xref: /openbmc/linux/fs/gfs2/rgrp.c (revision 4bce6fce)
1 /*
2  * Copyright (C) Sistina Software, Inc.  1997-2003 All rights reserved.
3  * Copyright (C) 2004-2008 Red Hat, Inc.  All rights reserved.
4  *
5  * This copyrighted material is made available to anyone wishing to use,
6  * modify, copy, or redistribute it subject to the terms and conditions
7  * of the GNU General Public License version 2.
8  */
9 
10 #define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
11 
12 #include <linux/slab.h>
13 #include <linux/spinlock.h>
14 #include <linux/completion.h>
15 #include <linux/buffer_head.h>
16 #include <linux/fs.h>
17 #include <linux/gfs2_ondisk.h>
18 #include <linux/prefetch.h>
19 #include <linux/blkdev.h>
20 #include <linux/rbtree.h>
21 #include <linux/random.h>
22 
23 #include "gfs2.h"
24 #include "incore.h"
25 #include "glock.h"
26 #include "glops.h"
27 #include "lops.h"
28 #include "meta_io.h"
29 #include "quota.h"
30 #include "rgrp.h"
31 #include "super.h"
32 #include "trans.h"
33 #include "util.h"
34 #include "log.h"
35 #include "inode.h"
36 #include "trace_gfs2.h"
37 
38 #define BFITNOENT ((u32)~0)
39 #define NO_BLOCK ((u64)~0)
40 
41 #if BITS_PER_LONG == 32
42 #define LBITMASK   (0x55555555UL)
43 #define LBITSKIP55 (0x55555555UL)
44 #define LBITSKIP00 (0x00000000UL)
45 #else
46 #define LBITMASK   (0x5555555555555555UL)
47 #define LBITSKIP55 (0x5555555555555555UL)
48 #define LBITSKIP00 (0x0000000000000000UL)
49 #endif
50 
51 /*
52  * These routines are used by the resource group routines (rgrp.c)
53  * to keep track of block allocation.  Each block is represented by two
54  * bits.  So, each byte represents GFS2_NBBY (i.e. 4) blocks.
55  *
56  * 0 = Free
57  * 1 = Used (not metadata)
58  * 2 = Unlinked (still in use) inode
59  * 3 = Used (metadata)
60  */
61 
62 struct gfs2_extent {
63 	struct gfs2_rbm rbm;
64 	u32 len;
65 };
66 
67 static const char valid_change[16] = {
68 	        /* current */
69 	/* n */ 0, 1, 1, 1,
70 	/* e */ 1, 0, 0, 0,
71 	/* w */ 0, 0, 0, 1,
72 	        1, 0, 0, 0
73 };
74 
75 static int gfs2_rbm_find(struct gfs2_rbm *rbm, u8 state, u32 *minext,
76 			 const struct gfs2_inode *ip, bool nowrap,
77 			 const struct gfs2_alloc_parms *ap);
78 
79 
80 /**
81  * gfs2_setbit - Set a bit in the bitmaps
82  * @rbm: The position of the bit to set
83  * @do_clone: Also set the clone bitmap, if it exists
84  * @new_state: the new state of the block
85  *
86  */
87 
88 static inline void gfs2_setbit(const struct gfs2_rbm *rbm, bool do_clone,
89 			       unsigned char new_state)
90 {
91 	unsigned char *byte1, *byte2, *end, cur_state;
92 	struct gfs2_bitmap *bi = rbm_bi(rbm);
93 	unsigned int buflen = bi->bi_len;
94 	const unsigned int bit = (rbm->offset % GFS2_NBBY) * GFS2_BIT_SIZE;
95 
96 	byte1 = bi->bi_bh->b_data + bi->bi_offset + (rbm->offset / GFS2_NBBY);
97 	end = bi->bi_bh->b_data + bi->bi_offset + buflen;
98 
99 	BUG_ON(byte1 >= end);
100 
101 	cur_state = (*byte1 >> bit) & GFS2_BIT_MASK;
102 
103 	if (unlikely(!valid_change[new_state * 4 + cur_state])) {
104 		pr_warn("buf_blk = 0x%x old_state=%d, new_state=%d\n",
105 			rbm->offset, cur_state, new_state);
106 		pr_warn("rgrp=0x%llx bi_start=0x%x\n",
107 			(unsigned long long)rbm->rgd->rd_addr, bi->bi_start);
108 		pr_warn("bi_offset=0x%x bi_len=0x%x\n",
109 			bi->bi_offset, bi->bi_len);
110 		dump_stack();
111 		gfs2_consist_rgrpd(rbm->rgd);
112 		return;
113 	}
114 	*byte1 ^= (cur_state ^ new_state) << bit;
115 
116 	if (do_clone && bi->bi_clone) {
117 		byte2 = bi->bi_clone + bi->bi_offset + (rbm->offset / GFS2_NBBY);
118 		cur_state = (*byte2 >> bit) & GFS2_BIT_MASK;
119 		*byte2 ^= (cur_state ^ new_state) << bit;
120 	}
121 }
122 
123 /**
124  * gfs2_testbit - test a bit in the bitmaps
125  * @rbm: The bit to test
126  *
127  * Returns: The two bit block state of the requested bit
128  */
129 
130 static inline u8 gfs2_testbit(const struct gfs2_rbm *rbm)
131 {
132 	struct gfs2_bitmap *bi = rbm_bi(rbm);
133 	const u8 *buffer = bi->bi_bh->b_data + bi->bi_offset;
134 	const u8 *byte;
135 	unsigned int bit;
136 
137 	byte = buffer + (rbm->offset / GFS2_NBBY);
138 	bit = (rbm->offset % GFS2_NBBY) * GFS2_BIT_SIZE;
139 
140 	return (*byte >> bit) & GFS2_BIT_MASK;
141 }
142 
143 /**
144  * gfs2_bit_search
145  * @ptr: Pointer to bitmap data
146  * @mask: Mask to use (normally 0x55555.... but adjusted for search start)
147  * @state: The state we are searching for
148  *
149  * We xor the bitmap data with a patter which is the bitwise opposite
150  * of what we are looking for, this gives rise to a pattern of ones
151  * wherever there is a match. Since we have two bits per entry, we
152  * take this pattern, shift it down by one place and then and it with
153  * the original. All the even bit positions (0,2,4, etc) then represent
154  * successful matches, so we mask with 0x55555..... to remove the unwanted
155  * odd bit positions.
156  *
157  * This allows searching of a whole u64 at once (32 blocks) with a
158  * single test (on 64 bit arches).
159  */
160 
161 static inline u64 gfs2_bit_search(const __le64 *ptr, u64 mask, u8 state)
162 {
163 	u64 tmp;
164 	static const u64 search[] = {
165 		[0] = 0xffffffffffffffffULL,
166 		[1] = 0xaaaaaaaaaaaaaaaaULL,
167 		[2] = 0x5555555555555555ULL,
168 		[3] = 0x0000000000000000ULL,
169 	};
170 	tmp = le64_to_cpu(*ptr) ^ search[state];
171 	tmp &= (tmp >> 1);
172 	tmp &= mask;
173 	return tmp;
174 }
175 
176 /**
177  * rs_cmp - multi-block reservation range compare
178  * @blk: absolute file system block number of the new reservation
179  * @len: number of blocks in the new reservation
180  * @rs: existing reservation to compare against
181  *
182  * returns: 1 if the block range is beyond the reach of the reservation
183  *         -1 if the block range is before the start of the reservation
184  *          0 if the block range overlaps with the reservation
185  */
186 static inline int rs_cmp(u64 blk, u32 len, struct gfs2_blkreserv *rs)
187 {
188 	u64 startblk = gfs2_rbm_to_block(&rs->rs_rbm);
189 
190 	if (blk >= startblk + rs->rs_free)
191 		return 1;
192 	if (blk + len - 1 < startblk)
193 		return -1;
194 	return 0;
195 }
196 
197 /**
198  * gfs2_bitfit - Search an rgrp's bitmap buffer to find a bit-pair representing
199  *       a block in a given allocation state.
200  * @buf: the buffer that holds the bitmaps
201  * @len: the length (in bytes) of the buffer
202  * @goal: start search at this block's bit-pair (within @buffer)
203  * @state: GFS2_BLKST_XXX the state of the block we're looking for.
204  *
205  * Scope of @goal and returned block number is only within this bitmap buffer,
206  * not entire rgrp or filesystem.  @buffer will be offset from the actual
207  * beginning of a bitmap block buffer, skipping any header structures, but
208  * headers are always a multiple of 64 bits long so that the buffer is
209  * always aligned to a 64 bit boundary.
210  *
211  * The size of the buffer is in bytes, but is it assumed that it is
212  * always ok to read a complete multiple of 64 bits at the end
213  * of the block in case the end is no aligned to a natural boundary.
214  *
215  * Return: the block number (bitmap buffer scope) that was found
216  */
217 
218 static u32 gfs2_bitfit(const u8 *buf, const unsigned int len,
219 		       u32 goal, u8 state)
220 {
221 	u32 spoint = (goal << 1) & ((8*sizeof(u64)) - 1);
222 	const __le64 *ptr = ((__le64 *)buf) + (goal >> 5);
223 	const __le64 *end = (__le64 *)(buf + ALIGN(len, sizeof(u64)));
224 	u64 tmp;
225 	u64 mask = 0x5555555555555555ULL;
226 	u32 bit;
227 
228 	/* Mask off bits we don't care about at the start of the search */
229 	mask <<= spoint;
230 	tmp = gfs2_bit_search(ptr, mask, state);
231 	ptr++;
232 	while(tmp == 0 && ptr < end) {
233 		tmp = gfs2_bit_search(ptr, 0x5555555555555555ULL, state);
234 		ptr++;
235 	}
236 	/* Mask off any bits which are more than len bytes from the start */
237 	if (ptr == end && (len & (sizeof(u64) - 1)))
238 		tmp &= (((u64)~0) >> (64 - 8*(len & (sizeof(u64) - 1))));
239 	/* Didn't find anything, so return */
240 	if (tmp == 0)
241 		return BFITNOENT;
242 	ptr--;
243 	bit = __ffs64(tmp);
244 	bit /= 2;	/* two bits per entry in the bitmap */
245 	return (((const unsigned char *)ptr - buf) * GFS2_NBBY) + bit;
246 }
247 
248 /**
249  * gfs2_rbm_from_block - Set the rbm based upon rgd and block number
250  * @rbm: The rbm with rgd already set correctly
251  * @block: The block number (filesystem relative)
252  *
253  * This sets the bi and offset members of an rbm based on a
254  * resource group and a filesystem relative block number. The
255  * resource group must be set in the rbm on entry, the bi and
256  * offset members will be set by this function.
257  *
258  * Returns: 0 on success, or an error code
259  */
260 
261 static int gfs2_rbm_from_block(struct gfs2_rbm *rbm, u64 block)
262 {
263 	u64 rblock = block - rbm->rgd->rd_data0;
264 
265 	if (WARN_ON_ONCE(rblock > UINT_MAX))
266 		return -EINVAL;
267 	if (block >= rbm->rgd->rd_data0 + rbm->rgd->rd_data)
268 		return -E2BIG;
269 
270 	rbm->bii = 0;
271 	rbm->offset = (u32)(rblock);
272 	/* Check if the block is within the first block */
273 	if (rbm->offset < rbm_bi(rbm)->bi_blocks)
274 		return 0;
275 
276 	/* Adjust for the size diff between gfs2_meta_header and gfs2_rgrp */
277 	rbm->offset += (sizeof(struct gfs2_rgrp) -
278 			sizeof(struct gfs2_meta_header)) * GFS2_NBBY;
279 	rbm->bii = rbm->offset / rbm->rgd->rd_sbd->sd_blocks_per_bitmap;
280 	rbm->offset -= rbm->bii * rbm->rgd->rd_sbd->sd_blocks_per_bitmap;
281 	return 0;
282 }
283 
284 /**
285  * gfs2_rbm_incr - increment an rbm structure
286  * @rbm: The rbm with rgd already set correctly
287  *
288  * This function takes an existing rbm structure and increments it to the next
289  * viable block offset.
290  *
291  * Returns: If incrementing the offset would cause the rbm to go past the
292  *          end of the rgrp, true is returned, otherwise false.
