]>
Commit | Line | Data |
---|---|---|
34dc7c2f BB |
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 | /* | |
572e2857 | 22 | * Copyright (c) 2005, 2010, Oracle and/or its affiliates. All rights reserved. |
34dc7c2f BB |
23 | */ |
24 | ||
25 | #ifndef _SYS_FS_ZFS_ZNODE_H | |
26 | #define _SYS_FS_ZFS_ZNODE_H | |
27 | ||
34dc7c2f BB |
28 | #ifdef _KERNEL |
29 | #include <sys/isa_defs.h> | |
30 | #include <sys/types32.h> | |
31 | #include <sys/attr.h> | |
32 | #include <sys/list.h> | |
33 | #include <sys/dmu.h> | |
428870ff | 34 | #include <sys/sa.h> |
34dc7c2f BB |
35 | #include <sys/zfs_vfsops.h> |
36 | #include <sys/rrwlock.h> | |
428870ff | 37 | #include <sys/zfs_sa.h> |
572e2857 | 38 | #include <sys/zfs_stat.h> |
34dc7c2f BB |
39 | #endif |
40 | #include <sys/zfs_acl.h> | |
41 | #include <sys/zil.h> | |
42 | ||
43 | #ifdef __cplusplus | |
44 | extern "C" { | |
45 | #endif | |
46 | ||
47 | /* | |
48 | * Additional file level attributes, that are stored | |
49 | * in the upper half of zp_flags | |
50 | */ | |
c5b3a7bb RC |
51 | #define ZFS_READONLY 0x0000000100000000ull |
52 | #define ZFS_HIDDEN 0x0000000200000000ull | |
53 | #define ZFS_SYSTEM 0x0000000400000000ull | |
54 | #define ZFS_ARCHIVE 0x0000000800000000ull | |
55 | #define ZFS_IMMUTABLE 0x0000001000000000ull | |
56 | #define ZFS_NOUNLINK 0x0000002000000000ull | |
57 | #define ZFS_APPENDONLY 0x0000004000000000ull | |
58 | #define ZFS_NODUMP 0x0000008000000000ull | |
59 | #define ZFS_OPAQUE 0x0000010000000000ull | |
60 | #define ZFS_AV_QUARANTINED 0x0000020000000000ull | |
61 | #define ZFS_AV_MODIFIED 0x0000040000000000ull | |
62 | #define ZFS_REPARSE 0x0000080000000000ull | |
63 | #define ZFS_OFFLINE 0x0000100000000000ull | |
64 | #define ZFS_SPARSE 0x0000200000000000ull | |
34dc7c2f | 65 | |
428870ff | 66 | #define ZFS_ATTR_SET(zp, attr, value, pflags, tx) \ |
34dc7c2f BB |
67 | { \ |
68 | if (value) \ | |
428870ff | 69 | pflags |= attr; \ |
34dc7c2f | 70 | else \ |
428870ff BB |
71 | pflags &= ~attr; \ |
72 | VERIFY(0 == sa_update(zp->z_sa_hdl, SA_ZPL_FLAGS(zp->z_zfsvfs), \ | |
73 | &pflags, sizeof (pflags), tx)); \ | |
34dc7c2f BB |
74 | } |
75 | ||
76 | /* | |
77 | * Define special zfs pflags | |
78 | */ | |
79 | #define ZFS_XATTR 0x1 /* is an extended attribute */ | |
80 | #define ZFS_INHERIT_ACE 0x2 /* ace has inheritable ACEs */ | |
81 | #define ZFS_ACL_TRIVIAL 0x4 /* files ACL is trivial */ | |
82 | #define ZFS_ACL_OBJ_ACE 0x8 /* ACL has CMPLX Object ACE */ | |
83 | #define ZFS_ACL_PROTECTED 0x10 /* ACL protected */ | |
84 | #define ZFS_ACL_DEFAULTED 0x20 /* ACL should be defaulted */ | |
85 | #define ZFS_ACL_AUTO_INHERIT 0x40 /* ACL should be inherited */ | |
