Blame view
mm/vmalloc.c
89.4 KB
457c89965 treewide: Add SPD... |
1 |
// SPDX-License-Identifier: GPL-2.0-only |
1da177e4c Linux-2.6.12-rc2 |
2 3 4 5 6 7 8 |
/* * linux/mm/vmalloc.c * * Copyright (C) 1993 Linus Torvalds * Support of BIGMEM added by Gerhard Wichert, Siemens AG, July 1999 * SMP-safe vmalloc/vfree/ioremap, Tigran Aivazian <tigran@veritas.com>, May 2000 * Major rework to support vmap/vunmap, Christoph Hellwig, SGI, August 2002 |
930fc45a4 [PATCH] vmalloc_node |
9 |
* Numa awareness, Christoph Lameter, SGI, June 2005 |
1da177e4c Linux-2.6.12-rc2 |
10 |
*/ |
db64fe022 mm: rewrite vmap ... |
11 |
#include <linux/vmalloc.h> |
1da177e4c Linux-2.6.12-rc2 |
12 13 14 |
#include <linux/mm.h> #include <linux/module.h> #include <linux/highmem.h> |
c3edc4010 sched/headers: Mo... |
15 |
#include <linux/sched/signal.h> |
1da177e4c Linux-2.6.12-rc2 |
16 17 18 |
#include <linux/slab.h> #include <linux/spinlock.h> #include <linux/interrupt.h> |
5f6a6a9c4 proc: move /proc/... |
19 |
#include <linux/proc_fs.h> |
a10aa5798 vmalloc: show vma... |
20 |
#include <linux/seq_file.h> |
868b104d7 mm/vmalloc: Add f... |
21 |
#include <linux/set_memory.h> |
3ac7fe5a4 infrastructure to... |
22 |
#include <linux/debugobjects.h> |
230169693 vmallocinfo: add ... |
23 |
#include <linux/kallsyms.h> |
db64fe022 mm: rewrite vmap ... |
24 |
#include <linux/list.h> |
4da56b99d mm/vmap: Add a no... |
25 |
#include <linux/notifier.h> |
db64fe022 mm: rewrite vmap ... |
26 27 28 |
#include <linux/rbtree.h> #include <linux/radix-tree.h> #include <linux/rcupdate.h> |
f0aa66179 vmalloc: implemen... |
29 |
#include <linux/pfn.h> |
89219d37a kmemleak: Add the... |
30 |
#include <linux/kmemleak.h> |
60063497a atomic: use <linu... |
31 |
#include <linux/atomic.h> |
3b32123d7 mm: use macros fr... |
32 |
#include <linux/compiler.h> |
32fcfd407 make vfree() safe... |
33 |
#include <linux/llist.h> |
0f616be12 mm: change __get_... |
34 |
#include <linux/bitops.h> |
68ad4a330 mm/vmalloc.c: kee... |
35 |
#include <linux/rbtree_augmented.h> |
3b32123d7 mm: use macros fr... |
36 |
|
7c0f6ba68 Replace <asm/uacc... |
37 |
#include <linux/uaccess.h> |
1da177e4c Linux-2.6.12-rc2 |
38 |
#include <asm/tlbflush.h> |
2dca6999e mm, perf_event: M... |
39 |
#include <asm/shmparam.h> |
1da177e4c Linux-2.6.12-rc2 |
40 |
|
dd56b0464 mm: page_alloc: h... |
41 |
#include "internal.h" |
32fcfd407 make vfree() safe... |
42 43 44 45 46 47 48 49 50 51 52 |
struct vfree_deferred { struct llist_head list; struct work_struct wq; }; static DEFINE_PER_CPU(struct vfree_deferred, vfree_deferred); static void __vunmap(const void *, int); static void free_work(struct work_struct *w) { struct vfree_deferred *p = container_of(w, struct vfree_deferred, wq); |
894e58c14 mm/vmalloc.c: don... |
53 54 55 56 |
struct llist_node *t, *llnode; llist_for_each_safe(llnode, t, llist_del_all(&p->list)) __vunmap((void *)llnode, 1); |
32fcfd407 make vfree() safe... |
57 |
} |
db64fe022 mm: rewrite vmap ... |
58 |
/*** Page table manipulation functions ***/ |
b221385bc [PATCH] mm/: make... |
59 |
|
1da177e4c Linux-2.6.12-rc2 |
60 61 62 63 64 65 66 67 68 69 |
static void vunmap_pte_range(pmd_t *pmd, unsigned long addr, unsigned long end) { pte_t *pte; pte = pte_offset_kernel(pmd, addr); do { pte_t ptent = ptep_get_and_clear(&init_mm, addr, pte); WARN_ON(!pte_none(ptent) && !pte_present(ptent)); } while (pte++, addr += PAGE_SIZE, addr != end); } |
db64fe022 mm: rewrite vmap ... |
70 |
static void vunmap_pmd_range(pud_t *pud, unsigned long addr, unsigned long end) |
1da177e4c Linux-2.6.12-rc2 |
71 72 73 74 75 76 77 |
{ pmd_t *pmd; unsigned long next; pmd = pmd_offset(pud, addr); do { next = pmd_addr_end(addr, end); |
b9820d8f3 mm: change vunmap... |
78 79 |
if (pmd_clear_huge(pmd)) continue; |
1da177e4c Linux-2.6.12-rc2 |
80 81 82 83 84 |
if (pmd_none_or_clear_bad(pmd)) continue; vunmap_pte_range(pmd, addr, next); } while (pmd++, addr = next, addr != end); } |
c2febafc6 mm: convert gener... |
85 |
static void vunmap_pud_range(p4d_t *p4d, unsigned long addr, unsigned long end) |
1da177e4c Linux-2.6.12-rc2 |
86 87 88 |
{ pud_t *pud; unsigned long next; |
c2febafc6 mm: convert gener... |
89 |
pud = pud_offset(p4d, addr); |
1da177e4c Linux-2.6.12-rc2 |
90 91 |
do { next = pud_addr_end(addr, end); |
b9820d8f3 mm: change vunmap... |
92 93 |
if (pud_clear_huge(pud)) continue; |
1da177e4c Linux-2.6.12-rc2 |
94 95 96 97 98 |
if (pud_none_or_clear_bad(pud)) continue; vunmap_pmd_range(pud, addr, next); } while (pud++, addr = next, addr != end); } |
c2febafc6 mm: convert gener... |
99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 |
static void vunmap_p4d_range(pgd_t *pgd, unsigned long addr, unsigned long end) { p4d_t *p4d; unsigned long next; p4d = p4d_offset(pgd, addr); do { next = p4d_addr_end(addr, end); if (p4d_clear_huge(p4d)) continue; if (p4d_none_or_clear_bad(p4d)) continue; vunmap_pud_range(p4d, addr, next); } while (p4d++, addr = next, addr != end); } |
db64fe022 mm: rewrite vmap ... |
114 |
static void vunmap_page_range(unsigned long addr, unsigned long end) |
1da177e4c Linux-2.6.12-rc2 |
115 116 117 |
{ pgd_t *pgd; unsigned long next; |
1da177e4c Linux-2.6.12-rc2 |
118 119 120 |
BUG_ON(addr >= end); pgd = pgd_offset_k(addr); |
1da177e4c Linux-2.6.12-rc2 |
121 122 123 124 |
do { next = pgd_addr_end(addr, end); if (pgd_none_or_clear_bad(pgd)) continue; |
c2febafc6 mm: convert gener... |
125 |
vunmap_p4d_range(pgd, addr, next); |
1da177e4c Linux-2.6.12-rc2 |
126 |
} while (pgd++, addr = next, addr != end); |
1da177e4c Linux-2.6.12-rc2 |
127 128 129 |
} static int vmap_pte_range(pmd_t *pmd, unsigned long addr, |
db64fe022 mm: rewrite vmap ... |
130 |
unsigned long end, pgprot_t prot, struct page **pages, int *nr) |
1da177e4c Linux-2.6.12-rc2 |
131 132 |
{ pte_t *pte; |
db64fe022 mm: rewrite vmap ... |
133 134 135 136 |
/* * nr is a running index into the array which helps higher level * callers keep track of where we're up to. */ |
872fec16d [PATCH] mm: init_... |
137 |
pte = pte_alloc_kernel(pmd, addr); |
1da177e4c Linux-2.6.12-rc2 |
138 139 140 |
if (!pte) return -ENOMEM; do { |
db64fe022 mm: rewrite vmap ... |
141 142 143 144 145 |
struct page *page = pages[*nr]; if (WARN_ON(!pte_none(*pte))) return -EBUSY; if (WARN_ON(!page)) |
1da177e4c Linux-2.6.12-rc2 |
146 147 |
return -ENOMEM; set_pte_at(&init_mm, addr, pte, mk_pte(page, prot)); |
db64fe022 mm: rewrite vmap ... |
148 |
(*nr)++; |
1da177e4c Linux-2.6.12-rc2 |
149 150 151 |
} while (pte++, addr += PAGE_SIZE, addr != end); return 0; } |
db64fe022 mm: rewrite vmap ... |
152 153 |
static int vmap_pmd_range(pud_t *pud, unsigned long addr, unsigned long end, pgprot_t prot, struct page **pages, int *nr) |
1da177e4c Linux-2.6.12-rc2 |
154 155 156 157 158 159 160 161 162 |
{ pmd_t *pmd; unsigned long next; pmd = pmd_alloc(&init_mm, pud, addr); if (!pmd) return -ENOMEM; do { next = pmd_addr_end(addr, end); |
db64fe022 mm: rewrite vmap ... |
163 |
if (vmap_pte_range(pmd, addr, next, prot, pages, nr)) |
1da177e4c Linux-2.6.12-rc2 |
164 165 166 167 |
return -ENOMEM; } while (pmd++, addr = next, addr != end); return 0; } |
c2febafc6 mm: convert gener... |
168 |
static int vmap_pud_range(p4d_t *p4d, unsigned long addr, |
db64fe022 mm: rewrite vmap ... |
169 |
unsigned long end, pgprot_t prot, struct page **pages, int *nr) |
1da177e4c Linux-2.6.12-rc2 |
170 171 172 |
{ pud_t *pud; unsigned long next; |
c2febafc6 mm: convert gener... |
173 |
pud = pud_alloc(&init_mm, p4d, addr); |
1da177e4c Linux-2.6.12-rc2 |
174 175 176 177 |
if (!pud) return -ENOMEM; do { next = pud_addr_end(addr, end); |
db64fe022 mm: rewrite vmap ... |
178 |
if (vmap_pmd_range(pud, addr, next, prot, pages, nr)) |
1da177e4c Linux-2.6.12-rc2 |
179 180 181 182 |
return -ENOMEM; } while (pud++, addr = next, addr != end); return 0; } |
c2febafc6 mm: convert gener... |
183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 |
static int vmap_p4d_range(pgd_t *pgd, unsigned long addr, unsigned long end, pgprot_t prot, struct page **pages, int *nr) { p4d_t *p4d; unsigned long next; p4d = p4d_alloc(&init_mm, pgd, addr); if (!p4d) return -ENOMEM; do { next = p4d_addr_end(addr, end); if (vmap_pud_range(p4d, addr, next, prot, pages, nr)) return -ENOMEM; } while (p4d++, addr = next, addr != end); return 0; } |
db64fe022 mm: rewrite vmap ... |
199 200 201 202 203 204 |
/* * Set up page tables in kva (addr, end). The ptes shall have prot "prot", and * will have pfns corresponding to the "pages" array. * * Ie. pte at addr+N*PAGE_SIZE shall point to pfn corresponding to pages[N] */ |
8fc489850 vmalloc: add un/m... |
205 206 |
static int vmap_page_range_noflush(unsigned long start, unsigned long end, pgprot_t prot, struct page **pages) |
1da177e4c Linux-2.6.12-rc2 |
207 208 209 |
{ pgd_t *pgd; unsigned long next; |
2e4e27c7d vmalloc.c: fix fl... |
210 |
unsigned long addr = start; |
db64fe022 mm: rewrite vmap ... |
211 212 |
int err = 0; int nr = 0; |
1da177e4c Linux-2.6.12-rc2 |
213 214 215 |
BUG_ON(addr >= end); pgd = pgd_offset_k(addr); |
1da177e4c Linux-2.6.12-rc2 |
216 217 |
do { next = pgd_addr_end(addr, end); |
c2febafc6 mm: convert gener... |
218 |
err = vmap_p4d_range(pgd, addr, next, prot, pages, &nr); |
1da177e4c Linux-2.6.12-rc2 |
219 |
if (err) |
bf88c8c83 vmalloc.c: fix do... |
220 |
return err; |
1da177e4c Linux-2.6.12-rc2 |
221 |
} while (pgd++, addr = next, addr != end); |
db64fe022 mm: rewrite vmap ... |
222 |
|
db64fe022 mm: rewrite vmap ... |
223 |
return nr; |
1da177e4c Linux-2.6.12-rc2 |
224 |
} |
8fc489850 vmalloc: add un/m... |
225 226 227 228 229 230 231 232 233 |
static int vmap_page_range(unsigned long start, unsigned long end, pgprot_t prot, struct page **pages) { int ret; ret = vmap_page_range_noflush(start, end, prot, pages); flush_cache_vmap(start, end); return ret; } |
81ac3ad90 kcore: register m... |
234 |
int is_vmalloc_or_module_addr(const void *x) |
73bdf0a60 Introduce is_vmal... |
235 236 |
{ /* |
ab4f2ee13 [ARM] fix naming ... |
237 |
* ARM, x86-64 and sparc64 put modules in a special place, |
73bdf0a60 Introduce is_vmal... |
238 239 240 241 242 243 244 245 246 247 |
* and fall back on vmalloc() if that fails. Others * just put it in the vmalloc space. */ #if defined(CONFIG_MODULES) && defined(MODULES_VADDR) unsigned long addr = (unsigned long)x; if (addr >= MODULES_VADDR && addr < MODULES_END) return 1; #endif return is_vmalloc_addr(x); } |
48667e7a4 Move vmalloc_to_p... |
248 |
/* |
add688fbd Revert "mm/vmallo... |
249 |
* Walk a vmap address to the struct page it maps. |
48667e7a4 Move vmalloc_to_p... |
250 |
*/ |
add688fbd Revert "mm/vmallo... |
251 |
struct page *vmalloc_to_page(const void *vmalloc_addr) |
48667e7a4 Move vmalloc_to_p... |
252 253 |
{ unsigned long addr = (unsigned long) vmalloc_addr; |
add688fbd Revert "mm/vmallo... |
254 |
struct page *page = NULL; |
48667e7a4 Move vmalloc_to_p... |
255 |
pgd_t *pgd = pgd_offset_k(addr); |
c2febafc6 mm: convert gener... |
256 257 258 259 |
p4d_t *p4d; pud_t *pud; pmd_t *pmd; pte_t *ptep, pte; |
48667e7a4 Move vmalloc_to_p... |
260 |
|
7aa413def x86, MM: virtual ... |
261 262 263 264 |
/* * XXX we might need to change this if we add VIRTUAL_BUG_ON for * architectures that do not vmalloc module space */ |
73bdf0a60 Introduce is_vmal... |
265 |
VIRTUAL_BUG_ON(!is_vmalloc_or_module_addr(vmalloc_addr)); |
59ea74633 MM: virtual addre... |
266 |
|
c2febafc6 mm: convert gener... |
267 268 269 270 271 272 |
if (pgd_none(*pgd)) return NULL; p4d = p4d_offset(pgd, addr); if (p4d_none(*p4d)) return NULL; pud = pud_offset(p4d, addr); |
029c54b09 mm/vmalloc.c: hug... |
273 274 275 276 277 278 279 280 281 282 283 |
/* * Don't dereference bad PUD or PMD (below) entries. This will also * identify huge mappings, which we may encounter on architectures * that define CONFIG_HAVE_ARCH_HUGE_VMAP=y. Such regions will be * identified as vmalloc addresses by is_vmalloc_addr(), but are * not [unambiguously] associated with a struct page, so there is * no correct value to return for them. */ WARN_ON_ONCE(pud_bad(*pud)); if (pud_none(*pud) || pud_bad(*pud)) |
c2febafc6 mm: convert gener... |
284 285 |
return NULL; pmd = pmd_offset(pud, addr); |
029c54b09 mm/vmalloc.c: hug... |
286 287 |
WARN_ON_ONCE(pmd_bad(*pmd)); if (pmd_none(*pmd) || pmd_bad(*pmd)) |
c2febafc6 mm: convert gener... |
288 289 290 291 292 293 294 |
return NULL; ptep = pte_offset_map(pmd, addr); pte = *ptep; if (pte_present(pte)) page = pte_page(pte); pte_unmap(ptep); |
add688fbd Revert "mm/vmallo... |
295 |
return page; |
48667e7a4 Move vmalloc_to_p... |
296 |
} |
add688fbd Revert "mm/vmallo... |
297 |
EXPORT_SYMBOL(vmalloc_to_page); |
48667e7a4 Move vmalloc_to_p... |
298 299 |
/* |
add688fbd Revert "mm/vmallo... |
300 |
* Map a vmalloc()-space virtual address to the physical page frame number. |
48667e7a4 Move vmalloc_to_p... |
301 |
*/ |
add688fbd Revert "mm/vmallo... |
302 |
unsigned long vmalloc_to_pfn(const void *vmalloc_addr) |
48667e7a4 Move vmalloc_to_p... |
303 |
{ |
add688fbd Revert "mm/vmallo... |
304 |
return page_to_pfn(vmalloc_to_page(vmalloc_addr)); |
48667e7a4 Move vmalloc_to_p... |
305 |
} |
add688fbd Revert "mm/vmallo... |
306 |
EXPORT_SYMBOL(vmalloc_to_pfn); |
48667e7a4 Move vmalloc_to_p... |
307 |
|
db64fe022 mm: rewrite vmap ... |
308 309 |
/*** Global kva allocator ***/ |
bb850f4da mm/vmap: add DEBU... |
310 |
#define DEBUG_AUGMENT_PROPAGATE_CHECK 0 |
a6cf4e0fe mm/vmap: add DEBU... |
311 |
#define DEBUG_AUGMENT_LOWEST_MATCH_CHECK 0 |
bb850f4da mm/vmap: add DEBU... |
312 |
|
db64fe022 mm: rewrite vmap ... |
313 |
|
db64fe022 mm: rewrite vmap ... |
314 |
static DEFINE_SPINLOCK(vmap_area_lock); |
f1c4069e1 mm, vmalloc: expo... |
315 316 |
/* Export for kexec only */ LIST_HEAD(vmap_area_list); |
80c4bd7a5 mm/vmalloc: keep ... |
317 |
static LLIST_HEAD(vmap_purge_list); |
89699605f mm: vmap area cache |
318 |
static struct rb_root vmap_area_root = RB_ROOT; |
68ad4a330 mm/vmalloc.c: kee... |
319 |
static bool vmap_initialized __read_mostly; |
89699605f mm: vmap area cache |
320 |
|
68ad4a330 mm/vmalloc.c: kee... |
321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 |
/* * This kmem_cache is used for vmap_area objects. Instead of * allocating from slab we reuse an object from this cache to * make things faster. Especially in "no edge" splitting of * free block. */ static struct kmem_cache *vmap_area_cachep; /* * This linked list is used in pair with free_vmap_area_root. * It gives O(1) access to prev/next to perform fast coalescing. */ static LIST_HEAD(free_vmap_area_list); /* * This augment red-black tree represents the free vmap space. * All vmap_area objects in this tree are sorted by va->va_start * address. It is used for allocation and merging when a vmap * object is released. * * Each vmap_area node contains a maximum available free block * of its sub-tree, right or left. Therefore it is possible to * find a lowest match of free area. */ static struct rb_root free_vmap_area_root = RB_ROOT; |
82dd23e84 mm/vmalloc.c: pre... |
346 347 348 349 350 351 |
/* * Preload a CPU with one object for "no edge" split case. The * aim is to get rid of allocations from the atomic context, thus * to use more permissive allocation masks. */ static DEFINE_PER_CPU(struct vmap_area *, ne_fit_preload_node); |
68ad4a330 mm/vmalloc.c: kee... |
352 353 354 355 356 357 358 359 360 361 362 363 364 365 |
static __always_inline unsigned long va_size(struct vmap_area *va) { return (va->va_end - va->va_start); } static __always_inline unsigned long get_subtree_max_size(struct rb_node *node) { struct vmap_area *va; va = rb_entry_safe(node, struct vmap_area, rb_node); return va ? va->subtree_max_size : 0; } |
89699605f mm: vmap area cache |
366 |
|
68ad4a330 mm/vmalloc.c: kee... |
367 368 369 370 371 372 373 374 375 376 |
/* * Gets called when remove the node and rotate. */ static __always_inline unsigned long compute_subtree_max_size(struct vmap_area *va) { return max3(va_size(va), get_subtree_max_size(va->rb_node.rb_left), get_subtree_max_size(va->rb_node.rb_right)); } |
315cc066b augmented rbtree:... |
377 378 |
RB_DECLARE_CALLBACKS_MAX(static, free_vmap_area_rb_augment_cb, struct vmap_area, rb_node, unsigned long, subtree_max_size, va_size) |
68ad4a330 mm/vmalloc.c: kee... |
379 380 381 382 |
static void purge_vmap_area_lazy(void); static BLOCKING_NOTIFIER_HEAD(vmap_notify_list); static unsigned long lazy_max_pages(void); |
db64fe022 mm: rewrite vmap ... |
383 |
|
97105f0ab mm: vmalloc: show... |
384 385 386 387 388 389 |
static atomic_long_t nr_vmalloc_pages; unsigned long vmalloc_nr_pages(void) { return atomic_long_read(&nr_vmalloc_pages); } |
db64fe022 mm: rewrite vmap ... |
390 |
static struct vmap_area *__find_vmap_area(unsigned long addr) |
1da177e4c Linux-2.6.12-rc2 |
391 |
{ |
db64fe022 mm: rewrite vmap ... |
392 393 394 395 396 397 398 399 |
struct rb_node *n = vmap_area_root.rb_node; while (n) { struct vmap_area *va; va = rb_entry(n, struct vmap_area, rb_node); if (addr < va->va_start) n = n->rb_left; |
cef2ac3f6 vmalloc: make fin... |
400 |
else if (addr >= va->va_end) |
db64fe022 mm: rewrite vmap ... |
401 402 403 404 405 406 407 |
n = n->rb_right; else return va; } return NULL; } |
68ad4a330 mm/vmalloc.c: kee... |
408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 |
/* * This function returns back addresses of parent node * and its left or right link for further processing. */ static __always_inline struct rb_node ** find_va_links(struct vmap_area *va, struct rb_root *root, struct rb_node *from, struct rb_node **parent) { struct vmap_area *tmp_va; struct rb_node **link; if (root) { link = &root->rb_node; if (unlikely(!*link)) { *parent = NULL; return link; } } else { link = &from; } |
db64fe022 mm: rewrite vmap ... |
429 |
|
68ad4a330 mm/vmalloc.c: kee... |
430 431 432 433 434 435 436 |
