dir.c 30.8 KB
Newer Older
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
/*
  FUSE: Filesystem in Userspace
  Copyright (C) 2001-2005  Miklos Szeredi <miklos@szeredi.hu>

  This program can be distributed under the terms of the GNU GPL.
  See the file COPYING.
*/

#include "fuse_i.h"

#include <linux/pagemap.h>
#include <linux/file.h>
#include <linux/gfp.h>
#include <linux/sched.h>
#include <linux/namei.h>

17 18 19 20 21 22 23 24 25
/*
 * FUSE caches dentries and attributes with separate timeout.  The
 * time in jiffies until the dentry/attributes are valid is stored in
 * dentry->d_time and fuse_inode->i_time respectively.
 */

/*
 * Calculate the time in jiffies until a dentry/attributes are valid
 */
26 27 28 29 30 31 32
static inline unsigned long time_to_jiffies(unsigned long sec,
					    unsigned long nsec)
{
	struct timespec ts = {sec, nsec};
	return jiffies + timespec_to_jiffies(&ts);
}

33 34 35 36
/*
 * Set dentry and possibly attribute timeouts from the lookup/mk*
 * replies
 */
37 38 39
static void fuse_change_timeout(struct dentry *entry, struct fuse_entry_out *o)
{
	entry->d_time = time_to_jiffies(o->entry_valid, o->entry_valid_nsec);
40 41 42 43 44
	if (entry->d_inode)
		get_fuse_inode(entry->d_inode)->i_time =
			time_to_jiffies(o->attr_valid, o->attr_valid_nsec);
}

45 46 47 48
/*
 * Mark the attributes as stale, so that at the next call to
 * ->getattr() they will be fetched from userspace
 */
49 50 51 52 53
void fuse_invalidate_attr(struct inode *inode)
{
	get_fuse_inode(inode)->i_time = jiffies - 1;
}

54 55 56 57 58 59 60 61
/*
 * Just mark the entry as stale, so that a next attempt to look it up
 * will result in a new lookup call to userspace
 *
 * This is called when a dentry is about to become negative and the
 * timeout is unknown (unlink, rmdir, rename and in some cases
 * lookup)
 */
62 63 64 65 66
static void fuse_invalidate_entry_cache(struct dentry *entry)
{
	entry->d_time = jiffies - 1;
}

67 68 69 70
/*
 * Same as fuse_invalidate_entry_cache(), but also try to remove the
 * dentry from the hash
 */
71 72 73 74
static void fuse_invalidate_entry(struct dentry *entry)
{
	d_invalidate(entry);
	fuse_invalidate_entry_cache(entry);
75 76
}

77 78 79 80 81 82 83 84 85 86 87 88 89 90 91
static void fuse_lookup_init(struct fuse_req *req, struct inode *dir,
			     struct dentry *entry,
			     struct fuse_entry_out *outarg)
{
	req->in.h.opcode = FUSE_LOOKUP;
	req->in.h.nodeid = get_node_id(dir);
	req->inode = dir;
	req->in.numargs = 1;
	req->in.args[0].size = entry->d_name.len + 1;
	req->in.args[0].value = entry->d_name.name;
	req->out.numargs = 1;
	req->out.args[0].size = sizeof(struct fuse_entry_out);
	req->out.args[0].value = outarg;
}

92 93 94 95 96 97 98 99 100
/*
 * Check whether the dentry is still valid
 *
 * If the entry validity timeout has expired and the dentry is
 * positive, try to redo the lookup.  If the lookup results in a
 * different inode, then let the VFS invalidate the dentry and redo
 * the lookup once more.  If the lookup results in the same inode,
 * then refresh the attributes, timeouts and mark the dentry valid.
 */
101 102
static int fuse_dentry_revalidate(struct dentry *entry, struct nameidata *nd)
{
103 104 105
	struct inode *inode = entry->d_inode;

	if (inode && is_bad_inode(inode))
106 107 108 109
		return 0;
	else if (time_after(jiffies, entry->d_time)) {
		int err;
		struct fuse_entry_out outarg;
110 111 112
		struct fuse_conn *fc;
		struct fuse_req *req;

113
		/* Doesn't hurt to "reset" the validity timeout */
114 115 116 117 118 119
		fuse_invalidate_entry_cache(entry);
		if (!inode)
			return 0;

		fc = get_fuse_conn(inode);
		req = fuse_get_request(fc);
120 121 122 123
		if (!req)
			return 0;

		fuse_lookup_init(req, entry->d_parent->d_inode, entry, &outarg);
124
		request_send(fc, req);
125
		err = req->out.h.error;
126
		if (!err) {
127
			struct fuse_inode *fi = get_fuse_inode(inode);
128 129 130 131 132 133
			if (outarg.nodeid != get_node_id(inode)) {
				fuse_send_forget(fc, req, outarg.nodeid, 1);
				return 0;
			}
			fi->nlookup ++;
		}
134
		fuse_put_request(fc, req);
135
		if (err || (outarg.attr.mode ^ inode->i_mode) & S_IFMT)
136 137 138
			return 0;

		fuse_change_attributes(inode, &outarg.attr);
139
		fuse_change_timeout(entry, &outarg);
140 141 142 143
	}
	return 1;
}

144 145 146 147
/*
 * Check if there's already a hashed alias of this directory inode.
 * If yes, then lookup and mkdir must not create a new alias.
 */
148 149 150 151 152 153 154 155 156 157 158 159
static int dir_alias(struct inode *inode)
{
	if (S_ISDIR(inode->i_mode)) {
		struct dentry *alias = d_find_alias(inode);
		if (alias) {
			dput(alias);
			return 1;
		}
	}
	return 0;
}

