summaryrefslogtreecommitdiff
path: root/drivers/hwtracing/coresight/coresight-etm-perf.c
blob: 755125f7917f2cb6aa369fd88d28d763e688a9d6 (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
/*
 * Copyright(C) 2015 Linaro Limited. All rights reserved.
 * Author: Mathieu Poirier <mathieu.poirier@linaro.org>
 *
 * This program is free software; you can redistribute it and/or modify it
 * under the terms of the GNU General Public License version 2 as published by
 * the Free Software Foundation.
 *
 * This program is distributed in the hope that it will be useful, but WITHOUT
 * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
 * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
 * more details.
 *
 * You should have received a copy of the GNU General Public License along with
 * this program.  If not, see <http://www.gnu.org/licenses/>.
 */

#include <linux/coresight.h>
#include <linux/coresight-pmu.h>
#include <linux/cpumask.h>
#include <linux/device.h>
#include <linux/list.h>
#include <linux/mm.h>
#include <linux/init.h>
#include <linux/perf_event.h>
#include <linux/slab.h>
#include <linux/types.h>
#include <linux/workqueue.h>

#include "coresight-priv.h"

static struct pmu etm_pmu;
static bool etm_perf_up;

/**
 * struct etm_event_data - Coresight specifics associated to an event
 * @work:		Handle to free allocated memory outside IRQ context.
 * @mask:		Hold the CPU(s) this event was set for.
 * @snk_config:		The sink configuration.
 * @path:		An array of path, each slot for one CPU.
 */
struct etm_event_data {
	struct work_struct work;
	cpumask_t mask;
	void *snk_config;
	struct list_head **path;
};

static DEFINE_PER_CPU(struct perf_output_handle, ctx_handle);
static DEFINE_PER_CPU(struct coresight_device *, csdev_src);

/* ETMv3.5/PTM's ETMCR is 'config' */
PMU_FORMAT_ATTR(cycacc,		"config:" __stringify(ETM_OPT_CYCACC));
PMU_FORMAT_ATTR(timestamp,	"config:" __stringify(ETM_OPT_TS));

static struct attribute *etm_config_formats_attr[] = {
	&format_attr_cycacc.attr,
	&format_attr_timestamp.attr,
	NULL,
};

static struct attribute_group etm_pmu_format_group = {
	.name   = "format",
	.attrs  = etm_config_formats_attr,
};

static const struct attribute_group *etm_pmu_attr_groups[] = {
	&etm_pmu_format_group,
	NULL,
};

static void etm_event_read(struct perf_event *event) {}

static int etm_event_init(struct perf_event *event)
{
	if (event->attr.type != etm_pmu.type)
		return -ENOENT;

	return 0;
}

static void free_event_data(struct work_struct *work)
{
	int cpu;
	cpumask_t *mask;
	struct etm_event_data *event_data;
	struct coresight_device *sink;

	event_data = container_of(work, struct etm_event_data, work);
	mask = &event_data->mask;
	/*
	 * First deal with the sink configuration.  See comment in
	 * etm_setup_aux() about why we take the first available path.
	 */
	if (event_data->snk_config) {
		cpu = cpumask_first(mask);
		sink = coresight_get_sink(event_data->path[cpu]);
		if (sink_ops(sink)->free_buffer)
			sink_ops(sink)->free_buffer(event_data->snk_config);
	}

	for_each_cpu(cpu, mask) {
		if (event_data->path[cpu])
			coresight_release_path(event_data->path[cpu]);
	}

	kfree(event_data->path);
	kfree(event_data);
}

static void *alloc_event_data(int cpu)
{
	int size;
	cpumask_t *mask;
	struct etm_event_data *event_data;

	/* First get memory for the session's data */
	event_data = kzalloc(sizeof(struct etm_event_data), GFP_KERNEL);
	if (!event_data)
		return NULL;

