fs/gfs2/lops.c

   1 /*
   2  * Copyright (C) Sistina Software, Inc.  1997-2003 All rights reserved.
   3  * Copyright (C) 2004-2006 Red Hat, Inc.  All rights reserved.
   4  *
   5  * This copyrighted material is made available to anyone wishing to use,
   6  * modify, copy, or redistribute it subject to the terms and conditions
   7  * of the GNU General Public License v.2.
   8  */
   9
  10 #include <linux/sched.h>
  11 #include <linux/slab.h>
  12 #include <linux/spinlock.h>
  13 #include <linux/completion.h>
  14 #include <linux/buffer_head.h>
  15 #include <linux/gfs2_ondisk.h>
  16
  17 #include "gfs2.h"
  18 #include "lm_interface.h"
  19 #include "incore.h"
  20 #include "glock.h"
  21 #include "log.h"
  22 #include "lops.h"
  23 #include "meta_io.h"
  24 #include "recovery.h"
  25 #include "rgrp.h"
  26 #include "trans.h"
  27 #include "util.h"
  28
  29 static void glock_lo_add(struct gfs2_sbd *sdp, struct gfs2_log_element *le)
  30 {
  31         struct gfs2_glock *gl;
  32         struct gfs2_trans *tr = current->journal_info;
  33
  34         tr->tr_touched = 1;
  35
  36         if (!list_empty(&le->le_list))
  37                 return;
  38
  39         gl = container_of(le, struct gfs2_glock, gl_le);
  40         if (gfs2_assert_withdraw(sdp, gfs2_glock_is_held_excl(gl)))
  41                 return;
  42         gfs2_glock_hold(gl);
  43         set_bit(GLF_DIRTY, &gl->gl_flags);
  44
  45         gfs2_log_lock(sdp);
  46         sdp->sd_log_num_gl++;
  47         list_add(&le->le_list, &sdp->sd_log_le_gl);
  48         gfs2_log_unlock(sdp);
  49 }
  50
  51 static void glock_lo_after_commit(struct gfs2_sbd *sdp, struct gfs2_ail *ai)
  52 {
  53         struct list_head *head = &sdp->sd_log_le_gl;
  54         struct gfs2_glock *gl;
  55
  56         while (!list_empty(head)) {
  57                 gl = list_entry(head->next, struct gfs2_glock, gl_le.le_list);
  58                 list_del_init(&gl->gl_le.le_list);
  59                 sdp->sd_log_num_gl--;
  60
  61                 gfs2_assert_withdraw(sdp, gfs2_glock_is_held_excl(gl));
  62                 gfs2_glock_put(gl);
  63         }
  64         gfs2_assert_warn(sdp, !sdp->sd_log_num_gl);
  65 }
  66
  67 static void buf_lo_add(struct gfs2_sbd *sdp, struct gfs2_log_element *le)
  68 {
  69         struct gfs2_bufdata *bd = container_of(le, struct gfs2_bufdata, bd_le);
  70         struct gfs2_trans *tr;
  71
  72         if (!list_empty(&bd->bd_list_tr))
  73                 return;
  74
  75         tr = current->journal_info;
  76         tr->tr_touched = 1;
  77         tr->tr_num_buf++;
  78         list_add(&bd->bd_list_tr, &tr->tr_list_buf);
  79
  80         if (!list_empty(&le->le_list))
  81                 return;
  82
  83         gfs2_trans_add_gl(bd->bd_gl);
  84
  85         gfs2_meta_check(sdp, bd->bd_bh);
  86         gfs2_pin(sdp, bd->bd_bh);
  87
  88         gfs2_log_lock(sdp);
  89         sdp->sd_log_num_buf++;
  90         list_add(&le->le_list, &sdp->sd_log_le_buf);
  91         gfs2_log_unlock(sdp);
  92
  93         tr->tr_num_buf_new++;
  94 }
  95
  96 static void buf_lo_incore_commit(struct gfs2_sbd *sdp, struct gfs2_trans *tr)
  97 {
  98         struct list_head *head = &tr->tr_list_buf;
  99         struct gfs2_bufdata *bd;
 100
 101         while (!list_empty(head)) {
 102                 bd = list_entry(head->next, struct gfs2_bufdata, bd_list_tr);
 103                 list_del_init(&bd->bd_list_tr);
 104                 tr->tr_num_buf--;
 105         }
 106         gfs2_assert_warn(sdp, !tr->tr_num_buf);
 107 }
 108
 109 static void buf_lo_before_commit(struct gfs2_sbd *sdp)
 110 {
 111         struct buffer_head *bh;
 112         struct gfs2_log_descriptor *ld;
 113         struct gfs2_bufdata *bd1 = NULL, *bd2;
