-
Notifications
You must be signed in to change notification settings - Fork 0
/
asm_386.s
1679 lines (1480 loc) · 38.2 KB
/
asm_386.s
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
// Copyright 2009 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
#include "go_asm.h"
#include "go_tls.h"
#include "funcdata.h"
#include "textflag.h"
TEXT runtime·rt0_go(SB),NOSPLIT,$0
// copy arguments forward on an even stack
MOVL argc+0(FP), AX
MOVL argv+4(FP), BX
SUBL $128, SP // plenty of scratch
ANDL $~15, SP
MOVL AX, 120(SP) // save argc, argv away
MOVL BX, 124(SP)
// set default stack bounds.
// _cgo_init may update stackguard.
MOVL $runtime·g0(SB), BP
LEAL (-64*1024+104)(SP), BX
MOVL BX, g_stackguard0(BP)
MOVL BX, g_stackguard1(BP)
MOVL BX, (g_stack+stack_lo)(BP)
MOVL SP, (g_stack+stack_hi)(BP)
// find out information about the processor we're on
MOVL $0, AX
CPUID
CMPL AX, $0
JE nocpuinfo
// Figure out how to serialize RDTSC.
// On Intel processors LFENCE is enough. AMD requires MFENCE.
// Don't know about the rest, so let's do MFENCE.
CMPL BX, $0x756E6547 // "Genu"
JNE notintel
CMPL DX, $0x49656E69 // "ineI"
JNE notintel
CMPL CX, $0x6C65746E // "ntel"
JNE notintel
MOVB $1, runtime·lfenceBeforeRdtsc(SB)
notintel:
MOVL $1, AX
CPUID
MOVL CX, runtime·cpuid_ecx(SB)
MOVL DX, runtime·cpuid_edx(SB)
nocpuinfo:
// if there is an _cgo_init, call it to let it
// initialize and to set up GS. if not,
// we set up GS ourselves.
MOVL _cgo_init(SB), AX
TESTL AX, AX
JZ needtls
MOVL $setg_gcc<>(SB), BX
MOVL BX, 4(SP)
MOVL BP, 0(SP)
CALL AX
// update stackguard after _cgo_init
MOVL $runtime·g0(SB), CX
MOVL (g_stack+stack_lo)(CX), AX
ADDL $const__StackGuard, AX
MOVL AX, g_stackguard0(CX)
MOVL AX, g_stackguard1(CX)
// skip runtime·ldt0setup(SB) and tls test after _cgo_init for non-windows
CMPL runtime·iswindows(SB), $0
JEQ ok
needtls:
// skip runtime·ldt0setup(SB) and tls test on Plan 9 in all cases
CMPL runtime·isplan9(SB), $1
JEQ ok
// set up %gs
CALL runtime·ldt0setup(SB)
// store through it, to make sure it works
get_tls(BX)
MOVL $0x123, g(BX)
MOVL runtime·tls0(SB), AX
CMPL AX, $0x123
JEQ ok
MOVL AX, 0 // abort
ok:
// set up m and g "registers"
get_tls(BX)
LEAL runtime·g0(SB), CX
MOVL CX, g(BX)
LEAL runtime·m0(SB), AX
// save m->g0 = g0
MOVL CX, m_g0(AX)
// save g0->m = m0
MOVL AX, g_m(CX)
CALL runtime·emptyfunc(SB) // fault if stack check is wrong
// convention is D is always cleared
CLD
CALL runtime·check(SB)
// saved argc, argv
MOVL 120(SP), AX
MOVL AX, 0(SP)
MOVL 124(SP), AX
MOVL AX, 4(SP)
CALL runtime·args(SB)
CALL runtime·osinit(SB)
CALL runtime·schedinit(SB)
// create a new goroutine to start program
PUSHL $runtime·mainPC(SB) // entry
PUSHL $0 // arg size
CALL runtime·newproc(SB)
POPL AX
POPL AX
// start this M
CALL runtime·mstart(SB)
INT $3
RET
DATA runtime·mainPC+0(SB)/4,$runtime·main(SB)
GLOBL runtime·mainPC(SB),RODATA,$4
TEXT runtime·breakpoint(SB),NOSPLIT,$0-0
INT $3
RET
TEXT runtime·asminit(SB),NOSPLIT,$0-0
// Linux and MinGW start the FPU in extended double precision.
// Other operating systems use double precision.
// Change to double precision to match them,
// and to match other hardware that only has double.
