Blame view
mm/vmalloc.c
100 KB
457c89965 treewide: Add SPD... |
1 |
// SPDX-License-Identifier: GPL-2.0-only |
1da177e4c Linux-2.6.12-rc2 |
2 |
/* |
1da177e4c Linux-2.6.12-rc2 |
3 4 5 6 |
* 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 |
7 |
* Numa awareness, Christoph Lameter, SGI, June 2005 |
d758ffe6b mm/vmalloc: updat... |
8 |
* Improving global KVA allocator, Uladzislau Rezki, Sony, May 2019 |
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> |
c3edc4010 sched/headers: Mo... |
14 |
#include <linux/sched/signal.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> |
868b104d7 mm/vmalloc: Add f... |
20 |
#include <linux/set_memory.h> |
3ac7fe5a4 infrastructure to... |
21 |
#include <linux/debugobjects.h> |
230169693 vmallocinfo: add ... |
22 |
#include <linux/kallsyms.h> |
db64fe022 mm: rewrite vmap ... |
23 |
#include <linux/list.h> |
4da56b99d mm/vmap: Add a no... |
24 |
#include <linux/notifier.h> |
db64fe022 mm: rewrite vmap ... |
25 |
#include <linux/rbtree.h> |
0f14599c6 vmalloc: convert ... |
26 |
#include <linux/xarray.h> |
5da96bdd9 mm/vmalloc: inclu... |
27 |
#include <linux/io.h> |
db64fe022 mm: rewrite vmap ... |
28 |
#include <linux/rcupdate.h> |
f0aa66179 vmalloc: implemen... |
29 |
#include <linux/pfn.h> |
89219d37a kmemleak: Add the... |
30 |
#include <linux/kmemleak.h> |
60063497a atomic: use <linu... |
31 |
#include <linux/atomic.h> |
3b32123d7 mm: use macros fr... |
32 |
#include <linux/compiler.h> |
32fcfd407 make vfree() safe... |
33 |
#include <linux/llist.h> |
0f616be12 mm: change __get_... |
34 |
#include <linux/bitops.h> |
68ad4a330 mm/vmalloc.c: kee... |
35 |
#include <linux/rbtree_augmented.h> |
bdebd6a28 vmalloc: fix rema... |
36 |
#include <linux/overflow.h> |
c0eb315ad mm/vmalloc: fix H... |
37 |
#include <linux/pgtable.h> |
7c0f6ba68 Replace <asm/uacc... |
38 |
#include <linux/uaccess.h> |
f7ee1f13d mm/vmalloc: enabl... |
39 |
#include <linux/hugetlb.h> |
1da177e4c Linux-2.6.12-rc2 |
40 |
#include <asm/tlbflush.h> |
2dca6999e mm, perf_event: M... |
41 |
#include <asm/shmparam.h> |
1da177e4c Linux-2.6.12-rc2 |
42 |
|
dd56b0464 mm: page_alloc: h... |
43 |
#include "internal.h" |
2a681cfa5 mm: move p?d_allo... |
44 |
#include "pgalloc-track.h" |
dd56b0464 mm: page_alloc: h... |
45 |
|
82a70ce04 mm: move ioremap_... |
46 47 48 49 50 51 52 53 54 55 56 57 |
#ifdef CONFIG_HAVE_ARCH_HUGE_VMAP static unsigned int __ro_after_init ioremap_max_page_shift = BITS_PER_LONG - 1; static int __init set_nohugeiomap(char *str) { ioremap_max_page_shift = PAGE_SHIFT; return 0; } early_param("nohugeiomap", set_nohugeiomap); #else /* CONFIG_HAVE_ARCH_HUGE_VMAP */ static const unsigned int ioremap_max_page_shift = PAGE_SHIFT; #endif /* CONFIG_HAVE_ARCH_HUGE_VMAP */ |
121e6f325 mm/vmalloc: hugep... |
58 59 60 61 62 63 64 65 66 67 68 69 |
#ifdef CONFIG_HAVE_ARCH_HUGE_VMALLOC static bool __ro_after_init vmap_allow_huge = true; static int __init set_nohugevmalloc(char *str) { vmap_allow_huge = false; return 0; } early_param("nohugevmalloc", set_nohugevmalloc); #else /* CONFIG_HAVE_ARCH_HUGE_VMALLOC */ static const bool vmap_allow_huge = false; #endif /* CONFIG_HAVE_ARCH_HUGE_VMALLOC */ |
186525bd6 mm, x86/mm: Untan... |
70 71 72 73 74 75 76 |
bool is_vmalloc_addr(const void *x) { unsigned long addr = (unsigned long)x; return addr >= VMALLOC_START && addr < VMALLOC_END; } EXPORT_SYMBOL(is_vmalloc_addr); |
32fcfd407 make vfree() safe... |
77 78 79 80 81 82 83 84 85 86 87 |
struct vfree_deferred { struct llist_head list; struct work_struct wq; }; static DEFINE_PER_CPU(struct vfree_deferred, vfree_deferred); static void __vunmap(const void *, int); static void free_work(struct work_struct *w) { struct vfree_deferred *p = container_of(w, struct vfree_deferred, wq); |
894e58c14 mm/vmalloc.c: don... |
88 89 90 91 |
struct llist_node *t, *llnode; llist_for_each_safe(llnode, t, llist_del_all(&p->list)) __vunmap((void *)llnode, 1); |
32fcfd407 make vfree() safe... |
92 |
} |
db64fe022 mm: rewrite vmap ... |
93 |
/*** Page table manipulation functions ***/ |
5e9e3d777 mm: move vmap_ran... |
94 95 |
static int vmap_pte_range(pmd_t *pmd, unsigned long addr, unsigned long end, phys_addr_t phys_addr, pgprot_t prot, |
f7ee1f13d mm/vmalloc: enabl... |
96 |
unsigned int max_page_shift, pgtbl_mod_mask *mask) |
5e9e3d777 mm: move vmap_ran... |
97 98 99 |
{ pte_t *pte; u64 pfn; |
f7ee1f13d mm/vmalloc: enabl... |
100 |
unsigned long size = PAGE_SIZE; |
5e9e3d777 mm: move vmap_ran... |
101 102 103 104 105 106 107 |
pfn = phys_addr >> PAGE_SHIFT; pte = pte_alloc_kernel_track(pmd, addr, mask); if (!pte) return -ENOMEM; do { BUG_ON(!pte_none(*pte)); |
f7ee1f13d mm/vmalloc: enabl... |
108 109 110 111 112 113 114 115 116 117 118 119 120 |
#ifdef CONFIG_HUGETLB_PAGE size = arch_vmap_pte_range_map_size(addr, end, pfn, max_page_shift); if (size != PAGE_SIZE) { pte_t entry = pfn_pte(pfn, prot); entry = pte_mkhuge(entry); entry = arch_make_huge_pte(entry, ilog2(size), 0); set_huge_pte_at(&init_mm, addr, pte, entry); pfn += PFN_DOWN(size); continue; } #endif |
5e9e3d777 mm: move vmap_ran... |
121 122 |
set_pte_at(&init_mm, addr, pte, pfn_pte(pfn, prot)); pfn++; |
f7ee1f13d mm/vmalloc: enabl... |
123 |
} while (pte += PFN_DOWN(size), addr += size, addr != end); |
5e9e3d777 mm: move vmap_ran... |
124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 |
*mask |= PGTBL_PTE_MODIFIED; return 0; } static int vmap_try_huge_pmd(pmd_t *pmd, unsigned long addr, unsigned long end, phys_addr_t phys_addr, pgprot_t prot, unsigned int max_page_shift) { if (max_page_shift < PMD_SHIFT) return 0; if (!arch_vmap_pmd_supported(prot)) return 0; if ((end - addr) != PMD_SIZE) return 0; if (!IS_ALIGNED(addr, PMD_SIZE)) return 0; if (!IS_ALIGNED(phys_addr, PMD_SIZE)) return 0; if (pmd_present(*pmd) && !pmd_free_pte_page(pmd, addr)) return 0; return pmd_set_huge(pmd, phys_addr, prot); } static int vmap_pmd_range(pud_t *pud, unsigned long addr, unsigned long end, phys_addr_t phys_addr, pgprot_t prot, unsigned int max_page_shift, pgtbl_mod_mask *mask) { pmd_t *pmd; unsigned long next; pmd = pmd_alloc_track(&init_mm, pud, addr, mask); if (!pmd) return -ENOMEM; do { next = pmd_addr_end(addr, end); if (vmap_try_huge_pmd(pmd, addr, next, phys_addr, prot, max_page_shift)) { *mask |= PGTBL_PMD_MODIFIED; continue; } |
f7ee1f13d mm/vmalloc: enabl... |
171 |
if (vmap_pte_range(pmd, addr, next, phys_addr, prot, max_page_shift, mask)) |
5e9e3d777 mm: move vmap_ran... |
172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 |
return -ENOMEM; } while (pmd++, phys_addr += (next - addr), addr = next, addr != end); return 0; } static int vmap_try_huge_pud(pud_t *pud, unsigned long addr, unsigned long end, phys_addr_t phys_addr, pgprot_t prot, unsigned int max_page_shift) { if (max_page_shift < PUD_SHIFT) return 0; if (!arch_vmap_pud_supported(prot)) return 0; if ((end - addr) != PUD_SIZE) return 0; if (!IS_ALIGNED(addr, PUD_SIZE)) return 0; if (!IS_ALIGNED(phys_addr, PUD_SIZE)) return 0; if (pud_present(*pud) && !pud_free_pmd_page(pud, addr)) return 0; return pud_set_huge(pud, phys_addr, prot); } static int vmap_pud_range(p4d_t *p4d, unsigned long addr, unsigned long end, phys_addr_t phys_addr, pgprot_t prot, unsigned int max_page_shift, pgtbl_mod_mask *mask) { pud_t *pud; unsigned long next; pud = pud_alloc_track(&init_mm, p4d, addr, mask); if (!pud) return -ENOMEM; do { next = pud_addr_end(addr, end); if (vmap_try_huge_pud(pud, addr, next, phys_addr, prot, max_page_shift)) { *mask |= PGTBL_PUD_MODIFIED; continue; } if (vmap_pmd_range(pud, addr, next, phys_addr, prot, max_page_shift, mask)) return -ENOMEM; } while (pud++, phys_addr += (next - addr), addr = next, addr != end); return 0; } static int vmap_try_huge_p4d(p4d_t *p4d, unsigned long addr, unsigned long end, phys_addr_t phys_addr, pgprot_t prot, unsigned int max_page_shift) { if (max_page_shift < P4D_SHIFT) return 0; if (!arch_vmap_p4d_supported(prot)) return 0; if ((end - addr) != P4D_SIZE) return 0; if (!IS_ALIGNED(addr, P4D_SIZE)) return 0; if (!IS_ALIGNED(phys_addr, P4D_SIZE)) return 0; if (p4d_present(*p4d) && !p4d_free_pud_page(p4d, addr)) return 0; return p4d_set_huge(p4d, phys_addr, prot); } static int vmap_p4d_range(pgd_t *pgd, unsigned long addr, unsigned long end, phys_addr_t phys_addr, pgprot_t prot, unsigned int max_page_shift, pgtbl_mod_mask *mask) { p4d_t *p4d; unsigned long next; p4d = p4d_alloc_track(&init_mm, pgd, addr, mask); if (!p4d) return -ENOMEM; do { next = p4d_addr_end(addr, end); if (vmap_try_huge_p4d(p4d, addr, next, phys_addr, prot, max_page_shift)) { *mask |= PGTBL_P4D_MODIFIED; continue; } if (vmap_pud_range(p4d, addr, next, phys_addr, prot, max_page_shift, mask)) return -ENOMEM; } while (p4d++, phys_addr += (next - addr), addr = next, addr != end); return 0; } |
5d87510de mm/vmalloc: add v... |
278 |
static int vmap_range_noflush(unsigned long addr, unsigned long end, |
5e9e3d777 mm: move vmap_ran... |
279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 |
phys_addr_t phys_addr, pgprot_t prot, unsigned int max_page_shift) { pgd_t *pgd; unsigned long start; unsigned long next; int err; pgtbl_mod_mask mask = 0; might_sleep(); BUG_ON(addr >= end); start = addr; pgd = pgd_offset_k(addr); do { next = pgd_addr_end(addr, end); err = vmap_p4d_range(pgd, addr, next, phys_addr, prot, max_page_shift, &mask); if (err) break; } while (pgd++, phys_addr += (next - addr), addr = next, addr != end); |
5e9e3d777 mm: move vmap_ran... |
300 301 302 303 304 |
if (mask & ARCH_PAGE_TABLE_SYNC_MASK) arch_sync_kernel_mappings(start, end); return err; } |
b221385bc [PATCH] mm/: make... |
305 |
|
82a70ce04 mm: move ioremap_... |
306 307 |
int ioremap_page_range(unsigned long addr, unsigned long end, phys_addr_t phys_addr, pgprot_t prot) |
5d87510de mm/vmalloc: add v... |
308 309 |
{ int err; |
319e703c7 LF-4658 Revert "m... |
310 |
err = vmap_range_noflush(addr, end, phys_addr, prot, |
82a70ce04 mm: move ioremap_... |
311 |
ioremap_max_page_shift); |
5d87510de mm/vmalloc: add v... |
312 |
flush_cache_vmap(addr, end); |
5d87510de mm/vmalloc: add v... |
313 314 |
return err; } |
c34c51708 mm: Re-export ior... |
315 |
EXPORT_SYMBOL_GPL(ioremap_page_range); |
5d87510de mm/vmalloc: add v... |
316 |
|
2ba3e6947 mm/vmalloc: track... |
317 318 |
static void vunmap_pte_range(pmd_t *pmd, unsigned long addr, unsigned long end, pgtbl_mod_mask *mask) |
1da177e4c Linux-2.6.12-rc2 |
319 320 321 322 323 324 325 326 |
{ 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); |
2ba3e6947 mm/vmalloc: track... |
327 |
*mask |= PGTBL_PTE_MODIFIED; |
1da177e4c Linux-2.6.12-rc2 |
328 |
} |
2ba3e6947 mm/vmalloc: track... |
329 330 |
static void vunmap_pmd_range(pud_t *pud, unsigned long addr, unsigned long end, pgtbl_mod_mask *mask) |
1da177e4c Linux-2.6.12-rc2 |
331 332 333 |
{ pmd_t *pmd; unsigned long next; |
2ba3e6947 mm/vmalloc: track... |
334 |
int cleared; |
1da177e4c Linux-2.6.12-rc2 |
335 336 337 338 |
pmd = pmd_offset(pud, addr); do { next = pmd_addr_end(addr, end); |
2ba3e6947 mm/vmalloc: track... |
339 340 341 342 343 344 |
cleared = pmd_clear_huge(pmd); if (cleared || pmd_bad(*pmd)) *mask |= PGTBL_PMD_MODIFIED; if (cleared) |
b9820d8f3 mm: change vunmap... |
345 |
continue; |
1da177e4c Linux-2.6.12-rc2 |
346 347 |
if (pmd_none_or_clear_bad(pmd)) continue; |
2ba3e6947 mm/vmalloc: track... |
348 |
vunmap_pte_range(pmd, addr, next, mask); |
e47110e90 mm/vunmap: add co... |
349 350 |
cond_resched(); |
1da177e4c Linux-2.6.12-rc2 |
351 352 |
} while (pmd++, addr = next, addr != end); } |
2ba3e6947 mm/vmalloc: track... |
353 354 |
static void vunmap_pud_range(p4d_t *p4d, unsigned long addr, unsigned long end, pgtbl_mod_mask *mask) |
1da177e4c Linux-2.6.12-rc2 |
355 356 357 |
{ pud_t *pud; unsigned long next; |
2ba3e6947 mm/vmalloc: track... |
358 |
int cleared; |
1da177e4c Linux-2.6.12-rc2 |
359 |
|
c2febafc6 mm: convert gener... |
360 |
pud = pud_offset(p4d, addr); |
1da177e4c Linux-2.6.12-rc2 |
361 362 |
do { next = pud_addr_end(addr, end); |
2ba3e6947 mm/vmalloc: track... |
363 364 365 366 367 368 |
cleared = pud_clear_huge(pud); if (cleared || pud_bad(*pud)) *mask |= PGTBL_PUD_MODIFIED; if (cleared) |
b9820d8f3 mm: change vunmap... |
369 |
continue; |
1da177e4c Linux-2.6.12-rc2 |
370 371 |
if (pud_none_or_clear_bad(pud)) continue; |
2ba3e6947 mm/vmalloc: track... |
372 |
vunmap_pmd_range(pud, addr, next, mask); |
1da177e4c Linux-2.6.12-rc2 |
373 374 |
} while (pud++, addr = next, addr != end); } |
2ba3e6947 mm/vmalloc: track... |
375 376 |
static void vunmap_p4d_range(pgd_t *pgd, unsigned long addr, unsigned long end, pgtbl_mod_mask *mask) |
c2febafc6 mm: convert gener... |
377 378 379 |
{ p4d_t *p4d; unsigned long next; |
2ba3e6947 mm/vmalloc: track... |
380 |
int cleared; |
c2febafc6 mm: convert gener... |
381 382 383 384 |
p4d = p4d_offset(pgd, addr); do { next = p4d_addr_end(addr, end); |
2ba3e6947 mm/vmalloc: track... |
385 386 387 388 389 390 |
cleared = p4d_clear_huge(p4d); if (cleared || p4d_bad(*p4d)) *mask |= PGTBL_P4D_MODIFIED; if (cleared) |
c2febafc6 mm: convert gener... |
391 392 393 |
continue; if (p4d_none_or_clear_bad(p4d)) continue; |
2ba3e6947 mm/vmalloc: track... |
394 |
vunmap_pud_range(p4d, addr, next, mask); |
c2febafc6 mm: convert gener... |
395 396 |
} while (p4d++, addr = next, addr != end); } |
4ad0ae8c6 mm/vmalloc: remov... |
397 398 399 |
/* * vunmap_range_noflush is similar to vunmap_range, but does not * flush caches or TLBs. |
b521c43f5 mm: remove vmap_p... |
400 |
* |
4ad0ae8c6 mm/vmalloc: remov... |
401 402 403 404 405 |
* The caller is responsible for calling flush_cache_vmap() before calling * this function, and flush_tlb_kernel_range after it has returned * successfully (and before the addresses are expected to cause a page fault * or be re-mapped for something else, if TLB flushes are being delayed or * coalesced). |
b521c43f5 mm: remove vmap_p... |
406 |
* |
4ad0ae8c6 mm/vmalloc: remov... |
407 |
* This is an internal function only. Do not use outside mm/. |
b521c43f5 mm: remove vmap_p... |
408 |
*/ |
4ad0ae8c6 mm/vmalloc: remov... |
409 |
void vunmap_range_noflush(unsigned long start, unsigned long end) |
1da177e4c Linux-2.6.12-rc2 |
410 |
{ |
1da177e4c Linux-2.6.12-rc2 |
411 |
unsigned long next; |
b521c43f5 mm: remove vmap_p... |
412 |
pgd_t *pgd; |
2ba3e6947 mm/vmalloc: track... |
413 414 |
unsigned long addr = start; pgtbl_mod_mask mask = 0; |
1da177e4c Linux-2.6.12-rc2 |
415 416 417 |
BUG_ON(addr >= end); pgd = pgd_offset_k(addr); |
1da177e4c Linux-2.6.12-rc2 |
418 419 |
do { next = pgd_addr_end(addr, end); |
2ba3e6947 mm/vmalloc: track... |
420 421 |
if (pgd_bad(*pgd)) mask |= PGTBL_PGD_MODIFIED; |
1da177e4c Linux-2.6.12-rc2 |
422 423 |
if (pgd_none_or_clear_bad(pgd)) continue; |
2ba3e6947 mm/vmalloc: track... |
424 |
vunmap_p4d_range(pgd, addr, next, &mask); |
1da177e4c Linux-2.6.12-rc2 |
425 |
} while (pgd++, addr = next, addr != end); |
2ba3e6947 mm/vmalloc: track... |
426 427 428 |
if (mask & ARCH_PAGE_TABLE_SYNC_MASK) arch_sync_kernel_mappings(start, end); |
1da177e4c Linux-2.6.12-rc2 |
429 |
} |
4ad0ae8c6 mm/vmalloc: remov... |
430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 |
/** * vunmap_range - unmap kernel virtual addresses * @addr: start of the VM area to unmap * @end: end of the VM area to unmap (non-inclusive) * * Clears any present PTEs in the virtual address range, flushes TLBs and * caches. Any subsequent access to the address before it has been re-mapped * is a kernel bug. */ void vunmap_range(unsigned long addr, unsigned long end) { flush_cache_vunmap(addr, end); vunmap_range_noflush(addr, end); flush_tlb_kernel_range(addr, end); } |
0a2648840 mm/vmalloc: renam... |
445 |
static int vmap_pages_pte_range(pmd_t *pmd, unsigned long addr, |
2ba3e6947 mm/vmalloc: track... |
446 447 |
unsigned long end, pgprot_t prot, struct page **pages, int *nr, pgtbl_mod_mask *mask) |
1da177e4c Linux-2.6.12-rc2 |
448 449 |
{ pte_t *pte; |
db64fe022 mm: rewrite vmap ... |
450 451 452 453 |
/* * nr is a running index into the array which helps higher level * callers keep track of where we're up to. */ |
2ba3e6947 mm/vmalloc: track... |
454 |
pte = pte_alloc_kernel_track(pmd, addr, mask); |
1da177e4c Linux-2.6.12-rc2 |
455 456 457 |
if (!pte) return -ENOMEM; do { |
db64fe022 mm: rewrite vmap ... |
458 459 460 461 462 |
struct page *page = pages[*nr]; if (WARN_ON(!pte_none(*pte))) return -EBUSY; if (WARN_ON(!page)) |
1da177e4c Linux-2.6.12-rc2 |
463 464 |
return -ENOMEM; set_pte_at(&init_mm, addr, pte, mk_pte(page, prot)); |
db64fe022 mm: rewrite vmap ... |
465 |
(*nr)++; |
1da177e4c Linux-2.6.12-rc2 |
466 |
} while (pte++, addr += PAGE_SIZE, addr != end); |
2ba3e6947 mm/vmalloc: track... |
467 |
*mask |= PGTBL_PTE_MODIFIED; |
1da177e4c Linux-2.6.12-rc2 |
468 469 |
return 0; } |
0a2648840 mm/vmalloc: renam... |
470 |
static int vmap_pages_pmd_range(pud_t *pud, unsigned long addr, |
2ba3e6947 mm/vmalloc: track... |
471 472 |
unsigned long end, pgprot_t prot, struct page **pages, int *nr, pgtbl_mod_mask *mask) |
1da177e4c Linux-2.6.12-rc2 |
473 474 475 |
{ pmd_t *pmd; unsigned long next; |
2ba3e6947 mm/vmalloc: track... |
476 |
pmd = pmd_alloc_track(&init_mm, pud, addr, mask); |
1da177e4c Linux-2.6.12-rc2 |
477 478 479 480 |
if (!pmd) return -ENOMEM; do { next = pmd_addr_end(addr, end); |
0a2648840 mm/vmalloc: renam... |
481 |
if (vmap_pages_pte_range(pmd, addr, next, prot, pages, nr, mask)) |
1da177e4c Linux-2.6.12-rc2 |
482 483 484 485 |
return -ENOMEM; } while (pmd++, addr = next, addr != end); return 0; } |
0a2648840 mm/vmalloc: renam... |
486 |
static int vmap_pages_pud_range(p4d_t *p4d, unsigned long addr, |
2ba3e6947 mm/vmalloc: track... |
487 488 |
unsigned long end, pgprot_t prot, struct page **pages, int *nr, pgtbl_mod_mask *mask) |
1da177e4c Linux-2.6.12-rc2 |
489 490 491 |
{ pud_t *pud; unsigned long next; |
2ba3e6947 mm/vmalloc: track... |
492 |
pud = pud_alloc_track(&init_mm, p4d, addr, mask); |
1da177e4c Linux-2.6.12-rc2 |
493 494 495 496 |
if (!pud) return -ENOMEM; do { next = pud_addr_end(addr, end); |
0a2648840 mm/vmalloc: renam... |
497 |
if (vmap_pages_pmd_range(pud, addr, next, prot, pages, nr, mask)) |
1da177e4c Linux-2.6.12-rc2 |
498 499 500 501 |
return -ENOMEM; } while (pud++, addr = next, addr != end); return 0; } |
0a2648840 mm/vmalloc: renam... |
502 |
static int vmap_pages_p4d_range(pgd_t *pgd, unsigned long addr, |
2ba3e6947 mm/vmalloc: track... |
503 504 |
unsigned long end, pgprot_t prot, struct page **pages, int *nr, pgtbl_mod_mask *mask) |
c2febafc6 mm: convert gener... |
505 506 507 |
{ p4d_t *p4d; unsigned long next; |
2ba3e6947 mm/vmalloc: track... |
508 |
p4d = p4d_alloc_track(&init_mm, pgd, addr, mask); |
c2febafc6 mm: convert gener... |
509 510 511 512 |
if (!p4d) return -ENOMEM; do { next = p4d_addr_end(addr, end); |
0a2648840 mm/vmalloc: renam... |
513 |
if (vmap_pages_pud_range(p4d, addr, next, prot, pages, nr, mask)) |
c2febafc6 mm: convert gener... |
514 515 516 517 |
return -ENOMEM; } while (p4d++, addr = next, addr != end); return 0; } |
121e6f325 mm/vmalloc: hugep... |
518 519 |
static int vmap_small_pages_range_noflush(unsigned long addr, unsigned long end, pgprot_t prot, struct page **pages) |
1da177e4c Linux-2.6.12-rc2 |
520 |
{ |
2ba3e6947 mm/vmalloc: track... |
521 |
unsigned long start = addr; |
b521c43f5 mm: remove vmap_p... |
522 |
pgd_t *pgd; |
