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
|
// SPDX-License-Identifier: GPL-2.0
/*
* Machine check handler
*
* Copyright IBM Corp. 2000, 2009
* Author(s): Ingo Adlung <adlung@de.ibm.com>,
* Martin Schwidefsky <schwidefsky@de.ibm.com>,
* Cornelia Huck <cornelia.huck@de.ibm.com>,
* Heiko Carstens <heiko.carstens@de.ibm.com>,
*/
#include <linux/kernel_stat.h>
#include <linux/init.h>
#include <linux/errno.h>
#include <linux/hardirq.h>
#include <linux/log2.h>
#include <linux/kprobes.h>
#include <linux/kmemleak.h>
#include <linux/time.h>
#include <linux/module.h>
#include <linux/sched/signal.h>
#include <linux/export.h>
#include <asm/lowcore.h>
#include <asm/smp.h>
#include <asm/stp.h>
#include <asm/cputime.h>
#include <asm/nmi.h>
#include <asm/crw.h>
#include <asm/switch_to.h>
#include <asm/ctl_reg.h>
#include <asm/asm-offsets.h>
#include <linux/kvm_host.h>
struct mcck_struct {
unsigned int kill_task : 1;
unsigned int channel_report : 1;
unsigned int warning : 1;
unsigned int stp_queue : 1;
unsigned long mcck_code;
};
static DEFINE_PER_CPU(struct mcck_struct, cpu_mcck);
static struct kmem_cache *mcesa_cache;
static unsigned long mcesa_origin_lc;
static inline int nmi_needs_mcesa(void)
{
return MACHINE_HAS_VX || MACHINE_HAS_GS;
}
static inline unsigned long nmi_get_mcesa_size(void)
{
if (MACHINE_HAS_GS)
return MCESA_MAX_SIZE;
return MCESA_MIN_SIZE;
}
/*
* The initial machine check extended save area for the boot CPU.
* It will be replaced by nmi_init() with an allocated structure.
* The structure is required for machine check happening early in
* the boot process.
*/
static struct mcesa boot_mcesa __initdata __aligned(MCESA_MAX_SIZE);
void __init nmi_alloc_boot_cpu(struct lowcore *lc)
{
if (!nmi_needs_mcesa())
return;
lc->mcesad = (unsigned long) &boot_mcesa;
if (MACHINE_HAS_GS)
lc->mcesad |= ilog2(MCESA_MAX_SIZE);
}
static int __init nmi_init(void)
{
unsigned long origin, cr0, size;
if (!nmi_needs_mcesa())
return 0;
size = nmi_get_mcesa_size();
if (size > MCESA_MIN_SIZE)
mcesa_origin_lc = ilog2(size);
/* create slab cache for the machine-check-extended-save-areas */
mcesa_cache = kmem_cache_create("nmi_save_areas", size, size, 0, NULL);
if (!mcesa_cache)
panic("Couldn't create nmi save area cache");
origin = (unsigned long) kmem_cache_alloc(mcesa_cache, GFP_KERNEL);
if (!origin)
panic("Couldn't allocate nmi save area");
/* The pointer is stored with mcesa_bits ORed in */
kmemleak_not_leak((void *) origin);
__ctl_store(cr0, 0, 0);
__ctl_clear_bit(0, 28); /* disable lowcore protection */
/* Replace boot_mcesa on the boot CPU */
S390_lowcore.mcesad = origin | mcesa_origin_lc;
__ctl_load(cr0, 0, 0);
return 0;
}
early_initcall(nmi_init);
int nmi_alloc_per_cpu(struct lowcore *lc)
{
unsigned long origin;
if (!nmi_needs_mcesa())
return 0;
origin = (unsigned long) kmem_cache_alloc(mcesa_cache, GFP_KERNEL);
if (!origin)
return -ENOMEM;
/* The pointer is stored with mcesa_bits ORed in */
kmemleak_not_leak((void *) origin);
lc->mcesad = origin | mcesa_origin_lc;
return 0;
}
void nmi_free_per_cpu(struct lowcore *lc)
{
if (!nmi_needs_mcesa())
return;
kmem_cache_free(mcesa_cache, (void *)(lc->mcesad & MCESA_ORIGIN_MASK));
}
static notrace void s390_handle_damage(void)
{
smp_emergency_stop();
disabled_wait();
while (1);
}
NOKPROBE_SYMBOL(s390_handle_damage);
/*
* Main machine check handler function. Will be called with interrupts enabled
* or disabled and machine checks enabled or disabled.
*/
void s390_handle_mcck(void)
{
unsigned long flags;
struct mcck_struct mcck;
/*
* Disable machine checks and get the current state of accumulated
* machine checks. Afterwards delete the old state and enable machine
* checks again.