86 | #define ZFS_BONUS_SCANSTAMP 0x80 /* Scanstamp in bonus area */ | |
45d1cae3 | 87 | #define ZFS_NO_EXECS_DENIED 0x100 /* exec was given to everyone */ |
34dc7c2f | 88 | |
428870ff BB |
89 | #define SA_ZPL_ATIME(z) z->z_attr_table[ZPL_ATIME] |
90 | #define SA_ZPL_MTIME(z) z->z_attr_table[ZPL_MTIME] | |
91 | #define SA_ZPL_CTIME(z) z->z_attr_table[ZPL_CTIME] | |
92 | #define SA_ZPL_CRTIME(z) z->z_attr_table[ZPL_CRTIME] | |
93 | #define SA_ZPL_GEN(z) z->z_attr_table[ZPL_GEN] | |
94 | #define SA_ZPL_DACL_ACES(z) z->z_attr_table[ZPL_DACL_ACES] | |
95 | #define SA_ZPL_XATTR(z) z->z_attr_table[ZPL_XATTR] | |
96 | #define SA_ZPL_SYMLINK(z) z->z_attr_table[ZPL_SYMLINK] | |
97 | #define SA_ZPL_RDEV(z) z->z_attr_table[ZPL_RDEV] | |
98 | #define SA_ZPL_SCANSTAMP(z) z->z_attr_table[ZPL_SCANSTAMP] | |
99 | #define SA_ZPL_UID(z) z->z_attr_table[ZPL_UID] | |
100 | #define SA_ZPL_GID(z) z->z_attr_table[ZPL_GID] | |
101 | #define SA_ZPL_PARENT(z) z->z_attr_table[ZPL_PARENT] | |
102 | #define SA_ZPL_LINKS(z) z->z_attr_table[ZPL_LINKS] | |
103 | #define SA_ZPL_MODE(z) z->z_attr_table[ZPL_MODE] | |
104 | #define SA_ZPL_DACL_COUNT(z) z->z_attr_table[ZPL_DACL_COUNT] | |
105 | #define SA_ZPL_FLAGS(z) z->z_attr_table[ZPL_FLAGS] | |
106 | #define SA_ZPL_SIZE(z) z->z_attr_table[ZPL_SIZE] | |
107 | #define SA_ZPL_ZNODE_ACL(z) z->z_attr_table[ZPL_ZNODE_ACL] | |
108 | #define SA_ZPL_PAD(z) z->z_attr_table[ZPL_PAD] | |
109 | ||
34dc7c2f BB |
110 | /* |
111 | * Is ID ephemeral? | |
112 | */ | |
113 | #define IS_EPHEMERAL(x) (x > MAXUID) | |
114 | ||
115 | /* | |
116 | * Should we use FUIDs? | |
117 | */ | |
428870ff | 118 | #define USE_FUIDS(version, os) (version >= ZPL_VERSION_FUID && \ |
34dc7c2f | 119 | spa_version(dmu_objset_spa(os)) >= SPA_VERSION_FUID) |
428870ff BB |
120 | #define USE_SA(version, os) (version >= ZPL_VERSION_SA && \ |
121 | spa_version(dmu_objset_spa(os)) >= SPA_VERSION_SA) | |
34dc7c2f BB |
122 | |
123 | #define MASTER_NODE_OBJ 1 | |
124 | ||
125 | /* | |
126 | * Special attributes for master node. | |
9babb374 BB |
127 | * "userquota@" and "groupquota@" are also valid (from |
128 | * zfs_userquota_prop_prefixes[]). | |
34dc7c2f BB |
129 | */ |
130 | #define ZFS_FSID "FSID" | |
131 | #define ZFS_UNLINKED_SET "DELETE_QUEUE" | |
132 | #define ZFS_ROOT_OBJ "ROOT" | |
133 | #define ZPL_VERSION_STR "VERSION" | |
134 | #define ZFS_FUID_TABLES "FUID" | |
9babb374 | 135 | #define ZFS_SHARES_DIR "SHARES" |
428870ff | 136 | #define ZFS_SA_ATTRS "SA_ATTRS" |
34dc7c2f BB |
137 | |
138 | #define ZFS_MAX_BLOCKSIZE (SPA_MAXBLOCKSIZE) | |
139 | ||
140 | /* Path component length */ | |
141 | /* | |
142 | * The generic fs code uses MAXNAMELEN to represent | |
143 | * what the largest component length is. Unfortunately, | |
