Blame view
fs/xfs/xfs_inode.c
104 KB
0b61f8a40 xfs: convert to S... |
1 |
// SPDX-License-Identifier: GPL-2.0 |
1da177e4c Linux-2.6.12-rc2 |
2 |
/* |
3e57ecf64 [XFS] Add paramet... |
3 |
* Copyright (c) 2000-2006 Silicon Graphics, Inc. |
7b7187698 [XFS] Update lice... |
4 |
* All Rights Reserved. |
1da177e4c Linux-2.6.12-rc2 |
5 |
*/ |
f0e282806 xfs: convert to n... |
6 |
#include <linux/iversion.h> |
40ebd81d1 [XFS] Use kernel-... |
7 |
|
1da177e4c Linux-2.6.12-rc2 |
8 |
#include "xfs.h" |
a844f4510 [XFS] Remove xfs_... |
9 |
#include "xfs_fs.h" |
70a9883c5 xfs: create a sha... |
10 |
#include "xfs_shared.h" |
239880ef6 xfs: decouple log... |
11 12 13 |
#include "xfs_format.h" #include "xfs_log_format.h" #include "xfs_trans_resv.h" |
1da177e4c Linux-2.6.12-rc2 |
14 |
#include "xfs_sb.h" |
1da177e4c Linux-2.6.12-rc2 |
15 |
#include "xfs_mount.h" |
3ab78df2a xfs: rework xfs_b... |
16 |
#include "xfs_defer.h" |
a4fbe6ab1 xfs: decouple ino... |
17 |
#include "xfs_inode.h" |
c24b5dfad xfs: kill xfs_vno... |
18 |
#include "xfs_dir2.h" |
c24b5dfad xfs: kill xfs_vno... |
19 |
#include "xfs_attr.h" |
239880ef6 xfs: decouple log... |
20 21 |
#include "xfs_trans_space.h" #include "xfs_trans.h" |
1da177e4c Linux-2.6.12-rc2 |
22 |
#include "xfs_buf_item.h" |
a844f4510 [XFS] Remove xfs_... |
23 |
#include "xfs_inode_item.h" |
a844f4510 [XFS] Remove xfs_... |
24 25 |
#include "xfs_ialloc.h" #include "xfs_bmap.h" |
689881145 xfs: create xfs_b... |
26 |
#include "xfs_bmap_util.h" |
e9e899a2a xfs: move error i... |
27 |
#include "xfs_errortag.h" |
1da177e4c Linux-2.6.12-rc2 |
28 |
#include "xfs_error.h" |
1da177e4c Linux-2.6.12-rc2 |
29 |
#include "xfs_quota.h" |
2a82b8be8 [XFS] Concurrent ... |
30 |
#include "xfs_filestream.h" |
0b1b213fc xfs: event tracin... |
31 |
#include "xfs_trace.h" |
33479e054 xfs: remove xfs_i... |
32 |
#include "xfs_icache.h" |
c24b5dfad xfs: kill xfs_vno... |
33 |
#include "xfs_symlink.h" |
239880ef6 xfs: decouple log... |
34 35 |
#include "xfs_trans_priv.h" #include "xfs_log.h" |
a4fbe6ab1 xfs: decouple ino... |
36 |
#include "xfs_bmap_btree.h" |
aa8968f22 xfs: cancel CoW r... |
37 |
#include "xfs_reflink.h" |
1da177e4c Linux-2.6.12-rc2 |
38 |
|
1da177e4c Linux-2.6.12-rc2 |
39 |
kmem_zone_t *xfs_inode_zone; |
1da177e4c Linux-2.6.12-rc2 |
40 41 |
/* |
8f04c47aa xfs: split xfs_it... |
42 |
* Used in xfs_itruncate_extents(). This is the maximum number of extents |
1da177e4c Linux-2.6.12-rc2 |
43 44 45 |
* freed from a file in a single transaction. */ #define XFS_ITRUNC_MAX_EXTENTS 2 |
54d7b5c1d xfs: use vfs inod... |
46 47 |
STATIC int xfs_iunlink(struct xfs_trans *, struct xfs_inode *); STATIC int xfs_iunlink_remove(struct xfs_trans *, struct xfs_inode *); |
ab2974311 xfs: allow linkat... |
48 |
|
2a0ec1d9e xfs: move xfs_get... |
49 50 51 52 53 54 55 |
/* * helper function to extract extent size hint from inode */ xfs_extlen_t xfs_get_extsz_hint( struct xfs_inode *ip) { |
bdb2ed2db xfs: ignore exten... |
56 57 58 59 60 61 |
/* * No point in aligning allocations if we need to COW to actually * write to them. */ if (xfs_is_always_cow_inode(ip)) return 0; |
2a0ec1d9e xfs: move xfs_get... |
62 63 64 65 66 67 |
if ((ip->i_d.di_flags & XFS_DIFLAG_EXTSIZE) && ip->i_d.di_extsize) return ip->i_d.di_extsize; if (XFS_IS_REALTIME_INODE(ip)) return ip->i_mount->m_sb.sb_rextsize; return 0; } |
fa96acadf xfs: move inode l... |
68 |
/* |
f7ca35227 xfs: create a sep... |
69 70 |
* Helper function to extract CoW extent size hint from inode. * Between the extent size hint and the CoW extent size hint, we |
e153aa799 xfs: set a defaul... |
71 72 |
* return the greater of the two. If the value is zero (automatic), * use the default size. |
f7ca35227 xfs: create a sep... |
73 74 75 76 77 78 79 80 81 82 83 |
*/ xfs_extlen_t xfs_get_cowextsz_hint( struct xfs_inode *ip) { xfs_extlen_t a, b; a = 0; if (ip->i_d.di_flags2 & XFS_DIFLAG2_COWEXTSIZE) a = ip->i_d.di_cowextsize; b = xfs_get_extsz_hint(ip); |
e153aa799 xfs: set a defaul... |
84 85 86 87 |
a = max(a, b); if (a == 0) return XFS_DEFAULT_COWEXTSZ_HINT; return a; |
f7ca35227 xfs: create a sep... |
88 89 90 |
} /* |
efa70be16 xfs: add xfs_iloc... |
91 92 93 94 95 96 97 98 99 100 |
* These two are wrapper routines around the xfs_ilock() routine used to * centralize some grungy code. They are used in places that wish to lock the * inode solely for reading the extents. The reason these places can't just * call xfs_ilock(ip, XFS_ILOCK_SHARED) is that the inode lock also guards to * bringing in of the extents from disk for a file in b-tree format. If the * inode is in b-tree format, then we need to lock the inode exclusively until * the extents are read in. Locking it exclusively all the time would limit * our parallelism unnecessarily, though. What we do instead is check to see * if the extents have been read in yet, and only lock the inode exclusively * if they have not. |
fa96acadf xfs: move inode l... |
101 |
* |
efa70be16 xfs: add xfs_iloc... |
