inode.c 38.5 KB
Newer Older
David Teigland's avatar
David Teigland committed
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
/*
 * Copyright (C) Sistina Software, Inc.  1997-2003 All rights reserved.
 * Copyright (C) 2004-2005 Red Hat, Inc.  All rights reserved.
 *
 * This copyrighted material is made available to anyone wishing to use,
 * modify, copy, or redistribute it subject to the terms and conditions
 * of the GNU General Public License v.2.
 */

#include <linux/sched.h>
#include <linux/slab.h>
#include <linux/spinlock.h>
#include <linux/completion.h>
#include <linux/buffer_head.h>
#include <linux/posix_acl.h>
#include <linux/sort.h>
17
#include <linux/gfs2_ondisk.h>
David Teigland's avatar
David Teigland committed
18 19 20
#include <asm/semaphore.h>

#include "gfs2.h"
21 22
#include "lm_interface.h"
#include "incore.h"
David Teigland's avatar
David Teigland committed
23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38
#include "acl.h"
#include "bmap.h"
#include "dir.h"
#include "eattr.h"
#include "glock.h"
#include "glops.h"
#include "inode.h"
#include "log.h"
#include "meta_io.h"
#include "ops_address.h"
#include "ops_file.h"
#include "ops_inode.h"
#include "quota.h"
#include "rgrp.h"
#include "trans.h"
#include "unlinked.h"
39
#include "util.h"
David Teigland's avatar
David Teigland committed
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

/**
 * inode_attr_in - Copy attributes from the dinode into the VFS inode
 * @ip: The GFS2 inode (with embedded disk inode data)
 * @inode:  The Linux VFS inode
 *
 */

static void inode_attr_in(struct gfs2_inode *ip, struct inode *inode)
{
	inode->i_ino = ip->i_num.no_formal_ino;

	switch (ip->i_di.di_mode & S_IFMT) {
	case S_IFBLK:
	case S_IFCHR:
		inode->i_rdev = MKDEV(ip->i_di.di_major, ip->i_di.di_minor);
		break;
	default:
		inode->i_rdev = 0;
		break;
	};

	inode->i_mode = ip->i_di.di_mode;
	inode->i_nlink = ip->i_di.di_nlink;
	inode->i_uid = ip->i_di.di_uid;
	inode->i_gid = ip->i_di.di_gid;
	i_size_write(inode, ip->i_di.di_size);
	inode->i_atime.tv_sec = ip->i_di.di_atime;
	inode->i_mtime.tv_sec = ip->i_di.di_mtime;
	inode->i_ctime.tv_sec = ip->i_di.di_ctime;
	inode->i_atime.tv_nsec = 0;
	inode->i_mtime.tv_nsec = 0;
	inode->i_ctime.tv_nsec = 0;
	inode->i_blksize = PAGE_SIZE;
	inode->i_blocks = ip->i_di.di_blocks <<
		(ip->i_sbd->sd_sb.sb_bsize_shift - GFS2_BASIC_BLOCK_SHIFT);

	if (ip->i_di.di_flags & GFS2_DIF_IMMUTABLE)
		inode->i_flags |= S_IMMUTABLE;
	else
		inode->i_flags &= ~S_IMMUTABLE;

	if (ip->i_di.di_flags & GFS2_DIF_APPENDONLY)
		inode->i_flags |= S_APPEND;
	else
		inode->i_flags &= ~S_APPEND;
}

/**
 * gfs2_inode_attr_in - Copy attributes from the dinode into the VFS inode
 * @ip: The GFS2 inode (with embedded disk inode data)
 *
 */

void gfs2_inode_attr_in(struct gfs2_inode *ip)
{
	struct inode *inode;

	inode = gfs2_ip2v_lookup(ip);
	if (inode) {
		inode_attr_in(ip, inode);
		iput(inode);
	}
}

/**
 * gfs2_inode_attr_out - Copy attributes from VFS inode into the dinode
 * @ip: The GFS2 inode
 *
 * Only copy out the attributes that we want the VFS layer
 * to be able to modify.
 */

void gfs2_inode_attr_out(struct gfs2_inode *ip)
{
	struct inode *inode = ip->i_vnode;

	gfs2_assert_withdraw(ip->i_sbd,
		(ip->i_di.di_mode & S_IFMT) == (inode->i_mode & S_IFMT));
	ip->i_di.di_mode = inode->i_mode;
	ip->i_di.di_uid = inode->i_uid;
	ip->i_di.di_gid = inode->i_gid;
	ip->i_di.di_atime = inode->i_atime.tv_sec;
	ip->i_di.di_mtime = inode->i_mtime.tv_sec;
	ip->i_di.di_ctime = inode->i_ctime.tv_sec;
}

/**
 * gfs2_ip2v_lookup - Get the struct inode for a struct gfs2_inode
 * @ip: the struct gfs2_inode to get the struct inode for
 *
 * Returns: A VFS inode, or NULL if none
 */

struct inode *gfs2_ip2v_lookup(struct gfs2_inode *ip)
{
	struct inode *inode = NULL;

	gfs2_assert_warn(ip->i_sbd, test_bit(GIF_MIN_INIT, &ip->i_flags));

	spin_lock(&ip->i_spin);
	if (ip->i_vnode)
		inode = igrab(ip->i_vnode);
	spin_unlock(&ip->i_spin);

	return inode;
}

/**
 * gfs2_ip2v - Get/Create a struct inode for a struct gfs2_inode
 * @ip: the struct gfs2_inode to get the struct inode for
 *
 * Returns: A VFS inode, or NULL if no mem
 */

struct inode *gfs2_ip2v(struct gfs2_inode *ip)
{
	struct inode *inode, *tmp;

	inode = gfs2_ip2v_lookup(ip);
	if (inode)
		return inode;

	tmp = new_inode(ip->i_sbd->sd_vfs);
	if (!tmp)
		return NULL;

	inode_attr_in(ip, tmp);

	if (S_ISREG(ip->i_di.di_mode)) {
		tmp->i_op = &gfs2_file_iops;
		tmp->i_fop = &gfs2_file_fops;
		tmp->i_mapping->a_ops = &gfs2_file_aops;
	} else if (S_ISDIR(ip->i_di.di_mode)) {
		tmp->i_op = &gfs2_dir_iops;
		tmp->i_fop = &gfs2_dir_fops;
	} else if (S_ISLNK(ip->i_di.di_mode)) {
		tmp->i_op = &gfs2_symlink_iops;
	} else {
		tmp->i_op = &gfs2_dev_iops;
		init_special_inode(tmp, tmp->i_mode, tmp->i_rdev);
	}

183
	tmp->u.generic_ip = NULL;
David Teigland's avatar
David Teigland committed
184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202

	for (;;) {
		spin_lock(&ip->i_spin);
		if (!ip->i_vnode)
			break;
		inode = igrab(ip->i_vnode);
		spin_unlock(&ip->i_spin);

		if (inode) {
			iput(tmp);
			return inode;
		}
		yield();
	}

	inode = tmp;

	gfs2_inode_hold(ip);
	ip->i_vnode = inode;
203
	inode->u.generic_ip = ip;
David Teigland's avatar
David Teigland committed
204 205 206 207 208 209 210 211 212 213

	spin_unlock(&ip->i_spin);

	insert_inode_hash(inode);

	return inode;
}

static int iget_test(struct inode *inode, void *opaque)
{
214
	struct gfs2_inode *ip = inode->u.generic_ip;
David Teigland's avatar
David Teigland committed
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
	struct gfs2_inum *inum = (struct gfs2_inum *)opaque;

	if (ip && ip->i_num.no_addr == inum->no_addr)
		return 1;

	return 0;
}

struct inode *gfs2_iget(struct super_block *sb, struct gfs2_inum *inum)
{
	return ilookup5(sb, (unsigned long)inum->no_formal_ino,
			iget_test, inum);
}

void gfs2_inode_min_init(struct gfs2_inode *ip, unsigned int type)
{
	spin_lock(&ip->i_spin);
	if (!test_and_set_bit(GIF_MIN_INIT, &ip->i_flags)) {
		ip->i_di.di_nlink = 1;
		ip->i_di.di_mode = DT2IF(type);
	}
	spin_unlock(&ip->i_spin);
}

