xattr.c 27.1 KB
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 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 1034 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 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106
/*
 *   Copyright (C) International Business Machines  Corp., 2000-2004
 *   Copyright (C) Christoph Hellwig, 2002
 *
 *   This program is free software;  you can redistribute it and/or modify
 *   it under the terms of the GNU General Public License as published by
 *   the Free Software Foundation; either version 2 of the License, or
 *   (at your option) any later version.
 *
 *   This program is distributed in the hope that it will 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.
 *
 *   You should have received a copy of the GNU General Public License
 *   along with this program;  if not, write to the Free Software
 *   Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
 */

#include <linux/capability.h>
#include <linux/fs.h>
#include <linux/xattr.h>
#include <linux/posix_acl_xattr.h>
#include <linux/slab.h>
#include <linux/quotaops.h>
#include <linux/security.h>
#include "jfs_incore.h"
#include "jfs_superblock.h"
#include "jfs_dmap.h"
#include "jfs_debug.h"
#include "jfs_dinode.h"
#include "jfs_extent.h"
#include "jfs_metapage.h"
#include "jfs_xattr.h"
#include "jfs_acl.h"

/*
 *	jfs_xattr.c: extended attribute service
 *
 * Overall design --
 *
 * Format:
 *
 *   Extended attribute lists (jfs_ea_list) consist of an overall size (32 bit
 *   value) and a variable (0 or more) number of extended attribute
 *   entries.  Each extended attribute entry (jfs_ea) is a <name,value> double
 *   where <name> is constructed from a null-terminated ascii string
 *   (1 ... 255 bytes in the name) and <value> is arbitrary 8 bit data
 *   (1 ... 65535 bytes).  The in-memory format is
 *
 *   0       1        2        4                4 + namelen + 1
 *   +-------+--------+--------+----------------+-------------------+
 *   | Flags | Name   | Value  | Name String \0 | Data . . . .      |
 *   |       | Length | Length |                |                   |
 *   +-------+--------+--------+----------------+-------------------+
 *
 *   A jfs_ea_list then is structured as
 *
 *   0            4                   4 + EA_SIZE(ea1)
 *   +------------+-------------------+--------------------+-----
 *   | Overall EA | First FEA Element | Second FEA Element | .....
 *   | List Size  |                   |                    |
 *   +------------+-------------------+--------------------+-----
 *
 *   On-disk:
 *
 *	FEALISTs are stored on disk using blocks allocated by dbAlloc() and
 *	written directly. An EA list may be in-lined in the inode if there is
 *	sufficient room available.
 */

struct ea_buffer {
	int flag;		/* Indicates what storage xattr points to */
	int max_size;		/* largest xattr that fits in current buffer */
	dxd_t new_ea;		/* dxd to replace ea when modifying xattr */
	struct metapage *mp;	/* metapage containing ea list */
	struct jfs_ea_list *xattr;	/* buffer containing ea list */
};

/*
 * ea_buffer.flag values
 */
#define EA_INLINE	0x0001
#define EA_EXTENT	0x0002
#define EA_NEW		0x0004
#define EA_MALLOC	0x0008


static int is_known_namespace(const char *name)
{
	if (strncmp(name, XATTR_SYSTEM_PREFIX, XATTR_SYSTEM_PREFIX_LEN) &&
	    strncmp(name, XATTR_USER_PREFIX, XATTR_USER_PREFIX_LEN) &&
	    strncmp(name, XATTR_SECURITY_PREFIX, XATTR_SECURITY_PREFIX_LEN) &&
	    strncmp(name, XATTR_TRUSTED_PREFIX, XATTR_TRUSTED_PREFIX_LEN))
		return false;

	return true;
}

/*
 * These three routines are used to recognize on-disk extended attributes
 * that are in a recognized namespace.  If the attribute is not recognized,
 * "os2." is prepended to the name
 */
static int is_os2_xattr(struct jfs_ea *ea)
{
	return !is_known_namespace(ea->name);
}

static inline int name_size(struct jfs_ea *ea)
{
	if (is_os2_xattr(ea))
		return ea->namelen + XATTR_OS2_PREFIX_LEN;
	else
		return ea->namelen;
}

static inline int copy_name(char *buffer, struct jfs_ea *ea)
{
	int len = ea->namelen;

	if (is_os2_xattr(ea)) {
		memcpy(buffer, XATTR_OS2_PREFIX, XATTR_OS2_PREFIX_LEN);
		buffer += XATTR_OS2_PREFIX_LEN;
		len += XATTR_OS2_PREFIX_LEN;
	}
	memcpy(buffer, ea->name, ea->namelen);
	buffer[ea->namelen] = 0;

	return len;
}

/* Forward references */
static void ea_release(struct inode *inode, struct ea_buffer *ea_buf);

