summaryrefslogtreecommitdiffstats
path: root/sys/arm/include/atomic.h
blob: 57fdf717aa81e7264f331452177b73c6bec7725e (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
1001
1002
1003
1004
1005
1006
1007
1008
1009
1010
1011
1012
1013
1014
1015
1016
1017
1018
1019
1020
1021
1022
1023
1024
1025
1026
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
1040
1041
1042
1043
1044
1045
1046
1047
1048
1049
1050
1051
1052
1053
1054
1055
1056
1057
1058
1059
1060
1061
1062
1063
1064
1065
1066
1067
1068
1069
1070
1071
1072
1073
1074
1075
1076
1077
1078
1079
1080
1081
1082
1083
1084
1085
1086
1087
1088
1089
1090
1091
1092
1093
1094
1095
1096
1097
1098
1099
1100
1101
1102
1103
1104
1105
1106
1107
1108
1109
1110
1111
1112
1113
1114
1115
1116
1117
1118
1119
1120
1121
1122
1123
1124
1125
1126
1127
1128
1129
1130
1131
1132
1133
1134
/* $NetBSD: atomic.h,v 1.1 2002/10/19 12:22:34 bsh Exp $ */

/*-
 * Copyright (C) 2003-2004 Olivier Houchard
 * Copyright (C) 1994-1997 Mark Brinicombe
 * Copyright (C) 1994 Brini
 * All rights reserved.
 *
 * This code is derived from software written for Brini by Mark Brinicombe
 *
 * 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 Brini.
 * 4. The name of Brini may not be used to endorse or promote products
 *    derived from this software without specific prior written permission.
 *
 * THIS SOFTWARE IS PROVIDED BY BRINI ``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 BRINI 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.
 *
 * $FreeBSD$
 */

#ifndef	_MACHINE_ATOMIC_H_
#define	_MACHINE_ATOMIC_H_

#include <sys/types.h>
#include <machine/armreg.h>

#ifndef _KERNEL
#include <machine/sysarch.h>
#else
#include <machine/cpuconf.h>
#endif

#if defined (__ARM_ARCH_7__) || defined (__ARM_ARCH_7A__)
#define isb()  __asm __volatile("isb" : : : "memory")
#define dsb()  __asm __volatile("dsb" : : : "memory")
#define dmb()  __asm __volatile("dmb" : : : "memory")
#elif defined (__ARM_ARCH_6__) || defined (__ARM_ARCH_6J__) || \
  defined (__ARM_ARCH_6K__) || defined (__ARM_ARCH_6T2__) || \
  defined (__ARM_ARCH_6Z__) || defined (__ARM_ARCH_6ZK__)
#define isb()  __asm __volatile("mcr p15, 0, %0, c7, c5, 4" : : "r" (0) : "memory")
#define dsb()  __asm __volatile("mcr p15, 0, %0, c7, c10, 4" : : "r" (0) : "memory")
#define dmb()  __asm __volatile("mcr p15, 0, %0, c7, c10, 5" : : "r" (0) : "memory")
#else
#define isb()  __asm __volatile("mcr p15, 0, %0, c7, c5, 4" : : "r" (0) : "memory")
#define dsb()  __asm __volatile("mcr p15, 0, %0, c7, c10, 4" : : "r" (0) : "memory")
#define dmb()  dsb()
#endif

#define mb()   dmb()
#define wmb()  dmb()
#define rmb()  dmb()



/*
 * It would be nice to use _HAVE_ARMv6_INSTRUCTIONS from machine/asm.h
 * here, but that header can't be included here because this is C
 * code.  I would like to move the _HAVE_ARMv6_INSTRUCTIONS definition
 * out of asm.h so it can be used in both asm and C code. - kientzle@
 */
#if defined (__ARM_ARCH_7__) || \
	defined (__ARM_ARCH_7A__)  || \
	defined (__ARM_ARCH_6__)   || \
	defined (__ARM_ARCH_6J__)  || \
	defined (__ARM_ARCH_6K__)  || \
	defined (__ARM_ARCH_6T2__) || \
	defined (__ARM_ARCH_6Z__)  || \
	defined (__ARM_ARCH_6ZK__)
#define	ARM_HAVE_ATOMIC64

static __inline void
__do_dmb(void)
{

#if defined (__ARM_ARCH_7__) || defined (__ARM_ARCH_7A__)
	__asm __volatile("dmb" : : : "memory");
#else
	__asm __volatile("mcr p15, 0, r0, c7, c10, 5" : : : "memory");
#endif
}

#define ATOMIC_ACQ_REL_LONG(NAME)					\
static __inline void							\
atomic_##NAME##_acq_long(__volatile u_long *p, u_long v)		\
{									\
	atomic_##NAME##_long(p, v);					\
	__do_dmb();							\
}									\
									\
static __inline  void							\
atomic_##NAME##_rel_long(__volatile u_long *p, u_long v)		\
{									\
	__do_dmb();							\
	atomic_##NAME##_long(p, v);					\
}

