base.c 83.1 KB
Newer Older
L
Linus Torvalds 已提交
1 2 3 4 5 6 7 8 9 10 11 12 13
/*
 *  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
14 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
 *
 *
 *  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
48 49 50 51 52 53 54 55
 */

#include <asm/uaccess.h>

#include <linux/errno.h>
#include <linux/time.h>
#include <linux/proc_fs.h>
#include <linux/stat.h>
56
#include <linux/task_io_accounting_ops.h>
L
Linus Torvalds 已提交
57
#include <linux/init.h>
58
#include <linux/capability.h>
L
Linus Torvalds 已提交
59
#include <linux/file.h>
A
Al Viro 已提交
60
#include <linux/fdtable.h>
L
Linus Torvalds 已提交
61 62 63
#include <linux/string.h>
#include <linux/seq_file.h>
#include <linux/namei.h>
64
#include <linux/mnt_namespace.h>
L
Linus Torvalds 已提交
65
#include <linux/mm.h>
D
David Rientjes 已提交
66
#include <linux/swap.h>
67
#include <linux/rcupdate.h>
L
Linus Torvalds 已提交
68
#include <linux/kallsyms.h>
K
Ken Chen 已提交
69
#include <linux/stacktrace.h>
70
#include <linux/resource.h>
K
Kees Cook 已提交
71
#include <linux/module.h>
L
Linus Torvalds 已提交
72 73 74
#include <linux/mount.h>
#include <linux/security.h>
#include <linux/ptrace.h>
75
#include <linux/tracehook.h>
A
Andrew Morton 已提交
76
#include <linux/printk.h>
77
#include <linux/cgroup.h>
L
Linus Torvalds 已提交
78 79
#include <linux/cpuset.h>
#include <linux/audit.h>
A
Al Viro 已提交
80
#include <linux/poll.h>
81
#include <linux/nsproxy.h>
82
#include <linux/oom.h>
83
#include <linux/elf.h>
84
#include <linux/pid_namespace.h>
85
#include <linux/user_namespace.h>
86
#include <linux/fs_struct.h>
87
#include <linux/slab.h>
88
#include <linux/flex_array.h>
89
#include <linux/posix-timers.h>
90 91 92
#ifdef CONFIG_HARDWALL
#include <asm/hardwall.h>
#endif
93
#include <trace/events/oom.h>
L
Linus Torvalds 已提交
94
#include "internal.h"
95
#include "fd.h"
L
Linus Torvalds 已提交
96

97 98 99 100 101 102 103 104 105 106
/* 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.
 */

L
Linus Torvalds 已提交
107
struct pid_entry {
108
	const char *name;
109
	int len;
A
Al Viro 已提交
110
	umode_t mode;
111
	const struct inode_operations *iop;
112
	const struct file_operations *fop;
113
	union proc_op op;
L
Linus Torvalds 已提交
114 115
};

116
#define NOD(NAME, MODE, IOP, FOP, OP) {			\
117
	.name = (NAME),					\
118
	.len  = sizeof(NAME) - 1,			\
119 120 121 122 123 124
	.mode = MODE,					\
	.iop  = IOP,					\
	.fop  = FOP,					\
	.op   = OP,					\
}

A
Alexey Dobriyan 已提交
125 126 127
#define DIR(NAME, MODE, iops, fops)	\
	NOD(NAME, (S_IFDIR|(MODE)), &iops, &fops, {} )
#define LNK(NAME, get_link)					\
128
	NOD(NAME, (S_IFLNK|S_IRWXUGO),				\
129
		&proc_pid_link_inode_operations, NULL,		\
A
Alexey Dobriyan 已提交
130 131 132 133
		{ .proc_get_link = get_link } )
#define REG(NAME, MODE, fops)				\
	NOD(NAME, (S_IFREG|(MODE)), NULL, &fops, {})
#define ONE(NAME, MODE, show)				\
134 135
	NOD(NAME, (S_IFREG|(MODE)), 			\
		NULL, &proc_single_file_operations,	\
A
Alexey Dobriyan 已提交
136
		{ .proc_show = show } )
L
Linus Torvalds 已提交
137

138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156
/*
 * Count the number of hardlinks for the pid_entry table, excluding the .
 * and .. links.
 */
static unsigned int pid_entry_count_dirs(const struct pid_entry *entries,
	unsigned int n)
{
	unsigned int i;
	unsigned int count;

	count = 0;
	for (i = 0; i < n; ++i) {
		if (S_ISDIR(entries[i].mode))
			++count;
	}

	return count;
}

157
static int get_task_root(struct task_struct *task, struct path *root)
L
Linus Torvalds 已提交
158
{
H
Hugh Dickins 已提交
159 160
	int result = -ENOENT;

161
	task_lock(task);
162 163
	if (task->fs) {
		get_fs_root(task->fs, root);
H
Hugh Dickins 已提交
164 165
		result = 0;
	}
166
	task_unlock(task);
H
Hugh Dickins 已提交
167
	return result;
168 169
}

170
static int proc_cwd_link(struct dentry *dentry, struct path *path)
171
{
172
	struct task_struct *task = get_proc_task(d_inode(dentry));
173
	int result = -ENOENT;
174 175

	if (task) {
176 177 178 179 180 181
		task_lock(task);
		if (task->fs) {
			get_fs_pwd(task->fs, path);
			result = 0;
		}
		task_unlock(task);
182 183
		put_task_struct(task);
	}
L
Linus Torvalds 已提交
184 185 186
	return result;
}

187
static int proc_root_link(struct dentry *dentry, struct path *path)
L
Linus Torvalds 已提交
188
{
189
	struct task_struct *task = get_proc_task(d_inode(dentry));
L
Linus Torvalds 已提交
190
	int result = -ENOENT;
191 192

	if (task) {
193
		result = get_task_root(task, path);
194 195
		put_task_struct(task);
	}
L
Linus Torvalds 已提交
196 197 198
	return result;
}

199 200
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
201
{
202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245
	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;
	}

	page = (char *)__get_free_page(GFP_TEMPORARY);
	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;

246 247 248 249 250
	/* Empty ARGV. */
	if (len1 == 0) {
		rv = 0;
		goto out_free_page;
	}
251
	/*
252 253
	 * Inherently racy -- command line shares address space
	 * with code and data.
254
	 */
255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395
	rv = access_remote_vm(mm, arg_end - 1, &c, 1, 0);
	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);
			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;

			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 maybe
		 * extends into ENVP.
		 */
		if (len1 + len2 <= *pos)
			goto skip_argv_envp;
		if (len1 <= *pos)
			goto skip_argv;

		p = arg_start + *pos;
		len = len1 - *pos;
		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;
		}
skip_argv:
		/*
		 * Command line (1 string) occupies ARGV and
		 * extends into ENVP.
		 */
		if (len1 <= *pos) {
			p = env_start + *pos - len1;
			len = len1 + len2 - *pos;
		} else {
			p = env_start;
			len = len2;
		}
		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;

			/* Find EOS. */
			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;
		}
skip_argv_envp:
		;
	}

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
396 397
}

398 399 400 401 402
static const struct file_operations proc_pid_cmdline_ops = {
	.read	= proc_pid_cmdline_read,
	.llseek	= generic_file_llseek,
};

403 404
static int proc_pid_auxv(struct seq_file *m, struct pid_namespace *ns,
			 struct pid *pid, struct task_struct *task)
L
Linus Torvalds 已提交
405
{
406
	struct mm_struct *mm = mm_access(task, PTRACE_MODE_READ_FSCREDS);
407
	if (mm && !IS_ERR(mm)) {
L
Linus Torvalds 已提交
408
		unsigned int nwords = 0;
H
Hannes Eder 已提交
409
		do {
L
Linus Torvalds 已提交
410
			nwords += 2;
H
Hannes Eder 已提交
411
		} while (mm->saved_auxv[nwords - 2] != 0); /* AT_NULL */
412
		seq_write(m, mm->saved_auxv, nwords * sizeof(mm->saved_auxv[0]));
L
Linus Torvalds 已提交
413
		mmput(mm);
414 415 416
		return 0;
	} else
		return PTR_ERR(mm);
L
Linus Torvalds 已提交
417 418 419 420 421 422 423 424
}


#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.
 */
425 426
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
427
{
A
Alexey Dobriyan 已提交
428
	unsigned long wchan;
429
	char symname[KSYM_NAME_LEN];
L
Linus Torvalds 已提交
430 431 432

	wchan = get_wchan(task);

433 434
	if (wchan && ptrace_may_access(task, PTRACE_MODE_READ_FSCREDS)
			&& !lookup_symbol_name(wchan, symname))
435
		seq_printf(m, "%s", symname);
436
	else
437
		seq_putc(m, '0');
438 439

	return 0;
L
Linus Torvalds 已提交
440 441 442
}
#endif /* CONFIG_KALLSYMS */

443 444 445 446 447
static int lock_trace(struct task_struct *task)
{
	int err = mutex_lock_killable(&task->signal->cred_guard_mutex);
	if (err)
		return err;
448
	if (!ptrace_may_access(task, PTRACE_MODE_ATTACH_FSCREDS)) {
449 450 451 452 453 454 455 456 457 458 459
		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
460 461 462 463 464 465 466 467 468
#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;
469
	int err;
K
Ken Chen 已提交
470 471 472 473 474 475 476 477 478 479 480
	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;

481 482 483 484 485
	err = lock_trace(task);
	if (!err) {
		save_stack_trace_tsk(task, &trace);

		for (i = 0; i < trace.nr_entries; i++) {
486
			seq_printf(m, "[<%pK>] %pS\n",
487 488 489
				   (void *)entries[i], (void *)entries[i]);
		}
		unlock_trace(task);
K
Ken Chen 已提交
490 491 492
	}
	kfree(entries);

493
	return err;
K
Ken Chen 已提交
494 495 496
}
#endif

497
#ifdef CONFIG_SCHED_INFO
L
Linus Torvalds 已提交
498 499 500
/*
 * Provides /proc/PID/schedstat
 */
501 502
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
503
{
504 505 506 507
	if (unlikely(!sched_info_on()))
		seq_printf(m, "0 0 0\n");
	else
		seq_printf(m, "%llu %llu %lu\n",
508 509 510 511 512
		   (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
513 514 515
}
#endif

A
Arjan van de Ven 已提交
516 517 518 519
#ifdef CONFIG_LATENCYTOP
static int lstats_show_proc(struct seq_file *m, void *v)
{
	int i;
520 521
	struct inode *inode = m->private;
	struct task_struct *task = get_proc_task(inode);
A
Arjan van de Ven 已提交
522

523 524 525
	if (!task)
		return -ESRCH;
	seq_puts(m, "Latency Top version : v0.1\n");
A
Arjan van de Ven 已提交
526
	for (i = 0; i < 32; i++) {
527 528
		struct latency_record *lr = &task->latency_record[i];
		if (lr->backtrace[0]) {
A
Arjan van de Ven 已提交
529
			int q;
530 531
			seq_printf(m, "%i %li %li",
				   lr->count, lr->time, lr->max);
A
Arjan van de Ven 已提交
532
			for (q = 0; q < LT_BACKTRACEDEPTH; q++) {
533 534
				unsigned long bt = lr->backtrace[q];
				if (!bt)
A
Arjan van de Ven 已提交
535
					break;
536
				if (bt == ULONG_MAX)
A
Arjan van de Ven 已提交
537
					break;
538
				seq_printf(m, " %ps", (void *)bt);
A
Arjan van de Ven 已提交
539
			}
540
			seq_putc(m, '\n');
A
Arjan van de Ven 已提交
541 542 543
		}

	}
544
	put_task_struct(task);
A
Arjan van de Ven 已提交
545 546 547 548 549
	return 0;
}

static int lstats_open(struct inode *inode, struct file *file)
{
550
	return single_open(file, lstats_show_proc, inode);
551 552
}

A
Arjan van de Ven 已提交
553 554 555
static ssize_t lstats_write(struct file *file, const char __user *buf,
			    size_t count, loff_t *offs)
{
A
Al Viro 已提交
556
	struct task_struct *task = get_proc_task(file_inode(file));
A
Arjan van de Ven 已提交
557

