mmap.c 85.9 KB
Newer Older
Linus Torvalds's avatar
Linus Torvalds committed
1 2 3 4 5
/*
 * mm/mmap.c
 *
 * Written by obz.
 *
Alan Cox's avatar
Alan Cox committed
6
 * Address space accounting code	<alan@lxorguk.ukuu.org.uk>
Linus Torvalds's avatar
Linus Torvalds committed
7 8
 */

9
#include <linux/kernel.h>
Linus Torvalds's avatar
Linus Torvalds committed
10
#include <linux/slab.h>
11
#include <linux/backing-dev.h>
Linus Torvalds's avatar
Linus Torvalds committed
12 13 14 15 16 17
#include <linux/mm.h>
#include <linux/shm.h>
#include <linux/mman.h>
#include <linux/pagemap.h>
#include <linux/swap.h>
#include <linux/syscalls.h>
18
#include <linux/capability.h>
Linus Torvalds's avatar
Linus Torvalds committed
19 20 21 22 23 24 25
#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>
26
#include <linux/export.h>
Linus Torvalds's avatar
Linus Torvalds committed
27 28 29
#include <linux/mount.h>
#include <linux/mempolicy.h>
#include <linux/rmap.h>
Andrea Arcangeli's avatar
Andrea Arcangeli committed
30
#include <linux/mmu_notifier.h>
31
#include <linux/perf_event.h>
Al Viro's avatar
Al Viro committed
32
#include <linux/audit.h>
Andrea Arcangeli's avatar
Andrea Arcangeli committed
33
#include <linux/khugepaged.h>
34
#include <linux/uprobes.h>
35
#include <linux/rbtree_augmented.h>
36
#include <linux/sched/sysctl.h>
37 38
#include <linux/notifier.h>
#include <linux/memory.h>
Linus Torvalds's avatar
Linus Torvalds committed
39 40 41 42

#include <asm/uaccess.h>
#include <asm/cacheflush.h>
#include <asm/tlb.h>
43
#include <asm/mmu_context.h>
Linus Torvalds's avatar
Linus Torvalds committed
44

45 46
#include "internal.h"

47 48 49 50
#ifndef arch_mmap_check
#define arch_mmap_check(addr, len, flags)	(0)
#endif

51 52 53 54
#ifndef arch_rebalance_pgtables
#define arch_rebalance_pgtables(addr, len)		(addr)
#endif

55 56 57 58
static void unmap_region(struct mm_struct *mm,
		struct vm_area_struct *vma, struct vm_area_struct *prev,
		unsigned long start, unsigned long end);

Linus Torvalds's avatar
Linus Torvalds committed
59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78
/* 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
};

79 80
pgprot_t vm_get_page_prot(unsigned long vm_flags)
{
81 82 83
	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)));
84 85 86
}
EXPORT_SYMBOL(vm_get_page_prot);

87 88
int sysctl_overcommit_memory __read_mostly = OVERCOMMIT_GUESS;  /* heuristic overcommit */
int sysctl_overcommit_ratio __read_mostly = 50;	/* default is 50% */
89
int sysctl_max_map_count __read_mostly = DEFAULT_MAX_MAP_COUNT;
90
unsigned long sysctl_user_reserve_kbytes __read_mostly = 1UL << 17; /* 128MB */
91
unsigned long sysctl_admin_reserve_kbytes __read_mostly = 1UL << 13; /* 8MB */
92 93 94 95 96
/*
 * 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;
Linus Torvalds's avatar
Linus Torvalds committed
97

98 99 100 101 102 103 104 105 106 107 108 109 110 111
/*
 * 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);

Linus Torvalds's avatar
Linus Torvalds committed
112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127
/*
 * 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.
 */
128
int __vm_enough_memory(struct mm_struct *mm, long pages, int cap_sys_admin)
Linus Torvalds's avatar
Linus Torvalds committed
129
{
130
	unsigned long free, allowed, reserve;
Linus Torvalds's avatar
Linus Torvalds committed
131 132 133 134 135 136 137 138 139 140

	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) {
141 142 143 144 145 146 147 148 149 150
		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);
Linus Torvalds's avatar
Linus Torvalds committed
151

152
		free += get_nr_swap_pages();
Linus Torvalds's avatar
Linus Torvalds committed
153 154 155 156 157 158 159

		/*
		 * 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
		 */
160
		free += global_page_state(NR_SLAB_RECLAIMABLE);
Linus Torvalds's avatar
Linus Torvalds committed
161

162 163 164
		/*
		 * Leave reserved pages. The pages are not for anonymous pages.
		 */
165
		if (free <= totalreserve_pages)
166 167
			goto error;
		else
168
			free -= totalreserve_pages;
169 170