121e6f325 mm/vmalloc: hugep... |
523 |
unsigned long next; |
db64fe022 mm: rewrite vmap ... |
524 525 |
int err = 0; int nr = 0; |
2ba3e6947 mm/vmalloc: track... |
526 |
pgtbl_mod_mask mask = 0; |
1da177e4c Linux-2.6.12-rc2 |
527 528 529 |
BUG_ON(addr >= end); pgd = pgd_offset_k(addr); |
1da177e4c Linux-2.6.12-rc2 |
530 531 |
do { next = pgd_addr_end(addr, end); |
2ba3e6947 mm/vmalloc: track... |
532 533 |
if (pgd_bad(*pgd)) mask |= PGTBL_PGD_MODIFIED; |
0a2648840 mm/vmalloc: renam... |
534 |
err = vmap_pages_p4d_range(pgd, addr, next, prot, pages, &nr, &mask); |
1da177e4c Linux-2.6.12-rc2 |
535 |
if (err) |
bf88c8c83 vmalloc.c: fix do... |
536 |
return err; |
1da177e4c Linux-2.6.12-rc2 |
537 |
} while (pgd++, addr = next, addr != end); |
db64fe022 mm: rewrite vmap ... |
538 |
|
2ba3e6947 mm/vmalloc: track... |
539 540 |
if (mask & ARCH_PAGE_TABLE_SYNC_MASK) arch_sync_kernel_mappings(start, end); |
60bb44652 mm: don't return ... |
541 |
return 0; |
1da177e4c Linux-2.6.12-rc2 |
542 |
} |
b67177ecd mm/vmalloc: remov... |
543 544 545 546 547 548 549 550 551 552 |
/* * vmap_pages_range_noflush is similar to vmap_pages_range, but does not * flush caches. * * The caller is responsible for calling flush_cache_vmap() after this * function returns successfully and before the addresses are accessed. * * This is an internal function only. Do not use outside mm/. */ int vmap_pages_range_noflush(unsigned long addr, unsigned long end, |
121e6f325 mm/vmalloc: hugep... |
553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 |
pgprot_t prot, struct page **pages, unsigned int page_shift) { unsigned int i, nr = (end - addr) >> PAGE_SHIFT; WARN_ON(page_shift < PAGE_SHIFT); if (!IS_ENABLED(CONFIG_HAVE_ARCH_HUGE_VMALLOC) || page_shift == PAGE_SHIFT) return vmap_small_pages_range_noflush(addr, end, prot, pages); for (i = 0; i < nr; i += 1U << (page_shift - PAGE_SHIFT)) { int err; err = vmap_range_noflush(addr, addr + (1UL << page_shift), __pa(page_address(pages[i])), prot, page_shift); if (err) return err; addr += 1UL << page_shift; } return 0; } |
121e6f325 mm/vmalloc: hugep... |
577 |
/** |
b67177ecd mm/vmalloc: remov... |
578 |
* vmap_pages_range - map pages to a kernel virtual address |
121e6f325 mm/vmalloc: hugep... |
579 |
* @addr: start of the VM area to map |
b67177ecd mm/vmalloc: remov... |
580 |
* @end: end of the VM area to map (non-inclusive) |
121e6f325 mm/vmalloc: hugep... |
581 |
* @prot: page protection flags to use |
b67177ecd mm/vmalloc: remov... |
582 583 584 |
* @pages: pages to map (always PAGE_SIZE pages) * @page_shift: maximum shift that the pages may be mapped with, @pages must * be aligned and contiguous up to at least this shift. |
121e6f325 mm/vmalloc: hugep... |
585 586 587 588 |
* * RETURNS: * 0 on success, -errno on failure. */ |
b67177ecd mm/vmalloc: remov... |
589 590 |
static int vmap_pages_range(unsigned long addr, unsigned long end, pgprot_t prot, struct page **pages, unsigned int page_shift) |
8fc489850 vmalloc: add un/m... |
591 |
{ |
b67177ecd mm/vmalloc: remov... |
592 |
int err; |
8fc489850 vmalloc: add un/m... |
593 |
|
b67177ecd mm/vmalloc: remov... |
594 595 596 |
err = vmap_pages_range_noflush(addr, end, prot, pages, page_shift); flush_cache_vmap(addr, end); return err; |
8fc489850 vmalloc: add un/m... |
597 |
} |
81ac3ad90 kcore: register m... |
598 |
int is_vmalloc_or_module_addr(const void *x) |
73bdf0a60 Introduce is_vmal... |
599 600 |
{ /* |
ab4f2ee13 [ARM] fix naming ... |
601 |
* ARM, x86-64 and sparc64 put modules in a special place, |
73bdf0a60 Introduce is_vmal... |
602 603 604 605 606 607 608 609 610 611 |
* 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... |
612 |
/* |
c0eb315ad mm/vmalloc: fix H... |
613 614 615 |
* Walk a vmap address to the struct page it maps. Huge vmap mappings will * return the tail page that corresponds to the base page address, which * matches small vmap mappings. |
48667e7a4 Move vmalloc_to_p... |
616 |
*/ |
add688fbd Revert "mm/vmallo... |
617 |
struct page *vmalloc_to_page(const void *vmalloc_addr) |
48667e7a4 Move vmalloc_to_p... |
618 619 |
{ unsigned long addr = (unsigned long) vmalloc_addr; |
add688fbd Revert "mm/vmallo... |
620 |
struct page *page = NULL; |
48667e7a4 Move vmalloc_to_p... |
621 |
pgd_t *pgd = pgd_offset_k(addr); |
c2febafc6 mm: convert gener... |
622 623 624 625 |
p4d_t *p4d; pud_t *pud; pmd_t *pmd; pte_t *ptep, pte; |
48667e7a4 Move vmalloc_to_p... |
626 |
|
7aa413def x86, MM: virtual ... |
627 628 629 630 |
/* * 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... |
631 |
VIRTUAL_BUG_ON(!is_vmalloc_or_module_addr(vmalloc_addr)); |
59ea74633 MM: virtual addre... |
632 |
|
c2febafc6 mm: convert gener... |
633 634 |
if (pgd_none(*pgd)) return NULL; |
c0eb315ad mm/vmalloc: fix H... |
635 636 637 638 |
if (WARN_ON_ONCE(pgd_leaf(*pgd))) return NULL; /* XXX: no allowance for huge pgd */ if (WARN_ON_ONCE(pgd_bad(*pgd))) return NULL; |
c2febafc6 mm: convert gener... |
639 640 641 |
p4d = p4d_offset(pgd, addr); if (p4d_none(*p4d)) return NULL; |
c0eb315ad mm/vmalloc: fix H... |
642 643 644 645 |
if (p4d_leaf(*p4d)) return p4d_page(*p4d) + ((addr & ~P4D_MASK) >> PAGE_SHIFT); if (WARN_ON_ONCE(p4d_bad(*p4d))) return NULL; |
029c54b09 mm/vmalloc.c: hug... |
646 |
|
c0eb315ad mm/vmalloc: fix H... |
647 648 649 650 651 652 |
pud = pud_offset(p4d, addr); if (pud_none(*pud)) return NULL; if (pud_leaf(*pud)) return pud_page(*pud) + ((addr & ~PUD_MASK) >> PAGE_SHIFT); if (WARN_ON_ONCE(pud_bad(*pud))) |
c2febafc6 mm: convert gener... |
653 |
return NULL; |
c0eb315ad mm/vmalloc: fix H... |
654 |
|
c2febafc6 mm: convert gener... |
655 |
pmd = pmd_offset(pud, addr); |
c0eb315ad mm/vmalloc: fix H... |
656 657 658 659 660 |
if (pmd_none(*pmd)) return NULL; if (pmd_leaf(*pmd)) return pmd_page(*pmd) + ((addr & ~PMD_MASK) >> PAGE_SHIFT); if (WARN_ON_ONCE(pmd_bad(*pmd))) |
c2febafc6 mm: convert gener... |
661 662 663 664 665 666 667 |
return NULL; ptep = pte_offset_map(pmd, addr); pte = *ptep; if (pte_present(pte)) page = pte_page(pte); pte_unmap(ptep); |
c0eb315ad mm/vmalloc: fix H... |
668 |
|
add688fbd Revert "mm/vmallo... |
669 |
return page; |
48667e7a4 Move vmalloc_to_p... |
670 |
} |
add688fbd Revert "mm/vmallo... |
671 |
EXPORT_SYMBOL(vmalloc_to_page); |
48667e7a4 Move vmalloc_to_p... |
672 673 |
/* |
add688fbd Revert "mm/vmallo... |
674 |
* Map a vmalloc()-space virtual address to the physical page frame number. |
48667e7a4 Move vmalloc_to_p... |
675 |
*/ |
add688fbd Revert "mm/vmallo... |
676 |
unsigned long vmalloc_to_pfn(const void *vmalloc_addr) |
48667e7a4 Move vmalloc_to_p... |
677 |
{ |
add688fbd Revert "mm/vmallo... |
678 |
return page_to_pfn(vmalloc_to_page(vmalloc_addr)); |
48667e7a4 Move vmalloc_to_p... |
679 |
} |
add688fbd Revert "mm/vmallo... |
680 |
EXPORT_SYMBOL(vmalloc_to_pfn); |
48667e7a4 Move vmalloc_to_p... |
681 |
|
db64fe022 mm: rewrite vmap ... |
682 683 |
/*** Global kva allocator ***/ |
bb850f4da mm/vmap: add DEBU... |
684 |
#define DEBUG_AUGMENT_PROPAGATE_CHECK 0 |
a6cf4e0fe mm/vmap: add DEBU... |
685 |
#define DEBUG_AUGMENT_LOWEST_MATCH_CHECK 0 |
bb850f4da mm/vmap: add DEBU... |
686 |
|
db64fe022 mm: rewrite vmap ... |
687 |
|
db64fe022 mm: rewrite vmap ... |
688 |
static DEFINE_SPINLOCK(vmap_area_lock); |
e36176be1 mm/vmalloc: rewor... |
689 |
static DEFINE_SPINLOCK(free_vmap_area_lock); |
f1c4069e1 mm, vmalloc: expo... |
690 691 |
/* Export for kexec only */ LIST_HEAD(vmap_area_list); |
89699605f mm: vmap area cache |
692 |
static struct rb_root vmap_area_root = RB_ROOT; |
68ad4a330 mm/vmalloc.c: kee... |
693 |
static bool vmap_initialized __read_mostly; |
89699605f mm: vmap area cache |
694 |
|
96e2db456 mm/vmalloc: rewor... |
695 696 697 |
static struct rb_root purge_vmap_area_root = RB_ROOT; static LIST_HEAD(purge_vmap_area_list); static DEFINE_SPINLOCK(purge_vmap_area_lock); |
68ad4a330 mm/vmalloc.c: kee... |
698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 |
/* * This kmem_cache is used for vmap_area objects. Instead of * allocating from slab we reuse an object from this cache to * make things faster. Especially in "no edge" splitting of * free block. */ static struct kmem_cache *vmap_area_cachep; /* * This linked list is used in pair with free_vmap_area_root. * It gives O(1) access to prev/next to perform fast coalescing. */ static LIST_HEAD(free_vmap_area_list); /* * This augment red-black tree represents the free vmap space. * All vmap_area objects in this tree are sorted by va->va_start * address. It is used for allocation and merging when a vmap * object is released. * * Each vmap_area node contains a maximum available free block * of its sub-tree, right or left. Therefore it is possible to * find a lowest match of free area. */ static struct rb_root free_vmap_area_root = RB_ROOT; |
82dd23e84 mm/vmalloc.c: pre... |
723 724 725 726 727 728 |
/* * Preload a CPU with one object for "no edge" split case. The * aim is to get rid of allocations from the atomic context, thus * to use more permissive allocation masks. */ static DEFINE_PER_CPU(struct vmap_area *, ne_fit_preload_node); |
68ad4a330 mm/vmalloc.c: kee... |
729 730 731 732 733 734 735 736 737 738 739 740 741 742 |
static __always_inline unsigned long va_size(struct vmap_area *va) { return (va->va_end - va->va_start); } static __always_inline unsigned long get_subtree_max_size(struct rb_node *node) { struct vmap_area *va; va = rb_entry_safe(node, struct vmap_area, rb_node); return va ? va->subtree_max_size : 0; } |
89699605f mm: vmap area cache |
743 |
|
68ad4a330 mm/vmalloc.c: kee... |
744 745 746 747 748 749 750 751 752 753 |
/* * Gets called when remove the node and rotate. */ static __always_inline unsigned long compute_subtree_max_size(struct vmap_area *va) { return max3(va_size(va), get_subtree_max_size(va->rb_node.rb_left), get_subtree_max_size(va->rb_node.rb_right)); } |
315cc066b augmented rbtree:... |
754 755 |
RB_DECLARE_CALLBACKS_MAX(static, free_vmap_area_rb_augment_cb, struct vmap_area, rb_node, unsigned long, subtree_max_size, va_size) |
68ad4a330 mm/vmalloc.c: kee... |
756 757 758 759 |
static void purge_vmap_area_lazy(void); static BLOCKING_NOTIFIER_HEAD(vmap_notify_list); static unsigned long lazy_max_pages(void); |
db64fe022 mm: rewrite vmap ... |
760 |
|
97105f0ab mm: vmalloc: show... |
761 762 763 764 765 766 |
static atomic_long_t nr_vmalloc_pages; unsigned long vmalloc_nr_pages(void) { return atomic_long_read(&nr_vmalloc_pages); } |
f181234a5 mm/vmalloc: fix w... |
767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 |
static struct vmap_area *find_vmap_area_exceed_addr(unsigned long addr) { struct vmap_area *va = NULL; struct rb_node *n = vmap_area_root.rb_node; while (n) { struct vmap_area *tmp; tmp = rb_entry(n, struct vmap_area, rb_node); if (tmp->va_end > addr) { va = tmp; if (tmp->va_start <= addr) break; n = n->rb_left; } else n = n->rb_right; } return va; } |
db64fe022 mm: rewrite vmap ... |
788 |
static struct vmap_area *__find_vmap_area(unsigned long addr) |
1da177e4c Linux-2.6.12-rc2 |
789 |
{ |
db64fe022 mm: rewrite vmap ... |
790 791 792 793 794 795 796 797 |
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... |
798 |
else if (addr >= va->va_end) |
db64fe022 mm: rewrite vmap ... |
799 800 801 802 803 804 805 |
n = n->rb_right; else return va; } return NULL; } |
68ad4a330 mm/vmalloc.c: kee... |
806 807 808 |
/* * This function returns back addresses of parent node * and its left or right link for further processing. |
9c801f61d mm/vmalloc.c: rem... |
809 810 811 812 |
* * Otherwise NULL is returned. In that case all further * steps regarding inserting of conflicting overlap range * have to be declined and actually considered as a bug. |
68ad4a330 mm/vmalloc.c: kee... |
813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 |
*/ static __always_inline struct rb_node ** find_va_links(struct vmap_area *va, struct rb_root *root, struct rb_node *from, struct rb_node **parent) { struct vmap_area *tmp_va; struct rb_node **link; if (root) { link = &root->rb_node; if (unlikely(!*link)) { *parent = NULL; return link; } } else { link = &from; } |
db64fe022 mm: rewrite vmap ... |
831 |
|
68ad4a330 mm/vmalloc.c: kee... |
832 833 834 835 836 837 838 |
/* * Go to the bottom of the tree. When we hit the last point * we end up with parent rb_node and correct direction, i name * it link, where the new va->rb_node will be attached to. */ do { tmp_va = rb_entry(*link, struct vmap_area, rb_node); |
db64fe022 mm: rewrite vmap ... |
839 |
|
68ad4a330 mm/vmalloc.c: kee... |
840 841 842 843 844 845 846 847 848 849 850 |
/* * During the traversal we also do some sanity check. * Trigger the BUG() if there are sides(left/right) * or full overlaps. */ if (va->va_start < tmp_va->va_end && va->va_end <= tmp_va->va_start) link = &(*link)->rb_left; else if (va->va_end > tmp_va->va_start && va->va_start >= tmp_va->va_end) link = &(*link)->rb_right; |
9c801f61d mm/vmalloc.c: rem... |
851 852 853 854 855 856 857 |
else { WARN(1, "vmalloc bug: 0x%lx-0x%lx overlaps with 0x%lx-0x%lx ", va->va_start, va->va_end, tmp_va->va_start, tmp_va->va_end); return NULL; } |
68ad4a330 mm/vmalloc.c: kee... |
858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 |
} while (*link); *parent = &tmp_va->rb_node; return link; } static __always_inline struct list_head * get_va_next_sibling(struct rb_node *parent, struct rb_node **link) { struct list_head *list; if (unlikely(!parent)) /* * The red-black tree where we try to find VA neighbors * before merging or inserting is empty, i.e. it means * there is no free vmap space. Normally it does not * happen but we handle this case anyway. */ return NULL; list = &rb_entry(parent, struct vmap_area, rb_node)->list; return (&parent->rb_right == link ? list->next : list); } static __always_inline void link_va(struct vmap_area *va, struct rb_root *root, struct rb_node *parent, struct rb_node **link, struct list_head *head) { /* * VA is still not in the list, but we can * identify its future previous list_head node. */ if (likely(parent)) { head = &rb_entry(parent, struct vmap_area, rb_node)->list; if (&parent->rb_right != link) head = head->prev; |
db64fe022 mm: rewrite vmap ... |
894 |
} |
68ad4a330 mm/vmalloc.c: kee... |
895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 |
/* Insert to the rb-tree */ rb_link_node(&va->rb_node, parent, link); if (root == &free_vmap_area_root) { /* * Some explanation here. Just perform simple insertion * to the tree. We do not set va->subtree_max_size to * its current size before calling rb_insert_augmented(). * It is because of we populate the tree from the bottom * to parent levels when the node _is_ in the tree. * * Therefore we set subtree_max_size to zero after insertion, * to let __augment_tree_propagate_from() puts everything to * the correct order later on. */ rb_insert_augmented(&va->rb_node, root, &free_vmap_area_rb_augment_cb); va->subtree_max_size = 0; } else { rb_insert_color(&va->rb_node, root); } |
db64fe022 mm: rewrite vmap ... |
915 |
|
68ad4a330 mm/vmalloc.c: kee... |
916 917 |
/* Address-sort this list */ list_add(&va->list, head); |
db64fe022 mm: rewrite vmap ... |
918 |
} |
68ad4a330 mm/vmalloc.c: kee... |
919 920 921 |
static __always_inline void unlink_va(struct vmap_area *va, struct rb_root *root) { |
460e42d19 mm/vmalloc.c: swi... |
922 923 |
if (WARN_ON(RB_EMPTY_NODE(&va->rb_node))) return; |
db64fe022 mm: rewrite vmap ... |
924 |
|
460e42d19 mm/vmalloc.c: swi... |
925 926 927 928 929 930 931 932 |
if (root == &free_vmap_area_root) rb_erase_augmented(&va->rb_node, root, &free_vmap_area_rb_augment_cb); else rb_erase(&va->rb_node, root); list_del(&va->list); RB_CLEAR_NODE(&va->rb_node); |
68ad4a330 mm/vmalloc.c: kee... |
933 |
} |
bb850f4da mm/vmap: add DEBU... |
934 935 |
#if DEBUG_AUGMENT_PROPAGATE_CHECK static void |
da27c9ed1 mm/vmalloc: simpl... |
936 |
augment_tree_propagate_check(void) |
bb850f4da mm/vmap: add DEBU... |
937 938 |
{ struct vmap_area *va; |
da27c9ed1 mm/vmalloc: simpl... |
939 |
unsigned long computed_size; |
bb850f4da mm/vmap: add DEBU... |
940 |
|
da27c9ed1 mm/vmalloc: simpl... |
941 942 943 944 945 946 |
list_for_each_entry(va, &free_vmap_area_list, list) { computed_size = compute_subtree_max_size(va); if (computed_size != va->subtree_max_size) pr_emerg("tree is corrupted: %lu, %lu ", va_size(va), va->subtree_max_size); |
bb850f4da mm/vmap: add DEBU... |
947 |
} |
bb850f4da mm/vmap: add DEBU... |
948 949 |
} #endif |
68ad4a330 mm/vmalloc.c: kee... |
950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 |
/* * This function populates subtree_max_size from bottom to upper * levels starting from VA point. The propagation must be done * when VA size is modified by changing its va_start/va_end. Or * in case of newly inserting of VA to the tree. * * It means that __augment_tree_propagate_from() must be called: * - After VA has been inserted to the tree(free path); * - After VA has been shrunk(allocation path); * - After VA has been increased(merging path). * * Please note that, it does not mean that upper parent nodes * and their subtree_max_size are recalculated all the time up * to the root node. * * 4--8 * /\ * / \ * / \ * 2--2 8--8 * * For example if we modify the node 4, shrinking it to 2, then * no any modification is required. If we shrink the node 2 to 1 * its subtree_max_size is updated only, and set to 1. If we shrink * the node 8 to 6, then its subtree_max_size is set to 6 and parent * node becomes 4--6. */ static __always_inline void augment_tree_propagate_from(struct vmap_area *va) { |
15ae144f7 mm/vmalloc: switc... |
980 981 982 983 984 985 |
/* * Populate the tree from bottom towards the root until * the calculated maximum available size of checked node * is equal to its current one. */ free_vmap_area_rb_augment_cb_propagate(&va->rb_node, NULL); |
bb850f4da mm/vmap: add DEBU... |
986 987 |
#if DEBUG_AUGMENT_PROPAGATE_CHECK |
da27c9ed1 mm/vmalloc: simpl... |
988 |
augment_tree_propagate_check(); |
bb850f4da mm/vmap: add DEBU... |
989 |
#endif |
68ad4a330 mm/vmalloc.c: kee... |
990 991 992 993 994 995 996 997 998 999 |
} static void insert_vmap_area(struct vmap_area *va, struct rb_root *root, struct list_head *head) { struct rb_node **link; struct rb_node *parent; link = find_va_links(va, root, NULL, &parent); |
9c801f61d mm/vmalloc.c: rem... |
1000 1001 |
if (link) link_va(va, root, parent, link, head); |
68ad4a330 mm/vmalloc.c: kee... |
1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 |
} static void insert_vmap_area_augment(struct vmap_area *va, struct rb_node *from, struct rb_root *root, struct list_head *head) { struct rb_node **link; struct rb_node *parent; if (from) link = find_va_links(va, NULL, from, &parent); else link = find_va_links(va, root, NULL, &parent); |
9c801f61d mm/vmalloc.c: rem... |
1016 1017 1018 1019 |
if (link) { link_va(va, root, parent, link, head); augment_tree_propagate_from(va); } |
68ad4a330 mm/vmalloc.c: kee... |
1020 1021 1022 1023 1024 1025 1026 |
} /* * Merge de-allocated chunk of VA memory with previous * and next free blocks. If coalesce is not done a new * free area is inserted. If VA has been merged, it is * freed. |
9c801f61d mm/vmalloc.c: rem... |
1027 1028 1029 1030 1031 |
* * Please note, it can return NULL in case of overlap * ranges, followed by WARN() report. Despite it is a * buggy behaviour, a system can be alive and keep * ongoing. |
