task_mmu.c 19.1 KB
Newer Older
L
Linus Torvalds 已提交
1 2 3 4
#include <linux/mm.h>
#include <linux/hugetlb.h>
#include <linux/mount.h>
#include <linux/seq_file.h>
M
Mauricio Lin 已提交
5
#include <linux/highmem.h>
K
Kees Cook 已提交
6
#include <linux/ptrace.h>
7 8
#include <linux/pagemap.h>
#include <linux/mempolicy.h>
9 10
#include <linux/swap.h>
#include <linux/swapops.h>
M
Mauricio Lin 已提交
11

L
Linus Torvalds 已提交
12 13
#include <asm/elf.h>
#include <asm/uaccess.h>
M
Mauricio Lin 已提交
14
#include <asm/tlbflush.h>
L
Linus Torvalds 已提交
15 16
#include "internal.h"

17
void task_mem(struct seq_file *m, struct mm_struct *mm)
L
Linus Torvalds 已提交
18 19
{
	unsigned long data, text, lib;
20 21 22 23 24 25 26 27 28 29 30 31 32 33 34
	unsigned long hiwater_vm, total_vm, hiwater_rss, total_rss;

	/*
	 * Note: to minimize their overhead, mm maintains hiwater_vm and
	 * hiwater_rss only when about to *lower* total_vm or rss.  Any
	 * collector of these hiwater stats must therefore get total_vm
	 * and rss too, which will usually be the higher.  Barriers? not
	 * worth the effort, such snapshots can always be inconsistent.
	 */
	hiwater_vm = total_vm = mm->total_vm;
	if (hiwater_vm < mm->hiwater_vm)
		hiwater_vm = mm->hiwater_vm;
	hiwater_rss = total_rss = get_mm_rss(mm);
	if (hiwater_rss < mm->hiwater_rss)
		hiwater_rss = mm->hiwater_rss;
L
Linus Torvalds 已提交
35 36 37 38

	data = mm->total_vm - mm->shared_vm - mm->stack_vm;
	text = (PAGE_ALIGN(mm->end_code) - (mm->start_code & PAGE_MASK)) >> 10;
	lib = (mm->exec_vm << (PAGE_SHIFT-10)) - text;
39
	seq_printf(m,
40
		"VmPeak:\t%8lu kB\n"
L
Linus Torvalds 已提交
41 42
		"VmSize:\t%8lu kB\n"
		"VmLck:\t%8lu kB\n"
43
		"VmHWM:\t%8lu kB\n"
L
Linus Torvalds 已提交
44 45 46 47 48 49
		"VmRSS:\t%8lu kB\n"
		"VmData:\t%8lu kB\n"
		"VmStk:\t%8lu kB\n"
		"VmExe:\t%8lu kB\n"
		"VmLib:\t%8lu kB\n"
		"VmPTE:\t%8lu kB\n",
50 51
		hiwater_vm << (PAGE_SHIFT-10),
		(total_vm - mm->reserved_vm) << (PAGE_SHIFT-10),
L
Linus Torvalds 已提交
52
		mm->locked_vm << (PAGE_SHIFT-10),
53 54
		hiwater_rss << (PAGE_SHIFT-10),
		total_rss << (PAGE_SHIFT-10),
L
Linus Torvalds 已提交
55 56 57 58 59 60 61 62 63 64 65 66 67
		data << (PAGE_SHIFT-10),
		mm->stack_vm << (PAGE_SHIFT-10), text, lib,
		(PTRS_PER_PTE*sizeof(pte_t)*mm->nr_ptes) >> 10);
}

unsigned long task_vsize(struct mm_struct *mm)
{
	return PAGE_SIZE * mm->total_vm;
}

int task_statm(struct mm_struct *mm, int *shared, int *text,
	       int *data, int *resident)
{
68
	*shared = get_mm_counter(mm, file_rss);
L
Linus Torvalds 已提交
69 70 71
	*text = (PAGE_ALIGN(mm->end_code) - (mm->start_code & PAGE_MASK))
								>> PAGE_SHIFT;
	*data = mm->total_vm - mm->shared_vm;
72
	*resident = *shared + get_mm_counter(mm, anon_rss);
L
Linus Torvalds 已提交
73 74 75 76 77 78 79 80 81 82 83
	return mm->total_vm;
}

static void pad_len_spaces(struct seq_file *m, int len)
{
	len = 25 + sizeof(void*) * 6 - len;
	if (len < 1)
		len = 1;
	seq_printf(m, "%*c", len, ' ');
}