		/*
171
		 * Reserve some for root
172
		 */
Linus Torvalds's avatar
Linus Torvalds committed
173
		if (!cap_sys_admin)
174
			free -= sysctl_admin_reserve_kbytes >> (PAGE_SHIFT - 10);
Linus Torvalds's avatar
Linus Torvalds committed
175 176 177

		if (free > pages)
			return 0;
178 179

		goto error;
Linus Torvalds's avatar
Linus Torvalds committed
180 181 182 183 184
	}

	allowed = (totalram_pages - hugetlb_total_pages())
	       	* sysctl_overcommit_ratio / 100;
	/*
185
	 * Reserve some for root
Linus Torvalds's avatar
Linus Torvalds committed
186 187
	 */
	if (!cap_sys_admin)
188
		allowed -= sysctl_admin_reserve_kbytes >> (PAGE_SHIFT - 10);
Linus Torvalds's avatar
Linus Torvalds committed
189 190
	allowed += total_swap_pages;

191 192 193 194 195 196 197
	/*
	 * Don't let a single process grow so big a user can't recover
	 */
	if (mm) {
		reserve = sysctl_user_reserve_kbytes >> (PAGE_SHIFT - 10);
		allowed -= min(mm->total_vm / 32, reserve);
	}
Linus Torvalds's avatar
Linus Torvalds committed
198

199
	if (percpu_counter_read_positive(&vm_committed_as) < allowed)
Linus Torvalds's avatar
Linus Torvalds committed
200
		return 0;
201
error:
Linus Torvalds's avatar
Linus Torvalds committed
202 203 204 205 206 207
	vm_unacct_memory(pages);

	return -ENOMEM;
}

/*
208
 * Requires inode->i_mapping->i_mmap_mutex
Linus Torvalds's avatar
Linus Torvalds committed
209 210 211 212 213
 */
static void __remove_shared_vm_struct(struct vm_area_struct *vma,
		struct file *file, struct address_space *mapping)
{
	if (vma->vm_flags & VM_DENYWRITE)
Al Viro's avatar
Al Viro committed
214
		atomic_inc(&file_inode(file)->i_writecount);
Linus Torvalds's avatar
Linus Torvalds committed
215 216 217 218 219
	if (vma->vm_flags & VM_SHARED)
		mapping->i_mmap_writable--;

	flush_dcache_mmap_lock(mapping);
	if (unlikely(vma->vm_flags & VM_NONLINEAR))
220
		list_del_init(&vma->shared.nonlinear);
Linus Torvalds's avatar
Linus Torvalds committed
221
	else
222
		vma_interval_tree_remove(vma, &mapping->i_mmap);
Linus Torvalds's avatar
Linus Torvalds committed
223 224 225 226
	flush_dcache_mmap_unlock(mapping);
}

/*
227
 * Unlink a file-based vm structure from its interval tree, to hide
228
 * vma from rmap and vmtruncate before freeing its page tables.
Linus Torvalds's avatar
Linus Torvalds committed
229
 */
230
void unlink_file_vma(struct vm_area_struct *vma)
Linus Torvalds's avatar
Linus Torvalds committed
231 232 233 234 235
{
	struct file *file = vma->vm_file;

	if (file) {
		struct address_space *mapping = file->f_mapping;
236
		mutex_lock(&mapping->i_mmap_mutex);
Linus Torvalds's avatar
Linus Torvalds committed
237
		__remove_shared_vm_struct(vma, file, mapping);
238
		mutex_unlock(&mapping->i_mmap_mutex);
Linus Torvalds's avatar
Linus Torvalds committed
239
	}
240 241 242 243 244 245 246 247 248 249
}

/*
 * 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();
Linus Torvalds's avatar
Linus Torvalds committed
250 251
	if (vma->vm_ops && vma->vm_ops->close)
		vma->vm_ops->close(vma);
252
	if (vma->vm_file)
253
		fput(vma->vm_file);
254
	mpol_put(vma_policy(vma));
Linus Torvalds's avatar
Linus Torvalds committed
255
	kmem_cache_free(vm_area_cachep, vma);
256
	return next;
Linus Torvalds's avatar
Linus Torvalds committed
257 258
}

259 260
static unsigned long do_brk(unsigned long addr, unsigned long len);

261
SYSCALL_DEFINE1(brk, unsigned long, brk)
Linus Torvalds's avatar
Linus Torvalds committed
262 263 264 265
{
	unsigned long rlim, retval;
	unsigned long newbrk, oldbrk;
	struct mm_struct *mm = current->mm;
266
	unsigned long min_brk;
267
	bool populate;
Linus Torvalds's avatar
Linus Torvalds committed
268 269 270

	down_write(&mm->mmap_sem);

271
#ifdef CONFIG_COMPAT_BRK
272 273 274 275 276
	/*
	 * 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
	 */
277
	if (current->brk_randomized)
278 279 280
		min_brk = mm->start_brk;
	else
		min_brk = mm->end_data;
281 282 283 284
#else
	min_brk = mm->start_brk;
#endif
	if (brk < min_brk)
Linus Torvalds's avatar
Linus Torvalds committed
285
		goto out;
Ram Gupta's avatar
Ram Gupta committed
286 287 288 289 290 291 292

