]>
Commit | Line | Data |
---|---|---|
7336d0e6 | 1 | // SPDX-License-Identifier: GPL-2.0-only |
b3b94faa DT |
2 | /* |
3 | * Copyright (C) Sistina Software, Inc. 1997-2003 All rights reserved. | |
fe6c991c | 4 | * Copyright (C) 2004-2008 Red Hat, Inc. All rights reserved. |
b3b94faa DT |
5 | */ |
6 | ||
d77d1b58 JP |
7 | #define pr_fmt(fmt) KBUILD_MODNAME ": " fmt |
8 | ||
b3b94faa DT |
9 | #include <linux/slab.h> |
10 | #include <linux/spinlock.h> | |
11 | #include <linux/completion.h> | |
12 | #include <linux/buffer_head.h> | |
f42faf4f | 13 | #include <linux/fs.h> |
5c676f6d | 14 | #include <linux/gfs2_ondisk.h> |
1f466a47 | 15 | #include <linux/prefetch.h> |
f15ab561 | 16 | #include <linux/blkdev.h> |
7c9ca621 | 17 | #include <linux/rbtree.h> |
9dbe9610 | 18 | #include <linux/random.h> |
b3b94faa DT |
19 | |
20 | #include "gfs2.h" | |
5c676f6d | 21 | #include "incore.h" |
b3b94faa DT |
22 | #include "glock.h" |
23 | #include "glops.h" | |
b3b94faa DT |
24 | #include "lops.h" |
25 | #include "meta_io.h" | |
26 | #include "quota.h" | |
27 | #include "rgrp.h" | |
28 | #include "super.h" | |
29 | #include "trans.h" | |
5c676f6d | 30 | #include "util.h" |
172e045a | 31 | #include "log.h" |
c8cdf479 | 32 | #include "inode.h" |
63997775 | 33 | #include "trace_gfs2.h" |
850d2d91 | 34 | #include "dir.h" |
b3b94faa | 35 | |
2c1e52aa | 36 | #define BFITNOENT ((u32)~0) |
6760bdcd | 37 | #define NO_BLOCK ((u64)~0) |
88c8ab1f SW |
38 | |
39 | /* | |
40 | * These routines are used by the resource group routines (rgrp.c) | |
41 | * to keep track of block allocation. Each block is represented by two | |
feaa7bba SW |
42 | * bits. So, each byte represents GFS2_NBBY (i.e. 4) blocks. |
43 | * | |
44 | * 0 = Free | |
45 | * 1 = Used (not metadata) | |
46 | * 2 = Unlinked (still in use) inode | |
47 | * 3 = Used (metadata) | |
88c8ab1f SW |
48 | */ |
49 | ||
5ce13431 BP |
50 | struct gfs2_extent { |
51 | struct gfs2_rbm rbm; | |
52 | u32 len; | |
53 | }; | |
54 | ||
88c8ab1f SW |
55 | static const char valid_change[16] = { |
56 | /* current */ | |
feaa7bba | 57 | /* n */ 0, 1, 1, 1, |
88c8ab1f | 58 | /* e */ 1, 0, 0, 0, |
feaa7bba | 59 | /* w */ 0, 0, 0, 1, |
88c8ab1f SW |
60 | 1, 0, 0, 0 |
61 | }; | |
62 | ||
5ce13431 | 63 | static int gfs2_rbm_find(struct gfs2_rbm *rbm, u8 state, u32 *minext, |
8381e602 | 64 | const struct gfs2_inode *ip, bool nowrap); |
ff7f4cb4 SW |
65 | |
66 | ||
88c8ab1f SW |
67 | /** |
68 | * gfs2_setbit - Set a bit in the bitmaps | |
3e6339dd SW |
69 | * @rbm: The position of the bit to set |
70 | * @do_clone: Also set the clone bitmap, if it exists | |
88c8ab1f SW |
71 | * @new_state: the new state of the block |
72 | * | |
73 | */ | |
74 | ||
3e6339dd | 75 | static inline void gfs2_setbit(const struct gfs2_rbm *rbm, bool do_clone, |
06344b91 | 76 | unsigned char new_state) |
88c8ab1f | 77 | { |
b45e41d7 | 78 | unsigned char *byte1, *byte2, *end, cur_state; |
e579ed4f | 79 | struct gfs2_bitmap *bi = rbm_bi(rbm); |
281b4952 | 80 | unsigned int buflen = bi->bi_bytes; |
3e6339dd | 81 | const unsigned int bit = (rbm->offset % GFS2_NBBY) * GFS2_BIT_SIZE; |
88c8ab1f | 82 | |
e579ed4f BP |
83 | byte1 = bi->bi_bh->b_data + bi->bi_offset + (rbm->offset / GFS2_NBBY); |
84 | end = bi->bi_bh->b_data + bi->bi_offset + buflen; | |
88c8ab1f | 85 | |
b45e41d7 | 86 | BUG_ON(byte1 >= end); |
88c8ab1f | 87 | |
b45e41d7 | 88 | cur_state = (*byte1 >> bit) & GFS2_BIT_MASK; |
88c8ab1f | 89 | |
b45e41d7 | 90 | if (unlikely(!valid_change[new_state * 4 + cur_state])) { |
e54c78a2 BP |
91 | struct gfs2_sbd *sdp = rbm->rgd->rd_sbd; |
92 | ||
93 | fs_warn(sdp, "buf_blk = 0x%x old_state=%d, new_state=%d\n", | |
d77d1b58 | 94 | rbm->offset, cur_state, new_state); |
e54c78a2 BP |
95 | fs_warn(sdp, "rgrp=0x%llx bi_start=0x%x biblk: 0x%llx\n", |
96 | (unsigned long long)rbm->rgd->rd_addr, bi->bi_start, | |
97 | (unsigned long long)bi->bi_bh->b_blocknr); | |
281b4952 AG |
98 | fs_warn(sdp, "bi_offset=0x%x bi_bytes=0x%x block=0x%llx\n", |
99 | bi->bi_offset, bi->bi_bytes, | |
e54c78a2 | 100 | (unsigned long long)gfs2_rbm_to_block(rbm)); |
95c8e17f | 101 | dump_stack(); |
3e6339dd | 102 | gfs2_consist_rgrpd(rbm->rgd); |
b45e41d7 SW |
103 | return; |
104 | } | |
105 | *byte1 ^= (cur_state ^ new_state) << bit; | |
106 | ||
e579ed4f BP |
107 | if (do_clone && bi->bi_clone) { |
108 | byte2 = bi->bi_clone + bi->bi_offset + (rbm->offset / GFS2_NBBY); | |
b45e41d7 SW |
109 | cur_state = (*byte2 >> bit) & GFS2_BIT_MASK; |
110 | *byte2 ^= (cur_state ^ new_state) << bit; | |
111 | } | |
88c8ab1f SW |
112 | } |
113 | ||
114 | /** | |
115 | * gfs2_testbit - test a bit in the bitmaps | |
c04a2ef3 | 116 | * @rbm: The bit to test |
dffe12a8 BP |
117 | * @use_clone: If true, test the clone bitmap, not the official bitmap. |
118 | * | |
119 | * Some callers like gfs2_unaligned_extlen need to test the clone bitmaps, | |
120 | * not the "real" bitmaps, to avoid allocating recently freed blocks. | |
88c8ab1f | 121 | * |
c04a2ef3 | 122 | * Returns: The two bit block state of the requested bit |
88c8ab1f SW |
123 | */ |
124 | ||
dffe12a8 | 125 | static inline u8 gfs2_testbit(const struct gfs2_rbm *rbm, bool use_clone) |
88c8ab1f | 126 | { |
e579ed4f | 127 | struct gfs2_bitmap *bi = rbm_bi(rbm); |
dffe12a8 | 128 | const u8 *buffer; |
c04a2ef3 | 129 | const u8 *byte; |
88c8ab1f SW |
130 | unsigned int bit; |
131 | ||
dffe12a8 BP |
132 | if (use_clone && bi->bi_clone) |
133 | buffer = bi->bi_clone; | |
134 | else | |
135 | buffer = bi->bi_bh->b_data; | |
136 | buffer += bi->bi_offset; | |
c04a2ef3 SW |
137 | byte = buffer + (rbm->offset / GFS2_NBBY); |
138 | bit = (rbm->offset % GFS2_NBBY) * GFS2_BIT_SIZE; | |
88c8ab1f | 139 | |
c04a2ef3 | 140 | return (*byte >> bit) & GFS2_BIT_MASK; |
88c8ab1f SW |
141 | } |
142 | ||
223b2b88 SW |
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[] = { | |
075ac448 HE |
165 | [0] = 0xffffffffffffffffULL, |
166 | [1] = 0xaaaaaaaaaaaaaaaaULL, | |
167 | [2] = 0x5555555555555555ULL, | |
168 | [3] = 0x0000000000000000ULL, | |
223b2b88 SW |
169 | }; |
170 | tmp = le64_to_cpu(*ptr) ^ search[state]; | |
171 | tmp &= (tmp >> 1); | |
172 | tmp &= mask; | |
173 | return tmp; | |
174 | } | |
175 | ||
8e2e0047 BP |
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 | { | |
4a993fb1 | 188 | u64 startblk = gfs2_rbm_to_block(&rs->rs_rbm); |
8e2e0047 BP |
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 | ||
88c8ab1f SW |
197 | /** |
198 | * gfs2_bitfit - Search an rgrp's bitmap buffer to find a bit-pair representing | |
199 | * a block in a given allocation state. | |
886b1416 | 200 | * @buf: the buffer that holds the bitmaps |
223b2b88 | 201 | * @len: the length (in bytes) of the buffer |
88c8ab1f | 202 | * @goal: start search at this block's bit-pair (within @buffer) |
223b2b88 | 203 | * @state: GFS2_BLKST_XXX the state of the block we're looking for. |
88c8ab1f SW |
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 | |
223b2b88 SW |
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 | |
fd589a8f | 212 | * always ok to read a complete multiple of 64 bits at the end |
223b2b88 | 213 | * of the block in case the end is no aligned to a natural boundary. |
88c8ab1f SW |
214 | * |
215 | * Return: the block number (bitmap buffer scope) that was found | |
216 | */ | |
217 | ||
02ab1721 HE |
218 | static u32 gfs2_bitfit(const u8 *buf, const unsigned int len, |
219 | u32 goal, u8 state) | |
88c8ab1f | 220 | { |
223b2b88 SW |
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; | |
075ac448 | 225 | u64 mask = 0x5555555555555555ULL; |
223b2b88 SW |
226 | u32 bit; |
227 | ||
223b2b88 SW |
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) { | |
075ac448 | 233 | tmp = gfs2_bit_search(ptr, 0x5555555555555555ULL, state); |
223b2b88 | 234 | ptr++; |
1f466a47 | 235 | } |
223b2b88 SW |
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--; | |
d8bd504a | 243 | bit = __ffs64(tmp); |
223b2b88 SW |
244 | bit /= 2; /* two bits per entry in the bitmap */ |
245 | return (((const unsigned char *)ptr - buf) * GFS2_NBBY) + bit; | |
88c8ab1f SW |
246 | } |
247 | ||
ff7f4cb4 SW |
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 | { | |
3548fce1 | 263 | if (!rgrp_contains_block(rbm->rgd, block)) |
ff7f4cb4 | 264 | return -E2BIG; |
e579ed4f | 265 | rbm->bii = 0; |
3548fce1 | 266 | rbm->offset = block - rbm->rgd->rd_data0; |
a68a0a35 | 267 | /* Check if the block is within the first block */ |
e579ed4f | 268 | if (rbm->offset < rbm_bi(rbm)->bi_blocks) |
a68a0a35 BP |
269 | return 0; |
270 | ||
271 | /* Adjust for the size diff between gfs2_meta_header and gfs2_rgrp */ | |
272 | rbm->offset += (sizeof(struct gfs2_rgrp) - | |
273 | sizeof(struct gfs2_meta_header)) * GFS2_NBBY; | |
e579ed4f BP |
274 | rbm->bii = rbm->offset / rbm->rgd->rd_sbd->sd_blocks_per_bitmap; |
275 | rbm->offset -= rbm->bii * rbm->rgd->rd_sbd->sd_blocks_per_bitmap; | |
ff7f4cb4 SW |
276 | return 0; |
277 | } | |
278 | ||
149ed7f5 BP |
279 | /** |
280 | * gfs2_rbm_incr - increment an rbm structure | |
281 | * @rbm: The rbm with rgd already set correctly | |
282 | * | |
283 | * This function takes an existing rbm structure and increments it to the next | |
284 | * viable block offset. | |
285 | * | |
286 | * Returns: If incrementing the offset would cause the rbm to go past the | |
287 | * end of the rgrp, true is returned, otherwise false. | |
288 | * | |
289 | */ | |
290 | ||
291 | static bool gfs2_rbm_incr(struct gfs2_rbm *rbm) | |
292 | { | |
293 | if (rbm->offset + 1 < rbm_bi(rbm)->bi_blocks) { /* in the same bitmap */ | |
294 | rbm->offset++; | |
295 | return false; | |
296 | } | |
297 | if (rbm->bii == rbm->rgd->rd_length - 1) /* at the last bitmap */ | |
298 | return true; | |
299 | ||
300 | rbm->offset = 0; | |
301 | rbm->bii++; | |
302 | return false; | |
303 | } | |
304 | ||
ff7f4cb4 SW |
305 | /** |
306 | * gfs2_unaligned_extlen - Look for free blocks which are not byte aligned | |
307 | * @rbm: Position to search (value/result) | |
308 | * @n_unaligned: Number of unaligned blocks to check | |
309 | * @len: Decremented for each block found (terminate on zero) | |
310 | * | |
311 | * Returns: true if a non-free block is encountered | |
312 | */ | |
313 | ||
314 | static bool gfs2_unaligned_extlen(struct gfs2_rbm *rbm, u32 n_unaligned, u32 *len) | |
315 | { | |
ff7f4cb4 SW |
316 | u32 n; |
317 | u8 res; | |
318 | ||
319 | for (n = 0; n < n_unaligned; n++) { | |
dffe12a8 | 320 | res = gfs2_testbit(rbm, true); |
ff7f4cb4 SW |
321 | if (res != GFS2_BLKST_FREE) |
322 | return true; | |
323 | (*len)--; | |
324 | if (*len == 0) | |
325 | return true; | |
149ed7f5 | 326 | if (gfs2_rbm_incr(rbm)) |
ff7f4cb4 SW |
327 | return true; |
328 | } | |
329 | ||
330 | return false; | |
331 | } | |
332 | ||
333 | /** | |
334 | * gfs2_free_extlen - Return extent length of free blocks | |
27ff6a0f | 335 | * @rrbm: Starting position |
ff7f4cb4 SW |
336 | * @len: Max length to check |
337 | * | |
338 | * Starting at the block specified by the rbm, see how many free blocks | |
339 | * there are, not reading more than len blocks ahead. This can be done | |
340 | * using memchr_inv when the blocks are byte aligned, but has to be done | |
341 | * on a block by block basis in case of unaligned blocks. Also this | |
342 | * function can cope with bitmap boundaries (although it must stop on | |
343 | * a resource group boundary) | |
344 | * | |
345 | * Returns: Number of free blocks in the extent | |
346 | */ | |
347 | ||
348 | static u32 gfs2_free_extlen(const struct gfs2_rbm *rrbm, u32 len) | |
349 | { | |
350 | struct gfs2_rbm rbm = *rrbm; | |
351 | u32 n_unaligned = rbm.offset & 3; | |
352 | u32 size = len; | |
353 | u32 bytes; | |
354 | u32 chunk_size; | |
355 | u8 *ptr, *start, *end; | |
356 | u64 block; | |
e579ed4f | 357 | struct gfs2_bitmap *bi; |
ff7f4cb4 SW |
358 | |
359 | if (n_unaligned && | |
360 | gfs2_unaligned_extlen(&rbm, 4 - n_unaligned, &len)) | |
361 | goto out; | |
362 | ||
3701530a | 363 | n_unaligned = len & 3; |
ff7f4cb4 SW |
364 | /* Start is now byte aligned */ |
365 | while (len > 3) { | |
e579ed4f BP |
366 | bi = rbm_bi(&rbm); |
367 | start = bi->bi_bh->b_data; | |
368 | if (bi->bi_clone) | |
369 | start = bi->bi_clone; | |
e579ed4f | 370 | start += bi->bi_offset; |
281b4952 | 371 | end = start + bi->bi_bytes; |
ff7f4cb4 SW |
372 | BUG_ON(rbm.offset & 3); |
373 | start += (rbm.offset / GFS2_NBBY); | |
374 | bytes = min_t(u32, len / GFS2_NBBY, (end - start)); | |
375 | ptr = memchr_inv(start, 0, bytes); | |
376 | chunk_size = ((ptr == NULL) ? bytes : (ptr - start)); | |
377 | chunk_size *= GFS2_NBBY; | |
378 | BUG_ON(len < chunk_size); | |
379 | len -= chunk_size; | |
380 | block = gfs2_rbm_to_block(&rbm); | |
15bd50ad BP |
381 | if (gfs2_rbm_from_block(&rbm, block + chunk_size)) { |
382 | n_unaligned = 0; | |
ff7f4cb4 | 383 | break; |
15bd50ad BP |
384 | } |
385 | if (ptr) { | |
386 | n_unaligned = 3; | |
387 | break; | |
388 | } | |
ff7f4cb4 SW |
389 | n_unaligned = len & 3; |
