xfs_iops.c 19.9 KB
Newer Older
L
Linus Torvalds 已提交
1
/*
2 3
 * Copyright (c) 2000-2005 Silicon Graphics, Inc.
 * All Rights Reserved.
L
Linus Torvalds 已提交
4
 *
5 6
 * This program is free software; you can redistribute it and/or
 * modify it under the terms of the GNU General Public License as
L
Linus Torvalds 已提交
7 8
 * published by the Free Software Foundation.
 *
9 10 11 12
 * This program is distributed in the hope that it would be useful,
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
 * GNU General Public License for more details.
L
Linus Torvalds 已提交
13
 *
14 15 16
 * You should have received a copy of the GNU General Public License
 * along with this program; if not, write the Free Software Foundation,
 * Inc.,  51 Franklin St, Fifth Floor, Boston, MA  02110-1301  USA
L
Linus Torvalds 已提交
17 18 19
 */
#include "xfs.h"
#include "xfs_fs.h"
20
#include "xfs_acl.h"
21
#include "xfs_bit.h"
L
Linus Torvalds 已提交
22
#include "xfs_log.h"
23
#include "xfs_inum.h"
L
Linus Torvalds 已提交
24 25 26 27 28 29 30 31
#include "xfs_trans.h"
#include "xfs_sb.h"
#include "xfs_ag.h"
#include "xfs_dir2.h"
#include "xfs_alloc.h"
#include "xfs_quota.h"
#include "xfs_mount.h"
#include "xfs_bmap_btree.h"
32
#include "xfs_alloc_btree.h"
L
Linus Torvalds 已提交
33 34
#include "xfs_ialloc_btree.h"
#include "xfs_dir2_sf.h"
35
#include "xfs_attr_sf.h"
L
Linus Torvalds 已提交
36 37 38
#include "xfs_dinode.h"
#include "xfs_inode.h"
#include "xfs_bmap.h"
39 40
#include "xfs_btree.h"
#include "xfs_ialloc.h"
L
Linus Torvalds 已提交
41 42 43 44 45 46 47
#include "xfs_rtalloc.h"
#include "xfs_error.h"
#include "xfs_itable.h"
#include "xfs_rw.h"
#include "xfs_attr.h"
#include "xfs_buf_item.h"
#include "xfs_utils.h"
48
#include "xfs_vnodeops.h"
C
Christoph Hellwig 已提交
49
#include "xfs_trace.h"
L
Linus Torvalds 已提交
50

51
#include <linux/capability.h>
L
Linus Torvalds 已提交
52 53
#include <linux/xattr.h>
#include <linux/namei.h>
54
#include <linux/posix_acl.h>
55
#include <linux/security.h>
D
David Chinner 已提交
56
#include <linux/falloc.h>
E
Eric Sandeen 已提交
57
#include <linux/fiemap.h>
58
#include <linux/slab.h>
L
Linus Torvalds 已提交
59

60
/*
61 62 63
 * Bring the timestamps in the XFS inode uptodate.
 *
 * Used before writing the inode to disk.
64 65
 */
void
66
xfs_synchronize_times(
67 68
	xfs_inode_t	*ip)
{
69
	struct inode	*inode = VFS_I(ip);
70

71 72 73 74 75 76
	ip->i_d.di_atime.t_sec = (__int32_t)inode->i_atime.tv_sec;
	ip->i_d.di_atime.t_nsec = (__int32_t)inode->i_atime.tv_nsec;
	ip->i_d.di_ctime.t_sec = (__int32_t)inode->i_ctime.tv_sec;
	ip->i_d.di_ctime.t_nsec = (__int32_t)inode->i_ctime.tv_nsec;
	ip->i_d.di_mtime.t_sec = (__int32_t)inode->i_mtime.tv_sec;
	ip->i_d.di_mtime.t_nsec = (__int32_t)inode->i_mtime.tv_nsec;
77 78
}

79
/*
80
 * If the linux inode is valid, mark it dirty.
81 82 83 84 85 86 87
 * Used when commiting a dirty inode into a transaction so that
 * the inode will get written back by the linux code
 */
void
xfs_mark_inode_dirty_sync(
	xfs_inode_t	*ip)
{
88
	struct inode	*inode = VFS_I(ip);
89

90
	if (!(inode->i_state & (I_WILL_FREE|I_FREEING|I_CLEAR)))
91
		mark_inode_dirty_sync(inode);
92 93
}

