
pushl %ecx
ALIGN_TEXT
3:
movl 0(%esi),%eax
movl 32(%esi),%eax
movl 64(%esi),%eax
movl 96(%esi),%eax
movl 128(%esi),%eax
movl 160(%esi),%eax
movl 192(%esi),%eax
movl 224(%esi),%eax
addl $256,%esi
subl $256,%ecx
cmpl $256,%ecx
jae 3b
popl %ecx
popl %esi
5:
ALIGN_TEXT
large_i586_bcopy_loop:
fildq 0(%esi)
fildq 8(%esi)
fildq 16(%esi)
fildq 24(%esi)
fildq 32(%esi)
fildq 40(%esi)
fildq 48(%esi)
fildq 56(%esi)
fistpq 56(%edi)
fistpq 48(%edi)
fistpq 40(%edi)
fistpq 32(%edi)
fistpq 24(%edi)
fistpq 16(%edi)
fistpq 8(%edi)
fistpq 0(%edi)
addl $64,%esi
addl $64,%edi
subl $64,%ecx
cmpl $64,%ecx
jae large_i586_bcopy_loop
popl %eax
addl %eax,%ecx
cmpl $64,%ecx
jae 4b
cmpl $0,PCPU(FPCURTHREAD)
je i586_bc2
/* XXX check that the condition for cases 1-2 stayed false. */
i586_bcopy_oops:
int $3
jmp i586_bcopy_oops
frstor 0(%esp)
addl $108,%esp
i586_bc2:
lmsw %dx
movb $0xfe,kernel_fpu_lock
/*
* This is a duplicate of the main part of generic_bcopy. See the comments
* there. Jumping into generic_bcopy would cost a whole 0-1 cycles and
* would mess up high resolution profiling.
*/
ALIGN_TEXT
small_i586_bcopy:
shrl $2,%ecx
cld
rep
movsl
movl 20(%esp),%ecx
andl $3,%ecx
rep
movsb
popl %edi
popl %esi
ret
ALIGN_TEXT
1:
addl %ecx,%edi
addl %ecx,%esi
decl %edi
decl %esi
andl $3,%ecx
std
rep
movsb
movl 20(%esp),%ecx
shrl $2,%ecx
subl $3,%esi
subl $3,%edi
rep
movsl
popl %edi
popl %esi
cld
ret
#endif /* I586_CPU && defined(DEV_NPX) */
/*
* Note: memcpy does not support overlapping copies
*/
ENTRY(memcpy)
pushl %edi
pushl %esi
movl 12(%esp),%edi
movl 16(%esp),%esi
movl 20(%esp),%ecx
movl %edi,%eax
shrl $2,%ecx /* copy by 32-bit words */
cld /* nope, copy forwards */
rep
movsl
movl 20(%esp),%ecx
andl $3,%ecx /* any bytes left? */
rep
movsb
popl %esi
popl %edi
ret
/*****************************************************************************/
/* copyout and fubyte family */
/*****************************************************************************/
/*
* Access user memory from inside the kernel. These routines and possibly
* the math- and DOS emulators should be the only places that do this.
*
* We have to access the memory with user's permissions, so use a segment
* selector with RPL 3. For writes to user space we have to additionally
* check the PTE for write permission, because the 386 does not check
* write permissions when we are executing with EPL 0. The 486 does check
* this if the WP bit is set in CR0, so we can use a simpler version here.
*
* These routines set curpcb->onfault for the time they execute. When a
* protection violation occurs inside the functions, the trap handler
* returns to *curpcb->onfault instead of the function.
*/
/*
* copyout(from_kernel, to_user, len) - MP SAFE (if not I386_CPU)
*/
ENTRY(copyout)
MEXITCOUNT
jmp *copyout_vector
ENTRY(generic_copyout)
movl PCPU(CURPCB),%eax
movl $copyout_fault,PCB_ONFAULT(%eax)
pushl %esi
pushl %edi
pushl %ebx
movl 16(%esp),%esi
movl 20(%esp),%edi
movl 24(%esp),%ebx
testl %ebx,%ebx /* anything to do? */
jz done_copyout
/*
* Check explicitly for non-user addresses. If 486 write protection
* is being used, this check is essential because we are in kernel
* mode so the h/w does not provide any protection against writing
* kernel addresses.
