Blame view
mm/mmap.c
98.2 KB
1da177e4c
|
1 2 3 4 5 |
/* * mm/mmap.c * * Written by obz. * |
046c68842
|
6 |
* Address space accounting code <alan@lxorguk.ukuu.org.uk> |
1da177e4c
|
7 |
*/ |
b1de0d139
|
8 |
#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt |
e8420a8ec
|
9 |
#include <linux/kernel.h> |
1da177e4c
|
10 |
#include <linux/slab.h> |
4af3c9cc4
|
11 |
#include <linux/backing-dev.h> |
1da177e4c
|
12 |
#include <linux/mm.h> |
615d6e875
|
13 |
#include <linux/vmacache.h> |
1da177e4c
|
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
|
19 |
#include <linux/capability.h> |
1da177e4c
|
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
|
26 |
#include <linux/shmem_fs.h> |
1da177e4c
|
27 |
#include <linux/profile.h> |
b95f1b31b
|
28 |
#include <linux/export.h> |
1da177e4c
|
29 30 31 |
#include <linux/mount.h> #include <linux/mempolicy.h> #include <linux/rmap.h> |
cddb8a5c1
|
32 |
#include <linux/mmu_notifier.h> |
82f71ae4a
|
33 |
#include <linux/mmdebug.h> |
cdd6c482c
|
34 |
#include <linux/perf_event.h> |
120a795da
|
35 |
#include <linux/audit.h> |
b15d00b6a
|
36 |
#include <linux/khugepaged.h> |
2b1444983
|
37 |
#include <linux/uprobes.h> |
d37371870
|
38 |
#include <linux/rbtree_augmented.h> |
1640879af
|
39 40 |
#include <linux/notifier.h> #include <linux/memory.h> |
b1de0d139
|
41 |
#include <linux/printk.h> |
19a809afe
|
42 |
#include <linux/userfaultfd_k.h> |
d977d56ce
|
43 |
#include <linux/moduleparam.h> |
62b5f7d01
|
44 |
#include <linux/pkeys.h> |
212925802
|
45 |
#include <linux/oom.h> |
1da177e4c
|
46 |
|
7c0f6ba68
|
47 |
#include <linux/uaccess.h> |
1da177e4c
|
48 49 |
#include <asm/cacheflush.h> #include <asm/tlb.h> |
d6dd61c83
|
50 |
#include <asm/mmu_context.h> |
1da177e4c
|
51 |
|
42b777281
|
52 |
#include "internal.h" |
3a4597568
|
53 54 55 |
#ifndef arch_mmap_check #define arch_mmap_check(addr, len, flags) (0) #endif |
d07e22597
|
56 57 58 59 60 61 62 63 64 65 |
#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
|
66 |
static bool ignore_rlimit_data; |
d977d56ce
|
67 |
core_param(ignore_rlimit_data, ignore_rlimit_data, bool, 0644); |
d07e22597
|
68 |
|
e0da382c9
|
69 70 71 |
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
|
72 73 74 75 76 77 78 79 80 |
/* 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
|
81 |
* |
1da177e4c
|
82 83 84 85 |
* 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 * |
cab15ce60
|
86 87 88 89 90 |
* On arm64, PROT_EXEC has the following behaviour for both MAP_SHARED and * MAP_PRIVATE: * r: (no) no * w: (no) no * x: (yes) yes |
1da177e4c
|
91 |
*/ |
ac34ceaf1
|
92 |
pgprot_t protection_map[16] __ro_after_init = { |
1da177e4c
|
93 94 95 |
__P000, __P001, __P010, __P011, __P100, __P101, __P110, __P111, __S000, __S001, __S010, __S011, __S100, __S101, __S110, __S111 }; |
804af2cf6
|
96 97 |
pgprot_t vm_get_page_prot(unsigned long vm_flags) { |
b845f313d
|
98 99 100 |
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
|
101 102 |
} EXPORT_SYMBOL(vm_get_page_prot); |
64e455079
|
103 104 105 106 107 108 109 110 111 |
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; |
6d2329f88
|
112 |
pgprot_t vm_page_prot; |
64e455079
|
113 |
|
6d2329f88
|
114 115 |
vm_page_prot = vm_pgprot_modify(vma->vm_page_prot, vm_flags); if (vma_wants_writenotify(vma, vm_page_prot)) { |
64e455079
|
116 |
vm_flags &= ~VM_SHARED; |
6d2329f88
|
117 |
vm_page_prot = vm_pgprot_modify(vm_page_prot, vm_flags); |
64e455079
|
118 |
} |
6d2329f88
|
119 120 |
/* remove_protection_ptes reads vma->vm_page_prot without mmap_sem */ WRITE_ONCE(vma->vm_page_prot, vm_page_prot); |
64e455079
|
121 |
} |
1da177e4c
|
122 |
/* |
c8c06efa8
|
123 |
* Requires inode->i_mapping->i_mmap_rwsem |
1da177e4c
|
124 125 126 127 128 |
*/ 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
|
129 |
atomic_inc(&file_inode(file)->i_writecount); |
1da177e4c
|
130 |
if (vma->vm_flags & VM_SHARED) |
4bb5f5d93
|
131 |
mapping_unmap_writable(mapping); |
1da177e4c
|
132 133 |
flush_dcache_mmap_lock(mapping); |
27ba0644e
|
134 |
vma_interval_tree_remove(vma, &mapping->i_mmap); |
1da177e4c
|
135 136 137 138 |
flush_dcache_mmap_unlock(mapping); } /* |
6b2dbba8b
|
139 |
* Unlink a file-based vm structure from its interval tree, to hide |
a8fb5618d
|
140 |
* vma from rmap and vmtruncate before freeing its page tables. |
1da177e4c
|
141 |
*/ |
a8fb5618d
|
142 |
void unlink_file_vma(struct vm_area_struct *vma) |
1da177e4c
|
143 144 |
{ struct file *file = vma->vm_file; |
1da177e4c
|
145 146 |
if (file) { struct address_space *mapping = file->f_mapping; |
83cde9e8b
|
147 |
i_mmap_lock_write(mapping); |
1da177e4c
|
148 |
__remove_shared_vm_struct(vma, file, mapping); |
83cde9e8b
|
149 |
i_mmap_unlock_write(mapping); |
1da177e4c
|
150 |
} |
a8fb5618d
|
151 152 153 154 155 156 157 158 |
} /* * 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
|
159 |
might_sleep(); |
1da177e4c
|
160 161 |
if (vma->vm_ops && vma->vm_ops->close) vma->vm_ops->close(vma); |
e9714acf8
|
162 |
if (vma->vm_file) |
a8fb5618d
|
163 |
fput(vma->vm_file); |
f0be3d32b
|
164 |
mpol_put(vma_policy(vma)); |
1da177e4c
|
165 |
kmem_cache_free(vm_area_cachep, vma); |
a8fb5618d
|
166 |
return next; |
1da177e4c
|
167 |
} |
81ebc9dec
|
168 169 |
static int do_brk_flags(unsigned long addr, unsigned long request, unsigned long flags, struct list_head *uf); |
6a6160a7b
|
170 |
SYSCALL_DEFINE1(brk, unsigned long, brk) |
1da177e4c
|
171 |
{ |
8764b338b
|
172 |
unsigned long retval; |
1da177e4c
|
173 174 |
unsigned long newbrk, oldbrk; struct mm_struct *mm = current->mm; |
1be7107fb
|
175 |
struct vm_area_struct *next; |
a5b4592cf
|
176 |
unsigned long min_brk; |
128557ffe
|
177 |
bool populate; |
897ab3e0c
|
178 |
LIST_HEAD(uf); |
1da177e4c
|
179 |
|
dc0ef0df7
|
180 181 |
if (down_write_killable(&mm->mmap_sem)) return -EINTR; |
1da177e4c
|
182 |
|
a5b4592cf
|
183 |
#ifdef CONFIG_COMPAT_BRK |
5520e8948
|
184 185 186 187 188 |
/* * 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
|
189 |
if (current->brk_randomized) |
5520e8948
|
190 191 192 |
min_brk = mm->start_brk; else min_brk = mm->end_data; |
a5b4592cf
|
193 194 195 196 |
#else min_brk = mm->start_brk; #endif if (brk < min_brk) |
1da177e4c
|
197 |
goto out; |
1e624196f
|
198 199 200 201 202 203 204 |
/* * 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
|
205 206 |
if (check_data_rlimit(rlimit(RLIMIT_DATA), brk, mm->start_brk, mm->end_data, mm->start_data)) |
1e624196f
|
207 |
goto out; |
1da177e4c
|
208 209 210 211 212 213 214 |
newbrk = PAGE_ALIGN(brk); oldbrk = PAGE_ALIGN(mm->brk); if (oldbrk == newbrk) goto set_brk; /* Always allow shrinking brk. */ if (brk <= mm->brk) { |
897ab3e0c
|
215 |
if (!do_munmap(mm, newbrk, oldbrk-newbrk, &uf)) |
1da177e4c
|
216 217 218 |
goto set_brk; goto out; } |
1da177e4c
|
219 |
/* Check against existing mmap mappings. */ |
1be7107fb
|
220 221 |
next = find_vma(mm, oldbrk); if (next && newbrk + PAGE_SIZE > vm_start_gap(next)) |
1da177e4c
|
222 223 224 |
goto out; /* Ok, looks good - let it rip. */ |
81ebc9dec
|
225 |
if (do_brk_flags(oldbrk, newbrk-oldbrk, 0, &uf) < 0) |
1da177e4c
|
226 |
goto out; |
128557ffe
|
227 |
|
1da177e4c
|
228 229 |
set_brk: mm->brk = brk; |
128557ffe
|
230 231 |
populate = newbrk > oldbrk && (mm->def_flags & VM_LOCKED) != 0; up_write(&mm->mmap_sem); |
897ab3e0c
|
232 |
userfaultfd_unmap_complete(mm, &uf); |
128557ffe
|
233 234 235 |
if (populate) mm_populate(oldbrk, newbrk - oldbrk); return brk; |
1da177e4c
|
236 237 238 239 240 |
out: retval = mm->brk; up_write(&mm->mmap_sem); return retval; } |
d37371870
|
241 242 |
static long vma_compute_subtree_gap(struct vm_area_struct *vma) { |
1be7107fb
|
243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 |
unsigned long max, prev_end, subtree_gap; /* * Note: in the rare case of a VM_GROWSDOWN above a VM_GROWSUP, we * allow two stack_guard_gaps between them here, and when choosing * an unmapped area; whereas when expanding we only require one. * That's a little inconsistent, but keeps the code here simpler. */ max = vm_start_gap(vma); if (vma->vm_prev) { prev_end = vm_end_gap(vma->vm_prev); if (max > prev_end) max -= prev_end; else max = 0; } |
d37371870
|
259 260 261 262 263 264 265 266 267 268 269 270 271 272 |
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
|
273 |
#ifdef CONFIG_DEBUG_VM_RB |
acf128d04
|
274 |
static int browse_rb(struct mm_struct *mm) |
1da177e4c
|
275 |
{ |
acf128d04
|
276 |
struct rb_root *root = &mm->mm_rb; |
5a0768f64
|
277 |
int i = 0, j, bug = 0; |
1da177e4c
|
278 279 280 281 282 283 |
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
|
284 |
if (vma->vm_start < prev) { |
ff26f70f4
|
285 286 287 |
pr_emerg("vm_start %lx < prev %lx ", vma->vm_start, prev); |
5a0768f64
|
288 289 290 |
bug = 1; } if (vma->vm_start < pend) { |
ff26f70f4
|
291 292 293 |
pr_emerg("vm_start %lx < pend %lx ", vma->vm_start, pend); |
5a0768f64
|
294 295 296 |
bug = 1; } if (vma->vm_start > vma->vm_end) { |
ff26f70f4
|
297 298 299 |
pr_emerg("vm_start %lx > vm_end %lx ", vma->vm_start, vma->vm_end); |
5a0768f64
|
300 301 |
bug = 1; } |
acf128d04
|
302 |
spin_lock(&mm->page_table_lock); |
5a0768f64
|
303 |
if (vma->rb_subtree_gap != vma_compute_subtree_gap(vma)) { |
8542bdfc6
|
304 305 |
pr_emerg("free gap %lx, correct %lx ", |
5a0768f64
|
306 307 308 309 |
vma->rb_subtree_gap, vma_compute_subtree_gap(vma)); bug = 1; } |
acf128d04
|
310 |
spin_unlock(&mm->page_table_lock); |
1da177e4c
|
311 312 |
i++; pn = nd; |
d1af65d13
|
313 314 |
prev = vma->vm_start; pend = vma->vm_end; |
1da177e4c
|
315 316 |
} j = 0; |
5a0768f64
|
317 |
for (nd = pn; nd; nd = rb_prev(nd)) |
1da177e4c
|
318 |
j++; |
5a0768f64
|
319 |
if (i != j) { |
8542bdfc6
|
320 321 |
pr_emerg("backwards %d, forwards %d ", j, i); |
5a0768f64
|
322 |
bug = 1; |
1da177e4c
|
323 |
} |
5a0768f64
|
324 |
return bug ? -1 : i; |
1da177e4c
|
325 |
} |
d37371870
|
326 327 328 329 330 331 332 |
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
|
333 334 335 |
VM_BUG_ON_VMA(vma != ignore && vma->rb_subtree_gap != vma_compute_subtree_gap(vma), vma); |
1da177e4c
|
336 |
} |
1da177e4c
|
337 |
} |
eafd4dc4d
|
338 |
static void validate_mm(struct mm_struct *mm) |
1da177e4c
|
339 340 341 |
{ int bug = 0; int i = 0; |
5a0768f64
|
342 |
unsigned long highest_address = 0; |
ed8ea8150
|
343 |
struct vm_area_struct *vma = mm->mmap; |
ff26f70f4
|
344 |
|
ed8ea8150
|
345 |
while (vma) { |
12352d3ca
|
346 |
struct anon_vma *anon_vma = vma->anon_vma; |
ed8ea8150
|
347 |
struct anon_vma_chain *avc; |
ff26f70f4
|
348 |
|
12352d3ca
|
349 350 351 352 353 354 |
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); } |
1be7107fb
|
355 |
highest_address = vm_end_gap(vma); |
ed8ea8150
|
356 |
vma = vma->vm_next; |
1da177e4c
|
357 358 |
i++; } |
5a0768f64
|
359 |
if (i != mm->map_count) { |
8542bdfc6
|
360 361 |
pr_emerg("map_count %d vm_next %d ", mm->map_count, i); |
5a0768f64
|
362 363 364 |
bug = 1; } if (highest_address != mm->highest_vm_end) { |
8542bdfc6
|
365 366 |
pr_emerg("mm->highest_vm_end %lx, found %lx ", |
ff26f70f4
|
367 |
mm->highest_vm_end, highest_address); |
5a0768f64
|
368 369 |
bug = 1; } |
acf128d04
|
370 |
i = browse_rb(mm); |
5a0768f64
|
371 |
if (i != mm->map_count) { |
ff26f70f4
|
372 373 374 |
if (i != -1) pr_emerg("map_count %d rb %d ", mm->map_count, i); |
5a0768f64
|
375 376 |
bug = 1; } |
96dad67ff
|
377 |
VM_BUG_ON_MM(bug, mm); |
1da177e4c
|
378 379 |
} #else |
d37371870
|
380 |
#define validate_mm_rb(root, ignore) do { } while (0) |
1da177e4c
|
381 382 |
#define validate_mm(mm) do { } while (0) #endif |
d37371870
|
383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 |
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); } |
8f26e0b17
|
408 |
static void __vma_rb_erase(struct vm_area_struct *vma, struct rb_root *root) |
d37371870
|
409 410 |
{ /* |
d37371870
|
411 412 413 414 415 416 |
* 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); } |