	/*
	 * 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
	 */
Jiri Slaby's avatar
Jiri Slaby committed
293
	rlim = rlimit(RLIMIT_DATA);
Jiri Kosina's avatar
Jiri Kosina committed
294 295
	if (rlim < RLIM_INFINITY && (brk - mm->start_brk) +
			(mm->end_data - mm->start_data) > rlim)
Ram Gupta's avatar
Ram Gupta committed
296 297
		goto out;

Linus Torvalds's avatar
Linus Torvalds committed
298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316
	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;
317

Linus Torvalds's avatar
Linus Torvalds committed
318 319
set_brk:
	mm->brk = brk;
320 321 322 323 324 325
	populate = newbrk > oldbrk && (mm->def_flags & VM_LOCKED) != 0;
	up_write(&mm->mmap_sem);
	if (populate)
		mm_populate(oldbrk, newbrk - oldbrk);
	return brk;

Linus Torvalds's avatar
Linus Torvalds committed
326 327 328 329 330 331
out:
	retval = mm->brk;
	up_write(&mm->mmap_sem);
	return retval;
}

332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352
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;
}

353
#ifdef CONFIG_DEBUG_VM_RB
Linus Torvalds's avatar
Linus Torvalds committed
354 355
static int browse_rb(struct rb_root *root)
{
356
	int i = 0, j, bug = 0;
Linus Torvalds's avatar
Linus Torvalds committed
357 358 359 360 361 362
	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);
363 364 365 366 367
		if (vma->vm_start < prev) {
			printk("vm_start %lx prev %lx\n", vma->vm_start, prev);
			bug = 1;
		}
		if (vma->vm_start < pend) {
Linus Torvalds's avatar
Linus Torvalds committed
368
			printk("vm_start %lx pend %lx\n", vma->vm_start, pend);
369 370 371 372 373 374 375 376 377 378 379 380 381
			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;
		}
Linus Torvalds's avatar
Linus Torvalds committed
382 383
		i++;
		pn = nd;
384 385
		prev = vma->vm_start;
		pend = vma->vm_end;
Linus Torvalds's avatar
Linus Torvalds committed
386 387
	}
	j = 0;
388
	for (nd = pn; nd; nd = rb_prev(nd))
Linus Torvalds's avatar
Linus Torvalds committed
389
		j++;
390 391 392
	if (i != j) {
		printk("backwards %d, forwards %d\n", j, i);
		bug = 1;
Linus Torvalds's avatar
Linus Torvalds committed
393
	}
394
	return bug ? -1 : i;
Linus Torvalds's avatar
Linus Torvalds committed
395 396
}

397 398 399 400 401 402 403 404 405
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));
Linus Torvalds's avatar
Linus Torvalds committed
406 407 408 409 410 411 412
	}
}

void validate_mm(struct mm_struct *mm)
{
	int bug = 0;
	int i = 0;
413
	unsigned long highest_address = 0;
414 415 416
	struct vm_area_struct *vma = mm->mmap;
	while (vma) {
		struct anon_vma_chain *avc;
417
		vma_lock_anon_vma(vma);
418 419
		list_for_each_entry(avc, &vma->anon_vma_chain, same_vma)
			anon_vma_interval_tree_verify(avc);
420
		vma_unlock_anon_vma(vma);
421
		highest_address = vma->vm_end;
422
		vma = vma->vm_next;
Linus Torvalds's avatar
Linus Torvalds committed
423 424
		i++;
	}
425 426 427 428 429 430 431 432 433
	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;
	}
Linus Torvalds's avatar
Linus Torvalds committed
434
	i = browse_rb(&mm->mm_rb);
435 436 437 438
	if (i != mm->map_count) {
		printk("map_count %d rb %d\n", mm->map_count, i);
		bug = 1;
	}
439
	BUG_ON(bug);
Linus Torvalds's avatar
Linus Torvalds committed
440 441
}
#else
442
#define validate_mm_rb(root, ignore) do { } while (0)
Linus Torvalds's avatar
Linus Torvalds committed
443 444 445
#define validate_mm(mm) do { } while (0)
#endif

446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487
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);
}

488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519
/*
 * 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);
}

520 521 522
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)
Linus Torvalds's avatar
Linus Torvalds committed
523
{
524
	struct rb_node **__rb_link, *__rb_parent, *rb_prev;
Linus Torvalds's avatar
Linus Torvalds committed
525 526 527 528 529 530 531 532 533 534 535

	__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) {
536 537 538
			/* Fail if an existing vma overlaps the area */
			if (vma_tmp->vm_start < end)
				return -ENOMEM;
Linus Torvalds's avatar
Linus Torvalds committed
539 540 541 542 543 544 545 546 547 548 549 550
			__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;
551
	return 0;
Linus Torvalds's avatar
Linus Torvalds committed
552 553
}

554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581
static unsigned long count_vma_pages_range(struct mm_struct *mm,
		unsigned long addr, unsigned long end)
{
	unsigned long nr_pages = 0;
	struct vm_area_struct *vma;

	/* Find first overlaping mapping */
	vma = find_vma_intersection(mm, addr, end);
	if (!vma)
		return 0;

	nr_pages = (min(end, vma->vm_end) -
		max(addr, vma->vm_start)) >> PAGE_SHIFT;

