Blame view
mm/vmalloc.c
89.7 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> |
f4f235309 vmalloc: fix rema... |
36 |
#include <linux/overflow.h> |
3b32123d7 mm: use macros fr... |
37 |
|
7c0f6ba68 Replace <asm/uacc... |
38 |
#include <linux/uaccess.h> |
1da177e4c Linux-2.6.12-rc2 |
39 |
#include <asm/tlbflush.h> |
2dca6999e mm, perf_event: M... |
40 |
#include <asm/shmparam.h> |
1da177e4c Linux-2.6.12-rc2 |
41 |
|
dd56b0464 mm: page_alloc: h... |
42 |
#include "internal.h" |
32fcfd407 make vfree() safe... |
43 44 45 46 47 48 49 50 51 52 53 |
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... |
54 55 56 57 |
struct llist_node *t, *llnode; llist_for_each_safe(llnode, t, llist_del_all(&p->list)) __vunmap((void *)llnode, 1); |
32fcfd407 make vfree() safe... |
58 |
} |
db64fe022 mm: rewrite vmap ... |
59 |
/*** Page table manipulation functions ***/ |
b221385bc [PATCH] mm/: make... |
60 |
|
1da177e4c Linux-2.6.12-rc2 |
61 62 63 64 65 66 67 68 69 70 |
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 ... |
71 |
static void vunmap_pmd_range(pud_t *pud, unsigned long addr, unsigned long end) |
1da177e4c Linux-2.6.12-rc2 |
72 73 74 75 76 77 78 |
{ pmd_t *pmd; unsigned long next; pmd = pmd_offset(pud, addr); do { next = pmd_addr_end(addr, end); |
b9820d8f3 mm: change vunmap... |
79 80 |
if (pmd_clear_huge(pmd)) continue; |
1da177e4c Linux-2.6.12-rc2 |
81 82 83 |
if (pmd_none_or_clear_bad(pmd)) continue; vunmap_pte_range(pmd, addr, next); |
75aaa8fa7 mm/vunmap: add co... |
84 85 |
cond_resched(); |
1da177e4c Linux-2.6.12-rc2 |
86 87 |
} while (pmd++, addr = next, addr != end); } |
c2febafc6 mm: convert gener... |
88 |
static void vunmap_pud_range(p4d_t *p4d, unsigned long addr, unsigned long end) |
1da177e4c Linux-2.6.12-rc2 |
89 90 91 |
{ pud_t *pud; unsigned long next; |
c2febafc6 mm: convert gener... |
92 |
pud = pud_offset(p4d, addr); |
1da177e4c Linux-2.6.12-rc2 |
93 94 |
do { next = pud_addr_end(addr, end); |
b9820d8f3 mm: change vunmap... |
95 96 |
if (pud_clear_huge(pud)) continue; |
1da177e4c Linux-2.6.12-rc2 |
97 98 99 100 101 |
if (pud_none_or_clear_bad(pud)) continue; vunmap_pmd_range(pud, addr, next); } while (pud++, addr = next, addr != end); } |
c2febafc6 mm: convert gener... |
102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 |
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 ... |
117 |
static void vunmap_page_range(unsigned long addr, unsigned long end) |
1da177e4c Linux-2.6.12-rc2 |
118 119 120 |
{ pgd_t *pgd; unsigned long next; |
1da177e4c Linux-2.6.12-rc2 |
121 122 123 |
BUG_ON(addr >= end); pgd = pgd_offset_k(addr); |
1da177e4c Linux-2.6.12-rc2 |
124 125 126 127 |
do { next = pgd_addr_end(addr, end); if (pgd_none_or_clear_bad(pgd)) continue; |
c2febafc6 mm: convert gener... |
128 |
vunmap_p4d_range(pgd, addr, next); |
1da177e4c Linux-2.6.12-rc2 |
129 |
} while (pgd++, addr = next, addr != end); |
1da177e4c Linux-2.6.12-rc2 |
130 131 132 |
} static int vmap_pte_range(pmd_t *pmd, unsigned long addr, |
db64fe022 mm: rewrite vmap ... |
133 |
unsigned long end, pgprot_t prot, struct page **pages, int *nr) |
1da177e4c Linux-2.6.12-rc2 |
134 135 |
{ pte_t *pte; |
db64fe022 mm: rewrite vmap ... |
136 137 138 139 |
/* * 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_... |
140 |
pte = pte_alloc_kernel(pmd, addr); |
1da177e4c Linux-2.6.12-rc2 |
141 142 143 |
if (!pte) return -ENOMEM; do { |
db64fe022 mm: rewrite vmap ... |
144 145 146 147 148 |
struct page *page = pages[*nr]; if (WARN_ON(!pte_none(*pte))) return -EBUSY; if (WARN_ON(!page)) |
1da177e4c Linux-2.6.12-rc2 |
149 150 |
return -ENOMEM; set_pte_at(&init_mm, addr, pte, mk_pte(page, prot)); |
db64fe022 mm: rewrite vmap ... |
151 |
(*nr)++; |
1da177e4c Linux-2.6.12-rc2 |
152 153 154 |
} while (pte++, addr += PAGE_SIZE, addr != end); return 0; } |
db64fe022 mm: rewrite vmap ... |
155 156 |
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 |
157 158 159 160 161 162 163 164 165 |
{ 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 ... |
166 |
if (vmap_pte_range(pmd, addr, next, prot, pages, nr)) |
1da177e4c Linux-2.6.12-rc2 |
167 168 169 170 |
return -ENOMEM; } while (pmd++, addr = next, addr != end); return 0; } |
c2febafc6 mm: convert gener... |
171 |
static int vmap_pud_range(p4d_t *p4d, unsigned long addr, |
db64fe022 mm: rewrite vmap ... |
172 |
unsigned long end, pgprot_t prot, struct page **pages, int *nr) |
1da177e4c Linux-2.6.12-rc2 |
173 174 175 |
{ pud_t *pud; unsigned long next; |
c2febafc6 mm: convert gener... |
176 |
pud = pud_alloc(&init_mm, p4d, addr); |
1da177e4c Linux-2.6.12-rc2 |
177 178 179 180 |
if (!pud) return -ENOMEM; do { next = pud_addr_end(addr, end); |
db64fe022 mm: rewrite vmap ... |
181 |
if (vmap_pmd_range(pud, addr, next, prot, pages, nr)) |
1da177e4c Linux-2.6.12-rc2 |
182 183 184 185 |
return -ENOMEM; } while (pud++, addr = next, addr != end); return 0; } |
c2febafc6 mm: convert gener... |
186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 |
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 ... |
202 203 204 205 206 207 |
/* * 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... |
208 209 |
static int vmap_page_range_noflush(unsigned long start, unsigned long end, pgprot_t prot, struct page **pages) |
1da177e4c Linux-2.6.12-rc2 |
210 211 212 |
{ pgd_t *pgd; unsigned long next; |
2e4e27c7d vmalloc.c: fix fl... |
213 |
unsigned long addr = start; |
db64fe022 mm: rewrite vmap ... |
214 215 |
int err = 0; int nr = 0; |
1da177e4c Linux-2.6.12-rc2 |
216 217 218 |
BUG_ON(addr >= end); pgd = pgd_offset_k(addr); |
1da177e4c Linux-2.6.12-rc2 |
219 220 |
do { next = pgd_addr_end(addr, end); |
c2febafc6 mm: convert gener... |
221 |
err = vmap_p4d_range(pgd, addr, next, prot, pages, &nr); |
1da177e4c Linux-2.6.12-rc2 |
222 |
if (err) |
bf88c8c83 vmalloc.c: fix do... |
223 |
return err; |
1da177e4c Linux-2.6.12-rc2 |
224 |
} while (pgd++, addr = next, addr != end); |
db64fe022 mm: rewrite vmap ... |
225 |
|
db64fe022 mm: rewrite vmap ... |
226 |
return nr; |
1da177e4c Linux-2.6.12-rc2 |
227 |
} |
8fc489850 vmalloc: add un/m... |
228 229 230 231 232 233 234 235 236 |
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... |
237 |
int is_vmalloc_or_module_addr(const void *x) |
73bdf0a60 Introduce is_vmal... |
238 239 |
{ /* |
ab4f2ee13 [ARM] fix naming ... |
240 |
* ARM, x86-64 and sparc64 put modules in a special place, |
73bdf0a60 Introduce is_vmal... |
241 242 243 244 245 246 247 248 249 250 |
* 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... |
251 |
/* |
add688fbd Revert "mm/vmallo... |
252 |
* Walk a vmap address to the struct page it maps. |
48667e7a4 Move vmalloc_to_p... |
253 |
*/ |
add688fbd Revert "mm/vmallo... |
254 |
struct page *vmalloc_to_page(const void *vmalloc_addr) |
48667e7a4 Move vmalloc_to_p... |
255 256 |
{ unsigned long addr = (unsigned long) vmalloc_addr; |
add688fbd Revert "mm/vmallo... |
257 |
struct page *page = NULL; |
48667e7a4 Move vmalloc_to_p... |
258 |
pgd_t *pgd = pgd_offset_k(addr); |
c2febafc6 mm: convert gener... |
259 260 261 262 |
p4d_t *p4d; pud_t *pud; pmd_t *pmd; pte_t *ptep, pte; |
48667e7a4 Move vmalloc_to_p... |
263 |
|
7aa413def x86, MM: virtual ... |
264 265 266 267 |
/* * 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... |
268 |
VIRTUAL_BUG_ON(!is_vmalloc_or_module_addr(vmalloc_addr)); |
59ea74633 MM: virtual addre... |
269 |
|
c2febafc6 mm: convert gener... |
270 271 272 273 274 275 |
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... |
276 277 278 279 280 281 282 283 284 285 286 |
/* * 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... |
287 288 |
return NULL; pmd = pmd_offset(pud, addr); |
029c54b09 mm/vmalloc.c: hug... |
289 290 |
WARN_ON_ONCE(pmd_bad(*pmd)); if (pmd_none(*pmd) || pmd_bad(*pmd)) |
c2febafc6 mm: convert gener... |
291 292 293 294 295 296 297 |
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... |
298 |
return page; |
48667e7a4 Move vmalloc_to_p... |
299 |
} |
add688fbd Revert "mm/vmallo... |
300 |
EXPORT_SYMBOL(vmalloc_to_page); |
48667e7a4 Move vmalloc_to_p... |
301 302 |
/* |
add688fbd Revert "mm/vmallo... |
303 |
* Map a vmalloc()-space virtual address to the physical page frame number. |
48667e7a4 Move vmalloc_to_p... |
304 |
*/ |
add688fbd Revert "mm/vmallo... |
305 |
unsigned long vmalloc_to_pfn(const void *vmalloc_addr) |
48667e7a4 Move vmalloc_to_p... |
306 |
{ |
add688fbd Revert "mm/vmallo... |
307 |
return page_to_pfn(vmalloc_to_page(vmalloc_addr)); |
48667e7a4 Move vmalloc_to_p... |
308 |
} |
add688fbd Revert "mm/vmallo... |
309 |
EXPORT_SYMBOL(vmalloc_to_pfn); |
48667e7a4 Move vmalloc_to_p... |
310 |
|
db64fe022 mm: rewrite vmap ... |
311 312 |
/*** Global kva allocator ***/ |
bb850f4da mm/vmap: add DEBU... |
313 |
#define DEBUG_AUGMENT_PROPAGATE_CHECK 0 |
a6cf4e0fe mm/vmap: add DEBU... |
314 |
#define DEBUG_AUGMENT_LOWEST_MATCH_CHECK 0 |
bb850f4da mm/vmap: add DEBU... |
315 |
|
db64fe022 mm: rewrite vmap ... |
316 |
|
db64fe022 mm: rewrite vmap ... |
317 |
static DEFINE_SPINLOCK(vmap_area_lock); |
f1c4069e1 mm, vmalloc: expo... |
318 319 |
/* Export for kexec only */ LIST_HEAD(vmap_area_list); |
80c4bd7a5 mm/vmalloc: keep ... |
320 |
static LLIST_HEAD(vmap_purge_list); |
89699605f mm: vmap area cache |
321 |
static struct rb_root vmap_area_root = RB_ROOT; |
68ad4a330 mm/vmalloc.c: kee... |
322 |
static bool vmap_initialized __read_mostly; |
89699605f mm: vmap area cache |
323 |
|
68ad4a330 mm/vmalloc.c: kee... |
324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 |
/* * 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... |
349 350 351 352 353 354 |
/* * 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... |
355 356 357 358 359 360 361 362 363 364 365 366 367 368 |
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 |
369 |
|
68ad4a330 mm/vmalloc.c: kee... |
370 371 372 373 374 375 376 377 378 379 |
/* * 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:... |
380 381 |
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... |
382 383 384 385 |
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 ... |
386 |
|
97105f0ab mm: vmalloc: show... |
387 388 389 390 391 392 |
static atomic_long_t nr_vmalloc_pages; unsigned long vmalloc_nr_pages(void) { return atomic_long_read(&nr_vmalloc_pages); } |
db64fe022 mm: rewrite vmap ... |
393 |
static struct vmap_area *__find_vmap_area(unsigned long addr) |
1da177e4c Linux-2.6.12-rc2 |
394 |
{ |
db64fe022 mm: rewrite vmap ... |
395 396 397 398 399 400 401 402 |
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... |
403 |
else if (addr >= va->va_end) |
db64fe022 mm: rewrite vmap ... |
404 405 406 407 408 409 410 |
n = n->rb_right; else return va; } return NULL; } |
68ad4a330 mm/vmalloc.c: kee... |
411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 |
/* * 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 ... |
432 |
|
68ad4a330 mm/vmalloc.c: kee... |
433 434 435 436 437 438 439 |
/* * 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 ... |
440 |
|
68ad4a330 mm/vmalloc.c: kee... |
441 442 443 444 445 446 447 448 449 450 451 |
/* * 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 ... |
452 453 |
else BUG(); |
68ad4a330 mm/vmalloc.c: kee... |
454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 |
} 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 ... |
490 |
} |
68ad4a330 mm/vmalloc.c: kee... |
491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 |
/* 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 ... |
511 |
|
68ad4a330 mm/vmalloc.c: kee... |
512 513 |
/* Address-sort this list */ list_add(&va->list, head); |
db64fe022 mm: rewrite vmap ... |
514 |
} |
68ad4a330 mm/vmalloc.c: kee... |
515 516 517 |
static __always_inline void unlink_va(struct vmap_area *va, struct rb_root *root) { |
460e42d19 mm/vmalloc.c: swi... |