68ad4a330 mm/vmalloc.c: kee... |
1032 |
*/ |
3c5c3cfb9 kasan: support ba... |
1033 |
static __always_inline struct vmap_area * |
68ad4a330 mm/vmalloc.c: kee... |
1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 |
merge_or_add_vmap_area(struct vmap_area *va, struct rb_root *root, struct list_head *head) { struct vmap_area *sibling; struct list_head *next; struct rb_node **link; struct rb_node *parent; bool merged = false; /* * Find a place in the tree where VA potentially will be * inserted, unless it is merged with its sibling/siblings. */ link = find_va_links(va, root, NULL, &parent); |
9c801f61d mm/vmalloc.c: rem... |
1048 1049 |
if (!link) return NULL; |
68ad4a330 mm/vmalloc.c: kee... |
1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 |
/* * Get next node of VA to check if merging can be done. */ next = get_va_next_sibling(parent, link); if (unlikely(next == NULL)) goto insert; /* * start end * | | * |<------VA------>|<-----Next----->| * | | * start end */ if (next != head) { sibling = list_entry(next, struct vmap_area, list); if (sibling->va_start == va->va_end) { sibling->va_start = va->va_start; |
68ad4a330 mm/vmalloc.c: kee... |
1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 |
/* Free vmap_area object. */ kmem_cache_free(vmap_area_cachep, va); /* Point to the new merged area. */ va = sibling; merged = true; } } /* * start end * | | * |<-----Prev----->|<------VA------>| * | | * start end */ if (next->prev != head) { sibling = list_entry(next->prev, struct vmap_area, list); if (sibling->va_end == va->va_start) { |
5dd786409 mm/vmalloc: simpl... |
1088 1089 1090 1091 1092 1093 1094 |
/* * If both neighbors are coalesced, it is important * to unlink the "next" node first, followed by merging * with "previous" one. Otherwise the tree might not be * fully populated if a sibling's augmented value is * "normalized" because of rotation operations. */ |
54f63d9d8 mm/vmalloc.c: get... |
1095 1096 |
if (merged) unlink_va(va, root); |
68ad4a330 mm/vmalloc.c: kee... |
1097 |
|
5dd786409 mm/vmalloc: simpl... |
1098 |
sibling->va_end = va->va_end; |
68ad4a330 mm/vmalloc.c: kee... |
1099 1100 |
/* Free vmap_area object. */ kmem_cache_free(vmap_area_cachep, va); |
3c5c3cfb9 kasan: support ba... |
1101 1102 1103 1104 |
/* Point to the new merged area. */ va = sibling; merged = true; |
68ad4a330 mm/vmalloc.c: kee... |
1105 1106 1107 1108 |
} } insert: |
5dd786409 mm/vmalloc: simpl... |
1109 |
if (!merged) |
68ad4a330 mm/vmalloc.c: kee... |
1110 |
link_va(va, root, parent, link, head); |
3c5c3cfb9 kasan: support ba... |
1111 |
|
96e2db456 mm/vmalloc: rewor... |
1112 1113 1114 1115 1116 1117 1118 1119 1120 1121 |
return va; } static __always_inline struct vmap_area * merge_or_add_vmap_area_augment(struct vmap_area *va, struct rb_root *root, struct list_head *head) { va = merge_or_add_vmap_area(va, root, head); if (va) augment_tree_propagate_from(va); |
3c5c3cfb9 kasan: support ba... |
1122 |
return va; |
68ad4a330 mm/vmalloc.c: kee... |
1123 1124 1125 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151 1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 1179 1180 1181 1182 1183 |
} static __always_inline bool is_within_this_va(struct vmap_area *va, unsigned long size, unsigned long align, unsigned long vstart) { unsigned long nva_start_addr; if (va->va_start > vstart) nva_start_addr = ALIGN(va->va_start, align); else nva_start_addr = ALIGN(vstart, align); /* Can be overflowed due to big size or alignment. */ if (nva_start_addr + size < nva_start_addr || nva_start_addr < vstart) return false; return (nva_start_addr + size <= va->va_end); } /* * Find the first free block(lowest start address) in the tree, * that will accomplish the request corresponding to passing * parameters. */ static __always_inline struct vmap_area * find_vmap_lowest_match(unsigned long size, unsigned long align, unsigned long vstart) { struct vmap_area *va; struct rb_node *node; unsigned long length; /* Start from the root. */ node = free_vmap_area_root.rb_node; /* Adjust the search size for alignment overhead. */ length = size + align - 1; while (node) { va = rb_entry(node, struct vmap_area, rb_node); if (get_subtree_max_size(node->rb_left) >= length && vstart < va->va_start) { node = node->rb_left; } else { if (is_within_this_va(va, size, align, vstart)) return va; /* * Does not make sense to go deeper towards the right * sub-tree if it does not have a free block that is * equal or bigger to the requested search length. */ if (get_subtree_max_size(node->rb_right) >= length) { node = node->rb_right; continue; } /* |
3806b0414 mm/vmalloc.c: fix... |
1184 |
* OK. We roll back and find the first right sub-tree, |
68ad4a330 mm/vmalloc.c: kee... |
1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199 1200 1201 1202 1203 |
* that will satisfy the search criteria. It can happen * only once due to "vstart" restriction. */ while ((node = rb_parent(node))) { va = rb_entry(node, struct vmap_area, rb_node); if (is_within_this_va(va, size, align, vstart)) return va; if (get_subtree_max_size(node->rb_right) >= length && vstart <= va->va_start) { node = node->rb_right; break; } } } } return NULL; } |
a6cf4e0fe mm/vmap: add DEBU... |
1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 1230 1231 1232 1233 1234 1235 1236 1237 1238 1239 1240 1241 |
#if DEBUG_AUGMENT_LOWEST_MATCH_CHECK #include <linux/random.h> static struct vmap_area * find_vmap_lowest_linear_match(unsigned long size, unsigned long align, unsigned long vstart) { struct vmap_area *va; list_for_each_entry(va, &free_vmap_area_list, list) { if (!is_within_this_va(va, size, align, vstart)) continue; return va; } return NULL; } static void find_vmap_lowest_match_check(unsigned long size) { struct vmap_area *va_1, *va_2; unsigned long vstart; unsigned int rnd; get_random_bytes(&rnd, sizeof(rnd)); vstart = VMALLOC_START + rnd; va_1 = find_vmap_lowest_match(size, 1, vstart); va_2 = find_vmap_lowest_linear_match(size, 1, vstart); if (va_1 != va_2) pr_emerg("not lowest: t: 0x%p, l: 0x%p, v: 0x%lx ", va_1, va_2, vstart); } #endif |
68ad4a330 mm/vmalloc.c: kee... |
1242 1243 1244 1245 1246 1247 1248 1249 1250 1251 1252 1253 1254 1255 1256 1257 1258 1259 1260 1261 1262 1263 1264 1265 1266 1267 1268 1269 1270 1271 1272 1273 1274 1275 1276 1277 1278 1279 1280 |
enum fit_type { NOTHING_FIT = 0, FL_FIT_TYPE = 1, /* full fit */ LE_FIT_TYPE = 2, /* left edge fit */ RE_FIT_TYPE = 3, /* right edge fit */ NE_FIT_TYPE = 4 /* no edge fit */ }; static __always_inline enum fit_type classify_va_fit_type(struct vmap_area *va, unsigned long nva_start_addr, unsigned long size) { enum fit_type type; /* Check if it is within VA. */ if (nva_start_addr < va->va_start || nva_start_addr + size > va->va_end) return NOTHING_FIT; /* Now classify. */ if (va->va_start == nva_start_addr) { if (va->va_end == nva_start_addr + size) type = FL_FIT_TYPE; else type = LE_FIT_TYPE; } else if (va->va_end == nva_start_addr + size) { type = RE_FIT_TYPE; } else { type = NE_FIT_TYPE; } return type; } static __always_inline int adjust_va_to_fit_type(struct vmap_area *va, unsigned long nva_start_addr, unsigned long size, enum fit_type type) { |
2c9292336 mm/vmalloc.c: avo... |
1281 |
struct vmap_area *lva = NULL; |
68ad4a330 mm/vmalloc.c: kee... |
1282 1283 1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 1299 1300 1301 1302 1303 1304 1305 1306 1307 1308 1309 1310 1311 1312 1313 1314 1315 1316 1317 1318 |
if (type == FL_FIT_TYPE) { /* * No need to split VA, it fully fits. * * | | * V NVA V * |---------------| */ unlink_va(va, &free_vmap_area_root); kmem_cache_free(vmap_area_cachep, va); } else if (type == LE_FIT_TYPE) { /* * Split left edge of fit VA. * * | | * V NVA V R * |-------|-------| */ va->va_start += size; } else if (type == RE_FIT_TYPE) { /* * Split right edge of fit VA. * * | | * L V NVA V * |-------|-------| */ va->va_end = nva_start_addr; } else if (type == NE_FIT_TYPE) { /* * Split no edge of fit VA. * * | | * L V NVA V R * |---|-------|---| */ |
82dd23e84 mm/vmalloc.c: pre... |
1319 1320 1321 1322 1323 1324 1325 1326 1327 1328 1329 1330 1331 |
lva = __this_cpu_xchg(ne_fit_preload_node, NULL); if (unlikely(!lva)) { /* * For percpu allocator we do not do any pre-allocation * and leave it as it is. The reason is it most likely * never ends up with NE_FIT_TYPE splitting. In case of * percpu allocations offsets and sizes are aligned to * fixed align request, i.e. RE_FIT_TYPE and FL_FIT_TYPE * are its main fitting cases. * * There are a few exceptions though, as an example it is * a first allocation (early boot up) when we have "one" * big free space that has to be split. |
060650a2a mm/vmalloc: add m... |
1332 1333 1334 1335 1336 1337 1338 1339 1340 1341 1342 1343 1344 |
* * Also we can hit this path in case of regular "vmap" * allocations, if "this" current CPU was not preloaded. * See the comment in alloc_vmap_area() why. If so, then * GFP_NOWAIT is used instead to get an extra object for * split purpose. That is rare and most time does not * occur. * * What happens if an allocation gets failed. Basically, * an "overflow" path is triggered to purge lazily freed * areas to free some memory, then, the "retry" path is * triggered to repeat one more time. See more details * in alloc_vmap_area() function. |
82dd23e84 mm/vmalloc.c: pre... |
1345 1346 1347 1348 1349 |
*/ lva = kmem_cache_alloc(vmap_area_cachep, GFP_NOWAIT); if (!lva) return -1; } |
68ad4a330 mm/vmalloc.c: kee... |
1350 1351 1352 1353 1354 1355 1356 1357 1358 1359 1360 1361 1362 1363 1364 1365 1366 |
/* * Build the remainder. */ lva->va_start = va->va_start; lva->va_end = nva_start_addr; /* * Shrink this VA to remaining size. */ va->va_start = nva_start_addr + size; } else { return -1; } if (type != FL_FIT_TYPE) { augment_tree_propagate_from(va); |
2c9292336 mm/vmalloc.c: avo... |
1367 |
if (lva) /* type == NE_FIT_TYPE */ |
68ad4a330 mm/vmalloc.c: kee... |
1368 1369 1370 1371 1372 1373 1374 1375 1376 1377 1378 1379 1380 |
insert_vmap_area_augment(lva, &va->rb_node, &free_vmap_area_root, &free_vmap_area_list); } return 0; } /* * Returns a start address of the newly allocated area, if success. * Otherwise a vend is returned that indicates failure. */ static __always_inline unsigned long __alloc_vmap_area(unsigned long size, unsigned long align, |
cacca6baf mm/vmalloc.c: rem... |
1381 |
unsigned long vstart, unsigned long vend) |
68ad4a330 mm/vmalloc.c: kee... |
1382 1383 1384 1385 1386 1387 1388 1389 1390 1391 1392 1393 1394 1395 1396 1397 1398 1399 1400 1401 1402 1403 1404 1405 1406 1407 1408 1409 |
{ unsigned long nva_start_addr; struct vmap_area *va; enum fit_type type; int ret; va = find_vmap_lowest_match(size, align, vstart); if (unlikely(!va)) return vend; if (va->va_start > vstart) nva_start_addr = ALIGN(va->va_start, align); else nva_start_addr = ALIGN(vstart, align); /* Check the "vend" restriction. */ if (nva_start_addr + size > vend) return vend; /* Classify what we have found. */ type = classify_va_fit_type(va, nva_start_addr, size); if (WARN_ON_ONCE(type == NOTHING_FIT)) return vend; /* Update the free vmap_area. */ ret = adjust_va_to_fit_type(va, nva_start_addr, size, type); if (ret) return vend; |
a6cf4e0fe mm/vmap: add DEBU... |
1410 1411 1412 |
#if DEBUG_AUGMENT_LOWEST_MATCH_CHECK find_vmap_lowest_match_check(size); #endif |
68ad4a330 mm/vmalloc.c: kee... |
1413 1414 |
return nva_start_addr; } |
4da56b99d mm/vmap: Add a no... |
1415 |
|
db64fe022 mm: rewrite vmap ... |
1416 |
/* |
d98c9e83b kasan: fix crashe... |
1417 1418 1419 1420 1421 1422 1423 1424 1425 1426 1427 1428 1429 1430 1431 |
* Free a region of KVA allocated by alloc_vmap_area */ static void free_vmap_area(struct vmap_area *va) { /* * Remove from the busy tree/list. */ spin_lock(&vmap_area_lock); unlink_va(va, &vmap_area_root); spin_unlock(&vmap_area_lock); /* * Insert/Merge it back to the free tree/list. */ spin_lock(&free_vmap_area_lock); |
96e2db456 mm/vmalloc: rewor... |
1432 |
merge_or_add_vmap_area_augment(va, &free_vmap_area_root, &free_vmap_area_list); |
d98c9e83b kasan: fix crashe... |
1433 1434 |
spin_unlock(&free_vmap_area_lock); } |
187f8cc45 mm/vmalloc: refac... |
1435 1436 1437 1438 1439 1440 1441 1442 1443 1444 1445 1446 1447 1448 1449 1450 1451 1452 1453 1454 1455 1456 |
static inline void preload_this_cpu_lock(spinlock_t *lock, gfp_t gfp_mask, int node) { struct vmap_area *va = NULL; /* * Preload this CPU with one extra vmap_area object. It is used * when fit type of free area is NE_FIT_TYPE. It guarantees that * a CPU that does an allocation is preloaded. * * We do it in non-atomic context, thus it allows us to use more * permissive allocation masks to be more stable under low memory * condition and high memory pressure. */ if (!this_cpu_read(ne_fit_preload_node)) va = kmem_cache_alloc_node(vmap_area_cachep, gfp_mask, node); spin_lock(lock); if (va && __this_cpu_cmpxchg(ne_fit_preload_node, NULL, va)) kmem_cache_free(vmap_area_cachep, va); } |
d98c9e83b kasan: fix crashe... |
1457 |
/* |
db64fe022 mm: rewrite vmap ... |
1458 1459 1460 1461 1462 1463 1464 1465 |
* 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) { |
187f8cc45 mm/vmalloc: refac... |
1466 |
struct vmap_area *va; |
12e376a6f mm/vmalloc: remov... |
1467 |
unsigned long freed; |
1da177e4c Linux-2.6.12-rc2 |
1468 |
unsigned long addr; |
db64fe022 mm: rewrite vmap ... |
1469 |
int purged = 0; |
d98c9e83b kasan: fix crashe... |
1470 |
int ret; |
db64fe022 mm: rewrite vmap ... |
1471 |
|
7766970cc mm: vmap fix over... |
1472 |
BUG_ON(!size); |
891c49abf mm/vmalloc: use o... |
1473 |
BUG_ON(offset_in_page(size)); |
89699605f mm: vmap area cache |
1474 |
BUG_ON(!is_power_of_2(align)); |
db64fe022 mm: rewrite vmap ... |
1475 |
|
68ad4a330 mm/vmalloc.c: kee... |
1476 1477 |
if (unlikely(!vmap_initialized)) return ERR_PTR(-EBUSY); |
5803ed292 mm: mark all call... |
1478 |
might_sleep(); |
f07116d77 mm/vmalloc: respe... |
1479 |
gfp_mask = gfp_mask & GFP_RECLAIM_MASK; |
4da56b99d mm/vmap: Add a no... |
1480 |
|
f07116d77 mm/vmalloc: respe... |
1481 |
va = kmem_cache_alloc_node(vmap_area_cachep, gfp_mask, node); |
db64fe022 mm: rewrite vmap ... |
1482 1483 |
if (unlikely(!va)) return ERR_PTR(-ENOMEM); |
7f88f88f8 mm: kmemleak: avo... |
1484 1485 1486 1487 |
/* * Only scan the relevant parts containing pointers to other objects * to avoid false negatives. */ |
f07116d77 mm/vmalloc: respe... |
1488 |
kmemleak_scan_area(&va->rb_node, SIZE_MAX, gfp_mask); |
7f88f88f8 mm: kmemleak: avo... |
1489 |
|
db64fe022 mm: rewrite vmap ... |
1490 |
retry: |
187f8cc45 mm/vmalloc: refac... |
1491 1492 1493 |
preload_this_cpu_lock(&free_vmap_area_lock, gfp_mask, node); addr = __alloc_vmap_area(size, align, vstart, vend); spin_unlock(&free_vmap_area_lock); |
89699605f mm: vmap area cache |
1494 |
|
afd07389d mm/vmalloc.c: fix... |
1495 |
/* |
68ad4a330 mm/vmalloc.c: kee... |
1496 1497 |
* If an allocation fails, the "vend" address is * returned. Therefore trigger the overflow path. |
afd07389d mm/vmalloc.c: fix... |
1498 |
*/ |
68ad4a330 mm/vmalloc.c: kee... |
1499 |
if (unlikely(addr == vend)) |
89699605f mm: vmap area cache |
1500 |
goto overflow; |
db64fe022 mm: rewrite vmap ... |
1501 1502 1503 |
va->va_start = addr; va->va_end = addr + size; |
688fcbfc0 mm/vmalloc: modif... |
1504 |
va->vm = NULL; |
68ad4a330 mm/vmalloc.c: kee... |
1505 |
|
e36176be1 mm/vmalloc: rewor... |
1506 1507 |
spin_lock(&vmap_area_lock); insert_vmap_area(va, &vmap_area_root, &vmap_area_list); |
db64fe022 mm: rewrite vmap ... |
1508 |
spin_unlock(&vmap_area_lock); |
61e165578 mm/vmalloc.c: use... |
1509 |
BUG_ON(!IS_ALIGNED(va->va_start, align)); |
89699605f mm: vmap area cache |
1510 1511 |
BUG_ON(va->va_start < vstart); BUG_ON(va->va_end > vend); |
d98c9e83b kasan: fix crashe... |
1512 1513 1514 1515 1516 |
ret = kasan_populate_vmalloc(addr, size); if (ret) { free_vmap_area(va); return ERR_PTR(ret); } |
db64fe022 mm: rewrite vmap ... |
1517 |
return va; |
89699605f mm: vmap area cache |
1518 1519 |
overflow: |
89699605f mm: vmap area cache |
1520 1521 1522 1523 1524 |
if (!purged) { purge_vmap_area_lazy(); purged = 1; goto retry; } |
4da56b99d mm/vmap: Add a no... |
1525 |
|
12e376a6f mm/vmalloc: remov... |
1526 1527 1528 1529 1530 1531 |
freed = 0; blocking_notifier_call_chain(&vmap_notify_list, 0, &freed); if (freed > 0) { purged = 0; goto retry; |
4da56b99d mm/vmap: Add a no... |
1532 |
} |
03497d761 mm: Silence vmap(... |
1533 |
if (!(gfp_mask & __GFP_NOWARN) && printk_ratelimit()) |
756a025f0 mm: coalesce spli... |
1534 1535 1536 |
pr_warn("vmap allocation for size %lu failed: use vmalloc=<size> to increase size ", size); |
68ad4a330 mm/vmalloc.c: kee... |
1537 1538 |
kmem_cache_free(vmap_area_cachep, va); |
89699605f mm: vmap area cache |
1539 |
return ERR_PTR(-EBUSY); |
db64fe022 mm: rewrite vmap ... |
1540 |
} |
4da56b99d mm/vmap: Add a no... |
1541 1542 1543 1544 1545 1546 1547 1548 1549 1550 1551 |
int register_vmap_purge_notifier(struct notifier_block *nb) { return blocking_notifier_chain_register(&vmap_notify_list, nb); } EXPORT_SYMBOL_GPL(register_vmap_purge_notifier); int unregister_vmap_purge_notifier(struct notifier_block *nb) { return blocking_notifier_chain_unregister(&vmap_notify_list, nb); } EXPORT_SYMBOL_GPL(unregister_vmap_purge_notifier); |
e36176be1 mm/vmalloc: rewor... |
1552 |
/* |
db64fe022 mm: rewrite vmap ... |
1553 1554 1555 1556 1557 1558 1559 1560 1561 1562 1563 1564 1565 1566 1567 1568 1569 1570 1571 1572 1573 1574 1575 |
* lazy_max_pages is the maximum amount of virtual address space we gather up * before attempting to purge with a TLB flush. * * There is a tradeoff here: a larger number will cover more kernel page tables * and take slightly longer to purge, but it will linearly reduce the number of * global TLB flushes that must be performed. It would seem natural to scale * this number up linearly with the number of CPUs (because vmapping activity * could also scale linearly with the number of CPUs), however it is likely * that in practice, workloads might be constrained in other ways that mean * vmap activity will not scale linearly with CPUs. Also, I want to be * conservative and not introduce a big latency on huge systems, so go with * a less aggressive log scale. It will still be an improvement over the old * code, and it will be simple to change the scale factor if we find that it * becomes a problem on bigger systems. */ static unsigned long lazy_max_pages(void) { unsigned int log; log = fls(num_online_cpus()); return log * (32UL * 1024 * 1024 / PAGE_SIZE); } |
4d36e6f80 mm/vmalloc.c: con... |
1576 |
static atomic_long_t vmap_lazy_nr = ATOMIC_LONG_INIT(0); |
db64fe022 mm: rewrite vmap ... |
1577 |
|
0574ecd14 mm: refactor __pu... |
1578 |
/* |
f0953a1bb mm: fix typos in ... |
1579 |
* Serialize vmap purging. There is no actual critical section protected |
0574ecd14 mm: refactor __pu... |
1580 1581 1582 |
* by this look, but we want to avoid concurrent calls for performance * reasons and to make the pcpu_get_vm_areas more deterministic. */ |
f9e099776 mm: turn vmap_pur... |
1583 |
static DEFINE_MUTEX(vmap_purge_lock); |
0574ecd14 mm: refactor __pu... |
1584 |
|
02b709df8 mm: purge fragmen... |
1585 1586 |
/* for per-CPU blocks */ static void purge_fragmented_blocks_allcpus(void); |
5da96bdd9 mm/vmalloc: inclu... |