#define	ATOMIC_ACQ_REL(NAME, WIDTH)					\
static __inline  void							\
atomic_##NAME##_acq_##WIDTH(__volatile uint##WIDTH##_t *p, uint##WIDTH##_t v)\
{									\
	atomic_##NAME##_##WIDTH(p, v);					\
	__do_dmb();							\
}									\
									\
static __inline  void							\
atomic_##NAME##_rel_##WIDTH(__volatile uint##WIDTH##_t *p, uint##WIDTH##_t v)\
{									\
	__do_dmb();							\
	atomic_##NAME##_##WIDTH(p, v);					\
}

static __inline void
atomic_set_32(volatile uint32_t *address, uint32_t setmask)
{
	uint32_t tmp = 0, tmp2 = 0;

	__asm __volatile("1: ldrex %0, [%2]\n"
	    		    "orr %0, %0, %3\n"
			    "strex %1, %0, [%2]\n"
			    "cmp %1, #0\n"
	                    "it ne\n"
			    "bne	1b\n"
			   : "=&r" (tmp), "+r" (tmp2)
			   , "+r" (address), "+r" (setmask) : : "cc", "memory");
			     
}

static __inline void
atomic_set_64(volatile uint64_t *p, uint64_t val)
{
	uint64_t tmp;
	uint32_t exflag;

	__asm __volatile(
		"1:          \n"
		"   ldrexd   %[tmp], [%[ptr]]\n"
		"   orr      %Q[tmp], %Q[val]\n"
		"   orr      %R[tmp], %R[val]\n"
		"   strexd   %[exf], %[tmp], [%[ptr]]\n"
		"   teq      %[exf], #0\n"
		"   it ne    \n"
		"   bne      1b\n"
		:   [exf]    "=&r"  (exflag), 
		    [tmp]    "=&r"  (tmp)
		:   [ptr]    "r"    (p), 
		    [val]    "r"    (val)
		:   "cc", "memory");
}

static __inline void
atomic_set_long(volatile u_long *address, u_long setmask)
{
	u_long tmp = 0, tmp2 = 0;

	__asm __volatile("1: ldrex %0, [%2]\n"
	    		    "orr %0, %0, %3\n"
			    "strex %1, %0, [%2]\n"
			    "cmp %1, #0\n"
	                    "it ne\n"
			    "bne	1b\n"
			   : "=&r" (tmp), "+r" (tmp2)
			   , "+r" (address), "+r" (setmask) : : "cc", "memory");
			     
}

static __inline void
atomic_clear_32(volatile uint32_t *address, uint32_t setmask)
{
	uint32_t tmp = 0, tmp2 = 0;

	__asm __volatile("1: ldrex %0, [%2]\n"
	    		    "bic %0, %0, %3\n"
			    "strex %1, %0, [%2]\n"
			    "cmp %1, #0\n"
	                    "it ne\n"
			    "bne	1b\n"
			   : "=&r" (tmp), "+r" (tmp2)
			   ,"+r" (address), "+r" (setmask) : : "cc", "memory");
}

static __inline void
atomic_clear_64(volatile uint64_t *p, uint64_t val)
{
	uint64_t tmp;
	uint32_t exflag;

	__asm __volatile(
		"1:          \n"
		"   ldrexd   %[tmp], [%[ptr]]\n"
		"   bic      %Q[tmp], %Q[val]\n"
		"   bic      %R[tmp], %R[val]\n"
		"   strexd   %[exf], %[tmp], [%[ptr]]\n"
		"   teq      %[exf], #0\n"
		"   it ne    \n"
		"   bne      1b\n"
		:   [exf]    "=&r"  (exflag), 
		    [tmp]    "=&r"  (tmp)
		:   [ptr]    "r"    (p), 
		    [val]    "r"    (val)
		:   "cc", "memory");
}

static __inline void
atomic_clear_long(volatile u_long *address, u_long setmask)
{
	u_long tmp = 0, tmp2 = 0;

	__asm __volatile("1: ldrex %0, [%2]\n"
	    		    "bic %0, %0, %3\n"
			    "strex %1, %0, [%2]\n"
			    "cmp %1, #0\n"
	                    "it ne\n"
			    "bne	1b\n"
			   : "=&r" (tmp), "+r" (tmp2)
			   ,"+r" (address), "+r" (setmask) : : "cc", "memory");
}

static __inline u_int32_t
atomic_cmpset_32(volatile u_int32_t *p, volatile u_int32_t cmpval, volatile u_int32_t newval)
{
	uint32_t ret;
	
