rgrp.c 40.4 KB
Newer Older
D
David Teigland 已提交
1 2
/*
 * Copyright (C) Sistina Software, Inc.  1997-2003 All rights reserved.
3
 * Copyright (C) 2004-2008 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
 */

#include <linux/slab.h>
#include <linux/spinlock.h>
#include <linux/completion.h>
#include <linux/buffer_head.h>
14
#include <linux/fs.h>
15
#include <linux/gfs2_ondisk.h>
16
#include <linux/lm_interface.h>
17
#include <linux/prefetch.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 28
#include "glock.h"
#include "glops.h"
#include "lops.h"
#include "meta_io.h"
#include "quota.h"
#include "rgrp.h"
#include "super.h"
#include "trans.h"
29
#include "util.h"
30
#include "log.h"
31
#include "inode.h"
32
#include "ops_address.h"
D
David Teigland 已提交
33

S
Steven Whitehouse 已提交
34
#define BFITNOENT ((u32)~0)
35
#define NO_BLOCK ((u64)~0)
36

37 38 39 40 41 42 43 44 45 46
#if BITS_PER_LONG == 32
#define LBITMASK   (0x55555555UL)
#define LBITSKIP55 (0x55555555UL)
#define LBITSKIP00 (0x00000000UL)
#else
#define LBITMASK   (0x5555555555555555UL)
#define LBITSKIP55 (0x5555555555555555UL)
#define LBITSKIP00 (0x0000000000000000UL)
#endif

47 48 49
/*
 * These routines are used by the resource group routines (rgrp.c)
 * to keep track of block allocation.  Each block is represented by two
50 51 52 53 54 55
 * bits.  So, each byte represents GFS2_NBBY (i.e. 4) blocks.
 *
 * 0 = Free
 * 1 = Used (not metadata)
 * 2 = Unlinked (still in use) inode
 * 3 = Used (metadata)
56 57 58 59
 */

static const char valid_change[16] = {
	        /* current */
60
	/* n */ 0, 1, 1, 1,
61
	/* e */ 1, 0, 0, 0,
62
	/* w */ 0, 0, 0, 1,
63 64 65
	        1, 0, 0, 0
};

66
static u32 rgblk_search(struct gfs2_rgrpd *rgd, u32 goal,
67 68
                        unsigned char old_state, unsigned char new_state,
			unsigned int *n);
69

70 71 72 73 74 75 76 77 78
/**
 * gfs2_setbit - Set a bit in the bitmaps
 * @buffer: the buffer that holds the bitmaps
 * @buflen: the length (in bytes) of the buffer
 * @block: the block to set
 * @new_state: the new state of the block
 *
 */

79 80 81 82
static inline void gfs2_setbit(struct gfs2_rgrpd *rgd, unsigned char *buf1,
			       unsigned char *buf2, unsigned int offset,
			       unsigned int buflen, u32 block,
			       unsigned char new_state)
83
{
84 85
	unsigned char *byte1, *byte2, *end, cur_state;
	const unsigned int bit = (block % GFS2_NBBY) * GFS2_BIT_SIZE;
86

87 88
	byte1 = buf1 + offset + (block / GFS2_NBBY);
	end = buf1 + offset + buflen;
89

90
	BUG_ON(byte1 >= end);
91

92
	cur_state = (*byte1 >> bit) & GFS2_BIT_MASK;
93

94
	if (unlikely(!valid_change[new_state * 4 + cur_state])) {
95
		gfs2_consist_rgrpd(rgd);
96 97 98 99 100 101 102 103 104
		return;
	}
	*byte1 ^= (cur_state ^ new_state) << bit;

	if (buf2) {
		byte2 = buf2 + offset + (block / GFS2_NBBY);
		cur_state = (*byte2 >> bit) & GFS2_BIT_MASK;
		*byte2 ^= (cur_state ^ new_state) << bit;
	}
105 106 107 108 109 110 111 112 113 114
}

/**
 * gfs2_testbit - test a bit in the bitmaps
 * @buffer: the buffer that holds the bitmaps
 * @buflen: the length (in bytes) of the buffer
 * @block: the block to read
 *
 */

115 116 117
static inline unsigned char gfs2_testbit(struct gfs2_rgrpd *rgd,
					 const unsigned char *buffer,
					 unsigned int buflen, u32 block)
118
{
119 120
	const unsigned char *byte, *end;
	unsigned char cur_state;
121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139
	unsigned int bit;

	byte = buffer + (block / GFS2_NBBY);
	bit = (block % GFS2_NBBY) * GFS2_BIT_SIZE;
	end = buffer + buflen;

	gfs2_assert(rgd->rd_sbd, byte < end);

	cur_state = (*byte >> bit) & GFS2_BIT_MASK;

	return cur_state;
}

/**
 * gfs2_bitfit - Search an rgrp's bitmap buffer to find a bit-pair representing
 *       a block in a given allocation state.
 * @buffer: the buffer that holds the bitmaps
 * @buflen: the length (in bytes) of the buffer
 * @goal: start search at this block's bit-pair (within @buffer)
140
 * @old_state: GFS2_BLKST_XXX the state of the block we're looking for.
141 142 143 144 145 146 147 148
 *
 * Scope of @goal and returned block number is only within this bitmap buffer,
 * not entire rgrp or filesystem.  @buffer will be offset from the actual
 * beginning of a bitmap block buffer, skipping any header structures.
 *
 * Return: the block number (bitmap buffer scope) that was found
 */

149 150
static u32 gfs2_bitfit(const u8 *buffer, unsigned int buflen, u32 goal,
		       u8 old_state)
151
{
152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175
	const u8 *byte, *start, *end;
	int bit, startbit;
	u32 g1, g2, misaligned;
	unsigned long *plong;
	unsigned long lskipval;

	lskipval = (old_state & GFS2_BLKST_USED) ? LBITSKIP00 : LBITSKIP55;
	g1 = (goal / GFS2_NBBY);
	start = buffer + g1;
	byte = start;
        end = buffer + buflen;
	g2 = ALIGN(g1, sizeof(unsigned long));
	plong = (unsigned long *)(buffer + g2);
	startbit = bit = (goal % GFS2_NBBY) * GFS2_BIT_SIZE;
	misaligned = g2 - g1;
	if (!misaligned)
		goto ulong_aligned;
/* parse the bitmap a byte at a time */
misaligned:
	while (byte < end) {
		if (((*byte >> bit) & GFS2_BIT_MASK) == old_state) {
			return goal +
				(((byte - start) * GFS2_NBBY) +
				 ((bit - startbit) >> 1));
176 177
		}
		bit += GFS2_BIT_SIZE;
178
		if (bit >= GFS2_NBBY * GFS2_BIT_SIZE) {
179 180
			bit = 0;
			byte++;
181 182 183 184 185
			misaligned--;
			if (!misaligned) {
				plong = (unsigned long *)byte;
				goto ulong_aligned;
			}
186 187
		}
	}
188
	return BFITNOENT;
189

190 191 192 193 194 195 196 197
/* parse the bitmap a unsigned long at a time */
ulong_aligned:
	/* Stop at "end - 1" or else prefetch can go past the end and segfault.
	   We could "if" it but we'd lose some of the performance gained.
	   This way will only slow down searching the very last 4/8 bytes
	   depending on architecture.  I've experimented with several ways
	   of writing this section such as using an else before the goto
	   but this one seems to be the fastest. */
198
	while ((unsigned char *)plong < end - sizeof(unsigned long)) {
199 200 201 202 203 204 205 206 207 208
		prefetch(plong + 1);
		if (((*plong) & LBITMASK) != lskipval)
			break;
		plong++;
	}
	if ((unsigned char *)plong < end) {
		byte = (const u8 *)plong;
		misaligned += sizeof(unsigned long) - 1;
		goto misaligned;
	}
209 210 211 212 213 214 215 216 217 218 219 220
	return BFITNOENT;
}

/**
 * gfs2_bitcount - count the number of bits in a certain state
 * @buffer: the buffer that holds the bitmaps
 * @buflen: the length (in bytes) of the buffer
 * @state: the state of the block we're looking for
 *
 * Returns: The number of bits
 */