PUSHL $0x27F
FLDCW 0(SP)
POPL AX
RET
/*
* go-routine
*/
// void gosave(Gobuf*)
// save state in Gobuf; setjmp
TEXT runtime·gosave(SB), NOSPLIT, $0-4
MOVL buf+0(FP), AX // gobuf
LEAL buf+0(FP), BX // caller's SP
MOVL BX, gobuf_sp(AX)
MOVL 0(SP), BX // caller's PC
MOVL BX, gobuf_pc(AX)
MOVL $0, gobuf_ret(AX)
MOVL $0, gobuf_ctxt(AX)
get_tls(CX)
MOVL g(CX), BX
MOVL BX, gobuf_g(AX)
RET
// void gogo(Gobuf*)
// restore state from Gobuf; longjmp
TEXT runtime·gogo(SB), NOSPLIT, $0-4
MOVL buf+0(FP), BX // gobuf
MOVL gobuf_g(BX), DX
MOVL 0(DX), CX // make sure g != nil
get_tls(CX)
MOVL DX, g(CX)
MOVL gobuf_sp(BX), SP // restore SP
MOVL gobuf_ret(BX), AX
MOVL gobuf_ctxt(BX), DX
MOVL $0, gobuf_sp(BX) // clear to help garbage collector
MOVL $0, gobuf_ret(BX)
MOVL $0, gobuf_ctxt(BX)
MOVL gobuf_pc(BX), BX
JMP BX
// func mcall(fn func(*g))
// Switch to m->g0's stack, call fn(g).
// Fn must never return. It should gogo(&g->sched)
// to keep running g.
TEXT runtime·mcall(SB), NOSPLIT, $0-4
MOVL fn+0(FP), DI
get_tls(CX)
MOVL g(CX), AX // save state in g->sched
MOVL 0(SP), BX // caller's PC
MOVL BX, (g_sched+gobuf_pc)(AX)
LEAL fn+0(FP), BX // caller's SP
MOVL BX, (g_sched+gobuf_sp)(AX)
MOVL AX, (g_sched+gobuf_g)(AX)
// switch to m->g0 & its stack, call fn
MOVL g(CX), BX
MOVL g_m(BX), BX
MOVL m_g0(BX), SI
CMPL SI, AX // if g == m->g0 call badmcall
JNE 3(PC)
MOVL $runtime·badmcall(SB), AX
JMP AX
MOVL SI, g(CX) // g = m->g0
MOVL (g_sched+gobuf_sp)(SI), SP // sp = m->g0->sched.sp
PUSHL AX
MOVL DI, DX
MOVL 0(DI), DI
CALL DI
POPL AX
MOVL $runtime·badmcall2(SB), AX
JMP AX
RET
// systemstack_switch is a dummy routine that systemstack leaves at the bottom
// of the G stack. We need to distinguish the routine that
// lives at the bottom of the G stack from the one that lives
// at the top of the system stack because the one at the top of
// the system stack terminates the stack walk (see topofstack()).
TEXT runtime·systemstack_switch(SB), NOSPLIT, $0-0
RET
// func systemstack(fn func())
TEXT runtime·systemstack(SB), NOSPLIT, $0-4
MOVL fn+0(FP), DI // DI = fn
get_tls(CX)
MOVL g(CX), AX // AX = g
MOVL g_m(AX), BX // BX = m
MOVL m_gsignal(BX), DX // DX = gsignal
CMPL AX, DX
JEQ noswitch
MOVL m_g0(BX), DX // DX = g0
CMPL AX, DX
JEQ noswitch
MOVL m_curg(BX), BP
CMPL AX, BP
JEQ switch
// Bad: g is not gsignal, not g0, not curg. What is it?
// Hide call from linker nosplit analysis.
MOVL $runtime·badsystemstack(SB), AX
CALL AX
switch:
// save our state in g->sched. Pretend to
// be systemstack_switch if the G stack is scanned.