518 519 |
if (WARN_ON(RB_EMPTY_NODE(&va->rb_node))) return; |
db64fe022 mm: rewrite vmap ... |
520 |
|
460e42d19 mm/vmalloc.c: swi... |
521 522 523 524 525 526 527 528 |
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... |
529 |
} |
bb850f4da mm/vmap: add DEBU... |
530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 |
#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... |
572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 |
/* * 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... |
621 622 623 624 |
#if DEBUG_AUGMENT_PROPAGATE_CHECK augment_tree_propagate_check(free_vmap_area_root.rb_node); #endif |
68ad4a330 mm/vmalloc.c: kee... |
625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 |
} 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... |
698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 |
/* 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... |
721 722 |
if (merged) unlink_va(va, root); |
68ad4a330 mm/vmalloc.c: kee... |
723 724 725 |
/* Free vmap_area object. */ kmem_cache_free(vmap_area_cachep, va); |
68ad4a330 mm/vmalloc.c: kee... |
726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 |
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... |
796 |
* OK. We roll back and find the first right sub-tree, |
68ad4a330 mm/vmalloc.c: kee... |
797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 |
* 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... |
816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 |
#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... |
854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 |
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... |
893 |
struct vmap_area *lva = NULL; |
68ad4a330 mm/vmalloc.c: kee... |
894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 |
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... |
931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 |
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... |
949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 |
/* * 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... |
966 |
if (lva) /* type == NE_FIT_TYPE */ |
68ad4a330 mm/vmalloc.c: kee... |
967 968 969 970 971 972 973 974 975 976 977 978 979 |
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... |
980 |
unsigned long vstart, unsigned long vend) |
68ad4a330 mm/vmalloc.c: kee... |
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 1006 1007 1008 |
{ 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... |
1009 1010 1011 |
#if DEBUG_AUGMENT_LOWEST_MATCH_CHECK find_vmap_lowest_match_check(size); #endif |
68ad4a330 mm/vmalloc.c: kee... |
1012 1013 |
return nva_start_addr; } |
4da56b99d mm/vmap: Add a no... |
1014 |
|
db64fe022 mm: rewrite vmap ... |
1015 1016 1017 1018 1019 1020 1021 1022 1023 |
/* * 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... |
1024 |
struct vmap_area *va, *pva; |
1da177e4c Linux-2.6.12-rc2 |
1025 |
unsigned long addr; |
db64fe022 mm: rewrite vmap ... |
1026 |
int purged = 0; |
7766970cc mm: vmap fix over... |
1027 |
BUG_ON(!size); |
891c49abf mm/vmalloc: use o... |
1028 |
BUG_ON(offset_in_page(size)); |
89699605f mm: vmap area cache |
1029 |
BUG_ON(!is_power_of_2(align)); |
db64fe022 mm: rewrite vmap ... |
1030 |
|
68ad4a330 mm/vmalloc.c: kee... |
1031 1032 |
if (unlikely(!vmap_initialized)) return ERR_PTR(-EBUSY); |
5803ed292 mm: mark all call... |
1033 |
might_sleep(); |
4da56b99d mm/vmap: Add a no... |
1034 |
|
68ad4a330 mm/vmalloc.c: kee... |
1035 |
va = kmem_cache_alloc_node(vmap_area_cachep, |
db64fe022 mm: rewrite vmap ... |
1036 1037 1038 |
gfp_mask & GFP_RECLAIM_MASK, node); if (unlikely(!va)) return ERR_PTR(-ENOMEM); |
7f88f88f8 mm: kmemleak: avo... |
1039 1040 1041 1042 1043 |
/* * 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 ... |
1044 |
retry: |
82dd23e84 mm/vmalloc.c: pre... |
1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 |
/* * 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 ... |
1068 |
spin_lock(&vmap_area_lock); |
82dd23e84 mm/vmalloc.c: pre... |
1069 |
preempt_enable(); |
89699605f mm: vmap area cache |
1070 |
|
afd07389d mm/vmalloc.c: fix... |
1071 |
/* |
68ad4a330 mm/vmalloc.c: kee... |
1072 1073 |
* If an allocation fails, the "vend" address is * returned. Therefore trigger the overflow path. |
afd07389d mm/vmalloc.c: fix... |
1074 |
*/ |
cacca6baf mm/vmalloc.c: rem... |
1075 |
addr = __alloc_vmap_area(size, align, vstart, vend); |
68ad4a330 mm/vmalloc.c: kee... |
1076 |
if (unlikely(addr == vend)) |
89699605f mm: vmap area cache |
1077 |
goto overflow; |
db64fe022 mm: rewrite vmap ... |
1078 1079 1080 |
va->va_start = addr; va->va_end = addr + size; |
688fcbfc0 mm/vmalloc: modif... |
1081 |
va->vm = NULL; |
68ad4a330 mm/vmalloc.c: kee... |
1082 |
insert_vmap_area(va, &vmap_area_root, &vmap_area_list); |
db64fe022 mm: rewrite vmap ... |
1083 |
spin_unlock(&vmap_area_lock); |
61e165578 mm/vmalloc.c: use... |
1084 |
BUG_ON(!IS_ALIGNED(va->va_start, align)); |
89699605f mm: vmap area cache |
1085 1086 |
BUG_ON(va->va_start < vstart); BUG_ON(va->va_end > vend); |
db64fe022 mm: rewrite vmap ... |
1087 |
return va; |
89699605f mm: vmap area cache |
1088 1089 1090 1091 1092 1093 1094 1095 |
overflow: spin_unlock(&vmap_area_lock); if (!purged) { purge_vmap_area_lazy(); purged = 1; goto retry; } |
4da56b99d mm/vmap: Add a no... |
1096 1097 1098 1099 1100 1101 1102 1103 1104 |
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(... |
1105 |
if (!(gfp_mask & __GFP_NOWARN) && printk_ratelimit()) |
756a025f0 mm: coalesce spli... |
1106 1107 1108 |
pr_warn("vmap allocation for size %lu failed: use vmalloc=<size> to increase size ", size); |
68ad4a330 mm/vmalloc.c: kee... |
1109 1110 |
kmem_cache_free(vmap_area_cachep, va); |
89699605f mm: vmap area cache |
1111 |
return ERR_PTR(-EBUSY); |
db64fe022 mm: rewrite vmap ... |
1112 |
} |
4da56b99d mm/vmap: Add a no... |
1113 1114 1115 1116 1117 1118 1119 1120 1121 1122 1123 |
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 ... |
1124 1125 |
static void __free_vmap_area(struct vmap_area *va) { |
ca23e405e vmalloc: implemen... |
1126 |
/* |
68ad4a330 mm/vmalloc.c: kee... |
1127 |
* Remove from the busy tree/list. |
ca23e405e vmalloc: implemen... |
1128 |
*/ |
68ad4a330 mm/vmalloc.c: kee... |
1129 |
unlink_va(va, &vmap_area_root); |
ca23e405e vmalloc: implemen... |
1130 |
|
68ad4a330 mm/vmalloc.c: kee... |
1131 1132 1133 1134 1135 |
/* * 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 ... |
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 1177 1178 1179 |
} /* * 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... |
1180 |
static atomic_long_t vmap_lazy_nr = ATOMIC_LONG_INIT(0); |
db64fe022 mm: rewrite vmap ... |
1181 |
|
0574ecd14 mm: refactor __pu... |
1182 1183 1184 1185 1186 |
/* * 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... |
1187 |
static DEFINE_MUTEX(vmap_purge_lock); |
0574ecd14 mm: refactor __pu... |
1188 |
|
02b709df8 mm: purge fragmen... |
1189 1190 |
/* for per-CPU blocks */ static void purge_fragmented_blocks_allcpus(void); |
db64fe022 mm: rewrite vmap ... |
1191 |
/* |
3ee48b6af mm, x86: Saving v... |
1192 1193 1194 1195 1196 |
* 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... |
1197 |
atomic_long_set(&vmap_lazy_nr, lazy_max_pages()+1); |
3ee48b6af mm, x86: Saving v... |
1198 1199 1200 |
} /* |
db64fe022 mm: rewrite vmap ... |
1201 |
* Purges all lazily-freed vmap areas. |
db64fe022 mm: rewrite vmap ... |
1202 |
*/ |
0574ecd14 mm: refactor __pu... |
1203 |
static bool __purge_vmap_area_lazy(unsigned long start, unsigned long end) |
db64fe022 mm: rewrite vmap ... |
1204 |
{ |
4d36e6f80 mm/vmalloc.c: con... |
1205 |
unsigned long resched_threshold; |
80c4bd7a5 mm/vmalloc: keep ... |
1206 |
struct llist_node *valist; |
db64fe022 mm: rewrite vmap ... |
1207 |
struct vmap_area *va; |
cbb766766 mm: fix lazy vmap... |
1208 |
struct vmap_area *n_va; |
db64fe022 mm: rewrite vmap ... |
1209 |
|
0574ecd14 mm: refactor __pu... |
1210 |
lockdep_assert_held(&vmap_purge_lock); |
02b709df8 mm: purge fragmen... |
1211 |
|
80c4bd7a5 mm/vmalloc: keep ... |
1212 |
valist = llist_del_all(&vmap_purge_list); |
68571be99 mm/vmalloc.c: add... |
1213 1214 1215 1216 |
if (unlikely(valist == NULL)) return false; /* |
3f8fd02b1 mm/vmalloc: Sync ... |
1217 1218 1219 |
* First make sure the mappings are removed from all page-tables * before they are freed. */ |
66f28e110 x86/mm: split vma... |
1220 |
vmalloc_sync_unmappings(); |
3f8fd02b1 mm/vmalloc: Sync ... |
1221 1222 |
/* |
68571be99 mm/vmalloc.c: add... |
1223 1224 1225 |
* TODO: to calculate a flush range without looping. * The list can be up to lazy_max_pages() elements. */ |
80c4bd7a5 mm/vmalloc: keep ... |
1226 |
llist_for_each_entry(va, valist, purge_list) { |
0574ecd14 mm: refactor __pu... |
1227 1228 1229 1230 |
if (va->va_start < start) start = va->va_start; if (va->va_end > end) end = va->va_end; |
db64fe022 mm: rewrite vmap ... |
1231 |
} |
db64fe022 mm: rewrite vmap ... |
1232 |
|
0574ecd14 mm: refactor __pu... |
1233 |
flush_tlb_kernel_range(start, end); |
4d36e6f80 mm/vmalloc.c: con... |
1234 |
resched_threshold = lazy_max_pages() << 1; |
db64fe022 mm: rewrite vmap ... |
1235 |
|
0574ecd14 mm: refactor __pu... |
1236 |
spin_lock(&vmap_area_lock); |
763b218dd mm: add preempt p... |
1237 |
llist_for_each_entry_safe(va, n_va, valist, purge_list) { |
4d36e6f80 mm/vmalloc.c: con... |
1238 |
unsigned long nr = (va->va_end - va->va_start) >> PAGE_SHIFT; |
763b218dd mm: add preempt p... |
1239 |
|
dd3b8353b mm/vmalloc: do no... |
1240 1241 1242 1243 1244 1245 1246 |
/* * 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... |
1247 |
atomic_long_sub(nr, &vmap_lazy_nr); |
68571be99 mm/vmalloc.c: add... |
1248 |
|
4d36e6f80 mm/vmalloc.c: con... |
1249 |
if (atomic_long_read(&vmap_lazy_nr) < resched_threshold) |
68571be99 mm/vmalloc.c: add... |
1250 |
cond_resched_lock(&vmap_area_lock); |
763b218dd mm: add preempt p... |
1251 |
} |
0574ecd14 mm: refactor __pu... |
1252 1253 |
spin_unlock(&vmap_area_lock); return true; |
db64fe022 mm: rewrite vmap ... |
1254 1255 1256 |
} /* |
496850e5f mm: vmalloc failu... |
1257 1258 1259 1260 1261 |
* 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... |
1262 |
if (mutex_trylock(&vmap_purge_lock)) { |
0574ecd14 mm: refactor __pu... |
1263 |
__purge_vmap_area_lazy(ULONG_MAX, 0); |
f9e099776 mm: turn vmap_pur... |
1264 |
mutex_unlock(&vmap_purge_lock); |
0574ecd14 mm: refactor __pu... |
1265 |
} |
496850e5f mm: vmalloc failu... |
1266 1267 1268 |
} /* |
db64fe022 mm: rewrite vmap ... |
1269 1270 1271 1272 |
* Kick off a purge of the outstanding lazy areas. */ static void purge_vmap_area_lazy(void) { |
f9e099776 mm: turn vmap_pur... |
1273 |
mutex_lock(&vmap_purge_lock); |
0574ecd14 mm: refactor __pu... |
1274 1275 |
purge_fragmented_blocks_allcpus(); __purge_vmap_area_lazy(ULONG_MAX, 0); |
f9e099776 mm: turn vmap_pur... |
1276 |
mutex_unlock(&vmap_purge_lock); |
db64fe022 mm: rewrite vmap ... |
1277 1278 1279 |
} /* |
64141da58 vmalloc: eagerly ... |
1280 1281 1282 |
* 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 ... |
1283 |
*/ |
64141da58 vmalloc: eagerly ... |
1284 |
static void free_vmap_area_noflush(struct vmap_area *va) |
db64fe022 mm: rewrite vmap ... |
1285 |
{ |
4d36e6f80 mm/vmalloc.c: con... |
1286 |
unsigned long nr_lazy; |
80c4bd7a5 mm/vmalloc: keep ... |
1287 |
|
dd3b8353b mm/vmalloc: do no... |
1288 1289 1290 |
spin_lock(&vmap_area_lock); unlink_va(va, &vmap_area_root); spin_unlock(&vmap_area_lock); |
4d36e6f80 mm/vmalloc.c: con... |
1291 1292 |
nr_lazy = atomic_long_add_return((va->va_end - va->va_start) >> PAGE_SHIFT, &vmap_lazy_nr); |
80c4bd7a5 mm/vmalloc: keep ... |
1293 1294 1295 1296 1297 |
/* 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... |
