bmap.c 28.5 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 13 14
 */

#include <linux/sched.h>
#include <linux/slab.h>
#include <linux/spinlock.h>
#include <linux/completion.h>
#include <linux/buffer_head.h>
15
#include <linux/gfs2_ondisk.h>
16
#include <linux/crc32.h>
17
#include <linux/lm_interface.h>
D
David Teigland 已提交
18 19

#include "gfs2.h"
20
#include "incore.h"
D
David Teigland 已提交
21 22 23 24 25 26 27
#include "bmap.h"
#include "glock.h"
#include "inode.h"
#include "meta_io.h"
#include "quota.h"
#include "rgrp.h"
#include "trans.h"
28
#include "dir.h"
29
#include "util.h"
S
Steven Whitehouse 已提交
30
#include "ops_address.h"
D
David Teigland 已提交
31 32 33 34 35 36 37 38 39 40

/* 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 {
	__u16 mp_list[GFS2_MAX_META_HEIGHT];
};

typedef int (*block_call_t) (struct gfs2_inode *ip, struct buffer_head *dibh,
41 42
			     struct buffer_head *bh, u64 *top,
			     u64 *bottom, unsigned int height,
D
David Teigland 已提交
43 44 45 46 47 48 49
			     void *data);

struct strip_mine {
	int sm_first;
	unsigned int sm_height;
};

50 51 52 53 54 55 56 57 58 59 60
/**
 * 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
 * @private: any locked page held by the caller process
 *
 * Returns: errno
 */

static int gfs2_unstuffer_page(struct gfs2_inode *ip, struct buffer_head *dibh,
61
			       u64 block, struct page *page)
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
{
	struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode);
	struct inode *inode = &ip->i_inode;
	struct buffer_head *bh;
	int release = 0;

	if (!page || page->index) {
		page = grab_cache_page(inode->i_mapping, 0);
		if (!page)
			return -ENOMEM;
		release = 1;
	}

	if (!PageUptodate(page)) {
		void *kaddr = kmap(page);

		memcpy(kaddr, dibh->b_data + sizeof(struct gfs2_dinode),
		       ip->i_di.di_size);
		memset(kaddr + ip->i_di.di_size, 0,
		       PAGE_CACHE_SIZE - ip->i_di.di_size);
		kunmap(page);

		SetPageUptodate(page);
	}

	if (!page_has_buffers(page))
		create_empty_buffers(page, 1 << inode->i_blkbits,
				     (1 << BH_Uptodate));

	bh = page_buffers(page);

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

	set_buffer_uptodate(bh);
97
	if (sdp->sd_args.ar_data == GFS2_DATA_ORDERED || gfs2_is_jdata(ip))
98 99 100 101 102 103 104 105 106 107 108
		gfs2_trans_add_bh(ip->i_gl, bh, 0);
	mark_buffer_dirty(bh);

	if (release) {
		unlock_page(page);
		page_cache_release(page);
	}

	return 0;
}

D
David Teigland 已提交
109 110 111 112 113 114 115 116 117 118 119 120
/**
 * gfs2_unstuff_dinode - Unstuff a dinode when the data has grown too big
 * @ip: The GFS2 inode to unstuff
 * @unstuffer: the routine that handles unstuffing a non-zero length file
 * @private: private data for the unstuffer
 *
 * 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
 */

121
int gfs2_unstuff_dinode(struct gfs2_inode *ip, struct page *page)
D
David Teigland 已提交
122 123
{
	struct buffer_head *bh, *dibh;
124
	u64 block = 0;
125
	int isdir = gfs2_is_dir(ip);
D
David Teigland 已提交
126 127 128 129 130 131 132 133 134 135 136 137
	int error;

	down_write(&ip->i_rw_mutex);

	error = gfs2_meta_inode_buffer(ip, &dibh);
	if (error)
		goto out;
		
	if (ip->i_di.di_size) {
		/* Get a free block, fill it with the stuffed data,
		   and write it out to disk */

138
		if (isdir) {
D
David Teigland 已提交
139 140
			block = gfs2_alloc_meta(ip);

141
			error = gfs2_dir_get_new_buffer(ip, block, &bh);
D
David Teigland 已提交
142 143 144 145 146 147 148 149 150
			if (error)
				goto out_brelse;
			gfs2_buffer_copy_tail(bh,
					      sizeof(struct gfs2_meta_header),
					      dibh, sizeof(struct gfs2_dinode));
			brelse(bh);
		} else {
			block = gfs2_alloc_data(ip);

151
			error = gfs2_unstuffer_page(ip, dibh, block, page);
D
David Teigland 已提交
152 153 154 155 156 157 158
			if (error)
				goto out_brelse;
		}
	}

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

159
	gfs2_trans_add_bh(ip->i_gl, dibh, 1);
D
David Teigland 已提交
160 161 162 163

	gfs2_buffer_clear_tail(dibh, sizeof(struct gfs2_dinode));

	if (ip->i_di.di_size) {
164
		*(u64 *)(dibh->b_data + sizeof(struct gfs2_dinode)) =
165
			cpu_to_be64(block);
D
David Teigland 已提交
166 167 168 169 170 171 172
		ip->i_di.di_blocks++;
	}

	ip->i_di.di_height = 1;

	gfs2_dinode_out(&ip->i_di, dibh->b_data);