293  *
294  */
295 
296 static bool gfs2_rbm_incr(struct gfs2_rbm *rbm)
297 {
298 	if (rbm->offset + 1 < rbm_bi(rbm)->bi_blocks) { /* in the same bitmap */
299 		rbm->offset++;
300 		return false;
301 	}
302 	if (rbm->bii == rbm->rgd->rd_length - 1) /* at the last bitmap */
303 		return true;
304 
305 	rbm->offset = 0;
306 	rbm->bii++;
307 	return false;
308 }
309 
310 /**
311  * gfs2_unaligned_extlen - Look for free blocks which are not byte aligned
312  * @rbm: Position to search (value/result)
313  * @n_unaligned: Number of unaligned blocks to check
314  * @len: Decremented for each block found (terminate on zero)
315  *
316  * Returns: true if a non-free block is encountered
317  */
318 
319 static bool gfs2_unaligned_extlen(struct gfs2_rbm *rbm, u32 n_unaligned, u32 *len)
320 {
321 	u32 n;
322 	u8 res;
323 
324 	for (n = 0; n < n_unaligned; n++) {
325 		res = gfs2_testbit(rbm);
326 		if (res != GFS2_BLKST_FREE)
327 			return true;
328 		(*len)--;
329 		if (*len == 0)
330 			return true;
331 		if (gfs2_rbm_incr(rbm))
332 			return true;
333 	}
334 
335 	return false;
336 }
337 
338 /**
339  * gfs2_free_extlen - Return extent length of free blocks
340  * @rrbm: Starting position
341  * @len: Max length to check
342  *
343  * Starting at the block specified by the rbm, see how many free blocks
344  * there are, not reading more than len blocks ahead. This can be done
345  * using memchr_inv when the blocks are byte aligned, but has to be done
346  * on a block by block basis in case of unaligned blocks. Also this
347  * function can cope with bitmap boundaries (although it must stop on
348  * a resource group boundary)
349  *
350  * Returns: Number of free blocks in the extent
351  */
352 
353 static u32 gfs2_free_extlen(const struct gfs2_rbm *rrbm, u32 len)
354 {
355 	struct gfs2_rbm rbm = *rrbm;
356 	u32 n_unaligned = rbm.offset & 3;
357 	u32 size = len;
358 	u32 bytes;
359 	u32 chunk_size;
360 	u8 *ptr, *start, *end;
361 	u64 block;
362 	struct gfs2_bitmap *bi;
363 
364 	if (n_unaligned &&
365 	    gfs2_unaligned_extlen(&rbm, 4 - n_unaligned, &len))
366 		goto out;
367 
368 	n_unaligned = len & 3;
369 	/* Start is now byte aligned */
370 	while (len > 3) {
371 		bi = rbm_bi(&rbm);
372 		start = bi->bi_bh->b_data;
373 		if (bi->bi_clone)
374 			start = bi->bi_clone;
375 		end = start + bi->bi_bh->b_size;
376 		start += bi->bi_offset;
377 		BUG_ON(rbm.offset & 3);
378 		start += (rbm.offset / GFS2_NBBY);
379 		bytes = min_t(u32, len / GFS2_NBBY, (end - start));
380 		ptr = memchr_inv(start, 0, bytes);
381 		chunk_size = ((ptr == NULL) ? bytes : (ptr - start));
382 		chunk_size *= GFS2_NBBY;
383 		BUG_ON(len < chunk_size);
384 		len -= chunk_size;
385 		block = gfs2_rbm_to_block(&rbm);
386 		if (gfs2_rbm_from_block(&rbm, block + chunk_size)) {
387 			n_unaligned = 0;
388 			break;
389 		}
390 		if (ptr) {
391 			n_unaligned = 3;
392 			break;
393 		}
394 		n_unaligned = len & 3;
395 	}
396 
397 	/* Deal with any bits left over at the end */
398 	if (n_unaligned)
399 		gfs2_unaligned_extlen(&rbm, n_unaligned, &len);
400 out:
401 	return size - len;
402 }
403 
404 /**
405  * gfs2_bitcount - count the number of bits in a certain state
406  * @rgd: the resource group descriptor
407  * @buffer: the buffer that holds the bitmaps
408  * @buflen: the length (in bytes) of the buffer
409  * @state: the state of the block we're looking for
410  *
411  * Returns: The number of bits
412  */
413 
414 static u32 gfs2_bitcount(struct gfs2_rgrpd *rgd, const u8 *buffer,
415 			 unsigned int buflen, u8 state)
416 {
417 	const u8 *byte = buffer;
418 	const u8 *end = buffer + buflen;
419 	const u8 state1 = state << 2;
420 	const u8 state2 = state << 4;
421 	const u8 state3 = state << 6;
422 	u32 count = 0;
423 
424 	for (; byte < end; byte++) {
425 		if (((*byte) & 0x03) == state)
426 			count++;
427 		if (((*byte) & 0x0C) == state1)
428 			count++;
429 		if (((*byte) & 0x30) == state2)
430 			count++;
431 		if (((*byte) & 0xC0) == state3)
432 			count++;
433 	}
434 
435 	return count;
436 }
437 
438 /**
439  * gfs2_rgrp_verify - Verify that a resource group is consistent
440  * @rgd: the rgrp
441  *
442  */
443 
444 void gfs2_rgrp_verify(struct gfs2_rgrpd *rgd)
445 {
446 	struct gfs2_sbd *sdp = rgd->rd_sbd;
447 	struct gfs2_bitmap *bi = NULL;
448 	u32 length = rgd->rd_length;
449 	u32 count[4], tmp;
450 	int buf, x;
451 
452 	memset(count, 0, 4 * sizeof(u32));
453 
454 	/* Count # blocks in each of 4 possible allocation states */
455 	for (buf = 0; buf < length; buf++) {
456 		bi = rgd->rd_bits + buf;
457 		for (x = 0; x < 4; x++)
458 			count[x] += gfs2_bitcount(rgd,
459 						  bi->bi_bh->b_data +
460 						  bi->bi_offset,
461 						  bi->bi_len, x);
462 	}
463 
464 	if (count[0] != rgd->rd_free) {
465 		if (gfs2_consist_rgrpd(rgd))
466 			fs_err(sdp, "free data mismatch:  %u != %u\n",
467 			       count[0], rgd->rd_free);
468 		return;
469 	}
470 
471 	tmp = rgd->rd_data - rgd->rd_free - rgd->rd_dinodes;
472 	if (count[1] != tmp) {
473 		if (gfs2_consist_rgrpd(rgd))
474 			fs_err(sdp, "used data mismatch:  %u != %u\n",
475 			       count[1], tmp);
476 		return;
477 	}
478 
479 	if (count[2] + count[3] != rgd->rd_dinodes) {
480 		if (gfs2_consist_rgrpd(rgd))
481 			fs_err(sdp, "used metadata mismatch:  %u != %u\n",
482 			       count[2] + count[3], rgd->rd_dinodes);
483 		return;
484 	}
485 }
486 
487 static inline int rgrp_contains_block(struct gfs2_rgrpd *rgd, u64 block)
488 {
489 	u64 first = rgd->rd_data0;
490 	u64 last = first + rgd->rd_data;
491 	return first <= block && block < last;
492 }
493 
494 /**
495  * gfs2_blk2rgrpd - Find resource group for a given data/meta block number
496  * @sdp: The GFS2 superblock
497  * @blk: The data block number
498  * @exact: True if this needs to be an exact match
499  *
500  * Returns: The resource group, or NULL if not found
501  */
502 
503 struct gfs2_rgrpd *gfs2_blk2rgrpd(struct gfs2_sbd *sdp, u64 blk, bool exact)
504 {
505 	struct rb_node *n, *next;
506 	struct gfs2_rgrpd *cur;
507 
508 	spin_lock(&sdp->sd_rindex_spin);
509 	n = sdp->sd_rindex_tree.rb_node;
510 	while (n) {
511 		cur = rb_entry(n, struct gfs2_rgrpd, rd_node);
512 		next = NULL;
513 		if (blk < cur->rd_addr)
514 			next = n->rb_left;
515 		else if (blk >= cur->rd_data0 + cur->rd_data)
516 			next = n->rb_right;
517 		if (next == NULL) {
518 			spin_unlock(&sdp->sd_rindex_spin);
519 			if (exact) {
520 				if (blk < cur->rd_addr)
521 					return NULL;
522 				if (blk >= cur->rd_data0 + cur->rd_data)
523 					return NULL;
524 			}
525 			return cur;
526 		}
527 		n = next;
528 	}
529 	spin_unlock(&sdp->sd_rindex_spin);
530 
531 	return NULL;
532 }
533 
534 /**
535  * gfs2_rgrpd_get_first - get the first Resource Group in the filesystem
536  * @sdp: The GFS2 superblock
537  *
538  * Returns: The first rgrp in the filesystem
539  */
540 
541 struct gfs2_rgrpd *gfs2_rgrpd_get_first(struct gfs2_sbd *sdp)
542 {
543 	const struct rb_node *n;
544 	struct gfs2_rgrpd *rgd;
545 
546 	spin_lock(&sdp->sd_rindex_spin);
547 	n = rb_first(&sdp->sd_rindex_tree);
548 	rgd = rb_entry(n, struct gfs2_rgrpd, rd_node);
549 	spin_unlock(&sdp->sd_rindex_spin);
550 
551 	return rgd;
552 }
553 
554 /**
555  * gfs2_rgrpd_get_next - get the next RG
556  * @rgd: the resource group descriptor
557  *
558  * Returns: The next rgrp
559  */
560 
561 struct gfs2_rgrpd *gfs2_rgrpd_get_next(struct gfs2_rgrpd *rgd)
562 {
563 	struct gfs2_sbd *sdp = rgd->rd_sbd;
564 	const struct rb_node *n;
565 
566 	spin_lock(&sdp->sd_rindex_spin);
567 	n = rb_next(&rgd->rd_node);
568 	if (n == NULL)
569 		n = rb_first(&sdp->sd_rindex_tree);
570 
571 	if (unlikely(&rgd->rd_node == n)) {
572 		spin_unlock(&sdp->sd_rindex_spin);
573 		return NULL;
574 	}
575 	rgd = rb_entry(n, struct gfs2_rgrpd, rd_node);
576 	spin_unlock(&sdp->sd_rindex_spin);
577 	return rgd;
578 }
579 
580 void check_and_update_goal(struct gfs2_inode *ip)
581 {
582 	struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode);
583 	if (!ip->i_goal || gfs2_blk2rgrpd(sdp, ip->i_goal, 1) == NULL)
584 		ip->i_goal = ip->i_no_addr;
585 }
586 
587 void gfs2_free_clones(struct gfs2_rgrpd *rgd)
588 {
589 	int x;
590 
591 	for (x = 0; x < rgd->rd_length; x++) {
592 		struct gfs2_bitmap *bi = rgd->rd_bits + x;
593 		kfree(bi->bi_clone);
594 		bi->bi_clone = NULL;
595 	}
596 }
597 
598 /**
599  * gfs2_rs_alloc - make sure we have a reservation assigned to the inode
600  * @ip: the inode for this reservation
601  */
602 int gfs2_rs_alloc(struct gfs2_inode *ip)
603 {
604 	int error = 0;
605 
606 	down_write(&ip->i_rw_mutex);
607 	if (ip->i_res)
608 		goto out;
609 
610 	ip->i_res = kmem_cache_zalloc(gfs2_rsrv_cachep, GFP_NOFS);
611 	if (!ip->i_res) {
612 		error = -ENOMEM;
613 		goto out;
614 	}
615 
616 	RB_CLEAR_NODE(&ip->i_res->rs_node);
617 out:
618 	up_write(&ip->i_rw_mutex);
619 	return error;
620 }
621 
622 static void dump_rs(struct seq_file *seq, const struct gfs2_blkreserv *rs)
623 {
624 	gfs2_print_dbg(seq, "  B: n:%llu s:%llu b:%u f:%u\n",
625 		       (unsigned long long)rs->rs_inum,
626 		       (unsigned long long)gfs2_rbm_to_block(&rs->rs_rbm),
627 		       rs->rs_rbm.offset, rs->rs_free);
628 }
629 
630 /**
631  * __rs_deltree - remove a multi-block reservation from the rgd tree
632  * @rs: The reservation to remove
633  *
634  */
635 static void __rs_deltree(struct gfs2_blkreserv *rs)
636 {
637 	struct gfs2_rgrpd *rgd;
638 
639 	if (!gfs2_rs_active(rs))
640 		return;
641 
642 	rgd = rs->rs_rbm.rgd;
643 	trace_gfs2_rs(rs, TRACE_RS_TREEDEL);
644 	rb_erase(&rs->rs_node, &rgd->rd_rstree);
645 	RB_CLEAR_NODE(&rs->rs_node);
646 
647 	if (rs->rs_free) {
648 		struct gfs2_bitmap *bi = rbm_bi(&rs->rs_rbm);
649 
650 		/* return reserved blocks to the rgrp */
651 		BUG_ON(rs->rs_rbm.rgd->rd_reserved < rs->rs_free);
652 		rs->rs_rbm.rgd->rd_reserved -= rs->rs_free;
653 		/* The rgrp extent failure point is likely not to increase;
654 		   it will only do so if the freed blocks are somehow
655 		   contiguous with a span of free blocks that follows. Still,
656 		   it will force the number to be recalculated later. */
657 		rgd->rd_extfail_pt += rs->rs_free;
658 		rs->rs_free = 0;
659 		clear_bit(GBF_FULL, &bi->bi_flags);
660 	}
661 }
662 
663 /**
664  * gfs2_rs_deltree - remove a multi-block reservation from the rgd tree
665  * @rs: The reservation to remove
666  *
667  */
668 void gfs2_rs_deltree(struct gfs2_blkreserv *rs)
669 {
670 	struct gfs2_rgrpd *rgd;
671 
672 	rgd = rs->rs_rbm.rgd;
673 	if (rgd) {
674 		spin_lock(&rgd->rd_rsspin);
675 		__rs_deltree(rs);
676 		spin_unlock(&rgd->rd_rsspin);
677 	}
678 }
679 
680 /**
681  * gfs2_rs_delete - delete a multi-block reservation
682  * @ip: The inode for this reservation
683  * @wcount: The inode's write count, or NULL
684  *
685  */
686 void gfs2_rs_delete(struct gfs2_inode *ip, atomic_t *wcount)
687 {
688 	down_write(&ip->i_rw_mutex);
689 	if (ip->i_res && ((wcount == NULL) || (atomic_read(wcount) <= 1))) {
690 		gfs2_rs_deltree(ip->i_res);
691 		BUG_ON(ip->i_res->rs_free);
692 		kmem_cache_free(gfs2_rsrv_cachep, ip->i_res);
693 		ip->i_res = NULL;
694 	}
695 	up_write(&ip->i_rw_mutex);
696 }
697 
698 /**
699  * return_all_reservations - return all reserved blocks back to the rgrp.
700  * @rgd: the rgrp that needs its space back
701  *
702  * We previously reserved a bunch of blocks for allocation. Now we need to
703  * give them back. This leave the reservation structures in tact, but removes
704  * all of their corresponding "no-fly zones".