1298 |
try_purge_vmap_area_lazy(); |
db64fe022 mm: rewrite vmap ... |
1299 |
} |
b29acbdcf mm: vmalloc fix l... |
1300 1301 1302 1303 1304 1305 |
/* * 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... |
1306 |
unmap_vmap_area(va); |
d30dce351 mm, debug_pageall... |
1307 |
if (debug_pagealloc_enabled_static()) |
82a2e924f mm: vmalloc: clea... |
1308 |
flush_tlb_kernel_range(va->va_start, va->va_end); |
c8eef01e2 mm: remove free_u... |
1309 |
free_vmap_area_noflush(va); |
b29acbdcf mm: vmalloc fix l... |
1310 |
} |
db64fe022 mm: rewrite vmap ... |
1311 1312 1313 1314 1315 1316 1317 1318 1319 1320 |
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 ... |
1321 1322 1323 1324 1325 1326 1327 1328 1329 1330 1331 1332 1333 1334 1335 1336 1337 1338 1339 1340 1341 1342 1343 |
/*** 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... |
1344 1345 1346 1347 |
#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 ... |
1348 1349 1350 1351 1352 1353 |
#define VMAP_BLOCK_SIZE (VMAP_BBMAP_BITS * PAGE_SIZE) struct vmap_block_queue { spinlock_t lock; struct list_head free; |
db64fe022 mm: rewrite vmap ... |
1354 1355 1356 1357 1358 |
}; struct vmap_block { spinlock_t lock; struct vmap_area *va; |
db64fe022 mm: rewrite vmap ... |
1359 |
unsigned long free, dirty; |
7d61bfe8f mm/vmalloc: get r... |
1360 |
unsigned long dirty_min, dirty_max; /*< dirty range */ |
de5604231 mm: percpu-vmap f... |
1361 1362 |
struct list_head free_list; struct rcu_head rcu_head; |
02b709df8 mm: purge fragmen... |
1363 |
struct list_head purge; |
db64fe022 mm: rewrite vmap ... |
1364 1365 1366 1367 1368 1369 1370 1371 1372 1373 1374 1375 1376 1377 1378 1379 1380 1381 1382 1383 1384 1385 1386 1387 1388 1389 |
}; /* 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... |
1390 1391 1392 1393 1394 1395 1396 1397 1398 1399 1400 1401 1402 1403 1404 |
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:... |
1405 |
* Return: virtual address in a newly allocated block or ERR_PTR(-errno) |
cf725ce27 mm/vmalloc: occup... |
1406 1407 |
*/ static void *new_vmap_block(unsigned int order, gfp_t gfp_mask) |
db64fe022 mm: rewrite vmap ... |
1408 1409 1410 1411 1412 1413 |
{ struct vmap_block_queue *vbq; struct vmap_block *vb; struct vmap_area *va; unsigned long vb_idx; int node, err; |
cf725ce27 mm/vmalloc: occup... |
1414 |
void *vaddr; |
db64fe022 mm: rewrite vmap ... |
1415 1416 1417 1418 1419 1420 1421 1422 1423 1424 1425 |
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... |
1426 |
if (IS_ERR(va)) { |
db64fe022 mm: rewrite vmap ... |
1427 |
kfree(vb); |
e7d863407 mm: use ERR_CAST |
1428 |
return ERR_CAST(va); |
db64fe022 mm: rewrite vmap ... |
1429 1430 1431 1432 1433 1434 1435 1436 |
} err = radix_tree_preload(gfp_mask); if (unlikely(err)) { kfree(vb); free_vmap_area(va); return ERR_PTR(err); } |
cf725ce27 mm/vmalloc: occup... |
1437 |
vaddr = vmap_block_vaddr(va->va_start, 0); |
db64fe022 mm: rewrite vmap ... |
1438 1439 |
spin_lock_init(&vb->lock); vb->va = va; |
cf725ce27 mm/vmalloc: occup... |
1440 1441 1442 |
/* 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 ... |
1443 |
vb->dirty = 0; |
7d61bfe8f mm/vmalloc: get r... |
1444 1445 |
vb->dirty_min = VMAP_BBMAP_BITS; vb->dirty_max = 0; |
db64fe022 mm: rewrite vmap ... |
1446 |
INIT_LIST_HEAD(&vb->free_list); |
db64fe022 mm: rewrite vmap ... |
1447 1448 1449 1450 1451 1452 1453 1454 1455 |
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 ... |
1456 |
spin_lock(&vbq->lock); |
68ac546f2 mm/vmalloc: fix p... |
1457 |
list_add_tail_rcu(&vb->free_list, &vbq->free); |
db64fe022 mm: rewrite vmap ... |
1458 |
spin_unlock(&vbq->lock); |
3f04ba859 vmalloc: fix use ... |
1459 |
put_cpu_var(vmap_block_queue); |
db64fe022 mm: rewrite vmap ... |
1460 |
|
cf725ce27 mm/vmalloc: occup... |
1461 |
return vaddr; |
db64fe022 mm: rewrite vmap ... |
1462 |
} |
db64fe022 mm: rewrite vmap ... |
1463 1464 1465 1466 |
static void free_vmap_block(struct vmap_block *vb) { struct vmap_block *tmp; unsigned long vb_idx; |
db64fe022 mm: rewrite vmap ... |
1467 1468 1469 1470 1471 |
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 ... |
1472 |
free_vmap_area_noflush(vb->va); |
22a3c7d18 vmalloc,rcu: Conv... |
1473 |
kfree_rcu(vb, rcu_head); |
db64fe022 mm: rewrite vmap ... |
1474 |
} |
02b709df8 mm: purge fragmen... |
1475 1476 1477 1478 1479 1480 1481 1482 1483 1484 1485 1486 1487 1488 1489 1490 1491 |
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... |
1492 1493 |
vb->dirty_min = 0; vb->dirty_max = VMAP_BBMAP_BITS; |
02b709df8 mm: purge fragmen... |
1494 1495 1496 1497 1498 1499 1500 1501 1502 1503 1504 1505 1506 1507 1508 |
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... |
1509 1510 1511 1512 1513 1514 1515 |
static void purge_fragmented_blocks_allcpus(void) { int cpu; for_each_possible_cpu(cpu) purge_fragmented_blocks(cpu); } |
db64fe022 mm: rewrite vmap ... |
1516 1517 1518 1519 |
static void *vb_alloc(unsigned long size, gfp_t gfp_mask) { struct vmap_block_queue *vbq; struct vmap_block *vb; |
cf725ce27 mm/vmalloc: occup... |
1520 |
void *vaddr = NULL; |
db64fe022 mm: rewrite vmap ... |
1521 |
unsigned int order; |
891c49abf mm/vmalloc: use o... |
1522 |
BUG_ON(offset_in_page(size)); |
db64fe022 mm: rewrite vmap ... |
1523 |
BUG_ON(size > PAGE_SIZE*VMAP_MAX_ALLOC); |
aa91c4d89 mm: make vb_alloc... |
1524 1525 1526 1527 1528 1529 1530 1531 |
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 ... |
1532 |
order = get_order(size); |
db64fe022 mm: rewrite vmap ... |
1533 1534 1535 |
rcu_read_lock(); vbq = &get_cpu_var(vmap_block_queue); list_for_each_entry_rcu(vb, &vbq->free, free_list) { |
cf725ce27 mm/vmalloc: occup... |
1536 |
unsigned long pages_off; |
db64fe022 mm: rewrite vmap ... |
1537 1538 |
spin_lock(&vb->lock); |
cf725ce27 mm/vmalloc: occup... |
1539 1540 1541 1542 |
if (vb->free < (1UL << order)) { spin_unlock(&vb->lock); continue; } |
02b709df8 mm: purge fragmen... |
1543 |
|
cf725ce27 mm/vmalloc: occup... |
1544 1545 |
pages_off = VMAP_BBMAP_BITS - vb->free; vaddr = vmap_block_vaddr(vb->va->va_start, pages_off); |
02b709df8 mm: purge fragmen... |
1546 1547 1548 1549 1550 1551 |
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... |
1552 |
|
02b709df8 mm: purge fragmen... |
1553 1554 |
spin_unlock(&vb->lock); break; |
db64fe022 mm: rewrite vmap ... |
1555 |
} |
02b709df8 mm: purge fragmen... |
1556 |
|
3f04ba859 vmalloc: fix use ... |
1557 |
put_cpu_var(vmap_block_queue); |
db64fe022 mm: rewrite vmap ... |
1558 |
rcu_read_unlock(); |
cf725ce27 mm/vmalloc: occup... |
1559 1560 1561 |
/* Allocate new block if nothing was found */ if (!vaddr) vaddr = new_vmap_block(order, gfp_mask); |
db64fe022 mm: rewrite vmap ... |
1562 |
|
cf725ce27 mm/vmalloc: occup... |
1563 |
return vaddr; |
db64fe022 mm: rewrite vmap ... |
1564 1565 1566 1567 1568 1569 1570 1571 |
} 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... |
1572 |
BUG_ON(offset_in_page(size)); |
db64fe022 mm: rewrite vmap ... |
1573 |
BUG_ON(size > PAGE_SIZE*VMAP_MAX_ALLOC); |
b29acbdcf mm: vmalloc fix l... |
1574 1575 |
flush_cache_vunmap((unsigned long)addr, (unsigned long)addr + size); |
db64fe022 mm: rewrite vmap ... |
1576 1577 1578 |
order = get_order(size); offset = (unsigned long)addr & (VMAP_BLOCK_SIZE - 1); |
7d61bfe8f mm/vmalloc: get r... |
1579 |
offset >>= PAGE_SHIFT; |
db64fe022 mm: rewrite vmap ... |
1580 1581 1582 1583 1584 1585 |
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 ... |
1586 |
vunmap_page_range((unsigned long)addr, (unsigned long)addr + size); |
d30dce351 mm, debug_pageall... |
1587 |
if (debug_pagealloc_enabled_static()) |
82a2e924f mm: vmalloc: clea... |
1588 1589 |
flush_tlb_kernel_range((unsigned long)addr, (unsigned long)addr + size); |
db64fe022 mm: rewrite vmap ... |
1590 |
spin_lock(&vb->lock); |
7d61bfe8f mm/vmalloc: get r... |
1591 1592 1593 1594 |
/* 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... |
1595 |
|
db64fe022 mm: rewrite vmap ... |
1596 1597 |
vb->dirty += 1UL << order; if (vb->dirty == VMAP_BBMAP_BITS) { |
de5604231 mm: percpu-vmap f... |
1598 |
BUG_ON(vb->free); |
db64fe022 mm: rewrite vmap ... |
1599 1600 1601 1602 1603 |
spin_unlock(&vb->lock); free_vmap_block(vb); } else spin_unlock(&vb->lock); } |
868b104d7 mm/vmalloc: Add f... |
1604 |
static void _vm_unmap_aliases(unsigned long start, unsigned long end, int flush) |
db64fe022 mm: rewrite vmap ... |
1605 |
{ |
db64fe022 mm: rewrite vmap ... |
1606 |
int cpu; |
db64fe022 mm: rewrite vmap ... |
1607 |
|
9b4633340 vmap: cope with v... |
1608 1609 |
if (unlikely(!vmap_initialized)) return; |
5803ed292 mm: mark all call... |
1610 |
might_sleep(); |
db64fe022 mm: rewrite vmap ... |
1611 1612 1613 1614 1615 1616 |
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 ... |
1617 |
spin_lock(&vb->lock); |
7d61bfe8f mm/vmalloc: get r... |
1618 1619 |
if (vb->dirty) { unsigned long va_start = vb->va->va_start; |
db64fe022 mm: rewrite vmap ... |
1620 |
unsigned long s, e; |
b136be5e0 mm, vmalloc: use ... |
1621 |
|
7d61bfe8f mm/vmalloc: get r... |
1622 1623 |
s = va_start + (vb->dirty_min << PAGE_SHIFT); e = va_start + (vb->dirty_max << PAGE_SHIFT); |
db64fe022 mm: rewrite vmap ... |
1624 |
|
7d61bfe8f mm/vmalloc: get r... |
1625 1626 |
start = min(s, start); end = max(e, end); |
db64fe022 mm: rewrite vmap ... |
1627 |
|
7d61bfe8f mm/vmalloc: get r... |
1628 |
flush = 1; |
db64fe022 mm: rewrite vmap ... |
1629 1630 1631 1632 1633 |
} spin_unlock(&vb->lock); } rcu_read_unlock(); } |
f9e099776 mm: turn vmap_pur... |
1634 |
mutex_lock(&vmap_purge_lock); |
0574ecd14 mm: refactor __pu... |
1635 1636 1637 |
purge_fragmented_blocks_allcpus(); if (!__purge_vmap_area_lazy(start, end) && flush) flush_tlb_kernel_range(start, end); |
f9e099776 mm: turn vmap_pur... |
1638 |
mutex_unlock(&vmap_purge_lock); |
db64fe022 mm: rewrite vmap ... |
1639 |
} |
868b104d7 mm/vmalloc: Add f... |
1640 1641 1642 1643 1644 1645 1646 1647 1648 1649 1650 1651 1652 1653 1654 1655 1656 1657 1658 1659 1660 |
/** * 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 ... |
1661 1662 1663 1664 1665 1666 1667 1668 1669 |
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 ... |
1670 |
unsigned long size = (unsigned long)count << PAGE_SHIFT; |
db64fe022 mm: rewrite vmap ... |
1671 |
unsigned long addr = (unsigned long)mem; |
9c3acf604 mm: remove free_u... |
1672 |
struct vmap_area *va; |
db64fe022 mm: rewrite vmap ... |
1673 |
|
5803ed292 mm: mark all call... |
1674 |
might_sleep(); |
db64fe022 mm: rewrite vmap ... |
1675 1676 1677 |
BUG_ON(!addr); BUG_ON(addr < VMALLOC_START); BUG_ON(addr > VMALLOC_END); |
a1c0b1a07 mm/vmalloc: use P... |
1678 |
BUG_ON(!PAGE_ALIGNED(addr)); |
db64fe022 mm: rewrite vmap ... |
1679 |
|
9c3acf604 mm: remove free_u... |
1680 |
if (likely(count <= VMAP_MAX_ALLOC)) { |
05e3ff950 mm: vmalloc: pass... |
1681 |
debug_check_no_locks_freed(mem, size); |
db64fe022 mm: rewrite vmap ... |
1682 |
vb_free(mem, size); |
9c3acf604 mm: remove free_u... |
1683 1684 1685 1686 1687 |
return; } va = find_vmap_area(addr); BUG_ON(!va); |
05e3ff950 mm: vmalloc: pass... |
1688 1689 |
debug_check_no_locks_freed((void *)va->va_start, (va->va_end - va->va_start)); |
9c3acf604 mm: remove free_u... |
1690 |
free_unmap_vmap_area(va); |
db64fe022 mm: rewrite vmap ... |
1691 1692 1693 1694 1695 1696 1697 1698 1699 |
} 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... |
1700 |
* |
364376383 mm/vmalloc.c: enh... |
1701 1702 1703 1704 1705 1706 |
* 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... |
1707 |
* Returns: a pointer to the address that has been mapped, or %NULL on failure |
db64fe022 mm: rewrite vmap ... |
1708 1709 1710 |
*/ void *vm_map_ram(struct page **pages, unsigned int count, int node, pgprot_t prot) { |
65ee03c4b mm: fix overflow ... |