	/* Make sure nothing disappears under us */
	get_online_cpus();
	size = num_online_cpus();

	mask = &event_data->mask;
	if (cpu != -1)
		cpumask_set_cpu(cpu, mask);
	else
		cpumask_copy(mask, cpu_online_mask);
	put_online_cpus();

	/*
	 * Each CPU has a single path between source and destination.  As such
	 * allocate an array using CPU numbers as indexes.  That way a path
	 * for any CPU can easily be accessed at any given time.  We proceed
	 * the same way for sessions involving a single CPU.  The cost of
	 * unused memory when dealing with single CPU trace scenarios is small
	 * compared to the cost of searching through an optimized array.
	 */
	event_data->path = kcalloc(size,
				   sizeof(struct list_head *), GFP_KERNEL);
	if (!event_data->path) {
		kfree(event_data);
		return NULL;
	}

	return event_data;
}

static void etm_free_aux(void *data)
{
	struct etm_event_data *event_data = data;

	schedule_work(&event_data->work);
}

static void *etm_setup_aux(int event_cpu, void **pages,
			   int nr_pages, bool overwrite)
{
	int cpu;
	cpumask_t *mask;
	struct coresight_device *sink;
	struct etm_event_data *event_data = NULL;

	event_data = alloc_event_data(event_cpu);
	if (!event_data)
		return NULL;

	INIT_WORK(&event_data->work, free_event_data);

	mask = &event_data->mask;

	/* Setup the path for each CPU in a trace session */
	for_each_cpu(cpu, mask) {
		struct coresight_device *csdev;

		csdev = per_cpu(csdev_src, cpu);
		if (!csdev)
			goto err;

		/*
		 * Building a path doesn't enable it, it simply builds a
		 * list of devices from source to sink that can be
		 * referenced later when the path is actually needed.
		 */
		event_data->path[cpu] = coresight_build_path(csdev);
		if (!event_data->path[cpu])
			goto err;
	}

	/*
	 * In theory nothing prevent tracers in a trace session from being
	 * associated with different sinks, nor having a sink per tracer.  But
	 * until we have HW with this kind of topology and a way to convey
	 * sink assignement from the perf cmd line we need to assume tracers
	 * in a trace session are using the same sink.  Therefore pick the sink
	 * found at the end of the first available path.
	 */
	cpu = cpumask_first(mask);
	/* Grab the sink at the end of the path */
	sink = coresight_get_sink(event_data->path[cpu]);
	if (!sink)
		goto err;

	if (!sink_ops(sink)->alloc_buffer)
		goto err;

	/* Get the AUX specific data from the sink buffer */
	event_data->snk_config =
			sink_ops(sink)->alloc_buffer(sink, cpu, pages,
						     nr_pages, overwrite);
	if (!event_data->snk_config)
		goto err;

out:
	return event_data;

err:
	etm_free_aux(event_data);
	event_data = NULL;
	goto out;
}

static void etm_event_start(struct perf_event *event, int flags)
{
	int cpu = smp_processor_id();
	struct etm_event_data *event_data;
	struct perf_output_handle *handle = this_cpu_ptr(&ctx_handle);
	struct coresight_device *sink, *csdev = per_cpu(csdev_src, cpu);

	if (!csdev)
		goto fail;

	/*
	 * Deal with the ring buffer API and get a handle on the
	 * session's information.
	 */
	event_data = perf_aux_output_begin(handle, event);
	if (!event_data)
		goto fail;

	/* We need a sink, no need to continue without one */
	sink = coresight_get_sink(event_data->path[cpu]);
	if (WARN_ON_ONCE(!sink || !sink_ops(sink)->set_buffer))
		goto fail_end_stop;

	/* Configure the sink */
	if (sink_ops(sink)->set_buffer(sink, handle,
				       event_data->snk_config))
		goto fail_end_stop;

	/* Nothing will happen without a path */
	if (coresight_enable_path(event_data->path[cpu], CS_MODE_PERF))
		goto fail_end_stop;

	/* Tell the perf core the event is alive */
	event->hw.state = 0;