 114         unsigned int total = sdp->sd_log_num_buf;
 115         unsigned int offset = sizeof(struct gfs2_log_descriptor);
 116         unsigned int limit;
 117         unsigned int num;
 118         unsigned n;
 119         __be64 *ptr;
 120
 121         offset += (sizeof(__be64) - 1);
 122         offset &= ~(sizeof(__be64) - 1);
 123         limit = (sdp->sd_sb.sb_bsize - offset)/sizeof(__be64);
 124         /* for 4k blocks, limit = 503 */
 125
 126         bd1 = bd2 = list_prepare_entry(bd1, &sdp->sd_log_le_buf, bd_le.le_list);
 127         while(total) {
 128                 num = total;
 129                 if (total > limit)
 130                         num = limit;
 131                 bh = gfs2_log_get_buf(sdp);
 132                 sdp->sd_log_num_hdrs++;
 133                 ld = (struct gfs2_log_descriptor *)bh->b_data;
 134                 ptr = (__be64 *)(bh->b_data + offset);
 135                 ld->ld_header.mh_magic = cpu_to_be32(GFS2_MAGIC);
 136                 ld->ld_header.mh_type = cpu_to_be32(GFS2_METATYPE_LD);
 137                 ld->ld_header.mh_format = cpu_to_be32(GFS2_FORMAT_LD);
 138                 ld->ld_type = cpu_to_be32(GFS2_LOG_DESC_METADATA);
 139                 ld->ld_length = cpu_to_be32(num + 1);
 140                 ld->ld_data1 = cpu_to_be32(num);
 141                 ld->ld_data2 = cpu_to_be32(0);
 142                 memset(ld->ld_reserved, 0, sizeof(ld->ld_reserved));
 143
 144                 n = 0;
 145                 list_for_each_entry_continue(bd1, &sdp->sd_log_le_buf,
 146                                              bd_le.le_list) {
 147                         *ptr++ = cpu_to_be64(bd1->bd_bh->b_blocknr);
 148                         if (++n >= num)
 149                                 break;
 150                 }
 151
 152                 set_buffer_dirty(bh);
 153                 ll_rw_block(WRITE, 1, &bh);
 154
 155                 n = 0;
 156                 list_for_each_entry_continue(bd2, &sdp->sd_log_le_buf,
 157                                              bd_le.le_list) {
 158                         bh = gfs2_log_fake_buf(sdp, bd2->bd_bh);
 159                         set_buffer_dirty(bh);
 160                         ll_rw_block(WRITE, 1, &bh);
 161                         if (++n >= num)
 162                                 break;
 163                 }
 164
 165                 total -= num;
 166         }
 167 }
 168
 169 static void buf_lo_after_commit(struct gfs2_sbd *sdp, struct gfs2_ail *ai)
 170 {
 171         struct list_head *head = &sdp->sd_log_le_buf;
 172         struct gfs2_bufdata *bd;
 173
 174         while (!list_empty(head)) {
 175                 bd = list_entry(head->next, struct gfs2_bufdata, bd_le.le_list);
 176                 list_del_init(&bd->bd_le.le_list);
 177                 sdp->sd_log_num_buf--;
 178
 179                 gfs2_unpin(sdp, bd->bd_bh, ai);
 180         }
 181         gfs2_assert_warn(sdp, !sdp->sd_log_num_buf);
 182 }
 183
 184 static void buf_lo_before_scan(struct gfs2_jdesc *jd,
 185                                struct gfs2_log_header *head, int pass)
 186 {
 187         struct gfs2_sbd *sdp = GFS2_SB(jd->jd_inode);
 188
 189         if (pass != 0)
 190                 return;
 191
 192         sdp->sd_found_blocks = 0;
 193         sdp->sd_replayed_blocks = 0;
 194 }
 195
 196 static int buf_lo_scan_elements(struct gfs2_jdesc *jd, unsigned int start,
 197                                 struct gfs2_log_descriptor *ld, __be64 *ptr,
 198                                 int pass)
 199 {
 200         struct gfs2_inode *ip = GFS2_I(jd->jd_inode);
 201         struct gfs2_sbd *sdp = GFS2_SB(jd->jd_inode);
 202         struct gfs2_glock *gl = ip->i_gl;
 203         unsigned int blks = be32_to_cpu(ld->ld_data1);
 204         struct buffer_head *bh_log, *bh_ip;
