iova.c 26.8 KB
Newer Older
1
/*
2
 * Copyright © 2006-2009, Intel Corporation.
3
 *
4
5
6
7
8
9
10
11
12
13
14
15
 * This program is free software; you can redistribute it and/or modify it
 * under the terms and conditions of the GNU General Public License,
 * version 2, as published by the Free Software Foundation.
 *
 * This program is distributed in the hope 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, write to the Free Software Foundation, Inc., 59 Temple
 * Place - Suite 330, Boston, MA 02111-1307 USA.
16
 *
17
 * Author: Anil S Keshavamurthy <anil.s.keshavamurthy@intel.com>
18
19
 */

Kay, Allen M's avatar
Kay, Allen M committed
20
#include <linux/iova.h>
21
#include <linux/module.h>
22
#include <linux/slab.h>
23
24
#include <linux/smp.h>
#include <linux/bitops.h>
25
#include <linux/cpu.h>
26

27
28
29
/* The anchor node sits above the top of the usable address space */
#define IOVA_ANCHOR	~0UL

30
31
32
33
34
35
36
37
static bool iova_rcache_insert(struct iova_domain *iovad,
			       unsigned long pfn,
			       unsigned long size);
static unsigned long iova_rcache_get(struct iova_domain *iovad,
				     unsigned long size,
				     unsigned long limit_pfn);
static void init_iova_rcaches(struct iova_domain *iovad);
static void free_iova_rcaches(struct iova_domain *iovad);
38
static void fq_destroy_all_entries(struct iova_domain *iovad);
39
static void fq_flush_timeout(struct timer_list *t);
40

41
void
42
init_iova_domain(struct iova_domain *iovad, unsigned long granule,
43
	unsigned long start_pfn)
44
{
45
46
47
48
49
50
51
	/*
	 * IOVA granularity will normally be equal to the smallest
	 * supported IOMMU page size; both *must* be capable of
	 * representing individual CPU pages exactly.
	 */
	BUG_ON((granule > PAGE_SIZE) || !is_power_of_2(granule));

52
53
	spin_lock_init(&iovad->iova_rbtree_lock);
	iovad->rbroot = RB_ROOT;
54
55
	iovad->cached_node = &iovad->anchor.node;
	iovad->cached32_node = &iovad->anchor.node;
56
	iovad->granule = granule;
57
	iovad->start_pfn = start_pfn;
58
	iovad->dma_32bit_pfn = 1UL << (32 - iova_shift(iovad));
59
	iovad->free_32bit_pfns = iovad->dma_32bit_pfn - start_pfn;
60
61
	iovad->flush_cb = NULL;
	iovad->fq = NULL;
62
63
64
	iovad->anchor.pfn_lo = iovad->anchor.pfn_hi = IOVA_ANCHOR;
	rb_link_node(&iovad->anchor.node, NULL, &iovad->rbroot.rb_node);
	rb_insert_color(&iovad->anchor.node, &iovad->rbroot);
65
	init_iova_rcaches(iovad);
66
}
Sakari Ailus's avatar
Sakari Ailus committed
67
EXPORT_SYMBOL_GPL(init_iova_domain);
68

69
70
71
72
73
static void free_iova_flush_queue(struct iova_domain *iovad)
{
	if (!iovad->fq)
		return;

Joerg Roedel's avatar
Joerg Roedel committed
74
75
76
	if (timer_pending(&iovad->fq_timer))
		del_timer(&iovad->fq_timer);

77
	fq_destroy_all_entries(iovad);
Joerg Roedel's avatar
Joerg Roedel committed
78

79
80
81
82
83
84
85
86
87
88
89
90
	free_percpu(iovad->fq);

	iovad->fq         = NULL;
	iovad->flush_cb   = NULL;
	iovad->entry_dtor = NULL;
}

int init_iova_flush_queue(struct iova_domain *iovad,
			  iova_flush_cb flush_cb, iova_entry_dtor entry_dtor)
{
	int cpu;

91
92
93
	atomic64_set(&iovad->fq_flush_start_cnt,  0);
	atomic64_set(&iovad->fq_flush_finish_cnt, 0);

94
95
96
97
98
99
100
101
102
103
104
105
106
	iovad->fq = alloc_percpu(struct iova_fq);
	if (!iovad->fq)
		return -ENOMEM;

	iovad->flush_cb   = flush_cb;
	iovad->entry_dtor = entry_dtor;

	for_each_possible_cpu(cpu) {
		struct iova_fq *fq;

		fq = per_cpu_ptr(iovad->fq, cpu);
		fq->head = 0;
		fq->tail = 0;
107
108

		spin_lock_init(&fq->lock);
109
110
	}

111
	timer_setup(&iovad->fq_timer, fq_flush_timeout, 0);
Joerg Roedel's avatar
Joerg Roedel committed
112
113
	atomic_set(&iovad->fq_timer_on, 0);