84 85 86 87 88 89 90 91
static void vma_stop(struct proc_maps_private *priv, struct vm_area_struct *vma)
{
	if (vma && vma != priv->tail_vma) {
		struct mm_struct *mm = vma->vm_mm;
		up_read(&mm->mmap_sem);
		mmput(mm);
	}
}
92

93
static void *m_start(struct seq_file *m, loff_t *pos)
M
Mauricio Lin 已提交
94
{
95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121
	struct proc_maps_private *priv = m->private;
	unsigned long last_addr = m->version;
	struct mm_struct *mm;
	struct vm_area_struct *vma, *tail_vma = NULL;
	loff_t l = *pos;

	/* Clear the per syscall fields in priv */
	priv->task = NULL;
	priv->tail_vma = NULL;

	/*
	 * We remember last_addr rather than next_addr to hit with
	 * mmap_cache most of the time. We have zero last_addr at
	 * the beginning and also after lseek. We will have -1 last_addr
	 * after the end of the vmas.
	 */

	if (last_addr == -1UL)
		return NULL;

	priv->task = get_pid_task(priv->pid, PIDTYPE_PID);
	if (!priv->task)
		return NULL;

	mm = mm_for_maps(priv->task);
	if (!mm)
		return NULL;
122
	down_read(&mm->mmap_sem);
123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183

	tail_vma = get_gate_vma(priv->task);
	priv->tail_vma = tail_vma;

	/* Start with last addr hint */
	vma = find_vma(mm, last_addr);
	if (last_addr && vma) {
		vma = vma->vm_next;
		goto out;
	}

	/*
	 * Check the vma index is within the range and do
	 * sequential scan until m_index.
	 */
	vma = NULL;
	if ((unsigned long)l < mm->map_count) {
		vma = mm->mmap;
		while (l-- && vma)
			vma = vma->vm_next;
		goto out;
	}

	if (l != mm->map_count)
		tail_vma = NULL; /* After gate vma */

out:
	if (vma)
		return vma;

	/* End of vmas has been reached */
	m->version = (tail_vma != NULL)? 0: -1UL;
	up_read(&mm->mmap_sem);
	mmput(mm);
	return tail_vma;
}

static void *m_next(struct seq_file *m, void *v, loff_t *pos)
{
	struct proc_maps_private *priv = m->private;
	struct vm_area_struct *vma = v;
	struct vm_area_struct *tail_vma = priv->tail_vma;

	(*pos)++;
	if (vma && (vma != tail_vma) && vma->vm_next)
		return vma->vm_next;
	vma_stop(priv, vma);
	return (vma != tail_vma)? tail_vma: NULL;
}

static void m_stop(struct seq_file *m, void *v)
{
	struct proc_maps_private *priv = m->private;
	struct vm_area_struct *vma = v;

	vma_stop(priv, vma);
	if (priv->task)
		put_task_struct(priv->task);
}

static int do_maps_open(struct inode *inode, struct file *file,
184
			const struct seq_operations *ops)
185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200
{
	struct proc_maps_private *priv;
	int ret = -ENOMEM;
	priv = kzalloc(sizeof(*priv), GFP_KERNEL);
	if (priv) {
		priv->pid = proc_pid(inode);
		ret = seq_open(file, ops);
		if (!ret) {
			struct seq_file *m = file->private_data;
			m->private = priv;
		} else {
			kfree(priv);
		}
	}
	return ret;
}
M
Mauricio Lin 已提交
201

