Blame view
mm/vmalloc.c
69 KB
1da177e4c Linux-2.6.12-rc2 |
1 2 3 4 5 6 7 |
/* * 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 |
8 |
* Numa awareness, Christoph Lameter, SGI, June 2005 |
1da177e4c Linux-2.6.12-rc2 |
9 |
*/ |
db64fe022 mm: rewrite vmap ... |
10 |
#include <linux/vmalloc.h> |
1da177e4c Linux-2.6.12-rc2 |
11 12 13 |
#include <linux/mm.h> #include <linux/module.h> #include <linux/highmem.h> |
d43c36dc6 headers: remove s... |
14 |
#include <linux/sched.h> |
1da177e4c Linux-2.6.12-rc2 |
15 16 17 |
#include <linux/slab.h> #include <linux/spinlock.h> #include <linux/interrupt.h> |
5f6a6a9c4 proc: move /proc/... |
18 |
#include <linux/proc_fs.h> |
a10aa5798 vmalloc: show vma... |
19 |
#include <linux/seq_file.h> |
3ac7fe5a4 infrastructure to... |
20 |
#include <linux/debugobjects.h> |
230169693 vmallocinfo: add ... |
21 |
#include <linux/kallsyms.h> |
db64fe022 mm: rewrite vmap ... |
22 23 24 25 |
#include <linux/list.h> #include <linux/rbtree.h> #include <linux/radix-tree.h> #include <linux/rcupdate.h> |
f0aa66179 vmalloc: implemen... |
26 |
#include <linux/pfn.h> |
89219d37a kmemleak: Add the... |
27 |
#include <linux/kmemleak.h> |
60063497a atomic: use <linu... |
28 |
#include <linux/atomic.h> |
3b32123d7 mm: use macros fr... |
29 |
#include <linux/compiler.h> |
32fcfd407 make vfree() safe... |
30 |
#include <linux/llist.h> |
0f616be12 mm: change __get_... |
31 |
#include <linux/bitops.h> |
3b32123d7 mm: use macros fr... |
32 |
|
1da177e4c Linux-2.6.12-rc2 |
33 34 |
#include <asm/uaccess.h> #include <asm/tlbflush.h> |
2dca6999e mm, perf_event: M... |
35 |
#include <asm/shmparam.h> |
1da177e4c Linux-2.6.12-rc2 |
36 |
|
32fcfd407 make vfree() safe... |
37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 |
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); struct llist_node *llnode = llist_del_all(&p->list); while (llnode) { void *p = llnode; llnode = llist_next(llnode); __vunmap(p, 1); } } |
db64fe022 mm: rewrite vmap ... |
55 |
/*** Page table manipulation functions ***/ |
b221385bc [PATCH] mm/: make... |
56 |
|
1da177e4c Linux-2.6.12-rc2 |
57 58 59 60 61 62 63 64 65 66 |
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 ... |
67 |
static void vunmap_pmd_range(pud_t *pud, unsigned long addr, unsigned long end) |
1da177e4c Linux-2.6.12-rc2 |
68 69 70 71 72 73 74 |
{ pmd_t *pmd; unsigned long next; pmd = pmd_offset(pud, addr); do { next = pmd_addr_end(addr, end); |
b9820d8f3 mm: change vunmap... |
75 76 |
if (pmd_clear_huge(pmd)) continue; |
1da177e4c Linux-2.6.12-rc2 |
77 78 79 80 81 |
if (pmd_none_or_clear_bad(pmd)) continue; vunmap_pte_range(pmd, addr, next); } while (pmd++, addr = next, addr != end); } |
db64fe022 mm: rewrite vmap ... |
82 |
static void vunmap_pud_range(pgd_t *pgd, unsigned long addr, unsigned long end) |
1da177e4c Linux-2.6.12-rc2 |
83 84 85 86 87 88 89 |
{ pud_t *pud; unsigned long next; pud = pud_offset(pgd, addr); do { next = pud_addr_end(addr, end); |
b9820d8f3 mm: change vunmap... |
90 91 |
if (pud_clear_huge(pud)) continue; |
1da177e4c Linux-2.6.12-rc2 |
92 93 94 95 96 |
if (pud_none_or_clear_bad(pud)) continue; vunmap_pmd_range(pud, addr, next); } while (pud++, addr = next, addr != end); } |
db64fe022 mm: rewrite vmap ... |
97 |
static void vunmap_page_range(unsigned long addr, unsigned long end) |
1da177e4c Linux-2.6.12-rc2 |
98 99 100 |
{ pgd_t *pgd; unsigned long next; |
1da177e4c Linux-2.6.12-rc2 |
101 102 103 |
BUG_ON(addr >= end); pgd = pgd_offset_k(addr); |
1da177e4c Linux-2.6.12-rc2 |
104 105 106 107 108 109 |
do { next = pgd_addr_end(addr, end); if (pgd_none_or_clear_bad(pgd)) continue; vunmap_pud_range(pgd, addr, next); } while (pgd++, addr = next, addr != end); |
1da177e4c Linux-2.6.12-rc2 |
110 111 112 |
} static int vmap_pte_range(pmd_t *pmd, unsigned long addr, |
db64fe022 mm: rewrite vmap ... |
113 |
unsigned long end, pgprot_t prot, struct page **pages, int *nr) |
1da177e4c Linux-2.6.12-rc2 |
114 115 |
{ pte_t *pte; |
db64fe022 mm: rewrite vmap ... |
116 117 118 119 |
/* * 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_... |
120 |
pte = pte_alloc_kernel(pmd, addr); |
1da177e4c Linux-2.6.12-rc2 |
121 122 123 |
if (!pte) return -ENOMEM; do { |
db64fe022 mm: rewrite vmap ... |
124 125 126 127 128 |
struct page *page = pages[*nr]; if (WARN_ON(!pte_none(*pte))) return -EBUSY; if (WARN_ON(!page)) |
1da177e4c Linux-2.6.12-rc2 |
129 130 |
return -ENOMEM; set_pte_at(&init_mm, addr, pte, mk_pte(page, prot)); |
db64fe022 mm: rewrite vmap ... |
131 |
(*nr)++; |
1da177e4c Linux-2.6.12-rc2 |
132 133 134 |
} while (pte++, addr += PAGE_SIZE, addr != end); return 0; } |
db64fe022 mm: rewrite vmap ... |
135 136 |
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 |
137 138 139 140 141 142 143 144 145 |
{ 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 ... |
146 |
if (vmap_pte_range(pmd, addr, next, prot, pages, nr)) |
1da177e4c Linux-2.6.12-rc2 |
147 148 149 150 |
return -ENOMEM; } while (pmd++, addr = next, addr != end); return 0; } |
db64fe022 mm: rewrite vmap ... |
151 152 |
static int vmap_pud_range(pgd_t *pgd, unsigned long addr, unsigned long end, pgprot_t prot, struct page **pages, int *nr) |
1da177e4c Linux-2.6.12-rc2 |
153 154 155 156 157 158 159 160 161 |
{ pud_t *pud; unsigned long next; pud = pud_alloc(&init_mm, pgd, addr); if (!pud) return -ENOMEM; do { next = pud_addr_end(addr, end); |
db64fe022 mm: rewrite vmap ... |
162 |
if (vmap_pmd_range(pud, addr, next, prot, pages, nr)) |
1da177e4c Linux-2.6.12-rc2 |
163 164 165 166 |
return -ENOMEM; } while (pud++, addr = next, addr != end); return 0; } |
db64fe022 mm: rewrite vmap ... |
167 168 169 170 171 172 |
/* * 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... |
173 174 |
static int vmap_page_range_noflush(unsigned long start, unsigned long end, pgprot_t prot, struct page **pages) |
1da177e4c Linux-2.6.12-rc2 |
175 176 177 |
{ pgd_t *pgd; unsigned long next; |
2e4e27c7d vmalloc.c: fix fl... |
178 |
unsigned long addr = start; |
db64fe022 mm: rewrite vmap ... |
179 180 |
int err = 0; int nr = 0; |
1da177e4c Linux-2.6.12-rc2 |
181 182 183 |
BUG_ON(addr >= end); pgd = pgd_offset_k(addr); |
1da177e4c Linux-2.6.12-rc2 |
184 185 |
do { next = pgd_addr_end(addr, end); |
db64fe022 mm: rewrite vmap ... |
186 |
err = vmap_pud_range(pgd, addr, next, prot, pages, &nr); |
1da177e4c Linux-2.6.12-rc2 |
187 |
if (err) |
bf88c8c83 vmalloc.c: fix do... |
188 |
return err; |
1da177e4c Linux-2.6.12-rc2 |
189 |
} while (pgd++, addr = next, addr != end); |
db64fe022 mm: rewrite vmap ... |
190 |
|
db64fe022 mm: rewrite vmap ... |
191 |
return nr; |
1da177e4c Linux-2.6.12-rc2 |
192 |
} |
8fc489850 vmalloc: add un/m... |
193 194 195 196 197 198 199 200 201 |
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... |
202 |
int is_vmalloc_or_module_addr(const void *x) |
73bdf0a60 Introduce is_vmal... |
203 204 |
{ /* |
ab4f2ee13 [ARM] fix naming ... |
205 |
* ARM, x86-64 and sparc64 put modules in a special place, |
73bdf0a60 Introduce is_vmal... |
206 207 208 209 210 211 212 213 214 215 |
* 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... |
216 |
/* |
add688fbd Revert "mm/vmallo... |
217 |
* Walk a vmap address to the struct page it maps. |
48667e7a4 Move vmalloc_to_p... |
218 |
*/ |
add688fbd Revert "mm/vmallo... |
219 |
struct page *vmalloc_to_page(const void *vmalloc_addr) |
48667e7a4 Move vmalloc_to_p... |
220 221 |
{ unsigned long addr = (unsigned long) vmalloc_addr; |
add688fbd Revert "mm/vmallo... |
222 |
struct page *page = NULL; |
48667e7a4 Move vmalloc_to_p... |
223 |
pgd_t *pgd = pgd_offset_k(addr); |
48667e7a4 Move vmalloc_to_p... |
224 |
|
7aa413def x86, MM: virtual ... |
225 226 227 228 |
/* * 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... |
229 |
VIRTUAL_BUG_ON(!is_vmalloc_or_module_addr(vmalloc_addr)); |
59ea74633 MM: virtual addre... |
230 |
|
48667e7a4 Move vmalloc_to_p... |
231 |
if (!pgd_none(*pgd)) { |
db64fe022 mm: rewrite vmap ... |
232 |
pud_t *pud = pud_offset(pgd, addr); |
48667e7a4 Move vmalloc_to_p... |
233 |
if (!pud_none(*pud)) { |
db64fe022 mm: rewrite vmap ... |
234 |
pmd_t *pmd = pmd_offset(pud, addr); |
48667e7a4 Move vmalloc_to_p... |
235 |
if (!pmd_none(*pmd)) { |
db64fe022 mm: rewrite vmap ... |
236 |
pte_t *ptep, pte; |
48667e7a4 Move vmalloc_to_p... |
237 238 239 |
ptep = pte_offset_map(pmd, addr); pte = *ptep; if (pte_present(pte)) |
add688fbd Revert "mm/vmallo... |
240 |
page = pte_page(pte); |
48667e7a4 Move vmalloc_to_p... |
241 242 243 244 |
pte_unmap(ptep); } } } |
add688fbd Revert "mm/vmallo... |
245 |
return page; |
48667e7a4 Move vmalloc_to_p... |
246 |
} |
add688fbd Revert "mm/vmallo... |
247 |
EXPORT_SYMBOL(vmalloc_to_page); |
48667e7a4 Move vmalloc_to_p... |
248 249 |
/* |
add688fbd Revert "mm/vmallo... |
250 |
* Map a vmalloc()-space virtual address to the physical page frame number. |
48667e7a4 Move vmalloc_to_p... |
251 |
*/ |
add688fbd Revert "mm/vmallo... |
252 |
unsigned long vmalloc_to_pfn(const void *vmalloc_addr) |
48667e7a4 Move vmalloc_to_p... |
253 |
{ |
add688fbd Revert "mm/vmallo... |
254 |
return page_to_pfn(vmalloc_to_page(vmalloc_addr)); |
48667e7a4 Move vmalloc_to_p... |
255 |
} |
add688fbd Revert "mm/vmallo... |
256 |
EXPORT_SYMBOL(vmalloc_to_pfn); |
48667e7a4 Move vmalloc_to_p... |
257 |
|
db64fe022 mm: rewrite vmap ... |
258 259 260 261 262 263 |
/*** Global kva allocator ***/ #define VM_LAZY_FREE 0x01 #define VM_LAZY_FREEING 0x02 #define VM_VM_AREA 0x04 |
db64fe022 mm: rewrite vmap ... |
264 |
static DEFINE_SPINLOCK(vmap_area_lock); |
f1c4069e1 mm, vmalloc: expo... |
265 266 |
/* Export for kexec only */ LIST_HEAD(vmap_area_list); |
89699605f mm: vmap area cache |
267 268 269 270 271 272 273 |
static struct rb_root vmap_area_root = RB_ROOT; /* The vmap cache globals are protected by vmap_area_lock */ static struct rb_node *free_vmap_cache; static unsigned long cached_hole_size; static unsigned long cached_vstart; static unsigned long cached_align; |
ca23e405e vmalloc: implemen... |
274 |
static unsigned long vmap_area_pcpu_hole; |
db64fe022 mm: rewrite vmap ... |
275 276 |
static struct vmap_area *__find_vmap_area(unsigned long addr) |
1da177e4c Linux-2.6.12-rc2 |
277 |
{ |
db64fe022 mm: rewrite vmap ... |
278 279 280 281 282 283 284 285 |
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... |
286 |
else if (addr >= va->va_end) |
db64fe022 mm: rewrite vmap ... |
287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 |
n = n->rb_right; else return va; } return NULL; } static void __insert_vmap_area(struct vmap_area *va) { struct rb_node **p = &vmap_area_root.rb_node; struct rb_node *parent = NULL; struct rb_node *tmp; while (*p) { |
170168d0a vmalloc: rename t... |
302 |
struct vmap_area *tmp_va; |
db64fe022 mm: rewrite vmap ... |
303 304 |
parent = *p; |
170168d0a vmalloc: rename t... |
305 306 |
tmp_va = rb_entry(parent, struct vmap_area, rb_node); if (va->va_start < tmp_va->va_end) |
db64fe022 mm: rewrite vmap ... |
307 |
p = &(*p)->rb_left; |
170168d0a vmalloc: rename t... |
308 |
else if (va->va_end > tmp_va->va_start) |
db64fe022 mm: rewrite vmap ... |
309 310 311 312 313 314 315 |
