super.c 58.2 KB
Newer Older
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
/*
 * This file is part of UBIFS.
 *
 * Copyright (C) 2006-2008 Nokia Corporation.
 *
 * This program is free software; you can redistribute it and/or modify it
 * under the terms of the GNU General Public License version 2 as published by
 * the Free Software Foundation.
 *
 * 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., 51
 * Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
 *
 * Authors: Artem Bityutskiy (Битюцкий Артём)
 *          Adrian Hunter
 */

/*
 * This file implements UBIFS initialization and VFS superblock operations. Some
 * initialization stuff which is rather large and complex is placed at
 * corresponding subsystems, but most of it is here.
 */

#include <linux/init.h>
#include <linux/slab.h>
#include <linux/module.h>
#include <linux/ctype.h>
#include <linux/kthread.h>
#include <linux/parser.h>
#include <linux/seq_file.h>
#include <linux/mount.h>
A
Artem Bityutskiy 已提交
37
#include <linux/math64.h>
38
#include <linux/writeback.h>
39
#include <linux/smp_lock.h>
40 41
#include "ubifs.h"

42 43 44 45 46 47
/*
 * Maximum amount of memory we may 'kmalloc()' without worrying that we are
 * allocating too much.
 */
#define UBIFS_KMALLOC_OK (128*1024)

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
/* Slab cache for UBIFS inodes */
struct kmem_cache *ubifs_inode_slab;

/* UBIFS TNC shrinker description */
static struct shrinker ubifs_shrinker_info = {
	.shrink = ubifs_shrinker,
	.seeks = DEFAULT_SEEKS,
};

/**
 * validate_inode - validate inode.
 * @c: UBIFS file-system description object
 * @inode: the inode to validate
 *
 * This is a helper function for 'ubifs_iget()' which validates various fields
 * of a newly built inode to make sure they contain sane values and prevent
 * possible vulnerabilities. Returns zero if the inode is all right and
 * a non-zero error code if not.
 */
static int validate_inode(struct ubifs_info *c, const struct inode *inode)
{
	int err;
	const struct ubifs_inode *ui = ubifs_inode(inode);

	if (inode->i_size > c->max_inode_sz) {
		ubifs_err("inode is too large (%lld)",
			  (long long)inode->i_size);
		return 1;
	}

	if (ui->compr_type < 0 || ui->compr_type >= UBIFS_COMPR_TYPES_CNT) {
		ubifs_err("unknown compression type %d", ui->compr_type);
		return 2;
	}

	if (ui->xattr_names + ui->xattr_cnt > XATTR_LIST_MAX)
		return 3;

	if (ui->data_len < 0 || ui->data_len > UBIFS_MAX_INO_DATA)
		return 4;

	if (ui->xattr && (inode->i_mode & S_IFMT) != S_IFREG)
		return 5;

	if (!ubifs_compr_present(ui->compr_type)) {
		ubifs_warn("inode %lu uses '%s' compression, but it was not "
			   "compiled in", inode->i_ino,
			   ubifs_compr_name(ui->compr_type));
	}

	err = dbg_check_dir_size(c, inode);
	return err;
}

struct inode *ubifs_iget(struct super_block *sb, unsigned long inum)
{
	int err;
	union ubifs_key key;
	struct ubifs_ino_node *ino;
	struct ubifs_info *c = sb->s_fs_info;
	struct inode *inode;
	struct ubifs_inode *ui;

	dbg_gen("inode %lu", inum);

	inode = iget_locked(sb, inum);
	if (!inode)
		return ERR_PTR(-ENOMEM);
	if (!(inode->i_state & I_NEW))
		return inode;
	ui = ubifs_inode(inode);

	ino = kmalloc(UBIFS_MAX_INO_NODE_SZ, GFP_NOFS);
	if (!ino) {
		err = -ENOMEM;
		goto out;
	}

	ino_key_init(c, &key, inode->i_ino);

	err = ubifs_tnc_lookup(c, &key, ino);
	if (err)
		goto out_ino;

	inode->i_flags |= (S_NOCMTIME | S_NOATIME);
	inode->i_nlink = le32_to_cpu(ino->nlink);
	inode->i_uid   = le32_to_cpu(ino->uid);
	inode->i_gid   = le32_to_cpu(ino->gid);
	inode->i_atime.tv_sec  = (int64_t)le64_to_cpu(ino->atime_sec);
	inode->i_atime.tv_nsec = le32_to_cpu(ino->atime_nsec);
	inode->i_mtime.tv_sec  = (int64_t)le64_to_cpu(ino->mtime_sec);
	inode->i_mtime.tv_nsec = le32_to_cpu(ino->mtime_nsec);
	inode->i_ctime.tv_sec  = (int64_t)le64_to_cpu(ino->ctime_sec);
	inode->i_ctime.tv_nsec = le32_to_cpu(ino->ctime_nsec);
	inode->i_mode = le32_to_cpu(ino->mode);
	inode->i_size = le64_to_cpu(ino->size);

	ui->data_len    = le32_to_cpu(ino->data_len);
	ui->flags       = le32_to_cpu(ino->flags);
	ui->compr_type  = le16_to_cpu(ino->compr_type);
	ui->creat_sqnum = le64_to_cpu(ino->creat_sqnum);
	ui->xattr_cnt   = le32_to_cpu(ino->xattr_cnt);
	ui->xattr_size  = le32_to_cpu(ino->xattr_size);
	ui->xattr_names = le32_to_cpu(ino->xattr_names);
	ui->synced_i_size = ui->ui_size = inode->i_size;

	ui->xattr = (ui->flags & UBIFS_XATTR_FL) ? 1 : 0;

	err = validate_inode(c, inode);
	if (err)
		goto out_invalid;

A
Artem Bityutskiy 已提交
160
	/* Disable read-ahead */
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
	inode->i_mapping->backing_dev_info = &c->bdi;

	switch (inode->i_mode & S_IFMT) {
	case S_IFREG:
		inode->i_mapping->a_ops = &ubifs_file_address_operations;
		inode->i_op = &ubifs_file_inode_operations;
		inode->i_fop = &ubifs_file_operations;
		if (ui->xattr) {
			ui->data = kmalloc(ui->data_len + 1, GFP_NOFS);
			if (!ui->data) {
				err = -ENOMEM;
				goto out_ino;
			}
			memcpy(ui->data, ino->data, ui->data_len);
			((char *)ui->data)[ui->data_len] = '\0';
		} else if (ui->data_len != 0) {
			err = 10;
			goto out_invalid;
		}
		break;
	case S_IFDIR:
		inode->i_op  = &ubifs_dir_inode_operations;
		inode->i_fop = &ubifs_dir_operations;
		if (ui->data_len != 0) {
			err = 11;
			goto out_invalid;
		}
		break;
	case S_IFLNK:
		inode->i_op = &ubifs_symlink_inode_operations;
		if (ui->data_len <= 0 || ui->data_len > UBIFS_MAX_INO_DATA) {
			err = 12;
			goto out_invalid;
		}
		ui->data = kmalloc(ui->data_len + 1, GFP_NOFS);
		if (!ui->data) {
			err = -ENOMEM;
			goto out_ino;
		}
		memcpy(ui->data, ino->data, ui->data_len);
		((char *)ui->data)[ui->data_len] = '\0';
		break;
	case S_IFBLK:
	case S_IFCHR:
	{
		dev_t rdev;
		union ubifs_dev_desc *dev;

		ui->data = kmalloc(sizeof(union ubifs_dev_desc), GFP_NOFS);
		if (!ui->data) {
			err = -ENOMEM;
			goto out_ino;
		}

		dev = (union ubifs_dev_desc *)ino->data;
		if (ui->data_len == sizeof(dev->new))
			rdev = new_decode_dev(le32_to_cpu(dev->new));
		else if (ui->data_len == sizeof(dev->huge))
			rdev = huge_decode_dev(le64_to_cpu(dev->huge));
		else {
			err = 13;
			goto out_invalid;
		}
		memcpy(ui->data, ino->data, ui->data_len);
		inode->i_op = &ubifs_file_inode_operations;
		init_special_inode(inode, inode->i_mode, rdev);
		break;
	}
	case S_IFSOCK:
	case S_IFIFO:
		inode->i_op = &ubifs_file_inode_operations;
		init_special_inode(inode, inode->i_mode, 0);
		if (ui->data_len != 0) {
			err = 14;
			goto out_invalid;
		}
		break;
	default:
		err = 15;
		goto out_invalid;
	}

	kfree(ino);
	ubifs_set_inode_flags(inode);
	unlock_new_inode(inode);
	return inode;

out_invalid:
	ubifs_err("inode %lu validation failed, error %d", inode->i_ino, err);
	dbg_dump_node(c, ino);
	dbg_dump_inode(c, inode);
	err = -EINVAL;
out_ino:
	kfree(ino);
out:
	ubifs_err("failed to read inode %lu, error %d", inode->i_ino, err);
	iget_failed(inode);
	return ERR_PTR(err);
}

static struct inode *ubifs_alloc_inode(struct super_block *sb)
{
	struct ubifs_inode *ui;

	ui = kmem_cache_alloc(ubifs_inode_slab, GFP_NOFS);
	if (!ui)
		return NULL;

	memset((void *)ui + sizeof(struct inode), 0,
	       sizeof(struct ubifs_inode) - sizeof(struct inode));
	mutex_init(&ui->ui_mutex);
	spin_lock_init(&ui->ui_lock);
	return &ui->vfs_inode;
};

static void ubifs_destroy_inode(struct inode *inode)
{
	struct ubifs_inode *ui = ubifs_inode(inode);

	kfree(ui->data);
	kmem_cache_free(ubifs_inode_slab, inode);
}

/*
 * Note, Linux write-back code calls this without 'i_mutex'.
 */
static int ubifs_write_inode(struct inode *inode, int wait)
{
289
	int err = 0;
290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309
	struct ubifs_info *c = inode->i_sb->s_fs_info;
	struct ubifs_inode *ui = ubifs_inode(inode);

	ubifs_assert(!ui->xattr);
	if (is_bad_inode(inode))
		return 0;

	mutex_lock(&ui->ui_mutex);
	/*
	 * Due to races between write-back forced by budgeting
	 * (see 'sync_some_inodes()') and pdflush write-back, the inode may
	 * have already been synchronized, do not do this again. This might
	 * also happen if it was synchronized in an VFS operation, e.g.
	 * 'ubifs_link()'.
	 */
	if (!ui->dirty) {
		mutex_unlock(&ui->ui_mutex);
		return 0;
	}

310 311 312 313 314 315 316
	/*
	 * As an optimization, do not write orphan inodes to the media just
	 * because this is not needed.
	 */
	dbg_gen("inode %lu, mode %#x, nlink %u",
		inode->i_ino, (int)inode->i_mode, inode->i_nlink);
	if (inode->i_nlink) {
317
		err = ubifs_jnl_write_inode(c, inode);
318 319 320 321
		if (err)
			ubifs_err("can't write inode %lu, error %d",
				  inode->i_ino, err);
	}
322 323 324 325 326 327 328 329 330 331 332

	ui->dirty = 0;
	mutex_unlock(&ui->ui_mutex);
	ubifs_release_dirty_inode_budget(c, ui);
	return err;
}

static void ubifs_delete_inode(struct inode *inode)
{
	int err;
	struct ubifs_info *c = inode->i_sb->s_fs_info;
333
	struct ubifs_inode *ui = ubifs_inode(inode);
334