1711 |
unsigned long size = (unsigned long)count << PAGE_SHIFT; |
db64fe022 mm: rewrite vmap ... |
1712 1713 1714 1715 1716 1717 1718 1719 1720 1721 1722 1723 1724 1725 1726 1727 1728 1729 1730 1731 1732 1733 1734 1735 1736 |
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... |
1737 |
static struct vm_struct *vmlist __initdata; |
92eac1681 docs/mm: vmalloc:... |
1738 |
|
f0aa66179 vmalloc: implemen... |
1739 |
/** |
be9b7335e mm: add vm_area_a... |
1740 1741 1742 1743 1744 1745 1746 1747 1748 1749 1750 1751 1752 1753 1754 1755 1756 1757 1758 1759 1760 1761 1762 1763 1764 1765 |
* 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... |
1766 1767 |
* vm_area_register_early - register vmap area early during boot * @vm: vm_struct to register |
c0c0a2937 vmalloc: add @ali... |
1768 |
* @align: requested alignment |
f0aa66179 vmalloc: implemen... |
1769 1770 1771 1772 1773 1774 1775 1776 |
* * 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... |
1777 |
void __init vm_area_register_early(struct vm_struct *vm, size_t align) |
f0aa66179 vmalloc: implemen... |
1778 1779 |
{ static size_t vm_init_off __initdata; |
c0c0a2937 vmalloc: add @ali... |
1780 1781 1782 1783 |
unsigned long addr; addr = ALIGN(VMALLOC_START + vm_init_off, align); vm_init_off = PFN_ALIGN(addr + vm->size) - VMALLOC_START; |
f0aa66179 vmalloc: implemen... |
1784 |
|
c0c0a2937 vmalloc: add @ali... |
1785 |
vm->addr = (void *)addr; |
f0aa66179 vmalloc: implemen... |
1786 |
|
be9b7335e mm: add vm_area_a... |
1787 |
vm_area_add_early(vm); |
f0aa66179 vmalloc: implemen... |
1788 |
} |
68ad4a330 mm/vmalloc.c: kee... |
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 1826 1827 1828 |
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 ... |
1829 1830 |
void __init vmalloc_init(void) { |
822c18f2e alpha: fix vmallo... |
1831 1832 |
struct vmap_area *va; struct vm_struct *tmp; |
db64fe022 mm: rewrite vmap ... |
1833 |
int i; |
68ad4a330 mm/vmalloc.c: kee... |
1834 1835 1836 1837 |
/* * Create the cache for vmap_area objects. */ vmap_area_cachep = KMEM_CACHE(vmap_area, SLAB_PANIC); |
db64fe022 mm: rewrite vmap ... |
1838 1839 |
for_each_possible_cpu(i) { struct vmap_block_queue *vbq; |
32fcfd407 make vfree() safe... |
1840 |
struct vfree_deferred *p; |
db64fe022 mm: rewrite vmap ... |
1841 1842 1843 1844 |
vbq = &per_cpu(vmap_block_queue, i); spin_lock_init(&vbq->lock); INIT_LIST_HEAD(&vbq->free); |
32fcfd407 make vfree() safe... |
1845 1846 1847 |
p = &per_cpu(vfree_deferred, i); init_llist_head(&p->list); INIT_WORK(&p->wq, free_work); |
db64fe022 mm: rewrite vmap ... |
1848 |
} |
9b4633340 vmap: cope with v... |
1849 |
|
822c18f2e alpha: fix vmallo... |
1850 1851 |
/* Import existing vmlist entries. */ for (tmp = vmlist; tmp; tmp = tmp->next) { |
68ad4a330 mm/vmalloc.c: kee... |
1852 1853 1854 |
va = kmem_cache_zalloc(vmap_area_cachep, GFP_NOWAIT); if (WARN_ON_ONCE(!va)) continue; |
822c18f2e alpha: fix vmallo... |
1855 1856 |
va->va_start = (unsigned long)tmp->addr; va->va_end = va->va_start + tmp->size; |
dbda591d9 mm: fix faulty in... |
1857 |
va->vm = tmp; |
68ad4a330 mm/vmalloc.c: kee... |
1858 |
insert_vmap_area(va, &vmap_area_root, &vmap_area_list); |
822c18f2e alpha: fix vmallo... |
1859 |
} |
ca23e405e vmalloc: implemen... |
1860 |
|
68ad4a330 mm/vmalloc.c: kee... |
1861 1862 1863 1864 |
/* * Now we can initialize a free vmap space. */ vmap_init_free_space(); |
9b4633340 vmap: cope with v... |
1865 |
vmap_initialized = true; |
db64fe022 mm: rewrite vmap ... |
1866 |
} |
8fc489850 vmalloc: add un/m... |
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 1907 1908 1909 |
/** * 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... |
1910 |
EXPORT_SYMBOL_GPL(unmap_kernel_range_noflush); |
8fc489850 vmalloc: add un/m... |
1911 1912 1913 1914 1915 1916 1917 1918 1919 |
/** * 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 ... |
1920 1921 1922 |
void unmap_kernel_range(unsigned long addr, unsigned long size) { unsigned long end = addr + size; |
f6fcba701 vmalloc: call flu... |
1923 1924 |
flush_cache_vunmap(addr, end); |
db64fe022 mm: rewrite vmap ... |
1925 1926 1927 |
vunmap_page_range(addr, end); flush_tlb_kernel_range(addr, end); } |
93ef6d6ca mm/vmalloc.c: exp... |
1928 |
EXPORT_SYMBOL_GPL(unmap_kernel_range); |
db64fe022 mm: rewrite vmap ... |
1929 |
|
f6f8ed473 mm/vmalloc.c: cle... |
1930 |
int map_vm_area(struct vm_struct *area, pgprot_t prot, struct page **pages) |
db64fe022 mm: rewrite vmap ... |
1931 1932 |
{ unsigned long addr = (unsigned long)area->addr; |
762216ab4 mm/vmalloc: use w... |
1933 |
unsigned long end = addr + get_vm_area_size(area); |
db64fe022 mm: rewrite vmap ... |
1934 |
int err; |
f6f8ed473 mm/vmalloc.c: cle... |
1935 |
err = vmap_page_range(addr, end, prot, pages); |
db64fe022 mm: rewrite vmap ... |
1936 |
|
f6f8ed473 mm/vmalloc.c: cle... |
1937 |
return err > 0 ? 0 : err; |
db64fe022 mm: rewrite vmap ... |
1938 1939 |
} EXPORT_SYMBOL_GPL(map_vm_area); |
f5252e009 mm: avoid null po... |
1940 |
static void setup_vmalloc_vm(struct vm_struct *vm, struct vmap_area *va, |
5e6cafc83 mm: vmalloc: use ... |
1941 |
unsigned long flags, const void *caller) |
cf88c7900 vmalloc: separate... |
1942 |
{ |
c69480ade mm, vmalloc: prot... |
1943 |
spin_lock(&vmap_area_lock); |
cf88c7900 vmalloc: separate... |
1944 1945 1946 1947 |
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... |
1948 |
va->vm = vm; |
c69480ade mm, vmalloc: prot... |
1949 |
spin_unlock(&vmap_area_lock); |
f5252e009 mm: avoid null po... |
1950 |
} |
cf88c7900 vmalloc: separate... |
1951 |
|
20fc02b47 mm/vmalloc.c: ren... |
1952 |
static void clear_vm_uninitialized_flag(struct vm_struct *vm) |
f5252e009 mm: avoid null po... |
1953 |
{ |
d4033afdf mm, vmalloc: iter... |
1954 |
/* |
20fc02b47 mm/vmalloc.c: ren... |
1955 |
* Before removing VM_UNINITIALIZED, |
d4033afdf mm, vmalloc: iter... |
1956 1957 1958 1959 |
* we should make sure that vm has proper values. * Pair with smp_rmb() in show_numa_info(). */ smp_wmb(); |
20fc02b47 mm/vmalloc.c: ren... |
1960 |
vm->flags &= ~VM_UNINITIALIZED; |
cf88c7900 vmalloc: separate... |
1961 |
} |
db64fe022 mm: rewrite vmap ... |
1962 |
static struct vm_struct *__get_vm_area_node(unsigned long size, |
2dca6999e mm, perf_event: M... |
1963 |
unsigned long align, unsigned long flags, unsigned long start, |
5e6cafc83 mm: vmalloc: use ... |
1964 |
unsigned long end, int node, gfp_t gfp_mask, const void *caller) |
db64fe022 mm: rewrite vmap ... |
1965 |
{ |
0006526d7 mm/vmalloc.c: rem... |
1966 |
struct vmap_area *va; |
db64fe022 mm: rewrite vmap ... |
1967 |
struct vm_struct *area; |
1da177e4c Linux-2.6.12-rc2 |
1968 |
|
52fd24ca1 [PATCH] __vmalloc... |
1969 |
BUG_ON(in_interrupt()); |
1da177e4c Linux-2.6.12-rc2 |
1970 |
size = PAGE_ALIGN(size); |
31be83095 [PATCH] Fix stran... |
1971 1972 |
if (unlikely(!size)) return NULL; |
1da177e4c Linux-2.6.12-rc2 |
1973 |
|
252e5c6e2 mm/vmalloc.c: fix... |
1974 1975 1976 |
if (flags & VM_IOREMAP) align = 1ul << clamp_t(int, get_count_order_long(size), PAGE_SHIFT, IOREMAP_MAX_ORDER); |
cf88c7900 vmalloc: separate... |
1977 |
area = kzalloc_node(sizeof(*area), gfp_mask & GFP_RECLAIM_MASK, node); |
1da177e4c Linux-2.6.12-rc2 |
1978 1979 |
if (unlikely(!area)) return NULL; |
71394fe50 mm: vmalloc: add ... |
1980 1981 |
if (!(flags & VM_NO_GUARD)) size += PAGE_SIZE; |
1da177e4c Linux-2.6.12-rc2 |
1982 |
|
db64fe022 mm: rewrite vmap ... |
1983 1984 1985 1986 |
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 |
1987 |
} |
1da177e4c Linux-2.6.12-rc2 |
1988 |
|
d82b1d857 mm, vmalloc: only... |
1989 |
setup_vmalloc_vm(area, va, flags, caller); |
f5252e009 mm: avoid null po... |
1990 |
|
1da177e4c Linux-2.6.12-rc2 |
1991 |
return area; |
1da177e4c Linux-2.6.12-rc2 |
1992 |
} |
930fc45a4 [PATCH] vmalloc_node |
1993 1994 1995 |
struct vm_struct *__get_vm_area(unsigned long size, unsigned long flags, unsigned long start, unsigned long end) { |
00ef2d2f8 mm: use NUMA_NO_NODE |
1996 1997 |
return __get_vm_area_node(size, 1, flags, start, end, NUMA_NO_NODE, GFP_KERNEL, __builtin_return_address(0)); |
930fc45a4 [PATCH] vmalloc_node |
1998 |
} |
5992b6dac lguest: export sy... |
1999 |
EXPORT_SYMBOL_GPL(__get_vm_area); |
930fc45a4 [PATCH] vmalloc_node |
2000 |
|
c29686129 vmalloc: add __ge... |
2001 2002 |
struct vm_struct *__get_vm_area_caller(unsigned long size, unsigned long flags, unsigned long start, unsigned long end, |
5e6cafc83 mm: vmalloc: use ... |
2003 |
const void *caller) |
c29686129 vmalloc: add __ge... |
2004 |
{ |
00ef2d2f8 mm: use NUMA_NO_NODE |
2005 2006 |
return __get_vm_area_node(size, 1, flags, start, end, NUMA_NO_NODE, GFP_KERNEL, caller); |
c29686129 vmalloc: add __ge... |
2007 |
} |
1da177e4c Linux-2.6.12-rc2 |
2008 |
/** |
92eac1681 docs/mm: vmalloc:... |
2009 2010 2011 |
* 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 |
2012 |
* |
92eac1681 docs/mm: vmalloc:... |
2013 2014 2015 |
* 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:... |
2016 2017 |
* * Return: the area descriptor on success or %NULL on failure. |
1da177e4c Linux-2.6.12-rc2 |
2018 2019 2020 |
*/ struct vm_struct *get_vm_area(unsigned long size, unsigned long flags) { |
2dca6999e mm, perf_event: M... |
2021 |
return __get_vm_area_node(size, 1, flags, VMALLOC_START, VMALLOC_END, |
00ef2d2f8 mm: use NUMA_NO_NODE |
2022 2023 |
NUMA_NO_NODE, GFP_KERNEL, __builtin_return_address(0)); |
230169693 vmallocinfo: add ... |
2024 2025 2026 |
} struct vm_struct *get_vm_area_caller(unsigned long size, unsigned long flags, |
5e6cafc83 mm: vmalloc: use ... |
2027 |
const void *caller) |
230169693 vmallocinfo: add ... |
2028 |
{ |
2dca6999e mm, perf_event: M... |
2029 |
return __get_vm_area_node(size, 1, flags, VMALLOC_START, VMALLOC_END, |
00ef2d2f8 mm: use NUMA_NO_NODE |
2030 |
NUMA_NO_NODE, GFP_KERNEL, caller); |
1da177e4c Linux-2.6.12-rc2 |
2031 |
} |
e9da6e990 ARM: dma-mapping:... |
2032 |
/** |
92eac1681 docs/mm: vmalloc:... |
2033 2034 |
* find_vm_area - find a continuous kernel virtual area * @addr: base address |
e9da6e990 ARM: dma-mapping:... |
2035 |
* |
92eac1681 docs/mm: vmalloc:... |
2036 2037 2038 |
* 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:... |
2039 2040 |
* * Return: pointer to the found area or %NULL on faulure |
e9da6e990 ARM: dma-mapping:... |
2041 2042 |
*/ struct vm_struct *find_vm_area(const void *addr) |
833423143 [PATCH] mm: intro... |
2043 |
{ |
db64fe022 mm: rewrite vmap ... |
2044 |
struct vmap_area *va; |
833423143 [PATCH] mm: intro... |
2045 |
|
db64fe022 mm: rewrite vmap ... |
2046 |
va = find_vmap_area((unsigned long)addr); |
688fcbfc0 mm/vmalloc: modif... |
2047 2048 |
if (!va) return NULL; |
1da177e4c Linux-2.6.12-rc2 |
2049 |
|
688fcbfc0 mm/vmalloc: modif... |
2050 |
return va->vm; |
1da177e4c Linux-2.6.12-rc2 |
2051 |
} |
7856dfeb2 [PATCH] x86_64: F... |
2052 |
/** |
92eac1681 docs/mm: vmalloc:... |
2053 2054 |
* remove_vm_area - find and remove a continuous kernel virtual area * @addr: base address |
7856dfeb2 [PATCH] x86_64: F... |
2055 |
* |
92eac1681 docs/mm: vmalloc:... |
2056 2057 2058 |
* 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:... |
2059 2060 |
* * Return: pointer to the found area or %NULL on faulure |
7856dfeb2 [PATCH] x86_64: F... |
2061 |
*/ |
b3bdda02a vmalloc: add cons... |
2062 |
struct vm_struct *remove_vm_area(const void *addr) |
7856dfeb2 [PATCH] x86_64: F... |
2063 |
{ |
db64fe022 mm: rewrite vmap ... |
2064 |
struct vmap_area *va; |
5803ed292 mm: mark all call... |
2065 |
might_sleep(); |
dd3b8353b mm/vmalloc: do no... |
2066 2067 |
spin_lock(&vmap_area_lock); va = __find_vmap_area((unsigned long)addr); |
688fcbfc0 mm/vmalloc: modif... |
2068 |
if (va && va->vm) { |
db1aecafe mm/vmalloc.c: cha... |
2069 |
struct vm_struct *vm = va->vm; |
