]>
Commit | Line | Data |
---|---|---|
1da177e4 LT |
1 | /* |
2 | * Copyright (c) 2000-2002 Silicon Graphics, Inc. All Rights Reserved. | |
3 | * | |
4 | * This program is free software; you can redistribute it and/or modify it | |
5 | * under the terms of version 2 of the GNU General Public License as | |
6 | * published by the Free Software Foundation. | |
7 | * | |
8 | * This program is distributed in the hope that it would be useful, but | |
9 | * WITHOUT ANY WARRANTY; without even the implied warranty of | |
10 | * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. | |
11 | * | |
12 | * Further, this software is distributed without any warranty that it is | |
13 | * free of the rightful claim of any third person regarding infringement | |
14 | * or the like. Any license provided herein, whether implied or | |
15 | * otherwise, applies only to this software file. Patent licenses, if | |
16 | * any, provided herein do not apply to combinations of this program with | |
17 | * other software, or any other product whatsoever. | |
18 | * | |
19 | * You should have received a copy of the GNU General Public License along | |
20 | * with this program; if not, write the Free Software Foundation, Inc., 59 | |
21 | * Temple Place - Suite 330, Boston MA 02111-1307, USA. | |
22 | * | |
23 | * Contact information: Silicon Graphics, Inc., 1600 Amphitheatre Pkwy, | |
24 | * Mountain View, CA 94043, or: | |
25 | * | |
26 | * http://www.sgi.com | |
27 | * | |
28 | * For further information regarding this notice, see: | |
29 | * | |
30 | * http://oss.sgi.com/projects/GenInfo/SGIGPLNoticeExplan/ | |
31 | */ | |
1da177e4 | 32 | #include "xfs.h" |
a844f451 | 33 | #include "xfs_fs.h" |
1da177e4 | 34 | #include "xfs_types.h" |
1da177e4 | 35 | #include "xfs_log.h" |
a844f451 | 36 | #include "xfs_inum.h" |
1da177e4 LT |
37 | #include "xfs_trans.h" |
38 | #include "xfs_sb.h" | |
39 | #include "xfs_dir.h" | |
40 | #include "xfs_dmapi.h" | |
41 | #include "xfs_mount.h" | |
42 | #include "xfs_trans_priv.h" | |
43 | #include "xfs_error.h" | |
44 | ||
45 | STATIC void xfs_ail_insert(xfs_ail_entry_t *, xfs_log_item_t *); | |
46 | STATIC xfs_log_item_t * xfs_ail_delete(xfs_ail_entry_t *, xfs_log_item_t *); | |
47 | STATIC xfs_log_item_t * xfs_ail_min(xfs_ail_entry_t *); | |
48 | STATIC xfs_log_item_t * xfs_ail_next(xfs_ail_entry_t *, xfs_log_item_t *); | |
49 | ||
50 | #ifdef DEBUG | |
51 | STATIC void xfs_ail_check(xfs_ail_entry_t *); | |
52 | #else | |
53 | #define xfs_ail_check(a) | |
54 | #endif /* DEBUG */ | |
55 | ||
56 | ||
57 | /* | |
58 | * This is called by the log manager code to determine the LSN | |
59 | * of the tail of the log. This is exactly the LSN of the first | |
60 | * item in the AIL. If the AIL is empty, then this function | |
61 | * returns 0. | |
62 | * | |
63 | * We need the AIL lock in order to get a coherent read of the | |
64 | * lsn of the last item in the AIL. | |
65 | */ | |
66 | xfs_lsn_t | |
67 | xfs_trans_tail_ail( | |
68 | xfs_mount_t *mp) | |
69 | { | |
70 | xfs_lsn_t lsn; | |
71 | xfs_log_item_t *lip; | |
72 | SPLDECL(s); | |
73 | ||
74 | AIL_LOCK(mp,s); | |
75 | lip = xfs_ail_min(&(mp->m_ail)); | |
76 | if (lip == NULL) { | |
77 | lsn = (xfs_lsn_t)0; | |
78 | } else { | |
79 | lsn = lip->li_lsn; | |
80 | } | |
81 | AIL_UNLOCK(mp, s); | |
82 | ||
83 | return lsn; | |
84 | } | |
85 | ||
86 | /* | |
87 | * xfs_trans_push_ail | |
