forked from torvalds/linux
-
Notifications
You must be signed in to change notification settings - Fork 0
/
mca.c
2128 lines (1874 loc) · 60.5 KB
/
mca.c
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
/*
* File: mca.c
* Purpose: Generic MCA handling layer
*
* Copyright (C) 2003 Hewlett-Packard Co
* David Mosberger-Tang <[email protected]>
*
* Copyright (C) 2002 Dell Inc.
* Copyright (C) Matt Domsch <[email protected]>
*
* Copyright (C) 2002 Intel
* Copyright (C) Jenna Hall <[email protected]>
*
* Copyright (C) 2001 Intel
* Copyright (C) Fred Lewis <[email protected]>
*
* Copyright (C) 2000 Intel
* Copyright (C) Chuck Fleckenstein <[email protected]>
*
* Copyright (C) 1999, 2004-2008 Silicon Graphics, Inc.
* Copyright (C) Vijay Chander <[email protected]>
*
* Copyright (C) 2006 FUJITSU LIMITED
* Copyright (C) Hidetoshi Seto <[email protected]>
*
* 2000-03-29 Chuck Fleckenstein <[email protected]>
* Fixed PAL/SAL update issues, began MCA bug fixes, logging issues,
* added min save state dump, added INIT handler.
*
* 2001-01-03 Fred Lewis <[email protected]>
* Added setup of CMCI and CPEI IRQs, logging of corrected platform
* errors, completed code for logging of corrected & uncorrected
* machine check errors, and updated for conformance with Nov. 2000
* revision of the SAL 3.0 spec.
*
* 2002-01-04 Jenna Hall <[email protected]>
* Aligned MCA stack to 16 bytes, added platform vs. CPU error flag,
* set SAL default return values, changed error record structure to
* linked list, added init call to sal_get_state_info_size().
*
* 2002-03-25 Matt Domsch <[email protected]>
* GUID cleanups.
*
* 2003-04-15 David Mosberger-Tang <[email protected]>
* Added INIT backtrace support.
*
* 2003-12-08 Keith Owens <[email protected]>
* smp_call_function() must not be called from interrupt context
* (can deadlock on tasklist_lock).
* Use keventd to call smp_call_function().
*
* 2004-02-01 Keith Owens <[email protected]>
* Avoid deadlock when using printk() for MCA and INIT records.
* Delete all record printing code, moved to salinfo_decode in user
* space. Mark variables and functions static where possible.
* Delete dead variables and functions. Reorder to remove the need
* for forward declarations and to consolidate related code.
*
* 2005-08-12 Keith Owens <[email protected]>
* Convert MCA/INIT handlers to use per event stacks and SAL/OS
* state.
*
* 2005-10-07 Keith Owens <[email protected]>
* Add notify_die() hooks.
*
* 2006-09-15 Hidetoshi Seto <[email protected]>
* Add printing support for MCA/INIT.
*
* 2007-04-27 Russ Anderson <[email protected]>
* Support multiple cpus going through OS_MCA in the same event.
*/
#include <linux/jiffies.h>
#include <linux/types.h>
#include <linux/init.h>
#include <linux/sched.h>
#include <linux/interrupt.h>
#include <linux/irq.h>
#include <linux/bootmem.h>
#include <linux/acpi.h>
#include <linux/timer.h>
#include <linux/module.h>
#include <linux/kernel.h>
#include <linux/smp.h>
#include <linux/workqueue.h>
#include <linux/cpumask.h>
#include <linux/kdebug.h>
#include <linux/cpu.h>
#include <asm/delay.h>
#include <asm/machvec.h>
#include <asm/meminit.h>
#include <asm/page.h>
#include <asm/ptrace.h>
#include <asm/system.h>
#include <asm/sal.h>
#include <asm/mca.h>
#include <asm/kexec.h>
#include <asm/irq.h>
#include <asm/hw_irq.h>
#include <asm/tlb.h>
#include "mca_drv.h"
#include "entry.h"
#if defined(IA64_MCA_DEBUG_INFO)
