aboutsummaryrefslogtreecommitdiffstats
path: root/arch/sh/kernel/perf_event.c
blob: 10b14e3a7eb8142bd88a81ddd9c3a9cb96941a36 (plain) (blame)
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
/*
 * Performance event support framework for SuperH hardware counters.
 *
 *  Copyright (C) 2009  Paul Mundt
 *
 * Heavily based on the x86 and PowerPC implementations.
 *
 * x86:
 *  Copyright (C) 2008 Thomas Gleixner <tglx@linutronix.de>
 *  Copyright (C) 2008-2009 Red Hat, Inc., Ingo Molnar
 *  Copyright (C) 2009 Jaswinder Singh Rajput
 *  Copyright (C) 2009 Advanced Micro Devices, Inc., Robert Richter
 *  Copyright (C) 2008-2009 Red Hat, Inc., Peter Zijlstra <pzijlstr@redhat.com>
 *  Copyright (C) 2009 Intel Corporation, <markus.t.metzger@intel.com>
 *
 * ppc:
 *  Copyright 2008-2009 Paul Mackerras, IBM Corporation.
 *
 * This file is subject to the terms and conditions of the GNU General Public
 * License.  See the file "COPYING" in the main directory of this archive
 * for more details.
 */
#include <linux/kernel.h>
#include <linux/init.h>
#include <linux/io.h>
#include <linux/irq.h>
#include <linux/perf_event.h>
#include <linux/export.h>
#include <asm/processor.h>

struct cpu_hw_events {
	struct perf_event	*events[MAX_HWEVENTS];
	unsigned long		used_mask[BITS_TO_LONGS(MAX_HWEVENTS)];
	unsigned long		active_mask[BITS_TO_LONGS(MAX_HWEVENTS)];
};

DEFINE_PER_CPU(struct cpu_hw_events, cpu_hw_events);

static struct sh_pmu *sh_pmu __read_mostly;

/* Number of perf_events counting hardware events */
static atomic_t num_events;
/* Used to avoid races in calling reserve/release_pmc_hardware */
static DEFINE_MUTEX(pmc_reserve_mutex);

/*
 * Stub these out for now, do something more profound later.
 */
int reserve_pmc_hardware(void)
{
	return 0;
}

void release_pmc_hardware(void)
{
}

static inline int sh_pmu_initialized(void)
{
	return !!sh_pmu;
}

const char *perf_pmu_name(void)
{
	if (!sh_pmu)
		return NULL;

	return sh_pmu->name;
}
EXPORT_SYMBOL_GPL(perf_pmu_name);

int perf_num_counters(void)
{
	if (!sh_pmu)
		return 0;

	return sh_pmu->num_events;
}
EXPORT_SYMBOL_GPL(perf_num_counters);

/*
 * Release the PMU if this is the last perf_event.
 */
static void hw_perf_event_destroy(struct perf_event *event)
{
	if (!atomic_add_unless(&num_events, -1, 1)) {
		mutex_lock(&pmc_reserve_mutex);
		if (atomic_dec_return(&num_events) == 0)
			release_pmc_hardware();
		mutex_unlock(&pmc_reserve_mutex);
	}
}

static int hw_perf_cache_event(int config, int *evp)
{
	unsigned long type, op, result;
	int ev;

	if (!sh_pmu->cache_events)
		return -EINVAL;

	/* unpack config */
	type = config & 0xff;
	op = (config >> 8) & 0xff;
	result = (config >> 16) & 0xff;

	if (type >= PERF_COUNT_HW_CACHE_MAX ||
	    op >= PERF_COUNT_HW_CACHE_OP_MAX ||
	    result >= PERF_COUNT_HW_CACHE_RESULT_MAX)
		return -EINVAL;

	ev = (*sh_pmu->cache_events)[type][op][result];
	if (ev == 0)
		return -EOPNOTSUPP;
	if (ev == -1)
		return -EINVAL;
	*evp = ev;
	return 0;
}

static int __hw_perf_event_init(struct perf_event *event)
{
	struct perf_event_attr *attr = &event->attr;
	struct hw_perf_event *hwc = &event->hw;
	int config = -1;
	int err;

	if (!sh_pmu_initialized())
		return -ENODEV;