88 | * | |
89 | * This routine is called to move the tail of the AIL | |
90 | * forward. It does this by trying to flush items in the AIL | |
91 | * whose lsns are below the given threshold_lsn. | |
92 | * | |
93 | * The routine returns the lsn of the tail of the log. | |
94 | */ | |
95 | xfs_lsn_t | |
96 | xfs_trans_push_ail( | |
97 | xfs_mount_t *mp, | |
98 | xfs_lsn_t threshold_lsn) | |
99 | { | |
100 | xfs_lsn_t lsn; | |
101 | xfs_log_item_t *lip; | |
102 | int gen; | |
103 | int restarts; | |
104 | int lock_result; | |
105 | int flush_log; | |
106 | SPLDECL(s); | |
107 | ||
108 | #define XFS_TRANS_PUSH_AIL_RESTARTS 10 | |
109 | ||
110 | AIL_LOCK(mp,s); | |
111 | lip = xfs_trans_first_ail(mp, &gen); | |
112 | if (lip == NULL || XFS_FORCED_SHUTDOWN(mp)) { | |
113 | /* | |
114 | * Just return if the AIL is empty. | |
115 | */ | |
116 | AIL_UNLOCK(mp, s); | |
117 | return (xfs_lsn_t)0; | |
118 | } | |
119 | ||
120 | XFS_STATS_INC(xs_push_ail); | |
121 | ||
122 | /* | |
123 | * While the item we are looking at is below the given threshold | |
124 | * try to flush it out. Make sure to limit the number of times | |
125 | * we allow xfs_trans_next_ail() to restart scanning from the | |
126 | * beginning of the list. We'd like not to stop until we've at least | |
127 | * tried to push on everything in the AIL with an LSN less than | |
128 | * the given threshold. However, we may give up before that if | |
129 | * we realize that we've been holding the AIL_LOCK for 'too long', | |
130 | * blocking interrupts. Currently, too long is < 500us roughly. | |
131 | */ | |
132 | flush_log = 0; | |
133 | restarts = 0; | |
134 | while (((restarts < XFS_TRANS_PUSH_AIL_RESTARTS) && | |
135 | (XFS_LSN_CMP(lip->li_lsn, threshold_lsn) < 0))) { | |
136 | /* | |
137 | * If we can lock the item without sleeping, unlock | |
138 | * the AIL lock and flush the item. Then re-grab the | |
139 | * AIL lock so we can look for the next item on the | |
140 | * AIL. Since we unlock the AIL while we flush the | |
141 | * item, the next routine may start over again at the | |
142 | * the beginning of the list if anything has changed. | |
143 | * That is what the generation count is for. | |
144 | * | |
145 | * If we can't lock the item, either its holder will flush | |
146 | * it or it is already being flushed or it is being relogged. | |
147 | * In any of these case it is being taken care of and we | |
148 | * can just skip to the next item in the list. | |
149 | */ | |
150 | lock_result = IOP_TRYLOCK(lip); | |
151 | switch (lock_result) { | |
152 | case XFS_ITEM_SUCCESS: | |
153 | AIL_UNLOCK(mp, s); | |
154 | XFS_STATS_INC(xs_push_ail_success); | |
155 | IOP_PUSH(lip); | |
156 | AIL_LOCK(mp,s); | |
157 | break; | |
158 | ||
159 | case XFS_ITEM_PUSHBUF: | |
160 | AIL_UNLOCK(mp, s); | |
161 | XFS_STATS_INC(xs_push_ail_pushbuf); | |
162 | #ifdef XFSRACEDEBUG | |
163 | delay_for_intr(); | |
164 | delay(300); | |
165 | #endif | |
166 | ASSERT(lip->li_ops->iop_pushbuf); | |
167 | ASSERT(lip); | |
168 | IOP_PUSHBUF(lip); | |
169 | AIL_LOCK(mp,s); | |
170 | break; | |
171 | ||
172 | case XFS_ITEM_PINNED: | |
173 | XFS_STATS_INC(xs_push_ail_pinned); | |
174 | flush_log = 1; | |
175 | break; | |
176 | ||
177 | case XFS_ITEM_LOCKED: | |
178 | XFS_STATS_INC(xs_push_ail_locked); | |
179 | break; | |
180 | ||
181 | case XFS_ITEM_FLUSHING: | |