558 559
	if (!task)
		return -ESRCH;
A
Arjan van de Ven 已提交
560
	clear_all_latency_tracing(task);
561
	put_task_struct(task);
A
Arjan van de Ven 已提交
562 563 564 565 566 567 568 569 570

	return count;
}

static const struct file_operations proc_lstats_operations = {
	.open		= lstats_open,
	.read		= seq_read,
	.write		= lstats_write,
	.llseek		= seq_lseek,
571
	.release	= single_release,
A
Arjan van de Ven 已提交
572 573 574 575
};

#endif

576 577
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
578
{
579
	unsigned long totalpages = totalram_pages + total_swap_pages;
580
	unsigned long points = 0;
L
Linus Torvalds 已提交
581

582
	read_lock(&tasklist_lock);
583
	if (pid_alive(task))
584 585
		points = oom_badness(task, NULL, NULL, totalpages) *
						1000 / totalpages;
586
	read_unlock(&tasklist_lock);
587 588 589
	seq_printf(m, "%lu\n", points);

	return 0;
L
Linus Torvalds 已提交
590 591
}

592
struct limit_names {
593 594
	const char *name;
	const char *unit;
595 596 597
};

static const struct limit_names lnames[RLIM_NLIMITS] = {
598
	[RLIMIT_CPU] = {"Max cpu time", "seconds"},
599 600 601 602 603 604 605 606 607 608 609 610 611 612
	[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},
613
	[RLIMIT_RTTIME] = {"Max realtime timeout", "us"},
614 615 616
};

/* Display limits for a process */
617 618
static int proc_pid_limits(struct seq_file *m, struct pid_namespace *ns,
			   struct pid *pid, struct task_struct *task)
619 620 621 622 623 624
{
	unsigned int i;
	unsigned long flags;

	struct rlimit rlim[RLIM_NLIMITS];

625
	if (!lock_task_sighand(task, &flags))
626 627 628 629 630 631 632
		return 0;
	memcpy(rlim, task->signal->rlim, sizeof(struct rlimit) * RLIM_NLIMITS);
	unlock_task_sighand(task, &flags);

	/*
	 * print the file header
	 */
633
       seq_printf(m, "%-25s %-20s %-20s %-10s\n",
634
		  "Limit", "Soft Limit", "Hard Limit", "Units");
635 636 637

	for (i = 0; i < RLIM_NLIMITS; i++) {
		if (rlim[i].rlim_cur == RLIM_INFINITY)
638
			seq_printf(m, "%-25s %-20s ",
639
				   lnames[i].name, "unlimited");
640
		else
641
			seq_printf(m, "%-25s %-20lu ",
642
				   lnames[i].name, rlim[i].rlim_cur);
643 644

		if (rlim[i].rlim_max == RLIM_INFINITY)
645
			seq_printf(m, "%-20s ", "unlimited");
646
		else
647
			seq_printf(m, "%-20lu ", rlim[i].rlim_max);
648 649

		if (lnames[i].unit)
650
			seq_printf(m, "%-10s\n", lnames[i].unit);
651
		else
652
			seq_putc(m, '\n');
653 654
	}

655
	return 0;
656 657
}

R
Roland McGrath 已提交
658
#ifdef CONFIG_HAVE_ARCH_TRACEHOOK
659 660
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
661 662 663
{
	long nr;
	unsigned long args[6], sp, pc;
664 665 666
	int res;

	res = lock_trace(task);
667 668
	if (res)
		return res;
R
Roland McGrath 已提交
669 670

	if (task_current_syscall(task, &nr, args, 6, &sp, &pc))
671
		seq_puts(m, "running\n");
672
	else if (nr < 0)
673
		seq_printf(m, "%ld 0x%lx 0x%lx\n", nr, sp, pc);
674
	else
675
		seq_printf(m,
R
Roland McGrath 已提交
676 677 678 679
		       "%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);
680
	unlock_trace(task);
681 682

	return 0;
R
Roland McGrath 已提交
683 684 685
}
#endif /* CONFIG_HAVE_ARCH_TRACEHOOK */

L
Linus Torvalds 已提交
686 687 688 689 690
/************************************************************************/
/*                       Here the fs part begins                        */
/************************************************************************/

/* permission checks */
691
static int proc_fd_access_allowed(struct inode *inode)
L
Linus Torvalds 已提交
692
{
693 694
	struct task_struct *task;
	int allowed = 0;
695 696 697
	/* 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.
698 699
	 */
	task = get_proc_task(inode);
700
	if (task) {
701
		allowed = ptrace_may_access(task, PTRACE_MODE_READ_FSCREDS);
702
		put_task_struct(task);
703
	}
704
	return allowed;
L
Linus Torvalds 已提交
705 706
}

707
int proc_setattr(struct dentry *dentry, struct iattr *attr)
708 709
{
	int error;
710
	struct inode *inode = d_inode(dentry);
711 712 713 714 715

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

	error = inode_change_ok(inode, attr);
C
Christoph Hellwig 已提交
716 717 718 719 720 721
	if (error)
		return error;

	setattr_copy(inode, attr);
	mark_inode_dirty(inode);
	return 0;
722 723
}

724 725 726 727 728 729 730 731 732 733 734 735
/*
 * 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;
736
	return ptrace_may_access(task, PTRACE_MODE_READ_FSCREDS);
737 738 739 740 741 742 743 744 745 746
}


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);
747 748
	if (!task)
		return -ESRCH;
749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769
	has_perms = has_pid_permissions(pid, task, 1);
	put_task_struct(task);

	if (!has_perms) {
		if (pid->hide_pid == 2) {
			/*
			 * 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);
}



770
static const struct inode_operations proc_def_inode_operations = {
771 772 773
	.setattr	= proc_setattr,
};

774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795
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
796
	return single_open(filp, proc_single_show, inode);
797 798 799 800 801 802 803 804 805
}

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
806 807

struct mm_struct *proc_mem_open(struct inode *inode, unsigned int mode)
L
Linus Torvalds 已提交
808
{
O
Oleg Nesterov 已提交
809 810
	struct task_struct *task = get_proc_task(inode);
	struct mm_struct *mm = ERR_PTR(-ESRCH);
811

O
Oleg Nesterov 已提交
812
	if (task) {
813
		mm = mm_access(task, mode | PTRACE_MODE_FSCREDS);
O
Oleg Nesterov 已提交
814
		put_task_struct(task);
815

O
Oleg Nesterov 已提交
816 817 818 819 820 821 822 823 824 825 826 827 828 829
		if (!IS_ERR_OR_NULL(mm)) {
			/* ensure this mm_struct can't be freed */
			atomic_inc(&mm->mm_count);
			/* 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);
830 831 832 833 834

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

	file->private_data = mm;
L
Linus Torvalds 已提交
835 836 837
	return 0;
}

838 839
static int mem_open(struct inode *inode, struct file *file)
{
840 841 842 843 844 845
	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;
846 847
}

848 849
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
850
{
851
	struct mm_struct *mm = file->private_data;
852 853
	unsigned long addr = *ppos;
	ssize_t copied;
L
Linus Torvalds 已提交
854 855
	char *page;

856 857
	if (!mm)
		return 0;
858

859 860
	page = (char *)__get_free_page(GFP_TEMPORARY);
	if (!page)
861
		return -ENOMEM;
L
Linus Torvalds 已提交
862

863
	copied = 0;
864 865 866
	if (!atomic_inc_not_zero(&mm->mm_users))
		goto free;

L
Linus Torvalds 已提交
867
	while (count > 0) {
868
		int this_len = min_t(int, count, PAGE_SIZE);
L
Linus Torvalds 已提交
869

870
		if (write && copy_from_user(page, buf, this_len)) {
L
Linus Torvalds 已提交
871 872 873
			copied = -EFAULT;
			break;
		}
874 875 876

		this_len = access_remote_vm(mm, addr, page, this_len, write);
		if (!this_len) {
L
Linus Torvalds 已提交
877 878 879 880
			if (!copied)
				copied = -EIO;
			break;
		}
881 882 883 884 885 886 887 888 889 890

		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
891
	}
892
	*ppos = addr;
893

894 895
	mmput(mm);
free:
896
	free_page((unsigned long) page);
L
Linus Torvalds 已提交
897 898 899
	return copied;
}

900 901 902 903 904 905 906 907 908 909 910 911
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);
}

912
loff_t mem_lseek(struct file *file, loff_t offset, int orig)
L
Linus Torvalds 已提交
913 914 915 916 917 918 919 920 921 922 923 924 925 926 927
{
	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;
}

928 929 930
static int mem_release(struct inode *inode, struct file *file)
{
	struct mm_struct *mm = file->private_data;
931
	if (mm)
932
		mmdrop(mm);
933 934 935
	return 0;
}

936
static const struct file_operations proc_mem_operations = {
L
Linus Torvalds 已提交
937 938 939 940
	.llseek		= mem_lseek,
	.read		= mem_read,
	.write		= mem_write,
	.open		= mem_open,
941
	.release	= mem_release,
L
Linus Torvalds 已提交
942 943
};

944 945 946 947 948
static int environ_open(struct inode *inode, struct file *file)
{
	return __mem_open(inode, file, PTRACE_MODE_READ);
}

949 950 951 952 953
static ssize_t environ_read(struct file *file, char __user *buf,
			size_t count, loff_t *ppos)
{
	char *page;
	unsigned long src = *ppos;
954 955
	int ret = 0;
	struct mm_struct *mm = file->private_data;
956
	unsigned long env_start, env_end;
957

958 959
	/* Ensure the process spawned far enough to have an environment. */
	if (!mm || !mm->env_end)
960
		return 0;
961 962 963

	page = (char *)__get_free_page(GFP_TEMPORARY);
	if (!page)
964
		return -ENOMEM;
965

A
Al Viro 已提交
966
	ret = 0;
967 968
	if (!atomic_inc_not_zero(&mm->mm_users))
		goto free;
969 970 971 972 973 974

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

975
	while (count > 0) {
976 977
		size_t this_len, max_len;
		int retval;
978

979
		if (src >= (env_end - env_start))
980 981
			break;

982
		this_len = env_end - (env_start + src);
983 984 985

		max_len = min_t(size_t, PAGE_SIZE, count);
		this_len = min(max_len, this_len);
986

987
		retval = access_remote_vm(mm, (env_start + src),
988 989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006
			page, this_len, 0);

		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);
1007 1008

free:
1009 1010 1011 1012 1013
	free_page((unsigned long) page);
	return ret;
}

static const struct file_operations proc_environ_operations = {
1014
	.open		= environ_open,
1015
	.read		= environ_read,
1016
	.llseek		= generic_file_llseek,
1017
	.release	= mem_release,
1018 1019
};

1020 1021 1022
static ssize_t oom_adj_read(struct file *file, char __user *buf, size_t count,
			    loff_t *ppos)
{
A
Al Viro 已提交
1023
	struct task_struct *task = get_proc_task(file_inode(file));
1024 1025 1026 1027 1028 1029
	char buffer[PROC_NUMBUF];
	int oom_adj = OOM_ADJUST_MIN;
	size_t len;

	if (!task)
		return -ESRCH;
1030 1031 1032 1033 1034
	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;
1035 1036 1037 1038 1039
	put_task_struct(task);
	len = snprintf(buffer, sizeof(buffer), "%d\n", oom_adj);
	return simple_read_from_buffer(buf, count, ppos, buffer, len);
}

1040 1041 1042
static int __set_oom_adj(struct file *file, int oom_adj, bool legacy)
{
	static DEFINE_MUTEX(oom_adj_mutex);
1043
	struct mm_struct *mm = NULL;
1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072
	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;
		}
	}

1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089
	/*
	 * 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;
				atomic_inc(&mm->mm_count);
			}
			task_unlock(p);
		}
	}

1090 1091 1092 1093
	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);
1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120 1121

	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);
	}
1122 1123 1124 1125 1126
err_unlock:
	mutex_unlock(&oom_adj_mutex);
	put_task_struct(task);
	return err;
}
1127

1128 1129 1130 1131 1132 1133 1134 1135 1136 1137
/*
 * /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.
 */
1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151 1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170
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;