 205         uint64_t blkno;
 206         int error = 0;
 207
 208         if (pass != 1 || be32_to_cpu(ld->ld_type) != GFS2_LOG_DESC_METADATA)
 209                 return 0;
 210
 211         gfs2_replay_incr_blk(sdp, &start);
 212
 213         for (; blks; gfs2_replay_incr_blk(sdp, &start), blks--) {
 214                 blkno = be64_to_cpu(*ptr++);
 215
 216                 sdp->sd_found_blocks++;
 217
 218                 if (gfs2_revoke_check(sdp, blkno, start))
 219                         continue;
 220
 221                 error = gfs2_replay_read_block(jd, start, &bh_log);
 222                 if (error)
 223                         return error;
 224
 225                 bh_ip = gfs2_meta_new(gl, blkno);
 226                 memcpy(bh_ip->b_data, bh_log->b_data, bh_log->b_size);
 227
 228                 if (gfs2_meta_check(sdp, bh_ip))
 229                         error = -EIO;
 230                 else
 231                         mark_buffer_dirty(bh_ip);
 232
 233                 brelse(bh_log);
 234                 brelse(bh_ip);
 235
 236                 if (error)
 237                         break;
 238
 239                 sdp->sd_replayed_blocks++;
 240         }
 241
 242         return error;
 243 }
 244
 245 static void buf_lo_after_scan(struct gfs2_jdesc *jd, int error, int pass)
 246 {
 247         struct gfs2_inode *ip = GFS2_I(jd->jd_inode);
 248         struct gfs2_sbd *sdp = GFS2_SB(jd->jd_inode);
 249
 250         if (error) {
 251                 gfs2_meta_sync(ip->i_gl,
 252                                DIO_START | DIO_WAIT);
 253                 return;
 254         }
 255         if (pass != 1)
 256                 return;
 257
 258         gfs2_meta_sync(ip->i_gl, DIO_START | DIO_WAIT);
 259
 260         fs_info(sdp, "jid=%u: Replayed %u of %u blocks\n",
 261                 jd->jd_jid, sdp->sd_replayed_blocks, sdp->sd_found_blocks);
 262 }
 263
 264 static void revoke_lo_add(struct gfs2_sbd *sdp, struct gfs2_log_element *le)
 265 {
 266         struct gfs2_trans *tr;
 267
 268         tr = current->journal_info;
 269         tr->tr_touched = 1;
 270         tr->tr_num_revoke++;
 271
 272         gfs2_log_lock(sdp);
 273         sdp->sd_log_num_revoke++;
 274         list_add(&le->le_list, &sdp->sd_log_le_revoke);
 275         gfs2_log_unlock(sdp);
 276 }
 277
 278 static void revoke_lo_before_commit(struct gfs2_sbd *sdp)
 279 {
 280         struct gfs2_log_descriptor *ld;
 281         struct gfs2_meta_header *mh;
 282         struct buffer_head *bh;
 283         unsigned int offset;
 284         struct list_head *head = &sdp->sd_log_le_revoke;
 285         struct gfs2_revoke *rv;
 286
 287         if (!sdp->sd_log_num_revoke)
 288                 return;
 289
 290         bh = gfs2_log_get_buf(sdp);
 291         ld = (struct gfs2_log_descriptor *)bh->b_data;
 292         ld->ld_header.mh_magic = cpu_to_be32(GFS2_MAGIC);
 293         ld->ld_header.mh_type = cpu_to_be32(GFS2_METATYPE_LD);
 294         ld->ld_header.mh_format = cpu_to_be32(GFS2_FORMAT_LD);
 295         ld->ld_type = cpu_to_be32(GFS2_LOG_DESC_REVOKE);
 296         ld->ld_length = cpu_to_be32(gfs2_struct2blk(sdp, sdp->sd_log_num_revoke,
 297                                                     sizeof(uint64_t)));
 298         ld->ld_data1 = cpu_to_be32(sdp->sd_log_num_revoke);
 299         ld->ld_data2 = cpu_to_be32(0);
 300         memset(ld->ld_reserved, 0, sizeof(ld->ld_reserved));
 301         offset = sizeof(struct gfs2_log_descriptor);
 302
 303         while (!list_empty(head)) {
 304                 rv = list_entry(head->next, struct gfs2_revoke, rv_le.le_list);
 305                 list_del_init(&rv->rv_le.le_list);
 306                 sdp->sd_log_num_revoke--;
 307
 308                 if (offset + sizeof(uint64_t) > sdp->sd_sb.sb_bsize) {