221 222
static u32 gfs2_bitcount(struct gfs2_rgrpd *rgd, const u8 *buffer,
			 unsigned int buflen, u8 state)
223
{
224 225 226 227 228
	const u8 *byte = buffer;
	const u8 *end = buffer + buflen;
	const u8 state1 = state << 2;
	const u8 state2 = state << 4;
	const u8 state3 = state << 6;
229
	u32 count = 0;
230 231 232 233 234 235 236 237 238 239 240 241 242 243 244

	for (; byte < end; byte++) {
		if (((*byte) & 0x03) == state)
			count++;
		if (((*byte) & 0x0C) == state1)
			count++;
		if (((*byte) & 0x30) == state2)
			count++;
		if (((*byte) & 0xC0) == state3)
			count++;
	}

	return count;
}

D
David Teigland 已提交
245 246 247 248 249 250 251 252 253 254 255
/**
 * gfs2_rgrp_verify - Verify that a resource group is consistent
 * @sdp: the filesystem
 * @rgd: the rgrp
 *
 */

void gfs2_rgrp_verify(struct gfs2_rgrpd *rgd)
{
	struct gfs2_sbd *sdp = rgd->rd_sbd;
	struct gfs2_bitmap *bi = NULL;
256
	u32 length = rgd->rd_length;
257
	u32 count[4], tmp;
D
David Teigland 已提交
258 259
	int buf, x;

260
	memset(count, 0, 4 * sizeof(u32));
D
David Teigland 已提交
261 262 263 264 265 266 267 268 269 270 271

	/* Count # blocks in each of 4 possible allocation states */
	for (buf = 0; buf < length; buf++) {
		bi = rgd->rd_bits + buf;
		for (x = 0; x < 4; x++)
			count[x] += gfs2_bitcount(rgd,
						  bi->bi_bh->b_data +
						  bi->bi_offset,
						  bi->bi_len, x);
	}

272
	if (count[0] != rgd->rd_free) {
D
David Teigland 已提交
273 274
		if (gfs2_consist_rgrpd(rgd))
			fs_err(sdp, "free data mismatch:  %u != %u\n",
275
			       count[0], rgd->rd_free);
D
David Teigland 已提交
276 277 278
		return;
	}

279
	tmp = rgd->rd_data -
280
		rgd->rd_free -
D
David Teigland 已提交
281
		rgd->rd_rg.rg_dinodes;
282
	if (count[1] + count[2] != tmp) {
D
David Teigland 已提交
283 284 285 286 287 288
		if (gfs2_consist_rgrpd(rgd))
			fs_err(sdp, "used data mismatch:  %u != %u\n",
			       count[1], tmp);
		return;
	}

289
	if (count[3] != rgd->rd_rg.rg_dinodes) {
D
David Teigland 已提交
290
		if (gfs2_consist_rgrpd(rgd))
291 292
			fs_err(sdp, "used metadata mismatch:  %u != %u\n",
			       count[3], rgd->rd_rg.rg_dinodes);
D
David Teigland 已提交
293 294 295
		return;
	}

296
	if (count[2] > count[3]) {
D
David Teigland 已提交
297
		if (gfs2_consist_rgrpd(rgd))
298 299
			fs_err(sdp, "unlinked inodes > inodes:  %u\n",
			       count[2]);
D
David Teigland 已提交
300 301
		return;
	}
302

D
David Teigland 已提交
303 304
}

305
static inline int rgrp_contains_block(struct gfs2_rgrpd *rgd, u64 block)
D
David Teigland 已提交
306
{
307 308
	u64 first = rgd->rd_data0;
	u64 last = first + rgd->rd_data;
309
	return first <= block && block < last;
D
David Teigland 已提交
310 311 312 313 314 315 316 317 318 319
}

/**
 * gfs2_blk2rgrpd - Find resource group for a given data/meta block number
 * @sdp: The GFS2 superblock
 * @n: The data block number
 *
 * Returns: The resource group, or NULL if not found
 */

320
struct gfs2_rgrpd *gfs2_blk2rgrpd(struct gfs2_sbd *sdp, u64 blk)
D
David Teigland 已提交
321 322 323 324 325 326
{
	struct gfs2_rgrpd *rgd;

	spin_lock(&sdp->sd_rindex_spin);

	list_for_each_entry(rgd, &sdp->sd_rindex_mru_list, rd_list_mru) {
327
		if (rgrp_contains_block(rgd, blk)) {
D
David Teigland 已提交
328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384
			list_move(&rgd->rd_list_mru, &sdp->sd_rindex_mru_list);
			spin_unlock(&sdp->sd_rindex_spin);
			return rgd;
		}
	}

	spin_unlock(&sdp->sd_rindex_spin);

	return NULL;
}

/**
 * gfs2_rgrpd_get_first - get the first Resource Group in the filesystem
 * @sdp: The GFS2 superblock
 *
 * Returns: The first rgrp in the filesystem
 */

struct gfs2_rgrpd *gfs2_rgrpd_get_first(struct gfs2_sbd *sdp)
{
	gfs2_assert(sdp, !list_empty(&sdp->sd_rindex_list));
	return list_entry(sdp->sd_rindex_list.next, struct gfs2_rgrpd, rd_list);
}

/**
 * gfs2_rgrpd_get_next - get the next RG
 * @rgd: A RG
 *
 * Returns: The next rgrp
 */

struct gfs2_rgrpd *gfs2_rgrpd_get_next(struct gfs2_rgrpd *rgd)
{
	if (rgd->rd_list.next == &rgd->rd_sbd->sd_rindex_list)
		return NULL;
	return list_entry(rgd->rd_list.next, struct gfs2_rgrpd, rd_list);
}

static void clear_rgrpdi(struct gfs2_sbd *sdp)
{
	struct list_head *head;
	struct gfs2_rgrpd *rgd;
	struct gfs2_glock *gl;

	spin_lock(&sdp->sd_rindex_spin);
	sdp->sd_rindex_forward = NULL;
	spin_unlock(&sdp->sd_rindex_spin);

	head = &sdp->sd_rindex_list;
	while (!list_empty(head)) {
		rgd = list_entry(head->next, struct gfs2_rgrpd, rd_list);
		gl = rgd->rd_gl;

		list_del(&rgd->rd_list);
		list_del(&rgd->rd_list_mru);

		if (gl) {
385
			gl->gl_object = NULL;
D
David Teigland 已提交
386 387 388 389
			gfs2_glock_put(gl);
		}

		kfree(rgd->rd_bits);
390
		kmem_cache_free(gfs2_rgrpd_cachep, rgd);
D
David Teigland 已提交
391 392 393 394 395
	}
}

void gfs2_clear_rgrpd(struct gfs2_sbd *sdp)
{
396
	mutex_lock(&sdp->sd_rindex_mutex);
D
David Teigland 已提交
397
	clear_rgrpdi(sdp);
398
	mutex_unlock(&sdp->sd_rindex_mutex);
D
David Teigland 已提交
399 400
}

401 402 403 404 405 406 407 408 409
static void gfs2_rindex_print(const struct gfs2_rgrpd *rgd)
{
	printk(KERN_INFO "  ri_addr = %llu\n", (unsigned long long)rgd->rd_addr);
	printk(KERN_INFO "  ri_length = %u\n", rgd->rd_length);
	printk(KERN_INFO "  ri_data0 = %llu\n", (unsigned long long)rgd->rd_data0);
	printk(KERN_INFO "  ri_data = %u\n", rgd->rd_data);
	printk(KERN_INFO "  ri_bitbytes = %u\n", rgd->rd_bitbytes);
}

D
David Teigland 已提交
410 411 412 413 414 415 416 417 418 419 420 421 422
/**
 * gfs2_compute_bitstructs - Compute the bitmap sizes
 * @rgd: The resource group descriptor
 *
 * Calculates bitmap descriptors, one for each block that contains bitmap data
 *
 * Returns: errno
 */

static int compute_bitstructs(struct gfs2_rgrpd *rgd)
{
	struct gfs2_sbd *sdp = rgd->rd_sbd;
	struct gfs2_bitmap *bi;
423
	u32 length = rgd->rd_length; /* # blocks in hdr & bitmap */
424
	u32 bytes_left, bytes;
D
David Teigland 已提交
425 426
	int x;

427 428 429
	if (!length)
		return -EINVAL;

430
	rgd->rd_bits = kcalloc(length, sizeof(struct gfs2_bitmap), GFP_NOFS);
D
David Teigland 已提交
431 432 433
	if (!rgd->rd_bits)
		return -ENOMEM;

434
	bytes_left = rgd->rd_bitbytes;
D
David Teigland 已提交
435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454

	for (x = 0; x < length; x++) {
		bi = rgd->rd_bits + x;