390 | } | |
391 | ||
392 | /* Deal with any bits left over at the end */ | |
393 | if (n_unaligned) | |
394 | gfs2_unaligned_extlen(&rbm, n_unaligned, &len); | |
395 | out: | |
396 | return size - len; | |
397 | } | |
398 | ||
88c8ab1f SW |
399 | /** |
400 | * gfs2_bitcount - count the number of bits in a certain state | |
886b1416 | 401 | * @rgd: the resource group descriptor |
88c8ab1f SW |
402 | * @buffer: the buffer that holds the bitmaps |
403 | * @buflen: the length (in bytes) of the buffer | |
404 | * @state: the state of the block we're looking for | |
405 | * | |
406 | * Returns: The number of bits | |
407 | */ | |
408 | ||
110acf38 SW |
409 | static u32 gfs2_bitcount(struct gfs2_rgrpd *rgd, const u8 *buffer, |
410 | unsigned int buflen, u8 state) | |
88c8ab1f | 411 | { |
110acf38 SW |
412 | const u8 *byte = buffer; |
413 | const u8 *end = buffer + buflen; | |
414 | const u8 state1 = state << 2; | |
415 | const u8 state2 = state << 4; | |
416 | const u8 state3 = state << 6; | |
cd915493 | 417 | u32 count = 0; |
88c8ab1f SW |
418 | |
419 | for (; byte < end; byte++) { | |
420 | if (((*byte) & 0x03) == state) | |
421 | count++; | |
422 | if (((*byte) & 0x0C) == state1) | |
423 | count++; | |
424 | if (((*byte) & 0x30) == state2) | |
425 | count++; | |
426 | if (((*byte) & 0xC0) == state3) | |
427 | count++; | |
428 | } | |
429 | ||
430 | return count; | |
431 | } | |
432 | ||
b3b94faa DT |
433 | /** |
434 | * gfs2_rgrp_verify - Verify that a resource group is consistent | |
b3b94faa DT |
435 | * @rgd: the rgrp |
436 | * | |
437 | */ | |
438 | ||
439 | void gfs2_rgrp_verify(struct gfs2_rgrpd *rgd) | |
440 | { | |
441 | struct gfs2_sbd *sdp = rgd->rd_sbd; | |
442 | struct gfs2_bitmap *bi = NULL; | |
bb8d8a6f | 443 | u32 length = rgd->rd_length; |
cd915493 | 444 | u32 count[4], tmp; |
b3b94faa DT |
445 | int buf, x; |
446 | ||
cd915493 | 447 | memset(count, 0, 4 * sizeof(u32)); |
b3b94faa DT |
448 | |
449 | /* Count # blocks in each of 4 possible allocation states */ | |
450 | for (buf = 0; buf < length; buf++) { | |
451 | bi = rgd->rd_bits + buf; | |
452 | for (x = 0; x < 4; x++) | |
453 | count[x] += gfs2_bitcount(rgd, | |
454 | bi->bi_bh->b_data + | |
455 | bi->bi_offset, | |
281b4952 | 456 | bi->bi_bytes, x); |
b3b94faa DT |
457 | } |
458 | ||
cfc8b549 | 459 | if (count[0] != rgd->rd_free) { |
8dc88ac6 AG |
460 | gfs2_lm(sdp, "free data mismatch: %u != %u\n", |
461 | count[0], rgd->rd_free); | |
462 | gfs2_consist_rgrpd(rgd); | |
b3b94faa DT |
463 | return; |
464 | } | |
465 | ||
73f74948 | 466 | tmp = rgd->rd_data - rgd->rd_free - rgd->rd_dinodes; |
6b946170 | 467 | if (count[1] != tmp) { |
8dc88ac6 AG |
468 | gfs2_lm(sdp, "used data mismatch: %u != %u\n", |
469 | count[1], tmp); | |
470 | gfs2_consist_rgrpd(rgd); | |
b3b94faa DT |
471 | return; |
472 | } | |
473 | ||
6b946170 | 474 | if (count[2] + count[3] != rgd->rd_dinodes) { |
8dc88ac6 AG |
475 | gfs2_lm(sdp, "used metadata mismatch: %u != %u\n", |
476 | count[2] + count[3], rgd->rd_dinodes); | |
477 | gfs2_consist_rgrpd(rgd); | |
b3b94faa DT |
478 | return; |
479 | } | |
b3b94faa DT |
480 | } |
481 | ||
b3b94faa DT |
482 | /** |
483 | * gfs2_blk2rgrpd - Find resource group for a given data/meta block number | |
484 | * @sdp: The GFS2 superblock | |
886b1416 BP |
485 | * @blk: The data block number |
486 | * @exact: True if this needs to be an exact match | |
b3b94faa | 487 | * |
90bcab99 SW |
488 | * The @exact argument should be set to true by most callers. The exception |
489 | * is when we need to match blocks which are not represented by the rgrp | |
490 | * bitmap, but which are part of the rgrp (i.e. padding blocks) which are | |
491 | * there for alignment purposes. Another way of looking at it is that @exact | |
492 | * matches only valid data/metadata blocks, but with @exact false, it will | |
493 | * match any block within the extent of the rgrp. | |
494 | * | |
b3b94faa DT |
495 | * Returns: The resource group, or NULL if not found |
496 | */ | |
497 | ||
66fc061b | 498 | struct gfs2_rgrpd *gfs2_blk2rgrpd(struct gfs2_sbd *sdp, u64 blk, bool exact) |
b3b94faa | 499 | { |
66fc061b | 500 | struct rb_node *n, *next; |
f75bbfb4 | 501 | struct gfs2_rgrpd *cur; |
b3b94faa DT |
502 | |
503 | spin_lock(&sdp->sd_rindex_spin); | |
66fc061b SW |
504 | n = sdp->sd_rindex_tree.rb_node; |
505 | while (n) { | |
506 | cur = rb_entry(n, struct gfs2_rgrpd, rd_node); | |
507 | next = NULL; | |
7c9ca621 | 508 | if (blk < cur->rd_addr) |
66fc061b | 509 | next = n->rb_left; |
f75bbfb4 | 510 | else if (blk >= cur->rd_data0 + cur->rd_data) |
66fc061b SW |
511 | next = n->rb_right; |
512 | if (next == NULL) { | |
b3b94faa | 513 | spin_unlock(&sdp->sd_rindex_spin); |
66fc061b SW |
514 | if (exact) { |
515 | if (blk < cur->rd_addr) | |
516 | return NULL; | |
517 | if (blk >= cur->rd_data0 + cur->rd_data) | |
518 | return NULL; | |
519 | } | |
7c9ca621 | 520 | return cur; |
b3b94faa | 521 | } |
66fc061b | 522 | n = next; |
b3b94faa | 523 | } |
b3b94faa DT |
524 | spin_unlock(&sdp->sd_rindex_spin); |
525 | ||
526 | return NULL; | |
527 | } | |
528 | ||
529 | /** | |
530 | * gfs2_rgrpd_get_first - get the first Resource Group in the filesystem | |
531 | * @sdp: The GFS2 superblock | |
532 | * | |
533 | * Returns: The first rgrp in the filesystem | |
534 | */ | |
535 | ||
536 | struct gfs2_rgrpd *gfs2_rgrpd_get_first(struct gfs2_sbd *sdp) | |
537 | { | |
7c9ca621 BP |
538 | const struct rb_node *n; |
539 | struct gfs2_rgrpd *rgd; | |
540 | ||
8339ee54 | 541 | spin_lock(&sdp->sd_rindex_spin); |
7c9ca621 BP |
542 | n = rb_first(&sdp->sd_rindex_tree); |
543 | rgd = rb_entry(n, struct gfs2_rgrpd, rd_node); | |
8339ee54 | 544 | spin_unlock(&sdp->sd_rindex_spin); |
7c9ca621 BP |
545 | |
546 | return rgd; | |
b3b94faa DT |
547 | } |
548 | ||
549 | /** | |
550 | * gfs2_rgrpd_get_next - get the next RG | |
886b1416 | 551 | * @rgd: the resource group descriptor |
b3b94faa DT |
552 | * |
553 | * Returns: The next rgrp | |
554 | */ | |
555 | ||
556 | struct gfs2_rgrpd *gfs2_rgrpd_get_next(struct gfs2_rgrpd *rgd) | |
557 | { | |
7c9ca621 BP |
558 | struct gfs2_sbd *sdp = rgd->rd_sbd; |
559 | const struct rb_node *n; | |
560 | ||
561 | spin_lock(&sdp->sd_rindex_spin); | |
562 | n = rb_next(&rgd->rd_node); | |
563 | if (n == NULL) | |
564 | n = rb_first(&sdp->sd_rindex_tree); | |
565 | ||
566 | if (unlikely(&rgd->rd_node == n)) { | |
567 | spin_unlock(&sdp->sd_rindex_spin); | |
b3b94faa | 568 | return NULL; |
7c9ca621 BP |
569 | } |
570 | rgd = rb_entry(n, struct gfs2_rgrpd, rd_node); | |
571 | spin_unlock(&sdp->sd_rindex_spin); | |
572 | return rgd; | |
b3b94faa DT |
573 | } |
574 | ||
00a158be AD |
575 | void check_and_update_goal(struct gfs2_inode *ip) |
576 | { | |
577 | struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode); | |
578 | if (!ip->i_goal || gfs2_blk2rgrpd(sdp, ip->i_goal, 1) == NULL) | |
579 | ip->i_goal = ip->i_no_addr; | |
580 | } | |
581 | ||
8339ee54 SW |
582 | void gfs2_free_clones(struct gfs2_rgrpd *rgd) |
583 | { | |
584 | int x; | |
585 | ||
586 | for (x = 0; x < rgd->rd_length; x++) { | |
587 | struct gfs2_bitmap *bi = rgd->rd_bits + x; | |
588 | kfree(bi->bi_clone); | |
589 | bi->bi_clone = NULL; | |
590 | } | |
591 | } | |
592 | ||
3792ce97 BP |
593 | static void dump_rs(struct seq_file *seq, const struct gfs2_blkreserv *rs, |
594 | const char *fs_id_buf) | |
8e2e0047 | 595 | { |
f85c10e2 BP |
596 | struct gfs2_inode *ip = container_of(rs, struct gfs2_inode, i_res); |
597 | ||
3792ce97 | 598 | gfs2_print_dbg(seq, "%s B: n:%llu s:%llu b:%u f:%u\n", fs_id_buf, |
f85c10e2 | 599 | (unsigned long long)ip->i_no_addr, |
9e733d39 | 600 | (unsigned long long)gfs2_rbm_to_block(&rs->rs_rbm), |
4a993fb1 | 601 | rs->rs_rbm.offset, rs->rs_free); |
8e2e0047 BP |
602 | } |
603 | ||
0a305e49 | 604 | /** |
8e2e0047 BP |
605 | * __rs_deltree - remove a multi-block reservation from the rgd tree |
606 | * @rs: The reservation to remove | |
607 | * | |
608 | */ | |
20095218 | 609 | static void __rs_deltree(struct gfs2_blkreserv *rs) |
8e2e0047 BP |
610 | { |
611 | struct gfs2_rgrpd *rgd; | |
612 | ||
613 | if (!gfs2_rs_active(rs)) | |
614 | return; | |
615 | ||
4a993fb1 | 616 | rgd = rs->rs_rbm.rgd; |
9e733d39 | 617 | trace_gfs2_rs(rs, TRACE_RS_TREEDEL); |
4a993fb1 | 618 | rb_erase(&rs->rs_node, &rgd->rd_rstree); |
24d634e8 | 619 | RB_CLEAR_NODE(&rs->rs_node); |
8e2e0047 BP |
620 | |
621 | if (rs->rs_free) { | |
ec23df2b AG |
622 | u64 last_block = gfs2_rbm_to_block(&rs->rs_rbm) + |
623 | rs->rs_free - 1; | |
624 | struct gfs2_rbm last_rbm = { .rgd = rs->rs_rbm.rgd, }; | |
625 | struct gfs2_bitmap *start, *last; | |
e579ed4f | 626 | |
20095218 | 627 | /* return reserved blocks to the rgrp */ |
4a993fb1 SW |
628 | BUG_ON(rs->rs_rbm.rgd->rd_reserved < rs->rs_free); |
629 | rs->rs_rbm.rgd->rd_reserved -= rs->rs_free; | |
5ea5050c BP |
630 | /* The rgrp extent failure point is likely not to increase; |
631 | it will only do so if the freed blocks are somehow | |
632 | contiguous with a span of free blocks that follows. Still, | |
633 | it will force the number to be recalculated later. */ | |
634 | rgd->rd_extfail_pt += rs->rs_free; | |
8e2e0047 | 635 | rs->rs_free = 0; |
ec23df2b AG |
636 | if (gfs2_rbm_from_block(&last_rbm, last_block)) |
637 | return; | |
638 | start = rbm_bi(&rs->rs_rbm); | |
639 | last = rbm_bi(&last_rbm); | |
640 | do | |
641 | clear_bit(GBF_FULL, &start->bi_flags); | |
642 | while (start++ != last); | |
8e2e0047 | 643 | } |
8e2e0047 BP |
644 | } |
645 | ||
646 | /** | |
647 | * gfs2_rs_deltree - remove a multi-block reservation from the rgd tree | |
648 | * @rs: The reservation to remove | |
649 | * | |
650 | */ | |
20095218 | 651 | void gfs2_rs_deltree(struct gfs2_blkreserv *rs) |
8e2e0047 BP |
652 | { |
653 | struct gfs2_rgrpd *rgd; | |
654 | ||
4a993fb1 SW |
655 | rgd = rs->rs_rbm.rgd; |
656 | if (rgd) { | |
657 | spin_lock(&rgd->rd_rsspin); | |
20095218 | 658 | __rs_deltree(rs); |
44f52122 | 659 | BUG_ON(rs->rs_free); |
4a993fb1 SW |
660 | spin_unlock(&rgd->rd_rsspin); |
661 | } | |
8e2e0047 BP |
662 | } |
663 | ||
664 | /** | |
1595548f | 665 | * gfs2_rs_delete - delete a multi-block reservation |
0a305e49 | 666 | * @ip: The inode for this reservation |
af5c2697 | 667 | * @wcount: The inode's write count, or NULL |
0a305e49 BP |
668 | * |
669 | */ | |
1595548f | 670 | void gfs2_rs_delete(struct gfs2_inode *ip, atomic_t *wcount) |
0a305e49 BP |
671 | { |
672 | down_write(&ip->i_rw_mutex); | |
44f52122 | 673 | if ((wcount == NULL) || (atomic_read(wcount) <= 1)) |
a097dc7e | 674 | gfs2_rs_deltree(&ip->i_res); |
0a305e49 BP |
675 | up_write(&ip->i_rw_mutex); |
676 | } | |
677 | ||
8e2e0047 BP |
678 | /** |
679 | * return_all_reservations - return all reserved blocks back to the rgrp. | |
680 | * @rgd: the rgrp that needs its space back | |
681 | * | |
682 | * We previously reserved a bunch of blocks for allocation. Now we need to | |
683 | * give them back. This leave the reservation structures in tact, but removes | |
684 | * all of their corresponding "no-fly zones". | |
685 | */ | |
686 | static void return_all_reservations(struct gfs2_rgrpd *rgd) | |
687 | { | |
688 | struct rb_node *n; | |
689 | struct gfs2_blkreserv *rs; | |
690 | ||
691 | spin_lock(&rgd->rd_rsspin); | |
692 | while ((n = rb_first(&rgd->rd_rstree))) { | |
693 | rs = rb_entry(n, struct gfs2_blkreserv, rs_node); | |
20095218 | 694 | __rs_deltree(rs); |
8e2e0047 BP |
695 | } |
696 | spin_unlock(&rgd->rd_rsspin); | |
697 | } | |
698 | ||
8339ee54 | 699 | void gfs2_clear_rgrpd(struct gfs2_sbd *sdp) |
b3b94faa | 700 | { |
7c9ca621 | 701 | struct rb_node *n; |
b3b94faa DT |
702 | struct gfs2_rgrpd *rgd; |
703 | struct gfs2_glock *gl; | |
704 | ||
7c9ca621 BP |
705 | while ((n = rb_first(&sdp->sd_rindex_tree))) { |
706 | rgd = rb_entry(n, struct gfs2_rgrpd, rd_node); | |
b3b94faa DT |
707 | gl = rgd->rd_gl; |
708 | ||
7c9ca621 | 709 | rb_erase(n, &sdp->sd_rindex_tree); |
b3b94faa DT |
710 | |
711 | if (gl) { | |
b3422cac BP |
712 | if (gl->gl_state != LM_ST_UNLOCKED) { |
713 | gfs2_glock_cb(gl, LM_ST_UNLOCKED); | |
714 | flush_delayed_work(&gl->gl_work); | |
715 | } | |
10283ea5 | 716 | gfs2_rgrp_brelse(rgd); |
b3422cac | 717 | glock_clear_object(gl, rgd); |
b3b94faa DT |
718 | gfs2_glock_put(gl); |
719 | } | |
720 | ||
8339ee54 | 721 | gfs2_free_clones(rgd); |
b3b94faa | 722 | kfree(rgd->rd_bits); |
36e4ad03 | 723 | rgd->rd_bits = NULL; |
8e2e0047 | 724 | return_all_reservations(rgd); |
6bdd9be6 | 725 | kmem_cache_free(gfs2_rgrpd_cachep, rgd); |
b3b94faa DT |
726 | } |
727 | } | |
728 | ||
b3b94faa DT |
729 | /** |
730 | * gfs2_compute_bitstructs - Compute the bitmap sizes | |
731 | * @rgd: The resource group descriptor | |
732 | * | |
733 | * Calculates bitmap descriptors, one for each block that contains bitmap data | |
734 | * | |
735 | * Returns: errno | |
736 | */ | |
737 | ||
738 | static int compute_bitstructs(struct gfs2_rgrpd *rgd) | |
739 | { | |
740 | struct gfs2_sbd *sdp = rgd->rd_sbd; | |