160 161 162 163 164
static inline int invalid_nodeid(u64 nodeid)
{
	return !nodeid || nodeid == FUSE_ROOT_ID;
}

165 166 167 168
static struct dentry_operations fuse_dentry_operations = {
	.d_revalidate	= fuse_dentry_revalidate,
};

169 170
static struct dentry *fuse_lookup(struct inode *dir, struct dentry *entry,
				  struct nameidata *nd)
171 172 173 174 175 176 177 178
{
	int err;
	struct fuse_entry_out outarg;
	struct inode *inode = NULL;
	struct fuse_conn *fc = get_fuse_conn(dir);
	struct fuse_req *req;

	if (entry->d_name.len > FUSE_NAME_MAX)
179
		return ERR_PTR(-ENAMETOOLONG);
180 181 182

	req = fuse_get_request(fc);
	if (!req)
183
		return ERR_PTR(-EINTR);
184 185 186 187

	fuse_lookup_init(req, dir, entry, &outarg);
	request_send(fc, req);
	err = req->out.h.error;
188
	if (!err && outarg.nodeid && invalid_nodeid(outarg.nodeid))
189
		err = -EIO;
190
	if (!err && outarg.nodeid) {
191
		inode = fuse_iget(dir->i_sb, outarg.nodeid, outarg.generation,
192
				  &outarg.attr);
193
		if (!inode) {
194
			fuse_send_forget(fc, req, outarg.nodeid, 1);
195
			return ERR_PTR(-ENOMEM);
196 197 198 199
		}
	}
	fuse_put_request(fc, req);
	if (err && err != -ENOENT)
200
		return ERR_PTR(err);
201

202 203 204
	if (inode && dir_alias(inode)) {
		iput(inode);
		return ERR_PTR(-EIO);
205
	}
206
	d_add(entry, inode);
207
	entry->d_op = &fuse_dentry_operations;
208
	if (!err)
209
		fuse_change_timeout(entry, &outarg);
210 211
	else
		fuse_invalidate_entry_cache(entry);
212
	return NULL;
213 214
}

215 216 217 218 219 220
/*
 * Atomic create+open operation
 *
 * If the filesystem doesn't support this, then fall back to separate
 * 'mknod' + 'open' requests.
 */
M
Miklos Szeredi 已提交
221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277
static int fuse_create_open(struct inode *dir, struct dentry *entry, int mode,
			    struct nameidata *nd)
{
	int err;
	struct inode *inode;
	struct fuse_conn *fc = get_fuse_conn(dir);
	struct fuse_req *req;
	struct fuse_open_in inarg;
	struct fuse_open_out outopen;
	struct fuse_entry_out outentry;
	struct fuse_file *ff;
	struct file *file;
	int flags = nd->intent.open.flags - 1;

	err = -ENOSYS;
	if (fc->no_create)
		goto out;

	err = -ENAMETOOLONG;
	if (entry->d_name.len > FUSE_NAME_MAX)
		goto out;

	err = -EINTR;
	req = fuse_get_request(fc);
	if (!req)
		goto out;

	ff = fuse_file_alloc();
	if (!ff)
		goto out_put_request;

	flags &= ~O_NOCTTY;
	memset(&inarg, 0, sizeof(inarg));
	inarg.flags = flags;
	inarg.mode = mode;
	req->in.h.opcode = FUSE_CREATE;
	req->in.h.nodeid = get_node_id(dir);
	req->inode = dir;
	req->in.numargs = 2;
	req->in.args[0].size = sizeof(inarg);
	req->in.args[0].value = &inarg;
	req->in.args[1].size = entry->d_name.len + 1;
	req->in.args[1].value = entry->d_name.name;
	req->out.numargs = 2;
	req->out.args[0].size = sizeof(outentry);
	req->out.args[0].value = &outentry;
	req->out.args[1].size = sizeof(outopen);
	req->out.args[1].value = &outopen;
	request_send(fc, req);
	err = req->out.h.error;
	if (err) {
		if (err == -ENOSYS)
			fc->no_create = 1;
		goto out_free_ff;
	}

	err = -EIO;
278
	if (!S_ISREG(outentry.attr.mode) || invalid_nodeid(outentry.nodeid))
M
Miklos Szeredi 已提交
279 280 281 282 283 284 285 286
		goto out_free_ff;

	inode = fuse_iget(dir->i_sb, outentry.nodeid, outentry.generation,
			  &outentry.attr);
	err = -ENOMEM;
	if (!inode) {
		flags &= ~(O_CREAT | O_EXCL | O_TRUNC);
		ff->fh = outopen.fh;
287 288 289
		/* Special release, with inode = NULL, this will
		   trigger a 'forget' request when the release is
		   complete */
M
Miklos Szeredi 已提交
290 291 292 293 294
		fuse_send_release(fc, ff, outentry.nodeid, NULL, flags, 0);
		goto out_put_request;
	}
	fuse_put_request(fc, req);
	d_instantiate(entry, inode);
295
	fuse_change_timeout(entry, &outentry);
M
Miklos Szeredi 已提交
296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312
	file = lookup_instantiate_filp(nd, entry, generic_file_open);
	if (IS_ERR(file)) {
		ff->fh = outopen.fh;
		fuse_send_release(fc, ff, outentry.nodeid, inode, flags, 0);
		return PTR_ERR(file);
	}
	fuse_finish_open(inode, file, ff, &outopen);
	return 0;

 out_free_ff:
	fuse_file_free(ff);
 out_put_request:
	fuse_put_request(fc, req);
 out:
	return err;
}

