mmap.c 70.1 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>
L
Linus Torvalds 已提交
34 35 36 37

#include <asm/uaccess.h>
#include <asm/cacheflush.h>
#include <asm/tlb.h>
38
#include <asm/mmu_context.h>
L
Linus Torvalds 已提交
39

40 41
#include "internal.h"

42 43 44 45
#ifndef arch_mmap_check
#define arch_mmap_check(addr, len, flags)	(0)
#endif

46 47 48 49
#ifndef arch_rebalance_pgtables
#define arch_rebalance_pgtables(addr, len)		(addr)
#endif

50 51 52 53
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
54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79
/*
 * WARNING: the debugging will use recursive algorithms so never enable this
 * unless you know what you are doing.
 */
#undef DEBUG_MM_RB

/* 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
};

80 81
pgprot_t vm_get_page_prot(unsigned long vm_flags)
{
82 83 84
	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)));
85 86 87
}
EXPORT_SYMBOL(vm_get_page_prot);

88 89
int sysctl_overcommit_memory __read_mostly = OVERCOMMIT_GUESS;  /* heuristic overcommit */
int sysctl_overcommit_ratio __read_mostly = 50;	/* default is 50% */
90
int sysctl_max_map_count __read_mostly = DEFAULT_MAX_MAP_COUNT;
91 92 93 94 95
/*
 * 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
96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112

/*
 * 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.
 */
113
int __vm_enough_memory(struct mm_struct *mm, long pages, int cap_sys_admin)
L
Linus Torvalds 已提交
114 115 116 117 118 119 120 121 122 123 124 125
{
	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) {
126 127 128 129 130 131 132 133 134 135
		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
136 137 138 139 140 141 142 143 144

		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
		 */
145
		free += global_page_state(NR_SLAB_RECLAIMABLE);
L
Linus Torvalds 已提交
146

147 148 149
		/*
		 * Leave reserved pages. The pages are not for anonymous pages.
		 */
150
		if (free <= totalreserve_pages)
151 152
			goto error;
		else
153
			free -= totalreserve_pages;
154 155 156 157

		/*
		 * Leave the last 3% for root
		 */
L
Linus Torvalds 已提交
158
		if (!cap_sys_admin)
159
			free -= free / 32;
L
Linus Torvalds 已提交
160 161 162

		if (free > pages)
			return 0;
163 164

		goto error;
L
Linus Torvalds 已提交
165 166 167 168 169 170 171 172 173 174 175 176 177
	}

	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
178 179
	if (mm)
		allowed -= mm->total_vm / 32;
L
Linus Torvalds 已提交
180

181
	if (percpu_counter_read_positive(&vm_committed_as) < allowed)
L
Linus Torvalds 已提交
182
		return 0;
183
error:
L
Linus Torvalds 已提交
184 185 186 187 188 189
	vm_unacct_memory(pages);

	return -ENOMEM;
}

/*
190
 * Requires inode->i_mapping->i_mmap_mutex
L
Linus Torvalds 已提交
191 192 193 194 195
 */
static void __remove_shared_vm_struct(struct vm_area_struct *vma,
		struct file *file, struct address_space *mapping)
{
	if (vma->vm_flags & VM_DENYWRITE)
196
		atomic_inc(&file->f_path.dentry->d_inode->i_writecount);
L
Linus Torvalds 已提交
197 198 199 200 201 202 203 204 205 206 207 208
	if (vma->vm_flags & VM_SHARED)
		mapping->i_mmap_writable--;

	flush_dcache_mmap_lock(mapping);
	if (unlikely(vma->vm_flags & VM_NONLINEAR))
		list_del_init(&vma->shared.vm_set.list);
	else
		vma_prio_tree_remove(vma, &mapping->i_mmap);
	flush_dcache_mmap_unlock(mapping);
}

/*
209 210
 * Unlink a file-based vm structure from its prio_tree, to hide
 * vma from rmap and vmtruncate before freeing its page tables.
L
Linus Torvalds 已提交
211
 */
212
void unlink_file_vma(struct vm_area_struct *vma)
L
Linus Torvalds 已提交
213 214 215 216 217
{
	struct file *file = vma->vm_file;

	if (file) {
		struct address_space *mapping = file->f_mapping;
218
		mutex_lock(&mapping->i_mmap_mutex);
L
Linus Torvalds 已提交
219
		__remove_shared_vm_struct(vma, file, mapping);
220
		mutex_unlock(&mapping->i_mmap_mutex);
L
Linus Torvalds 已提交
221
	}
222 223 224 225 226 227 228 229 230 231
}

/*
 * 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
232 233
	if (vma->vm_ops && vma->vm_ops->close)
		vma->vm_ops->close(vma);
234
	if (vma->vm_file)
235
		fput(vma->vm_file);
236
	mpol_put(vma_policy(vma));
L
Linus Torvalds 已提交
237
	kmem_cache_free(vm_area_cachep, vma);
238
	return next;
L
Linus Torvalds 已提交
239 240
}

241 242
static unsigned long do_brk(unsigned long addr, unsigned long len);

243
SYSCALL_DEFINE1(brk, unsigned long, brk)
L
Linus Torvalds 已提交
244 245 246 247
{
	unsigned long rlim, retval;
	unsigned long newbrk, oldbrk;
	struct mm_struct *mm = current->mm;
248
	unsigned long min_brk;
L
Linus Torvalds 已提交
249 250 251

	down_write(&mm->mmap_sem);

252
#ifdef CONFIG_COMPAT_BRK
253 254 255 256 257
	/*
	 * 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
	 */
258
	if (current->brk_randomized)
259 260 261
		min_brk = mm->start_brk;
	else
		min_brk = mm->end_data;
262 263 264 265
#else
	min_brk = mm->start_brk;
#endif
	if (brk < min_brk)
L
Linus Torvalds 已提交
266
		goto out;
R
Ram Gupta 已提交
267 268 269 270 271 272 273

	/*
	 * 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
274
	rlim = rlimit(RLIMIT_DATA);
J
Jiri Kosina 已提交
275 276
	if (rlim < RLIM_INFINITY && (brk - mm->start_brk) +
			(mm->end_data - mm->start_data) > rlim)
R
Ram Gupta 已提交
277 278
		goto out;

L
Linus Torvalds 已提交
279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323
	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;
}

#ifdef DEBUG_MM_RB
static int browse_rb(struct rb_root *root)
{
	int i = 0, j;
	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);
		if (vma->vm_start < prev)
			printk("vm_start %lx prev %lx\n", vma->vm_start, prev), i = -1;
		if (vma->vm_start < pend)
			printk("vm_start %lx pend %lx\n", vma->vm_start, pend);
		if (vma->vm_start > vma->vm_end)
			printk("vm_end %lx < vm_start %lx\n", vma->vm_end, vma->vm_start);
		i++;
		pn = nd;
D
David Miller 已提交
324 325
		prev = vma->vm_start;
		pend = vma->vm_end;
L
Linus Torvalds 已提交
326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349
	}
	j = 0;
	for (nd = pn; nd; nd = rb_prev(nd)) {
		j++;
	}
	if (i != j)
		printk("backwards %d, forwards %d\n", j, i), i = 0;
	return i;
}

void validate_mm(struct mm_struct *mm)
{
	int bug = 0;
	int i = 0;
	struct vm_area_struct *tmp = mm->mmap;
	while (tmp) {
		tmp = tmp->vm_next;
		i++;
	}
	if (i != mm->map_count)
		printk("map_count %d vm_next %d\n", mm->map_count, i), bug = 1;
	i = browse_rb(&mm->mm_rb);
	if (i != mm->map_count)
		printk("map_count %d rb %d\n", mm->map_count, i), bug = 1;
350
	BUG_ON(bug);
L
Linus Torvalds 已提交
351 352 353 354 355
}
#else
#define validate_mm(mm) do { } while (0)
#endif

356 357 358
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
359
{
360
	struct rb_node **__rb_link, *__rb_parent, *rb_prev;
L
Linus Torvalds 已提交
361 362 363 364 365 366 367 368 369 370 371

	__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) {
372 373 374
			/* Fail if an existing vma overlaps the area */
			if (vma_tmp->vm_start < end)
				return -ENOMEM;
L
Linus Torvalds 已提交
375 376 377 378 379 380 381 382 383 384 385 386
			__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;
387
	return 0;
L
Linus Torvalds 已提交
388 389 390 391 392 393 394 395 396
}

void __vma_link_rb(struct mm_struct *mm, struct vm_area_struct *vma,
		struct rb_node **rb_link, struct rb_node *rb_parent)
{
	rb_link_node(&vma->vm_rb, rb_parent, rb_link);
	rb_insert_color(&vma->vm_rb, &mm->mm_rb);
}