1171
	err = __set_oom_adj(file, oom_adj, true);
1172 1173 1174 1175 1176 1177 1178 1179 1180 1181
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
1182 1183 1184
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
1185
	struct task_struct *task = get_proc_task(file_inode(file));
D
David Rientjes 已提交
1186
	char buffer[PROC_NUMBUF];
1187
	short oom_score_adj = OOM_SCORE_ADJ_MIN;
D
David Rientjes 已提交
1188 1189 1190 1191
	size_t len;

	if (!task)
		return -ESRCH;
1192
	oom_score_adj = task->signal->oom_score_adj;
D
David Rientjes 已提交
1193
	put_task_struct(task);
1194
	len = snprintf(buffer, sizeof(buffer), "%hd\n", oom_score_adj);
D
David Rientjes 已提交
1195 1196 1197 1198 1199 1200 1201
	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
1202
	int oom_score_adj;
D
David Rientjes 已提交
1203 1204 1205 1206 1207
	int err;

	memset(buffer, 0, sizeof(buffer));
	if (count > sizeof(buffer) - 1)
		count = sizeof(buffer) - 1;
1208 1209 1210 1211
	if (copy_from_user(buffer, buf, count)) {
		err = -EFAULT;
		goto out;
	}
D
David Rientjes 已提交
1212

A
Alexey Dobriyan 已提交
1213
	err = kstrtoint(strstrip(buffer), 0, &oom_score_adj);
D
David Rientjes 已提交
1214
	if (err)
1215
		goto out;
D
David Rientjes 已提交
1216
	if (oom_score_adj < OOM_SCORE_ADJ_MIN ||
1217 1218 1219 1220
			oom_score_adj > OOM_SCORE_ADJ_MAX) {
		err = -EINVAL;
		goto out;
	}
D
David Rientjes 已提交
1221

1222
	err = __set_oom_adj(file, oom_score_adj, false);
1223 1224
out:
	return err < 0 ? err : count;
D
David Rientjes 已提交
1225 1226 1227 1228 1229
}

static const struct file_operations proc_oom_score_adj_operations = {
	.read		= oom_score_adj_read,
	.write		= oom_score_adj_write,
1230
	.llseek		= default_llseek,
D
David Rientjes 已提交
1231 1232
};

L
Linus Torvalds 已提交
1233 1234 1235 1236 1237
#ifdef CONFIG_AUDITSYSCALL
#define TMPBUFLEN 21
static ssize_t proc_loginuid_read(struct file * file, char __user * buf,
				  size_t count, loff_t *ppos)
{
A
Al Viro 已提交
1238
	struct inode * inode = file_inode(file);
1239
	struct task_struct *task = get_proc_task(inode);
L
Linus Torvalds 已提交
1240 1241 1242
	ssize_t length;
	char tmpbuf[TMPBUFLEN];

1243 1244
	if (!task)
		return -ESRCH;
L
Linus Torvalds 已提交
1245
	length = scnprintf(tmpbuf, TMPBUFLEN, "%u",
1246 1247
			   from_kuid(file->f_cred->user_ns,
				     audit_get_loginuid(task)));
1248
	put_task_struct(task);
L
Linus Torvalds 已提交
1249 1250 1251 1252 1253 1254
	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
1255
	struct inode * inode = file_inode(file);
L
Linus Torvalds 已提交
1256
	uid_t loginuid;
1257
	kuid_t kloginuid;
1258
	int rv;
L
Linus Torvalds 已提交
1259

1260 1261 1262
	rcu_read_lock();
	if (current != pid_task(proc_pid(inode), PIDTYPE_PID)) {
		rcu_read_unlock();
L
Linus Torvalds 已提交
1263
		return -EPERM;
1264 1265
	}
	rcu_read_unlock();
L
Linus Torvalds 已提交
1266 1267 1268 1269 1270 1271

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

1272 1273 1274
	rv = kstrtou32_from_user(buf, count, 10, &loginuid);
	if (rv < 0)
		return rv;
1275 1276 1277 1278 1279 1280

	/* 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);
1281 1282
		if (!uid_valid(kloginuid))
			return -EINVAL;
1283 1284
	}

1285 1286 1287 1288
	rv = audit_set_loginuid(kloginuid);
	if (rv < 0)
		return rv;
	return count;
L
Linus Torvalds 已提交
1289 1290
}

1291
static const struct file_operations proc_loginuid_operations = {
L
Linus Torvalds 已提交
1292 1293
	.read		= proc_loginuid_read,
	.write		= proc_loginuid_write,
1294
	.llseek		= generic_file_llseek,
L
Linus Torvalds 已提交
1295
};
1296 1297 1298 1299

static ssize_t proc_sessionid_read(struct file * file, char __user * buf,
				  size_t count, loff_t *ppos)
{
A
Al Viro 已提交
1300
	struct inode * inode = file_inode(file);
1301 1302 1303 1304 1305 1306 1307 1308 1309 1310 1311 1312 1313 1314
	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,
1315
	.llseek		= generic_file_llseek,
1316
};
L
Linus Torvalds 已提交
1317 1318
#endif

1319 1320 1321 1322
#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
1323
	struct task_struct *task = get_proc_task(file_inode(file));
1324 1325 1326 1327 1328 1329 1330 1331 1332 1333
	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);
1334 1335

	return simple_read_from_buffer(buf, count, ppos, buffer, len);
1336 1337 1338 1339 1340 1341
}

static ssize_t proc_fault_inject_write(struct file * file,
			const char __user * buf, size_t count, loff_t *ppos)
{
	struct task_struct *task;
1342
	char buffer[PROC_NUMBUF];
1343
	int make_it_fail;
1344
	int rv;
1345 1346 1347 1348 1349 1350 1351 1352

	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;
1353 1354 1355
	rv = kstrtoint(strstrip(buffer), 0, &make_it_fail);
	if (rv < 0)
		return rv;
1356 1357 1358
	if (make_it_fail < 0 || make_it_fail > 1)
		return -EINVAL;

A
Al Viro 已提交
1359
	task = get_proc_task(file_inode(file));
1360 1361 1362 1363
	if (!task)
		return -ESRCH;
	task->make_it_fail = make_it_fail;
	put_task_struct(task);
1364 1365

	return count;
1366 1367
}

1368
static const struct file_operations proc_fault_inject_operations = {
1369 1370
	.read		= proc_fault_inject_read,
	.write		= proc_fault_inject_write,
1371
	.llseek		= generic_file_llseek,
1372 1373 1374
};
#endif

A
Arjan van de Ven 已提交
1375

I
Ingo Molnar 已提交
1376 1377 1378 1379 1380 1381 1382 1383 1384 1385 1386 1387 1388 1389 1390 1391 1392 1393 1394 1395 1396 1397 1398
#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;
	struct task_struct *p;

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

	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
1399
	struct inode *inode = file_inode(file);
I
Ingo Molnar 已提交
1400 1401 1402 1403 1404 1405 1406 1407 1408 1409 1410 1411 1412 1413
	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
1414
	return single_open(filp, sched_show, inode);
I
Ingo Molnar 已提交
1415 1416 1417 1418 1419 1420 1421
}

static const struct file_operations proc_pid_sched_operations = {
	.open		= sched_open,
	.read		= seq_read,
	.write		= sched_write,
	.llseek		= seq_lseek,
1422
	.release	= single_release,
I
Ingo Molnar 已提交
1423 1424 1425 1426
};

#endif

1427 1428 1429 1430 1431 1432 1433 1434 1435 1436 1437 1438 1439 1440 1441 1442 1443 1444 1445 1446 1447 1448 1449
#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
1450
	struct inode *inode = file_inode(file);
1451 1452
	struct task_struct *p;
	char buffer[PROC_NUMBUF];
A
Alexey Dobriyan 已提交
1453
	int nice;
1454 1455 1456 1457 1458 1459 1460 1461
	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
1462 1463 1464
	err = kstrtoint(strstrip(buffer), 0, &nice);
	if (err < 0)
		return err;
1465 1466 1467 1468 1469

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

1470
	err = proc_sched_autogroup_set_nice(p, nice);
1471 1472 1473 1474 1475 1476 1477 1478 1479 1480 1481 1482 1483 1484 1485 1486 1487 1488 1489 1490 1491 1492 1493 1494 1495 1496 1497 1498 1499 1500 1501
	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 */

1502 1503 1504
static ssize_t comm_write(struct file *file, const char __user *buf,
				size_t count, loff_t *offset)
{
A
Al Viro 已提交
1505
	struct inode *inode = file_inode(file);
1506 1507
	struct task_struct *p;
	char buffer[TASK_COMM_LEN];
1508
	const size_t maxlen = sizeof(buffer) - 1;
1509 1510

	memset(buffer, 0, sizeof(buffer));
1511
	if (copy_from_user(buffer, buf, count > maxlen ? maxlen : count))
1512 1513 1514 1515 1516 1517 1518 1519 1520 1521 1522 1523 1524 1525 1526 1527 1528 1529 1530 1531 1532 1533 1534 1535 1536 1537 1538 1539 1540 1541 1542 1543 1544 1545 1546 1547
		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
1548
	return single_open(filp, comm_show, inode);
1549 1550 1551 1552 1553 1554 1555 1556 1557 1558
}

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,
};

1559
static int proc_exe_link(struct dentry *dentry, struct path *exe_path)
M
Matt Helsley 已提交
1560 1561 1562 1563 1564
{
	struct task_struct *task;
	struct mm_struct *mm;
	struct file *exe_file;

1565
	task = get_proc_task(d_inode(dentry));
M
Matt Helsley 已提交
1566 1567 1568 1569 1570 1571 1572 1573 1574 1575 1576 1577 1578 1579 1580 1581 1582
	if (!task)
		return -ENOENT;
	mm = get_task_mm(task);
	put_task_struct(task);
	if (!mm)
		return -ENOENT;
	exe_file = get_mm_exe_file(mm);
	mmput(mm);
	if (exe_file) {
		*exe_path = exe_file->f_path;
		path_get(&exe_file->f_path);
		fput(exe_file);
		return 0;
	} else
		return -ENOENT;
}

1583
static const char *proc_pid_get_link(struct dentry *dentry,
1584 1585
				     struct inode *inode,
				     struct delayed_call *done)
L
Linus Torvalds 已提交
1586
{
1587
	struct path path;
L
Linus Torvalds 已提交
1588 1589
	int error = -EACCES;

1590 1591 1592
	if (!dentry)
		return ERR_PTR(-ECHILD);

1593 1594
	/* Are we allowed to snoop on the tasks file descriptors? */
	if (!proc_fd_access_allowed(inode))
L
Linus Torvalds 已提交
1595 1596
		goto out;

1597 1598 1599 1600
	error = PROC_I(inode)->op.proc_get_link(dentry, &path);
	if (error)
		goto out;

1601
	nd_jump_link(&path);
1602
	return NULL;
L
Linus Torvalds 已提交
1603
out:
1604
	return ERR_PTR(error);
L
Linus Torvalds 已提交
1605 1606
}

1607
static int do_proc_readlink(struct path *path, char __user *buffer, int buflen)
L
Linus Torvalds 已提交
1608
{
1609
	char *tmp = (char*)__get_free_page(GFP_TEMPORARY);
1610
	char *pathname;
L
Linus Torvalds 已提交
1611 1612 1613 1614
	int len;

	if (!tmp)
		return -ENOMEM;
1615

1616
	pathname = d_path(path, tmp, PAGE_SIZE);
1617 1618
	len = PTR_ERR(pathname);
	if (IS_ERR(pathname))
L
Linus Torvalds 已提交
1619
		goto out;
1620
	len = tmp + PAGE_SIZE - 1 - pathname;
L
Linus Torvalds 已提交
1621 1622 1623

	if (len > buflen)
		len = buflen;
1624
	if (copy_to_user(buffer, pathname, len))
L
Linus Torvalds 已提交
1625 1626 1627 1628 1629 1630 1631 1632 1633
		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;
1634
	struct inode *inode = d_inode(dentry);
1635
	struct path path;
L
Linus Torvalds 已提交
1636

1637 1638
	/* Are we allowed to snoop on the tasks file descriptors? */
	if (!proc_fd_access_allowed(inode))
L
Linus Torvalds 已提交
1639 1640
		goto out;

1641
	error = PROC_I(inode)->op.proc_get_link(dentry, &path);
L
Linus Torvalds 已提交
1642 1643 1644
	if (error)
		goto out;

1645 1646
	error = do_proc_readlink(&path, buffer, buflen);
	path_put(&path);
L
Linus Torvalds 已提交
1647 1648 1649 1650
out:
	return error;
}

1651
const struct inode_operations proc_pid_link_inode_operations = {
L
Linus Torvalds 已提交
1652
	.readlink	= proc_pid_readlink,
1653
	.get_link	= proc_pid_get_link,
1654
	.setattr	= proc_setattr,
L
Linus Torvalds 已提交
1655 1656
};