MOVL $runtime·systemstack_switch(SB), (g_sched+gobuf_pc)(AX)
MOVL SP, (g_sched+gobuf_sp)(AX)
MOVL AX, (g_sched+gobuf_g)(AX)
// switch to g0
MOVL DX, g(CX)
MOVL (g_sched+gobuf_sp)(DX), BX
// make it look like mstart called systemstack on g0, to stop traceback
SUBL $4, BX
MOVL $runtime·mstart(SB), DX
MOVL DX, 0(BX)
MOVL BX, SP
// call target function
MOVL DI, DX
MOVL 0(DI), DI
CALL DI
// switch back to g
get_tls(CX)
MOVL g(CX), AX
MOVL g_m(AX), BX
MOVL m_curg(BX), AX
MOVL AX, g(CX)
MOVL (g_sched+gobuf_sp)(AX), SP
MOVL $0, (g_sched+gobuf_sp)(AX)
RET
noswitch:
// already on system stack, just call directly
MOVL DI, DX
MOVL 0(DI), DI
CALL DI
RET
/*
* support for morestack
*/
// Called during function prolog when more stack is needed.
//
// The traceback routines see morestack on a g0 as being
// the top of a stack (for example, morestack calling newstack
// calling the scheduler calling newm calling gc), so we must
// record an argument size. For that purpose, it has no arguments.
TEXT runtime·morestack(SB),NOSPLIT,$0-0
// Cannot grow scheduler stack (m->g0).
get_tls(CX)
MOVL g(CX), BX
MOVL g_m(BX), BX
MOVL m_g0(BX), SI
CMPL g(CX), SI
JNE 2(PC)
INT $3
// Cannot grow signal stack.
MOVL m_gsignal(BX), SI
CMPL g(CX), SI
JNE 2(PC)
INT $3
// Called from f.
// Set m->morebuf to f's caller.
MOVL 4(SP), DI // f's caller's PC
MOVL DI, (m_morebuf+gobuf_pc)(BX)
LEAL 8(SP), CX // f's caller's SP
MOVL CX, (m_morebuf+gobuf_sp)(BX)
get_tls(CX)
MOVL g(CX), SI
MOVL SI, (m_morebuf+gobuf_g)(BX)
// Set g->sched to context in f.
MOVL 0(SP), AX // f's PC
MOVL AX, (g_sched+gobuf_pc)(SI)
MOVL SI, (g_sched+gobuf_g)(SI)
LEAL 4(SP), AX // f's SP
MOVL AX, (g_sched+gobuf_sp)(SI)
MOVL DX, (g_sched+gobuf_ctxt)(SI)
// Call newstack on m->g0's stack.
MOVL m_g0(BX), BP
MOVL BP, g(CX)
MOVL (g_sched+gobuf_sp)(BP), AX
MOVL -4(AX), BX // fault if CALL would, before smashing SP
MOVL AX, SP
CALL runtime·newstack(SB)
MOVL $0, 0x1003 // crash if newstack returns
RET
TEXT runtime·morestack_noctxt(SB),NOSPLIT,$0-0
MOVL $0, DX
JMP runtime·morestack(SB)
TEXT runtime·stackBarrier(SB),NOSPLIT,$0
// We came here via a RET to an overwritten return PC.
// AX may be live. Other registers are available.
// Get the original return PC, g.stkbar[g.stkbarPos].savedLRVal.
get_tls(CX)
MOVL g(CX), CX
MOVL (g_stkbar+slice_array)(CX), DX
MOVL g_stkbarPos(CX), BX
IMULL $stkbar__size, BX // Too big for SIB.
MOVL stkbar_savedLRVal(DX)(BX*1), BX
// Record that this stack barrier was hit.
ADDL $1, g_stkbarPos(CX)
// Jump to the original return PC.
JMP BX
// reflectcall: call a function with the given argument list
// func call(argtype *_type, f *FuncVal, arg *byte, argsize, retoffset uint32).
// we don't have variable-sized frames, so we use a small number
// of constant-sized-frame functions to encode a few bits of size in the pc.
// Caution: ugly multiline assembly macros in your future!
#define DISPATCH(NAME,MAXSIZE) \
CMPL CX, $MAXSIZE; \
JA 3(PC); \
MOVL $NAME(SB), AX; \
JMP AX
// Note: can't just "JMP NAME(SB)" - bad inlining results.