		/* small rgrp; bitmap stored completely in header block */
		if (length == 1) {
			bytes = bytes_left;
			bi->bi_offset = sizeof(struct gfs2_rgrp);
			bi->bi_start = 0;
			bi->bi_len = bytes;
		/* header block */
		} else if (x == 0) {
			bytes = sdp->sd_sb.sb_bsize - sizeof(struct gfs2_rgrp);
			bi->bi_offset = sizeof(struct gfs2_rgrp);
			bi->bi_start = 0;
			bi->bi_len = bytes;
		/* last block */
		} else if (x + 1 == length) {
			bytes = bytes_left;
			bi->bi_offset = sizeof(struct gfs2_meta_header);
455
			bi->bi_start = rgd->rd_bitbytes - bytes_left;
D
David Teigland 已提交
456 457 458
			bi->bi_len = bytes;
		/* other blocks */
		} else {
459 460
			bytes = sdp->sd_sb.sb_bsize -
				sizeof(struct gfs2_meta_header);
D
David Teigland 已提交
461
			bi->bi_offset = sizeof(struct gfs2_meta_header);
462
			bi->bi_start = rgd->rd_bitbytes - bytes_left;
D
David Teigland 已提交
463 464 465 466 467 468 469 470 471 472 473
			bi->bi_len = bytes;
		}

		bytes_left -= bytes;
	}

	if (bytes_left) {
		gfs2_consist_rgrpd(rgd);
		return -EIO;
	}
	bi = rgd->rd_bits + (length - 1);
474
	if ((bi->bi_start + bi->bi_len) * GFS2_NBBY != rgd->rd_data) {
D
David Teigland 已提交
475
		if (gfs2_consist_rgrpd(rgd)) {
476
			gfs2_rindex_print(rgd);
D
David Teigland 已提交
477 478 479 480 481 482 483 484 485
			fs_err(sdp, "start=%u len=%u offset=%u\n",
			       bi->bi_start, bi->bi_len, bi->bi_offset);
		}
		return -EIO;
	}

	return 0;
}

486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503
/**
 * gfs2_ri_total - Total up the file system space, according to the rindex.
 *
 */
u64 gfs2_ri_total(struct gfs2_sbd *sdp)
{
	u64 total_data = 0;	
	struct inode *inode = sdp->sd_rindex;
	struct gfs2_inode *ip = GFS2_I(inode);
	char buf[sizeof(struct gfs2_rindex)];
	struct file_ra_state ra_state;
	int error, rgrps;

	mutex_lock(&sdp->sd_rindex_mutex);
	file_ra_state_init(&ra_state, inode->i_mapping);
	for (rgrps = 0;; rgrps++) {
		loff_t pos = rgrps * sizeof(struct gfs2_rindex);

504
		if (pos + sizeof(struct gfs2_rindex) >= ip->i_disksize)
505 506 507 508 509
			break;
		error = gfs2_internal_read(ip, &ra_state, buf, &pos,
					   sizeof(struct gfs2_rindex));
		if (error != sizeof(struct gfs2_rindex))
			break;
510
		total_data += be32_to_cpu(((struct gfs2_rindex *)buf)->ri_data);
511 512 513 514 515
	}
	mutex_unlock(&sdp->sd_rindex_mutex);
	return total_data;
}

516 517 518 519 520 521 522 523 524 525 526
static void gfs2_rindex_in(struct gfs2_rgrpd *rgd, const void *buf)
{
	const struct gfs2_rindex *str = buf;

	rgd->rd_addr = be64_to_cpu(str->ri_addr);
	rgd->rd_length = be32_to_cpu(str->ri_length);
	rgd->rd_data0 = be64_to_cpu(str->ri_data0);
	rgd->rd_data = be32_to_cpu(str->ri_data);
	rgd->rd_bitbytes = be32_to_cpu(str->ri_bitbytes);
}

D
David Teigland 已提交
527
/**
528
 * read_rindex_entry - Pull in a new resource index entry from the disk
D
David Teigland 已提交
529 530
 * @gl: The glock covering the rindex inode
 *
531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552
 * Returns: 0 on success, error code otherwise
 */

static int read_rindex_entry(struct gfs2_inode *ip,
			     struct file_ra_state *ra_state)
{
	struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode);
	loff_t pos = sdp->sd_rgrps * sizeof(struct gfs2_rindex);
	char buf[sizeof(struct gfs2_rindex)];
	int error;
	struct gfs2_rgrpd *rgd;

	error = gfs2_internal_read(ip, ra_state, buf, &pos,
				   sizeof(struct gfs2_rindex));
	if (!error)
		return 0;
	if (error != sizeof(struct gfs2_rindex)) {
		if (error > 0)
			error = -EIO;
		return error;
	}

553
	rgd = kmem_cache_zalloc(gfs2_rgrpd_cachep, GFP_NOFS);
554 555 556 557 558 559 560 561 562 563 564
	error = -ENOMEM;
	if (!rgd)
		return error;

	mutex_init(&rgd->rd_mutex);
	lops_init_le(&rgd->rd_le, &gfs2_rg_lops);
	rgd->rd_sbd = sdp;

	list_add_tail(&rgd->rd_list, &sdp->sd_rindex_list);
	list_add_tail(&rgd->rd_list_mru, &sdp->sd_rindex_mru_list);

565
	gfs2_rindex_in(rgd, buf);
566 567 568 569
	error = compute_bitstructs(rgd);
	if (error)
		return error;

570
	error = gfs2_glock_get(sdp, rgd->rd_addr,
571 572 573 574 575
			       &gfs2_rgrp_glops, CREATE, &rgd->rd_gl);
	if (error)
		return error;

	rgd->rd_gl->gl_object = rgd;
576
	rgd->rd_flags &= ~GFS2_RDF_UPTODATE;
577
	rgd->rd_flags |= GFS2_RDF_CHECK;
578 579 580 581 582 583 584
	return error;
}

/**
 * gfs2_ri_update - Pull in a new resource index from the disk
 * @ip: pointer to the rindex inode
 *
D
David Teigland 已提交
585 586 587 588 589
 * Returns: 0 on successful update, error code otherwise
 */

static int gfs2_ri_update(struct gfs2_inode *ip)
{
590 591
	struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode);
	struct inode *inode = &ip->i_inode;
592
	struct file_ra_state ra_state;
593
	u64 rgrp_count = ip->i_disksize;
D
David Teigland 已提交
594 595
	int error;

596
	if (do_div(rgrp_count, sizeof(struct gfs2_rindex))) {
D
David Teigland 已提交
597 598 599 600 601 602
		gfs2_consist_inode(ip);
		return -EIO;
	}

	clear_rgrpdi(sdp);

603
	file_ra_state_init(&ra_state, inode->i_mapping);
604
	for (sdp->sd_rgrps = 0; sdp->sd_rgrps < rgrp_count; sdp->sd_rgrps++) {
605 606 607 608
		error = read_rindex_entry(ip, &ra_state);
		if (error) {
			clear_rgrpdi(sdp);
			return error;
D
David Teigland 已提交
609
		}
610
	}
D
David Teigland 已提交
611

612
	sdp->sd_rindex_uptodate = 1;
613 614
	return 0;
}
D
David Teigland 已提交
615

616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631
/**
 * gfs2_ri_update_special - Pull in a new resource index from the disk
 *
 * This is a special version that's safe to call from gfs2_inplace_reserve_i.
 * In this case we know that we don't have any resource groups in memory yet.
 *
 * @ip: pointer to the rindex inode
 *
 * Returns: 0 on successful update, error code otherwise
 */
static int gfs2_ri_update_special(struct gfs2_inode *ip)
{
	struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode);
	struct inode *inode = &ip->i_inode;
	struct file_ra_state ra_state;
	int error;
D
David Teigland 已提交
632

633 634 635 636
	file_ra_state_init(&ra_state, inode->i_mapping);
	for (sdp->sd_rgrps = 0;; sdp->sd_rgrps++) {
		/* Ignore partials */
		if ((sdp->sd_rgrps + 1) * sizeof(struct gfs2_rindex) >
637
		    ip->i_disksize)
638 639 640 641 642 643
			break;
		error = read_rindex_entry(ip, &ra_state);
		if (error) {
			clear_rgrpdi(sdp);
			return error;
		}
D
David Teigland 已提交
644 645
	}

646
	sdp->sd_rindex_uptodate = 1;
D
David Teigland 已提交
647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669
	return 0;
}

/**
 * gfs2_rindex_hold - Grab a lock on the rindex
 * @sdp: The GFS2 superblock
 * @ri_gh: the glock holder
 *
 * We grab a lock on the rindex inode to make sure that it doesn't
 * change whilst we are performing an operation. We keep this lock
 * for quite long periods of time compared to other locks. This
 * doesn't matter, since it is shared and it is very, very rarely
 * accessed in the exclusive mode (i.e. only when expanding the filesystem).
 *
 * This makes sure that we're using the latest copy of the resource index
 * special file, which might have been updated if someone expanded the
 * filesystem (via gfs2_grow utility), which adds new resource groups.
 *
 * Returns: 0 on success, error code otherwise
 */

int gfs2_rindex_hold(struct gfs2_sbd *sdp, struct gfs2_holder *ri_gh)
{
670
	struct gfs2_inode *ip = GFS2_I(sdp->sd_rindex);
D
David Teigland 已提交
671 672 673 674 675 676 677 678
	struct gfs2_glock *gl = ip->i_gl;
	int error;

	error = gfs2_glock_nq_init(gl, LM_ST_SHARED, 0, ri_gh);
	if (error)
		return error;

