Blame view
fs/xfs/xfs_log_cil.c
27.7 KB
71e330b59 xfs: Introduce de... |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 |
/* * Copyright (c) 2010 Red Hat, Inc. All Rights Reserved. * * This program is free software; you can redistribute it and/or * modify it under the terms of the GNU General Public License as * published by the Free Software Foundation. * * This program is distributed in the hope that it would be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write the Free Software Foundation, * Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA */ #include "xfs.h" #include "xfs_fs.h" |
239880ef6 xfs: decouple log... |
20 |
#include "xfs_log_format.h" |
70a9883c5 xfs: create a sha... |
21 |
#include "xfs_shared.h" |
239880ef6 xfs: decouple log... |
22 |
#include "xfs_trans_resv.h" |
71e330b59 xfs: Introduce de... |
23 24 |
#include "xfs_sb.h" #include "xfs_ag.h" |
71e330b59 xfs: Introduce de... |
25 26 27 |
#include "xfs_mount.h" #include "xfs_error.h" #include "xfs_alloc.h" |
efc27b525 xfs: move busy ex... |
28 |
#include "xfs_extent_busy.h" |
e84661aa8 xfs: add online d... |
29 |
#include "xfs_discard.h" |
239880ef6 xfs: decouple log... |
30 31 32 33 |
#include "xfs_trans.h" #include "xfs_trans_priv.h" #include "xfs_log.h" #include "xfs_log_priv.h" |
71e330b59 xfs: Introduce de... |
34 35 |
/* |
71e330b59 xfs: Introduce de... |
36 37 38 39 40 41 42 43 44 45 46 47 48 |
* Allocate a new ticket. Failing to get a new ticket makes it really hard to * recover, so we don't allow failure here. Also, we allocate in a context that * we don't want to be issuing transactions from, so we need to tell the * allocation code this as well. * * We don't reserve any space for the ticket - we are going to steal whatever * space we require from transactions as they commit. To ensure we reserve all * the space required, we need to set the current reservation of the ticket to * zero so that we know to steal the initial transaction overhead from the * first transaction commit. */ static struct xlog_ticket * xlog_cil_ticket_alloc( |
f7bdf03a9 xfs: rename log s... |
49 |
struct xlog *log) |
71e330b59 xfs: Introduce de... |
50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 |
{ struct xlog_ticket *tic; tic = xlog_ticket_alloc(log, 0, 1, XFS_TRANSACTION, 0, KM_SLEEP|KM_NOFS); tic->t_trans_type = XFS_TRANS_CHECKPOINT; /* * set the current reservation to zero so we know to steal the basic * transaction overhead reservation from the first transaction commit. */ tic->t_curr_res = 0; return tic; } /* * After the first stage of log recovery is done, we know where the head and * tail of the log are. We need this log initialisation done before we can * initialise the first CIL checkpoint context. * * Here we allocate a log ticket to track space usage during a CIL push. This * ticket is passed to xlog_write() directly so that we don't slowly leak log * space by failing to account for space used by log headers and additional * region headers for split regions. */ void xlog_cil_init_post_recovery( |
f7bdf03a9 xfs: rename log s... |
77 |
struct xlog *log) |
71e330b59 xfs: Introduce de... |
78 |
{ |
71e330b59 xfs: Introduce de... |
79 80 81 82 83 84 85 |
log->l_cilp->xc_ctx->ticket = xlog_cil_ticket_alloc(log); log->l_cilp->xc_ctx->sequence = 1; log->l_cilp->xc_ctx->commit_lsn = xlog_assign_lsn(log->l_curr_cycle, log->l_curr_block); } /* |
991aaf65f xfs: Combine CIL ... |
86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 |
* Prepare the log item for insertion into the CIL. Calculate the difference in * log space and vectors it will consume, and if it is a new item pin it as * well. */ STATIC void xfs_cil_prepare_item( struct xlog *log, struct xfs_log_vec *lv, struct xfs_log_vec *old_lv, int *diff_len, int *diff_iovecs) { /* Account for the new LV being passed in */ if (lv->lv_buf_len != XFS_LOG_VEC_ORDERED) { *diff_len += lv->lv_buf_len; *diff_iovecs += lv->lv_niovecs; } /* * If there is no old LV, this is the first time we've seen the item in * this CIL context and so we need to pin it. If we are replacing the * old_lv, then remove the space it accounts for and free it. */ if (!old_lv) lv->lv_item->li_ops->iop_pin(lv->lv_item); else if (old_lv != lv) { ASSERT(lv->lv_buf_len != XFS_LOG_VEC_ORDERED); *diff_len -= old_lv->lv_buf_len; *diff_iovecs -= old_lv->lv_niovecs; kmem_free(old_lv); } /* attach new log vector to log item */ lv->lv_item->li_lv = lv; /* * If this is the first time the item is being committed to the * CIL, store the sequence number on the log item so we can * tell in future commits whether this is the first checkpoint * the item is being committed into. */ if (!lv->lv_item->li_seq) lv->lv_item->li_seq = log->l_cilp->xc_ctx->sequence; } /* |
71e330b59 xfs: Introduce de... |
133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 |
