summaryrefslogtreecommitdiff
path: root/drivers/gpu/drm/ttm/ttm_memory.c
blob: 072c281a6bb5bd07ed491ad67b6265684eb0c7fb (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
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
/**************************************************************************
 *
 * Copyright (c) 2006-2009 VMware, Inc., Palo Alto, CA., USA
 * All Rights Reserved.
 *
 * 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, sub license, 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 NON-INFRINGEMENT. IN NO EVENT SHALL
 * THE COPYRIGHT HOLDERS, AUTHORS AND/OR ITS SUPPLIERS 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 "ttm/ttm_memory.h"
#include "ttm/ttm_module.h"
#include <linux/spinlock.h>
#include <linux/sched.h>
#include <linux/wait.h>
#include <linux/mm.h>
#include <linux/module.h>

#define TTM_MEMORY_ALLOC_RETRIES 4

struct ttm_mem_zone {
	struct kobject kobj;
	struct ttm_mem_global *glob;
	const char *name;
	uint64_t zone_mem;
	uint64_t emer_mem;
	uint64_t max_mem;
	uint64_t swap_limit;
	uint64_t used_mem;
};

static struct attribute ttm_mem_sys = {
	.name = "zone_memory",
	.mode = S_IRUGO
};
static struct attribute ttm_mem_emer = {
	.name = "emergency_memory",
	.mode = S_IRUGO | S_IWUSR
};
static struct attribute ttm_mem_max = {
	.name = "available_memory",
	.mode = S_IRUGO | S_IWUSR
};
static struct attribute ttm_mem_swap = {
	.name = "swap_limit",
	.mode = S_IRUGO | S_IWUSR
};
static struct attribute ttm_mem_used = {
	.name = "used_memory",
	.mode = S_IRUGO
};

static void ttm_mem_zone_kobj_release(struct kobject *kobj)
{
	struct ttm_mem_zone *zone =
		container_of(kobj, struct ttm_mem_zone, kobj);

	printk(KERN_INFO TTM_PFX
	       "Zone %7s: Used memory at exit: %llu kiB.\n",
	       zone->name, (unsigned long long) zone->used_mem >> 10);
	kfree(zone);
}

static ssize_t ttm_mem_zone_show(struct kobject *kobj,
				 struct attribute *attr,
				 char *buffer)
{
	struct ttm_mem_zone *zone =
		container_of(kobj, struct ttm_mem_zone, kobj);
	uint64_t val = 0;

	spin_lock(&zone->glob->lock);
	if (attr == &ttm_mem_sys)
		val = zone->zone_mem;
	else if (attr == &ttm_mem_emer)
		val = zone->emer_mem;
	else if (attr == &ttm_mem_max)
		val = zone->max_mem;
	else if (attr == &ttm_mem_swap)
		val = zone->swap_limit;
	else if (attr == &ttm_mem_used)
		val = zone->used_mem;
	spin_unlock(&zone->glob->lock);

	return snprintf(buffer, PAGE_SIZE, "%llu\n",
			(unsigned long long) val >> 10);
}

static void ttm_check_swapping(struct ttm_mem_global *glob);

static ssize_t ttm_mem_zone_store(struct kobject *kobj,
				  struct attribute *attr,
				  const char *buffer,
				  size_t size)
{
	struct ttm_mem_zone *zone =
		container_of(kobj, struct ttm_mem_zone, kobj);
	int chars;
	unsigned long val;
	uint64_t val64;

	chars = sscanf(buffer, "%lu", &val);
	if (chars == 0)
		return size;

	val64 = val;
	val64 <<= 10;

	spin_lock(&zone->glob->lock);
	if (val64 > zone->zone_mem)
		val64 = zone->zone_mem;
	if (attr == &ttm_mem_emer) {
		zone->emer_mem = val64;
		if (zone->max_mem > val64)
			zone->max_mem = val64;
	} else if (attr == &ttm_mem_max) {
		zone->max_mem = val64;
		if (zone->emer_mem < val64)
			zone->emer_mem = val64;
	} else if (attr == &ttm_mem_swap)
		zone->swap_limit = val64;
	spin_unlock(&zone->glob->lock);

	ttm_check_swapping(zone->glob);

	return size;
}

static struct attribute *ttm_mem_zone_attrs[] = {
	&ttm_mem_sys,
	&ttm_mem_emer,
	&ttm_mem_max,
	&ttm_mem_swap,
	&ttm_mem_used,
	NULL
};