p = &(*p)->rb_right; else BUG(); } rb_link_node(&va->rb_node, parent, p); rb_insert_color(&va->rb_node, &vmap_area_root); |
4341fa454 mm, vmalloc: remo... |
316 |
/* address-sort this list */ |
db64fe022 mm: rewrite vmap ... |
317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 |
tmp = rb_prev(&va->rb_node); if (tmp) { struct vmap_area *prev; prev = rb_entry(tmp, struct vmap_area, rb_node); list_add_rcu(&va->list, &prev->list); } else list_add_rcu(&va->list, &vmap_area_list); } static void purge_vmap_area_lazy(void); /* * 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) { struct vmap_area *va; struct rb_node *n; |
1da177e4c Linux-2.6.12-rc2 |
339 |
unsigned long addr; |
db64fe022 mm: rewrite vmap ... |
340 |
int purged = 0; |
89699605f mm: vmap area cache |
341 |
struct vmap_area *first; |
db64fe022 mm: rewrite vmap ... |
342 |
|
7766970cc mm: vmap fix over... |
343 |
BUG_ON(!size); |
db64fe022 mm: rewrite vmap ... |
344 |
BUG_ON(size & ~PAGE_MASK); |
89699605f mm: vmap area cache |
345 |
BUG_ON(!is_power_of_2(align)); |
db64fe022 mm: rewrite vmap ... |
346 |
|
db64fe022 mm: rewrite vmap ... |
347 348 349 350 |
va = kmalloc_node(sizeof(struct vmap_area), gfp_mask & GFP_RECLAIM_MASK, node); if (unlikely(!va)) return ERR_PTR(-ENOMEM); |
7f88f88f8 mm: kmemleak: avo... |
351 352 353 354 355 |
/* * 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 ... |
356 357 |
retry: spin_lock(&vmap_area_lock); |
89699605f mm: vmap area cache |
358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 |
/* * Invalidate cache if we have more permissive parameters. * cached_hole_size notes the largest hole noticed _below_ * the vmap_area cached in free_vmap_cache: if size fits * into that hole, we want to scan from vstart to reuse * the hole instead of allocating above free_vmap_cache. * Note that __free_vmap_area may update free_vmap_cache * without updating cached_hole_size or cached_align. */ if (!free_vmap_cache || size < cached_hole_size || vstart < cached_vstart || align < cached_align) { nocache: cached_hole_size = 0; free_vmap_cache = NULL; } /* record if we encounter less permissive parameters */ cached_vstart = vstart; cached_align = align; /* find starting point for our search */ if (free_vmap_cache) { first = rb_entry(free_vmap_cache, struct vmap_area, rb_node); |
248ac0e19 mm/vmalloc: remov... |
382 |
addr = ALIGN(first->va_end, align); |
89699605f mm: vmap area cache |
383 384 |
if (addr < vstart) goto nocache; |
bcb615a81 mm/vmalloc.c: fix... |
385 |
if (addr + size < addr) |
89699605f mm: vmap area cache |
386 387 388 389 |
goto overflow; } else { addr = ALIGN(vstart, align); |
bcb615a81 mm/vmalloc.c: fix... |
390 |
if (addr + size < addr) |
89699605f mm: vmap area cache |
391 392 393 394 395 396 |
goto overflow; n = vmap_area_root.rb_node; first = NULL; while (n) { |
db64fe022 mm: rewrite vmap ... |
397 398 399 |
struct vmap_area *tmp; tmp = rb_entry(n, struct vmap_area, rb_node); if (tmp->va_end >= addr) { |
db64fe022 mm: rewrite vmap ... |
400 |
first = tmp; |
89699605f mm: vmap area cache |
401 402 403 404 |
if (tmp->va_start <= addr) break; n = n->rb_left; } else |
db64fe022 mm: rewrite vmap ... |
405 |
n = n->rb_right; |
89699605f mm: vmap area cache |
406 |
} |
db64fe022 mm: rewrite vmap ... |
407 408 409 |
if (!first) goto found; |
db64fe022 mm: rewrite vmap ... |
410 |
} |
89699605f mm: vmap area cache |
411 412 |
/* from the starting point, walk areas until a suitable hole is found */ |
248ac0e19 mm/vmalloc: remov... |
413 |
while (addr + size > first->va_start && addr + size <= vend) { |
89699605f mm: vmap area cache |
414 415 |
if (addr + cached_hole_size < first->va_start) cached_hole_size = first->va_start - addr; |
248ac0e19 mm/vmalloc: remov... |
416 |
addr = ALIGN(first->va_end, align); |
bcb615a81 mm/vmalloc.c: fix... |
417 |
if (addr + size < addr) |
89699605f mm: vmap area cache |
418 |
goto overflow; |
92ca922f0 vmalloc: walk vma... |
419 |
if (list_is_last(&first->list, &vmap_area_list)) |
89699605f mm: vmap area cache |
420 |
goto found; |
92ca922f0 vmalloc: walk vma... |
421 422 423 |
first = list_entry(first->list.next, struct vmap_area, list); |
db64fe022 mm: rewrite vmap ... |
424 |
} |
89699605f mm: vmap area cache |
425 426 427 |
found: if (addr + size > vend) goto overflow; |
db64fe022 mm: rewrite vmap ... |
428 429 430 431 432 |
va->va_start = addr; va->va_end = addr + size; va->flags = 0; __insert_vmap_area(va); |
89699605f mm: vmap area cache |
433 |
free_vmap_cache = &va->rb_node; |
db64fe022 mm: rewrite vmap ... |
434 |
spin_unlock(&vmap_area_lock); |
89699605f mm: vmap area cache |
435 436 437 |
BUG_ON(va->va_start & (align-1)); BUG_ON(va->va_start < vstart); BUG_ON(va->va_end > vend); |
db64fe022 mm: rewrite vmap ... |
438 |
return va; |
89699605f mm: vmap area cache |
439 440 441 442 443 444 445 446 447 |
overflow: spin_unlock(&vmap_area_lock); if (!purged) { purge_vmap_area_lazy(); purged = 1; goto retry; } if (printk_ratelimit()) |
0cbc8533b mm/vmalloc.c: rep... |
448 |
pr_warn("vmap allocation for size %lu failed: " |
89699605f mm: vmap area cache |
449 450 451 452 |
"use vmalloc=<size> to increase size. ", size); kfree(va); return ERR_PTR(-EBUSY); |
db64fe022 mm: rewrite vmap ... |
453 |
} |
db64fe022 mm: rewrite vmap ... |
454 455 456 |
static void __free_vmap_area(struct vmap_area *va) { BUG_ON(RB_EMPTY_NODE(&va->rb_node)); |
89699605f mm: vmap area cache |
457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 |
if (free_vmap_cache) { if (va->va_end < cached_vstart) { free_vmap_cache = NULL; } else { struct vmap_area *cache; cache = rb_entry(free_vmap_cache, struct vmap_area, rb_node); if (va->va_start <= cache->va_start) { free_vmap_cache = rb_prev(&va->rb_node); /* * We don't try to update cached_hole_size or * cached_align, but it won't go very wrong. */ } } } |
db64fe022 mm: rewrite vmap ... |
473 474 475 |
rb_erase(&va->rb_node, &vmap_area_root); RB_CLEAR_NODE(&va->rb_node); list_del_rcu(&va->list); |
ca23e405e vmalloc: implemen... |
476 477 478 479 480 481 482 483 |
/* * Track the highest possible candidate for pcpu area * allocation. Areas outside of vmalloc area can be returned * here too, consider only end addresses which fall inside * vmalloc area proper. */ if (va->va_end > VMALLOC_START && va->va_end <= VMALLOC_END) vmap_area_pcpu_hole = max(vmap_area_pcpu_hole, va->va_end); |
14769de93 vmalloc,rcu: Conv... |
484 |
kfree_rcu(va, rcu_head); |
db64fe022 mm: rewrite vmap ... |
485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 |
} /* * 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); } |
cd52858c7 mm: vmalloc make ... |
504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 |
static void vmap_debug_free_range(unsigned long start, unsigned long end) { /* * Unmap page tables and force a TLB flush immediately if * CONFIG_DEBUG_PAGEALLOC is set. This catches use after free * bugs similarly to those in linear kernel virtual address * space after a page has been freed. * * All the lazy freeing logic is still retained, in order to * minimise intrusiveness of this debugging feature. * * This is going to be *slow* (linear kernel virtual address * debugging doesn't do a broadcast TLB flush so it is a lot * faster). */ #ifdef CONFIG_DEBUG_PAGEALLOC vunmap_page_range(start, end); flush_tlb_kernel_range(start, end); #endif } |
db64fe022 mm: rewrite vmap ... |
524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 |
/* * 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); } static atomic_t vmap_lazy_nr = ATOMIC_INIT(0); |
02b709df8 mm: purge fragmen... |
550 551 |
/* for per-CPU blocks */ static void purge_fragmented_blocks_allcpus(void); |
db64fe022 mm: rewrite vmap ... |
552 |
/* |
3ee48b6af mm, x86: Saving v... |
553 554 555 556 557 558 559 560 561 |
* called before a call to iounmap() if the caller wants vm_area_struct's * immediately freed. */ void set_iounmap_nonlazy(void) { atomic_set(&vmap_lazy_nr, lazy_max_pages()+1); } /* |
db64fe022 mm: rewrite vmap ... |
562 563 564 565 566 567 568 569 570 571 572 573 |
* Purges all lazily-freed vmap areas. * * If sync is 0 then don't purge if there is already a purge in progress. * If force_flush is 1, then flush kernel TLBs between *start and *end even * if we found no lazy vmap areas to unmap (callers can use this to optimise * their own TLB flushing). * Returns with *start = min(*start, lowest purged address) * *end = max(*end, highest purged address) */ static void __purge_vmap_area_lazy(unsigned long *start, unsigned long *end, int sync, int force_flush) { |
46666d8ac revert "mm: vmall... |
574 |
static DEFINE_SPINLOCK(purge_lock); |
db64fe022 mm: rewrite vmap ... |
575 576 |
LIST_HEAD(valist); struct vmap_area *va; |
cbb766766 mm: fix lazy vmap... |
577 |
struct vmap_area *n_va; |
db64fe022 mm: rewrite vmap ... |
578 579 580 581 582 583 584 585 |
int nr = 0; /* * If sync is 0 but force_flush is 1, we'll go sync anyway but callers * should not expect such behaviour. This just simplifies locking for * the case that isn't actually used at the moment anyway. */ if (!sync && !force_flush) { |
46666d8ac revert "mm: vmall... |
586 |
if (!spin_trylock(&purge_lock)) |
db64fe022 mm: rewrite vmap ... |
587 588 |
return; } else |
46666d8ac revert "mm: vmall... |
589 |
spin_lock(&purge_lock); |
db64fe022 mm: rewrite vmap ... |
590 |
|
02b709df8 mm: purge fragmen... |
591 592 |
if (sync) purge_fragmented_blocks_allcpus(); |
db64fe022 mm: rewrite vmap ... |
593 594 595 596 597 598 599 600 |
rcu_read_lock(); list_for_each_entry_rcu(va, &vmap_area_list, list) { if (va->flags & VM_LAZY_FREE) { if (va->va_start < *start) *start = va->va_start; if (va->va_end > *end) *end = va->va_end; nr += (va->va_end - va->va_start) >> PAGE_SHIFT; |
db64fe022 mm: rewrite vmap ... |
601 602 603 604 605 606 |
list_add_tail(&va->purge_list, &valist); va->flags |= VM_LAZY_FREEING; va->flags &= ~VM_LAZY_FREE; } } rcu_read_unlock(); |
88f500443 vmalloc: remove B... |
607 |
if (nr) |
db64fe022 mm: rewrite vmap ... |
608 |
atomic_sub(nr, &vmap_lazy_nr); |
db64fe022 mm: rewrite vmap ... |
609 610 611 612 613 614 |
if (nr || force_flush) flush_tlb_kernel_range(*start, *end); if (nr) { spin_lock(&vmap_area_lock); |
cbb766766 mm: fix lazy vmap... |
615 |
list_for_each_entry_safe(va, n_va, &valist, purge_list) |
db64fe022 mm: rewrite vmap ... |
616 617 618 |
__free_vmap_area(va); spin_unlock(&vmap_area_lock); } |
46666d8ac revert "mm: vmall... |
619 |
spin_unlock(&purge_lock); |
db64fe022 mm: rewrite vmap ... |
620 621 622 |
} /* |
496850e5f mm: vmalloc failu... |
623 624 625 626 627 628 629 630 631 632 633 |
* 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) { unsigned long start = ULONG_MAX, end = 0; __purge_vmap_area_lazy(&start, &end, 0, 0); } /* |
db64fe022 mm: rewrite vmap ... |
634 635 636 637 638 |
* Kick off a purge of the outstanding lazy areas. */ static void purge_vmap_area_lazy(void) { unsigned long start = ULONG_MAX, end = 0; |
496850e5f mm: vmalloc failu... |
639 |
__purge_vmap_area_lazy(&start, &end, 1, 0); |
db64fe022 mm: rewrite vmap ... |
640 641 642 |
} /* |
64141da58 vmalloc: eagerly ... |
643 644 645 |
* 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 ... |
646 |
*/ |
64141da58 vmalloc: eagerly ... |