144 | * this length includes the terminating NULL. ZFS needs | |
145 | * to tell the users via pathconf() and statvfs() what the | |
146 | * true maximum length of a component is, excluding the NULL. | |
147 | */ | |
148 | #define ZFS_MAXNAMELEN (MAXNAMELEN - 1) | |
149 | ||
150 | /* | |
151 | * Convert mode bits (zp_mode) to BSD-style DT_* values for storing in | |
152 | * the directory entries. | |
153 | */ | |
154 | #define IFTODT(mode) (((mode) & S_IFMT) >> 12) | |
155 | ||
156 | /* | |
157 | * The directory entry has the type (currently unused on Solaris) in the | |
158 | * top 4 bits, and the object number in the low 48 bits. The "middle" | |
159 | * 12 bits are unused. | |
160 | */ | |
161 | #define ZFS_DIRENT_TYPE(de) BF64_GET(de, 60, 4) | |
162 | #define ZFS_DIRENT_OBJ(de) BF64_GET(de, 0, 48) | |
163 | ||
34dc7c2f BB |
164 | /* |
165 | * Directory entry locks control access to directory entries. | |
166 | * They are used to protect creates, deletes, and renames. | |
167 | * Each directory znode has a mutex and a list of locked names. | |
168 | */ | |
169 | #ifdef _KERNEL | |
170 | typedef struct zfs_dirlock { | |
171 | char *dl_name; /* directory entry being locked */ | |
172 | uint32_t dl_sharecnt; /* 0 if exclusive, > 0 if shared */ | |
428870ff | 173 | uint8_t dl_namelock; /* 1 if z_name_lock is NOT held */ |
34dc7c2f BB |
174 | uint16_t dl_namesize; /* set if dl_name was allocated */ |
175 | kcondvar_t dl_cv; /* wait for entry to be unlocked */ | |
176 | struct znode *dl_dzp; /* directory znode */ | |
177 | struct zfs_dirlock *dl_next; /* next in z_dirlocks list */ | |
178 | } zfs_dirlock_t; | |
179 | ||
180 | typedef struct znode { | |
181 | struct zfsvfs *z_zfsvfs; | |
182 | vnode_t *z_vnode; | |
183 | uint64_t z_id; /* object ID for this znode */ | |
184 | kmutex_t z_lock; /* znode modification lock */ | |
34dc7c2f BB |
185 | krwlock_t z_parent_lock; /* parent lock for directories */ |
186 | krwlock_t z_name_lock; /* "master" lock for dirent locks */ | |
187 | zfs_dirlock_t *z_dirlocks; /* directory entry lock list */ | |
188 | kmutex_t z_range_lock; /* protects changes to z_range_avl */ | |
189 | avl_tree_t z_range_avl; /* avl tree of file range locks */ | |
190 | uint8_t z_unlinked; /* file has been unlinked */ | |
191 | uint8_t z_atime_dirty; /* atime needs to be synced */ | |
192 | uint8_t z_zn_prefetch; /* Prefetch znodes? */ | |
572e2857 | 193 | uint8_t z_moved; /* Has this znode been moved? */ |
34dc7c2f BB |
194 | uint_t z_blksz; /* block size in bytes */ |
195 | uint_t z_seq; /* modification sequence number */ | |
196 | uint64_t z_mapcnt; /* number of pages mapped to file */ | |
428870ff BB |
