rgrp.c 45.5 KB
Newer Older
D
David Teigland 已提交
1 2
/*
 * Copyright (C) Sistina Software, Inc.  1997-2003 All rights reserved.
3
 * Copyright (C) 2004-2008 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/prefetch.h>
17
#include <linux/blkdev.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"
S
Steven Whitehouse 已提交
32
#include "trace_gfs2.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
	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;
}

134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155
/**
 * gfs2_bit_search
 * @ptr: Pointer to bitmap data
 * @mask: Mask to use (normally 0x55555.... but adjusted for search start)
 * @state: The state we are searching for
 *
 * We xor the bitmap data with a patter which is the bitwise opposite
 * of what we are looking for, this gives rise to a pattern of ones
 * wherever there is a match. Since we have two bits per entry, we
 * take this pattern, shift it down by one place and then and it with
 * the original. All the even bit positions (0,2,4, etc) then represent
 * successful matches, so we mask with 0x55555..... to remove the unwanted
 * odd bit positions.
 *
 * This allows searching of a whole u64 at once (32 blocks) with a
 * single test (on 64 bit arches).
 */

static inline u64 gfs2_bit_search(const __le64 *ptr, u64 mask, u8 state)
{
	u64 tmp;
	static const u64 search[] = {
156 157 158 159
		[0] = 0xffffffffffffffffULL,
		[1] = 0xaaaaaaaaaaaaaaaaULL,
		[2] = 0x5555555555555555ULL,
		[3] = 0x0000000000000000ULL,
160 161 162 163 164 165 166
	};
	tmp = le64_to_cpu(*ptr) ^ search[state];
	tmp &= (tmp >> 1);
	tmp &= mask;
	return tmp;
}

167 168 169 170
/**
 * 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
171
 * @len: the length (in bytes) of the buffer
172
 * @goal: start search at this block's bit-pair (within @buffer)
173
 * @state: GFS2_BLKST_XXX the state of the block we're looking for.
174 175 176
 *
 * 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
177 178 179 180 181
 * beginning of a bitmap block buffer, skipping any header structures, but
 * headers are always a multiple of 64 bits long so that the buffer is
 * always aligned to a 64 bit boundary.
 *
 * The size of the buffer is in bytes, but is it assumed that it is
182
 * always ok to read a complete multiple of 64 bits at the end
183
 * of the block in case the end is no aligned to a natural boundary.
184 185 186 187
 *
 * Return: the block number (bitmap buffer scope) that was found
 */

188 189
static u32 gfs2_bitfit(const u8 *buf, const unsigned int len,
		       u32 goal, u8 state)
190
{
191 192 193 194
	u32 spoint = (goal << 1) & ((8*sizeof(u64)) - 1);
	const __le64 *ptr = ((__le64 *)buf) + (goal >> 5);
	const __le64 *end = (__le64 *)(buf + ALIGN(len, sizeof(u64)));
	u64 tmp;
195
	u64 mask = 0x5555555555555555ULL;
196 197 198 199 200 201 202 203 204
	u32 bit;

	BUG_ON(state > 3);

	/* Mask off bits we don't care about at the start of the search */
	mask <<= spoint;
	tmp = gfs2_bit_search(ptr, mask, state);
	ptr++;
	while(tmp == 0 && ptr < end) {
205
		tmp = gfs2_bit_search(ptr, 0x5555555555555555ULL, state);
206
		ptr++;
207
	}
208 209 210 211 212 213 214
	/* Mask off any bits which are more than len bytes from the start */
	if (ptr == end && (len & (sizeof(u64) - 1)))
		tmp &= (((u64)~0) >> (64 - 8*(len & (sizeof(u64) - 1))));
	/* Didn't find anything, so return */
	if (tmp == 0)
		return BFITNOENT;
	ptr--;
215
	bit = __ffs64(tmp);
216 217
	bit /= 2;	/* two bits per entry in the bitmap */
	return (((const unsigned char *)ptr - buf) * GFS2_NBBY) + bit;
218 219 220 221 222 223 224 225 226 227 228
}

/**
 * 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
 */

229 230
static u32 gfs2_bitcount(struct gfs2_rgrpd *rgd, const u8 *buffer,
			 unsigned int buflen, u8 state)
231
{
232 233 234 235 236
	const u8 *byte = buffer;
	const u8 *end = buffer + buflen;
	const u8 state1 = state << 2;
	const u8 state2 = state << 4;
	const u8 state3 = state << 6;
237
	u32 count = 0;
238 239 240 241 242 243 244 245 246 247 248 249 250 251 252

	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
253 254 255 256 257 258 259 260 261 262 263
/**
 * 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;
264
	u32 length = rgd->rd_length;
265
	u32 count[4], tmp;
D
David Teigland 已提交
266 267
	int buf, x;

268
	memset(count, 0, 4 * sizeof(u32));
D
David Teigland 已提交
269 270 271 272 273 274 275 276 277 278 279

	/* 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);
	}

280
	if (count[0] != rgd->rd_free) {
D
David Teigland 已提交
281 282
		if (gfs2_consist_rgrpd(rgd))
			fs_err(sdp, "free data mismatch:  %u != %u\n",
283
			       count[0], rgd->rd_free);
D
David Teigland 已提交
284 285 286
		return;
	}

287
	tmp = rgd->rd_data - rgd->rd_free - rgd->rd_dinodes;
288
	if (count[1] != tmp) {
D
David Teigland 已提交
289 290 291 292 293 294
		if (gfs2_consist_rgrpd(rgd))
			fs_err(sdp, "used data mismatch:  %u != %u\n",
			       count[1], tmp);
		return;
	}

295
	if (count[2] + count[3] != rgd->rd_dinodes) {
D
David Teigland 已提交
296
		if (gfs2_consist_rgrpd(rgd))
297
			fs_err(sdp, "used metadata mismatch:  %u != %u\n",
298
			       count[2] + count[3], rgd->rd_dinodes);
D
David Teigland 已提交
299 300 301 302
		return;
	}
}

303
static inline int rgrp_contains_block(struct gfs2_rgrpd *rgd, u64 block)
D
David Teigland 已提交
304
{
305 306
	u64 first = rgd->rd_data0;
	u64 last = first + rgd->rd_data;
307
	return first <= block && block < last;
D
David Teigland 已提交
308 309 310 311 312 313 314 315 316 317
}