741 | struct gfs2_bitmap *bi; | |
bb8d8a6f | 742 | u32 length = rgd->rd_length; /* # blocks in hdr & bitmap */ |
cd915493 | 743 | u32 bytes_left, bytes; |
b3b94faa DT |
744 | int x; |
745 | ||
feaa7bba SW |
746 | if (!length) |
747 | return -EINVAL; | |
748 | ||
dd894be8 | 749 | rgd->rd_bits = kcalloc(length, sizeof(struct gfs2_bitmap), GFP_NOFS); |
b3b94faa DT |
750 | if (!rgd->rd_bits) |
751 | return -ENOMEM; | |
752 | ||
bb8d8a6f | 753 | bytes_left = rgd->rd_bitbytes; |
b3b94faa DT |
754 | |
755 | for (x = 0; x < length; x++) { | |
756 | bi = rgd->rd_bits + x; | |
757 | ||
60a0b8f9 | 758 | bi->bi_flags = 0; |
b3b94faa DT |
759 | /* small rgrp; bitmap stored completely in header block */ |
760 | if (length == 1) { | |
761 | bytes = bytes_left; | |
762 | bi->bi_offset = sizeof(struct gfs2_rgrp); | |
763 | bi->bi_start = 0; | |
281b4952 | 764 | bi->bi_bytes = bytes; |
7e230f57 | 765 | bi->bi_blocks = bytes * GFS2_NBBY; |
b3b94faa DT |
766 | /* header block */ |
767 | } else if (x == 0) { | |
768 | bytes = sdp->sd_sb.sb_bsize - sizeof(struct gfs2_rgrp); | |
769 | bi->bi_offset = sizeof(struct gfs2_rgrp); | |
770 | bi->bi_start = 0; | |
281b4952 | 771 | bi->bi_bytes = bytes; |
7e230f57 | 772 | bi->bi_blocks = bytes * GFS2_NBBY; |
b3b94faa DT |
773 | /* last block */ |
774 | } else if (x + 1 == length) { | |
775 | bytes = bytes_left; | |
776 | bi->bi_offset = sizeof(struct gfs2_meta_header); | |
bb8d8a6f | 777 | bi->bi_start = rgd->rd_bitbytes - bytes_left; |
281b4952 | 778 | bi->bi_bytes = bytes; |
7e230f57 | 779 | bi->bi_blocks = bytes * GFS2_NBBY; |
b3b94faa DT |
780 | /* other blocks */ |
781 | } else { | |
568f4c96 SW |
782 | bytes = sdp->sd_sb.sb_bsize - |
783 | sizeof(struct gfs2_meta_header); | |
b3b94faa | 784 | bi->bi_offset = sizeof(struct gfs2_meta_header); |
bb8d8a6f | 785 | bi->bi_start = rgd->rd_bitbytes - bytes_left; |
281b4952 | 786 | bi->bi_bytes = bytes; |
7e230f57 | 787 | bi->bi_blocks = bytes * GFS2_NBBY; |
b3b94faa DT |
788 | } |
789 | ||
790 | bytes_left -= bytes; | |
791 | } | |
792 | ||
793 | if (bytes_left) { | |
794 | gfs2_consist_rgrpd(rgd); | |
795 | return -EIO; | |
796 | } | |
797 | bi = rgd->rd_bits + (length - 1); | |
281b4952 | 798 | if ((bi->bi_start + bi->bi_bytes) * GFS2_NBBY != rgd->rd_data) { |
8dc88ac6 AG |
799 | gfs2_lm(sdp, |
800 | "ri_addr = %llu\n" | |
801 | "ri_length = %u\n" | |
802 | "ri_data0 = %llu\n" | |
803 | "ri_data = %u\n" | |
804 | "ri_bitbytes = %u\n" | |
805 | "start=%u len=%u offset=%u\n", | |
806 | (unsigned long long)rgd->rd_addr, | |
807 | rgd->rd_length, | |
808 | (unsigned long long)rgd->rd_data0, | |
809 | rgd->rd_data, | |
810 | rgd->rd_bitbytes, | |
811 | bi->bi_start, bi->bi_bytes, bi->bi_offset); | |
812 | gfs2_consist_rgrpd(rgd); | |
b3b94faa DT |
813 | return -EIO; |
814 | } | |
815 | ||
816 | return 0; | |
817 | } | |
818 | ||
7ae8fa84 RP |
819 | /** |
820 | * gfs2_ri_total - Total up the file system space, according to the rindex. | |
886b1416 | 821 | * @sdp: the filesystem |
7ae8fa84 RP |
822 | * |
823 | */ | |
824 | u64 gfs2_ri_total(struct gfs2_sbd *sdp) | |
825 | { | |
826 | u64 total_data = 0; | |
827 | struct inode *inode = sdp->sd_rindex; | |
828 | struct gfs2_inode *ip = GFS2_I(inode); | |
7ae8fa84 | 829 | char buf[sizeof(struct gfs2_rindex)]; |
7ae8fa84 RP |
830 | int error, rgrps; |
831 | ||
7ae8fa84 RP |
832 | for (rgrps = 0;; rgrps++) { |
833 | loff_t pos = rgrps * sizeof(struct gfs2_rindex); | |
834 | ||
bcd7278d | 835 | if (pos + sizeof(struct gfs2_rindex) > i_size_read(inode)) |
7ae8fa84 | 836 | break; |
4306629e | 837 | error = gfs2_internal_read(ip, buf, &pos, |
7ae8fa84 RP |
838 | sizeof(struct gfs2_rindex)); |
839 | if (error != sizeof(struct gfs2_rindex)) | |
840 | break; | |
bb8d8a6f | 841 | total_data += be32_to_cpu(((struct gfs2_rindex *)buf)->ri_data); |
7ae8fa84 | 842 | } |
7ae8fa84 RP |
843 | return total_data; |
844 | } | |
845 | ||
6aad1c3d | 846 | static int rgd_insert(struct gfs2_rgrpd *rgd) |
7c9ca621 BP |
847 | { |
848 | struct gfs2_sbd *sdp = rgd->rd_sbd; | |
849 | struct rb_node **newn = &sdp->sd_rindex_tree.rb_node, *parent = NULL; | |
850 | ||
851 | /* Figure out where to put new node */ | |
852 | while (*newn) { | |
853 | struct gfs2_rgrpd *cur = rb_entry(*newn, struct gfs2_rgrpd, | |
854 | rd_node); | |
855 | ||
856 | parent = *newn; | |
857 | if (rgd->rd_addr < cur->rd_addr) | |
858 | newn = &((*newn)->rb_left); | |
859 | else if (rgd->rd_addr > cur->rd_addr) | |
860 | newn = &((*newn)->rb_right); | |
861 | else | |
6aad1c3d | 862 | return -EEXIST; |
7c9ca621 BP |
863 | } |
864 | ||
865 | rb_link_node(&rgd->rd_node, parent, newn); | |
866 | rb_insert_color(&rgd->rd_node, &sdp->sd_rindex_tree); | |
6aad1c3d BP |
867 | sdp->sd_rgrps++; |
868 | return 0; | |
7c9ca621 BP |
869 | } |
870 | ||
b3b94faa | 871 | /** |
6c53267f | 872 | * read_rindex_entry - Pull in a new resource index entry from the disk |
4306629e | 873 | * @ip: Pointer to the rindex inode |
b3b94faa | 874 | * |
8339ee54 | 875 | * Returns: 0 on success, > 0 on EOF, error code otherwise |
6c53267f RP |
876 | */ |
877 | ||
4306629e | 878 | static int read_rindex_entry(struct gfs2_inode *ip) |
6c53267f RP |
879 | { |
880 | struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode); | |
881 | loff_t pos = sdp->sd_rgrps * sizeof(struct gfs2_rindex); | |
8339ee54 | 882 | struct gfs2_rindex buf; |
6c53267f RP |
883 | int error; |
884 | struct gfs2_rgrpd *rgd; | |
885 | ||
8339ee54 SW |
886 | if (pos >= i_size_read(&ip->i_inode)) |
887 | return 1; | |
888 | ||
4306629e | 889 | error = gfs2_internal_read(ip, (char *)&buf, &pos, |
6c53267f | 890 | sizeof(struct gfs2_rindex)); |
8339ee54 SW |
891 | |
892 | if (error != sizeof(struct gfs2_rindex)) | |
893 | return (error == 0) ? 1 : error; | |
6c53267f | 894 | |
6bdd9be6 | 895 | rgd = kmem_cache_zalloc(gfs2_rgrpd_cachep, GFP_NOFS); |
6c53267f RP |
896 | error = -ENOMEM; |
897 | if (!rgd) | |
898 | return error; | |
899 | ||
6c53267f | 900 | rgd->rd_sbd = sdp; |
8339ee54 SW |
901 | rgd->rd_addr = be64_to_cpu(buf.ri_addr); |
902 | rgd->rd_length = be32_to_cpu(buf.ri_length); | |
903 | rgd->rd_data0 = be64_to_cpu(buf.ri_data0); | |
904 | rgd->rd_data = be32_to_cpu(buf.ri_data); | |
905 | rgd->rd_bitbytes = be32_to_cpu(buf.ri_bitbytes); | |
8e2e0047 | 906 | spin_lock_init(&rgd->rd_rsspin); |
7c9ca621 | 907 | |
6c53267f RP |
908 | error = compute_bitstructs(rgd); |
909 | if (error) | |
8339ee54 | 910 | goto fail; |
6c53267f | 911 | |
bb8d8a6f | 912 | error = gfs2_glock_get(sdp, rgd->rd_addr, |
6c53267f RP |
913 | &gfs2_rgrp_glops, CREATE, &rgd->rd_gl); |
914 | if (error) | |
8339ee54 | 915 | goto fail; |
6c53267f | 916 | |
4e2f8849 | 917 | rgd->rd_rgl = (struct gfs2_rgrp_lvb *)rgd->rd_gl->gl_lksb.sb_lvbptr; |
0e27c18c | 918 | rgd->rd_flags &= ~(GFS2_RDF_UPTODATE | GFS2_RDF_PREFERRED); |
7c9ca621 BP |
919 | if (rgd->rd_data > sdp->sd_max_rg_data) |
920 | sdp->sd_max_rg_data = rgd->rd_data; | |
8339ee54 | 921 | spin_lock(&sdp->sd_rindex_spin); |
6aad1c3d | 922 | error = rgd_insert(rgd); |
8339ee54 | 923 | spin_unlock(&sdp->sd_rindex_spin); |
36e4ad03 | 924 | if (!error) { |
6f6597ba | 925 | glock_set_object(rgd->rd_gl, rgd); |
6aad1c3d | 926 | return 0; |
36e4ad03 | 927 | } |
6aad1c3d BP |
928 | |
929 | error = 0; /* someone else read in the rgrp; free it and ignore it */ | |
c1ac539e | 930 | gfs2_glock_put(rgd->rd_gl); |
8339ee54 SW |
931 | |
932 | fail: | |
933 | kfree(rgd->rd_bits); | |
36e4ad03 | 934 | rgd->rd_bits = NULL; |
8339ee54 | 935 | kmem_cache_free(gfs2_rgrpd_cachep, rgd); |
6c53267f RP |
936 | return error; |
937 | } | |
938 | ||
0e27c18c BP |
939 | /** |
940 | * set_rgrp_preferences - Run all the rgrps, selecting some we prefer to use | |
941 | * @sdp: the GFS2 superblock | |
942 | * | |
943 | * The purpose of this function is to select a subset of the resource groups | |
944 | * and mark them as PREFERRED. We do it in such a way that each node prefers | |
945 | * to use a unique set of rgrps to minimize glock contention. | |
946 | */ | |
947 | static void set_rgrp_preferences(struct gfs2_sbd *sdp) | |
948 | { | |
949 | struct gfs2_rgrpd *rgd, *first; | |
950 | int i; | |
951 | ||
952 | /* Skip an initial number of rgrps, based on this node's journal ID. | |
953 | That should start each node out on its own set. */ | |
954 | rgd = gfs2_rgrpd_get_first(sdp); | |
955 | for (i = 0; i < sdp->sd_lockstruct.ls_jid; i++) | |
956 | rgd = gfs2_rgrpd_get_next(rgd); | |
957 | first = rgd; | |
958 | ||
959 | do { | |
960 | rgd->rd_flags |= GFS2_RDF_PREFERRED; | |
961 | for (i = 0; i < sdp->sd_journals; i++) { | |
962 | rgd = gfs2_rgrpd_get_next(rgd); | |
959b6717 | 963 | if (!rgd || rgd == first) |
0e27c18c BP |
964 | break; |
965 | } | |
959b6717 | 966 | } while (rgd && rgd != first); |
0e27c18c BP |
967 | } |
968 | ||
6c53267f RP |
969 | /** |
970 | * gfs2_ri_update - Pull in a new resource index from the disk | |
971 | * @ip: pointer to the rindex inode | |
972 | * | |
b3b94faa DT |
973 | * Returns: 0 on successful update, error code otherwise |
974 | */ | |
975 | ||
8339ee54 | 976 | static int gfs2_ri_update(struct gfs2_inode *ip) |
b3b94faa | 977 | { |
feaa7bba | 978 | struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode); |
b3b94faa DT |
979 | int error; |
980 | ||
8339ee54 | 981 | do { |
4306629e | 982 | error = read_rindex_entry(ip); |
8339ee54 SW |
983 | } while (error == 0); |
984 | ||
985 | if (error < 0) | |
986 | return error; | |
b3b94faa | 987 | |
0e27c18c BP |
988 | set_rgrp_preferences(sdp); |
989 | ||
cf45b752 | 990 | sdp->sd_rindex_uptodate = 1; |
6c53267f RP |
991 | return 0; |
992 | } | |
b3b94faa | 993 | |
b3b94faa | 994 | /** |
8339ee54 | 995 | * gfs2_rindex_update - Update the rindex if required |
b3b94faa | 996 | * @sdp: The GFS2 superblock |
b3b94faa DT |
997 | * |
998 | * We grab a lock on the rindex inode to make sure that it doesn't | |
999 | * change whilst we are performing an operation. We keep this lock | |
1000 | * for quite long periods of time compared to other locks. This | |
1001 | * doesn't matter, since it is shared and it is very, very rarely | |
1002 | * accessed in the exclusive mode (i.e. only when expanding the filesystem). | |
1003 | * | |
1004 | * This makes sure that we're using the latest copy of the resource index | |
1005 | * special file, which might have been updated if someone expanded the | |
1006 | * filesystem (via gfs2_grow utility), which adds new resource groups. | |
1007 | * | |
8339ee54 | 1008 | * Returns: 0 on succeess, error code otherwise |
b3b94faa DT |
1009 | */ |
1010 | ||
8339ee54 | 1011 | int gfs2_rindex_update(struct gfs2_sbd *sdp) |
b3b94faa | 1012 | { |
feaa7bba | 1013 | struct gfs2_inode *ip = GFS2_I(sdp->sd_rindex); |
b3b94faa | 1014 | struct gfs2_glock *gl = ip->i_gl; |
8339ee54 SW |
1015 | struct gfs2_holder ri_gh; |
1016 | int error = 0; | |
a365fbf3 | 1017 | int unlock_required = 0; |
b3b94faa DT |
1018 | |
1019 | /* Read new copy from disk if we don't have the latest */ | |
cf45b752 | 1020 | if (!sdp->sd_rindex_uptodate) { |
a365fbf3 SW |
1021 | if (!gfs2_glock_is_locked_by_me(gl)) { |
1022 | error = gfs2_glock_nq_init(gl, LM_ST_SHARED, 0, &ri_gh); | |
1023 | if (error) | |
6aad1c3d | 1024 | return error; |
a365fbf3 SW |
1025 | unlock_required = 1; |
1026 | } | |
8339ee54 | 1027 | if (!sdp->sd_rindex_uptodate) |
b3b94faa | 1028 | error = gfs2_ri_update(ip); |
a365fbf3 SW |
1029 | if (unlock_required) |
1030 | gfs2_glock_dq_uninit(&ri_gh); | |
b3b94faa DT |
1031 | } |
1032 | ||
1033 | return error; | |
1034 | } | |
1035 | ||
42d52e38 | 1036 | static void gfs2_rgrp_in(struct gfs2_rgrpd *rgd, const void *buf) |
bb8d8a6f SW |
1037 | { |
1038 | const struct gfs2_rgrp *str = buf; | |
42d52e38 | 1039 | u32 rg_flags; |
bb8d8a6f | 1040 | |
42d52e38 | 1041 | rg_flags = be32_to_cpu(str->rg_flags); |
09010978 | 1042 | rg_flags &= ~GFS2_RDF_MASK; |
1ce97e56 SW |
1043 | rgd->rd_flags &= GFS2_RDF_MASK; |
1044 | rgd->rd_flags |= rg_flags; | |
cfc8b549 | 1045 | rgd->rd_free = be32_to_cpu(str->rg_free); |
73f74948 | 1046 | rgd->rd_dinodes = be32_to_cpu(str->rg_dinodes); |
d8b71f73 | 1047 | rgd->rd_igeneration = be64_to_cpu(str->rg_igeneration); |
166725d9 | 1048 | /* rd_data0, rd_data and rd_bitbytes already set from rindex */ |
bb8d8a6f SW |
1049 | } |
1050 | ||
3f30f929 BP |
1051 | static void gfs2_rgrp_ondisk2lvb(struct gfs2_rgrp_lvb *rgl, const void *buf) |
1052 | { | |
1053 | const struct gfs2_rgrp *str = buf; | |
1054 | ||
1055 | rgl->rl_magic = cpu_to_be32(GFS2_MAGIC); | |
1056 | rgl->rl_flags = str->rg_flags; | |
1057 | rgl->rl_free = str->rg_free; | |
1058 | rgl->rl_dinodes = str->rg_dinodes; | |
1059 | rgl->rl_igeneration = str->rg_igeneration; | |
1060 | rgl->__pad = 0UL; | |
1061 | } | |
1062 | ||
42d52e38 | 1063 | static void gfs2_rgrp_out(struct gfs2_rgrpd *rgd, void *buf) |
bb8d8a6f | 1064 | { |
65adc273 | 1065 | struct gfs2_rgrpd *next = gfs2_rgrpd_get_next(rgd); |
bb8d8a6f | 1066 | struct gfs2_rgrp *str = buf; |