	__asm __volatile("1: ldrex %0, [%1]\n"
	                 "cmp %0, %2\n"
	                 "itt ne\n"
			 "movne %0, #0\n"
			 "bne 2f\n"
			 "strex %0, %3, [%1]\n"
			 "cmp %0, #0\n"
	                 "ite eq\n"
			 "moveq %0, #1\n"
			 "bne	1b\n"
			 "2:"
			 : "=&r" (ret)
			 ,"+r" (p), "+r" (cmpval), "+r" (newval) : : "cc",
			 "memory");
	return (ret);
}

static __inline int
atomic_cmpset_64(volatile uint64_t *p, uint64_t cmpval, uint64_t newval)
{
	uint64_t tmp;
	uint32_t ret;

	__asm __volatile(
		"1:          \n"
		"   ldrexd   %[tmp], [%[ptr]]\n"
		"   teq      %Q[tmp], %Q[cmpval]\n"
		"   itee eq  \n"
		"   teqeq    %R[tmp], %R[cmpval]\n"
		"   movne    %[ret], #0\n"
		"   bne      2f\n"
		"   strexd   %[ret], %[newval], [%[ptr]]\n"
		"   teq      %[ret], #0\n"
		"   it ne    \n"
		"   bne      1b\n"
		"   mov      %[ret], #1\n"
		"2:          \n"
		:   [ret]    "=&r"  (ret), 
		    [tmp]    "=&r"  (tmp)
		:   [ptr]    "r"    (p), 
		    [cmpval] "r"    (cmpval), 
		    [newval] "r"    (newval)
		:   "cc", "memory");
	return (ret);
}

static __inline u_long
atomic_cmpset_long(volatile u_long *p, volatile u_long cmpval, volatile u_long newval)
{
	u_long ret;
	
	__asm __volatile("1: ldrex %0, [%1]\n"
	                 "cmp %0, %2\n"
	                 "itt ne\n"
			 "movne %0, #0\n"
			 "bne 2f\n"
			 "strex %0, %3, [%1]\n"
			 "cmp %0, #0\n"
	                 "ite eq\n"
			 "moveq %0, #1\n"
			 "bne	1b\n"
			 "2:"
			 : "=&r" (ret)
			 ,"+r" (p), "+r" (cmpval), "+r" (newval) : : "cc",
			 "memory");
	return (ret);
}

static __inline u_int32_t
atomic_cmpset_acq_32(volatile u_int32_t *p, volatile u_int32_t cmpval, volatile u_int32_t newval)
{
	u_int32_t ret = atomic_cmpset_32(p, cmpval, newval);

	__do_dmb();
	return (ret);
}

static __inline uint64_t
atomic_cmpset_acq_64(volatile uint64_t *p, volatile uint64_t cmpval, volatile uint64_t newval)
{
	uint64_t ret = atomic_cmpset_64(p, cmpval, newval);

	__do_dmb();
	return (ret);
}

static __inline u_long
atomic_cmpset_acq_long(volatile u_long *p, volatile u_long cmpval, volatile u_long newval)
{
	u_long ret = atomic_cmpset_long(p, cmpval, newval);

	__do_dmb();
	return (ret);
}

static __inline u_int32_t
atomic_cmpset_rel_32(volatile u_int32_t *p, volatile u_int32_t cmpval, volatile u_int32_t newval)
{
	
	__do_dmb();
	return (atomic_cmpset_32(p, cmpval, newval));
}

static __inline uint64_t
atomic_cmpset_rel_64(volatile uint64_t *p, volatile uint64_t cmpval, volatile uint64_t newval)
{
	
	__do_dmb();
	return (atomic_cmpset_64(p, cmpval, newval));
}

static __inline u_long
atomic_cmpset_rel_long(volatile u_long *p, volatile u_long cmpval, volatile u_long newval)
{
	
	__do_dmb();
	return (atomic_cmpset_long(p, cmpval, newval));
}


static __inline void
atomic_add_32(volatile u_int32_t *p, u_int32_t val)
{
	uint32_t tmp = 0, tmp2 = 0;

	__asm __volatile("1: ldrex %0, [%2]\n"
	    		    "add %0, %0, %3\n"
			    "strex %1, %0, [%2]\n"
			    "cmp %1, #0\n"
	                    "it ne\n"
			    "bne	1b\n"
			    : "=&r" (tmp), "+r" (tmp2)
			    ,"+r" (p), "+r" (val) : : "cc", "memory");
}

static __inline void
atomic_add_64(volatile uint64_t *p, uint64_t val)
{
	uint64_t tmp;
	uint32_t exflag;