TEXT reflect·call(SB), NOSPLIT, $0-0
JMP ·reflectcall(SB)
TEXT ·reflectcall(SB), NOSPLIT, $0-20
MOVL argsize+12(FP), CX
DISPATCH(runtime·call16, 16)
DISPATCH(runtime·call32, 32)
DISPATCH(runtime·call64, 64)
DISPATCH(runtime·call128, 128)
DISPATCH(runtime·call256, 256)
DISPATCH(runtime·call512, 512)
DISPATCH(runtime·call1024, 1024)
DISPATCH(runtime·call2048, 2048)
DISPATCH(runtime·call4096, 4096)
DISPATCH(runtime·call8192, 8192)
DISPATCH(runtime·call16384, 16384)
DISPATCH(runtime·call32768, 32768)
DISPATCH(runtime·call65536, 65536)
DISPATCH(runtime·call131072, 131072)
DISPATCH(runtime·call262144, 262144)
DISPATCH(runtime·call524288, 524288)
DISPATCH(runtime·call1048576, 1048576)
DISPATCH(runtime·call2097152, 2097152)
DISPATCH(runtime·call4194304, 4194304)
DISPATCH(runtime·call8388608, 8388608)
DISPATCH(runtime·call16777216, 16777216)
DISPATCH(runtime·call33554432, 33554432)
DISPATCH(runtime·call67108864, 67108864)
DISPATCH(runtime·call134217728, 134217728)
DISPATCH(runtime·call268435456, 268435456)
DISPATCH(runtime·call536870912, 536870912)
DISPATCH(runtime·call1073741824, 1073741824)
MOVL $runtime·badreflectcall(SB), AX
JMP AX
#define CALLFN(NAME,MAXSIZE) \
TEXT NAME(SB), WRAPPER, $MAXSIZE-20; \
NO_LOCAL_POINTERS; \
/* copy arguments to stack */ \
MOVL argptr+8(FP), SI; \
MOVL argsize+12(FP), CX; \
MOVL SP, DI; \
REP;MOVSB; \
/* call function */ \
MOVL f+4(FP), DX; \
MOVL (DX), AX; \
PCDATA $PCDATA_StackMapIndex, $0; \
CALL AX; \
/* copy return values back */ \
MOVL argptr+8(FP), DI; \
MOVL argsize+12(FP), CX; \
MOVL retoffset+16(FP), BX; \
MOVL SP, SI; \
ADDL BX, DI; \
ADDL BX, SI; \
SUBL BX, CX; \
REP;MOVSB; \
/* execute write barrier updates */ \
MOVL argtype+0(FP), DX; \
MOVL argptr+8(FP), DI; \
MOVL argsize+12(FP), CX; \
MOVL retoffset+16(FP), BX; \
MOVL DX, 0(SP); \
MOVL DI, 4(SP); \
MOVL CX, 8(SP); \
MOVL BX, 12(SP); \
CALL runtime·callwritebarrier(SB); \
RET
CALLFN(·call16, 16)
CALLFN(·call32, 32)
CALLFN(·call64, 64)
CALLFN(·call128, 128)
CALLFN(·call256, 256)
CALLFN(·call512, 512)
CALLFN(·call1024, 1024)
CALLFN(·call2048, 2048)
CALLFN(·call4096, 4096)
CALLFN(·call8192, 8192)
CALLFN(·call16384, 16384)
CALLFN(·call32768, 32768)
CALLFN(·call65536, 65536)
CALLFN(·call131072, 131072)
CALLFN(·call262144, 262144)
CALLFN(·call524288, 524288)
CALLFN(·call1048576, 1048576)
CALLFN(·call2097152, 2097152)
CALLFN(·call4194304, 4194304)
CALLFN(·call8388608, 8388608)
CALLFN(·call16777216, 16777216)
CALLFN(·call33554432, 33554432)
CALLFN(·call67108864, 67108864)
CALLFN(·call134217728, 134217728)
CALLFN(·call268435456, 268435456)
CALLFN(·call536870912, 536870912)
CALLFN(·call1073741824, 1073741824)
// bool cas(int32 *val, int32 old, int32 new)
// Atomically:
// if(*val == old){
// *val = new;
// return 1;
// }else
// return 0;
TEXT runtime·cas(SB), NOSPLIT, $0-13
MOVL ptr+0(FP), BX
MOVL old+4(FP), AX
MOVL new+8(FP), CX
LOCK
CMPXCHGL CX, 0(BX)
SETEQ ret+12(FP)
RET
TEXT runtime·casuintptr(SB), NOSPLIT, $0-13
JMP runtime·cas(SB)
