Blame view

mm/mmap.c 89.7 KB
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1
2
3
4
5
  /*
   * mm/mmap.c
   *
   * Written by obz.
   *
046c68842   Alan Cox   mm: update my add...
6
   * Address space accounting code	<alan@lxorguk.ukuu.org.uk>
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
7
   */
b1de0d139   Mitchel Humpherys   mm: convert some ...
8
  #define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
e8420a8ec   Cyril Hrubis   mm/mmap: check fo...
9
  #include <linux/kernel.h>
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
10
  #include <linux/slab.h>
4af3c9cc4   Alexey Dobriyan   Drop some headers...
11
  #include <linux/backing-dev.h>
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
12
  #include <linux/mm.h>
615d6e875   Davidlohr Bueso   mm: per-thread vm...
13
  #include <linux/vmacache.h>
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
14
15
16
17
18
  #include <linux/shm.h>
  #include <linux/mman.h>
  #include <linux/pagemap.h>
  #include <linux/swap.h>
  #include <linux/syscalls.h>
c59ede7b7   Randy.Dunlap   [PATCH] move capa...
19
  #include <linux/capability.h>
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
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>
c01d5b300   Hugh Dickins   shmem: get_unmapp...
26
  #include <linux/shmem_fs.h>
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
27
  #include <linux/profile.h>
b95f1b31b   Paul Gortmaker   mm: Map most file...
28
  #include <linux/export.h>
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
29
30
31
  #include <linux/mount.h>
  #include <linux/mempolicy.h>
  #include <linux/rmap.h>
cddb8a5c1   Andrea Arcangeli   mmu-notifiers: core
32
  #include <linux/mmu_notifier.h>
82f71ae4a   Konstantin Khlebnikov   mm: catch memory ...
33
  #include <linux/mmdebug.h>
cdd6c482c   Ingo Molnar   perf: Do the big ...
34
  #include <linux/perf_event.h>
120a795da   Al Viro   audit mmap
35
  #include <linux/audit.h>
b15d00b6a   Andrea Arcangeli   thp: khugepaged v...
36
  #include <linux/khugepaged.h>
2b1444983   Srikar Dronamraju   uprobes, mm, x86:...
37
  #include <linux/uprobes.h>
d37371870   Michel Lespinasse   mm: augment vma r...
38
  #include <linux/rbtree_augmented.h>
1640879af   Andrew Shewmaker   mm: reinititalise...
39
40
  #include <linux/notifier.h>
  #include <linux/memory.h>
b1de0d139   Mitchel Humpherys   mm: convert some ...
41
  #include <linux/printk.h>
19a809afe   Andrea Arcangeli   userfaultfd: teac...
42
  #include <linux/userfaultfd_k.h>
d977d56ce   Konstantin Khlebnikov   mm: warn about Vm...
43
  #include <linux/moduleparam.h>
62b5f7d01   Dave Hansen   mm/core, x86/mm/p...
44
  #include <linux/pkeys.h>
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
45
46
47
48
  
  #include <asm/uaccess.h>
  #include <asm/cacheflush.h>
  #include <asm/tlb.h>
d6dd61c83   Jeremy Fitzhardinge   [PATCH] x86: PARA...
49
  #include <asm/mmu_context.h>
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
50

42b777281   Jan Beulich   mm: remove double...
51
  #include "internal.h"
3a4597568   Kirill Korotaev   [PATCH] IA64,spar...
52
53
54
  #ifndef arch_mmap_check
  #define arch_mmap_check(addr, len, flags)	(0)
  #endif
d07e22597   Daniel Cashman   mm: mmap: add new...
55
56
57
58
59
60
61
62
63
64
  #ifdef CONFIG_HAVE_ARCH_MMAP_RND_BITS
  const int mmap_rnd_bits_min = CONFIG_ARCH_MMAP_RND_BITS_MIN;
  const int mmap_rnd_bits_max = CONFIG_ARCH_MMAP_RND_BITS_MAX;
  int mmap_rnd_bits __read_mostly = CONFIG_ARCH_MMAP_RND_BITS;
  #endif
  #ifdef CONFIG_HAVE_ARCH_MMAP_RND_COMPAT_BITS
  const int mmap_rnd_compat_bits_min = CONFIG_ARCH_MMAP_RND_COMPAT_BITS_MIN;
  const int mmap_rnd_compat_bits_max = CONFIG_ARCH_MMAP_RND_COMPAT_BITS_MAX;
  int mmap_rnd_compat_bits __read_mostly = CONFIG_ARCH_MMAP_RND_COMPAT_BITS;
  #endif
f4fcd5584   Konstantin Khlebnikov   mm: enable RLIMIT...
65
  static bool ignore_rlimit_data;
d977d56ce   Konstantin Khlebnikov   mm: warn about Vm...
66
  core_param(ignore_rlimit_data, ignore_rlimit_data, bool, 0644);
d07e22597   Daniel Cashman   mm: mmap: add new...
67

e0da382c9   Hugh Dickins   [PATCH] freepgt: ...
68
69
70
  static void unmap_region(struct mm_struct *mm,
  		struct vm_area_struct *vma, struct vm_area_struct *prev,
  		unsigned long start, unsigned long end);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
71
72
73
74
75
76
77
78
79
  /* 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
cc71aba34   vishnu.ps   mm/mmap.c: whites...
80
   *
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
81
82
83
84
85
86
87
88
89
   * 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
  };
804af2cf6   Hugh Dickins   [AGPGART] remove ...
90
91
  pgprot_t vm_get_page_prot(unsigned long vm_flags)
  {
b845f313d   Dave Kleikamp   mm: Allow archite...
92
93
94
  	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)));
804af2cf6   Hugh Dickins   [AGPGART] remove ...
95
96
  }
  EXPORT_SYMBOL(vm_get_page_prot);
64e455079   Peter Feiner   mm: softdirty: en...
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
  static pgprot_t vm_pgprot_modify(pgprot_t oldprot, unsigned long vm_flags)
  {
  	return pgprot_modify(oldprot, vm_get_page_prot(vm_flags));
  }
  
  /* Update vma->vm_page_prot to reflect vma->vm_flags. */
  void vma_set_page_prot(struct vm_area_struct *vma)
  {
  	unsigned long vm_flags = vma->vm_flags;
  
  	vma->vm_page_prot = vm_pgprot_modify(vma->vm_page_prot, vm_flags);
  	if (vma_wants_writenotify(vma)) {
  		vm_flags &= ~VM_SHARED;
  		vma->vm_page_prot = vm_pgprot_modify(vma->vm_page_prot,
  						     vm_flags);
  	}
  }
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
114
  /*
c8c06efa8   Davidlohr Bueso   mm: convert i_mma...
115
   * Requires inode->i_mapping->i_mmap_rwsem
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
116
117
118
119
120
   */
  static void __remove_shared_vm_struct(struct vm_area_struct *vma,
  		struct file *file, struct address_space *mapping)
  {
  	if (vma->vm_flags & VM_DENYWRITE)
496ad9aa8   Al Viro   new helper: file_...
121
  		atomic_inc(&file_inode(file)->i_writecount);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
122
  	if (vma->vm_flags & VM_SHARED)
4bb5f5d93   David Herrmann   mm: allow drivers...
123
  		mapping_unmap_writable(mapping);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
124
125
  
  	flush_dcache_mmap_lock(mapping);
27ba0644e   Kirill A. Shutemov   rmap: drop suppor...
126
  	vma_interval_tree_remove(vma, &mapping->i_mmap);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
127
128
129
130
  	flush_dcache_mmap_unlock(mapping);
  }
  
  /*
6b2dbba8b   Michel Lespinasse   mm: replace vma p...
131
   * Unlink a file-based vm structure from its interval tree, to hide
a8fb5618d   Hugh Dickins   [PATCH] mm: unlin...
132
   * vma from rmap and vmtruncate before freeing its page tables.
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
133
   */
a8fb5618d   Hugh Dickins   [PATCH] mm: unlin...
134
  void unlink_file_vma(struct vm_area_struct *vma)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
135
136
  {
  	struct file *file = vma->vm_file;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
137
138
  	if (file) {
  		struct address_space *mapping = file->f_mapping;
83cde9e8b   Davidlohr Bueso   mm: use new helpe...
139
  		i_mmap_lock_write(mapping);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
140
  		__remove_shared_vm_struct(vma, file, mapping);
83cde9e8b   Davidlohr Bueso   mm: use new helpe...
141
  		i_mmap_unlock_write(mapping);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
142
  	}
a8fb5618d   Hugh Dickins   [PATCH] mm: unlin...
143
144
145
146
147
148
149
150
  }
  
  /*
   * 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;
a8fb5618d   Hugh Dickins   [PATCH] mm: unlin...
151
  	might_sleep();
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
152
153
  	if (vma->vm_ops && vma->vm_ops->close)
  		vma->vm_ops->close(vma);
e9714acf8   Konstantin Khlebnikov   mm: kill vma flag...
154
  	if (vma->vm_file)
a8fb5618d   Hugh Dickins   [PATCH] mm: unlin...
155
  		fput(vma->vm_file);
f0be3d32b   Lee Schermerhorn   mempolicy: rename...
156
  	mpol_put(vma_policy(vma));
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
157
  	kmem_cache_free(vm_area_cachep, vma);
a8fb5618d   Hugh Dickins   [PATCH] mm: unlin...
158
  	return next;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
159
  }
5d22fc25d   Linus Torvalds   mm: remove more I...
160
  static int do_brk(unsigned long addr, unsigned long len);
e4eb1ff61   Linus Torvalds   VM: add "vm_brk()...
161

6a6160a7b   Heiko Carstens   [CVE-2009-0029] S...
162
  SYSCALL_DEFINE1(brk, unsigned long, brk)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
163
  {
8764b338b   Cyrill Gorcunov   mm: use may_adjus...
164
  	unsigned long retval;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
165
166
  	unsigned long newbrk, oldbrk;
  	struct mm_struct *mm = current->mm;
a5b4592cf   Jiri Kosina   brk: make sys_brk...
167
  	unsigned long min_brk;
128557ffe   Michel Lespinasse   mm: use mm_popula...
168
  	bool populate;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
169

dc0ef0df7   Michal Hocko   mm: make mmap_sem...
170
171
  	if (down_write_killable(&mm->mmap_sem))
  		return -EINTR;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
172

a5b4592cf   Jiri Kosina   brk: make sys_brk...
173
  #ifdef CONFIG_COMPAT_BRK
5520e8948   Jiri Kosina   brk: fix min_brk ...
174
175
176
177
178
  	/*
  	 * 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
  	 */
4471a675d   Jiri Kosina   brk: COMPAT_BRK: ...
179
  	if (current->brk_randomized)
5520e8948   Jiri Kosina   brk: fix min_brk ...
180
181
182
  		min_brk = mm->start_brk;
  	else
  		min_brk = mm->end_data;
a5b4592cf   Jiri Kosina   brk: make sys_brk...
183
184
185
186
  #else
  	min_brk = mm->start_brk;
  #endif
  	if (brk < min_brk)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
187
  		goto out;
1e624196f   Ram Gupta   [PATCH] mm: fix b...
188
189
190
191
192
193
194
  
  	/*
  	 * 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
  	 */
8764b338b   Cyrill Gorcunov   mm: use may_adjus...
195
196
  	if (check_data_rlimit(rlimit(RLIMIT_DATA), brk, mm->start_brk,
  			      mm->end_data, mm->start_data))
1e624196f   Ram Gupta   [PATCH] mm: fix b...
197
  		goto out;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
198
199
200
201
202
203
204
205
206
207
208
  	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;
  	}
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
209
210
211
212
213
  	/* Check against existing mmap mappings. */
  	if (find_vma_intersection(mm, oldbrk, newbrk+PAGE_SIZE))
  		goto out;
  
  	/* Ok, looks good - let it rip. */
5d22fc25d   Linus Torvalds   mm: remove more I...
214
  	if (do_brk(oldbrk, newbrk-oldbrk) < 0)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
215
  		goto out;
128557ffe   Michel Lespinasse   mm: use mm_popula...
216

1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
217
218
  set_brk:
  	mm->brk = brk;
128557ffe   Michel Lespinasse   mm: use mm_popula...
219
220
221
222
223
  	populate = newbrk > oldbrk && (mm->def_flags & VM_LOCKED) != 0;
  	up_write(&mm->mmap_sem);
  	if (populate)
  		mm_populate(oldbrk, newbrk - oldbrk);
  	return brk;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
224
225
226
227
228
  out:
  	retval = mm->brk;
  	up_write(&mm->mmap_sem);
  	return retval;
  }
d37371870   Michel Lespinasse   mm: augment vma r...
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
  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;
  }
ed8ea8150   Michel Lespinasse   mm: add CONFIG_DE...
249
  #ifdef CONFIG_DEBUG_VM_RB
acf128d04   Andrea Arcangeli   mm: validate_mm b...
250
  static int browse_rb(struct mm_struct *mm)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
251
  {
acf128d04   Andrea Arcangeli   mm: validate_mm b...
252
  	struct rb_root *root = &mm->mm_rb;
5a0768f64   Michel Lespinasse   mm: check rb_subt...
253
  	int i = 0, j, bug = 0;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
254
255
256
257
258
259
  	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);
5a0768f64   Michel Lespinasse   mm: check rb_subt...
260
  		if (vma->vm_start < prev) {
ff26f70f4   Andrew Morton   mm/mmap.c: clean ...
261
262
263
  			pr_emerg("vm_start %lx < prev %lx
  ",
  				  vma->vm_start, prev);
5a0768f64   Michel Lespinasse   mm: check rb_subt...
264
265
266
  			bug = 1;
  		}
  		if (vma->vm_start < pend) {
ff26f70f4   Andrew Morton   mm/mmap.c: clean ...
267
268
269
  			pr_emerg("vm_start %lx < pend %lx
  ",
  				  vma->vm_start, pend);
5a0768f64   Michel Lespinasse   mm: check rb_subt...
270
271
272
  			bug = 1;
  		}
  		if (vma->vm_start > vma->vm_end) {
ff26f70f4   Andrew Morton   mm/mmap.c: clean ...
273
274
275
  			pr_emerg("vm_start %lx > vm_end %lx
  ",
  				  vma->vm_start, vma->vm_end);
5a0768f64   Michel Lespinasse   mm: check rb_subt...
276
277
  			bug = 1;
  		}
acf128d04   Andrea Arcangeli   mm: validate_mm b...
278
  		spin_lock(&mm->page_table_lock);
5a0768f64   Michel Lespinasse   mm: check rb_subt...
279
  		if (vma->rb_subtree_gap != vma_compute_subtree_gap(vma)) {
8542bdfc6   Sasha Levin   mm/mmap.c: use pr...
280
281
  			pr_emerg("free gap %lx, correct %lx
  ",
5a0768f64   Michel Lespinasse   mm: check rb_subt...
282
283
284
285
  			       vma->rb_subtree_gap,
  			       vma_compute_subtree_gap(vma));
  			bug = 1;
  		}
acf128d04   Andrea Arcangeli   mm: validate_mm b...
286
  		spin_unlock(&mm->page_table_lock);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
287
288
  		i++;
  		pn = nd;
d1af65d13   David Miller   [PATCH] Bug in MM...
289
290
  		prev = vma->vm_start;
  		pend = vma->vm_end;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
291
292
  	}
  	j = 0;
5a0768f64   Michel Lespinasse   mm: check rb_subt...
293
  	for (nd = pn; nd; nd = rb_prev(nd))
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
294
  		j++;
5a0768f64   Michel Lespinasse   mm: check rb_subt...
295
  	if (i != j) {
8542bdfc6   Sasha Levin   mm/mmap.c: use pr...
296
297
  		pr_emerg("backwards %d, forwards %d
  ", j, i);
5a0768f64   Michel Lespinasse   mm: check rb_subt...
298
  		bug = 1;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
299
  	}
5a0768f64   Michel Lespinasse   mm: check rb_subt...
300
  	return bug ? -1 : i;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
301
  }
d37371870   Michel Lespinasse   mm: augment vma r...
302
303
304
305
306
307
308
  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);
96dad67ff   Sasha Levin   mm: use VM_BUG_ON...
309
310
311
  		VM_BUG_ON_VMA(vma != ignore &&
  			vma->rb_subtree_gap != vma_compute_subtree_gap(vma),
  			vma);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
312
  	}
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
313
  }
eafd4dc4d   Rashika Kheria   mm/mmap.c: mark f...
314
  static void validate_mm(struct mm_struct *mm)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
315
316
317
  {
  	int bug = 0;
  	int i = 0;
5a0768f64   Michel Lespinasse   mm: check rb_subt...
318
  	unsigned long highest_address = 0;
ed8ea8150   Michel Lespinasse   mm: add CONFIG_DE...
319
  	struct vm_area_struct *vma = mm->mmap;
ff26f70f4   Andrew Morton   mm/mmap.c: clean ...
320

ed8ea8150   Michel Lespinasse   mm: add CONFIG_DE...
321
  	while (vma) {
12352d3ca   Konstantin Khlebnikov   mm: replace vma_l...
322
  		struct anon_vma *anon_vma = vma->anon_vma;
ed8ea8150   Michel Lespinasse   mm: add CONFIG_DE...
323
  		struct anon_vma_chain *avc;
ff26f70f4   Andrew Morton   mm/mmap.c: clean ...
324

12352d3ca   Konstantin Khlebnikov   mm: replace vma_l...
325
326
327
328
329
330
  		if (anon_vma) {
  			anon_vma_lock_read(anon_vma);
  			list_for_each_entry(avc, &vma->anon_vma_chain, same_vma)
  				anon_vma_interval_tree_verify(avc);
  			anon_vma_unlock_read(anon_vma);
  		}
5a0768f64   Michel Lespinasse   mm: check rb_subt...
331
  		highest_address = vma->vm_end;
ed8ea8150   Michel Lespinasse   mm: add CONFIG_DE...
332
  		vma = vma->vm_next;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
333
334
  		i++;
  	}
5a0768f64   Michel Lespinasse   mm: check rb_subt...
335
  	if (i != mm->map_count) {
8542bdfc6   Sasha Levin   mm/mmap.c: use pr...
336
337
  		pr_emerg("map_count %d vm_next %d
  ", mm->map_count, i);
5a0768f64   Michel Lespinasse   mm: check rb_subt...
338
339
340
  		bug = 1;
  	}
  	if (highest_address != mm->highest_vm_end) {
8542bdfc6   Sasha Levin   mm/mmap.c: use pr...
341
342
  		pr_emerg("mm->highest_vm_end %lx, found %lx
  ",
ff26f70f4   Andrew Morton   mm/mmap.c: clean ...
343
  			  mm->highest_vm_end, highest_address);
5a0768f64   Michel Lespinasse   mm: check rb_subt...
344
345
  		bug = 1;
  	}
acf128d04   Andrea Arcangeli   mm: validate_mm b...
346
  	i = browse_rb(mm);
5a0768f64   Michel Lespinasse   mm: check rb_subt...
347
  	if (i != mm->map_count) {
ff26f70f4   Andrew Morton   mm/mmap.c: clean ...
348
349
350
  		if (i != -1)
  			pr_emerg("map_count %d rb %d
  ", mm->map_count, i);
5a0768f64   Michel Lespinasse   mm: check rb_subt...
351
352
  		bug = 1;
  	}
96dad67ff   Sasha Levin   mm: use VM_BUG_ON...
353
  	VM_BUG_ON_MM(bug, mm);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
354
355
  }
  #else
d37371870   Michel Lespinasse   mm: augment vma r...
356
  #define validate_mm_rb(root, ignore) do { } while (0)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
357
358
  #define validate_mm(mm) do { } while (0)
  #endif
d37371870   Michel Lespinasse   mm: augment vma r...
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
  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);
  }
bf181b9f9   Michel Lespinasse   mm anon rmap: rep...
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
  /*
   * 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);
  }
6597d7833   Hugh Dickins   mm/mmap.c: replac...
431
432
433
  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)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
434
  {
6597d7833   Hugh Dickins   mm/mmap.c: replac...
435
  	struct rb_node **__rb_link, *__rb_parent, *rb_prev;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
436
437
438
  
  	__rb_link = &mm->mm_rb.rb_node;
  	rb_prev = __rb_parent = NULL;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
439
440
441
442
443
444
445
446
  