173
out_brelse:
D
David Teigland 已提交
174
	brelse(dibh);
175
out:
D
David Teigland 已提交
176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192
	up_write(&ip->i_rw_mutex);
	return error;
}

/**
 * calc_tree_height - Calculate the height of a metadata tree
 * @ip: The GFS2 inode
 * @size: The proposed size of the file
 *
 * Work out how tall a metadata tree needs to be in order to accommodate a
 * file of a particular size. If size is less than the current size of
 * the inode, then the current size of the inode is used instead of the
 * supplied one.
 *
 * Returns: the height the tree should be
 */

193
static unsigned int calc_tree_height(struct gfs2_inode *ip, u64 size)
D
David Teigland 已提交
194
{
195
	struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode);
196
	u64 *arr;
D
David Teigland 已提交
197 198 199 200 201
	unsigned int max, height;

	if (ip->i_di.di_size > size)
		size = ip->i_di.di_size;

202
	if (gfs2_is_dir(ip)) {
D
David Teigland 已提交
203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225
		arr = sdp->sd_jheightsize;
		max = sdp->sd_max_jheight;
	} else {
		arr = sdp->sd_heightsize;
		max = sdp->sd_max_height;
	}

	for (height = 0; height < max; height++)
		if (arr[height] >= size)
			break;

	return height;
}

/**
 * build_height - Build a metadata tree of the requested height
 * @ip: The GFS2 inode
 * @height: The height to build to
 *
 *
 * Returns: errno
 */

226
static int build_height(struct inode *inode, unsigned height)
D
David Teigland 已提交
227
{
228
	struct gfs2_inode *ip = GFS2_I(inode);
229 230 231
	unsigned new_height = height - ip->i_di.di_height;
	struct buffer_head *dibh;
	struct buffer_head *blocks[GFS2_MAX_META_HEIGHT];
D
David Teigland 已提交
232
	int error;
233 234 235
	u64 *bp;
	u64 bn;
	unsigned n;
D
David Teigland 已提交
236

237 238
	if (height <= ip->i_di.di_height)
		return 0;
D
David Teigland 已提交
239

240 241 242
	error = gfs2_meta_inode_buffer(ip, &dibh);
	if (error)
		return error;
D
David Teigland 已提交
243

244 245 246 247 248 249 250 251 252 253 254
	for(n = 0; n < new_height; n++) {
		bn = gfs2_alloc_meta(ip);
		blocks[n] = gfs2_meta_new(ip->i_gl, bn);
		gfs2_trans_add_bh(ip->i_gl, blocks[n], 1);
	}
	
	n = 0;
	bn = blocks[0]->b_blocknr;
	if (new_height > 1) {
		for(; n < new_height-1; n++) {
			gfs2_metatype_set(blocks[n], GFS2_METATYPE_IN,
D
David Teigland 已提交
255
					  GFS2_FORMAT_IN);
256 257 258 259 260 261 262
			gfs2_buffer_clear_tail(blocks[n],
					       sizeof(struct gfs2_meta_header));
			bp = (u64 *)(blocks[n]->b_data +
				     sizeof(struct gfs2_meta_header));
			*bp = cpu_to_be64(blocks[n+1]->b_blocknr);
			brelse(blocks[n]);
			blocks[n] = NULL;
D
David Teigland 已提交
263 264
		}
	}
265 266 267 268 269 270 271 272 273 274 275 276 277
	gfs2_metatype_set(blocks[n], GFS2_METATYPE_IN, GFS2_FORMAT_IN);
	gfs2_buffer_copy_tail(blocks[n], sizeof(struct gfs2_meta_header),
			      dibh, sizeof(struct gfs2_dinode));
	brelse(blocks[n]);
	gfs2_trans_add_bh(ip->i_gl, dibh, 1);
	gfs2_buffer_clear_tail(dibh, sizeof(struct gfs2_dinode));
	bp = (u64 *)(dibh->b_data + sizeof(struct gfs2_dinode));
	*bp = cpu_to_be64(bn);
	ip->i_di.di_height += new_height;
	ip->i_di.di_blocks += new_height;
	gfs2_dinode_out(&ip->i_di, dibh->b_data);
	brelse(dibh);
	return error;
D
David Teigland 已提交
278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338
}

/**
 * find_metapath - Find path through the metadata tree
 * @ip: The inode pointer
 * @mp: The metapath to return the result in
 * @block: The disk block to look up
 *
 *   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                 |
 *             |                                      |
 *             |                                      |
 *             ----------------------------------------
 *
 */