1657 1658 1659

/* building an inode */

1660
struct inode *proc_pid_make_inode(struct super_block * sb, struct task_struct *task)
1661 1662 1663
{
	struct inode * inode;
	struct proc_inode *ei;
1664
	const struct cred *cred;
L
Linus Torvalds 已提交
1665

1666
	/* We need a new inode */
L
Linus Torvalds 已提交
1667

1668 1669 1670 1671 1672 1673
	inode = new_inode(sb);
	if (!inode)
		goto out;

	/* Common stuff */
	ei = PROC_I(inode);
1674
	inode->i_ino = get_next_ino();
1675 1676 1677 1678 1679 1680
	inode->i_mtime = inode->i_atime = inode->i_ctime = CURRENT_TIME;
	inode->i_op = &proc_def_inode_operations;

	/*
	 * grab the reference to task.
	 */
1681
	ei->pid = get_task_pid(task, PIDTYPE_PID);
1682 1683 1684 1685
	if (!ei->pid)
		goto out_unlock;

	if (task_dumpable(task)) {
1686 1687 1688 1689 1690
		rcu_read_lock();
		cred = __task_cred(task);
		inode->i_uid = cred->euid;
		inode->i_gid = cred->egid;
		rcu_read_unlock();
L
Linus Torvalds 已提交
1691
	}
1692 1693
	security_task_to_inode(task, inode);

L
Linus Torvalds 已提交
1694
out:
1695 1696 1697 1698 1699
	return inode;

out_unlock:
	iput(inode);
	return NULL;
L
Linus Torvalds 已提交
1700 1701
}

1702
int pid_getattr(struct vfsmount *mnt, struct dentry *dentry, struct kstat *stat)
L
Linus Torvalds 已提交
1703
{
1704
	struct inode *inode = d_inode(dentry);
1705
	struct task_struct *task;
1706
	const struct cred *cred;
1707
	struct pid_namespace *pid = dentry->d_sb->s_fs_info;
1708

1709
	generic_fillattr(inode, stat);
L
Linus Torvalds 已提交
1710

1711
	rcu_read_lock();
1712 1713
	stat->uid = GLOBAL_ROOT_UID;
	stat->gid = GLOBAL_ROOT_GID;
1714 1715
	task = pid_task(proc_pid(inode), PIDTYPE_PID);
	if (task) {
1716 1717 1718 1719 1720 1721 1722 1723
		if (!has_pid_permissions(pid, task, 2)) {
			rcu_read_unlock();
			/*
			 * This doesn't prevent learning whether PID exists,
			 * it only makes getattr() consistent with readdir().
			 */
			return -ENOENT;
		}
1724 1725
		if ((inode->i_mode == (S_IFDIR|S_IRUGO|S_IXUGO)) ||
		    task_dumpable(task)) {
1726 1727 1728
			cred = __task_cred(task);
			stat->uid = cred->euid;
			stat->gid = cred->egid;
L
Linus Torvalds 已提交
1729 1730
		}
	}
1731
	rcu_read_unlock();
A
Alan Cox 已提交
1732
	return 0;
L
Linus Torvalds 已提交
1733 1734 1735 1736 1737 1738 1739 1740 1741 1742 1743
}

/* 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.
1744 1745 1746 1747 1748 1749 1750
 *
 * 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.
L
Linus Torvalds 已提交
1751
 */
1752
int pid_revalidate(struct dentry *dentry, unsigned int flags)
L
Linus Torvalds 已提交
1753
{
1754 1755
	struct inode *inode;
	struct task_struct *task;
1756 1757
	const struct cred *cred;

1758
	if (flags & LOOKUP_RCU)
1759 1760
		return -ECHILD;

1761
	inode = d_inode(dentry);
1762 1763
	task = get_proc_task(inode);

1764 1765 1766
	if (task) {
		if ((inode->i_mode == (S_IFDIR|S_IRUGO|S_IXUGO)) ||
		    task_dumpable(task)) {
1767 1768 1769 1770 1771
			rcu_read_lock();
			cred = __task_cred(task);
			inode->i_uid = cred->euid;
			inode->i_gid = cred->egid;
			rcu_read_unlock();
L
Linus Torvalds 已提交
1772
		} else {
1773 1774
			inode->i_uid = GLOBAL_ROOT_UID;
			inode->i_gid = GLOBAL_ROOT_GID;
L
Linus Torvalds 已提交
1775
		}
L
Linus Torvalds 已提交
1776
		inode->i_mode &= ~(S_ISUID | S_ISGID);
L
Linus Torvalds 已提交
1777
		security_task_to_inode(task, inode);
1778
		put_task_struct(task);
L
Linus Torvalds 已提交
1779 1780 1781 1782 1783
		return 1;
	}
	return 0;
}

1784 1785 1786 1787 1788
static inline bool proc_inode_is_dead(struct inode *inode)
{
	return !proc_pid(inode)->tasks[PIDTYPE_PID].first;
}

1789 1790 1791 1792 1793 1794
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.
	 */
1795
	return proc_inode_is_dead(d_inode(dentry));
1796 1797
}

1798
const struct dentry_operations pid_dentry_operations =
1799 1800 1801 1802 1803 1804 1805
{
	.d_revalidate	= pid_revalidate,
	.d_delete	= pid_delete_dentry,
};

/* Lookups */

1806 1807 1808 1809 1810 1811 1812 1813 1814 1815 1816 1817
/*
 * 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
1818
bool proc_fill_cache(struct file *file, struct dir_context *ctx,
1819
	const char *name, int len,
1820
	instantiate_t instantiate, struct task_struct *task, const void *ptr)
1821
{
A
Al Viro 已提交
1822
	struct dentry *child, *dir = file->f_path.dentry;
1823
	struct qstr qname = QSTR_INIT(name, len);
1824
	struct inode *inode;
1825 1826
	unsigned type;
	ino_t ino;
1827

1828
	child = d_hash_and_lookup(dir, &qname);
1829
	if (!child) {
1830 1831 1832
		DECLARE_WAIT_QUEUE_HEAD_ONSTACK(wq);
		child = d_alloc_parallel(dir, &qname, &wq);
		if (IS_ERR(child))
1833
			goto end_instantiate;
1834 1835 1836 1837 1838 1839 1840
		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;
			}
1841 1842
		}
	}
1843
	inode = d_inode(child);
1844 1845
	ino = inode->i_ino;
	type = inode->i_mode >> 12;
1846
	dput(child);
A
Al Viro 已提交
1847
	return dir_emit(ctx, name, len, ino, type);
1848 1849 1850

end_instantiate:
	return dir_emit(ctx, name, len, 1, DT_UNKNOWN);
1851 1852
}

1853 1854 1855 1856 1857 1858 1859 1860 1861 1862 1863 1864 1865
/*
 * 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)
{
	if (sscanf(dentry->d_name.name, "%lx-%lx", start, end) != 2)
		return -EINVAL;

	return 0;
}

1866
static int map_files_d_revalidate(struct dentry *dentry, unsigned int flags)
1867 1868 1869 1870 1871 1872 1873 1874 1875
{
	unsigned long vm_start, vm_end;
	bool exact_vma_exists = false;
	struct mm_struct *mm = NULL;
	struct task_struct *task;
	const struct cred *cred;
	struct inode *inode;
	int status = 0;

1876
	if (flags & LOOKUP_RCU)
1877 1878
		return -ECHILD;

1879
	inode = d_inode(dentry);
1880 1881 1882 1883
	task = get_proc_task(inode);
	if (!task)
		goto out_notask;

1884
	mm = mm_access(task, PTRACE_MODE_READ_FSCREDS);
1885
	if (IS_ERR_OR_NULL(mm))
1886 1887 1888 1889 1890 1891 1892 1893 1894 1895 1896 1897 1898 1899 1900 1901 1902 1903
		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) {
		if (task_dumpable(task)) {
			rcu_read_lock();
			cred = __task_cred(task);
			inode->i_uid = cred->euid;
			inode->i_gid = cred->egid;
			rcu_read_unlock();
		} else {
1904 1905
			inode->i_uid = GLOBAL_ROOT_UID;
			inode->i_gid = GLOBAL_ROOT_GID;
1906 1907 1908 1909 1910 1911 1912 1913 1914 1915 1916 1917 1918 1919 1920 1921 1922
		}
		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,
};

1923
static int map_files_get_link(struct dentry *dentry, struct path *path)
1924 1925 1926 1927 1928 1929 1930 1931
{
	unsigned long vm_start, vm_end;
	struct vm_area_struct *vma;
	struct task_struct *task;
	struct mm_struct *mm;
	int rc;

	rc = -ENOENT;
1932
	task = get_proc_task(d_inode(dentry));
1933 1934 1935 1936 1937 1938 1939 1940 1941 1942 1943 1944
	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;

1945
	rc = -ENOENT;
1946 1947 1948 1949 1950 1951 1952 1953 1954 1955 1956 1957 1958 1959 1960 1961
	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 {
1962
	fmode_t		mode;
1963 1964 1965 1966
	unsigned long	len;
	unsigned char	name[4*sizeof(long)+2]; /* max: %lx-%lx\0 */
};

1967 1968 1969 1970 1971 1972
/*
 * 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 *
1973
proc_map_files_get_link(struct dentry *dentry,
1974 1975
			struct inode *inode,
		        struct delayed_call *done)
1976 1977 1978 1979
{
	if (!capable(CAP_SYS_ADMIN))
		return ERR_PTR(-EPERM);

1980
	return proc_pid_get_link(dentry, inode, done);
1981 1982 1983
}

/*
1984
 * Identical to proc_pid_link_inode_operations except for get_link()
1985 1986 1987
 */
static const struct inode_operations proc_map_files_link_inode_operations = {
	.readlink	= proc_pid_readlink,
1988
	.get_link	= proc_map_files_get_link,
1989 1990 1991
	.setattr	= proc_setattr,
};

1992
static int
1993 1994 1995
proc_map_files_instantiate(struct inode *dir, struct dentry *dentry,
			   struct task_struct *task, const void *ptr)
{
1996
	fmode_t mode = (fmode_t)(unsigned long)ptr;
1997 1998 1999 2000 2001
	struct proc_inode *ei;
	struct inode *inode;

	inode = proc_pid_make_inode(dir->i_sb, task);
	if (!inode)
2002
		return -ENOENT;
2003 2004

	ei = PROC_I(inode);
2005
	ei->op.proc_get_link = map_files_get_link;
2006

2007
	inode->i_op = &proc_map_files_link_inode_operations;
2008 2009 2010
	inode->i_size = 64;
	inode->i_mode = S_IFLNK;

2011
	if (mode & FMODE_READ)
2012
		inode->i_mode |= S_IRUSR;
2013
	if (mode & FMODE_WRITE)
2014 2015 2016 2017 2018
		inode->i_mode |= S_IWUSR;

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

2019
	return 0;
2020 2021 2022
}

static struct dentry *proc_map_files_lookup(struct inode *dir,
A
Al Viro 已提交
2023
		struct dentry *dentry, unsigned int flags)
2024 2025 2026 2027
{
	unsigned long vm_start, vm_end;
	struct vm_area_struct *vma;
	struct task_struct *task;
2028
	int result;
2029 2030
	struct mm_struct *mm;

2031
	result = -ENOENT;
2032 2033 2034 2035
	task = get_proc_task(dir);
	if (!task)
		goto out;

2036
	result = -EACCES;
2037
	if (!ptrace_may_access(task, PTRACE_MODE_READ_FSCREDS))
2038 2039
		goto out_put_task;

2040
	result = -ENOENT;
2041
	if (dname_to_vma_addr(dentry, &vm_start, &vm_end))
2042
		goto out_put_task;
2043 2044 2045

	mm = get_task_mm(task);
	if (!mm)
2046
		goto out_put_task;
2047 2048 2049 2050 2051 2052

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

2053 2054 2055
	if (vma->vm_file)
		result = proc_map_files_instantiate(dir, dentry, task,
				(void *)(unsigned long)vma->vm_file->f_mode);
2056 2057 2058 2059 2060 2061 2062

out_no_vma:
	up_read(&mm->mmap_sem);
	mmput(mm);
out_put_task:
	put_task_struct(task);
out:
2063
	return ERR_PTR(result);
2064 2065 2066 2067 2068 2069 2070 2071 2072
}