102 |
* The functions return a value which should be given to the corresponding |
01f4f3277 xfs: remove xfs_i... |
103 |
* xfs_iunlock() call. |
fa96acadf xfs: move inode l... |
104 105 |
*/ uint |
309ecac8e xfs: rename xfs_i... |
106 107 |
xfs_ilock_data_map_shared( struct xfs_inode *ip) |
fa96acadf xfs: move inode l... |
108 |
{ |
309ecac8e xfs: rename xfs_i... |
109 |
uint lock_mode = XFS_ILOCK_SHARED; |
fa96acadf xfs: move inode l... |
110 |
|
f7e67b20e xfs: move the for... |
111 |
if (ip->i_df.if_format == XFS_DINODE_FMT_BTREE && |
309ecac8e xfs: rename xfs_i... |
112 |
(ip->i_df.if_flags & XFS_IFEXTENTS) == 0) |
fa96acadf xfs: move inode l... |
113 |
lock_mode = XFS_ILOCK_EXCL; |
fa96acadf xfs: move inode l... |
114 |
xfs_ilock(ip, lock_mode); |
fa96acadf xfs: move inode l... |
115 116 |
return lock_mode; } |
efa70be16 xfs: add xfs_iloc... |
117 118 119 |
uint xfs_ilock_attr_map_shared( struct xfs_inode *ip) |
fa96acadf xfs: move inode l... |
120 |
{ |
efa70be16 xfs: add xfs_iloc... |
121 |
uint lock_mode = XFS_ILOCK_SHARED; |
f7e67b20e xfs: move the for... |
122 123 |
if (ip->i_afp && ip->i_afp->if_format == XFS_DINODE_FMT_BTREE && |
efa70be16 xfs: add xfs_iloc... |
124 125 126 127 |
(ip->i_afp->if_flags & XFS_IFEXTENTS) == 0) lock_mode = XFS_ILOCK_EXCL; xfs_ilock(ip, lock_mode); return lock_mode; |
fa96acadf xfs: move inode l... |
128 129 130 |
} /* |
655232183 xfs: remove i_iol... |
131 132 133 |
* In addition to i_rwsem in the VFS inode, the xfs inode contains 2 * multi-reader locks: i_mmap_lock and the i_lock. This routine allows * various combinations of the locks to be obtained. |
fa96acadf xfs: move inode l... |
134 |
* |
653c60b63 xfs: introduce mm... |
135 136 |
* The 3 locks should always be ordered so that the IO lock is obtained first, * the mmap lock second and the ilock last in order to prevent deadlock. |
fa96acadf xfs: move inode l... |
137 |
* |
653c60b63 xfs: introduce mm... |
138 139 |
* Basic locking order: * |
655232183 xfs: remove i_iol... |
140 |
* i_rwsem -> i_mmap_lock -> page_lock -> i_ilock |
653c60b63 xfs: introduce mm... |
141 |
* |
c1e8d7c6a mmap locking API:... |
142 |
* mmap_lock locking order: |
653c60b63 xfs: introduce mm... |
143 |
* |
c1e8d7c6a mmap locking API:... |
144 145 |
* i_rwsem -> page lock -> mmap_lock * mmap_lock -> i_mmap_lock -> page_lock |
653c60b63 xfs: introduce mm... |
146 |
* |
c1e8d7c6a mmap locking API:... |
147 |
* The difference in mmap_lock locking order mean that we cannot hold the |
653c60b63 xfs: introduce mm... |
148 |
* i_mmap_lock over syscall based read(2)/write(2) based IO. These IO paths can |
c1e8d7c6a mmap locking API:... |
149 |
* fault in pages during copy in/out (for buffered IO) or require the mmap_lock |
653c60b63 xfs: introduce mm... |
150 |
* in get_user_pages() to map the user pages into the kernel address space for |
655232183 xfs: remove i_iol... |
151 |
* direct IO. Similarly the i_rwsem cannot be taken inside a page fault because |
c1e8d7c6a mmap locking API:... |
152 |
* page faults already hold the mmap_lock. |
653c60b63 xfs: introduce mm... |
153 154 |
* * Hence to serialise fully against both syscall and mmap based IO, we need to |
655232183 xfs: remove i_iol... |
155 |
* take both the i_rwsem and the i_mmap_lock. These locks should *only* be both |
653c60b63 xfs: introduce mm... |
156 157 158 |
* taken in places where we need to invalidate the page cache in a race * free manner (e.g. truncate, hole punch and other extent manipulation * functions). |
fa96acadf xfs: move inode l... |
159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 |
*/ void xfs_ilock( xfs_inode_t *ip, uint lock_flags) { trace_xfs_ilock(ip, lock_flags, _RET_IP_); /* * You can't set both SHARED and EXCL for the same lock, * and only XFS_IOLOCK_SHARED, XFS_IOLOCK_EXCL, XFS_ILOCK_SHARED, * and XFS_ILOCK_EXCL are valid values to set in lock_flags. */ ASSERT((lock_flags & (XFS_IOLOCK_SHARED | XFS_IOLOCK_EXCL)) != (XFS_IOLOCK_SHARED | XFS_IOLOCK_EXCL)); |
653c60b63 xfs: introduce mm... |
174 175 |
ASSERT((lock_flags & (XFS_MMAPLOCK_SHARED | XFS_MMAPLOCK_EXCL)) != (XFS_MMAPLOCK_SHARED | XFS_MMAPLOCK_EXCL)); |
fa96acadf xfs: move inode l... |
176 177 |
ASSERT((lock_flags & (XFS_ILOCK_SHARED | XFS_ILOCK_EXCL)) != (XFS_ILOCK_SHARED | XFS_ILOCK_EXCL)); |
0952c8183 xfs: clean up ino... |
178 |
ASSERT((lock_flags & ~(XFS_LOCK_MASK | XFS_LOCK_SUBCLASS_MASK)) == 0); |
fa96acadf xfs: move inode l... |
179 |
|
655232183 xfs: remove i_iol... |
180 181 182 183 184 185 186 |
if (lock_flags & XFS_IOLOCK_EXCL) { down_write_nested(&VFS_I(ip)->i_rwsem, XFS_IOLOCK_DEP(lock_flags)); } else if (lock_flags & XFS_IOLOCK_SHARED) { down_read_nested(&VFS_I(ip)->i_rwsem, XFS_IOLOCK_DEP(lock_flags)); } |
fa96acadf xfs: move inode l... |
187 |
|
653c60b63 xfs: introduce mm... |
188 189 190 191 |
if (lock_flags & XFS_MMAPLOCK_EXCL) mrupdate_nested(&ip->i_mmaplock, XFS_MMAPLOCK_DEP(lock_flags)); else if (lock_flags & XFS_MMAPLOCK_SHARED) mraccess_nested(&ip->i_mmaplock, XFS_MMAPLOCK_DEP(lock_flags)); |
