base.c 85.7 KB
Newer Older
1
// SPDX-License-Identifier: GPL-2.0
L
Linus Torvalds 已提交
2 3 4 5 6 7 8 9 10 11 12 13 14
/*
 *  linux/fs/proc/base.c
 *
 *  Copyright (C) 1991, 1992 Linus Torvalds
 *
 *  proc base directory handling functions
 *
 *  1999, Al Viro. Rewritten. Now it covers the whole per-process part.
 *  Instead of using magical inumbers to determine the kind of object
 *  we allocate and fill in-core inodes upon lookup. They don't even
 *  go into icache. We cache the reference to task_struct upon lookup too.
 *  Eventually it should become a filesystem in its own. We don't use the
 *  rest of procfs anymore.
M
Mauricio Lin 已提交
15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48
 *
 *
 *  Changelog:
 *  17-Jan-2005
 *  Allan Bezerra
 *  Bruna Moreira <bruna.moreira@indt.org.br>
 *  Edjard Mota <edjard.mota@indt.org.br>
 *  Ilias Biris <ilias.biris@indt.org.br>
 *  Mauricio Lin <mauricio.lin@indt.org.br>
 *
 *  Embedded Linux Lab - 10LE Instituto Nokia de Tecnologia - INdT
 *
 *  A new process specific entry (smaps) included in /proc. It shows the
 *  size of rss for each memory area. The maps entry lacks information
 *  about physical memory size (rss) for each mapped file, i.e.,
 *  rss information for executables and library files.
 *  This additional information is useful for any tools that need to know
 *  about physical memory consumption for a process specific library.
 *
 *  Changelog:
 *  21-Feb-2005
 *  Embedded Linux Lab - 10LE Instituto Nokia de Tecnologia - INdT
 *  Pud inclusion in the page table walking.
 *
 *  ChangeLog:
 *  10-Mar-2005
 *  10LE Instituto Nokia de Tecnologia - INdT:
 *  A better way to walks through the page table as suggested by Hugh Dickins.
 *
 *  Simo Piiroinen <simo.piiroinen@nokia.com>:
 *  Smaps information related to shared, private, clean and dirty pages.
 *
 *  Paul Mundt <paul.mundt@nokia.com>:
 *  Overall revision about smaps.
L
Linus Torvalds 已提交
49 50
 */

51
#include <linux/uaccess.h>
L
Linus Torvalds 已提交
52 53 54 55 56

#include <linux/errno.h>
#include <linux/time.h>
#include <linux/proc_fs.h>
#include <linux/stat.h>
57
#include <linux/task_io_accounting_ops.h>
L
Linus Torvalds 已提交
58
#include <linux/init.h>
59
#include <linux/capability.h>
L
Linus Torvalds 已提交
60
#include <linux/file.h>
A
Al Viro 已提交
61
#include <linux/fdtable.h>
L
Linus Torvalds 已提交
62 63 64
#include <linux/string.h>
#include <linux/seq_file.h>
#include <linux/namei.h>
65
#include <linux/mnt_namespace.h>
L
Linus Torvalds 已提交
66
#include <linux/mm.h>
D
David Rientjes 已提交
67
#include <linux/swap.h>
68
#include <linux/rcupdate.h>
L
Linus Torvalds 已提交
69
#include <linux/kallsyms.h>
K
Ken Chen 已提交
70
#include <linux/stacktrace.h>
71
#include <linux/resource.h>
K
Kees Cook 已提交
72
#include <linux/module.h>
L
Linus Torvalds 已提交
73 74 75
#include <linux/mount.h>
#include <linux/security.h>
#include <linux/ptrace.h>
76
#include <linux/tracehook.h>
A
Andrew Morton 已提交
77
#include <linux/printk.h>
78
#include <linux/cgroup.h>
L
Linus Torvalds 已提交
79 80
#include <linux/cpuset.h>
#include <linux/audit.h>
A
Al Viro 已提交
81
#include <linux/poll.h>
82
#include <linux/nsproxy.h>
83
#include <linux/oom.h>
84
#include <linux/elf.h>
85
#include <linux/pid_namespace.h>
86
#include <linux/user_namespace.h>
87
#include <linux/fs_struct.h>
88
#include <linux/slab.h>
89
#include <linux/sched/autogroup.h>
90
#include <linux/sched/mm.h>
91
#include <linux/sched/coredump.h>
92
#include <linux/sched/debug.h>
93
#include <linux/sched/stat.h>
94
#include <linux/flex_array.h>
95
#include <linux/posix-timers.h>
96 97 98
#ifdef CONFIG_HARDWALL
#include <asm/hardwall.h>
#endif
99
#include <trace/events/oom.h>
L
Linus Torvalds 已提交
100
#include "internal.h"
101
#include "fd.h"
L
Linus Torvalds 已提交
102

103 104
#include "../../lib/kstrtox.h"

105 106 107 108 109 110 111 112 113 114
/* NOTE:
 *	Implementing inode permission operations in /proc is almost
 *	certainly an error.  Permission checks need to happen during
 *	each system call not at open time.  The reason is that most of
 *	what we wish to check for permissions in /proc varies at runtime.
 *
 *	The classic example of a problem is opening file descriptors
 *	in /proc for a task before it execs a suid executable.
 */

115 116 117
static u8 nlink_tid;
static u8 nlink_tgid;

L
Linus Torvalds 已提交
118
struct pid_entry {
119
	const char *name;
120
	unsigned int len;
A
Al Viro 已提交
121
	umode_t mode;
122
	const struct inode_operations *iop;
123
	const struct file_operations *fop;
124
	union proc_op op;
L
Linus Torvalds 已提交
125 126
};

127
#define NOD(NAME, MODE, IOP, FOP, OP) {			\
128
	.name = (NAME),					\
129
	.len  = sizeof(NAME) - 1,			\
130 131 132 133 134 135
	.mode = MODE,					\
	.iop  = IOP,					\
	.fop  = FOP,					\
	.op   = OP,					\
}

A
Alexey Dobriyan 已提交
136 137 138
#define DIR(NAME, MODE, iops, fops)	\
	NOD(NAME, (S_IFDIR|(MODE)), &iops, &fops, {} )
#define LNK(NAME, get_link)					\
139
	NOD(NAME, (S_IFLNK|S_IRWXUGO),				\
140
		&proc_pid_link_inode_operations, NULL,		\
A
Alexey Dobriyan 已提交
141 142 143 144
		{ .proc_get_link = get_link } )
#define REG(NAME, MODE, fops)				\
	NOD(NAME, (S_IFREG|(MODE)), NULL, &fops, {})
#define ONE(NAME, MODE, show)				\
145 146
	NOD(NAME, (S_IFREG|(MODE)), 			\
		NULL, &proc_single_file_operations,	\
A
Alexey Dobriyan 已提交
147
		{ .proc_show = show } )
L
Linus Torvalds 已提交
148

149 150 151 152
/*
 * Count the number of hardlinks for the pid_entry table, excluding the .
 * and .. links.
 */
153
static unsigned int __init pid_entry_nlink(const struct pid_entry *entries,
154 155 156 157 158
	unsigned int n)
{
	unsigned int i;
	unsigned int count;

159
	count = 2;
160 161 162 163 164 165 166 167
	for (i = 0; i < n; ++i) {
		if (S_ISDIR(entries[i].mode))
			++count;
	}

	return count;
}

168
static int get_task_root(struct task_struct *task, struct path *root)
L
Linus Torvalds 已提交
169
{
H
Hugh Dickins 已提交
170 171
	int result = -ENOENT;

172
	task_lock(task);
173 174
	if (task->fs) {
		get_fs_root(task->fs, root);
H
Hugh Dickins 已提交
175 176
		result = 0;
	}
177
	task_unlock(task);
H
Hugh Dickins 已提交
178
	return result;
179 180
}

181
static int proc_cwd_link(struct dentry *dentry, struct path *path)
182
{
183
	struct task_struct *task = get_proc_task(d_inode(dentry));
184
	int result = -ENOENT;
185 186

	if (task) {
187 188 189 190 191 192
		task_lock(task);
		if (task->fs) {
			get_fs_pwd(task->fs, path);
			result = 0;
		}
		task_unlock(task);
193 194
		put_task_struct(task);
	}
L
Linus Torvalds 已提交
195 196 197
	return result;
}

198
static int proc_root_link(struct dentry *dentry, struct path *path)
L
Linus Torvalds 已提交
199
{
200
	struct task_struct *task = get_proc_task(d_inode(dentry));
L
Linus Torvalds 已提交
201
	int result = -ENOENT;
202 203

	if (task) {
204
		result = get_task_root(task, path);
205 206
		put_task_struct(task);
	}
L
Linus Torvalds 已提交
207 208 209
	return result;
}

210 211
static ssize_t proc_pid_cmdline_read(struct file *file, char __user *buf,
				     size_t _count, loff_t *pos)
L
Linus Torvalds 已提交
212
{
213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237
	struct task_struct *tsk;
	struct mm_struct *mm;
	char *page;
	unsigned long count = _count;
	unsigned long arg_start, arg_end, env_start, env_end;
	unsigned long len1, len2, len;
	unsigned long p;
	char c;
	ssize_t rv;

	BUG_ON(*pos < 0);

	tsk = get_proc_task(file_inode(file));
	if (!tsk)
		return -ESRCH;
	mm = get_task_mm(tsk);
	put_task_struct(tsk);
	if (!mm)
		return 0;
	/* Check if process spawned far enough to have cmdline. */
	if (!mm->env_end) {
		rv = 0;
		goto out_mmput;
	}

238
	page = (char *)__get_free_page(GFP_KERNEL);
239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256
	if (!page) {
		rv = -ENOMEM;
		goto out_mmput;
	}

	down_read(&mm->mmap_sem);
	arg_start = mm->arg_start;
	arg_end = mm->arg_end;
	env_start = mm->env_start;
	env_end = mm->env_end;
	up_read(&mm->mmap_sem);

	BUG_ON(arg_start > arg_end);
	BUG_ON(env_start > env_end);

	len1 = arg_end - arg_start;
	len2 = env_end - env_start;

257 258 259 260 261
	/* Empty ARGV. */
	if (len1 == 0) {
		rv = 0;
		goto out_free_page;
	}
262
	/*
263 264
	 * Inherently racy -- command line shares address space
	 * with code and data.
265
	 */
266
	rv = access_remote_vm(mm, arg_end - 1, &c, 1, 0);
267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283
	if (rv <= 0)
		goto out_free_page;

	rv = 0;

	if (c == '\0') {
		/* Command line (set of strings) occupies whole ARGV. */
		if (len1 <= *pos)
			goto out_free_page;

		p = arg_start + *pos;
		len = len1 - *pos;
		while (count > 0 && len > 0) {
			unsigned int _count;
			int nr_read;

			_count = min3(count, len, PAGE_SIZE);
284
			nr_read = access_remote_vm(mm, p, page, _count, 0);
285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305
			if (nr_read < 0)
				rv = nr_read;
			if (nr_read <= 0)
				goto out_free_page;

			if (copy_to_user(buf, page, nr_read)) {
				rv = -EFAULT;
				goto out_free_page;
			}

			p	+= nr_read;
			len	-= nr_read;
			buf	+= nr_read;
			count	-= nr_read;
			rv	+= nr_read;
		}
	} else {
		/*
		 * Command line (1 string) occupies ARGV and
		 * extends into ENVP.
		 */
306 307 308 309 310 311 312 313 314 315 316 317 318 319
		struct {
			unsigned long p;
			unsigned long len;
		} cmdline[2] = {
			{ .p = arg_start, .len = len1 },
			{ .p = env_start, .len = len2 },
		};
		loff_t pos1 = *pos;
		unsigned int i;

		i = 0;
		while (i < 2 && pos1 >= cmdline[i].len) {
			pos1 -= cmdline[i].len;
			i++;
320
		}
321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359
		while (i < 2) {
			p = cmdline[i].p + pos1;
			len = cmdline[i].len - pos1;
			while (count > 0 && len > 0) {
				unsigned int _count, l;
				int nr_read;
				bool final;

				_count = min3(count, len, PAGE_SIZE);
				nr_read = access_remote_vm(mm, p, page, _count, 0);
				if (nr_read < 0)
					rv = nr_read;
				if (nr_read <= 0)
					goto out_free_page;

				/*
				 * Command line can be shorter than whole ARGV
				 * even if last "marker" byte says it is not.
				 */
				final = false;
				l = strnlen(page, nr_read);
				if (l < nr_read) {
					nr_read = l;
					final = true;
				}

				if (copy_to_user(buf, page, nr_read)) {
					rv = -EFAULT;
					goto out_free_page;
				}

				p	+= nr_read;
				len	-= nr_read;
				buf	+= nr_read;
				count	-= nr_read;
				rv	+= nr_read;

				if (final)
					goto out_free_page;
360 361
			}

362 363 364
			/* Only first chunk can be read partially. */
			pos1 = 0;
			i++;
365 366 367 368 369 370 371 372 373 374
		}
	}

out_free_page:
	free_page((unsigned long)page);
out_mmput:
	mmput(mm);
	if (rv > 0)
		*pos += rv;
	return rv;
L
Linus Torvalds 已提交
375 376
}

377 378 379 380 381
static const struct file_operations proc_pid_cmdline_ops = {
	.read	= proc_pid_cmdline_read,
	.llseek	= generic_file_llseek,
};

L
Linus Torvalds 已提交
382 383 384 385 386
#ifdef CONFIG_KALLSYMS
/*
 * Provides a wchan file via kallsyms in a proper one-value-per-file format.
 * Returns the resolved symbol.  If that fails, simply return the address.
 */
387 388
static int proc_pid_wchan(struct seq_file *m, struct pid_namespace *ns,
			  struct pid *pid, struct task_struct *task)
L
Linus Torvalds 已提交
389
{
A
Alexey Dobriyan 已提交
390
	unsigned long wchan;
391
	char symname[KSYM_NAME_LEN];
L
Linus Torvalds 已提交
392 393 394

	wchan = get_wchan(task);

395 396
	if (wchan && ptrace_may_access(task, PTRACE_MODE_READ_FSCREDS)
			&& !lookup_symbol_name(wchan, symname))
397
		seq_printf(m, "%s", symname);
398
	else
399
		seq_putc(m, '0');
400 401

	return 0;
L
Linus Torvalds 已提交
402 403 404
}
#endif /* CONFIG_KALLSYMS */

405 406 407 408 409
static int lock_trace(struct task_struct *task)
{
	int err = mutex_lock_killable(&task->signal->cred_guard_mutex);
	if (err)
		return err;
410
	if (!ptrace_may_access(task, PTRACE_MODE_ATTACH_FSCREDS)) {
411 412 413 414 415 416 417 418 419 420 421
		mutex_unlock(&task->signal->cred_guard_mutex);
		return -EPERM;
	}
	return 0;
}

static void unlock_trace(struct task_struct *task)
{
	mutex_unlock(&task->signal->cred_guard_mutex);
}

K
Ken Chen 已提交
422 423 424 425 426 427 428 429 430
#ifdef CONFIG_STACKTRACE

#define MAX_STACK_TRACE_DEPTH	64

static int proc_pid_stack(struct seq_file *m, struct pid_namespace *ns,
			  struct pid *pid, struct task_struct *task)
{
	struct stack_trace trace;
	unsigned long *entries;
431
	int err;
K
Ken Chen 已提交
432 433 434 435 436 437 438 439 440 441 442
	int i;

	entries = kmalloc(MAX_STACK_TRACE_DEPTH * sizeof(*entries), GFP_KERNEL);
	if (!entries)
		return -ENOMEM;

	trace.nr_entries	= 0;
	trace.max_entries	= MAX_STACK_TRACE_DEPTH;
	trace.entries		= entries;
	trace.skip		= 0;

443 444 445 446 447
	err = lock_trace(task);
	if (!err) {
		save_stack_trace_tsk(task, &trace);

		for (i = 0; i < trace.nr_entries; i++) {
448
			seq_printf(m, "[<0>] %pB\n", (void *)entries[i]);
449 450
		}
		unlock_trace(task);
K
Ken Chen 已提交
451 452 453
	}
	kfree(entries);

454
	return err;
K
Ken Chen 已提交
455 456 457
}
#endif

458
#ifdef CONFIG_SCHED_INFO
L
Linus Torvalds 已提交
459 460 461
/*
 * Provides /proc/PID/schedstat
 */
462 463
static int proc_pid_schedstat(struct seq_file *m, struct pid_namespace *ns,
			      struct pid *pid, struct task_struct *task)
L
Linus Torvalds 已提交
464
{
465 466 467 468
	if (unlikely(!sched_info_on()))
		seq_printf(m, "0 0 0\n");
	else
		seq_printf(m, "%llu %llu %lu\n",
469 470 471 472 473
		   (unsigned long long)task->se.sum_exec_runtime,
		   (unsigned long long)task->sched_info.run_delay,
		   task->sched_info.pcount);

	return 0;
L
Linus Torvalds 已提交
474 475 476
}
#endif

A
Arjan van de Ven 已提交
477 478 479 480
#ifdef CONFIG_LATENCYTOP
static int lstats_show_proc(struct seq_file *m, void *v)
{
	int i;
481 482
	struct inode *inode = m->private;
	struct task_struct *task = get_proc_task(inode);
A
Arjan van de Ven 已提交
483

484 485 486
	if (!task)
		return -ESRCH;
	seq_puts(m, "Latency Top version : v0.1\n");
A
Arjan van de Ven 已提交
487
	for (i = 0; i < 32; i++) {
488 489
		struct latency_record *lr = &task->latency_record[i];
		if (lr->backtrace[0]) {
A
Arjan van de Ven 已提交
490
			int q;
491 492
			seq_printf(m, "%i %li %li",
				   lr->count, lr->time, lr->max);
A
Arjan van de Ven 已提交
493
			for (q = 0; q < LT_BACKTRACEDEPTH; q++) {
494 495
				unsigned long bt = lr->backtrace[q];
				if (!bt)
A
Arjan van de Ven 已提交
496
					break;
497
				if (bt == ULONG_MAX)
A
Arjan van de Ven 已提交
498
					break;
499
				seq_printf(m, " %ps", (void *)bt);
A
Arjan van de Ven 已提交
500
			}
501
			seq_putc(m, '\n');
A
Arjan van de Ven 已提交
502 503 504
		}