f5252e009 mm: avoid null po... |
2070 |
|
c69480ade mm, vmalloc: prot... |
2071 |
va->vm = NULL; |
c69480ade mm, vmalloc: prot... |
2072 |
spin_unlock(&vmap_area_lock); |
a5af5aa8b kasan, module, vm... |
2073 |
kasan_free_shadow(vm); |
dd32c2799 vmalloc: unmap vm... |
2074 |
free_unmap_vmap_area(va); |
dd32c2799 vmalloc: unmap vm... |
2075 |
|
db64fe022 mm: rewrite vmap ... |
2076 2077 |
return vm; } |
dd3b8353b mm/vmalloc: do no... |
2078 2079 |
spin_unlock(&vmap_area_lock); |
db64fe022 mm: rewrite vmap ... |
2080 |
return NULL; |
7856dfeb2 [PATCH] x86_64: F... |
2081 |
} |
868b104d7 mm/vmalloc: Add f... |
2082 2083 2084 2085 2086 2087 2088 2089 2090 2091 2092 2093 2094 |
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... |
2095 2096 |
unsigned long start = ULONG_MAX, end = 0; int flush_reset = area->flags & VM_FLUSH_RESET_PERMS; |
31e67340c mm/vmalloc: Avoid... |
2097 |
int flush_dmap = 0; |
868b104d7 mm/vmalloc: Add f... |
2098 |
int i; |
868b104d7 mm/vmalloc: Add f... |
2099 2100 2101 2102 2103 2104 2105 2106 2107 2108 2109 2110 2111 2112 2113 2114 2115 2116 2117 2118 2119 |
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... |
2120 2121 |
unsigned long addr = (unsigned long)page_address(area->pages[i]); if (addr) { |
868b104d7 mm/vmalloc: Add f... |
2122 |
start = min(addr, start); |
8e41f8726 mm/vmalloc: Fix c... |
2123 |
end = max(addr + PAGE_SIZE, end); |
31e67340c mm/vmalloc: Avoid... |
2124 |
flush_dmap = 1; |
868b104d7 mm/vmalloc: Add f... |
2125 2126 2127 2128 2129 2130 2131 2132 2133 |
} } /* * 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... |
2134 |
_vm_unmap_aliases(start, end, flush_dmap); |
868b104d7 mm/vmalloc: Add f... |
2135 2136 |
set_area_direct_map(area, set_direct_map_default_noflush); } |
b3bdda02a vmalloc: add cons... |
2137 |
static void __vunmap(const void *addr, int deallocate_pages) |
1da177e4c Linux-2.6.12-rc2 |
2138 2139 2140 2141 2142 |
{ struct vm_struct *area; if (!addr) return; |
e69e9d4ae vmalloc: introduc... |
2143 2144 |
if (WARN(!PAGE_ALIGNED(addr), "Trying to vfree() bad address (%p) ", |
ab15d9b4c mm/vmalloc.c: unb... |
2145 |
addr)) |
1da177e4c Linux-2.6.12-rc2 |
2146 |
return; |
1da177e4c Linux-2.6.12-rc2 |
2147 |
|
6ade20327 mm/vmalloc.c: don... |
2148 |
area = find_vm_area(addr); |
1da177e4c Linux-2.6.12-rc2 |
2149 |
if (unlikely(!area)) { |
4c8573e25 Use WARN() in mm/... |
2150 2151 |
WARN(1, KERN_ERR "Trying to vfree() nonexistent vm area (%p) ", |
1da177e4c Linux-2.6.12-rc2 |
2152 |
addr); |
1da177e4c Linux-2.6.12-rc2 |
2153 2154 |
return; } |
05e3ff950 mm: vmalloc: pass... |
2155 2156 |
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: ... |
2157 |
|
868b104d7 mm/vmalloc: Add f... |
2158 |
vm_remove_mappings(area, deallocate_pages); |
1da177e4c Linux-2.6.12-rc2 |
2159 2160 2161 2162 |
if (deallocate_pages) { int i; for (i = 0; i < area->nr_pages; i++) { |
bf53d6f8f vmalloc: clean up... |
2163 2164 2165 |
struct page *page = area->pages[i]; BUG_ON(!page); |
4949148ad mm: charge/unchar... |
2166 |
__free_pages(page, 0); |
1da177e4c Linux-2.6.12-rc2 |
2167 |
} |
97105f0ab mm: vmalloc: show... |
2168 |
atomic_long_sub(area->nr_pages, &nr_vmalloc_pages); |
1da177e4c Linux-2.6.12-rc2 |
2169 |
|
244d63ee3 mm, vmalloc: remo... |
2170 |
kvfree(area->pages); |
1da177e4c Linux-2.6.12-rc2 |
2171 2172 2173 2174 2175 |
} kfree(area); return; } |
bf22e37a6 mm: add vfree_ato... |
2176 2177 2178 2179 2180 2181 2182 2183 2184 2185 2186 2187 2188 2189 2190 2191 |
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:... |
2192 2193 |
* vfree_atomic - release memory allocated by vmalloc() * @addr: memory base address |
bf22e37a6 mm: add vfree_ato... |
2194 |
* |
92eac1681 docs/mm: vmalloc:... |
2195 2196 |
* This one is just like vfree() but can be called in any atomic context * except NMIs. |
bf22e37a6 mm: add vfree_ato... |
2197 2198 2199 2200 2201 2202 2203 2204 2205 2206 2207 |
*/ void vfree_atomic(const void *addr) { BUG_ON(in_nmi()); kmemleak_free(addr); if (!addr) return; __vfree_deferred(addr); } |
c67dc6247 mm/vmalloc: do no... |
2208 2209 2210 2211 2212 2213 2214 |
static void __vfree(const void *addr) { if (unlikely(in_interrupt())) __vfree_deferred(addr); else __vunmap(addr, 1); } |
1da177e4c Linux-2.6.12-rc2 |
2215 |
/** |
92eac1681 docs/mm: vmalloc:... |
2216 2217 |
* vfree - release memory allocated by vmalloc() * @addr: memory base address |
1da177e4c Linux-2.6.12-rc2 |
2218 |
* |
92eac1681 docs/mm: vmalloc:... |
2219 2220 2221 |
* 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 |
2222 |
* |
92eac1681 docs/mm: vmalloc:... |
2223 2224 2225 |
* 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... |
2226 |
* |
92eac1681 docs/mm: vmalloc:... |
2227 |
* May sleep if called *not* from interrupt context. |
3ca4ea3a7 mm/vmalloc.c: imp... |
2228 |
* |
92eac1681 docs/mm: vmalloc:... |
2229 |
* NOTE: assumes that the object at @addr has a size >= sizeof(llist_node) |
1da177e4c Linux-2.6.12-rc2 |
2230 |
*/ |
b3bdda02a vmalloc: add cons... |
2231 |
void vfree(const void *addr) |
1da177e4c Linux-2.6.12-rc2 |
2232 |
{ |
32fcfd407 make vfree() safe... |
2233 |
BUG_ON(in_nmi()); |
89219d37a kmemleak: Add the... |
2234 2235 |
kmemleak_free(addr); |
a8dda165e vfree: add debug ... |
2236 |
might_sleep_if(!in_interrupt()); |
32fcfd407 make vfree() safe... |
2237 2238 |
if (!addr) return; |
c67dc6247 mm/vmalloc: do no... |
2239 2240 |
__vfree(addr); |
1da177e4c Linux-2.6.12-rc2 |
2241 |
} |
1da177e4c Linux-2.6.12-rc2 |
2242 2243 2244 |
EXPORT_SYMBOL(vfree); /** |
92eac1681 docs/mm: vmalloc:... |
2245 2246 |
* vunmap - release virtual mapping obtained by vmap() * @addr: memory base address |
1da177e4c Linux-2.6.12-rc2 |
2247 |
* |
92eac1681 docs/mm: vmalloc:... |
2248 2249 |
* 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 |
2250 |
* |
92eac1681 docs/mm: vmalloc:... |
2251 |
* Must not be called in interrupt context. |
1da177e4c Linux-2.6.12-rc2 |
2252 |
*/ |
b3bdda02a vmalloc: add cons... |
2253 |
void vunmap(const void *addr) |
1da177e4c Linux-2.6.12-rc2 |
2254 2255 |
{ BUG_ON(in_interrupt()); |
34754b69a x86: make vmap ye... |
2256 |
might_sleep(); |
32fcfd407 make vfree() safe... |
2257 2258 |
if (addr) __vunmap(addr, 0); |
1da177e4c Linux-2.6.12-rc2 |
2259 |
} |
1da177e4c Linux-2.6.12-rc2 |
2260 2261 2262 |
EXPORT_SYMBOL(vunmap); /** |
92eac1681 docs/mm: vmalloc:... |
2263 2264 2265 2266 2267 2268 2269 2270 |
* 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:... |
2271 2272 |
* * Return: the address of the area or %NULL on failure |
1da177e4c Linux-2.6.12-rc2 |
2273 2274 |
*/ void *vmap(struct page **pages, unsigned int count, |
92eac1681 docs/mm: vmalloc:... |
2275 |
unsigned long flags, pgprot_t prot) |
1da177e4c Linux-2.6.12-rc2 |
2276 2277 |
{ struct vm_struct *area; |
65ee03c4b mm: fix overflow ... |
2278 |
unsigned long size; /* In bytes */ |
1da177e4c Linux-2.6.12-rc2 |
2279 |
|
34754b69a x86: make vmap ye... |
2280 |
might_sleep(); |
ca79b0c21 mm: convert total... |
2281 |
if (count > totalram_pages()) |
1da177e4c Linux-2.6.12-rc2 |
2282 |
return NULL; |
65ee03c4b mm: fix overflow ... |
2283 2284 |
size = (unsigned long)count << PAGE_SHIFT; area = get_vm_area_caller(size, flags, __builtin_return_address(0)); |
1da177e4c Linux-2.6.12-rc2 |
2285 2286 |
if (!area) return NULL; |
230169693 vmallocinfo: add ... |
2287 |
|
f6f8ed473 mm/vmalloc.c: cle... |
2288 |
if (map_vm_area(area, prot, pages)) { |
1da177e4c Linux-2.6.12-rc2 |
2289 2290 2291 2292 2293 2294 |
vunmap(area->addr); return NULL; } return area->addr; } |
1da177e4c Linux-2.6.12-rc2 |
2295 |
EXPORT_SYMBOL(vmap); |
8594a21cf mm, vmalloc: fix ... |
2296 2297 2298 |
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... |
2299 |
static void *__vmalloc_area_node(struct vm_struct *area, gfp_t gfp_mask, |
3722e13cf mm/vmalloc: don't... |
2300 |
pgprot_t prot, int node) |
1da177e4c Linux-2.6.12-rc2 |
2301 2302 2303 |
{ struct page **pages; unsigned int nr_pages, array_size, i; |
930f036b4 mm, vmalloc: cons... |
2304 |
const gfp_t nested_gfp = (gfp_mask & GFP_RECLAIM_MASK) | __GFP_ZERO; |
704b862f9 mm/vmalloc.c: don... |
2305 2306 2307 2308 |
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 |
2309 |
|
762216ab4 mm/vmalloc: use w... |
2310 |
nr_pages = get_vm_area_size(area) >> PAGE_SHIFT; |
1da177e4c Linux-2.6.12-rc2 |
2311 |
array_size = (nr_pages * sizeof(struct page *)); |
1da177e4c Linux-2.6.12-rc2 |
2312 |
/* Please note that the recursion is strictly bounded. */ |
8757d5fa6 [PATCH] mm: fix o... |
2313 |
if (array_size > PAGE_SIZE) { |
704b862f9 mm/vmalloc.c: don... |
2314 |
pages = __vmalloc_node(array_size, 1, nested_gfp|highmem_mask, |
3722e13cf mm/vmalloc: don't... |
2315 |
PAGE_KERNEL, node, area->caller); |
286e1ea3a [PATCH] vmalloc()... |
2316 |
} else { |
976d6dfbb vmalloc(): adjust... |
2317 |
pages = kmalloc_node(array_size, nested_gfp, node); |
286e1ea3a [PATCH] vmalloc()... |
2318 |
} |
7ea362427 mm/vmalloc.c: mov... |
2319 2320 |
if (!pages) { |
1da177e4c Linux-2.6.12-rc2 |
2321 2322 2323 2324 |
remove_vm_area(area->addr); kfree(area); return NULL; } |
1da177e4c Linux-2.6.12-rc2 |
2325 |
|
7ea362427 mm/vmalloc.c: mov... |
2326 2327 |
area->pages = pages; area->nr_pages = nr_pages; |
1da177e4c Linux-2.6.12-rc2 |
2328 |
for (i = 0; i < area->nr_pages; i++) { |
bf53d6f8f vmalloc: clean up... |
2329 |
struct page *page; |
4b90951c0 mm/vmalloc: use N... |
2330 |
if (node == NUMA_NO_NODE) |
704b862f9 mm/vmalloc.c: don... |
2331 |
page = alloc_page(alloc_mask|highmem_mask); |
930fc45a4 [PATCH] vmalloc_node |
2332 |
else |
704b862f9 mm/vmalloc.c: don... |
2333 |
page = alloc_pages_node(node, alloc_mask|highmem_mask, 0); |
bf53d6f8f vmalloc: clean up... |
2334 2335 |
if (unlikely(!page)) { |
1da177e4c Linux-2.6.12-rc2 |
2336 2337 |
/* Successfully allocated i pages, free them in __vunmap() */ area->nr_pages = i; |
97105f0ab mm: vmalloc: show... |
2338 |
atomic_long_add(area->nr_pages, &nr_vmalloc_pages); |
1da177e4c Linux-2.6.12-rc2 |
2339 2340 |
goto fail; } |
bf53d6f8f vmalloc: clean up... |
2341 |
area->pages[i] = page; |
704b862f9 mm/vmalloc.c: don... |
2342 |
if (gfpflags_allow_blocking(gfp_mask|highmem_mask)) |
660654f90 mm/vmalloc.c: add... |
2343 |
cond_resched(); |
1da177e4c Linux-2.6.12-rc2 |
2344 |
} |
97105f0ab mm: vmalloc: show... |
2345 |
atomic_long_add(area->nr_pages, &nr_vmalloc_pages); |
1da177e4c Linux-2.6.12-rc2 |
2346 |
|
f6f8ed473 mm/vmalloc.c: cle... |
2347 |
if (map_vm_area(area, prot, pages)) |
1da177e4c Linux-2.6.12-rc2 |
2348 2349 2350 2351 |
goto fail; return area->addr; fail: |
a8e99259e mm, page_alloc: w... |
2352 |
warn_alloc(gfp_mask, NULL, |
7877cdcc3 mm: consolidate w... |
2353 |
"vmalloc: allocation failure, allocated %ld of %ld bytes", |
22943ab11 mm: print vmalloc... |
2354 |
(area->nr_pages*PAGE_SIZE), area->size); |
c67dc6247 mm/vmalloc: do no... |
2355 |
__vfree(area->addr); |
1da177e4c Linux-2.6.12-rc2 |
2356 2357 2358 2359 |
return NULL; } /** |
92eac1681 docs/mm: vmalloc:... |
2360 2361 2362 2363 2364 2365 2366 2367 2368 2369 2370 2371 2372 2373 |
* __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:... |
2374 2375 |
* * Return: the address of the area or %NULL on failure |
1da177e4c Linux-2.6.12-rc2 |
2376 |
*/ |
d0a21265d mm: unify module_... |
2377 2378 |
void *__vmalloc_node_range(unsigned long size, unsigned long align, unsigned long start, unsigned long end, gfp_t gfp_mask, |
cb9e3c292 mm: vmalloc: pass... |
2379 2380 |
pgprot_t prot, unsigned long vm_flags, int node, const void *caller) |
1da177e4c Linux-2.6.12-rc2 |
2381 2382 |
{ struct vm_struct *area; |
89219d37a kmemleak: Add the... |
2383 2384 |
void *addr; unsigned long real_size = size; |
1da177e4c Linux-2.6.12-rc2 |
2385 2386 |
size = PAGE_ALIGN(size); |
ca79b0c21 mm: convert total... |
2387 |