	/* Read new copy from disk if we don't have the latest */
679
	if (!sdp->sd_rindex_uptodate) {
680
		mutex_lock(&sdp->sd_rindex_mutex);
681
		if (!sdp->sd_rindex_uptodate) {
D
David Teigland 已提交
682 683 684 685
			error = gfs2_ri_update(ip);
			if (error)
				gfs2_glock_dq_uninit(ri_gh);
		}
686
		mutex_unlock(&sdp->sd_rindex_mutex);
D
David Teigland 已提交
687 688 689 690 691
	}

	return error;
}

692
static void gfs2_rgrp_in(struct gfs2_rgrpd *rgd, const void *buf)
693 694
{
	const struct gfs2_rgrp *str = buf;
695 696
	struct gfs2_rgrp_host *rg = &rgd->rd_rg;
	u32 rg_flags;
697

698 699 700 701 702
	rg_flags = be32_to_cpu(str->rg_flags);
	if (rg_flags & GFS2_RGF_NOALLOC)
		rgd->rd_flags |= GFS2_RDF_NOALLOC;
	else
		rgd->rd_flags &= ~GFS2_RDF_NOALLOC;
703
	rgd->rd_free = be32_to_cpu(str->rg_free);
704
	rg->rg_dinodes = be32_to_cpu(str->rg_dinodes);
705
	rgd->rd_igeneration = be64_to_cpu(str->rg_igeneration);
706 707
}

708
static void gfs2_rgrp_out(struct gfs2_rgrpd *rgd, void *buf)
709 710
{
	struct gfs2_rgrp *str = buf;
711 712
	struct gfs2_rgrp_host *rg = &rgd->rd_rg;
	u32 rg_flags = 0;
713

714 715 716
	if (rgd->rd_flags & GFS2_RDF_NOALLOC)
		rg_flags |= GFS2_RGF_NOALLOC;
	str->rg_flags = cpu_to_be32(rg_flags);
717
	str->rg_free = cpu_to_be32(rgd->rd_free);
718 719
	str->rg_dinodes = cpu_to_be32(rg->rg_dinodes);
	str->__pad = cpu_to_be32(0);
720
	str->rg_igeneration = cpu_to_be64(rgd->rd_igeneration);
721 722 723
	memset(&str->rg_reserved, 0, sizeof(str->rg_reserved));
}

D
David Teigland 已提交
724 725 726 727 728 729 730 731 732 733 734 735 736 737
/**
 * gfs2_rgrp_bh_get - Read in a RG's header and bitmaps
 * @rgd: the struct gfs2_rgrpd describing the RG to read in
 *
 * Read in all of a Resource Group's header and bitmap blocks.
 * Caller must eventually call gfs2_rgrp_relse() to free the bitmaps.
 *
 * Returns: errno
 */

int gfs2_rgrp_bh_get(struct gfs2_rgrpd *rgd)
{
	struct gfs2_sbd *sdp = rgd->rd_sbd;
	struct gfs2_glock *gl = rgd->rd_gl;
738
	unsigned int length = rgd->rd_length;
D
David Teigland 已提交
739 740 741 742
	struct gfs2_bitmap *bi;
	unsigned int x, y;
	int error;

743
	mutex_lock(&rgd->rd_mutex);
D
David Teigland 已提交
744 745 746 747 748

	spin_lock(&sdp->sd_rindex_spin);
	if (rgd->rd_bh_count) {
		rgd->rd_bh_count++;
		spin_unlock(&sdp->sd_rindex_spin);
749
		mutex_unlock(&rgd->rd_mutex);
D
David Teigland 已提交
750 751 752 753 754 755
		return 0;
	}
	spin_unlock(&sdp->sd_rindex_spin);

	for (x = 0; x < length; x++) {
		bi = rgd->rd_bits + x;
756
		error = gfs2_meta_read(gl, rgd->rd_addr + x, 0, &bi->bi_bh);
D
David Teigland 已提交
757 758 759 760 761 762
		if (error)
			goto fail;
	}

	for (y = length; y--;) {
		bi = rgd->rd_bits + y;
S
Steven Whitehouse 已提交
763
		error = gfs2_meta_wait(sdp, bi->bi_bh);
D
David Teigland 已提交
764 765
		if (error)
			goto fail;
766
		if (gfs2_metatype_check(sdp, bi->bi_bh, y ? GFS2_METATYPE_RB :
D
David Teigland 已提交
767 768 769 770 771 772
					      GFS2_METATYPE_RG)) {
			error = -EIO;
			goto fail;
		}
	}

773
	if (!(rgd->rd_flags & GFS2_RDF_UPTODATE)) {
774
		gfs2_rgrp_in(rgd, (rgd->rd_bits[0].bi_bh)->b_data);
775
		rgd->rd_flags |= GFS2_RDF_UPTODATE;
D
David Teigland 已提交
776 777 778
	}

	spin_lock(&sdp->sd_rindex_spin);
779
	rgd->rd_free_clone = rgd->rd_free;
D
David Teigland 已提交
780 781 782
	rgd->rd_bh_count++;
	spin_unlock(&sdp->sd_rindex_spin);

783
	mutex_unlock(&rgd->rd_mutex);
D
David Teigland 已提交
784 785 786

	return 0;

787
fail:
D
David Teigland 已提交
788 789 790 791 792 793
	while (x--) {
		bi = rgd->rd_bits + x;
		brelse(bi->bi_bh);
		bi->bi_bh = NULL;
		gfs2_assert_warn(sdp, !bi->bi_clone);
	}
794
	mutex_unlock(&rgd->rd_mutex);
D
David Teigland 已提交
795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817

	return error;
}

void gfs2_rgrp_bh_hold(struct gfs2_rgrpd *rgd)
{
	struct gfs2_sbd *sdp = rgd->rd_sbd;

	spin_lock(&sdp->sd_rindex_spin);
	gfs2_assert_warn(rgd->rd_sbd, rgd->rd_bh_count);
	rgd->rd_bh_count++;
	spin_unlock(&sdp->sd_rindex_spin);
}

/**
 * gfs2_rgrp_bh_put - Release RG bitmaps read in with gfs2_rgrp_bh_get()
 * @rgd: the struct gfs2_rgrpd describing the RG to read in
 *
 */

void gfs2_rgrp_bh_put(struct gfs2_rgrpd *rgd)
{
	struct gfs2_sbd *sdp = rgd->rd_sbd;
818
	int x, length = rgd->rd_length;
D
David Teigland 已提交
819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840

	spin_lock(&sdp->sd_rindex_spin);
	gfs2_assert_warn(rgd->rd_sbd, rgd->rd_bh_count);
	if (--rgd->rd_bh_count) {
		spin_unlock(&sdp->sd_rindex_spin);
		return;
	}

	for (x = 0; x < length; x++) {
		struct gfs2_bitmap *bi = rgd->rd_bits + x;
		kfree(bi->bi_clone);
		bi->bi_clone = NULL;
		brelse(bi->bi_bh);
		bi->bi_bh = NULL;
	}

	spin_unlock(&sdp->sd_rindex_spin);
}

void gfs2_rgrp_repolish_clones(struct gfs2_rgrpd *rgd)
{
	struct gfs2_sbd *sdp = rgd->rd_sbd;
841
	unsigned int length = rgd->rd_length;
D
David Teigland 已提交
842 843 844 845 846 847 848
	unsigned int x;

	for (x = 0; x < length; x++) {
		struct gfs2_bitmap *bi = rgd->rd_bits + x;
		if (!bi->bi_clone)
			continue;
		memcpy(bi->bi_clone + bi->bi_offset,
849
		       bi->bi_bh->b_data + bi->bi_offset, bi->bi_len);
D
David Teigland 已提交
850 851 852
	}

	spin_lock(&sdp->sd_rindex_spin);
853
	rgd->rd_free_clone = rgd->rd_free;
D
David Teigland 已提交
854 855 856 857 858 859 860 861 862 863 864 865
	spin_unlock(&sdp->sd_rindex_spin);
}

/**
 * gfs2_alloc_get - get the struct gfs2_alloc structure for an inode
 * @ip: the incore GFS2 inode structure
 *
 * Returns: the struct gfs2_alloc
 */

struct gfs2_alloc *gfs2_alloc_get(struct gfs2_inode *ip)
{
866 867 868
	BUG_ON(ip->i_alloc != NULL);
	ip->i_alloc = kzalloc(sizeof(struct gfs2_alloc), GFP_KERNEL);
	return ip->i_alloc;
D
David Teigland 已提交
869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886
}