# define IA64_MCA_DEBUG(fmt...) printk(fmt)
#else
# define IA64_MCA_DEBUG(fmt...)
#endif
#define NOTIFY_INIT(event, regs, arg, spin) \
do { \
if ((notify_die((event), "INIT", (regs), (arg), 0, 0) \
== NOTIFY_STOP) && ((spin) == 1)) \
ia64_mca_spin(__func__); \
} while (0)
#define NOTIFY_MCA(event, regs, arg, spin) \
do { \
if ((notify_die((event), "MCA", (regs), (arg), 0, 0) \
== NOTIFY_STOP) && ((spin) == 1)) \
ia64_mca_spin(__func__); \
} while (0)
/* Used by mca_asm.S */
DEFINE_PER_CPU(u64, ia64_mca_data); /* == __per_cpu_mca[smp_processor_id()] */
DEFINE_PER_CPU(u64, ia64_mca_per_cpu_pte); /* PTE to map per-CPU area */
DEFINE_PER_CPU(u64, ia64_mca_pal_pte); /* PTE to map PAL code */
DEFINE_PER_CPU(u64, ia64_mca_pal_base); /* vaddr PAL code granule */
DEFINE_PER_CPU(u64, ia64_mca_tr_reload); /* Flag for TR reload */
unsigned long __per_cpu_mca[NR_CPUS];
/* In mca_asm.S */
extern void ia64_os_init_dispatch_monarch (void);
extern void ia64_os_init_dispatch_slave (void);
static int monarch_cpu = -1;
static ia64_mc_info_t ia64_mc_info;
#define MAX_CPE_POLL_INTERVAL (15*60*HZ) /* 15 minutes */
#define MIN_CPE_POLL_INTERVAL (2*60*HZ) /* 2 minutes */
#define CMC_POLL_INTERVAL (1*60*HZ) /* 1 minute */
#define CPE_HISTORY_LENGTH 5
#define CMC_HISTORY_LENGTH 5
#ifdef CONFIG_ACPI
static struct timer_list cpe_poll_timer;
#endif
static struct timer_list cmc_poll_timer;
/*
* This variable tells whether we are currently in polling mode.
* Start with this in the wrong state so we won't play w/ timers
* before the system is ready.
*/
static int cmc_polling_enabled = 1;
/*
* Clearing this variable prevents CPE polling from getting activated
* in mca_late_init. Use it if your system doesn't provide a CPEI,
* but encounters problems retrieving CPE logs. This should only be
* necessary for debugging.
*/
static int cpe_poll_enabled = 1;
extern void salinfo_log_wakeup(int type, u8 *buffer, u64 size, int irqsafe);
static int mca_init __initdata;
/*
* limited & delayed printing support for MCA/INIT handler
*/
#define mprintk(fmt...) ia64_mca_printk(fmt)
#define MLOGBUF_SIZE (512+256*NR_CPUS)
#define MLOGBUF_MSGMAX 256
static char mlogbuf[MLOGBUF_SIZE];
static DEFINE_SPINLOCK(mlogbuf_wlock); /* mca context only */
static DEFINE_SPINLOCK(mlogbuf_rlock); /* normal context only */
static unsigned long mlogbuf_start;
static unsigned long mlogbuf_end;
static unsigned int mlogbuf_finished = 0;
static unsigned long mlogbuf_timestamp = 0;
static int loglevel_save = -1;
#define BREAK_LOGLEVEL(__console_loglevel) \
oops_in_progress = 1; \
if (loglevel_save < 0) \
loglevel_save = __console_loglevel; \
__console_loglevel = 15;
#define RESTORE_LOGLEVEL(__console_loglevel) \
if (loglevel_save >= 0) { \
__console_loglevel = loglevel_save; \
loglevel_save = -1; \
} \
mlogbuf_finished = 0; \
oops_in_progress = 0;
/*
* Push messages into buffer, print them later if not urgent.
*/
void ia64_mca_printk(const char *fmt, ...)
{
va_list args;
int printed_len;
char temp_buf[MLOGBUF_MSGMAX];
char *p;
va_start(args, fmt);
printed_len = vscnprintf(temp_buf, sizeof(temp_buf), fmt, args);
va_end(args);
/* Copy the output into mlogbuf */
if (oops_in_progress) {
/* mlogbuf was abandoned, use printk directly instead. */
printk(temp_buf);
} else {
spin_lock(&mlogbuf_wlock);
for (p = temp_buf; *p; p++) {
unsigned long next = (mlogbuf_end + 1) % MLOGBUF_SIZE;
if (next != mlogbuf_start) {
mlogbuf[mlogbuf_end] = *p;
mlogbuf_end = next;
} else {
/* buffer full */
break;
}
}
mlogbuf[mlogbuf_end] = '\0';
spin_unlock(&mlogbuf_wlock);
}
}
EXPORT_SYMBOL(ia64_mca_printk);
/*
* Print buffered messages.
