hugetlb.c 49.8 KB
Newer Older
Linus Torvalds's avatar
Linus Torvalds committed
1
2
3
4
5
6
7
8
9
10
11
12
/*
 * Generic hugetlb support.
 * (C) William Irwin, April 2004
 */
#include <linux/gfp.h>
#include <linux/list.h>
#include <linux/init.h>
#include <linux/module.h>
#include <linux/mm.h>
#include <linux/sysctl.h>
#include <linux/highmem.h>
#include <linux/nodemask.h>
David Gibson's avatar
David Gibson committed
13
#include <linux/pagemap.h>
14
#include <linux/mempolicy.h>
15
#include <linux/cpuset.h>
16
#include <linux/mutex.h>
17

David Gibson's avatar
David Gibson committed
18
19
20
21
#include <asm/page.h>
#include <asm/pgtable.h>

#include <linux/hugetlb.h>
22
#include "internal.h"
Linus Torvalds's avatar
Linus Torvalds committed
23
24

const unsigned long hugetlb_zero = 0, hugetlb_infinity = ~0UL;
25
26
static gfp_t htlb_alloc_mask = GFP_HIGHUSER;
unsigned long hugepages_treat_as_movable;
27

28
29
30
31
32
33
34
35
36
37
static int max_hstate;
unsigned int default_hstate_idx;
struct hstate hstates[HUGE_MAX_HSTATE];

/* for command line parsing */
static struct hstate * __initdata parsed_hstate;
static unsigned long __initdata default_hstate_max_huge_pages;

#define for_each_hstate(h) \
	for ((h) = hstates; (h) < &hstates[max_hstate]; (h)++)
38

39
40
41
42
/*
 * Protects updates to hugepage_freelists, nr_huge_pages, and free_huge_pages
 */
static DEFINE_SPINLOCK(hugetlb_lock);
43

44
45
46
/*
 * Region tracking -- allows tracking of reservations and instantiated pages
 *                    across the pages in a mapping.
47
48
49
50
51
52
53
54
55
56
 *
 * The region data structures are protected by a combination of the mmap_sem
 * and the hugetlb_instantion_mutex.  To access or modify a region the caller
 * must either hold the mmap_sem for write, or the mmap_sem for read and
 * the hugetlb_instantiation mutex:
 *
 * 	down_write(&mm->mmap_sem);
 * or
 * 	down_read(&mm->mmap_sem);
 * 	mutex_lock(&hugetlb_instantiation_mutex);
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
 */
struct file_region {
	struct list_head link;
	long from;
	long to;
};

static long region_add(struct list_head *head, long f, long t)
{
	struct file_region *rg, *nrg, *trg;

	/* Locate the region we are either in or before. */
	list_for_each_entry(rg, head, link)
		if (f <= rg->to)
			break;

	/* Round our left edge to the current segment if it encloses us. */
	if (f > rg->from)
		f = rg->from;

	/* Check for and consume any regions we now overlap with. */
	nrg = rg;
	list_for_each_entry_safe(rg, trg, rg->link.prev, link) {
		if (&rg->link == head)
			break;
		if (rg->from > t)
			break;

		/* If this area reaches higher then extend our area to
		 * include it completely.  If this is not the first area
		 * which we intend to reuse, free it. */
		if (rg->to > t)
			t = rg->to;
		if (rg != nrg) {
			list_del(&rg->link);
			kfree(rg);
		}
	}
	nrg->from = f;
	nrg->to = t;
	return 0;
}

static long region_chg(struct list_head *head, long f, long t)
{
	struct file_region *rg, *nrg;
	long chg = 0;

	/* Locate the region we are before or in. */
	list_for_each_entry(rg, head, link)
		if (f <= rg->to)
			break;

	/* If we are below the current region then a new region is required.
	 * Subtle, allocate a new region at the position but make it zero
	 * size such that we can guarantee to record the reservation. */
	if (&rg->link == head || t < rg->from) {
		nrg = kmalloc(sizeof(*nrg), GFP_KERNEL);
		if (!nrg)
			return -ENOMEM;
		nrg->from = f;
		nrg->to   = f;
		INIT_LIST_HEAD(&nrg->link);
		list_add(&nrg->link, rg->link.prev);

		return t - f;
	}

	/* Round our left edge to the current segment if it encloses us. */
	if (f > rg->from)
		f = rg->from;
	chg = t - f;

	/* Check for and consume any regions we now overlap with. */
	list_for_each_entry(rg, rg->link.prev, link) {
		if (&rg->link == head)
			break;
		if (rg->from > t)
			return chg;

		/* We overlap with this area, if it extends futher than
		 * us then we must extend ourselves.  Account for its
		 * existing reservation. */
		if (rg->to > t) {
			chg += rg->to - t;
			t = rg->to;
		}
		chg -= rg->to - rg->from;
	}
	return chg;
}

static long region_truncate(struct list_head *head, long end)
{
	struct file_region *rg, *trg;
	long chg = 0;

	/* Locate the region we are either in or before. */
	list_for_each_entry(rg, head, link)
		if (end <= rg->to)
			break;
	if (&rg->link == head)
		return 0;

	/* If we are in the middle of a region then adjust it. */
	if (end > rg->from) {
		chg = rg->to - end;
		rg->to = end;
		rg = list_entry(rg->link.next, typeof(*rg), link);
	}

	/* Drop any remaining regions. */
	list_for_each_entry_safe(rg, trg, rg->link.prev, link) {
		if (&rg->link == head)
			break;
		chg += rg->to - rg->from;
		list_del(&rg->link);
		kfree(rg);
	}
	return chg;
}

179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
static long region_count(struct list_head *head, long f, long t)
{
	struct file_region *rg;
	long chg = 0;

	/* Locate each segment we overlap with, and count that overlap. */
	list_for_each_entry(rg, head, link) {
		int seg_from;
		int seg_to;

		if (rg->to <= f)
			continue;
		if (rg->from >= t)
			break;

		seg_from = max(rg->from, f);
		seg_to = min(rg->to, t);

		chg += seg_to - seg_from;
	}

	return chg;
}

203
204
205
206
/*
 * Convert the address within this vma to the page offset within
 * the mapping, in pagecache page units; huge pages here.
 */
207
208
static pgoff_t vma_hugecache_offset(struct hstate *h,
			struct vm_area_struct *vma, unsigned long address)
209
{
210
211
	return ((address - vma->vm_start) >> huge_page_shift(h)) +
			(vma->vm_pgoff >> huge_page_order(h));
212
213
}

