diff options
Diffstat (limited to 'fs/xfs/xfs_log_recover.c')
| -rw-r--r-- | fs/xfs/xfs_log_recover.c | 959 |
1 files changed, 502 insertions, 457 deletions
diff --git a/fs/xfs/xfs_log_recover.c b/fs/xfs/xfs_log_recover.c index 981af0f6504..00cd7f3a8f5 100644 --- a/fs/xfs/xfs_log_recover.c +++ b/fs/xfs/xfs_log_recover.c @@ -179,7 +179,7 @@ xlog_bread_noalign( xfs_warn(log->l_mp, "Invalid block length (0x%x) for buffer", nbblks); XFS_ERROR_REPORT(__func__, XFS_ERRLEVEL_HIGH, log->l_mp); - return EFSCORRUPTED; + return -EFSCORRUPTED; } blk_no = round_down(blk_no, log->l_sectBBsize); @@ -193,12 +193,8 @@ xlog_bread_noalign( bp->b_io_length = nbblks; bp->b_error = 0; - if (XFS_FORCED_SHUTDOWN(log->l_mp)) - return XFS_ERROR(EIO); - - xfs_buf_iorequest(bp); - error = xfs_buf_iowait(bp); - if (error) + error = xfs_buf_submit_wait(bp); + if (error && !XFS_FORCED_SHUTDOWN(log->l_mp)) xfs_buf_ioerror_alert(bp, __func__); return error; } @@ -268,7 +264,7 @@ xlog_bwrite( xfs_warn(log->l_mp, "Invalid block length (0x%x) for buffer", nbblks); XFS_ERROR_REPORT(__func__, XFS_ERRLEVEL_HIGH, log->l_mp); - return EFSCORRUPTED; + return -EFSCORRUPTED; } blk_no = round_down(blk_no, log->l_sectBBsize); @@ -330,14 +326,14 @@ xlog_header_check_recover( xlog_header_check_dump(mp, head); XFS_ERROR_REPORT("xlog_header_check_recover(1)", XFS_ERRLEVEL_HIGH, mp); - return XFS_ERROR(EFSCORRUPTED); + return -EFSCORRUPTED; } else if (unlikely(!uuid_equal(&mp->m_sb.sb_uuid, &head->h_fs_uuid))) { xfs_warn(mp, "dirty log entry has mismatched uuid - can't recover"); xlog_header_check_dump(mp, head); XFS_ERROR_REPORT("xlog_header_check_recover(2)", XFS_ERRLEVEL_HIGH, mp); - return XFS_ERROR(EFSCORRUPTED); + return -EFSCORRUPTED; } return 0; } @@ -364,7 +360,7 @@ xlog_header_check_mount( xlog_header_check_dump(mp, head); XFS_ERROR_REPORT("xlog_header_check_mount", XFS_ERRLEVEL_HIGH, mp); - return XFS_ERROR(EFSCORRUPTED); + return -EFSCORRUPTED; } return 0; } @@ -378,12 +374,14 @@ xlog_recover_iodone( * We're not going to bother about retrying * this during recovery. One strike! */ - xfs_buf_ioerror_alert(bp, __func__); - xfs_force_shutdown(bp->b_target->bt_mount, - SHUTDOWN_META_IO_ERROR); + if (!XFS_FORCED_SHUTDOWN(bp->b_target->bt_mount)) { + xfs_buf_ioerror_alert(bp, __func__); + xfs_force_shutdown(bp->b_target->bt_mount, + SHUTDOWN_META_IO_ERROR); + } } bp->b_iodone = NULL; - xfs_buf_ioend(bp, 0); + xfs_buf_ioend(bp); } /* @@ -462,7 +460,7 @@ xlog_find_verify_cycle( while (!(bp = xlog_get_bp(log, bufblks))) { bufblks >>= 1; if (bufblks < log->l_sectBBsize) - return ENOMEM; + return -ENOMEM; } for (i = start_blk; i < start_blk + nbblks; i += bufblks) { @@ -524,7 +522,7 @@ xlog_find_verify_log_record( if (!(bp = xlog_get_bp(log, num_blks))) { if (!(bp = xlog_get_bp(log, 1))) - return ENOMEM; + return -ENOMEM; smallmem = 1; } else { error = xlog_bread(log, start_blk, num_blks, bp, &offset); @@ -539,7 +537,7 @@ xlog_find_verify_log_record( xfs_warn(log->l_mp, "Log inconsistent (didn't find previous header)"); ASSERT(0); - error = XFS_ERROR(EIO); + error = -EIO; goto out; } @@ -564,7 +562,7 @@ xlog_find_verify_log_record( * will be called again for the end of the physical log. */ if (i == -1) { - error = -1; + error = 1; goto out; } @@ -628,7 +626,12 @@ xlog_find_head( int error, log_bbnum = log->l_logBBsize; /* Is the end of the log device zeroed? */ - if ((error = xlog_find_zeroed(log, &first_blk)) == -1) { + error = xlog_find_zeroed(log, &first_blk); + if (error < 0) { + xfs_warn(log->l_mp, "empty log check failed"); + return error; + } + if (error == 1) { *return_head_blk = first_blk; /* Is the whole lot zeroed? */ @@ -641,15 +644,12 @@ xlog_find_head( } return 0; - } else if (error) { - xfs_warn(log->l_mp, "empty log check failed"); - return error; } first_blk = 0; /* get cycle # of 1st block */ bp = xlog_get_bp(log, 1); if (!bp) - return ENOMEM; + return -ENOMEM; error = xlog_bread(log, 0, 1, bp, &offset); if (error) @@ -818,29 +818,29 @@ validate_head: start_blk = head_blk - num_scan_bblks; /* don't read head_blk */ /* start ptr at last block ptr before head_blk */ - if ((error = xlog_find_verify_log_record(log, start_blk, - &head_blk, 0)) == -1) { - error = XFS_ERROR(EIO); - goto bp_err; - } else if (error) + error = xlog_find_verify_log_record(log, start_blk, &head_blk, 0); + if (error == 1) + error = -EIO; + if (error) goto bp_err; } else { start_blk = 0; ASSERT(head_blk <= INT_MAX); - if ((error = xlog_find_verify_log_record(log, start_blk, - &head_blk, 0)) == -1) { + error = xlog_find_verify_log_record(log, start_blk, &head_blk, 0); + if (error < 0) + goto bp_err; + if (error == 1) { /* We hit the beginning of the log during our search */ start_blk = log_bbnum - (num_scan_bblks - head_blk); new_blk = log_bbnum; ASSERT(start_blk <= INT_MAX && (xfs_daddr_t) log_bbnum-start_blk >= 0); ASSERT(head_blk <= INT_MAX); - if ((error = xlog_find_verify_log_record(log, - start_blk, &new_blk, - (int)head_blk)) == -1) { - error = XFS_ERROR(EIO); - goto bp_err; - } else if (error) + error = xlog_find_verify_log_record(log, start_blk, + &new_blk, (int)head_blk); + if (error == 1) + error = -EIO; + if (error) goto bp_err; if (new_blk != log_bbnum) head_blk = new_blk; @@ -911,7 +911,7 @@ xlog_find_tail( bp = xlog_get_bp(log, 