/* * Go to the bottom of the tree. When we hit the last point * we end up with parent rb_node and correct direction, i name * it link, where the new va->rb_node will be attached to. */ do { tmp_va = rb_entry(*link, struct vmap_area, rb_node); |
db64fe022 mm: rewrite vmap ... |
437 |
|
68ad4a330 mm/vmalloc.c: kee... |
438 439 440 441 442 443 444 445 446 447 448 |
/* * During the traversal we also do some sanity check. * Trigger the BUG() if there are sides(left/right) * or full overlaps. */ if (va->va_start < tmp_va->va_end && va->va_end <= tmp_va->va_start) link = &(*link)->rb_left; else if (va->va_end > tmp_va->va_start && va->va_start >= tmp_va->va_end) link = &(*link)->rb_right; |
db64fe022 mm: rewrite vmap ... |
449 450 |
else BUG(); |
68ad4a330 mm/vmalloc.c: kee... |
451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 |
} while (*link); *parent = &tmp_va->rb_node; return link; } static __always_inline struct list_head * get_va_next_sibling(struct rb_node *parent, struct rb_node **link) { struct list_head *list; if (unlikely(!parent)) /* * The red-black tree where we try to find VA neighbors * before merging or inserting is empty, i.e. it means * there is no free vmap space. Normally it does not * happen but we handle this case anyway. */ return NULL; list = &rb_entry(parent, struct vmap_area, rb_node)->list; return (&parent->rb_right == link ? list->next : list); } static __always_inline void link_va(struct vmap_area *va, struct rb_root *root, struct rb_node *parent, struct rb_node **link, struct list_head *head) { /* * VA is still not in the list, but we can * identify its future previous list_head node. */ if (likely(parent)) { head = &rb_entry(parent, struct vmap_area, rb_node)->list; if (&parent->rb_right != link) head = head->prev; |
db64fe022 mm: rewrite vmap ... |
487 |
} |
68ad4a330 mm/vmalloc.c: kee... |
488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 |
/* Insert to the rb-tree */ rb_link_node(&va->rb_node, parent, link); if (root == &free_vmap_area_root) { /* * Some explanation here. Just perform simple insertion * to the tree. We do not set va->subtree_max_size to * its current size before calling rb_insert_augmented(). * It is because of we populate the tree from the bottom * to parent levels when the node _is_ in the tree. * * Therefore we set subtree_max_size to zero after insertion, * to let __augment_tree_propagate_from() puts everything to * the correct order later on. */ rb_insert_augmented(&va->rb_node, root, &free_vmap_area_rb_augment_cb); va->subtree_max_size = 0; } else { rb_insert_color(&va->rb_node, root); } |
db64fe022 mm: rewrite vmap ... |
508 |
|
68ad4a330 mm/vmalloc.c: kee... |
509 510 |
/* Address-sort this list */ list_add(&va->list, head); |
db64fe022 mm: rewrite vmap ... |
511 |
} |
68ad4a330 mm/vmalloc.c: kee... |
512 513 514 |
static __always_inline void unlink_va(struct vmap_area *va, struct rb_root *root) { |
460e42d19 mm/vmalloc.c: swi... |
515 516 |
if (WARN_ON(RB_EMPTY_NODE(&va->rb_node))) return; |
db64fe022 mm: rewrite vmap ... |
517 |
|
460e42d19 mm/vmalloc.c: swi... |
518 519 520 521 522 523 524 525 |
if (root == &free_vmap_area_root) rb_erase_augmented(&va->rb_node, root, &free_vmap_area_rb_augment_cb); else rb_erase(&va->rb_node, root); list_del(&va->list); RB_CLEAR_NODE(&va->rb_node); |
68ad4a330 mm/vmalloc.c: kee... |
526 |
} |
bb850f4da mm/vmap: add DEBU... |
527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 |
#if DEBUG_AUGMENT_PROPAGATE_CHECK static void augment_tree_propagate_check(struct rb_node *n) { struct vmap_area *va; struct rb_node *node; unsigned long size; bool found = false; if (n == NULL) return; va = rb_entry(n, struct vmap_area, rb_node); size = va->subtree_max_size; node = n; while (node) { va = rb_entry(node, struct vmap_area, rb_node); if (get_subtree_max_size(node->rb_left) == size) { node = node->rb_left; } else { if (va_size(va) == size) { found = true; break; } node = node->rb_right; } } if (!found) { va = rb_entry(n, struct vmap_area, rb_node); pr_emerg("tree is corrupted: %lu, %lu ", va_size(va), va->subtree_max_size); } augment_tree_propagate_check(n->rb_left); augment_tree_propagate_check(n->rb_right); } #endif |
68ad4a330 mm/vmalloc.c: kee... |
569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 |
/* * This function populates subtree_max_size from bottom to upper * levels starting from VA point. The propagation must be done * when VA size is modified by changing its va_start/va_end. Or * in case of newly inserting of VA to the tree. * * It means that __augment_tree_propagate_from() must be called: * - After VA has been inserted to the tree(free path); * - After VA has been shrunk(allocation path); * - After VA has been increased(merging path). * * Please note that, it does not mean that upper parent nodes * and their subtree_max_size are recalculated all the time up * to the root node. * * 4--8 * /\ * / \ * / \ * 2--2 8--8 * * For example if we modify the node 4, shrinking it to 2, then * no any modification is required. If we shrink the node 2 to 1 * its subtree_max_size is updated only, and set to 1. If we shrink * the node 8 to 6, then its subtree_max_size is set to 6 and parent * node becomes 4--6. */ static __always_inline void augment_tree_propagate_from(struct vmap_area *va) { struct rb_node *node = &va->rb_node; unsigned long new_va_sub_max_size; while (node) { va = rb_entry(node, struct vmap_area, rb_node); new_va_sub_max_size = compute_subtree_max_size(va); /* * If the newly calculated maximum available size of the * subtree is equal to the current one, then it means that * the tree is propagated correctly. So we have to stop at * this point to save cycles. */ if (va->subtree_max_size == new_va_sub_max_size) break; va->subtree_max_size = new_va_sub_max_size; node = rb_parent(&va->rb_node); } |
bb850f4da mm/vmap: add DEBU... |
618 619 620 621 |
#if DEBUG_AUGMENT_PROPAGATE_CHECK augment_tree_propagate_check(free_vmap_area_root.rb_node); #endif |
68ad4a330 mm/vmalloc.c: kee... |
622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 |
} static void insert_vmap_area(struct vmap_area *va, struct rb_root *root, struct list_head *head) { struct rb_node **link; struct rb_node *parent; link = find_va_links(va, root, NULL, &parent); link_va(va, root, parent, link, head); } static void insert_vmap_area_augment(struct vmap_area *va, struct rb_node *from, struct rb_root *root, struct list_head *head) { struct rb_node **link; struct rb_node *parent; if (from) link = find_va_links(va, NULL, from, &parent); else link = find_va_links(va, root, NULL, &parent); link_va(va, root, parent, link, head); augment_tree_propagate_from(va); } /* * Merge de-allocated chunk of VA memory with previous * and next free blocks. If coalesce is not done a new * free area is inserted. If VA has been merged, it is * freed. */ static __always_inline void merge_or_add_vmap_area(struct vmap_area *va, struct rb_root *root, struct list_head *head) { struct vmap_area *sibling; struct list_head *next; struct rb_node **link; struct rb_node *parent; bool merged = false; /* * Find a place in the tree where VA potentially will be * inserted, unless it is merged with its sibling/siblings. */ link = find_va_links(va, root, NULL, &parent); /* * Get next node of VA to check if merging can be done. */ next = get_va_next_sibling(parent, link); if (unlikely(next == NULL)) goto insert; /* * start end * | | * |<------VA------>|<-----Next----->| * | | * start end */ if (next != head) { sibling = list_entry(next, struct vmap_area, list); if (sibling->va_start == va->va_end) { sibling->va_start = va->va_start; /* Check and update the tree if needed. */ augment_tree_propagate_from(sibling); |
68ad4a330 mm/vmalloc.c: kee... |
695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 |
/* Free vmap_area object. */ kmem_cache_free(vmap_area_cachep, va); /* Point to the new merged area. */ va = sibling; merged = true; } } /* * start end * | | * |<-----Prev----->|<------VA------>| * | | * start end */ if (next->prev != head) { sibling = list_entry(next->prev, struct vmap_area, list); if (sibling->va_end == va->va_start) { sibling->va_end = va->va_end; /* Check and update the tree if needed. */ augment_tree_propagate_from(sibling); |
54f63d9d8 mm/vmalloc.c: get... |
718 719 |
if (merged) unlink_va(va, root); |
68ad4a330 mm/vmalloc.c: kee... |
720 721 722 |
/* Free vmap_area object. */ kmem_cache_free(vmap_area_cachep, va); |
68ad4a330 mm/vmalloc.c: kee... |
723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 |
return; } } insert: if (!merged) { link_va(va, root, parent, link, head); augment_tree_propagate_from(va); } } static __always_inline bool is_within_this_va(struct vmap_area *va, unsigned long size, unsigned long align, unsigned long vstart) { unsigned long nva_start_addr; if (va->va_start > vstart) nva_start_addr = ALIGN(va->va_start, align); else nva_start_addr = ALIGN(vstart, align); /* Can be overflowed due to big size or alignment. */ if (nva_start_addr + size < nva_start_addr || nva_start_addr < vstart) return false; return (nva_start_addr + size <= va->va_end); } /* * Find the first free block(lowest start address) in the tree, * that will accomplish the request corresponding to passing * parameters. */ static __always_inline struct vmap_area * find_vmap_lowest_match(unsigned long size, unsigned long align, unsigned long vstart) { struct vmap_area *va; struct rb_node *node; unsigned long length; /* Start from the root. */ node = free_vmap_area_root.rb_node; /* Adjust the search size for alignment overhead. */ length = size + align - 1; while (node) { va = rb_entry(node, struct vmap_area, rb_node); if (get_subtree_max_size(node->rb_left) >= length && vstart < va->va_start) { node = node->rb_left; } else { if (is_within_this_va(va, size, align, vstart)) return va; /* * Does not make sense to go deeper towards the right * sub-tree if it does not have a free block that is * equal or bigger to the requested search length. */ if (get_subtree_max_size(node->rb_right) >= length) { node = node->rb_right; continue; } /* |
3806b0414 mm/vmalloc.c: fix... |
793 |
* OK. We roll back and find the first right sub-tree, |
68ad4a330 mm/vmalloc.c: kee... |
794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 |
* that will satisfy the search criteria. It can happen * only once due to "vstart" restriction. */ while ((node = rb_parent(node))) { va = rb_entry(node, struct vmap_area, rb_node); if (is_within_this_va(va, size, align, vstart)) return va; if (get_subtree_max_size(node->rb_right) >= length && vstart <= va->va_start) { node = node->rb_right; break; } } } } return NULL; } |
a6cf4e0fe mm/vmap: add DEBU... |
813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 |
#if DEBUG_AUGMENT_LOWEST_MATCH_CHECK #include <linux/random.h> static struct vmap_area * find_vmap_lowest_linear_match(unsigned long size, unsigned long align, unsigned long vstart) { struct vmap_area *va; list_for_each_entry(va, &free_vmap_area_list, list) { if (!is_within_this_va(va, size, align, vstart)) continue; return va; } return NULL; } static void find_vmap_lowest_match_check(unsigned long size) { struct vmap_area *va_1, *va_2; unsigned long vstart; unsigned int rnd; get_random_bytes(&rnd, sizeof(rnd)); vstart = VMALLOC_START + rnd; va_1 = find_vmap_lowest_match(size, 1, vstart); va_2 = find_vmap_lowest_linear_match(size, 1, vstart); if (va_1 != va_2) pr_emerg("not lowest: t: 0x%p, l: 0x%p, v: 0x%lx ", va_1, va_2, vstart); } #endif |
68ad4a330 mm/vmalloc.c: kee... |
851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 |
enum fit_type { NOTHING_FIT = 0, FL_FIT_TYPE = 1, /* full fit */ LE_FIT_TYPE = 2, /* left edge fit */ RE_FIT_TYPE = 3, /* right edge fit */ NE_FIT_TYPE = 4 /* no edge fit */ }; static __always_inline enum fit_type classify_va_fit_type(struct vmap_area *va, unsigned long nva_start_addr, unsigned long size) { enum fit_type type; /* Check if it is within VA. */ if (nva_start_addr < va->va_start || nva_start_addr + size > va->va_end) return NOTHING_FIT; /* Now classify. */ if (va->va_start == nva_start_addr) { if (va->va_end == nva_start_addr + size) type = FL_FIT_TYPE; else type = LE_FIT_TYPE; } else if (va->va_end == nva_start_addr + size) { type = RE_FIT_TYPE; } else { type = NE_FIT_TYPE; } return type; } static __always_inline int adjust_va_to_fit_type(struct vmap_area *va, unsigned long nva_start_addr, unsigned long size, enum fit_type type) { |
2c9292336 mm/vmalloc.c: avo... |
890 |
struct vmap_area *lva = NULL; |
68ad4a330 mm/vmalloc.c: kee... |
891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 |
if (type == FL_FIT_TYPE) { /* * No need to split VA, it fully fits. * * | | * V NVA V * |---------------| */ unlink_va(va, &free_vmap_area_root); kmem_cache_free(vmap_area_cachep, va); } else if (type == LE_FIT_TYPE) { /* * Split left edge of fit VA. * * | | * V NVA V R * |-------|-------| */ va->va_start += size; } else if (type == RE_FIT_TYPE) { /* * Split right edge of fit VA. * * | | * L V NVA V * |-------|-------| */ va->va_end = nva_start_addr; } else if (type == NE_FIT_TYPE) { /* * Split no edge of fit VA. * * | | * L V NVA V R * |---|-------|---| */ |
82dd23e84 mm/vmalloc.c: pre... |
928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 |
lva = __this_cpu_xchg(ne_fit_preload_node, NULL); if (unlikely(!lva)) { /* * For percpu allocator we do not do any pre-allocation * and leave it as it is. The reason is it most likely * never ends up with NE_FIT_TYPE splitting. In case of * percpu allocations offsets and sizes are aligned to * fixed align request, i.e. RE_FIT_TYPE and FL_FIT_TYPE * are its main fitting cases. * * There are a few exceptions though, as an example it is * a first allocation (early boot up) when we have "one" * big free space that has to be split. */ lva = kmem_cache_alloc(vmap_area_cachep, GFP_NOWAIT); if (!lva) return -1; } |
68ad4a330 mm/vmalloc.c: kee... |
946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 |
/* * Build the remainder. */ lva->va_start = va->va_start; lva->va_end = nva_start_addr; /* * Shrink this VA to remaining size. */ va->va_start = nva_start_addr + size; } else { return -1; } if (type != FL_FIT_TYPE) { augment_tree_propagate_from(va); |
2c9292336 mm/vmalloc.c: avo... |
963 |
if (lva) /* type == NE_FIT_TYPE */ |
68ad4a330 mm/vmalloc.c: kee... |
964 965 966 967 968 969 970 971 972 973 974 975 976 |
insert_vmap_area_augment(lva, &va->rb_node, &free_vmap_area_root, &free_vmap_area_list); } return 0; } /* * Returns a start address of the newly allocated area, if success. * Otherwise a vend is returned that indicates failure. */ static __always_inline unsigned long __alloc_vmap_area(unsigned long size, unsigned long align, |
cacca6baf mm/vmalloc.c: rem... |
977 |
unsigned long vstart, unsigned long vend) |
68ad4a330 mm/vmalloc.c: kee... |
978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 |
{ unsigned long nva_start_addr; struct vmap_area *va; enum fit_type type; int ret; va = find_vmap_lowest_match(size, align, vstart); if (unlikely(!va)) return vend; if (va->va_start > vstart) nva_start_addr = ALIGN(va->va_start, align); else nva_start_addr = ALIGN(vstart, align); /* Check the "vend" restriction. */ if (nva_start_addr + size > vend) return vend; /* Classify what we have found. */ type = classify_va_fit_type(va, nva_start_addr, size); if (WARN_ON_ONCE(type == NOTHING_FIT)) return vend; /* Update the free vmap_area. */ ret = adjust_va_to_fit_type(va, nva_start_addr, size, type); if (ret) return vend; |
a6cf4e0fe mm/vmap: add DEBU... |
1006 1007 1008 |
#if DEBUG_AUGMENT_LOWEST_MATCH_CHECK find_vmap_lowest_match_check(size); #endif |
68ad4a330 mm/vmalloc.c: kee... |
1009 1010 |
return nva_start_addr; } |
4da56b99d mm/vmap: Add a no... |
1011 |
|
db64fe022 mm: rewrite vmap ... |
1012 1013 1014 1015 1016 1017 1018 1019 1020 |
/* * Allocate a region of KVA of the specified size and alignment, within the * vstart and vend. */ static struct vmap_area *alloc_vmap_area(unsigned long size, unsigned long align, unsigned long vstart, unsigned long vend, int node, gfp_t gfp_mask) { |
82dd23e84 mm/vmalloc.c: pre... |
1021 |
struct vmap_area *va, *pva; |
1da177e4c Linux-2.6.12-rc2 |
1022 |
unsigned long addr; |
db64fe022 mm: rewrite vmap ... |
1023 |
int purged = 0; |
7766970cc mm: vmap fix over... |
1024 |
BUG_ON(!size); |
891c49abf mm/vmalloc: use o... |
1025 |
BUG_ON(offset_in_page(size)); |
89699605f mm: vmap area cache |
1026 |
BUG_ON(!is_power_of_2(align)); |
db64fe022 mm: rewrite vmap ... |
1027 |
|
68ad4a330 mm/vmalloc.c: kee... |
1028 1029 |
if (unlikely(!vmap_initialized)) return ERR_PTR(-EBUSY); |
5803ed292 mm: mark all call... |
1030 |
might_sleep(); |
4da56b99d mm/vmap: Add a no... |
1031 |
|
68ad4a330 mm/vmalloc.c: kee... |
1032 |
va = kmem_cache_alloc_node(vmap_area_cachep, |
db64fe022 mm: rewrite vmap ... |
1033 1034 1035 |
gfp_mask & GFP_RECLAIM_MASK, node); if (unlikely(!va)) return ERR_PTR(-ENOMEM); |
7f88f88f8 mm: kmemleak: avo... |
1036 1037 1038 1039 1040 |
/* * Only scan the relevant parts containing pointers to other objects * to avoid false negatives. */ kmemleak_scan_area(&va->rb_node, SIZE_MAX, gfp_mask & GFP_RECLAIM_MASK); |
db64fe022 mm: rewrite vmap ... |
1041 |
retry: |
82dd23e84 mm/vmalloc.c: pre... |
1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 |
/* * Preload this CPU with one extra vmap_area object to ensure * that we have it available when fit type of free area is * NE_FIT_TYPE. * * The preload is done in non-atomic context, thus it allows us * to use more permissive allocation masks to be more stable under * low memory condition and high memory pressure. * * Even if it fails we do not really care about that. Just proceed * as it is. "overflow" path will refill the cache we allocate from. */ preempt_disable(); if (!__this_cpu_read(ne_fit_preload_node)) { preempt_enable(); pva = kmem_cache_alloc_node(vmap_area_cachep, GFP_KERNEL, node); preempt_disable(); if (__this_cpu_cmpxchg(ne_fit_preload_node, NULL, pva)) { if (pva) kmem_cache_free(vmap_area_cachep, pva); } } |