static struct sysfs_ops ttm_mem_zone_ops = {
	.show = &ttm_mem_zone_show,
	.store = &ttm_mem_zone_store
};

static struct kobj_type ttm_mem_zone_kobj_type = {
	.release = &ttm_mem_zone_kobj_release,
	.sysfs_ops = &ttm_mem_zone_ops,
	.default_attrs = ttm_mem_zone_attrs,
};

static void ttm_mem_global_kobj_release(struct kobject *kobj)
{
	struct ttm_mem_global *glob =
		container_of(kobj, struct ttm_mem_global, kobj);

	kfree(glob);
}

static struct kobj_type ttm_mem_glob_kobj_type = {
	.release = &ttm_mem_global_kobj_release,
};

static bool ttm_zones_above_swap_target(struct ttm_mem_global *glob,
					bool from_wq, uint64_t extra)
{
	unsigned int i;
	struct ttm_mem_zone *zone;
	uint64_t target;

	for (i = 0; i < glob->num_zones; ++i) {
		zone = glob->zones[i];

		if (from_wq)
			target = zone->swap_limit;
		else if (capable(CAP_SYS_ADMIN))
			target = zone->emer_mem;
		else
			target = zone->max_mem;

		target = (extra > target) ? 0ULL : target;

		if (zone->used_mem > target)
			return true;
	}
	return false;
}

/**
 * At this point we only support a single shrink callback.
 * Extend this if needed, perhaps using a linked list of callbacks.
 * Note that this function is reentrant:
 * many threads may try to swap out at any given time.
 */

static void ttm_shrink(struct ttm_mem_global *glob, bool from_wq,
		       uint64_t extra)
{
	int ret;
	struct ttm_mem_shrink *shrink;

	spin_lock(&glob->lock);
	if (glob->shrink == NULL)
		goto out;

	while (ttm_zones_above_swap_target(glob, from_wq, extra)) {
		shrink = glob->shrink;
		spin_unlock(&glob->lock);
		ret = shrink->do_shrink(shrink);
		spin_lock(&glob->lock);
		if (unlikely(ret != 0))
			goto out;
	}
out:
	spin_unlock(&glob->lock);
}



static void ttm_shrink_work(struct work_struct *work)
{
	struct ttm_mem_global *glob =
	    container_of(work, struct ttm_mem_global, work);

	ttm_shrink(glob, true, 0ULL);
}

static int ttm_mem_init_kernel_zone(struct ttm_mem_global *glob,
				    const struct sysinfo *si)
{
	struct ttm_mem_zone *zone = kzalloc(sizeof(*zone), GFP_KERNEL);
	uint64_t mem;
	int ret;

	if (unlikely(!zone))
		return -ENOMEM;

	mem = si->totalram - si->totalhigh;
	mem *= si->mem_unit;

	zone->name = "kernel";
	zone->zone_mem = mem;
	zone->max_mem = mem >> 1;
	zone->emer_mem = (mem >> 1) + (mem >> 2);
	zone->swap_limit = zone->max_mem - (mem >> 3);
	zone->used_mem = 0;
	zone->glob = glob;
	glob->zone_kernel = zone;
	kobject_init(&zone->kobj, &ttm_mem_zone_kobj_type);
	ret = kobject_add(&zone->kobj, &glob->kobj, zone->name);
	if (unlikely(ret != 0)) {
		kobject_put(&zone->kobj);
		return ret;
	}
	glob->zones[glob->num_zones++] = zone;
	return 0;
}

#ifdef CONFIG_HIGHMEM
static int ttm_mem_init_highmem_zone(struct ttm_mem_global *glob,
				     const struct sysinfo *si)
{
	struct ttm_mem_zone *zone = kzalloc(sizeof(*zone), GFP_KERNEL);
	uint64_t mem;
	int ret;

	if (unlikely(!zone))
		return -ENOMEM;

	if (si->totalhigh == 0)
		return 0;

	mem = si->totalram;
	mem *= si->mem_unit;