if (!size || (size >> PAGE_SHIFT) > totalram_pages()) |
de7d2b567 mm/vmalloc.c: rep... |
2388 |
goto fail; |
1da177e4c Linux-2.6.12-rc2 |
2389 |
|
cb9e3c292 mm: vmalloc: pass... |
2390 2391 |
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 |
2392 |
if (!area) |
de7d2b567 mm/vmalloc.c: rep... |
2393 |
goto fail; |
1da177e4c Linux-2.6.12-rc2 |
2394 |
|
3722e13cf mm/vmalloc: don't... |
2395 |
addr = __vmalloc_area_node(area, gfp_mask, prot, node); |
1368edf06 mm: vmalloc: chec... |
2396 |
if (!addr) |
b82225f3f revert mm/vmalloc... |
2397 |
return NULL; |
89219d37a kmemleak: Add the... |
2398 2399 |
/* |
20fc02b47 mm/vmalloc.c: ren... |
2400 2401 |
* In this function, newly allocated vm_struct has VM_UNINITIALIZED * flag. It means that vm_struct is not fully initialized. |
4341fa454 mm, vmalloc: remo... |
2402 |
* Now, it is fully initialized, so remove this flag here. |
f5252e009 mm: avoid null po... |
2403 |
*/ |
20fc02b47 mm/vmalloc.c: ren... |
2404 |
clear_vm_uninitialized_flag(area); |
f5252e009 mm: avoid null po... |
2405 |
|
94f4a1618 mm: kmemleak: tre... |
2406 |
kmemleak_vmalloc(area, size, gfp_mask); |
89219d37a kmemleak: Add the... |
2407 2408 |
return addr; |
de7d2b567 mm/vmalloc.c: rep... |
2409 2410 |
fail: |
a8e99259e mm, page_alloc: w... |
2411 |
warn_alloc(gfp_mask, NULL, |
7877cdcc3 mm: consolidate w... |
2412 |
"vmalloc: allocation failure: %lu bytes", real_size); |
de7d2b567 mm/vmalloc.c: rep... |
2413 |
return NULL; |
1da177e4c Linux-2.6.12-rc2 |
2414 |
} |
153178edc vmalloc: export _... |
2415 2416 2417 2418 2419 2420 2421 2422 |
/* * 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_... |
2423 |
/** |
92eac1681 docs/mm: vmalloc:... |
2424 2425 2426 2427 2428 2429 2430 |
* __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[... |
2431 |
* |
92eac1681 docs/mm: vmalloc:... |
2432 2433 2434 |
* 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[... |
2435 |
* |
92eac1681 docs/mm: vmalloc:... |
2436 2437 |
* Reclaim modifiers in @gfp_mask - __GFP_NORETRY, __GFP_RETRY_MAYFAIL * and __GFP_NOFAIL are not supported |
a7c3e901a mm: introduce kv[... |
2438 |
* |
92eac1681 docs/mm: vmalloc:... |
2439 2440 |
* Any use of gfp flags outside of GFP_KERNEL should be consulted * with mm people. |
a862f68a8 docs/core-api/mm:... |
2441 2442 |
* * Return: pointer to the allocated memory or %NULL on error |
d0a21265d mm: unify module_... |
2443 |
*/ |
8594a21cf mm, vmalloc: fix ... |
2444 |
static void *__vmalloc_node(unsigned long size, unsigned long align, |
d0a21265d mm: unify module_... |
2445 |
gfp_t gfp_mask, pgprot_t prot, |
5e6cafc83 mm: vmalloc: use ... |
2446 |
int node, const void *caller) |
d0a21265d mm: unify module_... |
2447 2448 |
{ return __vmalloc_node_range(size, align, VMALLOC_START, VMALLOC_END, |
cb9e3c292 mm: vmalloc: pass... |
2449 |
gfp_mask, prot, 0, node, caller); |
d0a21265d mm: unify module_... |
2450 |
} |
930fc45a4 [PATCH] vmalloc_node |
2451 2452 |
void *__vmalloc(unsigned long size, gfp_t gfp_mask, pgprot_t prot) { |
00ef2d2f8 mm: use NUMA_NO_NODE |
2453 |
return __vmalloc_node(size, 1, gfp_mask, prot, NUMA_NO_NODE, |
230169693 vmallocinfo: add ... |
2454 |
__builtin_return_address(0)); |
930fc45a4 [PATCH] vmalloc_node |
2455 |
} |
1da177e4c Linux-2.6.12-rc2 |
2456 |
EXPORT_SYMBOL(__vmalloc); |
8594a21cf mm, vmalloc: fix ... |
2457 2458 2459 2460 2461 2462 2463 2464 2465 2466 2467 2468 2469 |
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 |
2470 |
/** |
92eac1681 docs/mm: vmalloc:... |
2471 2472 2473 2474 2475 |
* 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 |
2476 |
* |
92eac1681 docs/mm: vmalloc:... |
2477 2478 |
* For tight control over page level allocator and protection flags * use __vmalloc() instead. |
a862f68a8 docs/core-api/mm:... |
2479 2480 |
* * Return: pointer to the allocated memory or %NULL on error |
1da177e4c Linux-2.6.12-rc2 |
2481 2482 2483 |
*/ void *vmalloc(unsigned long size) { |
00ef2d2f8 mm: use NUMA_NO_NODE |
2484 |
return __vmalloc_node_flags(size, NUMA_NO_NODE, |
19809c2da mm, vmalloc: use ... |
2485 |
GFP_KERNEL); |
1da177e4c Linux-2.6.12-rc2 |
2486 |
} |
1da177e4c Linux-2.6.12-rc2 |
2487 |
EXPORT_SYMBOL(vmalloc); |
930fc45a4 [PATCH] vmalloc_node |
2488 |
/** |
92eac1681 docs/mm: vmalloc:... |
2489 2490 2491 2492 2493 2494 2495 2496 2497 |
* 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:... |
2498 2499 |
* * Return: pointer to the allocated memory or %NULL on error |
e1ca7788d mm: add vzalloc()... |
2500 2501 2502 |
*/ void *vzalloc(unsigned long size) { |
00ef2d2f8 mm: use NUMA_NO_NODE |
2503 |
return __vmalloc_node_flags(size, NUMA_NO_NODE, |
19809c2da mm, vmalloc: use ... |
2504 |
GFP_KERNEL | __GFP_ZERO); |
e1ca7788d mm: add vzalloc()... |
2505 2506 2507 2508 |
} EXPORT_SYMBOL(vzalloc); /** |
ead04089b [PATCH] Fix kerne... |
2509 2510 |
* vmalloc_user - allocate zeroed virtually contiguous memory for userspace * @size: allocation size |
833423143 [PATCH] mm: intro... |
2511 |
* |
ead04089b [PATCH] Fix kerne... |
2512 2513 |
* The resulting memory area is zeroed so it can be mapped to userspace * without leaking data. |
a862f68a8 docs/core-api/mm:... |
2514 2515 |
* * Return: pointer to the allocated memory or %NULL on error |
833423143 [PATCH] mm: intro... |
2516 2517 2518 |
*/ void *vmalloc_user(unsigned long size) { |
bc84c5352 mm/vmalloc: pass ... |
2519 2520 2521 2522 |
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... |
2523 2524 2525 2526 |
} EXPORT_SYMBOL(vmalloc_user); /** |
92eac1681 docs/mm: vmalloc:... |
2527 2528 2529 |
* vmalloc_node - allocate memory on a specific node * @size: allocation size * @node: numa node |
930fc45a4 [PATCH] vmalloc_node |
2530 |
* |
92eac1681 docs/mm: vmalloc:... |
2531 2532 |
* Allocate enough pages to cover @size from the page level * allocator and map them into contiguous kernel virtual space. |
930fc45a4 [PATCH] vmalloc_node |
2533 |
* |
92eac1681 docs/mm: vmalloc:... |
2534 2535 |
* For tight control over page level allocator and protection flags * use __vmalloc() instead. |
a862f68a8 docs/core-api/mm:... |
2536 2537 |
* * Return: pointer to the allocated memory or %NULL on error |
930fc45a4 [PATCH] vmalloc_node |
2538 2539 2540 |
*/ void *vmalloc_node(unsigned long size, int node) { |
19809c2da mm, vmalloc: use ... |
2541 |
return __vmalloc_node(size, 1, GFP_KERNEL, PAGE_KERNEL, |
230169693 vmallocinfo: add ... |
2542 |
node, __builtin_return_address(0)); |
930fc45a4 [PATCH] vmalloc_node |
2543 2544 |
} EXPORT_SYMBOL(vmalloc_node); |
e1ca7788d mm: add vzalloc()... |
2545 2546 2547 2548 2549 2550 2551 2552 2553 2554 2555 |
/** * 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:... |
2556 2557 |
* * Return: pointer to the allocated memory or %NULL on error |
e1ca7788d mm: add vzalloc()... |
2558 2559 2560 2561 |
*/ void *vzalloc_node(unsigned long size, int node) { return __vmalloc_node_flags(size, node, |
19809c2da mm, vmalloc: use ... |
2562 |
GFP_KERNEL | __GFP_ZERO); |
e1ca7788d mm: add vzalloc()... |
2563 2564 |
} EXPORT_SYMBOL(vzalloc_node); |
1da177e4c Linux-2.6.12-rc2 |
2565 |
/** |
92eac1681 docs/mm: vmalloc:... |
2566 2567 |
* vmalloc_exec - allocate virtually contiguous, executable memory * @size: allocation size |
1da177e4c Linux-2.6.12-rc2 |
2568 |
* |
92eac1681 docs/mm: vmalloc:... |
2569 2570 2571 |
* 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 |
2572 |
* |
92eac1681 docs/mm: vmalloc:... |
2573 2574 |
* For tight control over page level allocator and protection flags * use __vmalloc() instead. |
a862f68a8 docs/core-api/mm:... |
2575 2576 |
* * Return: pointer to the allocated memory or %NULL on error |
1da177e4c Linux-2.6.12-rc2 |
2577 |
*/ |
1da177e4c Linux-2.6.12-rc2 |
2578 2579 |
void *vmalloc_exec(unsigned long size) { |
868b104d7 mm/vmalloc: Add f... |
2580 2581 2582 |
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 |
2583 |
} |
0d08e0d3a [PATCH] x86-64: F... |
2584 |
#if defined(CONFIG_64BIT) && defined(CONFIG_ZONE_DMA32) |
698d0831b vmalloc: fix __GF... |
2585 |
#define GFP_VMALLOC32 (GFP_DMA32 | GFP_KERNEL) |
0d08e0d3a [PATCH] x86-64: F... |
2586 |
#elif defined(CONFIG_64BIT) && defined(CONFIG_ZONE_DMA) |
698d0831b vmalloc: fix __GF... |
2587 |
#define GFP_VMALLOC32 (GFP_DMA | GFP_KERNEL) |
0d08e0d3a [PATCH] x86-64: F... |
2588 |
#else |
698d0831b vmalloc: fix __GF... |
2589 2590 2591 2592 2593 |
/* * 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... |
2594 |
#endif |
1da177e4c Linux-2.6.12-rc2 |
2595 |
/** |
92eac1681 docs/mm: vmalloc:... |
2596 2597 |
* vmalloc_32 - allocate virtually contiguous memory (32bit addressable) * @size: allocation size |
1da177e4c Linux-2.6.12-rc2 |
2598 |
* |
92eac1681 docs/mm: vmalloc:... |
2599 2600 |
* 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:... |
2601 2602 |
* * Return: pointer to the allocated memory or %NULL on error |
1da177e4c Linux-2.6.12-rc2 |
2603 2604 2605 |
*/ void *vmalloc_32(unsigned long size) { |
2dca6999e mm, perf_event: M... |
2606 |
return __vmalloc_node(size, 1, GFP_VMALLOC32, PAGE_KERNEL, |
00ef2d2f8 mm: use NUMA_NO_NODE |
2607 |
NUMA_NO_NODE, __builtin_return_address(0)); |
1da177e4c Linux-2.6.12-rc2 |
2608 |
} |
1da177e4c Linux-2.6.12-rc2 |
2609 |
EXPORT_SYMBOL(vmalloc_32); |
833423143 [PATCH] mm: intro... |
2610 |
/** |
ead04089b [PATCH] Fix kerne... |
2611 |
* vmalloc_32_user - allocate zeroed virtually contiguous 32bit memory |
92eac1681 docs/mm: vmalloc:... |
2612 |
* @size: allocation size |
ead04089b [PATCH] Fix kerne... |
2613 2614 2615 |
* * The resulting memory area is 32bit addressable and zeroed so it can be * mapped to userspace without leaking data. |
a862f68a8 docs/core-api/mm:... |
2616 2617 |
* * Return: pointer to the allocated memory or %NULL on error |
833423143 [PATCH] mm: intro... |
2618 2619 2620 |
*/ void *vmalloc_32_user(unsigned long size) { |
bc84c5352 mm/vmalloc: pass ... |
2621 2622 2623 2624 |
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... |
2625 2626 |
} EXPORT_SYMBOL(vmalloc_32_user); |
d0107eb07 kcore: fix vread/... |
2627 2628 2629 2630 2631 2632 2633 2634 2635 2636 2637 2638 |
/* * 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... |
2639 |
offset = offset_in_page(addr); |
d0107eb07 kcore: fix vread/... |
2640 2641 2642 2643 2644 2645 2646 2647 2648 2649 2650 2651 2652 2653 2654 2655 |
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... |
2656 |
void *map = kmap_atomic(p); |
d0107eb07 kcore: fix vread/... |
2657 |
memcpy(buf, map + offset, length); |
9b04c5fec mm: remove the se... |
2658 |
kunmap_atomic(map); |
d0107eb07 kcore: fix vread/... |
2659 2660 2661 2662 2663 2664 2665 2666 2667 2668 2669 2670 2671 2672 2673 2674 2675 2676 |
} 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... |
2677 |
offset = offset_in_page(addr); |
d0107eb07 kcore: fix vread/... |
2678 2679 2680 2681 2682 2683 2684 2685 2686 2687 2688 2689 2690 2691 2692 2693 |
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... |
2694 |
void *map = kmap_atomic(p); |
d0107eb07 kcore: fix vread/... |
2695 |
memcpy(map + offset, buf, length); |
9b04c5fec mm: remove the se... |
2696 |
kunmap_atomic(map); |
d0107eb07 kcore: fix vread/... |
2697 2698 2699 2700 2701 2702 2703 2704 2705 2706 |
} addr += length; buf += length; copied += length; count -= length; } return copied; } /** |
92eac1681 docs/mm: vmalloc:... |
2707 2708 2709 2710 2711 |
* 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:... |
2712 2713 2714 2715 2716 2717 2718 2719 2720 2721 2722 2723 |
* 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... |
2724 |
* any information, as /dev/kmem. |
a862f68a8 docs/core-api/mm:... |
2725 2726 2727 2728 |
* * 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/... |
2729 |
*/ |
1da177e4c Linux-2.6.12-rc2 |
2730 2731 |
long vread(char *buf, char *addr, unsigned long count) { |