1); if (!bp) - return ENOMEM; + return -ENOMEM; if (*head_blk == 0) { /* special case */ error = xlog_bread(log, 0, 1, bp, &offset); if (error) @@ -961,7 +961,7 @@ xlog_find_tail( xfs_warn(log->l_mp, "%s: couldn't find sync record", __func__); xlog_put_bp(bp); ASSERT(0); - return XFS_ERROR(EIO); + return -EIO; } /* find blk_no of tail of log */ @@ -1092,8 +1092,8 @@ done: * * Return: * 0 => the log is completely written to - * -1 => use *blk_no as the first block of the log - * >0 => error has occurred + * 1 => use *blk_no as the first block of the log + * <0 => error has occurred */ STATIC int xlog_find_zeroed( @@ -1112,7 +1112,7 @@ xlog_find_zeroed( /* check totally zeroed log */ bp = xlog_get_bp(log, 1); if (!bp) - return ENOMEM; + return -ENOMEM; error = xlog_bread(log, 0, 1, bp, &offset); if (error) goto bp_err; @@ -1121,7 +1121,7 @@ xlog_find_zeroed( if (first_cycle == 0) { /* completely zeroed log */ *blk_no = 0; xlog_put_bp(bp); - return -1; + return 1; } /* check partially zeroed log */ @@ -1141,7 +1141,7 @@ xlog_find_zeroed( */ xfs_warn(log->l_mp, "Log inconsistent or not a log (last==0, first!=1)"); - error = XFS_ERROR(EINVAL); + error = -EINVAL; goto bp_err; } @@ -1179,19 +1179,18 @@ xlog_find_zeroed( * Potentially backup over partial log record write. We don't need * to search the end of the log because we know it is zero. */ - if ((error = xlog_find_verify_log_record(log, start_blk, - &last_blk, 0)) == -1) { - error = XFS_ERROR(EIO); - goto bp_err; - } else if (error) - goto bp_err; + error = xlog_find_verify_log_record(log, start_blk, &last_blk, 0); + if (error == 1) + error = -EIO; + if (error) + goto bp_err; *blk_no = last_blk; bp_err: xlog_put_bp(bp); if (error) return error; - return -1; + return 1; } /* @@ -1251,7 +1250,7 @@ xlog_write_log_records( while (!(bp = xlog_get_bp(log, bufblks))) { bufblks >>= 1; if (bufblks < sectbb) - return ENOMEM; + return -ENOMEM; } /* We may need to do a read at the start to fill in part of @@ -1354,7 +1353,7 @@ xlog_clear_stale_blocks( if (unlikely(head_block < tail_block || head_block >= log->l_logBBsize)) { XFS_ERROR_REPORT("xlog_clear_stale_blocks(1)", XFS_ERRLEVEL_LOW, log->l_mp); - return XFS_ERROR(EFSCORRUPTED); + return -EFSCORRUPTED; } tail_distance = tail_block + (log->l_logBBsize - head_block); } else { @@ -1366,7 +1365,7 @@ xlog_clear_stale_blocks( if (unlikely(head_block >= tail_block || head_cycle != (tail_cycle + 1))){ XFS_ERROR_REPORT("xlog_clear_stale_blocks(2)", XFS_ERRLEVEL_LOW, log->l_mp); - return XFS_ERROR(EFSCORRUPTED); + return -EFSCORRUPTED; } tail_distance = tail_block - head_block; } @@ -1444,160 +1443,6 @@ xlog_clear_stale_blocks( ****************************************************************************** */ -STATIC xlog_recover_t * -xlog_recover_find_tid( - struct hlist_head *head, - xlog_tid_t tid) -{ - xlog_recover_t *trans; - - hlist_for_each_entry(trans, head, r_list) { - if (trans->r_log_tid == tid) - return trans; - } - return NULL; -} - -STATIC void -xlog_recover_new_tid( - struct hlist_head *head, - xlog_tid_t tid, - xfs_lsn_t lsn) -{ - xlog_recover_t *trans; - - trans = kmem_zalloc(sizeof(xlog_recover_t), KM_SLEEP); - trans->r_log_tid = tid; - trans->r_lsn = lsn; - INIT_LIST_HEAD(&trans->r_itemq); - - INIT_HLIST_NODE(&trans->r_list); - hlist_add_head(&trans->r_list, head); -} - -STATIC void -xlog_recover_add_item( - struct list_head *head) -{ - xlog_recover_item_t *item; - - item = kmem_zalloc(sizeof(xlog_recover_item_t), KM_SLEEP); - INIT_LIST_HEAD(&item->ri_list); - list_add_tail(&item->ri_list, head); -} - -STATIC int -xlog_recover_add_to_cont_trans( - struct xlog *log, - struct xlog_recover *trans, - xfs_caddr_t dp, - int len) -{ - xlog_recover_item_t *item; - xfs_caddr_t ptr, old_ptr; - int old_len; - - if (list_empty(&trans->r_itemq)) { - /* finish copying rest of trans header */ - xlog_recover_add_item(&trans->r_itemq); - ptr = (xfs_caddr_t) &trans->r_theader + - sizeof(xfs_trans_header_t) - len; - memcpy(ptr, dp, len); /* d, s, l */ - return 0; - } - /* take the tail entry */ - item = list_entry(trans->r_itemq.prev, xlog_recover_item_t, ri_list); - - old_ptr = item->ri_buf[item->ri_cnt-1].i_addr; - old_len = item->ri_buf[item->ri_cnt-1].i_len; - - ptr = kmem_realloc(old_ptr, len+old_len, old_len, KM_SLEEP); - memcpy(&ptr[old_len], dp, len); /* d, s, l */ - item->ri_buf[item->ri_cnt-1].i_len += len; - item->ri_buf[item->ri_cnt-1].i_addr = ptr; - trace_xfs_log_recover_item_add_cont(log, trans, item, 0); - return 0; -} - -/* - * The next region to add is the start of a new region. It could be - * a whole region or it could be the first part of a new region. Because - * of this, the assumption here is that the type and size fields of all - * format structures fit into the first 32 bits of the structure. - * - * This works because all regions must be 32 bit aligned. Therefore, we - * either have both fields or we have neither field. In the case we have - * neither field, the data part of the region is zero length. We only have - * a log_op_header and can throw away the header since a new one will appear - * later. If we have at least 4 bytes, then we can determine how many regions - * will appear in the current log item. - */ -STATIC int -xlog_recover_add_to_trans( - struct xlog *log, - struct xlog_recover *trans, - xfs_caddr_t dp, - int len) -{ - xfs_inode_log_format_t *in_f; /* any will do */ - xlog_recover_item_t *item; - xfs_caddr_t ptr; - - if (!len) - return 0; - if (list_empty(&trans->r_itemq)) { - /* we need to catch log corruptions here */ - if (*(uint *)dp != XFS_TRANS_HEADER_MAGIC) { - xfs_warn(log->l_mp, "%s: bad header magic number", - __func__); - ASSERT(0); - return XFS_ERROR(EIO); - } - if (len == sizeof(xfs_trans_header_t)) - xlog_recover_add_item(&trans->r_itemq); - memcpy(&trans->r_theader, dp, len); /* d, s, l */ - return 0; - } - - ptr = kmem_alloc(len, KM_SLEEP); - memcpy(ptr, dp, len); - in_f = (xfs_inode_log_format_t *)ptr; - - /* take the tail entry */ - item = list_entry(trans->r_itemq.prev, xlog_recover_item_t, ri_list); - if (item->ri_total != 0 && - item->ri_total == item->ri_cnt) { - /* tail item is in use, get a new one */ - xlog_recover_add_item(&trans->r_itemq); - item = list_entry(trans->r_itemq.prev, - xlog_recover_item_t, ri_list); - } - - if (item->ri_total == 0) { /* first region to be added */ - if (in_f->ilf_size == 0 || - in_f->ilf_size > XLOG_MAX_REGIONS_IN_ITEM) { - xfs_warn(log->l_mp, - "bad number of regions (%d) in inode log format", - in_f->ilf_size); - ASSERT(0); - kmem_free(ptr); - return XFS_ERROR(EIO); - } - - item->ri_total = in_f->ilf_size; - item->ri_buf = - kmem_zalloc(item->ri_total * sizeof(xfs_log_iovec_t), - KM_SLEEP); - } - ASSERT(item->ri_total > item->ri_cnt); - /* Description region is ri_buf[0] */ - item->ri_buf[item->ri_cnt].i_addr = ptr; - item->ri_buf[item->ri_cnt].i_len = len; - item->ri_cnt++; - trace_xfs_log_recover_item_add(log, trans, item, 0); - return 0; -} - /* * Sort the log items in the transaction. * @@ -1702,7 +1547,7 @@ xlog_recover_reorder_trans( */ if (!list_empty(&sort_list)) list_splice_init(&sort_list, &trans->r_itemq); - error = XFS_ERROR(EIO); + error = -EIO; goto out; } } @@ -1943,7 +1788,7 @@ xlog_recover_do_inode_buffer( item, bp); XFS_ERROR_REPORT("xlog_recover_do_inode_buf", XFS_ERRLEVEL_LOW, mp); - return XFS_ERROR(EFSCORRUPTED); + return -EFSCORRUPTED; } buffer_nextp = (xfs_agino_t *)xfs_buf_offset(bp, @@ -2125,6 +1970,17 @@ xlog_recover_validate_buf_type( __uint16_t magic16; __uint16_t magicda; + /* + * We can only do post recovery validation on items on CRC enabled + * fielsystems as we need to know when the buffer was written to be able + * to determine if we should have replayed the item. If we replay old + * metadata over a newer buffer, then it will enter a temporarily + * inconsistent state resulting in verification failures. Hence for now + * just avoid the verification stage for non-crc filesystems + */ + if (!xfs_sb_version_hascrc(&mp->m_sb)) + return; + magic32 = be32_to_cpu(*(__be32 *)bp->b_addr); magic16 = be16_to_cpu(*(__be16*)bp->b_addr); magicda = be16_to_cpu(info->magic); @@ -2162,8 +2018,6 @@ xlog_recover_validate_buf_type( bp->b_ops = &xfs_agf_buf_ops; break; case XFS_BLFT_AGFL_BUF: - if (!xfs_sb_version_hascrc(&mp->m_sb)) - break; if (magic32 != XFS_AGFL_MAGIC) { xfs_warn(mp, "Bad AGFL block magic!"); ASSERT(0); @@ -2196,10 +2050,6 @@ xlog_recover_validate_buf_type( #endif break; case XFS_BLFT_DINO_BUF: - /* - * we get here with inode allocation buffers, not buffers that - * track unlinked list changes. - */ if (magic16 != XFS_DINODE_MAGIC) { xfs_warn(mp, "Bad INODE block magic!"); ASSERT(0); @@ -2279,8 +2129,6 @@ xlog_recover_validate_buf_type( bp->b_ops = &xfs_attr3_leaf_buf_ops; break; case XFS_BLFT_ATTR_RMT_BUF: - if (!xfs_sb_version_hascrc(&mp->m_sb)) - break; if (magic32 != XFS_ATTR3_RMT_MAGIC) { xfs_warn(mp, "Bad attr remote magic!"); ASSERT(0); @@ -2387,16 +2235,7 @@ xlog_recover_do_reg_buffer( /* Shouldn't be any more regions */ ASSERT(i == item->ri_total); - /* - * We can only do post recovery validation on items on CRC enabled - * fielsystems as we need to know when the buffer was written to be able - * to determine if we should have replayed the item. If we replay old - * metadata over a newer buffer, then it will enter a temporarily - * inconsistent state resulting in verification failures. Hence for now - * just avoid the verification stage for non-crc filesystems - */ - if (xfs_sb_version_hascrc(&mp->m_sb)) - xlog_recover_validate_buf_type(mp, bp, buf_f); + xlog_recover_validate_buf_type(mp, bp, buf_f); } /* @@ -2404,8 +2243,11 @@ xlog_recover_do_reg_buffer( * Simple algorithm: if we have found a QUOTAOFF log item of the same type * (ie. USR or GRP), then just toss this buffer away; don't recover it. * Else, treat it as a regular buffer and do recovery. + * + * Return false if the buffer was tossed and true if we recovered the buffer to + * indicate to the caller if the buffer needs writing. */ -STATIC void +STATIC bool xlog_recover_do_dquot_buffer( struct xfs_mount *mp, struct xlog *log, @@ -2420,9 +2262,8 @@ xlog_recover_do_dquot_buffer( /* * Filesystems are required to send in quota flags at mount time. */ - if (mp->m_qflags == 0) { - return; - } + if (!mp->m_qflags) + return false; type = 0; if (buf_f->blf_flags & XFS_BLF_UDQUOT_BUF) @@ -2435,9 +2276,10 @@ xlog_recover_do_dquot_buffer( * This type of quotas was turned off, so ignore this buffer */ if (log->l_quotaoffs_flag & type) - return; + return false; xlog_recover_do_reg_buffer(mp, item, bp, buf_f); + return true; } /* @@ -2496,7 +2338,7 @@ xlog_recover_buffer_pass2( bp = xfs_buf_read(mp->m_ddev_targp, buf_f->blf_blkno, buf_f->blf_len, buf_flags, NULL); if (!bp) - return XFS_ERROR(ENOMEM); + return -ENOMEM; error = bp->b_error; if (error) { xfs_buf_ioerror_alert(bp, "xlog_recover_do..