	}
505
	put_task_struct(task);
A
Arjan van de Ven 已提交
506 507 508 509 510
	return 0;
}

static int lstats_open(struct inode *inode, struct file *file)
{
511
	return single_open(file, lstats_show_proc, inode);
512 513
}

A
Arjan van de Ven 已提交
514 515 516
static ssize_t lstats_write(struct file *file, const char __user *buf,
			    size_t count, loff_t *offs)
{
A
Al Viro 已提交
517
	struct task_struct *task = get_proc_task(file_inode(file));
A
Arjan van de Ven 已提交
518

519 520
	if (!task)
		return -ESRCH;
A
Arjan van de Ven 已提交
521
	clear_all_latency_tracing(task);
522
	put_task_struct(task);
A
Arjan van de Ven 已提交
523 524 525 526 527 528 529 530 531

	return count;
}

static const struct file_operations proc_lstats_operations = {
	.open		= lstats_open,
	.read		= seq_read,
	.write		= lstats_write,
	.llseek		= seq_lseek,
532
	.release	= single_release,
A
Arjan van de Ven 已提交
533 534 535 536
};

#endif

537 538
static int proc_oom_score(struct seq_file *m, struct pid_namespace *ns,
			  struct pid *pid, struct task_struct *task)
L
Linus Torvalds 已提交
539
{
540
	unsigned long totalpages = totalram_pages + total_swap_pages;
541
	unsigned long points = 0;
L
Linus Torvalds 已提交
542

543 544
	points = oom_badness(task, NULL, NULL, totalpages) *
					1000 / totalpages;
545 546 547
	seq_printf(m, "%lu\n", points);

	return 0;
L
Linus Torvalds 已提交
548 549
}

550
struct limit_names {
551 552
	const char *name;
	const char *unit;
553 554 555
};

static const struct limit_names lnames[RLIM_NLIMITS] = {
556
	[RLIMIT_CPU] = {"Max cpu time", "seconds"},
557 558 559 560 561 562 563 564 565 566 567 568 569 570
	[RLIMIT_FSIZE] = {"Max file size", "bytes"},
	[RLIMIT_DATA] = {"Max data size", "bytes"},
	[RLIMIT_STACK] = {"Max stack size", "bytes"},
	[RLIMIT_CORE] = {"Max core file size", "bytes"},
	[RLIMIT_RSS] = {"Max resident set", "bytes"},
	[RLIMIT_NPROC] = {"Max processes", "processes"},
	[RLIMIT_NOFILE] = {"Max open files", "files"},
	[RLIMIT_MEMLOCK] = {"Max locked memory", "bytes"},
	[RLIMIT_AS] = {"Max address space", "bytes"},
	[RLIMIT_LOCKS] = {"Max file locks", "locks"},
	[RLIMIT_SIGPENDING] = {"Max pending signals", "signals"},
	[RLIMIT_MSGQUEUE] = {"Max msgqueue size", "bytes"},
	[RLIMIT_NICE] = {"Max nice priority", NULL},
	[RLIMIT_RTPRIO] = {"Max realtime priority", NULL},
571
	[RLIMIT_RTTIME] = {"Max realtime timeout", "us"},
572 573 574
};

/* Display limits for a process */
575 576
static int proc_pid_limits(struct seq_file *m, struct pid_namespace *ns,
			   struct pid *pid, struct task_struct *task)
577 578 579 580 581 582
{
	unsigned int i;
	unsigned long flags;

	struct rlimit rlim[RLIM_NLIMITS];

583
	if (!lock_task_sighand(task, &flags))
584 585 586 587 588 589 590
		return 0;
	memcpy(rlim, task->signal->rlim, sizeof(struct rlimit) * RLIM_NLIMITS);
	unlock_task_sighand(task, &flags);

	/*
	 * print the file header
	 */
591
       seq_printf(m, "%-25s %-20s %-20s %-10s\n",
592
		  "Limit", "Soft Limit", "Hard Limit", "Units");
593 594 595

	for (i = 0; i < RLIM_NLIMITS; i++) {
		if (rlim[i].rlim_cur == RLIM_INFINITY)
596
			seq_printf(m, "%-25s %-20s ",
597
				   lnames[i].name, "unlimited");
598
		else
599
			seq_printf(m, "%-25s %-20lu ",
600
				   lnames[i].name, rlim[i].rlim_cur);
601 602

		if (rlim[i].rlim_max == RLIM_INFINITY)
603
			seq_printf(m, "%-20s ", "unlimited");
604
		else
605
			seq_printf(m, "%-20lu ", rlim[i].rlim_max);
606 607

		if (lnames[i].unit)
608
			seq_printf(m, "%-10s\n", lnames[i].unit);
609
		else
610
			seq_putc(m, '\n');
611 612
	}

613
	return 0;
614 615
}

R
Roland McGrath 已提交
616
#ifdef CONFIG_HAVE_ARCH_TRACEHOOK
617 618
static int proc_pid_syscall(struct seq_file *m, struct pid_namespace *ns,
			    struct pid *pid, struct task_struct *task)
R
Roland McGrath 已提交
619 620 621
{
	long nr;
	unsigned long args[6], sp, pc;
622 623 624
	int res;

	res = lock_trace(task);
625 626
	if (res)
		return res;
R
Roland McGrath 已提交
627 628

	if (task_current_syscall(task, &nr, args, 6, &sp, &pc))
629
		seq_puts(m, "running\n");
630
	else if (nr < 0)
631
		seq_printf(m, "%ld 0x%lx 0x%lx\n", nr, sp, pc);
632
	else
633
		seq_printf(m,
R
Roland McGrath 已提交
634 635 636 637
		       "%ld 0x%lx 0x%lx 0x%lx 0x%lx 0x%lx 0x%lx 0x%lx 0x%lx\n",
		       nr,
		       args[0], args[1], args[2], args[3], args[4], args[5],
		       sp, pc);
638
	unlock_trace(task);
639 640

	return 0;
R
Roland McGrath 已提交
641 642 643
}
#endif /* CONFIG_HAVE_ARCH_TRACEHOOK */

L
Linus Torvalds 已提交
644 645 646 647 648
/************************************************************************/
/*                       Here the fs part begins                        */
/************************************************************************/

/* permission checks */
649
static int proc_fd_access_allowed(struct inode *inode)
L
Linus Torvalds 已提交
650
{
651 652
	struct task_struct *task;
	int allowed = 0;
653 654 655
	/* Allow access to a task's file descriptors if it is us or we
	 * may use ptrace attach to the process and find out that
	 * information.
656 657
	 */
	task = get_proc_task(inode);
658
	if (task) {
659
		allowed = ptrace_may_access(task, PTRACE_MODE_READ_FSCREDS);
660
		put_task_struct(task);
661
	}
662
	return allowed;
L
Linus Torvalds 已提交
663 664
}

665
int proc_setattr(struct dentry *dentry, struct iattr *attr)
666 667
{
	int error;
668
	struct inode *inode = d_inode(dentry);
669 670 671 672

	if (attr->ia_valid & ATTR_MODE)
		return -EPERM;

673
	error = setattr_prepare(dentry, attr);
C
Christoph Hellwig 已提交
674 675 676 677 678 679
	if (error)
		return error;

	setattr_copy(inode, attr);
	mark_inode_dirty(inode);
	return 0;
680 681
}

682 683 684 685 686 687 688 689 690 691 692 693
/*
 * May current process learn task's sched/cmdline info (for hide_pid_min=1)
 * or euid/egid (for hide_pid_min=2)?
 */
static bool has_pid_permissions(struct pid_namespace *pid,
				 struct task_struct *task,
				 int hide_pid_min)
{
	if (pid->hide_pid < hide_pid_min)
		return true;
	if (in_group_p(pid->pid_gid))
		return true;
694
	return ptrace_may_access(task, PTRACE_MODE_READ_FSCREDS);
695 696 697 698 699 700 701 702 703 704
}


static int proc_pid_permission(struct inode *inode, int mask)
{
	struct pid_namespace *pid = inode->i_sb->s_fs_info;
	struct task_struct *task;
	bool has_perms;

	task = get_proc_task(inode);
705 706
	if (!task)
		return -ESRCH;
707
	has_perms = has_pid_permissions(pid, task, HIDEPID_NO_ACCESS);
708 709 710
	put_task_struct(task);

	if (!has_perms) {
711
		if (pid->hide_pid == HIDEPID_INVISIBLE) {
712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727
			/*
			 * Let's make getdents(), stat(), and open()
			 * consistent with each other.  If a process
			 * may not stat() a file, it shouldn't be seen
			 * in procfs at all.
			 */
			return -ENOENT;
		}

		return -EPERM;
	}
	return generic_permission(inode, mask);
}



728
static const struct inode_operations proc_def_inode_operations = {
729 730 731
	.setattr	= proc_setattr,
};

732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753
static int proc_single_show(struct seq_file *m, void *v)
{
	struct inode *inode = m->private;
	struct pid_namespace *ns;
	struct pid *pid;
	struct task_struct *task;
	int ret;

	ns = inode->i_sb->s_fs_info;
	pid = proc_pid(inode);
	task = get_pid_task(pid, PIDTYPE_PID);
	if (!task)
		return -ESRCH;

	ret = PROC_I(inode)->op.proc_show(m, ns, pid, task);

	put_task_struct(task);
	return ret;
}

static int proc_single_open(struct inode *inode, struct file *filp)
{
J
Jovi Zhang 已提交
754
	return single_open(filp, proc_single_show, inode);
755 756 757 758 759 760 761 762 763
}

static const struct file_operations proc_single_file_operations = {
	.open		= proc_single_open,
	.read		= seq_read,
	.llseek		= seq_lseek,
	.release	= single_release,
};

O
Oleg Nesterov 已提交
764 765

struct mm_struct *proc_mem_open(struct inode *inode, unsigned int mode)
L
Linus Torvalds 已提交
766
{
O
Oleg Nesterov 已提交
767 768
	struct task_struct *task = get_proc_task(inode);
	struct mm_struct *mm = ERR_PTR(-ESRCH);
769

O
Oleg Nesterov 已提交
770
	if (task) {
771
		mm = mm_access(task, mode | PTRACE_MODE_FSCREDS);
O
Oleg Nesterov 已提交
772
		put_task_struct(task);
773

O
Oleg Nesterov 已提交
774 775
		if (!IS_ERR_OR_NULL(mm)) {
			/* ensure this mm_struct can't be freed */
V
Vegard Nossum 已提交
776
			mmgrab(mm);
O
Oleg Nesterov 已提交
777 778 779 780 781 782 783 784 785 786 787
			/* but do not pin its memory */
			mmput(mm);
		}
	}

	return mm;
}

static int __mem_open(struct inode *inode, struct file *file, unsigned int mode)
{
	struct mm_struct *mm = proc_mem_open(inode, mode);
788 789 790 791 792

	if (IS_ERR(mm))
		return PTR_ERR(mm);

	file->private_data = mm;
L
Linus Torvalds 已提交
793 794 795
	return 0;
}

796 797
static int mem_open(struct inode *inode, struct file *file)
{
798 799 800 801 802 803
	int ret = __mem_open(inode, file, PTRACE_MODE_ATTACH);

	/* OK to pass negative loff_t, we can catch out-of-range */
	file->f_mode |= FMODE_UNSIGNED_OFFSET;

	return ret;
804 805
}

806 807
static ssize_t mem_rw(struct file *file, char __user *buf,
			size_t count, loff_t *ppos, int write)
L
Linus Torvalds 已提交
808
{
809
	struct mm_struct *mm = file->private_data;
810 811
	unsigned long addr = *ppos;
	ssize_t copied;
L
Linus Torvalds 已提交
812
	char *page;
813
	unsigned int flags;
L
Linus Torvalds 已提交
814

815 816
	if (!mm)
		return 0;
817

818
	page = (char *)__get_free_page(GFP_KERNEL);
819
	if (!page)
820
		return -ENOMEM;
L
Linus Torvalds 已提交
821

822
	copied = 0;
V
Vegard Nossum 已提交
823
	if (!mmget_not_zero(mm))
824 825
		goto free;

L
Linus Torvalds 已提交
826
	flags = FOLL_FORCE | (write ? FOLL_WRITE : 0);
827

L
Linus Torvalds 已提交
828
	while (count > 0) {
829
		int this_len = min_t(int, count, PAGE_SIZE);
L
Linus Torvalds 已提交
830

831
		if (write && copy_from_user(page, buf, this_len)) {
L
Linus Torvalds 已提交
832 833 834
			copied = -EFAULT;
			break;
		}
835

836
		this_len = access_remote_vm(mm, addr, page, this_len, flags);
837
		if (!this_len) {
L
Linus Torvalds 已提交
838 839 840 841
			if (!copied)
				copied = -EIO;
			break;
		}
842 843 844 845 846 847 848 849 850 851

		if (!write && copy_to_user(buf, page, this_len)) {
			copied = -EFAULT;
			break;
		}

		buf += this_len;
		addr += this_len;
		copied += this_len;
		count -= this_len;
L
Linus Torvalds 已提交
852
	}
853
	*ppos = addr;
854

855 856
	mmput(mm);
free:
857
	free_page((unsigned long) page);
L
Linus Torvalds 已提交
858 859 860
	return copied;
}

861 862 863 864 865 866 867 868 869 870 871 872
static ssize_t mem_read(struct file *file, char __user *buf,
			size_t count, loff_t *ppos)
{
	return mem_rw(file, buf, count, ppos, 0);
}

static ssize_t mem_write(struct file *file, const char __user *buf,
			 size_t count, loff_t *ppos)
{
	return mem_rw(file, (char __user*)buf, count, ppos, 1);
}

873
loff_t mem_lseek(struct file *file, loff_t offset, int orig)
L
Linus Torvalds 已提交
874 875 876 877 878 879 880 881 882 883 884 885 886 887 888
{
	switch (orig) {
	case 0:
		file->f_pos = offset;
		break;
	case 1:
		file->f_pos += offset;
		break;
	default:
		return -EINVAL;
	}
	force_successful_syscall_return();
	return file->f_pos;
}

889 890 891
static int mem_release(struct inode *inode, struct file *file)
{
	struct mm_struct *mm = file->private_data;
892
	if (mm)
893
		mmdrop(mm);
894 895 896
	return 0;
}

897
static const struct file_operations proc_mem_operations = {
L
Linus Torvalds 已提交
898 899 900 901
	.llseek		= mem_lseek,
	.read		= mem_read,
	.write		= mem_write,
	.open		= mem_open,
902
	.release	= mem_release,
L
Linus Torvalds 已提交
903 904
};

905 906 907 908 909
static int environ_open(struct inode *inode, struct file *file)
{
	return __mem_open(inode, file, PTRACE_MODE_READ);
}

910 911 912 913 914
static ssize_t environ_read(struct file *file, char __user *buf,
			size_t count, loff_t *ppos)
{
	char *page;
	unsigned long src = *ppos;
915 916
	int ret = 0;
	struct mm_struct *mm = file->private_data;
917
	unsigned long env_start, env_end;
918

919 920
	/* Ensure the process spawned far enough to have an environment. */
	if (!mm || !mm->env_end)
921
		return 0;
922

923
	page = (char *)__get_free_page(GFP_KERNEL);
924
	if (!page)
925
		return -ENOMEM;
926

A
Al Viro 已提交
927
	ret = 0;
V
Vegard Nossum 已提交
928
	if (!mmget_not_zero(mm))
929
		goto free;
930 931 932 933 934 935

	down_read(&mm->mmap_sem);
	env_start = mm->env_start;
	env_end = mm->env_end;
	up_read(&mm->mmap_sem);

936
	while (count > 0) {
937 938
		size_t this_len, max_len;
		int retval;
939

940
		if (src >= (env_end - env_start))
941 942
			break;

943
		this_len = env_end - (env_start + src);
944 945 946

		max_len = min_t(size_t, PAGE_SIZE, count);
		this_len = min(max_len, this_len);
947

948
		retval = access_remote_vm(mm, (env_start + src), page, this_len, 0);
949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966

		if (retval <= 0) {
			ret = retval;
			break;
		}

		if (copy_to_user(buf, page, retval)) {
			ret = -EFAULT;
			break;
		}

		ret += retval;
		src += retval;
		buf += retval;
		count -= retval;
	}
	*ppos = src;
	mmput(mm);
967 968

free:
969 970 971 972 973
	free_page((unsigned long) page);
	return ret;
}

static const struct file_operations proc_environ_operations = {
974
	.open		= environ_open,
975
	.read		= environ_read,
976
	.llseek		= generic_file_llseek,
977
	.release	= mem_release,
978 979
};

A
Al Viro 已提交
980 981 982 983 984 985 986 987 988 989
static int auxv_open(struct inode *inode, struct file *file)
{
	return __mem_open(inode, file, PTRACE_MODE_READ_FSCREDS);
}

static ssize_t auxv_read(struct file *file, char __user *buf,
			size_t count, loff_t *ppos)
{
	struct mm_struct *mm = file->private_data;
	unsigned int nwords = 0;
990 991 992

	if (!mm)
		return 0;
A
Al Viro 已提交
993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006
	do {
		nwords += 2;
	} while (mm->saved_auxv[nwords - 2] != 0); /* AT_NULL */
	return simple_read_from_buffer(buf, count, ppos, mm->saved_auxv,
				       nwords * sizeof(mm->saved_auxv[0]));
}

static const struct file_operations proc_auxv_operations = {
	.open		= auxv_open,
	.read		= auxv_read,
	.llseek		= generic_file_llseek,
	.release	= mem_release,
};