/**
 * 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
 */

318
struct gfs2_rgrpd *gfs2_blk2rgrpd(struct gfs2_sbd *sdp, u64 blk)
D
David Teigland 已提交
319 320 321 322 323 324
{
	struct gfs2_rgrpd *rgd;

	spin_lock(&sdp->sd_rindex_spin);

	list_for_each_entry(rgd, &sdp->sd_rindex_mru_list, rd_list_mru) {
325
		if (rgrp_contains_block(rgd, blk)) {
D
David Teigland 已提交
326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382
			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) {
383
			gl->gl_object = NULL;
D
David Teigland 已提交
384 385 386 387
			gfs2_glock_put(gl);
		}

		kfree(rgd->rd_bits);
388
		kmem_cache_free(gfs2_rgrpd_cachep, rgd);
D
David Teigland 已提交
389 390 391 392 393
	}
}

void gfs2_clear_rgrpd(struct gfs2_sbd *sdp)
{
394
	mutex_lock(&sdp->sd_rindex_mutex);
D
David Teigland 已提交
395
	clear_rgrpdi(sdp);
396
	mutex_unlock(&sdp->sd_rindex_mutex);
D
David Teigland 已提交
397 398
}

399 400 401 402 403 404 405 406 407
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
408 409 410 411 412 413 414 415 416 417 418 419 420
/**
 * 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;
421
	u32 length = rgd->rd_length; /* # blocks in hdr & bitmap */
422
	u32 bytes_left, bytes;
D
David Teigland 已提交
423 424
	int x;

425 426 427
	if (!length)
		return -EINVAL;

428
	rgd->rd_bits = kcalloc(length, sizeof(struct gfs2_bitmap), GFP_NOFS);
D
David Teigland 已提交
429 430 431
	if (!rgd->rd_bits)
		return -ENOMEM;

432
	bytes_left = rgd->rd_bitbytes;
D
David Teigland 已提交
433 434 435 436

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

437
		bi->bi_flags = 0;
D
David Teigland 已提交
438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453
		/* 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);
454
			bi->bi_start = rgd->rd_bitbytes - bytes_left;
D
David Teigland 已提交
455 456 457
			bi->bi_len = bytes;
		/* other blocks */
		} else {
458 459
			bytes = sdp->sd_sb.sb_bsize -
				sizeof(struct gfs2_meta_header);
D
David Teigland 已提交
460
			bi->bi_offset = sizeof(struct gfs2_meta_header);
461
			bi->bi_start = rgd->rd_bitbytes - bytes_left;
D
David Teigland 已提交
462 463 464 465 466 467 468 469 470 471 472
			bi->bi_len = bytes;
		}

		bytes_left -= bytes;
	}

	if (bytes_left) {
		gfs2_consist_rgrpd(rgd);
		return -EIO;
	}
	bi = rgd->rd_bits + (length - 1);
473
	if ((bi->bi_start + bi->bi_len) * GFS2_NBBY != rgd->rd_data) {
D
David Teigland 已提交
474
		if (gfs2_consist_rgrpd(rgd)) {
475
			gfs2_rindex_print(rgd);
D
David Teigland 已提交
476 477 478 479 480 481 482 483 484
			fs_err(sdp, "start=%u len=%u offset=%u\n",
			       bi->bi_start, bi->bi_len, bi->bi_offset);
		}
		return -EIO;
	}

	return 0;
}

485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502
/**
 * 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);

503
		if (pos + sizeof(struct gfs2_rindex) >= ip->i_disksize)
504 505 506 507 508
			break;
		error = gfs2_internal_read(ip, &ra_state, buf, &pos,
					   sizeof(struct gfs2_rindex));
		if (error != sizeof(struct gfs2_rindex))
			break;
509
		total_data += be32_to_cpu(((struct gfs2_rindex *)buf)->ri_data);
510 511 512 513 514
	}
	mutex_unlock(&sdp->sd_rindex_mutex);
	return total_data;
}

515 516 517 518 519 520 521 522 523 524 525
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
526
/**
527
 * read_rindex_entry - Pull in a new resource index entry from the disk
D
David Teigland 已提交
528 529
 * @gl: The glock covering the rindex inode
 *
530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551
 * 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;
	}

552
	rgd = kmem_cache_zalloc(gfs2_rgrpd_cachep, GFP_NOFS);
553 554 555 556 557 558 559 560 561 562 563
	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);

564
	gfs2_rindex_in(rgd, buf);
565 566 567 568
	error = compute_bitstructs(rgd);
	if (error)
		return error;

569
	error = gfs2_glock_get(sdp, rgd->rd_addr,
570 571 572 573 574
			       &gfs2_rgrp_glops, CREATE, &rgd->rd_gl);
	if (error)
		return error;

	rgd->rd_gl->gl_object = rgd;
575
	rgd->rd_flags &= ~GFS2_RDF_UPTODATE;
576 577 578 579 580 581 582
	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
583 584 585 586 587
 * Returns: 0 on successful update, error code otherwise
 */

static int gfs2_ri_update(struct gfs2_inode *ip)
{
588 589
	struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode);
	struct inode *inode = &ip->i_inode;
590
	struct file_ra_state ra_state;
591
	u64 rgrp_count = ip->i_disksize;
D
David Teigland 已提交
592 593
	int error;

S
Steven Whitehouse 已提交
594
	do_div(rgrp_count, sizeof(struct gfs2_rindex));
D
David Teigland 已提交
595 596
	clear_rgrpdi(sdp);

597
	file_ra_state_init(&ra_state, inode->i_mapping);
598
	for (sdp->sd_rgrps = 0; sdp->sd_rgrps < rgrp_count; sdp->sd_rgrps++) {
599 600 601 602
		error = read_rindex_entry(ip, &ra_state);
		if (error) {
			clear_rgrpdi(sdp);
			return error;
D
David Teigland 已提交
603
		}
604
	}
D
David Teigland 已提交
605

606
	sdp->sd_rindex_uptodate = 1;
607 608
	return 0;
}
D
David Teigland 已提交
609

610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625
/**
 * 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
626

627 628 629 630
	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) >
631
		    ip->i_disksize)
632 633 634 635 636 637
			break;
		error = read_rindex_entry(ip, &ra_state);
		if (error) {
			clear_rgrpdi(sdp);
			return error;
		}
D
David Teigland 已提交
638 639
	}

