mmap.c 80.6 KB
Newer Older
L
Linus Torvalds 已提交
1 2 3 4 5
/*
 * mm/mmap.c
 *
 * Written by obz.
 *
A
Alan Cox 已提交
6
 * Address space accounting code	<alan@lxorguk.ukuu.org.uk>
L
Linus Torvalds 已提交
7 8 9
 */

#include <linux/slab.h>
A
Alexey Dobriyan 已提交
10
#include <linux/backing-dev.h>
L
Linus Torvalds 已提交
11 12 13 14 15 16
#include <linux/mm.h>
#include <linux/shm.h>
#include <linux/mman.h>
#include <linux/pagemap.h>
#include <linux/swap.h>
#include <linux/syscalls.h>
17
#include <linux/capability.h>
L
Linus Torvalds 已提交
18 19 20 21 22 23 24
#include <linux/init.h>
#include <linux/file.h>
#include <linux/fs.h>
#include <linux/personality.h>
#include <linux/security.h>
#include <linux/hugetlb.h>
#include <linux/profile.h>
25
#include <linux/export.h>
L
Linus Torvalds 已提交
26 27 28
#include <linux/mount.h>
#include <linux/mempolicy.h>
#include <linux/rmap.h>
A
Andrea Arcangeli 已提交
29
#include <linux/mmu_notifier.h>
30
#include <linux/perf_event.h>
A
Al Viro 已提交
31
#include <linux/audit.h>
A
Andrea Arcangeli 已提交
32
#include <linux/khugepaged.h>
33
#include <linux/uprobes.h>
34
#include <linux/rbtree_augmented.h>
L
Linus Torvalds 已提交
35 36 37 38

#include <asm/uaccess.h>
#include <asm/cacheflush.h>
#include <asm/tlb.h>
39
#include <asm/mmu_context.h>
L
Linus Torvalds 已提交
40

41 42
#include "internal.h"

43 44 45 46
#ifndef arch_mmap_check
#define arch_mmap_check(addr, len, flags)	(0)
#endif

47 48 49 50
#ifndef arch_rebalance_pgtables
#define arch_rebalance_pgtables(addr, len)		(addr)
#endif

51 52 53 54
static void unmap_region(struct mm_struct *mm,
		struct vm_area_struct *vma, struct vm_area_struct *prev,
		unsigned long start, unsigned long end);

L
Linus Torvalds 已提交
55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74
/* description of effects of mapping type and prot in current implementation.
 * this is due to the limited x86 page protection hardware.  The expected
 * behavior is in parens:
 *
 * map_type	prot
 *		PROT_NONE	PROT_READ	PROT_WRITE	PROT_EXEC
 * MAP_SHARED	r: (no) no	r: (yes) yes	r: (no) yes	r: (no) yes
 *		w: (no) no	w: (no) no	w: (yes) yes	w: (no) no
 *		x: (no) no	x: (no) yes	x: (no) yes	x: (yes) yes
 *		
 * MAP_PRIVATE	r: (no) no	r: (yes) yes	r: (no) yes	r: (no) yes
 *		w: (no) no	w: (no) no	w: (copy) copy	w: (no) no
 *		x: (no) no	x: (no) yes	x: (no) yes	x: (yes) yes
 *
 */
pgprot_t protection_map[16] = {
	__P000, __P001, __P010, __P011, __P100, __P101, __P110, __P111,
	__S000, __S001, __S010, __S011, __S100, __S101, __S110, __S111
};

75 76
pgprot_t vm_get_page_prot(unsigned long vm_flags)
{
77 78 79
	return __pgprot(pgprot_val(protection_map[vm_flags &
				(VM_READ|VM_WRITE|VM_EXEC|VM_SHARED)]) |
			pgprot_val(arch_vm_get_page_prot(vm_flags)));
80 81 82
}
EXPORT_SYMBOL(vm_get_page_prot);

83 84
int sysctl_overcommit_memory __read_mostly = OVERCOMMIT_GUESS;  /* heuristic overcommit */
int sysctl_overcommit_ratio __read_mostly = 50;	/* default is 50% */
85
int sysctl_max_map_count __read_mostly = DEFAULT_MAX_MAP_COUNT;
86 87 88 89 90
/*
 * Make sure vm_committed_as in one cacheline and not cacheline shared with
 * other variables. It can be updated by several CPUs frequently.
 */
struct percpu_counter vm_committed_as ____cacheline_aligned_in_smp;
L
Linus Torvalds 已提交
91

92 93 94 95 96 97 98 99 100 101 102 103 104 105
/*
 * The global memory commitment made in the system can be a metric
 * that can be used to drive ballooning decisions when Linux is hosted
 * as a guest. On Hyper-V, the host implements a policy engine for dynamically
 * balancing memory across competing virtual machines that are hosted.
 * Several metrics drive this policy engine including the guest reported
 * memory commitment.
 */
unsigned long vm_memory_committed(void)
{
	return percpu_counter_read_positive(&vm_committed_as);
}
EXPORT_SYMBOL_GPL(vm_memory_committed);

L
Linus Torvalds 已提交
106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121
/*
 * Check that a process has enough memory to allocate a new virtual
 * mapping. 0 means there is enough memory for the allocation to
 * succeed and -ENOMEM implies there is not.
 *
 * We currently support three overcommit policies, which are set via the
 * vm.overcommit_memory sysctl.  See Documentation/vm/overcommit-accounting
 *
 * Strict overcommit modes added 2002 Feb 26 by Alan Cox.
 * Additional code 2002 Jul 20 by Robert Love.
 *
 * cap_sys_admin is 1 if the process has admin privileges, 0 otherwise.
 *
 * Note this is a helper function intended to be used by LSMs which
 * wish to use this logic.
 */
122
int __vm_enough_memory(struct mm_struct *mm, long pages, int cap_sys_admin)
L
Linus Torvalds 已提交
123 124 125 126 127 128 129 130 131 132 133 134
{
	unsigned long free, allowed;

	vm_acct_memory(pages);

	/*
	 * Sometimes we want to use more memory than we have
	 */
	if (sysctl_overcommit_memory == OVERCOMMIT_ALWAYS)
		return 0;

	if (sysctl_overcommit_memory == OVERCOMMIT_GUESS) {
135 136 137 138 139 140 141 142 143 144
		free = global_page_state(NR_FREE_PAGES);
		free += global_page_state(NR_FILE_PAGES);

		/*
		 * shmem pages shouldn't be counted as free in this
		 * case, they can't be purged, only swapped out, and
		 * that won't affect the overall amount of available
		 * memory in the system.
		 */
		free -= global_page_state(NR_SHMEM);
L
Linus Torvalds 已提交
145 146 147 148 149 150 151 152 153

		free += nr_swap_pages;

		/*
		 * Any slabs which are created with the
		 * SLAB_RECLAIM_ACCOUNT flag claim to have contents
		 * which are reclaimable, under pressure.  The dentry
		 * cache and most inode caches should fall into this
		 */
154
		free += global_page_state(NR_SLAB_RECLAIMABLE);
L
Linus Torvalds 已提交
155

156 157 158
		/*
		 * Leave reserved pages. The pages are not for anonymous pages.
		 */
159
		if (free <= totalreserve_pages)
160 161
			goto error;
		else
162
			free -= totalreserve_pages;
163 164 165 166

		/*
		 * Leave the last 3% for root
		 */
L
Linus Torvalds 已提交
167
		if (!cap_sys_admin)
168
			free -= free / 32;
L
Linus Torvalds 已提交
169 170 171

		if (free > pages)
			return 0;
172 173

		goto error;
L
Linus Torvalds 已提交
174 175 176 177 178 179 180 181 182 183 184 185 186
	}

	allowed = (totalram_pages - hugetlb_total_pages())
	       	* sysctl_overcommit_ratio / 100;
	/*
	 * Leave the last 3% for root
	 */
	if (!cap_sys_admin)
		allowed -= allowed / 32;
	allowed += total_swap_pages;

	/* Don't let a single process grow too big:
	   leave 3% of the size of this process for other processes */
A
Alan Cox 已提交
187 188
	if (mm)
		allowed -= mm->total_vm / 32;
L
Linus Torvalds 已提交
189

190
	if (percpu_counter_read_positive(&vm_committed_as) < allowed)
L
Linus Torvalds 已提交
191
		return 0;
192
error:
L
Linus Torvalds 已提交
193 194 195 196 197 198
	vm_unacct_memory(pages);

	return -ENOMEM;
}

/*
199
 * Requires inode->i_mapping->i_mmap_mutex
L
Linus Torvalds 已提交
200 201 202 203 204
 */
static void __remove_shared_vm_struct(struct vm_area_struct *vma,
		struct file *file, struct address_space *mapping)
{
	if (vma->vm_flags & VM_DENYWRITE)
205
		atomic_inc(&file->f_path.dentry->d_inode->i_writecount);
L
Linus Torvalds 已提交
206 207 208 209 210
	if (vma->vm_flags & VM_SHARED)
		mapping->i_mmap_writable--;

	flush_dcache_mmap_lock(mapping);
	if (unlikely(vma->vm_flags & VM_NONLINEAR))
211
		list_del_init(&vma->shared.nonlinear);
L
Linus Torvalds 已提交
212
	else
213
		vma_interval_tree_remove(vma, &mapping->i_mmap);
L
Linus Torvalds 已提交
214 215 216 217
	flush_dcache_mmap_unlock(mapping);
}

/*
218
 * Unlink a file-based vm structure from its interval tree, to hide
219
 * vma from rmap and vmtruncate before freeing its page tables.
L
Linus Torvalds 已提交
220
 */
221
void unlink_file_vma(struct vm_area_struct *vma)
L
Linus Torvalds 已提交
222 223 224 225 226
{
	struct file *file = vma->vm_file;

	if (file) {
		struct address_space *mapping = file->f_mapping;
227
		mutex_lock(&mapping->i_mmap_mutex);
L
Linus Torvalds 已提交
228
		__remove_shared_vm_struct(vma, file, mapping);
229
		mutex_unlock(&mapping->i_mmap_mutex);
L
Linus Torvalds 已提交
230
	}
231 232 233 234 235 236 237 238 239 240
}

/*
 * Close a vm structure and free it, returning the next.
 */
static struct vm_area_struct *remove_vma(struct vm_area_struct *vma)
{
	struct vm_area_struct *next = vma->vm_next;

	might_sleep();
L
Linus Torvalds 已提交
241 242
	if (vma->vm_ops && vma->vm_ops->close)
		vma->vm_ops->close(vma);
243
	if (vma->vm_file)
244
		fput(vma->vm_file);
245
	mpol_put(vma_policy(vma));
L
Linus Torvalds 已提交
246
	kmem_cache_free(vm_area_cachep, vma);
247
	return next;
L
Linus Torvalds 已提交
248 249
}

250 251
static unsigned long do_brk(unsigned long addr, unsigned long len);

252
SYSCALL_DEFINE1(brk, unsigned long, brk)
L
Linus Torvalds 已提交
253 254 255 256
{
	unsigned long rlim, retval;
	unsigned long newbrk, oldbrk;
	struct mm_struct *mm = current->mm;
257
	unsigned long min_brk;
L
Linus Torvalds 已提交
258 259 260

	down_write(&mm->mmap_sem);

261
#ifdef CONFIG_COMPAT_BRK
262 263 264 265 266
	/*
	 * CONFIG_COMPAT_BRK can still be overridden by setting
	 * randomize_va_space to 2, which will still cause mm->start_brk
	 * to be arbitrarily shifted
	 */
267
	if (current->brk_randomized)
268 269 270
		min_brk = mm->start_brk;
	else
		min_brk = mm->end_data;
271 272 273 274
#else
	min_brk = mm->start_brk;
#endif
	if (brk < min_brk)
L
Linus Torvalds 已提交
275
		goto out;
R
Ram Gupta 已提交
276 277 278 279 280 281 282

	/*
	 * Check against rlimit here. If this check is done later after the test
	 * of oldbrk with newbrk then it can escape the test and let the data
	 * segment grow beyond its set limit the in case where the limit is
	 * not page aligned -Ram Gupta
	 */
J
Jiri Slaby 已提交
283
	rlim = rlimit(RLIMIT_DATA);
J
Jiri Kosina 已提交
284 285
	if (rlim < RLIM_INFINITY && (brk - mm->start_brk) +
			(mm->end_data - mm->start_data) > rlim)
R
Ram Gupta 已提交
286 287
		goto out;

L
Linus Torvalds 已提交
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
	newbrk = PAGE_ALIGN(brk);
	oldbrk = PAGE_ALIGN(mm->brk);
	if (oldbrk == newbrk)
		goto set_brk;

	/* Always allow shrinking brk. */
	if (brk <= mm->brk) {
		if (!do_munmap(mm, newbrk, oldbrk-newbrk))
			goto set_brk;
		goto out;
	}

	/* Check against existing mmap mappings. */
	if (find_vma_intersection(mm, oldbrk, newbrk+PAGE_SIZE))
		goto out;

	/* Ok, looks good - let it rip. */
	if (do_brk(oldbrk, newbrk-oldbrk) != oldbrk)
		goto out;
set_brk:
	mm->brk = brk;
out:
	retval = mm->brk;
	up_write(&mm->mmap_sem);
	return retval;
}

315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335
static long vma_compute_subtree_gap(struct vm_area_struct *vma)
{
	unsigned long max, subtree_gap;
	max = vma->vm_start;
	if (vma->vm_prev)
		max -= vma->vm_prev->vm_end;
	if (vma->vm_rb.rb_left) {
		subtree_gap = rb_entry(vma->vm_rb.rb_left,
				struct vm_area_struct, vm_rb)->rb_subtree_gap;
		if (subtree_gap > max)
			max = subtree_gap;
	}
	if (vma->vm_rb.rb_right) {
		subtree_gap = rb_entry(vma->vm_rb.rb_right,
				struct vm_area_struct, vm_rb)->rb_subtree_gap;
		if (subtree_gap > max)
			max = subtree_gap;
	}
	return max;
}

336
#ifdef CONFIG_DEBUG_VM_RB
L
Linus Torvalds 已提交
337 338
static int browse_rb(struct rb_root *root)
{
339
	int i = 0, j, bug = 0;
L
Linus Torvalds 已提交
340 341 342 343 344 345
	struct rb_node *nd, *pn = NULL;
	unsigned long prev = 0, pend = 0;

	for (nd = rb_first(root); nd; nd = rb_next(nd)) {
		struct vm_area_struct *vma;
		vma = rb_entry(nd, struct vm_area_struct, vm_rb);
346 347 348 349 350
		if (vma->vm_start < prev) {
			printk("vm_start %lx prev %lx\n", vma->vm_start, prev);
			bug = 1;
		}
		if (vma->vm_start < pend) {
L
Linus Torvalds 已提交
351
			printk("vm_start %lx pend %lx\n", vma->vm_start, pend);
352 353 354 355 356 357 358 359 360 361 362 363 364
			bug = 1;
		}
		if (vma->vm_start > vma->vm_end) {
			printk("vm_end %lx < vm_start %lx\n",
				vma->vm_end, vma->vm_start);
			bug = 1;
		}
		if (vma->rb_subtree_gap != vma_compute_subtree_gap(vma)) {
			printk("free gap %lx, correct %lx\n",
			       vma->rb_subtree_gap,
			       vma_compute_subtree_gap(vma));
			bug = 1;
		}
L
Linus Torvalds 已提交
365 366
		i++;
		pn = nd;
D
David Miller 已提交
367 368
		prev = vma->vm_start;
		pend = vma->vm_end;
L
Linus Torvalds 已提交
369 370
	}
	j = 0;
371
	for (nd = pn; nd; nd = rb_prev(nd))
L
Linus Torvalds 已提交
372
		j++;
373 374 375
	if (i != j) {
		printk("backwards %d, forwards %d\n", j, i);
		bug = 1;
L
Linus Torvalds 已提交
376
	}
377
	return bug ? -1 : i;
L
Linus Torvalds 已提交
378 379
}