114
115
116
117
	return 0;
}
EXPORT_SYMBOL_GPL(init_iova_flush_queue);

118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
static void insert_update_32bit_pfn(struct iova_domain *iovad, struct iova *new)
{
	unsigned long pfn_lo = new->pfn_lo;
	unsigned long pfn_hi = min(new->pfn_hi, iovad->dma_32bit_pfn);

	if (pfn_lo >= iovad->dma_32bit_pfn)
		iovad->free_32bit_pfns -= pfn_hi - pfn_lo + 1;
}

static void delete_update_32bit_pfn(struct iova_domain *iovad, struct iova *free)
{
	unsigned long pfn_lo = free->pfn_lo;
	unsigned long pfn_hi = min(free->pfn_hi, iovad->dma_32bit_pfn);

	if (pfn_lo >= iovad->dma_32bit_pfn)
		iovad->free_32bit_pfns += pfn_hi - pfn_lo + 1;
}

136
static struct rb_node *
137
__get_cached_rbnode(struct iova_domain *iovad, unsigned long limit_pfn)
138
{
139
140
	if (limit_pfn <= iovad->dma_32bit_pfn)
		return iovad->cached32_node;
141

142
	return iovad->cached_node;
143
144
145
}

static void
146
__cached_rbnode_insert_update(struct iova_domain *iovad, struct iova *new)
147
{
148
149
150
151
	if (new->pfn_hi < iovad->dma_32bit_pfn)
		iovad->cached32_node = &new->node;
	else
		iovad->cached_node = &new->node;
152
153

	insert_update_32bit_pfn(iovad, new);
154
155
156
157
158
159
160
}

static void
__cached_rbnode_delete_update(struct iova_domain *iovad, struct iova *free)
{
	struct iova *cached_iova;

161
162
	cached_iova = rb_entry(iovad->cached32_node, struct iova, node);
	if (free->pfn_hi < iovad->dma_32bit_pfn &&
163
	    free->pfn_lo >= cached_iova->pfn_lo)
164
165
166
		iovad->cached32_node = rb_next(&free->node);

	cached_iova = rb_entry(iovad->cached_node, struct iova, node);
167
	if (free->pfn_lo >= cached_iova->pfn_lo)
168
		iovad->cached_node = rb_next(&free->node);
169
170

	delete_update_32bit_pfn(iovad, free);
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
/* Insert the iova into domain rbtree by holding writer lock */
static void
iova_insert_rbtree(struct rb_root *root, struct iova *iova,
		   struct rb_node *start)
{
	struct rb_node **new, *parent = NULL;

	new = (start) ? &start : &(root->rb_node);
	/* Figure out where to put new node */
	while (*new) {
		struct iova *this = rb_entry(*new, struct iova, node);

		parent = *new;

		if (iova->pfn_lo < this->pfn_lo)
			new = &((*new)->rb_left);
		else if (iova->pfn_lo > this->pfn_lo)
			new = &((*new)->rb_right);
		else {
			WARN_ON(1); /* this should not happen */
			return;
		}
	}
	/* Add new node and rebalance tree. */
	rb_link_node(&iova->node, parent, new);
	rb_insert_color(&iova->node, root);
}

mark gross's avatar
mark gross committed
201
202
203
static int __alloc_and_insert_iova_range(struct iova_domain *iovad,
		unsigned long size, unsigned long limit_pfn,
			struct iova *new, bool size_aligned)
204
{
205
206
	struct rb_node *curr, *prev;
	struct iova *curr_iova;
207
	unsigned long flags;
208
	unsigned long new_pfn;
209
210
211
212
	unsigned long align_mask = ~0UL;

	if (size_aligned)
		align_mask <<= fls_long(size - 1);
213
214

	spin_lock_irqsave(&iovad->iova_rbtree_lock, flags);
215
216
217
218
219
220

	/* Fail fast if a 32-bit allocation cannot possibly succeed */
	if (limit_pfn <= iovad->dma_32bit_pfn && iovad->free_32bit_pfns < size)
		goto no_mem;

	/* Walk the tree backwards */
221
222
223
224
225
	curr = __get_cached_rbnode(iovad, limit_pfn);
	curr_iova = rb_entry(curr, struct iova, node);
	do {
		limit_pfn = min(limit_pfn, curr_iova->pfn_lo);
		new_pfn = (limit_pfn - size) & align_mask;
mark gross's avatar
mark gross committed
226
		prev = curr;
227
		curr = rb_prev(curr);
228
229
		curr_iova = rb_entry(curr, struct iova, node);
	} while (curr && new_pfn <= curr_iova->pfn_hi);
230