705  */
706 static void return_all_reservations(struct gfs2_rgrpd *rgd)
707 {
708 	struct rb_node *n;
709 	struct gfs2_blkreserv *rs;
710 
711 	spin_lock(&rgd->rd_rsspin);
712 	while ((n = rb_first(&rgd->rd_rstree))) {
713 		rs = rb_entry(n, struct gfs2_blkreserv, rs_node);
714 		__rs_deltree(rs);
715 	}
716 	spin_unlock(&rgd->rd_rsspin);
717 }
718 
719 void gfs2_clear_rgrpd(struct gfs2_sbd *sdp)
720 {
721 	struct rb_node *n;
722 	struct gfs2_rgrpd *rgd;
723 	struct gfs2_glock *gl;
724 
725 	while ((n = rb_first(&sdp->sd_rindex_tree))) {
726 		rgd = rb_entry(n, struct gfs2_rgrpd, rd_node);
727 		gl = rgd->rd_gl;
728 
729 		rb_erase(n, &sdp->sd_rindex_tree);
730 
731 		if (gl) {
732 			spin_lock(&gl->gl_spin);
733 			gl->gl_object = NULL;
734 			spin_unlock(&gl->gl_spin);
735 			gfs2_glock_add_to_lru(gl);
736 			gfs2_glock_put(gl);
737 		}
738 
739 		gfs2_free_clones(rgd);
740 		kfree(rgd->rd_bits);
741 		return_all_reservations(rgd);
742 		kmem_cache_free(gfs2_rgrpd_cachep, rgd);
743 	}
744 }
745 
746 static void gfs2_rindex_print(const struct gfs2_rgrpd *rgd)
747 {
748 	pr_info("ri_addr = %llu\n", (unsigned long long)rgd->rd_addr);
749 	pr_info("ri_length = %u\n", rgd->rd_length);
750 	pr_info("ri_data0 = %llu\n", (unsigned long long)rgd->rd_data0);
751 	pr_info("ri_data = %u\n", rgd->rd_data);
752 	pr_info("ri_bitbytes = %u\n", rgd->rd_bitbytes);
753 }
754 
755 /**
756  * gfs2_compute_bitstructs - Compute the bitmap sizes
757  * @rgd: The resource group descriptor
758  *
759  * Calculates bitmap descriptors, one for each block that contains bitmap data
760  *
761  * Returns: errno
762  */
763 
764 static int compute_bitstructs(struct gfs2_rgrpd *rgd)
765 {
766 	struct gfs2_sbd *sdp = rgd->rd_sbd;
767 	struct gfs2_bitmap *bi;
768 	u32 length = rgd->rd_length; /* # blocks in hdr & bitmap */
769 	u32 bytes_left, bytes;
770 	int x;
771 
772 	if (!length)
773 		return -EINVAL;
774 
775 	rgd->rd_bits = kcalloc(length, sizeof(struct gfs2_bitmap), GFP_NOFS);
776 	if (!rgd->rd_bits)
777 		return -ENOMEM;
778 
779 	bytes_left = rgd->rd_bitbytes;
780 
781 	for (x = 0; x < length; x++) {
782 		bi = rgd->rd_bits + x;
783 
784 		bi->bi_flags = 0;
785 		/* small rgrp; bitmap stored completely in header block */
786 		if (length == 1) {
787 			bytes = bytes_left;
788 			bi->bi_offset = sizeof(struct gfs2_rgrp);
789 			bi->bi_start = 0;
790 			bi->bi_len = bytes;
791 			bi->bi_blocks = bytes * GFS2_NBBY;
792 		/* header block */
793 		} else if (x == 0) {
794 			bytes = sdp->sd_sb.sb_bsize - sizeof(struct gfs2_rgrp);
795 			bi->bi_offset = sizeof(struct gfs2_rgrp);
796 			bi->bi_start = 0;
797 			bi->bi_len = bytes;
798 			bi->bi_blocks = bytes * GFS2_NBBY;
799 		/* last block */
800 		} else if (x + 1 == length) {
801 			bytes = bytes_left;
802 			bi->bi_offset = sizeof(struct gfs2_meta_header);
803 			bi->bi_start = rgd->rd_bitbytes - bytes_left;
804 			bi->bi_len = bytes;
805 			bi->bi_blocks = bytes * GFS2_NBBY;
806 		/* other blocks */
807 		} else {
808 			bytes = sdp->sd_sb.sb_bsize -
809 				sizeof(struct gfs2_meta_header);
810 			bi->bi_offset = sizeof(struct gfs2_meta_header);
811 			bi->bi_start = rgd->rd_bitbytes - bytes_left;
812 			bi->bi_len = bytes;
813 			bi->bi_blocks = bytes * GFS2_NBBY;
814 		}
815 
816 		bytes_left -= bytes;
817 	}
818 
819 	if (bytes_left) {
820 		gfs2_consist_rgrpd(rgd);
821 		return -EIO;
822 	}
823 	bi = rgd->rd_bits + (length - 1);
824 	if ((bi->bi_start + bi->bi_len) * GFS2_NBBY != rgd->rd_data) {
825 		if (gfs2_consist_rgrpd(rgd)) {
826 			gfs2_rindex_print(rgd);
827 			fs_err(sdp, "start=%u len=%u offset=%u\n",
828 			       bi->bi_start, bi->bi_len, bi->bi_offset);
829 		}
830 		return -EIO;
831 	}
832 
833 	return 0;
834 }
835 
836 /**
837  * gfs2_ri_total - Total up the file system space, according to the rindex.
838  * @sdp: the filesystem
839  *
840  */
841 u64 gfs2_ri_total(struct gfs2_sbd *sdp)
842 {
843 	u64 total_data = 0;
844 	struct inode *inode = sdp->sd_rindex;
845 	struct gfs2_inode *ip = GFS2_I(inode);
846 	char buf[sizeof(struct gfs2_rindex)];
847 	int error, rgrps;
848 
849 	for (rgrps = 0;; rgrps++) {
850 		loff_t pos = rgrps * sizeof(struct gfs2_rindex);
851 
852 		if (pos + sizeof(struct gfs2_rindex) > i_size_read(inode))
853 			break;
854 		error = gfs2_internal_read(ip, buf, &pos,
855 					   sizeof(struct gfs2_rindex));
856 		if (error != sizeof(struct gfs2_rindex))
857 			break;
858 		total_data += be32_to_cpu(((struct gfs2_rindex *)buf)->ri_data);
859 	}
860 	return total_data;
861 }
862 
863 static int rgd_insert(struct gfs2_rgrpd *rgd)
864 {
865 	struct gfs2_sbd *sdp = rgd->rd_sbd;
866 	struct rb_node **newn = &sdp->sd_rindex_tree.rb_node, *parent = NULL;
867 
868 	/* Figure out where to put new node */
869 	while (*newn) {
870 		struct gfs2_rgrpd *cur = rb_entry(*newn, struct gfs2_rgrpd,
871 						  rd_node);
872 
873 		parent = *newn;
874 		if (rgd->rd_addr < cur->rd_addr)
875 			newn = &((*newn)->rb_left);
876 		else if (rgd->rd_addr > cur->rd_addr)
877 			newn = &((*newn)->rb_right);
878 		else
879 			return -EEXIST;
880 	}
881 
882 	rb_link_node(&rgd->rd_node, parent, newn);
883 	rb_insert_color(&rgd->rd_node, &sdp->sd_rindex_tree);
884 	sdp->sd_rgrps++;
885 	return 0;
886 }
887 
888 /**
889  * read_rindex_entry - Pull in a new resource index entry from the disk
890  * @ip: Pointer to the rindex inode
891  *
892  * Returns: 0 on success, > 0 on EOF, error code otherwise
893  */
894 
895 static int read_rindex_entry(struct gfs2_inode *ip)
896 {
897 	struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode);
898 	const unsigned bsize = sdp->sd_sb.sb_bsize;
899 	loff_t pos = sdp->sd_rgrps * sizeof(struct gfs2_rindex);
900 	struct gfs2_rindex buf;
901 	int error;
902 	struct gfs2_rgrpd *rgd;
903 
904 	if (pos >= i_size_read(&ip->i_inode))
905 		return 1;
906 
907 	error = gfs2_internal_read(ip, (char *)&buf, &pos,
908 				   sizeof(struct gfs2_rindex));
909 
910 	if (error != sizeof(struct gfs2_rindex))
911 		return (error == 0) ? 1 : error;
912 
913 	rgd = kmem_cache_zalloc(gfs2_rgrpd_cachep, GFP_NOFS);
914 	error = -ENOMEM;
915 	if (!rgd)
916 		return error;
917 
918 	rgd->rd_sbd = sdp;
919 	rgd->rd_addr = be64_to_cpu(buf.ri_addr);
920 	rgd->rd_length = be32_to_cpu(buf.ri_length);
921 	rgd->rd_data0 = be64_to_cpu(buf.ri_data0);
922 	rgd->rd_data = be32_to_cpu(buf.ri_data);
923 	rgd->rd_bitbytes = be32_to_cpu(buf.ri_bitbytes);
924 	spin_lock_init(&rgd->rd_rsspin);
925 
926 	error = compute_bitstructs(rgd);
927 	if (error)
928 		goto fail;
929 
930 	error = gfs2_glock_get(sdp, rgd->rd_addr,
931 			       &gfs2_rgrp_glops, CREATE, &rgd->rd_gl);
932 	if (error)
933 		goto fail;
934 
935 	rgd->rd_gl->gl_object = rgd;
936 	rgd->rd_gl->gl_vm.start = rgd->rd_addr * bsize;
937 	rgd->rd_gl->gl_vm.end = rgd->rd_gl->gl_vm.start + (rgd->rd_length * bsize) - 1;
938 	rgd->rd_rgl = (struct gfs2_rgrp_lvb *)rgd->rd_gl->gl_lksb.sb_lvbptr;
939 	rgd->rd_flags &= ~(GFS2_RDF_UPTODATE | GFS2_RDF_PREFERRED);
940 	if (rgd->rd_data > sdp->sd_max_rg_data)
941 		sdp->sd_max_rg_data = rgd->rd_data;
942 	spin_lock(&sdp->sd_rindex_spin);
943 	error = rgd_insert(rgd);
944 	spin_unlock(&sdp->sd_rindex_spin);
945 	if (!error)
946 		return 0;
947 
948 	error = 0; /* someone else read in the rgrp; free it and ignore it */
949 	gfs2_glock_put(rgd->rd_gl);
950 
951 fail:
952 	kfree(rgd->rd_bits);
953 	kmem_cache_free(gfs2_rgrpd_cachep, rgd);
954 	return error;
955 }
956 
957 /**
958  * set_rgrp_preferences - Run all the rgrps, selecting some we prefer to use
959  * @sdp: the GFS2 superblock
960  *
961  * The purpose of this function is to select a subset of the resource groups
962  * and mark them as PREFERRED. We do it in such a way that each node prefers
963  * to use a unique set of rgrps to minimize glock contention.
964  */
965 static void set_rgrp_preferences(struct gfs2_sbd *sdp)
966 {
967 	struct gfs2_rgrpd *rgd, *first;
968 	int i;
969 
970 	/* Skip an initial number of rgrps, based on this node's journal ID.
971 	   That should start each node out on its own set. */
972 	rgd = gfs2_rgrpd_get_first(sdp);
973 	for (i = 0; i < sdp->sd_lockstruct.ls_jid; i++)
974 		rgd = gfs2_rgrpd_get_next(rgd);
975 	first = rgd;
976 
977 	do {
978 		rgd->rd_flags |= GFS2_RDF_PREFERRED;
979 		for (i = 0; i < sdp->sd_journals; i++) {
980 			rgd = gfs2_rgrpd_get_next(rgd);
981 			if (rgd == first)
982 				break;
983 		}
984 	} while (rgd != first);
985 }
986 
987 /**
988  * gfs2_ri_update - Pull in a new resource index from the disk
989  * @ip: pointer to the rindex inode
990  *
991  * Returns: 0 on successful update, error code otherwise
992  */
993 
994 static int gfs2_ri_update(struct gfs2_inode *ip)
995 {
996 	struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode);
997 	int error;
998 
999 	do {
1000 		error = read_rindex_entry(ip);
1001 	} while (error == 0);
1002 
1003 	if (error < 0)
1004 		return error;
1005 
1006 	set_rgrp_preferences(sdp);
1007 
1008 	sdp->sd_rindex_uptodate = 1;
1009 	return 0;
1010 }
1011 
1012 /**
1013  * gfs2_rindex_update - Update the rindex if required
1014  * @sdp: The GFS2 superblock
1015  *
1016  * We grab a lock on the rindex inode to make sure that it doesn't
1017  * change whilst we are performing an operation. We keep this lock
1018  * for quite long periods of time compared to other locks. This
1019  * doesn't matter, since it is shared and it is very, very rarely
1020  * accessed in the exclusive mode (i.e. only when expanding the filesystem).
1021  *
1022  * This makes sure that we're using the latest copy of the resource index
1023  * special file, which might have been updated if someone expanded the
1024  * filesystem (via gfs2_grow utility), which adds new resource groups.