 309                         set_buffer_dirty(bh);
 310                         ll_rw_block(WRITE, 1, &bh);
 311
 312                         bh = gfs2_log_get_buf(sdp);
 313                         mh = (struct gfs2_meta_header *)bh->b_data;
 314                         mh->mh_magic = cpu_to_be32(GFS2_MAGIC);
 315                         mh->mh_type = cpu_to_be32(GFS2_METATYPE_LB);
 316                         mh->mh_format = cpu_to_be32(GFS2_FORMAT_LB);
 317                         offset = sizeof(struct gfs2_meta_header);
 318                 }
 319
 320                 *(__be64 *)(bh->b_data + offset) = cpu_to_be64(rv->rv_blkno);
 321                 kfree(rv);
 322
 323                 offset += sizeof(uint64_t);
 324         }
 325         gfs2_assert_withdraw(sdp, !sdp->sd_log_num_revoke);
 326
 327         set_buffer_dirty(bh);
 328         ll_rw_block(WRITE, 1, &bh);
 329 }
 330
 331 static void revoke_lo_before_scan(struct gfs2_jdesc *jd,
 332                                   struct gfs2_log_header *head, int pass)
 333 {
 334         struct gfs2_sbd *sdp = GFS2_SB(jd->jd_inode);
 335
 336         if (pass != 0)
 337                 return;
 338
 339         sdp->sd_found_revokes = 0;
 340         sdp->sd_replay_tail = head->lh_tail;
 341 }
 342
 343 static int revoke_lo_scan_elements(struct gfs2_jdesc *jd, unsigned int start,
 344                                    struct gfs2_log_descriptor *ld, __be64 *ptr,
 345                                    int pass)
 346 {
 347         struct gfs2_sbd *sdp = GFS2_SB(jd->jd_inode);
 348         unsigned int blks = be32_to_cpu(ld->ld_length);
 349         unsigned int revokes = be32_to_cpu(ld->ld_data1);
 350         struct buffer_head *bh;
 351         unsigned int offset;
 352         uint64_t blkno;
 353         int first = 1;
 354         int error;
 355
 356         if (pass != 0 || be32_to_cpu(ld->ld_type) != GFS2_LOG_DESC_REVOKE)
 357                 return 0;
 358
 359         offset = sizeof(struct gfs2_log_descriptor);
 360
 361         for (; blks; gfs2_replay_incr_blk(sdp, &start), blks--) {
 362                 error = gfs2_replay_read_block(jd, start, &bh);
 363                 if (error)
 364                         return error;
 365
 366                 if (!first)
 367                         gfs2_metatype_check(sdp, bh, GFS2_METATYPE_LB);
 368
 369                 while (offset + sizeof(uint64_t) <= sdp->sd_sb.sb_bsize) {
 370                         blkno = be64_to_cpu(*(__be64 *)(bh->b_data + offset));
 371
 372                         error = gfs2_revoke_add(sdp, blkno, start);
 373                         if (error < 0)
 374                                 return error;
 375                         else if (error)
 376                                 sdp->sd_found_revokes++;
 377
 378                         if (!--revokes)
 379                                 break;
 380                         offset += sizeof(uint64_t);
 381                 }
 382
 383                 brelse(bh);
 384                 offset = sizeof(struct gfs2_meta_header);
 385                 first = 0;
 386         }
 387
 388         return 0;
 389 }
 390
 391 static void revoke_lo_after_scan(struct gfs2_jdesc *jd, int error, int pass)
 392 {
 393         struct gfs2_sbd *sdp = GFS2_SB(jd->jd_inode);
 394
 395         if (error) {
 396                 gfs2_revoke_clean(sdp);
 397                 return;
 398         }
 399         if (pass != 1)
 400                 return;
 401
 402         fs_info(sdp, "jid=%u: Found %u revoke tags\n",
 403                 jd->jd_jid, sdp->sd_found_revokes);
 404
 405         gfs2_revoke_clean(sdp);
 406 }
 407
 408 static void rg_lo_add(struct gfs2_sbd *sdp, struct gfs2_log_element *le)
 409 {
 410         struct gfs2_rgrpd *rgd;
