bmap.c 45.3 KB
Newer Older
D
David Teigland 已提交
1 2
/*
 * Copyright (C) Sistina Software, Inc.  1997-2003 All rights reserved.
3
 * Copyright (C) 2004-2006 Red Hat, Inc.  All rights reserved.
D
David Teigland 已提交
4 5 6
 *
 * This copyrighted material is made available to anyone wishing to use,
 * modify, copy, or redistribute it subject to the terms and conditions
7
 * of the GNU General Public License version 2.
D
David Teigland 已提交
8 9 10 11 12
 */

#include <linux/spinlock.h>
#include <linux/completion.h>
#include <linux/buffer_head.h>
13
#include <linux/blkdev.h>
14
#include <linux/gfs2_ondisk.h>
15
#include <linux/crc32.h>
D
David Teigland 已提交
16 17

#include "gfs2.h"
18
#include "incore.h"
D
David Teigland 已提交
19 20 21 22 23 24
#include "bmap.h"
#include "glock.h"
#include "inode.h"
#include "meta_io.h"
#include "quota.h"
#include "rgrp.h"
25
#include "log.h"
26
#include "super.h"
D
David Teigland 已提交
27
#include "trans.h"
28
#include "dir.h"
29
#include "util.h"
S
Steven Whitehouse 已提交
30
#include "trace_gfs2.h"
D
David Teigland 已提交
31 32 33 34 35 36

/* This doesn't need to be that large as max 64 bit pointers in a 4k
 * block is 512, so __u16 is fine for that. It saves stack space to
 * keep it small.
 */
struct metapath {
37
	struct buffer_head *mp_bh[GFS2_MAX_META_HEIGHT];
D
David Teigland 已提交
38 39 40
	__u16 mp_list[GFS2_MAX_META_HEIGHT];
};

41 42 43 44 45
/**
 * gfs2_unstuffer_page - unstuff a stuffed inode into a block cached by a page
 * @ip: the inode
 * @dibh: the dinode buffer
 * @block: the block number that was allocated
S
Steven Whitehouse 已提交
46
 * @page: The (optional) page. This is looked up if @page is NULL
47 48 49 50 51
 *
 * Returns: errno
 */

static int gfs2_unstuffer_page(struct gfs2_inode *ip, struct buffer_head *dibh,
52
			       u64 block, struct page *page)
53 54 55 56 57 58
{
	struct inode *inode = &ip->i_inode;
	struct buffer_head *bh;
	int release = 0;

	if (!page || page->index) {
59
		page = find_or_create_page(inode->i_mapping, 0, GFP_NOFS);
60 61 62 63 64 65 66
		if (!page)
			return -ENOMEM;
		release = 1;
	}

	if (!PageUptodate(page)) {
		void *kaddr = kmap(page);
67 68 69 70
		u64 dsize = i_size_read(inode);
 
		if (dsize > (dibh->b_size - sizeof(struct gfs2_dinode)))
			dsize = dibh->b_size - sizeof(struct gfs2_dinode);
71

72
		memcpy(kaddr, dibh->b_data + sizeof(struct gfs2_dinode), dsize);
73
		memset(kaddr + dsize, 0, PAGE_SIZE - dsize);
74 75 76 77 78 79
		kunmap(page);

		SetPageUptodate(page);
	}

	if (!page_has_buffers(page))
F
Fabian Frederick 已提交
80 81
		create_empty_buffers(page, BIT(inode->i_blkbits),
				     BIT(BH_Uptodate));
82 83 84 85 86 87 88

	bh = page_buffers(page);

	if (!buffer_mapped(bh))
		map_bh(bh, inode->i_sb, block);

	set_buffer_uptodate(bh);
89 90
	if (!gfs2_is_jdata(ip))
		mark_buffer_dirty(bh);
91
	if (!gfs2_is_writeback(ip))
92
		gfs2_trans_add_data(ip->i_gl, bh);
93 94 95

	if (release) {
		unlock_page(page);
96
		put_page(page);
97 98 99 100 101
	}

	return 0;
}

D
David Teigland 已提交
102 103 104
/**
 * gfs2_unstuff_dinode - Unstuff a dinode when the data has grown too big
 * @ip: The GFS2 inode to unstuff
S
Steven Whitehouse 已提交
105
 * @page: The (optional) page. This is looked up if the @page is NULL
D
David Teigland 已提交
106 107 108 109 110 111 112
 *
 * This routine unstuffs a dinode and returns it to a "normal" state such
 * that the height can be grown in the traditional way.
 *
 * Returns: errno
 */

113
int gfs2_unstuff_dinode(struct gfs2_inode *ip, struct page *page)
D
David Teigland 已提交
114 115
{
	struct buffer_head *bh, *dibh;
116
	struct gfs2_dinode *di;
117
	u64 block = 0;
118
	int isdir = gfs2_is_dir(ip);
D
David Teigland 已提交
119 120 121 122 123 124 125
	int error;

	down_write(&ip->i_rw_mutex);

	error = gfs2_meta_inode_buffer(ip, &dibh);
	if (error)
		goto out;
126

S
Steven Whitehouse 已提交
127
	if (i_size_read(&ip->i_inode)) {
D
David Teigland 已提交
128 129 130
		/* Get a free block, fill it with the stuffed data,
		   and write it out to disk */

131
		unsigned int n = 1;
132
		error = gfs2_alloc_blocks(ip, &block, &n, 0, NULL);
133 134
		if (error)
			goto out_brelse;
135
		if (isdir) {
136
			gfs2_trans_add_unrevoke(GFS2_SB(&ip->i_inode), block, 1);
137
			error = gfs2_dir_get_new_buffer(ip, block, &bh);
D
David Teigland 已提交
138 139
			if (error)
				goto out_brelse;
140
			gfs2_buffer_copy_tail(bh, sizeof(struct gfs2_meta_header),
D
David Teigland 已提交
141 142 143
					      dibh, sizeof(struct gfs2_dinode));
			brelse(bh);
		} else {
144
			error = gfs2_unstuffer_page(ip, dibh, block, page);
D
David Teigland 已提交
145 146 147 148 149 150 151
			if (error)
				goto out_brelse;
		}
	}

	/*  Set up the pointer to the new block  */

152
	gfs2_trans_add_meta(ip->i_gl, dibh);
153
	di = (struct gfs2_dinode *)dibh->b_data;
D
David Teigland 已提交
154 155
	gfs2_buffer_clear_tail(dibh, sizeof(struct gfs2_dinode));

S
Steven Whitehouse 已提交
156
	if (i_size_read(&ip->i_inode)) {
157
		*(__be64 *)(di + 1) = cpu_to_be64(block);
158 159
		gfs2_add_inode_blocks(&ip->i_inode, 1);
		di->di_blocks = cpu_to_be64(gfs2_get_inode_blocks(&ip->i_inode));
D
David Teigland 已提交
160 161
	}

162
	ip->i_height = 1;
163
	di->di_height = cpu_to_be16(1);
D
David Teigland 已提交
164

165
out_brelse:
D
David Teigland 已提交
166
	brelse(dibh);
167
out:
D
David Teigland 已提交
168 169 170 171 172 173 174
	up_write(&ip->i_rw_mutex);
	return error;
}


/**
 * find_metapath - Find path through the metadata tree
175
 * @sdp: The superblock
D
David Teigland 已提交
176 177
 * @mp: The metapath to return the result in
 * @block: The disk block to look up
178
 * @height: The pre-calculated height of the metadata tree
D
David Teigland 已提交
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
 *
 *   This routine returns a struct metapath structure that defines a path
 *   through the metadata of inode "ip" to get to block "block".
 *
 *   Example:
 *   Given:  "ip" is a height 3 file, "offset" is 101342453, and this is a
 *   filesystem with a blocksize of 4096.
 *
 *   find_metapath() would return a struct metapath structure set to:
 *   mp_offset = 101342453, mp_height = 3, mp_list[0] = 0, mp_list[1] = 48,
 *   and mp_list[2] = 165.
 *
 *   That means that in order to get to the block containing the byte at
 *   offset 101342453, we would load the indirect block pointed to by pointer
 *   0 in the dinode.  We would then load the indirect block pointed to by
 *   pointer 48 in that indirect block.  We would then load the data block
 *   pointed to by pointer 165 in that indirect block.
 *
 *             ----------------------------------------
 *             | Dinode |                             |
 *             |        |                            4|
 *             |        |0 1 2 3 4 5                 9|
 *             |        |                            6|
 *             ----------------------------------------
 *                       |
 *                       |
 *                       V
 *             ----------------------------------------
 *             | Indirect Block                       |
 *             |                                     5|
 *             |            4 4 4 4 4 5 5            1|
 *             |0           5 6 7 8 9 0 1            2|
 *             ----------------------------------------
 *                                |
 *                                |
 *                                V
 *             ----------------------------------------
 *             | Indirect Block                       |
 *             |                         1 1 1 1 1   5|
 *             |                         6 6 6 6 6   1|
 *             |0                        3 4 5 6 7   2|
 *             ----------------------------------------
 *                                           |
 *                                           |
 *                                           V
 *             ----------------------------------------
 *             | Data block containing offset         |
 *             |            101342453                 |
 *             |                                      |
 *             |                                      |
 *             ----------------------------------------
 *
 */