(read#1)"); @@ -2504,23 +2346,44 @@ xlog_recover_buffer_pass2( } /* - * recover the buffer only if we get an LSN from it and it's less than + * Recover the buffer only if we get an LSN from it and it's less than * the lsn of the transaction we are replaying. + * + * Note that we have to be extremely careful of readahead here. + * Readahead does not attach verfiers to the buffers so if we don't + * actually do any replay after readahead because of the LSN we found + * in the buffer if more recent than that current transaction then we + * need to attach the verifier directly. Failure to do so can lead to + * future recovery actions (e.g. EFI and unlinked list recovery) can + * operate on the buffers and they won't get the verifier attached. This + * can lead to blocks on disk having the correct content but a stale + * CRC. + * + * It is safe to assume these clean buffers are currently up to date. + * If the buffer is dirtied by a later transaction being replayed, then + * the verifier will be reset to match whatever recover turns that + * buffer into. */ lsn = xlog_recover_get_buf_lsn(mp, bp); - if (lsn && lsn != -1 && XFS_LSN_CMP(lsn, current_lsn) >= 0) + if (lsn && lsn != -1 && XFS_LSN_CMP(lsn, current_lsn) >= 0) { + xlog_recover_validate_buf_type(mp, bp, buf_f); goto out_release; + } if (buf_f->blf_flags & XFS_BLF_INODE_BUF) { error = xlog_recover_do_inode_buffer(mp, item, bp, buf_f); + if (error) + goto out_release; } else if (buf_f->blf_flags & (XFS_BLF_UDQUOT_BUF|XFS_BLF_PDQUOT_BUF|XFS_BLF_GDQUOT_BUF)) { - xlog_recover_do_dquot_buffer(mp, log, item, bp, buf_f); + bool dirty; + + dirty = xlog_recover_do_dquot_buffer(mp, log, item, bp, buf_f); + if (!dirty) + goto out_release; } else { xlog_recover_do_reg_buffer(mp, item, bp, buf_f); } - if (error) - goto out_release; /* * Perform delayed write on the buffer. Asynchronous writes will be @@ -2598,7 +2461,7 @@ xfs_recover_inode_owner_change( ip = xfs_inode_alloc(mp, in_f->ilf_ino); if (!ip) - return ENOMEM; + return -ENOMEM; /* instantiate the inode */ xfs_dinode_from_disk(&ip->i_d, dip); @@ -2676,7 +2539,7 @@ xlog_recover_inode_pass2( bp = xfs_buf_read(mp->m_ddev_targp, in_f->ilf_blkno, in_f->ilf_len, 0, &xfs_inode_buf_ops); if (!bp) { - error = ENOMEM; + error = -ENOMEM; goto error; } error = bp->b_error; @@ -2697,7 +2560,7 @@ xlog_recover_inode_pass2( __func__, dip, bp, in_f->ilf_ino); XFS_ERROR_REPORT("xlog_recover_inode_pass2(1)", XFS_ERRLEVEL_LOW, mp); - error = EFSCORRUPTED; + error = -EFSCORRUPTED; goto out_release; } dicp = item->ri_buf[1].i_addr; @@ -2707,7 +2570,7 @@ xlog_recover_inode_pass2( __func__, item, in_f->ilf_ino); XFS_ERROR_REPORT("xlog_recover_inode_pass2(2)", XFS_ERRLEVEL_LOW, mp); - error = EFSCORRUPTED; + error = -EFSCORRUPTED; goto out_release; } @@ -2764,7 +2627,7 @@ xlog_recover_inode_pass2( "%s: Bad regular inode log record, rec ptr 0x%p, " "ino ptr = 0x%p, ino bp = 0x%p, ino %Ld", __func__, item, dip, bp, in_f->ilf_ino); - error = EFSCORRUPTED; + error = -EFSCORRUPTED; goto out_release; } } else if (unlikely(S_ISDIR(dicp->di_mode))) { @@ -2777,7 +2640,7 @@ xlog_recover_inode_pass2( "%s: Bad dir inode log record, rec ptr 0x%p, " "ino ptr = 0x%p, ino bp = 0x%p, ino %Ld", __func__, item, dip, bp, in_f->ilf_ino); - error = EFSCORRUPTED; + error = -EFSCORRUPTED; goto out_release; } } @@ -2790,7 +2653,7 @@ xlog_recover_inode_pass2( __func__, item, dip, bp, in_f->ilf_ino, dicp->di_nextents + dicp->di_anextents, dicp->di_nblocks); - error = EFSCORRUPTED; + error = -EFSCORRUPTED; goto out_release; } if (unlikely(dicp->di_forkoff > mp->m_sb.sb_inodesize)) { @@ -2800,7 +2663,7 @@ xlog_recover_inode_pass2( "%s: Bad inode log record, rec ptr 0x%p, dino ptr 0x%p, " "dino bp 0x%p, ino %Ld, forkoff 0x%x", __func__, item, dip, bp, in_f->ilf_ino, dicp->di_forkoff); - error = EFSCORRUPTED; + error = -EFSCORRUPTED; goto out_release; } isize = xfs_icdinode_size(dicp->di_version); @@ -2810,7 +2673,7 @@ xlog_recover_inode_pass2( xfs_alert(mp, "%s: Bad inode log record length %d, rec ptr 0x%p", __func__, item->ri_buf[1].i_len, item); - error = EFSCORRUPTED; + error = -EFSCORRUPTED; goto out_release; } @@ -2898,7 +2761,7 @@ xlog_recover_inode_pass2( default: xfs_warn(log->l_mp, "%s: Invalid flag", __func__); ASSERT(0); - error = EIO; + error = -EIO; goto out_release; } } @@ -2919,7 +2782,7 @@ out_release: error: if (need_free) kmem_free(in_f); - return XFS_ERROR(error); + return error; } /* @@ -2946,7 +2809,7 @@ xlog_recover_quotaoff_pass1( if (qoff_f->qf_flags & XFS_GQUOTA_ACCT) log->l_quotaoffs_flag |= XFS_DQ_GROUP; - return (0); + return 0; } /* @@ -2971,17 +2834,17 @@ xlog_recover_dquot_pass2( * Filesystems are required to send in quota flags at mount time. */ if (mp->m_qflags == 0) - return (0); + return 0; recddq = item->ri_buf[1].i_addr; if (recddq == NULL) { xfs_alert(log->l_mp, "NULL dquot in %s.", __func__); - return XFS_ERROR(EIO); + return -EIO; } if (item->ri_buf[1].i_len < sizeof(xfs_disk_dquot_t)) { xfs_alert(log->l_mp, "dquot