about summary refs log tree commit diff
path: root/sysdeps/unix/sysv/linux/sh/atomic-machine.h
blob: 89993f1670d8f61762cd7451060934cd8f3309a4 (plain) (blame)
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
/* Atomic operations used inside libc.  Linux/SH version.
   Copyright (C) 2003-2023 Free Software Foundation, Inc.
   This file is part of the GNU C Library.

   The GNU C Library is free software; you can redistribute it and/or
   modify it under the terms of the GNU Lesser General Public
   License as published by the Free Software Foundation; either
   version 2.1 of the License, or (at your option) any later version.

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

   You should have received a copy of the GNU Lesser General Public
   License along with the GNU C Library; if not, see
   <https://www.gnu.org/licenses/>.  */

#define __HAVE_64B_ATOMICS 0
#define USE_ATOMIC_COMPILER_BUILTINS 0

/* XXX Is this actually correct?  */
#define ATOMIC_EXCHANGE_USES_CAS 1

/* SH kernel has implemented a gUSA ("g" User Space Atomicity) support
   for the user space atomicity. The atomicity macros use this scheme.

  Reference:
    Niibe Yutaka, "gUSA: Simple and Efficient User Space Atomicity
    Emulation with Little Kernel Modification", Linux Conference 2002,
    Japan. http://lc.linux.or.jp/lc2002/papers/niibe0919h.pdf (in
    Japanese).

    B.N. Bershad, D. Redell, and J. Ellis, "Fast Mutual Exclusion for
    Uniprocessors",  Proceedings of the Fifth Architectural Support for
    Programming Languages and Operating Systems (ASPLOS), pp. 223-233,
    October 1992. http://www.cs.washington.edu/homes/bershad/Papers/Rcs.ps

  SuperH ABI:
      r15:    -(size of atomic instruction sequence) < 0
      r0:     end point
      r1:     saved stack pointer
*/

#define __arch_compare_and_exchange_val_8_acq(mem, newval, oldval) \
  ({ __typeof (*(mem)) __result; \
     __asm __volatile ("\
	mova 1f,r0\n\
	.align 2\n\
	mov r15,r1\n\
	mov #(0f-1f),r15\n\
     0: mov.b @%1,%0\n\
	cmp/eq %0,%3\n\
	bf 1f\n\
	mov.b %2,@%1\n\
     1: mov r1,r15"\
	: "=&r" (__result) : "u" (mem), "u" (newval), "u" (oldval) \
	: "r0", "r1", "t", "memory"); \
     __result; })

#define __arch_compare_and_exchange_val_16_acq(mem, newval, oldval) \
  ({ __typeof (*(mem)) __result; \
     __asm __volatile ("\
	mova 1f,r0\n\
	mov r15,r1\n\
	.align 2\n\
	mov #(0f-1f),r15\n\
	mov #-8,r15\n\
     0: mov.w @%1,%0\n\
	cmp/eq %0,%3\n\
	bf 1f\n\
	mov.w %2,@%1\n\
     1: mov r1,r15"\
	: "=&r" (__result) : "u" (mem), "u" (newval), "u" (oldval) \
	: "r0", "r1", "t", "memory"); \
     __result; })

#define __arch_compare_and_exchange_val_32_acq(mem, newval, oldval) \
  ({ __typeof (*(mem)) __result; \
     __asm __volatile ("\
	mova 1f,r0\n\
	.align 2\n\
	mov r15,r1\n\
	mov #(0f-1f),r15\n\
     0: mov.l @%1,%0\n\
	cmp/eq %0,%3\n\
	bf 1f\n\
	mov.l %2,@%1\n\
     1: mov r1,r15"\
	: "=&r" (__result) : "u" (mem), "u" (newval), "u" (oldval) \
	: "r0", "r1", "t", "memory"); \
     __result; })

/* XXX We do not really need 64-bit compare-and-exchange.  At least
   not in the moment.  Using it would mean causing portability
   problems since not many other 32-bit architectures have support for
   such an operation.  So don't define any code for now.  */