380 381 382 383 384 385 386 387 388
static void validate_mm_rb(struct rb_root *root, struct vm_area_struct *ignore)
{
	struct rb_node *nd;

	for (nd = rb_first(root); nd; nd = rb_next(nd)) {
		struct vm_area_struct *vma;
		vma = rb_entry(nd, struct vm_area_struct, vm_rb);
		BUG_ON(vma != ignore &&
		       vma->rb_subtree_gap != vma_compute_subtree_gap(vma));
L
Linus Torvalds 已提交
389 390 391 392 393 394 395
	}
}

void validate_mm(struct mm_struct *mm)
{
	int bug = 0;
	int i = 0;
396
	unsigned long highest_address = 0;
397 398 399
	struct vm_area_struct *vma = mm->mmap;
	while (vma) {
		struct anon_vma_chain *avc;
400
		vma_lock_anon_vma(vma);
401 402
		list_for_each_entry(avc, &vma->anon_vma_chain, same_vma)
			anon_vma_interval_tree_verify(avc);
403
		vma_unlock_anon_vma(vma);
404
		highest_address = vma->vm_end;
405
		vma = vma->vm_next;
L
Linus Torvalds 已提交
406 407
		i++;
	}
408 409 410 411 412 413 414 415 416
	if (i != mm->map_count) {
		printk("map_count %d vm_next %d\n", mm->map_count, i);
		bug = 1;
	}
	if (highest_address != mm->highest_vm_end) {
		printk("mm->highest_vm_end %lx, found %lx\n",
		       mm->highest_vm_end, highest_address);
		bug = 1;
	}
L
Linus Torvalds 已提交
417
	i = browse_rb(&mm->mm_rb);
418 419 420 421
	if (i != mm->map_count) {
		printk("map_count %d rb %d\n", mm->map_count, i);
		bug = 1;
	}
422
	BUG_ON(bug);
L
Linus Torvalds 已提交
423 424
}
#else
425
#define validate_mm_rb(root, ignore) do { } while (0)
L
Linus Torvalds 已提交
426 427 428
#define validate_mm(mm) do { } while (0)
#endif

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
RB_DECLARE_CALLBACKS(static, vma_gap_callbacks, struct vm_area_struct, vm_rb,
		     unsigned long, rb_subtree_gap, vma_compute_subtree_gap)

/*
 * Update augmented rbtree rb_subtree_gap values after vma->vm_start or
 * vma->vm_prev->vm_end values changed, without modifying the vma's position
 * in the rbtree.
 */
static void vma_gap_update(struct vm_area_struct *vma)
{
	/*
	 * As it turns out, RB_DECLARE_CALLBACKS() already created a callback
	 * function that does exacltly what we want.
	 */
	vma_gap_callbacks_propagate(&vma->vm_rb, NULL);
}

static inline void vma_rb_insert(struct vm_area_struct *vma,
				 struct rb_root *root)
{
	/* All rb_subtree_gap values must be consistent prior to insertion */
	validate_mm_rb(root, NULL);

	rb_insert_augmented(&vma->vm_rb, root, &vma_gap_callbacks);
}

static void vma_rb_erase(struct vm_area_struct *vma, struct rb_root *root)
{
	/*
	 * All rb_subtree_gap values must be consistent prior to erase,
	 * with the possible exception of the vma being erased.
	 */
	validate_mm_rb(root, vma);

	/*
	 * Note rb_erase_augmented is a fairly large inline function,
	 * so make sure we instantiate it only once with our desired
	 * augmented rbtree callbacks.
	 */
	rb_erase_augmented(&vma->vm_rb, root, &vma_gap_callbacks);
}

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
/*
 * vma has some anon_vma assigned, and is already inserted on that
 * anon_vma's interval trees.
 *
 * Before updating the vma's vm_start / vm_end / vm_pgoff fields, the
 * vma must be removed from the anon_vma's interval trees using
 * anon_vma_interval_tree_pre_update_vma().
 *
 * After the update, the vma will be reinserted using
 * anon_vma_interval_tree_post_update_vma().
 *
 * The entire update must be protected by exclusive mmap_sem and by
 * the root anon_vma's mutex.
 */
static inline void
anon_vma_interval_tree_pre_update_vma(struct vm_area_struct *vma)
{
	struct anon_vma_chain *avc;

	list_for_each_entry(avc, &vma->anon_vma_chain, same_vma)
		anon_vma_interval_tree_remove(avc, &avc->anon_vma->rb_root);
}

static inline void
anon_vma_interval_tree_post_update_vma(struct vm_area_struct *vma)
{
	struct anon_vma_chain *avc;

	list_for_each_entry(avc, &vma->anon_vma_chain, same_vma)
		anon_vma_interval_tree_insert(avc, &avc->anon_vma->rb_root);
}

503 504 505
static int find_vma_links(struct mm_struct *mm, unsigned long addr,
		unsigned long end, struct vm_area_struct **pprev,
		struct rb_node ***rb_link, struct rb_node **rb_parent)
L
Linus Torvalds 已提交
506
{
507
	struct rb_node **__rb_link, *__rb_parent, *rb_prev;
L
Linus Torvalds 已提交
508 509 510 511 512 513 514 515 516 517 518

	__rb_link = &mm->mm_rb.rb_node;
	rb_prev = __rb_parent = NULL;

	while (*__rb_link) {
		struct vm_area_struct *vma_tmp;

		__rb_parent = *__rb_link;
		vma_tmp = rb_entry(__rb_parent, struct vm_area_struct, vm_rb);

		if (vma_tmp->vm_end > addr) {
519 520 521
			/* Fail if an existing vma overlaps the area */
			if (vma_tmp->vm_start < end)
				return -ENOMEM;
L
Linus Torvalds 已提交
522 523 524 525 526 527 528 529 530 531 532 533
			__rb_link = &__rb_parent->rb_left;
		} else {
			rb_prev = __rb_parent;
			__rb_link = &__rb_parent->rb_right;
		}
	}

	*pprev = NULL;
	if (rb_prev)
		*pprev = rb_entry(rb_prev, struct vm_area_struct, vm_rb);
	*rb_link = __rb_link;
	*rb_parent = __rb_parent;
534
	return 0;
L
Linus Torvalds 已提交
535 536 537 538 539
}

void __vma_link_rb(struct mm_struct *mm, struct vm_area_struct *vma,
		struct rb_node **rb_link, struct rb_node *rb_parent)
{
540 541 542 543 544 545 546 547 548 549 550 551 552 553 554
	/* Update tracking information for the gap following the new vma. */
	if (vma->vm_next)
		vma_gap_update(vma->vm_next);
	else
		mm->highest_vm_end = vma->vm_end;

	/*
	 * vma->vm_prev wasn't known when we followed the rbtree to find the
	 * correct insertion point for that vma. As a result, we could not
	 * update the vma vm_rb parents rb_subtree_gap values on the way down.
	 * So, we first insert the vma with a zero rb_subtree_gap value
	 * (to be consistent with what we did on the way down), and then
	 * immediately update the gap to the correct value. Finally we
	 * rebalance the rbtree after all augmented values have been set.
	 */
L
Linus Torvalds 已提交
555
	rb_link_node(&vma->vm_rb, rb_parent, rb_link);
556 557 558
	vma->rb_subtree_gap = 0;
	vma_gap_update(vma);
	vma_rb_insert(vma, &mm->mm_rb);
L
Linus Torvalds 已提交
559 560
}

D
Denys Vlasenko 已提交
561
static void __vma_link_file(struct vm_area_struct *vma)
L
Linus Torvalds 已提交
562
{
Z
ZhenwenXu 已提交
563
	struct file *file;
L
Linus Torvalds 已提交
564 565 566 567 568 569

	file = vma->vm_file;
	if (file) {
		struct address_space *mapping = file->f_mapping;

		if (vma->vm_flags & VM_DENYWRITE)
570
			atomic_dec(&file->f_path.dentry->d_inode->i_writecount);
L
Linus Torvalds 已提交
571 572 573 574 575 576 577
		if (vma->vm_flags & VM_SHARED)
			mapping->i_mmap_writable++;

		flush_dcache_mmap_lock(mapping);
		if (unlikely(vma->vm_flags & VM_NONLINEAR))
			vma_nonlinear_insert(vma, &mapping->i_mmap_nonlinear);
		else
578
			vma_interval_tree_insert(vma, &mapping->i_mmap);
L
Linus Torvalds 已提交
579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600
		flush_dcache_mmap_unlock(mapping);
	}
}

static void
__vma_link(struct mm_struct *mm, struct vm_area_struct *vma,
	struct vm_area_struct *prev, struct rb_node **rb_link,
	struct rb_node *rb_parent)
{
	__vma_link_list(mm, vma, prev, rb_parent);
	__vma_link_rb(mm, vma, rb_link, rb_parent);
}

static void vma_link(struct mm_struct *mm, struct vm_area_struct *vma,
			struct vm_area_struct *prev, struct rb_node **rb_link,
			struct rb_node *rb_parent)
{
	struct address_space *mapping = NULL;

	if (vma->vm_file)
		mapping = vma->vm_file->f_mapping;

601
	if (mapping)
602
		mutex_lock(&mapping->i_mmap_mutex);
L
Linus Torvalds 已提交
603 604 605 606 607

	__vma_link(mm, vma, prev, rb_link, rb_parent);
	__vma_link_file(vma);

	if (mapping)
608
		mutex_unlock(&mapping->i_mmap_mutex);
L
Linus Torvalds 已提交
609 610 611 612 613 614

	mm->map_count++;
	validate_mm(mm);
}

/*
615
 * Helper for vma_adjust() in the split_vma insert case: insert a vma into the
616
 * mm's list and rbtree.  It has already been inserted into the interval tree.
L
Linus Torvalds 已提交
617
 */
Z
ZhenwenXu 已提交
618
static void __insert_vm_struct(struct mm_struct *mm, struct vm_area_struct *vma)
L
Linus Torvalds 已提交
619
{
620
	struct vm_area_struct *prev;
Z
ZhenwenXu 已提交
621
	struct rb_node **rb_link, *rb_parent;
L
Linus Torvalds 已提交
622

623 624 625
	if (find_vma_links(mm, vma->vm_start, vma->vm_end,
			   &prev, &rb_link, &rb_parent))
		BUG();
L
Linus Torvalds 已提交
626 627 628 629 630 631 632 633
	__vma_link(mm, vma, prev, rb_link, rb_parent);
	mm->map_count++;
}

static inline void
__vma_unlink(struct mm_struct *mm, struct vm_area_struct *vma,
		struct vm_area_struct *prev)
{
634
	struct vm_area_struct *next;
635

636 637
	vma_rb_erase(vma, &mm->mm_rb);
	prev->vm_next = next = vma->vm_next;
638 639
	if (next)
		next->vm_prev = prev;
L
Linus Torvalds 已提交
640 641 642 643 644 645 646 647 648 649 650
	if (mm->mmap_cache == vma)
		mm->mmap_cache = prev;
}

/*
 * We cannot adjust vm_start, vm_end, vm_pgoff fields of a vma that
 * is already present in an i_mmap tree without adjusting the tree.
 * The following helper function should be used when such adjustments
 * are necessary.  The "insert" vma (if any) is to be inserted
 * before we drop the necessary locks.
 */
651
int vma_adjust(struct vm_area_struct *vma, unsigned long start,
L
Linus Torvalds 已提交
652 653 654 655 656 657
	unsigned long end, pgoff_t pgoff, struct vm_area_struct *insert)
{
	struct mm_struct *mm = vma->vm_mm;
	struct vm_area_struct *next = vma->vm_next;
	struct vm_area_struct *importer = NULL;
	struct address_space *mapping = NULL;
658
	struct rb_root *root = NULL;
659
	struct anon_vma *anon_vma = NULL;
L
Linus Torvalds 已提交
660
	struct file *file = vma->vm_file;
661
	bool start_changed = false, end_changed = false;
L
Linus Torvalds 已提交
662 663 664 665
	long adjust_next = 0;
	int remove_next = 0;

	if (next && !insert) {
666 667
		struct vm_area_struct *exporter = NULL;

L
Linus Torvalds 已提交
668 669 670 671 672 673 674
		if (end >= next->vm_end) {
			/*
			 * vma expands, overlapping all the next, and
			 * perhaps the one after too (mprotect case 6).
			 */
again:			remove_next = 1 + (end > next->vm_end);
			end = next->vm_end;
675
			exporter = next;
L
Linus Torvalds 已提交
676 677 678 679 680 681 682
			importer = vma;
		} else if (end > next->vm_start) {
			/*
			 * vma expands, overlapping part of the next:
			 * mprotect case 5 shifting the boundary up.
			 */
			adjust_next = (end - next->vm_start) >> PAGE_SHIFT;
683
			exporter = next;
L
Linus Torvalds 已提交
684 685 686 687 688 689 690 691
			importer = vma;
		} else if (end < vma->vm_end) {
			/*
			 * vma shrinks, and !insert tells it's not
			 * split_vma inserting another: so it must be
			 * mprotect case 4 shifting the boundary down.
			 */
			adjust_next = - ((vma->vm_end - end) >> PAGE_SHIFT);
692
			exporter = vma;
L
Linus Torvalds 已提交
693 694 695
			importer = next;
		}

696 697 698 699 700
		/*
		 * Easily overlooked: when mprotect shifts the boundary,
		 * make sure the expanding vma has anon_vma set if the
		 * shrinking vma had, to cover any anon pages imported.
		 */
701 702
		if (exporter && exporter->anon_vma && !importer->anon_vma) {
			if (anon_vma_clone(importer, exporter))
703
				return -ENOMEM;
704
			importer->anon_vma = exporter->anon_vma;
705 706 707
		}
	}

L
Linus Torvalds 已提交
708 709
	if (file) {
		mapping = file->f_mapping;
710
		if (!(vma->vm_flags & VM_NONLINEAR)) {
L
Linus Torvalds 已提交
711
			root = &mapping->i_mmap;
712
			uprobe_munmap(vma, vma->vm_start, vma->vm_end);
713 714

			if (adjust_next)
715 716
				uprobe_munmap(next, next->vm_start,
							next->vm_end);
717 718
		}

719
		mutex_lock(&mapping->i_mmap_mutex);
L
Linus Torvalds 已提交
720 721
		if (insert) {
			/*
722
			 * Put into interval tree now, so instantiated pages
L
Linus Torvalds 已提交
723 724 725 726 727 728 729 730
			 * are visible to arm/parisc __flush_dcache_page
			 * throughout; but we cannot insert into address
			 * space until vma start or end is updated.
			 */
			__vma_link_file(insert);
		}
	}

731 732
	vma_adjust_trans_huge(vma, start, end, adjust_next);

733 734 735 736
	anon_vma = vma->anon_vma;
	if (!anon_vma && adjust_next)
		anon_vma = next->anon_vma;
	if (anon_vma) {
737 738
		VM_BUG_ON(adjust_next && next->anon_vma &&
			  anon_vma != next->anon_vma);
739
		anon_vma_lock_write(anon_vma);
740 741 742 743
		anon_vma_interval_tree_pre_update_vma(vma);
		if (adjust_next)
			anon_vma_interval_tree_pre_update_vma(next);
	}
744

L
Linus Torvalds 已提交
745 746
	if (root) {
		flush_dcache_mmap_lock(mapping);
747
		vma_interval_tree_remove(vma, root);
L
Linus Torvalds 已提交
748
		if (adjust_next)
749
			vma_interval_tree_remove(next, root);
L
Linus Torvalds 已提交
750 751
	}

752 753 754 755 756 757 758 759
	if (start != vma->vm_start) {
		vma->vm_start = start;
		start_changed = true;
	}
	if (end != vma->vm_end) {
		vma->vm_end = end;
		end_changed = true;
	}
L
Linus Torvalds 已提交
760 761 762 763 764 765 766 767
	vma->vm_pgoff = pgoff;
	if (adjust_next) {
		next->vm_start += adjust_next << PAGE_SHIFT;
		next->vm_pgoff += adjust_next;
	}

	if (root) {
		if (adjust_next)
768 769
			vma_interval_tree_insert(next, root);
		vma_interval_tree_insert(vma, root);
L
Linus Torvalds 已提交
770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787
		flush_dcache_mmap_unlock(mapping);
	}

	if (remove_next) {
		/*
		 * vma_merge has merged next into vma, and needs
		 * us to remove next before dropping the locks.
		 */
		__vma_unlink(mm, next, vma);
		if (file)
			__remove_shared_vm_struct(next, file, mapping);
	} else if (insert) {
		/*
		 * split_vma has split insert from vma, and needs
		 * us to insert it before dropping the locks
		 * (it may either follow vma or precede it).
		 */
		__insert_vm_struct(mm, insert);
788 789 790 791 792 793 794 795 796
	} else {
		if (start_changed)
			vma_gap_update(vma);
		if (end_changed) {
			if (!next)
				mm->highest_vm_end = end;
			else if (!adjust_next)
				vma_gap_update(next);
		}
L
Linus Torvalds 已提交
797 798
	}