1025  *
1026  * Returns: 0 on succeess, error code otherwise
1027  */
1028 
1029 int gfs2_rindex_update(struct gfs2_sbd *sdp)
1030 {
1031 	struct gfs2_inode *ip = GFS2_I(sdp->sd_rindex);
1032 	struct gfs2_glock *gl = ip->i_gl;
1033 	struct gfs2_holder ri_gh;
1034 	int error = 0;
1035 	int unlock_required = 0;
1036 
1037 	/* Read new copy from disk if we don't have the latest */
1038 	if (!sdp->sd_rindex_uptodate) {
1039 		if (!gfs2_glock_is_locked_by_me(gl)) {
1040 			error = gfs2_glock_nq_init(gl, LM_ST_SHARED, 0, &ri_gh);
1041 			if (error)
1042 				return error;
1043 			unlock_required = 1;
1044 		}
1045 		if (!sdp->sd_rindex_uptodate)
1046 			error = gfs2_ri_update(ip);
1047 		if (unlock_required)
1048 			gfs2_glock_dq_uninit(&ri_gh);
1049 	}
1050 
1051 	return error;
1052 }
1053 
1054 static void gfs2_rgrp_in(struct gfs2_rgrpd *rgd, const void *buf)
1055 {
1056 	const struct gfs2_rgrp *str = buf;
1057 	u32 rg_flags;
1058 
1059 	rg_flags = be32_to_cpu(str->rg_flags);
1060 	rg_flags &= ~GFS2_RDF_MASK;
1061 	rgd->rd_flags &= GFS2_RDF_MASK;
1062 	rgd->rd_flags |= rg_flags;
1063 	rgd->rd_free = be32_to_cpu(str->rg_free);
1064 	rgd->rd_dinodes = be32_to_cpu(str->rg_dinodes);
1065 	rgd->rd_igeneration = be64_to_cpu(str->rg_igeneration);
1066 }
1067 
1068 static void gfs2_rgrp_out(struct gfs2_rgrpd *rgd, void *buf)
1069 {
1070 	struct gfs2_rgrp *str = buf;
1071 
1072 	str->rg_flags = cpu_to_be32(rgd->rd_flags & ~GFS2_RDF_MASK);
1073 	str->rg_free = cpu_to_be32(rgd->rd_free);
1074 	str->rg_dinodes = cpu_to_be32(rgd->rd_dinodes);
1075 	str->__pad = cpu_to_be32(0);
1076 	str->rg_igeneration = cpu_to_be64(rgd->rd_igeneration);
1077 	memset(&str->rg_reserved, 0, sizeof(str->rg_reserved));
1078 }
1079 
1080 static int gfs2_rgrp_lvb_valid(struct gfs2_rgrpd *rgd)
1081 {
1082 	struct gfs2_rgrp_lvb *rgl = rgd->rd_rgl;
1083 	struct gfs2_rgrp *str = (struct gfs2_rgrp *)rgd->rd_bits[0].bi_bh->b_data;
1084 
1085 	if (rgl->rl_flags != str->rg_flags || rgl->rl_free != str->rg_free ||
1086 	    rgl->rl_dinodes != str->rg_dinodes ||
1087 	    rgl->rl_igeneration != str->rg_igeneration)
1088 		return 0;
1089 	return 1;
1090 }
1091 
1092 static void gfs2_rgrp_ondisk2lvb(struct gfs2_rgrp_lvb *rgl, const void *buf)
1093 {
1094 	const struct gfs2_rgrp *str = buf;
1095 
1096 	rgl->rl_magic = cpu_to_be32(GFS2_MAGIC);
1097 	rgl->rl_flags = str->rg_flags;
1098 	rgl->rl_free = str->rg_free;
1099 	rgl->rl_dinodes = str->rg_dinodes;
1100 	rgl->rl_igeneration = str->rg_igeneration;
1101 	rgl->__pad = 0UL;
1102 }
1103 
1104 static void update_rgrp_lvb_unlinked(struct gfs2_rgrpd *rgd, u32 change)
1105 {
1106 	struct gfs2_rgrp_lvb *rgl = rgd->rd_rgl;
1107 	u32 unlinked = be32_to_cpu(rgl->rl_unlinked) + change;
1108 	rgl->rl_unlinked = cpu_to_be32(unlinked);
1109 }
1110 
1111 static u32 count_unlinked(struct gfs2_rgrpd *rgd)
1112 {
1113 	struct gfs2_bitmap *bi;
1114 	const u32 length = rgd->rd_length;
1115 	const u8 *buffer = NULL;
1116 	u32 i, goal, count = 0;
1117 
1118 	for (i = 0, bi = rgd->rd_bits; i < length; i++, bi++) {
1119 		goal = 0;
1120 		buffer = bi->bi_bh->b_data + bi->bi_offset;
1121 		WARN_ON(!buffer_uptodate(bi->bi_bh));
1122 		while (goal < bi->bi_len * GFS2_NBBY) {
1123 			goal = gfs2_bitfit(buffer, bi->bi_len, goal,
1124 					   GFS2_BLKST_UNLINKED);
1125 			if (goal == BFITNOENT)
1126 				break;
1127 			count++;
1128 			goal++;
1129 		}
1130 	}
1131 
1132 	return count;
1133 }
1134 
1135 
1136 /**
1137  * gfs2_rgrp_bh_get - Read in a RG's header and bitmaps
1138  * @rgd: the struct gfs2_rgrpd describing the RG to read in
1139  *
1140  * Read in all of a Resource Group's header and bitmap blocks.
1141  * Caller must eventually call gfs2_rgrp_relse() to free the bitmaps.
1142  *
1143  * Returns: errno
1144  */
1145 
1146 static int gfs2_rgrp_bh_get(struct gfs2_rgrpd *rgd)
1147 {
1148 	struct gfs2_sbd *sdp = rgd->rd_sbd;
1149 	struct gfs2_glock *gl = rgd->rd_gl;
1150 	unsigned int length = rgd->rd_length;
1151 	struct gfs2_bitmap *bi;
1152 	unsigned int x, y;
1153 	int error;
1154 
1155 	if (rgd->rd_bits[0].bi_bh != NULL)
1156 		return 0;
1157 
1158 	for (x = 0; x < length; x++) {
1159 		bi = rgd->rd_bits + x;
1160 		error = gfs2_meta_read(gl, rgd->rd_addr + x, 0, &bi->bi_bh);
1161 		if (error)
1162 			goto fail;
1163 	}
1164 
1165 	for (y = length; y--;) {
1166 		bi = rgd->rd_bits + y;
1167 		error = gfs2_meta_wait(sdp, bi->bi_bh);
1168 		if (error)
1169 			goto fail;
1170 		if (gfs2_metatype_check(sdp, bi->bi_bh, y ? GFS2_METATYPE_RB :
1171 					      GFS2_METATYPE_RG)) {
1172 			error = -EIO;
1173 			goto fail;
1174 		}
1175 	}
1176 
1177 	if (!(rgd->rd_flags & GFS2_RDF_UPTODATE)) {
1178 		for (x = 0; x < length; x++)
1179 			clear_bit(GBF_FULL, &rgd->rd_bits[x].bi_flags);
1180 		gfs2_rgrp_in(rgd, (rgd->rd_bits[0].bi_bh)->b_data);
1181 		rgd->rd_flags |= (GFS2_RDF_UPTODATE | GFS2_RDF_CHECK);
1182 		rgd->rd_free_clone = rgd->rd_free;
1183 		/* max out the rgrp allocation failure point */
1184 		rgd->rd_extfail_pt = rgd->rd_free;
1185 	}
1186 	if (cpu_to_be32(GFS2_MAGIC) != rgd->rd_rgl->rl_magic) {
1187 		rgd->rd_rgl->rl_unlinked = cpu_to_be32(count_unlinked(rgd));
1188 		gfs2_rgrp_ondisk2lvb(rgd->rd_rgl,
1189 				     rgd->rd_bits[0].bi_bh->b_data);
1190 	}
1191 	else if (sdp->sd_args.ar_rgrplvb) {
1192 		if (!gfs2_rgrp_lvb_valid(rgd)){
1193 			gfs2_consist_rgrpd(rgd);
1194 			error = -EIO;
1195 			goto fail;
1196 		}
1197 		if (rgd->rd_rgl->rl_unlinked == 0)
1198 			rgd->rd_flags &= ~GFS2_RDF_CHECK;
1199 	}
1200 	return 0;
1201 
1202 fail:
1203 	while (x--) {
1204 		bi = rgd->rd_bits + x;
1205 		brelse(bi->bi_bh);
1206 		bi->bi_bh = NULL;
1207 		gfs2_assert_warn(sdp, !bi->bi_clone);
1208 	}
1209 
1210 	return error;
1211 }
1212 
1213 static int update_rgrp_lvb(struct gfs2_rgrpd *rgd)
1214 {
1215 	u32 rl_flags;
1216 
1217 	if (rgd->rd_flags & GFS2_RDF_UPTODATE)
1218 		return 0;
1219 
1220 	if (cpu_to_be32(GFS2_MAGIC) != rgd->rd_rgl->rl_magic)
1221 		return gfs2_rgrp_bh_get(rgd);
1222 
1223 	rl_flags = be32_to_cpu(rgd->rd_rgl->rl_flags);
1224 	rl_flags &= ~GFS2_RDF_MASK;
1225 	rgd->rd_flags &= GFS2_RDF_MASK;
1226 	rgd->rd_flags |= (rl_flags | GFS2_RDF_UPTODATE | GFS2_RDF_CHECK);
1227 	if (rgd->rd_rgl->rl_unlinked == 0)
1228 		rgd->rd_flags &= ~GFS2_RDF_CHECK;
1229 	rgd->rd_free = be32_to_cpu(rgd->rd_rgl->rl_free);
1230 	rgd->rd_free_clone = rgd->rd_free;
1231 	rgd->rd_dinodes = be32_to_cpu(rgd->rd_rgl->rl_dinodes);
1232 	rgd->rd_igeneration = be64_to_cpu(rgd->rd_rgl->rl_igeneration);
1233 	return 0;
1234 }
1235 
1236 int gfs2_rgrp_go_lock(struct gfs2_holder *gh)
1237 {
1238 	struct gfs2_rgrpd *rgd = gh->gh_gl->gl_object;
1239 	struct gfs2_sbd *sdp = rgd->rd_sbd;
1240 
1241 	if (gh->gh_flags & GL_SKIP && sdp->sd_args.ar_rgrplvb)
1242 		return 0;
1243 	return gfs2_rgrp_bh_get(rgd);
1244 }
1245 
1246 /**
1247  * gfs2_rgrp_go_unlock - Release RG bitmaps read in with gfs2_rgrp_bh_get()
1248  * @gh: The glock holder for the resource group
1249  *
1250  */
1251 
1252 void gfs2_rgrp_go_unlock(struct gfs2_holder *gh)
1253 {
1254 	struct gfs2_rgrpd *rgd = gh->gh_gl->gl_object;
1255 	int x, length = rgd->rd_length;
1256 
1257 	for (x = 0; x < length; x++) {
1258 		struct gfs2_bitmap *bi = rgd->rd_bits + x;
1259 		if (bi->bi_bh) {
1260 			brelse(bi->bi_bh);
1261 			bi->bi_bh = NULL;
1262 		}
1263 	}
1264 
1265 }
1266 
1267 int gfs2_rgrp_send_discards(struct gfs2_sbd *sdp, u64 offset,
1268 			     struct buffer_head *bh,
1269 			     const struct gfs2_bitmap *bi, unsigned minlen, u64 *ptrimmed)
1270 {
1271 	struct super_block *sb = sdp->sd_vfs;
1272 	u64 blk;
1273 	sector_t start = 0;
1274 	sector_t nr_blks = 0;
1275 	int rv;
1276 	unsigned int x;
1277 	u32 trimmed = 0;
1278 	u8 diff;
1279 
1280 	for (x = 0; x < bi->bi_len; x++) {
1281 		const u8 *clone = bi->bi_clone ? bi->bi_clone : bi->bi_bh->b_data;
1282 		clone += bi->bi_offset;
1283 		clone += x;
1284 		if (bh) {
1285 			const u8 *orig = bh->b_data + bi->bi_offset + x;
1286 			diff = ~(*orig | (*orig >> 1)) & (*clone | (*clone >> 1));
1287 		} else {
1288 			diff = ~(*clone | (*clone >> 1));
1289 		}
1290 		diff &= 0x55;
1291 		if (diff == 0)
1292 			continue;
1293 		blk = offset + ((bi->bi_start + x) * GFS2_NBBY);
1294 		while(diff) {
1295 			if (diff & 1) {
1296 				if (nr_blks == 0)
1297 					goto start_new_extent;
1298 				if ((start + nr_blks) != blk) {
1299 					if (nr_blks >= minlen) {
1300 						rv = sb_issue_discard(sb,
1301 							start, nr_blks,
1302 							GFP_NOFS, 0);
1303 						if (rv)
1304 							goto fail;
1305 						trimmed += nr_blks;
1306 					}
1307 					nr_blks = 0;
1308 start_new_extent:
1309 					start = blk;
1310 				}
1311 				nr_blks++;
1312 			}
1313 			diff >>= 2;
1314 			blk++;
1315 		}
1316 	}
1317 	if (nr_blks >= minlen) {
1318 		rv = sb_issue_discard(sb, start, nr_blks, GFP_NOFS, 0);
1319 		if (rv)
1320 			goto fail;
1321 		trimmed += nr_blks;
1322 	}
1323 	if (ptrimmed)
1324 		*ptrimmed = trimmed;
1325 	return 0;
1326 
1327 fail:
1328 	if (sdp->sd_args.ar_discard)
1329 		fs_warn(sdp, "error %d on discard request, turning discards off for this filesystem", rv);
1330 	sdp->sd_args.ar_discard = 0;
1331 	return -EIO;
1332 }
1333 
1334 /**
1335  * gfs2_fitrim - Generate discard requests for unused bits of the filesystem
1336  * @filp: Any file on the filesystem
1337  * @argp: Pointer to the arguments (also used to pass result)
1338  *
1339  * Returns: 0 on success, otherwise error code
1340  */
1341 
1342 int gfs2_fitrim(struct file *filp, void __user *argp)
1343 {
1344 	struct inode *inode = file_inode(filp);
1345 	struct gfs2_sbd *sdp = GFS2_SB(inode);
1346 	struct request_queue *q = bdev_get_queue(sdp->sd_vfs->s_bdev);
1347 	struct buffer_head *bh;
1348 	struct gfs2_rgrpd *rgd;
1349 	struct gfs2_rgrpd *rgd_end;
1350 	struct gfs2_holder gh;
1351 	struct fstrim_range r;
1352 	int ret = 0;
1353 	u64 amt;
1354 	u64 trimmed = 0;
1355 	u64 start, end, minlen;
1356 	unsigned int x;