D
Denys Vlasenko 已提交
397
static void __vma_link_file(struct vm_area_struct *vma)
L
Linus Torvalds 已提交
398
{
Z
ZhenwenXu 已提交
399
	struct file *file;
L
Linus Torvalds 已提交
400 401 402 403 404 405

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

		if (vma->vm_flags & VM_DENYWRITE)
406
			atomic_dec(&file->f_path.dentry->d_inode->i_writecount);
L
Linus Torvalds 已提交
407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436
		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
			vma_prio_tree_insert(vma, &mapping->i_mmap);
		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;

437
	if (mapping)
438
		mutex_lock(&mapping->i_mmap_mutex);
L
Linus Torvalds 已提交
439 440 441 442 443

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

	if (mapping)
444
		mutex_unlock(&mapping->i_mmap_mutex);
L
Linus Torvalds 已提交
445 446 447 448 449 450

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

/*
451 452
 * Helper for vma_adjust() in the split_vma insert case: insert a vma into the
 * mm's list and rbtree.  It has already been inserted into the prio_tree.
L
Linus Torvalds 已提交
453
 */
Z
ZhenwenXu 已提交
454
static void __insert_vm_struct(struct mm_struct *mm, struct vm_area_struct *vma)
L
Linus Torvalds 已提交
455
{
456
	struct vm_area_struct *prev;
Z
ZhenwenXu 已提交
457
	struct rb_node **rb_link, *rb_parent;
L
Linus Torvalds 已提交
458

459 460 461
	if (find_vma_links(mm, vma->vm_start, vma->vm_end,
			   &prev, &rb_link, &rb_parent))
		BUG();
L
Linus Torvalds 已提交
462 463 464 465 466 467 468 469
	__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)
{
470 471 472 473 474
	struct vm_area_struct *next = vma->vm_next;

	prev->vm_next = next;
	if (next)
		next->vm_prev = prev;
L
Linus Torvalds 已提交
475 476 477 478 479 480 481 482 483 484 485 486
	rb_erase(&vma->vm_rb, &mm->mm_rb);
	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.
 */
487
int vma_adjust(struct vm_area_struct *vma, unsigned long start,
L
Linus Torvalds 已提交
488 489 490 491 492 493 494
	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;
	struct prio_tree_root *root = NULL;
495
	struct anon_vma *anon_vma = NULL;
L
Linus Torvalds 已提交
496 497 498 499 500
	struct file *file = vma->vm_file;
	long adjust_next = 0;
	int remove_next = 0;

	if (next && !insert) {
501 502
		struct vm_area_struct *exporter = NULL;

L
Linus Torvalds 已提交
503 504 505 506 507 508 509
		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;
510
			exporter = next;
L
Linus Torvalds 已提交
511 512 513 514 515 516 517
			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;
518
			exporter = next;
L
Linus Torvalds 已提交
519 520 521 522 523 524 525 526
			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);
527
			exporter = vma;
L
Linus Torvalds 已提交
528 529 530
			importer = next;
		}

531 532 533 534 535
		/*
		 * 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.
		 */
536 537
		if (exporter && exporter->anon_vma && !importer->anon_vma) {
			if (anon_vma_clone(importer, exporter))
538
				return -ENOMEM;
539
			importer->anon_vma = exporter->anon_vma;
540 541 542
		}
	}

L
Linus Torvalds 已提交
543 544
	if (file) {
		mapping = file->f_mapping;
545
		if (!(vma->vm_flags & VM_NONLINEAR)) {
L
Linus Torvalds 已提交
546
			root = &mapping->i_mmap;
547
			uprobe_munmap(vma, vma->vm_start, vma->vm_end);
548 549

			if (adjust_next)
550 551
				uprobe_munmap(next, next->vm_start,
							next->vm_end);
552 553
		}

554
		mutex_lock(&mapping->i_mmap_mutex);
L
Linus Torvalds 已提交
555 556 557 558 559 560 561 562 563 564 565
		if (insert) {
			/*
			 * Put into prio_tree now, so instantiated pages
			 * 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);
		}
	}

566 567
	vma_adjust_trans_huge(vma, start, end, adjust_next);

568 569 570 571 572 573
	/*
	 * When changing only vma->vm_end, we don't really need anon_vma
	 * lock. This is a fairly rare case by itself, but the anon_vma
	 * lock may be shared between many sibling processes.  Skipping
	 * the lock for brk adjustments makes a difference sometimes.
	 */
574
	if (vma->anon_vma && (importer || start != vma->vm_start)) {
575
		anon_vma = vma->anon_vma;
576 577 578 579 580
		VM_BUG_ON(adjust_next && next->anon_vma &&
			  anon_vma != next->anon_vma);
	} else if (adjust_next && next->anon_vma)
		anon_vma = next->anon_vma;
	if (anon_vma)
581 582
		anon_vma_lock(anon_vma);

L
Linus Torvalds 已提交
583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621
	if (root) {
		flush_dcache_mmap_lock(mapping);
		vma_prio_tree_remove(vma, root);
		if (adjust_next)
			vma_prio_tree_remove(next, root);
	}

	vma->vm_start = start;
	vma->vm_end = end;
	vma->vm_pgoff = pgoff;
	if (adjust_next) {
		next->vm_start += adjust_next << PAGE_SHIFT;
		next->vm_pgoff += adjust_next;
	}

	if (root) {
		if (adjust_next)
			vma_prio_tree_insert(next, root);
		vma_prio_tree_insert(vma, root);
		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);
	}

622 623
	if (anon_vma)
		anon_vma_unlock(anon_vma);
L
Linus Torvalds 已提交
624
	if (mapping)
625
		mutex_unlock(&mapping->i_mmap_mutex);
L
Linus Torvalds 已提交
626

627
	if (root) {
628
		uprobe_mmap(vma);
629 630

		if (adjust_next)
631
			uprobe_mmap(next);
632 633
	}

L
Linus Torvalds 已提交
634
	if (remove_next) {
M
Matt Helsley 已提交
635
		if (file) {
636
			uprobe_munmap(next, next->vm_start, next->vm_end);
L
Linus Torvalds 已提交
637
			fput(file);
M
Matt Helsley 已提交
638
		}
639 640
		if (next->anon_vma)
			anon_vma_merge(vma, next);
L
Linus Torvalds 已提交
641
		mm->map_count--;
642
		mpol_put(vma_policy(next));
L
Linus Torvalds 已提交
643 644 645 646 647 648 649 650 651 652 653
		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.
		 */
		if (remove_next == 2) {
			next = vma->vm_next;
			goto again;
		}
	}
654
	if (insert && file)
655
		uprobe_mmap(insert);
L
Linus Torvalds 已提交
656 657

	validate_mm(mm);
658 659

	return 0;
L
Linus Torvalds 已提交
660 661 662 663 664 665 666 667 668
}

/*
 * 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)
{
669
	if (vma->vm_flags ^ vm_flags)
L
Linus Torvalds 已提交
670 671 672 673 674 675 676 677 678
		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
679 680
					struct anon_vma *anon_vma2,
					struct vm_area_struct *vma)
L
Linus Torvalds 已提交
681
{
S
Shaohua Li 已提交
682 683 684 685 686 687 688 689
	/*
	 * 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
690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707
}

/*
 * 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
708
	    is_mergeable_anon_vma(anon_vma, vma->anon_vma, vma)) {
L
Linus Torvalds 已提交
709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726
		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
727
	    is_mergeable_anon_vma(anon_vma, vma->anon_vma, vma)) {
L
Linus Torvalds 已提交
728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772
		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;
773
	int err;
L
Linus Torvalds 已提交
774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804

	/*
	 * 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
805
						      next->anon_vma, NULL)) {
L
Linus Torvalds 已提交
806
							/* cases 1, 6 */
807
			err = vma_adjust(prev, prev->vm_start,
L
Linus Torvalds 已提交
808 809
				next->vm_end, prev->vm_pgoff, NULL);
		} else					/* cases 2, 5, 7 */
810
			err = vma_adjust(prev, prev->vm_start,
L
Linus Torvalds 已提交
811
				end, prev->vm_pgoff, NULL);
812 813
		if (err)
			return NULL;
A
Andrea Arcangeli 已提交
814
		khugepaged_enter_vma_merge(prev);
L
Linus Torvalds 已提交
815 816 817 818 819 820 821 822 823 824 825
		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 */
826
			err = vma_adjust(prev, prev->vm_start,
L
Linus Torvalds 已提交
827 828
				addr, prev->vm_pgoff, NULL);
		else					/* cases 3, 8 */
829
			err = vma_adjust(area, addr, next->vm_end,
L
Linus Torvalds 已提交
830
				next->vm_pgoff - pglen, NULL);
831 832
		if (err)
			return NULL;
A
Andrea Arcangeli 已提交
833
		khugepaged_enter_vma_merge(area);
L
Linus Torvalds 已提交
834 835 836 837 838 839
		return area;
	}

	return NULL;
}

840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894
/*
 * 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
895 896 897 898 899 900 901 902 903 904
/*
 * 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)
{
905
	struct anon_vma *anon_vma;
L
Linus Torvalds 已提交
906 907 908 909 910 911
	struct vm_area_struct *near;

