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
|
// SPDX-License-Identifier: GPL-2.0+
/*
* Octeon Watchdog driver
*
* Copyright (C) 2007-2017 Cavium, Inc.
*
* Converted to use WATCHDOG_CORE by Aaro Koskinen <aaro.koskinen@iki.fi>.
*
* Some parts derived from wdt.c
*
* (c) Copyright 1996-1997 Alan Cox <alan@lxorguk.ukuu.org.uk>,
* All Rights Reserved.
*
* Neither Alan Cox nor CymruNet Ltd. admit liability nor provide
* warranty for any of this software. This material is provided
* "AS-IS" and at no charge.
*
* (c) Copyright 1995 Alan Cox <alan@lxorguk.ukuu.org.uk>
*
* The OCTEON watchdog has a maximum timeout of 2^32 * io_clock.
* For most systems this is less than 10 seconds, so to allow for
* software to request longer watchdog heartbeats, we maintain software
* counters to count multiples of the base rate. If the system locks
* up in such a manner that we can not run the software counters, the
* only result is a watchdog reset sooner than was requested. But
* that is OK, because in this case userspace would likely not be able
* to do anything anyhow.
*
* The hardware watchdog interval we call the period. The OCTEON
* watchdog goes through several stages, after the first period an
* irq is asserted, then if it is not reset, after the next period NMI
* is asserted, then after an additional period a chip wide soft reset.
* So for the software counters, we reset watchdog after each period
* and decrement the counter. But for the last two periods we need to
* let the watchdog progress to the NMI stage so we disable the irq
* and let it proceed. Once in the NMI, we print the register state
* to the serial port and then wait for the reset.
*
* A watchdog is maintained for each CPU in the system, that way if
* one CPU suffers a lockup, we also get a register dump and reset.
* The userspace ping resets the watchdog on all CPUs.
*
* Before userspace opens the watchdog device, we still run the
* watchdogs to catch any lockups that may be kernel related.
*
*/
#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
#include <linux/interrupt.h>
#include <linux/watchdog.h>
#include <linux/cpumask.h>
#include <linux/module.h>
#include <linux/delay.h>
#include <linux/cpu.h>
#include <linux/irq.h>
#include <asm/mipsregs.h>
#include <asm/uasm.h>
#include <asm/octeon/octeon.h>
#include <asm/octeon/cvmx-boot-vector.h>
#include <asm/octeon/cvmx-ciu2-defs.h>
#include <asm/octeon/cvmx-rst-defs.h>
/* Watchdog interrupt major block number (8 MSBs of intsn) */
#define WD_BLOCK_NUMBER 0x01
static int divisor;
/* The count needed to achieve timeout_sec. */
static unsigned int timeout_cnt;
/* The maximum period supported. */
static unsigned int max_timeout_sec;
/* The current period. */
static unsigned int timeout_sec;
/* Set to non-zero when userspace countdown mode active */
static bool do_countdown;
static unsigned int countdown_reset;
static unsigned int per_cpu_countdown[NR_CPUS];
static cpumask_t irq_enabled_cpus;
#define WD_TIMO 60 /* Default heartbeat = 60 seconds */
#define CVMX_GSERX_SCRATCH(offset) (CVMX_ADD_IO_SEG(0x0001180090000020ull) + ((offset) & 15) * 0x1000000ull)
static int heartbeat = WD_TIMO;
module_param(heartbeat, int, 0444);
MODULE_PARM_DESC(heartbeat,
"Watchdog heartbeat in seconds. (0 < heartbeat, default="
__MODULE_STRING(WD_TIMO) ")");
static bool nowayout = WATCHDOG_NOWAYOUT;
module_param(nowayout, bool, 0444);
MODULE_PARM_DESC(nowayout,
"Watchdog cannot be stopped once started (default="
__MODULE_STRING(WATCHDOG_NOWAYOUT) ")");
static int disable;
module_param(disable, int, 0444);
MODULE_PARM_DESC(disable,
"Disable the watchdog entirely (default=0)");
static struct cvmx_boot_vector_element *octeon_wdt_bootvector;
void octeon_wdt_nmi_stage2(void);
static int cpu2core(int cpu)
{
#ifdef CONFIG_SMP
return cpu_logical_map(cpu) & 0x3f;
#else
return cvmx_get_core_num();
#endif
}
/**
* Poke the watchdog when an interrupt is received
*
* @cpl:
* @dev_id:
*
* Returns
*/
static irqreturn_t octeon_wdt_poke_irq(int cpl, void *dev_id)