214
215
216
217
218
219
220
/*
 * Flags for MAP_PRIVATE reservations.  These are stored in the bottom
 * bits of the reservation map pointer, which are always clear due to
 * alignment.
 */
#define HPAGE_RESV_OWNER    (1UL << 0)
#define HPAGE_RESV_UNMAPPED (1UL << 1)
221
#define HPAGE_RESV_MASK (HPAGE_RESV_OWNER | HPAGE_RESV_UNMAPPED)
222

223
224
225
226
227
228
229
230
231
/*
 * These helpers are used to track how many pages are reserved for
 * faults in a MAP_PRIVATE mapping. Only the process that called mmap()
 * is guaranteed to have their future faults succeed.
 *
 * With the exception of reset_vma_resv_huge_pages() which is called at fork(),
 * the reserve counters are updated with the hugetlb_lock held. It is safe
 * to reset the VMA at fork() time as it is not in use yet and there is no
 * chance of the global counters getting corrupted as a result of the values.
232
233
234
235
236
237
238
239
240
 *
 * The private mapping reservation is represented in a subtly different
 * manner to a shared mapping.  A shared mapping has a region map associated
 * with the underlying file, this region map represents the backing file
 * pages which have ever had a reservation assigned which this persists even
 * after the page is instantiated.  A private mapping has a region map
 * associated with the original mmap which is attached to all VMAs which
 * reference it, this region map represents those offsets which have consumed
 * reservation ie. where pages have been instantiated.
241
 */
242
243
244
245
246
247
248
249
250
251
252
static unsigned long get_vma_private_data(struct vm_area_struct *vma)
{
	return (unsigned long)vma->vm_private_data;
}

static void set_vma_private_data(struct vm_area_struct *vma,
							unsigned long value)
{
	vma->vm_private_data = (void *)value;
}

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
struct resv_map {
	struct kref refs;
	struct list_head regions;
};

struct resv_map *resv_map_alloc(void)
{
	struct resv_map *resv_map = kmalloc(sizeof(*resv_map), GFP_KERNEL);
	if (!resv_map)
		return NULL;

	kref_init(&resv_map->refs);
	INIT_LIST_HEAD(&resv_map->regions);

	return resv_map;
}

void resv_map_release(struct kref *ref)
{
	struct resv_map *resv_map = container_of(ref, struct resv_map, refs);

	/* Clear out any active regions before we release the map. */
	region_truncate(&resv_map->regions, 0);
	kfree(resv_map);
}

static struct resv_map *vma_resv_map(struct vm_area_struct *vma)
280
281
282
{
	VM_BUG_ON(!is_vm_hugetlb_page(vma));
	if (!(vma->vm_flags & VM_SHARED))
283
284
		return (struct resv_map *)(get_vma_private_data(vma) &
							~HPAGE_RESV_MASK);
285
286
287
	return 0;
}

288
static void set_vma_resv_map(struct vm_area_struct *vma, struct resv_map *map)
289
290
291
292
{
	VM_BUG_ON(!is_vm_hugetlb_page(vma));
	VM_BUG_ON(vma->vm_flags & VM_SHARED);

293
294
	set_vma_private_data(vma, (get_vma_private_data(vma) &
				HPAGE_RESV_MASK) | (unsigned long)map);
295
296
297
298
299
}

static void set_vma_resv_flags(struct vm_area_struct *vma, unsigned long flags)
{
	VM_BUG_ON(!is_vm_hugetlb_page(vma));
300
301
302
	VM_BUG_ON(vma->vm_flags & VM_SHARED);

	set_vma_private_data(vma, get_vma_private_data(vma) | flags);
303
304
305
306
307
}

static int is_vma_resv_set(struct vm_area_struct *vma, unsigned long flag)
{
	VM_BUG_ON(!is_vm_hugetlb_page(vma));
308
309

	return (get_vma_private_data(vma) & flag) != 0;
310
311
312
}

/* Decrement the reserved pages in the hugepage pool by one */
313
314
static void decrement_hugepage_resv_vma(struct hstate *h,
			struct vm_area_struct *vma)
315
{
316
317
318
	if (vma->vm_flags & VM_NORESERVE)
		return;

319
320
	if (vma->vm_flags & VM_SHARED) {
		/* Shared mappings always use reserves */
321
		h->resv_huge_pages--;
322
	} else if (is_vma_resv_set(vma, HPAGE_RESV_OWNER)) {
323
324
325
326
		/*
		 * Only the process that called mmap() has reserves for
		 * private mappings.
		 */
327
		h->resv_huge_pages--;
328
329
330
	}
}

331
/* Reset counters to 0 and clear all HPAGE_RESV_* flags */
332
333
334
335
336
337
338
339
340
341
342
343
void reset_vma_resv_huge_pages(struct vm_area_struct *vma)
{
	VM_BUG_ON(!is_vm_hugetlb_page(vma));
	if (!(vma->vm_flags & VM_SHARED))
		vma->vm_private_data = (void *)0;
}

/* Returns true if the VMA has associated reserve pages */
static int vma_has_private_reserves(struct vm_area_struct *vma)
{
	if (vma->vm_flags & VM_SHARED)
		return 0;
344
	if (!is_vma_resv_set(vma, HPAGE_RESV_OWNER))
345
346
347
348
		return 0;
	return 1;
}

349
350
static void clear_huge_page(struct page *page,
			unsigned long addr, unsigned long sz)
351
352
353
354
{
	int i;

	might_sleep();
355
	for (i = 0; i < sz/PAGE_SIZE; i++) {
356
		cond_resched();
357
		clear_user_highpage(page + i, addr + i * PAGE_SIZE);
358
359
360
361
	}
}

static void copy_huge_page(struct page *dst, struct page *src,
362
			   unsigned long addr, struct vm_area_struct *vma)
363
364
{
	int i;
365
	struct hstate *h = hstate_vma(vma);
366
367

	might_sleep();
368
	for (i = 0; i < pages_per_huge_page(h); i++) {
369
		cond_resched();
370
		copy_user_highpage(dst + i, src + i, addr + i*PAGE_SIZE, vma);
371
372
373
	}
}

