generic.c 19.4 KB
Newer Older
L
Linus Torvalds 已提交
1 2 3 4 5 6 7 8 9 10 11 12 13 14
/*
 * proc/fs/generic.c --- generic routines for the proc-fs
 *
 * This file contains generic proc-fs routines for handling
 * directories and files.
 * 
 * Copyright (C) 1991, 1992 Linus Torvalds.
 * Copyright (C) 1997 Theodore Ts'o
 */

#include <linux/errno.h>
#include <linux/time.h>
#include <linux/proc_fs.h>
#include <linux/stat.h>
C
Christoph Hellwig 已提交
15
#include <linux/mm.h>
L
Linus Torvalds 已提交
16
#include <linux/module.h>
17
#include <linux/slab.h>
L
Linus Torvalds 已提交
18 19 20 21 22
#include <linux/mount.h>
#include <linux/init.h>
#include <linux/idr.h>
#include <linux/namei.h>
#include <linux/bitops.h>
23
#include <linux/spinlock.h>
24
#include <linux/completion.h>
L
Linus Torvalds 已提交
25 26
#include <asm/uaccess.h>

27 28
#include "internal.h"

29 30
DEFINE_SPINLOCK(proc_subdir_lock);

31
static int proc_match(unsigned int len, const char *name, struct proc_dir_entry *de)
L
Linus Torvalds 已提交
32 33 34 35 36 37 38 39 40 41
{
	if (de->namelen != len)
		return 0;
	return !memcmp(name, de->name, len);
}

/* buffer size is one page but our output routines use some slack for overruns */
#define PROC_BLOCK_SIZE	(PAGE_SIZE - 1024)

static ssize_t
42
__proc_file_read(struct file *file, char __user *buf, size_t nbytes,
L
Linus Torvalds 已提交
43 44
	       loff_t *ppos)
{
45
	struct inode * inode = file->f_path.dentry->d_inode;
L
Linus Torvalds 已提交
46 47 48 49 50 51
	char 	*page;
	ssize_t	retval=0;
	int	eof=0;
	ssize_t	n, count;
	char	*start;
	struct proc_dir_entry * dp;
L
Linus Torvalds 已提交
52 53 54 55 56 57 58 59 60 61 62 63
	unsigned long long pos;

	/*
	 * Gaah, please just use "seq_file" instead. The legacy /proc
	 * interfaces cut loff_t down to off_t for reads, and ignore
	 * the offset entirely for writes..
	 */
	pos = *ppos;
	if (pos > MAX_NON_LFS)
		return 0;
	if (nbytes > MAX_NON_LFS - pos)
		nbytes = MAX_NON_LFS - pos;
L
Linus Torvalds 已提交
64 65

	dp = PDE(inode);
66
	if (!(page = (char*) __get_free_page(GFP_TEMPORARY)))
L
Linus Torvalds 已提交
67 68 69 70 71 72
		return -ENOMEM;

	while ((nbytes > 0) && !eof) {
		count = min_t(size_t, PROC_BLOCK_SIZE, nbytes);

		start = NULL;
73
		if (dp->read_proc) {
L
Linus Torvalds 已提交
74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186
			/*
			 * How to be a proc read function
			 * ------------------------------
			 * Prototype:
			 *    int f(char *buffer, char **start, off_t offset,
			 *          int count, int *peof, void *dat)
			 *
			 * Assume that the buffer is "count" bytes in size.
			 *
			 * If you know you have supplied all the data you
			 * have, set *peof.
			 *
			 * You have three ways to return data:
			 * 0) Leave *start = NULL.  (This is the default.)
			 *    Put the data of the requested offset at that
			 *    offset within the buffer.  Return the number (n)
			 *    of bytes there are from the beginning of the
			 *    buffer up to the last byte of data.  If the
			 *    number of supplied bytes (= n - offset) is 
			 *    greater than zero and you didn't signal eof
			 *    and the reader is prepared to take more data
			 *    you will be called again with the requested
			 *    offset advanced by the number of bytes 
			 *    absorbed.  This interface is useful for files
			 *    no larger than the buffer.
			 * 1) Set *start = an unsigned long value less than
			 *    the buffer address but greater than zero.
			 *    Put the data of the requested offset at the
			 *    beginning of the buffer.  Return the number of
			 *    bytes of data placed there.  If this number is
			 *    greater than zero and you didn't signal eof
			 *    and the reader is prepared to take more data
			 *    you will be called again with the requested
			 *    offset advanced by *start.  This interface is
			 *    useful when you have a large file consisting
			 *    of a series of blocks which you want to count
			 *    and return as wholes.
			 *    (Hack by Paul.Russell@rustcorp.com.au)
			 * 2) Set *start = an address within the buffer.
			 *    Put the data of the requested offset at *start.
			 *    Return the number of bytes of data placed there.
			 *    If this number is greater than zero and you
			 *    didn't signal eof and the reader is prepared to
			 *    take more data you will be called again with the
			 *    requested offset advanced by the number of bytes
			 *    absorbed.
			 */
			n = dp->read_proc(page, &start, *ppos,
					  count, &eof, dp->data);
		} else
			break;