TEXT runtime·atomicloaduintptr(SB), NOSPLIT, $0-8
JMP runtime·atomicload(SB)
TEXT runtime·atomicloaduint(SB), NOSPLIT, $0-8
JMP runtime·atomicload(SB)
TEXT runtime·atomicstoreuintptr(SB), NOSPLIT, $0-8
JMP runtime·atomicstore(SB)
// bool runtime·cas64(uint64 *val, uint64 old, uint64 new)
// Atomically:
// if(*val == *old){
// *val = new;
// return 1;
// } else {
// return 0;
// }
TEXT runtime·cas64(SB), NOSPLIT, $0-21
MOVL ptr+0(FP), BP
MOVL old_lo+4(FP), AX
MOVL old_hi+8(FP), DX
MOVL new_lo+12(FP), BX
MOVL new_hi+16(FP), CX
LOCK
CMPXCHG8B 0(BP)
SETEQ ret+20(FP)
RET
// bool casp(void **p, void *old, void *new)
// Atomically:
// if(*p == old){
// *p = new;
// return 1;
// }else
// return 0;
TEXT runtime·casp1(SB), NOSPLIT, $0-13
MOVL ptr+0(FP), BX
MOVL old+4(FP), AX
MOVL new+8(FP), CX
LOCK
CMPXCHGL CX, 0(BX)
SETEQ ret+12(FP)
RET
// uint32 xadd(uint32 volatile *val, int32 delta)
// Atomically:
// *val += delta;
// return *val;
TEXT runtime·xadd(SB), NOSPLIT, $0-12
MOVL ptr+0(FP), BX
MOVL delta+4(FP), AX
MOVL AX, CX
LOCK
XADDL AX, 0(BX)
ADDL CX, AX
MOVL AX, ret+8(FP)
RET
TEXT runtime·xchg(SB), NOSPLIT, $0-12
MOVL ptr+0(FP), BX
MOVL new+4(FP), AX
XCHGL AX, 0(BX)
MOVL AX, ret+8(FP)
RET
TEXT runtime·xchguintptr(SB), NOSPLIT, $0-12
JMP runtime·xchg(SB)
TEXT runtime·procyield(SB),NOSPLIT,$0-0
MOVL cycles+0(FP), AX
again:
PAUSE
SUBL $1, AX
JNZ again
RET
TEXT runtime·atomicstorep1(SB), NOSPLIT, $0-8
MOVL ptr+0(FP), BX
MOVL val+4(FP), AX
XCHGL AX, 0(BX)
RET
TEXT runtime·atomicstore(SB), NOSPLIT, $0-8
MOVL ptr+0(FP), BX
MOVL val+4(FP), AX
XCHGL AX, 0(BX)
RET
// uint64 atomicload64(uint64 volatile* addr);
TEXT runtime·atomicload64(SB), NOSPLIT, $0-12
MOVL ptr+0(FP), AX
TESTL $7, AX
JZ 2(PC)
MOVL 0, AX // crash with nil ptr deref
LEAL ret_lo+4(FP), BX
// MOVQ (%EAX), %MM0
BYTE $0x0f; BYTE $0x6f; BYTE $0x00
// MOVQ %MM0, 0(%EBX)
BYTE $0x0f; BYTE $0x7f; BYTE $0x03
// EMMS
BYTE $0x0F; BYTE $0x77
RET
// void runtime·atomicstore64(uint64 volatile* addr, uint64 v);
TEXT runtime·atomicstore64(SB), NOSPLIT, $0-12
MOVL ptr+0(FP), AX
TESTL $7, AX
JZ 2(PC)
MOVL 0, AX // crash with nil ptr deref
// MOVQ and EMMS were introduced on the Pentium MMX.
// MOVQ 0x8(%ESP), %MM0
BYTE $0x0f; BYTE $0x6f; BYTE $0x44; BYTE $0x24; BYTE $0x08
// MOVQ %MM0, (%EAX)
BYTE $0x0f; BYTE $0x7f; BYTE $0x00
// EMMS
BYTE $0x0F; BYTE $0x77
// This is essentially a no-op, but it provides required memory fencing.
// It can be replaced with MFENCE, but MFENCE was introduced only on the Pentium4 (SSE2).
MOVL $0, AX
LOCK
XADDL AX, (SP)
RET
// void runtime·atomicor8(byte volatile*, byte);
TEXT runtime·atomicor8(SB), NOSPLIT, $0-5
MOVL ptr+0(FP), AX
MOVB val+4(FP), BX
LOCK
ORB BX, (AX)
RET
// void runtime·atomicand8(byte volatile*, byte);
TEXT runtime·atomicand8(SB), NOSPLIT, $0-5
MOVL ptr+0(FP), AX
MOVB val+4(FP), BX
LOCK
ANDB BX, (AX)
RET
TEXT ·publicationBarrier(SB),NOSPLIT,$0-0
// Stores are already ordered on x86, so this is just a
// compile barrier.