640
	sdp->sd_rindex_uptodate = 1;
D
David Teigland 已提交
641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663
	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)
{
664
	struct gfs2_inode *ip = GFS2_I(sdp->sd_rindex);
D
David Teigland 已提交
665 666 667 668 669 670 671 672
	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 */
673
	if (!sdp->sd_rindex_uptodate) {
674
		mutex_lock(&sdp->sd_rindex_mutex);
675
		if (!sdp->sd_rindex_uptodate) {
D
David Teigland 已提交
676 677 678 679
			error = gfs2_ri_update(ip);
			if (error)
				gfs2_glock_dq_uninit(ri_gh);
		}
680
		mutex_unlock(&sdp->sd_rindex_mutex);
D
David Teigland 已提交
681 682 683 684 685
	}

	return error;
}

686
static void gfs2_rgrp_in(struct gfs2_rgrpd *rgd, const void *buf)
687 688
{
	const struct gfs2_rgrp *str = buf;
689
	u32 rg_flags;
690

691
	rg_flags = be32_to_cpu(str->rg_flags);
692
	rg_flags &= ~GFS2_RDF_MASK;
693 694
	rgd->rd_flags &= GFS2_RDF_MASK;
	rgd->rd_flags |= rg_flags;
695
	rgd->rd_free = be32_to_cpu(str->rg_free);
696
	rgd->rd_dinodes = be32_to_cpu(str->rg_dinodes);
697
	rgd->rd_igeneration = be64_to_cpu(str->rg_igeneration);
698 699
}

700
static void gfs2_rgrp_out(struct gfs2_rgrpd *rgd, void *buf)
701 702 703
{
	struct gfs2_rgrp *str = buf;

704
	str->rg_flags = cpu_to_be32(rgd->rd_flags & ~GFS2_RDF_MASK);
705
	str->rg_free = cpu_to_be32(rgd->rd_free);
706
	str->rg_dinodes = cpu_to_be32(rgd->rd_dinodes);
707
	str->__pad = cpu_to_be32(0);
708
	str->rg_igeneration = cpu_to_be64(rgd->rd_igeneration);
709 710 711
	memset(&str->rg_reserved, 0, sizeof(str->rg_reserved));
}

D
David Teigland 已提交
712 713 714 715 716 717 718 719 720 721 722 723 724 725
/**
 * 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;
726
	unsigned int length = rgd->rd_length;
D
David Teigland 已提交
727 728 729 730
	struct gfs2_bitmap *bi;
	unsigned int x, y;
	int error;

731
	mutex_lock(&rgd->rd_mutex);
D
David Teigland 已提交
732 733 734 735 736

	spin_lock(&sdp->sd_rindex_spin);
	if (rgd->rd_bh_count) {
		rgd->rd_bh_count++;
		spin_unlock(&sdp->sd_rindex_spin);
737
		mutex_unlock(&rgd->rd_mutex);
D
David Teigland 已提交
738 739 740 741 742 743
		return 0;
	}
	spin_unlock(&sdp->sd_rindex_spin);

	for (x = 0; x < length; x++) {
		bi = rgd->rd_bits + x;
744
		error = gfs2_meta_read(gl, rgd->rd_addr + x, 0, &bi->bi_bh);
D
David Teigland 已提交
745 746 747 748 749 750
		if (error)
			goto fail;
	}

	for (y = length; y--;) {
		bi = rgd->rd_bits + y;
S
Steven Whitehouse 已提交
751
		error = gfs2_meta_wait(sdp, bi->bi_bh);
D
David Teigland 已提交
752 753
		if (error)
			goto fail;
754
		if (gfs2_metatype_check(sdp, bi->bi_bh, y ? GFS2_METATYPE_RB :
D
David Teigland 已提交
755 756 757 758 759 760
					      GFS2_METATYPE_RG)) {
			error = -EIO;
			goto fail;
		}
	}

761
	if (!(rgd->rd_flags & GFS2_RDF_UPTODATE)) {
762 763
		for (x = 0; x < length; x++)
			clear_bit(GBF_FULL, &rgd->rd_bits[x].bi_flags);
764
		gfs2_rgrp_in(rgd, (rgd->rd_bits[0].bi_bh)->b_data);
765
		rgd->rd_flags |= (GFS2_RDF_UPTODATE | GFS2_RDF_CHECK);
D
David Teigland 已提交
766 767 768
	}

	spin_lock(&sdp->sd_rindex_spin);
769
	rgd->rd_free_clone = rgd->rd_free;
D
David Teigland 已提交
770 771 772
	rgd->rd_bh_count++;
	spin_unlock(&sdp->sd_rindex_spin);

773
	mutex_unlock(&rgd->rd_mutex);
D
David Teigland 已提交
774 775 776

	return 0;

777
fail:
D
David Teigland 已提交
778 779 780 781 782 783
	while (x--) {
		bi = rgd->rd_bits + x;
		brelse(bi->bi_bh);
		bi->bi_bh = NULL;
		gfs2_assert_warn(sdp, !bi->bi_clone);
	}
784
	mutex_unlock(&rgd->rd_mutex);
D
David Teigland 已提交
785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807

	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;
808
	int x, length = rgd->rd_length;
D
David Teigland 已提交
809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827

	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);
}

828 829 830 831 832 833
static void gfs2_rgrp_send_discards(struct gfs2_sbd *sdp, u64 offset,
				    const struct gfs2_bitmap *bi)
{
	struct super_block *sb = sdp->sd_vfs;
	struct block_device *bdev = sb->s_bdev;
	const unsigned int sects_per_blk = sdp->sd_sb.sb_bsize /
834
					   bdev_logical_block_size(sb->s_bdev);
835
	u64 blk;
836
	sector_t start = 0;
837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855
	sector_t nr_sects = 0;
	int rv;
	unsigned int x;

	for (x = 0; x < bi->bi_len; x++) {
		const u8 *orig = bi->bi_bh->b_data + bi->bi_offset + x;
		const u8 *clone = bi->bi_clone + bi->bi_offset + x;
		u8 diff = ~(*orig | (*orig >> 1)) & (*clone | (*clone >> 1));
		diff &= 0x55;
		if (diff == 0)
			continue;
		blk = offset + ((bi->bi_start + x) * GFS2_NBBY);
		blk *= sects_per_blk; /* convert to sectors */
		while(diff) {
			if (diff & 1) {
				if (nr_sects == 0)
					goto start_new_extent;
				if ((start + nr_sects) != blk) {
					rv = blkdev_issue_discard(bdev, start,
856 857
							    nr_sects, GFP_NOFS,
							    DISCARD_FL_BARRIER);
858 859 860 861 862 863 864 865 866 867 868 869 870
					if (rv)
						goto fail;
					nr_sects = 0;
start_new_extent:
					start = blk;
				}
				nr_sects += sects_per_blk;
			}
			diff >>= 2;
			blk += sects_per_blk;
		}
	}
	if (nr_sects) {
871 872
		rv = blkdev_issue_discard(bdev, start, nr_sects, GFP_NOFS,
					 DISCARD_FL_BARRIER);
873 874 875 876 877 878 879 880 881
		if (rv)
			goto fail;
	}
	return;
fail:
	fs_warn(sdp, "error %d on discard request, turning discards off for this filesystem", rv);
	sdp->sd_args.ar_discard = 0;
}

