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
|
/*
* 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/time.h>
#include <linux/module.h>
#include <asm/lowcore.h>
#include <asm/smp.h>
#include <asm/etr.h>
#include <asm/cputime.h>
#include <asm/nmi.h>
#include <asm/crw.h>
#include <asm/switch_to.h>
struct mcck_struct {
int kill_task;
int channel_report;
int warning;
unsigned long long mcck_code;
};
static DEFINE_PER_CPU(struct mcck_struct, cpu_mcck);
static void s390_handle_damage(char *msg)
{
smp_send_stop();
disabled_wait((unsigned long) __builtin_return_address(0));
while (1);
}
/*
* 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 = __get_cpu_var(cpu_mcck);
memset(&__get_cpu_var(cpu_mcck), 0, sizeof(struct mcck_struct));
clear_cpu_flag(CIF_MCCK_PENDING);
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.kill_task) {
local_irq_enable();
printk(KERN_EMERG "mcck: Terminating task because of machine "
"malfunction (code 0x%016llx).\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 registers could be validated
* returns 1 otherwise
*/
static int notrace s390_revalidate_registers(struct mci *mci)
{
int kill_task;
u64 zero;
void *fpt_save_area, *fpt_creg_save_area;
kill_task = 0;
zero = 0;
if (!mci->gr) {
/*
* General purpose registers couldn't be restored and have
* unknown contents. Process needs to be terminated.
*/
kill_task = 1;
}
if (!mci->fp) {
/*
* Floating point registers can't be restored and
* therefore the process needs to be terminated.
*/
kill_task = 1;
}
#ifndef CONFIG_64BIT
asm volatile(
" ld 0,0(%0)\n"
" ld 2,8(%0)\n"
" ld 4,16(%0)\n"
" ld 6,24(%0)"
: : "a" (&S390_lowcore.floating_pt_save_area));
#endif
if (MACHINE_HAS_IEEE) {
#ifdef CONFIG_64BIT
fpt_save_area = &S390_lowcore.floating_pt_save_area;
fpt_creg_save_area = &S390_lowcore.fpt_creg_save_area;
#else
fpt_save_area = (void *) S390_lowcore.extended_save_area_addr;
fpt_creg_save_area = fpt_save_area + 128;
#endif
if (!mci->fc) {
/*
* Floating point control register can't be restored.
* Task will be terminated.
*/
asm volatile("lfpc 0(%0)" : : "a" (&zero), "m" (zero));
kill_task = 1;
} else
asm volatile("lfpc 0(%0)" : : "a" (fpt_creg_save_area));
asm volatile(
" ld 0,0(%0)\n"
" ld 1,8(%0)\n"
" ld 2,16(%0)\n"
" ld 3,24(%0)\n"
" ld 4,32(%0)\n"
" ld 5,40(%0)\n"
" ld 6,48(%0)\n"
" ld 7,56(%0)\n"
" ld 8,64(%0)\n"
" ld 9,72(%0)\n"
" ld 10,80(%0)\n"
" ld 11,88(%0)\n"
" ld 12,96(%0)\n"
" ld 13,104(%0)\n"
" ld 14,112(%0)\n"
" ld 15,120(%0)\n"
: : "a" (fpt_save_area));
}
#ifdef CONFIG_64BIT
/* Revalidate vector registers */
if (MACHINE_HAS_VX && current->thread.vxrs) {
if (!mci->vr) {
/*
* Vector registers can't be restored and therefore
* the process needs to be terminated.
*/
kill_task = 1;
}
restore_vx_regs((__vector128 *)
S390_lowcore.vector_save_area_addr);
}
#endif
/* Revalidate access registers */
asm volatile(
" lam 0,15,0(%0)"
: : "a" (&S390_lowcore.access_regs_save_area));
if (!mci->ar) {
/*
* Access registers have unknown contents.
* Terminating task.
*/
kill_task = 1;
}
/* Revalidate control registers */
if (!mci->cr) {
/*
* Control registers have unknown contents.