647 |
static void free_vmap_area_noflush(struct vmap_area *va) |
db64fe022 mm: rewrite vmap ... |
648 649 650 651 |
{ va->flags |= VM_LAZY_FREE; atomic_add((va->va_end - va->va_start) >> PAGE_SHIFT, &vmap_lazy_nr); if (unlikely(atomic_read(&vmap_lazy_nr) > lazy_max_pages())) |
496850e5f mm: vmalloc failu... |
652 |
try_purge_vmap_area_lazy(); |
db64fe022 mm: rewrite vmap ... |
653 |
} |
b29acbdcf mm: vmalloc fix l... |
654 |
/* |
64141da58 vmalloc: eagerly ... |
655 656 657 658 659 660 661 662 663 664 |
* Free and unmap a vmap area, caller ensuring flush_cache_vunmap had been * called for the correct range previously. */ static void free_unmap_vmap_area_noflush(struct vmap_area *va) { unmap_vmap_area(va); free_vmap_area_noflush(va); } /* |
b29acbdcf mm: vmalloc fix l... |
665 666 667 668 669 670 671 |
* 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); free_unmap_vmap_area_noflush(va); } |
db64fe022 mm: rewrite vmap ... |
672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 |
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; } static void free_unmap_vmap_area_addr(unsigned long addr) { struct vmap_area *va; va = find_vmap_area(addr); BUG_ON(!va); free_unmap_vmap_area(va); } /*** 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... |
716 717 718 719 |
#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 ... |
720 721 |
#define VMAP_BLOCK_SIZE (VMAP_BBMAP_BITS * PAGE_SIZE) |
9b4633340 vmap: cope with v... |
722 |
static bool vmap_initialized __read_mostly = false; |
db64fe022 mm: rewrite vmap ... |
723 724 725 |
struct vmap_block_queue { spinlock_t lock; struct list_head free; |
db64fe022 mm: rewrite vmap ... |
726 727 728 729 730 |
}; struct vmap_block { spinlock_t lock; struct vmap_area *va; |
db64fe022 mm: rewrite vmap ... |
731 |
unsigned long free, dirty; |
7d61bfe8f mm/vmalloc: get r... |
732 |
unsigned long dirty_min, dirty_max; /*< dirty range */ |
de5604231 mm: percpu-vmap f... |
733 734 |
struct list_head free_list; struct rcu_head rcu_head; |
02b709df8 mm: purge fragmen... |
735 |
struct list_head purge; |
db64fe022 mm: rewrite vmap ... |
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 |
}; /* 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... |
762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 |
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 * * Returns: virtual address in a newly allocated block or ERR_PTR(-errno) */ static void *new_vmap_block(unsigned int order, gfp_t gfp_mask) |
db64fe022 mm: rewrite vmap ... |
780 781 782 783 784 785 |
{ struct vmap_block_queue *vbq; struct vmap_block *vb; struct vmap_area *va; unsigned long vb_idx; int node, err; |
cf725ce27 mm/vmalloc: occup... |
786 |
void *vaddr; |
db64fe022 mm: rewrite vmap ... |
787 788 789 790 791 792 793 794 795 796 797 |
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... |
798 |
if (IS_ERR(va)) { |
db64fe022 mm: rewrite vmap ... |
799 |
kfree(vb); |
e7d863407 mm: use ERR_CAST |
800 |
return ERR_CAST(va); |
db64fe022 mm: rewrite vmap ... |
801 802 803 804 805 806 807 808 |
} err = radix_tree_preload(gfp_mask); if (unlikely(err)) { kfree(vb); free_vmap_area(va); return ERR_PTR(err); } |
cf725ce27 mm/vmalloc: occup... |
809 |
vaddr = vmap_block_vaddr(va->va_start, 0); |
db64fe022 mm: rewrite vmap ... |
810 811 |
spin_lock_init(&vb->lock); vb->va = va; |
cf725ce27 mm/vmalloc: occup... |
812 813 814 |
/* 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 ... |
815 |
vb->dirty = 0; |
7d61bfe8f mm/vmalloc: get r... |
816 817 |
vb->dirty_min = VMAP_BBMAP_BITS; vb->dirty_max = 0; |
db64fe022 mm: rewrite vmap ... |
818 |
INIT_LIST_HEAD(&vb->free_list); |
db64fe022 mm: rewrite vmap ... |
819 820 821 822 823 824 825 826 827 |
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 ... |
828 |
spin_lock(&vbq->lock); |
68ac546f2 mm/vmalloc: fix p... |
829 |
list_add_tail_rcu(&vb->free_list, &vbq->free); |
db64fe022 mm: rewrite vmap ... |
830 |
spin_unlock(&vbq->lock); |
3f04ba859 vmalloc: fix use ... |
831 |
put_cpu_var(vmap_block_queue); |
db64fe022 mm: rewrite vmap ... |
832 |
|
cf725ce27 mm/vmalloc: occup... |
833 |
return vaddr; |
db64fe022 mm: rewrite vmap ... |
834 |
} |
db64fe022 mm: rewrite vmap ... |
835 836 837 838 |
static void free_vmap_block(struct vmap_block *vb) { struct vmap_block *tmp; unsigned long vb_idx; |
db64fe022 mm: rewrite vmap ... |
839 840 841 842 843 |
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 ... |
844 |
free_vmap_area_noflush(vb->va); |
22a3c7d18 vmalloc,rcu: Conv... |
845 |
kfree_rcu(vb, rcu_head); |
db64fe022 mm: rewrite vmap ... |
846 |
} |
02b709df8 mm: purge fragmen... |
847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 |
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... |
864 865 |
vb->dirty_min = 0; vb->dirty_max = VMAP_BBMAP_BITS; |
02b709df8 mm: purge fragmen... |
866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 |
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... |
881 882 883 884 885 886 887 |
static void purge_fragmented_blocks_allcpus(void) { int cpu; for_each_possible_cpu(cpu) purge_fragmented_blocks(cpu); } |
db64fe022 mm: rewrite vmap ... |
888 889 890 891 |
static void *vb_alloc(unsigned long size, gfp_t gfp_mask) { struct vmap_block_queue *vbq; struct vmap_block *vb; |
cf725ce27 mm/vmalloc: occup... |
892 |
void *vaddr = NULL; |
db64fe022 mm: rewrite vmap ... |
893 894 895 896 |
unsigned int order; BUG_ON(size & ~PAGE_MASK); BUG_ON(size > PAGE_SIZE*VMAP_MAX_ALLOC); |
aa91c4d89 mm: make vb_alloc... |
897 898 899 900 901 902 903 904 |
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 ... |
905 |
order = get_order(size); |
db64fe022 mm: rewrite vmap ... |
906 907 908 |
rcu_read_lock(); vbq = &get_cpu_var(vmap_block_queue); list_for_each_entry_rcu(vb, &vbq->free, free_list) { |
cf725ce27 mm/vmalloc: occup... |
909 |
unsigned long pages_off; |
db64fe022 mm: rewrite vmap ... |
910 911 |
spin_lock(&vb->lock); |
cf725ce27 mm/vmalloc: occup... |
912 913 914 915 |
if (vb->free < (1UL << order)) { spin_unlock(&vb->lock); continue; } |
02b709df8 mm: purge fragmen... |
916 |
|
cf725ce27 mm/vmalloc: occup... |
917 918 |
pages_off = VMAP_BBMAP_BITS - vb->free; vaddr = vmap_block_vaddr(vb->va->va_start, pages_off); |
02b709df8 mm: purge fragmen... |
919 920 921 922 923 924 |
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... |
925 |
|
02b709df8 mm: purge fragmen... |
926 927 |
spin_unlock(&vb->lock); break; |
db64fe022 mm: rewrite vmap ... |
928 |
} |
02b709df8 mm: purge fragmen... |
929 |
|
3f04ba859 vmalloc: fix use ... |
930 |
put_cpu_var(vmap_block_queue); |
db64fe022 mm: rewrite vmap ... |
931 |
rcu_read_unlock(); |
cf725ce27 mm/vmalloc: occup... |
932 933 934 |
/* Allocate new block if nothing was found */ if (!vaddr) vaddr = new_vmap_block(order, gfp_mask); |
db64fe022 mm: rewrite vmap ... |
935 |
|
cf725ce27 mm/vmalloc: occup... |
936 |
return vaddr; |
db64fe022 mm: rewrite vmap ... |
937 938 939 940 941 942 943 944 945 946 947 |
} static void vb_free(const void *addr, unsigned long size) { unsigned long offset; unsigned long vb_idx; unsigned int order; struct vmap_block *vb; BUG_ON(size & ~PAGE_MASK); BUG_ON(size > PAGE_SIZE*VMAP_MAX_ALLOC); |
b29acbdcf mm: vmalloc fix l... |
948 949 |
flush_cache_vunmap((unsigned long)addr, (unsigned long)addr + size); |
db64fe022 mm: rewrite vmap ... |
950 951 952 |
order = get_order(size); offset = (unsigned long)addr & (VMAP_BLOCK_SIZE - 1); |
7d61bfe8f mm/vmalloc: get r... |
953 |
offset >>= PAGE_SHIFT; |
db64fe022 mm: rewrite vmap ... |
954 955 956 957 958 959 |
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 ... |
960 |
vunmap_page_range((unsigned long)addr, (unsigned long)addr + size); |
db64fe022 mm: rewrite vmap ... |
961 |
spin_lock(&vb->lock); |
7d61bfe8f mm/vmalloc: get r... |
962 963 964 965 |
/* 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... |
966 |
|
db64fe022 mm: rewrite vmap ... |
967 968 |
vb->dirty += 1UL << order; if (vb->dirty == VMAP_BBMAP_BITS) { |
de5604231 mm: percpu-vmap f... |
969 |
BUG_ON(vb->free); |
db64fe022 mm: rewrite vmap ... |
970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 |
spin_unlock(&vb->lock); free_vmap_block(vb); } else spin_unlock(&vb->lock); } /** * 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 cpu; int flush = 0; |
9b4633340 vmap: cope with v... |
994 995 |
if (unlikely(!vmap_initialized)) return; |
db64fe022 mm: rewrite vmap ... |
996 997 998 999 1000 1001 |
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 ... |
1002 |
spin_lock(&vb->lock); |
7d61bfe8f mm/vmalloc: get r... |
1003 1004 |
if (vb->dirty) { unsigned long va_start = vb->va->va_start; |
db64fe022 mm: rewrite vmap ... |
1005 |
unsigned long s, e; |
b136be5e0 mm, vmalloc: use ... |
1006 |
|
7d61bfe8f mm/vmalloc: get r... |
1007 1008 |
s = va_start + (vb->dirty_min << PAGE_SHIFT); e = va_start + (vb->dirty_max << PAGE_SHIFT); |
db64fe022 mm: rewrite vmap ... |
1009 |
|
7d61bfe8f mm/vmalloc: get r... |
1010 1011 |
start = min(s, start); end = max(e, end); |
db64fe022 mm: rewrite vmap ... |
1012 |
|
7d61bfe8f mm/vmalloc: get r... |
1013 |
flush = 1; |
db64fe022 mm: rewrite vmap ... |
1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 |
} spin_unlock(&vb->lock); } rcu_read_unlock(); } __purge_vmap_area_lazy(&start, &end, 1, flush); } 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) { unsigned long size = count << PAGE_SHIFT; unsigned long addr = (unsigned long)mem; BUG_ON(!addr); BUG_ON(addr < VMALLOC_START); BUG_ON(addr > VMALLOC_END); BUG_ON(addr & (PAGE_SIZE-1)); debug_check_no_locks_freed(mem, size); |
cd52858c7 mm: vmalloc make ... |
1040 |
vmap_debug_free_range(addr, addr+size); |
db64fe022 mm: rewrite vmap ... |
1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 |
if (likely(count <= VMAP_MAX_ALLOC)) vb_free(mem, size); else free_unmap_vmap_area_addr(addr); } 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... |
1055 |
* |
364376383 mm/vmalloc.c: enh... |
1056 1057 1058 1059 1060 1061 |
* 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... |
1062 |
* Returns: a pointer to the address that has been mapped, or %NULL on failure |
db64fe022 mm: rewrite vmap ... |
1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 |
*/ void *vm_map_ram(struct page **pages, unsigned int count, int node, pgprot_t prot) { unsigned long size = count << PAGE_SHIFT; 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... |
1092 |
static struct vm_struct *vmlist __initdata; |
f0aa66179 vmalloc: implemen... |
1093 |
/** |
be9b7335e mm: add vm_area_a... |
1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 |
* 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... |
1120 1121 |
* vm_area_register_early - register vmap area early during boot * @vm: vm_struct to register |
c0c0a2937 vmalloc: add @ali... |
1122 |
* @align: requested alignment |
f0aa66179 vmalloc: implemen... |
1123 1124 1125 1126 1127 1128 1129 1130 |
* * 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... |
1131 |
void __init vm_area_register_early(struct vm_struct *vm, size_t align) |
f0aa66179 vmalloc: implemen... |
1132 1133 |
{ static size_t vm_init_off __initdata; |
c0c0a2937 vmalloc: add @ali... |