*/
/*
* First, prevent address wrapping.
*/
movl %edi,%eax
addl %ebx,%eax
jc copyout_fault
/*
* XXX STOP USING VM_MAXUSER_ADDRESS.
* It is an end address, not a max, so every time it is used correctly it
* looks like there is an off by one error, and of course it caused an off
* by one error in several places.
*/
cmpl $VM_MAXUSER_ADDRESS,%eax
ja copyout_fault
#ifdef I386_CPU
/*
* We have to check each PTE for user write permission.
* The checking may cause a page fault, so it is important to set
* up everything for return via copyout_fault before here.
*/
/* compute number of pages */
movl %edi,%ecx
andl $PAGE_MASK,%ecx
addl %ebx,%ecx
decl %ecx
shrl $IDXSHIFT+2,%ecx
incl %ecx
/* compute PTE offset for start address */
movl %edi,%edx
shrl $IDXSHIFT,%edx
andb $0xfc,%dl
1:
/* check PTE for each page */
leal PTmap(%edx),%eax
shrl $IDXSHIFT,%eax
andb $0xfc,%al
testb $PG_V,PTmap(%eax) /* PTE page must be valid */
je 4f
movb PTmap(%edx),%al
andb $PG_V|PG_RW|PG_U,%al /* page must be valid and user writable */
cmpb $PG_V|PG_RW|PG_U,%al
je 2f
4:
/* simulate a trap */
pushl %edx
pushl %ecx
shll $IDXSHIFT,%edx
pushl %edx
call trapwrite /* trapwrite(addr) */
popl %edx
popl %ecx
popl %edx
testl %eax,%eax /* if not ok, return EFAULT */
jnz copyout_fault
2:
addl $4,%edx
decl %ecx
jnz 1b /* check next page */
#endif /* I386_CPU */
/* bcopy(%esi, %edi, %ebx) */
movl %ebx,%ecx
#if defined(I586_CPU) && defined(DEV_NPX)
ALIGN_TEXT
slow_copyout:
#endif
shrl $2,%ecx
cld
rep
movsl
movb %bl,%cl
andb $3,%cl
rep
movsb
done_copyout:
popl %ebx
popl %edi
popl %esi
xorl %eax,%eax
movl PCPU(CURPCB),%edx
movl %eax,PCB_ONFAULT(%edx)
ret
ALIGN_TEXT
copyout_fault:
popl %ebx
popl %edi
popl %esi
movl PCPU(CURPCB),%edx
movl $0,PCB_ONFAULT(%edx)
movl $EFAULT,%eax
ret
#if defined(I586_CPU) && defined(DEV_NPX)
ENTRY(i586_copyout)
/*
* Duplicated from generic_copyout. Could be done a bit better.
*/
movl PCPU(CURPCB),%eax
movl $copyout_fault,PCB_ONFAULT(%eax)
pushl %esi
pushl %edi
pushl %ebx
movl 16(%esp),%esi
movl 20(%esp),%edi
movl 24(%esp),%ebx
testl %ebx,%ebx /* anything to do? */
jz done_copyout
/*
* Check explicitly for non-user addresses. If 486 write protection
* is being used, this check is essential because we are in kernel
* mode so the h/w does not provide any protection against writing
* kernel addresses.
*/
/*
* First, prevent address wrapping.
*/
movl %edi,%eax
addl %ebx,%eax
jc copyout_fault
/*
* XXX STOP USING VM_MAXUSER_ADDRESS.
* It is an end address, not a max, so every time it is used correctly it
* looks like there is an off by one error, and of course it caused an off
* by one error in several places.
*/
cmpl $VM_MAXUSER_ADDRESS,%eax
ja copyout_fault
/* bcopy(%esi, %edi, %ebx) */
3:
movl %ebx,%ecx
/*
* End of duplicated code.