233 234
static void find_metapath(const struct gfs2_sbd *sdp, u64 block,
			  struct metapath *mp, unsigned int height)
D
David Teigland 已提交
235 236 237
{
	unsigned int i;

238
	for (i = height; i--;)
B
Bob Peterson 已提交
239
		mp->mp_list[i] = do_div(block, sdp->sd_inptrs);
D
David Teigland 已提交
240 241 242

}

243
static inline unsigned int metapath_branch_start(const struct metapath *mp)
244
{
245 246 247
	if (mp->mp_list[0] == 0)
		return 2;
	return 1;
248 249
}

B
Bob Peterson 已提交
250 251 252 253 254 255 256 257 258 259 260 261 262
/**
 * metaptr1 - Return the first possible metadata pointer in a metaath buffer
 * @height: The metadata height (0 = dinode)
 * @mp: The metapath
 */
static inline __be64 *metaptr1(unsigned int height, const struct metapath *mp)
{
	struct buffer_head *bh = mp->mp_bh[height];
	if (height == 0)
		return ((__be64 *)(bh->b_data + sizeof(struct gfs2_dinode)));
	return ((__be64 *)(bh->b_data + sizeof(struct gfs2_meta_header)));
}

D
David Teigland 已提交
263 264 265 266 267 268 269 270 271 272
/**
 * metapointer - Return pointer to start of metadata in a buffer
 * @height: The metadata height (0 = dinode)
 * @mp: The metapath
 *
 * Return a pointer to the block number of the next height of the metadata
 * tree given a buffer containing the pointer to the current height of the
 * metadata tree.
 */

273
static inline __be64 *metapointer(unsigned int height, const struct metapath *mp)
D
David Teigland 已提交
274
{
B
Bob Peterson 已提交
275 276
	__be64 *p = metaptr1(height, mp);
	return p + mp->mp_list[height];
D
David Teigland 已提交
277 278
}

279 280 281 282 283 284 285 286 287 288 289 290 291 292 293
static void gfs2_metapath_ra(struct gfs2_glock *gl,
			     const struct buffer_head *bh, const __be64 *pos)
{
	struct buffer_head *rabh;
	const __be64 *endp = (const __be64 *)(bh->b_data + bh->b_size);
	const __be64 *t;

	for (t = pos; t < endp; t++) {
		if (!*t)
			continue;

		rabh = gfs2_getbuf(gl, be64_to_cpu(*t), CREATE);
		if (trylock_buffer(rabh)) {
			if (!buffer_uptodate(rabh)) {
				rabh->b_end_io = end_buffer_read_sync;
294 295 296
				submit_bh(REQ_OP_READ,
					  REQ_RAHEAD | REQ_META | REQ_PRIO,
					  rabh);
297 298 299 300 301 302 303 304
				continue;
			}
			unlock_buffer(rabh);
		}
		brelse(rabh);
	}
}

B
Bob Peterson 已提交
305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321
/**
 * lookup_mp_height - helper function for lookup_metapath
 * @ip: the inode
 * @mp: the metapath
 * @h: the height which needs looking up
 */
static int lookup_mp_height(struct gfs2_inode *ip, struct metapath *mp, int h)
{
	__be64 *ptr = metapointer(h, mp);
	u64 dblock = be64_to_cpu(*ptr);

	if (!dblock)
		return h + 1;

	return gfs2_meta_indirect_buffer(ip, h + 1, dblock, &mp->mp_bh[h + 1]);
}

D
David Teigland 已提交
322
/**
323 324
 * lookup_metapath - Walk the metadata tree to a specific point
 * @ip: The inode
D
David Teigland 已提交
325 326
 * @mp: The metapath
 *
327 328 329 330 331 332 333 334
 * Assumes that the inode's buffer has already been looked up and
 * hooked onto mp->mp_bh[0] and that the metapath has been initialised
 * by find_metapath().
 *
 * If this function encounters part of the tree which has not been
 * allocated, it returns the current height of the tree at the point
 * at which it found the unallocated block. Blocks which are found are
 * added to the mp->mp_bh[] list.
D
David Teigland 已提交
335
 *
336
 * Returns: error or height of metadata tree
D
David Teigland 已提交
337 338
 */

339
static int lookup_metapath(struct gfs2_inode *ip, struct metapath *mp)
340 341 342
{
	unsigned int end_of_metadata = ip->i_height - 1;
	unsigned int x;
343
	int ret;
344 345

	for (x = 0; x < end_of_metadata; x++) {
B
Bob Peterson 已提交
346
		ret = lookup_mp_height(ip, mp, x);
347 348 349 350
		if (ret)
			return ret;
	}

351
	return ip->i_height;
352 353
}

B
Bob Peterson 已提交
354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382
/**
 * fillup_metapath - fill up buffers for the metadata path to a specific height
 * @ip: The inode
 * @mp: The metapath
 * @h: The height to which it should be mapped
 *
 * Similar to lookup_metapath, but does lookups for a range of heights
 *
 * Returns: error or height of metadata tree
 */

static int fillup_metapath(struct gfs2_inode *ip, struct metapath *mp, int h)
{
	unsigned int start_h = h - 1;
	int ret;

	if (h) {
		/* find the first buffer we need to look up. */
		while (start_h > 0 && mp->mp_bh[start_h] == NULL)
			start_h--;
		for (; start_h < h; start_h++) {
			ret = lookup_mp_height(ip, mp, start_h);
			if (ret)
				return ret;
		}
	}
	return ip->i_height;
}

383
static inline void release_metapath(struct metapath *mp)
384 385 386
{
	int i;

387 388 389 390 391
	for (i = 0; i < GFS2_MAX_META_HEIGHT; i++) {
		if (mp->mp_bh[i] == NULL)
			break;
		brelse(mp->mp_bh[i]);
	}
392 393
}

394 395 396 397 398 399 400 401 402 403 404 405 406 407 408
/**
 * gfs2_extent_length - Returns length of an extent of blocks
 * @start: Start of the buffer
 * @len: Length of the buffer in bytes
 * @ptr: Current position in the buffer
 * @limit: Max extent length to return (0 = unlimited)
 * @eob: Set to 1 if we hit "end of block"
 *
 * If the first block is zero (unallocated) it will return the number of
 * unallocated blocks in the extent, otherwise it will return the number
 * of contiguous blocks in the extent.
 *
 * Returns: The length of the extent (minimum of one block)
 */

409
static inline unsigned int gfs2_extent_length(void *start, unsigned int len, __be64 *ptr, size_t limit, int *eob)
410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429
{
	const __be64 *end = (start + len);
	const __be64 *first = ptr;
	u64 d = be64_to_cpu(*ptr);

	*eob = 0;
	do {
		ptr++;
		if (ptr >= end)
			break;
		if (limit && --limit == 0)
			break;
		if (d)
			d++;
	} while(be64_to_cpu(*ptr) == d);
	if (ptr >= end)
		*eob = 1;
	return (ptr - first);
}

430
static inline void bmap_lock(struct gfs2_inode *ip, int create)
431 432 433 434 435 436 437
{
	if (create)
		down_write(&ip->i_rw_mutex);
	else
		down_read(&ip->i_rw_mutex);
}

438
static inline void bmap_unlock(struct gfs2_inode *ip, int create)
439 440 441 442 443 444 445
{
	if (create)
		up_write(&ip->i_rw_mutex);
	else
		up_read(&ip->i_rw_mutex);
}

446 447 448 449 450 451 452 453 454 455
static inline __be64 *gfs2_indirect_init(struct metapath *mp,
					 struct gfs2_glock *gl, unsigned int i,
					 unsigned offset, u64 bn)
{
	__be64 *ptr = (__be64 *)(mp->mp_bh[i - 1]->b_data +
		       ((i > 1) ? sizeof(struct gfs2_meta_header) :
				 sizeof(struct gfs2_dinode)));
	BUG_ON(i < 1);
	BUG_ON(mp->mp_bh[i] != NULL);
	mp->mp_bh[i] = gfs2_meta_new(gl, bn);
456
	gfs2_trans_add_meta(gl, mp->mp_bh[i]);
457 458 459 460 461 462 463 464 465 466 467 468 469 470
	gfs2_metatype_set(mp->mp_bh[i], GFS2_METATYPE_IN, GFS2_FORMAT_IN);
	gfs2_buffer_clear_tail(mp->mp_bh[i], sizeof(struct gfs2_meta_header));
	ptr += offset;
	*ptr = cpu_to_be64(bn);
	return ptr;
}

enum alloc_state {
	ALLOC_DATA = 0,
	ALLOC_GROW_DEPTH = 1,
	ALLOC_GROW_HEIGHT = 2,
	/* ALLOC_UNSTUFF = 3,   TBD and rather complicated */
};

