1 /* 2 * CDDL HEADER START 3 * 4 * The contents of this file are subject to the terms of the 5 * Common Development and Distribution License (the "License"). 6 * You may not use this file except in compliance with the License. 7 * 8 * You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE 9 * or http://www.opensolaris.org/os/licensing. 10 * See the License for the specific language governing permissions 11 * and limitations under the License. 12 * 13 * When distributing Covered Code, include this CDDL HEADER in each 14 * file and include the License file at usr/src/OPENSOLARIS.LICENSE. 15 * If applicable, add the following below this CDDL HEADER, with the 16 * fields enclosed by brackets "[]" replaced with your own identifying 17 * information: Portions Copyright [yyyy] [name of copyright owner] 18 * 19 * CDDL HEADER END 20 */ 21 /* 22 * Copyright (c) 2005, 2010, Oracle and/or its affiliates. All rights reserved. 23 */ 24 25 /* Portions Copyright 2010 Robert Milkowski */ 26 27 #ifndef _SYS_ZIL_IMPL_H 28 #define _SYS_ZIL_IMPL_H 29 30 #include <sys/zil.h> 31 #include <sys/dmu_objset.h> 32 33 #ifdef __cplusplus 34 extern "C" { 35 #endif 36 37 /* 38 * Log write buffer. 39 */ 40 typedef struct lwb { 41 zilog_t *lwb_zilog; /* back pointer to log struct */ 42 blkptr_t lwb_blk; /* on disk address of this log blk */ 43 int lwb_nused; /* # used bytes in buffer */ 44 int lwb_sz; /* size of block and buffer */ 45 char *lwb_buf; /* log write buffer */ 46 zio_t *lwb_zio; /* zio for this buffer */ 47 dmu_tx_t *lwb_tx; /* tx for log block allocation */ 48 uint64_t lwb_max_txg; /* highest txg in this lwb */ 49 list_node_t lwb_node; /* zilog->zl_lwb_list linkage */ 50 } lwb_t; 51 52 /* 53 * Intent log transaction lists 54 */ 55 typedef struct itxs { 56 list_t i_sync_list; /* list of synchronous itxs */ 57 avl_tree_t i_async_tree; /* tree of foids for async itxs */ 58 } itxs_t; 59 60 typedef struct itxg { 61 kmutex_t itxg_lock; /* lock for this structure */ 62 uint64_t itxg_txg; /* txg for this chain */ 63 uint64_t itxg_sod; /* total size on disk for this txg */ 64 itxs_t *itxg_itxs; /* sync and async itxs */ 65 } itxg_t; 66 67 /* for async nodes we build up an AVL tree of lists of async itxs per file */ 68 typedef struct itx_async_node { 69 uint64_t ia_foid; /* file object id */ 70 list_t ia_list; /* list of async itxs for this foid */ 71 avl_node_t ia_node; /* AVL tree linkage */ 72 } itx_async_node_t; 73 74 /* 75 * Vdev flushing: during a zil_commit(), we build up an AVL tree of the vdevs 76 * we've touched so we know which ones need a write cache flush at the end. 77 */ 78 typedef struct zil_vdev_node { 79 uint64_t zv_vdev; /* vdev to be flushed */ 80 avl_node_t zv_node; /* AVL tree linkage */ 81 } zil_vdev_node_t; 82 83 #define ZIL_PREV_BLKS 16 84 85 /* 86 * Stable storage intent log management structure. One per dataset. 87 */ 88 struct zilog { 89 kmutex_t zl_lock; /* protects most zilog_t fields */ 90 struct dsl_pool *zl_dmu_pool; /* DSL pool */ 91 spa_t *zl_spa; /* handle for read/write log */ 92 const zil_header_t *zl_header; /* log header buffer */ 93 objset_t *zl_os; /* object set we're logging */ 94 zil_get_data_t *zl_get_data; /* callback to get object content */ 95 zio_t *zl_root_zio; /* log writer root zio */ 96 uint64_t zl_lr_seq; /* on-disk log record sequence number */ 97 uint64_t zl_commit_lr_seq; /* last committed on-disk lr seq */ 98 uint64_t zl_destroy_txg; /* txg of last zil_destroy() */ 99 uint64_t zl_replayed_seq[TXG_SIZE]; /* last replayed rec seq */ 100 uint64_t zl_replaying_seq; /* current replay seq number */ 101 uint32_t zl_suspend; /* log suspend count */ 102 kcondvar_t zl_cv_writer; /* log writer thread completion */ 103 kcondvar_t zl_cv_suspend; /* log suspend completion */ 104 uint8_t zl_suspending; /* log is currently suspending */ 105 uint8_t zl_keep_first; /* keep first log block in destroy */ 106 uint8_t zl_replay; /* replaying records while set */ 107 uint8_t zl_stop_sync; /* for debugging */ 108 uint8_t zl_writer; /* boolean: write setup in progress */ 109 uint8_t zl_logbias; /* latency or throughput */ 110 uint8_t zl_sync; /* synchronous or asynchronous */ 111 int zl_parse_error; /* last zil_parse() error */ 112 uint64_t zl_parse_blk_seq; /* highest blk seq on last parse */ 113 uint64_t zl_parse_lr_seq; /* highest lr seq on last parse */ 114 uint64_t zl_parse_blk_count; /* number of blocks parsed */ 115 uint64_t zl_parse_lr_count; /* number of log records parsed */ 116 uint64_t zl_next_batch; /* next batch number */ 117 uint64_t zl_com_batch; /* committed batch number */ 118 kcondvar_t zl_cv_batch[2]; /* batch condition variables */ 119 itxg_t zl_itxg[TXG_SIZE]; /* intent log txg chains */ 120 list_t zl_itx_commit_list; /* itx list to be committed */ 121 uint64_t zl_itx_list_sz; /* total size of records on list */ 122 uint64_t zl_cur_used; /* current commit log size used */ 123 list_t zl_lwb_list; /* in-flight log write list */ 124 kmutex_t zl_vdev_lock; /* protects zl_vdev_tree */ 125 avl_tree_t zl_vdev_tree; /* vdevs to flush in zil_commit() */ 126 taskq_t *zl_clean_taskq; /* runs lwb and itx clean tasks */ 127 avl_tree_t zl_bp_tree; /* track bps during log parse */ 128 clock_t zl_replay_time; /* lbolt of when replay started */ 129 uint64_t zl_replay_blks; /* number of log blocks replayed */ 130 zil_header_t zl_old_header; /* debugging aid */ 131 uint_t zl_prev_blks[ZIL_PREV_BLKS]; /* size - sector rounded */ 132 uint_t zl_prev_rotor; /* rotor for zl_prev[] */ 133 }; 134 135 typedef struct zil_bp_node { 136 dva_t zn_dva; 137 avl_node_t zn_node; 138 } zil_bp_node_t; 139 140 #define ZIL_MAX_LOG_DATA (SPA_MAXBLOCKSIZE - sizeof (zil_chain_t) - \ 141 sizeof (lr_write_t)) 142 143 #ifdef __cplusplus 144 } 145 #endif 146 147 #endif /* _SYS_ZIL_IMPL_H */ 148