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

912 913 914
	anon_vma = reusable_anon_vma(near, vma, near);
	if (anon_vma)
		return anon_vma;
L
Linus Torvalds 已提交
915
try_prev:
916
	near = vma->vm_prev;
L
Linus Torvalds 已提交
917 918 919
	if (!near)
		goto none;

920 921 922
	anon_vma = reusable_anon_vma(near, near, vma);
	if (anon_vma)
		return anon_vma;
L
Linus Torvalds 已提交
923 924 925 926 927 928 929 930 931 932 933 934 935
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
936
void vm_stat_account(struct mm_struct *mm, unsigned long flags,
L
Linus Torvalds 已提交
937 938 939 940 941
						struct file *file, long pages)
{
	const unsigned long stack_flags
		= VM_STACK_FLAGS & (VM_GROWSUP|VM_GROWSDOWN);

942 943
	mm->total_vm += pages;

L
Linus Torvalds 已提交
944 945 946 947 948 949 950 951 952
	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
953 954 955 956 957 958 959 960 961 962 963 964 965
/*
 * 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
966
/*
967
 * The caller must hold down_write(&current->mm->mmap_sem).
L
Linus Torvalds 已提交
968 969
 */

970
unsigned long do_mmap_pgoff(struct file *file, unsigned long addr,
L
Linus Torvalds 已提交
971 972 973 974 975
			unsigned long len, unsigned long prot,
			unsigned long flags, unsigned long pgoff)
{
	struct mm_struct * mm = current->mm;
	struct inode *inode;
976
	vm_flags_t vm_flags;
L
Linus Torvalds 已提交
977 978 979 980 981 982 983 984

	/*
	 * 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))
985
		if (!(file && (file->f_path.mnt->mnt_flags & MNT_NOEXEC)))
L
Linus Torvalds 已提交
986 987 988 989 990
			prot |= PROT_EXEC;

	if (!len)
		return -EINVAL;

991 992 993
	if (!(flags & MAP_FIXED))
		addr = round_hint_to_min(addr);

L
Linus Torvalds 已提交
994 995
	/* Careful about overflows.. */
	len = PAGE_ALIGN(len);
996
	if (!len)
L
Linus Torvalds 已提交
997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020
		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
1021
	if (flags & MAP_LOCKED)
L
Linus Torvalds 已提交
1022 1023
		if (!can_do_mlock())
			return -EPERM;
1024

L
Linus Torvalds 已提交
1025 1026 1027
	/* mlock MCL_FUTURE? */
	if (vm_flags & VM_LOCKED) {
		unsigned long locked, lock_limit;
1028 1029
		locked = len >> PAGE_SHIFT;
		locked += mm->locked_vm;
J
Jiri Slaby 已提交
1030
		lock_limit = rlimit(RLIMIT_MEMLOCK);
1031
		lock_limit >>= PAGE_SHIFT;
L
Linus Torvalds 已提交
1032 1033 1034 1035
		if (locked > lock_limit && !capable(CAP_IPC_LOCK))
			return -EAGAIN;
	}

1036
	inode = file ? file->f_path.dentry->d_inode : NULL;
L
Linus Torvalds 已提交
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

	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;
1065
			if (file->f_path.mnt->mnt_flags & MNT_NOEXEC) {
L
Linus Torvalds 已提交
1066 1067 1068 1069 1070 1071 1072
				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
1073 1074 1075 1076 1077 1078 1079 1080
			break;

		default:
			return -EINVAL;
		}
	} else {
		switch (flags & MAP_TYPE) {
		case MAP_SHARED:
1081 1082 1083 1084
			/*
			 * Ignore pgoff.
			 */
			pgoff = 0;
L
Linus Torvalds 已提交
1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097
			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;
		}
	}

1098
	return mmap_region(file, addr, len, flags, vm_flags, pgoff);
M
Miklos Szeredi 已提交
1099
}
1100

H
Hugh Dickins 已提交
1101 1102 1103 1104 1105 1106 1107 1108
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
1109
		audit_mmap_fd(fd, flags);
H
Hugh Dickins 已提交
1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120 1121 1122
		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
		 */
1123 1124 1125
		file = hugetlb_file_setup(HUGETLB_ANON_FILE, addr, len,
						VM_NORESERVE, &user,
						HUGETLB_ANONHUGE_INODE);
H
Hugh Dickins 已提交
1126 1127 1128 1129 1130 1131
		if (IS_ERR(file))
			return PTR_ERR(file);
	}

	flags &= ~(MAP_EXECUTABLE | MAP_DENYWRITE);

A
Al Viro 已提交
1132
	retval = vm_mmap_pgoff(file, addr, len, prot, flags, pgoff);
H
Hugh Dickins 已提交
1133 1134 1135 1136 1137 1138
	if (file)
		fput(file);
out:
	return retval;
}

C
Christoph Hellwig 已提交
1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151 1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162
#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
1163 1164 1165 1166 1167 1168 1169 1170
/*
 * 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)
{
1171
	vm_flags_t vm_flags = vma->vm_flags;
A
Alexey Dobriyan 已提交
1172 1173 1174 1175 1176 1177 1178 1179 1180 1181 1182

	/* 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) !=
1183
	    pgprot_val(vm_get_page_prot(vm_flags)))
A
Alexey Dobriyan 已提交
1184 1185 1186
		return 0;

	/* Specialty mapping? */
1187
	if (vm_flags & VM_PFNMAP)
A
Alexey Dobriyan 已提交
1188 1189 1190 1191 1192 1193 1194
		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);
}

1195 1196
/*
 * We account for memory if it's a private writeable mapping,
1197
 * not hugepages and VM_NORESERVE wasn't set.
1198
 */
1199
static inline int accountable_mapping(struct file *file, vm_flags_t vm_flags)
1200
{
1201 1202 1203 1204 1205 1206 1207
	/*
	 * 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;

1208 1209 1210
	return (vm_flags & (VM_NORESERVE | VM_SHARED | VM_WRITE)) == VM_WRITE;
}

M
Miklos Szeredi 已提交
1211 1212
unsigned long mmap_region(struct file *file, unsigned long addr,
			  unsigned long len, unsigned long flags,
1213
			  vm_flags_t vm_flags, unsigned long pgoff)
M
Miklos Szeredi 已提交
1214 1215 1216 1217 1218 1219 1220 1221 1222
{
	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
1223 1224 1225
	/* Clear old maps */
	error = -ENOMEM;
munmap_back:
1226
	if (find_vma_links(mm, addr, addr + len, &prev, &rb_link, &rb_parent)) {
L
Linus Torvalds 已提交
1227 1228 1229 1230 1231 1232
		if (do_munmap(mm, addr, len))
			return -ENOMEM;
		goto munmap_back;
	}

	/* Check against address space limit. */
A
akpm@osdl.org 已提交
1233
	if (!may_expand_vm(mm, len >> PAGE_SHIFT))
L
Linus Torvalds 已提交
1234 1235
		return -ENOMEM;

1236 1237
	/*
	 * Set 'VM_NORESERVE' if we should not account for the
1238
	 * memory use of this mapping.
1239
	 */
1240 1241 1242 1243 1244 1245 1246 1247 1248
	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;
	}
1249

1250 1251 1252
	/*
	 * Private writable mapping: check memory availability
	 */
1253
	if (accountable_mapping(file, vm_flags)) {
1254
		charged = len >> PAGE_SHIFT;
1255
		if (security_vm_enough_memory_mm(mm, charged))
1256 1257
			return -ENOMEM;
		vm_flags |= VM_ACCOUNT;
L
Linus Torvalds 已提交
1258 1259 1260
	}

	/*
1261
	 * Can we just expand an old mapping?
L
Linus Torvalds 已提交
1262
	 */
1263 1264 1265
	vma = vma_merge(mm, prev, addr, addr + len, vm_flags, NULL, file, pgoff, NULL);
	if (vma)
		goto out;
L
Linus Torvalds 已提交
1266 1267 1268 1269 1270 1271

