summaryrefslogtreecommitdiff
path: root/drivers/gpu/drm/i915/gt/intel_hangcheck.c
blob: 6bcfa6456c4537befb7248e184165e157245eee7 (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
/*
 * Copyright © 2016 Intel Corporation
 *
 * Permission is hereby granted, free of charge, to any person obtaining a
 * copy of this software and associated documentation files (the "Software"),
 * to deal in the Software without restriction, including without limitation
 * the rights to use, copy, modify, merge, publish, distribute, sublicense,
 * and/or sell copies of the Software, and to permit persons to whom the
 * Software is furnished to do so, subject to the following conditions:
 *
 * The above copyright notice and this permission notice (including the next
 * paragraph) shall be included in all copies or substantial portions of the
 * Software.
 *
 * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
 * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
 * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
 * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
 * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
 * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
 * IN THE SOFTWARE.
 *
 */

#include "intel_reset.h"
#include "i915_drv.h"

struct hangcheck {
	u64 acthd;
	u32 ring;
	u32 head;
	enum intel_engine_hangcheck_action action;
	unsigned long action_timestamp;
	int deadlock;
	struct intel_instdone instdone;
	bool wedged:1;
	bool stalled:1;
};

static bool instdone_unchanged(u32 current_instdone, u32 *old_instdone)
{
	u32 tmp = current_instdone | *old_instdone;
	bool unchanged;

	unchanged = tmp == *old_instdone;
	*old_instdone |= tmp;

	return unchanged;
}

static bool subunits_stuck(struct intel_engine_cs *engine)
{
	struct drm_i915_private *dev_priv = engine->i915;
	struct intel_instdone instdone;
	struct intel_instdone *accu_instdone = &engine->hangcheck.instdone;
	bool stuck;
	int slice;
	int subslice;

	if (engine->id != RCS0)
		return true;

	intel_engine_get_instdone(engine, &instdone);

	/* There might be unstable subunit states even when
	 * actual head is not moving. Filter out the unstable ones by
	 * accumulating the undone -> done transitions and only
	 * consider those as progress.
	 */
	stuck = instdone_unchanged(instdone.instdone,
				   &accu_instdone->instdone);
	stuck &= instdone_unchanged(instdone.slice_common,
				    &accu_instdone->slice_common);

	for_each_instdone_slice_subslice(dev_priv, slice, subslice) {
		stuck &= instdone_unchanged(instdone.sampler[slice][subslice],
					    &accu_instdone->sampler[slice][subslice]);
		stuck &= instdone_unchanged(instdone.row[slice][subslice],
					    &accu_instdone->row[slice][subslice]);
	}

	return stuck;
}

static enum intel_engine_hangcheck_action
head_stuck(struct intel_engine_cs *engine, u64 acthd)
{
	if (acthd != engine->hangcheck.acthd) {

		/* Clear subunit states on head movement */
		memset(&engine->hangcheck.instdone, 0,
		       sizeof(engine->hangcheck.instdone));

		return ENGINE_ACTIVE_HEAD;
	}

	if (!subunits_stuck(engine))
		return ENGINE_ACTIVE_SUBUNITS;

	return ENGINE_DEAD;
}

static enum intel_engine_hangcheck_action
engine_stuck(struct intel_engine_cs *engine, u64 acthd)
{
	struct drm_i915_private *dev_priv = engine->i915;
	enum intel_engine_hangcheck_action ha;
	u32 tmp;

	ha = head_stuck(engine, acthd);
	if (ha != ENGINE_DEAD)
		return ha;

	if (IS_GEN(dev_priv, 2))
		return ENGINE_DEAD;

	/* Is the chip hanging on a WAIT_FOR_EVENT?
	 * If so we can simply poke the RB_WAIT bit
	 * and break the hang. This should work on
	 * all but the second generation chipsets.
	 */
	tmp = ENGINE_READ(engine, RING_CTL);
	if (tmp & RING_WAIT) {
		i915_handle_error(dev_priv, engine->mask, 0,
				  "stuck wait on %s", engine->name);
		ENGINE_WRITE(engine, RING_CTL, tmp);
		return ENGINE_WAIT_KICK;
	}

	return ENGINE_DEAD;
}

static void hangcheck_load_sample(struct intel_engine_cs *engine,
				  struct hangcheck *hc)
{
	hc->acthd = intel_engine_get_active_head(engine);
	hc->ring = ENGINE_READ(engine, RING_START);
	hc->head = ENGINE_READ(engine, RING_HEAD);
}

static void hangcheck_store_sample(struct intel_engine_cs *engine,
				   const struct hangcheck *hc)
{
	engine->hangcheck.acthd = hc->acthd;
	engine->hangcheck.last_ring = hc->ring;
	engine->hangcheck.last_head = hc->head;
}

static enum intel_engine_hangcheck_action
hangcheck_get_action(struct intel_engine_cs *engine,
		     const struct hangcheck *hc)
{
	if (intel_engine_is_idle(engine))
		return ENGINE_IDLE;

	if (engine->hangcheck.last_ring != hc->ring)
		return ENGINE_ACTIVE_SEQNO;

	if (engine->hangcheck.last_head != hc->head)
		return ENGINE_ACTIVE_SEQNO;

	return engine_stuck(engine, hc->acthd);
}

static void hangcheck_accumulate_sample(struct intel_engine_cs *engine,
					struct hangcheck *hc)
{
	unsigned long timeout = I915_ENGINE_DEAD_TIMEOUT;

	hc->action = hangcheck_get_action(engine, hc);

