summaryrefslogtreecommitdiffstats
path: root/kernel/arch/s390/oprofile/init.c
blob: bc927a09a172b6d6699cf60610cde600d4c73e74 (plain)
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
/*
 * S390 Version
 *   Copyright IBM Corp. 2002, 2011
 *   Author(s): Thomas Spatzier (tspat@de.ibm.com)
 *   Author(s): Mahesh Salgaonkar (mahesh@linux.vnet.ibm.com)
 *   Author(s): Heinz Graalfs (graalfs@linux.vnet.ibm.com)
 *   Author(s): Andreas Krebbel (krebbel@linux.vnet.ibm.com)
 *
 * @remark Copyright 2002-2011 OProfile authors
 */

#include <linux/oprofile.h>
#include <linux/perf_event.h>
#include <linux/init.h>
#include <linux/errno.h>
#include <linux/fs.h>
#include <linux/module.h>
#include <asm/processor.h>

#include "../../../drivers/oprofile/oprof.h"

extern void s390_backtrace(struct pt_regs * const regs, unsigned int depth);

#include "hwsampler.h"
#include "op_counter.h"

#define DEFAULT_INTERVAL	4127518

#define DEFAULT_SDBT_BLOCKS	1
#define DEFAULT_SDB_BLOCKS	511

static unsigned long oprofile_hw_interval = DEFAULT_INTERVAL;
static unsigned long oprofile_min_interval;
static unsigned long oprofile_max_interval;

static unsigned long oprofile_sdbt_blocks = DEFAULT_SDBT_BLOCKS;
static unsigned long oprofile_sdb_blocks = DEFAULT_SDB_BLOCKS;

static int hwsampler_enabled;
static int hwsampler_running;	/* start_mutex must be held to change */
static int hwsampler_available;

static struct oprofile_operations timer_ops;

struct op_counter_config counter_config;

enum __force_cpu_type {
	reserved = 0,		/* do not force */
	timer,
};
static int force_cpu_type;

static int set_cpu_type(const char *str, struct kernel_param *kp)
{
	if (!strcmp(str, "timer")) {
		force_cpu_type = timer;
		printk(KERN_INFO "oprofile: forcing timer to be returned "
		                 "as cpu type\n");
	} else {
		force_cpu_type = 0;
	}

	return 0;
}
module_param_call(cpu_type, set_cpu_type, NULL, NULL, 0);
MODULE_PARM_DESC(cpu_type, "Force legacy basic mode sampling"
		           "(report cpu_type \"timer\"");

static int __oprofile_hwsampler_start(void)
{
	int retval;

	retval = hwsampler_allocate(oprofile_sdbt_blocks, oprofile_sdb_blocks);
	if (retval)
		return retval;

	retval = hwsampler_start_all(oprofile_hw_interval);
	if (retval)
		hwsampler_deallocate();

	return retval;
}

static int oprofile_hwsampler_start(void)
{
	int retval;

	hwsampler_running = hwsampler_enabled;

	if (!hwsampler_running)
		return timer_ops.start();

	retval = perf_reserve_sampling();
	if (retval)
		return retval;

	retval = __oprofile_hwsampler_start();
	if (retval)
		perf_release_sampling();

	return retval;
}

static void oprofile_hwsampler_stop(void)
{
	if (!hwsampler_running) {
		timer_ops.stop();
		return;
	}

	hwsampler_stop_all();
	hwsampler_deallocate();
	perf_release_sampling();
	return;
}

/*
 * File ops used for:
 * /dev/oprofile/0/enabled
 * /dev/oprofile/hwsampling/hwsampler  (cpu_type = timer)
 */

static ssize_t hwsampler_read(struct file *file, char __user *buf,
		size_t count, loff_t *offset)
{
	return oprofilefs_ulong_to_user(hwsampler_enabled, buf, count, offset);
}

static ssize_t hwsampler_write(struct file *file, char const __user *buf,
		size_t count, loff_t *offset)
{
	unsigned long val;
	int retval;

	if (*offset)
		return -EINVAL;

	retval = oprofilefs_ulong_from_user(&val, buf, count);
	if (retval <= 0)
		return retval;

	if (val != 0 && val != 1)
		return -EINVAL;

	if (oprofile_started)
		/*
		 * save to do without locking as we set
		 * hwsampler_running in start() when start_mutex is
		 * held
		 */
		return -EBUSY;

	hwsampler_enabled = val;

	return count;
}

static const struct file_operations hwsampler_fops = {
	.read		= hwsampler_read,
	.write		= hwsampler_write,
};

