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 /*
  27  * Copyright (c) 2013 by Delphix. All rights reserved.
  28  */
  29 
  30 #ifndef _SYS_METASLAB_IMPL_H
  31 #define _SYS_METASLAB_IMPL_H
  32 
  33 #include <sys/metaslab.h>
  34 #include <sys/space_map.h>
  35 #include <sys/vdev.h>
  36 #include <sys/txg.h>
  37 #include <sys/avl.h>
  38 
  39 #ifdef  __cplusplus
  40 extern "C" {
  41 #endif
  42 
  43 struct metaslab_class {
  44         spa_t                   *mc_spa;
  45         metaslab_group_t        *mc_rotor;
  46         space_map_ops_t         *mc_ops;
  47         uint64_t                mc_aliquot;
  48         uint64_t                mc_alloc_groups; /* # of allocatable groups */
  49         uint64_t                mc_alloc;       /* total allocated space */
  50         uint64_t                mc_deferred;    /* total deferred frees */
  51         uint64_t                mc_space;       /* total space (alloc + free) */
  52         uint64_t                mc_dspace;      /* total deflated space */
  53 };
  54 
  55 struct metaslab_group {
  56         kmutex_t                mg_lock;
  57         avl_tree_t              mg_metaslab_tree;
  58         uint64_t                mg_aliquot;
  59         uint64_t                mg_bonus_area;
  60         uint64_t                mg_alloc_failures;
  61         boolean_t               mg_allocatable;         /* can we allocate? */
  62         uint64_t                mg_free_capacity;       /* percentage free */
  63         int64_t                 mg_bias;
  64         int64_t                 mg_activation_count;
  65         metaslab_class_t        *mg_class;
  66         vdev_t                  *mg_vd;
  67         metaslab_group_t        *mg_prev;
  68         metaslab_group_t        *mg_next;
  69 };
  70 
  71 /*
  72  * Each metaslab maintains an in-core free map (ms_map) that contains the
  73  * current list of free segments. As blocks are allocated, the allocated
  74  * segment is removed from the ms_map and added to a per txg allocation map.
  75  * As blocks are freed, they are added to the per txg free map. These per
  76  * txg maps allow us to process all allocations and frees in syncing context
  77  * where it is safe to update the on-disk space maps.
  78  *
  79  * Each metaslab's free space is tracked in a space map object in the MOS,
  80  * which is only updated in syncing context. Each time we sync a txg,
  81  * we append the allocs and frees from that txg to the space map object.
  82  * When the txg is done syncing, metaslab_sync_done() updates ms_smo
  83  * to ms_smo_syncing. Everything in ms_smo is always safe to allocate.
  84  *
  85  * To load the in-core free map we read the space map object from disk.
  86  * This object contains a series of alloc and free records that are
  87  * combined to make up the list of all free segments in this metaslab. These
  88  * segments are represented in-core by the ms_map and are stored in an
  89  * AVL tree.
  90  *
  91  * As the space map objects grows (as a result of the appends) it will
  92  * eventually become space-inefficient. When the space map object is
  93  * zfs_condense_pct/100 times the size of the minimal on-disk representation,
  94  * we rewrite it in its minimized form.
  95  */
  96 struct metaslab {
  97         kmutex_t        ms_lock;        /* metaslab lock                */
  98         space_map_obj_t ms_smo;         /* synced space map object      */
  99         space_map_obj_t ms_smo_syncing; /* syncing space map object     */
 100         space_map_t     *ms_allocmap[TXG_SIZE]; /* allocated this txg   */
 101         space_map_t     *ms_freemap[TXG_SIZE];  /* freed this txg       */
 102         space_map_t     *ms_defermap[TXG_DEFER_SIZE];   /* deferred frees */
 103         space_map_t     *ms_map;        /* in-core free space map       */
 104         int64_t         ms_deferspace;  /* sum of ms_defermap[] space   */
 105         uint64_t        ms_weight;      /* weight vs. others in group   */
 106         metaslab_group_t *ms_group;     /* metaslab group               */
 107         avl_node_t      ms_group_node;  /* node in metaslab group tree  */
 108         txg_node_t      ms_txg_node;    /* per-txg dirty metaslab links */
 109 };
 110 
 111 #ifdef  __cplusplus
 112 }
 113 #endif
 114 
 115 #endif  /* _SYS_METASLAB_IMPL_H */