# define __arch_compare_and_exchange_val_64_acq(mem, newval, oldval) \
  (abort (), (__typeof (*mem)) 0)

#define atomic_exchange_and_add(mem, value) \
  ({ __typeof (*(mem)) __result, __tmp, __value = (value); \
     if (sizeof (*(mem)) == 1) \
       __asm __volatile ("\
	  mova 1f,r0\n\
	  .align 2\n\
	  mov r15,r1\n\
	  mov #(0f-1f),r15\n\
       0: mov.b @%2,%0\n\
	  mov %1,r2\n\
	  add %0,r2\n\
	  mov.b r2,@%2\n\
       1: mov r1,r15"\
	: "=&r" (__result), "=&r" (__tmp) : "u" (mem), "1" (__value) \
	: "r0", "r1", "r2", "memory");		       \
     else if (sizeof (*(mem)) == 2) \
       __asm __volatile ("\
	  mova 1f,r0\n\
	  .align 2\n\
	  mov r15,r1\n\
	  mov #(0f-1f),r15\n\
       0: mov.w @%2,%0\n\
	  mov %1,r2\n\
	  add %0,r2\n\
	  mov.w r2,@%2\n\
       1: mov r1,r15"\
	: "=&r" (__result), "=&r" (__tmp) : "u" (mem), "1" (__value) \
	: "r0", "r1", "r2", "memory"); \
     else if (sizeof (*(mem)) == 4) \
       __asm __volatile ("\
	  mova 1f,r0\n\
	  .align 2\n\
	  mov r15,r1\n\
	  mov #(0f-1f),r15\n\
       0: mov.l @%2,%0\n\
	  mov %1,r2\n\
	  add %0,r2\n\
	  mov.l r2,@%2\n\
       1: mov r1,r15"\
	: "=&r" (__result), "=&r" (__tmp) : "u" (mem), "1" (__value) \
	: "r0", "r1", "r2", "memory"); \
     else \
       { \
	 __typeof (mem) memp = (mem); \
	 do \
	   __result = *memp; \
	 while (__arch_compare_and_exchange_val_64_acq \
		 (memp,	__result + __value, __result) == __result); \
	 (void) __value; \
       } \
     __result; })

#define atomic_add(mem, value) \
  (void) ({ __typeof (*(mem)) __tmp, __value = (value); \
	    if (sizeof (*(mem)) == 1) \
	      __asm __volatile ("\
		mova 1f,r0\n\
		mov r15,r1\n\
		.align 2\n\
		mov #(0f-1f),r15\n\
	     0: mov.b @%1,r2\n\
		add %0,r2\n\
		mov.b r2,@%1\n\
	     1: mov r1,r15"\
		: "=&r" (__tmp) : "u" (mem), "0" (__value) \
		: "r0", "r1", "r2", "memory"); \
	    else if (sizeof (*(mem)) == 2) \
	      __asm __volatile ("\
		mova 1f,r0\n\
		mov r15,r1\n\
		.align 2\n\
		mov #(0f-1f),r15\n\
	     0: mov.w @%1,r2\n\
		add %0,r2\n\
		mov.w r2,@%1\n\
	     1: mov r1,r15"\
		: "=&r" (__tmp) : "u" (mem), "0" (__value) \
		: "r0", "r1", "r2", "memory"); \
	    else if (sizeof (*(mem)) == 4) \
	      __asm __volatile ("\
		mova 1f,r0\n\
		mov r15,r1\n\
		.align 2\n\
		mov #(0f-1f),r15\n\
	     0: mov.l @%1,r2\n\
		add %0,r2\n\
		mov.l r2,@%1\n\
	     1: mov r1,r15"\
		: "=&r" (__tmp) : "u" (mem), "0" (__value) \
		: "r0", "r1", "r2", "memory"); \
	    else \
	      { \
		__typeof (*(mem)) oldval; \
		__typeof (mem) memp = (mem); \
		do \
		  oldval = *memp; \
		while (__arch_compare_and_exchange_val_64_acq \
			(memp, oldval + __value, oldval) == oldval); \
		(void) __value; \
	      } \
	    })