1587 |
#ifdef CONFIG_X86_64 |
db64fe022 mm: rewrite vmap ... |
1588 |
/* |
3ee48b6af mm, x86: Saving v... |
1589 1590 1591 1592 1593 |
* called before a call to iounmap() if the caller wants vm_area_struct's * immediately freed. */ void set_iounmap_nonlazy(void) { |
4d36e6f80 mm/vmalloc.c: con... |
1594 |
atomic_long_set(&vmap_lazy_nr, lazy_max_pages()+1); |
3ee48b6af mm, x86: Saving v... |
1595 |
} |
5da96bdd9 mm/vmalloc: inclu... |
1596 |
#endif /* CONFIG_X86_64 */ |
3ee48b6af mm, x86: Saving v... |
1597 1598 |
/* |
db64fe022 mm: rewrite vmap ... |
1599 |
* Purges all lazily-freed vmap areas. |
db64fe022 mm: rewrite vmap ... |
1600 |
*/ |
0574ecd14 mm: refactor __pu... |
1601 |
static bool __purge_vmap_area_lazy(unsigned long start, unsigned long end) |
db64fe022 mm: rewrite vmap ... |
1602 |
{ |
4d36e6f80 mm/vmalloc.c: con... |
1603 |
unsigned long resched_threshold; |
96e2db456 mm/vmalloc: rewor... |
1604 1605 |
struct list_head local_pure_list; struct vmap_area *va, *n_va; |
db64fe022 mm: rewrite vmap ... |
1606 |
|
0574ecd14 mm: refactor __pu... |
1607 |
lockdep_assert_held(&vmap_purge_lock); |
02b709df8 mm: purge fragmen... |
1608 |
|
96e2db456 mm/vmalloc: rewor... |
1609 1610 1611 1612 1613 1614 |
spin_lock(&purge_vmap_area_lock); purge_vmap_area_root = RB_ROOT; list_replace_init(&purge_vmap_area_list, &local_pure_list); spin_unlock(&purge_vmap_area_lock); if (unlikely(list_empty(&local_pure_list))) |
68571be99 mm/vmalloc.c: add... |
1615 |
return false; |
96e2db456 mm/vmalloc: rewor... |
1616 1617 1618 1619 1620 1621 1622 |
start = min(start, list_first_entry(&local_pure_list, struct vmap_area, list)->va_start); end = max(end, list_last_entry(&local_pure_list, struct vmap_area, list)->va_end); |
db64fe022 mm: rewrite vmap ... |
1623 |
|
0574ecd14 mm: refactor __pu... |
1624 |
flush_tlb_kernel_range(start, end); |
4d36e6f80 mm/vmalloc.c: con... |
1625 |
resched_threshold = lazy_max_pages() << 1; |
db64fe022 mm: rewrite vmap ... |
1626 |
|
e36176be1 mm/vmalloc: rewor... |
1627 |
spin_lock(&free_vmap_area_lock); |
96e2db456 mm/vmalloc: rewor... |
1628 |
list_for_each_entry_safe(va, n_va, &local_pure_list, list) { |
4d36e6f80 mm/vmalloc.c: con... |
1629 |
unsigned long nr = (va->va_end - va->va_start) >> PAGE_SHIFT; |
3c5c3cfb9 kasan: support ba... |
1630 1631 |
unsigned long orig_start = va->va_start; unsigned long orig_end = va->va_end; |
763b218dd mm: add preempt p... |
1632 |
|
dd3b8353b mm/vmalloc: do no... |
1633 1634 1635 1636 1637 |
/* * Finally insert or merge lazily-freed area. It is * detached and there is no need to "unlink" it from * anything. */ |
96e2db456 mm/vmalloc: rewor... |
1638 1639 |
va = merge_or_add_vmap_area_augment(va, &free_vmap_area_root, &free_vmap_area_list); |
3c5c3cfb9 kasan: support ba... |
1640 |
|
9c801f61d mm/vmalloc.c: rem... |
1641 1642 |
if (!va) continue; |
3c5c3cfb9 kasan: support ba... |
1643 1644 1645 |
if (is_vmalloc_or_module_addr((void *)orig_start)) kasan_release_vmalloc(orig_start, orig_end, va->va_start, va->va_end); |
dd3b8353b mm/vmalloc: do no... |
1646 |
|
4d36e6f80 mm/vmalloc.c: con... |
1647 |
atomic_long_sub(nr, &vmap_lazy_nr); |
68571be99 mm/vmalloc.c: add... |
1648 |
|
4d36e6f80 mm/vmalloc.c: con... |
1649 |
if (atomic_long_read(&vmap_lazy_nr) < resched_threshold) |
e36176be1 mm/vmalloc: rewor... |
1650 |
cond_resched_lock(&free_vmap_area_lock); |
763b218dd mm: add preempt p... |
1651 |
} |
e36176be1 mm/vmalloc: rewor... |
1652 |
spin_unlock(&free_vmap_area_lock); |
0574ecd14 mm: refactor __pu... |
1653 |
return true; |
db64fe022 mm: rewrite vmap ... |
1654 1655 1656 |
} /* |
496850e5f mm: vmalloc failu... |
1657 1658 1659 1660 1661 |
* Kick off a purge of the outstanding lazy areas. Don't bother if somebody * is already purging. */ static void try_purge_vmap_area_lazy(void) { |
f9e099776 mm: turn vmap_pur... |
1662 |
if (mutex_trylock(&vmap_purge_lock)) { |
0574ecd14 mm: refactor __pu... |
1663 |
__purge_vmap_area_lazy(ULONG_MAX, 0); |
f9e099776 mm: turn vmap_pur... |
1664 |
mutex_unlock(&vmap_purge_lock); |
0574ecd14 mm: refactor __pu... |
1665 |
} |
496850e5f mm: vmalloc failu... |
1666 1667 1668 |
} /* |
db64fe022 mm: rewrite vmap ... |
1669 1670 1671 1672 |
* Kick off a purge of the outstanding lazy areas. */ static void purge_vmap_area_lazy(void) { |
f9e099776 mm: turn vmap_pur... |
1673 |
mutex_lock(&vmap_purge_lock); |
0574ecd14 mm: refactor __pu... |
1674 1675 |
purge_fragmented_blocks_allcpus(); __purge_vmap_area_lazy(ULONG_MAX, 0); |
f9e099776 mm: turn vmap_pur... |
1676 |
mutex_unlock(&vmap_purge_lock); |
db64fe022 mm: rewrite vmap ... |
1677 1678 1679 |
} /* |
64141da58 vmalloc: eagerly ... |
1680 1681 1682 |
* 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 ... |
1683 |
*/ |
64141da58 vmalloc: eagerly ... |
1684 |
static void free_vmap_area_noflush(struct vmap_area *va) |
db64fe022 mm: rewrite vmap ... |
1685 |
{ |
4d36e6f80 mm/vmalloc.c: con... |
1686 |
unsigned long nr_lazy; |
80c4bd7a5 mm/vmalloc: keep ... |
1687 |
|
dd3b8353b mm/vmalloc: do no... |
1688 1689 1690 |
spin_lock(&vmap_area_lock); unlink_va(va, &vmap_area_root); spin_unlock(&vmap_area_lock); |
4d36e6f80 mm/vmalloc.c: con... |
1691 1692 |
nr_lazy = atomic_long_add_return((va->va_end - va->va_start) >> PAGE_SHIFT, &vmap_lazy_nr); |
80c4bd7a5 mm/vmalloc: keep ... |
1693 |
|
96e2db456 mm/vmalloc: rewor... |
1694 1695 1696 1697 1698 1699 1700 |
/* * Merge or place it to the purge tree/list. */ spin_lock(&purge_vmap_area_lock); merge_or_add_vmap_area(va, &purge_vmap_area_root, &purge_vmap_area_list); spin_unlock(&purge_vmap_area_lock); |
80c4bd7a5 mm/vmalloc: keep ... |
1701 |
|
96e2db456 mm/vmalloc: rewor... |
1702 |
/* After this point, we may free va at any time */ |
80c4bd7a5 mm/vmalloc: keep ... |
1703 |
if (unlikely(nr_lazy > lazy_max_pages())) |
496850e5f mm: vmalloc failu... |
1704 |
try_purge_vmap_area_lazy(); |
db64fe022 mm: rewrite vmap ... |
1705 |
} |
b29acbdcf mm: vmalloc fix l... |
1706 1707 1708 1709 1710 1711 |
/* * 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); |
4ad0ae8c6 mm/vmalloc: remov... |
1712 |
vunmap_range_noflush(va->va_start, va->va_end); |
8e57f8acb mm, debug_pageall... |
1713 |
if (debug_pagealloc_enabled_static()) |
82a2e924f mm: vmalloc: clea... |
1714 |
flush_tlb_kernel_range(va->va_start, va->va_end); |
c8eef01e2 mm: remove free_u... |
1715 |
free_vmap_area_noflush(va); |
b29acbdcf mm: vmalloc fix l... |
1716 |
} |
db64fe022 mm: rewrite vmap ... |
1717 1718 1719 1720 1721 1722 1723 1724 1725 1726 |
static struct vmap_area *find_vmap_area(unsigned long addr) { struct vmap_area *va; spin_lock(&vmap_area_lock); va = __find_vmap_area(addr); spin_unlock(&vmap_area_lock); return va; } |
db64fe022 mm: rewrite vmap ... |
1727 1728 1729 1730 1731 1732 1733 1734 1735 1736 1737 1738 1739 1740 1741 1742 1743 1744 1745 1746 1747 1748 1749 |
/*** 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... |
1750 1751 1752 1753 |
#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 ... |
1754 1755 1756 1757 1758 1759 |
#define VMAP_BLOCK_SIZE (VMAP_BBMAP_BITS * PAGE_SIZE) struct vmap_block_queue { spinlock_t lock; struct list_head free; |
db64fe022 mm: rewrite vmap ... |
1760 1761 1762 1763 1764 |
}; struct vmap_block { spinlock_t lock; struct vmap_area *va; |
db64fe022 mm: rewrite vmap ... |
1765 |
unsigned long free, dirty; |
7d61bfe8f mm/vmalloc: get r... |
1766 |
unsigned long dirty_min, dirty_max; /*< dirty range */ |
de5604231 mm: percpu-vmap f... |
1767 1768 |
struct list_head free_list; struct rcu_head rcu_head; |
02b709df8 mm: purge fragmen... |
1769 |
struct list_head purge; |
db64fe022 mm: rewrite vmap ... |
1770 1771 1772 1773 1774 1775 |
}; /* Queue of free and dirty vmap blocks, for allocation and flushing purposes */ static DEFINE_PER_CPU(struct vmap_block_queue, vmap_block_queue); /* |
0f14599c6 vmalloc: convert ... |
1776 |
* XArray of vmap blocks, indexed by address, to quickly find a vmap block |
db64fe022 mm: rewrite vmap ... |
1777 1778 1779 |
* 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. */ |
0f14599c6 vmalloc: convert ... |
1780 |
static DEFINE_XARRAY(vmap_blocks); |
db64fe022 mm: rewrite vmap ... |
1781 1782 1783 1784 1785 1786 1787 1788 1789 1790 1791 1792 1793 1794 |
/* * 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... |
1795 1796 1797 1798 1799 1800 1801 1802 1803 1804 1805 1806 1807 1808 1809 |
static void *vmap_block_vaddr(unsigned long va_start, unsigned long pages_off) { unsigned long addr; addr = va_start + (pages_off << PAGE_SHIFT); BUG_ON(addr_to_vb_idx(addr) != addr_to_vb_idx(va_start)); return (void *)addr; } /** * new_vmap_block - allocates new vmap_block and occupies 2^order pages in this * block. Of course pages number can't exceed VMAP_BBMAP_BITS * @order: how many 2^order pages should be occupied in newly allocated block * @gfp_mask: flags for the page level allocator * |
a862f68a8 docs/core-api/mm:... |
1810 |
* Return: virtual address in a newly allocated block or ERR_PTR(-errno) |
cf725ce27 mm/vmalloc: occup... |
1811 1812 |
*/ static void *new_vmap_block(unsigned int order, gfp_t gfp_mask) |
db64fe022 mm: rewrite vmap ... |
1813 1814 1815 1816 1817 1818 |
{ struct vmap_block_queue *vbq; struct vmap_block *vb; struct vmap_area *va; unsigned long vb_idx; int node, err; |
cf725ce27 mm/vmalloc: occup... |
1819 |
void *vaddr; |
db64fe022 mm: rewrite vmap ... |
1820 1821 1822 1823 1824 1825 1826 1827 1828 1829 1830 |
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... |
1831 |
if (IS_ERR(va)) { |
db64fe022 mm: rewrite vmap ... |
1832 |
kfree(vb); |
e7d863407 mm: use ERR_CAST |
1833 |
return ERR_CAST(va); |
db64fe022 mm: rewrite vmap ... |
1834 |
} |
cf725ce27 mm/vmalloc: occup... |
1835 |
vaddr = vmap_block_vaddr(va->va_start, 0); |
db64fe022 mm: rewrite vmap ... |
1836 1837 |
spin_lock_init(&vb->lock); vb->va = va; |
cf725ce27 mm/vmalloc: occup... |
1838 1839 1840 |
/* 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 ... |
1841 |
vb->dirty = 0; |
7d61bfe8f mm/vmalloc: get r... |
1842 1843 |
vb->dirty_min = VMAP_BBMAP_BITS; vb->dirty_max = 0; |
db64fe022 mm: rewrite vmap ... |
1844 |
INIT_LIST_HEAD(&vb->free_list); |
db64fe022 mm: rewrite vmap ... |
1845 1846 |
vb_idx = addr_to_vb_idx(va->va_start); |
0f14599c6 vmalloc: convert ... |
1847 1848 1849 1850 1851 1852 |
err = xa_insert(&vmap_blocks, vb_idx, vb, gfp_mask); if (err) { kfree(vb); free_vmap_area(va); return ERR_PTR(err); } |
db64fe022 mm: rewrite vmap ... |
1853 1854 |
vbq = &get_cpu_var(vmap_block_queue); |
db64fe022 mm: rewrite vmap ... |
1855 |
spin_lock(&vbq->lock); |
68ac546f2 mm/vmalloc: fix p... |
1856 |
list_add_tail_rcu(&vb->free_list, &vbq->free); |
db64fe022 mm: rewrite vmap ... |
1857 |
spin_unlock(&vbq->lock); |
3f04ba859 vmalloc: fix use ... |
1858 |
put_cpu_var(vmap_block_queue); |
db64fe022 mm: rewrite vmap ... |
1859 |
|
cf725ce27 mm/vmalloc: occup... |
1860 |
return vaddr; |
db64fe022 mm: rewrite vmap ... |
1861 |
} |
db64fe022 mm: rewrite vmap ... |
1862 1863 1864 |
static void free_vmap_block(struct vmap_block *vb) { struct vmap_block *tmp; |
db64fe022 mm: rewrite vmap ... |
1865 |
|
0f14599c6 vmalloc: convert ... |
1866 |
tmp = xa_erase(&vmap_blocks, addr_to_vb_idx(vb->va->va_start)); |
db64fe022 mm: rewrite vmap ... |
1867 |
BUG_ON(tmp != vb); |
64141da58 vmalloc: eagerly ... |
1868 |
free_vmap_area_noflush(vb->va); |
22a3c7d18 vmalloc,rcu: Conv... |
1869 |
kfree_rcu(vb, rcu_head); |
db64fe022 mm: rewrite vmap ... |
1870 |
} |
02b709df8 mm: purge fragmen... |
1871 1872 1873 1874 1875 1876 1877 1878 1879 1880 1881 1882 1883 1884 1885 1886 1887 |
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... |
1888 1889 |
vb->dirty_min = 0; vb->dirty_max = VMAP_BBMAP_BITS; |
02b709df8 mm: purge fragmen... |
1890 1891 1892 1893 1894 1895 1896 1897 1898 1899 1900 1901 1902 1903 1904 |
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... |
1905 1906 1907 1908 1909 1910 1911 |
static void purge_fragmented_blocks_allcpus(void) { int cpu; for_each_possible_cpu(cpu) purge_fragmented_blocks(cpu); } |
db64fe022 mm: rewrite vmap ... |
1912 1913 1914 1915 |
static void *vb_alloc(unsigned long size, gfp_t gfp_mask) { struct vmap_block_queue *vbq; struct vmap_block *vb; |
cf725ce27 mm/vmalloc: occup... |
1916 |
void *vaddr = NULL; |
db64fe022 mm: rewrite vmap ... |
1917 |
unsigned int order; |
891c49abf mm/vmalloc: use o... |
1918 |
BUG_ON(offset_in_page(size)); |
db64fe022 mm: rewrite vmap ... |
1919 |
BUG_ON(size > PAGE_SIZE*VMAP_MAX_ALLOC); |
aa91c4d89 mm: make vb_alloc... |
1920 1921 1922 1923 1924 1925 1926 1927 |
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 ... |
1928 |
order = get_order(size); |
db64fe022 mm: rewrite vmap ... |
1929 1930 1931 |
rcu_read_lock(); vbq = &get_cpu_var(vmap_block_queue); list_for_each_entry_rcu(vb, &vbq->free, free_list) { |
cf725ce27 mm/vmalloc: occup... |
1932 |
unsigned long pages_off; |
db64fe022 mm: rewrite vmap ... |
1933 1934 |
spin_lock(&vb->lock); |
cf725ce27 mm/vmalloc: occup... |
1935 1936 1937 1938 |
if (vb->free < (1UL << order)) { spin_unlock(&vb->lock); continue; } |
02b709df8 mm: purge fragmen... |
1939 |
|
cf725ce27 mm/vmalloc: occup... |
1940 1941 |
pages_off = VMAP_BBMAP_BITS - vb->free; vaddr = vmap_block_vaddr(vb->va->va_start, pages_off); |
02b709df8 mm: purge fragmen... |
1942 1943 1944 1945 1946 1947 |
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... |
1948 |
|
02b709df8 mm: purge fragmen... |
1949 1950 |
spin_unlock(&vb->lock); break; |
db64fe022 mm: rewrite vmap ... |
1951 |
} |
02b709df8 mm: purge fragmen... |
1952 |
|
3f04ba859 vmalloc: fix use ... |
1953 |
put_cpu_var(vmap_block_queue); |
db64fe022 mm: rewrite vmap ... |
1954 |
rcu_read_unlock(); |
cf725ce27 mm/vmalloc: occup... |
1955 1956 1957 |
/* Allocate new block if nothing was found */ if (!vaddr) vaddr = new_vmap_block(order, gfp_mask); |
db64fe022 mm: rewrite vmap ... |
1958 |
|
cf725ce27 mm/vmalloc: occup... |
1959 |
return vaddr; |
db64fe022 mm: rewrite vmap ... |
1960 |
} |
78a0e8c48 mm: pass addr as ... |
1961 |
static void vb_free(unsigned long addr, unsigned long size) |
db64fe022 mm: rewrite vmap ... |
1962 1963 |
{ unsigned long offset; |
db64fe022 mm: rewrite vmap ... |
1964 1965 |
unsigned int order; struct vmap_block *vb; |
891c49abf mm/vmalloc: use o... |
1966 |
BUG_ON(offset_in_page(size)); |
db64fe022 mm: rewrite vmap ... |
1967 |
BUG_ON(size > PAGE_SIZE*VMAP_MAX_ALLOC); |
b29acbdcf mm: vmalloc fix l... |
1968 |
|
78a0e8c48 mm: pass addr as ... |
1969 |
flush_cache_vunmap(addr, addr + size); |
b29acbdcf mm: vmalloc fix l... |
1970 |
|
db64fe022 mm: rewrite vmap ... |
1971 |
order = get_order(size); |
78a0e8c48 mm: pass addr as ... |
1972 |
offset = (addr & (VMAP_BLOCK_SIZE - 1)) >> PAGE_SHIFT; |
0f14599c6 vmalloc: convert ... |
1973 |
vb = xa_load(&vmap_blocks, addr_to_vb_idx(addr)); |
db64fe022 mm: rewrite vmap ... |
1974 |
|
4ad0ae8c6 mm/vmalloc: remov... |
1975 |
vunmap_range_noflush(addr, addr + size); |
64141da58 vmalloc: eagerly ... |
1976 |
|
8e57f8acb mm, debug_pageall... |
1977 |
if (debug_pagealloc_enabled_static()) |
78a0e8c48 mm: pass addr as ... |
1978 |
flush_tlb_kernel_range(addr, addr + size); |
82a2e924f mm: vmalloc: clea... |
1979 |
|
db64fe022 mm: rewrite vmap ... |
1980 |
spin_lock(&vb->lock); |
7d61bfe8f mm/vmalloc: get r... |
1981 1982 1983 1984 |
/* 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... |
1985 |
|
db64fe022 mm: rewrite vmap ... |
1986 1987 |
vb->dirty += 1UL << order; if (vb->dirty == VMAP_BBMAP_BITS) { |
de5604231 mm: percpu-vmap f... |
1988 |
BUG_ON(vb->free); |
db64fe022 mm: rewrite vmap ... |
1989 1990 1991 1992 1993 |
spin_unlock(&vb->lock); free_vmap_block(vb); } else spin_unlock(&vb->lock); } |
868b104d7 mm/vmalloc: Add f... |
1994 |
static void _vm_unmap_aliases(unsigned long start, unsigned long end, int flush) |
db64fe022 mm: rewrite vmap ... |
1995 |
{ |
db64fe022 mm: rewrite vmap ... |
1996 |
int cpu; |
db64fe022 mm: rewrite vmap ... |
1997 |
|
9b4633340 vmap: cope with v... |
1998 1999 |
if (unlikely(!vmap_initialized)) return; |
5803ed292 mm: mark all call... |
2000 |
might_sleep(); |
db64fe022 mm: rewrite vmap ... |
2001 2002 2003 2004 2005 2006 |
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 ... |
2007 |
spin_lock(&vb->lock); |
ad216c031 mm: vmalloc: prev... |
2008 |
if (vb->dirty && vb->dirty != VMAP_BBMAP_BITS) { |
7d61bfe8f mm/vmalloc: get r... |
2009 |
unsigned long va_start = vb->va->va_start; |
db64fe022 mm: rewrite vmap ... |
2010 |
unsigned long s, e; |
b136be5e0 mm, vmalloc: use ... |
2011 |
|
7d61bfe8f mm/vmalloc: get r... |
2012 2013 |
s = va_start + (vb->dirty_min << PAGE_SHIFT); e = va_start + (vb->dirty_max << PAGE_SHIFT); |
db64fe022 mm: rewrite vmap ... |
2014 |
|
7d61bfe8f mm/vmalloc: get r... |
2015 2016 |
start = min(s, start); end = max(e, end); |
db64fe022 mm: rewrite vmap ... |
2017 |
|
7d61bfe8f mm/vmalloc: get r... |
2018 |
flush = 1; |
db64fe022 mm: rewrite vmap ... |
2019 2020 2021 2022 2023 |
} spin_unlock(&vb->lock); } rcu_read_unlock(); } |
f9e099776 mm: turn vmap_pur... |
2024 |
mutex_lock(&vmap_purge_lock); |
0574ecd14 mm: refactor __pu... |
2025 2026 2027 |
purge_fragmented_blocks_allcpus(); if (!__purge_vmap_area_lazy(start, end) && flush) flush_tlb_kernel_range(start, end); |
f9e099776 mm: turn vmap_pur... |
2028 |
mutex_unlock(&vmap_purge_lock); |
db64fe022 mm: rewrite vmap ... |
2029 |
} |
868b104d7 mm/vmalloc: Add f... |
2030 2031 2032 2033 2034 2035 2036 2037 2038 2039 2040 2041 2042 2043 2044 2045 2046 2047 2048 2049 2050 |
/** * vm_unmap_aliases - unmap outstanding lazy aliases in the vmap layer * * The vmap/vmalloc layer lazily flushes kernel virtual mappings primarily * to amortize TLB flushing overheads. What this means is that any page you * have now, may, in a former life, have been mapped into kernel virtual * address by the vmap layer and so there might be some CPUs with TLB entries * still referencing that page (additional to the regular 1:1 kernel mapping). * * vm_unmap_aliases flushes all such lazy mappings. After it returns, we can * be sure that none of the pages we have control over will have any aliases * from the vmap layer. */ void vm_unmap_aliases(void) { unsigned long start = ULONG_MAX, end = 0; int flush = 0; _vm_unmap_aliases(start, end, flush); } |
db64fe022 mm: rewrite vmap ... |
2051 2052 2053 2054 2055 2056 2057 2058 2059 |
EXPORT_SYMBOL_GPL(vm_unmap_aliases); /** * vm_unmap_ram - unmap linear kernel address space set up by vm_map_ram * @mem: the pointer returned by vm_map_ram * @count: the count passed to that vm_map_ram call (cannot unmap partial) */ void vm_unmap_ram(const void *mem, unsigned int count) { |
65ee03c4b mm: fix overflow ... |
2060 |
unsigned long size = (unsigned long)count << PAGE_SHIFT; |
db64fe022 mm: rewrite vmap ... |
2061 |
unsigned long addr = (unsigned long)mem; |
9c3acf604 mm: remove free_u... |
2062 |
struct vmap_area *va; |
db64fe022 mm: rewrite vmap ... |
2063 |
|
5803ed292 mm: mark all call... |
2064 |
might_sleep(); |
db64fe022 mm: rewrite vmap ... |
2065 2066 2067 |
BUG_ON(!addr); BUG_ON(addr < VMALLOC_START); BUG_ON(addr > VMALLOC_END); |
a1c0b1a07 mm/vmalloc: use P... |
2068 |
BUG_ON(!PAGE_ALIGNED(addr)); |
db64fe022 mm: rewrite vmap ... |
2069 |
|
d98c9e83b kasan: fix crashe... |
2070 |
kasan_poison_vmalloc(mem, size); |
9c3acf604 mm: remove free_u... |