  	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) {
6597d7833   Hugh Dickins   mm/mmap.c: replac...
447
448
449
  			/* Fail if an existing vma overlaps the area */
  			if (vma_tmp->vm_start < end)
  				return -ENOMEM;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
450
451
452
453
454
455
456
457
458
459
460
461
  			__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;
6597d7833   Hugh Dickins   mm/mmap.c: replac...
462
  	return 0;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
463
  }
e8420a8ec   Cyril Hrubis   mm/mmap: check fo...
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
  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;
  }
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
491
492
493
  void __vma_link_rb(struct mm_struct *mm, struct vm_area_struct *vma,
  		struct rb_node **rb_link, struct rb_node *rb_parent)
  {
d37371870   Michel Lespinasse   mm: augment vma r...
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
  	/* 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.
  	 */
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
509
  	rb_link_node(&vma->vm_rb, rb_parent, rb_link);
d37371870   Michel Lespinasse   mm: augment vma r...
510
511
512
  	vma->rb_subtree_gap = 0;
  	vma_gap_update(vma);
  	vma_rb_insert(vma, &mm->mm_rb);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
513
  }
cb8f488c3   Denys Vlasenko   mmap.c: deinline ...
514
  static void __vma_link_file(struct vm_area_struct *vma)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
515
  {
48aae4255   ZhenwenXu   mm/mmap.c: fix co...
516
  	struct file *file;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
517
518
519
520
521
522
  
  	file = vma->vm_file;
  	if (file) {
  		struct address_space *mapping = file->f_mapping;
  
  		if (vma->vm_flags & VM_DENYWRITE)
496ad9aa8   Al Viro   new helper: file_...
523
  			atomic_dec(&file_inode(file)->i_writecount);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
524
  		if (vma->vm_flags & VM_SHARED)
4bb5f5d93   David Herrmann   mm: allow drivers...
525
  			atomic_inc(&mapping->i_mmap_writable);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
526
527
  
  		flush_dcache_mmap_lock(mapping);
27ba0644e   Kirill A. Shutemov   rmap: drop suppor...
528
  		vma_interval_tree_insert(vma, &mapping->i_mmap);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
529
530
531
532
533
534
535
536
537
538
539
  		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);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
540
541
542
543
544
545
546
  }
  
  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;
64ac4940d   Huang Shijie   mm/mmap.c: remove...
547
  	if (vma->vm_file) {
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
548
  		mapping = vma->vm_file->f_mapping;
83cde9e8b   Davidlohr Bueso   mm: use new helpe...
549
  		i_mmap_lock_write(mapping);
64ac4940d   Huang Shijie   mm/mmap.c: remove...
550
  	}
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
551
552
553
  
  	__vma_link(mm, vma, prev, rb_link, rb_parent);
  	__vma_link_file(vma);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
554
  	if (mapping)
83cde9e8b   Davidlohr Bueso   mm: use new helpe...
555
  		i_mmap_unlock_write(mapping);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
556
557
558
559
560
561
  
  	mm->map_count++;
  	validate_mm(mm);
  }
  
  /*
88f6b4c32   Kautuk Consul   mmap.c: fix comme...
562
   * Helper for vma_adjust() in the split_vma insert case: insert a vma into the
6b2dbba8b   Michel Lespinasse   mm: replace vma p...
563
   * mm's list and rbtree.  It has already been inserted into the interval tree.
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
564
   */
48aae4255   ZhenwenXu   mm/mmap.c: fix co...
565
  static void __insert_vm_struct(struct mm_struct *mm, struct vm_area_struct *vma)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
566
  {
6597d7833   Hugh Dickins   mm/mmap.c: replac...
567
  	struct vm_area_struct *prev;
48aae4255   ZhenwenXu   mm/mmap.c: fix co...
568
  	struct rb_node **rb_link, *rb_parent;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
569

6597d7833   Hugh Dickins   mm/mmap.c: replac...
570
571
572
  	if (find_vma_links(mm, vma->vm_start, vma->vm_end,
  			   &prev, &rb_link, &rb_parent))
  		BUG();
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
573
574
575
576
577
578
579
580
  	__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)
  {
d37371870   Michel Lespinasse   mm: augment vma r...
581
  	struct vm_area_struct *next;
297c5eee3   Linus Torvalds   mm: make the vma ...
582

d37371870   Michel Lespinasse   mm: augment vma r...
583
584
  	vma_rb_erase(vma, &mm->mm_rb);
  	prev->vm_next = next = vma->vm_next;
297c5eee3   Linus Torvalds   mm: make the vma ...
585
586
  	if (next)
  		next->vm_prev = prev;
615d6e875   Davidlohr Bueso   mm: per-thread vm...
587
588
589
  
  	/* Kill the cache */
  	vmacache_invalidate(mm);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
590
591
592
593
594
595
596
597
598
  }
  
  /*
   * 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.
   */
5beb49305   Rik van Riel   mm: change anon_v...
599
  int vma_adjust(struct vm_area_struct *vma, unsigned long start,
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
600
601
602
603
  	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;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
604
  	struct address_space *mapping = NULL;
6b2dbba8b   Michel Lespinasse   mm: replace vma p...
605
  	struct rb_root *root = NULL;
012f18004   Rik van Riel   mm: always lock t...
606
  	struct anon_vma *anon_vma = NULL;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
607
  	struct file *file = vma->vm_file;
d37371870   Michel Lespinasse   mm: augment vma r...
608
  	bool start_changed = false, end_changed = false;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
609
610
611
612
  	long adjust_next = 0;
  	int remove_next = 0;
  
  	if (next && !insert) {
734537c9c   Kirill A. Shutemov   mm: fix use-after...
613
  		struct vm_area_struct *exporter = NULL, *importer = NULL;
287d97ac0   Linus Torvalds   vma_adjust: fix t...
614

1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
615
616
617
618
619
  		if (end >= next->vm_end) {
  			/*
  			 * vma expands, overlapping all the next, and
  			 * perhaps the one after too (mprotect case 6).
  			 */
734537c9c   Kirill A. Shutemov   mm: fix use-after...
620
  			remove_next = 1 + (end > next->vm_end);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
621
  			end = next->vm_end;
287d97ac0   Linus Torvalds   vma_adjust: fix t...
622
  			exporter = next;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
623
  			importer = vma;
734537c9c   Kirill A. Shutemov   mm: fix use-after...
624
625
626
627
628
629
630
  
  			/*
  			 * If next doesn't have anon_vma, import from vma after
  			 * next, if the vma overlaps with it.
  			 */
  			if (remove_next == 2 && next && !next->anon_vma)
  				exporter = next->vm_next;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
631
632
633
634
635
636
  		} 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;
287d97ac0   Linus Torvalds   vma_adjust: fix t...
637
  			exporter = next;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
638
639
640
641
642
643
644
  			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.
  			 */
cc71aba34   vishnu.ps   mm/mmap.c: whites...
645
  			adjust_next = -((vma->vm_end - end) >> PAGE_SHIFT);
287d97ac0   Linus Torvalds   vma_adjust: fix t...
646
  			exporter = vma;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
647
648
  			importer = next;
  		}
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
649

5beb49305   Rik van Riel   mm: change anon_v...
650
651
652
653
654
  		/*
  		 * 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.
  		 */
287d97ac0   Linus Torvalds   vma_adjust: fix t...
655
  		if (exporter && exporter->anon_vma && !importer->anon_vma) {
c4ea95d7c   Daniel Forrest   mm: fix anon_vma_...
656
  			int error;
b800c91a0   Konstantin Khlebnikov   mm: fix corner ca...
657
  			importer->anon_vma = exporter->anon_vma;
c4ea95d7c   Daniel Forrest   mm: fix anon_vma_...
658
  			error = anon_vma_clone(importer, exporter);
3fe89b3e2   Leon Yu   mm: fix anon_vma-...
659
  			if (error)
c4ea95d7c   Daniel Forrest   mm: fix anon_vma_...
660
  				return error;
5beb49305   Rik van Riel   mm: change anon_v...
661
662
  		}
  	}
734537c9c   Kirill A. Shutemov   mm: fix use-after...
663
  again:
37f9f5595   Kirill A. Shutemov   thp: run vma_adju...
664
  	vma_adjust_trans_huge(vma, start, end, adjust_next);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
665
666
  	if (file) {
  		mapping = file->f_mapping;
27ba0644e   Kirill A. Shutemov   rmap: drop suppor...
667
668
  		root = &mapping->i_mmap;
  		uprobe_munmap(vma, vma->vm_start, vma->vm_end);
682968e0c   Srikar Dronamraju   uprobes/core: Opt...
669

27ba0644e   Kirill A. Shutemov   rmap: drop suppor...
670
671
  		if (adjust_next)
  			uprobe_munmap(next, next->vm_start, next->vm_end);
682968e0c   Srikar Dronamraju   uprobes/core: Opt...
672

83cde9e8b   Davidlohr Bueso   mm: use new helpe...
673
  		i_mmap_lock_write(mapping);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
674
  		if (insert) {
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
675
  			/*
6b2dbba8b   Michel Lespinasse   mm: replace vma p...
676
  			 * Put into interval tree now, so instantiated pages
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
677
678
679
680
681
682
683
  			 * 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);
  		}
  	}
bf181b9f9   Michel Lespinasse   mm anon rmap: rep...
684
685
686
687
  	anon_vma = vma->anon_vma;
  	if (!anon_vma && adjust_next)
  		anon_vma = next->anon_vma;
  	if (anon_vma) {
81d1b09c6   Sasha Levin   mm: convert a few...
688
689
  		VM_BUG_ON_VMA(adjust_next && next->anon_vma &&
  			  anon_vma != next->anon_vma, next);
4fc3f1d66   Ingo Molnar   mm/rmap, migratio...
690
  		anon_vma_lock_write(anon_vma);
bf181b9f9   Michel Lespinasse   mm anon rmap: rep...
691
692
693
694
  		anon_vma_interval_tree_pre_update_vma(vma);
  		if (adjust_next)
  			anon_vma_interval_tree_pre_update_vma(next);
  	}
012f18004   Rik van Riel   mm: always lock t...
695

1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
696
697
  	if (root) {
  		flush_dcache_mmap_lock(mapping);
6b2dbba8b   Michel Lespinasse   mm: replace vma p...
698
  		vma_interval_tree_remove(vma, root);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
699
  		if (adjust_next)
6b2dbba8b   Michel Lespinasse   mm: replace vma p...
700
  			vma_interval_tree_remove(next, root);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
701
  	}
d37371870   Michel Lespinasse   mm: augment vma r...
702
703
704
705
706
707
708
709
  	if (start != vma->vm_start) {
  		vma->vm_start = start;
  		start_changed = true;
  	}
  	if (end != vma->vm_end) {
  		vma->vm_end = end;
  		end_changed = true;
  	}
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
710
711
712
713
714
715
716
717
  	vma->vm_pgoff = pgoff;
  	if (adjust_next) {
  		next->vm_start += adjust_next << PAGE_SHIFT;
  		next->vm_pgoff += adjust_next;
  	}
  
  	if (root) {
  		if (adjust_next)
6b2dbba8b   Michel Lespinasse   mm: replace vma p...
718
719
  			vma_interval_tree_insert(next, root);
  		vma_interval_tree_insert(vma, root);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
720
721
722
723
724
725
726
727
728
729
730
  		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);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
731
732
733
734
735
736
737
  	} 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);
d37371870   Michel Lespinasse   mm: augment vma r...
738
739
740
741
742
743
744
745
746
  	} 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);
  		}
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
747
  	}
bf181b9f9   Michel Lespinasse   mm anon rmap: rep...
748
749
750
751
  	if (anon_vma) {
  		anon_vma_interval_tree_post_update_vma(vma);
  		if (adjust_next)
  			anon_vma_interval_tree_post_update_vma(next);
08b52706d   Konstantin Khlebnikov   mm/rmap: rename a...
752
  		anon_vma_unlock_write(anon_vma);
bf181b9f9   Michel Lespinasse   mm anon rmap: rep...
753
  	}
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
754
  	if (mapping)
83cde9e8b   Davidlohr Bueso   mm: use new helpe...
755
  		i_mmap_unlock_write(mapping);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
756

2b1444983   Srikar Dronamraju   uprobes, mm, x86:...
757
  	if (root) {
7b2d81d48   Ingo Molnar   uprobes/core: Cle...
758
  		uprobe_mmap(vma);
2b1444983   Srikar Dronamraju   uprobes, mm, x86:...
759
760
  
  		if (adjust_next)
7b2d81d48   Ingo Molnar   uprobes/core: Cle...
761
  			uprobe_mmap(next);
2b1444983   Srikar Dronamraju   uprobes, mm, x86:...
762
  	}
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
763
  	if (remove_next) {
925d1c401   Matt Helsley   procfs task exe s...
764
  		if (file) {
cbc91f71b   Srikar Dronamraju   uprobes/core: Dec...
765
  			uprobe_munmap(next, next->vm_start, next->vm_end);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
766
  			fput(file);
925d1c401   Matt Helsley   procfs task exe s...
767
  		}
5beb49305   Rik van Riel   mm: change anon_v...
768
769
  		if (next->anon_vma)
  			anon_vma_merge(vma, next);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
770
  		mm->map_count--;
3964acd0d   Oleg Nesterov   mm: mempolicy: fi...
771
  		mpol_put(vma_policy(next));
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
772
773
774
775
776
777
  		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.
  		 */
d37371870   Michel Lespinasse   mm: augment vma r...
778
  		next = vma->vm_next;
734537c9c   Kirill A. Shutemov   mm: fix use-after...
779
780
781
  		if (remove_next == 2) {
  			remove_next = 1;
  			end = next->vm_end;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
782
  			goto again;
734537c9c   Kirill A. Shutemov   mm: fix use-after...
783
  		}
d37371870   Michel Lespinasse   mm: augment vma r...
784
785
786
787
  		else if (next)
  			vma_gap_update(next);
  		else
  			mm->highest_vm_end = end;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
788
  	}
2b1444983   Srikar Dronamraju   uprobes, mm, x86:...
789
  	if (insert && file)
7b2d81d48   Ingo Molnar   uprobes/core: Cle...
790
  		uprobe_mmap(insert);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
791
792
  
  	validate_mm(mm);
5beb49305   Rik van Riel   mm: change anon_v...
793
794
  
  	return 0;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
795
796
797
798
799
800
  }
  
  /*
   * 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.
   */
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
801
  static inline int is_mergeable_vma(struct vm_area_struct *vma,
19a809afe   Andrea Arcangeli   userfaultfd: teac...
802
803
  				struct file *file, unsigned long vm_flags,
  				struct vm_userfaultfd_ctx vm_userfaultfd_ctx)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
804
  {
34228d473   Cyrill Gorcunov   mm: ignore VM_SOF...
805
806
807
808
809
810
811
812
813
  	/*
  	 * VM_SOFTDIRTY should not prevent from VMA merging, if we
  	 * match the flags but dirty bit -- the caller should mark
  	 * merged VMA as dirty. If dirty bit won't be excluded from
  	 * comparison, we increase pressue on the memory system forcing
  	 * the kernel to generate new VMAs when old one could be
  	 * extended instead.
  	 */
  	if ((vma->vm_flags ^ vm_flags) & ~VM_SOFTDIRTY)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
814
815
816
817
818
  		return 0;
  	if (vma->vm_file != file)
  		return 0;
  	if (vma->vm_ops && vma->vm_ops->close)
  		return 0;
19a809afe   Andrea Arcangeli   userfaultfd: teac...
819
820
  	if (!is_mergeable_vm_userfaultfd_ctx(vma, vm_userfaultfd_ctx))
  		return 0;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
821
822
823
824
  	return 1;
  }
  
  static inline int is_mergeable_anon_vma(struct anon_vma *anon_vma1,
965f55dea   Shaohua Li   mmap: avoid mergi...
825
826
  					struct anon_vma *anon_vma2,
  					struct vm_area_struct *vma)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
827
  {
965f55dea   Shaohua Li   mmap: avoid mergi...
828
829
830
831
832
833
834
835
  	/*
  	 * 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;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
  }
  
  /*
   * 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,
19a809afe   Andrea Arcangeli   userfaultfd: teac...
851
852
853
  		     struct anon_vma *anon_vma, struct file *file,
  		     pgoff_t vm_pgoff,
  		     struct vm_userfaultfd_ctx vm_userfaultfd_ctx)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
854
  {
19a809afe   Andrea Arcangeli   userfaultfd: teac...
855
  	if (is_mergeable_vma(vma, file, vm_flags, vm_userfaultfd_ctx) &&
965f55dea   Shaohua Li   mmap: avoid mergi...
856
  	    is_mergeable_anon_vma(anon_vma, vma->anon_vma, vma)) {
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
  		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,
19a809afe   Andrea Arcangeli   userfaultfd: teac...
872
873
874
  		    struct anon_vma *anon_vma, struct file *file,
  		    pgoff_t vm_pgoff,
  		    struct vm_userfaultfd_ctx vm_userfaultfd_ctx)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
875
  {
19a809afe   Andrea Arcangeli   userfaultfd: teac...
876
  	if (is_mergeable_vma(vma, file, vm_flags, vm_userfaultfd_ctx) &&
965f55dea   Shaohua Li   mmap: avoid mergi...
877
  	    is_mergeable_anon_vma(anon_vma, vma->anon_vma, vma)) {
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
878
  		pgoff_t vm_pglen;
d6e932177   Libin   mm: use vma_pages...
879
  		vm_pglen = vma_pages(vma);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
  		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,
cc71aba34   vishnu.ps   mm/mmap.c: whites...
918
  			struct anon_vma *anon_vma, struct file *file,
19a809afe   Andrea Arcangeli   userfaultfd: teac...
919
920
  			pgoff_t pgoff, struct mempolicy *policy,
  			struct vm_userfaultfd_ctx vm_userfaultfd_ctx)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
921
922
923
  {
  	pgoff_t pglen = (end - addr) >> PAGE_SHIFT;
  	struct vm_area_struct *area, *next;
5beb49305   Rik van Riel   mm: change anon_v...
924
  	int err;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
  
  	/*
  	 * 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 &&
cc71aba34   vishnu.ps   mm/mmap.c: whites...
945
  			mpol_equal(vma_policy(prev), policy) &&
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
946
  			can_vma_merge_after(prev, vm_flags,
19a809afe   Andrea Arcangeli   userfaultfd: teac...
947
948
  					    anon_vma, file, pgoff,
  					    vm_userfaultfd_ctx)) {
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
949
950
951
952
953
954
  		/*
  		 * 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,
19a809afe   Andrea Arcangeli   userfaultfd: teac...
955
956
957
  						     anon_vma, file,
  						     pgoff+pglen,
  						     vm_userfaultfd_ctx) &&
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
958
  				is_mergeable_anon_vma(prev->anon_vma,
965f55dea   Shaohua Li   mmap: avoid mergi...
959
  						      next->anon_vma, NULL)) {
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
960
  							/* cases 1, 6 */
5beb49305   Rik van Riel   mm: change anon_v...
961
  			err = vma_adjust(prev, prev->vm_start,
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
962
963
  				next->vm_end, prev->vm_pgoff, NULL);
  		} else					/* cases 2, 5, 7 */
5beb49305   Rik van Riel   mm: change anon_v...
964
  			err = vma_adjust(prev, prev->vm_start,
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
965
  				end, prev->vm_pgoff, NULL);
5beb49305   Rik van Riel   mm: change anon_v...
966
967
  		if (err)
  			return NULL;
6d50e60cd   David Rientjes   mm, thp: fix coll...
968
  		khugepaged_enter_vma_merge(prev, vm_flags);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
969
970
971
972
973
974
975
  		return prev;
  	}
  