e81ce85f9 mm, vmalloc: iter... |
2732 2733 |
struct vmap_area *va; struct vm_struct *vm; |
1da177e4c Linux-2.6.12-rc2 |
2734 |
char *vaddr, *buf_start = buf; |
d0107eb07 kcore: fix vread/... |
2735 |
unsigned long buflen = count; |
1da177e4c Linux-2.6.12-rc2 |
2736 2737 2738 2739 2740 |
unsigned long n; /* Don't allow overflow */ if ((unsigned long) addr + count < count) count = -(unsigned long) addr; |
e81ce85f9 mm, vmalloc: iter... |
2741 2742 2743 2744 |
spin_lock(&vmap_area_lock); list_for_each_entry(va, &vmap_area_list, list) { if (!count) break; |
688fcbfc0 mm/vmalloc: modif... |
2745 |
if (!va->vm) |
e81ce85f9 mm, vmalloc: iter... |
2746 2747 2748 2749 |
continue; vm = va->vm; vaddr = (char *) vm->addr; |
762216ab4 mm/vmalloc: use w... |
2750 |
if (addr >= vaddr + get_vm_area_size(vm)) |
1da177e4c Linux-2.6.12-rc2 |
2751 2752 2753 2754 2755 2756 2757 2758 2759 |
continue; while (addr < vaddr) { if (count == 0) goto finished; *buf = '\0'; buf++; addr++; count--; } |
762216ab4 mm/vmalloc: use w... |
2760 |
n = vaddr + get_vm_area_size(vm) - addr; |
d0107eb07 kcore: fix vread/... |
2761 2762 |
if (n > count) n = count; |
e81ce85f9 mm, vmalloc: iter... |
2763 |
if (!(vm->flags & VM_IOREMAP)) |
d0107eb07 kcore: fix vread/... |
2764 2765 2766 2767 2768 2769 |
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 |
2770 2771 |
} finished: |
e81ce85f9 mm, vmalloc: iter... |
2772 |
spin_unlock(&vmap_area_lock); |
d0107eb07 kcore: fix vread/... |
2773 2774 2775 2776 2777 2778 2779 2780 |
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 |
2781 |
} |
d0107eb07 kcore: fix vread/... |
2782 |
/** |
92eac1681 docs/mm: vmalloc:... |
2783 2784 2785 2786 2787 |
* 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:... |
2788 2789 2790 2791 2792 2793 2794 2795 2796 2797 2798 2799 |
* 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... |
2800 |
* any information, as /dev/kmem. |
a862f68a8 docs/core-api/mm:... |
2801 2802 2803 2804 |
* * 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/... |
2805 |
*/ |
1da177e4c Linux-2.6.12-rc2 |
2806 2807 |
long vwrite(char *buf, char *addr, unsigned long count) { |
e81ce85f9 mm, vmalloc: iter... |
2808 2809 |
struct vmap_area *va; struct vm_struct *vm; |
d0107eb07 kcore: fix vread/... |
2810 2811 2812 |
char *vaddr; unsigned long n, buflen; int copied = 0; |
1da177e4c Linux-2.6.12-rc2 |
2813 2814 2815 2816 |
/* Don't allow overflow */ if ((unsigned long) addr + count < count) count = -(unsigned long) addr; |
d0107eb07 kcore: fix vread/... |
2817 |
buflen = count; |
1da177e4c Linux-2.6.12-rc2 |
2818 |
|
e81ce85f9 mm, vmalloc: iter... |
2819 2820 2821 2822 |
spin_lock(&vmap_area_lock); list_for_each_entry(va, &vmap_area_list, list) { if (!count) break; |
688fcbfc0 mm/vmalloc: modif... |
2823 |
if (!va->vm) |
e81ce85f9 mm, vmalloc: iter... |
2824 2825 2826 2827 |
continue; vm = va->vm; vaddr = (char *) vm->addr; |
762216ab4 mm/vmalloc: use w... |
2828 |
if (addr >= vaddr + get_vm_area_size(vm)) |
1da177e4c Linux-2.6.12-rc2 |
2829 2830 2831 2832 2833 2834 2835 2836 |
continue; while (addr < vaddr) { if (count == 0) goto finished; buf++; addr++; count--; } |
762216ab4 mm/vmalloc: use w... |
2837 |
n = vaddr + get_vm_area_size(vm) - addr; |
d0107eb07 kcore: fix vread/... |
2838 2839 |
if (n > count) n = count; |
e81ce85f9 mm, vmalloc: iter... |
2840 |
if (!(vm->flags & VM_IOREMAP)) { |
d0107eb07 kcore: fix vread/... |
2841 2842 2843 2844 2845 2846 |
aligned_vwrite(buf, addr, n); copied++; } buf += n; addr += n; count -= n; |
1da177e4c Linux-2.6.12-rc2 |
2847 2848 |
} finished: |
e81ce85f9 mm, vmalloc: iter... |
2849 |
spin_unlock(&vmap_area_lock); |
d0107eb07 kcore: fix vread/... |
2850 2851 2852 |
if (!copied) return 0; return buflen; |
1da177e4c Linux-2.6.12-rc2 |
2853 |
} |
833423143 [PATCH] mm: intro... |
2854 2855 |
/** |
92eac1681 docs/mm: vmalloc:... |
2856 2857 2858 2859 |
* 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 |
f4f235309 vmalloc: fix rema... |
2860 |
* @pgoff: offset from @kaddr to start at |
92eac1681 docs/mm: vmalloc:... |
2861 |
* @size: size of map area |
7682486b3 mm: fix various k... |
2862 |
* |
92eac1681 docs/mm: vmalloc:... |
2863 |
* Returns: 0 for success, -Exxx on failure |
833423143 [PATCH] mm: intro... |
2864 |
* |
92eac1681 docs/mm: vmalloc:... |
2865 2866 2867 2868 |
* 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... |
2869 |
* |
92eac1681 docs/mm: vmalloc:... |
2870 |
* Similar to remap_pfn_range() (see mm/memory.c) |
833423143 [PATCH] mm: intro... |
2871 |
*/ |
e69e9d4ae vmalloc: introduc... |
2872 |
int remap_vmalloc_range_partial(struct vm_area_struct *vma, unsigned long uaddr, |
f4f235309 vmalloc: fix rema... |
2873 2874 |
void *kaddr, unsigned long pgoff, unsigned long size) |
833423143 [PATCH] mm: intro... |
2875 2876 |
{ struct vm_struct *area; |
f4f235309 vmalloc: fix rema... |
2877 2878 2879 2880 2881 |
unsigned long off; unsigned long end_index; if (check_shl_overflow(pgoff, PAGE_SHIFT, &off)) return -EINVAL; |
833423143 [PATCH] mm: intro... |
2882 |
|
e69e9d4ae vmalloc: introduc... |
2883 2884 2885 |
size = PAGE_ALIGN(size); if (!PAGE_ALIGNED(uaddr) || !PAGE_ALIGNED(kaddr)) |
833423143 [PATCH] mm: intro... |
2886 |
return -EINVAL; |
e69e9d4ae vmalloc: introduc... |
2887 |
area = find_vm_area(kaddr); |
833423143 [PATCH] mm: intro... |
2888 |
if (!area) |
db64fe022 mm: rewrite vmap ... |
2889 |
return -EINVAL; |
833423143 [PATCH] mm: intro... |
2890 |
|
fe9041c24 vmalloc: lift the... |
2891 |
if (!(area->flags & (VM_USERMAP | VM_DMA_COHERENT))) |
db64fe022 mm: rewrite vmap ... |
2892 |
return -EINVAL; |
833423143 [PATCH] mm: intro... |
2893 |
|
f4f235309 vmalloc: fix rema... |
2894 2895 |
if (check_add_overflow(size, off, &end_index) || end_index > get_vm_area_size(area)) |
db64fe022 mm: rewrite vmap ... |
2896 |
return -EINVAL; |
f4f235309 vmalloc: fix rema... |
2897 |
kaddr += off; |
833423143 [PATCH] mm: intro... |
2898 |
|
833423143 [PATCH] mm: intro... |
2899 |
do { |
e69e9d4ae vmalloc: introduc... |
2900 |
struct page *page = vmalloc_to_page(kaddr); |
db64fe022 mm: rewrite vmap ... |
2901 |
int ret; |
833423143 [PATCH] mm: intro... |
2902 2903 2904 2905 2906 |
ret = vm_insert_page(vma, uaddr, page); if (ret) return ret; uaddr += PAGE_SIZE; |
e69e9d4ae vmalloc: introduc... |
2907 2908 2909 |
kaddr += PAGE_SIZE; size -= PAGE_SIZE; } while (size > 0); |
833423143 [PATCH] mm: intro... |
2910 |
|
314e51b98 mm: kill vma flag... |
2911 |
vma->vm_flags |= VM_DONTEXPAND | VM_DONTDUMP; |
833423143 [PATCH] mm: intro... |
2912 |
|
db64fe022 mm: rewrite vmap ... |
2913 |
return 0; |
833423143 [PATCH] mm: intro... |
2914 |
} |
e69e9d4ae vmalloc: introduc... |
2915 2916 2917 |
EXPORT_SYMBOL(remap_vmalloc_range_partial); /** |
92eac1681 docs/mm: vmalloc:... |
2918 2919 2920 2921 |
* 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... |
2922 |
* |
92eac1681 docs/mm: vmalloc:... |
2923 |
* Returns: 0 for success, -Exxx on failure |
e69e9d4ae vmalloc: introduc... |
2924 |
* |
92eac1681 docs/mm: vmalloc:... |
2925 2926 2927 |
* 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... |
2928 |
* |
92eac1681 docs/mm: vmalloc:... |
2929 |
* Similar to remap_pfn_range() (see mm/memory.c) |
e69e9d4ae vmalloc: introduc... |
2930 2931 2932 2933 2934 |
*/ int remap_vmalloc_range(struct vm_area_struct *vma, void *addr, unsigned long pgoff) { return remap_vmalloc_range_partial(vma, vma->vm_start, |
f4f235309 vmalloc: fix rema... |
2935 |
addr, pgoff, |
e69e9d4ae vmalloc: introduc... |
2936 2937 |
vma->vm_end - vma->vm_start); } |
833423143 [PATCH] mm: intro... |
2938 |
EXPORT_SYMBOL(remap_vmalloc_range); |
1eeb66a1b move die notifier... |
2939 |
/* |
66f28e110 x86/mm: split vma... |
2940 2941 |
* Implement stubs for vmalloc_sync_[un]mappings () if the architecture chose * not to have one. |
3f8fd02b1 mm/vmalloc: Sync ... |
2942 2943 2944 |
* * 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... |
2945 |
*/ |
66f28e110 x86/mm: split vma... |
2946 |
void __weak vmalloc_sync_mappings(void) |
1eeb66a1b move die notifier... |
2947 2948 |
{ } |
5f4352fbf Allocate and free... |
2949 |
|
66f28e110 x86/mm: split vma... |
2950 2951 2952 |
void __weak vmalloc_sync_unmappings(void) { } |
5f4352fbf Allocate and free... |
2953 |
|
8b1e0f81f mm/pgtable: drop ... |
2954 |
static int f(pte_t *pte, unsigned long addr, void *data) |
5f4352fbf Allocate and free... |
2955 |
{ |
cd12909cb xen: map foreign ... |
2956 2957 2958 2959 2960 2961 |
pte_t ***p = data; if (p) { *(*p) = pte; (*p)++; } |
5f4352fbf Allocate and free... |
2962 2963 2964 2965 |
return 0; } /** |
92eac1681 docs/mm: vmalloc:... |
2966 2967 2968 |
* 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... |
2969 |
* |
92eac1681 docs/mm: vmalloc:... |
2970 |
* Returns: NULL on failure, vm_struct on success |
5f4352fbf Allocate and free... |
2971 |
* |
92eac1681 docs/mm: vmalloc:... |
2972 2973 2974 |
* 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 ... |
2975 |
* |
92eac1681 docs/mm: vmalloc:... |
2976 2977 |
* If @ptes is non-NULL, pointers to the PTEs (in init_mm) * allocated for the VM area are returned. |
5f4352fbf Allocate and free... |
2978 |
*/ |
cd12909cb xen: map foreign ... |
2979 |
struct vm_struct *alloc_vm_area(size_t size, pte_t **ptes) |
5f4352fbf Allocate and free... |
2980 2981 |
{ struct vm_struct *area; |
230169693 vmallocinfo: add ... |
2982 2983 |
area = get_vm_area_caller(size, VM_IOREMAP, __builtin_return_address(0)); |
5f4352fbf Allocate and free... |
2984 2985 2986 2987 2988 2989 2990 2991 |
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 ... |
2992 |
size, f, ptes ? &ptes : NULL)) { |
5f4352fbf Allocate and free... |
2993 2994 2995 |
free_vm_area(area); return NULL; } |
5f4352fbf Allocate and free... |
2996 2997 2998 2999 3000 3001 3002 3003 3004 3005 3006 3007 |
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... |
3008 |
|
4f8b02b4e vmalloc: pcpu_get... |
3009 |
#ifdef CONFIG_SMP |
ca23e405e vmalloc: implemen... |
3010 3011 |
static struct vmap_area *node_to_va(struct rb_node *n) { |
4583e7731 mm/vmalloc.c: use... |
3012 |
return rb_entry_safe(n, struct vmap_area, rb_node); |
ca23e405e vmalloc: implemen... |
3013 3014 3015 |
} /** |
68ad4a330 mm/vmalloc.c: kee... |
3016 3017 |
* pvm_find_va_enclose_addr - find the vmap_area @addr belongs to * @addr: target address |
ca23e405e vmalloc: implemen... |
3018 |
* |
68ad4a330 mm/vmalloc.c: kee... |
3019 3020 3021 3022 |
* 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... |
3023 |
*/ |
68ad4a330 mm/vmalloc.c: kee... |
3024 3025 |
static struct vmap_area * pvm_find_va_enclose_addr(unsigned long addr) |
ca23e405e vmalloc: implemen... |
3026 |
{ |
68ad4a330 mm/vmalloc.c: kee... |
3027 3028 3029 3030 3031 |
struct vmap_area *va, *tmp; struct rb_node *n; n = free_vmap_area_root.rb_node; va = NULL; |
ca23e405e vmalloc: implemen... |
3032 3033 |
while (n) { |
68ad4a330 mm/vmalloc.c: kee... |
3034 3035 3036 3037 3038 |
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... |
3039 |
n = n->rb_right; |
68ad4a330 mm/vmalloc.c: kee... |
3040 3041 3042 |
} else { n = n->rb_left; } |
ca23e405e vmalloc: implemen... |
3043 |
} |
68ad4a330 mm/vmalloc.c: kee... |
3044 |
return va; |
ca23e405e vmalloc: implemen... |
3045 3046 3047 |
} /** |
68ad4a330 mm/vmalloc.c: kee... |
3048 3049 3050 3051 3052 |
* 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... |
3053 |
* |
68ad4a330 mm/vmalloc.c: kee... |
3054 |
* Returns: determined end address within vmap_area |
ca23e405e vmalloc: implemen... |
3055 |
*/ |
68ad4a330 mm/vmalloc.c: kee... |
3056 3057 |
static unsigned long pvm_determine_end_from_reverse(struct vmap_area **va, unsigned long align) |
ca23e405e vmalloc: implemen... |
3058 |
{ |
68ad4a330 mm/vmalloc.c: kee... |
3059 |