2071 |
if (likely(count <= VMAP_MAX_ALLOC)) { |
05e3ff950 mm: vmalloc: pass... |
2072 |
debug_check_no_locks_freed(mem, size); |
78a0e8c48 mm: pass addr as ... |
2073 |
vb_free(addr, size); |
9c3acf604 mm: remove free_u... |
2074 2075 2076 2077 2078 |
return; } va = find_vmap_area(addr); BUG_ON(!va); |
05e3ff950 mm: vmalloc: pass... |
2079 2080 |
debug_check_no_locks_freed((void *)va->va_start, (va->va_end - va->va_start)); |
9c3acf604 mm: remove free_u... |
2081 |
free_unmap_vmap_area(va); |
db64fe022 mm: rewrite vmap ... |
2082 2083 2084 2085 2086 2087 2088 2089 |
} 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 |
e99c97ade mm: fix kernel-do... |
2090 |
* |
364376383 mm/vmalloc.c: enh... |
2091 2092 2093 2094 2095 2096 |
* 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... |
2097 |
* Returns: a pointer to the address that has been mapped, or %NULL on failure |
db64fe022 mm: rewrite vmap ... |
2098 |
*/ |
d4efd79a8 mm: remove the pr... |
2099 |
void *vm_map_ram(struct page **pages, unsigned int count, int node) |
db64fe022 mm: rewrite vmap ... |
2100 |
{ |
65ee03c4b mm: fix overflow ... |
2101 |
unsigned long size = (unsigned long)count << PAGE_SHIFT; |
db64fe022 mm: rewrite vmap ... |
2102 2103 2104 2105 2106 2107 2108 2109 2110 2111 2112 2113 2114 2115 2116 2117 2118 2119 |
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; } |
d98c9e83b kasan: fix crashe... |
2120 2121 |
kasan_unpoison_vmalloc(mem, size); |
b67177ecd mm/vmalloc: remov... |
2122 2123 |
if (vmap_pages_range(addr, addr + size, PAGE_KERNEL, pages, PAGE_SHIFT) < 0) { |
db64fe022 mm: rewrite vmap ... |
2124 2125 2126 |
vm_unmap_ram(mem, count); return NULL; } |
b67177ecd mm/vmalloc: remov... |
2127 |
|
db64fe022 mm: rewrite vmap ... |
2128 2129 2130 |
return mem; } EXPORT_SYMBOL(vm_map_ram); |
4341fa454 mm, vmalloc: remo... |
2131 |
static struct vm_struct *vmlist __initdata; |
92eac1681 docs/mm: vmalloc:... |
2132 |
|
121e6f325 mm/vmalloc: hugep... |
2133 2134 2135 2136 2137 2138 2139 2140 2141 2142 2143 2144 2145 2146 2147 2148 2149 |
static inline unsigned int vm_area_page_order(struct vm_struct *vm) { #ifdef CONFIG_HAVE_ARCH_HUGE_VMALLOC return vm->page_order; #else return 0; #endif } static inline void set_vm_area_page_order(struct vm_struct *vm, unsigned int order) { #ifdef CONFIG_HAVE_ARCH_HUGE_VMALLOC vm->page_order = order; #else BUG_ON(order != 0); #endif } |
f0aa66179 vmalloc: implemen... |
2150 |
/** |
be9b7335e mm: add vm_area_a... |
2151 2152 2153 2154 2155 2156 2157 2158 2159 2160 2161 2162 2163 2164 2165 2166 2167 2168 2169 2170 2171 2172 2173 2174 2175 2176 |
* 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... |
2177 2178 |
* vm_area_register_early - register vmap area early during boot * @vm: vm_struct to register |
c0c0a2937 vmalloc: add @ali... |
2179 |
* @align: requested alignment |
f0aa66179 vmalloc: implemen... |
2180 2181 2182 2183 2184 2185 2186 2187 |
* * 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... |
2188 |
void __init vm_area_register_early(struct vm_struct *vm, size_t align) |
f0aa66179 vmalloc: implemen... |
2189 2190 |
{ static size_t vm_init_off __initdata; |
c0c0a2937 vmalloc: add @ali... |
2191 2192 2193 2194 |
unsigned long addr; addr = ALIGN(VMALLOC_START + vm_init_off, align); vm_init_off = PFN_ALIGN(addr + vm->size) - VMALLOC_START; |
f0aa66179 vmalloc: implemen... |
2195 |
|
c0c0a2937 vmalloc: add @ali... |
2196 |
vm->addr = (void *)addr; |
f0aa66179 vmalloc: implemen... |
2197 |
|
be9b7335e mm: add vm_area_a... |
2198 |
vm_area_add_early(vm); |
f0aa66179 vmalloc: implemen... |
2199 |
} |
68ad4a330 mm/vmalloc.c: kee... |
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 |
static void vmap_init_free_space(void) { unsigned long vmap_start = 1; const unsigned long vmap_end = ULONG_MAX; struct vmap_area *busy, *free; /* * B F B B B F * -|-----|.....|-----|-----|-----|.....|- * | The KVA space | * |<--------------------------------->| */ list_for_each_entry(busy, &vmap_area_list, list) { if (busy->va_start - vmap_start > 0) { free = kmem_cache_zalloc(vmap_area_cachep, GFP_NOWAIT); if (!WARN_ON_ONCE(!free)) { free->va_start = vmap_start; free->va_end = busy->va_start; insert_vmap_area_augment(free, NULL, &free_vmap_area_root, &free_vmap_area_list); } } vmap_start = busy->va_end; } if (vmap_end - vmap_start > 0) { free = kmem_cache_zalloc(vmap_area_cachep, GFP_NOWAIT); if (!WARN_ON_ONCE(!free)) { free->va_start = vmap_start; free->va_end = vmap_end; insert_vmap_area_augment(free, NULL, &free_vmap_area_root, &free_vmap_area_list); } } } |
db64fe022 mm: rewrite vmap ... |
2240 2241 |
void __init vmalloc_init(void) { |
822c18f2e alpha: fix vmallo... |
2242 2243 |
struct vmap_area *va; struct vm_struct *tmp; |
db64fe022 mm: rewrite vmap ... |
2244 |
int i; |
68ad4a330 mm/vmalloc.c: kee... |
2245 2246 2247 2248 |
/* * Create the cache for vmap_area objects. */ vmap_area_cachep = KMEM_CACHE(vmap_area, SLAB_PANIC); |
db64fe022 mm: rewrite vmap ... |
2249 2250 |
for_each_possible_cpu(i) { struct vmap_block_queue *vbq; |
32fcfd407 make vfree() safe... |
2251 |
struct vfree_deferred *p; |
db64fe022 mm: rewrite vmap ... |
2252 2253 2254 2255 |
vbq = &per_cpu(vmap_block_queue, i); spin_lock_init(&vbq->lock); INIT_LIST_HEAD(&vbq->free); |
32fcfd407 make vfree() safe... |
2256 2257 2258 |
p = &per_cpu(vfree_deferred, i); init_llist_head(&p->list); INIT_WORK(&p->wq, free_work); |
db64fe022 mm: rewrite vmap ... |
2259 |
} |
9b4633340 vmap: cope with v... |
2260 |
|
822c18f2e alpha: fix vmallo... |
2261 2262 |
/* Import existing vmlist entries. */ for (tmp = vmlist; tmp; tmp = tmp->next) { |
68ad4a330 mm/vmalloc.c: kee... |
2263 2264 2265 |
va = kmem_cache_zalloc(vmap_area_cachep, GFP_NOWAIT); if (WARN_ON_ONCE(!va)) continue; |
822c18f2e alpha: fix vmallo... |
2266 2267 |
va->va_start = (unsigned long)tmp->addr; va->va_end = va->va_start + tmp->size; |
dbda591d9 mm: fix faulty in... |
2268 |
va->vm = tmp; |
68ad4a330 mm/vmalloc.c: kee... |
2269 |
insert_vmap_area(va, &vmap_area_root, &vmap_area_list); |
822c18f2e alpha: fix vmallo... |
2270 |
} |
ca23e405e vmalloc: implemen... |
2271 |
|
68ad4a330 mm/vmalloc.c: kee... |
2272 2273 2274 2275 |
/* * Now we can initialize a free vmap space. */ vmap_init_free_space(); |
9b4633340 vmap: cope with v... |
2276 |
vmap_initialized = true; |
db64fe022 mm: rewrite vmap ... |
2277 |
} |
e36176be1 mm/vmalloc: rewor... |
2278 2279 |
static inline void setup_vmalloc_vm_locked(struct vm_struct *vm, struct vmap_area *va, unsigned long flags, const void *caller) |
cf88c7900 vmalloc: separate... |
2280 |
{ |
cf88c7900 vmalloc: separate... |
2281 2282 2283 2284 |
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... |
2285 |
va->vm = vm; |
e36176be1 mm/vmalloc: rewor... |
2286 2287 2288 2289 2290 2291 2292 |
} static void setup_vmalloc_vm(struct vm_struct *vm, struct vmap_area *va, unsigned long flags, const void *caller) { spin_lock(&vmap_area_lock); setup_vmalloc_vm_locked(vm, va, flags, caller); |
c69480ade mm, vmalloc: prot... |
2293 |
spin_unlock(&vmap_area_lock); |
f5252e009 mm: avoid null po... |
2294 |
} |
cf88c7900 vmalloc: separate... |
2295 |
|
20fc02b47 mm/vmalloc.c: ren... |
2296 |
static void clear_vm_uninitialized_flag(struct vm_struct *vm) |
f5252e009 mm: avoid null po... |
2297 |
{ |
d4033afdf mm, vmalloc: iter... |
2298 |
/* |
20fc02b47 mm/vmalloc.c: ren... |
2299 |
* Before removing VM_UNINITIALIZED, |
d4033afdf mm, vmalloc: iter... |
2300 2301 2302 2303 |
* we should make sure that vm has proper values. * Pair with smp_rmb() in show_numa_info(). */ smp_wmb(); |
20fc02b47 mm/vmalloc.c: ren... |
2304 |
vm->flags &= ~VM_UNINITIALIZED; |
cf88c7900 vmalloc: separate... |
2305 |
} |
db64fe022 mm: rewrite vmap ... |
2306 |
static struct vm_struct *__get_vm_area_node(unsigned long size, |
7ca3027b7 mm/vmalloc: unbre... |
2307 2308 2309 |
unsigned long align, unsigned long shift, unsigned long flags, unsigned long start, unsigned long end, int node, gfp_t gfp_mask, const void *caller) |
db64fe022 mm: rewrite vmap ... |
2310 |
{ |
0006526d7 mm/vmalloc.c: rem... |
2311 |
struct vmap_area *va; |
db64fe022 mm: rewrite vmap ... |
2312 |
struct vm_struct *area; |
d98c9e83b kasan: fix crashe... |
2313 |
unsigned long requested_size = size; |
1da177e4c Linux-2.6.12-rc2 |
2314 |
|
52fd24ca1 [PATCH] __vmalloc... |
2315 |
BUG_ON(in_interrupt()); |
7ca3027b7 mm/vmalloc: unbre... |
2316 |
size = ALIGN(size, 1ul << shift); |
31be83095 [PATCH] Fix stran... |
2317 2318 |
if (unlikely(!size)) return NULL; |
1da177e4c Linux-2.6.12-rc2 |
2319 |
|
252e5c6e2 mm/vmalloc.c: fix... |
2320 2321 2322 |
if (flags & VM_IOREMAP) align = 1ul << clamp_t(int, get_count_order_long(size), PAGE_SHIFT, IOREMAP_MAX_ORDER); |
cf88c7900 vmalloc: separate... |
2323 |
area = kzalloc_node(sizeof(*area), gfp_mask & GFP_RECLAIM_MASK, node); |
1da177e4c Linux-2.6.12-rc2 |
2324 2325 |
if (unlikely(!area)) return NULL; |
71394fe50 mm: vmalloc: add ... |
2326 2327 |
if (!(flags & VM_NO_GUARD)) size += PAGE_SIZE; |
1da177e4c Linux-2.6.12-rc2 |
2328 |
|
db64fe022 mm: rewrite vmap ... |
2329 2330 2331 2332 |
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 |
2333 |
} |
1da177e4c Linux-2.6.12-rc2 |
2334 |
|
d98c9e83b kasan: fix crashe... |
2335 |
kasan_unpoison_vmalloc((void *)va->va_start, requested_size); |
f5252e009 mm: avoid null po... |
2336 |
|
d98c9e83b kasan: fix crashe... |
2337 |
setup_vmalloc_vm(area, va, flags, caller); |
3c5c3cfb9 kasan: support ba... |
2338 |
|
1da177e4c Linux-2.6.12-rc2 |
2339 |
return area; |
1da177e4c Linux-2.6.12-rc2 |
2340 |
} |
c29686129 vmalloc: add __ge... |
2341 2342 |
struct vm_struct *__get_vm_area_caller(unsigned long size, unsigned long flags, unsigned long start, unsigned long end, |
5e6cafc83 mm: vmalloc: use ... |
2343 |
const void *caller) |
c29686129 vmalloc: add __ge... |
2344 |
{ |
7ca3027b7 mm/vmalloc: unbre... |
2345 2346 |
return __get_vm_area_node(size, 1, PAGE_SHIFT, flags, start, end, NUMA_NO_NODE, GFP_KERNEL, caller); |
c29686129 vmalloc: add __ge... |
2347 |
} |
1102b657d mm: vmalloc: Expo... |
2348 |
EXPORT_SYMBOL_GPL(__get_vm_area_caller); |
c29686129 vmalloc: add __ge... |
2349 |
|
1da177e4c Linux-2.6.12-rc2 |
2350 |
/** |
92eac1681 docs/mm: vmalloc:... |
2351 2352 2353 |
* get_vm_area - reserve a contiguous kernel virtual area * @size: size of the area * @flags: %VM_IOREMAP for I/O mappings or VM_ALLOC |
1da177e4c Linux-2.6.12-rc2 |
2354 |
* |
92eac1681 docs/mm: vmalloc:... |
2355 2356 2357 |
* Search an area of @size in the kernel virtual mapping area, * and reserved it for out purposes. Returns the area descriptor * on success or %NULL on failure. |
a862f68a8 docs/core-api/mm:... |
2358 2359 |
* * Return: the area descriptor on success or %NULL on failure. |
1da177e4c Linux-2.6.12-rc2 |
2360 2361 2362 |
*/ struct vm_struct *get_vm_area(unsigned long size, unsigned long flags) { |
7ca3027b7 mm/vmalloc: unbre... |
2363 2364 |
return __get_vm_area_node(size, 1, PAGE_SHIFT, flags, VMALLOC_START, VMALLOC_END, |
00ef2d2f8 mm: use NUMA_NO_NODE |
2365 2366 |
NUMA_NO_NODE, GFP_KERNEL, __builtin_return_address(0)); |
230169693 vmallocinfo: add ... |
2367 2368 2369 |
} struct vm_struct *get_vm_area_caller(unsigned long size, unsigned long flags, |
5e6cafc83 mm: vmalloc: use ... |
2370 |
const void *caller) |
230169693 vmallocinfo: add ... |
2371 |
{ |
7ca3027b7 mm/vmalloc: unbre... |
2372 2373 |
return __get_vm_area_node(size, 1, PAGE_SHIFT, flags, VMALLOC_START, VMALLOC_END, |
00ef2d2f8 mm: use NUMA_NO_NODE |
2374 |
NUMA_NO_NODE, GFP_KERNEL, caller); |
1da177e4c Linux-2.6.12-rc2 |
2375 |
} |
e9da6e990 ARM: dma-mapping:... |
2376 |
/** |
92eac1681 docs/mm: vmalloc:... |
2377 2378 |
* find_vm_area - find a continuous kernel virtual area * @addr: base address |
e9da6e990 ARM: dma-mapping:... |
2379 |
* |
92eac1681 docs/mm: vmalloc:... |
2380 2381 2382 |
* Search for the kernel VM area starting at @addr, and return it. * It is up to the caller to do all required locking to keep the returned * pointer valid. |
a862f68a8 docs/core-api/mm:... |
2383 |
* |
74640617e mm/vmalloc.c: fix... |
2384 |
* Return: the area descriptor on success or %NULL on failure. |
e9da6e990 ARM: dma-mapping:... |
2385 2386 |
*/ struct vm_struct *find_vm_area(const void *addr) |
833423143 [PATCH] mm: intro... |
2387 |
{ |
db64fe022 mm: rewrite vmap ... |
2388 |
struct vmap_area *va; |
833423143 [PATCH] mm: intro... |
2389 |
|
db64fe022 mm: rewrite vmap ... |
2390 |
va = find_vmap_area((unsigned long)addr); |
688fcbfc0 mm/vmalloc: modif... |
2391 2392 |
if (!va) return NULL; |
1da177e4c Linux-2.6.12-rc2 |
2393 |
|
688fcbfc0 mm/vmalloc: modif... |
2394 |
return va->vm; |
1da177e4c Linux-2.6.12-rc2 |
2395 |
} |
7856dfeb2 [PATCH] x86_64: F... |
2396 |
/** |
92eac1681 docs/mm: vmalloc:... |
2397 2398 |
* remove_vm_area - find and remove a continuous kernel virtual area * @addr: base address |
7856dfeb2 [PATCH] x86_64: F... |
2399 |
* |
92eac1681 docs/mm: vmalloc:... |
2400 2401 2402 |
* Search for the kernel VM area starting at @addr, and remove it. * This function returns the found VM area, but using it is NOT safe * on SMP machines, except for its size or flags. |
a862f68a8 docs/core-api/mm:... |
2403 |
* |
74640617e mm/vmalloc.c: fix... |
2404 |
* Return: the area descriptor on success or %NULL on failure. |
7856dfeb2 [PATCH] x86_64: F... |
2405 |
*/ |
b3bdda02a vmalloc: add cons... |
2406 |
struct vm_struct *remove_vm_area(const void *addr) |
7856dfeb2 [PATCH] x86_64: F... |
2407 |
{ |
db64fe022 mm: rewrite vmap ... |
2408 |
struct vmap_area *va; |
5803ed292 mm: mark all call... |
2409 |
might_sleep(); |
dd3b8353b mm/vmalloc: do no... |
2410 2411 |
spin_lock(&vmap_area_lock); va = __find_vmap_area((unsigned long)addr); |
688fcbfc0 mm/vmalloc: modif... |
2412 |
if (va && va->vm) { |
db1aecafe mm/vmalloc.c: cha... |
2413 |
struct vm_struct *vm = va->vm; |
f5252e009 mm: avoid null po... |
2414 |
|
c69480ade mm, vmalloc: prot... |
2415 |
va->vm = NULL; |
c69480ade mm, vmalloc: prot... |
2416 |
spin_unlock(&vmap_area_lock); |
a5af5aa8b kasan, module, vm... |
2417 |
kasan_free_shadow(vm); |
dd32c2799 vmalloc: unmap vm... |
2418 |
free_unmap_vmap_area(va); |
dd32c2799 vmalloc: unmap vm... |
2419 |
|
db64fe022 mm: rewrite vmap ... |
2420 2421 |
return vm; } |
dd3b8353b mm/vmalloc: do no... |
2422 2423 |
spin_unlock(&vmap_area_lock); |
db64fe022 mm: rewrite vmap ... |
2424 |
return NULL; |
7856dfeb2 [PATCH] x86_64: F... |
2425 |
} |
868b104d7 mm/vmalloc: Add f... |
2426 2427 2428 2429 |
static inline void set_area_direct_map(const struct vm_struct *area, int (*set_direct_map)(struct page *page)) { int i; |
121e6f325 mm/vmalloc: hugep... |
2430 |
/* HUGE_VMALLOC passes small pages to set_direct_map */ |
868b104d7 mm/vmalloc: Add f... |
2431 2432 2433 2434 2435 2436 2437 2438 |
for (i = 0; i < area->nr_pages; i++) if (page_address(area->pages[i])) set_direct_map(area->pages[i]); } /* Handle removing and resetting vm mappings related to the vm_struct. */ static void vm_remove_mappings(struct vm_struct *area, int deallocate_pages) { |
868b104d7 mm/vmalloc: Add f... |
2439 |
unsigned long start = ULONG_MAX, end = 0; |
121e6f325 mm/vmalloc: hugep... |
2440 |
unsigned int page_order = vm_area_page_order(area); |
868b104d7 mm/vmalloc: Add f... |
2441 |
int flush_reset = area->flags & VM_FLUSH_RESET_PERMS; |
31e67340c mm/vmalloc: Avoid... |
2442 |
int flush_dmap = 0; |
868b104d7 mm/vmalloc: Add f... |
2443 |
int i; |
868b104d7 mm/vmalloc: Add f... |
2444 2445 2446 2447 2448 2449 2450 2451 2452 2453 2454 2455 2456 2457 2458 2459 2460 2461 2462 2463 |
remove_vm_area(area->addr); /* If this is not VM_FLUSH_RESET_PERMS memory, no need for the below. */ if (!flush_reset) return; /* * If not deallocating pages, just do the flush of the VM area and * return. */ if (!deallocate_pages) { vm_unmap_aliases(); return; } /* * If execution gets here, flush the vm mapping and reset the direct * map. Find the start and end range of the direct mappings to make sure * the vm_unmap_aliases() flush includes the direct map. */ |
121e6f325 mm/vmalloc: hugep... |
2464 |
for (i = 0; i < area->nr_pages; i += 1U << page_order) { |
8e41f8726 mm/vmalloc: Fix c... |
2465 2466 |
unsigned long addr = (unsigned long)page_address(area->pages[i]); if (addr) { |
121e6f325 mm/vmalloc: hugep... |
2467 2468 2469 |
unsigned long page_size; page_size = PAGE_SIZE << page_order; |
868b104d7 mm/vmalloc: Add f... |
2470 |
start = min(addr, start); |
121e6f325 mm/vmalloc: hugep... |
2471 |
end = max(addr + page_size, end); |
31e67340c mm/vmalloc: Avoid... |
2472 |
flush_dmap = 1; |
868b104d7 mm/vmalloc: Add f... |
2473 2474 2475 2476 2477 2478 2479 2480 2481 |
} } /* * Set direct map to something invalid so that it won't be cached if * there are any accesses after the TLB flush, then flush the TLB and * reset the direct map permissions to the default. */ set_area_direct_map(area, set_direct_map_invalid_noflush); |
31e67340c mm/vmalloc: Avoid... |
2482 |
_vm_unmap_aliases(start, end, flush_dmap); |
868b104d7 mm/vmalloc: Add f... |
2483 2484 |
set_area_direct_map(area, set_direct_map_default_noflush); } |
b3bdda02a vmalloc: add cons... |
2485 |
static void __vunmap(const void *addr, int deallocate_pages) |
1da177e4c Linux-2.6.12-rc2 |
2486 2487 2488 2489 2490 |
{ struct vm_struct *area; if (!addr) return; |
e69e9d4ae vmalloc: introduc... |
2491 2492 |
if (WARN(!PAGE_ALIGNED(addr), "Trying to vfree() bad address (%p) ", |
ab15d9b4c mm/vmalloc.c: unb... |
2493 |
addr)) |
1da177e4c Linux-2.6.12-rc2 |
2494 |
return; |
1da177e4c Linux-2.6.12-rc2 |
2495 |
|
6ade20327 mm/vmalloc.c: don... |
2496 |
area = find_vm_area(addr); |
1da177e4c Linux-2.6.12-rc2 |
2497 |
if (unlikely(!area)) { |
4c8573e25 Use WARN() in mm/... |
2498 2499 |
WARN(1, KERN_ERR "Trying to vfree() nonexistent vm area (%p) ", |
1da177e4c Linux-2.6.12-rc2 |
2500 |
addr); |
1da177e4c Linux-2.6.12-rc2 |
2501 2502 |
return; } |
05e3ff950 mm: vmalloc: pass... |
2503 2504 |
debug_check_no_locks_freed(area->addr, get_vm_area_size(area)); debug_check_no_obj_freed(area->addr, get_vm_area_size(area)); |
9a11b49a8 [PATCH] lockdep: ... |
2505 |
|
c041098c6 mm/vmalloc.c: fix... |
2506 |
kasan_poison_vmalloc(area->addr, get_vm_area_size(area)); |
3c5c3cfb9 kasan: support ba... |
2507 |
|
868b104d7 mm/vmalloc: Add f... |
2508 |
vm_remove_mappings(area, deallocate_pages); |
1da177e4c Linux-2.6.12-rc2 |
2509 |
if (deallocate_pages) { |
121e6f325 mm/vmalloc: hugep... |
2510 |
unsigned int page_order = vm_area_page_order(area); |
1da177e4c Linux-2.6.12-rc2 |
2511 |
int i; |
121e6f325 mm/vmalloc: hugep... |
2512 |
for (i = 0; i < area->nr_pages; i += 1U << page_order) { |
bf53d6f8f vmalloc: clean up... |
2513 2514 2515 |
struct page *page = area->pages[i]; BUG_ON(!page); |
121e6f325 mm/vmalloc: hugep... |
2516 |
__free_pages(page, page_order); |
a850e932d mm: vmalloc: add ... |
2517 |
cond_resched(); |
1da177e4c Linux-2.6.12-rc2 |
2518 |
} |
97105f0ab mm: vmalloc: show... |
2519 |
atomic_long_sub(area->nr_pages, &nr_vmalloc_pages); |
1da177e4c Linux-2.6.12-rc2 |
2520 |
|
244d63ee3 mm, vmalloc: remo... |
2521 |
kvfree(area->pages); |
1da177e4c Linux-2.6.12-rc2 |
2522 2523 2524 |
} kfree(area); |
1da177e4c Linux-2.6.12-rc2 |
2525 |
} |
bf22e37a6 mm: add vfree_ato... |
2526 2527 2528 2529 2530 2531 2532 |