202
static void show_map_vma(struct seq_file *m, struct vm_area_struct *vma)
L
Linus Torvalds 已提交
203
{
M
Mauricio Lin 已提交
204 205 206
	struct mm_struct *mm = vma->vm_mm;
	struct file *file = vma->vm_file;
	int flags = vma->vm_flags;
L
Linus Torvalds 已提交
207
	unsigned long ino = 0;
208
	unsigned long long pgoff = 0;
L
Linus Torvalds 已提交
209 210 211 212
	dev_t dev = 0;
	int len;

	if (file) {
213
		struct inode *inode = vma->vm_file->f_path.dentry->d_inode;
L
Linus Torvalds 已提交
214 215
		dev = inode->i_sb->s_dev;
		ino = inode->i_ino;
216
		pgoff = ((loff_t)vma->vm_pgoff) << PAGE_SHIFT;
L
Linus Torvalds 已提交
217 218
	}

219
	seq_printf(m, "%08lx-%08lx %c%c%c%c %08llx %02x:%02x %lu %n",
M
Mauricio Lin 已提交
220 221
			vma->vm_start,
			vma->vm_end,
L
Linus Torvalds 已提交
222 223 224 225
			flags & VM_READ ? 'r' : '-',
			flags & VM_WRITE ? 'w' : '-',
			flags & VM_EXEC ? 'x' : '-',
			flags & VM_MAYSHARE ? 's' : 'p',
226
			pgoff,
L
Linus Torvalds 已提交
227 228 229 230 231 232
			MAJOR(dev), MINOR(dev), ino, &len);

	/*
	 * Print the dentry name for named mappings, and a
	 * special [heap] marker for the heap:
	 */
M
Mauricio Lin 已提交
233
	if (file) {
L
Linus Torvalds 已提交
234
		pad_len_spaces(m, len);
235
		seq_path(m, &file->f_path, "\n");
L
Linus Torvalds 已提交
236
	} else {
237 238 239 240
		const char *name = arch_vma_name(vma);
		if (!name) {
			if (mm) {
				if (vma->vm_start <= mm->start_brk &&
M
Mauricio Lin 已提交
241
						vma->vm_end >= mm->brk) {
242 243 244 245
					name = "[heap]";
				} else if (vma->vm_start <= mm->start_stack &&
					   vma->vm_end >= mm->start_stack) {
					name = "[stack]";
L
Linus Torvalds 已提交
246
				}
247 248
			} else {
				name = "[vdso]";
L
Linus Torvalds 已提交
249
			}
250 251
		}
		if (name) {
L
Linus Torvalds 已提交
252
			pad_len_spaces(m, len);
253
			seq_puts(m, name);
L
Linus Torvalds 已提交
254 255 256
		}
	}
	seq_putc(m, '\n');
257 258 259 260 261 262 263 264 265
}

static int show_map(struct seq_file *m, void *v)
{
	struct vm_area_struct *vma = v;
	struct proc_maps_private *priv = m->private;
	struct task_struct *task = priv->task;

	show_map_vma(m, vma);
M
Mauricio Lin 已提交
266 267 268

	if (m->count < m->size)  /* vma is copied successfully */
		m->version = (vma != get_gate_vma(task))? vma->vm_start: 0;
L
Linus Torvalds 已提交
269 270 271
	return 0;
}

272
static const struct seq_operations proc_pid_maps_op = {
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
	.start	= m_start,
	.next	= m_next,
	.stop	= m_stop,
	.show	= show_map
};

static int maps_open(struct inode *inode, struct file *file)
{
	return do_maps_open(inode, file, &proc_pid_maps_op);
}

const struct file_operations proc_maps_operations = {
	.open		= maps_open,
	.read		= seq_read,
	.llseek		= seq_lseek,
	.release	= seq_release_private,
};