94 95 96 97 98 99 100 101 102 103
void
xfs_mark_inode_dirty(
	xfs_inode_t	*ip)
{
	struct inode	*inode = VFS_I(ip);

	if (!(inode->i_state & (I_WILL_FREE|I_FREEING|I_CLEAR)))
		mark_inode_dirty(inode);
}

104 105 106 107 108 109 110 111 112 113
/*
 * Change the requested timestamp in the given inode.
 * We don't lock across timestamp updates, and we don't log them but
 * we do record the fact that there is dirty information in core.
 */
void
xfs_ichgtime(
	xfs_inode_t	*ip,
	int		flags)
{
114
	struct inode	*inode = VFS_I(ip);
115
	timespec_t	tv;
C
Christoph Hellwig 已提交
116
	int		sync_it = 0;
117

C
Christoph Hellwig 已提交
118 119 120 121
	tv = current_fs_time(inode->i_sb);

	if ((flags & XFS_ICHGTIME_MOD) &&
	    !timespec_equal(&inode->i_mtime, &tv)) {
122
		inode->i_mtime = tv;
C
Christoph Hellwig 已提交
123
		sync_it = 1;
124
	}
C
Christoph Hellwig 已提交
125 126
	if ((flags & XFS_ICHGTIME_CHG) &&
	    !timespec_equal(&inode->i_ctime, &tv)) {
127
		inode->i_ctime = tv;
C
Christoph Hellwig 已提交
128
		sync_it = 1;
129 130 131
	}

	/*
132 133
	 * Update complete - now make sure everyone knows that the inode
	 * is dirty.
134
	 */
135
	if (sync_it)
136
		xfs_mark_inode_dirty_sync(ip);
137 138
}

139 140 141 142 143 144 145
/*
 * Hook in SELinux.  This is not quite correct yet, what we really need
 * here (as we do for default ACLs) is a mechanism by which creation of
 * these attrs can be journalled at inode creation time (along with the
 * inode, of course, such that log replay can't cause these to be lost).
 */
STATIC int
146
xfs_init_security(
147
	struct inode	*inode,
148 149
	struct inode	*dir)
{
150
	struct xfs_inode *ip = XFS_I(inode);
151 152
	size_t		length;
	void		*value;
153
	unsigned char	*name;
154 155
	int		error;

156
	error = security_inode_init_security(inode, dir, (char **)&name,
157
					     &value, &length);
158 159 160 161 162 163
	if (error) {
		if (error == -EOPNOTSUPP)
			return 0;
		return -error;
	}

164
	error = xfs_attr_set(ip, name, value, length, ATTR_SECURE);
165 166 167 168 169 170

	kfree(name);
	kfree(value);
	return error;
}

171 172 173 174 175 176 177 178 179
static void
xfs_dentry_to_name(
	struct xfs_name	*namep,
	struct dentry	*dentry)
{
	namep->name = dentry->d_name.name;
	namep->len = dentry->d_name.len;
}

180
STATIC void
181
xfs_cleanup_inode(
182
	struct inode	*dir,
183
	struct inode	*inode,
184
	struct dentry	*dentry)
185
{
186
	struct xfs_name	teardown;
187 188

	/* Oh, the horror.
189
	 * If we can't add the ACL or we fail in
190
	 * xfs_init_security we must back out.
191 192
	 * ENOSPC can hit here, among other things.
	 */
193
	xfs_dentry_to_name(&teardown, dentry);
194

195
	xfs_remove(XFS_I(dir), &teardown, XFS_I(inode));
196
	iput(inode);
197 198
}

L
Linus Torvalds 已提交
199
STATIC int
200
xfs_vn_mknod(
L
Linus Torvalds 已提交
201 202 203 204 205
	struct inode	*dir,
	struct dentry	*dentry,
	int		mode,
	dev_t		rdev)
{
C
Christoph Hellwig 已提交
206
	struct inode	*inode;
207
	struct xfs_inode *ip = NULL;
208
	struct posix_acl *default_acl = NULL;
209
	struct xfs_name	name;
L
Linus Torvalds 已提交
210 211 212 213 214 215
	int		error;

	/*
	 * Irix uses Missed'em'V split, but doesn't want to see
	 * the upper 5 bits of (14bit) major.
	 */
216 217 218 219 220 221 222
	if (S_ISCHR(mode) || S_ISBLK(mode)) {
		if (unlikely(!sysv_valid_dev(rdev) || MAJOR(rdev) & ~0x1ff))
			return -EINVAL;
		rdev = sysv_encode_dev(rdev);
	} else {
		rdev = 0;
	}
L
Linus Torvalds 已提交
223