* Can't recover and therefore stopping machine.
*/
s390_handle_damage("invalid control registers.");
} else {
#ifdef CONFIG_64BIT
asm volatile(
" lctlg 0,15,0(%0)"
: : "a" (&S390_lowcore.cregs_save_area));
#else
asm volatile(
" lctl 0,15,0(%0)"
: : "a" (&S390_lowcore.cregs_save_area));
#endif
}
/*
* We don't even try to revalidate the TOD register, since we simply
* can't write something sensible into that register.
*/
#ifdef CONFIG_64BIT
/*
* See if we can revalidate the TOD programmable register with its
* old contents (should be zero) otherwise set it to zero.
*/
if (!mci->pr)
asm volatile(
" sr 0,0\n"
" sckpf"
: : : "0", "cc");
else
asm volatile(
" l 0,0(%0)\n"
" sckpf"
: : "a" (&S390_lowcore.tod_progreg_save_area)
: "0", "cc");
#endif
/* Revalidate clock comparator register */
set_clock_comparator(S390_lowcore.clock_comparator);
/* 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("old psw invalid.");
if (!mci->ms || !mci->pm || !mci->ia)
kill_task = 1;
return kill_task;
}
#define MAX_IPD_COUNT 29
#define MAX_IPD_TIME (5 * 60 * USEC_PER_SEC) /* 5 minutes */
#define ED_STP_ISLAND 6 /* External damage STP island check */
#define ED_STP_SYNC 7 /* External damage STP sync check */
#define ED_ETR_SYNC 12 /* External damage ETR sync check */
#define ED_ETR_SWITCH 13 /* External damage ETR switch to local */
/*
* machine check handler.
*/
void notrace s390_do_machine_check(struct pt_regs *regs)
{
static int ipd_count;
static DEFINE_SPINLOCK(ipd_lock);
static unsigned long long last_ipd;
struct mcck_struct *mcck;
unsigned long long tmp;
struct mci *mci;
int umode;
nmi_enter();
inc_irq_stat(NMI_NMI);
mci = (struct mci *) &S390_lowcore.mcck_interruption_code;
mcck = &__get_cpu_var(cpu_mcck);
umode = user_mode(regs);
if (mci->sd) {
/* System damage -> stopping machine */
s390_handle_damage("received system damage machine check.");
}
if (mci->pd) {
if (mci->b) {
/* Processing backup -> verify if we can survive this */
u64 z_mcic, o_mcic, t_mcic;
#ifdef CONFIG_64BIT
z_mcic = (1ULL<<63 | 1ULL<<59 | 1ULL<<29);
o_mcic = (1ULL<<43 | 1ULL<<42 | 1ULL<<41 | 1ULL<<40 |
1ULL<<36 | 1ULL<<35 | 1ULL<<34 | 1ULL<<32 |
1ULL<<30 | 1ULL<<21 | 1ULL<<20 | 1ULL<<17 |
1ULL<<16);
#else
z_mcic = (1ULL<<63 | 1ULL<<59 | 1ULL<<57 | 1ULL<<50 |
1ULL<<29);
o_mcic = (1ULL<<43 | 1ULL<<42 | 1ULL<<41 | 1ULL<<40 |
1ULL<<36 | 1ULL<<35 | 1ULL<<34 | 1ULL<<32 |
1ULL<<30 | 1ULL<<20 | 1ULL<<17 | 1ULL<<16);
#endif
t_mcic = *(u64 *)mci;
if (((t_mcic & z_mcic) != 0) ||
((t_mcic & o_mcic) != o_mcic)) {
s390_handle_damage("processing backup machine "
"check with damage.");
}
/*
* Nullifying exigent condition, therefore we might
* retry this instruction.
*/
spin_lock(&ipd_lock);
tmp = get_tod_clock();
if (((tmp - last_ipd) >> 12) < MAX_IPD_TIME)
|