374
static void enqueue_huge_page(struct hstate *h, struct page *page)
Linus Torvalds's avatar
Linus Torvalds committed
375
376
{
	int nid = page_to_nid(page);
377
378
379
	list_add(&page->lru, &h->hugepage_freelists[nid]);
	h->free_huge_pages++;
	h->free_huge_pages_node[nid]++;
Linus Torvalds's avatar
Linus Torvalds committed
380
381
}

382
static struct page *dequeue_huge_page(struct hstate *h)
383
384
385
386
387
{
	int nid;
	struct page *page = NULL;

	for (nid = 0; nid < MAX_NUMNODES; ++nid) {
388
389
		if (!list_empty(&h->hugepage_freelists[nid])) {
			page = list_entry(h->hugepage_freelists[nid].next,
390
391
					  struct page, lru);
			list_del(&page->lru);
392
393
			h->free_huge_pages--;
			h->free_huge_pages_node[nid]--;
394
395
396
397
398
399
			break;
		}
	}
	return page;
}

400
401
static struct page *dequeue_huge_page_vma(struct hstate *h,
				struct vm_area_struct *vma,
402
				unsigned long address, int avoid_reserve)
Linus Torvalds's avatar
Linus Torvalds committed
403
{
404
	int nid;
Linus Torvalds's avatar
Linus Torvalds committed
405
	struct page *page = NULL;
406
	struct mempolicy *mpol;
407
	nodemask_t *nodemask;
408
	struct zonelist *zonelist = huge_zonelist(vma, address,
409
					htlb_alloc_mask, &mpol, &nodemask);
410
411
	struct zone *zone;
	struct zoneref *z;
Linus Torvalds's avatar
Linus Torvalds committed
412

413
414
415
416
417
418
	/*
	 * A child process with MAP_PRIVATE mappings created by their parent
	 * have no page reserves. This check ensures that reservations are
	 * not "stolen". The child may still get SIGKILLed
	 */
	if (!vma_has_private_reserves(vma) &&
419
			h->free_huge_pages - h->resv_huge_pages == 0)
420
421
		return NULL;

422
	/* If reserves cannot be used, ensure enough pages are in the pool */
423
	if (avoid_reserve && h->free_huge_pages - h->resv_huge_pages == 0)
424
425
		return NULL;

426
427
	for_each_zone_zonelist_nodemask(zone, z, zonelist,
						MAX_NR_ZONES - 1, nodemask) {
428
429
		nid = zone_to_nid(zone);
		if (cpuset_zone_allowed_softwall(zone, htlb_alloc_mask) &&
430
431
		    !list_empty(&h->hugepage_freelists[nid])) {
			page = list_entry(h->hugepage_freelists[nid].next,
432
433
					  struct page, lru);
			list_del(&page->lru);
434
435
			h->free_huge_pages--;
			h->free_huge_pages_node[nid]--;
436
437

			if (!avoid_reserve)
438
				decrement_hugepage_resv_vma(h, vma);
439

Ken Chen's avatar
Ken Chen committed
440
			break;
441
		}
Linus Torvalds's avatar
Linus Torvalds committed
442
	}
443
	mpol_cond_put(mpol);
Linus Torvalds's avatar
Linus Torvalds committed
444
445
446
	return page;
}

447
static void update_and_free_page(struct hstate *h, struct page *page)
448
449
{
	int i;
450
451
452
453

	h->nr_huge_pages--;
	h->nr_huge_pages_node[page_to_nid(page)]--;
	for (i = 0; i < pages_per_huge_page(h); i++) {
454
455
456
457
458
459
		page[i].flags &= ~(1 << PG_locked | 1 << PG_error | 1 << PG_referenced |
				1 << PG_dirty | 1 << PG_active | 1 << PG_reserved |
				1 << PG_private | 1<< PG_writeback);
	}
	set_compound_page_dtor(page, NULL);
	set_page_refcounted(page);
460
	arch_release_hugepage(page);
461
	__free_pages(page, huge_page_order(h));
462
463
}

464
465
466
467
468
469
470
471
472
473
474
struct hstate *size_to_hstate(unsigned long size)
{
	struct hstate *h;

	for_each_hstate(h) {
		if (huge_page_size(h) == size)
			return h;
	}
	return NULL;
}

475
476
static void free_huge_page(struct page *page)
{
477
478
479
480
	/*
	 * Can't pass hstate in here because it is called from the
	 * compound page destructor.
	 */
481
	struct hstate *h = page_hstate(page);
482
	int nid = page_to_nid(page);
483
	struct address_space *mapping;
484

485
	mapping = (struct address_space *) page_private(page);
486
	set_page_private(page, 0);
487
	BUG_ON(page_count(page));
488
489
490
	INIT_LIST_HEAD(&page->lru);

	spin_lock(&hugetlb_lock);
491
492
493
494
	if (h->surplus_huge_pages_node[nid]) {
		update_and_free_page(h, page);
		h->surplus_huge_pages--;
		h->surplus_huge_pages_node[nid]--;
495
	} else {
496
		enqueue_huge_page(h, page);
497
	}
498
	spin_unlock(&hugetlb_lock);
499
	if (mapping)
500
		hugetlb_put_quota(mapping, 1);
501
502
}

503
504
505
506
507
/*
 * Increment or decrement surplus_huge_pages.  Keep node-specific counters
 * balanced by operating on them in a round-robin fashion.
 * Returns 1 if an adjustment was made.
 */
508
static int adjust_pool_surplus(struct hstate *h, int delta)
509
510
511
512
513
514
515
516
517
518
519
520
{
	static int prev_nid;
	int nid = prev_nid;
	int ret = 0;

	VM_BUG_ON(delta != -1 && delta != 1);
	do {
		nid = next_node(nid, node_online_map);
		if (nid == MAX_NUMNODES)
			nid = first_node(node_online_map);

		/* To shrink on this node, there must be a surplus page */
521
		if (delta < 0 && !h->surplus_huge_pages_node[nid])
522
523
			continue;
		/* Surplus cannot exceed the total number of pages */
524
525
		if (delta > 0 && h->surplus_huge_pages_node[nid] >=
						h->nr_huge_pages_node[nid])
526
527
			continue;