1134 1135 1136 1137 |
unsigned long addr; addr = ALIGN(VMALLOC_START + vm_init_off, align); vm_init_off = PFN_ALIGN(addr + vm->size) - VMALLOC_START; |
f0aa66179 vmalloc: implemen... |
1138 |
|
c0c0a2937 vmalloc: add @ali... |
1139 |
vm->addr = (void *)addr; |
f0aa66179 vmalloc: implemen... |
1140 |
|
be9b7335e mm: add vm_area_a... |
1141 |
vm_area_add_early(vm); |
f0aa66179 vmalloc: implemen... |
1142 |
} |
db64fe022 mm: rewrite vmap ... |
1143 1144 |
void __init vmalloc_init(void) { |
822c18f2e alpha: fix vmallo... |
1145 1146 |
struct vmap_area *va; struct vm_struct *tmp; |
db64fe022 mm: rewrite vmap ... |
1147 1148 1149 1150 |
int i; for_each_possible_cpu(i) { struct vmap_block_queue *vbq; |
32fcfd407 make vfree() safe... |
1151 |
struct vfree_deferred *p; |
db64fe022 mm: rewrite vmap ... |
1152 1153 1154 1155 |
vbq = &per_cpu(vmap_block_queue, i); spin_lock_init(&vbq->lock); INIT_LIST_HEAD(&vbq->free); |
32fcfd407 make vfree() safe... |
1156 1157 1158 |
p = &per_cpu(vfree_deferred, i); init_llist_head(&p->list); INIT_WORK(&p->wq, free_work); |
db64fe022 mm: rewrite vmap ... |
1159 |
} |
9b4633340 vmap: cope with v... |
1160 |
|
822c18f2e alpha: fix vmallo... |
1161 1162 |
/* Import existing vmlist entries. */ for (tmp = vmlist; tmp; tmp = tmp->next) { |
43ebdac42 vmalloc: use kzal... |
1163 |
va = kzalloc(sizeof(struct vmap_area), GFP_NOWAIT); |
dbda591d9 mm: fix faulty in... |
1164 |
va->flags = VM_VM_AREA; |
822c18f2e alpha: fix vmallo... |
1165 1166 |
va->va_start = (unsigned long)tmp->addr; va->va_end = va->va_start + tmp->size; |
dbda591d9 mm: fix faulty in... |
1167 |
va->vm = tmp; |
822c18f2e alpha: fix vmallo... |
1168 1169 |
__insert_vmap_area(va); } |
ca23e405e vmalloc: implemen... |
1170 1171 |
vmap_area_pcpu_hole = VMALLOC_END; |
9b4633340 vmap: cope with v... |
1172 |
vmap_initialized = true; |
db64fe022 mm: rewrite vmap ... |
1173 |
} |
8fc489850 vmalloc: add un/m... |
1174 1175 1176 1177 1178 1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199 1200 1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 |
/** * 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... |
1217 |
EXPORT_SYMBOL_GPL(unmap_kernel_range_noflush); |
8fc489850 vmalloc: add un/m... |
1218 1219 1220 1221 1222 1223 1224 1225 1226 |
/** * 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 ... |
1227 1228 1229 |
void unmap_kernel_range(unsigned long addr, unsigned long size) { unsigned long end = addr + size; |
f6fcba701 vmalloc: call flu... |
1230 1231 |
flush_cache_vunmap(addr, end); |
db64fe022 mm: rewrite vmap ... |
1232 1233 1234 |
vunmap_page_range(addr, end); flush_tlb_kernel_range(addr, end); } |
93ef6d6ca mm/vmalloc.c: exp... |
1235 |
EXPORT_SYMBOL_GPL(unmap_kernel_range); |
db64fe022 mm: rewrite vmap ... |
1236 |
|
f6f8ed473 mm/vmalloc.c: cle... |
1237 |
int map_vm_area(struct vm_struct *area, pgprot_t prot, struct page **pages) |
db64fe022 mm: rewrite vmap ... |
1238 1239 |
{ unsigned long addr = (unsigned long)area->addr; |
762216ab4 mm/vmalloc: use w... |
1240 |
unsigned long end = addr + get_vm_area_size(area); |
db64fe022 mm: rewrite vmap ... |
1241 |
int err; |
f6f8ed473 mm/vmalloc.c: cle... |
1242 |
err = vmap_page_range(addr, end, prot, pages); |
db64fe022 mm: rewrite vmap ... |
1243 |
|
f6f8ed473 mm/vmalloc.c: cle... |
1244 |
return err > 0 ? 0 : err; |
db64fe022 mm: rewrite vmap ... |
1245 1246 |
} EXPORT_SYMBOL_GPL(map_vm_area); |
f5252e009 mm: avoid null po... |
1247 |
static void setup_vmalloc_vm(struct vm_struct *vm, struct vmap_area *va, |
5e6cafc83 mm: vmalloc: use ... |
1248 |
unsigned long flags, const void *caller) |
cf88c7900 vmalloc: separate... |
1249 |
{ |
c69480ade mm, vmalloc: prot... |
1250 |
spin_lock(&vmap_area_lock); |
cf88c7900 vmalloc: separate... |
1251 1252 1253 1254 |
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... |
1255 |
va->vm = vm; |
cf88c7900 vmalloc: separate... |
1256 |
va->flags |= VM_VM_AREA; |
c69480ade mm, vmalloc: prot... |
1257 |
spin_unlock(&vmap_area_lock); |
f5252e009 mm: avoid null po... |
1258 |
} |
cf88c7900 vmalloc: separate... |
1259 |
|
20fc02b47 mm/vmalloc.c: ren... |
1260 |
static void clear_vm_uninitialized_flag(struct vm_struct *vm) |
f5252e009 mm: avoid null po... |
1261 |
{ |
d4033afdf mm, vmalloc: iter... |
1262 |
/* |
20fc02b47 mm/vmalloc.c: ren... |
1263 |
* Before removing VM_UNINITIALIZED, |
d4033afdf mm, vmalloc: iter... |
1264 1265 1266 1267 |
* we should make sure that vm has proper values. * Pair with smp_rmb() in show_numa_info(). */ smp_wmb(); |
20fc02b47 mm/vmalloc.c: ren... |
1268 |
vm->flags &= ~VM_UNINITIALIZED; |
cf88c7900 vmalloc: separate... |
1269 |
} |
db64fe022 mm: rewrite vmap ... |
1270 |
static struct vm_struct *__get_vm_area_node(unsigned long size, |
2dca6999e mm, perf_event: M... |
1271 |
unsigned long align, unsigned long flags, unsigned long start, |
5e6cafc83 mm: vmalloc: use ... |
1272 |
unsigned long end, int node, gfp_t gfp_mask, const void *caller) |
db64fe022 mm: rewrite vmap ... |
1273 |
{ |
0006526d7 mm/vmalloc.c: rem... |
1274 |
struct vmap_area *va; |
db64fe022 mm: rewrite vmap ... |
1275 |
struct vm_struct *area; |
1da177e4c Linux-2.6.12-rc2 |
1276 |
|
52fd24ca1 [PATCH] __vmalloc... |
1277 |
BUG_ON(in_interrupt()); |
0f2d4a8e2 mm, vmalloc: use ... |
1278 |
if (flags & VM_IOREMAP) |
0f616be12 mm: change __get_... |
1279 1280 |
align = 1ul << clamp_t(int, fls_long(size), PAGE_SHIFT, IOREMAP_MAX_ORDER); |
db64fe022 mm: rewrite vmap ... |
1281 |
|
1da177e4c Linux-2.6.12-rc2 |
1282 |
size = PAGE_ALIGN(size); |
31be83095 [PATCH] Fix stran... |
1283 1284 |
if (unlikely(!size)) return NULL; |
1da177e4c Linux-2.6.12-rc2 |
1285 |
|
cf88c7900 vmalloc: separate... |
1286 |
area = kzalloc_node(sizeof(*area), gfp_mask & GFP_RECLAIM_MASK, node); |
1da177e4c Linux-2.6.12-rc2 |
1287 1288 |
if (unlikely(!area)) return NULL; |
71394fe50 mm: vmalloc: add ... |
1289 1290 |
if (!(flags & VM_NO_GUARD)) size += PAGE_SIZE; |
1da177e4c Linux-2.6.12-rc2 |
1291 |
|
db64fe022 mm: rewrite vmap ... |
1292 1293 1294 1295 |
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 |
1296 |
} |
1da177e4c Linux-2.6.12-rc2 |
1297 |
|
d82b1d857 mm, vmalloc: only... |
1298 |
setup_vmalloc_vm(area, va, flags, caller); |
f5252e009 mm: avoid null po... |
1299 |
|
1da177e4c Linux-2.6.12-rc2 |
1300 |
return area; |
1da177e4c Linux-2.6.12-rc2 |
1301 |
} |
930fc45a4 [PATCH] vmalloc_node |
1302 1303 1304 |
struct vm_struct *__get_vm_area(unsigned long size, unsigned long flags, unsigned long start, unsigned long end) { |
00ef2d2f8 mm: use NUMA_NO_NODE |
1305 1306 |
return __get_vm_area_node(size, 1, flags, start, end, NUMA_NO_NODE, GFP_KERNEL, __builtin_return_address(0)); |
930fc45a4 [PATCH] vmalloc_node |
1307 |
} |
5992b6dac lguest: export sy... |
1308 |
EXPORT_SYMBOL_GPL(__get_vm_area); |
930fc45a4 [PATCH] vmalloc_node |
1309 |
|
c29686129 vmalloc: add __ge... |
1310 1311 |
struct vm_struct *__get_vm_area_caller(unsigned long size, unsigned long flags, unsigned long start, unsigned long end, |
5e6cafc83 mm: vmalloc: use ... |
1312 |
const void *caller) |
c29686129 vmalloc: add __ge... |
1313 |
{ |
00ef2d2f8 mm: use NUMA_NO_NODE |
1314 1315 |
return __get_vm_area_node(size, 1, flags, start, end, NUMA_NO_NODE, GFP_KERNEL, caller); |
c29686129 vmalloc: add __ge... |
1316 |
} |
1da177e4c Linux-2.6.12-rc2 |
1317 |
/** |
183ff22bb spelling fixes: mm/ |
1318 |
* get_vm_area - reserve a contiguous kernel virtual area |
1da177e4c Linux-2.6.12-rc2 |
1319 1320 1321 1322 1323 1324 1325 1326 1327 |
* @size: size of the area * @flags: %VM_IOREMAP for I/O mappings or VM_ALLOC * * 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. */ struct vm_struct *get_vm_area(unsigned long size, unsigned long flags) { |
2dca6999e mm, perf_event: M... |
1328 |
return __get_vm_area_node(size, 1, flags, VMALLOC_START, VMALLOC_END, |
00ef2d2f8 mm: use NUMA_NO_NODE |
1329 1330 |
NUMA_NO_NODE, GFP_KERNEL, __builtin_return_address(0)); |
230169693 vmallocinfo: add ... |
1331 1332 1333 |
} struct vm_struct *get_vm_area_caller(unsigned long size, unsigned long flags, |
5e6cafc83 mm: vmalloc: use ... |
1334 |
const void *caller) |
230169693 vmallocinfo: add ... |
1335 |
{ |
2dca6999e mm, perf_event: M... |
1336 |
return __get_vm_area_node(size, 1, flags, VMALLOC_START, VMALLOC_END, |
00ef2d2f8 mm: use NUMA_NO_NODE |
1337 |
NUMA_NO_NODE, GFP_KERNEL, caller); |
1da177e4c Linux-2.6.12-rc2 |
1338 |
} |
e9da6e990 ARM: dma-mapping:... |
1339 1340 1341 1342 1343 1344 1345 1346 1347 |
/** * find_vm_area - find a continuous kernel virtual area * @addr: base address * * 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. */ struct vm_struct *find_vm_area(const void *addr) |
833423143 [PATCH] mm: intro... |
1348 |
{ |
db64fe022 mm: rewrite vmap ... |
1349 |
struct vmap_area *va; |
833423143 [PATCH] mm: intro... |
1350 |
|
db64fe022 mm: rewrite vmap ... |
1351 1352 |
va = find_vmap_area((unsigned long)addr); if (va && va->flags & VM_VM_AREA) |
db1aecafe mm/vmalloc.c: cha... |
1353 |
return va->vm; |
1da177e4c Linux-2.6.12-rc2 |
1354 |
|
1da177e4c Linux-2.6.12-rc2 |
1355 |
return NULL; |
1da177e4c Linux-2.6.12-rc2 |
1356 |
} |
7856dfeb2 [PATCH] x86_64: F... |
1357 |
/** |
183ff22bb spelling fixes: mm/ |
1358 |
* remove_vm_area - find and remove a continuous kernel virtual area |
7856dfeb2 [PATCH] x86_64: F... |
1359 1360 1361 1362 1363 1364 |
* @addr: base address * * 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. */ |
b3bdda02a vmalloc: add cons... |
1365 |
struct vm_struct *remove_vm_area(const void *addr) |
7856dfeb2 [PATCH] x86_64: F... |
1366 |
{ |
db64fe022 mm: rewrite vmap ... |
1367 1368 1369 1370 |
struct vmap_area *va; va = find_vmap_area((unsigned long)addr); if (va && va->flags & VM_VM_AREA) { |
db1aecafe mm/vmalloc.c: cha... |
1371 |
struct vm_struct *vm = va->vm; |
f5252e009 mm: avoid null po... |
1372 |
|
c69480ade mm, vmalloc: prot... |
1373 1374 1375 1376 |
spin_lock(&vmap_area_lock); va->vm = NULL; va->flags &= ~VM_VM_AREA; spin_unlock(&vmap_area_lock); |
dd32c2799 vmalloc: unmap vm... |
1377 |
vmap_debug_free_range(va->va_start, va->va_end); |
a5af5aa8b kasan, module, vm... |
1378 |
kasan_free_shadow(vm); |
dd32c2799 vmalloc: unmap vm... |
1379 1380 |
free_unmap_vmap_area(va); vm->size -= PAGE_SIZE; |
db64fe022 mm: rewrite vmap ... |
1381 1382 1383 |
return vm; } return NULL; |
7856dfeb2 [PATCH] x86_64: F... |
1384 |
} |
b3bdda02a vmalloc: add cons... |
1385 |
static void __vunmap(const void *addr, int deallocate_pages) |
1da177e4c Linux-2.6.12-rc2 |
1386 1387 1388 1389 1390 |
{ struct vm_struct *area; if (!addr) return; |
e69e9d4ae vmalloc: introduc... |
1391 1392 |
if (WARN(!PAGE_ALIGNED(addr), "Trying to vfree() bad address (%p) ", |
ab15d9b4c mm/vmalloc.c: unb... |
1393 |
addr)) |
1da177e4c Linux-2.6.12-rc2 |
1394 |
return; |
1da177e4c Linux-2.6.12-rc2 |
1395 1396 1397 |
area = remove_vm_area(addr); if (unlikely(!area)) { |
4c8573e25 Use WARN() in mm/... |
1398 1399 |
WARN(1, KERN_ERR "Trying to vfree() nonexistent vm area (%p) ", |
1da177e4c Linux-2.6.12-rc2 |
1400 |
addr); |
1da177e4c Linux-2.6.12-rc2 |
1401 1402 |
return; } |
9a11b49a8 [PATCH] lockdep: ... |
1403 |