197 | uint64_t z_gen; /* generation (cached) */ |
198 | uint64_t z_size; /* file size (cached) */ | |
199 | uint64_t z_atime[2]; /* atime (cached) */ | |
200 | uint64_t z_links; /* file links (cached) */ | |
201 | uint64_t z_pflags; /* pflags (cached) */ | |
572e2857 BB |
202 | uint64_t z_uid; /* uid fuid (cached) */ |
203 | uint64_t z_gid; /* gid fuid (cached) */ | |
428870ff | 204 | mode_t z_mode; /* mode (cached) */ |
34dc7c2f BB |
205 | uint32_t z_sync_cnt; /* synchronous open count */ |
206 | kmutex_t z_acl_lock; /* acl data lock */ | |
45d1cae3 | 207 | zfs_acl_t *z_acl_cached; /* cached acl */ |
34dc7c2f | 208 | list_node_t z_link_node; /* all znodes in fs link */ |
428870ff BB |
209 | sa_handle_t *z_sa_hdl; /* handle to sa data */ |
210 | boolean_t z_is_sa; /* are we native sa? */ | |
34dc7c2f BB |
211 | } znode_t; |
212 | ||
213 | ||
214 | /* | |
215 | * Range locking rules | |
216 | * -------------------- | |
217 | * 1. When truncating a file (zfs_create, zfs_setattr, zfs_space) the whole | |
218 | * file range needs to be locked as RL_WRITER. Only then can the pages be | |
219 | * freed etc and zp_size reset. zp_size must be set within range lock. | |
220 | * 2. For writes and punching holes (zfs_write & zfs_space) just the range | |
221 | * being written or freed needs to be locked as RL_WRITER. | |
222 | * Multiple writes at the end of the file must coordinate zp_size updates | |
223 | * to ensure data isn't lost. A compare and swap loop is currently used | |
224 | * to ensure the file size is at least the offset last written. | |
225 | * 3. For reads (zfs_read, zfs_get_data & zfs_putapage) just the range being | |
226 | * read needs to be locked as RL_READER. A check against zp_size can then | |
227 | * be made for reading beyond end of file. | |
228 | */ | |
229 | ||
230 | /* | |
231 | * Convert between znode pointers and vnode pointers | |
232 | */ | |
233 | #define ZTOV(ZP) ((ZP)->z_vnode) | |
234 | #define VTOZ(VP) ((znode_t *)(VP)->v_data) | |
235 | ||
236 | /* | |
237 | * ZFS_ENTER() is called on entry to each ZFS vnode and vfs operation. | |
238 | * ZFS_EXIT() must be called before exitting the vop. | |
239 | * ZFS_VERIFY_ZP() verifies the znode is valid. | |
240 | */ | |
241 | #define ZFS_ENTER(zfsvfs) \ | |
242 | { \ | |
243 | rrw_enter(&(zfsvfs)->z_teardown_lock, RW_READER, FTAG); \ | |
244 | if ((zfsvfs)->z_unmounted) { \ | |
245 | ZFS_EXIT(zfsvfs); \ | |
246 | return (EIO); \ | |
247 | } \ | |
248 | } | |
249 | ||
250 | #define ZFS_EXIT(zfsvfs) rrw_exit(&(zfsvfs)->z_teardown_lock, FTAG) | |
251 | ||
252 | #define ZFS_VERIFY_ZP(zp) \ | |
428870ff | 253 | if ((zp)->z_sa_hdl == NULL) { \ |
34dc7c2f BB |
254 | ZFS_EXIT((zp)->z_zfsvfs); \ |