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
2073
proc_map_files_readdir(struct file *file, struct dir_context *ctx)
2074 2075 2076 2077
{
	struct vm_area_struct *vma;
	struct task_struct *task;
	struct mm_struct *mm;
A
Al Viro 已提交
2078 2079 2080 2081
	unsigned long nr_files, pos, i;
	struct flex_array *fa = NULL;
	struct map_files_info info;
	struct map_files_info *p;
2082 2083 2084
	int ret;

	ret = -ENOENT;
A
Al Viro 已提交
2085
	task = get_proc_task(file_inode(file));
2086 2087 2088 2089
	if (!task)
		goto out;

	ret = -EACCES;
2090
	if (!ptrace_may_access(task, PTRACE_MODE_READ_FSCREDS))
2091 2092 2093
		goto out_put_task;

	ret = 0;
A
Al Viro 已提交
2094 2095
	if (!dir_emit_dots(file, ctx))
		goto out_put_task;
2096

A
Al Viro 已提交
2097 2098 2099 2100
	mm = get_task_mm(task);
	if (!mm)
		goto out_put_task;
	down_read(&mm->mmap_sem);
2101

A
Al Viro 已提交
2102
	nr_files = 0;
2103

A
Al Viro 已提交
2104 2105 2106 2107 2108 2109 2110 2111 2112
	/*
	 * 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().
	 */
2113

A
Al Viro 已提交
2114 2115 2116 2117 2118 2119 2120 2121 2122 2123 2124 2125 2126 2127 2128 2129
	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;
2130
		}
A
Al Viro 已提交
2131 2132 2133 2134 2135 2136 2137 2138 2139 2140 2141 2142 2143
		for (i = 0, vma = mm->mmap, pos = 2; vma;
				vma = vma->vm_next) {
			if (!vma->vm_file)
				continue;
			if (++pos <= ctx->pos)
				continue;

			info.mode = vma->vm_file->f_mode;
			info.len = snprintf(info.name,
					sizeof(info.name), "%lx-%lx",
					vma->vm_start, vma->vm_end);
			if (flex_array_put(fa, i++, &info, GFP_KERNEL))
				BUG();
2144 2145
		}
	}
A
Al Viro 已提交
2146 2147 2148 2149 2150 2151 2152 2153 2154 2155 2156
	up_read(&mm->mmap_sem);

	for (i = 0; i < nr_files; i++) {
		p = flex_array_get(fa, i);
		if (!proc_fill_cache(file, ctx,
				      p->name, p->len,
				      proc_map_files_instantiate,
				      task,
				      (void *)(unsigned long)p->mode))
			break;
		ctx->pos++;
2157
	}
A
Al Viro 已提交
2158 2159 2160
	if (fa)
		flex_array_free(fa);
	mmput(mm);
2161 2162 2163 2164 2165 2166 2167 2168 2169

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

static const struct file_operations proc_map_files_operations = {
	.read		= generic_read_dir,
2170 2171
	.iterate_shared	= proc_map_files_readdir,
	.llseek		= generic_file_llseek,
2172 2173
};

2174
#ifdef CONFIG_CHECKPOINT_RESTORE
2175 2176 2177 2178
struct timers_private {
	struct pid *pid;
	struct task_struct *task;
	struct sighand_struct *sighand;
2179
	struct pid_namespace *ns;
2180 2181 2182 2183 2184 2185 2186 2187 2188 2189 2190 2191 2192 2193 2194 2195 2196 2197 2198 2199 2200 2201 2202 2203 2204 2205 2206 2207 2208 2209 2210 2211 2212 2213 2214 2215 2216 2217 2218 2219 2220 2221
	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;
2222 2223
	struct timers_private *tp = m->private;
	int notify;
2224
	static const char * const nstr[] = {
2225 2226 2227 2228
		[SIGEV_SIGNAL] = "signal",
		[SIGEV_NONE] = "none",
		[SIGEV_THREAD] = "thread",
	};
2229 2230

	timer = list_entry((struct list_head *)v, struct k_itimer, list);
2231 2232
	notify = timer->it_sigev_notify;

2233
	seq_printf(m, "ID: %d\n", timer->it_id);
2234 2235 2236
	seq_printf(m, "signal: %d/%p\n",
		   timer->sigq->info.si_signo,
		   timer->sigq->info.si_value.sival_ptr);
2237
	seq_printf(m, "notify: %s/%s.%d\n",
2238 2239 2240
		   nstr[notify & ~SIGEV_THREAD_ID],
		   (notify & SIGEV_THREAD_ID) ? "tid" : "pid",
		   pid_nr_ns(timer->it_pid, tp->ns));
2241
	seq_printf(m, "ClockID: %d\n", timer->it_clock);
2242 2243 2244 2245 2246 2247 2248 2249 2250 2251 2252 2253 2254 2255 2256 2257 2258 2259 2260 2261 2262

	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);
2263
	tp->ns = inode->i_sb->s_fs_info;
2264 2265 2266 2267 2268 2269 2270 2271 2272
	return 0;
}

static const struct file_operations proc_timers_operations = {
	.open		= proc_timers_open,
	.read		= seq_read,
	.llseek		= seq_lseek,
	.release	= seq_release_private,
};
2273
#endif
2274

2275 2276 2277 2278 2279 2280 2281 2282 2283 2284 2285 2286 2287 2288 2289 2290 2291 2292 2293 2294 2295 2296 2297 2298 2299 2300 2301 2302 2303 2304 2305 2306 2307 2308 2309 2310 2311 2312 2313 2314 2315 2316 2317 2318 2319 2320 2321 2322 2323 2324 2325 2326 2327 2328 2329 2330 2331 2332 2333 2334 2335 2336 2337 2338 2339 2340
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;

	if (ptrace_may_access(p, PTRACE_MODE_ATTACH_FSCREDS)) {
		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);
	} else
		count = -EPERM;

	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;
	int err =  0;

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

	if (ptrace_may_access(p, PTRACE_MODE_ATTACH_FSCREDS)) {
		task_lock(p);
		seq_printf(m, "%llu\n", p->timer_slack_ns);
		task_unlock(p);
	} else
		err = -EPERM;

	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,
};

2341
static int proc_pident_instantiate(struct inode *dir,
2342
	struct dentry *dentry, struct task_struct *task, const void *ptr)
2343
{
2344
	const struct pid_entry *p = ptr;
2345 2346 2347
	struct inode *inode;
	struct proc_inode *ei;

2348
	inode = proc_pid_make_inode(dir->i_sb, task);
2349 2350 2351 2352 2353 2354
	if (!inode)
		goto out;

	ei = PROC_I(inode);
	inode->i_mode = p->mode;
	if (S_ISDIR(inode->i_mode))
M
Miklos Szeredi 已提交
2355
		set_nlink(inode, 2);	/* Use getattr to fix if necessary */
2356 2357 2358 2359 2360
	if (p->iop)
		inode->i_op = p->iop;
	if (p->fop)
		inode->i_fop = p->fop;
	ei->op = p->op;
2361
	d_set_d_op(dentry, &pid_dentry_operations);
2362 2363
	d_add(dentry, inode);
	/* Close the race of the process dying before we return the dentry */
2364
	if (pid_revalidate(dentry, 0))
2365
		return 0;
2366
out:
2367
	return -ENOENT;
2368 2369
}

L
Linus Torvalds 已提交
2370 2371
static struct dentry *proc_pident_lookup(struct inode *dir, 
					 struct dentry *dentry,
2372
					 const struct pid_entry *ents,
2373
					 unsigned int nents)
L
Linus Torvalds 已提交
2374
{
2375
	int error;
2376
	struct task_struct *task = get_proc_task(dir);
2377
	const struct pid_entry *p, *last;
L
Linus Torvalds 已提交
2378

2379
	error = -ENOENT;
L
Linus Torvalds 已提交
2380

2381 2382
	if (!task)
		goto out_no_task;
L
Linus Torvalds 已提交
2383

2384 2385 2386 2387
	/*
	 * Yes, it does not scale. And it should not. Don't add
	 * new entries into /proc/<tgid>/ without very good reasons.
	 */
2388 2389
	last = &ents[nents - 1];
	for (p = ents; p <= last; p++) {
L
Linus Torvalds 已提交
2390 2391 2392 2393 2394
		if (p->len != dentry->d_name.len)
			continue;
		if (!memcmp(dentry->d_name.name, p->name, p->len))
			break;
	}
2395
	if (p > last)
L
Linus Torvalds 已提交
2396 2397
		goto out;

2398
	error = proc_pident_instantiate(dir, dentry, task, p);
L
Linus Torvalds 已提交
2399
out:
2400 2401
	put_task_struct(task);
out_no_task:
2402
	return ERR_PTR(error);
L
Linus Torvalds 已提交
2403 2404
}

A
Al Viro 已提交
2405
static int proc_pident_readdir(struct file *file, struct dir_context *ctx,
2406
		const struct pid_entry *ents, unsigned int nents)
2407
{
A
Al Viro 已提交
2408 2409
	struct task_struct *task = get_proc_task(file_inode(file));
	const struct pid_entry *p;
2410 2411

	if (!task)
A
Al Viro 已提交
2412
		return -ENOENT;
2413

A
Al Viro 已提交
2414 2415 2416 2417 2418
	if (!dir_emit_dots(file, ctx))
		goto out;

	if (ctx->pos >= nents + 2)
		goto out;
2419

A
Al Viro 已提交
2420 2421 2422 2423 2424 2425
	for (p = ents + (ctx->pos - 2); p <= ents + nents - 1; p++) {
		if (!proc_fill_cache(file, ctx, p->name, p->len,
				proc_pident_instantiate, task, p))
			break;
		ctx->pos++;
	}
2426
out:
2427
	put_task_struct(task);
A
Al Viro 已提交
2428
	return 0;
L
Linus Torvalds 已提交
2429 2430
}

2431 2432 2433 2434
#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
2435
	struct inode * inode = file_inode(file);
2436
	char *p = NULL;
2437 2438 2439 2440
	ssize_t length;
	struct task_struct *task = get_proc_task(inode);

	if (!task)
2441
		return -ESRCH;
2442 2443

	length = security_getprocattr(task,
2444
				      (char*)file->f_path.dentry->d_name.name,
2445
				      &p);
2446
	put_task_struct(task);
2447 2448 2449
	if (length > 0)
		length = simple_read_from_buffer(buf, count, ppos, p, length);
	kfree(p);
2450
	return length;
L
Linus Torvalds 已提交
2451 2452
}

2453 2454 2455
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
2456
	struct inode * inode = file_inode(file);
2457
	void *page;
2458 2459 2460 2461 2462 2463 2464 2465 2466 2467 2468 2469 2470 2471
	ssize_t length;
	struct task_struct *task = get_proc_task(inode);

	length = -ESRCH;
	if (!task)
		goto out_no_task;
	if (count > PAGE_SIZE)
		count = PAGE_SIZE;

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

2472 2473 2474
	page = memdup_user(buf, count);
	if (IS_ERR(page)) {
		length = PTR_ERR(page);
2475
		goto out;
2476
	}
2477

2478
	/* Guard against adverse ptrace interaction */
2479
	length = mutex_lock_interruptible(&task->signal->cred_guard_mutex);
2480 2481 2482
	if (length < 0)
		goto out_free;

2483
	length = security_setprocattr(task,
2484
				      (char*)file->f_path.dentry->d_name.name,
2485
				      page, count);
2486
	mutex_unlock(&task->signal->cred_guard_mutex);
2487
out_free:
2488
	kfree(page);
2489 2490 2491 2492 2493 2494
out:
	put_task_struct(task);
out_no_task:
	return length;
}

2495
static const struct file_operations proc_pid_attr_operations = {
2496 2497
	.read		= proc_pid_attr_read,
	.write		= proc_pid_attr_write,
2498
	.llseek		= generic_file_llseek,
2499 2500
};

2501
static const struct pid_entry attr_dir_stuff[] = {
A
Alexey Dobriyan 已提交
2502 2503 2504 2505 2506 2507
	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),
2508 2509
};

A
Al Viro 已提交
2510
static int proc_attr_dir_readdir(struct file *file, struct dir_context *ctx)
2511
{
A
Al Viro 已提交
2512 2513
	return proc_pident_readdir(file, ctx, 
				   attr_dir_stuff, ARRAY_SIZE(attr_dir_stuff));
2514 2515
}