 411         struct gfs2_trans *tr = current->journal_info;
 412
 413         tr->tr_touched = 1;
 414
 415         if (!list_empty(&le->le_list))
 416                 return;
 417
 418         rgd = container_of(le, struct gfs2_rgrpd, rd_le);
 419         gfs2_rgrp_bh_hold(rgd);
 420
 421         gfs2_log_lock(sdp);
 422         sdp->sd_log_num_rg++;
 423         list_add(&le->le_list, &sdp->sd_log_le_rg);
 424         gfs2_log_unlock(sdp);
 425 }
 426
 427 static void rg_lo_after_commit(struct gfs2_sbd *sdp, struct gfs2_ail *ai)
 428 {
 429         struct list_head *head = &sdp->sd_log_le_rg;
 430         struct gfs2_rgrpd *rgd;
 431
 432         while (!list_empty(head)) {
 433                 rgd = list_entry(head->next, struct gfs2_rgrpd, rd_le.le_list);
 434                 list_del_init(&rgd->rd_le.le_list);
 435                 sdp->sd_log_num_rg--;
 436
 437                 gfs2_rgrp_repolish_clones(rgd);
 438                 gfs2_rgrp_bh_put(rgd);
 439         }
 440         gfs2_assert_warn(sdp, !sdp->sd_log_num_rg);
 441 }
 442
 443 /**
 444  * databuf_lo_add - Add a databuf to the transaction.
 445  *
 446  * This is used in two distinct cases:
 447  * i) In ordered write mode
 448  *    We put the data buffer on a list so that we can ensure that its
 449  *    synced to disk at the right time
 450  * ii) In journaled data mode
 451  *    We need to journal the data block in the same way as metadata in
 452  *    the functions above. The difference is that here we have a tag
 453  *    which is two __be64's being the block number (as per meta data)
 454  *    and a flag which says whether the data block needs escaping or
 455  *    not. This means we need a new log entry for each 251 or so data
 456  *    blocks, which isn't an enormous overhead but twice as much as
 457  *    for normal metadata blocks.
 458  */
 459 static void databuf_lo_add(struct gfs2_sbd *sdp, struct gfs2_log_element *le)
 460 {
 461         struct gfs2_bufdata *bd = container_of(le, struct gfs2_bufdata, bd_le);
 462         struct gfs2_trans *tr = current->journal_info;
 463         struct address_space *mapping = bd->bd_bh->b_page->mapping;
 464         struct gfs2_inode *ip = GFS2_I(mapping->host);
 465
 466         tr->tr_touched = 1;
 467         if (!list_empty(&bd->bd_list_tr) &&
 468             (ip->i_di.di_flags & GFS2_DIF_JDATA)) {
 469                 tr->tr_num_buf++;
 470                 list_add(&bd->bd_list_tr, &tr->tr_list_buf);
 471                 gfs2_pin(sdp, bd->bd_bh);
 472                 tr->tr_num_buf_new++;
 473         }
 474         gfs2_trans_add_gl(bd->bd_gl);
 475         gfs2_log_lock(sdp);
 476         if (!list_empty(&le->le_list)) {
 477                 if (ip->i_di.di_flags & GFS2_DIF_JDATA)
 478                         sdp->sd_log_num_jdata++;
 479                 sdp->sd_log_num_databuf++;
 480                 list_add(&le->le_list, &sdp->sd_log_le_databuf);
 481         }
 482         gfs2_log_unlock(sdp);
 483 }
 484
 485 static int gfs2_check_magic(struct buffer_head *bh)
 486 {
 487         struct page *page = bh->b_page;
 488         void *kaddr;
 489         __be32 *ptr;
 490         int rv = 0;
 491
 492         kaddr = kmap_atomic(page, KM_USER0);
 493         ptr = kaddr + bh_offset(bh);
 494         if (*ptr == cpu_to_be32(GFS2_MAGIC))
 495                 rv = 1;
 496         kunmap_atomic(page, KM_USER0);
 497
 498         return rv;
 499 }
 500
 501 /**
 502  * databuf_lo_before_commit - Scan the data buffers, writing as we go
 503  *
 504  * Here we scan through the lists of buffers and make the assumption
 505  * that any buffer thats been pinned is being journaled, and that
 506  * any unpinned buffer is an ordered write data buffer and therefore
 507  * will be written back rather than journaled.