528
529
		h->surplus_huge_pages += delta;
		h->surplus_huge_pages_node[nid] += delta;
530
531
532
533
534
535
536
537
		ret = 1;
		break;
	} while (nid != prev_nid);

	prev_nid = nid;
	return ret;
}

538
static void prep_new_huge_page(struct hstate *h, struct page *page, int nid)
539
540
541
{
	set_compound_page_dtor(page, free_huge_page);
	spin_lock(&hugetlb_lock);
542
543
	h->nr_huge_pages++;
	h->nr_huge_pages_node[nid]++;
544
545
546
547
	spin_unlock(&hugetlb_lock);
	put_page(page); /* free it into the hugepage allocator */
}

548
static struct page *alloc_fresh_huge_page_node(struct hstate *h, int nid)
Linus Torvalds's avatar
Linus Torvalds committed
549
550
{
	struct page *page;
551

552
	page = alloc_pages_node(nid,
553
554
		htlb_alloc_mask|__GFP_COMP|__GFP_THISNODE|
						__GFP_REPEAT|__GFP_NOWARN,
555
		huge_page_order(h));
Linus Torvalds's avatar
Linus Torvalds committed
556
	if (page) {
557
558
		if (arch_prepare_hugepage(page)) {
			__free_pages(page, HUGETLB_PAGE_ORDER);
559
			return NULL;
560
		}
561
		prep_new_huge_page(h, page, nid);
Linus Torvalds's avatar
Linus Torvalds committed
562
	}
563
564
565
566

	return page;
}

567
static int alloc_fresh_huge_page(struct hstate *h)
568
569
570
571
572
573
{
	struct page *page;
	int start_nid;
	int next_nid;
	int ret = 0;

574
	start_nid = h->hugetlb_next_nid;
575
576

	do {
577
		page = alloc_fresh_huge_page_node(h, h->hugetlb_next_nid);
578
579
580
581
582
583
584
585
586
587
588
589
590
		if (page)
			ret = 1;
		/*
		 * Use a helper variable to find the next node and then
		 * copy it back to hugetlb_next_nid afterwards:
		 * otherwise there's a window in which a racer might
		 * pass invalid nid MAX_NUMNODES to alloc_pages_node.
		 * But we don't need to use a spin_lock here: it really
		 * doesn't matter if occasionally a racer chooses the
		 * same nid as we do.  Move nid forward in the mask even
		 * if we just successfully allocated a hugepage so that
		 * the next caller gets hugepages on the next node.
		 */
591
		next_nid = next_node(h->hugetlb_next_nid, node_online_map);
592
593
		if (next_nid == MAX_NUMNODES)
			next_nid = first_node(node_online_map);
594
595
		h->hugetlb_next_nid = next_nid;
	} while (!page && h->hugetlb_next_nid != start_nid);
596

597
598
599
600
601
	if (ret)
		count_vm_event(HTLB_BUDDY_PGALLOC);
	else
		count_vm_event(HTLB_BUDDY_PGALLOC_FAIL);

602
	return ret;
Linus Torvalds's avatar
Linus Torvalds committed
603
604
}

605
606
static struct page *alloc_buddy_huge_page(struct hstate *h,
			struct vm_area_struct *vma, unsigned long address)
607
608
{
	struct page *page;
609
	unsigned int nid;
610

611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
	/*
	 * Assume we will successfully allocate the surplus page to
	 * prevent racing processes from causing the surplus to exceed
	 * overcommit
	 *
	 * This however introduces a different race, where a process B
	 * tries to grow the static hugepage pool while alloc_pages() is
	 * called by process A. B will only examine the per-node
	 * counters in determining if surplus huge pages can be
	 * converted to normal huge pages in adjust_pool_surplus(). A
	 * won't be able to increment the per-node counter, until the
	 * lock is dropped by B, but B doesn't drop hugetlb_lock until
	 * no more huge pages can be converted from surplus to normal
	 * state (and doesn't try to convert again). Thus, we have a
	 * case where a surplus huge page exists, the pool is grown, and
	 * the surplus huge page still exists after, even though it
	 * should just have been converted to a normal huge page. This
	 * does not leak memory, though, as the hugepage will be freed
	 * once it is out of use. It also does not allow the counters to
	 * go out of whack in adjust_pool_surplus() as we don't modify
	 * the node values until we've gotten the hugepage and only the
	 * per-node value is checked there.
	 */
	spin_lock(&hugetlb_lock);
635
	if (h->surplus_huge_pages >= h->nr_overcommit_huge_pages) {
636
637
638
		spin_unlock(&hugetlb_lock);
		return NULL;
	} else {
639
640
		h->nr_huge_pages++;
		h->surplus_huge_pages++;
641
642
643
	}
	spin_unlock(&hugetlb_lock);

644
645
	page = alloc_pages(htlb_alloc_mask|__GFP_COMP|
					__GFP_REPEAT|__GFP_NOWARN,
646
					huge_page_order(h));
647
648

	spin_lock(&hugetlb_lock);
649
	if (page) {
650
651
652
653
654
655
		/*
		 * This page is now managed by the hugetlb allocator and has
		 * no users -- drop the buddy allocator's reference.
		 */
		put_page_testzero(page);
		VM_BUG_ON(page_count(page));
656
		nid = page_to_nid(page);
657
		set_compound_page_dtor(page, free_huge_page);
658
659
660
		/*
		 * We incremented the global counters already
		 */
661
662
		h->nr_huge_pages_node[nid]++;
		h->surplus_huge_pages_node[nid]++;
663
		__count_vm_event(HTLB_BUDDY_PGALLOC);
664
	} else {
665
666
		h->nr_huge_pages--;
		h->surplus_huge_pages--;
667
		__count_vm_event(HTLB_BUDDY_PGALLOC_FAIL);
668
	}
669
	spin_unlock(&hugetlb_lock);
670
671
672
673

	return page;
}

674
675
676
677
/*
 * Increase the hugetlb pool such that it can accomodate a reservation
 * of size 'delta'.
 */
678
static int gather_surplus_pages(struct hstate *h, int delta)
679
680
681
682
683
684
{
	struct list_head surplus_list;
	struct page *page, *tmp;
	int ret, i;
	int needed, allocated;

685
	needed = (h->resv_huge_pages + delta) - h->free_huge_pages;
686
	if (needed <= 0) {
687
		h->resv_huge_pages += delta;
688
		return 0;
689
	}
690
691
692
693
694
695
696
697

	allocated = 0;
	INIT_LIST_HEAD(&surplus_list);

	ret = -ENOMEM;
retry:
	spin_unlock(&hugetlb_lock);
	for (i = 0; i < needed; i++) {
698
		page = alloc_buddy_huge_page(h, NULL, 0);
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
		if (!page) {
			/*
			 * We were not able to allocate enough pages to
			 * satisfy the entire reservation so we free what
			 * we've allocated so far.
			 */
			spin_lock(&hugetlb_lock);
			needed = 0;
			goto free;
		}

		list_add(&page->lru, &surplus_list);
	}
	allocated += needed;

