vmalloc.c 59.2 KB
Newer Older
L
Linus Torvalds 已提交
1 2 3 4 5 6 7
/*
 *  linux/mm/vmalloc.c
 *
 *  Copyright (C) 1993  Linus Torvalds
 *  Support of BIGMEM added by Gerhard Wichert, Siemens AG, July 1999
 *  SMP-safe vmalloc/vfree/ioremap, Tigran Aivazian <tigran@veritas.com>, May 2000
 *  Major rework to support vmap/vunmap, Christoph Hellwig, SGI, August 2002
C
Christoph Lameter 已提交
8
 *  Numa awareness, Christoph Lameter, SGI, June 2005
L
Linus Torvalds 已提交
9 10
 */

N
Nick Piggin 已提交
11
#include <linux/vmalloc.h>
L
Linus Torvalds 已提交
12 13 14 15 16 17
#include <linux/mm.h>
#include <linux/module.h>
#include <linux/highmem.h>
#include <linux/slab.h>
#include <linux/spinlock.h>
#include <linux/interrupt.h>
18
#include <linux/proc_fs.h>
19
#include <linux/seq_file.h>
20
#include <linux/debugobjects.h>
21
#include <linux/kallsyms.h>
N
Nick Piggin 已提交
22 23 24 25
#include <linux/list.h>
#include <linux/rbtree.h>
#include <linux/radix-tree.h>
#include <linux/rcupdate.h>
26
#include <linux/pfn.h>
27
#include <linux/kmemleak.h>
28
#include <linux/highmem.h>
N
Nick Piggin 已提交
29
#include <asm/atomic.h>
L
Linus Torvalds 已提交
30 31 32 33
#include <asm/uaccess.h>
#include <asm/tlbflush.h>


N
Nick Piggin 已提交
34
/*** Page table manipulation functions ***/
A
Adrian Bunk 已提交
35

L
Linus Torvalds 已提交
36 37 38 39 40 41 42 43 44 45 46
static void vunmap_pte_range(pmd_t *pmd, unsigned long addr, unsigned long end)
{
	pte_t *pte;

	pte = pte_offset_kernel(pmd, addr);
	do {
		pte_t ptent = ptep_get_and_clear(&init_mm, addr, pte);
		WARN_ON(!pte_none(ptent) && !pte_present(ptent));
	} while (pte++, addr += PAGE_SIZE, addr != end);
}

N
Nick Piggin 已提交
47
static void vunmap_pmd_range(pud_t *pud, unsigned long addr, unsigned long end)
L
Linus Torvalds 已提交
48 49 50 51 52 53 54 55 56 57 58 59 60
{
	pmd_t *pmd;
	unsigned long next;

	pmd = pmd_offset(pud, addr);
	do {
		next = pmd_addr_end(addr, end);
		if (pmd_none_or_clear_bad(pmd))
			continue;
		vunmap_pte_range(pmd, addr, next);
	} while (pmd++, addr = next, addr != end);
}

N
Nick Piggin 已提交
61
static void vunmap_pud_range(pgd_t *pgd, unsigned long addr, unsigned long end)
L
Linus Torvalds 已提交
62 63 64 65 66 67 68 69 70 71 72 73 74
{
	pud_t *pud;
	unsigned long next;

	pud = pud_offset(pgd, addr);
	do {
		next = pud_addr_end(addr, end);
		if (pud_none_or_clear_bad(pud))
			continue;
		vunmap_pmd_range(pud, addr, next);
	} while (pud++, addr = next, addr != end);
}

N
Nick Piggin 已提交
75
static void vunmap_page_range(unsigned long addr, unsigned long end)
L
Linus Torvalds 已提交
76 77 78 79 80 81 82 83 84 85 86 87 88 89 90
{
	pgd_t *pgd;
	unsigned long next;

	BUG_ON(addr >= end);
	pgd = pgd_offset_k(addr);
	do {
		next = pgd_addr_end(addr, end);
		if (pgd_none_or_clear_bad(pgd))
			continue;
		vunmap_pud_range(pgd, addr, next);
	} while (pgd++, addr = next, addr != end);
}

static int vmap_pte_range(pmd_t *pmd, unsigned long addr,
N
Nick Piggin 已提交
91
		unsigned long end, pgprot_t prot, struct page **pages, int *nr)
L
Linus Torvalds 已提交
92 93 94
{
	pte_t *pte;

N
Nick Piggin 已提交
95 96 97 98 99
	/*
	 * nr is a running index into the array which helps higher level
	 * callers keep track of where we're up to.
	 */

H
Hugh Dickins 已提交
100
	pte = pte_alloc_kernel(pmd, addr);
L
Linus Torvalds 已提交
101 102 103
	if (!pte)
		return -ENOMEM;
	do {
N
Nick Piggin 已提交
104 105 106 107 108
		struct page *page = pages[*nr];

		if (WARN_ON(!pte_none(*pte)))
			return -EBUSY;
		if (WARN_ON(!page))
L
Linus Torvalds 已提交
109 110
			return -ENOMEM;
		set_pte_at(&init_mm, addr, pte, mk_pte(page, prot));
N
Nick Piggin 已提交
111
		(*nr)++;
L
Linus Torvalds 已提交
112 113 114 115
	} while (pte++, addr += PAGE_SIZE, addr != end);
	return 0;
}

N
Nick Piggin 已提交
116 117
static int vmap_pmd_range(pud_t *pud, unsigned long addr,
		unsigned long end, pgprot_t prot, struct page **pages, int *nr)
L
Linus Torvalds 已提交
118 119 120 121 122 123 124 125 126
{
	pmd_t *pmd;
	unsigned long next;

	pmd = pmd_alloc(&init_mm, pud, addr);
	if (!pmd)
		return -ENOMEM;
	do {
		next = pmd_addr_end(addr, end);
N
Nick Piggin 已提交
127
		if (vmap_pte_range(pmd, addr, next, prot, pages, nr))
L
Linus Torvalds 已提交
128 129 130 131 132
			return -ENOMEM;
	} while (pmd++, addr = next, addr != end);
	return 0;
}

N
Nick Piggin 已提交
133 134
static int vmap_pud_range(pgd_t *pgd, unsigned long addr,
		unsigned long end, pgprot_t prot, struct page **pages, int *nr)
L
Linus Torvalds 已提交
135 136 137 138 139 140 141 142 143
{
	pud_t *pud;
	unsigned long next;

	pud = pud_alloc(&init_mm, pgd, addr);
	if (!pud)
		return -ENOMEM;
	do {
		next = pud_addr_end(addr, end);
N
Nick Piggin 已提交
144
		if (vmap_pmd_range(pud, addr, next, prot, pages, nr))
L
Linus Torvalds 已提交
145 146 147 148 149
			return -ENOMEM;
	} while (pud++, addr = next, addr != end);
	return 0;
}

N
Nick Piggin 已提交
150 151 152 153 154 155
/*
 * Set up page tables in kva (addr, end). The ptes shall have prot "prot", and
 * will have pfns corresponding to the "pages" array.
 *
 * Ie. pte at addr+N*PAGE_SIZE shall point to pfn corresponding to pages[N]
 */
156 157
static int vmap_page_range_noflush(unsigned long start, unsigned long end,
				   pgprot_t prot, struct page **pages)
L
Linus Torvalds 已提交
158 159 160
{
	pgd_t *pgd;
	unsigned long next;
161
	unsigned long addr = start;
N
Nick Piggin 已提交
162 163
	int err = 0;
	int nr = 0;
L
Linus Torvalds 已提交
164 165 166 167 168

	BUG_ON(addr >= end);
	pgd = pgd_offset_k(addr);
	do {
		next = pgd_addr_end(addr, end);
N
Nick Piggin 已提交
169
		err = vmap_pud_range(pgd, addr, next, prot, pages, &nr);
L
Linus Torvalds 已提交
170
		if (err)
171
			return err;
L
Linus Torvalds 已提交
172
	} while (pgd++, addr = next, addr != end);
N
Nick Piggin 已提交
173 174

	return nr;
L
Linus Torvalds 已提交
175 176
}

177 178 179 180 181 182 183 184 185 186
static int vmap_page_range(unsigned long start, unsigned long end,
			   pgprot_t prot, struct page **pages)
{
	int ret;

	ret = vmap_page_range_noflush(start, end, prot, pages);
	flush_cache_vmap(start, end);
	return ret;
}

187 188 189
static inline int is_vmalloc_or_module_addr(const void *x)
{
	/*
190
	 * ARM, x86-64 and sparc64 put modules in a special place,
191 192 193 194 195 196 197 198 199 200 201
	 * 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);
}

202
/*
N
Nick Piggin 已提交
203
 * Walk a vmap address to the struct page it maps.
204
 */
205
struct page *vmalloc_to_page(const void *vmalloc_addr)
206 207 208 209 210
{
	unsigned long addr = (unsigned long) vmalloc_addr;
	struct page *page = NULL;
	pgd_t *pgd = pgd_offset_k(addr);

211 212 213 214
	/*
	 * XXX we might need to change this if we add VIRTUAL_BUG_ON for
	 * architectures that do not vmalloc module space
	 */
215
	VIRTUAL_BUG_ON(!is_vmalloc_or_module_addr(vmalloc_addr));
J
Jiri Slaby 已提交
216

217
	if (!pgd_none(*pgd)) {
N
Nick Piggin 已提交
218
		pud_t *pud = pud_offset(pgd, addr);
219
		if (!pud_none(*pud)) {
N
Nick Piggin 已提交
220
			pmd_t *pmd = pmd_offset(pud, addr);
221
			if (!pmd_none(*pmd)) {
N
Nick Piggin 已提交
222 223
				pte_t *ptep, pte;

224 225 226 227 228 229 230 231 232 233 234 235 236 237 238
				ptep = pte_offset_map(pmd, addr);
				pte = *ptep;
				if (pte_present(pte))
					page = pte_page(pte);
				pte_unmap(ptep);
			}
		}
	}
	return page;
}
EXPORT_SYMBOL(vmalloc_to_page);

/*
 * Map a vmalloc()-space virtual address to the physical page frame number.
 */
239
unsigned long vmalloc_to_pfn(const void *vmalloc_addr)
240 241 242 243 244
{
	return page_to_pfn(vmalloc_to_page(vmalloc_addr));
}
EXPORT_SYMBOL(vmalloc_to_pfn);

N
Nick Piggin 已提交
245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265

/*** Global kva allocator ***/

#define VM_LAZY_FREE	0x01
#define VM_LAZY_FREEING	0x02
#define VM_VM_AREA	0x04

struct vmap_area {
	unsigned long va_start;
	unsigned long va_end;
	unsigned long flags;
	struct rb_node rb_node;		/* address sorted rbtree */
	struct list_head list;		/* address sorted list */
	struct list_head purge_list;	/* "lazy purge" list */
	void *private;
	struct rcu_head rcu_head;
};

static DEFINE_SPINLOCK(vmap_area_lock);
static struct rb_root vmap_area_root = RB_ROOT;
static LIST_HEAD(vmap_area_list);
266
static unsigned long vmap_area_pcpu_hole;
N
Nick Piggin 已提交
267 268

static struct vmap_area *__find_vmap_area(unsigned long addr)
L
Linus Torvalds 已提交
269
{
N
Nick Piggin 已提交
270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331
	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;
		else if (addr > va->va_start)
			n = n->rb_right;
		else
			return va;
	}

	return NULL;
}

static void __insert_vmap_area(struct vmap_area *va)
{
	struct rb_node **p = &vmap_area_root.rb_node;
	struct rb_node *parent = NULL;
	struct rb_node *tmp;

	while (*p) {
		struct vmap_area *tmp;

		parent = *p;
		tmp = rb_entry(parent, struct vmap_area, rb_node);
		if (va->va_start < tmp->va_end)
			p = &(*p)->rb_left;
		else if (va->va_end > tmp->va_start)
			p = &(*p)->rb_right;
		else
			BUG();
	}

	rb_link_node(&va->rb_node, parent, p);
	rb_insert_color(&va->rb_node, &vmap_area_root);