debug_check_no_locks_freed(addr, area->size); |
3ac7fe5a4 infrastructure to... |
1404 |
debug_check_no_obj_freed(addr, area->size); |
9a11b49a8 [PATCH] lockdep: ... |
1405 |
|
1da177e4c Linux-2.6.12-rc2 |
1406 1407 1408 1409 |
if (deallocate_pages) { int i; for (i = 0; i < area->nr_pages; i++) { |
bf53d6f8f vmalloc: clean up... |
1410 1411 1412 1413 |
struct page *page = area->pages[i]; BUG_ON(!page); __free_page(page); |
1da177e4c Linux-2.6.12-rc2 |
1414 |
} |
8757d5fa6 [PATCH] mm: fix o... |
1415 |
if (area->flags & VM_VPAGES) |
1da177e4c Linux-2.6.12-rc2 |
1416 1417 1418 1419 1420 1421 1422 1423 |
vfree(area->pages); else kfree(area->pages); } kfree(area); return; } |
32fcfd407 make vfree() safe... |
1424 |
|
1da177e4c Linux-2.6.12-rc2 |
1425 1426 |
/** * vfree - release memory allocated by vmalloc() |
1da177e4c Linux-2.6.12-rc2 |
1427 1428 |
* @addr: memory base address * |
183ff22bb spelling fixes: mm/ |
1429 |
* Free the virtually continuous memory area starting at @addr, as |
80e93effc [PATCH] update kf... |
1430 1431 |
* obtained from vmalloc(), vmalloc_32() or __vmalloc(). If @addr is * NULL, no operation is performed. |
1da177e4c Linux-2.6.12-rc2 |
1432 |
* |
32fcfd407 make vfree() safe... |
1433 1434 1435 |
* 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... |
1436 1437 |
* * NOTE: assumes that the object at *addr has a size >= sizeof(llist_node) |
1da177e4c Linux-2.6.12-rc2 |
1438 |
*/ |
b3bdda02a vmalloc: add cons... |
1439 |
void vfree(const void *addr) |
1da177e4c Linux-2.6.12-rc2 |
1440 |
{ |
32fcfd407 make vfree() safe... |
1441 |
BUG_ON(in_nmi()); |
89219d37a kmemleak: Add the... |
1442 1443 |
kmemleak_free(addr); |
32fcfd407 make vfree() safe... |
1444 1445 1446 |
if (!addr) return; if (unlikely(in_interrupt())) { |
7c8e0181e mm: replace __get... |
1447 |
struct vfree_deferred *p = this_cpu_ptr(&vfree_deferred); |
59d3132f8 vfree: don't sche... |
1448 1449 |
if (llist_add((struct llist_node *)addr, &p->list)) schedule_work(&p->wq); |
32fcfd407 make vfree() safe... |
1450 1451 |
} else __vunmap(addr, 1); |
1da177e4c Linux-2.6.12-rc2 |
1452 |
} |
1da177e4c Linux-2.6.12-rc2 |
1453 1454 1455 1456 |
EXPORT_SYMBOL(vfree); /** * vunmap - release virtual mapping obtained by vmap() |
1da177e4c Linux-2.6.12-rc2 |
1457 1458 1459 1460 1461 |
* @addr: memory base address * * Free the virtually contiguous memory area starting at @addr, * which was created from the page array passed to vmap(). * |
80e93effc [PATCH] update kf... |
1462 |
* Must not be called in interrupt context. |
1da177e4c Linux-2.6.12-rc2 |
1463 |
*/ |
b3bdda02a vmalloc: add cons... |
1464 |
void vunmap(const void *addr) |
1da177e4c Linux-2.6.12-rc2 |
1465 1466 |
{ BUG_ON(in_interrupt()); |
34754b69a x86: make vmap ye... |
1467 |
might_sleep(); |
32fcfd407 make vfree() safe... |
1468 1469 |
if (addr) __vunmap(addr, 0); |
1da177e4c Linux-2.6.12-rc2 |
1470 |
} |
1da177e4c Linux-2.6.12-rc2 |
1471 1472 1473 1474 |
EXPORT_SYMBOL(vunmap); /** * vmap - map an array of pages into virtually contiguous space |
1da177e4c Linux-2.6.12-rc2 |
1475 1476 1477 1478 1479 1480 1481 1482 1483 1484 1485 1486 |
* @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. */ void *vmap(struct page **pages, unsigned int count, unsigned long flags, pgprot_t prot) { struct vm_struct *area; |
34754b69a x86: make vmap ye... |
1487 |
might_sleep(); |
4481374ce mm: replace vario... |
1488 |
if (count > totalram_pages) |
1da177e4c Linux-2.6.12-rc2 |
1489 |
return NULL; |
230169693 vmallocinfo: add ... |
1490 1491 |
area = get_vm_area_caller((count << PAGE_SHIFT), flags, __builtin_return_address(0)); |
1da177e4c Linux-2.6.12-rc2 |
1492 1493 |
if (!area) return NULL; |
230169693 vmallocinfo: add ... |
1494 |
|
f6f8ed473 mm/vmalloc.c: cle... |
1495 |
if (map_vm_area(area, prot, pages)) { |
1da177e4c Linux-2.6.12-rc2 |
1496 1497 1498 1499 1500 1501 |
vunmap(area->addr); return NULL; } return area->addr; } |
1da177e4c Linux-2.6.12-rc2 |
1502 |
EXPORT_SYMBOL(vmap); |
2dca6999e mm, perf_event: M... |
1503 1504 |
static void *__vmalloc_node(unsigned long size, unsigned long align, gfp_t gfp_mask, pgprot_t prot, |
5e6cafc83 mm: vmalloc: use ... |
1505 |
int node, const void *caller); |
e31d9eb5c make __vmalloc_ar... |
1506 |
static void *__vmalloc_area_node(struct vm_struct *area, gfp_t gfp_mask, |
3722e13cf mm/vmalloc: don't... |
1507 |
pgprot_t prot, int node) |
1da177e4c Linux-2.6.12-rc2 |
1508 |
{ |
22943ab11 mm: print vmalloc... |
1509 |
const int order = 0; |
1da177e4c Linux-2.6.12-rc2 |
1510 1511 |
struct page **pages; unsigned int nr_pages, array_size, i; |
930f036b4 mm, vmalloc: cons... |
1512 1513 |
const gfp_t nested_gfp = (gfp_mask & GFP_RECLAIM_MASK) | __GFP_ZERO; const gfp_t alloc_mask = gfp_mask | __GFP_NOWARN; |
1da177e4c Linux-2.6.12-rc2 |
1514 |
|
762216ab4 mm/vmalloc: use w... |
1515 |
nr_pages = get_vm_area_size(area) >> PAGE_SHIFT; |
1da177e4c Linux-2.6.12-rc2 |
1516 1517 1518 1519 |
array_size = (nr_pages * sizeof(struct page *)); area->nr_pages = nr_pages; /* Please note that the recursion is strictly bounded. */ |
8757d5fa6 [PATCH] mm: fix o... |
1520 |
if (array_size > PAGE_SIZE) { |
976d6dfbb vmalloc(): adjust... |
1521 |
pages = __vmalloc_node(array_size, 1, nested_gfp|__GFP_HIGHMEM, |
3722e13cf mm/vmalloc: don't... |
1522 |
PAGE_KERNEL, node, area->caller); |
8757d5fa6 [PATCH] mm: fix o... |
1523 |
area->flags |= VM_VPAGES; |
286e1ea3a [PATCH] vmalloc()... |
1524 |
} else { |
976d6dfbb vmalloc(): adjust... |
1525 |
pages = kmalloc_node(array_size, nested_gfp, node); |
286e1ea3a [PATCH] vmalloc()... |
1526 |
} |
1da177e4c Linux-2.6.12-rc2 |
1527 1528 1529 1530 1531 1532 |
area->pages = pages; if (!area->pages) { remove_vm_area(area->addr); kfree(area); return NULL; } |
1da177e4c Linux-2.6.12-rc2 |
1533 1534 |
for (i = 0; i < area->nr_pages; i++) { |
bf53d6f8f vmalloc: clean up... |
1535 |
struct page *page; |
4b90951c0 mm/vmalloc: use N... |
1536 |
if (node == NUMA_NO_NODE) |
930f036b4 mm, vmalloc: cons... |
1537 |
page = alloc_page(alloc_mask); |
930fc45a4 [PATCH] vmalloc_node |
1538 |
else |
930f036b4 mm, vmalloc: cons... |
1539 |
page = alloc_pages_node(node, alloc_mask, order); |
bf53d6f8f vmalloc: clean up... |
1540 1541 |
if (unlikely(!page)) { |
1da177e4c Linux-2.6.12-rc2 |
1542 1543 1544 1545 |
/* Successfully allocated i pages, free them in __vunmap() */ area->nr_pages = i; goto fail; } |
bf53d6f8f vmalloc: clean up... |
1546 |
area->pages[i] = page; |
660654f90 mm/vmalloc.c: add... |
1547 1548 |
if (gfp_mask & __GFP_WAIT) cond_resched(); |
1da177e4c Linux-2.6.12-rc2 |
1549 |
} |
f6f8ed473 mm/vmalloc.c: cle... |
1550 |
if (map_vm_area(area, prot, pages)) |
1da177e4c Linux-2.6.12-rc2 |
1551 1552 1553 1554 |
goto fail; return area->addr; fail: |
3ee9a4f08 mm: neaten warn_a... |
1555 1556 1557 |
warn_alloc_failed(gfp_mask, order, "vmalloc: allocation failure, allocated %ld of %ld bytes ", |
22943ab11 mm: print vmalloc... |
1558 |
(area->nr_pages*PAGE_SIZE), area->size); |
1da177e4c Linux-2.6.12-rc2 |
1559 1560 1561 1562 1563 |
vfree(area->addr); return NULL; } /** |
d0a21265d mm: unify module_... |
1564 |
* __vmalloc_node_range - allocate virtually contiguous memory |
1da177e4c Linux-2.6.12-rc2 |
1565 |
* @size: allocation size |
2dca6999e mm, perf_event: M... |
1566 |
* @align: desired alignment |
d0a21265d mm: unify module_... |
1567 1568 |
* @start: vm area range start * @end: vm area range end |
1da177e4c Linux-2.6.12-rc2 |
1569 1570 |
* @gfp_mask: flags for the page level allocator * @prot: protection mask for the allocated pages |
cb9e3c292 mm: vmalloc: pass... |
1571 |
* @vm_flags: additional vm area flags (e.g. %VM_NO_GUARD) |
00ef2d2f8 mm: use NUMA_NO_NODE |
1572 |
* @node: node to use for allocation or NUMA_NO_NODE |
c85d194bf docbook: fix vmal... |
1573 |
* @caller: caller's return address |
1da177e4c Linux-2.6.12-rc2 |
1574 1575 1576 1577 1578 |
* * 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. */ |
d0a21265d mm: unify module_... |
1579 1580 |
void *__vmalloc_node_range(unsigned long size, unsigned long align, unsigned long start, unsigned long end, gfp_t gfp_mask, |
cb9e3c292 mm: vmalloc: pass... |
1581 1582 |
pgprot_t prot, unsigned long vm_flags, int node, const void *caller) |
1da177e4c Linux-2.6.12-rc2 |
1583 1584 |
{ struct vm_struct *area; |
89219d37a kmemleak: Add the... |
1585 1586 |
void *addr; unsigned long real_size = size; |
1da177e4c Linux-2.6.12-rc2 |
1587 1588 |
size = PAGE_ALIGN(size); |
4481374ce mm: replace vario... |
1589 |
if (!size || (size >> PAGE_SHIFT) > totalram_pages) |
de7d2b567 mm/vmalloc.c: rep... |
1590 |
goto fail; |
1da177e4c Linux-2.6.12-rc2 |
1591 |
|
cb9e3c292 mm: vmalloc: pass... |
1592 1593 |
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 |
1594 |
if (!area) |
de7d2b567 mm/vmalloc.c: rep... |
1595 |
goto fail; |
1da177e4c Linux-2.6.12-rc2 |
1596 |
|
3722e13cf mm/vmalloc: don't... |
1597 |
addr = __vmalloc_area_node(area, gfp_mask, prot, node); |
1368edf06 mm: vmalloc: chec... |
1598 |
if (!addr) |
b82225f3f revert mm/vmalloc... |
1599 |
return NULL; |
89219d37a kmemleak: Add the... |
1600 1601 |
/* |
20fc02b47 mm/vmalloc.c: ren... |
1602 1603 |
* In this function, newly allocated vm_struct has VM_UNINITIALIZED * flag. It means that vm_struct is not fully initialized. |
4341fa454 mm, vmalloc: remo... |
1604 |
* Now, it is fully initialized, so remove this flag here. |
f5252e009 mm: avoid null po... |
1605 |
*/ |
20fc02b47 mm/vmalloc.c: ren... |
1606 |
clear_vm_uninitialized_flag(area); |
f5252e009 mm: avoid null po... |
1607 1608 |
/* |
7f88f88f8 mm: kmemleak: avo... |
1609 1610 1611 |
* A ref_count = 2 is needed because vm_struct allocated in * __get_vm_area_node() contains a reference to the virtual address of * the vmalloc'ed block. |
89219d37a kmemleak: Add the... |
1612 |
*/ |
7f88f88f8 mm: kmemleak: avo... |
1613 |
kmemleak_alloc(addr, real_size, 2, gfp_mask); |
89219d37a kmemleak: Add the... |
1614 1615 |
return addr; |
de7d2b567 mm/vmalloc.c: rep... |
1616 1617 1618 1619 1620 1621 1622 |
fail: warn_alloc_failed(gfp_mask, 0, "vmalloc: allocation failure: %lu bytes ", real_size); return NULL; |
1da177e4c Linux-2.6.12-rc2 |
1623 |
} |
d0a21265d mm: unify module_... |
1624 1625 1626 1627 1628 1629 |
/** * __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 |
00ef2d2f8 mm: use NUMA_NO_NODE |
1630 |
* @node: node to use for allocation or NUMA_NO_NODE |
d0a21265d mm: unify module_... |
1631 1632 1633 1634 1635 1636 1637 1638 |
* @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. */ static void *__vmalloc_node(unsigned long size, unsigned long align, gfp_t gfp_mask, pgprot_t prot, |
5e6cafc83 mm: vmalloc: use ... |
1639 |
int node, const void *caller) |
d0a21265d mm: unify module_... |
1640 1641 |
{ return __vmalloc_node_range(size, align, VMALLOC_START, VMALLOC_END, |
cb9e3c292 mm: vmalloc: pass... |
1642 |
gfp_mask, prot, 0, node, caller); |
d0a21265d mm: unify module_... |
1643 |
} |
930fc45a4 [PATCH] vmalloc_node |
1644 1645 |
void *__vmalloc(unsigned long size, gfp_t gfp_mask, pgprot_t prot) { |
00ef2d2f8 mm: use NUMA_NO_NODE |
1646 |
return __vmalloc_node(size, 1, gfp_mask, prot, NUMA_NO_NODE, |
230169693 vmallocinfo: add ... |