339
static void find_metapath(struct gfs2_inode *ip, u64 block,
340
			  struct metapath *mp)
D
David Teigland 已提交
341
{
342
	struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode);
343
	u64 b = block;
D
David Teigland 已提交
344 345 346
	unsigned int i;

	for (i = ip->i_di.di_height; i--;)
347
		mp->mp_list[i] = do_div(b, sdp->sd_inptrs);
D
David Teigland 已提交
348 349 350 351 352 353 354 355 356 357 358 359 360 361

}

/**
 * metapointer - Return pointer to start of metadata in a buffer
 * @bh: The 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.
 */

S
Steven Whitehouse 已提交
362 363
static inline u64 *metapointer(struct buffer_head *bh, int *boundary,
			       unsigned int height, const struct metapath *mp)
D
David Teigland 已提交
364 365 366
{
	unsigned int head_size = (height > 0) ?
		sizeof(struct gfs2_meta_header) : sizeof(struct gfs2_dinode);
S
Steven Whitehouse 已提交
367 368 369
	u64 *ptr;
	*boundary = 0;
	ptr = ((u64 *)(bh->b_data + head_size)) + mp->mp_list[height];
370
	if (ptr + 1 == (u64 *)(bh->b_data + bh->b_size))
S
Steven Whitehouse 已提交
371 372
		*boundary = 1;
	return ptr;
D
David Teigland 已提交
373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390
}

/**
 * lookup_block - Get the next metadata block in metadata tree
 * @ip: The GFS2 inode
 * @bh: Buffer containing the pointers to metadata blocks
 * @height: The height of the tree (0 = dinode)
 * @mp: The metapath
 * @create: Non-zero if we may create a new meatdata block
 * @new: Used to indicate if we did create a new metadata block
 * @block: the returned disk block number
 *
 * Given a metatree, complete to a particular height, checks to see if the next
 * height of the tree exists. If not the next height of the tree is created.
 * The block number of the next height of the metadata tree is returned.
 *
 */

S
Steven Whitehouse 已提交
391 392
static int lookup_block(struct gfs2_inode *ip, struct buffer_head *bh,
			unsigned int height, struct metapath *mp, int create,
393
			int *new, u64 *block)
D
David Teigland 已提交
394
{
S
Steven Whitehouse 已提交
395
	int boundary;
396
	u64 *ptr = metapointer(bh, &boundary, height, mp);
D
David Teigland 已提交
397 398 399

	if (*ptr) {
		*block = be64_to_cpu(*ptr);
S
Steven Whitehouse 已提交
400
		return boundary;
D
David Teigland 已提交
401 402 403 404 405
	}

	*block = 0;

	if (!create)
S
Steven Whitehouse 已提交
406
		return 0;
D
David Teigland 已提交
407

S
Steven Whitehouse 已提交
408
	if (height == ip->i_di.di_height - 1 && !gfs2_is_dir(ip))
D
David Teigland 已提交
409 410 411 412
		*block = gfs2_alloc_data(ip);
	else
		*block = gfs2_alloc_meta(ip);

413
	gfs2_trans_add_bh(ip->i_gl, bh, 1);
D
David Teigland 已提交
414 415 416 417 418

	*ptr = cpu_to_be64(*block);
	ip->i_di.di_blocks++;

	*new = 1;
S
Steven Whitehouse 已提交
419
	return 0;
D
David Teigland 已提交
420 421 422
}

/**
S
Steven Whitehouse 已提交
423 424
 * gfs2_block_pointers - Map a block from an inode to a disk block
 * @inode: The inode
D
David Teigland 已提交
425
 * @lblock: The logical block number
426
 * @map_bh: The bh to be mapped
S
Steven Whitehouse 已提交
427
 * @mp: metapath to use
D
David Teigland 已提交
428 429 430 431 432 433 434
 *
 * Find the block number on the current device which corresponds to an
 * inode's block. If the block had to be created, "new" will be set.
 *
 * Returns: errno
 */

435 436 437
static int gfs2_block_pointers(struct inode *inode, u64 lblock, int create,
			       struct buffer_head *bh_map, struct metapath *mp,
			       unsigned int maxlen)
D
David Teigland 已提交
438
{
439 440
	struct gfs2_inode *ip = GFS2_I(inode);
	struct gfs2_sbd *sdp = GFS2_SB(inode);
D
David Teigland 已提交
441 442 443 444 445 446
	struct buffer_head *bh;
	unsigned int bsize;
	unsigned int height;
	unsigned int end_of_metadata;
	unsigned int x;
	int error = 0;
447 448 449
	int new = 0;
	u64 dblock = 0;
	int boundary;
D
David Teigland 已提交
450 451

	if (gfs2_assert_warn(sdp, !gfs2_is_stuffed(ip)))
452
		return 0;
D
David Teigland 已提交
453

454
	bsize = gfs2_is_dir(ip) ? sdp->sd_jbsize : sdp->sd_sb.sb_bsize;
D
David Teigland 已提交
455 456 457 458

	height = calc_tree_height(ip, (lblock + 1) * bsize);
	if (ip->i_di.di_height < height) {
		if (!create)
459
			return 0;
D
David Teigland 已提交
460