224 225 226 227
	if (IS_POSIXACL(dir)) {
		default_acl = xfs_get_acl(dir, ACL_TYPE_DEFAULT);
		if (IS_ERR(default_acl))
			return -PTR_ERR(default_acl);
L
Linus Torvalds 已提交
228

229 230
		if (!default_acl)
			mode &= ~current_umask();
231
	}
L
Linus Torvalds 已提交
232

233 234
	xfs_dentry_to_name(&name, dentry);
	error = xfs_create(XFS_I(dir), &name, mode, rdev, &ip, NULL);
C
Christoph Hellwig 已提交
235 236
	if (unlikely(error))
		goto out_free_acl;
237

238
	inode = VFS_I(ip);
239 240

	error = xfs_init_security(inode, dir);
C
Christoph Hellwig 已提交
241 242 243 244
	if (unlikely(error))
		goto out_cleanup_inode;

	if (default_acl) {
245
		error = -xfs_inherit_acl(inode, default_acl);
C
Christoph Hellwig 已提交
246 247
		if (unlikely(error))
			goto out_cleanup_inode;
248
		posix_acl_release(default_acl);
L
Linus Torvalds 已提交
249 250 251
	}


C
Christoph Hellwig 已提交
252 253 254 255
	d_instantiate(dentry, inode);
	return -error;

 out_cleanup_inode:
256
	xfs_cleanup_inode(dir, inode, dentry);
C
Christoph Hellwig 已提交
257
 out_free_acl:
258
	posix_acl_release(default_acl);
L
Linus Torvalds 已提交
259 260 261 262
	return -error;
}

STATIC int
263
xfs_vn_create(
L
Linus Torvalds 已提交
264 265 266 267 268
	struct inode	*dir,
	struct dentry	*dentry,
	int		mode,
	struct nameidata *nd)
{
269
	return xfs_vn_mknod(dir, dentry, mode, 0);
L
Linus Torvalds 已提交
270 271 272
}

STATIC int
273
xfs_vn_mkdir(
L
Linus Torvalds 已提交
274 275 276 277
	struct inode	*dir,
	struct dentry	*dentry,
	int		mode)
{
278
	return xfs_vn_mknod(dir, dentry, mode|S_IFDIR, 0);
L
Linus Torvalds 已提交
279 280 281
}

STATIC struct dentry *
282
xfs_vn_lookup(
L
Linus Torvalds 已提交
283 284 285 286
	struct inode	*dir,
	struct dentry	*dentry,
	struct nameidata *nd)
{
287
	struct xfs_inode *cip;
288
	struct xfs_name	name;
L
Linus Torvalds 已提交
289 290 291 292 293
	int		error;

	if (dentry->d_name.len >= MAXNAMELEN)
		return ERR_PTR(-ENAMETOOLONG);

294
	xfs_dentry_to_name(&name, dentry);
295
	error = xfs_lookup(XFS_I(dir), &name, &cip, NULL);
296
	if (unlikely(error)) {
L
Linus Torvalds 已提交
297 298 299 300 301 302
		if (unlikely(error != ENOENT))
			return ERR_PTR(-error);
		d_add(dentry, NULL);
		return NULL;
	}

303
	return d_splice_alias(VFS_I(cip), dentry);
L
Linus Torvalds 已提交
304 305
}

306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325
STATIC struct dentry *
xfs_vn_ci_lookup(
	struct inode	*dir,
	struct dentry	*dentry,
	struct nameidata *nd)
{
	struct xfs_inode *ip;
	struct xfs_name	xname;
	struct xfs_name ci_name;
	struct qstr	dname;
	int		error;

	if (dentry->d_name.len >= MAXNAMELEN)
		return ERR_PTR(-ENAMETOOLONG);

	xfs_dentry_to_name(&xname, dentry);
	error = xfs_lookup(XFS_I(dir), &xname, &ip, &ci_name);
	if (unlikely(error)) {
		if (unlikely(error != ENOENT))
			return ERR_PTR(-error);
326 327 328 329 330
		/*
		 * call d_add(dentry, NULL) here when d_drop_negative_children
		 * is called in xfs_vn_mknod (ie. allow negative dentries
		 * with CI filesystems).
		 */
331 332 333 334 335
		return NULL;
	}