	/* Iterate over the rest of the overlaps */
	for (vma = vma->vm_next; vma; vma = vma->vm_next) {
		unsigned long overlap_len;

		if (vma->vm_start > end)
			break;

		overlap_len = min(end, vma->vm_end) - vma->vm_start;
		nr_pages += overlap_len >> PAGE_SHIFT;
	}

	return nr_pages;
}

Linus Torvalds's avatar
Linus Torvalds committed
582 583 584
void __vma_link_rb(struct mm_struct *mm, struct vm_area_struct *vma,
		struct rb_node **rb_link, struct rb_node *rb_parent)
{
585 586 587 588 589 590 591 592 593 594 595 596 597 598 599
	/* 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.
	 */
Linus Torvalds's avatar
Linus Torvalds committed
600
	rb_link_node(&vma->vm_rb, rb_parent, rb_link);
601 602 603
	vma->rb_subtree_gap = 0;
	vma_gap_update(vma);
	vma_rb_insert(vma, &mm->mm_rb);
Linus Torvalds's avatar
Linus Torvalds committed
604 605
}

606
static void __vma_link_file(struct vm_area_struct *vma)
Linus Torvalds's avatar
Linus Torvalds committed
607
{
ZhenwenXu's avatar
ZhenwenXu committed
608
	struct file *file;
Linus Torvalds's avatar
Linus Torvalds committed
609 610 611 612 613 614

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

		if (vma->vm_flags & VM_DENYWRITE)
Al Viro's avatar
Al Viro committed
615
			atomic_dec(&file_inode(file)->i_writecount);
Linus Torvalds's avatar
Linus Torvalds committed
616 617 618 619 620 621 622
		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
623
			vma_interval_tree_insert(vma, &mapping->i_mmap);
Linus Torvalds's avatar
Linus Torvalds committed
624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645
		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;

646
	if (mapping)
647
		mutex_lock(&mapping->i_mmap_mutex);
Linus Torvalds's avatar
Linus Torvalds committed
648 649 650 651 652

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

	if (mapping)
653
		mutex_unlock(&mapping->i_mmap_mutex);
Linus Torvalds's avatar
Linus Torvalds committed
654 655 656 657 658 659

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

/*
660
 * Helper for vma_adjust() in the split_vma insert case: insert a vma into the
661
 * mm's list and rbtree.  It has already been inserted into the interval tree.
Linus Torvalds's avatar
Linus Torvalds committed
662
 */
ZhenwenXu's avatar
ZhenwenXu committed
663
static void __insert_vm_struct(struct mm_struct *mm, struct vm_area_struct *vma)
Linus Torvalds's avatar
Linus Torvalds committed
664
{
665
	struct vm_area_struct *prev;
ZhenwenXu's avatar
ZhenwenXu committed
666
	struct rb_node **rb_link, *rb_parent;
Linus Torvalds's avatar
Linus Torvalds committed
667

668 669 670
	if (find_vma_links(mm, vma->vm_start, vma->vm_end,
			   &prev, &rb_link, &rb_parent))
		BUG();
Linus Torvalds's avatar
Linus Torvalds committed
671 672 673 674 675 676 677 678
	__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)
{
679
	struct vm_area_struct *next;
680

681 682
	vma_rb_erase(vma, &mm->mm_rb);
	prev->vm_next = next = vma->vm_next;
683 684
	if (next)
		next->vm_prev = prev;
Linus Torvalds's avatar
Linus Torvalds committed
685 686 687 688 689 690 691 692 693 694 695
	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.
 */
696
int vma_adjust(struct vm_area_struct *vma, unsigned long start,
Linus Torvalds's avatar
Linus Torvalds committed
697 698 699 700 701 702
	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;
703
	struct rb_root *root = NULL;
704
	struct anon_vma *anon_vma = NULL;
Linus Torvalds's avatar
Linus Torvalds committed
705
	struct file *file = vma->vm_file;
706
	bool start_changed = false, end_changed = false;
Linus Torvalds's avatar
Linus Torvalds committed
707 708 709 710
	long adjust_next = 0;
	int remove_next = 0;

	if (next && !insert) {
711 712
		struct vm_area_struct *exporter = NULL;

Linus Torvalds's avatar
Linus Torvalds committed
713 714 715 716 717 718 719
		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;
720
			exporter = next;
Linus Torvalds's avatar
Linus Torvalds committed
721 722 723 724 725 726 727
			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;
728
			exporter = next;
Linus Torvalds's avatar
Linus Torvalds committed
729 730 731 732 733 734 735 736
			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);
737
			exporter = vma;
Linus Torvalds's avatar
Linus Torvalds committed
738 739 740
			importer = next;
		}

741 742 743 744 745
		/*
		 * 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.
		 */
746 747
		if (exporter && exporter->anon_vma && !importer->anon_vma) {
			if (anon_vma_clone(importer, exporter))
748
				return -ENOMEM;
749
			importer->anon_vma = exporter->anon_vma;
750 751 752
		}
	}