db64fe022 mm: rewrite vmap ... |
1065 |
spin_lock(&vmap_area_lock); |
82dd23e84 mm/vmalloc.c: pre... |
1066 |
preempt_enable(); |
89699605f mm: vmap area cache |
1067 |
|
afd07389d mm/vmalloc.c: fix... |
1068 |
/* |
68ad4a330 mm/vmalloc.c: kee... |
1069 1070 |
* If an allocation fails, the "vend" address is * returned. Therefore trigger the overflow path. |
afd07389d mm/vmalloc.c: fix... |
1071 |
*/ |
cacca6baf mm/vmalloc.c: rem... |
1072 |
addr = __alloc_vmap_area(size, align, vstart, vend); |
68ad4a330 mm/vmalloc.c: kee... |
1073 |
if (unlikely(addr == vend)) |
89699605f mm: vmap area cache |
1074 |
goto overflow; |
db64fe022 mm: rewrite vmap ... |
1075 1076 1077 |
va->va_start = addr; va->va_end = addr + size; |
688fcbfc0 mm/vmalloc: modif... |
1078 |
va->vm = NULL; |
68ad4a330 mm/vmalloc.c: kee... |
1079 |
insert_vmap_area(va, &vmap_area_root, &vmap_area_list); |
db64fe022 mm: rewrite vmap ... |
1080 |
spin_unlock(&vmap_area_lock); |
61e165578 mm/vmalloc.c: use... |
1081 |
BUG_ON(!IS_ALIGNED(va->va_start, align)); |
89699605f mm: vmap area cache |
1082 1083 |
BUG_ON(va->va_start < vstart); BUG_ON(va->va_end > vend); |
db64fe022 mm: rewrite vmap ... |
1084 |
return va; |
89699605f mm: vmap area cache |
1085 1086 1087 1088 1089 1090 1091 1092 |
overflow: spin_unlock(&vmap_area_lock); if (!purged) { purge_vmap_area_lazy(); purged = 1; goto retry; } |
4da56b99d mm/vmap: Add a no... |
1093 1094 1095 1096 1097 1098 1099 1100 1101 |
if (gfpflags_allow_blocking(gfp_mask)) { unsigned long freed = 0; blocking_notifier_call_chain(&vmap_notify_list, 0, &freed); if (freed > 0) { purged = 0; goto retry; } } |
03497d761 mm: Silence vmap(... |
1102 |
if (!(gfp_mask & __GFP_NOWARN) && printk_ratelimit()) |
756a025f0 mm: coalesce spli... |
1103 1104 1105 |
pr_warn("vmap allocation for size %lu failed: use vmalloc=<size> to increase size ", size); |
68ad4a330 mm/vmalloc.c: kee... |
1106 1107 |
kmem_cache_free(vmap_area_cachep, va); |
89699605f mm: vmap area cache |
1108 |
return ERR_PTR(-EBUSY); |
db64fe022 mm: rewrite vmap ... |
1109 |
} |
4da56b99d mm/vmap: Add a no... |
1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120 |
int register_vmap_purge_notifier(struct notifier_block *nb) { return blocking_notifier_chain_register(&vmap_notify_list, nb); } EXPORT_SYMBOL_GPL(register_vmap_purge_notifier); int unregister_vmap_purge_notifier(struct notifier_block *nb) { return blocking_notifier_chain_unregister(&vmap_notify_list, nb); } EXPORT_SYMBOL_GPL(unregister_vmap_purge_notifier); |
db64fe022 mm: rewrite vmap ... |
1121 1122 |
static void __free_vmap_area(struct vmap_area *va) { |
ca23e405e vmalloc: implemen... |
1123 |
/* |
68ad4a330 mm/vmalloc.c: kee... |
1124 |
* Remove from the busy tree/list. |
ca23e405e vmalloc: implemen... |
1125 |
*/ |
68ad4a330 mm/vmalloc.c: kee... |
1126 |
unlink_va(va, &vmap_area_root); |
ca23e405e vmalloc: implemen... |
1127 |
|
68ad4a330 mm/vmalloc.c: kee... |
1128 1129 1130 1131 1132 |
/* * Merge VA with its neighbors, otherwise just add it. */ merge_or_add_vmap_area(va, &free_vmap_area_root, &free_vmap_area_list); |
db64fe022 mm: rewrite vmap ... |
1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151 1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 |
} /* * Free a region of KVA allocated by alloc_vmap_area */ static void free_vmap_area(struct vmap_area *va) { spin_lock(&vmap_area_lock); __free_vmap_area(va); spin_unlock(&vmap_area_lock); } /* * Clear the pagetable entries of a given vmap_area */ static void unmap_vmap_area(struct vmap_area *va) { vunmap_page_range(va->va_start, va->va_end); } /* * lazy_max_pages is the maximum amount of virtual address space we gather up * before attempting to purge with a TLB flush. * * There is a tradeoff here: a larger number will cover more kernel page tables * and take slightly longer to purge, but it will linearly reduce the number of * global TLB flushes that must be performed. It would seem natural to scale * this number up linearly with the number of CPUs (because vmapping activity * could also scale linearly with the number of CPUs), however it is likely * that in practice, workloads might be constrained in other ways that mean * vmap activity will not scale linearly with CPUs. Also, I want to be * conservative and not introduce a big latency on huge systems, so go with * a less aggressive log scale. It will still be an improvement over the old * code, and it will be simple to change the scale factor if we find that it * becomes a problem on bigger systems. */ static unsigned long lazy_max_pages(void) { unsigned int log; log = fls(num_online_cpus()); return log * (32UL * 1024 * 1024 / PAGE_SIZE); } |
4d36e6f80 mm/vmalloc.c: con... |
1177 |
static atomic_long_t vmap_lazy_nr = ATOMIC_LONG_INIT(0); |
db64fe022 mm: rewrite vmap ... |
1178 |
|
0574ecd14 mm: refactor __pu... |
1179 1180 1181 1182 1183 |
/* * Serialize vmap purging. There is no actual criticial section protected * by this look, but we want to avoid concurrent calls for performance * reasons and to make the pcpu_get_vm_areas more deterministic. */ |
f9e099776 mm: turn vmap_pur... |
1184 |
static DEFINE_MUTEX(vmap_purge_lock); |
0574ecd14 mm: refactor __pu... |
1185 |
|
02b709df8 mm: purge fragmen... |
1186 1187 |
/* for per-CPU blocks */ static void purge_fragmented_blocks_allcpus(void); |
db64fe022 mm: rewrite vmap ... |
1188 |
/* |
3ee48b6af mm, x86: Saving v... |
1189 1190 1191 1192 1193 |
* called before a call to iounmap() if the caller wants vm_area_struct's * immediately freed. */ void set_iounmap_nonlazy(void) { |
4d36e6f80 mm/vmalloc.c: con... |
1194 |
atomic_long_set(&vmap_lazy_nr, lazy_max_pages()+1); |
3ee48b6af mm, x86: Saving v... |
1195 1196 1197 |
} /* |
db64fe022 mm: rewrite vmap ... |
1198 |
* Purges all lazily-freed vmap areas. |
db64fe022 mm: rewrite vmap ... |
1199 |
*/ |
0574ecd14 mm: refactor __pu... |
1200 |
static bool __purge_vmap_area_lazy(unsigned long start, unsigned long end) |
db64fe022 mm: rewrite vmap ... |
1201 |
{ |
4d36e6f80 mm/vmalloc.c: con... |
1202 |
unsigned long resched_threshold; |
80c4bd7a5 mm/vmalloc: keep ... |
1203 |
struct llist_node *valist; |
db64fe022 mm: rewrite vmap ... |
1204 |
struct vmap_area *va; |
cbb766766 mm: fix lazy vmap... |
1205 |
struct vmap_area *n_va; |
db64fe022 mm: rewrite vmap ... |
1206 |
|
0574ecd14 mm: refactor __pu... |
1207 |
lockdep_assert_held(&vmap_purge_lock); |
02b709df8 mm: purge fragmen... |
1208 |
|
80c4bd7a5 mm/vmalloc: keep ... |
1209 |
valist = llist_del_all(&vmap_purge_list); |
68571be99 mm/vmalloc.c: add... |
1210 1211 1212 1213 |
if (unlikely(valist == NULL)) return false; /* |
3f8fd02b1 mm/vmalloc: Sync ... |
1214 1215 1216 1217 1218 1219 |
* First make sure the mappings are removed from all page-tables * before they are freed. */ vmalloc_sync_all(); /* |
68571be99 mm/vmalloc.c: add... |
1220 1221 1222 |
* TODO: to calculate a flush range without looping. * The list can be up to lazy_max_pages() elements. */ |
80c4bd7a5 mm/vmalloc: keep ... |
1223 |
llist_for_each_entry(va, valist, purge_list) { |
0574ecd14 mm: refactor __pu... |
1224 1225 1226 1227 |
if (va->va_start < start) start = va->va_start; if (va->va_end > end) end = va->va_end; |
db64fe022 mm: rewrite vmap ... |
1228 |
} |
db64fe022 mm: rewrite vmap ... |
1229 |
|
0574ecd14 mm: refactor __pu... |
1230 |
flush_tlb_kernel_range(start, end); |
4d36e6f80 mm/vmalloc.c: con... |
1231 |
resched_threshold = lazy_max_pages() << 1; |
db64fe022 mm: rewrite vmap ... |
1232 |
|
0574ecd14 mm: refactor __pu... |
1233 |
spin_lock(&vmap_area_lock); |
763b218dd mm: add preempt p... |
1234 |
llist_for_each_entry_safe(va, n_va, valist, purge_list) { |
4d36e6f80 mm/vmalloc.c: con... |
1235 |
unsigned long nr = (va->va_end - va->va_start) >> PAGE_SHIFT; |
763b218dd mm: add preempt p... |
1236 |
|
dd3b8353b mm/vmalloc: do no... |
1237 1238 1239 1240 1241 1242 1243 |
/* * Finally insert or merge lazily-freed area. It is * detached and there is no need to "unlink" it from * anything. */ merge_or_add_vmap_area(va, &free_vmap_area_root, &free_vmap_area_list); |
4d36e6f80 mm/vmalloc.c: con... |
1244 |
atomic_long_sub(nr, &vmap_lazy_nr); |
68571be99 mm/vmalloc.c: add... |
1245 |
|
4d36e6f80 mm/vmalloc.c: con... |
1246 |
if (atomic_long_read(&vmap_lazy_nr) < resched_threshold) |
68571be99 mm/vmalloc.c: add... |
1247 |
cond_resched_lock(&vmap_area_lock); |
763b218dd mm: add preempt p... |
1248 |
} |
0574ecd14 mm: refactor __pu... |
1249 1250 |
spin_unlock(&vmap_area_lock); return true; |
db64fe022 mm: rewrite vmap ... |
1251 1252 1253 |
} /* |
496850e5f mm: vmalloc failu... |
1254 1255 1256 1257 1258 |
* Kick off a purge of the outstanding lazy areas. Don't bother if somebody * is already purging. */ static void try_purge_vmap_area_lazy(void) { |
f9e099776 mm: turn vmap_pur... |
1259 |
if (mutex_trylock(&vmap_purge_lock)) { |
0574ecd14 mm: refactor __pu... |
1260 |
__purge_vmap_area_lazy(ULONG_MAX, 0); |
f9e099776 mm: turn vmap_pur... |
1261 |
mutex_unlock(&vmap_purge_lock); |
0574ecd14 mm: refactor __pu... |
1262 |
} |
496850e5f mm: vmalloc failu... |
1263 1264 1265 |
} /* |
db64fe022 mm: rewrite vmap ... |
1266 1267 1268 1269 |
* Kick off a purge of the outstanding lazy areas. */ static void purge_vmap_area_lazy(void) { |
f9e099776 mm: turn vmap_pur... |
1270 |
mutex_lock(&vmap_purge_lock); |
0574ecd14 mm: refactor __pu... |
1271 1272 |
purge_fragmented_blocks_allcpus(); __purge_vmap_area_lazy(ULONG_MAX, 0); |
f9e099776 mm: turn vmap_pur... |
1273 |
mutex_unlock(&vmap_purge_lock); |
db64fe022 mm: rewrite vmap ... |
1274 1275 1276 |
} /* |
64141da58 vmalloc: eagerly ... |
1277 1278 1279 |
* Free a vmap area, caller ensuring that the area has been unmapped * and flush_cache_vunmap had been called for the correct range * previously. |
db64fe022 mm: rewrite vmap ... |
1280 |
*/ |
64141da58 vmalloc: eagerly ... |
1281 |
static void free_vmap_area_noflush(struct vmap_area *va) |
db64fe022 mm: rewrite vmap ... |
1282 |
{ |
4d36e6f80 mm/vmalloc.c: con... |
1283 |
unsigned long nr_lazy; |
80c4bd7a5 mm/vmalloc: keep ... |
1284 |
|
dd3b8353b mm/vmalloc: do no... |
1285 1286 1287 |
spin_lock(&vmap_area_lock); unlink_va(va, &vmap_area_root); spin_unlock(&vmap_area_lock); |
4d36e6f80 mm/vmalloc.c: con... |
1288 1289 |
nr_lazy = atomic_long_add_return((va->va_end - va->va_start) >> PAGE_SHIFT, &vmap_lazy_nr); |
80c4bd7a5 mm/vmalloc: keep ... |
1290 1291 1292 1293 1294 |
/* After this point, we may free va at any time */ llist_add(&va->purge_list, &vmap_purge_list); if (unlikely(nr_lazy > lazy_max_pages())) |
496850e5f mm: vmalloc failu... |
1295 |
try_purge_vmap_area_lazy(); |
db64fe022 mm: rewrite vmap ... |
1296 |
} |
b29acbdcf mm: vmalloc fix l... |
1297 1298 1299 1300 1301 1302 |
/* * Free and unmap a vmap area */ static void free_unmap_vmap_area(struct vmap_area *va) { flush_cache_vunmap(va->va_start, va->va_end); |
c8eef01e2 mm: remove free_u... |
1303 |
unmap_vmap_area(va); |
d30dce351 mm, debug_pageall... |
1304 |
if (debug_pagealloc_enabled_static()) |
82a2e924f mm: vmalloc: clea... |
1305 |
flush_tlb_kernel_range(va->va_start, va->va_end); |
c8eef01e2 mm: remove free_u... |
1306 |
free_vmap_area_noflush(va); |
b29acbdcf mm: vmalloc fix l... |
1307 |
} |
db64fe022 mm: rewrite vmap ... |
1308 1309 1310 1311 1312 1313 1314 1315 1316 1317 |
static struct vmap_area *find_vmap_area(unsigned long addr) { struct vmap_area *va; spin_lock(&vmap_area_lock); va = __find_vmap_area(addr); spin_unlock(&vmap_area_lock); return va; } |
db64fe022 mm: rewrite vmap ... |
1318 1319 1320 1321 1322 1323 1324 1325 1326 1327 1328 1329 1330 1331 1332 1333 1334 1335 1336 1337 1338 1339 1340 |
/*** Per cpu kva allocator ***/ /* * vmap space is limited especially on 32 bit architectures. Ensure there is * room for at least 16 percpu vmap blocks per CPU. */ /* * If we had a constant VMALLOC_START and VMALLOC_END, we'd like to be able * to #define VMALLOC_SPACE (VMALLOC_END-VMALLOC_START). Guess * instead (we just need a rough idea) */ #if BITS_PER_LONG == 32 #define VMALLOC_SPACE (128UL*1024*1024) #else #define VMALLOC_SPACE (128UL*1024*1024*1024) #endif #define VMALLOC_PAGES (VMALLOC_SPACE / PAGE_SIZE) #define VMAP_MAX_ALLOC BITS_PER_LONG /* 256K with 4K pages */ #define VMAP_BBMAP_BITS_MAX 1024 /* 4MB with 4K pages */ #define VMAP_BBMAP_BITS_MIN (VMAP_MAX_ALLOC*2) #define VMAP_MIN(x, y) ((x) < (y) ? (x) : (y)) /* can't use min() */ #define VMAP_MAX(x, y) ((x) > (y) ? (x) : (y)) /* can't use max() */ |
f982f9151 mm: fix wrong vma... |
1341 1342 1343 1344 |
#define VMAP_BBMAP_BITS \ VMAP_MIN(VMAP_BBMAP_BITS_MAX, \ VMAP_MAX(VMAP_BBMAP_BITS_MIN, \ VMALLOC_PAGES / roundup_pow_of_two(NR_CPUS) / 16)) |
db64fe022 mm: rewrite vmap ... |
1345 1346 1347 1348 1349 1350 |
#define VMAP_BLOCK_SIZE (VMAP_BBMAP_BITS * PAGE_SIZE) struct vmap_block_queue { spinlock_t lock; struct list_head free; |
db64fe022 mm: rewrite vmap ... |
1351 1352 1353 1354 1355 |
}; struct vmap_block { spinlock_t lock; struct vmap_area *va; |
db64fe022 mm: rewrite vmap ... |
1356 |
unsigned long free, dirty; |
7d61bfe8f mm/vmalloc: get r... |
1357 |
unsigned long dirty_min, dirty_max; /*< dirty range */ |
de5604231 mm: percpu-vmap f... |
1358 1359 |
struct list_head free_list; struct rcu_head rcu_head; |
02b709df8 mm: purge fragmen... |
1360 |
struct list_head purge; |
db64fe022 mm: rewrite vmap ... |
1361 1362 1363 1364 1365 1366 1367 1368 1369 1370 1371 1372 1373 1374 1375 1376 1377 1378 1379 1380 1381 1382 1383 1384 1385 1386 |
}; /* Queue of free and dirty vmap blocks, for allocation and flushing purposes */ static DEFINE_PER_CPU(struct vmap_block_queue, vmap_block_queue); /* * Radix tree of vmap blocks, indexed by address, to quickly find a vmap block * in the free path. Could get rid of this if we change the API to return a * "cookie" from alloc, to be passed to free. But no big deal yet. */ static DEFINE_SPINLOCK(vmap_block_tree_lock); static RADIX_TREE(vmap_block_tree, GFP_ATOMIC); /* * We should probably have a fallback mechanism to allocate virtual memory * out of partially filled vmap blocks. However vmap block sizing should be * fairly reasonable according to the vmalloc size, so it shouldn't be a * big problem. */ static unsigned long addr_to_vb_idx(unsigned long addr) { addr -= VMALLOC_START & ~(VMAP_BLOCK_SIZE-1); addr /= VMAP_BLOCK_SIZE; return addr; } |
cf725ce27 mm/vmalloc: occup... |
1387 1388 1389 1390 1391 1392 1393 1394 1395 1396 1397 1398 1399 1400 1401 |
static void *vmap_block_vaddr(unsigned long va_start, unsigned long pages_off) { unsigned long addr; addr = va_start + (pages_off << PAGE_SHIFT); BUG_ON(addr_to_vb_idx(addr) != addr_to_vb_idx(va_start)); return (void *)addr; } /** * new_vmap_block - allocates new vmap_block and occupies 2^order pages in this * block. Of course pages number can't exceed VMAP_BBMAP_BITS * @order: how many 2^order pages should be occupied in newly allocated block * @gfp_mask: flags for the page level allocator * |
a862f68a8 docs/core-api/mm:... |
1402 |
* Return: virtual address in a newly allocated block or ERR_PTR(-errno) |
cf725ce27 mm/vmalloc: occup... |
1403 1404 |
*/ static void *new_vmap_block(unsigned int order, gfp_t gfp_mask) |
db64fe022 mm: rewrite vmap ... |
1405 1406 1407 1408 1409 1410 |
{ struct vmap_block_queue *vbq; struct vmap_block *vb; struct vmap_area *va; unsigned long vb_idx; int node, err; |
cf725ce27 mm/vmalloc: occup... |
1411 |
void *vaddr; |
db64fe022 mm: rewrite vmap ... |
1412 1413 1414 1415 1416 1417 1418 1419 1420 1421 1422 |
node = numa_node_id(); vb = kmalloc_node(sizeof(struct vmap_block), gfp_mask & GFP_RECLAIM_MASK, node); if (unlikely(!vb)) return ERR_PTR(-ENOMEM); va = alloc_vmap_area(VMAP_BLOCK_SIZE, VMAP_BLOCK_SIZE, VMALLOC_START, VMALLOC_END, node, gfp_mask); |
ddf9c6d47 vmalloc: remove r... |
1423 |
if (IS_ERR(va)) { |
db64fe022 mm: rewrite vmap ... |
1424 |
kfree(vb); |
e7d863407 mm: use ERR_CAST |
1425 |
return ERR_CAST(va); |
db64fe022 mm: rewrite vmap ... |
1426 1427 1428 1429 1430 1431 1432 1433 |
} err = radix_tree_preload(gfp_mask); if (unlikely(err)) { kfree(vb); free_vmap_area(va); return ERR_PTR(err); } |
cf725ce27 mm/vmalloc: occup... |
1434 |
vaddr = vmap_block_vaddr(va->va_start, 0); |
db64fe022 mm: rewrite vmap ... |
1435 1436 |
spin_lock_init(&vb->lock); vb->va = va; |
cf725ce27 mm/vmalloc: occup... |
1437 1438 1439 |
/* At least something should be left free */ BUG_ON(VMAP_BBMAP_BITS <= (1UL << order)); vb->free = VMAP_BBMAP_BITS - (1UL << order); |
db64fe022 mm: rewrite vmap ... |
1440 |
vb->dirty = 0; |
7d61bfe8f mm/vmalloc: get r... |
1441 1442 |
vb->dirty_min = VMAP_BBMAP_BITS; vb->dirty_max = 0; |
db64fe022 mm: rewrite vmap ... |
1443 |
INIT_LIST_HEAD(&vb->free_list); |
db64fe022 mm: rewrite vmap ... |
1444 1445 1446 1447 1448 1449 1450 1451 1452 |
vb_idx = addr_to_vb_idx(va->va_start); spin_lock(&vmap_block_tree_lock); err = radix_tree_insert(&vmap_block_tree, vb_idx, vb); spin_unlock(&vmap_block_tree_lock); BUG_ON(err); radix_tree_preload_end(); vbq = &get_cpu_var(vmap_block_queue); |
db64fe022 mm: rewrite vmap ... |
1453 |
spin_lock(&vbq->lock); |
68ac546f2 mm/vmalloc: fix p... |
1454 |
list_add_tail_rcu(&vb->free_list, &vbq->free); |
db64fe022 mm: rewrite vmap ... |
1455 |
spin_unlock(&vbq->lock); |
3f04ba859 vmalloc: fix use ... |
1456 |
put_cpu_var(vmap_block_queue); |
db64fe022 mm: rewrite vmap ... |
1457 |
|
cf725ce27 mm/vmalloc: occup... |
1458 |
return vaddr; |
db64fe022 mm: rewrite vmap ... |
1459 |
} |
db64fe022 mm: rewrite vmap ... |
1460 1461 1462 1463 |
static void free_vmap_block(struct vmap_block *vb) { struct vmap_block *tmp; unsigned long vb_idx; |