	__asm __volatile(
		"1:          \n"
		"   ldrexd   %[tmp], [%[ptr]]\n"
		"   adds     %Q[tmp], %Q[val]\n"
		"   adc      %R[tmp], %R[val]\n"
		"   strexd   %[exf], %[tmp], [%[ptr]]\n"
		"   teq      %[exf], #0\n"
		"   it ne    \n"
		"   bne      1b\n"
		:   [exf]    "=&r"  (exflag), 
		    [tmp]    "=&r"  (tmp)
		:   [ptr]    "r"    (p), 
		    [val]    "r"    (val)
		:   "cc", "memory");
}

static __inline void
atomic_add_long(volatile u_long *p, u_long val)
{
	u_long tmp = 0, tmp2 = 0;

	__asm __volatile("1: ldrex %0, [%2]\n"
	    		    "add %0, %0, %3\n"
			    "strex %1, %0, [%2]\n"
			    "cmp %1, #0\n"
	                    "it ne\n"
			    "bne	1b\n"
			    : "=&r" (tmp), "+r" (tmp2)
			    ,"+r" (p), "+r" (val) : : "cc", "memory");
}

static __inline void
atomic_subtract_32(volatile u_int32_t *p, u_int32_t val)
{
	uint32_t tmp = 0, tmp2 = 0;

	__asm __volatile("1: ldrex %0, [%2]\n"
	    		    "sub %0, %0, %3\n"
			    "strex %1, %0, [%2]\n"
			    "cmp %1, #0\n"
	                    "it ne\n"
			    "bne	1b\n"
			    : "=&r" (tmp), "+r" (tmp2)
			    ,"+r" (p), "+r" (val) : : "cc", "memory");
}

static __inline void
atomic_subtract_64(volatile uint64_t *p, uint64_t val)
{
	uint64_t tmp;
	uint32_t exflag;

	__asm __volatile(
		"1:          \n"
		"   ldrexd   %[tmp], [%[ptr]]\n"
		"   subs     %Q[tmp], %Q[val]\n"
		"   sbc      %R[tmp], %R[val]\n"
		"   strexd   %[exf], %[tmp], [%[ptr]]\n"
		"   teq      %[exf], #0\n"
		"   it ne    \n"
		"   bne      1b\n"
		:   [exf]    "=&r"  (exflag), 
		    [tmp]    "=&r"  (tmp)
		:   [ptr]    "r"    (p), 
		    [val]    "r"    (val)
		:   "cc", "memory");
}

static __inline void
atomic_subtract_long(volatile u_long *p, u_long val)
{
	u_long tmp = 0, tmp2 = 0;

	__asm __volatile("1: ldrex %0, [%2]\n"
	    		    "sub %0, %0, %3\n"
			    "strex %1, %0, [%2]\n"
			    "cmp %1, #0\n"
	                    "it ne\n"
			    "bne	1b\n"
			    : "=&r" (tmp), "+r" (tmp2)
			    ,"+r" (p), "+r" (val) : : "cc", "memory");
}

ATOMIC_ACQ_REL(clear, 32)
ATOMIC_ACQ_REL(add, 32)
ATOMIC_ACQ_REL(subtract, 32)
ATOMIC_ACQ_REL(set, 32)
ATOMIC_ACQ_REL(clear, 64)
ATOMIC_ACQ_REL(add, 64)
ATOMIC_ACQ_REL(subtract, 64)
ATOMIC_ACQ_REL(set, 64)
ATOMIC_ACQ_REL_LONG(clear)
ATOMIC_ACQ_REL_LONG(add)
ATOMIC_ACQ_REL_LONG(subtract)
ATOMIC_ACQ_REL_LONG(set)

#undef ATOMIC_ACQ_REL
#undef ATOMIC_ACQ_REL_LONG

static __inline uint32_t
atomic_fetchadd_32(volatile uint32_t *p, uint32_t val)
{
	uint32_t tmp = 0, tmp2 = 0, ret = 0;

	__asm __volatile("1: ldrex %0, [%3]\n"
	    		    "add %1, %0, %4\n"
			    "strex %2, %1, [%3]\n"
			    "cmp %2, #0\n"
	                    "it ne\n"
			    "bne	1b\n"
			   : "+r" (ret), "=&r" (tmp), "+r" (tmp2)
			   ,"+r" (p), "+r" (val) : : "cc", "memory");
	return (ret);
}

static __inline uint32_t
atomic_readandclear_32(volatile u_int32_t *p)
{
	uint32_t ret, tmp = 0, tmp2 = 0;