* NOTE: call this after returning normal context. (ex. from salinfod)
*/
void ia64_mlogbuf_dump(void)
{
char temp_buf[MLOGBUF_MSGMAX];
char *p;
unsigned long index;
unsigned long flags;
unsigned int printed_len;
/* Get output from mlogbuf */
while (mlogbuf_start != mlogbuf_end) {
temp_buf[0] = '\0';
p = temp_buf;
printed_len = 0;
spin_lock_irqsave(&mlogbuf_rlock, flags);
index = mlogbuf_start;
while (index != mlogbuf_end) {
*p = mlogbuf[index];
index = (index + 1) % MLOGBUF_SIZE;
if (!*p)
break;
p++;
if (++printed_len >= MLOGBUF_MSGMAX - 1)
break;
}
*p = '\0';
if (temp_buf[0])
printk(temp_buf);
mlogbuf_start = index;
mlogbuf_timestamp = 0;
spin_unlock_irqrestore(&mlogbuf_rlock, flags);
}
}
EXPORT_SYMBOL(ia64_mlogbuf_dump);
/*
* Call this if system is going to down or if immediate flushing messages to
* console is required. (ex. recovery was failed, crash dump is going to be
* invoked, long-wait rendezvous etc.)
* NOTE: this should be called from monarch.
*/
static void ia64_mlogbuf_finish(int wait)
{
BREAK_LOGLEVEL(console_loglevel);
spin_lock_init(&mlogbuf_rlock);
ia64_mlogbuf_dump();
printk(KERN_EMERG "mlogbuf_finish: printing switched to urgent mode, "
"MCA/INIT might be dodgy or fail.\n");
if (!wait)
return;
/* wait for console */
printk("Delaying for 5 seconds...\n");
udelay(5*1000000);
mlogbuf_finished = 1;
}
/*
* Print buffered messages from INIT context.
*/
static void ia64_mlogbuf_dump_from_init(void)
{
if (mlogbuf_finished)
return;
if (mlogbuf_timestamp &&
time_before(jiffies, mlogbuf_timestamp + 30 * HZ)) {
printk(KERN_ERR "INIT: mlogbuf_dump is interrupted by INIT "
" and the system seems to be messed up.\n");
ia64_mlogbuf_finish(0);
return;
}
if (!spin_trylock(&mlogbuf_rlock)) {
printk(KERN_ERR "INIT: mlogbuf_dump is interrupted by INIT. "
"Generated messages other than stack dump will be "
"buffered to mlogbuf and will be printed later.\n");
printk(KERN_ERR "INIT: If messages would not printed after "
"this INIT, wait 30sec and assert INIT again.\n");
if (!mlogbuf_timestamp)
mlogbuf_timestamp = jiffies;
return;
}
spin_unlock(&mlogbuf_rlock);
ia64_mlogbuf_dump();
}
static void inline
ia64_mca_spin(const char *func)
{
if (monarch_cpu == smp_processor_id())
ia64_mlogbuf_finish(0);
mprintk(KERN_EMERG "%s: spinning here, not returning to SAL\n", func);
while (1)
cpu_relax();
}
/*
* IA64_MCA log support
*/
#define IA64_MAX_LOGS 2 /* Double-buffering for nested MCAs */
#define IA64_MAX_LOG_TYPES 4 /* MCA, INIT, CMC, CPE */
typedef struct ia64_state_log_s
{
spinlock_t isl_lock;
int isl_index;
unsigned long isl_count;
ia64_err_rec_t *isl_log[IA64_MAX_LOGS]; /* need space to store header + error log */
} ia64_state_log_t;
static ia64_state_log_t ia64_state_log[IA64_MAX_LOG_TYPES];
#define IA64_LOG_ALLOCATE(it, size) \
{ia64_state_log[it].isl_log[IA64_LOG_CURR_INDEX(it)] = \
(ia64_err_rec_t *)alloc_bootmem(size); \
ia64_state_log[it].isl_log[IA64_LOG_NEXT_INDEX(it)] = \
(ia64_err_rec_t *)alloc_bootmem(size);}
#define IA64_LOG_LOCK_INIT(it) spin_lock_init(&ia64_state_log[it].isl_lock)
#define IA64_LOG_LOCK(it) spin_lock_irqsave(&ia64_state_log[it].isl_lock, s)
#define IA64_LOG_UNLOCK(it) spin_unlock_irqrestore(&ia64_state_log[it].isl_lock,s)
#define IA64_LOG_NEXT_INDEX(it) ia64_state_log[it].isl_index