1007 1008 1009
static ssize_t oom_adj_read(struct file *file, char __user *buf, size_t count,
			    loff_t *ppos)
{
A
Al Viro 已提交
1010
	struct task_struct *task = get_proc_task(file_inode(file));
1011 1012 1013 1014 1015 1016
	char buffer[PROC_NUMBUF];
	int oom_adj = OOM_ADJUST_MIN;
	size_t len;

	if (!task)
		return -ESRCH;
1017 1018 1019 1020 1021
	if (task->signal->oom_score_adj == OOM_SCORE_ADJ_MAX)
		oom_adj = OOM_ADJUST_MAX;
	else
		oom_adj = (task->signal->oom_score_adj * -OOM_DISABLE) /
			  OOM_SCORE_ADJ_MAX;
1022 1023 1024 1025 1026
	put_task_struct(task);
	len = snprintf(buffer, sizeof(buffer), "%d\n", oom_adj);
	return simple_read_from_buffer(buf, count, ppos, buffer, len);
}

1027 1028 1029
static int __set_oom_adj(struct file *file, int oom_adj, bool legacy)
{
	static DEFINE_MUTEX(oom_adj_mutex);
1030
	struct mm_struct *mm = NULL;
1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059
	struct task_struct *task;
	int err = 0;

	task = get_proc_task(file_inode(file));
	if (!task)
		return -ESRCH;

	mutex_lock(&oom_adj_mutex);
	if (legacy) {
		if (oom_adj < task->signal->oom_score_adj &&
				!capable(CAP_SYS_RESOURCE)) {
			err = -EACCES;
			goto err_unlock;
		}
		/*
		 * /proc/pid/oom_adj is provided for legacy purposes, ask users to use
		 * /proc/pid/oom_score_adj instead.
		 */
		pr_warn_once("%s (%d): /proc/%d/oom_adj is deprecated, please use /proc/%d/oom_score_adj instead.\n",
			  current->comm, task_pid_nr(current), task_pid_nr(task),
			  task_pid_nr(task));
	} else {
		if ((short)oom_adj < task->signal->oom_score_adj_min &&
				!capable(CAP_SYS_RESOURCE)) {
			err = -EACCES;
			goto err_unlock;
		}
	}

1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070
	/*
	 * Make sure we will check other processes sharing the mm if this is
	 * not vfrok which wants its own oom_score_adj.
	 * pin the mm so it doesn't go away and get reused after task_unlock
	 */
	if (!task->vfork_done) {
		struct task_struct *p = find_lock_task_mm(task);

		if (p) {
			if (atomic_read(&p->mm->mm_users) > 1) {
				mm = p->mm;
V
Vegard Nossum 已提交
1071
				mmgrab(mm);
1072 1073 1074 1075 1076
			}
			task_unlock(p);
		}
	}

1077 1078 1079 1080
	task->signal->oom_score_adj = oom_adj;
	if (!legacy && has_capability_noaudit(current, CAP_SYS_RESOURCE))
		task->signal->oom_score_adj_min = (short)oom_adj;
	trace_oom_score_adj_update(task);
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 1107 1108

	if (mm) {
		struct task_struct *p;

		rcu_read_lock();
		for_each_process(p) {
			if (same_thread_group(task, p))
				continue;

			/* do not touch kernel threads or the global init */
			if (p->flags & PF_KTHREAD || is_global_init(p))
				continue;

			task_lock(p);
			if (!p->vfork_done && process_shares_mm(p, mm)) {
				pr_info("updating oom_score_adj for %d (%s) from %d to %d because it shares mm with %d (%s). Report if this is unexpected.\n",
						task_pid_nr(p), p->comm,
						p->signal->oom_score_adj, oom_adj,
						task_pid_nr(task), task->comm);
				p->signal->oom_score_adj = oom_adj;
				if (!legacy && has_capability_noaudit(current, CAP_SYS_RESOURCE))
					p->signal->oom_score_adj_min = (short)oom_adj;
			}
			task_unlock(p);
		}
		rcu_read_unlock();
		mmdrop(mm);
	}
1109 1110 1111 1112 1113
err_unlock:
	mutex_unlock(&oom_adj_mutex);
	put_task_struct(task);
	return err;
}
1114

1115 1116 1117 1118 1119 1120 1121 1122 1123 1124
/*
 * /proc/pid/oom_adj exists solely for backwards compatibility with previous
 * kernels.  The effective policy is defined by oom_score_adj, which has a
 * different scale: oom_adj grew exponentially and oom_score_adj grows linearly.
 * Values written to oom_adj are simply mapped linearly to oom_score_adj.
 * Processes that become oom disabled via oom_adj will still be oom disabled
 * with this implementation.
 *
 * oom_adj cannot be removed since existing userspace binaries use it.
 */
1125 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151 1152 1153 1154 1155 1156 1157
static ssize_t oom_adj_write(struct file *file, const char __user *buf,
			     size_t count, loff_t *ppos)
{
	char buffer[PROC_NUMBUF];
	int oom_adj;
	int err;

	memset(buffer, 0, sizeof(buffer));
	if (count > sizeof(buffer) - 1)
		count = sizeof(buffer) - 1;
	if (copy_from_user(buffer, buf, count)) {
		err = -EFAULT;
		goto out;
	}

	err = kstrtoint(strstrip(buffer), 0, &oom_adj);
	if (err)
		goto out;
	if ((oom_adj < OOM_ADJUST_MIN || oom_adj > OOM_ADJUST_MAX) &&
	     oom_adj != OOM_DISABLE) {
		err = -EINVAL;
		goto out;
	}

	/*
	 * Scale /proc/pid/oom_score_adj appropriately ensuring that a maximum
	 * value is always attainable.
	 */
	if (oom_adj == OOM_ADJUST_MAX)
		oom_adj = OOM_SCORE_ADJ_MAX;
	else
		oom_adj = (oom_adj * OOM_SCORE_ADJ_MAX) / -OOM_DISABLE;

1158
	err = __set_oom_adj(file, oom_adj, true);
1159 1160 1161 1162 1163 1164 1165 1166 1167 1168
out:
	return err < 0 ? err : count;
}

static const struct file_operations proc_oom_adj_operations = {
	.read		= oom_adj_read,
	.write		= oom_adj_write,
	.llseek		= generic_file_llseek,
};

D
David Rientjes 已提交
1169 1170 1171
static ssize_t oom_score_adj_read(struct file *file, char __user *buf,
					size_t count, loff_t *ppos)
{
A
Al Viro 已提交
1172
	struct task_struct *task = get_proc_task(file_inode(file));
D
David Rientjes 已提交
1173
	char buffer[PROC_NUMBUF];
1174
	short oom_score_adj = OOM_SCORE_ADJ_MIN;
D
David Rientjes 已提交
1175 1176 1177 1178
	size_t len;

	if (!task)
		return -ESRCH;
1179
	oom_score_adj = task->signal->oom_score_adj;
D
David Rientjes 已提交
1180
	put_task_struct(task);
1181
	len = snprintf(buffer, sizeof(buffer), "%hd\n", oom_score_adj);
D
David Rientjes 已提交
1182 1183 1184 1185 1186 1187 1188
	return simple_read_from_buffer(buf, count, ppos, buffer, len);
}

static ssize_t oom_score_adj_write(struct file *file, const char __user *buf,
					size_t count, loff_t *ppos)
{
	char buffer[PROC_NUMBUF];
A
Alexey Dobriyan 已提交
1189
	int oom_score_adj;
D
David Rientjes 已提交
1190 1191 1192 1193 1194
	int err;

	memset(buffer, 0, sizeof(buffer));
	if (count > sizeof(buffer) - 1)
		count = sizeof(buffer) - 1;
1195 1196 1197 1198
	if (copy_from_user(buffer, buf, count)) {
		err = -EFAULT;
		goto out;
	}
D
David Rientjes 已提交
1199

A
Alexey Dobriyan 已提交
1200
	err = kstrtoint(strstrip(buffer), 0, &oom_score_adj);
D
David Rientjes 已提交
1201
	if (err)
1202
		goto out;
D
David Rientjes 已提交
1203
	if (oom_score_adj < OOM_SCORE_ADJ_MIN ||
1204 1205 1206 1207
			oom_score_adj > OOM_SCORE_ADJ_MAX) {
		err = -EINVAL;
		goto out;
	}
D
David Rientjes 已提交
1208

1209
	err = __set_oom_adj(file, oom_score_adj, false);
1210 1211
out:
	return err < 0 ? err : count;
D
David Rientjes 已提交
1212 1213 1214 1215 1216
}

static const struct file_operations proc_oom_score_adj_operations = {
	.read		= oom_score_adj_read,
	.write		= oom_score_adj_write,
1217
	.llseek		= default_llseek,
D
David Rientjes 已提交
1218 1219
};

L
Linus Torvalds 已提交
1220
#ifdef CONFIG_AUDITSYSCALL
1221
#define TMPBUFLEN 11
L
Linus Torvalds 已提交
1222 1223 1224
static ssize_t proc_loginuid_read(struct file * file, char __user * buf,
				  size_t count, loff_t *ppos)
{
A
Al Viro 已提交
1225
	struct inode * inode = file_inode(file);
1226
	struct task_struct *task = get_proc_task(inode);
L
Linus Torvalds 已提交
1227 1228 1229
	ssize_t length;
	char tmpbuf[TMPBUFLEN];

1230 1231
	if (!task)
		return -ESRCH;
L
Linus Torvalds 已提交
1232
	length = scnprintf(tmpbuf, TMPBUFLEN, "%u",
1233 1234
			   from_kuid(file->f_cred->user_ns,
				     audit_get_loginuid(task)));
1235
	put_task_struct(task);
L
Linus Torvalds 已提交
1236 1237 1238 1239 1240 1241
	return simple_read_from_buffer(buf, count, ppos, tmpbuf, length);
}

static ssize_t proc_loginuid_write(struct file * file, const char __user * buf,
				   size_t count, loff_t *ppos)
{
A
Al Viro 已提交
1242
	struct inode * inode = file_inode(file);
L
Linus Torvalds 已提交
1243
	uid_t loginuid;
1244
	kuid_t kloginuid;
1245
	int rv;
L
Linus Torvalds 已提交
1246

1247 1248 1249
	rcu_read_lock();
	if (current != pid_task(proc_pid(inode), PIDTYPE_PID)) {
		rcu_read_unlock();
L
Linus Torvalds 已提交
1250
		return -EPERM;
1251 1252
	}
	rcu_read_unlock();
L
Linus Torvalds 已提交
1253 1254 1255 1256 1257 1258

	if (*ppos != 0) {
		/* No partial writes. */
		return -EINVAL;
	}

1259 1260 1261
	rv = kstrtou32_from_user(buf, count, 10, &loginuid);
	if (rv < 0)
		return rv;
1262 1263 1264 1265 1266 1267

	/* is userspace tring to explicitly UNSET the loginuid? */
	if (loginuid == AUDIT_UID_UNSET) {
		kloginuid = INVALID_UID;
	} else {
		kloginuid = make_kuid(file->f_cred->user_ns, loginuid);
1268 1269
		if (!uid_valid(kloginuid))
			return -EINVAL;
1270 1271
	}

1272 1273 1274 1275
	rv = audit_set_loginuid(kloginuid);
	if (rv < 0)
		return rv;
	return count;
L
Linus Torvalds 已提交
1276 1277
}

1278
static const struct file_operations proc_loginuid_operations = {
L
Linus Torvalds 已提交
1279 1280
	.read		= proc_loginuid_read,
	.write		= proc_loginuid_write,
1281
	.llseek		= generic_file_llseek,
L
Linus Torvalds 已提交
1282
};
1283 1284 1285 1286

static ssize_t proc_sessionid_read(struct file * file, char __user * buf,
				  size_t count, loff_t *ppos)
{
A
Al Viro 已提交
1287
	struct inode * inode = file_inode(file);
1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 1299 1300 1301
	struct task_struct *task = get_proc_task(inode);
	ssize_t length;
	char tmpbuf[TMPBUFLEN];

	if (!task)
		return -ESRCH;
	length = scnprintf(tmpbuf, TMPBUFLEN, "%u",
				audit_get_sessionid(task));
	put_task_struct(task);
	return simple_read_from_buffer(buf, count, ppos, tmpbuf, length);
}

static const struct file_operations proc_sessionid_operations = {
	.read		= proc_sessionid_read,
1302
	.llseek		= generic_file_llseek,
1303
};
L
Linus Torvalds 已提交
1304 1305
#endif

1306 1307 1308 1309
#ifdef CONFIG_FAULT_INJECTION
static ssize_t proc_fault_inject_read(struct file * file, char __user * buf,
				      size_t count, loff_t *ppos)
{
A
Al Viro 已提交
1310
	struct task_struct *task = get_proc_task(file_inode(file));
1311 1312 1313 1314 1315 1316 1317 1318 1319 1320
	char buffer[PROC_NUMBUF];
	size_t len;
	int make_it_fail;

	if (!task)
		return -ESRCH;
	make_it_fail = task->make_it_fail;
	put_task_struct(task);

	len = snprintf(buffer, sizeof(buffer), "%i\n", make_it_fail);
1321 1322

	return simple_read_from_buffer(buf, count, ppos, buffer, len);
1323 1324 1325 1326 1327 1328
}

static ssize_t proc_fault_inject_write(struct file * file,
			const char __user * buf, size_t count, loff_t *ppos)
{
	struct task_struct *task;
1329
	char buffer[PROC_NUMBUF];
1330
	int make_it_fail;
1331
	int rv;
1332 1333 1334 1335 1336 1337 1338 1339

	if (!capable(CAP_SYS_RESOURCE))
		return -EPERM;
	memset(buffer, 0, sizeof(buffer));
	if (count > sizeof(buffer) - 1)
		count = sizeof(buffer) - 1;
	if (copy_from_user(buffer, buf, count))
		return -EFAULT;
1340 1341 1342
	rv = kstrtoint(strstrip(buffer), 0, &make_it_fail);
	if (rv < 0)
		return rv;
1343 1344 1345
	if (make_it_fail < 0 || make_it_fail > 1)
		return -EINVAL;

A
Al Viro 已提交
1346
	task = get_proc_task(file_inode(file));
1347 1348 1349 1350
	if (!task)
		return -ESRCH;
	task->make_it_fail = make_it_fail;
	put_task_struct(task);
1351 1352

	return count;
1353 1354
}

1355
static const struct file_operations proc_fault_inject_operations = {
1356 1357
	.read		= proc_fault_inject_read,
	.write		= proc_fault_inject_write,
1358
	.llseek		= generic_file_llseek,
1359
};
1360 1361 1362 1363 1364

static ssize_t proc_fail_nth_write(struct file *file, const char __user *buf,
				   size_t count, loff_t *ppos)
{
	struct task_struct *task;
1365 1366
	int err;
	unsigned int n;
1367

1368 1369 1370 1371
	err = kstrtouint_from_user(buf, count, 0, &n);
	if (err)
		return err;

1372 1373 1374
	task = get_proc_task(file_inode(file));
	if (!task)
		return -ESRCH;
1375
	task->fail_nth = n;
1376
	put_task_struct(task);
1377

1378 1379 1380 1381 1382 1383 1384
	return count;
}

static ssize_t proc_fail_nth_read(struct file *file, char __user *buf,
				  size_t count, loff_t *ppos)
{
	struct task_struct *task;
1385 1386
	char numbuf[PROC_NUMBUF];
	ssize_t len;
1387 1388 1389 1390

	task = get_proc_task(file_inode(file));
	if (!task)
		return -ESRCH;
1391
	len = snprintf(numbuf, sizeof(numbuf), "%u\n", task->fail_nth);
1392
	len = simple_read_from_buffer(buf, count, ppos, numbuf, len);
1393
	put_task_struct(task);
1394 1395

	return len;
1396 1397 1398 1399 1400 1401
}

static const struct file_operations proc_fail_nth_operations = {
	.read		= proc_fail_nth_read,
	.write		= proc_fail_nth_write,
};
1402 1403
#endif

A
Arjan van de Ven 已提交
1404

I
Ingo Molnar 已提交
1405 1406 1407 1408 1409 1410 1411
#ifdef CONFIG_SCHED_DEBUG
/*
 * Print out various scheduling related per-task fields:
 */
static int sched_show(struct seq_file *m, void *v)
{
	struct inode *inode = m->private;
1412
	struct pid_namespace *ns = inode->i_sb->s_fs_info;
I
Ingo Molnar 已提交
1413 1414 1415 1416 1417
	struct task_struct *p;

	p = get_proc_task(inode);
	if (!p)
		return -ESRCH;
1418
	proc_sched_show_task(p, ns, m);
I
Ingo Molnar 已提交
1419 1420 1421 1422 1423 1424 1425 1426 1427 1428

	put_task_struct(p);

	return 0;
}

static ssize_t
sched_write(struct file *file, const char __user *buf,
	    size_t count, loff_t *offset)
{
A
Al Viro 已提交
1429
	struct inode *inode = file_inode(file);
I
Ingo Molnar 已提交
1430 1431 1432 1433 1434 1435 1436 1437 1438 1439 1440 1441 1442 1443
	struct task_struct *p;

	p = get_proc_task(inode);
	if (!p)
		return -ESRCH;
	proc_sched_set_task(p);

	put_task_struct(p);

	return count;
}

static int sched_open(struct inode *inode, struct file *filp)
{
J
Jovi Zhang 已提交
1444
	return single_open(filp, sched_show, inode);
I
Ingo Molnar 已提交
1445 1446 1447 1448 1449 1450 1451
}

static const struct file_operations proc_pid_sched_operations = {
	.open		= sched_open,
	.read		= seq_read,
	.write		= sched_write,
	.llseek		= seq_lseek,
1452
	.release	= single_release,
I
Ingo Molnar 已提交
1453 1454 1455 1456
};