	__asm __volatile("1: ldrex %0, [%3]\n"
	    		 "mov %1, #0\n"
			 "strex %2, %1, [%3]\n"
			 "cmp %2, #0\n"
	                 "it ne\n"
			 "bne 1b\n"
			 : "=r" (ret), "=&r" (tmp), "+r" (tmp2)
			 ,"+r" (p) : : "cc", "memory");
	return (ret);
}

static __inline uint32_t
atomic_load_acq_32(volatile uint32_t *p)
{
	uint32_t v;

	v = *p;
	__do_dmb();
	return (v);
}

static __inline void
atomic_store_rel_32(volatile uint32_t *p, uint32_t v)
{
	
	__do_dmb();
	*p = v;
}

static __inline uint64_t
atomic_fetchadd_64(volatile uint64_t *p, uint64_t val)
{
	uint64_t ret, tmp;
	uint32_t exflag;

	__asm __volatile(
		"1:          \n"
		"   ldrexd   %[ret], [%[ptr]]\n"
		"   adds     %Q[tmp], %Q[ret], %Q[val]\n"
		"   adc      %R[tmp], %R[ret], %R[val]\n"
		"   strexd   %[exf], %[tmp], [%[ptr]]\n"
		"   teq      %[exf], #0\n"
		"   it ne    \n"
		"   bne      1b\n"
		:   [ret]    "=&r"  (ret),
		    [exf]    "=&r"  (exflag),
		    [tmp]    "=&r"  (tmp)
		:   [ptr]    "r"    (p), 
		    [val]    "r"    (val)
		:   "cc", "memory");
	return (ret);
}

static __inline uint64_t
atomic_readandclear_64(volatile uint64_t *p)
{
	uint64_t ret, tmp;
	uint32_t exflag;

	__asm __volatile(
		"1:          \n"
		"   ldrexd   %[ret], [%[ptr]]\n"
		"   mov      %Q[tmp], #0\n"
		"   mov      %R[tmp], #0\n"
		"   strexd   %[exf], %[tmp], [%[ptr]]\n"
		"   teq      %[exf], #0\n"
		"   it ne    \n"
		"   bne      1b\n"
		:   [ret]    "=&r"  (ret),
		    [exf]    "=&r"  (exflag),
		    [tmp]    "=&r"  (tmp)
		:   [ptr]    "r"    (p)
		:   "cc", "memory");
	return (ret);
}

static __inline uint64_t
atomic_load_64(volatile uint64_t *p)
{
	uint64_t ret;

	/*
	 * The only way to atomically load 64 bits is with LDREXD which puts the
	 * exclusive monitor into the exclusive state, so reset it to open state
	 * with CLREX because we don't actually need to store anything.
	 */
	__asm __volatile(
		"1:          \n"
		"   ldrexd   %[ret], [%[ptr]]\n"
		"   clrex    \n"
		:   [ret]    "=&r"  (ret)
		:   [ptr]    "r"    (p)
		:   "cc", "memory");
	return (ret);
}

static __inline uint64_t
atomic_load_acq_64(volatile uint64_t *p)
{
	uint64_t ret;

	ret = atomic_load_64(p);
	__do_dmb();
	return (ret);
}

static __inline void
atomic_store_64(volatile uint64_t *p, uint64_t val)
{
	uint64_t tmp;
	uint32_t exflag;

	/*
	 * The only way to atomically store 64 bits is with STREXD, which will
	 * succeed only if paired up with a preceeding LDREXD using the same
	 * address, so we read and discard the existing value before storing.
	 */
	__asm __volatile(
		"1:          \n"
		"   ldrexd   %[tmp], [%[ptr]]\n"
		"   strexd   %[exf], %[val], [%[ptr]]\n"
		"   teq      %[exf], #0\n"
		"   it ne    \n"
		"   bne      1b\n"
		:   [tmp]    "=&r"  (tmp),
		    [exf]    "=&r"  (exflag)
		:   [ptr]    "r"    (p),
		    [val]    "r"    (val)
		:   "cc", "memory");
}

static __inline void
atomic_store_rel_64(volatile uint64_t *p, uint64_t val)
{

	__do_dmb();
	atomic_store_64(p, val);
}

static __inline u_long
atomic_fetchadd_long(volatile u_long *p, u_long val)
{
	u_long tmp = 0, tmp2 = 0, ret = 0;

	__asm __volatile("1: ldrex %0, [%3]\n"
	    		    "add %1, %0, %4\n"
			    "strex %2, %1, [%3]\n"
			    "cmp %2, #0\n"
	                    "it ne\n"
			    "bne	1b\n"
			   : "+r" (ret), "=&r" (tmp), "+r" (tmp2)
			   ,"+r" (p), "+r" (val) : : "cc", "memory");
	return (ret);
}

static __inline u_long
atomic_readandclear_long(volatile u_long *p)
{
	u_long ret, tmp = 0, tmp2 = 0;