too small (%d) in %s.", item->ri_buf[1].i_len, __func__); - return XFS_ERROR(EIO); + return -EIO; } /* @@ -2990,7 +2853,7 @@ xlog_recover_dquot_pass2( type = recddq->d_flags & (XFS_DQ_USER | XFS_DQ_PROJ | XFS_DQ_GROUP); ASSERT(type); if (log->l_quotaoffs_flag & type) - return (0); + return 0; /* * At this point we know that quota was _not_ turned off. @@ -3007,12 +2870,19 @@ xlog_recover_dquot_pass2( error = xfs_dqcheck(mp, recddq, dq_f->qlf_id, 0, XFS_QMOPT_DOWARN, "xlog_recover_dquot_pass2 (log copy)"); if (error) - return XFS_ERROR(EIO); + return -EIO; ASSERT(dq_f->qlf_len == 1); + /* + * At this point we are assuming that the dquots have been allocated + * and hence the buffer has valid dquots stamped in it. It should, + * therefore, pass verifier validation. If the dquot is bad, then the + * we'll return an error here, so we don't need to specifically check + * the dquot in the buffer after the verifier has run. + */ error = xfs_trans_read_buf(mp, NULL, mp->m_ddev_targp, dq_f->qlf_blkno, XFS_FSB_TO_BB(mp, dq_f->qlf_len), 0, &bp, - NULL); + &xfs_dquot_buf_ops); if (error) return error; @@ -3020,18 +2890,6 @@ xlog_recover_dquot_pass2( ddq = (xfs_disk_dquot_t *)xfs_buf_offset(bp, dq_f->qlf_boffset); /* - * At least the magic num portion should be on disk because this - * was among a chunk of dquots created earlier, and we did some - * minimal initialization then. - */ - error = xfs_dqcheck(mp, ddq, dq_f->qlf_id, 0, XFS_QMOPT_DOWARN, - "xlog_recover_dquot_pass2"); - if (error) { - xfs_buf_relse(bp); - return XFS_ERROR(EIO); - } - - /* * If the dquot has an LSN in it, recover the dquot only if it's less * than the lsn of the transaction we are replaying. */ @@ -3178,38 +3036,38 @@ xlog_recover_do_icreate_pass2( icl = (struct xfs_icreate_log *)item->ri_buf[0].i_addr; if (icl->icl_type != XFS_LI_ICREATE) { xfs_warn(log->l_mp, "xlog_recover_do_icreate_trans: bad type"); - return EINVAL; + return -EINVAL; } if (icl->icl_size != 1) { xfs_warn(log->l_mp, "xlog_recover_do_icreate_trans: bad icl size"); - return EINVAL; + return -EINVAL; } agno = be32_to_cpu(icl->icl_ag); if (agno >= mp->m_sb.sb_agcount) { xfs_warn(log->l_mp, "xlog_recover_do_icreate_trans: bad agno"); - return EINVAL; + return -EINVAL; } agbno = be32_to_cpu(icl->icl_agbno); if (!agbno || agbno == NULLAGBLOCK || agbno >= mp->m_sb.sb_agblocks) { xfs_warn(log->l_mp, "xlog_recover_do_icreate_trans: bad agbno"); - return EINVAL; + return -EINVAL; } isize = be32_to_cpu(icl->icl_isize); if (isize != mp->m_sb.sb_inodesize) { xfs_warn(log->l_mp, "xlog_recover_do_icreate_trans: bad isize"); - return EINVAL; + return -EINVAL; } count = be32_to_cpu(icl->icl_count); if (!count) { xfs_warn(log->l_mp, "xlog_recover_do_icreate_trans: bad count"); - return EINVAL; + return -EINVAL; } length = be32_to_cpu(icl->icl_length); if (!length || length >= mp->m_sb.sb_agblocks) { xfs_warn(log->l_mp, "xlog_recover_do_icreate_trans: bad length"); - return EINVAL; + return -EINVAL; } /* existing allocation is fixed value */ @@ -3218,7 +3076,7 @@ xlog_recover_do_icreate_pass2( if (count != mp->m_ialloc_inos || length != mp->m_ialloc_blks) { xfs_warn(log->l_mp, "xlog_recover_do_icreate_trans: bad count 2"); - return EINVAL; + return -EINVAL; } /* @@ -3240,31 +3098,6 @@ xlog_recover_do_icreate_pass2( return 0; } -/* - * Free up any resources allocated by the transaction - * - * Remember that EFIs, EFDs, and IUNLINKs are handled later. - */ -STATIC void -xlog_recover_free_trans( - struct xlog_recover *trans) -{ - xlog_recover_item_t *item, *n; - int i; - - list_for_each_entry_safe(item, n, &trans->r_itemq, ri_list) { - /* Free the regions in the item. */ - list_del(&item->ri_list); - for (i = 0; i < item->ri_cnt; i++) - kmem_free(item->ri_buf[i].i_addr); - /* Free the item itself */ - kmem_free(item->ri_buf); - kmem_free(item); - } - /* Free the transaction recover structure */ - kmem_free(trans); -} - STATIC void xlog_recover_buffer_ra_pass2( struct xlog *log, @@ -3389,7 +3222,7 @@ xlog_recover_commit_pass1( xfs_warn(log->l_mp, "%s: invalid item type (%d)", __func__, ITEM_TYPE(item)); ASSERT(0); - return XFS_ERROR(EIO); + return -EIO; } } @@ -3425,7 +3258,7 @@ xlog_recover_commit_pass2( xfs_warn(log->l_mp, "%s: invalid item type (%d)", __func__, ITEM_TYPE(item)); ASSERT(0); - return XFS_ERROR(EIO); + return -EIO; } } @@ -3514,22 +3347,309 @@ out: if (!list_empty(&done_list)) list_splice_init(&done_list, &trans->r_itemq); - xlog_recover_free_trans(trans); - error2 = xfs_buf_delwri_submit(&buffer_list); return error ? error : error2; } +STATIC void +xlog_recover_add_item( + struct list_head *head) +{ + xlog_recover_item_t *item; + + item = kmem_zalloc(sizeof(xlog_recover_item_t), KM_SLEEP); + INIT_LIST_HEAD(&item->ri_list); + list_add_tail(&item->ri_list, head); +} + STATIC int -xlog_recover_unmount_trans( - struct xlog *log) +xlog_recover_add_to_cont_trans( + struct xlog *log, + struct xlog_recover *trans, + xfs_caddr_t dp, + int len) { - /* Do nothing now */ - xfs_warn(log->l_mp, "%s: Unmount LR", __func__); + xlog_recover_item_t *item; + xfs_caddr_t ptr, old_ptr; + int old_len; + + if (list_empty(&trans->r_itemq)) { + /* finish copying rest of trans header */ + xlog_recover_add_item(&trans->r_itemq); + ptr = (xfs_caddr_t) &trans->r_theader + + sizeof(xfs_trans_header_t) - len; + memcpy(ptr, dp, len); + return 0; + } + /* take the tail entry */ + item = list_entry(trans->r_itemq.prev, xlog_recover_item_t, ri_list); + + old_ptr = item->ri_buf[item->ri_cnt-1].i_addr; + old_len = item->ri_buf[item->ri_cnt-1].i_len; + + ptr = kmem_realloc(old_ptr, len+old_len, old_len, KM_SLEEP); + memcpy(&ptr[old_len], dp, len); + item->ri_buf[item->ri_cnt-1].i_len += len; + item->ri_buf[item->ri_cnt-1].i_addr = ptr; + trace_xfs_log_recover_item_add_cont(log, trans, item, 0); + return 0; +} + +/* + * The next region to add is the start of a new region. It could be + * a whole region or it could be the first part of a new region. Because + * of this, the assumption here is that the type and size fields of all + * format structures fit into the first 32 bits of the structure. + * + * This works because all regions must be 32 bit aligned. Therefore, we + * either have both fields or we have neither field. In the case we have + * neither field, the data part of the region is zero length. We only have + * a log_op_header and can throw away the header since a new one will appear + * later. If we have at least 4 bytes, then we can determine how many regions + * will appear in the current log item. + */ +STATIC int +xlog_recover_add_to_trans( + struct xlog *log, + struct xlog_recover *trans, + xfs_caddr_t dp, + int len) +{ + xfs_inode_log_format_t *in_f; /* any will do */ + xlog_recover_item_t *item; + xfs_caddr_t ptr; + + if (!len) + return 0; + if (list_empty(&trans->r_itemq)) { + /* we need to catch log corruptions here */ + if (*(uint *)dp != XFS_TRANS_HEADER_MAGIC) { + xfs_warn(log->l_mp, "%s: bad header magic number", + __func__); + ASSERT(0); + return -EIO; + } + if (len == sizeof(xfs_trans_header_t)) + xlog_recover_add_item(&trans->r_itemq); + memcpy(&trans->r_theader, dp, len); + return 0; + } + + ptr = kmem_alloc(len, KM_SLEEP); + memcpy(ptr, dp, len); + in_f = (xfs_inode_log_format_t *)ptr; + + /* take the tail entry */ + item = list_entry(trans->r_itemq.prev, xlog_recover_item_t, ri_list); + if (item->ri_total != 0 && + item->ri_total == item->ri_cnt) { + /* tail item is in use, get a new one */ + xlog_recover_add_item(&trans->r_itemq); + item = list_entry(trans->r_itemq.prev, + xlog_recover_item_t, ri_list); + } + + if (item->ri_total == 0) { /* first region to be added */ + if (in_f->ilf_size == 0 || + in_f->ilf_size > XLOG_MAX_REGIONS_IN_ITEM) { + xfs_warn(log->l_mp, + "bad number of regions (%d) in inode log format", + in_f->ilf_size); + ASSERT(0); + kmem_free(ptr); + return -EIO; + } + + item->ri_total = in_f->ilf_size; + item->ri_buf = + kmem_zalloc(item->ri_total * sizeof(xfs_log_iovec_t), + KM_SLEEP); + } + ASSERT(item->ri_total > item->ri_cnt); + /* Description region is ri_buf[0] */ + item->ri_buf[item->ri_cnt].i_addr = ptr; + item->ri_buf[item->ri_cnt].i_len = len; + item->ri_cnt++; + trace_xfs_log_recover_item_add(log, trans, item, 0); return 0; } /* + * Free up any resources allocated by the transaction + * + * Remember that EFIs, EFDs, and IUNLINKs are handled later. + */ +STATIC void +xlog_recover_free_trans( + struct xlog_recover *trans) +{ + xlog_recover_item_t *item, *n; + int i; + + list_for_each_entry_safe(item, n, &trans->r_itemq, ri_list) { + /* Free the regions in the item. */ + list_del(&item->ri_list); + for (i = 0; i < item->ri_cnt; i++) + kmem_free(item->ri_buf[i].i_addr); + /* Free the item itself */ + kmem_free(item->ri_buf); + kmem_free(item); + } + /* Free the transaction recover structure */ + kmem_free(trans); +} + +/* + * On error or completion, trans is freed. + */ +STATIC int +xlog_recovery_process_trans( + struct xlog *log, + struct xlog_recover *trans, + xfs_caddr_t dp, + unsigned int len, + unsigned int flags, + int pass) +{ + int error = 0; + bool freeit = false; + + /* mask off ophdr transaction container flags */ + flags &= ~XLOG_END_TRANS; + if (flags & XLOG_WAS_CONT_TRANS) + flags &= ~XLOG_CONTINUE_TRANS; + + /* + * Callees must not free the trans structure. We'll decide if we need to + * free it or not based on the operation being done and it's result. + */ + switch (flags) { + /* expected flag values */ + case 0: + case XLOG_CONTINUE_TRANS: + error = xlog_recover_add_to_trans(log, trans, dp, len); + break; + case XLOG_WAS_CONT_TRANS: + error = xlog_recover_add_to_cont_trans(log, trans, dp, len); + break; + case XLOG_COMMIT_TRANS: + error = xlog_recover_commit_trans(log, trans, pass); + /* success or fail, we are now done with this transaction. */ + freeit = true; + break; + + /* unexpected flag values */ + case XLOG_UNMOUNT_TRANS: + /* just skip trans */ + xfs_warn(log->l_mp, "%s: Unmount LR", __func__); + freeit = true; + break; + case XLOG_START_TRANS: + default: + xfs_warn(log->l_mp, "%s: bad flag 0x%x", __func__, flags); + ASSERT(0); + error = -EIO; + break; + } + if (error || freeit) + xlog_recover_free_trans(trans); + return error; +} + +/* + * Lookup the transaction recovery structure associated with the ID in the + * current ophdr. If the transaction doesn't exist and the start flag is set in + * the ophdr, then allocate a new transaction for future ID matches to find. + * Either way, return what we found during the lookup - an existing transaction + * or nothing. + */ +STATIC struct xlog_recover * +xlog_recover_ophdr_to_trans( + struct hlist_head rhash[], + struct xlog_rec_header *rhead, + struct xlog_op_header *ohead) +{ + struct xlog_recover *trans; + xlog_tid_t tid; + struct hlist_head *rhp; + + tid = be32_to_cpu(ohead->oh_tid); + rhp = &rhash[XLOG_RHASH(tid)]; + hlist_for_each_entry(trans, rhp, r_list) { + if (trans->r_log_tid == tid) + return trans; + } + + /* + * skip over non-start transaction headers - we could be + * processing slack space before the next transaction starts + */ + if (!