#define IA64_LOG_CURR_INDEX(it) 1 - ia64_state_log[it].isl_index
#define IA64_LOG_INDEX_INC(it) \
{ia64_state_log[it].isl_index = 1 - ia64_state_log[it].isl_index; \
ia64_state_log[it].isl_count++;}
#define IA64_LOG_INDEX_DEC(it) \
ia64_state_log[it].isl_index = 1 - ia64_state_log[it].isl_index
#define IA64_LOG_NEXT_BUFFER(it) (void *)((ia64_state_log[it].isl_log[IA64_LOG_NEXT_INDEX(it)]))
#define IA64_LOG_CURR_BUFFER(it) (void *)((ia64_state_log[it].isl_log[IA64_LOG_CURR_INDEX(it)]))
#define IA64_LOG_COUNT(it) ia64_state_log[it].isl_count
/*
* ia64_log_init
* Reset the OS ia64 log buffer
* Inputs : info_type (SAL_INFO_TYPE_{MCA,INIT,CMC,CPE})
* Outputs : None
*/
static void __init
ia64_log_init(int sal_info_type)
{
u64 max_size = 0;
IA64_LOG_NEXT_INDEX(sal_info_type) = 0;
IA64_LOG_LOCK_INIT(sal_info_type);
// SAL will tell us the maximum size of any error record of this type
max_size = ia64_sal_get_state_info_size(sal_info_type);
if (!max_size)
/* alloc_bootmem() doesn't like zero-sized allocations! */
return;
// set up OS data structures to hold error info
IA64_LOG_ALLOCATE(sal_info_type, max_size);
memset(IA64_LOG_CURR_BUFFER(sal_info_type), 0, max_size);
memset(IA64_LOG_NEXT_BUFFER(sal_info_type), 0, max_size);
}
/*
* ia64_log_get
*
* Get the current MCA log from SAL and copy it into the OS log buffer.
*
* Inputs : info_type (SAL_INFO_TYPE_{MCA,INIT,CMC,CPE})
* irq_safe whether you can use printk at this point
* Outputs : size (total record length)
* *buffer (ptr to error record)
*
*/
static u64
ia64_log_get(int sal_info_type, u8 **buffer, int irq_safe)
{
sal_log_record_header_t *log_buffer;
u64 total_len = 0;
unsigned long s;
IA64_LOG_LOCK(sal_info_type);
/* Get the process state information */
log_buffer = IA64_LOG_NEXT_BUFFER(sal_info_type);
total_len = ia64_sal_get_state_info(sal_info_type, (u64 *)log_buffer);
if (total_len) {
IA64_LOG_INDEX_INC(sal_info_type);
IA64_LOG_UNLOCK(sal_info_type);
if (irq_safe) {
IA64_MCA_DEBUG("%s: SAL error record type %d retrieved. Record length = %ld\n",
__func__, sal_info_type, total_len);
}
*buffer = (u8 *) log_buffer;
return total_len;
} else {
IA64_LOG_UNLOCK(sal_info_type);
return 0;
}
}
/*
* ia64_mca_log_sal_error_record
*
* This function retrieves a specified error record type from SAL
* and wakes up any processes waiting for error records.
*
* Inputs : sal_info_type (Type of error record MCA/CMC/CPE)
* FIXME: remove MCA and irq_safe.
*/
static void
ia64_mca_log_sal_error_record(int sal_info_type)
{
u8 *buffer;
sal_log_record_header_t *rh;
u64 size;
int irq_safe = sal_info_type != SAL_INFO_TYPE_MCA;
#ifdef IA64_MCA_DEBUG_INFO
static const char * const rec_name[] = { "MCA", "INIT", "CMC", "CPE" };
#endif
size = ia64_log_get(sal_info_type, &buffer, irq_safe);
if (!size)
return;
salinfo_log_wakeup(sal_info_type, buffer, size, irq_safe);
if (irq_safe)
IA64_MCA_DEBUG("CPU %d: SAL log contains %s error record\n",
smp_processor_id(),
sal_info_type < ARRAY_SIZE(rec_name) ? rec_name[sal_info_type] : "UNKNOWN");
/* Clear logs from corrected errors in case there's no user-level logger */
rh = (sal_log_record_header_t *)buffer;
if (rh->severity == sal_log_severity_corrected)
ia64_sal_clear_state_info(sal_info_type);
}
/*
* search_mca_table
* See if the MCA surfaced in an instruction range
* that has been tagged as recoverable.
