Print this page
NEX-9200 Improve the scalability of attribute locking in zfs_zget
Reviewed by: Joyce McIntosh <joyce.mcintosh@nexenta.com>
Reviewed by: Sanjay Nadkarni <sanjay.nadkarni@nexenta.com>
NEX-9436 Rate limiting controls (was QoS) per ZFS dataset, updates from demo
Reviewed by: Gordon Ross <gordon.ross@nexenta.com>
Reviewed by: Rob Gittins <rob.gittins@nexenta.com>
NEX-8972 Async-delete side-effect that may cause unmount EBUSY
Reviewed by: Alek Pinchuk <alek@nexenta.com>
Reviewed by: Sanjay Nadkarni <sanjay.nadkarni@nexenta.com>
NEX-8852 Quality-of-Service (QoS) controls per NFS share
Reviewed by: Rob Gittins <rob.gittins@nexenta.com>
Reviewed by: Evan Layton <evan.layton@nexenta.com>
Reviewed by: Sanjay Nadkarni <sanjay.nadkarni@nexenta.com>
NEX-3762 Appliance crashes with a NULL pointer dereference during a zpool export when a zfs_vn_rele_taskq thread attempts to check a bogus rwlock from rw_write_held
Reviewed by: Josef 'Jeff' Sipek <josef.sipek@nexenta.com>
OS-80 support for vdev and CoS properties for the new I/O scheduler
OS-95 lint warning introduced by OS-61
Fixup merge results
re #13204 rb4280 zfs receive/rollback deadlock
re #6815 rb1758 need WORM in nza-kernel (4.0)


   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 #ifndef _SYS_FS_ZFS_VFSOPS_H
  26 #define _SYS_FS_ZFS_VFSOPS_H
  27 
  28 #include <sys/isa_defs.h>
  29 #include <sys/types32.h>
  30 #include <sys/list.h>
  31 #include <sys/vfs.h>
  32 #include <sys/zil.h>
  33 #include <sys/sa.h>
  34 #include <sys/rrwlock.h>
  35 #include <sys/zfs_ioctl.h>
  36 
  37 #ifdef  __cplusplus
  38 extern "C" {
  39 #endif
  40 
  41 typedef struct zfsvfs zfsvfs_t;
  42 struct znode;
  43 























  44 struct zfsvfs {
  45         vfs_t           *z_vfs;         /* generic fs struct */
  46         zfsvfs_t        *z_parent;      /* parent fs */
  47         objset_t        *z_os;          /* objset reference */
  48         uint64_t        z_root;         /* id of root znode */
  49         uint64_t        z_unlinkedobj;  /* id of unlinked zapobj */
  50         uint64_t        z_max_blksz;    /* maximum block size for files */
  51         uint64_t        z_fuid_obj;     /* fuid table object number */
  52         uint64_t        z_fuid_size;    /* fuid table size */
  53         avl_tree_t      z_fuid_idx;     /* fuid tree keyed by index */
  54         avl_tree_t      z_fuid_domain;  /* fuid tree keyed by domain */
  55         krwlock_t       z_fuid_lock;    /* fuid lock */
  56         boolean_t       z_fuid_loaded;  /* fuid tables are loaded */
  57         boolean_t       z_fuid_dirty;   /* need to sync fuid table ? */
  58         struct zfs_fuid_info    *z_fuid_replay; /* fuid info for replay */
  59         zilog_t         *z_log;         /* intent log pointer */
  60         uint_t          z_acl_mode;     /* acl chmod/mode behavior */
  61         uint_t          z_acl_inherit;  /* acl inheritance behavior */
  62         zfs_case_t      z_case;         /* case-sense */
  63         boolean_t       z_utf8;         /* utf8-only */
  64         int             z_norm;         /* normalization flags */
  65         boolean_t       z_atime;        /* enable atimes mount option */
  66         boolean_t       z_unmounted;    /* unmounted */
  67         rrmlock_t       z_teardown_lock;
  68         krwlock_t       z_teardown_inactive_lock;
  69         list_t          z_all_znodes;   /* all vnodes in the fs */
  70         kmutex_t        z_znodes_lock;  /* lock for z_all_znodes */