{
int cpu = raw_smp_processor_id();
unsigned int core = cpu2core(cpu);
int node = cpu_to_node(cpu);
if (do_countdown) {
if (per_cpu_countdown[cpu] > 0) {
/* We're alive, poke the watchdog */
cvmx_write_csr_node(node, CVMX_CIU_PP_POKEX(core), 1);
per_cpu_countdown[cpu]--;
} else {
/* Bad news, you are about to reboot. */
disable_irq_nosync(cpl);
cpumask_clear_cpu(cpu, &irq_enabled_cpus);
}
} else {
/* Not open, just ping away... */
cvmx_write_csr_node(node, CVMX_CIU_PP_POKEX(core), 1);
}
return IRQ_HANDLED;
}
/* From setup.c */
extern int prom_putchar(char c);
/**
* Write a string to the uart
*
* @str: String to write
*/
static void octeon_wdt_write_string(const char *str)
{
/* Just loop writing one byte at a time */
while (*str)
prom_putchar(*str++);
}
/**
* Write a hex number out of the uart
*
* @value: Number to display
* @digits: Number of digits to print (1 to 16)
*/
static void octeon_wdt_write_hex(u64 value, int digits)
{
int d;
int v;
for (d = 0; d < digits; d++) {
v = (value >> ((digits - d - 1) * 4)) & 0xf;
if (v >= 10)
prom_putchar('a' + v - 10);
else
prom_putchar('0' + v);
}
}
static const char reg_name[][3] = {
"$0", "at", "v0", "v1", "a0", "a1", "a2", "a3",
"a4", "a5", "a6", "a7", "t0", "t1", "t2", "t3",
"s0", "s1", "s2", "s3", "s4", "s5", "s6", "s7",
"t8", "t9", "k0", "k1", "gp", "sp", "s8", "ra"
};
/**
* NMI stage 3 handler. NMIs are handled in the following manner:
* 1) The first NMI handler enables CVMSEG and transfers from
* the bootbus region into normal memory. It is careful to not
* destroy any registers.
* 2) The second stage handler uses CVMSEG to save the registers
* and create a stack for C code. It then calls the third level
* handler with one argument, a pointer to the register values.
* 3) The third, and final, level handler is the following C
* function that prints out some useful infomration.
*
* @reg: Pointer to register state before the NMI
*/
void octeon_wdt_nmi_stage3(u64 reg[32])
{
u64 i;
unsigned int coreid = cvmx_get_core_num();
/*
* Save status and cause early to get them before any changes
* might happen.
*/
u64 cp0_cause = read_c0_cause();
u64 cp0_status = read_c0_status();
u64 cp0_error_epc = read_c0_errorepc();
u64 cp0_epc = read_c0_epc();
/* Delay so output from all cores output is not jumbled together. */
udelay(85000 * coreid);
octeon_wdt_write_string("\r\n*** NMI Watchdog interrupt on Core 0x");
octeon_wdt_write_hex(coreid, 2);
octeon_wdt_write_string(" ***\r\n");
for (i = 0; i < 32; i++) {
octeon_wdt_write_string("\t");
octeon_wdt_write_string(reg_name[i]);
octeon_wdt_write_string("\t0x");
octeon_wdt_write_hex(reg[i], 16);
if (i & 1)
octeon_wdt_write_string("\r\n");
}
octeon_wdt_write_string("\terr_epc\t0x");
octeon_wdt_write_hex(cp0_error_epc, 16);
octeon_wdt_write_string("\tepc\t0x");
octeon_wdt_write_hex(cp0_epc, 16);
octeon_wdt_write_string("\r\n");
octeon_wdt_write_string("\tstatus\t0x");
octeon_wdt_write_hex(cp0_status, 16);
octeon_wdt_write_string("\tcause\t0x");
octeon_wdt_write_hex(cp0_cause, 16);
octeon_wdt_write_string("\r\n");
/* The CIU register is different for each Octeon model. */
if (OCTEON_IS_MODEL(OCTEON_CN68XX)) {
octeon_wdt_write_string("\tsrc_wd\t0x");
octeon_wdt_write_hex(cvmx_read_csr(CVMX_CIU2_SRC_PPX_IP2_WDOG(coreid)), 16);
octeon_wdt_write_string("\ten_wd\t0x");
octeon_wdt_write_hex(cvmx_read_csr(CVMX_CIU2_EN_PPX_IP2_WDOG(coreid)), 16);
octeon_wdt_write_string("\r\n");
octeon_wdt_write_string("\tsrc_rml\t0x");
octeon_wdt_write_hex(cvmx_read_csr(CVMX_CIU2_SRC_PPX_IP2_RML(coreid)), 16);
octeon_wdt_write_string("\ten_rml\t0x");
octeon_wdt_write_hex(cvmx_read_csr(CVMX_CIU2_EN_PPX_IP2_RML(coreid)), 16);
octeon_wdt_write_string("\r\n");
octeon_wdt_write_string("\tsum\t0x");
octeon_wdt_write_hex(cvmx_read_csr(CVMX_CIU2_SUM_PPX_IP2(coreid)), 16);
octeon_wdt_write_string("\r\n");
} else if (!octeon_has_feature(OCTEON_FEATURE_CIU3)) {
octeon_wdt_write_string("\tsum0\t0x");
octeon_wdt_write_hex(cvmx_read_csr(CVMX_CIU_INTX_SUM0(coreid * 2)), <
|