* Format log item into a flat buffers * * For delayed logging, we need to hold a formatted buffer containing all the * changes on the log item. This enables us to relog the item in memory and * write it out asynchronously without needing to relock the object that was * modified at the time it gets written into the iclog. * * This function builds a vector for the changes in each log item in the * transaction. It then works out the length of the buffer needed for each log * item, allocates them and formats the vector for the item into the buffer. * The buffer is then attached to the log item are then inserted into the * Committed Item List for tracking until the next checkpoint is written out. * * We don't set up region headers during this process; we simply copy the * regions into the flat buffer. We can do this because we still have to do a * formatting step to write the regions into the iclog buffer. Writing the * ophdrs during the iclog write means that we can support splitting large * regions across iclog boundares without needing a change in the format of the * item/region encapsulation. * * Hence what we need to do now is change the rewrite the vector array to point * to the copied region inside the buffer we just allocated. This allows us to * format the regions into the iclog as though they are being formatted * directly out of the objects themselves. */ |
991aaf65f xfs: Combine CIL ... |
158 159 160 161 162 163 |
static void xlog_cil_insert_format_items( struct xlog *log, struct xfs_trans *tp, int *diff_len, int *diff_iovecs) |
71e330b59 xfs: Introduce de... |
164 |
{ |
0244b9603 xfs: cleanup the ... |
165 |
struct xfs_log_item_desc *lidp; |
71e330b59 xfs: Introduce de... |
166 |
|
0244b9603 xfs: cleanup the ... |
167 168 169 170 |
/* Bail out if we didn't find a log item. */ if (list_empty(&tp->t_items)) { ASSERT(0); |
991aaf65f xfs: Combine CIL ... |
171 |
return; |
0244b9603 xfs: cleanup the ... |
172 173 174 |
} list_for_each_entry(lidp, &tp->t_items, lid_trans) { |
166d13688 xfs: return log i... |
175 |
struct xfs_log_item *lip = lidp->lid_item; |
7492c5b42 xfs: Reduce alloc... |
176 |
struct xfs_log_vec *lv; |
991aaf65f xfs: Combine CIL ... |
177 |
struct xfs_log_vec *old_lv; |
7492c5b42 xfs: Reduce alloc... |
178 179 180 |
int niovecs = 0; int nbytes = 0; int buf_size; |
fd63875cc xfs: Introduce or... |
181 |
bool ordered = false; |
71e330b59 xfs: Introduce de... |
182 |
|
0244b9603 xfs: cleanup the ... |
183 184 185 |
/* Skip items which aren't dirty in this transaction. */ if (!(lidp->lid_flags & XFS_LID_DIRTY)) continue; |
166d13688 xfs: return log i... |
186 187 |
/* get number of vecs and size of data to be stored */ lip->li_ops->iop_size(lip, &niovecs, &nbytes); |
0244b9603 xfs: cleanup the ... |
188 |
/* Skip items that do not have any vectors for writing */ |
b39342134 xfs: remove the l... |
189 |
if (!niovecs) |
0244b9603 xfs: cleanup the ... |
190 |
continue; |
fd63875cc xfs: Introduce or... |
191 192 193 194 195 196 197 198 |
/* * Ordered items need to be tracked but we do not wish to write * them. We need a logvec to track the object, but we do not * need an iovec or buffer to be allocated for copying data. */ if (niovecs == XFS_LOG_VEC_ORDERED) { ordered = true; niovecs = 0; |
7492c5b42 xfs: Reduce alloc... |
199 |
nbytes = 0; |
fd63875cc xfs: Introduce or... |
200 |
} |
bde7cff67 xfs: format log i... |
201 202 203 |
/* * We 64-bit align the length of each iovec so that the start * of the next one is naturally aligned. We'll need to |
3895e51f6 xfs: ensure corre... |
204 205 206 |
* account for that slack space here. Then round nbytes up * to 64-bit alignment so that the initial buffer alignment is * easy to calculate and verify. |
bde7cff67 xfs: format log i... |
207 208 |
*/ nbytes += niovecs * sizeof(uint64_t); |
3895e51f6 xfs: ensure corre... |
209 |
nbytes = round_up(nbytes, sizeof(uint64_t)); |
bde7cff67 xfs: format log i... |
210 |
|
991aaf65f xfs: Combine CIL ... |
211 212 |
/* grab the old item if it exists for reservation accounting */ old_lv = lip->li_lv; |
3895e51f6 xfs: ensure corre... |
213 214 215 216 217 218 219 220 221 |
/* * The data buffer needs to start 64-bit aligned, so round up * that space to ensure we can align it appropriately and not * overrun the buffer. */ buf_size = nbytes + round_up((sizeof(struct xfs_log_vec) + niovecs * sizeof(struct xfs_log_iovec)), sizeof(uint64_t)); |
0244b9603 xfs: cleanup the ... |
222 |
|
f5baac354 xfs: avoid CIL al... |
223 224 225 226 227 228 229 230 |
/* compare to existing item size */ if (lip->li_lv && buf_size <= lip->li_lv->lv_size) { /* same or smaller, optimise common overwrite case */ lv = lip->li_lv; lv->lv_next = NULL; if (ordered) goto insert; |
991aaf65f xfs: Combine CIL ... |
231 232 233 234 235 236 |
/* * set the item up as though it is a new insertion so * that the space reservation accounting is correct. */ *diff_iovecs -= lv->lv_niovecs; *diff_len -= lv->lv_buf_len; |
9597df6b2 xfs: remove dupli... |