	/*
	 * 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
1272
	vma = kmem_cache_zalloc(vm_area_cachep, GFP_KERNEL);
L
Linus Torvalds 已提交
1273 1274 1275 1276 1277 1278 1279 1280 1281
	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;
1282
	vma->vm_page_prot = vm_get_page_prot(vm_flags);
L
Linus Torvalds 已提交
1283
	vma->vm_pgoff = pgoff;
1284
	INIT_LIST_HEAD(&vma->anon_vma_chain);
L
Linus Torvalds 已提交
1285

1286 1287
	error = -EINVAL;	/* when rejecting VM_GROWSDOWN|VM_GROWSUP */

L
Linus Torvalds 已提交
1288 1289 1290 1291 1292 1293 1294 1295 1296
	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
1297
		vma->vm_file = get_file(file);
L
Linus Torvalds 已提交
1298 1299 1300
		error = file->f_op->mmap(file, vma);
		if (error)
			goto unmap_and_free_vma;
1301 1302 1303 1304 1305 1306 1307 1308 1309

		/* Can addr have changed??
		 *
		 * Answer: Yes, several device drivers can do it in their
		 *         f_op->mmap method. -DaveM
		 */
		addr = vma->vm_start;
		pgoff = vma->vm_pgoff;
		vm_flags = vma->vm_flags;
L
Linus Torvalds 已提交
1310
	} else if (vm_flags & VM_SHARED) {
1311 1312
		if (unlikely(vm_flags & (VM_GROWSDOWN|VM_GROWSUP)))
			goto free_vma;
L
Linus Torvalds 已提交
1313 1314 1315 1316 1317
		error = shmem_zero_setup(vma);
		if (error)
			goto free_vma;
	}

1318 1319 1320 1321 1322 1323 1324 1325 1326 1327
	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.
		 */
1328
		vma->vm_page_prot = vm_get_page_prot(vm_flags & ~VM_SHARED);
1329 1330 1331
		if (pgprot_val(pprot) == pgprot_val(pgprot_noncached(pprot)))
			vma->vm_page_prot = pgprot_noncached(vma->vm_page_prot);
	}
1332

1333 1334
	vma_link(mm, vma, prev, rb_link, rb_parent);
	file = vma->vm_file;
1335 1336 1337 1338 1339

	/* Once vma denies write, undo our temporary denial count */
	if (correct_wcount)
		atomic_inc(&inode->i_writecount);
out:
1340
	perf_event_mmap(vma);
1341

1342
	vm_stat_account(mm, vm_flags, file, len >> PAGE_SHIFT);
L
Linus Torvalds 已提交
1343
	if (vm_flags & VM_LOCKED) {
1344 1345
		if (!mlock_vma_pages_range(vma, addr, addr + len))
			mm->locked_vm += (len >> PAGE_SHIFT);
1346
	} else if ((flags & MAP_POPULATE) && !(flags & MAP_NONBLOCK))
1347
		make_pages_present(addr, addr + len);
1348

1349 1350
	if (file)
		uprobe_mmap(vma);
1351

L
Linus Torvalds 已提交
1352 1353 1354 1355 1356 1357 1358 1359 1360
	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. */
1361 1362
	unmap_region(mm, vma, prev, vma->vm_start, vma->vm_end);
	charged = 0;
L
Linus Torvalds 已提交
1363 1364 1365 1366 1367 1368 1369 1370 1371 1372 1373 1374 1375 1376 1377 1378 1379 1380 1381 1382 1383 1384 1385 1386 1387 1388 1389 1390 1391 1392 1393
free_vma:
	kmem_cache_free(vm_area_cachep, vma);
unacct_error:
	if (charged)
		vm_unacct_memory(charged);
	return error;
}

/* 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;
	unsigned long start_addr;

	if (len > TASK_SIZE)
		return -ENOMEM;

1394 1395 1396
	if (flags & MAP_FIXED)
		return addr;

L
Linus Torvalds 已提交
1397 1398 1399 1400 1401 1402 1403
	if (addr) {
		addr = PAGE_ALIGN(addr);
		vma = find_vma(mm, addr);
		if (TASK_SIZE - len >= addr &&
		    (!vma || addr + len <= vma->vm_start))
			return addr;
	}
1404 1405 1406 1407 1408 1409
	if (len > mm->cached_hole_size) {
	        start_addr = addr = mm->free_area_cache;
	} else {
	        start_addr = addr = TASK_UNMAPPED_BASE;
	        mm->cached_hole_size = 0;
	}
L
Linus Torvalds 已提交
1410 1411 1412 1413 1414 1415 1416 1417 1418 1419

full_search:
	for (vma = find_vma(mm, addr); ; vma = vma->vm_next) {
		/* At this point:  (!vma || addr < vma->vm_end). */
		if (TASK_SIZE - len < addr) {
			/*
			 * Start a new search - just in case we missed
			 * some holes.
			 */
			if (start_addr != TASK_UNMAPPED_BASE) {
1420 1421 1422
				addr = TASK_UNMAPPED_BASE;
			        start_addr = addr;
				mm->cached_hole_size = 0;
L
Linus Torvalds 已提交
1423 1424 1425 1426 1427 1428 1429 1430 1431 1432 1433
				goto full_search;
			}
			return -ENOMEM;
		}
		if (!vma || addr + len <= vma->vm_start) {
			/*
			 * Remember the place where we stopped the search:
			 */
			mm->free_area_cache = addr + len;
			return addr;
		}
1434 1435
		if (addr + mm->cached_hole_size < vma->vm_start)
		        mm->cached_hole_size = vma->vm_start - addr;
L
Linus Torvalds 已提交
1436 1437 1438 1439 1440
		addr = vma->vm_end;
	}
}
#endif	

1441
void arch_unmap_area(struct mm_struct *mm, unsigned long addr)
L
Linus Torvalds 已提交
1442 1443 1444 1445
{
	/*
	 * Is this a new hole at the lowest possible address?
	 */
1446
	if (addr >= TASK_UNMAPPED_BASE && addr < mm->free_area_cache)
1447
		mm->free_area_cache = addr;
L
Linus Torvalds 已提交
1448 1449 1450 1451 1452 1453 1454 1455 1456 1457 1458 1459 1460 1461
}

/*
 * 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;
1462
	unsigned long addr = addr0, start_addr;
L
Linus Torvalds 已提交
1463 1464 1465 1466 1467

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

1468 1469 1470
	if (flags & MAP_FIXED)
		return addr;

L
Linus Torvalds 已提交
1471 1472 1473 1474 1475 1476 1477 1478 1479
	/* 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;
	}

1480 1481 1482 1483 1484 1485
	/* check if free_area_cache is useful for us */
	if (len <= mm->cached_hole_size) {
 	        mm->cached_hole_size = 0;
 		mm->free_area_cache = mm->mmap_base;
 	}

1486
try_again:
L
Linus Torvalds 已提交
1487
	/* either no address requested or can't fit in requested address hole */
1488
	start_addr = addr = mm->free_area_cache;
1489

1490 1491
	if (addr < len)
		goto fail;
L
Linus Torvalds 已提交
1492

1493
	addr -= len;
L
Linus Torvalds 已提交
1494 1495 1496 1497 1498 1499 1500 1501 1502 1503 1504
	do {
		/*
		 * Lookup failure means no vma is above this address,
		 * else if new region fits below vma->vm_start,
		 * return with success:
		 */
		vma = find_vma(mm, addr);
		if (!vma || addr+len <= vma->vm_start)
			/* remember the address as a hint for next time */
			return (mm->free_area_cache = addr);

1505 1506 1507 1508
 		/* remember the largest hole we saw so far */
 		if (addr + mm->cached_hole_size < vma->vm_start)
 		        mm->cached_hole_size = vma->vm_start - addr;

L
Linus Torvalds 已提交
1509 1510
		/* try just below the current vma->vm_start */
		addr = vma->vm_start-len;
1511
	} while (len < vma->vm_start);
L
Linus Torvalds 已提交
1512

1513 1514 1515 1516 1517 1518 1519 1520 1521 1522 1523 1524 1525 1526 1527
fail:
	/*
	 * if hint left us with no space for the requested
	 * mapping then try again:
	 *
	 * Note: this is different with the case of bottomup
	 * which does the fully line-search, but we use find_vma
	 * here that causes some holes skipped.
	 */
	if (start_addr != mm->mmap_base) {
		mm->free_area_cache = mm->mmap_base;
		mm->cached_hole_size = 0;
		goto try_again;
	}