	/* address-sort this list so it is usable like the vmlist */
	tmp = rb_prev(&va->rb_node);
	if (tmp) {
		struct vmap_area *prev;
		prev = rb_entry(tmp, struct vmap_area, rb_node);
		list_add_rcu(&va->list, &prev->list);
	} else
		list_add_rcu(&va->list, &vmap_area_list);
}

static void purge_vmap_area_lazy(void);

/*
 * Allocate a region of KVA of the specified size and alignment, within the
 * vstart and vend.
 */
static struct vmap_area *alloc_vmap_area(unsigned long size,
				unsigned long align,
				unsigned long vstart, unsigned long vend,
				int node, gfp_t gfp_mask)
{
	struct vmap_area *va;
	struct rb_node *n;
L
Linus Torvalds 已提交
332
	unsigned long addr;
N
Nick Piggin 已提交
333 334
	int purged = 0;

N
Nick Piggin 已提交
335
	BUG_ON(!size);
N
Nick Piggin 已提交
336 337 338 339 340 341 342 343
	BUG_ON(size & ~PAGE_MASK);

	va = kmalloc_node(sizeof(struct vmap_area),
			gfp_mask & GFP_RECLAIM_MASK, node);
	if (unlikely(!va))
		return ERR_PTR(-ENOMEM);

retry:
G
Glauber Costa 已提交
344 345
	addr = ALIGN(vstart, align);

N
Nick Piggin 已提交
346
	spin_lock(&vmap_area_lock);
N
Nick Piggin 已提交
347 348 349
	if (addr + size - 1 < addr)
		goto overflow;

N
Nick Piggin 已提交
350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378
	/* XXX: could have a last_hole cache */
	n = vmap_area_root.rb_node;
	if (n) {
		struct vmap_area *first = NULL;

		do {
			struct vmap_area *tmp;
			tmp = rb_entry(n, struct vmap_area, rb_node);
			if (tmp->va_end >= addr) {
				if (!first && tmp->va_start < addr + size)
					first = tmp;
				n = n->rb_left;
			} else {
				first = tmp;
				n = n->rb_right;
			}
		} while (n);

		if (!first)
			goto found;

		if (first->va_end < addr) {
			n = rb_next(&first->rb_node);
			if (n)
				first = rb_entry(n, struct vmap_area, rb_node);
			else
				goto found;
		}

N
Nick Piggin 已提交
379
		while (addr + size > first->va_start && addr + size <= vend) {
N
Nick Piggin 已提交
380
			addr = ALIGN(first->va_end + PAGE_SIZE, align);
N
Nick Piggin 已提交
381 382
			if (addr + size - 1 < addr)
				goto overflow;
N
Nick Piggin 已提交
383 384 385 386 387 388 389 390 391 392

			n = rb_next(&first->rb_node);
			if (n)
				first = rb_entry(n, struct vmap_area, rb_node);
			else
				goto found;
		}
	}
found:
	if (addr + size > vend) {
N
Nick Piggin 已提交
393
overflow:
N
Nick Piggin 已提交
394 395 396 397 398 399 400
		spin_unlock(&vmap_area_lock);
		if (!purged) {
			purge_vmap_area_lazy();
			purged = 1;
			goto retry;
		}
		if (printk_ratelimit())
G
Glauber Costa 已提交
401 402 403
			printk(KERN_WARNING
				"vmap allocation for size %lu failed: "
				"use vmalloc=<size> to increase size.\n", size);
404
		kfree(va);
N
Nick Piggin 已提交
405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432
		return ERR_PTR(-EBUSY);
	}

	BUG_ON(addr & (align-1));

	va->va_start = addr;
	va->va_end = addr + size;
	va->flags = 0;
	__insert_vmap_area(va);
	spin_unlock(&vmap_area_lock);

	return va;
}

static void rcu_free_va(struct rcu_head *head)
{
	struct vmap_area *va = container_of(head, struct vmap_area, rcu_head);

	kfree(va);
}

static void __free_vmap_area(struct vmap_area *va)
{
	BUG_ON(RB_EMPTY_NODE(&va->rb_node));
	rb_erase(&va->rb_node, &vmap_area_root);
	RB_CLEAR_NODE(&va->rb_node);
	list_del_rcu(&va->list);

433 434 435 436 437 438 439 440 441
	/*
	 * Track the highest possible candidate for pcpu area
	 * allocation.  Areas outside of vmalloc area can be returned
	 * here too, consider only end addresses which fall inside
	 * vmalloc area proper.
	 */
	if (va->va_end > VMALLOC_START && va->va_end <= VMALLOC_END)
		vmap_area_pcpu_hole = max(vmap_area_pcpu_hole, va->va_end);

N
Nick Piggin 已提交
442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462
	call_rcu(&va->rcu_head, rcu_free_va);
}

/*
 * Free a region of KVA allocated by alloc_vmap_area
 */
static void free_vmap_area(struct vmap_area *va)
{
	spin_lock(&vmap_area_lock);
	__free_vmap_area(va);
	spin_unlock(&vmap_area_lock);
}

/*
 * Clear the pagetable entries of a given vmap_area
 */
static void unmap_vmap_area(struct vmap_area *va)
{
	vunmap_page_range(va->va_start, va->va_end);
}

463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483
static void vmap_debug_free_range(unsigned long start, unsigned long end)
{
	/*
	 * Unmap page tables and force a TLB flush immediately if
	 * CONFIG_DEBUG_PAGEALLOC is set. This catches use after free
	 * bugs similarly to those in linear kernel virtual address
	 * space after a page has been freed.
	 *
	 * All the lazy freeing logic is still retained, in order to
	 * minimise intrusiveness of this debugging feature.
	 *
	 * This is going to be *slow* (linear kernel virtual address
	 * debugging doesn't do a broadcast TLB flush so it is a lot
	 * faster).
	 */
#ifdef CONFIG_DEBUG_PAGEALLOC
	vunmap_page_range(start, end);
	flush_tlb_kernel_range(start, end);
#endif
}

N
Nick Piggin 已提交
484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523
/*
 * lazy_max_pages is the maximum amount of virtual address space we gather up
 * before attempting to purge with a TLB flush.
 *
 * There is a tradeoff here: a larger number will cover more kernel page tables
 * and take slightly longer to purge, but it will linearly reduce the number of
 * global TLB flushes that must be performed. It would seem natural to scale
 * this number up linearly with the number of CPUs (because vmapping activity
 * could also scale linearly with the number of CPUs), however it is likely
 * that in practice, workloads might be constrained in other ways that mean
 * vmap activity will not scale linearly with CPUs. Also, I want to be
 * conservative and not introduce a big latency on huge systems, so go with
 * a less aggressive log scale. It will still be an improvement over the old
 * code, and it will be simple to change the scale factor if we find that it
 * becomes a problem on bigger systems.
 */
static unsigned long lazy_max_pages(void)
{
	unsigned int log;

	log = fls(num_online_cpus());

	return log * (32UL * 1024 * 1024 / PAGE_SIZE);
}

static atomic_t vmap_lazy_nr = ATOMIC_INIT(0);

/*
 * Purges all lazily-freed vmap areas.
 *
 * If sync is 0 then don't purge if there is already a purge in progress.
 * If force_flush is 1, then flush kernel TLBs between *start and *end even
 * if we found no lazy vmap areas to unmap (callers can use this to optimise
 * their own TLB flushing).
 * Returns with *start = min(*start, lowest purged address)
 *              *end = max(*end, highest purged address)
 */
static void __purge_vmap_area_lazy(unsigned long *start, unsigned long *end,
					int sync, int force_flush)
{
524
	static DEFINE_SPINLOCK(purge_lock);
N
Nick Piggin 已提交
525 526
	LIST_HEAD(valist);
	struct vmap_area *va;
527
	struct vmap_area *n_va;
N
Nick Piggin 已提交
528 529 530 531 532 533 534 535
	int nr = 0;

	/*
	 * If sync is 0 but force_flush is 1, we'll go sync anyway but callers
	 * should not expect such behaviour. This just simplifies locking for
	 * the case that isn't actually used at the moment anyway.
	 */
	if (!sync && !force_flush) {
536
		if (!spin_trylock(&purge_lock))
N
Nick Piggin 已提交
537 538
			return;
	} else
539
		spin_lock(&purge_lock);
N
Nick Piggin 已提交
540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566

	rcu_read_lock();
	list_for_each_entry_rcu(va, &vmap_area_list, list) {
		if (va->flags & VM_LAZY_FREE) {
			if (va->va_start < *start)
				*start = va->va_start;
			if (va->va_end > *end)
				*end = va->va_end;
			nr += (va->va_end - va->va_start) >> PAGE_SHIFT;
			unmap_vmap_area(va);
			list_add_tail(&va->purge_list, &valist);
			va->flags |= VM_LAZY_FREEING;
			va->flags &= ~VM_LAZY_FREE;
		}
	}
	rcu_read_unlock();

	if (nr) {
		BUG_ON(nr > atomic_read(&vmap_lazy_nr));
		atomic_sub(nr, &vmap_lazy_nr);
	}

	if (nr || force_flush)
		flush_tlb_kernel_range(*start, *end);

	if (nr) {
		spin_lock(&vmap_area_lock);
567
		list_for_each_entry_safe(va, n_va, &valist, purge_list)
N
Nick Piggin 已提交
568 569 570
			__free_vmap_area(va);
		spin_unlock(&vmap_area_lock);
	}
571
	spin_unlock(&purge_lock);
N
Nick Piggin 已提交
572 573
}

N
Nick Piggin 已提交
574 575 576 577 578 579 580 581 582 583 584
/*
 * Kick off a purge of the outstanding lazy areas. Don't bother if somebody
 * is already purging.
 */
static void try_purge_vmap_area_lazy(void)
{
	unsigned long start = ULONG_MAX, end = 0;

	__purge_vmap_area_lazy(&start, &end, 0, 0);
}

N
Nick Piggin 已提交
585 586 587 588 589 590 591
/*
 * Kick off a purge of the outstanding lazy areas.
 */
static void purge_vmap_area_lazy(void)
{
	unsigned long start = ULONG_MAX, end = 0;

N
Nick Piggin 已提交
592
	__purge_vmap_area_lazy(&start, &end, 1, 0);
N
Nick Piggin 已提交
593 594 595
}