	/* if exact match, just splice and exit */
	if (!ci_name.name)
336
		return d_splice_alias(VFS_I(ip), dentry);
337 338 339 340

	/* else case-insensitive match... */
	dname.name = ci_name.name;
	dname.len = ci_name.len;
341
	dentry = d_add_ci(dentry, VFS_I(ip), &dname);
342 343 344 345
	kmem_free(ci_name.name);
	return dentry;
}

L
Linus Torvalds 已提交
346
STATIC int
347
xfs_vn_link(
L
Linus Torvalds 已提交
348 349 350 351
	struct dentry	*old_dentry,
	struct inode	*dir,
	struct dentry	*dentry)
{
352
	struct inode	*inode = old_dentry->d_inode;
353
	struct xfs_name	name;
L
Linus Torvalds 已提交
354 355
	int		error;

356
	xfs_dentry_to_name(&name, dentry);
L
Linus Torvalds 已提交
357

358
	error = xfs_link(XFS_I(dir), XFS_I(inode), &name);
359
	if (unlikely(error))
360 361
		return -error;

362
	atomic_inc(&inode->i_count);
363 364
	d_instantiate(dentry, inode);
	return 0;
L
Linus Torvalds 已提交
365 366 367
}

STATIC int
368
xfs_vn_unlink(
L
Linus Torvalds 已提交
369 370 371
	struct inode	*dir,
	struct dentry	*dentry)
{
372
	struct xfs_name	name;
L
Linus Torvalds 已提交
373 374
	int		error;

375
	xfs_dentry_to_name(&name, dentry);
L
Linus Torvalds 已提交
376

377 378 379 380 381 382 383 384 385 386 387 388
	error = -xfs_remove(XFS_I(dir), &name, XFS_I(dentry->d_inode));
	if (error)
		return error;

	/*
	 * With unlink, the VFS makes the dentry "negative": no inode,
	 * but still hashed. This is incompatible with case-insensitive
	 * mode, so invalidate (unhash) the dentry in CI-mode.
	 */
	if (xfs_sb_version_hasasciici(&XFS_M(dir->i_sb)->m_sb))
		d_invalidate(dentry);
	return 0;
L
Linus Torvalds 已提交
389 390 391
}

STATIC int
392
xfs_vn_symlink(
L
Linus Torvalds 已提交
393 394 395 396
	struct inode	*dir,
	struct dentry	*dentry,
	const char	*symname)
{
397 398
	struct inode	*inode;
	struct xfs_inode *cip = NULL;
399
	struct xfs_name	name;
L
Linus Torvalds 已提交
400
	int		error;
401
	mode_t		mode;
L
Linus Torvalds 已提交
402

403
	mode = S_IFLNK |
A
Al Viro 已提交
404
		(irix_symlink_mode ? 0777 & ~current_umask() : S_IRWXUGO);
405
	xfs_dentry_to_name(&name, dentry);
L
Linus Torvalds 已提交
406

407
	error = xfs_symlink(XFS_I(dir), &name, symname, mode, &cip, NULL);
408 409 410
	if (unlikely(error))
		goto out;

411
	inode = VFS_I(cip);
412 413 414 415 416 417 418 419 420

	error = xfs_init_security(inode, dir);
	if (unlikely(error))
		goto out_cleanup_inode;

	d_instantiate(dentry, inode);
	return 0;

 out_cleanup_inode:
421
	xfs_cleanup_inode(dir, inode, dentry);
422
 out:
L
Linus Torvalds 已提交
423 424 425 426
	return -error;
}

STATIC int
427
xfs_vn_rename(
L
Linus Torvalds 已提交
428 429 430 431 432 433
	struct inode	*odir,
	struct dentry	*odentry,
	struct inode	*ndir,
	struct dentry	*ndentry)
{
	struct inode	*new_inode = ndentry->d_inode;
434 435
	struct xfs_name	oname;
	struct xfs_name	nname;
L
Linus Torvalds 已提交
436

437 438 439
	xfs_dentry_to_name(&oname, odentry);
	xfs_dentry_to_name(&nname, ndentry);

440
	return -xfs_rename(XFS_I(odir), &oname, XFS_I(odentry->d_inode),
441 442
			   XFS_I(ndir), &nname, new_inode ?
			   			XFS_I(new_inode) : NULL);
L
Linus Torvalds 已提交
443 444 445 446 447 448 449
}