L
Linus Torvalds 已提交
1528 1529 1530 1531 1532 1533
	/*
	 * 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.
	 */
1534 1535
	mm->cached_hole_size = ~0UL;
  	mm->free_area_cache = TASK_UNMAPPED_BASE;
L
Linus Torvalds 已提交
1536 1537 1538 1539 1540
	addr = arch_get_unmapped_area(filp, addr0, len, pgoff, flags);
	/*
	 * Restore the topdown base:
	 */
	mm->free_area_cache = mm->mmap_base;
1541
	mm->cached_hole_size = ~0UL;
L
Linus Torvalds 已提交
1542 1543 1544 1545 1546

	return addr;
}
#endif

1547
void arch_unmap_area_topdown(struct mm_struct *mm, unsigned long addr)
L
Linus Torvalds 已提交
1548 1549 1550 1551
{
	/*
	 * Is this a new hole at the highest possible address?
	 */
1552 1553
	if (addr > mm->free_area_cache)
		mm->free_area_cache = addr;
L
Linus Torvalds 已提交
1554 1555

	/* dont allow allocations above current base */
1556 1557
	if (mm->free_area_cache > mm->mmap_base)
		mm->free_area_cache = mm->mmap_base;
L
Linus Torvalds 已提交
1558 1559 1560 1561 1562 1563
}

unsigned long
get_unmapped_area(struct file *file, unsigned long addr, unsigned long len,
		unsigned long pgoff, unsigned long flags)
{
1564 1565 1566
	unsigned long (*get_area)(struct file *, unsigned long,
				  unsigned long, unsigned long, unsigned long);

1567 1568 1569 1570 1571 1572 1573 1574
	unsigned long error = arch_mmap_check(addr, len, flags);
	if (error)
		return error;

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

1575 1576 1577 1578 1579 1580
	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
1581

1582 1583 1584 1585
	if (addr > TASK_SIZE - len)
		return -ENOMEM;
	if (addr & ~PAGE_MASK)
		return -EINVAL;
1586

1587 1588 1589
	addr = arch_rebalance_pgtables(addr, len);
	error = security_mmap_addr(addr);
	return error ? error : addr;
L
Linus Torvalds 已提交
1590 1591 1592 1593 1594
}

EXPORT_SYMBOL(get_unmapped_area);

/* Look up the first VMA which satisfies  addr < vm_end,  NULL if none. */
Z
ZhenwenXu 已提交
1595
struct vm_area_struct *find_vma(struct mm_struct *mm, unsigned long addr)
L
Linus Torvalds 已提交
1596 1597 1598
{
	struct vm_area_struct *vma = NULL;

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
	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
1624
		}
1625 1626
		if (vma)
			mm->mmap_cache = vma;
L
Linus Torvalds 已提交
1627 1628 1629 1630 1631 1632
	}
	return vma;
}

EXPORT_SYMBOL(find_vma);

K
KOSAKI Motohiro 已提交
1633 1634 1635
/*
 * Same as find_vma, but also return a pointer to the previous VMA in *pprev.
 */
L
Linus Torvalds 已提交
1636 1637 1638 1639
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
1640
	struct vm_area_struct *vma;
L
Linus Torvalds 已提交
1641

K
KOSAKI Motohiro 已提交
1642
	vma = find_vma(mm, addr);
M
Mikulas Patocka 已提交
1643 1644 1645 1646 1647 1648 1649 1650 1651 1652
	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
1653
	return vma;
L
Linus Torvalds 已提交
1654 1655 1656 1657 1658 1659 1660
}

/*
 * 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
1661
static int acct_stack_growth(struct vm_area_struct *vma, unsigned long size, unsigned long grow)
L
Linus Torvalds 已提交
1662 1663 1664
{
	struct mm_struct *mm = vma->vm_mm;
	struct rlimit *rlim = current->signal->rlim;
1665
	unsigned long new_start;
L
Linus Torvalds 已提交
1666 1667

	/* address space limit tests */
A
akpm@osdl.org 已提交
1668
	if (!may_expand_vm(mm, grow))
L
Linus Torvalds 已提交
1669 1670 1671
		return -ENOMEM;

	/* Stack limit test */
J
Jiri Slaby 已提交
1672
	if (size > ACCESS_ONCE(rlim[RLIMIT_STACK].rlim_cur))
L
Linus Torvalds 已提交
1673 1674 1675 1676 1677 1678 1679
		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
1680 1681
		limit = ACCESS_ONCE(rlim[RLIMIT_MEMLOCK].rlim_cur);
		limit >>= PAGE_SHIFT;
L
Linus Torvalds 已提交
1682 1683 1684 1685
		if (locked > limit && !capable(CAP_IPC_LOCK))
			return -ENOMEM;
	}

1686 1687 1688 1689 1690 1691
	/* 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
1692 1693 1694 1695
	/*
	 * Overcommit..  This must be the final test, as it will
	 * update security statistics.
	 */
1696
	if (security_vm_enough_memory_mm(mm, grow))
L
Linus Torvalds 已提交
1697 1698 1699 1700 1701
		return -ENOMEM;

	/* Ok, everything looks good - let it rip */
	if (vma->vm_flags & VM_LOCKED)
		mm->locked_vm += grow;
1702
	vm_stat_account(mm, vma->vm_flags, vma->vm_file, grow);
L
Linus Torvalds 已提交
1703 1704 1705
	return 0;
}

H
Hugh Dickins 已提交
1706
#if defined(CONFIG_STACK_GROWSUP) || defined(CONFIG_IA64)
L
Linus Torvalds 已提交
1707
/*
H
Hugh Dickins 已提交
1708 1709
 * 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
1710
 */
H
Hugh Dickins 已提交
1711
int expand_upwards(struct vm_area_struct *vma, unsigned long address)
L
Linus Torvalds 已提交
1712 1713 1714 1715 1716 1717 1718 1719 1720 1721 1722 1723
{
	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;
1724
	vma_lock_anon_vma(vma);
L
Linus Torvalds 已提交
1725 1726 1727 1728 1729

	/*
	 * 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.
1730
	 * Also guard against wrapping around to address 0.
L
Linus Torvalds 已提交
1731
	 */
1732 1733 1734
	if (address < PAGE_ALIGN(address+4))
		address = PAGE_ALIGN(address+4);
	else {
1735
		vma_unlock_anon_vma(vma);
1736 1737
		return -ENOMEM;
	}
L
Linus Torvalds 已提交
1738 1739 1740 1741 1742 1743 1744 1745 1746
	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;

1747 1748 1749 1750 1751 1752 1753
		error = -ENOMEM;
		if (vma->vm_pgoff + (size >> PAGE_SHIFT) >= vma->vm_pgoff) {
			error = acct_stack_growth(vma, size, grow);
			if (!error) {
				vma->vm_end = address;
				perf_event_mmap(vma);
			}
1754
		}
L
Linus Torvalds 已提交
1755
	}
1756
	vma_unlock_anon_vma(vma);
A
Andrea Arcangeli 已提交
1757
	khugepaged_enter_vma_merge(vma);
L
Linus Torvalds 已提交
1758 1759
	return error;
}
H
Hugh Dickins 已提交
1760 1761
#endif /* CONFIG_STACK_GROWSUP || CONFIG_IA64 */

L
Linus Torvalds 已提交
1762 1763 1764
/*
 * vma is the first one with address < vma->vm_start.  Have to extend vma.
 */
1765
int expand_downwards(struct vm_area_struct *vma,
1766
				   unsigned long address)
L
Linus Torvalds 已提交
1767 1768 1769 1770 1771 1772 1773 1774 1775
{
	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;
1776 1777

	address &= PAGE_MASK;
1778
	error = security_mmap_addr(address);
1779 1780 1781
	if (error)
		return error;

1782
	vma_lock_anon_vma(vma);
L
Linus Torvalds 已提交
1783 1784 1785 1786 1787 1788 1789 1790 1791 1792 1793 1794 1795 1796

	/*
	 * 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;

1797 1798 1799 1800 1801 1802 1803 1804
		error = -ENOMEM;
		if (grow <= vma->vm_pgoff) {
			error = acct_stack_growth(vma, size, grow);
			if (!error) {
				vma->vm_start = address;
				vma->vm_pgoff -= grow;
				perf_event_mmap(vma);
			}
L
Linus Torvalds 已提交
1805 1806
		}
	}
1807
	vma_unlock_anon_vma(vma);
A
Andrea Arcangeli 已提交
1808
	khugepaged_enter_vma_merge(vma);
L
Linus Torvalds 已提交
1809 1810 1811
	return error;
}