/*
 * NAME: ea_write_inline
 *
 * FUNCTION: Attempt to write an EA inline if area is available
 *
 * PRE CONDITIONS:
 *	Already verified that the specified EA is small enough to fit inline
 *
 * PARAMETERS:
 *	ip	- Inode pointer
 *	ealist	- EA list pointer
 *	size	- size of ealist in bytes
 *	ea	- dxd_t structure to be filled in with necessary EA information
 *		  if we successfully copy the EA inline
 *
 * NOTES:
 *	Checks if the inode's inline area is available.  If so, copies EA inline
 *	and sets <ea> fields appropriately.  Otherwise, returns failure, EA will
 *	have to be put into an extent.
 *
 * RETURNS: 0 for successful copy to inline area; -1 if area not available
 */
static int ea_write_inline(struct inode *ip, struct jfs_ea_list *ealist,
			   int size, dxd_t * ea)
{
	struct jfs_inode_info *ji = JFS_IP(ip);

	/*
	 * Make sure we have an EA -- the NULL EA list is valid, but you
	 * can't copy it!
	 */
	if (ealist && size > sizeof (struct jfs_ea_list)) {
		assert(size <= sizeof (ji->i_inline_ea));

		/*
		 * See if the space is available or if it is already being
		 * used for an inline EA.
		 */
		if (!(ji->mode2 & INLINEEA) && !(ji->ea.flag & DXD_INLINE))
			return -EPERM;

		DXDsize(ea, size);
		DXDlength(ea, 0);
		DXDaddress(ea, 0);
		memcpy(ji->i_inline_ea, ealist, size);
		ea->flag = DXD_INLINE;
		ji->mode2 &= ~INLINEEA;
	} else {
		ea->flag = 0;
		DXDsize(ea, 0);
		DXDlength(ea, 0);
		DXDaddress(ea, 0);

		/* Free up INLINE area */
		if (ji->ea.flag & DXD_INLINE)
			ji->mode2 |= INLINEEA;
	}

	return 0;
}

/*
 * NAME: ea_write
 *
 * FUNCTION: Write an EA for an inode
 *
 * PRE CONDITIONS: EA has been verified
 *
 * PARAMETERS:
 *	ip	- Inode pointer
 *	ealist	- EA list pointer
 *	size	- size of ealist in bytes
 *	ea	- dxd_t structure to be filled in appropriately with where the
 *		  EA was copied
 *
 * NOTES: Will write EA inline if able to, otherwise allocates blocks for an
 *	extent and synchronously writes it to those blocks.
 *
 * RETURNS: 0 for success; Anything else indicates failure
 */
static int ea_write(struct inode *ip, struct jfs_ea_list *ealist, int size,
		       dxd_t * ea)
{
	struct super_block *sb = ip->i_sb;
	struct jfs_inode_info *ji = JFS_IP(ip);
	struct jfs_sb_info *sbi = JFS_SBI(sb);
	int nblocks;
	s64 blkno;
	int rc = 0, i;
	char *cp;
	s32 nbytes, nb;
	s32 bytes_to_write;
	struct metapage *mp;

	/*
	 * Quick check to see if this is an in-linable EA.  Short EAs
	 * and empty EAs are all in-linable, provided the space exists.
	 */
	if (!ealist || size <= sizeof (ji->i_inline_ea)) {
		if (!ea_write_inline(ip, ealist, size, ea))
			return 0;
	}

	/* figure out how many blocks we need */
	nblocks = (size + (sb->s_blocksize - 1)) >> sb->s_blocksize_bits;

	/* Allocate new blocks to quota. */
	rc = dquot_alloc_block(ip, nblocks);
	if (rc)
		return rc;

	rc = dbAlloc(ip, INOHINT(ip), nblocks, &blkno);
	if (rc) {
		/*Rollback quota allocation. */
		dquot_free_block(ip, nblocks);
		return rc;
	}

	/*
	 * Now have nblocks worth of storage to stuff into the FEALIST.
	 * loop over the FEALIST copying data into the buffer one page at
	 * a time.
	 */
	cp = (char *) ealist;
	nbytes = size;
	for (i = 0; i < nblocks; i += sbi->nbperpage) {
		/*
		 * Determine how many bytes for this request, and round up to
		 * the nearest aggregate block size
		 */
		nb = min(PSIZE, nbytes);
		bytes_to_write =
		    ((((nb + sb->s_blocksize - 1)) >> sb->s_blocksize_bits))
		    << sb->s_blocksize_bits;

		if (!(mp = get_metapage(ip, blkno + i, bytes_to_write, 1))) {
			rc = -EIO;
			goto failed;
		}

		memcpy(mp->data, cp, nb);