313 314 315
/*
 * Code shared between mknod, mkdir, symlink and link
 */
316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334
static int create_new_entry(struct fuse_conn *fc, struct fuse_req *req,
			    struct inode *dir, struct dentry *entry,
			    int mode)
{
	struct fuse_entry_out outarg;
	struct inode *inode;
	int err;

	req->in.h.nodeid = get_node_id(dir);
	req->inode = dir;
	req->out.numargs = 1;
	req->out.args[0].size = sizeof(outarg);
	req->out.args[0].value = &outarg;
	request_send(fc, req);
	err = req->out.h.error;
	if (err) {
		fuse_put_request(fc, req);
		return err;
	}
335
	if (invalid_nodeid(outarg.nodeid)) {
336 337 338
		fuse_put_request(fc, req);
		return -EIO;
	}
339 340 341 342 343 344 345 346 347
	inode = fuse_iget(dir->i_sb, outarg.nodeid, outarg.generation,
			  &outarg.attr);
	if (!inode) {
		fuse_send_forget(fc, req, outarg.nodeid, 1);
		return -ENOMEM;
	}
	fuse_put_request(fc, req);

	/* Don't allow userspace to do really stupid things... */
348
	if (((inode->i_mode ^ mode) & S_IFMT) || dir_alias(inode)) {
349 350 351 352 353
		iput(inode);
		return -EIO;
	}

	d_instantiate(entry, inode);
354
	fuse_change_timeout(entry, &outarg);
355 356 357 358 359 360 361 362 363 364 365
	fuse_invalidate_attr(dir);
	return 0;
}

static int fuse_mknod(struct inode *dir, struct dentry *entry, int mode,
		      dev_t rdev)
{
	struct fuse_mknod_in inarg;
	struct fuse_conn *fc = get_fuse_conn(dir);
	struct fuse_req *req = fuse_get_request(fc);
	if (!req)
366
		return -EINTR;
367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382

	memset(&inarg, 0, sizeof(inarg));
	inarg.mode = mode;
	inarg.rdev = new_encode_dev(rdev);
	req->in.h.opcode = FUSE_MKNOD;
	req->in.numargs = 2;
	req->in.args[0].size = sizeof(inarg);
	req->in.args[0].value = &inarg;
	req->in.args[1].size = entry->d_name.len + 1;
	req->in.args[1].value = entry->d_name.name;
	return create_new_entry(fc, req, dir, entry, mode);
}

static int fuse_create(struct inode *dir, struct dentry *entry, int mode,
		       struct nameidata *nd)
{
M
Miklos Szeredi 已提交
383 384 385 386 387 388
	if (nd && (nd->flags & LOOKUP_CREATE)) {
		int err = fuse_create_open(dir, entry, mode, nd);
		if (err != -ENOSYS)
			return err;
		/* Fall back on mknod */
	}
389 390 391 392 393 394 395 396 397
	return fuse_mknod(dir, entry, mode, 0);
}

static int fuse_mkdir(struct inode *dir, struct dentry *entry, int mode)
{
	struct fuse_mkdir_in inarg;
	struct fuse_conn *fc = get_fuse_conn(dir);
	struct fuse_req *req = fuse_get_request(fc);
	if (!req)
398
		return -EINTR;
399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422

	memset(&inarg, 0, sizeof(inarg));
	inarg.mode = mode;
	req->in.h.opcode = FUSE_MKDIR;
	req->in.numargs = 2;
	req->in.args[0].size = sizeof(inarg);
	req->in.args[0].value = &inarg;
	req->in.args[1].size = entry->d_name.len + 1;
	req->in.args[1].value = entry->d_name.name;
	return create_new_entry(fc, req, dir, entry, S_IFDIR);
}

static int fuse_symlink(struct inode *dir, struct dentry *entry,
			const char *link)
{
	struct fuse_conn *fc = get_fuse_conn(dir);
	unsigned len = strlen(link) + 1;
	struct fuse_req *req;

	if (len > FUSE_SYMLINK_MAX)
		return -ENAMETOOLONG;

	req = fuse_get_request(fc);
	if (!req)
423
		return -EINTR;
424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439

	req->in.h.opcode = FUSE_SYMLINK;
	req->in.numargs = 2;
	req->in.args[0].size = entry->d_name.len + 1;
	req->in.args[0].value = entry->d_name.name;
	req->in.args[1].size = len;
	req->in.args[1].value = link;
	return create_new_entry(fc, req, dir, entry, S_IFLNK);
}

static int fuse_unlink(struct inode *dir, struct dentry *entry)
{
	int err;
	struct fuse_conn *fc = get_fuse_conn(dir);
	struct fuse_req *req = fuse_get_request(fc);
	if (!req)
440
		return -EINTR;
441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459

	req->in.h.opcode = FUSE_UNLINK;
	req->in.h.nodeid = get_node_id(dir);
	req->inode = dir;
	req->in.numargs = 1;
	req->in.args[0].size = entry->d_name.len + 1;
	req->in.args[0].value = entry->d_name.name;
	request_send(fc, req);
	err = req->out.h.error;
	fuse_put_request(fc, req);
	if (!err) {
		struct inode *inode = entry->d_inode;