231
232
	if (limit_pfn < size || new_pfn < iovad->start_pfn)
		goto no_mem;
233
234

	/* pfn_lo will point to size aligned address if size_aligned is set */
235
	new->pfn_lo = new_pfn;
236
	new->pfn_hi = new->pfn_lo + size - 1;
237

238
239
	/* If we have 'prev', it's a valid place to start the insertion. */
	iova_insert_rbtree(&iovad->rbroot, new, prev);
240
	__cached_rbnode_insert_update(iovad, new);
mark gross's avatar
mark gross committed
241

242
243
	spin_unlock_irqrestore(&iovad->iova_rbtree_lock, flags);
	return 0;
244
245
246
247

no_mem:
	spin_unlock_irqrestore(&iovad->iova_rbtree_lock, flags);
	return -ENOMEM;
248
249
}

250
251
252
253
254
255
static struct kmem_cache *iova_cache;
static unsigned int iova_cache_users;
static DEFINE_MUTEX(iova_cache_mutex);

struct iova *alloc_iova_mem(void)
{
256
	return kmem_cache_zalloc(iova_cache, GFP_ATOMIC);
257
258
259
260
261
}
EXPORT_SYMBOL(alloc_iova_mem);

void free_iova_mem(struct iova *iova)
{
262
263
	if (iova->pfn_lo != IOVA_ANCHOR)
		kmem_cache_free(iova_cache, iova);
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
}
EXPORT_SYMBOL(free_iova_mem);

int iova_cache_get(void)
{
	mutex_lock(&iova_cache_mutex);
	if (!iova_cache_users) {
		iova_cache = kmem_cache_create(
			"iommu_iova", sizeof(struct iova), 0,
			SLAB_HWCACHE_ALIGN, NULL);
		if (!iova_cache) {
			mutex_unlock(&iova_cache_mutex);
			printk(KERN_ERR "Couldn't create iova cache\n");
			return -ENOMEM;
		}
	}

	iova_cache_users++;
	mutex_unlock(&iova_cache_mutex);

	return 0;
}
Sakari Ailus's avatar
Sakari Ailus committed
286
EXPORT_SYMBOL_GPL(iova_cache_get);
287
288
289
290
291
292
293
294
295
296
297
298
299

void iova_cache_put(void)
{
	mutex_lock(&iova_cache_mutex);
	if (WARN_ON(!iova_cache_users)) {
		mutex_unlock(&iova_cache_mutex);
		return;
	}
	iova_cache_users--;
	if (!iova_cache_users)
		kmem_cache_destroy(iova_cache);
	mutex_unlock(&iova_cache_mutex);
}
Sakari Ailus's avatar
Sakari Ailus committed
300
EXPORT_SYMBOL_GPL(iova_cache_put);
301

302
303
/**
 * alloc_iova - allocates an iova
Masanari Iida's avatar
Masanari Iida committed
304
305
306
307
 * @iovad: - iova domain in question
 * @size: - size of page frames to allocate
 * @limit_pfn: - max limit address
 * @size_aligned: - set if size_aligned address range is required
308
309
 * This function allocates an iova in the range iovad->start_pfn to limit_pfn,
 * searching top-down from limit_pfn to iovad->start_pfn. If the size_aligned
310
311
 * flag is set then the allocated address iova->pfn_lo will be naturally
 * aligned on roundup_power_of_two(size).
312
313
314
 */
struct iova *
alloc_iova(struct iova_domain *iovad, unsigned long size,
315
316
	unsigned long limit_pfn,
	bool size_aligned)
317
318
319
320
321
322
323
324
{
	struct iova *new_iova;
	int ret;

	new_iova = alloc_iova_mem();
	if (!new_iova)
		return NULL;

325
	ret = __alloc_and_insert_iova_range(iovad, size, limit_pfn + 1,
mark gross's avatar
mark gross committed
326
			new_iova, size_aligned);
327
328
329
330
331
332
333
334

	if (ret) {
		free_iova_mem(new_iova);
		return NULL;
	}

	return new_iova;
}
Sakari Ailus's avatar
Sakari Ailus committed
335
EXPORT_SYMBOL_GPL(alloc_iova);
336

337
338
static struct iova *
private_find_iova(struct iova_domain *iovad, unsigned long pfn)
339
{
340
341
342
	struct rb_node *node = iovad->rbroot.rb_node;

	assert_spin_locked(&iovad->iova_rbtree_lock);
343
344

	while (node) {
Geliang Tang's avatar
Geliang Tang committed
345
		struct iova *iova = rb_entry(node, struct iova, node);
346
347
348

		if (pfn < iova->pfn_lo)
			node = node->rb_left;
349
		else if (pfn > iova->pfn_hi)
350
			node = node->rb_right;
351
352
		else
			return iova;	/* pfn falls within iova's range */
353
354
355
356
	}

	return NULL;
}
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

static void private_free_iova(struct iova_domain *iovad, struct iova *iova)
{
	assert_spin_locked(&iovad->iova_rbtree_lock);
	__cached_rbnode_delete_update(iovad, iova);
	rb_erase(&iova->node, &iovad->rbroot);
	free_iova_mem(iova);
}