461
		error = build_height(inode, height);
D
David Teigland 已提交
462
		if (error)
463
			return error;
D
David Teigland 已提交
464 465
	}

S
Steven Whitehouse 已提交
466
	find_metapath(ip, lblock, mp);
D
David Teigland 已提交
467 468 469 470
	end_of_metadata = ip->i_di.di_height - 1;

	error = gfs2_meta_inode_buffer(ip, &bh);
	if (error)
471
		return error;
D
David Teigland 已提交
472 473

	for (x = 0; x < end_of_metadata; x++) {
474
		lookup_block(ip, bh, x, mp, create, &new, &dblock);
D
David Teigland 已提交
475
		brelse(bh);
476 477
		if (!dblock)
			return 0;
D
David Teigland 已提交
478

479
		error = gfs2_meta_indirect_buffer(ip, x+1, dblock, new, &bh);
D
David Teigland 已提交
480
		if (error)
481
			return error;
D
David Teigland 已提交
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
	boundary = lookup_block(ip, bh, end_of_metadata, mp, create, &new, &dblock);
	clear_buffer_mapped(bh_map);
	clear_buffer_new(bh_map);
	clear_buffer_boundary(bh_map);

	if (dblock) {
		map_bh(bh_map, inode->i_sb, dblock);
		if (boundary)
			set_buffer_boundary(bh);
		if (new) {
			struct buffer_head *dibh;
			error = gfs2_meta_inode_buffer(ip, &dibh);
			if (!error) {
				gfs2_trans_add_bh(ip->i_gl, dibh, 1);
				gfs2_dinode_out(&ip->i_di, dibh->b_data);
				brelse(dibh);
			}
			set_buffer_new(bh_map);
			goto out_brelse;
		}
		while(--maxlen && !buffer_boundary(bh_map)) {
			u64 eblock;

			mp->mp_list[end_of_metadata]++;
			boundary = lookup_block(ip, bh, end_of_metadata, mp, 0, &new, &eblock);
			if (eblock != ++dblock)
				break;
511
			bh_map->b_size += (1 << inode->i_blkbits);
512 513
			if (boundary)
				set_buffer_boundary(bh_map);
D
David Teigland 已提交
514 515
		}
	}
516 517 518
out_brelse:
	brelse(bh);
	return 0;
S
Steven Whitehouse 已提交
519
}
D
David Teigland 已提交
520

S
Steven Whitehouse 已提交
521 522 523

static inline void bmap_lock(struct inode *inode, int create)
{
524
	struct gfs2_inode *ip = GFS2_I(inode);
S
Steven Whitehouse 已提交
525 526 527 528 529 530 531 532
	if (create)
		down_write(&ip->i_rw_mutex);
	else
		down_read(&ip->i_rw_mutex);
}

static inline void bmap_unlock(struct inode *inode, int create)
{
533
	struct gfs2_inode *ip = GFS2_I(inode);
D
David Teigland 已提交
534 535 536 537
	if (create)
		up_write(&ip->i_rw_mutex);
	else
		up_read(&ip->i_rw_mutex);
S
Steven Whitehouse 已提交
538
}
D
David Teigland 已提交
539

540 541
int gfs2_block_map(struct inode *inode, u64 lblock, int create,
		   struct buffer_head *bh, unsigned int maxlen)
S
Steven Whitehouse 已提交
542 543
{
	struct metapath mp;
544
	int ret;
S
Steven Whitehouse 已提交
545 546

	bmap_lock(inode, create);
547
	ret = gfs2_block_pointers(inode, lblock, create, bh, &mp, maxlen);
S
Steven Whitehouse 已提交
548
	bmap_unlock(inode, create);
549
	return ret;
S
Steven Whitehouse 已提交
550 551 552 553 554
}

int gfs2_extent_map(struct inode *inode, u64 lblock, int *new, u64 *dblock, unsigned *extlen)
{
	struct metapath mp;
555 556
	struct buffer_head bh = { .b_state = 0, .b_blocknr = 0, .b_size = 0 };
	int ret;
S
Steven Whitehouse 已提交
557 558 559 560 561 562 563
	int create = *new;

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

	bmap_lock(inode, create);
564
	ret = gfs2_block_pointers(inode, lblock, create, &bh, &mp, *extlen);
S
Steven Whitehouse 已提交
565
	bmap_unlock(inode, create);
566 567 568 569 570 571 572
	*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
573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593
}

/**
 * recursive_scan - recursively scan through the end of a file
 * @ip: the inode
 * @dibh: the dinode buffer
 * @mp: the path through the metadata to the point to start
 * @height: the height the recursion is at
 * @block: the indirect block to look at
 * @first: 1 if this is the first block
 * @bc: the call to make for each piece of metadata
 * @data: data opaque to this function to pass to @bc
 *
 * When this is first called @height and @block should be zero and
 * @first should be 1.
 *
 * Returns: errno
 */

static int recursive_scan(struct gfs2_inode *ip, struct buffer_head *dibh,
			  struct metapath *mp, unsigned int height,
594
			  u64 block, int first, block_call_t bc,
D
David Teigland 已提交
595 596
			  void *data)
{
597
	struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode);
D
David Teigland 已提交
598
	struct buffer_head *bh = NULL;
599 600
	u64 *top, *bottom;
	u64 bn;
D
David Teigland 已提交
601 602 603 604 605 606 607 608 609
	int error;
	int mh_size = sizeof(struct gfs2_meta_header);

	if (!height) {
		error = gfs2_meta_inode_buffer(ip, &bh);
		if (error)
			return error;
		dibh = bh;