335
	if (ui->xattr)
336 337 338 339 340 341 342
		/*
		 * Extended attribute inode deletions are fully handled in
		 * 'ubifs_removexattr()'. These inodes are special and have
		 * limited usage, so there is nothing to do here.
		 */
		goto out;

A
Artem Bityutskiy 已提交
343
	dbg_gen("inode %lu, mode %#x", inode->i_ino, (int)inode->i_mode);
344 345 346 347 348 349 350
	ubifs_assert(!atomic_read(&inode->i_count));
	ubifs_assert(inode->i_nlink == 0);

	truncate_inode_pages(&inode->i_data, 0);
	if (is_bad_inode(inode))
		goto out;

351
	ui->ui_size = inode->i_size = 0;
A
Artem Bityutskiy 已提交
352
	err = ubifs_jnl_delete_inode(c, inode);
353 354 355
	if (err)
		/*
		 * Worst case we have a lost orphan inode wasting space, so a
A
Artem Bityutskiy 已提交
356
		 * simple error message is OK here.
357
		 */
A
Artem Bityutskiy 已提交
358 359 360
		ubifs_err("can't delete inode %lu, error %d",
			  inode->i_ino, err);

361
out:
362 363
	if (ui->dirty)
		ubifs_release_dirty_inode_budget(c, ui);
364 365 366 367 368
	else {
		/* We've deleted something - clean the "no space" flags */
		c->nospace = c->nospace_rp = 0;
		smp_wmb();
	}
369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386
	clear_inode(inode);
}

static void ubifs_dirty_inode(struct inode *inode)
{
	struct ubifs_inode *ui = ubifs_inode(inode);

	ubifs_assert(mutex_is_locked(&ui->ui_mutex));
	if (!ui->dirty) {
		ui->dirty = 1;
		dbg_gen("inode %lu",  inode->i_ino);
	}
}

static int ubifs_statfs(struct dentry *dentry, struct kstatfs *buf)
{
	struct ubifs_info *c = dentry->d_sb->s_fs_info;
	unsigned long long free;
A
Artem Bityutskiy 已提交
387
	__le32 *uuid = (__le32 *)c->uuid;
388

389
	free = ubifs_get_free_space(c);
390 391 392 393 394 395 396 397 398 399 400 401 402 403
	dbg_gen("free space %lld bytes (%lld blocks)",
		free, free >> UBIFS_BLOCK_SHIFT);

	buf->f_type = UBIFS_SUPER_MAGIC;
	buf->f_bsize = UBIFS_BLOCK_SIZE;
	buf->f_blocks = c->block_cnt;
	buf->f_bfree = free >> UBIFS_BLOCK_SHIFT;
	if (free > c->report_rp_size)
		buf->f_bavail = (free - c->report_rp_size) >> UBIFS_BLOCK_SHIFT;
	else
		buf->f_bavail = 0;
	buf->f_files = 0;
	buf->f_ffree = 0;
	buf->f_namelen = UBIFS_MAX_NLEN;
A
Artem Bityutskiy 已提交
404 405
	buf->f_fsid.val[0] = le32_to_cpu(uuid[0]) ^ le32_to_cpu(uuid[2]);
	buf->f_fsid.val[1] = le32_to_cpu(uuid[1]) ^ le32_to_cpu(uuid[3]);
406
	ubifs_assert(buf->f_bfree <= c->block_cnt);
407 408 409 410 411 412 413 414 415 416 417 418
	return 0;
}

static int ubifs_show_options(struct seq_file *s, struct vfsmount *mnt)
{
	struct ubifs_info *c = mnt->mnt_sb->s_fs_info;

	if (c->mount_opts.unmount_mode == 2)
		seq_printf(s, ",fast_unmount");
	else if (c->mount_opts.unmount_mode == 1)
		seq_printf(s, ",norm_unmount");

A
Adrian Hunter 已提交
419 420 421 422 423
	if (c->mount_opts.bulk_read == 2)
		seq_printf(s, ",bulk_read");
	else if (c->mount_opts.bulk_read == 1)
		seq_printf(s, ",no_bulk_read");

424 425 426 427 428
	if (c->mount_opts.chk_data_crc == 2)
		seq_printf(s, ",chk_data_crc");
	else if (c->mount_opts.chk_data_crc == 1)
		seq_printf(s, ",no_chk_data_crc");

429
	if (c->mount_opts.override_compr) {
H
Hunter Adrian 已提交
430 431
		seq_printf(s, ",compr=%s",
			   ubifs_compr_name(c->mount_opts.compr_type));
432 433
	}

434 435 436 437 438
	return 0;
}

static int ubifs_sync_fs(struct super_block *sb, int wait)
{
A
Artem Bityutskiy 已提交
439
	int i, err;
440
	struct ubifs_info *c = sb->s_fs_info;
441

442
	/*
A
Artem Bityutskiy 已提交
443 444
	 * Zero @wait is just an advisory thing to help the file system shove
	 * lots of data into the queues, and there will be the second
445 446
	 * '->sync_fs()' call, with non-zero @wait.
	 */
A
Artem Bityutskiy 已提交
447 448
	if (!wait)
		return 0;
449

450 451 452 453 454 455 456 457 458
	/*
	 * VFS calls '->sync_fs()' before synchronizing all dirty inodes and
	 * pages, so synchronize them first, then commit the journal. Strictly
	 * speaking, it is not necessary to commit the journal here,
	 * synchronizing write-buffers would be enough. But committing makes
	 * UBIFS free space predictions much more accurate, so we want to let
	 * the user be able to get more accurate results of 'statfs()' after
	 * they synchronize the file system.
	 */
459
	sync_inodes_sb(sb);
460

461 462 463 464 465 466 467 468 469 470
	/*
	 * Synchronize write buffers, because 'ubifs_run_commit()' does not
	 * do this if it waits for an already running commit.
	 */
	for (i = 0; i < c->jhead_cnt; i++) {
		err = ubifs_wbuf_sync(&c->jheads[i].wbuf);
		if (err)
			return err;
	}

A
Artem Bityutskiy 已提交
471 472 473
	err = ubifs_run_commit(c);
	if (err)
		return err;
A
Artem Bityutskiy 已提交
474

A
Artem Bityutskiy 已提交
475
	return ubi_sync(c->vi.ubi_num);
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
}

/**
 * init_constants_early - initialize UBIFS constants.
 * @c: UBIFS file-system description object
 *
 * This function initialize UBIFS constants which do not need the superblock to
 * be read. It also checks that the UBI volume satisfies basic UBIFS
 * requirements. Returns zero in case of success and a negative error code in
 * case of failure.
 */
static int init_constants_early(struct ubifs_info *c)
{
	if (c->vi.corrupted) {
		ubifs_warn("UBI volume is corrupted - read-only mode");
		c->ro_media = 1;
	}

	if (c->di.ro_mode) {
		ubifs_msg("read-only UBI device");
		c->ro_media = 1;
	}

	if (c->vi.vol_type == UBI_STATIC_VOLUME) {
		ubifs_msg("static UBI volume - read-only mode");
		c->ro_media = 1;
	}

	c->leb_cnt = c->vi.size;
	c->leb_size = c->vi.usable_leb_size;
	c->half_leb_size = c->leb_size / 2;
	c->min_io_size = c->di.min_io_size;
	c->min_io_shift = fls(c->min_io_size) - 1;

	if (c->leb_size < UBIFS_MIN_LEB_SZ) {
		ubifs_err("too small LEBs (%d bytes), min. is %d bytes",
			  c->leb_size, UBIFS_MIN_LEB_SZ);
		return -EINVAL;
	}

	if (c->leb_cnt < UBIFS_MIN_LEB_CNT) {
		ubifs_err("too few LEBs (%d), min. is %d",
			  c->leb_cnt, UBIFS_MIN_LEB_CNT);
		return -EINVAL;
	}

	if (!is_power_of_2(c->min_io_size)) {
		ubifs_err("bad min. I/O size %d", c->min_io_size);
		return -EINVAL;
	}

	/*
	 * UBIFS aligns all node to 8-byte boundary, so to make function in
	 * io.c simpler, assume minimum I/O unit size to be 8 bytes if it is
	 * less than 8.
	 */
	if (c->min_io_size < 8) {
		c->min_io_size = 8;
		c->min_io_shift = 3;
	}

	c->ref_node_alsz = ALIGN(UBIFS_REF_NODE_SZ, c->min_io_size);
	c->mst_node_alsz = ALIGN(UBIFS_MST_NODE_SZ, c->min_io_size);

	/*
	 * Initialize node length ranges which are mostly needed for node
	 * length validation.
	 */
	c->ranges[UBIFS_PAD_NODE].len  = UBIFS_PAD_NODE_SZ;
	c->ranges[UBIFS_SB_NODE].len   = UBIFS_SB_NODE_SZ;
	c->ranges[UBIFS_MST_NODE].len  = UBIFS_MST_NODE_SZ;
	c->ranges[UBIFS_REF_NODE].len  = UBIFS_REF_NODE_SZ;
	c->ranges[UBIFS_TRUN_NODE].len = UBIFS_TRUN_NODE_SZ;
	c->ranges[UBIFS_CS_NODE].len   = UBIFS_CS_NODE_SZ;

	c->ranges[UBIFS_INO_NODE].min_len  = UBIFS_INO_NODE_SZ;
	c->ranges[UBIFS_INO_NODE].max_len  = UBIFS_MAX_INO_NODE_SZ;
	c->ranges[UBIFS_ORPH_NODE].min_len =
				UBIFS_ORPH_NODE_SZ + sizeof(__le64);
	c->ranges[UBIFS_ORPH_NODE].max_len = c->leb_size;
	c->ranges[UBIFS_DENT_NODE].min_len = UBIFS_DENT_NODE_SZ;
	c->ranges[UBIFS_DENT_NODE].max_len = UBIFS_MAX_DENT_NODE_SZ;
	c->ranges[UBIFS_XENT_NODE].min_len = UBIFS_XENT_NODE_SZ;
	c->ranges[UBIFS_XENT_NODE].max_len = UBIFS_MAX_XENT_NODE_SZ;
	c->ranges[UBIFS_DATA_NODE].min_len = UBIFS_DATA_NODE_SZ;
	c->ranges[UBIFS_DATA_NODE].max_len = UBIFS_MAX_DATA_NODE_SZ;
	/*
	 * Minimum indexing node size is amended later when superblock is
	 * read and the key length is known.
	 */
	c->ranges[UBIFS_IDX_NODE].min_len = UBIFS_IDX_NODE_SZ + UBIFS_BRANCH_SZ;
	/*
	 * Maximum indexing node size is amended later when superblock is
	 * read and the fanout is known.
	 */
	c->ranges[UBIFS_IDX_NODE].max_len = INT_MAX;

	/*
574 575
	 * Initialize dead and dark LEB space watermarks. See gc.c for comments
	 * about these values.
576 577 578 579
	 */
	c->dead_wm = ALIGN(MIN_WRITE_SZ, c->min_io_size);
	c->dark_wm = ALIGN(UBIFS_MAX_NODE_SZ, c->min_io_size);