850d2d91 | 1067 | u32 crc; |
bb8d8a6f | 1068 | |
09010978 | 1069 | str->rg_flags = cpu_to_be32(rgd->rd_flags & ~GFS2_RDF_MASK); |
cfc8b549 | 1070 | str->rg_free = cpu_to_be32(rgd->rd_free); |
73f74948 | 1071 | str->rg_dinodes = cpu_to_be32(rgd->rd_dinodes); |
65adc273 AP |
1072 | if (next == NULL) |
1073 | str->rg_skip = 0; | |
1074 | else if (next->rd_addr > rgd->rd_addr) | |
1075 | str->rg_skip = cpu_to_be32(next->rd_addr - rgd->rd_addr); | |
d8b71f73 | 1076 | str->rg_igeneration = cpu_to_be64(rgd->rd_igeneration); |
166725d9 AP |
1077 | str->rg_data0 = cpu_to_be64(rgd->rd_data0); |
1078 | str->rg_data = cpu_to_be32(rgd->rd_data); | |
1079 | str->rg_bitbytes = cpu_to_be32(rgd->rd_bitbytes); | |
850d2d91 AP |
1080 | str->rg_crc = 0; |
1081 | crc = gfs2_disk_hash(buf, sizeof(struct gfs2_rgrp)); | |
1082 | str->rg_crc = cpu_to_be32(crc); | |
166725d9 | 1083 | |
bb8d8a6f | 1084 | memset(&str->rg_reserved, 0, sizeof(str->rg_reserved)); |
3f30f929 | 1085 | gfs2_rgrp_ondisk2lvb(rgd->rd_rgl, buf); |
bb8d8a6f SW |
1086 | } |
1087 | ||
90306c41 BM |
1088 | static int gfs2_rgrp_lvb_valid(struct gfs2_rgrpd *rgd) |
1089 | { | |
1090 | struct gfs2_rgrp_lvb *rgl = rgd->rd_rgl; | |
1091 | struct gfs2_rgrp *str = (struct gfs2_rgrp *)rgd->rd_bits[0].bi_bh->b_data; | |
f29e62ee | 1092 | struct gfs2_sbd *sdp = rgd->rd_sbd; |
72244b6b | 1093 | int valid = 1; |
90306c41 | 1094 | |
72244b6b | 1095 | if (rgl->rl_flags != str->rg_flags) { |
f29e62ee BP |
1096 | fs_warn(sdp, "GFS2: rgd: %llu lvb flag mismatch %u/%u", |
1097 | (unsigned long long)rgd->rd_addr, | |
72244b6b BP |
1098 | be32_to_cpu(rgl->rl_flags), be32_to_cpu(str->rg_flags)); |
1099 | valid = 0; | |
1100 | } | |
1101 | if (rgl->rl_free != str->rg_free) { | |
f29e62ee BP |
1102 | fs_warn(sdp, "GFS2: rgd: %llu lvb free mismatch %u/%u", |
1103 | (unsigned long long)rgd->rd_addr, | |
1104 | be32_to_cpu(rgl->rl_free), be32_to_cpu(str->rg_free)); | |
72244b6b BP |
1105 | valid = 0; |
1106 | } | |
1107 | if (rgl->rl_dinodes != str->rg_dinodes) { | |
f29e62ee BP |
1108 | fs_warn(sdp, "GFS2: rgd: %llu lvb dinode mismatch %u/%u", |
1109 | (unsigned long long)rgd->rd_addr, | |
1110 | be32_to_cpu(rgl->rl_dinodes), | |
1111 | be32_to_cpu(str->rg_dinodes)); | |
72244b6b BP |
1112 | valid = 0; |
1113 | } | |
1114 | if (rgl->rl_igeneration != str->rg_igeneration) { | |
f29e62ee BP |
1115 | fs_warn(sdp, "GFS2: rgd: %llu lvb igen mismatch %llu/%llu", |
1116 | (unsigned long long)rgd->rd_addr, | |
1117 | (unsigned long long)be64_to_cpu(rgl->rl_igeneration), | |
1118 | (unsigned long long)be64_to_cpu(str->rg_igeneration)); | |
72244b6b BP |
1119 | valid = 0; |
1120 | } | |
1121 | return valid; | |
90306c41 BM |
1122 | } |
1123 | ||
90306c41 BM |
1124 | static u32 count_unlinked(struct gfs2_rgrpd *rgd) |
1125 | { | |
1126 | struct gfs2_bitmap *bi; | |
1127 | const u32 length = rgd->rd_length; | |
1128 | const u8 *buffer = NULL; | |
1129 | u32 i, goal, count = 0; | |
1130 | ||
1131 | for (i = 0, bi = rgd->rd_bits; i < length; i++, bi++) { | |
1132 | goal = 0; | |
1133 | buffer = bi->bi_bh->b_data + bi->bi_offset; | |
1134 | WARN_ON(!buffer_uptodate(bi->bi_bh)); | |
281b4952 AG |
1135 | while (goal < bi->bi_blocks) { |
1136 | goal = gfs2_bitfit(buffer, bi->bi_bytes, goal, | |
90306c41 BM |
1137 | GFS2_BLKST_UNLINKED); |
1138 | if (goal == BFITNOENT) | |
1139 | break; | |
1140 | count++; | |
1141 | goal++; | |
1142 | } | |
1143 | } | |
1144 | ||
1145 | return count; | |
1146 | } | |
1147 | ||
1148 | ||
b3b94faa | 1149 | /** |
90306c41 BM |
1150 | * gfs2_rgrp_bh_get - Read in a RG's header and bitmaps |
1151 | * @rgd: the struct gfs2_rgrpd describing the RG to read in | |
b3b94faa DT |
1152 | * |
1153 | * Read in all of a Resource Group's header and bitmap blocks. | |
10283ea5 | 1154 | * Caller must eventually call gfs2_rgrp_brelse() to free the bitmaps. |
b3b94faa DT |
1155 | * |
1156 | * Returns: errno | |
1157 | */ | |
1158 | ||
c2b0b30e | 1159 | static int gfs2_rgrp_bh_get(struct gfs2_rgrpd *rgd) |
b3b94faa DT |
1160 | { |
1161 | struct gfs2_sbd *sdp = rgd->rd_sbd; | |
1162 | struct gfs2_glock *gl = rgd->rd_gl; | |
bb8d8a6f | 1163 | unsigned int length = rgd->rd_length; |
b3b94faa DT |
1164 | struct gfs2_bitmap *bi; |
1165 | unsigned int x, y; | |
1166 | int error; | |
1167 | ||
90306c41 BM |
1168 | if (rgd->rd_bits[0].bi_bh != NULL) |
1169 | return 0; | |
1170 | ||
b3b94faa DT |
1171 | for (x = 0; x < length; x++) { |
1172 | bi = rgd->rd_bits + x; | |
c8d57703 | 1173 | error = gfs2_meta_read(gl, rgd->rd_addr + x, 0, 0, &bi->bi_bh); |
b3b94faa DT |
1174 | if (error) |
1175 | goto fail; | |
1176 | } | |
1177 | ||
1178 | for (y = length; y--;) { | |
1179 | bi = rgd->rd_bits + y; | |
7276b3b0 | 1180 | error = gfs2_meta_wait(sdp, bi->bi_bh); |
b3b94faa DT |
1181 | if (error) |
1182 | goto fail; | |
feaa7bba | 1183 | if (gfs2_metatype_check(sdp, bi->bi_bh, y ? GFS2_METATYPE_RB : |
b3b94faa DT |
1184 | GFS2_METATYPE_RG)) { |
1185 | error = -EIO; | |
1186 | goto fail; | |
1187 | } | |
1188 | } | |
1189 | ||
cf45b752 | 1190 | if (!(rgd->rd_flags & GFS2_RDF_UPTODATE)) { |
60a0b8f9 SW |
1191 | for (x = 0; x < length; x++) |
1192 | clear_bit(GBF_FULL, &rgd->rd_bits[x].bi_flags); | |
42d52e38 | 1193 | gfs2_rgrp_in(rgd, (rgd->rd_bits[0].bi_bh)->b_data); |
1ce97e56 | 1194 | rgd->rd_flags |= (GFS2_RDF_UPTODATE | GFS2_RDF_CHECK); |
7c9ca621 | 1195 | rgd->rd_free_clone = rgd->rd_free; |
5ea5050c BP |
1196 | /* max out the rgrp allocation failure point */ |
1197 | rgd->rd_extfail_pt = rgd->rd_free; | |
b3b94faa | 1198 | } |
951b4bd5 | 1199 | if (cpu_to_be32(GFS2_MAGIC) != rgd->rd_rgl->rl_magic) { |
90306c41 BM |
1200 | rgd->rd_rgl->rl_unlinked = cpu_to_be32(count_unlinked(rgd)); |
1201 | gfs2_rgrp_ondisk2lvb(rgd->rd_rgl, | |
1202 | rgd->rd_bits[0].bi_bh->b_data); | |
1203 | } | |
1204 | else if (sdp->sd_args.ar_rgrplvb) { | |
1205 | if (!gfs2_rgrp_lvb_valid(rgd)){ | |
1206 | gfs2_consist_rgrpd(rgd); | |
1207 | error = -EIO; | |
1208 | goto fail; | |
1209 | } | |
1210 | if (rgd->rd_rgl->rl_unlinked == 0) | |
1211 | rgd->rd_flags &= ~GFS2_RDF_CHECK; | |
1212 | } | |
b3b94faa DT |
1213 | return 0; |
1214 | ||
feaa7bba | 1215 | fail: |
b3b94faa DT |
1216 | while (x--) { |
1217 | bi = rgd->rd_bits + x; | |
1218 | brelse(bi->bi_bh); | |
1219 | bi->bi_bh = NULL; | |
1220 | gfs2_assert_warn(sdp, !bi->bi_clone); | |
1221 | } | |
b3b94faa DT |
1222 | |
1223 | return error; | |
1224 | } | |
1225 | ||
c2b0b30e | 1226 | static int update_rgrp_lvb(struct gfs2_rgrpd *rgd) |
90306c41 BM |
1227 | { |
1228 | u32 rl_flags; | |
1229 | ||
1230 | if (rgd->rd_flags & GFS2_RDF_UPTODATE) | |
1231 | return 0; | |
1232 | ||
951b4bd5 | 1233 | if (cpu_to_be32(GFS2_MAGIC) != rgd->rd_rgl->rl_magic) |
90306c41 BM |
1234 | return gfs2_rgrp_bh_get(rgd); |
1235 | ||
1236 | rl_flags = be32_to_cpu(rgd->rd_rgl->rl_flags); | |
1237 | rl_flags &= ~GFS2_RDF_MASK; | |
1238 | rgd->rd_flags &= GFS2_RDF_MASK; | |
4f36cb36 | 1239 | rgd->rd_flags |= (rl_flags | GFS2_RDF_CHECK); |
90306c41 BM |
1240 | if (rgd->rd_rgl->rl_unlinked == 0) |
1241 | rgd->rd_flags &= ~GFS2_RDF_CHECK; | |
1242 | rgd->rd_free = be32_to_cpu(rgd->rd_rgl->rl_free); | |
1243 | rgd->rd_free_clone = rgd->rd_free; | |
1244 | rgd->rd_dinodes = be32_to_cpu(rgd->rd_rgl->rl_dinodes); | |
1245 | rgd->rd_igeneration = be64_to_cpu(rgd->rd_rgl->rl_igeneration); | |
1246 | return 0; | |
1247 | } | |
1248 | ||
1249 | int gfs2_rgrp_go_lock(struct gfs2_holder *gh) | |
1250 | { | |
1251 | struct gfs2_rgrpd *rgd = gh->gh_gl->gl_object; | |
1252 | struct gfs2_sbd *sdp = rgd->rd_sbd; | |
1253 | ||
1254 | if (gh->gh_flags & GL_SKIP && sdp->sd_args.ar_rgrplvb) | |
1255 | return 0; | |
8b127d04 | 1256 | return gfs2_rgrp_bh_get(rgd); |
90306c41 BM |
1257 | } |
1258 | ||
b3b94faa | 1259 | /** |
39b0f1e9 BP |
1260 | * gfs2_rgrp_brelse - Release RG bitmaps read in with gfs2_rgrp_bh_get() |
1261 | * @rgd: The resource group | |
b3b94faa DT |
1262 | * |
1263 | */ | |
1264 | ||
39b0f1e9 | 1265 | void gfs2_rgrp_brelse(struct gfs2_rgrpd *rgd) |
b3b94faa | 1266 | { |
bb8d8a6f | 1267 | int x, length = rgd->rd_length; |
b3b94faa | 1268 | |
b3b94faa DT |
1269 | for (x = 0; x < length; x++) { |
1270 | struct gfs2_bitmap *bi = rgd->rd_bits + x; | |
90306c41 BM |
1271 | if (bi->bi_bh) { |
1272 | brelse(bi->bi_bh); | |
1273 | bi->bi_bh = NULL; | |
1274 | } | |
b3b94faa | 1275 | } |
39b0f1e9 BP |
1276 | } |
1277 | ||
66fc061b | 1278 | int gfs2_rgrp_send_discards(struct gfs2_sbd *sdp, u64 offset, |
7c9ca621 | 1279 | struct buffer_head *bh, |
66fc061b | 1280 | const struct gfs2_bitmap *bi, unsigned minlen, u64 *ptrimmed) |
f15ab561 SW |
1281 | { |
1282 | struct super_block *sb = sdp->sd_vfs; | |
f15ab561 | 1283 | u64 blk; |
64d576ba | 1284 | sector_t start = 0; |
b2c87cae | 1285 | sector_t nr_blks = 0; |
f15ab561 SW |
1286 | int rv; |
1287 | unsigned int x; | |
66fc061b SW |
1288 | u32 trimmed = 0; |
1289 | u8 diff; | |
f15ab561 | 1290 | |
281b4952 | 1291 | for (x = 0; x < bi->bi_bytes; x++) { |
66fc061b SW |
1292 | const u8 *clone = bi->bi_clone ? bi->bi_clone : bi->bi_bh->b_data; |
1293 | clone += bi->bi_offset; | |
1294 | clone += x; | |
1295 | if (bh) { | |
1296 | const u8 *orig = bh->b_data + bi->bi_offset + x; | |
1297 | diff = ~(*orig | (*orig >> 1)) & (*clone | (*clone >> 1)); | |
1298 | } else { | |
1299 | diff = ~(*clone | (*clone >> 1)); | |
1300 | } | |
f15ab561 SW |
1301 | diff &= 0x55; |
1302 | if (diff == 0) | |
1303 | continue; | |
1304 | blk = offset + ((bi->bi_start + x) * GFS2_NBBY); | |
f15ab561 SW |
1305 | while(diff) { |
1306 | if (diff & 1) { | |
b2c87cae | 1307 | if (nr_blks == 0) |
f15ab561 | 1308 | goto start_new_extent; |
b2c87cae BP |
1309 | if ((start + nr_blks) != blk) { |
1310 | if (nr_blks >= minlen) { | |
1311 | rv = sb_issue_discard(sb, | |
1312 | start, nr_blks, | |
66fc061b SW |
1313 | GFP_NOFS, 0); |
1314 | if (rv) | |
1315 | goto fail; | |
b2c87cae | 1316 | trimmed += nr_blks; |
66fc061b | 1317 | } |
b2c87cae | 1318 | nr_blks = 0; |
f15ab561 SW |
1319 | start_new_extent: |
1320 | start = blk; | |
1321 | } | |
b2c87cae | 1322 | nr_blks++; |
f15ab561 SW |
1323 | } |
1324 | diff >>= 2; | |
b2c87cae | 1325 | blk++; |
f15ab561 SW |
1326 | } |
1327 | } | |
b2c87cae BP |
1328 | if (nr_blks >= minlen) { |
1329 | rv = sb_issue_discard(sb, start, nr_blks, GFP_NOFS, 0); | |
f15ab561 SW |
1330 | if (rv) |
1331 | goto fail; | |
b2c87cae | 1332 | trimmed += nr_blks; |
f15ab561 | 1333 | } |
66fc061b SW |
1334 | if (ptrimmed) |
1335 | *ptrimmed = trimmed; | |
1336 | return 0; | |
1337 | ||
f15ab561 | 1338 | fail: |
66fc061b | 1339 | if (sdp->sd_args.ar_discard) |
af38816e | 1340 | fs_warn(sdp, "error %d on discard request, turning discards off for this filesystem\n", rv); |
f15ab561 | 1341 | sdp->sd_args.ar_discard = 0; |
66fc061b SW |
1342 | return -EIO; |
1343 | } | |
1344 | ||
1345 | /** | |
1346 | * gfs2_fitrim - Generate discard requests for unused bits of the filesystem | |
1347 | * @filp: Any file on the filesystem | |
1348 | * @argp: Pointer to the arguments (also used to pass result) | |
1349 | * | |
1350 | * Returns: 0 on success, otherwise error code | |
1351 | */ | |
1352 | ||
1353 | int gfs2_fitrim(struct file *filp, void __user *argp) | |
1354 | { | |
496ad9aa | 1355 | struct inode *inode = file_inode(filp); |
66fc061b SW |
1356 | struct gfs2_sbd *sdp = GFS2_SB(inode); |
1357 | struct request_queue *q = bdev_get_queue(sdp->sd_vfs->s_bdev); | |
1358 | struct buffer_head *bh; | |
1359 | struct gfs2_rgrpd *rgd; | |
1360 | struct gfs2_rgrpd *rgd_end; | |
1361 | struct gfs2_holder gh; | |
1362 | struct fstrim_range r; | |
1363 | int ret = 0; | |
1364 | u64 amt; | |
1365 | u64 trimmed = 0; | |
076f0faa | 1366 | u64 start, end, minlen; |
66fc061b | 1367 | unsigned int x; |
076f0faa | 1368 | unsigned bs_shift = sdp->sd_sb.sb_bsize_shift; |
66fc061b SW |
1369 | |
1370 | if (!capable(CAP_SYS_ADMIN)) | |
1371 | return -EPERM; | |
1372 | ||
1373 | if (!blk_queue_discard(q)) | |
1374 | return -EOPNOTSUPP; | |
1375 | ||
3a238ade | 1376 | if (copy_from_user(&r, argp, sizeof(r))) |
66fc061b SW |
1377 | return -EFAULT; |
1378 | ||
5e2f7d61 BP |
1379 | ret = gfs2_rindex_update(sdp); |
1380 | if (ret) | |
1381 | return ret; | |
1382 | ||
076f0faa LC |
1383 | start = r.start >> bs_shift; |
1384 | end = start + (r.len >> bs_shift); | |
1385 | minlen = max_t(u64, r.minlen, | |
1386 | q->limits.discard_granularity) >> bs_shift; | |
1387 | ||
6a98c333 AD |
1388 | if (end <= start || minlen > sdp->sd_max_rg_data) |
1389 | return -EINVAL; | |
1390 | ||
076f0faa | 1391 | rgd = gfs2_blk2rgrpd(sdp, start, 0); |
6a98c333 | 1392 | rgd_end = gfs2_blk2rgrpd(sdp, end, 0); |
076f0faa | 1393 | |
6a98c333 AD |
1394 | if ((gfs2_rgrpd_get_first(sdp) == gfs2_rgrpd_get_next(rgd_end)) |
1395 | && (start > rgd_end->rd_data0 + rgd_end->rd_data)) | |
1396 | return -EINVAL; /* start is beyond the end of the fs */ | |
66fc061b SW |
1397 | |
1398 | while (1) { | |
1399 | ||
1400 | ret = gfs2_glock_nq_init(rgd->rd_gl, LM_ST_EXCLUSIVE, 0, &gh); | |