799 800 801 802
	if (anon_vma) {
		anon_vma_interval_tree_post_update_vma(vma);
		if (adjust_next)
			anon_vma_interval_tree_post_update_vma(next);
803
		anon_vma_unlock(anon_vma);
804
	}
L
Linus Torvalds 已提交
805
	if (mapping)
806
		mutex_unlock(&mapping->i_mmap_mutex);
L
Linus Torvalds 已提交
807

808
	if (root) {
809
		uprobe_mmap(vma);
810 811

		if (adjust_next)
812
			uprobe_mmap(next);
813 814
	}

L
Linus Torvalds 已提交
815
	if (remove_next) {
M
Matt Helsley 已提交
816
		if (file) {
817
			uprobe_munmap(next, next->vm_start, next->vm_end);
L
Linus Torvalds 已提交
818
			fput(file);
M
Matt Helsley 已提交
819
		}
820 821
		if (next->anon_vma)
			anon_vma_merge(vma, next);
L
Linus Torvalds 已提交
822
		mm->map_count--;
823
		mpol_put(vma_policy(next));
L
Linus Torvalds 已提交
824 825 826 827 828 829
		kmem_cache_free(vm_area_cachep, next);
		/*
		 * In mprotect's case 6 (see comments on vma_merge),
		 * we must remove another next too. It would clutter
		 * up the code too much to do both in one go.
		 */
830 831
		next = vma->vm_next;
		if (remove_next == 2)
L
Linus Torvalds 已提交
832
			goto again;
833 834 835 836
		else if (next)
			vma_gap_update(next);
		else
			mm->highest_vm_end = end;
L
Linus Torvalds 已提交
837
	}
838
	if (insert && file)
839
		uprobe_mmap(insert);
L
Linus Torvalds 已提交
840 841

	validate_mm(mm);
842 843

	return 0;
L
Linus Torvalds 已提交
844 845 846 847 848 849 850 851 852
}

/*
 * If the vma has a ->close operation then the driver probably needs to release
 * per-vma resources, so we don't attempt to merge those.
 */
static inline int is_mergeable_vma(struct vm_area_struct *vma,
			struct file *file, unsigned long vm_flags)
{
853
	if (vma->vm_flags ^ vm_flags)
L
Linus Torvalds 已提交
854 855 856 857 858 859 860 861 862
		return 0;
	if (vma->vm_file != file)
		return 0;
	if (vma->vm_ops && vma->vm_ops->close)
		return 0;
	return 1;
}

static inline int is_mergeable_anon_vma(struct anon_vma *anon_vma1,
S
Shaohua Li 已提交
863 864
					struct anon_vma *anon_vma2,
					struct vm_area_struct *vma)
L
Linus Torvalds 已提交
865
{
S
Shaohua Li 已提交
866 867 868 869 870 871 872 873
	/*
	 * The list_is_singular() test is to avoid merging VMA cloned from
	 * parents. This can improve scalability caused by anon_vma lock.
	 */
	if ((!anon_vma1 || !anon_vma2) && (!vma ||
		list_is_singular(&vma->anon_vma_chain)))
		return 1;
	return anon_vma1 == anon_vma2;
L
Linus Torvalds 已提交
874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891
}

/*
 * Return true if we can merge this (vm_flags,anon_vma,file,vm_pgoff)
 * in front of (at a lower virtual address and file offset than) the vma.
 *
 * We cannot merge two vmas if they have differently assigned (non-NULL)
 * anon_vmas, nor if same anon_vma is assigned but offsets incompatible.
 *
 * We don't check here for the merged mmap wrapping around the end of pagecache
 * indices (16TB on ia32) because do_mmap_pgoff() does not permit mmap's which
 * wrap, nor mmaps which cover the final page at index -1UL.
 */
static int
can_vma_merge_before(struct vm_area_struct *vma, unsigned long vm_flags,
	struct anon_vma *anon_vma, struct file *file, pgoff_t vm_pgoff)
{
	if (is_mergeable_vma(vma, file, vm_flags) &&
S
Shaohua Li 已提交
892
	    is_mergeable_anon_vma(anon_vma, vma->anon_vma, vma)) {
L
Linus Torvalds 已提交
893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910
		if (vma->vm_pgoff == vm_pgoff)
			return 1;
	}
	return 0;
}

/*
 * Return true if we can merge this (vm_flags,anon_vma,file,vm_pgoff)
 * beyond (at a higher virtual address and file offset than) the vma.
 *
 * We cannot merge two vmas if they have differently assigned (non-NULL)
 * anon_vmas, nor if same anon_vma is assigned but offsets incompatible.
 */
static int
can_vma_merge_after(struct vm_area_struct *vma, unsigned long vm_flags,
	struct anon_vma *anon_vma, struct file *file, pgoff_t vm_pgoff)
{
	if (is_mergeable_vma(vma, file, vm_flags) &&
S
Shaohua Li 已提交
911
	    is_mergeable_anon_vma(anon_vma, vma->anon_vma, vma)) {
L
Linus Torvalds 已提交
912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956
		pgoff_t vm_pglen;
		vm_pglen = (vma->vm_end - vma->vm_start) >> PAGE_SHIFT;
		if (vma->vm_pgoff + vm_pglen == vm_pgoff)
			return 1;
	}
	return 0;
}

/*
 * Given a mapping request (addr,end,vm_flags,file,pgoff), figure out
 * whether that can be merged with its predecessor or its successor.
 * Or both (it neatly fills a hole).
 *
 * In most cases - when called for mmap, brk or mremap - [addr,end) is
 * certain not to be mapped by the time vma_merge is called; but when
 * called for mprotect, it is certain to be already mapped (either at
 * an offset within prev, or at the start of next), and the flags of
 * this area are about to be changed to vm_flags - and the no-change
 * case has already been eliminated.
 *
 * The following mprotect cases have to be considered, where AAAA is
 * the area passed down from mprotect_fixup, never extending beyond one
 * vma, PPPPPP is the prev vma specified, and NNNNNN the next vma after:
 *
 *     AAAA             AAAA                AAAA          AAAA
 *    PPPPPPNNNNNN    PPPPPPNNNNNN    PPPPPPNNNNNN    PPPPNNNNXXXX
 *    cannot merge    might become    might become    might become
 *                    PPNNNNNNNNNN    PPPPPPPPPPNN    PPPPPPPPPPPP 6 or
 *    mmap, brk or    case 4 below    case 5 below    PPPPPPPPXXXX 7 or
 *    mremap move:                                    PPPPNNNNNNNN 8
 *        AAAA
 *    PPPP    NNNN    PPPPPPPPPPPP    PPPPPPPPNNNN    PPPPNNNNNNNN
 *    might become    case 1 below    case 2 below    case 3 below
 *
 * Odd one out? Case 8, because it extends NNNN but needs flags of XXXX:
 * mprotect_fixup updates vm_flags & vm_page_prot on successful return.
 */
struct vm_area_struct *vma_merge(struct mm_struct *mm,
			struct vm_area_struct *prev, unsigned long addr,
			unsigned long end, unsigned long vm_flags,
		     	struct anon_vma *anon_vma, struct file *file,
			pgoff_t pgoff, struct mempolicy *policy)
{
	pgoff_t pglen = (end - addr) >> PAGE_SHIFT;
	struct vm_area_struct *area, *next;
957
	int err;
L
Linus Torvalds 已提交
958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988

	/*
	 * We later require that vma->vm_flags == vm_flags,
	 * so this tests vma->vm_flags & VM_SPECIAL, too.
	 */
	if (vm_flags & VM_SPECIAL)
		return NULL;

	if (prev)
		next = prev->vm_next;
	else
		next = mm->mmap;
	area = next;
	if (next && next->vm_end == end)		/* cases 6, 7, 8 */
		next = next->vm_next;

	/*
	 * Can it merge with the predecessor?
	 */
	if (prev && prev->vm_end == addr &&
  			mpol_equal(vma_policy(prev), policy) &&
			can_vma_merge_after(prev, vm_flags,
						anon_vma, file, pgoff)) {
		/*
		 * OK, it can.  Can we now merge in the successor as well?
		 */
		if (next && end == next->vm_start &&
				mpol_equal(policy, vma_policy(next)) &&
				can_vma_merge_before(next, vm_flags,
					anon_vma, file, pgoff+pglen) &&
				is_mergeable_anon_vma(prev->anon_vma,
S
Shaohua Li 已提交
989
						      next->anon_vma, NULL)) {
L
Linus Torvalds 已提交
990
							/* cases 1, 6 */
991
			err = vma_adjust(prev, prev->vm_start,
L
Linus Torvalds 已提交
992 993
				next->vm_end, prev->vm_pgoff, NULL);
		} else					/* cases 2, 5, 7 */
994
			err = vma_adjust(prev, prev->vm_start,
L
Linus Torvalds 已提交
995
				end, prev->vm_pgoff, NULL);
996 997
		if (err)
			return NULL;
A
Andrea Arcangeli 已提交
998
		khugepaged_enter_vma_merge(prev);
L
Linus Torvalds 已提交
999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009
		return prev;
	}

	/*
	 * Can this new request be merged in front of next?
	 */
	if (next && end == next->vm_start &&
 			mpol_equal(policy, vma_policy(next)) &&
			can_vma_merge_before(next, vm_flags,
					anon_vma, file, pgoff+pglen)) {
		if (prev && addr < prev->vm_end)	/* case 4 */
1010
			err = vma_adjust(prev, prev->vm_start,
L
Linus Torvalds 已提交
1011 1012
				addr, prev->vm_pgoff, NULL);
		else					/* cases 3, 8 */
1013
			err = vma_adjust(area, addr, next->vm_end,
L
Linus Torvalds 已提交
1014
				next->vm_pgoff - pglen, NULL);
1015 1016
		if (err)
			return NULL;
A
Andrea Arcangeli 已提交
1017
		khugepaged_enter_vma_merge(area);
L
Linus Torvalds 已提交
1018 1019 1020 1021 1022 1023
		return area;
	}

	return NULL;
}

1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078
/*
 * Rough compatbility check to quickly see if it's even worth looking
 * at sharing an anon_vma.
 *
 * They need to have the same vm_file, and the flags can only differ
 * in things that mprotect may change.
 *
 * NOTE! The fact that we share an anon_vma doesn't _have_ to mean that
 * we can merge the two vma's. For example, we refuse to merge a vma if
 * there is a vm_ops->close() function, because that indicates that the
 * driver is doing some kind of reference counting. But that doesn't
 * really matter for the anon_vma sharing case.
 */
static int anon_vma_compatible(struct vm_area_struct *a, struct vm_area_struct *b)
{
	return a->vm_end == b->vm_start &&
		mpol_equal(vma_policy(a), vma_policy(b)) &&
		a->vm_file == b->vm_file &&
		!((a->vm_flags ^ b->vm_flags) & ~(VM_READ|VM_WRITE|VM_EXEC)) &&
		b->vm_pgoff == a->vm_pgoff + ((b->vm_start - a->vm_start) >> PAGE_SHIFT);
}

/*
 * Do some basic sanity checking to see if we can re-use the anon_vma
 * from 'old'. The 'a'/'b' vma's are in VM order - one of them will be
 * the same as 'old', the other will be the new one that is trying
 * to share the anon_vma.
 *
 * NOTE! This runs with mm_sem held for reading, so it is possible that
 * the anon_vma of 'old' is concurrently in the process of being set up
 * by another page fault trying to merge _that_. But that's ok: if it
 * is being set up, that automatically means that it will be a singleton
 * acceptable for merging, so we can do all of this optimistically. But
 * we do that ACCESS_ONCE() to make sure that we never re-load the pointer.
 *
 * IOW: that the "list_is_singular()" test on the anon_vma_chain only
 * matters for the 'stable anon_vma' case (ie the thing we want to avoid
 * is to return an anon_vma that is "complex" due to having gone through
 * a fork).
 *
 * We also make sure that the two vma's are compatible (adjacent,
 * and with the same memory policies). That's all stable, even with just
 * a read lock on the mm_sem.
 */
static struct anon_vma *reusable_anon_vma(struct vm_area_struct *old, struct vm_area_struct *a, struct vm_area_struct *b)
{
	if (anon_vma_compatible(a, b)) {
		struct anon_vma *anon_vma = ACCESS_ONCE(old->anon_vma);

		if (anon_vma && list_is_singular(&old->anon_vma_chain))
			return anon_vma;
	}
	return NULL;
}

L
Linus Torvalds 已提交
1079 1080 1081 1082 1083 1084 1085 1086 1087 1088
/*
 * find_mergeable_anon_vma is used by anon_vma_prepare, to check
 * neighbouring vmas for a suitable anon_vma, before it goes off
 * to allocate a new anon_vma.  It checks because a repetitive
 * sequence of mprotects and faults may otherwise lead to distinct
 * anon_vmas being allocated, preventing vma merge in subsequent
 * mprotect.
 */
struct anon_vma *find_mergeable_anon_vma(struct vm_area_struct *vma)
{
1089
	struct anon_vma *anon_vma;
L
Linus Torvalds 已提交
1090 1091 1092 1093 1094 1095
	struct vm_area_struct *near;

	near = vma->vm_next;
	if (!near)
		goto try_prev;

1096 1097 1098
	anon_vma = reusable_anon_vma(near, vma, near);
	if (anon_vma)
		return anon_vma;
L
Linus Torvalds 已提交
1099
try_prev:
1100
	near = vma->vm_prev;
L
Linus Torvalds 已提交
1101 1102 1103
	if (!near)
		goto none;

1104 1105 1106
	anon_vma = reusable_anon_vma(near, near, vma);
	if (anon_vma)
		return anon_vma;
L
Linus Torvalds 已提交
1107 1108 1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119
none:
	/*
	 * There's no absolute need to look only at touching neighbours:
	 * we could search further afield for "compatible" anon_vmas.
	 * But it would probably just be a waste of time searching,
	 * or lead to too many vmas hanging off the same anon_vma.
	 * We're trying to allow mprotect remerging later on,
	 * not trying to minimize memory used for anon_vmas.
	 */
	return NULL;
}

#ifdef CONFIG_PROC_FS
1120
void vm_stat_account(struct mm_struct *mm, unsigned long flags,
L
Linus Torvalds 已提交
1121 1122 1123 1124 1125
						struct file *file, long pages)
{
	const unsigned long stack_flags
		= VM_STACK_FLAGS & (VM_GROWSUP|VM_GROWSDOWN);

1126 1127
	mm->total_vm += pages;

L
Linus Torvalds 已提交
1128 1129 1130 1131 1132 1133 1134 1135 1136
	if (file) {
		mm->shared_vm += pages;
		if ((flags & (VM_EXEC|VM_WRITE)) == VM_EXEC)
			mm->exec_vm += pages;
	} else if (flags & stack_flags)
		mm->stack_vm += pages;
}
#endif /* CONFIG_PROC_FS */

A
Al Viro 已提交
1137 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149
/*
 * If a hint addr is less than mmap_min_addr change hint to be as
 * low as possible but still greater than mmap_min_addr
 */
static inline unsigned long round_hint_to_min(unsigned long hint)
{
	hint &= PAGE_MASK;
	if (((void *)hint != NULL) &&
	    (hint < mmap_min_addr))
		return PAGE_ALIGN(mmap_min_addr);
	return hint;
}

L
Linus Torvalds 已提交
1150
/*
1151
 * The caller must hold down_write(&current->mm->mmap_sem).
L
Linus Torvalds 已提交
1152 1153
 */

1154
unsigned long do_mmap_pgoff(struct file *file, unsigned long addr,
L
Linus Torvalds 已提交
1155 1156 1157 1158 1159
			unsigned long len, unsigned long prot,
			unsigned long flags, unsigned long pgoff)
{
	struct mm_struct * mm = current->mm;
	struct inode *inode;
1160
	vm_flags_t vm_flags;
L
Linus Torvalds 已提交
1161 1162 1163 1164 1165 1166 1167 1168

	/*
	 * Does the application expect PROT_READ to imply PROT_EXEC?
	 *
	 * (the exception is when the underlying filesystem is noexec
	 *  mounted, in which case we dont add PROT_EXEC.)
	 */
	if ((prot & PROT_READ) && (current->personality & READ_IMPLIES_EXEC))
1169
		if (!(file && (file->f_path.mnt->mnt_flags & MNT_NOEXEC)))
L
Linus Torvalds 已提交
1170 1171 1172 1173 1174
			prot |= PROT_EXEC;

	if (!len)
		return -EINVAL;