		if (n == 0)   /* end of file */
			break;
		if (n < 0) {  /* error */
			if (retval == 0)
				retval = n;
			break;
		}

		if (start == NULL) {
			if (n > PAGE_SIZE) {
				printk(KERN_ERR
				       "proc_file_read: Apparent buffer overflow!\n");
				n = PAGE_SIZE;
			}
			n -= *ppos;
			if (n <= 0)
				break;
			if (n > count)
				n = count;
			start = page + *ppos;
		} else if (start < page) {
			if (n > PAGE_SIZE) {
				printk(KERN_ERR
				       "proc_file_read: Apparent buffer overflow!\n");
				n = PAGE_SIZE;
			}
			if (n > count) {
				/*
				 * Don't reduce n because doing so might
				 * cut off part of a data block.
				 */
				printk(KERN_WARNING
				       "proc_file_read: Read count exceeded\n");
			}
		} else /* start >= page */ {
			unsigned long startoff = (unsigned long)(start - page);
			if (n > (PAGE_SIZE - startoff)) {
				printk(KERN_ERR
				       "proc_file_read: Apparent buffer overflow!\n");
				n = PAGE_SIZE - startoff;
			}
			if (n > count)
				n = count;
		}
		
 		n -= copy_to_user(buf, start < page ? page : start, n);
		if (n == 0) {
			if (retval == 0)
				retval = -EFAULT;
			break;
		}

		*ppos += start < page ? (unsigned long)start : n;
		nbytes -= n;
		buf += n;
		retval += n;
	}
	free_page((unsigned long) page);
	return retval;
}

187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207
static ssize_t
proc_file_read(struct file *file, char __user *buf, size_t nbytes,
	       loff_t *ppos)
{
	struct proc_dir_entry *pde = PDE(file->f_path.dentry->d_inode);
	ssize_t rv = -EIO;

	spin_lock(&pde->pde_unload_lock);
	if (!pde->proc_fops) {
		spin_unlock(&pde->pde_unload_lock);
		return rv;
	}
	pde->pde_users++;
	spin_unlock(&pde->pde_unload_lock);

	rv = __proc_file_read(file, buf, nbytes, ppos);

	pde_users_dec(pde);
	return rv;
}

L
Linus Torvalds 已提交
208 209 210 211
static ssize_t
proc_file_write(struct file *file, const char __user *buffer,
		size_t count, loff_t *ppos)
{
212 213 214 215 216 217 218 219 220 221 222
	struct proc_dir_entry *pde = PDE(file->f_path.dentry->d_inode);
	ssize_t rv = -EIO;

	if (pde->write_proc) {
		spin_lock(&pde->pde_unload_lock);
		if (!pde->proc_fops) {
			spin_unlock(&pde->pde_unload_lock);
			return rv;
		}
		pde->pde_users++;
		spin_unlock(&pde->pde_unload_lock);
L
Linus Torvalds 已提交
223