1401 | if (ret) | |
1402 | goto out; | |
1403 | ||
1404 | if (!(rgd->rd_flags & GFS2_RGF_TRIMMED)) { | |
1405 | /* Trim each bitmap in the rgrp */ | |
1406 | for (x = 0; x < rgd->rd_length; x++) { | |
1407 | struct gfs2_bitmap *bi = rgd->rd_bits + x; | |
076f0faa LC |
1408 | ret = gfs2_rgrp_send_discards(sdp, |
1409 | rgd->rd_data0, NULL, bi, minlen, | |
1410 | &amt); | |
66fc061b SW |
1411 | if (ret) { |
1412 | gfs2_glock_dq_uninit(&gh); | |
1413 | goto out; | |
1414 | } | |
1415 | trimmed += amt; | |
1416 | } | |
1417 | ||
1418 | /* Mark rgrp as having been trimmed */ | |
1419 | ret = gfs2_trans_begin(sdp, RES_RG_HDR, 0); | |
1420 | if (ret == 0) { | |
1421 | bh = rgd->rd_bits[0].bi_bh; | |
1422 | rgd->rd_flags |= GFS2_RGF_TRIMMED; | |
350a9b0a | 1423 | gfs2_trans_add_meta(rgd->rd_gl, bh); |
66fc061b SW |
1424 | gfs2_rgrp_out(rgd, bh->b_data); |
1425 | gfs2_trans_end(sdp); | |
1426 | } | |
1427 | } | |
1428 | gfs2_glock_dq_uninit(&gh); | |
1429 | ||
1430 | if (rgd == rgd_end) | |
1431 | break; | |
1432 | ||
1433 | rgd = gfs2_rgrpd_get_next(rgd); | |
1434 | } | |
1435 | ||
1436 | out: | |
6a98c333 | 1437 | r.len = trimmed << bs_shift; |
3a238ade | 1438 | if (copy_to_user(argp, &r, sizeof(r))) |
66fc061b SW |
1439 | return -EFAULT; |
1440 | ||
1441 | return ret; | |
f15ab561 SW |
1442 | } |
1443 | ||
8e2e0047 BP |
1444 | /** |
1445 | * rs_insert - insert a new multi-block reservation into the rgrp's rb_tree | |
8e2e0047 | 1446 | * @ip: the inode structure |
8e2e0047 | 1447 | * |
8e2e0047 | 1448 | */ |
ff7f4cb4 | 1449 | static void rs_insert(struct gfs2_inode *ip) |
8e2e0047 BP |
1450 | { |
1451 | struct rb_node **newn, *parent = NULL; | |
1452 | int rc; | |
a097dc7e | 1453 | struct gfs2_blkreserv *rs = &ip->i_res; |
4a993fb1 | 1454 | struct gfs2_rgrpd *rgd = rs->rs_rbm.rgd; |
ff7f4cb4 | 1455 | u64 fsblock = gfs2_rbm_to_block(&rs->rs_rbm); |
8e2e0047 | 1456 | |
8e2e0047 | 1457 | BUG_ON(gfs2_rs_active(rs)); |
c743ffd0 | 1458 | |
ff7f4cb4 SW |
1459 | spin_lock(&rgd->rd_rsspin); |
1460 | newn = &rgd->rd_rstree.rb_node; | |
8e2e0047 BP |
1461 | while (*newn) { |
1462 | struct gfs2_blkreserv *cur = | |
1463 | rb_entry(*newn, struct gfs2_blkreserv, rs_node); | |
1464 | ||
1465 | parent = *newn; | |
ff7f4cb4 | 1466 | rc = rs_cmp(fsblock, rs->rs_free, cur); |
8e2e0047 BP |
1467 | if (rc > 0) |
1468 | newn = &((*newn)->rb_right); | |
1469 | else if (rc < 0) | |
1470 | newn = &((*newn)->rb_left); | |
1471 | else { | |
1472 | spin_unlock(&rgd->rd_rsspin); | |
ff7f4cb4 SW |
1473 | WARN_ON(1); |
1474 | return; | |
8e2e0047 BP |
1475 | } |
1476 | } | |
1477 | ||
8e2e0047 BP |
1478 | rb_link_node(&rs->rs_node, parent, newn); |
1479 | rb_insert_color(&rs->rs_node, &rgd->rd_rstree); | |
1480 | ||
8e2e0047 | 1481 | /* Do our rgrp accounting for the reservation */ |
ff7f4cb4 | 1482 | rgd->rd_reserved += rs->rs_free; /* blocks reserved */ |
8e2e0047 | 1483 | spin_unlock(&rgd->rd_rsspin); |
9e733d39 | 1484 | trace_gfs2_rs(rs, TRACE_RS_INSERT); |
8e2e0047 BP |
1485 | } |
1486 | ||
f6753df3 BP |
1487 | /** |
1488 | * rgd_free - return the number of free blocks we can allocate. | |
1489 | * @rgd: the resource group | |
1490 | * | |
1491 | * This function returns the number of free blocks for an rgrp. | |
1492 | * That's the clone-free blocks (blocks that are free, not including those | |
1493 | * still being used for unlinked files that haven't been deleted.) | |
1494 | * | |
1495 | * It also subtracts any blocks reserved by someone else, but does not | |
1496 | * include free blocks that are still part of our current reservation, | |
1497 | * because obviously we can (and will) allocate them. | |
1498 | */ | |
1499 | static inline u32 rgd_free(struct gfs2_rgrpd *rgd, struct gfs2_blkreserv *rs) | |
1500 | { | |
1501 | u32 tot_reserved, tot_free; | |
1502 | ||
1503 | if (WARN_ON_ONCE(rgd->rd_reserved < rs->rs_free)) | |
1504 | return 0; | |
1505 | tot_reserved = rgd->rd_reserved - rs->rs_free; | |
1506 | ||
1507 | if (rgd->rd_free_clone < tot_reserved) | |
1508 | tot_reserved = 0; | |
1509 | ||
1510 | tot_free = rgd->rd_free_clone - tot_reserved; | |
1511 | ||
1512 | return tot_free; | |
1513 | } | |
1514 | ||
8e2e0047 | 1515 | /** |
ff7f4cb4 | 1516 | * rg_mblk_search - find a group of multiple free blocks to form a reservation |
8e2e0047 | 1517 | * @rgd: the resource group descriptor |
8e2e0047 | 1518 | * @ip: pointer to the inode for which we're reserving blocks |
7b9cff46 | 1519 | * @ap: the allocation parameters |
8e2e0047 | 1520 | * |
8e2e0047 BP |
1521 | */ |
1522 | ||
ff7f4cb4 | 1523 | static void rg_mblk_search(struct gfs2_rgrpd *rgd, struct gfs2_inode *ip, |
7b9cff46 | 1524 | const struct gfs2_alloc_parms *ap) |
8e2e0047 | 1525 | { |
ff7f4cb4 SW |
1526 | struct gfs2_rbm rbm = { .rgd = rgd, }; |
1527 | u64 goal; | |
a097dc7e | 1528 | struct gfs2_blkreserv *rs = &ip->i_res; |
ff7f4cb4 | 1529 | u32 extlen; |
f6753df3 | 1530 | u32 free_blocks = rgd_free(rgd, rs); |
ff7f4cb4 | 1531 | int ret; |
af21ca8e | 1532 | struct inode *inode = &ip->i_inode; |
8e2e0047 | 1533 | |
af21ca8e BP |
1534 | if (S_ISDIR(inode->i_mode)) |
1535 | extlen = 1; | |
1536 | else { | |
21f09c43 | 1537 | extlen = max_t(u32, atomic_read(&ip->i_sizehint), ap->target); |
ad899458 | 1538 | extlen = clamp(extlen, (u32)RGRP_RSRV_MINBLKS, free_blocks); |
af21ca8e | 1539 | } |
ff7f4cb4 | 1540 | if ((rgd->rd_free_clone < rgd->rd_reserved) || (free_blocks < extlen)) |
c743ffd0 SW |
1541 | return; |
1542 | ||
8e2e0047 BP |
1543 | /* Find bitmap block that contains bits for goal block */ |
1544 | if (rgrp_contains_block(rgd, ip->i_goal)) | |
ff7f4cb4 | 1545 | goal = ip->i_goal; |
8e2e0047 | 1546 | else |
ff7f4cb4 | 1547 | goal = rgd->rd_last_alloc + rgd->rd_data0; |
8e2e0047 | 1548 | |
ff7f4cb4 SW |
1549 | if (WARN_ON(gfs2_rbm_from_block(&rbm, goal))) |
1550 | return; | |
8e2e0047 | 1551 | |
8381e602 | 1552 | ret = gfs2_rbm_find(&rbm, GFS2_BLKST_FREE, &extlen, ip, true); |
ff7f4cb4 SW |
1553 | if (ret == 0) { |
1554 | rs->rs_rbm = rbm; | |
1555 | rs->rs_free = extlen; | |
ff7f4cb4 | 1556 | rs_insert(ip); |
13d2eb01 BP |
1557 | } else { |
1558 | if (goal == rgd->rd_last_alloc + rgd->rd_data0) | |
1559 | rgd->rd_last_alloc = 0; | |
8e2e0047 | 1560 | } |
b3e47ca0 BP |
1561 | } |
1562 | ||
5b924ae2 SW |
1563 | /** |
1564 | * gfs2_next_unreserved_block - Return next block that is not reserved | |
1565 | * @rgd: The resource group | |
1566 | * @block: The starting block | |
ff7f4cb4 | 1567 | * @length: The required length |
5b924ae2 SW |
1568 | * @ip: Ignore any reservations for this inode |
1569 | * | |
1570 | * If the block does not appear in any reservation, then return the | |
1571 | * block number unchanged. If it does appear in the reservation, then | |
1572 | * keep looking through the tree of reservations in order to find the | |
1573 | * first block number which is not reserved. | |
1574 | */ | |
1575 | ||
1576 | static u64 gfs2_next_unreserved_block(struct gfs2_rgrpd *rgd, u64 block, | |
ff7f4cb4 | 1577 | u32 length, |
5b924ae2 SW |
1578 | const struct gfs2_inode *ip) |
1579 | { | |
1580 | struct gfs2_blkreserv *rs; | |
1581 | struct rb_node *n; | |
1582 | int rc; | |
1583 | ||
1584 | spin_lock(&rgd->rd_rsspin); | |
ff7f4cb4 | 1585 | n = rgd->rd_rstree.rb_node; |
5b924ae2 SW |
1586 | while (n) { |
1587 | rs = rb_entry(n, struct gfs2_blkreserv, rs_node); | |
ff7f4cb4 | 1588 | rc = rs_cmp(block, length, rs); |
5b924ae2 SW |
1589 | if (rc < 0) |
1590 | n = n->rb_left; | |
1591 | else if (rc > 0) | |
1592 | n = n->rb_right; | |
1593 | else | |
1594 | break; | |
1595 | } | |
1596 | ||
1597 | if (n) { | |
a097dc7e | 1598 | while ((rs_cmp(block, length, rs) == 0) && (&ip->i_res != rs)) { |
5b924ae2 | 1599 | block = gfs2_rbm_to_block(&rs->rs_rbm) + rs->rs_free; |
ff7f4cb4 | 1600 | n = n->rb_right; |
5b924ae2 SW |
1601 | if (n == NULL) |
1602 | break; | |
1603 | rs = rb_entry(n, struct gfs2_blkreserv, rs_node); | |
1604 | } | |
1605 | } | |
1606 | ||
1607 | spin_unlock(&rgd->rd_rsspin); | |
1608 | return block; | |
1609 | } | |
1610 | ||
5b924ae2 SW |
1611 | /** |
1612 | * gfs2_reservation_check_and_update - Check for reservations during block alloc | |
1613 | * @rbm: The current position in the resource group | |
ff7f4cb4 SW |
1614 | * @ip: The inode for which we are searching for blocks |
1615 | * @minext: The minimum extent length | |
5ce13431 | 1616 | * @maxext: A pointer to the maximum extent structure |
5b924ae2 SW |
1617 | * |
1618 | * This checks the current position in the rgrp to see whether there is | |
1619 | * a reservation covering this block. If not then this function is a | |
1620 | * no-op. If there is, then the position is moved to the end of the | |
1621 | * contiguous reservation(s) so that we are pointing at the first | |
1622 | * non-reserved block. | |
1623 | * | |
1624 | * Returns: 0 if no reservation, 1 if @rbm has changed, otherwise an error | |
1625 | */ | |
1626 | ||
1627 | static int gfs2_reservation_check_and_update(struct gfs2_rbm *rbm, | |
ff7f4cb4 | 1628 | const struct gfs2_inode *ip, |
5ce13431 BP |
1629 | u32 minext, |
1630 | struct gfs2_extent *maxext) | |
5b924ae2 SW |
1631 | { |
1632 | u64 block = gfs2_rbm_to_block(rbm); | |
ff7f4cb4 | 1633 | u32 extlen = 1; |
5b924ae2 SW |
1634 | u64 nblock; |
1635 | int ret; | |
1636 | ||
ff7f4cb4 SW |
1637 | /* |
1638 | * If we have a minimum extent length, then skip over any extent | |
1639 | * which is less than the min extent length in size. | |
1640 | */ | |
1641 | if (minext) { | |
1642 | extlen = gfs2_free_extlen(rbm, minext); | |
5ce13431 | 1643 | if (extlen <= maxext->len) |
ff7f4cb4 SW |
1644 | goto fail; |
1645 | } | |
1646 | ||
1647 | /* | |
1648 | * Check the extent which has been found against the reservations | |
1649 | * and skip if parts of it are already reserved | |
1650 | */ | |
1651 | nblock = gfs2_next_unreserved_block(rbm->rgd, block, extlen, ip); | |
5ce13431 BP |
1652 | if (nblock == block) { |
1653 | if (!minext || extlen >= minext) | |
1654 | return 0; | |
1655 | ||
1656 | if (extlen > maxext->len) { | |
1657 | maxext->len = extlen; | |
1658 | maxext->rbm = *rbm; | |
1659 | } | |
ff7f4cb4 | 1660 | fail: |
5ce13431 BP |
1661 | nblock = block + extlen; |
1662 | } | |
5b924ae2 SW |
1663 | ret = gfs2_rbm_from_block(rbm, nblock); |
1664 | if (ret < 0) | |
1665 | return ret; | |
1666 | return 1; | |
1667 | } | |
1668 | ||
1669 | /** | |
1670 | * gfs2_rbm_find - Look for blocks of a particular state | |
1671 | * @rbm: Value/result starting position and final position | |
1672 | * @state: The state which we want to find | |
5ce13431 BP |
1673 | * @minext: Pointer to the requested extent length (NULL for a single block) |
1674 | * This is updated to be the actual reservation size. | |
5b924ae2 SW |
1675 | * @ip: If set, check for reservations |
1676 | * @nowrap: Stop looking at the end of the rgrp, rather than wrapping | |
1677 | * around until we've reached the starting point. | |
1678 | * | |
1679 | * Side effects: | |
1680 | * - If looking for free blocks, we set GBF_FULL on each bitmap which | |
1681 | * has no free blocks in it. | |
5ea5050c BP |
1682 | * - If looking for free blocks, we set rd_extfail_pt on each rgrp which |
1683 | * has come up short on a free block search. | |
5b924ae2 SW |
1684 | * |
1685 | * Returns: 0 on success, -ENOSPC if there is no block of the requested state | |
1686 | */ | |
1687 | ||
5ce13431 | 1688 | static int gfs2_rbm_find(struct gfs2_rbm *rbm, u8 state, u32 *minext, |
8381e602 | 1689 | const struct gfs2_inode *ip, bool nowrap) |
5b924ae2 | 1690 | { |
71921ef8 | 1691 | bool scan_from_start = rbm->bii == 0 && rbm->offset == 0; |
5b924ae2 | 1692 | struct buffer_head *bh; |
71921ef8 | 1693 | int last_bii; |
5b924ae2 SW |
1694 | u32 offset; |
1695 | u8 *buffer; | |
71921ef8 | 1696 | bool wrapped = false; |
5b924ae2 | 1697 | int ret; |
e579ed4f | 1698 | struct gfs2_bitmap *bi; |
5ce13431 | 1699 | struct gfs2_extent maxext = { .rbm.rgd = rbm->rgd, }; |
5b924ae2 | 1700 | |
71921ef8 AG |
1701 | /* |
1702 | * Determine the last bitmap to search. If we're not starting at the | |
1703 | * beginning of a bitmap, we need to search that bitmap twice to scan | |
1704 | * the entire resource group. | |
5b924ae2 | 1705 | */ |
71921ef8 | 1706 | last_bii = rbm->bii - (rbm->offset == 0); |
5b924ae2 SW |
1707 | |
1708 | while(1) { | |
e579ed4f | 1709 | bi = rbm_bi(rbm); |
e79e0e14 BP |
1710 | if ((ip == NULL || !gfs2_rs_active(&ip->i_res)) && |
1711 | test_bit(GBF_FULL, &bi->bi_flags) && | |
5b924ae2 SW |
1712 | (state == GFS2_BLKST_FREE)) |
1713 | goto next_bitmap; | |
1714 | ||
e579ed4f BP |
1715 | bh = bi->bi_bh; |
1716 | buffer = bh->b_data + bi->bi_offset; | |
5b924ae2 | 1717 | WARN_ON(!buffer_uptodate(bh)); |
e579ed4f BP |
1718 | if (state != GFS2_BLKST_UNLINKED && bi->bi_clone) |
1719 | buffer = bi->bi_clone + bi->bi_offset; | |
281b4952 | 1720 | offset = gfs2_bitfit(buffer, bi->bi_bytes, rbm->offset, state); |
71921ef8 AG |
1721 | if (offset == BFITNOENT) { |