D
David Teigland 已提交
882 883 884
void gfs2_rgrp_repolish_clones(struct gfs2_rgrpd *rgd)
{
	struct gfs2_sbd *sdp = rgd->rd_sbd;
885
	unsigned int length = rgd->rd_length;
D
David Teigland 已提交
886 887 888 889 890 891
	unsigned int x;

	for (x = 0; x < length; x++) {
		struct gfs2_bitmap *bi = rgd->rd_bits + x;
		if (!bi->bi_clone)
			continue;
892 893
		if (sdp->sd_args.ar_discard)
			gfs2_rgrp_send_discards(sdp, rgd->rd_data0, bi);
894
		clear_bit(GBF_FULL, &bi->bi_flags);
D
David Teigland 已提交
895
		memcpy(bi->bi_clone + bi->bi_offset,
896
		       bi->bi_bh->b_data + bi->bi_offset, bi->bi_len);
D
David Teigland 已提交
897 898 899
	}

	spin_lock(&sdp->sd_rindex_spin);
900
	rgd->rd_free_clone = rgd->rd_free;
D
David Teigland 已提交
901 902 903 904 905 906 907 908 909 910 911 912
	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)
{
913
	BUG_ON(ip->i_alloc != NULL);
914
	ip->i_alloc = kzalloc(sizeof(struct gfs2_alloc), GFP_NOFS);
915
	return ip->i_alloc;
D
David Teigland 已提交
916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933
}

/**
 * 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;

934
	if (rgd->rd_flags & (GFS2_RGF_NOALLOC | GFS2_RDF_ERROR))
935 936
		return 0;

D
David Teigland 已提交
937 938 939 940 941 942 943 944 945 946
	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;
}

947 948 949 950 951 952 953
/**
 * try_rgrp_unlink - Look for any unlinked, allocated, but unused inodes
 * @rgd: The rgrp
 *
 * Returns: The inode, if one has been found
 */

954 955
static struct inode *try_rgrp_unlink(struct gfs2_rgrpd *rgd, u64 *last_unlinked,
				     u64 skip)
956 957
{
	struct inode *inode;
958
	u32 goal = 0, block;
959
	u64 no_addr;
960
	struct gfs2_sbd *sdp = rgd->rd_sbd;
961
	unsigned int n;
962 963

	for(;;) {
964 965
		if (goal >= rgd->rd_data)
			break;
966
		down_write(&sdp->sd_log_flush_lock);
967
		n = 1;
968
		block = rgblk_search(rgd, goal, GFS2_BLKST_UNLINKED,
969
				     GFS2_BLKST_UNLINKED, &n);
970
		up_write(&sdp->sd_log_flush_lock);
971
		if (block == BFITNOENT)
972
			break;
973 974 975
		/* rgblk_search can return a block < goal, so we need to
		   keep it marching forward. */
		no_addr = block + rgd->rd_data0;
976
		goal++;
977
		if (*last_unlinked != NO_BLOCK && no_addr <= *last_unlinked)
978
			continue;
979 980
		if (no_addr == skip)
			continue;
981 982
		*last_unlinked = no_addr;
		inode = gfs2_inode_lookup(rgd->rd_sbd->sd_vfs, DT_UNKNOWN,
983
					  no_addr, -1, 1);
984 985 986 987 988 989 990 991
		if (!IS_ERR(inode))
			return inode;
	}

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

D
David Teigland 已提交
992 993 994 995 996 997 998
/**
 * recent_rgrp_next - get next RG from "recent" list
 * @cur_rgd: current rgrp
 *
 * Returns: The next rgrp in the recent list
 */

999
static struct gfs2_rgrpd *recent_rgrp_next(struct gfs2_rgrpd *cur_rgd)
D
David Teigland 已提交
1000 1001 1002 1003 1004 1005
{
	struct gfs2_sbd *sdp = cur_rgd->rd_sbd;
	struct list_head *head;
	struct gfs2_rgrpd *rgd;

	spin_lock(&sdp->sd_rindex_spin);
1006 1007 1008 1009
	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
1010
	}
1011
	rgd = list_entry(cur_rgd->rd_list_mru.next, struct gfs2_rgrpd, rd_list_mru);
D
David Teigland 已提交
1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035
	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
1036
		for (x = 0, rgd = gfs2_rgrpd_get_first(sdp); x < rg;
D
David Teigland 已提交
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
		     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
 */

1072
static struct inode *get_local_rgrp(struct gfs2_inode *ip, u64 *last_unlinked)
D
David Teigland 已提交
1073
{
1074
	struct inode *inode = NULL;
1075
	struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode);
D
David Teigland 已提交
1076
	struct gfs2_rgrpd *rgd, *begin = NULL;
1077
	struct gfs2_alloc *al = ip->i_alloc;
D
David Teigland 已提交
1078 1079 1080
	int flags = LM_FLAG_TRY;
	int skipped = 0;
	int loops = 0;
1081
	int error, rg_locked;
D
David Teigland 已提交
1082

1083
	rgd = gfs2_blk2rgrpd(sdp, ip->i_goal);
D
David Teigland 已提交
1084 1085

	while (rgd) {
1086 1087 1088 1089 1090 1091 1092 1093 1094
		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
1095 1096 1097 1098
		switch (error) {
		case 0:
			if (try_rgrp_fit(rgd, al))
				goto out;
1099
			if (rgd->rd_flags & GFS2_RDF_CHECK)
1100
				inode = try_rgrp_unlink(rgd, last_unlinked, ip->i_no_addr);
1101 1102
			if (!rg_locked)
				gfs2_glock_dq_uninit(&al->al_rgd_gh);
1103 1104
			if (inode)
				return inode;
1105
			/* fall through */
D
David Teigland 已提交
1106
		case GLR_TRYFAILED:
1107
			rgd = recent_rgrp_next(rgd);
D
David Teigland 已提交
1108 1109 1110
			break;

		default:
1111
			return ERR_PTR(error);
D
David Teigland 已提交
1112 1113 1114 1115 1116 1117 1118 1119
		}
	}