237 238 239 240 241 242 243 244 245 246 247 248 |
} else { /* allocate new data chunk */ lv = kmem_zalloc(buf_size, KM_SLEEP|KM_NOFS); lv->lv_item = lip; lv->lv_size = buf_size; if (ordered) { /* track as an ordered logvec */ ASSERT(lip->li_lv == NULL); lv->lv_buf_len = XFS_LOG_VEC_ORDERED; goto insert; } lv->lv_iovecp = (struct xfs_log_iovec *)&lv[1]; |
f5baac354 xfs: avoid CIL al... |
249 |
} |
9597df6b2 xfs: remove dupli... |
250 |
/* Ensure the lv is set up according to ->iop_size */ |
7492c5b42 xfs: Reduce alloc... |
251 |
lv->lv_niovecs = niovecs; |
71e330b59 xfs: Introduce de... |
252 |
|
7492c5b42 xfs: Reduce alloc... |
253 |
/* The allocated data region lies beyond the iovec region */ |
bde7cff67 xfs: format log i... |
254 |
lv->lv_buf_len = 0; |
7492c5b42 xfs: Reduce alloc... |
255 |
lv->lv_buf = (char *)lv + buf_size - nbytes; |
3895e51f6 xfs: ensure corre... |
256 |
ASSERT(IS_ALIGNED((unsigned long)lv->lv_buf, sizeof(uint64_t))); |
bde7cff67 xfs: format log i... |
257 |
lip->li_ops->iop_format(lip, lv); |
7492c5b42 xfs: Reduce alloc... |
258 |
insert: |
f5baac354 xfs: avoid CIL al... |
259 |
ASSERT(lv->lv_buf_len <= nbytes); |
991aaf65f xfs: Combine CIL ... |
260 |
xfs_cil_prepare_item(log, lv, old_lv, diff_len, diff_iovecs); |
3b93c7aae xfs: don't do mem... |
261 |
} |
d1583a383 xfs: reduce the n... |
262 263 264 265 266 267 |
} /* * Insert the log items into the CIL and calculate the difference in space * consumed by the item. Add the space to the checkpoint ticket and calculate * if the change requires additional log metadata. If it does, take that space |
42b2aa86c treewide: Fix typ... |
268 |
* as well. Remove the amount of space we added to the checkpoint ticket from |
d1583a383 xfs: reduce the n... |
269 270 |
* the current transaction ticket so that the accounting works out correctly. */ |
3b93c7aae xfs: don't do mem... |
271 272 |
static void xlog_cil_insert_items( |
f7bdf03a9 xfs: rename log s... |
273 |
struct xlog *log, |
991aaf65f xfs: Combine CIL ... |
274 |
struct xfs_trans *tp) |
3b93c7aae xfs: don't do mem... |
275 |
{ |
d1583a383 xfs: reduce the n... |
276 277 |
struct xfs_cil *cil = log->l_cilp; struct xfs_cil_ctx *ctx = cil->xc_ctx; |
991aaf65f xfs: Combine CIL ... |
278 |
struct xfs_log_item_desc *lidp; |
d1583a383 xfs: reduce the n... |
279 280 281 |
int len = 0; int diff_iovecs = 0; int iclog_space; |
3b93c7aae xfs: don't do mem... |
282 |
|
991aaf65f xfs: Combine CIL ... |
283 |
ASSERT(tp); |
d1583a383 xfs: reduce the n... |
284 285 |
/* |
d1583a383 xfs: reduce the n... |
286 287 288 |
* We can do this safely because the context can't checkpoint until we * are done so it doesn't matter exactly how we update the CIL. */ |
991aaf65f xfs: Combine CIL ... |
289 290 291 292 293 294 295 |
xlog_cil_insert_format_items(log, tp, &len, &diff_iovecs); /* * Now (re-)position everything modified at the tail of the CIL. * We do this here so we only need to take the CIL lock once during * the transaction commit. */ |
d1583a383 xfs: reduce the n... |
296 |
spin_lock(&cil->xc_cil_lock); |
991aaf65f xfs: Combine CIL ... |
297 298 |
list_for_each_entry(lidp, &tp->t_items, lid_trans) { struct xfs_log_item *lip = lidp->lid_item; |
d1583a383 xfs: reduce the n... |
299 |
|
991aaf65f xfs: Combine CIL ... |
300 301 302 |
/* Skip items which aren't dirty in this transaction. */ if (!(lidp->lid_flags & XFS_LID_DIRTY)) continue; |
fd63875cc xfs: Introduce or... |
303 |
|
991aaf65f xfs: Combine CIL ... |
304 |
list_move_tail(&lip->li_cil, &cil->xc_cil); |
fd63875cc xfs: Introduce or... |
305 |
} |
d1583a383 xfs: reduce the n... |
306 |
|
fd63875cc xfs: Introduce or... |
307 308 |
/* account for space used by new iovec headers */ len += diff_iovecs * sizeof(xlog_op_header_t); |
d1583a383 xfs: reduce the n... |
309 |
ctx->nvecs += diff_iovecs; |
991aaf65f xfs: Combine CIL ... |
310 311 312 |
/* attach the transaction to the CIL if it has any busy extents */ if (!list_empty(&tp->t_busy)) list_splice_init(&tp->t_busy, &ctx->busy_extents); |
d1583a383 xfs: reduce the n... |
313 314 315 316 317 318 319 320 |
/* * Now transfer enough transaction reservation to the context ticket * for the checkpoint. The context ticket is special - the unit * reservation has to grow as well as the current reservation as we * steal from tickets so we can correctly determine the space used * during the transaction commit. */ if (ctx->ticket->t_curr_res == 0) { |
d1583a383 xfs: reduce the n... |
321 |
ctx->ticket->t_curr_res = ctx->ticket->t_unit_res; |
991aaf65f xfs: Combine CIL ... |
322 |
tp->t_ticket->t_curr_res -= ctx->ticket->t_unit_res; |
d1583a383 xfs: reduce the n... |
323 324 325 326 327 328 329 330 331 332 333 334 335 |