  71         vnode_t         *z_ctldir;      /* .zfs directory pointer */
  72         boolean_t       z_show_ctldir;  /* expose .zfs in the root dir */
  73         boolean_t       z_issnap;       /* true if this is a snapshot */
  74         boolean_t       z_vscan;        /* virus scan on/off */
  75         boolean_t       z_use_fuids;    /* version allows fuids */
  76         boolean_t       z_replay;       /* set during ZIL replay */
  77         boolean_t       z_use_sa;       /* version allow system attributes */
  78         uint64_t        z_version;      /* ZPL version */
  79         uint64_t        z_shares_dir;   /* hidden shares dir */
  80         kmutex_t        z_lock;
  81         uint64_t        z_userquota_obj;
  82         uint64_t        z_groupquota_obj;
  83         uint64_t        z_replay_eof;   /* New end of file - replay only */
  84         sa_attr_type_t  *z_attr_table;  /* SA attr mapping->id */
  85 #define ZFS_OBJ_MTX_SZ  64
  86         kmutex_t        z_hold_mtx[ZFS_OBJ_MTX_SZ];     /* znode hold locks */








  87 };
  88 
  89 /*
  90  * Normal filesystems (those not under .zfs/snapshot) have a total
  91  * file ID size limited to 12 bytes (including the length field) due to
  92  * NFSv2 protocol's limitation of 32 bytes for a filehandle.  For historical
  93  * reasons, this same limit is being imposed by the Solaris NFSv3 implementation
  94  * (although the NFSv3 protocol actually permits a maximum of 64 bytes).  It
  95  * is not possible to expand beyond 12 bytes without abandoning support
  96  * of NFSv2.
  97  *
  98  * For normal filesystems, we partition up the available space as follows:
  99  *      2 bytes         fid length (required)
 100  *      6 bytes         object number (48 bits)
 101  *      4 bytes         generation number (32 bits)
 102  *
 103  * We reserve only 48 bits for the object number, as this is the limit
 104  * currently defined and imposed by the DMU.
 105  */
 106 typedef struct zfid_short {