8f26e0b17
|
417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 |
static __always_inline void vma_rb_erase_ignore(struct vm_area_struct *vma, struct rb_root *root, struct vm_area_struct *ignore) { /* * All rb_subtree_gap values must be consistent prior to erase, * with the possible exception of the "next" vma being erased if * next->vm_start was reduced. */ validate_mm_rb(root, ignore); __vma_rb_erase(vma, root); } static __always_inline 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); __vma_rb_erase(vma, root); } |
bf181b9f9
|
442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 |
/* * 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
|
473 474 475 |
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
|
476 |
{ |
6597d7833
|
477 |
struct rb_node **__rb_link, *__rb_parent, *rb_prev; |
1da177e4c
|
478 479 480 |
__rb_link = &mm->mm_rb.rb_node; rb_prev = __rb_parent = NULL; |
1da177e4c
|
481 482 483 484 485 486 487 488 |
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
|
489 490 491 |
/* Fail if an existing vma overlaps the area */ if (vma_tmp->vm_start < end) return -ENOMEM; |
1da177e4c
|
492 493 494 495 496 497 498 499 500 501 502 503 |
__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
|
504 |
return 0; |
1da177e4c
|
505 |
} |
e8420a8ec
|
506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 |
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
|
533 534 535 |
void __vma_link_rb(struct mm_struct *mm, struct vm_area_struct *vma, struct rb_node **rb_link, struct rb_node *rb_parent) { |
d37371870
|
536 537 538 539 |
/* Update tracking information for the gap following the new vma. */ if (vma->vm_next) vma_gap_update(vma->vm_next); else |
1be7107fb
|
540 |
mm->highest_vm_end = vm_end_gap(vma); |
d37371870
|
541 542 543 544 545 546 547 548 549 550 |
/* * 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
|
551 |
rb_link_node(&vma->vm_rb, rb_parent, rb_link); |
d37371870
|
552 553 554 |
vma->rb_subtree_gap = 0; vma_gap_update(vma); vma_rb_insert(vma, &mm->mm_rb); |
1da177e4c
|
555 |
} |
cb8f488c3
|
556 |
static void __vma_link_file(struct vm_area_struct *vma) |
1da177e4c
|
557 |
{ |
48aae4255
|
558 |
struct file *file; |
1da177e4c
|
559 560 561 562 563 564 |
file = vma->vm_file; if (file) { struct address_space *mapping = file->f_mapping; if (vma->vm_flags & VM_DENYWRITE) |
496ad9aa8
|
565 |
atomic_dec(&file_inode(file)->i_writecount); |
1da177e4c
|
566 |
if (vma->vm_flags & VM_SHARED) |
4bb5f5d93
|
567 |
atomic_inc(&mapping->i_mmap_writable); |
1da177e4c
|
568 569 |
flush_dcache_mmap_lock(mapping); |
27ba0644e
|
570 |
vma_interval_tree_insert(vma, &mapping->i_mmap); |
1da177e4c
|
571 572 573 574 575 576 577 578 579 580 581 |
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
|
582 583 584 585 586 587 588 |
} 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
|
589 |
if (vma->vm_file) { |
1da177e4c
|
590 |
mapping = vma->vm_file->f_mapping; |
83cde9e8b
|
591 |
i_mmap_lock_write(mapping); |
64ac4940d
|
592 |
} |
1da177e4c
|
593 594 595 |
__vma_link(mm, vma, prev, rb_link, rb_parent); __vma_link_file(vma); |
1da177e4c
|
596 |
if (mapping) |
83cde9e8b
|
597 |
i_mmap_unlock_write(mapping); |
1da177e4c
|
598 599 600 601 602 603 |
mm->map_count++; validate_mm(mm); } /* |
88f6b4c32
|
604 |
* Helper for vma_adjust() in the split_vma insert case: insert a vma into the |
6b2dbba8b
|
605 |
* mm's list and rbtree. It has already been inserted into the interval tree. |
1da177e4c
|
606 |
*/ |
48aae4255
|
607 |
static void __insert_vm_struct(struct mm_struct *mm, struct vm_area_struct *vma) |
1da177e4c
|
608 |
{ |
6597d7833
|
609 |
struct vm_area_struct *prev; |
48aae4255
|
610 |
struct rb_node **rb_link, *rb_parent; |
1da177e4c
|
611 |
|
6597d7833
|
612 613 614 |
if (find_vma_links(mm, vma->vm_start, vma->vm_end, &prev, &rb_link, &rb_parent)) BUG(); |
1da177e4c
|
615 616 617 |
__vma_link(mm, vma, prev, rb_link, rb_parent); mm->map_count++; } |
e86f15ee6
|
618 619 620 |
static __always_inline void __vma_unlink_common(struct mm_struct *mm, struct vm_area_struct *vma, struct vm_area_struct *prev, |
8f26e0b17
|
621 622 |
bool has_prev, struct vm_area_struct *ignore) |
1da177e4c
|
623 |
{ |
d37371870
|
624 |
struct vm_area_struct *next; |
297c5eee3
|
625 |
|
8f26e0b17
|
626 |
vma_rb_erase_ignore(vma, &mm->mm_rb, ignore); |
e86f15ee6
|
627 628 629 630 631 632 633 634 635 636 |
next = vma->vm_next; if (has_prev) prev->vm_next = next; else { prev = vma->vm_prev; if (prev) prev->vm_next = next; else mm->mmap = next; } |
297c5eee3
|
637 638 |
if (next) next->vm_prev = prev; |
615d6e875
|
639 640 641 |
/* Kill the cache */ vmacache_invalidate(mm); |
1da177e4c
|
642 |
} |
e86f15ee6
|
643 644 645 646 |
static inline void __vma_unlink_prev(struct mm_struct *mm, struct vm_area_struct *vma, struct vm_area_struct *prev) { |
8f26e0b17
|
647 |
__vma_unlink_common(mm, vma, prev, true, vma); |
e86f15ee6
|
648 |
} |
1da177e4c
|
649 650 651 652 653 654 655 |
/* * 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. */ |
e86f15ee6
|
656 657 658 |
int __vma_adjust(struct vm_area_struct *vma, unsigned long start, unsigned long end, pgoff_t pgoff, struct vm_area_struct *insert, struct vm_area_struct *expand) |
1da177e4c
|
659 660 |
{ struct mm_struct *mm = vma->vm_mm; |
e86f15ee6
|
661 |
struct vm_area_struct *next = vma->vm_next, *orig_vma = vma; |
1da177e4c
|
662 |
struct address_space *mapping = NULL; |
f808c13fd
|
663 |
struct rb_root_cached *root = NULL; |
012f18004
|
664 |
struct anon_vma *anon_vma = NULL; |
1da177e4c
|
665 |
struct file *file = vma->vm_file; |
d37371870
|
666 |
bool start_changed = false, end_changed = false; |
1da177e4c
|
667 668 669 670 |
long adjust_next = 0; int remove_next = 0; if (next && !insert) { |
734537c9c
|
671 |
struct vm_area_struct *exporter = NULL, *importer = NULL; |
287d97ac0
|
672 |
|
1da177e4c
|
673 674 675 676 |
if (end >= next->vm_end) { /* * vma expands, overlapping all the next, and * perhaps the one after too (mprotect case 6). |
86d12e471
|
677 |
* The only other cases that gets here are |
e86f15ee6
|
678 |
* case 1, case 7 and case 8. |
1da177e4c
|
679 |
*/ |
e86f15ee6
|
680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 |
if (next == expand) { /* * The only case where we don't expand "vma" * and we expand "next" instead is case 8. */ VM_WARN_ON(end != next->vm_end); /* * remove_next == 3 means we're * removing "vma" and that to do so we * swapped "vma" and "next". */ remove_next = 3; VM_WARN_ON(file != next->vm_file); swap(vma, next); } else { VM_WARN_ON(expand != vma); /* * case 1, 6, 7, remove_next == 2 is case 6, * remove_next == 1 is case 1 or 7. */ remove_next = 1 + (end > next->vm_end); VM_WARN_ON(remove_next == 2 && end != next->vm_next->vm_end); VM_WARN_ON(remove_next == 1 && end != next->vm_end); /* trim end to next, for case 6 first pass */ end = next->vm_end; } |
287d97ac0
|
708 |
exporter = next; |
1da177e4c
|
709 |
importer = vma; |
734537c9c
|
710 711 712 713 714 |
/* * If next doesn't have anon_vma, import from vma after * next, if the vma overlaps with it. */ |
97a42cd43
|
715 |
if (remove_next == 2 && !next->anon_vma) |
734537c9c
|
716 |
exporter = next->vm_next; |
1da177e4c
|
717 718 719 720 721 722 |
} 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
|
723 |
exporter = next; |
1da177e4c
|
724 |
importer = vma; |
e86f15ee6
|
725 |
VM_WARN_ON(expand != importer); |
1da177e4c
|
726 727 728 729 730 731 |
} 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
|
732 |
adjust_next = -((vma->vm_end - end) >> PAGE_SHIFT); |
287d97ac0
|
733 |
exporter = vma; |
1da177e4c
|
734 |
importer = next; |
e86f15ee6
|
735 |
VM_WARN_ON(expand != importer); |
1da177e4c
|
736 |
} |
1da177e4c
|
737 |
|
5beb49305
|
738 739 740 741 742 |
/* * 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
|
743 |
if (exporter && exporter->anon_vma && !importer->anon_vma) { |
c4ea95d7c
|
744 |
int error; |
b800c91a0
|
745 |
importer->anon_vma = exporter->anon_vma; |
c4ea95d7c
|
746 |
error = anon_vma_clone(importer, exporter); |
3fe89b3e2
|
747 |
if (error) |
c4ea95d7c
|
748 |
return error; |
5beb49305
|
749 750 |
} } |
734537c9c
|
751 |
again: |
e86f15ee6
|
752 |
vma_adjust_trans_huge(orig_vma, start, end, adjust_next); |
37f9f5595
|
753 |
|
1da177e4c
|
754 755 |
if (file) { mapping = file->f_mapping; |
27ba0644e
|
756 757 |
root = &mapping->i_mmap; uprobe_munmap(vma, vma->vm_start, vma->vm_end); |
682968e0c
|
758 |
|
27ba0644e
|
759 760 |
if (adjust_next) uprobe_munmap(next, next->vm_start, next->vm_end); |
682968e0c
|
761 |
|
83cde9e8b
|
762 |
i_mmap_lock_write(mapping); |
1da177e4c
|
763 |
if (insert) { |
1da177e4c
|
764 |
/* |
6b2dbba8b
|
765 |
* Put into interval tree now, so instantiated pages |
1da177e4c
|
766 767 768 769 770 771 772 |
* 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
|
773 774 775 776 |
anon_vma = vma->anon_vma; if (!anon_vma && adjust_next) anon_vma = next->anon_vma; if (anon_vma) { |
e86f15ee6
|
777 778 |
VM_WARN_ON(adjust_next && next->anon_vma && anon_vma != next->anon_vma); |
4fc3f1d66
|
779 |
anon_vma_lock_write(anon_vma); |
bf181b9f9
|
780 781 782 783 |
anon_vma_interval_tree_pre_update_vma(vma); if (adjust_next) anon_vma_interval_tree_pre_update_vma(next); } |
012f18004
|
784 |
|
1da177e4c
|
785 786 |
if (root) { flush_dcache_mmap_lock(mapping); |
6b2dbba8b
|
787 |
vma_interval_tree_remove(vma, root); |
1da177e4c
|
788 |
if (adjust_next) |
6b2dbba8b
|
789 |
vma_interval_tree_remove(next, root); |
1da177e4c
|
790 |
} |
d37371870
|
791 792 793 794 795 796 797 798 |
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
|
799 800 801 802 803 804 805 806 |
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
|
807 808 |
vma_interval_tree_insert(next, root); vma_interval_tree_insert(vma, root); |
1da177e4c
|
809 810 811 812 813 814 815 816 |
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. */ |
e86f15ee6
|
817 818 819 |
if (remove_next != 3) __vma_unlink_prev(mm, next, vma); else |
8f26e0b17
|
820 821 822 823 824 825 826 827 828 829 |
/* * vma is not before next if they've been * swapped. * * pre-swap() next->vm_start was reduced so * tell validate_mm_rb to ignore pre-swap() * "next" (which is stored in post-swap() * "vma"). */ __vma_unlink_common(mm, next, NULL, false, vma); |
1da177e4c
|
830 831 |
if (file) __remove_shared_vm_struct(next, file, mapping); |
1da177e4c
|
832 833 834 835 836 837 838 |
} 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
|
839 840 841 842 843 |
} else { if (start_changed) vma_gap_update(vma); if (end_changed) { if (!next) |
1be7107fb
|
844 |
mm->highest_vm_end = vm_end_gap(vma); |
d37371870
|
845 846 847 |
else if (!adjust_next) vma_gap_update(next); } |
1da177e4c
|
848 |
} |
bf181b9f9
|
849 850 851 852 |
if (anon_vma) { anon_vma_interval_tree_post_update_vma(vma); if (adjust_next) anon_vma_interval_tree_post_update_vma(next); |
08b52706d
|
853 |
anon_vma_unlock_write(anon_vma); |
bf181b9f9
|
854 |
} |
1da177e4c
|
855 |
if (mapping) |
83cde9e8b
|
856 |
i_mmap_unlock_write(mapping); |
1da177e4c
|
857 |
|
2b1444983
|
858 |
if (root) { |
7b2d81d48
|
859 |
uprobe_mmap(vma); |
2b1444983
|
860 861 |
if (adjust_next) |
7b2d81d48
|
862 |
uprobe_mmap(next); |
2b1444983
|
863 |
} |
1da177e4c
|
864 |
if (remove_next) { |
925d1c401
|
865 |
if (file) { |
cbc91f71b
|
866 |
uprobe_munmap(next, next->vm_start, next->vm_end); |
1da177e4c
|
867 |
fput(file); |
925d1c401
|
868 |
} |
5beb49305
|
869 870 |
if (next->anon_vma) anon_vma_merge(vma, next); |
1da177e4c
|
871 |
mm->map_count--; |
3964acd0d
|
872 |
mpol_put(vma_policy(next)); |
1da177e4c
|
873 874 875 876 877 878 |
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. */ |
e86f15ee6
|
879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 |