} /* do we need space for more log record headers? */ iclog_space = log->l_iclog_size - log->l_iclog_hsize; if (len > 0 && (ctx->space_used / iclog_space != (ctx->space_used + len) / iclog_space)) { int hdrs; hdrs = (len + iclog_space - 1) / iclog_space; /* need to take into account split region headers, too */ hdrs *= log->l_iclog_hsize + sizeof(struct xlog_op_header); ctx->ticket->t_unit_res += hdrs; ctx->ticket->t_curr_res += hdrs; |
991aaf65f xfs: Combine CIL ... |
336 337 |
tp->t_ticket->t_curr_res -= hdrs; ASSERT(tp->t_ticket->t_curr_res >= len); |
d1583a383 xfs: reduce the n... |
338 |
} |
991aaf65f xfs: Combine CIL ... |
339 |
tp->t_ticket->t_curr_res -= len; |
d1583a383 xfs: reduce the n... |
340 341 342 |
ctx->space_used += len; spin_unlock(&cil->xc_cil_lock); |
71e330b59 xfs: Introduce de... |
343 344 345 346 347 348 349 350 351 352 |
} static void xlog_cil_free_logvec( struct xfs_log_vec *log_vector) { struct xfs_log_vec *lv; for (lv = log_vector; lv; ) { struct xfs_log_vec *next = lv->lv_next; |
71e330b59 xfs: Introduce de... |
353 354 355 356 357 358 |
kmem_free(lv); lv = next; } } /* |
71e330b59 xfs: Introduce de... |
359 360 361 362 363 364 365 366 367 368 |
* Mark all items committed and clear busy extents. We free the log vector * chains in a separate pass so that we unpin the log items as quickly as * possible. */ static void xlog_cil_committed( void *args, int abort) { struct xfs_cil_ctx *ctx = args; |
e84661aa8 xfs: add online d... |
369 |
struct xfs_mount *mp = ctx->cil->xc_log->l_mp; |
71e330b59 xfs: Introduce de... |
370 |
|
0e57f6a36 xfs: bulk AIL ins... |
371 372 |
xfs_trans_committed_bulk(ctx->cil->xc_log->l_ailp, ctx->lv_chain, ctx->start_lsn, abort); |
71e330b59 xfs: Introduce de... |
373 |
|
4ecbfe637 xfs: clean up bus... |
374 375 |
xfs_extent_busy_sort(&ctx->busy_extents); xfs_extent_busy_clear(mp, &ctx->busy_extents, |
e84661aa8 xfs: add online d... |
376 |
(mp->m_flags & XFS_MOUNT_DISCARD) && !abort); |
71e330b59 xfs: Introduce de... |
377 |
|
4bb928cdb xfs: split the CI... |
378 |
spin_lock(&ctx->cil->xc_push_lock); |
71e330b59 xfs: Introduce de... |
379 |
list_del(&ctx->committing); |
4bb928cdb xfs: split the CI... |
380 |
spin_unlock(&ctx->cil->xc_push_lock); |
71e330b59 xfs: Introduce de... |
381 382 |
xlog_cil_free_logvec(ctx->lv_chain); |
e84661aa8 xfs: add online d... |
383 384 385 386 387 |
if (!list_empty(&ctx->busy_extents)) { ASSERT(mp->m_flags & XFS_MOUNT_DISCARD); xfs_discard_extents(mp, &ctx->busy_extents); |
4ecbfe637 xfs: clean up bus... |
388 |
xfs_extent_busy_clear(mp, &ctx->busy_extents, false); |
e84661aa8 xfs: add online d... |
389 |
} |
71e330b59 xfs: Introduce de... |
390 391 392 393 |
kmem_free(ctx); } /* |
a44f13edf xfs: Reduce log f... |
394 395 396 397 398 399 400 401 402 403 404 405 |
* Push the Committed Item List to the log. If @push_seq flag is zero, then it * is a background flush and so we can chose to ignore it. Otherwise, if the * current sequence is the same as @push_seq we need to do a flush. If * @push_seq is less than the current sequence, then it has already been * flushed and we don't need to do anything - the caller will wait for it to * complete if necessary. * * @push_seq is a value rather than a flag because that allows us to do an * unlocked check of the sequence number for a match. Hence we can allows log * forces to run racily and not issue pushes for the same sequence twice. If we * get a race between multiple pushes for the same sequence they will block on * the first one and then abort, hence avoiding needless pushes. |
71e330b59 xfs: Introduce de... |
406 |
*/ |
a44f13edf xfs: Reduce log f... |
407 |
STATIC int |
71e330b59 xfs: Introduce de... |
408 |
xlog_cil_push( |
f7bdf03a9 xfs: rename log s... |
409 |
struct xlog *log) |
71e330b59 xfs: Introduce de... |
410 411 412 413 414 415 416 |
{ struct xfs_cil *cil = log->l_cilp; struct xfs_log_vec *lv; struct xfs_cil_ctx *ctx; struct xfs_cil_ctx *new_ctx; struct xlog_in_core *commit_iclog; struct xlog_ticket *tic; |
71e330b59 xfs: Introduce de... |
417 |
int num_iovecs; |
71e330b59 xfs: Introduce de... |
418 419 420 421 422 |
int error = 0; struct xfs_trans_header thdr; struct xfs_log_iovec lhdr; struct xfs_log_vec lvhdr = { NULL }; xfs_lsn_t commit_lsn; |
4c2d542f2 xfs: Do backgroun... |
423 |
xfs_lsn_t push_seq; |
71e330b59 xfs: Introduce de... |
424 425 426 |
if (!cil) return 0; |
71e330b59 xfs: Introduce de... |
427 428 |
new_ctx = kmem_zalloc(sizeof(*new_ctx), KM_SLEEP|KM_NOFS); new_ctx->ticket = xlog_cil_ticket_alloc(log); |
4c2d542f2 xfs: Do backgroun... |
429 |
down_write(&cil->xc_ctx_lock); |
71e330b59 xfs: Introduce de... |
430 |
ctx = cil->xc_ctx; |
4bb928cdb xfs: split the CI... |
431 |
spin_lock(&cil->xc_push_lock); |
4c2d542f2 xfs: Do backgroun... |
432 433 |
push_seq = cil->xc_push_seq; ASSERT(push_seq <= ctx->sequence); |
71e330b59 xfs: Introduce de... |
434 |
|
4c2d542f2 xfs: Do backgroun... |
435 436 437 438 439 440 441 |
/* * Check if we've anything to push. If there is nothing, then we don't * move on to a new sequence number and so we have to be able to push * this sequence again later. */ if (list_empty(&cil->xc_cil)) { cil->xc_push_seq = 0; |