B
Bob Peterson 已提交
471 472 473 474 475 476 477
static inline unsigned int hptrs(struct gfs2_sbd *sdp, const unsigned int hgt)
{
	if (hgt)
		return sdp->sd_inptrs;
	return sdp->sd_diptrs;
}

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
/**
 * gfs2_bmap_alloc - Build a metadata tree of the requested height
 * @inode: The GFS2 inode
 * @lblock: The logical starting block of the extent
 * @bh_map: This is used to return the mapping details
 * @mp: The metapath
 * @sheight: The starting height (i.e. whats already mapped)
 * @height: The height to build to
 * @maxlen: The max number of data blocks to alloc
 *
 * In this routine we may have to alloc:
 *   i) Indirect blocks to grow the metadata tree height
 *  ii) Indirect blocks to fill in lower part of the metadata tree
 * iii) Data blocks
 *
 * The function is in two parts. The first part works out the total
 * number of blocks which we need. The second part does the actual
 * allocation asking for an extent at a time (if enough contiguous free
 * blocks are available, there will only be one request per bmap call)
 * and uses the state machine to initialise the blocks in order.
 *
 * Returns: errno on error
 */

static int gfs2_bmap_alloc(struct inode *inode, const sector_t lblock,
			   struct buffer_head *bh_map, struct metapath *mp,
			   const unsigned int sheight,
			   const unsigned int height,
506
			   const size_t maxlen)
507 508 509
{
	struct gfs2_inode *ip = GFS2_I(inode);
	struct gfs2_sbd *sdp = GFS2_SB(inode);
510
	struct super_block *sb = sdp->sd_vfs;
511 512
	struct buffer_head *dibh = mp->mp_bh[0];
	u64 bn, dblock = 0;
513
	unsigned n, i, blks, alloced = 0, iblks = 0, branch_start = 0;
514 515 516
	unsigned dblks = 0;
	unsigned ptrs_per_blk;
	const unsigned end_of_metadata = height - 1;
517
	int ret;
518 519 520 521 522 523 524 525
	int eob = 0;
	enum alloc_state state;
	__be64 *ptr;
	__be64 zero_bn = 0;

	BUG_ON(sheight < 1);
	BUG_ON(dibh == NULL);

526
	gfs2_trans_add_meta(ip->i_gl, dibh);
527 528 529 530 531 532 533 534 535 536 537 538 539

	if (height == sheight) {
		struct buffer_head *bh;
		/* Bottom indirect block exists, find unalloced extent size */
		ptr = metapointer(end_of_metadata, mp);
		bh = mp->mp_bh[end_of_metadata];
		dblks = gfs2_extent_length(bh->b_data, bh->b_size, ptr, maxlen,
					   &eob);
		BUG_ON(dblks < 1);
		state = ALLOC_DATA;
	} else {
		/* Need to allocate indirect blocks */
		ptrs_per_blk = height > 1 ? sdp->sd_inptrs : sdp->sd_diptrs;
540 541
		dblks = min(maxlen, (size_t)(ptrs_per_blk -
					     mp->mp_list[end_of_metadata]));
542 543 544 545 546 547 548 549
		if (height == ip->i_height) {
			/* Writing into existing tree, extend tree down */
			iblks = height - sheight;
			state = ALLOC_GROW_DEPTH;
		} else {
			/* Building up tree height */
			state = ALLOC_GROW_HEIGHT;
			iblks = height - ip->i_height;
550 551
			branch_start = metapath_branch_start(mp);
			iblks += (height - branch_start);
552 553 554 555 556 557 558 559
		}
	}

	/* start of the second part of the function (state machine) */

	blks = dblks + iblks;
	i = sheight;
	do {
560
		int error;
561
		n = blks - alloced;
562
		error = gfs2_alloc_blocks(ip, &bn, &n, 0, NULL);
563 564
		if (error)
			return error;
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
		alloced += n;
		if (state != ALLOC_DATA || gfs2_is_jdata(ip))
			gfs2_trans_add_unrevoke(sdp, bn, n);
		switch (state) {
		/* Growing height of tree */
		case ALLOC_GROW_HEIGHT:
			if (i == 1) {
				ptr = (__be64 *)(dibh->b_data +
						 sizeof(struct gfs2_dinode));
				zero_bn = *ptr;
			}
			for (; i - 1 < height - ip->i_height && n > 0; i++, n--)
				gfs2_indirect_init(mp, ip->i_gl, i, 0, bn++);
			if (i - 1 == height - ip->i_height) {
				i--;
				gfs2_buffer_copy_tail(mp->mp_bh[i],
						sizeof(struct gfs2_meta_header),
						dibh, sizeof(struct gfs2_dinode));
				gfs2_buffer_clear_tail(dibh,
						sizeof(struct gfs2_dinode) +
						sizeof(__be64));
				ptr = (__be64 *)(mp->mp_bh[i]->b_data +
					sizeof(struct gfs2_meta_header));
				*ptr = zero_bn;
				state = ALLOC_GROW_DEPTH;
590
				for(i = branch_start; i < height; i++) {
591 592 593 594 595
					if (mp->mp_bh[i] == NULL)
						break;
					brelse(mp->mp_bh[i]);
					mp->mp_bh[i] = NULL;
				}
596
				i = branch_start;
597 598 599 600 601 602
			}
			if (n == 0)
				break;
		/* Branching from existing tree */
		case ALLOC_GROW_DEPTH:
			if (i > 1 && i < height)
603
				gfs2_trans_add_meta(ip->i_gl, mp->mp_bh[i-1]);
604 605 606 607 608 609 610 611 612 613 614
			for (; i < height && n > 0; i++, n--)
				gfs2_indirect_init(mp, ip->i_gl, i,
						   mp->mp_list[i-1], bn++);
			if (i == height)
				state = ALLOC_DATA;
			if (n == 0)
				break;
		/* Tree complete, adding data blocks */
		case ALLOC_DATA:
			BUG_ON(n > dblks);
			BUG_ON(mp->mp_bh[end_of_metadata] == NULL);
615
			gfs2_trans_add_meta(ip->i_gl, mp->mp_bh[end_of_metadata]);
616 617 618 619 620
			dblks = n;
			ptr = metapointer(end_of_metadata, mp);
			dblock = bn;
			while (n-- > 0)
				*ptr++ = cpu_to_be64(bn++);
621 622 623 624 625 626 627 628 629
			if (buffer_zeronew(bh_map)) {
				ret = sb_issue_zeroout(sb, dblock, dblks,
						       GFP_NOFS);
				if (ret) {
					fs_err(sdp,
					       "Failed to zero data buffers\n");
					clear_buffer_zeronew(bh_map);
				}
			}
630 631
			break;
		}
632
	} while ((state != ALLOC_DATA) || !dblock);
633 634 635 636 637 638 639 640 641 642

	ip->i_height = height;
	gfs2_add_inode_blocks(&ip->i_inode, alloced);
	gfs2_dinode_out(ip, mp->mp_bh[0]->b_data);
	map_bh(bh_map, inode->i_sb, dblock);
	bh_map->b_size = dblks << inode->i_blkbits;
	set_buffer_new(bh_map);
	return 0;
}

D
David Teigland 已提交
643
/**
644
 * gfs2_block_map - Map a block from an inode to a disk block
S
Steven Whitehouse 已提交
645
 * @inode: The inode
D
David Teigland 已提交
646
 * @lblock: The logical block number
647
 * @bh_map: The bh to be mapped
648
 * @create: True if its ok to alloc blocks to satify the request
D
David Teigland 已提交
649
 *
650 651 652
 * Sets buffer_mapped() if successful, sets buffer_boundary() if a
 * read of metadata will be required before the next block can be
 * mapped. Sets buffer_new() if new blocks were allocated.
D
David Teigland 已提交
653 654 655 656
 *
 * Returns: errno
 */

657 658
int gfs2_block_map(struct inode *inode, sector_t lblock,
		   struct buffer_head *bh_map, int create)
D
David Teigland 已提交
659
{
660 661
	struct gfs2_inode *ip = GFS2_I(inode);
	struct gfs2_sbd *sdp = GFS2_SB(inode);
662
	unsigned int bsize = sdp->sd_sb.sb_bsize;
663
	const size_t maxlen = bh_map->b_size >> inode->i_blkbits;
664
	const u64 *arr = sdp->sd_heightsize;
665 666 667 668 669 670 671 672
	__be64 *ptr;
	u64 size;
	struct metapath mp;
	int ret;
	int eob;
	unsigned int len;
	struct buffer_head *bh;
	u8 height;
S
Steven Whitehouse 已提交
673