A
Artem Bityutskiy 已提交
580 581 582 583 584 585
	/*
	 * Calculate how many bytes would be wasted at the end of LEB if it was
	 * fully filled with data nodes of maximum size. This is used in
	 * calculations when reporting free space.
	 */
	c->leb_overhead = c->leb_size % UBIFS_MAX_DATA_NODE_SZ;
586

A
Adrian Hunter 已提交
587
	/* Buffer size for bulk-reads */
A
Artem Bityutskiy 已提交
588 589 590
	c->max_bu_buf_len = UBIFS_MAX_BULK_READ * UBIFS_MAX_DATA_NODE_SZ;
	if (c->max_bu_buf_len > c->leb_size)
		c->max_bu_buf_len = c->leb_size;
591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614
	return 0;
}

/**
 * bud_wbuf_callback - bud LEB write-buffer synchronization call-back.
 * @c: UBIFS file-system description object
 * @lnum: LEB the write-buffer was synchronized to
 * @free: how many free bytes left in this LEB
 * @pad: how many bytes were padded
 *
 * This is a callback function which is called by the I/O unit when the
 * write-buffer is synchronized. We need this to correctly maintain space
 * accounting in bud logical eraseblocks. This function returns zero in case of
 * success and a negative error code in case of failure.
 *
 * This function actually belongs to the journal, but we keep it here because
 * we want to keep it static.
 */
static int bud_wbuf_callback(struct ubifs_info *c, int lnum, int free, int pad)
{
	return ubifs_update_one_lp(c, lnum, free, pad, 0, 0);
}

/*
615
 * init_constants_sb - initialize UBIFS constants.
616 617 618 619 620 621 622
 * @c: UBIFS file-system description object
 *
 * This is a helper function which initializes various UBIFS constants after
 * the superblock has been read. It also checks various UBIFS parameters and
 * makes sure they are all right. Returns zero in case of success and a
 * negative error code in case of failure.
 */
623
static int init_constants_sb(struct ubifs_info *c)
624 625
{
	int tmp, err;
A
Artem Bityutskiy 已提交
626
	long long tmp64;
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

	c->main_bytes = (long long)c->main_lebs * c->leb_size;
	c->max_znode_sz = sizeof(struct ubifs_znode) +
				c->fanout * sizeof(struct ubifs_zbranch);

	tmp = ubifs_idx_node_sz(c, 1);
	c->ranges[UBIFS_IDX_NODE].min_len = tmp;
	c->min_idx_node_sz = ALIGN(tmp, 8);

	tmp = ubifs_idx_node_sz(c, c->fanout);
	c->ranges[UBIFS_IDX_NODE].max_len = tmp;
	c->max_idx_node_sz = ALIGN(tmp, 8);

	/* Make sure LEB size is large enough to fit full commit */
	tmp = UBIFS_CS_NODE_SZ + UBIFS_REF_NODE_SZ * c->jhead_cnt;
	tmp = ALIGN(tmp, c->min_io_size);
	if (tmp > c->leb_size) {
		dbg_err("too small LEB size %d, at least %d needed",
			c->leb_size, tmp);
		return -EINVAL;
	}

	/*
	 * Make sure that the log is large enough to fit reference nodes for
	 * all buds plus one reserved LEB.
	 */
A
Artem Bityutskiy 已提交
653 654
	tmp64 = c->max_bud_bytes + c->leb_size - 1;
	c->max_bud_cnt = div_u64(tmp64, c->leb_size);
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
	tmp = (c->ref_node_alsz * c->max_bud_cnt + c->leb_size - 1);
	tmp /= c->leb_size;
	tmp += 1;
	if (c->log_lebs < tmp) {
		dbg_err("too small log %d LEBs, required min. %d LEBs",
			c->log_lebs, tmp);
		return -EINVAL;
	}

	/*
	 * When budgeting we assume worst-case scenarios when the pages are not
	 * be compressed and direntries are of the maximum size.
	 *
	 * Note, data, which may be stored in inodes is budgeted separately, so
	 * it is not included into 'c->inode_budget'.
	 */
	c->page_budget = UBIFS_MAX_DATA_NODE_SZ * UBIFS_BLOCKS_PER_PAGE;
	c->inode_budget = UBIFS_INO_NODE_SZ;
	c->dent_budget = UBIFS_MAX_DENT_NODE_SZ;

	/*
	 * When the amount of flash space used by buds becomes
	 * 'c->max_bud_bytes', UBIFS just blocks all writers and starts commit.
	 * The writers are unblocked when the commit is finished. To avoid
	 * writers to be blocked UBIFS initiates background commit in advance,
	 * when number of bud bytes becomes above the limit defined below.
	 */
	c->bg_bud_bytes = (c->max_bud_bytes * 13) >> 4;

	/*
	 * Ensure minimum journal size. All the bytes in the journal heads are
	 * considered to be used, when calculating the current journal usage.
	 * Consequently, if the journal is too small, UBIFS will treat it as
	 * always full.
	 */
A
Artem Bityutskiy 已提交
690
	tmp64 = (long long)(c->jhead_cnt + 1) * c->leb_size + 1;
691 692 693 694 695 696 697 698 699
	if (c->bg_bud_bytes < tmp64)
		c->bg_bud_bytes = tmp64;
	if (c->max_bud_bytes < tmp64 + c->leb_size)
		c->max_bud_bytes = tmp64 + c->leb_size;

	err = ubifs_calc_lpt_geom(c);
	if (err)
		return err;

700 701
	/* Initialize effective LEB size used in budgeting calculations */
	c->idx_leb_size = c->leb_size - c->max_idx_node_sz;
702 703 704 705 706 707 708 709 710 711 712 713 714 715 716
	return 0;
}

/*
 * init_constants_master - initialize UBIFS constants.
 * @c: UBIFS file-system description object
 *
 * This is a helper function which initializes various UBIFS constants after
 * the master node has been read. It also checks various UBIFS parameters and
 * makes sure they are all right.
 */
static void init_constants_master(struct ubifs_info *c)
{
	long long tmp64;

717
	c->min_idx_lebs = ubifs_calc_min_idx_lebs(c);
718
	c->report_rp_size = ubifs_reported_space(c, c->rp_size);
719 720 721 722 723 724

	/*
	 * Calculate total amount of FS blocks. This number is not used
	 * internally because it does not make much sense for UBIFS, but it is
	 * necessary to report something for the 'statfs()' call.
	 *
725
	 * Subtract the LEB reserved for GC, the LEB which is reserved for
726 727
	 * deletions, minimum LEBs for the index, and assume only one journal
	 * head is available.
728
	 */
729
	tmp64 = c->main_lebs - 1 - 1 - MIN_INDEX_LEBS - c->jhead_cnt + 1;
A
Artem Bityutskiy 已提交
730
	tmp64 *= (long long)c->leb_size - c->leb_overhead;
731 732 733 734 735 736 737 738
	tmp64 = ubifs_reported_space(c, tmp64);
	c->block_cnt = tmp64 >> UBIFS_BLOCK_SHIFT;
}

/**
 * take_gc_lnum - reserve GC LEB.
 * @c: UBIFS file-system description object
 *
739 740 741 742 743 744
 * This function ensures that the LEB reserved for garbage collection is marked
 * as "taken" in lprops. We also have to set free space to LEB size and dirty
 * space to zero, because lprops may contain out-of-date information if the
 * file-system was un-mounted before it has been committed. This function
 * returns zero in case of success and a negative error code in case of
 * failure.
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
 */
static int take_gc_lnum(struct ubifs_info *c)
{
	int err;

	if (c->gc_lnum == -1) {
		ubifs_err("no LEB for GC");
		return -EINVAL;
	}

	/* And we have to tell lprops that this LEB is taken */
	err = ubifs_change_one_lp(c, c->gc_lnum, c->leb_size, 0,
				  LPROPS_TAKEN, 0, 0);
	return err;
}

/**
 * alloc_wbufs - allocate write-buffers.
 * @c: UBIFS file-system description object
 *
 * This helper function allocates and initializes UBIFS write-buffers. Returns
 * zero in case of success and %-ENOMEM in case of failure.
 */
static int alloc_wbufs(struct ubifs_info *c)
{
	int i, err;

	c->jheads = kzalloc(c->jhead_cnt * sizeof(struct ubifs_jhead),
			   GFP_KERNEL);
	if (!c->jheads)
		return -ENOMEM;

	/* Initialize journal heads */
	for (i = 0; i < c->jhead_cnt; i++) {
		INIT_LIST_HEAD(&c->jheads[i].buds_list);
		err = ubifs_wbuf_init(c, &c->jheads[i].wbuf);
		if (err)
			return err;

		c->jheads[i].wbuf.sync_callback = &bud_wbuf_callback;
		c->jheads[i].wbuf.jhead = i;
	}

	c->jheads[BASEHD].wbuf.dtype = UBI_SHORTTERM;
	/*
	 * Garbage Collector head likely contains long-term data and
	 * does not need to be synchronized by timer.
	 */
	c->jheads[GCHD].wbuf.dtype = UBI_LONGTERM;
794
	c->jheads[GCHD].wbuf.no_timer = 1;
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

	return 0;
}

/**
 * free_wbufs - free write-buffers.
 * @c: UBIFS file-system description object
 */
static void free_wbufs(struct ubifs_info *c)
{
	int i;

	if (c->jheads) {
		for (i = 0; i < c->jhead_cnt; i++) {
			kfree(c->jheads[i].wbuf.buf);
			kfree(c->jheads[i].wbuf.inodes);
		}
		kfree(c->jheads);
		c->jheads = NULL;
	}
}

/**
 * free_orphans - free orphans.
 * @c: UBIFS file-system description object
 */
static void free_orphans(struct ubifs_info *c)
{
	struct ubifs_orphan *orph;

	while (c->orph_dnext) {
		orph = c->orph_dnext;
		c->orph_dnext = orph->dnext;
		list_del(&orph->list);
		kfree(orph);
	}

	while (!list_empty(&c->orph_list)) {
		orph = list_entry(c->orph_list.next, struct ubifs_orphan, list);
		list_del(&orph->list);
		kfree(orph);
		dbg_err("orphan list not empty at unmount");
	}

	vfree(c->orph_buf);
	c->orph_buf = NULL;
}

/**
 * free_buds - free per-bud objects.
 * @c: UBIFS file-system description object
 */
static void free_buds(struct ubifs_info *c)
{
	struct rb_node *this = c->buds.rb_node;
	struct ubifs_bud *bud;

	while (this) {
		if (this->rb_left)
			this = this->rb_left;
		else if (this->rb_right)
			this = this->rb_right;
		else {
			bud = rb_entry(this, struct ubifs_bud, rb);
			this = rb_parent(this);
			if (this) {
				if (this->rb_left == &bud->rb)
					this->rb_left = NULL;
				else
					this->rb_right = NULL;
			}
			kfree(bud);
		}
	}
}