/**
 * try_rgrp_fit - See if a given reservation will fit in a given RG
 * @rgd: the RG data
 * @al: the struct gfs2_alloc structure describing the reservation
 *
 * If there's room for the requested blocks to be allocated from the RG:
 *   Sets the $al_rgd field in @al.
 *
 * Returns: 1 on success (it fits), 0 on failure (it doesn't fit)
 */

static int try_rgrp_fit(struct gfs2_rgrpd *rgd, struct gfs2_alloc *al)
{
	struct gfs2_sbd *sdp = rgd->rd_sbd;
	int ret = 0;

887
	if (rgd->rd_flags & GFS2_RDF_NOALLOC)
888 889
		return 0;

D
David Teigland 已提交
890 891 892 893 894 895 896 897 898 899
	spin_lock(&sdp->sd_rindex_spin);
	if (rgd->rd_free_clone >= al->al_requested) {
		al->al_rgd = rgd;
		ret = 1;
	}
	spin_unlock(&sdp->sd_rindex_spin);

	return ret;
}

900 901 902 903 904 905 906 907 908 909
/**
 * try_rgrp_unlink - Look for any unlinked, allocated, but unused inodes
 * @rgd: The rgrp
 *
 * Returns: The inode, if one has been found
 */

static struct inode *try_rgrp_unlink(struct gfs2_rgrpd *rgd, u64 *last_unlinked)
{
	struct inode *inode;
910
	u32 goal = 0, block;
911
	u64 no_addr;
912
	struct gfs2_sbd *sdp = rgd->rd_sbd;
913
	unsigned int n;
914 915

	for(;;) {
916 917
		if (goal >= rgd->rd_data)
			break;
918
		down_write(&sdp->sd_log_flush_lock);
919
		n = 1;
920
		block = rgblk_search(rgd, goal, GFS2_BLKST_UNLINKED,
921
				     GFS2_BLKST_UNLINKED, &n);
922
		up_write(&sdp->sd_log_flush_lock);
923
		if (block == BFITNOENT)
924
			break;
925 926 927
		/* rgblk_search can return a block < goal, so we need to
		   keep it marching forward. */
		no_addr = block + rgd->rd_data0;
928
		goal++;
929
		if (*last_unlinked != NO_BLOCK && no_addr <= *last_unlinked)
930
			continue;
931 932
		*last_unlinked = no_addr;
		inode = gfs2_inode_lookup(rgd->rd_sbd->sd_vfs, DT_UNKNOWN,
933
					  no_addr, -1, 1);
934 935 936 937 938 939 940 941
		if (!IS_ERR(inode))
			return inode;
	}

	rgd->rd_flags &= ~GFS2_RDF_CHECK;
	return NULL;
}

D
David Teigland 已提交
942 943 944 945 946 947 948
/**
 * recent_rgrp_next - get next RG from "recent" list
 * @cur_rgd: current rgrp
 *
 * Returns: The next rgrp in the recent list
 */

949
static struct gfs2_rgrpd *recent_rgrp_next(struct gfs2_rgrpd *cur_rgd)
D
David Teigland 已提交
950 951 952 953 954 955
{
	struct gfs2_sbd *sdp = cur_rgd->rd_sbd;
	struct list_head *head;
	struct gfs2_rgrpd *rgd;

	spin_lock(&sdp->sd_rindex_spin);
956 957 958 959
	head = &sdp->sd_rindex_mru_list;
	if (unlikely(cur_rgd->rd_list_mru.next == head)) {
		spin_unlock(&sdp->sd_rindex_spin);
		return NULL;
D
David Teigland 已提交
960
	}
961
	rgd = list_entry(cur_rgd->rd_list_mru.next, struct gfs2_rgrpd, rd_list_mru);
D
David Teigland 已提交
962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985
	spin_unlock(&sdp->sd_rindex_spin);
	return rgd;
}

/**
 * forward_rgrp_get - get an rgrp to try next from full list
 * @sdp: The GFS2 superblock
 *
 * Returns: The rgrp to try next
 */

static struct gfs2_rgrpd *forward_rgrp_get(struct gfs2_sbd *sdp)
{
	struct gfs2_rgrpd *rgd;
	unsigned int journals = gfs2_jindex_size(sdp);
	unsigned int rg = 0, x;

	spin_lock(&sdp->sd_rindex_spin);

	rgd = sdp->sd_rindex_forward;
	if (!rgd) {
		if (sdp->sd_rgrps >= journals)
			rg = sdp->sd_rgrps * sdp->sd_jdesc->jd_jid / journals;

S
Steven Whitehouse 已提交
986
		for (x = 0, rgd = gfs2_rgrpd_get_first(sdp); x < rg;
D
David Teigland 已提交
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
		     x++, rgd = gfs2_rgrpd_get_next(rgd))
			/* Do Nothing */;

		sdp->sd_rindex_forward = rgd;
	}

	spin_unlock(&sdp->sd_rindex_spin);

	return rgd;
}

/**
 * forward_rgrp_set - set the forward rgrp pointer
 * @sdp: the filesystem
 * @rgd: The new forward rgrp
 *
 */

static void forward_rgrp_set(struct gfs2_sbd *sdp, struct gfs2_rgrpd *rgd)
{
	spin_lock(&sdp->sd_rindex_spin);
	sdp->sd_rindex_forward = rgd;
	spin_unlock(&sdp->sd_rindex_spin);
}

/**
 * get_local_rgrp - Choose and lock a rgrp for allocation
 * @ip: the inode to reserve space for
 * @rgp: the chosen and locked rgrp
 *
 * Try to acquire rgrp in way which avoids contending with others.
 *
 * Returns: errno
 */

1022
static struct inode *get_local_rgrp(struct gfs2_inode *ip, u64 *last_unlinked)
D
David Teigland 已提交
1023
{
1024
	struct inode *inode = NULL;
1025
	struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode);
D
David Teigland 已提交
1026
	struct gfs2_rgrpd *rgd, *begin = NULL;
1027
	struct gfs2_alloc *al = ip->i_alloc;
D
David Teigland 已提交
1028 1029 1030
	int flags = LM_FLAG_TRY;
	int skipped = 0;
	int loops = 0;
1031
	int error, rg_locked;
D
David Teigland 已提交
1032

1033
	rgd = gfs2_blk2rgrpd(sdp, ip->i_goal);
D
David Teigland 已提交
1034 1035

	while (rgd) {
1036 1037 1038 1039 1040 1041 1042 1043 1044
		rg_locked = 0;

		if (gfs2_glock_is_locked_by_me(rgd->rd_gl)) {
			rg_locked = 1;
			error = 0;
		} else {
			error = gfs2_glock_nq_init(rgd->rd_gl, LM_ST_EXCLUSIVE,
						   LM_FLAG_TRY, &al->al_rgd_gh);
		}
D
David Teigland 已提交
1045 1046 1047 1048
		switch (error) {
		case 0:
			if (try_rgrp_fit(rgd, al))
				goto out;
1049 1050
			if (rgd->rd_flags & GFS2_RDF_CHECK)
				inode = try_rgrp_unlink(rgd, last_unlinked);
1051 1052
			if (!rg_locked)
				gfs2_glock_dq_uninit(&al->al_rgd_gh);
1053 1054
			if (inode)
				return inode;
1055
			/* fall through */
D
David Teigland 已提交
1056
		case GLR_TRYFAILED:
1057
			rgd = recent_rgrp_next(rgd);
D
David Teigland 已提交
1058 1059 1060
			break;

		default:
1061
			return ERR_PTR(error);
D
David Teigland 已提交
1062 1063 1064 1065 1066 1067 1068 1069
		}
	}

	/* Go through full list of rgrps */

	begin = rgd = forward_rgrp_get(sdp);

	for (;;) {
1070 1071 1072 1073 1074 1075 1076 1077 1078
		rg_locked = 0;

		if (gfs2_glock_is_locked_by_me(rgd->rd_gl)) {
			rg_locked = 1;
			error = 0;
		} else {
			error = gfs2_glock_nq_init(rgd->rd_gl, LM_ST_EXCLUSIVE, flags,
						   &al->al_rgd_gh);
		}
D
David Teigland 已提交
1079 1080 1081 1082
		switch (error) {
		case 0:
			if (try_rgrp_fit(rgd, al))
				goto out;
1083 1084
			if (rgd->rd_flags & GFS2_RDF_CHECK)
				inode = try_rgrp_unlink(rgd, last_unlinked);
1085 1086
			if (!rg_locked)
				gfs2_glock_dq_uninit(&al->al_rgd_gh);
1087 1088
			if (inode)
				return inode;
D
David Teigland 已提交
1089 1090 1091 1092 1093 1094 1095
			break;

		case GLR_TRYFAILED:
			skipped++;
			break;

		default:
1096
			return ERR_PTR(error);
D
David Teigland 已提交
1097 1098 1099 1100 1101 1102 1103
		}

		rgd = gfs2_rgrpd_get_next(rgd);
		if (!rgd)
			rgd = gfs2_rgrpd_get_first(sdp);

		if (rgd == begin) {
1104
			if (++loops >= 3)
1105
				return ERR_PTR(-ENOSPC);
1106 1107
			if (!skipped)
				loops++;
D
David Teigland 已提交
1108
			flags = 0;
1109 1110
			if (loops == 2)
				gfs2_log_flush(sdp, NULL);
D
David Teigland 已提交
1111 1112 1113
		}
	}