*/
cmpl $1024,%ecx
jb slow_copyout
pushl %ecx
call fastmove
addl $4,%esp
jmp done_copyout
#endif /* I586_CPU && defined(DEV_NPX) */
/*
* copyin(from_user, to_kernel, len) - MP SAFE
*/
ENTRY(copyin)
MEXITCOUNT
jmp *copyin_vector
ENTRY(generic_copyin)
movl PCPU(CURPCB),%eax
movl $copyin_fault,PCB_ONFAULT(%eax)
pushl %esi
pushl %edi
movl 12(%esp),%esi /* caddr_t from */
movl 16(%esp),%edi /* caddr_t to */
movl 20(%esp),%ecx /* size_t len */
/*
* make sure address is valid
*/
movl %esi,%edx
addl %ecx,%edx
jc copyin_fault
cmpl $VM_MAXUSER_ADDRESS,%edx
ja copyin_fault
#if defined(I586_CPU) && defined(DEV_NPX)
ALIGN_TEXT
slow_copyin:
#endif
movb %cl,%al
shrl $2,%ecx /* copy longword-wise */
cld
rep
movsl
movb %al,%cl
andb $3,%cl /* copy remaining bytes */
rep
movsb
#if defined(I586_CPU) && defined(DEV_NPX)
ALIGN_TEXT
done_copyin:
#endif
popl %edi
popl %esi
xorl %eax,%eax
movl PCPU(CURPCB),%edx
movl %eax,PCB_ONFAULT(%edx)
ret
ALIGN_TEXT
copyin_fault:
popl %edi
popl %esi
movl PCPU(CURPCB),%edx
movl $0,PCB_ONFAULT(%edx)
movl $EFAULT,%eax
ret
#if defined(I586_CPU) && defined(DEV_NPX)
ENTRY(i586_copyin)
/*
* Duplicated from generic_copyin. Could be done a bit better.
*/
movl PCPU(CURPCB),%eax
movl $copyin_fault,PCB_ONFAULT(%eax)
pushl %esi
pushl %edi
movl 12(%esp),%esi /* caddr_t from */
movl 16(%esp),%edi /* caddr_t to */
movl 20(%esp),%ecx /* size_t len */
/*
* make sure address is valid
*/
movl %esi,%edx
addl %ecx,%edx
jc copyin_fault
cmpl $VM_MAXUSER_ADDRESS,%edx
ja copyin_fault
/*
* End of duplicated code.
*/
cmpl $1024,%ecx
jb slow_copyin
pushl %ebx /* XXX prepare for fastmove_fault */
pushl %ecx
call fastmove
addl $8,%esp
jmp done_copyin
#endif /* I586_CPU && defined(DEV_NPX) */
#if defined(I586_CPU) && defined(DEV_NPX)
/* fastmove(src, dst, len)
src in %esi
dst in %edi
len in %ecx XXX changed to on stack for profiling
uses %eax and %edx for tmp. storage
*/
/* XXX use ENTRY() to get profiling. fastmove() is actually a non-entry. */
ENTRY(fastmove)
pushl %ebp
movl %esp,%ebp
subl $PCB_SAVE87_SIZE+3*4,%esp
movl 8(%ebp),%ecx
cmpl $63,%ecx
jbe fastmove_tail
testl $7,%esi /* check if src addr is multiple of 8 */
jnz fastmove_tail
testl $7,%edi /* check if dst addr is multiple of 8 */
jnz fastmove_tail
/* XXX grab FPU context atomically. */
cli
/* if (fpcurthread != NULL) { */
cmpl $0,PCPU(FPCURTHREAD)
je 6f
/* fnsave(&curpcb->pcb_savefpu); */
movl PCPU(CURPCB),%eax
fnsave PCB_SAVEFPU(%eax)
/* FPCURTHREAD = NULL; */
movl $0,PCPU(FPCURTHREAD)
/* } */
6:
/* now we own the FPU. */
/*
* The process' FP state is saved in the pcb, but if we get
* switched, the cpu_switch() will store our FP state in the
* pcb. It should be possible to avoid all the copying for
* this, e.g., by setting a flag to tell cpu_switch() to
* save the state somewhere else.