	/* Go through full list of rgrps */

	begin = rgd = forward_rgrp_get(sdp);

	for (;;) {
1120 1121 1122 1123 1124 1125 1126 1127 1128
		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
1129 1130 1131 1132
		switch (error) {
		case 0:
			if (try_rgrp_fit(rgd, al))
				goto out;
1133
			if (rgd->rd_flags & GFS2_RDF_CHECK)
1134
				inode = try_rgrp_unlink(rgd, last_unlinked, ip->i_no_addr);
1135 1136
			if (!rg_locked)
				gfs2_glock_dq_uninit(&al->al_rgd_gh);
1137 1138
			if (inode)
				return inode;
D
David Teigland 已提交
1139 1140 1141 1142 1143 1144 1145
			break;

		case GLR_TRYFAILED:
			skipped++;
			break;

		default:
1146
			return ERR_PTR(error);
D
David Teigland 已提交
1147 1148 1149 1150 1151 1152 1153
		}

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

		if (rgd == begin) {
1154
			if (++loops >= 3)
1155
				return ERR_PTR(-ENOSPC);
1156 1157
			if (!skipped)
				loops++;
D
David Teigland 已提交
1158
			flags = 0;
1159 1160
			if (loops == 2)
				gfs2_log_flush(sdp, NULL);
D
David Teigland 已提交
1161 1162 1163
		}
	}

1164
out:
D
David Teigland 已提交
1165
	if (begin) {
1166 1167 1168
		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
1169 1170 1171 1172 1173 1174
		rgd = gfs2_rgrpd_get_next(rgd);
		if (!rgd)
			rgd = gfs2_rgrpd_get_first(sdp);
		forward_rgrp_set(sdp, rgd);
	}

1175
	return NULL;
D
David Teigland 已提交
1176 1177 1178 1179 1180 1181 1182 1183 1184 1185 1186
}

/**
 * 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)
{
1187
	struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode);
1188
	struct gfs2_alloc *al = ip->i_alloc;
1189
	struct inode *inode;
1190
	int error = 0;
1191
	u64 last_unlinked = NO_BLOCK;
D
David Teigland 已提交
1192 1193 1194 1195

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

1196
try_again:
1197 1198 1199 1200 1201
	/* 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: */
1202
		error = gfs2_ri_update_special(ip);
1203

D
David Teigland 已提交
1204 1205 1206
	if (error)
		return error;

1207 1208
	inode = get_local_rgrp(ip, &last_unlinked);
	if (inode) {
1209 1210
		if (ip != GFS2_I(sdp->sd_rindex))
			gfs2_glock_dq_uninit(&al->al_ri_gh);
1211 1212 1213 1214 1215
		if (IS_ERR(inode))
			return PTR_ERR(inode);
		iput(inode);
		gfs2_log_flush(sdp, NULL);
		goto try_again;
D
David Teigland 已提交
1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 1230 1231 1232
	}

	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)
{
1233
	struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode);
1234
	struct gfs2_alloc *al = ip->i_alloc;
D
David Teigland 已提交
1235 1236 1237 1238 1239 1240 1241 1242

	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;
1243 1244
	if (al->al_rgd_gh.gh_gl)
		gfs2_glock_dq_uninit(&al->al_rgd_gh);
1245 1246
	if (ip != GFS2_I(sdp->sd_rindex))
		gfs2_glock_dq_uninit(&al->al_ri_gh);
D
David Teigland 已提交
1247 1248 1249 1250 1251 1252 1253 1254 1255 1256
}

/**
 * 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_*)
 */

1257
static unsigned char gfs2_get_block_type(struct gfs2_rgrpd *rgd, u64 block)
D
David Teigland 已提交
1258 1259
{
	struct gfs2_bitmap *bi = NULL;
1260
	u32 length, rgrp_block, buf_block;
D
David Teigland 已提交
1261 1262 1263
	unsigned int buf;
	unsigned char type;

1264 1265
	length = rgd->rd_length;
	rgrp_block = block - rgd->rd_data0;
D
David Teigland 已提交
1266 1267 1268 1269 1270 1271 1272 1273 1274 1275

	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;

1276
	type = gfs2_testbit(rgd, bi->bi_bh->b_data + bi->bi_offset,
D
David Teigland 已提交
1277 1278 1279 1280 1281 1282 1283 1284 1285 1286 1287 1288
			   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
1289
 * @n: The extent length
D
David Teigland 已提交
1290 1291 1292 1293 1294 1295 1296 1297 1298 1299 1300 1301 1302 1303
 *
 * 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
 */

1304
static u32 rgblk_search(struct gfs2_rgrpd *rgd, u32 goal,
1305 1306
			unsigned char old_state, unsigned char new_state,
			unsigned int *n)
D
David Teigland 已提交
1307 1308
{
	struct gfs2_bitmap *bi = NULL;
1309
	const u32 length = rgd->rd_length;
1310
	u32 blk = BFITNOENT;
D
David Teigland 已提交
1311
	unsigned int buf, x;
1312
	const unsigned int elen = *n;
1313
	const u8 *buffer = NULL;
D
David Teigland 已提交
1314

1315
	*n = 0;
D
David Teigland 已提交
1316 1317 1318
	/* Find bitmap block that contains bits for goal block */
	for (buf = 0; buf < length; buf++) {
		bi = rgd->rd_bits + buf;
1319 1320 1321 1322 1323
		/* Convert scope of "goal" from rgrp-wide to within found bit block */
		if (goal < (bi->bi_start + bi->bi_len) * GFS2_NBBY) {
			goal -= bi->bi_start * GFS2_NBBY;
			goto do_search;
		}
D
David Teigland 已提交
1324
	}
1325 1326
	buf = 0;
	goal = 0;
D
David Teigland 已提交
1327

1328
do_search:
D
David Teigland 已提交
1329 1330 1331 1332 1333 1334
	/* 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++) {
1335 1336 1337 1338 1339 1340
		bi = rgd->rd_bits + buf;

		if (test_bit(GBF_FULL, &bi->bi_flags) &&
		    (old_state == GFS2_BLKST_FREE))
			goto skip;