2516
static const struct file_operations proc_attr_dir_operations = {
L
Linus Torvalds 已提交
2517
	.read		= generic_read_dir,
2518 2519
	.iterate_shared	= proc_attr_dir_readdir,
	.llseek		= generic_file_llseek,
L
Linus Torvalds 已提交
2520 2521
};

2522
static struct dentry *proc_attr_dir_lookup(struct inode *dir,
A
Al Viro 已提交
2523
				struct dentry *dentry, unsigned int flags)
2524
{
2525 2526
	return proc_pident_lookup(dir, dentry,
				  attr_dir_stuff, ARRAY_SIZE(attr_dir_stuff));
2527 2528
}

2529
static const struct inode_operations proc_attr_dir_inode_operations = {
2530
	.lookup		= proc_attr_dir_lookup,
2531
	.getattr	= pid_getattr,
2532
	.setattr	= proc_setattr,
L
Linus Torvalds 已提交
2533 2534
};

2535 2536
#endif

C
Christoph Hellwig 已提交
2537
#ifdef CONFIG_ELF_CORE
2538 2539 2540
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
2541
	struct task_struct *task = get_proc_task(file_inode(file));
2542 2543 2544 2545 2546 2547 2548 2549 2550 2551 2552 2553 2554 2555 2556 2557 2558 2559 2560 2561 2562 2563 2564 2565 2566 2567 2568 2569 2570 2571 2572 2573 2574 2575 2576
	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;

2577 2578 2579
	ret = kstrtouint_from_user(buf, count, 0, &val);
	if (ret < 0)
		return ret;
2580 2581

	ret = -ESRCH;
A
Al Viro 已提交
2582
	task = get_proc_task(file_inode(file));
2583 2584 2585 2586 2587 2588
	if (!task)
		goto out_no_task;

	mm = get_task_mm(task);
	if (!mm)
		goto out_no_mm;
2589
	ret = 0;
2590 2591 2592 2593 2594 2595 2596 2597 2598 2599 2600 2601

	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:
2602 2603 2604
	if (ret < 0)
		return ret;
	return count;
2605 2606 2607 2608 2609
}

static const struct file_operations proc_coredump_filter_operations = {
	.read		= proc_coredump_filter_read,
	.write		= proc_coredump_filter_write,
2610
	.llseek		= generic_file_llseek,
2611 2612 2613
};
#endif

2614
#ifdef CONFIG_TASK_IO_ACCOUNTING
2615
static int do_io_accounting(struct task_struct *task, struct seq_file *m, int whole)
2616
{
2617
	struct task_io_accounting acct = task->ioac;
2618
	unsigned long flags;
2619
	int result;
2620

2621 2622 2623 2624
	result = mutex_lock_killable(&task->signal->cred_guard_mutex);
	if (result)
		return result;

2625
	if (!ptrace_may_access(task, PTRACE_MODE_READ_FSCREDS)) {
2626 2627 2628
		result = -EACCES;
		goto out_unlock;
	}
2629

2630 2631 2632 2633 2634 2635 2636 2637
	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);
2638
	}
2639 2640 2641 2642 2643 2644 2645 2646 2647 2648 2649 2650 2651 2652 2653 2654 2655
	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;

2656 2657 2658
out_unlock:
	mutex_unlock(&task->signal->cred_guard_mutex);
	return result;
2659 2660
}

2661 2662
static int proc_tid_io_accounting(struct seq_file *m, struct pid_namespace *ns,
				  struct pid *pid, struct task_struct *task)
2663
{
2664
	return do_io_accounting(task, m, 0);
2665
}
2666

2667 2668
static int proc_tgid_io_accounting(struct seq_file *m, struct pid_namespace *ns,
				   struct pid *pid, struct task_struct *task)
2669
{
2670
	return do_io_accounting(task, m, 1);
2671 2672
}
#endif /* CONFIG_TASK_IO_ACCOUNTING */
2673

2674 2675
#ifdef CONFIG_USER_NS
static int proc_id_map_open(struct inode *inode, struct file *file,
F
Fabian Frederick 已提交
2676
	const struct seq_operations *seq_ops)
2677 2678 2679 2680 2681 2682 2683 2684 2685 2686 2687 2688 2689 2690 2691 2692 2693 2694 2695 2696 2697 2698 2699 2700 2701 2702 2703 2704 2705 2706 2707 2708 2709 2710 2711 2712 2713 2714 2715 2716 2717 2718 2719 2720 2721 2722 2723 2724
{
	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);
}

2725 2726 2727 2728 2729
static int proc_projid_map_open(struct inode *inode, struct file *file)
{
	return proc_id_map_open(inode, file, &proc_projid_seq_operations);
}

2730 2731 2732 2733 2734 2735 2736 2737 2738 2739 2740 2741 2742 2743 2744
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,
};
2745 2746 2747 2748 2749 2750 2751 2752

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,
};
2753 2754 2755 2756 2757 2758 2759 2760 2761 2762 2763 2764 2765 2766 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

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,
};
2804 2805
#endif /* CONFIG_USER_NS */

2806 2807 2808
static int proc_pid_personality(struct seq_file *m, struct pid_namespace *ns,
				struct pid *pid, struct task_struct *task)
{
2809 2810 2811 2812 2813 2814
	int err = lock_trace(task);
	if (!err) {
		seq_printf(m, "%08x\n", task->personality);
		unlock_trace(task);
	}
	return err;
2815 2816
}

2817 2818 2819
/*
 * Thread groups
 */
2820
static const struct file_operations proc_task_operations;
2821
static const struct inode_operations proc_task_inode_operations;
2822

2823
static const struct pid_entry tgid_base_stuff[] = {
A
Alexey Dobriyan 已提交
2824 2825
	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),
2826
	DIR("map_files",  S_IRUSR|S_IXUSR, proc_map_files_inode_operations, proc_map_files_operations),
A
Alexey Dobriyan 已提交
2827
	DIR("fdinfo",     S_IRUSR|S_IXUSR, proc_fdinfo_inode_operations, proc_fdinfo_operations),
2828
	DIR("ns",	  S_IRUSR|S_IXUGO, proc_ns_dir_inode_operations, proc_ns_dir_operations),
A
Andrew Morton 已提交
2829
#ifdef CONFIG_NET
A
Alexey Dobriyan 已提交
2830
	DIR("net",        S_IRUGO|S_IXUGO, proc_net_inode_operations, proc_net_operations),
A
Andrew Morton 已提交
2831
#endif
A
Alexey Dobriyan 已提交
2832
	REG("environ",    S_IRUSR, proc_environ_operations),
2833
	ONE("auxv",       S_IRUSR, proc_pid_auxv),
A
Alexey Dobriyan 已提交
2834
	ONE("status",     S_IRUGO, proc_pid_status),
2835
	ONE("personality", S_IRUSR, proc_pid_personality),
2836
	ONE("limits",	  S_IRUGO, proc_pid_limits),
I
Ingo Molnar 已提交
2837
#ifdef CONFIG_SCHED_DEBUG
A
Alexey Dobriyan 已提交
2838
	REG("sched",      S_IRUGO|S_IWUSR, proc_pid_sched_operations),
2839 2840 2841
#endif
#ifdef CONFIG_SCHED_AUTOGROUP
	REG("autogroup",  S_IRUGO|S_IWUSR, proc_pid_sched_autogroup_operations),
R
Roland McGrath 已提交
2842
#endif
2843
	REG("comm",      S_IRUGO|S_IWUSR, proc_pid_set_comm_operations),
R
Roland McGrath 已提交
2844
#ifdef CONFIG_HAVE_ARCH_TRACEHOOK
2845
	ONE("syscall",    S_IRUSR, proc_pid_syscall),
I
Ingo Molnar 已提交
2846
#endif
2847
	REG("cmdline",    S_IRUGO, proc_pid_cmdline_ops),
A
Alexey Dobriyan 已提交
2848 2849
	ONE("stat",       S_IRUGO, proc_tgid_stat),
	ONE("statm",      S_IRUGO, proc_pid_statm),
2850
	REG("maps",       S_IRUGO, proc_pid_maps_operations),
2851
#ifdef CONFIG_NUMA
2852
	REG("numa_maps",  S_IRUGO, proc_pid_numa_maps_operations),
2853
#endif
A
Alexey Dobriyan 已提交
2854 2855 2856 2857 2858 2859 2860
	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),
2861
#ifdef CONFIG_PROC_PAGE_MONITOR
A
Alexey Dobriyan 已提交
2862
	REG("clear_refs", S_IWUSR, proc_clear_refs_operations),
2863
	REG("smaps",      S_IRUGO, proc_pid_smaps_operations),
2864
	REG("pagemap",    S_IRUSR, proc_pagemap_operations),
2865 2866
#endif
#ifdef CONFIG_SECURITY
A
Alexey Dobriyan 已提交
2867
	DIR("attr",       S_IRUGO|S_IXUGO, proc_attr_dir_inode_operations, proc_attr_dir_operations),
2868 2869
#endif
#ifdef CONFIG_KALLSYMS
2870
	ONE("wchan",      S_IRUGO, proc_pid_wchan),
2871
#endif
K
Ken Chen 已提交
2872
#ifdef CONFIG_STACKTRACE
2873
	ONE("stack",      S_IRUSR, proc_pid_stack),
2874
#endif
2875
#ifdef CONFIG_SCHED_INFO
2876
	ONE("schedstat",  S_IRUGO, proc_pid_schedstat),
2877
#endif
A
Arjan van de Ven 已提交
2878
#ifdef CONFIG_LATENCYTOP
A
Alexey Dobriyan 已提交
2879
	REG("latency",  S_IRUGO, proc_lstats_operations),
A
Arjan van de Ven 已提交
2880
#endif
2881
#ifdef CONFIG_PROC_PID_CPUSET
Z
Zefan Li 已提交
2882
	ONE("cpuset",     S_IRUGO, proc_cpuset_show),
2883 2884
#endif
#ifdef CONFIG_CGROUPS
Z
Zefan Li 已提交
2885
	ONE("cgroup",  S_IRUGO, proc_cgroup_show),
2886
#endif
2887
	ONE("oom_score",  S_IRUGO, proc_oom_score),
2888
	REG("oom_adj",    S_IRUGO|S_IWUSR, proc_oom_adj_operations),
D
David Rientjes 已提交
2889
	REG("oom_score_adj", S_IRUGO|S_IWUSR, proc_oom_score_adj_operations),
2890
#ifdef CONFIG_AUDITSYSCALL
A
Alexey Dobriyan 已提交
2891 2892
	REG("loginuid",   S_IWUSR|S_IRUGO, proc_loginuid_operations),
	REG("sessionid",  S_IRUGO, proc_sessionid_operations),
2893
#endif
2894
#ifdef CONFIG_FAULT_INJECTION
A
Alexey Dobriyan 已提交
2895
	REG("make-it-fail", S_IRUGO|S_IWUSR, proc_fault_inject_operations),
2896
#endif
C
Christoph Hellwig 已提交
2897
#ifdef CONFIG_ELF_CORE
A
Alexey Dobriyan 已提交
2898
	REG("coredump_filter", S_IRUGO|S_IWUSR, proc_coredump_filter_operations),
2899
#endif
2900
#ifdef CONFIG_TASK_IO_ACCOUNTING
2901
	ONE("io",	S_IRUSR, proc_tgid_io_accounting),
2902
#endif
2903
#ifdef CONFIG_HARDWALL
2904
	ONE("hardwall",   S_IRUGO, proc_pid_hardwall),
2905
#endif
2906 2907 2908
#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),
2909
	REG("projid_map", S_IRUGO|S_IWUSR, proc_projid_map_operations),
2910
	REG("setgroups",  S_IRUGO|S_IWUSR, proc_setgroups_operations),
2911
#endif
2912 2913 2914
#ifdef CONFIG_CHECKPOINT_RESTORE
	REG("timers",	  S_IRUGO, proc_timers_operations),
#endif
2915
	REG("timerslack_ns", S_IRUGO|S_IWUGO, proc_pid_set_timerslack_ns_operations),
2916
};
L
Linus Torvalds 已提交
2917

A
Al Viro 已提交
2918
static int proc_tgid_base_readdir(struct file *file, struct dir_context *ctx)
L
Linus Torvalds 已提交
2919
{
A
Al Viro 已提交
2920 2921
	return proc_pident_readdir(file, ctx,
				   tgid_base_stuff, ARRAY_SIZE(tgid_base_stuff));
L
Linus Torvalds 已提交
2922 2923
}