*/
/* tmp = curpcb->pcb_savefpu; */
movl %ecx,-12(%ebp)
movl %esi,-8(%ebp)
movl %edi,-4(%ebp)
movl %esp,%edi
movl PCPU(CURPCB),%esi
addl $PCB_SAVEFPU,%esi
cld
movl $PCB_SAVE87_SIZE>>2,%ecx
rep
movsl
movl -12(%ebp),%ecx
movl -8(%ebp),%esi
movl -4(%ebp),%edi
/* stop_emulating(); */
clts
/* fpcurthread = curthread; */
movl PCPU(CURTHREAD),%eax
movl %eax,PCPU(FPCURTHREAD)
movl PCPU(CURPCB),%eax
/* XXX end of atomic FPU context grab. */
sti
movl $fastmove_fault,PCB_ONFAULT(%eax)
4:
movl %ecx,-12(%ebp)
cmpl $1792,%ecx
jbe 2f
movl $1792,%ecx
2:
subl %ecx,-12(%ebp)
cmpl $256,%ecx
jb 5f
movl %ecx,-8(%ebp)
movl %esi,-4(%ebp)
ALIGN_TEXT
3:
movl 0(%esi),%eax
movl 32(%esi),%eax
movl 64(%esi),%eax
movl 96(%esi),%eax
movl 128(%esi),%eax
movl 160(%esi),%eax
movl 192(%esi),%eax
movl 224(%esi),%eax
addl $256,%esi
subl $256,%ecx
cmpl $256,%ecx
jae 3b
movl -8(%ebp),%ecx
movl -4(%ebp),%esi
5:
ALIGN_TEXT
fastmove_loop:
fildq 0(%esi)
fildq 8(%esi)
fildq 16(%esi)
fildq 24(%esi)
fildq 32(%esi)
fildq 40(%esi)
fildq 48(%esi)
fildq 56(%esi)
fistpq 56(%edi)
fistpq 48(%edi)
fistpq 40(%edi)
fistpq 32(%edi)
fistpq 24(%edi)
fistpq 16(%edi)
fistpq 8(%edi)
fistpq 0(%edi)
addl $-64,%ecx
addl $64,%esi
addl $64,%edi
cmpl $63,%ecx
ja fastmove_loop
movl -12(%ebp),%eax
addl %eax,%ecx
cmpl $64,%ecx
jae 4b
/* XXX ungrab FPU context atomically. */
cli
/* curpcb->pcb_savefpu = tmp; */
movl %ecx,-12(%ebp)
movl %esi,-8(%ebp)
movl %edi,-4(%ebp)
movl PCPU(CURPCB),%edi
addl $PCB_SAVEFPU,%edi
movl %esp,%esi
cld
movl $PCB_SAVE87_SIZE>>2,%ecx
rep
movsl
movl -12(%ebp),%ecx
movl -8(%ebp),%esi
movl -4(%ebp),%edi
/* start_emulating(); */
smsw %ax
orb $CR0_TS,%al
lmsw %ax
/* fpcurthread = NULL; */
movl $0,PCPU(FPCURTHREAD)
/* XXX end of atomic FPU context ungrab. */
sti
ALIGN_TEXT
fastmove_tail:
movl PCPU(CURPCB),%eax
movl $fastmove_tail_fault,PCB_ONFAULT(%eax)
movb %cl,%al
shrl $2,%ecx /* copy longword-wise */
cld
rep
movsl
movb %al,%cl
andb $3,%cl /* copy remaining bytes */
rep
movsb
movl %ebp,%esp
popl %ebp
ret
ALIGN_TEXT
fastmove_fault:
/* XXX ungrab FPU context atomically. */
cli
movl PCPU(CURPCB),%edi
addl $PCB_SAVEFPU,%edi
movl %esp,%esi
cld
movl $PCB_SAVE87_SIZE>>2,%ecx
rep
movsl
smsw %ax
orb $CR0_TS,%al
lmsw %ax
movl $0,PCPU(FPCURTHREAD)
/* XXX end of atomic FPU context ungrab. */
sti
fastmove_tail_fault:
movl %ebp,%esp
popl %ebp
addl $8,%esp
popl %ebx
popl %edi
popl %esi
movl PCPU(CURPCB),%edx
movl $0,PCB_ONFAULT(%edx)
movl $EFAULT,%eax
ret
#endif /* I586_CPU && defined(DEV_NPX) */
/*
* fu{byte,sword,word} - MP SAFE
*
* Fetch a byte (sword, word) from user memory
*/
ENTRY(fuword)
movl PCPU(CURPCB),%ecx
movl $fusufault,PCB_ONFAULT(%ecx)
movl 4(%esp),%edx /* from */
cmpl $VM_MAXUSER_ADDRESS-4,%edx /* verify address is valid */
ja fusufault
movl (%edx),%eax
movl $0,PCB_ONFAULT(%ecx)
ret
ENTRY(fuword32)
jmp fuword
/*
* These two routines are called from the profiling code, potentially
* at interrupt time. If they fail, that's okay, good things will
* happen later. Fail all the time for now - until the trap code is
* able to deal with this.
