expfs.c 12.2 KB
Newer Older
1 2 3 4 5 6 7 8
/*
 * Copyright (C) Neil Brown 2002
 * Copyright (C) Christoph Hellwig 2007
 *
 * This file contains the code mapping from inodes to NFS file handles,
 * and for mapping back from file handles to dentries.
 *
 * For details on why we do all the strange and hairy things in here
9
 * take a look at Documentation/filesystems/nfs/Exporting.
10
 */
11
#include <linux/exportfs.h>
Linus Torvalds's avatar
Linus Torvalds committed
12 13 14
#include <linux/fs.h>
#include <linux/file.h>
#include <linux/module.h>
15
#include <linux/mount.h>
Linus Torvalds's avatar
Linus Torvalds committed
16
#include <linux/namei.h>
17
#include <linux/sched.h>
Linus Torvalds's avatar
Linus Torvalds committed
18

19
#define dprintk(fmt, args...) do{}while(0)
Linus Torvalds's avatar
Linus Torvalds committed
20 21


22
static int get_name(struct vfsmount *mnt, struct dentry *dentry, char *name,
23 24 25
		struct dentry *child);


26 27
static int exportfs_get_name(struct vfsmount *mnt, struct dentry *dir,
		char *name, struct dentry *child)
28
{
29
	const struct export_operations *nop = dir->d_sb->s_export_op;
30 31 32 33

	if (nop->get_name)
		return nop->get_name(dir, name, child);
	else
34
		return get_name(mnt, dir, name, child);
35
}
Linus Torvalds's avatar
Linus Torvalds committed
36

37 38 39
/*
 * Check if the dentry or any of it's aliases is acceptable.
 */
40 41 42 43 44 45 46
static struct dentry *
find_acceptable_alias(struct dentry *result,
		int (*acceptable)(void *context, struct dentry *dentry),
		void *context)
{
	struct dentry *dentry, *toput = NULL;

47 48 49
	if (acceptable(context, result))
		return result;

Nick Piggin's avatar
Nick Piggin committed
50
	spin_lock(&dcache_inode_lock);
51 52
	list_for_each_entry(dentry, &result->d_inode->i_dentry, d_alias) {
		dget_locked(dentry);
Nick Piggin's avatar
Nick Piggin committed
53
		spin_unlock(&dcache_inode_lock);
54 55 56 57 58 59
		if (toput)
			dput(toput);
		if (dentry != result && acceptable(context, dentry)) {
			dput(result);
			return dentry;
		}
Nick Piggin's avatar
Nick Piggin committed
60
		spin_lock(&dcache_inode_lock);
61 62
		toput = dentry;
	}
Nick Piggin's avatar
Nick Piggin committed
63
	spin_unlock(&dcache_inode_lock);
64 65 66 67 68 69

	if (toput)
		dput(toput);
	return NULL;
}

70 71 72 73 74 75 76
/*
 * Find root of a disconnected subtree and return a reference to it.
 */
static struct dentry *
find_disconnected_root(struct dentry *dentry)
{
	dget(dentry);
77 78 79 80 81 82 83 84
	while (!IS_ROOT(dentry)) {
		struct dentry *parent = dget_parent(dentry);

		if (!(parent->d_flags & DCACHE_DISCONNECTED)) {
			dput(parent);
			break;
		}

85 86 87 88 89 90
		dput(dentry);
		dentry = parent;
	}
	return dentry;
}

91 92
/*
 * Make sure target_dir is fully connected to the dentry tree.
Linus Torvalds's avatar
Linus Torvalds committed
93
 *
94
 * It may already be, as the flag isn't always updated when connection happens.
Linus Torvalds's avatar
Linus Torvalds committed
95
 */
96
static int
97
reconnect_path(struct vfsmount *mnt, struct dentry *target_dir, char *nbuf)
Linus Torvalds's avatar
Linus Torvalds committed
98
{
99 100
	int noprogress = 0;
	int err = -ESTALE;
Linus Torvalds's avatar
Linus Torvalds committed
101 102

	/*
103
	 * It is possible that a confused file system might not let us complete
Linus Torvalds's avatar
Linus Torvalds committed
104 105 106 107 108 109 110
	 * the path to the root.  For example, if get_parent returns a directory
	 * in which we cannot find a name for the child.  While this implies a
	 * very sick filesystem we don't want it to cause knfsd to spin.  Hence
	 * the noprogress counter.  If we go through the loop 10 times (2 is
	 * probably enough) without getting anywhere, we just give up
	 */
	while (target_dir->d_flags & DCACHE_DISCONNECTED && noprogress++ < 10) {
111
		struct dentry *pd = find_disconnected_root(target_dir);
Linus Torvalds's avatar
Linus Torvalds committed
112 113 114 115 116 117 118