1341 1342
		/* The GFS2_BLKST_UNLINKED state doesn't apply to the clone
		   bitmaps, so we must search the originals for that. */
1343
		buffer = bi->bi_bh->b_data + bi->bi_offset;
1344
		if (old_state != GFS2_BLKST_UNLINKED && bi->bi_clone)
1345 1346 1347
			buffer = bi->bi_clone + bi->bi_offset;

		blk = gfs2_bitfit(buffer, bi->bi_len, goal, old_state);
D
David Teigland 已提交
1348 1349 1350
		if (blk != BFITNOENT)
			break;

1351 1352 1353
		if ((goal == 0) && (old_state == GFS2_BLKST_FREE))
			set_bit(GBF_FULL, &bi->bi_flags);

D
David Teigland 已提交
1354
		/* Try next bitmap block (wrap back to rgrp header if at end) */
1355 1356 1357
skip:
		buf++;
		buf %= length;
D
David Teigland 已提交
1358 1359 1360
		goal = 0;
	}

1361 1362 1363 1364 1365 1366 1367 1368 1369 1370 1371 1372 1373 1374 1375 1376 1377
	if (blk == BFITNOENT)
		return blk;
	*n = 1;
	if (old_state == new_state)
		goto out;

	gfs2_trans_add_bh(rgd->rd_gl, bi->bi_bh, 1);
	gfs2_setbit(rgd, bi->bi_bh->b_data, bi->bi_clone, bi->bi_offset,
		    bi->bi_len, blk, new_state);
	goal = blk;
	while (*n < elen) {
		goal++;
		if (goal >= (bi->bi_len * GFS2_NBBY))
			break;
		if (gfs2_testbit(rgd, buffer, bi->bi_len, goal) !=
		    GFS2_BLKST_FREE)
			break;
1378
		gfs2_setbit(rgd, bi->bi_bh->b_data, bi->bi_clone, bi->bi_offset,
1379 1380
			    bi->bi_len, goal, new_state);
		(*n)++;
1381
	}
1382 1383
out:
	return (bi->bi_start * GFS2_NBBY) + blk;
D
David Teigland 已提交
1384 1385 1386 1387 1388 1389 1390 1391 1392 1393 1394 1395
}

/**
 * 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)
 */

1396 1397
static struct gfs2_rgrpd *rgblk_free(struct gfs2_sbd *sdp, u64 bstart,
				     u32 blen, unsigned char new_state)
D
David Teigland 已提交
1398 1399 1400
{
	struct gfs2_rgrpd *rgd;
	struct gfs2_bitmap *bi = NULL;
1401
	u32 length, rgrp_blk, buf_blk;
D
David Teigland 已提交
1402 1403 1404 1405 1406
	unsigned int buf;

	rgd = gfs2_blk2rgrpd(sdp, bstart);
	if (!rgd) {
		if (gfs2_consist(sdp))
1407
			fs_err(sdp, "block = %llu\n", (unsigned long long)bstart);
D
David Teigland 已提交
1408 1409 1410
		return NULL;
	}

1411
	length = rgd->rd_length;
D
David Teigland 已提交
1412

1413
	rgrp_blk = bstart - rgd->rd_data0;
D
David Teigland 已提交
1414 1415 1416 1417 1418 1419 1420 1421 1422 1423 1424 1425 1426 1427 1428

	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,
1429
					       GFP_NOFS | __GFP_NOFAIL);
D
David Teigland 已提交
1430 1431 1432 1433
			memcpy(bi->bi_clone + bi->bi_offset,
			       bi->bi_bh->b_data + bi->bi_offset,
			       bi->bi_len);
		}
1434
		gfs2_trans_add_bh(rgd->rd_gl, bi->bi_bh, 1);
1435
		gfs2_setbit(rgd, bi->bi_bh->b_data, NULL, bi->bi_offset,
D
David Teigland 已提交
1436 1437 1438 1439 1440 1441 1442
			    bi->bi_len, buf_blk, new_state);
	}

	return rgd;
}

/**
1443 1444 1445 1446 1447 1448 1449 1450 1451 1452 1453 1454 1455 1456 1457 1458 1459
 * gfs2_rgrp_dump - print out an rgrp
 * @seq: The iterator
 * @gl: The glock in question
 *
 */

int gfs2_rgrp_dump(struct seq_file *seq, const struct gfs2_glock *gl)
{
	const struct gfs2_rgrpd *rgd = gl->gl_object;
	if (rgd == NULL)
		return 0;
	gfs2_print_dbg(seq, " R: n:%llu f:%02x b:%u/%u i:%u\n",
		       (unsigned long long)rgd->rd_addr, rgd->rd_flags,
		       rgd->rd_free, rgd->rd_free_clone, rgd->rd_dinodes);
	return 0;
}

1460 1461 1462 1463
static void gfs2_rgrp_error(struct gfs2_rgrpd *rgd)
{
	struct gfs2_sbd *sdp = rgd->rd_sbd;
	fs_warn(sdp, "rgrp %llu has an error, marking it readonly until umount\n",
S
Steven Whitehouse 已提交
1464
		(unsigned long long)rgd->rd_addr);
1465 1466 1467 1468 1469
	fs_warn(sdp, "umount on all nodes and run fsck.gfs2 to fix the error\n");
	gfs2_rgrp_dump(NULL, rgd->rd_gl);
	rgd->rd_flags |= GFS2_RDF_ERROR;
}

1470 1471
/**
 * gfs2_alloc_block - Allocate one or more blocks
1472
 * @ip: the inode to allocate the block for
1473 1474
 * @bn: Used to return the starting block number
 * @n: requested number of blocks/extent length (value/result)
D
David Teigland 已提交
1475
 *
1476
 * Returns: 0 or error
D
David Teigland 已提交
1477 1478
 */

1479
int gfs2_alloc_block(struct gfs2_inode *ip, u64 *bn, unsigned int *n)
D
David Teigland 已提交
1480
{
1481
	struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode);
1482
	struct buffer_head *dibh;
1483
	struct gfs2_alloc *al = ip->i_alloc;
D
David Teigland 已提交
1484
	struct gfs2_rgrpd *rgd = al->al_rgd;
1485 1486
	u32 goal, blk;
	u64 block;
1487
	int error;
D
David Teigland 已提交
1488

1489 1490
	if (rgrp_contains_block(rgd, ip->i_goal))
		goal = ip->i_goal - rgd->rd_data0;
D
David Teigland 已提交
1491
	else
1492
		goal = rgd->rd_last_alloc;
D
David Teigland 已提交
1493