1722 | if (state == GFS2_BLKST_FREE && rbm->offset == 0) | |
1723 | set_bit(GBF_FULL, &bi->bi_flags); | |
1724 | goto next_bitmap; | |
1725 | } | |
5b924ae2 SW |
1726 | rbm->offset = offset; |
1727 | if (ip == NULL) | |
1728 | return 0; | |
1729 | ||
5ce13431 BP |
1730 | ret = gfs2_reservation_check_and_update(rbm, ip, |
1731 | minext ? *minext : 0, | |
1732 | &maxext); | |
5b924ae2 SW |
1733 | if (ret == 0) |
1734 | return 0; | |
71921ef8 | 1735 | if (ret > 0) |
8d8b752a | 1736 | goto next_iter; |
5d50d532 | 1737 | if (ret == -E2BIG) { |
e579ed4f | 1738 | rbm->bii = 0; |
5d50d532 | 1739 | rbm->offset = 0; |
5d50d532 SW |
1740 | goto res_covered_end_of_rgrp; |
1741 | } | |
5b924ae2 SW |
1742 | return ret; |
1743 | ||
5b924ae2 SW |
1744 | next_bitmap: /* Find next bitmap in the rgrp */ |
1745 | rbm->offset = 0; | |
e579ed4f BP |
1746 | rbm->bii++; |
1747 | if (rbm->bii == rbm->rgd->rd_length) | |
1748 | rbm->bii = 0; | |
5d50d532 | 1749 | res_covered_end_of_rgrp: |
71921ef8 AG |
1750 | if (rbm->bii == 0) { |
1751 | if (wrapped) | |
1752 | break; | |
1753 | wrapped = true; | |
1754 | if (nowrap) | |
1755 | break; | |
1756 | } | |
8d8b752a | 1757 | next_iter: |
71921ef8 AG |
1758 | /* Have we scanned the entire resource group? */ |
1759 | if (wrapped && rbm->bii > last_bii) | |
5b924ae2 SW |
1760 | break; |
1761 | } | |
1762 | ||
5ce13431 BP |
1763 | if (minext == NULL || state != GFS2_BLKST_FREE) |
1764 | return -ENOSPC; | |
1765 | ||
5ea5050c BP |
1766 | /* If the extent was too small, and it's smaller than the smallest |
1767 | to have failed before, remember for future reference that it's | |
1768 | useless to search this rgrp again for this amount or more. */ | |
71921ef8 AG |
1769 | if (wrapped && (scan_from_start || rbm->bii > last_bii) && |
1770 | *minext < rbm->rgd->rd_extfail_pt) | |
5ea5050c BP |
1771 | rbm->rgd->rd_extfail_pt = *minext; |
1772 | ||
5ce13431 BP |
1773 | /* If the maximum extent we found is big enough to fulfill the |
1774 | minimum requirements, use it anyway. */ | |
1775 | if (maxext.len) { | |
1776 | *rbm = maxext.rbm; | |
1777 | *minext = maxext.len; | |
1778 | return 0; | |
1779 | } | |
1780 | ||
5b924ae2 SW |
1781 | return -ENOSPC; |
1782 | } | |
1783 | ||
c8cdf479 SW |
1784 | /** |
1785 | * try_rgrp_unlink - Look for any unlinked, allocated, but unused inodes | |
1786 | * @rgd: The rgrp | |
886b1416 BP |
1787 | * @last_unlinked: block address of the last dinode we unlinked |
1788 | * @skip: block address we should explicitly not unlink | |
c8cdf479 | 1789 | * |
1a0eae88 BP |
1790 | * Returns: 0 if no error |
1791 | * The inode, if one has been found, in inode. | |
c8cdf479 SW |
1792 | */ |
1793 | ||
044b9414 | 1794 | static void try_rgrp_unlink(struct gfs2_rgrpd *rgd, u64 *last_unlinked, u64 skip) |
c8cdf479 | 1795 | { |
5b924ae2 | 1796 | u64 block; |
5f3eae75 | 1797 | struct gfs2_sbd *sdp = rgd->rd_sbd; |
044b9414 SW |
1798 | struct gfs2_glock *gl; |
1799 | struct gfs2_inode *ip; | |
1800 | int error; | |
1801 | int found = 0; | |
e579ed4f | 1802 | struct gfs2_rbm rbm = { .rgd = rgd, .bii = 0, .offset = 0 }; |
c8cdf479 | 1803 | |
5b924ae2 | 1804 | while (1) { |
5ce13431 | 1805 | error = gfs2_rbm_find(&rbm, GFS2_BLKST_UNLINKED, NULL, NULL, |
8381e602 | 1806 | true); |
5b924ae2 SW |
1807 | if (error == -ENOSPC) |
1808 | break; | |
1809 | if (WARN_ON_ONCE(error)) | |
24c73873 | 1810 | break; |
b3e47ca0 | 1811 | |
5b924ae2 SW |
1812 | block = gfs2_rbm_to_block(&rbm); |
1813 | if (gfs2_rbm_from_block(&rbm, block + 1)) | |
1814 | break; | |
1815 | if (*last_unlinked != NO_BLOCK && block <= *last_unlinked) | |
c8cdf479 | 1816 | continue; |
5b924ae2 | 1817 | if (block == skip) |
1e19a195 | 1818 | continue; |
5b924ae2 | 1819 | *last_unlinked = block; |
044b9414 | 1820 | |
5ea31bc0 | 1821 | error = gfs2_glock_get(sdp, block, &gfs2_iopen_glops, CREATE, &gl); |
044b9414 SW |
1822 | if (error) |
1823 | continue; | |
1824 | ||
1825 | /* If the inode is already in cache, we can ignore it here | |
1826 | * because the existing inode disposal code will deal with | |
1827 | * it when all refs have gone away. Accessing gl_object like | |
1828 | * this is not safe in general. Here it is ok because we do | |
1829 | * not dereference the pointer, and we only need an approx | |
1830 | * answer to whether it is NULL or not. | |
1831 | */ | |
1832 | ip = gl->gl_object; | |
1833 | ||
a0e3cc65 | 1834 | if (ip || !gfs2_queue_delete_work(gl, 0)) |
044b9414 SW |
1835 | gfs2_glock_put(gl); |
1836 | else | |
1837 | found++; | |
1838 | ||
1839 | /* Limit reclaim to sensible number of tasks */ | |
44ad37d6 | 1840 | if (found > NR_CPUS) |
044b9414 | 1841 | return; |
c8cdf479 SW |
1842 | } |
1843 | ||
1844 | rgd->rd_flags &= ~GFS2_RDF_CHECK; | |
044b9414 | 1845 | return; |
c8cdf479 SW |
1846 | } |
1847 | ||
bcd97c06 SW |
1848 | /** |
1849 | * gfs2_rgrp_congested - Use stats to figure out whether an rgrp is congested | |
1850 | * @rgd: The rgrp in question | |
1851 | * @loops: An indication of how picky we can be (0=very, 1=less so) | |
1852 | * | |
1853 | * This function uses the recently added glock statistics in order to | |
1854 | * figure out whether a parciular resource group is suffering from | |
1855 | * contention from multiple nodes. This is done purely on the basis | |
1856 | * of timings, since this is the only data we have to work with and | |
1857 | * our aim here is to reject a resource group which is highly contended | |
1858 | * but (very important) not to do this too often in order to ensure that | |
1859 | * we do not land up introducing fragmentation by changing resource | |
1860 | * groups when not actually required. | |
1861 | * | |
1862 | * The calculation is fairly simple, we want to know whether the SRTTB | |
1863 | * (i.e. smoothed round trip time for blocking operations) to acquire | |
1864 | * the lock for this rgrp's glock is significantly greater than the | |
1865 | * time taken for resource groups on average. We introduce a margin in | |
1866 | * the form of the variable @var which is computed as the sum of the two | |
1867 | * respective variences, and multiplied by a factor depending on @loops | |
1868 | * and whether we have a lot of data to base the decision on. This is | |
1869 | * then tested against the square difference of the means in order to | |
1870 | * decide whether the result is statistically significant or not. | |
1871 | * | |
1872 | * Returns: A boolean verdict on the congestion status | |
1873 | */ | |
1874 | ||
1875 | static bool gfs2_rgrp_congested(const struct gfs2_rgrpd *rgd, int loops) | |
1876 | { | |
1877 | const struct gfs2_glock *gl = rgd->rd_gl; | |
15562c43 | 1878 | const struct gfs2_sbd *sdp = gl->gl_name.ln_sbd; |
bcd97c06 | 1879 | struct gfs2_lkstats *st; |
4d207133 BH |
1880 | u64 r_dcount, l_dcount; |
1881 | u64 l_srttb, a_srttb = 0; | |
bcd97c06 | 1882 | s64 srttb_diff; |
4d207133 BH |
1883 | u64 sqr_diff; |
1884 | u64 var; | |
0166b197 | 1885 | int cpu, nonzero = 0; |
bcd97c06 SW |
1886 | |
1887 | preempt_disable(); | |
f4a3ae93 BP |
1888 | for_each_present_cpu(cpu) { |
1889 | st = &per_cpu_ptr(sdp->sd_lkstats, cpu)->lkstats[LM_TYPE_RGRP]; | |
0166b197 BP |
1890 | if (st->stats[GFS2_LKS_SRTTB]) { |
1891 | a_srttb += st->stats[GFS2_LKS_SRTTB]; | |
1892 | nonzero++; | |
1893 | } | |
f4a3ae93 | 1894 | } |
bcd97c06 | 1895 | st = &this_cpu_ptr(sdp->sd_lkstats)->lkstats[LM_TYPE_RGRP]; |
0166b197 BP |
1896 | if (nonzero) |
1897 | do_div(a_srttb, nonzero); | |
bcd97c06 SW |
1898 | r_dcount = st->stats[GFS2_LKS_DCOUNT]; |
1899 | var = st->stats[GFS2_LKS_SRTTVARB] + | |
1900 | gl->gl_stats.stats[GFS2_LKS_SRTTVARB]; | |
1901 | preempt_enable(); | |
1902 | ||
1903 | l_srttb = gl->gl_stats.stats[GFS2_LKS_SRTTB]; | |
1904 | l_dcount = gl->gl_stats.stats[GFS2_LKS_DCOUNT]; | |
1905 | ||
f4a3ae93 | 1906 | if ((l_dcount < 1) || (r_dcount < 1) || (a_srttb == 0)) |
bcd97c06 SW |
1907 | return false; |
1908 | ||
f4a3ae93 | 1909 | srttb_diff = a_srttb - l_srttb; |
bcd97c06 SW |
1910 | sqr_diff = srttb_diff * srttb_diff; |
1911 | ||
1912 | var *= 2; | |
1913 | if (l_dcount < 8 || r_dcount < 8) | |
1914 | var *= 2; | |
1915 | if (loops == 1) | |
1916 | var *= 2; | |
1917 | ||
1918 | return ((srttb_diff < 0) && (sqr_diff > var)); | |
1919 | } | |
1920 | ||
1921 | /** | |
1922 | * gfs2_rgrp_used_recently | |
1923 | * @rs: The block reservation with the rgrp to test | |
1924 | * @msecs: The time limit in milliseconds | |
1925 | * | |
1926 | * Returns: True if the rgrp glock has been used within the time limit | |
1927 | */ | |
1928 | static bool gfs2_rgrp_used_recently(const struct gfs2_blkreserv *rs, | |
1929 | u64 msecs) | |
1930 | { | |
1931 | u64 tdiff; | |
1932 | ||
1933 | tdiff = ktime_to_ns(ktime_sub(ktime_get_real(), | |
1934 | rs->rs_rbm.rgd->rd_gl->gl_dstamp)); | |
1935 | ||
1936 | return tdiff > (msecs * 1000 * 1000); | |
1937 | } | |
1938 | ||
9dbe9610 SW |
1939 | static u32 gfs2_orlov_skip(const struct gfs2_inode *ip) |
1940 | { | |
1941 | const struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode); | |
1942 | u32 skip; | |
1943 | ||
1944 | get_random_bytes(&skip, sizeof(skip)); | |
1945 | return skip % sdp->sd_rgrps; | |
1946 | } | |
1947 | ||
c743ffd0 SW |
1948 | static bool gfs2_select_rgrp(struct gfs2_rgrpd **pos, const struct gfs2_rgrpd *begin) |
1949 | { | |
1950 | struct gfs2_rgrpd *rgd = *pos; | |
aa8920c9 | 1951 | struct gfs2_sbd *sdp = rgd->rd_sbd; |
c743ffd0 SW |
1952 | |
1953 | rgd = gfs2_rgrpd_get_next(rgd); | |
1954 | if (rgd == NULL) | |
aa8920c9 | 1955 | rgd = gfs2_rgrpd_get_first(sdp); |
c743ffd0 SW |
1956 | *pos = rgd; |
1957 | if (rgd != begin) /* If we didn't wrap */ | |
1958 | return true; | |
1959 | return false; | |
1960 | } | |
1961 | ||
0e27c18c BP |
1962 | /** |
1963 | * fast_to_acquire - determine if a resource group will be fast to acquire | |
1964 | * | |
1965 | * If this is one of our preferred rgrps, it should be quicker to acquire, | |
1966 | * because we tried to set ourselves up as dlm lock master. | |
1967 | */ | |
1968 | static inline int fast_to_acquire(struct gfs2_rgrpd *rgd) | |
1969 | { | |
1970 | struct gfs2_glock *gl = rgd->rd_gl; | |
1971 | ||
1972 | if (gl->gl_state != LM_ST_UNLOCKED && list_empty(&gl->gl_holders) && | |
1973 | !test_bit(GLF_DEMOTE_IN_PROGRESS, &gl->gl_flags) && | |
1974 | !test_bit(GLF_DEMOTE, &gl->gl_flags)) | |
1975 | return 1; | |
1976 | if (rgd->rd_flags & GFS2_RDF_PREFERRED) | |
1977 | return 1; | |
1978 | return 0; | |
1979 | } | |
1980 | ||
b3b94faa | 1981 | /** |
666d1d8a | 1982 | * gfs2_inplace_reserve - Reserve space in the filesystem |
b3b94faa | 1983 | * @ip: the inode to reserve space for |
7b9cff46 | 1984 | * @ap: the allocation parameters |
b3b94faa | 1985 | * |
25435e5e AD |
1986 | * We try our best to find an rgrp that has at least ap->target blocks |
1987 | * available. After a couple of passes (loops == 2), the prospects of finding | |
1988 | * such an rgrp diminish. At this stage, we return the first rgrp that has | |
243fea4d | 1989 | * at least ap->min_target blocks available. Either way, we set ap->allowed to |
25435e5e AD |
1990 | * the number of blocks available in the chosen rgrp. |
1991 | * | |
1992 | * Returns: 0 on success, | |
1993 | * -ENOMEM if a suitable rgrp can't be found | |
1994 | * errno otherwise | |
b3b94faa DT |
1995 | */ |
1996 | ||
25435e5e | 1997 | int gfs2_inplace_reserve(struct gfs2_inode *ip, struct gfs2_alloc_parms *ap) |
b3b94faa | 1998 | { |
feaa7bba | 1999 | struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode); |
8e2e0047 | 2000 | struct gfs2_rgrpd *begin = NULL; |
a097dc7e | 2001 | struct gfs2_blkreserv *rs = &ip->i_res; |
bcd97c06 | 2002 | int error = 0, rg_locked, flags = 0; |
666d1d8a | 2003 | u64 last_unlinked = NO_BLOCK; |
7c9ca621 | 2004 | int loops = 0; |
f6753df3 | 2005 | u32 free_blocks, skip = 0; |
b3b94faa | 2006 | |
90306c41 BM |
2007 | if (sdp->sd_args.ar_rgrplvb) |
2008 | flags |= GL_SKIP; | |
7b9cff46 | 2009 | if (gfs2_assert_warn(sdp, ap->target)) |
c743ffd0 | 2010 | return -EINVAL; |
8e2e0047 | 2011 | if (gfs2_rs_active(rs)) { |
4a993fb1 | 2012 | begin = rs->rs_rbm.rgd; |
b7eba890 AG |
2013 | } else if (rs->rs_rbm.rgd && |
2014 | rgrp_contains_block(rs->rs_rbm.rgd, ip->i_goal)) { | |
2015 | begin = rs->rs_rbm.rgd; | |
8e2e0047 | 2016 | } else { |
00a158be | 2017 | check_and_update_goal(ip); |
4a993fb1 | 2018 | rs->rs_rbm.rgd = begin = gfs2_blk2rgrpd(sdp, ip->i_goal, 1); |
8e2e0047 | 2019 | } |
7b9cff46 | 2020 | if (S_ISDIR(ip->i_inode.i_mode) && (ap->aflags & GFS2_AF_ORLOV)) |
9dbe9610 | 2021 | skip = gfs2_orlov_skip(ip); |
4a993fb1 | 2022 | if (rs->rs_rbm.rgd == NULL) |
7c9ca621 BP |
2023 | return -EBADSLT; |
2024 | ||
2025 | while (loops < 3) { | |
c743ffd0 SW |
2026 | rg_locked = 1; |
2027 | ||
2028 | if (!gfs2_glock_is_locked_by_me(rs->rs_rbm.rgd->rd_gl)) { | |
2029 | rg_locked = 0; | |
9dbe9610 SW |
2030 | if (skip && skip--) |
2031 | goto next_rgrp; | |
0e27c18c BP |
2032 | if (!gfs2_rs_active(rs)) { |
2033 | if (loops == 0 && | |