/**
 * gfs2_inode_refresh - Refresh the incore copy of the dinode
 * @ip: The GFS2 inode
 *
 * Returns: errno
 */

int gfs2_inode_refresh(struct gfs2_inode *ip)
{
	struct buffer_head *dibh;
	int error;

	error = gfs2_meta_inode_buffer(ip, &dibh);
	if (error)
		return error;

	if (gfs2_metatype_check(ip->i_sbd, dibh, GFS2_METATYPE_DI)) {
		brelse(dibh);
		return -EIO;
	}

	spin_lock(&ip->i_spin);
	gfs2_dinode_in(&ip->i_di, dibh->b_data);
	set_bit(GIF_MIN_INIT, &ip->i_flags);
	spin_unlock(&ip->i_spin);

	brelse(dibh);

	if (ip->i_num.no_addr != ip->i_di.di_num.no_addr) {
		if (gfs2_consist_inode(ip))
			gfs2_dinode_print(&ip->i_di);
		return -EIO;
	}
	if (ip->i_num.no_formal_ino != ip->i_di.di_num.no_formal_ino)
		return -ESTALE;

	ip->i_vn = ip->i_gl->gl_vn;

	return 0;
}

/**
 * inode_create - create a struct gfs2_inode
 * @i_gl: The glock covering the inode
 * @inum: The inode number
 * @io_gl: the iopen glock to acquire/hold (using holder in new gfs2_inode)
 * @io_state: the state the iopen glock should be acquired in
 * @ipp: pointer to put the returned inode in
 *
 * Returns: errno
 */

static int inode_create(struct gfs2_glock *i_gl, struct gfs2_inum *inum,
			struct gfs2_glock *io_gl, unsigned int io_state,
			struct gfs2_inode **ipp)
{
	struct gfs2_sbd *sdp = i_gl->gl_sbd;
	struct gfs2_inode *ip;
	int error = 0;

	ip = kmem_cache_alloc(gfs2_inode_cachep, GFP_KERNEL);
	if (!ip)
		return -ENOMEM;
	memset(ip, 0, sizeof(struct gfs2_inode));

	ip->i_num = *inum;

	atomic_set(&ip->i_count, 1);

	ip->i_vn = i_gl->gl_vn - 1;

	ip->i_gl = i_gl;
	ip->i_sbd = sdp;

	spin_lock_init(&ip->i_spin);
	init_rwsem(&ip->i_rw_mutex);

	ip->i_greedy = gfs2_tune_get(sdp, gt_greedy_default);

	error = gfs2_glock_nq_init(io_gl,
				   io_state, GL_LOCAL_EXCL | GL_EXACT,
				   &ip->i_iopen_gh);
	if (error)
		goto fail;
	ip->i_iopen_gh.gh_owner = NULL;

	spin_lock(&io_gl->gl_spin);
	gfs2_glock_hold(i_gl);
327
	io_gl->gl_object = i_gl;
David Teigland's avatar
David Teigland committed
328 329 330
	spin_unlock(&io_gl->gl_spin);

	gfs2_glock_hold(i_gl);
331
	i_gl->gl_object = ip;
David Teigland's avatar
David Teigland committed
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

	atomic_inc(&sdp->sd_inode_count);

	*ipp = ip;

	return 0;

 fail:
	gfs2_meta_cache_flush(ip);
	kmem_cache_free(gfs2_inode_cachep, ip);
	*ipp = NULL;

	return error;
}

/**
 * gfs2_inode_get - Create or get a reference on an inode
 * @i_gl: The glock covering the inode
 * @inum: The inode number
 * @create:
 * @ipp: pointer to put the returned inode in
 *
 * Returns: errno
 */

int gfs2_inode_get(struct gfs2_glock *i_gl, struct gfs2_inum *inum, int create,
		   struct gfs2_inode **ipp)
{
	struct gfs2_sbd *sdp = i_gl->gl_sbd;
	struct gfs2_glock *io_gl;
	int error = 0;

	gfs2_glmutex_lock(i_gl);

366
	*ipp = i_gl->gl_object;
David Teigland's avatar
David Teigland committed
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
	if (*ipp) {
		error = -ESTALE;
		if ((*ipp)->i_num.no_formal_ino != inum->no_formal_ino)
			goto out;
		atomic_inc(&(*ipp)->i_count);
		error = 0;
		goto out;
	}

	if (!create)
		goto out;

	error = gfs2_glock_get(sdp, inum->no_addr, &gfs2_iopen_glops,
			       CREATE, &io_gl);
	if (!error) {
		error = inode_create(i_gl, inum, io_gl, LM_ST_SHARED, ipp);
		gfs2_glock_put(io_gl);
	}

 out:
	gfs2_glmutex_unlock(i_gl);

	return error;
}

void gfs2_inode_hold(struct gfs2_inode *ip)
{
	gfs2_assert(ip->i_sbd, atomic_read(&ip->i_count) > 0);
	atomic_inc(&ip->i_count);
}

void gfs2_inode_put(struct gfs2_inode *ip)
{
	gfs2_assert(ip->i_sbd, atomic_read(&ip->i_count) > 0);
	atomic_dec(&ip->i_count);
}

void gfs2_inode_destroy(struct gfs2_inode *ip)
{
	struct gfs2_sbd *sdp = ip->i_sbd;
	struct gfs2_glock *io_gl = ip->i_iopen_gh.gh_gl;
	struct gfs2_glock *i_gl = ip->i_gl;

	gfs2_assert_warn(sdp, !atomic_read(&ip->i_count));
411
	gfs2_assert(sdp, io_gl->gl_object == i_gl);
David Teigland's avatar
David Teigland committed
412 413

	spin_lock(&io_gl->gl_spin);
414
	io_gl->gl_object = NULL;
David Teigland's avatar
David Teigland committed
415 416 417 418 419 420 421 422
	gfs2_glock_put(i_gl);
	spin_unlock(&io_gl->gl_spin);

	gfs2_glock_dq_uninit(&ip->i_iopen_gh);

	gfs2_meta_cache_flush(ip);
	kmem_cache_free(gfs2_inode_cachep, ip);

423
	i_gl->gl_object = NULL;
David Teigland's avatar
David Teigland committed
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
	gfs2_glock_put(i_gl);

	atomic_dec(&sdp->sd_inode_count);
}

static int dinode_dealloc(struct gfs2_inode *ip, struct gfs2_unlinked *ul)
{
	struct gfs2_sbd *sdp = ip->i_sbd;
	struct gfs2_alloc *al;
	struct gfs2_rgrpd *rgd;
	int error;

	if (ip->i_di.di_blocks != 1) {
		if (gfs2_consist_inode(ip))
			gfs2_dinode_print(&ip->i_di);
		return -EIO;
	}

	al = gfs2_alloc_get(ip);

	error = gfs2_quota_hold(ip, NO_QUOTA_CHANGE, NO_QUOTA_CHANGE);
	if (error)
		goto out;

	error = gfs2_rindex_hold(sdp, &al->al_ri_gh);
	if (error)
		goto out_qs;

	rgd = gfs2_blk2rgrpd(sdp, ip->i_num.no_addr);
	if (!rgd) {
		gfs2_consist_inode(ip);
		error = -EIO;
		goto out_rindex_relse;
	}

	error = gfs2_glock_nq_init(rgd->rd_gl, LM_ST_EXCLUSIVE, 0,
				   &al->al_rgd_gh);
	if (error)
		goto out_rindex_relse;

	error = gfs2_trans_begin(sdp, RES_RG_BIT + RES_UNLINKED +
				 RES_STATFS + RES_QUOTA, 1);
	if (error)
		goto out_rg_gunlock;

	gfs2_trans_add_gl(ip->i_gl);

	gfs2_free_di(rgd, ip);

	error = gfs2_unlinked_ondisk_rm(sdp, ul);

	gfs2_trans_end(sdp);
	clear_bit(GLF_STICKY, &ip->i_gl->gl_flags);

 out_rg_gunlock:
	gfs2_glock_dq_uninit(&al->al_rgd_gh);

 out_rindex_relse:
	gfs2_glock_dq_uninit(&al->al_ri_gh);

 out_qs:
	gfs2_quota_unhold(ip);

 out:
	gfs2_alloc_put(ip);

	return error;
}

/**
 * inode_dealloc - Deallocate all on-disk blocks for an inode (dinode)
 * @sdp: the filesystem
 * @inum: the inode number to deallocate
 * @io_gh: a holder for the iopen glock for this inode
 *
 * Returns: errno
 */

static int inode_dealloc(struct gfs2_sbd *sdp, struct gfs2_unlinked *ul,
			 struct gfs2_holder *io_gh)
{
	struct gfs2_inode *ip;
	struct gfs2_holder i_gh;
	int error;

	error = gfs2_glock_nq_num(sdp,
				  ul->ul_ut.ut_inum.no_addr, &gfs2_inode_glops,
				  LM_ST_EXCLUSIVE, 0, &i_gh);
	if (error)
		return error;