if (remove_next != 3) { /* * If "next" was removed and vma->vm_end was * expanded (up) over it, in turn * "next->vm_prev->vm_end" changed and the * "vma->vm_next" gap must be updated. */ next = vma->vm_next; } else { /* * For the scope of the comment "next" and * "vma" considered pre-swap(): if "vma" was * removed, next->vm_start was expanded (down) * over it and the "next" gap must be updated. * Because of the swap() the post-swap() "vma" * actually points to pre-swap() "next" * (post-swap() "next" as opposed is now a * dangling pointer). */ next = vma; } |
734537c9c
|
900 901 902 |
if (remove_next == 2) { remove_next = 1; end = next->vm_end; |
1da177e4c
|
903 |
goto again; |
734537c9c
|
904 |
} |
d37371870
|
905 906 |
else if (next) vma_gap_update(next); |
fb8c41e9a
|
907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 |
else { /* * If remove_next == 2 we obviously can't * reach this path. * * If remove_next == 3 we can't reach this * path because pre-swap() next is always not * NULL. pre-swap() "next" is not being * removed and its next->vm_end is not altered * (and furthermore "end" already matches * next->vm_end in remove_next == 3). * * We reach this only in the remove_next == 1 * case if the "next" vma that was removed was * the highest vma of the mm. However in such * case next->vm_end == "end" and the extended * "vma" has vma->vm_end == next->vm_end so * mm->highest_vm_end doesn't need any update * in remove_next == 1 case. */ |
1be7107fb
|
927 |
VM_WARN_ON(mm->highest_vm_end != vm_end_gap(vma)); |
fb8c41e9a
|
928 |
} |
1da177e4c
|
929 |
} |
2b1444983
|
930 |
if (insert && file) |
7b2d81d48
|
931 |
uprobe_mmap(insert); |
1da177e4c
|
932 933 |
validate_mm(mm); |
5beb49305
|
934 935 |
return 0; |
1da177e4c
|
936 937 938 939 940 941 |
} /* * 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
|
942 |
static inline int is_mergeable_vma(struct vm_area_struct *vma, |
19a809afe
|
943 944 |
struct file *file, unsigned long vm_flags, struct vm_userfaultfd_ctx vm_userfaultfd_ctx) |
1da177e4c
|
945 |
{ |
34228d473
|
946 947 948 949 950 951 952 953 954 |
/* * 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
|
955 956 957 958 959 |
return 0; if (vma->vm_file != file) return 0; if (vma->vm_ops && vma->vm_ops->close) return 0; |
19a809afe
|
960 961 |
if (!is_mergeable_vm_userfaultfd_ctx(vma, vm_userfaultfd_ctx)) return 0; |
1da177e4c
|
962 963 964 965 |
return 1; } static inline int is_mergeable_anon_vma(struct anon_vma *anon_vma1, |
965f55dea
|
966 967 |
struct anon_vma *anon_vma2, struct vm_area_struct *vma) |
1da177e4c
|
968 |
{ |
965f55dea
|
969 970 971 972 973 974 975 976 |
/* * 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
|
977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 |
} /* * 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
|
992 993 994 |
struct anon_vma *anon_vma, struct file *file, pgoff_t vm_pgoff, struct vm_userfaultfd_ctx vm_userfaultfd_ctx) |
1da177e4c
|
995 |
{ |
19a809afe
|
996 |
if (is_mergeable_vma(vma, file, vm_flags, vm_userfaultfd_ctx) && |
965f55dea
|
997 |
is_mergeable_anon_vma(anon_vma, vma->anon_vma, vma)) { |
1da177e4c
|
998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 |
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
|
1013 1014 1015 |
struct anon_vma *anon_vma, struct file *file, pgoff_t vm_pgoff, struct vm_userfaultfd_ctx vm_userfaultfd_ctx) |
1da177e4c
|
1016 |
{ |
19a809afe
|
1017 |
if (is_mergeable_vma(vma, file, vm_flags, vm_userfaultfd_ctx) && |
965f55dea
|
1018 |
is_mergeable_anon_vma(anon_vma, vma->anon_vma, vma)) { |
1da177e4c
|
1019 |
pgoff_t vm_pglen; |
d6e932177
|
1020 |
vm_pglen = vma_pages(vma); |
1da177e4c
|
1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 |
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 |
e86f15ee6
|
1048 |
* mremap move: PPPPXXXXXXXX 8 |
1da177e4c
|
1049 1050 1051 1052 |
* AAAA * PPPP NNNN PPPPPPPPPPPP PPPPPPPPNNNN PPPPNNNNNNNN * might become case 1 below case 2 below case 3 below * |
e86f15ee6
|
1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 |
* It is important for case 8 that the the vma NNNN overlapping the * region AAAA is never going to extended over XXXX. Instead XXXX must * be extended in region AAAA and NNNN must be removed. This way in * all cases where vma_merge succeeds, the moment vma_adjust drops the * rmap_locks, the properties of the merged vma will be already * correct for the whole merged range. Some of those properties like * vm_page_prot/vm_flags may be accessed by rmap_walks and they must * be correct for the whole merged range immediately after the * rmap_locks are released. Otherwise if XXXX would be removed and * NNNN would be extended over the XXXX range, remove_migration_ptes * or other rmap walkers (if working on addresses beyond the "end" * parameter) may establish ptes with the wrong permissions of NNNN * instead of the right permissions of XXXX. |
1da177e4c
|
1066 1067 1068 1069 |
*/ 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
|
1070 |
struct anon_vma *anon_vma, struct file *file, |
19a809afe
|
1071 1072 |
pgoff_t pgoff, struct mempolicy *policy, struct vm_userfaultfd_ctx vm_userfaultfd_ctx) |
1da177e4c
|
1073 1074 1075 |
{ pgoff_t pglen = (end - addr) >> PAGE_SHIFT; struct vm_area_struct *area, *next; |
5beb49305
|
1076 |
int err; |
1da177e4c
|
1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 |
/* * 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; |
e86f15ee6
|
1090 |
if (area && area->vm_end == end) /* cases 6, 7, 8 */ |
1da177e4c
|
1091 |
next = next->vm_next; |
e86f15ee6
|
1092 1093 1094 1095 |
/* verify some invariant that must be enforced by the caller */ VM_WARN_ON(prev && addr <= prev->vm_start); VM_WARN_ON(area && end > area->vm_end); VM_WARN_ON(addr >= end); |
1da177e4c
|
1096 1097 1098 1099 |
/* * Can it merge with the predecessor? */ if (prev && prev->vm_end == addr && |
cc71aba34
|
1100 |
mpol_equal(vma_policy(prev), policy) && |
1da177e4c
|
1101 |
can_vma_merge_after(prev, vm_flags, |
19a809afe
|
1102 1103 |
anon_vma, file, pgoff, vm_userfaultfd_ctx)) { |
1da177e4c
|
1104 1105 1106 1107 1108 1109 |
/* * 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
|
1110 1111 1112 |
anon_vma, file, pgoff+pglen, vm_userfaultfd_ctx) && |
1da177e4c
|
1113 |
is_mergeable_anon_vma(prev->anon_vma, |
965f55dea
|
1114 |
next->anon_vma, NULL)) { |
1da177e4c
|
1115 |
/* cases 1, 6 */ |
e86f15ee6
|
1116 1117 1118 |
err = __vma_adjust(prev, prev->vm_start, next->vm_end, prev->vm_pgoff, NULL, prev); |
1da177e4c
|
1119 |
} else /* cases 2, 5, 7 */ |
e86f15ee6
|
1120 1121 |
err = __vma_adjust(prev, prev->vm_start, end, prev->vm_pgoff, NULL, prev); |
5beb49305
|
1122 1123 |
if (err) return NULL; |
6d50e60cd
|
1124 |
khugepaged_enter_vma_merge(prev, vm_flags); |
1da177e4c
|
1125 1126 1127 1128 1129 1130 1131 |
return prev; } /* * Can this new request be merged in front of next? */ if (next && end == next->vm_start && |
cc71aba34
|
1132 |
mpol_equal(policy, vma_policy(next)) && |
1da177e4c
|
1133 |
can_vma_merge_before(next, vm_flags, |
19a809afe
|
1134 1135 |
anon_vma, file, pgoff+pglen, vm_userfaultfd_ctx)) { |
1da177e4c
|
1136 |
if (prev && addr < prev->vm_end) /* case 4 */ |
e86f15ee6
|
1137 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 |
err = __vma_adjust(prev, prev->vm_start, addr, prev->vm_pgoff, NULL, next); else { /* cases 3, 8 */ err = __vma_adjust(area, addr, next->vm_end, next->vm_pgoff - pglen, NULL, next); /* * In case 3 area is already equal to next and * this is a noop, but in case 8 "area" has * been removed and next was expanded over it. */ area = next; } |
5beb49305
|
1149 1150 |
if (err) return NULL; |
6d50e60cd
|
1151 |
khugepaged_enter_vma_merge(area, vm_flags); |
1da177e4c
|
1152 1153 1154 1155 1156 1157 1158 |
return area; } return NULL; } /* |
d0e9fe175
|
1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 |
* 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
|
1176 |
!((a->vm_flags ^ b->vm_flags) & ~(VM_READ|VM_WRITE|VM_EXEC|VM_SOFTDIRTY)) && |
d0e9fe175
|
1177 1178 1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 |
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
|
1191 |
* we do that READ_ONCE() to make sure that we never re-load the pointer. |
d0e9fe175
|
1192 1193 1194 1195 1196 1197 1198 1199 1200 1201 1202 1203 1204 |
* * 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
|
1205 |
struct anon_vma *anon_vma = READ_ONCE(old->anon_vma); |
d0e9fe175
|
1206 1207 1208 1209 1210 1211 1212 1213 |
if (anon_vma && list_is_singular(&old->anon_vma_chain)) return anon_vma; } return NULL; } /* |
1da177e4c
|
1214 1215 1216 1217 1218 1219 1220 1221 1222 |
* 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
|
1223 |
struct anon_vma *anon_vma; |
1da177e4c
|
1224 |
struct vm_area_struct *near; |
1da177e4c
|
1225 1226 1227 1228 |
near = vma->vm_next; if (!near) goto try_prev; |
d0e9fe175
|
1229 1230 1231 |
anon_vma = reusable_anon_vma(near, vma, near); if (anon_vma) return anon_vma; |
1da177e4c
|
1232 |
try_prev: |
9be34c9d5
|
1233 |
near = vma->vm_prev; |
1da177e4c
|
1234 1235 |
if (!near) goto none; |
d0e9fe175
|
1236 1237 1238 |
anon_vma = reusable_anon_vma(near, near, vma); if (anon_vma) return anon_vma; |
1da177e4c
|
1239 1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 |
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
|
1250 |
/* |
404015308
|
1251 1252 1253 1254 1255 1256 1257 1258 1259 1260 1261 |
* 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
|
1262 1263 1264 1265 1266 1267 1268 1269 1270 1271 1272 1273 1274 1275 1276 1277 1278 |
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; } |
16d7ceb04
|
1279 1280 1281 |
static inline u64 file_mmap_size_max(struct file *file, struct inode *inode) { if (S_ISREG(inode->i_mode)) |
af760b568
|
1282 |
return MAX_LFS_FILESIZE; |
16d7ceb04
|
1283 1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 1299 1300 1301 1302 1303 1304 1305 1306 |
if (S_ISBLK(inode->i_mode)) return MAX_LFS_FILESIZE; /* Special "we do even unsigned file positions" case */ if (file->f_mode & FMODE_UNSIGNED_OFFSET) return 0; /* Yes, random drivers might want more. But I'm tired of buggy drivers */ return ULONG_MAX; } static inline bool file_mmap_ok(struct file *file, struct inode *inode, unsigned long pgoff, unsigned long len) { u64 maxsize = file_mmap_size_max(file, inode); if (maxsize && len > maxsize) return false; maxsize -= len; if (pgoff > maxsize >> PAGE_SHIFT) return false; return true; } |
404015308
|
1307 |
/* |
27f5de796
|
1308 |
* The caller must hold down_write(¤t->mm->mmap_sem). |
1da177e4c
|
1309 |
*/ |
1fcfd8db7
|
1310 |
unsigned long do_mmap(struct file *file, unsigned long addr, |
1da177e4c
|
1311 |
unsigned long len, unsigned long prot, |
1fcfd8db7
|
1312 |
unsigned long flags, vm_flags_t vm_flags, |
897ab3e0c
|
1313 1314 |
unsigned long pgoff, unsigned long *populate, struct list_head *uf) |
1da177e4c
|
1315 |
{ |
cc71aba34
|
1316 |
struct mm_struct *mm = current->mm; |
62b5f7d01
|
1317 |
int pkey = 0; |
1da177e4c
|
1318 |
|
41badc15c
|
1319 |
*populate = 0; |
bebeb3d68
|
1320 |
|
e37609bb3
|
1321 1322 |
if (!len) return -EINVAL; |
1da177e4c
|
1323 1324 1325 1326 1327 1328 1329 |
/* * 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
|
1330 |
if (!(file && path_noexec(&file->f_path))) |
1da177e4c
|
1331 |
prot |= PROT_EXEC; |
7cd94146c
|
1332 1333 |
if (!(flags & MAP_FIXED)) addr = round_hint_to_min(addr); |
1da177e4c
|
1334 1335 |
/* Careful about overflows.. */ len = PAGE_ALIGN(len); |
9206de95b
|
1336 |
if (!len) |
1da177e4c
|
1337 1338 1339 1340 |
return -ENOMEM; /* offset overflow? */ if ((pgoff + (len >> PAGE_SHIFT)) < pgoff) |
cc71aba34
|
1341 |
return -EOVERFLOW; |
1da177e4c
|
1342 1343 1344 1345 1346 1347 1348 1349 1350 |
/* 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
|
1351 |
if (offset_in_page(addr)) |
1da177e4c
|
1352 |
return addr; |
62b5f7d01
|
1353 1354 1355 1356 1357 |
if (prot == PROT_EXEC) { pkey = execute_only_pkey(mm); if (pkey < 0) pkey = 0; } |
1da177e4c
|
1358 1359 1360 1361 |
/* 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