 508  */
 509 static void databuf_lo_before_commit(struct gfs2_sbd *sdp)
 510 {
 511         LIST_HEAD(started);
 512         struct gfs2_bufdata *bd1 = NULL, *bd2, *bdt;
 513         struct buffer_head *bh = NULL;
 514         unsigned int offset = sizeof(struct gfs2_log_descriptor);
 515         struct gfs2_log_descriptor *ld;
 516         unsigned int limit;
 517         unsigned int total_dbuf = sdp->sd_log_num_databuf;
 518         unsigned int total_jdata = sdp->sd_log_num_jdata;
 519         unsigned int num, n;
 520         __be64 *ptr = NULL;
 521
 522         offset += (2*sizeof(__be64) - 1);
 523         offset &= ~(2*sizeof(__be64) - 1);
 524         limit = (sdp->sd_sb.sb_bsize - offset)/sizeof(__be64);
 525
 526         /*
 527          * Start writing ordered buffers, write journaled buffers
 528          * into the log along with a header
 529          */
 530         gfs2_log_lock(sdp);
 531         bd2 = bd1 = list_prepare_entry(bd1, &sdp->sd_log_le_databuf,
 532                                        bd_le.le_list);
 533         while(total_dbuf) {
 534                 num = total_jdata;
 535                 if (num > limit)
 536                         num = limit;
 537                 n = 0;
 538                 list_for_each_entry_safe_continue(bd1, bdt,
 539                                                   &sdp->sd_log_le_databuf,
 540                                                   bd_le.le_list) {
 541                         /* An ordered write buffer */
 542                         if (bd1->bd_bh && !buffer_pinned(bd1->bd_bh)) {
 543                                 list_move(&bd1->bd_le.le_list, &started);
 544                                 if (bd1 == bd2) {
 545                                         bd2 = NULL;
 546                                         bd2 = list_prepare_entry(bd2,
 547                                                         &sdp->sd_log_le_databuf,
 548                                                         bd_le.le_list);
 549                                 }
 550                                 total_dbuf--;
 551                                 if (bd1->bd_bh) {
 552                                         get_bh(bd1->bd_bh);
 553                                         if (buffer_dirty(bd1->bd_bh)) {
 554                                                 gfs2_log_unlock(sdp);
 555                                                 wait_on_buffer(bd1->bd_bh);
 556                                                 ll_rw_block(WRITE, 1,
 557                                                             &bd1->bd_bh);
 558                                                 gfs2_log_lock(sdp);
 559                                         }
 560                                         brelse(bd1->bd_bh);
 561                                         continue;
 562                                 }
 563                                 continue;
 564                         } else if (bd1->bd_bh) { /* A journaled buffer */
 565                                 int magic;
 566                                 gfs2_log_unlock(sdp);
 567                                 if (!bh) {
 568                                         bh = gfs2_log_get_buf(sdp);
 569                                         sdp->sd_log_num_hdrs++;
 570                                         ld = (struct gfs2_log_descriptor *)
 571                                              bh->b_data;
 572                                         ptr = (__be64 *)(bh->b_data + offset);
 573                                         ld->ld_header.mh_magic =
 574                                                 cpu_to_be32(GFS2_MAGIC);
 575                                         ld->ld_header.mh_type =
 576                                                 cpu_to_be32(GFS2_METATYPE_LD);
 577                                         ld->ld_header.mh_format =
 578                                                 cpu_to_be32(GFS2_FORMAT_LD);
 579                                         ld->ld_type =
 580                                                 cpu_to_be32(GFS2_LOG_DESC_JDATA);
 581                                         ld->ld_length = cpu_to_be32(num + 1);
 582                                         ld->ld_data1 = cpu_to_be32(num);
 583                                         ld->ld_data2 = cpu_to_be32(0);
 584                                         memset(ld->ld_reserved, 0, sizeof(ld->ld_reserved));