/*
 * careful here - this function can get called recursively, so
 * we need to be very careful about how much stack we use.
 * uio is kmalloced for this reason...
 */
450
STATIC void *
451
xfs_vn_follow_link(
L
Linus Torvalds 已提交
452 453 454 455
	struct dentry		*dentry,
	struct nameidata	*nd)
{
	char			*link;
456
	int			error = -ENOMEM;
L
Linus Torvalds 已提交
457

458
	link = kmalloc(MAXPATHLEN+1, GFP_KERNEL);
459 460
	if (!link)
		goto out_err;
L
Linus Torvalds 已提交
461

462
	error = -xfs_readlink(XFS_I(dentry->d_inode), link);
463 464
	if (unlikely(error))
		goto out_kfree;
L
Linus Torvalds 已提交
465 466

	nd_set_link(nd, link);
467
	return NULL;
468 469 470 471 472 473

 out_kfree:
	kfree(link);
 out_err:
	nd_set_link(nd, ERR_PTR(error));
	return NULL;
L
Linus Torvalds 已提交
474 475
}

476
STATIC void
477
xfs_vn_put_link(
478 479 480
	struct dentry	*dentry,
	struct nameidata *nd,
	void		*p)
L
Linus Torvalds 已提交
481
{
482 483
	char		*s = nd_get_link(nd);

L
Linus Torvalds 已提交
484 485 486 487 488
	if (!IS_ERR(s))
		kfree(s);
}

STATIC int
489
xfs_vn_getattr(
490 491 492
	struct vfsmount		*mnt,
	struct dentry		*dentry,
	struct kstat		*stat)
L
Linus Torvalds 已提交
493
{
494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510
	struct inode		*inode = dentry->d_inode;
	struct xfs_inode	*ip = XFS_I(inode);
	struct xfs_mount	*mp = ip->i_mount;

	xfs_itrace_entry(ip);

	if (XFS_FORCED_SHUTDOWN(mp))
		return XFS_ERROR(EIO);

	stat->size = XFS_ISIZE(ip);
	stat->dev = inode->i_sb->s_dev;
	stat->mode = ip->i_d.di_mode;
	stat->nlink = ip->i_d.di_nlink;
	stat->uid = ip->i_d.di_uid;
	stat->gid = ip->i_d.di_gid;
	stat->ino = ip->i_ino;
	stat->atime = inode->i_atime;
511 512
	stat->mtime = inode->i_mtime;
	stat->ctime = inode->i_ctime;
513 514 515 516 517 518 519 520 521 522 523 524
	stat->blocks =
		XFS_FSB_TO_BB(mp, ip->i_d.di_nblocks + ip->i_delayed_blks);


	switch (inode->i_mode & S_IFMT) {
	case S_IFBLK:
	case S_IFCHR:
		stat->blksize = BLKDEV_IOSIZE;
		stat->rdev = MKDEV(sysv_major(ip->i_df.if_u2.if_rdev) & 0x1ff,
				   sysv_minor(ip->i_df.if_u2.if_rdev));
		break;
	default:
525
		if (XFS_IS_REALTIME_INODE(ip)) {
526 527 528 529 530 531 532 533 534 535 536
			/*
			 * If the file blocks are being allocated from a
			 * realtime volume, then return the inode's realtime
			 * extent size or the realtime volume's extent size.
			 */
			stat->blksize =
				xfs_get_extsz_hint(ip) << mp->m_sb.sb_blocklog;
		} else
			stat->blksize = xfs_preferred_iosize(mp);
		stat->rdev = 0;
		break;
537
	}
538 539

	return 0;
L
Linus Torvalds 已提交
540 541 542
}

STATIC int
543
xfs_vn_setattr(
L
Linus Torvalds 已提交
544
	struct dentry	*dentry,
545
	struct iattr	*iattr)
L
Linus Torvalds 已提交
546
{
C
Christoph Hellwig 已提交
547
	return -xfs_setattr(XFS_I(dentry->d_inode), iattr, 0);
L
Linus Torvalds 已提交
548 549
}