674
	BUG_ON(maxlen == 0);
D
David Teigland 已提交
675

B
Bob Peterson 已提交
676
	memset(&mp, 0, sizeof(mp));
677
	bmap_lock(ip, create);
678 679 680
	clear_buffer_mapped(bh_map);
	clear_buffer_new(bh_map);
	clear_buffer_boundary(bh_map);
S
Steven Whitehouse 已提交
681
	trace_gfs2_bmap(ip, bh_map, lblock, create, 1);
682 683 684 685
	if (gfs2_is_dir(ip)) {
		bsize = sdp->sd_jbsize;
		arr = sdp->sd_jheightsize;
	}
686

687 688 689
	ret = gfs2_meta_inode_buffer(ip, &mp.mp_bh[0]);
	if (ret)
		goto out;
D
David Teigland 已提交
690

691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715
	height = ip->i_height;
	size = (lblock + 1) * bsize;
	while (size > arr[height])
		height++;
	find_metapath(sdp, lblock, &mp, height);
	ret = 1;
	if (height > ip->i_height || gfs2_is_stuffed(ip))
		goto do_alloc;
	ret = lookup_metapath(ip, &mp);
	if (ret < 0)
		goto out;
	if (ret != ip->i_height)
		goto do_alloc;
	ptr = metapointer(ip->i_height - 1, &mp);
	if (*ptr == 0)
		goto do_alloc;
	map_bh(bh_map, inode->i_sb, be64_to_cpu(*ptr));
	bh = mp.mp_bh[ip->i_height - 1];
	len = gfs2_extent_length(bh->b_data, bh->b_size, ptr, maxlen, &eob);
	bh_map->b_size = (len << inode->i_blkbits);
	if (eob)
		set_buffer_boundary(bh_map);
	ret = 0;
out:
	release_metapath(&mp);
S
Steven Whitehouse 已提交
716
	trace_gfs2_bmap(ip, bh_map, lblock, create, ret);
717 718
	bmap_unlock(ip, create);
	return ret;
719

720 721 722 723 724 725
do_alloc:
	/* All allocations are done here, firstly check create flag */
	if (!create) {
		BUG_ON(gfs2_is_stuffed(ip));
		ret = 0;
		goto out;
D
David Teigland 已提交
726
	}
727 728 729 730

	/* At this point ret is the tree depth of already allocated blocks */
	ret = gfs2_bmap_alloc(inode, lblock, bh_map, &mp, ret, height, maxlen);
	goto out;
S
Steven Whitehouse 已提交
731 732
}

733 734 735
/*
 * Deprecated: do not use in new code
 */
S
Steven Whitehouse 已提交
736 737
int gfs2_extent_map(struct inode *inode, u64 lblock, int *new, u64 *dblock, unsigned *extlen)
{
738
	struct buffer_head bh = { .b_state = 0, .b_blocknr = 0 };
739
	int ret;
S
Steven Whitehouse 已提交
740 741 742 743 744 745
	int create = *new;

	BUG_ON(!extlen);
	BUG_ON(!dblock);
	BUG_ON(!new);

F
Fabian Frederick 已提交
746
	bh.b_size = BIT(inode->i_blkbits + (create ? 0 : 5));
747
	ret = gfs2_block_map(inode, lblock, &bh, create);
748 749 750 751 752 753 754
	*extlen = bh.b_size >> inode->i_blkbits;
	*dblock = bh.b_blocknr;
	if (buffer_new(&bh))
		*new = 1;
	else
		*new = 0;
	return ret;
D
David Teigland 已提交
755 756
}

S
Steven Whitehouse 已提交
757 758 759 760 761
/**
 * gfs2_block_truncate_page - Deal with zeroing out data for truncate
 *
 * This is partly borrowed from ext3.
 */
S
Steven Whitehouse 已提交
762
static int gfs2_block_truncate_page(struct address_space *mapping, loff_t from)
S
Steven Whitehouse 已提交
763 764 765
{
	struct inode *inode = mapping->host;
	struct gfs2_inode *ip = GFS2_I(inode);
766 767
	unsigned long index = from >> PAGE_SHIFT;
	unsigned offset = from & (PAGE_SIZE-1);
S
Steven Whitehouse 已提交
768 769 770 771 772
	unsigned blocksize, iblock, length, pos;
	struct buffer_head *bh;
	struct page *page;
	int err;

773
	page = find_or_create_page(mapping, index, GFP_NOFS);
S
Steven Whitehouse 已提交
774 775 776 777 778
	if (!page)
		return 0;

	blocksize = inode->i_sb->s_blocksize;
	length = blocksize - (offset & (blocksize - 1));
779
	iblock = index << (PAGE_SHIFT - inode->i_sb->s_blocksize_bits);
S
Steven Whitehouse 已提交
780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795

	if (!page_has_buffers(page))
		create_empty_buffers(page, blocksize, 0);

	/* Find the buffer that contains "offset" */
	bh = page_buffers(page);
	pos = blocksize;
	while (offset >= pos) {
		bh = bh->b_this_page;
		iblock++;
		pos += blocksize;
	}

	err = 0;

	if (!buffer_mapped(bh)) {
796
		gfs2_block_map(inode, iblock, bh, 0);
S
Steven Whitehouse 已提交
797 798 799 800 801 802 803 804 805 806 807
		/* unmapped? It's a hole - nothing to do */
		if (!buffer_mapped(bh))
			goto unlock;
	}

	/* Ok, it's mapped. Make sure it's up-to-date */
	if (PageUptodate(page))
		set_buffer_uptodate(bh);

	if (!buffer_uptodate(bh)) {
		err = -EIO;
808
		ll_rw_block(REQ_OP_READ, 0, 1, &bh);
S
Steven Whitehouse 已提交
809 810 811 812
		wait_on_buffer(bh);
		/* Uhhuh. Read error. Complain and punt. */
		if (!buffer_uptodate(bh))
			goto unlock;
813
		err = 0;
S
Steven Whitehouse 已提交
814 815
	}

816
	if (!gfs2_is_writeback(ip))
817
		gfs2_trans_add_data(ip->i_gl, bh);
S
Steven Whitehouse 已提交
818

819
	zero_user(page, offset, length);
820
	mark_buffer_dirty(bh);
S
Steven Whitehouse 已提交
821 822
unlock:
	unlock_page(page);
823
	put_page(page);
S
Steven Whitehouse 已提交
824 825 826
	return err;
}

827 828
#define GFS2_JTRUNC_REVOKES 8192

829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850
/**
 * gfs2_journaled_truncate - Wrapper for truncate_pagecache for jdata files
 * @inode: The inode being truncated
 * @oldsize: The original (larger) size
 * @newsize: The new smaller size
 *
 * With jdata files, we have to journal a revoke for each block which is
 * truncated. As a result, we need to split this into separate transactions
 * if the number of pages being truncated gets too large.
 */

static int gfs2_journaled_truncate(struct inode *inode, u64 oldsize, u64 newsize)
{
	struct gfs2_sbd *sdp = GFS2_SB(inode);
	u64 max_chunk = GFS2_JTRUNC_REVOKES * sdp->sd_vfs->s_blocksize;
	u64 chunk;
	int error;

	while (oldsize != newsize) {
		chunk = oldsize - newsize;
		if (chunk > max_chunk)
			chunk = max_chunk;
851
		truncate_pagecache(inode, oldsize - chunk);
852 853 854 855 856 857 858 859 860 861
		oldsize -= chunk;
		gfs2_trans_end(sdp);
		error = gfs2_trans_begin(sdp, RES_DINODE, GFS2_JTRUNC_REVOKES);
		if (error)
			return error;
	}

	return 0;
}

S
Steven Whitehouse 已提交
862
static int trunc_start(struct inode *inode, u64 oldsize, u64 newsize)
D
David Teigland 已提交
863
{
S
Steven Whitehouse 已提交
864 865 866
	struct gfs2_inode *ip = GFS2_I(inode);
	struct gfs2_sbd *sdp = GFS2_SB(inode);
	struct address_space *mapping = inode->i_mapping;
D
David Teigland 已提交
867 868 869 870
	struct buffer_head *dibh;
	int journaled = gfs2_is_jdata(ip);
	int error;

871 872 873 874
	if (journaled)
		error = gfs2_trans_begin(sdp, RES_DINODE + RES_JDATA, GFS2_JTRUNC_REVOKES);
	else
		error = gfs2_trans_begin(sdp, RES_DINODE, 0);
D
David Teigland 已提交
875 876 877 878 879 880 881
	if (error)
		return error;