*
* Inputs
* first First address range to check
* last Last address range to check
* ip Instruction pointer, address we are looking for
*
* Return value:
* 1 on Success (in the table)/ 0 on Failure (not in the table)
*/
int
search_mca_table (const struct mca_table_entry *first,
const struct mca_table_entry *last,
unsigned long ip)
{
const struct mca_table_entry *curr;
u64 curr_start, curr_end;
curr = first;
while (curr <= last) {
curr_start = (u64) &curr->start_addr + curr->start_addr;
curr_end = (u64) &curr->end_addr + curr->end_addr;
if ((ip >= curr_start) && (ip <= curr_end)) {
return 1;
}
curr++;
}
return 0;
}
/* Given an address, look for it in the mca tables. */
int mca_recover_range(unsigned long addr)
{
extern struct mca_table_entry __start___mca_table[];
extern struct mca_table_entry __stop___mca_table[];
return search_mca_table(__start___mca_table, __stop___mca_table-1, addr);
}
EXPORT_SYMBOL_GPL(mca_recover_range);
#ifdef CONFIG_ACPI
int cpe_vector = -1;
int ia64_cpe_irq = -1;
static irqreturn_t
ia64_mca_cpe_int_handler (int cpe_irq, void *arg)
{
static unsigned long cpe_history[CPE_HISTORY_LENGTH];
static int index;
static DEFINE_SPINLOCK(cpe_history_lock);
IA64_MCA_DEBUG("%s: received interrupt vector = %#x on CPU %d\n",
__func__, cpe_irq, smp_processor_id());
/* SAL spec states this should run w/ interrupts enabled */
local_irq_enable();
spin_lock(&cpe_history_lock);
if (!cpe_poll_enabled && cpe_vector >= 0) {
int i, count = 1; /* we know 1 happened now */
unsigned long now = jiffies;
for (i = 0; i < CPE_HISTORY_LENGTH; i++) {
if (now - cpe_history[i] <= HZ)
count++;
}
IA64_MCA_DEBUG(KERN_INFO "CPE threshold %d/%d\n", count, CPE_HISTORY_LENGTH);
if (count >= CPE_HISTORY_LENGTH) {
cpe_poll_enabled = 1;
spin_unlock(&cpe_history_lock);
disable_irq_nosync(local_vector_to_irq(IA64_CPE_VECTOR));
/*
* Corrected errors will still be corrected, but
* make sure there's a log somewhere that indicates
* something is generating more than we can handle.
*/
printk(KERN_WARNING "WARNING: Switching to polling CPE handler; error records may be lost\n");
mod_timer(&cpe_poll_timer, jiffies + MIN_CPE_POLL_INTERVAL);
/* lock already released, get out now */
goto out;
} else {
cpe_history[index++] = now;
if (index == CPE_HISTORY_LENGTH)
index = 0;
}
}
spin_unlock(&cpe_history_lock);
out:
/* Get the CPE error record and log it */
ia64_mca_log_sal_error_record(SAL_INFO_TYPE_CPE);
return IRQ_HANDLED;
}
#endif /* CONFIG_ACPI */
#ifdef CONFIG_ACPI
/*
* ia64_mca_register_cpev
*
* Register the corrected platform error vector with SAL.
*
* Inputs
* cpev Corrected Platform Error Vector number
*
* Outputs
* None
*/
void
ia64_mca_register_cpev (int cpev)
{
/* Register the CPE interrupt vector with SAL */
struct ia64_sal_retval isrv;
isrv = ia64_sal_mc_set_params(SAL_MC_PARAM_CPE_INT, SAL_MC_PARAM_MECHANISM_INT, cpev, 0, 0);
if (isrv.status) {
printk(KERN_ERR "Failed to register Corrected Platform "
"Error interrupt vector with SAL (status %ld)\n", isrv.status);
return;
}
IA64_MCA_DEBUG("%s: corrected platform error "
"vector %#x registered\n", __func__, cpev);
}
#endif /* CONFIG_ACPI */
/*
* ia64_mca_cmc_vector_setup
*
* Setup the corrected machine check vector register in the processor.