1114
out:
D
David Teigland 已提交
1115
	if (begin) {
1116 1117 1118
		spin_lock(&sdp->sd_rindex_spin);
		list_move(&rgd->rd_list_mru, &sdp->sd_rindex_mru_list);
		spin_unlock(&sdp->sd_rindex_spin);
D
David Teigland 已提交
1119 1120 1121 1122 1123 1124
		rgd = gfs2_rgrpd_get_next(rgd);
		if (!rgd)
			rgd = gfs2_rgrpd_get_first(sdp);
		forward_rgrp_set(sdp, rgd);
	}

1125
	return NULL;
D
David Teigland 已提交
1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136
}

/**
 * gfs2_inplace_reserve_i - Reserve space in the filesystem
 * @ip: the inode to reserve space for
 *
 * Returns: errno
 */

int gfs2_inplace_reserve_i(struct gfs2_inode *ip, char *file, unsigned int line)
{
1137
	struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode);
1138
	struct gfs2_alloc *al = ip->i_alloc;
1139
	struct inode *inode;
1140
	int error = 0;
1141
	u64 last_unlinked = NO_BLOCK;
D
David Teigland 已提交
1142 1143 1144 1145

	if (gfs2_assert_warn(sdp, al->al_requested))
		return -EINVAL;

1146
try_again:
1147 1148 1149 1150 1151
	/* We need to hold the rindex unless the inode we're using is
	   the rindex itself, in which case it's already held. */
	if (ip != GFS2_I(sdp->sd_rindex))
		error = gfs2_rindex_hold(sdp, &al->al_ri_gh);
	else if (!sdp->sd_rgrps) /* We may not have the rindex read in, so: */
1152
		error = gfs2_ri_update_special(ip);
1153

D
David Teigland 已提交
1154 1155 1156
	if (error)
		return error;

1157 1158
	inode = get_local_rgrp(ip, &last_unlinked);
	if (inode) {
1159 1160
		if (ip != GFS2_I(sdp->sd_rindex))
			gfs2_glock_dq_uninit(&al->al_ri_gh);
1161 1162 1163 1164 1165
		if (IS_ERR(inode))
			return PTR_ERR(inode);
		iput(inode);
		gfs2_log_flush(sdp, NULL);
		goto try_again;
D
David Teigland 已提交
1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 1179 1180 1181 1182
	}

	al->al_file = file;
	al->al_line = line;

	return 0;
}

/**
 * gfs2_inplace_release - release an inplace reservation
 * @ip: the inode the reservation was taken out on
 *
 * Release a reservation made by gfs2_inplace_reserve().
 */

void gfs2_inplace_release(struct gfs2_inode *ip)
{
1183
	struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode);
1184
	struct gfs2_alloc *al = ip->i_alloc;
D
David Teigland 已提交
1185 1186 1187 1188 1189 1190 1191 1192

	if (gfs2_assert_warn(sdp, al->al_alloced <= al->al_requested) == -1)
		fs_warn(sdp, "al_alloced = %u, al_requested = %u "
			     "al_file = %s, al_line = %u\n",
		             al->al_alloced, al->al_requested, al->al_file,
			     al->al_line);

	al->al_rgd = NULL;
1193 1194
	if (al->al_rgd_gh.gh_gl)
		gfs2_glock_dq_uninit(&al->al_rgd_gh);
1195 1196
	if (ip != GFS2_I(sdp->sd_rindex))
		gfs2_glock_dq_uninit(&al->al_ri_gh);
D
David Teigland 已提交
1197 1198 1199 1200 1201 1202 1203 1204 1205 1206
}

/**
 * gfs2_get_block_type - Check a block in a RG is of given type
 * @rgd: the resource group holding the block
 * @block: the block number
 *
 * Returns: The block type (GFS2_BLKST_*)
 */

1207
unsigned char gfs2_get_block_type(struct gfs2_rgrpd *rgd, u64 block)
D
David Teigland 已提交
1208 1209
{
	struct gfs2_bitmap *bi = NULL;
1210
	u32 length, rgrp_block, buf_block;
D
David Teigland 已提交
1211 1212 1213
	unsigned int buf;
	unsigned char type;

1214 1215
	length = rgd->rd_length;
	rgrp_block = block - rgd->rd_data0;
D
David Teigland 已提交
1216 1217 1218 1219 1220 1221 1222 1223 1224 1225

	for (buf = 0; buf < length; buf++) {
		bi = rgd->rd_bits + buf;
		if (rgrp_block < (bi->bi_start + bi->bi_len) * GFS2_NBBY)
			break;
	}

	gfs2_assert(rgd->rd_sbd, buf < length);
	buf_block = rgrp_block - bi->bi_start * GFS2_NBBY;

1226
	type = gfs2_testbit(rgd, bi->bi_bh->b_data + bi->bi_offset,
D
David Teigland 已提交
1227 1228 1229 1230 1231 1232 1233 1234 1235 1236 1237 1238
			   bi->bi_len, buf_block);

	return type;
}

/**
 * rgblk_search - find a block in @old_state, change allocation
 *           state to @new_state
 * @rgd: the resource group descriptor
 * @goal: the goal block within the RG (start here to search for avail block)
 * @old_state: GFS2_BLKST_XXX the before-allocation state to find
 * @new_state: GFS2_BLKST_XXX the after-allocation block state
1239
 * @n: The extent length
D
David Teigland 已提交
1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 1251 1252 1253
 *
 * Walk rgrp's bitmap to find bits that represent a block in @old_state.
 * Add the found bitmap buffer to the transaction.
 * Set the found bits to @new_state to change block's allocation state.
 *
 * This function never fails, because we wouldn't call it unless we
 * know (from reservation results, etc.) that a block is available.
 *
 * Scope of @goal and returned block is just within rgrp, not the whole
 * filesystem.
 *
 * Returns:  the block number allocated
 */

1254
static u32 rgblk_search(struct gfs2_rgrpd *rgd, u32 goal,
1255 1256
			unsigned char old_state, unsigned char new_state,
			unsigned int *n)
D
David Teigland 已提交
1257 1258
{
	struct gfs2_bitmap *bi = NULL;
1259
	const u32 length = rgd->rd_length;
1260
	u32 blk = 0;
D
David Teigland 已提交
1261
	unsigned int buf, x;
1262 1263
	const unsigned int elen = *n;
	const u8 *buffer;
D
David Teigland 已提交
1264

1265
	*n = 0;
D
David Teigland 已提交
1266 1267 1268 1269 1270 1271 1272 1273 1274 1275 1276 1277 1278 1279 1280 1281 1282 1283
	/* Find bitmap block that contains bits for goal block */
	for (buf = 0; buf < length; buf++) {
		bi = rgd->rd_bits + buf;
		if (goal < (bi->bi_start + bi->bi_len) * GFS2_NBBY)
			break;
	}

	gfs2_assert(rgd->rd_sbd, buf < length);

	/* Convert scope of "goal" from rgrp-wide to within found bit block */
	goal -= bi->bi_start * GFS2_NBBY;

	/* Search (up to entire) bitmap in this rgrp for allocatable block.
	   "x <= length", instead of "x < length", because we typically start
	   the search in the middle of a bit block, but if we can't find an
	   allocatable block anywhere else, we want to be able wrap around and
	   search in the first part of our first-searched bit block.  */
	for (x = 0; x <= length; x++) {
1284 1285
		/* The GFS2_BLKST_UNLINKED state doesn't apply to the clone
		   bitmaps, so we must search the originals for that. */
1286
		buffer = bi->bi_bh->b_data + bi->bi_offset;
1287
		if (old_state != GFS2_BLKST_UNLINKED && bi->bi_clone)
1288 1289 1290
			buffer = bi->bi_clone + bi->bi_offset;

		blk = gfs2_bitfit(buffer, bi->bi_len, goal, old_state);
D
David Teigland 已提交
1291 1292 1293 1294 1295 1296 1297 1298 1299
		if (blk != BFITNOENT)
			break;