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

882
	gfs2_trans_add_meta(ip->i_gl, dibh);
S
Steven Whitehouse 已提交
883

D
David Teigland 已提交
884
	if (gfs2_is_stuffed(ip)) {
S
Steven Whitehouse 已提交
885
		gfs2_buffer_clear_tail(dibh, sizeof(struct gfs2_dinode) + newsize);
D
David Teigland 已提交
886
	} else {
S
Steven Whitehouse 已提交
887 888 889 890
		if (newsize & (u64)(sdp->sd_sb.sb_bsize - 1)) {
			error = gfs2_block_truncate_page(mapping, newsize);
			if (error)
				goto out_brelse;
D
David Teigland 已提交
891
		}
S
Steven Whitehouse 已提交
892
		ip->i_diskflags |= GFS2_DIF_TRUNC_IN_PROG;
D
David Teigland 已提交
893 894
	}

S
Steven Whitehouse 已提交
895
	i_size_write(inode, newsize);
896
	ip->i_inode.i_mtime = ip->i_inode.i_ctime = current_time(&ip->i_inode);
S
Steven Whitehouse 已提交
897
	gfs2_dinode_out(ip, dibh->b_data);
D
David Teigland 已提交
898

899 900 901
	if (journaled)
		error = gfs2_journaled_truncate(inode, oldsize, newsize);
	else
902
		truncate_pagecache(inode, newsize);
903 904 905 906 907 908

	if (error) {
		brelse(dibh);
		return error;
	}

S
Steven Whitehouse 已提交
909 910
out_brelse:
	brelse(dibh);
911
out:
D
David Teigland 已提交
912 913 914 915
	gfs2_trans_end(sdp);
	return error;
}

B
Bob Peterson 已提交
916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939
/**
 * sweep_bh_for_rgrps - find an rgrp in a meta buffer and free blocks therein
 * @ip: inode
 * @rg_gh: holder of resource group glock
 * @mp: current metapath fully populated with buffers
 * @btotal: place to keep count of total blocks freed
 * @hgt: height we're processing
 * @first: true if this is the first call to this function for this height
 *
 * We sweep a metadata buffer (provided by the metapath) for blocks we need to
 * free, and free them all. However, we do it one rgrp at a time. If this
 * block has references to multiple rgrps, we break it into individual
 * transactions. This allows other processes to use the rgrps while we're
 * focused on a single one, for better concurrency / performance.
 * At every transaction boundary, we rewrite the inode into the journal.
 * That way the bitmaps are kept consistent with the inode and we can recover
 * if we're interrupted by power-outages.
 *
 * Returns: 0, or return code if an error occurred.
 *          *btotal has the total number of blocks freed
 */
static int sweep_bh_for_rgrps(struct gfs2_inode *ip, struct gfs2_holder *rd_gh,
			      const struct metapath *mp, u32 *btotal, int hgt,
			      bool preserve1)
D
David Teigland 已提交
940
{
941
	struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode);
B
Bob Peterson 已提交
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
	struct gfs2_rgrpd *rgd;
	struct gfs2_trans *tr;
	struct buffer_head *bh = mp->mp_bh[hgt];
	__be64 *top, *bottom, *p;
	int blks_outside_rgrp;
	u64 bn, bstart, isize_blks;
	s64 blen; /* needs to be s64 or gfs2_add_inode_blocks breaks */
	int meta = ((hgt != ip->i_height - 1) ? 1 : 0);
	int ret = 0;
	bool buf_in_tr = false; /* buffer was added to transaction */

	if (gfs2_metatype_check(sdp, bh,
				(hgt ? GFS2_METATYPE_IN : GFS2_METATYPE_DI)))
		return -EIO;

more_rgrps:
	blks_outside_rgrp = 0;
	bstart = 0;
	blen = 0;
	top = metapointer(hgt, mp); /* first ptr from metapath */
	/* If we're keeping some data at the truncation point, we've got to
	   preserve the metadata tree by adding 1 to the starting metapath. */
	if (preserve1)
		top++;

	bottom = (__be64 *)(bh->b_data + bh->b_size);

	for (p = top; p < bottom; p++) {
		if (!*p)
			continue;
		bn = be64_to_cpu(*p);
		if (gfs2_holder_initialized(rd_gh)) {
974
			rgd = gfs2_glock2rgrp(rd_gh->gh_gl);
B
Bob Peterson 已提交
975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075
			gfs2_assert_withdraw(sdp,
				     gfs2_glock_is_locked_by_me(rd_gh->gh_gl));
		} else {
			rgd = gfs2_blk2rgrpd(sdp, bn, false);
			ret = gfs2_glock_nq_init(rgd->rd_gl, LM_ST_EXCLUSIVE,
						 0, rd_gh);
			if (ret)
				goto out;

			/* Must be done with the rgrp glock held: */
			if (gfs2_rs_active(&ip->i_res) &&
			    rgd == ip->i_res.rs_rbm.rgd)
				gfs2_rs_deltree(&ip->i_res);
		}

		if (!rgrp_contains_block(rgd, bn)) {
			blks_outside_rgrp++;
			continue;
		}

		/* The size of our transactions will be unknown until we
		   actually process all the metadata blocks that relate to
		   the rgrp. So we estimate. We know it can't be more than
		   the dinode's i_blocks and we don't want to exceed the
		   journal flush threshold, sd_log_thresh2. */
		if (current->journal_info == NULL) {
			unsigned int jblocks_rqsted, revokes;

			jblocks_rqsted = rgd->rd_length + RES_DINODE +
				RES_INDIRECT;
			isize_blks = gfs2_get_inode_blocks(&ip->i_inode);
			if (isize_blks > atomic_read(&sdp->sd_log_thresh2))
				jblocks_rqsted +=
					atomic_read(&sdp->sd_log_thresh2);
			else
				jblocks_rqsted += isize_blks;
			revokes = jblocks_rqsted;
			if (meta)
				revokes += hptrs(sdp, hgt);
			else if (ip->i_depth)
				revokes += sdp->sd_inptrs;
			ret = gfs2_trans_begin(sdp, jblocks_rqsted, revokes);
			if (ret)
				goto out_unlock;
			down_write(&ip->i_rw_mutex);
		}
		/* check if we will exceed the transaction blocks requested */
		tr = current->journal_info;
		if (tr->tr_num_buf_new + RES_STATFS +
		    RES_QUOTA >= atomic_read(&sdp->sd_log_thresh2)) {
			/* We set blks_outside_rgrp to ensure the loop will
			   be repeated for the same rgrp, but with a new
			   transaction. */
			blks_outside_rgrp++;
			/* This next part is tricky. If the buffer was added
			   to the transaction, we've already set some block
			   pointers to 0, so we better follow through and free
			   them, or we will introduce corruption (so break).
			   This may be impossible, or at least rare, but I
			   decided to cover the case regardless.

			   If the buffer was not added to the transaction
			   (this call), doing so would exceed our transaction
			   size, so we need to end the transaction and start a
			   new one (so goto). */

			if (buf_in_tr)
				break;
			goto out_unlock;
		}

		gfs2_trans_add_meta(ip->i_gl, bh);
		buf_in_tr = true;
		*p = 0;
		if (bstart + blen == bn) {
			blen++;
			continue;
		}
		if (bstart) {
			__gfs2_free_blocks(ip, bstart, (u32)blen, meta);
			(*btotal) += blen;
			gfs2_add_inode_blocks(&ip->i_inode, -blen);
		}
		bstart = bn;
		blen = 1;
	}
	if (bstart) {
		__gfs2_free_blocks(ip, bstart, (u32)blen, meta);
		(*btotal) += blen;
		gfs2_add_inode_blocks(&ip->i_inode, -blen);
	}
out_unlock:
	if (!ret && blks_outside_rgrp) { /* If buffer still has non-zero blocks
					    outside the rgrp we just processed,
					    do it all over again. */
		if (current->journal_info) {
			struct buffer_head *dibh = mp->mp_bh[0];

			/* Every transaction boundary, we rewrite the dinode
			   to keep its di_blocks current in case of failure. */
			ip->i_inode.i_mtime = ip->i_inode.i_ctime =
1076
				current_time(&ip->i_inode);
B
Bob Peterson 已提交
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
			gfs2_trans_add_meta(ip->i_gl, dibh);
			gfs2_dinode_out(ip, dibh->b_data);
			up_write(&ip->i_rw_mutex);
			gfs2_trans_end(sdp);
		}
		gfs2_glock_dq_uninit(rd_gh);
		cond_resched();
		goto more_rgrps;
	}
out:
	return ret;
}