*/
ALTENTRY(suswintr)
ENTRY(fuswintr)
movl $-1,%eax
ret
/*
* fuword16 - MP SAFE
*/
ENTRY(fuword16)
movl PCPU(CURPCB),%ecx
movl $fusufault,PCB_ONFAULT(%ecx)
movl 4(%esp),%edx
cmpl $VM_MAXUSER_ADDRESS-2,%edx
ja fusufault
movzwl (%edx),%eax
movl $0,PCB_ONFAULT(%ecx)
ret
/*
* fubyte - MP SAFE
*/
ENTRY(fubyte)
movl PCPU(CURPCB),%ecx
movl $fusufault,PCB_ONFAULT(%ecx)
movl 4(%esp),%edx
cmpl $VM_MAXUSER_ADDRESS-1,%edx
ja fusufault
movzbl (%edx),%eax
movl $0,PCB_ONFAULT(%ecx)
ret
ALIGN_TEXT
fusufault:
movl PCPU(CURPCB),%ecx
xorl %eax,%eax
movl %eax,PCB_ONFAULT(%ecx)
decl %eax
ret
/*
* su{byte,sword,word} - MP SAFE (if not I386_CPU)
*
* Write a byte (word, longword) to user memory
*/
ENTRY(suword)
movl PCPU(CURPCB),%ecx
movl $fusufault,PCB_ONFAULT(%ecx)
movl 4(%esp),%edx
#ifdef I386_CPU
/* XXX - page boundary crossing is still not handled */
movl %edx,%eax
shrl $IDXSHIFT,%edx
andb $0xfc,%dl
leal PTmap(%edx),%ecx
shrl $IDXSHIFT,%ecx
andb $0xfc,%cl
testb $PG_V,PTmap(%ecx) /* PTE page must be valid */
je 4f
movb PTmap(%edx),%dl
andb $PG_V|PG_RW|PG_U,%dl /* page must be valid and user writable */
cmpb $PG_V|PG_RW|PG_U,%dl
je 1f
4:
/* simulate a trap */
pushl %eax
call trapwrite
popl %edx /* remove junk parameter from stack */
testl %eax,%eax
jnz fusufault
1:
movl 4(%esp),%edx
#endif
cmpl $VM_MAXUSER_ADDRESS-4,%edx /* verify address validity */
ja fusufault
movl 8(%esp),%eax
movl %eax,(%edx)
xorl %eax,%eax
movl PCPU(CURPCB),%ecx
movl %eax,PCB_ONFAULT(%ecx)
ret
ENTRY(suword32)
jmp suword
/*
* suword16 - MP SAFE (if not I386_CPU)
*/
ENTRY(suword16)
movl PCPU(CURPCB),%ecx
movl $fusufault,PCB_ONFAULT(%ecx)
movl 4(%esp),%edx
#ifdef I386_CPU
/* XXX - page boundary crossing is still not handled */
movl %edx,%eax
shrl $IDXSHIFT,%edx
andb $0xfc,%dl
leal PTmap(%edx),%ecx
shrl $IDXSHIFT,%ecx
andb $0xfc,%cl
testb $PG_V,PTmap(%ecx) /* PTE page must be valid */
je 4f
movb PTmap(%edx),%dl
andb $PG_V|PG_RW|PG_U,%dl /* page must be valid and user writable */
cmpb $PG_V|PG_RW|PG_U,%dl
je 1f
4:
/* simulate a trap */
pushl %eax
call trapwrite
popl %edx /* remove junk parameter from stack */
testl %eax,%eax
jnz fusufault
1:
movl 4(%esp),%edx
#endif
cmpl $VM_MAXUSER_ADDRESS-2,%edx /* verify address validity */
ja fusufault
movw 8(%esp),%ax
movw %ax,(%edx)
xorl %eax,%eax
movl PCPU(CURPCB),%ecx /* restore trashed register */
movl %eax,PCB_ONFAULT(%ecx)
ret
/*
* subyte - MP SAFE (if not I386_CPU)
*/
ENTRY(subyte)
movl PCPU(CURPCB),%ecx
movl $fusufault,PCB_ONFAULT(%ecx)
movl 4(%esp),%edx
#ifdef I386_CPU