* (The interrupt is masked on boot. ia64_mca_late_init unmask this.)
* This function is invoked on a per-processor basis.
*
* Inputs
* None
*
* Outputs
* None
*/
void __cpuinit
ia64_mca_cmc_vector_setup (void)
{
cmcv_reg_t cmcv;
cmcv.cmcv_regval = 0;
cmcv.cmcv_mask = 1; /* Mask/disable interrupt at first */
cmcv.cmcv_vector = IA64_CMC_VECTOR;
ia64_setreg(_IA64_REG_CR_CMCV, cmcv.cmcv_regval);
IA64_MCA_DEBUG("%s: CPU %d corrected machine check vector %#x registered.\n",
__func__, smp_processor_id(), IA64_CMC_VECTOR);
IA64_MCA_DEBUG("%s: CPU %d CMCV = %#016lx\n",
__func__, smp_processor_id(), ia64_getreg(_IA64_REG_CR_CMCV));
}
/*
* ia64_mca_cmc_vector_disable
*
* Mask the corrected machine check vector register in the processor.
* This function is invoked on a per-processor basis.
*
* Inputs
* dummy(unused)
*
* Outputs
* None
*/
static void
ia64_mca_cmc_vector_disable (void *dummy)
{
cmcv_reg_t cmcv;
cmcv.cmcv_regval = ia64_getreg(_IA64_REG_CR_CMCV);
cmcv.cmcv_mask = 1; /* Mask/disable interrupt */
ia64_setreg(_IA64_REG_CR_CMCV, cmcv.cmcv_regval);
IA64_MCA_DEBUG("%s: CPU %d corrected machine check vector %#x disabled.\n",
__func__, smp_processor_id(), cmcv.cmcv_vector);
}
/*
* ia64_mca_cmc_vector_enable
*
* Unmask the corrected machine check vector register in the processor.
* This function is invoked on a per-processor basis.
*
* Inputs
* dummy(unused)
*
* Outputs
* None
*/
static void
ia64_mca_cmc_vector_enable (void *dummy)
{
cmcv_reg_t cmcv;
cmcv.cmcv_regval = ia64_getreg(_IA64_REG_CR_CMCV);
cmcv.cmcv_mask = 0; /* Unmask/enable interrupt */
ia64_setreg(_IA64_REG_CR_CMCV, cmcv.cmcv_regval);
IA64_MCA_DEBUG("%s: CPU %d corrected machine check vector %#x enabled.\n",
__func__, smp_processor_id(), cmcv.cmcv_vector);
}
/*
* ia64_mca_cmc_vector_disable_keventd
*
* Called via keventd (smp_call_function() is not safe in interrupt context) to
* disable the cmc interrupt vector.
*/
static void
ia64_mca_cmc_vector_disable_keventd(struct work_struct *unused)
{
on_each_cpu(ia64_mca_cmc_vector_disable, NULL, 0);
}
/*
* ia64_mca_cmc_vector_enable_keventd
*
* Called via keventd (smp_call_function() is not safe in interrupt context) to
* enable the cmc interrupt vector.
*/
static void
ia64_mca_cmc_vector_enable_keventd(struct work_struct *unused)
{
on_each_cpu(ia64_mca_cmc_vector_enable, NULL, 0);
}
/*
* ia64_mca_wakeup
*
* Send an inter-cpu interrupt to wake-up a particular cpu.
*
* Inputs : cpuid
* Outputs : None
*/
static void
ia64_mca_wakeup(int cpu)
{
platform_send_ipi(cpu, IA64_MCA_WAKEUP_VECTOR, IA64_IPI_DM_INT, 0);
}
/*
* ia64_mca_wakeup_all
*
* Wakeup all the slave cpus which have rendez'ed previously.