	/* We reacquire the iopen lock here to avoid a race with the NFS server
	   calling gfs2_read_inode() with the inode number of a inode we're in
	   the process of deallocating.  And we can't keep our hold on the lock
	   from inode_dealloc_init() for deadlock reasons. */

	gfs2_holder_reinit(LM_ST_EXCLUSIVE, LM_FLAG_TRY, io_gh);
	error = gfs2_glock_nq(io_gh);
	switch (error) {
	case 0:
		break;
	case GLR_TRYFAILED:
		error = 1;
	default:
		goto out;
	}

531
	gfs2_assert_warn(sdp, !i_gh.gh_gl->gl_object);
David Teigland's avatar
David Teigland committed
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
	error = inode_create(i_gh.gh_gl, &ul->ul_ut.ut_inum, io_gh->gh_gl,
			     LM_ST_EXCLUSIVE, &ip);

	gfs2_glock_dq(io_gh);

	if (error)
		goto out;

	error = gfs2_inode_refresh(ip);
	if (error)
		goto out_iput;

	if (ip->i_di.di_nlink) {
		if (gfs2_consist_inode(ip))
			gfs2_dinode_print(&ip->i_di);
		error = -EIO;
		goto out_iput;
	}

	if (S_ISDIR(ip->i_di.di_mode) &&
	    (ip->i_di.di_flags & GFS2_DIF_EXHASH)) {
		error = gfs2_dir_exhash_dealloc(ip);
		if (error)
			goto out_iput;
	}

	if (ip->i_di.di_eattr) {
		error = gfs2_ea_dealloc(ip);
		if (error)
			goto out_iput;
	}

	if (!gfs2_is_stuffed(ip)) {
		error = gfs2_file_dealloc(ip);
		if (error)
			goto out_iput;
	}

	error = dinode_dealloc(ip, ul);
	if (error)
		goto out_iput;

 out_iput:
	gfs2_glmutex_lock(i_gh.gh_gl);
	gfs2_inode_put(ip);
	gfs2_inode_destroy(ip);
	gfs2_glmutex_unlock(i_gh.gh_gl);

 out:
	gfs2_glock_dq_uninit(&i_gh);

	return error;
}

/**
 * try_inode_dealloc - Try to deallocate an inode and all its blocks
 * @sdp: the filesystem
 *
 * Returns: 0 on success, -errno on error, 1 on busy (inode open)
 */

static int try_inode_dealloc(struct gfs2_sbd *sdp, struct gfs2_unlinked *ul)
{
	struct gfs2_holder io_gh;
	int error = 0;

	gfs2_try_toss_inode(sdp, &ul->ul_ut.ut_inum);

	error = gfs2_glock_nq_num(sdp,
				  ul->ul_ut.ut_inum.no_addr, &gfs2_iopen_glops,
				  LM_ST_EXCLUSIVE, LM_FLAG_TRY_1CB, &io_gh);
	switch (error) {
	case 0:
		break;
	case GLR_TRYFAILED:
		return 1;
	default:
		return error;
	}

	gfs2_glock_dq(&io_gh);
	error = inode_dealloc(sdp, ul, &io_gh);
	gfs2_holder_uninit(&io_gh);

	return error;
}

static int inode_dealloc_uninit(struct gfs2_sbd *sdp, struct gfs2_unlinked *ul)
{
	struct gfs2_rgrpd *rgd;
	struct gfs2_holder ri_gh, rgd_gh;
	int error;

	error = gfs2_rindex_hold(sdp, &ri_gh);
	if (error)
		return error;

	rgd = gfs2_blk2rgrpd(sdp, ul->ul_ut.ut_inum.no_addr);
	if (!rgd) {
		gfs2_consist(sdp);
		error = -EIO;
		goto out;
	}

	error = gfs2_glock_nq_init(rgd->rd_gl, LM_ST_EXCLUSIVE, 0, &rgd_gh);
	if (error)
		goto out;

	error = gfs2_trans_begin(sdp,
				 RES_RG_BIT + RES_UNLINKED + RES_STATFS,
				 0);
	if (error)
		goto out_gunlock;

	gfs2_free_uninit_di(rgd, ul->ul_ut.ut_inum.no_addr);
	gfs2_unlinked_ondisk_rm(sdp, ul);

	gfs2_trans_end(sdp);

 out_gunlock:
	gfs2_glock_dq_uninit(&rgd_gh);
 out:
	gfs2_glock_dq_uninit(&ri_gh);

	return error;
}

int gfs2_inode_dealloc(struct gfs2_sbd *sdp, struct gfs2_unlinked *ul)
{
	if (ul->ul_ut.ut_flags & GFS2_UTF_UNINIT)
		return inode_dealloc_uninit(sdp, ul);
	else
		return try_inode_dealloc(sdp, ul);
}

/**
 * gfs2_change_nlink - Change nlink count on inode
 * @ip: The GFS2 inode
 * @diff: The change in the nlink count required
 *
 * Returns: errno
 */

int gfs2_change_nlink(struct gfs2_inode *ip, int diff)
{
	struct buffer_head *dibh;
	uint32_t nlink;
	int error;

	nlink = ip->i_di.di_nlink + diff;

	/* If we are reducing the nlink count, but the new value ends up being
	   bigger than the old one, we must have underflowed. */
	if (diff < 0 && nlink > ip->i_di.di_nlink) {
		if (gfs2_consist_inode(ip))
			gfs2_dinode_print(&ip->i_di);
		return -EIO;
	}

	error = gfs2_meta_inode_buffer(ip, &dibh);
	if (error)
		return error;

	ip->i_di.di_nlink = nlink;
	ip->i_di.di_ctime = get_seconds();

698
	gfs2_trans_add_bh(ip->i_gl, dibh, 1);
David Teigland's avatar
David Teigland committed
699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717
	gfs2_dinode_out(&ip->i_di, dibh->b_data);
	brelse(dibh);

	return 0;
}

/**
 * gfs2_lookupi - Look up a filename in a directory and return its inode
 * @d_gh: An initialized holder for the directory glock
 * @name: The name of the inode to look for
 * @is_root: If 1, ignore the caller's permissions
 * @i_gh: An uninitialized holder for the new inode glock
 *
 * There will always be a vnode (Linux VFS inode) for the d_gh inode unless
 * @is_root is true.
 *
 * Returns: errno
 */

718 719
int gfs2_lookupi(struct inode *dir, struct qstr *name, int is_root,
		 struct inode **inodep)
David Teigland's avatar
David Teigland committed
720
{
721
	struct gfs2_inode *ipp;
722
	struct gfs2_inode *dip = dir->u.generic_ip;
David Teigland's avatar
David Teigland committed
723 724 725 726 727
	struct gfs2_sbd *sdp = dip->i_sbd;
	struct gfs2_holder d_gh;
	struct gfs2_inum inum;
	unsigned int type;
	struct gfs2_glock *gl;
728 729 730
	int error = 0;