182 | XFS_STATS_INC(xs_push_ail_flushing); | |
183 | break; | |
184 | ||
185 | default: | |
186 | ASSERT(0); | |
187 | break; | |
188 | } | |
189 | ||
190 | lip = xfs_trans_next_ail(mp, lip, &gen, &restarts); | |
191 | if (lip == NULL) { | |
192 | break; | |
193 | } | |
194 | if (XFS_FORCED_SHUTDOWN(mp)) { | |
195 | /* | |
196 | * Just return if we shut down during the last try. | |
197 | */ | |
198 | AIL_UNLOCK(mp, s); | |
199 | return (xfs_lsn_t)0; | |
200 | } | |
201 | ||
202 | } | |
203 | ||
204 | if (flush_log) { | |
205 | /* | |
206 | * If something we need to push out was pinned, then | |
207 | * push out the log so it will become unpinned and | |
208 | * move forward in the AIL. | |
209 | */ | |
210 | AIL_UNLOCK(mp, s); | |
211 | XFS_STATS_INC(xs_push_ail_flush); | |
212 | xfs_log_force(mp, (xfs_lsn_t)0, XFS_LOG_FORCE); | |
213 | AIL_LOCK(mp, s); | |
214 | } | |
215 | ||
216 | lip = xfs_ail_min(&(mp->m_ail)); | |
217 | if (lip == NULL) { | |
218 | lsn = (xfs_lsn_t)0; | |
219 | } else { | |
220 | lsn = lip->li_lsn; | |
221 | } | |
222 | ||
223 | AIL_UNLOCK(mp, s); | |
224 | return lsn; | |
225 | } /* xfs_trans_push_ail */ | |
226 | ||
227 | ||
228 | /* | |
229 | * This is to be called when an item is unlocked that may have | |
230 | * been in the AIL. It will wake up the first member of the AIL | |
231 | * wait list if this item's unlocking might allow it to progress. | |
232 | * If the item is in the AIL, then we need to get the AIL lock | |
233 | * while doing our checking so we don't race with someone going | |
234 | * to sleep waiting for this event in xfs_trans_push_ail(). | |
235 | */ | |
236 | void | |
237 | xfs_trans_unlocked_item( | |
238 | xfs_mount_t *mp, | |
239 | xfs_log_item_t *lip) | |
240 | { | |
241 | xfs_log_item_t *min_lip; | |
242 | ||
243 | /* | |
244 | * If we're forcibly shutting down, we may have | |
245 | * unlocked log items arbitrarily. The last thing | |
246 | * we want to do is to move the tail of the log | |
247 | * over some potentially valid data. | |
248 | */ | |
249 | if (!(lip->li_flags & XFS_LI_IN_AIL) || | |
250 | XFS_FORCED_SHUTDOWN(mp)) { | |
251 | return; | |
252 | } | |
253 | ||
254 | /* | |
255 | * This is the one case where we can call into xfs_ail_min() | |
256 | * without holding the AIL lock because we only care about the | |
257 | * case where we are at the tail of the AIL. If the object isn't | |
258 | * at the tail, it doesn't matter what result we get back. This | |
259 | * is slightly racy because since we were just unlocked, we could | |
260 | * go to sleep between the call to xfs_ail_min and the call to | |
261 | * xfs_log_move_tail, have someone else lock us, commit to us disk, | |
262 | * move us out of the tail of the AIL, and then we wake up. However, | |
263 | * the call to xfs_log_move_tail() doesn't do anything if there's | |
264 | * not enough free space to wake people up so we're safe calling it. | |
265 | */ | |
266 | min_lip = xfs_ail_min(&mp->m_ail); | |
267 | ||
268 | if (min_lip == lip) | |
269 | xfs_log_move_tail(mp, 1); | |
270 | } /* xfs_trans_unlocked_item */ | |
271 | ||
272 | ||
273 | /* | |
274 | * Update the position of the item in the AIL with the new | |
275 | * lsn. If it is not yet in the AIL, add it. Otherwise, move | |
276 | * it to its new position by removing it and re-adding it. | |
277 | * | |
278 | * Wakeup anyone with an lsn less than the item's lsn. If the item | |