/**
 * check_volume_empty - check if the UBI volume is empty.
 * @c: UBIFS file-system description object
 *
 * This function checks if the UBIFS volume is empty by looking if its LEBs are
 * mapped or not. The result of checking is stored in the @c->empty variable.
 * Returns zero in case of success and a negative error code in case of
 * failure.
 */
static int check_volume_empty(struct ubifs_info *c)
{
	int lnum, err;

	c->empty = 1;
	for (lnum = 0; lnum < c->leb_cnt; lnum++) {
		err = ubi_is_mapped(c->ubi, lnum);
		if (unlikely(err < 0))
			return err;
		if (err == 1) {
			c->empty = 0;
			break;
		}

		cond_resched();
	}

	return 0;
}

/*
 * UBIFS mount options.
 *
 * Opt_fast_unmount: do not run a journal commit before un-mounting
 * Opt_norm_unmount: run a journal commit before un-mounting
A
Adrian Hunter 已提交
905 906
 * Opt_bulk_read: enable bulk-reads
 * Opt_no_bulk_read: disable bulk-reads
907 908
 * Opt_chk_data_crc: check CRCs when reading data nodes
 * Opt_no_chk_data_crc: do not check CRCs when reading data nodes
909
 * Opt_override_compr: override default compressor
910 911 912 913 914
 * Opt_err: just end of array marker
 */
enum {
	Opt_fast_unmount,
	Opt_norm_unmount,
A
Adrian Hunter 已提交
915 916
	Opt_bulk_read,
	Opt_no_bulk_read,
917 918
	Opt_chk_data_crc,
	Opt_no_chk_data_crc,
919
	Opt_override_compr,
920 921 922
	Opt_err,
};

923
static const match_table_t tokens = {
924 925
	{Opt_fast_unmount, "fast_unmount"},
	{Opt_norm_unmount, "norm_unmount"},
A
Adrian Hunter 已提交
926 927
	{Opt_bulk_read, "bulk_read"},
	{Opt_no_bulk_read, "no_bulk_read"},
928 929
	{Opt_chk_data_crc, "chk_data_crc"},
	{Opt_no_chk_data_crc, "no_chk_data_crc"},
930
	{Opt_override_compr, "compr=%s"},
931 932 933
	{Opt_err, NULL},
};

934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954
/**
 * parse_standard_option - parse a standard mount option.
 * @option: the option to parse
 *
 * Normally, standard mount options like "sync" are passed to file-systems as
 * flags. However, when a "rootflags=" kernel boot parameter is used, they may
 * be present in the options string. This function tries to deal with this
 * situation and parse standard options. Returns 0 if the option was not
 * recognized, and the corresponding integer flag if it was.
 *
 * UBIFS is only interested in the "sync" option, so do not check for anything
 * else.
 */
static int parse_standard_option(const char *option)
{
	ubifs_msg("parse %s", option);
	if (!strcmp(option, "sync"))
		return MS_SYNCHRONOUS;
	return 0;
}

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
/**
 * ubifs_parse_options - parse mount parameters.
 * @c: UBIFS file-system description object
 * @options: parameters to parse
 * @is_remount: non-zero if this is FS re-mount
 *
 * This function parses UBIFS mount options and returns zero in case success
 * and a negative error code in case of failure.
 */
static int ubifs_parse_options(struct ubifs_info *c, char *options,
			       int is_remount)
{
	char *p;
	substring_t args[MAX_OPT_ARGS];

	if (!options)
		return 0;

	while ((p = strsep(&options, ","))) {
		int token;

		if (!*p)
			continue;

		token = match_token(p, tokens, args);
		switch (token) {
A
Artem Bityutskiy 已提交
981 982
		/*
		 * %Opt_fast_unmount and %Opt_norm_unmount options are ignored.
A
Artem Bityutskiy 已提交
983
		 * We accept them in order to be backward-compatible. But this
A
Artem Bityutskiy 已提交
984 985
		 * should be removed at some point.
		 */
986 987 988 989 990 991
		case Opt_fast_unmount:
			c->mount_opts.unmount_mode = 2;
			break;
		case Opt_norm_unmount:
			c->mount_opts.unmount_mode = 1;
			break;
A
Adrian Hunter 已提交
992 993 994 995 996 997 998 999
		case Opt_bulk_read:
			c->mount_opts.bulk_read = 2;
			c->bulk_read = 1;
			break;
		case Opt_no_bulk_read:
			c->mount_opts.bulk_read = 1;
			c->bulk_read = 0;
			break;
1000 1001 1002 1003 1004 1005 1006 1007
		case Opt_chk_data_crc:
			c->mount_opts.chk_data_crc = 2;
			c->no_chk_data_crc = 0;
			break;
		case Opt_no_chk_data_crc:
			c->mount_opts.chk_data_crc = 1;
			c->no_chk_data_crc = 1;
			break;
1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029
		case Opt_override_compr:
		{
			char *name = match_strdup(&args[0]);

			if (!name)
				return -ENOMEM;
			if (!strcmp(name, "none"))
				c->mount_opts.compr_type = UBIFS_COMPR_NONE;
			else if (!strcmp(name, "lzo"))
				c->mount_opts.compr_type = UBIFS_COMPR_LZO;
			else if (!strcmp(name, "zlib"))
				c->mount_opts.compr_type = UBIFS_COMPR_ZLIB;
			else {
				ubifs_err("unknown compressor \"%s\"", name);
				kfree(name);
				return -EINVAL;
			}
			kfree(name);
			c->mount_opts.override_compr = 1;
			c->default_compr = c->mount_opts.compr_type;
			break;
		}
1030
		default:
1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043
		{
			unsigned long flag;
			struct super_block *sb = c->vfs_sb;

			flag = parse_standard_option(p);
			if (!flag) {
				ubifs_err("unrecognized mount option \"%s\" "
					  "or missing value", p);
				return -EINVAL;
			}
			sb->s_flags |= flag;
			break;
		}
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
		}
	}

	return 0;
}

/**
 * destroy_journal - destroy journal data structures.
 * @c: UBIFS file-system description object
 *
 * This function destroys journal data structures including those that may have
 * been created by recovery functions.
 */
static void destroy_journal(struct ubifs_info *c)
{
	while (!list_empty(&c->unclean_leb_list)) {
		struct ubifs_unclean_leb *ucleb;

		ucleb = list_entry(c->unclean_leb_list.next,
				   struct ubifs_unclean_leb, list);
		list_del(&ucleb->list);
		kfree(ucleb);
	}
	while (!list_empty(&c->old_buds)) {
		struct ubifs_bud *bud;

		bud = list_entry(c->old_buds.next, struct ubifs_bud, list);
		list_del(&bud->list);
		kfree(bud);
	}
	ubifs_destroy_idx_gc(c);
	ubifs_destroy_size_tree(c);
	ubifs_tnc_close(c);
	free_buds(c);
}

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
/**
 * bu_init - initialize bulk-read information.
 * @c: UBIFS file-system description object
 */
static void bu_init(struct ubifs_info *c)
{
	ubifs_assert(c->bulk_read == 1);

	if (c->bu.buf)
		return; /* Already initialized */

again:
	c->bu.buf = kmalloc(c->max_bu_buf_len, GFP_KERNEL | __GFP_NOWARN);
	if (!c->bu.buf) {
		if (c->max_bu_buf_len > UBIFS_KMALLOC_OK) {
			c->max_bu_buf_len = UBIFS_KMALLOC_OK;
			goto again;
		}

		/* Just disable bulk-read */
		ubifs_warn("Cannot allocate %d bytes of memory for bulk-read, "
			   "disabling it", c->max_bu_buf_len);
		c->mount_opts.bulk_read = 1;
		c->bulk_read = 0;
		return;
	}
}

1108 1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120 1121
/**
 * check_free_space - check if there is enough free space to mount.
 * @c: UBIFS file-system description object
 *
 * This function makes sure UBIFS has enough free space to be mounted in
 * read/write mode. UBIFS must always have some free space to allow deletions.
 */
static int check_free_space(struct ubifs_info *c)
{
	ubifs_assert(c->dark_wm > 0);
	if (c->lst.total_free + c->lst.total_dirty < c->dark_wm) {
		ubifs_err("insufficient free space to mount in read/write mode");
		dbg_dump_budg(c);
		dbg_dump_lprops(c);
1122
		return -ENOSPC;
1123 1124 1125 1126
	}
	return 0;
}

1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147
/**
 * mount_ubifs - mount UBIFS file-system.
 * @c: UBIFS file-system description object
 *
 * This function mounts UBIFS file system. Returns zero in case of success and
 * a negative error code in case of failure.
 *
 * Note, the function does not de-allocate resources it it fails half way
 * through, and the caller has to do this instead.
 */
static int mount_ubifs(struct ubifs_info *c)
{
	struct super_block *sb = c->vfs_sb;
	int err, mounted_read_only = (sb->s_flags & MS_RDONLY);
	long long x;
	size_t sz;

	err = init_constants_early(c);
	if (err)
		return err;

1148 1149 1150
	err = ubifs_debugging_init(c);
	if (err)
		return err;
1151 1152 1153 1154 1155 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 1190 1191 1192

	err = check_volume_empty(c);
	if (err)
		goto out_free;

	if (c->empty && (mounted_read_only || c->ro_media)) {
		/*
		 * This UBI volume is empty, and read-only, or the file system
		 * is mounted read-only - we cannot format it.
		 */
		ubifs_err("can't format empty UBI volume: read-only %s",
			  c->ro_media ? "UBI volume" : "mount");
		err = -EROFS;
		goto out_free;
	}

	if (c->ro_media && !mounted_read_only) {
		ubifs_err("cannot mount read-write - read-only media");
		err = -EROFS;
		goto out_free;
	}

	/*
	 * The requirement for the buffer is that it should fit indexing B-tree
	 * height amount of integers. We assume the height if the TNC tree will
	 * never exceed 64.
	 */
	err = -ENOMEM;
	c->bottom_up_buf = kmalloc(BOTTOM_UP_HEIGHT * sizeof(int), GFP_KERNEL);
	if (!c->bottom_up_buf)
		goto out_free;

	c->sbuf = vmalloc(c->leb_size);
	if (!c->sbuf)
		goto out_free;

	if (!mounted_read_only) {
		c->ileb_buf = vmalloc(c->leb_size);
		if (!c->ileb_buf)
			goto out_free;
	}

1193 1194 1195 1196 1197 1198 1199
	if (c->bulk_read == 1)
		bu_init(c);

	/*
	 * We have to check all CRCs, even for data nodes, when we mount the FS
	 * (specifically, when we are replaying).
	 */
1200 1201
	c->always_chk_crc = 1;

1202 1203 1204 1205 1206
	err = ubifs_read_superblock(c);
	if (err)
		goto out_free;

	/*
1207
	 * Make sure the compressor which is set as default in the superblock
1208
	 * or overridden by mount options is actually compiled in.
1209 1210
	 */
	if (!ubifs_compr_present(c->default_compr)) {
1211 1212
		ubifs_err("'compressor \"%s\" is not compiled in",
			  ubifs_compr_name(c->default_compr));
1213
		err = -ENOTSUPP;
1214
		goto out_free;
1215 1216
	}