/*
 * File ops used for:
 * /dev/oprofile/0/count
 * /dev/oprofile/hwsampling/hw_interval  (cpu_type = timer)
 *
 * Make sure that the value is within the hardware range.
 */

static ssize_t hw_interval_read(struct file *file, char __user *buf,
				size_t count, loff_t *offset)
{
	return oprofilefs_ulong_to_user(oprofile_hw_interval, buf,
					count, offset);
}

static ssize_t hw_interval_write(struct file *file, char const __user *buf,
				 size_t count, loff_t *offset)
{
	unsigned long val;
	int retval;

	if (*offset)
		return -EINVAL;
	retval = oprofilefs_ulong_from_user(&val, buf, count);
	if (retval <= 0)
		return retval;
	if (val < oprofile_min_interval)
		oprofile_hw_interval = oprofile_min_interval;
	else if (val > oprofile_max_interval)
		oprofile_hw_interval = oprofile_max_interval;
	else
		oprofile_hw_interval = val;

	return count;
}

static const struct file_operations hw_interval_fops = {
	.read		= hw_interval_read,
	.write		= hw_interval_write,
};

/*
 * File ops used for:
 * /dev/oprofile/0/event
 * Only a single event with number 0 is supported with this counter.
 *
 * /dev/oprofile/0/unit_mask
 * This is a dummy file needed by the user space tools.
 * No value other than 0 is accepted or returned.
 */

static ssize_t hwsampler_zero_read(struct file *file, char __user *buf,
				    size_t count, loff_t *offset)
{
	return oprofilefs_ulong_to_user(0, buf, count, offset);
}

static ssize_t hwsampler_zero_write(struct file *file, char const __user *buf,
				     size_t count, loff_t *offset)
{
	unsigned long val;
	int retval;

	if (*offset)
		return -EINVAL;

	retval = oprofilefs_ulong_from_user(&val, buf, count);
	if (retval <= 0)
		return retval;
	if (val != 0)
		return -EINVAL;
	return count;
}

static const struct file_operations zero_fops = {
	.read		= hwsampler_zero_read,
	.write		= hwsampler_zero_write,
};

/* /dev/oprofile/0/kernel file ops.  */

static ssize_t hwsampler_kernel_read(struct file *file, char __user *buf,
				     size_t count, loff_t *offset)
{
	return oprofilefs_ulong_to_user(counter_config.kernel,
					buf, count, offset);
}

static ssize_t hwsampler_kernel_write(struct file *file, char const __user *buf,
				      size_t count, loff_t *offset)
{
	unsigned long val;
	int retval;

	if (*offset)
		return -EINVAL;

	retval = oprofilefs_ulong_from_user(&val, buf, count);
	if (retval <= 0)
		return retval;

	if (val != 0 && val != 1)
		return -EINVAL;

	counter_config.kernel = val;

	return count;
}

static const struct file_operations kernel_fops = {
	.read		= hwsampler_kernel_read,
	.write		= hwsampler_kernel_write,
};

/* /dev/oprofile/0/user file ops. */

static ssize_t hwsampler_user_read(struct file *file, char __user *buf,
				   size_t count, loff_t *offset)
{
	return oprofilefs_ulong_to_user(counter_config.user,
					buf, count, offset);
}

static ssize_t hwsampler_user_write(struct file *file, char const __user *buf,
				    size_t count, loff_t *offset)
{
	unsigned long val;
	int retval;

	if (*offset)
		return -EINVAL;

	retval = oprofilefs_ulong_from_user(&val, buf, count);
	if (retval <= 0)
		return retval;

	if (val != 0 && val != 1)
		return -EINVAL;

	counter_config.user = val;

	return count;
}

static const struct file_operations user_fops = {
	.read		= hwsampler_user_read,
	.write		= hwsampler_user_write,
};


/*
 * File ops used for: /dev/oprofile/timer/enabled
 * The value always has to be the inverted value of hwsampler_enabled. So
 * no separate variable is created. That way we do not need locking.
 */

static ssize_t timer_enabled_read(struct file *file, char __user *buf,
				  size_t count, loff_t *offset)
{
	return oprofilefs_ulong_to_user(!hwsampler_enabled, buf, count, offset);
}

static ssize_t timer_enabled_write(struct file *file, char const __user *buf,
				   size_t count, loff_t *offset)
{
	unsigned long val;
	int retval;

	if (*offset)
		return -EINVAL;

	retval = oprofilefs_ulong_from_user(&val, buf, count);
	if (retval <= 0)
		return retval;

	if (val != 0 && val != 1)
		return -EINVAL;

	/* Timer cannot be disabled without having hardware sampling.  */
	if (val == 0 && !hwsampler_available)
		return -EINVAL;

	if (oprofile_started)
		/*
		 * save to do without locking as we set
		 * hwsampler_running in start() when start_mutex is
		 * held
		 */
		return -EBUSY;

	hwsampler_enabled = !val;

	return count;
}

static const struct file_operations timer_enabled_fops = {
	.read		= timer_enabled_read,
	.write		= timer_enabled_write,
};


static int oprofile_create_hwsampling_files(struct dentry *root)
{
	struct dentry *dir;

	dir = oprofilefs_mkdir(root, "timer");
	if (!dir)
		return -EINVAL;

	oprofilefs_create_file(dir, "enabled", &timer_enabled_fops);

	if (!hwsampler_available)
		return 0;