	/* Finally enable the tracer */
	if (source_ops(csdev)->enable(csdev, &event->attr, CS_MODE_PERF))
		goto fail_end_stop;

out:
	return;

fail_end_stop:
	perf_aux_output_end(handle, 0, true);
fail:
	event->hw.state = PERF_HES_STOPPED;
	goto out;
}

static void etm_event_stop(struct perf_event *event, int mode)
{
	bool lost;
	int cpu = smp_processor_id();
	unsigned long size;
	struct coresight_device *sink, *csdev = per_cpu(csdev_src, cpu);
	struct perf_output_handle *handle = this_cpu_ptr(&ctx_handle);
	struct etm_event_data *event_data = perf_get_aux(handle);

	if (event->hw.state == PERF_HES_STOPPED)
		return;

	if (!csdev)
		return;

	sink = coresight_get_sink(event_data->path[cpu]);
	if (!sink)
		return;

	/* stop tracer */
	source_ops(csdev)->disable(csdev);

	/* tell the core */
	event->hw.state = PERF_HES_STOPPED;

	if (mode & PERF_EF_UPDATE) {
		if (WARN_ON_ONCE(handle->event != event))
			return;

		/* update trace information */
		if (!sink_ops(sink)->update_buffer)
			return;

		sink_ops(sink)->update_buffer(sink, handle,
					      event_data->snk_config);

		if (!sink_ops(sink)->reset_buffer)
			return;

		size = sink_ops(sink)->reset_buffer(sink, handle,
						    event_data->snk_config,
						    &lost);

		perf_aux_output_end(handle, size, lost);
	}

	/* Disabling the path make its elements available to other sessions */
	coresight_disable_path(event_data->path[cpu]);
}

static int etm_event_add(struct perf_event *event, int mode)
{
	int ret = 0;
	struct hw_perf_event *hwc = &event->hw;

	if (mode & PERF_EF_START) {
		etm_event_start(event, 0);
		if (hwc->state & PERF_HES_STOPPED)
			ret = -EINVAL;
	} else {
		hwc->state = PERF_HES_STOPPED;
	}

	return ret;
}

static void etm_event_del(struct perf_event *event, int mode)
{
	etm_event_stop(event, PERF_EF_UPDATE);
}

int etm_perf_symlink(struct coresight_device *csdev, bool link)
{
	char entry[sizeof("cpu9999999")];
	int ret = 0, cpu = source_ops(csdev)->cpu_id(csdev);
	struct device *pmu_dev = etm_pmu.dev;
	struct device *cs_dev = &csdev->dev;

	sprintf(entry, "cpu%d", cpu);

	if (!etm_perf_up)
		return -EPROBE_DEFER;

	if (link) {
		ret = sysfs_create_link(&pmu_dev->kobj, &cs_dev->kobj, entry);
		if (ret)
			return ret;
		per_cpu(csdev_src, cpu) = csdev;
	} else {
		sysfs_remove_link(&pmu_dev->kobj, entry);
		per_cpu(csdev_src, cpu) = NULL;
	}

	return 0;
}

static int __init etm_perf_init(void)
{
	int ret;

	etm_pmu.capabilities	= PERF_PMU_CAP_EXCLUSIVE;

	etm_pmu.attr_groups	= etm_pmu_attr_groups;
	etm_pmu.task_ctx_nr	= perf_sw_context;
	etm_pmu.read		= etm_event_read;
	etm_pmu.event_init	= etm_event_init;
	etm_pmu.setup_aux	= etm_setup_aux;
	etm_pmu.free_aux	= etm_free_aux;
	etm_pmu.start		= etm_event_start;
	etm_pmu.stop		= etm_event_stop;
	etm_pmu.add		= etm_event_add;
	etm_pmu.del		= etm_event_del;

	ret = perf_pmu_register(&etm_pmu, CORESIGHT_ETM_PMU_NAME, -1);
	if (ret == 0)
		etm_perf_up = true;

	return ret;
}
device_initcall(etm_perf_init);