*/
local_irq_save(flags);
local_mcck_disable();
mcck = *this_cpu_ptr(&cpu_mcck);
memset(this_cpu_ptr(&cpu_mcck), 0, sizeof(mcck));
local_mcck_enable();
local_irq_restore(flags);
if (mcck.channel_report)
crw_handle_channel_report();
/*
* A warning may remain for a prolonged period on the bare iron.
* (actually until the machine is powered off, or the problem is gone)
* So we just stop listening for the WARNING MCH and avoid continuously
* being interrupted. One caveat is however, that we must do this per
* processor and cannot use the smp version of ctl_clear_bit().
* On VM we only get one interrupt per virtally presented machinecheck.
* Though one suffices, we may get one interrupt per (virtual) cpu.
*/
if (mcck.warning) { /* WARNING pending ? */
static int mchchk_wng_posted = 0;
/* Use single cpu clear, as we cannot handle smp here. */
__ctl_clear_bit(14, 24); /* Disable WARNING MCH */
if (xchg(&mchchk_wng_posted, 1) == 0)
kill_cad_pid(SIGPWR, 1);
}
if (mcck.stp_queue)
stp_queue_work();
if (mcck.kill_task) {
local_irq_enable();
printk(KERN_EMERG "mcck: Terminating task because of machine "
"malfunction (code 0x%016lx).\n", mcck.mcck_code);
printk(KERN_EMERG "mcck: task: %s, pid: %d.\n",
current->comm, current->pid);
do_exit(SIGSEGV);
}
}
EXPORT_SYMBOL_GPL(s390_handle_mcck);
/*
* returns 0 if all required registers are available
* returns 1 otherwise
*/
static int notrace s390_check_registers(union mci mci, int umode)
{
union ctlreg2 cr2;
int kill_task;
kill_task = 0;
if (!mci.gr) {
/*
* General purpose registers couldn't be restored and have
* unknown contents. Stop system or terminate process.
*/
if (!umode)
s390_handle_damage();
kill_task = 1;
}
/* Check control registers */
if (!mci.cr) {
/*
* Control registers have unknown contents.
* Can't recover and therefore stopping machine.
*/
s390_handle_damage();
}
if (!mci.fp) {
/*
* Floating point registers can't be restored. If the
* kernel currently uses floating point registers the
* system is stopped. If the process has its floating
* pointer registers loaded it is terminated.
*/
if (S390_lowcore.fpu_flags & KERNEL_VXR_V0V7)
s390_handle_damage();
if (!test_cpu_flag(CIF_FPU))
kill_task = 1;
}
if (!mci.fc) {
/*
* Floating point control register can't be restored.
* If the kernel currently uses the floating pointer
* registers and needs the FPC register the system is
* stopped. If the process has its floating pointer
* registers loaded it is terminated.
*/
if (S390_lowcore.fpu_flags & KERNEL_FPC)
s390_handle_damage();
if (!test_cpu_flag(CIF_FPU))
kill_task = 1;
}
if (MACHINE_HAS_VX) {
if (!mci.vr) {
/*
* Vector registers can't be restored. If the kernel
* currently uses vector registers the system is
* stopped. If the process has its vector registers
* loaded it is terminated.
*/
if (S390_lowcore.fpu_flags & KERNEL_VXR)
s390_handle_damage();
if (!test_cpu_flag(CIF_FPU))
kill_task = 1;
}
}
/* Check if access registers are valid */
if (!mci.ar) {
/*
* Access registers have unknown contents.
* Terminating task.
*/
kill_task = 1;
}
/* Check guarded storage registers */
cr2.val = S390_lowcore.cregs_save_area[2];
if (cr2.gse) {
if (!mci.gs) {
/*
* Guarded storage register can't be restored and
* the current processes uses guarded storage.
* It has to be terminated.
*/
kill_task = 1;
}
}
/* Check if old PSW is valid */
if (!mci.wp) {
/*
* Can't tell if we come from user or kernel mode
* -> stopping machine.
*/
s390_handle_damage();
}
/* Check for invalid kernel instruction address */
if (!mci.ia && !umode) {
/*
* The instruction address got lost while running
* in the kernel -> stopping machine.
*/
s390_handle_damage();
}
if (!mci.ms || !mci.pm || !mci.ia)
kill_task = 1;
return kill_task;
}
NOKPROBE_SYMBOL(s390_check_registers);
/*
* Backup the guest's machine check info to its description block
*/
static void notrace s390_backup_mcck_info(struct pt_regs *regs)
{
struct mcck_volatile_info *mcck_backup;
struct sie_page *sie_page;
/* r14 contains the sie block, which was set in sie64a */
struct kvm_s390_sie_block *sie_block =
(struct kvm_s390_sie_block *) regs->gprs[14];
if (sie_block == NULL
|