		/*
		 * We really need a way to propagate errors for
		 * forced writes like this one.  --hch
		 *
		 * (__write_metapage => release_metapage => flush_metapage)
		 */
#ifdef _JFS_FIXME
		if ((rc = flush_metapage(mp))) {
			/*
			 * the write failed -- this means that the buffer
			 * is still assigned and the blocks are not being
			 * used.  this seems like the best error recovery
			 * we can get ...
			 */
			goto failed;
		}
#else
		flush_metapage(mp);
#endif

		cp += PSIZE;
		nbytes -= nb;
	}

	ea->flag = DXD_EXTENT;
	DXDsize(ea, le32_to_cpu(ealist->size));
	DXDlength(ea, nblocks);
	DXDaddress(ea, blkno);

	/* Free up INLINE area */
	if (ji->ea.flag & DXD_INLINE)
		ji->mode2 |= INLINEEA;

	return 0;

      failed:
	/* Rollback quota allocation. */
	dquot_free_block(ip, nblocks);

	dbFree(ip, blkno, nblocks);
	return rc;
}

/*
 * NAME: ea_read_inline
 *
 * FUNCTION: Read an inlined EA into user's buffer
 *
 * PARAMETERS:
 *	ip	- Inode pointer
 *	ealist	- Pointer to buffer to fill in with EA
 *
 * RETURNS: 0
 */
static int ea_read_inline(struct inode *ip, struct jfs_ea_list *ealist)
{
	struct jfs_inode_info *ji = JFS_IP(ip);
	int ea_size = sizeDXD(&ji->ea);

	if (ea_size == 0) {
		ealist->size = 0;
		return 0;
	}

	/* Sanity Check */
	if ((sizeDXD(&ji->ea) > sizeof (ji->i_inline_ea)))
		return -EIO;
	if (le32_to_cpu(((struct jfs_ea_list *) &ji->i_inline_ea)->size)
	    != ea_size)
		return -EIO;

	memcpy(ealist, ji->i_inline_ea, ea_size);
	return 0;
}

/*
 * NAME: ea_read
 *
 * FUNCTION: copy EA data into user's buffer
 *
 * PARAMETERS:
 *	ip	- Inode pointer
 *	ealist	- Pointer to buffer to fill in with EA
 *
 * NOTES:  If EA is inline calls ea_read_inline() to copy EA.
 *
 * RETURNS: 0 for success; other indicates failure
 */
static int ea_read(struct inode *ip, struct jfs_ea_list *ealist)
{
	struct super_block *sb = ip->i_sb;
	struct jfs_inode_info *ji = JFS_IP(ip);
	struct jfs_sb_info *sbi = JFS_SBI(sb);
	int nblocks;
	s64 blkno;
	char *cp = (char *) ealist;
	int i;
	int nbytes, nb;
	s32 bytes_to_read;
	struct metapage *mp;

	/* quick check for in-line EA */
	if (ji->ea.flag & DXD_INLINE)
		return ea_read_inline(ip, ealist);

	nbytes = sizeDXD(&ji->ea);
	if (!nbytes) {
		jfs_error(sb, "nbytes is 0\n");
		return -EIO;
	}

	/*
	 * Figure out how many blocks were allocated when this EA list was
	 * originally written to disk.
	 */
	nblocks = lengthDXD(&ji->ea) << sbi->l2nbperpage;
	blkno = addressDXD(&ji->ea) << sbi->l2nbperpage;

	/*
	 * I have found the disk blocks which were originally used to store
	 * the FEALIST.  now i loop over each contiguous block copying the
	 * data into the buffer.
	 */
	for (i = 0; i < nblocks; i += sbi->nbperpage) {
		/*
		 * Determine how many bytes for this request, and round up to
		 * the nearest aggregate block size
		 */
		nb = min(PSIZE, nbytes);
		bytes_to_read =
		    ((((nb + sb->s_blocksize - 1)) >> sb->s_blocksize_bits))
		    << sb->s_blocksize_bits;

		if (!(mp = read_metapage(ip, blkno + i, bytes_to_read, 1)))
			return -EIO;

		memcpy(cp, mp->data, nb);
		release_metapage(mp);