Linus Torvalds's avatar
Linus Torvalds committed
753 754
	if (file) {
		mapping = file->f_mapping;
755
		if (!(vma->vm_flags & VM_NONLINEAR)) {
Linus Torvalds's avatar
Linus Torvalds committed
756
			root = &mapping->i_mmap;
757
			uprobe_munmap(vma, vma->vm_start, vma->vm_end);
758 759

			if (adjust_next)
760 761
				uprobe_munmap(next, next->vm_start,
							next->vm_end);
762 763
		}

764
		mutex_lock(&mapping->i_mmap_mutex);
Linus Torvalds's avatar
Linus Torvalds committed
765 766
		if (insert) {
			/*
767
			 * Put into interval tree now, so instantiated pages
Linus Torvalds's avatar
Linus Torvalds committed
768 769 770 771 772 773 774 775
			 * 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);
		}
	}

776 777
	vma_adjust_trans_huge(vma, start, end, adjust_next);

778 779 780 781
	anon_vma = vma->anon_vma;
	if (!anon_vma && adjust_next)
		anon_vma = next->anon_vma;
	if (anon_vma) {
782 783
		VM_BUG_ON(adjust_next && next->anon_vma &&
			  anon_vma != next->anon_vma);
784
		anon_vma_lock_write(anon_vma);
785 786 787 788
		anon_vma_interval_tree_pre_update_vma(vma);
		if (adjust_next)
			anon_vma_interval_tree_pre_update_vma(next);
	}
789

Linus Torvalds's avatar
Linus Torvalds committed
790 791
	if (root) {
		flush_dcache_mmap_lock(mapping);
792
		vma_interval_tree_remove(vma, root);
Linus Torvalds's avatar
Linus Torvalds committed
793
		if (adjust_next)
794
			vma_interval_tree_remove(next, root);
Linus Torvalds's avatar
Linus Torvalds committed
795 796
	}

797 798 799 800 801 802 803 804
	if (start != vma->vm_start) {
		vma->vm_start = start;
		start_changed = true;
	}
	if (end != vma->vm_end) {
		vma->vm_end = end;
		end_changed = true;
	}
Linus Torvalds's avatar
Linus Torvalds committed
805 806 807 808 809 810 811 812
	vma->vm_pgoff = pgoff;
	if (adjust_next) {
		next->vm_start += adjust_next << PAGE_SHIFT;
		next->vm_pgoff += adjust_next;
	}

	if (root) {
		if (adjust_next)
813 814
			vma_interval_tree_insert(next, root);
		vma_interval_tree_insert(vma, root);
Linus Torvalds's avatar
Linus Torvalds committed
815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832
		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);
833 834 835 836 837 838 839 840 841
	} 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);
		}
Linus Torvalds's avatar
Linus Torvalds committed
842 843
	}

844 845 846 847
	if (anon_vma) {
		anon_vma_interval_tree_post_update_vma(vma);
		if (adjust_next)
			anon_vma_interval_tree_post_update_vma(next);
848
		anon_vma_unlock_write(anon_vma);
849
	}
Linus Torvalds's avatar
Linus Torvalds committed
850
	if (mapping)
851
		mutex_unlock(&mapping->i_mmap_mutex);
Linus Torvalds's avatar
Linus Torvalds committed
852

853
	if (root) {
854
		uprobe_mmap(vma);
855 856

		if (adjust_next)
857
			uprobe_mmap(next);
858 859
	}

Linus Torvalds's avatar
Linus Torvalds committed
860
	if (remove_next) {
Matt Helsley's avatar
Matt Helsley committed
861
		if (file) {
862
			uprobe_munmap(next, next->vm_start, next->vm_end);
Linus Torvalds's avatar
Linus Torvalds committed
863
			fput(file);
Matt Helsley's avatar
Matt Helsley committed
864
		}
865 866
		if (next->anon_vma)
			anon_vma_merge(vma, next);
Linus Torvalds's avatar
Linus Torvalds committed
867
		mm->map_count--;
868
		vma_set_policy(vma, vma_policy(next));
Linus Torvalds's avatar
Linus Torvalds committed
869 870 871 872 873 874
		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.
		 */
875 876
		next = vma->vm_next;
		if (remove_next == 2)
Linus Torvalds's avatar
Linus Torvalds committed
877
			goto again;
878 879 880 881
		else if (next)
			vma_gap_update(next);
		else
			mm->highest_vm_end = end;
Linus Torvalds's avatar
Linus Torvalds committed
882
	}
883
	if (insert && file)
884
		uprobe_mmap(insert);
Linus Torvalds's avatar
Linus Torvalds committed
885 886

	validate_mm(mm);
887 888

	return 0;
Linus Torvalds's avatar
Linus Torvalds committed
889 890 891 892 893 894 895 896 897
}