610 611
		top = (u64 *)(bh->b_data + sizeof(struct gfs2_dinode)) + mp->mp_list[0];
		bottom = (u64 *)(bh->b_data + sizeof(struct gfs2_dinode)) + sdp->sd_diptrs;
D
David Teigland 已提交
612 613 614 615 616
	} else {
		error = gfs2_meta_indirect_buffer(ip, height, block, 0, &bh);
		if (error)
			return error;

617
		top = (u64 *)(bh->b_data + mh_size) +
J
Jan Engelhardt 已提交
618
				  (first ? mp->mp_list[height] : 0);
D
David Teigland 已提交
619

620
		bottom = (u64 *)(bh->b_data + mh_size) + sdp->sd_inptrs;
D
David Teigland 已提交
621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639
	}

	error = bc(ip, dibh, bh, top, bottom, height, data);
	if (error)
		goto out;

	if (height < ip->i_di.di_height - 1)
		for (; top < bottom; top++, first = 0) {
			if (!*top)
				continue;

			bn = be64_to_cpu(*top);

			error = recursive_scan(ip, dibh, mp, height + 1, bn,
					       first, bc, data);
			if (error)
				break;
		}

640
out:
D
David Teigland 已提交
641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658
	brelse(bh);
	return error;
}

/**
 * do_strip - Look for a layer a particular layer of the file and strip it off
 * @ip: the inode
 * @dibh: the dinode buffer
 * @bh: A buffer of pointers
 * @top: The first pointer in the buffer
 * @bottom: One more than the last pointer
 * @height: the height this buffer is at
 * @data: a pointer to a struct strip_mine
 *
 * Returns: errno
 */

static int do_strip(struct gfs2_inode *ip, struct buffer_head *dibh,
659
		    struct buffer_head *bh, u64 *top, u64 *bottom,
D
David Teigland 已提交
660 661
		    unsigned int height, void *data)
{
662 663
	struct strip_mine *sm = data;
	struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode);
D
David Teigland 已提交
664
	struct gfs2_rgrp_list rlist;
665 666 667
	u64 bn, bstart;
	u32 blen;
	u64 *p;
D
David Teigland 已提交
668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684
	unsigned int rg_blocks = 0;
	int metadata;
	unsigned int revokes = 0;
	int x;
	int error;

	if (!*top)
		sm->sm_first = 0;

	if (height != sm->sm_height)
		return 0;

	if (sm->sm_first) {
		top++;
		sm->sm_first = 0;
	}

685
	metadata = (height != ip->i_di.di_height - 1);
D
David Teigland 已提交
686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722
	if (metadata)
		revokes = (height) ? sdp->sd_inptrs : sdp->sd_diptrs;

	error = gfs2_rindex_hold(sdp, &ip->i_alloc.al_ri_gh);
	if (error)
		return error;

	memset(&rlist, 0, sizeof(struct gfs2_rgrp_list));
	bstart = 0;
	blen = 0;

	for (p = top; p < bottom; p++) {
		if (!*p)
			continue;

		bn = be64_to_cpu(*p);

		if (bstart + blen == bn)
			blen++;
		else {
			if (bstart)
				gfs2_rlist_add(sdp, &rlist, bstart);

			bstart = bn;
			blen = 1;
		}
	}

	if (bstart)
		gfs2_rlist_add(sdp, &rlist, bstart);
	else
		goto out; /* Nothing to do */

	gfs2_rlist_alloc(&rlist, LM_ST_EXCLUSIVE, 0);

	for (x = 0; x < rlist.rl_rgrps; x++) {
		struct gfs2_rgrpd *rgd;
723
		rgd = rlist.rl_ghs[x].gh_gl->gl_object;
D
David Teigland 已提交
724 725 726 727 728 729 730 731 732 733 734 735 736 737 738
		rg_blocks += rgd->rd_ri.ri_length;
	}

	error = gfs2_glock_nq_m(rlist.rl_rgrps, rlist.rl_ghs);
	if (error)
		goto out_rlist;

	error = gfs2_trans_begin(sdp, rg_blocks + RES_DINODE +
				 RES_INDIRECT + RES_STATFS + RES_QUOTA,
				 revokes);
	if (error)
		goto out_rg_gunlock;

	down_write(&ip->i_rw_mutex);