  	/*
  	 * Can this new request be merged in front of next?
  	 */
  	if (next && end == next->vm_start &&
cc71aba34   vishnu.ps   mm/mmap.c: whites...
976
  			mpol_equal(policy, vma_policy(next)) &&
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
977
  			can_vma_merge_before(next, vm_flags,
19a809afe   Andrea Arcangeli   userfaultfd: teac...
978
979
  					     anon_vma, file, pgoff+pglen,
  					     vm_userfaultfd_ctx)) {
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
980
  		if (prev && addr < prev->vm_end)	/* case 4 */
5beb49305   Rik van Riel   mm: change anon_v...
981
  			err = vma_adjust(prev, prev->vm_start,
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
982
983
  				addr, prev->vm_pgoff, NULL);
  		else					/* cases 3, 8 */
5beb49305   Rik van Riel   mm: change anon_v...
984
  			err = vma_adjust(area, addr, next->vm_end,
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
985
  				next->vm_pgoff - pglen, NULL);
5beb49305   Rik van Riel   mm: change anon_v...
986
987
  		if (err)
  			return NULL;
6d50e60cd   David Rientjes   mm, thp: fix coll...
988
  		khugepaged_enter_vma_merge(area, vm_flags);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
989
990
991
992
993
994
995
  		return area;
  	}
  
  	return NULL;
  }
  
  /*
d0e9fe175   Linus Torvalds   Simplify and comm...
996
997
998
999
1000
1001
1002
1003
1004
1005
1006
1007
1008
1009
1010
1011
1012
   * 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 &&
34228d473   Cyrill Gorcunov   mm: ignore VM_SOF...
1013
  		!((a->vm_flags ^ b->vm_flags) & ~(VM_READ|VM_WRITE|VM_EXEC|VM_SOFTDIRTY)) &&
d0e9fe175   Linus Torvalds   Simplify and comm...
1014
1015
1016
1017
1018
1019
1020
1021
1022
1023
1024
1025
1026
1027
  		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
4db0c3c29   Jason Low   mm: remove rest o...
1028
   * we do that READ_ONCE() to make sure that we never re-load the pointer.
d0e9fe175   Linus Torvalds   Simplify and comm...
1029
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
1040
1041
   *
   * 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)) {
4db0c3c29   Jason Low   mm: remove rest o...
1042
  		struct anon_vma *anon_vma = READ_ONCE(old->anon_vma);
d0e9fe175   Linus Torvalds   Simplify and comm...
1043
1044
1045
1046
1047
1048
1049
1050
  
  		if (anon_vma && list_is_singular(&old->anon_vma_chain))
  			return anon_vma;
  	}
  	return NULL;
  }
  
  /*
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1051
1052
1053
1054
1055
1056
1057
1058
1059
   * 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)
  {
d0e9fe175   Linus Torvalds   Simplify and comm...
1060
  	struct anon_vma *anon_vma;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1061
  	struct vm_area_struct *near;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1062
1063
1064
1065
  
  	near = vma->vm_next;
  	if (!near)
  		goto try_prev;
d0e9fe175   Linus Torvalds   Simplify and comm...
1066
1067
1068
  	anon_vma = reusable_anon_vma(near, vma, near);
  	if (anon_vma)
  		return anon_vma;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1069
  try_prev:
9be34c9d5   Linus Torvalds   mm: get rid of th...
1070
  	near = vma->vm_prev;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1071
1072
  	if (!near)
  		goto none;
d0e9fe175   Linus Torvalds   Simplify and comm...
1073
1074
1075
  	anon_vma = reusable_anon_vma(near, near, vma);
  	if (anon_vma)
  		return anon_vma;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1076
1077
1078
1079
1080
1081
1082
1083
1084
1085
1086
  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;
  }
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1087
  /*
404015308   Al Viro   security: trim se...
1088
1089
1090
1091
1092
1093
1094
1095
1096
1097
1098
   * 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;
  }
363ee17f0   Davidlohr Bueso   mm/mmap.c: add ml...
1099
1100
1101
1102
1103
1104
1105
1106
1107
1108
1109
1110
1111
1112
1113
1114
1115
  static inline int mlock_future_check(struct mm_struct *mm,
  				     unsigned long flags,
  				     unsigned long len)
  {
  	unsigned long locked, lock_limit;
  
  	/*  mlock MCL_FUTURE? */
  	if (flags & VM_LOCKED) {
  		locked = len >> PAGE_SHIFT;
  		locked += mm->locked_vm;
  		lock_limit = rlimit(RLIMIT_MEMLOCK);
  		lock_limit >>= PAGE_SHIFT;
  		if (locked > lock_limit && !capable(CAP_IPC_LOCK))
  			return -EAGAIN;
  	}
  	return 0;
  }
404015308   Al Viro   security: trim se...
1116
  /*
27f5de796   Jianjun Kong   mm: Fix problem o...
1117
   * The caller must hold down_write(&current->mm->mmap_sem).
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1118
   */
1fcfd8db7   Oleg Nesterov   mm, mpx: add "vm_...
1119
  unsigned long do_mmap(struct file *file, unsigned long addr,
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1120
  			unsigned long len, unsigned long prot,
1fcfd8db7   Oleg Nesterov   mm, mpx: add "vm_...
1121
1122
  			unsigned long flags, vm_flags_t vm_flags,
  			unsigned long pgoff, unsigned long *populate)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1123
  {
cc71aba34   vishnu.ps   mm/mmap.c: whites...
1124
  	struct mm_struct *mm = current->mm;
62b5f7d01   Dave Hansen   mm/core, x86/mm/p...
1125
  	int pkey = 0;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1126

41badc15c   Michel Lespinasse   mm: make do_mmap_...
1127
  	*populate = 0;
bebeb3d68   Michel Lespinasse   mm: introduce mm_...
1128

e37609bb3   Piotr Kwapulinski   mm/mmap.c: optimi...
1129
1130
  	if (!len)
  		return -EINVAL;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1131
1132
1133
1134
1135
1136
1137
  	/*
  	 * 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))
90f8572b0   Eric W. Biederman   vfs: Commit to ne...
1138
  		if (!(file && path_noexec(&file->f_path)))
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1139
  			prot |= PROT_EXEC;
7cd94146c   Eric Paris   Security: round m...
1140
1141
  	if (!(flags & MAP_FIXED))
  		addr = round_hint_to_min(addr);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1142
1143
  	/* Careful about overflows.. */
  	len = PAGE_ALIGN(len);
9206de95b   Al Viro   Take arch_mmap_ch...
1144
  	if (!len)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1145
1146
1147
1148
  		return -ENOMEM;
  
  	/* offset overflow? */
  	if ((pgoff + (len >> PAGE_SHIFT)) < pgoff)
cc71aba34   vishnu.ps   mm/mmap.c: whites...
1149
  		return -EOVERFLOW;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1150
1151
1152
1153
1154
1155
1156
1157
1158
  
  	/* 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);
de1741a13   Alexander Kuleshov   mm/mmap: use offs...
1159
  	if (offset_in_page(addr))
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1160
  		return addr;
62b5f7d01   Dave Hansen   mm/core, x86/mm/p...
1161
1162
1163
1164
1165
  	if (prot == PROT_EXEC) {
  		pkey = execute_only_pkey(mm);
  		if (pkey < 0)
  			pkey = 0;
  	}
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1166
1167
1168
1169
  	/* 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.
  	 */
62b5f7d01   Dave Hansen   mm/core, x86/mm/p...
1170
  	vm_flags |= calc_vm_prot_bits(prot, pkey) | calc_vm_flag_bits(flags) |
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1171
  			mm->def_flags | VM_MAYREAD | VM_MAYWRITE | VM_MAYEXEC;
cdf7b3418   Huang Shijie   mmap: remove unne...
1172
  	if (flags & MAP_LOCKED)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1173
1174
  		if (!can_do_mlock())
  			return -EPERM;
ba470de43   Rik van Riel   mmap: handle mloc...
1175

363ee17f0   Davidlohr Bueso   mm/mmap.c: add ml...
1176
1177
  	if (mlock_future_check(mm, vm_flags, len))
  		return -EAGAIN;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1178

1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1179
  	if (file) {
077bf22b5   Oleg Nesterov   mm: do_mmap_pgoff...
1180
  		struct inode *inode = file_inode(file);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1181
1182
1183
1184
1185
1186
1187
1188
1189
1190
1191
1192
1193
1194
1195
  		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.
  			 */
d7a06983a   Jeff Layton   locks: fix locks_...
1196
  			if (locks_verify_locked(file))
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1197
1198
1199
1200
1201
1202
1203
1204
1205
1206
  				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;
90f8572b0   Eric W. Biederman   vfs: Commit to ne...
1207
  			if (path_noexec(&file->f_path)) {
80c5606c3   Linus Torvalds   Fix VM_MAYEXEC ca...
1208
1209
1210
1211
  				if (vm_flags & VM_EXEC)
  					return -EPERM;
  				vm_flags &= ~VM_MAYEXEC;
  			}
80c5606c3   Linus Torvalds   Fix VM_MAYEXEC ca...
1212

72c2d5319   Al Viro   file->f_op is nev...
1213
  			if (!file->f_op->mmap)
80c5606c3   Linus Torvalds   Fix VM_MAYEXEC ca...
1214
  				return -ENODEV;
b2c56e4f7   Oleg Nesterov   mm: shift VM_GROW...
1215
1216
  			if (vm_flags & (VM_GROWSDOWN|VM_GROWSUP))
  				return -EINVAL;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1217
1218
1219
1220
1221
1222
1223
1224
  			break;
  
  		default:
  			return -EINVAL;
  		}
  	} else {
  		switch (flags & MAP_TYPE) {
  		case MAP_SHARED:
b2c56e4f7   Oleg Nesterov   mm: shift VM_GROW...
1225
1226
  			if (vm_flags & (VM_GROWSDOWN|VM_GROWSUP))
  				return -EINVAL;
ce3639426   Tejun Heo   mmap: fix petty b...
1227
1228
1229
1230
  			/*
  			 * Ignore pgoff.
  			 */
  			pgoff = 0;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1231
1232
1233
1234
1235
1236
1237
1238
1239
1240
1241
1242
  			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;
  		}
  	}
c22c0d634   Michel Lespinasse   mm: remove flags ...
1243
1244
1245
1246
1247
1248
1249
1250
1251
1252
1253
1254
1255
1256
1257
  	/*
  	 * 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);
09a9f1d27   Michel Lespinasse   Revert "mm: intro...
1258
1259
1260
  	if (!IS_ERR_VALUE(addr) &&
  	    ((vm_flags & VM_LOCKED) ||
  	     (flags & (MAP_POPULATE | MAP_NONBLOCK)) == MAP_POPULATE))
41badc15c   Michel Lespinasse   mm: make do_mmap_...
1261
  		*populate = len;
bebeb3d68   Michel Lespinasse   mm: introduce mm_...
1262
  	return addr;
0165ab443   Miklos Szeredi   split mmap
1263
  }
6be5ceb02   Linus Torvalds   VM: add "vm_mmap(...
1264

66f0dc481   Hugh Dickins   mm: move sys_mmap...
1265
1266
1267
1268
1269
  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;
1e3ee14b9   Chen Gang   mm/mmap.c: do not...
1270
  	unsigned long retval;
66f0dc481   Hugh Dickins   mm: move sys_mmap...
1271
1272
  
  	if (!(flags & MAP_ANONYMOUS)) {
120a795da   Al Viro   audit mmap
1273
  		audit_mmap_fd(fd, flags);
66f0dc481   Hugh Dickins   mm: move sys_mmap...
1274
1275
  		file = fget(fd);
  		if (!file)
1e3ee14b9   Chen Gang   mm/mmap.c: do not...
1276
  			return -EBADF;
af73e4d95   Naoya Horiguchi   hugetlbfs: fix mm...
1277
1278
  		if (is_file_hugepages(file))
  			len = ALIGN(len, huge_page_size(hstate_file(file)));
493af5780   Jörn Engel   mmap: allow MAP_H...
1279
1280
1281
  		retval = -EINVAL;
  		if (unlikely(flags & MAP_HUGETLB && !is_file_hugepages(file)))
  			goto out_fput;
66f0dc481   Hugh Dickins   mm: move sys_mmap...
1282
1283
  	} else if (flags & MAP_HUGETLB) {
  		struct user_struct *user = NULL;
c103a4dc4   Andrew Morton   ipc/shmc.c: elimi...
1284
  		struct hstate *hs;
af73e4d95   Naoya Horiguchi   hugetlbfs: fix mm...
1285

c103a4dc4   Andrew Morton   ipc/shmc.c: elimi...
1286
  		hs = hstate_sizelog((flags >> MAP_HUGE_SHIFT) & SHM_HUGE_MASK);
091d0d55b   Li Zefan   shm: fix null poi...
1287
1288
1289
1290
  		if (!hs)
  			return -EINVAL;
  
  		len = ALIGN(len, huge_page_size(hs));
66f0dc481   Hugh Dickins   mm: move sys_mmap...
1291
1292
1293
1294
1295
1296
  		/*
  		 * 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
  		 */
af73e4d95   Naoya Horiguchi   hugetlbfs: fix mm...
1297
  		file = hugetlb_file_setup(HUGETLB_ANON_FILE, len,
42d7395fe   Andi Kleen   mm: support more ...
1298
1299
1300
  				VM_NORESERVE,
  				&user, HUGETLB_ANONHUGE_INODE,
  				(flags >> MAP_HUGE_SHIFT) & MAP_HUGE_MASK);
66f0dc481   Hugh Dickins   mm: move sys_mmap...
1301
1302
1303
1304
1305
  		if (IS_ERR(file))
  			return PTR_ERR(file);
  	}
  
  	flags &= ~(MAP_EXECUTABLE | MAP_DENYWRITE);
9fbeb5ab5   Michal Hocko   mm: make vm_mmap ...
1306
  	retval = vm_mmap_pgoff(file, addr, len, prot, flags, pgoff);
493af5780   Jörn Engel   mmap: allow MAP_H...
1307
  out_fput:
66f0dc481   Hugh Dickins   mm: move sys_mmap...
1308
1309
  	if (file)
  		fput(file);
66f0dc481   Hugh Dickins   mm: move sys_mmap...
1310
1311
  	return retval;
  }
a4679373c   Christoph Hellwig   Add generic sys_o...
1312
1313
1314
1315
1316
1317
1318
1319
1320
1321
1322
1323
1324
1325
1326
1327
  #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;
de1741a13   Alexander Kuleshov   mm/mmap: use offs...
1328
  	if (offset_in_page(a.offset))
a4679373c   Christoph Hellwig   Add generic sys_o...
1329
1330
1331
1332
1333
1334
  		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 */
4e950f6f0   Alexey Dobriyan   Remove fs.h from ...
1335
1336
1337
1338
1339
1340
1341
1342
  /*
   * 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)
  {
ca16d140a   KOSAKI Motohiro   mm: don't access ...
1343
  	vm_flags_t vm_flags = vma->vm_flags;
8a04446ab   Kirill A. Shutemov   mm, dax: VMA with...
1344
  	const struct vm_operations_struct *vm_ops = vma->vm_ops;
4e950f6f0   Alexey Dobriyan   Remove fs.h from ...
1345
1346
1347
1348
1349
1350
  
  	/* 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? */
8a04446ab   Kirill A. Shutemov   mm, dax: VMA with...
1351
  	if (vm_ops && (vm_ops->page_mkwrite || vm_ops->pfn_mkwrite))
4e950f6f0   Alexey Dobriyan   Remove fs.h from ...
1352
  		return 1;
64e455079   Peter Feiner   mm: softdirty: en...
1353
1354
  	/* The open routine did something to the protections that pgprot_modify
  	 * won't preserve? */
4e950f6f0   Alexey Dobriyan   Remove fs.h from ...
1355
  	if (pgprot_val(vma->vm_page_prot) !=
64e455079   Peter Feiner   mm: softdirty: en...
1356
  	    pgprot_val(vm_pgprot_modify(vma->vm_page_prot, vm_flags)))
4e950f6f0   Alexey Dobriyan   Remove fs.h from ...
1357
  		return 0;
64e455079   Peter Feiner   mm: softdirty: en...
1358
1359
1360
  	/* Do we need to track softdirty? */
  	if (IS_ENABLED(CONFIG_MEM_SOFT_DIRTY) && !(vm_flags & VM_SOFTDIRTY))
  		return 1;
4e950f6f0   Alexey Dobriyan   Remove fs.h from ...
1361
  	/* Specialty mapping? */
4b6e1e370   Konstantin Khlebnikov   mm: kill vma flag...
1362
  	if (vm_flags & VM_PFNMAP)
4e950f6f0   Alexey Dobriyan   Remove fs.h from ...
1363
1364
1365
1366
1367
1368
  		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);
  }
fc8744adc   Linus Torvalds   Stop playing sill...
1369
1370
  /*
   * We account for memory if it's a private writeable mapping,
5a6fe1259   Mel Gorman   Do not account fo...
1371
   * not hugepages and VM_NORESERVE wasn't set.
fc8744adc   Linus Torvalds   Stop playing sill...
1372
   */
ca16d140a   KOSAKI Motohiro   mm: don't access ...
1373
  static inline int accountable_mapping(struct file *file, vm_flags_t vm_flags)
fc8744adc   Linus Torvalds   Stop playing sill...
1374
  {
5a6fe1259   Mel Gorman   Do not account fo...
1375
1376
1377
1378
1379
1380
  	/*
  	 * 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;
fc8744adc   Linus Torvalds   Stop playing sill...
1381
1382
  	return (vm_flags & (VM_NORESERVE | VM_SHARED | VM_WRITE)) == VM_WRITE;
  }
0165ab443   Miklos Szeredi   split mmap
1383
  unsigned long mmap_region(struct file *file, unsigned long addr,
c22c0d634   Michel Lespinasse   mm: remove flags ...
1384
  		unsigned long len, vm_flags_t vm_flags, unsigned long pgoff)
0165ab443   Miklos Szeredi   split mmap
1385
1386
1387
  {
  	struct mm_struct *mm = current->mm;
  	struct vm_area_struct *vma, *prev;
0165ab443   Miklos Szeredi   split mmap
1388
1389
1390
  	int error;
  	struct rb_node **rb_link, *rb_parent;
  	unsigned long charged = 0;
0165ab443   Miklos Szeredi   split mmap
1391

e8420a8ec   Cyril Hrubis   mm/mmap: check fo...
1392
  	/* Check against address space limit. */
846383359   Konstantin Khlebnikov   mm: rework virtua...
1393
  	if (!may_expand_vm(mm, vm_flags, len >> PAGE_SHIFT)) {
e8420a8ec   Cyril Hrubis   mm/mmap: check fo...
1394
1395
1396
1397
1398
1399
  		unsigned long nr_pages;
  
  		/*
  		 * MAP_FIXED may remove pages of mappings that intersects with
  		 * requested mapping. Account for the pages it would unmap.
  		 */
e8420a8ec   Cyril Hrubis   mm/mmap: check fo...
1400
  		nr_pages = count_vma_pages_range(mm, addr, addr + len);
846383359   Konstantin Khlebnikov   mm: rework virtua...
1401
1402
  		if (!may_expand_vm(mm, vm_flags,
  					(len >> PAGE_SHIFT) - nr_pages))
e8420a8ec   Cyril Hrubis   mm/mmap: check fo...
1403
1404
  			return -ENOMEM;
  	}
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1405
  	/* Clear old maps */
9fcd14571   Rasmus Villemoes   mm/mmap.c: use wh...
1406
1407
  	while (find_vma_links(mm, addr, addr + len, &prev, &rb_link,
  			      &rb_parent)) {
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1408
1409
  		if (do_munmap(mm, addr, len))
  			return -ENOMEM;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1410
  	}
fc8744adc   Linus Torvalds   Stop playing sill...
1411
  	/*
fc8744adc   Linus Torvalds   Stop playing sill...
1412
1413
  	 * Private writable mapping: check memory availability
  	 */
5a6fe1259   Mel Gorman   Do not account fo...
1414
  	if (accountable_mapping(file, vm_flags)) {
fc8744adc   Linus Torvalds   Stop playing sill...
1415
  		charged = len >> PAGE_SHIFT;
191c54244   Al Viro   mm: collapse secu...
1416
  		if (security_vm_enough_memory_mm(mm, charged))
fc8744adc   Linus Torvalds   Stop playing sill...
1417
1418
  			return -ENOMEM;
  		vm_flags |= VM_ACCOUNT;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1419
1420
1421
  	}
  
  	/*
de33c8db5   Linus Torvalds   Fix OOPS in mmap_...
1422
  	 * Can we just expand an old mapping?
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1423
  	 */
19a809afe   Andrea Arcangeli   userfaultfd: teac...
1424
1425
  	vma = vma_merge(mm, prev, addr, addr + len, vm_flags,
  			NULL, file, pgoff, NULL, NULL_VM_UFFD_CTX);
de33c8db5   Linus Torvalds   Fix OOPS in mmap_...
1426
1427
  	if (vma)
  		goto out;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1428
1429
1430
1431
1432
1433
  