static inline void __vfree_deferred(const void *addr) { /* * Use raw_cpu_ptr() because this can be called from preemptible * context. Preemption is absolutely fine here, because the llist_add() * implementation is lockless, so it works even if we are adding to |
73221d888 mm/vmalloc: fix a... |
2533 |
* another cpu's list. schedule_work() should be fine with this too. |
bf22e37a6 mm: add vfree_ato... |
2534 2535 2536 2537 2538 2539 2540 2541 |
*/ struct vfree_deferred *p = raw_cpu_ptr(&vfree_deferred); if (llist_add((struct llist_node *)addr, &p->list)) schedule_work(&p->wq); } /** |
92eac1681 docs/mm: vmalloc:... |
2542 2543 |
* vfree_atomic - release memory allocated by vmalloc() * @addr: memory base address |
bf22e37a6 mm: add vfree_ato... |
2544 |
* |
92eac1681 docs/mm: vmalloc:... |
2545 2546 |
* This one is just like vfree() but can be called in any atomic context * except NMIs. |
bf22e37a6 mm: add vfree_ato... |
2547 2548 2549 2550 2551 2552 2553 2554 2555 2556 2557 |
*/ void vfree_atomic(const void *addr) { BUG_ON(in_nmi()); kmemleak_free(addr); if (!addr) return; __vfree_deferred(addr); } |
c67dc6247 mm/vmalloc: do no... |
2558 2559 2560 2561 2562 2563 2564 |
static void __vfree(const void *addr) { if (unlikely(in_interrupt())) __vfree_deferred(addr); else __vunmap(addr, 1); } |
1da177e4c Linux-2.6.12-rc2 |
2565 |
/** |
fa307474c mm: update the do... |
2566 2567 |
* vfree - Release memory allocated by vmalloc() * @addr: Memory base address |
1da177e4c Linux-2.6.12-rc2 |
2568 |
* |
fa307474c mm: update the do... |
2569 2570 2571 2572 |
* Free the virtually continuous memory area starting at @addr, as obtained * from one of the vmalloc() family of APIs. This will usually also free the * physical memory underlying the virtual allocation, but that memory is * reference counted, so it will not be freed until the last user goes away. |
1da177e4c Linux-2.6.12-rc2 |
2573 |
* |
fa307474c mm: update the do... |
2574 |
* If @addr is NULL, no operation is performed. |
c9fcee513 mm/vmalloc.c: add... |
2575 |
* |
fa307474c mm: update the do... |
2576 |
* Context: |
92eac1681 docs/mm: vmalloc:... |
2577 |
* May sleep if called *not* from interrupt context. |
fa307474c mm: update the do... |
2578 2579 |
* Must not be called in NMI context (strictly speaking, it could be * if we have CONFIG_ARCH_HAVE_NMI_SAFE_CMPXCHG, but making the calling |
f0953a1bb mm: fix typos in ... |
2580 |
* conventions for vfree() arch-dependent would be a really bad idea). |
1da177e4c Linux-2.6.12-rc2 |
2581 |
*/ |
b3bdda02a vmalloc: add cons... |
2582 |
void vfree(const void *addr) |
1da177e4c Linux-2.6.12-rc2 |
2583 |
{ |
32fcfd407 make vfree() safe... |
2584 |
BUG_ON(in_nmi()); |
89219d37a kmemleak: Add the... |
2585 2586 |
kmemleak_free(addr); |
a8dda165e vfree: add debug ... |
2587 |
might_sleep_if(!in_interrupt()); |
32fcfd407 make vfree() safe... |
2588 2589 |
if (!addr) return; |
c67dc6247 mm/vmalloc: do no... |
2590 2591 |
__vfree(addr); |
1da177e4c Linux-2.6.12-rc2 |
2592 |
} |
1da177e4c Linux-2.6.12-rc2 |
2593 2594 2595 |
EXPORT_SYMBOL(vfree); /** |
92eac1681 docs/mm: vmalloc:... |
2596 2597 |
* vunmap - release virtual mapping obtained by vmap() * @addr: memory base address |
1da177e4c Linux-2.6.12-rc2 |
2598 |
* |
92eac1681 docs/mm: vmalloc:... |
2599 2600 |
* Free the virtually contiguous memory area starting at @addr, * which was created from the page array passed to vmap(). |
1da177e4c Linux-2.6.12-rc2 |
2601 |
* |
92eac1681 docs/mm: vmalloc:... |
2602 |
* Must not be called in interrupt context. |
1da177e4c Linux-2.6.12-rc2 |
2603 |
*/ |
b3bdda02a vmalloc: add cons... |
2604 |
void vunmap(const void *addr) |
1da177e4c Linux-2.6.12-rc2 |
2605 2606 |
{ BUG_ON(in_interrupt()); |
34754b69a x86: make vmap ye... |
2607 |
might_sleep(); |
32fcfd407 make vfree() safe... |
2608 2609 |
if (addr) __vunmap(addr, 0); |
1da177e4c Linux-2.6.12-rc2 |
2610 |
} |
1da177e4c Linux-2.6.12-rc2 |
2611 2612 2613 |
EXPORT_SYMBOL(vunmap); /** |
92eac1681 docs/mm: vmalloc:... |
2614 2615 2616 2617 2618 2619 |
* vmap - map an array of pages into virtually contiguous space * @pages: array of page pointers * @count: number of pages to map * @flags: vm_area->flags * @prot: page protection for the mapping * |
b944afc9d mm: add a VM_MAP_... |
2620 2621 2622 2623 2624 |
* Maps @count pages from @pages into contiguous kernel virtual space. * If @flags contains %VM_MAP_PUT_PAGES the ownership of the pages array itself * (which must be kmalloc or vmalloc memory) and one reference per pages in it * are transferred from the caller to vmap(), and will be freed / dropped when * vfree() is called on the return value. |
a862f68a8 docs/core-api/mm:... |
2625 2626 |
* * Return: the address of the area or %NULL on failure |
1da177e4c Linux-2.6.12-rc2 |
2627 2628 |
*/ void *vmap(struct page **pages, unsigned int count, |
92eac1681 docs/mm: vmalloc:... |
2629 |
unsigned long flags, pgprot_t prot) |
1da177e4c Linux-2.6.12-rc2 |
2630 2631 |
{ struct vm_struct *area; |
b67177ecd mm/vmalloc: remov... |
2632 |
unsigned long addr; |
65ee03c4b mm: fix overflow ... |
2633 |
unsigned long size; /* In bytes */ |
1da177e4c Linux-2.6.12-rc2 |
2634 |
|
34754b69a x86: make vmap ye... |
2635 |
might_sleep(); |
ca79b0c21 mm: convert total... |
2636 |
if (count > totalram_pages()) |
1da177e4c Linux-2.6.12-rc2 |
2637 |
return NULL; |
65ee03c4b mm: fix overflow ... |
2638 2639 |
size = (unsigned long)count << PAGE_SHIFT; area = get_vm_area_caller(size, flags, __builtin_return_address(0)); |
1da177e4c Linux-2.6.12-rc2 |
2640 2641 |
if (!area) return NULL; |
230169693 vmallocinfo: add ... |
2642 |
|
b67177ecd mm/vmalloc: remov... |
2643 2644 2645 |
addr = (unsigned long)area->addr; if (vmap_pages_range(addr, addr + size, pgprot_nx(prot), pages, PAGE_SHIFT) < 0) { |
1da177e4c Linux-2.6.12-rc2 |
2646 2647 2648 |
vunmap(area->addr); return NULL; } |
c22ee5284 mm/vmalloc.c: fix... |
2649 |
if (flags & VM_MAP_PUT_PAGES) { |
b944afc9d mm: add a VM_MAP_... |
2650 |
area->pages = pages; |
c22ee5284 mm/vmalloc.c: fix... |
2651 2652 |
area->nr_pages = count; } |
1da177e4c Linux-2.6.12-rc2 |
2653 2654 |
return area->addr; } |
1da177e4c Linux-2.6.12-rc2 |
2655 |
EXPORT_SYMBOL(vmap); |
3e9a9e256 mm: add a vmap_pf... |
2656 2657 2658 2659 2660 2661 2662 2663 2664 2665 2666 2667 2668 2669 2670 2671 2672 2673 2674 2675 2676 2677 2678 2679 2680 2681 2682 2683 2684 2685 2686 2687 2688 2689 2690 2691 2692 2693 2694 2695 2696 2697 2698 2699 |
#ifdef CONFIG_VMAP_PFN struct vmap_pfn_data { unsigned long *pfns; pgprot_t prot; unsigned int idx; }; static int vmap_pfn_apply(pte_t *pte, unsigned long addr, void *private) { struct vmap_pfn_data *data = private; if (WARN_ON_ONCE(pfn_valid(data->pfns[data->idx]))) return -EINVAL; *pte = pte_mkspecial(pfn_pte(data->pfns[data->idx++], data->prot)); return 0; } /** * vmap_pfn - map an array of PFNs into virtually contiguous space * @pfns: array of PFNs * @count: number of pages to map * @prot: page protection for the mapping * * Maps @count PFNs from @pfns into contiguous kernel virtual space and returns * the start address of the mapping. */ void *vmap_pfn(unsigned long *pfns, unsigned int count, pgprot_t prot) { struct vmap_pfn_data data = { .pfns = pfns, .prot = pgprot_nx(prot) }; struct vm_struct *area; area = get_vm_area_caller(count * PAGE_SIZE, VM_IOREMAP, __builtin_return_address(0)); if (!area) return NULL; if (apply_to_page_range(&init_mm, (unsigned long)area->addr, count * PAGE_SIZE, vmap_pfn_apply, &data)) { free_vm_area(area); return NULL; } return area->addr; } EXPORT_SYMBOL_GPL(vmap_pfn); #endif /* CONFIG_VMAP_PFN */ |
12b9f873a mm/vmalloc: fallb... |
2700 2701 |
static inline unsigned int vm_area_alloc_pages(gfp_t gfp, int nid, |
343ab8178 mm/vmalloc: use b... |
2702 |
unsigned int order, unsigned int nr_pages, struct page **pages) |
12b9f873a mm/vmalloc: fallb... |
2703 2704 |
{ unsigned int nr_allocated = 0; |
ffb29b1c2 mm/vmalloc: fix n... |
2705 2706 |
struct page *page; int i; |
12b9f873a mm/vmalloc: fallb... |
2707 2708 2709 2710 2711 2712 2713 |
/* * For order-0 pages we make use of bulk allocator, if * the page array is partly or not at all populated due * to fails, fallback to a single page allocator that is * more permissive. */ |
ffb29b1c2 mm/vmalloc: fix n... |
2714 |
if (!order && nid != NUMA_NO_NODE) { |
343ab8178 mm/vmalloc: use b... |
2715 2716 2717 2718 2719 2720 2721 2722 2723 2724 2725 2726 2727 2728 2729 2730 2731 2732 2733 2734 2735 2736 2737 2738 |
while (nr_allocated < nr_pages) { unsigned int nr, nr_pages_request; /* * A maximum allowed request is hard-coded and is 100 * pages per call. That is done in order to prevent a * long preemption off scenario in the bulk-allocator * so the range is [1:100]. */ nr_pages_request = min(100U, nr_pages - nr_allocated); nr = alloc_pages_bulk_array_node(gfp, nid, nr_pages_request, pages + nr_allocated); nr_allocated += nr; cond_resched(); /* * If zero or pages were obtained partly, * fallback to a single page allocator. */ if (nr != nr_pages_request) break; } |
ffb29b1c2 mm/vmalloc: fix n... |
2739 |
} else if (order) |
12b9f873a mm/vmalloc: fallb... |
2740 2741 2742 2743 2744 2745 2746 |
/* * Compound pages required for remap_vmalloc_page if * high-order pages. */ gfp |= __GFP_COMP; /* High-order pages or fallback path if "bulk" fails. */ |
12b9f873a mm/vmalloc: fallb... |
2747 |
|
ffb29b1c2 mm/vmalloc: fix n... |
2748 2749 2750 2751 2752 |
while (nr_allocated < nr_pages) { if (nid == NUMA_NO_NODE) page = alloc_pages(gfp, order); else page = alloc_pages_node(nid, gfp, order); |
12b9f873a mm/vmalloc: fallb... |
2753 2754 2755 2756 2757 2758 2759 2760 2761 2762 |
if (unlikely(!page)) break; /* * Careful, we allocate and map page-order pages, but * tracking is done per PAGE_SIZE page so as to keep the * vm_struct APIs independent of the physical/mapped size. */ for (i = 0; i < (1U << order); i++) pages[nr_allocated + i] = page + i; |
12e376a6f mm/vmalloc: remov... |
2763 |
cond_resched(); |
12b9f873a mm/vmalloc: fallb... |
2764 2765 2766 2767 2768 |
nr_allocated += 1U << order; } return nr_allocated; } |
e31d9eb5c make __vmalloc_ar... |
2769 |
static void *__vmalloc_area_node(struct vm_struct *area, gfp_t gfp_mask, |
121e6f325 mm/vmalloc: hugep... |
2770 2771 |
pgprot_t prot, unsigned int page_shift, int node) |
1da177e4c Linux-2.6.12-rc2 |
2772 |
{ |
930f036b4 mm, vmalloc: cons... |
2773 |
const gfp_t nested_gfp = (gfp_mask & GFP_RECLAIM_MASK) | __GFP_ZERO; |
121e6f325 mm/vmalloc: hugep... |
2774 2775 |
unsigned long addr = (unsigned long)area->addr; unsigned long size = get_vm_area_size(area); |
34fe65371 mm/vmalloc.c:__vm... |
2776 |
unsigned long array_size; |
121e6f325 mm/vmalloc: hugep... |
2777 2778 |
unsigned int nr_small_pages = size >> PAGE_SHIFT; unsigned int page_order; |
1da177e4c Linux-2.6.12-rc2 |
2779 |
|
121e6f325 mm/vmalloc: hugep... |
2780 |
array_size = (unsigned long)nr_small_pages * sizeof(struct page *); |
f255935b9 mm: cleanup the g... |
2781 2782 2783 |
gfp_mask |= __GFP_NOWARN; if (!(gfp_mask & (GFP_DMA | GFP_DMA32))) gfp_mask |= __GFP_HIGHMEM; |
1da177e4c Linux-2.6.12-rc2 |
2784 |
|
1da177e4c Linux-2.6.12-rc2 |
2785 |
/* Please note that the recursion is strictly bounded. */ |
8757d5fa6 [PATCH] mm: fix o... |
2786 |
if (array_size > PAGE_SIZE) { |
5c1f4e690 mm/vmalloc: switc... |
2787 |
area->pages = __vmalloc_node(array_size, 1, nested_gfp, node, |
f255935b9 mm: cleanup the g... |
2788 |
area->caller); |
286e1ea3a [PATCH] vmalloc()... |
2789 |
} else { |
5c1f4e690 mm/vmalloc: switc... |
2790 |
area->pages = kmalloc_node(array_size, nested_gfp, node); |
286e1ea3a [PATCH] vmalloc()... |
2791 |
} |
7ea362427 mm/vmalloc.c: mov... |
2792 |
|
5c1f4e690 mm/vmalloc: switc... |
2793 |
if (!area->pages) { |
d70bec8cc mm/vmalloc: impro... |
2794 |
warn_alloc(gfp_mask, NULL, |
f4bdfeaf1 mm/vmalloc: remov... |
2795 2796 |
"vmalloc error: size %lu, failed to allocated page array size %lu", nr_small_pages * PAGE_SIZE, array_size); |
cd61413ba mm/vmalloc: print... |
2797 |
free_vm_area(area); |
1da177e4c Linux-2.6.12-rc2 |
2798 2799 |
return NULL; } |
1da177e4c Linux-2.6.12-rc2 |
2800 |
|
121e6f325 mm/vmalloc: hugep... |
2801 |
set_vm_area_page_order(area, page_shift - PAGE_SHIFT); |
121e6f325 mm/vmalloc: hugep... |
2802 |
page_order = vm_area_page_order(area); |
bf53d6f8f vmalloc: clean up... |
2803 |
|
12b9f873a mm/vmalloc: fallb... |
2804 2805 |
area->nr_pages = vm_area_alloc_pages(gfp_mask, node, page_order, nr_small_pages, area->pages); |
5c1f4e690 mm/vmalloc: switc... |
2806 |
|
97105f0ab mm: vmalloc: show... |
2807 |
atomic_long_add(area->nr_pages, &nr_vmalloc_pages); |
1da177e4c Linux-2.6.12-rc2 |
2808 |
|
5c1f4e690 mm/vmalloc: switc... |
2809 2810 2811 2812 2813 2814 |
/* * If not enough pages were obtained to accomplish an * allocation request, free them via __vfree() if any. */ if (area->nr_pages != nr_small_pages) { warn_alloc(gfp_mask, NULL, |
f4bdfeaf1 mm/vmalloc: remov... |
2815 |
"vmalloc error: size %lu, page order %u, failed to allocate pages", |
5c1f4e690 mm/vmalloc: switc... |
2816 2817 2818 |
area->nr_pages * PAGE_SIZE, page_order); goto fail; } |
12b9f873a mm/vmalloc: fallb... |
2819 2820 |
if (vmap_pages_range(addr, addr + size, prot, area->pages, page_shift) < 0) { |
d70bec8cc mm/vmalloc: impro... |
2821 |
warn_alloc(gfp_mask, NULL, |
f4bdfeaf1 mm/vmalloc: remov... |
2822 2823 |
"vmalloc error: size %lu, failed to map pages", area->nr_pages * PAGE_SIZE); |
1da177e4c Linux-2.6.12-rc2 |
2824 |
goto fail; |
d70bec8cc mm/vmalloc: impro... |
2825 |
} |
ed1f324c5 mm: remove map_vm... |
2826 |
|
1da177e4c Linux-2.6.12-rc2 |
2827 2828 2829 |
return area->addr; fail: |
c67dc6247 mm/vmalloc: do no... |
2830 |
__vfree(area->addr); |
1da177e4c Linux-2.6.12-rc2 |
2831 2832 2833 2834 |
return NULL; } /** |
92eac1681 docs/mm: vmalloc:... |
2835 2836 2837 2838 2839 2840 2841 2842 2843 2844 2845 2846 2847 2848 |
* __vmalloc_node_range - allocate virtually contiguous memory * @size: allocation size * @align: desired alignment * @start: vm area range start * @end: vm area range end * @gfp_mask: flags for the page level allocator * @prot: protection mask for the allocated pages * @vm_flags: additional vm area flags (e.g. %VM_NO_GUARD) * @node: node to use for allocation or NUMA_NO_NODE * @caller: caller's return address * * Allocate enough pages to cover @size from the page level * allocator with @gfp_mask flags. Map them into contiguous * kernel virtual space, using a pagetable protection of @prot. |
a862f68a8 docs/core-api/mm:... |
2849 2850 |
* * Return: the address of the area or %NULL on failure |
1da177e4c Linux-2.6.12-rc2 |
2851 |
*/ |
d0a21265d mm: unify module_... |
2852 2853 |
void *__vmalloc_node_range(unsigned long size, unsigned long align, unsigned long start, unsigned long end, gfp_t gfp_mask, |
cb9e3c292 mm: vmalloc: pass... |
2854 2855 |
pgprot_t prot, unsigned long vm_flags, int node, const void *caller) |
1da177e4c Linux-2.6.12-rc2 |
2856 2857 |
{ struct vm_struct *area; |
89219d37a kmemleak: Add the... |
2858 2859 |
void *addr; unsigned long real_size = size; |
121e6f325 mm/vmalloc: hugep... |
2860 2861 |
unsigned long real_align = align; unsigned int shift = PAGE_SHIFT; |
1da177e4c Linux-2.6.12-rc2 |
2862 |
|
d70bec8cc mm/vmalloc: impro... |
2863 2864 2865 2866 2867 |
if (WARN_ON_ONCE(!size)) return NULL; if ((size >> PAGE_SHIFT) > totalram_pages()) { warn_alloc(gfp_mask, NULL, |
f4bdfeaf1 mm/vmalloc: remov... |
2868 2869 |
"vmalloc error: size %lu, exceeds total pages", real_size); |
d70bec8cc mm/vmalloc: impro... |
2870 |
return NULL; |
121e6f325 mm/vmalloc: hugep... |
2871 |
} |
3382bbee0 mm/vmalloc: enabl... |
2872 |
if (vmap_allow_huge && !(vm_flags & VM_NO_HUGE_VMAP)) { |
121e6f325 mm/vmalloc: hugep... |
2873 |
unsigned long size_per_node; |
1da177e4c Linux-2.6.12-rc2 |
2874 |
|
121e6f325 mm/vmalloc: hugep... |
2875 2876 2877 2878 2879 2880 2881 2882 2883 2884 |
/* * Try huge pages. Only try for PAGE_KERNEL allocations, * others like modules don't yet expect huge pages in * their allocations due to apply_to_page_range not * supporting them. */ size_per_node = size; if (node == NUMA_NO_NODE) size_per_node /= num_online_nodes(); |
3382bbee0 mm/vmalloc: enabl... |
2885 |
if (arch_vmap_pmd_supported(prot) && size_per_node >= PMD_SIZE) |
121e6f325 mm/vmalloc: hugep... |
2886 |
shift = PMD_SHIFT; |
3382bbee0 mm/vmalloc: enabl... |
2887 2888 2889 2890 2891 |
else shift = arch_vmap_pte_supported_shift(size_per_node); align = max(real_align, 1UL << shift); size = ALIGN(real_size, 1UL << shift); |
121e6f325 mm/vmalloc: hugep... |
2892 2893 2894 |
} again: |
7ca3027b7 mm/vmalloc: unbre... |
2895 2896 2897 |
area = __get_vm_area_node(real_size, align, shift, VM_ALLOC | VM_UNINITIALIZED | vm_flags, start, end, node, gfp_mask, caller); |
d70bec8cc mm/vmalloc: impro... |
2898 2899 |
if (!area) { warn_alloc(gfp_mask, NULL, |
f4bdfeaf1 mm/vmalloc: remov... |
2900 2901 |
"vmalloc error: size %lu, vm_struct allocation failed", real_size); |
de7d2b567 mm/vmalloc.c: rep... |
2902 |
goto fail; |
d70bec8cc mm/vmalloc: impro... |
2903 |
} |
1da177e4c Linux-2.6.12-rc2 |
2904 |
|
121e6f325 mm/vmalloc: hugep... |
2905 |
addr = __vmalloc_area_node(area, gfp_mask, prot, shift, node); |
1368edf06 mm: vmalloc: chec... |
2906 |
if (!addr) |
121e6f325 mm/vmalloc: hugep... |
2907 |
goto fail; |
89219d37a kmemleak: Add the... |
2908 2909 |
/* |
20fc02b47 mm/vmalloc.c: ren... |
2910 2911 |
* In this function, newly allocated vm_struct has VM_UNINITIALIZED * flag. It means that vm_struct is not fully initialized. |
4341fa454 mm, vmalloc: remo... |
2912 |
* Now, it is fully initialized, so remove this flag here. |
f5252e009 mm: avoid null po... |
2913 |
*/ |
20fc02b47 mm/vmalloc.c: ren... |
2914 |
clear_vm_uninitialized_flag(area); |
f5252e009 mm: avoid null po... |
2915 |
|
7ca3027b7 mm/vmalloc: unbre... |
2916 |
size = PAGE_ALIGN(size); |
f1675103e mm: defer kmemlea... |
2917 2918 |
if (!(vm_flags & VM_DEFER_KMEMLEAK)) kmemleak_vmalloc(area, size, gfp_mask); |
89219d37a kmemleak: Add the... |
2919 2920 |
return addr; |
de7d2b567 mm/vmalloc.c: rep... |
2921 2922 |
fail: |
121e6f325 mm/vmalloc: hugep... |
2923 2924 2925 2926 2927 2928 |
if (shift > PAGE_SHIFT) { shift = PAGE_SHIFT; align = real_align; size = real_size; goto again; } |
de7d2b567 mm/vmalloc.c: rep... |
2929 |
return NULL; |
1da177e4c Linux-2.6.12-rc2 |
2930 |
} |
d0a21265d mm: unify module_... |
2931 |
/** |
92eac1681 docs/mm: vmalloc:... |
2932 2933 2934 2935 |
* __vmalloc_node - allocate virtually contiguous memory * @size: allocation size * @align: desired alignment * @gfp_mask: flags for the page level allocator |
92eac1681 docs/mm: vmalloc:... |
2936 2937 |
* @node: node to use for allocation or NUMA_NO_NODE * @caller: caller's return address |
a7c3e901a mm: introduce kv[... |
2938 |
* |
f38fcb9c1 mm: remove the pr... |
2939 2940 |
* Allocate enough pages to cover @size from the page level allocator with * @gfp_mask flags. Map them into contiguous kernel virtual space. |
a7c3e901a mm: introduce kv[... |
2941 |
* |
92eac1681 docs/mm: vmalloc:... |
2942 2943 |
* Reclaim modifiers in @gfp_mask - __GFP_NORETRY, __GFP_RETRY_MAYFAIL * and __GFP_NOFAIL are not supported |
a7c3e901a mm: introduce kv[... |
2944 |
* |
92eac1681 docs/mm: vmalloc:... |
2945 2946 |