	/*
	 * All of the on-chip counters are "limited", in that they have
	 * no interrupts, and are therefore unable to do sampling without
	 * further work and timer assistance.
	 */
	if (hwc->sample_period)
		return -EINVAL;

	/*
	 * See if we need to reserve the counter.
	 *
	 * If no events are currently in use, then we have to take a
	 * mutex to ensure that we don't race with another task doing
	 * reserve_pmc_hardware or release_pmc_hardware.
	 */
	err = 0;
	if (!atomic_inc_not_zero(&num_events)) {
		mutex_lock(&pmc_reserve_mutex);
		if (atomic_read(&num_events) == 0 &&
		    reserve_pmc_hardware())
			err = -EBUSY;
		else
			atomic_inc(&num_events);
		mutex_unlock(&pmc_reserve_mutex);
	}

	if (err)
		return err;

	event->destroy = hw_perf_event_destroy;

	switch (attr->type) {
	case PERF_TYPE_RAW:
		config = attr->config & sh_pmu->raw_event_mask;
		break;
	case PERF_TYPE_HW_CACHE:
		err = hw_perf_cache_event(attr->config, &config);
		if (err)
			return err;
		break;
	case PERF_TYPE_HARDWARE:
		if (attr->config >= sh_pmu->max_events)
			return -EINVAL;

		config = sh_pmu->event_map(attr->config);
		break;
	}

	if (config == -1)
		return -EINVAL;

	hwc->config |= config;

	return 0;
}

static void sh_perf_event_update(struct perf_event *event,
				   struct hw_perf_event *hwc, int idx)
{
	u64 prev_raw_count, new_raw_count;
	s64 delta;
	int shift = 0;

	/*
	 * Depending on the counter configuration, they may or may not
	 * be chained, in which case the previous counter value can be
	 * updated underneath us if the lower-half overflows.
	 *
	 * Our tactic to handle this is to first atomically read and
	 * exchange a new raw count - then add that new-prev delta
	 * count to the generic counter atomically.
	 *
	 * As there is no interrupt associated with the overflow events,
	 * this is the simplest approach for maintaining consistency.
	 */
again:
	prev_raw_count = local64_read(&hwc->prev_count);
	new_raw_count = sh_pmu->read(idx);

	if (local64_cmpxchg(&hwc->prev_count, prev_raw_count,
			     new_raw_count) != prev_raw_count)
		goto again;

	/*
	 * Now we have the new raw value and have updated the prev
	 * timestamp already. We can now calculate the elapsed delta
	 * (counter-)time and add that to the generic counter.
	 *
	 * Careful, not all hw sign-extends above the physical width
	 * of the count.
	 */
	delta = (new_raw_count << shift) - (prev_raw_count << shift);
	delta >>= shift;

	local64_add(delta, &event->count);
}

static void sh_pmu_stop(struct perf_event *event, int flags)
{
	struct cpu_hw_events *cpuc = &__get_cpu_var(cpu_hw_events);
	struct hw_perf_event *hwc = &event->hw;
	int idx = hwc->idx;

	if (!(event->hw.state & PERF_HES_STOPPED)) {
		sh_pmu->disable(hwc, idx);
		cpuc->events[idx] = NULL;
		event->hw.state |= PERF_HES_STOPPED;
	}

	if ((flags & PERF_EF_UPDATE) && !(event->hw.state & PERF_HES_UPTODATE)) {
		sh_perf_event_update(event, &event->hw, idx);
		event->hw.state |= PERF_HES_UPTODATE;
	}
}

static void sh_pmu_start(struct perf_event *event, int flags)
{
	struct cpu_hw_events *cpuc = &__get_cpu_var(cpu_hw_events);
	struct hw_perf_event *hwc = &event->hw;
	int idx = hwc->idx;

	if (WARN_ON_ONCE(idx == -1))
		return;