#endif

1457 1458 1459 1460 1461 1462 1463 1464 1465 1466 1467 1468 1469 1470 1471 1472 1473 1474 1475 1476 1477 1478 1479
#ifdef CONFIG_SCHED_AUTOGROUP
/*
 * Print out autogroup related information:
 */
static int sched_autogroup_show(struct seq_file *m, void *v)
{
	struct inode *inode = m->private;
	struct task_struct *p;

	p = get_proc_task(inode);
	if (!p)
		return -ESRCH;
	proc_sched_autogroup_show_task(p, m);

	put_task_struct(p);

	return 0;
}

static ssize_t
sched_autogroup_write(struct file *file, const char __user *buf,
	    size_t count, loff_t *offset)
{
A
Al Viro 已提交
1480
	struct inode *inode = file_inode(file);
1481 1482
	struct task_struct *p;
	char buffer[PROC_NUMBUF];
A
Alexey Dobriyan 已提交
1483
	int nice;
1484 1485 1486 1487 1488 1489 1490 1491
	int err;

	memset(buffer, 0, sizeof(buffer));
	if (count > sizeof(buffer) - 1)
		count = sizeof(buffer) - 1;
	if (copy_from_user(buffer, buf, count))
		return -EFAULT;

A
Alexey Dobriyan 已提交
1492 1493 1494
	err = kstrtoint(strstrip(buffer), 0, &nice);
	if (err < 0)
		return err;
1495 1496 1497 1498 1499

	p = get_proc_task(inode);
	if (!p)
		return -ESRCH;

1500
	err = proc_sched_autogroup_set_nice(p, nice);
1501 1502 1503 1504 1505 1506 1507 1508 1509 1510 1511 1512 1513 1514 1515 1516 1517 1518 1519 1520 1521 1522 1523 1524 1525 1526 1527 1528 1529 1530 1531
	if (err)
		count = err;

	put_task_struct(p);

	return count;
}

static int sched_autogroup_open(struct inode *inode, struct file *filp)
{
	int ret;

	ret = single_open(filp, sched_autogroup_show, NULL);
	if (!ret) {
		struct seq_file *m = filp->private_data;

		m->private = inode;
	}
	return ret;
}

static const struct file_operations proc_pid_sched_autogroup_operations = {
	.open		= sched_autogroup_open,
	.read		= seq_read,
	.write		= sched_autogroup_write,
	.llseek		= seq_lseek,
	.release	= single_release,
};

#endif /* CONFIG_SCHED_AUTOGROUP */

1532 1533 1534
static ssize_t comm_write(struct file *file, const char __user *buf,
				size_t count, loff_t *offset)
{
A
Al Viro 已提交
1535
	struct inode *inode = file_inode(file);
1536 1537
	struct task_struct *p;
	char buffer[TASK_COMM_LEN];
1538
	const size_t maxlen = sizeof(buffer) - 1;
1539 1540

	memset(buffer, 0, sizeof(buffer));
1541
	if (copy_from_user(buffer, buf, count > maxlen ? maxlen : count))
1542 1543 1544 1545 1546 1547 1548 1549 1550 1551 1552 1553 1554 1555 1556 1557 1558 1559 1560 1561 1562 1563 1564 1565 1566 1567 1568 1569 1570 1571 1572 1573 1574 1575 1576 1577
		return -EFAULT;

	p = get_proc_task(inode);
	if (!p)
		return -ESRCH;

	if (same_thread_group(current, p))
		set_task_comm(p, buffer);
	else
		count = -EINVAL;

	put_task_struct(p);

	return count;
}

static int comm_show(struct seq_file *m, void *v)
{
	struct inode *inode = m->private;
	struct task_struct *p;

	p = get_proc_task(inode);
	if (!p)
		return -ESRCH;

	task_lock(p);
	seq_printf(m, "%s\n", p->comm);
	task_unlock(p);

	put_task_struct(p);

	return 0;
}

static int comm_open(struct inode *inode, struct file *filp)
{
J
Jovi Zhang 已提交
1578
	return single_open(filp, comm_show, inode);
1579 1580 1581 1582 1583 1584 1585 1586 1587 1588
}

static const struct file_operations proc_pid_set_comm_operations = {
	.open		= comm_open,
	.read		= seq_read,
	.write		= comm_write,
	.llseek		= seq_lseek,
	.release	= single_release,
};

1589
static int proc_exe_link(struct dentry *dentry, struct path *exe_path)
M
Matt Helsley 已提交
1590 1591 1592 1593
{
	struct task_struct *task;
	struct file *exe_file;

1594
	task = get_proc_task(d_inode(dentry));
M
Matt Helsley 已提交
1595 1596
	if (!task)
		return -ENOENT;
M
Mateusz Guzik 已提交
1597
	exe_file = get_task_exe_file(task);
M
Matt Helsley 已提交
1598 1599 1600 1601 1602 1603 1604 1605 1606 1607
	put_task_struct(task);
	if (exe_file) {
		*exe_path = exe_file->f_path;
		path_get(&exe_file->f_path);
		fput(exe_file);
		return 0;
	} else
		return -ENOENT;
}

1608
static const char *proc_pid_get_link(struct dentry *dentry,
1609 1610
				     struct inode *inode,
				     struct delayed_call *done)
L
Linus Torvalds 已提交
1611
{
1612
	struct path path;
L
Linus Torvalds 已提交
1613 1614
	int error = -EACCES;

1615 1616 1617
	if (!dentry)
		return ERR_PTR(-ECHILD);

1618 1619
	/* Are we allowed to snoop on the tasks file descriptors? */
	if (!proc_fd_access_allowed(inode))
L
Linus Torvalds 已提交
1620 1621
		goto out;

1622 1623 1624 1625
	error = PROC_I(inode)->op.proc_get_link(dentry, &path);
	if (error)
		goto out;

1626
	nd_jump_link(&path);
1627
	return NULL;
L
Linus Torvalds 已提交
1628
out:
1629
	return ERR_PTR(error);
L
Linus Torvalds 已提交
1630 1631
}

1632
static int do_proc_readlink(struct path *path, char __user *buffer, int buflen)
L
Linus Torvalds 已提交
1633
{
1634
	char *tmp = (char *)__get_free_page(GFP_KERNEL);
1635
	char *pathname;
L
Linus Torvalds 已提交
1636 1637 1638 1639
	int len;

	if (!tmp)
		return -ENOMEM;
1640

1641
	pathname = d_path(path, tmp, PAGE_SIZE);
1642 1643
	len = PTR_ERR(pathname);
	if (IS_ERR(pathname))
L
Linus Torvalds 已提交
1644
		goto out;
1645
	len = tmp + PAGE_SIZE - 1 - pathname;
L
Linus Torvalds 已提交
1646 1647 1648

	if (len > buflen)
		len = buflen;
1649
	if (copy_to_user(buffer, pathname, len))
L
Linus Torvalds 已提交
1650 1651 1652 1653 1654 1655 1656 1657 1658
		len = -EFAULT;
 out:
	free_page((unsigned long)tmp);
	return len;
}

static int proc_pid_readlink(struct dentry * dentry, char __user * buffer, int buflen)
{
	int error = -EACCES;
1659
	struct inode *inode = d_inode(dentry);
1660
	struct path path;
L
Linus Torvalds 已提交
1661

1662 1663
	/* Are we allowed to snoop on the tasks file descriptors? */
	if (!proc_fd_access_allowed(inode))
L
Linus Torvalds 已提交
1664 1665
		goto out;

1666
	error = PROC_I(inode)->op.proc_get_link(dentry, &path);
L
Linus Torvalds 已提交
1667 1668 1669
	if (error)
		goto out;

1670 1671
	error = do_proc_readlink(&path, buffer, buflen);
	path_put(&path);
L
Linus Torvalds 已提交
1672 1673 1674 1675
out:
	return error;
}

1676
const struct inode_operations proc_pid_link_inode_operations = {
L
Linus Torvalds 已提交
1677
	.readlink	= proc_pid_readlink,
1678
	.get_link	= proc_pid_get_link,
1679
	.setattr	= proc_setattr,
L
Linus Torvalds 已提交
1680 1681
};

1682 1683 1684

/* building an inode */

A
Al Viro 已提交
1685
void task_dump_owner(struct task_struct *task, umode_t mode,
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
		     kuid_t *ruid, kgid_t *rgid)
{
	/* Depending on the state of dumpable compute who should own a
	 * proc file for a task.
	 */
	const struct cred *cred;
	kuid_t uid;
	kgid_t gid;

	/* Default to the tasks effective ownership */
	rcu_read_lock();
	cred = __task_cred(task);
	uid = cred->euid;
	gid = cred->egid;
	rcu_read_unlock();

	/*
	 * Before the /proc/pid/status file was created the only way to read
	 * the effective uid of a /process was to stat /proc/pid.  Reading
	 * /proc/pid/status is slow enough that procps and other packages
	 * kept stating /proc/pid.  To keep the rules in /proc simple I have
	 * made this apply to all per process world readable and executable
	 * directories.
	 */
	if (mode != (S_IFDIR|S_IRUGO|S_IXUGO)) {
		struct mm_struct *mm;
		task_lock(task);
		mm = task->mm;
		/* Make non-dumpable tasks owned by some root */
		if (mm) {
			if (get_dumpable(mm) != SUID_DUMP_USER) {
				struct user_namespace *user_ns = mm->user_ns;

				uid = make_kuid(user_ns, 0);
				if (!uid_valid(uid))
					uid = GLOBAL_ROOT_UID;

				gid = make_kgid(user_ns, 0);
				if (!gid_valid(gid))
					gid = GLOBAL_ROOT_GID;
			}
		} else {
			uid = GLOBAL_ROOT_UID;
			gid = GLOBAL_ROOT_GID;
		}
		task_unlock(task);
	}
	*ruid = uid;
	*rgid = gid;
}

1737 1738
struct inode *proc_pid_make_inode(struct super_block * sb,
				  struct task_struct *task, umode_t mode)
1739 1740 1741
{
	struct inode * inode;
	struct proc_inode *ei;
L
Linus Torvalds 已提交
1742

1743
	/* We need a new inode */
L
Linus Torvalds 已提交
1744

1745 1746 1747 1748 1749 1750
	inode = new_inode(sb);
	if (!inode)
		goto out;

	/* Common stuff */
	ei = PROC_I(inode);
1751
	inode->i_mode = mode;
1752
	inode->i_ino = get_next_ino();
1753
	inode->i_mtime = inode->i_atime = inode->i_ctime = current_time(inode);
1754 1755 1756 1757 1758
	inode->i_op = &proc_def_inode_operations;

	/*
	 * grab the reference to task.
	 */
1759
	ei->pid = get_task_pid(task, PIDTYPE_PID);
1760 1761 1762
	if (!ei->pid)
		goto out_unlock;

1763
	task_dump_owner(task, 0, &inode->i_uid, &inode->i_gid);
1764 1765
	security_task_to_inode(task, inode);

L
Linus Torvalds 已提交
1766
out:
1767 1768 1769 1770 1771
	return inode;

out_unlock:
	iput(inode);
	return NULL;
L
Linus Torvalds 已提交
1772 1773
}

1774 1775
int pid_getattr(const struct path *path, struct kstat *stat,
		u32 request_mask, unsigned int query_flags)
L
Linus Torvalds 已提交
1776
{
1777
	struct inode *inode = d_inode(path->dentry);
1778
	struct task_struct *task;
1779
	struct pid_namespace *pid = path->dentry->d_sb->s_fs_info;
1780

1781
	generic_fillattr(inode, stat);
L
Linus Torvalds 已提交
1782

1783
	rcu_read_lock();
1784 1785
	stat->uid = GLOBAL_ROOT_UID;
	stat->gid = GLOBAL_ROOT_GID;
1786 1787
	task = pid_task(proc_pid(inode), PIDTYPE_PID);
	if (task) {
1788
		if (!has_pid_permissions(pid, task, HIDEPID_INVISIBLE)) {
1789 1790 1791 1792 1793 1794 1795
			rcu_read_unlock();
			/*
			 * This doesn't prevent learning whether PID exists,
			 * it only makes getattr() consistent with readdir().
			 */
			return -ENOENT;
		}
1796
		task_dump_owner(task, inode->i_mode, &stat->uid, &stat->gid);
L
Linus Torvalds 已提交
1797
	}
1798
	rcu_read_unlock();
A
Alan Cox 已提交
1799
	return 0;
L
Linus Torvalds 已提交
1800 1801 1802 1803 1804 1805 1806 1807 1808 1809 1810
}

/* dentry stuff */

/*
 *	Exceptional case: normally we are not allowed to unhash a busy
 * directory. In this case, however, we can do it - no aliasing problems
 * due to the way we treat inodes.
 *
 * Rewrite the inode's ownerships here because the owning task may have
 * performed a setuid(), etc.
1811
 *
L
Linus Torvalds 已提交
1812
 */
1813
int pid_revalidate(struct dentry *dentry, unsigned int flags)
L
Linus Torvalds 已提交
1814
{
1815 1816
	struct inode *inode;
	struct task_struct *task;
1817

1818
	if (flags & LOOKUP_RCU)
1819 1820
		return -ECHILD;

1821
	inode = d_inode(dentry);
1822 1823
	task = get_proc_task(inode);

1824
	if (task) {
1825 1826
		task_dump_owner(task, inode->i_mode, &inode->i_uid, &inode->i_gid);

L
Linus Torvalds 已提交
1827
		inode->i_mode &= ~(S_ISUID | S_ISGID);
L
Linus Torvalds 已提交
1828
		security_task_to_inode(task, inode);
1829
		put_task_struct(task);
L
Linus Torvalds 已提交
1830 1831 1832 1833 1834
		return 1;
	}
	return 0;
}

1835 1836 1837 1838 1839
static inline bool proc_inode_is_dead(struct inode *inode)
{
	return !proc_pid(inode)->tasks[PIDTYPE_PID].first;
}

1840 1841 1842 1843 1844 1845
int pid_delete_dentry(const struct dentry *dentry)
{
	/* Is the task we represent dead?
	 * If so, then don't put the dentry on the lru list,
	 * kill it immediately.
	 */
1846
	return proc_inode_is_dead(d_inode(dentry));
1847 1848
}

1849
const struct dentry_operations pid_dentry_operations =
1850 1851 1852 1853 1854 1855 1856
{
	.d_revalidate	= pid_revalidate,
	.d_delete	= pid_delete_dentry,
};

/* Lookups */

1857 1858 1859 1860 1861 1862 1863 1864 1865 1866 1867 1868
/*
 * Fill a directory entry.
 *
 * If possible create the dcache entry and derive our inode number and
 * file type from dcache entry.
 *
 * Since all of the proc inode numbers are dynamically generated, the inode
 * numbers do not exist until the inode is cache.  This means creating the
 * the dcache entry in readdir is necessary to keep the inode numbers
 * reported by readdir in sync with the inode numbers reported
 * by stat.
 */
A
Al Viro 已提交
1869
bool proc_fill_cache(struct file *file, struct dir_context *ctx,
1870
	const char *name, int len,
1871
	instantiate_t instantiate, struct task_struct *task, const void *ptr)
1872
{
A
Al Viro 已提交
1873
	struct dentry *child, *dir = file->f_path.dentry;
1874
	struct qstr qname = QSTR_INIT(name, len);
1875
	struct inode *inode;
1876 1877
	unsigned type;
	ino_t ino;
1878

1879
	child = d_hash_and_lookup(dir, &qname);
1880
	if (!child) {
1881 1882 1883
		DECLARE_WAIT_QUEUE_HEAD_ONSTACK(wq);
		child = d_alloc_parallel(dir, &qname, &wq);
		if (IS_ERR(child))
1884
			goto end_instantiate;
1885 1886 1887 1888 1889 1890 1891
		if (d_in_lookup(child)) {
			int err = instantiate(d_inode(dir), child, task, ptr);
			d_lookup_done(child);
			if (err < 0) {
				dput(child);
				goto end_instantiate;
			}
1892 1893
		}
	}
1894
	inode = d_inode(child);
1895 1896
	ino = inode->i_ino;
	type = inode->i_mode >> 12;
1897
	dput(child);
A
Al Viro 已提交
1898
	return dir_emit(ctx, name, len, ino, type);
1899 1900 1901

end_instantiate:
	return dir_emit(ctx, name, len, 1, DT_UNKNOWN);
1902 1903
}

1904 1905 1906 1907 1908 1909 1910
/*
 * dname_to_vma_addr - maps a dentry name into two unsigned longs
 * which represent vma start and end addresses.
 */
static int dname_to_vma_addr(struct dentry *dentry,
			     unsigned long *start, unsigned long *end)
{
1911 1912 1913 1914 1915 1916 1917 1918 1919 1920 1921 1922
	const char *str = dentry->d_name.name;
	unsigned long long sval, eval;
	unsigned int len;

	len = _parse_integer(str, 16, &sval);
	if (len & KSTRTOX_OVERFLOW)
		return -EINVAL;
	if (sval != (unsigned long)sval)
		return -EINVAL;
	str += len;

	if (*str != '-')
1923
		return -EINVAL;
1924 1925 1926 1927 1928 1929 1930 1931 1932 1933 1934 1935 1936 1937
	str++;

	len = _parse_integer(str, 16, &eval);
	if (len & KSTRTOX_OVERFLOW)
		return -EINVAL;
	if (eval != (unsigned long)eval)
		return -EINVAL;
	str += len;

	if (*str != '\0')
		return -EINVAL;

	*start = sval;
	*end = eval;
1938 1939 1940 1941

	return 0;
}

1942
static int map_files_d_revalidate(struct dentry *dentry, unsigned int flags)
1943 1944 1945 1946 1947 1948 1949 1950
{
	unsigned long vm_start, vm_end;
	bool exact_vma_exists = false;
	struct mm_struct *mm = NULL;
	struct task_struct *task;
	struct inode *inode;
	int status = 0;

1951
	if (flags & LOOKUP_RCU)
1952 1953
		return -ECHILD;

1954
	inode = d_inode(dentry);
1955 1956 1957 1958
	task = get_proc_task(inode);
	if (!task)
		goto out_notask;