/**
 * find_nonnull_ptr - find a non-null pointer given a metapath and height
 * assumes the metapath is valid (with buffers) out to height h
 * @mp: starting metapath
 * @h: desired height to search
 *
 * Returns: true if a non-null pointer was found in the metapath buffer
 *          false if all remaining pointers are NULL in the buffer
 */
static bool find_nonnull_ptr(struct gfs2_sbd *sdp, struct metapath *mp,
			     unsigned int h)
{
	__be64 *ptr;
	unsigned int ptrs = hptrs(sdp, h) - 1;

	while (true) {
		ptr = metapointer(h, mp);
1107 1108 1109 1110 1111 1112 1113
		if (*ptr) { /* if we have a non-null pointer */
			/* Now zero the metapath after the current height. */
			h++;
			if (h < GFS2_MAX_META_HEIGHT)
				memset(&mp->mp_list[h], 0,
				       (GFS2_MAX_META_HEIGHT - h) *
				       sizeof(mp->mp_list[0]));
B
Bob Peterson 已提交
1114
			return true;
1115
		}
B
Bob Peterson 已提交
1116 1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127 1128 1129

		if (mp->mp_list[h] < ptrs)
			mp->mp_list[h]++;
		else
			return false; /* no more pointers in this buffer */
	}
}

enum dealloc_states {
	DEALLOC_MP_FULL = 0,    /* Strip a metapath with all buffers read in */
	DEALLOC_MP_LOWER = 1,   /* lower the metapath strip height */
	DEALLOC_FILL_MP = 2,  /* Fill in the metapath to the given height. */
	DEALLOC_DONE = 3,       /* process complete */
};
D
David Teigland 已提交
1130

1131 1132 1133 1134 1135 1136 1137
static bool mp_eq_to_hgt(struct metapath *mp, __u16 *nbof, unsigned int h)
{
	if (memcmp(mp->mp_list, nbof, h * sizeof(mp->mp_list[0])))
		return false;
	return true;
}

B
Bob Peterson 已提交
1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151 1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165
/**
 * trunc_dealloc - truncate a file down to a desired size
 * @ip: inode to truncate
 * @newsize: The desired size of the file
 *
 * This function truncates a file to newsize. It works from the
 * bottom up, and from the right to the left. In other words, it strips off
 * the highest layer (data) before stripping any of the metadata. Doing it
 * this way is best in case the operation is interrupted by power failure, etc.
 * The dinode is rewritten in every transaction to guarantee integrity.
 */
static int trunc_dealloc(struct gfs2_inode *ip, u64 newsize)
{
	struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode);
	struct metapath mp;
	struct buffer_head *dibh, *bh;
	struct gfs2_holder rd_gh;
	u64 lblock;
	__u16 nbof[GFS2_MAX_META_HEIGHT]; /* new beginning of truncation */
	unsigned int strip_h = ip->i_height - 1;
	u32 btotal = 0;
	int ret, state;
	int mp_h; /* metapath buffers are read in to this height */
	sector_t last_ra = 0;
	u64 prev_bnr = 0;
	bool preserve1; /* need to preserve the first meta pointer? */

	if (!newsize)
D
David Teigland 已提交
1166
		lblock = 0;
1167
	else
B
Bob Peterson 已提交
1168
		lblock = (newsize - 1) >> sdp->sd_sb.sb_bsize_shift;
D
David Teigland 已提交
1169

B
Bob Peterson 已提交
1170
	memset(&mp, 0, sizeof(mp));
1171
	find_metapath(sdp, lblock, &mp, ip->i_height);
D
David Teigland 已提交
1172

B
Bob Peterson 已提交
1173 1174 1175 1176 1177
	memcpy(&nbof, &mp.mp_list, sizeof(nbof));

	ret = gfs2_meta_inode_buffer(ip, &dibh);
	if (ret)
		return ret;
D
David Teigland 已提交
1178

B
Bob Peterson 已提交
1179 1180 1181 1182 1183 1184
	mp.mp_bh[0] = dibh;
	ret = lookup_metapath(ip, &mp);
	if (ret == ip->i_height)
		state = DEALLOC_MP_FULL; /* We have a complete metapath */
	else
		state = DEALLOC_FILL_MP; /* deal with partial metapath */
D
David Teigland 已提交
1185

B
Bob Peterson 已提交
1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199 1200 1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214
	ret = gfs2_rindex_update(sdp);
	if (ret)
		goto out_metapath;

	ret = gfs2_quota_hold(ip, NO_UID_QUOTA_CHANGE, NO_GID_QUOTA_CHANGE);
	if (ret)
		goto out_metapath;
	gfs2_holder_mark_uninitialized(&rd_gh);

	mp_h = strip_h;

	while (state != DEALLOC_DONE) {
		switch (state) {
		/* Truncate a full metapath at the given strip height.
		 * Note that strip_h == mp_h in order to be in this state. */
		case DEALLOC_MP_FULL:
			if (mp_h > 0) { /* issue read-ahead on metadata */
				__be64 *top;

				bh = mp.mp_bh[mp_h - 1];
				if (bh->b_blocknr != last_ra) {
					last_ra = bh->b_blocknr;
					top = metaptr1(mp_h - 1, &mp);
					gfs2_metapath_ra(ip->i_gl, bh, top);
				}
			}
			/* If we're truncating to a non-zero size and the mp is
			   at the beginning of file for the strip height, we
			   need to preserve the first metadata pointer. */
1215
			preserve1 = (newsize && mp_eq_to_hgt(&mp, nbof, mp_h));
B
Bob Peterson 已提交
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 1241 1242 1243 1244 1245 1246 1247 1248 1249 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 1287 1288 1289 1290 1291 1292 1293 1294
			bh = mp.mp_bh[mp_h];
			gfs2_assert_withdraw(sdp, bh);
			if (gfs2_assert_withdraw(sdp,
						 prev_bnr != bh->b_blocknr)) {
				printk(KERN_EMERG "GFS2: fsid=%s:inode %llu, "
				       "block:%llu, i_h:%u, s_h:%u, mp_h:%u\n",
				       sdp->sd_fsname,
				       (unsigned long long)ip->i_no_addr,
				       prev_bnr, ip->i_height, strip_h, mp_h);
			}
			prev_bnr = bh->b_blocknr;
			ret = sweep_bh_for_rgrps(ip, &rd_gh, &mp, &btotal,
						 mp_h, preserve1);
			/* If we hit an error or just swept dinode buffer,
			   just exit. */
			if (ret || !mp_h) {
				state = DEALLOC_DONE;
				break;
			}
			state = DEALLOC_MP_LOWER;
			break;

		/* lower the metapath strip height */
		case DEALLOC_MP_LOWER:
			/* We're done with the current buffer, so release it,
			   unless it's the dinode buffer. Then back up to the
			   previous pointer. */
			if (mp_h) {
				brelse(mp.mp_bh[mp_h]);
				mp.mp_bh[mp_h] = NULL;
			}
			/* If we can't get any lower in height, we've stripped
			   off all we can. Next step is to back up and start
			   stripping the previous level of metadata. */
			if (mp_h == 0) {
				strip_h--;
				memcpy(&mp.mp_list, &nbof, sizeof(nbof));
				mp_h = strip_h;
				state = DEALLOC_FILL_MP;
				break;
			}
			mp.mp_list[mp_h] = 0;
			mp_h--; /* search one metadata height down */
			if (mp.mp_list[mp_h] >= hptrs(sdp, mp_h) - 1)
				break; /* loop around in the same state */
			mp.mp_list[mp_h]++;
			/* Here we've found a part of the metapath that is not
			 * allocated. We need to search at that height for the
			 * next non-null pointer. */
			if (find_nonnull_ptr(sdp, &mp, mp_h)) {
				state = DEALLOC_FILL_MP;
				mp_h++;
			}
			/* No more non-null pointers at this height. Back up
			   to the previous height and try again. */
			break; /* loop around in the same state */

		/* Fill the metapath with buffers to the given height. */
		case DEALLOC_FILL_MP:
			/* Fill the buffers out to the current height. */
			ret = fillup_metapath(ip, &mp, mp_h);
			if (ret < 0)
				goto out;

			/* If buffers found for the entire strip height */
			if ((ret == ip->i_height) && (mp_h == strip_h)) {
				state = DEALLOC_MP_FULL;
				break;
			}
			if (ret < ip->i_height) /* We have a partial height */
				mp_h = ret - 1;

			/* If we find a non-null block pointer, crawl a bit
			   higher up in the metapath and try again, otherwise
			   we need to look lower for a new starting point. */
			if (find_nonnull_ptr(sdp, &mp, mp_h))
				mp_h++;
			else
				state = DEALLOC_MP_LOWER;
D
David Teigland 已提交
1295
			break;
B
Bob Peterson 已提交
1296
		}
D
David Teigland 已提交
1297 1298
	}