movl %edx,%eax
shrl $IDXSHIFT,%edx
andb $0xfc,%dl
leal PTmap(%edx),%ecx
shrl $IDXSHIFT,%ecx
andb $0xfc,%cl
testb $PG_V,PTmap(%ecx) /* PTE page must be valid */
je 4f
movb PTmap(%edx),%dl
andb $PG_V|PG_RW|PG_U,%dl /* page must be valid and user writable */
cmpb $PG_V|PG_RW|PG_U,%dl
je 1f
4:
/* simulate a trap */
pushl %eax
call trapwrite
popl %edx /* remove junk parameter from stack */
testl %eax,%eax
jnz fusufault
1:
movl 4(%esp),%edx
#endif
cmpl $VM_MAXUSER_ADDRESS-1,%edx /* verify address validity */
ja fusufault
movb 8(%esp),%al
movb %al,(%edx)
xorl %eax,%eax
movl PCPU(CURPCB),%ecx /* restore trashed register */
movl %eax,PCB_ONFAULT(%ecx)
ret
/*
* copyinstr(from, to, maxlen, int *lencopied) - MP SAFE
*
* copy a string from from to to, stop when a 0 character is reached.
* return ENAMETOOLONG if string is longer than maxlen, and
* EFAULT on protection violations. If lencopied is non-zero,
* return the actual length in *lencopied.
*/
ENTRY(copyinstr)
pushl %esi
pushl %edi
movl PCPU(CURPCB),%ecx
movl $cpystrflt,PCB_ONFAULT(%ecx)
movl 12(%esp),%esi /* %esi = from */
movl 16(%esp),%edi /* %edi = to */
movl 20(%esp),%edx /* %edx = maxlen */
movl $VM_MAXUSER_ADDRESS,%eax
/* make sure 'from' is within bounds */
subl %esi,%eax
jbe cpystrflt
/* restrict maxlen to <= VM_MAXUSER_ADDRESS-from */
cmpl %edx,%eax
jae 1f
movl %eax,%edx
movl %eax,20(%esp)
1:
incl %edx
cld
2:
decl %edx
jz 3f
lodsb
stosb
orb %al,%al
jnz 2b
/* Success -- 0 byte reached */
decl %edx
xorl %eax,%eax
jmp cpystrflt_x
3:
/* edx is zero - return ENAMETOOLONG or EFAULT */
cmpl $VM_MAXUSER_ADDRESS,%esi
jae cpystrflt
4:
movl $ENAMETOOLONG,%eax
jmp cpystrflt_x
cpystrflt:
movl $EFAULT,%eax
cpystrflt_x:
/* set *lencopied and return %eax */
movl PCPU(CURPCB),%ecx
movl $0,PCB_ONFAULT(%ecx)
movl 20(%esp),%ecx
subl %edx,%ecx
movl 24(%esp),%edx
testl %edx,%edx
jz 1f
movl %ecx,(%edx)
1:
popl %edi
popl %esi
ret
/*
* copystr(from, to, maxlen, int *lencopied) - MP SAFE
*/
ENTRY(copystr)
pushl %esi
pushl %edi
movl 12(%esp),%esi /* %esi = from */
movl 16(%esp),%edi /* %edi = to */
movl 20(%esp),%edx /* %edx = maxlen */
incl %edx
cld
1:
decl %edx
jz 4f
lodsb
stosb
orb %al,%al
jnz 1b
/* Success -- 0 byte reached */
decl %edx
xorl %eax,%eax
jmp 6f
4:
/* edx is zero -- return ENAMETOOLONG */
movl $ENAMETOOLONG,%eax
6:
/* set *lencopied and return %eax */
movl 20(%esp),%ecx
subl %edx,%ecx
movl 24(%esp),%edx
testl %edx,%edx
jz 7f
movl %ecx,(%edx)
7:
popl %edi
popl %esi
ret
ENTRY(bcmp)
pushl %edi
pushl %esi
movl 12(%esp),%edi