1357 	unsigned bs_shift = sdp->sd_sb.sb_bsize_shift;
1358 
1359 	if (!capable(CAP_SYS_ADMIN))
1360 		return -EPERM;
1361 
1362 	if (!blk_queue_discard(q))
1363 		return -EOPNOTSUPP;
1364 
1365 	if (copy_from_user(&r, argp, sizeof(r)))
1366 		return -EFAULT;
1367 
1368 	ret = gfs2_rindex_update(sdp);
1369 	if (ret)
1370 		return ret;
1371 
1372 	start = r.start >> bs_shift;
1373 	end = start + (r.len >> bs_shift);
1374 	minlen = max_t(u64, r.minlen,
1375 		       q->limits.discard_granularity) >> bs_shift;
1376 
1377 	if (end <= start || minlen > sdp->sd_max_rg_data)
1378 		return -EINVAL;
1379 
1380 	rgd = gfs2_blk2rgrpd(sdp, start, 0);
1381 	rgd_end = gfs2_blk2rgrpd(sdp, end, 0);
1382 
1383 	if ((gfs2_rgrpd_get_first(sdp) == gfs2_rgrpd_get_next(rgd_end))
1384 	    && (start > rgd_end->rd_data0 + rgd_end->rd_data))
1385 		return -EINVAL; /* start is beyond the end of the fs */
1386 
1387 	while (1) {
1388 
1389 		ret = gfs2_glock_nq_init(rgd->rd_gl, LM_ST_EXCLUSIVE, 0, &gh);
1390 		if (ret)
1391 			goto out;
1392 
1393 		if (!(rgd->rd_flags & GFS2_RGF_TRIMMED)) {
1394 			/* Trim each bitmap in the rgrp */
1395 			for (x = 0; x < rgd->rd_length; x++) {
1396 				struct gfs2_bitmap *bi = rgd->rd_bits + x;
1397 				ret = gfs2_rgrp_send_discards(sdp,
1398 						rgd->rd_data0, NULL, bi, minlen,
1399 						&amt);
1400 				if (ret) {
1401 					gfs2_glock_dq_uninit(&gh);
1402 					goto out;
1403 				}
1404 				trimmed += amt;
1405 			}
1406 
1407 			/* Mark rgrp as having been trimmed */
1408 			ret = gfs2_trans_begin(sdp, RES_RG_HDR, 0);
1409 			if (ret == 0) {
1410 				bh = rgd->rd_bits[0].bi_bh;
1411 				rgd->rd_flags |= GFS2_RGF_TRIMMED;
1412 				gfs2_trans_add_meta(rgd->rd_gl, bh);
1413 				gfs2_rgrp_out(rgd, bh->b_data);
1414 				gfs2_rgrp_ondisk2lvb(rgd->rd_rgl, bh->b_data);
1415 				gfs2_trans_end(sdp);
1416 			}
1417 		}
1418 		gfs2_glock_dq_uninit(&gh);
1419 
1420 		if (rgd == rgd_end)
1421 			break;
1422 
1423 		rgd = gfs2_rgrpd_get_next(rgd);
1424 	}
1425 
1426 out:
1427 	r.len = trimmed << bs_shift;
1428 	if (copy_to_user(argp, &r, sizeof(r)))
1429 		return -EFAULT;
1430 
1431 	return ret;
1432 }
1433 
1434 /**
1435  * rs_insert - insert a new multi-block reservation into the rgrp's rb_tree
1436  * @ip: the inode structure
1437  *
1438  */
1439 static void rs_insert(struct gfs2_inode *ip)
1440 {
1441 	struct rb_node **newn, *parent = NULL;
1442 	int rc;
1443 	struct gfs2_blkreserv *rs = ip->i_res;
1444 	struct gfs2_rgrpd *rgd = rs->rs_rbm.rgd;
1445 	u64 fsblock = gfs2_rbm_to_block(&rs->rs_rbm);
1446 
1447 	BUG_ON(gfs2_rs_active(rs));
1448 
1449 	spin_lock(&rgd->rd_rsspin);
1450 	newn = &rgd->rd_rstree.rb_node;
1451 	while (*newn) {
1452 		struct gfs2_blkreserv *cur =
1453 			rb_entry(*newn, struct gfs2_blkreserv, rs_node);
1454 
1455 		parent = *newn;
1456 		rc = rs_cmp(fsblock, rs->rs_free, cur);
1457 		if (rc > 0)
1458 			newn = &((*newn)->rb_right);
1459 		else if (rc < 0)
1460 			newn = &((*newn)->rb_left);
1461 		else {
1462 			spin_unlock(&rgd->rd_rsspin);
1463 			WARN_ON(1);
1464 			return;
1465 		}
1466 	}
1467 
1468 	rb_link_node(&rs->rs_node, parent, newn);
1469 	rb_insert_color(&rs->rs_node, &rgd->rd_rstree);
1470 
1471 	/* Do our rgrp accounting for the reservation */
1472 	rgd->rd_reserved += rs->rs_free; /* blocks reserved */
1473 	spin_unlock(&rgd->rd_rsspin);
1474 	trace_gfs2_rs(rs, TRACE_RS_INSERT);
1475 }
1476 
1477 /**
1478  * rg_mblk_search - find a group of multiple free blocks to form a reservation
1479  * @rgd: the resource group descriptor
1480  * @ip: pointer to the inode for which we're reserving blocks
1481  * @ap: the allocation parameters
1482  *
1483  */
1484 
1485 static void rg_mblk_search(struct gfs2_rgrpd *rgd, struct gfs2_inode *ip,
1486 			   const struct gfs2_alloc_parms *ap)
1487 {
1488 	struct gfs2_rbm rbm = { .rgd = rgd, };
1489 	u64 goal;
1490 	struct gfs2_blkreserv *rs = ip->i_res;
1491 	u32 extlen;
1492 	u32 free_blocks = rgd->rd_free_clone - rgd->rd_reserved;
1493 	int ret;
1494 	struct inode *inode = &ip->i_inode;
1495 
1496 	if (S_ISDIR(inode->i_mode))
1497 		extlen = 1;
1498 	else {
1499 		extlen = max_t(u32, atomic_read(&rs->rs_sizehint), ap->target);
1500 		extlen = clamp(extlen, RGRP_RSRV_MINBLKS, free_blocks);
1501 	}
1502 	if ((rgd->rd_free_clone < rgd->rd_reserved) || (free_blocks < extlen))
1503 		return;
1504 
1505 	/* Find bitmap block that contains bits for goal block */
1506 	if (rgrp_contains_block(rgd, ip->i_goal))
1507 		goal = ip->i_goal;
1508 	else
1509 		goal = rgd->rd_last_alloc + rgd->rd_data0;
1510 
1511 	if (WARN_ON(gfs2_rbm_from_block(&rbm, goal)))
1512 		return;
1513 
1514 	ret = gfs2_rbm_find(&rbm, GFS2_BLKST_FREE, &extlen, ip, true, ap);
1515 	if (ret == 0) {
1516 		rs->rs_rbm = rbm;
1517 		rs->rs_free = extlen;
1518 		rs->rs_inum = ip->i_no_addr;
1519 		rs_insert(ip);
1520 	} else {
1521 		if (goal == rgd->rd_last_alloc + rgd->rd_data0)
1522 			rgd->rd_last_alloc = 0;
1523 	}
1524 }
1525 
1526 /**
1527  * gfs2_next_unreserved_block - Return next block that is not reserved
1528  * @rgd: The resource group
1529  * @block: The starting block
1530  * @length: The required length
1531  * @ip: Ignore any reservations for this inode
1532  *
1533  * If the block does not appear in any reservation, then return the
1534  * block number unchanged. If it does appear in the reservation, then
1535  * keep looking through the tree of reservations in order to find the
1536  * first block number which is not reserved.
1537  */
1538 
1539 static u64 gfs2_next_unreserved_block(struct gfs2_rgrpd *rgd, u64 block,
1540 				      u32 length,
1541 				      const struct gfs2_inode *ip)
1542 {
1543 	struct gfs2_blkreserv *rs;
1544 	struct rb_node *n;
1545 	int rc;
1546 
1547 	spin_lock(&rgd->rd_rsspin);
1548 	n = rgd->rd_rstree.rb_node;
1549 	while (n) {
1550 		rs = rb_entry(n, struct gfs2_blkreserv, rs_node);
1551 		rc = rs_cmp(block, length, rs);
1552 		if (rc < 0)
1553 			n = n->rb_left;
1554 		else if (rc > 0)
1555 			n = n->rb_right;
1556 		else
1557 			break;
1558 	}
1559 
1560 	if (n) {
1561 		while ((rs_cmp(block, length, rs) == 0) && (ip->i_res != rs)) {
1562 			block = gfs2_rbm_to_block(&rs->rs_rbm) + rs->rs_free;
1563 			n = n->rb_right;
1564 			if (n == NULL)
1565 				break;
1566 			rs = rb_entry(n, struct gfs2_blkreserv, rs_node);
1567 		}
1568 	}
1569 
1570 	spin_unlock(&rgd->rd_rsspin);
1571 	return block;
1572 }
1573 
1574 /**
1575  * gfs2_reservation_check_and_update - Check for reservations during block alloc
1576  * @rbm: The current position in the resource group
1577  * @ip: The inode for which we are searching for blocks
1578  * @minext: The minimum extent length
1579  * @maxext: A pointer to the maximum extent structure
1580  *
1581  * This checks the current position in the rgrp to see whether there is
1582  * a reservation covering this block. If not then this function is a
1583  * no-op. If there is, then the position is moved to the end of the
1584  * contiguous reservation(s) so that we are pointing at the first
1585  * non-reserved block.
1586  *
1587  * Returns: 0 if no reservation, 1 if @rbm has changed, otherwise an error
1588  */
1589 
1590 static int gfs2_reservation_check_and_update(struct gfs2_rbm *rbm,
1591 					     const struct gfs2_inode *ip,
1592 					     u32 minext,
1593 					     struct gfs2_extent *maxext)
1594 {
1595 	u64 block = gfs2_rbm_to_block(rbm);
1596 	u32 extlen = 1;
1597 	u64 nblock;
1598 	int ret;
1599 
1600 	/*
1601 	 * If we have a minimum extent length, then skip over any extent
1602 	 * which is less than the min extent length in size.
1603 	 */
1604 	if (minext) {
1605 		extlen = gfs2_free_extlen(rbm, minext);
1606 		if (extlen <= maxext->len)
1607 			goto fail;
1608 	}
1609 
1610 	/*
1611 	 * Check the extent which has been found against the reservations
1612 	 * and skip if parts of it are already reserved
1613 	 */
1614 	nblock = gfs2_next_unreserved_block(rbm->rgd, block, extlen, ip);
1615 	if (nblock == block) {
1616 		if (!minext || extlen >= minext)
1617 			return 0;
1618 
1619 		if (extlen > maxext->len) {
1620 			maxext->len = extlen;
1621 			maxext->rbm = *rbm;
1622 		}
1623 fail:
1624 		nblock = block + extlen;
1625 	}
1626 	ret = gfs2_rbm_from_block(rbm, nblock);
1627 	if (ret < 0)
1628 		return ret;
1629 	return 1;
1630 }
1631 
1632 /**
1633  * gfs2_rbm_find - Look for blocks of a particular state
1634  * @rbm: Value/result starting position and final position
1635  * @state: The state which we want to find
1636  * @minext: Pointer to the requested extent length (NULL for a single block)
1637  *          This is updated to be the actual reservation size.
1638  * @ip: If set, check for reservations
1639  * @nowrap: Stop looking at the end of the rgrp, rather than wrapping
1640  *          around until we've reached the starting point.
1641  * @ap: the allocation parameters
1642  *
1643  * Side effects:
1644  * - If looking for free blocks, we set GBF_FULL on each bitmap which
1645  *   has no free blocks in it.
1646  * - If looking for free blocks, we set rd_extfail_pt on each rgrp which
1647  *   has come up short on a free block search.
1648  *
1649  * Returns: 0 on success, -ENOSPC if there is no block of the requested state
1650  */
1651 
1652 static int gfs2_rbm_find(struct gfs2_rbm *rbm, u8 state, u32 *minext,
1653 			 const struct gfs2_inode *ip, bool nowrap,
1654 			 const struct gfs2_alloc_parms *ap)
1655 {
1656 	struct buffer_head *bh;
1657 	int initial_bii;
1658 	u32 initial_offset;
1659 	int first_bii = rbm->bii;
1660 	u32 first_offset = rbm->offset;
1661 	u32 offset;
1662 	u8 *buffer;
1663 	int n = 0;
1664 	int iters = rbm->rgd->rd_length;
1665 	int ret;
1666 	struct gfs2_bitmap *bi;
1667 	struct gfs2_extent maxext = { .rbm.rgd = rbm->rgd, };
1668 
1669 	/* If we are not starting at the beginning of a bitmap, then we
1670 	 * need to add one to the bitmap count to ensure that we search
1671 	 * the starting bitmap twice.