		/* Set nlink to zero so the inode can be cleared, if
                   the inode does have more links this will be
                   discovered at the next lookup/getattr */
		inode->i_nlink = 0;
		fuse_invalidate_attr(inode);
		fuse_invalidate_attr(dir);
460
		fuse_invalidate_entry_cache(entry);
461 462 463 464 465 466 467 468 469 470 471
	} else if (err == -EINTR)
		fuse_invalidate_entry(entry);
	return err;
}

static int fuse_rmdir(struct inode *dir, struct dentry *entry)
{
	int err;
	struct fuse_conn *fc = get_fuse_conn(dir);
	struct fuse_req *req = fuse_get_request(fc);
	if (!req)
472
		return -EINTR;
473 474 475 476 477 478 479 480 481 482 483 484 485

	req->in.h.opcode = FUSE_RMDIR;
	req->in.h.nodeid = get_node_id(dir);
	req->inode = dir;
	req->in.numargs = 1;
	req->in.args[0].size = entry->d_name.len + 1;
	req->in.args[0].value = entry->d_name.name;
	request_send(fc, req);
	err = req->out.h.error;
	fuse_put_request(fc, req);
	if (!err) {
		entry->d_inode->i_nlink = 0;
		fuse_invalidate_attr(dir);
486
		fuse_invalidate_entry_cache(entry);
487 488 489 490 491 492 493 494 495 496 497 498 499
	} else if (err == -EINTR)
		fuse_invalidate_entry(entry);
	return err;
}

static int fuse_rename(struct inode *olddir, struct dentry *oldent,
		       struct inode *newdir, struct dentry *newent)
{
	int err;
	struct fuse_rename_in inarg;
	struct fuse_conn *fc = get_fuse_conn(olddir);
	struct fuse_req *req = fuse_get_request(fc);
	if (!req)
500
		return -EINTR;
501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521

	memset(&inarg, 0, sizeof(inarg));
	inarg.newdir = get_node_id(newdir);
	req->in.h.opcode = FUSE_RENAME;
	req->in.h.nodeid = get_node_id(olddir);
	req->inode = olddir;
	req->inode2 = newdir;
	req->in.numargs = 3;
	req->in.args[0].size = sizeof(inarg);
	req->in.args[0].value = &inarg;
	req->in.args[1].size = oldent->d_name.len + 1;
	req->in.args[1].value = oldent->d_name.name;
	req->in.args[2].size = newent->d_name.len + 1;
	req->in.args[2].value = newent->d_name.name;
	request_send(fc, req);
	err = req->out.h.error;
	fuse_put_request(fc, req);
	if (!err) {
		fuse_invalidate_attr(olddir);
		if (olddir != newdir)
			fuse_invalidate_attr(newdir);
522 523 524 525

		/* newent will end up negative */
		if (newent->d_inode)
			fuse_invalidate_entry_cache(newent);
526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548
	} else if (err == -EINTR) {
		/* If request was interrupted, DEITY only knows if the
		   rename actually took place.  If the invalidation
		   fails (e.g. some process has CWD under the renamed
		   directory), then there can be inconsistency between
		   the dcache and the real filesystem.  Tough luck. */
		fuse_invalidate_entry(oldent);
		if (newent->d_inode)
			fuse_invalidate_entry(newent);
	}

	return err;
}

static int fuse_link(struct dentry *entry, struct inode *newdir,
		     struct dentry *newent)
{
	int err;
	struct fuse_link_in inarg;
	struct inode *inode = entry->d_inode;
	struct fuse_conn *fc = get_fuse_conn(inode);
	struct fuse_req *req = fuse_get_request(fc);
	if (!req)
549
		return -EINTR;
550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571

	memset(&inarg, 0, sizeof(inarg));
	inarg.oldnodeid = get_node_id(inode);
	req->in.h.opcode = FUSE_LINK;
	req->inode2 = inode;
	req->in.numargs = 2;
	req->in.args[0].size = sizeof(inarg);
	req->in.args[0].value = &inarg;
	req->in.args[1].size = newent->d_name.len + 1;
	req->in.args[1].value = newent->d_name.name;
	err = create_new_entry(fc, req, newdir, newent, inode->i_mode);
	/* Contrary to "normal" filesystems it can happen that link
	   makes two "logical" inodes point to the same "physical"
	   inode.  We invalidate the attributes of the old one, so it
	   will reflect changes in the backing inode (link count,
	   etc.)
	*/
	if (!err || err == -EINTR)
		fuse_invalidate_attr(inode);
	return err;
}

572 573 574 575 576 577 578
int fuse_do_getattr(struct inode *inode)
{
	int err;
	struct fuse_attr_out arg;
	struct fuse_conn *fc = get_fuse_conn(inode);
	struct fuse_req *req = fuse_get_request(fc);
	if (!req)
579
		return -EINTR;
580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603

	req->in.h.opcode = FUSE_GETATTR;
	req->in.h.nodeid = get_node_id(inode);
	req->inode = inode;
	req->out.numargs = 1;
	req->out.args[0].size = sizeof(arg);
	req->out.args[0].value = &arg;
	request_send(fc, req);
	err = req->out.h.error;
	fuse_put_request(fc, req);
	if (!err) {
		if ((inode->i_mode ^ arg.attr.mode) & S_IFMT) {
			make_bad_inode(inode);
			err = -EIO;
		} else {
			struct fuse_inode *fi = get_fuse_inode(inode);
			fuse_change_attributes(inode, &arg.attr);
			fi->i_time = time_to_jiffies(arg.attr_valid,
						     arg.attr_valid_nsec);
		}
	}
	return err;
}