RET
// void jmpdefer(fn, sp);
// called from deferreturn.
// 1. pop the caller
// 2. sub 5 bytes from the callers return
// 3. jmp to the argument
TEXT runtime·jmpdefer(SB), NOSPLIT, $0-8
MOVL fv+0(FP), DX // fn
MOVL argp+4(FP), BX // caller sp
LEAL -4(BX), SP // caller sp after CALL
SUBL $5, (SP) // return to CALL again
MOVL 0(DX), BX
JMP BX // but first run the deferred function
// Save state of caller into g->sched.
TEXT gosave<>(SB),NOSPLIT,$0
PUSHL AX
PUSHL BX
get_tls(BX)
MOVL g(BX), BX
LEAL arg+0(FP), AX
MOVL AX, (g_sched+gobuf_sp)(BX)
MOVL -4(AX), AX
MOVL AX, (g_sched+gobuf_pc)(BX)
MOVL $0, (g_sched+gobuf_ret)(BX)
MOVL $0, (g_sched+gobuf_ctxt)(BX)
POPL BX
POPL AX
RET
// func asmcgocall(fn, arg unsafe.Pointer) int32
// Call fn(arg) on the scheduler stack,
// aligned appropriately for the gcc ABI.
// See cgocall.go for more details.
TEXT ·asmcgocall(SB),NOSPLIT,$0-12
MOVL fn+0(FP), AX
MOVL arg+4(FP), BX
MOVL SP, DX
// Figure out if we need to switch to m->g0 stack.
// We get called to create new OS threads too, and those
// come in on the m->g0 stack already.
get_tls(CX)
MOVL g(CX), BP
MOVL g_m(BP), BP
MOVL m_g0(BP), SI
MOVL g(CX), DI
CMPL SI, DI
JEQ 4(PC)
CALL gosave<>(SB)
MOVL SI, g(CX)
MOVL (g_sched+gobuf_sp)(SI), SP
// Now on a scheduling stack (a pthread-created stack).
SUBL $32, SP
ANDL $~15, SP // alignment, perhaps unnecessary
MOVL DI, 8(SP) // save g
MOVL (g_stack+stack_hi)(DI), DI
SUBL DX, DI
MOVL DI, 4(SP) // save depth in stack (can't just save SP, as stack might be copied during a callback)
MOVL BX, 0(SP) // first argument in x86-32 ABI
CALL AX
// Restore registers, g, stack pointer.
get_tls(CX)
MOVL 8(SP), DI
MOVL (g_stack+stack_hi)(DI), SI
SUBL 4(SP), SI
MOVL DI, g(CX)
MOVL SI, SP
MOVL AX, ret+8(FP)
RET
// cgocallback(void (*fn)(void*), void *frame, uintptr framesize)
// Turn the fn into a Go func (by taking its address) and call
// cgocallback_gofunc.
TEXT runtime·cgocallback(SB),NOSPLIT,$12-12
LEAL fn+0(FP), AX
MOVL AX, 0(SP)
MOVL frame+4(FP), AX
MOVL AX, 4(SP)
MOVL framesize+8(FP), AX
MOVL AX, 8(SP)
MOVL $runtime·cgocallback_gofunc(SB), AX
CALL AX
RET
// cgocallback_gofunc(FuncVal*, void *frame, uintptr framesize)
// See cgocall.go for more details.
TEXT ·cgocallback_gofunc(SB),NOSPLIT,$12-12
NO_LOCAL_POINTERS
// If g is nil, Go did not create the current thread.
// Call needm to obtain one for temporary use.
// In this case, we're running on the thread stack, so there's
// lots of space, but the linker doesn't know. Hide the call from
// the linker analysis by using an indirect call through AX.
get_tls(CX)
#ifdef GOOS_windows
MOVL $0, BP
CMPL CX, $0
JEQ 2(PC) // TODO
#endif
MOVL g(CX), BP
CMPL BP, $0
JEQ needm
MOVL g_m(BP), BP
MOVL BP, DX // saved copy of oldm
JMP havem
needm:
MOVL $0, 0(SP)
MOVL $runtime·needm(SB), AX
CALL AX
MOVL 0(SP), DX
get_tls(CX)
MOVL g(CX), BP
MOVL g_m(BP), BP
// Set m->sched.sp = SP, so that if a panic happens
// during the function we are about to execute, it will
// have a valid SP to run on the g0 stack.