739 740
	gfs2_trans_add_bh(ip->i_gl, dibh, 1);
	gfs2_trans_add_bh(ip->i_gl, bh, 1);
D
David Teigland 已提交
741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784

	bstart = 0;
	blen = 0;

	for (p = top; p < bottom; p++) {
		if (!*p)
			continue;

		bn = be64_to_cpu(*p);

		if (bstart + blen == bn)
			blen++;
		else {
			if (bstart) {
				if (metadata)
					gfs2_free_meta(ip, bstart, blen);
				else
					gfs2_free_data(ip, bstart, blen);
			}

			bstart = bn;
			blen = 1;
		}

		*p = 0;
		if (!ip->i_di.di_blocks)
			gfs2_consist_inode(ip);
		ip->i_di.di_blocks--;
	}
	if (bstart) {
		if (metadata)
			gfs2_free_meta(ip, bstart, blen);
		else
			gfs2_free_data(ip, bstart, blen);
	}

	ip->i_di.di_mtime = ip->i_di.di_ctime = get_seconds();

	gfs2_dinode_out(&ip->i_di, dibh->b_data);

	up_write(&ip->i_rw_mutex);

	gfs2_trans_end(sdp);

785
out_rg_gunlock:
D
David Teigland 已提交
786
	gfs2_glock_dq_m(rlist.rl_rgrps, rlist.rl_ghs);
787
out_rlist:
D
David Teigland 已提交
788
	gfs2_rlist_free(&rlist);
789
out:
D
David Teigland 已提交
790 791 792 793 794 795 796 797 798 799 800 801 802 803
	gfs2_glock_dq_uninit(&ip->i_alloc.al_ri_gh);
	return error;
}

/**
 * do_grow - Make a file look bigger than it is
 * @ip: the inode
 * @size: the size to set the file to
 *
 * Called with an exclusive lock on @ip.
 *
 * Returns: errno
 */

804
static int do_grow(struct gfs2_inode *ip, u64 size)
D
David Teigland 已提交
805
{
806
	struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode);
D
David Teigland 已提交
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
	struct gfs2_alloc *al;
	struct buffer_head *dibh;
	unsigned int h;
	int error;

	al = gfs2_alloc_get(ip);

	error = gfs2_quota_lock(ip, NO_QUOTA_CHANGE, NO_QUOTA_CHANGE);
	if (error)
		goto out;

	error = gfs2_quota_check(ip, ip->i_di.di_uid, ip->i_di.di_gid);
	if (error)
		goto out_gunlock_q;

	al->al_requested = sdp->sd_max_height + RES_DATA;

	error = gfs2_inplace_reserve(ip);
	if (error)
		goto out_gunlock_q;

	error = gfs2_trans_begin(sdp,
			sdp->sd_max_height + al->al_rgd->rd_ri.ri_length +
			RES_JDATA + RES_DINODE + RES_STATFS + RES_QUOTA, 0);
	if (error)
		goto out_ipres;

	if (size > sdp->sd_sb.sb_bsize - sizeof(struct gfs2_dinode)) {
		if (gfs2_is_stuffed(ip)) {
836
			error = gfs2_unstuff_dinode(ip, NULL);
D
David Teigland 已提交
837 838 839 840 841 842 843
			if (error)
				goto out_end_trans;
		}

		h = calc_tree_height(ip, size);
		if (ip->i_di.di_height < h) {
			down_write(&ip->i_rw_mutex);
844
			error = build_height(&ip->i_inode, h);
D
David Teigland 已提交
845 846 847 848 849 850 851 852 853 854 855 856 857
			up_write(&ip->i_rw_mutex);
			if (error)
				goto out_end_trans;
		}
	}

	ip->i_di.di_size = size;
	ip->i_di.di_mtime = ip->i_di.di_ctime = get_seconds();

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

858
	gfs2_trans_add_bh(ip->i_gl, dibh, 1);
D
David Teigland 已提交
859 860 861
	gfs2_dinode_out(&ip->i_di, dibh->b_data);
	brelse(dibh);

862
out_end_trans:
D
David Teigland 已提交
863
	gfs2_trans_end(sdp);
864
out_ipres:
D
David Teigland 已提交
865
	gfs2_inplace_release(ip);
866
out_gunlock_q:
D
David Teigland 已提交
867
	gfs2_quota_unlock(ip);
868
out:
D
David Teigland 已提交
869 870 871 872
	gfs2_alloc_put(ip);
	return error;
}

S
Steven Whitehouse 已提交
873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948

/**
 * gfs2_block_truncate_page - Deal with zeroing out data for truncate
 *
 * This is partly borrowed from ext3.
 */
static int gfs2_block_truncate_page(struct address_space *mapping)
{
	struct inode *inode = mapping->host;
	struct gfs2_inode *ip = GFS2_I(inode);
	struct gfs2_sbd *sdp = GFS2_SB(inode);
	loff_t from = inode->i_size;
	unsigned long index = from >> PAGE_CACHE_SHIFT;
	unsigned offset = from & (PAGE_CACHE_SIZE-1);
	unsigned blocksize, iblock, length, pos;
	struct buffer_head *bh;
	struct page *page;
	void *kaddr;
	int err;

	page = grab_cache_page(mapping, index);
	if (!page)
		return 0;

	blocksize = inode->i_sb->s_blocksize;
	length = blocksize - (offset & (blocksize - 1));
	iblock = index << (PAGE_CACHE_SHIFT - inode->i_sb->s_blocksize_bits);