1175 1176 1177
	if (!(flags & MAP_FIXED))
		addr = round_hint_to_min(addr);

L
Linus Torvalds 已提交
1178 1179
	/* Careful about overflows.. */
	len = PAGE_ALIGN(len);
1180
	if (!len)
L
Linus Torvalds 已提交
1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199 1200 1201 1202 1203 1204
		return -ENOMEM;

	/* offset overflow? */
	if ((pgoff + (len >> PAGE_SHIFT)) < pgoff)
               return -EOVERFLOW;

	/* Too many mappings? */
	if (mm->map_count > sysctl_max_map_count)
		return -ENOMEM;

	/* Obtain the address to map to. we verify (or select) it and ensure
	 * that it represents a valid section of the address space.
	 */
	addr = get_unmapped_area(file, addr, len, pgoff, flags);
	if (addr & ~PAGE_MASK)
		return addr;

	/* Do simple checking here so the lower-level routines won't have
	 * to. we assume access permissions have been handled by the open
	 * of the memory object, so we don't do any here.
	 */
	vm_flags = calc_vm_prot_bits(prot) | calc_vm_flag_bits(flags) |
			mm->def_flags | VM_MAYREAD | VM_MAYWRITE | VM_MAYEXEC;

H
Huang Shijie 已提交
1205
	if (flags & MAP_LOCKED)
L
Linus Torvalds 已提交
1206 1207
		if (!can_do_mlock())
			return -EPERM;
1208

L
Linus Torvalds 已提交
1209 1210 1211
	/* mlock MCL_FUTURE? */
	if (vm_flags & VM_LOCKED) {
		unsigned long locked, lock_limit;
1212 1213
		locked = len >> PAGE_SHIFT;
		locked += mm->locked_vm;
J
Jiri Slaby 已提交
1214
		lock_limit = rlimit(RLIMIT_MEMLOCK);
1215
		lock_limit >>= PAGE_SHIFT;
L
Linus Torvalds 已提交
1216 1217 1218 1219
		if (locked > lock_limit && !capable(CAP_IPC_LOCK))
			return -EAGAIN;
	}

1220
	inode = file ? file->f_path.dentry->d_inode : NULL;
L
Linus Torvalds 已提交
1221 1222 1223 1224 1225 1226 1227 1228 1229 1230 1231 1232 1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 1243 1244 1245 1246 1247 1248

	if (file) {
		switch (flags & MAP_TYPE) {
		case MAP_SHARED:
			if ((prot&PROT_WRITE) && !(file->f_mode&FMODE_WRITE))
				return -EACCES;

			/*
			 * Make sure we don't allow writing to an append-only
			 * file..
			 */
			if (IS_APPEND(inode) && (file->f_mode & FMODE_WRITE))
				return -EACCES;

			/*
			 * Make sure there are no mandatory locks on the file.
			 */
			if (locks_verify_locked(inode))
				return -EAGAIN;

			vm_flags |= VM_SHARED | VM_MAYSHARE;
			if (!(file->f_mode & FMODE_WRITE))
				vm_flags &= ~(VM_MAYWRITE | VM_SHARED);

			/* fall through */
		case MAP_PRIVATE:
			if (!(file->f_mode & FMODE_READ))
				return -EACCES;
1249
			if (file->f_path.mnt->mnt_flags & MNT_NOEXEC) {
L
Linus Torvalds 已提交
1250 1251 1252 1253 1254 1255 1256
				if (vm_flags & VM_EXEC)
					return -EPERM;
				vm_flags &= ~VM_MAYEXEC;
			}

			if (!file->f_op || !file->f_op->mmap)
				return -ENODEV;
L
Linus Torvalds 已提交
1257 1258 1259 1260 1261 1262 1263 1264
			break;

		default:
			return -EINVAL;
		}
	} else {
		switch (flags & MAP_TYPE) {
		case MAP_SHARED:
1265 1266 1267 1268
			/*
			 * Ignore pgoff.
			 */
			pgoff = 0;
L
Linus Torvalds 已提交
1269 1270 1271 1272 1273 1274 1275 1276 1277 1278 1279 1280 1281
			vm_flags |= VM_SHARED | VM_MAYSHARE;
			break;
		case MAP_PRIVATE:
			/*
			 * Set pgoff according to addr for anon_vma.
			 */
			pgoff = addr >> PAGE_SHIFT;
			break;
		default:
			return -EINVAL;
		}
	}

1282
	return mmap_region(file, addr, len, flags, vm_flags, pgoff);
M
Miklos Szeredi 已提交
1283
}
1284

H
Hugh Dickins 已提交
1285 1286 1287 1288 1289 1290 1291 1292
SYSCALL_DEFINE6(mmap_pgoff, unsigned long, addr, unsigned long, len,
		unsigned long, prot, unsigned long, flags,
		unsigned long, fd, unsigned long, pgoff)
{
	struct file *file = NULL;
	unsigned long retval = -EBADF;

	if (!(flags & MAP_ANONYMOUS)) {
A
Al Viro 已提交
1293
		audit_mmap_fd(fd, flags);
H
Hugh Dickins 已提交
1294 1295 1296 1297 1298 1299 1300 1301 1302 1303 1304 1305 1306
		if (unlikely(flags & MAP_HUGETLB))
			return -EINVAL;
		file = fget(fd);
		if (!file)
			goto out;
	} else if (flags & MAP_HUGETLB) {
		struct user_struct *user = NULL;
		/*
		 * VM_NORESERVE is used because the reservations will be
		 * taken when vm_ops->mmap() is called
		 * A dummy user value is used because we are not locking
		 * memory so no accounting is necessary
		 */
1307
		file = hugetlb_file_setup(HUGETLB_ANON_FILE, addr, len,
1308 1309 1310
				VM_NORESERVE,
				&user, HUGETLB_ANONHUGE_INODE,
				(flags >> MAP_HUGE_SHIFT) & MAP_HUGE_MASK);
H
Hugh Dickins 已提交
1311 1312 1313 1314 1315 1316
		if (IS_ERR(file))
			return PTR_ERR(file);
	}

	flags &= ~(MAP_EXECUTABLE | MAP_DENYWRITE);

A
Al Viro 已提交
1317
	retval = vm_mmap_pgoff(file, addr, len, prot, flags, pgoff);
H
Hugh Dickins 已提交
1318 1319 1320 1321 1322 1323
	if (file)
		fput(file);
out:
	return retval;
}

C
Christoph Hellwig 已提交
1324 1325 1326 1327 1328 1329 1330 1331 1332 1333 1334 1335 1336 1337 1338 1339 1340 1341 1342 1343 1344 1345 1346 1347
#ifdef __ARCH_WANT_SYS_OLD_MMAP
struct mmap_arg_struct {
	unsigned long addr;
	unsigned long len;
	unsigned long prot;
	unsigned long flags;
	unsigned long fd;
	unsigned long offset;
};

SYSCALL_DEFINE1(old_mmap, struct mmap_arg_struct __user *, arg)
{
	struct mmap_arg_struct a;

	if (copy_from_user(&a, arg, sizeof(a)))
		return -EFAULT;
	if (a.offset & ~PAGE_MASK)
		return -EINVAL;

	return sys_mmap_pgoff(a.addr, a.len, a.prot, a.flags, a.fd,
			      a.offset >> PAGE_SHIFT);
}
#endif /* __ARCH_WANT_SYS_OLD_MMAP */

A
Alexey Dobriyan 已提交
1348 1349 1350 1351 1352 1353 1354 1355
/*
 * Some shared mappigns will want the pages marked read-only
 * to track write events. If so, we'll downgrade vm_page_prot
 * to the private version (using protection_map[] without the
 * VM_SHARED bit).
 */
int vma_wants_writenotify(struct vm_area_struct *vma)
{
1356
	vm_flags_t vm_flags = vma->vm_flags;
A
Alexey Dobriyan 已提交
1357 1358 1359 1360 1361 1362 1363 1364 1365 1366 1367

	/* If it was private or non-writable, the write bit is already clear */
	if ((vm_flags & (VM_WRITE|VM_SHARED)) != ((VM_WRITE|VM_SHARED)))
		return 0;

	/* The backer wishes to know when pages are first written to? */
	if (vma->vm_ops && vma->vm_ops->page_mkwrite)
		return 1;

	/* The open routine did something to the protections already? */
	if (pgprot_val(vma->vm_page_prot) !=
1368
	    pgprot_val(vm_get_page_prot(vm_flags)))
A
Alexey Dobriyan 已提交
1369 1370 1371
		return 0;

	/* Specialty mapping? */
1372
	if (vm_flags & VM_PFNMAP)
A
Alexey Dobriyan 已提交
1373 1374 1375 1376 1377 1378 1379
		return 0;

	/* Can the mapping track the dirty pages? */
	return vma->vm_file && vma->vm_file->f_mapping &&
		mapping_cap_account_dirty(vma->vm_file->f_mapping);
}

1380 1381
/*
 * We account for memory if it's a private writeable mapping,
1382
 * not hugepages and VM_NORESERVE wasn't set.
1383
 */
1384
static inline int accountable_mapping(struct file *file, vm_flags_t vm_flags)
1385
{
1386 1387 1388 1389 1390 1391 1392
	/*
	 * hugetlb has its own accounting separate from the core VM
	 * VM_HUGETLB may not be set yet so we cannot check for that flag.
	 */
	if (file && is_file_hugepages(file))
		return 0;

1393 1394 1395
	return (vm_flags & (VM_NORESERVE | VM_SHARED | VM_WRITE)) == VM_WRITE;
}

M
Miklos Szeredi 已提交
1396 1397
unsigned long mmap_region(struct file *file, unsigned long addr,
			  unsigned long len, unsigned long flags,
1398
			  vm_flags_t vm_flags, unsigned long pgoff)
M
Miklos Szeredi 已提交
1399 1400 1401 1402 1403 1404 1405 1406 1407
{
	struct mm_struct *mm = current->mm;
	struct vm_area_struct *vma, *prev;
	int correct_wcount = 0;
	int error;
	struct rb_node **rb_link, *rb_parent;
	unsigned long charged = 0;
	struct inode *inode =  file ? file->f_path.dentry->d_inode : NULL;

L
Linus Torvalds 已提交
1408 1409 1410
	/* Clear old maps */
	error = -ENOMEM;
munmap_back:
1411
	if (find_vma_links(mm, addr, addr + len, &prev, &rb_link, &rb_parent)) {
L
Linus Torvalds 已提交
1412 1413 1414 1415 1416 1417
		if (do_munmap(mm, addr, len))
			return -ENOMEM;
		goto munmap_back;
	}

	/* Check against address space limit. */
A
akpm@osdl.org 已提交
1418
	if (!may_expand_vm(mm, len >> PAGE_SHIFT))
L
Linus Torvalds 已提交
1419 1420
		return -ENOMEM;

1421 1422
	/*
	 * Set 'VM_NORESERVE' if we should not account for the
1423
	 * memory use of this mapping.
1424
	 */
1425 1426 1427 1428 1429 1430 1431 1432 1433
	if ((flags & MAP_NORESERVE)) {
		/* We honor MAP_NORESERVE if allowed to overcommit */
		if (sysctl_overcommit_memory != OVERCOMMIT_NEVER)
			vm_flags |= VM_NORESERVE;

		/* hugetlb applies strict overcommit unless MAP_NORESERVE */
		if (file && is_file_hugepages(file))
			vm_flags |= VM_NORESERVE;
	}
1434

1435 1436 1437
	/*
	 * Private writable mapping: check memory availability
	 */
1438
	if (accountable_mapping(file, vm_flags)) {
1439
		charged = len >> PAGE_SHIFT;
1440
		if (security_vm_enough_memory_mm(mm, charged))
1441 1442
			return -ENOMEM;
		vm_flags |= VM_ACCOUNT;
L
Linus Torvalds 已提交
1443 1444 1445
	}

	/*
1446
	 * Can we just expand an old mapping?
L
Linus Torvalds 已提交
1447
	 */
1448 1449 1450
	vma = vma_merge(mm, prev, addr, addr + len, vm_flags, NULL, file, pgoff, NULL);
	if (vma)
		goto out;
L
Linus Torvalds 已提交
1451 1452 1453 1454 1455 1456

	/*
	 * Determine the object being mapped and call the appropriate
	 * specific mapper. the address has already been validated, but
	 * not unmapped, but the maps are removed from the list.
	 */
P
Pekka Enberg 已提交
1457
	vma = kmem_cache_zalloc(vm_area_cachep, GFP_KERNEL);
L
Linus Torvalds 已提交
1458 1459 1460 1461 1462 1463 1464 1465 1466
	if (!vma) {
		error = -ENOMEM;
		goto unacct_error;
	}

	vma->vm_mm = mm;
	vma->vm_start = addr;
	vma->vm_end = addr + len;
	vma->vm_flags = vm_flags;
1467
	vma->vm_page_prot = vm_get_page_prot(vm_flags);
L
Linus Torvalds 已提交
1468
	vma->vm_pgoff = pgoff;
1469
	INIT_LIST_HEAD(&vma->anon_vma_chain);
L
Linus Torvalds 已提交
1470

1471 1472
	error = -EINVAL;	/* when rejecting VM_GROWSDOWN|VM_GROWSUP */

L
Linus Torvalds 已提交
1473 1474 1475 1476 1477 1478 1479 1480 1481
	if (file) {
		if (vm_flags & (VM_GROWSDOWN|VM_GROWSUP))
			goto free_vma;
		if (vm_flags & VM_DENYWRITE) {
			error = deny_write_access(file);
			if (error)
				goto free_vma;
			correct_wcount = 1;
		}
A
Al Viro 已提交
1482
		vma->vm_file = get_file(file);
L
Linus Torvalds 已提交
1483 1484 1485
		error = file->f_op->mmap(file, vma);
		if (error)
			goto unmap_and_free_vma;
1486 1487 1488 1489 1490

		/* Can addr have changed??
		 *
		 * Answer: Yes, several device drivers can do it in their
		 *         f_op->mmap method. -DaveM
1491 1492
		 * Bug: If addr is changed, prev, rb_link, rb_parent should
		 *      be updated for vma_link()
1493
		 */
1494 1495
		WARN_ON_ONCE(addr != vma->vm_start);

1496 1497 1498
		addr = vma->vm_start;
		pgoff = vma->vm_pgoff;
		vm_flags = vma->vm_flags;
L
Linus Torvalds 已提交
1499
	} else if (vm_flags & VM_SHARED) {
1500 1501
		if (unlikely(vm_flags & (VM_GROWSDOWN|VM_GROWSUP)))
			goto free_vma;
L
Linus Torvalds 已提交
1502 1503 1504 1505 1506
		error = shmem_zero_setup(vma);
		if (error)
			goto free_vma;
	}

1507 1508 1509 1510 1511 1512 1513 1514 1515 1516
	if (vma_wants_writenotify(vma)) {
		pgprot_t pprot = vma->vm_page_prot;

		/* Can vma->vm_page_prot have changed??
		 *
		 * Answer: Yes, drivers may have changed it in their
		 *         f_op->mmap method.
		 *
		 * Ensures that vmas marked as uncached stay that way.
		 */
1517
		vma->vm_page_prot = vm_get_page_prot(vm_flags & ~VM_SHARED);
1518 1519 1520
		if (pgprot_val(pprot) == pgprot_val(pgprot_noncached(pprot)))
			vma->vm_page_prot = pgprot_noncached(vma->vm_page_prot);
	}
1521

1522 1523
	vma_link(mm, vma, prev, rb_link, rb_parent);
	file = vma->vm_file;
1524 1525 1526 1527 1528

	/* Once vma denies write, undo our temporary denial count */
	if (correct_wcount)
		atomic_inc(&inode->i_writecount);
out:
1529
	perf_event_mmap(vma);
1530

1531
	vm_stat_account(mm, vm_flags, file, len >> PAGE_SHIFT);
L
Linus Torvalds 已提交
1532
	if (vm_flags & VM_LOCKED) {
1533 1534
		if (!mlock_vma_pages_range(vma, addr, addr + len))
			mm->locked_vm += (len >> PAGE_SHIFT);
1535
	} else if ((flags & MAP_POPULATE) && !(flags & MAP_NONBLOCK))
1536
		make_pages_present(addr, addr + len);
1537

1538 1539
	if (file)
		uprobe_mmap(vma);
1540

L
Linus Torvalds 已提交
1541 1542 1543 1544 1545 1546 1547 1548 1549
	return addr;

unmap_and_free_vma:
	if (correct_wcount)
		atomic_inc(&inode->i_writecount);
	vma->vm_file = NULL;
	fput(file);