// The next few lines (after the havem label)
// will save this SP onto the stack and then write
// the same SP back to m->sched.sp. That seems redundant,
// but if an unrecovered panic happens, unwindm will
// restore the g->sched.sp from the stack location
// and then systemstack will try to use it. If we don't set it here,
// that restored SP will be uninitialized (typically 0) and
// will not be usable.
MOVL m_g0(BP), SI
MOVL SP, (g_sched+gobuf_sp)(SI)
havem:
// Now there's a valid m, and we're running on its m->g0.
// Save current m->g0->sched.sp on stack and then set it to SP.
// Save current sp in m->g0->sched.sp in preparation for
// switch back to m->curg stack.
// NOTE: unwindm knows that the saved g->sched.sp is at 0(SP).
MOVL m_g0(BP), SI
MOVL (g_sched+gobuf_sp)(SI), AX
MOVL AX, 0(SP)
MOVL SP, (g_sched+gobuf_sp)(SI)
// Switch to m->curg stack and call runtime.cgocallbackg.
// Because we are taking over the execution of m->curg
// but *not* resuming what had been running, we need to
// save that information (m->curg->sched) so we can restore it.
// We can restore m->curg->sched.sp easily, because calling
// runtime.cgocallbackg leaves SP unchanged upon return.
// To save m->curg->sched.pc, we push it onto the stack.
// This has the added benefit that it looks to the traceback
// routine like cgocallbackg is going to return to that
// PC (because the frame we allocate below has the same
// size as cgocallback_gofunc's frame declared above)
// so that the traceback will seamlessly trace back into
// the earlier calls.
//
// In the new goroutine, 0(SP) holds the saved oldm (DX) register.
// 4(SP) and 8(SP) are unused.
MOVL m_curg(BP), SI
MOVL SI, g(CX)
MOVL (g_sched+gobuf_sp)(SI), DI // prepare stack as DI
MOVL (g_sched+gobuf_pc)(SI), BP
MOVL BP, -4(DI)
LEAL -(4+12)(DI), SP
MOVL DX, 0(SP)
CALL runtime·cgocallbackg(SB)
MOVL 0(SP), DX
// Restore g->sched (== m->curg->sched) from saved values.
get_tls(CX)
MOVL g(CX), SI
MOVL 12(SP), BP
MOVL BP, (g_sched+gobuf_pc)(SI)
LEAL (12+4)(SP), DI
MOVL DI, (g_sched+gobuf_sp)(SI)
// Switch back to m->g0's stack and restore m->g0->sched.sp.
// (Unlike m->curg, the g0 goroutine never uses sched.pc,
// so we do not have to restore it.)
MOVL g(CX), BP
MOVL g_m(BP), BP
MOVL m_g0(BP), SI
MOVL SI, g(CX)
MOVL (g_sched+gobuf_sp)(SI), SP
MOVL 0(SP), AX
MOVL AX, (g_sched+gobuf_sp)(SI)
// If the m on entry was nil, we called needm above to borrow an m
// for the duration of the call. Since the call is over, return it with dropm.
CMPL DX, $0
JNE 3(PC)
MOVL $runtime·dropm(SB), AX
CALL AX
// Done!
RET
// void setg(G*); set g. for use by needm.
TEXT runtime·setg(SB), NOSPLIT, $0-4
MOVL gg+0(FP), BX
#ifdef GOOS_windows
CMPL BX, $0
JNE settls
MOVL $0, 0x14(FS)
RET
settls:
MOVL g_m(BX), AX
LEAL m_tls(AX), AX
MOVL AX, 0x14(FS)
#endif
get_tls(CX)
MOVL BX, g(CX)
RET
// void setg_gcc(G*); set g. for use by gcc
TEXT setg_gcc<>(SB), NOSPLIT, $0
get_tls(AX)
MOVL gg+0(FP), DX
MOVL DX, g(AX)
RET
// check that SP is in range [g->stack.lo, g->stack.hi)
TEXT runtime·stackcheck(SB), NOSPLIT, $0-0
get_tls(CX)
MOVL g(CX), AX
CMPL (g_stack+stack_hi)(AX), SP
JHI 2(PC)
INT $3
CMPL SP, (g_stack+stack_lo)(AX)
JHI 2(PC)
INT $3
RET
TEXT runtime·getcallerpc(SB),NOSPLIT,$4-8
MOVL argp+0(FP),AX // addr of first arg
MOVL -4(AX),AX // get calling pc
CMPL AX, runtime·stackBarrierPC(SB)
JNE nobar
// Get original return PC.