2034 | !fast_to_acquire(rs->rs_rbm.rgd)) | |
2035 | goto next_rgrp; | |
2036 | if ((loops < 2) && | |
2037 | gfs2_rgrp_used_recently(rs, 1000) && | |
2038 | gfs2_rgrp_congested(rs->rs_rbm.rgd, loops)) | |
2039 | goto next_rgrp; | |
2040 | } | |
4a993fb1 | 2041 | error = gfs2_glock_nq_init(rs->rs_rbm.rgd->rd_gl, |
8e2e0047 | 2042 | LM_ST_EXCLUSIVE, flags, |
21f09c43 | 2043 | &ip->i_rgd_gh); |
c743ffd0 SW |
2044 | if (unlikely(error)) |
2045 | return error; | |
bcd97c06 SW |
2046 | if (!gfs2_rs_active(rs) && (loops < 2) && |
2047 | gfs2_rgrp_congested(rs->rs_rbm.rgd, loops)) | |
2048 | goto skip_rgrp; | |
c743ffd0 | 2049 | if (sdp->sd_args.ar_rgrplvb) { |
4a993fb1 | 2050 | error = update_rgrp_lvb(rs->rs_rbm.rgd); |
c743ffd0 | 2051 | if (unlikely(error)) { |
21f09c43 | 2052 | gfs2_glock_dq_uninit(&ip->i_rgd_gh); |
90306c41 BM |
2053 | return error; |
2054 | } | |
2055 | } | |
292c8c14 | 2056 | } |
666d1d8a | 2057 | |
243fea4d | 2058 | /* Skip unusable resource groups */ |
5ea5050c BP |
2059 | if ((rs->rs_rbm.rgd->rd_flags & (GFS2_RGF_NOALLOC | |
2060 | GFS2_RDF_ERROR)) || | |
25435e5e | 2061 | (loops == 0 && ap->target > rs->rs_rbm.rgd->rd_extfail_pt)) |
c743ffd0 SW |
2062 | goto skip_rgrp; |
2063 | ||
2064 | if (sdp->sd_args.ar_rgrplvb) | |
2065 | gfs2_rgrp_bh_get(rs->rs_rbm.rgd); | |
2066 | ||
2067 | /* Get a reservation if we don't already have one */ | |
2068 | if (!gfs2_rs_active(rs)) | |
7b9cff46 | 2069 | rg_mblk_search(rs->rs_rbm.rgd, ip, ap); |
c743ffd0 SW |
2070 | |
2071 | /* Skip rgrps when we can't get a reservation on first pass */ | |
2072 | if (!gfs2_rs_active(rs) && (loops < 1)) | |
2073 | goto check_rgrp; | |
2074 | ||
2075 | /* If rgrp has enough free space, use it */ | |
f6753df3 BP |
2076 | free_blocks = rgd_free(rs->rs_rbm.rgd, rs); |
2077 | if (free_blocks >= ap->target || | |
25435e5e | 2078 | (loops == 2 && ap->min_target && |
f6753df3 BP |
2079 | free_blocks >= ap->min_target)) { |
2080 | ap->allowed = free_blocks; | |
c743ffd0 | 2081 | return 0; |
b3b94faa | 2082 | } |
c743ffd0 SW |
2083 | check_rgrp: |
2084 | /* Check for unlinked inodes which can be reclaimed */ | |
2085 | if (rs->rs_rbm.rgd->rd_flags & GFS2_RDF_CHECK) | |
2086 | try_rgrp_unlink(rs->rs_rbm.rgd, &last_unlinked, | |
2087 | ip->i_no_addr); | |
2088 | skip_rgrp: | |
1330edbe BP |
2089 | /* Drop reservation, if we couldn't use reserved rgrp */ |
2090 | if (gfs2_rs_active(rs)) | |
2091 | gfs2_rs_deltree(rs); | |
2092 | ||
c743ffd0 SW |
2093 | /* Unlock rgrp if required */ |
2094 | if (!rg_locked) | |
21f09c43 | 2095 | gfs2_glock_dq_uninit(&ip->i_rgd_gh); |
c743ffd0 SW |
2096 | next_rgrp: |
2097 | /* Find the next rgrp, and continue looking */ | |
2098 | if (gfs2_select_rgrp(&rs->rs_rbm.rgd, begin)) | |
2099 | continue; | |
9dbe9610 SW |
2100 | if (skip) |
2101 | continue; | |
c743ffd0 SW |
2102 | |
2103 | /* If we've scanned all the rgrps, but found no free blocks | |
2104 | * then this checks for some less likely conditions before | |
2105 | * trying again. | |
2106 | */ | |
c743ffd0 SW |
2107 | loops++; |
2108 | /* Check that fs hasn't grown if writing to rindex */ | |
2109 | if (ip == GFS2_I(sdp->sd_rindex) && !sdp->sd_rindex_uptodate) { | |
2110 | error = gfs2_ri_update(ip); | |
2111 | if (error) | |
2112 | return error; | |
2113 | } | |
2114 | /* Flushing the log may release space */ | |
2115 | if (loops == 2) | |
805c0907 BP |
2116 | gfs2_log_flush(sdp, NULL, GFS2_LOG_HEAD_FLUSH_NORMAL | |
2117 | GFS2_LFC_INPLACE_RESERVE); | |
b3b94faa | 2118 | } |
b3b94faa | 2119 | |
c743ffd0 | 2120 | return -ENOSPC; |
b3b94faa DT |
2121 | } |
2122 | ||
2123 | /** | |
2124 | * gfs2_inplace_release - release an inplace reservation | |
2125 | * @ip: the inode the reservation was taken out on | |
2126 | * | |
2127 | * Release a reservation made by gfs2_inplace_reserve(). | |
2128 | */ | |
2129 | ||
2130 | void gfs2_inplace_release(struct gfs2_inode *ip) | |
2131 | { | |
21f09c43 AG |
2132 | if (gfs2_holder_initialized(&ip->i_rgd_gh)) |
2133 | gfs2_glock_dq_uninit(&ip->i_rgd_gh); | |
b3b94faa DT |
2134 | } |
2135 | ||
b3e47ca0 BP |
2136 | /** |
2137 | * gfs2_alloc_extent - allocate an extent from a given bitmap | |
4a993fb1 | 2138 | * @rbm: the resource group information |
b3e47ca0 | 2139 | * @dinode: TRUE if the first block we allocate is for a dinode |
c04a2ef3 | 2140 | * @n: The extent length (value/result) |
b3e47ca0 | 2141 | * |
c04a2ef3 | 2142 | * Add the bitmap buffer to the transaction. |
b3e47ca0 | 2143 | * Set the found bits to @new_state to change block's allocation state. |
b3e47ca0 | 2144 | */ |
c04a2ef3 | 2145 | static void gfs2_alloc_extent(const struct gfs2_rbm *rbm, bool dinode, |
4a993fb1 | 2146 | unsigned int *n) |
b3e47ca0 | 2147 | { |
c04a2ef3 | 2148 | struct gfs2_rbm pos = { .rgd = rbm->rgd, }; |
b3e47ca0 | 2149 | const unsigned int elen = *n; |
c04a2ef3 SW |
2150 | u64 block; |
2151 | int ret; | |
b3e47ca0 | 2152 | |
c04a2ef3 SW |
2153 | *n = 1; |
2154 | block = gfs2_rbm_to_block(rbm); | |
e579ed4f | 2155 | gfs2_trans_add_meta(rbm->rgd->rd_gl, rbm_bi(rbm)->bi_bh); |
3e6339dd | 2156 | gfs2_setbit(rbm, true, dinode ? GFS2_BLKST_DINODE : GFS2_BLKST_USED); |
c04a2ef3 | 2157 | block++; |
60a0b8f9 | 2158 | while (*n < elen) { |
c04a2ef3 | 2159 | ret = gfs2_rbm_from_block(&pos, block); |
dffe12a8 | 2160 | if (ret || gfs2_testbit(&pos, true) != GFS2_BLKST_FREE) |
60a0b8f9 | 2161 | break; |
e579ed4f | 2162 | gfs2_trans_add_meta(pos.rgd->rd_gl, rbm_bi(&pos)->bi_bh); |
3e6339dd | 2163 | gfs2_setbit(&pos, true, GFS2_BLKST_USED); |
60a0b8f9 | 2164 | (*n)++; |
c04a2ef3 | 2165 | block++; |
c8cdf479 | 2166 | } |
b3b94faa DT |
2167 | } |
2168 | ||
2169 | /** | |
2170 | * rgblk_free - Change alloc state of given block(s) | |
2171 | * @sdp: the filesystem | |
0ddeded4 | 2172 | * @rgd: the resource group the blocks are in |
b3b94faa DT |
2173 | * @bstart: the start of a run of blocks to free |
2174 | * @blen: the length of the block run (all must lie within ONE RG!) | |
2175 | * @new_state: GFS2_BLKST_XXX the after-allocation block state | |
b3b94faa DT |
2176 | */ |
2177 | ||
0ddeded4 AG |
2178 | static void rgblk_free(struct gfs2_sbd *sdp, struct gfs2_rgrpd *rgd, |
2179 | u64 bstart, u32 blen, unsigned char new_state) | |
b3b94faa | 2180 | { |
3b1d0b9d | 2181 | struct gfs2_rbm rbm; |
d24e0569 | 2182 | struct gfs2_bitmap *bi, *bi_prev = NULL; |
b3b94faa | 2183 | |
0ddeded4 | 2184 | rbm.rgd = rgd; |
f654683d | 2185 | if (WARN_ON_ONCE(gfs2_rbm_from_block(&rbm, bstart))) |
0ddeded4 | 2186 | return; |
b3b94faa | 2187 | while (blen--) { |
e579ed4f | 2188 | bi = rbm_bi(&rbm); |
d24e0569 BP |
2189 | if (bi != bi_prev) { |
2190 | if (!bi->bi_clone) { | |
2191 | bi->bi_clone = kmalloc(bi->bi_bh->b_size, | |
2192 | GFP_NOFS | __GFP_NOFAIL); | |
2193 | memcpy(bi->bi_clone + bi->bi_offset, | |
2194 | bi->bi_bh->b_data + bi->bi_offset, | |
281b4952 | 2195 | bi->bi_bytes); |
d24e0569 BP |
2196 | } |
2197 | gfs2_trans_add_meta(rbm.rgd->rd_gl, bi->bi_bh); | |
2198 | bi_prev = bi; | |
b3b94faa | 2199 | } |
3e6339dd | 2200 | gfs2_setbit(&rbm, false, new_state); |
d24e0569 | 2201 | gfs2_rbm_incr(&rbm); |
b3b94faa | 2202 | } |
b3b94faa DT |
2203 | } |
2204 | ||
2205 | /** | |
09010978 SW |
2206 | * gfs2_rgrp_dump - print out an rgrp |
2207 | * @seq: The iterator | |
0e539ca1 | 2208 | * @rgd: The rgrp in question |
3792ce97 | 2209 | * @fs_id_buf: pointer to file system id (if requested) |
09010978 SW |
2210 | * |
2211 | */ | |
2212 | ||
0e539ca1 | 2213 | void gfs2_rgrp_dump(struct seq_file *seq, struct gfs2_rgrpd *rgd, |
3792ce97 | 2214 | const char *fs_id_buf) |
09010978 | 2215 | { |
8e2e0047 BP |
2216 | struct gfs2_blkreserv *trs; |
2217 | const struct rb_node *n; | |
2218 | ||
3792ce97 BP |
2219 | gfs2_print_dbg(seq, "%s R: n:%llu f:%02x b:%u/%u i:%u r:%u e:%u\n", |
2220 | fs_id_buf, | |
09010978 | 2221 | (unsigned long long)rgd->rd_addr, rgd->rd_flags, |
8e2e0047 | 2222 | rgd->rd_free, rgd->rd_free_clone, rgd->rd_dinodes, |
5ea5050c | 2223 | rgd->rd_reserved, rgd->rd_extfail_pt); |
72244b6b BP |
2224 | if (rgd->rd_sbd->sd_args.ar_rgrplvb) { |
2225 | struct gfs2_rgrp_lvb *rgl = rgd->rd_rgl; | |
2226 | ||
3792ce97 | 2227 | gfs2_print_dbg(seq, "%s L: f:%02x b:%u i:%u\n", fs_id_buf, |
72244b6b BP |
2228 | be32_to_cpu(rgl->rl_flags), |
2229 | be32_to_cpu(rgl->rl_free), | |
2230 | be32_to_cpu(rgl->rl_dinodes)); | |
2231 | } | |
8e2e0047 BP |
2232 | spin_lock(&rgd->rd_rsspin); |
2233 | for (n = rb_first(&rgd->rd_rstree); n; n = rb_next(&trs->rs_node)) { | |
2234 | trs = rb_entry(n, struct gfs2_blkreserv, rs_node); | |
3792ce97 | 2235 | dump_rs(seq, trs, fs_id_buf); |
8e2e0047 BP |
2236 | } |
2237 | spin_unlock(&rgd->rd_rsspin); | |
09010978 SW |
2238 | } |
2239 | ||
6050b9c7 SW |
2240 | static void gfs2_rgrp_error(struct gfs2_rgrpd *rgd) |
2241 | { | |
2242 | struct gfs2_sbd *sdp = rgd->rd_sbd; | |
98fb0574 | 2243 | char fs_id_buf[sizeof(sdp->sd_fsname) + 7]; |
3792ce97 | 2244 | |
6050b9c7 | 2245 | fs_warn(sdp, "rgrp %llu has an error, marking it readonly until umount\n", |
86d00636 | 2246 | (unsigned long long)rgd->rd_addr); |
6050b9c7 | 2247 | fs_warn(sdp, "umount on all nodes and run fsck.gfs2 to fix the error\n"); |
3792ce97 | 2248 | sprintf(fs_id_buf, "fsid=%s: ", sdp->sd_fsname); |
0e539ca1 | 2249 | gfs2_rgrp_dump(NULL, rgd, fs_id_buf); |
6050b9c7 SW |
2250 | rgd->rd_flags |= GFS2_RDF_ERROR; |
2251 | } | |
2252 | ||
8e2e0047 | 2253 | /** |
5b924ae2 SW |
2254 | * gfs2_adjust_reservation - Adjust (or remove) a reservation after allocation |
2255 | * @ip: The inode we have just allocated blocks for | |
2256 | * @rbm: The start of the allocated blocks | |
2257 | * @len: The extent length | |
8e2e0047 | 2258 | * |
5b924ae2 SW |
2259 | * Adjusts a reservation after an allocation has taken place. If the |
2260 | * reservation does not match the allocation, or if it is now empty | |
2261 | * then it is removed. | |
8e2e0047 | 2262 | */ |
5b924ae2 SW |
2263 | |
2264 | static void gfs2_adjust_reservation(struct gfs2_inode *ip, | |
2265 | const struct gfs2_rbm *rbm, unsigned len) | |
8e2e0047 | 2266 | { |
a097dc7e | 2267 | struct gfs2_blkreserv *rs = &ip->i_res; |
5b924ae2 SW |
2268 | struct gfs2_rgrpd *rgd = rbm->rgd; |
2269 | unsigned rlen; | |
2270 | u64 block; | |
2271 | int ret; | |
8e2e0047 | 2272 | |
5b924ae2 SW |
2273 | spin_lock(&rgd->rd_rsspin); |
2274 | if (gfs2_rs_active(rs)) { | |
2275 | if (gfs2_rbm_eq(&rs->rs_rbm, rbm)) { | |
2276 | block = gfs2_rbm_to_block(rbm); | |
2277 | ret = gfs2_rbm_from_block(&rs->rs_rbm, block + len); | |
2278 | rlen = min(rs->rs_free, len); | |
2279 | rs->rs_free -= rlen; | |
2280 | rgd->rd_reserved -= rlen; | |
9e733d39 | 2281 | trace_gfs2_rs(rs, TRACE_RS_CLAIM); |
5b924ae2 SW |
2282 | if (rs->rs_free && !ret) |
2283 | goto out; | |
1a855033 BP |
2284 | /* We used up our block reservation, so we should |
2285 | reserve more blocks next time. */ | |
21f09c43 | 2286 | atomic_add(RGRP_RSRV_ADDBLKS, &ip->i_sizehint); |
5b924ae2 | 2287 | } |
20095218 | 2288 | __rs_deltree(rs); |
8e2e0047 | 2289 | } |
5b924ae2 SW |
2290 | out: |
2291 | spin_unlock(&rgd->rd_rsspin); | |
8e2e0047 BP |
2292 | } |
2293 | ||
9e07f2cb SW |
2294 | /** |
2295 | * gfs2_set_alloc_start - Set starting point for block allocation | |
2296 | * @rbm: The rbm which will be set to the required location | |
2297 | * @ip: The gfs2 inode | |
2298 | * @dinode: Flag to say if allocation includes a new inode | |
2299 | * | |
2300 | * This sets the starting point from the reservation if one is active | |
2301 | * otherwise it falls back to guessing a start point based on the | |
2302 | * inode's goal block or the last allocation point in the rgrp. | |
2303 | */ | |
2304 | ||
2305 | static void gfs2_set_alloc_start(struct gfs2_rbm *rbm, | |
2306 | const struct gfs2_inode *ip, bool dinode) | |
2307 | { | |
2308 | u64 goal; | |
2309 | ||
a097dc7e BP |
2310 | if (gfs2_rs_active(&ip->i_res)) { |
2311 | *rbm = ip->i_res.rs_rbm; | |
9e07f2cb SW |
2312 | return; |
2313 | } | |
2314 | ||
2315 | if (!dinode && rgrp_contains_block(rbm->rgd, ip->i_goal)) | |
2316 | goal = ip->i_goal; | |
2317 | else | |
2318 | goal = rbm->rgd->rd_last_alloc + rbm->rgd->rd_data0; | |
2319 | ||
f654683d AG |
2320 | if (WARN_ON_ONCE(gfs2_rbm_from_block(rbm, goal))) { |
2321 | rbm->bii = 0; | |
2322 | rbm->offset = 0; | |
2323 | } | |
9e07f2cb SW |
2324 | } |
2325 | ||
09010978 | 2326 | /** |
6e87ed0f | 2327 | * gfs2_alloc_blocks - Allocate one or more blocks of data and/or a dinode |
1639431a | 2328 | * @ip: the inode to allocate the block for |
09010978 | 2329 | * @bn: Used to return the starting block number |
8e2e0047 | 2330 | * @nblocks: requested number of blocks/extent length (value/result) |
6e87ed0f | 2331 | * @dinode: 1 if we're allocating a dinode block, else 0 |
3c5d785a | 2332 | * @generation: the generation number of the inode |
b3b94faa | 2333 | * |
09010978 | 2334 | * Returns: 0 or error |
b3b94faa DT |
2335 | */ |
2336 | ||
6a8099ed | 2337 | int gfs2_alloc_blocks(struct gfs2_inode *ip, u64 *bn, unsigned int *nblocks, |
