spa_impl.h revision 35a5a3587fd94b666239c157d3722745250ccbd7
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 2009 Sun Microsystems, Inc.  All rights reserved.
23 * Use is subject to license terms.
24 */
25
26#ifndef _SYS_SPA_IMPL_H
27#define	_SYS_SPA_IMPL_H
28
29#include <sys/spa.h>
30#include <sys/vdev.h>
31#include <sys/metaslab.h>
32#include <sys/dmu.h>
33#include <sys/dsl_pool.h>
34#include <sys/uberblock_impl.h>
35#include <sys/zfs_context.h>
36#include <sys/avl.h>
37#include <sys/refcount.h>
38#include <sys/bplist.h>
39
40#ifdef	__cplusplus
41extern "C" {
42#endif
43
44typedef struct spa_error_entry {
45	zbookmark_t	se_bookmark;
46	char		*se_name;
47	avl_node_t	se_avl;
48} spa_error_entry_t;
49
50typedef struct spa_history_phys {
51	uint64_t sh_pool_create_len;	/* ending offset of zpool create */
52	uint64_t sh_phys_max_off;	/* physical EOF */
53	uint64_t sh_bof;		/* logical BOF */
54	uint64_t sh_eof;		/* logical EOF */
55	uint64_t sh_records_lost;	/* num of records overwritten */
56} spa_history_phys_t;
57
58struct spa_aux_vdev {
59	uint64_t	sav_object;		/* MOS object for device list */
60	nvlist_t	*sav_config;		/* cached device config */
61	vdev_t		**sav_vdevs;		/* devices */
62	int		sav_count;		/* number devices */
63	boolean_t	sav_sync;		/* sync the device list */
64	nvlist_t	**sav_pending;		/* pending device additions */
65	uint_t		sav_npending;		/* # pending devices */
66};
67
68typedef struct spa_config_lock {
69	kmutex_t	scl_lock;
70	kthread_t	*scl_writer;
71	int		scl_write_wanted;
72	kcondvar_t	scl_cv;
73	refcount_t	scl_count;
74} spa_config_lock_t;
75
76typedef struct spa_config_dirent {
77	list_node_t	scd_link;
78	char		*scd_path;
79} spa_config_dirent_t;
80
81enum zio_taskq_type {
82	ZIO_TASKQ_ISSUE = 0,
83	ZIO_TASKQ_ISSUE_HIGH,
84	ZIO_TASKQ_INTERRUPT,
85	ZIO_TASKQ_INTERRUPT_HIGH,
86	ZIO_TASKQ_TYPES
87};
88
89/*
90 * State machine for the zpool-pooname process.  The states transitions
91 * are done as follows:
92 *
93 *	From		   To			Routine
94 *	PROC_NONE	-> PROC_CREATED		spa_activate()
95 *	PROC_CREATED	-> PROC_ACTIVE		spa_thread()
96 *	PROC_ACTIVE	-> PROC_DEACTIVATE	spa_deactivate()
97 *	PROC_DEACTIVATE	-> PROC_GONE		spa_thread()
98 *	PROC_GONE	-> PROC_NONE		spa_deactivate()
99 */
100typedef enum spa_proc_state {
101	SPA_PROC_NONE,		/* spa_proc = &p0, no process created */
102	SPA_PROC_CREATED,	/* spa_activate() has proc, is waiting */
103	SPA_PROC_ACTIVE,	/* taskqs created, spa_proc set */
104	SPA_PROC_DEACTIVATE,	/* spa_deactivate() requests process exit */
105	SPA_PROC_GONE		/* spa_thread() is exiting, spa_proc = &p0 */
106} spa_proc_state_t;
107
108struct spa {
109	/*
110	 * Fields protected by spa_namespace_lock.