279 | * we move in the AIL is the minimum one, update the tail lsn in the | |
280 | * log manager. | |
281 | * | |
282 | * Increment the AIL's generation count to indicate that the tree | |
283 | * has changed. | |
284 | * | |
285 | * This function must be called with the AIL lock held. The lock | |
286 | * is dropped before returning, so the caller must pass in the | |
287 | * cookie returned by AIL_LOCK. | |
288 | */ | |
289 | void | |
290 | xfs_trans_update_ail( | |
291 | xfs_mount_t *mp, | |
292 | xfs_log_item_t *lip, | |
293 | xfs_lsn_t lsn, | |
294 | unsigned long s) | |
295 | { | |
296 | xfs_ail_entry_t *ailp; | |
297 | xfs_log_item_t *dlip=NULL; | |
298 | xfs_log_item_t *mlip; /* ptr to minimum lip */ | |
299 | ||
300 | ailp = &(mp->m_ail); | |
301 | mlip = xfs_ail_min(ailp); | |
302 | ||
303 | if (lip->li_flags & XFS_LI_IN_AIL) { | |
304 | dlip = xfs_ail_delete(ailp, lip); | |
305 | ASSERT(dlip == lip); | |
306 | } else { | |
307 | lip->li_flags |= XFS_LI_IN_AIL; | |
308 | } | |
309 | ||
310 | lip->li_lsn = lsn; | |
311 | ||
312 | xfs_ail_insert(ailp, lip); | |
313 | mp->m_ail_gen++; | |
314 | ||
315 | if (mlip == dlip) { | |
316 | mlip = xfs_ail_min(&(mp->m_ail)); | |
317 | AIL_UNLOCK(mp, s); | |
318 | xfs_log_move_tail(mp, mlip->li_lsn); | |
319 | } else { | |
320 | AIL_UNLOCK(mp, s); | |
321 | } | |
322 | ||
323 | ||
324 | } /* xfs_trans_update_ail */ | |
325 | ||
326 | /* | |
327 | * Delete the given item from the AIL. It must already be in | |
328 | * the AIL. | |
329 | * | |
330 | * Wakeup anyone with an lsn less than item's lsn. If the item | |
331 | * we delete in the AIL is the minimum one, update the tail lsn in the | |
332 | * log manager. | |
333 | * | |
334 | * Clear the IN_AIL flag from the item, reset its lsn to 0, and | |
335 | * bump the AIL's generation count to indicate that the tree | |
336 | * has changed. | |
337 | * | |
338 | * This function must be called with the AIL lock held. The lock | |
339 | * is dropped before returning, so the caller must pass in the | |
340 | * cookie returned by AIL_LOCK. | |
341 | */ | |
342 | void | |
343 | xfs_trans_delete_ail( | |
344 | xfs_mount_t *mp, | |
345 | xfs_log_item_t *lip, | |
346 | unsigned long s) | |
347 | { | |
348 | xfs_ail_entry_t *ailp; | |
349 | xfs_log_item_t *dlip; | |
350 | xfs_log_item_t *mlip; | |
351 | ||
352 | if (lip->li_flags & XFS_LI_IN_AIL) { | |
353 | ailp = &(mp->m_ail); | |
354 | mlip = xfs_ail_min(ailp); | |
355 | dlip = xfs_ail_delete(ailp, lip); | |
356 | ASSERT(dlip == lip); | |
357 | ||
358 | ||
359 | lip->li_flags &= ~XFS_LI_IN_AIL; | |
360 | lip->li_lsn = 0; | |
361 | mp->m_ail_gen++; | |
362 | ||
363 | if (mlip == dlip) { | |
364 | mlip = xfs_ail_min(&(mp->m_ail)); | |
365 | AIL_UNLOCK(mp, s); | |
366 | xfs_log_move_tail(mp, (mlip ? mlip->li_lsn : 0)); | |
367 | } else { | |
368 | AIL_UNLOCK(mp, s); | |
369 | } | |
370 | } | |
371 | else { | |
372 | /* | |
373 | * If the file system is not being shutdown, we are in | |
374 | * serious trouble if we get to this stage. | |
375 | */ | |
376 | if (XFS_FORCED_SHUTDOWN(mp)) | |
377 | AIL_UNLOCK(mp, s); | |
378 | else { | |
379 | xfs_cmn_err(XFS_PTAG_AILDELETE, CE_ALERT, mp, | |
380 | "xfs_trans_delete_ail: attempting to delete a log item that is not in the AIL"); | |
1da177e4 | 381 | AIL_UNLOCK(mp, s); |
6f948fbd | 382 | xfs_force_shutdown(mp, XFS_CORRUPT_INCORE); |