  	/*
  	 * 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.
  	 */
c5e3b83e9   Pekka Enberg   [PATCH] mm: use k...
1434
  	vma = kmem_cache_zalloc(vm_area_cachep, GFP_KERNEL);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1435
1436
1437
1438
  	if (!vma) {
  		error = -ENOMEM;
  		goto unacct_error;
  	}
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1439
1440
1441
1442
1443
  
  	vma->vm_mm = mm;
  	vma->vm_start = addr;
  	vma->vm_end = addr + len;
  	vma->vm_flags = vm_flags;
3ed75eb8f   Coly Li   setup vma->vm_pag...
1444
  	vma->vm_page_prot = vm_get_page_prot(vm_flags);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1445
  	vma->vm_pgoff = pgoff;
5beb49305   Rik van Riel   mm: change anon_v...
1446
  	INIT_LIST_HEAD(&vma->anon_vma_chain);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1447
1448
  
  	if (file) {
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1449
1450
1451
1452
  		if (vm_flags & VM_DENYWRITE) {
  			error = deny_write_access(file);
  			if (error)
  				goto free_vma;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1453
  		}
4bb5f5d93   David Herrmann   mm: allow drivers...
1454
1455
1456
1457
1458
1459
1460
1461
1462
1463
1464
  		if (vm_flags & VM_SHARED) {
  			error = mapping_map_writable(file->f_mapping);
  			if (error)
  				goto allow_write_and_free_vma;
  		}
  
  		/* ->mmap() can change vma->vm_file, but must guarantee that
  		 * vma_link() below can deny write-access if VM_DENYWRITE is set
  		 * and map writably if VM_SHARED is set. This usually means the
  		 * new file must not have been exposed to user-space, yet.
  		 */
cb0942b81   Al Viro   make get_file() r...
1465
  		vma->vm_file = get_file(file);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1466
1467
1468
  		error = file->f_op->mmap(file, vma);
  		if (error)
  			goto unmap_and_free_vma;
f8dbf0a7a   Huang Shijie   mmap: save some c...
1469
1470
1471
1472
1473
  
  		/* Can addr have changed??
  		 *
  		 * Answer: Yes, several device drivers can do it in their
  		 *         f_op->mmap method. -DaveM
2897b4d29   Joonsoo Kim   mm: WARN_ON_ONCE ...
1474
1475
  		 * Bug: If addr is changed, prev, rb_link, rb_parent should
  		 *      be updated for vma_link()
f8dbf0a7a   Huang Shijie   mmap: save some c...
1476
  		 */
2897b4d29   Joonsoo Kim   mm: WARN_ON_ONCE ...
1477
  		WARN_ON_ONCE(addr != vma->vm_start);
f8dbf0a7a   Huang Shijie   mmap: save some c...
1478
  		addr = vma->vm_start;
f8dbf0a7a   Huang Shijie   mmap: save some c...
1479
  		vm_flags = vma->vm_flags;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1480
1481
1482
1483
1484
  	} else if (vm_flags & VM_SHARED) {
  		error = shmem_zero_setup(vma);
  		if (error)
  			goto free_vma;
  	}
de33c8db5   Linus Torvalds   Fix OOPS in mmap_...
1485
  	vma_link(mm, vma, prev, rb_link, rb_parent);
4d3d5b41a   Oleg Nesterov   mmap_region: clea...
1486
  	/* Once vma denies write, undo our temporary denial count */
4bb5f5d93   David Herrmann   mm: allow drivers...
1487
1488
1489
1490
1491
1492
  	if (file) {
  		if (vm_flags & VM_SHARED)
  			mapping_unmap_writable(file->f_mapping);
  		if (vm_flags & VM_DENYWRITE)
  			allow_write_access(file);
  	}
e86867720   Oleg Nesterov   mm: mmap_region: ...
1493
  	file = vma->vm_file;
4d3d5b41a   Oleg Nesterov   mmap_region: clea...
1494
  out:
cdd6c482c   Ingo Molnar   perf: Do the big ...
1495
  	perf_event_mmap(vma);
0a4a93919   Peter Zijlstra   perf_counter: exe...
1496

846383359   Konstantin Khlebnikov   mm: rework virtua...
1497
  	vm_stat_account(mm, vm_flags, len >> PAGE_SHIFT);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1498
  	if (vm_flags & VM_LOCKED) {
bebeb3d68   Michel Lespinasse   mm: introduce mm_...
1499
1500
  		if (!((vm_flags & VM_SPECIAL) || is_vm_hugetlb_page(vma) ||
  					vma == get_gate_vma(current->mm)))
06f9d8c2b   KOSAKI Motohiro   mm: mlock_vma_pag...
1501
  			mm->locked_vm += (len >> PAGE_SHIFT);
bebeb3d68   Michel Lespinasse   mm: introduce mm_...
1502
  		else
de60f5f10   Eric B Munson   mm: introduce VM_...
1503
  			vma->vm_flags &= VM_LOCKED_CLEAR_MASK;
bebeb3d68   Michel Lespinasse   mm: introduce mm_...
1504
  	}
2b1444983   Srikar Dronamraju   uprobes, mm, x86:...
1505

c7a3a88c9   Oleg Nesterov   uprobes: Fix mmap...
1506
1507
  	if (file)
  		uprobe_mmap(vma);
2b1444983   Srikar Dronamraju   uprobes, mm, x86:...
1508

d9104d1ca   Cyrill Gorcunov   mm: track vma cha...
1509
1510
1511
1512
1513
1514
1515
1516
  	/*
  	 * New (or expanded) vma always get soft dirty status.
  	 * Otherwise user-space soft-dirty page tracker won't
  	 * be able to distinguish situation when vma area unmapped,
  	 * then new mapped in-place (which must be aimed as
  	 * a completely new data area).
  	 */
  	vma->vm_flags |= VM_SOFTDIRTY;
64e455079   Peter Feiner   mm: softdirty: en...
1517
  	vma_set_page_prot(vma);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1518
1519
1520
  	return addr;
  
  unmap_and_free_vma:
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1521
1522
1523
1524
  	vma->vm_file = NULL;
  	fput(file);
  
  	/* Undo any partial mapping done by a device driver. */
e0da382c9   Hugh Dickins   [PATCH] freepgt: ...
1525
1526
  	unmap_region(mm, vma, prev, vma->vm_start, vma->vm_end);
  	charged = 0;
4bb5f5d93   David Herrmann   mm: allow drivers...
1527
1528
1529
1530
1531
  	if (vm_flags & VM_SHARED)
  		mapping_unmap_writable(file->f_mapping);
  allow_write_and_free_vma:
  	if (vm_flags & VM_DENYWRITE)
  		allow_write_access(file);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1532
1533
1534
1535
1536
1537
1538
  free_vma:
  	kmem_cache_free(vm_area_cachep, vma);
  unacct_error:
  	if (charged)
  		vm_unacct_memory(charged);
  	return error;
  }
db4fbfb95   Michel Lespinasse   mm: vm_unmapped_a...
1539
1540
1541
1542
1543
1544
1545
1546
1547
1548
1549
1550
1551
1552
1553
1554
1555
1556
1557
1558
1559
1560
1561
1562
1563
1564
1565
1566
1567
1568
1569
1570
1571
1572
1573
1574
1575
1576
1577
1578
1579
1580
1581
1582
1583
1584
1585
1586
1587
1588
1589
1590
1591
1592
1593
1594
1595
1596
1597
1598
1599
1600
1601
1602
1603
1604
1605
1606
1607
1608
1609
1610
1611
1612
1613
1614
1615
1616
1617
1618
1619
1620
1621
1622
1623
1624
1625
1626
1627
1628
1629
1630
1631
1632
1633
1634
1635
1636
1637
1638
1639
1640
1641
1642
1643
1644
1645
1646
1647
1648
1649
1650
1651
1652
1653
1654
1655
1656
1657
1658
1659
1660
1661
1662
1663
1664
1665
1666
1667
1668
1669
1670
1671
1672
1673
1674
1675
1676
1677
1678
1679
1680
1681
1682
1683
1684
1685
1686
1687
1688
1689
1690
1691
1692
1693
1694
1695
1696
1697
1698
1699
1700
1701
1702
1703
1704
1705
1706
1707
1708
1709
1710
1711
1712
1713
1714
1715
1716
1717
1718
1719
1720
1721
1722
1723
1724
1725
1726
1727
1728
1729
1730
1731
1732
1733
1734
1735
1736
1737
  unsigned long unmapped_area(struct vm_unmapped_area_info *info)
  {
  	/*
  	 * We implement the search by looking for an rbtree node that
  	 * immediately follows a suitable gap. That is,
  	 * - gap_start = vma->vm_prev->vm_end <= info->high_limit - length;
  	 * - gap_end   = vma->vm_start        >= info->low_limit  + length;
  	 * - gap_end - gap_start >= length
  	 */
  
  	struct mm_struct *mm = current->mm;
  	struct vm_area_struct *vma;
  	unsigned long length, low_limit, high_limit, gap_start, gap_end;
  
  	/* Adjust search length to account for worst case alignment overhead */
  	length = info->length + info->align_mask;
  	if (length < info->length)
  		return -ENOMEM;
  
  	/* Adjust search limits by the desired length */
  	if (info->high_limit < length)
  		return -ENOMEM;
  	high_limit = info->high_limit - length;
  
  	if (info->low_limit > high_limit)
  		return -ENOMEM;
  	low_limit = info->low_limit + length;
  
  	/* Check if rbtree root looks promising */
  	if (RB_EMPTY_ROOT(&mm->mm_rb))
  		goto check_highest;
  	vma = rb_entry(mm->mm_rb.rb_node, struct vm_area_struct, vm_rb);
  	if (vma->rb_subtree_gap < length)
  		goto check_highest;
  
  	while (true) {
  		/* Visit left subtree if it looks promising */
  		gap_end = vma->vm_start;
  		if (gap_end >= low_limit && vma->vm_rb.rb_left) {
  			struct vm_area_struct *left =
  				rb_entry(vma->vm_rb.rb_left,
  					 struct vm_area_struct, vm_rb);
  			if (left->rb_subtree_gap >= length) {
  				vma = left;
  				continue;
  			}
  		}
  
  		gap_start = vma->vm_prev ? vma->vm_prev->vm_end : 0;
  check_current:
  		/* Check if current node has a suitable gap */
  		if (gap_start > high_limit)
  			return -ENOMEM;
  		if (gap_end >= low_limit && gap_end - gap_start >= length)
  			goto found;
  
  		/* Visit right subtree if it looks promising */
  		if (vma->vm_rb.rb_right) {
  			struct vm_area_struct *right =
  				rb_entry(vma->vm_rb.rb_right,
  					 struct vm_area_struct, vm_rb);
  			if (right->rb_subtree_gap >= length) {
  				vma = right;
  				continue;
  			}
  		}
  
  		/* Go back up the rbtree to find next candidate node */
  		while (true) {
  			struct rb_node *prev = &vma->vm_rb;
  			if (!rb_parent(prev))
  				goto check_highest;
  			vma = rb_entry(rb_parent(prev),
  				       struct vm_area_struct, vm_rb);
  			if (prev == vma->vm_rb.rb_left) {
  				gap_start = vma->vm_prev->vm_end;
  				gap_end = vma->vm_start;
  				goto check_current;
  			}
  		}
  	}
  
  check_highest:
  	/* Check highest gap, which does not precede any rbtree node */
  	gap_start = mm->highest_vm_end;
  	gap_end = ULONG_MAX;  /* Only for VM_BUG_ON below */
  	if (gap_start > high_limit)
  		return -ENOMEM;
  
  found:
  	/* We found a suitable gap. Clip it with the original low_limit. */
  	if (gap_start < info->low_limit)
  		gap_start = info->low_limit;
  
  	/* Adjust gap address to the desired alignment */
  	gap_start += (info->align_offset - gap_start) & info->align_mask;
  
  	VM_BUG_ON(gap_start + info->length > info->high_limit);
  	VM_BUG_ON(gap_start + info->length > gap_end);
  	return gap_start;
  }
  
  unsigned long unmapped_area_topdown(struct vm_unmapped_area_info *info)
  {
  	struct mm_struct *mm = current->mm;
  	struct vm_area_struct *vma;
  	unsigned long length, low_limit, high_limit, gap_start, gap_end;
  
  	/* Adjust search length to account for worst case alignment overhead */
  	length = info->length + info->align_mask;
  	if (length < info->length)
  		return -ENOMEM;
  
  	/*
  	 * Adjust search limits by the desired length.
  	 * See implementation comment at top of unmapped_area().
  	 */
  	gap_end = info->high_limit;
  	if (gap_end < length)
  		return -ENOMEM;
  	high_limit = gap_end - length;
  
  	if (info->low_limit > high_limit)
  		return -ENOMEM;
  	low_limit = info->low_limit + length;
  
  	/* Check highest gap, which does not precede any rbtree node */
  	gap_start = mm->highest_vm_end;
  	if (gap_start <= high_limit)
  		goto found_highest;
  
  	/* Check if rbtree root looks promising */
  	if (RB_EMPTY_ROOT(&mm->mm_rb))
  		return -ENOMEM;
  	vma = rb_entry(mm->mm_rb.rb_node, struct vm_area_struct, vm_rb);
  	if (vma->rb_subtree_gap < length)
  		return -ENOMEM;
  
  	while (true) {
  		/* Visit right subtree if it looks promising */
  		gap_start = vma->vm_prev ? vma->vm_prev->vm_end : 0;
  		if (gap_start <= high_limit && vma->vm_rb.rb_right) {
  			struct vm_area_struct *right =
  				rb_entry(vma->vm_rb.rb_right,
  					 struct vm_area_struct, vm_rb);
  			if (right->rb_subtree_gap >= length) {
  				vma = right;
  				continue;
  			}
  		}
  
  check_current:
  		/* Check if current node has a suitable gap */
  		gap_end = vma->vm_start;
  		if (gap_end < low_limit)
  			return -ENOMEM;
  		if (gap_start <= high_limit && gap_end - gap_start >= length)
  			goto found;
  
  		/* Visit left subtree if it looks promising */
  		if (vma->vm_rb.rb_left) {
  			struct vm_area_struct *left =
  				rb_entry(vma->vm_rb.rb_left,
  					 struct vm_area_struct, vm_rb);
  			if (left->rb_subtree_gap >= length) {
  				vma = left;
  				continue;
  			}
  		}
  
  		/* Go back up the rbtree to find next candidate node */
  		while (true) {
  			struct rb_node *prev = &vma->vm_rb;
  			if (!rb_parent(prev))
  				return -ENOMEM;
  			vma = rb_entry(rb_parent(prev),
  				       struct vm_area_struct, vm_rb);
  			if (prev == vma->vm_rb.rb_right) {
  				gap_start = vma->vm_prev ?
  					vma->vm_prev->vm_end : 0;
  				goto check_current;
  			}
  		}
  	}
  
  found:
  	/* We found a suitable gap. Clip it with the original high_limit. */
  	if (gap_end > info->high_limit)
  		gap_end = info->high_limit;
  
  found_highest:
  	/* Compute highest gap address at the desired alignment */
  	gap_end -= info->length;
  	gap_end -= (gap_end - info->align_offset) & info->align_mask;
  
  	VM_BUG_ON(gap_end < info->low_limit);
  	VM_BUG_ON(gap_end < gap_start);
  	return gap_end;
  }
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1738
1739
1740
1741
1742
1743
1744
1745
1746
1747
1748
1749
1750
1751
1752
1753
1754
1755
  /* 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;
db4fbfb95   Michel Lespinasse   mm: vm_unmapped_a...
1756
  	struct vm_unmapped_area_info info;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1757

2afc745f3   Akira Takeuchi   mm: ensure get_un...
1758
  	if (len > TASK_SIZE - mmap_min_addr)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1759
  		return -ENOMEM;
06abdfb47   Benjamin Herrenschmidt   get_unmapped_area...
1760
1761
  	if (flags & MAP_FIXED)
  		return addr;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1762
1763
1764
  	if (addr) {
  		addr = PAGE_ALIGN(addr);
  		vma = find_vma(mm, addr);
2afc745f3   Akira Takeuchi   mm: ensure get_un...
1765
  		if (TASK_SIZE - len >= addr && addr >= mmap_min_addr &&
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1766
1767
1768
  		    (!vma || addr + len <= vma->vm_start))
  			return addr;
  	}
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1769

db4fbfb95   Michel Lespinasse   mm: vm_unmapped_a...
1770
1771
  	info.flags = 0;
  	info.length = len;
4e99b0213   Heiko Carstens   mmap: arch_get_un...
1772
  	info.low_limit = mm->mmap_base;
db4fbfb95   Michel Lespinasse   mm: vm_unmapped_a...
1773
1774
1775
  	info.high_limit = TASK_SIZE;
  	info.align_mask = 0;
  	return vm_unmapped_area(&info);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1776
  }
cc71aba34   vishnu.ps   mm/mmap.c: whites...
1777
  #endif
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1778

1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1779
1780
1781
1782
1783
1784
1785
1786
1787
1788
1789
1790
  /*
   * 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;
db4fbfb95   Michel Lespinasse   mm: vm_unmapped_a...
1791
1792
  	unsigned long addr = addr0;
  	struct vm_unmapped_area_info info;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1793
1794
  
  	/* requested length too big for entire address space */
2afc745f3   Akira Takeuchi   mm: ensure get_un...
1795
  	if (len > TASK_SIZE - mmap_min_addr)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1796
  		return -ENOMEM;
06abdfb47   Benjamin Herrenschmidt   get_unmapped_area...
1797
1798
  	if (flags & MAP_FIXED)
  		return addr;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1799
1800
1801
1802
  	/* requesting a specific address */
  	if (addr) {
  		addr = PAGE_ALIGN(addr);
  		vma = find_vma(mm, addr);
2afc745f3   Akira Takeuchi   mm: ensure get_un...
1803
  		if (TASK_SIZE - len >= addr && addr >= mmap_min_addr &&
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1804
1805
1806
  				(!vma || addr + len <= vma->vm_start))
  			return addr;
  	}
db4fbfb95   Michel Lespinasse   mm: vm_unmapped_a...
1807
1808
  	info.flags = VM_UNMAPPED_AREA_TOPDOWN;
  	info.length = len;
2afc745f3   Akira Takeuchi   mm: ensure get_un...
1809
  	info.low_limit = max(PAGE_SIZE, mmap_min_addr);
db4fbfb95   Michel Lespinasse   mm: vm_unmapped_a...
1810
1811
1812
  	info.high_limit = mm->mmap_base;
  	info.align_mask = 0;
  	addr = vm_unmapped_area(&info);
b716ad953   Xiao Guangrong   mm: search from f...
1813

1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1814
1815
1816
1817
1818
1819
  	/*
  	 * 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.
  	 */
de1741a13   Alexander Kuleshov   mm/mmap: use offs...
1820
  	if (offset_in_page(addr)) {
db4fbfb95   Michel Lespinasse   mm: vm_unmapped_a...
1821
1822
1823
1824
1825
1826
  		VM_BUG_ON(addr != -ENOMEM);
  		info.flags = 0;
  		info.low_limit = TASK_UNMAPPED_BASE;
  		info.high_limit = TASK_SIZE;
  		addr = vm_unmapped_area(&info);
  	}
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1827
1828
1829
1830
  
  	return addr;
  }
  #endif
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1831
1832
1833
1834
  unsigned long
  get_unmapped_area(struct file *file, unsigned long addr, unsigned long len,
  		unsigned long pgoff, unsigned long flags)
  {
06abdfb47   Benjamin Herrenschmidt   get_unmapped_area...
1835
1836
  	unsigned long (*get_area)(struct file *, unsigned long,
  				  unsigned long, unsigned long, unsigned long);
9206de95b   Al Viro   Take arch_mmap_ch...
1837
1838
1839
1840
1841
1842
1843
  	unsigned long error = arch_mmap_check(addr, len, flags);
  	if (error)
  		return error;
  