/**
 * find_iova - finds an iova for a given pfn
 * @iovad: - iova domain in question.
 * @pfn: - page frame number
 * This function finds and returns an iova belonging to the
 * given doamin which matches the given pfn.
 */
struct iova *find_iova(struct iova_domain *iovad, unsigned long pfn)
{
	unsigned long flags;
	struct iova *iova;

	/* Take the lock so that no other thread is manipulating the rbtree */
	spin_lock_irqsave(&iovad->iova_rbtree_lock, flags);
	iova = private_find_iova(iovad, pfn);
	spin_unlock_irqrestore(&iovad->iova_rbtree_lock, flags);
	return iova;
}
Sakari Ailus's avatar
Sakari Ailus committed
384
EXPORT_SYMBOL_GPL(find_iova);
385
386
387
388
389
390
391
392
393
394
395
396
397

/**
 * __free_iova - frees the given iova
 * @iovad: iova domain in question.
 * @iova: iova in question.
 * Frees the given iova belonging to the giving domain
 */
void
__free_iova(struct iova_domain *iovad, struct iova *iova)
{
	unsigned long flags;

	spin_lock_irqsave(&iovad->iova_rbtree_lock, flags);
398
	private_free_iova(iovad, iova);
399
400
	spin_unlock_irqrestore(&iovad->iova_rbtree_lock, flags);
}
Sakari Ailus's avatar
Sakari Ailus committed
401
EXPORT_SYMBOL_GPL(__free_iova);
402
403
404
405
406
407
408
409
410
411
412
413

/**
 * free_iova - finds and frees the iova for a given pfn
 * @iovad: - iova domain in question.
 * @pfn: - pfn that is allocated previously
 * This functions finds an iova for a given pfn and then
 * frees the iova from that domain.
 */
void
free_iova(struct iova_domain *iovad, unsigned long pfn)
{
	struct iova *iova = find_iova(iovad, pfn);
414

415
416
417
418
	if (iova)
		__free_iova(iovad, iova);

}
Sakari Ailus's avatar
Sakari Ailus committed
419
EXPORT_SYMBOL_GPL(free_iova);
420

421
422
423
424
425
/**
 * alloc_iova_fast - allocates an iova from rcache
 * @iovad: - iova domain in question
 * @size: - size of page frames to allocate
 * @limit_pfn: - max limit address
426
 * @flush_rcache: - set to flush rcache on regular allocation failure
427
 * This function tries to satisfy an iova allocation from the rcache,
428
429
 * and falls back to regular allocation on failure. If regular allocation
 * fails too and the flush_rcache flag is set then the rcache will be flushed.
430
431
432
*/
unsigned long
alloc_iova_fast(struct iova_domain *iovad, unsigned long size,
433
		unsigned long limit_pfn, bool flush_rcache)
434
435
436
437
{
	unsigned long iova_pfn;
	struct iova *new_iova;

438
	iova_pfn = iova_rcache_get(iovad, size, limit_pfn + 1);
439
440
441
442
443
444
445
446
	if (iova_pfn)
		return iova_pfn;

retry:
	new_iova = alloc_iova(iovad, size, limit_pfn, true);
	if (!new_iova) {
		unsigned int cpu;

447
		if (!flush_rcache)
448
449
450
			return 0;

		/* Try replenishing IOVAs by flushing rcache. */
451
		flush_rcache = false;
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
		for_each_online_cpu(cpu)
			free_cpu_cached_iovas(cpu, iovad);
		goto retry;
	}

	return new_iova->pfn_lo;
}
EXPORT_SYMBOL_GPL(alloc_iova_fast);

/**
 * free_iova_fast - free iova pfn range into rcache
 * @iovad: - iova domain in question.
 * @pfn: - pfn that is allocated previously
 * @size: - # of pages in range
 * This functions frees an iova range by trying to put it into the rcache,
 * falling back to regular iova deallocation via free_iova() if this fails.
 */
void
free_iova_fast(struct iova_domain *iovad, unsigned long pfn, unsigned long size)
{
	if (iova_rcache_insert(iovad, pfn, size))
		return;

	free_iova(iovad, pfn);
}
EXPORT_SYMBOL_GPL(free_iova_fast);