/*
596 597
 * Free and unmap a vmap area, caller ensuring flush_cache_vunmap had been
 * called for the correct range previously.
N
Nick Piggin 已提交
598
 */
599
static void free_unmap_vmap_area_noflush(struct vmap_area *va)
N
Nick Piggin 已提交
600 601 602 603
{
	va->flags |= VM_LAZY_FREE;
	atomic_add((va->va_end - va->va_start) >> PAGE_SHIFT, &vmap_lazy_nr);
	if (unlikely(atomic_read(&vmap_lazy_nr) > lazy_max_pages()))
N
Nick Piggin 已提交
604
		try_purge_vmap_area_lazy();
N
Nick Piggin 已提交
605 606
}

607 608 609 610 611 612 613 614 615
/*
 * Free and unmap a vmap area
 */
static void free_unmap_vmap_area(struct vmap_area *va)
{
	flush_cache_vunmap(va->va_start, va->va_end);
	free_unmap_vmap_area_noflush(va);
}

N
Nick Piggin 已提交
616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665
static struct vmap_area *find_vmap_area(unsigned long addr)
{
	struct vmap_area *va;

	spin_lock(&vmap_area_lock);
	va = __find_vmap_area(addr);
	spin_unlock(&vmap_area_lock);

	return va;
}

static void free_unmap_vmap_area_addr(unsigned long addr)
{
	struct vmap_area *va;

	va = find_vmap_area(addr);
	BUG_ON(!va);
	free_unmap_vmap_area(va);
}


/*** Per cpu kva allocator ***/

/*
 * vmap space is limited especially on 32 bit architectures. Ensure there is
 * room for at least 16 percpu vmap blocks per CPU.
 */
/*
 * If we had a constant VMALLOC_START and VMALLOC_END, we'd like to be able
 * to #define VMALLOC_SPACE		(VMALLOC_END-VMALLOC_START). Guess
 * instead (we just need a rough idea)
 */
#if BITS_PER_LONG == 32
#define VMALLOC_SPACE		(128UL*1024*1024)
#else
#define VMALLOC_SPACE		(128UL*1024*1024*1024)
#endif

#define VMALLOC_PAGES		(VMALLOC_SPACE / PAGE_SIZE)
#define VMAP_MAX_ALLOC		BITS_PER_LONG	/* 256K with 4K pages */
#define VMAP_BBMAP_BITS_MAX	1024	/* 4MB with 4K pages */
#define VMAP_BBMAP_BITS_MIN	(VMAP_MAX_ALLOC*2)
#define VMAP_MIN(x, y)		((x) < (y) ? (x) : (y)) /* can't use min() */
#define VMAP_MAX(x, y)		((x) > (y) ? (x) : (y)) /* can't use max() */
#define VMAP_BBMAP_BITS		VMAP_MIN(VMAP_BBMAP_BITS_MAX,		\
					VMAP_MAX(VMAP_BBMAP_BITS_MIN,	\
						VMALLOC_PAGES / NR_CPUS / 16))

#define VMAP_BLOCK_SIZE		(VMAP_BBMAP_BITS * PAGE_SIZE)

666 667
static bool vmap_initialized __read_mostly = false;

N
Nick Piggin 已提交
668 669 670 671 672 673 674 675 676 677 678 679 680 681 682
struct vmap_block_queue {
	spinlock_t lock;
	struct list_head free;
	struct list_head dirty;
	unsigned int nr_dirty;
};

struct vmap_block {
	spinlock_t lock;
	struct vmap_area *va;
	struct vmap_block_queue *vbq;
	unsigned long free, dirty;
	DECLARE_BITMAP(alloc_map, VMAP_BBMAP_BITS);
	DECLARE_BITMAP(dirty_map, VMAP_BBMAP_BITS);
	union {
683
		struct list_head free_list;
N
Nick Piggin 已提交
684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779
		struct rcu_head rcu_head;
	};
};

/* Queue of free and dirty vmap blocks, for allocation and flushing purposes */
static DEFINE_PER_CPU(struct vmap_block_queue, vmap_block_queue);

/*
 * Radix tree of vmap blocks, indexed by address, to quickly find a vmap block
 * in the free path. Could get rid of this if we change the API to return a
 * "cookie" from alloc, to be passed to free. But no big deal yet.
 */
static DEFINE_SPINLOCK(vmap_block_tree_lock);
static RADIX_TREE(vmap_block_tree, GFP_ATOMIC);

/*
 * We should probably have a fallback mechanism to allocate virtual memory
 * out of partially filled vmap blocks. However vmap block sizing should be
 * fairly reasonable according to the vmalloc size, so it shouldn't be a
 * big problem.
 */

static unsigned long addr_to_vb_idx(unsigned long addr)
{
	addr -= VMALLOC_START & ~(VMAP_BLOCK_SIZE-1);
	addr /= VMAP_BLOCK_SIZE;
	return addr;
}

static struct vmap_block *new_vmap_block(gfp_t gfp_mask)
{
	struct vmap_block_queue *vbq;
	struct vmap_block *vb;
	struct vmap_area *va;
	unsigned long vb_idx;
	int node, err;

	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);
	if (unlikely(IS_ERR(va))) {
		kfree(vb);
		return ERR_PTR(PTR_ERR(va));
	}

	err = radix_tree_preload(gfp_mask);
	if (unlikely(err)) {
		kfree(vb);
		free_vmap_area(va);
		return ERR_PTR(err);
	}

	spin_lock_init(&vb->lock);
	vb->va = va;
	vb->free = VMAP_BBMAP_BITS;
	vb->dirty = 0;
	bitmap_zero(vb->alloc_map, VMAP_BBMAP_BITS);
	bitmap_zero(vb->dirty_map, VMAP_BBMAP_BITS);
	INIT_LIST_HEAD(&vb->free_list);

	vb_idx = addr_to_vb_idx(va->va_start);
	spin_lock(&vmap_block_tree_lock);
	err = radix_tree_insert(&vmap_block_tree, vb_idx, vb);
	spin_unlock(&vmap_block_tree_lock);
	BUG_ON(err);
	radix_tree_preload_end();

	vbq = &get_cpu_var(vmap_block_queue);
	vb->vbq = vbq;
	spin_lock(&vbq->lock);
	list_add(&vb->free_list, &vbq->free);
	spin_unlock(&vbq->lock);
	put_cpu_var(vmap_cpu_blocks);

	return vb;
}

static void rcu_free_vb(struct rcu_head *head)
{
	struct vmap_block *vb = container_of(head, struct vmap_block, rcu_head);

	kfree(vb);
}

static void free_vmap_block(struct vmap_block *vb)
{
	struct vmap_block *tmp;
	unsigned long vb_idx;

780
	BUG_ON(!list_empty(&vb->free_list));
N
Nick Piggin 已提交
781 782 783 784 785 786 787

	vb_idx = addr_to_vb_idx(vb->va->va_start);
	spin_lock(&vmap_block_tree_lock);
	tmp = radix_tree_delete(&vmap_block_tree, vb_idx);
	spin_unlock(&vmap_block_tree_lock);
	BUG_ON(tmp != vb);

788
	free_unmap_vmap_area_noflush(vb->va);
N
Nick Piggin 已提交
789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849
	call_rcu(&vb->rcu_head, rcu_free_vb);
}

static void *vb_alloc(unsigned long size, gfp_t gfp_mask)
{
	struct vmap_block_queue *vbq;
	struct vmap_block *vb;
	unsigned long addr = 0;
	unsigned int order;

	BUG_ON(size & ~PAGE_MASK);
	BUG_ON(size > PAGE_SIZE*VMAP_MAX_ALLOC);
	order = get_order(size);

again:
	rcu_read_lock();
	vbq = &get_cpu_var(vmap_block_queue);
	list_for_each_entry_rcu(vb, &vbq->free, free_list) {
		int i;

		spin_lock(&vb->lock);
		i = bitmap_find_free_region(vb->alloc_map,
						VMAP_BBMAP_BITS, order);

		if (i >= 0) {
			addr = vb->va->va_start + (i << PAGE_SHIFT);
			BUG_ON(addr_to_vb_idx(addr) !=
					addr_to_vb_idx(vb->va->va_start));
			vb->free -= 1UL << order;
			if (vb->free == 0) {
				spin_lock(&vbq->lock);
				list_del_init(&vb->free_list);
				spin_unlock(&vbq->lock);
			}
			spin_unlock(&vb->lock);
			break;
		}
		spin_unlock(&vb->lock);
	}
	put_cpu_var(vmap_cpu_blocks);
	rcu_read_unlock();

	if (!addr) {
		vb = new_vmap_block(gfp_mask);
		if (IS_ERR(vb))
			return vb;
		goto again;
	}

	return (void *)addr;
}

static void vb_free(const void *addr, unsigned long size)
{
	unsigned long offset;
	unsigned long vb_idx;
	unsigned int order;
	struct vmap_block *vb;

	BUG_ON(size & ~PAGE_MASK);
	BUG_ON(size > PAGE_SIZE*VMAP_MAX_ALLOC);
850 851 852

	flush_cache_vunmap((unsigned long)addr, (unsigned long)addr + size);

N
Nick Piggin 已提交
853 854 855 856 857 858 859 860 861 862 863 864
	order = get_order(size);

	offset = (unsigned long)addr & (VMAP_BLOCK_SIZE - 1);

	vb_idx = addr_to_vb_idx((unsigned long)addr);
	rcu_read_lock();
	vb = radix_tree_lookup(&vmap_block_tree, vb_idx);
	rcu_read_unlock();
	BUG_ON(!vb);

	spin_lock(&vb->lock);
	bitmap_allocate_region(vb->dirty_map, offset >> PAGE_SHIFT, order);
865

N
Nick Piggin 已提交
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
	vb->dirty += 1UL << order;
	if (vb->dirty == VMAP_BBMAP_BITS) {
		BUG_ON(vb->free || !list_empty(&vb->free_list));
		spin_unlock(&vb->lock);
		free_vmap_block(vb);
	} else
		spin_unlock(&vb->lock);
}

/**
 * vm_unmap_aliases - unmap outstanding lazy aliases in the vmap layer
 *
 * The vmap/vmalloc layer lazily flushes kernel virtual mappings primarily
 * to amortize TLB flushing overheads. What this means is that any page you
 * have now, may, in a former life, have been mapped into kernel virtual
 * address by the vmap layer and so there might be some CPUs with TLB entries
 * still referencing that page (additional to the regular 1:1 kernel mapping).
 *
 * vm_unmap_aliases flushes all such lazy mappings. After it returns, we can
 * be sure that none of the pages we have control over will have any aliases
 * from the vmap layer.
 */
void vm_unmap_aliases(void)
{
	unsigned long start = ULONG_MAX, end = 0;
	int cpu;
	int flush = 0;

894 895 896
	if (unlikely(!vmap_initialized))
		return;

N
Nick Piggin 已提交
897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951
	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) {
			int i;

			spin_lock(&vb->lock);
			i = find_first_bit(vb->dirty_map, VMAP_BBMAP_BITS);
			while (i < VMAP_BBMAP_BITS) {
				unsigned long s, e;
				int j;
				j = find_next_zero_bit(vb->dirty_map,
					VMAP_BBMAP_BITS, i);

				s = vb->va->va_start + (i << PAGE_SHIFT);
				e = vb->va->va_start + (j << PAGE_SHIFT);
				vunmap_page_range(s, e);
				flush = 1;

				if (s < start)
					start = s;
				if (e > end)
					end = e;

				i = j;
				i = find_next_bit(vb->dirty_map,
							VMAP_BBMAP_BITS, i);
			}
			spin_unlock(&vb->lock);
		}
		rcu_read_unlock();
	}