B
Bob Peterson 已提交
1299 1300 1301 1302 1303 1304 1305 1306 1307 1308 1309
	if (btotal) {
		if (current->journal_info == NULL) {
			ret = gfs2_trans_begin(sdp, RES_DINODE + RES_STATFS +
					       RES_QUOTA, 0);
			if (ret)
				goto out;
			down_write(&ip->i_rw_mutex);
		}
		gfs2_statfs_change(sdp, 0, +btotal, 0);
		gfs2_quota_change(ip, -(s64)btotal, ip->i_inode.i_uid,
				  ip->i_inode.i_gid);
1310
		ip->i_inode.i_mtime = ip->i_inode.i_ctime = current_time(&ip->i_inode);
B
Bob Peterson 已提交
1311 1312 1313 1314 1315
		gfs2_trans_add_meta(ip->i_gl, dibh);
		gfs2_dinode_out(ip, dibh->b_data);
		up_write(&ip->i_rw_mutex);
		gfs2_trans_end(sdp);
	}
D
David Teigland 已提交
1316

B
Bob Peterson 已提交
1317 1318 1319 1320 1321 1322 1323 1324 1325 1326 1327 1328
out:
	if (gfs2_holder_initialized(&rd_gh))
		gfs2_glock_dq_uninit(&rd_gh);
	if (current->journal_info) {
		up_write(&ip->i_rw_mutex);
		gfs2_trans_end(sdp);
		cond_resched();
	}
	gfs2_quota_unhold(ip);
out_metapath:
	release_metapath(&mp);
	return ret;
D
David Teigland 已提交
1329 1330 1331 1332
}

static int trunc_end(struct gfs2_inode *ip)
{
1333
	struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode);
D
David Teigland 已提交
1334 1335 1336 1337 1338 1339 1340 1341 1342 1343 1344 1345 1346
	struct buffer_head *dibh;
	int error;

	error = gfs2_trans_begin(sdp, RES_DINODE, 0);
	if (error)
		return error;

	down_write(&ip->i_rw_mutex);

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

S
Steven Whitehouse 已提交
1347
	if (!i_size_read(&ip->i_inode)) {
1348
		ip->i_height = 0;
1349
		ip->i_goal = ip->i_no_addr;
D
David Teigland 已提交
1350
		gfs2_buffer_clear_tail(dibh, sizeof(struct gfs2_dinode));
1351
		gfs2_ordered_del_inode(ip);
D
David Teigland 已提交
1352
	}
1353
	ip->i_inode.i_mtime = ip->i_inode.i_ctime = current_time(&ip->i_inode);
1354
	ip->i_diskflags &= ~GFS2_DIF_TRUNC_IN_PROG;
D
David Teigland 已提交
1355

1356
	gfs2_trans_add_meta(ip->i_gl, dibh);
1357
	gfs2_dinode_out(ip, dibh->b_data);
D
David Teigland 已提交
1358 1359
	brelse(dibh);

1360
out:
D
David Teigland 已提交
1361 1362 1363 1364 1365 1366 1367
	up_write(&ip->i_rw_mutex);
	gfs2_trans_end(sdp);
	return error;
}

/**
 * do_shrink - make a file smaller
S
Steven Whitehouse 已提交
1368 1369 1370
 * @inode: the inode
 * @oldsize: the current inode size
 * @newsize: the size to make the file
D
David Teigland 已提交
1371
 *
S
Steven Whitehouse 已提交
1372 1373
 * Called with an exclusive lock on @inode. The @size must
 * be equal to or smaller than the current inode size.
D
David Teigland 已提交
1374 1375 1376 1377
 *
 * Returns: errno
 */

S
Steven Whitehouse 已提交
1378
static int do_shrink(struct inode *inode, u64 oldsize, u64 newsize)
D
David Teigland 已提交
1379
{
S
Steven Whitehouse 已提交
1380
	struct gfs2_inode *ip = GFS2_I(inode);
D
David Teigland 已提交
1381 1382
	int error;

S
Steven Whitehouse 已提交
1383
	error = trunc_start(inode, oldsize, newsize);
D
David Teigland 已提交
1384 1385
	if (error < 0)
		return error;
S
Steven Whitehouse 已提交
1386
	if (gfs2_is_stuffed(ip))
D
David Teigland 已提交
1387 1388
		return 0;

S
Steven Whitehouse 已提交
1389 1390
	error = trunc_dealloc(ip, newsize);
	if (error == 0)
D
David Teigland 已提交
1391 1392 1393 1394 1395
		error = trunc_end(ip);

	return error;
}

S
Steven Whitehouse 已提交
1396
void gfs2_trim_blocks(struct inode *inode)
W
Wendy Cheng 已提交
1397
{
S
Steven Whitehouse 已提交
1398 1399 1400 1401 1402 1403 1404 1405 1406 1407 1408 1409 1410 1411 1412 1413 1414 1415 1416 1417 1418
	u64 size = inode->i_size;
	int ret;

	ret = do_shrink(inode, size, size);
	WARN_ON(ret != 0);
}

/**
 * do_grow - Touch and update inode size
 * @inode: The inode
 * @size: The new size
 *
 * This function updates the timestamps on the inode and
 * may also increase the size of the inode. This function
 * must not be called with @size any smaller than the current
 * inode size.
 *
 * Although it is not strictly required to unstuff files here,
 * earlier versions of GFS2 have a bug in the stuffed file reading
 * code which will result in a buffer overrun if the size is larger
 * than the max stuffed file size. In order to prevent this from
L
Lucas De Marchi 已提交
1419
 * occurring, such files are unstuffed, but in other cases we can
S
Steven Whitehouse 已提交
1420 1421 1422 1423 1424 1425 1426 1427 1428
 * just update the inode size directly.
 *
 * Returns: 0 on success, or -ve on error
 */

static int do_grow(struct inode *inode, u64 size)
{
	struct gfs2_inode *ip = GFS2_I(inode);
	struct gfs2_sbd *sdp = GFS2_SB(inode);
1429
	struct gfs2_alloc_parms ap = { .target = 1, };
W
Wendy Cheng 已提交
1430 1431
	struct buffer_head *dibh;
	int error;
1432
	int unstuff = 0;
W
Wendy Cheng 已提交
1433

S
Steven Whitehouse 已提交
1434 1435
	if (gfs2_is_stuffed(ip) &&
	    (size > (sdp->sd_sb.sb_bsize - sizeof(struct gfs2_dinode)))) {
1436
		error = gfs2_quota_lock_check(ip, &ap);
S
Steven Whitehouse 已提交
1437
		if (error)
1438
			return error;
S
Steven Whitehouse 已提交
1439

1440
		error = gfs2_inplace_reserve(ip, &ap);
S
Steven Whitehouse 已提交
1441 1442
		if (error)
			goto do_grow_qunlock;
1443
		unstuff = 1;
S
Steven Whitehouse 已提交
1444 1445
	}

1446 1447 1448
	error = gfs2_trans_begin(sdp, RES_DINODE + RES_STATFS + RES_RG_BIT +
				 (sdp->sd_args.ar_quota == GFS2_QUOTA_OFF ?
				  0 : RES_QUOTA), 0);
W
Wendy Cheng 已提交
1449
	if (error)
S
Steven Whitehouse 已提交
1450
		goto do_grow_release;
W
Wendy Cheng 已提交
1451

1452
	if (unstuff) {
S
Steven Whitehouse 已提交
1453 1454 1455 1456
		error = gfs2_unstuff_dinode(ip, NULL);
		if (error)
			goto do_end_trans;
	}
W
Wendy Cheng 已提交
1457 1458 1459

	error = gfs2_meta_inode_buffer(ip, &dibh);
	if (error)
S
Steven Whitehouse 已提交
1460
		goto do_end_trans;
W
Wendy Cheng 已提交
1461

S
Steven Whitehouse 已提交
1462
	i_size_write(inode, size);
1463
	ip->i_inode.i_mtime = ip->i_inode.i_ctime = current_time(&ip->i_inode);
1464
	gfs2_trans_add_meta(ip->i_gl, dibh);
W
Wendy Cheng 已提交
1465 1466 1467
	gfs2_dinode_out(ip, dibh->b_data);
	brelse(dibh);

S
Steven Whitehouse 已提交
1468
do_end_trans:
W
Wendy Cheng 已提交
1469
	gfs2_trans_end(sdp);
S
Steven Whitehouse 已提交
1470
do_grow_release:
1471
	if (unstuff) {
S
Steven Whitehouse 已提交
1472 1473 1474 1475
		gfs2_inplace_release(ip);
do_grow_qunlock:
		gfs2_quota_unlock(ip);
	}
W
Wendy Cheng 已提交
1476 1477 1478
	return error;
}

D
David Teigland 已提交
1479
/**
S
Steven Whitehouse 已提交
1480 1481 1482
 * gfs2_setattr_size - make a file a given size
 * @inode: the inode
 * @newsize: the size to make the file
D
David Teigland 已提交
1483
 *
S
Steven Whitehouse 已提交
1484 1485 1486
 * The file size can grow, shrink, or stay the same size. This
 * is called holding i_mutex and an exclusive glock on the inode
 * in question.
D
David Teigland 已提交
1487 1488 1489 1490
 *
 * Returns: errno
 */