4bb928cdb xfs: split the CI... |
442 |
spin_unlock(&cil->xc_push_lock); |
a44f13edf xfs: Reduce log f... |
443 |
goto out_skip; |
4c2d542f2 xfs: Do backgroun... |
444 |
} |
4bb928cdb xfs: split the CI... |
445 |
spin_unlock(&cil->xc_push_lock); |
4c2d542f2 xfs: Do backgroun... |
446 |
|
a44f13edf xfs: Reduce log f... |
447 448 |
/* check for a previously pushed seqeunce */ |
4c2d542f2 xfs: Do backgroun... |
449 |
if (push_seq < cil->xc_ctx->sequence) |
df806158b xfs: enable backg... |
450 |
goto out_skip; |
71e330b59 xfs: Introduce de... |
451 452 453 454 455 456 457 |
/* * pull all the log vectors off the items in the CIL, and * remove the items from the CIL. We don't need the CIL lock * here because it's only needed on the transaction commit * side which is currently locked out by the flush lock. */ lv = NULL; |
71e330b59 xfs: Introduce de... |
458 |
num_iovecs = 0; |
71e330b59 xfs: Introduce de... |
459 460 |
while (!list_empty(&cil->xc_cil)) { struct xfs_log_item *item; |
71e330b59 xfs: Introduce de... |
461 462 463 464 465 466 467 468 469 470 |
item = list_first_entry(&cil->xc_cil, struct xfs_log_item, li_cil); list_del_init(&item->li_cil); if (!ctx->lv_chain) ctx->lv_chain = item->li_lv; else lv->lv_next = item->li_lv; lv = item->li_lv; item->li_lv = NULL; |
71e330b59 xfs: Introduce de... |
471 |
num_iovecs += lv->lv_niovecs; |
71e330b59 xfs: Introduce de... |
472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 |
} /* * initialise the new context and attach it to the CIL. Then attach * the current context to the CIL committing lsit so it can be found * during log forces to extract the commit lsn of the sequence that * needs to be forced. */ INIT_LIST_HEAD(&new_ctx->committing); INIT_LIST_HEAD(&new_ctx->busy_extents); new_ctx->sequence = ctx->sequence + 1; new_ctx->cil = cil; cil->xc_ctx = new_ctx; /* * The switch is now done, so we can drop the context lock and move out * of a shared context. We can't just go straight to the commit record, * though - we need to synchronise with previous and future commits so * that the commit records are correctly ordered in the log to ensure * that we process items during log IO completion in the correct order. * * For example, if we get an EFI in one checkpoint and the EFD in the * next (e.g. due to log forces), we do not want the checkpoint with * the EFD to be committed before the checkpoint with the EFI. Hence * we must strictly order the commit records of the checkpoints so * that: a) the checkpoint callbacks are attached to the iclogs in the * correct order; and b) the checkpoints are replayed in correct order * in log recovery. * * Hence we need to add this context to the committing context list so * that higher sequences will wait for us to write out a commit record * before they do. |
f876e4460 xfs: always do lo... |
504 505 506 507 508 509 |
* * xfs_log_force_lsn requires us to mirror the new sequence into the cil * structure atomically with the addition of this sequence to the * committing list. This also ensures that we can do unlocked checks * against the current sequence in log forces without risking * deferencing a freed context pointer. |
71e330b59 xfs: Introduce de... |
510 |
*/ |
4bb928cdb xfs: split the CI... |
511 |
spin_lock(&cil->xc_push_lock); |
f876e4460 xfs: always do lo... |
512 |
cil->xc_current_sequence = new_ctx->sequence; |
71e330b59 xfs: Introduce de... |
513 |
list_add(&ctx->committing, &cil->xc_committing); |
4bb928cdb xfs: split the CI... |
514 |
spin_unlock(&cil->xc_push_lock); |
71e330b59 xfs: Introduce de... |
515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 |
up_write(&cil->xc_ctx_lock); /* * Build a checkpoint transaction header and write it to the log to * begin the transaction. We need to account for the space used by the * transaction header here as it is not accounted for in xlog_write(). * * The LSN we need to pass to the log items on transaction commit is * the LSN reported by the first log vector write. If we use the commit * record lsn then we can move the tail beyond the grant write head. */ tic = ctx->ticket; thdr.th_magic = XFS_TRANS_HEADER_MAGIC; thdr.th_type = XFS_TRANS_CHECKPOINT; thdr.th_tid = tic->t_tid; thdr.th_num_items = num_iovecs; |
4e0d5f926 xfs: fix the xfs_... |
531 |
lhdr.i_addr = &thdr; |
71e330b59 xfs: Introduce de... |
532 533 534 535 536 537 538 539 540 541 |
lhdr.i_len = sizeof(xfs_trans_header_t); lhdr.i_type = XLOG_REG_TYPE_TRANSHDR; tic->t_curr_res -= lhdr.i_len + sizeof(xlog_op_header_t); lvhdr.lv_niovecs = 1; lvhdr.lv_iovecp = &lhdr; lvhdr.lv_next = ctx->lv_chain; error = xlog_write(log, &lvhdr, tic, &ctx->start_lsn, NULL, 0); if (error) |
7db37c5e6 xfs: fix log tick... |
542 |
goto out_abort_free_ticket; |
71e330b59 xfs: Introduce de... |
543 544 545 546 547 548 |