fa96acadf xfs: move inode l... |
192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 |
if (lock_flags & XFS_ILOCK_EXCL) mrupdate_nested(&ip->i_lock, XFS_ILOCK_DEP(lock_flags)); else if (lock_flags & XFS_ILOCK_SHARED) mraccess_nested(&ip->i_lock, XFS_ILOCK_DEP(lock_flags)); } /* * This is just like xfs_ilock(), except that the caller * is guaranteed not to sleep. It returns 1 if it gets * the requested locks and 0 otherwise. If the IO lock is * obtained but the inode lock cannot be, then the IO lock * is dropped before returning. * * ip -- the inode being locked * lock_flags -- this parameter indicates the inode's locks to be * to be locked. See the comment for xfs_ilock() for a list * of valid values. */ int xfs_ilock_nowait( xfs_inode_t *ip, uint lock_flags) { trace_xfs_ilock_nowait(ip, lock_flags, _RET_IP_); /* * You can't set both SHARED and EXCL for the same lock, * and only XFS_IOLOCK_SHARED, XFS_IOLOCK_EXCL, XFS_ILOCK_SHARED, * and XFS_ILOCK_EXCL are valid values to set in lock_flags. */ ASSERT((lock_flags & (XFS_IOLOCK_SHARED | XFS_IOLOCK_EXCL)) != (XFS_IOLOCK_SHARED | XFS_IOLOCK_EXCL)); |
653c60b63 xfs: introduce mm... |
224 225 |
ASSERT((lock_flags & (XFS_MMAPLOCK_SHARED | XFS_MMAPLOCK_EXCL)) != (XFS_MMAPLOCK_SHARED | XFS_MMAPLOCK_EXCL)); |
fa96acadf xfs: move inode l... |
226 227 |
ASSERT((lock_flags & (XFS_ILOCK_SHARED | XFS_ILOCK_EXCL)) != (XFS_ILOCK_SHARED | XFS_ILOCK_EXCL)); |
0952c8183 xfs: clean up ino... |
228 |
ASSERT((lock_flags & ~(XFS_LOCK_MASK | XFS_LOCK_SUBCLASS_MASK)) == 0); |
fa96acadf xfs: move inode l... |
229 230 |
if (lock_flags & XFS_IOLOCK_EXCL) { |
655232183 xfs: remove i_iol... |
231 |
if (!down_write_trylock(&VFS_I(ip)->i_rwsem)) |
fa96acadf xfs: move inode l... |
232 233 |
goto out; } else if (lock_flags & XFS_IOLOCK_SHARED) { |
655232183 xfs: remove i_iol... |
234 |
if (!down_read_trylock(&VFS_I(ip)->i_rwsem)) |
fa96acadf xfs: move inode l... |
235 236 |
goto out; } |
653c60b63 xfs: introduce mm... |
237 238 239 240 241 242 243 244 |
if (lock_flags & XFS_MMAPLOCK_EXCL) { if (!mrtryupdate(&ip->i_mmaplock)) goto out_undo_iolock; } else if (lock_flags & XFS_MMAPLOCK_SHARED) { if (!mrtryaccess(&ip->i_mmaplock)) goto out_undo_iolock; } |
fa96acadf xfs: move inode l... |
245 246 |
if (lock_flags & XFS_ILOCK_EXCL) { if (!mrtryupdate(&ip->i_lock)) |
653c60b63 xfs: introduce mm... |
247 |
goto out_undo_mmaplock; |
fa96acadf xfs: move inode l... |
248 249 |
} else if (lock_flags & XFS_ILOCK_SHARED) { if (!mrtryaccess(&ip->i_lock)) |
653c60b63 xfs: introduce mm... |
250 |
goto out_undo_mmaplock; |
fa96acadf xfs: move inode l... |
251 252 |
} return 1; |
653c60b63 xfs: introduce mm... |
253 254 255 256 257 258 |
out_undo_mmaplock: if (lock_flags & XFS_MMAPLOCK_EXCL) mrunlock_excl(&ip->i_mmaplock); else if (lock_flags & XFS_MMAPLOCK_SHARED) mrunlock_shared(&ip->i_mmaplock); out_undo_iolock: |
fa96acadf xfs: move inode l... |
259 |
if (lock_flags & XFS_IOLOCK_EXCL) |
655232183 xfs: remove i_iol... |
260 |
up_write(&VFS_I(ip)->i_rwsem); |
fa96acadf xfs: move inode l... |
261 |
else if (lock_flags & XFS_IOLOCK_SHARED) |
655232183 xfs: remove i_iol... |
262 |
up_read(&VFS_I(ip)->i_rwsem); |
653c60b63 xfs: introduce mm... |
263 |
out: |
fa96acadf xfs: move inode l... |
264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 |
return 0; } /* * xfs_iunlock() is used to drop the inode locks acquired with * xfs_ilock() and xfs_ilock_nowait(). The caller must pass * in the flags given to xfs_ilock() or xfs_ilock_nowait() so * that we know which locks to drop. * * ip -- the inode being unlocked * lock_flags -- this parameter indicates the inode's locks to be * to be unlocked. See the comment for xfs_ilock() for a list * of valid values for this parameter. * */ void xfs_iunlock( xfs_inode_t *ip, uint lock_flags) { /* * You can't set both SHARED and EXCL for the same lock, * and only XFS_IOLOCK_SHARED, XFS_IOLOCK_EXCL, XFS_ILOCK_SHARED, * and XFS_ILOCK_EXCL are valid values to set in lock_flags. */ ASSERT((lock_flags & (XFS_IOLOCK_SHARED | XFS_IOLOCK_EXCL)) != (XFS_IOLOCK_SHARED | XFS_IOLOCK_EXCL)); |
653c60b63 xfs: introduce mm... |
291 292 |
ASSERT((lock_flags & (XFS_MMAPLOCK_SHARED | XFS_MMAPLOCK_EXCL)) != (XFS_MMAPLOCK_SHARED | XFS_MMAPLOCK_EXCL)); |
fa96acadf xfs: move inode l... |
293 294 |
ASSERT((lock_flags & (XFS_ILOCK_SHARED | XFS_ILOCK_EXCL)) != (XFS_ILOCK_SHARED | XFS_ILOCK_EXCL)); |
0952c8183 xfs: clean up ino... |
295 |
ASSERT((lock_flags & ~(XFS_LOCK_MASK | XFS_LOCK_SUBCLASS_MASK)) == 0); |
fa96acadf xfs: move inode l... |
296 297 298 |
ASSERT(lock_flags != 0); if (lock_flags & XFS_IOLOCK_EXCL) |
655232183 xfs: remove i_iol... |
299 |
up_write(&VFS_I(ip)->i_rwsem); |
fa96acadf xfs: move inode l... |
300 |
else if (lock_flags & XFS_IOLOCK_SHARED) |
655232183 xfs: remove i_iol... |
301 |
up_read(&VFS_I(ip)->i_rwsem); |
fa96acadf xfs: move inode l... |
302 |
|
653c60b63 xfs: introduce mm... |