550 551 552 553 554
/*
 * block_truncate_page can return an error, but we can't propagate it
 * at all here. Leave a complaint + stack trace in the syslog because
 * this could be bad. If it is bad, we need to propagate the error further.
 */
L
Linus Torvalds 已提交
555
STATIC void
556
xfs_vn_truncate(
L
Linus Torvalds 已提交
557 558
	struct inode	*inode)
{
559 560 561 562
	int	error;
	error = block_truncate_page(inode->i_mapping, inode->i_size,
							xfs_get_blocks);
	WARN_ON(error);
L
Linus Torvalds 已提交
563 564
}

D
David Chinner 已提交
565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586
STATIC long
xfs_vn_fallocate(
	struct inode	*inode,
	int		mode,
	loff_t		offset,
	loff_t		len)
{
	long		error;
	loff_t		new_size = 0;
	xfs_flock64_t	bf;
	xfs_inode_t	*ip = XFS_I(inode);

	/* preallocation on directories not yet supported */
	error = -ENODEV;
	if (S_ISDIR(inode->i_mode))
		goto out_error;

	bf.l_whence = 0;
	bf.l_start = offset;
	bf.l_len = len;

	xfs_ilock(ip, XFS_IOLOCK_EXCL);
587 588 589 590 591 592 593 594 595 596

	/* check the new inode size is valid before allocating */
	if (!(mode & FALLOC_FL_KEEP_SIZE) &&
	    offset + len > i_size_read(inode)) {
		new_size = offset + len;
		error = inode_newsize_ok(inode, new_size);
		if (error)
			goto out_unlock;
	}

597 598
	error = -xfs_change_file_space(ip, XFS_IOC_RESVSP, &bf,
				       0, XFS_ATTR_NOLOCK);
599 600
	if (error)
		goto out_unlock;
D
David Chinner 已提交
601 602 603

	/* Change file size if needed */
	if (new_size) {
604
		struct iattr iattr;
D
David Chinner 已提交
605

606 607
		iattr.ia_valid = ATTR_SIZE;
		iattr.ia_size = new_size;
608
		error = -xfs_setattr(ip, &iattr, XFS_ATTR_NOLOCK);
D
David Chinner 已提交
609 610
	}

611
out_unlock:
D
David Chinner 已提交
612 613 614 615
	xfs_iunlock(ip, XFS_IOLOCK_EXCL);
out_error:
	return error;
}
L
Linus Torvalds 已提交
616

E
Eric Sandeen 已提交
617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683
#define XFS_FIEMAP_FLAGS	(FIEMAP_FLAG_SYNC|FIEMAP_FLAG_XATTR)

/*
 * Call fiemap helper to fill in user data.
 * Returns positive errors to xfs_getbmap.
 */
STATIC int
xfs_fiemap_format(
	void			**arg,
	struct getbmapx		*bmv,
	int			*full)
{
	int			error;
	struct fiemap_extent_info *fieinfo = *arg;
	u32			fiemap_flags = 0;
	u64			logical, physical, length;

	/* Do nothing for a hole */
	if (bmv->bmv_block == -1LL)
		return 0;

	logical = BBTOB(bmv->bmv_offset);
	physical = BBTOB(bmv->bmv_block);
	length = BBTOB(bmv->bmv_length);

	if (bmv->bmv_oflags & BMV_OF_PREALLOC)
		fiemap_flags |= FIEMAP_EXTENT_UNWRITTEN;
	else if (bmv->bmv_oflags & BMV_OF_DELALLOC) {
		fiemap_flags |= FIEMAP_EXTENT_DELALLOC;
		physical = 0;   /* no block yet */
	}
	if (bmv->bmv_oflags & BMV_OF_LAST)
		fiemap_flags |= FIEMAP_EXTENT_LAST;

	error = fiemap_fill_next_extent(fieinfo, logical, physical,
					length, fiemap_flags);
	if (error > 0) {
		error = 0;
		*full = 1;	/* user array now full */
	}

	return -error;
}

STATIC int
xfs_vn_fiemap(
	struct inode		*inode,
	struct fiemap_extent_info *fieinfo,
	u64			start,
	u64			length)
{
	xfs_inode_t		*ip = XFS_I(inode);
	struct getbmapx		bm;
	int			error;

	error = fiemap_check_flags(fieinfo, XFS_FIEMAP_FLAGS);
	if (error)
		return error;

	/* Set up bmap header for xfs internal routine */
	bm.bmv_offset = BTOBB(start);
	/* Special case for whole file */
	if (length == FIEMAP_MAX_OFFSET)
		bm.bmv_length = -1LL;
	else
		bm.bmv_length = BTOBB(length);