479
480
481
482
483
#define fq_ring_for_each(i, fq) \
	for ((i) = (fq)->head; (i) != (fq)->tail; (i) = ((i) + 1) % IOVA_FQ_SIZE)

static inline bool fq_full(struct iova_fq *fq)
{
484
	assert_spin_locked(&fq->lock);
485
486
487
488
489
490
491
	return (((fq->tail + 1) % IOVA_FQ_SIZE) == fq->head);
}

static inline unsigned fq_ring_add(struct iova_fq *fq)
{
	unsigned idx = fq->tail;

492
493
	assert_spin_locked(&fq->lock);

494
495
496
497
498
499
500
	fq->tail = (idx + 1) % IOVA_FQ_SIZE;

	return idx;
}

static void fq_ring_free(struct iova_domain *iovad, struct iova_fq *fq)
{
501
	u64 counter = atomic64_read(&iovad->fq_flush_finish_cnt);
502
503
	unsigned idx;

504
505
	assert_spin_locked(&fq->lock);

506
507
	fq_ring_for_each(idx, fq) {

508
509
510
		if (fq->entries[idx].counter >= counter)
			break;

511
512
513
514
515
516
		if (iovad->entry_dtor)
			iovad->entry_dtor(fq->entries[idx].data);

		free_iova_fast(iovad,
			       fq->entries[idx].iova_pfn,
			       fq->entries[idx].pages);
517
518

		fq->head = (fq->head + 1) % IOVA_FQ_SIZE;
519
	}
520
}
521

522
523
524
525
526
static void iova_domain_flush(struct iova_domain *iovad)
{
	atomic64_inc(&iovad->fq_flush_start_cnt);
	iovad->flush_cb(iovad);
	atomic64_inc(&iovad->fq_flush_finish_cnt);
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
}

static void fq_destroy_all_entries(struct iova_domain *iovad)
{
	int cpu;

	/*
	 * This code runs when the iova_domain is being detroyed, so don't
	 * bother to free iovas, just call the entry_dtor on all remaining
	 * entries.
	 */
	if (!iovad->entry_dtor)
		return;

	for_each_possible_cpu(cpu) {
		struct iova_fq *fq = per_cpu_ptr(iovad->fq, cpu);
		int idx;

		fq_ring_for_each(idx, fq)
			iovad->entry_dtor(fq->entries[idx].data);
	}
}

550
static void fq_flush_timeout(struct timer_list *t)
Joerg Roedel's avatar
Joerg Roedel committed
551
{
552
	struct iova_domain *iovad = from_timer(iovad, t, fq_timer);
Joerg Roedel's avatar
Joerg Roedel committed
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
	int cpu;

	atomic_set(&iovad->fq_timer_on, 0);
	iova_domain_flush(iovad);

	for_each_possible_cpu(cpu) {
		unsigned long flags;
		struct iova_fq *fq;

		fq = per_cpu_ptr(iovad->fq, cpu);
		spin_lock_irqsave(&fq->lock, flags);
		fq_ring_free(iovad, fq);
		spin_unlock_irqrestore(&fq->lock, flags);
	}
}

569
570
571
572
void queue_iova(struct iova_domain *iovad,
		unsigned long pfn, unsigned long pages,
		unsigned long data)
{
573
	struct iova_fq *fq = raw_cpu_ptr(iovad->fq);
574
	unsigned long flags;
575
576
	unsigned idx;

577
578
	spin_lock_irqsave(&fq->lock, flags);

579
580
581
582
583
584
585
	/*
	 * First remove all entries from the flush queue that have already been
	 * flushed out on another CPU. This makes the fq_full() check below less
	 * likely to be true.
	 */
	fq_ring_free(iovad, fq);

586
	if (fq_full(fq)) {
587
		iova_domain_flush(iovad);
588
589
590
591
592
593
594
595
		fq_ring_free(iovad, fq);
	}

	idx = fq_ring_add(fq);

	fq->entries[idx].iova_pfn = pfn;
	fq->entries[idx].pages    = pages;
	fq->entries[idx].data     = data;
596
	fq->entries[idx].counter  = atomic64_read(&iovad->fq_flush_start_cnt);
597

598
	spin_unlock_irqrestore(&fq->lock, flags);
Joerg Roedel's avatar
Joerg Roedel committed
599
600
601
602

	if (atomic_cmpxchg(&iovad->fq_timer_on, 0, 1) == 0)
		mod_timer(&iovad->fq_timer,
			  jiffies + msecs_to_jiffies(IOVA_FQ_TIMEOUT));
603
604
605
}
EXPORT_SYMBOL_GPL(queue_iova);

606
607
608
609
610
611
612
/**
 * put_iova_domain - destroys the iova doamin
 * @iovad: - iova domain in question.
 * All the iova's in that domain are destroyed.
 */
void put_iova_domain(struct iova_domain *iovad)
{
613
	struct iova *iova, *tmp;
614