604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632
/*
 * Calling into a user-controlled filesystem gives the filesystem
 * daemon ptrace-like capabilities over the requester process.  This
 * means, that the filesystem daemon is able to record the exact
 * filesystem operations performed, and can also control the behavior
 * of the requester process in otherwise impossible ways.  For example
 * it can delay the operation for arbitrary length of time allowing
 * DoS against the requester.
 *
 * For this reason only those processes can call into the filesystem,
 * for which the owner of the mount has ptrace privilege.  This
 * excludes processes started by other users, suid or sgid processes.
 */
static int fuse_allow_task(struct fuse_conn *fc, struct task_struct *task)
{
	if (fc->flags & FUSE_ALLOW_OTHER)
		return 1;

	if (task->euid == fc->user_id &&
	    task->suid == fc->user_id &&
	    task->uid == fc->user_id &&
	    task->egid == fc->group_id &&
	    task->sgid == fc->group_id &&
	    task->gid == fc->group_id)
		return 1;

	return 0;
}

633 634 635 636 637 638 639 640 641
/*
 * Check whether the inode attributes are still valid
 *
 * If the attribute validity timeout has expired, then fetch the fresh
 * attributes with a 'getattr' request
 *
 * I'm not sure why cached attributes are never returned for the root
 * inode, this is probably being too cautious.
 */
642 643 644 645 646 647
static int fuse_revalidate(struct dentry *entry)
{
	struct inode *inode = entry->d_inode;
	struct fuse_inode *fi = get_fuse_inode(inode);
	struct fuse_conn *fc = get_fuse_conn(inode);

648 649 650 651
	if (!fuse_allow_task(fc, current))
		return -EACCES;
	if (get_node_id(inode) != FUSE_ROOT_ID &&
	    time_before_eq(jiffies, fi->i_time))
652 653 654 655 656
		return 0;

	return fuse_do_getattr(inode);
}

M
Miklos Szeredi 已提交
657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688
static int fuse_access(struct inode *inode, int mask)
{
	struct fuse_conn *fc = get_fuse_conn(inode);
	struct fuse_req *req;
	struct fuse_access_in inarg;
	int err;

	if (fc->no_access)
		return 0;

	req = fuse_get_request(fc);
	if (!req)
		return -EINTR;

	memset(&inarg, 0, sizeof(inarg));
	inarg.mask = mask;
	req->in.h.opcode = FUSE_ACCESS;
	req->in.h.nodeid = get_node_id(inode);
	req->inode = inode;
	req->in.numargs = 1;
	req->in.args[0].size = sizeof(inarg);
	req->in.args[0].value = &inarg;
	request_send(fc, req);
	err = req->out.h.error;
	fuse_put_request(fc, req);
	if (err == -ENOSYS) {
		fc->no_access = 1;
		err = 0;
	}
	return err;
}

689 690 691 692 693 694 695 696 697 698 699 700 701
/*
 * Check permission.  The two basic access models of FUSE are:
 *
 * 1) Local access checking ('default_permissions' mount option) based
 * on file mode.  This is the plain old disk filesystem permission
 * modell.
 *
 * 2) "Remote" access checking, where server is responsible for
 * checking permission in each inode operation.  An exception to this
 * is if ->permission() was invoked from sys_access() in which case an
 * access request is sent.  Execute permission is still checked
 * locally based on file mode.
 */
702 703 704 705
static int fuse_permission(struct inode *inode, int mask, struct nameidata *nd)
{
	struct fuse_conn *fc = get_fuse_conn(inode);

706
	if (!fuse_allow_task(fc, current))
707
		return -EACCES;
M
Miklos Szeredi 已提交
708 709 710 711 712 713 714 715 716 717 718 719
	else if (fc->flags & FUSE_DEFAULT_PERMISSIONS) {
		int err = generic_permission(inode, mask, NULL);

		/* If permission is denied, try to refresh file
		   attributes.  This is also needed, because the root
		   node will at first have no permissions */
		if (err == -EACCES) {
		 	err = fuse_do_getattr(inode);
			if (!err)
				err = generic_permission(inode, mask, NULL);
		}

720 721 722 723
		/* Note: the opposite of the above test does not
		   exist.  So if permissions are revoked this won't be
		   noticed immediately, only after the attribute
		   timeout has expired */
M
Miklos Szeredi 已提交
724 725 726

		return err;
	} else {
727 728 729
		int mode = inode->i_mode;
		if ((mask & MAY_EXEC) && !S_ISDIR(mode) && !(mode & S_IXUGO))
			return -EACCES;
M
Miklos Szeredi 已提交
730 731 732

		if (nd && (nd->flags & LOOKUP_ACCESS))
			return fuse_access(inode, mask);
733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761
		return 0;
	}
}

static int parse_dirfile(char *buf, size_t nbytes, struct file *file,
			 void *dstbuf, filldir_t filldir)
{
	while (nbytes >= FUSE_NAME_OFFSET) {
		struct fuse_dirent *dirent = (struct fuse_dirent *) buf;
		size_t reclen = FUSE_DIRENT_SIZE(dirent);
		int over;
		if (!dirent->namelen || dirent->namelen > FUSE_NAME_MAX)
			return -EIO;
		if (reclen > nbytes)
			break;

		over = filldir(dstbuf, dirent->name, dirent->namelen,
			       file->f_pos, dirent->ino, dirent->type);
		if (over)
			break;

		buf += reclen;
		nbytes -= reclen;
		file->f_pos = dirent->off;
	}

	return 0;
}

762 763 764
static inline size_t fuse_send_readdir(struct fuse_req *req, struct file *file,
				       struct inode *inode, loff_t pos,
				       size_t count)
765
{
766
	return fuse_send_read_common(req, file, inode, pos, count, 1);
767 768
}