/* * now that we've written the checkpoint into the log, strictly * order the commit records so replay will get them in the right order. */ restart: |
4bb928cdb xfs: split the CI... |
549 |
spin_lock(&cil->xc_push_lock); |
71e330b59 xfs: Introduce de... |
550 551 552 553 554 555 556 557 558 559 560 561 |
list_for_each_entry(new_ctx, &cil->xc_committing, committing) { /* * Higher sequences will wait for this one so skip them. * Don't wait for own own sequence, either. */ if (new_ctx->sequence >= ctx->sequence) continue; if (!new_ctx->commit_lsn) { /* * It is still being pushed! Wait for the push to * complete, then start again from the beginning. */ |
4bb928cdb xfs: split the CI... |
562 |
xlog_wait(&cil->xc_commit_wait, &cil->xc_push_lock); |
71e330b59 xfs: Introduce de... |
563 564 565 |
goto restart; } } |
4bb928cdb xfs: split the CI... |
566 |
spin_unlock(&cil->xc_push_lock); |
71e330b59 xfs: Introduce de... |
567 |
|
7db37c5e6 xfs: fix log tick... |
568 |
/* xfs_log_done always frees the ticket on error. */ |
71e330b59 xfs: Introduce de... |
569 |
commit_lsn = xfs_log_done(log->l_mp, tic, &commit_iclog, 0); |
7db37c5e6 xfs: fix log tick... |
570 |
if (commit_lsn == -1) |
71e330b59 xfs: Introduce de... |
571 572 573 574 575 576 577 578 579 580 581 582 583 584 |
goto out_abort; /* attach all the transactions w/ busy extents to iclog */ ctx->log_cb.cb_func = xlog_cil_committed; ctx->log_cb.cb_arg = ctx; error = xfs_log_notify(log->l_mp, commit_iclog, &ctx->log_cb); if (error) goto out_abort; /* * now the checkpoint commit is complete and we've attached the * callbacks to the iclog we can assign the commit LSN to the context * and wake up anyone who is waiting for the commit to complete. */ |
4bb928cdb xfs: split the CI... |
585 |
spin_lock(&cil->xc_push_lock); |
71e330b59 xfs: Introduce de... |
586 |
ctx->commit_lsn = commit_lsn; |
eb40a8750 xfs: use wait que... |
587 |
wake_up_all(&cil->xc_commit_wait); |
4bb928cdb xfs: split the CI... |
588 |
spin_unlock(&cil->xc_push_lock); |
71e330b59 xfs: Introduce de... |
589 590 591 592 593 594 595 596 597 |
/* release the hounds! */ return xfs_log_release_iclog(log->l_mp, commit_iclog); out_skip: up_write(&cil->xc_ctx_lock); xfs_log_ticket_put(new_ctx->ticket); kmem_free(new_ctx); return 0; |
7db37c5e6 xfs: fix log tick... |
598 599 |
out_abort_free_ticket: xfs_log_ticket_put(tic); |
71e330b59 xfs: Introduce de... |
600 601 602 603 |
out_abort: xlog_cil_committed(ctx, XFS_LI_ABORTED); return XFS_ERROR(EIO); } |
4c2d542f2 xfs: Do backgroun... |
604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 |
static void xlog_cil_push_work( struct work_struct *work) { struct xfs_cil *cil = container_of(work, struct xfs_cil, xc_push_work); xlog_cil_push(cil->xc_log); } /* * We need to push CIL every so often so we don't cache more than we can fit in * the log. The limit really is that a checkpoint can't be more than half the * log (the current checkpoint is not allowed to overwrite the previous * checkpoint), but commit latency and memory usage limit this to a smaller * size. */ static void xlog_cil_push_background( |
f7bdf03a9 xfs: rename log s... |
622 |
struct xlog *log) |
4c2d542f2 xfs: Do backgroun... |
623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 |
{ struct xfs_cil *cil = log->l_cilp; /* * The cil won't be empty because we are called while holding the * context lock so whatever we added to the CIL will still be there */ ASSERT(!list_empty(&cil->xc_cil)); /* * don't do a background push if we haven't used up all the * space available yet. */ if (cil->xc_ctx->space_used < XLOG_CIL_SPACE_LIMIT(log)) return; |
4bb928cdb xfs: split the CI... |
638 |
spin_lock(&cil->xc_push_lock); |
4c2d542f2 xfs: Do backgroun... |
639 640 641 642 |
if (cil->xc_push_seq < cil->xc_current_sequence) { cil->xc_push_seq = cil->xc_current_sequence; queue_work(log->l_mp->m_cil_workqueue, &cil->xc_push_work); } |
4bb928cdb xfs: split the CI... |
643 |
spin_unlock(&cil->xc_push_lock); |
4c2d542f2 xfs: Do backgroun... |
644 645 |
} |
f876e4460 xfs: always do lo... |
646 647 648 649 650 651 |
/* * xlog_cil_push_now() is used to trigger an immediate CIL push to the sequence * number that is passed. When it returns, the work will be queued for * @push_seq, but it won't be completed. The caller is expected to do any * waiting for push_seq to complete if it is required. */ |
4c2d542f2 xfs: Do backgroun... |
652 |
static void |
f876e4460 xfs: always do lo... |
653 |
xlog_cil_push_now( |
f7bdf03a9 xfs: rename log s... |
654 |
struct xlog *log, |
4c2d542f2 xfs: Do backgroun... |
655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 |
xfs_lsn_t push_seq) { struct xfs_cil *cil = log->l_cilp; if (!cil) return; ASSERT(push_seq && push_seq <= cil->xc_current_sequence); /* start on any pending background push to minimise wait time on it */ flush_work(&cil->xc_push_work); /* * If the CIL is empty or we've already pushed the sequence then * there's no work we need to do. */ |