1672 	 */
1673 	if (rbm->offset != 0)
1674 		iters++;
1675 
1676 	while(1) {
1677 		bi = rbm_bi(rbm);
1678 		if (test_bit(GBF_FULL, &bi->bi_flags) &&
1679 		    (state == GFS2_BLKST_FREE))
1680 			goto next_bitmap;
1681 
1682 		bh = bi->bi_bh;
1683 		buffer = bh->b_data + bi->bi_offset;
1684 		WARN_ON(!buffer_uptodate(bh));
1685 		if (state != GFS2_BLKST_UNLINKED && bi->bi_clone)
1686 			buffer = bi->bi_clone + bi->bi_offset;
1687 		initial_offset = rbm->offset;
1688 		offset = gfs2_bitfit(buffer, bi->bi_len, rbm->offset, state);
1689 		if (offset == BFITNOENT)
1690 			goto bitmap_full;
1691 		rbm->offset = offset;
1692 		if (ip == NULL)
1693 			return 0;
1694 
1695 		initial_bii = rbm->bii;
1696 		ret = gfs2_reservation_check_and_update(rbm, ip,
1697 							minext ? *minext : 0,
1698 							&maxext);
1699 		if (ret == 0)
1700 			return 0;
1701 		if (ret > 0) {
1702 			n += (rbm->bii - initial_bii);
1703 			goto next_iter;
1704 		}
1705 		if (ret == -E2BIG) {
1706 			rbm->bii = 0;
1707 			rbm->offset = 0;
1708 			n += (rbm->bii - initial_bii);
1709 			goto res_covered_end_of_rgrp;
1710 		}
1711 		return ret;
1712 
1713 bitmap_full:	/* Mark bitmap as full and fall through */
1714 		if ((state == GFS2_BLKST_FREE) && initial_offset == 0) {
1715 			struct gfs2_bitmap *bi = rbm_bi(rbm);
1716 			set_bit(GBF_FULL, &bi->bi_flags);
1717 		}
1718 
1719 next_bitmap:	/* Find next bitmap in the rgrp */
1720 		rbm->offset = 0;
1721 		rbm->bii++;
1722 		if (rbm->bii == rbm->rgd->rd_length)
1723 			rbm->bii = 0;
1724 res_covered_end_of_rgrp:
1725 		if ((rbm->bii == 0) && nowrap)
1726 			break;
1727 		n++;
1728 next_iter:
1729 		if (n >= iters)
1730 			break;
1731 	}
1732 
1733 	if (minext == NULL || state != GFS2_BLKST_FREE)
1734 		return -ENOSPC;
1735 
1736 	/* If the extent was too small, and it's smaller than the smallest
1737 	   to have failed before, remember for future reference that it's
1738 	   useless to search this rgrp again for this amount or more. */
1739 	if ((first_offset == 0) && (first_bii == 0) &&
1740 	    (*minext < rbm->rgd->rd_extfail_pt))
1741 		rbm->rgd->rd_extfail_pt = *minext;
1742 
1743 	/* If the maximum extent we found is big enough to fulfill the
1744 	   minimum requirements, use it anyway. */
1745 	if (maxext.len) {
1746 		*rbm = maxext.rbm;
1747 		*minext = maxext.len;
1748 		return 0;
1749 	}
1750 
1751 	return -ENOSPC;
1752 }
1753 
1754 /**
1755  * try_rgrp_unlink - Look for any unlinked, allocated, but unused inodes
1756  * @rgd: The rgrp
1757  * @last_unlinked: block address of the last dinode we unlinked
1758  * @skip: block address we should explicitly not unlink
1759  *
1760  * Returns: 0 if no error
1761  *          The inode, if one has been found, in inode.
1762  */
1763 
1764 static void try_rgrp_unlink(struct gfs2_rgrpd *rgd, u64 *last_unlinked, u64 skip)
1765 {
1766 	u64 block;
1767 	struct gfs2_sbd *sdp = rgd->rd_sbd;
1768 	struct gfs2_glock *gl;
1769 	struct gfs2_inode *ip;
1770 	int error;
1771 	int found = 0;
1772 	struct gfs2_rbm rbm = { .rgd = rgd, .bii = 0, .offset = 0 };
1773 
1774 	while (1) {
1775 		down_write(&sdp->sd_log_flush_lock);
1776 		error = gfs2_rbm_find(&rbm, GFS2_BLKST_UNLINKED, NULL, NULL,
1777 				      true, NULL);
1778 		up_write(&sdp->sd_log_flush_lock);
1779 		if (error == -ENOSPC)
1780 			break;
1781 		if (WARN_ON_ONCE(error))
1782 			break;
1783 
1784 		block = gfs2_rbm_to_block(&rbm);
1785 		if (gfs2_rbm_from_block(&rbm, block + 1))
1786 			break;
1787 		if (*last_unlinked != NO_BLOCK && block <= *last_unlinked)
1788 			continue;
1789 		if (block == skip)
1790 			continue;
1791 		*last_unlinked = block;
1792 
1793 		error = gfs2_glock_get(sdp, block, &gfs2_inode_glops, CREATE, &gl);
1794 		if (error)
1795 			continue;
1796 
1797 		/* If the inode is already in cache, we can ignore it here
1798 		 * because the existing inode disposal code will deal with
1799 		 * it when all refs have gone away. Accessing gl_object like
1800 		 * this is not safe in general. Here it is ok because we do
1801 		 * not dereference the pointer, and we only need an approx
1802 		 * answer to whether it is NULL or not.
1803 		 */
1804 		ip = gl->gl_object;
1805 
1806 		if (ip || queue_work(gfs2_delete_workqueue, &gl->gl_delete) == 0)
1807 			gfs2_glock_put(gl);
1808 		else
1809 			found++;
1810 
1811 		/* Limit reclaim to sensible number of tasks */
1812 		if (found > NR_CPUS)
1813 			return;
1814 	}
1815 
1816 	rgd->rd_flags &= ~GFS2_RDF_CHECK;
1817 	return;
1818 }
1819 
1820 /**
1821  * gfs2_rgrp_congested - Use stats to figure out whether an rgrp is congested
1822  * @rgd: The rgrp in question
1823  * @loops: An indication of how picky we can be (0=very, 1=less so)
1824  *
1825  * This function uses the recently added glock statistics in order to
1826  * figure out whether a parciular resource group is suffering from
1827  * contention from multiple nodes. This is done purely on the basis
1828  * of timings, since this is the only data we have to work with and
1829  * our aim here is to reject a resource group which is highly contended
1830  * but (very important) not to do this too often in order to ensure that
1831  * we do not land up introducing fragmentation by changing resource
1832  * groups when not actually required.
1833  *
1834  * The calculation is fairly simple, we want to know whether the SRTTB
1835  * (i.e. smoothed round trip time for blocking operations) to acquire
1836  * the lock for this rgrp's glock is significantly greater than the
1837  * time taken for resource groups on average. We introduce a margin in
1838  * the form of the variable @var which is computed as the sum of the two
1839  * respective variences, and multiplied by a factor depending on @loops
1840  * and whether we have a lot of data to base the decision on. This is
1841  * then tested against the square difference of the means in order to
1842  * decide whether the result is statistically significant or not.
1843  *
1844  * Returns: A boolean verdict on the congestion status
1845  */
1846 
1847 static bool gfs2_rgrp_congested(const struct gfs2_rgrpd *rgd, int loops)
1848 {
1849 	const struct gfs2_glock *gl = rgd->rd_gl;
1850 	const struct gfs2_sbd *sdp = gl->gl_sbd;
1851 	struct gfs2_lkstats *st;
1852 	s64 r_dcount, l_dcount;
1853 	s64 r_srttb, l_srttb;
1854 	s64 srttb_diff;
1855 	s64 sqr_diff;
1856 	s64 var;
1857 
1858 	preempt_disable();
1859 	st = &this_cpu_ptr(sdp->sd_lkstats)->lkstats[LM_TYPE_RGRP];
1860 	r_srttb = st->stats[GFS2_LKS_SRTTB];
1861 	r_dcount = st->stats[GFS2_LKS_DCOUNT];
1862 	var = st->stats[GFS2_LKS_SRTTVARB] +
1863 	      gl->gl_stats.stats[GFS2_LKS_SRTTVARB];
1864 	preempt_enable();
1865 
1866 	l_srttb = gl->gl_stats.stats[GFS2_LKS_SRTTB];
1867 	l_dcount = gl->gl_stats.stats[GFS2_LKS_DCOUNT];
1868 
1869 	if ((l_dcount < 1) || (r_dcount < 1) || (r_srttb == 0))
1870 		return false;
1871 
1872 	srttb_diff = r_srttb - l_srttb;
1873 	sqr_diff = srttb_diff * srttb_diff;
1874 
1875 	var *= 2;
1876 	if (l_dcount < 8 || r_dcount < 8)
1877 		var *= 2;
1878 	if (loops == 1)
1879 		var *= 2;
1880 
1881 	return ((srttb_diff < 0) && (sqr_diff > var));
1882 }
1883 
1884 /**
1885  * gfs2_rgrp_used_recently
1886  * @rs: The block reservation with the rgrp to test
1887  * @msecs: The time limit in milliseconds
1888  *
1889  * Returns: True if the rgrp glock has been used within the time limit
1890  */
1891 static bool gfs2_rgrp_used_recently(const struct gfs2_blkreserv *rs,
1892 				    u64 msecs)
1893 {
1894 	u64 tdiff;
1895 
1896 	tdiff = ktime_to_ns(ktime_sub(ktime_get_real(),
1897                             rs->rs_rbm.rgd->rd_gl->gl_dstamp));
1898 
1899 	return tdiff > (msecs * 1000 * 1000);
1900 }
1901 
1902 static u32 gfs2_orlov_skip(const struct gfs2_inode *ip)
1903 {
1904 	const struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode);
1905 	u32 skip;
1906 
1907 	get_random_bytes(&skip, sizeof(skip));
1908 	return skip % sdp->sd_rgrps;
1909 }
1910 
1911 static bool gfs2_select_rgrp(struct gfs2_rgrpd **pos, const struct gfs2_rgrpd *begin)
1912 {
1913 	struct gfs2_rgrpd *rgd = *pos;
1914 	struct gfs2_sbd *sdp = rgd->rd_sbd;
1915 
1916 	rgd = gfs2_rgrpd_get_next(rgd);
1917 	if (rgd == NULL)
1918 		rgd = gfs2_rgrpd_get_first(sdp);
1919 	*pos = rgd;
1920 	if (rgd != begin) /* If we didn't wrap */
1921 		return true;
1922 	return false;
1923 }
1924 
1925 /**
1926  * fast_to_acquire - determine if a resource group will be fast to acquire
1927  *
1928  * If this is one of our preferred rgrps, it should be quicker to acquire,
1929  * because we tried to set ourselves up as dlm lock master.
1930  */
1931 static inline int fast_to_acquire(struct gfs2_rgrpd *rgd)
1932 {
1933 	struct gfs2_glock *gl = rgd->rd_gl;
1934 
1935 	if (gl->gl_state != LM_ST_UNLOCKED && list_empty(&gl->gl_holders) &&
1936 	    !test_bit(GLF_DEMOTE_IN_PROGRESS, &gl->gl_flags) &&
1937 	    !test_bit(GLF_DEMOTE, &gl->gl_flags))
1938 		return 1;
1939 	if (rgd->rd_flags & GFS2_RDF_PREFERRED)
1940 		return 1;
1941 	return 0;
1942 }
1943 
1944 /**
1945  * gfs2_inplace_reserve - Reserve space in the filesystem
1946  * @ip: the inode to reserve space for
1947  * @ap: the allocation parameters
1948  *
1949  * We try our best to find an rgrp that has at least ap->target blocks
1950  * available. After a couple of passes (loops == 2), the prospects of finding
1951  * such an rgrp diminish. At this stage, we return the first rgrp that has
1952  * atleast ap->min_target blocks available. Either way, we set ap->allowed to
1953  * the number of blocks available in the chosen rgrp.
1954  *
1955  * Returns: 0 on success,
1956  *          -ENOMEM if a suitable rgrp can't be found
1957  *          errno otherwise
1958  */
1959 
1960 int gfs2_inplace_reserve(struct gfs2_inode *ip, struct gfs2_alloc_parms *ap)
1961 {
1962 	struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode);
1963 	struct gfs2_rgrpd *begin = NULL;
1964 	struct gfs2_blkreserv *rs = ip->i_res;
1965 	int error = 0, rg_locked, flags = 0;
1966 	u64 last_unlinked = NO_BLOCK;
1967 	int loops = 0;
1968 	u32 skip = 0;
1969 
1970 	if (sdp->sd_args.ar_rgrplvb)
1971 		flags |= GL_SKIP;
1972 	if (gfs2_assert_warn(sdp, ap->target))
1973 		return -EINVAL;
1974 	if (gfs2_rs_active(rs)) {
1975 		begin = rs->rs_rbm.rgd;
1976 	} else if (ip->i_rgd && rgrp_contains_block(ip->i_rgd, ip->i_goal)) {
1977 		rs->rs_rbm.rgd = begin = ip->i_rgd;
1978 	} else {
1979 		check_and_update_goal(ip);
1980 		rs->rs_rbm.rgd = begin = gfs2_blk2rgrpd(sdp, ip->i_goal, 1);
1981 	}
1982 	if (S_ISDIR(ip->i_inode.i_mode) && (ap->aflags & GFS2_AF_ORLOV))
1983 		skip = gfs2_orlov_skip(ip);
1984 	if (rs->rs_rbm.rgd == NULL)
1985 		return -EBADSLT;
1986 
1987 	while (loops < 3) {
1988 		rg_locked = 1;
1989 
1990 		if (!gfs2_glock_is_locked_by_me(rs->rs_rbm.rgd->rd_gl)) {
1991 			rg_locked = 0;
1992 			if (skip && skip--)
1993 				goto next_rgrp;
1994 			if (!gfs2_rs_active(rs)) {
1995 				if (loops == 0 &&
1996 				    !fast_to_acquire(rs->rs_rbm.rgd))
1997 					goto next_rgrp;
1998 				if ((loops < 2) &&
1999 				    gfs2_rgrp_used_recently(rs, 1000) &&
2000 				    gfs2_rgrp_congested(rs->rs_rbm.rgd, loops))
2001 					goto next_rgrp;
2002 			}
2003 			error = gfs2_glock_nq_init(rs->rs_rbm.rgd->rd_gl,
2004 						   LM_ST_EXCLUSIVE, flags,
2005 						   &rs->rs_rgd_gh);
2006 			if (unlikely(error))
2007 				return error;
2008 			if (!gfs2_rs_active(rs) && (loops < 2) &&
2009 			    gfs2_rgrp_congested(rs->rs_rbm.rgd, loops))
2010 				goto skip_rgrp;
2011 			if (sdp->sd_args.ar_rgrplvb) {
2012 				error = update_rgrp_lvb(rs->rs_rbm.rgd);
2013 				if (unlikely(error)) {
2014 					gfs2_glock_dq_uninit(&rs->rs_rgd_gh);
2015 					return error;
2016 				}
2017 			}
2018 		}
2019 
2020 		/* Skip unuseable resource groups */
2021 		if ((rs->rs_rbm.rgd->rd_flags & (GFS2_RGF_NOALLOC |
2022 						 GFS2_RDF_ERROR)) ||
2023 		    (loops == 0 && ap->target > rs->rs_rbm.rgd->rd_extfail_pt))
2024 			goto skip_rgrp;
2025 
2026 		if (sdp->sd_args.ar_rgrplvb)
2027 			gfs2_rgrp_bh_get(rs->rs_rbm.rgd);
2028 
2029 		/* Get a reservation if we don't already have one */
2030 		if (!gfs2_rs_active(rs))
2031 			rg_mblk_search(rs->rs_rbm.rgd, ip, ap);
2032 
2033 		/* Skip rgrps when we can't get a reservation on first pass */
2034 		if (!gfs2_rs_active(rs) && (loops < 1))
2035 			goto check_rgrp;
2036 
2037 		/* If rgrp has enough free space, use it */
2038 		if (rs->rs_rbm.rgd->rd_free_clone >= ap->target ||
2039 		    (loops == 2 && ap->min_target &&
2040 		     rs->rs_rbm.rgd->rd_free_clone >= ap->min_target)) {
2041 			ip->i_rgd = rs->rs_rbm.rgd;
2042 			ap->allowed = ip->i_rgd->rd_free_clone;
2043 			return 0;
2044 		}
2045 check_rgrp:
2046 		/* Check for unlinked inodes which can be reclaimed */
2047 		if (rs->rs_rbm.rgd->rd_flags & GFS2_RDF_CHECK)
2048 			try_rgrp_unlink(rs->rs_rbm.rgd, &last_unlinked,
2049 					ip->i_no_addr);
2050 skip_rgrp:
2051 		/* Drop reservation, if we couldn't use reserved rgrp */
2052 		if (gfs2_rs_active(rs))
2053 			gfs2_rs_deltree(rs);
2054 
2055 		/* Unlock rgrp if required */
2056 		if (!rg_locked)
2057 			gfs2_glock_dq_uninit(&rs->rs_rgd_gh);
2058 next_rgrp:
2059 		/* Find the next rgrp, and continue looking */
2060 		if (gfs2_select_rgrp(&rs->rs_rbm.rgd, begin))
2061 			continue;
2062 		if (skip)
2063 			continue;
2064 
2065 		/* If we've scanned all the rgrps, but found no free blocks
2066 		 * then this checks for some less likely conditions before
2067 		 * trying again.