303 304 305 306 |
if (lock_flags & XFS_MMAPLOCK_EXCL) mrunlock_excl(&ip->i_mmaplock); else if (lock_flags & XFS_MMAPLOCK_SHARED) mrunlock_shared(&ip->i_mmaplock); |
fa96acadf xfs: move inode l... |
307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 |
if (lock_flags & XFS_ILOCK_EXCL) mrunlock_excl(&ip->i_lock); else if (lock_flags & XFS_ILOCK_SHARED) mrunlock_shared(&ip->i_lock); trace_xfs_iunlock(ip, lock_flags, _RET_IP_); } /* * give up write locks. the i/o lock cannot be held nested * if it is being demoted. */ void xfs_ilock_demote( xfs_inode_t *ip, uint lock_flags) { |
653c60b63 xfs: introduce mm... |
324 325 326 |
ASSERT(lock_flags & (XFS_IOLOCK_EXCL|XFS_MMAPLOCK_EXCL|XFS_ILOCK_EXCL)); ASSERT((lock_flags & ~(XFS_IOLOCK_EXCL|XFS_MMAPLOCK_EXCL|XFS_ILOCK_EXCL)) == 0); |
fa96acadf xfs: move inode l... |
327 328 329 |
if (lock_flags & XFS_ILOCK_EXCL) mrdemote(&ip->i_lock); |
653c60b63 xfs: introduce mm... |
330 331 |
if (lock_flags & XFS_MMAPLOCK_EXCL) mrdemote(&ip->i_mmaplock); |
fa96acadf xfs: move inode l... |
332 |
if (lock_flags & XFS_IOLOCK_EXCL) |
655232183 xfs: remove i_iol... |
333 |
downgrade_write(&VFS_I(ip)->i_rwsem); |
fa96acadf xfs: move inode l... |
334 335 336 |
trace_xfs_ilock_demote(ip, lock_flags, _RET_IP_); } |
742ae1e35 xfs: introduce CO... |
337 |
#if defined(DEBUG) || defined(XFS_WARN) |
fa96acadf xfs: move inode l... |
338 339 340 341 342 343 344 345 346 347 |
int xfs_isilocked( xfs_inode_t *ip, uint lock_flags) { if (lock_flags & (XFS_ILOCK_EXCL|XFS_ILOCK_SHARED)) { if (!(lock_flags & XFS_ILOCK_SHARED)) return !!ip->i_lock.mr_writer; return rwsem_is_locked(&ip->i_lock.mr_lock); } |
653c60b63 xfs: introduce mm... |
348 349 350 351 352 |
if (lock_flags & (XFS_MMAPLOCK_EXCL|XFS_MMAPLOCK_SHARED)) { if (!(lock_flags & XFS_MMAPLOCK_SHARED)) return !!ip->i_mmaplock.mr_writer; return rwsem_is_locked(&ip->i_mmaplock.mr_lock); } |
fa96acadf xfs: move inode l... |
353 354 |
if (lock_flags & (XFS_IOLOCK_EXCL|XFS_IOLOCK_SHARED)) { if (!(lock_flags & XFS_IOLOCK_SHARED)) |
655232183 xfs: remove i_iol... |
355 356 357 |
return !debug_locks || lockdep_is_held_type(&VFS_I(ip)->i_rwsem, 0); return rwsem_is_locked(&VFS_I(ip)->i_rwsem); |
fa96acadf xfs: move inode l... |
358 359 360 361 362 363 |
} ASSERT(0); return 0; } #endif |
b6a9947ef xfs: lockdep anno... |
364 365 366 367 368 369 370 |
/* * xfs_lockdep_subclass_ok() is only used in an ASSERT, so is only called when * DEBUG or XFS_WARN is set. And MAX_LOCKDEP_SUBCLASSES is then only defined * when CONFIG_LOCKDEP is set. Hence the complex define below to avoid build * errors and warnings. */ #if (defined(DEBUG) || defined(XFS_WARN)) && defined(CONFIG_LOCKDEP) |
3403ccc0c xfs: inode lockde... |
371 372 373 374 375 376 377 378 379 |
static bool xfs_lockdep_subclass_ok( int subclass) { return subclass < MAX_LOCKDEP_SUBCLASSES; } #else #define xfs_lockdep_subclass_ok(subclass) (true) #endif |
c24b5dfad xfs: kill xfs_vno... |
380 |
/* |
653c60b63 xfs: introduce mm... |
381 |
* Bump the subclass so xfs_lock_inodes() acquires each lock with a different |
0952c8183 xfs: clean up ino... |
382 383 384 |
* value. This can be called for any type of inode lock combination, including * parent locking. Care must be taken to ensure we don't overrun the subclass * storage fields in the class mask we build. |
c24b5dfad xfs: kill xfs_vno... |
385 386 387 388 |
*/ static inline int xfs_lock_inumorder(int lock_mode, int subclass) { |
0952c8183 xfs: clean up ino... |
389 390 391 392 |
int class = 0; ASSERT(!(lock_mode & (XFS_ILOCK_PARENT | XFS_ILOCK_RTBITMAP | XFS_ILOCK_RTSUM))); |
3403ccc0c xfs: inode lockde... |
393 |
ASSERT(xfs_lockdep_subclass_ok(subclass)); |
0952c8183 xfs: clean up ino... |
394 |
|
653c60b63 xfs: introduce mm... |
395 |
if (lock_mode & (XFS_IOLOCK_SHARED|XFS_IOLOCK_EXCL)) { |
0952c8183 xfs: clean up ino... |
396 |
ASSERT(subclass <= XFS_IOLOCK_MAX_SUBCLASS); |
0952c8183 xfs: clean up ino... |
397 |
class += subclass << XFS_IOLOCK_SHIFT; |
653c60b63 xfs: introduce mm... |
398 399 400 |
} if (lock_mode & (XFS_MMAPLOCK_SHARED|XFS_MMAPLOCK_EXCL)) { |
0952c8183 xfs: clean up ino... |
401 402 |
ASSERT(subclass <= XFS_MMAPLOCK_MAX_SUBCLASS); class += subclass << XFS_MMAPLOCK_SHIFT; |
653c60b63 xfs: introduce mm... |
403 |
} |
0952c8183 xfs: clean up ino... |
404 405 406 407 |
if (lock_mode & (XFS_ILOCK_SHARED|XFS_ILOCK_EXCL)) { ASSERT(subclass <= XFS_ILOCK_MAX_SUBCLASS); class += subclass << XFS_ILOCK_SHIFT; } |
c24b5dfad xfs: kill xfs_vno... |
408 |
|
0952c8183 xfs: clean up ino... |
409 |
return (lock_mode & ~XFS_LOCK_SUBCLASS_MASK) | class; |
c24b5dfad xfs: kill xfs_vno... |
410 411 412 |
} /* |
95afcf5c7 xfs: clean up ino... |
413 414 |
* The following routine will lock n inodes in exclusive mode. We assume the * caller calls us with the inodes in i_ino order. |
c24b5dfad xfs: kill xfs_vno... |
415 |
* |
95afcf5c7 xfs: clean up ino... |
416 417 418 419 420 |