	/* Undo any partial mapping done by a device driver. */
1550 1551
	unmap_region(mm, vma, prev, vma->vm_start, vma->vm_end);
	charged = 0;
L
Linus Torvalds 已提交
1552 1553 1554 1555 1556 1557 1558 1559
free_vma:
	kmem_cache_free(vm_area_cachep, vma);
unacct_error:
	if (charged)
		vm_unacct_memory(charged);
	return error;
}

1560 1561 1562 1563 1564 1565 1566 1567 1568 1569 1570 1571 1572 1573 1574 1575 1576 1577 1578 1579 1580 1581 1582 1583 1584 1585 1586 1587 1588 1589 1590 1591 1592 1593 1594 1595 1596 1597 1598 1599 1600 1601 1602 1603 1604 1605 1606 1607 1608 1609 1610 1611 1612 1613 1614 1615 1616 1617 1618 1619 1620 1621 1622 1623 1624 1625 1626 1627 1628 1629 1630 1631 1632 1633 1634 1635 1636 1637 1638 1639 1640 1641 1642 1643 1644 1645 1646 1647 1648 1649 1650 1651 1652 1653 1654 1655 1656 1657 1658 1659 1660 1661 1662 1663 1664 1665 1666 1667 1668 1669 1670 1671 1672 1673 1674 1675 1676 1677 1678 1679 1680 1681 1682 1683 1684 1685 1686 1687 1688 1689 1690 1691 1692 1693 1694 1695 1696 1697 1698 1699 1700 1701 1702 1703 1704 1705 1706 1707 1708 1709 1710 1711 1712 1713 1714 1715 1716 1717 1718 1719 1720 1721 1722 1723 1724 1725 1726 1727 1728 1729 1730 1731 1732 1733 1734 1735 1736 1737 1738 1739 1740 1741 1742 1743 1744 1745 1746 1747 1748 1749 1750 1751 1752 1753 1754 1755 1756 1757 1758 1759
unsigned long unmapped_area(struct vm_unmapped_area_info *info)
{
	/*
	 * We implement the search by looking for an rbtree node that
	 * immediately follows a suitable gap. That is,
	 * - gap_start = vma->vm_prev->vm_end <= info->high_limit - length;
	 * - gap_end   = vma->vm_start        >= info->low_limit  + length;
	 * - gap_end - gap_start >= length
	 */

	struct mm_struct *mm = current->mm;
	struct vm_area_struct *vma;
	unsigned long length, low_limit, high_limit, gap_start, gap_end;

	/* Adjust search length to account for worst case alignment overhead */
	length = info->length + info->align_mask;
	if (length < info->length)
		return -ENOMEM;

	/* Adjust search limits by the desired length */
	if (info->high_limit < length)
		return -ENOMEM;
	high_limit = info->high_limit - length;

	if (info->low_limit > high_limit)
		return -ENOMEM;
	low_limit = info->low_limit + length;

	/* Check if rbtree root looks promising */
	if (RB_EMPTY_ROOT(&mm->mm_rb))
		goto check_highest;
	vma = rb_entry(mm->mm_rb.rb_node, struct vm_area_struct, vm_rb);
	if (vma->rb_subtree_gap < length)
		goto check_highest;

	while (true) {
		/* Visit left subtree if it looks promising */
		gap_end = vma->vm_start;
		if (gap_end >= low_limit && vma->vm_rb.rb_left) {
			struct vm_area_struct *left =
				rb_entry(vma->vm_rb.rb_left,
					 struct vm_area_struct, vm_rb);
			if (left->rb_subtree_gap >= length) {
				vma = left;
				continue;
			}
		}

		gap_start = vma->vm_prev ? vma->vm_prev->vm_end : 0;
check_current:
		/* Check if current node has a suitable gap */
		if (gap_start > high_limit)
			return -ENOMEM;
		if (gap_end >= low_limit && gap_end - gap_start >= length)
			goto found;

		/* Visit right subtree if it looks promising */
		if (vma->vm_rb.rb_right) {
			struct vm_area_struct *right =
				rb_entry(vma->vm_rb.rb_right,
					 struct vm_area_struct, vm_rb);
			if (right->rb_subtree_gap >= length) {
				vma = right;
				continue;
			}
		}

		/* Go back up the rbtree to find next candidate node */
		while (true) {
			struct rb_node *prev = &vma->vm_rb;
			if (!rb_parent(prev))
				goto check_highest;
			vma = rb_entry(rb_parent(prev),
				       struct vm_area_struct, vm_rb);
			if (prev == vma->vm_rb.rb_left) {
				gap_start = vma->vm_prev->vm_end;
				gap_end = vma->vm_start;
				goto check_current;
			}
		}
	}

check_highest:
	/* Check highest gap, which does not precede any rbtree node */
	gap_start = mm->highest_vm_end;
	gap_end = ULONG_MAX;  /* Only for VM_BUG_ON below */
	if (gap_start > high_limit)
		return -ENOMEM;

found:
	/* We found a suitable gap. Clip it with the original low_limit. */
	if (gap_start < info->low_limit)
		gap_start = info->low_limit;

	/* Adjust gap address to the desired alignment */
	gap_start += (info->align_offset - gap_start) & info->align_mask;

	VM_BUG_ON(gap_start + info->length > info->high_limit);
	VM_BUG_ON(gap_start + info->length > gap_end);
	return gap_start;
}

unsigned long unmapped_area_topdown(struct vm_unmapped_area_info *info)
{
	struct mm_struct *mm = current->mm;
	struct vm_area_struct *vma;
	unsigned long length, low_limit, high_limit, gap_start, gap_end;

	/* Adjust search length to account for worst case alignment overhead */
	length = info->length + info->align_mask;
	if (length < info->length)
		return -ENOMEM;

	/*
	 * Adjust search limits by the desired length.
	 * See implementation comment at top of unmapped_area().
	 */
	gap_end = info->high_limit;
	if (gap_end < length)
		return -ENOMEM;
	high_limit = gap_end - length;

	if (info->low_limit > high_limit)
		return -ENOMEM;
	low_limit = info->low_limit + length;

	/* Check highest gap, which does not precede any rbtree node */
	gap_start = mm->highest_vm_end;
	if (gap_start <= high_limit)
		goto found_highest;

	/* Check if rbtree root looks promising */
	if (RB_EMPTY_ROOT(&mm->mm_rb))
		return -ENOMEM;
	vma = rb_entry(mm->mm_rb.rb_node, struct vm_area_struct, vm_rb);
	if (vma->rb_subtree_gap < length)
		return -ENOMEM;

	while (true) {
		/* Visit right subtree if it looks promising */
		gap_start = vma->vm_prev ? vma->vm_prev->vm_end : 0;
		if (gap_start <= high_limit && vma->vm_rb.rb_right) {
			struct vm_area_struct *right =
				rb_entry(vma->vm_rb.rb_right,
					 struct vm_area_struct, vm_rb);
			if (right->rb_subtree_gap >= length) {
				vma = right;
				continue;
			}
		}

check_current:
		/* Check if current node has a suitable gap */
		gap_end = vma->vm_start;
		if (gap_end < low_limit)
			return -ENOMEM;
		if (gap_start <= high_limit && gap_end - gap_start >= length)
			goto found;

		/* Visit left subtree if it looks promising */
		if (vma->vm_rb.rb_left) {
			struct vm_area_struct *left =
				rb_entry(vma->vm_rb.rb_left,
					 struct vm_area_struct, vm_rb);
			if (left->rb_subtree_gap >= length) {
				vma = left;
				continue;
			}
		}

		/* Go back up the rbtree to find next candidate node */
		while (true) {
			struct rb_node *prev = &vma->vm_rb;
			if (!rb_parent(prev))
				return -ENOMEM;
			vma = rb_entry(rb_parent(prev),
				       struct vm_area_struct, vm_rb);
			if (prev == vma->vm_rb.rb_right) {
				gap_start = vma->vm_prev ?
					vma->vm_prev->vm_end : 0;
				goto check_current;
			}
		}
	}

found:
	/* We found a suitable gap. Clip it with the original high_limit. */
	if (gap_end > info->high_limit)
		gap_end = info->high_limit;

found_highest:
	/* Compute highest gap address at the desired alignment */
	gap_end -= info->length;
	gap_end -= (gap_end - info->align_offset) & info->align_mask;

	VM_BUG_ON(gap_end < info->low_limit);
	VM_BUG_ON(gap_end < gap_start);
	return gap_end;
}

L
Linus Torvalds 已提交
1760 1761 1762 1763 1764 1765 1766 1767 1768 1769 1770 1771 1772 1773 1774 1775 1776 1777
/* Get an address range which is currently unmapped.
 * For shmat() with addr=0.
 *
 * Ugly calling convention alert:
 * Return value with the low bits set means error value,
 * ie
 *	if (ret & ~PAGE_MASK)
 *		error = ret;
 *
 * This function "knows" that -ENOMEM has the bits set.
 */
#ifndef HAVE_ARCH_UNMAPPED_AREA
unsigned long
arch_get_unmapped_area(struct file *filp, unsigned long addr,
		unsigned long len, unsigned long pgoff, unsigned long flags)
{
	struct mm_struct *mm = current->mm;
	struct vm_area_struct *vma;
1778
	struct vm_unmapped_area_info info;
L
Linus Torvalds 已提交
1779 1780 1781 1782

	if (len > TASK_SIZE)
		return -ENOMEM;

1783 1784 1785
	if (flags & MAP_FIXED)
		return addr;

L
Linus Torvalds 已提交
1786 1787 1788 1789 1790 1791 1792 1793
	if (addr) {
		addr = PAGE_ALIGN(addr);
		vma = find_vma(mm, addr);
		if (TASK_SIZE - len >= addr &&
		    (!vma || addr + len <= vma->vm_start))
			return addr;
	}

1794 1795 1796 1797 1798 1799
	info.flags = 0;
	info.length = len;
	info.low_limit = TASK_UNMAPPED_BASE;
	info.high_limit = TASK_SIZE;
	info.align_mask = 0;
	return vm_unmapped_area(&info);
L
Linus Torvalds 已提交
1800 1801 1802
}
#endif	

1803
void arch_unmap_area(struct mm_struct *mm, unsigned long addr)
L
Linus Torvalds 已提交
1804 1805 1806 1807
{
	/*
	 * Is this a new hole at the lowest possible address?
	 */
1808
	if (addr >= TASK_UNMAPPED_BASE && addr < mm->free_area_cache)
1809
		mm->free_area_cache = addr;
L
Linus Torvalds 已提交
1810 1811 1812 1813 1814 1815 1816 1817 1818 1819 1820 1821 1822 1823
}

/*
 * This mmap-allocator allocates new areas top-down from below the
 * stack's low limit (the base):
 */
#ifndef HAVE_ARCH_UNMAPPED_AREA_TOPDOWN
unsigned long
arch_get_unmapped_area_topdown(struct file *filp, const unsigned long addr0,
			  const unsigned long len, const unsigned long pgoff,
			  const unsigned long flags)
{
	struct vm_area_struct *vma;
	struct mm_struct *mm = current->mm;
1824 1825
	unsigned long addr = addr0;
	struct vm_unmapped_area_info info;
L
Linus Torvalds 已提交
1826 1827 1828 1829 1830

	/* requested length too big for entire address space */
	if (len > TASK_SIZE)
		return -ENOMEM;

1831 1832 1833
	if (flags & MAP_FIXED)
		return addr;

L
Linus Torvalds 已提交
1834 1835 1836 1837 1838 1839 1840 1841 1842
	/* requesting a specific address */
	if (addr) {
		addr = PAGE_ALIGN(addr);
		vma = find_vma(mm, addr);
		if (TASK_SIZE - len >= addr &&
				(!vma || addr + len <= vma->vm_start))
			return addr;
	}

1843 1844 1845 1846 1847 1848
	info.flags = VM_UNMAPPED_AREA_TOPDOWN;
	info.length = len;
	info.low_limit = PAGE_SIZE;
	info.high_limit = mm->mmap_base;
	info.align_mask = 0;
	addr = vm_unmapped_area(&info);
1849

L
Linus Torvalds 已提交
1850 1851 1852 1853 1854 1855
	/*
	 * A failed mmap() very likely causes application failure,
	 * so fall back to the bottom-up function here. This scenario
	 * can happen with large stack limits and large mmap()
	 * allocations.
	 */
1856 1857 1858 1859 1860 1861 1862
	if (addr & ~PAGE_MASK) {
		VM_BUG_ON(addr != -ENOMEM);
		info.flags = 0;
		info.low_limit = TASK_UNMAPPED_BASE;
		info.high_limit = TASK_SIZE;
		addr = vm_unmapped_area(&info);
	}
L
Linus Torvalds 已提交
1863 1864 1865 1866 1867

	return addr;
}
#endif

1868
void arch_unmap_area_topdown(struct mm_struct *mm, unsigned long addr)
L
Linus Torvalds 已提交
1869 1870 1871 1872
{
	/*
	 * Is this a new hole at the highest possible address?
	 */
1873 1874
	if (addr > mm->free_area_cache)
		mm->free_area_cache = addr;
L
Linus Torvalds 已提交
1875 1876

	/* dont allow allocations above current base */
1877 1878
	if (mm->free_area_cache > mm->mmap_base)
		mm->free_area_cache = mm->mmap_base;
L
Linus Torvalds 已提交
1879 1880 1881 1882 1883 1884
}

unsigned long
get_unmapped_area(struct file *file, unsigned long addr, unsigned long len,
		unsigned long pgoff, unsigned long flags)
{
1885 1886 1887
	unsigned long (*get_area)(struct file *, unsigned long,
				  unsigned long, unsigned long, unsigned long);

1888 1889 1890 1891 1892 1893 1894 1895
	unsigned long error = arch_mmap_check(addr, len, flags);
	if (error)
		return error;

	/* Careful about overflows.. */
	if (len > TASK_SIZE)
		return -ENOMEM;

1896 1897 1898 1899 1900 1901
	get_area = current->mm->get_unmapped_area;
	if (file && file->f_op && file->f_op->get_unmapped_area)
		get_area = file->f_op->get_unmapped_area;
	addr = get_area(file, addr, len, pgoff, flags);
	if (IS_ERR_VALUE(addr))
		return addr;
L
Linus Torvalds 已提交
1902

1903 1904 1905 1906
	if (addr > TASK_SIZE - len)
		return -ENOMEM;
	if (addr & ~PAGE_MASK)
		return -EINVAL;
1907

1908 1909 1910
	addr = arch_rebalance_pgtables(addr, len);
	error = security_mmap_addr(addr);
	return error ? error : addr;
L
Linus Torvalds 已提交
1911 1912 1913 1914 1915
}

EXPORT_SYMBOL(get_unmapped_area);

/* Look up the first VMA which satisfies  addr < vm_end,  NULL if none. */
Z
ZhenwenXu 已提交
1916
struct vm_area_struct *find_vma(struct mm_struct *mm, unsigned long addr)
L
Linus Torvalds 已提交
1917 1918 1919
{
	struct vm_area_struct *vma = NULL;

1920 1921 1922 1923 1924 1925 1926 1927 1928 1929 1930 1931 1932 1933 1934 1935 1936 1937 1938 1939 1940 1941 1942 1943 1944
	if (WARN_ON_ONCE(!mm))		/* Remove this in linux-3.6 */
		return NULL;

	/* Check the cache first. */
	/* (Cache hit rate is typically around 35%.) */
	vma = mm->mmap_cache;
	if (!(vma && vma->vm_end > addr && vma->vm_start <= addr)) {
		struct rb_node *rb_node;

		rb_node = mm->mm_rb.rb_node;
		vma = NULL;

		while (rb_node) {
			struct vm_area_struct *vma_tmp;

			vma_tmp = rb_entry(rb_node,
					   struct vm_area_struct, vm_rb);

			if (vma_tmp->vm_end > addr) {
				vma = vma_tmp;
				if (vma_tmp->vm_start <= addr)
					break;
				rb_node = rb_node->rb_left;
			} else
				rb_node = rb_node->rb_right;
L
Linus Torvalds 已提交
1945
		}
1946 1947
		if (vma)
			mm->mmap_cache = vma;
L
Linus Torvalds 已提交
1948 1949 1950 1951 1952 1953
	}
	return vma;
}

EXPORT_SYMBOL(find_vma);