		cp += PSIZE;
		nbytes -= nb;
	}

	return 0;
}

/*
 * NAME: ea_get
 *
 * FUNCTION: Returns buffer containing existing extended attributes.
 *	     The size of the buffer will be the larger of the existing
 *	     attributes size, or min_size.
 *
 *	     The buffer, which may be inlined in the inode or in the
 *	     page cache must be release by calling ea_release or ea_put
 *
 * PARAMETERS:
 *	inode	- Inode pointer
 *	ea_buf	- Structure to be populated with ealist and its metadata
 *	min_size- minimum size of buffer to be returned
 *
 * RETURNS: 0 for success; Other indicates failure
 */
static int ea_get(struct inode *inode, struct ea_buffer *ea_buf, int min_size)
{
	struct jfs_inode_info *ji = JFS_IP(inode);
	struct super_block *sb = inode->i_sb;
	int size;
	int ea_size = sizeDXD(&ji->ea);
	int blocks_needed, current_blocks;
	s64 blkno;
	int rc;
	int quota_allocation = 0;

	/* When fsck.jfs clears a bad ea, it doesn't clear the size */
	if (ji->ea.flag == 0)
		ea_size = 0;

	if (ea_size == 0) {
		if (min_size == 0) {
			ea_buf->flag = 0;
			ea_buf->max_size = 0;
			ea_buf->xattr = NULL;
			return 0;
		}
		if ((min_size <= sizeof (ji->i_inline_ea)) &&
		    (ji->mode2 & INLINEEA)) {
			ea_buf->flag = EA_INLINE | EA_NEW;
			ea_buf->max_size = sizeof (ji->i_inline_ea);
			ea_buf->xattr = (struct jfs_ea_list *) ji->i_inline_ea;
			DXDlength(&ea_buf->new_ea, 0);
			DXDaddress(&ea_buf->new_ea, 0);
			ea_buf->new_ea.flag = DXD_INLINE;
			DXDsize(&ea_buf->new_ea, min_size);
			return 0;
		}
		current_blocks = 0;
	} else if (ji->ea.flag & DXD_INLINE) {
		if (min_size <= sizeof (ji->i_inline_ea)) {
			ea_buf->flag = EA_INLINE;
			ea_buf->max_size = sizeof (ji->i_inline_ea);
			ea_buf->xattr = (struct jfs_ea_list *) ji->i_inline_ea;
			goto size_check;
		}
		current_blocks = 0;
	} else {
		if (!(ji->ea.flag & DXD_EXTENT)) {
			jfs_error(sb, "invalid ea.flag\n");
			return -EIO;
		}
		current_blocks = (ea_size + sb->s_blocksize - 1) >>
		    sb->s_blocksize_bits;
	}
	size = max(min_size, ea_size);

	if (size > PSIZE) {
		/*
		 * To keep the rest of the code simple.  Allocate a
		 * contiguous buffer to work with
		 */
		ea_buf->xattr = kmalloc(size, GFP_KERNEL);
		if (ea_buf->xattr == NULL)
			return -ENOMEM;

		ea_buf->flag = EA_MALLOC;
		ea_buf->max_size = (size + sb->s_blocksize - 1) &
		    ~(sb->s_blocksize - 1);

		if (ea_size == 0)
			return 0;

		if ((rc = ea_read(inode, ea_buf->xattr))) {
			kfree(ea_buf->xattr);
			ea_buf->xattr = NULL;
			return rc;
		}
		goto size_check;
	}
	blocks_needed = (min_size + sb->s_blocksize - 1) >>
	    sb->s_blocksize_bits;

	if (blocks_needed > current_blocks) {
		/* Allocate new blocks to quota. */
		rc = dquot_alloc_block(inode, blocks_needed);
		if (rc)
			return -EDQUOT;

		quota_allocation = blocks_needed;

		rc = dbAlloc(inode, INOHINT(inode), (s64) blocks_needed,
			     &blkno);
		if (rc)
			goto clean_up;

		DXDlength(&ea_buf->new_ea, blocks_needed);
		DXDaddress(&ea_buf->new_ea, blkno);
		ea_buf->new_ea.flag = DXD_EXTENT;
		DXDsize(&ea_buf->new_ea, min_size);

		ea_buf->flag = EA_EXTENT | EA_NEW;