	__asm __volatile("1: ldrex %0, [%3]\n"
	    		 "mov %1, #0\n"
			 "strex %2, %1, [%3]\n"
			 "cmp %2, #0\n"
	                 "it ne\n"
			 "bne 1b\n"
			 : "=r" (ret), "=&r" (tmp), "+r" (tmp2)
			 ,"+r" (p) : : "cc", "memory");
	return (ret);
}

static __inline u_long
atomic_load_acq_long(volatile u_long *p)
{
	u_long v;

	v = *p;
	__do_dmb();
	return (v);
}

static __inline void
atomic_store_rel_long(volatile u_long *p, u_long v)
{
	
	__do_dmb();
	*p = v;
}
#else /* < armv6 */

#define __with_interrupts_disabled(expr) \
	do {						\
		u_int cpsr_save, tmp;			\
							\
		__asm __volatile(			\
			"mrs  %0, cpsr;"		\
			"orr  %1, %0, %2;"		\
			"msr  cpsr_fsxc, %1;"		\
			: "=r" (cpsr_save), "=r" (tmp)	\
			: "I" (PSR_I | PSR_F)		\
		        : "cc" );		\
		(expr);				\
		 __asm __volatile(		\
			"msr  cpsr_fsxc, %0"	\
			: /* no output */	\
			: "r" (cpsr_save)	\
			: "cc" );		\
	} while(0)

static __inline uint32_t
__swp(uint32_t val, volatile uint32_t *ptr)
{
	__asm __volatile("swp	%0, %2, [%3]"
	    : "=&r" (val), "=m" (*ptr)
	    : "r" (val), "r" (ptr), "m" (*ptr)
	    : "memory");
	return (val);
}


#ifdef _KERNEL
#define	ARM_HAVE_ATOMIC64

static __inline void
atomic_set_32(volatile uint32_t *address, uint32_t setmask)
{
	__with_interrupts_disabled(*address |= setmask);
}

static __inline void
atomic_set_64(volatile uint64_t *address, uint64_t setmask)
{
	__with_interrupts_disabled(*address |= setmask);
}

static __inline void
atomic_clear_32(volatile uint32_t *address, uint32_t clearmask)
{
	__with_interrupts_disabled(*address &= ~clearmask);
}

static __inline void
atomic_clear_64(volatile uint64_t *address, uint64_t clearmask)
{
	__with_interrupts_disabled(*address &= ~clearmask);
}

static __inline u_int32_t
atomic_cmpset_32(volatile u_int32_t *p, volatile u_int32_t cmpval, volatile u_int32_t newval)
{
	int ret;
	
	__with_interrupts_disabled(
	 {
	    	if (*p == cmpval) {
			*p = newval;
			ret = 1;
		} else {
			ret = 0;
		}
	});
	return (ret);
}

static __inline u_int64_t
atomic_cmpset_64(volatile u_int64_t *p, volatile u_int64_t cmpval, volatile u_int64_t newval)
{
	int ret;
	
	__with_interrupts_disabled(
	 {
	    	if (*p == cmpval) {
			*p = newval;
			ret = 1;
		} else {
			ret = 0;
		}
	});
	return (ret);
}

static __inline void
atomic_add_32(volatile u_int32_t *p, u_int32_t val)
{
	__with_interrupts_disabled(*p += val);
}

static __inline void
atomic_add_64(volatile u_int64_t *p, u_int64_t val)
{
	__with_interrupts_disabled(*p += val);
}

static __inline void
atomic_subtract_32(volatile u_int32_t *p, u_int32_t val)
{
	__with_interrupts_disabled(*p -= val);
}

static __inline void
atomic_subtract_64(volatile u_int64_t *p, u_int64_t val)
{
	__with_interrupts_disabled(*p -= val);
}

static __inline uint32_t
atomic_fetchadd_32(volatile uint32_t *p, uint32_t v)
{
	uint32_t value;

	__with_interrupts_disabled(
	{
	    	value = *p;
		*p += v;
	});
	return (value);
}

static __inline uint64_t
atomic_fetchadd_64(volatile uint64_t *p, uint64_t v)
{
	uint64_t value;

	__with_interrupts_disabled(
	{
	    	value = *p;
		*p += v;
	});
	return (value);
}

static __inline uint64_t
atomic_load_64(volatile uint64_t *p)
{
	uint64_t value;

	__with_interrupts_disabled(value = *p);
	return (value);
}

static __inline void
atomic_store_64(volatile uint64_t *p, uint64_t value)
{
	__with_interrupts_disabled(*p = value);
}

#else /* !_KERNEL */

static __inline u_int32_t
atomic_cmpset_32(volatile u_int32_t *p, volatile u_int32_t cmpval, volatile u_int32_t newval)
{
	register int done, ras_start = ARM_RAS_START;