* Any use of gfp flags outside of GFP_KERNEL should be consulted * with mm people. |
a862f68a8 docs/core-api/mm:... |
2947 2948 |
* * Return: pointer to the allocated memory or %NULL on error |
d0a21265d mm: unify module_... |
2949 |
*/ |
2b9059489 mm: remove __vmal... |
2950 |
void *__vmalloc_node(unsigned long size, unsigned long align, |
f38fcb9c1 mm: remove the pr... |
2951 |
gfp_t gfp_mask, int node, const void *caller) |
d0a21265d mm: unify module_... |
2952 2953 |
{ return __vmalloc_node_range(size, align, VMALLOC_START, VMALLOC_END, |
f38fcb9c1 mm: remove the pr... |
2954 |
gfp_mask, PAGE_KERNEL, 0, node, caller); |
d0a21265d mm: unify module_... |
2955 |
} |
c3f896dcf mm: switch the te... |
2956 2957 2958 2959 2960 2961 2962 2963 |
/* * This is only for performance analysis of vmalloc and stress purpose. * It is required by vmalloc test module, therefore do not use it other * than that. */ #ifdef CONFIG_TEST_VMALLOC_MODULE EXPORT_SYMBOL_GPL(__vmalloc_node); #endif |
d0a21265d mm: unify module_... |
2964 |
|
88dca4ca5 mm: remove the pg... |
2965 |
void *__vmalloc(unsigned long size, gfp_t gfp_mask) |
930fc45a4 [PATCH] vmalloc_node |
2966 |
{ |
f38fcb9c1 mm: remove the pr... |
2967 |
return __vmalloc_node(size, 1, gfp_mask, NUMA_NO_NODE, |
230169693 vmallocinfo: add ... |
2968 |
__builtin_return_address(0)); |
930fc45a4 [PATCH] vmalloc_node |
2969 |
} |
1da177e4c Linux-2.6.12-rc2 |
2970 2971 2972 |
EXPORT_SYMBOL(__vmalloc); /** |
92eac1681 docs/mm: vmalloc:... |
2973 2974 2975 2976 2977 |
* vmalloc - allocate virtually contiguous memory * @size: allocation size * * Allocate enough pages to cover @size from the page level * allocator and map them into contiguous kernel virtual space. |
1da177e4c Linux-2.6.12-rc2 |
2978 |
* |
92eac1681 docs/mm: vmalloc:... |
2979 2980 |
* For tight control over page level allocator and protection flags * use __vmalloc() instead. |
a862f68a8 docs/core-api/mm:... |
2981 2982 |
* * Return: pointer to the allocated memory or %NULL on error |
1da177e4c Linux-2.6.12-rc2 |
2983 2984 2985 |
*/ void *vmalloc(unsigned long size) { |
4d39d7285 mm: remove both i... |
2986 2987 |
return __vmalloc_node(size, 1, GFP_KERNEL, NUMA_NO_NODE, __builtin_return_address(0)); |
1da177e4c Linux-2.6.12-rc2 |
2988 |
} |
1da177e4c Linux-2.6.12-rc2 |
2989 |
EXPORT_SYMBOL(vmalloc); |
930fc45a4 [PATCH] vmalloc_node |
2990 |
/** |
15a64f5a8 mm/vmalloc: add v... |
2991 2992 2993 2994 2995 2996 2997 2998 2999 3000 3001 3002 3003 3004 3005 3006 3007 |
* vmalloc_no_huge - allocate virtually contiguous memory using small pages * @size: allocation size * * Allocate enough non-huge pages to cover @size from the page level * allocator and map them into contiguous kernel virtual space. * * Return: pointer to the allocated memory or %NULL on error */ void *vmalloc_no_huge(unsigned long size) { return __vmalloc_node_range(size, 1, VMALLOC_START, VMALLOC_END, GFP_KERNEL, PAGE_KERNEL, VM_NO_HUGE_VMAP, NUMA_NO_NODE, __builtin_return_address(0)); } EXPORT_SYMBOL(vmalloc_no_huge); /** |
92eac1681 docs/mm: vmalloc:... |
3008 3009 3010 3011 3012 3013 3014 3015 3016 |
* vzalloc - allocate virtually contiguous memory with zero fill * @size: allocation size * * Allocate enough pages to cover @size from the page level * allocator and map them into contiguous kernel virtual space. * The memory allocated is set to zero. * * For tight control over page level allocator and protection flags * use __vmalloc() instead. |
a862f68a8 docs/core-api/mm:... |
3017 3018 |
* * Return: pointer to the allocated memory or %NULL on error |
e1ca7788d mm: add vzalloc()... |
3019 3020 3021 |
*/ void *vzalloc(unsigned long size) { |
4d39d7285 mm: remove both i... |
3022 3023 |
return __vmalloc_node(size, 1, GFP_KERNEL | __GFP_ZERO, NUMA_NO_NODE, __builtin_return_address(0)); |
e1ca7788d mm: add vzalloc()... |
3024 3025 3026 3027 |
} EXPORT_SYMBOL(vzalloc); /** |
ead04089b [PATCH] Fix kerne... |
3028 3029 |
* vmalloc_user - allocate zeroed virtually contiguous memory for userspace * @size: allocation size |
833423143 [PATCH] mm: intro... |
3030 |
* |
ead04089b [PATCH] Fix kerne... |
3031 3032 |
* The resulting memory area is zeroed so it can be mapped to userspace * without leaking data. |
a862f68a8 docs/core-api/mm:... |
3033 3034 |
* * Return: pointer to the allocated memory or %NULL on error |
833423143 [PATCH] mm: intro... |
3035 3036 3037 |
*/ void *vmalloc_user(unsigned long size) { |
bc84c5352 mm/vmalloc: pass ... |
3038 3039 3040 3041 |
return __vmalloc_node_range(size, SHMLBA, VMALLOC_START, VMALLOC_END, GFP_KERNEL | __GFP_ZERO, PAGE_KERNEL, VM_USERMAP, NUMA_NO_NODE, __builtin_return_address(0)); |
833423143 [PATCH] mm: intro... |
3042 3043 3044 3045 |
} EXPORT_SYMBOL(vmalloc_user); /** |
92eac1681 docs/mm: vmalloc:... |
3046 3047 3048 |
* vmalloc_node - allocate memory on a specific node * @size: allocation size * @node: numa node |
930fc45a4 [PATCH] vmalloc_node |
3049 |
* |
92eac1681 docs/mm: vmalloc:... |
3050 3051 |
* Allocate enough pages to cover @size from the page level * allocator and map them into contiguous kernel virtual space. |
930fc45a4 [PATCH] vmalloc_node |
3052 |
* |
92eac1681 docs/mm: vmalloc:... |
3053 3054 |
* For tight control over page level allocator and protection flags * use __vmalloc() instead. |
a862f68a8 docs/core-api/mm:... |
3055 3056 |
* * Return: pointer to the allocated memory or %NULL on error |
930fc45a4 [PATCH] vmalloc_node |
3057 3058 3059 |
*/ void *vmalloc_node(unsigned long size, int node) { |
f38fcb9c1 mm: remove the pr... |
3060 3061 |
return __vmalloc_node(size, 1, GFP_KERNEL, node, __builtin_return_address(0)); |
930fc45a4 [PATCH] vmalloc_node |
3062 3063 |
} EXPORT_SYMBOL(vmalloc_node); |
e1ca7788d mm: add vzalloc()... |
3064 3065 3066 3067 3068 3069 3070 3071 3072 |
/** * 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. * |
a862f68a8 docs/core-api/mm:... |
3073 |
* Return: pointer to the allocated memory or %NULL on error |
e1ca7788d mm: add vzalloc()... |
3074 3075 3076 |
*/ void *vzalloc_node(unsigned long size, int node) { |
4d39d7285 mm: remove both i... |
3077 3078 |
return __vmalloc_node(size, 1, GFP_KERNEL | __GFP_ZERO, node, __builtin_return_address(0)); |
e1ca7788d mm: add vzalloc()... |
3079 3080 |
} EXPORT_SYMBOL(vzalloc_node); |
0d08e0d3a [PATCH] x86-64: F... |
3081 |
#if defined(CONFIG_64BIT) && defined(CONFIG_ZONE_DMA32) |
698d0831b vmalloc: fix __GF... |
3082 |
#define GFP_VMALLOC32 (GFP_DMA32 | GFP_KERNEL) |
0d08e0d3a [PATCH] x86-64: F... |
3083 |
#elif defined(CONFIG_64BIT) && defined(CONFIG_ZONE_DMA) |
698d0831b vmalloc: fix __GF... |
3084 |
#define GFP_VMALLOC32 (GFP_DMA | GFP_KERNEL) |
0d08e0d3a [PATCH] x86-64: F... |
3085 |
#else |
698d0831b vmalloc: fix __GF... |
3086 3087 3088 3089 |
/* * 64b systems should always have either DMA or DMA32 zones. For others * GFP_DMA32 should do the right thing and use the normal zone. */ |
68d68ff6e mm/mempool: minor... |
3090 |
#define GFP_VMALLOC32 (GFP_DMA32 | GFP_KERNEL) |
0d08e0d3a [PATCH] x86-64: F... |
3091 |
#endif |
1da177e4c Linux-2.6.12-rc2 |
3092 |
/** |
92eac1681 docs/mm: vmalloc:... |
3093 3094 |
* vmalloc_32 - allocate virtually contiguous memory (32bit addressable) * @size: allocation size |
1da177e4c Linux-2.6.12-rc2 |
3095 |
* |
92eac1681 docs/mm: vmalloc:... |
3096 3097 |
* Allocate enough 32bit PA addressable pages to cover @size from the * page level allocator and map them into contiguous kernel virtual space. |
a862f68a8 docs/core-api/mm:... |
3098 3099 |
* * Return: pointer to the allocated memory or %NULL on error |
1da177e4c Linux-2.6.12-rc2 |
3100 3101 3102 |
*/ void *vmalloc_32(unsigned long size) { |
f38fcb9c1 mm: remove the pr... |
3103 3104 |
return __vmalloc_node(size, 1, GFP_VMALLOC32, NUMA_NO_NODE, __builtin_return_address(0)); |
1da177e4c Linux-2.6.12-rc2 |
3105 |
} |
1da177e4c Linux-2.6.12-rc2 |
3106 |
EXPORT_SYMBOL(vmalloc_32); |
833423143 [PATCH] mm: intro... |
3107 |
/** |
ead04089b [PATCH] Fix kerne... |
3108 |
* vmalloc_32_user - allocate zeroed virtually contiguous 32bit memory |
92eac1681 docs/mm: vmalloc:... |
3109 |
* @size: allocation size |
ead04089b [PATCH] Fix kerne... |
3110 3111 3112 |
* * The resulting memory area is 32bit addressable and zeroed so it can be * mapped to userspace without leaking data. |
a862f68a8 docs/core-api/mm:... |
3113 3114 |
* * Return: pointer to the allocated memory or %NULL on error |
833423143 [PATCH] mm: intro... |
3115 3116 3117 |
*/ void *vmalloc_32_user(unsigned long size) { |
bc84c5352 mm/vmalloc: pass ... |
3118 3119 3120 3121 |
return __vmalloc_node_range(size, SHMLBA, VMALLOC_START, VMALLOC_END, GFP_VMALLOC32 | __GFP_ZERO, PAGE_KERNEL, VM_USERMAP, NUMA_NO_NODE, __builtin_return_address(0)); |
833423143 [PATCH] mm: intro... |
3122 3123 |
} EXPORT_SYMBOL(vmalloc_32_user); |
d0107eb07 kcore: fix vread/... |
3124 3125 3126 3127 3128 3129 3130 3131 3132 3133 3134 3135 |
/* * small helper routine , copy contents to buf from addr. * If the page is not present, fill zero. */ static int aligned_vread(char *buf, char *addr, unsigned long count) { struct page *p; int copied = 0; while (count) { unsigned long offset, length; |
891c49abf mm/vmalloc: use o... |
3136 |
offset = offset_in_page(addr); |
d0107eb07 kcore: fix vread/... |
3137 3138 3139 3140 3141 3142 3143 |
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 |
f0953a1bb mm: fix typos in ... |
3144 |
* overhead of vmalloc()/vfree() calls for this _debug_ |
d0107eb07 kcore: fix vread/... |
3145 3146 3147 3148 |
* interface, rarely used. Instead of that, we'll use * kmap() and get small overhead in this access function. */ if (p) { |
f7c8ce44e mm/vmalloc: remov... |
3149 |
/* We can expect USER0 is not used -- see vread() */ |
9b04c5fec mm: remove the se... |
3150 |
void *map = kmap_atomic(p); |
d0107eb07 kcore: fix vread/... |
3151 |
memcpy(buf, map + offset, length); |
9b04c5fec mm: remove the se... |
3152 |
kunmap_atomic(map); |
d0107eb07 kcore: fix vread/... |
3153 3154 3155 3156 3157 3158 3159 3160 3161 3162 |
} else memset(buf, 0, length); addr += length; buf += length; copied += length; count -= length; } return copied; } |
d0107eb07 kcore: fix vread/... |
3163 |
/** |
92eac1681 docs/mm: vmalloc:... |
3164 3165 3166 3167 3168 |
* vread() - read vmalloc area in a safe way. * @buf: buffer for reading data * @addr: vm address. * @count: number of bytes to be read. * |
92eac1681 docs/mm: vmalloc:... |
3169 3170 3171 3172 3173 3174 3175 3176 3177 3178 3179 3180 |
* This function checks that addr is a valid vmalloc'ed area, and * copy data from that area to a given buffer. If the given memory range * of [addr...addr+count) includes some valid address, data is copied to * proper area of @buf. If there are memory holes, they'll be zero-filled. * IOREMAP area is treated as memory hole and no copy is done. * * If [addr...addr+count) doesn't includes any intersects with alive * vm_struct area, returns 0. @buf should be kernel's buffer. * * Note: In usual ops, vread() is never necessary because the caller * should know vmalloc() area is valid and can use memcpy(). * This is for routines which have to access vmalloc area without |
bbcd53c96 drivers/char: rem... |
3181 |
* any information, as /proc/kcore. |
a862f68a8 docs/core-api/mm:... |
3182 3183 3184 3185 |
* * Return: number of bytes for which addr and buf should be increased * (same number as @count) or %0 if [addr...addr+count) doesn't * include any intersection with valid vmalloc area |
d0107eb07 kcore: fix vread/... |
3186 |
*/ |
1da177e4c Linux-2.6.12-rc2 |
3187 3188 |
long vread(char *buf, char *addr, unsigned long count) { |
e81ce85f9 mm, vmalloc: iter... |
3189 3190 |
struct vmap_area *va; struct vm_struct *vm; |
1da177e4c Linux-2.6.12-rc2 |
3191 |
char *vaddr, *buf_start = buf; |
d0107eb07 kcore: fix vread/... |
3192 |
unsigned long buflen = count; |
1da177e4c Linux-2.6.12-rc2 |
3193 3194 3195 3196 3197 |
unsigned long n; /* Don't allow overflow */ if ((unsigned long) addr + count < count) count = -(unsigned long) addr; |
e81ce85f9 mm, vmalloc: iter... |
3198 |
spin_lock(&vmap_area_lock); |
f181234a5 mm/vmalloc: fix w... |
3199 |
va = find_vmap_area_exceed_addr((unsigned long)addr); |
f608788cd mm/vmalloc: use r... |
3200 3201 |
if (!va) goto finished; |
f181234a5 mm/vmalloc: fix w... |
3202 3203 3204 3205 |
/* no intersects with alive vmap_area */ if ((unsigned long)addr + count <= va->va_start) goto finished; |
f608788cd mm/vmalloc: use r... |
3206 |
list_for_each_entry_from(va, &vmap_area_list, list) { |
e81ce85f9 mm, vmalloc: iter... |
3207 3208 |
if (!count) break; |
688fcbfc0 mm/vmalloc: modif... |
3209 |
if (!va->vm) |
e81ce85f9 mm, vmalloc: iter... |
3210 3211 3212 3213 |
continue; vm = va->vm; vaddr = (char *) vm->addr; |
762216ab4 mm/vmalloc: use w... |
3214 |
if (addr >= vaddr + get_vm_area_size(vm)) |
1da177e4c Linux-2.6.12-rc2 |
3215 3216 3217 3218 3219 3220 3221 3222 3223 |
continue; while (addr < vaddr) { if (count == 0) goto finished; *buf = '\0'; buf++; addr++; count--; } |
762216ab4 mm/vmalloc: use w... |
3224 |
n = vaddr + get_vm_area_size(vm) - addr; |
d0107eb07 kcore: fix vread/... |
3225 3226 |
if (n > count) n = count; |
e81ce85f9 mm, vmalloc: iter... |
3227 |
if (!(vm->flags & VM_IOREMAP)) |
d0107eb07 kcore: fix vread/... |
3228 3229 3230 3231 3232 3233 |
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 |
3234 3235 |
} finished: |
e81ce85f9 mm, vmalloc: iter... |
3236 |
spin_unlock(&vmap_area_lock); |
d0107eb07 kcore: fix vread/... |
3237 3238 3239 3240 3241 3242 3243 3244 |
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 |
3245 |
} |
d0107eb07 kcore: fix vread/... |
3246 |
/** |
92eac1681 docs/mm: vmalloc:... |
3247 3248 3249 3250 |
* 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 |
bdebd6a28 vmalloc: fix rema... |
3251 |
* @pgoff: offset from @kaddr to start at |
92eac1681 docs/mm: vmalloc:... |
3252 |
* @size: size of map area |
7682486b3 mm: fix various k... |
3253 |
* |
92eac1681 docs/mm: vmalloc:... |
3254 |
* Returns: 0 for success, -Exxx on failure |
833423143 [PATCH] mm: intro... |
3255 |
* |
92eac1681 docs/mm: vmalloc:... |
3256 3257 3258 3259 |
* 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... |
3260 |
* |
92eac1681 docs/mm: vmalloc:... |
3261 |
* Similar to remap_pfn_range() (see mm/memory.c) |
833423143 [PATCH] mm: intro... |
3262 |
*/ |
e69e9d4ae vmalloc: introduc... |
3263 |
int remap_vmalloc_range_partial(struct vm_area_struct *vma, unsigned long uaddr, |
bdebd6a28 vmalloc: fix rema... |
3264 3265 |
void *kaddr, unsigned long pgoff, unsigned long size) |
833423143 [PATCH] mm: intro... |
3266 3267 |
{ struct vm_struct *area; |
bdebd6a28 vmalloc: fix rema... |
3268 3269 3270 3271 3272 |
unsigned long off; unsigned long end_index; if (check_shl_overflow(pgoff, PAGE_SHIFT, &off)) return -EINVAL; |
833423143 [PATCH] mm: intro... |
3273 |
|
e69e9d4ae vmalloc: introduc... |
3274 3275 3276 |
size = PAGE_ALIGN(size); if (!PAGE_ALIGNED(uaddr) || !PAGE_ALIGNED(kaddr)) |
833423143 [PATCH] mm: intro... |
3277 |
return -EINVAL; |
e69e9d4ae vmalloc: introduc... |
3278 |
area = find_vm_area(kaddr); |
833423143 [PATCH] mm: intro... |
3279 |
if (!area) |
db64fe022 mm: rewrite vmap ... |
3280 |
return -EINVAL; |
833423143 [PATCH] mm: intro... |
3281 |
|
fe9041c24 vmalloc: lift the... |
3282 |
if (!(area->flags & (VM_USERMAP | VM_DMA_COHERENT))) |
db64fe022 mm: rewrite vmap ... |
3283 |
return -EINVAL; |
833423143 [PATCH] mm: intro... |
3284 |
|
bdebd6a28 vmalloc: fix rema... |
3285 3286 |
if (check_add_overflow(size, off, &end_index) || end_index > get_vm_area_size(area)) |
db64fe022 mm: rewrite vmap ... |
3287 |
return -EINVAL; |
bdebd6a28 vmalloc: fix rema... |
3288 |
kaddr += off; |
833423143 [PATCH] mm: intro... |
3289 |
|
833423143 [PATCH] mm: intro... |
3290 |
do { |
e69e9d4ae vmalloc: introduc... |
3291 |
struct page *page = vmalloc_to_page(kaddr); |
db64fe022 mm: rewrite vmap ... |
3292 |
int ret; |
833423143 [PATCH] mm: intro... |
3293 3294 3295 3296 3297 |
ret = vm_insert_page(vma, uaddr, page); if (ret) return ret; uaddr += PAGE_SIZE; |
e69e9d4ae vmalloc: introduc... |
3298 3299 3300 |
kaddr += PAGE_SIZE; size -= PAGE_SIZE; } while (size > 0); |
833423143 [PATCH] mm: intro... |
3301 |
|
314e51b98 mm: kill vma flag... |
3302 |
vma->vm_flags |= VM_DONTEXPAND | VM_DONTDUMP; |
833423143 [PATCH] mm: intro... |
3303 |
|
db64fe022 mm: rewrite vmap ... |
3304 |
return 0; |
833423143 [PATCH] mm: intro... |
3305 |
} |
e69e9d4ae vmalloc: introduc... |
3306 3307 |
/** |
92eac1681 docs/mm: vmalloc:... |
3308 3309 3310 3311 |
* remap_vmalloc_range - map vmalloc pages to userspace * @vma: vma to cover (map full range of vma) * @addr: vmalloc memory * @pgoff: number of pages into addr before first page to map |
e69e9d4ae vmalloc: introduc... |
3312 |
* |
92eac1681 docs/mm: vmalloc:... |
3313 |
* Returns: 0 for success, -Exxx on failure |
e69e9d4ae vmalloc: introduc... |
3314 |
* |
92eac1681 docs/mm: vmalloc:... |
3315 3316 3317 |
* This function checks that addr is a valid vmalloc'ed area, and * that it is big enough to cover the vma. Will return failure if * that criteria isn't met. |
e69e9d4ae vmalloc: introduc... |
3318 |
* |
92eac1681 docs/mm: vmalloc:... |
3319 |
* Similar to remap_pfn_range() (see mm/memory.c) |
e69e9d4ae vmalloc: introduc... |
3320 3321 3322 3323 3324 |
*/ int remap_vmalloc_range(struct vm_area_struct *vma, void *addr, unsigned long pgoff) { return remap_vmalloc_range_partial(vma, vma->vm_start, |
bdebd6a28 vmalloc: fix rema... |
3325 |
addr, pgoff, |
e69e9d4ae vmalloc: introduc... |
3326 3327 |
vma->vm_end - vma->vm_start); } |
833423143 [PATCH] mm: intro... |
3328 |
EXPORT_SYMBOL(remap_vmalloc_range); |
5f4352fbf Allocate and free... |
3329 3330 3331 3332 3333 3334 3335 3336 |
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... |
3337 |
|
4f8b02b4e vmalloc: pcpu_get... |
3338 |
#ifdef CONFIG_SMP |
ca23e405e vmalloc: implemen... |
3339 3340 |
static struct vmap_area *node_to_va(struct rb_node *n) { |
4583e7731 mm/vmalloc.c: use... |
3341 |
return rb_entry_safe(n, struct vmap_area, rb_node); |
ca23e405e vmalloc: implemen... |
3342 3343 3344 |
} /** |
68ad4a330 mm/vmalloc.c: kee... |
3345 3346 |
* pvm_find_va_enclose_addr - find the vmap_area @addr belongs to * @addr: target address |
ca23e405e vmalloc: implemen... |
3347 |
* |
68ad4a330 mm/vmalloc.c: kee... |
3348 3349 3350 3351 |
* Returns: vmap_area if it is found. If there is no such area * the first highest(reverse order) vmap_area is returned * i.e. va->va_start < addr && va->va_end < addr or NULL * if there are no any areas before @addr. |
ca23e405e vmalloc: implemen... |
3352 |
*/ |
68ad4a330 mm/vmalloc.c: kee... |
3353 3354 |
static struct vmap_area * pvm_find_va_enclose_addr(unsigned long addr) |
ca23e405e vmalloc: implemen... |
3355 |
{ |
68ad4a330 mm/vmalloc.c: kee... |
3356 3357 3358 3359 3360 |
struct vmap_area *va, *tmp; struct rb_node *n; n = free_vmap_area_root.rb_node; va = NULL; |
ca23e405e vmalloc: implemen... |
3361 3362 |
while (n) { |
68ad4a330 mm/vmalloc.c: kee... |
3363 3364 3365 3366 3367 |
tmp = rb_entry(n, struct vmap_area, rb_node); if (tmp->va_start <= addr) { va = tmp; if (tmp->va_end >= addr) break; |
ca23e405e vmalloc: implemen... |
3368 |
n = n->rb_right; |
68ad4a330 mm/vmalloc.c: kee... |
3369 3370 3371 |
} else { n = n->rb_left; } |
ca23e405e vmalloc: implemen... |
3372 |