2068 		 */
2069 		loops++;
2070 		/* Check that fs hasn't grown if writing to rindex */
2071 		if (ip == GFS2_I(sdp->sd_rindex) && !sdp->sd_rindex_uptodate) {
2072 			error = gfs2_ri_update(ip);
2073 			if (error)
2074 				return error;
2075 		}
2076 		/* Flushing the log may release space */
2077 		if (loops == 2)
2078 			gfs2_log_flush(sdp, NULL, NORMAL_FLUSH);
2079 	}
2080 
2081 	return -ENOSPC;
2082 }
2083 
2084 /**
2085  * gfs2_inplace_release - release an inplace reservation
2086  * @ip: the inode the reservation was taken out on
2087  *
2088  * Release a reservation made by gfs2_inplace_reserve().
2089  */
2090 
2091 void gfs2_inplace_release(struct gfs2_inode *ip)
2092 {
2093 	struct gfs2_blkreserv *rs = ip->i_res;
2094 
2095 	if (rs->rs_rgd_gh.gh_gl)
2096 		gfs2_glock_dq_uninit(&rs->rs_rgd_gh);
2097 }
2098 
2099 /**
2100  * gfs2_get_block_type - Check a block in a RG is of given type
2101  * @rgd: the resource group holding the block
2102  * @block: the block number
2103  *
2104  * Returns: The block type (GFS2_BLKST_*)
2105  */
2106 
2107 static unsigned char gfs2_get_block_type(struct gfs2_rgrpd *rgd, u64 block)
2108 {
2109 	struct gfs2_rbm rbm = { .rgd = rgd, };
2110 	int ret;
2111 
2112 	ret = gfs2_rbm_from_block(&rbm, block);
2113 	WARN_ON_ONCE(ret != 0);
2114 
2115 	return gfs2_testbit(&rbm);
2116 }
2117 
2118 
2119 /**
2120  * gfs2_alloc_extent - allocate an extent from a given bitmap
2121  * @rbm: the resource group information
2122  * @dinode: TRUE if the first block we allocate is for a dinode
2123  * @n: The extent length (value/result)
2124  *
2125  * Add the bitmap buffer to the transaction.
2126  * Set the found bits to @new_state to change block's allocation state.
2127  */
2128 static void gfs2_alloc_extent(const struct gfs2_rbm *rbm, bool dinode,
2129 			     unsigned int *n)
2130 {
2131 	struct gfs2_rbm pos = { .rgd = rbm->rgd, };
2132 	const unsigned int elen = *n;
2133 	u64 block;
2134 	int ret;
2135 
2136 	*n = 1;
2137 	block = gfs2_rbm_to_block(rbm);
2138 	gfs2_trans_add_meta(rbm->rgd->rd_gl, rbm_bi(rbm)->bi_bh);
2139 	gfs2_setbit(rbm, true, dinode ? GFS2_BLKST_DINODE : GFS2_BLKST_USED);
2140 	block++;
2141 	while (*n < elen) {
2142 		ret = gfs2_rbm_from_block(&pos, block);
2143 		if (ret || gfs2_testbit(&pos) != GFS2_BLKST_FREE)
2144 			break;
2145 		gfs2_trans_add_meta(pos.rgd->rd_gl, rbm_bi(&pos)->bi_bh);
2146 		gfs2_setbit(&pos, true, GFS2_BLKST_USED);
2147 		(*n)++;
2148 		block++;
2149 	}
2150 }
2151 
2152 /**
2153  * rgblk_free - Change alloc state of given block(s)
2154  * @sdp: the filesystem
2155  * @bstart: the start of a run of blocks to free
2156  * @blen: the length of the block run (all must lie within ONE RG!)
2157  * @new_state: GFS2_BLKST_XXX the after-allocation block state
2158  *
2159  * Returns:  Resource group containing the block(s)
2160  */
2161 
2162 static struct gfs2_rgrpd *rgblk_free(struct gfs2_sbd *sdp, u64 bstart,
2163 				     u32 blen, unsigned char new_state)
2164 {
2165 	struct gfs2_rbm rbm;
2166 	struct gfs2_bitmap *bi, *bi_prev = NULL;
2167 
2168 	rbm.rgd = gfs2_blk2rgrpd(sdp, bstart, 1);
2169 	if (!rbm.rgd) {
2170 		if (gfs2_consist(sdp))
2171 			fs_err(sdp, "block = %llu\n", (unsigned long long)bstart);
2172 		return NULL;
2173 	}
2174 
2175 	gfs2_rbm_from_block(&rbm, bstart);
2176 	while (blen--) {
2177 		bi = rbm_bi(&rbm);
2178 		if (bi != bi_prev) {
2179 			if (!bi->bi_clone) {
2180 				bi->bi_clone = kmalloc(bi->bi_bh->b_size,
2181 						      GFP_NOFS | __GFP_NOFAIL);
2182 				memcpy(bi->bi_clone + bi->bi_offset,
2183 				       bi->bi_bh->b_data + bi->bi_offset,
2184 				       bi->bi_len);
2185 			}
2186 			gfs2_trans_add_meta(rbm.rgd->rd_gl, bi->bi_bh);
2187 			bi_prev = bi;
2188 		}
2189 		gfs2_setbit(&rbm, false, new_state);
2190 		gfs2_rbm_incr(&rbm);
2191 	}
2192 
2193 	return rbm.rgd;
2194 }
2195 
2196 /**
2197  * gfs2_rgrp_dump - print out an rgrp
2198  * @seq: The iterator
2199  * @gl: The glock in question
2200  *
2201  */
2202 
2203 void gfs2_rgrp_dump(struct seq_file *seq, const struct gfs2_glock *gl)
2204 {
2205 	struct gfs2_rgrpd *rgd = gl->gl_object;
2206 	struct gfs2_blkreserv *trs;
2207 	const struct rb_node *n;
2208 
2209 	if (rgd == NULL)
2210 		return;
2211 	gfs2_print_dbg(seq, " R: n:%llu f:%02x b:%u/%u i:%u r:%u e:%u\n",
2212 		       (unsigned long long)rgd->rd_addr, rgd->rd_flags,
2213 		       rgd->rd_free, rgd->rd_free_clone, rgd->rd_dinodes,
2214 		       rgd->rd_reserved, rgd->rd_extfail_pt);
2215 	spin_lock(&rgd->rd_rsspin);
2216 	for (n = rb_first(&rgd->rd_rstree); n; n = rb_next(&trs->rs_node)) {
2217 		trs = rb_entry(n, struct gfs2_blkreserv, rs_node);
2218 		dump_rs(seq, trs);
2219 	}
2220 	spin_unlock(&rgd->rd_rsspin);
2221 }
2222 
2223 static void gfs2_rgrp_error(struct gfs2_rgrpd *rgd)
2224 {
2225 	struct gfs2_sbd *sdp = rgd->rd_sbd;
2226 	fs_warn(sdp, "rgrp %llu has an error, marking it readonly until umount\n",
2227 		(unsigned long long)rgd->rd_addr);
2228 	fs_warn(sdp, "umount on all nodes and run fsck.gfs2 to fix the error\n");
2229 	gfs2_rgrp_dump(NULL, rgd->rd_gl);
2230 	rgd->rd_flags |= GFS2_RDF_ERROR;
2231 }
2232 
2233 /**
2234  * gfs2_adjust_reservation - Adjust (or remove) a reservation after allocation
2235  * @ip: The inode we have just allocated blocks for
2236  * @rbm: The start of the allocated blocks
2237  * @len: The extent length
2238  *
2239  * Adjusts a reservation after an allocation has taken place. If the
2240  * reservation does not match the allocation, or if it is now empty
2241  * then it is removed.
2242  */
2243 
2244 static void gfs2_adjust_reservation(struct gfs2_inode *ip,
2245 				    const struct gfs2_rbm *rbm, unsigned len)
2246 {
2247 	struct gfs2_blkreserv *rs = ip->i_res;
2248 	struct gfs2_rgrpd *rgd = rbm->rgd;
2249 	unsigned rlen;
2250 	u64 block;
2251 	int ret;
2252 
2253 	spin_lock(&rgd->rd_rsspin);
2254 	if (gfs2_rs_active(rs)) {
2255 		if (gfs2_rbm_eq(&rs->rs_rbm, rbm)) {
2256 			block = gfs2_rbm_to_block(rbm);
2257 			ret = gfs2_rbm_from_block(&rs->rs_rbm, block + len);
2258 			rlen = min(rs->rs_free, len);
2259 			rs->rs_free -= rlen;
2260 			rgd->rd_reserved -= rlen;
2261 			trace_gfs2_rs(rs, TRACE_RS_CLAIM);
2262 			if (rs->rs_free && !ret)
2263 				goto out;
2264 			/* We used up our block reservation, so we should
2265 			   reserve more blocks next time. */
2266 			atomic_add(RGRP_RSRV_ADDBLKS, &rs->rs_sizehint);
2267 		}
2268 		__rs_deltree(rs);
2269 	}
2270 out:
2271 	spin_unlock(&rgd->rd_rsspin);
2272 }
2273 
2274 /**
2275  * gfs2_set_alloc_start - Set starting point for block allocation
2276  * @rbm: The rbm which will be set to the required location
2277  * @ip: The gfs2 inode
2278  * @dinode: Flag to say if allocation includes a new inode
2279  *
2280  * This sets the starting point from the reservation if one is active
2281  * otherwise it falls back to guessing a start point based on the
2282  * inode's goal block or the last allocation point in the rgrp.