2924
static const struct file_operations proc_tgid_base_operations = {
L
Linus Torvalds 已提交
2925
	.read		= generic_read_dir,
2926 2927
	.iterate_shared	= proc_tgid_base_readdir,
	.llseek		= generic_file_llseek,
L
Linus Torvalds 已提交
2928 2929
};

A
Al Viro 已提交
2930 2931
static struct dentry *proc_tgid_base_lookup(struct inode *dir, struct dentry *dentry, unsigned int flags)
{
2932 2933
	return proc_pident_lookup(dir, dentry,
				  tgid_base_stuff, ARRAY_SIZE(tgid_base_stuff));
L
Linus Torvalds 已提交
2934 2935
}

2936
static const struct inode_operations proc_tgid_base_inode_operations = {
2937
	.lookup		= proc_tgid_base_lookup,
2938
	.getattr	= pid_getattr,
2939
	.setattr	= proc_setattr,
2940
	.permission	= proc_pid_permission,
L
Linus Torvalds 已提交
2941 2942
};

2943
static void proc_flush_task_mnt(struct vfsmount *mnt, pid_t pid, pid_t tgid)
L
Linus Torvalds 已提交
2944
{
2945
	struct dentry *dentry, *leader, *dir;
2946
	char buf[PROC_NUMBUF];
2947 2948 2949
	struct qstr name;

	name.name = buf;
2950
	name.len = snprintf(buf, sizeof(buf), "%d", pid);
2951
	/* no ->d_hash() rejects on procfs */
2952
	dentry = d_hash_and_lookup(mnt->mnt_root, &name);
2953
	if (dentry) {
2954
		d_invalidate(dentry);
2955 2956
		dput(dentry);
	}
L
Linus Torvalds 已提交
2957

2958 2959 2960
	if (pid == tgid)
		return;

2961
	name.name = buf;
2962 2963
	name.len = snprintf(buf, sizeof(buf), "%d", tgid);
	leader = d_hash_and_lookup(mnt->mnt_root, &name);
2964 2965
	if (!leader)
		goto out;
L
Linus Torvalds 已提交
2966

2967 2968 2969 2970 2971 2972 2973
	name.name = "task";
	name.len = strlen(name.name);
	dir = d_hash_and_lookup(leader, &name);
	if (!dir)
		goto out_put_leader;

	name.name = buf;
2974
	name.len = snprintf(buf, sizeof(buf), "%d", pid);
2975 2976
	dentry = d_hash_and_lookup(dir, &name);
	if (dentry) {
2977
		d_invalidate(dentry);
2978
		dput(dentry);
L
Linus Torvalds 已提交
2979
	}
2980 2981 2982 2983 2984 2985

	dput(dir);
out_put_leader:
	dput(leader);
out:
	return;
L
Linus Torvalds 已提交
2986 2987
}

2988 2989 2990 2991 2992
/**
 * 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
2993
 * proc (proc_mnt) and from all the namespaces' procs this task was seen
2994 2995 2996 2997 2998 2999 3000 3001 3002 3003 3004 3005 3006 3007 3008 3009 3010
 * 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.
3011 3012 3013 3014
 */

void proc_flush_task(struct task_struct *task)
{
3015
	int i;
3016
	struct pid *pid, *tgid;
3017 3018 3019
	struct upid *upid;

	pid = task_pid(task);
3020
	tgid = task_tgid(task);
3021

3022
	for (i = 0; i <= pid->level; i++) {
3023 3024
		upid = &pid->numbers[i];
		proc_flush_task_mnt(upid->ns->proc_mnt, upid->nr,
3025
					tgid->numbers[i].nr);
3026
	}
3027 3028
}

3029 3030 3031
static int proc_pid_instantiate(struct inode *dir,
				   struct dentry * dentry,
				   struct task_struct *task, const void *ptr)
3032 3033 3034
{
	struct inode *inode;

3035
	inode = proc_pid_make_inode(dir->i_sb, task);
3036 3037 3038 3039 3040 3041 3042
	if (!inode)
		goto out;

	inode->i_mode = S_IFDIR|S_IRUGO|S_IXUGO;
	inode->i_op = &proc_tgid_base_inode_operations;
	inode->i_fop = &proc_tgid_base_operations;
	inode->i_flags|=S_IMMUTABLE;
3043

M
Miklos Szeredi 已提交
3044 3045
	set_nlink(inode, 2 + pid_entry_count_dirs(tgid_base_stuff,
						  ARRAY_SIZE(tgid_base_stuff)));
3046

3047
	d_set_d_op(dentry, &pid_dentry_operations);
3048 3049 3050

	d_add(dentry, inode);
	/* Close the race of the process dying before we return the dentry */
3051
	if (pid_revalidate(dentry, 0))
3052
		return 0;
3053
out:
3054
	return -ENOENT;
3055 3056
}

A
Al Viro 已提交
3057
struct dentry *proc_pid_lookup(struct inode *dir, struct dentry * dentry, unsigned int flags)
L
Linus Torvalds 已提交
3058
{
A
Alexey Dobriyan 已提交
3059
	int result = -ENOENT;
L
Linus Torvalds 已提交
3060 3061
	struct task_struct *task;
	unsigned tgid;
3062
	struct pid_namespace *ns;
L
Linus Torvalds 已提交
3063

3064
	tgid = name_to_int(&dentry->d_name);
L
Linus Torvalds 已提交
3065 3066 3067
	if (tgid == ~0U)
		goto out;

3068
	ns = dentry->d_sb->s_fs_info;
3069
	rcu_read_lock();
3070
	task = find_task_by_pid_ns(tgid, ns);
L
Linus Torvalds 已提交
3071 3072
	if (task)
		get_task_struct(task);
3073
	rcu_read_unlock();
L
Linus Torvalds 已提交
3074 3075 3076
	if (!task)
		goto out;

3077
	result = proc_pid_instantiate(dir, dentry, task, NULL);
L
Linus Torvalds 已提交
3078 3079
	put_task_struct(task);
out:
3080
	return ERR_PTR(result);
L
Linus Torvalds 已提交
3081 3082 3083
}

/*
3084
 * Find the first task with tgid >= tgid
3085
 *
L
Linus Torvalds 已提交
3086
 */
3087 3088
struct tgid_iter {
	unsigned int tgid;
3089
	struct task_struct *task;
3090 3091 3092
};
static struct tgid_iter next_tgid(struct pid_namespace *ns, struct tgid_iter iter)
{
3093
	struct pid *pid;
L
Linus Torvalds 已提交
3094

3095 3096
	if (iter.task)
		put_task_struct(iter.task);
3097
	rcu_read_lock();
3098
retry:
3099 3100
	iter.task = NULL;
	pid = find_ge_pid(iter.tgid, ns);
3101
	if (pid) {
3102 3103
		iter.tgid = pid_nr_ns(pid, ns);
		iter.task = pid_task(pid, PIDTYPE_PID);
3104 3105 3106 3107 3108 3109 3110 3111 3112 3113 3114 3115
		/* 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.
		 */
3116 3117
		if (!iter.task || !has_group_leader_pid(iter.task)) {
			iter.tgid += 1;
3118
			goto retry;
3119 3120
		}
		get_task_struct(iter.task);
3121
	}
3122
	rcu_read_unlock();
3123
	return iter;
L
Linus Torvalds 已提交
3124 3125
}

3126
#define TGID_OFFSET (FIRST_PROCESS_ENTRY + 2)
3127

L
Linus Torvalds 已提交
3128
/* for the /proc/ directory itself, after non-process stuff has been done */
A
Al Viro 已提交
3129
int proc_pid_readdir(struct file *file, struct dir_context *ctx)
L
Linus Torvalds 已提交
3130
{
3131
	struct tgid_iter iter;
A
Al Viro 已提交
3132
	struct pid_namespace *ns = file_inode(file)->i_sb->s_fs_info;
A
Al Viro 已提交
3133
	loff_t pos = ctx->pos;
L
Linus Torvalds 已提交
3134

3135
	if (pos >= PID_MAX_LIMIT + TGID_OFFSET)
A
Al Viro 已提交
3136
		return 0;
L
Linus Torvalds 已提交
3137

3138
	if (pos == TGID_OFFSET - 2) {
3139
		struct inode *inode = d_inode(ns->proc_self);
3140
		if (!dir_emit(ctx, "self", 4, inode->i_ino, DT_LNK))
A
Al Viro 已提交
3141
			return 0;
3142 3143 3144
		ctx->pos = pos = pos + 1;
	}
	if (pos == TGID_OFFSET - 1) {
3145
		struct inode *inode = d_inode(ns->proc_thread_self);
3146 3147 3148
		if (!dir_emit(ctx, "thread-self", 11, inode->i_ino, DT_LNK))
			return 0;
		ctx->pos = pos = pos + 1;
3149
	}
3150
	iter.tgid = pos - TGID_OFFSET;
3151 3152 3153 3154
	iter.task = NULL;
	for (iter = next_tgid(ns, iter);
	     iter.task;
	     iter.tgid += 1, iter = next_tgid(ns, iter)) {
A
Al Viro 已提交
3155 3156 3157 3158
		char name[PROC_NUMBUF];
		int len;
		if (!has_pid_permissions(ns, iter.task, 2))
			continue;
3159

A
Al Viro 已提交
3160 3161 3162 3163
		len = snprintf(name, sizeof(name), "%d", iter.tgid);
		ctx->pos = iter.tgid + TGID_OFFSET;
		if (!proc_fill_cache(file, ctx, name, len,
				     proc_pid_instantiate, iter.task, NULL)) {
3164
			put_task_struct(iter.task);
A
Al Viro 已提交
3165
			return 0;
L
Linus Torvalds 已提交
3166
		}
3167
	}
A
Al Viro 已提交
3168
	ctx->pos = PID_MAX_LIMIT + TGID_OFFSET;
3169 3170
	return 0;
}
L
Linus Torvalds 已提交
3171

3172 3173 3174 3175 3176 3177 3178 3179 3180 3181 3182 3183 3184 3185 3186 3187 3188 3189 3190 3191 3192 3193 3194 3195 3196 3197 3198 3199 3200 3201 3202 3203 3204 3205 3206 3207 3208 3209
/*
 * 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,
};