db64fe022 mm: rewrite vmap ... |
1464 1465 1466 1467 1468 |
vb_idx = addr_to_vb_idx(vb->va->va_start); spin_lock(&vmap_block_tree_lock); tmp = radix_tree_delete(&vmap_block_tree, vb_idx); spin_unlock(&vmap_block_tree_lock); BUG_ON(tmp != vb); |
64141da58 vmalloc: eagerly ... |
1469 |
free_vmap_area_noflush(vb->va); |
22a3c7d18 vmalloc,rcu: Conv... |
1470 |
kfree_rcu(vb, rcu_head); |
db64fe022 mm: rewrite vmap ... |
1471 |
} |
02b709df8 mm: purge fragmen... |
1472 1473 1474 1475 1476 1477 1478 1479 1480 1481 1482 1483 1484 1485 1486 1487 1488 |
static void purge_fragmented_blocks(int cpu) { LIST_HEAD(purge); struct vmap_block *vb; struct vmap_block *n_vb; struct vmap_block_queue *vbq = &per_cpu(vmap_block_queue, cpu); rcu_read_lock(); list_for_each_entry_rcu(vb, &vbq->free, free_list) { if (!(vb->free + vb->dirty == VMAP_BBMAP_BITS && vb->dirty != VMAP_BBMAP_BITS)) continue; spin_lock(&vb->lock); if (vb->free + vb->dirty == VMAP_BBMAP_BITS && vb->dirty != VMAP_BBMAP_BITS) { vb->free = 0; /* prevent further allocs after releasing lock */ vb->dirty = VMAP_BBMAP_BITS; /* prevent purging it again */ |
7d61bfe8f mm/vmalloc: get r... |
1489 1490 |
vb->dirty_min = 0; vb->dirty_max = VMAP_BBMAP_BITS; |
02b709df8 mm: purge fragmen... |
1491 1492 1493 1494 1495 1496 1497 1498 1499 1500 1501 1502 1503 1504 1505 |
spin_lock(&vbq->lock); list_del_rcu(&vb->free_list); spin_unlock(&vbq->lock); spin_unlock(&vb->lock); list_add_tail(&vb->purge, &purge); } else spin_unlock(&vb->lock); } rcu_read_unlock(); list_for_each_entry_safe(vb, n_vb, &purge, purge) { list_del(&vb->purge); free_vmap_block(vb); } } |
02b709df8 mm: purge fragmen... |
1506 1507 1508 1509 1510 1511 1512 |
static void purge_fragmented_blocks_allcpus(void) { int cpu; for_each_possible_cpu(cpu) purge_fragmented_blocks(cpu); } |
db64fe022 mm: rewrite vmap ... |
1513 1514 1515 1516 |
static void *vb_alloc(unsigned long size, gfp_t gfp_mask) { struct vmap_block_queue *vbq; struct vmap_block *vb; |
cf725ce27 mm/vmalloc: occup... |
1517 |
void *vaddr = NULL; |
db64fe022 mm: rewrite vmap ... |
1518 |
unsigned int order; |
891c49abf mm/vmalloc: use o... |
1519 |
BUG_ON(offset_in_page(size)); |
db64fe022 mm: rewrite vmap ... |
1520 |
BUG_ON(size > PAGE_SIZE*VMAP_MAX_ALLOC); |
aa91c4d89 mm: make vb_alloc... |
1521 1522 1523 1524 1525 1526 1527 1528 |
if (WARN_ON(size == 0)) { /* * Allocating 0 bytes isn't what caller wants since * get_order(0) returns funny result. Just warn and terminate * early. */ return NULL; } |
db64fe022 mm: rewrite vmap ... |
1529 |
order = get_order(size); |
db64fe022 mm: rewrite vmap ... |
1530 1531 1532 |
rcu_read_lock(); vbq = &get_cpu_var(vmap_block_queue); list_for_each_entry_rcu(vb, &vbq->free, free_list) { |
cf725ce27 mm/vmalloc: occup... |
1533 |
unsigned long pages_off; |
db64fe022 mm: rewrite vmap ... |
1534 1535 |
spin_lock(&vb->lock); |
cf725ce27 mm/vmalloc: occup... |
1536 1537 1538 1539 |
if (vb->free < (1UL << order)) { spin_unlock(&vb->lock); continue; } |
02b709df8 mm: purge fragmen... |
1540 |
|
cf725ce27 mm/vmalloc: occup... |
1541 1542 |
pages_off = VMAP_BBMAP_BITS - vb->free; vaddr = vmap_block_vaddr(vb->va->va_start, pages_off); |
02b709df8 mm: purge fragmen... |
1543 1544 1545 1546 1547 1548 |
vb->free -= 1UL << order; if (vb->free == 0) { spin_lock(&vbq->lock); list_del_rcu(&vb->free_list); spin_unlock(&vbq->lock); } |
cf725ce27 mm/vmalloc: occup... |
1549 |
|
02b709df8 mm: purge fragmen... |
1550 1551 |
spin_unlock(&vb->lock); break; |
db64fe022 mm: rewrite vmap ... |
1552 |
} |
02b709df8 mm: purge fragmen... |
1553 |
|
3f04ba859 vmalloc: fix use ... |
1554 |
put_cpu_var(vmap_block_queue); |
db64fe022 mm: rewrite vmap ... |
1555 |
rcu_read_unlock(); |
cf725ce27 mm/vmalloc: occup... |
1556 1557 1558 |
/* Allocate new block if nothing was found */ if (!vaddr) vaddr = new_vmap_block(order, gfp_mask); |
db64fe022 mm: rewrite vmap ... |
1559 |
|
cf725ce27 mm/vmalloc: occup... |
1560 |
return vaddr; |
db64fe022 mm: rewrite vmap ... |
1561 1562 1563 1564 1565 1566 1567 1568 |
} static void vb_free(const void *addr, unsigned long size) { unsigned long offset; unsigned long vb_idx; unsigned int order; struct vmap_block *vb; |
891c49abf mm/vmalloc: use o... |
1569 |
BUG_ON(offset_in_page(size)); |
db64fe022 mm: rewrite vmap ... |
1570 |
BUG_ON(size > PAGE_SIZE*VMAP_MAX_ALLOC); |
b29acbdcf mm: vmalloc fix l... |
1571 1572 |
flush_cache_vunmap((unsigned long)addr, (unsigned long)addr + size); |
db64fe022 mm: rewrite vmap ... |
1573 1574 1575 |
order = get_order(size); offset = (unsigned long)addr & (VMAP_BLOCK_SIZE - 1); |
7d61bfe8f mm/vmalloc: get r... |
1576 |
offset >>= PAGE_SHIFT; |
db64fe022 mm: rewrite vmap ... |
1577 1578 1579 1580 1581 1582 |
vb_idx = addr_to_vb_idx((unsigned long)addr); rcu_read_lock(); vb = radix_tree_lookup(&vmap_block_tree, vb_idx); rcu_read_unlock(); BUG_ON(!vb); |
64141da58 vmalloc: eagerly ... |
1583 |
vunmap_page_range((unsigned long)addr, (unsigned long)addr + size); |
d30dce351 mm, debug_pageall... |
1584 |
if (debug_pagealloc_enabled_static()) |
82a2e924f mm: vmalloc: clea... |
1585 1586 |
flush_tlb_kernel_range((unsigned long)addr, (unsigned long)addr + size); |
db64fe022 mm: rewrite vmap ... |
1587 |
spin_lock(&vb->lock); |
7d61bfe8f mm/vmalloc: get r... |
1588 1589 1590 1591 |
/* Expand dirty range */ vb->dirty_min = min(vb->dirty_min, offset); vb->dirty_max = max(vb->dirty_max, offset + (1UL << order)); |
d086817dc vmap: remove need... |
1592 |
|
db64fe022 mm: rewrite vmap ... |
1593 1594 |
vb->dirty += 1UL << order; if (vb->dirty == VMAP_BBMAP_BITS) { |
de5604231 mm: percpu-vmap f... |
1595 |
BUG_ON(vb->free); |
db64fe022 mm: rewrite vmap ... |
1596 1597 1598 1599 1600 |
spin_unlock(&vb->lock); free_vmap_block(vb); } else spin_unlock(&vb->lock); } |
868b104d7 mm/vmalloc: Add f... |
1601 |
static void _vm_unmap_aliases(unsigned long start, unsigned long end, int flush) |
db64fe022 mm: rewrite vmap ... |
1602 |
{ |
db64fe022 mm: rewrite vmap ... |
1603 |
int cpu; |
db64fe022 mm: rewrite vmap ... |
1604 |
|
9b4633340 vmap: cope with v... |
1605 1606 |
if (unlikely(!vmap_initialized)) return; |
5803ed292 mm: mark all call... |
1607 |
might_sleep(); |
db64fe022 mm: rewrite vmap ... |
1608 1609 1610 1611 1612 1613 |
for_each_possible_cpu(cpu) { struct vmap_block_queue *vbq = &per_cpu(vmap_block_queue, cpu); struct vmap_block *vb; rcu_read_lock(); list_for_each_entry_rcu(vb, &vbq->free, free_list) { |
db64fe022 mm: rewrite vmap ... |
1614 |
spin_lock(&vb->lock); |
7d61bfe8f mm/vmalloc: get r... |
1615 1616 |
if (vb->dirty) { unsigned long va_start = vb->va->va_start; |
db64fe022 mm: rewrite vmap ... |
1617 |
unsigned long s, e; |
b136be5e0 mm, vmalloc: use ... |
1618 |
|
7d61bfe8f mm/vmalloc: get r... |
1619 1620 |
s = va_start + (vb->dirty_min << PAGE_SHIFT); e = va_start + (vb->dirty_max << PAGE_SHIFT); |
db64fe022 mm: rewrite vmap ... |
1621 |
|
7d61bfe8f mm/vmalloc: get r... |
1622 1623 |
start = min(s, start); end = max(e, end); |
db64fe022 mm: rewrite vmap ... |
1624 |
|
7d61bfe8f mm/vmalloc: get r... |
1625 |
flush = 1; |
db64fe022 mm: rewrite vmap ... |
1626 1627 1628 1629 1630 |
} spin_unlock(&vb->lock); } rcu_read_unlock(); } |
f9e099776 mm: turn vmap_pur... |
1631 |
mutex_lock(&vmap_purge_lock); |
0574ecd14 mm: refactor __pu... |
1632 1633 1634 |
purge_fragmented_blocks_allcpus(); if (!__purge_vmap_area_lazy(start, end) && flush) flush_tlb_kernel_range(start, end); |
f9e099776 mm: turn vmap_pur... |
1635 |
mutex_unlock(&vmap_purge_lock); |
db64fe022 mm: rewrite vmap ... |
1636 |
} |
868b104d7 mm/vmalloc: Add f... |
1637 1638 1639 1640 1641 1642 1643 1644 1645 1646 1647 1648 1649 1650 1651 1652 1653 1654 1655 1656 1657 |
/** * vm_unmap_aliases - unmap outstanding lazy aliases in the vmap layer * * The vmap/vmalloc layer lazily flushes kernel virtual mappings primarily * to amortize TLB flushing overheads. What this means is that any page you * have now, may, in a former life, have been mapped into kernel virtual * address by the vmap layer and so there might be some CPUs with TLB entries * still referencing that page (additional to the regular 1:1 kernel mapping). * * vm_unmap_aliases flushes all such lazy mappings. After it returns, we can * be sure that none of the pages we have control over will have any aliases * from the vmap layer. */ void vm_unmap_aliases(void) { unsigned long start = ULONG_MAX, end = 0; int flush = 0; _vm_unmap_aliases(start, end, flush); } |
db64fe022 mm: rewrite vmap ... |
1658 1659 1660 1661 1662 1663 1664 1665 1666 |
EXPORT_SYMBOL_GPL(vm_unmap_aliases); /** * vm_unmap_ram - unmap linear kernel address space set up by vm_map_ram * @mem: the pointer returned by vm_map_ram * @count: the count passed to that vm_map_ram call (cannot unmap partial) */ void vm_unmap_ram(const void *mem, unsigned int count) { |
65ee03c4b mm: fix overflow ... |
1667 |
unsigned long size = (unsigned long)count << PAGE_SHIFT; |
db64fe022 mm: rewrite vmap ... |
1668 |
unsigned long addr = (unsigned long)mem; |
9c3acf604 mm: remove free_u... |
1669 |
struct vmap_area *va; |
db64fe022 mm: rewrite vmap ... |
1670 |
|
5803ed292 mm: mark all call... |
1671 |
might_sleep(); |
db64fe022 mm: rewrite vmap ... |
1672 1673 1674 |
BUG_ON(!addr); BUG_ON(addr < VMALLOC_START); BUG_ON(addr > VMALLOC_END); |
a1c0b1a07 mm/vmalloc: use P... |
1675 |
BUG_ON(!PAGE_ALIGNED(addr)); |
db64fe022 mm: rewrite vmap ... |
1676 |
|
9c3acf604 mm: remove free_u... |
1677 |
if (likely(count <= VMAP_MAX_ALLOC)) { |
05e3ff950 mm: vmalloc: pass... |
1678 |
debug_check_no_locks_freed(mem, size); |
db64fe022 mm: rewrite vmap ... |
1679 |
vb_free(mem, size); |
9c3acf604 mm: remove free_u... |
1680 1681 1682 1683 1684 |
return; } va = find_vmap_area(addr); BUG_ON(!va); |
05e3ff950 mm: vmalloc: pass... |
1685 1686 |
debug_check_no_locks_freed((void *)va->va_start, (va->va_end - va->va_start)); |
9c3acf604 mm: remove free_u... |
1687 |
free_unmap_vmap_area(va); |
db64fe022 mm: rewrite vmap ... |
1688 1689 1690 1691 1692 1693 1694 1695 1696 |
} EXPORT_SYMBOL(vm_unmap_ram); /** * vm_map_ram - map pages linearly into kernel virtual address (vmalloc space) * @pages: an array of pointers to the pages to be mapped * @count: number of pages * @node: prefer to allocate data structures on this node * @prot: memory protection to use. PAGE_KERNEL for regular RAM |
e99c97ade mm: fix kernel-do... |
1697 |
* |
364376383 mm/vmalloc.c: enh... |
1698 1699 1700 1701 1702 1703 |
* If you use this function for less than VMAP_MAX_ALLOC pages, it could be * faster than vmap so it's good. But if you mix long-life and short-life * objects with vm_map_ram(), it could consume lots of address space through * fragmentation (especially on a 32bit machine). You could see failures in * the end. Please use this function for short-lived objects. * |
e99c97ade mm: fix kernel-do... |
1704 |
* Returns: a pointer to the address that has been mapped, or %NULL on failure |
db64fe022 mm: rewrite vmap ... |
1705 1706 1707 |
*/ void *vm_map_ram(struct page **pages, unsigned int count, int node, pgprot_t prot) { |
65ee03c4b mm: fix overflow ... |
1708 |
unsigned long size = (unsigned long)count << PAGE_SHIFT; |
db64fe022 mm: rewrite vmap ... |
1709 1710 1711 1712 1713 1714 1715 1716 1717 1718 1719 1720 1721 1722 1723 1724 1725 1726 1727 1728 1729 1730 1731 1732 1733 |
unsigned long addr; void *mem; if (likely(count <= VMAP_MAX_ALLOC)) { mem = vb_alloc(size, GFP_KERNEL); if (IS_ERR(mem)) return NULL; addr = (unsigned long)mem; } else { struct vmap_area *va; va = alloc_vmap_area(size, PAGE_SIZE, VMALLOC_START, VMALLOC_END, node, GFP_KERNEL); if (IS_ERR(va)) return NULL; addr = va->va_start; mem = (void *)addr; } if (vmap_page_range(addr, addr + size, prot, pages) < 0) { vm_unmap_ram(mem, count); return NULL; } return mem; } EXPORT_SYMBOL(vm_map_ram); |
4341fa454 mm, vmalloc: remo... |
1734 |
static struct vm_struct *vmlist __initdata; |
92eac1681 docs/mm: vmalloc:... |
1735 |
|
f0aa66179 vmalloc: implemen... |
1736 |
/** |
be9b7335e mm: add vm_area_a... |
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 |
* vm_area_add_early - add vmap area early during boot * @vm: vm_struct to add * * This function is used to add fixed kernel vm area to vmlist before * vmalloc_init() is called. @vm->addr, @vm->size, and @vm->flags * should contain proper values and the other fields should be zero. * * DO NOT USE THIS FUNCTION UNLESS YOU KNOW WHAT YOU'RE DOING. */ void __init vm_area_add_early(struct vm_struct *vm) { struct vm_struct *tmp, **p; BUG_ON(vmap_initialized); for (p = &vmlist; (tmp = *p) != NULL; p = &tmp->next) { if (tmp->addr >= vm->addr) { BUG_ON(tmp->addr < vm->addr + vm->size); break; } else BUG_ON(tmp->addr + tmp->size > vm->addr); } vm->next = *p; *p = vm; } /** |
f0aa66179 vmalloc: implemen... |
1763 1764 |
* vm_area_register_early - register vmap area early during boot * @vm: vm_struct to register |
c0c0a2937 vmalloc: add @ali... |
1765 |
* @align: requested alignment |
f0aa66179 vmalloc: implemen... |
1766 1767 1768 1769 1770 1771 1772 1773 |
* * This function is used to register kernel vm area before * vmalloc_init() is called. @vm->size and @vm->flags should contain * proper values on entry and other fields should be zero. On return, * vm->addr contains the allocated address. * * DO NOT USE THIS FUNCTION UNLESS YOU KNOW WHAT YOU'RE DOING. */ |
c0c0a2937 vmalloc: add @ali... |
1774 |
void __init vm_area_register_early(struct vm_struct *vm, size_t align) |
f0aa66179 vmalloc: implemen... |
1775 1776 |
{ static size_t vm_init_off __initdata; |
c0c0a2937 vmalloc: add @ali... |
1777 1778 1779 1780 |
unsigned long addr; addr = ALIGN(VMALLOC_START + vm_init_off, align); vm_init_off = PFN_ALIGN(addr + vm->size) - VMALLOC_START; |
f0aa66179 vmalloc: implemen... |
1781 |
|
c0c0a2937 vmalloc: add @ali... |
1782 |
vm->addr = (void *)addr; |
f0aa66179 vmalloc: implemen... |
1783 |
|
be9b7335e mm: add vm_area_a... |
1784 |
vm_area_add_early(vm); |
f0aa66179 vmalloc: implemen... |
1785 |
} |
68ad4a330 mm/vmalloc.c: kee... |
1786 1787 1788 1789 1790 1791 1792 1793 1794 1795 1796 1797 1798 1799 1800 1801 1802 1803 1804 1805 1806 1807 1808 1809 1810 1811 1812 1813 1814 1815 1816 1817 1818 1819 1820 1821 1822 1823 1824 1825 |
static void vmap_init_free_space(void) { unsigned long vmap_start = 1; const unsigned long vmap_end = ULONG_MAX; struct vmap_area *busy, *free; /* * B F B B B F * -|-----|.....|-----|-----|-----|.....|- * | The KVA space | * |<--------------------------------->| */ list_for_each_entry(busy, &vmap_area_list, list) { if (busy->va_start - vmap_start > 0) { free = kmem_cache_zalloc(vmap_area_cachep, GFP_NOWAIT); if (!WARN_ON_ONCE(!free)) { free->va_start = vmap_start; free->va_end = busy->va_start; insert_vmap_area_augment(free, NULL, &free_vmap_area_root, &free_vmap_area_list); } } vmap_start = busy->va_end; } if (vmap_end - vmap_start > 0) { free = kmem_cache_zalloc(vmap_area_cachep, GFP_NOWAIT); if (!WARN_ON_ONCE(!free)) { free->va_start = vmap_start; free->va_end = vmap_end; insert_vmap_area_augment(free, NULL, &free_vmap_area_root, &free_vmap_area_list); } } } |
db64fe022 mm: rewrite vmap ... |
1826 1827 |
void __init vmalloc_init(void) { |
822c18f2e alpha: fix vmallo... |
1828 1829 |
struct vmap_area *va; struct vm_struct *tmp; |
db64fe022 mm: rewrite vmap ... |
1830 |
int i; |
68ad4a330 mm/vmalloc.c: kee... |
1831 1832 1833 1834 |
/* * Create the cache for vmap_area objects. */ vmap_area_cachep = KMEM_CACHE(vmap_area, SLAB_PANIC); |
db64fe022 mm: rewrite vmap ... |
1835 1836 |
for_each_possible_cpu(i) { struct vmap_block_queue *vbq; |
32fcfd407 make vfree() safe... |
1837 |
struct vfree_deferred *p; |
db64fe022 mm: rewrite vmap ... |
1838 1839 1840 1841 |
vbq = &per_cpu(vmap_block_queue, i); spin_lock_init(&vbq->lock); INIT_LIST_HEAD(&vbq->free); |
32fcfd407 make vfree() safe... |
1842 1843 1844 |
p = &per_cpu(vfree_deferred, i); init_llist_head(&p->list); INIT_WORK(&p->wq, free_work); |
db64fe022 mm: rewrite vmap ... |
1845 |
} |
9b4633340 vmap: cope with v... |
1846 |
|
822c18f2e alpha: fix vmallo... |
1847 1848 |
/* Import existing vmlist entries. */ for (tmp = vmlist; tmp; tmp = tmp->next) { |
68ad4a330 mm/vmalloc.c: kee... |
1849 1850 1851 |
va = kmem_cache_zalloc(vmap_area_cachep, GFP_NOWAIT); if (WARN_ON_ONCE(!va)) continue; |
822c18f2e alpha: fix vmallo... |
1852 1853 |
va->va_start = (unsigned long)tmp->addr; va->va_end = va->va_start + tmp->size; |
dbda591d9 mm: fix faulty in... |
1854 |
va->vm = tmp; |
68ad4a330 mm/vmalloc.c: kee... |
1855 |
insert_vmap_area(va, &vmap_area_root, &vmap_area_list); |
822c18f2e alpha: fix vmallo... |
1856 |
} |
ca23e405e vmalloc: implemen... |
1857 |
|
68ad4a330 mm/vmalloc.c: kee... |
1858 1859 1860 1861 |
/* * Now we can initialize a free vmap space. */ vmap_init_free_space(); |
9b4633340 vmap: cope with v... |
1862 |
vmap_initialized = true; |
db64fe022 mm: rewrite vmap ... |
1863 |
} |
8fc489850 vmalloc: add un/m... |
1864 1865 1866 1867 1868 1869 1870 1871 1872 1873 1874 1875 1876 1877 1878 1879 1880 1881 1882 1883 1884 1885 1886 1887 1888 1889 1890 1891 1892 1893 1894 1895 1896 1897 1898 1899 1900 1901 1902 1903 1904 1905 1906 |