/*
 * 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)
{
898
	if (vma->vm_flags ^ vm_flags)
Linus Torvalds's avatar
Linus Torvalds committed
899 900 901 902 903 904 905 906 907
		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,
Shaohua Li's avatar
Shaohua Li committed
908 909
					struct anon_vma *anon_vma2,
					struct vm_area_struct *vma)
Linus Torvalds's avatar
Linus Torvalds committed
910
{
Shaohua Li's avatar
Shaohua Li committed
911 912 913 914 915 916 917 918
	/*
	 * 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;
Linus Torvalds's avatar
Linus Torvalds committed
919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936
}

/*
 * 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) &&
Shaohua Li's avatar
Shaohua Li committed
937
	    is_mergeable_anon_vma(anon_vma, vma->anon_vma, vma)) {
Linus Torvalds's avatar
Linus Torvalds committed
938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955
		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) &&
Shaohua Li's avatar
Shaohua Li committed
956
	    is_mergeable_anon_vma(anon_vma, vma->anon_vma, vma)) {
Linus Torvalds's avatar
Linus Torvalds committed
957
		pgoff_t vm_pglen;
958
		vm_pglen = vma_pages(vma);
Linus Torvalds's avatar
Linus Torvalds committed
959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000 1001
		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;
1002
	int err;
Linus Torvalds's avatar
Linus Torvalds committed
1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033

	/*
	 * 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,
Shaohua Li's avatar
Shaohua Li committed
1034
						      next->anon_vma, NULL)) {
Linus Torvalds's avatar
Linus Torvalds committed
1035
							/* cases 1, 6 */
1036
			err = vma_adjust(prev, prev->vm_start,
Linus Torvalds's avatar
Linus Torvalds committed
1037 1038
				next->vm_end, prev->vm_pgoff, NULL);
		} else					/* cases 2, 5, 7 */
1039
			err = vma_adjust(prev, prev->vm_start,
Linus Torvalds's avatar
Linus Torvalds committed
1040
				end, prev->vm_pgoff, NULL);
1041 1042
		if (err)
			return NULL;
Andrea Arcangeli's avatar
Andrea Arcangeli committed
1043
		khugepaged_enter_vma_merge(prev);
Linus Torvalds's avatar
Linus Torvalds committed
1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054
		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 */
1055
			err = vma_adjust(prev, prev->vm_start,
Linus Torvalds's avatar
Linus Torvalds committed
1056 1057
				addr, prev->vm_pgoff, NULL);
		else					/* cases 3, 8 */
1058
			err = vma_adjust(area, addr, next->vm_end,
Linus Torvalds's avatar
Linus Torvalds committed
1059
				next->vm_pgoff - pglen, NULL);
1060 1061
		if (err)
			return NULL;
Andrea Arcangeli's avatar
Andrea Arcangeli committed
1062
		khugepaged_enter_vma_merge(area);
Linus Torvalds's avatar
Linus Torvalds committed
1063 1064 1065 1066 1067 1068
		return area;
	}

	return NULL;
}

1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120 1121 1122 1123
/*
 * 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;
}

Linus Torvalds's avatar
Linus Torvalds committed
1124 1125 1126 1127 1128 1129 1130 1131 1132 1133
/*
 * 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)
{
1134
	struct anon_vma *anon_vma;
Linus Torvalds's avatar
Linus Torvalds committed
1135 1136 1137 1138 1139 1140
	struct vm_area_struct *near;

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

1141 1142 1143
	anon_vma = reusable_anon_vma(near, vma, near);
	if (anon_vma)
		return anon_vma;
Linus Torvalds's avatar
Linus Torvalds committed
1144
try_prev:
1145
	near = vma->vm_prev;
Linus Torvalds's avatar
Linus Torvalds committed
1146 1147 1148
	if (!near)
		goto none;

1149 1150 1151
	anon_vma = reusable_anon_vma(near, near, vma);
	if (anon_vma)
		return anon_vma;
Linus Torvalds's avatar
Linus Torvalds committed
1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164
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
1165
void vm_stat_account(struct mm_struct *mm, unsigned long flags,
Linus Torvalds's avatar
Linus Torvalds committed
1166 1167 1168 1169 1170
						struct file *file, long pages)
{
	const unsigned long stack_flags
		= VM_STACK_FLAGS & (VM_GROWSUP|VM_GROWSDOWN);

1171 1172
	mm->total_vm += pages;

Linus Torvalds's avatar
Linus Torvalds committed
1173 1174 1175 1176 1177 1178 1179 1180 1181
	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 */

Al Viro's avatar
Al Viro committed
1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194
/*
 * 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;
}

Linus Torvalds's avatar
Linus Torvalds committed
1195
/*
1196
 * The caller must hold down_write(&current->mm->mmap_sem).
Linus Torvalds's avatar
Linus Torvalds committed
1197 1198
 */

1199
unsigned long do_mmap_pgoff(struct file *file, unsigned long addr,
Linus Torvalds's avatar
Linus Torvalds committed
1200
			unsigned long len, unsigned long prot,
1201
			unsigned long flags, unsigned long pgoff,
1202
			unsigned long *populate)
Linus Torvalds's avatar
Linus Torvalds committed
1203 1204 1205
{
	struct mm_struct * mm = current->mm;
	struct inode *inode;
1206
	vm_flags_t vm_flags;
Linus Torvalds's avatar
Linus Torvalds committed
1207