1647 |
__builtin_return_address(0)); |
930fc45a4 [PATCH] vmalloc_node |
1648 |
} |
1da177e4c Linux-2.6.12-rc2 |
1649 |
EXPORT_SYMBOL(__vmalloc); |
e1ca7788d mm: add vzalloc()... |
1650 1651 1652 1653 1654 1655 |
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)); } |
1da177e4c Linux-2.6.12-rc2 |
1656 1657 |
/** * vmalloc - allocate virtually contiguous memory |
1da177e4c Linux-2.6.12-rc2 |
1658 |
* @size: allocation size |
1da177e4c Linux-2.6.12-rc2 |
1659 1660 1661 |
* Allocate enough pages to cover @size from the page level * allocator and map them into contiguous kernel virtual space. * |
c1c8897f8 Spelling fix: "co... |
1662 |
* For tight control over page level allocator and protection flags |
1da177e4c Linux-2.6.12-rc2 |
1663 1664 1665 1666 |
* use __vmalloc() instead. */ void *vmalloc(unsigned long size) { |
00ef2d2f8 mm: use NUMA_NO_NODE |
1667 1668 |
return __vmalloc_node_flags(size, NUMA_NO_NODE, GFP_KERNEL | __GFP_HIGHMEM); |
1da177e4c Linux-2.6.12-rc2 |
1669 |
} |
1da177e4c Linux-2.6.12-rc2 |
1670 |
EXPORT_SYMBOL(vmalloc); |
930fc45a4 [PATCH] vmalloc_node |
1671 |
/** |
e1ca7788d mm: add vzalloc()... |
1672 1673 1674 1675 1676 1677 1678 1679 1680 1681 1682 |
* 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. */ void *vzalloc(unsigned long size) { |
00ef2d2f8 mm: use NUMA_NO_NODE |
1683 |
return __vmalloc_node_flags(size, NUMA_NO_NODE, |
e1ca7788d mm: add vzalloc()... |
1684 1685 1686 1687 1688 |
GFP_KERNEL | __GFP_HIGHMEM | __GFP_ZERO); } EXPORT_SYMBOL(vzalloc); /** |
ead04089b [PATCH] Fix kerne... |
1689 1690 |
* vmalloc_user - allocate zeroed virtually contiguous memory for userspace * @size: allocation size |
833423143 [PATCH] mm: intro... |
1691 |
* |
ead04089b [PATCH] Fix kerne... |
1692 1693 |
* The resulting memory area is zeroed so it can be mapped to userspace * without leaking data. |
833423143 [PATCH] mm: intro... |
1694 1695 1696 1697 1698 |
*/ void *vmalloc_user(unsigned long size) { struct vm_struct *area; void *ret; |
2dca6999e mm, perf_event: M... |
1699 1700 |
ret = __vmalloc_node(size, SHMLBA, GFP_KERNEL | __GFP_HIGHMEM | __GFP_ZERO, |
00ef2d2f8 mm: use NUMA_NO_NODE |
1701 1702 |
PAGE_KERNEL, NUMA_NO_NODE, __builtin_return_address(0)); |
2b4ac44e7 [PATCH] vmalloc: ... |
1703 |
if (ret) { |
db64fe022 mm: rewrite vmap ... |
1704 |
area = find_vm_area(ret); |
2b4ac44e7 [PATCH] vmalloc: ... |
1705 |
area->flags |= VM_USERMAP; |
2b4ac44e7 [PATCH] vmalloc: ... |
1706 |
} |
833423143 [PATCH] mm: intro... |
1707 1708 1709 1710 1711 |
return ret; } EXPORT_SYMBOL(vmalloc_user); /** |
930fc45a4 [PATCH] vmalloc_node |
1712 |
* vmalloc_node - allocate memory on a specific node |
930fc45a4 [PATCH] vmalloc_node |
1713 |
* @size: allocation size |
d44e0780b [PATCH] kernel-do... |
1714 |
* @node: numa node |
930fc45a4 [PATCH] vmalloc_node |
1715 1716 1717 1718 |
* * Allocate enough pages to cover @size from the page level * allocator and map them into contiguous kernel virtual space. * |
c1c8897f8 Spelling fix: "co... |
1719 |
* For tight control over page level allocator and protection flags |
930fc45a4 [PATCH] vmalloc_node |
1720 1721 1722 1723 |
* use __vmalloc() instead. */ void *vmalloc_node(unsigned long size, int node) { |
2dca6999e mm, perf_event: M... |
1724 |
return __vmalloc_node(size, 1, GFP_KERNEL | __GFP_HIGHMEM, PAGE_KERNEL, |
230169693 vmallocinfo: add ... |
1725 |
node, __builtin_return_address(0)); |
930fc45a4 [PATCH] vmalloc_node |
1726 1727 |
} EXPORT_SYMBOL(vmalloc_node); |
e1ca7788d mm: add vzalloc()... |
1728 1729 1730 1731 1732 1733 1734 1735 1736 1737 1738 1739 1740 1741 1742 1743 1744 1745 |
/** * 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. */ void *vzalloc_node(unsigned long size, int node) { return __vmalloc_node_flags(size, node, GFP_KERNEL | __GFP_HIGHMEM | __GFP_ZERO); } EXPORT_SYMBOL(vzalloc_node); |
4dc3b16ba [PATCH] DocBook: ... |
1746 1747 1748 |
#ifndef PAGE_KERNEL_EXEC # define PAGE_KERNEL_EXEC PAGE_KERNEL #endif |
1da177e4c Linux-2.6.12-rc2 |
1749 1750 |
/** * vmalloc_exec - allocate virtually contiguous, executable memory |
1da177e4c Linux-2.6.12-rc2 |
1751 1752 1753 1754 1755 1756 |
* @size: allocation size * * Kernel-internal function to allocate enough pages to cover @size * the page level allocator and map them into contiguous and * executable kernel virtual space. * |
c1c8897f8 Spelling fix: "co... |
1757 |
* For tight control over page level allocator and protection flags |
1da177e4c Linux-2.6.12-rc2 |
1758 1759 |
* use __vmalloc() instead. */ |
1da177e4c Linux-2.6.12-rc2 |
1760 1761 |
void *vmalloc_exec(unsigned long size) { |
2dca6999e mm, perf_event: M... |
1762 |
return __vmalloc_node(size, 1, GFP_KERNEL | __GFP_HIGHMEM, PAGE_KERNEL_EXEC, |
00ef2d2f8 mm: use NUMA_NO_NODE |
1763 |
NUMA_NO_NODE, __builtin_return_address(0)); |
1da177e4c Linux-2.6.12-rc2 |
1764 |
} |
0d08e0d3a [PATCH] x86-64: F... |
1765 |
#if defined(CONFIG_64BIT) && defined(CONFIG_ZONE_DMA32) |
7ac674f52 vmalloc_32 should... |
1766 |
#define GFP_VMALLOC32 GFP_DMA32 | GFP_KERNEL |
0d08e0d3a [PATCH] x86-64: F... |
1767 |
#elif defined(CONFIG_64BIT) && defined(CONFIG_ZONE_DMA) |
7ac674f52 vmalloc_32 should... |
1768 |
#define GFP_VMALLOC32 GFP_DMA | GFP_KERNEL |
0d08e0d3a [PATCH] x86-64: F... |
1769 1770 1771 |
#else #define GFP_VMALLOC32 GFP_KERNEL #endif |
1da177e4c Linux-2.6.12-rc2 |
1772 1773 |
/** * vmalloc_32 - allocate virtually contiguous memory (32bit addressable) |
1da177e4c Linux-2.6.12-rc2 |
1774 1775 1776 1777 1778 1779 1780 |
* @size: allocation size * * Allocate enough 32bit PA addressable pages to cover @size from the * page level allocator and map them into contiguous kernel virtual space. */ void *vmalloc_32(unsigned long size) { |
2dca6999e mm, perf_event: M... |
1781 |
return __vmalloc_node(size, 1, GFP_VMALLOC32, PAGE_KERNEL, |
00ef2d2f8 mm: use NUMA_NO_NODE |
1782 |
NUMA_NO_NODE, __builtin_return_address(0)); |
1da177e4c Linux-2.6.12-rc2 |
1783 |
} |
1da177e4c Linux-2.6.12-rc2 |
1784 |
EXPORT_SYMBOL(vmalloc_32); |
833423143 [PATCH] mm: intro... |
1785 |
/** |
ead04089b [PATCH] Fix kerne... |
1786 |
* vmalloc_32_user - allocate zeroed virtually contiguous 32bit memory |
833423143 [PATCH] mm: intro... |
1787 |
* @size: allocation size |
ead04089b [PATCH] Fix kerne... |
1788 1789 1790 |
* * The resulting memory area is 32bit addressable and zeroed so it can be * mapped to userspace without leaking data. |
833423143 [PATCH] mm: intro... |
1791 1792 1793 1794 1795 |
*/ void *vmalloc_32_user(unsigned long size) { struct vm_struct *area; void *ret; |
2dca6999e mm, perf_event: M... |
1796 |
ret = __vmalloc_node(size, 1, GFP_VMALLOC32 | __GFP_ZERO, PAGE_KERNEL, |
00ef2d2f8 mm: use NUMA_NO_NODE |
1797 |
NUMA_NO_NODE, __builtin_return_address(0)); |
2b4ac44e7 [PATCH] vmalloc: ... |
1798 |
if (ret) { |
db64fe022 mm: rewrite vmap ... |
1799 |
area = find_vm_area(ret); |
2b4ac44e7 [PATCH] vmalloc: ... |
1800 |
area->flags |= VM_USERMAP; |
2b4ac44e7 [PATCH] vmalloc: ... |
1801 |
} |
833423143 [PATCH] mm: intro... |
1802 1803 1804 |
return ret; } EXPORT_SYMBOL(vmalloc_32_user); |
d0107eb07 kcore: fix vread/... |
1805 1806 1807 1808 1809 1810 1811 1812 1813 1814 1815 1816 1817 1818 1819 1820 1821 1822 1823 1824 1825 1826 1827 1828 1829 1830 1831 1832 1833 1834 |
/* * 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; offset = (unsigned long)addr & ~PAGE_MASK; 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... |
1835 |
void *map = kmap_atomic(p); |
d0107eb07 kcore: fix vread/... |
1836 |
memcpy(buf, map + offset, length); |
9b04c5fec mm: remove the se... |
1837 |
kunmap_atomic(map); |
d0107eb07 kcore: fix vread/... |
1838 1839 1840 1841 1842 1843 1844 1845 1846 1847 1848 1849 1850 1851 1852 1853 1854 1855 1856 1857 1858 1859 1860 1861 1862 1863 1864 1865 1866 1867 1868 1869 1870 1871 1872 1873 |
} 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; offset = (unsigned long)addr & ~PAGE_MASK; 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... |
1874 |
void *map = kmap_atomic(p); |
d0107eb07 kcore: fix vread/... |
1875 |
memcpy(map + offset, buf, length); |
9b04c5fec mm: remove the se... |
1876 |
kunmap_atomic(map); |
d0107eb07 kcore: fix vread/... |
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 |
} addr += length; buf += length; copied += length; count -= length; } return copied; } /** * vread() - read vmalloc area in a safe way. * @buf: buffer for reading data * @addr: vm address. * @count: number of bytes to be read. * * Returns # of bytes which addr and buf should be increased. * (same number to @count). Returns 0 if [addr...addr+count) doesn't * includes any intersect with alive vmalloc area. * * 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 |
a8e5202d0 vmalloc: remove K... |
1903 |
* vm_struct area, returns 0. @buf should be kernel's buffer. |
d0107eb07 kcore: fix vread/... |
1904 1905 1906 1907 1908 1909 1910 |
* * 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 * any informaion, as /dev/kmem. * */ |
1da177e4c Linux-2.6.12-rc2 |
1911 1912 |
long vread(char *buf, char *addr, unsigned long count) { |
e81ce85f9 mm, vmalloc: iter... |
1913 1914 |
struct vmap_area *va; struct vm_struct *vm; |
1da177e4c Linux-2.6.12-rc2 |
1915 |
char *vaddr, *buf_start = buf; |
d0107eb07 kcore: fix vread/... |
1916 |
unsigned long buflen = count; |
1da177e4c Linux-2.6.12-rc2 |
1917 1918 1919 1920 1921 |
unsigned long n; /* Don't allow overflow */ if ((unsigned long) addr + count < count) count = -(unsigned long) addr; |
e81ce85f9 mm, vmalloc: iter... |
1922 1923 1924 1925 1926 1927 1928 1929 1930 1931 |
spin_lock(&vmap_area_lock); list_for_each_entry(va, &vmap_area_list, list) { if (!count) break; if (!(va->flags & VM_VM_AREA)) continue; vm = va->vm; vaddr = (char *) vm->addr; |
762216ab4 mm/vmalloc: use w... |
1932 |
if (addr >= vaddr + get_vm_area_size(vm)) |
1da177e4c Linux-2.6.12-rc2 |
1933 1934 1935 1936 1937 1938 1939 1940 1941 |
continue; while (addr < vaddr) { if (count == 0) goto finished; *buf = '\0'; buf++; addr++; count--; } |
762216ab4 mm/vmalloc: use w... |
1942 |
n = vaddr + get_vm_area_size(vm) - addr; |
d0107eb07 kcore: fix vread/... |
1943 1944 |
if (n > count) n = count; |
e81ce85f9 mm, vmalloc: iter... |
1945 |
if (!(vm->flags & VM_IOREMAP)) |
d0107eb07 kcore: fix vread/... |
1946 1947 1948 1949 1950 1951 |
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 |
1952 1953 |
} finished: |
e81ce85f9 mm, vmalloc: iter... |
1954 |
spin_unlock(&vmap_area_lock); |
d0107eb07 kcore: fix vread/... |
1955 1956 1957 1958 1959 1960 1961 1962 |
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 |
1963 |
} |
d0107eb07 kcore: fix vread/... |
1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 |
/** * vwrite() - write vmalloc area in a safe way. * @buf: buffer for source data * @addr: vm address. * @count: number of bytes to be read. * * Returns # of bytes which addr and buf should be incresed. * (same number to @count). * If [addr...addr+count) doesn't includes any intersect with valid * vmalloc area, returns 0. * * 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 |
a8e5202d0 vmalloc: remove K... |
1982 |
* vm_struct area, returns 0. @buf should be kernel's buffer. |