		if (!IS_ROOT(pd)) {
			/* must have found a connected parent - great */
			spin_lock(&pd->d_lock);
			pd->d_flags &= ~DCACHE_DISCONNECTED;
			spin_unlock(&pd->d_lock);
			noprogress = 0;
119
		} else if (pd == mnt->mnt_sb->s_root) {
Linus Torvalds's avatar
Linus Torvalds committed
120 121 122 123 124 125
			printk(KERN_ERR "export: Eeek filesystem root is not connected, impossible\n");
			spin_lock(&pd->d_lock);
			pd->d_flags &= ~DCACHE_DISCONNECTED;
			spin_unlock(&pd->d_lock);
			noprogress = 0;
		} else {
126 127 128 129 130 131 132 133 134 135 136 137 138 139
			/*
			 * We have hit the top of a disconnected path, try to
			 * find parent and connect.
			 *
			 * Racing with some other process renaming a directory
			 * isn't much of a problem here.  If someone renames
			 * the directory, it will end up properly connected,
			 * which is what we want
			 *
			 * Getting the parent can't be supported generically,
			 * the locking is too icky.
			 *
			 * Instead we just return EACCES.  If server reboots
			 * or inodes get flushed, you lose
Linus Torvalds's avatar
Linus Torvalds committed
140
			 */
141
			struct dentry *ppd = ERR_PTR(-EACCES);
Linus Torvalds's avatar
Linus Torvalds committed
142 143
			struct dentry *npd;

144
			mutex_lock(&pd->d_inode->i_mutex);
145 146
			if (mnt->mnt_sb->s_export_op->get_parent)
				ppd = mnt->mnt_sb->s_export_op->get_parent(pd);
147
			mutex_unlock(&pd->d_inode->i_mutex);
Linus Torvalds's avatar
Linus Torvalds committed
148 149 150

			if (IS_ERR(ppd)) {
				err = PTR_ERR(ppd);
151
				dprintk("%s: get_parent of %ld failed, err %d\n",
152
					__func__, pd->d_inode->i_ino, err);
Linus Torvalds's avatar
Linus Torvalds committed
153 154 155
				dput(pd);
				break;
			}
156

157
			dprintk("%s: find name of %lu in %lu\n", __func__,
158
				pd->d_inode->i_ino, ppd->d_inode->i_ino);
159
			err = exportfs_get_name(mnt, ppd, nbuf, pd);
Linus Torvalds's avatar
Linus Torvalds committed
160 161 162 163 164 165 166 167 168 169
			if (err) {
				dput(ppd);
				dput(pd);
				if (err == -ENOENT)
					/* some race between get_parent and
					 * get_name?  just try again
					 */
					continue;
				break;
			}
170
			dprintk("%s: found name: %s\n", __func__, nbuf);
171
			mutex_lock(&ppd->d_inode->i_mutex);
Linus Torvalds's avatar
Linus Torvalds committed
172
			npd = lookup_one_len(nbuf, ppd, strlen(nbuf));
173
			mutex_unlock(&ppd->d_inode->i_mutex);
Linus Torvalds's avatar
Linus Torvalds committed
174 175
			if (IS_ERR(npd)) {
				err = PTR_ERR(npd);
176
				dprintk("%s: lookup failed: %d\n",
177
					__func__, err);
Linus Torvalds's avatar
Linus Torvalds committed
178 179 180 181 182 183 184 185 186 187 188 189
				dput(ppd);
				dput(pd);
				break;
			}
			/* we didn't really want npd, we really wanted
			 * a side-effect of the lookup.
			 * hopefully, npd == pd, though it isn't really
			 * a problem if it isn't
			 */
			if (npd == pd)
				noprogress = 0;
			else
190
				printk("%s: npd != pd\n", __func__);
Linus Torvalds's avatar
Linus Torvalds committed
191 192 193 194 195 196 197 198 199 200 201 202 203 204 205
			dput(npd);
			dput(ppd);
			if (IS_ROOT(pd)) {
				/* something went wrong, we have to give up */
				dput(pd);
				break;
			}
		}
		dput(pd);
	}

	if (target_dir->d_flags & DCACHE_DISCONNECTED) {
		/* something went wrong - oh-well */
		if (!err)
			err = -ESTALE;
206
		return err;
Linus Torvalds's avatar
Linus Torvalds committed
207
	}
208 209 210 211

	return 0;
}

Linus Torvalds's avatar
Linus Torvalds committed
212 213 214 215 216 217 218 219 220 221 222 223 224
struct getdents_callback {
	char *name;		/* name that was found. It already points to a
				   buffer NAME_MAX+1 is size */
	unsigned long ino;	/* the inum we are looking for */
	int found;		/* inode matched? */
	int sequence;		/* sequence counter */
};