1812 1813 1814 1815 1816 1817 1818 1819 1820 1821 1822 1823 1824 1825 1826
#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;
1827
	if (!prev || expand_stack(prev, addr))
1828
		return NULL;
1829
	if (prev->vm_flags & VM_LOCKED) {
1830
		mlock_vma_pages_range(prev, addr, prev->vm_end);
1831
	}
1832 1833 1834 1835 1836 1837 1838 1839
	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
1840 1841 1842 1843 1844 1845 1846 1847 1848 1849 1850 1851 1852 1853 1854 1855 1856
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;
1857
	if (vma->vm_flags & VM_LOCKED) {
1858
		mlock_vma_pages_range(vma, addr, start);
1859
	}
L
Linus Torvalds 已提交
1860 1861 1862 1863 1864
	return vma;
}
#endif

/*
1865
 * Ok - we have the memory areas we should free on the vma list,
L
Linus Torvalds 已提交
1866
 * so release them, and do the vma updates.
1867 1868
 *
 * Called with the mm semaphore held.
L
Linus Torvalds 已提交
1869
 */
1870
static void remove_vma_list(struct mm_struct *mm, struct vm_area_struct *vma)
L
Linus Torvalds 已提交
1871
{
1872 1873
	unsigned long nr_accounted = 0;

1874 1875
	/* Update high watermark before we lower total_vm */
	update_hiwater_vm(mm);
L
Linus Torvalds 已提交
1876
	do {
1877 1878
		long nrpages = vma_pages(vma);

1879 1880
		if (vma->vm_flags & VM_ACCOUNT)
			nr_accounted += nrpages;
1881
		vm_stat_account(mm, vma->vm_flags, vma->vm_file, -nrpages);
1882
		vma = remove_vma(vma);
1883
	} while (vma);
1884
	vm_unacct_memory(nr_accounted);
L
Linus Torvalds 已提交
1885 1886 1887 1888 1889 1890
	validate_mm(mm);
}

/*
 * Get rid of page table information in the indicated region.
 *
1891
 * Called with the mm semaphore held.
L
Linus Torvalds 已提交
1892 1893
 */
static void unmap_region(struct mm_struct *mm,
1894 1895
		struct vm_area_struct *vma, struct vm_area_struct *prev,
		unsigned long start, unsigned long end)
L
Linus Torvalds 已提交
1896
{
1897
	struct vm_area_struct *next = prev? prev->vm_next: mm->mmap;
P
Peter Zijlstra 已提交
1898
	struct mmu_gather tlb;
L
Linus Torvalds 已提交
1899 1900

	lru_add_drain();
P
Peter Zijlstra 已提交
1901
	tlb_gather_mmu(&tlb, mm, 0);
1902
	update_hiwater_rss(mm);
1903
	unmap_vmas(&tlb, vma, start, end);
P
Peter Zijlstra 已提交
1904 1905 1906
	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
1907 1908 1909 1910 1911 1912 1913 1914 1915 1916 1917 1918
}

/*
 * 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;
1919
	unsigned long addr;
L
Linus Torvalds 已提交
1920 1921

	insertion_point = (prev ? &prev->vm_next : &mm->mmap);
1922
	vma->vm_prev = NULL;
L
Linus Torvalds 已提交
1923 1924 1925 1926 1927 1928 1929
	do {
		rb_erase(&vma->vm_rb, &mm->mm_rb);
		mm->map_count--;
		tail_vma = vma;
		vma = vma->vm_next;
	} while (vma && vma->vm_start < end);
	*insertion_point = vma;
1930 1931
	if (vma)
		vma->vm_prev = prev;
L
Linus Torvalds 已提交
1932
	tail_vma->vm_next = NULL;
1933 1934 1935 1936 1937
	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
1938 1939 1940 1941
	mm->mmap_cache = NULL;		/* Kill the cache. */
}

/*
1942 1943
 * __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
1944
 */
1945
static int __split_vma(struct mm_struct * mm, struct vm_area_struct * vma,
L
Linus Torvalds 已提交
1946 1947 1948 1949
	      unsigned long addr, int new_below)
{
	struct mempolicy *pol;
	struct vm_area_struct *new;
1950
	int err = -ENOMEM;
L
Linus Torvalds 已提交
1951

1952 1953
	if (is_vm_hugetlb_page(vma) && (addr &
					~(huge_page_mask(hstate_vma(vma)))))
L
Linus Torvalds 已提交
1954 1955
		return -EINVAL;

1956
	new = kmem_cache_alloc(vm_area_cachep, GFP_KERNEL);
L
Linus Torvalds 已提交
1957
	if (!new)
1958
		goto out_err;
L
Linus Torvalds 已提交
1959 1960 1961 1962

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

1963 1964
	INIT_LIST_HEAD(&new->anon_vma_chain);

L
Linus Torvalds 已提交
1965 1966 1967 1968 1969 1970 1971
	if (new_below)
		new->vm_end = addr;
	else {
		new->vm_start = addr;
		new->vm_pgoff += ((addr - vma->vm_start) >> PAGE_SHIFT);
	}

1972
	pol = mpol_dup(vma_policy(vma));
L
Linus Torvalds 已提交
1973
	if (IS_ERR(pol)) {
1974 1975
		err = PTR_ERR(pol);
		goto out_free_vma;
L
Linus Torvalds 已提交
1976 1977 1978
	}
	vma_set_policy(new, pol);

1979 1980 1981
	if (anon_vma_clone(new, vma))
		goto out_free_mpol;

1982
	if (new->vm_file)
L
Linus Torvalds 已提交
1983 1984 1985 1986 1987 1988
		get_file(new->vm_file);

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

	if (new_below)
1989
		err = vma_adjust(vma, addr, vma->vm_end, vma->vm_pgoff +
L
Linus Torvalds 已提交
1990 1991
			((addr - new->vm_start) >> PAGE_SHIFT), new);
	else
1992
		err = vma_adjust(vma, vma->vm_start, addr, vma->vm_pgoff, new);
L
Linus Torvalds 已提交
1993

1994 1995 1996 1997 1998
	/* Success. */
	if (!err)
		return 0;

	/* Clean everything up if vma_adjust failed. */
1999 2000
	if (new->vm_ops && new->vm_ops->close)
		new->vm_ops->close(new);
2001
	if (new->vm_file)
2002
		fput(new->vm_file);
2003
	unlink_anon_vmas(new);
2004 2005 2006 2007 2008 2009
 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
2010 2011
}

2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024
/*
 * 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
2025 2026 2027 2028 2029 2030 2031 2032
/* 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;
2033
	struct vm_area_struct *vma, *prev, *last;
L
Linus Torvalds 已提交
2034 2035 2036 2037 2038 2039 2040 2041

	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 */
2042
	vma = find_vma(mm, start);
2043
	if (!vma)
L
Linus Torvalds 已提交
2044
		return 0;
2045
	prev = vma->vm_prev;
2046
	/* we have  start < vma->vm_end  */
L
Linus Torvalds 已提交
2047 2048 2049

	/* if it doesn't overlap, we have nothing.. */
	end = start + len;
2050
	if (vma->vm_start >= end)
L
Linus Torvalds 已提交
2051 2052 2053 2054 2055 2056 2057 2058 2059
		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.
	 */
2060
	if (start > vma->vm_start) {
2061 2062 2063 2064 2065 2066 2067 2068 2069 2070 2071
		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
2072 2073
		if (error)
			return error;
2074
		prev = vma;
L
Linus Torvalds 已提交
2075 2076 2077 2078 2079
	}

	/* Does it split the last one? */
	last = find_vma(mm, end);
	if (last && end > last->vm_start) {
2080
		int error = __split_vma(mm, last, end, 1);
L
Linus Torvalds 已提交
2081 2082 2083
		if (error)
			return error;
	}
2084
	vma = prev? prev->vm_next: mm->mmap;
L
Linus Torvalds 已提交
2085

2086 2087 2088 2089 2090 2091 2092 2093 2094 2095 2096 2097 2098 2099
	/*
	 * 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
2100 2101 2102
	/*
	 * Remove the vma's, and unmap the actual pages
	 */
2103 2104
	detach_vmas_to_be_unmapped(mm, vma, prev, end);
	unmap_region(mm, vma, prev, start, end);
L
Linus Torvalds 已提交
2105 2106