movl 16(%esp),%esi
movl 20(%esp),%edx
xorl %eax,%eax
movl %edx,%ecx
shrl $2,%ecx
cld /* compare forwards */
repe
cmpsl
jne 1f
movl %edx,%ecx
andl $3,%ecx
repe
cmpsb
je 2f
1:
incl %eax
2:
popl %esi
popl %edi
ret
/*
* Handling of special 386 registers and descriptor tables etc
*/
/* void lgdt(struct region_descriptor *rdp); */
ENTRY(lgdt)
/* reload the descriptor table */
movl 4(%esp),%eax
lgdt (%eax)
/* flush the prefetch q */
jmp 1f
nop
1:
/* reload "stale" selectors */
movl $KDSEL,%eax
mov %ax,%ds
mov %ax,%es
mov %ax,%gs
mov %ax,%ss
movl $KPSEL,%eax
mov %ax,%fs
/* reload code selector by turning return into intersegmental return */
movl (%esp),%eax
pushl %eax
movl $KCSEL,4(%esp)
lret
/*
* void lidt(struct region_descriptor *rdp);
*/
ENTRY(lidt)
movl 4(%esp),%eax
lidt (%eax)
ret
/*
* void lldt(u_short sel)
*/
ENTRY(lldt)
lldt 4(%esp)
ret
/*
* void ltr(u_short sel)
*/
ENTRY(ltr)
ltr 4(%esp)
ret
/* ssdtosd(*ssdp,*sdp) */
ENTRY(ssdtosd)
pushl %ebx
movl 8(%esp),%ecx
movl 8(%ecx),%ebx
shll $16,%ebx
movl (%ecx),%edx
roll $16,%edx
movb %dh,%bl
movb %dl,%bh
rorl $8,%ebx
movl 4(%ecx),%eax
movw %ax,%dx
andl $0xf0000,%eax
orl %eax,%ebx
movl 12(%esp),%ecx
movl %edx,(%ecx)
movl %ebx,4(%ecx)
popl %ebx
ret
/* void reset_dbregs() */
ENTRY(reset_dbregs)
movl $0,%eax
movl %eax,%dr7 /* disable all breapoints first */
movl %eax,%dr0
movl %eax,%dr1
movl %eax,%dr2
movl %eax,%dr3
movl %eax,%dr6
ret
/*****************************************************************************/
/* setjump, longjump */
/*****************************************************************************/
ENTRY(setjmp)
movl 4(%esp),%eax
movl %ebx,(%eax) /* save ebx */
movl %esp,4(%eax) /* save esp */
movl %ebp,8(%eax) /* save ebp */
movl %esi,12(%eax) /* save esi */
movl %edi,16(%eax) /* save edi */
movl (%esp),%edx /* get rta */
movl %edx,20(%eax) /* save eip */
xorl %eax,%eax /* return(0); */
ret
ENTRY(longjmp)
movl 4(%esp),%eax
movl (%eax),%ebx /* restore ebx */
movl 4(%eax),%esp /* restore esp */
movl 8(%eax),%ebp /* restore ebp */
movl 12(%eax),%esi /* restore esi */
movl 16(%eax),%edi /* restore edi */
movl 20(%eax),%edx /* get rta */
movl %edx,(%esp) /* put in return frame */
xorl %eax,%eax /* return(1); */
incl %eax
ret
/*
* Support for BB-profiling (gcc -a). The kernbb program will extract
* the data from the kernel.
*/
.data
ALIGN_DATA
.globl bbhead
bbhead:
.long 0
.text
NON_GPROF_ENTRY(__bb_init_func)
movl 4(%esp),%eax
movl $1,(%eax)
movl bbhead,%edx
movl %edx,16(%eax)
movl %eax,bbhead
NON_GPROF_RET
|