* We need to detect deadlock where an inode that we lock is in the AIL and we * start waiting for another inode that is locked by a thread in a long running * transaction (such as truncate). This can result in deadlock since the long * running trans might need to wait for the inode we just locked in order to * push the tail and free space in the log. |
0952c8183 xfs: clean up ino... |
421 422 423 424 425 |
* * xfs_lock_inodes() can only be used to lock one type of lock at a time - * the iolock, the mmaplock or the ilock, but not more than one at a time. If we * lock more than one at a time, lockdep will report false positives saying we * have violated locking orders. |
c24b5dfad xfs: kill xfs_vno... |
426 |
*/ |
0d5a75e9e xfs: make several... |
427 |
static void |
c24b5dfad xfs: kill xfs_vno... |
428 |
xfs_lock_inodes( |
efe2330fd xfs: remove the x... |
429 430 431 |
struct xfs_inode **ips, int inodes, uint lock_mode) |
c24b5dfad xfs: kill xfs_vno... |
432 |
{ |
efe2330fd xfs: remove the x... |
433 434 |
int attempts = 0, i, j, try_lock; struct xfs_log_item *lp; |
c24b5dfad xfs: kill xfs_vno... |
435 |
|
0952c8183 xfs: clean up ino... |
436 437 438 |
/* * Currently supports between 2 and 5 inodes with exclusive locking. We * support an arbitrary depth of locking here, but absolute limits on |
b63da6c8d xfs: delete dupli... |
439 |
* inodes depend on the type of locking and the limits placed by |
0952c8183 xfs: clean up ino... |
440 441 442 |
* lockdep annotations in xfs_lock_inumorder. These are all checked by * the asserts. */ |
95afcf5c7 xfs: clean up ino... |
443 |
ASSERT(ips && inodes >= 2 && inodes <= 5); |
0952c8183 xfs: clean up ino... |
444 445 446 447 |
ASSERT(lock_mode & (XFS_IOLOCK_EXCL | XFS_MMAPLOCK_EXCL | XFS_ILOCK_EXCL)); ASSERT(!(lock_mode & (XFS_IOLOCK_SHARED | XFS_MMAPLOCK_SHARED | XFS_ILOCK_SHARED))); |
0952c8183 xfs: clean up ino... |
448 449 450 451 452 453 454 455 456 |
ASSERT(!(lock_mode & XFS_MMAPLOCK_EXCL) || inodes <= XFS_MMAPLOCK_MAX_SUBCLASS + 1); ASSERT(!(lock_mode & XFS_ILOCK_EXCL) || inodes <= XFS_ILOCK_MAX_SUBCLASS + 1); if (lock_mode & XFS_IOLOCK_EXCL) { ASSERT(!(lock_mode & (XFS_MMAPLOCK_EXCL | XFS_ILOCK_EXCL))); } else if (lock_mode & XFS_MMAPLOCK_EXCL) ASSERT(!(lock_mode & XFS_ILOCK_EXCL)); |
c24b5dfad xfs: kill xfs_vno... |
457 458 459 |
try_lock = 0; i = 0; |
c24b5dfad xfs: kill xfs_vno... |
460 461 462 |
again: for (; i < inodes; i++) { ASSERT(ips[i]); |
95afcf5c7 xfs: clean up ino... |
463 |
if (i && (ips[i] == ips[i - 1])) /* Already locked */ |
c24b5dfad xfs: kill xfs_vno... |
464 465 466 |
continue; /* |
95afcf5c7 xfs: clean up ino... |
467 468 |
* If try_lock is not set yet, make sure all locked inodes are * not in the AIL. If any are, set try_lock to be used later. |
c24b5dfad xfs: kill xfs_vno... |
469 |
*/ |
c24b5dfad xfs: kill xfs_vno... |
470 471 |
if (!try_lock) { for (j = (i - 1); j >= 0 && !try_lock; j--) { |
b3b14aacc xfs: don't cast i... |
472 |
lp = &ips[j]->i_itemp->ili_item; |
22525c17e xfs: log item fla... |
473 |
if (lp && test_bit(XFS_LI_IN_AIL, &lp->li_flags)) |
c24b5dfad xfs: kill xfs_vno... |
474 |
try_lock++; |
c24b5dfad xfs: kill xfs_vno... |
475 476 477 478 479 480 481 482 483 |
} } /* * If any of the previous locks we have locked is in the AIL, * we must TRY to get the second and subsequent locks. If * we can't get any, we must release all we have * and try again. */ |
95afcf5c7 xfs: clean up ino... |
484 485 486 487 488 489 490 491 492 |
if (!try_lock) { xfs_ilock(ips[i], xfs_lock_inumorder(lock_mode, i)); continue; } /* try_lock means we have an inode locked that is in the AIL. */ ASSERT(i != 0); if (xfs_ilock_nowait(ips[i], xfs_lock_inumorder(lock_mode, i))) continue; |
c24b5dfad xfs: kill xfs_vno... |
493 |
|
95afcf5c7 xfs: clean up ino... |
494 495 496 497 498 499 |
/* * Unlock all previous guys and try again. xfs_iunlock will try * to push the tail if the inode is in the AIL. */ attempts++; for (j = i - 1; j >= 0; j--) { |
c24b5dfad xfs: kill xfs_vno... |
500 |
/* |
95afcf5c7 xfs: clean up ino... |
501 502 503 |
* Check to see if we've already unlocked this one. Not * the first one going back, and the inode ptr is the * same. |
c24b5dfad xfs: kill xfs_vno... |
504 |
*/ |
95afcf5c7 xfs: clean up ino... |
505 506 |
if (j != (i - 1) && ips[j] == ips[j + 1]) continue; |
c24b5dfad xfs: kill xfs_vno... |
507 |
|
95afcf5c7 xfs: clean up ino... |
508 509 |
xfs_iunlock(ips[j], lock_mode); } |
c24b5dfad xfs: kill xfs_vno... |
510 |
|
95afcf5c7 xfs: clean up ino... |
511 512 |
if ((attempts % 5) == 0) { delay(1); /* Don't just spin the CPU */ |
c24b5dfad xfs: kill xfs_vno... |
513 |
} |
95afcf5c7 xfs: clean up ino... |
514 515 516 |
i = 0; try_lock = 0; goto again; |
c24b5dfad xfs: kill xfs_vno... |
517 |
} |
c24b5dfad xfs: kill xfs_vno... |
518 519 520 |
} /* |
653c60b63 xfs: introduce mm... |
521 |
* xfs_lock_two_inodes() can only be used to lock one type of lock at a time - |
7c2d238ac xfs: allow xfs_lo... |
522 523 524 525 526 |