/** * map_kernel_range_noflush - map kernel VM area with the specified pages * @addr: start of the VM area to map * @size: size of the VM area to map * @prot: page protection flags to use * @pages: pages to map * * Map PFN_UP(@size) pages at @addr. The VM area @addr and @size * specify should have been allocated using get_vm_area() and its * friends. * * NOTE: * This function does NOT do any cache flushing. The caller is * responsible for calling flush_cache_vmap() on to-be-mapped areas * before calling this function. * * RETURNS: * The number of pages mapped on success, -errno on failure. */ int map_kernel_range_noflush(unsigned long addr, unsigned long size, pgprot_t prot, struct page **pages) { return vmap_page_range_noflush(addr, addr + size, prot, pages); } /** * unmap_kernel_range_noflush - unmap kernel VM area * @addr: start of the VM area to unmap * @size: size of the VM area to unmap * * Unmap PFN_UP(@size) pages at @addr. The VM area @addr and @size * specify should have been allocated using get_vm_area() and its * friends. * * NOTE: * This function does NOT do any cache flushing. The caller is * responsible for calling flush_cache_vunmap() on to-be-mapped areas * before calling this function and flush_tlb_kernel_range() after. */ void unmap_kernel_range_noflush(unsigned long addr, unsigned long size) { vunmap_page_range(addr, addr + size); } |
81e88fdc4 ACPI, APEI, Gener... |
1907 |
EXPORT_SYMBOL_GPL(unmap_kernel_range_noflush); |
8fc489850 vmalloc: add un/m... |
1908 1909 1910 1911 1912 1913 1914 1915 1916 |
/** * unmap_kernel_range - unmap kernel VM area and flush cache and TLB * @addr: start of the VM area to unmap * @size: size of the VM area to unmap * * Similar to unmap_kernel_range_noflush() but flushes vcache before * the unmapping and tlb after. */ |
db64fe022 mm: rewrite vmap ... |
1917 1918 1919 |
void unmap_kernel_range(unsigned long addr, unsigned long size) { unsigned long end = addr + size; |
f6fcba701 vmalloc: call flu... |
1920 1921 |
flush_cache_vunmap(addr, end); |
db64fe022 mm: rewrite vmap ... |
1922 1923 1924 |
vunmap_page_range(addr, end); flush_tlb_kernel_range(addr, end); } |
93ef6d6ca mm/vmalloc.c: exp... |
1925 |
EXPORT_SYMBOL_GPL(unmap_kernel_range); |
db64fe022 mm: rewrite vmap ... |
1926 |
|
f6f8ed473 mm/vmalloc.c: cle... |
1927 |
int map_vm_area(struct vm_struct *area, pgprot_t prot, struct page **pages) |
db64fe022 mm: rewrite vmap ... |
1928 1929 |
{ unsigned long addr = (unsigned long)area->addr; |
762216ab4 mm/vmalloc: use w... |
1930 |
unsigned long end = addr + get_vm_area_size(area); |
db64fe022 mm: rewrite vmap ... |
1931 |
int err; |
f6f8ed473 mm/vmalloc.c: cle... |
1932 |
err = vmap_page_range(addr, end, prot, pages); |
db64fe022 mm: rewrite vmap ... |
1933 |
|
f6f8ed473 mm/vmalloc.c: cle... |
1934 |
return err > 0 ? 0 : err; |
db64fe022 mm: rewrite vmap ... |
1935 1936 |
} EXPORT_SYMBOL_GPL(map_vm_area); |
f5252e009 mm: avoid null po... |
1937 |
static void setup_vmalloc_vm(struct vm_struct *vm, struct vmap_area *va, |
5e6cafc83 mm: vmalloc: use ... |
1938 |
unsigned long flags, const void *caller) |
cf88c7900 vmalloc: separate... |
1939 |
{ |
c69480ade mm, vmalloc: prot... |
1940 |
spin_lock(&vmap_area_lock); |
cf88c7900 vmalloc: separate... |
1941 1942 1943 1944 |
vm->flags = flags; vm->addr = (void *)va->va_start; vm->size = va->va_end - va->va_start; vm->caller = caller; |
db1aecafe mm/vmalloc.c: cha... |
1945 |
va->vm = vm; |
c69480ade mm, vmalloc: prot... |
1946 |
spin_unlock(&vmap_area_lock); |
f5252e009 mm: avoid null po... |
1947 |
} |
cf88c7900 vmalloc: separate... |
1948 |
|
20fc02b47 mm/vmalloc.c: ren... |
1949 |
static void clear_vm_uninitialized_flag(struct vm_struct *vm) |
f5252e009 mm: avoid null po... |
1950 |
{ |
d4033afdf mm, vmalloc: iter... |
1951 |
/* |
20fc02b47 mm/vmalloc.c: ren... |
1952 |
* Before removing VM_UNINITIALIZED, |
d4033afdf mm, vmalloc: iter... |
1953 1954 1955 1956 |
* we should make sure that vm has proper values. * Pair with smp_rmb() in show_numa_info(). */ smp_wmb(); |
20fc02b47 mm/vmalloc.c: ren... |
1957 |
vm->flags &= ~VM_UNINITIALIZED; |
cf88c7900 vmalloc: separate... |
1958 |
} |
db64fe022 mm: rewrite vmap ... |
1959 |
static struct vm_struct *__get_vm_area_node(unsigned long size, |
2dca6999e mm, perf_event: M... |
1960 |
unsigned long align, unsigned long flags, unsigned long start, |
5e6cafc83 mm: vmalloc: use ... |
1961 |
unsigned long end, int node, gfp_t gfp_mask, const void *caller) |
db64fe022 mm: rewrite vmap ... |
1962 |
{ |
0006526d7 mm/vmalloc.c: rem... |
1963 |
struct vmap_area *va; |
db64fe022 mm: rewrite vmap ... |
1964 |
struct vm_struct *area; |
1da177e4c Linux-2.6.12-rc2 |
1965 |
|
52fd24ca1 [PATCH] __vmalloc... |
1966 |
BUG_ON(in_interrupt()); |
1da177e4c Linux-2.6.12-rc2 |
1967 |
size = PAGE_ALIGN(size); |
31be83095 [PATCH] Fix stran... |
1968 1969 |
if (unlikely(!size)) return NULL; |
1da177e4c Linux-2.6.12-rc2 |
1970 |
|
252e5c6e2 mm/vmalloc.c: fix... |
1971 1972 1973 |
if (flags & VM_IOREMAP) align = 1ul << clamp_t(int, get_count_order_long(size), PAGE_SHIFT, IOREMAP_MAX_ORDER); |
cf88c7900 vmalloc: separate... |
1974 |
area = kzalloc_node(sizeof(*area), gfp_mask & GFP_RECLAIM_MASK, node); |
1da177e4c Linux-2.6.12-rc2 |
1975 1976 |
if (unlikely(!area)) return NULL; |
71394fe50 mm: vmalloc: add ... |
1977 1978 |
if (!(flags & VM_NO_GUARD)) size += PAGE_SIZE; |
1da177e4c Linux-2.6.12-rc2 |
1979 |
|
db64fe022 mm: rewrite vmap ... |
1980 1981 1982 1983 |
va = alloc_vmap_area(size, align, start, end, node, gfp_mask); if (IS_ERR(va)) { kfree(area); return NULL; |
1da177e4c Linux-2.6.12-rc2 |
1984 |
} |
1da177e4c Linux-2.6.12-rc2 |
1985 |
|
d82b1d857 mm, vmalloc: only... |
1986 |
setup_vmalloc_vm(area, va, flags, caller); |
f5252e009 mm: avoid null po... |
1987 |
|
1da177e4c Linux-2.6.12-rc2 |
1988 |
return area; |
1da177e4c Linux-2.6.12-rc2 |
1989 |
} |
930fc45a4 [PATCH] vmalloc_node |
1990 1991 1992 |
struct vm_struct *__get_vm_area(unsigned long size, unsigned long flags, unsigned long start, unsigned long end) { |
00ef2d2f8 mm: use NUMA_NO_NODE |
1993 1994 |
return __get_vm_area_node(size, 1, flags, start, end, NUMA_NO_NODE, GFP_KERNEL, __builtin_return_address(0)); |
930fc45a4 [PATCH] vmalloc_node |
1995 |
} |
5992b6dac lguest: export sy... |
1996 |
EXPORT_SYMBOL_GPL(__get_vm_area); |
930fc45a4 [PATCH] vmalloc_node |
1997 |
|
c29686129 vmalloc: add __ge... |
1998 1999 |
struct vm_struct *__get_vm_area_caller(unsigned long size, unsigned long flags, unsigned long start, unsigned long end, |
5e6cafc83 mm: vmalloc: use ... |
2000 |
const void *caller) |
c29686129 vmalloc: add __ge... |
2001 |
{ |
00ef2d2f8 mm: use NUMA_NO_NODE |
2002 2003 |
return __get_vm_area_node(size, 1, flags, start, end, NUMA_NO_NODE, GFP_KERNEL, caller); |
c29686129 vmalloc: add __ge... |
2004 |
} |
1da177e4c Linux-2.6.12-rc2 |
2005 |
/** |
92eac1681 docs/mm: vmalloc:... |
2006 2007 2008 |
* get_vm_area - reserve a contiguous kernel virtual area * @size: size of the area * @flags: %VM_IOREMAP for I/O mappings or VM_ALLOC |
1da177e4c Linux-2.6.12-rc2 |
2009 |
* |
92eac1681 docs/mm: vmalloc:... |
2010 2011 2012 |
* Search an area of @size in the kernel virtual mapping area, * and reserved it for out purposes. Returns the area descriptor * on success or %NULL on failure. |
a862f68a8 docs/core-api/mm:... |
2013 2014 |
* * Return: the area descriptor on success or %NULL on failure. |
1da177e4c Linux-2.6.12-rc2 |
2015 2016 2017 |
*/ struct vm_struct *get_vm_area(unsigned long size, unsigned long flags) { |
2dca6999e mm, perf_event: M... |
2018 |
return __get_vm_area_node(size, 1, flags, VMALLOC_START, VMALLOC_END, |
00ef2d2f8 mm: use NUMA_NO_NODE |
2019 2020 |
NUMA_NO_NODE, GFP_KERNEL, __builtin_return_address(0)); |
230169693 vmallocinfo: add ... |
2021 2022 2023 |
} struct vm_struct *get_vm_area_caller(unsigned long size, unsigned long flags, |
5e6cafc83 mm: vmalloc: use ... |
2024 |
const void *caller) |
230169693 vmallocinfo: add ... |
2025 |
{ |
2dca6999e mm, perf_event: M... |
2026 |
return __get_vm_area_node(size, 1, flags, VMALLOC_START, VMALLOC_END, |
00ef2d2f8 mm: use NUMA_NO_NODE |
2027 |
NUMA_NO_NODE, GFP_KERNEL, caller); |
1da177e4c Linux-2.6.12-rc2 |
2028 |
} |
e9da6e990 ARM: dma-mapping:... |
2029 |
/** |
92eac1681 docs/mm: vmalloc:... |
2030 2031 |
* find_vm_area - find a continuous kernel virtual area * @addr: base address |
e9da6e990 ARM: dma-mapping:... |
2032 |
* |
92eac1681 docs/mm: vmalloc:... |
2033 2034 2035 |
* Search for the kernel VM area starting at @addr, and return it. * It is up to the caller to do all required locking to keep the returned * pointer valid. |
a862f68a8 docs/core-api/mm:... |
2036 2037 |
* * Return: pointer to the found area or %NULL on faulure |
e9da6e990 ARM: dma-mapping:... |
2038 2039 |
*/ struct vm_struct *find_vm_area(const void *addr) |
833423143 [PATCH] mm: intro... |
2040 |
{ |
db64fe022 mm: rewrite vmap ... |
2041 |
struct vmap_area *va; |
833423143 [PATCH] mm: intro... |
2042 |
|
db64fe022 mm: rewrite vmap ... |
2043 |
va = find_vmap_area((unsigned long)addr); |
688fcbfc0 mm/vmalloc: modif... |
2044 2045 |
if (!va) return NULL; |
1da177e4c Linux-2.6.12-rc2 |
2046 |
|
688fcbfc0 mm/vmalloc: modif... |
2047 |
return va->vm; |
1da177e4c Linux-2.6.12-rc2 |
2048 |
} |
7856dfeb2 [PATCH] x86_64: F... |
2049 |
/** |
92eac1681 docs/mm: vmalloc:... |
2050 2051 |
* remove_vm_area - find and remove a continuous kernel virtual area * @addr: base address |
7856dfeb2 [PATCH] x86_64: F... |
2052 |
* |
92eac1681 docs/mm: vmalloc:... |
2053 2054 2055 |
* Search for the kernel VM area starting at @addr, and remove it. * This function returns the found VM area, but using it is NOT safe * on SMP machines, except for its size or flags. |
a862f68a8 docs/core-api/mm:... |
2056 2057 |
* * Return: pointer to the found area or %NULL on faulure |
7856dfeb2 [PATCH] x86_64: F... |
2058 |
*/ |
b3bdda02a vmalloc: add cons... |
2059 |
struct vm_struct *remove_vm_area(const void *addr) |
7856dfeb2 [PATCH] x86_64: F... |
2060 |
{ |
db64fe022 mm: rewrite vmap ... |
2061 |
struct vmap_area *va; |
5803ed292 mm: mark all call... |
2062 |
might_sleep(); |
dd3b8353b mm/vmalloc: do no... |
2063 2064 |
spin_lock(&vmap_area_lock); va = __find_vmap_area((unsigned long)addr); |
688fcbfc0 mm/vmalloc: modif... |
2065 |
if (va && va->vm) { |
db1aecafe mm/vmalloc.c: cha... |
2066 |
struct vm_struct *vm = va->vm; |
f5252e009 mm: avoid null po... |
2067 |
|
c69480ade mm, vmalloc: prot... |
2068 |
va->vm = NULL; |
c69480ade mm, vmalloc: prot... |
2069 |
spin_unlock(&vmap_area_lock); |
a5af5aa8b kasan, module, vm... |
2070 |
kasan_free_shadow(vm); |
dd32c2799 vmalloc: unmap vm... |
2071 |
free_unmap_vmap_area(va); |
dd32c2799 vmalloc: unmap vm... |
2072 |
|
db64fe022 mm: rewrite vmap ... |
2073 2074 |
return vm; } |
dd3b8353b mm/vmalloc: do no... |
2075 2076 |
spin_unlock(&vmap_area_lock); |
db64fe022 mm: rewrite vmap ... |
2077 |
return NULL; |
7856dfeb2 [PATCH] x86_64: F... |
2078 |
} |
868b104d7 mm/vmalloc: Add f... |
2079 2080 2081 2082 2083 2084 2085 2086 2087 2088 2089 2090 2091 |
static inline void set_area_direct_map(const struct vm_struct *area, int (*set_direct_map)(struct page *page)) { int i; for (i = 0; i < area->nr_pages; i++) if (page_address(area->pages[i])) set_direct_map(area->pages[i]); } /* Handle removing and resetting vm mappings related to the vm_struct. */ static void vm_remove_mappings(struct vm_struct *area, int deallocate_pages) { |
868b104d7 mm/vmalloc: Add f... |
2092 2093 |
unsigned long start = ULONG_MAX, end = 0; int flush_reset = area->flags & VM_FLUSH_RESET_PERMS; |
31e67340c mm/vmalloc: Avoid... |
2094 |
int flush_dmap = 0; |
868b104d7 mm/vmalloc: Add f... |
2095 |
int i; |
868b104d7 mm/vmalloc: Add f... |
2096 2097 2098 2099 2100 2101 2102 2103 2104 2105 2106 2107 2108 2109 2110 2111 2112 2113 2114 2115 2116 |
remove_vm_area(area->addr); /* If this is not VM_FLUSH_RESET_PERMS memory, no need for the below. */ if (!flush_reset) return; /* * If not deallocating pages, just do the flush of the VM area and * return. */ if (!deallocate_pages) { vm_unmap_aliases(); return; } /* * If execution gets here, flush the vm mapping and reset the direct * map. Find the start and end range of the direct mappings to make sure * the vm_unmap_aliases() flush includes the direct map. */ for (i = 0; i < area->nr_pages; i++) { |
8e41f8726 mm/vmalloc: Fix c... |
2117 2118 |
unsigned long addr = (unsigned long)page_address(area->pages[i]); if (addr) { |
868b104d7 mm/vmalloc: Add f... |
2119 |
start = min(addr, start); |
8e41f8726 mm/vmalloc: Fix c... |
2120 |
end = max(addr + PAGE_SIZE, end); |
31e67340c mm/vmalloc: Avoid... |
2121 |
flush_dmap = 1; |
868b104d7 mm/vmalloc: Add f... |
2122 2123 2124 2125 2126 2127 2128 2129 2130 |
} } /* * Set direct map to something invalid so that it won't be cached if * there are any accesses after the TLB flush, then flush the TLB and * reset the direct map permissions to the default. */ set_area_direct_map(area, set_direct_map_invalid_noflush); |
31e67340c mm/vmalloc: Avoid... |
2131 |
_vm_unmap_aliases(start, end, flush_dmap); |
868b104d7 mm/vmalloc: Add f... |
2132 2133 |
set_area_direct_map(area, set_direct_map_default_noflush); } |
b3bdda02a vmalloc: add cons... |
2134 |
static void __vunmap(const void *addr, int deallocate_pages) |
1da177e4c Linux-2.6.12-rc2 |
2135 2136 2137 2138 2139 |
{ struct vm_struct *area; if (!addr) return; |
e69e9d4ae vmalloc: introduc... |
2140 2141 |
if (WARN(!PAGE_ALIGNED(addr), "Trying to vfree() bad address (%p) ", |
ab15d9b4c mm/vmalloc.c: unb... |
2142 |
addr)) |
1da177e4c Linux-2.6.12-rc2 |
2143 |
return; |
1da177e4c Linux-2.6.12-rc2 |
2144 |
|
6ade20327 mm/vmalloc.c: don... |
2145 |
area = find_vm_area(addr); |
1da177e4c Linux-2.6.12-rc2 |
2146 |
if (unlikely(!area)) { |
4c8573e25 Use WARN() in mm/... |
2147 2148 |
WARN(1, KERN_ERR "Trying to vfree() nonexistent vm area (%p) ", |
1da177e4c Linux-2.6.12-rc2 |
2149 |
addr); |
1da177e4c Linux-2.6.12-rc2 |
2150 2151 |
return; } |
05e3ff950 mm: vmalloc: pass... |
2152 2153 |
debug_check_no_locks_freed(area->addr, get_vm_area_size(area)); debug_check_no_obj_freed(area->addr, get_vm_area_size(area)); |
9a11b49a8 [PATCH] lockdep: ... |
2154 |
|
868b104d7 mm/vmalloc: Add f... |
2155 |
vm_remove_mappings(area, deallocate_pages); |
1da177e4c Linux-2.6.12-rc2 |
2156 2157 2158 2159 |
if (deallocate_pages) { int i; for (i = 0; i < area->nr_pages; i++) { |
bf53d6f8f vmalloc: clean up... |
2160 2161 2162 |
struct page *page = area->pages[i]; BUG_ON(!page); |
4949148ad mm: charge/unchar... |
2163 |
__free_pages(page, 0); |
1da177e4c Linux-2.6.12-rc2 |
2164 |
} |
97105f0ab mm: vmalloc: show... |
2165 |
atomic_long_sub(area->nr_pages, &nr_vmalloc_pages); |
1da177e4c Linux-2.6.12-rc2 |
2166 |
|
244d63ee3 mm, vmalloc: remo... |
2167 |
kvfree(area->pages); |
1da177e4c Linux-2.6.12-rc2 |
2168 2169 2170 2171 2172 |
} kfree(area); return; } |
bf22e37a6 mm: add vfree_ato... |
2173 2174 2175 2176 2177 2178 2179 2180 2181 2182 2183 2184 2185 2186 2187 2188 |
static inline void __vfree_deferred(const void *addr) { /* * Use raw_cpu_ptr() because this can be called from preemptible * context. Preemption is absolutely fine here, because the llist_add() * implementation is lockless, so it works even if we are adding to * nother cpu's list. schedule_work() should be fine with this too. */ struct vfree_deferred *p = raw_cpu_ptr(&vfree_deferred); if (llist_add((struct llist_node *)addr, &p->list)) schedule_work(&p->wq); } /** |
92eac1681 docs/mm: vmalloc:... |
2189 2190 |
* vfree_atomic - release memory allocated by vmalloc() * @addr: memory base address |
bf22e37a6 mm: add vfree_ato... |
2191 |
* |
92eac1681 docs/mm: vmalloc:... |
2192 2193 |
* This one is just like vfree() but can be called in any atomic context * except NMIs. |
bf22e37a6 mm: add vfree_ato... |
2194 2195 2196 2197 2198 2199 2200 2201 2202 2203 2204 |
*/ void vfree_atomic(const void *addr) { BUG_ON(in_nmi()); kmemleak_free(addr); if (!addr) return; __vfree_deferred(addr); } |
c67dc6247 mm/vmalloc: do no... |
2205 2206 2207 2208 2209 2210 2211 |
static void __vfree(const void *addr) { if (unlikely(in_interrupt())) __vfree_deferred(addr); else __vunmap(addr, 1); } |
1da177e4c Linux-2.6.12-rc2 |
2212 |
/** |
92eac1681 docs/mm: vmalloc:... |
2213 2214 |
* vfree - release memory allocated by vmalloc() * @addr: memory base address |
1da177e4c Linux-2.6.12-rc2 |
2215 |
* |
92eac1681 docs/mm: vmalloc:... |
2216 2217 2218 |
* Free the virtually continuous memory area starting at @addr, as * obtained from vmalloc(), vmalloc_32() or __vmalloc(). If @addr is * NULL, no operation is performed. |
1da177e4c Linux-2.6.12-rc2 |
2219 |
* |
92eac1681 docs/mm: vmalloc:... |
2220 2221 2222 |
* Must not be called in NMI context (strictly speaking, only if we don't * have CONFIG_ARCH_HAVE_NMI_SAFE_CMPXCHG, but making the calling * conventions for vfree() arch-depenedent would be a really bad idea) |
c9fcee513 mm/vmalloc.c: add... |
2223 |
* |
92eac1681 docs/mm: vmalloc:... |
2224 |
* May sleep if called *not* from interrupt context. |
3ca4ea3a7 mm/vmalloc.c: imp... |
2225 |
* |
92eac1681 docs/mm: vmalloc:... |
2226 |
* NOTE: assumes that the object at @addr has a size >= sizeof(llist_node) |
1da177e4c Linux-2.6.12-rc2 |
2227 |
*/ |
b3bdda02a vmalloc: add cons... |
2228 |
void vfree(const void *addr) |
1da177e4c Linux-2.6.12-rc2 |
2229 |
{ |
32fcfd407 make vfree() safe... |
2230 |
BUG_ON(in_nmi()); |
89219d37a kmemleak: Add the... |
2231 2232 |
kmemleak_free(addr); |
a8dda165e vfree: add debug ... |
2233 |
might_sleep_if(!in_interrupt()); |
32fcfd407 make vfree() safe... |
2234 2235 |
if (!addr) return; |
c67dc6247 mm/vmalloc: do no... |
2236 2237 |
__vfree(addr); |
1da177e4c Linux-2.6.12-rc2 |
2238 |
} |
1da177e4c Linux-2.6.12-rc2 |
2239 2240 2241 |
EXPORT_SYMBOL(vfree); /** |
92eac1681 docs/mm: vmalloc:... |
2242 2243 |
* vunmap - release virtual mapping obtained by vmap() * @addr: memory base address |
1da177e4c Linux-2.6.12-rc2 |
2244 |
* |
92eac1681 docs/mm: vmalloc:... |
2245 2246 |
* Free the virtually contiguous memory area starting at @addr, * which was created from the page array passed to vmap(). |
1da177e4c Linux-2.6.12-rc2 |
2247 |
* |
92eac1681 docs/mm: vmalloc:... |
2248 |
* Must not be called in interrupt context. |
1da177e4c Linux-2.6.12-rc2 |
2249 |
*/ |
b3bdda02a vmalloc: add cons... |
2250 |
void vunmap(const void *addr) |
1da177e4c Linux-2.6.12-rc2 |
2251 2252 |