	zone->name = "highmem";
	zone->zone_mem = mem;
	zone->max_mem = mem >> 1;
	zone->emer_mem = (mem >> 1) + (mem >> 2);
	zone->swap_limit = zone->max_mem - (mem >> 3);
	zone->used_mem = 0;
	zone->glob = glob;
	glob->zone_highmem = zone;
	kobject_init(&zone->kobj, &ttm_mem_zone_kobj_type);
	ret = kobject_add(&zone->kobj, &glob->kobj, zone->name);
	if (unlikely(ret != 0)) {
		kobject_put(&zone->kobj);
		return ret;
	}
	glob->zones[glob->num_zones++] = zone;
	return 0;
}
#else
static int ttm_mem_init_dma32_zone(struct ttm_mem_global *glob,
				   const struct sysinfo *si)
{
	struct ttm_mem_zone *zone = kzalloc(sizeof(*zone), GFP_KERNEL);
	uint64_t mem;
	int ret;

	if (unlikely(!zone))
		return -ENOMEM;

	mem = si->totalram;
	mem *= si->mem_unit;

	/**
	 * No special dma32 zone needed.
	 */

	if (mem <= ((uint64_t) 1ULL << 32))
		return 0;

	/*
	 * Limit max dma32 memory to 4GB for now
	 * until we can figure out how big this
	 * zone really is.
	 */

	mem = ((uint64_t) 1ULL << 32);
	zone->name = "dma32";
	zone->zone_mem = mem;
	zone->max_mem = mem >> 1;
	zone->emer_mem = (mem >> 1) + (mem >> 2);
	zone->swap_limit = zone->max_mem - (mem >> 3);
	zone->used_mem = 0;
	zone->glob = glob;
	glob->zone_dma32 = zone;
	kobject_init(&zone->kobj, &ttm_mem_zone_kobj_type);
	ret = kobject_add(&zone->kobj, &glob->kobj, zone->name);
	if (unlikely(ret != 0)) {
		kobject_put(&zone->kobj);
		return ret;
	}
	glob->zones[glob->num_zones++] = zone;
	return 0;
}
#endif

int ttm_mem_global_init(struct ttm_mem_global *glob)
{
	struct sysinfo si;
	int ret;
	int i;
	struct ttm_mem_zone *zone;

	spin_lock_init(&glob->lock);
	glob->swap_queue = create_singlethread_workqueue("ttm_swap");
	INIT_WORK(&glob->work, ttm_shrink_work);
	init_waitqueue_head(&glob->queue);
	kobject_init(&glob->kobj, &ttm_mem_glob_kobj_type);
	ret = kobject_add(&glob->kobj,
			  ttm_get_kobj(),
			  "memory_accounting");
	if (unlikely(ret != 0)) {
		kobject_put(&glob->kobj);
		return ret;
	}

	si_meminfo(&si);

	ret = ttm_mem_init_kernel_zone(glob, &si);
	if (unlikely(ret != 0))
		goto out_no_zone;
#ifdef CONFIG_HIGHMEM
	ret = ttm_mem_init_highmem_zone(glob, &si);
	if (unlikely(ret != 0))
		goto out_no_zone;
#else
	ret = ttm_mem_init_dma32_zone(glob, &si);
	if (unlikely(ret != 0))
		goto out_no_zone;
#endif
	for (i = 0; i < glob->num_zones; ++i) {
		zone = glob->zones[i];
		printk(KERN_INFO TTM_PFX
		       "Zone %7s: Available graphics memory: %llu kiB.\n",
		       zone->name, (unsigned long long) zone->max_mem >> 10);
	}
	return 0;
out_no_zone:
	ttm_mem_global_release(glob);
	return ret;
}
EXPORT_SYMBOL(ttm_mem_global_init);

void ttm_mem_global_release(struct ttm_mem_global *glob)
{
	unsigned int i;
	struct ttm_mem_zone *zone;

	flush_workqueue(glob->swap_queue);
	destroy_workqueue(glob->swap_queue);
	glob->swap_queue = NULL;
	for (i = 0; i < glob->num_zones; ++i) {
		zone = glob->zones[i];
		kobject_del(&zone->kobj);
		kobject_put(&zone->kobj);
	}
	kobject_del(&glob->kobj);
	kobject_put(&glob->kobj);
}
EXPORT_SYMBOL(ttm_mem_global_release);

static void ttm_check_swapping(struct ttm_mem_global *glob)
{
	bool needs_swapping = false;
	unsigned int i;
	struct ttm_mem_zone *zone;

	spin_lock(&glob->lock);
	for (i = 0; i < glob->num_zones; ++i) {
		zone = glob->zones[i];
		if (zone->used_mem > zone->swap_limit) {
			needs_swapping = true;
			break;
		}
	}

	spin_unlock(&glob->lock);

	if (unlikely(needs_swapping))
		(void)queue_work(glob->swap_queue, &glob->work);