  	/* Careful about overflows.. */
  	if (len > TASK_SIZE)
  		return -ENOMEM;
06abdfb47   Benjamin Herrenschmidt   get_unmapped_area...
1844
  	get_area = current->mm->get_unmapped_area;
c01d5b300   Hugh Dickins   shmem: get_unmapp...
1845
1846
1847
1848
1849
1850
1851
1852
1853
1854
1855
1856
  	if (file) {
  		if (file->f_op->get_unmapped_area)
  			get_area = file->f_op->get_unmapped_area;
  	} else if (flags & MAP_SHARED) {
  		/*
  		 * mmap_region() will call shmem_zero_setup() to create a file,
  		 * so use shmem's get_unmapped_area in case it can be huge.
  		 * do_mmap_pgoff() will clear pgoff, so match alignment.
  		 */
  		pgoff = 0;
  		get_area = shmem_get_unmapped_area;
  	}
06abdfb47   Benjamin Herrenschmidt   get_unmapped_area...
1857
1858
1859
  	addr = get_area(file, addr, len, pgoff, flags);
  	if (IS_ERR_VALUE(addr))
  		return addr;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1860

07ab67c8d   Linus Torvalds   Fix get_unmapped_...
1861
1862
  	if (addr > TASK_SIZE - len)
  		return -ENOMEM;
de1741a13   Alexander Kuleshov   mm/mmap: use offs...
1863
  	if (offset_in_page(addr))
07ab67c8d   Linus Torvalds   Fix get_unmapped_...
1864
  		return -EINVAL;
06abdfb47   Benjamin Herrenschmidt   get_unmapped_area...
1865

9ac4ed4bd   Al Viro   move security_mma...
1866
1867
  	error = security_mmap_addr(addr);
  	return error ? error : addr;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1868
1869
1870
1871
1872
  }
  
  EXPORT_SYMBOL(get_unmapped_area);
  
  /* Look up the first VMA which satisfies  addr < vm_end,  NULL if none. */
48aae4255   ZhenwenXu   mm/mmap.c: fix co...
1873
  struct vm_area_struct *find_vma(struct mm_struct *mm, unsigned long addr)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1874
  {
615d6e875   Davidlohr Bueso   mm: per-thread vm...
1875
1876
  	struct rb_node *rb_node;
  	struct vm_area_struct *vma;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1877

841e31e5c   Rajman Mekaco   mm/mmap.c: find_v...
1878
  	/* Check the cache first. */
615d6e875   Davidlohr Bueso   mm: per-thread vm...
1879
1880
1881
  	vma = vmacache_find(mm, addr);
  	if (likely(vma))
  		return vma;
841e31e5c   Rajman Mekaco   mm/mmap.c: find_v...
1882

615d6e875   Davidlohr Bueso   mm: per-thread vm...
1883
  	rb_node = mm->mm_rb.rb_node;
841e31e5c   Rajman Mekaco   mm/mmap.c: find_v...
1884

615d6e875   Davidlohr Bueso   mm: per-thread vm...
1885
1886
1887
1888
1889
1890
1891
1892
1893
1894
1895
1896
  	while (rb_node) {
  		struct vm_area_struct *tmp;
  
  		tmp = rb_entry(rb_node, struct vm_area_struct, vm_rb);
  
  		if (tmp->vm_end > addr) {
  			vma = tmp;
  			if (tmp->vm_start <= addr)
  				break;
  			rb_node = rb_node->rb_left;
  		} else
  			rb_node = rb_node->rb_right;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1897
  	}
615d6e875   Davidlohr Bueso   mm: per-thread vm...
1898
1899
1900
  
  	if (vma)
  		vmacache_update(addr, vma);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1901
1902
1903
1904
  	return vma;
  }
  
  EXPORT_SYMBOL(find_vma);
6bd4837de   KOSAKI Motohiro   mm: simplify find...
1905
1906
  /*
   * Same as find_vma, but also return a pointer to the previous VMA in *pprev.
6bd4837de   KOSAKI Motohiro   mm: simplify find...
1907
   */
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1908
1909
1910
1911
  struct vm_area_struct *
  find_vma_prev(struct mm_struct *mm, unsigned long addr,
  			struct vm_area_struct **pprev)
  {
6bd4837de   KOSAKI Motohiro   mm: simplify find...
1912
  	struct vm_area_struct *vma;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1913

6bd4837de   KOSAKI Motohiro   mm: simplify find...
1914
  	vma = find_vma(mm, addr);
83cd904d2   Mikulas Patocka   mm: fix find_vma_...
1915
1916
1917
1918
1919
1920
1921
1922
1923
1924
  	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;
  		}
  	}
6bd4837de   KOSAKI Motohiro   mm: simplify find...
1925
  	return vma;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1926
1927
1928
1929
1930
1931
1932
  }
  
  /*
   * Verify that the stack growth is acceptable and
   * update accounting. This is shared with both the
   * grow-up and grow-down cases.
   */
48aae4255   ZhenwenXu   mm/mmap.c: fix co...
1933
  static int acct_stack_growth(struct vm_area_struct *vma, unsigned long size, unsigned long grow)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1934
1935
1936
  {
  	struct mm_struct *mm = vma->vm_mm;
  	struct rlimit *rlim = current->signal->rlim;
690eac53d   Linus Torvalds   mm: Don't count t...
1937
  	unsigned long new_start, actual_size;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1938
1939
  
  	/* address space limit tests */
846383359   Konstantin Khlebnikov   mm: rework virtua...
1940
  	if (!may_expand_vm(mm, vma->vm_flags, grow))
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1941
1942
1943
  		return -ENOMEM;
  
  	/* Stack limit test */
690eac53d   Linus Torvalds   mm: Don't count t...
1944
1945
1946
  	actual_size = size;
  	if (size && (vma->vm_flags & (VM_GROWSUP | VM_GROWSDOWN)))
  		actual_size -= PAGE_SIZE;
4db0c3c29   Jason Low   mm: remove rest o...
1947
  	if (actual_size > READ_ONCE(rlim[RLIMIT_STACK].rlim_cur))
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1948
1949
1950
1951
1952
1953
1954
  		return -ENOMEM;
  
  	/* mlock limit tests */
  	if (vma->vm_flags & VM_LOCKED) {
  		unsigned long locked;
  		unsigned long limit;
  		locked = mm->locked_vm + grow;
4db0c3c29   Jason Low   mm: remove rest o...
1955
  		limit = READ_ONCE(rlim[RLIMIT_MEMLOCK].rlim_cur);
59e99e5b9   Jiri Slaby   mm: use rlimit he...
1956
  		limit >>= PAGE_SHIFT;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1957
1958
1959
  		if (locked > limit && !capable(CAP_IPC_LOCK))
  			return -ENOMEM;
  	}
0d59a01bc   Adam Litke   [PATCH] Don't all...
1960
1961
1962
1963
1964
  	/* 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;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1965
1966
1967
1968
  	/*
  	 * Overcommit..  This must be the final test, as it will
  	 * update security statistics.
  	 */
05fa199d4   Hugh Dickins   mm: pass correct ...
1969
  	if (security_vm_enough_memory_mm(mm, grow))
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1970
  		return -ENOMEM;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1971
1972
  	return 0;
  }
46dea3d09   Hugh Dickins   [PATCH] mm: ia64 ...
1973
  #if defined(CONFIG_STACK_GROWSUP) || defined(CONFIG_IA64)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1974
  /*
46dea3d09   Hugh Dickins   [PATCH] mm: ia64 ...
1975
1976
   * 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.
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1977
   */
46dea3d09   Hugh Dickins   [PATCH] mm: ia64 ...
1978
  int expand_upwards(struct vm_area_struct *vma, unsigned long address)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1979
  {
093578147   Oleg Nesterov   mm: add the "stru...
1980
  	struct mm_struct *mm = vma->vm_mm;
12352d3ca   Konstantin Khlebnikov   mm: replace vma_l...
1981
  	int error = 0;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1982
1983
1984
  
  	if (!(vma->vm_flags & VM_GROWSUP))
  		return -EFAULT;
12352d3ca   Konstantin Khlebnikov   mm: replace vma_l...
1985
1986
1987
1988
1989
1990
1991
  	/* Guard against wrapping around to address 0. */
  	if (address < PAGE_ALIGN(address+4))
  		address = PAGE_ALIGN(address+4);
  	else
  		return -ENOMEM;
  
  	/* We must make sure the anon_vma is allocated. */
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1992
1993
  	if (unlikely(anon_vma_prepare(vma)))
  		return -ENOMEM;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1994
1995
1996
1997
1998
1999
  
  	/*
  	 * 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.
  	 */
12352d3ca   Konstantin Khlebnikov   mm: replace vma_l...
2000
  	anon_vma_lock_write(vma->anon_vma);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2001
2002
2003
2004
2005
2006
2007
  
  	/* 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;
42c36f63a   Hugh Dickins   vm: fix vm_pgoff ...
2008
2009
2010
2011
  		error = -ENOMEM;
  		if (vma->vm_pgoff + (size >> PAGE_SHIFT) >= vma->vm_pgoff) {
  			error = acct_stack_growth(vma, size, grow);
  			if (!error) {
4128997b5   Michel Lespinasse   mm: protect again...
2012
2013
2014
2015
2016
  				/*
  				 * vma_gap_update() doesn't support concurrent
  				 * updates, but we only hold a shared mmap_sem
  				 * lock here, so we need to protect against
  				 * concurrent vma expansions.
12352d3ca   Konstantin Khlebnikov   mm: replace vma_l...
2017
  				 * anon_vma_lock_write() doesn't help here, as
4128997b5   Michel Lespinasse   mm: protect again...
2018
2019
2020
2021
2022
  				 * we don't guarantee that all growable vmas
  				 * in a mm share the same root anon vma.
  				 * So, we reuse mm->page_table_lock to guard
  				 * against concurrent vma expansions.
  				 */
093578147   Oleg Nesterov   mm: add the "stru...
2023
  				spin_lock(&mm->page_table_lock);
87e8827b3   Oleg Nesterov   mm: fix the racy ...
2024
  				if (vma->vm_flags & VM_LOCKED)
093578147   Oleg Nesterov   mm: add the "stru...
2025
  					mm->locked_vm += grow;
846383359   Konstantin Khlebnikov   mm: rework virtua...
2026
  				vm_stat_account(mm, vma->vm_flags, grow);
bf181b9f9   Michel Lespinasse   mm anon rmap: rep...
2027
  				anon_vma_interval_tree_pre_update_vma(vma);
42c36f63a   Hugh Dickins   vm: fix vm_pgoff ...
2028
  				vma->vm_end = address;
bf181b9f9   Michel Lespinasse   mm anon rmap: rep...
2029
  				anon_vma_interval_tree_post_update_vma(vma);
d37371870   Michel Lespinasse   mm: augment vma r...
2030
2031
2032
  				if (vma->vm_next)
  					vma_gap_update(vma->vm_next);
  				else
093578147   Oleg Nesterov   mm: add the "stru...
2033
2034
  					mm->highest_vm_end = address;
  				spin_unlock(&mm->page_table_lock);
4128997b5   Michel Lespinasse   mm: protect again...
2035

42c36f63a   Hugh Dickins   vm: fix vm_pgoff ...
2036
2037
  				perf_event_mmap(vma);
  			}
3af9e8592   Eric B Munson   perf: Add non-exe...
2038
  		}
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2039
  	}
12352d3ca   Konstantin Khlebnikov   mm: replace vma_l...
2040
  	anon_vma_unlock_write(vma->anon_vma);
6d50e60cd   David Rientjes   mm, thp: fix coll...
2041
  	khugepaged_enter_vma_merge(vma, vma->vm_flags);
093578147   Oleg Nesterov   mm: add the "stru...
2042
  	validate_mm(mm);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2043
2044
  	return error;
  }
46dea3d09   Hugh Dickins   [PATCH] mm: ia64 ...
2045
  #endif /* CONFIG_STACK_GROWSUP || CONFIG_IA64 */
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2046
2047
2048
  /*
   * vma is the first one with address < vma->vm_start.  Have to extend vma.
   */
d05f3169c   Michal Hocko   mm: make expand_d...
2049
  int expand_downwards(struct vm_area_struct *vma,
b6a2fea39   Ollie Wild   mm: variable leng...
2050
  				   unsigned long address)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2051
  {
093578147   Oleg Nesterov   mm: add the "stru...
2052
  	struct mm_struct *mm = vma->vm_mm;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2053
  	int error;
8869477a4   Eric Paris   security: protect...
2054
  	address &= PAGE_MASK;
e5467859f   Al Viro   split ->file_mmap...
2055
  	error = security_mmap_addr(address);
8869477a4   Eric Paris   security: protect...
2056
2057
  	if (error)
  		return error;
12352d3ca   Konstantin Khlebnikov   mm: replace vma_l...
2058
2059
2060
  	/* We must make sure the anon_vma is allocated. */
  	if (unlikely(anon_vma_prepare(vma)))
  		return -ENOMEM;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2061
2062
2063
2064
2065
2066
  
  	/*
  	 * 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.
  	 */
12352d3ca   Konstantin Khlebnikov   mm: replace vma_l...
2067
  	anon_vma_lock_write(vma->anon_vma);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2068
2069
2070
2071
2072
2073
2074
  
  	/* 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;
a626ca6a6   Linus Torvalds   vm: fix vm_pgoff ...
2075
2076
2077
2078
  		error = -ENOMEM;
  		if (grow <= vma->vm_pgoff) {
  			error = acct_stack_growth(vma, size, grow);
  			if (!error) {
4128997b5   Michel Lespinasse   mm: protect again...
2079
2080
2081
2082
2083
  				/*
  				 * vma_gap_update() doesn't support concurrent
  				 * updates, but we only hold a shared mmap_sem
  				 * lock here, so we need to protect against
  				 * concurrent vma expansions.
12352d3ca   Konstantin Khlebnikov   mm: replace vma_l...
2084
  				 * anon_vma_lock_write() doesn't help here, as
4128997b5   Michel Lespinasse   mm: protect again...
2085
2086
2087
2088
2089
  				 * we don't guarantee that all growable vmas
  				 * in a mm share the same root anon vma.
  				 * So, we reuse mm->page_table_lock to guard
  				 * against concurrent vma expansions.
  				 */
093578147   Oleg Nesterov   mm: add the "stru...
2090
  				spin_lock(&mm->page_table_lock);
87e8827b3   Oleg Nesterov   mm: fix the racy ...
2091
  				if (vma->vm_flags & VM_LOCKED)
093578147   Oleg Nesterov   mm: add the "stru...
2092
  					mm->locked_vm += grow;
846383359   Konstantin Khlebnikov   mm: rework virtua...
2093
  				vm_stat_account(mm, vma->vm_flags, grow);
bf181b9f9   Michel Lespinasse   mm anon rmap: rep...
2094
  				anon_vma_interval_tree_pre_update_vma(vma);
a626ca6a6   Linus Torvalds   vm: fix vm_pgoff ...
2095
2096
  				vma->vm_start = address;
  				vma->vm_pgoff -= grow;
bf181b9f9   Michel Lespinasse   mm anon rmap: rep...
2097
  				anon_vma_interval_tree_post_update_vma(vma);
d37371870   Michel Lespinasse   mm: augment vma r...
2098
  				vma_gap_update(vma);
093578147   Oleg Nesterov   mm: add the "stru...
2099
  				spin_unlock(&mm->page_table_lock);
4128997b5   Michel Lespinasse   mm: protect again...
2100

a626ca6a6   Linus Torvalds   vm: fix vm_pgoff ...
2101
2102
  				perf_event_mmap(vma);
  			}
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2103
2104
  		}
  	}
12352d3ca   Konstantin Khlebnikov   mm: replace vma_l...
2105
  	anon_vma_unlock_write(vma->anon_vma);
6d50e60cd   David Rientjes   mm, thp: fix coll...
2106
  	khugepaged_enter_vma_merge(vma, vma->vm_flags);
093578147   Oleg Nesterov   mm: add the "stru...
2107
  	validate_mm(mm);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2108
2109
  	return error;
  }
098849643   Linus Torvalds   mm: do not grow t...
2110
2111
2112
2113
2114
2115
2116
2117
2118
2119
2120
  /*
   * Note how expand_stack() refuses to expand the stack all the way to
   * abut the next virtual mapping, *unless* that mapping itself is also
   * a stack mapping. We want to leave room for a guard page, after all
   * (the guard page itself is not added here, that is done by the
   * actual page faulting logic)
   *
   * This matches the behavior of the guard page logic (see mm/memory.c:
   * check_stack_guard_page()), which only allows the guard page to be
   * removed under these circumstances.
   */
b6a2fea39   Ollie Wild   mm: variable leng...
2121
2122
2123
  #ifdef CONFIG_STACK_GROWSUP
  int expand_stack(struct vm_area_struct *vma, unsigned long address)
  {
098849643   Linus Torvalds   mm: do not grow t...
2124
2125
2126
2127
2128
2129
2130
2131
  	struct vm_area_struct *next;
  
  	address &= PAGE_MASK;
  	next = vma->vm_next;
  	if (next && next->vm_start == address + PAGE_SIZE) {
  		if (!(next->vm_flags & VM_GROWSUP))
  			return -ENOMEM;
  	}
b6a2fea39   Ollie Wild   mm: variable leng...
2132
2133
2134
2135
2136
2137
2138
2139
2140
2141
2142
2143
  	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;
1c1271850   Denys Vlasenko   parisc: fix find_...
2144
  	if (!prev || expand_stack(prev, addr))
b6a2fea39   Ollie Wild   mm: variable leng...
2145
  		return NULL;
cea10a19b   Michel Lespinasse   mm: directly use ...
2146
  	if (prev->vm_flags & VM_LOCKED)
fc05f5662   Kirill A. Shutemov   mm: rename __mloc...
2147
  		populate_vma_page_range(prev, addr, prev->vm_end, NULL);
b6a2fea39   Ollie Wild   mm: variable leng...
2148
2149
2150
2151
2152
  	return prev;
  }
  #else
  int expand_stack(struct vm_area_struct *vma, unsigned long address)
  {
098849643   Linus Torvalds   mm: do not grow t...
2153
2154
2155
2156
2157
2158
2159
2160
  	struct vm_area_struct *prev;
  
  	address &= PAGE_MASK;
  	prev = vma->vm_prev;
  	if (prev && prev->vm_end == address) {
  		if (!(prev->vm_flags & VM_GROWSDOWN))
  			return -ENOMEM;
  	}
b6a2fea39   Ollie Wild   mm: variable leng...
2161
2162
  	return expand_downwards(vma, address);
  }
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2163
  struct vm_area_struct *
cc71aba34   vishnu.ps   mm/mmap.c: whites...
2164
  find_extend_vma(struct mm_struct *mm, unsigned long addr)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2165
  {
cc71aba34   vishnu.ps   mm/mmap.c: whites...
2166
  	struct vm_area_struct *vma;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2167
2168
2169
  	unsigned long start;
  
  	addr &= PAGE_MASK;
cc71aba34   vishnu.ps   mm/mmap.c: whites...
2170
  	vma = find_vma(mm, addr);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2171
2172
2173
2174
2175
2176
2177
2178
2179
  	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;
cea10a19b   Michel Lespinasse   mm: directly use ...
2180
  	if (vma->vm_flags & VM_LOCKED)
fc05f5662   Kirill A. Shutemov   mm: rename __mloc...
2181
  		populate_vma_page_range(vma, addr, start, NULL);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2182
2183
2184
  	return vma;
  }
  #endif
e1d6d01ab   Jesse Barnes   mm: export find_e...
2185
  EXPORT_SYMBOL_GPL(find_extend_vma);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2186
  /*
2c0b38146   Hugh Dickins   [PATCH] mm: remov...
2187
   * Ok - we have the memory areas we should free on the vma list,
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2188
   * so release them, and do the vma updates.
2c0b38146   Hugh Dickins   [PATCH] mm: remov...
2189
2190
   *
   * Called with the mm semaphore held.
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2191
   */
2c0b38146   Hugh Dickins   [PATCH] mm: remov...
2192
  static void remove_vma_list(struct mm_struct *mm, struct vm_area_struct *vma)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2193
  {
4f74d2c8e   Linus Torvalds   vm: remove 'nr_ac...
2194
  	unsigned long nr_accounted = 0;
365e9c87a   Hugh Dickins   [PATCH] mm: updat...
2195
2196
  	/* Update high watermark before we lower total_vm */
  	update_hiwater_vm(mm);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2197
  	do {
2c0b38146   Hugh Dickins   [PATCH] mm: remov...
2198
  		long nrpages = vma_pages(vma);
4f74d2c8e   Linus Torvalds   vm: remove 'nr_ac...
2199
2200
  		if (vma->vm_flags & VM_ACCOUNT)
  			nr_accounted += nrpages;
846383359   Konstantin Khlebnikov   mm: rework virtua...
2201
  		vm_stat_account(mm, vma->vm_flags, -nrpages);
a8fb5618d   Hugh Dickins   [PATCH] mm: unlin...
2202
  		vma = remove_vma(vma);
146425a31   Hugh Dickins   [PATCH] freepgt: ...
2203
  	} while (vma);
4f74d2c8e   Linus Torvalds   vm: remove 'nr_ac...
2204
  	vm_unacct_memory(nr_accounted);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2205
2206
2207
2208
2209
2210
  	validate_mm(mm);
  }
  