	*inodep = NULL;
David Teigland's avatar
David Teigland committed
731 732 733 734 735

	if (!name->len || name->len > GFS2_FNAMESIZE)
		return -ENAMETOOLONG;

	if (gfs2_filecmp(name, ".", 1) ||
736
	    (gfs2_filecmp(name, "..", 2) && dir == sdp->sd_root_dir)) {
David Teigland's avatar
David Teigland committed
737
		gfs2_inode_hold(dip);
738 739
		ipp = dip;
		goto done;
David Teigland's avatar
David Teigland committed
740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760
	}

	error = gfs2_glock_nq_init(dip->i_gl, LM_ST_SHARED, 0, &d_gh);
	if (error)
		return error;

	if (!is_root) {
		error = gfs2_repermission(dip->i_vnode, MAY_EXEC, NULL);
		if (error)
			goto out;
	}

	error = gfs2_dir_search(dip, name, &inum, &type);
	if (error)
		goto out;

	error = gfs2_glock_get(sdp, inum.no_addr, &gfs2_inode_glops,
			       CREATE, &gl);
	if (error)
		goto out;

761
	error = gfs2_inode_get(gl, &inum, CREATE, &ipp);
David Teigland's avatar
David Teigland committed
762
	if (!error)
763
		gfs2_inode_min_init(ipp, type);
David Teigland's avatar
David Teigland committed
764 765 766

	gfs2_glock_put(gl);

767
out:
David Teigland's avatar
David Teigland committed
768
	gfs2_glock_dq_uninit(&d_gh);
769 770 771 772 773 774 775
done:
	if (error == 0) {
		*inodep = gfs2_ip2v(ipp);
		if (!*inodep)
			error = -ENOMEM;
		gfs2_inode_put(ipp);
	}
David Teigland's avatar
David Teigland committed
776 777 778 779 780
	return error;
}

static int pick_formal_ino_1(struct gfs2_sbd *sdp, uint64_t *formal_ino)
{
781
	struct gfs2_inode *ip = sdp->sd_ir_inode->u.generic_ip;
David Teigland's avatar
David Teigland committed
782 783 784 785 786 787 788
	struct buffer_head *bh;
	struct gfs2_inum_range ir;
	int error;

	error = gfs2_trans_begin(sdp, RES_DINODE, 0);
	if (error)
		return error;
789
	mutex_lock(&sdp->sd_inum_mutex);
David Teigland's avatar
David Teigland committed
790 791 792

	error = gfs2_meta_inode_buffer(ip, &bh);
	if (error) {
793
		mutex_unlock(&sdp->sd_inum_mutex);
David Teigland's avatar
David Teigland committed
794 795 796 797 798 799 800 801 802
		gfs2_trans_end(sdp);
		return error;
	}

	gfs2_inum_range_in(&ir, bh->b_data + sizeof(struct gfs2_dinode));

	if (ir.ir_length) {
		*formal_ino = ir.ir_start++;
		ir.ir_length--;
803
		gfs2_trans_add_bh(ip->i_gl, bh, 1);
David Teigland's avatar
David Teigland committed
804 805 806
		gfs2_inum_range_out(&ir,
				    bh->b_data + sizeof(struct gfs2_dinode));
		brelse(bh);
807
		mutex_unlock(&sdp->sd_inum_mutex);
David Teigland's avatar
David Teigland committed
808 809 810 811 812 813
		gfs2_trans_end(sdp);
		return 0;
	}

	brelse(bh);

814
	mutex_unlock(&sdp->sd_inum_mutex);
David Teigland's avatar
David Teigland committed
815 816 817 818 819 820 821
	gfs2_trans_end(sdp);

	return 1;
}

static int pick_formal_ino_2(struct gfs2_sbd *sdp, uint64_t *formal_ino)
{
822 823
	struct gfs2_inode *ip = sdp->sd_ir_inode->u.generic_ip;
	struct gfs2_inode *m_ip = sdp->sd_inum_inode->u.generic_ip;
David Teigland's avatar
David Teigland committed
824 825 826 827 828 829 830 831 832 833 834 835
	struct gfs2_holder gh;
	struct buffer_head *bh;
	struct gfs2_inum_range ir;
	int error;

	error = gfs2_glock_nq_init(m_ip->i_gl, LM_ST_EXCLUSIVE, 0, &gh);
	if (error)
		return error;

	error = gfs2_trans_begin(sdp, 2 * RES_DINODE, 0);
	if (error)
		goto out;
836
	mutex_lock(&sdp->sd_inum_mutex);
David Teigland's avatar
David Teigland committed
837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859

	error = gfs2_meta_inode_buffer(ip, &bh);
	if (error)
		goto out_end_trans;
	
	gfs2_inum_range_in(&ir, bh->b_data + sizeof(struct gfs2_dinode));

	if (!ir.ir_length) {
		struct buffer_head *m_bh;
		uint64_t x, y;

		error = gfs2_meta_inode_buffer(m_ip, &m_bh);
		if (error)
			goto out_brelse;

		x = *(uint64_t *)(m_bh->b_data + sizeof(struct gfs2_dinode));
		x = y = be64_to_cpu(x);
		ir.ir_start = x;
		ir.ir_length = GFS2_INUM_QUANTUM;
		x += GFS2_INUM_QUANTUM;
		if (x < y)
			gfs2_consist_inode(m_ip);
		x = cpu_to_be64(x);
860
		gfs2_trans_add_bh(m_ip->i_gl, m_bh, 1);
David Teigland's avatar
David Teigland committed
861 862 863 864 865 866 867 868
		*(uint64_t *)(m_bh->b_data + sizeof(struct gfs2_dinode)) = x;

		brelse(m_bh);
	}

	*formal_ino = ir.ir_start++;
	ir.ir_length--;

869
	gfs2_trans_add_bh(ip->i_gl, bh, 1);
David Teigland's avatar
David Teigland committed
870 871 872 873 874 875
	gfs2_inum_range_out(&ir, bh->b_data + sizeof(struct gfs2_dinode));

 out_brelse:
	brelse(bh);

 out_end_trans:
876
	mutex_unlock(&sdp->sd_inum_mutex);
David Teigland's avatar
David Teigland committed
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
	gfs2_trans_end(sdp);

 out:
	gfs2_glock_dq_uninit(&gh);

	return error;
}

static int pick_formal_ino(struct gfs2_sbd *sdp, uint64_t *inum)
{
	int error;

	error = pick_formal_ino_1(sdp, inum);
	if (error <= 0)
		return error;

	error = pick_formal_ino_2(sdp, inum);

	return error;
}

/**
 * create_ok - OK to create a new on-disk inode here?
 * @dip:  Directory in which dinode is to be created
 * @name:  Name of new dinode
 * @mode:
 *
 * Returns: errno
 */

static int create_ok(struct gfs2_inode *dip, struct qstr *name,
		     unsigned int mode)
{
	int error;

	error = gfs2_repermission(dip->i_vnode, MAY_WRITE | MAY_EXEC, NULL);
	if (error)
		return error;

	/*  Don't create entries in an unlinked directory  */
	if (!dip->i_di.di_nlink)
		return -EPERM;

	error = gfs2_dir_search(dip, name, NULL, NULL);
	switch (error) {
	case -ENOENT:
		error = 0;
		break;
	case 0:
		return -EEXIST;
	default:
		return error;
	}

	if (dip->i_di.di_entries == (uint32_t)-1)
		return -EFBIG;
	if (S_ISDIR(mode) && dip->i_di.di_nlink == (uint32_t)-1)
		return -EMLINK;

	return 0;
}

static void munge_mode_uid_gid(struct gfs2_inode *dip, unsigned int *mode,
			       unsigned int *uid, unsigned int *gid)
{
	if (dip->i_sbd->sd_args.ar_suiddir &&
	    (dip->i_di.di_mode & S_ISUID) &&
	    dip->i_di.di_uid) {
		if (S_ISDIR(*mode))
			*mode |= S_ISUID;
		else if (dip->i_di.di_uid != current->fsuid)
			*mode &= ~07111;
		*uid = dip->i_di.di_uid;
	} else
		*uid = current->fsuid;

	if (dip->i_di.di_mode & S_ISGID) {
		if (S_ISDIR(*mode))
			*mode |= S_ISGID;
		*gid = dip->i_di.di_gid;
	} else
		*gid = current->fsgid;
}