/*
 * A rather strange filldir function to capture
 * the name matching the specified inode number.
 */
static int filldir_one(void * __buf, const char * name, int len,
225
			loff_t pos, u64 ino, unsigned int d_type)
Linus Torvalds's avatar
Linus Torvalds committed
226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248
{
	struct getdents_callback *buf = __buf;
	int result = 0;

	buf->sequence++;
	if (buf->ino == ino) {
		memcpy(buf->name, name, len);
		buf->name[len] = '\0';
		buf->found = 1;
		result = -1;
	}
	return result;
}

/**
 * get_name - default export_operations->get_name function
 * @dentry: the directory in which to find a name
 * @name:   a pointer to a %NAME_MAX+1 char buffer to store the name
 * @child:  the dentry for the child directory.
 *
 * calls readdir on the parent until it finds an entry with
 * the same inode number as the child, and returns that.
 */
249 250
static int get_name(struct vfsmount *mnt, struct dentry *dentry,
		char *name, struct dentry *child)
Linus Torvalds's avatar
Linus Torvalds committed
251
{
252
	const struct cred *cred = current_cred();
Linus Torvalds's avatar
Linus Torvalds committed
253 254 255 256 257 258 259 260 261 262 263 264 265 266
	struct inode *dir = dentry->d_inode;
	int error;
	struct file *file;
	struct getdents_callback buffer;

	error = -ENOTDIR;
	if (!dir || !S_ISDIR(dir->i_mode))
		goto out;
	error = -EINVAL;
	if (!dir->i_fop)
		goto out;
	/*
	 * Open the directory ...
	 */
267
	file = dentry_open(dget(dentry), mntget(mnt), O_RDONLY, cred);
Linus Torvalds's avatar
Linus Torvalds committed
268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283
	error = PTR_ERR(file);
	if (IS_ERR(file))
		goto out;

	error = -EINVAL;
	if (!file->f_op->readdir)
		goto out_close;

	buffer.name = name;
	buffer.ino = child->d_inode->i_ino;
	buffer.found = 0;
	buffer.sequence = 0;
	while (1) {
		int old_seq = buffer.sequence;

		error = vfs_readdir(file, filldir_one, &buffer);
284 285 286 287
		if (buffer.found) {
			error = 0;
			break;
		}
Linus Torvalds's avatar
Linus Torvalds committed
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

		if (error < 0)
			break;

		error = -ENOENT;
		if (old_seq == buffer.sequence)
			break;
	}

out_close:
	fput(file);
out:
	return error;
}

/**
 * export_encode_fh - default export_operations->encode_fh function
 * @dentry:  the dentry to encode
 * @fh:      where to store the file handle fragment
 * @max_len: maximum length to store there
 * @connectable: whether to store parent information
 *
 * This default encode_fh function assumes that the 32 inode number
 * is suitable for locating an inode, and that the generation number
 * can be used to check that it is still valid.  It places them in the
 * filehandle fragment where export_decode_fh expects to find them.
 */
Christoph Hellwig's avatar
Christoph Hellwig committed
315 316
static int export_encode_fh(struct dentry *dentry, struct fid *fid,
		int *max_len, int connectable)
Linus Torvalds's avatar
Linus Torvalds committed
317 318 319
{
	struct inode * inode = dentry->d_inode;
	int len = *max_len;
Christoph Hellwig's avatar
Christoph Hellwig committed
320
	int type = FILEID_INO32_GEN;
Linus Torvalds's avatar
Linus Torvalds committed
321 322 323 324 325
	
	if (len < 2 || (connectable && len < 4))
		return 255;

	len = 2;
Christoph Hellwig's avatar
Christoph Hellwig committed
326 327
	fid->i32.ino = inode->i_ino;
	fid->i32.gen = inode->i_generation;
Linus Torvalds's avatar
Linus Torvalds committed
328 329 330 331 332
	if (connectable && !S_ISDIR(inode->i_mode)) {
		struct inode *parent;

		spin_lock(&dentry->d_lock);
		parent = dentry->d_parent->d_inode;
Christoph Hellwig's avatar
Christoph Hellwig committed
333 334
		fid->i32.parent_ino = parent->i_ino;
		fid->i32.parent_gen = parent->i_generation;
Linus Torvalds's avatar
Linus Torvalds committed
335 336
		spin_unlock(&dentry->d_lock);
		len = 4;
Christoph Hellwig's avatar
Christoph Hellwig committed
337
		type = FILEID_INO32_GEN_PARENT;
Linus Torvalds's avatar
Linus Torvalds committed
338 339 340 341 342
	}
	*max_len = len;
	return type;
}

Christoph Hellwig's avatar
Christoph Hellwig committed
343
int exportfs_encode_fh(struct dentry *dentry, struct fid *fid, int *max_len,
344 345
		int connectable)
{
346
	const struct export_operations *nop = dentry->d_sb->s_export_op;
347
	int error;
348