	/*
	 * After retaking hugetlb_lock, we need to recalculate 'needed'
	 * because either resv_huge_pages or free_huge_pages may have changed.
	 */
	spin_lock(&hugetlb_lock);
719
720
	needed = (h->resv_huge_pages + delta) -
			(h->free_huge_pages + allocated);
721
722
723
724
725
726
727
	if (needed > 0)
		goto retry;

	/*
	 * The surplus_list now contains _at_least_ the number of extra pages
	 * needed to accomodate the reservation.  Add the appropriate number
	 * of pages to the hugetlb pool and free the extras back to the buddy
728
729
730
	 * allocator.  Commit the entire reservation here to prevent another
	 * process from stealing the pages as they are added to the pool but
	 * before they are reserved.
731
732
	 */
	needed += allocated;
733
	h->resv_huge_pages += delta;
734
735
	ret = 0;
free:
736
	/* Free the needed pages to the hugetlb pool */
737
	list_for_each_entry_safe(page, tmp, &surplus_list, lru) {
738
739
		if ((--needed) < 0)
			break;
740
		list_del(&page->lru);
741
		enqueue_huge_page(h, page);
742
743
744
745
746
747
748
	}

	/* Free unnecessary surplus pages to the buddy allocator */
	if (!list_empty(&surplus_list)) {
		spin_unlock(&hugetlb_lock);
		list_for_each_entry_safe(page, tmp, &surplus_list, lru) {
			list_del(&page->lru);
749
			/*
750
751
752
			 * The page has a reference count of zero already, so
			 * call free_huge_page directly instead of using
			 * put_page.  This must be done with hugetlb_lock
753
754
755
			 * unlocked which is safe because free_huge_page takes
			 * hugetlb_lock before deciding how to free the page.
			 */
756
			free_huge_page(page);
757
		}
758
		spin_lock(&hugetlb_lock);
759
760
761
762
763
764
765
766
767
768
	}

	return ret;
}

/*
 * When releasing a hugetlb pool reservation, any surplus pages that were
 * allocated to satisfy the reservation must be explicitly freed if they were
 * never used.
 */
769
770
static void return_unused_surplus_pages(struct hstate *h,
					unsigned long unused_resv_pages)
771
772
773
774
775
{
	static int nid = -1;
	struct page *page;
	unsigned long nr_pages;

776
777
778
779
780
781
782
783
	/*
	 * We want to release as many surplus pages as possible, spread
	 * evenly across all nodes. Iterate across all nodes until we
	 * can no longer free unreserved surplus pages. This occurs when
	 * the nodes with surplus pages have no free pages.
	 */
	unsigned long remaining_iterations = num_online_nodes();

784
	/* Uncommit the reservation */
785
	h->resv_huge_pages -= unused_resv_pages;
786

787
	nr_pages = min(unused_resv_pages, h->surplus_huge_pages);
788

789
	while (remaining_iterations-- && nr_pages) {
790
791
792
793
		nid = next_node(nid, node_online_map);
		if (nid == MAX_NUMNODES)
			nid = first_node(node_online_map);

794
		if (!h->surplus_huge_pages_node[nid])
795
796
			continue;

797
798
		if (!list_empty(&h->hugepage_freelists[nid])) {
			page = list_entry(h->hugepage_freelists[nid].next,
799
800
					  struct page, lru);
			list_del(&page->lru);
801
802
803
804
805
			update_and_free_page(h, page);
			h->free_huge_pages--;
			h->free_huge_pages_node[nid]--;
			h->surplus_huge_pages--;
			h->surplus_huge_pages_node[nid]--;
806
			nr_pages--;
807
			remaining_iterations = num_online_nodes();
808
809
810
811
		}
	}
}

812
813
814
815
816
817
818
819
820
/*
 * Determine if the huge page at addr within the vma has an associated
 * reservation.  Where it does not we will need to logically increase
 * reservation and actually increase quota before an allocation can occur.
 * Where any new reservation would be required the reservation change is
 * prepared, but not committed.  Once the page has been quota'd allocated
 * an instantiated the change should be committed via vma_commit_reservation.
 * No action is required on failure.
 */
821
822
static int vma_needs_reservation(struct hstate *h,
			struct vm_area_struct *vma, unsigned long addr)
823
824
825
826
827
{
	struct address_space *mapping = vma->vm_file->f_mapping;
	struct inode *inode = mapping->host;

	if (vma->vm_flags & VM_SHARED) {
828
		pgoff_t idx = vma_hugecache_offset(h, vma, addr);
829
830
831
		return region_chg(&inode->i_mapping->private_list,
							idx, idx + 1);

832
833
	} else if (!is_vma_resv_set(vma, HPAGE_RESV_OWNER)) {
		return 1;
834

835
836
	} else  {
		int err;
837
		pgoff_t idx = vma_hugecache_offset(h, vma, addr);
838
839
840
841
842
843
844
		struct resv_map *reservations = vma_resv_map(vma);

		err = region_chg(&reservations->regions, idx, idx + 1);
		if (err < 0)
			return err;
		return 0;
	}
845
}
846
847
static void vma_commit_reservation(struct hstate *h,
			struct vm_area_struct *vma, unsigned long addr)
848
849
850
851
852
{
	struct address_space *mapping = vma->vm_file->f_mapping;
	struct inode *inode = mapping->host;

	if (vma->vm_flags & VM_SHARED) {
853
		pgoff_t idx = vma_hugecache_offset(h, vma, addr);
854
		region_add(&inode->i_mapping->private_list, idx, idx + 1);
855
856

	} else if (is_vma_resv_set(vma, HPAGE_RESV_OWNER)) {
857
		pgoff_t idx = vma_hugecache_offset(h, vma, addr);
858
859
860
861
		struct resv_map *reservations = vma_resv_map(vma);

		/* Mark this page used in the map. */
		region_add(&reservations->regions, idx, idx + 1);
862
863
864
	}
}