		ea_buf->mp = get_metapage(inode, blkno,
					  blocks_needed << sb->s_blocksize_bits,
					  1);
		if (ea_buf->mp == NULL) {
			dbFree(inode, blkno, (s64) blocks_needed);
			rc = -EIO;
			goto clean_up;
		}
		ea_buf->xattr = ea_buf->mp->data;
		ea_buf->max_size = (min_size + sb->s_blocksize - 1) &
		    ~(sb->s_blocksize - 1);
		if (ea_size == 0)
			return 0;
		if ((rc = ea_read(inode, ea_buf->xattr))) {
			discard_metapage(ea_buf->mp);
			dbFree(inode, blkno, (s64) blocks_needed);
			goto clean_up;
		}
		goto size_check;
	}
	ea_buf->flag = EA_EXTENT;
	ea_buf->mp = read_metapage(inode, addressDXD(&ji->ea),
				   lengthDXD(&ji->ea) << sb->s_blocksize_bits,
				   1);
	if (ea_buf->mp == NULL) {
		rc = -EIO;
		goto clean_up;
	}
	ea_buf->xattr = ea_buf->mp->data;
	ea_buf->max_size = (ea_size + sb->s_blocksize - 1) &
	    ~(sb->s_blocksize - 1);

      size_check:
	if (EALIST_SIZE(ea_buf->xattr) != ea_size) {
		printk(KERN_ERR "ea_get: invalid extended attribute\n");
		print_hex_dump(KERN_ERR, "", DUMP_PREFIX_ADDRESS, 16, 1,
				     ea_buf->xattr, ea_size, 1);
		ea_release(inode, ea_buf);
		rc = -EIO;
		goto clean_up;
	}

	return ea_size;

      clean_up:
	/* Rollback quota allocation */
	if (quota_allocation)
		dquot_free_block(inode, quota_allocation);

	return (rc);
}

static void ea_release(struct inode *inode, struct ea_buffer *ea_buf)
{
	if (ea_buf->flag & EA_MALLOC)
		kfree(ea_buf->xattr);
	else if (ea_buf->flag & EA_EXTENT) {
		assert(ea_buf->mp);
		release_metapage(ea_buf->mp);

		if (ea_buf->flag & EA_NEW)
			dbFree(inode, addressDXD(&ea_buf->new_ea),
			       lengthDXD(&ea_buf->new_ea));
	}
}

static int ea_put(tid_t tid, struct inode *inode, struct ea_buffer *ea_buf,
		  int new_size)
{
	struct jfs_inode_info *ji = JFS_IP(inode);
	unsigned long old_blocks, new_blocks;
	int rc = 0;

	if (new_size == 0) {
		ea_release(inode, ea_buf);
		ea_buf = NULL;
	} else if (ea_buf->flag & EA_INLINE) {
		assert(new_size <= sizeof (ji->i_inline_ea));
		ji->mode2 &= ~INLINEEA;
		ea_buf->new_ea.flag = DXD_INLINE;
		DXDsize(&ea_buf->new_ea, new_size);
		DXDaddress(&ea_buf->new_ea, 0);
		DXDlength(&ea_buf->new_ea, 0);
	} else if (ea_buf->flag & EA_MALLOC) {
		rc = ea_write(inode, ea_buf->xattr, new_size, &ea_buf->new_ea);
		kfree(ea_buf->xattr);
	} else if (ea_buf->flag & EA_NEW) {
		/* We have already allocated a new dxd */
		flush_metapage(ea_buf->mp);
	} else {
		/* ->xattr must point to original ea's metapage */
		rc = ea_write(inode, ea_buf->xattr, new_size, &ea_buf->new_ea);
		discard_metapage(ea_buf->mp);
	}
	if (rc)
		return rc;

	old_blocks = new_blocks = 0;

	if (ji->ea.flag & DXD_EXTENT) {
		invalidate_dxd_metapages(inode, ji->ea);
		old_blocks = lengthDXD(&ji->ea);
	}

	if (ea_buf) {
		txEA(tid, inode, &ji->ea, &ea_buf->new_ea);
		if (ea_buf->new_ea.flag & DXD_EXTENT) {
			new_blocks = lengthDXD(&ea_buf->new_ea);
			if (ji->ea.flag & DXD_INLINE)
				ji->mode2 |= INLINEEA;
		}
		ji->ea = ea_buf->new_ea;
	} else {
		txEA(tid, inode, &ji->ea, NULL);
		if (ji->ea.flag & DXD_INLINE)
			ji->mode2 |= INLINEEA;
		ji->ea.flag = 0;
		ji->ea.size = 0;
	}