K
KOSAKI Motohiro 已提交
1954 1955 1956
/*
 * Same as find_vma, but also return a pointer to the previous VMA in *pprev.
 */
L
Linus Torvalds 已提交
1957 1958 1959 1960
struct vm_area_struct *
find_vma_prev(struct mm_struct *mm, unsigned long addr,
			struct vm_area_struct **pprev)
{
K
KOSAKI Motohiro 已提交
1961
	struct vm_area_struct *vma;
L
Linus Torvalds 已提交
1962

K
KOSAKI Motohiro 已提交
1963
	vma = find_vma(mm, addr);
M
Mikulas Patocka 已提交
1964 1965 1966 1967 1968 1969 1970 1971 1972 1973
	if (vma) {
		*pprev = vma->vm_prev;
	} else {
		struct rb_node *rb_node = mm->mm_rb.rb_node;
		*pprev = NULL;
		while (rb_node) {
			*pprev = rb_entry(rb_node, struct vm_area_struct, vm_rb);
			rb_node = rb_node->rb_right;
		}
	}
K
KOSAKI Motohiro 已提交
1974
	return vma;
L
Linus Torvalds 已提交
1975 1976 1977 1978 1979 1980 1981
}

/*
 * Verify that the stack growth is acceptable and
 * update accounting. This is shared with both the
 * grow-up and grow-down cases.
 */
Z
ZhenwenXu 已提交
1982
static int acct_stack_growth(struct vm_area_struct *vma, unsigned long size, unsigned long grow)
L
Linus Torvalds 已提交
1983 1984 1985
{
	struct mm_struct *mm = vma->vm_mm;
	struct rlimit *rlim = current->signal->rlim;
1986
	unsigned long new_start;
L
Linus Torvalds 已提交
1987 1988

	/* address space limit tests */
A
akpm@osdl.org 已提交
1989
	if (!may_expand_vm(mm, grow))
L
Linus Torvalds 已提交
1990 1991 1992
		return -ENOMEM;

	/* Stack limit test */
J
Jiri Slaby 已提交
1993
	if (size > ACCESS_ONCE(rlim[RLIMIT_STACK].rlim_cur))
L
Linus Torvalds 已提交
1994 1995 1996 1997 1998 1999 2000
		return -ENOMEM;

	/* mlock limit tests */
	if (vma->vm_flags & VM_LOCKED) {
		unsigned long locked;
		unsigned long limit;
		locked = mm->locked_vm + grow;
J
Jiri Slaby 已提交
2001 2002
		limit = ACCESS_ONCE(rlim[RLIMIT_MEMLOCK].rlim_cur);
		limit >>= PAGE_SHIFT;
L
Linus Torvalds 已提交
2003 2004 2005 2006
		if (locked > limit && !capable(CAP_IPC_LOCK))
			return -ENOMEM;
	}

2007 2008 2009 2010 2011 2012
	/* Check to ensure the stack will not grow into a hugetlb-only region */
	new_start = (vma->vm_flags & VM_GROWSUP) ? vma->vm_start :
			vma->vm_end - size;
	if (is_hugepage_only_range(vma->vm_mm, new_start, size))
		return -EFAULT;

L
Linus Torvalds 已提交
2013 2014 2015 2016
	/*
	 * Overcommit..  This must be the final test, as it will
	 * update security statistics.
	 */
2017
	if (security_vm_enough_memory_mm(mm, grow))
L
Linus Torvalds 已提交
2018 2019 2020 2021 2022
		return -ENOMEM;

	/* Ok, everything looks good - let it rip */
	if (vma->vm_flags & VM_LOCKED)
		mm->locked_vm += grow;
2023
	vm_stat_account(mm, vma->vm_flags, vma->vm_file, grow);
L
Linus Torvalds 已提交
2024 2025 2026
	return 0;
}

H
Hugh Dickins 已提交
2027
#if defined(CONFIG_STACK_GROWSUP) || defined(CONFIG_IA64)
L
Linus Torvalds 已提交
2028
/*
H
Hugh Dickins 已提交
2029 2030
 * PA-RISC uses this for its stack; IA64 for its Register Backing Store.
 * vma is the last one with address > vma->vm_end.  Have to extend vma.
L
Linus Torvalds 已提交
2031
 */
H
Hugh Dickins 已提交
2032
int expand_upwards(struct vm_area_struct *vma, unsigned long address)
L
Linus Torvalds 已提交
2033 2034 2035 2036 2037 2038 2039 2040 2041 2042 2043 2044
{
	int error;

	if (!(vma->vm_flags & VM_GROWSUP))
		return -EFAULT;

	/*
	 * We must make sure the anon_vma is allocated
	 * so that the anon_vma locking is not a noop.
	 */
	if (unlikely(anon_vma_prepare(vma)))
		return -ENOMEM;
2045
	vma_lock_anon_vma(vma);
L
Linus Torvalds 已提交
2046 2047 2048 2049 2050

	/*
	 * vma->vm_start/vm_end cannot change under us because the caller
	 * is required to hold the mmap_sem in read mode.  We need the
	 * anon_vma lock to serialize against concurrent expand_stacks.
2051
	 * Also guard against wrapping around to address 0.
L
Linus Torvalds 已提交
2052
	 */
2053 2054 2055
	if (address < PAGE_ALIGN(address+4))
		address = PAGE_ALIGN(address+4);
	else {
2056
		vma_unlock_anon_vma(vma);
2057 2058
		return -ENOMEM;
	}
L
Linus Torvalds 已提交
2059 2060 2061 2062 2063 2064 2065 2066 2067
	error = 0;

	/* Somebody else might have raced and expanded it already */
	if (address > vma->vm_end) {
		unsigned long size, grow;

		size = address - vma->vm_start;
		grow = (address - vma->vm_end) >> PAGE_SHIFT;

2068 2069 2070 2071
		error = -ENOMEM;
		if (vma->vm_pgoff + (size >> PAGE_SHIFT) >= vma->vm_pgoff) {
			error = acct_stack_growth(vma, size, grow);
			if (!error) {
2072 2073 2074 2075 2076 2077 2078 2079 2080 2081 2082 2083
				/*
				 * vma_gap_update() doesn't support concurrent
				 * updates, but we only hold a shared mmap_sem
				 * lock here, so we need to protect against
				 * concurrent vma expansions.
				 * vma_lock_anon_vma() doesn't help here, as
				 * we don't guarantee that all growable vmas
				 * in a mm share the same root anon vma.
				 * So, we reuse mm->page_table_lock to guard
				 * against concurrent vma expansions.
				 */
				spin_lock(&vma->vm_mm->page_table_lock);
2084
				anon_vma_interval_tree_pre_update_vma(vma);
2085
				vma->vm_end = address;
2086
				anon_vma_interval_tree_post_update_vma(vma);
2087 2088 2089 2090
				if (vma->vm_next)
					vma_gap_update(vma->vm_next);
				else
					vma->vm_mm->highest_vm_end = address;
2091 2092
				spin_unlock(&vma->vm_mm->page_table_lock);

2093 2094
				perf_event_mmap(vma);
			}
2095
		}
L
Linus Torvalds 已提交
2096
	}
2097
	vma_unlock_anon_vma(vma);
A
Andrea Arcangeli 已提交
2098
	khugepaged_enter_vma_merge(vma);
2099
	validate_mm(vma->vm_mm);
L
Linus Torvalds 已提交
2100 2101
	return error;
}
H
Hugh Dickins 已提交
2102 2103
#endif /* CONFIG_STACK_GROWSUP || CONFIG_IA64 */

L
Linus Torvalds 已提交
2104 2105 2106
/*
 * vma is the first one with address < vma->vm_start.  Have to extend vma.
 */
2107
int expand_downwards(struct vm_area_struct *vma,
2108
				   unsigned long address)
L
Linus Torvalds 已提交
2109 2110 2111 2112 2113 2114 2115 2116 2117
{
	int error;

	/*
	 * We must make sure the anon_vma is allocated
	 * so that the anon_vma locking is not a noop.
	 */
	if (unlikely(anon_vma_prepare(vma)))
		return -ENOMEM;
2118 2119

	address &= PAGE_MASK;
2120
	error = security_mmap_addr(address);
2121 2122 2123
	if (error)
		return error;

2124
	vma_lock_anon_vma(vma);
L
Linus Torvalds 已提交
2125 2126 2127 2128 2129 2130 2131 2132 2133 2134 2135 2136 2137 2138

	/*
	 * vma->vm_start/vm_end cannot change under us because the caller
	 * is required to hold the mmap_sem in read mode.  We need the
	 * anon_vma lock to serialize against concurrent expand_stacks.
	 */

	/* Somebody else might have raced and expanded it already */
	if (address < vma->vm_start) {
		unsigned long size, grow;

		size = vma->vm_end - address;
		grow = (vma->vm_start - address) >> PAGE_SHIFT;

2139 2140 2141 2142
		error = -ENOMEM;
		if (grow <= vma->vm_pgoff) {
			error = acct_stack_growth(vma, size, grow);
			if (!error) {
2143 2144 2145 2146 2147 2148 2149 2150 2151 2152 2153 2154
				/*
				 * vma_gap_update() doesn't support concurrent
				 * updates, but we only hold a shared mmap_sem
				 * lock here, so we need to protect against
				 * concurrent vma expansions.
				 * vma_lock_anon_vma() doesn't help here, as
				 * we don't guarantee that all growable vmas
				 * in a mm share the same root anon vma.
				 * So, we reuse mm->page_table_lock to guard
				 * against concurrent vma expansions.
				 */
				spin_lock(&vma->vm_mm->page_table_lock);
2155
				anon_vma_interval_tree_pre_update_vma(vma);
2156 2157
				vma->vm_start = address;
				vma->vm_pgoff -= grow;
2158
				anon_vma_interval_tree_post_update_vma(vma);
2159
				vma_gap_update(vma);
2160 2161
				spin_unlock(&vma->vm_mm->page_table_lock);

2162 2163
				perf_event_mmap(vma);
			}
L
Linus Torvalds 已提交
2164 2165
		}
	}
2166
	vma_unlock_anon_vma(vma);
A
Andrea Arcangeli 已提交
2167
	khugepaged_enter_vma_merge(vma);
2168
	validate_mm(vma->vm_mm);
L
Linus Torvalds 已提交
2169 2170 2171
	return error;
}

2172 2173 2174 2175 2176 2177 2178 2179 2180 2181 2182 2183 2184 2185 2186
#ifdef CONFIG_STACK_GROWSUP
int expand_stack(struct vm_area_struct *vma, unsigned long address)
{
	return expand_upwards(vma, address);
}

struct vm_area_struct *
find_extend_vma(struct mm_struct *mm, unsigned long addr)
{
	struct vm_area_struct *vma, *prev;

	addr &= PAGE_MASK;
	vma = find_vma_prev(mm, addr, &prev);
	if (vma && (vma->vm_start <= addr))
		return vma;
2187
	if (!prev || expand_stack(prev, addr))
2188
		return NULL;
2189
	if (prev->vm_flags & VM_LOCKED) {
2190
		mlock_vma_pages_range(prev, addr, prev->vm_end);
2191
	}
2192 2193 2194 2195 2196 2197 2198 2199
	return prev;
}
#else
int expand_stack(struct vm_area_struct *vma, unsigned long address)
{
	return expand_downwards(vma, address);
}

L
Linus Torvalds 已提交
2200 2201 2202 2203 2204 2205 2206 2207 2208 2209 2210 2211 2212 2213 2214 2215 2216
struct vm_area_struct *
find_extend_vma(struct mm_struct * mm, unsigned long addr)
{
	struct vm_area_struct * vma;
	unsigned long start;

	addr &= PAGE_MASK;
	vma = find_vma(mm,addr);
	if (!vma)
		return NULL;
	if (vma->vm_start <= addr)
		return vma;
	if (!(vma->vm_flags & VM_GROWSDOWN))
		return NULL;
	start = vma->vm_start;
	if (expand_stack(vma, addr))
		return NULL;
2217
	if (vma->vm_flags & VM_LOCKED) {
2218
		mlock_vma_pages_range(vma, addr, start);
2219
	}
L
Linus Torvalds 已提交
2220 2221 2222 2223 2224
	return vma;
}
#endif

/*
2225
 * Ok - we have the memory areas we should free on the vma list,
L
Linus Torvalds 已提交
2226
 * so release them, and do the vma updates.
2227 2228
 *
 * Called with the mm semaphore held.
L
Linus Torvalds 已提交
2229
 */
2230
static void remove_vma_list(struct mm_struct *mm, struct vm_area_struct *vma)
L
Linus Torvalds 已提交
2231
{
2232 2233
	unsigned long nr_accounted = 0;

2234 2235
	/* Update high watermark before we lower total_vm */
	update_hiwater_vm(mm);
L
Linus Torvalds 已提交
2236
	do {
2237 2238
		long nrpages = vma_pages(vma);

2239 2240
		if (vma->vm_flags & VM_ACCOUNT)
			nr_accounted += nrpages;
2241
		vm_stat_account(mm, vma->vm_flags, vma->vm_file, -nrpages);
2242
		vma = remove_vma(vma);
2243
	} while (vma);
2244
	vm_unacct_memory(nr_accounted);
L
Linus Torvalds 已提交
2245 2246 2247 2248 2249 2250
	validate_mm(mm);
}

/*
 * Get rid of page table information in the indicated region.
 *
2251
 * Called with the mm semaphore held.
L
Linus Torvalds 已提交
2252 2253
 */
static void unmap_region(struct mm_struct *mm,
2254 2255
		struct vm_area_struct *vma, struct vm_area_struct *prev,
		unsigned long start, unsigned long end)
L
Linus Torvalds 已提交
2256
{
2257
	struct vm_area_struct *next = prev? prev->vm_next: mm->mmap;
P
Peter Zijlstra 已提交
2258
	struct mmu_gather tlb;
L
Linus Torvalds 已提交
2259 2260

	lru_add_drain();
P
Peter Zijlstra 已提交
2261
	tlb_gather_mmu(&tlb, mm, 0);
2262
	update_hiwater_rss(mm);
2263
	unmap_vmas(&tlb, vma, start, end);
P
Peter Zijlstra 已提交
2264 2265 2266
	free_pgtables(&tlb, vma, prev ? prev->vm_end : FIRST_USER_ADDRESS,
				 next ? next->vm_start : 0);
	tlb_finish_mmu(&tlb, start, end);
L
Linus Torvalds 已提交
2267 2268 2269 2270 2271 2272 2273 2274 2275 2276 2277 2278
}

/*
 * Create a list of vma's touched by the unmap, removing them from the mm's
 * vma list as we go..
 */
static void
detach_vmas_to_be_unmapped(struct mm_struct *mm, struct vm_area_struct *vma,
	struct vm_area_struct *prev, unsigned long end)
{
	struct vm_area_struct **insertion_point;
	struct vm_area_struct *tail_vma = NULL;
2279
	unsigned long addr;
L
Linus Torvalds 已提交
2280 2281

	insertion_point = (prev ? &prev->vm_next : &mm->mmap);
2282
	vma->vm_prev = NULL;
L
Linus Torvalds 已提交
2283
	do {
2284
		vma_rb_erase(vma, &mm->mm_rb);
L
Linus Torvalds 已提交
2285 2286 2287 2288 2289
		mm->map_count--;
		tail_vma = vma;
		vma = vma->vm_next;
	} while (vma && vma->vm_start < end);
	*insertion_point = vma;
2290
	if (vma) {
2291
		vma->vm_prev = prev;
2292 2293 2294
		vma_gap_update(vma);
	} else
		mm->highest_vm_end = prev ? prev->vm_end : 0;
L
Linus Torvalds 已提交
2295
	tail_vma->vm_next = NULL;
2296 2297 2298 2299 2300
	if (mm->unmap_area == arch_unmap_area)
		addr = prev ? prev->vm_end : mm->mmap_base;
	else
		addr = vma ?  vma->vm_start : mm->mmap_base;
	mm->unmap_area(mm, addr);
L
Linus Torvalds 已提交
2301 2302 2303 2304
	mm->mmap_cache = NULL;		/* Kill the cache. */
}

/*
2305 2306
 * __split_vma() bypasses sysctl_max_map_count checking.  We use this on the
 * munmap path where it doesn't make sense to fail.
L
Linus Torvalds 已提交
2307
 */
2308
static int __split_vma(struct mm_struct * mm, struct vm_area_struct * vma,
L
Linus Torvalds 已提交
2309 2310 2311 2312
	      unsigned long addr, int new_below)
{
	struct mempolicy *pol;
	struct vm_area_struct *new;
2313
	int err = -ENOMEM;
L
Linus Torvalds 已提交
2314