769
static int fuse_readdir(struct file *file, void *dstbuf, filldir_t filldir)
770
{
771 772 773
	int err;
	size_t nbytes;
	struct page *page;
774 775
	struct inode *inode = file->f_dentry->d_inode;
	struct fuse_conn *fc = get_fuse_conn(inode);
776
	struct fuse_req *req = fuse_get_request(fc);
777
	if (!req)
778
		return -EINTR;
779

780 781 782 783 784 785 786 787
	page = alloc_page(GFP_KERNEL);
	if (!page) {
		fuse_put_request(fc, req);
		return -ENOMEM;
	}
	req->num_pages = 1;
	req->pages[0] = page;
	nbytes = fuse_send_readdir(req, file, inode, file->f_pos, PAGE_SIZE);
788 789 790
	err = req->out.h.error;
	fuse_put_request(fc, req);
	if (!err)
791 792
		err = parse_dirfile(page_address(page), nbytes, file, dstbuf,
				    filldir);
793

794
	__free_page(page);
795
	fuse_invalidate_attr(inode); /* atime changed */
796
	return err;
797 798 799 800 801 802 803 804 805 806
}

static char *read_link(struct dentry *dentry)
{
	struct inode *inode = dentry->d_inode;
	struct fuse_conn *fc = get_fuse_conn(inode);
	struct fuse_req *req = fuse_get_request(fc);
	char *link;

	if (!req)
807
		return ERR_PTR(-EINTR);
808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828

	link = (char *) __get_free_page(GFP_KERNEL);
	if (!link) {
		link = ERR_PTR(-ENOMEM);
		goto out;
	}
	req->in.h.opcode = FUSE_READLINK;
	req->in.h.nodeid = get_node_id(inode);
	req->inode = inode;
	req->out.argvar = 1;
	req->out.numargs = 1;
	req->out.args[0].size = PAGE_SIZE - 1;
	req->out.args[0].value = link;
	request_send(fc, req);
	if (req->out.h.error) {
		free_page((unsigned long) link);
		link = ERR_PTR(req->out.h.error);
	} else
		link[req->out.args[0].size] = '\0';
 out:
	fuse_put_request(fc, req);
829
	fuse_invalidate_attr(inode); /* atime changed */
830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851
	return link;
}

static void free_link(char *link)
{
	if (!IS_ERR(link))
		free_page((unsigned long) link);
}

static void *fuse_follow_link(struct dentry *dentry, struct nameidata *nd)
{
	nd_set_link(nd, read_link(dentry));
	return NULL;
}

static void fuse_put_link(struct dentry *dentry, struct nameidata *nd, void *c)
{
	free_link(nd_get_link(nd));
}

static int fuse_dir_open(struct inode *inode, struct file *file)
{
852
	return fuse_open_common(inode, file, 1);
853 854 855 856
}

static int fuse_dir_release(struct inode *inode, struct file *file)
{
857
	return fuse_release_common(inode, file, 1);
858 859
}

860 861 862 863 864 865
static int fuse_dir_fsync(struct file *file, struct dentry *de, int datasync)
{
	/* nfsd can call this with no file */
	return file ? fuse_fsync_common(file, de, datasync, 1) : 0;
}

866
static void iattr_to_fattr(struct iattr *iattr, struct fuse_setattr_in *arg)
867 868 869 870
{
	unsigned ivalid = iattr->ia_valid;

	if (ivalid & ATTR_MODE)
871
		arg->valid |= FATTR_MODE,   arg->mode = iattr->ia_mode;
872
	if (ivalid & ATTR_UID)
873
		arg->valid |= FATTR_UID,    arg->uid = iattr->ia_uid;
874
	if (ivalid & ATTR_GID)
875
		arg->valid |= FATTR_GID,    arg->gid = iattr->ia_gid;
876
	if (ivalid & ATTR_SIZE)
877
		arg->valid |= FATTR_SIZE,   arg->size = iattr->ia_size;
878 879
	/* You can only _set_ these together (they may change by themselves) */
	if ((ivalid & (ATTR_ATIME | ATTR_MTIME)) == (ATTR_ATIME | ATTR_MTIME)) {
880 881 882 883 884 885 886 887
		arg->valid |= FATTR_ATIME | FATTR_MTIME;
		arg->atime = iattr->ia_atime.tv_sec;
		arg->mtime = iattr->ia_mtime.tv_sec;
	}
	if (ivalid & ATTR_FILE) {
		struct fuse_file *ff = iattr->ia_file->private_data;
		arg->valid |= FATTR_FH;
		arg->fh = ff->fh;
888 889 890
	}
}

891 892 893 894 895 896 897 898 899
/*
 * Set attributes, and at the same time refresh them.
 *
 * Truncation is slightly complicated, because the 'truncate' request
 * may fail, in which case we don't want to touch the mapping.
 * vmtruncate() doesn't allow for this case.  So do the rlimit
 * checking by hand and call vmtruncate() only after the file has
 * actually been truncated.
 */
900 901 902 903 904 905 906 907 908 909 910
static int fuse_setattr(struct dentry *entry, struct iattr *attr)
{
	struct inode *inode = entry->d_inode;
	struct fuse_conn *fc = get_fuse_conn(inode);
	struct fuse_inode *fi = get_fuse_inode(inode);
	struct fuse_req *req;
	struct fuse_setattr_in inarg;
	struct fuse_attr_out outarg;
	int err;
	int is_truncate = 0;

