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
|
/***************************************************************************
* __________ __ ___.
* Open \______ \ ____ ____ | | _\_ |__ _______ ___
* Source | _// _ \_/ ___\| |/ /| __ \ / _ \ \/ /
* Jukebox | | ( <_> ) \___| < | \_\ ( <_> > < <
* Firmware |____|_ /\____/ \___ >__|_ \|___ /\____/__/\_ \
* \/ \/ \/ \/ \/
* $Id$
*
* Copyright (C) 2006,2007 by Greg White
*
* This program is free software; you can redistribute it and/or
* modify it under the terms of the GNU General Public License
* as published by the Free Software Foundation; either version 2
* of the License, or (at your option) any later version.
*
* This software is distributed on an "AS IS" basis, WITHOUT WARRANTY OF ANY
* KIND, either express or implied.
*
****************************************************************************/
#include "config.h"
#include "cpu.h"
/* Used by ARMv4 & ARMv5 CPUs with cp15 register and MMU */
#if CONFIG_CPU == TCC7801 || CONFIG_CPU == AT91SAM9260
/* MMU present but unused */
#define HAVE_TEST_AND_CLEAN_CACHE
#elif CONFIG_CPU == DM320 || CONFIG_CPU == AS3525v2
#define USE_MMU
#define HAVE_TEST_AND_CLEAN_CACHE
#elif CONFIG_CPU == AS3525
#define USE_MMU
#define CACHE_SIZE 8
#elif CONFIG_CPU == S3C2440
#define USE_MMU
#define CACHE_SIZE 16
#elif CONFIG_CPU == S5L8701
/* MMU not present */
#define CACHE_SIZE 4
#else
#error Cache settings unknown for this CPU !
#endif /* CPU specific configuration */
@ Index format: 31:26 = index, N:5 = segment, remainder = SBZ
@ assume 64-way set associative separate I/D caches
@ CACHE_SIZE = N (kB) = N*2^10 B
@ number of lines = N*2^(10-CACHEALIGN_BITS)
@ Index bits = 6
@ Segment loops = N*2^(10-CACHEALIGN_BITS-6) = N*2^(4-CACHEALIGN_BITS)
@ Segment loops = N/2^(CACHEALIGN_BITS - 4)
@ Segment loops = N/(1<<(CACHEALIGN_BITS - 4))
#ifdef CACHE_SIZE
#if CACHEALIGN_BITS == 4
#define INDEX_STEPS CACHE_SIZE
#elif CACHEALIGN_BITS == 5
#define INDEX_STEPS (CACHE_SIZE/2)
#endif /* CACHEALIGN_BITS */
@ assume 64-way set associative separate I/D caches (log2(64) == 6)
@ Index format: 31:26 = index, M:N = segment, remainder = SBZ
@ Segment bits = log2(cache size in bytes / cache line size in byte) - Index bits (== 6)
@ N = CACHEALIGN_BITS
#endif /* CACHE_SIZE */
#ifdef USE_MMU
/** MMU setup **/
/*
* void ttb_init(void);
*/
.section .text, "ax", %progbits
.align 2
.global ttb_init
.type ttb_init, %function
ttb_init:
ldr r0, =TTB_BASE_ADDR @
mvn r1, #0 @
mcr p15, 0, r0, c2, c0, 0 @ Set the TTB base address
mcr p15, 0, r1, c3, c0, 0 @ Set all domains to manager status
bx lr @
.size ttb_init, .-ttb_init
/*
* void map_section(unsigned int pa, unsigned int va, int mb, int flags);
*/
.section .text, "ax", %progbits
.align 2
.global map_section
.type map_section, %function
map_section:
@ align to 1MB
@ pa &= (-1 << 20);
mov r0, r0, lsr #20
mov r0, r0, lsl #20
@ pa |= (flags | 0x412);
@ bit breakdown:
@ 10: superuser - r/w, user - no access
@ 4: should be "1"
@ 3,2: Cache flags (flags (r3))
@ 1: Section signature
orr r0, r0, r3
orr r0, r0, #0x410
orr r0, r0, #0x2
@ unsigned int* ttbPtr = TTB_BASE + (va >> 20);
@ sections are 1MB size
mov r1, r1, lsr #20
ldr r3, =TTB_BASE_ADDR
add r1, r3, r1, lsl #0x2
@ Add MB to pa, flags are already present in pa, but addition
@ should not effect them
@
@ for( ; mb>0; mb--, pa += (1 << 20))
@ {
@ *(ttbPtr++) = pa;
@ }
cmp r2, #0
bxle lr
mov r3, #0x0
1: @ loop
str r0, [r1], #4
add r0, r0, #0x100000
add r3, r3, #0x1
cmp r2, r3
bne 1b @ loop
bx lr
.size map_section, .-map_section
/*
* void enable_mmu(void);
*/
.section .text, "ax", %progbits
.align 2
.global enable_mmu
.type enable_mmu, %function
enable_mmu:
mov r0, #0 @
mcr p15, 0, r0, c8, c7, 0 @ invalidate TLB
mcr p15, 0, r0, c7, c7,0 @ invalidate both i and dcache
mrc p15, 0, r0, c1, c0, 0 @
orr r0, r0, #1 @ enable mmu bit, i and dcache
orr r0, r0, #1<<2 @ enable dcache
orr r0, r0, #1<<12 @ enable icache
mcr p15, 0, r0, c1, c0, 0 @
nop @
nop @
nop @
nop @
bx lr @
.size enable_mmu, .-enable_mmu
.ltorg
#endif /* USE_MMU */
/** Cache coherency **/
/*
* Invalidate DCache for this range
* will do write back
* void invalidate_dcache_range(const void *base, unsigned int size);
*/
.section .text, "ax", %progbits
.align 2
.global invalidate_dcache_range
.type invalidate_dcache_range, %function
@ MVA format: 31:5 = Modified virtual address, 4:0 = SBZ
invalidate_dcache_range:
add r1, r0, r1 @ size -> end
cmp r1, r0 @ end <= start?