865
static struct page *alloc_huge_page(struct vm_area_struct *vma,
866
				    unsigned long addr, int avoid_reserve)
Linus Torvalds's avatar
Linus Torvalds committed
867
{
868
	struct hstate *h = hstate_vma(vma);
869
	struct page *page;
870
871
	struct address_space *mapping = vma->vm_file->f_mapping;
	struct inode *inode = mapping->host;
872
	unsigned int chg;
873
874
875
876
877

	/*
	 * Processes that did not create the mapping will have no reserves and
	 * will not have accounted against quota. Check that the quota can be
	 * made before satisfying the allocation
878
879
	 * MAP_NORESERVE mappings may also need pages and quota allocated
	 * if no reserve mapping overlaps.
880
	 */
881
	chg = vma_needs_reservation(h, vma, addr);
882
883
884
	if (chg < 0)
		return ERR_PTR(chg);
	if (chg)
885
886
		if (hugetlb_get_quota(inode->i_mapping, chg))
			return ERR_PTR(-ENOSPC);
Linus Torvalds's avatar
Linus Torvalds committed
887
888

	spin_lock(&hugetlb_lock);
889
	page = dequeue_huge_page_vma(h, vma, addr, avoid_reserve);
Linus Torvalds's avatar
Linus Torvalds committed
890
	spin_unlock(&hugetlb_lock);
891

Ken Chen's avatar
Ken Chen committed
892
	if (!page) {
893
		page = alloc_buddy_huge_page(h, vma, addr);
Ken Chen's avatar
Ken Chen committed
894
		if (!page) {
895
			hugetlb_put_quota(inode->i_mapping, chg);
Ken Chen's avatar
Ken Chen committed
896
897
898
			return ERR_PTR(-VM_FAULT_OOM);
		}
	}
899

900
901
	set_page_refcounted(page);
	set_page_private(page, (unsigned long) mapping);
902

903
	vma_commit_reservation(h, vma, addr);
904

905
	return page;
906
907
}

908
static void __init hugetlb_init_one_hstate(struct hstate *h)
Linus Torvalds's avatar
Linus Torvalds committed
909
910
{
	unsigned long i;
911

Linus Torvalds's avatar
Linus Torvalds committed
912
	for (i = 0; i < MAX_NUMNODES; ++i)
913
		INIT_LIST_HEAD(&h->hugepage_freelists[i]);
Linus Torvalds's avatar
Linus Torvalds committed
914

915
	h->hugetlb_next_nid = first_node(node_online_map);
916

917
	for (i = 0; i < h->max_huge_pages; ++i) {
918
		if (!alloc_fresh_huge_page(h))
Linus Torvalds's avatar
Linus Torvalds committed
919
920
			break;
	}
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
	h->max_huge_pages = h->free_huge_pages = h->nr_huge_pages = i;
}

static void __init hugetlb_init_hstates(void)
{
	struct hstate *h;

	for_each_hstate(h) {
		hugetlb_init_one_hstate(h);
	}
}

static void __init report_hugepages(void)
{
	struct hstate *h;

	for_each_hstate(h) {
		printk(KERN_INFO "Total HugeTLB memory allocated, "
				"%ld %dMB pages\n",
				h->free_huge_pages,
				1 << (h->order + PAGE_SHIFT - 20));
	}
}

static int __init hugetlb_init(void)
{
	BUILD_BUG_ON(HPAGE_SHIFT == 0);

	if (!size_to_hstate(HPAGE_SIZE)) {
		hugetlb_add_hstate(HUGETLB_PAGE_ORDER);
		parsed_hstate->max_huge_pages = default_hstate_max_huge_pages;
	}
	default_hstate_idx = size_to_hstate(HPAGE_SIZE) - hstates;

	hugetlb_init_hstates();

	report_hugepages();

Linus Torvalds's avatar
Linus Torvalds committed
959
960
961
962
	return 0;
}
module_init(hugetlb_init);

963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
/* Should be called on processing a hugepagesz=... option */
void __init hugetlb_add_hstate(unsigned order)
{
	struct hstate *h;
	if (size_to_hstate(PAGE_SIZE << order)) {
		printk(KERN_WARNING "hugepagesz= specified twice, ignoring\n");
		return;
	}
	BUG_ON(max_hstate >= HUGE_MAX_HSTATE);
	BUG_ON(order == 0);
	h = &hstates[max_hstate++];
	h->order = order;
	h->mask = ~((1ULL << (order + PAGE_SHIFT)) - 1);
	hugetlb_init_one_hstate(h);
	parsed_hstate = h;
}

Linus Torvalds's avatar
Linus Torvalds committed
980
981
static int __init hugetlb_setup(char *s)
{
982
983
984
985
986
987
988
989
990
991
992
993
994
995
	unsigned long *mhp;

	/*
	 * !max_hstate means we haven't parsed a hugepagesz= parameter yet,
	 * so this hugepages= parameter goes to the "default hstate".
	 */
	if (!max_hstate)
		mhp = &default_hstate_max_huge_pages;
	else
		mhp = &parsed_hstate->max_huge_pages;

	if (sscanf(s, "%lu", mhp) <= 0)
		*mhp = 0;

Linus Torvalds's avatar
Linus Torvalds committed
996
997
998
999
	return 1;
}
__setup("hugepages=", hugetlb_setup);

1000
1001
1002
1003
1004
1005
1006
1007
1008
1009
1010
static unsigned int cpuset_mems_nr(unsigned int *array)
{
	int node;
	unsigned int nr = 0;

	for_each_node_mask(node, cpuset_current_mems_allowed)
		nr += array[node];

	return nr;
}

Linus Torvalds's avatar
Linus Torvalds committed
1011
1012
#ifdef CONFIG_SYSCTL
#ifdef CONFIG_HIGHMEM
1013
static void try_to_free_low(struct hstate *h, unsigned long count)
Linus Torvalds's avatar
Linus Torvalds committed
1014
{
1015
1016
	int i;