684
	/* We add one because in getbmap world count includes the header */
T
Tao Ma 已提交
685 686 687 688
	bm.bmv_count = !fieinfo->fi_extents_max ? MAXEXTNUM :
					fieinfo->fi_extents_max + 1;
	bm.bmv_count = min_t(__s32, bm.bmv_count,
			     (PAGE_SIZE * 16 / sizeof(struct getbmapx)));
E
Eric Sandeen 已提交
689 690 691 692 693 694 695 696 697 698 699 700 701
	bm.bmv_iflags = BMV_IF_PREALLOC;
	if (fieinfo->fi_flags & FIEMAP_FLAG_XATTR)
		bm.bmv_iflags |= BMV_IF_ATTRFORK;
	if (!(fieinfo->fi_flags & FIEMAP_FLAG_SYNC))
		bm.bmv_iflags |= BMV_IF_DELALLOC;

	error = xfs_getbmap(ip, &bm, xfs_fiemap_format, fieinfo);
	if (error)
		return -error;

	return 0;
}

702
static const struct inode_operations xfs_inode_operations = {
703
	.check_acl		= xfs_check_acl,
704 705 706
	.truncate		= xfs_vn_truncate,
	.getattr		= xfs_vn_getattr,
	.setattr		= xfs_vn_setattr,
707 708 709
	.setxattr		= generic_setxattr,
	.getxattr		= generic_getxattr,
	.removexattr		= generic_removexattr,
710
	.listxattr		= xfs_vn_listxattr,
D
David Chinner 已提交
711
	.fallocate		= xfs_vn_fallocate,
E
Eric Sandeen 已提交
712
	.fiemap			= xfs_vn_fiemap,
L
Linus Torvalds 已提交
713 714
};

715
static const struct inode_operations xfs_dir_inode_operations = {
716 717 718 719 720 721
	.create			= xfs_vn_create,
	.lookup			= xfs_vn_lookup,
	.link			= xfs_vn_link,
	.unlink			= xfs_vn_unlink,
	.symlink		= xfs_vn_symlink,
	.mkdir			= xfs_vn_mkdir,
722 723 724 725 726 727 728
	/*
	 * Yes, XFS uses the same method for rmdir and unlink.
	 *
	 * There are some subtile differences deeper in the code,
	 * but we use S_ISDIR to check for those.
	 */
	.rmdir			= xfs_vn_unlink,
729 730
	.mknod			= xfs_vn_mknod,
	.rename			= xfs_vn_rename,
731
	.check_acl		= xfs_check_acl,
732 733
	.getattr		= xfs_vn_getattr,
	.setattr		= xfs_vn_setattr,
734 735 736
	.setxattr		= generic_setxattr,
	.getxattr		= generic_getxattr,
	.removexattr		= generic_removexattr,
737
	.listxattr		= xfs_vn_listxattr,
L
Linus Torvalds 已提交
738 739
};

740
static const struct inode_operations xfs_dir_ci_inode_operations = {
741 742 743 744 745 746
	.create			= xfs_vn_create,
	.lookup			= xfs_vn_ci_lookup,
	.link			= xfs_vn_link,
	.unlink			= xfs_vn_unlink,
	.symlink		= xfs_vn_symlink,
	.mkdir			= xfs_vn_mkdir,
747 748 749 750 751 752 753
	/*
	 * Yes, XFS uses the same method for rmdir and unlink.
	 *
	 * There are some subtile differences deeper in the code,
	 * but we use S_ISDIR to check for those.
	 */
	.rmdir			= xfs_vn_unlink,
754 755
	.mknod			= xfs_vn_mknod,
	.rename			= xfs_vn_rename,
756
	.check_acl		= xfs_check_acl,
757 758
	.getattr		= xfs_vn_getattr,
	.setattr		= xfs_vn_setattr,
759 760 761
	.setxattr		= generic_setxattr,
	.getxattr		= generic_getxattr,
	.removexattr		= generic_removexattr,
762 763 764
	.listxattr		= xfs_vn_listxattr,
};

765
static const struct inode_operations xfs_symlink_inode_operations = {
L
Linus Torvalds 已提交
766
	.readlink		= generic_readlink,
767 768
	.follow_link		= xfs_vn_follow_link,
	.put_link		= xfs_vn_put_link,
769
	.check_acl		= xfs_check_acl,
770 771
	.getattr		= xfs_vn_getattr,
	.setattr		= xfs_vn_setattr,
772 773 774
	.setxattr		= generic_setxattr,
	.getxattr		= generic_getxattr,
	.removexattr		= generic_removexattr,
775
	.listxattr		= xfs_vn_listxattr,
L
Linus Torvalds 已提交
776
};
777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807