|
1362 |
vm_flags |= calc_vm_prot_bits(prot, pkey) | calc_vm_flag_bits(flags) | |
1da177e4c
|
1363 |
mm->def_flags | VM_MAYREAD | VM_MAYWRITE | VM_MAYEXEC; |
cdf7b3418
|
1364 |
if (flags & MAP_LOCKED) |
1da177e4c
|
1365 1366 |
if (!can_do_mlock()) return -EPERM; |
ba470de43
|
1367 |
|
363ee17f0
|
1368 1369 |
if (mlock_future_check(mm, vm_flags, len)) return -EAGAIN; |
1da177e4c
|
1370 |
|
1da177e4c
|
1371 |
if (file) { |
077bf22b5
|
1372 |
struct inode *inode = file_inode(file); |
16d7ceb04
|
1373 1374 |
if (!file_mmap_ok(file, inode, pgoff, len)) return -EOVERFLOW; |
1da177e4c
|
1375 1376 1377 1378 1379 1380 1381 1382 1383 1384 1385 1386 1387 1388 1389 |
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
|
1390 |
if (locks_verify_locked(file)) |
1da177e4c
|
1391 1392 1393 1394 1395 1396 1397 1398 1399 1400 |
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
|
1401 |
if (path_noexec(&file->f_path)) { |
80c5606c3
|
1402 1403 1404 1405 |
if (vm_flags & VM_EXEC) return -EPERM; vm_flags &= ~VM_MAYEXEC; } |
80c5606c3
|
1406 |
|
72c2d5319
|
1407 |
if (!file->f_op->mmap) |
80c5606c3
|
1408 |
return -ENODEV; |
b2c56e4f7
|
1409 1410 |
if (vm_flags & (VM_GROWSDOWN|VM_GROWSUP)) return -EINVAL; |
1da177e4c
|
1411 1412 1413 1414 1415 1416 1417 1418 |
break; default: return -EINVAL; } } else { switch (flags & MAP_TYPE) { case MAP_SHARED: |
b2c56e4f7
|
1419 1420 |
if (vm_flags & (VM_GROWSDOWN|VM_GROWSUP)) return -EINVAL; |
ce3639426
|
1421 1422 1423 1424 |
/* * Ignore pgoff. */ pgoff = 0; |
1da177e4c
|
1425 1426 1427 1428 1429 1430 1431 1432 1433 1434 1435 1436 |
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
|
1437 1438 1439 1440 1441 1442 1443 1444 1445 1446 1447 1448 1449 |
/* * 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; } |
897ab3e0c
|
1450 |
addr = mmap_region(file, addr, len, vm_flags, pgoff, uf); |
09a9f1d27
|
1451 1452 1453 |
if (!IS_ERR_VALUE(addr) && ((vm_flags & VM_LOCKED) || (flags & (MAP_POPULATE | MAP_NONBLOCK)) == MAP_POPULATE)) |
41badc15c
|
1454 |
*populate = len; |
bebeb3d68
|
1455 |
return addr; |
0165ab443
|
1456 |
} |
6be5ceb02
|
1457 |
|
66f0dc481
|
1458 1459 1460 1461 1462 |
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
|
1463 |
unsigned long retval; |
66f0dc481
|
1464 1465 |
if (!(flags & MAP_ANONYMOUS)) { |
120a795da
|
1466 |
audit_mmap_fd(fd, flags); |
66f0dc481
|
1467 1468 |
file = fget(fd); if (!file) |
1e3ee14b9
|
1469 |
return -EBADF; |
af73e4d95
|
1470 1471 |
if (is_file_hugepages(file)) len = ALIGN(len, huge_page_size(hstate_file(file))); |
493af5780
|
1472 1473 1474 |
retval = -EINVAL; if (unlikely(flags & MAP_HUGETLB && !is_file_hugepages(file))) goto out_fput; |
66f0dc481
|
1475 1476 |
} else if (flags & MAP_HUGETLB) { struct user_struct *user = NULL; |
c103a4dc4
|
1477 |
struct hstate *hs; |
af73e4d95
|
1478 |
|
20ac28933
|
1479 |
hs = hstate_sizelog((flags >> MAP_HUGE_SHIFT) & MAP_HUGE_MASK); |
091d0d55b
|
1480 1481 1482 1483 |
if (!hs) return -EINVAL; len = ALIGN(len, huge_page_size(hs)); |
66f0dc481
|
1484 1485 1486 1487 1488 1489 |
/* * 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
|
1490 |
file = hugetlb_file_setup(HUGETLB_ANON_FILE, len, |
42d7395fe
|
1491 1492 1493 |
VM_NORESERVE, &user, HUGETLB_ANONHUGE_INODE, (flags >> MAP_HUGE_SHIFT) & MAP_HUGE_MASK); |
66f0dc481
|
1494 1495 1496 1497 1498 |
if (IS_ERR(file)) return PTR_ERR(file); } flags &= ~(MAP_EXECUTABLE | MAP_DENYWRITE); |
9fbeb5ab5
|
1499 |
retval = vm_mmap_pgoff(file, addr, len, prot, flags, pgoff); |
493af5780
|
1500 |
out_fput: |
66f0dc481
|
1501 1502 |
if (file) fput(file); |
66f0dc481
|
1503 1504 |
return retval; } |
a4679373c
|
1505 1506 1507 1508 1509 1510 1511 1512 1513 1514 1515 1516 1517 1518 1519 1520 |
#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
|
1521 |
if (offset_in_page(a.offset)) |
a4679373c
|
1522 1523 1524 1525 1526 1527 |
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
|
1528 1529 1530 1531 1532 1533 |
/* * 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). */ |
6d2329f88
|
1534 |
int vma_wants_writenotify(struct vm_area_struct *vma, pgprot_t vm_page_prot) |
4e950f6f0
|
1535 |
{ |
ca16d140a
|
1536 |
vm_flags_t vm_flags = vma->vm_flags; |
8a04446ab
|
1537 |
const struct vm_operations_struct *vm_ops = vma->vm_ops; |
4e950f6f0
|
1538 1539 1540 1541 1542 1543 |
/* 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
|
1544 |
if (vm_ops && (vm_ops->page_mkwrite || vm_ops->pfn_mkwrite)) |
4e950f6f0
|
1545 |
return 1; |
64e455079
|
1546 1547 |
/* The open routine did something to the protections that pgprot_modify * won't preserve? */ |
6d2329f88
|
1548 1549 |
if (pgprot_val(vm_page_prot) != pgprot_val(vm_pgprot_modify(vm_page_prot, vm_flags))) |
4e950f6f0
|
1550 |
return 0; |
64e455079
|
1551 1552 1553 |
/* Do we need to track softdirty? */ if (IS_ENABLED(CONFIG_MEM_SOFT_DIRTY) && !(vm_flags & VM_SOFTDIRTY)) return 1; |
4e950f6f0
|
1554 |
/* Specialty mapping? */ |
4b6e1e370
|
1555 |
if (vm_flags & VM_PFNMAP) |
4e950f6f0
|
1556 1557 1558 1559 1560 1561 |
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
|
1562 1563 |
/* * We account for memory if it's a private writeable mapping, |
5a6fe1259
|
1564 |
* not hugepages and VM_NORESERVE wasn't set. |
fc8744adc
|
1565 |
*/ |
ca16d140a
|
1566 |
static inline int accountable_mapping(struct file *file, vm_flags_t vm_flags) |
fc8744adc
|
1567 |
{ |
5a6fe1259
|
1568 1569 1570 1571 1572 1573 |
/* * 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
|
1574 1575 |
return (vm_flags & (VM_NORESERVE | VM_SHARED | VM_WRITE)) == VM_WRITE; } |
0165ab443
|
1576 |
unsigned long mmap_region(struct file *file, unsigned long addr, |
897ab3e0c
|
1577 1578 |
unsigned long len, vm_flags_t vm_flags, unsigned long pgoff, struct list_head *uf) |
0165ab443
|
1579 1580 1581 |
{ struct mm_struct *mm = current->mm; struct vm_area_struct *vma, *prev; |
0165ab443
|
1582 1583 1584 |
int error; struct rb_node **rb_link, *rb_parent; unsigned long charged = 0; |
0165ab443
|
1585 |
|
e8420a8ec
|
1586 |
/* Check against address space limit. */ |
846383359
|
1587 |
if (!may_expand_vm(mm, vm_flags, len >> PAGE_SHIFT)) { |
e8420a8ec
|
1588 1589 1590 1591 1592 1593 |
unsigned long nr_pages; /* * MAP_FIXED may remove pages of mappings that intersects with * requested mapping. Account for the pages it would unmap. */ |
e8420a8ec
|
1594 |
nr_pages = count_vma_pages_range(mm, addr, addr + len); |
846383359
|
1595 1596 |
if (!may_expand_vm(mm, vm_flags, (len >> PAGE_SHIFT) - nr_pages)) |
e8420a8ec
|
1597 1598 |
return -ENOMEM; } |
1da177e4c
|
1599 |
/* Clear old maps */ |
9fcd14571
|
1600 1601 |
while (find_vma_links(mm, addr, addr + len, &prev, &rb_link, &rb_parent)) { |
897ab3e0c
|
1602 |
if (do_munmap(mm, addr, len, uf)) |
1da177e4c
|
1603 |
return -ENOMEM; |
1da177e4c
|
1604 |
} |
fc8744adc
|
1605 |
/* |
fc8744adc
|
1606 1607 |
* Private writable mapping: check memory availability */ |
5a6fe1259
|
1608 |
if (accountable_mapping(file, vm_flags)) { |
fc8744adc
|
1609 |
charged = len >> PAGE_SHIFT; |
191c54244
|
1610 |
if (security_vm_enough_memory_mm(mm, charged)) |
fc8744adc
|
1611 1612 |
return -ENOMEM; vm_flags |= VM_ACCOUNT; |
1da177e4c
|
1613 1614 1615 |
} /* |
de33c8db5
|
1616 |
* Can we just expand an old mapping? |
1da177e4c
|
1617 |
*/ |
19a809afe
|
1618 1619 |
vma = vma_merge(mm, prev, addr, addr + len, vm_flags, NULL, file, pgoff, NULL, NULL_VM_UFFD_CTX); |
de33c8db5
|
1620 1621 |
if (vma) goto out; |
1da177e4c
|
1622 1623 1624 1625 1626 1627 |
/* * 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
|
1628 |
vma = kmem_cache_zalloc(vm_area_cachep, GFP_KERNEL); |
1da177e4c
|
1629 1630 1631 1632 |
if (!vma) { error = -ENOMEM; goto unacct_error; } |
1da177e4c
|
1633 1634 1635 1636 1637 |
vma->vm_mm = mm; vma->vm_start = addr; vma->vm_end = addr + len; vma->vm_flags = vm_flags; |
3ed75eb8f
|
1638 |
vma->vm_page_prot = vm_get_page_prot(vm_flags); |
1da177e4c
|
1639 |
vma->vm_pgoff = pgoff; |
5beb49305
|
1640 |
INIT_LIST_HEAD(&vma->anon_vma_chain); |
1da177e4c
|
1641 1642 |
if (file) { |
1da177e4c
|
1643 1644 1645 1646 |
if (vm_flags & VM_DENYWRITE) { error = deny_write_access(file); if (error) goto free_vma; |
1da177e4c
|
1647 |
} |
4bb5f5d93
|
1648 1649 1650 1651 1652 1653 1654 1655 1656 1657 1658 |
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
|
1659 |
vma->vm_file = get_file(file); |
f74ac0152
|
1660 |
error = call_mmap(file, vma); |
1da177e4c
|
1661 1662 |
if (error) goto unmap_and_free_vma; |
f8dbf0a7a
|
1663 1664 1665 1666 1667 |
/* Can addr have changed?? * * Answer: Yes, several device drivers can do it in their * f_op->mmap method. -DaveM |
2897b4d29
|
1668 1669 |
* Bug: If addr is changed, prev, rb_link, rb_parent should * be updated for vma_link() |
f8dbf0a7a
|
1670 |
*/ |
2897b4d29
|
1671 |
WARN_ON_ONCE(addr != vma->vm_start); |
f8dbf0a7a
|
1672 |
addr = vma->vm_start; |
f8dbf0a7a
|
1673 |
vm_flags = vma->vm_flags; |
1da177e4c
|
1674 1675 1676 1677 1678 |
} else if (vm_flags & VM_SHARED) { error = shmem_zero_setup(vma); if (error) goto free_vma; } |
de33c8db5
|
1679 |
vma_link(mm, vma, prev, rb_link, rb_parent); |
4d3d5b41a
|
1680 |
/* Once vma denies write, undo our temporary denial count */ |
4bb5f5d93
|
1681 1682 1683 1684 1685 1686 |
if (file) { if (vm_flags & VM_SHARED) mapping_unmap_writable(file->f_mapping); if (vm_flags & VM_DENYWRITE) allow_write_access(file); } |
e86867720
|
1687 |
file = vma->vm_file; |
4d3d5b41a
|
1688 |
out: |
cdd6c482c
|
1689 |
perf_event_mmap(vma); |
0a4a93919
|
1690 |
|
846383359
|
1691 |
vm_stat_account(mm, vm_flags, len >> PAGE_SHIFT); |
1da177e4c
|
1692 |
if (vm_flags & VM_LOCKED) { |
bebeb3d68
|
1693 1694 |
if (!((vm_flags & VM_SPECIAL) || is_vm_hugetlb_page(vma) || vma == get_gate_vma(current->mm))) |
06f9d8c2b
|
1695 |
mm->locked_vm += (len >> PAGE_SHIFT); |
bebeb3d68
|
1696 |
else |
de60f5f10
|
1697 |
vma->vm_flags &= VM_LOCKED_CLEAR_MASK; |
bebeb3d68
|
1698 |
} |
2b1444983
|
1699 |
|
c7a3a88c9
|
1700 1701 |
if (file) uprobe_mmap(vma); |
2b1444983
|
1702 |
|
d9104d1ca
|
1703 1704 1705 1706 1707 1708 1709 1710 |
/* * 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
|
1711 |
vma_set_page_prot(vma); |
1da177e4c
|
1712 1713 1714 |
return addr; unmap_and_free_vma: |
1da177e4c
|
1715 1716 1717 1718 |
vma->vm_file = NULL; fput(file); /* Undo any partial mapping done by a device driver. */ |
e0da382c9
|
1719 1720 |
unmap_region(mm, vma, prev, vma->vm_start, vma->vm_end); charged = 0; |
4bb5f5d93
|
1721 1722 1723 1724 1725 |
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
|
1726 1727 1728 1729 1730 1731 1732 |
free_vma: kmem_cache_free(vm_area_cachep, vma); unacct_error: if (charged) vm_unacct_memory(charged); return error; } |
db4fbfb95
|
1733 1734 1735 1736 1737 1738 1739 1740 1741 1742 1743 1744 1745 1746 1747 1748 1749 1750 1751 1752 1753 1754 1755 1756 1757 1758 1759 1760 1761 1762 1763 1764 1765 1766 1767 1768 1769 |
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 */ |
1be7107fb
|
1770 |
gap_end = vm_start_gap(vma); |
db4fbfb95
|
1771 1772 1773 1774 1775 1776 1777 1778 1779 |
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; } } |
1be7107fb
|
1780 |
gap_start = vma->vm_prev ? vm_end_gap(vma->vm_prev) : 0; |
db4fbfb95
|
1781 1782 1783 1784 |
check_current: /* Check if current node has a suitable gap */ if (gap_start > high_limit) return -ENOMEM; |
f4cb767d7
|
1785 1786 |
if (gap_end >= low_limit && gap_end > gap_start && gap_end - gap_start >= length) |
db4fbfb95
|
1787 1788 1789 1790 1791 1792 1793 1794 1795 1796 1797 1798 1799 1800 1801 1802 1803 1804 1805 1806 1807 |
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) { |
1be7107fb
|
1808 1809 |
gap_start = vm_end_gap(vma->vm_prev); gap_end = vm_start_gap(vma); |
db4fbfb95
|
1810 1811 1812 1813 1814 1815 1816 1817 1818 1819 1820 1821 1822 1823 1824 1825 1826 1827 1828 1829 1830 1831 1832 1833 1834 1835 1836 1837 1838 1839 1840 1841 1842 1843 1844 1845 1846 1847 1848 1849 1850 1851 1852 1853 1854 1855 1856 1857 1858 1859 1860 1861 1862 1863 1864 1865 1866 1867 1868 1869 1870 1871 1872 |