  /*
   * Get rid of page table information in the indicated region.
   *
f10df6860   Paolo 'Blaisorblade' Giarrusso   [PATCH] fix locki...
2211
   * Called with the mm semaphore held.
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2212
2213
   */
  static void unmap_region(struct mm_struct *mm,
e0da382c9   Hugh Dickins   [PATCH] freepgt: ...
2214
2215
  		struct vm_area_struct *vma, struct vm_area_struct *prev,
  		unsigned long start, unsigned long end)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2216
  {
cc71aba34   vishnu.ps   mm/mmap.c: whites...
2217
  	struct vm_area_struct *next = prev ? prev->vm_next : mm->mmap;
d16dfc550   Peter Zijlstra   mm: mmu_gather re...
2218
  	struct mmu_gather tlb;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2219
2220
  
  	lru_add_drain();
2b047252d   Linus Torvalds   Fix TLB gather vi...
2221
  	tlb_gather_mmu(&tlb, mm, start, end);
365e9c87a   Hugh Dickins   [PATCH] mm: updat...
2222
  	update_hiwater_rss(mm);
4f74d2c8e   Linus Torvalds   vm: remove 'nr_ac...
2223
  	unmap_vmas(&tlb, vma, start, end);
d16dfc550   Peter Zijlstra   mm: mmu_gather re...
2224
  	free_pgtables(&tlb, vma, prev ? prev->vm_end : FIRST_USER_ADDRESS,
6ee8630e0   Hugh Dickins   mm: allow arch co...
2225
  				 next ? next->vm_start : USER_PGTABLES_CEILING);
d16dfc550   Peter Zijlstra   mm: mmu_gather re...
2226
  	tlb_finish_mmu(&tlb, start, end);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2227
2228
2229
2230
2231
2232
2233
2234
2235
2236
2237
2238
2239
2240
  }
  
  /*
   * 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;
  
  	insertion_point = (prev ? &prev->vm_next : &mm->mmap);
297c5eee3   Linus Torvalds   mm: make the vma ...
2241
  	vma->vm_prev = NULL;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2242
  	do {
d37371870   Michel Lespinasse   mm: augment vma r...
2243
  		vma_rb_erase(vma, &mm->mm_rb);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2244
2245
2246
2247
2248
  		mm->map_count--;
  		tail_vma = vma;
  		vma = vma->vm_next;
  	} while (vma && vma->vm_start < end);
  	*insertion_point = vma;
d37371870   Michel Lespinasse   mm: augment vma r...
2249
  	if (vma) {
297c5eee3   Linus Torvalds   mm: make the vma ...
2250
  		vma->vm_prev = prev;
d37371870   Michel Lespinasse   mm: augment vma r...
2251
2252
2253
  		vma_gap_update(vma);
  	} else
  		mm->highest_vm_end = prev ? prev->vm_end : 0;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2254
  	tail_vma->vm_next = NULL;
615d6e875   Davidlohr Bueso   mm: per-thread vm...
2255
2256
2257
  
  	/* Kill the cache */
  	vmacache_invalidate(mm);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2258
2259
2260
  }
  
  /*
659ace584   KOSAKI Motohiro   mmap: don't retur...
2261
2262
   * __split_vma() bypasses sysctl_max_map_count checking.  We use this on the
   * munmap path where it doesn't make sense to fail.
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2263
   */
cc71aba34   vishnu.ps   mm/mmap.c: whites...
2264
  static int __split_vma(struct mm_struct *mm, struct vm_area_struct *vma,
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2265
2266
  	      unsigned long addr, int new_below)
  {
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2267
  	struct vm_area_struct *new;
e39758912   Chen Gang   mm/mmap.c: simpli...
2268
  	int err;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2269

a55164389   Andi Kleen   hugetlb: modular ...
2270
2271
  	if (is_vm_hugetlb_page(vma) && (addr &
  					~(huge_page_mask(hstate_vma(vma)))))
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2272
  		return -EINVAL;
e94b17660   Christoph Lameter   [PATCH] slab: rem...
2273
  	new = kmem_cache_alloc(vm_area_cachep, GFP_KERNEL);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2274
  	if (!new)
e39758912   Chen Gang   mm/mmap.c: simpli...
2275
  		return -ENOMEM;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2276
2277
2278
  
  	/* most fields are the same, copy all, and then fixup */
  	*new = *vma;
5beb49305   Rik van Riel   mm: change anon_v...
2279
  	INIT_LIST_HEAD(&new->anon_vma_chain);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2280
2281
2282
2283
2284
2285
  	if (new_below)
  		new->vm_end = addr;
  	else {
  		new->vm_start = addr;
  		new->vm_pgoff += ((addr - vma->vm_start) >> PAGE_SHIFT);
  	}
ef0855d33   Oleg Nesterov   mm: mempolicy: tu...
2286
2287
  	err = vma_dup_policy(vma, new);
  	if (err)
5beb49305   Rik van Riel   mm: change anon_v...
2288
  		goto out_free_vma;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2289

c4ea95d7c   Daniel Forrest   mm: fix anon_vma_...
2290
2291
  	err = anon_vma_clone(new, vma);
  	if (err)
5beb49305   Rik van Riel   mm: change anon_v...
2292
  		goto out_free_mpol;
e9714acf8   Konstantin Khlebnikov   mm: kill vma flag...
2293
  	if (new->vm_file)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2294
2295
2296
2297
2298
2299
  		get_file(new->vm_file);
  
  	if (new->vm_ops && new->vm_ops->open)
  		new->vm_ops->open(new);
  
  	if (new_below)
5beb49305   Rik van Riel   mm: change anon_v...
2300
  		err = vma_adjust(vma, addr, vma->vm_end, vma->vm_pgoff +
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2301
2302
  			((addr - new->vm_start) >> PAGE_SHIFT), new);
  	else
5beb49305   Rik van Riel   mm: change anon_v...
2303
  		err = vma_adjust(vma, vma->vm_start, addr, vma->vm_pgoff, new);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2304

5beb49305   Rik van Riel   mm: change anon_v...
2305
2306
2307
2308
2309
  	/* Success. */
  	if (!err)
  		return 0;
  
  	/* Clean everything up if vma_adjust failed. */
589275338   Rik van Riel   mmap: check ->vm_...
2310
2311
  	if (new->vm_ops && new->vm_ops->close)
  		new->vm_ops->close(new);
e9714acf8   Konstantin Khlebnikov   mm: kill vma flag...
2312
  	if (new->vm_file)
5beb49305   Rik van Riel   mm: change anon_v...
2313
  		fput(new->vm_file);
2aeadc30d   Andrea Arcangeli   mmap: call unlink...
2314
  	unlink_anon_vmas(new);
5beb49305   Rik van Riel   mm: change anon_v...
2315
   out_free_mpol:
ef0855d33   Oleg Nesterov   mm: mempolicy: tu...
2316
  	mpol_put(vma_policy(new));
5beb49305   Rik van Riel   mm: change anon_v...
2317
2318
   out_free_vma:
  	kmem_cache_free(vm_area_cachep, new);
5beb49305   Rik van Riel   mm: change anon_v...
2319
  	return err;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2320
  }
659ace584   KOSAKI Motohiro   mmap: don't retur...
2321
2322
2323
2324
2325
2326
2327
2328
2329
2330
2331
2332
  /*
   * 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);
  }
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2333
2334
2335
2336
2337
2338
2339
2340
  /* 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;
146425a31   Hugh Dickins   [PATCH] freepgt: ...
2341
  	struct vm_area_struct *vma, *prev, *last;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2342

de1741a13   Alexander Kuleshov   mm/mmap: use offs...
2343
  	if ((offset_in_page(start)) || start > TASK_SIZE || len > TASK_SIZE-start)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2344
  		return -EINVAL;
cc71aba34   vishnu.ps   mm/mmap.c: whites...
2345
2346
  	len = PAGE_ALIGN(len);
  	if (len == 0)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2347
2348
2349
  		return -EINVAL;
  
  	/* Find the first overlapping VMA */
9be34c9d5   Linus Torvalds   mm: get rid of th...
2350
  	vma = find_vma(mm, start);
146425a31   Hugh Dickins   [PATCH] freepgt: ...
2351
  	if (!vma)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2352
  		return 0;
9be34c9d5   Linus Torvalds   mm: get rid of th...
2353
  	prev = vma->vm_prev;
146425a31   Hugh Dickins   [PATCH] freepgt: ...
2354
  	/* we have  start < vma->vm_end  */
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2355
2356
2357
  
  	/* if it doesn't overlap, we have nothing.. */
  	end = start + len;
146425a31   Hugh Dickins   [PATCH] freepgt: ...
2358
  	if (vma->vm_start >= end)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2359
2360
2361
2362
2363
2364
2365
2366
2367
  		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.
  	 */
146425a31   Hugh Dickins   [PATCH] freepgt: ...
2368
  	if (start > vma->vm_start) {
659ace584   KOSAKI Motohiro   mmap: don't retur...
2369
2370
2371
2372
2373
2374
2375
2376
2377
2378
2379
  		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);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2380
2381
  		if (error)
  			return error;
146425a31   Hugh Dickins   [PATCH] freepgt: ...
2382
  		prev = vma;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2383
2384
2385
2386
2387
  	}
  
  	/* Does it split the last one? */
  	last = find_vma(mm, end);
  	if (last && end > last->vm_start) {
659ace584   KOSAKI Motohiro   mmap: don't retur...
2388
  		int error = __split_vma(mm, last, end, 1);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2389
2390
2391
  		if (error)
  			return error;
  	}
cc71aba34   vishnu.ps   mm/mmap.c: whites...
2392
  	vma = prev ? prev->vm_next : mm->mmap;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2393
2394
  
  	/*
ba470de43   Rik van Riel   mmap: handle mloc...
2395
2396
2397
2398
2399
2400
2401
2402
2403
2404
2405
2406
2407
2408
  	 * 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;
  		}
  	}
  
  	/*
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2409
2410
  	 * Remove the vma's, and unmap the actual pages
  	 */
146425a31   Hugh Dickins   [PATCH] freepgt: ...
2411
2412
  	detach_vmas_to_be_unmapped(mm, vma, prev, end);
  	unmap_region(mm, vma, prev, start, end);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2413

1de4fa14e   Dave Hansen   x86, mpx: Cleanup...
2414
  	arch_unmap(mm, vma, start, end);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2415
  	/* Fix up all other VM information */
2c0b38146   Hugh Dickins   [PATCH] mm: remov...
2416
  	remove_vma_list(mm, vma);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2417
2418
2419
  
  	return 0;
  }
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2420

bfce281c2   Al Viro   kill mm argument ...
2421
  int vm_munmap(unsigned long start, size_t len)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2422
2423
  {
  	int ret;
bfce281c2   Al Viro   kill mm argument ...
2424
  	struct mm_struct *mm = current->mm;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2425

ae7987835   Michal Hocko   mm: make vm_munma...
2426
2427
  	if (down_write_killable(&mm->mmap_sem))
  		return -EINTR;
a46ef99d8   Linus Torvalds   VM: add "vm_munma...
2428
  	ret = do_munmap(mm, start, len);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2429
2430
2431
  	up_write(&mm->mmap_sem);
  	return ret;
  }
a46ef99d8   Linus Torvalds   VM: add "vm_munma...
2432
2433
2434
2435
  EXPORT_SYMBOL(vm_munmap);
  
  SYSCALL_DEFINE2(munmap, unsigned long, addr, size_t, len)
  {
dc0ef0df7   Michal Hocko   mm: make mmap_sem...
2436
2437
  	int ret;
  	struct mm_struct *mm = current->mm;
a46ef99d8   Linus Torvalds   VM: add "vm_munma...
2438
  	profile_munmap(addr);
dc0ef0df7   Michal Hocko   mm: make mmap_sem...
2439
2440
2441
2442
2443
  	if (down_write_killable(&mm->mmap_sem))
  		return -EINTR;
  	ret = do_munmap(mm, addr, len);
  	up_write(&mm->mmap_sem);
  	return ret;
a46ef99d8   Linus Torvalds   VM: add "vm_munma...
2444
  }
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2445

c8d78c182   Kirill A. Shutemov   mm: replace remap...
2446
2447
2448
2449
2450
2451
2452
2453
2454
2455
2456
2457
2458
  
  /*
   * Emulation of deprecated remap_file_pages() syscall.
   */
  SYSCALL_DEFINE5(remap_file_pages, unsigned long, start, unsigned long, size,
  		unsigned long, prot, unsigned long, pgoff, unsigned long, flags)
  {
  
  	struct mm_struct *mm = current->mm;
  	struct vm_area_struct *vma;
  	unsigned long populate = 0;
  	unsigned long ret = -EINVAL;
  	struct file *file;
756a025f0   Joe Perches   mm: coalesce spli...
2459
2460
2461
  	pr_warn_once("%s (%d) uses deprecated remap_file_pages() syscall. See Documentation/vm/remap_file_pages.txt.
  ",
  		     current->comm, current->pid);
c8d78c182   Kirill A. Shutemov   mm: replace remap...
2462
2463
2464
2465
2466
2467
2468
2469
2470
2471
2472
2473
  
  	if (prot)
  		return ret;
  	start = start & PAGE_MASK;
  	size = size & PAGE_MASK;
  
  	if (start + size <= start)
  		return ret;
  
  	/* Does pgoff wrap? */
  	if (pgoff + (size >> PAGE_SHIFT) < pgoff)
  		return ret;
dc0ef0df7   Michal Hocko   mm: make mmap_sem...
2474
2475
  	if (down_write_killable(&mm->mmap_sem))
  		return -EINTR;
c8d78c182   Kirill A. Shutemov   mm: replace remap...
2476
2477
2478
2479
  	vma = find_vma(mm, start);
  
  	if (!vma || !(vma->vm_flags & VM_SHARED))
  		goto out;
48f7df329   Kirill A. Shutemov   mm: fix regressio...
2480
  	if (start < vma->vm_start)
c8d78c182   Kirill A. Shutemov   mm: replace remap...
2481
  		goto out;
48f7df329   Kirill A. Shutemov   mm: fix regressio...
2482
2483
2484
2485
2486
2487
2488
2489
2490
2491
2492
2493
2494
2495
2496
2497
2498
2499
2500
2501
  	if (start + size > vma->vm_end) {
  		struct vm_area_struct *next;
  
  		for (next = vma->vm_next; next; next = next->vm_next) {
  			/* hole between vmas ? */
  			if (next->vm_start != next->vm_prev->vm_end)
  				goto out;
  
  			if (next->vm_file != vma->vm_file)
  				goto out;
  
  			if (next->vm_flags != vma->vm_flags)
  				goto out;
  
  			if (start + size <= next->vm_end)
  				break;
  		}
  
  		if (!next)
  			goto out;
c8d78c182   Kirill A. Shutemov   mm: replace remap...
2502
2503
2504
2505
2506
2507
2508
2509
2510
  	}
  
  	prot |= vma->vm_flags & VM_READ ? PROT_READ : 0;
  	prot |= vma->vm_flags & VM_WRITE ? PROT_WRITE : 0;
  	prot |= vma->vm_flags & VM_EXEC ? PROT_EXEC : 0;
  
  	flags &= MAP_NONBLOCK;
  	flags |= MAP_SHARED | MAP_FIXED | MAP_POPULATE;
  	if (vma->vm_flags & VM_LOCKED) {
48f7df329   Kirill A. Shutemov   mm: fix regressio...
2511
  		struct vm_area_struct *tmp;
c8d78c182   Kirill A. Shutemov   mm: replace remap...
2512
  		flags |= MAP_LOCKED;
48f7df329   Kirill A. Shutemov   mm: fix regressio...
2513

c8d78c182   Kirill A. Shutemov   mm: replace remap...
2514
  		/* drop PG_Mlocked flag for over-mapped range */
48f7df329   Kirill A. Shutemov   mm: fix regressio...
2515
2516
  		for (tmp = vma; tmp->vm_start >= start + size;
  				tmp = tmp->vm_next) {
9a73f61bd   Kirill A. Shutemov   thp, mlock: do no...
2517
2518
2519
2520
2521
  			/*
  			 * Split pmd and munlock page on the border
  			 * of the range.
  			 */
  			vma_adjust_trans_huge(tmp, start, start + size, 0);
48f7df329   Kirill A. Shutemov   mm: fix regressio...
2522
2523
2524
2525
  			munlock_vma_pages_range(tmp,
  					max(tmp->vm_start, start),
  					min(tmp->vm_end, start + size));
  		}
c8d78c182   Kirill A. Shutemov   mm: replace remap...
2526
2527
2528
2529
2530
2531
2532
2533
2534
2535
2536
2537
2538
2539
  	}
  
  	file = get_file(vma->vm_file);
  	ret = do_mmap_pgoff(vma->vm_file, start, size,
  			prot, flags, pgoff, &populate);
  	fput(file);
  out:
  	up_write(&mm->mmap_sem);
  	if (populate)
  		mm_populate(ret, populate);
  	if (!IS_ERR_VALUE(ret))
  		ret = 0;
  	return ret;
  }
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2540
2541
  static inline void verify_mm_writelocked(struct mm_struct *mm)
  {
a241ec65a   Paul E. McKenney   [PATCH] RCU tortu...
2542
  #ifdef CONFIG_DEBUG_VM
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2543
2544
2545
2546
2547
2548
2549
2550
2551
2552
2553
2554
  	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.
   */
ba093a6d9   Kees Cook   mm: refuse wrappe...
2555
  static int do_brk(unsigned long addr, unsigned long request)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2556
  {
cc71aba34   vishnu.ps   mm/mmap.c: whites...
2557
2558
  	struct mm_struct *mm = current->mm;
  	struct vm_area_struct *vma, *prev;
ba093a6d9   Kees Cook   mm: refuse wrappe...
2559
  	unsigned long flags, len;
cc71aba34   vishnu.ps   mm/mmap.c: whites...
2560
  	struct rb_node **rb_link, *rb_parent;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2561
  	pgoff_t pgoff = addr >> PAGE_SHIFT;
3a4597568   Kirill Korotaev   [PATCH] IA64,spar...
2562
  	int error;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2563

ba093a6d9   Kees Cook   mm: refuse wrappe...
2564
2565
2566
  	len = PAGE_ALIGN(request);
  	if (len < request)
  		return -ENOMEM;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2567
  	if (!len)
5d22fc25d   Linus Torvalds   mm: remove more I...
2568
  		return 0;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2569

3a4597568   Kirill Korotaev   [PATCH] IA64,spar...
2570
  	flags = VM_DATA_DEFAULT_FLAGS | VM_ACCOUNT | mm->def_flags;
2c6a10161   Al Viro   switch do_brk() t...
2571
  	error = get_unmapped_area(NULL, addr, len, 0, MAP_FIXED);
de1741a13   Alexander Kuleshov   mm/mmap: use offs...
2572
  	if (offset_in_page(error))
3a4597568   Kirill Korotaev   [PATCH] IA64,spar...
2573
  		return error;
363ee17f0   Davidlohr Bueso   mm/mmap.c: add ml...
2574
2575
2576
  	error = mlock_future_check(mm, mm->def_flags, len);
  	if (error)
  		return error;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2577
2578
2579
2580
2581
2582
2583
2584
2585
2586
  