224 225 226 227 228
		/* FIXME: does this routine need ppos?  probably... */
		rv = pde->write_proc(file, buffer, count, pde->data);
		pde_users_dec(pde);
	}
	return rv;
L
Linus Torvalds 已提交
229 230 231 232 233 234
}


static loff_t
proc_file_lseek(struct file *file, loff_t offset, int orig)
{
L
Linus Torvalds 已提交
235 236 237 238 239 240 241 242 243 244 245
	loff_t retval = -EINVAL;
	switch (orig) {
	case 1:
		offset += file->f_pos;
	/* fallthrough */
	case 0:
		if (offset < 0 || offset > MAX_NON_LFS)
			break;
		file->f_pos = retval = offset;
	}
	return retval;
L
Linus Torvalds 已提交
246 247
}

248 249 250 251 252 253
static const struct file_operations proc_file_operations = {
	.llseek		= proc_file_lseek,
	.read		= proc_file_read,
	.write		= proc_file_write,
};

L
Linus Torvalds 已提交
254 255 256 257 258 259 260 261
static int proc_notify_change(struct dentry *dentry, struct iattr *iattr)
{
	struct inode *inode = dentry->d_inode;
	struct proc_dir_entry *de = PDE(inode);
	int error;

	error = inode_change_ok(inode, iattr);
	if (error)
C
Christoph Hellwig 已提交
262
		return error;
L
Linus Torvalds 已提交
263

C
Christoph Hellwig 已提交
264 265 266 267 268 269 270 271 272
	if ((iattr->ia_valid & ATTR_SIZE) &&
	    iattr->ia_size != i_size_read(inode)) {
		error = vmtruncate(inode, iattr->ia_size);
		if (error)
			return error;
	}

	setattr_copy(inode, iattr);
	mark_inode_dirty(inode);
L
Linus Torvalds 已提交
273 274 275 276
	
	de->uid = inode->i_uid;
	de->gid = inode->i_gid;
	de->mode = inode->i_mode;
C
Christoph Hellwig 已提交
277
	return 0;
L
Linus Torvalds 已提交
278 279
}

M
Miklos Szeredi 已提交
280 281 282 283 284 285
static int proc_getattr(struct vfsmount *mnt, struct dentry *dentry,
			struct kstat *stat)
{
	struct inode *inode = dentry->d_inode;
	struct proc_dir_entry *de = PROC_I(inode)->pde;
	if (de && de->nlink)
M
Miklos Szeredi 已提交
286
		set_nlink(inode, de->nlink);
M
Miklos Szeredi 已提交
287 288 289 290 291

	generic_fillattr(inode, stat);
	return 0;
}

292
static const struct inode_operations proc_file_inode_operations = {
L
Linus Torvalds 已提交
293 294 295 296 297 298 299 300
	.setattr	= proc_notify_change,
};

/*
 * This function parses a name such as "tty/driver/serial", and
 * returns the struct proc_dir_entry for "/proc/tty/driver", and
 * returns "serial" in residual.
 */
301 302
static int __xlate_proc_name(const char *name, struct proc_dir_entry **ret,
			     const char **residual)
L
Linus Torvalds 已提交
303 304 305
{
	const char     		*cp = name, *next;
	struct proc_dir_entry	*de;
306
	unsigned int		len;
L
Linus Torvalds 已提交
307

308 309 310 311
	de = *ret;
	if (!de)
		de = &proc_root;

L
Linus Torvalds 已提交
312 313 314 315 316 317 318 319 320 321
	while (1) {
		next = strchr(cp, '/');
		if (!next)
			break;

		len = next - cp;
		for (de = de->subdir; de ; de = de->next) {
			if (proc_match(len, cp, de))
				break;
		}
322 323
		if (!de) {
			WARN(1, "name '%s'\n", name);
324
			return -ENOENT;
325
		}
L
Linus Torvalds 已提交
326 327 328 329
		cp += len + 1;
	}
	*residual = cp;
	*ret = de;
330 331 332 333 334 335 336 337 338 339
	return 0;
}

static int xlate_proc_name(const char *name, struct proc_dir_entry **ret,
			   const char **residual)
{
	int rv;

	spin_lock(&proc_subdir_lock);
	rv = __xlate_proc_name(name, ret, residual);
340
	spin_unlock(&proc_subdir_lock);
341
	return rv;
L
Linus Torvalds 已提交
342 343
}