349
	if (nop->encode_fh)
Christoph Hellwig's avatar
Christoph Hellwig committed
350
		error = nop->encode_fh(dentry, fid->raw, max_len, connectable);
351
	else
Christoph Hellwig's avatar
Christoph Hellwig committed
352
		error = export_encode_fh(dentry, fid, max_len, connectable);
353 354

	return error;
355 356 357
}
EXPORT_SYMBOL_GPL(exportfs_encode_fh);

Christoph Hellwig's avatar
Christoph Hellwig committed
358 359 360
struct dentry *exportfs_decode_fh(struct vfsmount *mnt, struct fid *fid,
		int fh_len, int fileid_type,
		int (*acceptable)(void *, struct dentry *), void *context)
361
{
362
	const struct export_operations *nop = mnt->mnt_sb->s_export_op;
363
	struct dentry *result, *alias;
364
	char nbuf[NAME_MAX+1];
365
	int err;
366

367 368 369 370
	/*
	 * Try to get any dentry for the given file handle from the filesystem.
	 */
	result = nop->fh_to_dentry(mnt->mnt_sb, fid, fh_len, fileid_type);
371 372
	if (!result)
		result = ERR_PTR(-ESTALE);
373 374 375 376 377 378 379 380 381 382 383 384 385
	if (IS_ERR(result))
		return result;

	if (S_ISDIR(result->d_inode->i_mode)) {
		/*
		 * This request is for a directory.
		 *
		 * On the positive side there is only one dentry for each
		 * directory inode.  On the negative side this implies that we
		 * to ensure our dentry is connected all the way up to the
		 * filesystem root.
		 */
		if (result->d_flags & DCACHE_DISCONNECTED) {
386
			err = reconnect_path(mnt, result, nbuf);
387 388 389 390 391 392 393 394 395 396
			if (err)
				goto err_result;
		}

		if (!acceptable(context, result)) {
			err = -EACCES;
			goto err_result;
		}

		return result;
397
	} else {
398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425
		/*
		 * It's not a directory.  Life is a little more complicated.
		 */
		struct dentry *target_dir, *nresult;

		/*
		 * See if either the dentry we just got from the filesystem
		 * or any alias for it is acceptable.  This is always true
		 * if this filesystem is exported without the subtreecheck
		 * option.  If the filesystem is exported with the subtree
		 * check option there's a fair chance we need to look at
		 * the parent directory in the file handle and make sure
		 * it's connected to the filesystem root.
		 */
		alias = find_acceptable_alias(result, acceptable, context);
		if (alias)
			return alias;

		/*
		 * Try to extract a dentry for the parent directory from the
		 * file handle.  If this fails we'll have to give up.
		 */
		err = -ESTALE;
		if (!nop->fh_to_parent)
			goto err_result;

		target_dir = nop->fh_to_parent(mnt->mnt_sb, fid,
				fh_len, fileid_type);
426 427
		if (!target_dir)
			goto err_result;
428 429 430 431 432 433 434 435 436
		err = PTR_ERR(target_dir);
		if (IS_ERR(target_dir))
			goto err_result;

		/*
		 * And as usual we need to make sure the parent directory is
		 * connected to the filesystem root.  The VFS really doesn't
		 * like disconnected directories..
		 */
437
		err = reconnect_path(mnt, target_dir, nbuf);
438 439 440 441 442 443 444 445 446 447
		if (err) {
			dput(target_dir);
			goto err_result;
		}

		/*
		 * Now that we've got both a well-connected parent and a
		 * dentry for the inode we're after, make sure that our
		 * inode is actually connected to the parent.
		 */
448
		err = exportfs_get_name(mnt, target_dir, nbuf, result);
449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479
		if (!err) {
			mutex_lock(&target_dir->d_inode->i_mutex);
			nresult = lookup_one_len(nbuf, target_dir,
						 strlen(nbuf));
			mutex_unlock(&target_dir->d_inode->i_mutex);
			if (!IS_ERR(nresult)) {
				if (nresult->d_inode) {
					dput(result);
					result = nresult;
				} else
					dput(nresult);
			}
		}

		/*
		 * At this point we are done with the parent, but it's pinned
		 * by the child dentry anyway.
		 */
		dput(target_dir);

		/*
		 * And finally make sure the dentry is actually acceptable
		 * to NFSD.
		 */
		alias = find_acceptable_alias(result, acceptable, context);
		if (!alias) {
			err = -EACCES;
			goto err_result;
		}

		return alias;
480 481
	}

482 483 484
 err_result:
	dput(result);
	return ERR_PTR(err);
485 486 487
}
EXPORT_SYMBOL_GPL(exportfs_decode_fh);

Linus Torvalds's avatar
Linus Torvalds committed
488
MODULE_LICENSE("GPL");