{ BUG_ON(in_interrupt()); |
34754b69a x86: make vmap ye... |
2253 |
might_sleep(); |
32fcfd407 make vfree() safe... |
2254 2255 |
if (addr) __vunmap(addr, 0); |
1da177e4c Linux-2.6.12-rc2 |
2256 |
} |
1da177e4c Linux-2.6.12-rc2 |
2257 2258 2259 |
EXPORT_SYMBOL(vunmap); /** |
92eac1681 docs/mm: vmalloc:... |
2260 2261 2262 2263 2264 2265 2266 2267 |
* vmap - map an array of pages into virtually contiguous space * @pages: array of page pointers * @count: number of pages to map * @flags: vm_area->flags * @prot: page protection for the mapping * * Maps @count pages from @pages into contiguous kernel virtual * space. |
a862f68a8 docs/core-api/mm:... |
2268 2269 |
* * Return: the address of the area or %NULL on failure |
1da177e4c Linux-2.6.12-rc2 |
2270 2271 |
*/ void *vmap(struct page **pages, unsigned int count, |
92eac1681 docs/mm: vmalloc:... |
2272 |
unsigned long flags, pgprot_t prot) |
1da177e4c Linux-2.6.12-rc2 |
2273 2274 |
{ struct vm_struct *area; |
65ee03c4b mm: fix overflow ... |
2275 |
unsigned long size; /* In bytes */ |
1da177e4c Linux-2.6.12-rc2 |
2276 |
|
34754b69a x86: make vmap ye... |
2277 |
might_sleep(); |
ca79b0c21 mm: convert total... |
2278 |
if (count > totalram_pages()) |
1da177e4c Linux-2.6.12-rc2 |
2279 |
return NULL; |
65ee03c4b mm: fix overflow ... |
2280 2281 |
size = (unsigned long)count << PAGE_SHIFT; area = get_vm_area_caller(size, flags, __builtin_return_address(0)); |
1da177e4c Linux-2.6.12-rc2 |
2282 2283 |
if (!area) return NULL; |
230169693 vmallocinfo: add ... |
2284 |
|
f6f8ed473 mm/vmalloc.c: cle... |
2285 |
if (map_vm_area(area, prot, pages)) { |
1da177e4c Linux-2.6.12-rc2 |
2286 2287 2288 2289 2290 2291 |
vunmap(area->addr); return NULL; } return area->addr; } |
1da177e4c Linux-2.6.12-rc2 |
2292 |
EXPORT_SYMBOL(vmap); |
8594a21cf mm, vmalloc: fix ... |
2293 2294 2295 |
static void *__vmalloc_node(unsigned long size, unsigned long align, gfp_t gfp_mask, pgprot_t prot, int node, const void *caller); |
e31d9eb5c make __vmalloc_ar... |
2296 |
static void *__vmalloc_area_node(struct vm_struct *area, gfp_t gfp_mask, |
3722e13cf mm/vmalloc: don't... |
2297 |
pgprot_t prot, int node) |
1da177e4c Linux-2.6.12-rc2 |
2298 2299 2300 |
{ struct page **pages; unsigned int nr_pages, array_size, i; |
930f036b4 mm, vmalloc: cons... |
2301 |
const gfp_t nested_gfp = (gfp_mask & GFP_RECLAIM_MASK) | __GFP_ZERO; |
704b862f9 mm/vmalloc.c: don... |
2302 2303 2304 2305 |
const gfp_t alloc_mask = gfp_mask | __GFP_NOWARN; const gfp_t highmem_mask = (gfp_mask & (GFP_DMA | GFP_DMA32)) ? 0 : __GFP_HIGHMEM; |
1da177e4c Linux-2.6.12-rc2 |
2306 |
|
762216ab4 mm/vmalloc: use w... |
2307 |
nr_pages = get_vm_area_size(area) >> PAGE_SHIFT; |
1da177e4c Linux-2.6.12-rc2 |
2308 |
array_size = (nr_pages * sizeof(struct page *)); |
1da177e4c Linux-2.6.12-rc2 |
2309 |
/* Please note that the recursion is strictly bounded. */ |
8757d5fa6 [PATCH] mm: fix o... |
2310 |
if (array_size > PAGE_SIZE) { |
704b862f9 mm/vmalloc.c: don... |
2311 |
pages = __vmalloc_node(array_size, 1, nested_gfp|highmem_mask, |
3722e13cf mm/vmalloc: don't... |
2312 |
PAGE_KERNEL, node, area->caller); |
286e1ea3a [PATCH] vmalloc()... |
2313 |
} else { |
976d6dfbb vmalloc(): adjust... |
2314 |
pages = kmalloc_node(array_size, nested_gfp, node); |
286e1ea3a [PATCH] vmalloc()... |
2315 |
} |
7ea362427 mm/vmalloc.c: mov... |
2316 2317 |
if (!pages) { |
1da177e4c Linux-2.6.12-rc2 |
2318 2319 2320 2321 |
remove_vm_area(area->addr); kfree(area); return NULL; } |
1da177e4c Linux-2.6.12-rc2 |
2322 |
|
7ea362427 mm/vmalloc.c: mov... |
2323 2324 |
area->pages = pages; area->nr_pages = nr_pages; |
1da177e4c Linux-2.6.12-rc2 |
2325 |
for (i = 0; i < area->nr_pages; i++) { |
bf53d6f8f vmalloc: clean up... |
2326 |
struct page *page; |
4b90951c0 mm/vmalloc: use N... |
2327 |
if (node == NUMA_NO_NODE) |
704b862f9 mm/vmalloc.c: don... |
2328 |
page = alloc_page(alloc_mask|highmem_mask); |
930fc45a4 [PATCH] vmalloc_node |
2329 |
else |
704b862f9 mm/vmalloc.c: don... |
2330 |
page = alloc_pages_node(node, alloc_mask|highmem_mask, 0); |
bf53d6f8f vmalloc: clean up... |
2331 2332 |
if (unlikely(!page)) { |
1da177e4c Linux-2.6.12-rc2 |
2333 2334 |
/* Successfully allocated i pages, free them in __vunmap() */ area->nr_pages = i; |
97105f0ab mm: vmalloc: show... |
2335 |
atomic_long_add(area->nr_pages, &nr_vmalloc_pages); |
1da177e4c Linux-2.6.12-rc2 |
2336 2337 |
goto fail; } |
bf53d6f8f vmalloc: clean up... |
2338 |
area->pages[i] = page; |
704b862f9 mm/vmalloc.c: don... |
2339 |
if (gfpflags_allow_blocking(gfp_mask|highmem_mask)) |
660654f90 mm/vmalloc.c: add... |
2340 |
cond_resched(); |
1da177e4c Linux-2.6.12-rc2 |
2341 |
} |
97105f0ab mm: vmalloc: show... |
2342 |
atomic_long_add(area->nr_pages, &nr_vmalloc_pages); |
1da177e4c Linux-2.6.12-rc2 |
2343 |
|
f6f8ed473 mm/vmalloc.c: cle... |
2344 |
if (map_vm_area(area, prot, pages)) |
1da177e4c Linux-2.6.12-rc2 |
2345 2346 2347 2348 |
goto fail; return area->addr; fail: |
a8e99259e mm, page_alloc: w... |
2349 |
warn_alloc(gfp_mask, NULL, |
7877cdcc3 mm: consolidate w... |
2350 |
"vmalloc: allocation failure, allocated %ld of %ld bytes", |
22943ab11 mm: print vmalloc... |
2351 |
(area->nr_pages*PAGE_SIZE), area->size); |
c67dc6247 mm/vmalloc: do no... |
2352 |
__vfree(area->addr); |
1da177e4c Linux-2.6.12-rc2 |
2353 2354 2355 2356 |
return NULL; } /** |
92eac1681 docs/mm: vmalloc:... |
2357 2358 2359 2360 2361 2362 2363 2364 2365 2366 2367 2368 2369 2370 |
* __vmalloc_node_range - allocate virtually contiguous memory * @size: allocation size * @align: desired alignment * @start: vm area range start * @end: vm area range end * @gfp_mask: flags for the page level allocator * @prot: protection mask for the allocated pages * @vm_flags: additional vm area flags (e.g. %VM_NO_GUARD) * @node: node to use for allocation or NUMA_NO_NODE * @caller: caller's return address * * Allocate enough pages to cover @size from the page level * allocator with @gfp_mask flags. Map them into contiguous * kernel virtual space, using a pagetable protection of @prot. |
a862f68a8 docs/core-api/mm:... |
2371 2372 |
* * Return: the address of the area or %NULL on failure |
1da177e4c Linux-2.6.12-rc2 |
2373 |
*/ |
d0a21265d mm: unify module_... |
2374 2375 |
void *__vmalloc_node_range(unsigned long size, unsigned long align, unsigned long start, unsigned long end, gfp_t gfp_mask, |
cb9e3c292 mm: vmalloc: pass... |
2376 2377 |
pgprot_t prot, unsigned long vm_flags, int node, const void *caller) |
1da177e4c Linux-2.6.12-rc2 |
2378 2379 |
{ struct vm_struct *area; |
89219d37a kmemleak: Add the... |
2380 2381 |
void *addr; unsigned long real_size = size; |
1da177e4c Linux-2.6.12-rc2 |
2382 2383 |
size = PAGE_ALIGN(size); |
ca79b0c21 mm: convert total... |
2384 |
if (!size || (size >> PAGE_SHIFT) > totalram_pages()) |
de7d2b567 mm/vmalloc.c: rep... |
2385 |
goto fail; |
1da177e4c Linux-2.6.12-rc2 |
2386 |
|
cb9e3c292 mm: vmalloc: pass... |
2387 2388 |
area = __get_vm_area_node(size, align, VM_ALLOC | VM_UNINITIALIZED | vm_flags, start, end, node, gfp_mask, caller); |
1da177e4c Linux-2.6.12-rc2 |
2389 |
if (!area) |
de7d2b567 mm/vmalloc.c: rep... |
2390 |
goto fail; |
1da177e4c Linux-2.6.12-rc2 |
2391 |
|
3722e13cf mm/vmalloc: don't... |
2392 |
addr = __vmalloc_area_node(area, gfp_mask, prot, node); |
1368edf06 mm: vmalloc: chec... |
2393 |
if (!addr) |
b82225f3f revert mm/vmalloc... |
2394 |
return NULL; |
89219d37a kmemleak: Add the... |
2395 2396 |
/* |
20fc02b47 mm/vmalloc.c: ren... |
2397 2398 |
* In this function, newly allocated vm_struct has VM_UNINITIALIZED * flag. It means that vm_struct is not fully initialized. |
4341fa454 mm, vmalloc: remo... |
2399 |
* Now, it is fully initialized, so remove this flag here. |
f5252e009 mm: avoid null po... |
2400 |
*/ |
20fc02b47 mm/vmalloc.c: ren... |
2401 |
clear_vm_uninitialized_flag(area); |
f5252e009 mm: avoid null po... |
2402 |
|
94f4a1618 mm: kmemleak: tre... |
2403 |
kmemleak_vmalloc(area, size, gfp_mask); |
89219d37a kmemleak: Add the... |
2404 2405 |
return addr; |
de7d2b567 mm/vmalloc.c: rep... |
2406 2407 |
fail: |
a8e99259e mm, page_alloc: w... |
2408 |
warn_alloc(gfp_mask, NULL, |
7877cdcc3 mm: consolidate w... |
2409 |
"vmalloc: allocation failure: %lu bytes", real_size); |
de7d2b567 mm/vmalloc.c: rep... |
2410 |
return NULL; |
1da177e4c Linux-2.6.12-rc2 |
2411 |
} |
153178edc vmalloc: export _... |
2412 2413 2414 2415 2416 2417 2418 2419 |
/* * This is only for performance analysis of vmalloc and stress purpose. * It is required by vmalloc test module, therefore do not use it other * than that. */ #ifdef CONFIG_TEST_VMALLOC_MODULE EXPORT_SYMBOL_GPL(__vmalloc_node_range); #endif |
d0a21265d mm: unify module_... |
2420 |
/** |
92eac1681 docs/mm: vmalloc:... |
2421 2422 2423 2424 2425 2426 2427 |
* __vmalloc_node - allocate virtually contiguous memory * @size: allocation size * @align: desired alignment * @gfp_mask: flags for the page level allocator * @prot: protection mask for the allocated pages * @node: node to use for allocation or NUMA_NO_NODE * @caller: caller's return address |
a7c3e901a mm: introduce kv[... |
2428 |
* |
92eac1681 docs/mm: vmalloc:... |
2429 2430 2431 |
* Allocate enough pages to cover @size from the page level * allocator with @gfp_mask flags. Map them into contiguous * kernel virtual space, using a pagetable protection of @prot. |
a7c3e901a mm: introduce kv[... |
2432 |
* |
92eac1681 docs/mm: vmalloc:... |
2433 2434 |
* Reclaim modifiers in @gfp_mask - __GFP_NORETRY, __GFP_RETRY_MAYFAIL * and __GFP_NOFAIL are not supported |
a7c3e901a mm: introduce kv[... |
2435 |
* |
92eac1681 docs/mm: vmalloc:... |
2436 2437 |
* Any use of gfp flags outside of GFP_KERNEL should be consulted * with mm people. |
a862f68a8 docs/core-api/mm:... |
2438 2439 |
* * Return: pointer to the allocated memory or %NULL on error |
d0a21265d mm: unify module_... |
2440 |
*/ |
8594a21cf mm, vmalloc: fix ... |
2441 |
static void *__vmalloc_node(unsigned long size, unsigned long align, |
d0a21265d mm: unify module_... |
2442 |
gfp_t gfp_mask, pgprot_t prot, |
5e6cafc83 mm: vmalloc: use ... |
2443 |
int node, const void *caller) |
d0a21265d mm: unify module_... |
2444 2445 |
{ return __vmalloc_node_range(size, align, VMALLOC_START, VMALLOC_END, |
cb9e3c292 mm: vmalloc: pass... |
2446 |
gfp_mask, prot, 0, node, caller); |
d0a21265d mm: unify module_... |
2447 |
} |
930fc45a4 [PATCH] vmalloc_node |
2448 2449 |
void *__vmalloc(unsigned long size, gfp_t gfp_mask, pgprot_t prot) { |
00ef2d2f8 mm: use NUMA_NO_NODE |
2450 |
return __vmalloc_node(size, 1, gfp_mask, prot, NUMA_NO_NODE, |
230169693 vmallocinfo: add ... |
2451 |
__builtin_return_address(0)); |
930fc45a4 [PATCH] vmalloc_node |
2452 |
} |
1da177e4c Linux-2.6.12-rc2 |
2453 |
EXPORT_SYMBOL(__vmalloc); |
8594a21cf mm, vmalloc: fix ... |
2454 2455 2456 2457 2458 2459 2460 2461 2462 2463 2464 2465 2466 |
static inline void *__vmalloc_node_flags(unsigned long size, int node, gfp_t flags) { return __vmalloc_node(size, 1, flags, PAGE_KERNEL, node, __builtin_return_address(0)); } void *__vmalloc_node_flags_caller(unsigned long size, int node, gfp_t flags, void *caller) { return __vmalloc_node(size, 1, flags, PAGE_KERNEL, node, caller); } |
1da177e4c Linux-2.6.12-rc2 |
2467 |
/** |
92eac1681 docs/mm: vmalloc:... |
2468 2469 2470 2471 2472 |
* vmalloc - allocate virtually contiguous memory * @size: allocation size * * Allocate enough pages to cover @size from the page level * allocator and map them into contiguous kernel virtual space. |
1da177e4c Linux-2.6.12-rc2 |
2473 |
* |
92eac1681 docs/mm: vmalloc:... |
2474 2475 |
* For tight control over page level allocator and protection flags * use __vmalloc() instead. |
a862f68a8 docs/core-api/mm:... |
2476 2477 |
* * Return: pointer to the allocated memory or %NULL on error |
1da177e4c Linux-2.6.12-rc2 |
2478 2479 2480 |
*/ void *vmalloc(unsigned long size) { |
00ef2d2f8 mm: use NUMA_NO_NODE |
2481 |
return __vmalloc_node_flags(size, NUMA_NO_NODE, |
19809c2da mm, vmalloc: use ... |
2482 |
GFP_KERNEL); |
1da177e4c Linux-2.6.12-rc2 |
2483 |
} |
1da177e4c Linux-2.6.12-rc2 |
2484 |
EXPORT_SYMBOL(vmalloc); |
930fc45a4 [PATCH] vmalloc_node |
2485 |
/** |
92eac1681 docs/mm: vmalloc:... |
2486 2487 2488 2489 2490 2491 2492 2493 2494 |
* vzalloc - allocate virtually contiguous memory with zero fill * @size: allocation size * * Allocate enough pages to cover @size from the page level * allocator and map them into contiguous kernel virtual space. * The memory allocated is set to zero. * * For tight control over page level allocator and protection flags * use __vmalloc() instead. |
a862f68a8 docs/core-api/mm:... |
2495 2496 |
* * Return: pointer to the allocated memory or %NULL on error |
e1ca7788d mm: add vzalloc()... |
2497 2498 2499 |
*/ void *vzalloc(unsigned long size) { |
00ef2d2f8 mm: use NUMA_NO_NODE |
2500 |
return __vmalloc_node_flags(size, NUMA_NO_NODE, |
19809c2da mm, vmalloc: use ... |
2501 |
GFP_KERNEL | __GFP_ZERO); |
e1ca7788d mm: add vzalloc()... |
2502 2503 2504 2505 |
} EXPORT_SYMBOL(vzalloc); /** |
ead04089b [PATCH] Fix kerne... |
2506 2507 |
* vmalloc_user - allocate zeroed virtually contiguous memory for userspace * @size: allocation size |
833423143 [PATCH] mm: intro... |
2508 |
* |
ead04089b [PATCH] Fix kerne... |
2509 2510 |
* The resulting memory area is zeroed so it can be mapped to userspace * without leaking data. |
a862f68a8 docs/core-api/mm:... |
2511 2512 |
* * Return: pointer to the allocated memory or %NULL on error |
833423143 [PATCH] mm: intro... |
2513 2514 2515 |
*/ void *vmalloc_user(unsigned long size) { |
bc84c5352 mm/vmalloc: pass ... |
2516 2517 2518 2519 |
return __vmalloc_node_range(size, SHMLBA, VMALLOC_START, VMALLOC_END, GFP_KERNEL | __GFP_ZERO, PAGE_KERNEL, VM_USERMAP, NUMA_NO_NODE, __builtin_return_address(0)); |
833423143 [PATCH] mm: intro... |
2520 2521 2522 2523 |
} EXPORT_SYMBOL(vmalloc_user); /** |
92eac1681 docs/mm: vmalloc:... |
2524 2525 2526 |
* vmalloc_node - allocate memory on a specific node * @size: allocation size * @node: numa node |
930fc45a4 [PATCH] vmalloc_node |
2527 |
* |
92eac1681 docs/mm: vmalloc:... |
2528 2529 |
* Allocate enough pages to cover @size from the page level * allocator and map them into contiguous kernel virtual space. |
930fc45a4 [PATCH] vmalloc_node |
2530 |
* |
92eac1681 docs/mm: vmalloc:... |
2531 2532 |
* For tight control over page level allocator and protection flags * use __vmalloc() instead. |
a862f68a8 docs/core-api/mm:... |
2533 2534 |
* * Return: pointer to the allocated memory or %NULL on error |
930fc45a4 [PATCH] vmalloc_node |
2535 2536 2537 |
*/ void *vmalloc_node(unsigned long size, int node) { |
19809c2da mm, vmalloc: use ... |
2538 |
return __vmalloc_node(size, 1, GFP_KERNEL, PAGE_KERNEL, |
230169693 vmallocinfo: add ... |
2539 |
node, __builtin_return_address(0)); |
930fc45a4 [PATCH] vmalloc_node |
2540 2541 |
} EXPORT_SYMBOL(vmalloc_node); |
e1ca7788d mm: add vzalloc()... |
2542 2543 2544 2545 2546 2547 2548 2549 2550 2551 2552 |
/** * vzalloc_node - allocate memory on a specific node with zero fill * @size: allocation size * @node: numa node * * Allocate enough pages to cover @size from the page level * allocator and map them into contiguous kernel virtual space. * The memory allocated is set to zero. * * For tight control over page level allocator and protection flags * use __vmalloc_node() instead. |
a862f68a8 docs/core-api/mm:... |
2553 2554 |
* * Return: pointer to the allocated memory or %NULL on error |
e1ca7788d mm: add vzalloc()... |
2555 2556 2557 2558 |
*/ void *vzalloc_node(unsigned long size, int node) { return __vmalloc_node_flags(size, node, |
19809c2da mm, vmalloc: use ... |
2559 |
GFP_KERNEL | __GFP_ZERO); |
e1ca7788d mm: add vzalloc()... |
2560 2561 |
} EXPORT_SYMBOL(vzalloc_node); |
1da177e4c Linux-2.6.12-rc2 |
2562 |
/** |
92eac1681 docs/mm: vmalloc:... |
2563 2564 |
* vmalloc_exec - allocate virtually contiguous, executable memory * @size: allocation size |
1da177e4c Linux-2.6.12-rc2 |
2565 |
* |
92eac1681 docs/mm: vmalloc:... |
2566 2567 2568 |
* Kernel-internal function to allocate enough pages to cover @size * the page level allocator and map them into contiguous and * executable kernel virtual space. |
1da177e4c Linux-2.6.12-rc2 |
2569 |
* |
92eac1681 docs/mm: vmalloc:... |
2570 2571 |
* For tight control over page level allocator and protection flags * use __vmalloc() instead. |
a862f68a8 docs/core-api/mm:... |
2572 2573 |
* * Return: pointer to the allocated memory or %NULL on error |
1da177e4c Linux-2.6.12-rc2 |
2574 |
*/ |
1da177e4c Linux-2.6.12-rc2 |
2575 2576 |
void *vmalloc_exec(unsigned long size) { |
868b104d7 mm/vmalloc: Add f... |
2577 2578 2579 |
return __vmalloc_node_range(size, 1, VMALLOC_START, VMALLOC_END, GFP_KERNEL, PAGE_KERNEL_EXEC, VM_FLUSH_RESET_PERMS, NUMA_NO_NODE, __builtin_return_address(0)); |
1da177e4c Linux-2.6.12-rc2 |
2580 |
} |
0d08e0d3a [PATCH] x86-64: F... |
2581 |
#if defined(CONFIG_64BIT) && defined(CONFIG_ZONE_DMA32) |
698d0831b vmalloc: fix __GF... |
2582 |
#define GFP_VMALLOC32 (GFP_DMA32 | GFP_KERNEL) |
0d08e0d3a [PATCH] x86-64: F... |
2583 |
#elif defined(CONFIG_64BIT) && defined(CONFIG_ZONE_DMA) |
698d0831b vmalloc: fix __GF... |
2584 |
#define GFP_VMALLOC32 (GFP_DMA | GFP_KERNEL) |
0d08e0d3a [PATCH] x86-64: F... |
2585 |
#else |
698d0831b vmalloc: fix __GF... |
2586 2587 2588 2589 2590 |
/* * 64b systems should always have either DMA or DMA32 zones. For others * GFP_DMA32 should do the right thing and use the normal zone. */ #define GFP_VMALLOC32 GFP_DMA32 | GFP_KERNEL |
0d08e0d3a [PATCH] x86-64: F... |
2591 |
#endif |
1da177e4c Linux-2.6.12-rc2 |
2592 |
/** |
92eac1681 docs/mm: vmalloc:... |
2593 2594 |
* vmalloc_32 - allocate virtually contiguous memory (32bit addressable) * @size: allocation size |