1959
	mm = mm_access(task, PTRACE_MODE_READ_FSCREDS);
1960
	if (IS_ERR_OR_NULL(mm))
1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971
		goto out;

	if (!dname_to_vma_addr(dentry, &vm_start, &vm_end)) {
		down_read(&mm->mmap_sem);
		exact_vma_exists = !!find_exact_vma(mm, vm_start, vm_end);
		up_read(&mm->mmap_sem);
	}

	mmput(mm);

	if (exact_vma_exists) {
1972 1973
		task_dump_owner(task, 0, &inode->i_uid, &inode->i_gid);

1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989
		security_task_to_inode(task, inode);
		status = 1;
	}

out:
	put_task_struct(task);

out_notask:
	return status;
}

static const struct dentry_operations tid_map_files_dentry_operations = {
	.d_revalidate	= map_files_d_revalidate,
	.d_delete	= pid_delete_dentry,
};

1990
static int map_files_get_link(struct dentry *dentry, struct path *path)
1991 1992 1993 1994 1995 1996 1997 1998
{
	unsigned long vm_start, vm_end;
	struct vm_area_struct *vma;
	struct task_struct *task;
	struct mm_struct *mm;
	int rc;

	rc = -ENOENT;
1999
	task = get_proc_task(d_inode(dentry));
2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011
	if (!task)
		goto out;

	mm = get_task_mm(task);
	put_task_struct(task);
	if (!mm)
		goto out;

	rc = dname_to_vma_addr(dentry, &vm_start, &vm_end);
	if (rc)
		goto out_mmput;

2012
	rc = -ENOENT;
2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028
	down_read(&mm->mmap_sem);
	vma = find_exact_vma(mm, vm_start, vm_end);
	if (vma && vma->vm_file) {
		*path = vma->vm_file->f_path;
		path_get(path);
		rc = 0;
	}
	up_read(&mm->mmap_sem);

out_mmput:
	mmput(mm);
out:
	return rc;
}

struct map_files_info {
2029 2030
	unsigned long	start;
	unsigned long	end;
2031
	fmode_t		mode;
2032 2033
};

2034 2035 2036 2037 2038 2039
/*
 * Only allow CAP_SYS_ADMIN to follow the links, due to concerns about how the
 * symlinks may be used to bypass permissions on ancestor directories in the
 * path to the file in question.
 */
static const char *
2040
proc_map_files_get_link(struct dentry *dentry,
2041 2042
			struct inode *inode,
		        struct delayed_call *done)
2043 2044 2045 2046
{
	if (!capable(CAP_SYS_ADMIN))
		return ERR_PTR(-EPERM);

2047
	return proc_pid_get_link(dentry, inode, done);
2048 2049 2050
}

/*
2051
 * Identical to proc_pid_link_inode_operations except for get_link()
2052 2053 2054
 */
static const struct inode_operations proc_map_files_link_inode_operations = {
	.readlink	= proc_pid_readlink,
2055
	.get_link	= proc_map_files_get_link,
2056 2057 2058
	.setattr	= proc_setattr,
};

2059
static int
2060 2061 2062
proc_map_files_instantiate(struct inode *dir, struct dentry *dentry,
			   struct task_struct *task, const void *ptr)
{
2063
	fmode_t mode = (fmode_t)(unsigned long)ptr;
2064 2065 2066
	struct proc_inode *ei;
	struct inode *inode;

2067 2068 2069
	inode = proc_pid_make_inode(dir->i_sb, task, S_IFLNK |
				    ((mode & FMODE_READ ) ? S_IRUSR : 0) |
				    ((mode & FMODE_WRITE) ? S_IWUSR : 0));
2070
	if (!inode)
2071
		return -ENOENT;
2072 2073

	ei = PROC_I(inode);
2074
	ei->op.proc_get_link = map_files_get_link;
2075

2076
	inode->i_op = &proc_map_files_link_inode_operations;
2077 2078 2079 2080 2081
	inode->i_size = 64;

	d_set_d_op(dentry, &tid_map_files_dentry_operations);
	d_add(dentry, inode);

2082
	return 0;
2083 2084 2085
}

static struct dentry *proc_map_files_lookup(struct inode *dir,
A
Al Viro 已提交
2086
		struct dentry *dentry, unsigned int flags)
2087 2088 2089 2090
{
	unsigned long vm_start, vm_end;
	struct vm_area_struct *vma;
	struct task_struct *task;
2091
	int result;
2092 2093
	struct mm_struct *mm;

2094
	result = -ENOENT;
2095 2096 2097 2098
	task = get_proc_task(dir);
	if (!task)
		goto out;

2099
	result = -EACCES;
2100
	if (!ptrace_may_access(task, PTRACE_MODE_READ_FSCREDS))
2101 2102
		goto out_put_task;

2103
	result = -ENOENT;
2104
	if (dname_to_vma_addr(dentry, &vm_start, &vm_end))
2105
		goto out_put_task;
2106 2107 2108

	mm = get_task_mm(task);
	if (!mm)
2109
		goto out_put_task;
2110 2111 2112 2113 2114 2115

	down_read(&mm->mmap_sem);
	vma = find_exact_vma(mm, vm_start, vm_end);
	if (!vma)
		goto out_no_vma;

2116 2117 2118
	if (vma->vm_file)
		result = proc_map_files_instantiate(dir, dentry, task,
				(void *)(unsigned long)vma->vm_file->f_mode);
2119 2120 2121 2122 2123 2124 2125

out_no_vma:
	up_read(&mm->mmap_sem);
	mmput(mm);
out_put_task:
	put_task_struct(task);
out:
2126
	return ERR_PTR(result);
2127 2128 2129 2130 2131 2132 2133 2134 2135
}

static const struct inode_operations proc_map_files_inode_operations = {
	.lookup		= proc_map_files_lookup,
	.permission	= proc_fd_permission,
	.setattr	= proc_setattr,
};

static int
A
Al Viro 已提交
2136
proc_map_files_readdir(struct file *file, struct dir_context *ctx)
2137 2138 2139 2140
{
	struct vm_area_struct *vma;
	struct task_struct *task;
	struct mm_struct *mm;
A
Al Viro 已提交
2141 2142 2143 2144
	unsigned long nr_files, pos, i;
	struct flex_array *fa = NULL;
	struct map_files_info info;
	struct map_files_info *p;
2145 2146 2147
	int ret;

	ret = -ENOENT;
A
Al Viro 已提交
2148
	task = get_proc_task(file_inode(file));
2149 2150 2151 2152
	if (!task)
		goto out;

	ret = -EACCES;
2153
	if (!ptrace_may_access(task, PTRACE_MODE_READ_FSCREDS))
2154 2155 2156
		goto out_put_task;

	ret = 0;
A
Al Viro 已提交
2157 2158
	if (!dir_emit_dots(file, ctx))
		goto out_put_task;
2159

A
Al Viro 已提交
2160 2161 2162 2163
	mm = get_task_mm(task);
	if (!mm)
		goto out_put_task;
	down_read(&mm->mmap_sem);
2164

A
Al Viro 已提交
2165
	nr_files = 0;
2166

A
Al Viro 已提交
2167 2168 2169 2170 2171 2172 2173 2174 2175
	/*
	 * We need two passes here:
	 *
	 *  1) Collect vmas of mapped files with mmap_sem taken
	 *  2) Release mmap_sem and instantiate entries
	 *
	 * otherwise we get lockdep complained, since filldir()
	 * routine might require mmap_sem taken in might_fault().
	 */
2176

A
Al Viro 已提交
2177 2178 2179 2180 2181 2182 2183 2184 2185 2186 2187 2188 2189 2190 2191 2192
	for (vma = mm->mmap, pos = 2; vma; vma = vma->vm_next) {
		if (vma->vm_file && ++pos > ctx->pos)
			nr_files++;
	}

	if (nr_files) {
		fa = flex_array_alloc(sizeof(info), nr_files,
					GFP_KERNEL);
		if (!fa || flex_array_prealloc(fa, 0, nr_files,
						GFP_KERNEL)) {
			ret = -ENOMEM;
			if (fa)
				flex_array_free(fa);
			up_read(&mm->mmap_sem);
			mmput(mm);
			goto out_put_task;
2193
		}
A
Al Viro 已提交
2194 2195 2196 2197 2198 2199 2200
		for (i = 0, vma = mm->mmap, pos = 2; vma;
				vma = vma->vm_next) {
			if (!vma->vm_file)
				continue;
			if (++pos <= ctx->pos)
				continue;

2201 2202
			info.start = vma->vm_start;
			info.end = vma->vm_end;
A
Al Viro 已提交
2203 2204 2205
			info.mode = vma->vm_file->f_mode;
			if (flex_array_put(fa, i++, &info, GFP_KERNEL))
				BUG();
2206 2207
		}
	}
A
Al Viro 已提交
2208 2209 2210
	up_read(&mm->mmap_sem);

	for (i = 0; i < nr_files; i++) {
2211 2212 2213
		char buf[4 * sizeof(long) + 2];	/* max: %lx-%lx\0 */
		unsigned int len;

A
Al Viro 已提交
2214
		p = flex_array_get(fa, i);
2215
		len = snprintf(buf, sizeof(buf), "%lx-%lx", p->start, p->end);
A
Al Viro 已提交
2216
		if (!proc_fill_cache(file, ctx,
2217
				      buf, len,
A
Al Viro 已提交
2218 2219 2220 2221 2222
				      proc_map_files_instantiate,
				      task,
				      (void *)(unsigned long)p->mode))
			break;
		ctx->pos++;
2223
	}
A
Al Viro 已提交
2224 2225 2226
	if (fa)
		flex_array_free(fa);
	mmput(mm);
2227 2228 2229 2230 2231 2232 2233 2234 2235

out_put_task:
	put_task_struct(task);
out:
	return ret;
}

static const struct file_operations proc_map_files_operations = {
	.read		= generic_read_dir,
2236 2237
	.iterate_shared	= proc_map_files_readdir,
	.llseek		= generic_file_llseek,
2238 2239
};

2240
#if defined(CONFIG_CHECKPOINT_RESTORE) && defined(CONFIG_POSIX_TIMERS)
2241 2242 2243 2244
struct timers_private {
	struct pid *pid;
	struct task_struct *task;
	struct sighand_struct *sighand;
2245
	struct pid_namespace *ns;
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 2272 2273 2274 2275 2276 2277 2278 2279 2280 2281 2282 2283 2284 2285 2286 2287
	unsigned long flags;
};

static void *timers_start(struct seq_file *m, loff_t *pos)
{
	struct timers_private *tp = m->private;

	tp->task = get_pid_task(tp->pid, PIDTYPE_PID);
	if (!tp->task)
		return ERR_PTR(-ESRCH);

	tp->sighand = lock_task_sighand(tp->task, &tp->flags);
	if (!tp->sighand)
		return ERR_PTR(-ESRCH);

	return seq_list_start(&tp->task->signal->posix_timers, *pos);
}

static void *timers_next(struct seq_file *m, void *v, loff_t *pos)
{
	struct timers_private *tp = m->private;
	return seq_list_next(v, &tp->task->signal->posix_timers, pos);
}

static void timers_stop(struct seq_file *m, void *v)
{
	struct timers_private *tp = m->private;

	if (tp->sighand) {
		unlock_task_sighand(tp->task, &tp->flags);
		tp->sighand = NULL;
	}

	if (tp->task) {
		put_task_struct(tp->task);
		tp->task = NULL;
	}
}

static int show_timer(struct seq_file *m, void *v)
{
	struct k_itimer *timer;
2288 2289
	struct timers_private *tp = m->private;
	int notify;
2290
	static const char * const nstr[] = {
2291 2292 2293 2294
		[SIGEV_SIGNAL] = "signal",
		[SIGEV_NONE] = "none",
		[SIGEV_THREAD] = "thread",
	};
2295 2296

	timer = list_entry((struct list_head *)v, struct k_itimer, list);
2297 2298
	notify = timer->it_sigev_notify;

2299
	seq_printf(m, "ID: %d\n", timer->it_id);
2300
	seq_printf(m, "signal: %d/%px\n",
2301 2302
		   timer->sigq->info.si_signo,
		   timer->sigq->info.si_value.sival_ptr);
2303
	seq_printf(m, "notify: %s/%s.%d\n",
2304 2305 2306
		   nstr[notify & ~SIGEV_THREAD_ID],
		   (notify & SIGEV_THREAD_ID) ? "tid" : "pid",
		   pid_nr_ns(timer->it_pid, tp->ns));
2307
	seq_printf(m, "ClockID: %d\n", timer->it_clock);
2308 2309 2310 2311 2312 2313 2314 2315 2316 2317 2318 2319 2320 2321 2322 2323 2324 2325 2326 2327 2328

	return 0;
}

static const struct seq_operations proc_timers_seq_ops = {
	.start	= timers_start,
	.next	= timers_next,
	.stop	= timers_stop,
	.show	= show_timer,
};

static int proc_timers_open(struct inode *inode, struct file *file)
{
	struct timers_private *tp;

	tp = __seq_open_private(file, &proc_timers_seq_ops,
			sizeof(struct timers_private));
	if (!tp)
		return -ENOMEM;

	tp->pid = proc_pid(inode);
2329
	tp->ns = inode->i_sb->s_fs_info;
2330 2331 2332 2333 2334 2335 2336 2337 2338
	return 0;
}

static const struct file_operations proc_timers_operations = {
	.open		= proc_timers_open,
	.read		= seq_read,
	.llseek		= seq_lseek,
	.release	= seq_release_private,
};
2339
#endif
2340

2341 2342 2343 2344 2345 2346 2347 2348 2349 2350 2351 2352 2353 2354 2355 2356
static ssize_t timerslack_ns_write(struct file *file, const char __user *buf,
					size_t count, loff_t *offset)
{
	struct inode *inode = file_inode(file);
	struct task_struct *p;
	u64 slack_ns;
	int err;

	err = kstrtoull_from_user(buf, count, 10, &slack_ns);
	if (err < 0)
		return err;

	p = get_proc_task(inode);
	if (!p)
		return -ESRCH;

2357 2358 2359 2360 2361
	if (p != current) {
		if (!capable(CAP_SYS_NICE)) {
			count = -EPERM;
			goto out;
		}
2362

2363 2364 2365 2366 2367
		err = security_task_setscheduler(p);
		if (err) {
			count = err;
			goto out;
		}
2368 2369
	}

2370 2371 2372 2373 2374 2375 2376 2377
	task_lock(p);
	if (slack_ns == 0)
		p->timer_slack_ns = p->default_timer_slack_ns;
	else
		p->timer_slack_ns = slack_ns;
	task_unlock(p);

out:
2378 2379 2380 2381 2382 2383 2384 2385 2386
	put_task_struct(p);

	return count;
}

static int timerslack_ns_show(struct seq_file *m, void *v)
{
	struct inode *inode = m->private;
	struct task_struct *p;
2387
	int err = 0;
2388 2389 2390 2391 2392

	p = get_proc_task(inode);
	if (!p)
		return -ESRCH;

2393
	if (p != current) {
2394

2395 2396 2397 2398 2399 2400 2401 2402
		if (!capable(CAP_SYS_NICE)) {
			err = -EPERM;
			goto out;
		}
		err = security_task_getscheduler(p);
		if (err)
			goto out;
	}
2403

2404 2405 2406 2407 2408
	task_lock(p);
	seq_printf(m, "%llu\n", p->timer_slack_ns);
	task_unlock(p);

out:
2409 2410 2411 2412 2413 2414 2415 2416 2417 2418 2419 2420 2421 2422 2423 2424 2425 2426
	put_task_struct(p);

	return err;
}

static int timerslack_ns_open(struct inode *inode, struct file *filp)
{
	return single_open(filp, timerslack_ns_show, inode);
}

static const struct file_operations proc_pid_set_timerslack_ns_operations = {
	.open		= timerslack_ns_open,
	.read		= seq_read,
	.write		= timerslack_ns_write,
	.llseek		= seq_lseek,
	.release	= single_release,
};

2427
static int proc_pident_instantiate(struct inode *dir,
2428
	struct dentry *dentry, struct task_struct *task, const void *ptr)
2429
{
2430
	const struct pid_entry *p = ptr;
2431 2432 2433
	struct inode *inode;
	struct proc_inode *ei;

2434
	inode = proc_pid_make_inode(dir->i_sb, task, p->mode);
2435 2436 2437 2438 2439
	if (!inode)
		goto out;

	ei = PROC_I(inode);
	if (S_ISDIR(inode->i_mode))
M
Miklos Szeredi 已提交
2440
		set_nlink(inode, 2);	/* Use getattr to fix if necessary */
2441 2442 2443 2444 2445
	if (p->iop)
		inode->i_op = p->iop;
	if (p->fop)
		inode->i_fop = p->fop;
	ei->op = p->op;
2446
	d_set_d_op(dentry, &pid_dentry_operations);
2447 2448
	d_add(dentry, inode);
	/* Close the race of the process dying before we return the dentry */
2449
	if (pid_revalidate(dentry, 0))
2450
		return 0;
2451
out:
2452
	return -ENOENT;
2453 2454
}

L
Linus Torvalds 已提交
2455 2456
static struct dentry *proc_pident_lookup(struct inode *dir, 
					 struct dentry *dentry,
2457
					 const struct pid_entry *ents,
2458
					 unsigned int nents)
L
Linus Torvalds 已提交
2459
{
2460
	int error;
2461
	struct task_struct *task = get_proc_task(dir);
2462
	const struct pid_entry *p, *last;
L
Linus Torvalds 已提交
2463