1208
	*populate = 0;
1209

Linus Torvalds's avatar
Linus Torvalds committed
1210 1211 1212 1213 1214 1215 1216
	/*
	 * 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))
1217
		if (!(file && (file->f_path.mnt->mnt_flags & MNT_NOEXEC)))
Linus Torvalds's avatar
Linus Torvalds committed
1218 1219 1220 1221 1222
			prot |= PROT_EXEC;

	if (!len)
		return -EINVAL;

1223 1224 1225
	if (!(flags & MAP_FIXED))
		addr = round_hint_to_min(addr);

Linus Torvalds's avatar
Linus Torvalds committed
1226 1227
	/* Careful about overflows.. */
	len = PAGE_ALIGN(len);
1228
	if (!len)
Linus Torvalds's avatar
Linus Torvalds committed
1229 1230 1231 1232 1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 1251 1252
		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;

Huang Shijie's avatar
Huang Shijie committed
1253
	if (flags & MAP_LOCKED)
Linus Torvalds's avatar
Linus Torvalds committed
1254 1255
		if (!can_do_mlock())
			return -EPERM;
1256

Linus Torvalds's avatar
Linus Torvalds committed
1257 1258 1259
	/* mlock MCL_FUTURE? */
	if (vm_flags & VM_LOCKED) {
		unsigned long locked, lock_limit;
1260 1261
		locked = len >> PAGE_SHIFT;
		locked += mm->locked_vm;
Jiri Slaby's avatar
Jiri Slaby committed
1262
		lock_limit = rlimit(RLIMIT_MEMLOCK);
1263
		lock_limit >>= PAGE_SHIFT;
Linus Torvalds's avatar
Linus Torvalds committed
1264 1265 1266 1267
		if (locked > lock_limit && !capable(CAP_IPC_LOCK))
			return -EAGAIN;
	}

Al Viro's avatar
Al Viro committed
1268
	inode = file ? file_inode(file) : NULL;
Linus Torvalds's avatar
Linus Torvalds committed
1269 1270 1271 1272 1273 1274 1275 1276 1277 1278 1279 1280 1281 1282 1283 1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296

	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;
1297
			if (file->f_path.mnt->mnt_flags & MNT_NOEXEC) {
Linus Torvalds's avatar
Linus Torvalds committed
1298 1299 1300 1301 1302 1303 1304
				if (vm_flags & VM_EXEC)
					return -EPERM;
				vm_flags &= ~VM_MAYEXEC;
			}

			if (!file->f_op || !file->f_op->mmap)
				return -ENODEV;
Linus Torvalds's avatar
Linus Torvalds committed
1305 1306 1307 1308 1309 1310 1311 1312
			break;

		default:
			return -EINVAL;
		}
	} else {
		switch (flags & MAP_TYPE) {
		case MAP_SHARED:
1313 1314 1315 1316
			/*
			 * Ignore pgoff.
			 */
			pgoff = 0;
Linus Torvalds's avatar
Linus Torvalds committed
1317 1318 1319 1320 1321 1322 1323 1324 1325 1326 1327 1328 1329
			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;
		}
	}

1330 1331 1332 1333 1334 1335 1336 1337 1338 1339 1340 1341 1342 1343 1344
	/*
	 * Set 'VM_NORESERVE' if we should not account for the
	 * memory use of this mapping.
	 */
	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;
	}

	addr = mmap_region(file, addr, len, vm_flags, pgoff);
1345 1346 1347
	if (!IS_ERR_VALUE(addr) &&
	    ((vm_flags & VM_LOCKED) ||
	     (flags & (MAP_POPULATE | MAP_NONBLOCK)) == MAP_POPULATE))
1348
		*populate = len;
1349
	return addr;
Miklos Szeredi's avatar
Miklos Szeredi committed
1350
}
1351

1352 1353 1354 1355 1356 1357 1358 1359
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)) {
Al Viro's avatar
Al Viro committed
1360
		audit_mmap_fd(fd, flags);
1361 1362 1363
		file = fget(fd);
		if (!file)
			goto out;
1364 1365
		if (is_file_hugepages(file))
			len = ALIGN(len, huge_page_size(hstate_file(file)));
1366 1367 1368
		retval = -EINVAL;
		if (unlikely(flags & MAP_HUGETLB && !is_file_hugepages(file)))
			goto out_fput;
1369 1370
	} else if (flags & MAP_HUGETLB) {
		struct user_struct *user = NULL;
1371
		struct hstate *hs;
1372