  	/*
  	 * 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
  	 */
9fcd14571   Rasmus Villemoes   mm/mmap.c: use wh...
2587
2588
  	while (find_vma_links(mm, addr, addr + len, &prev, &rb_link,
  			      &rb_parent)) {
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2589
2590
  		if (do_munmap(mm, addr, len))
  			return -ENOMEM;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2591
2592
2593
  	}
  
  	/* Check against address space limits *after* clearing old maps... */
846383359   Konstantin Khlebnikov   mm: rework virtua...
2594
  	if (!may_expand_vm(mm, flags, len >> PAGE_SHIFT))
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2595
2596
2597
2598
  		return -ENOMEM;
  
  	if (mm->map_count > sysctl_max_map_count)
  		return -ENOMEM;
191c54244   Al Viro   mm: collapse secu...
2599
  	if (security_vm_enough_memory_mm(mm, len >> PAGE_SHIFT))
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2600
  		return -ENOMEM;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2601
  	/* Can we just expand an old private anonymous mapping? */
ba470de43   Rik van Riel   mmap: handle mloc...
2602
  	vma = vma_merge(mm, prev, addr, addr + len, flags,
19a809afe   Andrea Arcangeli   userfaultfd: teac...
2603
  			NULL, NULL, pgoff, NULL, NULL_VM_UFFD_CTX);
ba470de43   Rik van Riel   mmap: handle mloc...
2604
  	if (vma)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2605
2606
2607
2608
2609
  		goto out;
  
  	/*
  	 * create a vma struct for an anonymous mapping
  	 */
c5e3b83e9   Pekka Enberg   [PATCH] mm: use k...
2610
  	vma = kmem_cache_zalloc(vm_area_cachep, GFP_KERNEL);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2611
2612
2613
2614
  	if (!vma) {
  		vm_unacct_memory(len >> PAGE_SHIFT);
  		return -ENOMEM;
  	}
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2615

5beb49305   Rik van Riel   mm: change anon_v...
2616
  	INIT_LIST_HEAD(&vma->anon_vma_chain);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2617
2618
2619
2620
2621
  	vma->vm_mm = mm;
  	vma->vm_start = addr;
  	vma->vm_end = addr + len;
  	vma->vm_pgoff = pgoff;
  	vma->vm_flags = flags;
3ed75eb8f   Coly Li   setup vma->vm_pag...
2622
  	vma->vm_page_prot = vm_get_page_prot(flags);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2623
2624
  	vma_link(mm, vma, prev, rb_link, rb_parent);
  out:
3af9e8592   Eric B Munson   perf: Add non-exe...
2625
  	perf_event_mmap(vma);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2626
  	mm->total_vm += len >> PAGE_SHIFT;
846383359   Konstantin Khlebnikov   mm: rework virtua...
2627
  	mm->data_vm += len >> PAGE_SHIFT;
128557ffe   Michel Lespinasse   mm: use mm_popula...
2628
2629
  	if (flags & VM_LOCKED)
  		mm->locked_vm += (len >> PAGE_SHIFT);
d9104d1ca   Cyrill Gorcunov   mm: track vma cha...
2630
  	vma->vm_flags |= VM_SOFTDIRTY;
5d22fc25d   Linus Torvalds   mm: remove more I...
2631
  	return 0;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2632
  }
5d22fc25d   Linus Torvalds   mm: remove more I...
2633
  int vm_brk(unsigned long addr, unsigned long len)
e4eb1ff61   Linus Torvalds   VM: add "vm_brk()...
2634
2635
  {
  	struct mm_struct *mm = current->mm;
5d22fc25d   Linus Torvalds   mm: remove more I...
2636
  	int ret;
128557ffe   Michel Lespinasse   mm: use mm_popula...
2637
  	bool populate;
e4eb1ff61   Linus Torvalds   VM: add "vm_brk()...
2638

2d6c92824   Michal Hocko   mm: make vm_brk k...
2639
2640
  	if (down_write_killable(&mm->mmap_sem))
  		return -EINTR;
e4eb1ff61   Linus Torvalds   VM: add "vm_brk()...
2641
  	ret = do_brk(addr, len);
128557ffe   Michel Lespinasse   mm: use mm_popula...
2642
  	populate = ((mm->def_flags & VM_LOCKED) != 0);
e4eb1ff61   Linus Torvalds   VM: add "vm_brk()...
2643
  	up_write(&mm->mmap_sem);
5d22fc25d   Linus Torvalds   mm: remove more I...
2644
  	if (populate && !ret)
128557ffe   Michel Lespinasse   mm: use mm_popula...
2645
  		mm_populate(addr, len);
e4eb1ff61   Linus Torvalds   VM: add "vm_brk()...
2646
2647
2648
  	return ret;
  }
  EXPORT_SYMBOL(vm_brk);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2649
2650
2651
2652
  
  /* Release all mmaps. */
  void exit_mmap(struct mm_struct *mm)
  {
d16dfc550   Peter Zijlstra   mm: mmu_gather re...
2653
  	struct mmu_gather tlb;
ba470de43   Rik van Riel   mmap: handle mloc...
2654
  	struct vm_area_struct *vma;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2655
  	unsigned long nr_accounted = 0;
d6dd61c83   Jeremy Fitzhardinge   [PATCH] x86: PARA...
2656
  	/* mm's last user has gone, and its about to be pulled down */
cddb8a5c1   Andrea Arcangeli   mmu-notifiers: core
2657
  	mmu_notifier_release(mm);
d6dd61c83   Jeremy Fitzhardinge   [PATCH] x86: PARA...
2658

ba470de43   Rik van Riel   mmap: handle mloc...
2659
2660
2661
2662
2663
2664
2665
2666
  	if (mm->locked_vm) {
  		vma = mm->mmap;
  		while (vma) {
  			if (vma->vm_flags & VM_LOCKED)
  				munlock_vma_pages_all(vma);
  			vma = vma->vm_next;
  		}
  	}
9480c53e9   Jeremy Fitzhardinge   mm: rearrange exi...
2667
2668
  
  	arch_exit_mmap(mm);
ba470de43   Rik van Riel   mmap: handle mloc...
2669
  	vma = mm->mmap;
9480c53e9   Jeremy Fitzhardinge   mm: rearrange exi...
2670
2671
  	if (!vma)	/* Can happen if dup_mmap() received an OOM */
  		return;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2672
  	lru_add_drain();
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2673
  	flush_cache_mm(mm);
2b047252d   Linus Torvalds   Fix TLB gather vi...
2674
  	tlb_gather_mmu(&tlb, mm, 0, -1);
901608d90   Oleg Nesterov   mm: introduce get...
2675
  	/* update_hiwater_rss(mm) here? but nobody should be looking */
e0da382c9   Hugh Dickins   [PATCH] freepgt: ...
2676
  	/* Use -1 here to ensure all VMAs in the mm are unmapped */
4f74d2c8e   Linus Torvalds   vm: remove 'nr_ac...
2677
  	unmap_vmas(&tlb, vma, 0, -1);
9ba692948   Hugh Dickins   ksm: fix oom dead...
2678

6ee8630e0   Hugh Dickins   mm: allow arch co...
2679
  	free_pgtables(&tlb, vma, FIRST_USER_ADDRESS, USER_PGTABLES_CEILING);
853f5e264   Al Viro   VM: don't bother ...
2680
  	tlb_finish_mmu(&tlb, 0, -1);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2681

1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2682
  	/*
8f4f8c164   Hugh Dickins   [PATCH] mm: unlin...
2683
2684
  	 * Walk the list again, actually closing and freeing it,
  	 * with preemption enabled, without holding any MM locks.
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2685
  	 */
4f74d2c8e   Linus Torvalds   vm: remove 'nr_ac...
2686
2687
2688
  	while (vma) {
  		if (vma->vm_flags & VM_ACCOUNT)
  			nr_accounted += vma_pages(vma);
a8fb5618d   Hugh Dickins   [PATCH] mm: unlin...
2689
  		vma = remove_vma(vma);
4f74d2c8e   Linus Torvalds   vm: remove 'nr_ac...
2690
2691
  	}
  	vm_unacct_memory(nr_accounted);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2692
2693
2694
2695
  }
  
  /* Insert vm structure into process list sorted by address
   * and into the inode's i_mmap tree.  If vm_file is non-NULL
c8c06efa8   Davidlohr Bueso   mm: convert i_mma...
2696
   * then i_mmap_rwsem is taken here.
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2697
   */
6597d7833   Hugh Dickins   mm/mmap.c: replac...
2698
  int insert_vm_struct(struct mm_struct *mm, struct vm_area_struct *vma)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2699
  {
6597d7833   Hugh Dickins   mm/mmap.c: replac...
2700
2701
  	struct vm_area_struct *prev;
  	struct rb_node **rb_link, *rb_parent;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2702

c9d13f5fc   Chen Gang   mm/mmap.c:insert_...
2703
2704
2705
2706
2707
2708
  	if (find_vma_links(mm, vma->vm_start, vma->vm_end,
  			   &prev, &rb_link, &rb_parent))
  		return -ENOMEM;
  	if ((vma->vm_flags & VM_ACCOUNT) &&
  	     security_vm_enough_memory_mm(mm, vma_pages(vma)))
  		return -ENOMEM;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2709
2710
2711
2712
2713
2714
2715
2716
2717
2718
2719
2720
  	/*
  	 * 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.
  	 */
8a9cc3b55   Oleg Nesterov   mmap: fix the usa...
2721
  	if (vma_is_anonymous(vma)) {
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2722
2723
2724
  		BUG_ON(vma->anon_vma);
  		vma->vm_pgoff = vma->vm_start >> PAGE_SHIFT;
  	}
2b1444983   Srikar Dronamraju   uprobes, mm, x86:...
2725

1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2726
2727
2728
2729
2730
2731
2732
2733
2734
  	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,
38a76013a   Michel Lespinasse   mm: avoid taking ...
2735
2736
  	unsigned long addr, unsigned long len, pgoff_t pgoff,
  	bool *need_rmap_locks)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2737
2738
2739
2740
2741
2742
  {
  	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;
948f017b0   Andrea Arcangeli   mremap: enforce r...
2743
  	bool faulted_in_anon_vma = true;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2744
2745
2746
2747
2748
  
  	/*
  	 * If anonymous vma has not yet been faulted, update new pgoff
  	 * to match new location, to increase its chance of merging.
  	 */
ce75799b8   Oleg Nesterov   mremap: fix the w...
2749
  	if (unlikely(vma_is_anonymous(vma) && !vma->anon_vma)) {
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2750
  		pgoff = addr >> PAGE_SHIFT;
948f017b0   Andrea Arcangeli   mremap: enforce r...
2751
2752
  		faulted_in_anon_vma = false;
  	}
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2753

6597d7833   Hugh Dickins   mm/mmap.c: replac...
2754
2755
  	if (find_vma_links(mm, addr, addr + len, &prev, &rb_link, &rb_parent))
  		return NULL;	/* should never get here */
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2756
  	new_vma = vma_merge(mm, prev, addr, addr + len, vma->vm_flags,
19a809afe   Andrea Arcangeli   userfaultfd: teac...
2757
2758
  			    vma->anon_vma, vma->vm_file, pgoff, vma_policy(vma),
  			    vma->vm_userfaultfd_ctx);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2759
2760
2761
2762
  	if (new_vma) {
  		/*
  		 * Source vma may have been merged into new_vma
  		 */
948f017b0   Andrea Arcangeli   mremap: enforce r...
2763
2764
2765
2766
2767
2768
2769
2770
2771
2772
2773
2774
2775
2776
  		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.
  			 */
81d1b09c6   Sasha Levin   mm: convert a few...
2777
  			VM_BUG_ON_VMA(faulted_in_anon_vma, new_vma);
38a76013a   Michel Lespinasse   mm: avoid taking ...
2778
  			*vmap = vma = new_vma;
108d6642a   Michel Lespinasse   mm anon rmap: rem...
2779
  		}
38a76013a   Michel Lespinasse   mm: avoid taking ...
2780
  		*need_rmap_locks = (new_vma->vm_pgoff <= vma->vm_pgoff);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2781
  	} else {
e94b17660   Christoph Lameter   [PATCH] slab: rem...
2782
  		new_vma = kmem_cache_alloc(vm_area_cachep, GFP_KERNEL);
e39758912   Chen Gang   mm/mmap.c: simpli...
2783
2784
2785
2786
2787
2788
2789
2790
2791
2792
2793
2794
2795
2796
2797
2798
2799
  		if (!new_vma)
  			goto out;
  		*new_vma = *vma;
  		new_vma->vm_start = addr;
  		new_vma->vm_end = addr + len;
  		new_vma->vm_pgoff = pgoff;
  		if (vma_dup_policy(vma, new_vma))
  			goto out_free_vma;
  		INIT_LIST_HEAD(&new_vma->anon_vma_chain);
  		if (anon_vma_clone(new_vma, vma))
  			goto out_free_mempol;
  		if (new_vma->vm_file)
  			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);
  		*need_rmap_locks = false;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2800
2801
  	}
  	return new_vma;
5beb49305   Rik van Riel   mm: change anon_v...
2802

e39758912   Chen Gang   mm/mmap.c: simpli...
2803
  out_free_mempol:
ef0855d33   Oleg Nesterov   mm: mempolicy: tu...
2804
  	mpol_put(vma_policy(new_vma));
e39758912   Chen Gang   mm/mmap.c: simpli...
2805
  out_free_vma:
5beb49305   Rik van Riel   mm: change anon_v...
2806
  	kmem_cache_free(vm_area_cachep, new_vma);
e39758912   Chen Gang   mm/mmap.c: simpli...
2807
  out:
5beb49305   Rik van Riel   mm: change anon_v...
2808
  	return NULL;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2809
  }
119f657c7   Andrew Morton   [PATCH] RLIMIT_AS...
2810
2811
2812
2813
2814
  
  /*
   * Return true if the calling process may expand its vm space by the passed
   * number of pages
   */
846383359   Konstantin Khlebnikov   mm: rework virtua...
2815
  bool may_expand_vm(struct mm_struct *mm, vm_flags_t flags, unsigned long npages)
119f657c7   Andrew Morton   [PATCH] RLIMIT_AS...
2816
  {
846383359   Konstantin Khlebnikov   mm: rework virtua...
2817
2818
  	if (mm->total_vm + npages > rlimit(RLIMIT_AS) >> PAGE_SHIFT)
  		return false;
119f657c7   Andrew Morton   [PATCH] RLIMIT_AS...
2819

d977d56ce   Konstantin Khlebnikov   mm: warn about Vm...
2820
2821
  	if (is_data_mapping(flags) &&
  	    mm->data_vm + npages > rlimit(RLIMIT_DATA) >> PAGE_SHIFT) {
f4fcd5584   Konstantin Khlebnikov   mm: enable RLIMIT...
2822
2823
2824
2825
2826
2827
2828
  		/* Workaround for Valgrind */
  		if (rlimit(RLIMIT_DATA) == 0 &&
  		    mm->data_vm + npages <= rlimit_max(RLIMIT_DATA) >> PAGE_SHIFT)
  			return true;
  		if (!ignore_rlimit_data) {
  			pr_warn_once("%s (%d): VmData %lu exceed data ulimit %lu. Update limits or use boot option ignore_rlimit_data.
  ",
d977d56ce   Konstantin Khlebnikov   mm: warn about Vm...
2829
2830
2831
  				     current->comm, current->pid,
  				     (mm->data_vm + npages) << PAGE_SHIFT,
  				     rlimit(RLIMIT_DATA));
d977d56ce   Konstantin Khlebnikov   mm: warn about Vm...
2832
  			return false;
f4fcd5584   Konstantin Khlebnikov   mm: enable RLIMIT...
2833
  		}
d977d56ce   Konstantin Khlebnikov   mm: warn about Vm...
2834
  	}
119f657c7   Andrew Morton   [PATCH] RLIMIT_AS...
2835

846383359   Konstantin Khlebnikov   mm: rework virtua...
2836
2837
2838
2839
2840
2841
  	return true;
  }
  
  void vm_stat_account(struct mm_struct *mm, vm_flags_t flags, long npages)
  {
  	mm->total_vm += npages;
d977d56ce   Konstantin Khlebnikov   mm: warn about Vm...
2842
  	if (is_exec_mapping(flags))
846383359   Konstantin Khlebnikov   mm: rework virtua...
2843
  		mm->exec_vm += npages;
d977d56ce   Konstantin Khlebnikov   mm: warn about Vm...
2844
  	else if (is_stack_mapping(flags))
846383359   Konstantin Khlebnikov   mm: rework virtua...
2845
  		mm->stack_vm += npages;
d977d56ce   Konstantin Khlebnikov   mm: warn about Vm...
2846
  	else if (is_data_mapping(flags))
846383359   Konstantin Khlebnikov   mm: rework virtua...
2847
  		mm->data_vm += npages;
119f657c7   Andrew Morton   [PATCH] RLIMIT_AS...
2848
  }
fa5dc22f8   Roland McGrath   [PATCH] Add insta...
2849

a62c34bd2   Andy Lutomirski   x86, mm: Improve ...
2850
2851
2852
2853
2854
2855
2856
2857
2858
2859
2860
2861
2862
2863
  static int special_mapping_fault(struct vm_area_struct *vma,
  				 struct vm_fault *vmf);
  
  /*
   * Having a close hook prevents vma merging regardless of flags.
   */
  static void special_mapping_close(struct vm_area_struct *vma)
  {
  }
  
  static const char *special_mapping_name(struct vm_area_struct *vma)
  {
  	return ((struct vm_special_mapping *)vma->vm_private_data)->name;
  }
b059a453b   Dmitry Safonov   x86/vdso: Add mre...
2864
2865
2866
2867
2868
2869
2870
2871
  static int special_mapping_mremap(struct vm_area_struct *new_vma)
  {
  	struct vm_special_mapping *sm = new_vma->vm_private_data;
  
  	if (sm->mremap)
  		return sm->mremap(sm, new_vma);
  	return 0;
  }
a62c34bd2   Andy Lutomirski   x86, mm: Improve ...
2872
2873
2874
  static const struct vm_operations_struct special_mapping_vmops = {
  	.close = special_mapping_close,
  	.fault = special_mapping_fault,
b059a453b   Dmitry Safonov   x86/vdso: Add mre...
2875
  	.mremap = special_mapping_mremap,
a62c34bd2   Andy Lutomirski   x86, mm: Improve ...
2876
2877
2878
2879
2880
2881
2882
  	.name = special_mapping_name,
  };
  
  static const struct vm_operations_struct legacy_special_mapping_vmops = {
  	.close = special_mapping_close,
  	.fault = special_mapping_fault,
  };
fa5dc22f8   Roland McGrath   [PATCH] Add insta...
2883

b1d0e4f53   Nick Piggin   mm: special mappi...
2884
2885
  static int special_mapping_fault(struct vm_area_struct *vma,
  				struct vm_fault *vmf)
fa5dc22f8   Roland McGrath   [PATCH] Add insta...
2886
  {
b1d0e4f53   Nick Piggin   mm: special mappi...
2887
  	pgoff_t pgoff;
fa5dc22f8   Roland McGrath   [PATCH] Add insta...
2888
  	struct page **pages;
f872f5400   Andy Lutomirski   mm: Add a vm_spec...
2889
  	if (vma->vm_ops == &legacy_special_mapping_vmops) {
a62c34bd2   Andy Lutomirski   x86, mm: Improve ...
2890
  		pages = vma->vm_private_data;
f872f5400   Andy Lutomirski   mm: Add a vm_spec...
2891
2892
2893
2894
2895
2896
2897
2898
  	} else {
  		struct vm_special_mapping *sm = vma->vm_private_data;
  
  		if (sm->fault)
  			return sm->fault(sm, vma, vmf);
  
  		pages = sm->pages;
  	}
a62c34bd2   Andy Lutomirski   x86, mm: Improve ...
2899

8a9cc3b55   Oleg Nesterov   mmap: fix the usa...
2900
  	for (pgoff = vmf->pgoff; pgoff && *pages; ++pages)
b1d0e4f53   Nick Piggin   mm: special mappi...
2901
  		pgoff--;
fa5dc22f8   Roland McGrath   [PATCH] Add insta...
2902
2903
2904
2905
  
  	if (*pages) {
  		struct page *page = *pages;
  		get_page(page);
b1d0e4f53   Nick Piggin   mm: special mappi...
2906
2907
  		vmf->page = page;
  		return 0;
fa5dc22f8   Roland McGrath   [PATCH] Add insta...
2908
  	}
b1d0e4f53   Nick Piggin   mm: special mappi...
2909
  	return VM_FAULT_SIGBUS;
fa5dc22f8   Roland McGrath   [PATCH] Add insta...
2910
  }
a62c34bd2   Andy Lutomirski   x86, mm: Improve ...
2911
2912
2913
  static struct vm_area_struct *__install_special_mapping(
  	struct mm_struct *mm,
  	unsigned long addr, unsigned long len,
27f28b972   Chen Gang   mm/mmap.c: change...
2914
2915
  	unsigned long vm_flags, void *priv,
  	const struct vm_operations_struct *ops)
fa5dc22f8   Roland McGrath   [PATCH] Add insta...
2916
  {
462e635e5   Tavis Ormandy   install_special_m...
2917
  	int ret;
fa5dc22f8   Roland McGrath   [PATCH] Add insta...
2918
2919
2920
2921
  	struct vm_area_struct *vma;
  
  	vma = kmem_cache_zalloc(vm_area_cachep, GFP_KERNEL);
  	if (unlikely(vma == NULL))
3935ed6a3   Stefani Seibold   mm: Add new func ...
2922
  		return ERR_PTR(-ENOMEM);
fa5dc22f8   Roland McGrath   [PATCH] Add insta...
2923