4bb928cdb xfs: split the CI... |
671 |
spin_lock(&cil->xc_push_lock); |
4c2d542f2 xfs: Do backgroun... |
672 |
if (list_empty(&cil->xc_cil) || push_seq <= cil->xc_push_seq) { |
4bb928cdb xfs: split the CI... |
673 |
spin_unlock(&cil->xc_push_lock); |
4c2d542f2 xfs: Do backgroun... |
674 675 676 677 |
return; } cil->xc_push_seq = push_seq; |
f876e4460 xfs: always do lo... |
678 |
queue_work(log->l_mp->m_cil_workqueue, &cil->xc_push_work); |
4bb928cdb xfs: split the CI... |
679 |
spin_unlock(&cil->xc_push_lock); |
4c2d542f2 xfs: Do backgroun... |
680 |
} |
2c6e24ce1 xfs: prevent dead... |
681 682 683 684 685 686 687 688 689 690 691 692 693 |
bool xlog_cil_empty( struct xlog *log) { struct xfs_cil *cil = log->l_cilp; bool empty = false; spin_lock(&cil->xc_push_lock); if (list_empty(&cil->xc_cil)) empty = true; spin_unlock(&cil->xc_push_lock); return empty; } |
71e330b59 xfs: Introduce de... |
694 |
/* |
a44f13edf xfs: Reduce log f... |
695 696 697 698 699 700 701 702 |
* Commit a transaction with the given vector to the Committed Item List. * * To do this, we need to format the item, pin it in memory if required and * account for the space used by the transaction. Once we have done that we * need to release the unused reservation for the transaction, attach the * transaction to the checkpoint context so we carry the busy extents through * to checkpoint completion, and then unlock all the items in the transaction. * |
a44f13edf xfs: Reduce log f... |
703 704 705 706 |
* Called with the context lock already held in read mode to lock out * background commit, returns without it held once background commits are * allowed again. */ |
c6f972644 xfs: convert xfs_... |
707 |
void |
a44f13edf xfs: Reduce log f... |
708 709 710 |
xfs_log_commit_cil( struct xfs_mount *mp, struct xfs_trans *tp, |
a44f13edf xfs: Reduce log f... |
711 712 713 |
xfs_lsn_t *commit_lsn, int flags) { |
f7bdf03a9 xfs: rename log s... |
714 |
struct xlog *log = mp->m_log; |
991aaf65f xfs: Combine CIL ... |
715 |
struct xfs_cil *cil = log->l_cilp; |
a44f13edf xfs: Reduce log f... |
716 |
int log_flags = 0; |
a44f13edf xfs: Reduce log f... |
717 718 719 |
if (flags & XFS_TRANS_RELEASE_LOG_RES) log_flags = XFS_LOG_REL_PERM_RESERV; |
f5baac354 xfs: avoid CIL al... |
720 |
/* lock out background commit */ |
991aaf65f xfs: Combine CIL ... |
721 |
down_read(&cil->xc_ctx_lock); |
f5baac354 xfs: avoid CIL al... |
722 |
|
991aaf65f xfs: Combine CIL ... |
723 |
xlog_cil_insert_items(log, tp); |
a44f13edf xfs: Reduce log f... |
724 725 726 |
/* check we didn't blow the reservation */ if (tp->t_ticket->t_curr_res < 0) |
991aaf65f xfs: Combine CIL ... |
727 |
xlog_print_tic_res(mp, tp->t_ticket); |
a44f13edf xfs: Reduce log f... |
728 |
|
991aaf65f xfs: Combine CIL ... |
729 730 731 |
tp->t_commit_lsn = cil->xc_ctx->sequence; if (commit_lsn) *commit_lsn = tp->t_commit_lsn; |
a44f13edf xfs: Reduce log f... |
732 |
|
a44f13edf xfs: Reduce log f... |
733 734 735 736 737 738 739 740 741 742 743 744 745 746 |
xfs_log_done(mp, tp->t_ticket, NULL, log_flags); xfs_trans_unreserve_and_mod_sb(tp); /* * Once all the items of the transaction have been copied to the CIL, * the items can be unlocked and freed. * * This needs to be done before we drop the CIL context lock because we * have to update state in the log items and unlock them before they go * to disk. If we don't, then the CIL checkpoint can race with us and * we can run checkpoint completion before we've updated and unlocked * the log items. This affects (at least) processing of stale buffers, * inodes and EFIs. */ |
991aaf65f xfs: Combine CIL ... |
747 |
xfs_trans_free_items(tp, tp->t_commit_lsn, 0); |
a44f13edf xfs: Reduce log f... |
748 |
|
4c2d542f2 xfs: Do backgroun... |
749 |
xlog_cil_push_background(log); |
a44f13edf xfs: Reduce log f... |
750 |
|
991aaf65f xfs: Combine CIL ... |
751 |
up_read(&cil->xc_ctx_lock); |
a44f13edf xfs: Reduce log f... |
752 753 754 |
} /* |
71e330b59 xfs: Introduce de... |
755 756 757 758 759 760 761 762 |
* Conditionally push the CIL based on the sequence passed in. * * We only need to push if we haven't already pushed the sequence * number given. Hence the only time we will trigger a push here is * if the push sequence is the same as the current context. * * We return the current commit lsn to allow the callers to determine if a * iclog flush is necessary following this call. |
71e330b59 xfs: Introduce de... |
763 764 |
*/ xfs_lsn_t |
a44f13edf xfs: Reduce log f... |
765 |
xlog_cil_force_lsn( |
f7bdf03a9 xfs: rename log s... |
766 |
struct xlog *log, |
a44f13edf xfs: Reduce log f... |
767 |
xfs_lsn_t sequence) |
71e330b59 xfs: Introduce de... |
768 769 770 771 |
{ struct xfs_cil *cil = log->l_cilp; struct xfs_cil_ctx *ctx; xfs_lsn_t commit_lsn = NULLCOMMITLSN; |
a44f13edf xfs: Reduce log f... |
772 773 774 775 776 777 778 |