M
Miklos Szeredi 已提交
911 912 913 914 915 916
	if (fc->flags & FUSE_DEFAULT_PERMISSIONS) {
		err = inode_change_ok(inode, attr);
		if (err)
			return err;
	}

917 918 919 920 921 922 923 924 925 926 927 928
	if (attr->ia_valid & ATTR_SIZE) {
		unsigned long limit;
		is_truncate = 1;
		limit = current->signal->rlim[RLIMIT_FSIZE].rlim_cur;
		if (limit != RLIM_INFINITY && attr->ia_size > (loff_t) limit) {
			send_sig(SIGXFSZ, current, 0);
			return -EFBIG;
		}
	}

	req = fuse_get_request(fc);
	if (!req)
929
		return -EINTR;
930 931

	memset(&inarg, 0, sizeof(inarg));
932
	iattr_to_fattr(attr, &inarg);
933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965
	req->in.h.opcode = FUSE_SETATTR;
	req->in.h.nodeid = get_node_id(inode);
	req->inode = inode;
	req->in.numargs = 1;
	req->in.args[0].size = sizeof(inarg);
	req->in.args[0].value = &inarg;
	req->out.numargs = 1;
	req->out.args[0].size = sizeof(outarg);
	req->out.args[0].value = &outarg;
	request_send(fc, req);
	err = req->out.h.error;
	fuse_put_request(fc, req);
	if (!err) {
		if ((inode->i_mode ^ outarg.attr.mode) & S_IFMT) {
			make_bad_inode(inode);
			err = -EIO;
		} else {
			if (is_truncate) {
				loff_t origsize = i_size_read(inode);
				i_size_write(inode, outarg.attr.size);
				if (origsize > outarg.attr.size)
					vmtruncate(inode, outarg.attr.size);
			}
			fuse_change_attributes(inode, &outarg.attr);
			fi->i_time = time_to_jiffies(outarg.attr_valid,
						     outarg.attr_valid_nsec);
		}
	} else if (err == -EINTR)
		fuse_invalidate_attr(inode);

	return err;
}

966 967 968 969 970 971 972 973 974 975 976
static int fuse_getattr(struct vfsmount *mnt, struct dentry *entry,
			struct kstat *stat)
{
	struct inode *inode = entry->d_inode;
	int err = fuse_revalidate(entry);
	if (!err)
		generic_fillattr(inode, stat);

	return err;
}

977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993
static int fuse_setxattr(struct dentry *entry, const char *name,
			 const void *value, size_t size, int flags)
{
	struct inode *inode = entry->d_inode;
	struct fuse_conn *fc = get_fuse_conn(inode);
	struct fuse_req *req;
	struct fuse_setxattr_in inarg;
	int err;

	if (size > FUSE_XATTR_SIZE_MAX)
		return -E2BIG;

	if (fc->no_setxattr)
		return -EOPNOTSUPP;

	req = fuse_get_request(fc);
	if (!req)
994
		return -EINTR;
995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033

	memset(&inarg, 0, sizeof(inarg));
	inarg.size = size;
	inarg.flags = flags;
	req->in.h.opcode = FUSE_SETXATTR;
	req->in.h.nodeid = get_node_id(inode);
	req->inode = inode;
	req->in.numargs = 3;
	req->in.args[0].size = sizeof(inarg);
	req->in.args[0].value = &inarg;
	req->in.args[1].size = strlen(name) + 1;
	req->in.args[1].value = name;
	req->in.args[2].size = size;
	req->in.args[2].value = value;
	request_send(fc, req);
	err = req->out.h.error;
	fuse_put_request(fc, req);
	if (err == -ENOSYS) {
		fc->no_setxattr = 1;
		err = -EOPNOTSUPP;
	}
	return err;
}

static ssize_t fuse_getxattr(struct dentry *entry, const char *name,
			     void *value, size_t size)
{
	struct inode *inode = entry->d_inode;
	struct fuse_conn *fc = get_fuse_conn(inode);
	struct fuse_req *req;
	struct fuse_getxattr_in inarg;
	struct fuse_getxattr_out outarg;
	ssize_t ret;

	if (fc->no_getxattr)
		return -EOPNOTSUPP;

	req = fuse_get_request(fc);
	if (!req)
1034
		return -EINTR;
1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083

	memset(&inarg, 0, sizeof(inarg));
	inarg.size = size;
	req->in.h.opcode = FUSE_GETXATTR;
	req->in.h.nodeid = get_node_id(inode);
	req->inode = inode;
	req->in.numargs = 2;
	req->in.args[0].size = sizeof(inarg);
	req->in.args[0].value = &inarg;
	req->in.args[1].size = strlen(name) + 1;
	req->in.args[1].value = name;
	/* This is really two different operations rolled into one */
	req->out.numargs = 1;
	if (size) {
		req->out.argvar = 1;
		req->out.args[0].size = size;
		req->out.args[0].value = value;
	} else {
		req->out.args[0].size = sizeof(outarg);
		req->out.args[0].value = &outarg;
	}
	request_send(fc, req);
	ret = req->out.h.error;
	if (!ret)
		ret = size ? req->out.args[0].size : outarg.size;
	else {
		if (ret == -ENOSYS) {
			fc->no_getxattr = 1;
			ret = -EOPNOTSUPP;
		}
	}
	fuse_put_request(fc, req);
	return ret;
}