3210 3211 3212
/*
 * Tasks
 */
3213
static const struct pid_entry tid_base_stuff[] = {
A
Alexey Dobriyan 已提交
3214
	DIR("fd",        S_IRUSR|S_IXUSR, proc_fd_inode_operations, proc_fd_operations),
J
Jerome Marchand 已提交
3215
	DIR("fdinfo",    S_IRUSR|S_IXUSR, proc_fdinfo_inode_operations, proc_fdinfo_operations),
3216
	DIR("ns",	 S_IRUSR|S_IXUGO, proc_ns_dir_inode_operations, proc_ns_dir_operations),
3217 3218 3219
#ifdef CONFIG_NET
	DIR("net",        S_IRUGO|S_IXUGO, proc_net_inode_operations, proc_net_operations),
#endif
A
Alexey Dobriyan 已提交
3220
	REG("environ",   S_IRUSR, proc_environ_operations),
3221
	ONE("auxv",      S_IRUSR, proc_pid_auxv),
A
Alexey Dobriyan 已提交
3222
	ONE("status",    S_IRUGO, proc_pid_status),
3223
	ONE("personality", S_IRUSR, proc_pid_personality),
3224
	ONE("limits",	 S_IRUGO, proc_pid_limits),
I
Ingo Molnar 已提交
3225
#ifdef CONFIG_SCHED_DEBUG
A
Alexey Dobriyan 已提交
3226
	REG("sched",     S_IRUGO|S_IWUSR, proc_pid_sched_operations),
R
Roland McGrath 已提交
3227
#endif
3228 3229 3230
	NOD("comm",      S_IFREG|S_IRUGO|S_IWUSR,
			 &proc_tid_comm_inode_operations,
			 &proc_pid_set_comm_operations, {}),
R
Roland McGrath 已提交
3231
#ifdef CONFIG_HAVE_ARCH_TRACEHOOK
3232
	ONE("syscall",   S_IRUSR, proc_pid_syscall),
I
Ingo Molnar 已提交
3233
#endif
3234
	REG("cmdline",   S_IRUGO, proc_pid_cmdline_ops),
A
Alexey Dobriyan 已提交
3235 3236
	ONE("stat",      S_IRUGO, proc_tid_stat),
	ONE("statm",     S_IRUGO, proc_pid_statm),
3237
	REG("maps",      S_IRUGO, proc_tid_maps_operations),
3238
#ifdef CONFIG_PROC_CHILDREN
3239 3240
	REG("children",  S_IRUGO, proc_tid_children_operations),
#endif
3241
#ifdef CONFIG_NUMA
3242
	REG("numa_maps", S_IRUGO, proc_tid_numa_maps_operations),
3243
#endif
A
Alexey Dobriyan 已提交
3244 3245 3246 3247 3248 3249
	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),
3250
#ifdef CONFIG_PROC_PAGE_MONITOR
A
Alexey Dobriyan 已提交
3251
	REG("clear_refs", S_IWUSR, proc_clear_refs_operations),
3252
	REG("smaps",     S_IRUGO, proc_tid_smaps_operations),
3253
	REG("pagemap",    S_IRUSR, proc_pagemap_operations),
3254 3255
#endif
#ifdef CONFIG_SECURITY
A
Alexey Dobriyan 已提交
3256
	DIR("attr",      S_IRUGO|S_IXUGO, proc_attr_dir_inode_operations, proc_attr_dir_operations),
3257 3258
#endif
#ifdef CONFIG_KALLSYMS
3259
	ONE("wchan",     S_IRUGO, proc_pid_wchan),
3260
#endif
K
Ken Chen 已提交
3261
#ifdef CONFIG_STACKTRACE
3262
	ONE("stack",      S_IRUSR, proc_pid_stack),
3263
#endif
3264
#ifdef CONFIG_SCHED_INFO
3265
	ONE("schedstat", S_IRUGO, proc_pid_schedstat),
3266
#endif
A
Arjan van de Ven 已提交
3267
#ifdef CONFIG_LATENCYTOP
A
Alexey Dobriyan 已提交
3268
	REG("latency",  S_IRUGO, proc_lstats_operations),
A
Arjan van de Ven 已提交
3269
#endif
3270
#ifdef CONFIG_PROC_PID_CPUSET
Z
Zefan Li 已提交
3271
	ONE("cpuset",    S_IRUGO, proc_cpuset_show),
3272 3273
#endif
#ifdef CONFIG_CGROUPS
Z
Zefan Li 已提交
3274
	ONE("cgroup",  S_IRUGO, proc_cgroup_show),
3275
#endif
3276
	ONE("oom_score", S_IRUGO, proc_oom_score),
3277
	REG("oom_adj",   S_IRUGO|S_IWUSR, proc_oom_adj_operations),
D
David Rientjes 已提交
3278
	REG("oom_score_adj", S_IRUGO|S_IWUSR, proc_oom_score_adj_operations),
3279
#ifdef CONFIG_AUDITSYSCALL
A
Alexey Dobriyan 已提交
3280
	REG("loginuid",  S_IWUSR|S_IRUGO, proc_loginuid_operations),
3281
	REG("sessionid",  S_IRUGO, proc_sessionid_operations),
3282
#endif
3283
#ifdef CONFIG_FAULT_INJECTION
A
Alexey Dobriyan 已提交
3284
	REG("make-it-fail", S_IRUGO|S_IWUSR, proc_fault_inject_operations),
3285
#endif
3286
#ifdef CONFIG_TASK_IO_ACCOUNTING
3287
	ONE("io",	S_IRUSR, proc_tid_io_accounting),
3288
#endif
3289
#ifdef CONFIG_HARDWALL
3290
	ONE("hardwall",   S_IRUGO, proc_pid_hardwall),
3291
#endif
3292 3293 3294
#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),
3295
	REG("projid_map", S_IRUGO|S_IWUSR, proc_projid_map_operations),
3296
	REG("setgroups",  S_IRUGO|S_IWUSR, proc_setgroups_operations),
3297
#endif
3298 3299
};

A
Al Viro 已提交
3300
static int proc_tid_base_readdir(struct file *file, struct dir_context *ctx)
3301
{
A
Al Viro 已提交
3302 3303
	return proc_pident_readdir(file, ctx,
				   tid_base_stuff, ARRAY_SIZE(tid_base_stuff));
3304 3305
}

A
Al Viro 已提交
3306 3307
static struct dentry *proc_tid_base_lookup(struct inode *dir, struct dentry *dentry, unsigned int flags)
{
3308 3309
	return proc_pident_lookup(dir, dentry,
				  tid_base_stuff, ARRAY_SIZE(tid_base_stuff));
3310 3311
}

3312
static const struct file_operations proc_tid_base_operations = {
3313
	.read		= generic_read_dir,
3314 3315
	.iterate_shared	= proc_tid_base_readdir,
	.llseek		= generic_file_llseek,
3316 3317
};

3318
static const struct inode_operations proc_tid_base_inode_operations = {
3319 3320 3321 3322 3323
	.lookup		= proc_tid_base_lookup,
	.getattr	= pid_getattr,
	.setattr	= proc_setattr,
};

3324
static int proc_task_instantiate(struct inode *dir,
3325
	struct dentry *dentry, struct task_struct *task, const void *ptr)
3326 3327
{
	struct inode *inode;
3328
	inode = proc_pid_make_inode(dir->i_sb, task);
3329 3330 3331 3332 3333 3334 3335

	if (!inode)
		goto out;
	inode->i_mode = S_IFDIR|S_IRUGO|S_IXUGO;
	inode->i_op = &proc_tid_base_inode_operations;
	inode->i_fop = &proc_tid_base_operations;
	inode->i_flags|=S_IMMUTABLE;
3336

M
Miklos Szeredi 已提交
3337 3338
	set_nlink(inode, 2 + pid_entry_count_dirs(tid_base_stuff,
						  ARRAY_SIZE(tid_base_stuff)));
3339

3340
	d_set_d_op(dentry, &pid_dentry_operations);
3341 3342 3343

	d_add(dentry, inode);
	/* Close the race of the process dying before we return the dentry */
3344
	if (pid_revalidate(dentry, 0))
3345
		return 0;
3346
out:
3347
	return -ENOENT;
3348 3349
}

A
Al Viro 已提交
3350
static struct dentry *proc_task_lookup(struct inode *dir, struct dentry * dentry, unsigned int flags)
3351
{
3352
	int result = -ENOENT;
3353 3354 3355
	struct task_struct *task;
	struct task_struct *leader = get_proc_task(dir);
	unsigned tid;
3356
	struct pid_namespace *ns;
3357 3358 3359 3360

	if (!leader)
		goto out_no_task;

3361
	tid = name_to_int(&dentry->d_name);
3362 3363 3364
	if (tid == ~0U)
		goto out;

3365
	ns = dentry->d_sb->s_fs_info;
3366
	rcu_read_lock();
3367
	task = find_task_by_pid_ns(tid, ns);
3368 3369 3370 3371 3372
	if (task)
		get_task_struct(task);
	rcu_read_unlock();
	if (!task)
		goto out;
3373
	if (!same_thread_group(leader, task))
3374 3375
		goto out_drop_task;

3376
	result = proc_task_instantiate(dir, dentry, task, NULL);
3377 3378 3379 3380 3381
out_drop_task:
	put_task_struct(task);
out:
	put_task_struct(leader);
out_no_task:
3382
	return ERR_PTR(result);
3383 3384
}

3385 3386 3387 3388 3389 3390 3391 3392 3393 3394 3395 3396
/*
 * 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.
 */
3397 3398
static struct task_struct *first_tid(struct pid *pid, int tid, loff_t f_pos,
					struct pid_namespace *ns)
3399
{
3400
	struct task_struct *pos, *task;
3401 3402 3403 3404
	unsigned long nr = f_pos;

	if (nr != f_pos)	/* 32bit overflow? */
		return NULL;
L
Linus Torvalds 已提交
3405

3406
	rcu_read_lock();
3407 3408 3409 3410 3411
	task = pid_task(pid, PIDTYPE_PID);
	if (!task)
		goto fail;

	/* Attempt to start with the tid of a thread */
3412
	if (tid && nr) {
3413
		pos = find_task_by_pid_ns(tid, ns);
3414
		if (pos && same_thread_group(pos, task))
O
Oleg Nesterov 已提交
3415
			goto found;
3416
	}
L
Linus Torvalds 已提交
3417

3418
	/* If nr exceeds the number of threads there is nothing todo */
3419
	if (nr >= get_nr_threads(task))
3420
		goto fail;
L
Linus Torvalds 已提交
3421

O
Oleg Nesterov 已提交
3422 3423
	/* If we haven't found our starting place yet start
	 * with the leader and walk nr threads forward.
3424
	 */
3425
	pos = task = task->group_leader;
3426
	do {
3427
		if (!nr--)
3428
			goto found;
3429
	} while_each_thread(task, pos);
3430 3431 3432
fail:
	pos = NULL;
	goto out;
O
Oleg Nesterov 已提交
3433 3434 3435
found:
	get_task_struct(pos);
out:
3436
	rcu_read_unlock();
3437 3438 3439 3440 3441 3442 3443 3444 3445 3446 3447
	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
3448
	struct task_struct *pos = NULL;
3449
	rcu_read_lock();
O
Oleg Nesterov 已提交
3450
	if (pid_alive(start)) {
3451
		pos = next_thread(start);
O
Oleg Nesterov 已提交
3452 3453 3454 3455 3456
		if (thread_group_leader(pos))
			pos = NULL;
		else
			get_task_struct(pos);
	}
3457
	rcu_read_unlock();
3458 3459
	put_task_struct(start);
	return pos;
L
Linus Torvalds 已提交
3460 3461 3462
}

/* for the /proc/TGID/task/ directories */
A
Al Viro 已提交
3463
static int proc_task_readdir(struct file *file, struct dir_context *ctx)
L
Linus Torvalds 已提交
3464
{
3465 3466
	struct inode *inode = file_inode(file);
	struct task_struct *task;
3467
	struct pid_namespace *ns;
A
Al Viro 已提交
3468
	int tid;
L
Linus Torvalds 已提交
3469

3470
	if (proc_inode_is_dead(inode))
A
Al Viro 已提交
3471
		return -ENOENT;
L
Linus Torvalds 已提交
3472

A
Al Viro 已提交
3473
	if (!dir_emit_dots(file, ctx))
3474
		return 0;
L
Linus Torvalds 已提交
3475

3476 3477 3478
	/* 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
3479
	ns = inode->i_sb->s_fs_info;
A
Al Viro 已提交
3480 3481
	tid = (int)file->f_version;
	file->f_version = 0;
3482
	for (task = first_tid(proc_pid(inode), tid, ctx->pos - 2, ns);
3483
	     task;
A
Al Viro 已提交
3484 3485 3486
	     task = next_tid(task), ctx->pos++) {
		char name[PROC_NUMBUF];
		int len;
3487
		tid = task_pid_nr_ns(task, ns);
A
Al Viro 已提交
3488 3489 3490
		len = snprintf(name, sizeof(name), "%d", tid);
		if (!proc_fill_cache(file, ctx, name, len,
				proc_task_instantiate, task, NULL)) {
3491 3492
			/* returning this tgid failed, save it as the first
			 * pid for the next readir call */
A
Al Viro 已提交
3493
			file->f_version = (u64)tid;
3494
			put_task_struct(task);
L
Linus Torvalds 已提交
3495
			break;
3496
		}
L
Linus Torvalds 已提交
3497
	}
3498

A
Al Viro 已提交
3499
	return 0;
L
Linus Torvalds 已提交
3500
}
3501 3502 3503

static int proc_task_getattr(struct vfsmount *mnt, struct dentry *dentry, struct kstat *stat)
{
3504
	struct inode *inode = d_inode(dentry);
3505
	struct task_struct *p = get_proc_task(inode);
3506 3507
	generic_fillattr(inode, stat);

3508 3509 3510
	if (p) {
		stat->nlink += get_nr_threads(p);
		put_task_struct(p);
3511 3512 3513 3514
	}

	return 0;
}
3515

3516
static const struct inode_operations proc_task_inode_operations = {
3517 3518 3519
	.lookup		= proc_task_lookup,
	.getattr	= proc_task_getattr,
	.setattr	= proc_setattr,
3520
	.permission	= proc_pid_permission,
3521 3522
};

3523
static const struct file_operations proc_task_operations = {
3524
	.read		= generic_read_dir,
3525 3526
	.iterate_shared	= proc_task_readdir,
	.llseek		= generic_file_llseek,
3527
};