344
static DEFINE_IDA(proc_inum_ida);
L
Linus Torvalds 已提交
345 346
static DEFINE_SPINLOCK(proc_inum_lock); /* protects the above */

347
#define PROC_DYNAMIC_FIRST 0xF0000000U
L
Linus Torvalds 已提交
348 349 350 351 352 353 354

/*
 * Return an inode number between PROC_DYNAMIC_FIRST and
 * 0xffffffff, or zero on failure.
 */
static unsigned int get_inode_number(void)
{
355
	unsigned int i;
L
Linus Torvalds 已提交
356 357 358
	int error;

retry:
359
	if (ida_pre_get(&proc_inum_ida, GFP_KERNEL) == 0)
L
Linus Torvalds 已提交
360 361 362
		return 0;

	spin_lock(&proc_inum_lock);
363
	error = ida_get_new(&proc_inum_ida, &i);
L
Linus Torvalds 已提交
364 365 366 367 368 369
	spin_unlock(&proc_inum_lock);
	if (error == -EAGAIN)
		goto retry;
	else if (error)
		return 0;

370 371
	if (i > UINT_MAX - PROC_DYNAMIC_FIRST) {
		spin_lock(&proc_inum_lock);
372
		ida_remove(&proc_inum_ida, i);
373
		spin_unlock(&proc_inum_lock);
374
		return 0;
375 376
	}
	return PROC_DYNAMIC_FIRST + i;
L
Linus Torvalds 已提交
377 378 379 380 381
}

static void release_inode_number(unsigned int inum)
{
	spin_lock(&proc_inum_lock);
382
	ida_remove(&proc_inum_ida, inum - PROC_DYNAMIC_FIRST);
L
Linus Torvalds 已提交
383 384 385
	spin_unlock(&proc_inum_lock);
}

386
static void *proc_follow_link(struct dentry *dentry, struct nameidata *nd)
L
Linus Torvalds 已提交
387 388
{
	nd_set_link(nd, PDE(dentry->d_inode)->data);
389
	return NULL;
L
Linus Torvalds 已提交
390 391
}

392
static const struct inode_operations proc_link_inode_operations = {
L
Linus Torvalds 已提交
393 394 395 396 397 398 399 400 401 402
	.readlink	= generic_readlink,
	.follow_link	= proc_follow_link,
};

/*
 * As some entries in /proc are volatile, we want to 
 * get rid of unused dentries.  This could be made 
 * smarter: we could keep a "volatile" flag in the 
 * inode to indicate which ones to keep.
 */
N
Nick Piggin 已提交
403
static int proc_delete_dentry(const struct dentry * dentry)
L
Linus Torvalds 已提交
404 405 406 407
{
	return 1;
}

A
Al Viro 已提交
408
static const struct dentry_operations proc_dentry_operations =
L
Linus Torvalds 已提交
409 410 411 412 413 414 415 416
{
	.d_delete	= proc_delete_dentry,
};

/*
 * Don't create negative dentries here, return -ENOENT by hand
 * instead.
 */
417 418
struct dentry *proc_lookup_de(struct proc_dir_entry *de, struct inode *dir,
		struct dentry *dentry)
L
Linus Torvalds 已提交
419 420 421 422
{
	struct inode *inode = NULL;
	int error = -ENOENT;

423
	spin_lock(&proc_subdir_lock);
424 425 426 427
	for (de = de->subdir; de ; de = de->next) {
		if (de->namelen != dentry->d_name.len)
			continue;
		if (!memcmp(dentry->d_name.name, de->name, de->namelen)) {
428
			pde_get(de);
429
			spin_unlock(&proc_subdir_lock);
430
			error = -ENOMEM;
A
Alexey Dobriyan 已提交
431
			inode = proc_get_inode(dir->i_sb, de);
432
			goto out_unlock;
L
Linus Torvalds 已提交
433 434
		}
	}
435
	spin_unlock(&proc_subdir_lock);
436
out_unlock:
L
Linus Torvalds 已提交
437 438

	if (inode) {
439
		d_set_d_op(dentry, &proc_dentry_operations);
L
Linus Torvalds 已提交
440 441 442
		d_add(dentry, inode);
		return NULL;
	}
443
	if (de)
444
		pde_put(de);
L
Linus Torvalds 已提交
445 446 447
	return ERR_PTR(error);
}