1da177e4c Linux-2.6.12-rc2 |
2595 |
* |
92eac1681 docs/mm: vmalloc:... |
2596 2597 |
* Allocate enough 32bit PA addressable pages to cover @size from the * page level allocator and map them into contiguous kernel virtual space. |
a862f68a8 docs/core-api/mm:... |
2598 2599 |
* * Return: pointer to the allocated memory or %NULL on error |
1da177e4c Linux-2.6.12-rc2 |
2600 2601 2602 |
*/ void *vmalloc_32(unsigned long size) { |
2dca6999e mm, perf_event: M... |
2603 |
return __vmalloc_node(size, 1, GFP_VMALLOC32, PAGE_KERNEL, |
00ef2d2f8 mm: use NUMA_NO_NODE |
2604 |
NUMA_NO_NODE, __builtin_return_address(0)); |
1da177e4c Linux-2.6.12-rc2 |
2605 |
} |
1da177e4c Linux-2.6.12-rc2 |
2606 |
EXPORT_SYMBOL(vmalloc_32); |
833423143 [PATCH] mm: intro... |
2607 |
/** |
ead04089b [PATCH] Fix kerne... |
2608 |
* vmalloc_32_user - allocate zeroed virtually contiguous 32bit memory |
92eac1681 docs/mm: vmalloc:... |
2609 |
* @size: allocation size |
ead04089b [PATCH] Fix kerne... |
2610 2611 2612 |
* * The resulting memory area is 32bit addressable and zeroed so it can be * mapped to userspace without leaking data. |
a862f68a8 docs/core-api/mm:... |
2613 2614 |
* * Return: pointer to the allocated memory or %NULL on error |
833423143 [PATCH] mm: intro... |
2615 2616 2617 |
*/ void *vmalloc_32_user(unsigned long size) { |
bc84c5352 mm/vmalloc: pass ... |
2618 2619 2620 2621 |
return __vmalloc_node_range(size, SHMLBA, VMALLOC_START, VMALLOC_END, GFP_VMALLOC32 | __GFP_ZERO, PAGE_KERNEL, VM_USERMAP, NUMA_NO_NODE, __builtin_return_address(0)); |
833423143 [PATCH] mm: intro... |
2622 2623 |
} EXPORT_SYMBOL(vmalloc_32_user); |
d0107eb07 kcore: fix vread/... |
2624 2625 2626 2627 2628 2629 2630 2631 2632 2633 2634 2635 |
/* * small helper routine , copy contents to buf from addr. * If the page is not present, fill zero. */ static int aligned_vread(char *buf, char *addr, unsigned long count) { struct page *p; int copied = 0; while (count) { unsigned long offset, length; |
891c49abf mm/vmalloc: use o... |
2636 |
offset = offset_in_page(addr); |
d0107eb07 kcore: fix vread/... |
2637 2638 2639 2640 2641 2642 2643 2644 2645 2646 2647 2648 2649 2650 2651 2652 |
length = PAGE_SIZE - offset; if (length > count) length = count; p = vmalloc_to_page(addr); /* * To do safe access to this _mapped_ area, we need * lock. But adding lock here means that we need to add * overhead of vmalloc()/vfree() calles for this _debug_ * interface, rarely used. Instead of that, we'll use * kmap() and get small overhead in this access function. */ if (p) { /* * we can expect USER0 is not used (see vread/vwrite's * function description) */ |
9b04c5fec mm: remove the se... |
2653 |
void *map = kmap_atomic(p); |
d0107eb07 kcore: fix vread/... |
2654 |
memcpy(buf, map + offset, length); |
9b04c5fec mm: remove the se... |
2655 |
kunmap_atomic(map); |
d0107eb07 kcore: fix vread/... |
2656 2657 2658 2659 2660 2661 2662 2663 2664 2665 2666 2667 2668 2669 2670 2671 2672 2673 |
} else memset(buf, 0, length); addr += length; buf += length; copied += length; count -= length; } return copied; } static int aligned_vwrite(char *buf, char *addr, unsigned long count) { struct page *p; int copied = 0; while (count) { unsigned long offset, length; |
891c49abf mm/vmalloc: use o... |
2674 |
offset = offset_in_page(addr); |
d0107eb07 kcore: fix vread/... |
2675 2676 2677 2678 2679 2680 2681 2682 2683 2684 2685 2686 2687 2688 2689 2690 |
length = PAGE_SIZE - offset; if (length > count) length = count; p = vmalloc_to_page(addr); /* * To do safe access to this _mapped_ area, we need * lock. But adding lock here means that we need to add * overhead of vmalloc()/vfree() calles for this _debug_ * interface, rarely used. Instead of that, we'll use * kmap() and get small overhead in this access function. */ if (p) { /* * we can expect USER0 is not used (see vread/vwrite's * function description) */ |
9b04c5fec mm: remove the se... |
2691 |
void *map = kmap_atomic(p); |
d0107eb07 kcore: fix vread/... |
2692 |
memcpy(map + offset, buf, length); |
9b04c5fec mm: remove the se... |
2693 |
kunmap_atomic(map); |
d0107eb07 kcore: fix vread/... |
2694 2695 2696 2697 2698 2699 2700 2701 2702 2703 |
} addr += length; buf += length; copied += length; count -= length; } return copied; } /** |
92eac1681 docs/mm: vmalloc:... |
2704 2705 2706 2707 2708 |
* vread() - read vmalloc area in a safe way. * @buf: buffer for reading data * @addr: vm address. * @count: number of bytes to be read. * |
92eac1681 docs/mm: vmalloc:... |
2709 2710 2711 2712 2713 2714 2715 2716 2717 2718 2719 2720 |
* This function checks that addr is a valid vmalloc'ed area, and * copy data from that area to a given buffer. If the given memory range * of [addr...addr+count) includes some valid address, data is copied to * proper area of @buf. If there are memory holes, they'll be zero-filled. * IOREMAP area is treated as memory hole and no copy is done. * * If [addr...addr+count) doesn't includes any intersects with alive * vm_struct area, returns 0. @buf should be kernel's buffer. * * Note: In usual ops, vread() is never necessary because the caller * should know vmalloc() area is valid and can use memcpy(). * This is for routines which have to access vmalloc area without |
d9009d67f mm/vmalloc.c: spe... |
2721 |
* any information, as /dev/kmem. |
a862f68a8 docs/core-api/mm:... |
2722 2723 2724 2725 |
* * Return: number of bytes for which addr and buf should be increased * (same number as @count) or %0 if [addr...addr+count) doesn't * include any intersection with valid vmalloc area |
d0107eb07 kcore: fix vread/... |
2726 |
*/ |
1da177e4c Linux-2.6.12-rc2 |
2727 2728 |
long vread(char *buf, char *addr, unsigned long count) { |
e81ce85f9 mm, vmalloc: iter... |
2729 2730 |
struct vmap_area *va; struct vm_struct *vm; |
1da177e4c Linux-2.6.12-rc2 |
2731 |
char *vaddr, *buf_start = buf; |
d0107eb07 kcore: fix vread/... |
2732 |
unsigned long buflen = count; |
1da177e4c Linux-2.6.12-rc2 |
2733 2734 2735 2736 2737 |
unsigned long n; /* Don't allow overflow */ if ((unsigned long) addr + count < count) count = -(unsigned long) addr; |
e81ce85f9 mm, vmalloc: iter... |
2738 2739 2740 2741 |
spin_lock(&vmap_area_lock); list_for_each_entry(va, &vmap_area_list, list) { if (!count) break; |
688fcbfc0 mm/vmalloc: modif... |
2742 |
if (!va->vm) |
e81ce85f9 mm, vmalloc: iter... |
2743 2744 2745 2746 |
continue; vm = va->vm; vaddr = (char *) vm->addr; |
762216ab4 mm/vmalloc: use w... |
2747 |
if (addr >= vaddr + get_vm_area_size(vm)) |
1da177e4c Linux-2.6.12-rc2 |
2748 2749 2750 2751 2752 2753 2754 2755 2756 |
continue; while (addr < vaddr) { if (count == 0) goto finished; *buf = '\0'; buf++; addr++; count--; } |
762216ab4 mm/vmalloc: use w... |
2757 |
n = vaddr + get_vm_area_size(vm) - addr; |
d0107eb07 kcore: fix vread/... |
2758 2759 |
if (n > count) n = count; |
e81ce85f9 mm, vmalloc: iter... |
2760 |
if (!(vm->flags & VM_IOREMAP)) |
d0107eb07 kcore: fix vread/... |
2761 2762 2763 2764 2765 2766 |
aligned_vread(buf, addr, n); else /* IOREMAP area is treated as memory hole */ memset(buf, 0, n); buf += n; addr += n; count -= n; |
1da177e4c Linux-2.6.12-rc2 |
2767 2768 |
} finished: |
e81ce85f9 mm, vmalloc: iter... |
2769 |
spin_unlock(&vmap_area_lock); |
d0107eb07 kcore: fix vread/... |
2770 2771 2772 2773 2774 2775 2776 2777 |
if (buf == buf_start) return 0; /* zero-fill memory holes */ if (buf != buf_start + buflen) memset(buf, 0, buflen - (buf - buf_start)); return buflen; |
1da177e4c Linux-2.6.12-rc2 |
2778 |
} |
d0107eb07 kcore: fix vread/... |
2779 |
/** |
92eac1681 docs/mm: vmalloc:... |
2780 2781 2782 2783 2784 |
* vwrite() - write vmalloc area in a safe way. * @buf: buffer for source data * @addr: vm address. * @count: number of bytes to be read. * |
92eac1681 docs/mm: vmalloc:... |
2785 2786 2787 2788 2789 2790 2791 2792 2793 2794 2795 2796 |
* This function checks that addr is a valid vmalloc'ed area, and * copy data from a buffer to the given addr. If specified range of * [addr...addr+count) includes some valid address, data is copied from * proper area of @buf. If there are memory holes, no copy to hole. * IOREMAP area is treated as memory hole and no copy is done. * * If [addr...addr+count) doesn't includes any intersects with alive * vm_struct area, returns 0. @buf should be kernel's buffer. * * Note: In usual ops, vwrite() is never necessary because the caller * should know vmalloc() area is valid and can use memcpy(). * This is for routines which have to access vmalloc area without |
d9009d67f mm/vmalloc.c: spe... |
2797 |
* any information, as /dev/kmem. |
a862f68a8 docs/core-api/mm:... |
2798 2799 2800 2801 |
* * Return: number of bytes for which addr and buf should be * increased (same number as @count) or %0 if [addr...addr+count) * doesn't include any intersection with valid vmalloc area |
d0107eb07 kcore: fix vread/... |
2802 |
*/ |
1da177e4c Linux-2.6.12-rc2 |
2803 2804 |
long vwrite(char *buf, char *addr, unsigned long count) { |
e81ce85f9 mm, vmalloc: iter... |
2805 2806 |
struct vmap_area *va; struct vm_struct *vm; |
d0107eb07 kcore: fix vread/... |
2807 2808 2809 |
char *vaddr; unsigned long n, buflen; int copied = 0; |
1da177e4c Linux-2.6.12-rc2 |
2810 2811 2812 2813 |
/* Don't allow overflow */ if ((unsigned long) addr + count < count) count = -(unsigned long) addr; |
d0107eb07 kcore: fix vread/... |
2814 |
buflen = count; |
1da177e4c Linux-2.6.12-rc2 |
2815 |
|
e81ce85f9 mm, vmalloc: iter... |
2816 2817 2818 2819 |
spin_lock(&vmap_area_lock); list_for_each_entry(va, &vmap_area_list, list) { if (!count) break; |
688fcbfc0 mm/vmalloc: modif... |
2820 |
if (!va->vm) |
e81ce85f9 mm, vmalloc: iter... |
2821 2822 2823 2824 |
continue; vm = va->vm; vaddr = (char *) vm->addr; |
762216ab4 mm/vmalloc: use w... |
2825 |
if (addr >= vaddr + get_vm_area_size(vm)) |
1da177e4c Linux-2.6.12-rc2 |
2826 2827 2828 2829 2830 2831 2832 2833 |
continue; while (addr < vaddr) { if (count == 0) goto finished; buf++; addr++; count--; } |
762216ab4 mm/vmalloc: use w... |
2834 |
n = vaddr + get_vm_area_size(vm) - addr; |
d0107eb07 kcore: fix vread/... |
2835 2836 |
if (n > count) n = count; |
e81ce85f9 mm, vmalloc: iter... |
2837 |
if (!(vm->flags & VM_IOREMAP)) { |
d0107eb07 kcore: fix vread/... |
2838 2839 2840 2841 2842 2843 |
aligned_vwrite(buf, addr, n); copied++; } buf += n; addr += n; count -= n; |
1da177e4c Linux-2.6.12-rc2 |
2844 2845 |
} finished: |
e81ce85f9 mm, vmalloc: iter... |
2846 |
spin_unlock(&vmap_area_lock); |
d0107eb07 kcore: fix vread/... |
2847 2848 2849 |
if (!copied) return 0; return buflen; |
1da177e4c Linux-2.6.12-rc2 |
2850 |
} |
833423143 [PATCH] mm: intro... |
2851 2852 |
/** |
92eac1681 docs/mm: vmalloc:... |
2853 2854 2855 2856 2857 |
* remap_vmalloc_range_partial - map vmalloc pages to userspace * @vma: vma to cover * @uaddr: target user address to start at * @kaddr: virtual address of vmalloc kernel memory * @size: size of map area |
7682486b3 mm: fix various k... |
2858 |
* |
92eac1681 docs/mm: vmalloc:... |
2859 |
* Returns: 0 for success, -Exxx on failure |
833423143 [PATCH] mm: intro... |
2860 |
* |
92eac1681 docs/mm: vmalloc:... |
2861 2862 2863 2864 |
* This function checks that @kaddr is a valid vmalloc'ed area, * and that it is big enough to cover the range starting at * @uaddr in @vma. Will return failure if that criteria isn't * met. |
833423143 [PATCH] mm: intro... |
2865 |
* |
92eac1681 docs/mm: vmalloc:... |
2866 |
* Similar to remap_pfn_range() (see mm/memory.c) |
833423143 [PATCH] mm: intro... |
2867 |
*/ |
e69e9d4ae vmalloc: introduc... |
2868 2869 |
int remap_vmalloc_range_partial(struct vm_area_struct *vma, unsigned long uaddr, void *kaddr, unsigned long size) |
833423143 [PATCH] mm: intro... |
2870 2871 |
{ struct vm_struct *area; |
833423143 [PATCH] mm: intro... |
2872 |
|
e69e9d4ae vmalloc: introduc... |
2873 2874 2875 |
size = PAGE_ALIGN(size); if (!PAGE_ALIGNED(uaddr) || !PAGE_ALIGNED(kaddr)) |
833423143 [PATCH] mm: intro... |
2876 |
return -EINVAL; |
e69e9d4ae vmalloc: introduc... |
2877 |
area = find_vm_area(kaddr); |
833423143 [PATCH] mm: intro... |
2878 |
if (!area) |
db64fe022 mm: rewrite vmap ... |
2879 |
return -EINVAL; |
833423143 [PATCH] mm: intro... |
2880 |
|
fe9041c24 vmalloc: lift the... |
2881 |
if (!(area->flags & (VM_USERMAP | VM_DMA_COHERENT))) |
db64fe022 mm: rewrite vmap ... |
2882 |
return -EINVAL; |
833423143 [PATCH] mm: intro... |
2883 |
|
401592d2e mm/vmalloc: fix s... |
2884 |
if (kaddr + size > area->addr + get_vm_area_size(area)) |
db64fe022 mm: rewrite vmap ... |
2885 |
return -EINVAL; |
833423143 [PATCH] mm: intro... |
2886 |
|
833423143 [PATCH] mm: intro... |
2887 |
do { |
e69e9d4ae vmalloc: introduc... |
2888 |
struct page *page = vmalloc_to_page(kaddr); |
db64fe022 mm: rewrite vmap ... |
2889 |
int ret; |
833423143 [PATCH] mm: intro... |
2890 2891 2892 2893 2894 |
ret = vm_insert_page(vma, uaddr, page); if (ret) return ret; uaddr += PAGE_SIZE; |
e69e9d4ae vmalloc: introduc... |
2895 2896 2897 |
kaddr += PAGE_SIZE; size -= PAGE_SIZE; } while (size > 0); |
833423143 [PATCH] mm: intro... |
2898 |
|
314e51b98 mm: kill vma flag... |
2899 |
vma->vm_flags |= VM_DONTEXPAND | VM_DONTDUMP; |
833423143 [PATCH] mm: intro... |
2900 |
|
db64fe022 mm: rewrite vmap ... |
2901 |
return 0; |
833423143 [PATCH] mm: intro... |
2902 |
} |
e69e9d4ae vmalloc: introduc... |
2903 2904 2905 |
EXPORT_SYMBOL(remap_vmalloc_range_partial); /** |
92eac1681 docs/mm: vmalloc:... |
2906 2907 2908 2909 |
* remap_vmalloc_range - map vmalloc pages to userspace * @vma: vma to cover (map full range of vma) * @addr: vmalloc memory * @pgoff: number of pages into addr before first page to map |
e69e9d4ae vmalloc: introduc... |
2910 |
* |
92eac1681 docs/mm: vmalloc:... |
2911 |
* Returns: 0 for success, -Exxx on failure |
e69e9d4ae vmalloc: introduc... |
2912 |
* |
92eac1681 docs/mm: vmalloc:... |
2913 2914 2915 |
* This function checks that addr is a valid vmalloc'ed area, and * that it is big enough to cover the vma. Will return failure if * that criteria isn't met. |
e69e9d4ae vmalloc: introduc... |
2916 |
* |
92eac1681 docs/mm: vmalloc:... |
2917 |
* Similar to remap_pfn_range() (see mm/memory.c) |
e69e9d4ae vmalloc: introduc... |
2918 2919 2920 2921 2922 2923 2924 2925 |
*/ int remap_vmalloc_range(struct vm_area_struct *vma, void *addr, unsigned long pgoff) { return remap_vmalloc_range_partial(vma, vma->vm_start, addr + (pgoff << PAGE_SHIFT), vma->vm_end - vma->vm_start); } |
833423143 [PATCH] mm: intro... |
2926 |
EXPORT_SYMBOL(remap_vmalloc_range); |
1eeb66a1b move die notifier... |
2927 2928 2929 |
/* * Implement a stub for vmalloc_sync_all() if the architecture chose not to * have one. |
3f8fd02b1 mm/vmalloc: Sync ... |
2930 2931 2932 |
* * The purpose of this function is to make sure the vmalloc area * mappings are identical in all page-tables in the system. |
1eeb66a1b move die notifier... |
2933 |
*/ |
3b32123d7 mm: use macros fr... |
2934 |
void __weak vmalloc_sync_all(void) |
1eeb66a1b move die notifier... |
2935 2936 |
{ } |
5f4352fbf Allocate and free... |
2937 |
|
8b1e0f81f mm/pgtable: drop ... |
2938 |
static int f(pte_t *pte, unsigned long addr, void *data) |
5f4352fbf Allocate and free... |
2939 |
{ |
cd12909cb xen: map foreign ... |
2940 2941 2942 2943 2944 2945 |
pte_t ***p = data; if (p) { *(*p) = pte; (*p)++; } |
5f4352fbf Allocate and free... |
2946 2947 2948 2949 |
return 0; } /** |
92eac1681 docs/mm: vmalloc:... |
2950 2951 2952 |
* alloc_vm_area - allocate a range of kernel address space * @size: size of the area * @ptes: returns the PTEs for the address space |
7682486b3 mm: fix various k... |
2953 |
* |
92eac1681 docs/mm: vmalloc:... |
2954 |
* Returns: NULL on failure, vm_struct on success |
5f4352fbf Allocate and free... |
2955 |
* |
92eac1681 docs/mm: vmalloc:... |
2956 2957 2958 |
* This function reserves a range of kernel address space, and * allocates pagetables to map that range. No actual mappings * are created. |
cd12909cb xen: map foreign ... |
2959 |
* |
92eac1681 docs/mm: vmalloc:... |
2960 2961 |
* If @ptes is non-NULL, pointers to the PTEs (in init_mm) * allocated for the VM area are returned. |
5f4352fbf Allocate and free... |
2962 |
*/ |
cd12909cb xen: map foreign ... |
2963 |
struct vm_struct *alloc_vm_area(size_t size, pte_t **ptes) |
5f4352fbf Allocate and free... |
2964 2965 |
{ struct vm_struct *area; |
230169693 vmallocinfo: add ... |
2966 2967 |
area = get_vm_area_caller(size, VM_IOREMAP, __builtin_return_address(0)); |
5f4352fbf Allocate and free... |
2968 2969 2970 2971 2972 2973 2974 2975 |
if (area == NULL) return NULL; /* * This ensures that page tables are constructed for this region * of kernel virtual address space and mapped into init_mm. */ if (apply_to_page_range(&init_mm, (unsigned long)area->addr, |
cd12909cb xen: map foreign ... |
2976 |
size, f, ptes ? &ptes : NULL)) { |
5f4352fbf Allocate and free... |
2977 2978 2979 |
free_vm_area(area); return NULL; } |
5f4352fbf Allocate and free... |
2980 2981 2982 2983 2984 2985 2986 2987 2988 2989 2990 2991 |
return area; } EXPORT_SYMBOL_GPL(alloc_vm_area); void free_vm_area(struct vm_struct *area) { struct vm_struct *ret; ret = remove_vm_area(area->addr); BUG_ON(ret != area); kfree(area); } EXPORT_SYMBOL_GPL(free_vm_area); |
a10aa5798 vmalloc: show vma... |
2992 |
|
4f8b02b4e vmalloc: pcpu_get... |
2993 |
#ifdef CONFIG_SMP |
ca23e405e vmalloc: implemen... |
2994 2995 |
static struct vmap_area *node_to_va(struct rb_node *n) { |
4583e7731 mm/vmalloc.c: use... |
2996 |
return rb_entry_safe(n, struct vmap_area, rb_node); |
ca23e405e vmalloc: implemen... |
2997 2998 2999 |
} /** |
68ad4a330 mm/vmalloc.c: kee... |
3000 3001 |
* pvm_find_va_enclose_addr - find the vmap_area @addr belongs to * @addr: target address |
ca23e405e vmalloc: implemen... |
3002 |
* |