111	 */
112	char		spa_name[MAXNAMELEN];	/* pool name */
113	avl_node_t	spa_avl;		/* node in spa_namespace_avl */
114	nvlist_t	*spa_config;		/* last synced config */
115	nvlist_t	*spa_config_syncing;	/* currently syncing config */
116	uint64_t	spa_config_txg;		/* txg of last config change */
117	int		spa_sync_pass;		/* iterate-to-convergence */
118	pool_state_t	spa_state;		/* pool state */
119	int		spa_inject_ref;		/* injection references */
120	uint8_t		spa_sync_on;		/* sync threads are running */
121	spa_load_state_t spa_load_state;	/* current load operation */
122	boolean_t	spa_load_verbatim;	/* load the given config? */
123	taskq_t		*spa_zio_taskq[ZIO_TYPES][ZIO_TASKQ_TYPES];
124	dsl_pool_t	*spa_dsl_pool;
125	metaslab_class_t *spa_normal_class;	/* normal data class */
126	metaslab_class_t *spa_log_class;	/* intent log data class */
127	uint64_t	spa_first_txg;		/* first txg after spa_open() */
128	uint64_t	spa_final_txg;		/* txg of export/destroy */
129	uint64_t	spa_freeze_txg;		/* freeze pool at this txg */
130	uint64_t	spa_load_max_txg;	/* best initial ub_txg */
131	uint64_t	spa_claim_max_txg;	/* highest claimed birth txg */
132	objset_t	*spa_meta_objset;	/* copy of dp->dp_meta_objset */
133	txg_list_t	spa_vdev_txg_list;	/* per-txg dirty vdev list */
134	vdev_t		*spa_root_vdev;		/* top-level vdev container */
135	uint64_t	spa_load_guid;		/* initial guid for spa_load */
136	list_t		spa_config_dirty_list;	/* vdevs with dirty config */
137	list_t		spa_state_dirty_list;	/* vdevs with dirty state */
138	spa_aux_vdev_t	spa_spares;		/* hot spares */
139	spa_aux_vdev_t	spa_l2cache;		/* L2ARC cache devices */
140	uint64_t	spa_config_object;	/* MOS object for pool config */
141	uint64_t	spa_config_generation;	/* config generation number */
142	uint64_t	spa_syncing_txg;	/* txg currently syncing */
143	uint64_t	spa_deferred_bplist_obj; /* object for deferred frees */
144	bplist_t	spa_deferred_bplist;	/* deferred-free bplist */
145	bplist_t	spa_free_bplist[TXG_SIZE]; /* bplist of stuff to free */
146	uberblock_t	spa_ubsync;		/* last synced uberblock */
147	uberblock_t	spa_uberblock;		/* current uberblock */
148	boolean_t	spa_extreme_rewind;	/* rewind past deferred frees */
149	kmutex_t	spa_scrub_lock;		/* resilver/scrub lock */
150	uint64_t	spa_scrub_inflight;	/* in-flight scrub I/Os */
151	uint64_t	spa_scrub_maxinflight;	/* max in-flight scrub I/Os */
152	uint64_t	spa_scrub_errors;	/* scrub I/O error count */
153	kcondvar_t	spa_scrub_io_cv;	/* scrub I/O completion */
154	uint8_t		spa_scrub_active;	/* active or suspended? */
155	uint8_t		spa_scrub_type;		/* type of scrub we're doing */
156	uint8_t		spa_scrub_finished;	/* indicator to rotate logs */
157	uint8_t		spa_scrub_started;	/* started since last boot */
158	uint8_t		spa_scrub_reopen;	/* scrub doing vdev_reopen */
159	kmutex_t	spa_async_lock;		/* protect async state */
160	kthread_t	*spa_async_thread;	/* thread doing async task */
161	int		spa_async_suspended;	/* async tasks suspended */
162	kcondvar_t	spa_async_cv;		/* wait for thread_exit() */
163	uint16_t	spa_async_tasks;	/* async task mask */
164	char		*spa_root;		/* alternate root directory */
165	uint64_t	spa_ena;		/* spa-wide ereport ENA */
166	int		spa_last_open_failed;	/* error if last open failed */