	/* reinitialize default values */
	hwsampler_enabled = 1;
	counter_config.kernel = 1;
	counter_config.user = 1;

	if (!force_cpu_type) {
		/*
		 * Create the counter file system.  A single virtual
		 * counter is created which can be used to
		 * enable/disable hardware sampling dynamically from
		 * user space.  The user space will configure a single
		 * counter with a single event.  The value of 'event'
		 * and 'unit_mask' are not evaluated by the kernel code
		 * and can only be set to 0.
		 */

		dir = oprofilefs_mkdir(root, "0");
		if (!dir)
			return -EINVAL;

		oprofilefs_create_file(dir, "enabled", &hwsampler_fops);
		oprofilefs_create_file(dir, "event", &zero_fops);
		oprofilefs_create_file(dir, "count", &hw_interval_fops);
		oprofilefs_create_file(dir, "unit_mask", &zero_fops);
		oprofilefs_create_file(dir, "kernel", &kernel_fops);
		oprofilefs_create_file(dir, "user", &user_fops);
		oprofilefs_create_ulong(dir, "hw_sdbt_blocks",
					&oprofile_sdbt_blocks);

	} else {
		/*
		 * Hardware sampling can be used but the cpu_type is
		 * forced to timer in order to deal with legacy user
		 * space tools.  The /dev/oprofile/hwsampling fs is
		 * provided in that case.
		 */
		dir = oprofilefs_mkdir(root, "hwsampling");
		if (!dir)
			return -EINVAL;

		oprofilefs_create_file(dir, "hwsampler",
				       &hwsampler_fops);
		oprofilefs_create_file(dir, "hw_interval",
				       &hw_interval_fops);
		oprofilefs_create_ro_ulong(dir, "hw_min_interval",
					   &oprofile_min_interval);
		oprofilefs_create_ro_ulong(dir, "hw_max_interval",
					   &oprofile_max_interval);
		oprofilefs_create_ulong(dir, "hw_sdbt_blocks",
					&oprofile_sdbt_blocks);
	}
	return 0;
}

static int oprofile_hwsampler_init(struct oprofile_operations *ops)
{
	/*
	 * Initialize the timer mode infrastructure as well in order
	 * to be able to switch back dynamically.  oprofile_timer_init
	 * is not supposed to fail.
	 */
	if (oprofile_timer_init(ops))
		BUG();

	memcpy(&timer_ops, ops, sizeof(timer_ops));
	ops->create_files = oprofile_create_hwsampling_files;

	/*
	 * If the user space tools do not support newer cpu types,
	 * the force_cpu_type module parameter
	 * can be used to always return \"timer\" as cpu type.
	 */
	if (force_cpu_type != timer) {
		struct cpuid id;

		get_cpu_id (&id);

		switch (id.machine) {
		case 0x2097: case 0x2098: ops->cpu_type = "s390/z10"; break;
		case 0x2817: case 0x2818: ops->cpu_type = "s390/z196"; break;
		case 0x2827: case 0x2828: ops->cpu_type = "s390/zEC12"; break;
		default: return -ENODEV;
		}
	}

	if (hwsampler_setup())
		return -ENODEV;

	/*
	 * Query the range for the sampling interval from the
	 * hardware.
	 */
	oprofile_min_interval = hwsampler_query_min_interval();
	if (oprofile_min_interval == 0)
		return -ENODEV;
	oprofile_max_interval = hwsampler_query_max_interval();
	if (oprofile_max_interval == 0)
		return -ENODEV;

	/* The initial value should be sane */
	if (oprofile_hw_interval < oprofile_min_interval)
		oprofile_hw_interval = oprofile_min_interval;
	if (oprofile_hw_interval > oprofile_max_interval)
		oprofile_hw_interval = oprofile_max_interval;

	printk(KERN_INFO "oprofile: System z hardware sampling "
	       "facility found.\n");

	ops->start = oprofile_hwsampler_start;
	ops->stop = oprofile_hwsampler_stop;

	return 0;
}

static void oprofile_hwsampler_exit(void)
{
	hwsampler_shutdown();
}

int __init oprofile_arch_init(struct oprofile_operations *ops)
{
	ops->backtrace = s390_backtrace;

	/*
	 * -ENODEV is not reported to the caller.  The module itself
         * will use the timer mode sampling as fallback and this is
         * always available.
	 */
	hwsampler_available = oprofile_hwsampler_init(ops) == 0;

	return 0;
}

void oprofile_arch_exit(void)
{
	oprofile_hwsampler_exit();
}