	/* We always increment the progress
	 * if the engine is busy and still processing
	 * the same request, so that no single request
	 * can run indefinitely (such as a chain of
	 * batches). The only time we do not increment
	 * the hangcheck score on this ring, if this
	 * engine is in a legitimate wait for another
	 * engine. In that case the waiting engine is a
	 * victim and we want to be sure we catch the
	 * right culprit. Then every time we do kick
	 * the ring, make it as a progress as the seqno
	 * advancement might ensure and if not, it
	 * will catch the hanging engine.
	 */

	switch (hc->action) {
	case ENGINE_IDLE:
	case ENGINE_ACTIVE_SEQNO:
		/* Clear head and subunit states on seqno movement */
		hc->acthd = 0;

		memset(&engine->hangcheck.instdone, 0,
		       sizeof(engine->hangcheck.instdone));

		/* Intentional fall through */
	case ENGINE_WAIT_KICK:
	case ENGINE_WAIT:
		engine->hangcheck.action_timestamp = jiffies;
		break;

	case ENGINE_ACTIVE_HEAD:
	case ENGINE_ACTIVE_SUBUNITS:
		/*
		 * Seqno stuck with still active engine gets leeway,
		 * in hopes that it is just a long shader.
		 */
		timeout = I915_SEQNO_DEAD_TIMEOUT;
		break;

	case ENGINE_DEAD:
		break;

	default:
		MISSING_CASE(hc->action);
	}

	hc->stalled = time_after(jiffies,
				 engine->hangcheck.action_timestamp + timeout);
	hc->wedged = time_after(jiffies,
				 engine->hangcheck.action_timestamp +
				 I915_ENGINE_WEDGED_TIMEOUT);
}

static void hangcheck_declare_hang(struct drm_i915_private *i915,
				   intel_engine_mask_t hung,
				   intel_engine_mask_t stuck)
{
	struct intel_engine_cs *engine;
	intel_engine_mask_t tmp;
	char msg[80];
	int len;

	/* If some rings hung but others were still busy, only
	 * blame the hanging rings in the synopsis.
	 */
	if (stuck != hung)
		hung &= ~stuck;
	len = scnprintf(msg, sizeof(msg),
			"%s on ", stuck == hung ? "no progress" : "hang");
	for_each_engine_masked(engine, i915, hung, tmp)
		len += scnprintf(msg + len, sizeof(msg) - len,
				 "%s, ", engine->name);
	msg[len-2] = '\0';

	return i915_handle_error(i915, hung, I915_ERROR_CAPTURE, "%s", msg);
}

/*
 * This is called when the chip hasn't reported back with completed
 * batchbuffers in a long time. We keep track per ring seqno progress and
 * if there are no progress, hangcheck score for that ring is increased.
 * Further, acthd is inspected to see if the ring is stuck. On stuck case
 * we kick the ring. If we see no progress on three subsequent calls
 * we assume chip is wedged and try to fix it by resetting the chip.
 */
static void i915_hangcheck_elapsed(struct work_struct *work)
{
	struct drm_i915_private *dev_priv =
		container_of(work, typeof(*dev_priv),
			     gpu_error.hangcheck_work.work);
	intel_engine_mask_t hung = 0, stuck = 0, wedged = 0;
	struct intel_engine_cs *engine;
	enum intel_engine_id id;
	intel_wakeref_t wakeref;

	if (!i915_modparams.enable_hangcheck)
		return;

	if (!READ_ONCE(dev_priv->gt.awake))
		return;

	if (i915_terminally_wedged(dev_priv))
		return;

	wakeref = intel_runtime_pm_get_if_in_use(&dev_priv->runtime_pm);
	if (!wakeref)
		return;

	/* As enabling the GPU requires fairly extensive mmio access,
	 * periodically arm the mmio checker to see if we are triggering
	 * any invalid access.
	 */
	intel_uncore_arm_unclaimed_mmio_detection(&dev_priv->uncore);

	for_each_engine(engine, dev_priv, id) {
		struct hangcheck hc;

		intel_engine_signal_breadcrumbs(engine);

		hangcheck_load_sample(engine, &hc);
		hangcheck_accumulate_sample(engine, &hc);
		hangcheck_store_sample(engine, &hc);

		if (hc.stalled) {
			hung |= engine->mask;
			if (hc.action != ENGINE_DEAD)
				stuck |= engine->mask;
		}

		if (hc.wedged)
			wedged |= engine->mask;
	}

	if (GEM_SHOW_DEBUG() && (hung | stuck)) {
		struct drm_printer p = drm_debug_printer("hangcheck");

		for_each_engine(engine, dev_priv, id) {
			if (intel_engine_is_idle(engine))
				continue;

			intel_engine_dump(engine, &p, "%s\n", engine->name);
		}
	}

	if (wedged) {
		dev_err(dev_priv->drm.dev,
			"GPU recovery timed out,"
			" cancelling all in-flight rendering.\n");
		GEM_TRACE_DUMP();
		i915_gem_set_wedged(dev_priv);
	}

	if (hung)
		hangcheck_declare_hang(dev_priv, hung, stuck);

	intel_runtime_pm_put(&dev_priv->runtime_pm, wakeref);

	/* Reset timer in case GPU hangs without another request being added */
	i915_queue_hangcheck(dev_priv);
}

void intel_engine_init_hangcheck(struct intel_engine_cs *engine)
{
	memset(&engine->hangcheck, 0, sizeof(engine->hangcheck));
	engine->hangcheck.action_timestamp = jiffies;
}

void intel_hangcheck_init(struct drm_i915_private *i915)
{
	INIT_DELAYED_WORK(&i915->gpu_error.hangcheck_work,
			  i915_hangcheck_elapsed);
}

#if IS_ENABLED(CONFIG_DRM_I915_SELFTEST)
#include "selftest_hangcheck.c"
#endif