ASSERT(sequence <= cil->xc_current_sequence); /* * check to see if we need to force out the current context. * xlog_cil_push() handles racing pushes for the same sequence, * so no need to deal with it here. */ |
f876e4460 xfs: always do lo... |
779 780 |
restart: xlog_cil_push_now(log, sequence); |
71e330b59 xfs: Introduce de... |
781 782 783 |
/* * See if we can find a previous sequence still committing. |
71e330b59 xfs: Introduce de... |
784 785 786 787 |
* We need to wait for all previous sequence commits to complete * before allowing the force of push_seq to go ahead. Hence block * on commits for those as well. */ |
4bb928cdb xfs: split the CI... |
788 |
spin_lock(&cil->xc_push_lock); |
71e330b59 xfs: Introduce de... |
789 |
list_for_each_entry(ctx, &cil->xc_committing, committing) { |
a44f13edf xfs: Reduce log f... |
790 |
if (ctx->sequence > sequence) |
71e330b59 xfs: Introduce de... |
791 792 793 794 795 796 |
continue; if (!ctx->commit_lsn) { /* * It is still being pushed! Wait for the push to * complete, then start again from the beginning. */ |
4bb928cdb xfs: split the CI... |
797 |
xlog_wait(&cil->xc_commit_wait, &cil->xc_push_lock); |
71e330b59 xfs: Introduce de... |
798 799 |
goto restart; } |
a44f13edf xfs: Reduce log f... |
800 |
if (ctx->sequence != sequence) |
71e330b59 xfs: Introduce de... |
801 802 803 804 |
continue; /* found it! */ commit_lsn = ctx->commit_lsn; } |
f876e4460 xfs: always do lo... |
805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 |
/* * The call to xlog_cil_push_now() executes the push in the background. * Hence by the time we have got here it our sequence may not have been * pushed yet. This is true if the current sequence still matches the * push sequence after the above wait loop and the CIL still contains * dirty objects. * * When the push occurs, it will empty the CIL and * atomically increment the currect sequence past the push sequence and * move it into the committing list. Of course, if the CIL is clean at * the time of the push, it won't have pushed the CIL at all, so in that * case we should try the push for this sequence again from the start * just in case. */ if (sequence == cil->xc_current_sequence && !list_empty(&cil->xc_cil)) { spin_unlock(&cil->xc_push_lock); goto restart; } |
4bb928cdb xfs: split the CI... |
826 |
spin_unlock(&cil->xc_push_lock); |
71e330b59 xfs: Introduce de... |
827 828 |
return commit_lsn; } |
ccf7c23fc xfs: Ensure inode... |
829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 |
/* * Check if the current log item was first committed in this sequence. * We can't rely on just the log item being in the CIL, we have to check * the recorded commit sequence number. * * Note: for this to be used in a non-racy manner, it has to be called with * CIL flushing locked out. As a result, it should only be used during the * transaction commit process when deciding what to format into the item. */ bool xfs_log_item_in_current_chkpt( struct xfs_log_item *lip) { struct xfs_cil_ctx *ctx; |
ccf7c23fc xfs: Ensure inode... |
844 845 846 847 848 849 850 851 852 853 854 855 856 857 |
if (list_empty(&lip->li_cil)) return false; ctx = lip->li_mountp->m_log->l_cilp->xc_ctx; /* * li_seq is written on the first commit of a log item to record the * first checkpoint it is written to. Hence if it is different to the * current sequence, we're in a new checkpoint. */ if (XFS_LSN_CMP(lip->li_seq, ctx->sequence) != 0) return false; return true; } |
4c2d542f2 xfs: Do backgroun... |
858 859 860 861 862 863 |
/* * Perform initial CIL structure initialisation. */ int xlog_cil_init( |
f7bdf03a9 xfs: rename log s... |
864 |
struct xlog *log) |
4c2d542f2 xfs: Do backgroun... |
865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 |
{ struct xfs_cil *cil; struct xfs_cil_ctx *ctx; cil = kmem_zalloc(sizeof(*cil), KM_SLEEP|KM_MAYFAIL); if (!cil) return ENOMEM; ctx = kmem_zalloc(sizeof(*ctx), KM_SLEEP|KM_MAYFAIL); if (!ctx) { kmem_free(cil); return ENOMEM; } INIT_WORK(&cil->xc_push_work, xlog_cil_push_work); INIT_LIST_HEAD(&cil->xc_cil); INIT_LIST_HEAD(&cil->xc_committing); spin_lock_init(&cil->xc_cil_lock); |
4bb928cdb xfs: split the CI... |
883 |
spin_lock_init(&cil->xc_push_lock); |
4c2d542f2 xfs: Do backgroun... |
884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 |
init_rwsem(&cil->xc_ctx_lock); init_waitqueue_head(&cil->xc_commit_wait); INIT_LIST_HEAD(&ctx->committing); INIT_LIST_HEAD(&ctx->busy_extents); ctx->sequence = 1; ctx->cil = cil; cil->xc_ctx = ctx; cil->xc_current_sequence = ctx->sequence; cil->xc_log = log; log->l_cilp = cil; return 0; } void xlog_cil_destroy( |
f7bdf03a9 xfs: rename log s... |
901 |
struct xlog *log) |
4c2d542f2 xfs: Do backgroun... |
902 903 904 905 906 907 908 909 910 911 |
{ if (log->l_cilp->xc_ctx) { if (log->l_cilp->xc_ctx->ticket) xfs_log_ticket_put(log->l_cilp->xc_ctx->ticket); kmem_free(log->l_cilp->xc_ctx); } ASSERT(list_empty(&log->l_cilp->xc_cil)); kmem_free(log->l_cilp); } |