	/* Fix up all other VM information */
2107
	remove_vma_list(mm, vma);
L
Linus Torvalds 已提交
2108 2109 2110 2111

	return 0;
}

A
Al Viro 已提交
2112
int vm_munmap(unsigned long start, size_t len)
L
Linus Torvalds 已提交
2113 2114
{
	int ret;
A
Al Viro 已提交
2115
	struct mm_struct *mm = current->mm;
L
Linus Torvalds 已提交
2116 2117

	down_write(&mm->mmap_sem);
2118
	ret = do_munmap(mm, start, len);
L
Linus Torvalds 已提交
2119 2120 2121
	up_write(&mm->mmap_sem);
	return ret;
}
2122 2123 2124 2125 2126
EXPORT_SYMBOL(vm_munmap);

SYSCALL_DEFINE2(munmap, unsigned long, addr, size_t, len)
{
	profile_munmap(addr);
A
Al Viro 已提交
2127
	return vm_munmap(addr, len);
2128
}
L
Linus Torvalds 已提交
2129 2130 2131

static inline void verify_mm_writelocked(struct mm_struct *mm)
{
2132
#ifdef CONFIG_DEBUG_VM
L
Linus Torvalds 已提交
2133 2134 2135 2136 2137 2138 2139 2140 2141 2142 2143 2144
	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.
 */
2145
static unsigned long do_brk(unsigned long addr, unsigned long len)
L
Linus Torvalds 已提交
2146 2147 2148 2149 2150 2151
{
	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;
2152
	int error;
L
Linus Torvalds 已提交
2153 2154 2155 2156 2157

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

2158 2159
	flags = VM_DATA_DEFAULT_FLAGS | VM_ACCOUNT | mm->def_flags;

A
Al Viro 已提交
2160 2161
	error = get_unmapped_area(NULL, addr, len, 0, MAP_FIXED);
	if (error & ~PAGE_MASK)
2162 2163
		return error;

L
Linus Torvalds 已提交
2164 2165 2166 2167 2168
	/*
	 * mlock MCL_FUTURE?
	 */
	if (mm->def_flags & VM_LOCKED) {
		unsigned long locked, lock_limit;
2169 2170
		locked = len >> PAGE_SHIFT;
		locked += mm->locked_vm;
J
Jiri Slaby 已提交
2171
		lock_limit = rlimit(RLIMIT_MEMLOCK);
2172
		lock_limit >>= PAGE_SHIFT;
L
Linus Torvalds 已提交
2173 2174 2175 2176 2177 2178 2179 2180 2181 2182 2183 2184 2185 2186
		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:
2187
	if (find_vma_links(mm, addr, addr + len, &prev, &rb_link, &rb_parent)) {
L
Linus Torvalds 已提交
2188 2189 2190 2191 2192 2193
		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
2194
	if (!may_expand_vm(mm, len >> PAGE_SHIFT))
L
Linus Torvalds 已提交
2195 2196 2197 2198 2199
		return -ENOMEM;

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

2200
	if (security_vm_enough_memory_mm(mm, len >> PAGE_SHIFT))
L
Linus Torvalds 已提交
2201 2202 2203
		return -ENOMEM;

	/* Can we just expand an old private anonymous mapping? */
2204 2205 2206
	vma = vma_merge(mm, prev, addr, addr + len, flags,
					NULL, NULL, pgoff, NULL);
	if (vma)
L
Linus Torvalds 已提交
2207 2208 2209 2210 2211
		goto out;

	/*
	 * create a vma struct for an anonymous mapping
	 */
P
Pekka Enberg 已提交
2212
	vma = kmem_cache_zalloc(vm_area_cachep, GFP_KERNEL);
L
Linus Torvalds 已提交
2213 2214 2215 2216 2217
	if (!vma) {
		vm_unacct_memory(len >> PAGE_SHIFT);
		return -ENOMEM;
	}

2218
	INIT_LIST_HEAD(&vma->anon_vma_chain);
L
Linus Torvalds 已提交
2219 2220 2221 2222 2223
	vma->vm_mm = mm;
	vma->vm_start = addr;
	vma->vm_end = addr + len;
	vma->vm_pgoff = pgoff;
	vma->vm_flags = flags;
2224
	vma->vm_page_prot = vm_get_page_prot(flags);
L
Linus Torvalds 已提交
2225 2226
	vma_link(mm, vma, prev, rb_link, rb_parent);
out:
2227
	perf_event_mmap(vma);
L
Linus Torvalds 已提交
2228 2229
	mm->total_vm += len >> PAGE_SHIFT;
	if (flags & VM_LOCKED) {
2230 2231
		if (!mlock_vma_pages_range(vma, addr, addr + len))
			mm->locked_vm += (len >> PAGE_SHIFT);
L
Linus Torvalds 已提交
2232 2233 2234 2235
	}
	return addr;
}

2236 2237 2238 2239 2240 2241 2242 2243 2244 2245 2246
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
2247 2248 2249 2250

/* Release all mmaps. */
void exit_mmap(struct mm_struct *mm)
{
P
Peter Zijlstra 已提交
2251
	struct mmu_gather tlb;
2252
	struct vm_area_struct *vma;
L
Linus Torvalds 已提交
2253 2254
	unsigned long nr_accounted = 0;

2255
	/* mm's last user has gone, and its about to be pulled down */
A
Andrea Arcangeli 已提交
2256
	mmu_notifier_release(mm);
2257

2258 2259 2260 2261 2262 2263 2264 2265
	if (mm->locked_vm) {
		vma = mm->mmap;
		while (vma) {
			if (vma->vm_flags & VM_LOCKED)
				munlock_vma_pages_all(vma);
			vma = vma->vm_next;
		}
	}
2266 2267 2268

	arch_exit_mmap(mm);

2269
	vma = mm->mmap;
2270 2271 2272
	if (!vma)	/* Can happen if dup_mmap() received an OOM */
		return;

L
Linus Torvalds 已提交
2273 2274
	lru_add_drain();
	flush_cache_mm(mm);
P
Peter Zijlstra 已提交
2275
	tlb_gather_mmu(&tlb, mm, 1);
2276
	/* update_hiwater_rss(mm) here? but nobody should be looking */
2277
	/* Use -1 here to ensure all VMAs in the mm are unmapped */
2278
	unmap_vmas(&tlb, vma, 0, -1);
H
Hugh Dickins 已提交
2279

P
Peter Zijlstra 已提交
2280
	free_pgtables(&tlb, vma, FIRST_USER_ADDRESS, 0);
2281
	tlb_finish_mmu(&tlb, 0, -1);
L
Linus Torvalds 已提交
2282 2283

	/*
2284 2285
	 * Walk the list again, actually closing and freeing it,
	 * with preemption enabled, without holding any MM locks.
L
Linus Torvalds 已提交
2286
	 */
2287 2288 2289
	while (vma) {
		if (vma->vm_flags & VM_ACCOUNT)
			nr_accounted += vma_pages(vma);
2290
		vma = remove_vma(vma);
2291 2292
	}
	vm_unacct_memory(nr_accounted);
2293

2294
	WARN_ON(mm->nr_ptes > (FIRST_USER_ADDRESS+PMD_SIZE-1)>>PMD_SHIFT);
L
Linus Torvalds 已提交
2295 2296 2297 2298
}

/* Insert vm structure into process list sorted by address
 * and into the inode's i_mmap tree.  If vm_file is non-NULL
2299
 * then i_mmap_mutex is taken here.
L
Linus Torvalds 已提交
2300
 */
2301
int insert_vm_struct(struct mm_struct *mm, struct vm_area_struct *vma)
L
Linus Torvalds 已提交
2302
{
2303 2304
	struct vm_area_struct *prev;
	struct rb_node **rb_link, *rb_parent;
L
Linus Torvalds 已提交
2305 2306 2307 2308 2309 2310 2311 2312 2313 2314 2315 2316 2317 2318 2319 2320 2321

	/*
	 * 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;
	}
2322 2323
	if (find_vma_links(mm, vma->vm_start, vma->vm_end,
			   &prev, &rb_link, &rb_parent))
L
Linus Torvalds 已提交
2324
		return -ENOMEM;
2325
	if ((vma->vm_flags & VM_ACCOUNT) &&
2326
	     security_vm_enough_memory_mm(mm, vma_pages(vma)))
2327
		return -ENOMEM;
2328

L
Linus Torvalds 已提交
2329 2330 2331 2332 2333 2334 2335 2336 2337 2338 2339 2340 2341 2342 2343 2344 2345
	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,
	unsigned long addr, unsigned long len, pgoff_t pgoff)
{
	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;
2346
	bool faulted_in_anon_vma = true;
L
Linus Torvalds 已提交
2347 2348 2349 2350 2351