* the mmaplock or the ilock, but not more than one type at a time. If we lock * more than one at a time, lockdep will report false positives saying we have * violated locking orders. The iolock must be double-locked separately since * we use i_rwsem for that. We now support taking one lock EXCL and the other * SHARED. |
c24b5dfad xfs: kill xfs_vno... |
527 528 529 |
*/ void xfs_lock_two_inodes( |
7c2d238ac xfs: allow xfs_lo... |
530 531 532 533 |
struct xfs_inode *ip0, uint ip0_mode, struct xfs_inode *ip1, uint ip1_mode) |
c24b5dfad xfs: kill xfs_vno... |
534 |
{ |
7c2d238ac xfs: allow xfs_lo... |
535 536 |
struct xfs_inode *temp; uint mode_temp; |
c24b5dfad xfs: kill xfs_vno... |
537 |
int attempts = 0; |
efe2330fd xfs: remove the x... |
538 |
struct xfs_log_item *lp; |
c24b5dfad xfs: kill xfs_vno... |
539 |
|
7c2d238ac xfs: allow xfs_lo... |
540 541 542 543 544 545 546 547 548 549 550 551 |
ASSERT(hweight32(ip0_mode) == 1); ASSERT(hweight32(ip1_mode) == 1); ASSERT(!(ip0_mode & (XFS_IOLOCK_SHARED|XFS_IOLOCK_EXCL))); ASSERT(!(ip1_mode & (XFS_IOLOCK_SHARED|XFS_IOLOCK_EXCL))); ASSERT(!(ip0_mode & (XFS_MMAPLOCK_SHARED|XFS_MMAPLOCK_EXCL)) || !(ip0_mode & (XFS_ILOCK_SHARED|XFS_ILOCK_EXCL))); ASSERT(!(ip1_mode & (XFS_MMAPLOCK_SHARED|XFS_MMAPLOCK_EXCL)) || !(ip1_mode & (XFS_ILOCK_SHARED|XFS_ILOCK_EXCL))); ASSERT(!(ip1_mode & (XFS_MMAPLOCK_SHARED|XFS_MMAPLOCK_EXCL)) || !(ip0_mode & (XFS_ILOCK_SHARED|XFS_ILOCK_EXCL))); ASSERT(!(ip0_mode & (XFS_MMAPLOCK_SHARED|XFS_MMAPLOCK_EXCL)) || !(ip1_mode & (XFS_ILOCK_SHARED|XFS_ILOCK_EXCL))); |
653c60b63 xfs: introduce mm... |
552 |
|
c24b5dfad xfs: kill xfs_vno... |
553 554 555 556 557 558 |
ASSERT(ip0->i_ino != ip1->i_ino); if (ip0->i_ino > ip1->i_ino) { temp = ip0; ip0 = ip1; ip1 = temp; |
7c2d238ac xfs: allow xfs_lo... |
559 560 561 |
mode_temp = ip0_mode; ip0_mode = ip1_mode; ip1_mode = mode_temp; |
c24b5dfad xfs: kill xfs_vno... |
562 563 564 |
} again: |
7c2d238ac xfs: allow xfs_lo... |
565 |
xfs_ilock(ip0, xfs_lock_inumorder(ip0_mode, 0)); |
c24b5dfad xfs: kill xfs_vno... |
566 567 568 569 570 571 |
/* * If the first lock we have locked is in the AIL, we must TRY to get * the second lock. If we can't get it, we must release the first one * and try again. */ |
b3b14aacc xfs: don't cast i... |
572 |
lp = &ip0->i_itemp->ili_item; |
22525c17e xfs: log item fla... |
573 |
if (lp && test_bit(XFS_LI_IN_AIL, &lp->li_flags)) { |
7c2d238ac xfs: allow xfs_lo... |
574 575 |
if (!xfs_ilock_nowait(ip1, xfs_lock_inumorder(ip1_mode, 1))) { xfs_iunlock(ip0, ip0_mode); |
c24b5dfad xfs: kill xfs_vno... |
576 577 578 579 580 |
if ((++attempts % 5) == 0) delay(1); /* Don't just spin the CPU */ goto again; } } else { |
7c2d238ac xfs: allow xfs_lo... |
581 |
xfs_ilock(ip1, xfs_lock_inumorder(ip1_mode, 1)); |
c24b5dfad xfs: kill xfs_vno... |
582 583 |
} } |
1da177e4c Linux-2.6.12-rc2 |
584 585 |
STATIC uint _xfs_dic2xflags( |
c8ce540db xfs: remove doubl... |
586 |
uint16_t di_flags, |
58f88ca2d xfs: introduce pe... |
587 588 |
uint64_t di_flags2, bool has_attr) |
1da177e4c Linux-2.6.12-rc2 |
589 590 591 592 593 |
{ uint flags = 0; if (di_flags & XFS_DIFLAG_ANY) { if (di_flags & XFS_DIFLAG_REALTIME) |
e7b894810 xfs: use FS_XFLAG... |
594 |
flags |= FS_XFLAG_REALTIME; |
1da177e4c Linux-2.6.12-rc2 |
595 |
if (di_flags & XFS_DIFLAG_PREALLOC) |
e7b894810 xfs: use FS_XFLAG... |
596 |
flags |= FS_XFLAG_PREALLOC; |
1da177e4c Linux-2.6.12-rc2 |
597 |
if (di_flags & XFS_DIFLAG_IMMUTABLE) |
e7b894810 xfs: use FS_XFLAG... |
598 |
flags |= FS_XFLAG_IMMUTABLE; |
1da177e4c Linux-2.6.12-rc2 |
599 |
if (di_flags & XFS_DIFLAG_APPEND) |
e7b894810 xfs: use FS_XFLAG... |
600 |
flags |= FS_XFLAG_APPEND; |
1da177e4c Linux-2.6.12-rc2 |
601 |
if (di_flags & XFS_DIFLAG_SYNC) |
e7b894810 xfs: use FS_XFLAG... |
602 |
flags |= FS_XFLAG_SYNC; |
1da177e4c Linux-2.6.12-rc2 |
603 |
if (di_flags & XFS_DIFLAG_NOATIME) |
e7b894810 xfs: use FS_XFLAG... |
604 |
flags |= FS_XFLAG_NOATIME; |
1da177e4c Linux-2.6.12-rc2 |
605 |
if (di_flags & XFS_DIFLAG_NODUMP) |
e7b894810 xfs: use FS_XFLAG... |
606 |
flags |= FS_XFLAG_NODUMP; |
1da177e4c Linux-2.6.12-rc2 |
607 |
if (di_flags & XFS_DIFLAG_RTINHERIT) |
e7b894810 xfs: use FS_XFLAG... |
608 |
flags |= FS_XFLAG_RTINHERIT; |
1da177e4c Linux-2.6.12-rc2 |
609 |
if (di_flags & XFS_DIFLAG_PROJINHERIT) |
e7b894810 xfs: use FS_XFLAG... |
610 |
flags |= FS_XFLAG_PROJINHERIT; |
1da177e4c Linux-2.6.12-rc2 |
611 |
if (di_flags & XFS_DIFLAG_NOSYMLINKS) |
e7b894810 xfs: use FS_XFLAG... |
612 |
flags |= FS_XFLAG_NOSYMLINKS; |
dd9f438e3 [XFS] Implement t... |
613 |
if (di_flags & XFS_DIFLAG_EXTSIZE) |
e7b894810 xfs: use FS_XFLAG... |
614 |
flags |= FS_XFLAG_EXTSIZE; |
dd9f438e3 [XFS] Implement t... |
615 |
if (di_flags & XFS_DIFLAG_EXTSZINHERIT) |
e7b894810 xfs: use FS_XFLAG... |
616 |
flags |= FS_XFLAG_EXTSZINHERIT; |
d3446eac3 [XFS] Add degfram... |
617 |
if (di_flags & XFS_DIFLAG_NODEFRAG) |
e7b894810 xfs: use FS_XFLAG... |
618 |
flags |= FS_XFLAG_NODEFRAG; |
2a82b8be8 [XFS] Concurrent ... |
619 |
if (di_flags & XFS_DIFLAG_FILESTREAM) |