static int alloc_dinode(struct gfs2_inode *dip, struct gfs2_unlinked *ul)
{
	struct gfs2_sbd *sdp = dip->i_sbd;
	int error;

	gfs2_alloc_get(dip);

	dip->i_alloc.al_requested = RES_DINODE;
	error = gfs2_inplace_reserve(dip);
	if (error)
		goto out;

	error = gfs2_trans_begin(sdp, RES_RG_BIT + RES_UNLINKED +
				 RES_STATFS, 0);
	if (error)
		goto out_ipreserv;

	ul->ul_ut.ut_inum.no_addr = gfs2_alloc_di(dip);

	ul->ul_ut.ut_flags = GFS2_UTF_UNINIT;
	error = gfs2_unlinked_ondisk_add(sdp, ul);

	gfs2_trans_end(sdp);

 out_ipreserv:
	gfs2_inplace_release(dip);

 out:
	gfs2_alloc_put(dip);

	return error;
}

/**
 * init_dinode - Fill in a new dinode structure
 * @dip: the directory this inode is being created in
 * @gl: The glock covering the new inode
 * @inum: the inode number
 * @mode: the file permissions
 * @uid:
 * @gid:
 *
 */

static void init_dinode(struct gfs2_inode *dip, struct gfs2_glock *gl,
			struct gfs2_inum *inum, unsigned int mode,
			unsigned int uid, unsigned int gid)
{
	struct gfs2_sbd *sdp = dip->i_sbd;
1010
	struct gfs2_dinode *di;
David Teigland's avatar
David Teigland committed
1011 1012 1013
	struct buffer_head *dibh;

	dibh = gfs2_meta_new(gl, inum->no_addr);
1014
	gfs2_trans_add_bh(gl, dibh, 1);
David Teigland's avatar
David Teigland committed
1015 1016
	gfs2_metatype_set(dibh, GFS2_METATYPE_DI, GFS2_FORMAT_DI);
	gfs2_buffer_clear_tail(dibh, sizeof(struct gfs2_dinode));
1017 1018
	di = (struct gfs2_dinode *)dibh->b_data;

1019 1020
	di->di_num.no_formal_ino = cpu_to_be64(inum->no_formal_ino);
	di->di_num.no_addr = cpu_to_be64(inum->no_addr);
1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031
	di->di_mode = cpu_to_be32(mode);
	di->di_uid = cpu_to_be32(uid);
	di->di_gid = cpu_to_be32(gid);
	di->di_nlink = cpu_to_be32(0);
	di->di_size = cpu_to_be64(0);
	di->di_blocks = cpu_to_be64(1);
	di->di_atime = di->di_mtime = di->di_ctime = cpu_to_be64(get_seconds());
	di->di_major = di->di_minor = cpu_to_be32(0);
	di->di_goal_meta = di->di_goal_data = cpu_to_be64(inum->no_addr);
	di->__pad[0] = di->__pad[1] = 0;
	di->di_flags = cpu_to_be32(0);
David Teigland's avatar
David Teigland committed
1032 1033 1034 1035

	if (S_ISREG(mode)) {
		if ((dip->i_di.di_flags & GFS2_DIF_INHERIT_JDATA) ||
		    gfs2_tune_get(sdp, gt_new_files_jdata))
1036
			di->di_flags |= cpu_to_be32(GFS2_DIF_JDATA);
David Teigland's avatar
David Teigland committed
1037 1038
		if ((dip->i_di.di_flags & GFS2_DIF_INHERIT_DIRECTIO) ||
		    gfs2_tune_get(sdp, gt_new_files_directio))
1039
			di->di_flags |= cpu_to_be32(GFS2_DIF_DIRECTIO);
David Teigland's avatar
David Teigland committed
1040
	} else if (S_ISDIR(mode)) {
1041 1042 1043 1044
		di->di_flags |= cpu_to_be32(dip->i_di.di_flags &
					    GFS2_DIF_INHERIT_DIRECTIO);
		di->di_flags |= cpu_to_be32(dip->i_di.di_flags &
					    GFS2_DIF_INHERIT_JDATA);
David Teigland's avatar
David Teigland committed
1045 1046
	}

1047 1048 1049 1050 1051 1052 1053 1054 1055 1056
	di->__pad1 = 0;
	di->di_height = cpu_to_be32(0);
	di->__pad2 = 0;
	di->__pad3 = 0;
	di->di_depth = cpu_to_be16(0);
	di->di_entries = cpu_to_be32(0);
	memset(&di->__pad4, 0, sizeof(di->__pad4));
	di->di_eattr = cpu_to_be64(0);
	memset(&di->di_reserved, 0, sizeof(di->di_reserved));

David Teigland's avatar
David Teigland committed
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 1107 1108 1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151 1152 1153 1154
	brelse(dibh);
}

static int make_dinode(struct gfs2_inode *dip, struct gfs2_glock *gl,
		       unsigned int mode, struct gfs2_unlinked *ul)
{
	struct gfs2_sbd *sdp = dip->i_sbd;
	unsigned int uid, gid;
	int error;

	munge_mode_uid_gid(dip, &mode, &uid, &gid);

	gfs2_alloc_get(dip);

	error = gfs2_quota_lock(dip, uid, gid);
	if (error)
		goto out;

	error = gfs2_quota_check(dip, uid, gid);
	if (error)
		goto out_quota;

	error = gfs2_trans_begin(sdp, RES_DINODE + RES_UNLINKED +
				 RES_QUOTA, 0);
	if (error)
		goto out_quota;

	ul->ul_ut.ut_flags = 0;
	error = gfs2_unlinked_ondisk_munge(sdp, ul);

	init_dinode(dip, gl, &ul->ul_ut.ut_inum,
		     mode, uid, gid);

	gfs2_quota_change(dip, +1, uid, gid);

	gfs2_trans_end(sdp);

 out_quota:
	gfs2_quota_unlock(dip);

 out:
	gfs2_alloc_put(dip);

	return error;
}

static int link_dinode(struct gfs2_inode *dip, struct qstr *name,
		       struct gfs2_inode *ip, struct gfs2_unlinked *ul)
{
	struct gfs2_sbd *sdp = dip->i_sbd;
	struct gfs2_alloc *al;
	int alloc_required;
	struct buffer_head *dibh;
	int error;

	al = gfs2_alloc_get(dip);

	error = gfs2_quota_lock(dip, NO_QUOTA_CHANGE, NO_QUOTA_CHANGE);
	if (error)
		goto fail;

	error = gfs2_diradd_alloc_required(dip, name, &alloc_required);
	if (alloc_required) {
		error = gfs2_quota_check(dip, dip->i_di.di_uid,
					 dip->i_di.di_gid);
		if (error)
			goto fail_quota_locks;

		al->al_requested = sdp->sd_max_dirres;

		error = gfs2_inplace_reserve(dip);
		if (error)
			goto fail_quota_locks;

		error = gfs2_trans_begin(sdp,
					 sdp->sd_max_dirres +
					 al->al_rgd->rd_ri.ri_length +
					 2 * RES_DINODE + RES_UNLINKED +
					 RES_STATFS + RES_QUOTA, 0);
		if (error)
			goto fail_ipreserv;
	} else {
		error = gfs2_trans_begin(sdp,
					 RES_LEAF +
					 2 * RES_DINODE +
					 RES_UNLINKED, 0);
		if (error)
			goto fail_quota_locks;
	}

	error = gfs2_dir_add(dip, name, &ip->i_num, IF2DT(ip->i_di.di_mode));
	if (error)
		goto fail_end_trans;

	error = gfs2_meta_inode_buffer(ip, &dibh);
	if (error)
		goto fail_end_trans;
	ip->i_di.di_nlink = 1;
1155
	gfs2_trans_add_bh(ip->i_gl, dibh, 1);
David Teigland's avatar
David Teigland committed
1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189
	gfs2_dinode_out(&ip->i_di, dibh->b_data);
	brelse(dibh);

	error = gfs2_unlinked_ondisk_rm(sdp, ul);
	if (error)
		goto fail_end_trans;

	return 0;

 fail_end_trans:
	gfs2_trans_end(sdp);

 fail_ipreserv:
	if (dip->i_alloc.al_rgd)
		gfs2_inplace_release(dip);

 fail_quota_locks:
	gfs2_quota_unlock(dip);

 fail:
	gfs2_alloc_put(dip);

	return error;
}

/**
 * gfs2_createi - Create a new inode
 * @ghs: An array of two holders
 * @name: The name of the new file
 * @mode: the permissions on the new inode
 *
 * @ghs[0] is an initialized holder for the directory
 * @ghs[1] is the holder for the inode lock
 *
1190
 * If the return value is not NULL, the glocks on both the directory and the new
David Teigland's avatar
David Teigland committed
1191 1192 1193
 * file are held.  A transaction has been started and an inplace reservation
 * is held, as well.
 *
1194
 * Returns: An inode
David Teigland's avatar
David Teigland committed
1195 1196
 */