1494
	blk = rgblk_search(rgd, goal, GFS2_BLKST_FREE, GFS2_BLKST_USED, n);
1495 1496 1497 1498

	/* Since all blocks are reserved in advance, this shouldn't happen */
	if (blk == BFITNOENT)
		goto rgrp_error;
D
David Teigland 已提交
1499

1500
	rgd->rd_last_alloc = blk;
1501
	block = rgd->rd_data0 + blk;
1502
	ip->i_goal = block;
1503 1504 1505 1506 1507 1508 1509
	error = gfs2_meta_inode_buffer(ip, &dibh);
	if (error == 0) {
		struct gfs2_dinode *di = (struct gfs2_dinode *)dibh->b_data;
		gfs2_trans_add_bh(ip->i_gl, dibh, 1);
		di->di_goal_meta = di->di_goal_data = cpu_to_be64(ip->i_goal);
		brelse(dibh);
	}
1510 1511 1512
	if (rgd->rd_free < *n)
		goto rgrp_error;

1513
	rgd->rd_free -= *n;
D
David Teigland 已提交
1514

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

1518
	al->al_alloced += *n;
D
David Teigland 已提交
1519

1520
	gfs2_statfs_change(sdp, 0, -(s64)*n, 0);
1521
	gfs2_quota_change(ip, *n, ip->i_inode.i_uid, ip->i_inode.i_gid);
D
David Teigland 已提交
1522 1523

	spin_lock(&sdp->sd_rindex_spin);
1524
	rgd->rd_free_clone -= *n;
D
David Teigland 已提交
1525
	spin_unlock(&sdp->sd_rindex_spin);
S
Steven Whitehouse 已提交
1526
	trace_gfs2_block_alloc(ip, block, *n, GFS2_BLKST_USED);
1527 1528 1529 1530
	*bn = block;
	return 0;

rgrp_error:
1531
	gfs2_rgrp_error(rgd);
1532
	return -EIO;
D
David Teigland 已提交
1533 1534 1535 1536 1537
}

/**
 * gfs2_alloc_di - Allocate a dinode
 * @dip: the directory that the inode is going in
1538 1539
 * @bn: the block number which is allocated
 * @generation: the generation number of the inode
D
David Teigland 已提交
1540
 *
1541
 * Returns: 0 on success or error
D
David Teigland 已提交
1542 1543
 */

1544
int gfs2_alloc_di(struct gfs2_inode *dip, u64 *bn, u64 *generation)
D
David Teigland 已提交
1545
{
1546
	struct gfs2_sbd *sdp = GFS2_SB(&dip->i_inode);
1547
	struct gfs2_alloc *al = dip->i_alloc;
D
David Teigland 已提交
1548
	struct gfs2_rgrpd *rgd = al->al_rgd;
S
Steven Whitehouse 已提交
1549 1550
	u32 blk;
	u64 block;
1551
	unsigned int n = 1;
D
David Teigland 已提交
1552

1553
	blk = rgblk_search(rgd, rgd->rd_last_alloc,
1554
			   GFS2_BLKST_FREE, GFS2_BLKST_DINODE, &n);
D
David Teigland 已提交
1555

1556 1557 1558
	/* Since all blocks are reserved in advance, this shouldn't happen */
	if (blk == BFITNOENT)
		goto rgrp_error;
D
David Teigland 已提交
1559

1560
	rgd->rd_last_alloc = blk;
1561
	block = rgd->rd_data0 + blk;
1562 1563 1564
	if (rgd->rd_free == 0)
		goto rgrp_error;

1565
	rgd->rd_free--;
1566
	rgd->rd_dinodes++;
1567
	*generation = rgd->rd_igeneration++;
1568 1569
	if (*generation == 0)
		*generation = rgd->rd_igeneration++;
1570
	gfs2_trans_add_bh(rgd->rd_gl, rgd->rd_bits[0].bi_bh, 1);
1571
	gfs2_rgrp_out(rgd, rgd->rd_bits[0].bi_bh->b_data);
D
David Teigland 已提交
1572 1573 1574 1575

	al->al_alloced++;

	gfs2_statfs_change(sdp, 0, -1, +1);
1576
	gfs2_trans_add_unrevoke(sdp, block, 1);
D
David Teigland 已提交
1577 1578 1579 1580

	spin_lock(&sdp->sd_rindex_spin);
	rgd->rd_free_clone--;
	spin_unlock(&sdp->sd_rindex_spin);
S
Steven Whitehouse 已提交
1581
	trace_gfs2_block_alloc(dip, block, 1, GFS2_BLKST_DINODE);
1582 1583 1584 1585 1586 1587
	*bn = block;
	return 0;

rgrp_error:
	gfs2_rgrp_error(rgd);
	return -EIO;
D
David Teigland 已提交
1588 1589 1590 1591 1592 1593 1594 1595 1596 1597
}

/**
 * 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
 *
 */

1598
void gfs2_free_data(struct gfs2_inode *ip, u64 bstart, u32 blen)
D
David Teigland 已提交
1599
{
1600
	struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode);
D
David Teigland 已提交
1601 1602 1603 1604 1605
	struct gfs2_rgrpd *rgd;

	rgd = rgblk_free(sdp, bstart, blen, GFS2_BLKST_FREE);
	if (!rgd)
		return;
S
Steven Whitehouse 已提交
1606
	trace_gfs2_block_alloc(ip, bstart, blen, GFS2_BLKST_FREE);
1607
	rgd->rd_free += blen;
D
David Teigland 已提交
1608

1609
	gfs2_trans_add_bh(rgd->rd_gl, rgd->rd_bits[0].bi_bh, 1);
1610
	gfs2_rgrp_out(rgd, rgd->rd_bits[0].bi_bh->b_data);
D
David Teigland 已提交
1611 1612 1613 1614

	gfs2_trans_add_rg(rgd);

	gfs2_statfs_change(sdp, 0, +blen, 0);
1615
	gfs2_quota_change(ip, -(s64)blen, ip->i_inode.i_uid, ip->i_inode.i_gid);
D
David Teigland 已提交
1616 1617 1618 1619 1620 1621 1622 1623 1624 1625
}