448
struct dentry *proc_lookup(struct inode *dir, struct dentry *dentry,
A
Al Viro 已提交
449
		unsigned int flags)
450 451 452 453
{
	return proc_lookup_de(PDE(dir), dir, dentry);
}

L
Linus Torvalds 已提交
454 455 456 457 458 459 460 461 462
/*
 * This returns non-zero if at EOF, so that the /proc
 * root directory can use this and check if it should
 * continue with the <pid> entries..
 *
 * Note that the VFS-layer doesn't care about the return
 * value of the readdir() call, as long as it's non-negative
 * for success..
 */
463 464
int proc_readdir_de(struct proc_dir_entry *de, struct file *filp, void *dirent,
		filldir_t filldir)
L
Linus Torvalds 已提交
465 466 467
{
	unsigned int ino;
	int i;
468
	struct inode *inode = filp->f_path.dentry->d_inode;
L
Linus Torvalds 已提交
469 470 471 472 473 474 475 476 477 478 479 480 481
	int ret = 0;

	ino = inode->i_ino;
	i = filp->f_pos;
	switch (i) {
		case 0:
			if (filldir(dirent, ".", 1, i, ino, DT_DIR) < 0)
				goto out;
			i++;
			filp->f_pos++;
			/* fall through */
		case 1:
			if (filldir(dirent, "..", 2, i,
482
				    parent_ino(filp->f_path.dentry),
L
Linus Torvalds 已提交
483 484 485 486 487 488
				    DT_DIR) < 0)
				goto out;
			i++;
			filp->f_pos++;
			/* fall through */
		default:
489
			spin_lock(&proc_subdir_lock);
L
Linus Torvalds 已提交
490 491 492 493 494
			de = de->subdir;
			i -= 2;
			for (;;) {
				if (!de) {
					ret = 1;
495
					spin_unlock(&proc_subdir_lock);
L
Linus Torvalds 已提交
496 497 498 499 500 501 502 503 504
					goto out;
				}
				if (!i)
					break;
				de = de->next;
				i--;
			}

			do {
505 506
				struct proc_dir_entry *next;

507
				/* filldir passes info to user space */
508
				pde_get(de);
509
				spin_unlock(&proc_subdir_lock);
L
Linus Torvalds 已提交
510
				if (filldir(dirent, de->name, de->namelen, filp->f_pos,
511
					    de->low_ino, de->mode >> 12) < 0) {
512
					pde_put(de);
L
Linus Torvalds 已提交
513
					goto out;
514
				}
515
				spin_lock(&proc_subdir_lock);
L
Linus Torvalds 已提交
516
				filp->f_pos++;
517
				next = de->next;
518
				pde_put(de);
519
				de = next;
L
Linus Torvalds 已提交
520
			} while (de);
521
			spin_unlock(&proc_subdir_lock);
L
Linus Torvalds 已提交
522 523
	}
	ret = 1;
A
Alexey Dobriyan 已提交
524
out:
L
Linus Torvalds 已提交
525 526 527
	return ret;	
}

528 529 530 531 532 533 534
int proc_readdir(struct file *filp, void *dirent, filldir_t filldir)
{
	struct inode *inode = filp->f_path.dentry->d_inode;

	return proc_readdir_de(PDE(inode), filp, dirent, filldir);
}

L
Linus Torvalds 已提交
535 536 537 538 539
/*
 * These are the generic /proc directory operations. They
 * use the in-memory "struct proc_dir_entry" tree to parse
 * the /proc directory.
 */
540
static const struct file_operations proc_dir_operations = {
A
Alexey Dobriyan 已提交
541
	.llseek			= generic_file_llseek,
L
Linus Torvalds 已提交
542 543 544 545 546 547 548
	.read			= generic_read_dir,
	.readdir		= proc_readdir,
};