1197 1198
struct inode *gfs2_createi(struct gfs2_holder *ghs, struct qstr *name,
			   unsigned int mode)
David Teigland's avatar
David Teigland committed
1199
{
1200
	struct inode *inode;
1201
	struct gfs2_inode *dip = ghs->gh_gl->gl_object;
David Teigland's avatar
David Teigland committed
1202 1203 1204 1205 1206 1207
	struct gfs2_sbd *sdp = dip->i_sbd;
	struct gfs2_unlinked *ul;
	struct gfs2_inode *ip;
	int error;

	if (!name->len || name->len > GFS2_FNAMESIZE)
1208
		return ERR_PTR(-ENAMETOOLONG);
David Teigland's avatar
David Teigland committed
1209 1210 1211

	error = gfs2_unlinked_get(sdp, &ul);
	if (error)
1212
		return ERR_PTR(error);
David Teigland's avatar
David Teigland committed
1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 1230 1231 1232 1233 1234 1235 1236 1237 1238 1239 1240

	gfs2_holder_reinit(LM_ST_EXCLUSIVE, 0, ghs);
	error = gfs2_glock_nq(ghs);
	if (error)
		goto fail;

	error = create_ok(dip, name, mode);
	if (error)
		goto fail_gunlock;

	error = pick_formal_ino(sdp, &ul->ul_ut.ut_inum.no_formal_ino);
	if (error)
		goto fail_gunlock;

	error = alloc_dinode(dip, ul);
	if (error)
		goto fail_gunlock;

	if (ul->ul_ut.ut_inum.no_addr < dip->i_num.no_addr) {
		gfs2_glock_dq(ghs);

		error = gfs2_glock_nq_num(sdp,
					  ul->ul_ut.ut_inum.no_addr,
					  &gfs2_inode_glops,
					  LM_ST_EXCLUSIVE, GL_SKIP,
					  ghs + 1);
		if (error) {
			gfs2_unlinked_put(sdp, ul);
1241
			return ERR_PTR(error);
David Teigland's avatar
David Teigland committed
1242 1243 1244 1245 1246 1247 1248
		}

		gfs2_holder_reinit(LM_ST_EXCLUSIVE, 0, ghs);
		error = gfs2_glock_nq(ghs);
		if (error) {
			gfs2_glock_dq_uninit(ghs + 1);
			gfs2_unlinked_put(sdp, ul);
1249
			return ERR_PTR(error);
David Teigland's avatar
David Teigland committed
1250 1251 1252 1253 1254 1255 1256 1257 1258 1259 1260 1261 1262 1263 1264 1265 1266 1267 1268 1269 1270 1271 1272 1273 1274 1275 1276 1277 1278 1279 1280 1281 1282 1283 1284 1285 1286
		}

		error = create_ok(dip, name, mode);
		if (error)
			goto fail_gunlock2;
	} else {
		error = gfs2_glock_nq_num(sdp,
					  ul->ul_ut.ut_inum.no_addr,
					  &gfs2_inode_glops,
					  LM_ST_EXCLUSIVE, GL_SKIP,
					  ghs + 1);
		if (error)
			goto fail_gunlock;
	}

	error = make_dinode(dip, ghs[1].gh_gl, mode, ul);
	if (error)
		goto fail_gunlock2;

	error = gfs2_inode_get(ghs[1].gh_gl, &ul->ul_ut.ut_inum, CREATE, &ip);
	if (error)
		goto fail_gunlock2;

	error = gfs2_inode_refresh(ip);
	if (error)
		goto fail_iput;

	error = gfs2_acl_create(dip, ip);
	if (error)
		goto fail_iput;

	error = link_dinode(dip, name, ip, ul);
	if (error)
		goto fail_iput;

	gfs2_unlinked_put(sdp, ul);

1287 1288 1289 1290 1291
	inode = gfs2_ip2v(ip);
	gfs2_inode_put(ip);
	if (!inode)
		return ERR_PTR(-ENOMEM);
	return inode;
David Teigland's avatar
David Teigland committed
1292 1293 1294 1295 1296 1297 1298 1299 1300 1301 1302 1303 1304

 fail_iput:
	gfs2_inode_put(ip);

 fail_gunlock2:
	gfs2_glock_dq_uninit(ghs + 1);

 fail_gunlock:
	gfs2_glock_dq(ghs);

 fail:
	gfs2_unlinked_put(sdp, ul);

1305
	return ERR_PTR(error);
David Teigland's avatar
David Teigland committed
1306 1307 1308 1309 1310 1311 1312 1313 1314 1315 1316 1317 1318 1319 1320 1321 1322 1323 1324 1325 1326 1327 1328 1329 1330 1331 1332 1333 1334 1335 1336 1337 1338 1339 1340 1341 1342 1343 1344 1345 1346 1347 1348 1349 1350 1351 1352 1353 1354 1355 1356 1357 1358 1359 1360 1361 1362 1363 1364 1365 1366 1367 1368 1369 1370 1371 1372 1373 1374 1375 1376 1377 1378 1379 1380 1381 1382 1383 1384 1385 1386 1387 1388 1389 1390 1391 1392 1393 1394 1395 1396 1397 1398 1399 1400 1401 1402 1403 1404 1405 1406 1407 1408 1409 1410 1411 1412 1413 1414 1415 1416 1417 1418 1419 1420 1421 1422 1423 1424 1425 1426 1427 1428 1429 1430 1431 1432 1433 1434 1435 1436 1437 1438 1439 1440 1441 1442 1443 1444 1445 1446 1447 1448 1449 1450 1451 1452 1453 1454 1455 1456 1457 1458 1459 1460 1461 1462 1463 1464 1465 1466 1467 1468 1469
}

/**
 * gfs2_unlinki - Unlink a file
 * @dip: The inode of the directory
 * @name: The name of the file to be unlinked
 * @ip: The inode of the file to be removed
 *
 * Assumes Glocks on both dip and ip are held.
 *
 * Returns: errno
 */

int gfs2_unlinki(struct gfs2_inode *dip, struct qstr *name,
		 struct gfs2_inode *ip, struct gfs2_unlinked *ul)
{
	struct gfs2_sbd *sdp = dip->i_sbd;
	int error;

	error = gfs2_dir_del(dip, name);
	if (error)
		return error;

	error = gfs2_change_nlink(ip, -1);
	if (error)
		return error;

	/* If this inode is being unlinked from the directory structure,
	   we need to mark that in the log so that it isn't lost during
	   a crash. */

	if (!ip->i_di.di_nlink) {
		ul->ul_ut.ut_inum = ip->i_num;
		error = gfs2_unlinked_ondisk_add(sdp, ul);
		if (!error)
			set_bit(GLF_STICKY, &ip->i_gl->gl_flags);
	}

	return error;
}

/**
 * gfs2_rmdiri - Remove a directory
 * @dip: The parent directory of the directory to be removed
 * @name: The name of the directory to be removed
 * @ip: The GFS2 inode of the directory to be removed
 *
 * Assumes Glocks on dip and ip are held
 *
 * Returns: errno
 */

int gfs2_rmdiri(struct gfs2_inode *dip, struct qstr *name,
		struct gfs2_inode *ip, struct gfs2_unlinked *ul)
{
	struct gfs2_sbd *sdp = dip->i_sbd;
	struct qstr dotname;
	int error;

	if (ip->i_di.di_entries != 2) {
		if (gfs2_consist_inode(ip))
			gfs2_dinode_print(&ip->i_di);
		return -EIO;
	}

	error = gfs2_dir_del(dip, name);
	if (error)
		return error;

	error = gfs2_change_nlink(dip, -1);
	if (error)
		return error;

	dotname.len = 1;
	dotname.name = ".";
	error = gfs2_dir_del(ip, &dotname);
	if (error)
		return error;

	dotname.len = 2;
	dotname.name = "..";
	error = gfs2_dir_del(ip, &dotname);
	if (error)
		return error;

	error = gfs2_change_nlink(ip, -2);
	if (error)
		return error;