2283  */
2284 
2285 static void gfs2_set_alloc_start(struct gfs2_rbm *rbm,
2286 				 const struct gfs2_inode *ip, bool dinode)
2287 {
2288 	u64 goal;
2289 
2290 	if (gfs2_rs_active(ip->i_res)) {
2291 		*rbm = ip->i_res->rs_rbm;
2292 		return;
2293 	}
2294 
2295 	if (!dinode && rgrp_contains_block(rbm->rgd, ip->i_goal))
2296 		goal = ip->i_goal;
2297 	else
2298 		goal = rbm->rgd->rd_last_alloc + rbm->rgd->rd_data0;
2299 
2300 	gfs2_rbm_from_block(rbm, goal);
2301 }
2302 
2303 /**
2304  * gfs2_alloc_blocks - Allocate one or more blocks of data and/or a dinode
2305  * @ip: the inode to allocate the block for
2306  * @bn: Used to return the starting block number
2307  * @nblocks: requested number of blocks/extent length (value/result)
2308  * @dinode: 1 if we're allocating a dinode block, else 0
2309  * @generation: the generation number of the inode
2310  *
2311  * Returns: 0 or error
2312  */
2313 
2314 int gfs2_alloc_blocks(struct gfs2_inode *ip, u64 *bn, unsigned int *nblocks,
2315 		      bool dinode, u64 *generation)
2316 {
2317 	struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode);
2318 	struct buffer_head *dibh;
2319 	struct gfs2_rbm rbm = { .rgd = ip->i_rgd, };
2320 	unsigned int ndata;
2321 	u64 block; /* block, within the file system scope */
2322 	int error;
2323 
2324 	gfs2_set_alloc_start(&rbm, ip, dinode);
2325 	error = gfs2_rbm_find(&rbm, GFS2_BLKST_FREE, NULL, ip, false, NULL);
2326 
2327 	if (error == -ENOSPC) {
2328 		gfs2_set_alloc_start(&rbm, ip, dinode);
2329 		error = gfs2_rbm_find(&rbm, GFS2_BLKST_FREE, NULL, NULL, false,
2330 				      NULL);
2331 	}
2332 
2333 	/* Since all blocks are reserved in advance, this shouldn't happen */
2334 	if (error) {
2335 		fs_warn(sdp, "inum=%llu error=%d, nblocks=%u, full=%d fail_pt=%d\n",
2336 			(unsigned long long)ip->i_no_addr, error, *nblocks,
2337 			test_bit(GBF_FULL, &rbm.rgd->rd_bits->bi_flags),
2338 			rbm.rgd->rd_extfail_pt);
2339 		goto rgrp_error;
2340 	}
2341 
2342 	gfs2_alloc_extent(&rbm, dinode, nblocks);
2343 	block = gfs2_rbm_to_block(&rbm);
2344 	rbm.rgd->rd_last_alloc = block - rbm.rgd->rd_data0;
2345 	if (gfs2_rs_active(ip->i_res))
2346 		gfs2_adjust_reservation(ip, &rbm, *nblocks);
2347 	ndata = *nblocks;
2348 	if (dinode)
2349 		ndata--;
2350 
2351 	if (!dinode) {
2352 		ip->i_goal = block + ndata - 1;
2353 		error = gfs2_meta_inode_buffer(ip, &dibh);
2354 		if (error == 0) {
2355 			struct gfs2_dinode *di =
2356 				(struct gfs2_dinode *)dibh->b_data;
2357 			gfs2_trans_add_meta(ip->i_gl, dibh);
2358 			di->di_goal_meta = di->di_goal_data =
2359 				cpu_to_be64(ip->i_goal);
2360 			brelse(dibh);
2361 		}
2362 	}
2363 	if (rbm.rgd->rd_free < *nblocks) {
2364 		pr_warn("nblocks=%u\n", *nblocks);
2365 		goto rgrp_error;
2366 	}
2367 
2368 	rbm.rgd->rd_free -= *nblocks;
2369 	if (dinode) {
2370 		rbm.rgd->rd_dinodes++;
2371 		*generation = rbm.rgd->rd_igeneration++;
2372 		if (*generation == 0)
2373 			*generation = rbm.rgd->rd_igeneration++;
2374 	}
2375 
2376 	gfs2_trans_add_meta(rbm.rgd->rd_gl, rbm.rgd->rd_bits[0].bi_bh);
2377 	gfs2_rgrp_out(rbm.rgd, rbm.rgd->rd_bits[0].bi_bh->b_data);
2378 	gfs2_rgrp_ondisk2lvb(rbm.rgd->rd_rgl, rbm.rgd->rd_bits[0].bi_bh->b_data);
2379 
2380 	gfs2_statfs_change(sdp, 0, -(s64)*nblocks, dinode ? 1 : 0);
2381 	if (dinode)
2382 		gfs2_trans_add_unrevoke(sdp, block, *nblocks);
2383 
2384 	gfs2_quota_change(ip, *nblocks, ip->i_inode.i_uid, ip->i_inode.i_gid);
2385 
2386 	rbm.rgd->rd_free_clone -= *nblocks;
2387 	trace_gfs2_block_alloc(ip, rbm.rgd, block, *nblocks,
2388 			       dinode ? GFS2_BLKST_DINODE : GFS2_BLKST_USED);
2389 	*bn = block;
2390 	return 0;
2391 
2392 rgrp_error:
2393 	gfs2_rgrp_error(rbm.rgd);
2394 	return -EIO;
2395 }
2396 
2397 /**
2398  * __gfs2_free_blocks - free a contiguous run of block(s)
2399  * @ip: the inode these blocks are being freed from
2400  * @bstart: first block of a run of contiguous blocks
2401  * @blen: the length of the block run
2402  * @meta: 1 if the blocks represent metadata
2403  *
2404  */
2405 
2406 void __gfs2_free_blocks(struct gfs2_inode *ip, u64 bstart, u32 blen, int meta)
2407 {
2408 	struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode);
2409 	struct gfs2_rgrpd *rgd;
2410 
2411 	rgd = rgblk_free(sdp, bstart, blen, GFS2_BLKST_FREE);
2412 	if (!rgd)
2413 		return;
2414 	trace_gfs2_block_alloc(ip, rgd, bstart, blen, GFS2_BLKST_FREE);
2415 	rgd->rd_free += blen;
2416 	rgd->rd_flags &= ~GFS2_RGF_TRIMMED;
2417 	gfs2_trans_add_meta(rgd->rd_gl, rgd->rd_bits[0].bi_bh);
2418 	gfs2_rgrp_out(rgd, rgd->rd_bits[0].bi_bh->b_data);
2419 	gfs2_rgrp_ondisk2lvb(rgd->rd_rgl, rgd->rd_bits[0].bi_bh->b_data);
2420 
2421 	/* Directories keep their data in the metadata address space */
2422 	if (meta || ip->i_depth)
2423 		gfs2_meta_wipe(ip, bstart, blen);
2424 }
2425 
2426 /**
2427  * gfs2_free_meta - free a contiguous run of data block(s)
2428  * @ip: the inode these blocks are being freed from
2429  * @bstart: first block of a run of contiguous blocks
2430  * @blen: the length of the block run
2431  *
2432  */
2433 
2434 void gfs2_free_meta(struct gfs2_inode *ip, u64 bstart, u32 blen)
2435 {
2436 	struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode);
2437 
2438 	__gfs2_free_blocks(ip, bstart, blen, 1);
2439 	gfs2_statfs_change(sdp, 0, +blen, 0);
2440 	gfs2_quota_change(ip, -(s64)blen, ip->i_inode.i_uid, ip->i_inode.i_gid);
2441 }
2442 
2443 void gfs2_unlink_di(struct inode *inode)
2444 {
2445 	struct gfs2_inode *ip = GFS2_I(inode);
2446 	struct gfs2_sbd *sdp = GFS2_SB(inode);
2447 	struct gfs2_rgrpd *rgd;
2448 	u64 blkno = ip->i_no_addr;
2449 
2450 	rgd = rgblk_free(sdp, blkno, 1, GFS2_BLKST_UNLINKED);
2451 	if (!rgd)
2452 		return;
2453 	trace_gfs2_block_alloc(ip, rgd, blkno, 1, GFS2_BLKST_UNLINKED);
2454 	gfs2_trans_add_meta(rgd->rd_gl, rgd->rd_bits[0].bi_bh);
2455 	gfs2_rgrp_out(rgd, rgd->rd_bits[0].bi_bh->b_data);
2456 	gfs2_rgrp_ondisk2lvb(rgd->rd_rgl, rgd->rd_bits[0].bi_bh->b_data);
2457 	update_rgrp_lvb_unlinked(rgd, 1);
2458 }
2459 
2460 static void gfs2_free_uninit_di(struct gfs2_rgrpd *rgd, u64 blkno)
2461 {
2462 	struct gfs2_sbd *sdp = rgd->rd_sbd;
2463 	struct gfs2_rgrpd *tmp_rgd;
2464 
2465 	tmp_rgd = rgblk_free(sdp, blkno, 1, GFS2_BLKST_FREE);
2466 	if (!tmp_rgd)
2467 		return;
2468 	gfs2_assert_withdraw(sdp, rgd == tmp_rgd);
2469 
2470 	if (!rgd->rd_dinodes)
2471 		gfs2_consist_rgrpd(rgd);
2472 	rgd->rd_dinodes--;
2473 	rgd->rd_free++;
2474 
2475 	gfs2_trans_add_meta(rgd->rd_gl, rgd->rd_bits[0].bi_bh);
2476 	gfs2_rgrp_out(rgd, rgd->rd_bits[0].bi_bh->b_data);
2477 	gfs2_rgrp_ondisk2lvb(rgd->rd_rgl, rgd->rd_bits[0].bi_bh->b_data);
2478 	update_rgrp_lvb_unlinked(rgd, -1);
2479 
2480 	gfs2_statfs_change(sdp, 0, +1, -1);
2481 }
2482 
2483 
2484 void gfs2_free_di(struct gfs2_rgrpd *rgd, struct gfs2_inode *ip)
2485 {
2486 	gfs2_free_uninit_di(rgd, ip->i_no_addr);
2487 	trace_gfs2_block_alloc(ip, rgd, ip->i_no_addr, 1, GFS2_BLKST_FREE);
2488 	gfs2_quota_change(ip, -1, ip->i_inode.i_uid, ip->i_inode.i_gid);
2489 	gfs2_meta_wipe(ip, ip->i_no_addr, 1);
2490 }
2491 
2492 /**
2493  * gfs2_check_blk_type - Check the type of a block
2494  * @sdp: The superblock
2495  * @no_addr: The block number to check
2496  * @type: The block type we are looking for
2497  *
2498  * Returns: 0 if the block type matches the expected type
2499  *          -ESTALE if it doesn't match
2500  *          or -ve errno if something went wrong while checking
2501  */
2502 
2503 int gfs2_check_blk_type(struct gfs2_sbd *sdp, u64 no_addr, unsigned int type)
2504 {
2505 	struct gfs2_rgrpd *rgd;
2506 	struct gfs2_holder rgd_gh;
2507 	int error = -EINVAL;
2508 
2509 	rgd = gfs2_blk2rgrpd(sdp, no_addr, 1);
2510 	if (!rgd)
2511 		goto fail;
2512 
2513 	error = gfs2_glock_nq_init(rgd->rd_gl, LM_ST_SHARED, 0, &rgd_gh);
2514 	if (error)
2515 		goto fail;
2516 
2517 	if (gfs2_get_block_type(rgd, no_addr) != type)
2518 		error = -ESTALE;
2519 
2520 	gfs2_glock_dq_uninit(&rgd_gh);
2521 fail:
2522 	return error;
2523 }
2524 
2525 /**
2526  * gfs2_rlist_add - add a RG to a list of RGs
2527  * @ip: the inode
2528  * @rlist: the list of resource groups
2529  * @block: the block
2530  *
2531  * Figure out what RG a block belongs to and add that RG to the list
2532  *
2533  * FIXME: Don't use NOFAIL
2534  *
2535  */
2536 
2537 void gfs2_rlist_add(struct gfs2_inode *ip, struct gfs2_rgrp_list *rlist,
2538 		    u64 block)
2539 {
2540 	struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode);
2541 	struct gfs2_rgrpd *rgd;
2542 	struct gfs2_rgrpd **tmp;
2543 	unsigned int new_space;
2544 	unsigned int x;
2545 
2546 	if (gfs2_assert_warn(sdp, !rlist->rl_ghs))
2547 		return;
2548 
2549 	if (ip->i_rgd && rgrp_contains_block(ip->i_rgd, block))
2550 		rgd = ip->i_rgd;
2551 	else
2552 		rgd = gfs2_blk2rgrpd(sdp, block, 1);
2553 	if (!rgd) {
2554 		fs_err(sdp, "rlist_add: no rgrp for block %llu\n", (unsigned long long)block);
2555 		return;
2556 	}
2557 	ip->i_rgd = rgd;
2558 
2559 	for (x = 0; x < rlist->rl_rgrps; x++)
2560 		if (rlist->rl_rgd[x] == rgd)
2561 			return;
2562 
2563 	if (rlist->rl_rgrps == rlist->rl_space) {
2564 		new_space = rlist->rl_space + 10;
2565 
2566 		tmp = kcalloc(new_space, sizeof(struct gfs2_rgrpd *),
2567 			      GFP_NOFS | __GFP_NOFAIL);
2568 
2569 		if (rlist->rl_rgd) {
2570 			memcpy(tmp, rlist->rl_rgd,
2571 			       rlist->rl_space * sizeof(struct gfs2_rgrpd *));
2572 			kfree(rlist->rl_rgd);
2573 		}
2574 
2575 		rlist->rl_space = new_space;
2576 		rlist->rl_rgd = tmp;
2577 	}
2578 
2579 	rlist->rl_rgd[rlist->rl_rgrps++] = rgd;
2580 }
2581 
2582 /**
2583  * gfs2_rlist_alloc - all RGs have been added to the rlist, now allocate
2584  *      and initialize an array of glock holders for them
2585  * @rlist: the list of resource groups
2586  * @state: the lock state to acquire the RG lock in
2587  *
2588  * FIXME: Don't use NOFAIL
2589  *
2590  */
2591 
2592 void gfs2_rlist_alloc(struct gfs2_rgrp_list *rlist, unsigned int state)
2593 {
2594 	unsigned int x;
2595 
2596 	rlist->rl_ghs = kcalloc(rlist->rl_rgrps, sizeof(struct gfs2_holder),
2597 				GFP_NOFS | __GFP_NOFAIL);
2598 	for (x = 0; x < rlist->rl_rgrps; x++)
2599 		gfs2_holder_init(rlist->rl_rgd[x]->rd_gl,
2600 				state, 0,
2601 				&rlist->rl_ghs[x]);
2602 }
2603 
2604 /**
2605  * gfs2_rlist_free - free a resource group list
2606  * @rlist: the list of resource groups
2607  *
2608  */
2609 
2610 void gfs2_rlist_free(struct gfs2_rgrp_list *rlist)
2611 {
2612 	unsigned int x;
2613 
2614 	kfree(rlist->rl_rgd);
2615 
2616 	if (rlist->rl_ghs) {
2617 		for (x = 0; x < rlist->rl_rgrps; x++)
2618 			gfs2_holder_uninit(&rlist->rl_ghs[x]);
2619 		kfree(rlist->rl_ghs);
2620 		rlist->rl_ghs = NULL;
2621 	}
2622 }
2623 
2624