(ohead->oh_flags & XLOG_START_TRANS)) + return NULL; + + ASSERT(be32_to_cpu(ohead->oh_len) == 0); + + /* + * This is a new transaction so allocate a new recovery container to + * hold the recovery ops that will follow. + */ + trans = kmem_zalloc(sizeof(struct xlog_recover), KM_SLEEP); + trans->r_log_tid = tid; + trans->r_lsn = be64_to_cpu(rhead->h_lsn); + INIT_LIST_HEAD(&trans->r_itemq); + INIT_HLIST_NODE(&trans->r_list); + hlist_add_head(&trans->r_list, rhp); + + /* + * Nothing more to do for this ophdr. Items to be added to this new + * transaction will be in subsequent ophdr containers. + */ + return NULL; +} + +STATIC int +xlog_recover_process_ophdr( + struct xlog *log, + struct hlist_head rhash[], + struct xlog_rec_header *rhead, + struct xlog_op_header *ohead, + xfs_caddr_t dp, + xfs_caddr_t end, + int pass) +{ + struct xlog_recover *trans; + unsigned int len; + + /* Do we understand who wrote this op? */ + if (ohead->oh_clientid != XFS_TRANSACTION && + ohead->oh_clientid != XFS_LOG) { + xfs_warn(log->l_mp, "%s: bad clientid 0x%x", + __func__, ohead->oh_clientid); + ASSERT(0); + return -EIO; + } + + /* + * Check the ophdr contains all the data it is supposed to contain. + */ + len = be32_to_cpu(ohead->oh_len); + if (dp + len > end) { + xfs_warn(log->l_mp, "%s: bad length 0x%x", __func__, len); + WARN_ON(1); + return -EIO; + } + + trans = xlog_recover_ophdr_to_trans(rhash, rhead, ohead); + if (!trans) { + /* nothing to do, so skip over this ophdr */ + return 0; + } + + return xlog_recovery_process_trans(log, trans, dp, len, + ohead->oh_flags, pass); +} + +/* * There are two valid states of the r_state field. 0 indicates that the * transaction structure is in a normal state. We have either seen the * start of the transaction or the last operation we added was not a partial @@ -3546,86 +3666,30 @@ xlog_recover_process_data( xfs_caddr_t dp, int pass) { - xfs_caddr_t lp; + struct xlog_op_header *ohead; + xfs_caddr_t end; int num_logops; - xlog_op_header_t *ohead; - xlog_recover_t *trans; - xlog_tid_t tid; int error; - unsigned long hash; - uint flags; - lp = dp + be32_to_cpu(rhead->h_len); + end = dp + be32_to_cpu(rhead->h_len); num_logops = be32_to_cpu(rhead->h_num_logops); /* check the log format matches our own - else we can't recover */ if (xlog_header_check_recover(log->l_mp, rhead)) - return (XFS_ERROR(EIO)); - - while ((dp < lp) && num_logops) { - ASSERT(dp + sizeof(xlog_op_header_t) <= lp); - ohead = (xlog_op_header_t *)dp; - dp += sizeof(xlog_op_header_t); - if (ohead->oh_clientid != XFS_TRANSACTION && - ohead->oh_clientid != XFS_LOG) { - xfs_warn(log->l_mp, "%s: bad clientid 0x%x", - __func__, ohead->oh_clientid); - ASSERT(0); - return (XFS_ERROR(EIO)); - } - tid = be32_to_cpu(ohead->oh_tid); - hash = XLOG_RHASH(tid); - trans = xlog_recover_find_tid(&rhash[hash], tid); - if (trans == NULL) { /* not found; add new tid */ - if (ohead->oh_flags & XLOG_START_TRANS) - xlog_recover_new_tid(&rhash[hash], tid, - be64_to_cpu(rhead->h_lsn)); - } else { - if (dp + be32_to_cpu(ohead->oh_len) > lp) { - xfs_warn(log->l_mp, "%s: bad length 0x%x", - __func__, be32_to_cpu(ohead->oh_len)); - WARN_ON(1); - return (XFS_ERROR(EIO)); - } - flags = ohead->oh_flags & ~XLOG_END_TRANS; - if (flags & XLOG_WAS_CONT_TRANS) - flags &= ~XLOG_CONTINUE_TRANS; - switch (flags) { - case XLOG_COMMIT_TRANS: - error = xlog_recover_commit_trans(log, - trans, pass); - break; - case XLOG_UNMOUNT_TRANS: - error = xlog_recover_unmount_trans(log); - break; - case XLOG_WAS_CONT_TRANS: - error = xlog_recover_add_to_cont_trans(log, - trans, dp, - be32_to_cpu(ohead->oh_len)); - break; - case XLOG_START_TRANS: - xfs_warn(log->l_mp, "%s: bad transaction", - __func__); - ASSERT(0); - error = XFS_ERROR(EIO); - break; - case 0: - case XLOG_CONTINUE_TRANS: - error = xlog_recover_add_to_trans(log, trans, - dp, be32_to_cpu(ohead->oh_len)); - break; - default: - xfs_warn(log->l_mp, "%s: bad flag 0x%x", - __func__, flags); - ASSERT(0); - error = XFS_ERROR(EIO); - break; - } - if (error) { - xlog_recover_free_trans(trans); - return error; - } - } + return -EIO; + + while ((dp < end) && num_logops) { + + ohead = (struct xlog_op_header *)dp; + dp += sizeof(*ohead); + ASSERT(dp <= end); + + /* errors will abort recovery */ + error = xlog_recover_process_ophdr(log, rhash, rhead, ohead, + dp, end, pass); + if (error) + return error; + dp += be32_to_cpu(ohead->oh_len); num_logops--; } @@ -3669,7 +3733,7 @@ xlog_recover_process_efi( */ set_bit(XFS_EFI_RECOVERED, &efip->efi_flags); xfs_efi_release(efip, efip->efi_format.efi_nextents); - return XFS_ERROR(EIO); + return -EIO; } } @@ -3969,7 +4033,7 @@ xlog_unpack_data_crc( * CRC protection by punting an error back up the stack. */ if (xfs_sb_version_hascrc(&log->l_mp->m_sb)) - return EFSCORRUPTED; + return -EFSCORRUPTED; } return 0; @@ -4018,14 +4082,14 @@ xlog_valid_rec_header( if (unlikely(rhead->h_magicno != cpu_to_be32(XLOG_HEADER_MAGIC_NUM))) { XFS_ERROR_REPORT("xlog_valid_rec_header(1)", XFS_ERRLEVEL_LOW, log->l_mp); - return XFS_ERROR(EFSCORRUPTED); + return -EFSCORRUPTED; } if (unlikely( (!rhead->h_version || (be32_to_cpu(rhead->h_version) & (~XLOG_VERSION_OKBITS))))) { xfs_warn(log->l_mp, "%s: unrecognised log version (%d).", __func__, be32_to_cpu(rhead->h_version)); - return XFS_ERROR(EIO); + return -EIO; } /* LR body must have data or it wouldn't have been written */ @@ -4033,12 +4097,12 @@ xlog_valid_rec_header( if (unlikely( hlen <= 0 || hlen > INT_MAX )) { XFS_ERROR_REPORT("xlog_valid_rec_header(2)", XFS_ERRLEVEL_LOW, log->l_mp); - return XFS_ERROR(EFSCORRUPTED); + return -EFSCORRUPTED; } if (unlikely( blkno > log->l_logBBsize || blkno > INT_MAX )) { XFS_ERROR_REPORT("xlog_valid_rec_header(3)", XFS_ERRLEVEL_LOW, log->l_mp); - return XFS_ERROR(EFSCORRUPTED); + return -EFSCORRUPTED; } return 0; } @@ -4081,7 +4145,7 @@ xlog_do_recovery_pass( */ hbp = xlog_get_bp(log, 1); if (!hbp) - return ENOMEM; + return -ENOMEM; error = xlog_bread(log, tail_blk, 1, hbp, &offset); if (error) @@ -4110,49 +4174,21 @@ xlog_do_recovery_pass( } if (!hbp) - return ENOMEM; + return -ENOMEM; dbp = xlog_get_bp(log, BTOBB(h_size)); if (!dbp) { xlog_put_bp(hbp); - return ENOMEM; + return -ENOMEM; } memset(rhash, 0, sizeof(rhash)); - if (tail_blk <= head_blk) { - for (blk_no = tail_blk; blk_no < head_blk; ) { - error = xlog_bread(log, blk_no, hblks, hbp, &offset); - if (error) - goto bread_err2; - - rhead = (xlog_rec_header_t *)offset; - error = xlog_valid_rec_header(log, rhead, blk_no); - if (error) - goto bread_err2; - - /* blocks in data section */ - bblks = (int)BTOBB(be32_to_cpu(rhead->h_len)); - error = xlog_bread(log, blk_no + hblks, bblks, dbp, - &offset); - if (error) - goto bread_err2; - - error = xlog_unpack_data(rhead, offset, log); - if (error) - goto bread_err2; - - error = xlog_recover_process_data(log, - rhash, rhead, offset, pass); - if (error) - goto bread_err2; - blk_no += bblks + hblks; - } - } else { + blk_no = tail_blk; + if (tail_blk > head_blk) { /* * Perform recovery around the end of the physical log. * When the head is not on the same cycle number as the tail, - * we can't do a sequential recovery as above. + * we can't do a sequential recovery. */ - blk_no = tail_blk; while (blk_no < log->l_logBBsize) { /* * Check for header wrapping around physical end-of-log @@ -4266,34 +4302,35 @@ xlog_do_recovery_pass( ASSERT(blk_no >= log->l_logBBsize); blk_no -= log->l_logBBsize; + } - /* read first part of physical log */ - while (blk_no < head_blk) { - error = xlog_bread(log, blk_no, hblks, hbp, &offset); - if (error) - goto bread_err2; + /* read first part of physical log */ + while (blk_no < head_blk) { + error = xlog_bread(log, blk_no, hblks, hbp, &offset); + if (error) + goto bread_err2; - rhead = (xlog_rec_header_t *)offset; - error = xlog_valid_rec_header(log, rhead, blk_no); - if (error) - goto bread_err2; + rhead = (xlog_rec_header_t *)offset; + error = xlog_valid_rec_header(log, rhead, blk_no); + if (error) + goto bread_err2; - bblks = (int)BTOBB(be32_to_cpu(rhead->h_len)); - error = xlog_bread(log, blk_no+hblks, bblks, dbp, - &offset); - if (error) - goto bread_err2; + /* blocks in data section */ + bblks = (int)BTOBB(be32_to_cpu(rhead->h_len)); + error = xlog_bread(log, blk_no+hblks, bblks, dbp, + &offset); + if (error) + goto bread_err2; - error = xlog_unpack_data(rhead, offset, log); - if (error) - goto bread_err2; + error = xlog_unpack_data(rhead, offset, log); + if (error) + goto bread_err2; - error = xlog_recover_process_data(log, rhash, - rhead, offset, pass); - if (error) - goto bread_err2; - blk_no += bblks + hblks; - } + error = xlog_recover_process_data(log, rhash, + rhead, offset, pass); + if (error) + goto bread_err2; + blk_no += bblks + hblks; } bread_err2: @@ -4388,7 +4425,7 @@ xlog_do_recover( * If IO errors happened during recovery, bail out. */ if (XFS_FORCED_SHUTDOWN(log->l_mp)) { - return (EIO); + return -EIO; } /* @@ -4413,16 +4450,12 @@ xlog_do_recover( XFS_BUF_UNASYNC(bp); bp->b_ops = &xfs_sb_buf_ops; - if (XFS_FORCED_SHUTDOWN(log->l_mp)) { - xfs_buf_relse(bp); - return XFS_ERROR(EIO); - } - - xfs_buf_iorequest(bp); - error = xfs_buf_iowait(bp); + error = xfs_buf_submit_wait(bp); if (error) { - xfs_buf_ioerror_alert(bp, __func__); - ASSERT(0); + if (!XFS_FORCED_SHUTDOWN(log->l_mp)) { + xfs_buf_ioerror_alert(bp, __func__); + ASSERT(0); + } xfs_buf_relse(bp); return error; } @@ -4492,7 +4525,19 @@ xlog_recover( "Please recover the log on a kernel that supports the unknown features.", (log->l_mp->m_sb.sb_features_log_incompat & XFS_SB_FEAT_INCOMPAT_LOG_UNKNOWN)); - return EINVAL; + return -EINVAL; + } + + /* + * Delay log recovery if the debug hook is set. This is debug + * instrumention to coordinate simulation of I/O failures with + * log recovery. + */ + if (xfs_globals.log_recovery_delay) { + xfs_notice(log->l_mp, + "Delaying log recovery for %d seconds.", + xfs_globals.log_recovery_delay); + msleep(xfs_globals.log_recovery_delay * 1000); } xfs_notice(log->l_mp, "Starting recovery (logdev: %s)", |