	/* If old blocks exist, they must be removed from quota allocation. */
	if (old_blocks)
		dquot_free_block(inode, old_blocks);

	inode->i_ctime = CURRENT_TIME;

	return 0;
}

/*
 * Most of the permission checking is done by xattr_permission in the vfs.
 * We also need to verify that this is a namespace that we recognize.
 */
static int can_set_xattr(struct inode *inode, const char *name,
			 const void *value, size_t value_len)
{
	if (!strncmp(name, XATTR_OS2_PREFIX, XATTR_OS2_PREFIX_LEN)) {
		/*
		 * This makes sure that we aren't trying to set an
		 * attribute in a different namespace by prefixing it
		 * with "os2."
		 */
		if (is_known_namespace(name + XATTR_OS2_PREFIX_LEN))
			return -EOPNOTSUPP;
		return 0;
	}

	/*
	 * Don't allow setting an attribute in an unknown namespace.
	 */
	if (strncmp(name, XATTR_TRUSTED_PREFIX, XATTR_TRUSTED_PREFIX_LEN) &&
	    strncmp(name, XATTR_SECURITY_PREFIX, XATTR_SECURITY_PREFIX_LEN) &&
	    strncmp(name, XATTR_USER_PREFIX, XATTR_USER_PREFIX_LEN))
		return -EOPNOTSUPP;

	return 0;
}

int __jfs_setxattr(tid_t tid, struct inode *inode, const char *name,
		   const void *value, size_t value_len, int flags)
{
	struct jfs_ea_list *ealist;
	struct jfs_ea *ea, *old_ea = NULL, *next_ea = NULL;
	struct ea_buffer ea_buf;
	int old_ea_size = 0;
	int xattr_size;
	int new_size;
	int namelen = strlen(name);
	char *os2name = NULL;
	int found = 0;
	int rc;
	int length;

	if (strncmp(name, XATTR_OS2_PREFIX, XATTR_OS2_PREFIX_LEN) == 0) {
		os2name = kmalloc(namelen - XATTR_OS2_PREFIX_LEN + 1,
				  GFP_KERNEL);
		if (!os2name)
			return -ENOMEM;
		strcpy(os2name, name + XATTR_OS2_PREFIX_LEN);
		name = os2name;
		namelen -= XATTR_OS2_PREFIX_LEN;
	}

	down_write(&JFS_IP(inode)->xattr_sem);

	xattr_size = ea_get(inode, &ea_buf, 0);
	if (xattr_size < 0) {
		rc = xattr_size;
		goto out;
	}

      again:
	ealist = (struct jfs_ea_list *) ea_buf.xattr;
	new_size = sizeof (struct jfs_ea_list);

	if (xattr_size) {
		for (ea = FIRST_EA(ealist); ea < END_EALIST(ealist);
		     ea = NEXT_EA(ea)) {
			if ((namelen == ea->namelen) &&
			    (memcmp(name, ea->name, namelen) == 0)) {
				found = 1;
				if (flags & XATTR_CREATE) {
					rc = -EEXIST;
					goto release;
				}
				old_ea = ea;
				old_ea_size = EA_SIZE(ea);
				next_ea = NEXT_EA(ea);
			} else
				new_size += EA_SIZE(ea);
		}
	}

	if (!found) {
		if (flags & XATTR_REPLACE) {
			rc = -ENODATA;
			goto release;
		}
		if (value == NULL) {
			rc = 0;
			goto release;
		}
	}
	if (value)
		new_size += sizeof (struct jfs_ea) + namelen + 1 + value_len;

	if (new_size > ea_buf.max_size) {
		/*
		 * We need to allocate more space for merged ea list.
		 * We should only have loop to again: once.
		 */
		ea_release(inode, &ea_buf);
		xattr_size = ea_get(inode, &ea_buf, new_size);
		if (xattr_size < 0) {
			rc = xattr_size;
			goto out;
		}
		goto again;
	}

	/* Remove old ea of the same name */
	if (found) {
		/* number of bytes following target EA */
		length = (char *) END_EALIST(ealist) - (char *) next_ea;
		if (length > 0)
			memmove(old_ea, next_ea, length);
		xattr_size -= old_ea_size;
	}

	/* Add new entry to the end */
	if (value) {
		if (xattr_size == 0)
			/* Completely new ea list */
			xattr_size = sizeof (struct jfs_ea_list);

		/*
		 * The size of EA value is limitted by on-disk format up to
		 *  __le16, there would be an overflow if the size is equal
		 * to XATTR_SIZE_MAX (65536).  In order to avoid this issue,
		 * we can pre-checkup the value size against USHRT_MAX, and
		 * return -E2BIG in this case, which is consistent with the
		 * VFS setxattr interface.
		 */
		if (value_len >= USHRT_MAX) {
			rc = -E2BIG;
			goto release;
		}

		ea = (struct jfs_ea *) ((char *) ealist + xattr_size);
		ea->flag = 0;
		ea->namelen = namelen;
		ea->valuelen = (cpu_to_le16(value_len));
		memcpy(ea->name, name, namelen);
		ea->name[namelen] = 0;
		if (value_len)
			memcpy(&ea->name[namelen + 1], value, value_len);
		xattr_size += EA_SIZE(ea);
	}