615
	free_iova_flush_queue(iovad);
616
	free_iova_rcaches(iovad);
617
	rbtree_postorder_for_each_entry_safe(iova, tmp, &iovad->rbroot, node)
618
619
		free_iova_mem(iova);
}
Sakari Ailus's avatar
Sakari Ailus committed
620
EXPORT_SYMBOL_GPL(put_iova_domain);
621
622
623
624
625

static int
__is_range_overlap(struct rb_node *node,
	unsigned long pfn_lo, unsigned long pfn_hi)
{
Geliang Tang's avatar
Geliang Tang committed
626
	struct iova *iova = rb_entry(node, struct iova, node);
627
628
629
630
631
632

	if ((pfn_lo <= iova->pfn_hi) && (pfn_hi >= iova->pfn_lo))
		return 1;
	return 0;
}

633
634
635
636
637
638
639
640
641
642
643
644
645
646
static inline struct iova *
alloc_and_init_iova(unsigned long pfn_lo, unsigned long pfn_hi)
{
	struct iova *iova;

	iova = alloc_iova_mem();
	if (iova) {
		iova->pfn_lo = pfn_lo;
		iova->pfn_hi = pfn_hi;
	}

	return iova;
}

647
648
649
650
651
652
static struct iova *
__insert_new_range(struct iova_domain *iovad,
	unsigned long pfn_lo, unsigned long pfn_hi)
{
	struct iova *iova;

653
654
	iova = alloc_and_init_iova(pfn_lo, pfn_hi);
	if (iova)
655
		iova_insert_rbtree(&iovad->rbroot, iova, NULL);
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686

	return iova;
}

static void
__adjust_overlap_range(struct iova *iova,
	unsigned long *pfn_lo, unsigned long *pfn_hi)
{
	if (*pfn_lo < iova->pfn_lo)
		iova->pfn_lo = *pfn_lo;
	if (*pfn_hi > iova->pfn_hi)
		*pfn_lo = iova->pfn_hi + 1;
}

/**
 * reserve_iova - reserves an iova in the given range
 * @iovad: - iova domain pointer
 * @pfn_lo: - lower page frame address
 * @pfn_hi:- higher pfn adderss
 * This function allocates reserves the address range from pfn_lo to pfn_hi so
 * that this address is not dished out as part of alloc_iova.
 */
struct iova *
reserve_iova(struct iova_domain *iovad,
	unsigned long pfn_lo, unsigned long pfn_hi)
{
	struct rb_node *node;
	unsigned long flags;
	struct iova *iova;
	unsigned int overlap = 0;

687
688
689
690
	/* Don't allow nonsensical pfns */
	if (WARN_ON((pfn_hi | pfn_lo) > (ULLONG_MAX >> iova_shift(iovad))))
		return NULL;

691
	spin_lock_irqsave(&iovad->iova_rbtree_lock, flags);
692
693
	for (node = rb_first(&iovad->rbroot); node; node = rb_next(node)) {
		if (__is_range_overlap(node, pfn_lo, pfn_hi)) {
Geliang Tang's avatar
Geliang Tang committed
694
			iova = rb_entry(node, struct iova, node);
695
696
697
698
699
700
701
702
703
704
			__adjust_overlap_range(iova, &pfn_lo, &pfn_hi);
			if ((pfn_lo >= iova->pfn_lo) &&
				(pfn_hi <= iova->pfn_hi))
				goto finish;
			overlap = 1;

		} else if (overlap)
				break;
	}

Lucas De Marchi's avatar
Lucas De Marchi committed
705
	/* We are here either because this is the first reserver node
706
707
708
	 * or need to insert remaining non overlap addr range
	 */
	iova = __insert_new_range(iovad, pfn_lo, pfn_hi);
709
710
	if (iova)
		insert_update_32bit_pfn(iovad, iova);
711
712
finish:

713
	spin_unlock_irqrestore(&iovad->iova_rbtree_lock, flags);
714
715
	return iova;
}
Sakari Ailus's avatar
Sakari Ailus committed
716
EXPORT_SYMBOL_GPL(reserve_iova);
717
718
719
720
721
722
723
724
725
726
727
728
729
730

/**
 * copy_reserved_iova - copies the reserved between domains
 * @from: - source doamin from where to copy
 * @to: - destination domin where to copy
 * This function copies reserved iova's from one doamin to
 * other.
 */
void
copy_reserved_iova(struct iova_domain *from, struct iova_domain *to)
{
	unsigned long flags;
	struct rb_node *node;