bxls lr @
bic r0, r0, #31 @ Align start to cache line (down)
1: @ inv_start @
mcr p15, 0, r0, c7, c14, 1 @ Clean and invalidate line by MVA
add r0, r0, #32 @
cmp r1, r0 @
mcrhi p15, 0, r0, c7, c14, 1 @ Clean and invalidate line by MVA
addhi r0, r0, #32 @
cmphi r1, r0 @
mcrhi p15, 0, r0, c7, c14, 1 @ Clean and invalidate line by MVA
addhi r0, r0, #32 @
cmphi r1, r0 @
mcrhi p15, 0, r0, c7, c14, 1 @ Clean and invalidate line by MVA
addhi r0, r0, #32 @
cmphi r1, r0 @
mcrhi p15, 0, r0, c7, c14, 1 @ Clean and invalidate line by MVA
addhi r0, r0, #32 @
cmphi r1, r0 @
mcrhi p15, 0, r0, c7, c14, 1 @ Clean and invalidate line by MVA
addhi r0, r0, #32 @
cmphi r1, r0 @
mcrhi p15, 0, r0, c7, c14, 1 @ Clean and invalidate line by MVA
addhi r0, r0, #32 @
cmphi r1, r0 @
mcrhi p15, 0, r0, c7, c14, 1 @ Clean and invalidate line by MVA
addhi r0, r0, #32 @
cmphi r1, r0 @
bhi 1b @ inv_start @
mov r0, #0 @
mcr p15, 0, r0, c7, c10, 4 @ Drain write buffer
bx lr @
.size invalidate_dcache_range, .-invalidate_dcache_range
/*
* clean DCache for this range
* forces DCache writeback for the specified range
* void clean_dcache_range(const void *base, unsigned int size);
*/
.section .text, "ax", %progbits
.align 2
.global clean_dcache_range
.type clean_dcache_range, %function
@ MVA format: 31:5 = Modified virtual address, 4:0 = SBZ
clean_dcache_range:
add r1, r0, r1 @ size -> end
cmp r1, r0 @ end <= start?
bxls lr @
bic r0, r0, #31 @ Align start to cache line (down)
1: @ clean_start @
mcr p15, 0, r0, c7, c10, 1 @ Clean line by MVA
add r0, r0, #32 @
cmp r1, r0 @
mcrhi p15, 0, r0, c7, c10, 1 @ Clean line by MVA
addhi r0, r0, #32 @
cmphi r1, r0 @
mcrhi p15, 0, r0, c7, c10, 1 @ Clean line by MVA
addhi r0, r0, #32 @
cmphi r1, r0 @
mcrhi p15, 0, r0, c7, c10, 1 @ Clean line by MVA
addhi r0, r0, #32 @
cmphi r1, r0 @
mcrhi p15, 0, r0, c7, c10, 1 @ Clean line by MVA
addhi r0, r0, #32 @
cmphi r1, r0 @
mcrhi p15, 0, r0, c7, c10, 1 @ Clean line by MVA
addhi r0, r0, #32 @
cmphi r1, r0 @
mcrhi p15, 0, r0, c7, c10, 1 @ Clean line by MVA
addhi r0, r0, #32 @
cmphi r1, r0 @
mcrhi p15, 0, r0, c7, c10, 1 @ Clean line by MVA
addhi r0, r0, #32 @
cmphi r1, r0 @
bhi 1b @clean_start @
mov r0, #0 @
mcr p15, 0, r0, c7, c10, 4 @ Drain write buffer
bx lr @
.size clean_dcache_range, .-clean_dcache_range
/*
* Dump DCache for this range
* will *NOT* do write back except for buffer edges not on a line boundary
* void dump_dcache_range(const void *base, unsigned int size);
*/
.section .text, "ax", %progbits
.align 2
.global dump_dcache_range
.type dump_dcache_range, %function
@ MVA format: 31:5 = Modified virtual address, 4:0 = SBZ
dump_dcache_range:
add r1, r0, r1 @ size -> end
cmp r1, r0 @ end <= start?
bxls lr @
tst r0, #31 @ Check first line for bits set
bicne r0, r0, #31 @ Clear low five bits (down)
mcrne p15, 0, r0, c7, c14, 1 @ Clean and invalidate line by MVA
@ if not cache aligned
addne r0, r0, #32 @ Move to the next cache line
@
tst r1, #31 @ Check last line for bits set
bicne r1, r1, #31 @ Clear low five bits (down)
mcrne p15, 0, r1, c7, c14, 1 @ Clean and invalidate line by MVA
@ if not cache aligned
cmp r1, r0 @ end <= start now?