2315 2316
	if (is_vm_hugetlb_page(vma) && (addr &
					~(huge_page_mask(hstate_vma(vma)))))
L
Linus Torvalds 已提交
2317 2318
		return -EINVAL;

2319
	new = kmem_cache_alloc(vm_area_cachep, GFP_KERNEL);
L
Linus Torvalds 已提交
2320
	if (!new)
2321
		goto out_err;
L
Linus Torvalds 已提交
2322 2323 2324 2325

	/* most fields are the same, copy all, and then fixup */
	*new = *vma;

2326 2327
	INIT_LIST_HEAD(&new->anon_vma_chain);

L
Linus Torvalds 已提交
2328 2329 2330 2331 2332 2333 2334
	if (new_below)
		new->vm_end = addr;
	else {
		new->vm_start = addr;
		new->vm_pgoff += ((addr - vma->vm_start) >> PAGE_SHIFT);
	}

2335
	pol = mpol_dup(vma_policy(vma));
L
Linus Torvalds 已提交
2336
	if (IS_ERR(pol)) {
2337 2338
		err = PTR_ERR(pol);
		goto out_free_vma;
L
Linus Torvalds 已提交
2339 2340 2341
	}
	vma_set_policy(new, pol);

2342 2343 2344
	if (anon_vma_clone(new, vma))
		goto out_free_mpol;

2345
	if (new->vm_file)
L
Linus Torvalds 已提交
2346 2347 2348 2349 2350 2351
		get_file(new->vm_file);

	if (new->vm_ops && new->vm_ops->open)
		new->vm_ops->open(new);

	if (new_below)
2352
		err = vma_adjust(vma, addr, vma->vm_end, vma->vm_pgoff +
L
Linus Torvalds 已提交
2353 2354
			((addr - new->vm_start) >> PAGE_SHIFT), new);
	else
2355
		err = vma_adjust(vma, vma->vm_start, addr, vma->vm_pgoff, new);
L
Linus Torvalds 已提交
2356

2357 2358 2359 2360 2361
	/* Success. */
	if (!err)
		return 0;

	/* Clean everything up if vma_adjust failed. */
2362 2363
	if (new->vm_ops && new->vm_ops->close)
		new->vm_ops->close(new);
2364
	if (new->vm_file)
2365
		fput(new->vm_file);
2366
	unlink_anon_vmas(new);
2367 2368 2369 2370 2371 2372
 out_free_mpol:
	mpol_put(pol);
 out_free_vma:
	kmem_cache_free(vm_area_cachep, new);
 out_err:
	return err;
L
Linus Torvalds 已提交
2373 2374
}

2375 2376 2377 2378 2379 2380 2381 2382 2383 2384 2385 2386 2387
/*
 * Split a vma into two pieces at address 'addr', a new vma is allocated
 * either for the first part or the tail.
 */
int split_vma(struct mm_struct *mm, struct vm_area_struct *vma,
	      unsigned long addr, int new_below)
{
	if (mm->map_count >= sysctl_max_map_count)
		return -ENOMEM;

	return __split_vma(mm, vma, addr, new_below);
}

L
Linus Torvalds 已提交
2388 2389 2390 2391 2392 2393 2394 2395
/* Munmap is split into 2 main parts -- this part which finds
 * what needs doing, and the areas themselves, which do the
 * work.  This now handles partial unmappings.
 * Jeremy Fitzhardinge <jeremy@goop.org>
 */
int do_munmap(struct mm_struct *mm, unsigned long start, size_t len)
{
	unsigned long end;
2396
	struct vm_area_struct *vma, *prev, *last;
L
Linus Torvalds 已提交
2397 2398 2399 2400 2401 2402 2403 2404

	if ((start & ~PAGE_MASK) || start > TASK_SIZE || len > TASK_SIZE-start)
		return -EINVAL;

	if ((len = PAGE_ALIGN(len)) == 0)
		return -EINVAL;

	/* Find the first overlapping VMA */
2405
	vma = find_vma(mm, start);
2406
	if (!vma)
L
Linus Torvalds 已提交
2407
		return 0;
2408
	prev = vma->vm_prev;
2409
	/* we have  start < vma->vm_end  */
L
Linus Torvalds 已提交
2410 2411 2412

	/* if it doesn't overlap, we have nothing.. */
	end = start + len;
2413
	if (vma->vm_start >= end)
L
Linus Torvalds 已提交
2414 2415 2416 2417 2418 2419 2420 2421 2422
		return 0;

	/*
	 * If we need to split any vma, do it now to save pain later.
	 *
	 * Note: mremap's move_vma VM_ACCOUNT handling assumes a partially
	 * unmapped vm_area_struct will remain in use: so lower split_vma
	 * places tmp vma above, and higher split_vma places tmp vma below.
	 */
2423
	if (start > vma->vm_start) {
2424 2425 2426 2427 2428 2429 2430 2431 2432 2433 2434
		int error;

		/*
		 * Make sure that map_count on return from munmap() will
		 * not exceed its limit; but let map_count go just above
		 * its limit temporarily, to help free resources as expected.
		 */
		if (end < vma->vm_end && mm->map_count >= sysctl_max_map_count)
			return -ENOMEM;

		error = __split_vma(mm, vma, start, 0);
L
Linus Torvalds 已提交
2435 2436
		if (error)
			return error;
2437
		prev = vma;
L
Linus Torvalds 已提交
2438 2439 2440 2441 2442
	}

	/* Does it split the last one? */
	last = find_vma(mm, end);
	if (last && end > last->vm_start) {
2443
		int error = __split_vma(mm, last, end, 1);
L
Linus Torvalds 已提交
2444 2445 2446
		if (error)
			return error;
	}
2447
	vma = prev? prev->vm_next: mm->mmap;
L
Linus Torvalds 已提交
2448

2449 2450 2451 2452 2453 2454 2455 2456 2457 2458 2459 2460 2461 2462
	/*
	 * unlock any mlock()ed ranges before detaching vmas
	 */
	if (mm->locked_vm) {
		struct vm_area_struct *tmp = vma;
		while (tmp && tmp->vm_start < end) {
			if (tmp->vm_flags & VM_LOCKED) {
				mm->locked_vm -= vma_pages(tmp);
				munlock_vma_pages_all(tmp);
			}
			tmp = tmp->vm_next;
		}
	}

L
Linus Torvalds 已提交
2463 2464 2465
	/*
	 * Remove the vma's, and unmap the actual pages
	 */
2466 2467
	detach_vmas_to_be_unmapped(mm, vma, prev, end);
	unmap_region(mm, vma, prev, start, end);
L
Linus Torvalds 已提交
2468 2469

	/* Fix up all other VM information */
2470
	remove_vma_list(mm, vma);
L
Linus Torvalds 已提交
2471 2472 2473 2474

	return 0;
}

A
Al Viro 已提交
2475
int vm_munmap(unsigned long start, size_t len)
L
Linus Torvalds 已提交
2476 2477
{
	int ret;
A
Al Viro 已提交
2478
	struct mm_struct *mm = current->mm;
L
Linus Torvalds 已提交
2479 2480

	down_write(&mm->mmap_sem);
2481
	ret = do_munmap(mm, start, len);
L
Linus Torvalds 已提交
2482 2483 2484
	up_write(&mm->mmap_sem);
	return ret;
}
2485 2486 2487 2488 2489
EXPORT_SYMBOL(vm_munmap);

SYSCALL_DEFINE2(munmap, unsigned long, addr, size_t, len)
{
	profile_munmap(addr);
A
Al Viro 已提交
2490
	return vm_munmap(addr, len);
2491
}
L
Linus Torvalds 已提交
2492 2493 2494

static inline void verify_mm_writelocked(struct mm_struct *mm)
{
2495
#ifdef CONFIG_DEBUG_VM
L
Linus Torvalds 已提交
2496 2497 2498 2499 2500 2501 2502 2503 2504 2505 2506 2507
	if (unlikely(down_read_trylock(&mm->mmap_sem))) {
		WARN_ON(1);
		up_read(&mm->mmap_sem);
	}
#endif
}

/*
 *  this is really a simplified "do_mmap".  it only handles
 *  anonymous maps.  eventually we may be able to do some
 *  brk-specific accounting here.
 */
2508
static unsigned long do_brk(unsigned long addr, unsigned long len)
L
Linus Torvalds 已提交
2509 2510 2511 2512 2513 2514
{
	struct mm_struct * mm = current->mm;
	struct vm_area_struct * vma, * prev;
	unsigned long flags;
	struct rb_node ** rb_link, * rb_parent;
	pgoff_t pgoff = addr >> PAGE_SHIFT;
2515
	int error;
L
Linus Torvalds 已提交
2516 2517 2518 2519 2520

	len = PAGE_ALIGN(len);
	if (!len)
		return addr;

2521 2522
	flags = VM_DATA_DEFAULT_FLAGS | VM_ACCOUNT | mm->def_flags;

A
Al Viro 已提交
2523 2524
	error = get_unmapped_area(NULL, addr, len, 0, MAP_FIXED);
	if (error & ~PAGE_MASK)
2525 2526
		return error;

L
Linus Torvalds 已提交
2527 2528 2529 2530 2531
	/*
	 * mlock MCL_FUTURE?
	 */
	if (mm->def_flags & VM_LOCKED) {
		unsigned long locked, lock_limit;
2532 2533
		locked = len >> PAGE_SHIFT;
		locked += mm->locked_vm;
J
Jiri Slaby 已提交
2534
		lock_limit = rlimit(RLIMIT_MEMLOCK);
2535
		lock_limit >>= PAGE_SHIFT;
L
Linus Torvalds 已提交
2536 2537 2538 2539 2540 2541 2542 2543 2544 2545 2546 2547 2548 2549
		if (locked > lock_limit && !capable(CAP_IPC_LOCK))
			return -EAGAIN;
	}

	/*
	 * mm->mmap_sem is required to protect against another thread
	 * changing the mappings in case we sleep.
	 */
	verify_mm_writelocked(mm);

	/*
	 * Clear old maps.  this also does some error checking for us
	 */
 munmap_back:
2550
	if (find_vma_links(mm, addr, addr + len, &prev, &rb_link, &rb_parent)) {
L
Linus Torvalds 已提交
2551 2552 2553 2554 2555 2556
		if (do_munmap(mm, addr, len))
			return -ENOMEM;
		goto munmap_back;
	}

	/* Check against address space limits *after* clearing old maps... */
A
akpm@osdl.org 已提交
2557
	if (!may_expand_vm(mm, len >> PAGE_SHIFT))
L
Linus Torvalds 已提交
2558 2559 2560 2561 2562
		return -ENOMEM;

	if (mm->map_count > sysctl_max_map_count)
		return -ENOMEM;

2563
	if (security_vm_enough_memory_mm(mm, len >> PAGE_SHIFT))
L
Linus Torvalds 已提交
2564 2565 2566
		return -ENOMEM;

	/* Can we just expand an old private anonymous mapping? */
2567 2568 2569
	vma = vma_merge(mm, prev, addr, addr + len, flags,
					NULL, NULL, pgoff, NULL);
	if (vma)
L
Linus Torvalds 已提交
2570 2571 2572 2573 2574
		goto out;

	/*
	 * create a vma struct for an anonymous mapping
	 */
P
Pekka Enberg 已提交
2575
	vma = kmem_cache_zalloc(vm_area_cachep, GFP_KERNEL);
L
Linus Torvalds 已提交
2576 2577 2578 2579 2580
	if (!vma) {
		vm_unacct_memory(len >> PAGE_SHIFT);
		return -ENOMEM;
	}

2581
	INIT_LIST_HEAD(&vma->anon_vma_chain);
L
Linus Torvalds 已提交
2582 2583 2584 2585 2586
	vma->vm_mm = mm;
	vma->vm_start = addr;
	vma->vm_end = addr + len;
	vma->vm_pgoff = pgoff;
	vma->vm_flags = flags;
2587
	vma->vm_page_prot = vm_get_page_prot(flags);
L
Linus Torvalds 已提交
2588 2589
	vma_link(mm, vma, prev, rb_link, rb_parent);
out:
2590
	perf_event_mmap(vma);
L
Linus Torvalds 已提交
2591 2592
	mm->total_vm += len >> PAGE_SHIFT;
	if (flags & VM_LOCKED) {
2593 2594
		if (!mlock_vma_pages_range(vma, addr, addr + len))
			mm->locked_vm += (len >> PAGE_SHIFT);
L
Linus Torvalds 已提交
2595 2596 2597 2598
	}
	return addr;
}

2599 2600 2601 2602 2603 2604 2605 2606 2607 2608 2609
unsigned long vm_brk(unsigned long addr, unsigned long len)
{
	struct mm_struct *mm = current->mm;
	unsigned long ret;

	down_write(&mm->mmap_sem);
	ret = do_brk(addr, len);
	up_write(&mm->mmap_sem);
	return ret;
}
EXPORT_SYMBOL(vm_brk);
L
Linus Torvalds 已提交
2610 2611 2612 2613

/* Release all mmaps. */
void exit_mmap(struct mm_struct *mm)
{
P
Peter Zijlstra 已提交
2614
	struct mmu_gather tlb;
2615
	struct vm_area_struct *vma;
L
Linus Torvalds 已提交
2616 2617
	unsigned long nr_accounted = 0;

2618
	/* mm's last user has gone, and its about to be pulled down */
A
Andrea Arcangeli 已提交
2619
	mmu_notifier_release(mm);
2620

2621 2622 2623 2624 2625 2626 2627 2628
	if (mm->locked_vm) {
		vma = mm->mmap;
		while (vma) {
			if (vma->vm_flags & VM_LOCKED)
				munlock_vma_pages_all(vma);
			vma = vma->vm_next;
		}
	}
2629 2630 2631

	arch_exit_mmap(mm);

2632
	vma = mm->mmap;
2633 2634 2635
	if (!vma)	/* Can happen if dup_mmap() received an OOM */
		return;

L
Linus Torvalds 已提交
2636 2637
	lru_add_drain();
	flush_cache_mm(mm);
P
Peter Zijlstra 已提交
2638
	tlb_gather_mmu(&tlb, mm, 1);
2639
	/* update_hiwater_rss(mm) here? but nobody should be looking */
2640
	/* Use -1 here to ensure all VMAs in the mm are unmapped */
2641
	unmap_vmas(&tlb, vma, 0, -1);
H
Hugh Dickins 已提交
2642

P
Peter Zijlstra 已提交
2643
	free_pgtables(&tlb, vma, FIRST_USER_ADDRESS, 0);
2644
	tlb_finish_mmu(&tlb, 0, -1);
L
Linus Torvalds 已提交
2645 2646

	/*
2647 2648
	 * Walk the list again, actually closing and freeing it,
	 * with preemption enabled, without holding any MM locks.
L
Linus Torvalds 已提交
2649
	 */
2650 2651 2652
	while (vma) {
		if (vma->vm_flags & VM_ACCOUNT)
			nr_accounted += vma_pages(vma);
2653
		vma = remove_vma(vma);
2654 2655
	}
	vm_unacct_memory(nr_accounted);
2656

2657
	WARN_ON(mm->nr_ptes > (FIRST_USER_ADDRESS+PMD_SIZE-1)>>PMD_SHIFT);
L
Linus Torvalds 已提交
2658 2659 2660 2661
}

/* Insert vm structure into process list sorted by address
 * and into the inode's i_mmap tree.  If vm_file is non-NULL
2662
 * then i_mmap_mutex is taken here.
L
Linus Torvalds 已提交
2663
 */
2664
int insert_vm_struct(struct mm_struct *mm, struct vm_area_struct *vma)
L
Linus Torvalds 已提交
2665
{
2666 2667
	struct vm_area_struct *prev;
	struct rb_node **rb_link, *rb_parent;
L
Linus Torvalds 已提交
2668 2669 2670 2671 2672 2673 2674 2675 2676 2677 2678 2679 2680 2681 2682 2683 2684