6e87ed0f | 2338 | bool dinode, u64 *generation) |
b3b94faa | 2339 | { |
feaa7bba | 2340 | struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode); |
d9ba7615 | 2341 | struct buffer_head *dibh; |
b7eba890 | 2342 | struct gfs2_rbm rbm = { .rgd = ip->i_res.rs_rbm.rgd, }; |
6a8099ed | 2343 | unsigned int ndata; |
3c5d785a | 2344 | u64 block; /* block, within the file system scope */ |
d9ba7615 | 2345 | int error; |
b3b94faa | 2346 | |
9e07f2cb | 2347 | gfs2_set_alloc_start(&rbm, ip, dinode); |
8381e602 | 2348 | error = gfs2_rbm_find(&rbm, GFS2_BLKST_FREE, NULL, ip, false); |
62e252ee | 2349 | |
137834a6 | 2350 | if (error == -ENOSPC) { |
9e07f2cb | 2351 | gfs2_set_alloc_start(&rbm, ip, dinode); |
8381e602 | 2352 | error = gfs2_rbm_find(&rbm, GFS2_BLKST_FREE, NULL, NULL, false); |
137834a6 SW |
2353 | } |
2354 | ||
62e252ee | 2355 | /* Since all blocks are reserved in advance, this shouldn't happen */ |
5b924ae2 | 2356 | if (error) { |
5ea5050c | 2357 | fs_warn(sdp, "inum=%llu error=%d, nblocks=%u, full=%d fail_pt=%d\n", |
9e733d39 | 2358 | (unsigned long long)ip->i_no_addr, error, *nblocks, |
5ea5050c BP |
2359 | test_bit(GBF_FULL, &rbm.rgd->rd_bits->bi_flags), |
2360 | rbm.rgd->rd_extfail_pt); | |
62e252ee | 2361 | goto rgrp_error; |
8e2e0047 | 2362 | } |
62e252ee | 2363 | |
c04a2ef3 SW |
2364 | gfs2_alloc_extent(&rbm, dinode, nblocks); |
2365 | block = gfs2_rbm_to_block(&rbm); | |
c743ffd0 | 2366 | rbm.rgd->rd_last_alloc = block - rbm.rgd->rd_data0; |
a097dc7e | 2367 | if (gfs2_rs_active(&ip->i_res)) |
5b924ae2 | 2368 | gfs2_adjust_reservation(ip, &rbm, *nblocks); |
6a8099ed SW |
2369 | ndata = *nblocks; |
2370 | if (dinode) | |
2371 | ndata--; | |
b3e47ca0 | 2372 | |
3c5d785a | 2373 | if (!dinode) { |
6a8099ed | 2374 | ip->i_goal = block + ndata - 1; |
3c5d785a BP |
2375 | error = gfs2_meta_inode_buffer(ip, &dibh); |
2376 | if (error == 0) { | |
2377 | struct gfs2_dinode *di = | |
2378 | (struct gfs2_dinode *)dibh->b_data; | |
350a9b0a | 2379 | gfs2_trans_add_meta(ip->i_gl, dibh); |
3c5d785a BP |
2380 | di->di_goal_meta = di->di_goal_data = |
2381 | cpu_to_be64(ip->i_goal); | |
2382 | brelse(dibh); | |
2383 | } | |
d9ba7615 | 2384 | } |
4a993fb1 | 2385 | if (rbm.rgd->rd_free < *nblocks) { |
e54c78a2 | 2386 | fs_warn(sdp, "nblocks=%u\n", *nblocks); |
09010978 | 2387 | goto rgrp_error; |
8e2e0047 | 2388 | } |
09010978 | 2389 | |
4a993fb1 | 2390 | rbm.rgd->rd_free -= *nblocks; |
3c5d785a | 2391 | if (dinode) { |
4a993fb1 SW |
2392 | rbm.rgd->rd_dinodes++; |
2393 | *generation = rbm.rgd->rd_igeneration++; | |
3c5d785a | 2394 | if (*generation == 0) |
4a993fb1 | 2395 | *generation = rbm.rgd->rd_igeneration++; |
3c5d785a | 2396 | } |
b3b94faa | 2397 | |
350a9b0a | 2398 | gfs2_trans_add_meta(rbm.rgd->rd_gl, rbm.rgd->rd_bits[0].bi_bh); |
4a993fb1 | 2399 | gfs2_rgrp_out(rbm.rgd, rbm.rgd->rd_bits[0].bi_bh->b_data); |
b3b94faa | 2400 | |
6a8099ed | 2401 | gfs2_statfs_change(sdp, 0, -(s64)*nblocks, dinode ? 1 : 0); |
3c5d785a | 2402 | if (dinode) |
fbb27873 | 2403 | gfs2_trans_remove_revoke(sdp, block, *nblocks); |
6a8099ed | 2404 | |
fd4b4e04 | 2405 | gfs2_quota_change(ip, *nblocks, ip->i_inode.i_uid, ip->i_inode.i_gid); |
b3b94faa | 2406 | |
4a993fb1 SW |
2407 | rbm.rgd->rd_free_clone -= *nblocks; |
2408 | trace_gfs2_block_alloc(ip, rbm.rgd, block, *nblocks, | |
6e87ed0f | 2409 | dinode ? GFS2_BLKST_DINODE : GFS2_BLKST_USED); |
6050b9c7 SW |
2410 | *bn = block; |
2411 | return 0; | |
2412 | ||
2413 | rgrp_error: | |
4a993fb1 | 2414 | gfs2_rgrp_error(rbm.rgd); |
6050b9c7 | 2415 | return -EIO; |
b3b94faa DT |
2416 | } |
2417 | ||
2418 | /** | |
46fcb2ed | 2419 | * __gfs2_free_blocks - free a contiguous run of block(s) |
b3b94faa | 2420 | * @ip: the inode these blocks are being freed from |
0ddeded4 | 2421 | * @rgd: the resource group the blocks are in |
b3b94faa DT |
2422 | * @bstart: first block of a run of contiguous blocks |
2423 | * @blen: the length of the block run | |
46fcb2ed | 2424 | * @meta: 1 if the blocks represent metadata |
b3b94faa DT |
2425 | * |
2426 | */ | |
2427 | ||
0ddeded4 AG |
2428 | void __gfs2_free_blocks(struct gfs2_inode *ip, struct gfs2_rgrpd *rgd, |
2429 | u64 bstart, u32 blen, int meta) | |
b3b94faa | 2430 | { |
feaa7bba | 2431 | struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode); |
b3b94faa | 2432 | |
0ddeded4 | 2433 | rgblk_free(sdp, rgd, bstart, blen, GFS2_BLKST_FREE); |
41db1ab9 | 2434 | trace_gfs2_block_alloc(ip, rgd, bstart, blen, GFS2_BLKST_FREE); |
cfc8b549 | 2435 | rgd->rd_free += blen; |
66fc061b | 2436 | rgd->rd_flags &= ~GFS2_RGF_TRIMMED; |
350a9b0a | 2437 | gfs2_trans_add_meta(rgd->rd_gl, rgd->rd_bits[0].bi_bh); |
42d52e38 | 2438 | gfs2_rgrp_out(rgd, rgd->rd_bits[0].bi_bh->b_data); |
b3b94faa | 2439 | |
6d3117b4 | 2440 | /* Directories keep their data in the metadata address space */ |
68942870 BP |
2441 | if (meta || ip->i_depth || gfs2_is_jdata(ip)) |
2442 | gfs2_journal_wipe(ip, bstart, blen); | |
4c16c36a | 2443 | } |
b3b94faa | 2444 | |
4c16c36a BP |
2445 | /** |
2446 | * gfs2_free_meta - free a contiguous run of data block(s) | |
2447 | * @ip: the inode these blocks are being freed from | |
0ddeded4 | 2448 | * @rgd: the resource group the blocks are in |
4c16c36a BP |
2449 | * @bstart: first block of a run of contiguous blocks |
2450 | * @blen: the length of the block run | |
2451 | * | |
2452 | */ | |
2453 | ||
0ddeded4 AG |
2454 | void gfs2_free_meta(struct gfs2_inode *ip, struct gfs2_rgrpd *rgd, |
2455 | u64 bstart, u32 blen) | |
4c16c36a BP |
2456 | { |
2457 | struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode); | |
2458 | ||
0ddeded4 | 2459 | __gfs2_free_blocks(ip, rgd, bstart, blen, 1); |
b3b94faa | 2460 | gfs2_statfs_change(sdp, 0, +blen, 0); |
2933f925 | 2461 | gfs2_quota_change(ip, -(s64)blen, ip->i_inode.i_uid, ip->i_inode.i_gid); |
b3b94faa DT |
2462 | } |
2463 | ||
feaa7bba SW |
2464 | void gfs2_unlink_di(struct inode *inode) |
2465 | { | |
2466 | struct gfs2_inode *ip = GFS2_I(inode); | |
2467 | struct gfs2_sbd *sdp = GFS2_SB(inode); | |
2468 | struct gfs2_rgrpd *rgd; | |
dbb7cae2 | 2469 | u64 blkno = ip->i_no_addr; |
feaa7bba | 2470 | |
0ddeded4 | 2471 | rgd = gfs2_blk2rgrpd(sdp, blkno, true); |
feaa7bba SW |
2472 | if (!rgd) |
2473 | return; | |
0ddeded4 | 2474 | rgblk_free(sdp, rgd, blkno, 1, GFS2_BLKST_UNLINKED); |
41db1ab9 | 2475 | trace_gfs2_block_alloc(ip, rgd, blkno, 1, GFS2_BLKST_UNLINKED); |
350a9b0a | 2476 | gfs2_trans_add_meta(rgd->rd_gl, rgd->rd_bits[0].bi_bh); |
42d52e38 | 2477 | gfs2_rgrp_out(rgd, rgd->rd_bits[0].bi_bh->b_data); |
f5580d0f | 2478 | be32_add_cpu(&rgd->rd_rgl->rl_unlinked, 1); |
feaa7bba SW |
2479 | } |
2480 | ||
a18c78c5 | 2481 | void gfs2_free_di(struct gfs2_rgrpd *rgd, struct gfs2_inode *ip) |
b3b94faa DT |
2482 | { |
2483 | struct gfs2_sbd *sdp = rgd->rd_sbd; | |
b3b94faa | 2484 | |
0ddeded4 | 2485 | rgblk_free(sdp, rgd, ip->i_no_addr, 1, GFS2_BLKST_FREE); |
73f74948 | 2486 | if (!rgd->rd_dinodes) |
b3b94faa | 2487 | gfs2_consist_rgrpd(rgd); |
73f74948 | 2488 | rgd->rd_dinodes--; |
cfc8b549 | 2489 | rgd->rd_free++; |
b3b94faa | 2490 | |
350a9b0a | 2491 | gfs2_trans_add_meta(rgd->rd_gl, rgd->rd_bits[0].bi_bh); |
42d52e38 | 2492 | gfs2_rgrp_out(rgd, rgd->rd_bits[0].bi_bh->b_data); |
f5580d0f | 2493 | be32_add_cpu(&rgd->rd_rgl->rl_unlinked, -1); |
b3b94faa DT |
2494 | |
2495 | gfs2_statfs_change(sdp, 0, +1, -1); | |
41db1ab9 | 2496 | trace_gfs2_block_alloc(ip, rgd, ip->i_no_addr, 1, GFS2_BLKST_FREE); |
2933f925 | 2497 | gfs2_quota_change(ip, -1, ip->i_inode.i_uid, ip->i_inode.i_gid); |
68942870 | 2498 | gfs2_journal_wipe(ip, ip->i_no_addr, 1); |
b3b94faa DT |
2499 | } |
2500 | ||
acf7e244 SW |
2501 | /** |
2502 | * gfs2_check_blk_type - Check the type of a block | |
2503 | * @sdp: The superblock | |
2504 | * @no_addr: The block number to check | |
2505 | * @type: The block type we are looking for | |
2506 | * | |
2507 | * Returns: 0 if the block type matches the expected type | |
2508 | * -ESTALE if it doesn't match | |
2509 | * or -ve errno if something went wrong while checking | |
2510 | */ | |
2511 | ||
2512 | int gfs2_check_blk_type(struct gfs2_sbd *sdp, u64 no_addr, unsigned int type) | |
2513 | { | |
2514 | struct gfs2_rgrpd *rgd; | |
8339ee54 | 2515 | struct gfs2_holder rgd_gh; |
dffe12a8 | 2516 | struct gfs2_rbm rbm; |
58884c4d | 2517 | int error = -EINVAL; |
acf7e244 | 2518 | |
66fc061b | 2519 | rgd = gfs2_blk2rgrpd(sdp, no_addr, 1); |
acf7e244 | 2520 | if (!rgd) |
8339ee54 | 2521 | goto fail; |
acf7e244 SW |
2522 | |
2523 | error = gfs2_glock_nq_init(rgd->rd_gl, LM_ST_SHARED, 0, &rgd_gh); | |
2524 | if (error) | |
8339ee54 | 2525 | goto fail; |
acf7e244 | 2526 | |
dffe12a8 BP |
2527 | rbm.rgd = rgd; |
2528 | error = gfs2_rbm_from_block(&rbm, no_addr); | |
f654683d AG |
2529 | if (WARN_ON_ONCE(error)) |
2530 | goto fail; | |
dffe12a8 BP |
2531 | |
2532 | if (gfs2_testbit(&rbm, false) != type) | |
acf7e244 SW |
2533 | error = -ESTALE; |
2534 | ||
2535 | gfs2_glock_dq_uninit(&rgd_gh); | |
acf7e244 SW |
2536 | fail: |
2537 | return error; | |
2538 | } | |
2539 | ||
b3b94faa DT |
2540 | /** |
2541 | * gfs2_rlist_add - add a RG to a list of RGs | |
70b0c365 | 2542 | * @ip: the inode |
b3b94faa DT |
2543 | * @rlist: the list of resource groups |
2544 | * @block: the block | |
2545 | * | |
2546 | * Figure out what RG a block belongs to and add that RG to the list | |
2547 | * | |
2548 | * FIXME: Don't use NOFAIL | |
2549 | * | |
2550 | */ | |
2551 | ||
70b0c365 | 2552 | void gfs2_rlist_add(struct gfs2_inode *ip, struct gfs2_rgrp_list *rlist, |
cd915493 | 2553 | u64 block) |
b3b94faa | 2554 | { |
70b0c365 | 2555 | struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode); |
b3b94faa DT |
2556 | struct gfs2_rgrpd *rgd; |
2557 | struct gfs2_rgrpd **tmp; | |
2558 | unsigned int new_space; | |
2559 | unsigned int x; | |
2560 | ||
2561 | if (gfs2_assert_warn(sdp, !rlist->rl_ghs)) | |
2562 | return; | |
2563 | ||
03f8c41c AG |
2564 | /* |
2565 | * The resource group last accessed is kept in the last position. | |
2566 | */ | |
2567 | ||
2568 | if (rlist->rl_rgrps) { | |
2569 | rgd = rlist->rl_rgd[rlist->rl_rgrps - 1]; | |
2570 | if (rgrp_contains_block(rgd, block)) | |
2571 | return; | |
66fc061b | 2572 | rgd = gfs2_blk2rgrpd(sdp, block, 1); |
03f8c41c | 2573 | } else { |
b7eba890 | 2574 | rgd = ip->i_res.rs_rbm.rgd; |
03f8c41c AG |
2575 | if (!rgd || !rgrp_contains_block(rgd, block)) |
2576 | rgd = gfs2_blk2rgrpd(sdp, block, 1); | |
2577 | } | |
2578 | ||
b3b94faa | 2579 | if (!rgd) { |
03f8c41c AG |
2580 | fs_err(sdp, "rlist_add: no rgrp for block %llu\n", |
2581 | (unsigned long long)block); | |
b3b94faa DT |
2582 | return; |
2583 | } | |
2584 | ||
03f8c41c AG |
2585 | for (x = 0; x < rlist->rl_rgrps; x++) { |
2586 | if (rlist->rl_rgd[x] == rgd) { | |
2587 | swap(rlist->rl_rgd[x], | |
2588 | rlist->rl_rgd[rlist->rl_rgrps - 1]); | |
b3b94faa | 2589 | return; |
03f8c41c AG |
2590 | } |
2591 | } | |
b3b94faa DT |
2592 | |
2593 | if (rlist->rl_rgrps == rlist->rl_space) { | |
2594 | new_space = rlist->rl_space + 10; | |
2595 | ||
2596 | tmp = kcalloc(new_space, sizeof(struct gfs2_rgrpd *), | |
dd894be8 | 2597 | GFP_NOFS | __GFP_NOFAIL); |
b3b94faa DT |
2598 | |
2599 | if (rlist->rl_rgd) { | |
2600 | memcpy(tmp, rlist->rl_rgd, | |
2601 | rlist->rl_space * sizeof(struct gfs2_rgrpd *)); | |
2602 | kfree(rlist->rl_rgd); | |
2603 | } | |
2604 | ||
2605 | rlist->rl_space = new_space; | |
2606 | rlist->rl_rgd = tmp; | |
2607 | } | |
2608 | ||
2609 | rlist->rl_rgd[rlist->rl_rgrps++] = rgd; | |
2610 | } | |
2611 | ||
2612 | /** | |
2613 | * gfs2_rlist_alloc - all RGs have been added to the rlist, now allocate | |
2614 | * and initialize an array of glock holders for them | |
2615 | * @rlist: the list of resource groups | |
b3b94faa DT |
2616 | * |
2617 | * FIXME: Don't use NOFAIL | |
2618 | * | |
2619 | */ | |
2620 | ||
c3abc29e | 2621 | void gfs2_rlist_alloc(struct gfs2_rgrp_list *rlist) |
b3b94faa DT |
2622 | { |
2623 | unsigned int x; | |
2624 | ||
6da2ec56 KC |
2625 | rlist->rl_ghs = kmalloc_array(rlist->rl_rgrps, |
2626 | sizeof(struct gfs2_holder), | |
2627 | GFP_NOFS | __GFP_NOFAIL); | |
b3b94faa DT |
2628 | for (x = 0; x < rlist->rl_rgrps; x++) |
2629 | gfs2_holder_init(rlist->rl_rgd[x]->rd_gl, | |
c3abc29e | 2630 | LM_ST_EXCLUSIVE, 0, |
b3b94faa DT |
2631 | &rlist->rl_ghs[x]); |
2632 | } | |
2633 | ||
2634 | /** | |
2635 | * gfs2_rlist_free - free a resource group list | |
27ff6a0f | 2636 | * @rlist: the list of resource groups |
b3b94faa DT |
2637 | * |
2638 | */ | |
2639 | ||
2640 | void gfs2_rlist_free(struct gfs2_rgrp_list *rlist) | |
2641 | { | |
2642 | unsigned int x; | |
2643 | ||
2644 | kfree(rlist->rl_rgd); | |
2645 | ||
2646 | if (rlist->rl_ghs) { | |
2647 | for (x = 0; x < rlist->rl_rgrps; x++) | |
2648 | gfs2_holder_uninit(&rlist->rl_ghs[x]); | |
2649 | kfree(rlist->rl_ghs); | |
8e2e0047 | 2650 | rlist->rl_ghs = NULL; |
b3b94faa DT |
2651 | } |
2652 | } | |
2653 |