255 | return (EIO); \ | |
256 | } \ | |
257 | ||
258 | /* | |
259 | * Macros for dealing with dmu_buf_hold | |
260 | */ | |
261 | #define ZFS_OBJ_HASH(obj_num) ((obj_num) & (ZFS_OBJ_MTX_SZ - 1)) | |
b128c09f BB |
262 | #define ZFS_OBJ_MUTEX(zfsvfs, obj_num) \ |
263 | (&(zfsvfs)->z_hold_mtx[ZFS_OBJ_HASH(obj_num)]) | |
34dc7c2f | 264 | #define ZFS_OBJ_HOLD_ENTER(zfsvfs, obj_num) \ |
b128c09f BB |
265 | mutex_enter(ZFS_OBJ_MUTEX((zfsvfs), (obj_num))) |
266 | #define ZFS_OBJ_HOLD_TRYENTER(zfsvfs, obj_num) \ | |
267 | mutex_tryenter(ZFS_OBJ_MUTEX((zfsvfs), (obj_num))) | |
34dc7c2f | 268 | #define ZFS_OBJ_HOLD_EXIT(zfsvfs, obj_num) \ |
b128c09f | 269 | mutex_exit(ZFS_OBJ_MUTEX((zfsvfs), (obj_num))) |
34dc7c2f BB |
270 | |
271 | /* | |
272 | * Macros to encode/decode ZFS stored time values from/to struct timespec | |
273 | */ | |
274 | #define ZFS_TIME_ENCODE(tp, stmp) \ | |
275 | { \ | |
b128c09f | 276 | (stmp)[0] = (uint64_t)(tp)->tv_sec; \ |
34dc7c2f BB |
277 | (stmp)[1] = (uint64_t)(tp)->tv_nsec; \ |
278 | } | |
279 | ||
280 | #define ZFS_TIME_DECODE(tp, stmp) \ | |
281 | { \ | |
282 | (tp)->tv_sec = (time_t)(stmp)[0]; \ | |
283 | (tp)->tv_nsec = (long)(stmp)[1]; \ | |
284 | } | |
285 | ||
286 | /* | |
287 | * Timestamp defines | |
288 | */ | |
289 | #define ACCESSED (AT_ATIME) | |
290 | #define STATE_CHANGED (AT_CTIME) | |
291 | #define CONTENT_MODIFIED (AT_MTIME | AT_CTIME) | |
292 | ||
293 | #define ZFS_ACCESSTIME_STAMP(zfsvfs, zp) \ | |
294 | if ((zfsvfs)->z_atime && !((zfsvfs)->z_vfs->vfs_flag & VFS_RDONLY)) \ | |
428870ff | 295 | zfs_tstamp_update_setup(zp, ACCESSED, NULL, NULL, B_FALSE); |
34dc7c2f | 296 | |
b128c09f | 297 | extern int zfs_init_fs(zfsvfs_t *, znode_t **); |
34dc7c2f BB |
298 | extern void zfs_set_dataprop(objset_t *); |
299 | extern void zfs_create_fs(objset_t *os, cred_t *cr, nvlist_t *, | |
300 | dmu_tx_t *tx); | |
428870ff BB |
301 | extern void zfs_tstamp_update_setup(znode_t *, uint_t, uint64_t [2], |
302 | uint64_t [2], boolean_t); | |
34dc7c2f BB |
303 | extern void zfs_grow_blocksize(znode_t *, uint64_t, dmu_tx_t *); |
304 | extern int zfs_freesp(znode_t *, uint64_t, uint64_t, int, boolean_t); | |
305 | extern void zfs_znode_init(void); | |
306 | extern void zfs_znode_fini(void); | |
307 | extern int zfs_zget(zfsvfs_t *, uint64_t, znode_t **); | |
308 | extern int zfs_rezget(znode_t *); | |
309 | extern void zfs_zinactive(znode_t *); | |
310 | extern void zfs_znode_delete(znode_t *, dmu_tx_t *); | |
311 | extern void zfs_znode_free(znode_t *); | |
0bc8fd78 BB |
312 | extern void zfs_remove_op_tables(void); |
313 | extern int zfs_create_op_tables(void); | |
34dc7c2f BB |
314 | extern int zfs_sync(vfs_t *vfsp, short flag, cred_t *cr); |