1da177e4 LT |
383 | } |
384 | } | |
385 | } | |
386 | ||
387 | ||
388 | ||
389 | /* | |
390 | * Return the item in the AIL with the smallest lsn. | |
391 | * Return the current tree generation number for use | |
392 | * in calls to xfs_trans_next_ail(). | |
393 | */ | |
394 | xfs_log_item_t * | |
395 | xfs_trans_first_ail( | |
396 | xfs_mount_t *mp, | |
397 | int *gen) | |
398 | { | |
399 | xfs_log_item_t *lip; | |
400 | ||
401 | lip = xfs_ail_min(&(mp->m_ail)); | |
402 | *gen = (int)mp->m_ail_gen; | |
403 | ||
404 | return (lip); | |
405 | } | |
406 | ||
407 | /* | |
408 | * If the generation count of the tree has not changed since the | |
409 | * caller last took something from the AIL, then return the elmt | |
410 | * in the tree which follows the one given. If the count has changed, | |
411 | * then return the minimum elmt of the AIL and bump the restarts counter | |
412 | * if one is given. | |
413 | */ | |
414 | xfs_log_item_t * | |
415 | xfs_trans_next_ail( | |
416 | xfs_mount_t *mp, | |
417 | xfs_log_item_t *lip, | |
418 | int *gen, | |
419 | int *restarts) | |
420 | { | |
421 | xfs_log_item_t *nlip; | |
422 | ||
423 | ASSERT(mp && lip && gen); | |
424 | if (mp->m_ail_gen == *gen) { | |
425 | nlip = xfs_ail_next(&(mp->m_ail), lip); | |
426 | } else { | |
427 | nlip = xfs_ail_min(&(mp->m_ail)); | |
428 | *gen = (int)mp->m_ail_gen; | |
429 | if (restarts != NULL) { | |
430 | XFS_STATS_INC(xs_push_ail_restarts); | |
431 | (*restarts)++; | |
432 | } | |
433 | } | |
434 | ||
435 | return (nlip); | |
436 | } | |
437 | ||
438 | ||
439 | /* | |
440 | * The active item list (AIL) is a doubly linked list of log | |
441 | * items sorted by ascending lsn. The base of the list is | |
442 | * a forw/back pointer pair embedded in the xfs mount structure. | |
443 | * The base is initialized with both pointers pointing to the | |
444 | * base. This case always needs to be distinguished, because | |
445 | * the base has no lsn to look at. We almost always insert | |
446 | * at the end of the list, so on inserts we search from the | |
447 | * end of the list to find where the new item belongs. | |
448 | */ | |
449 | ||
450 | /* | |
451 | * Initialize the doubly linked list to point only to itself. | |
452 | */ | |
453 | void | |
454 | xfs_trans_ail_init( | |
455 | xfs_mount_t *mp) | |
456 | { | |
457 | mp->m_ail.ail_forw = (xfs_log_item_t*)&(mp->m_ail); | |
458 | mp->m_ail.ail_back = (xfs_log_item_t*)&(mp->m_ail); | |
459 | } | |
460 | ||
461 | /* | |
462 | * Insert the given log item into the AIL. | |
463 | * We almost always insert at the end of the list, so on inserts | |
464 | * we search from the end of the list to find where the | |
465 | * new item belongs. | |
466 | */ | |
467 | STATIC void | |
468 | xfs_ail_insert( | |
469 | xfs_ail_entry_t *base, | |
470 | xfs_log_item_t *lip) | |
471 | /* ARGSUSED */ | |
472 | { | |
473 | xfs_log_item_t *next_lip; | |
474 | ||
475 | /* | |
476 | * If the list is empty, just insert the item. | |
477 | */ | |
478 | if (base->ail_back == (xfs_log_item_t*)base) { | |
479 | base->ail_forw = lip; | |
480 | base->ail_back = lip; | |
481 | lip->li_ail.ail_forw = (xfs_log_item_t*)base; | |
482 | lip->li_ail.ail_back = (xfs_log_item_t*)base; | |
483 | return; | |
484 | } | |
485 | ||
486 | next_lip = base->ail_back; | |
487 | while ((next_lip != (xfs_log_item_t*)base) && | |