2464
	error = -ENOENT;
L
Linus Torvalds 已提交
2465

2466 2467
	if (!task)
		goto out_no_task;
L
Linus Torvalds 已提交
2468

2469 2470 2471 2472
	/*
	 * Yes, it does not scale. And it should not. Don't add
	 * new entries into /proc/<tgid>/ without very good reasons.
	 */
2473 2474
	last = &ents[nents];
	for (p = ents; p < last; p++) {
L
Linus Torvalds 已提交
2475 2476 2477 2478 2479
		if (p->len != dentry->d_name.len)
			continue;
		if (!memcmp(dentry->d_name.name, p->name, p->len))
			break;
	}
2480
	if (p >= last)
L
Linus Torvalds 已提交
2481 2482
		goto out;

2483
	error = proc_pident_instantiate(dir, dentry, task, p);
L
Linus Torvalds 已提交
2484
out:
2485 2486
	put_task_struct(task);
out_no_task:
2487
	return ERR_PTR(error);
L
Linus Torvalds 已提交
2488 2489
}

A
Al Viro 已提交
2490
static int proc_pident_readdir(struct file *file, struct dir_context *ctx,
2491
		const struct pid_entry *ents, unsigned int nents)
2492
{
A
Al Viro 已提交
2493 2494
	struct task_struct *task = get_proc_task(file_inode(file));
	const struct pid_entry *p;
2495 2496

	if (!task)
A
Al Viro 已提交
2497
		return -ENOENT;
2498

A
Al Viro 已提交
2499 2500 2501 2502 2503
	if (!dir_emit_dots(file, ctx))
		goto out;

	if (ctx->pos >= nents + 2)
		goto out;
2504

2505
	for (p = ents + (ctx->pos - 2); p < ents + nents; p++) {
A
Al Viro 已提交
2506 2507 2508 2509 2510
		if (!proc_fill_cache(file, ctx, p->name, p->len,
				proc_pident_instantiate, task, p))
			break;
		ctx->pos++;
	}
2511
out:
2512
	put_task_struct(task);
A
Al Viro 已提交
2513
	return 0;
L
Linus Torvalds 已提交
2514 2515
}

2516 2517 2518 2519
#ifdef CONFIG_SECURITY
static ssize_t proc_pid_attr_read(struct file * file, char __user * buf,
				  size_t count, loff_t *ppos)
{
A
Al Viro 已提交
2520
	struct inode * inode = file_inode(file);
2521
	char *p = NULL;
2522 2523 2524 2525
	ssize_t length;
	struct task_struct *task = get_proc_task(inode);

	if (!task)
2526
		return -ESRCH;
2527 2528

	length = security_getprocattr(task,
2529
				      (char*)file->f_path.dentry->d_name.name,
2530
				      &p);
2531
	put_task_struct(task);
2532 2533 2534
	if (length > 0)
		length = simple_read_from_buffer(buf, count, ppos, p, length);
	kfree(p);
2535
	return length;
L
Linus Torvalds 已提交
2536 2537
}

2538 2539 2540
static ssize_t proc_pid_attr_write(struct file * file, const char __user * buf,
				   size_t count, loff_t *ppos)
{
A
Al Viro 已提交
2541
	struct inode * inode = file_inode(file);
2542
	void *page;
2543 2544 2545 2546 2547 2548
	ssize_t length;
	struct task_struct *task = get_proc_task(inode);

	length = -ESRCH;
	if (!task)
		goto out_no_task;
2549 2550 2551 2552 2553 2554

	/* A task may only write its own attributes. */
	length = -EACCES;
	if (current != task)
		goto out;

2555 2556 2557 2558 2559 2560 2561 2562
	if (count > PAGE_SIZE)
		count = PAGE_SIZE;

	/* No partial writes. */
	length = -EINVAL;
	if (*ppos != 0)
		goto out;

2563 2564 2565
	page = memdup_user(buf, count);
	if (IS_ERR(page)) {
		length = PTR_ERR(page);
2566
		goto out;
2567
	}
2568

2569
	/* Guard against adverse ptrace interaction */
2570
	length = mutex_lock_interruptible(&current->signal->cred_guard_mutex);
2571 2572 2573
	if (length < 0)
		goto out_free;

2574
	length = security_setprocattr(file->f_path.dentry->d_name.name,
2575
				      page, count);
2576
	mutex_unlock(&current->signal->cred_guard_mutex);
2577
out_free:
2578
	kfree(page);
2579 2580 2581 2582 2583 2584
out:
	put_task_struct(task);
out_no_task:
	return length;
}

2585
static const struct file_operations proc_pid_attr_operations = {
2586 2587
	.read		= proc_pid_attr_read,
	.write		= proc_pid_attr_write,
2588
	.llseek		= generic_file_llseek,
2589 2590
};

2591
static const struct pid_entry attr_dir_stuff[] = {
A
Alexey Dobriyan 已提交
2592 2593 2594 2595 2596 2597
	REG("current",    S_IRUGO|S_IWUGO, proc_pid_attr_operations),
	REG("prev",       S_IRUGO,	   proc_pid_attr_operations),
	REG("exec",       S_IRUGO|S_IWUGO, proc_pid_attr_operations),
	REG("fscreate",   S_IRUGO|S_IWUGO, proc_pid_attr_operations),
	REG("keycreate",  S_IRUGO|S_IWUGO, proc_pid_attr_operations),
	REG("sockcreate", S_IRUGO|S_IWUGO, proc_pid_attr_operations),
2598 2599
};

A
Al Viro 已提交
2600
static int proc_attr_dir_readdir(struct file *file, struct dir_context *ctx)
2601
{
A
Al Viro 已提交
2602 2603
	return proc_pident_readdir(file, ctx, 
				   attr_dir_stuff, ARRAY_SIZE(attr_dir_stuff));
2604 2605
}

2606
static const struct file_operations proc_attr_dir_operations = {
L
Linus Torvalds 已提交
2607
	.read		= generic_read_dir,
2608 2609
	.iterate_shared	= proc_attr_dir_readdir,
	.llseek		= generic_file_llseek,
L
Linus Torvalds 已提交
2610 2611
};

2612
static struct dentry *proc_attr_dir_lookup(struct inode *dir,
A
Al Viro 已提交
2613
				struct dentry *dentry, unsigned int flags)
2614
{
2615 2616
	return proc_pident_lookup(dir, dentry,
				  attr_dir_stuff, ARRAY_SIZE(attr_dir_stuff));
2617 2618
}

2619
static const struct inode_operations proc_attr_dir_inode_operations = {
2620
	.lookup		= proc_attr_dir_lookup,
2621
	.getattr	= pid_getattr,
2622
	.setattr	= proc_setattr,
L
Linus Torvalds 已提交
2623 2624
};

2625 2626
#endif

C
Christoph Hellwig 已提交
2627
#ifdef CONFIG_ELF_CORE
2628 2629 2630
static ssize_t proc_coredump_filter_read(struct file *file, char __user *buf,
					 size_t count, loff_t *ppos)
{
A
Al Viro 已提交
2631
	struct task_struct *task = get_proc_task(file_inode(file));
2632 2633 2634 2635 2636 2637 2638 2639 2640 2641 2642 2643 2644 2645 2646 2647 2648 2649 2650 2651 2652 2653 2654 2655 2656 2657 2658 2659 2660 2661 2662 2663 2664 2665 2666
	struct mm_struct *mm;
	char buffer[PROC_NUMBUF];
	size_t len;
	int ret;

	if (!task)
		return -ESRCH;

	ret = 0;
	mm = get_task_mm(task);
	if (mm) {
		len = snprintf(buffer, sizeof(buffer), "%08lx\n",
			       ((mm->flags & MMF_DUMP_FILTER_MASK) >>
				MMF_DUMP_FILTER_SHIFT));
		mmput(mm);
		ret = simple_read_from_buffer(buf, count, ppos, buffer, len);
	}

	put_task_struct(task);

	return ret;
}

static ssize_t proc_coredump_filter_write(struct file *file,
					  const char __user *buf,
					  size_t count,
					  loff_t *ppos)
{
	struct task_struct *task;
	struct mm_struct *mm;
	unsigned int val;
	int ret;
	int i;
	unsigned long mask;

2667 2668 2669
	ret = kstrtouint_from_user(buf, count, 0, &val);
	if (ret < 0)
		return ret;
2670 2671

	ret = -ESRCH;
A
Al Viro 已提交
2672
	task = get_proc_task(file_inode(file));
2673 2674 2675 2676 2677 2678
	if (!task)
		goto out_no_task;

	mm = get_task_mm(task);
	if (!mm)
		goto out_no_mm;
2679
	ret = 0;
2680 2681 2682 2683 2684 2685 2686 2687 2688 2689 2690 2691

	for (i = 0, mask = 1; i < MMF_DUMP_FILTER_BITS; i++, mask <<= 1) {
		if (val & mask)
			set_bit(i + MMF_DUMP_FILTER_SHIFT, &mm->flags);
		else
			clear_bit(i + MMF_DUMP_FILTER_SHIFT, &mm->flags);
	}

	mmput(mm);
 out_no_mm:
	put_task_struct(task);
 out_no_task:
2692 2693 2694
	if (ret < 0)
		return ret;
	return count;
2695 2696 2697 2698 2699
}

static const struct file_operations proc_coredump_filter_operations = {
	.read		= proc_coredump_filter_read,
	.write		= proc_coredump_filter_write,
2700
	.llseek		= generic_file_llseek,
2701 2702 2703
};
#endif

2704
#ifdef CONFIG_TASK_IO_ACCOUNTING
2705
static int do_io_accounting(struct task_struct *task, struct seq_file *m, int whole)
2706
{
2707
	struct task_io_accounting acct = task->ioac;
2708
	unsigned long flags;
2709
	int result;
2710

2711 2712 2713 2714
	result = mutex_lock_killable(&task->signal->cred_guard_mutex);
	if (result)
		return result;

2715
	if (!ptrace_may_access(task, PTRACE_MODE_READ_FSCREDS)) {
2716 2717 2718
		result = -EACCES;
		goto out_unlock;
	}
2719

2720 2721 2722 2723 2724 2725 2726 2727
	if (whole && lock_task_sighand(task, &flags)) {
		struct task_struct *t = task;

		task_io_accounting_add(&acct, &task->signal->ioac);
		while_each_thread(task, t)
			task_io_accounting_add(&acct, &t->ioac);

		unlock_task_sighand(task, &flags);
2728
	}
2729 2730 2731 2732 2733 2734 2735 2736 2737 2738 2739 2740 2741 2742 2743 2744 2745
	seq_printf(m,
		   "rchar: %llu\n"
		   "wchar: %llu\n"
		   "syscr: %llu\n"
		   "syscw: %llu\n"
		   "read_bytes: %llu\n"
		   "write_bytes: %llu\n"
		   "cancelled_write_bytes: %llu\n",
		   (unsigned long long)acct.rchar,
		   (unsigned long long)acct.wchar,
		   (unsigned long long)acct.syscr,
		   (unsigned long long)acct.syscw,
		   (unsigned long long)acct.read_bytes,
		   (unsigned long long)acct.write_bytes,
		   (unsigned long long)acct.cancelled_write_bytes);
	result = 0;

2746 2747 2748
out_unlock:
	mutex_unlock(&task->signal->cred_guard_mutex);
	return result;
2749 2750
}

2751 2752
static int proc_tid_io_accounting(struct seq_file *m, struct pid_namespace *ns,
				  struct pid *pid, struct task_struct *task)
2753
{
2754
	return do_io_accounting(task, m, 0);
2755
}
2756

2757 2758
static int proc_tgid_io_accounting(struct seq_file *m, struct pid_namespace *ns,
				   struct pid *pid, struct task_struct *task)
2759
{
2760
	return do_io_accounting(task, m, 1);
2761 2762
}
#endif /* CONFIG_TASK_IO_ACCOUNTING */
2763

2764 2765
#ifdef CONFIG_USER_NS
static int proc_id_map_open(struct inode *inode, struct file *file,
F
Fabian Frederick 已提交
2766
	const struct seq_operations *seq_ops)
2767 2768 2769 2770 2771 2772 2773 2774 2775 2776 2777 2778 2779 2780 2781 2782 2783 2784 2785 2786 2787 2788 2789 2790 2791 2792 2793 2794 2795 2796 2797 2798 2799 2800 2801 2802 2803 2804 2805 2806 2807 2808 2809 2810 2811 2812 2813 2814
{
	struct user_namespace *ns = NULL;
	struct task_struct *task;
	struct seq_file *seq;
	int ret = -EINVAL;

	task = get_proc_task(inode);
	if (task) {
		rcu_read_lock();
		ns = get_user_ns(task_cred_xxx(task, user_ns));
		rcu_read_unlock();
		put_task_struct(task);
	}
	if (!ns)
		goto err;

	ret = seq_open(file, seq_ops);
	if (ret)
		goto err_put_ns;

	seq = file->private_data;
	seq->private = ns;

	return 0;
err_put_ns:
	put_user_ns(ns);
err:
	return ret;
}

static int proc_id_map_release(struct inode *inode, struct file *file)
{
	struct seq_file *seq = file->private_data;
	struct user_namespace *ns = seq->private;
	put_user_ns(ns);
	return seq_release(inode, file);
}

static int proc_uid_map_open(struct inode *inode, struct file *file)
{
	return proc_id_map_open(inode, file, &proc_uid_seq_operations);
}

static int proc_gid_map_open(struct inode *inode, struct file *file)
{
	return proc_id_map_open(inode, file, &proc_gid_seq_operations);
}

2815 2816 2817 2818 2819
static int proc_projid_map_open(struct inode *inode, struct file *file)
{
	return proc_id_map_open(inode, file, &proc_projid_seq_operations);
}

2820 2821 2822 2823 2824 2825 2826 2827 2828 2829 2830 2831 2832 2833 2834
static const struct file_operations proc_uid_map_operations = {
	.open		= proc_uid_map_open,
	.write		= proc_uid_map_write,
	.read		= seq_read,
	.llseek		= seq_lseek,
	.release	= proc_id_map_release,
};

static const struct file_operations proc_gid_map_operations = {
	.open		= proc_gid_map_open,
	.write		= proc_gid_map_write,
	.read		= seq_read,
	.llseek		= seq_lseek,
	.release	= proc_id_map_release,
};
2835 2836 2837 2838 2839 2840 2841 2842

static const struct file_operations proc_projid_map_operations = {
	.open		= proc_projid_map_open,
	.write		= proc_projid_map_write,
	.read		= seq_read,
	.llseek		= seq_lseek,
	.release	= proc_id_map_release,
};
2843 2844 2845 2846 2847 2848 2849 2850 2851 2852 2853 2854 2855 2856 2857 2858 2859 2860 2861 2862 2863 2864 2865 2866 2867 2868 2869 2870 2871 2872 2873 2874 2875 2876 2877 2878 2879 2880 2881 2882 2883 2884 2885 2886 2887 2888 2889 2890 2891 2892 2893

static int proc_setgroups_open(struct inode *inode, struct file *file)
{
	struct user_namespace *ns = NULL;
	struct task_struct *task;
	int ret;

	ret = -ESRCH;
	task = get_proc_task(inode);
	if (task) {
		rcu_read_lock();
		ns = get_user_ns(task_cred_xxx(task, user_ns));
		rcu_read_unlock();
		put_task_struct(task);
	}
	if (!ns)
		goto err;

	if (file->f_mode & FMODE_WRITE) {
		ret = -EACCES;
		if (!ns_capable(ns, CAP_SYS_ADMIN))
			goto err_put_ns;
	}

	ret = single_open(file, &proc_setgroups_show, ns);
	if (ret)
		goto err_put_ns;

	return 0;
err_put_ns:
	put_user_ns(ns);
err:
	return ret;
}

static int proc_setgroups_release(struct inode *inode, struct file *file)
{
	struct seq_file *seq = file->private_data;
	struct user_namespace *ns = seq->private;
	int ret = single_release(inode, file);
	put_user_ns(ns);
	return ret;
}

static const struct file_operations proc_setgroups_operations = {
	.open		= proc_setgroups_open,
	.write		= proc_setgroups_write,
	.read		= seq_read,
	.llseek		= seq_lseek,
	.release	= proc_setgroups_release,
};
2894 2895
#endif /* CONFIG_USER_NS */

2896 2897 2898
static int proc_pid_personality(struct seq_file *m, struct pid_namespace *ns,
				struct pid *pid, struct task_struct *task)
{
2899 2900 2901 2902 2903 2904
	int err = lock_trace(task);
	if (!err) {
		seq_printf(m, "%08x\n", task->personality);
		unlock_trace(task);
	}
	return err;
2905 2906
}

2907 2908 2909 2910 2911 2912 2913 2914 2915
#ifdef CONFIG_LIVEPATCH
static int proc_pid_patch_state(struct seq_file *m, struct pid_namespace *ns,
				struct pid *pid, struct task_struct *task)
{
	seq_printf(m, "%d\n", task->patch_state);
	return 0;
}
#endif /* CONFIG_LIVEPATCH */

2916 2917 2918
/*
 * Thread groups
 */
2919
static const struct file_operations proc_task_operations;
2920
static const struct inode_operations proc_task_inode_operations;
2921