   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  * Copyright 2016 Nexenta Systems, Inc.  All rights reserved.
  24  */
  25 
  26 #ifndef _SYS_FS_ZFS_VFSOPS_H
  27 #define _SYS_FS_ZFS_VFSOPS_H
  28 
  29 #include <sys/isa_defs.h>
  30 #include <sys/types32.h>
  31 #include <sys/list.h>
  32 #include <sys/vfs.h>
  33 #include <sys/zil.h>
  34 #include <sys/sa.h>
  35 #include <sys/rrwlock.h>
  36 #include <sys/zfs_ioctl.h>
  37 
  38 #ifdef  __cplusplus
  39 extern "C" {
  40 #endif
  41 
  42 typedef struct zfsvfs zfsvfs_t;
  43 struct znode;
  44 
  45 /*
  46  * ZFS Quality of Service (QoS) I/O throttling state,
  47  * per file system.  Limits the I/O rate in this FS.
  48  * See "Token Bucket" on Wikipedia
  49  */
  50 typedef struct zfs_rate_state {
  51         uint64_t rate_cap;              /* zero means no cap */
  52         int64_t rate_token_bucket;      /* bytes I/O allowed without waiting */
  53         hrtime_t rate_last_update;
  54         kmutex_t rate_lock;
  55         kcondvar_t rate_wait_cv;
  56         int rate_waiters;
  57 } zfs_rate_state_t;
  58 
  59 /*
  60  * Status of the zfs_unlinked_drain thread.
  61  */
  62 typedef enum drain_state {
  63         ZFS_DRAIN_SHUTDOWN = 0,
  64         ZFS_DRAIN_RUNNING,
  65         ZFS_DRAIN_SHUTDOWN_REQ
  66 } drain_state_t;
  67 
  68 struct zfsvfs {
  69         vfs_t           *z_vfs;         /* generic fs struct */
  70         zfsvfs_t        *z_parent;      /* parent fs */
  71         objset_t        *z_os;          /* objset reference */
  72         uint64_t        z_root;         /* id of root znode */
  73         uint64_t        z_unlinkedobj;  /* id of unlinked zapobj */
  74         uint64_t        z_max_blksz;    /* maximum block size for files */
  75         uint64_t        z_fuid_obj;     /* fuid table object number */
  76         uint64_t        z_fuid_size;    /* fuid table size */
  77         avl_tree_t      z_fuid_idx;     /* fuid tree keyed by index */
  78         avl_tree_t      z_fuid_domain;  /* fuid tree keyed by domain */
  79         krwlock_t       z_fuid_lock;    /* fuid lock */
  80         boolean_t       z_fuid_loaded;  /* fuid tables are loaded */
  81         boolean_t       z_fuid_dirty;   /* need to sync fuid table ? */
  82         struct zfs_fuid_info    *z_fuid_replay; /* fuid info for replay */
  83         zilog_t         *z_log;         /* intent log pointer */
  84         uint_t          z_acl_mode;     /* acl chmod/mode behavior */
  85         uint_t          z_acl_inherit;  /* acl inheritance behavior */
  86         zfs_case_t      z_case;         /* case-sense */
  87         boolean_t       z_utf8;         /* utf8-only */
  88         int             z_norm;         /* normalization flags */
  89         boolean_t       z_atime;        /* enable atimes mount option */
  90         boolean_t       z_unmounted;    /* unmounted */
  91         rrmlock_t       z_teardown_lock;
  92         krwlock_t       z_teardown_inactive_lock;
  93         list_t          z_all_znodes;   /* all vnodes in the fs */
  94         kmutex_t        z_znodes_lock;  /* lock for z_all_znodes */
  95         uint_t          z_znodes_freeing_cnt; /* number of znodes to be freed */
  96         vnode_t         *z_ctldir;      /* .zfs directory pointer */
  97         boolean_t       z_show_ctldir;  /* expose .zfs in the root dir */
  98         boolean_t       z_issnap;       /* true if this is a snapshot */
  99         boolean_t       z_vscan;        /* virus scan on/off */
 100         boolean_t       z_use_fuids;    /* version allows fuids */
 101         boolean_t       z_replay;       /* set during ZIL replay */
 102         boolean_t       z_use_sa;       /* version allow system attributes */
 103         uint64_t        z_version;      /* ZPL version */
 104         uint64_t        z_shares_dir;   /* hidden shares dir */
 105         kmutex_t        z_lock;
 106         uint64_t        z_userquota_obj;
 107         uint64_t        z_groupquota_obj;
 108         uint64_t        z_replay_eof;   /* New end of file - replay only */
 109         sa_attr_type_t  *z_attr_table;  /* SA attr mapping->id */
 110         boolean_t       z_isworm;       /* true if this is a WORM FS */
 111         /* true if suspend-resume cycle is in progress */
 112         boolean_t       z_busy;
 113         int     z_hold_mtx_sz; /* the size of z_hold_mtx array */
 114         kmutex_t        *z_hold_mtx;    /* znode hold locks */
 115         /* for controlling async zfs_unlinked_drain */
 116         kmutex_t        z_drain_lock;
 117         kcondvar_t      z_drain_cv;
 118         drain_state_t   z_drain_state;
 119         zfs_rate_state_t z_rate;
 120 };
 121 
 122 /*
 123  * Normal filesystems (those not under .zfs/snapshot) have a total
 124  * file ID size limited to 12 bytes (including the length field) due to
 125  * NFSv2 protocol's limitation of 32 bytes for a filehandle.  For historical
 126  * reasons, this same limit is being imposed by the Solaris NFSv3 implementation
 127  * (although the NFSv3 protocol actually permits a maximum of 64 bytes).  It
 128  * is not possible to expand beyond 12 bytes without abandoning support
 129  * of NFSv2.
 130  *
 131  * For normal filesystems, we partition up the available space as follows:
 132  *      2 bytes         fid length (required)
 133  *      6 bytes         object number (48 bits)
 134  *      4 bytes         generation number (32 bits)
 135  *
 136  * We reserve only 48 bits for the object number, as this is the limit
 137  * currently defined and imposed by the DMU.
 138  */
 139 typedef struct zfid_short {