Linus Torvalds's avatar
Linus Torvalds committed
1017
1018
	for (i = 0; i < MAX_NUMNODES; ++i) {
		struct page *page, *next;
1019
1020
1021
		struct list_head *freel = &h->hugepage_freelists[i];
		list_for_each_entry_safe(page, next, freel, lru) {
			if (count >= h->nr_huge_pages)
1022
				return;
Linus Torvalds's avatar
Linus Torvalds committed
1023
1024
1025
			if (PageHighMem(page))
				continue;
			list_del(&page->lru);
1026
			update_and_free_page(h, page);
1027
1028
			h->free_huge_pages--;
			h->free_huge_pages_node[page_to_nid(page)]--;
Linus Torvalds's avatar
Linus Torvalds committed
1029
1030
1031
1032
		}
	}
}
#else
1033
static inline void try_to_free_low(struct hstate *h, unsigned long count)
Linus Torvalds's avatar
Linus Torvalds committed
1034
1035
1036
1037
{
}
#endif

1038
#define persistent_huge_pages(h) (h->nr_huge_pages - h->surplus_huge_pages)
1039
static unsigned long set_max_huge_pages(struct hstate *h, unsigned long count)
Linus Torvalds's avatar
Linus Torvalds committed
1040
{
1041
	unsigned long min_count, ret;
Linus Torvalds's avatar
Linus Torvalds committed
1042

1043
1044
1045
1046
	/*
	 * Increase the pool size
	 * First take pages out of surplus state.  Then make up the
	 * remaining difference by allocating fresh huge pages.
1047
1048
1049
1050
1051
1052
	 *
	 * We might race with alloc_buddy_huge_page() here and be unable
	 * to convert a surplus huge page to a normal huge page. That is
	 * not critical, though, it just means the overall size of the
	 * pool might be one hugepage larger than it needs to be, but
	 * within all the constraints specified by the sysctls.
1053
	 */
Linus Torvalds's avatar
Linus Torvalds committed
1054
	spin_lock(&hugetlb_lock);
1055
1056
	while (h->surplus_huge_pages && count > persistent_huge_pages(h)) {
		if (!adjust_pool_surplus(h, -1))
1057
1058
1059
			break;
	}

1060
	while (count > persistent_huge_pages(h)) {
1061
1062
1063
1064
1065
1066
		/*
		 * If this allocation races such that we no longer need the
		 * page, free_huge_page will handle it by freeing the page
		 * and reducing the surplus.
		 */
		spin_unlock(&hugetlb_lock);
1067
		ret = alloc_fresh_huge_page(h);
1068
1069
1070
1071
1072
1073
1074
1075
1076
1077
1078
1079
		spin_lock(&hugetlb_lock);
		if (!ret)
			goto out;

	}

	/*
	 * Decrease the pool size
	 * First return free pages to the buddy allocator (being careful
	 * to keep enough around to satisfy reservations).  Then place
	 * pages into surplus state as needed so the pool will shrink
	 * to the desired size as pages become free.
1080
1081
1082
1083
1084
1085
1086
1087
	 *
	 * By placing pages into the surplus state independent of the
	 * overcommit value, we are allowing the surplus pool size to
	 * exceed overcommit. There are few sane options here. Since
	 * alloc_buddy_huge_page() is checking the global counter,
	 * though, we'll note that we're not allowed to exceed surplus
	 * and won't grow the pool anywhere else. Not until one of the
	 * sysctls are changed, or the surplus pages go out of use.
1088
	 */
1089
	min_count = h->resv_huge_pages + h->nr_huge_pages - h->free_huge_pages;
1090
	min_count = max(count, min_count);
1091
1092
1093
	try_to_free_low(h, min_count);
	while (min_count < persistent_huge_pages(h)) {
		struct page *page = dequeue_huge_page(h);
Linus Torvalds's avatar
Linus Torvalds committed
1094
1095
		if (!page)
			break;
1096
		update_and_free_page(h, page);
Linus Torvalds's avatar
Linus Torvalds committed
1097
	}
1098
1099
	while (count < persistent_huge_pages(h)) {
		if (!adjust_pool_surplus(h, 1))
1100
1101
1102
			break;
	}
out:
1103
	ret = persistent_huge_pages(h);
Linus Torvalds's avatar
Linus Torvalds committed
1104
	spin_unlock(&hugetlb_lock);
1105
	return ret;
Linus Torvalds's avatar
Linus Torvalds committed
1106
1107
1108
1109
1110
1111
}

int hugetlb_sysctl_handler(struct ctl_table *table, int write,
			   struct file *file, void __user *buffer,
			   size_t *length, loff_t *ppos)
{
1112
1113
1114
1115
1116
1117
1118
1119
	struct hstate *h = &default_hstate;
	unsigned long tmp;

	if (!write)
		tmp = h->max_huge_pages;

	table->data = &tmp;
	table->maxlen = sizeof(unsigned long);
Linus Torvalds's avatar
Linus Torvalds committed
1120
	proc_doulongvec_minmax(table, write, file, buffer, length, ppos);
1121
1122
1123
1124

	if (write)
		h->max_huge_pages = set_max_huge_pages(h, tmp);

Linus Torvalds's avatar
Linus Torvalds committed
1125
1126
	return 0;
}
1127
1128
1129
1130
1131
1132
1133
1134
1135
1136
1137
1138
1139

int hugetlb_treat_movable_handler(struct ctl_table *table, int write,
			struct file *file, void __user *buffer,
			size_t *length, loff_t *ppos)
{
	proc_dointvec(table, write, file, buffer, length, ppos);
	if (hugepages_treat_as_movable)
		htlb_alloc_mask = GFP_HIGHUSER_MOVABLE;
	else
		htlb_alloc_mask = GFP_HIGHUSER;
	return 0;
}

1140
1141
1142
1143
int hugetlb_overcommit_handler(struct ctl_table *table, int write,
			struct file *file, void __user *buffer,
			size_t *length, loff_t *ppos)
{
1144
	struct hstate *h = &default_hstate;
1145
1146
1147
1148
1149
1150
1151
	unsigned long tmp;

	if (!write)
		tmp = h->nr_overcommit_huge_pages;

	table->data = &tmp;
	table->maxlen = sizeof(unsigned long);
1152
	proc_doulongvec_minmax(table, write, file, buffer, length, ppos);
1153
1154
1155
1156
1157
1158
1159

	if (write) {
		spin_lock(&hugetlb_lock);
		h->nr_overcommit_huge_pages = tmp;
		spin_unlock(&hugetlb_lock);
	}