167	nvlist_t	*spa_failed_open_cfg;	/* cached config nvlist */
168	uint64_t	spa_last_ubsync_txg;	/* "best" uberblock txg */
169	uint64_t	spa_last_ubsync_txg_ts;	/* timestamp from that ub */
170	uint64_t	spa_load_txg;		/* ub txg that loaded */
171	uint64_t	spa_load_txg_ts;	/* timestamp from that ub */
172	uint64_t	spa_load_meta_errors;	/* verify metadata err count */
173	uint64_t	spa_load_data_errors;	/* verify data err count */
174	uint64_t	spa_verify_min_txg;	/* start txg of verify scrub */
175	kmutex_t	spa_errlog_lock;	/* error log lock */
176	uint64_t	spa_errlog_last;	/* last error log object */
177	uint64_t	spa_errlog_scrub;	/* scrub error log object */
178	kmutex_t	spa_errlist_lock;	/* error list/ereport lock */
179	avl_tree_t	spa_errlist_last;	/* last error list */
180	avl_tree_t	spa_errlist_scrub;	/* scrub error list */
181	uint64_t	spa_deflate;		/* should we deflate? */
182	uint64_t	spa_history;		/* history object */
183	kmutex_t	spa_history_lock;	/* history lock */
184	vdev_t		*spa_pending_vdev;	/* pending vdev additions */
185	kmutex_t	spa_props_lock;		/* property lock */
186	uint64_t	spa_pool_props_object;	/* object for properties */
187	uint64_t	spa_bootfs;		/* default boot filesystem */
188	uint64_t	spa_failmode;		/* failure mode for the pool */
189	uint64_t	spa_delegation;		/* delegation on/off */
190	list_t		spa_config_list;	/* previous cache file(s) */
191	zio_t		*spa_async_zio_root;	/* root of all async I/O */
192	zio_t		*spa_suspend_zio_root;	/* root of all suspended I/O */
193	kmutex_t	spa_suspend_lock;	/* protects suspend_zio_root */
194	kcondvar_t	spa_suspend_cv;		/* notification of resume */
195	uint8_t		spa_suspended;		/* pool is suspended */
196	uint8_t		spa_claiming;		/* pool is doing zil_claim() */
197	boolean_t	spa_is_root;		/* pool is root */
198	int		spa_minref;		/* num refs when first opened */
199	int		spa_mode;		/* FREAD | FWRITE */
200	spa_log_state_t spa_log_state;		/* log state */
201	uint64_t	spa_autoexpand;		/* lun expansion on/off */
202	ddt_t		*spa_ddt[ZIO_CHECKSUM_FUNCTIONS]; /* in-core DDTs */
203	uint64_t	spa_ddt_stat_object;	/* DDT statistics */
204	uint64_t	spa_dedup_ditto;	/* dedup ditto threshold */
205	uint64_t	spa_dedup_checksum;	/* default dedup checksum */
206	uint64_t	spa_dspace;		/* dspace in normal class */
207	kmutex_t	spa_vdev_top_lock;	/* dueling offline/remove */
208	kmutex_t	spa_proc_lock;		/* protects spa_proc* */
209	kcondvar_t	spa_proc_cv;		/* spa_proc_state transitions */
210	spa_proc_state_t spa_proc_state;	/* see definition */
211	struct proc	*spa_proc;		/* "zpool-poolname" process */
212	uint64_t	spa_did;		/* if procp != p0, did of t1 */
213	boolean_t	spa_autoreplace;	/* autoreplace set in open */
214	int		spa_vdev_locks;		/* locks grabbed */
215	/*
216	 * spa_refcnt & spa_config_lock must be the last elements
217	 * because refcount_t changes size based on compilation options.
218	 * In order for the MDB module to function correctly, the other
219	 * fields must remain in the same location.
220	 */
221	spa_config_lock_t spa_config_lock[SCL_LOCKS]; /* config changes */
222	refcount_t	spa_refcount;		/* number of opens */
223};
224
225extern const char *spa_config_path;
226
227#ifdef	__cplusplus
228}
229#endif
230
231#endif	/* _SYS_SPA_IMPL_H */
232