d0107eb07 kcore: fix vread/... |
1983 1984 1985 1986 1987 |
* * 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 * any informaion, as /dev/kmem. |
d0107eb07 kcore: fix vread/... |
1988 |
*/ |
1da177e4c Linux-2.6.12-rc2 |
1989 1990 |
long vwrite(char *buf, char *addr, unsigned long count) { |
e81ce85f9 mm, vmalloc: iter... |
1991 1992 |
struct vmap_area *va; struct vm_struct *vm; |
d0107eb07 kcore: fix vread/... |
1993 1994 1995 |
char *vaddr; unsigned long n, buflen; int copied = 0; |
1da177e4c Linux-2.6.12-rc2 |
1996 1997 1998 1999 |
/* Don't allow overflow */ if ((unsigned long) addr + count < count) count = -(unsigned long) addr; |
d0107eb07 kcore: fix vread/... |
2000 |
buflen = count; |
1da177e4c Linux-2.6.12-rc2 |
2001 |
|
e81ce85f9 mm, vmalloc: iter... |
2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 |
spin_lock(&vmap_area_lock); list_for_each_entry(va, &vmap_area_list, list) { if (!count) break; if (!(va->flags & VM_VM_AREA)) continue; vm = va->vm; vaddr = (char *) vm->addr; |
762216ab4 mm/vmalloc: use w... |
2012 |
if (addr >= vaddr + get_vm_area_size(vm)) |
1da177e4c Linux-2.6.12-rc2 |
2013 2014 2015 2016 2017 2018 2019 2020 |
continue; while (addr < vaddr) { if (count == 0) goto finished; buf++; addr++; count--; } |
762216ab4 mm/vmalloc: use w... |
2021 |
n = vaddr + get_vm_area_size(vm) - addr; |
d0107eb07 kcore: fix vread/... |
2022 2023 |
if (n > count) n = count; |
e81ce85f9 mm, vmalloc: iter... |
2024 |
if (!(vm->flags & VM_IOREMAP)) { |
d0107eb07 kcore: fix vread/... |
2025 2026 2027 2028 2029 2030 |
aligned_vwrite(buf, addr, n); copied++; } buf += n; addr += n; count -= n; |
1da177e4c Linux-2.6.12-rc2 |
2031 2032 |
} finished: |
e81ce85f9 mm, vmalloc: iter... |
2033 |
spin_unlock(&vmap_area_lock); |
d0107eb07 kcore: fix vread/... |
2034 2035 2036 |
if (!copied) return 0; return buflen; |
1da177e4c Linux-2.6.12-rc2 |
2037 |
} |
833423143 [PATCH] mm: intro... |
2038 2039 |
/** |
e69e9d4ae vmalloc: introduc... |
2040 2041 2042 2043 2044 |
* remap_vmalloc_range_partial - map vmalloc pages to userspace * @vma: vma to cover * @uaddr: target user address to start at * @kaddr: virtual address of vmalloc kernel memory * @size: size of map area |
7682486b3 mm: fix various k... |
2045 2046 |
* * Returns: 0 for success, -Exxx on failure |
833423143 [PATCH] mm: intro... |
2047 |
* |
e69e9d4ae vmalloc: introduc... |
2048 2049 2050 2051 |
* 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... |
2052 |
* |
72fd4a35a [PATCH] Numerous ... |
2053 |
* Similar to remap_pfn_range() (see mm/memory.c) |
833423143 [PATCH] mm: intro... |
2054 |
*/ |
e69e9d4ae vmalloc: introduc... |
2055 2056 |
int remap_vmalloc_range_partial(struct vm_area_struct *vma, unsigned long uaddr, void *kaddr, unsigned long size) |
833423143 [PATCH] mm: intro... |
2057 2058 |
{ struct vm_struct *area; |
833423143 [PATCH] mm: intro... |
2059 |
|
e69e9d4ae vmalloc: introduc... |
2060 2061 2062 |
size = PAGE_ALIGN(size); if (!PAGE_ALIGNED(uaddr) || !PAGE_ALIGNED(kaddr)) |
833423143 [PATCH] mm: intro... |
2063 |
return -EINVAL; |
e69e9d4ae vmalloc: introduc... |
2064 |
area = find_vm_area(kaddr); |
833423143 [PATCH] mm: intro... |
2065 |
if (!area) |
db64fe022 mm: rewrite vmap ... |
2066 |
return -EINVAL; |
833423143 [PATCH] mm: intro... |
2067 2068 |
if (!(area->flags & VM_USERMAP)) |
db64fe022 mm: rewrite vmap ... |
2069 |
return -EINVAL; |
833423143 [PATCH] mm: intro... |
2070 |
|
e69e9d4ae vmalloc: introduc... |
2071 |
if (kaddr + size > area->addr + area->size) |
db64fe022 mm: rewrite vmap ... |
2072 |
return -EINVAL; |
833423143 [PATCH] mm: intro... |
2073 |
|
833423143 [PATCH] mm: intro... |
2074 |
do { |
e69e9d4ae vmalloc: introduc... |
2075 |
struct page *page = vmalloc_to_page(kaddr); |
db64fe022 mm: rewrite vmap ... |
2076 |
int ret; |
833423143 [PATCH] mm: intro... |
2077 2078 2079 2080 2081 |
ret = vm_insert_page(vma, uaddr, page); if (ret) return ret; uaddr += PAGE_SIZE; |
e69e9d4ae vmalloc: introduc... |
2082 2083 2084 |
kaddr += PAGE_SIZE; size -= PAGE_SIZE; } while (size > 0); |
833423143 [PATCH] mm: intro... |
2085 |
|
314e51b98 mm: kill vma flag... |
2086 |
vma->vm_flags |= VM_DONTEXPAND | VM_DONTDUMP; |
833423143 [PATCH] mm: intro... |
2087 |
|
db64fe022 mm: rewrite vmap ... |
2088 |
return 0; |
833423143 [PATCH] mm: intro... |
2089 |
} |
e69e9d4ae vmalloc: introduc... |
2090 2091 2092 2093 2094 2095 2096 2097 2098 2099 2100 2101 2102 2103 2104 2105 2106 2107 2108 2109 2110 2111 2112 |
EXPORT_SYMBOL(remap_vmalloc_range_partial); /** * 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 * * Returns: 0 for success, -Exxx on failure * * 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. * * Similar to remap_pfn_range() (see mm/memory.c) */ int remap_vmalloc_range(struct vm_area_struct *vma, void *addr, unsigned long pgoff) { return remap_vmalloc_range_partial(vma, vma->vm_start, addr + (pgoff << PAGE_SHIFT), vma->vm_end - vma->vm_start); } |
833423143 [PATCH] mm: intro... |
2113 |
EXPORT_SYMBOL(remap_vmalloc_range); |
1eeb66a1b move die notifier... |
2114 2115 2116 2117 |
/* * Implement a stub for vmalloc_sync_all() if the architecture chose not to * have one. */ |
3b32123d7 mm: use macros fr... |
2118 |
void __weak vmalloc_sync_all(void) |
1eeb66a1b move die notifier... |
2119 2120 |
{ } |
5f4352fbf Allocate and free... |
2121 |
|
2f569afd9 CONFIG_HIGHPTE vs... |
2122 |
static int f(pte_t *pte, pgtable_t table, unsigned long addr, void *data) |
5f4352fbf Allocate and free... |
2123 |
{ |
cd12909cb xen: map foreign ... |
2124 2125 2126 2127 2128 2129 |
pte_t ***p = data; if (p) { *(*p) = pte; (*p)++; } |
5f4352fbf Allocate and free... |
2130 2131 2132 2133 2134 2135 |
return 0; } /** * alloc_vm_area - allocate a range of kernel address space * @size: size of the area |
cd12909cb xen: map foreign ... |
2136 |
* @ptes: returns the PTEs for the address space |
7682486b3 mm: fix various k... |
2137 2138 |
* * Returns: NULL on failure, vm_struct on success |
5f4352fbf Allocate and free... |
2139 2140 2141 |
* * This function reserves a range of kernel address space, and * allocates pagetables to map that range. No actual mappings |
cd12909cb xen: map foreign ... |
2142 2143 2144 2145 |
* are created. * * If @ptes is non-NULL, pointers to the PTEs (in init_mm) * allocated for the VM area are returned. |
5f4352fbf Allocate and free... |
2146 |
*/ |
cd12909cb xen: map foreign ... |
2147 |
struct vm_struct *alloc_vm_area(size_t size, pte_t **ptes) |
5f4352fbf Allocate and free... |
2148 2149 |
{ struct vm_struct *area; |
230169693 vmallocinfo: add ... |
2150 2151 |
area = get_vm_area_caller(size, VM_IOREMAP, __builtin_return_address(0)); |
5f4352fbf Allocate and free... |
2152 2153 2154 2155 2156 2157 2158 2159 |
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 ... |
2160 |
size, f, ptes ? &ptes : NULL)) { |
5f4352fbf Allocate and free... |
2161 2162 2163 |
free_vm_area(area); return NULL; } |
5f4352fbf Allocate and free... |
2164 2165 2166 2167 2168 2169 2170 2171 2172 2173 2174 2175 |
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... |
2176 |
|
4f8b02b4e vmalloc: pcpu_get... |
2177 |
#ifdef CONFIG_SMP |
ca23e405e vmalloc: implemen... |
2178 2179 2180 2181 2182 2183 2184 2185 2186 2187 2188 2189 2190 2191 2192 2193 2194 2195 2196 2197 2198 2199 2200 2201 2202 2203 2204 2205 2206 2207 2208 2209 2210 2211 2212 2213 2214 2215 2216 2217 2218 2219 2220 2221 2222 2223 2224 2225 2226 2227 2228 2229 2230 2231 2232 2233 2234 2235 2236 2237 2238 2239 2240 2241 2242 2243 2244 2245 2246 2247 2248 2249 2250 2251 2252 2253 2254 2255 2256 2257 2258 2259 2260 2261 2262 2263 2264 2265 2266 |
static struct vmap_area *node_to_va(struct rb_node *n) { return n ? rb_entry(n, struct vmap_area, rb_node) : NULL; } /** * pvm_find_next_prev - find the next and prev vmap_area surrounding @end * @end: target address * @pnext: out arg for the next vmap_area * @pprev: out arg for the previous vmap_area * * Returns: %true if either or both of next and prev are found, * %false if no vmap_area exists * * Find vmap_areas end addresses of which enclose @end. ie. if not * NULL, *pnext->va_end > @end and *pprev->va_end <= @end. */ static bool pvm_find_next_prev(unsigned long end, struct vmap_area **pnext, struct vmap_area **pprev) { struct rb_node *n = vmap_area_root.rb_node; struct vmap_area *va = NULL; while (n) { va = rb_entry(n, struct vmap_area, rb_node); if (end < va->va_end) n = n->rb_left; else if (end > va->va_end) n = n->rb_right; else break; } if (!va) return false; if (va->va_end > end) { *pnext = va; *pprev = node_to_va(rb_prev(&(*pnext)->rb_node)); } else { *pprev = va; *pnext = node_to_va(rb_next(&(*pprev)->rb_node)); } return true; } /** * pvm_determine_end - find the highest aligned address between two vmap_areas * @pnext: in/out arg for the next vmap_area * @pprev: in/out arg for the previous vmap_area * @align: alignment * * Returns: determined end address * * Find the highest aligned address between *@pnext and *@pprev below * VMALLOC_END. *@pnext and *@pprev are adjusted so that the aligned * down address is between the end addresses of the two vmap_areas. * * Please note that the address returned by this function may fall * inside *@pnext vmap_area. The caller is responsible for checking * that. */ static unsigned long pvm_determine_end(struct vmap_area **pnext, struct vmap_area **pprev, unsigned long align) { const unsigned long vmalloc_end = VMALLOC_END & ~(align - 1); unsigned long addr; if (*pnext) addr = min((*pnext)->va_start & ~(align - 1), vmalloc_end); else addr = vmalloc_end; while (*pprev && (*pprev)->va_end > addr) { *pnext = *pprev; *pprev = node_to_va(rb_prev(&(*pnext)->rb_node)); } return addr; } /** * 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... |
2267 2268 2269 2270 2271 2272 |
* * 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... |
2273 2274 2275 2276 |
* 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... |
2277 2278 2279 2280 2281 2282 2283 2284 2285 2286 |
* * Despite its complicated look, this allocator is rather simple. It * does everything top-down and scans areas from the end looking for * matching slot. While scanning, if any of the areas overlaps with * existing vmap_area, the base address is pulled down to fit the * area. Scanning is repeated till all the areas fit and then all * necessary data structres are inserted and the result is returned. */ struct vm_struct **pcpu_get_vm_areas(const unsigned long *offsets, const size_t *sizes, int nr_vms, |
ec3f64fc9 mm: remove gfp ma... |
2287 |
size_t align) |
ca23e405e vmalloc: implemen... |
2288 2289 2290 2291 2292 2293 2294 2295 |
{ const unsigned long vmalloc_start = ALIGN(VMALLOC_START, align); const unsigned long vmalloc_end = VMALLOC_END & ~(align - 1); struct vmap_area **vas, *prev, *next; struct vm_struct **vms; int area, area2, last_area, term_area; unsigned long base, start, end, last_end; bool purged = false; |
ca23e405e vmalloc: implemen... |
2296 2297 2298 2299 2300 2301 2302 2303 2304 2305 2306 2307 2308 2309 2310 2311 2312 2313 2314 2315 2316 2317 2318 2319 2320 2321 2322 2323 2324 2325 2326 |