		/* Try next bitmap block (wrap back to rgrp header if at end) */
		buf = (buf + 1) % length;
		bi = rgd->rd_bits + buf;
		goal = 0;
	}

1300
	if (blk != BFITNOENT && old_state != new_state) {
1301
		*n = 1;
1302
		gfs2_trans_add_bh(rgd->rd_gl, bi->bi_bh, 1);
1303
		gfs2_setbit(rgd, bi->bi_bh->b_data, bi->bi_clone, bi->bi_offset,
D
David Teigland 已提交
1304
			    bi->bi_len, blk, new_state);
1305 1306
		goal = blk;
		while (*n < elen) {
1307
			goal++;
1308
			if (goal >= (bi->bi_len * GFS2_NBBY))
1309 1310 1311 1312 1313
				break;
			if (gfs2_testbit(rgd, buffer, bi->bi_len, goal) !=
			    GFS2_BLKST_FREE)
				break;
			gfs2_setbit(rgd, bi->bi_bh->b_data, bi->bi_clone,
1314 1315 1316
				    bi->bi_offset, bi->bi_len, goal,
				    new_state);
			(*n)++;
1317
		}
1318
	}
D
David Teigland 已提交
1319

1320
	return (blk == BFITNOENT) ? blk : (bi->bi_start * GFS2_NBBY) + blk;
D
David Teigland 已提交
1321 1322 1323 1324 1325 1326 1327 1328 1329 1330 1331 1332
}

/**
 * rgblk_free - Change alloc state of given block(s)
 * @sdp: the filesystem
 * @bstart: the start of a run of blocks to free
 * @blen: the length of the block run (all must lie within ONE RG!)
 * @new_state: GFS2_BLKST_XXX the after-allocation block state
 *
 * Returns:  Resource group containing the block(s)
 */

1333 1334
static struct gfs2_rgrpd *rgblk_free(struct gfs2_sbd *sdp, u64 bstart,
				     u32 blen, unsigned char new_state)
D
David Teigland 已提交
1335 1336 1337
{
	struct gfs2_rgrpd *rgd;
	struct gfs2_bitmap *bi = NULL;
1338
	u32 length, rgrp_blk, buf_blk;
D
David Teigland 已提交
1339 1340 1341 1342 1343
	unsigned int buf;

	rgd = gfs2_blk2rgrpd(sdp, bstart);
	if (!rgd) {
		if (gfs2_consist(sdp))
1344
			fs_err(sdp, "block = %llu\n", (unsigned long long)bstart);
D
David Teigland 已提交
1345 1346 1347
		return NULL;
	}

1348
	length = rgd->rd_length;
D
David Teigland 已提交
1349

1350
	rgrp_blk = bstart - rgd->rd_data0;
D
David Teigland 已提交
1351 1352 1353 1354 1355 1356 1357 1358 1359 1360 1361 1362 1363 1364 1365

	while (blen--) {
		for (buf = 0; buf < length; buf++) {
			bi = rgd->rd_bits + buf;
			if (rgrp_blk < (bi->bi_start + bi->bi_len) * GFS2_NBBY)
				break;
		}

		gfs2_assert(rgd->rd_sbd, buf < length);

		buf_blk = rgrp_blk - bi->bi_start * GFS2_NBBY;
		rgrp_blk++;

		if (!bi->bi_clone) {
			bi->bi_clone = kmalloc(bi->bi_bh->b_size,
1366
					       GFP_NOFS | __GFP_NOFAIL);
D
David Teigland 已提交
1367 1368 1369 1370
			memcpy(bi->bi_clone + bi->bi_offset,
			       bi->bi_bh->b_data + bi->bi_offset,
			       bi->bi_len);
		}
1371
		gfs2_trans_add_bh(rgd->rd_gl, bi->bi_bh, 1);
1372
		gfs2_setbit(rgd, bi->bi_bh->b_data, NULL, bi->bi_offset,
D
David Teigland 已提交
1373 1374 1375 1376 1377 1378 1379
			    bi->bi_len, buf_blk, new_state);
	}

	return rgd;
}

/**
1380 1381
 * gfs2_alloc_block - Allocate a block
 * @ip: the inode to allocate the block for
D
David Teigland 已提交
1382 1383 1384 1385
 *
 * Returns: the allocated block
 */

1386
u64 gfs2_alloc_block(struct gfs2_inode *ip, unsigned int *n)
D
David Teigland 已提交
1387
{
1388
	struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode);
1389
	struct gfs2_alloc *al = ip->i_alloc;
D
David Teigland 已提交
1390
	struct gfs2_rgrpd *rgd = al->al_rgd;
1391 1392
	u32 goal, blk;
	u64 block;
D
David Teigland 已提交
1393

1394 1395
	if (rgrp_contains_block(rgd, ip->i_goal))
		goal = ip->i_goal - rgd->rd_data0;
D
David Teigland 已提交
1396
	else
1397
		goal = rgd->rd_last_alloc;
D
David Teigland 已提交
1398

1399
	blk = rgblk_search(rgd, goal, GFS2_BLKST_FREE, GFS2_BLKST_USED, n);
1400
	BUG_ON(blk == BFITNOENT);
D
David Teigland 已提交
1401

1402
	rgd->rd_last_alloc = blk;
1403
	block = rgd->rd_data0 + blk;
1404
	ip->i_goal = block;
D
David Teigland 已提交
1405

1406 1407
	gfs2_assert_withdraw(sdp, rgd->rd_free >= *n);
	rgd->rd_free -= *n;
D
David Teigland 已提交
1408

1409
	gfs2_trans_add_bh(rgd->rd_gl, rgd->rd_bits[0].bi_bh, 1);
1410
	gfs2_rgrp_out(rgd, rgd->rd_bits[0].bi_bh->b_data);
D
David Teigland 已提交
1411

1412
	al->al_alloced += *n;
D
David Teigland 已提交
1413

1414
	gfs2_statfs_change(sdp, 0, -(s64)*n, 0);
1415
	gfs2_quota_change(ip, *n, ip->i_inode.i_uid, ip->i_inode.i_gid);
D
David Teigland 已提交
1416 1417

	spin_lock(&sdp->sd_rindex_spin);
1418
	rgd->rd_free_clone -= *n;
D
David Teigland 已提交
1419 1420 1421 1422 1423 1424 1425 1426 1427 1428 1429 1430
	spin_unlock(&sdp->sd_rindex_spin);

	return block;
}

/**
 * gfs2_alloc_di - Allocate a dinode
 * @dip: the directory that the inode is going in
 *
 * Returns: the block allocated
 */

S
Steven Whitehouse 已提交
1431
u64 gfs2_alloc_di(struct gfs2_inode *dip, u64 *generation)
D
David Teigland 已提交
1432
{
1433
	struct gfs2_sbd *sdp = GFS2_SB(&dip->i_inode);
1434
	struct gfs2_alloc *al = dip->i_alloc;
D
David Teigland 已提交
1435
	struct gfs2_rgrpd *rgd = al->al_rgd;
S
Steven Whitehouse 已提交
1436 1437
	u32 blk;
	u64 block;
1438
	unsigned int n = 1;
D
David Teigland 已提交
1439

1440
	blk = rgblk_search(rgd, rgd->rd_last_alloc,
1441
			   GFS2_BLKST_FREE, GFS2_BLKST_DINODE, &n);
1442
	BUG_ON(blk == BFITNOENT);
D
David Teigland 已提交
1443

1444
	rgd->rd_last_alloc = blk;
D
David Teigland 已提交
1445

1446
	block = rgd->rd_data0 + blk;
D
David Teigland 已提交
1447

1448 1449
	gfs2_assert_withdraw(sdp, rgd->rd_free);
	rgd->rd_free--;
D
David Teigland 已提交
1450
	rgd->rd_rg.rg_dinodes++;
1451
	*generation = rgd->rd_igeneration++;
1452
	gfs2_trans_add_bh(rgd->rd_gl, rgd->rd_bits[0].bi_bh, 1);
1453
	gfs2_rgrp_out(rgd, rgd->rd_bits[0].bi_bh->b_data);
D
David Teigland 已提交
1454 1455 1456 1457

	al->al_alloced++;

	gfs2_statfs_change(sdp, 0, -1, +1);
1458
	gfs2_trans_add_unrevoke(sdp, block, 1);
D
David Teigland 已提交
1459 1460 1461 1462 1463 1464 1465 1466 1467 1468 1469 1470 1471 1472 1473 1474

	spin_lock(&sdp->sd_rindex_spin);
	rgd->rd_free_clone--;
	spin_unlock(&sdp->sd_rindex_spin);

	return block;
}

/**
 * gfs2_free_data - free a contiguous run of data block(s)
 * @ip: the inode these blocks are being freed from
 * @bstart: first block of a run of contiguous blocks
 * @blen: the length of the block run
 *
 */