2922
static const struct pid_entry tgid_base_stuff[] = {
A
Alexey Dobriyan 已提交
2923 2924
	DIR("task",       S_IRUGO|S_IXUGO, proc_task_inode_operations, proc_task_operations),
	DIR("fd",         S_IRUSR|S_IXUSR, proc_fd_inode_operations, proc_fd_operations),
2925
	DIR("map_files",  S_IRUSR|S_IXUSR, proc_map_files_inode_operations, proc_map_files_operations),
A
Alexey Dobriyan 已提交
2926
	DIR("fdinfo",     S_IRUSR|S_IXUSR, proc_fdinfo_inode_operations, proc_fdinfo_operations),
2927
	DIR("ns",	  S_IRUSR|S_IXUGO, proc_ns_dir_inode_operations, proc_ns_dir_operations),
A
Andrew Morton 已提交
2928
#ifdef CONFIG_NET
A
Alexey Dobriyan 已提交
2929
	DIR("net",        S_IRUGO|S_IXUGO, proc_net_inode_operations, proc_net_operations),
A
Andrew Morton 已提交
2930
#endif
A
Alexey Dobriyan 已提交
2931
	REG("environ",    S_IRUSR, proc_environ_operations),
A
Al Viro 已提交
2932
	REG("auxv",       S_IRUSR, proc_auxv_operations),
A
Alexey Dobriyan 已提交
2933
	ONE("status",     S_IRUGO, proc_pid_status),
2934
	ONE("personality", S_IRUSR, proc_pid_personality),
2935
	ONE("limits",	  S_IRUGO, proc_pid_limits),
I
Ingo Molnar 已提交
2936
#ifdef CONFIG_SCHED_DEBUG
A
Alexey Dobriyan 已提交
2937
	REG("sched",      S_IRUGO|S_IWUSR, proc_pid_sched_operations),
2938 2939 2940
#endif
#ifdef CONFIG_SCHED_AUTOGROUP
	REG("autogroup",  S_IRUGO|S_IWUSR, proc_pid_sched_autogroup_operations),
R
Roland McGrath 已提交
2941
#endif
2942
	REG("comm",      S_IRUGO|S_IWUSR, proc_pid_set_comm_operations),
R
Roland McGrath 已提交
2943
#ifdef CONFIG_HAVE_ARCH_TRACEHOOK
2944
	ONE("syscall",    S_IRUSR, proc_pid_syscall),
I
Ingo Molnar 已提交
2945
#endif
2946
	REG("cmdline",    S_IRUGO, proc_pid_cmdline_ops),
A
Alexey Dobriyan 已提交
2947 2948
	ONE("stat",       S_IRUGO, proc_tgid_stat),
	ONE("statm",      S_IRUGO, proc_pid_statm),
2949
	REG("maps",       S_IRUGO, proc_pid_maps_operations),
2950
#ifdef CONFIG_NUMA
2951
	REG("numa_maps",  S_IRUGO, proc_pid_numa_maps_operations),
2952
#endif
A
Alexey Dobriyan 已提交
2953 2954 2955 2956 2957 2958 2959
	REG("mem",        S_IRUSR|S_IWUSR, proc_mem_operations),
	LNK("cwd",        proc_cwd_link),
	LNK("root",       proc_root_link),
	LNK("exe",        proc_exe_link),
	REG("mounts",     S_IRUGO, proc_mounts_operations),
	REG("mountinfo",  S_IRUGO, proc_mountinfo_operations),
	REG("mountstats", S_IRUSR, proc_mountstats_operations),
2960
#ifdef CONFIG_PROC_PAGE_MONITOR
A
Alexey Dobriyan 已提交
2961
	REG("clear_refs", S_IWUSR, proc_clear_refs_operations),
2962
	REG("smaps",      S_IRUGO, proc_pid_smaps_operations),
2963
	REG("smaps_rollup", S_IRUGO, proc_pid_smaps_rollup_operations),
2964
	REG("pagemap",    S_IRUSR, proc_pagemap_operations),
2965 2966
#endif
#ifdef CONFIG_SECURITY
A
Alexey Dobriyan 已提交
2967
	DIR("attr",       S_IRUGO|S_IXUGO, proc_attr_dir_inode_operations, proc_attr_dir_operations),
2968 2969
#endif
#ifdef CONFIG_KALLSYMS
2970
	ONE("wchan",      S_IRUGO, proc_pid_wchan),
2971
#endif
K
Ken Chen 已提交
2972
#ifdef CONFIG_STACKTRACE
2973
	ONE("stack",      S_IRUSR, proc_pid_stack),
2974
#endif
2975
#ifdef CONFIG_SCHED_INFO
2976
	ONE("schedstat",  S_IRUGO, proc_pid_schedstat),
2977
#endif
A
Arjan van de Ven 已提交
2978
#ifdef CONFIG_LATENCYTOP
A
Alexey Dobriyan 已提交
2979
	REG("latency",  S_IRUGO, proc_lstats_operations),
A
Arjan van de Ven 已提交
2980
#endif
2981
#ifdef CONFIG_PROC_PID_CPUSET
Z
Zefan Li 已提交
2982
	ONE("cpuset",     S_IRUGO, proc_cpuset_show),
2983 2984
#endif
#ifdef CONFIG_CGROUPS
Z
Zefan Li 已提交
2985
	ONE("cgroup",  S_IRUGO, proc_cgroup_show),
2986
#endif
2987
	ONE("oom_score",  S_IRUGO, proc_oom_score),
2988
	REG("oom_adj",    S_IRUGO|S_IWUSR, proc_oom_adj_operations),
D
David Rientjes 已提交
2989
	REG("oom_score_adj", S_IRUGO|S_IWUSR, proc_oom_score_adj_operations),
2990
#ifdef CONFIG_AUDITSYSCALL
A
Alexey Dobriyan 已提交
2991 2992
	REG("loginuid",   S_IWUSR|S_IRUGO, proc_loginuid_operations),
	REG("sessionid",  S_IRUGO, proc_sessionid_operations),
2993
#endif
2994
#ifdef CONFIG_FAULT_INJECTION
A
Alexey Dobriyan 已提交
2995
	REG("make-it-fail", S_IRUGO|S_IWUSR, proc_fault_inject_operations),
2996
	REG("fail-nth", 0644, proc_fail_nth_operations),
2997
#endif
C
Christoph Hellwig 已提交
2998
#ifdef CONFIG_ELF_CORE
A
Alexey Dobriyan 已提交
2999
	REG("coredump_filter", S_IRUGO|S_IWUSR, proc_coredump_filter_operations),
3000
#endif
3001
#ifdef CONFIG_TASK_IO_ACCOUNTING
3002
	ONE("io",	S_IRUSR, proc_tgid_io_accounting),
3003
#endif
3004
#ifdef CONFIG_HARDWALL
3005
	ONE("hardwall",   S_IRUGO, proc_pid_hardwall),
3006
#endif
3007 3008 3009
#ifdef CONFIG_USER_NS
	REG("uid_map",    S_IRUGO|S_IWUSR, proc_uid_map_operations),
	REG("gid_map",    S_IRUGO|S_IWUSR, proc_gid_map_operations),
3010
	REG("projid_map", S_IRUGO|S_IWUSR, proc_projid_map_operations),
3011
	REG("setgroups",  S_IRUGO|S_IWUSR, proc_setgroups_operations),
3012
#endif
3013
#if defined(CONFIG_CHECKPOINT_RESTORE) && defined(CONFIG_POSIX_TIMERS)
3014 3015
	REG("timers",	  S_IRUGO, proc_timers_operations),
#endif
3016
	REG("timerslack_ns", S_IRUGO|S_IWUGO, proc_pid_set_timerslack_ns_operations),
3017 3018 3019
#ifdef CONFIG_LIVEPATCH
	ONE("patch_state",  S_IRUSR, proc_pid_patch_state),
#endif
3020
};
L
Linus Torvalds 已提交
3021

A
Al Viro 已提交
3022
static int proc_tgid_base_readdir(struct file *file, struct dir_context *ctx)
L
Linus Torvalds 已提交
3023
{
A
Al Viro 已提交
3024 3025
	return proc_pident_readdir(file, ctx,
				   tgid_base_stuff, ARRAY_SIZE(tgid_base_stuff));
L
Linus Torvalds 已提交
3026 3027
}

3028
static const struct file_operations proc_tgid_base_operations = {
L
Linus Torvalds 已提交
3029
	.read		= generic_read_dir,
3030 3031
	.iterate_shared	= proc_tgid_base_readdir,
	.llseek		= generic_file_llseek,
L
Linus Torvalds 已提交
3032 3033
};

A
Al Viro 已提交
3034 3035
static struct dentry *proc_tgid_base_lookup(struct inode *dir, struct dentry *dentry, unsigned int flags)
{
3036 3037
	return proc_pident_lookup(dir, dentry,
				  tgid_base_stuff, ARRAY_SIZE(tgid_base_stuff));
L
Linus Torvalds 已提交
3038 3039
}

3040
static const struct inode_operations proc_tgid_base_inode_operations = {
3041
	.lookup		= proc_tgid_base_lookup,
3042
	.getattr	= pid_getattr,
3043
	.setattr	= proc_setattr,
3044
	.permission	= proc_pid_permission,
L
Linus Torvalds 已提交
3045 3046
};

3047
static void proc_flush_task_mnt(struct vfsmount *mnt, pid_t pid, pid_t tgid)
L
Linus Torvalds 已提交
3048
{
3049
	struct dentry *dentry, *leader, *dir;
3050
	char buf[10 + 1];
3051 3052 3053
	struct qstr name;

	name.name = buf;
3054
	name.len = snprintf(buf, sizeof(buf), "%u", pid);
3055
	/* no ->d_hash() rejects on procfs */
3056
	dentry = d_hash_and_lookup(mnt->mnt_root, &name);
3057
	if (dentry) {
3058
		d_invalidate(dentry);
3059 3060
		dput(dentry);
	}
L
Linus Torvalds 已提交
3061

3062 3063 3064
	if (pid == tgid)
		return;

3065
	name.name = buf;
3066
	name.len = snprintf(buf, sizeof(buf), "%u", tgid);
3067
	leader = d_hash_and_lookup(mnt->mnt_root, &name);
3068 3069
	if (!leader)
		goto out;
L
Linus Torvalds 已提交
3070

3071 3072 3073 3074 3075 3076 3077
	name.name = "task";
	name.len = strlen(name.name);
	dir = d_hash_and_lookup(leader, &name);
	if (!dir)
		goto out_put_leader;

	name.name = buf;
3078
	name.len = snprintf(buf, sizeof(buf), "%u", pid);
3079 3080
	dentry = d_hash_and_lookup(dir, &name);
	if (dentry) {
3081
		d_invalidate(dentry);
3082
		dput(dentry);
L
Linus Torvalds 已提交
3083
	}
3084 3085 3086 3087 3088 3089

	dput(dir);
out_put_leader:
	dput(leader);
out:
	return;
L
Linus Torvalds 已提交
3090 3091
}

3092 3093 3094 3095 3096
/**
 * proc_flush_task -  Remove dcache entries for @task from the /proc dcache.
 * @task: task that should be flushed.
 *
 * When flushing dentries from proc, one needs to flush them from global
3097
 * proc (proc_mnt) and from all the namespaces' procs this task was seen
3098 3099 3100 3101 3102 3103 3104 3105 3106 3107 3108 3109 3110 3111 3112 3113 3114
 * in. This call is supposed to do all of this job.
 *
 * Looks in the dcache for
 * /proc/@pid
 * /proc/@tgid/task/@pid
 * if either directory is present flushes it and all of it'ts children
 * from the dcache.
 *
 * It is safe and reasonable to cache /proc entries for a task until
 * that task exits.  After that they just clog up the dcache with
 * useless entries, possibly causing useful dcache entries to be
 * flushed instead.  This routine is proved to flush those useless
 * dcache entries at process exit time.
 *
 * NOTE: This routine is just an optimization so it does not guarantee
 *       that no dcache entries will exist at process exit time it
 *       just makes it very unlikely that any will persist.
3115 3116 3117 3118
 */

void proc_flush_task(struct task_struct *task)
{
3119
	int i;
3120
	struct pid *pid, *tgid;
3121 3122 3123
	struct upid *upid;

	pid = task_pid(task);
3124
	tgid = task_tgid(task);
3125

3126
	for (i = 0; i <= pid->level; i++) {
3127 3128
		upid = &pid->numbers[i];
		proc_flush_task_mnt(upid->ns->proc_mnt, upid->nr,
3129
					tgid->numbers[i].nr);
3130
	}
3131 3132
}

3133 3134 3135
static int proc_pid_instantiate(struct inode *dir,
				   struct dentry * dentry,
				   struct task_struct *task, const void *ptr)
3136 3137 3138
{
	struct inode *inode;

3139
	inode = proc_pid_make_inode(dir->i_sb, task, S_IFDIR | S_IRUGO | S_IXUGO);
3140 3141 3142 3143 3144 3145
	if (!inode)
		goto out;

	inode->i_op = &proc_tgid_base_inode_operations;
	inode->i_fop = &proc_tgid_base_operations;
	inode->i_flags|=S_IMMUTABLE;
3146

3147
	set_nlink(inode, nlink_tgid);
3148

3149
	d_set_d_op(dentry, &pid_dentry_operations);
3150 3151 3152

	d_add(dentry, inode);
	/* Close the race of the process dying before we return the dentry */
3153
	if (pid_revalidate(dentry, 0))
3154
		return 0;
3155
out:
3156
	return -ENOENT;
3157 3158
}

A
Al Viro 已提交
3159
struct dentry *proc_pid_lookup(struct inode *dir, struct dentry * dentry, unsigned int flags)
L
Linus Torvalds 已提交
3160
{
A
Alexey Dobriyan 已提交
3161
	int result = -ENOENT;
L
Linus Torvalds 已提交
3162 3163
	struct task_struct *task;
	unsigned tgid;
3164
	struct pid_namespace *ns;
L
Linus Torvalds 已提交
3165

3166
	tgid = name_to_int(&dentry->d_name);
L
Linus Torvalds 已提交
3167 3168 3169
	if (tgid == ~0U)
		goto out;

3170
	ns = dentry->d_sb->s_fs_info;
3171
	rcu_read_lock();
3172
	task = find_task_by_pid_ns(tgid, ns);
L
Linus Torvalds 已提交
3173 3174
	if (task)
		get_task_struct(task);
3175
	rcu_read_unlock();
L
Linus Torvalds 已提交
3176 3177 3178
	if (!task)
		goto out;

3179
	result = proc_pid_instantiate(dir, dentry, task, NULL);
L
Linus Torvalds 已提交
3180 3181
	put_task_struct(task);
out:
3182
	return ERR_PTR(result);
L
Linus Torvalds 已提交
3183 3184 3185
}

/*
3186
 * Find the first task with tgid >= tgid
3187
 *
L
Linus Torvalds 已提交
3188
 */
3189 3190
struct tgid_iter {
	unsigned int tgid;
3191
	struct task_struct *task;
3192 3193 3194
};
static struct tgid_iter next_tgid(struct pid_namespace *ns, struct tgid_iter iter)
{
3195
	struct pid *pid;
L
Linus Torvalds 已提交
3196

3197 3198
	if (iter.task)
		put_task_struct(iter.task);
3199
	rcu_read_lock();
3200
retry:
3201 3202
	iter.task = NULL;
	pid = find_ge_pid(iter.tgid, ns);
3203
	if (pid) {
3204 3205
		iter.tgid = pid_nr_ns(pid, ns);
		iter.task = pid_task(pid, PIDTYPE_PID);
3206 3207 3208 3209 3210 3211 3212 3213 3214 3215 3216 3217
		/* What we to know is if the pid we have find is the
		 * pid of a thread_group_leader.  Testing for task
		 * being a thread_group_leader is the obvious thing
		 * todo but there is a window when it fails, due to
		 * the pid transfer logic in de_thread.
		 *
		 * So we perform the straight forward test of seeing
		 * if the pid we have found is the pid of a thread
		 * group leader, and don't worry if the task we have
		 * found doesn't happen to be a thread group leader.
		 * As we don't care in the case of readdir.
		 */
3218 3219
		if (!iter.task || !has_group_leader_pid(iter.task)) {
			iter.tgid += 1;
3220
			goto retry;
3221 3222
		}
		get_task_struct(iter.task);
3223
	}
3224
	rcu_read_unlock();
3225
	return iter;
L
Linus Torvalds 已提交
3226 3227
}

3228
#define TGID_OFFSET (FIRST_PROCESS_ENTRY + 2)
3229

L
Linus Torvalds 已提交
3230
/* for the /proc/ directory itself, after non-process stuff has been done */
A
Al Viro 已提交
3231
int proc_pid_readdir(struct file *file, struct dir_context *ctx)
L
Linus Torvalds 已提交
3232
{
3233
	struct tgid_iter iter;
A
Al Viro 已提交
3234
	struct pid_namespace *ns = file_inode(file)->i_sb->s_fs_info;
A
Al Viro 已提交
3235
	loff_t pos = ctx->pos;
L
Linus Torvalds 已提交
3236

3237
	if (pos >= PID_MAX_LIMIT + TGID_OFFSET)
A
Al Viro 已提交
3238
		return 0;
L
Linus Torvalds 已提交
3239

3240
	if (pos == TGID_OFFSET - 2) {
3241
		struct inode *inode = d_inode(ns->proc_self);
3242
		if (!dir_emit(ctx, "self", 4, inode->i_ino, DT_LNK))
A
Al Viro 已提交
3243
			return 0;
3244 3245 3246
		ctx->pos = pos = pos + 1;
	}
	if (pos == TGID_OFFSET - 1) {
3247
		struct inode *inode = d_inode(ns->proc_thread_self);
3248 3249 3250
		if (!dir_emit(ctx, "thread-self", 11, inode->i_ino, DT_LNK))
			return 0;
		ctx->pos = pos = pos + 1;
3251
	}
3252
	iter.tgid = pos - TGID_OFFSET;
3253 3254 3255 3256
	iter.task = NULL;
	for (iter = next_tgid(ns, iter);
	     iter.task;
	     iter.tgid += 1, iter = next_tgid(ns, iter)) {
3257
		char name[10 + 1];
A
Al Viro 已提交
3258
		int len;
3259 3260

		cond_resched();
3261
		if (!has_pid_permissions(ns, iter.task, HIDEPID_INVISIBLE))
A
Al Viro 已提交
3262
			continue;
3263

3264
		len = snprintf(name, sizeof(name), "%u", iter.tgid);
A
Al Viro 已提交
3265 3266 3267
		ctx->pos = iter.tgid + TGID_OFFSET;
		if (!proc_fill_cache(file, ctx, name, len,
				     proc_pid_instantiate, iter.task, NULL)) {
3268
			put_task_struct(iter.task);
A
Al Viro 已提交
3269
			return 0;
L
Linus Torvalds 已提交
3270
		}
3271
	}
A
Al Viro 已提交
3272
	ctx->pos = PID_MAX_LIMIT + TGID_OFFSET;
3273 3274
	return 0;
}
L
Linus Torvalds 已提交
3275