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 */ |
1be7107fb
|
1873 |
gap_start = vma->vm_prev ? vm_end_gap(vma->vm_prev) : 0; |
db4fbfb95
|
1874 1875 1876 1877 1878 1879 1880 1881 1882 1883 1884 1885 |
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 */ |
1be7107fb
|
1886 |
gap_end = vm_start_gap(vma); |
db4fbfb95
|
1887 1888 |
if (gap_end < low_limit) return -ENOMEM; |
f4cb767d7
|
1889 1890 |
if (gap_start <= high_limit && gap_end > gap_start && gap_end - gap_start >= length) |
db4fbfb95
|
1891 1892 1893 1894 1895 1896 1897 1898 1899 1900 1901 1902 1903 1904 1905 1906 1907 1908 1909 1910 1911 1912 |
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 ? |
1be7107fb
|
1913 |
vm_end_gap(vma->vm_prev) : 0; |
db4fbfb95
|
1914 1915 1916 1917 1918 1919 1920 1921 1922 1923 1924 1925 1926 1927 1928 1929 1930 1931 1932 |
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
|
1933 1934 1935 1936 1937 1938 1939 1940 1941 1942 1943 1944 1945 1946 1947 1948 1949 |
/* 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; |
1be7107fb
|
1950 |
struct vm_area_struct *vma, *prev; |
db4fbfb95
|
1951 |
struct vm_unmapped_area_info info; |
1da177e4c
|
1952 |
|
2afc745f3
|
1953 |
if (len > TASK_SIZE - mmap_min_addr) |
1da177e4c
|
1954 |
return -ENOMEM; |
06abdfb47
|
1955 1956 |
if (flags & MAP_FIXED) return addr; |
1da177e4c
|
1957 1958 |
if (addr) { addr = PAGE_ALIGN(addr); |
1be7107fb
|
1959 |
vma = find_vma_prev(mm, addr, &prev); |
2afc745f3
|
1960 |
if (TASK_SIZE - len >= addr && addr >= mmap_min_addr && |
1be7107fb
|
1961 1962 |
(!vma || addr + len <= vm_start_gap(vma)) && (!prev || addr >= vm_end_gap(prev))) |
1da177e4c
|
1963 1964 |
return addr; } |
1da177e4c
|
1965 |
|
db4fbfb95
|
1966 1967 |
info.flags = 0; info.length = len; |
4e99b0213
|
1968 |
info.low_limit = mm->mmap_base; |
db4fbfb95
|
1969 1970 1971 |
info.high_limit = TASK_SIZE; info.align_mask = 0; return vm_unmapped_area(&info); |
1da177e4c
|
1972 |
} |
cc71aba34
|
1973 |
#endif |
1da177e4c
|
1974 |
|
1da177e4c
|
1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 |
/* * 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) { |
1be7107fb
|
1985 |
struct vm_area_struct *vma, *prev; |
1da177e4c
|
1986 |
struct mm_struct *mm = current->mm; |
db4fbfb95
|
1987 1988 |
unsigned long addr = addr0; struct vm_unmapped_area_info info; |
1da177e4c
|
1989 1990 |
/* requested length too big for entire address space */ |
2afc745f3
|
1991 |
if (len > TASK_SIZE - mmap_min_addr) |
1da177e4c
|
1992 |
return -ENOMEM; |
06abdfb47
|
1993 1994 |
if (flags & MAP_FIXED) return addr; |
1da177e4c
|
1995 1996 1997 |
/* requesting a specific address */ if (addr) { addr = PAGE_ALIGN(addr); |
1be7107fb
|
1998 |
vma = find_vma_prev(mm, addr, &prev); |
2afc745f3
|
1999 |
if (TASK_SIZE - len >= addr && addr >= mmap_min_addr && |
1be7107fb
|
2000 2001 |
(!vma || addr + len <= vm_start_gap(vma)) && (!prev || addr >= vm_end_gap(prev))) |
1da177e4c
|
2002 2003 |
return addr; } |
db4fbfb95
|
2004 2005 |
info.flags = VM_UNMAPPED_AREA_TOPDOWN; info.length = len; |
2afc745f3
|
2006 |
info.low_limit = max(PAGE_SIZE, mmap_min_addr); |
db4fbfb95
|
2007 2008 2009 |
info.high_limit = mm->mmap_base; info.align_mask = 0; addr = vm_unmapped_area(&info); |
b716ad953
|
2010 |
|
1da177e4c
|
2011 2012 2013 2014 2015 2016 |
/* * 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
|
2017 |
if (offset_in_page(addr)) { |
db4fbfb95
|
2018 2019 2020 2021 2022 2023 |
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
|
2024 2025 2026 2027 |
return addr; } #endif |
1da177e4c
|
2028 2029 2030 2031 |
unsigned long get_unmapped_area(struct file *file, unsigned long addr, unsigned long len, unsigned long pgoff, unsigned long flags) { |
06abdfb47
|
2032 2033 |
unsigned long (*get_area)(struct file *, unsigned long, unsigned long, unsigned long, unsigned long); |
9206de95b
|
2034 2035 2036 2037 2038 2039 2040 |
unsigned long error = arch_mmap_check(addr, len, flags); if (error) return error; /* Careful about overflows.. */ if (len > TASK_SIZE) return -ENOMEM; |
06abdfb47
|
2041 |
get_area = current->mm->get_unmapped_area; |
c01d5b300
|
2042 2043 2044 2045 2046 2047 2048 2049 2050 2051 2052 2053 |
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
|
2054 2055 2056 |
addr = get_area(file, addr, len, pgoff, flags); if (IS_ERR_VALUE(addr)) return addr; |
1da177e4c
|
2057 |
|
07ab67c8d
|
2058 2059 |
if (addr > TASK_SIZE - len) return -ENOMEM; |
de1741a13
|
2060 |
if (offset_in_page(addr)) |
07ab67c8d
|
2061 |
return -EINVAL; |
06abdfb47
|
2062 |
|
9ac4ed4bd
|
2063 2064 |
error = security_mmap_addr(addr); return error ? error : addr; |
1da177e4c
|
2065 2066 2067 2068 2069 |
} EXPORT_SYMBOL(get_unmapped_area); /* Look up the first VMA which satisfies addr < vm_end, NULL if none. */ |
48aae4255
|
2070 |
struct vm_area_struct *find_vma(struct mm_struct *mm, unsigned long addr) |
1da177e4c
|
2071 |
{ |
615d6e875
|
2072 2073 |
struct rb_node *rb_node; struct vm_area_struct *vma; |
1da177e4c
|
2074 |
|
841e31e5c
|
2075 |
/* Check the cache first. */ |
615d6e875
|
2076 2077 2078 |
vma = vmacache_find(mm, addr); if (likely(vma)) return vma; |
841e31e5c
|
2079 |
|
615d6e875
|
2080 |
rb_node = mm->mm_rb.rb_node; |
841e31e5c
|
2081 |
|
615d6e875
|
2082 2083 2084 2085 2086 2087 2088 2089 2090 2091 2092 2093 |
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
|
2094 |
} |
615d6e875
|
2095 2096 2097 |
if (vma) vmacache_update(addr, vma); |
1da177e4c
|
2098 2099 2100 2101 |
return vma; } EXPORT_SYMBOL(find_vma); |
6bd4837de
|
2102 2103 |
/* * Same as find_vma, but also return a pointer to the previous VMA in *pprev. |
6bd4837de
|
2104 |
*/ |
1da177e4c
|
2105 2106 2107 2108 |
struct vm_area_struct * find_vma_prev(struct mm_struct *mm, unsigned long addr, struct vm_area_struct **pprev) { |
6bd4837de
|
2109 |
struct vm_area_struct *vma; |
1da177e4c
|
2110 |
|
6bd4837de
|
2111 |
vma = find_vma(mm, addr); |
83cd904d2
|
2112 2113 2114 2115 2116 2117 2118 2119 2120 2121 |
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
|
2122 |
return vma; |
1da177e4c
|
2123 2124 2125 2126 2127 2128 2129 |
} /* * Verify that the stack growth is acceptable and * update accounting. This is shared with both the * grow-up and grow-down cases. */ |
1be7107fb
|
2130 2131 |
static int acct_stack_growth(struct vm_area_struct *vma, unsigned long size, unsigned long grow) |
1da177e4c
|
2132 2133 |
{ struct mm_struct *mm = vma->vm_mm; |
1be7107fb
|
2134 |
unsigned long new_start; |
1da177e4c
|
2135 2136 |
/* address space limit tests */ |
846383359
|
2137 |
if (!may_expand_vm(mm, vma->vm_flags, grow)) |
1da177e4c
|
2138 2139 2140 |
return -ENOMEM; /* Stack limit test */ |
24c79d8e0
|
2141 |
if (size > rlimit(RLIMIT_STACK)) |
1da177e4c
|
2142 2143 2144 2145 2146 2147 2148 |
return -ENOMEM; /* mlock limit tests */ if (vma->vm_flags & VM_LOCKED) { unsigned long locked; unsigned long limit; locked = mm->locked_vm + grow; |
24c79d8e0
|
2149 |
limit = rlimit(RLIMIT_MEMLOCK); |
59e99e5b9
|
2150 |
limit >>= PAGE_SHIFT; |
1da177e4c
|
2151 2152 2153 |
if (locked > limit && !capable(CAP_IPC_LOCK)) return -ENOMEM; } |
0d59a01bc
|
2154 2155 2156 2157 2158 |
/* 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
|
2159 2160 2161 2162 |
/* * Overcommit.. This must be the final test, as it will * update security statistics. */ |
05fa199d4
|
2163 |
if (security_vm_enough_memory_mm(mm, grow)) |
1da177e4c
|
2164 |
return -ENOMEM; |
1da177e4c
|
2165 2166 |
return 0; } |
46dea3d09
|
2167 |
#if defined(CONFIG_STACK_GROWSUP) || defined(CONFIG_IA64) |
1da177e4c
|
2168 |
/* |
46dea3d09
|
2169 2170 |
* 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
|
2171 |
*/ |
46dea3d09
|
2172 |
int expand_upwards(struct vm_area_struct *vma, unsigned long address) |
1da177e4c
|
2173 |
{ |
093578147
|
2174 |
struct mm_struct *mm = vma->vm_mm; |
1be7107fb
|
2175 2176 |
struct vm_area_struct *next; unsigned long gap_addr; |
12352d3ca
|
2177 |
int error = 0; |
1da177e4c
|
2178 2179 2180 |
if (!(vma->vm_flags & VM_GROWSUP)) return -EFAULT; |
bd726c90b
|
2181 |
/* Guard against exceeding limits of the address space. */ |
1be7107fb
|
2182 |
address &= PAGE_MASK; |
37511fb5c
|
2183 |
if (address >= (TASK_SIZE & PAGE_MASK)) |
12352d3ca
|
2184 |
return -ENOMEM; |
bd726c90b
|
2185 |
address += PAGE_SIZE; |
12352d3ca
|
2186 |
|
1be7107fb
|
2187 2188 |
/* Enforce stack_guard_gap */ gap_addr = address + stack_guard_gap; |
bd726c90b
|
2189 2190 2191 2192 |
/* Guard against overflow */ if (gap_addr < address || gap_addr > TASK_SIZE) gap_addr = TASK_SIZE; |
1be7107fb
|
2193 |
next = vma->vm_next; |
561b5e070
|
2194 2195 |
if (next && next->vm_start < gap_addr && (next->vm_flags & (VM_WRITE|VM_READ|VM_EXEC))) { |
1be7107fb
|
2196 2197 2198 2199 |
if (!(next->vm_flags & VM_GROWSUP)) return -ENOMEM; /* Check that both stack segments have the same anon_vma? */ } |
12352d3ca
|
2200 |
/* We must make sure the anon_vma is allocated. */ |
1da177e4c
|
2201 2202 |
if (unlikely(anon_vma_prepare(vma))) return -ENOMEM; |
1da177e4c
|
2203 2204 2205 2206 2207 2208 |
/* * 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
|
2209 |
anon_vma_lock_write(vma->anon_vma); |
1da177e4c
|
2210 2211 2212 2213 2214 2215 2216 |
/* 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
|
2217 2218 2219 2220 |
error = -ENOMEM; if (vma->vm_pgoff + (size >> PAGE_SHIFT) >= vma->vm_pgoff) { error = acct_stack_growth(vma, size, grow); if (!error) { |
4128997b5
|
2221 2222 2223 2224 2225 |
/* * 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
|
2226 |
* anon_vma_lock_write() doesn't help here, as |
4128997b5
|
2227 2228 2229 2230 2231 |
* 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
|
2232 |
spin_lock(&mm->page_table_lock); |
87e8827b3
|
2233 |
if (vma->vm_flags & VM_LOCKED) |
093578147
|
2234 |
mm->locked_vm += grow; |
846383359
|
2235 |
vm_stat_account(mm, vma->vm_flags, grow); |
bf181b9f9
|
2236 |
anon_vma_interval_tree_pre_update_vma(vma); |
42c36f63a
|
2237 |
vma->vm_end = address; |
bf181b9f9
|
2238 |
anon_vma_interval_tree_post_update_vma(vma); |
d37371870
|
2239 2240 2241 |
if (vma->vm_next) vma_gap_update(vma->vm_next); else |
1be7107fb
|
2242 |
mm->highest_vm_end = vm_end_gap(vma); |
093578147
|
2243 |
spin_unlock(&mm->page_table_lock); |
4128997b5
|
2244 |
|
42c36f63a
|
2245 2246 |
perf_event_mmap(vma); } |
3af9e8592
|
2247 |
} |
1da177e4c
|
2248 |
} |
12352d3ca
|
2249 |
anon_vma_unlock_write(vma->anon_vma); |
6d50e60cd
|
2250 |
khugepaged_enter_vma_merge(vma, vma->vm_flags); |
093578147
|
2251 |
validate_mm(mm); |
1da177e4c
|
2252 2253 |
return error; } |
46dea3d09
|
2254 |
#endif /* CONFIG_STACK_GROWSUP || CONFIG_IA64 */ |
1da177e4c
|
2255 2256 2257 |
/* * vma is the first one with address < vma->vm_start. Have to extend vma. */ |
d05f3169c
|
2258 |
int expand_downwards(struct vm_area_struct *vma, |
b6a2fea39
|
2259 |
unsigned long address) |
1da177e4c
|
2260 |
{ |
093578147
|
2261 |
struct mm_struct *mm = vma->vm_mm; |
1be7107fb
|
2262 |
struct vm_area_struct *prev; |
1da177e4c
|
2263 |
int error; |
8869477a4
|
2264 |
address &= PAGE_MASK; |
e5467859f
|
2265 |
error = security_mmap_addr(address); |
8869477a4
|
2266 2267 |
if (error) return error; |
1be7107fb
|
2268 |
/* Enforce stack_guard_gap */ |
1be7107fb
|
2269 |
prev = vma->vm_prev; |
32e4e6d5c
|
2270 2271 |
/* Check that both stack segments have the same anon_vma? */ if (prev && !(prev->vm_flags & VM_GROWSDOWN) && |
561b5e070
|
2272 |
(prev->vm_flags & (VM_WRITE|VM_READ|VM_EXEC))) { |
32e4e6d5c
|
2273 |
if (address - prev->vm_end < stack_guard_gap) |
1be7107fb
|
2274 |
return -ENOMEM; |
1be7107fb
|
2275 |
} |
12352d3ca
|
2276 2277 2278 |
/* We must make sure the anon_vma is allocated. */ if (unlikely(anon_vma_prepare(vma))) return -ENOMEM; |
1da177e4c