731
	spin_lock_irqsave(&from->iova_rbtree_lock, flags);
732
	for (node = rb_first(&from->rbroot); node; node = rb_next(node)) {
Geliang Tang's avatar
Geliang Tang committed
733
		struct iova *iova = rb_entry(node, struct iova, node);
734
		struct iova *new_iova;
735

736
737
738
		if (iova->pfn_lo == IOVA_ANCHOR)
			continue;

739
740
741
742
743
		new_iova = reserve_iova(to, iova->pfn_lo, iova->pfn_hi);
		if (!new_iova)
			printk(KERN_ERR "Reserve iova range %lx@%lx failed\n",
				iova->pfn_lo, iova->pfn_lo);
	}
744
	spin_unlock_irqrestore(&from->iova_rbtree_lock, flags);
745
}
Sakari Ailus's avatar
Sakari Ailus committed
746
EXPORT_SYMBOL_GPL(copy_reserved_iova);
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770

struct iova *
split_and_remove_iova(struct iova_domain *iovad, struct iova *iova,
		      unsigned long pfn_lo, unsigned long pfn_hi)
{
	unsigned long flags;
	struct iova *prev = NULL, *next = NULL;

	spin_lock_irqsave(&iovad->iova_rbtree_lock, flags);
	if (iova->pfn_lo < pfn_lo) {
		prev = alloc_and_init_iova(iova->pfn_lo, pfn_lo - 1);
		if (prev == NULL)
			goto error;
	}
	if (iova->pfn_hi > pfn_hi) {
		next = alloc_and_init_iova(pfn_hi + 1, iova->pfn_hi);
		if (next == NULL)
			goto error;
	}

	__cached_rbnode_delete_update(iovad, iova);
	rb_erase(&iova->node, &iovad->rbroot);

	if (prev) {
771
		iova_insert_rbtree(&iovad->rbroot, prev, NULL);
772
773
774
		iova->pfn_lo = pfn_lo;
	}
	if (next) {
775
		iova_insert_rbtree(&iovad->rbroot, next, NULL);
776
777
778
779
780
781
782
783
784
785
786
787
		iova->pfn_hi = pfn_hi;
	}
	spin_unlock_irqrestore(&iovad->iova_rbtree_lock, flags);

	return iova;

error:
	spin_unlock_irqrestore(&iovad->iova_rbtree_lock, flags);
	if (prev)
		free_iova_mem(prev);
	return NULL;
}
788

789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
/*
 * Magazine caches for IOVA ranges.  For an introduction to magazines,
 * see the USENIX 2001 paper "Magazines and Vmem: Extending the Slab
 * Allocator to Many CPUs and Arbitrary Resources" by Bonwick and Adams.
 * For simplicity, we use a static magazine size and don't implement the
 * dynamic size tuning described in the paper.
 */

#define IOVA_MAG_SIZE 128

struct iova_magazine {
	unsigned long size;
	unsigned long pfns[IOVA_MAG_SIZE];
};

struct iova_cpu_rcache {
	spinlock_t lock;
	struct iova_magazine *loaded;
	struct iova_magazine *prev;
};

static struct iova_magazine *iova_magazine_alloc(gfp_t flags)
{
	return kzalloc(sizeof(struct iova_magazine), flags);
}

static void iova_magazine_free(struct iova_magazine *mag)
{
	kfree(mag);
}

static void
iova_magazine_free_pfns(struct iova_magazine *mag, struct iova_domain *iovad)
{
	unsigned long flags;
	int i;

	if (!mag)
		return;

	spin_lock_irqsave(&iovad->iova_rbtree_lock, flags);

	for (i = 0 ; i < mag->size; ++i) {
		struct iova *iova = private_find_iova(iovad, mag->pfns[i]);

		BUG_ON(!iova);
		private_free_iova(iovad, iova);
	}

	spin_unlock_irqrestore(&iovad->iova_rbtree_lock, flags);

	mag->size = 0;
}

static bool iova_magazine_full(struct iova_magazine *mag)
{
	return (mag && mag->size == IOVA_MAG_SIZE);
}

static bool iova_magazine_empty(struct iova_magazine *mag)
{
	return (!mag || mag->size == 0);
}

static unsigned long iova_magazine_pop(struct iova_magazine *mag,
				       unsigned long limit_pfn)
{
856
857
858
	int i;
	unsigned long pfn;

859
860
	BUG_ON(iova_magazine_empty(mag));

861
862
863
864
865
866
867
868
	/* Only fall back to the rbtree if we have no suitable pfns at all */
	for (i = mag->size - 1; mag->pfns[i] > limit_pfn; i--)
		if (i == 0)
			return 0;

	/* Swap it to pop it */
	pfn = mag->pfns[i];
	mag->pfns[i] = mag->pfns[--mag->size];
869