	/*
	 * If anonymous vma has not yet been faulted, update new pgoff
	 * to match new location, to increase its chance of merging.
	 */
2352
	if (unlikely(!vma->vm_file && !vma->anon_vma)) {
L
Linus Torvalds 已提交
2353
		pgoff = addr >> PAGE_SHIFT;
2354 2355
		faulted_in_anon_vma = false;
	}
L
Linus Torvalds 已提交
2356

2357 2358
	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
2359 2360 2361 2362 2363 2364
	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
		 */
2365 2366 2367 2368 2369 2370 2371 2372 2373 2374 2375 2376 2377 2378 2379
		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);
L
Linus Torvalds 已提交
2380
			*vmap = new_vma;
2381 2382
		} else
			anon_vma_moveto_tail(new_vma);
L
Linus Torvalds 已提交
2383
	} else {
2384
		new_vma = kmem_cache_alloc(vm_area_cachep, GFP_KERNEL);
L
Linus Torvalds 已提交
2385 2386
		if (new_vma) {
			*new_vma = *vma;
2387
			pol = mpol_dup(vma_policy(vma));
2388 2389 2390 2391 2392
			if (IS_ERR(pol))
				goto out_free_vma;
			INIT_LIST_HEAD(&new_vma->anon_vma_chain);
			if (anon_vma_clone(new_vma, vma))
				goto out_free_mempol;
L
Linus Torvalds 已提交
2393 2394 2395 2396
			vma_set_policy(new_vma, pol);
			new_vma->vm_start = addr;
			new_vma->vm_end = addr + len;
			new_vma->vm_pgoff = pgoff;
2397
			if (new_vma->vm_file)
L
Linus Torvalds 已提交
2398 2399 2400 2401 2402 2403 2404
				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);
		}
	}
	return new_vma;
2405 2406 2407 2408 2409 2410

 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
2411
}
A
akpm@osdl.org 已提交
2412 2413 2414 2415 2416 2417 2418 2419 2420 2421

/*
 * 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
2422
	lim = rlimit(RLIMIT_AS) >> PAGE_SHIFT;
A
akpm@osdl.org 已提交
2423 2424 2425 2426 2427

	if (cur + npages > lim)
		return 0;
	return 1;
}
2428 2429


N
Nick Piggin 已提交
2430 2431
static int special_mapping_fault(struct vm_area_struct *vma,
				struct vm_fault *vmf)
2432
{
N
Nick Piggin 已提交
2433
	pgoff_t pgoff;
2434 2435
	struct page **pages;

N
Nick Piggin 已提交
2436 2437 2438 2439 2440 2441 2442
	/*
	 * 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;
2443

N
Nick Piggin 已提交
2444 2445
	for (pages = vma->vm_private_data; pgoff && *pages; ++pages)
		pgoff--;
2446 2447 2448 2449

	if (*pages) {
		struct page *page = *pages;
		get_page(page);
N
Nick Piggin 已提交
2450 2451
		vmf->page = page;
		return 0;
2452 2453
	}

N
Nick Piggin 已提交
2454
	return VM_FAULT_SIGBUS;
2455 2456 2457 2458 2459 2460 2461 2462 2463
}

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

2464
static const struct vm_operations_struct special_mapping_vmops = {
2465
	.close = special_mapping_close,
N
Nick Piggin 已提交
2466
	.fault = special_mapping_fault,
2467 2468 2469 2470 2471 2472 2473 2474 2475 2476 2477 2478 2479 2480 2481
};

/*
 * 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)
{
2482
	int ret;
2483 2484 2485 2486 2487 2488
	struct vm_area_struct *vma;

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

2489
	INIT_LIST_HEAD(&vma->anon_vma_chain);
2490 2491 2492 2493
	vma->vm_mm = mm;
	vma->vm_start = addr;
	vma->vm_end = addr + len;

2494
	vma->vm_flags = vm_flags | mm->def_flags | VM_DONTEXPAND;
2495
	vma->vm_page_prot = vm_get_page_prot(vma->vm_flags);
2496 2497 2498 2499

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

2500 2501 2502
	ret = insert_vm_struct(mm, vma);
	if (ret)
		goto out;
2503 2504 2505

	mm->total_vm += len >> PAGE_SHIFT;

2506
	perf_event_mmap(vma);
2507

2508
	return 0;
2509 2510 2511 2512

out:
	kmem_cache_free(vm_area_cachep, vma);
	return ret;
2513
}
2514 2515 2516

static DEFINE_MUTEX(mm_all_locks_mutex);

2517
static void vm_lock_anon_vma(struct mm_struct *mm, struct anon_vma *anon_vma)
2518
{
2519
	if (!test_bit(0, (unsigned long *) &anon_vma->root->head.next)) {
2520 2521 2522 2523
		/*
		 * The LSB of head.next can't change from under us
		 * because we hold the mm_all_locks_mutex.
		 */
2524
		mutex_lock_nest_lock(&anon_vma->root->mutex, &mm->mmap_sem);
2525 2526
		/*
		 * We can safely modify head.next after taking the
2527
		 * anon_vma->root->mutex. If some other vma in this mm shares
2528 2529 2530 2531
		 * 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
2532
		 * anon_vma->root->mutex.
2533 2534
		 */
		if (__test_and_set_bit(0, (unsigned long *)
2535
				       &anon_vma->root->head.next))
2536 2537 2538 2539
			BUG();
	}
}

2540
static void vm_lock_mapping(struct mm_struct *mm, struct address_space *mapping)
2541 2542 2543 2544 2545 2546 2547 2548 2549 2550 2551 2552 2553
{
	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();
2554
		mutex_lock_nest_lock(&mapping->i_mmap_mutex, &mm->mmap_sem);
2555 2556 2557 2558 2559 2560 2561 2562 2563 2564 2565 2566 2567 2568 2569 2570 2571 2572 2573 2574 2575 2576 2577 2578 2579 2580
	}
}

/*
 * 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.
 *
 * The LSB in anon_vma->head.next and the AS_MM_ALL_LOCKS bitflag in
 * 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
2581
 * taking i_mmap_mutex or anon_vma->mutex outside the mmap_sem never
2582 2583 2584 2585 2586 2587 2588 2589 2590 2591 2592
 * 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;
2593
	struct anon_vma_chain *avc;
2594 2595 2596 2597 2598 2599 2600 2601 2602

	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)
2603
			vm_lock_mapping(mm, vma->vm_file->f_mapping);
2604
	}
2605 2606 2607 2608 2609

	for (vma = mm->mmap; vma; vma = vma->vm_next) {
		if (signal_pending(current))
			goto out_unlock;
		if (vma->anon_vma)
2610 2611
			list_for_each_entry(avc, &vma->anon_vma_chain, same_vma)
				vm_lock_anon_vma(mm, avc->anon_vma);
2612
	}
2613

2614
	return 0;
2615 2616

out_unlock:
2617 2618
	mm_drop_all_locks(mm);
	return -EINTR;
2619 2620 2621 2622
}

static void vm_unlock_anon_vma(struct anon_vma *anon_vma)
{
2623
	if (test_bit(0, (unsigned long *) &anon_vma->root->head.next)) {
2624 2625 2626 2627 2628 2629 2630 2631 2632 2633
		/*
		 * 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
		 * the vma so the users using the anon_vma->head will
		 * never see our bitflag.
		 *
		 * No need of atomic instructions here, head.next
		 * can't change from under us until we release the
2634
		 * anon_vma->root->mutex.
2635 2636
		 */
		if (!__test_and_clear_bit(0, (unsigned long *)
2637
					  &anon_vma->root->head.next))
2638
			BUG();
2639
		anon_vma_unlock(anon_vma);
2640 2641 2642 2643 2644 2645 2646 2647 2648 2649
	}
}

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.
		 */
2650
		mutex_unlock(&mapping->i_mmap_mutex);
2651 2652 2653 2654 2655 2656 2657 2658 2659 2660 2661 2662 2663
		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;
2664
	struct anon_vma_chain *avc;
2665 2666 2667 2668 2669 2670

	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)
2671 2672
			list_for_each_entry(avc, &vma->anon_vma_chain, same_vma)
				vm_unlock_anon_vma(avc->anon_vma);
2673 2674 2675 2676 2677 2678
		if (vma->vm_file && vma->vm_file->f_mapping)
			vm_unlock_mapping(vma->vm_file->f_mapping);
	}

	mutex_unlock(&mm_all_locks_mutex);
}
2679 2680 2681 2682 2683 2684

/*
 * initialise the VMA slab
 */
void __init mmap_init(void)
{
2685 2686 2687 2688
	int ret;

	ret = percpu_counter_init(&vm_committed_as, 0);
	VM_BUG_ON(ret);
2689
}