1160
1161
1162
	return 0;
}

Linus Torvalds's avatar
Linus Torvalds committed
1163
1164
1165
1166
#endif /* CONFIG_SYSCTL */

int hugetlb_report_meminfo(char *buf)
{
1167
	struct hstate *h = &default_hstate;
Linus Torvalds's avatar
Linus Torvalds committed
1168
1169
1170
	return sprintf(buf,
			"HugePages_Total: %5lu\n"
			"HugePages_Free:  %5lu\n"
1171
			"HugePages_Rsvd:  %5lu\n"
1172
			"HugePages_Surp:  %5lu\n"
Linus Torvalds's avatar
Linus Torvalds committed
1173
			"Hugepagesize:    %5lu kB\n",
1174
1175
1176
1177
1178
			h->nr_huge_pages,
			h->free_huge_pages,
			h->resv_huge_pages,
			h->surplus_huge_pages,
			1UL << (huge_page_order(h) + PAGE_SHIFT - 10));
Linus Torvalds's avatar
Linus Torvalds committed
1179
1180
1181
1182
}

int hugetlb_report_node_meminfo(int nid, char *buf)
{
1183
	struct hstate *h = &default_hstate;
Linus Torvalds's avatar
Linus Torvalds committed
1184
1185
	return sprintf(buf,
		"Node %d HugePages_Total: %5u\n"
1186
1187
		"Node %d HugePages_Free:  %5u\n"
		"Node %d HugePages_Surp:  %5u\n",
1188
1189
1190
		nid, h->nr_huge_pages_node[nid],
		nid, h->free_huge_pages_node[nid],
		nid, h->surplus_huge_pages_node[nid]);
Linus Torvalds's avatar
Linus Torvalds committed
1191
1192
1193
1194
1195
}

/* Return the number pages of memory we physically have, in PAGE_SIZE units. */
unsigned long hugetlb_total_pages(void)
{
1196
1197
	struct hstate *h = &default_hstate;
	return h->nr_huge_pages * pages_per_huge_page(h);
Linus Torvalds's avatar
Linus Torvalds committed
1198
1199
}

1200
static int hugetlb_acct_memory(struct hstate *h, long delta)
1201
1202
1203
1204
1205
1206
1207
1208
1209
1210
1211
1212
1213
1214
1215
1216
1217
1218
1219
1220
1221
1222
{
	int ret = -ENOMEM;

	spin_lock(&hugetlb_lock);
	/*
	 * When cpuset is configured, it breaks the strict hugetlb page
	 * reservation as the accounting is done on a global variable. Such
	 * reservation is completely rubbish in the presence of cpuset because
	 * the reservation is not checked against page availability for the
	 * current cpuset. Application can still potentially OOM'ed by kernel
	 * with lack of free htlb page in cpuset that the task is in.
	 * Attempt to enforce strict accounting with cpuset is almost
	 * impossible (or too ugly) because cpuset is too fluid that
	 * task or memory node can be dynamically moved between cpusets.
	 *
	 * The change of semantics for shared hugetlb mapping with cpuset is
	 * undesirable. However, in order to preserve some of the semantics,
	 * we fall back to check against current free page availability as
	 * a best attempt and hopefully to minimize the impact of changing
	 * semantics that cpuset has.
	 */
	if (delta > 0) {
1223
		if (gather_surplus_pages(h, delta) < 0)
1224
1225
			goto out;

1226
1227
		if (delta > cpuset_mems_nr(h->free_huge_pages_node)) {
			return_unused_surplus_pages(h, delta);
1228
1229
1230
1231
1232
1233
			goto out;
		}
	}

	ret = 0;
	if (delta < 0)
1234
		return_unused_surplus_pages(h, (unsigned long) -delta);
1235
1236
1237
1238
1239
1240

out:
	spin_unlock(&hugetlb_lock);
	return ret;
}

1241
1242
1243
1244
1245
1246
1247
1248
1249
1250
1251
1252
1253
1254
1255
1256
static void hugetlb_vm_op_open(struct vm_area_struct *vma)
{
	struct resv_map *reservations = vma_resv_map(vma);

	/*
	 * This new VMA should share its siblings reservation map if present.
	 * The VMA will only ever have a valid reservation map pointer where
	 * it is being copied for another still existing VMA.  As that VMA
	 * has a reference to the reservation map it cannot dissappear until
	 * after this open call completes.  It is therefore safe to take a
	 * new reference here without additional locking.
	 */
	if (reservations)
		kref_get(&reservations->refs);
}

1257
1258
static void hugetlb_vm_op_close(struct vm_area_struct *vma)
{
1259
	struct hstate *h = hstate_vma(vma);
1260
1261
1262
1263
1264
1265
	struct resv_map *reservations = vma_resv_map(vma);
	unsigned long reserve;
	unsigned long start;
	unsigned long end;

	if (reservations) {
1266
1267
		start = vma_hugecache_offset(h, vma, vma->vm_start);
		end = vma_hugecache_offset(h, vma, vma->vm_end);
1268
1269
1270
1271
1272
1273
1274

		reserve = (end - start) -
			region_count(&reservations->regions, start, end);

		kref_put(&reservations->refs, resv_map_release);

		if (reserve)
1275
			hugetlb_acct_memory(h, -reserve);
1276
	}
1277
1278
}

Linus Torvalds's avatar
Linus Torvalds committed
1279
1280
1281
1282
1283
1284
/*
 * We cannot handle pagefaults against hugetlb pages at all.  They cause
 * handle_mm_fault() to try to instantiate regular-sized pages in the
 * hugegpage VMA.  do_page_fault() is supposed to trap this, so BUG is we get
 * this far.
 */
Nick Piggin's avatar
Nick Piggin committed
1285
static int hugetlb_vm_op_fault(struct vm_area_struct *vma, struct vm_fault *vmf)
Linus Torvalds's avatar
Linus Torvalds committed
1286
1287
{
	BUG();
Nick Piggin's avatar
Nick Piggin committed
1288
	return 0;
Linus Torvalds's avatar
Linus Torvalds committed
1289
1290
1291
}

struct vm_operations_struct hugetlb_vm_ops = {
Nick Piggin's avatar
Nick Piggin committed
1292
	.fault = hugetlb_vm_op_fault,
1293
	.open = hugetlb_vm_op_open,
1294
	.close = hugetlb_vm_op_close,
Linus Torvalds's avatar
Linus Torvalds committed
1295
1296
};

1297
1298
static pte_t make_huge_pte(struct vm_area_struct *vma, struct page *page,
				int writable)
David Gibson's avatar
David Gibson committed
1299
1300
1301
{
	pte_t entry;