STATIC void
xfs_diflags_to_iflags(
	struct inode		*inode,
	struct xfs_inode	*ip)
{
	if (ip->i_d.di_flags & XFS_DIFLAG_IMMUTABLE)
		inode->i_flags |= S_IMMUTABLE;
	else
		inode->i_flags &= ~S_IMMUTABLE;
	if (ip->i_d.di_flags & XFS_DIFLAG_APPEND)
		inode->i_flags |= S_APPEND;
	else
		inode->i_flags &= ~S_APPEND;
	if (ip->i_d.di_flags & XFS_DIFLAG_SYNC)
		inode->i_flags |= S_SYNC;
	else
		inode->i_flags &= ~S_SYNC;
	if (ip->i_d.di_flags & XFS_DIFLAG_NOATIME)
		inode->i_flags |= S_NOATIME;
	else
		inode->i_flags &= ~S_NOATIME;
}

/*
 * Initialize the Linux inode, set up the operation vectors and
 * unlock the inode.
 *
 * When reading existing inodes from disk this is called directly
 * from xfs_iget, when creating a new inode it is called from
 * xfs_ialloc after setting up the inode.
808 809 810 811
 *
 * We are always called with an uninitialised linux inode here.
 * We need to initialise the necessary fields and take a reference
 * on it.
812 813 814 815 816
 */
void
xfs_setup_inode(
	struct xfs_inode	*ip)
{
817 818 819
	struct inode		*inode = &ip->i_vnode;

	inode->i_ino = ip->i_ino;
C
Christoph Hellwig 已提交
820
	inode->i_state = I_NEW;
821
	inode_add_to_lists(ip->i_mount->m_super, inode);
822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878

	inode->i_mode	= ip->i_d.di_mode;
	inode->i_nlink	= ip->i_d.di_nlink;
	inode->i_uid	= ip->i_d.di_uid;
	inode->i_gid	= ip->i_d.di_gid;

	switch (inode->i_mode & S_IFMT) {
	case S_IFBLK:
	case S_IFCHR:
		inode->i_rdev =
			MKDEV(sysv_major(ip->i_df.if_u2.if_rdev) & 0x1ff,
			      sysv_minor(ip->i_df.if_u2.if_rdev));
		break;
	default:
		inode->i_rdev = 0;
		break;
	}

	inode->i_generation = ip->i_d.di_gen;
	i_size_write(inode, ip->i_d.di_size);
	inode->i_atime.tv_sec	= ip->i_d.di_atime.t_sec;
	inode->i_atime.tv_nsec	= ip->i_d.di_atime.t_nsec;
	inode->i_mtime.tv_sec	= ip->i_d.di_mtime.t_sec;
	inode->i_mtime.tv_nsec	= ip->i_d.di_mtime.t_nsec;
	inode->i_ctime.tv_sec	= ip->i_d.di_ctime.t_sec;
	inode->i_ctime.tv_nsec	= ip->i_d.di_ctime.t_nsec;
	xfs_diflags_to_iflags(inode, ip);

	switch (inode->i_mode & S_IFMT) {
	case S_IFREG:
		inode->i_op = &xfs_inode_operations;
		inode->i_fop = &xfs_file_operations;
		inode->i_mapping->a_ops = &xfs_address_space_operations;
		break;
	case S_IFDIR:
		if (xfs_sb_version_hasasciici(&XFS_M(inode->i_sb)->m_sb))
			inode->i_op = &xfs_dir_ci_inode_operations;
		else
			inode->i_op = &xfs_dir_inode_operations;
		inode->i_fop = &xfs_dir_file_operations;
		break;
	case S_IFLNK:
		inode->i_op = &xfs_symlink_inode_operations;
		if (!(ip->i_df.if_flags & XFS_IFINLINE))
			inode->i_mapping->a_ops = &xfs_address_space_operations;
		break;
	default:
		inode->i_op = &xfs_inode_operations;
		init_special_inode(inode, inode->i_mode, inode->i_rdev);
		break;
	}

	xfs_iflags_clear(ip, XFS_INEW);
	barrier();

	unlock_new_inode(inode);
}