}

static void ttm_mem_global_free_zone(struct ttm_mem_global *glob,
				     struct ttm_mem_zone *single_zone,
				     uint64_t amount)
{
	unsigned int i;
	struct ttm_mem_zone *zone;

	spin_lock(&glob->lock);
	for (i = 0; i < glob->num_zones; ++i) {
		zone = glob->zones[i];
		if (single_zone && zone != single_zone)
			continue;
		zone->used_mem -= amount;
	}
	spin_unlock(&glob->lock);
}

void ttm_mem_global_free(struct ttm_mem_global *glob,
			 uint64_t amount)
{
	return ttm_mem_global_free_zone(glob, NULL, amount);
}

static int ttm_mem_global_reserve(struct ttm_mem_global *glob,
				  struct ttm_mem_zone *single_zone,
				  uint64_t amount, bool reserve)
{
	uint64_t limit;
	int ret = -ENOMEM;
	unsigned int i;
	struct ttm_mem_zone *zone;

	spin_lock(&glob->lock);
	for (i = 0; i < glob->num_zones; ++i) {
		zone = glob->zones[i];
		if (single_zone && zone != single_zone)
			continue;

		limit = (capable(CAP_SYS_ADMIN)) ?
			zone->emer_mem : zone->max_mem;

		if (zone->used_mem > limit)
			goto out_unlock;
	}

	if (reserve) {
		for (i = 0; i < glob->num_zones; ++i) {
			zone = glob->zones[i];
			if (single_zone && zone != single_zone)
				continue;
			zone->used_mem += amount;
		}
	}

	ret = 0;
out_unlock:
	spin_unlock(&glob->lock);
	ttm_check_swapping(glob);

	return ret;
}


static int ttm_mem_global_alloc_zone(struct ttm_mem_global *glob,
				     struct ttm_mem_zone *single_zone,
				     uint64_t memory,
				     bool no_wait, bool interruptible)
{
	int count = TTM_MEMORY_ALLOC_RETRIES;

	while (unlikely(ttm_mem_global_reserve(glob,
					       single_zone,
					       memory, true)
			!= 0)) {
		if (no_wait)
			return -ENOMEM;
		if (unlikely(count-- == 0))
			return -ENOMEM;
		ttm_shrink(glob, false, memory + (memory >> 2) + 16);
	}

	return 0;
}

int ttm_mem_global_alloc(struct ttm_mem_global *glob, uint64_t memory,
			 bool no_wait, bool interruptible)
{
	/**
	 * Normal allocations of kernel memory are registered in
	 * all zones.
	 */

	return ttm_mem_global_alloc_zone(glob, NULL, memory, no_wait,
					 interruptible);
}

int ttm_mem_global_alloc_page(struct ttm_mem_global *glob,
			      struct page *page,
			      bool no_wait, bool interruptible)
{

	struct ttm_mem_zone *zone = NULL;

	/**
	 * Page allocations may be registed in a single zone
	 * only if highmem or !dma32.
	 */

#ifdef CONFIG_HIGHMEM
	if (PageHighMem(page) && glob->zone_highmem != NULL)
		zone = glob->zone_highmem;
#else
	if (glob->zone_dma32 && page_to_pfn(page) > 0x00100000UL)
		zone = glob->zone_kernel;
#endif
	return ttm_mem_global_alloc_zone(glob, zone, PAGE_SIZE, no_wait,
					 interruptible);
}

void ttm_mem_global_free_page(struct ttm_mem_global *glob, struct page *page)
{
	struct ttm_mem_zone *zone = NULL;

#ifdef CONFIG_HIGHMEM
	if (PageHighMem(page) && glob->zone_highmem != NULL)
		zone = glob->zone_highmem;
#else
	if (glob->zone_dma32 && page_to_pfn(page) > 0x00100000UL)
		zone = glob->zone_kernel;
#endif
	ttm_mem_global_free_zone(glob, zone, PAGE_SIZE);
}


size_t ttm_round_pot(size_t size)
{
	if ((size & (size - 1)) == 0)
		return size;
	else if (size > PAGE_SIZE)
		return PAGE_ALIGN(size);
	else {
		size_t tmp_size = 4;

		while (tmp_size < size)
			tmp_size <<= 1;

		return tmp_size;
	}
	return 0;
}