|
2279 2280 2281 2282 2283 2284 |
/* * 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
|
2285 |
anon_vma_lock_write(vma->anon_vma); |
1da177e4c
|
2286 2287 2288 2289 2290 2291 2292 |
/* 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
|
2293 2294 2295 2296 |
error = -ENOMEM; if (grow <= vma->vm_pgoff) { error = acct_stack_growth(vma, size, grow); if (!error) { |
4128997b5
|
2297 2298 2299 2300 2301 |
/* * 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
|
2302 |
* anon_vma_lock_write() doesn't help here, as |
4128997b5
|
2303 2304 2305 2306 2307 |
* 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
|
2308 |
spin_lock(&mm->page_table_lock); |
87e8827b3
|
2309 |
if (vma->vm_flags & VM_LOCKED) |
093578147
|
2310 |
mm->locked_vm += grow; |
846383359
|
2311 |
vm_stat_account(mm, vma->vm_flags, grow); |
bf181b9f9
|
2312 |
anon_vma_interval_tree_pre_update_vma(vma); |
a626ca6a6
|
2313 2314 |
vma->vm_start = address; vma->vm_pgoff -= grow; |
bf181b9f9
|
2315 |
anon_vma_interval_tree_post_update_vma(vma); |
d37371870
|
2316 |
vma_gap_update(vma); |
093578147
|
2317 |
spin_unlock(&mm->page_table_lock); |
4128997b5
|
2318 |
|
a626ca6a6
|
2319 2320 |
perf_event_mmap(vma); } |
1da177e4c
|
2321 2322 |
} } |
12352d3ca
|
2323 |
anon_vma_unlock_write(vma->anon_vma); |
6d50e60cd
|
2324 |
khugepaged_enter_vma_merge(vma, vma->vm_flags); |
093578147
|
2325 |
validate_mm(mm); |
1da177e4c
|
2326 2327 |
return error; } |
1be7107fb
|
2328 2329 2330 2331 2332 2333 2334 2335 2336 2337 2338 2339 2340 2341 2342 |
/* enforced gap between the expanding stack and other mappings. */ unsigned long stack_guard_gap = 256UL<<PAGE_SHIFT; static int __init cmdline_parse_stack_guard_gap(char *p) { unsigned long val; char *endptr; val = simple_strtoul(p, &endptr, 10); if (!*endptr) stack_guard_gap = val << PAGE_SHIFT; return 0; } __setup("stack_guard_gap=", cmdline_parse_stack_guard_gap); |
b6a2fea39
|
2343 2344 2345 2346 2347 2348 2349 2350 2351 2352 2353 2354 2355 2356 2357 |
#ifdef CONFIG_STACK_GROWSUP int expand_stack(struct vm_area_struct *vma, unsigned long address) { return expand_upwards(vma, address); } struct vm_area_struct * find_extend_vma(struct mm_struct *mm, unsigned long addr) { struct vm_area_struct *vma, *prev; addr &= PAGE_MASK; vma = find_vma_prev(mm, addr, &prev); if (vma && (vma->vm_start <= addr)) return vma; |
1c1271850
|
2358 |
if (!prev || expand_stack(prev, addr)) |
b6a2fea39
|
2359 |
return NULL; |
cea10a19b
|
2360 |
if (prev->vm_flags & VM_LOCKED) |
fc05f5662
|
2361 |
populate_vma_page_range(prev, addr, prev->vm_end, NULL); |
b6a2fea39
|
2362 2363 2364 2365 2366 2367 2368 |
return prev; } #else int expand_stack(struct vm_area_struct *vma, unsigned long address) { return expand_downwards(vma, address); } |
1da177e4c
|
2369 |
struct vm_area_struct * |
cc71aba34
|
2370 |
find_extend_vma(struct mm_struct *mm, unsigned long addr) |
1da177e4c
|
2371 |
{ |
cc71aba34
|
2372 |
struct vm_area_struct *vma; |
1da177e4c
|
2373 2374 2375 |
unsigned long start; addr &= PAGE_MASK; |
cc71aba34
|
2376 |
vma = find_vma(mm, addr); |
1da177e4c
|
2377 2378 2379 2380 2381 2382 2383 2384 2385 |
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
|
2386 |
if (vma->vm_flags & VM_LOCKED) |
fc05f5662
|
2387 |
populate_vma_page_range(vma, addr, start, NULL); |
1da177e4c
|
2388 2389 2390 |
return vma; } #endif |
e1d6d01ab
|
2391 |
EXPORT_SYMBOL_GPL(find_extend_vma); |
1da177e4c
|
2392 |
/* |
2c0b38146
|
2393 |
* Ok - we have the memory areas we should free on the vma list, |
1da177e4c
|
2394 |
* so release them, and do the vma updates. |
2c0b38146
|
2395 2396 |
* * Called with the mm semaphore held. |
1da177e4c
|
2397 |
*/ |
2c0b38146
|
2398 |
static void remove_vma_list(struct mm_struct *mm, struct vm_area_struct *vma) |
1da177e4c
|
2399 |
{ |
4f74d2c8e
|
2400 |
unsigned long nr_accounted = 0; |
365e9c87a
|
2401 2402 |
/* Update high watermark before we lower total_vm */ update_hiwater_vm(mm); |
1da177e4c
|
2403 |
do { |
2c0b38146
|
2404 |
long nrpages = vma_pages(vma); |
4f74d2c8e
|
2405 2406 |
if (vma->vm_flags & VM_ACCOUNT) nr_accounted += nrpages; |
846383359
|
2407 |
vm_stat_account(mm, vma->vm_flags, -nrpages); |
a8fb5618d
|
2408 |
vma = remove_vma(vma); |
146425a31
|
2409 |
} while (vma); |
4f74d2c8e
|
2410 |
vm_unacct_memory(nr_accounted); |
1da177e4c
|
2411 2412 2413 2414 2415 2416 |
validate_mm(mm); } /* * Get rid of page table information in the indicated region. * |
f10df6860
|
2417 |
* Called with the mm semaphore held. |
1da177e4c
|
2418 2419 |
*/ static void unmap_region(struct mm_struct *mm, |
e0da382c9
|
2420 2421 |
struct vm_area_struct *vma, struct vm_area_struct *prev, unsigned long start, unsigned long end) |
1da177e4c
|
2422 |
{ |
cc71aba34
|
2423 |
struct vm_area_struct *next = prev ? prev->vm_next : mm->mmap; |
d16dfc550
|
2424 |
struct mmu_gather tlb; |
1da177e4c
|
2425 2426 |
lru_add_drain(); |
2b047252d
|
2427 |
tlb_gather_mmu(&tlb, mm, start, end); |
365e9c87a
|
2428 |
update_hiwater_rss(mm); |
4f74d2c8e
|
2429 |
unmap_vmas(&tlb, vma, start, end); |
d16dfc550
|
2430 |
free_pgtables(&tlb, vma, prev ? prev->vm_end : FIRST_USER_ADDRESS, |
6ee8630e0
|
2431 |
next ? next->vm_start : USER_PGTABLES_CEILING); |
d16dfc550
|
2432 |
tlb_finish_mmu(&tlb, start, end); |
1da177e4c
|
2433 2434 2435 2436 2437 2438 2439 2440 2441 2442 2443 2444 2445 2446 |
} /* * 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
|
2447 |
vma->vm_prev = NULL; |
1da177e4c
|
2448 |
do { |
d37371870
|
2449 |
vma_rb_erase(vma, &mm->mm_rb); |
1da177e4c
|
2450 2451 2452 2453 2454 |
mm->map_count--; tail_vma = vma; vma = vma->vm_next; } while (vma && vma->vm_start < end); *insertion_point = vma; |
d37371870
|
2455 |
if (vma) { |
297c5eee3
|
2456 |
vma->vm_prev = prev; |
d37371870
|
2457 2458 |
vma_gap_update(vma); } else |
1be7107fb
|
2459 |
mm->highest_vm_end = prev ? vm_end_gap(prev) : 0; |
1da177e4c
|
2460 |
tail_vma->vm_next = NULL; |
615d6e875
|
2461 2462 2463 |
/* Kill the cache */ vmacache_invalidate(mm); |
1da177e4c
|
2464 2465 2466 |
} /* |
def5efe03
|
2467 2468 |
* __split_vma() bypasses sysctl_max_map_count checking. We use this where it * has already been checked or doesn't make sense to fail. |
1da177e4c
|
2469 |
*/ |
def5efe03
|
2470 2471 |
int __split_vma(struct mm_struct *mm, struct vm_area_struct *vma, unsigned long addr, int new_below) |
1da177e4c
|
2472 |
{ |
1da177e4c
|
2473 |
struct vm_area_struct *new; |
e39758912
|
2474 |
int err; |
1da177e4c
|
2475 |
|
c6c78a1d4
|
2476 2477 2478 2479 2480 |
if (vma->vm_ops && vma->vm_ops->split) { err = vma->vm_ops->split(vma, addr); if (err) return err; } |
1da177e4c
|
2481 |
|
e94b17660
|
2482 |
new = kmem_cache_alloc(vm_area_cachep, GFP_KERNEL); |
1da177e4c
|
2483 |
if (!new) |
e39758912
|
2484 |
return -ENOMEM; |
1da177e4c
|
2485 2486 2487 |
/* most fields are the same, copy all, and then fixup */ *new = *vma; |
5beb49305
|
2488 |
INIT_LIST_HEAD(&new->anon_vma_chain); |
1da177e4c
|
2489 2490 2491 2492 2493 2494 |
if (new_below) new->vm_end = addr; else { new->vm_start = addr; new->vm_pgoff += ((addr - vma->vm_start) >> PAGE_SHIFT); } |
ef0855d33
|
2495 2496 |
err = vma_dup_policy(vma, new); if (err) |
5beb49305
|
2497 |
goto out_free_vma; |
1da177e4c
|
2498 |
|
c4ea95d7c
|
2499 2500 |
err = anon_vma_clone(new, vma); if (err) |
5beb49305
|
2501 |
goto out_free_mpol; |
e9714acf8
|
2502 |
if (new->vm_file) |
1da177e4c
|
2503 2504 2505 2506 2507 2508 |
get_file(new->vm_file); if (new->vm_ops && new->vm_ops->open) new->vm_ops->open(new); if (new_below) |
5beb49305
|
2509 |
err = vma_adjust(vma, addr, vma->vm_end, vma->vm_pgoff + |
1da177e4c
|
2510 2511 |
((addr - new->vm_start) >> PAGE_SHIFT), new); else |
5beb49305
|
2512 |
err = vma_adjust(vma, vma->vm_start, addr, vma->vm_pgoff, new); |
1da177e4c
|
2513 |
|
5beb49305
|
2514 2515 2516 2517 2518 |
/* Success. */ if (!err) return 0; /* Clean everything up if vma_adjust failed. */ |
589275338
|
2519 2520 |
if (new->vm_ops && new->vm_ops->close) new->vm_ops->close(new); |
e9714acf8
|
2521 |
if (new->vm_file) |
5beb49305
|
2522 |
fput(new->vm_file); |
2aeadc30d
|
2523 |
unlink_anon_vmas(new); |
5beb49305
|
2524 |
out_free_mpol: |
ef0855d33
|
2525 |
mpol_put(vma_policy(new)); |
5beb49305
|
2526 2527 |
out_free_vma: kmem_cache_free(vm_area_cachep, new); |
5beb49305
|
2528 |
return err; |
1da177e4c
|
2529 |
} |
659ace584
|
2530 2531 2532 2533 2534 2535 2536 2537 2538 2539 2540 2541 |
/* * 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
|
2542 2543 2544 2545 2546 |
/* 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> */ |
897ab3e0c
|
2547 2548 |
int do_munmap(struct mm_struct *mm, unsigned long start, size_t len, struct list_head *uf) |
1da177e4c
|
2549 2550 |
{ unsigned long end; |
146425a31
|
2551 |
struct vm_area_struct *vma, *prev, *last; |
1da177e4c
|
2552 |
|
de1741a13
|
2553 |
if ((offset_in_page(start)) || start > TASK_SIZE || len > TASK_SIZE-start) |
1da177e4c
|
2554 |
return -EINVAL; |
cc71aba34
|
2555 2556 |
len = PAGE_ALIGN(len); if (len == 0) |
1da177e4c
|
2557 2558 2559 |
return -EINVAL; /* Find the first overlapping VMA */ |
9be34c9d5
|
2560 |
vma = find_vma(mm, start); |
146425a31
|
2561 |
if (!vma) |
1da177e4c
|
2562 |
return 0; |
9be34c9d5
|
2563 |
prev = vma->vm_prev; |
146425a31
|
2564 |
/* we have start < vma->vm_end */ |
1da177e4c
|
2565 2566 2567 |
/* if it doesn't overlap, we have nothing.. */ end = start + len; |
146425a31
|
2568 |
if (vma->vm_start >= end) |
1da177e4c
|
2569 2570 2571 2572 2573 2574 2575 2576 2577 |
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
|
2578 |
if (start > vma->vm_start) { |
659ace584
|
2579 2580 2581 2582 2583 2584 2585 2586 2587 2588 2589 |
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
|
2590 2591 |
if (error) return error; |
146425a31
|
2592 |
prev = vma; |
1da177e4c
|
2593 2594 2595 2596 2597 |
} /* Does it split the last one? */ last = find_vma(mm, end); if (last && end > last->vm_start) { |
659ace584
|
2598 |
int error = __split_vma(mm, last, end, 1); |
1da177e4c
|
2599 2600 2601 |
if (error) return error; } |
cc71aba34
|
2602 |
vma = prev ? prev->vm_next : mm->mmap; |
1da177e4c
|
2603 |
|
2376dd7ce
|
2604 2605 2606 2607 2608 2609 2610 2611 2612 2613 2614 2615 2616 2617 |
if (unlikely(uf)) { /* * If userfaultfd_unmap_prep returns an error the vmas * will remain splitted, but userland will get a * highly unexpected error anyway. This is no * different than the case where the first of the two * __split_vma fails, but we don't undo the first * split, despite we could. This is unlikely enough * failure that it's not worth optimizing it for. */ int error = userfaultfd_unmap_prep(vma, start, end, uf); if (error) return error; } |
1da177e4c
|
2618 |
/* |
ba470de43
|
2619 2620 2621 2622 2623 2624 2625 2626 2627 2628 2629 2630 2631 2632 |
* 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
|
2633 2634 |
* Remove the vma's, and unmap the actual pages */ |
146425a31
|
2635 2636 |
detach_vmas_to_be_unmapped(mm, vma, prev, end); unmap_region(mm, vma, prev, start, end); |
1da177e4c
|
2637 |
|
1de4fa14e
|
2638 |
arch_unmap(mm, vma, start, end); |
1da177e4c
|
2639 |
/* Fix up all other VM information */ |
2c0b38146
|
2640 |
remove_vma_list(mm, vma); |
1da177e4c
|
2641 2642 2643 |
return 0; } |
1da177e4c
|
2644 |
|
bfce281c2
|
2645 |
int vm_munmap(unsigned long start, size_t len) |
1da177e4c
|
2646 2647 |
{ int ret; |
bfce281c2
|
2648 |
struct mm_struct *mm = current->mm; |
897ab3e0c
|
2649 |
LIST_HEAD(uf); |
1da177e4c
|
2650 |
|
ae7987835
|
2651 2652 |
if (down_write_killable(&mm->mmap_sem)) return -EINTR; |
897ab3e0c
|
2653 |
ret = do_munmap(mm, start, len, &uf); |
1da177e4c
|
2654 |
up_write(&mm->mmap_sem); |
897ab3e0c
|
2655 |
userfaultfd_unmap_complete(mm, &uf); |
1da177e4c
|
2656 2657 |
return ret; } |
a46ef99d8
|
2658 2659 2660 2661 2662 |
EXPORT_SYMBOL(vm_munmap); SYSCALL_DEFINE2(munmap, unsigned long, addr, size_t, len) { profile_munmap(addr); |