/*
 * proc directories can do almost nothing..
 */
549
static const struct inode_operations proc_dir_inode_operations = {
L
Linus Torvalds 已提交
550
	.lookup		= proc_lookup,
M
Miklos Szeredi 已提交
551
	.getattr	= proc_getattr,
L
Linus Torvalds 已提交
552 553 554 555 556 557
	.setattr	= proc_notify_change,
};

static int proc_register(struct proc_dir_entry * dir, struct proc_dir_entry * dp)
{
	unsigned int i;
558
	struct proc_dir_entry *tmp;
L
Linus Torvalds 已提交
559 560 561 562 563
	
	i = get_inode_number();
	if (i == 0)
		return -EAGAIN;
	dp->low_ino = i;
564

L
Linus Torvalds 已提交
565 566 567 568 569 570 571 572 573 574 575 576 577 578 579
	if (S_ISDIR(dp->mode)) {
		if (dp->proc_iops == NULL) {
			dp->proc_fops = &proc_dir_operations;
			dp->proc_iops = &proc_dir_inode_operations;
		}
		dir->nlink++;
	} else if (S_ISLNK(dp->mode)) {
		if (dp->proc_iops == NULL)
			dp->proc_iops = &proc_link_inode_operations;
	} else if (S_ISREG(dp->mode)) {
		if (dp->proc_fops == NULL)
			dp->proc_fops = &proc_file_operations;
		if (dp->proc_iops == NULL)
			dp->proc_iops = &proc_file_inode_operations;
	}
C
Changli Gao 已提交
580 581

	spin_lock(&proc_subdir_lock);
582 583 584

	for (tmp = dir->subdir; tmp; tmp = tmp->next)
		if (strcmp(tmp->name, dp->name) == 0) {
585
			WARN(1, KERN_WARNING "proc_dir_entry '%s/%s' already registered\n",
586
				dir->name, dp->name);
587 588 589
			break;
		}

C
Changli Gao 已提交
590 591 592 593 594
	dp->next = dir->subdir;
	dp->parent = dir;
	dir->subdir = dp;
	spin_unlock(&proc_subdir_lock);

L
Linus Torvalds 已提交
595 596 597
	return 0;
}

598
static struct proc_dir_entry *__proc_create(struct proc_dir_entry **parent,
L
Linus Torvalds 已提交
599
					  const char *name,
A
Al Viro 已提交
600
					  umode_t mode,
L
Linus Torvalds 已提交
601 602 603 604
					  nlink_t nlink)
{
	struct proc_dir_entry *ent = NULL;
	const char *fn = name;
605
	unsigned int len;
L
Linus Torvalds 已提交
606 607 608 609

	/* make sure name is valid */
	if (!name || !strlen(name)) goto out;

610
	if (xlate_proc_name(name, parent, &fn) != 0)
L
Linus Torvalds 已提交
611 612 613 614 615 616 617 618 619 620 621 622
		goto out;

	/* At this point there must not be any '/' characters beyond *fn */
	if (strchr(fn, '/'))
		goto out;

	len = strlen(fn);

	ent = kmalloc(sizeof(struct proc_dir_entry) + len + 1, GFP_KERNEL);
	if (!ent) goto out;

	memset(ent, 0, sizeof(struct proc_dir_entry));
623
	memcpy(ent->name, fn, len + 1);
L
Linus Torvalds 已提交
624 625 626
	ent->namelen = len;
	ent->mode = mode;
	ent->nlink = nlink;
627
	atomic_set(&ent->count, 1);
628 629 630
	ent->pde_users = 0;
	spin_lock_init(&ent->pde_unload_lock);
	ent->pde_unload_completion = NULL;
A
Alexey Dobriyan 已提交
631
	INIT_LIST_HEAD(&ent->pde_openers);
L
Linus Torvalds 已提交
632 633 634 635 636 637 638 639 640
 out:
	return ent;
}

struct proc_dir_entry *proc_symlink(const char *name,
		struct proc_dir_entry *parent, const char *dest)
{
	struct proc_dir_entry *ent;