	__purge_vmap_area_lazy(&start, &end, 1, flush);
}
EXPORT_SYMBOL_GPL(vm_unmap_aliases);

/**
 * vm_unmap_ram - unmap linear kernel address space set up by vm_map_ram
 * @mem: the pointer returned by vm_map_ram
 * @count: the count passed to that vm_map_ram call (cannot unmap partial)
 */
void vm_unmap_ram(const void *mem, unsigned int count)
{
	unsigned long size = count << PAGE_SHIFT;
	unsigned long addr = (unsigned long)mem;

	BUG_ON(!addr);
	BUG_ON(addr < VMALLOC_START);
	BUG_ON(addr > VMALLOC_END);
	BUG_ON(addr & (PAGE_SIZE-1));

	debug_check_no_locks_freed(mem, size);
952
	vmap_debug_free_range(addr, addr+size);
N
Nick Piggin 已提交
953 954 955 956 957 958 959 960 961 962 963 964 965 966

	if (likely(count <= VMAP_MAX_ALLOC))
		vb_free(mem, size);
	else
		free_unmap_vmap_area_addr(addr);
}
EXPORT_SYMBOL(vm_unmap_ram);

/**
 * vm_map_ram - map pages linearly into kernel virtual address (vmalloc space)
 * @pages: an array of pointers to the pages to be mapped
 * @count: number of pages
 * @node: prefer to allocate data structures on this node
 * @prot: memory protection to use. PAGE_KERNEL for regular RAM
967 968
 *
 * Returns: a pointer to the address that has been mapped, or %NULL on failure
N
Nick Piggin 已提交
969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998
 */
void *vm_map_ram(struct page **pages, unsigned int count, int node, pgprot_t prot)
{
	unsigned long size = count << PAGE_SHIFT;
	unsigned long addr;
	void *mem;

	if (likely(count <= VMAP_MAX_ALLOC)) {
		mem = vb_alloc(size, GFP_KERNEL);
		if (IS_ERR(mem))
			return NULL;
		addr = (unsigned long)mem;
	} else {
		struct vmap_area *va;
		va = alloc_vmap_area(size, PAGE_SIZE,
				VMALLOC_START, VMALLOC_END, node, GFP_KERNEL);
		if (IS_ERR(va))
			return NULL;

		addr = va->va_start;
		mem = (void *)addr;
	}
	if (vmap_page_range(addr, addr + size, prot, pages) < 0) {
		vm_unmap_ram(mem, count);
		return NULL;
	}
	return mem;
}
EXPORT_SYMBOL(vm_map_ram);

999 1000 1001
/**
 * vm_area_register_early - register vmap area early during boot
 * @vm: vm_struct to register
1002
 * @align: requested alignment
1003 1004 1005 1006 1007 1008 1009 1010
 *
 * 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.
 */
1011
void __init vm_area_register_early(struct vm_struct *vm, size_t align)
1012 1013
{
	static size_t vm_init_off __initdata;
1014 1015 1016 1017
	unsigned long addr;

	addr = ALIGN(VMALLOC_START + vm_init_off, align);
	vm_init_off = PFN_ALIGN(addr + vm->size) - VMALLOC_START;
1018

1019
	vm->addr = (void *)addr;
1020 1021 1022 1023 1024

	vm->next = vmlist;
	vmlist = vm;
}

N
Nick Piggin 已提交
1025 1026
void __init vmalloc_init(void)
{
I
Ivan Kokshaysky 已提交
1027 1028
	struct vmap_area *va;
	struct vm_struct *tmp;
N
Nick Piggin 已提交
1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039
	int i;

	for_each_possible_cpu(i) {
		struct vmap_block_queue *vbq;

		vbq = &per_cpu(vmap_block_queue, i);
		spin_lock_init(&vbq->lock);
		INIT_LIST_HEAD(&vbq->free);
		INIT_LIST_HEAD(&vbq->dirty);
		vbq->nr_dirty = 0;
	}
1040

I
Ivan Kokshaysky 已提交
1041 1042
	/* Import existing vmlist entries. */
	for (tmp = vmlist; tmp; tmp = tmp->next) {
1043
		va = kzalloc(sizeof(struct vmap_area), GFP_NOWAIT);
I
Ivan Kokshaysky 已提交
1044 1045 1046 1047 1048
		va->flags = tmp->flags | VM_VM_AREA;
		va->va_start = (unsigned long)tmp->addr;
		va->va_end = va->va_start + tmp->size;
		__insert_vmap_area(va);
	}
1049 1050 1051

	vmap_area_pcpu_hole = VMALLOC_END;

1052
	vmap_initialized = true;
N
Nick Piggin 已提交
1053 1054
}

1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106
/**
 * map_kernel_range_noflush - map kernel VM area with the specified pages
 * @addr: start of the VM area to map
 * @size: size of the VM area to map
 * @prot: page protection flags to use
 * @pages: pages to map
 *
 * Map PFN_UP(@size) pages at @addr.  The VM area @addr and @size
 * specify should have been allocated using get_vm_area() and its
 * friends.
 *
 * NOTE:
 * This function does NOT do any cache flushing.  The caller is
 * responsible for calling flush_cache_vmap() on to-be-mapped areas
 * before calling this function.
 *
 * RETURNS:
 * The number of pages mapped on success, -errno on failure.
 */
int map_kernel_range_noflush(unsigned long addr, unsigned long size,
			     pgprot_t prot, struct page **pages)
{
	return vmap_page_range_noflush(addr, addr + size, prot, pages);
}

/**
 * unmap_kernel_range_noflush - unmap kernel VM area
 * @addr: start of the VM area to unmap
 * @size: size of the VM area to unmap
 *
 * Unmap PFN_UP(@size) pages at @addr.  The VM area @addr and @size
 * specify should have been allocated using get_vm_area() and its
 * friends.
 *
 * NOTE:
 * This function does NOT do any cache flushing.  The caller is
 * responsible for calling flush_cache_vunmap() on to-be-mapped areas
 * before calling this function and flush_tlb_kernel_range() after.
 */
void unmap_kernel_range_noflush(unsigned long addr, unsigned long size)
{
	vunmap_page_range(addr, addr + size);
}

/**
 * unmap_kernel_range - unmap kernel VM area and flush cache and TLB
 * @addr: start of the VM area to unmap
 * @size: size of the VM area to unmap
 *
 * Similar to unmap_kernel_range_noflush() but flushes vcache before
 * the unmapping and tlb after.
 */
N
Nick Piggin 已提交
1107 1108 1109
void unmap_kernel_range(unsigned long addr, unsigned long size)
{
	unsigned long end = addr + size;
1110 1111

	flush_cache_vunmap(addr, end);
N
Nick Piggin 已提交
1112 1113 1114 1115 1116 1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135
	vunmap_page_range(addr, end);
	flush_tlb_kernel_range(addr, end);
}

int map_vm_area(struct vm_struct *area, pgprot_t prot, struct page ***pages)
{
	unsigned long addr = (unsigned long)area->addr;
	unsigned long end = addr + area->size - PAGE_SIZE;
	int err;

	err = vmap_page_range(addr, end, prot, *pages);
	if (err > 0) {
		*pages += err;
		err = 0;
	}

	return err;
}
EXPORT_SYMBOL_GPL(map_vm_area);

/*** Old vmalloc interfaces ***/
DEFINE_RWLOCK(vmlist_lock);
struct vm_struct *vmlist;

1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151 1152 1153 1154 1155 1156 1157
static void insert_vmalloc_vm(struct vm_struct *vm, struct vmap_area *va,
			      unsigned long flags, void *caller)
{
	struct vm_struct *tmp, **p;

	vm->flags = flags;
	vm->addr = (void *)va->va_start;
	vm->size = va->va_end - va->va_start;
	vm->caller = caller;
	va->private = vm;
	va->flags |= VM_VM_AREA;

	write_lock(&vmlist_lock);
	for (p = &vmlist; (tmp = *p) != NULL; p = &tmp->next) {
		if (tmp->addr >= vm->addr)
			break;
	}
	vm->next = *p;
	*p = vm;
	write_unlock(&vmlist_lock);
}

N
Nick Piggin 已提交
1158 1159 1160 1161 1162 1163 1164
static struct vm_struct *__get_vm_area_node(unsigned long size,
		unsigned long flags, unsigned long start, unsigned long end,
		int node, gfp_t gfp_mask, void *caller)
{
	static struct vmap_area *va;
	struct vm_struct *area;
	unsigned long align = 1;
L
Linus Torvalds 已提交
1165

1166
	BUG_ON(in_interrupt());
L
Linus Torvalds 已提交
1167 1168 1169 1170 1171 1172 1173 1174 1175 1176
	if (flags & VM_IOREMAP) {
		int bit = fls(size);

		if (bit > IOREMAP_MAX_ORDER)
			bit = IOREMAP_MAX_ORDER;
		else if (bit < PAGE_SHIFT)
			bit = PAGE_SHIFT;

		align = 1ul << bit;
	}
N
Nick Piggin 已提交
1177

L
Linus Torvalds 已提交
1178
	size = PAGE_ALIGN(size);
1179 1180
	if (unlikely(!size))
		return NULL;
L
Linus Torvalds 已提交
1181

1182
	area = kzalloc_node(sizeof(*area), gfp_mask & GFP_RECLAIM_MASK, node);
L
Linus Torvalds 已提交
1183 1184 1185 1186 1187 1188 1189 1190
	if (unlikely(!area))
		return NULL;

	/*
	 * We always allocate a guard page.
	 */
	size += PAGE_SIZE;

N
Nick Piggin 已提交
1191 1192 1193 1194
	va = alloc_vmap_area(size, align, start, end, node, gfp_mask);
	if (IS_ERR(va)) {
		kfree(area);
		return NULL;
L
Linus Torvalds 已提交
1195 1196
	}

1197
	insert_vmalloc_vm(area, va, flags, caller);
L
Linus Torvalds 已提交
1198 1199 1200
	return area;
}

C
Christoph Lameter 已提交
1201 1202 1203
struct vm_struct *__get_vm_area(unsigned long size, unsigned long flags,
				unsigned long start, unsigned long end)
{
1204 1205
	return __get_vm_area_node(size, flags, start, end, -1, GFP_KERNEL,
						__builtin_return_address(0));
C
Christoph Lameter 已提交
1206
}
1207
EXPORT_SYMBOL_GPL(__get_vm_area);
C
Christoph Lameter 已提交
1208

1209 1210 1211 1212 1213 1214 1215 1216
struct vm_struct *__get_vm_area_caller(unsigned long size, unsigned long flags,
				       unsigned long start, unsigned long end,
				       void *caller)
{
	return __get_vm_area_node(size, flags, start, end, -1, GFP_KERNEL,
				  caller);
}