 585                                 }
 586                                 magic = gfs2_check_magic(bd1->bd_bh);
 587                                 *ptr++ = cpu_to_be64(bd1->bd_bh->b_blocknr);
 588                                 *ptr++ = cpu_to_be64((__u64)magic);
 589                                 clear_buffer_escaped(bd1->bd_bh);
 590                                 if (unlikely(magic != 0))
 591                                         set_buffer_escaped(bd1->bd_bh);
 592                                 gfs2_log_lock(sdp);
 593                                 if (n++ > num)
 594                                         break;
 595                         }
 596                 }
 597                 gfs2_log_unlock(sdp);
 598                 if (bh) {
 599                         set_buffer_dirty(bh);
 600                         ll_rw_block(WRITE, 1, &bh);
 601                         bh = NULL;
 602                 }
 603                 n = 0;
 604                 gfs2_log_lock(sdp);
 605                 list_for_each_entry_continue(bd2, &sdp->sd_log_le_databuf,
 606                                              bd_le.le_list) {
 607                         if (!bd2->bd_bh)
 608                                 continue;
 609                         /* copy buffer if it needs escaping */
 610                         gfs2_log_unlock(sdp);
 611                         if (unlikely(buffer_escaped(bd2->bd_bh))) {
 612                                 void *kaddr;
 613                                 struct page *page = bd2->bd_bh->b_page;
 614                                 bh = gfs2_log_get_buf(sdp);
 615                                 kaddr = kmap_atomic(page, KM_USER0);
 616                                 memcpy(bh->b_data,
 617                                        kaddr + bh_offset(bd2->bd_bh),
 618                                        sdp->sd_sb.sb_bsize);
 619                                 kunmap_atomic(page, KM_USER0);
 620                                 *(__be32 *)bh->b_data = 0;
 621                         } else {
 622                                 bh = gfs2_log_fake_buf(sdp, bd2->bd_bh);
 623                         }
 624                         set_buffer_dirty(bh);
 625                         ll_rw_block(WRITE, 1, &bh);
 626                         gfs2_log_lock(sdp);
 627                         if (++n >= num)
 628                                 break;
 629                 }
 630                 bh = NULL;
 631                 total_dbuf -= num;
 632                 total_jdata -= num;
 633         }
 634         gfs2_log_unlock(sdp);
 635
 636         /* Wait on all ordered buffers */
 637         while (!list_empty(&started)) {
 638                 gfs2_log_lock(sdp);
 639                 bd1 = list_entry(started.next, struct gfs2_bufdata,
 640                                  bd_le.le_list);
 641                 list_del(&bd1->bd_le.le_list);
 642                 sdp->sd_log_num_databuf--;
 643
 644                 bh = bd1->bd_bh;
 645                 if (bh) {
 646                         bh->b_private = NULL;
 647                         gfs2_log_unlock(sdp);
 648                         wait_on_buffer(bh);
 649                         brelse(bh);
 650                 } else
 651                         gfs2_log_unlock(sdp);
 652
 653                 kmem_cache_free(gfs2_bufdata_cachep, bd1);
 654         }
 655
 656         /* We've removed all the ordered write bufs here, so only jdata left */
 657         gfs2_assert_warn(sdp, sdp->sd_log_num_databuf == sdp->sd_log_num_jdata);
 658 }
 659
 660 static int databuf_lo_scan_elements(struct gfs2_jdesc *jd, unsigned int start,
 661                                     struct gfs2_log_descriptor *ld,
 662                                     __be64 *ptr, int pass)
 663 {
 664         struct gfs2_inode *ip = GFS2_I(jd->jd_inode);
 665         struct gfs2_sbd *sdp = GFS2_SB(jd->jd_inode);
 666         struct gfs2_glock *gl = ip->i_gl;
 667         unsigned int blks = be32_to_cpu(ld->ld_data1);
 668         struct buffer_head *bh_log, *bh_ip;
 669         uint64_t blkno;
 670         uint64_t esc;
 671         int error = 0;
 672
 673         if (pass != 1 || be32_to_cpu(ld->ld_type) != GFS2_LOG_DESC_JDATA)
 674                 return 0;
 675
 676         gfs2_replay_incr_blk(sdp, &start);