#define atomic_add_negative(mem, value) \
  ({ unsigned char __result; \
     __typeof (*(mem)) __tmp, __value = (value); \
     if (sizeof (*(mem)) == 1) \
       __asm __volatile ("\
	  mova 1f,r0\n\
	  mov r15,r1\n\
	  .align 2\n\
	  mov #(0f-1f),r15\n\
       0: mov.b @%2,r2\n\
	  add %1,r2\n\
	  mov.b r2,@%2\n\
       1: mov r1,r15\n\
	  shal r2\n\
	  movt %0"\
	: "=r" (__result), "=&r" (__tmp) : "u" (mem), "1" (__value) \
	: "r0", "r1", "r2", "t", "memory"); \
     else if (sizeof (*(mem)) == 2) \
       __asm __volatile ("\
	  mova 1f,r0\n\
	  mov r15,r1\n\
	  .align 2\n\
	  mov #(0f-1f),r15\n\
       0: mov.w @%2,r2\n\
	  add %1,r2\n\
	  mov.w r2,@%2\n\
       1: mov r1,r15\n\
	  shal r2\n\
	  movt %0"\
	: "=r" (__result), "=&r" (__tmp) : "u" (mem), "1" (__value) \
	: "r0", "r1", "r2", "t", "memory"); \
     else if (sizeof (*(mem)) == 4) \
       __asm __volatile ("\
	  mova 1f,r0\n\
	  mov r15,r1\n\
	  .align 2\n\
	  mov #(0f-1f),r15\n\
       0: mov.l @%2,r2\n\
	  add %1,r2\n\
	  mov.l r2,@%2\n\
       1: mov r1,r15\n\
	  shal r2\n\
	  movt %0"\
	: "=r" (__result), "=&r" (__tmp) : "u" (mem), "1" (__value) \
	: "r0", "r1", "r2", "t", "memory"); \
     else \
       abort (); \
     __result; })

#define atomic_add_zero(mem, value) \
  ({ unsigned char __result; \
     __typeof (*(mem)) __tmp, __value = (value); \
     if (sizeof (*(mem)) == 1) \
       __asm __volatile ("\
	  mova 1f,r0\n\
	  mov r15,r1\n\
	  .align 2\n\
	  mov #(0f-1f),r15\n\
       0: mov.b @%2,r2\n\
	  add %1,r2\n\
	  mov.b r2,@%2\n\
       1: mov r1,r15\n\
	  tst r2,r2\n\
	  movt %0"\
	: "=r" (__result), "=&r" (__tmp) : "u" (mem), "1" (__value) \
	: "r0", "r1", "r2", "t", "memory"); \
     else if (sizeof (*(mem)) == 2) \
       __asm __volatile ("\
	  mova 1f,r0\n\
	  mov r15,r1\n\
	  .align 2\n\
	  mov #(0f-1f),r15\n\
       0: mov.w @%2,r2\n\
	  add %1,r2\n\
	  mov.w r2,@%2\n\
       1: mov r1,r15\n\
	  tst r2,r2\n\
	  movt %0"\
	: "=r" (__result), "=&r" (__tmp) : "u" (mem), "1" (__value) \
	: "r0", "r1", "r2", "t", "memory"); \
     else if (sizeof (*(mem)) == 4) \
       __asm __volatile ("\
	  mova 1f,r0\n\
	  mov r15,r1\n\
	  .align 2\n\
	  mov #(0f-1f),r15\n\
       0: mov.l @%2,r2\n\
	  add %1,r2\n\
	  mov.l r2,@%2\n\
       1: mov r1,r15\n\
	  tst r2,r2\n\
	  movt %0"\
	: "=r" (__result), "=&r" (__tmp) : "u" (mem), "1" (__value) \
	: "r0", "r1", "r2", "t", "memory"); \
     else \
       abort (); \
     __result; })