L
Linus Torvalds 已提交
1217
/**
S
Simon Arlott 已提交
1218
 *	get_vm_area  -  reserve a contiguous kernel virtual area
L
Linus Torvalds 已提交
1219 1220 1221 1222 1223 1224 1225 1226 1227
 *	@size:		size of the area
 *	@flags:		%VM_IOREMAP for I/O mappings or VM_ALLOC
 *
 *	Search an area of @size in the kernel virtual mapping area,
 *	and reserved it for out purposes.  Returns the area descriptor
 *	on success or %NULL on failure.
 */
struct vm_struct *get_vm_area(unsigned long size, unsigned long flags)
{
1228 1229 1230 1231 1232 1233 1234 1235 1236
	return __get_vm_area_node(size, flags, VMALLOC_START, VMALLOC_END,
				-1, GFP_KERNEL, __builtin_return_address(0));
}

struct vm_struct *get_vm_area_caller(unsigned long size, unsigned long flags,
				void *caller)
{
	return __get_vm_area_node(size, flags, VMALLOC_START, VMALLOC_END,
						-1, GFP_KERNEL, caller);
L
Linus Torvalds 已提交
1237 1238
}

1239 1240
struct vm_struct *get_vm_area_node(unsigned long size, unsigned long flags,
				   int node, gfp_t gfp_mask)
C
Christoph Lameter 已提交
1241
{
1242
	return __get_vm_area_node(size, flags, VMALLOC_START, VMALLOC_END, node,
1243
				  gfp_mask, __builtin_return_address(0));
C
Christoph Lameter 已提交
1244 1245
}

N
Nick Piggin 已提交
1246
static struct vm_struct *find_vm_area(const void *addr)
1247
{
N
Nick Piggin 已提交
1248
	struct vmap_area *va;
1249

N
Nick Piggin 已提交
1250 1251 1252
	va = find_vmap_area((unsigned long)addr);
	if (va && va->flags & VM_VM_AREA)
		return va->private;
L
Linus Torvalds 已提交
1253 1254 1255 1256

	return NULL;
}

1257
/**
S
Simon Arlott 已提交
1258
 *	remove_vm_area  -  find and remove a continuous kernel virtual area
1259 1260 1261 1262 1263 1264
 *	@addr:		base address
 *
 *	Search for the kernel VM area starting at @addr, and remove it.
 *	This function returns the found VM area, but using it is NOT safe
 *	on SMP machines, except for its size or flags.
 */
1265
struct vm_struct *remove_vm_area(const void *addr)
1266
{
N
Nick Piggin 已提交
1267 1268 1269 1270 1271 1272
	struct vmap_area *va;

	va = find_vmap_area((unsigned long)addr);
	if (va && va->flags & VM_VM_AREA) {
		struct vm_struct *vm = va->private;
		struct vm_struct *tmp, **p;
1273 1274 1275 1276 1277
		/*
		 * remove from list and disallow access to this vm_struct
		 * before unmap. (address range confliction is maintained by
		 * vmap.)
		 */
N
Nick Piggin 已提交
1278 1279 1280 1281 1282 1283
		write_lock(&vmlist_lock);
		for (p = &vmlist; (tmp = *p) != vm; p = &tmp->next)
			;
		*p = tmp->next;
		write_unlock(&vmlist_lock);

1284 1285 1286 1287
		vmap_debug_free_range(va->va_start, va->va_end);
		free_unmap_vmap_area(va);
		vm->size -= PAGE_SIZE;

N
Nick Piggin 已提交
1288 1289 1290
		return vm;
	}
	return NULL;
1291 1292
}

1293
static void __vunmap(const void *addr, int deallocate_pages)
L
Linus Torvalds 已提交
1294 1295 1296 1297 1298 1299 1300
{
	struct vm_struct *area;

	if (!addr)
		return;

	if ((PAGE_SIZE-1) & (unsigned long)addr) {
A
Arjan van de Ven 已提交
1301
		WARN(1, KERN_ERR "Trying to vfree() bad address (%p)\n", addr);
L
Linus Torvalds 已提交
1302 1303 1304 1305 1306
		return;
	}

	area = remove_vm_area(addr);
	if (unlikely(!area)) {
A
Arjan van de Ven 已提交
1307
		WARN(1, KERN_ERR "Trying to vfree() nonexistent vm area (%p)\n",
L
Linus Torvalds 已提交
1308 1309 1310 1311
				addr);
		return;
	}

1312
	debug_check_no_locks_freed(addr, area->size);
1313
	debug_check_no_obj_freed(addr, area->size);
1314

L
Linus Torvalds 已提交
1315 1316 1317 1318
	if (deallocate_pages) {
		int i;

		for (i = 0; i < area->nr_pages; i++) {
1319 1320 1321 1322
			struct page *page = area->pages[i];

			BUG_ON(!page);
			__free_page(page);
L
Linus Torvalds 已提交
1323 1324
		}

1325
		if (area->flags & VM_VPAGES)
L
Linus Torvalds 已提交
1326 1327 1328 1329 1330 1331 1332 1333 1334 1335 1336 1337 1338
			vfree(area->pages);
		else
			kfree(area->pages);
	}

	kfree(area);
	return;
}

/**
 *	vfree  -  release memory allocated by vmalloc()
 *	@addr:		memory base address
 *
S
Simon Arlott 已提交
1339
 *	Free the virtually continuous memory area starting at @addr, as
1340 1341
 *	obtained from vmalloc(), vmalloc_32() or __vmalloc(). If @addr is
 *	NULL, no operation is performed.
L
Linus Torvalds 已提交
1342
 *
1343
 *	Must not be called in interrupt context.
L
Linus Torvalds 已提交
1344
 */
1345
void vfree(const void *addr)
L
Linus Torvalds 已提交
1346 1347
{
	BUG_ON(in_interrupt());
1348 1349 1350

	kmemleak_free(addr);

L
Linus Torvalds 已提交
1351 1352 1353 1354 1355 1356 1357 1358 1359 1360 1361
	__vunmap(addr, 1);
}
EXPORT_SYMBOL(vfree);

/**
 *	vunmap  -  release virtual mapping obtained by vmap()
 *	@addr:		memory base address
 *
 *	Free the virtually contiguous memory area starting at @addr,
 *	which was created from the page array passed to vmap().
 *
1362
 *	Must not be called in interrupt context.
L
Linus Torvalds 已提交
1363
 */
1364
void vunmap(const void *addr)
L
Linus Torvalds 已提交
1365 1366
{
	BUG_ON(in_interrupt());
1367
	might_sleep();
L
Linus Torvalds 已提交
1368 1369 1370 1371 1372 1373 1374 1375 1376 1377 1378 1379 1380 1381 1382 1383 1384 1385 1386
	__vunmap(addr, 0);
}
EXPORT_SYMBOL(vunmap);

/**
 *	vmap  -  map an array of pages into virtually contiguous space
 *	@pages:		array of page pointers
 *	@count:		number of pages to map
 *	@flags:		vm_area->flags
 *	@prot:		page protection for the mapping
 *
 *	Maps @count pages from @pages into contiguous kernel virtual
 *	space.
 */
void *vmap(struct page **pages, unsigned int count,
		unsigned long flags, pgprot_t prot)
{
	struct vm_struct *area;

1387 1388
	might_sleep();

1389
	if (count > totalram_pages)
L
Linus Torvalds 已提交
1390 1391
		return NULL;

1392 1393
	area = get_vm_area_caller((count << PAGE_SHIFT), flags,
					__builtin_return_address(0));
L
Linus Torvalds 已提交
1394 1395
	if (!area)
		return NULL;
1396

L
Linus Torvalds 已提交
1397 1398 1399 1400 1401 1402 1403 1404 1405
	if (map_vm_area(area, prot, &pages)) {
		vunmap(area->addr);
		return NULL;
	}

	return area->addr;
}
EXPORT_SYMBOL(vmap);

N
Nick Piggin 已提交
1406 1407
static void *__vmalloc_node(unsigned long size, gfp_t gfp_mask, pgprot_t prot,
			    int node, void *caller);
A
Adrian Bunk 已提交
1408
static void *__vmalloc_area_node(struct vm_struct *area, gfp_t gfp_mask,
1409
				 pgprot_t prot, int node, void *caller)
L
Linus Torvalds 已提交
1410 1411 1412 1413 1414 1415 1416 1417 1418
{
	struct page **pages;
	unsigned int nr_pages, array_size, i;

	nr_pages = (area->size - PAGE_SIZE) >> PAGE_SHIFT;
	array_size = (nr_pages * sizeof(struct page *));

	area->nr_pages = nr_pages;
	/* Please note that the recursion is strictly bounded. */
1419
	if (array_size > PAGE_SIZE) {
1420
		pages = __vmalloc_node(array_size, gfp_mask | __GFP_ZERO,
1421
				PAGE_KERNEL, node, caller);
1422
		area->flags |= VM_VPAGES;
1423 1424
	} else {
		pages = kmalloc_node(array_size,
C
Christoph Lameter 已提交
1425
				(gfp_mask & GFP_RECLAIM_MASK) | __GFP_ZERO,
1426 1427
				node);
	}
L
Linus Torvalds 已提交
1428
	area->pages = pages;
1429
	area->caller = caller;
L
Linus Torvalds 已提交
1430 1431 1432 1433 1434 1435 1436
	if (!area->pages) {
		remove_vm_area(area->addr);
		kfree(area);
		return NULL;
	}

	for (i = 0; i < area->nr_pages; i++) {
1437 1438
		struct page *page;

C
Christoph Lameter 已提交
1439
		if (node < 0)
1440
			page = alloc_page(gfp_mask);
C
Christoph Lameter 已提交
1441
		else
1442 1443 1444
			page = alloc_pages_node(node, gfp_mask, 0);

		if (unlikely(!page)) {
L
Linus Torvalds 已提交
1445 1446 1447 1448
			/* Successfully allocated i pages, free them in __vunmap() */
			area->nr_pages = i;
			goto fail;
		}
1449
		area->pages[i] = page;
L
Linus Torvalds 已提交
1450 1451 1452 1453 1454 1455 1456 1457 1458 1459 1460
	}

	if (map_vm_area(area, prot, &pages))
		goto fail;
	return area->addr;

fail:
	vfree(area->addr);
	return NULL;
}

C
Christoph Lameter 已提交
1461 1462
void *__vmalloc_area(struct vm_struct *area, gfp_t gfp_mask, pgprot_t prot)
{
1463 1464 1465 1466 1467 1468 1469 1470 1471 1472 1473
	void *addr = __vmalloc_area_node(area, gfp_mask, prot, -1,
					 __builtin_return_address(0));

	/*
	 * A ref_count = 3 is needed because the vm_struct and vmap_area
	 * structures allocated in the __get_vm_area_node() function contain
	 * references to the virtual address of the vmalloc'ed block.
	 */
	kmemleak_alloc(addr, area->size - PAGE_SIZE, 3, gfp_mask);

	return addr;
C
Christoph Lameter 已提交
1474 1475
}

L
Linus Torvalds 已提交
1476
/**
C
Christoph Lameter 已提交
1477
 *	__vmalloc_node  -  allocate virtually contiguous memory
L
Linus Torvalds 已提交
1478 1479 1480
 *	@size:		allocation size
 *	@gfp_mask:	flags for the page level allocator
 *	@prot:		protection mask for the allocated pages
1481
 *	@node:		node to use for allocation or -1
1482
 *	@caller:	caller's return address
L
Linus Torvalds 已提交
1483 1484 1485 1486 1487
 *
 *	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.
 */
A
Adrian Bunk 已提交
1488
static void *__vmalloc_node(unsigned long size, gfp_t gfp_mask, pgprot_t prot,
1489
						int node, void *caller)
L
Linus Torvalds 已提交
1490 1491
{
	struct vm_struct *area;
1492 1493
	void *addr;
	unsigned long real_size = size;
L
Linus Torvalds 已提交
1494 1495

	size = PAGE_ALIGN(size);
1496
	if (!size || (size >> PAGE_SHIFT) > totalram_pages)
L
Linus Torvalds 已提交
1497 1498
		return NULL;