1: @ dump_start @
mcrhi p15, 0, r0, c7, c6, 1 @ Invalidate line by MVA
addhi r0, r0, #32 @
cmphi r1, r0 @
mcrhi p15, 0, r0, c7, c6, 1 @ Invalidate line by MVA
addhi r0, r0, #32 @
cmphi r1, r0 @
mcrhi p15, 0, r0, c7, c6, 1 @ Invalidate line by MVA
addhi r0, r0, #32 @
cmphi r1, r0 @
mcrhi p15, 0, r0, c7, c6, 1 @ Invalidate line by MVA
addhi r0, r0, #32 @
cmphi r1, r0 @
mcrhi p15, 0, r0, c7, c6, 1 @ Invalidate line by MVA
addhi r0, r0, #32 @
cmphi r1, r0 @
mcrhi p15, 0, r0, c7, c6, 1 @ Invalidate line by MVA
addhi r0, r0, #32 @
cmphi r1, r0 @
mcrhi p15, 0, r0, c7, c6, 1 @ Invalidate line by MVA
addhi r0, r0, #32 @
cmphi r1, r0 @
mcrhi p15, 0, r0, c7, c6, 1 @ Invalidate line by MVA
addhi r0, r0, #32 @
cmphi r1, r0 @
bhi 1b @ dump_start @
mov r0, #0 @
mcr p15, 0, r0, c7, c10, 4 @ Drain write buffer
bx lr @
.size dump_dcache_range, .-dump_dcache_range
/*
* Cleans entire DCache
* void clean_dcache(void);
*/
.section .text, "ax", %progbits
.align 2
.global clean_dcache
.type clean_dcache, %function
.global cpucache_flush @ Alias
clean_dcache:
cpucache_flush:
#ifdef HAVE_TEST_AND_CLEAN_CACHE
mrc p15, 0, r15, c7, c10, 3 @ test and clean dcache
bne clean_dcache
mov r1, #0
#else
mov r1, #0x00000000 @
1: @ clean_start @
mcr p15, 0, r1, c7, c10, 2 @ Clean entry by index
add r0, r1, #(1<<CACHEALIGN_BITS)
mcr p15, 0, r0, c7, c10, 2 @ Clean entry by index
.rept INDEX_STEPS - 2 /* 2 steps already executed */
add r0, r0, #(1<<CACHEALIGN_BITS)
mcr p15, 0, r0, c7, c10, 2 @ Clean entry by index
.endr
adds r1, r1, #0x04000000 @ will wrap to zero at loop end
bne 1b @ clean_start @
#endif /* HAVE_TEST_AND_CLEAN_CACHE */
mcr p15, 0, r1, c7, c10, 4 @ Drain write buffer
bx lr @
.size clean_dcache, .-clean_dcache
/*
* Invalidate entire DCache
* will do writeback
* void invalidate_dcache(void);
*/
.section .icode, "ax", %progbits
.align 2
.global invalidate_dcache
.type invalidate_dcache, %function
invalidate_dcache:
#ifdef HAVE_TEST_AND_CLEAN_CACHE
mrc p15, 0, r15, c7, c14, 3 @ test, clean and invalidate dcache
bne invalidate_dcache
mov r1, #0
#else
mov r1, #0x00000000 @
1: @ inv_start @
mcr p15, 0, r1, c7, c14, 2 @ Clean and invalidate entry by index
add r0, r1, #(1<<CACHEALIGN_BITS)
mcr p15, 0, r0, c7, c14, 2 @ Clean and invalidate entry by index
.rept INDEX_STEPS - 2 /* 2 steps already executed */
add r0, r0, #(1<<CACHEALIGN_BITS)
mcr p15, 0, r0, c7, c14, 2 @ Clean and invalidate entry by index
.endr
adds r1, r1, #0x04000000 @ will wrap to zero at loop end
bne 1b @ inv_start @
#endif /* HAVE_TEST_AND_CLEAN_CACHE */
mcr p15, 0, r1, c7, c10, 4 @ Drain write buffer
bx lr @
.size invalidate_dcache, .-invalidate_dcache
/*
* Invalidate entire ICache and DCache
* will do writeback
* void invalidate_idcache(void);
*/
.section .icode, "ax", %progbits
.align 2
.global invalidate_idcache
.type invalidate_idcache, %function
.global cpucache_invalidate @ Alias
invalidate_idcache:
cpucache_invalidate:
mov r2, lr @ save lr to r1, call uses r0 only
bl invalidate_dcache @ Clean and invalidate entire DCache
mcr p15, 0, r1, c7, c5, 0 @ Invalidate ICache (r1=0 from call)
bx r2
.size invalidate_idcache, .-invalidate_idcache
|