/* verify parameters and allocate data structures */ BUG_ON(align & ~PAGE_MASK || !is_power_of_2(align)); 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; for (area2 = 0; area2 < nr_vms; area2++) { unsigned long start2 = offsets[area2]; unsigned long end2 = start2 + sizes[area2]; if (area2 == area) continue; BUG_ON(start2 >= start && start2 < end); BUG_ON(end2 <= end && end2 > start); } } last_end = offsets[last_area] + sizes[last_area]; if (vmalloc_end - vmalloc_start < last_end) { WARN_ON(true); return NULL; } |
4d67d8605 mm: use kcalloc()... |
2327 2328 |
vms = kcalloc(nr_vms, sizeof(vms[0]), GFP_KERNEL); vas = kcalloc(nr_vms, sizeof(vas[0]), GFP_KERNEL); |
ca23e405e vmalloc: implemen... |
2329 |
if (!vas || !vms) |
f1db7afd9 mm/vmalloc.c: eli... |
2330 |
goto err_free2; |
ca23e405e vmalloc: implemen... |
2331 2332 |
for (area = 0; area < nr_vms; area++) { |
ec3f64fc9 mm: remove gfp ma... |
2333 2334 |
vas[area] = kzalloc(sizeof(struct vmap_area), GFP_KERNEL); vms[area] = kzalloc(sizeof(struct vm_struct), GFP_KERNEL); |
ca23e405e vmalloc: implemen... |
2335 2336 2337 2338 2339 2340 2341 2342 2343 2344 2345 2346 2347 2348 2349 2350 2351 2352 2353 2354 2355 2356 2357 2358 2359 2360 2361 2362 2363 2364 2365 2366 2367 2368 2369 2370 2371 2372 2373 2374 2375 2376 2377 2378 2379 2380 2381 2382 2383 2384 2385 2386 2387 2388 2389 2390 2391 2392 2393 2394 2395 2396 2397 2398 2399 2400 2401 2402 2403 2404 2405 2406 2407 2408 2409 2410 2411 2412 2413 2414 2415 2416 2417 2418 2419 |
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]; if (!pvm_find_next_prev(vmap_area_pcpu_hole, &next, &prev)) { base = vmalloc_end - last_end; goto found; } base = pvm_determine_end(&next, &prev, align) - end; while (true) { BUG_ON(next && next->va_end <= base + end); BUG_ON(prev && prev->va_end > base + end); /* * base might have underflowed, add last_end before * comparing. */ if (base + last_end < vmalloc_start + last_end) { spin_unlock(&vmap_area_lock); if (!purged) { purge_vmap_area_lazy(); purged = true; goto retry; } goto err_free; } /* * If next overlaps, move base downwards so that it's * right below next and then recheck. */ if (next && next->va_start < base + end) { base = pvm_determine_end(&next, &prev, align) - end; term_area = area; continue; } /* * If prev overlaps, shift down next and prev and move * base so that it's right below new next and then * recheck. */ if (prev && prev->va_end > base + start) { next = prev; prev = node_to_va(rb_prev(&next->rb_node)); base = pvm_determine_end(&next, &prev, align) - end; 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; start = offsets[area]; end = start + sizes[area]; pvm_find_next_prev(base + end, &next, &prev); } found: /* we've found a fitting base, insert all va's */ for (area = 0; area < nr_vms; area++) { struct vmap_area *va = vas[area]; va->va_start = base + offsets[area]; va->va_end = va->va_start + sizes[area]; __insert_vmap_area(va); } vmap_area_pcpu_hole = base + offsets[last_area]; spin_unlock(&vmap_area_lock); /* insert all vm's */ for (area = 0; area < nr_vms; area++) |
3645cb4a4 mm, vmalloc: call... |
2420 2421 |
setup_vmalloc_vm(vms[area], vas[area], VM_ALLOC, pcpu_get_vm_areas); |
ca23e405e vmalloc: implemen... |
2422 2423 2424 2425 2426 2427 |
kfree(vas); return vms; err_free: for (area = 0; area < nr_vms; area++) { |
f1db7afd9 mm/vmalloc.c: eli... |
2428 2429 |
kfree(vas[area]); kfree(vms[area]); |
ca23e405e vmalloc: implemen... |
2430 |
} |
f1db7afd9 mm/vmalloc.c: eli... |
2431 |
err_free2: |
ca23e405e vmalloc: implemen... |
2432 2433 2434 2435 2436 2437 2438 2439 2440 2441 2442 2443 2444 2445 2446 2447 2448 2449 2450 2451 |
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... |
2452 |
#endif /* CONFIG_SMP */ |
a10aa5798 vmalloc: show vma... |
2453 2454 2455 |
#ifdef CONFIG_PROC_FS static void *s_start(struct seq_file *m, loff_t *pos) |
d4033afdf mm, vmalloc: iter... |
2456 |
__acquires(&vmap_area_lock) |
a10aa5798 vmalloc: show vma... |
2457 2458 |
{ loff_t n = *pos; |
d4033afdf mm, vmalloc: iter... |
2459 |
struct vmap_area *va; |
a10aa5798 vmalloc: show vma... |
2460 |
|
d4033afdf mm, vmalloc: iter... |
2461 2462 2463 |
spin_lock(&vmap_area_lock); va = list_entry((&vmap_area_list)->next, typeof(*va), list); while (n > 0 && &va->list != &vmap_area_list) { |
a10aa5798 vmalloc: show vma... |
2464 |
n--; |
d4033afdf mm, vmalloc: iter... |
2465 |
va = list_entry(va->list.next, typeof(*va), list); |
a10aa5798 vmalloc: show vma... |
2466 |
} |
d4033afdf mm, vmalloc: iter... |
2467 2468 |
if (!n && &va->list != &vmap_area_list) return va; |
a10aa5798 vmalloc: show vma... |
2469 2470 2471 2472 2473 2474 2475 |
return NULL; } static void *s_next(struct seq_file *m, void *p, loff_t *pos) { |
d4033afdf mm, vmalloc: iter... |
2476 |
struct vmap_area *va = p, *next; |
a10aa5798 vmalloc: show vma... |
2477 2478 |
++*pos; |
d4033afdf mm, vmalloc: iter... |
2479 2480 2481 2482 2483 |
next = list_entry(va->list.next, typeof(*va), list); if (&next->list != &vmap_area_list) return next; return NULL; |
a10aa5798 vmalloc: show vma... |
2484 2485 2486 |
} static void s_stop(struct seq_file *m, void *p) |
d4033afdf mm, vmalloc: iter... |
2487 |
__releases(&vmap_area_lock) |
a10aa5798 vmalloc: show vma... |
2488 |
{ |
d4033afdf mm, vmalloc: iter... |
2489 |
spin_unlock(&vmap_area_lock); |
a10aa5798 vmalloc: show vma... |
2490 |
} |
a47a126ad vmallocinfo: add ... |
2491 2492 |
static void show_numa_info(struct seq_file *m, struct vm_struct *v) { |
e5adfffc8 mm: use IS_ENABLE... |
2493 |
if (IS_ENABLED(CONFIG_NUMA)) { |
a47a126ad vmallocinfo: add ... |
2494 2495 2496 2497 |
unsigned int nr, *counters = m->private; if (!counters) return; |
af12346cd mm/vmalloc: rever... |
2498 2499 |
if (v->flags & VM_UNINITIALIZED) return; |
7e5b528b4 mm/vmalloc.c: fix... |
2500 2501 |
/* Pair with smp_wmb() in clear_vm_uninitialized_flag() */ smp_rmb(); |
af12346cd mm/vmalloc: rever... |
2502 |
|
a47a126ad vmallocinfo: add ... |
2503 2504 2505 2506 2507 2508 2509 2510 2511 2512 |
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]); } } |
a10aa5798 vmalloc: show vma... |
2513 2514 |
static int s_show(struct seq_file *m, void *p) { |
d4033afdf mm, vmalloc: iter... |
2515 2516 |
struct vmap_area *va = p; struct vm_struct *v; |
c2ce8c142 mm/vmalloc: fix s... |
2517 2518 2519 2520 2521 |
/* * s_show can encounter race with remove_vm_area, !VM_VM_AREA on * behalf of vmap area is being tear down or vm_map_ram allocation. */ if (!(va->flags & VM_VM_AREA)) |
d4033afdf mm, vmalloc: iter... |
2522 |
return 0; |
d4033afdf mm, vmalloc: iter... |
2523 2524 |
v = va->vm; |
a10aa5798 vmalloc: show vma... |
2525 |
|
45ec16908 mm: use %pK for /... |
2526 |
seq_printf(m, "0x%pK-0x%pK %7ld", |
a10aa5798 vmalloc: show vma... |
2527 |
v->addr, v->addr + v->size, v->size); |
62c70bce8 mm: convert sprin... |
2528 2529 |
if (v->caller) seq_printf(m, " %pS", v->caller); |
230169693 vmallocinfo: add ... |
2530 |
|
a10aa5798 vmalloc: show vma... |
2531 2532 2533 2534 |
if (v->nr_pages) seq_printf(m, " pages=%d", v->nr_pages); if (v->phys_addr) |
ffa71f33a x86, ioremap: Fix... |
2535 |
seq_printf(m, " phys=%llx", (unsigned long long)v->phys_addr); |
a10aa5798 vmalloc: show vma... |
2536 2537 |
if (v->flags & VM_IOREMAP) |
f4527c908 mm/vmalloc.c: rep... |
2538 |
seq_puts(m, " ioremap"); |
a10aa5798 vmalloc: show vma... |
2539 2540 |
if (v->flags & VM_ALLOC) |
f4527c908 mm/vmalloc.c: rep... |
2541 |
seq_puts(m, " vmalloc"); |
a10aa5798 vmalloc: show vma... |
2542 2543 |
if (v->flags & VM_MAP) |
f4527c908 mm/vmalloc.c: rep... |
2544 |
seq_puts(m, " vmap"); |
a10aa5798 vmalloc: show vma... |
2545 2546 |
if (v->flags & VM_USERMAP) |
f4527c908 mm/vmalloc.c: rep... |
2547 |
seq_puts(m, " user"); |
a10aa5798 vmalloc: show vma... |
2548 2549 |
if (v->flags & VM_VPAGES) |
f4527c908 mm/vmalloc.c: rep... |
2550 |
seq_puts(m, " vpages"); |
a10aa5798 vmalloc: show vma... |
2551 |
|
a47a126ad vmallocinfo: add ... |
2552 |
show_numa_info(m, v); |
a10aa5798 vmalloc: show vma... |
2553 2554 2555 2556 |
seq_putc(m, ' '); return 0; } |
5f6a6a9c4 proc: move /proc/... |
2557 |
static const struct seq_operations vmalloc_op = { |
a10aa5798 vmalloc: show vma... |
2558 2559 2560 2561 2562 |
.start = s_start, .next = s_next, .stop = s_stop, .show = s_show, }; |
5f6a6a9c4 proc: move /proc/... |
2563 2564 2565 |
static int vmalloc_open(struct inode *inode, struct file *file) { |
703394c10 mm/vmalloc.c: use... |
2566 2567 2568 2569 2570 |
if (IS_ENABLED(CONFIG_NUMA)) return seq_open_private(file, &vmalloc_op, nr_node_ids * sizeof(unsigned int)); else return seq_open(file, &vmalloc_op); |
5f6a6a9c4 proc: move /proc/... |
2571 2572 2573 2574 2575 2576 2577 2578 2579 2580 2581 2582 2583 2584 2585 |
} static const struct file_operations proc_vmalloc_operations = { .open = vmalloc_open, .read = seq_read, .llseek = seq_lseek, .release = seq_release_private, }; static int __init proc_vmalloc_init(void) { proc_create("vmallocinfo", S_IRUSR, NULL, &proc_vmalloc_operations); return 0; } module_init(proc_vmalloc_init); |
db3808c1b mm, vmalloc: move... |
2586 2587 2588 |
void get_vmalloc_info(struct vmalloc_info *vmi) { |
f98782ddd mm, vmalloc: iter... |
2589 |
struct vmap_area *va; |
db3808c1b mm, vmalloc: move... |
2590 2591 2592 2593 |
unsigned long free_area_size; unsigned long prev_end; vmi->used = 0; |
f98782ddd mm, vmalloc: iter... |
2594 |
vmi->largest_chunk = 0; |
db3808c1b mm, vmalloc: move... |
2595 |
|
f98782ddd mm, vmalloc: iter... |
2596 |
prev_end = VMALLOC_START; |
db3808c1b mm, vmalloc: move... |
2597 |
|
474750aba vmalloc: use rcu ... |
2598 |
rcu_read_lock(); |
db3808c1b mm, vmalloc: move... |
2599 |
|
f98782ddd mm, vmalloc: iter... |
2600 2601 2602 2603 |
if (list_empty(&vmap_area_list)) { vmi->largest_chunk = VMALLOC_TOTAL; goto out; } |
db3808c1b mm, vmalloc: move... |
2604 |
|
474750aba vmalloc: use rcu ... |
2605 |
list_for_each_entry_rcu(va, &vmap_area_list, list) { |
f98782ddd mm, vmalloc: iter... |
2606 |
unsigned long addr = va->va_start; |
db3808c1b mm, vmalloc: move... |
2607 |
|
f98782ddd mm, vmalloc: iter... |
2608 2609 2610 2611 2612 2613 2614 |
/* * Some archs keep another range for modules in vmalloc space */ if (addr < VMALLOC_START) continue; if (addr >= VMALLOC_END) break; |
db3808c1b mm, vmalloc: move... |
2615 |
|
f98782ddd mm, vmalloc: iter... |
2616 2617 |
if (va->flags & (VM_LAZY_FREE | VM_LAZY_FREEING)) continue; |
db3808c1b mm, vmalloc: move... |
2618 |
|
f98782ddd mm, vmalloc: iter... |
2619 |
vmi->used += (va->va_end - va->va_start); |
db3808c1b mm, vmalloc: move... |
2620 |
|
f98782ddd mm, vmalloc: iter... |
2621 2622 2623 |
free_area_size = addr - prev_end; if (vmi->largest_chunk < free_area_size) vmi->largest_chunk = free_area_size; |
db3808c1b mm, vmalloc: move... |
2624 |
|
f98782ddd mm, vmalloc: iter... |
2625 |
prev_end = va->va_end; |
db3808c1b mm, vmalloc: move... |
2626 |
} |
f98782ddd mm, vmalloc: iter... |
2627 2628 2629 2630 2631 |
if (VMALLOC_END - prev_end > vmi->largest_chunk) vmi->largest_chunk = VMALLOC_END - prev_end; out: |
474750aba vmalloc: use rcu ... |
2632 |
rcu_read_unlock(); |
db3808c1b mm, vmalloc: move... |
2633 |
} |
a10aa5798 vmalloc: show vma... |
2634 |
#endif |