641
	ent = __proc_create(&parent, name,
L
Linus Torvalds 已提交
642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659
			  (S_IFLNK | S_IRUGO | S_IWUGO | S_IXUGO),1);

	if (ent) {
		ent->data = kmalloc((ent->size=strlen(dest))+1, GFP_KERNEL);
		if (ent->data) {
			strcpy((char*)ent->data,dest);
			if (proc_register(parent, ent) < 0) {
				kfree(ent->data);
				kfree(ent);
				ent = NULL;
			}
		} else {
			kfree(ent);
			ent = NULL;
		}
	}
	return ent;
}
H
Helight.Xu 已提交
660
EXPORT_SYMBOL(proc_symlink);
L
Linus Torvalds 已提交
661

A
Al Viro 已提交
662
struct proc_dir_entry *proc_mkdir_mode(const char *name, umode_t mode,
L
Linus Torvalds 已提交
663 664 665 666
		struct proc_dir_entry *parent)
{
	struct proc_dir_entry *ent;

667
	ent = __proc_create(&parent, name, S_IFDIR | mode, 2);
L
Linus Torvalds 已提交
668 669 670 671 672 673 674 675
	if (ent) {
		if (proc_register(parent, ent) < 0) {
			kfree(ent);
			ent = NULL;
		}
	}
	return ent;
}
676
EXPORT_SYMBOL(proc_mkdir_mode);
L
Linus Torvalds 已提交
677

678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694
struct proc_dir_entry *proc_net_mkdir(struct net *net, const char *name,
		struct proc_dir_entry *parent)
{
	struct proc_dir_entry *ent;

	ent = __proc_create(&parent, name, S_IFDIR | S_IRUGO | S_IXUGO, 2);
	if (ent) {
		ent->data = net;
		if (proc_register(parent, ent) < 0) {
			kfree(ent);
			ent = NULL;
		}
	}
	return ent;
}
EXPORT_SYMBOL_GPL(proc_net_mkdir);

L
Linus Torvalds 已提交
695 696 697 698 699
struct proc_dir_entry *proc_mkdir(const char *name,
		struct proc_dir_entry *parent)
{
	return proc_mkdir_mode(name, S_IRUGO | S_IXUGO, parent);
}
H
Helight.Xu 已提交
700
EXPORT_SYMBOL(proc_mkdir);
L
Linus Torvalds 已提交
701

A
Al Viro 已提交
702
struct proc_dir_entry *create_proc_entry(const char *name, umode_t mode,
L
Linus Torvalds 已提交
703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719
					 struct proc_dir_entry *parent)
{
	struct proc_dir_entry *ent;
	nlink_t nlink;

	if (S_ISDIR(mode)) {
		if ((mode & S_IALLUGO) == 0)
			mode |= S_IRUGO | S_IXUGO;
		nlink = 2;
	} else {
		if ((mode & S_IFMT) == 0)
			mode |= S_IFREG;
		if ((mode & S_IALLUGO) == 0)
			mode |= S_IRUGO;
		nlink = 1;
	}

720
	ent = __proc_create(&parent, name, mode, nlink);
L
Linus Torvalds 已提交
721 722 723 724 725 726 727 728
	if (ent) {
		if (proc_register(parent, ent) < 0) {
			kfree(ent);
			ent = NULL;
		}
	}
	return ent;
}
H
Helight.Xu 已提交
729
EXPORT_SYMBOL(create_proc_entry);
L
Linus Torvalds 已提交
730

A
Al Viro 已提交
731
struct proc_dir_entry *proc_create_data(const char *name, umode_t mode,
732 733 734
					struct proc_dir_entry *parent,
					const struct file_operations *proc_fops,
					void *data)
735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754
{
	struct proc_dir_entry *pde;
	nlink_t nlink;

	if (S_ISDIR(mode)) {
		if ((mode & S_IALLUGO) == 0)
			mode |= S_IRUGO | S_IXUGO;
		nlink = 2;
	} else {
		if ((mode & S_IFMT) == 0)
			mode |= S_IFREG;
		if ((mode & S_IALLUGO) == 0)
			mode |= S_IRUGO;
		nlink = 1;
	}

	pde = __proc_create(&parent, name, mode, nlink);
	if (!pde)
		goto out;
	pde->proc_fops = proc_fops;
755
	pde->data = data;
756 757 758 759 760 761 762 763
	if (proc_register(parent, pde) < 0)
		goto out_free;
	return pde;
out_free:
	kfree(pde);
out:
	return NULL;
}
H
Helight.Xu 已提交
764
EXPORT_SYMBOL(proc_create_data);
765