	if (flags & PERF_EF_RELOAD)
		WARN_ON_ONCE(!(event->hw.state & PERF_HES_UPTODATE));

	cpuc->events[idx] = event;
	event->hw.state = 0;
	sh_pmu->enable(hwc, idx);
}

static void sh_pmu_del(struct perf_event *event, int flags)
{
	struct cpu_hw_events *cpuc = &__get_cpu_var(cpu_hw_events);

	sh_pmu_stop(event, PERF_EF_UPDATE);
	__clear_bit(event->hw.idx, cpuc->used_mask);

	perf_event_update_userpage(event);
}

static int sh_pmu_add(struct perf_event *event, int flags)
{
	struct cpu_hw_events *cpuc = &__get_cpu_var(cpu_hw_events);
	struct hw_perf_event *hwc = &event->hw;
	int idx = hwc->idx;
	int ret = -EAGAIN;

	perf_pmu_disable(event->pmu);

	if (__test_and_set_bit(idx, cpuc->used_mask)) {
		idx = find_first_zero_bit(cpuc->used_mask, sh_pmu->num_events);
		if (idx == sh_pmu->num_events)
			goto out;

		__set_bit(idx, cpuc->used_mask);
		hwc->idx = idx;
	}

	sh_pmu->disable(hwc, idx);

	event->hw.state = PERF_HES_UPTODATE | PERF_HES_STOPPED;
	if (flags & PERF_EF_START)
		sh_pmu_start(event, PERF_EF_RELOAD);

	perf_event_update_userpage(event);
	ret = 0;
out:
	perf_pmu_enable(event->pmu);
	return ret;
}

static void sh_pmu_read(struct perf_event *event)
{
	sh_perf_event_update(event, &event->hw, event->hw.idx);
}

static int sh_pmu_event_init(struct perf_event *event)
{
	int err;

	switch (event->attr.type) {
	case PERF_TYPE_RAW:
	case PERF_TYPE_HW_CACHE:
	case PERF_TYPE_HARDWARE:
		err = __hw_perf_event_init(event);
		break;

	default:
		return -ENOENT;
	}

	if (unlikely(err)) {
		if (event->destroy)
			event->destroy(event);
	}

	return err;
}

static void sh_pmu_enable(struct pmu *pmu)
{
	if (!sh_pmu_initialized())
		return;

	sh_pmu->enable_all();
}

static void sh_pmu_disable(struct pmu *pmu)
{
	if (!sh_pmu_initialized())
		return;

	sh_pmu->disable_all();
}

static struct pmu pmu = {
	.pmu_enable	= sh_pmu_enable,
	.pmu_disable	= sh_pmu_disable,
	.event_init	= sh_pmu_event_init,
	.add		= sh_pmu_add,
	.del		= sh_pmu_del,
	.start		= sh_pmu_start,
	.stop		= sh_pmu_stop,
	.read		= sh_pmu_read,
};

static void sh_pmu_setup(int cpu)
{
	struct cpu_hw_events *cpuhw = &per_cpu(cpu_hw_events, cpu);

	memset(cpuhw, 0, sizeof(struct cpu_hw_events));
}

static int __cpuinit
sh_pmu_notifier(struct notifier_block *self, unsigned long action, void *hcpu)
{
	unsigned int cpu = (long)hcpu;

	switch (action & ~CPU_TASKS_FROZEN) {
	case CPU_UP_PREPARE:
		sh_pmu_setup(cpu);
		break;

	default:
		break;
	}

	return NOTIFY_OK;
}

int __cpuinit register_sh_pmu(struct sh_pmu *_pmu)
{
	if (sh_pmu)
		return -EBUSY;
	sh_pmu = _pmu;

	pr_info("Performance Events: %s support registered\n", _pmu->name);

	WARN_ON(_pmu->num_events > MAX_HWEVENTS);

	perf_pmu_register(&pmu, "cpu", PERF_TYPE_RAW);
	perf_cpu_notifier(sh_pmu_notifier);
	return 0;
}