/*
 * Proportional Set Size(PSS): my share of RSS.
 *
 * PSS of a process is the count of pages it has in memory, where each
 * page is divided by the number of processes sharing it.  So if a
 * process has 1000 pages all to itself, and 1000 shared with one other
 * process, its PSS will be 1500.
 *
 * To keep (accumulated) division errors low, we adopt a 64bit
 * fixed-point pss counter to minimize division errors. So (pss >>
 * PSS_SHIFT) would be the real byte count.
 *
 * A shift of 12 before division means (assuming 4K page size):
 * 	- 1M 3-user-pages add up to 8KB errors;
 * 	- supports mapcount up to 2^24, or 16M;
 * 	- supports PSS up to 2^52 bytes, or 4PB.
 */
#define PSS_SHIFT 12

310
#ifdef CONFIG_PROC_PAGE_MONITOR
P
Peter Zijlstra 已提交
311
struct mem_size_stats {
312 313 314 315 316 317 318
	struct vm_area_struct *vma;
	unsigned long resident;
	unsigned long shared_clean;
	unsigned long shared_dirty;
	unsigned long private_clean;
	unsigned long private_dirty;
	unsigned long referenced;
P
Peter Zijlstra 已提交
319
	unsigned long swap;
320 321 322
	u64 pss;
};

323
static int smaps_pte_range(pmd_t *pmd, unsigned long addr, unsigned long end,
D
Dave Hansen 已提交
324
			   struct mm_walk *walk)
M
Mauricio Lin 已提交
325
{
D
Dave Hansen 已提交
326
	struct mem_size_stats *mss = walk->private;
327
	struct vm_area_struct *vma = mss->vma;
M
Mauricio Lin 已提交
328
	pte_t *pte, ptent;
329
	spinlock_t *ptl;
M
Mauricio Lin 已提交
330
	struct page *page;
331
	int mapcount;
M
Mauricio Lin 已提交
332

333
	pte = pte_offset_map_lock(vma->vm_mm, pmd, addr, &ptl);
334
	for (; addr != end; pte++, addr += PAGE_SIZE) {
M
Mauricio Lin 已提交
335
		ptent = *pte;
P
Peter Zijlstra 已提交
336 337 338 339 340 341

		if (is_swap_pte(ptent)) {
			mss->swap += PAGE_SIZE;
			continue;
		}

342
		if (!pte_present(ptent))
M
Mauricio Lin 已提交
343 344 345
			continue;

		mss->resident += PAGE_SIZE;
N
Nick Piggin 已提交
346 347 348

		page = vm_normal_page(vma, addr, ptent);
		if (!page)
M
Mauricio Lin 已提交
349 350
			continue;

351 352 353
		/* Accumulate the size in pages that have been accessed. */
		if (pte_young(ptent) || PageReferenced(page))
			mss->referenced += PAGE_SIZE;
354 355
		mapcount = page_mapcount(page);
		if (mapcount >= 2) {
M
Mauricio Lin 已提交
356 357 358 359
			if (pte_dirty(ptent))
				mss->shared_dirty += PAGE_SIZE;
			else
				mss->shared_clean += PAGE_SIZE;
360
			mss->pss += (PAGE_SIZE << PSS_SHIFT) / mapcount;
M
Mauricio Lin 已提交
361 362 363 364 365
		} else {
			if (pte_dirty(ptent))
				mss->private_dirty += PAGE_SIZE;
			else
				mss->private_clean += PAGE_SIZE;
366
			mss->pss += (PAGE_SIZE << PSS_SHIFT);
M
Mauricio Lin 已提交
367
		}
368
	}
369 370
	pte_unmap_unlock(pte - 1, ptl);
	cond_resched();
371
	return 0;
M
Mauricio Lin 已提交
372 373 374 375
}

static int show_smap(struct seq_file *m, void *v)
{
376 377
	struct proc_maps_private *priv = m->private;
	struct task_struct *task = priv->task;
M
Mauricio Lin 已提交
378 379
	struct vm_area_struct *vma = v;
	struct mem_size_stats mss;
D
Dave Hansen 已提交
380 381 382 383 384
	struct mm_walk smaps_walk = {
		.pmd_entry = smaps_pte_range,
		.mm = vma->vm_mm,
		.private = &mss,
	};
M
Mauricio Lin 已提交
385 386

	memset(&mss, 0, sizeof mss);
387
	mss.vma = vma;
N
Nick Piggin 已提交
388
	if (vma->vm_mm && !is_vm_hugetlb_page(vma))
D
Dave Hansen 已提交
389
		walk_page_range(vma->vm_start, vma->vm_end, &smaps_walk);
390