*
* Inputs : None
* Outputs : None
*/
static void
ia64_mca_wakeup_all(void)
{
int cpu;
/* Clear the Rendez checkin flag for all cpus */
for_each_online_cpu(cpu) {
if (ia64_mc_info.imi_rendez_checkin[cpu] == IA64_MCA_RENDEZ_CHECKIN_DONE)
ia64_mca_wakeup(cpu);
}
}
/*
* ia64_mca_rendez_interrupt_handler
*
* This is handler used to put slave processors into spinloop
* while the monarch processor does the mca handling and later
* wake each slave up once the monarch is done. The state
* IA64_MCA_RENDEZ_CHECKIN_DONE indicates the cpu is rendez'ed
* in SAL. The state IA64_MCA_RENDEZ_CHECKIN_NOTDONE indicates
* the cpu has come out of OS rendezvous.
*
* Inputs : None
* Outputs : None
*/
static irqreturn_t
ia64_mca_rendez_int_handler(int rendez_irq, void *arg)
{
unsigned long flags;
int cpu = smp_processor_id();
struct ia64_mca_notify_die nd =
{ .sos = NULL, .monarch_cpu = &monarch_cpu };
/* Mask all interrupts */
local_irq_save(flags);
NOTIFY_MCA(DIE_MCA_RENDZVOUS_ENTER, get_irq_regs(), (long)&nd, 1);
ia64_mc_info.imi_rendez_checkin[cpu] = IA64_MCA_RENDEZ_CHECKIN_DONE;
/* Register with the SAL monarch that the slave has
* reached SAL
*/
ia64_sal_mc_rendez();
NOTIFY_MCA(DIE_MCA_RENDZVOUS_PROCESS, get_irq_regs(), (long)&nd, 1);
/* Wait for the monarch cpu to exit. */
while (monarch_cpu != -1)
cpu_relax(); /* spin until monarch leaves */
NOTIFY_MCA(DIE_MCA_RENDZVOUS_LEAVE, get_irq_regs(), (long)&nd, 1);
ia64_mc_info.imi_rendez_checkin[cpu] = IA64_MCA_RENDEZ_CHECKIN_NOTDONE;
/* Enable all interrupts */
local_irq_restore(flags);
return IRQ_HANDLED;
}
/*
* ia64_mca_wakeup_int_handler
*
* The interrupt handler for processing the inter-cpu interrupt to the
* slave cpu which was spinning in the rendez loop.
* Since this spinning is done by turning off the interrupts and
* polling on the wakeup-interrupt bit in the IRR, there is
* nothing useful to be done in the handler.
*
* Inputs : wakeup_irq (Wakeup-interrupt bit)
* arg (Interrupt handler specific argument)
* Outputs : None
*
*/
static irqreturn_t
ia64_mca_wakeup_int_handler(int wakeup_irq, void *arg)
{
return IRQ_HANDLED;
}
/* Function pointer for extra MCA recovery */
int (*ia64_mca_ucmc_extension)
(void*,struct ia64_sal_os_state*)
= NULL;
int
ia64_reg_MCA_extension(int (*fn)(void *, struct ia64_sal_os_state *))
{
if (ia64_mca_ucmc_extension)
return 1;
ia64_mca_ucmc_extension = fn;
return 0;
}
void
ia64_unreg_MCA_extension(void)
{
if (ia64_mca_ucmc_extension)
ia64_mca_ucmc_extension = NULL;
}
EXPORT_SYMBOL(ia64_reg_MCA_extension);
EXPORT_SYMBOL(ia64_unreg_MCA_extension);
static inline void
copy_reg(const u64 *fr, u64 fnat, u64 *tr, u64 *tnat)
{
u64 fslot, tslot, nat;
*tr = *fr;
fslot = ((unsigned long)fr >> 3) & 63;
tslot = ((unsigned long)tr >> 3) & 63;
*tnat &= ~(1UL << tslot);
nat = (fnat >> fslot) & 1;
*tnat |= (nat << tslot);
}
/* Change the comm field on the MCA/INT task to include the pid that
* was interrupted, it makes for easier debugging. If that pid was 0
* (swapper or nested MCA/INIT) then use the start of the previous comm
* field suffixed with its cpu.