	/*
	 * The vm_pgoff of a purely anonymous vma should be irrelevant
	 * until its first write fault, when page's anon_vma and index
	 * are set.  But now set the vm_pgoff it will almost certainly
	 * end up with (unless mremap moves it elsewhere before that
	 * first wfault), so /proc/pid/maps tells a consistent story.
	 *
	 * By setting it to reflect the virtual start address of the
	 * vma, merges and splits can happen in a seamless way, just
	 * using the existing file pgoff checks and manipulations.
	 * Similarly in do_mmap_pgoff and in do_brk.
	 */
	if (!vma->vm_file) {
		BUG_ON(vma->anon_vma);
		vma->vm_pgoff = vma->vm_start >> PAGE_SHIFT;
	}
2685 2686
	if (find_vma_links(mm, vma->vm_start, vma->vm_end,
			   &prev, &rb_link, &rb_parent))
L
Linus Torvalds 已提交
2687
		return -ENOMEM;
2688
	if ((vma->vm_flags & VM_ACCOUNT) &&
2689
	     security_vm_enough_memory_mm(mm, vma_pages(vma)))
2690
		return -ENOMEM;
2691

L
Linus Torvalds 已提交
2692 2693 2694 2695 2696 2697 2698 2699 2700
	vma_link(mm, vma, prev, rb_link, rb_parent);
	return 0;
}

/*
 * Copy the vma structure to a new location in the same mm,
 * prior to moving page table entries, to effect an mremap move.
 */
struct vm_area_struct *copy_vma(struct vm_area_struct **vmap,
2701 2702
	unsigned long addr, unsigned long len, pgoff_t pgoff,
	bool *need_rmap_locks)
L
Linus Torvalds 已提交
2703 2704 2705 2706 2707 2708 2709
{
	struct vm_area_struct *vma = *vmap;
	unsigned long vma_start = vma->vm_start;
	struct mm_struct *mm = vma->vm_mm;
	struct vm_area_struct *new_vma, *prev;
	struct rb_node **rb_link, *rb_parent;
	struct mempolicy *pol;
2710
	bool faulted_in_anon_vma = true;
L
Linus Torvalds 已提交
2711 2712 2713 2714 2715

	/*
	 * If anonymous vma has not yet been faulted, update new pgoff
	 * to match new location, to increase its chance of merging.
	 */
2716
	if (unlikely(!vma->vm_file && !vma->anon_vma)) {
L
Linus Torvalds 已提交
2717
		pgoff = addr >> PAGE_SHIFT;
2718 2719
		faulted_in_anon_vma = false;
	}
L
Linus Torvalds 已提交
2720

2721 2722
	if (find_vma_links(mm, addr, addr + len, &prev, &rb_link, &rb_parent))
		return NULL;	/* should never get here */
L
Linus Torvalds 已提交
2723 2724 2725 2726 2727 2728
	new_vma = vma_merge(mm, prev, addr, addr + len, vma->vm_flags,
			vma->anon_vma, vma->vm_file, pgoff, vma_policy(vma));
	if (new_vma) {
		/*
		 * Source vma may have been merged into new_vma
		 */
2729 2730 2731 2732 2733 2734 2735 2736 2737 2738 2739 2740 2741 2742 2743
		if (unlikely(vma_start >= new_vma->vm_start &&
			     vma_start < new_vma->vm_end)) {
			/*
			 * The only way we can get a vma_merge with
			 * self during an mremap is if the vma hasn't
			 * been faulted in yet and we were allowed to
			 * reset the dst vma->vm_pgoff to the
			 * destination address of the mremap to allow
			 * the merge to happen. mremap must change the
			 * vm_pgoff linearity between src and dst vmas
			 * (in turn preventing a vma_merge) to be
			 * safe. It is only safe to keep the vm_pgoff
			 * linear if there are no pages mapped yet.
			 */
			VM_BUG_ON(faulted_in_anon_vma);
2744
			*vmap = vma = new_vma;
2745
		}
2746
		*need_rmap_locks = (new_vma->vm_pgoff <= vma->vm_pgoff);
L
Linus Torvalds 已提交
2747
	} else {
2748
		new_vma = kmem_cache_alloc(vm_area_cachep, GFP_KERNEL);
L
Linus Torvalds 已提交
2749 2750
		if (new_vma) {
			*new_vma = *vma;
2751 2752 2753
			new_vma->vm_start = addr;
			new_vma->vm_end = addr + len;
			new_vma->vm_pgoff = pgoff;
2754
			pol = mpol_dup(vma_policy(vma));
2755 2756
			if (IS_ERR(pol))
				goto out_free_vma;
2757
			vma_set_policy(new_vma, pol);
2758 2759 2760
			INIT_LIST_HEAD(&new_vma->anon_vma_chain);
			if (anon_vma_clone(new_vma, vma))
				goto out_free_mempol;
2761
			if (new_vma->vm_file)
L
Linus Torvalds 已提交
2762 2763 2764 2765
				get_file(new_vma->vm_file);
			if (new_vma->vm_ops && new_vma->vm_ops->open)
				new_vma->vm_ops->open(new_vma);
			vma_link(mm, new_vma, prev, rb_link, rb_parent);
2766
			*need_rmap_locks = false;
L
Linus Torvalds 已提交
2767 2768 2769
		}
	}
	return new_vma;
2770 2771 2772 2773 2774 2775

 out_free_mempol:
	mpol_put(pol);
 out_free_vma:
	kmem_cache_free(vm_area_cachep, new_vma);
	return NULL;
L
Linus Torvalds 已提交
2776
}
A
akpm@osdl.org 已提交
2777 2778 2779 2780 2781 2782 2783 2784 2785 2786

/*
 * Return true if the calling process may expand its vm space by the passed
 * number of pages
 */
int may_expand_vm(struct mm_struct *mm, unsigned long npages)
{
	unsigned long cur = mm->total_vm;	/* pages */
	unsigned long lim;

J
Jiri Slaby 已提交
2787
	lim = rlimit(RLIMIT_AS) >> PAGE_SHIFT;
A
akpm@osdl.org 已提交
2788 2789 2790 2791 2792

	if (cur + npages > lim)
		return 0;
	return 1;
}
2793 2794


N
Nick Piggin 已提交
2795 2796
static int special_mapping_fault(struct vm_area_struct *vma,
				struct vm_fault *vmf)
2797
{
N
Nick Piggin 已提交
2798
	pgoff_t pgoff;
2799 2800
	struct page **pages;

N
Nick Piggin 已提交
2801 2802 2803 2804 2805 2806 2807
	/*
	 * special mappings have no vm_file, and in that case, the mm
	 * uses vm_pgoff internally. So we have to subtract it from here.
	 * We are allowed to do this because we are the mm; do not copy
	 * this code into drivers!
	 */
	pgoff = vmf->pgoff - vma->vm_pgoff;
2808

N
Nick Piggin 已提交
2809 2810
	for (pages = vma->vm_private_data; pgoff && *pages; ++pages)
		pgoff--;
2811 2812 2813 2814

	if (*pages) {
		struct page *page = *pages;
		get_page(page);
N
Nick Piggin 已提交
2815 2816
		vmf->page = page;
		return 0;
2817 2818
	}

N
Nick Piggin 已提交
2819
	return VM_FAULT_SIGBUS;
2820 2821 2822 2823 2824 2825 2826 2827 2828
}

/*
 * Having a close hook prevents vma merging regardless of flags.
 */
static void special_mapping_close(struct vm_area_struct *vma)
{
}

2829
static const struct vm_operations_struct special_mapping_vmops = {
2830
	.close = special_mapping_close,
N
Nick Piggin 已提交
2831
	.fault = special_mapping_fault,
2832 2833 2834 2835 2836 2837 2838 2839 2840 2841 2842 2843 2844 2845 2846
};

/*
 * Called with mm->mmap_sem held for writing.
 * Insert a new vma covering the given region, with the given flags.
 * Its pages are supplied by the given array of struct page *.
 * The array can be shorter than len >> PAGE_SHIFT if it's null-terminated.
 * The region past the last page supplied will always produce SIGBUS.
 * The array pointer and the pages it points to are assumed to stay alive
 * for as long as this mapping might exist.
 */
int install_special_mapping(struct mm_struct *mm,
			    unsigned long addr, unsigned long len,
			    unsigned long vm_flags, struct page **pages)
{
2847
	int ret;
2848 2849 2850 2851 2852 2853
	struct vm_area_struct *vma;

	vma = kmem_cache_zalloc(vm_area_cachep, GFP_KERNEL);
	if (unlikely(vma == NULL))
		return -ENOMEM;

2854
	INIT_LIST_HEAD(&vma->anon_vma_chain);
2855 2856 2857 2858
	vma->vm_mm = mm;
	vma->vm_start = addr;
	vma->vm_end = addr + len;

2859
	vma->vm_flags = vm_flags | mm->def_flags | VM_DONTEXPAND;
2860
	vma->vm_page_prot = vm_get_page_prot(vma->vm_flags);
2861 2862 2863 2864

	vma->vm_ops = &special_mapping_vmops;
	vma->vm_private_data = pages;

2865 2866 2867
	ret = insert_vm_struct(mm, vma);
	if (ret)
		goto out;
2868 2869 2870

	mm->total_vm += len >> PAGE_SHIFT;

2871
	perf_event_mmap(vma);
2872

2873
	return 0;
2874 2875 2876 2877

out:
	kmem_cache_free(vm_area_cachep, vma);
	return ret;
2878
}
2879 2880 2881

static DEFINE_MUTEX(mm_all_locks_mutex);

2882
static void vm_lock_anon_vma(struct mm_struct *mm, struct anon_vma *anon_vma)
2883
{
2884
	if (!test_bit(0, (unsigned long *) &anon_vma->root->rb_root.rb_node)) {
2885 2886 2887 2888
		/*
		 * The LSB of head.next can't change from under us
		 * because we hold the mm_all_locks_mutex.
		 */
2889
		down_write_nest_lock(&anon_vma->root->rwsem, &mm->mmap_sem);
2890 2891
		/*
		 * We can safely modify head.next after taking the
2892
		 * anon_vma->root->rwsem. If some other vma in this mm shares
2893 2894 2895 2896
		 * the same anon_vma we won't take it again.
		 *
		 * No need of atomic instructions here, head.next
		 * can't change from under us thanks to the
2897
		 * anon_vma->root->rwsem.
2898 2899
		 */
		if (__test_and_set_bit(0, (unsigned long *)
2900
				       &anon_vma->root->rb_root.rb_node))
2901 2902 2903 2904
			BUG();
	}
}

2905
static void vm_lock_mapping(struct mm_struct *mm, struct address_space *mapping)
2906 2907 2908 2909 2910 2911 2912 2913 2914 2915 2916 2917 2918
{
	if (!test_bit(AS_MM_ALL_LOCKS, &mapping->flags)) {
		/*
		 * AS_MM_ALL_LOCKS can't change from under us because
		 * we hold the mm_all_locks_mutex.
		 *
		 * Operations on ->flags have to be atomic because
		 * even if AS_MM_ALL_LOCKS is stable thanks to the
		 * mm_all_locks_mutex, there may be other cpus
		 * changing other bitflags in parallel to us.
		 */
		if (test_and_set_bit(AS_MM_ALL_LOCKS, &mapping->flags))
			BUG();
2919
		mutex_lock_nest_lock(&mapping->i_mmap_mutex, &mm->mmap_sem);
2920 2921 2922 2923 2924 2925 2926 2927 2928 2929 2930 2931 2932 2933 2934 2935 2936 2937 2938 2939 2940
	}
}

/*
 * This operation locks against the VM for all pte/vma/mm related
 * operations that could ever happen on a certain mm. This includes
 * vmtruncate, try_to_unmap, and all page faults.
 *
 * The caller must take the mmap_sem in write mode before calling
 * mm_take_all_locks(). The caller isn't allowed to release the
 * mmap_sem until mm_drop_all_locks() returns.
 *
 * mmap_sem in write mode is required in order to block all operations
 * that could modify pagetables and free pages without need of
 * altering the vma layout (for example populate_range() with
 * nonlinear vmas). It's also needed in write mode to avoid new
 * anon_vmas to be associated with existing vmas.
 *
 * A single task can't take more than one mm_take_all_locks() in a row
 * or it would deadlock.
 *
2941
 * The LSB in anon_vma->rb_root.rb_node and the AS_MM_ALL_LOCKS bitflag in
2942 2943 2944 2945
 * mapping->flags avoid to take the same lock twice, if more than one
 * vma in this mm is backed by the same anon_vma or address_space.
 *
 * We can take all the locks in random order because the VM code
2946
 * taking i_mmap_mutex or anon_vma->mutex outside the mmap_sem never
2947 2948 2949 2950 2951 2952 2953 2954 2955 2956 2957
 * takes more than one of them in a row. Secondly we're protected
 * against a concurrent mm_take_all_locks() by the mm_all_locks_mutex.
 *
 * mm_take_all_locks() and mm_drop_all_locks are expensive operations
 * that may have to take thousand of locks.
 *
 * mm_take_all_locks() can fail if it's interrupted by signals.
 */
int mm_take_all_locks(struct mm_struct *mm)
{
	struct vm_area_struct *vma;
2958
	struct anon_vma_chain *avc;
2959 2960 2961 2962 2963 2964 2965 2966 2967

	BUG_ON(down_read_trylock(&mm->mmap_sem));

	mutex_lock(&mm_all_locks_mutex);

	for (vma = mm->mmap; vma; vma = vma->vm_next) {
		if (signal_pending(current))
			goto out_unlock;
		if (vma->vm_file && vma->vm_file->f_mapping)
2968
			vm_lock_mapping(mm, vma->vm_file->f_mapping);
2969
	}
2970 2971 2972 2973 2974

	for (vma = mm->mmap; vma; vma = vma->vm_next) {
		if (signal_pending(current))
			goto out_unlock;
		if (vma->anon_vma)
2975 2976
			list_for_each_entry(avc, &vma->anon_vma_chain, same_vma)
				vm_lock_anon_vma(mm, avc->anon_vma);
2977
	}
2978

2979
	return 0;
2980 2981

out_unlock:
2982 2983
	mm_drop_all_locks(mm);
	return -EINTR;
2984 2985 2986 2987
}

static void vm_unlock_anon_vma(struct anon_vma *anon_vma)
{
2988
	if (test_bit(0, (unsigned long *) &anon_vma->root->rb_root.rb_node)) {
2989 2990 2991 2992 2993
		/*
		 * The LSB of head.next can't change to 0 from under
		 * us because we hold the mm_all_locks_mutex.
		 *
		 * We must however clear the bitflag before unlocking
2994
		 * the vma so the users using the anon_vma->rb_root will
2995 2996 2997 2998
		 * never see our bitflag.
		 *
		 * No need of atomic instructions here, head.next
		 * can't change from under us until we release the
2999
		 * anon_vma->root->rwsem.
3000 3001
		 */
		if (!__test_and_clear_bit(0, (unsigned long *)
3002
					  &anon_vma->root->rb_root.rb_node))
3003
			BUG();
3004
		anon_vma_unlock(anon_vma);
3005 3006 3007 3008 3009 3010 3011 3012 3013 3014
	}
}

static void vm_unlock_mapping(struct address_space *mapping)
{
	if (test_bit(AS_MM_ALL_LOCKS, &mapping->flags)) {
		/*
		 * AS_MM_ALL_LOCKS can't change to 0 from under us
		 * because we hold the mm_all_locks_mutex.
		 */
3015
		mutex_unlock(&mapping->i_mmap_mutex);
3016 3017 3018 3019 3020 3021 3022 3023 3024 3025 3026 3027 3028
		if (!test_and_clear_bit(AS_MM_ALL_LOCKS,
					&mapping->flags))
			BUG();
	}
}

/*
 * The mmap_sem cannot be released by the caller until
 * mm_drop_all_locks() returns.
 */
void mm_drop_all_locks(struct mm_struct *mm)
{
	struct vm_area_struct *vma;
3029
	struct anon_vma_chain *avc;
3030 3031 3032 3033 3034 3035

	BUG_ON(down_read_trylock(&mm->mmap_sem));
	BUG_ON(!mutex_is_locked(&mm_all_locks_mutex));

	for (vma = mm->mmap; vma; vma = vma->vm_next) {
		if (vma->anon_vma)
3036 3037
			list_for_each_entry(avc, &vma->anon_vma_chain, same_vma)
				vm_unlock_anon_vma(avc->anon_vma);
3038 3039 3040 3041 3042 3043
		if (vma->vm_file && vma->vm_file->f_mapping)
			vm_unlock_mapping(vma->vm_file->f_mapping);
	}

	mutex_unlock(&mm_all_locks_mutex);
}
3044 3045 3046 3047 3048 3049

/*
 * initialise the VMA slab
 */
void __init mmap_init(void)
{
3050 3051 3052 3053
	int ret;

	ret = percpu_counter_init(&vm_committed_as, 0);
	VM_BUG_ON(ret);
3054
}