	/* This inode is being unlinked from the directory structure and
	   we need to mark that in the log so that it isn't lost during
	   a crash. */

	ul->ul_ut.ut_inum = ip->i_num;
	error = gfs2_unlinked_ondisk_add(sdp, ul);
	if (!error)
		set_bit(GLF_STICKY, &ip->i_gl->gl_flags);

	return error;
}

/*
 * gfs2_unlink_ok - check to see that a inode is still in a directory
 * @dip: the directory
 * @name: the name of the file
 * @ip: the inode
 *
 * Assumes that the lock on (at least) @dip is held.
 *
 * Returns: 0 if the parent/child relationship is correct, errno if it isn't
 */

int gfs2_unlink_ok(struct gfs2_inode *dip, struct qstr *name,
		   struct gfs2_inode *ip)
{
	struct gfs2_inum inum;
	unsigned int type;
	int error;

	if (IS_IMMUTABLE(ip->i_vnode) || IS_APPEND(ip->i_vnode))
		return -EPERM;

	if ((dip->i_di.di_mode & S_ISVTX) &&
	    dip->i_di.di_uid != current->fsuid &&
	    ip->i_di.di_uid != current->fsuid &&
	    !capable(CAP_FOWNER))
		return -EPERM;

	if (IS_APPEND(dip->i_vnode))
		return -EPERM;

	error = gfs2_repermission(dip->i_vnode, MAY_WRITE | MAY_EXEC, NULL);
	if (error)
		return error;

	error = gfs2_dir_search(dip, name, &inum, &type);
	if (error)
		return error;

	if (!gfs2_inum_equal(&inum, &ip->i_num))
		return -ENOENT;

	if (IF2DT(ip->i_di.di_mode) != type) {
		gfs2_consist_inode(dip);
		return -EIO;
	}

	return 0;
}

/*
 * gfs2_ok_to_move - check if it's ok to move a directory to another directory
 * @this: move this
 * @to: to here
 *
 * Follow @to back to the root and make sure we don't encounter @this
 * Assumes we already hold the rename lock.
 *
 * Returns: errno
 */

int gfs2_ok_to_move(struct gfs2_inode *this, struct gfs2_inode *to)
{
	struct gfs2_sbd *sdp = this->i_sbd;
1470 1471
	struct inode *dir = to->i_vnode;
	struct inode *tmp;
David Teigland's avatar
David Teigland committed
1472 1473 1474 1475 1476 1477 1478
	struct qstr dotdot;
	int error = 0;

	memset(&dotdot, 0, sizeof(struct qstr));
	dotdot.name = "..";
	dotdot.len = 2;

1479
	igrab(dir);
David Teigland's avatar
David Teigland committed
1480 1481

	for (;;) {
1482
		if (dir == this->i_vnode) {
David Teigland's avatar
David Teigland committed
1483 1484 1485
			error = -EINVAL;
			break;
		}
1486
		if (dir == sdp->sd_root_dir) {
David Teigland's avatar
David Teigland committed
1487 1488 1489 1490
			error = 0;
			break;
		}

1491
		error = gfs2_lookupi(dir, &dotdot, 1, &tmp);
David Teigland's avatar
David Teigland committed
1492 1493 1494
		if (error)
			break;

1495 1496
		iput(dir);
		dir = tmp;
David Teigland's avatar
David Teigland committed
1497 1498
	}

1499
	iput(dir);
David Teigland's avatar
David Teigland committed
1500 1501 1502 1503 1504 1505 1506 1507 1508 1509 1510 1511 1512 1513 1514 1515 1516 1517 1518 1519 1520 1521 1522 1523 1524 1525 1526 1527 1528 1529 1530 1531 1532 1533 1534 1535 1536 1537 1538 1539 1540 1541 1542 1543 1544 1545 1546 1547 1548 1549 1550 1551 1552 1553 1554 1555 1556 1557 1558 1559 1560 1561 1562 1563 1564 1565 1566 1567 1568 1569 1570 1571 1572 1573 1574 1575 1576

	return error;
}

/**
 * gfs2_readlinki - return the contents of a symlink
 * @ip: the symlink's inode
 * @buf: a pointer to the buffer to be filled
 * @len: a pointer to the length of @buf
 *
 * If @buf is too small, a piece of memory is kmalloc()ed and needs
 * to be freed by the caller.
 *
 * Returns: errno
 */

int gfs2_readlinki(struct gfs2_inode *ip, char **buf, unsigned int *len)
{
	struct gfs2_holder i_gh;
	struct buffer_head *dibh;
	unsigned int x;
	int error;

	gfs2_holder_init(ip->i_gl, LM_ST_SHARED, GL_ATIME, &i_gh);
	error = gfs2_glock_nq_atime(&i_gh);
	if (error) {
		gfs2_holder_uninit(&i_gh);
		return error;
	}

	if (!ip->i_di.di_size) {
		gfs2_consist_inode(ip);
		error = -EIO;
		goto out;
	}

	error = gfs2_meta_inode_buffer(ip, &dibh);
	if (error)
		goto out;

	x = ip->i_di.di_size + 1;
	if (x > *len) {
		*buf = kmalloc(x, GFP_KERNEL);
		if (!*buf) {
			error = -ENOMEM;
			goto out_brelse;
		}
	}

	memcpy(*buf, dibh->b_data + sizeof(struct gfs2_dinode), x);
	*len = x;

 out_brelse:
	brelse(dibh);

 out:
	gfs2_glock_dq_uninit(&i_gh);

	return error;
}

/**
 * gfs2_glock_nq_atime - Acquire a hold on an inode's glock, and
 *       conditionally update the inode's atime
 * @gh: the holder to acquire
 *
 * Tests atime (access time) for gfs2_read, gfs2_readdir and gfs2_mmap
 * Update if the difference between the current time and the inode's current
 * atime is greater than an interval specified at mount.
 *
 * Returns: errno
 */

int gfs2_glock_nq_atime(struct gfs2_holder *gh)
{
	struct gfs2_glock *gl = gh->gh_gl;
	struct gfs2_sbd *sdp = gl->gl_sbd;
1577
	struct gfs2_inode *ip = gl->gl_object;
David Teigland's avatar
David Teigland committed
1578 1579 1580 1581 1582 1583 1584 1585 1586 1587 1588 1589 1590 1591 1592 1593 1594 1595 1596 1597 1598 1599 1600 1601 1602 1603 1604 1605 1606 1607 1608 1609 1610 1611 1612 1613 1614 1615 1616 1617 1618 1619 1620 1621 1622 1623 1624 1625 1626 1627
	int64_t curtime, quantum = gfs2_tune_get(sdp, gt_atime_quantum);
	unsigned int state;
	int flags;
	int error;

	if (gfs2_assert_warn(sdp, gh->gh_flags & GL_ATIME) ||
	    gfs2_assert_warn(sdp, !(gh->gh_flags & GL_ASYNC)) ||
	    gfs2_assert_warn(sdp, gl->gl_ops == &gfs2_inode_glops))
		return -EINVAL;

	state = gh->gh_state;
	flags = gh->gh_flags;

	error = gfs2_glock_nq(gh);
	if (error)
		return error;

	if (test_bit(SDF_NOATIME, &sdp->sd_flags) ||
	    (sdp->sd_vfs->s_flags & MS_RDONLY))
		return 0;

	curtime = get_seconds();
	if (curtime - ip->i_di.di_atime >= quantum) {
		gfs2_glock_dq(gh);
		gfs2_holder_reinit(LM_ST_EXCLUSIVE,
				  gh->gh_flags & ~LM_FLAG_ANY,
				  gh);
		error = gfs2_glock_nq(gh);
		if (error)
			return error;