*/
static void
ia64_mca_modify_comm(const struct task_struct *previous_current)
{
char *p, comm[sizeof(current->comm)];
if (previous_current->pid)
snprintf(comm, sizeof(comm), "%s %d",
current->comm, previous_current->pid);
else {
int l;
if ((p = strchr(previous_current->comm, ' ')))
l = p - previous_current->comm;
else
l = strlen(previous_current->comm);
snprintf(comm, sizeof(comm), "%s %*s %d",
current->comm, l, previous_current->comm,
task_thread_info(previous_current)->cpu);
}
memcpy(current->comm, comm, sizeof(current->comm));
}
/* On entry to this routine, we are running on the per cpu stack, see
* mca_asm.h. The original stack has not been touched by this event. Some of
* the original stack's registers will be in the RBS on this stack. This stack
* also contains a partial pt_regs and switch_stack, the rest of the data is in
* PAL minstate.
*
* The first thing to do is modify the original stack to look like a blocked
* task so we can run backtrace on the original task. Also mark the per cpu
* stack as current to ensure that we use the correct task state, it also means
* that we can do backtrace on the MCA/INIT handler code itself.
*/
static struct task_struct *
ia64_mca_modify_original_stack(struct pt_regs *regs,
const struct switch_stack *sw,
struct ia64_sal_os_state *sos,
const char *type)
{
char *p;
ia64_va va;
extern char ia64_leave_kernel[]; /* Need asm address, not function descriptor */
const pal_min_state_area_t *ms = sos->pal_min_state;
struct task_struct *previous_current;
struct pt_regs *old_regs;
struct switch_stack *old_sw;
unsigned size = sizeof(struct pt_regs) +
sizeof(struct switch_stack) + 16;
u64 *old_bspstore, *old_bsp;
u64 *new_bspstore, *new_bsp;
u64 old_unat, old_rnat, new_rnat, nat;
u64 slots, loadrs = regs->loadrs;
u64 r12 = ms->pmsa_gr[12-1], r13 = ms->pmsa_gr[13-1];
u64 ar_bspstore = regs->ar_bspstore;
u64 ar_bsp = regs->ar_bspstore + (loadrs >> 16);
const u64 *bank;
const char *msg;
int cpu = smp_processor_id();
previous_current = curr_task(cpu);
set_curr_task(cpu, current);
if ((p = strchr(current->comm, ' ')))
*p = '\0';
/* Best effort attempt to cope with MCA/INIT delivered while in
* physical mode.
*/
regs->cr_ipsr = ms->pmsa_ipsr;
if (ia64_psr(regs)->dt == 0) {
va.l = r12;
if (va.f.reg == 0) {
va.f.reg = 7;
r12 = va.l;
}
va.l = r13;
if (va.f.reg == 0) {
va.f.reg = 7;
r13 = va.l;
}
}
if (ia64_psr(regs)->rt == 0) {
va.l = ar_bspstore;
if (va.f.reg == 0) {
va.f.reg = 7;
ar_bspstore = va.l;
}
va.l = ar_bsp;
if (va.f.reg == 0) {
va.f.reg = 7;
ar_bsp = va.l;
}
}
/* mca_asm.S ia64_old_stack() cannot assume that the dirty registers
* have been copied to the old stack, the old stack may fail the
* validation tests below. So ia64_old_stack() must restore the dirty
* registers from the new stack. The old and new bspstore probably
* have different alignments, so loadrs calculated on the old bsp
* cannot be used to restore from the new bsp. Calculate a suitable
* loadrs for the new stack and save it in the new pt_regs, where
* ia64_old_stack() can get it.
*/
old_bspstore = (u64 *)ar_bspstore;
old_bsp = (u64 *)ar_bsp;
slots = ia64_rse_num_regs(old_bspstore, old_bsp);
new_bspstore = (u64 *)((u64)current + IA64_RBS_OFFSET);
new_bsp = ia64_rse_skip_regs(new_bspstore, slots);
regs->loadrs = (new_bsp - new_bspstore) * 8 << 16;
/* Verify the previous stack state before we change it */
if (user_mode(regs)) {
msg = "occurred in user space";
/* previous_current is guaranteed to be valid when the task was
* in user space, so ...
*/
ia64_mca_modify_comm(previous_current);
goto no_mod;
}
if (r13 != sos->prev_IA64_KR_CURRENT) {
msg = "inconsistent previous current and r13";
goto no_mod;
}
if (!mca_recover_range(ms->pmsa_iip)) {
if ((r12 - r13) >= KERNEL_STACK_SIZE) {
msg = "inconsistent r12 and r13";
goto no_mod;
}
if ((ar_bspstore - r13) >= KERNEL_STACK_SIZE) {
msg = "inconsistent ar.bspstore and r13";
goto no_mod;