68ad4a330 mm/vmalloc.c: kee... |
3003 3004 3005 3006 |
* Returns: vmap_area if it is found. If there is no such area * the first highest(reverse order) vmap_area is returned * i.e. va->va_start < addr && va->va_end < addr or NULL * if there are no any areas before @addr. |
ca23e405e vmalloc: implemen... |
3007 |
*/ |
68ad4a330 mm/vmalloc.c: kee... |
3008 3009 |
static struct vmap_area * pvm_find_va_enclose_addr(unsigned long addr) |
ca23e405e vmalloc: implemen... |
3010 |
{ |
68ad4a330 mm/vmalloc.c: kee... |
3011 3012 3013 3014 3015 |
struct vmap_area *va, *tmp; struct rb_node *n; n = free_vmap_area_root.rb_node; va = NULL; |
ca23e405e vmalloc: implemen... |
3016 3017 |
while (n) { |
68ad4a330 mm/vmalloc.c: kee... |
3018 3019 3020 3021 3022 |
tmp = rb_entry(n, struct vmap_area, rb_node); if (tmp->va_start <= addr) { va = tmp; if (tmp->va_end >= addr) break; |
ca23e405e vmalloc: implemen... |
3023 |
n = n->rb_right; |
68ad4a330 mm/vmalloc.c: kee... |
3024 3025 3026 |
} else { n = n->rb_left; } |
ca23e405e vmalloc: implemen... |
3027 |
} |
68ad4a330 mm/vmalloc.c: kee... |
3028 |
return va; |
ca23e405e vmalloc: implemen... |
3029 3030 3031 |
} /** |
68ad4a330 mm/vmalloc.c: kee... |
3032 3033 3034 3035 3036 |
* pvm_determine_end_from_reverse - find the highest aligned address * of free block below VMALLOC_END * @va: * in - the VA we start the search(reverse order); * out - the VA with the highest aligned end address. |
ca23e405e vmalloc: implemen... |
3037 |
* |
68ad4a330 mm/vmalloc.c: kee... |
3038 |
* Returns: determined end address within vmap_area |
ca23e405e vmalloc: implemen... |
3039 |
*/ |
68ad4a330 mm/vmalloc.c: kee... |
3040 3041 |
static unsigned long pvm_determine_end_from_reverse(struct vmap_area **va, unsigned long align) |
ca23e405e vmalloc: implemen... |
3042 |
{ |
68ad4a330 mm/vmalloc.c: kee... |
3043 |
unsigned long vmalloc_end = VMALLOC_END & ~(align - 1); |
ca23e405e vmalloc: implemen... |
3044 |
unsigned long addr; |
68ad4a330 mm/vmalloc.c: kee... |
3045 3046 3047 3048 3049 3050 3051 |
if (likely(*va)) { list_for_each_entry_from_reverse((*va), &free_vmap_area_list, list) { addr = min((*va)->va_end & ~(align - 1), vmalloc_end); if ((*va)->va_start < addr) return addr; } |
ca23e405e vmalloc: implemen... |
3052 |
} |
68ad4a330 mm/vmalloc.c: kee... |
3053 |
return 0; |
ca23e405e vmalloc: implemen... |
3054 3055 3056 3057 3058 3059 3060 3061 |
} /** * pcpu_get_vm_areas - allocate vmalloc areas for percpu allocator * @offsets: array containing offset of each area * @sizes: array containing size of each area * @nr_vms: the number of areas to allocate * @align: alignment, all entries in @offsets and @sizes must be aligned to this |
ca23e405e vmalloc: implemen... |
3062 3063 3064 3065 3066 3067 |
* * Returns: kmalloc'd vm_struct pointer array pointing to allocated * vm_structs on success, %NULL on failure * * Percpu allocator wants to use congruent vm areas so that it can * maintain the offsets among percpu areas. This function allocates |
ec3f64fc9 mm: remove gfp ma... |
3068 3069 3070 3071 |
* congruent vmalloc areas for it with GFP_KERNEL. These areas tend to * be scattered pretty far, distance between two areas easily going up * to gigabytes. To avoid interacting with regular vmallocs, these * areas are allocated from top. |
ca23e405e vmalloc: implemen... |
3072 |
* |
68ad4a330 mm/vmalloc.c: kee... |
3073 3074 3075 3076 3077 3078 |
* Despite its complicated look, this allocator is rather simple. It * does everything top-down and scans free blocks from the end looking * for matching base. While scanning, if any of the areas do not fit the * base address is pulled down to fit the area. Scanning is repeated till * all the areas fit and then all necessary data structures are inserted * and the result is returned. |
ca23e405e vmalloc: implemen... |
3079 3080 3081 |
*/ struct vm_struct **pcpu_get_vm_areas(const unsigned long *offsets, const size_t *sizes, int nr_vms, |
ec3f64fc9 mm: remove gfp ma... |
3082 |
size_t align) |
ca23e405e vmalloc: implemen... |
3083 3084 3085 |
{ const unsigned long vmalloc_start = ALIGN(VMALLOC_START, align); const unsigned long vmalloc_end = VMALLOC_END & ~(align - 1); |
68ad4a330 mm/vmalloc.c: kee... |
3086 |
struct vmap_area **vas, *va; |
ca23e405e vmalloc: implemen... |
3087 3088 |
struct vm_struct **vms; int area, area2, last_area, term_area; |
68ad4a330 mm/vmalloc.c: kee... |
3089 |
unsigned long base, start, size, end, last_end; |
ca23e405e vmalloc: implemen... |
3090 |
bool purged = false; |
68ad4a330 mm/vmalloc.c: kee... |
3091 |
enum fit_type type; |
ca23e405e vmalloc: implemen... |
3092 |
|
ca23e405e vmalloc: implemen... |
3093 |
/* verify parameters and allocate data structures */ |
891c49abf mm/vmalloc: use o... |
3094 |
BUG_ON(offset_in_page(align) || !is_power_of_2(align)); |
ca23e405e vmalloc: implemen... |
3095 3096 3097 3098 3099 3100 3101 3102 3103 3104 3105 |
for (last_area = 0, area = 0; area < nr_vms; area++) { start = offsets[area]; end = start + sizes[area]; /* is everything aligned properly? */ BUG_ON(!IS_ALIGNED(offsets[area], align)); BUG_ON(!IS_ALIGNED(sizes[area], align)); /* detect the area with the highest address */ if (start > offsets[last_area]) last_area = area; |
c568da282 mm/vmalloc.c: hal... |
3106 |
for (area2 = area + 1; area2 < nr_vms; area2++) { |
ca23e405e vmalloc: implemen... |
3107 3108 |
unsigned long start2 = offsets[area2]; unsigned long end2 = start2 + sizes[area2]; |
c568da282 mm/vmalloc.c: hal... |
3109 |
BUG_ON(start2 < end && start < end2); |
ca23e405e vmalloc: implemen... |
3110 3111 3112 3113 3114 3115 3116 3117 |
} } last_end = offsets[last_area] + sizes[last_area]; if (vmalloc_end - vmalloc_start < last_end) { WARN_ON(true); return NULL; } |
4d67d8605 mm: use kcalloc()... |
3118 3119 |
vms = kcalloc(nr_vms, sizeof(vms[0]), GFP_KERNEL); vas = kcalloc(nr_vms, sizeof(vas[0]), GFP_KERNEL); |
ca23e405e vmalloc: implemen... |
3120 |
if (!vas || !vms) |
f1db7afd9 mm/vmalloc.c: eli... |
3121 |
goto err_free2; |
ca23e405e vmalloc: implemen... |
3122 3123 |
for (area = 0; area < nr_vms; area++) { |
68ad4a330 mm/vmalloc.c: kee... |
3124 |
vas[area] = kmem_cache_zalloc(vmap_area_cachep, GFP_KERNEL); |
ec3f64fc9 mm: remove gfp ma... |
3125 |
vms[area] = kzalloc(sizeof(struct vm_struct), GFP_KERNEL); |
ca23e405e vmalloc: implemen... |
3126 3127 3128 3129 3130 3131 3132 3133 3134 3135 |
if (!vas[area] || !vms[area]) goto err_free; } retry: spin_lock(&vmap_area_lock); /* start scanning - we scan from the top, begin with the last area */ area = term_area = last_area; start = offsets[area]; end = start + sizes[area]; |
68ad4a330 mm/vmalloc.c: kee... |
3136 3137 |
va = pvm_find_va_enclose_addr(vmalloc_end); base = pvm_determine_end_from_reverse(&va, align) - end; |
ca23e405e vmalloc: implemen... |
3138 3139 |
while (true) { |
ca23e405e vmalloc: implemen... |
3140 3141 3142 3143 |
/* * base might have underflowed, add last_end before * comparing. */ |
68ad4a330 mm/vmalloc.c: kee... |
3144 3145 |
if (base + last_end < vmalloc_start + last_end) goto overflow; |
ca23e405e vmalloc: implemen... |
3146 3147 |
/* |
68ad4a330 mm/vmalloc.c: kee... |
3148 |
* Fitting base has not been found. |
ca23e405e vmalloc: implemen... |
3149 |
*/ |
68ad4a330 mm/vmalloc.c: kee... |
3150 3151 |
if (va == NULL) goto overflow; |
ca23e405e vmalloc: implemen... |
3152 3153 |
/* |
5336e52c9 mm/vmalloc.c: fix... |
3154 3155 3156 3157 3158 3159 3160 3161 3162 3163 |
* If required width exeeds current VA block, move * base downwards and then recheck. */ if (base + end > va->va_end) { base = pvm_determine_end_from_reverse(&va, align) - end; term_area = area; continue; } /* |
68ad4a330 mm/vmalloc.c: kee... |
3164 |
* If this VA does not fit, move base downwards and recheck. |
ca23e405e vmalloc: implemen... |
3165 |
*/ |
5336e52c9 mm/vmalloc.c: fix... |
3166 |
if (base + start < va->va_start) { |
68ad4a330 mm/vmalloc.c: kee... |
3167 3168 |
va = node_to_va(rb_prev(&va->rb_node)); base = pvm_determine_end_from_reverse(&va, align) - end; |
ca23e405e vmalloc: implemen... |
3169 3170 3171 3172 3173 3174 3175 3176 3177 3178 3179 |
term_area = area; continue; } /* * This area fits, move on to the previous one. If * the previous one is the terminal one, we're done. */ area = (area + nr_vms - 1) % nr_vms; if (area == term_area) break; |
68ad4a330 mm/vmalloc.c: kee... |
3180 |
|
ca23e405e vmalloc: implemen... |
3181 3182 |
start = offsets[area]; end = start + sizes[area]; |
68ad4a330 mm/vmalloc.c: kee... |
3183 |
va = pvm_find_va_enclose_addr(base + end); |
ca23e405e vmalloc: implemen... |
3184 |
} |
68ad4a330 mm/vmalloc.c: kee... |
3185 |
|
ca23e405e vmalloc: implemen... |
3186 3187 |
/* we've found a fitting base, insert all va's */ for (area = 0; area < nr_vms; area++) { |
68ad4a330 mm/vmalloc.c: kee... |
3188 |
int ret; |
ca23e405e vmalloc: implemen... |
3189 |
|
68ad4a330 mm/vmalloc.c: kee... |
3190 3191 |
start = base + offsets[area]; size = sizes[area]; |
ca23e405e vmalloc: implemen... |
3192 |
|
68ad4a330 mm/vmalloc.c: kee... |
3193 3194 3195 3196 3197 3198 3199 3200 3201 3202 3203 3204 3205 3206 3207 3208 3209 3210 3211 3212 3213 |
va = pvm_find_va_enclose_addr(start); if (WARN_ON_ONCE(va == NULL)) /* It is a BUG(), but trigger recovery instead. */ goto recovery; type = classify_va_fit_type(va, start, size); if (WARN_ON_ONCE(type == NOTHING_FIT)) /* It is a BUG(), but trigger recovery instead. */ goto recovery; ret = adjust_va_to_fit_type(va, start, size, type); if (unlikely(ret)) goto recovery; /* Allocated area. */ va = vas[area]; va->va_start = start; va->va_end = start + size; insert_vmap_area(va, &vmap_area_root, &vmap_area_list); } |
ca23e405e vmalloc: implemen... |
3214 3215 3216 3217 3218 |
spin_unlock(&vmap_area_lock); /* insert all vm's */ for (area = 0; area < nr_vms; area++) |
3645cb4a4 mm, vmalloc: call... |
3219 3220 |
setup_vmalloc_vm(vms[area], vas[area], VM_ALLOC, pcpu_get_vm_areas); |
ca23e405e vmalloc: implemen... |
3221 3222 3223 |
kfree(vas); return vms; |
68ad4a330 mm/vmalloc.c: kee... |
3224 3225 3226 3227 3228 3229 3230 3231 3232 3233 3234 3235 3236 3237 3238 3239 3240 3241 3242 3243 3244 3245 3246 3247 3248 3249 |
recovery: /* Remove previously inserted areas. */ while (area--) { __free_vmap_area(vas[area]); vas[area] = NULL; } overflow: spin_unlock(&vmap_area_lock); if (!purged) { purge_vmap_area_lazy(); purged = true; /* Before "retry", check if we recover. */ for (area = 0; area < nr_vms; area++) { if (vas[area]) continue; vas[area] = kmem_cache_zalloc( vmap_area_cachep, GFP_KERNEL); if (!vas[area]) goto err_free; } goto retry; } |
ca23e405e vmalloc: implemen... |
3250 3251 |
err_free: for (area = 0; area < nr_vms; area++) { |
68ad4a330 mm/vmalloc.c: kee... |
3252 3253 |
if (vas[area]) kmem_cache_free(vmap_area_cachep, vas[area]); |
f1db7afd9 mm/vmalloc.c: eli... |
3254 |
kfree(vms[area]); |
ca23e405e vmalloc: implemen... |
3255 |
} |
f1db7afd9 mm/vmalloc.c: eli... |
3256 |
err_free2: |
ca23e405e vmalloc: implemen... |
3257 3258 3259 3260 3261 3262 3263 3264 3265 3266 3267 3268 3269 3270 3271 3272 3273 3274 3275 3276 |
kfree(vas); kfree(vms); return NULL; } /** * pcpu_free_vm_areas - free vmalloc areas for percpu allocator * @vms: vm_struct pointer array returned by pcpu_get_vm_areas() * @nr_vms: the number of allocated areas * * Free vm_structs and the array allocated by pcpu_get_vm_areas(). */ void pcpu_free_vm_areas(struct vm_struct **vms, int nr_vms) { int i; for (i = 0; i < nr_vms; i++) free_vm_area(vms[i]); kfree(vms); } |
4f8b02b4e vmalloc: pcpu_get... |
3277 |
#endif /* CONFIG_SMP */ |
a10aa5798 vmalloc: show vma... |
3278 3279 3280 |
#ifdef CONFIG_PROC_FS static void *s_start(struct seq_file *m, loff_t *pos) |
d4033afdf mm, vmalloc: iter... |
3281 |
__acquires(&vmap_area_lock) |
a10aa5798 vmalloc: show vma... |
3282 |
{ |
d4033afdf mm, vmalloc: iter... |
3283 |
spin_lock(&vmap_area_lock); |
3f5000693 mm/vmalloc.c: sim... |
3284 |
return seq_list_start(&vmap_area_list, *pos); |
a10aa5798 vmalloc: show vma... |
3285 3286 3287 3288 |
} static void *s_next(struct seq_file *m, void *p, loff_t *pos) { |
3f5000693 mm/vmalloc.c: sim... |
3289 |
return seq_list_next(p, &vmap_area_list, pos); |
a10aa5798 vmalloc: show vma... |
3290 3291 3292 |
} static void s_stop(struct seq_file *m, void *p) |
d4033afdf mm, vmalloc: iter... |
3293 |
__releases(&vmap_area_lock) |
a10aa5798 vmalloc: show vma... |
3294 |
{ |
d4033afdf mm, vmalloc: iter... |
3295 |
spin_unlock(&vmap_area_lock); |
a10aa5798 vmalloc: show vma... |
3296 |
} |
a47a126ad vmallocinfo: add ... |
3297 3298 |
static void show_numa_info(struct seq_file *m, struct vm_struct *v) { |
e5adfffc8 mm: use IS_ENABLE... |
3299 |
if (IS_ENABLED(CONFIG_NUMA)) { |
a47a126ad vmallocinfo: add ... |
3300 3301 3302 3303 |
unsigned int nr, *counters = m->private; if (!counters) return; |
af12346cd mm/vmalloc: rever... |
3304 3305 |
if (v->flags & VM_UNINITIALIZED) return; |
7e5b528b4 mm/vmalloc.c: fix... |
3306 3307 |
/* Pair with smp_wmb() in clear_vm_uninitialized_flag() */ smp_rmb(); |
af12346cd mm/vmalloc: rever... |
3308 |
|
a47a126ad vmallocinfo: add ... |
3309 3310 3311 3312 3313 3314 3315 3316 3317 3318 |
memset(counters, 0, nr_node_ids * sizeof(unsigned int)); for (nr = 0; nr < v->nr_pages; nr++) counters[page_to_nid(v->pages[nr])]++; for_each_node_state(nr, N_HIGH_MEMORY) if (counters[nr]) seq_printf(m, " N%u=%u", nr, counters[nr]); } } |
dd3b8353b mm/vmalloc: do no... |
3319 3320 3321 3322 3323 3324 3325 3326 3327 3328 3329 3330 3331 3332 3333 3334 |
static void show_purge_info(struct seq_file *m) { struct llist_node *head; struct vmap_area *va; head = READ_ONCE(vmap_purge_list.first); if (head == NULL) return; llist_for_each_entry(va, head, purge_list) { seq_printf(m, "0x%pK-0x%pK %7ld unpurged vm_area ", (void *)va->va_start, (void *)va->va_end, va->va_end - va->va_start); } } |
a10aa5798 vmalloc: show vma... |
3335 3336 |
static int s_show(struct seq_file *m, void *p) { |
3f5000693 mm/vmalloc.c: sim... |
3337 |
struct vmap_area *va; |
d4033afdf mm, vmalloc: iter... |
3338 |
struct vm_struct *v; |
3f5000693 mm/vmalloc.c: sim... |
3339 |
va = list_entry(p, struct vmap_area, list); |
c2ce8c142 mm/vmalloc: fix s... |
3340 |
/* |
688fcbfc0 mm/vmalloc: modif... |
3341 3342 |
* s_show can encounter race with remove_vm_area, !vm on behalf * of vmap area is being tear down or vm_map_ram allocation. |
c2ce8c142 mm/vmalloc: fix s... |
3343 |
*/ |
688fcbfc0 mm/vmalloc: modif... |
3344 |
if (!va->vm) { |
dd3b8353b mm/vmalloc: do no... |
3345 3346 |
seq_printf(m, "0x%pK-0x%pK %7ld vm_map_ram ", |
78c72746f vmalloc: show laz... |
3347 |
(void *)va->va_start, (void *)va->va_end, |
dd3b8353b mm/vmalloc: do no... |
3348 |
va->va_end - va->va_start); |
78c72746f vmalloc: show laz... |
3349 |
|
d4033afdf mm, vmalloc: iter... |
3350 |
return 0; |
78c72746f vmalloc: show laz... |
3351 |
} |
d4033afdf mm, vmalloc: iter... |
3352 3353 |
v = va->vm; |
a10aa5798 vmalloc: show vma... |
3354 |
|
45ec16908 mm: use %pK for /... |
3355 |
seq_printf(m, "0x%pK-0x%pK %7ld", |
a10aa5798 vmalloc: show vma... |
3356 |
v->addr, v->addr + v->size, v->size); |
62c70bce8 mm: convert sprin... |
3357 3358 |
if (v->caller) seq_printf(m, " %pS", v->caller); |
230169693 vmallocinfo: add ... |
3359 |
|
a10aa5798 vmalloc: show vma... |
3360 3361 3362 3363 |
if (v->nr_pages) seq_printf(m, " pages=%d", v->nr_pages); if (v->phys_addr) |
199eaa05a mm: cleanups for ... |
3364 |
seq_printf(m, " phys=%pa", &v->phys_addr); |
a10aa5798 vmalloc: show vma... |
3365 3366 |
if (v->flags & VM_IOREMAP) |
f4527c908 mm/vmalloc.c: rep... |
3367 |
seq_puts(m, " ioremap"); |
a10aa5798 vmalloc: show vma... |
3368 3369 |
if (v->flags & VM_ALLOC) |
f4527c908 mm/vmalloc.c: rep... |
3370 |
seq_puts(m, " vmalloc"); |
a10aa5798 vmalloc: show vma... |
3371 3372 |
if (v->flags & VM_MAP) |
f4527c908 mm/vmalloc.c: rep... |
3373 |
seq_puts(m, " vmap"); |
a10aa5798 vmalloc: show vma... |
3374 3375 |
if (v->flags & VM_USERMAP) |
f4527c908 mm/vmalloc.c: rep... |
3376 |
seq_puts(m, " user"); |
a10aa5798 vmalloc: show vma... |
3377 |
|
fe9041c24 vmalloc: lift the... |
3378 3379 |
if (v->flags & VM_DMA_COHERENT) seq_puts(m, " dma-coherent"); |
244d63ee3 mm, vmalloc: remo... |
3380 |
if (is_vmalloc_addr(v->pages)) |
f4527c908 mm/vmalloc.c: rep... |
3381 |
seq_puts(m, " vpages"); |
a10aa5798 vmalloc: show vma... |
3382 |
|
a47a126ad vmallocinfo: add ... |
3383 |
show_numa_info(m, v); |
a10aa5798 vmalloc: show vma... |
3384 3385 |
seq_putc(m, ' '); |
dd3b8353b mm/vmalloc: do no... |
3386 3387 3388 3389 3390 3391 3392 3393 3394 |
/* * As a final step, dump "unpurged" areas. Note, * that entire "/proc/vmallocinfo" output will not * be address sorted, because the purge list is not * sorted. */ if (list_is_last(&va->list, &vmap_area_list)) show_purge_info(m); |
a10aa5798 vmalloc: show vma... |
3395 3396 |
return 0; } |
5f6a6a9c4 proc: move /proc/... |
3397 |
static const struct seq_operations vmalloc_op = { |
a10aa5798 vmalloc: show vma... |
3398 3399 3400 3401 3402 |
.start = s_start, .next = s_next, .stop = s_stop, .show = s_show, }; |
5f6a6a9c4 proc: move /proc/... |
3403 |
|
5f6a6a9c4 proc: move /proc/... |
3404 3405 |
static int __init proc_vmalloc_init(void) { |
fddda2b7b proc: introduce p... |
3406 |
if (IS_ENABLED(CONFIG_NUMA)) |
0825a6f98 mm: use octal not... |
3407 |
proc_create_seq_private("vmallocinfo", 0400, NULL, |
44414d82c proc: introduce p... |
3408 3409 |
&vmalloc_op, nr_node_ids * sizeof(unsigned int), NULL); |
fddda2b7b proc: introduce p... |
3410 |
else |
0825a6f98 mm: use octal not... |
3411 |
proc_create_seq("vmallocinfo", 0400, NULL, &vmalloc_op); |
5f6a6a9c4 proc: move /proc/... |
3412 3413 3414 |
return 0; } module_init(proc_vmalloc_init); |
db3808c1b mm, vmalloc: move... |
3415 |
|
a10aa5798 vmalloc: show vma... |
3416 |
#endif |