1499 1500 1501
	area = __get_vm_area_node(size, VM_ALLOC, VMALLOC_START, VMALLOC_END,
						node, gfp_mask, caller);

L
Linus Torvalds 已提交
1502 1503 1504
	if (!area)
		return NULL;

1505 1506 1507 1508 1509 1510 1511 1512 1513 1514
	addr = __vmalloc_area_node(area, gfp_mask, prot, node, caller);

	/*
	 * A ref_count = 3 is needed because the vm_struct and vmap_area
	 * structures allocated in the __get_vm_area_node() function contain
	 * references to the virtual address of the vmalloc'ed block.
	 */
	kmemleak_alloc(addr, real_size, 3, gfp_mask);

	return addr;
L
Linus Torvalds 已提交
1515 1516
}

C
Christoph Lameter 已提交
1517 1518
void *__vmalloc(unsigned long size, gfp_t gfp_mask, pgprot_t prot)
{
1519 1520
	return __vmalloc_node(size, gfp_mask, prot, -1,
				__builtin_return_address(0));
C
Christoph Lameter 已提交
1521
}
L
Linus Torvalds 已提交
1522 1523 1524 1525 1526 1527 1528 1529
EXPORT_SYMBOL(__vmalloc);

/**
 *	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.
 *
1530
 *	For tight control over page level allocator and protection flags
L
Linus Torvalds 已提交
1531 1532 1533 1534
 *	use __vmalloc() instead.
 */
void *vmalloc(unsigned long size)
{
1535 1536
	return __vmalloc_node(size, GFP_KERNEL | __GFP_HIGHMEM, PAGE_KERNEL,
					-1, __builtin_return_address(0));
L
Linus Torvalds 已提交
1537 1538 1539
}
EXPORT_SYMBOL(vmalloc);

1540
/**
1541 1542
 * vmalloc_user - allocate zeroed virtually contiguous memory for userspace
 * @size: allocation size
1543
 *
1544 1545
 * The resulting memory area is zeroed so it can be mapped to userspace
 * without leaking data.
1546 1547 1548 1549 1550 1551
 */
void *vmalloc_user(unsigned long size)
{
	struct vm_struct *area;
	void *ret;

G
Glauber Costa 已提交
1552 1553
	ret = __vmalloc_node(size, GFP_KERNEL | __GFP_HIGHMEM | __GFP_ZERO,
			     PAGE_KERNEL, -1, __builtin_return_address(0));
1554
	if (ret) {
N
Nick Piggin 已提交
1555
		area = find_vm_area(ret);
1556 1557
		area->flags |= VM_USERMAP;
	}
1558 1559 1560 1561
	return ret;
}
EXPORT_SYMBOL(vmalloc_user);

C
Christoph Lameter 已提交
1562 1563 1564
/**
 *	vmalloc_node  -  allocate memory on a specific node
 *	@size:		allocation size
1565
 *	@node:		numa node
C
Christoph Lameter 已提交
1566 1567 1568 1569
 *
 *	Allocate enough pages to cover @size from the page level
 *	allocator and map them into contiguous kernel virtual space.
 *
1570
 *	For tight control over page level allocator and protection flags
C
Christoph Lameter 已提交
1571 1572 1573 1574
 *	use __vmalloc() instead.
 */
void *vmalloc_node(unsigned long size, int node)
{
1575 1576
	return __vmalloc_node(size, GFP_KERNEL | __GFP_HIGHMEM, PAGE_KERNEL,
					node, __builtin_return_address(0));
C
Christoph Lameter 已提交
1577 1578 1579
}
EXPORT_SYMBOL(vmalloc_node);

1580 1581 1582 1583
#ifndef PAGE_KERNEL_EXEC
# define PAGE_KERNEL_EXEC PAGE_KERNEL
#endif

L
Linus Torvalds 已提交
1584 1585 1586 1587 1588 1589 1590 1591
/**
 *	vmalloc_exec  -  allocate virtually contiguous, executable memory
 *	@size:		allocation size
 *
 *	Kernel-internal function to allocate enough pages to cover @size
 *	the page level allocator and map them into contiguous and
 *	executable kernel virtual space.
 *
1592
 *	For tight control over page level allocator and protection flags
L
Linus Torvalds 已提交
1593 1594 1595 1596 1597
 *	use __vmalloc() instead.
 */

void *vmalloc_exec(unsigned long size)
{
G
Glauber Costa 已提交
1598 1599
	return __vmalloc_node(size, GFP_KERNEL | __GFP_HIGHMEM, PAGE_KERNEL_EXEC,
			      -1, __builtin_return_address(0));
L
Linus Torvalds 已提交
1600 1601
}

1602
#if defined(CONFIG_64BIT) && defined(CONFIG_ZONE_DMA32)
1603
#define GFP_VMALLOC32 GFP_DMA32 | GFP_KERNEL
1604
#elif defined(CONFIG_64BIT) && defined(CONFIG_ZONE_DMA)
1605
#define GFP_VMALLOC32 GFP_DMA | GFP_KERNEL
1606 1607 1608 1609
#else
#define GFP_VMALLOC32 GFP_KERNEL
#endif

L
Linus Torvalds 已提交
1610 1611 1612 1613 1614 1615 1616 1617 1618
/**
 *	vmalloc_32  -  allocate virtually contiguous memory (32bit addressable)
 *	@size:		allocation size
 *
 *	Allocate enough 32bit PA addressable pages to cover @size from the
 *	page level allocator and map them into contiguous kernel virtual space.
 */
void *vmalloc_32(unsigned long size)
{
G
Glauber Costa 已提交
1619 1620
	return __vmalloc_node(size, GFP_VMALLOC32, PAGE_KERNEL,
			      -1, __builtin_return_address(0));
L
Linus Torvalds 已提交
1621 1622 1623
}
EXPORT_SYMBOL(vmalloc_32);

1624
/**
1625
 * vmalloc_32_user - allocate zeroed virtually contiguous 32bit memory
1626
 *	@size:		allocation size
1627 1628 1629
 *
 * The resulting memory area is 32bit addressable and zeroed so it can be
 * mapped to userspace without leaking data.
1630 1631 1632 1633 1634 1635
 */
void *vmalloc_32_user(unsigned long size)
{
	struct vm_struct *area;
	void *ret;

G
Glauber Costa 已提交
1636 1637
	ret = __vmalloc_node(size, GFP_VMALLOC32 | __GFP_ZERO, PAGE_KERNEL,
			     -1, __builtin_return_address(0));
1638
	if (ret) {
N
Nick Piggin 已提交
1639
		area = find_vm_area(ret);
1640 1641
		area->flags |= VM_USERMAP;
	}
1642 1643 1644 1645
	return ret;
}
EXPORT_SYMBOL(vmalloc_32_user);

1646 1647 1648 1649 1650 1651 1652 1653 1654 1655 1656 1657 1658 1659 1660 1661 1662 1663 1664 1665 1666 1667 1668 1669 1670 1671 1672 1673 1674 1675 1676 1677 1678 1679 1680 1681 1682 1683 1684 1685 1686 1687 1688 1689 1690 1691 1692 1693 1694 1695 1696 1697 1698 1699 1700 1701 1702 1703 1704 1705 1706 1707 1708 1709 1710 1711 1712 1713 1714 1715 1716 1717 1718 1719 1720 1721 1722 1723 1724 1725 1726 1727 1728 1729 1730 1731 1732 1733 1734 1735 1736 1737 1738 1739 1740 1741 1742 1743 1744 1745 1746 1747 1748 1749 1750 1751 1752 1753 1754
/*
 * small helper routine , copy contents to buf from addr.
 * If the page is not present, fill zero.
 */

static int aligned_vread(char *buf, char *addr, unsigned long count)
{
	struct page *p;
	int copied = 0;

	while (count) {
		unsigned long offset, length;

		offset = (unsigned long)addr & ~PAGE_MASK;
		length = PAGE_SIZE - offset;
		if (length > count)
			length = count;
		p = vmalloc_to_page(addr);
		/*
		 * To do safe access to this _mapped_ area, we need
		 * lock. But adding lock here means that we need to add
		 * overhead of vmalloc()/vfree() calles for this _debug_
		 * interface, rarely used. Instead of that, we'll use
		 * kmap() and get small overhead in this access function.
		 */
		if (p) {
			/*
			 * we can expect USER0 is not used (see vread/vwrite's
			 * function description)
			 */
			void *map = kmap_atomic(p, KM_USER0);
			memcpy(buf, map + offset, length);
			kunmap_atomic(map, KM_USER0);
		} else
			memset(buf, 0, length);

		addr += length;
		buf += length;
		copied += length;
		count -= length;
	}
	return copied;
}

static int aligned_vwrite(char *buf, char *addr, unsigned long count)
{
	struct page *p;
	int copied = 0;

	while (count) {
		unsigned long offset, length;

		offset = (unsigned long)addr & ~PAGE_MASK;
		length = PAGE_SIZE - offset;
		if (length > count)
			length = count;
		p = vmalloc_to_page(addr);
		/*
		 * To do safe access to this _mapped_ area, we need
		 * lock. But adding lock here means that we need to add
		 * overhead of vmalloc()/vfree() calles for this _debug_
		 * interface, rarely used. Instead of that, we'll use
		 * kmap() and get small overhead in this access function.
		 */
		if (p) {
			/*
			 * we can expect USER0 is not used (see vread/vwrite's
			 * function description)
			 */
			void *map = kmap_atomic(p, KM_USER0);
			memcpy(map + offset, buf, length);
			kunmap_atomic(map, KM_USER0);
		}
		addr += length;
		buf += length;
		copied += length;
		count -= length;
	}
	return copied;
}

/**
 *	vread() -  read vmalloc area in a safe way.
 *	@buf:		buffer for reading data
 *	@addr:		vm address.
 *	@count:		number of bytes to be read.
 *
 *	Returns # of bytes which addr and buf should be increased.
 *	(same number to @count). Returns 0 if [addr...addr+count) doesn't
 *	includes any intersect with alive vmalloc area.
 *
 *	This function checks that addr is a valid vmalloc'ed area, and
 *	copy data from that area to a given buffer. If the given memory range
 *	of [addr...addr+count) includes some valid address, data is copied to
 *	proper area of @buf. If there are memory holes, they'll be zero-filled.
 *	IOREMAP area is treated as memory hole and no copy is done.
 *
 *	If [addr...addr+count) doesn't includes any intersects with alive
 *	vm_struct area, returns 0.
 *	@buf should be kernel's buffer. Because	this function uses KM_USER0,
 *	the caller should guarantee KM_USER0 is not used.
 *
 *	Note: In usual ops, vread() is never necessary because the caller
 *	should know vmalloc() area is valid and can use memcpy().
 *	This is for routines which have to access vmalloc area without
 *	any informaion, as /dev/kmem.
 *
 */

L
Linus Torvalds 已提交
1755 1756 1757 1758
long vread(char *buf, char *addr, unsigned long count)
{
	struct vm_struct *tmp;
	char *vaddr, *buf_start = buf;
1759
	unsigned long buflen = count;
L
Linus Torvalds 已提交
1760 1761 1762 1763 1764 1765 1766
	unsigned long n;