	__asm __volatile("1:\n"
	    "adr	%1, 1b\n"
	    "str	%1, [%0]\n"
	    "adr	%1, 2f\n"
	    "str	%1, [%0, #4]\n"
	    "ldr	%1, [%2]\n"
	    "cmp	%1, %3\n"
	    "streq	%4, [%2]\n"
	    "2:\n"
	    "mov	%1, #0\n"
	    "str	%1, [%0]\n"
	    "mov	%1, #0xffffffff\n"
	    "str	%1, [%0, #4]\n"
	    "moveq	%1, #1\n"
	    "movne	%1, #0\n"
	    : "+r" (ras_start), "=r" (done)
	    ,"+r" (p), "+r" (cmpval), "+r" (newval) : : "cc", "memory");
	return (done);
}

static __inline void
atomic_add_32(volatile u_int32_t *p, u_int32_t val)
{
	int start, ras_start = ARM_RAS_START;

	__asm __volatile("1:\n"
	    "adr	%1, 1b\n"
	    "str	%1, [%0]\n"
	    "adr	%1, 2f\n"
	    "str	%1, [%0, #4]\n"
	    "ldr	%1, [%2]\n"
	    "add	%1, %1, %3\n"
	    "str	%1, [%2]\n"
	    "2:\n"
	    "mov	%1, #0\n"
	    "str	%1, [%0]\n"
	    "mov	%1, #0xffffffff\n"
	    "str	%1, [%0, #4]\n"
	    : "+r" (ras_start), "=r" (start), "+r" (p), "+r" (val)
	    : : "memory");
}

static __inline void
atomic_subtract_32(volatile u_int32_t *p, u_int32_t val)
{
	int start, ras_start = ARM_RAS_START;

	__asm __volatile("1:\n"
	    "adr	%1, 1b\n"
	    "str	%1, [%0]\n"
	    "adr	%1, 2f\n"
	    "str	%1, [%0, #4]\n"
	    "ldr	%1, [%2]\n"
	    "sub	%1, %1, %3\n"
	    "str	%1, [%2]\n"
	    "2:\n"
	    "mov	%1, #0\n"
	    "str	%1, [%0]\n"
	    "mov	%1, #0xffffffff\n"
	    "str	%1, [%0, #4]\n"

	    : "+r" (ras_start), "=r" (start), "+r" (p), "+r" (val)
	    : : "memory");
}

static __inline void
atomic_set_32(volatile uint32_t *address, uint32_t setmask)
{
	int start, ras_start = ARM_RAS_START;

	__asm __volatile("1:\n"
	    "adr	%1, 1b\n"
	    "str	%1, [%0]\n"
	    "adr	%1, 2f\n"
	    "str	%1, [%0, #4]\n"
	    "ldr	%1, [%2]\n"
	    "orr	%1, %1, %3\n"
	    "str	%1, [%2]\n"
	    "2:\n"
	    "mov	%1, #0\n"
	    "str	%1, [%0]\n"
	    "mov	%1, #0xffffffff\n"
	    "str	%1, [%0, #4]\n"

	    : "+r" (ras_start), "=r" (start), "+r" (address), "+r" (setmask)
	    : : "memory");
}

static __inline void
atomic_clear_32(volatile uint32_t *address, uint32_t clearmask)
{
	int start, ras_start = ARM_RAS_START;

	__asm __volatile("1:\n"
	    "adr	%1, 1b\n"
	    "str	%1, [%0]\n"
	    "adr	%1, 2f\n"
	    "str	%1, [%0, #4]\n"
	    "ldr	%1, [%2]\n"
	    "bic	%1, %1, %3\n"
	    "str	%1, [%2]\n"
	    "2:\n"
	    "mov	%1, #0\n"
	    "str	%1, [%0]\n"
	    "mov	%1, #0xffffffff\n"
	    "str	%1, [%0, #4]\n"
	    : "+r" (ras_start), "=r" (start), "+r" (address), "+r" (clearmask)
	    : : "memory");

}

static __inline uint32_t
atomic_fetchadd_32(volatile uint32_t *p, uint32_t v)
{
	uint32_t start, tmp, ras_start = ARM_RAS_START;

	__asm __volatile("1:\n"
	    "adr	%1, 1b\n"
	    "str	%1, [%0]\n"
	    "adr	%1, 2f\n"
	    "str	%1, [%0, #4]\n"
	    "ldr	%1, [%3]\n"
	    "mov	%2, %1\n"
	    "add	%2, %2, %4\n"
	    "str	%2, [%3]\n"
	    "2:\n"
	    "mov	%2, #0\n"
	    "str	%2, [%0]\n"
	    "mov	%2, #0xffffffff\n"
	    "str	%2, [%0, #4]\n"
	    : "+r" (ras_start), "=r" (start), "=r" (tmp), "+r" (p), "+r" (v)
	    : : "memory");
	return (start);
}