1373
		hs = hstate_sizelog((flags >> MAP_HUGE_SHIFT) & SHM_HUGE_MASK);
1374 1375 1376 1377
		if (!hs)
			return -EINVAL;

		len = ALIGN(len, huge_page_size(hs));
1378 1379 1380 1381 1382 1383
		/*
		 * 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
		 */
1384
		file = hugetlb_file_setup(HUGETLB_ANON_FILE, len,
1385 1386 1387
				VM_NORESERVE,
				&user, HUGETLB_ANONHUGE_INODE,
				(flags >> MAP_HUGE_SHIFT) & MAP_HUGE_MASK);
1388 1389 1390 1391 1392 1393
		if (IS_ERR(file))
			return PTR_ERR(file);
	}

	flags &= ~(MAP_EXECUTABLE | MAP_DENYWRITE);

Al Viro's avatar
Al Viro committed
1394
	retval = vm_mmap_pgoff(file, addr, len, prot, flags, pgoff);
1395
out_fput:
1396 1397 1398 1399 1400 1401
	if (file)
		fput(file);
out:
	return retval;
}

1402 1403 1404 1405 1406 1407 1408 1409 1410 1411 1412 1413 1414 1415 1416 1417 1418 1419 1420 1421 1422 1423 1424 1425
#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 */

Alexey Dobriyan's avatar
Alexey Dobriyan committed
1426 1427 1428 1429 1430 1431 1432 1433
/*
 * 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)
{
1434
	vm_flags_t vm_flags = vma->vm_flags;
Alexey Dobriyan's avatar
Alexey Dobriyan committed
1435 1436 1437 1438 1439 1440 1441 1442 1443 1444 1445

	/* 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) !=
1446
	    pgprot_val(vm_get_page_prot(vm_flags)))
Alexey Dobriyan's avatar
Alexey Dobriyan committed
1447 1448 1449
		return 0;

	/* Specialty mapping? */
1450
	if (vm_flags & VM_PFNMAP)
Alexey Dobriyan's avatar
Alexey Dobriyan committed
1451 1452 1453 1454 1455 1456 1457
		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);
}

1458 1459
/*
 * We account for memory if it's a private writeable mapping,
1460
 * not hugepages and VM_NORESERVE wasn't set.
1461
 */
1462
static inline int accountable_mapping(struct file *file, vm_flags_t vm_flags)
1463
{
1464 1465 1466 1467 1468 1469 1470
	/*
	 * 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;

1471 1472 1473
	return (vm_flags & (VM_NORESERVE | VM_SHARED | VM_WRITE)) == VM_WRITE;
}

Miklos Szeredi's avatar
Miklos Szeredi committed
1474
unsigned long mmap_region(struct file *file, unsigned long addr,
1475
		unsigned long len, vm_flags_t vm_flags, unsigned long pgoff)
Miklos Szeredi's avatar
Miklos Szeredi committed
1476 1477 1478 1479 1480 1481 1482
{
	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;
Al Viro's avatar
Al Viro committed
1483
	struct inode *inode =  file ? file_inode(file) : NULL;
Miklos Szeredi's avatar
Miklos Szeredi committed
1484

1485 1486 1487 1488 1489 1490 1491 1492 1493 1494 1495 1496 1497 1498 1499 1500 1501
	/* Check against address space limit. */
	if (!may_expand_vm(mm, len >> PAGE_SHIFT)) {
		unsigned long nr_pages;

		/*
		 * MAP_FIXED may remove pages of mappings that intersects with
		 * requested mapping. Account for the pages it would unmap.
		 */
		if (!(vm_flags & MAP_FIXED))
			return -ENOMEM;

		nr_pages = count_vma_pages_range(mm, addr, addr + len);

		if (!may_expand_vm(mm, (len >> PAGE_SHIFT) - nr_pages))
			return -ENOMEM;
	}

Linus Torvalds's avatar
Linus Torvalds committed
1502 1503 1504
	/* Clear old maps */
	error = -ENOMEM;
munmap_back:
1505
	if (find_vma_links(mm, addr, addr + len, &prev, &rb_link, &rb_parent)) {
Linus Torvalds's avatar
Linus Torvalds committed
1506 1507 1508 1509 1510
		if (do_munmap(mm, addr, len))
			return -ENOMEM;
		goto munmap_back;
	}

1511 1512 1513
	/*
	 * Private writable mapping: check memory availability
	 */
1514
	if (accountable_mapping(file, vm_flags)) {
1515
		charged = len >> PAGE_SHIFT;
1516
		if (security_vm_enough_memory_mm(mm, charged))
1517 1518
			return -ENOMEM;
		vm_flags |= VM_ACCOUNT;
Linus Torvalds's avatar
Linus Torvalds committed
1519 1520 1521
	}

	/*
1522
	 * Can we just expand an old mapping?
Linus Torvalds's avatar
Linus Torvalds committed
1523
	 */
1524 1525 1526
	vma = vma_merge(mm, prev, addr, addr + len, vm_flags, NULL, file, pgoff, NULL);
	if (vma)
		goto out;
Linus Torvalds's avatar
Linus Torvalds committed
1527 1528 1529 1530 1531 1532

	/*
	 * 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.
	 */
1533
	vma = kmem_cache_zalloc(vm_area_cachep, GFP_KERNEL);
Linus Torvalds's avatar
Linus Torvalds committed
1534 1535 1536 1537 1538 1539 1540 1541 1542
	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;