1217
	err = init_constants_sb(c);
1218
	if (err)
1219
		goto out_free;
1220 1221 1222 1223 1224 1225

	sz = ALIGN(c->max_idx_node_sz, c->min_io_size);
	sz = ALIGN(sz + c->max_idx_node_sz, c->min_io_size);
	c->cbuf = kmalloc(sz, GFP_NOFS);
	if (!c->cbuf) {
		err = -ENOMEM;
1226
		goto out_free;
1227 1228
	}

1229
	sprintf(c->bgt_name, BGT_NAME_PATTERN, c->vi.ubi_num, c->vi.vol_id);
1230 1231 1232 1233 1234 1235
	if (!mounted_read_only) {
		err = alloc_wbufs(c);
		if (err)
			goto out_cbuf;

		/* Create background thread */
H
Hunter Adrian 已提交
1236
		c->bgt = kthread_create(ubifs_bg_thread, c, "%s", c->bgt_name);
1237 1238 1239 1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 1250
		if (IS_ERR(c->bgt)) {
			err = PTR_ERR(c->bgt);
			c->bgt = NULL;
			ubifs_err("cannot spawn \"%s\", error %d",
				  c->bgt_name, err);
			goto out_wbufs;
		}
		wake_up_process(c->bgt);
	}

	err = ubifs_read_master(c);
	if (err)
		goto out_master;

1251 1252
	init_constants_master(c);

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
	if ((c->mst_node->flags & cpu_to_le32(UBIFS_MST_DIRTY)) != 0) {
		ubifs_msg("recovery needed");
		c->need_recovery = 1;
		if (!mounted_read_only) {
			err = ubifs_recover_inl_heads(c, c->sbuf);
			if (err)
				goto out_master;
		}
	} else if (!mounted_read_only) {
		/*
		 * Set the "dirty" flag so that if we reboot uncleanly we
		 * will notice this immediately on the next mount.
		 */
		c->mst_node->flags |= cpu_to_le32(UBIFS_MST_DIRTY);
		err = ubifs_write_master(c);
		if (err)
			goto out_master;
	}

	err = ubifs_lpt_init(c, 1, !mounted_read_only);
	if (err)
		goto out_lpt;

	err = dbg_check_idx_size(c, c->old_idx_sz);
	if (err)
		goto out_lpt;

	err = ubifs_replay_journal(c);
	if (err)
		goto out_journal;

A
Artem Bityutskiy 已提交
1284 1285 1286
	/* Calculate 'min_idx_lebs' after journal replay */
	c->min_idx_lebs = ubifs_calc_min_idx_lebs(c);

1287 1288 1289 1290 1291 1292 1293
	err = ubifs_mount_orphans(c, c->need_recovery, mounted_read_only);
	if (err)
		goto out_orphans;

	if (!mounted_read_only) {
		int lnum;

1294 1295
		err = check_free_space(c);
		if (err)
1296 1297 1298 1299 1300 1301 1302 1303 1304 1305 1306 1307 1308 1309 1310 1311 1312
			goto out_orphans;

		/* Check for enough log space */
		lnum = c->lhead_lnum + 1;
		if (lnum >= UBIFS_LOG_LNUM + c->log_lebs)
			lnum = UBIFS_LOG_LNUM;
		if (lnum == c->ltail_lnum) {
			err = ubifs_consolidate_log(c);
			if (err)
				goto out_orphans;
		}

		if (c->need_recovery) {
			err = ubifs_recover_size(c);
			if (err)
				goto out_orphans;
			err = ubifs_rcvry_gc_commit(c);
1313
		} else {
1314
			err = take_gc_lnum(c);
1315 1316 1317 1318 1319 1320 1321 1322 1323 1324 1325
			if (err)
				goto out_orphans;

			/*
			 * GC LEB may contain garbage if there was an unclean
			 * reboot, and it should be un-mapped.
			 */
			err = ubifs_leb_unmap(c, c->gc_lnum);
			if (err)
				return err;
		}
1326 1327 1328 1329 1330 1331 1332 1333

		err = dbg_check_lprops(c);
		if (err)
			goto out_orphans;
	} else if (c->need_recovery) {
		err = ubifs_recover_size(c);
		if (err)
			goto out_orphans;
1334 1335 1336 1337 1338 1339 1340 1341 1342 1343
	} else {
		/*
		 * Even if we mount read-only, we have to set space in GC LEB
		 * to proper value because this affects UBIFS free space
		 * reporting. We do not want to have a situation when
		 * re-mounting from R/O to R/W changes amount of free space.
		 */
		err = take_gc_lnum(c);
		if (err)
			goto out_orphans;
1344 1345 1346 1347 1348 1349 1350 1351 1352 1353 1354 1355
	}

	spin_lock(&ubifs_infos_lock);
	list_add_tail(&c->infos_list, &ubifs_infos);
	spin_unlock(&ubifs_infos_lock);

	if (c->need_recovery) {
		if (mounted_read_only)
			ubifs_msg("recovery deferred");
		else {
			c->need_recovery = 0;
			ubifs_msg("recovery completed");
A
Artem Bityutskiy 已提交
1356 1357 1358 1359 1360 1361
			/*
			 * GC LEB has to be empty and taken at this point. But
			 * the journal head LEBs may also be accounted as
			 * "empty taken" if they are empty.
			 */
			ubifs_assert(c->lst.taken_empty_lebs > 0);
1362
		}
A
Artem Bityutskiy 已提交
1363
	} else
A
Artem Bityutskiy 已提交
1364
		ubifs_assert(c->lst.taken_empty_lebs > 0);
1365

A
Artem Bityutskiy 已提交
1366
	err = dbg_check_filesystem(c);
A
Artem Bityutskiy 已提交
1367 1368 1369
	if (err)
		goto out_infos;

A
Artem Bityutskiy 已提交
1370
	err = dbg_debugfs_init_fs(c);
1371 1372 1373
	if (err)
		goto out_infos;

1374 1375
	c->always_chk_crc = 0;

1376 1377
	ubifs_msg("mounted UBI device %d, volume %d, name \"%s\"",
		  c->vi.ubi_num, c->vi.vol_id, c->vi.name);
1378 1379 1380
	if (mounted_read_only)
		ubifs_msg("mounted read-only");
	x = (long long)c->main_lebs * c->leb_size;
1381 1382
	ubifs_msg("file system size:   %lld bytes (%lld KiB, %lld MiB, %d "
		  "LEBs)", x, x >> 10, x >> 20, c->main_lebs);
1383
	x = (long long)c->log_lebs * c->leb_size + c->max_bud_bytes;
1384 1385
	ubifs_msg("journal size:       %lld bytes (%lld KiB, %lld MiB, %d "
		  "LEBs)", x, x >> 10, x >> 20, c->log_lebs + c->max_bud_cnt);
A
Artem Bityutskiy 已提交
1386 1387 1388
	ubifs_msg("media format:       w%d/r%d (latest is w%d/r%d)",
		  c->fmt_version, c->ro_compat_version,
		  UBIFS_FORMAT_VERSION, UBIFS_RO_COMPAT_VERSION);
1389
	ubifs_msg("default compressor: %s", ubifs_compr_name(c->default_compr));
A
Artem Bityutskiy 已提交
1390
	ubifs_msg("reserved for root:  %llu bytes (%llu KiB)",
1391
		c->report_rp_size, c->report_rp_size >> 10);
1392 1393 1394 1395

	dbg_msg("compiled on:         " __DATE__ " at " __TIME__);
	dbg_msg("min. I/O unit size:  %d bytes", c->min_io_size);
	dbg_msg("LEB size:            %d bytes (%d KiB)",
1396
		c->leb_size, c->leb_size >> 10);
1397 1398 1399 1400 1401 1402 1403 1404 1405 1406 1407 1408 1409 1410 1411 1412 1413 1414 1415 1416 1417 1418 1419 1420
	dbg_msg("data journal heads:  %d",
		c->jhead_cnt - NONDATA_JHEADS_CNT);
	dbg_msg("UUID:                %02X%02X%02X%02X-%02X%02X"
	       "-%02X%02X-%02X%02X-%02X%02X%02X%02X%02X%02X",
	       c->uuid[0], c->uuid[1], c->uuid[2], c->uuid[3],
	       c->uuid[4], c->uuid[5], c->uuid[6], c->uuid[7],
	       c->uuid[8], c->uuid[9], c->uuid[10], c->uuid[11],
	       c->uuid[12], c->uuid[13], c->uuid[14], c->uuid[15]);
	dbg_msg("big_lpt              %d", c->big_lpt);
	dbg_msg("log LEBs:            %d (%d - %d)",
		c->log_lebs, UBIFS_LOG_LNUM, c->log_last);
	dbg_msg("LPT area LEBs:       %d (%d - %d)",
		c->lpt_lebs, c->lpt_first, c->lpt_last);
	dbg_msg("orphan area LEBs:    %d (%d - %d)",
		c->orph_lebs, c->orph_first, c->orph_last);
	dbg_msg("main area LEBs:      %d (%d - %d)",
		c->main_lebs, c->main_first, c->leb_cnt - 1);
	dbg_msg("index LEBs:          %d", c->lst.idx_lebs);
	dbg_msg("total index bytes:   %lld (%lld KiB, %lld MiB)",
		c->old_idx_sz, c->old_idx_sz >> 10, c->old_idx_sz >> 20);
	dbg_msg("key hash type:       %d", c->key_hash_type);
	dbg_msg("tree fanout:         %d", c->fanout);
	dbg_msg("reserved GC LEB:     %d", c->gc_lnum);
	dbg_msg("first main LEB:      %d", c->main_first);
1421 1422 1423 1424 1425 1426 1427 1428 1429 1430 1431
	dbg_msg("max. znode size      %d", c->max_znode_sz);
	dbg_msg("max. index node size %d", c->max_idx_node_sz);
	dbg_msg("node sizes:          data %zu, inode %zu, dentry %zu",
		UBIFS_DATA_NODE_SZ, UBIFS_INO_NODE_SZ, UBIFS_DENT_NODE_SZ);
	dbg_msg("node sizes:          trun %zu, sb %zu, master %zu",
		UBIFS_TRUN_NODE_SZ, UBIFS_SB_NODE_SZ, UBIFS_MST_NODE_SZ);
	dbg_msg("node sizes:          ref %zu, cmt. start %zu, orph %zu",
		UBIFS_REF_NODE_SZ, UBIFS_CS_NODE_SZ, UBIFS_ORPH_NODE_SZ);
	dbg_msg("max. node sizes:     data %zu, inode %zu dentry %zu",
	        UBIFS_MAX_DATA_NODE_SZ, UBIFS_MAX_INO_NODE_SZ,
		UBIFS_MAX_DENT_NODE_SZ);
1432 1433
	dbg_msg("dead watermark:      %d", c->dead_wm);
	dbg_msg("dark watermark:      %d", c->dark_wm);
1434
	dbg_msg("LEB overhead:        %d", c->leb_overhead);
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 1470
	x = (long long)c->main_lebs * c->dark_wm;
	dbg_msg("max. dark space:     %lld (%lld KiB, %lld MiB)",
		x, x >> 10, x >> 20);
	dbg_msg("maximum bud bytes:   %lld (%lld KiB, %lld MiB)",
		c->max_bud_bytes, c->max_bud_bytes >> 10,
		c->max_bud_bytes >> 20);
	dbg_msg("BG commit bud bytes: %lld (%lld KiB, %lld MiB)",
		c->bg_bud_bytes, c->bg_bud_bytes >> 10,
		c->bg_bud_bytes >> 20);
	dbg_msg("current bud bytes    %lld (%lld KiB, %lld MiB)",
		c->bud_bytes, c->bud_bytes >> 10, c->bud_bytes >> 20);
	dbg_msg("max. seq. number:    %llu", c->max_sqnum);
	dbg_msg("commit number:       %llu", c->cmt_no);

	return 0;

out_infos:
	spin_lock(&ubifs_infos_lock);
	list_del(&c->infos_list);
	spin_unlock(&ubifs_infos_lock);
out_orphans:
	free_orphans(c);
out_journal:
	destroy_journal(c);
out_lpt:
	ubifs_lpt_free(c, 0);
out_master:
	kfree(c->mst_node);
	kfree(c->rcvrd_mst_node);
	if (c->bgt)
		kthread_stop(c->bgt);
out_wbufs:
	free_wbufs(c);
out_cbuf:
	kfree(c->cbuf);
out_free:
1471
	kfree(c->bu.buf);
1472 1473 1474
	vfree(c->ileb_buf);
	vfree(c->sbuf);
	kfree(c->bottom_up_buf);
1475
	ubifs_debugging_exit(c);
1476 1477 1478 1479 1480 1481 1482 1483 1484 1485 1486 1487 1488 1489 1490 1491 1492
	return err;
}