766
static void free_proc_entry(struct proc_dir_entry *de)
L
Linus Torvalds 已提交
767
{
A
Alexey Dobriyan 已提交
768
	release_inode_number(de->low_ino);
L
Linus Torvalds 已提交
769

770
	if (S_ISLNK(de->mode))
L
Linus Torvalds 已提交
771 772 773 774
		kfree(de->data);
	kfree(de);
}

775 776 777 778 779 780
void pde_put(struct proc_dir_entry *pde)
{
	if (atomic_dec_and_test(&pde->count))
		free_proc_entry(pde);
}

L
Linus Torvalds 已提交
781 782 783 784 785 786
/*
 * Remove a /proc entry and free it if it's not currently in use.
 */
void remove_proc_entry(const char *name, struct proc_dir_entry *parent)
{
	struct proc_dir_entry **p;
787
	struct proc_dir_entry *de = NULL;
L
Linus Torvalds 已提交
788
	const char *fn = name;
789
	unsigned int len;
L
Linus Torvalds 已提交
790

791 792 793
	spin_lock(&proc_subdir_lock);
	if (__xlate_proc_name(name, &parent, &fn) != 0) {
		spin_unlock(&proc_subdir_lock);
794
		return;
795
	}
L
Linus Torvalds 已提交
796
	len = strlen(fn);
797

L
Linus Torvalds 已提交
798
	for (p = &parent->subdir; *p; p=&(*p)->next ) {
799 800 801 802 803 804 805 806
		if (proc_match(len, fn, *p)) {
			de = *p;
			*p = de->next;
			de->next = NULL;
			break;
		}
	}
	spin_unlock(&proc_subdir_lock);
807 808
	if (!de) {
		WARN(1, "name '%s'\n", name);
809
		return;
810
	}
811

812 813 814 815 816 817 818 819 820 821 822 823
	spin_lock(&de->pde_unload_lock);
	/*
	 * Stop accepting new callers into module. If you're
	 * dynamically allocating ->proc_fops, save a pointer somewhere.
	 */
	de->proc_fops = NULL;
	/* Wait until all existing callers into module are done. */
	if (de->pde_users > 0) {
		DECLARE_COMPLETION_ONSTACK(c);

		if (!de->pde_unload_completion)
			de->pde_unload_completion = &c;
824 825 826

		spin_unlock(&de->pde_unload_lock);

827 828
		wait_for_completion(de->pde_unload_completion);

829
		spin_lock(&de->pde_unload_lock);
830 831
	}

A
Alexey Dobriyan 已提交
832 833 834 835 836 837 838 839 840 841 842 843
	while (!list_empty(&de->pde_openers)) {
		struct pde_opener *pdeo;

		pdeo = list_first_entry(&de->pde_openers, struct pde_opener, lh);
		list_del(&pdeo->lh);
		spin_unlock(&de->pde_unload_lock);
		pdeo->release(pdeo->inode, pdeo->file);
		kfree(pdeo);
		spin_lock(&de->pde_unload_lock);
	}
	spin_unlock(&de->pde_unload_lock);

844 845 846
	if (S_ISDIR(de->mode))
		parent->nlink--;
	de->nlink = 0;
A
Arjan van de Ven 已提交
847
	WARN(de->subdir, KERN_WARNING "%s: removing non-empty directory "
848 849
			"'%s/%s', leaking at least '%s'\n", __func__,
			de->parent->name, de->name, de->subdir->name);
850
	pde_put(de);
L
Linus Torvalds 已提交
851
}
H
Helight.Xu 已提交
852
EXPORT_SYMBOL(remove_proc_entry);