846b1a0f1
|
2663 |
return vm_munmap(addr, len); |
a46ef99d8
|
2664 |
} |
1da177e4c
|
2665 |
|
c8d78c182
|
2666 2667 2668 2669 2670 2671 2672 2673 2674 2675 2676 2677 2678 |
/* * 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
|
2679 2680 2681 |
pr_warn_once("%s (%d) uses deprecated remap_file_pages() syscall. See Documentation/vm/remap_file_pages.txt. ", current->comm, current->pid); |
c8d78c182
|
2682 2683 2684 2685 2686 2687 2688 2689 2690 2691 2692 2693 |
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
|
2694 2695 |
if (down_write_killable(&mm->mmap_sem)) return -EINTR; |
c8d78c182
|
2696 2697 2698 2699 |
vma = find_vma(mm, start); if (!vma || !(vma->vm_flags & VM_SHARED)) goto out; |
48f7df329
|
2700 |
if (start < vma->vm_start) |
c8d78c182
|
2701 |
goto out; |
48f7df329
|
2702 2703 2704 2705 2706 2707 2708 2709 2710 2711 2712 2713 2714 2715 2716 2717 2718 2719 2720 2721 |
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
|
2722 2723 2724 2725 2726 2727 2728 2729 2730 |
} 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
|
2731 |
struct vm_area_struct *tmp; |
c8d78c182
|
2732 |
flags |= MAP_LOCKED; |
48f7df329
|
2733 |
|
c8d78c182
|
2734 |
/* drop PG_Mlocked flag for over-mapped range */ |
48f7df329
|
2735 2736 |
for (tmp = vma; tmp->vm_start >= start + size; tmp = tmp->vm_next) { |
9a73f61bd
|
2737 2738 2739 2740 2741 |
/* * Split pmd and munlock page on the border * of the range. */ vma_adjust_trans_huge(tmp, start, start + size, 0); |
48f7df329
|
2742 2743 2744 2745 |
munlock_vma_pages_range(tmp, max(tmp->vm_start, start), min(tmp->vm_end, start + size)); } |
c8d78c182
|
2746 2747 2748 2749 |
} file = get_file(vma->vm_file); ret = do_mmap_pgoff(vma->vm_file, start, size, |
897ab3e0c
|
2750 |
prot, flags, pgoff, &populate, NULL); |
c8d78c182
|
2751 2752 2753 2754 2755 2756 2757 2758 2759 |
fput(file); out: up_write(&mm->mmap_sem); if (populate) mm_populate(ret, populate); if (!IS_ERR_VALUE(ret)) ret = 0; return ret; } |
1da177e4c
|
2760 2761 |
static inline void verify_mm_writelocked(struct mm_struct *mm) { |
a241ec65a
|
2762 |
#ifdef CONFIG_DEBUG_VM |
1da177e4c
|
2763 2764 2765 2766 2767 2768 2769 2770 2771 2772 2773 2774 |
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. */ |
81ebc9dec
|
2775 |
static int do_brk_flags(unsigned long addr, unsigned long len, unsigned long flags, struct list_head *uf) |
1da177e4c
|
2776 |
{ |
cc71aba34
|
2777 2778 |
struct mm_struct *mm = current->mm; struct vm_area_struct *vma, *prev; |
cc71aba34
|
2779 |
struct rb_node **rb_link, *rb_parent; |
1da177e4c
|
2780 |
pgoff_t pgoff = addr >> PAGE_SHIFT; |
3a4597568
|
2781 |
int error; |
1da177e4c
|
2782 |
|
16e72e9b3
|
2783 2784 2785 2786 |
/* Until we need other flags, refuse anything except VM_EXEC. */ if ((flags & (~VM_EXEC)) != 0) return -EINVAL; flags |= VM_DATA_DEFAULT_FLAGS | VM_ACCOUNT | mm->def_flags; |
3a4597568
|
2787 |
|
2c6a10161
|
2788 |
error = get_unmapped_area(NULL, addr, len, 0, MAP_FIXED); |
de1741a13
|
2789 |
if (offset_in_page(error)) |
3a4597568
|
2790 |
return error; |
363ee17f0
|
2791 2792 2793 |
error = mlock_future_check(mm, mm->def_flags, len); if (error) return error; |
1da177e4c
|
2794 2795 2796 2797 2798 2799 2800 2801 2802 2803 |
/* * 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
|
2804 2805 |
while (find_vma_links(mm, addr, addr + len, &prev, &rb_link, &rb_parent)) { |
897ab3e0c
|
2806 |
if (do_munmap(mm, addr, len, uf)) |
1da177e4c
|
2807 |
return -ENOMEM; |
1da177e4c
|
2808 2809 2810 |
} /* Check against address space limits *after* clearing old maps... */ |
846383359
|
2811 |
if (!may_expand_vm(mm, flags, len >> PAGE_SHIFT)) |
1da177e4c
|
2812 2813 2814 2815 |
return -ENOMEM; if (mm->map_count > sysctl_max_map_count) return -ENOMEM; |
191c54244
|
2816 |
if (security_vm_enough_memory_mm(mm, len >> PAGE_SHIFT)) |
1da177e4c
|
2817 |
return -ENOMEM; |
1da177e4c
|
2818 |
/* Can we just expand an old private anonymous mapping? */ |
ba470de43
|
2819 |
vma = vma_merge(mm, prev, addr, addr + len, flags, |
19a809afe
|
2820 |
NULL, NULL, pgoff, NULL, NULL_VM_UFFD_CTX); |
ba470de43
|
2821 |
if (vma) |
1da177e4c
|
2822 2823 2824 2825 2826 |
goto out; /* * create a vma struct for an anonymous mapping */ |
c5e3b83e9
|
2827 |
vma = kmem_cache_zalloc(vm_area_cachep, GFP_KERNEL); |
1da177e4c
|
2828 2829 2830 2831 |
if (!vma) { vm_unacct_memory(len >> PAGE_SHIFT); return -ENOMEM; } |
1da177e4c
|
2832 |
|
5beb49305
|
2833 |
INIT_LIST_HEAD(&vma->anon_vma_chain); |
1da177e4c
|
2834 2835 2836 2837 2838 |
vma->vm_mm = mm; vma->vm_start = addr; vma->vm_end = addr + len; vma->vm_pgoff = pgoff; vma->vm_flags = flags; |
3ed75eb8f
|
2839 |
vma->vm_page_prot = vm_get_page_prot(flags); |
1da177e4c
|
2840 2841 |
vma_link(mm, vma, prev, rb_link, rb_parent); out: |
3af9e8592
|
2842 |
perf_event_mmap(vma); |
1da177e4c
|
2843 |
mm->total_vm += len >> PAGE_SHIFT; |
846383359
|
2844 |
mm->data_vm += len >> PAGE_SHIFT; |
128557ffe
|
2845 2846 |
if (flags & VM_LOCKED) mm->locked_vm += (len >> PAGE_SHIFT); |
d9104d1ca
|
2847 |
vma->vm_flags |= VM_SOFTDIRTY; |
5d22fc25d
|
2848 |
return 0; |
1da177e4c
|
2849 |
} |
81ebc9dec
|
2850 |
int vm_brk_flags(unsigned long addr, unsigned long request, unsigned long flags) |
e4eb1ff61
|
2851 2852 |
{ struct mm_struct *mm = current->mm; |
81ebc9dec
|
2853 |
unsigned long len; |
5d22fc25d
|
2854 |
int ret; |
128557ffe
|
2855 |
bool populate; |
897ab3e0c
|
2856 |
LIST_HEAD(uf); |
e4eb1ff61
|
2857 |
|
81ebc9dec
|
2858 2859 2860 2861 2862 |
len = PAGE_ALIGN(request); if (len < request) return -ENOMEM; if (!len) return 0; |
2d6c92824
|
2863 2864 |
if (down_write_killable(&mm->mmap_sem)) return -EINTR; |
897ab3e0c
|
2865 |
ret = do_brk_flags(addr, len, flags, &uf); |
128557ffe
|
2866 |
populate = ((mm->def_flags & VM_LOCKED) != 0); |
e4eb1ff61
|
2867 |
up_write(&mm->mmap_sem); |
897ab3e0c
|
2868 |
userfaultfd_unmap_complete(mm, &uf); |
5d22fc25d
|
2869 |
if (populate && !ret) |
128557ffe
|
2870 |
mm_populate(addr, len); |
e4eb1ff61
|
2871 2872 |
return ret; } |
16e72e9b3
|
2873 2874 2875 2876 2877 2878 |
EXPORT_SYMBOL(vm_brk_flags); int vm_brk(unsigned long addr, unsigned long len) { return vm_brk_flags(addr, len, 0); } |
e4eb1ff61
|
2879 |
EXPORT_SYMBOL(vm_brk); |
1da177e4c
|
2880 2881 2882 2883 |
/* Release all mmaps. */ void exit_mmap(struct mm_struct *mm) { |
d16dfc550
|
2884 |
struct mmu_gather tlb; |
ba470de43
|
2885 |
struct vm_area_struct *vma; |
1da177e4c
|
2886 |
unsigned long nr_accounted = 0; |
d6dd61c83
|
2887 |
/* mm's last user has gone, and its about to be pulled down */ |
cddb8a5c1
|
2888 |
mmu_notifier_release(mm); |
d6dd61c83
|
2889 |
|
2270dfcc4
|
2890 2891 2892 2893 2894 2895 2896 2897 2898 2899 2900 2901 2902 2903 2904 2905 2906 2907 2908 2909 2910 2911 2912 2913 2914 |
if (unlikely(mm_is_oom_victim(mm))) { /* * Manually reap the mm to free as much memory as possible. * Then, as the oom reaper does, set MMF_OOM_SKIP to disregard * this mm from further consideration. Taking mm->mmap_sem for * write after setting MMF_OOM_SKIP will guarantee that the oom * reaper will not run on this mm again after mmap_sem is * dropped. * * Nothing can be holding mm->mmap_sem here and the above call * to mmu_notifier_release(mm) ensures mmu notifier callbacks in * __oom_reap_task_mm() will not block. * * This needs to be done before calling munlock_vma_pages_all(), * which clears VM_LOCKED, otherwise the oom reaper cannot * reliably test it. */ mutex_lock(&oom_lock); __oom_reap_task_mm(mm); mutex_unlock(&oom_lock); set_bit(MMF_OOM_SKIP, &mm->flags); down_write(&mm->mmap_sem); up_write(&mm->mmap_sem); } |
ba470de43
|
2915 2916 2917 2918 2919 2920 2921 2922 |
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
|
2923 2924 |
arch_exit_mmap(mm); |
ba470de43
|
2925 |
vma = mm->mmap; |
9480c53e9
|
2926 2927 |
if (!vma) /* Can happen if dup_mmap() received an OOM */ return; |
1da177e4c
|
2928 |
lru_add_drain(); |
1da177e4c
|
2929 |
flush_cache_mm(mm); |
2b047252d
|
2930 |
tlb_gather_mmu(&tlb, mm, 0, -1); |
901608d90
|
2931 |
/* update_hiwater_rss(mm) here? but nobody should be looking */ |
e0da382c9
|
2932 |
/* Use -1 here to ensure all VMAs in the mm are unmapped */ |
4f74d2c8e
|
2933 |
unmap_vmas(&tlb, vma, 0, -1); |
6ee8630e0
|
2934 |
free_pgtables(&tlb, vma, FIRST_USER_ADDRESS, USER_PGTABLES_CEILING); |
853f5e264
|
2935 |
tlb_finish_mmu(&tlb, 0, -1); |
1da177e4c
|
2936 |
|
1da177e4c
|
2937 |
/* |
8f4f8c164
|
2938 2939 |
* Walk the list again, actually closing and freeing it, * with preemption enabled, without holding any MM locks. |
1da177e4c
|
2940 |
*/ |
4f74d2c8e
|
2941 2942 2943 |
while (vma) { if (vma->vm_flags & VM_ACCOUNT) nr_accounted += vma_pages(vma); |
a8fb5618d
|
2944 |
vma = remove_vma(vma); |
4f74d2c8e
|
2945 2946 |
} vm_unacct_memory(nr_accounted); |
1da177e4c
|
2947 2948 2949 2950 |
} /* Insert vm structure into process list sorted by address * and into the inode's i_mmap tree. If vm_file is non-NULL |
c8c06efa8
|
2951 |
* then i_mmap_rwsem is taken here. |
1da177e4c
|
2952 |
*/ |
6597d7833
|
2953 |
int insert_vm_struct(struct mm_struct *mm, struct vm_area_struct *vma) |
1da177e4c
|
2954 |
{ |
6597d7833
|
2955 2956 |
struct vm_area_struct *prev; struct rb_node **rb_link, *rb_parent; |
1da177e4c
|
2957 |
|
c9d13f5fc
|
2958 2959 2960 2961 2962 2963 |
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
|
2964 2965 2966 2967 2968 2969 2970 2971 2972 2973 2974 2975 |
/* * 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
|
2976 |
if (vma_is_anonymous(vma)) { |
1da177e4c
|
2977 2978 2979 |
BUG_ON(vma->anon_vma); vma->vm_pgoff = vma->vm_start >> PAGE_SHIFT; } |
2b1444983
|
2980 |
|
1da177e4c
|
2981 2982 2983 2984 2985 2986 2987 2988 2989 |
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
|
2990 2991 |
unsigned long addr, unsigned long len, pgoff_t pgoff, bool *need_rmap_locks) |
1da177e4c
|
2992 2993 2994 2995 2996 2997 |
{ 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
|
2998 |
bool faulted_in_anon_vma = true; |
1da177e4c
|
2999 3000 3001 3002 3003 |
/* * If anonymous vma has not yet been faulted, update new pgoff * to match new location, to increase its chance of merging. */ |
ce75799b8
|
3004 |
if (unlikely(vma_is_anonymous(vma) && !vma->anon_vma)) { |
1da177e4c
|
3005 |
pgoff = addr >> PAGE_SHIFT; |
948f017b0
|
3006 3007 |
faulted_in_anon_vma = false; } |
1da177e4c
|
3008 |
|
6597d7833
|
3009 3010 |
if (find_vma_links(mm, addr, addr + len, &prev, &rb_link, &rb_parent)) return NULL; /* should never get here */ |
1da177e4c
|
3011 |
new_vma = vma_merge(mm, prev, addr, addr + len, vma->vm_flags, |
19a809afe
|
3012 3013 |
vma->anon_vma, vma->vm_file, pgoff, vma_policy(vma), vma->vm_userfaultfd_ctx); |
1da177e4c
|
3014 3015 3016 3017 |
if (new_vma) { /* * Source vma may have been merged into new_vma */ |
948f017b0
|
3018 3019 3020 3021 3022 3023 3024 3025 3026 3027 3028 3029 3030 3031 |
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
|
3032 |
VM_BUG_ON_VMA(faulted_in_anon_vma, new_vma); |
38a76013a
|
3033 |
*vmap = vma = new_vma; |
108d6642a
|
3034 |
} |
38a76013a
|
3035 |
*need_rmap_locks = (new_vma->vm_pgoff <= vma->vm_pgoff); |
1da177e4c
|
3036 |
} else { |
e94b17660
|
3037 |
new_vma = kmem_cache_alloc(vm_area_cachep, GFP_KERNEL); |
e39758912
|
3038 3039 3040 3041 3042 3043 3044 3045 3046 3047 3048 3049 3050 3051 3052 3053 3054 |
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
|
3055 3056 |
} return new_vma; |
5beb49305
|
3057 |
|
e39758912
|
3058 |
out_free_mempol: |
ef0855d33
|
3059 |
mpol_put(vma_policy(new_vma)); |
e39758912
|
3060 |
out_free_vma: |
5beb49305
|
3061 |
kmem_cache_free(vm_area_cachep, new_vma); |
e39758912
|
3062 |
out: |
5beb49305
|
3063 |
return NULL; |
1da177e4c
|
3064 |
} |
119f657c7
|
3065 3066 3067 3068 3069 |
/* * Return true if the calling process may expand its vm space by the passed * number of pages */ |
846383359
|