5beb49305   Rik van Riel   mm: change anon_v...
2924
  	INIT_LIST_HEAD(&vma->anon_vma_chain);
fa5dc22f8   Roland McGrath   [PATCH] Add insta...
2925
2926
2927
  	vma->vm_mm = mm;
  	vma->vm_start = addr;
  	vma->vm_end = addr + len;
d9104d1ca   Cyrill Gorcunov   mm: track vma cha...
2928
  	vma->vm_flags = vm_flags | mm->def_flags | VM_DONTEXPAND | VM_SOFTDIRTY;
3ed75eb8f   Coly Li   setup vma->vm_pag...
2929
  	vma->vm_page_prot = vm_get_page_prot(vma->vm_flags);
fa5dc22f8   Roland McGrath   [PATCH] Add insta...
2930

a62c34bd2   Andy Lutomirski   x86, mm: Improve ...
2931
2932
  	vma->vm_ops = ops;
  	vma->vm_private_data = priv;
fa5dc22f8   Roland McGrath   [PATCH] Add insta...
2933

462e635e5   Tavis Ormandy   install_special_m...
2934
2935
2936
  	ret = insert_vm_struct(mm, vma);
  	if (ret)
  		goto out;
fa5dc22f8   Roland McGrath   [PATCH] Add insta...
2937

846383359   Konstantin Khlebnikov   mm: rework virtua...
2938
  	vm_stat_account(mm, vma->vm_flags, len >> PAGE_SHIFT);
fa5dc22f8   Roland McGrath   [PATCH] Add insta...
2939

cdd6c482c   Ingo Molnar   perf: Do the big ...
2940
  	perf_event_mmap(vma);
089dd79db   Peter Zijlstra   perf_counter: Gen...
2941

3935ed6a3   Stefani Seibold   mm: Add new func ...
2942
  	return vma;
462e635e5   Tavis Ormandy   install_special_m...
2943
2944
2945
  
  out:
  	kmem_cache_free(vm_area_cachep, vma);
3935ed6a3   Stefani Seibold   mm: Add new func ...
2946
2947
  	return ERR_PTR(ret);
  }
a62c34bd2   Andy Lutomirski   x86, mm: Improve ...
2948
2949
2950
2951
2952
2953
2954
2955
2956
2957
2958
2959
2960
2961
  /*
   * 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.
   */
  struct vm_area_struct *_install_special_mapping(
  	struct mm_struct *mm,
  	unsigned long addr, unsigned long len,
  	unsigned long vm_flags, const struct vm_special_mapping *spec)
  {
27f28b972   Chen Gang   mm/mmap.c: change...
2962
2963
  	return __install_special_mapping(mm, addr, len, vm_flags, (void *)spec,
  					&special_mapping_vmops);
a62c34bd2   Andy Lutomirski   x86, mm: Improve ...
2964
  }
3935ed6a3   Stefani Seibold   mm: Add new func ...
2965
2966
2967
2968
  int install_special_mapping(struct mm_struct *mm,
  			    unsigned long addr, unsigned long len,
  			    unsigned long vm_flags, struct page **pages)
  {
a62c34bd2   Andy Lutomirski   x86, mm: Improve ...
2969
  	struct vm_area_struct *vma = __install_special_mapping(
27f28b972   Chen Gang   mm/mmap.c: change...
2970
2971
  		mm, addr, len, vm_flags, (void *)pages,
  		&legacy_special_mapping_vmops);
3935ed6a3   Stefani Seibold   mm: Add new func ...
2972

14bd5b458   Duan Jiong   mm/mmap.c: replac...
2973
  	return PTR_ERR_OR_ZERO(vma);
fa5dc22f8   Roland McGrath   [PATCH] Add insta...
2974
  }
7906d00cd   Andrea Arcangeli   mmu-notifiers: ad...
2975
2976
  
  static DEFINE_MUTEX(mm_all_locks_mutex);
454ed842d   Peter Zijlstra   lockdep: annotate...
2977
  static void vm_lock_anon_vma(struct mm_struct *mm, struct anon_vma *anon_vma)
7906d00cd   Andrea Arcangeli   mmu-notifiers: ad...
2978
  {
bf181b9f9   Michel Lespinasse   mm anon rmap: rep...
2979
  	if (!test_bit(0, (unsigned long *) &anon_vma->root->rb_root.rb_node)) {
7906d00cd   Andrea Arcangeli   mmu-notifiers: ad...
2980
2981
2982
2983
  		/*
  		 * The LSB of head.next can't change from under us
  		 * because we hold the mm_all_locks_mutex.
  		 */
572043c90   Jiri Kosina   mm: mmap: annotat...
2984
  		down_write_nest_lock(&anon_vma->root->rwsem, &mm->mmap_sem);
7906d00cd   Andrea Arcangeli   mmu-notifiers: ad...
2985
2986
  		/*
  		 * We can safely modify head.next after taking the
5a505085f   Ingo Molnar   mm/rmap: Convert ...
2987
  		 * anon_vma->root->rwsem. If some other vma in this mm shares
7906d00cd   Andrea Arcangeli   mmu-notifiers: ad...
2988
2989
2990
2991
  		 * 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
5a505085f   Ingo Molnar   mm/rmap: Convert ...
2992
  		 * anon_vma->root->rwsem.
7906d00cd   Andrea Arcangeli   mmu-notifiers: ad...
2993
2994
  		 */
  		if (__test_and_set_bit(0, (unsigned long *)
bf181b9f9   Michel Lespinasse   mm anon rmap: rep...
2995
  				       &anon_vma->root->rb_root.rb_node))
7906d00cd   Andrea Arcangeli   mmu-notifiers: ad...
2996
2997
2998
  			BUG();
  	}
  }
454ed842d   Peter Zijlstra   lockdep: annotate...
2999
  static void vm_lock_mapping(struct mm_struct *mm, struct address_space *mapping)
7906d00cd   Andrea Arcangeli   mmu-notifiers: ad...
3000
3001
3002
3003
3004
3005
3006
3007
3008
3009
3010
3011
3012
  {
  	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();
c8c06efa8   Davidlohr Bueso   mm: convert i_mma...
3013
  		down_write_nest_lock(&mapping->i_mmap_rwsem, &mm->mmap_sem);
7906d00cd   Andrea Arcangeli   mmu-notifiers: ad...
3014
3015
3016
3017
3018
3019
3020
3021
3022
3023
3024
3025
3026
3027
  	}
  }
  
  /*
   * 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
27ba0644e   Kirill A. Shutemov   rmap: drop suppor...
3028
   * altering the vma layout. It's also needed in write mode to avoid new
7906d00cd   Andrea Arcangeli   mmu-notifiers: ad...
3029
3030
3031
3032
3033
   * 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.
   *
bf181b9f9   Michel Lespinasse   mm anon rmap: rep...
3034
   * The LSB in anon_vma->rb_root.rb_node and the AS_MM_ALL_LOCKS bitflag in
7906d00cd   Andrea Arcangeli   mmu-notifiers: ad...
3035
3036
3037
   * 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.
   *
88f306b68   Kirill A. Shutemov   mm: fix locking o...
3038
3039
3040
3041
3042
3043
3044
3045
3046
3047
   * We take locks in following order, accordingly to comment at beginning
   * of mm/rmap.c:
   *   - all hugetlbfs_i_mmap_rwsem_key locks (aka mapping->i_mmap_rwsem for
   *     hugetlb mapping);
   *   - all i_mmap_rwsem locks;
   *   - all anon_vma->rwseml
   *
   * We can take all locks within these types randomly because the VM code
   * doesn't nest them and we protected from parallel mm_take_all_locks() by
   * mm_all_locks_mutex.
7906d00cd   Andrea Arcangeli   mmu-notifiers: ad...
3048
3049
3050
3051
3052
3053
3054
3055
3056
   *
   * 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;
5beb49305   Rik van Riel   mm: change anon_v...
3057
  	struct anon_vma_chain *avc;
7906d00cd   Andrea Arcangeli   mmu-notifiers: ad...
3058
3059
3060
3061
3062
3063
3064
3065
  
  	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;
88f306b68   Kirill A. Shutemov   mm: fix locking o...
3066
3067
3068
3069
3070
3071
3072
3073
3074
3075
  		if (vma->vm_file && vma->vm_file->f_mapping &&
  				is_vm_hugetlb_page(vma))
  			vm_lock_mapping(mm, vma->vm_file->f_mapping);
  	}
  
  	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 &&
  				!is_vm_hugetlb_page(vma))
454ed842d   Peter Zijlstra   lockdep: annotate...
3076
  			vm_lock_mapping(mm, vma->vm_file->f_mapping);
7906d00cd   Andrea Arcangeli   mmu-notifiers: ad...
3077
  	}
7cd5a02f5   Peter Zijlstra   mm: fix mm_take_a...
3078
3079
3080
3081
3082
  
  	for (vma = mm->mmap; vma; vma = vma->vm_next) {
  		if (signal_pending(current))
  			goto out_unlock;
  		if (vma->anon_vma)
5beb49305   Rik van Riel   mm: change anon_v...
3083
3084
  			list_for_each_entry(avc, &vma->anon_vma_chain, same_vma)
  				vm_lock_anon_vma(mm, avc->anon_vma);
7906d00cd   Andrea Arcangeli   mmu-notifiers: ad...
3085
  	}
7cd5a02f5   Peter Zijlstra   mm: fix mm_take_a...
3086

584cff54e   Kautuk Consul   mm/mmap.c: elimin...
3087
  	return 0;
7906d00cd   Andrea Arcangeli   mmu-notifiers: ad...
3088
3089
  
  out_unlock:
584cff54e   Kautuk Consul   mm/mmap.c: elimin...
3090
3091
  	mm_drop_all_locks(mm);
  	return -EINTR;
7906d00cd   Andrea Arcangeli   mmu-notifiers: ad...
3092
3093
3094
3095
  }
  
  static void vm_unlock_anon_vma(struct anon_vma *anon_vma)
  {
bf181b9f9   Michel Lespinasse   mm anon rmap: rep...
3096
  	if (test_bit(0, (unsigned long *) &anon_vma->root->rb_root.rb_node)) {
7906d00cd   Andrea Arcangeli   mmu-notifiers: ad...
3097
3098
3099
3100
3101
  		/*
  		 * 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
bf181b9f9   Michel Lespinasse   mm anon rmap: rep...
3102
  		 * the vma so the users using the anon_vma->rb_root will
7906d00cd   Andrea Arcangeli   mmu-notifiers: ad...
3103
3104
3105
3106
  		 * never see our bitflag.
  		 *
  		 * No need of atomic instructions here, head.next
  		 * can't change from under us until we release the
5a505085f   Ingo Molnar   mm/rmap: Convert ...
3107
  		 * anon_vma->root->rwsem.
7906d00cd   Andrea Arcangeli   mmu-notifiers: ad...
3108
3109
  		 */
  		if (!__test_and_clear_bit(0, (unsigned long *)
bf181b9f9   Michel Lespinasse   mm anon rmap: rep...
3110
  					  &anon_vma->root->rb_root.rb_node))
7906d00cd   Andrea Arcangeli   mmu-notifiers: ad...
3111
  			BUG();
08b52706d   Konstantin Khlebnikov   mm/rmap: rename a...
3112
  		anon_vma_unlock_write(anon_vma);
7906d00cd   Andrea Arcangeli   mmu-notifiers: ad...
3113
3114
3115
3116
3117
3118
3119
3120
3121
3122
  	}
  }
  
  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.
  		 */
83cde9e8b   Davidlohr Bueso   mm: use new helpe...
3123
  		i_mmap_unlock_write(mapping);
7906d00cd   Andrea Arcangeli   mmu-notifiers: ad...
3124
3125
3126
3127
3128
3129
3130
3131
3132
3133
3134
3135
3136
  		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;
5beb49305   Rik van Riel   mm: change anon_v...
3137
  	struct anon_vma_chain *avc;
7906d00cd   Andrea Arcangeli   mmu-notifiers: ad...
3138
3139
3140
3141
3142
3143
  
  	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)
5beb49305   Rik van Riel   mm: change anon_v...
3144
3145
  			list_for_each_entry(avc, &vma->anon_vma_chain, same_vma)
  				vm_unlock_anon_vma(avc->anon_vma);
7906d00cd   Andrea Arcangeli   mmu-notifiers: ad...
3146
3147
3148
3149
3150
3151
  		if (vma->vm_file && vma->vm_file->f_mapping)
  			vm_unlock_mapping(vma->vm_file->f_mapping);
  	}
  
  	mutex_unlock(&mm_all_locks_mutex);
  }
8feae1311   David Howells   NOMMU: Make VMAs ...
3152
3153
3154
3155
3156
3157
  
  /*
   * initialise the VMA slab
   */
  void __init mmap_init(void)
  {
00a62ce91   KOSAKI Motohiro   mm: fix Committed...
3158
  	int ret;
908c7f194   Tejun Heo   percpu_counter: a...
3159
  	ret = percpu_counter_init(&vm_committed_as, 0, GFP_KERNEL);
00a62ce91   KOSAKI Motohiro   mm: fix Committed...
3160
  	VM_BUG_ON(ret);
8feae1311   David Howells   NOMMU: Make VMAs ...
3161
  }
c9b1d0981   Andrew Shewmaker   mm: limit growth ...
3162
3163
3164
3165
3166
3167
3168
3169
3170
3171
3172
  
  /*
   * Initialise sysctl_user_reserve_kbytes.
   *
   * This is intended to prevent a user from starting a single memory hogging
   * process, such that they cannot recover (kill the hog) in OVERCOMMIT_NEVER
   * mode.
   *
   * The default value is min(3% of free memory, 128MB)
   * 128MB is enough to recover with sshd/login, bash, and top/kill.
   */
1640879af   Andrew Shewmaker   mm: reinititalise...
3173
  static int init_user_reserve(void)
c9b1d0981   Andrew Shewmaker   mm: limit growth ...
3174
3175
3176
3177
3178
3179
3180
3181
  {
  	unsigned long free_kbytes;
  
  	free_kbytes = global_page_state(NR_FREE_PAGES) << (PAGE_SHIFT - 10);
  
  	sysctl_user_reserve_kbytes = min(free_kbytes / 32, 1UL << 17);
  	return 0;
  }
a64fb3cd6   Paul Gortmaker   mm: audit/fix non...
3182
  subsys_initcall(init_user_reserve);
4eeab4f55   Andrew Shewmaker   mm: replace hardc...
3183
3184
3185
3186
3187
3188
3189
3190
3191
3192
3193
  
  /*
   * Initialise sysctl_admin_reserve_kbytes.
   *
   * The purpose of sysctl_admin_reserve_kbytes is to allow the sys admin
   * to log in and kill a memory hogging process.
   *
   * Systems with more than 256MB will reserve 8MB, enough to recover
   * with sshd, bash, and top in OVERCOMMIT_GUESS. Smaller systems will
   * only reserve 3% of free pages by default.
   */
1640879af   Andrew Shewmaker   mm: reinititalise...
3194
  static int init_admin_reserve(void)
4eeab4f55   Andrew Shewmaker   mm: replace hardc...
3195
3196
3197
3198
3199
3200
3201
3202
  {
  	unsigned long free_kbytes;
  
  	free_kbytes = global_page_state(NR_FREE_PAGES) << (PAGE_SHIFT - 10);
  
  	sysctl_admin_reserve_kbytes = min(free_kbytes / 32, 1UL << 13);
  	return 0;
  }
a64fb3cd6   Paul Gortmaker   mm: audit/fix non...
3203
  subsys_initcall(init_admin_reserve);
1640879af   Andrew Shewmaker   mm: reinititalise...
3204
3205
3206
3207
3208
3209
3210
3211
3212
3213
3214
3215
3216
3217
3218
3219
3220
3221
3222
3223
3224
3225
3226
3227
3228
3229
3230
3231
3232
3233
3234
3235
3236
3237
3238
3239
3240
3241
3242
3243
3244
3245
3246
3247
3248
3249
3250
3251
3252
3253
3254
3255
3256
3257
3258
3259
3260
3261
3262
3263
3264
3265
3266
3267
3268
3269
3270
  
  /*
   * Reinititalise user and admin reserves if memory is added or removed.
   *
   * The default user reserve max is 128MB, and the default max for the
   * admin reserve is 8MB. These are usually, but not always, enough to
   * enable recovery from a memory hogging process using login/sshd, a shell,
   * and tools like top. It may make sense to increase or even disable the
   * reserve depending on the existence of swap or variations in the recovery
   * tools. So, the admin may have changed them.
   *
   * If memory is added and the reserves have been eliminated or increased above
   * the default max, then we'll trust the admin.
   *
   * If memory is removed and there isn't enough free memory, then we
   * need to reset the reserves.
   *
   * Otherwise keep the reserve set by the admin.
   */
  static int reserve_mem_notifier(struct notifier_block *nb,
  			     unsigned long action, void *data)
  {
  	unsigned long tmp, free_kbytes;
  
  	switch (action) {
  	case MEM_ONLINE:
  		/* Default max is 128MB. Leave alone if modified by operator. */
  		tmp = sysctl_user_reserve_kbytes;
  		if (0 < tmp && tmp < (1UL << 17))
  			init_user_reserve();
  
  		/* Default max is 8MB.  Leave alone if modified by operator. */
  		tmp = sysctl_admin_reserve_kbytes;
  		if (0 < tmp && tmp < (1UL << 13))
  			init_admin_reserve();
  
  		break;
  	case MEM_OFFLINE:
  		free_kbytes = global_page_state(NR_FREE_PAGES) << (PAGE_SHIFT - 10);
  
  		if (sysctl_user_reserve_kbytes > free_kbytes) {
  			init_user_reserve();
  			pr_info("vm.user_reserve_kbytes reset to %lu
  ",
  				sysctl_user_reserve_kbytes);
  		}
  
  		if (sysctl_admin_reserve_kbytes > free_kbytes) {
  			init_admin_reserve();
  			pr_info("vm.admin_reserve_kbytes reset to %lu
  ",
  				sysctl_admin_reserve_kbytes);
  		}
  		break;
  	default:
  		break;
  	}
  	return NOTIFY_OK;
  }
  
  static struct notifier_block reserve_mem_nb = {
  	.notifier_call = reserve_mem_notifier,
  };
  
  static int __meminit init_reserve_notifier(void)
  {
  	if (register_hotmemory_notifier(&reserve_mem_nb))
b1de0d139   Mitchel Humpherys   mm: convert some ...
3271
3272
  		pr_err("Failed registering memory add/remove notifier for admin reserve
  ");
1640879af   Andrew Shewmaker   mm: reinititalise...
3273
3274
3275
  
  	return 0;
  }
a64fb3cd6   Paul Gortmaker   mm: audit/fix non...
3276
  subsys_initcall(init_reserve_notifier);