 677         for (; blks; gfs2_replay_incr_blk(sdp, &start), blks--) {
 678                 blkno = be64_to_cpu(*ptr++);
 679                 esc = be64_to_cpu(*ptr++);
 680
 681                 sdp->sd_found_blocks++;
 682
 683                 if (gfs2_revoke_check(sdp, blkno, start))
 684                         continue;
 685
 686                 error = gfs2_replay_read_block(jd, start, &bh_log);
 687                 if (error)
 688                         return error;
 689
 690                 bh_ip = gfs2_meta_new(gl, blkno);
 691                 memcpy(bh_ip->b_data, bh_log->b_data, bh_log->b_size);
 692
 693                 /* Unescape */
 694                 if (esc) {
 695                         __be32 *eptr = (__be32 *)bh_ip->b_data;
 696                         *eptr = cpu_to_be32(GFS2_MAGIC);
 697                 }
 698                 mark_buffer_dirty(bh_ip);
 699
 700                 brelse(bh_log);
 701                 brelse(bh_ip);
 702                 if (error)
 703                         break;
 704
 705                 sdp->sd_replayed_blocks++;
 706         }
 707
 708         return error;
 709 }
 710
 711 /* FIXME: sort out accounting for log blocks etc. */
 712
 713 static void databuf_lo_after_scan(struct gfs2_jdesc *jd, int error, int pass)
 714 {
 715         struct gfs2_inode *ip = GFS2_I(jd->jd_inode);
 716         struct gfs2_sbd *sdp = GFS2_SB(jd->jd_inode);
 717
 718         if (error) {
 719                 gfs2_meta_sync(ip->i_gl,
 720                                DIO_START | DIO_WAIT);
 721                 return;
 722         }
 723         if (pass != 1)
 724                 return;
 725
 726         /* data sync? */
 727         gfs2_meta_sync(ip->i_gl, DIO_START | DIO_WAIT);
 728
 729         fs_info(sdp, "jid=%u: Replayed %u of %u data blocks\n",
 730                 jd->jd_jid, sdp->sd_replayed_blocks, sdp->sd_found_blocks);
 731 }
 732
 733 static void databuf_lo_after_commit(struct gfs2_sbd *sdp, struct gfs2_ail *ai)
 734 {
 735         struct list_head *head = &sdp->sd_log_le_databuf;
 736         struct gfs2_bufdata *bd;
 737
 738         while (!list_empty(head)) {
 739                 bd = list_entry(head->next, struct gfs2_bufdata, bd_le.le_list);
 740                 list_del(&bd->bd_le.le_list);
 741                 sdp->sd_log_num_databuf--;
 742                 sdp->sd_log_num_jdata--;
 743                 gfs2_unpin(sdp, bd->bd_bh, ai);
 744         }
 745         gfs2_assert_warn(sdp, !sdp->sd_log_num_databuf);
 746         gfs2_assert_warn(sdp, !sdp->sd_log_num_jdata);
 747 }
 748
 749
 750 const struct gfs2_log_operations gfs2_glock_lops = {
 751         .lo_add = glock_lo_add,
 752         .lo_after_commit = glock_lo_after_commit,
 753         .lo_name = "glock"
 754 };
 755
 756 const struct gfs2_log_operations gfs2_buf_lops = {
 757         .lo_add = buf_lo_add,
 758         .lo_incore_commit = buf_lo_incore_commit,
 759         .lo_before_commit = buf_lo_before_commit,
 760         .lo_after_commit = buf_lo_after_commit,
 761         .lo_before_scan = buf_lo_before_scan,
 762         .lo_scan_elements = buf_lo_scan_elements,
 763         .lo_after_scan = buf_lo_after_scan,
 764         .lo_name = "buf"
 765 };
 766
 767 const struct gfs2_log_operations gfs2_revoke_lops = {
 768         .lo_add = revoke_lo_add,
 769         .lo_before_commit = revoke_lo_before_commit,
 770         .lo_before_scan = revoke_lo_before_scan,
 771         .lo_scan_elements = revoke_lo_scan_elements,
 772         .lo_after_scan = revoke_lo_after_scan,
 773         .lo_name = "revoke"
 774 };
 775
 776 const struct gfs2_log_operations gfs2_rg_lops = {
 777         .lo_add = rg_lo_add,
 778         .lo_after_commit = rg_lo_after_commit,
 779         .lo_name = "rg"
 780 };
 781
 782 const struct gfs2_log_operations gfs2_databuf_lops = {
 783         .lo_add = databuf_lo_add,
 784         .lo_incore_commit = buf_lo_incore_commit,
 785         .lo_before_commit = databuf_lo_before_commit,
 786         .lo_after_commit = databuf_lo_after_commit,
 787         .lo_scan_elements = databuf_lo_scan_elements,
 788         .lo_after_scan = databuf_lo_after_scan,
 789         .lo_name = "databuf"
 790 };
 791
 792 const struct gfs2_log_operations *gfs2_log_ops[] = {
 793         &gfs2_glock_lops,
 794         &gfs2_buf_lops,
 795         &gfs2_revoke_lops,
 796         &gfs2_rg_lops,
 797         &gfs2_databuf_lops,
 798         NULL
 799 };
 800