CALL runtime·nextBarrierPC(SB)
MOVL 0(SP), AX
nobar:
MOVL AX, ret+4(FP)
RET
TEXT runtime·setcallerpc(SB),NOSPLIT,$4-8
MOVL argp+0(FP),AX // addr of first arg
MOVL pc+4(FP), BX
MOVL -4(AX), CX
CMPL CX, runtime·stackBarrierPC(SB)
JEQ setbar
MOVL BX, -4(AX) // set calling pc
RET
setbar:
// Set the stack barrier return PC.
MOVL BX, 0(SP)
CALL runtime·setNextBarrierPC(SB)
RET
TEXT runtime·getcallersp(SB), NOSPLIT, $0-8
MOVL argp+0(FP), AX
MOVL AX, ret+4(FP)
RET
// func cputicks() int64
TEXT runtime·cputicks(SB),NOSPLIT,$0-8
TESTL $0x4000000, runtime·cpuid_edx(SB) // no sse2, no mfence
JEQ done
CMPB runtime·lfenceBeforeRdtsc(SB), $1
JNE mfence
BYTE $0x0f; BYTE $0xae; BYTE $0xe8 // LFENCE
JMP done
mfence:
BYTE $0x0f; BYTE $0xae; BYTE $0xf0 // MFENCE
done:
RDTSC
MOVL AX, ret_lo+0(FP)
MOVL DX, ret_hi+4(FP)
RET
TEXT runtime·ldt0setup(SB),NOSPLIT,$16-0
// set up ldt 7 to point at tls0
// ldt 1 would be fine on Linux, but on OS X, 7 is as low as we can go.
// the entry number is just a hint. setldt will set up GS with what it used.
MOVL $7, 0(SP)
LEAL runtime·tls0(SB), AX
MOVL AX, 4(SP)
MOVL $32, 8(SP) // sizeof(tls array)
CALL runtime·setldt(SB)
RET
TEXT runtime·emptyfunc(SB),0,$0-0
RET
TEXT runtime·abort(SB),NOSPLIT,$0-0
INT $0x3
// memhash_varlen(p unsafe.Pointer, h seed) uintptr
// redirects to memhash(p, h, size) using the size
// stored in the closure.
TEXT runtime·memhash_varlen(SB),NOSPLIT,$16-12
GO_ARGS
NO_LOCAL_POINTERS
MOVL p+0(FP), AX
MOVL h+4(FP), BX
MOVL 4(DX), CX
MOVL AX, 0(SP)
MOVL BX, 4(SP)
MOVL CX, 8(SP)
CALL runtime·memhash(SB)
MOVL 12(SP), AX
MOVL AX, ret+8(FP)
RET
// hash function using AES hardware instructions
TEXT runtime·aeshash(SB),NOSPLIT,$0-16
MOVL p+0(FP), AX // ptr to data
MOVL s+8(FP), CX // size
LEAL ret+12(FP), DX
JMP runtime·aeshashbody(SB)
TEXT runtime·aeshashstr(SB),NOSPLIT,$0-12
MOVL p+0(FP), AX // ptr to string object
MOVL 4(AX), CX // length of string
MOVL (AX), AX // string data
LEAL ret+8(FP), DX
JMP runtime·aeshashbody(SB)
// AX: data
// CX: length
// DX: address to put return value
TEXT runtime·aeshashbody(SB),NOSPLIT,$0-0
MOVL h+4(FP), X6 // seed to low 64 bits of xmm6
PINSRD $2, CX, X6 // size to high 64 bits of xmm6
PSHUFHW $0, X6, X6 // replace size with its low 2 bytes repeated 4 times
MOVO runtime·aeskeysched(SB), X7
CMPL CX, $16
JB aes0to15
JE aes16
CMPL CX, $32
JBE aes17to32
CMPL CX, $64
JBE aes33to64
JMP aes65plus
aes0to15:
TESTL CX, CX
JE aes0
ADDL $16, AX
TESTW $0xff0, AX
JE endofpage
// 16 bytes loaded at this address won't cross
// a page boundary, so we can load it directly.
MOVOU -16(AX), X0
ADDL CX, CX
PAND masks<>(SB)(CX*8), X0
// scramble 3 times
AESENC X6, X0
AESENC X7, X0
AESENC X7, X0
MOVL X0, (DX)
RET