	/* DEBUG - If we did this right, these number match */
	if (xattr_size != new_size) {
		printk(KERN_ERR
		       "__jfs_setxattr: xattr_size = %d, new_size = %d\n",
		       xattr_size, new_size);

		rc = -EINVAL;
		goto release;
	}

	/*
	 * If we're left with an empty list, there's no ea
	 */
	if (new_size == sizeof (struct jfs_ea_list))
		new_size = 0;

	ealist->size = cpu_to_le32(new_size);

	rc = ea_put(tid, inode, &ea_buf, new_size);

	goto out;
      release:
	ea_release(inode, &ea_buf);
      out:
	up_write(&JFS_IP(inode)->xattr_sem);

	kfree(os2name);

	return rc;
}

int jfs_setxattr(struct dentry *dentry, const char *name, const void *value,
		 size_t value_len, int flags)
{
	struct inode *inode = dentry->d_inode;
	struct jfs_inode_info *ji = JFS_IP(inode);
	int rc;
	tid_t tid;

	/*
	 * If this is a request for a synthetic attribute in the system.*
	 * namespace use the generic infrastructure to resolve a handler
	 * for it via sb->s_xattr.
	 */
	if (!strncmp(name, XATTR_SYSTEM_PREFIX, XATTR_SYSTEM_PREFIX_LEN))
		return generic_setxattr(dentry, name, value, value_len, flags);

	if ((rc = can_set_xattr(inode, name, value, value_len)))
		return rc;

	if (value == NULL) {	/* empty EA, do not remove */
		value = "";
		value_len = 0;
	}

	tid = txBegin(inode->i_sb, 0);
	mutex_lock(&ji->commit_mutex);
	rc = __jfs_setxattr(tid, dentry->d_inode, name, value, value_len,
			    flags);
	if (!rc)
		rc = txCommit(tid, 1, &inode, 0);
	txEnd(tid);
	mutex_unlock(&ji->commit_mutex);

	return rc;
}

ssize_t __jfs_getxattr(struct inode *inode, const char *name, void *data,
		       size_t buf_size)
{
	struct jfs_ea_list *ealist;
	struct jfs_ea *ea;
	struct ea_buffer ea_buf;
	int xattr_size;
	ssize_t size;
	int namelen = strlen(name);
	char *value;

	down_read(&JFS_IP(inode)->xattr_sem);

	xattr_size = ea_get(inode, &ea_buf, 0);

	if (xattr_size < 0) {
		size = xattr_size;
		goto out;
	}

	if (xattr_size == 0)
		goto not_found;

	ealist = (struct jfs_ea_list *) ea_buf.xattr;

	/* Find the named attribute */
	for (ea = FIRST_EA(ealist); ea < END_EALIST(ealist); ea = NEXT_EA(ea))
		if ((namelen == ea->namelen) &&
		    memcmp(name, ea->name, namelen) == 0) {
			/* Found it */
			size = le16_to_cpu(ea->valuelen);
			if (!data)
				goto release;
			else if (size > buf_size) {
				size = -ERANGE;
				goto release;
			}
			value = ((char *) &ea->name) + ea->namelen + 1;
			memcpy(data, value, size);
			goto release;
		}
      not_found:
	size = -ENODATA;
      release:
	ea_release(inode, &ea_buf);
      out:
	up_read(&JFS_IP(inode)->xattr_sem);

	return size;
}

ssize_t jfs_getxattr(struct dentry *dentry, const char *name, void *data,
		     size_t buf_size)
{
	int err;