3276 3277 3278 3279 3280 3281 3282 3283 3284 3285 3286 3287 3288 3289 3290 3291 3292 3293 3294 3295 3296 3297 3298 3299 3300 3301 3302 3303 3304 3305 3306 3307 3308 3309 3310 3311 3312 3313
/*
 * proc_tid_comm_permission is a special permission function exclusively
 * used for the node /proc/<pid>/task/<tid>/comm.
 * It bypasses generic permission checks in the case where a task of the same
 * task group attempts to access the node.
 * The rationale behind this is that glibc and bionic access this node for
 * cross thread naming (pthread_set/getname_np(!self)). However, if
 * PR_SET_DUMPABLE gets set to 0 this node among others becomes uid=0 gid=0,
 * which locks out the cross thread naming implementation.
 * This function makes sure that the node is always accessible for members of
 * same thread group.
 */
static int proc_tid_comm_permission(struct inode *inode, int mask)
{
	bool is_same_tgroup;
	struct task_struct *task;

	task = get_proc_task(inode);
	if (!task)
		return -ESRCH;
	is_same_tgroup = same_thread_group(current, task);
	put_task_struct(task);

	if (likely(is_same_tgroup && !(mask & MAY_EXEC))) {
		/* This file (/proc/<pid>/task/<tid>/comm) can always be
		 * read or written by the members of the corresponding
		 * thread group.
		 */
		return 0;
	}

	return generic_permission(inode, mask);
}

static const struct inode_operations proc_tid_comm_inode_operations = {
		.permission = proc_tid_comm_permission,
};

3314 3315 3316
/*
 * Tasks
 */
3317
static const struct pid_entry tid_base_stuff[] = {
A
Alexey Dobriyan 已提交
3318
	DIR("fd",        S_IRUSR|S_IXUSR, proc_fd_inode_operations, proc_fd_operations),
J
Jerome Marchand 已提交
3319
	DIR("fdinfo",    S_IRUSR|S_IXUSR, proc_fdinfo_inode_operations, proc_fdinfo_operations),
3320
	DIR("ns",	 S_IRUSR|S_IXUGO, proc_ns_dir_inode_operations, proc_ns_dir_operations),
3321 3322 3323
#ifdef CONFIG_NET
	DIR("net",        S_IRUGO|S_IXUGO, proc_net_inode_operations, proc_net_operations),
#endif
A
Alexey Dobriyan 已提交
3324
	REG("environ",   S_IRUSR, proc_environ_operations),
A
Al Viro 已提交
3325
	REG("auxv",      S_IRUSR, proc_auxv_operations),
A
Alexey Dobriyan 已提交
3326
	ONE("status",    S_IRUGO, proc_pid_status),
3327
	ONE("personality", S_IRUSR, proc_pid_personality),
3328
	ONE("limits",	 S_IRUGO, proc_pid_limits),
I
Ingo Molnar 已提交
3329
#ifdef CONFIG_SCHED_DEBUG
A
Alexey Dobriyan 已提交
3330
	REG("sched",     S_IRUGO|S_IWUSR, proc_pid_sched_operations),
R
Roland McGrath 已提交
3331
#endif
3332 3333 3334
	NOD("comm",      S_IFREG|S_IRUGO|S_IWUSR,
			 &proc_tid_comm_inode_operations,
			 &proc_pid_set_comm_operations, {}),
R
Roland McGrath 已提交
3335
#ifdef CONFIG_HAVE_ARCH_TRACEHOOK
3336
	ONE("syscall",   S_IRUSR, proc_pid_syscall),
I
Ingo Molnar 已提交
3337
#endif
3338
	REG("cmdline",   S_IRUGO, proc_pid_cmdline_ops),
A
Alexey Dobriyan 已提交
3339 3340
	ONE("stat",      S_IRUGO, proc_tid_stat),
	ONE("statm",     S_IRUGO, proc_pid_statm),
3341
	REG("maps",      S_IRUGO, proc_tid_maps_operations),
3342
#ifdef CONFIG_PROC_CHILDREN
3343 3344
	REG("children",  S_IRUGO, proc_tid_children_operations),
#endif
3345
#ifdef CONFIG_NUMA
3346
	REG("numa_maps", S_IRUGO, proc_tid_numa_maps_operations),
3347
#endif
A
Alexey Dobriyan 已提交
3348 3349 3350 3351 3352 3353
	REG("mem",       S_IRUSR|S_IWUSR, proc_mem_operations),
	LNK("cwd",       proc_cwd_link),
	LNK("root",      proc_root_link),
	LNK("exe",       proc_exe_link),
	REG("mounts",    S_IRUGO, proc_mounts_operations),
	REG("mountinfo",  S_IRUGO, proc_mountinfo_operations),
3354
#ifdef CONFIG_PROC_PAGE_MONITOR
A
Alexey Dobriyan 已提交
3355
	REG("clear_refs", S_IWUSR, proc_clear_refs_operations),
3356
	REG("smaps",     S_IRUGO, proc_tid_smaps_operations),
3357
	REG("smaps_rollup", S_IRUGO, proc_pid_smaps_rollup_operations),
3358
	REG("pagemap",    S_IRUSR, proc_pagemap_operations),
3359 3360
#endif
#ifdef CONFIG_SECURITY
A
Alexey Dobriyan 已提交
3361
	DIR("attr",      S_IRUGO|S_IXUGO, proc_attr_dir_inode_operations, proc_attr_dir_operations),
3362 3363
#endif
#ifdef CONFIG_KALLSYMS
3364
	ONE("wchan",     S_IRUGO, proc_pid_wchan),
3365
#endif
K
Ken Chen 已提交
3366
#ifdef CONFIG_STACKTRACE
3367
	ONE("stack",      S_IRUSR, proc_pid_stack),
3368
#endif
3369
#ifdef CONFIG_SCHED_INFO
3370
	ONE("schedstat", S_IRUGO, proc_pid_schedstat),
3371
#endif
A
Arjan van de Ven 已提交
3372
#ifdef CONFIG_LATENCYTOP
A
Alexey Dobriyan 已提交
3373
	REG("latency",  S_IRUGO, proc_lstats_operations),
A
Arjan van de Ven 已提交
3374
#endif
3375
#ifdef CONFIG_PROC_PID_CPUSET
Z
Zefan Li 已提交
3376
	ONE("cpuset",    S_IRUGO, proc_cpuset_show),
3377 3378
#endif
#ifdef CONFIG_CGROUPS
Z
Zefan Li 已提交
3379
	ONE("cgroup",  S_IRUGO, proc_cgroup_show),
3380
#endif
3381
	ONE("oom_score", S_IRUGO, proc_oom_score),
3382
	REG("oom_adj",   S_IRUGO|S_IWUSR, proc_oom_adj_operations),
D
David Rientjes 已提交
3383
	REG("oom_score_adj", S_IRUGO|S_IWUSR, proc_oom_score_adj_operations),
3384
#ifdef CONFIG_AUDITSYSCALL
A
Alexey Dobriyan 已提交
3385
	REG("loginuid",  S_IWUSR|S_IRUGO, proc_loginuid_operations),
3386
	REG("sessionid",  S_IRUGO, proc_sessionid_operations),
3387
#endif
3388
#ifdef CONFIG_FAULT_INJECTION
A
Alexey Dobriyan 已提交
3389
	REG("make-it-fail", S_IRUGO|S_IWUSR, proc_fault_inject_operations),
3390
	REG("fail-nth", 0644, proc_fail_nth_operations),
3391
#endif
3392
#ifdef CONFIG_TASK_IO_ACCOUNTING
3393
	ONE("io",	S_IRUSR, proc_tid_io_accounting),
3394
#endif
3395
#ifdef CONFIG_HARDWALL
3396
	ONE("hardwall",   S_IRUGO, proc_pid_hardwall),
3397
#endif
3398 3399 3400
#ifdef CONFIG_USER_NS
	REG("uid_map",    S_IRUGO|S_IWUSR, proc_uid_map_operations),
	REG("gid_map",    S_IRUGO|S_IWUSR, proc_gid_map_operations),
3401
	REG("projid_map", S_IRUGO|S_IWUSR, proc_projid_map_operations),
3402
	REG("setgroups",  S_IRUGO|S_IWUSR, proc_setgroups_operations),
3403
#endif
3404 3405 3406
#ifdef CONFIG_LIVEPATCH
	ONE("patch_state",  S_IRUSR, proc_pid_patch_state),
#endif
3407 3408
};

A
Al Viro 已提交
3409
static int proc_tid_base_readdir(struct file *file, struct dir_context *ctx)
3410
{
A
Al Viro 已提交
3411 3412
	return proc_pident_readdir(file, ctx,
				   tid_base_stuff, ARRAY_SIZE(tid_base_stuff));
3413 3414
}

A
Al Viro 已提交
3415 3416
static struct dentry *proc_tid_base_lookup(struct inode *dir, struct dentry *dentry, unsigned int flags)
{
3417 3418
	return proc_pident_lookup(dir, dentry,
				  tid_base_stuff, ARRAY_SIZE(tid_base_stuff));
3419 3420
}

3421
static const struct file_operations proc_tid_base_operations = {
3422
	.read		= generic_read_dir,
3423 3424
	.iterate_shared	= proc_tid_base_readdir,
	.llseek		= generic_file_llseek,
3425 3426
};

3427
static const struct inode_operations proc_tid_base_inode_operations = {
3428 3429 3430 3431 3432
	.lookup		= proc_tid_base_lookup,
	.getattr	= pid_getattr,
	.setattr	= proc_setattr,
};

3433
static int proc_task_instantiate(struct inode *dir,
3434
	struct dentry *dentry, struct task_struct *task, const void *ptr)
3435 3436
{
	struct inode *inode;
3437
	inode = proc_pid_make_inode(dir->i_sb, task, S_IFDIR | S_IRUGO | S_IXUGO);
3438 3439 3440 3441 3442 3443

	if (!inode)
		goto out;
	inode->i_op = &proc_tid_base_inode_operations;
	inode->i_fop = &proc_tid_base_operations;
	inode->i_flags|=S_IMMUTABLE;
3444

3445
	set_nlink(inode, nlink_tid);
3446

3447
	d_set_d_op(dentry, &pid_dentry_operations);
3448 3449 3450

	d_add(dentry, inode);
	/* Close the race of the process dying before we return the dentry */
3451
	if (pid_revalidate(dentry, 0))
3452
		return 0;
3453
out:
3454
	return -ENOENT;
3455 3456
}

A
Al Viro 已提交
3457
static struct dentry *proc_task_lookup(struct inode *dir, struct dentry * dentry, unsigned int flags)
3458
{
3459
	int result = -ENOENT;
3460 3461 3462
	struct task_struct *task;
	struct task_struct *leader = get_proc_task(dir);
	unsigned tid;
3463
	struct pid_namespace *ns;
3464 3465 3466 3467

	if (!leader)
		goto out_no_task;

3468
	tid = name_to_int(&dentry->d_name);
3469 3470 3471
	if (tid == ~0U)
		goto out;

3472
	ns = dentry->d_sb->s_fs_info;
3473
	rcu_read_lock();
3474
	task = find_task_by_pid_ns(tid, ns);
3475 3476 3477 3478 3479
	if (task)
		get_task_struct(task);
	rcu_read_unlock();
	if (!task)
		goto out;
3480
	if (!same_thread_group(leader, task))
3481 3482
		goto out_drop_task;

3483
	result = proc_task_instantiate(dir, dentry, task, NULL);
3484 3485 3486 3487 3488
out_drop_task:
	put_task_struct(task);
out:
	put_task_struct(leader);
out_no_task:
3489
	return ERR_PTR(result);
3490 3491
}

3492 3493 3494 3495 3496 3497 3498 3499 3500 3501 3502 3503
/*
 * Find the first tid of a thread group to return to user space.
 *
 * Usually this is just the thread group leader, but if the users
 * buffer was too small or there was a seek into the middle of the
 * directory we have more work todo.
 *
 * In the case of a short read we start with find_task_by_pid.
 *
 * In the case of a seek we start with the leader and walk nr
 * threads past it.
 */
3504 3505
static struct task_struct *first_tid(struct pid *pid, int tid, loff_t f_pos,
					struct pid_namespace *ns)
3506
{
3507
	struct task_struct *pos, *task;
3508 3509 3510 3511
	unsigned long nr = f_pos;

	if (nr != f_pos)	/* 32bit overflow? */
		return NULL;
L
Linus Torvalds 已提交
3512

3513
	rcu_read_lock();
3514 3515 3516 3517 3518
	task = pid_task(pid, PIDTYPE_PID);
	if (!task)
		goto fail;

	/* Attempt to start with the tid of a thread */
3519
	if (tid && nr) {
3520
		pos = find_task_by_pid_ns(tid, ns);
3521
		if (pos && same_thread_group(pos, task))
O
Oleg Nesterov 已提交
3522
			goto found;
3523
	}
L
Linus Torvalds 已提交
3524

3525
	/* If nr exceeds the number of threads there is nothing todo */
3526
	if (nr >= get_nr_threads(task))
3527
		goto fail;
L
Linus Torvalds 已提交
3528

O
Oleg Nesterov 已提交
3529 3530
	/* If we haven't found our starting place yet start
	 * with the leader and walk nr threads forward.
3531
	 */
3532
	pos = task = task->group_leader;
3533
	do {
3534
		if (!nr--)
3535
			goto found;
3536
	} while_each_thread(task, pos);
3537 3538 3539
fail:
	pos = NULL;
	goto out;
O
Oleg Nesterov 已提交
3540 3541 3542
found:
	get_task_struct(pos);
out:
3543
	rcu_read_unlock();
3544 3545 3546 3547 3548 3549 3550 3551 3552 3553 3554
	return pos;
}

/*
 * Find the next thread in the thread list.
 * Return NULL if there is an error or no next thread.
 *
 * The reference to the input task_struct is released.
 */
static struct task_struct *next_tid(struct task_struct *start)
{
O
Oleg Nesterov 已提交
3555
	struct task_struct *pos = NULL;
3556
	rcu_read_lock();
O
Oleg Nesterov 已提交
3557
	if (pid_alive(start)) {
3558
		pos = next_thread(start);
O
Oleg Nesterov 已提交
3559 3560 3561 3562 3563
		if (thread_group_leader(pos))
			pos = NULL;
		else
			get_task_struct(pos);
	}
3564
	rcu_read_unlock();
3565 3566
	put_task_struct(start);
	return pos;
L
Linus Torvalds 已提交
3567 3568 3569
}

/* for the /proc/TGID/task/ directories */
A
Al Viro 已提交
3570
static int proc_task_readdir(struct file *file, struct dir_context *ctx)
L
Linus Torvalds 已提交
3571
{
3572 3573
	struct inode *inode = file_inode(file);
	struct task_struct *task;
3574
	struct pid_namespace *ns;
A
Al Viro 已提交
3575
	int tid;
L
Linus Torvalds 已提交
3576

3577
	if (proc_inode_is_dead(inode))
A
Al Viro 已提交
3578
		return -ENOENT;
L
Linus Torvalds 已提交
3579

A
Al Viro 已提交
3580
	if (!dir_emit_dots(file, ctx))
3581
		return 0;
L
Linus Torvalds 已提交
3582

3583 3584 3585
	/* f_version caches the tgid value that the last readdir call couldn't
	 * return. lseek aka telldir automagically resets f_version to 0.
	 */
A
Al Viro 已提交
3586
	ns = inode->i_sb->s_fs_info;
A
Al Viro 已提交
3587 3588
	tid = (int)file->f_version;
	file->f_version = 0;
3589
	for (task = first_tid(proc_pid(inode), tid, ctx->pos - 2, ns);
3590
	     task;
A
Al Viro 已提交
3591
	     task = next_tid(task), ctx->pos++) {
3592
		char name[10 + 1];
A
Al Viro 已提交
3593
		int len;
3594
		tid = task_pid_nr_ns(task, ns);
3595
		len = snprintf(name, sizeof(name), "%u", tid);
A
Al Viro 已提交
3596 3597
		if (!proc_fill_cache(file, ctx, name, len,
				proc_task_instantiate, task, NULL)) {
3598 3599
			/* returning this tgid failed, save it as the first
			 * pid for the next readir call */
A
Al Viro 已提交
3600
			file->f_version = (u64)tid;
3601
			put_task_struct(task);
L
Linus Torvalds 已提交
3602
			break;
3603
		}
L
Linus Torvalds 已提交
3604
	}
3605

A
Al Viro 已提交
3606
	return 0;
L
Linus Torvalds 已提交
3607
}
3608

3609 3610
static int proc_task_getattr(const struct path *path, struct kstat *stat,
			     u32 request_mask, unsigned int query_flags)
3611
{
3612
	struct inode *inode = d_inode(path->dentry);
3613
	struct task_struct *p = get_proc_task(inode);
3614 3615
	generic_fillattr(inode, stat);

3616 3617 3618
	if (p) {
		stat->nlink += get_nr_threads(p);
		put_task_struct(p);
3619 3620 3621 3622
	}

	return 0;
}
3623

3624
static const struct inode_operations proc_task_inode_operations = {
3625 3626 3627
	.lookup		= proc_task_lookup,
	.getattr	= proc_task_getattr,
	.setattr	= proc_setattr,
3628
	.permission	= proc_pid_permission,
3629 3630
};

3631
static const struct file_operations proc_task_operations = {
3632
	.read		= generic_read_dir,
3633 3634
	.iterate_shared	= proc_task_readdir,
	.llseek		= generic_file_llseek,
3635
};
3636 3637 3638 3639 3640 3641

void __init set_proc_pid_nlink(void)
{
	nlink_tid = pid_entry_nlink(tid_base_stuff, ARRAY_SIZE(tid_base_stuff));
	nlink_tgid = pid_entry_nlink(tgid_base_stuff, ARRAY_SIZE(tgid_base_stuff));
}