3070 |
bool may_expand_vm(struct mm_struct *mm, vm_flags_t flags, unsigned long npages) |
119f657c7
|
3071 |
{ |
846383359
|
3072 3073 |
if (mm->total_vm + npages > rlimit(RLIMIT_AS) >> PAGE_SHIFT) return false; |
119f657c7
|
3074 |
|
d977d56ce
|
3075 3076 |
if (is_data_mapping(flags) && mm->data_vm + npages > rlimit(RLIMIT_DATA) >> PAGE_SHIFT) { |
f4fcd5584
|
3077 3078 3079 3080 3081 3082 3083 |
/* 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
|
3084 3085 3086 |
current->comm, current->pid, (mm->data_vm + npages) << PAGE_SHIFT, rlimit(RLIMIT_DATA)); |
d977d56ce
|
3087 |
return false; |
f4fcd5584
|
3088 |
} |
d977d56ce
|
3089 |
} |
119f657c7
|
3090 |
|
846383359
|
3091 3092 3093 3094 3095 3096 |
return true; } void vm_stat_account(struct mm_struct *mm, vm_flags_t flags, long npages) { mm->total_vm += npages; |
d977d56ce
|
3097 |
if (is_exec_mapping(flags)) |
846383359
|
3098 |
mm->exec_vm += npages; |
d977d56ce
|
3099 |
else if (is_stack_mapping(flags)) |
846383359
|
3100 |
mm->stack_vm += npages; |
d977d56ce
|
3101 |
else if (is_data_mapping(flags)) |
846383359
|
3102 |
mm->data_vm += npages; |
119f657c7
|
3103 |
} |
fa5dc22f8
|
3104 |
|
11bac8000
|
3105 |
static int special_mapping_fault(struct vm_fault *vmf); |
a62c34bd2
|
3106 3107 3108 3109 3110 3111 3112 3113 3114 3115 3116 3117 |
/* * 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
|
3118 3119 3120 |
static int special_mapping_mremap(struct vm_area_struct *new_vma) { struct vm_special_mapping *sm = new_vma->vm_private_data; |
280e87e98
|
3121 3122 |
if (WARN_ON_ONCE(current->mm != new_vma->vm_mm)) return -EFAULT; |
b059a453b
|
3123 3124 |
if (sm->mremap) return sm->mremap(sm, new_vma); |
280e87e98
|
3125 |
|
b059a453b
|
3126 3127 |
return 0; } |
a62c34bd2
|
3128 3129 3130 |
static const struct vm_operations_struct special_mapping_vmops = { .close = special_mapping_close, .fault = special_mapping_fault, |
b059a453b
|
3131 |
.mremap = special_mapping_mremap, |
a62c34bd2
|
3132 3133 3134 3135 3136 3137 3138 |
.name = special_mapping_name, }; static const struct vm_operations_struct legacy_special_mapping_vmops = { .close = special_mapping_close, .fault = special_mapping_fault, }; |
fa5dc22f8
|
3139 |
|
11bac8000
|
3140 |
static int special_mapping_fault(struct vm_fault *vmf) |
fa5dc22f8
|
3141 |
{ |
11bac8000
|
3142 |
struct vm_area_struct *vma = vmf->vma; |
b1d0e4f53
|
3143 |
pgoff_t pgoff; |
fa5dc22f8
|
3144 |
struct page **pages; |
f872f5400
|
3145 |
if (vma->vm_ops == &legacy_special_mapping_vmops) { |
a62c34bd2
|
3146 |
pages = vma->vm_private_data; |
f872f5400
|
3147 3148 3149 3150 |
} else { struct vm_special_mapping *sm = vma->vm_private_data; if (sm->fault) |
11bac8000
|
3151 |
return sm->fault(sm, vmf->vma, vmf); |
f872f5400
|
3152 3153 3154 |
pages = sm->pages; } |
a62c34bd2
|
3155 |
|
8a9cc3b55
|
3156 |
for (pgoff = vmf->pgoff; pgoff && *pages; ++pages) |
b1d0e4f53
|
3157 |
pgoff--; |
fa5dc22f8
|
3158 3159 3160 3161 |
if (*pages) { struct page *page = *pages; get_page(page); |
b1d0e4f53
|
3162 3163 |
vmf->page = page; return 0; |
fa5dc22f8
|
3164 |
} |
b1d0e4f53
|
3165 |
return VM_FAULT_SIGBUS; |
fa5dc22f8
|
3166 |
} |
a62c34bd2
|
3167 3168 3169 |
static struct vm_area_struct *__install_special_mapping( struct mm_struct *mm, unsigned long addr, unsigned long len, |
27f28b972
|
3170 3171 |
unsigned long vm_flags, void *priv, const struct vm_operations_struct *ops) |
fa5dc22f8
|
3172 |
{ |
462e635e5
|
3173 |
int ret; |
fa5dc22f8
|
3174 3175 3176 3177 |
struct vm_area_struct *vma; vma = kmem_cache_zalloc(vm_area_cachep, GFP_KERNEL); if (unlikely(vma == NULL)) |
3935ed6a3
|
3178 |
return ERR_PTR(-ENOMEM); |
fa5dc22f8
|
3179 |
|
5beb49305
|
3180 |
INIT_LIST_HEAD(&vma->anon_vma_chain); |
fa5dc22f8
|
3181 3182 3183 |
vma->vm_mm = mm; vma->vm_start = addr; vma->vm_end = addr + len; |
d9104d1ca
|
3184 |
vma->vm_flags = vm_flags | mm->def_flags | VM_DONTEXPAND | VM_SOFTDIRTY; |
3ed75eb8f
|
3185 |
vma->vm_page_prot = vm_get_page_prot(vma->vm_flags); |
fa5dc22f8
|
3186 |
|
a62c34bd2
|
3187 3188 |
vma->vm_ops = ops; vma->vm_private_data = priv; |
fa5dc22f8
|
3189 |
|
462e635e5
|
3190 3191 3192 |
ret = insert_vm_struct(mm, vma); if (ret) goto out; |
fa5dc22f8
|
3193 |
|
846383359
|
3194 |
vm_stat_account(mm, vma->vm_flags, len >> PAGE_SHIFT); |
fa5dc22f8
|
3195 |
|
cdd6c482c
|
3196 |
perf_event_mmap(vma); |
089dd79db
|
3197 |
|
3935ed6a3
|
3198 |
return vma; |
462e635e5
|
3199 3200 3201 |
out: kmem_cache_free(vm_area_cachep, vma); |
3935ed6a3
|
3202 3203 |
return ERR_PTR(ret); } |
2eefd8789
|
3204 3205 3206 3207 3208 3209 3210 |
bool vma_is_special_mapping(const struct vm_area_struct *vma, const struct vm_special_mapping *sm) { return vma->vm_private_data == sm && (vma->vm_ops == &special_mapping_vmops || vma->vm_ops == &legacy_special_mapping_vmops); } |
a62c34bd2
|
3211 3212 3213 3214 3215 3216 3217 3218 3219 3220 3221 3222 3223 3224 |
/* * 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
|
3225 3226 |
return __install_special_mapping(mm, addr, len, vm_flags, (void *)spec, &special_mapping_vmops); |
a62c34bd2
|
3227 |
} |
3935ed6a3
|
3228 3229 3230 3231 |
int install_special_mapping(struct mm_struct *mm, unsigned long addr, unsigned long len, unsigned long vm_flags, struct page **pages) { |
a62c34bd2
|
3232 |
struct vm_area_struct *vma = __install_special_mapping( |
27f28b972
|
3233 3234 |
mm, addr, len, vm_flags, (void *)pages, &legacy_special_mapping_vmops); |
3935ed6a3
|
3235 |
|
14bd5b458
|
3236 |
return PTR_ERR_OR_ZERO(vma); |
fa5dc22f8
|
3237 |
} |
7906d00cd
|
3238 3239 |
static DEFINE_MUTEX(mm_all_locks_mutex); |
454ed842d
|
3240 |
static void vm_lock_anon_vma(struct mm_struct *mm, struct anon_vma *anon_vma) |
7906d00cd
|
3241 |
{ |
f808c13fd
|
3242 |
if (!test_bit(0, (unsigned long *) &anon_vma->root->rb_root.rb_root.rb_node)) { |
7906d00cd
|
3243 3244 3245 3246 |
/* * The LSB of head.next can't change from under us * because we hold the mm_all_locks_mutex. */ |
572043c90
|
3247 |
down_write_nest_lock(&anon_vma->root->rwsem, &mm->mmap_sem); |
7906d00cd
|
3248 3249 |
/* * We can safely modify head.next after taking the |
5a505085f
|
3250 |
* anon_vma->root->rwsem. If some other vma in this mm shares |
7906d00cd
|
3251 3252 3253 3254 |
* 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
|
3255 |
* anon_vma->root->rwsem. |
7906d00cd
|
3256 3257 |
*/ if (__test_and_set_bit(0, (unsigned long *) |
f808c13fd
|
3258 |
&anon_vma->root->rb_root.rb_root.rb_node)) |
7906d00cd
|
3259 3260 3261 |
BUG(); } } |
454ed842d
|
3262 |
static void vm_lock_mapping(struct mm_struct *mm, struct address_space *mapping) |
7906d00cd
|
3263 3264 3265 3266 3267 3268 3269 3270 3271 3272 3273 3274 3275 |
{ 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
|
3276 |
down_write_nest_lock(&mapping->i_mmap_rwsem, &mm->mmap_sem); |
7906d00cd
|
3277 3278 3279 3280 3281 3282 3283 3284 3285 3286 3287 3288 3289 3290 |
} } /* * 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
|
3291 |
* altering the vma layout. It's also needed in write mode to avoid new |
7906d00cd
|
3292 3293 3294 3295 3296 |
* 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
|
3297 |
* The LSB in anon_vma->rb_root.rb_node and the AS_MM_ALL_LOCKS bitflag in |
7906d00cd
|
3298 3299 3300 |
* 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
|
3301 3302 3303 3304 3305 3306 3307 3308 3309 3310 |
* 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
|
3311 3312 3313 3314 3315 3316 3317 3318 3319 |
* * 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
|
3320 |
struct anon_vma_chain *avc; |
7906d00cd
|
3321 3322 3323 3324 3325 3326 3327 3328 |
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
|
3329 3330 3331 3332 3333 3334 3335 3336 3337 3338 |
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
|
3339 |
vm_lock_mapping(mm, vma->vm_file->f_mapping); |
7906d00cd
|
3340 |
} |
7cd5a02f5
|
3341 3342 3343 3344 3345 |
for (vma = mm->mmap; vma; vma = vma->vm_next) { if (signal_pending(current)) goto out_unlock; if (vma->anon_vma) |
5beb49305
|
3346 3347 |
list_for_each_entry(avc, &vma->anon_vma_chain, same_vma) vm_lock_anon_vma(mm, avc->anon_vma); |
7906d00cd
|
3348 |
} |
7cd5a02f5
|
3349 |
|
584cff54e
|
3350 |
return 0; |
7906d00cd
|
3351 3352 |
out_unlock: |
584cff54e
|
3353 3354 |
mm_drop_all_locks(mm); return -EINTR; |
7906d00cd
|
3355 3356 3357 3358 |
} static void vm_unlock_anon_vma(struct anon_vma *anon_vma) { |
f808c13fd
|
3359 |
if (test_bit(0, (unsigned long *) &anon_vma->root->rb_root.rb_root.rb_node)) { |
7906d00cd
|
3360 3361 3362 3363 3364 |
/* * 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
|
3365 |
* the vma so the users using the anon_vma->rb_root will |
7906d00cd
|
3366 3367 3368 3369 |
* never see our bitflag. * * No need of atomic instructions here, head.next * can't change from under us until we release the |
5a505085f
|
3370 |
* anon_vma->root->rwsem. |
7906d00cd
|
3371 3372 |
*/ if (!__test_and_clear_bit(0, (unsigned long *) |
f808c13fd
|
3373 |
&anon_vma->root->rb_root.rb_root.rb_node)) |
7906d00cd
|
3374 |
BUG(); |
08b52706d
|
3375 |
anon_vma_unlock_write(anon_vma); |
7906d00cd
|
3376 3377 3378 3379 3380 3381 3382 3383 3384 3385 |
} } 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
|
3386 |
i_mmap_unlock_write(mapping); |
7906d00cd
|
3387 3388 3389 3390 3391 3392 3393 3394 3395 3396 3397 3398 3399 |
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
|
3400 |
struct anon_vma_chain *avc; |
7906d00cd
|
3401 3402 3403 3404 3405 3406 |
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
|
3407 3408 |
list_for_each_entry(avc, &vma->anon_vma_chain, same_vma) vm_unlock_anon_vma(avc->anon_vma); |
7906d00cd
|
3409 3410 3411 3412 3413 3414 |
if (vma->vm_file && vma->vm_file->f_mapping) vm_unlock_mapping(vma->vm_file->f_mapping); } mutex_unlock(&mm_all_locks_mutex); } |
8feae1311
|
3415 3416 |
/* |
3edf41d84
|
3417 |
* initialise the percpu counter for VM |
8feae1311
|
3418 3419 3420 |
*/ void __init mmap_init(void) { |
00a62ce91
|
3421 |
int ret; |
908c7f194
|
3422 |
ret = percpu_counter_init(&vm_committed_as, 0, GFP_KERNEL); |
00a62ce91
|
3423 |
VM_BUG_ON(ret); |
8feae1311
|
3424 |
} |
c9b1d0981
|
3425 3426 3427 3428 3429 3430 3431 3432 3433 3434 3435 |
/* * 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
|
3436 |
static int init_user_reserve(void) |
c9b1d0981
|
3437 3438 |
{ unsigned long free_kbytes; |
c41f012ad
|
3439 |
free_kbytes = global_zone_page_state(NR_FREE_PAGES) << (PAGE_SHIFT - 10); |
c9b1d0981
|
3440 3441 3442 3443 |
sysctl_user_reserve_kbytes = min(free_kbytes / 32, 1UL << 17); return 0; } |
a64fb3cd6
|
3444 |
subsys_initcall(init_user_reserve); |
4eeab4f55
|
3445 3446 3447 3448 3449 3450 3451 3452 3453 3454 3455 |
/* * 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
|
3456 |
static int init_admin_reserve(void) |
4eeab4f55
|
3457 3458 |
{ unsigned long free_kbytes; |
c41f012ad
|
3459 |
free_kbytes = global_zone_page_state(NR_FREE_PAGES) << (PAGE_SHIFT - 10); |
4eeab4f55
|
3460 3461 3462 3463 |
sysctl_admin_reserve_kbytes = min(free_kbytes / 32, 1UL << 13); return 0; } |
a64fb3cd6
|
3464 |
subsys_initcall(init_admin_reserve); |
1640879af
|
3465 3466 3467 3468 3469 3470 3471 3472 3473 3474 3475 3476 3477 3478 3479 3480 3481 3482 3483 3484 3485 3486 3487 3488 3489 3490 3491 3492 3493 3494 3495 3496 3497 3498 3499 3500 3501 3502 |
/* * 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: |
c41f012ad
|
3503 |
free_kbytes = global_zone_page_state(NR_FREE_PAGES) << (PAGE_SHIFT - 10); |
1640879af
|
3504 3505 3506 3507 3508 3509 3510 3511 3512 3513 3514 3515 3516 3517 3518 3519 3520 3521 3522 3523 3524 3525 3526 3527 3528 3529 3530 3531 |
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
|
3532 3533 |
pr_err("Failed registering memory add/remove notifier for admin reserve "); |
1640879af
|
3534 3535 3536 |
return 0; } |
a64fb3cd6
|
3537 |
subsys_initcall(init_reserve_notifier); |