	/* Don't allow overflow */
	if ((unsigned long) addr + count < count)
		count = -(unsigned long) addr;

	read_lock(&vmlist_lock);
1767
	for (tmp = vmlist; count && tmp; tmp = tmp->next) {
L
Linus Torvalds 已提交
1768 1769 1770 1771 1772 1773 1774 1775 1776 1777 1778 1779
		vaddr = (char *) tmp->addr;
		if (addr >= vaddr + tmp->size - PAGE_SIZE)
			continue;
		while (addr < vaddr) {
			if (count == 0)
				goto finished;
			*buf = '\0';
			buf++;
			addr++;
			count--;
		}
		n = vaddr + tmp->size - PAGE_SIZE - addr;
1780 1781 1782 1783 1784 1785 1786 1787 1788
		if (n > count)
			n = count;
		if (!(tmp->flags & VM_IOREMAP))
			aligned_vread(buf, addr, n);
		else /* IOREMAP area is treated as memory hole */
			memset(buf, 0, n);
		buf += n;
		addr += n;
		count -= n;
L
Linus Torvalds 已提交
1789 1790 1791
	}
finished:
	read_unlock(&vmlist_lock);
1792 1793 1794 1795 1796 1797 1798 1799

	if (buf == buf_start)
		return 0;
	/* zero-fill memory holes */
	if (buf != buf_start + buflen)
		memset(buf, 0, buflen - (buf - buf_start));

	return buflen;
L
Linus Torvalds 已提交
1800 1801
}

1802 1803 1804 1805 1806 1807 1808 1809 1810 1811 1812 1813 1814 1815 1816 1817 1818 1819 1820 1821 1822 1823 1824 1825 1826 1827 1828 1829 1830 1831
/**
 *	vwrite() -  write vmalloc area in a safe way.
 *	@buf:		buffer for source data
 *	@addr:		vm address.
 *	@count:		number of bytes to be read.
 *
 *	Returns # of bytes which addr and buf should be incresed.
 *	(same number to @count).
 *	If [addr...addr+count) doesn't includes any intersect with valid
 *	vmalloc area, returns 0.
 *
 *	This function checks that addr is a valid vmalloc'ed area, and
 *	copy data from a buffer to the given addr. If specified range of
 *	[addr...addr+count) includes some valid address, data is copied from
 *	proper area of @buf. If there are memory holes, no copy to hole.
 *	IOREMAP area is treated as memory hole and no copy is done.
 *
 *	If [addr...addr+count) doesn't includes any intersects with alive
 *	vm_struct area, returns 0.
 *	@buf should be kernel's buffer. Because	this function uses KM_USER0,
 *	the caller should guarantee KM_USER0 is not used.
 *
 *	Note: In usual ops, vwrite() is never necessary because the caller
 *	should know vmalloc() area is valid and can use memcpy().
 *	This is for routines which have to access vmalloc area without
 *	any informaion, as /dev/kmem.
 *
 *	The caller should guarantee KM_USER1 is not used.
 */

L
Linus Torvalds 已提交
1832 1833 1834
long vwrite(char *buf, char *addr, unsigned long count)
{
	struct vm_struct *tmp;
1835 1836 1837
	char *vaddr;
	unsigned long n, buflen;
	int copied = 0;
L
Linus Torvalds 已提交
1838 1839 1840 1841

	/* Don't allow overflow */
	if ((unsigned long) addr + count < count)
		count = -(unsigned long) addr;
1842
	buflen = count;
L
Linus Torvalds 已提交
1843 1844

	read_lock(&vmlist_lock);
1845
	for (tmp = vmlist; count && tmp; tmp = tmp->next) {
L
Linus Torvalds 已提交
1846 1847 1848 1849 1850 1851 1852 1853 1854 1855 1856
		vaddr = (char *) tmp->addr;
		if (addr >= vaddr + tmp->size - PAGE_SIZE)
			continue;
		while (addr < vaddr) {
			if (count == 0)
				goto finished;
			buf++;
			addr++;
			count--;
		}
		n = vaddr + tmp->size - PAGE_SIZE - addr;
1857 1858 1859 1860 1861 1862 1863 1864 1865
		if (n > count)
			n = count;
		if (!(tmp->flags & VM_IOREMAP)) {
			aligned_vwrite(buf, addr, n);
			copied++;
		}
		buf += n;
		addr += n;
		count -= n;
L
Linus Torvalds 已提交
1866 1867 1868
	}
finished:
	read_unlock(&vmlist_lock);
1869 1870 1871
	if (!copied)
		return 0;
	return buflen;
L
Linus Torvalds 已提交
1872
}
1873 1874 1875 1876 1877 1878

/**
 *	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
1879 1880
 *
 *	Returns:	0 for success, -Exxx on failure
1881 1882 1883 1884 1885
 *
 *	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.
 *
1886
 *	Similar to remap_pfn_range() (see mm/memory.c)
1887 1888 1889 1890 1891 1892 1893 1894 1895 1896 1897
 */
int remap_vmalloc_range(struct vm_area_struct *vma, void *addr,
						unsigned long pgoff)
{
	struct vm_struct *area;
	unsigned long uaddr = vma->vm_start;
	unsigned long usize = vma->vm_end - vma->vm_start;

	if ((PAGE_SIZE-1) & (unsigned long)addr)
		return -EINVAL;

N
Nick Piggin 已提交
1898
	area = find_vm_area(addr);
1899
	if (!area)
N
Nick Piggin 已提交
1900
		return -EINVAL;
1901 1902

	if (!(area->flags & VM_USERMAP))
N
Nick Piggin 已提交
1903
		return -EINVAL;
1904 1905

	if (usize + (pgoff << PAGE_SHIFT) > area->size - PAGE_SIZE)
N
Nick Piggin 已提交
1906
		return -EINVAL;
1907 1908 1909 1910

	addr += pgoff << PAGE_SHIFT;
	do {
		struct page *page = vmalloc_to_page(addr);
N
Nick Piggin 已提交
1911 1912
		int ret;

1913 1914 1915 1916 1917 1918 1919 1920 1921 1922 1923 1924
		ret = vm_insert_page(vma, uaddr, page);
		if (ret)
			return ret;

		uaddr += PAGE_SIZE;
		addr += PAGE_SIZE;
		usize -= PAGE_SIZE;
	} while (usize > 0);

	/* Prevent "things" like memory migration? VM_flags need a cleanup... */
	vma->vm_flags |= VM_RESERVED;

N
Nick Piggin 已提交
1925
	return 0;
1926 1927 1928
}
EXPORT_SYMBOL(remap_vmalloc_range);

1929 1930 1931 1932 1933 1934 1935
/*
 * Implement a stub for vmalloc_sync_all() if the architecture chose not to
 * have one.
 */
void  __attribute__((weak)) vmalloc_sync_all(void)
{
}
1936 1937


1938
static int f(pte_t *pte, pgtable_t table, unsigned long addr, void *data)
1939 1940 1941 1942 1943 1944 1945 1946
{
	/* apply_to_page_range() does all the hard work. */
	return 0;
}

/**
 *	alloc_vm_area - allocate a range of kernel address space
 *	@size:		size of the area
1947 1948
 *
 *	Returns:	NULL on failure, vm_struct on success
1949 1950 1951 1952 1953 1954 1955 1956 1957 1958 1959
 *
 *	This function reserves a range of kernel address space, and
 *	allocates pagetables to map that range.  No actual mappings
 *	are created.  If the kernel address space is not shared
 *	between processes, it syncs the pagetable across all
 *	processes.
 */
struct vm_struct *alloc_vm_area(size_t size)
{
	struct vm_struct *area;

1960 1961
	area = get_vm_area_caller(size, VM_IOREMAP,
				__builtin_return_address(0));
1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990
	if (area == NULL)
		return NULL;

	/*
	 * This ensures that page tables are constructed for this region
	 * of kernel virtual address space and mapped into init_mm.
	 */
	if (apply_to_page_range(&init_mm, (unsigned long)area->addr,
				area->size, f, NULL)) {
		free_vm_area(area);
		return NULL;
	}

	/* Make sure the pagetables are constructed in process kernel
	   mappings */
	vmalloc_sync_all();

	return area;
}
EXPORT_SYMBOL_GPL(alloc_vm_area);

void free_vm_area(struct vm_struct *area)
{
	struct vm_struct *ret;
	ret = remove_vm_area(area->addr);
	BUG_ON(ret != area);
	kfree(area);
}
EXPORT_SYMBOL_GPL(free_vm_area);
1991

1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 2034 2035 2036 2037 2038 2039 2040 2041 2042 2043 2044 2045 2046 2047 2048 2049 2050 2051 2052 2053 2054 2055 2056 2057 2058 2059 2060 2061 2062 2063 2064 2065 2066 2067 2068 2069 2070 2071 2072 2073 2074 2075 2076 2077 2078 2079 2080 2081 2082 2083 2084 2085 2086 2087 2088 2089 2090 2091 2092 2093 2094 2095 2096 2097 2098 2099 2100 2101 2102 2103 2104 2105 2106 2107 2108 2109 2110 2111 2112 2113 2114 2115 2116 2117 2118 2119 2120 2121 2122 2123 2124 2125 2126 2127 2128 2129 2130 2131 2132 2133 2134 2135 2136 2137 2138 2139 2140 2141 2142 2143 2144 2145 2146 2147 2148 2149 2150 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 2177 2178 2179 2180 2181 2182 2183 2184 2185 2186 2187 2188 2189 2190 2191 2192 2193 2194 2195 2196 2197 2198 2199 2200 2201 2202 2203 2204 2205 2206 2207 2208 2209 2210 2211 2212 2213 2214 2215 2216 2217 2218 2219 2220 2221 2222 2223 2224 2225 2226 2227 2228 2229 2230 2231 2232 2233 2234 2235 2236 2237 2238 2239 2240 2241 2242 2243 2244 2245 2246 2247 2248 2249 2250 2251 2252 2253 2254 2255 2256 2257 2258 2259 2260 2261 2262 2263 2264 2265 2266 2267 2268 2269 2270 2271
static struct vmap_area *node_to_va(struct rb_node *n)
{
	return n ? rb_entry(n, struct vmap_area, rb_node) : NULL;
}

/**
 * pvm_find_next_prev - find the next and prev vmap_area surrounding @end
 * @end: target address
 * @pnext: out arg for the next vmap_area
 * @pprev: out arg for the previous vmap_area
 *
 * Returns: %true if either or both of next and prev are found,
 *	    %false if no vmap_area exists
 *
 * Find vmap_areas end addresses of which enclose @end.  ie. if not
 * NULL, *pnext->va_end > @end and *pprev->va_end <= @end.
 */
static bool pvm_find_next_prev(unsigned long end,
			       struct vmap_area **pnext,
			       struct vmap_area **pprev)
{
	struct rb_node *n = vmap_area_root.rb_node;
	struct vmap_area *va = NULL;

	while (n) {
		va = rb_entry(n, struct vmap_area, rb_node);
		if (end < va->va_end)
			n = n->rb_left;
		else if (end > va->va_end)
			n = n->rb_right;
		else
			break;
	}

	if (!va)
		return false;

	if (va->va_end > end) {
		*pnext = va;
		*pprev = node_to_va(rb_prev(&(*pnext)->rb_node));
	} else {
		*pprev = va;
		*pnext = node_to_va(rb_next(&(*pprev)->rb_node));
	}
	return true;
}