S
Steven Whitehouse 已提交
1491
int gfs2_setattr_size(struct inode *inode, u64 newsize)
D
David Teigland 已提交
1492
{
1493
	struct gfs2_inode *ip = GFS2_I(inode);
S
Steven Whitehouse 已提交
1494 1495
	int ret;
	u64 oldsize;
D
David Teigland 已提交
1496

S
Steven Whitehouse 已提交
1497
	BUG_ON(!S_ISREG(inode->i_mode));
D
David Teigland 已提交
1498

S
Steven Whitehouse 已提交
1499 1500 1501
	ret = inode_newsize_ok(inode, newsize);
	if (ret)
		return ret;
D
David Teigland 已提交
1502

1503 1504
	inode_dio_wait(inode);

1505
	ret = gfs2_rsqa_alloc(ip);
1506
	if (ret)
1507
		goto out;
1508

S
Steven Whitehouse 已提交
1509
	oldsize = inode->i_size;
1510 1511 1512 1513
	if (newsize >= oldsize) {
		ret = do_grow(inode, newsize);
		goto out;
	}
S
Steven Whitehouse 已提交
1514

1515 1516
	ret = do_shrink(inode, oldsize, newsize);
out:
1517
	gfs2_rsqa_delete(ip, NULL);
1518
	return ret;
D
David Teigland 已提交
1519 1520 1521 1522 1523
}

int gfs2_truncatei_resume(struct gfs2_inode *ip)
{
	int error;
S
Steven Whitehouse 已提交
1524
	error = trunc_dealloc(ip, i_size_read(&ip->i_inode));
D
David Teigland 已提交
1525 1526 1527 1528 1529 1530 1531 1532 1533 1534
	if (!error)
		error = trunc_end(ip);
	return error;
}

int gfs2_file_dealloc(struct gfs2_inode *ip)
{
	return trunc_dealloc(ip, 0);
}

1535 1536 1537 1538 1539 1540 1541 1542 1543 1544 1545 1546 1547 1548 1549 1550 1551 1552 1553 1554 1555
/**
 * gfs2_free_journal_extents - Free cached journal bmap info
 * @jd: The journal
 *
 */

void gfs2_free_journal_extents(struct gfs2_jdesc *jd)
{
	struct gfs2_journal_extent *jext;

	while(!list_empty(&jd->extent_list)) {
		jext = list_entry(jd->extent_list.next, struct gfs2_journal_extent, list);
		list_del(&jext->list);
		kfree(jext);
	}
}

/**
 * gfs2_add_jextent - Add or merge a new extent to extent cache
 * @jd: The journal descriptor
 * @lblock: The logical block at start of new extent
1556
 * @dblock: The physical block at start of new extent
1557 1558 1559 1560 1561 1562 1563 1564 1565 1566 1567 1568 1569 1570 1571 1572 1573 1574 1575 1576 1577 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 1628 1629 1630 1631 1632 1633 1634 1635 1636 1637 1638 1639 1640 1641 1642 1643 1644 1645 1646 1647 1648 1649
 * @blocks: Size of extent in fs blocks
 *
 * Returns: 0 on success or -ENOMEM
 */

static int gfs2_add_jextent(struct gfs2_jdesc *jd, u64 lblock, u64 dblock, u64 blocks)
{
	struct gfs2_journal_extent *jext;

	if (!list_empty(&jd->extent_list)) {
		jext = list_entry(jd->extent_list.prev, struct gfs2_journal_extent, list);
		if ((jext->dblock + jext->blocks) == dblock) {
			jext->blocks += blocks;
			return 0;
		}
	}

	jext = kzalloc(sizeof(struct gfs2_journal_extent), GFP_NOFS);
	if (jext == NULL)
		return -ENOMEM;
	jext->dblock = dblock;
	jext->lblock = lblock;
	jext->blocks = blocks;
	list_add_tail(&jext->list, &jd->extent_list);
	jd->nr_extents++;
	return 0;
}

/**
 * gfs2_map_journal_extents - Cache journal bmap info
 * @sdp: The super block
 * @jd: The journal to map
 *
 * Create a reusable "extent" mapping from all logical
 * blocks to all physical blocks for the given journal.  This will save
 * us time when writing journal blocks.  Most journals will have only one
 * extent that maps all their logical blocks.  That's because gfs2.mkfs
 * arranges the journal blocks sequentially to maximize performance.
 * So the extent would map the first block for the entire file length.
 * However, gfs2_jadd can happen while file activity is happening, so
 * those journals may not be sequential.  Less likely is the case where
 * the users created their own journals by mounting the metafs and
 * laying it out.  But it's still possible.  These journals might have
 * several extents.
 *
 * Returns: 0 on success, or error on failure
 */

int gfs2_map_journal_extents(struct gfs2_sbd *sdp, struct gfs2_jdesc *jd)
{
	u64 lblock = 0;
	u64 lblock_stop;
	struct gfs2_inode *ip = GFS2_I(jd->jd_inode);
	struct buffer_head bh;
	unsigned int shift = sdp->sd_sb.sb_bsize_shift;
	u64 size;
	int rc;

	lblock_stop = i_size_read(jd->jd_inode) >> shift;
	size = (lblock_stop - lblock) << shift;
	jd->nr_extents = 0;
	WARN_ON(!list_empty(&jd->extent_list));

	do {
		bh.b_state = 0;
		bh.b_blocknr = 0;
		bh.b_size = size;
		rc = gfs2_block_map(jd->jd_inode, lblock, &bh, 0);
		if (rc || !buffer_mapped(&bh))
			goto fail;
		rc = gfs2_add_jextent(jd, lblock, bh.b_blocknr, bh.b_size >> shift);
		if (rc)
			goto fail;
		size -= bh.b_size;
		lblock += (bh.b_size >> ip->i_inode.i_blkbits);
	} while(size > 0);

	fs_info(sdp, "journal %d mapped with %u extents\n", jd->jd_jid,
		jd->nr_extents);
	return 0;

fail:
	fs_warn(sdp, "error %d mapping journal %u at offset %llu (extent %u)\n",
		rc, jd->jd_jid,
		(unsigned long long)(i_size_read(jd->jd_inode) - size),
		jd->nr_extents);
	fs_warn(sdp, "bmap=%d lblock=%llu block=%llu, state=0x%08lx, size=%llu\n",
		rc, (unsigned long long)lblock, (unsigned long long)bh.b_blocknr,
		bh.b_state, (unsigned long long)bh.b_size);
	gfs2_free_journal_extents(jd);
	return rc;
}

D
David Teigland 已提交
1650 1651 1652 1653 1654 1655
/**
 * gfs2_write_alloc_required - figure out if a write will require an allocation
 * @ip: the file being written to
 * @offset: the offset to write to
 * @len: the number of bytes being written
 *
1656
 * Returns: 1 if an alloc is required, 0 otherwise
D
David Teigland 已提交
1657 1658
 */

1659
int gfs2_write_alloc_required(struct gfs2_inode *ip, u64 offset,
1660
			      unsigned int len)
D
David Teigland 已提交
1661
{
1662
	struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode);
1663 1664 1665
	struct buffer_head bh;
	unsigned int shift;
	u64 lblock, lblock_stop, size;
1666
	u64 end_of_file;
D
David Teigland 已提交
1667 1668 1669 1670 1671 1672 1673

	if (!len)
		return 0;

	if (gfs2_is_stuffed(ip)) {
		if (offset + len >
		    sdp->sd_sb.sb_bsize - sizeof(struct gfs2_dinode))
1674
			return 1;
D
David Teigland 已提交
1675 1676 1677
		return 0;
	}

1678
	shift = sdp->sd_sb.sb_bsize_shift;
1679
	BUG_ON(gfs2_is_dir(ip));
S
Steven Whitehouse 已提交
1680
	end_of_file = (i_size_read(&ip->i_inode) + sdp->sd_sb.sb_bsize - 1) >> shift;
1681 1682 1683
	lblock = offset >> shift;
	lblock_stop = (offset + len + sdp->sd_sb.sb_bsize - 1) >> shift;
	if (lblock_stop > end_of_file)
1684
		return 1;
D
David Teigland 已提交
1685

1686 1687 1688 1689 1690 1691
	size = (lblock_stop - lblock) << shift;
	do {
		bh.b_state = 0;
		bh.b_size = size;
		gfs2_block_map(&ip->i_inode, lblock, &bh, 0);
		if (!buffer_mapped(&bh))
1692
			return 1;
1693 1694 1695
		size -= bh.b_size;
		lblock += (bh.b_size >> ip->i_inode.i_blkbits);
	} while(size > 0);
D
David Teigland 已提交
1696 1697 1698 1699

	return 0;
}