/**
 * 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
 *
 */

1626
void gfs2_free_meta(struct gfs2_inode *ip, u64 bstart, u32 blen)
D
David Teigland 已提交
1627
{
1628
	struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode);
D
David Teigland 已提交
1629 1630 1631 1632 1633
	struct gfs2_rgrpd *rgd;

	rgd = rgblk_free(sdp, bstart, blen, GFS2_BLKST_FREE);
	if (!rgd)
		return;
S
Steven Whitehouse 已提交
1634
	trace_gfs2_block_alloc(ip, bstart, blen, GFS2_BLKST_FREE);
1635
	rgd->rd_free += blen;
D
David Teigland 已提交
1636

1637
	gfs2_trans_add_bh(rgd->rd_gl, rgd->rd_bits[0].bi_bh, 1);
1638
	gfs2_rgrp_out(rgd, rgd->rd_bits[0].bi_bh->b_data);
D
David Teigland 已提交
1639 1640 1641 1642

	gfs2_trans_add_rg(rgd);

	gfs2_statfs_change(sdp, 0, +blen, 0);
1643
	gfs2_quota_change(ip, -(s64)blen, ip->i_inode.i_uid, ip->i_inode.i_gid);
D
David Teigland 已提交
1644 1645 1646
	gfs2_meta_wipe(ip, bstart, blen);
}

1647 1648 1649 1650 1651
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;
1652
	u64 blkno = ip->i_no_addr;
1653 1654 1655 1656

	rgd = rgblk_free(sdp, blkno, 1, GFS2_BLKST_UNLINKED);
	if (!rgd)
		return;
S
Steven Whitehouse 已提交
1657
	trace_gfs2_block_alloc(ip, blkno, 1, GFS2_BLKST_UNLINKED);
1658
	gfs2_trans_add_bh(rgd->rd_gl, rgd->rd_bits[0].bi_bh, 1);
1659
	gfs2_rgrp_out(rgd, rgd->rd_bits[0].bi_bh->b_data);
1660 1661 1662
	gfs2_trans_add_rg(rgd);
}

1663
static void gfs2_free_uninit_di(struct gfs2_rgrpd *rgd, u64 blkno)
D
David Teigland 已提交
1664 1665 1666 1667 1668 1669 1670 1671 1672
{
	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);

1673
	if (!rgd->rd_dinodes)
D
David Teigland 已提交
1674
		gfs2_consist_rgrpd(rgd);
1675
	rgd->rd_dinodes--;
1676
	rgd->rd_free++;
D
David Teigland 已提交
1677

1678
	gfs2_trans_add_bh(rgd->rd_gl, rgd->rd_bits[0].bi_bh, 1);
1679
	gfs2_rgrp_out(rgd, rgd->rd_bits[0].bi_bh->b_data);
D
David Teigland 已提交
1680 1681 1682 1683 1684 1685 1686 1687

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


void gfs2_free_di(struct gfs2_rgrpd *rgd, struct gfs2_inode *ip)
{
1688
	gfs2_free_uninit_di(rgd, ip->i_no_addr);
S
Steven Whitehouse 已提交
1689
	trace_gfs2_block_alloc(ip, ip->i_no_addr, 1, GFS2_BLKST_FREE);
1690
	gfs2_quota_change(ip, -1, ip->i_inode.i_uid, ip->i_inode.i_gid);
1691
	gfs2_meta_wipe(ip, ip->i_no_addr, 1);
D
David Teigland 已提交
1692 1693
}

1694 1695 1696 1697 1698 1699 1700 1701 1702 1703 1704 1705 1706 1707 1708
/**
 * gfs2_check_blk_type - Check the type of a block
 * @sdp: The superblock
 * @no_addr: The block number to check
 * @type: The block type we are looking for
 *
 * Returns: 0 if the block type matches the expected type
 *          -ESTALE if it doesn't match
 *          or -ve errno if something went wrong while checking
 */

int gfs2_check_blk_type(struct gfs2_sbd *sdp, u64 no_addr, unsigned int type)
{
	struct gfs2_rgrpd *rgd;
	struct gfs2_holder ri_gh, rgd_gh;
1709 1710
	struct gfs2_inode *ip = GFS2_I(sdp->sd_rindex);
	int ri_locked = 0;
1711 1712
	int error;

1713 1714 1715 1716 1717 1718
	if (!gfs2_glock_is_locked_by_me(ip->i_gl)) {
		error = gfs2_rindex_hold(sdp, &ri_gh);
		if (error)
			goto fail;
		ri_locked = 1;
	}
1719 1720 1721 1722 1723 1724 1725 1726 1727 1728 1729 1730 1731 1732 1733

	error = -EINVAL;
	rgd = gfs2_blk2rgrpd(sdp, no_addr);
	if (!rgd)
		goto fail_rindex;

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

	if (gfs2_get_block_type(rgd, no_addr) != type)
		error = -ESTALE;

	gfs2_glock_dq_uninit(&rgd_gh);
fail_rindex:
1734 1735
	if (ri_locked)
		gfs2_glock_dq_uninit(&ri_gh);
1736 1737 1738 1739
fail:
	return error;
}

D
David Teigland 已提交
1740 1741 1742 1743 1744 1745 1746 1747 1748 1749 1750 1751 1752
/**
 * 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,
1753
		    u64 block)
D
David Teigland 已提交
1754 1755 1756 1757 1758 1759 1760 1761 1762 1763 1764 1765
{
	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))
1766
			fs_err(sdp, "block = %llu\n", (unsigned long long)block);
D
David Teigland 已提交
1767 1768 1769 1770 1771 1772 1773 1774 1775 1776 1777
		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 *),
1778
			      GFP_NOFS | __GFP_NOFAIL);
D
David Teigland 已提交
1779 1780 1781 1782 1783 1784 1785 1786 1787 1788 1789 1790 1791 1792 1793 1794 1795 1796 1797 1798 1799 1800 1801 1802 1803

		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
 *
 */

1804
void gfs2_rlist_alloc(struct gfs2_rgrp_list *rlist, unsigned int state)
D
David Teigland 已提交
1805 1806 1807 1808
{
	unsigned int x;

	rlist->rl_ghs = kcalloc(rlist->rl_rgrps, sizeof(struct gfs2_holder),
1809
				GFP_NOFS | __GFP_NOFAIL);
D
David Teigland 已提交
1810 1811
	for (x = 0; x < rlist->rl_rgrps; x++)
		gfs2_holder_init(rlist->rl_rgd[x]->rd_gl,
1812
				state, 0,
D
David Teigland 已提交
1813 1814 1815 1816 1817 1818 1819 1820 1821 1822 1823 1824 1825 1826 1827 1828 1829 1830 1831 1832 1833 1834
				&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);
	}
}