391
	show_map_vma(m, vma);
392 393 394 395 396 397 398 399 400

	seq_printf(m,
		   "Size:           %8lu kB\n"
		   "Rss:            %8lu kB\n"
		   "Pss:            %8lu kB\n"
		   "Shared_Clean:   %8lu kB\n"
		   "Shared_Dirty:   %8lu kB\n"
		   "Private_Clean:  %8lu kB\n"
		   "Private_Dirty:  %8lu kB\n"
P
Peter Zijlstra 已提交
401
		   "Referenced:     %8lu kB\n"
402
		   "Swap:           %8lu kB\n"
403 404
		   "KernelPageSize: %8lu kB\n"
		   "MMUPageSize:    %8lu kB\n",
405 406 407 408 409 410 411
		   (vma->vm_end - vma->vm_start) >> 10,
		   mss.resident >> 10,
		   (unsigned long)(mss.pss >> (10 + PSS_SHIFT)),
		   mss.shared_clean  >> 10,
		   mss.shared_dirty  >> 10,
		   mss.private_clean >> 10,
		   mss.private_dirty >> 10,
P
Peter Zijlstra 已提交
412
		   mss.referenced >> 10,
413
		   mss.swap >> 10,
414 415
		   vma_kernel_pagesize(vma) >> 10,
		   vma_mmu_pagesize(vma) >> 10);
416

417 418 419
	if (m->count < m->size)  /* vma is copied successfully */
		m->version = (vma != get_gate_vma(task)) ? vma->vm_start : 0;
	return 0;
M
Mauricio Lin 已提交
420 421
}

422
static const struct seq_operations proc_pid_smaps_op = {
423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441
	.start	= m_start,
	.next	= m_next,
	.stop	= m_stop,
	.show	= show_smap
};

static int smaps_open(struct inode *inode, struct file *file)
{
	return do_maps_open(inode, file, &proc_pid_smaps_op);
}

const struct file_operations proc_smaps_operations = {
	.open		= smaps_open,
	.read		= seq_read,
	.llseek		= seq_lseek,
	.release	= seq_release_private,
};

static int clear_refs_pte_range(pmd_t *pmd, unsigned long addr,
D
Dave Hansen 已提交
442
				unsigned long end, struct mm_walk *walk)
443
{
D
Dave Hansen 已提交
444
	struct vm_area_struct *vma = walk->private;
445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467
	pte_t *pte, ptent;
	spinlock_t *ptl;
	struct page *page;

	pte = pte_offset_map_lock(vma->vm_mm, pmd, addr, &ptl);
	for (; addr != end; pte++, addr += PAGE_SIZE) {
		ptent = *pte;
		if (!pte_present(ptent))
			continue;

		page = vm_normal_page(vma, addr, ptent);
		if (!page)
			continue;

		/* Clear accessed and referenced bits. */
		ptep_test_and_clear_young(vma, addr, pte);
		ClearPageReferenced(page);
	}
	pte_unmap_unlock(pte - 1, ptl);
	cond_resched();
	return 0;
}

468 469
static ssize_t clear_refs_write(struct file *file, const char __user *buf,
				size_t count, loff_t *ppos)
470
{
471 472 473
	struct task_struct *task;
	char buffer[PROC_NUMBUF], *end;
	struct mm_struct *mm;
474 475
	struct vm_area_struct *vma;