		/* Verify that atime hasn't been updated while we were
		   trying to get exclusive lock. */

		curtime = get_seconds();
		if (curtime - ip->i_di.di_atime >= quantum) {
			struct buffer_head *dibh;

			error = gfs2_trans_begin(sdp, RES_DINODE, 0);
			if (error == -EROFS)
				return 0;
			if (error)
				goto fail;

			error = gfs2_meta_inode_buffer(ip, &dibh);
			if (error)
				goto fail_end_trans;

			ip->i_di.di_atime = curtime;

1628
			gfs2_trans_add_bh(ip->i_gl, dibh, 1);
David Teigland's avatar
David Teigland committed
1629 1630 1631 1632 1633 1634 1635 1636 1637 1638 1639 1640 1641 1642 1643 1644 1645 1646 1647 1648 1649 1650 1651 1652 1653 1654 1655 1656 1657 1658 1659 1660 1661 1662 1663 1664 1665 1666 1667 1668 1669 1670 1671 1672 1673 1674 1675 1676 1677 1678 1679 1680 1681 1682 1683 1684 1685 1686 1687 1688 1689 1690 1691 1692 1693 1694 1695 1696 1697 1698 1699 1700 1701 1702 1703 1704 1705 1706 1707 1708 1709 1710 1711 1712 1713 1714 1715 1716 1717 1718 1719 1720 1721 1722 1723 1724 1725 1726 1727 1728 1729 1730 1731 1732 1733 1734 1735 1736 1737 1738 1739 1740 1741 1742 1743 1744 1745 1746 1747 1748 1749 1750 1751 1752 1753 1754 1755 1756 1757 1758 1759 1760 1761 1762 1763 1764 1765 1766 1767 1768 1769 1770 1771 1772 1773 1774 1775 1776 1777 1778 1779 1780 1781 1782 1783 1784 1785 1786 1787 1788 1789 1790 1791 1792 1793 1794 1795 1796 1797 1798 1799 1800 1801 1802
			gfs2_dinode_out(&ip->i_di, dibh->b_data);
			brelse(dibh);

			gfs2_trans_end(sdp);
		}

		/* If someone else has asked for the glock,
		   unlock and let them have it. Then reacquire
		   in the original state. */
		if (gfs2_glock_is_blocking(gl)) {
			gfs2_glock_dq(gh);
			gfs2_holder_reinit(state, flags, gh);
			return gfs2_glock_nq(gh);
		}
	}

	return 0;

 fail_end_trans:
	gfs2_trans_end(sdp);

 fail:
	gfs2_glock_dq(gh);

	return error;
}

/**
 * glock_compare_atime - Compare two struct gfs2_glock structures for sort
 * @arg_a: the first structure
 * @arg_b: the second structure
 *
 * Returns: 1 if A > B
 *         -1 if A < B
 *          0 if A = B
 */

static int glock_compare_atime(const void *arg_a, const void *arg_b)
{
	struct gfs2_holder *gh_a = *(struct gfs2_holder **)arg_a;
	struct gfs2_holder *gh_b = *(struct gfs2_holder **)arg_b;
	struct lm_lockname *a = &gh_a->gh_gl->gl_name;
	struct lm_lockname *b = &gh_b->gh_gl->gl_name;
	int ret = 0;

	if (a->ln_number > b->ln_number)
		ret = 1;
	else if (a->ln_number < b->ln_number)
		ret = -1;
	else {
		if (gh_a->gh_state == LM_ST_SHARED &&
		    gh_b->gh_state == LM_ST_EXCLUSIVE)
			ret = 1;
		else if (gh_a->gh_state == LM_ST_SHARED &&
			 (gh_b->gh_flags & GL_ATIME))
			ret = 1;
	}

	return ret;
}

/**
 * gfs2_glock_nq_m_atime - acquire multiple glocks where one may need an
 *      atime update
 * @num_gh: the number of structures
 * @ghs: an array of struct gfs2_holder structures
 *
 * Returns: 0 on success (all glocks acquired),
 *          errno on failure (no glocks acquired)
 */

int gfs2_glock_nq_m_atime(unsigned int num_gh, struct gfs2_holder *ghs)
{
	struct gfs2_holder **p;
	unsigned int x;
	int error = 0;

	if (!num_gh)
		return 0;

	if (num_gh == 1) {
		ghs->gh_flags &= ~(LM_FLAG_TRY | GL_ASYNC);
		if (ghs->gh_flags & GL_ATIME)
			error = gfs2_glock_nq_atime(ghs);
		else
			error = gfs2_glock_nq(ghs);
		return error;
	}

	p = kcalloc(num_gh, sizeof(struct gfs2_holder *), GFP_KERNEL);
	if (!p)
		return -ENOMEM;

	for (x = 0; x < num_gh; x++)
		p[x] = &ghs[x];

	sort(p, num_gh, sizeof(struct gfs2_holder *), glock_compare_atime,NULL);

	for (x = 0; x < num_gh; x++) {
		p[x]->gh_flags &= ~(LM_FLAG_TRY | GL_ASYNC);

		if (p[x]->gh_flags & GL_ATIME)
			error = gfs2_glock_nq_atime(p[x]);
		else
			error = gfs2_glock_nq(p[x]);

		if (error) {
			while (x--)
				gfs2_glock_dq(p[x]);
			break;
		}
	}

	kfree(p);

	return error;
}

/**
 * gfs2_try_toss_vnode - See if we can toss a vnode from memory
 * @ip: the inode
 *
 * Returns:  1 if the vnode was tossed
 */

void gfs2_try_toss_vnode(struct gfs2_inode *ip)
{
	struct inode *inode;

	inode = gfs2_ip2v_lookup(ip);
	if (!inode)
		return;

	d_prune_aliases(inode);

	if (S_ISDIR(ip->i_di.di_mode)) {
		struct list_head *head = &inode->i_dentry;
		struct dentry *d = NULL;

		spin_lock(&dcache_lock);
		if (list_empty(head))
			spin_unlock(&dcache_lock);
		else {
			d = list_entry(head->next, struct dentry, d_alias);
			dget_locked(d);
			spin_unlock(&dcache_lock);

			if (have_submounts(d))
				dput(d);
			else {
				shrink_dcache_parent(d);
				dput(d);
				d_prune_aliases(inode);
			}
		}
	}

	inode->i_nlink = 0;
	iput(inode);
}


static int
__gfs2_setattr_simple(struct gfs2_inode *ip, struct iattr *attr)
{
	struct buffer_head *dibh;
	int error;

	error = gfs2_meta_inode_buffer(ip, &dibh);
	if (!error) {
		error = inode_setattr(ip->i_vnode, attr);
		gfs2_assert_warn(ip->i_sbd, !error);
		gfs2_inode_attr_out(ip);

1803
		gfs2_trans_add_bh(ip->i_gl, dibh, 1);
David Teigland's avatar
David Teigland committed
1804 1805 1806 1807 1808 1809 1810 1811 1812 1813 1814 1815 1816 1817 1818 1819 1820 1821 1822 1823
		gfs2_dinode_out(&ip->i_di, dibh->b_data);
		brelse(dibh);
	}
	return error;
}

/**
 * gfs2_setattr_simple -
 * @ip:
 * @attr:
 *
 * Called with a reference on the vnode.
 *
 * Returns: errno
 */

int gfs2_setattr_simple(struct gfs2_inode *ip, struct iattr *attr)
{
	int error;

1824
	if (current->journal_info)
David Teigland's avatar
David Teigland committed
1825 1826 1827 1828 1829 1830 1831 1832 1833 1834 1835 1836 1837 1838 1839 1840 1841 1842
		return __gfs2_setattr_simple(ip, attr);

	error = gfs2_trans_begin(ip->i_sbd, RES_DINODE, 0);
	if (error)
		return error;

	error = __gfs2_setattr_simple(ip, attr);

	gfs2_trans_end(ip->i_sbd);

	return error;
}

int gfs2_repermission(struct inode *inode, int mask, struct nameidata *nd)
{
	return permission(inode, mask, nd);
}