315 | extern dev_t zfs_cmpldev(uint64_t); | |
316 | extern int zfs_get_zplprop(objset_t *os, zfs_prop_t prop, uint64_t *value); | |
34dc7c2f BB |
317 | extern int zfs_get_stats(objset_t *os, nvlist_t *nv); |
318 | extern void zfs_znode_dmu_fini(znode_t *); | |
319 | ||
320 | extern void zfs_log_create(zilog_t *zilog, dmu_tx_t *tx, uint64_t txtype, | |
321 | znode_t *dzp, znode_t *zp, char *name, vsecattr_t *, zfs_fuid_info_t *, | |
322 | vattr_t *vap); | |
323 | extern int zfs_log_create_txtype(zil_create_t, vsecattr_t *vsecp, | |
324 | vattr_t *vap); | |
325 | extern void zfs_log_remove(zilog_t *zilog, dmu_tx_t *tx, uint64_t txtype, | |
572e2857 BB |
326 | znode_t *dzp, char *name, uint64_t foid); |
327 | #define ZFS_NO_OBJECT 0 /* no object id */ | |
34dc7c2f BB |
328 | extern void zfs_log_link(zilog_t *zilog, dmu_tx_t *tx, uint64_t txtype, |
329 | znode_t *dzp, znode_t *zp, char *name); | |
330 | extern void zfs_log_symlink(zilog_t *zilog, dmu_tx_t *tx, uint64_t txtype, | |
331 | znode_t *dzp, znode_t *zp, char *name, char *link); | |
332 | extern void zfs_log_rename(zilog_t *zilog, dmu_tx_t *tx, uint64_t txtype, | |
333 | znode_t *sdzp, char *sname, znode_t *tdzp, char *dname, znode_t *szp); | |
334 | extern void zfs_log_write(zilog_t *zilog, dmu_tx_t *tx, int txtype, | |
335 | znode_t *zp, offset_t off, ssize_t len, int ioflag); | |
336 | extern void zfs_log_truncate(zilog_t *zilog, dmu_tx_t *tx, int txtype, | |
337 | znode_t *zp, uint64_t off, uint64_t len); | |
338 | extern void zfs_log_setattr(zilog_t *zilog, dmu_tx_t *tx, int txtype, | |
339 | znode_t *zp, vattr_t *vap, uint_t mask_applied, zfs_fuid_info_t *fuidp); | |
340 | extern void zfs_log_acl(zilog_t *zilog, dmu_tx_t *tx, znode_t *zp, | |
341 | vsecattr_t *vsecp, zfs_fuid_info_t *fuidp); | |
428870ff | 342 | extern void zfs_xvattr_set(znode_t *zp, xvattr_t *xvap, dmu_tx_t *tx); |
34dc7c2f | 343 | extern void zfs_upgrade(zfsvfs_t *zfsvfs, dmu_tx_t *tx); |
9babb374 | 344 | extern int zfs_create_share_dir(zfsvfs_t *zfsvfs, dmu_tx_t *tx); |
34dc7c2f | 345 | |
60101509 | 346 | #if defined(HAVE_UIO_RW) |
b128c09f BB |
347 | extern caddr_t zfs_map_page(page_t *, enum seg_rw); |
348 | extern void zfs_unmap_page(page_t *, caddr_t); | |
60101509 | 349 | #endif /* HAVE_UIO_RW */ |
b128c09f | 350 | |
34dc7c2f BB |
351 | extern zil_get_data_t zfs_get_data; |
352 | extern zil_replay_func_t *zfs_replay_vector[TX_MAX_TYPE]; | |
353 | extern int zfsfstype; | |
354 | ||
355 | #endif /* _KERNEL */ | |
356 | ||
357 | extern int zfs_obj_to_path(objset_t *osp, uint64_t obj, char *buf, int len); | |
358 | ||
359 | #ifdef __cplusplus | |
360 | } | |
361 | #endif | |
362 | ||
363 | #endif /* _SYS_FS_ZFS_ZNODE_H */ |