} |
68ad4a330 mm/vmalloc.c: kee... |
3373 |
return va; |
ca23e405e vmalloc: implemen... |
3374 3375 3376 |
} /** |
68ad4a330 mm/vmalloc.c: kee... |
3377 3378 3379 3380 3381 |
* pvm_determine_end_from_reverse - find the highest aligned address * of free block below VMALLOC_END * @va: * in - the VA we start the search(reverse order); * out - the VA with the highest aligned end address. |
799fa85d6 mm/vmalloc: add '... |
3382 |
* @align: alignment for required highest address |
ca23e405e vmalloc: implemen... |
3383 |
* |
68ad4a330 mm/vmalloc.c: kee... |
3384 |
* Returns: determined end address within vmap_area |
ca23e405e vmalloc: implemen... |
3385 |
*/ |
68ad4a330 mm/vmalloc.c: kee... |
3386 3387 |
static unsigned long pvm_determine_end_from_reverse(struct vmap_area **va, unsigned long align) |
ca23e405e vmalloc: implemen... |
3388 |
{ |
68ad4a330 mm/vmalloc.c: kee... |
3389 |
unsigned long vmalloc_end = VMALLOC_END & ~(align - 1); |
ca23e405e vmalloc: implemen... |
3390 |
unsigned long addr; |
68ad4a330 mm/vmalloc.c: kee... |
3391 3392 3393 3394 3395 3396 3397 |
if (likely(*va)) { list_for_each_entry_from_reverse((*va), &free_vmap_area_list, list) { addr = min((*va)->va_end & ~(align - 1), vmalloc_end); if ((*va)->va_start < addr) return addr; } |
ca23e405e vmalloc: implemen... |
3398 |
} |
68ad4a330 mm/vmalloc.c: kee... |
3399 |
return 0; |
ca23e405e vmalloc: implemen... |
3400 3401 3402 3403 3404 3405 3406 3407 |
} /** * 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... |
3408 3409 3410 3411 3412 3413 |
* * 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... |
3414 3415 3416 3417 |
* 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... |
3418 |
* |
68ad4a330 mm/vmalloc.c: kee... |
3419 3420 3421 3422 3423 3424 |
* Despite its complicated look, this allocator is rather simple. It * does everything top-down and scans free blocks from the end looking * for matching base. While scanning, if any of the areas do not fit the * base address is pulled down to fit the area. Scanning is repeated till * all the areas fit and then all necessary data structures are inserted * and the result is returned. |
ca23e405e vmalloc: implemen... |
3425 3426 3427 |
*/ struct vm_struct **pcpu_get_vm_areas(const unsigned long *offsets, const size_t *sizes, int nr_vms, |
ec3f64fc9 mm: remove gfp ma... |
3428 |
size_t align) |
ca23e405e vmalloc: implemen... |
3429 3430 3431 |
{ const unsigned long vmalloc_start = ALIGN(VMALLOC_START, align); const unsigned long vmalloc_end = VMALLOC_END & ~(align - 1); |
68ad4a330 mm/vmalloc.c: kee... |
3432 |
struct vmap_area **vas, *va; |
ca23e405e vmalloc: implemen... |
3433 3434 |
struct vm_struct **vms; int area, area2, last_area, term_area; |
253a496d8 kasan: don't assu... |
3435 |
unsigned long base, start, size, end, last_end, orig_start, orig_end; |
ca23e405e vmalloc: implemen... |
3436 |
bool purged = false; |
68ad4a330 mm/vmalloc.c: kee... |
3437 |
enum fit_type type; |
ca23e405e vmalloc: implemen... |
3438 |
|
ca23e405e vmalloc: implemen... |
3439 |
/* verify parameters and allocate data structures */ |
891c49abf mm/vmalloc: use o... |
3440 |
BUG_ON(offset_in_page(align) || !is_power_of_2(align)); |
ca23e405e vmalloc: implemen... |
3441 3442 3443 3444 3445 3446 3447 3448 3449 3450 3451 |
for (last_area = 0, area = 0; area < nr_vms; area++) { start = offsets[area]; end = start + sizes[area]; /* is everything aligned properly? */ BUG_ON(!IS_ALIGNED(offsets[area], align)); BUG_ON(!IS_ALIGNED(sizes[area], align)); /* detect the area with the highest address */ if (start > offsets[last_area]) last_area = area; |
c568da282 mm/vmalloc.c: hal... |
3452 |
for (area2 = area + 1; area2 < nr_vms; area2++) { |
ca23e405e vmalloc: implemen... |
3453 3454 |
unsigned long start2 = offsets[area2]; unsigned long end2 = start2 + sizes[area2]; |
c568da282 mm/vmalloc.c: hal... |
3455 |
BUG_ON(start2 < end && start < end2); |
ca23e405e vmalloc: implemen... |
3456 3457 3458 3459 3460 3461 3462 3463 |
} } last_end = offsets[last_area] + sizes[last_area]; if (vmalloc_end - vmalloc_start < last_end) { WARN_ON(true); return NULL; } |
4d67d8605 mm: use kcalloc()... |
3464 3465 |
vms = kcalloc(nr_vms, sizeof(vms[0]), GFP_KERNEL); vas = kcalloc(nr_vms, sizeof(vas[0]), GFP_KERNEL); |
ca23e405e vmalloc: implemen... |
3466 |
if (!vas || !vms) |
f1db7afd9 mm/vmalloc.c: eli... |
3467 |
goto err_free2; |
ca23e405e vmalloc: implemen... |
3468 3469 |
for (area = 0; area < nr_vms; area++) { |
68ad4a330 mm/vmalloc.c: kee... |
3470 |
vas[area] = kmem_cache_zalloc(vmap_area_cachep, GFP_KERNEL); |
ec3f64fc9 mm: remove gfp ma... |
3471 |
vms[area] = kzalloc(sizeof(struct vm_struct), GFP_KERNEL); |
ca23e405e vmalloc: implemen... |
3472 3473 3474 3475 |
if (!vas[area] || !vms[area]) goto err_free; } retry: |
e36176be1 mm/vmalloc: rewor... |
3476 |
spin_lock(&free_vmap_area_lock); |
ca23e405e vmalloc: implemen... |
3477 3478 3479 3480 3481 |
/* start scanning - we scan from the top, begin with the last area */ area = term_area = last_area; start = offsets[area]; end = start + sizes[area]; |
68ad4a330 mm/vmalloc.c: kee... |
3482 3483 |
va = pvm_find_va_enclose_addr(vmalloc_end); base = pvm_determine_end_from_reverse(&va, align) - end; |
ca23e405e vmalloc: implemen... |
3484 3485 |
while (true) { |
ca23e405e vmalloc: implemen... |
3486 3487 3488 3489 |
/* * base might have underflowed, add last_end before * comparing. */ |
68ad4a330 mm/vmalloc.c: kee... |
3490 3491 |
if (base + last_end < vmalloc_start + last_end) goto overflow; |
ca23e405e vmalloc: implemen... |
3492 3493 |
/* |
68ad4a330 mm/vmalloc.c: kee... |
3494 |
* Fitting base has not been found. |
ca23e405e vmalloc: implemen... |
3495 |
*/ |
68ad4a330 mm/vmalloc.c: kee... |
3496 3497 |
if (va == NULL) goto overflow; |
ca23e405e vmalloc: implemen... |
3498 3499 |
/* |
d8cc323d9 mm/vmalloc: fix a... |
3500 |
* If required width exceeds current VA block, move |
5336e52c9 mm/vmalloc.c: fix... |
3501 3502 3503 3504 3505 3506 3507 3508 3509 |
* base downwards and then recheck. */ if (base + end > va->va_end) { base = pvm_determine_end_from_reverse(&va, align) - end; term_area = area; continue; } /* |
68ad4a330 mm/vmalloc.c: kee... |
3510 |
* If this VA does not fit, move base downwards and recheck. |
ca23e405e vmalloc: implemen... |
3511 |
*/ |
5336e52c9 mm/vmalloc.c: fix... |
3512 |
if (base + start < va->va_start) { |
68ad4a330 mm/vmalloc.c: kee... |
3513 3514 |
va = node_to_va(rb_prev(&va->rb_node)); base = pvm_determine_end_from_reverse(&va, align) - end; |
ca23e405e vmalloc: implemen... |
3515 3516 3517 3518 3519 3520 3521 3522 3523 3524 3525 |
term_area = area; continue; } /* * This area fits, move on to the previous one. If * the previous one is the terminal one, we're done. */ area = (area + nr_vms - 1) % nr_vms; if (area == term_area) break; |
68ad4a330 mm/vmalloc.c: kee... |
3526 |
|
ca23e405e vmalloc: implemen... |
3527 3528 |
start = offsets[area]; end = start + sizes[area]; |
68ad4a330 mm/vmalloc.c: kee... |
3529 |
va = pvm_find_va_enclose_addr(base + end); |
ca23e405e vmalloc: implemen... |
3530 |
} |
68ad4a330 mm/vmalloc.c: kee... |
3531 |
|
ca23e405e vmalloc: implemen... |
3532 3533 |
/* we've found a fitting base, insert all va's */ for (area = 0; area < nr_vms; area++) { |
68ad4a330 mm/vmalloc.c: kee... |
3534 |
int ret; |
ca23e405e vmalloc: implemen... |
3535 |
|
68ad4a330 mm/vmalloc.c: kee... |
3536 3537 |
start = base + offsets[area]; size = sizes[area]; |
ca23e405e vmalloc: implemen... |
3538 |
|
68ad4a330 mm/vmalloc.c: kee... |
3539 3540 3541 3542 3543 3544 3545 3546 3547 3548 3549 3550 3551 3552 3553 3554 3555 3556 |
va = pvm_find_va_enclose_addr(start); if (WARN_ON_ONCE(va == NULL)) /* It is a BUG(), but trigger recovery instead. */ goto recovery; type = classify_va_fit_type(va, start, size); if (WARN_ON_ONCE(type == NOTHING_FIT)) /* It is a BUG(), but trigger recovery instead. */ goto recovery; ret = adjust_va_to_fit_type(va, start, size, type); if (unlikely(ret)) goto recovery; /* Allocated area. */ va = vas[area]; va->va_start = start; va->va_end = start + size; |
68ad4a330 mm/vmalloc.c: kee... |
3557 |
} |
ca23e405e vmalloc: implemen... |
3558 |
|
e36176be1 mm/vmalloc: rewor... |
3559 |
spin_unlock(&free_vmap_area_lock); |
ca23e405e vmalloc: implemen... |
3560 |
|
253a496d8 kasan: don't assu... |
3561 3562 3563 3564 3565 3566 3567 3568 |
/* populate the kasan shadow space */ for (area = 0; area < nr_vms; area++) { if (kasan_populate_vmalloc(vas[area]->va_start, sizes[area])) goto err_free_shadow; kasan_unpoison_vmalloc((void *)vas[area]->va_start, sizes[area]); } |
ca23e405e vmalloc: implemen... |
3569 |
/* insert all vm's */ |
e36176be1 mm/vmalloc: rewor... |
3570 3571 3572 3573 3574 |
spin_lock(&vmap_area_lock); for (area = 0; area < nr_vms; area++) { insert_vmap_area(vas[area], &vmap_area_root, &vmap_area_list); setup_vmalloc_vm_locked(vms[area], vas[area], VM_ALLOC, |
3645cb4a4 mm, vmalloc: call... |
3575 |
pcpu_get_vm_areas); |
e36176be1 mm/vmalloc: rewor... |
3576 3577 |
} spin_unlock(&vmap_area_lock); |
ca23e405e vmalloc: implemen... |
3578 3579 3580 |
kfree(vas); return vms; |
68ad4a330 mm/vmalloc.c: kee... |
3581 |
recovery: |
e36176be1 mm/vmalloc: rewor... |
3582 3583 3584 3585 3586 3587 |
/* * Remove previously allocated areas. There is no * need in removing these areas from the busy tree, * because they are inserted only on the final step * and when pcpu_get_vm_areas() is success. */ |
68ad4a330 mm/vmalloc.c: kee... |
3588 |
while (area--) { |
253a496d8 kasan: don't assu... |
3589 3590 |
orig_start = vas[area]->va_start; orig_end = vas[area]->va_end; |
96e2db456 mm/vmalloc: rewor... |
3591 3592 |
va = merge_or_add_vmap_area_augment(vas[area], &free_vmap_area_root, &free_vmap_area_list); |
9c801f61d mm/vmalloc.c: rem... |
3593 3594 3595 |
if (va) kasan_release_vmalloc(orig_start, orig_end, va->va_start, va->va_end); |
68ad4a330 mm/vmalloc.c: kee... |
3596 3597 3598 3599 |
vas[area] = NULL; } overflow: |
e36176be1 mm/vmalloc: rewor... |
3600 |
spin_unlock(&free_vmap_area_lock); |
68ad4a330 mm/vmalloc.c: kee... |
3601 3602 3603 3604 3605 3606 3607 3608 3609 3610 3611 3612 3613 3614 3615 3616 3617 |
if (!purged) { purge_vmap_area_lazy(); purged = true; /* Before "retry", check if we recover. */ for (area = 0; area < nr_vms; area++) { if (vas[area]) continue; vas[area] = kmem_cache_zalloc( vmap_area_cachep, GFP_KERNEL); if (!vas[area]) goto err_free; } goto retry; } |
ca23e405e vmalloc: implemen... |
3618 3619 |
err_free: for (area = 0; area < nr_vms; area++) { |
68ad4a330 mm/vmalloc.c: kee... |
3620 3621 |
if (vas[area]) kmem_cache_free(vmap_area_cachep, vas[area]); |
f1db7afd9 mm/vmalloc.c: eli... |
3622 |
kfree(vms[area]); |
ca23e405e vmalloc: implemen... |
3623 |
} |
f1db7afd9 mm/vmalloc.c: eli... |
3624 |
err_free2: |
ca23e405e vmalloc: implemen... |
3625 3626 3627 |
kfree(vas); kfree(vms); return NULL; |
253a496d8 kasan: don't assu... |
3628 3629 3630 3631 3632 3633 3634 3635 3636 3637 3638 |
err_free_shadow: spin_lock(&free_vmap_area_lock); /* * We release all the vmalloc shadows, even the ones for regions that * hadn't been successfully added. This relies on kasan_release_vmalloc * being able to tolerate this case. */ for (area = 0; area < nr_vms; area++) { orig_start = vas[area]->va_start; orig_end = vas[area]->va_end; |
96e2db456 mm/vmalloc: rewor... |
3639 3640 |
va = merge_or_add_vmap_area_augment(vas[area], &free_vmap_area_root, &free_vmap_area_list); |
9c801f61d mm/vmalloc.c: rem... |
3641 3642 3643 |
if (va) kasan_release_vmalloc(orig_start, orig_end, va->va_start, va->va_end); |
253a496d8 kasan: don't assu... |
3644 3645 3646 3647 3648 3649 3650 |
vas[area] = NULL; kfree(vms[area]); } spin_unlock(&free_vmap_area_lock); kfree(vas); kfree(vms); return NULL; |
ca23e405e vmalloc: implemen... |
3651 3652 3653 3654 3655 3656 3657 3658 3659 3660 3661 3662 3663 3664 3665 3666 3667 |
} /** * 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... |
3668 |
#endif /* CONFIG_SMP */ |
a10aa5798 vmalloc: show vma... |
3669 |
|
5bb1bb353 mm: Don't build m... |
3670 |
#ifdef CONFIG_PRINTK |
98f180837 mm: Make mem_dump... |
3671 3672 3673 3674 3675 3676 3677 3678 |
bool vmalloc_dump_obj(void *object) { struct vm_struct *vm; void *objp = (void *)PAGE_ALIGN((unsigned long)object); vm = find_vm_area(objp); if (!vm) return false; |
bd34dcd41 mm: Make mem_obj_... |
3679 3680 3681 |
pr_cont(" %u-page vmalloc region starting at %#lx allocated at %pS ", vm->nr_pages, (unsigned long)vm->addr, vm->caller); |
98f180837 mm: Make mem_dump... |
3682 3683 |
return true; } |
5bb1bb353 mm: Don't build m... |
3684 |
#endif |
98f180837 mm: Make mem_dump... |
3685 |
|
a10aa5798 vmalloc: show vma... |
3686 3687 |
#ifdef CONFIG_PROC_FS static void *s_start(struct seq_file *m, loff_t *pos) |
e36176be1 mm/vmalloc: rewor... |
3688 |
__acquires(&vmap_purge_lock) |
d4033afdf mm, vmalloc: iter... |
3689 |
__acquires(&vmap_area_lock) |
a10aa5798 vmalloc: show vma... |
3690 |
{ |
e36176be1 mm/vmalloc: rewor... |
3691 |
mutex_lock(&vmap_purge_lock); |
d4033afdf mm, vmalloc: iter... |
3692 |
spin_lock(&vmap_area_lock); |
e36176be1 mm/vmalloc: rewor... |
3693 |
|
3f5000693 mm/vmalloc.c: sim... |
3694 |
return seq_list_start(&vmap_area_list, *pos); |
a10aa5798 vmalloc: show vma... |
3695 3696 3697 3698 |
} static void *s_next(struct seq_file *m, void *p, loff_t *pos) { |
3f5000693 mm/vmalloc.c: sim... |
3699 |
return seq_list_next(p, &vmap_area_list, pos); |
a10aa5798 vmalloc: show vma... |
3700 3701 3702 |
} static void s_stop(struct seq_file *m, void *p) |
d4033afdf mm, vmalloc: iter... |
3703 |
__releases(&vmap_area_lock) |
0a7dd4e90 mm/vmalloc: Fix u... |
3704 |
__releases(&vmap_purge_lock) |
a10aa5798 vmalloc: show vma... |
3705 |
{ |
d4033afdf mm, vmalloc: iter... |
3706 |
spin_unlock(&vmap_area_lock); |
0a7dd4e90 mm/vmalloc: Fix u... |
3707 |
mutex_unlock(&vmap_purge_lock); |
a10aa5798 vmalloc: show vma... |
3708 |
} |
a47a126ad vmallocinfo: add ... |
3709 3710 |
static void show_numa_info(struct seq_file *m, struct vm_struct *v) { |
e5adfffc8 mm: use IS_ENABLE... |
3711 |
if (IS_ENABLED(CONFIG_NUMA)) { |
a47a126ad vmallocinfo: add ... |
3712 3713 3714 3715 |
unsigned int nr, *counters = m->private; if (!counters) return; |
af12346cd mm/vmalloc: rever... |
3716 3717 |
if (v->flags & VM_UNINITIALIZED) return; |
7e5b528b4 mm/vmalloc.c: fix... |
3718 3719 |
/* Pair with smp_wmb() in clear_vm_uninitialized_flag() */ smp_rmb(); |
af12346cd mm/vmalloc: rever... |
3720 |
|
a47a126ad vmallocinfo: add ... |
3721 3722 3723 3724 3725 3726 3727 3728 3729 3730 |
memset(counters, 0, nr_node_ids * sizeof(unsigned int)); for (nr = 0; nr < v->nr_pages; nr++) counters[page_to_nid(v->pages[nr])]++; for_each_node_state(nr, N_HIGH_MEMORY) if (counters[nr]) seq_printf(m, " N%u=%u", nr, counters[nr]); } } |
dd3b8353b mm/vmalloc: do no... |
3731 3732 |
static void show_purge_info(struct seq_file *m) { |
dd3b8353b mm/vmalloc: do no... |
3733 |
struct vmap_area *va; |
96e2db456 mm/vmalloc: rewor... |
3734 3735 |
spin_lock(&purge_vmap_area_lock); list_for_each_entry(va, &purge_vmap_area_list, list) { |
dd3b8353b mm/vmalloc: do no... |
3736 3737 3738 3739 3740 |
seq_printf(m, "0x%pK-0x%pK %7ld unpurged vm_area ", (void *)va->va_start, (void *)va->va_end, va->va_end - va->va_start); } |
96e2db456 mm/vmalloc: rewor... |
3741 |
spin_unlock(&purge_vmap_area_lock); |
dd3b8353b mm/vmalloc: do no... |
3742 |
} |
a10aa5798 vmalloc: show vma... |
3743 3744 |
static int s_show(struct seq_file *m, void *p) { |
3f5000693 mm/vmalloc.c: sim... |
3745 |
struct vmap_area *va; |
d4033afdf mm, vmalloc: iter... |
3746 |
struct vm_struct *v; |
3f5000693 mm/vmalloc.c: sim... |
3747 |
va = list_entry(p, struct vmap_area, list); |
c2ce8c142 mm/vmalloc: fix s... |
3748 |
/* |
688fcbfc0 mm/vmalloc: modif... |
3749 3750 |
* s_show can encounter race with remove_vm_area, !vm on behalf * of vmap area is being tear down or vm_map_ram allocation. |
c2ce8c142 mm/vmalloc: fix s... |
3751 |
*/ |
688fcbfc0 mm/vmalloc: modif... |
3752 |
if (!va->vm) { |
dd3b8353b mm/vmalloc: do no... |
3753 3754 |
seq_printf(m, "0x%pK-0x%pK %7ld vm_map_ram ", |
78c72746f vmalloc: show laz... |
3755 |
(void *)va->va_start, (void *)va->va_end, |
dd3b8353b mm/vmalloc: do no... |
3756 |
va->va_end - va->va_start); |
78c72746f vmalloc: show laz... |
3757 |
|
d4033afdf mm, vmalloc: iter... |
3758 |
return 0; |
78c72746f vmalloc: show laz... |
3759 |
} |
d4033afdf mm, vmalloc: iter... |
3760 3761 |
v = va->vm; |
a10aa5798 vmalloc: show vma... |
3762 |
|
45ec16908 mm: use %pK for /... |
3763 |
seq_printf(m, "0x%pK-0x%pK %7ld", |
a10aa5798 vmalloc: show vma... |
3764 |
v->addr, v->addr + v->size, v->size); |
62c70bce8 mm: convert sprin... |
3765 3766 |
if (v->caller) seq_printf(m, " %pS", v->caller); |
230169693 vmallocinfo: add ... |
3767 |
|
a10aa5798 vmalloc: show vma... |
3768 3769 3770 3771 |
if (v->nr_pages) seq_printf(m, " pages=%d", v->nr_pages); if (v->phys_addr) |
199eaa05a mm: cleanups for ... |
3772 |
seq_printf(m, " phys=%pa", &v->phys_addr); |
a10aa5798 vmalloc: show vma... |
3773 3774 |
if (v->flags & VM_IOREMAP) |
f4527c908 mm/vmalloc.c: rep... |
3775 |
seq_puts(m, " ioremap"); |
a10aa5798 vmalloc: show vma... |
3776 3777 |
if (v->flags & VM_ALLOC) |
f4527c908 mm/vmalloc.c: rep... |
3778 |
seq_puts(m, " vmalloc"); |
a10aa5798 vmalloc: show vma... |
3779 3780 |
if (v->flags & VM_MAP) |
f4527c908 mm/vmalloc.c: rep... |
3781 |
seq_puts(m, " vmap"); |
a10aa5798 vmalloc: show vma... |
3782 3783 |
if (v->flags & VM_USERMAP) |
f4527c908 mm/vmalloc.c: rep... |
3784 |
seq_puts(m, " user"); |
a10aa5798 vmalloc: show vma... |
3785 |
|
fe9041c24 vmalloc: lift the... |
3786 3787 |
if (v->flags & VM_DMA_COHERENT) seq_puts(m, " dma-coherent"); |
244d63ee3 mm, vmalloc: remo... |
3788 |
if (is_vmalloc_addr(v->pages)) |
f4527c908 mm/vmalloc.c: rep... |
3789 |
seq_puts(m, " vpages"); |
a10aa5798 vmalloc: show vma... |
3790 |
|
a47a126ad vmallocinfo: add ... |
3791 |
show_numa_info(m, v); |
a10aa5798 vmalloc: show vma... |
3792 3793 |
seq_putc(m, ' '); |
dd3b8353b mm/vmalloc: do no... |
3794 3795 |
/* |
96e2db456 mm/vmalloc: rewor... |
3796 |
* As a final step, dump "unpurged" areas. |
dd3b8353b mm/vmalloc: do no... |
3797 3798 3799 |
*/ if (list_is_last(&va->list, &vmap_area_list)) show_purge_info(m); |
a10aa5798 vmalloc: show vma... |
3800 3801 |
return 0; } |
5f6a6a9c4 proc: move /proc/... |
3802 |
static const struct seq_operations vmalloc_op = { |
a10aa5798 vmalloc: show vma... |
3803 3804 3805 3806 3807 |
.start = s_start, .next = s_next, .stop = s_stop, .show = s_show, }; |
5f6a6a9c4 proc: move /proc/... |
3808 |
|
5f6a6a9c4 proc: move /proc/... |
3809 3810 |
static int __init proc_vmalloc_init(void) { |
fddda2b7b proc: introduce p... |
3811 |
if (IS_ENABLED(CONFIG_NUMA)) |
0825a6f98 mm: use octal not... |
3812 |
proc_create_seq_private("vmallocinfo", 0400, NULL, |
44414d82c proc: introduce p... |
3813 3814 |
&vmalloc_op, nr_node_ids * sizeof(unsigned int), NULL); |
fddda2b7b proc: introduce p... |
3815 |
else |
0825a6f98 mm: use octal not... |
3816 |
proc_create_seq("vmallocinfo", 0400, NULL, &vmalloc_op); |
5f6a6a9c4 proc: move /proc/... |
3817 3818 3819 |
return 0; } module_init(proc_vmalloc_init); |
db3808c1b mm, vmalloc: move... |
3820 |
|
a10aa5798 vmalloc: show vma... |
3821 |
#endif |