476 477 478 479 480 481 482 483 484 485 486 487 488 489
	memset(buffer, 0, sizeof(buffer));
	if (count > sizeof(buffer) - 1)
		count = sizeof(buffer) - 1;
	if (copy_from_user(buffer, buf, count))
		return -EFAULT;
	if (!simple_strtol(buffer, &end, 0))
		return -EINVAL;
	if (*end == '\n')
		end++;
	task = get_proc_task(file->f_path.dentry->d_inode);
	if (!task)
		return -ESRCH;
	mm = get_task_mm(task);
	if (mm) {
490 491 492 493
		struct mm_walk clear_refs_walk = {
			.pmd_entry = clear_refs_pte_range,
			.mm = mm,
		};
494
		down_read(&mm->mmap_sem);
D
Dave Hansen 已提交
495 496
		for (vma = mm->mmap; vma; vma = vma->vm_next) {
			clear_refs_walk.private = vma;
497
			if (!is_vm_hugetlb_page(vma))
D
Dave Hansen 已提交
498 499 500
				walk_page_range(vma->vm_start, vma->vm_end,
						&clear_refs_walk);
		}
501 502 503 504 505 506 507 508
		flush_tlb_mm(mm);
		up_read(&mm->mmap_sem);
		mmput(mm);
	}
	put_task_struct(task);
	if (end - buffer == 0)
		return -EIO;
	return end - buffer;
509 510
}

511 512 513 514
const struct file_operations proc_clear_refs_operations = {
	.write		= clear_refs_write,
};

515
struct pagemapread {
516
	u64 __user *out, *end;
517 518
};

519 520 521 522 523 524 525 526 527 528 529 530 531 532 533
#define PM_ENTRY_BYTES      sizeof(u64)
#define PM_STATUS_BITS      3
#define PM_STATUS_OFFSET    (64 - PM_STATUS_BITS)
#define PM_STATUS_MASK      (((1LL << PM_STATUS_BITS) - 1) << PM_STATUS_OFFSET)
#define PM_STATUS(nr)       (((nr) << PM_STATUS_OFFSET) & PM_STATUS_MASK)
#define PM_PSHIFT_BITS      6
#define PM_PSHIFT_OFFSET    (PM_STATUS_OFFSET - PM_PSHIFT_BITS)
#define PM_PSHIFT_MASK      (((1LL << PM_PSHIFT_BITS) - 1) << PM_PSHIFT_OFFSET)
#define PM_PSHIFT(x)        (((u64) (x) << PM_PSHIFT_OFFSET) & PM_PSHIFT_MASK)
#define PM_PFRAME_MASK      ((1LL << PM_PSHIFT_OFFSET) - 1)
#define PM_PFRAME(x)        ((x) & PM_PFRAME_MASK)

#define PM_PRESENT          PM_STATUS(4LL)
#define PM_SWAP             PM_STATUS(2LL)
#define PM_NOT_PRESENT      PM_PSHIFT(PAGE_SHIFT)
534 535 536 537 538 539 540
#define PM_END_OF_BUFFER    1

static int add_to_pagemap(unsigned long addr, u64 pfn,
			  struct pagemapread *pm)
{
	if (put_user(pfn, pm->out))
		return -EFAULT;
541 542 543
	pm->out++;
	if (pm->out >= pm->end)
		return PM_END_OF_BUFFER;
544 545 546 547
	return 0;
}

static int pagemap_pte_hole(unsigned long start, unsigned long end,
D
Dave Hansen 已提交
548
				struct mm_walk *walk)
549
{
D
Dave Hansen 已提交
550
	struct pagemapread *pm = walk->private;
551 552 553 554 555 556 557 558 559 560
	unsigned long addr;
	int err = 0;
	for (addr = start; addr < end; addr += PAGE_SIZE) {
		err = add_to_pagemap(addr, PM_NOT_PRESENT, pm);
		if (err)
			break;
	}
	return err;
}

561
static u64 swap_pte_to_pagemap_entry(pte_t pte)
562 563
{
	swp_entry_t e = pte_to_swp_entry(pte);
564
	return swp_type(e) | (swp_offset(e) << MAX_SWAPFILES_SHIFT);
565 566
}