488 | (XFS_LSN_CMP(next_lip->li_lsn, lip->li_lsn) > 0)) { | |
489 | next_lip = next_lip->li_ail.ail_back; | |
490 | } | |
491 | ASSERT((next_lip == (xfs_log_item_t*)base) || | |
492 | (XFS_LSN_CMP(next_lip->li_lsn, lip->li_lsn) <= 0)); | |
493 | lip->li_ail.ail_forw = next_lip->li_ail.ail_forw; | |
494 | lip->li_ail.ail_back = next_lip; | |
495 | next_lip->li_ail.ail_forw = lip; | |
496 | lip->li_ail.ail_forw->li_ail.ail_back = lip; | |
497 | ||
498 | xfs_ail_check(base); | |
499 | return; | |
500 | } | |
501 | ||
502 | /* | |
503 | * Delete the given item from the AIL. Return a pointer to the item. | |
504 | */ | |
505 | /*ARGSUSED*/ | |
506 | STATIC xfs_log_item_t * | |
507 | xfs_ail_delete( | |
508 | xfs_ail_entry_t *base, | |
509 | xfs_log_item_t *lip) | |
510 | /* ARGSUSED */ | |
511 | { | |
512 | lip->li_ail.ail_forw->li_ail.ail_back = lip->li_ail.ail_back; | |
513 | lip->li_ail.ail_back->li_ail.ail_forw = lip->li_ail.ail_forw; | |
514 | lip->li_ail.ail_forw = NULL; | |
515 | lip->li_ail.ail_back = NULL; | |
516 | ||
517 | xfs_ail_check(base); | |
518 | return lip; | |
519 | } | |
520 | ||
521 | /* | |
522 | * Return a pointer to the first item in the AIL. | |
523 | * If the AIL is empty, then return NULL. | |
524 | */ | |
525 | STATIC xfs_log_item_t * | |
526 | xfs_ail_min( | |
527 | xfs_ail_entry_t *base) | |
528 | /* ARGSUSED */ | |
529 | { | |
530 | register xfs_log_item_t *forw = base->ail_forw; | |
531 | if (forw == (xfs_log_item_t*)base) { | |
532 | return NULL; | |
533 | } | |
534 | return forw; | |
535 | } | |
536 | ||
537 | /* | |
538 | * Return a pointer to the item which follows | |
539 | * the given item in the AIL. If the given item | |
540 | * is the last item in the list, then return NULL. | |
541 | */ | |
542 | STATIC xfs_log_item_t * | |
543 | xfs_ail_next( | |
544 | xfs_ail_entry_t *base, | |
545 | xfs_log_item_t *lip) | |
546 | /* ARGSUSED */ | |
547 | { | |
548 | if (lip->li_ail.ail_forw == (xfs_log_item_t*)base) { | |
549 | return NULL; | |
550 | } | |
551 | return lip->li_ail.ail_forw; | |
552 | ||
553 | } | |
554 | ||
555 | #ifdef DEBUG | |
556 | /* | |
557 | * Check that the list is sorted as it should be. | |
558 | */ | |
559 | STATIC void | |
560 | xfs_ail_check( | |
561 | xfs_ail_entry_t *base) | |
562 | { | |
563 | xfs_log_item_t *lip; | |
564 | xfs_log_item_t *prev_lip; | |
565 | ||
566 | lip = base->ail_forw; | |
567 | if (lip == (xfs_log_item_t*)base) { | |
568 | /* | |
569 | * Make sure the pointers are correct when the list | |
570 | * is empty. | |
571 | */ | |
572 | ASSERT(base->ail_back == (xfs_log_item_t*)base); | |
573 | return; | |
574 | } | |
575 | ||
576 | /* | |
577 | * Walk the list checking forward and backward pointers, | |
578 | * lsn ordering, and that every entry has the XFS_LI_IN_AIL | |
579 | * flag set. | |
580 | */ | |
581 | prev_lip = (xfs_log_item_t*)base; | |
582 | while (lip != (xfs_log_item_t*)base) { | |
583 | if (prev_lip != (xfs_log_item_t*)base) { | |
584 | ASSERT(prev_lip->li_ail.ail_forw == lip); | |
585 | ASSERT(XFS_LSN_CMP(prev_lip->li_lsn, lip->li_lsn) <= 0); | |
586 | } | |
587 | ASSERT(lip->li_ail.ail_back == prev_lip); | |
588 | ASSERT((lip->li_flags & XFS_LI_IN_AIL) != 0); | |
589 | prev_lip = lip; | |
590 | lip = lip->li_ail.ail_forw; | |
591 | } | |
592 | ASSERT(lip == (xfs_log_item_t*)base); | |
593 | ASSERT(base->ail_back == prev_lip); | |
594 | } | |
595 | #endif /* DEBUG */ |