static ssize_t fuse_listxattr(struct dentry *entry, char *list, size_t size)
{
	struct inode *inode = entry->d_inode;
	struct fuse_conn *fc = get_fuse_conn(inode);
	struct fuse_req *req;
	struct fuse_getxattr_in inarg;
	struct fuse_getxattr_out outarg;
	ssize_t ret;

	if (fc->no_listxattr)
		return -EOPNOTSUPP;

	req = fuse_get_request(fc);
	if (!req)
1084
		return -EINTR;
1085 1086 1087 1088 1089 1090 1091 1092 1093 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 1122 1123 1124 1125 1126 1127 1128 1129

	memset(&inarg, 0, sizeof(inarg));
	inarg.size = size;
	req->in.h.opcode = FUSE_LISTXATTR;
	req->in.h.nodeid = get_node_id(inode);
	req->inode = inode;
	req->in.numargs = 1;
	req->in.args[0].size = sizeof(inarg);
	req->in.args[0].value = &inarg;
	/* This is really two different operations rolled into one */
	req->out.numargs = 1;
	if (size) {
		req->out.argvar = 1;
		req->out.args[0].size = size;
		req->out.args[0].value = list;
	} else {
		req->out.args[0].size = sizeof(outarg);
		req->out.args[0].value = &outarg;
	}
	request_send(fc, req);
	ret = req->out.h.error;
	if (!ret)
		ret = size ? req->out.args[0].size : outarg.size;
	else {
		if (ret == -ENOSYS) {
			fc->no_listxattr = 1;
			ret = -EOPNOTSUPP;
		}
	}
	fuse_put_request(fc, req);
	return ret;
}

static int fuse_removexattr(struct dentry *entry, const char *name)
{
	struct inode *inode = entry->d_inode;
	struct fuse_conn *fc = get_fuse_conn(inode);
	struct fuse_req *req;
	int err;

	if (fc->no_removexattr)
		return -EOPNOTSUPP;

	req = fuse_get_request(fc);
	if (!req)
1130
		return -EINTR;
1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147

	req->in.h.opcode = FUSE_REMOVEXATTR;
	req->in.h.nodeid = get_node_id(inode);
	req->inode = inode;
	req->in.numargs = 1;
	req->in.args[0].size = strlen(name) + 1;
	req->in.args[0].value = name;
	request_send(fc, req);
	err = req->out.h.error;
	fuse_put_request(fc, req);
	if (err == -ENOSYS) {
		fc->no_removexattr = 1;
		err = -EOPNOTSUPP;
	}
	return err;
}

1148 1149
static struct inode_operations fuse_dir_inode_operations = {
	.lookup		= fuse_lookup,
1150 1151 1152 1153 1154 1155 1156 1157 1158
	.mkdir		= fuse_mkdir,
	.symlink	= fuse_symlink,
	.unlink		= fuse_unlink,
	.rmdir		= fuse_rmdir,
	.rename		= fuse_rename,
	.link		= fuse_link,
	.setattr	= fuse_setattr,
	.create		= fuse_create,
	.mknod		= fuse_mknod,
1159 1160
	.permission	= fuse_permission,
	.getattr	= fuse_getattr,
1161 1162 1163 1164
	.setxattr	= fuse_setxattr,
	.getxattr	= fuse_getxattr,
	.listxattr	= fuse_listxattr,
	.removexattr	= fuse_removexattr,
1165 1166 1167
};

static struct file_operations fuse_dir_operations = {
M
Miklos Szeredi 已提交
1168
	.llseek		= generic_file_llseek,
1169 1170 1171 1172
	.read		= generic_read_dir,
	.readdir	= fuse_readdir,
	.open		= fuse_dir_open,
	.release	= fuse_dir_release,
1173
	.fsync		= fuse_dir_fsync,
1174 1175 1176
};

static struct inode_operations fuse_common_inode_operations = {
1177
	.setattr	= fuse_setattr,
1178 1179
	.permission	= fuse_permission,
	.getattr	= fuse_getattr,
1180 1181 1182 1183
	.setxattr	= fuse_setxattr,
	.getxattr	= fuse_getxattr,
	.listxattr	= fuse_listxattr,
	.removexattr	= fuse_removexattr,
1184 1185 1186
};

static struct inode_operations fuse_symlink_inode_operations = {
1187
	.setattr	= fuse_setattr,
1188 1189 1190 1191
	.follow_link	= fuse_follow_link,
	.put_link	= fuse_put_link,
	.readlink	= generic_readlink,
	.getattr	= fuse_getattr,
1192 1193 1194 1195
	.setxattr	= fuse_setxattr,
	.getxattr	= fuse_getxattr,
	.listxattr	= fuse_listxattr,
	.removexattr	= fuse_removexattr,
1196 1197 1198 1199 1200 1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212
};

void fuse_init_common(struct inode *inode)
{
	inode->i_op = &fuse_common_inode_operations;
}

void fuse_init_dir(struct inode *inode)
{
	inode->i_op = &fuse_dir_inode_operations;
	inode->i_fop = &fuse_dir_operations;
}

void fuse_init_symlink(struct inode *inode)
{
	inode->i_op = &fuse_symlink_inode_operations;
}