1475
void gfs2_free_data(struct gfs2_inode *ip, u64 bstart, u32 blen)
D
David Teigland 已提交
1476
{
1477
	struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode);
D
David Teigland 已提交
1478 1479 1480 1481 1482 1483
	struct gfs2_rgrpd *rgd;

	rgd = rgblk_free(sdp, bstart, blen, GFS2_BLKST_FREE);
	if (!rgd)
		return;

1484
	rgd->rd_free += blen;
D
David Teigland 已提交
1485

1486
	gfs2_trans_add_bh(rgd->rd_gl, rgd->rd_bits[0].bi_bh, 1);
1487
	gfs2_rgrp_out(rgd, rgd->rd_bits[0].bi_bh->b_data);
D
David Teigland 已提交
1488 1489 1490 1491

	gfs2_trans_add_rg(rgd);

	gfs2_statfs_change(sdp, 0, +blen, 0);
1492
	gfs2_quota_change(ip, -(s64)blen, ip->i_inode.i_uid, ip->i_inode.i_gid);
D
David Teigland 已提交
1493 1494 1495 1496 1497 1498 1499 1500 1501 1502
}

/**
 * gfs2_free_meta - free a contiguous run of data block(s)
 * @ip: the inode these blocks are being freed from
 * @bstart: first block of a run of contiguous blocks
 * @blen: the length of the block run
 *
 */

1503
void gfs2_free_meta(struct gfs2_inode *ip, u64 bstart, u32 blen)
D
David Teigland 已提交
1504
{
1505
	struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode);
D
David Teigland 已提交
1506 1507 1508 1509 1510 1511
	struct gfs2_rgrpd *rgd;

	rgd = rgblk_free(sdp, bstart, blen, GFS2_BLKST_FREE);
	if (!rgd)
		return;

1512
	rgd->rd_free += blen;
D
David Teigland 已提交
1513

1514
	gfs2_trans_add_bh(rgd->rd_gl, rgd->rd_bits[0].bi_bh, 1);
1515
	gfs2_rgrp_out(rgd, rgd->rd_bits[0].bi_bh->b_data);
D
David Teigland 已提交
1516 1517 1518 1519

	gfs2_trans_add_rg(rgd);

	gfs2_statfs_change(sdp, 0, +blen, 0);
1520
	gfs2_quota_change(ip, -(s64)blen, ip->i_inode.i_uid, ip->i_inode.i_gid);
D
David Teigland 已提交
1521 1522 1523
	gfs2_meta_wipe(ip, bstart, blen);
}

1524 1525 1526 1527 1528
void gfs2_unlink_di(struct inode *inode)
{
	struct gfs2_inode *ip = GFS2_I(inode);
	struct gfs2_sbd *sdp = GFS2_SB(inode);
	struct gfs2_rgrpd *rgd;
1529
	u64 blkno = ip->i_no_addr;
1530 1531 1532 1533 1534

	rgd = rgblk_free(sdp, blkno, 1, GFS2_BLKST_UNLINKED);
	if (!rgd)
		return;
	gfs2_trans_add_bh(rgd->rd_gl, rgd->rd_bits[0].bi_bh, 1);
1535
	gfs2_rgrp_out(rgd, rgd->rd_bits[0].bi_bh->b_data);
1536 1537 1538
	gfs2_trans_add_rg(rgd);
}

1539
static void gfs2_free_uninit_di(struct gfs2_rgrpd *rgd, u64 blkno)
D
David Teigland 已提交
1540 1541 1542 1543 1544 1545 1546 1547 1548 1549 1550 1551
{
	struct gfs2_sbd *sdp = rgd->rd_sbd;
	struct gfs2_rgrpd *tmp_rgd;

	tmp_rgd = rgblk_free(sdp, blkno, 1, GFS2_BLKST_FREE);
	if (!tmp_rgd)
		return;
	gfs2_assert_withdraw(sdp, rgd == tmp_rgd);

	if (!rgd->rd_rg.rg_dinodes)
		gfs2_consist_rgrpd(rgd);
	rgd->rd_rg.rg_dinodes--;
1552
	rgd->rd_free++;
D
David Teigland 已提交
1553

1554
	gfs2_trans_add_bh(rgd->rd_gl, rgd->rd_bits[0].bi_bh, 1);
1555
	gfs2_rgrp_out(rgd, rgd->rd_bits[0].bi_bh->b_data);
D
David Teigland 已提交
1556 1557 1558 1559 1560 1561 1562 1563

	gfs2_statfs_change(sdp, 0, +1, -1);
	gfs2_trans_add_rg(rgd);
}


void gfs2_free_di(struct gfs2_rgrpd *rgd, struct gfs2_inode *ip)
{
1564
	gfs2_free_uninit_di(rgd, ip->i_no_addr);
1565
	gfs2_quota_change(ip, -1, ip->i_inode.i_uid, ip->i_inode.i_gid);
1566
	gfs2_meta_wipe(ip, ip->i_no_addr, 1);
D
David Teigland 已提交
1567 1568 1569 1570 1571 1572 1573 1574 1575 1576 1577 1578 1579 1580 1581
}

/**
 * gfs2_rlist_add - add a RG to a list of RGs
 * @sdp: the filesystem
 * @rlist: the list of resource groups
 * @block: the block
 *
 * Figure out what RG a block belongs to and add that RG to the list
 *
 * FIXME: Don't use NOFAIL
 *
 */

void gfs2_rlist_add(struct gfs2_sbd *sdp, struct gfs2_rgrp_list *rlist,
1582
		    u64 block)
D
David Teigland 已提交
1583 1584 1585 1586 1587 1588 1589 1590 1591 1592 1593 1594
{
	struct gfs2_rgrpd *rgd;
	struct gfs2_rgrpd **tmp;
	unsigned int new_space;
	unsigned int x;

	if (gfs2_assert_warn(sdp, !rlist->rl_ghs))
		return;

	rgd = gfs2_blk2rgrpd(sdp, block);
	if (!rgd) {
		if (gfs2_consist(sdp))
1595
			fs_err(sdp, "block = %llu\n", (unsigned long long)block);
D
David Teigland 已提交
1596 1597 1598 1599 1600 1601 1602 1603 1604 1605 1606
		return;
	}

	for (x = 0; x < rlist->rl_rgrps; x++)
		if (rlist->rl_rgd[x] == rgd)
			return;

	if (rlist->rl_rgrps == rlist->rl_space) {
		new_space = rlist->rl_space + 10;

		tmp = kcalloc(new_space, sizeof(struct gfs2_rgrpd *),
1607
			      GFP_NOFS | __GFP_NOFAIL);
D
David Teigland 已提交
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

		if (rlist->rl_rgd) {
			memcpy(tmp, rlist->rl_rgd,
			       rlist->rl_space * sizeof(struct gfs2_rgrpd *));
			kfree(rlist->rl_rgd);
		}

		rlist->rl_space = new_space;
		rlist->rl_rgd = tmp;
	}

	rlist->rl_rgd[rlist->rl_rgrps++] = rgd;
}

/**
 * gfs2_rlist_alloc - all RGs have been added to the rlist, now allocate
 *      and initialize an array of glock holders for them
 * @rlist: the list of resource groups
 * @state: the lock state to acquire the RG lock in
 * @flags: the modifier flags for the holder structures
 *
 * FIXME: Don't use NOFAIL
 *
 */

1633
void gfs2_rlist_alloc(struct gfs2_rgrp_list *rlist, unsigned int state)
D
David Teigland 已提交
1634 1635 1636 1637
{
	unsigned int x;

	rlist->rl_ghs = kcalloc(rlist->rl_rgrps, sizeof(struct gfs2_holder),
1638
				GFP_NOFS | __GFP_NOFAIL);
D
David Teigland 已提交
1639 1640
	for (x = 0; x < rlist->rl_rgrps; x++)
		gfs2_holder_init(rlist->rl_rgd[x]->rd_gl,
1641
				state, 0,
D
David Teigland 已提交
1642 1643 1644 1645 1646 1647 1648 1649 1650 1651 1652 1653 1654 1655 1656 1657 1658 1659 1660 1661 1662 1663
				&rlist->rl_ghs[x]);
}

/**
 * gfs2_rlist_free - free a resource group list
 * @list: the list of resource groups
 *
 */

void gfs2_rlist_free(struct gfs2_rgrp_list *rlist)
{
	unsigned int x;

	kfree(rlist->rl_rgd);

	if (rlist->rl_ghs) {
		for (x = 0; x < rlist->rl_rgrps; x++)
			gfs2_holder_uninit(&rlist->rl_ghs[x]);
		kfree(rlist->rl_ghs);
	}
}