#endif /* _KERNEL */


static __inline uint32_t
atomic_readandclear_32(volatile u_int32_t *p)
{

	return (__swp(0, p));
}

#define atomic_cmpset_rel_32	atomic_cmpset_32
#define atomic_cmpset_acq_32	atomic_cmpset_32
#define atomic_set_rel_32	atomic_set_32
#define atomic_set_acq_32	atomic_set_32
#define atomic_clear_rel_32	atomic_clear_32
#define atomic_clear_acq_32	atomic_clear_32
#define atomic_add_rel_32	atomic_add_32
#define atomic_add_acq_32	atomic_add_32
#define atomic_subtract_rel_32	atomic_subtract_32
#define atomic_subtract_acq_32	atomic_subtract_32
#define atomic_store_rel_32	atomic_store_32
#define atomic_store_rel_long	atomic_store_long
#define atomic_load_acq_32	atomic_load_32
#define atomic_load_acq_long	atomic_load_long
#define atomic_add_acq_long		atomic_add_long
#define atomic_add_rel_long		atomic_add_long
#define atomic_subtract_acq_long	atomic_subtract_long
#define atomic_subtract_rel_long	atomic_subtract_long
#define atomic_clear_acq_long		atomic_clear_long
#define atomic_clear_rel_long		atomic_clear_long
#define atomic_set_acq_long		atomic_set_long
#define atomic_set_rel_long		atomic_set_long
#define atomic_cmpset_acq_long		atomic_cmpset_long
#define atomic_cmpset_rel_long		atomic_cmpset_long
#define atomic_load_acq_long		atomic_load_long
#undef __with_interrupts_disabled

static __inline void
atomic_add_long(volatile u_long *p, u_long v)
{

	atomic_add_32((volatile uint32_t *)p, v);
}

static __inline void
atomic_clear_long(volatile u_long *p, u_long v)
{

	atomic_clear_32((volatile uint32_t *)p, v);
}

static __inline int
atomic_cmpset_long(volatile u_long *dst, u_long old, u_long newe)
{

	return (atomic_cmpset_32((volatile uint32_t *)dst, old, newe));
}

static __inline u_long
atomic_fetchadd_long(volatile u_long *p, u_long v)
{

	return (atomic_fetchadd_32((volatile uint32_t *)p, v));
}

static __inline void
atomic_readandclear_long(volatile u_long *p)
{

	atomic_readandclear_32((volatile uint32_t *)p);
}

static __inline void
atomic_set_long(volatile u_long *p, u_long v)
{

	atomic_set_32((volatile uint32_t *)p, v);
}

static __inline void
atomic_subtract_long(volatile u_long *p, u_long v)
{

	atomic_subtract_32((volatile uint32_t *)p, v);
}



#endif /* Arch >= v6 */

static __inline int
atomic_load_32(volatile uint32_t *v)
{

	return (*v);
}

static __inline void
atomic_store_32(volatile uint32_t *dst, uint32_t src)
{
	*dst = src;
}

static __inline int
atomic_load_long(volatile u_long *v)
{

	return (*v);
}

static __inline void
atomic_store_long(volatile u_long *dst, u_long src)
{
	*dst = src;
}

#define atomic_clear_ptr		atomic_clear_32
#define atomic_set_ptr			atomic_set_32
#define atomic_cmpset_ptr		atomic_cmpset_32
#define atomic_cmpset_rel_ptr		atomic_cmpset_rel_32
#define atomic_cmpset_acq_ptr		atomic_cmpset_acq_32
#define atomic_store_ptr		atomic_store_32
#define atomic_store_rel_ptr		atomic_store_rel_32

#define atomic_add_int			atomic_add_32
#define atomic_add_acq_int		atomic_add_acq_32
#define atomic_add_rel_int		atomic_add_rel_32
#define atomic_subtract_int		atomic_subtract_32
#define atomic_subtract_acq_int		atomic_subtract_acq_32
#define atomic_subtract_rel_int		atomic_subtract_rel_32
#define atomic_clear_int		atomic_clear_32
#define atomic_clear_acq_int		atomic_clear_acq_32
#define atomic_clear_rel_int		atomic_clear_rel_32
#define atomic_set_int			atomic_set_32
#define atomic_set_acq_int		atomic_set_acq_32
#define atomic_set_rel_int		atomic_set_rel_32
#define atomic_cmpset_int		atomic_cmpset_32
#define atomic_cmpset_acq_int		atomic_cmpset_acq_32
#define atomic_cmpset_rel_int		atomic_cmpset_rel_32
#define atomic_fetchadd_int		atomic_fetchadd_32
#define atomic_readandclear_int		atomic_readandclear_32
#define atomic_load_acq_int		atomic_load_acq_32
#define atomic_store_rel_int		atomic_store_rel_32

#endif /* _MACHINE_ATOMIC_H_ */
OpenPOWER on IntegriCloud