567
static u64 pte_to_pagemap_entry(pte_t pte)
568
{
569
	u64 pme = 0;
570 571 572 573 574 575 576 577 578
	if (is_swap_pte(pte))
		pme = PM_PFRAME(swap_pte_to_pagemap_entry(pte))
			| PM_PSHIFT(PAGE_SHIFT) | PM_SWAP;
	else if (pte_present(pte))
		pme = PM_PFRAME(pte_pfn(pte))
			| PM_PSHIFT(PAGE_SHIFT) | PM_PRESENT;
	return pme;
}

579
static int pagemap_pte_range(pmd_t *pmd, unsigned long addr, unsigned long end,
D
Dave Hansen 已提交
580
			     struct mm_walk *walk)
581
{
582
	struct vm_area_struct *vma;
D
Dave Hansen 已提交
583
	struct pagemapread *pm = walk->private;
584 585 586
	pte_t *pte;
	int err = 0;

587 588
	/* find the first VMA at or above 'addr' */
	vma = find_vma(walk->mm, addr);
589 590
	for (; addr != end; addr += PAGE_SIZE) {
		u64 pfn = PM_NOT_PRESENT;
591 592 593 594 595 596 597 598 599 600 601 602 603 604 605

		/* check to see if we've left 'vma' behind
		 * and need a new, higher one */
		if (vma && (addr >= vma->vm_end))
			vma = find_vma(walk->mm, addr);

		/* check that 'vma' actually covers this address,
		 * and that it isn't a huge page vma */
		if (vma && (vma->vm_start <= addr) &&
		    !is_vm_hugetlb_page(vma)) {
			pte = pte_offset_map(pmd, addr);
			pfn = pte_to_pagemap_entry(*pte);
			/* unmap before userspace copy */
			pte_unmap(pte);
		}
606 607 608 609 610 611 612 613 614 615 616 617 618
		err = add_to_pagemap(addr, pfn, pm);
		if (err)
			return err;
	}

	cond_resched();

	return err;
}

/*
 * /proc/pid/pagemap - an array mapping virtual pages to pfns
 *
619 620 621 622 623 624 625 626 627 628 629 630 631 632
 * For each page in the address space, this file contains one 64-bit entry
 * consisting of the following:
 *
 * Bits 0-55  page frame number (PFN) if present
 * Bits 0-4   swap type if swapped
 * Bits 5-55  swap offset if swapped
 * Bits 55-60 page shift (page size = 1<<page shift)
 * Bit  61    reserved for future use
 * Bit  62    page swapped
 * Bit  63    page present
 *
 * If the page is not present but in swap, then the PFN contains an
 * encoding of the swap file number and the page's offset into the
 * swap. Unmapped pages return a null PFN. This allows determining
633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649
 * precisely which pages are mapped (or in swap) and comparing mapped
 * pages between processes.
 *
 * Efficient users of this interface will use /proc/pid/maps to
 * determine which areas of memory are actually mapped and llseek to
 * skip over unmapped regions.
 */
static ssize_t pagemap_read(struct file *file, char __user *buf,
			    size_t count, loff_t *ppos)
{
	struct task_struct *task = get_proc_task(file->f_path.dentry->d_inode);
	struct page **pages, *page;
	unsigned long uaddr, uend;
	struct mm_struct *mm;
	struct pagemapread pm;
	int pagecount;
	int ret = -ESRCH;
650
	struct mm_walk pagemap_walk = {};
651 652 653 654
	unsigned long src;
	unsigned long svpfn;
	unsigned long start_vaddr;
	unsigned long end_vaddr;
655 656 657 658 659

	if (!task)
		goto out;

	ret = -EACCES;
660
	if (!ptrace_may_access(task, PTRACE_MODE_READ))
661
		goto out_task;
662 663 664

	ret = -EINVAL;
	/* file position must be aligned */
665
	if ((*ppos % PM_ENTRY_BYTES) || (count % PM_ENTRY_BYTES))
666
		goto out_task;
667 668

	ret = 0;
669 670 671 672

	if (!count)
		goto out_task;