	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)) {
		gfs2_get_block(inode, iblock, bh, 0);
		/* 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;
		ll_rw_block(READ, 1, &bh);
		wait_on_buffer(bh);
		/* Uhhuh. Read error. Complain and punt. */
		if (!buffer_uptodate(bh))
			goto unlock;
	}

	if (sdp->sd_args.ar_data == GFS2_DATA_ORDERED || gfs2_is_jdata(ip))
		gfs2_trans_add_bh(ip->i_gl, bh, 0);

	kaddr = kmap_atomic(page, KM_USER0);
	memset(kaddr + offset, 0, length);
	flush_dcache_page(page);
	kunmap_atomic(kaddr, KM_USER0);

unlock:
	unlock_page(page);
	page_cache_release(page);
	return err;
}

949
static int trunc_start(struct gfs2_inode *ip, u64 size)
D
David Teigland 已提交
950
{
951
	struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode);
D
David Teigland 已提交
952 953 954 955 956
	struct buffer_head *dibh;
	int journaled = gfs2_is_jdata(ip);
	int error;

	error = gfs2_trans_begin(sdp,
J
Jan Engelhardt 已提交
957
				 RES_DINODE + (journaled ? RES_JDATA : 0), 0);
D
David Teigland 已提交
958 959 960 961 962 963 964 965 966 967
	if (error)
		return error;

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

	if (gfs2_is_stuffed(ip)) {
		ip->i_di.di_size = size;
		ip->i_di.di_mtime = ip->i_di.di_ctime = get_seconds();
968
		gfs2_trans_add_bh(ip->i_gl, dibh, 1);
D
David Teigland 已提交
969 970 971 972 973
		gfs2_dinode_out(&ip->i_di, dibh->b_data);
		gfs2_buffer_clear_tail(dibh, sizeof(struct gfs2_dinode) + size);
		error = 1;

	} else {
974
		if (size & (u64)(sdp->sd_sb.sb_bsize - 1))
975
			error = gfs2_block_truncate_page(ip->i_inode.i_mapping);
D
David Teigland 已提交
976 977 978 979 980

		if (!error) {
			ip->i_di.di_size = size;
			ip->i_di.di_mtime = ip->i_di.di_ctime = get_seconds();
			ip->i_di.di_flags |= GFS2_DIF_TRUNC_IN_PROG;
981
			gfs2_trans_add_bh(ip->i_gl, dibh, 1);
D
David Teigland 已提交
982 983 984 985 986 987
			gfs2_dinode_out(&ip->i_di, dibh->b_data);
		}
	}

	brelse(dibh);

988
out:
D
David Teigland 已提交
989 990 991 992
	gfs2_trans_end(sdp);
	return error;
}

993
static int trunc_dealloc(struct gfs2_inode *ip, u64 size)
D
David Teigland 已提交
994 995
{
	unsigned int height = ip->i_di.di_height;
996
	u64 lblock;
D
David Teigland 已提交
997 998 999 1000 1001
	struct metapath mp;
	int error;

	if (!size)
		lblock = 0;
1002
	else
1003
		lblock = (size - 1) >> GFS2_SB(&ip->i_inode)->sd_sb.sb_bsize_shift;
D
David Teigland 已提交
1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023

	find_metapath(ip, lblock, &mp);
	gfs2_alloc_get(ip);

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

	while (height--) {
		struct strip_mine sm;
		sm.sm_first = !!size;
		sm.sm_height = height;

		error = recursive_scan(ip, NULL, &mp, 0, 0, 1, do_strip, &sm);
		if (error)
			break;
	}

	gfs2_quota_unhold(ip);

1024
out:
D
David Teigland 已提交
1025 1026 1027 1028 1029 1030
	gfs2_alloc_put(ip);
	return error;
}

static int trunc_end(struct gfs2_inode *ip)
{
1031
	struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode);
D
David Teigland 已提交
1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054
	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;

	if (!ip->i_di.di_size) {
		ip->i_di.di_height = 0;
		ip->i_di.di_goal_meta =
			ip->i_di.di_goal_data =
			ip->i_num.no_addr;
		gfs2_buffer_clear_tail(dibh, sizeof(struct gfs2_dinode));
	}
	ip->i_di.di_mtime = ip->i_di.di_ctime = get_seconds();
	ip->i_di.di_flags &= ~GFS2_DIF_TRUNC_IN_PROG;

1055
	gfs2_trans_add_bh(ip->i_gl, dibh, 1);
D
David Teigland 已提交
1056 1057 1058
	gfs2_dinode_out(&ip->i_di, dibh->b_data);
	brelse(dibh);