unsigned long vmalloc_end = VMALLOC_END & ~(align - 1); |
ca23e405e vmalloc: implemen... |
3060 |
unsigned long addr; |
68ad4a330 mm/vmalloc.c: kee... |
3061 3062 3063 3064 3065 3066 3067 |
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... |
3068 |
} |
68ad4a330 mm/vmalloc.c: kee... |
3069 |
return 0; |
ca23e405e vmalloc: implemen... |
3070 3071 3072 3073 3074 3075 3076 3077 |
} /** * 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... |
3078 3079 3080 3081 3082 3083 |
* * 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... |
3084 3085 3086 3087 |
* 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... |
3088 |
* |
68ad4a330 mm/vmalloc.c: kee... |
3089 3090 3091 3092 3093 3094 |
* 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... |
3095 3096 3097 |
*/ struct vm_struct **pcpu_get_vm_areas(const unsigned long *offsets, const size_t *sizes, int nr_vms, |
ec3f64fc9 mm: remove gfp ma... |
3098 |
size_t align) |
ca23e405e vmalloc: implemen... |
3099 3100 3101 |
{ const unsigned long vmalloc_start = ALIGN(VMALLOC_START, align); const unsigned long vmalloc_end = VMALLOC_END & ~(align - 1); |
68ad4a330 mm/vmalloc.c: kee... |
3102 |
struct vmap_area **vas, *va; |
ca23e405e vmalloc: implemen... |
3103 3104 |
struct vm_struct **vms; int area, area2, last_area, term_area; |
68ad4a330 mm/vmalloc.c: kee... |
3105 |
unsigned long base, start, size, end, last_end; |
ca23e405e vmalloc: implemen... |
3106 |
bool purged = false; |
68ad4a330 mm/vmalloc.c: kee... |
3107 |
enum fit_type type; |
ca23e405e vmalloc: implemen... |
3108 |
|
ca23e405e vmalloc: implemen... |
3109 |
/* verify parameters and allocate data structures */ |
891c49abf mm/vmalloc: use o... |
3110 |
BUG_ON(offset_in_page(align) || !is_power_of_2(align)); |
ca23e405e vmalloc: implemen... |
3111 3112 3113 3114 3115 3116 3117 3118 3119 3120 3121 |
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... |
3122 |
for (area2 = area + 1; area2 < nr_vms; area2++) { |
ca23e405e vmalloc: implemen... |
3123 3124 |
unsigned long start2 = offsets[area2]; unsigned long end2 = start2 + sizes[area2]; |
c568da282 mm/vmalloc.c: hal... |
3125 |
BUG_ON(start2 < end && start < end2); |
ca23e405e vmalloc: implemen... |
3126 3127 3128 3129 3130 3131 3132 3133 |
} } last_end = offsets[last_area] + sizes[last_area]; if (vmalloc_end - vmalloc_start < last_end) { WARN_ON(true); return NULL; } |
4d67d8605 mm: use kcalloc()... |
3134 3135 |
vms = kcalloc(nr_vms, sizeof(vms[0]), GFP_KERNEL); vas = kcalloc(nr_vms, sizeof(vas[0]), GFP_KERNEL); |
ca23e405e vmalloc: implemen... |
3136 |
if (!vas || !vms) |
f1db7afd9 mm/vmalloc.c: eli... |
3137 |
goto err_free2; |
ca23e405e vmalloc: implemen... |
3138 3139 |
for (area = 0; area < nr_vms; area++) { |
68ad4a330 mm/vmalloc.c: kee... |
3140 |
vas[area] = kmem_cache_zalloc(vmap_area_cachep, GFP_KERNEL); |
ec3f64fc9 mm: remove gfp ma... |
3141 |
vms[area] = kzalloc(sizeof(struct vm_struct), GFP_KERNEL); |
ca23e405e vmalloc: implemen... |
3142 3143 3144 3145 3146 3147 3148 3149 3150 3151 |
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... |
3152 3153 |
va = pvm_find_va_enclose_addr(vmalloc_end); base = pvm_determine_end_from_reverse(&va, align) - end; |
ca23e405e vmalloc: implemen... |
3154 3155 |
while (true) { |
ca23e405e vmalloc: implemen... |
3156 3157 3158 3159 |
/* * base might have underflowed, add last_end before * comparing. */ |
68ad4a330 mm/vmalloc.c: kee... |
3160 3161 |
if (base + last_end < vmalloc_start + last_end) goto overflow; |
ca23e405e vmalloc: implemen... |
3162 3163 |
/* |
68ad4a330 mm/vmalloc.c: kee... |
3164 |
* Fitting base has not been found. |
ca23e405e vmalloc: implemen... |
3165 |
*/ |
68ad4a330 mm/vmalloc.c: kee... |
3166 3167 |
if (va == NULL) goto overflow; |
ca23e405e vmalloc: implemen... |
3168 3169 |
/* |
5336e52c9 mm/vmalloc.c: fix... |
3170 3171 3172 3173 3174 3175 3176 3177 3178 3179 |
* 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... |
3180 |
* If this VA does not fit, move base downwards and recheck. |
ca23e405e vmalloc: implemen... |
3181 |
*/ |
5336e52c9 mm/vmalloc.c: fix... |
3182 |
if (base + start < va->va_start) { |
68ad4a330 mm/vmalloc.c: kee... |
3183 3184 |
va = node_to_va(rb_prev(&va->rb_node)); base = pvm_determine_end_from_reverse(&va, align) - end; |
ca23e405e vmalloc: implemen... |
3185 3186 3187 3188 3189 3190 3191 3192 3193 3194 3195 |
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... |
3196 |
|
ca23e405e vmalloc: implemen... |
3197 3198 |
start = offsets[area]; end = start + sizes[area]; |
68ad4a330 mm/vmalloc.c: kee... |
3199 |
va = pvm_find_va_enclose_addr(base + end); |
ca23e405e vmalloc: implemen... |
3200 |
} |
68ad4a330 mm/vmalloc.c: kee... |
3201 |
|
ca23e405e vmalloc: implemen... |
3202 3203 |
/* we've found a fitting base, insert all va's */ for (area = 0; area < nr_vms; area++) { |
68ad4a330 mm/vmalloc.c: kee... |
3204 |
int ret; |
ca23e405e vmalloc: implemen... |
3205 |
|
68ad4a330 mm/vmalloc.c: kee... |
3206 3207 |
start = base + offsets[area]; size = sizes[area]; |
ca23e405e vmalloc: implemen... |
3208 |
|
68ad4a330 mm/vmalloc.c: kee... |
3209 3210 3211 3212 3213 3214 3215 3216 3217 3218 3219 3220 3221 3222 3223 3224 3225 3226 3227 3228 3229 |
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... |
3230 3231 3232 3233 3234 |
spin_unlock(&vmap_area_lock); /* insert all vm's */ for (area = 0; area < nr_vms; area++) |
3645cb4a4 mm, vmalloc: call... |
3235 3236 |
setup_vmalloc_vm(vms[area], vas[area], VM_ALLOC, pcpu_get_vm_areas); |
ca23e405e vmalloc: implemen... |
3237 3238 3239 |
kfree(vas); return vms; |
68ad4a330 mm/vmalloc.c: kee... |
3240 3241 3242 3243 3244 3245 3246 3247 3248 3249 3250 3251 3252 3253 3254 3255 3256 3257 3258 3259 3260 3261 3262 3263 3264 3265 |
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... |
3266 3267 |
err_free: for (area = 0; area < nr_vms; area++) { |
68ad4a330 mm/vmalloc.c: kee... |
3268 3269 |
if (vas[area]) kmem_cache_free(vmap_area_cachep, vas[area]); |
f1db7afd9 mm/vmalloc.c: eli... |
3270 |
kfree(vms[area]); |
ca23e405e vmalloc: implemen... |
3271 |
} |
f1db7afd9 mm/vmalloc.c: eli... |
3272 |
err_free2: |
ca23e405e vmalloc: implemen... |
3273 3274 3275 3276 3277 3278 3279 3280 3281 3282 3283 3284 3285 3286 3287 3288 3289 3290 3291 3292 |
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... |
3293 |
#endif /* CONFIG_SMP */ |
a10aa5798 vmalloc: show vma... |
3294 3295 3296 |
#ifdef CONFIG_PROC_FS static void *s_start(struct seq_file *m, loff_t *pos) |
d4033afdf mm, vmalloc: iter... |
3297 |
__acquires(&vmap_area_lock) |
a10aa5798 vmalloc: show vma... |
3298 |
{ |
d4033afdf mm, vmalloc: iter... |
3299 |
spin_lock(&vmap_area_lock); |
3f5000693 mm/vmalloc.c: sim... |
3300 |
return seq_list_start(&vmap_area_list, *pos); |
a10aa5798 vmalloc: show vma... |
3301 3302 3303 3304 |
} static void *s_next(struct seq_file *m, void *p, loff_t *pos) { |
3f5000693 mm/vmalloc.c: sim... |
3305 |
return seq_list_next(p, &vmap_area_list, pos); |
a10aa5798 vmalloc: show vma... |
3306 3307 3308 |
} static void s_stop(struct seq_file *m, void *p) |
d4033afdf mm, vmalloc: iter... |
3309 |
__releases(&vmap_area_lock) |
a10aa5798 vmalloc: show vma... |
3310 |
{ |
d4033afdf mm, vmalloc: iter... |
3311 |
spin_unlock(&vmap_area_lock); |
a10aa5798 vmalloc: show vma... |
3312 |
} |
a47a126ad vmallocinfo: add ... |
3313 3314 |
static void show_numa_info(struct seq_file *m, struct vm_struct *v) { |
e5adfffc8 mm: use IS_ENABLE... |
3315 |
if (IS_ENABLED(CONFIG_NUMA)) { |
a47a126ad vmallocinfo: add ... |
3316 3317 3318 3319 |
unsigned int nr, *counters = m->private; if (!counters) return; |
af12346cd mm/vmalloc: rever... |
3320 3321 |
if (v->flags & VM_UNINITIALIZED) return; |
7e5b528b4 mm/vmalloc.c: fix... |
3322 3323 |
/* Pair with smp_wmb() in clear_vm_uninitialized_flag() */ smp_rmb(); |
af12346cd mm/vmalloc: rever... |
3324 |
|
a47a126ad vmallocinfo: add ... |
3325 3326 3327 3328 3329 3330 3331 3332 3333 3334 |
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... |
3335 3336 3337 3338 3339 3340 3341 3342 3343 3344 3345 3346 3347 3348 3349 3350 |
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... |
3351 3352 |
static int s_show(struct seq_file *m, void *p) { |
3f5000693 mm/vmalloc.c: sim... |
3353 |
struct vmap_area *va; |
d4033afdf mm, vmalloc: iter... |
3354 |
struct vm_struct *v; |
3f5000693 mm/vmalloc.c: sim... |
3355 |
va = list_entry(p, struct vmap_area, list); |
c2ce8c142 mm/vmalloc: fix s... |
3356 |
/* |
688fcbfc0 mm/vmalloc: modif... |
3357 3358 |
* 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... |
3359 |
*/ |
688fcbfc0 mm/vmalloc: modif... |
3360 |
if (!va->vm) { |
dd3b8353b mm/vmalloc: do no... |
3361 3362 |
seq_printf(m, "0x%pK-0x%pK %7ld vm_map_ram ", |
78c72746f vmalloc: show laz... |
3363 |
(void *)va->va_start, (void *)va->va_end, |
dd3b8353b mm/vmalloc: do no... |
3364 |
va->va_end - va->va_start); |
78c72746f vmalloc: show laz... |
3365 |
|
d4033afdf mm, vmalloc: iter... |
3366 |
return 0; |
78c72746f vmalloc: show laz... |
3367 |
} |
d4033afdf mm, vmalloc: iter... |
3368 3369 |
v = va->vm; |
a10aa5798 vmalloc: show vma... |
3370 |
|
45ec16908 mm: use %pK for /... |
3371 |
seq_printf(m, "0x%pK-0x%pK %7ld", |
a10aa5798 vmalloc: show vma... |
3372 |
v->addr, v->addr + v->size, v->size); |
62c70bce8 mm: convert sprin... |
3373 3374 |
if (v->caller) seq_printf(m, " %pS", v->caller); |
230169693 vmallocinfo: add ... |
3375 |
|
a10aa5798 vmalloc: show vma... |
3376 3377 3378 3379 |
if (v->nr_pages) seq_printf(m, " pages=%d", v->nr_pages); if (v->phys_addr) |
199eaa05a mm: cleanups for ... |
3380 |
seq_printf(m, " phys=%pa", &v->phys_addr); |
a10aa5798 vmalloc: show vma... |
3381 3382 |
if (v->flags & VM_IOREMAP) |
f4527c908 mm/vmalloc.c: rep... |
3383 |
seq_puts(m, " ioremap"); |
a10aa5798 vmalloc: show vma... |
3384 3385 |
if (v->flags & VM_ALLOC) |
f4527c908 mm/vmalloc.c: rep... |
3386 |
seq_puts(m, " vmalloc"); |
a10aa5798 vmalloc: show vma... |
3387 3388 |
if (v->flags & VM_MAP) |
f4527c908 mm/vmalloc.c: rep... |
3389 |
seq_puts(m, " vmap"); |
a10aa5798 vmalloc: show vma... |
3390 3391 |
if (v->flags & VM_USERMAP) |
f4527c908 mm/vmalloc.c: rep... |
3392 |
seq_puts(m, " user"); |
a10aa5798 vmalloc: show vma... |
3393 |
|
fe9041c24 vmalloc: lift the... |
3394 3395 |
if (v->flags & VM_DMA_COHERENT) seq_puts(m, " dma-coherent"); |
244d63ee3 mm, vmalloc: remo... |
3396 |
if (is_vmalloc_addr(v->pages)) |
f4527c908 mm/vmalloc.c: rep... |
3397 |
seq_puts(m, " vpages"); |
a10aa5798 vmalloc: show vma... |
3398 |
|
a47a126ad vmallocinfo: add ... |
3399 |
show_numa_info(m, v); |
a10aa5798 vmalloc: show vma... |
3400 3401 |
seq_putc(m, ' '); |
dd3b8353b mm/vmalloc: do no... |
3402 3403 3404 3405 3406 3407 3408 3409 3410 |
/* * 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... |
3411 3412 |
return 0; } |
5f6a6a9c4 proc: move /proc/... |
3413 |
static const struct seq_operations vmalloc_op = { |
a10aa5798 vmalloc: show vma... |
3414 3415 3416 3417 3418 |
.start = s_start, .next = s_next, .stop = s_stop, .show = s_show, }; |
5f6a6a9c4 proc: move /proc/... |
3419 |
|
5f6a6a9c4 proc: move /proc/... |
3420 3421 |
static int __init proc_vmalloc_init(void) { |
fddda2b7b proc: introduce p... |
3422 |
if (IS_ENABLED(CONFIG_NUMA)) |
0825a6f98 mm: use octal not... |
3423 |
proc_create_seq_private("vmallocinfo", 0400, NULL, |
44414d82c proc: introduce p... |
3424 3425 |
&vmalloc_op, nr_node_ids * sizeof(unsigned int), NULL); |
fddda2b7b proc: introduce p... |
3426 |
else |
0825a6f98 mm: use octal not... |
3427 |
proc_create_seq("vmallocinfo", 0400, NULL, &vmalloc_op); |
5f6a6a9c4 proc: move /proc/... |
3428 3429 3430 |
return 0; } module_init(proc_vmalloc_init); |
db3808c1b mm, vmalloc: move... |
3431 |
|
a10aa5798 vmalloc: show vma... |
3432 |
#endif |