/**
 * pvm_determine_end - find the highest aligned address between two vmap_areas
 * @pnext: in/out arg for the next vmap_area
 * @pprev: in/out arg for the previous vmap_area
 * @align: alignment
 *
 * Returns: determined end address
 *
 * Find the highest aligned address between *@pnext and *@pprev below
 * VMALLOC_END.  *@pnext and *@pprev are adjusted so that the aligned
 * down address is between the end addresses of the two vmap_areas.
 *
 * Please note that the address returned by this function may fall
 * inside *@pnext vmap_area.  The caller is responsible for checking
 * that.
 */
static unsigned long pvm_determine_end(struct vmap_area **pnext,
				       struct vmap_area **pprev,
				       unsigned long align)
{
	const unsigned long vmalloc_end = VMALLOC_END & ~(align - 1);
	unsigned long addr;

	if (*pnext)
		addr = min((*pnext)->va_start & ~(align - 1), vmalloc_end);
	else
		addr = vmalloc_end;

	while (*pprev && (*pprev)->va_end > addr) {
		*pnext = *pprev;
		*pprev = node_to_va(rb_prev(&(*pnext)->rb_node));
	}

	return addr;
}

/**
 * pcpu_get_vm_areas - allocate vmalloc areas for percpu allocator
 * @offsets: array containing offset of each area
 * @sizes: array containing size of each area
 * @nr_vms: the number of areas to allocate
 * @align: alignment, all entries in @offsets and @sizes must be aligned to this
 * @gfp_mask: allocation mask
 *
 * 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
 * congruent vmalloc areas for it.  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.
 *
 * Despite its complicated look, this allocator is rather simple.  It
 * does everything top-down and scans areas from the end looking for
 * matching slot.  While scanning, if any of the areas overlaps with
 * existing vmap_area, the base address is pulled down to fit the
 * area.  Scanning is repeated till all the areas fit and then all
 * necessary data structres are inserted and the result is returned.
 */
struct vm_struct **pcpu_get_vm_areas(const unsigned long *offsets,
				     const size_t *sizes, int nr_vms,
				     size_t align, gfp_t gfp_mask)
{
	const unsigned long vmalloc_start = ALIGN(VMALLOC_START, align);
	const unsigned long vmalloc_end = VMALLOC_END & ~(align - 1);
	struct vmap_area **vas, *prev, *next;
	struct vm_struct **vms;
	int area, area2, last_area, term_area;
	unsigned long base, start, end, last_end;
	bool purged = false;

	gfp_mask &= GFP_RECLAIM_MASK;

	/* verify parameters and allocate data structures */
	BUG_ON(align & ~PAGE_MASK || !is_power_of_2(align));
	for (last_area = 0, area = 0; area < nr_vms; area++) {
		start = offsets[area];
		end = start + sizes[area];

		/* is everything aligned properly? */
		BUG_ON(!IS_ALIGNED(offsets[area], align));
		BUG_ON(!IS_ALIGNED(sizes[area], align));

		/* detect the area with the highest address */
		if (start > offsets[last_area])
			last_area = area;

		for (area2 = 0; area2 < nr_vms; area2++) {
			unsigned long start2 = offsets[area2];
			unsigned long end2 = start2 + sizes[area2];

			if (area2 == area)
				continue;

			BUG_ON(start2 >= start && start2 < end);
			BUG_ON(end2 <= end && end2 > start);
		}
	}
	last_end = offsets[last_area] + sizes[last_area];

	if (vmalloc_end - vmalloc_start < last_end) {
		WARN_ON(true);
		return NULL;
	}

	vms = kzalloc(sizeof(vms[0]) * nr_vms, gfp_mask);
	vas = kzalloc(sizeof(vas[0]) * nr_vms, gfp_mask);
	if (!vas || !vms)
		goto err_free;

	for (area = 0; area < nr_vms; area++) {
		vas[area] = kzalloc(sizeof(struct vmap_area), gfp_mask);
		vms[area] = kzalloc(sizeof(struct vm_struct), gfp_mask);
		if (!vas[area] || !vms[area])
			goto err_free;
	}
retry:
	spin_lock(&vmap_area_lock);

	/* start scanning - we scan from the top, begin with the last area */
	area = term_area = last_area;
	start = offsets[area];
	end = start + sizes[area];

	if (!pvm_find_next_prev(vmap_area_pcpu_hole, &next, &prev)) {
		base = vmalloc_end - last_end;
		goto found;
	}
	base = pvm_determine_end(&next, &prev, align) - end;

	while (true) {
		BUG_ON(next && next->va_end <= base + end);
		BUG_ON(prev && prev->va_end > base + end);

		/*
		 * base might have underflowed, add last_end before
		 * comparing.
		 */
		if (base + last_end < vmalloc_start + last_end) {
			spin_unlock(&vmap_area_lock);
			if (!purged) {
				purge_vmap_area_lazy();
				purged = true;
				goto retry;
			}
			goto err_free;
		}

		/*
		 * If next overlaps, move base downwards so that it's
		 * right below next and then recheck.
		 */
		if (next && next->va_start < base + end) {
			base = pvm_determine_end(&next, &prev, align) - end;
			term_area = area;
			continue;
		}

		/*
		 * If prev overlaps, shift down next and prev and move
		 * base so that it's right below new next and then
		 * recheck.
		 */
		if (prev && prev->va_end > base + start)  {
			next = prev;
			prev = node_to_va(rb_prev(&next->rb_node));
			base = pvm_determine_end(&next, &prev, align) - end;
			term_area = area;
			continue;
		}

		/*
		 * This area fits, move on to the previous one.  If
		 * the previous one is the terminal one, we're done.
		 */
		area = (area + nr_vms - 1) % nr_vms;
		if (area == term_area)
			break;
		start = offsets[area];
		end = start + sizes[area];
		pvm_find_next_prev(base + end, &next, &prev);
	}
found:
	/* we've found a fitting base, insert all va's */
	for (area = 0; area < nr_vms; area++) {
		struct vmap_area *va = vas[area];

		va->va_start = base + offsets[area];
		va->va_end = va->va_start + sizes[area];
		__insert_vmap_area(va);
	}

	vmap_area_pcpu_hole = base + offsets[last_area];

	spin_unlock(&vmap_area_lock);

	/* insert all vm's */
	for (area = 0; area < nr_vms; area++)
		insert_vmalloc_vm(vms[area], vas[area], VM_ALLOC,
				  pcpu_get_vm_areas);

	kfree(vas);
	return vms;

err_free:
	for (area = 0; area < nr_vms; area++) {
		if (vas)
			kfree(vas[area]);
		if (vms)
			kfree(vms[area]);
	}
	kfree(vas);
	kfree(vms);
	return NULL;
}

/**
 * pcpu_free_vm_areas - free vmalloc areas for percpu allocator
 * @vms: vm_struct pointer array returned by pcpu_get_vm_areas()
 * @nr_vms: the number of allocated areas
 *
 * Free vm_structs and the array allocated by pcpu_get_vm_areas().
 */
void pcpu_free_vm_areas(struct vm_struct **vms, int nr_vms)
{
	int i;

	for (i = 0; i < nr_vms; i++)
		free_vm_area(vms[i]);
	kfree(vms);
}
2272 2273 2274 2275 2276 2277 2278 2279 2280 2281 2282 2283 2284 2285 2286 2287 2288 2289 2290 2291 2292 2293 2294 2295 2296 2297 2298 2299 2300 2301 2302 2303 2304

#ifdef CONFIG_PROC_FS
static void *s_start(struct seq_file *m, loff_t *pos)
{
	loff_t n = *pos;
	struct vm_struct *v;

	read_lock(&vmlist_lock);
	v = vmlist;
	while (n > 0 && v) {
		n--;
		v = v->next;
	}
	if (!n)
		return v;

	return NULL;

}

static void *s_next(struct seq_file *m, void *p, loff_t *pos)
{
	struct vm_struct *v = p;

	++*pos;
	return v->next;
}

static void s_stop(struct seq_file *m, void *p)
{
	read_unlock(&vmlist_lock);
}

E
Eric Dumazet 已提交
2305 2306 2307 2308 2309 2310 2311 2312 2313 2314 2315 2316 2317 2318 2319 2320 2321 2322 2323
static void show_numa_info(struct seq_file *m, struct vm_struct *v)
{
	if (NUMA_BUILD) {
		unsigned int nr, *counters = m->private;

		if (!counters)
			return;

		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]);
	}
}

2324 2325 2326 2327 2328 2329 2330
static int s_show(struct seq_file *m, void *p)
{
	struct vm_struct *v = p;

	seq_printf(m, "0x%p-0x%p %7ld",
		v->addr, v->addr + v->size, v->size);

2331
	if (v->caller) {
H
Hugh Dickins 已提交
2332
		char buff[KSYM_SYMBOL_LEN];
2333 2334 2335 2336 2337 2338

		seq_putc(m, ' ');
		sprint_symbol(buff, (unsigned long)v->caller);
		seq_puts(m, buff);
	}

2339 2340 2341 2342 2343 2344 2345 2346 2347 2348 2349 2350 2351 2352 2353 2354 2355 2356 2357 2358 2359
	if (v->nr_pages)
		seq_printf(m, " pages=%d", v->nr_pages);

	if (v->phys_addr)
		seq_printf(m, " phys=%lx", v->phys_addr);

	if (v->flags & VM_IOREMAP)
		seq_printf(m, " ioremap");

	if (v->flags & VM_ALLOC)
		seq_printf(m, " vmalloc");

	if (v->flags & VM_MAP)
		seq_printf(m, " vmap");

	if (v->flags & VM_USERMAP)
		seq_printf(m, " user");

	if (v->flags & VM_VPAGES)
		seq_printf(m, " vpages");

E
Eric Dumazet 已提交
2360
	show_numa_info(m, v);
2361 2362 2363 2364
	seq_putc(m, '\n');
	return 0;
}

2365
static const struct seq_operations vmalloc_op = {
2366 2367 2368 2369 2370
	.start = s_start,
	.next = s_next,
	.stop = s_stop,
	.show = s_show,
};
2371 2372 2373 2374 2375 2376 2377 2378 2379 2380 2381 2382 2383 2384 2385 2386 2387 2388 2389 2390 2391 2392 2393 2394 2395 2396 2397 2398 2399 2400

static int vmalloc_open(struct inode *inode, struct file *file)
{
	unsigned int *ptr = NULL;
	int ret;

	if (NUMA_BUILD)
		ptr = kmalloc(nr_node_ids * sizeof(unsigned int), GFP_KERNEL);
	ret = seq_open(file, &vmalloc_op);
	if (!ret) {
		struct seq_file *m = file->private_data;
		m->private = ptr;
	} else
		kfree(ptr);
	return ret;
}

static const struct file_operations proc_vmalloc_operations = {
	.open		= vmalloc_open,
	.read		= seq_read,
	.llseek		= seq_lseek,
	.release	= seq_release_private,
};

static int __init proc_vmalloc_init(void)
{
	proc_create("vmallocinfo", S_IRUSR, NULL, &proc_vmalloc_operations);
	return 0;
}
module_init(proc_vmalloc_init);
2401 2402
#endif