e7b894810 xfs: use FS_XFLAG... |
620 |
flags |= FS_XFLAG_FILESTREAM; |
1da177e4c Linux-2.6.12-rc2 |
621 |
} |
58f88ca2d xfs: introduce pe... |
622 623 624 |
if (di_flags2 & XFS_DIFLAG2_ANY) { if (di_flags2 & XFS_DIFLAG2_DAX) flags |= FS_XFLAG_DAX; |
f7ca35227 xfs: create a sep... |
625 626 |
if (di_flags2 & XFS_DIFLAG2_COWEXTSIZE) flags |= FS_XFLAG_COWEXTSIZE; |
58f88ca2d xfs: introduce pe... |
627 628 629 630 |
} if (has_attr) flags |= FS_XFLAG_HASATTR; |
1da177e4c Linux-2.6.12-rc2 |
631 632 633 634 635 |
return flags; } uint xfs_ip2xflags( |
58f88ca2d xfs: introduce pe... |
636 |
struct xfs_inode *ip) |
1da177e4c Linux-2.6.12-rc2 |
637 |
{ |
58f88ca2d xfs: introduce pe... |
638 |
struct xfs_icdinode *dic = &ip->i_d; |
1da177e4c Linux-2.6.12-rc2 |
639 |
|
58f88ca2d xfs: introduce pe... |
640 |
return _xfs_dic2xflags(dic->di_flags, dic->di_flags2, XFS_IFORK_Q(ip)); |
1da177e4c Linux-2.6.12-rc2 |
641 |
} |
1da177e4c Linux-2.6.12-rc2 |
642 |
/* |
c24b5dfad xfs: kill xfs_vno... |
643 644 645 646 647 648 649 650 651 652 653 654 655 656 |
* Lookups up an inode from "name". If ci_name is not NULL, then a CI match * is allowed, otherwise it has to be an exact match. If a CI match is found, * ci_name->name will point to a the actual name (caller must free) or * will be set to NULL if an exact match is found. */ int xfs_lookup( xfs_inode_t *dp, struct xfs_name *name, xfs_inode_t **ipp, struct xfs_name *ci_name) { xfs_ino_t inum; int error; |
c24b5dfad xfs: kill xfs_vno... |
657 658 659 660 |
trace_xfs_lookup(dp, name); if (XFS_FORCED_SHUTDOWN(dp->i_mount)) |
2451337dd xfs: global error... |
661 |
return -EIO; |
c24b5dfad xfs: kill xfs_vno... |
662 |
|
c24b5dfad xfs: kill xfs_vno... |
663 |
error = xfs_dir_lookup(NULL, dp, name, &inum, ci_name); |
c24b5dfad xfs: kill xfs_vno... |
664 |
if (error) |
dbad7c993 xfs: stop holding... |
665 |
goto out_unlock; |
c24b5dfad xfs: kill xfs_vno... |
666 667 668 669 670 671 672 673 674 675 |
error = xfs_iget(dp->i_mount, NULL, inum, 0, 0, ipp); if (error) goto out_free_name; return 0; out_free_name: if (ci_name) kmem_free(ci_name->name); |
dbad7c993 xfs: stop holding... |
676 |
out_unlock: |
c24b5dfad xfs: kill xfs_vno... |
677 678 679 |
*ipp = NULL; return error; } |
8a569d717 xfs: refactor ino... |
680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 |
/* Propagate di_flags from a parent inode to a child inode. */ static void xfs_inode_inherit_flags( struct xfs_inode *ip, const struct xfs_inode *pip) { unsigned int di_flags = 0; umode_t mode = VFS_I(ip)->i_mode; if (S_ISDIR(mode)) { if (pip->i_d.di_flags & XFS_DIFLAG_RTINHERIT) di_flags |= XFS_DIFLAG_RTINHERIT; if (pip->i_d.di_flags & XFS_DIFLAG_EXTSZINHERIT) { di_flags |= XFS_DIFLAG_EXTSZINHERIT; ip->i_d.di_extsize = pip->i_d.di_extsize; } if (pip->i_d.di_flags & XFS_DIFLAG_PROJINHERIT) di_flags |= XFS_DIFLAG_PROJINHERIT; } else if (S_ISREG(mode)) { |
d4f2c14cc xfs: don't propag... |
699 700 |
if ((pip->i_d.di_flags & XFS_DIFLAG_RTINHERIT) && xfs_sb_version_hasrealtime(&ip->i_mount->m_sb)) |
8a569d717 xfs: refactor ino... |
701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 |
di_flags |= XFS_DIFLAG_REALTIME; if (pip->i_d.di_flags & XFS_DIFLAG_EXTSZINHERIT) { di_flags |= XFS_DIFLAG_EXTSIZE; ip->i_d.di_extsize = pip->i_d.di_extsize; } } if ((pip->i_d.di_flags & XFS_DIFLAG_NOATIME) && xfs_inherit_noatime) di_flags |= XFS_DIFLAG_NOATIME; if ((pip->i_d.di_flags & XFS_DIFLAG_NODUMP) && xfs_inherit_nodump) di_flags |= XFS_DIFLAG_NODUMP; if ((pip->i_d.di_flags & XFS_DIFLAG_SYNC) && xfs_inherit_sync) di_flags |= XFS_DIFLAG_SYNC; if ((pip->i_d.di_flags & XFS_DIFLAG_NOSYMLINKS) && xfs_inherit_nosymlinks) di_flags |= XFS_DIFLAG_NOSYMLINKS; if ((pip->i_d.di_flags & XFS_DIFLAG_NODEFRAG) && xfs_inherit_nodefrag) di_flags |= XFS_DIFLAG_NODEFRAG; if (pip->i_d.di_flags & XFS_DIFLAG_FILESTREAM) di_flags |= XFS_DIFLAG_FILESTREAM; ip->i_d.di_flags |= di_flags; } /* Propagate di_flags2 from a parent inode to a child inode. */ static void xfs_inode_inherit_flags2( struct xfs_inode *ip, const struct xfs_inode *pip) { if (pip->i_d.di_flags2 & XFS_DIFLAG2_COWEXTSIZE) { ip->i_d.di_flags2 |= XFS_DIFLAG2_COWEXTSIZE; ip->i_d.di_cowextsize = pip->i_d.di_cowextsize; } if (pip->i_d.di_flags2 & XFS_DIFLAG2_DAX) ip->i_d.di_flags2 |= XFS_DIFLAG2_DAX; } |
c24b5dfad xfs: kill xfs_vno... |
741 |
/* |
1da177e4c Linux-2.6.12-rc2 |
742 743 744 745 746 747 |
* Allocate an inode on disk and return a copy of its in-core version. * The in-core inode is locked exclusively. Set mode, nlink, and rdev * appropriately within the inode. The uid and gid for the inode are * set according to the contents of the given cred structure. * * Use xfs_dialloc() to allocate the on-disk inode. If xfs_dialloc() |
cd856db69 xfs: Update inode... |
748 749 750 751 |
* has a free inode available, call xfs_iget() to obtain the in-core * version of the allocated inode. Finally, fill in the inode and * log its initial contents. In this case, ialloc_context would be * set to NULL. |