673 674
	mm = get_task_mm(task);
	if (!mm)
675
		goto out_task;
676

677

678 679 680
	uaddr = (unsigned long)buf & PAGE_MASK;
	uend = (unsigned long)(buf + count);
	pagecount = (PAGE_ALIGN(uend) - uaddr) / PAGE_SIZE;
681 682 683 684 685
	ret = 0;
	if (pagecount == 0)
		goto out_mm;
	pages = kcalloc(pagecount, sizeof(struct page *), GFP_KERNEL);
	ret = -ENOMEM;
686
	if (!pages)
687
		goto out_mm;
688 689 690 691 692 693 694 695 696

	down_read(&current->mm->mmap_sem);
	ret = get_user_pages(current, current->mm, uaddr, pagecount,
			     1, 0, pages, NULL);
	up_read(&current->mm->mmap_sem);

	if (ret < 0)
		goto out_free;

697 698 699 700 701 702
	if (ret != pagecount) {
		pagecount = ret;
		ret = -EFAULT;
		goto out_pages;
	}

703 704
	pm.out = (u64 __user *)buf;
	pm.end = (u64 __user *)(buf + count);
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
	pagemap_walk.pmd_entry = pagemap_pte_range;
	pagemap_walk.pte_hole = pagemap_pte_hole;
	pagemap_walk.mm = mm;
	pagemap_walk.private = &pm;

	src = *ppos;
	svpfn = src / PM_ENTRY_BYTES;
	start_vaddr = svpfn << PAGE_SHIFT;
	end_vaddr = TASK_SIZE_OF(task);

	/* watch out for wraparound */
	if (svpfn > TASK_SIZE_OF(task) >> PAGE_SHIFT)
		start_vaddr = end_vaddr;

	/*
	 * The odds are that this will stop walking way
	 * before end_vaddr, because the length of the
	 * user buffer is tracked in "pm", and the walk
	 * will stop when we hit the end of the buffer.
	 */
	ret = walk_page_range(start_vaddr, end_vaddr, &pagemap_walk);
	if (ret == PM_END_OF_BUFFER)
		ret = 0;
	/* don't need mmap_sem for these, but this looks cleaner */
730
	*ppos += (char __user *)pm.out - buf;
731
	if (!ret)
732
		ret = (char __user *)pm.out - buf;
733

734
out_pages:
735 736 737 738 739 740 741 742
	for (; pagecount; pagecount--) {
		page = pages[pagecount-1];
		if (!PageReserved(page))
			SetPageDirty(page);
		page_cache_release(page);
	}
out_free:
	kfree(pages);
743 744
out_mm:
	mmput(mm);
745 746 747 748 749 750 751 752 753 754
out_task:
	put_task_struct(task);
out:
	return ret;
}

const struct file_operations proc_pagemap_operations = {
	.llseek		= mem_lseek, /* borrow this */
	.read		= pagemap_read,
};
755
#endif /* CONFIG_PROC_PAGE_MONITOR */
756

757
#ifdef CONFIG_NUMA
758
extern int show_numa_map(struct seq_file *m, void *v);
759

760
static const struct seq_operations proc_pid_numa_maps_op = {
761 762 763
        .start  = m_start,
        .next   = m_next,
        .stop   = m_stop,
A
Alexey Dobriyan 已提交
764
        .show   = show_numa_map,
765
};
766 767 768 769 770 771

static int numa_maps_open(struct inode *inode, struct file *file)
{
	return do_maps_open(inode, file, &proc_pid_numa_maps_op);
}

772
const struct file_operations proc_numa_maps_operations = {
773 774 775
	.open		= numa_maps_open,
	.read		= seq_read,
	.llseek		= seq_lseek,
776
	.release	= seq_release_private,
777
};
778
#endif