#define atomic_increment_and_test(mem) atomic_add_zero((mem), 1)
#define atomic_decrement_and_test(mem) atomic_add_zero((mem), -1)

#define atomic_bit_set(mem, bit) \
  (void) ({ unsigned int __mask = 1 << (bit); \
	    if (sizeof (*(mem)) == 1) \
	      __asm __volatile ("\
		mova 1f,r0\n\
		mov r15,r1\n\
		.align 2\n\
		mov #(0f-1f),r15\n\
	     0: mov.b @%0,r2\n\
		or %1,r2\n\
		mov.b r2,@%0\n\
	     1: mov r1,r15"\
		: : "u" (mem), "u" (__mask) \
		: "r0", "r1", "r2", "memory"); \
	    else if (sizeof (*(mem)) == 2) \
	      __asm __volatile ("\
		mova 1f,r0\n\
		mov r15,r1\n\
		.align 2\n\
		mov #(0f-1f),r15\n\
	     0: mov.w @%0,r2\n\
		or %1,r2\n\
		mov.w r2,@%0\n\
	     1: mov r1,r15"\
		: : "u" (mem), "u" (__mask) \
		: "r0", "r1", "r2", "memory"); \
	    else if (sizeof (*(mem)) == 4) \
	      __asm __volatile ("\
		mova 1f,r0\n\
		mov r15,r1\n\
		.align 2\n\
		mov #(0f-1f),r15\n\
	     0: mov.l @%0,r2\n\
		or %1,r2\n\
		mov.l r2,@%0\n\
	     1: mov r1,r15"\
		: : "u" (mem), "u" (__mask) \
		: "r0", "r1", "r2", "memory"); \
	    else \
	      abort (); \
	    })

#define atomic_bit_test_set(mem, bit) \
  ({ unsigned int __mask = 1 << (bit); \
     unsigned int __result = __mask; \
     if (sizeof (*(mem)) == 1) \
       __asm __volatile ("\
	  mova 1f,r0\n\
	  .align 2\n\
	  mov r15,r1\n\
	  mov #(0f-1f),r15\n\
       0: mov.b @%2,r2\n\
	  mov r2,r3\n\
	  or %1,r2\n\
	  mov.b r2,@%2\n\
       1: mov r1,r15\n\
	  and r3,%0"\
	: "=&r" (__result), "=&r" (__mask) \
	: "u" (mem), "0" (__result), "1" (__mask) \
	: "r0", "r1", "r2", "r3", "memory");	\
     else if (sizeof (*(mem)) == 2) \
       __asm __volatile ("\
	  mova 1f,r0\n\
	  .align 2\n\
	  mov r15,r1\n\
	  mov #(0f-1f),r15\n\
       0: mov.w @%2,r2\n\
	  mov r2,r3\n\
	  or %1,r2\n\
	  mov.w %1,@%2\n\
       1: mov r1,r15\n\
	  and r3,%0"\
	: "=&r" (__result), "=&r" (__mask) \
	: "u" (mem), "0" (__result), "1" (__mask) \
	: "r0", "r1", "r2", "r3", "memory"); \
     else if (sizeof (*(mem)) == 4) \
       __asm __volatile ("\
	  mova 1f,r0\n\
	  .align 2\n\
	  mov r15,r1\n\
	  mov #(0f-1f),r15\n\
       0: mov.l @%2,r2\n\
	  mov r2,r3\n\
	  or r2,%1\n\
	  mov.l %1,@%2\n\
       1: mov r1,r15\n\
	  and r3,%0"\
	: "=&r" (__result), "=&r" (__mask) \
	: "u" (mem), "0" (__result), "1" (__mask) \
	: "r0", "r1", "r2", "r3", "memory"); \
     else \
       abort (); \
     __result; })