/**
 * ubifs_umount - un-mount UBIFS file-system.
 * @c: UBIFS file-system description object
 *
 * Note, this function is called to free allocated resourced when un-mounting,
 * as well as free resources when an error occurred while we were half way
 * through mounting (error path cleanup function). So it has to make sure the
 * resource was actually allocated before freeing it.
 */
static void ubifs_umount(struct ubifs_info *c)
{
	dbg_gen("un-mounting UBI device %d, volume %d", c->vi.ubi_num,
		c->vi.vol_id);

A
Artem Bityutskiy 已提交
1493
	dbg_debugfs_exit_fs(c);
1494 1495 1496 1497 1498 1499 1500 1501 1502 1503 1504 1505 1506 1507 1508
	spin_lock(&ubifs_infos_lock);
	list_del(&c->infos_list);
	spin_unlock(&ubifs_infos_lock);

	if (c->bgt)
		kthread_stop(c->bgt);

	destroy_journal(c);
	free_wbufs(c);
	free_orphans(c);
	ubifs_lpt_free(c, 0);

	kfree(c->cbuf);
	kfree(c->rcvrd_mst_node);
	kfree(c->mst_node);
1509 1510
	kfree(c->bu.buf);
	vfree(c->ileb_buf);
1511 1512
	vfree(c->sbuf);
	kfree(c->bottom_up_buf);
1513
	ubifs_debugging_exit(c);
1514 1515 1516 1517 1518 1519 1520 1521 1522 1523 1524 1525 1526 1527
}

/**
 * ubifs_remount_rw - re-mount in read-write mode.
 * @c: UBIFS file-system description object
 *
 * UBIFS avoids allocating many unnecessary resources when mounted in read-only
 * mode. This function allocates the needed resources and re-mounts UBIFS in
 * read-write mode.
 */
static int ubifs_remount_rw(struct ubifs_info *c)
{
	int err, lnum;

A
Artem Bityutskiy 已提交
1528 1529 1530 1531 1532 1533 1534 1535 1536
	if (c->rw_incompat) {
		ubifs_err("the file-system is not R/W-compatible");
		ubifs_msg("on-flash format version is w%d/r%d, but software "
			  "only supports up to version w%d/r%d", c->fmt_version,
			  c->ro_compat_version, UBIFS_FORMAT_VERSION,
			  UBIFS_RO_COMPAT_VERSION);
		return -EROFS;
	}

1537
	mutex_lock(&c->umount_mutex);
1538
	dbg_save_space_info(c);
1539
	c->remounting_rw = 1;
1540
	c->always_chk_crc = 1;
1541

1542 1543
	err = check_free_space(c);
	if (err)
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
		goto out;

	if (c->old_leb_cnt != c->leb_cnt) {
		struct ubifs_sb_node *sup;

		sup = ubifs_read_sb_node(c);
		if (IS_ERR(sup)) {
			err = PTR_ERR(sup);
			goto out;
		}
		sup->leb_cnt = cpu_to_le32(c->leb_cnt);
		err = ubifs_write_sb_node(c, sup);
		if (err)
			goto out;
	}

	if (c->need_recovery) {
		ubifs_msg("completing deferred recovery");
		err = ubifs_write_rcvrd_mst_node(c);
		if (err)
			goto out;
		err = ubifs_recover_size(c);
		if (err)
			goto out;
		err = ubifs_clean_lebs(c, c->sbuf);
		if (err)
			goto out;
		err = ubifs_recover_inl_heads(c, c->sbuf);
		if (err)
			goto out;
1574 1575 1576 1577 1578 1579
	} else {
		/* A readonly mount is not allowed to have orphans */
		ubifs_assert(c->tot_orphans == 0);
		err = ubifs_clear_orphans(c);
		if (err)
			goto out;
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
	}

	if (!(c->mst_node->flags & cpu_to_le32(UBIFS_MST_DIRTY))) {
		c->mst_node->flags |= cpu_to_le32(UBIFS_MST_DIRTY);
		err = ubifs_write_master(c);
		if (err)
			goto out;
	}

	c->ileb_buf = vmalloc(c->leb_size);
	if (!c->ileb_buf) {
		err = -ENOMEM;
		goto out;
	}

	err = ubifs_lpt_init(c, 0, 1);
	if (err)
		goto out;

	err = alloc_wbufs(c);
	if (err)
		goto out;

	ubifs_create_buds_lists(c);

	/* Create background thread */
H
Hunter Adrian 已提交
1606
	c->bgt = kthread_create(ubifs_bg_thread, c, "%s", c->bgt_name);
1607 1608 1609 1610 1611
	if (IS_ERR(c->bgt)) {
		err = PTR_ERR(c->bgt);
		c->bgt = NULL;
		ubifs_err("cannot spawn \"%s\", error %d",
			  c->bgt_name, err);
1612
		goto out;
1613 1614 1615 1616
	}
	wake_up_process(c->bgt);

	c->orph_buf = vmalloc(c->leb_size);
1617 1618 1619 1620
	if (!c->orph_buf) {
		err = -ENOMEM;
		goto out;
	}
1621 1622 1623 1624 1625 1626 1627 1628 1629 1630 1631 1632 1633 1634

	/* Check for enough log space */
	lnum = c->lhead_lnum + 1;
	if (lnum >= UBIFS_LOG_LNUM + c->log_lebs)
		lnum = UBIFS_LOG_LNUM;
	if (lnum == c->ltail_lnum) {
		err = ubifs_consolidate_log(c);
		if (err)
			goto out;
	}

	if (c->need_recovery)
		err = ubifs_rcvry_gc_commit(c);
	else
1635
		err = ubifs_leb_unmap(c, c->gc_lnum);
1636 1637 1638 1639 1640 1641 1642 1643 1644 1645 1646
	if (err)
		goto out;

	if (c->need_recovery) {
		c->need_recovery = 0;
		ubifs_msg("deferred recovery completed");
	}

	dbg_gen("re-mounted read-write");
	c->vfs_sb->s_flags &= ~MS_RDONLY;
	c->remounting_rw = 0;
1647
	c->always_chk_crc = 0;
1648
	err = dbg_check_space_info(c);
1649
	mutex_unlock(&c->umount_mutex);
1650
	return err;
1651 1652 1653 1654 1655 1656 1657 1658 1659 1660 1661 1662 1663

out:
	vfree(c->orph_buf);
	c->orph_buf = NULL;
	if (c->bgt) {
		kthread_stop(c->bgt);
		c->bgt = NULL;
	}
	free_wbufs(c);
	vfree(c->ileb_buf);
	c->ileb_buf = NULL;
	ubifs_lpt_free(c, 1);
	c->remounting_rw = 0;
1664
	c->always_chk_crc = 0;
1665 1666 1667 1668 1669 1670 1671 1672
	mutex_unlock(&c->umount_mutex);
	return err;
}

/**
 * ubifs_remount_ro - re-mount in read-only mode.
 * @c: UBIFS file-system description object
 *
1673 1674
 * We assume VFS has stopped writing. Possibly the background thread could be
 * running a commit, however kthread_stop will wait in that case.
1675 1676 1677 1678 1679 1680
 */
static void ubifs_remount_ro(struct ubifs_info *c)
{
	int i, err;

	ubifs_assert(!c->need_recovery);
A
Artem Bityutskiy 已提交
1681
	ubifs_assert(!(c->vfs_sb->s_flags & MS_RDONLY));
A
Artem Bityutskiy 已提交
1682

1683 1684 1685 1686 1687 1688
	mutex_lock(&c->umount_mutex);
	if (c->bgt) {
		kthread_stop(c->bgt);
		c->bgt = NULL;
	}

1689 1690
	dbg_save_space_info(c);

1691 1692
	for (i = 0; i < c->jhead_cnt; i++) {
		ubifs_wbuf_sync(&c->jheads[i].wbuf);
A
Artem Bityutskiy 已提交
1693
		hrtimer_cancel(&c->jheads[i].wbuf.timer);
1694 1695
	}

A
Artem Bityutskiy 已提交
1696 1697 1698 1699 1700 1701 1702
	c->mst_node->flags &= ~cpu_to_le32(UBIFS_MST_DIRTY);
	c->mst_node->flags |= cpu_to_le32(UBIFS_MST_NO_ORPHS);
	c->mst_node->gc_lnum = cpu_to_le32(c->gc_lnum);
	err = ubifs_write_master(c);
	if (err)
		ubifs_ro_mode(c, err);

1703 1704 1705 1706 1707 1708
	free_wbufs(c);
	vfree(c->orph_buf);
	c->orph_buf = NULL;
	vfree(c->ileb_buf);
	c->ileb_buf = NULL;
	ubifs_lpt_free(c, 1);
1709 1710 1711
	err = dbg_check_space_info(c);
	if (err)
		ubifs_ro_mode(c, err);
1712 1713 1714 1715 1716 1717 1718 1719 1720 1721
	mutex_unlock(&c->umount_mutex);
}

static void ubifs_put_super(struct super_block *sb)
{
	int i;
	struct ubifs_info *c = sb->s_fs_info;

	ubifs_msg("un-mount UBI device %d, volume %d", c->vi.ubi_num,
		  c->vi.vol_id);
1722 1723 1724

	lock_kernel();

1725 1726 1727 1728 1729 1730 1731
	/*
	 * The following asserts are only valid if there has not been a failure
	 * of the media. For example, there will be dirty inodes if we failed
	 * to write them back because of I/O errors.
	 */
	ubifs_assert(atomic_long_read(&c->dirty_pg_cnt) == 0);
	ubifs_assert(c->budg_idx_growth == 0);
A
Artem Bityutskiy 已提交
1732
	ubifs_assert(c->budg_dd_growth == 0);
1733 1734 1735 1736 1737 1738 1739 1740 1741 1742 1743 1744 1745 1746 1747 1748 1749 1750 1751 1752 1753
	ubifs_assert(c->budg_data_growth == 0);