1059
out:
D
David Teigland 已提交
1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075
	up_write(&ip->i_rw_mutex);
	gfs2_trans_end(sdp);
	return error;
}

/**
 * do_shrink - make a file smaller
 * @ip: the inode
 * @size: the size to make the file
 * @truncator: function to truncate the last partial block
 *
 * Called with an exclusive lock on @ip.
 *
 * Returns: errno
 */

1076
static int do_shrink(struct gfs2_inode *ip, u64 size)
D
David Teigland 已提交
1077 1078 1079
{
	int error;

1080
	error = trunc_start(ip, size);
D
David Teigland 已提交
1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093
	if (error < 0)
		return error;
	if (error > 0)
		return 0;

	error = trunc_dealloc(ip, size);
	if (!error)
		error = trunc_end(ip);

	return error;
}

/**
1094
 * gfs2_truncatei - make a file a given size
D
David Teigland 已提交
1095 1096 1097 1098 1099 1100 1101 1102 1103
 * @ip: the inode
 * @size: the size to make the file
 * @truncator: function to truncate the last partial block
 *
 * The file size can grow, shrink, or stay the same size.
 *
 * Returns: errno
 */

1104
int gfs2_truncatei(struct gfs2_inode *ip, u64 size)
D
David Teigland 已提交
1105 1106 1107
{
	int error;

1108
	if (gfs2_assert_warn(GFS2_SB(&ip->i_inode), S_ISREG(ip->i_di.di_mode)))
D
David Teigland 已提交
1109 1110 1111 1112 1113
		return -EINVAL;

	if (size > ip->i_di.di_size)
		error = do_grow(ip, size);
	else
1114
		error = do_shrink(ip, size);
D
David Teigland 已提交
1115 1116 1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144

	return error;
}

int gfs2_truncatei_resume(struct gfs2_inode *ip)
{
	int error;
	error = trunc_dealloc(ip, ip->i_di.di_size);
	if (!error)
		error = trunc_end(ip);
	return error;
}

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

/**
 * gfs2_write_calc_reserv - calculate number of blocks needed to write to a file
 * @ip: the file
 * @len: the number of bytes to be written to the file
 * @data_blocks: returns the number of data blocks required
 * @ind_blocks: returns the number of indirect blocks required
 *
 */

void gfs2_write_calc_reserv(struct gfs2_inode *ip, unsigned int len,
			    unsigned int *data_blocks, unsigned int *ind_blocks)
{
1145
	struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode);
D
David Teigland 已提交
1146 1147
	unsigned int tmp;

1148
	if (gfs2_is_dir(ip)) {
1149
		*data_blocks = DIV_ROUND_UP(len, sdp->sd_jbsize) + 2;
D
David Teigland 已提交
1150 1151 1152 1153 1154 1155 1156
		*ind_blocks = 3 * (sdp->sd_max_jheight - 1);
	} else {
		*data_blocks = (len >> sdp->sd_sb.sb_bsize_shift) + 3;
		*ind_blocks = 3 * (sdp->sd_max_height - 1);
	}

	for (tmp = *data_blocks; tmp > sdp->sd_diptrs;) {
1157
		tmp = DIV_ROUND_UP(tmp, sdp->sd_inptrs);
D
David Teigland 已提交
1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171
		*ind_blocks += tmp;
	}
}

/**
 * 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
 * @alloc_required: set to 1 if an alloc is required, 0 otherwise
 *
 * Returns: errno
 */

1172
int gfs2_write_alloc_required(struct gfs2_inode *ip, u64 offset,
D
David Teigland 已提交
1173 1174
			      unsigned int len, int *alloc_required)
{
1175
	struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode);
1176 1177
	u64 lblock, lblock_stop, dblock;
	u32 extlen;
D
David Teigland 已提交
1178 1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192
	int new = 0;
	int error = 0;

	*alloc_required = 0;

	if (!len)
		return 0;

	if (gfs2_is_stuffed(ip)) {
		if (offset + len >
		    sdp->sd_sb.sb_bsize - sizeof(struct gfs2_dinode))
			*alloc_required = 1;
		return 0;
	}

1193
	if (gfs2_is_dir(ip)) {
D
David Teigland 已提交
1194 1195 1196 1197 1198 1199 1200 1201 1202 1203 1204 1205
		unsigned int bsize = sdp->sd_jbsize;
		lblock = offset;
		do_div(lblock, bsize);
		lblock_stop = offset + len + bsize - 1;
		do_div(lblock_stop, bsize);
	} else {
		unsigned int shift = sdp->sd_sb.sb_bsize_shift;
		lblock = offset >> shift;
		lblock_stop = (offset + len + sdp->sd_sb.sb_bsize - 1) >> shift;
	}

	for (; lblock < lblock_stop; lblock += extlen) {
1206
		error = gfs2_extent_map(&ip->i_inode, lblock, &new, &dblock, &extlen);
D
David Teigland 已提交
1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218
		if (error)
			return error;

		if (!dblock) {
			*alloc_required = 1;
			return 0;
		}
	}

	return 0;
}