870
	return pfn;
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
}

static void iova_magazine_push(struct iova_magazine *mag, unsigned long pfn)
{
	BUG_ON(iova_magazine_full(mag));

	mag->pfns[mag->size++] = pfn;
}

static void init_iova_rcaches(struct iova_domain *iovad)
{
	struct iova_cpu_rcache *cpu_rcache;
	struct iova_rcache *rcache;
	unsigned int cpu;
	int i;

	for (i = 0; i < IOVA_RANGE_CACHE_MAX_SIZE; ++i) {
		rcache = &iovad->rcaches[i];
		spin_lock_init(&rcache->lock);
		rcache->depot_size = 0;
		rcache->cpu_rcaches = __alloc_percpu(sizeof(*cpu_rcache), cache_line_size());
		if (WARN_ON(!rcache->cpu_rcaches))
			continue;
		for_each_possible_cpu(cpu) {
			cpu_rcache = per_cpu_ptr(rcache->cpu_rcaches, cpu);
			spin_lock_init(&cpu_rcache->lock);
			cpu_rcache->loaded = iova_magazine_alloc(GFP_KERNEL);
			cpu_rcache->prev = iova_magazine_alloc(GFP_KERNEL);
		}
	}
}

/*
 * Try inserting IOVA range starting with 'iova_pfn' into 'rcache', and
 * return true on success.  Can fail if rcache is full and we can't free
 * space, and free_iova() (our only caller) will then return the IOVA
 * range to the rbtree instead.
 */
static bool __iova_rcache_insert(struct iova_domain *iovad,
				 struct iova_rcache *rcache,
				 unsigned long iova_pfn)
{
	struct iova_magazine *mag_to_free = NULL;
	struct iova_cpu_rcache *cpu_rcache;
	bool can_insert = false;
	unsigned long flags;

918
	cpu_rcache = raw_cpu_ptr(rcache->cpu_rcaches);
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
	spin_lock_irqsave(&cpu_rcache->lock, flags);

	if (!iova_magazine_full(cpu_rcache->loaded)) {
		can_insert = true;
	} else if (!iova_magazine_full(cpu_rcache->prev)) {
		swap(cpu_rcache->prev, cpu_rcache->loaded);
		can_insert = true;
	} else {
		struct iova_magazine *new_mag = iova_magazine_alloc(GFP_ATOMIC);

		if (new_mag) {
			spin_lock(&rcache->lock);
			if (rcache->depot_size < MAX_GLOBAL_MAGS) {
				rcache->depot[rcache->depot_size++] =
						cpu_rcache->loaded;
			} else {
				mag_to_free = cpu_rcache->loaded;
			}
			spin_unlock(&rcache->lock);

			cpu_rcache->loaded = new_mag;
			can_insert = true;
		}
	}

	if (can_insert)
		iova_magazine_push(cpu_rcache->loaded, iova_pfn);

	spin_unlock_irqrestore(&cpu_rcache->lock, flags);

	if (mag_to_free) {
		iova_magazine_free_pfns(mag_to_free, iovad);
		iova_magazine_free(mag_to_free);
	}

	return can_insert;
}

static bool iova_rcache_insert(struct iova_domain *iovad, unsigned long pfn,
			       unsigned long size)
{
	unsigned int log_size = order_base_2(size);

	if (log_size >= IOVA_RANGE_CACHE_MAX_SIZE)
		return false;

	return __iova_rcache_insert(iovad, &iovad->rcaches[log_size], pfn);
}

/*
 * Caller wants to allocate a new IOVA range from 'rcache'.  If we can
 * satisfy the request, return a matching non-NULL range and remove
 * it from the 'rcache'.
 */
static unsigned long __iova_rcache_get(struct iova_rcache *rcache,
				       unsigned long limit_pfn)
{
	struct iova_cpu_rcache *cpu_rcache;
	unsigned long iova_pfn = 0;
	bool has_pfn = false;
	unsigned long flags;

981
	cpu_rcache = raw_cpu_ptr(rcache->cpu_rcaches);
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
	spin_lock_irqsave(&cpu_rcache->lock, flags);

	if (!iova_magazine_empty(cpu_rcache->loaded)) {
		has_pfn = true;
	} else if (!iova_magazine_empty(cpu_rcache->prev)) {
		swap(cpu_rcache->prev, cpu_rcache->loaded);
		has_pfn = true;
	} else {
		spin_lock(&rcache->lock);
		if (rcache->depot_size > 0) {
			iova_magazine_free(cpu_rcache->loaded);
			cpu_rcache->loaded = rcache->depot[--rcache->depot_size];
			has_pfn = true;
		}
		spin_unlock(&rcache->lock);
	}

	if (has_pfn)
		iova_pfn = iova_magazine_pop(cpu_rcache->loaded, limit_pfn);