	/*
	 * If this is a request for a synthetic attribute in the system.*
	 * namespace use the generic infrastructure to resolve a handler
	 * for it via sb->s_xattr.
	 */
	if (!strncmp(name, XATTR_SYSTEM_PREFIX, XATTR_SYSTEM_PREFIX_LEN))
		return generic_getxattr(dentry, name, data, buf_size);

	if (strncmp(name, XATTR_OS2_PREFIX, XATTR_OS2_PREFIX_LEN) == 0) {
		/*
		 * skip past "os2." prefix
		 */
		name += XATTR_OS2_PREFIX_LEN;
		/*
		 * Don't allow retrieving properly prefixed attributes
		 * by prepending them with "os2."
		 */
		if (is_known_namespace(name))
			return -EOPNOTSUPP;
	}

	err = __jfs_getxattr(dentry->d_inode, name, data, buf_size);

	return err;
}

/*
 * No special permissions are needed to list attributes except for trusted.*
 */
static inline int can_list(struct jfs_ea *ea)
{
	return (strncmp(ea->name, XATTR_TRUSTED_PREFIX,
			    XATTR_TRUSTED_PREFIX_LEN) ||
		capable(CAP_SYS_ADMIN));
}

ssize_t jfs_listxattr(struct dentry * dentry, char *data, size_t buf_size)
{
	struct inode *inode = dentry->d_inode;
	char *buffer;
	ssize_t size = 0;
	int xattr_size;
	struct jfs_ea_list *ealist;
	struct jfs_ea *ea;
	struct ea_buffer ea_buf;

	down_read(&JFS_IP(inode)->xattr_sem);

	xattr_size = ea_get(inode, &ea_buf, 0);
	if (xattr_size < 0) {
		size = xattr_size;
		goto out;
	}

	if (xattr_size == 0)
		goto release;

	ealist = (struct jfs_ea_list *) ea_buf.xattr;

	/* compute required size of list */
	for (ea = FIRST_EA(ealist); ea < END_EALIST(ealist); ea = NEXT_EA(ea)) {
		if (can_list(ea))
			size += name_size(ea) + 1;
	}

	if (!data)
		goto release;

	if (size > buf_size) {
		size = -ERANGE;
		goto release;
	}

	/* Copy attribute names to buffer */
	buffer = data;
	for (ea = FIRST_EA(ealist); ea < END_EALIST(ealist); ea = NEXT_EA(ea)) {
		if (can_list(ea)) {
			int namelen = copy_name(buffer, ea);
			buffer += namelen + 1;
		}
	}

      release:
	ea_release(inode, &ea_buf);
      out:
	up_read(&JFS_IP(inode)->xattr_sem);
	return size;
}

int jfs_removexattr(struct dentry *dentry, const char *name)
{
	struct inode *inode = dentry->d_inode;
	struct jfs_inode_info *ji = JFS_IP(inode);
	int rc;
	tid_t tid;

	/*
	 * If this is a request for a synthetic attribute in the system.*
	 * namespace use the generic infrastructure to resolve a handler
	 * for it via sb->s_xattr.
	 */
	if (!strncmp(name, XATTR_SYSTEM_PREFIX, XATTR_SYSTEM_PREFIX_LEN))
		return generic_removexattr(dentry, name);

	if ((rc = can_set_xattr(inode, name, NULL, 0)))
		return rc;

	tid = txBegin(inode->i_sb, 0);
	mutex_lock(&ji->commit_mutex);
	rc = __jfs_setxattr(tid, dentry->d_inode, name, NULL, 0, XATTR_REPLACE);
	if (!rc)
		rc = txCommit(tid, 1, &inode, 0);
	txEnd(tid);
	mutex_unlock(&ji->commit_mutex);

	return rc;
}

/*
 * List of handlers for synthetic system.* attributes.  All real ondisk
 * attributes are handled directly.
 */
const struct xattr_handler *jfs_xattr_handlers[] = {
#ifdef CONFIG_JFS_POSIX_ACL
	&posix_acl_access_xattr_handler,
	&posix_acl_default_xattr_handler,
#endif
	NULL,
};


#ifdef CONFIG_JFS_SECURITY
static int jfs_initxattrs(struct inode *inode, const struct xattr *xattr_array,
			  void *fs_info)
{
	const struct xattr *xattr;
	tid_t *tid = fs_info;
	char *name;
	int err = 0;

	for (xattr = xattr_array; xattr->name != NULL; xattr++) {
		name = kmalloc(XATTR_SECURITY_PREFIX_LEN +
			       strlen(xattr->name) + 1, GFP_NOFS);
		if (!name) {
			err = -ENOMEM;
			break;
		}
		strcpy(name, XATTR_SECURITY_PREFIX);
		strcpy(name + XATTR_SECURITY_PREFIX_LEN, xattr->name);

		err = __jfs_setxattr(*tid, inode, name,
				     xattr->value, xattr->value_len, 0);
		kfree(name);
		if (err < 0)
			break;
	}
	return err;
}

int jfs_init_security(tid_t tid, struct inode *inode, struct inode *dir,
		      const struct qstr *qstr)
{
	return security_inode_init_security(inode, dir, qstr,
					    &jfs_initxattrs, &tid);
}
#endif