	/*
	 * The 'c->umount_lock' prevents races between UBIFS memory shrinker
	 * and file system un-mount. Namely, it prevents the shrinker from
	 * picking this superblock for shrinking - it will be just skipped if
	 * the mutex is locked.
	 */
	mutex_lock(&c->umount_mutex);
	if (!(c->vfs_sb->s_flags & MS_RDONLY)) {
		/*
		 * First of all kill the background thread to make sure it does
		 * not interfere with un-mounting and freeing resources.
		 */
		if (c->bgt) {
			kthread_stop(c->bgt);
			c->bgt = NULL;
		}

		/* Synchronize write-buffers */
		if (c->jheads)
1754
			for (i = 0; i < c->jhead_cnt; i++)
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
				ubifs_wbuf_sync(&c->jheads[i].wbuf);

		/*
		 * On fatal errors c->ro_media is set to 1, in which case we do
		 * not write the master node.
		 */
		if (!c->ro_media) {
			/*
			 * We are being cleanly unmounted which means the
			 * orphans were killed - indicate this in the master
			 * node. Also save the reserved GC LEB number.
			 */
			int err;

			c->mst_node->flags &= ~cpu_to_le32(UBIFS_MST_DIRTY);
			c->mst_node->flags |= cpu_to_le32(UBIFS_MST_NO_ORPHS);
			c->mst_node->gc_lnum = cpu_to_le32(c->gc_lnum);
			err = ubifs_write_master(c);
			if (err)
				/*
				 * Recovery will attempt to fix the master area
				 * next mount, so we just print a message and
				 * continue to unmount normally.
				 */
				ubifs_err("failed to write master node, "
					  "error %d", err);
		}
	}

	ubifs_umount(c);
	bdi_destroy(&c->bdi);
	ubi_close_volume(c->ubi);
	mutex_unlock(&c->umount_mutex);
	kfree(c);
1789 1790

	unlock_kernel();
1791 1792 1793 1794 1795 1796 1797 1798 1799 1800 1801 1802 1803 1804
}

static int ubifs_remount_fs(struct super_block *sb, int *flags, char *data)
{
	int err;
	struct ubifs_info *c = sb->s_fs_info;

	dbg_gen("old flags %#lx, new flags %#x", sb->s_flags, *flags);

	err = ubifs_parse_options(c, data, 1);
	if (err) {
		ubifs_err("invalid or unknown remount parameter");
		return err;
	}
1805

1806
	lock_kernel();
1807
	if ((sb->s_flags & MS_RDONLY) && !(*flags & MS_RDONLY)) {
A
Artem Bityutskiy 已提交
1808
		if (c->ro_media) {
A
Adrian Hunter 已提交
1809
			ubifs_msg("cannot re-mount due to prior errors");
1810
			unlock_kernel();
1811
			return -EROFS;
A
Artem Bityutskiy 已提交
1812
		}
1813
		err = ubifs_remount_rw(c);
1814 1815
		if (err) {
			unlock_kernel();
1816
			return err;
1817
		}
A
Adrian Hunter 已提交
1818 1819 1820
	} else if (!(sb->s_flags & MS_RDONLY) && (*flags & MS_RDONLY)) {
		if (c->ro_media) {
			ubifs_msg("cannot re-mount due to prior errors");
1821
			unlock_kernel();
1822
			return -EROFS;
A
Adrian Hunter 已提交
1823
		}
1824
		ubifs_remount_ro(c);
A
Adrian Hunter 已提交
1825
	}
1826

1827 1828 1829 1830 1831 1832 1833 1834
	if (c->bulk_read == 1)
		bu_init(c);
	else {
		dbg_gen("disable bulk-read");
		kfree(c->bu.buf);
		c->bu.buf = NULL;
	}

A
Artem Bityutskiy 已提交
1835
	ubifs_assert(c->lst.taken_empty_lebs > 0);
1836
	unlock_kernel();
1837 1838 1839
	return 0;
}

A
Artem Bityutskiy 已提交
1840
const struct super_operations ubifs_super_operations = {
1841 1842 1843 1844 1845 1846 1847 1848 1849 1850 1851 1852 1853 1854 1855 1856 1857 1858 1859 1860 1861 1862 1863 1864 1865 1866 1867 1868 1869 1870 1871 1872 1873 1874 1875 1876 1877 1878 1879 1880 1881 1882 1883 1884 1885 1886 1887 1888 1889 1890 1891 1892 1893 1894 1895 1896 1897 1898 1899 1900 1901 1902 1903 1904 1905 1906 1907 1908 1909 1910 1911 1912 1913 1914 1915 1916 1917 1918 1919 1920 1921 1922 1923 1924 1925 1926
	.alloc_inode   = ubifs_alloc_inode,
	.destroy_inode = ubifs_destroy_inode,
	.put_super     = ubifs_put_super,
	.write_inode   = ubifs_write_inode,
	.delete_inode  = ubifs_delete_inode,
	.statfs        = ubifs_statfs,
	.dirty_inode   = ubifs_dirty_inode,
	.remount_fs    = ubifs_remount_fs,
	.show_options  = ubifs_show_options,
	.sync_fs       = ubifs_sync_fs,
};

/**
 * open_ubi - parse UBI device name string and open the UBI device.
 * @name: UBI volume name
 * @mode: UBI volume open mode
 *
 * There are several ways to specify UBI volumes when mounting UBIFS:
 * o ubiX_Y    - UBI device number X, volume Y;
 * o ubiY      - UBI device number 0, volume Y;
 * o ubiX:NAME - mount UBI device X, volume with name NAME;
 * o ubi:NAME  - mount UBI device 0, volume with name NAME.
 *
 * Alternative '!' separator may be used instead of ':' (because some shells
 * like busybox may interpret ':' as an NFS host name separator). This function
 * returns ubi volume object in case of success and a negative error code in
 * case of failure.
 */
static struct ubi_volume_desc *open_ubi(const char *name, int mode)
{
	int dev, vol;
	char *endptr;

	if (name[0] != 'u' || name[1] != 'b' || name[2] != 'i')
		return ERR_PTR(-EINVAL);

	/* ubi:NAME method */
	if ((name[3] == ':' || name[3] == '!') && name[4] != '\0')
		return ubi_open_volume_nm(0, name + 4, mode);

	if (!isdigit(name[3]))
		return ERR_PTR(-EINVAL);

	dev = simple_strtoul(name + 3, &endptr, 0);

	/* ubiY method */
	if (*endptr == '\0')
		return ubi_open_volume(0, dev, mode);

	/* ubiX_Y method */
	if (*endptr == '_' && isdigit(endptr[1])) {
		vol = simple_strtoul(endptr + 1, &endptr, 0);
		if (*endptr != '\0')
			return ERR_PTR(-EINVAL);
		return ubi_open_volume(dev, vol, mode);
	}

	/* ubiX:NAME method */
	if ((*endptr == ':' || *endptr == '!') && endptr[1] != '\0')
		return ubi_open_volume_nm(dev, ++endptr, mode);

	return ERR_PTR(-EINVAL);
}

static int ubifs_fill_super(struct super_block *sb, void *data, int silent)
{
	struct ubi_volume_desc *ubi = sb->s_fs_info;
	struct ubifs_info *c;
	struct inode *root;
	int err;

	c = kzalloc(sizeof(struct ubifs_info), GFP_KERNEL);
	if (!c)
		return -ENOMEM;

	spin_lock_init(&c->cnt_lock);
	spin_lock_init(&c->cs_lock);
	spin_lock_init(&c->buds_lock);
	spin_lock_init(&c->space_lock);
	spin_lock_init(&c->orphan_lock);
	init_rwsem(&c->commit_sem);
	mutex_init(&c->lp_mutex);
	mutex_init(&c->tnc_mutex);
	mutex_init(&c->log_mutex);
	mutex_init(&c->mst_mutex);
	mutex_init(&c->umount_mutex);
1927
	mutex_init(&c->bu_mutex);
1928 1929 1930 1931 1932 1933 1934 1935 1936 1937 1938 1939 1940 1941 1942 1943 1944 1945
	init_waitqueue_head(&c->cmt_wq);
	c->buds = RB_ROOT;
	c->old_idx = RB_ROOT;
	c->size_tree = RB_ROOT;
	c->orph_tree = RB_ROOT;
	INIT_LIST_HEAD(&c->infos_list);
	INIT_LIST_HEAD(&c->idx_gc);
	INIT_LIST_HEAD(&c->replay_list);
	INIT_LIST_HEAD(&c->replay_buds);
	INIT_LIST_HEAD(&c->uncat_list);
	INIT_LIST_HEAD(&c->empty_list);
	INIT_LIST_HEAD(&c->freeable_list);
	INIT_LIST_HEAD(&c->frdi_idx_list);
	INIT_LIST_HEAD(&c->unclean_leb_list);
	INIT_LIST_HEAD(&c->old_buds);
	INIT_LIST_HEAD(&c->orph_list);
	INIT_LIST_HEAD(&c->orph_new);

1946
	c->vfs_sb = sb;
1947 1948 1949 1950 1951 1952 1953 1954 1955 1956 1957 1958 1959 1960
	c->highest_inum = UBIFS_FIRST_INO;
	c->lhead_lnum = c->ltail_lnum = UBIFS_LOG_LNUM;

	ubi_get_volume_info(ubi, &c->vi);
	ubi_get_device_info(c->vi.ubi_num, &c->di);

	/* Re-open the UBI device in read-write mode */
	c->ubi = ubi_open_volume(c->vi.ubi_num, c->vi.vol_id, UBI_READWRITE);
	if (IS_ERR(c->ubi)) {
		err = PTR_ERR(c->ubi);
		goto out_free;
	}

	/*
A
Artem Bityutskiy 已提交
1961
	 * UBIFS provides 'backing_dev_info' in order to disable read-ahead. For
1962 1963
	 * UBIFS, I/O is not deferred, it is done immediately in readpage,
	 * which means the user would have to wait not just for their own I/O
A
Artem Bityutskiy 已提交
1964
	 * but the read-ahead I/O as well i.e. completely pointless.
1965 1966 1967 1968 1969 1970 1971 1972
	 *
	 * Read-ahead will be disabled because @c->bdi.ra_pages is 0.
	 */
	c->bdi.capabilities = BDI_CAP_MAP_COPY;
	c->bdi.unplug_io_fn = default_unplug_io_fn;
	err  = bdi_init(&c->bdi);
	if (err)
		goto out_close;
1973 1974
	err = bdi_register(&c->bdi, NULL, "ubifs_%d_%d",
			   c->vi.ubi_num, c->vi.vol_id);
J
Jens Axboe 已提交
1975 1976
	if (err)
		goto out_bdi;
1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029

	err = ubifs_parse_options(c, data, 0);
	if (err)
		goto out_bdi;

	sb->s_fs_info = c;
	sb->s_magic = UBIFS_SUPER_MAGIC;
	sb->s_blocksize = UBIFS_BLOCK_SIZE;
	sb->s_blocksize_bits = UBIFS_BLOCK_SHIFT;
	sb->s_maxbytes = c->max_inode_sz = key_max_inode_size(c);
	if (c->max_inode_sz > MAX_LFS_FILESIZE)
		sb->s_maxbytes = c->max_inode_sz = MAX_LFS_FILESIZE;
	sb->s_op = &ubifs_super_operations;

	mutex_lock(&c->umount_mutex);
	err = mount_ubifs(c);
	if (err) {
		ubifs_assert(err < 0);
		goto out_unlock;
	}

	/* Read the root inode */
	root = ubifs_iget(sb, UBIFS_ROOT_INO);
	if (IS_ERR(root)) {
		err = PTR_ERR(root);
		goto out_umount;
	}

	sb->s_root = d_alloc_root(root);
	if (!sb->s_root)
		goto out_iput;

	mutex_unlock(&c->umount_mutex);
	return 0;

out_iput:
	iput(root);
out_umount:
	ubifs_umount(c);
out_unlock:
	mutex_unlock(&c->umount_mutex);
out_bdi:
	bdi_destroy(&c->bdi);
out_close:
	ubi_close_volume(c->ubi);
out_free:
	kfree(c);
	return err;
}

static int sb_test(struct super_block *sb, void *data)
{
	dev_t *dev = data;
A
Artem Bityutskiy 已提交
2030
	struct ubifs_info *c = sb->s_fs_info;
2031

A
Artem Bityutskiy 已提交
2032
	return c->vi.cdev == *dev;
2033 2034 2035 2036 2037 2038 2039 2040 2041 2042 2043 2044 2045 2046 2047 2048 2049 2050 2051 2052 2053 2054 2055 2056 2057 2058 2059
}

static int ubifs_get_sb(struct file_system_type *fs_type, int flags,
			const char *name, void *data, struct vfsmount *mnt)
{
	struct ubi_volume_desc *ubi;
	struct ubi_volume_info vi;
	struct super_block *sb;
	int err;

	dbg_gen("name %s, flags %#x", name, flags);

	/*
	 * Get UBI device number and volume ID. Mount it read-only so far
	 * because this might be a new mount point, and UBI allows only one
	 * read-write user at a time.
	 */
	ubi = open_ubi(name, UBI_READONLY);
	if (IS_ERR(ubi)) {
		ubifs_err("cannot open \"%s\", error %d",
			  name, (int)PTR_ERR(ubi));
		return PTR_ERR(ubi);
	}
	ubi_get_volume_info(ubi, &vi);

	dbg_gen("opened ubi%d_%d", vi.ubi_num, vi.vol_id);

A
Artem Bityutskiy 已提交
2060
	sb = sget(fs_type, &sb_test, &set_anon_super, &vi.cdev);
2061 2062 2063 2064 2065 2066 2067 2068 2069 2070 2071 2072 2073 2074 2075 2076 2077 2078 2079 2080 2081 2082 2083 2084 2085 2086 2087 2088 2089
	if (IS_ERR(sb)) {
		err = PTR_ERR(sb);
		goto out_close;
	}

	if (sb->s_root) {
		/* A new mount point for already mounted UBIFS */
		dbg_gen("this ubi volume is already mounted");
		if ((flags ^ sb->s_flags) & MS_RDONLY) {
			err = -EBUSY;
			goto out_deact;
		}
	} else {
		sb->s_flags = flags;
		/*
		 * Pass 'ubi' to 'fill_super()' in sb->s_fs_info where it is
		 * replaced by 'c'.
		 */
		sb->s_fs_info = ubi;
		err = ubifs_fill_super(sb, data, flags & MS_SILENT ? 1 : 0);
		if (err)
			goto out_deact;
		/* We do not support atime */
		sb->s_flags |= MS_ACTIVE | MS_NOATIME;
	}

	/* 'fill_super()' opens ubi again so we must close it here */
	ubi_close_volume(ubi);

2090 2091
	simple_set_mnt(mnt, sb);
	return 0;
2092 2093

out_deact:
2094
	deactivate_locked_super(sb);
2095 2096 2097 2098 2099 2100 2101 2102 2103
out_close:
	ubi_close_volume(ubi);
	return err;
}

static struct file_system_type ubifs_fs_type = {
	.name    = "ubifs",
	.owner   = THIS_MODULE,
	.get_sb  = ubifs_get_sb,
A
Artem Bityutskiy 已提交
2104
	.kill_sb = kill_anon_super,
2105 2106 2107 2108 2109
};

/*
 * Inode slab cache constructor.
 */
2110
static void inode_slab_ctor(void *obj)
2111 2112 2113 2114 2115 2116 2117 2118 2119 2120 2121 2122 2123 2124 2125 2126 2127 2128 2129 2130 2131 2132 2133 2134 2135 2136 2137 2138 2139 2140 2141 2142 2143 2144 2145 2146 2147 2148 2149 2150 2151 2152 2153 2154 2155 2156 2157 2158
{
	struct ubifs_inode *ui = obj;
	inode_init_once(&ui->vfs_inode);
}

static int __init ubifs_init(void)
{
	int err;

	BUILD_BUG_ON(sizeof(struct ubifs_ch) != 24);

	/* Make sure node sizes are 8-byte aligned */
	BUILD_BUG_ON(UBIFS_CH_SZ        & 7);
	BUILD_BUG_ON(UBIFS_INO_NODE_SZ  & 7);
	BUILD_BUG_ON(UBIFS_DENT_NODE_SZ & 7);
	BUILD_BUG_ON(UBIFS_XENT_NODE_SZ & 7);
	BUILD_BUG_ON(UBIFS_DATA_NODE_SZ & 7);
	BUILD_BUG_ON(UBIFS_TRUN_NODE_SZ & 7);
	BUILD_BUG_ON(UBIFS_SB_NODE_SZ   & 7);
	BUILD_BUG_ON(UBIFS_MST_NODE_SZ  & 7);
	BUILD_BUG_ON(UBIFS_REF_NODE_SZ  & 7);
	BUILD_BUG_ON(UBIFS_CS_NODE_SZ   & 7);
	BUILD_BUG_ON(UBIFS_ORPH_NODE_SZ & 7);

	BUILD_BUG_ON(UBIFS_MAX_DENT_NODE_SZ & 7);
	BUILD_BUG_ON(UBIFS_MAX_XENT_NODE_SZ & 7);
	BUILD_BUG_ON(UBIFS_MAX_DATA_NODE_SZ & 7);
	BUILD_BUG_ON(UBIFS_MAX_INO_NODE_SZ  & 7);
	BUILD_BUG_ON(UBIFS_MAX_NODE_SZ      & 7);
	BUILD_BUG_ON(MIN_WRITE_SZ           & 7);

	/* Check min. node size */
	BUILD_BUG_ON(UBIFS_INO_NODE_SZ  < MIN_WRITE_SZ);
	BUILD_BUG_ON(UBIFS_DENT_NODE_SZ < MIN_WRITE_SZ);
	BUILD_BUG_ON(UBIFS_XENT_NODE_SZ < MIN_WRITE_SZ);
	BUILD_BUG_ON(UBIFS_TRUN_NODE_SZ < MIN_WRITE_SZ);

	BUILD_BUG_ON(UBIFS_MAX_DENT_NODE_SZ > UBIFS_MAX_NODE_SZ);
	BUILD_BUG_ON(UBIFS_MAX_XENT_NODE_SZ > UBIFS_MAX_NODE_SZ);
	BUILD_BUG_ON(UBIFS_MAX_DATA_NODE_SZ > UBIFS_MAX_NODE_SZ);
	BUILD_BUG_ON(UBIFS_MAX_INO_NODE_SZ  > UBIFS_MAX_NODE_SZ);

	/* Defined node sizes */
	BUILD_BUG_ON(UBIFS_SB_NODE_SZ  != 4096);
	BUILD_BUG_ON(UBIFS_MST_NODE_SZ != 512);
	BUILD_BUG_ON(UBIFS_INO_NODE_SZ != 160);
	BUILD_BUG_ON(UBIFS_REF_NODE_SZ != 64);

2159 2160 2161
	/*
	 * We use 2 bit wide bit-fields to store compression type, which should
	 * be amended if more compressors are added. The bit-fields are:
2162 2163
	 * @compr_type in 'struct ubifs_inode', @default_compr in
	 * 'struct ubifs_info' and @compr_type in 'struct ubifs_mount_opts'.
2164 2165 2166
	 */
	BUILD_BUG_ON(UBIFS_COMPR_TYPES_CNT > 4);

2167 2168 2169 2170 2171 2172 2173 2174 2175 2176 2177 2178 2179 2180 2181 2182 2183 2184 2185 2186 2187 2188 2189 2190 2191 2192 2193 2194
	/*
	 * We require that PAGE_CACHE_SIZE is greater-than-or-equal-to
	 * UBIFS_BLOCK_SIZE. It is assumed that both are powers of 2.
	 */
	if (PAGE_CACHE_SIZE < UBIFS_BLOCK_SIZE) {
		ubifs_err("VFS page cache size is %u bytes, but UBIFS requires"
			  " at least 4096 bytes",
			  (unsigned int)PAGE_CACHE_SIZE);
		return -EINVAL;
	}

	err = register_filesystem(&ubifs_fs_type);
	if (err) {
		ubifs_err("cannot register file system, error %d", err);
		return err;
	}

	err = -ENOMEM;
	ubifs_inode_slab = kmem_cache_create("ubifs_inode_slab",
				sizeof(struct ubifs_inode), 0,
				SLAB_MEM_SPREAD | SLAB_RECLAIM_ACCOUNT,
				&inode_slab_ctor);
	if (!ubifs_inode_slab)
		goto out_reg;

	register_shrinker(&ubifs_shrinker_info);

	err = ubifs_compressors_init();
A
Artem Bityutskiy 已提交
2195 2196 2197 2198
	if (err)
		goto out_shrinker;

	err = dbg_debugfs_init();
2199 2200 2201 2202 2203 2204
	if (err)
		goto out_compr;

	return 0;

out_compr:
A
Artem Bityutskiy 已提交
2205 2206
	ubifs_compressors_exit();
out_shrinker:
2207 2208 2209 2210 2211 2212 2213 2214 2215 2216 2217 2218 2219 2220
	unregister_shrinker(&ubifs_shrinker_info);
	kmem_cache_destroy(ubifs_inode_slab);
out_reg:
	unregister_filesystem(&ubifs_fs_type);
	return err;
}
/* late_initcall to let compressors initialize first */
late_initcall(ubifs_init);

static void __exit ubifs_exit(void)
{
	ubifs_assert(list_empty(&ubifs_infos));
	ubifs_assert(atomic_long_read(&ubifs_clean_zn_cnt) == 0);

A
Artem Bityutskiy 已提交
2221
	dbg_debugfs_exit();
2222 2223 2224 2225 2226 2227 2228 2229 2230 2231 2232
	ubifs_compressors_exit();
	unregister_shrinker(&ubifs_shrinker_info);
	kmem_cache_destroy(ubifs_inode_slab);
	unregister_filesystem(&ubifs_fs_type);
}
module_exit(ubifs_exit);

MODULE_LICENSE("GPL");
MODULE_VERSION(__stringify(UBIFS_VERSION));
MODULE_AUTHOR("Artem Bityutskiy, Adrian Hunter");
MODULE_DESCRIPTION("UBIFS - UBI File System");