1da177e4c Linux-2.6.12-rc2 |
752 |
* |
cd856db69 xfs: Update inode... |
753 754 755 756 757 |
* If xfs_dialloc() does not have an available inode, it will replenish * its supply by doing an allocation. Since we can only do one * allocation within a transaction without deadlocks, we must commit * the current transaction before returning the inode itself. * In this case, therefore, we will set ialloc_context and return. |
1da177e4c Linux-2.6.12-rc2 |
758 759 760 761 762 763 764 765 |
* The caller should then commit the current transaction, start a new * transaction, and call xfs_ialloc() again to actually get the inode. * * To ensure that some other process does not grab the inode that * was allocated during the first call to xfs_ialloc(), this routine * also returns the [locked] bp pointing to the head of the freelist * as ialloc_context. The caller should hold this buffer across * the commit and pass it back into this routine on the second call. |
b11f94d53 [XFS] Quota inode... |
766 767 768 769 770 |
* * If we are allocating quota inodes, we do not have a parent inode * to attach to or associate with (i.e. pip == NULL) because they * are not linked into the directory structure - they are attached * directly to the superblock - and so have no parent. |
1da177e4c Linux-2.6.12-rc2 |
771 |
*/ |
0d5a75e9e xfs: make several... |
772 |
static int |
1da177e4c Linux-2.6.12-rc2 |
773 774 775 |
xfs_ialloc( xfs_trans_t *tp, xfs_inode_t *pip, |
576b1d67c xfs: propagate um... |
776 |
umode_t mode, |
31b084aef [XFS] Fix up uses... |
777 |
xfs_nlink_t nlink, |
66f364649 xfs: remove if_rdev |
778 |
dev_t rdev, |
6743099ce xfs: Extend proje... |
779 |
prid_t prid, |
1da177e4c Linux-2.6.12-rc2 |
780 |
xfs_buf_t **ialloc_context, |
1da177e4c Linux-2.6.12-rc2 |
781 782 |
xfs_inode_t **ipp) { |
93848a999 xfs: add version ... |
783 |
struct xfs_mount *mp = tp->t_mountp; |
1da177e4c Linux-2.6.12-rc2 |
784 785 |
xfs_ino_t ino; xfs_inode_t *ip; |
1da177e4c Linux-2.6.12-rc2 |
786 787 |
uint flags; int error; |
95582b008 vfs: change inode... |
788 |
struct timespec64 tv; |
3987848c7 xfs: remove times... |
789 |
struct inode *inode; |
1da177e4c Linux-2.6.12-rc2 |
790 791 792 793 794 |
/* * Call the space management code to pick * the on-disk inode to be allocated. */ |
f59cf5c29 xfs: remove "no-a... |
795 |
error = xfs_dialloc(tp, pip ? pip->i_ino : 0, mode, |
08358906e xfs: remove the a... |
796 |
ialloc_context, &ino); |
bf904248a [XFS] Combine the... |
797 |
if (error) |
1da177e4c Linux-2.6.12-rc2 |
798 |
return error; |
08358906e xfs: remove the a... |
799 |
if (*ialloc_context || ino == NULLFSINO) { |
1da177e4c Linux-2.6.12-rc2 |
800 801 802 803 804 805 |
*ipp = NULL; return 0; } ASSERT(*ialloc_context == NULL); /* |
8b26984db xfs: validate all... |
806 807 808 809 810 811 812 813 814 815 816 817 |
* Protect against obviously corrupt allocation btree records. Later * xfs_iget checks will catch re-allocation of other active in-memory * and on-disk inodes. If we don't catch reallocating the parent inode * here we will deadlock in xfs_iget() so we have to do these checks * first. */ if ((pip && ino == pip->i_ino) || !xfs_verify_dir_ino(mp, ino)) { xfs_alert(mp, "Allocated a known in-use inode 0x%llx!", ino); return -EFSCORRUPTED; } /* |
1da177e4c Linux-2.6.12-rc2 |
818 819 820 821 |
* Get the in-core inode with the lock held exclusively. * This is because we're setting fields here we need * to prevent others from looking at until we're done. */ |
93848a999 xfs: add version ... |
822 |
error = xfs_iget(mp, tp, ino, XFS_IGET_CREATE, |
ec3ba85f4 xfs: more sensibl... |
823 |
XFS_ILOCK_EXCL, &ip); |
bf904248a [XFS] Combine the... |
824 |
if (error) |
1da177e4c Linux-2.6.12-rc2 |
825 |
return error; |
1da177e4c Linux-2.6.12-rc2 |
826 |
ASSERT(ip != NULL); |
3987848c7 xfs: remove times... |
827 |
inode = VFS_I(ip); |
c19b3b05a xfs: mode di_mode... |
828 |
inode->i_mode = mode; |
54d7b5c1d xfs: use vfs inod... |
829 |
set_nlink(inode, nlink); |
3d8f28215 xfs: ensure that ... |
830 |
inode->i_uid = current_fsuid(); |
66f364649 xfs: remove if_rdev |
831 |
inode->i_rdev = rdev; |
de7a866fd xfs: merge the pr... |
832 |
ip->i_d.di_projid = prid; |
1da177e4c Linux-2.6.12-rc2 |
833 |
|
bd186aa90 [XFS] kill the vf... |
834 |
if (pip && XFS_INHERIT_GID(pip)) { |
3d8f28215 xfs: ensure that ... |
835 |
inode->i_gid = VFS_I(pip)->i_gid; |
c19b3b05a xfs: mode di_mode... |
836 837 |
if ((VFS_I(pip)->i_mode & S_ISGID) && S_ISDIR(mode)) inode->i_mode |= S_ISGID; |
3d8f28215 xfs: ensure that ... |
838 839 |
} else { inode->i_gid = current_fsgid(); |
1da177e4c Linux-2.6.12-rc2 |
840 841 842 843 844 845 846 |
} /* * If the group ID of the new file does not match the effective group * ID or one of the supplementary group IDs, the S_ISGID bit is cleared * (and only if the irix_sgid_inherit compatibility variable is set). */ |
542951592 xfs: remove the i... |
847 848 |
if (irix_sgid_inherit && (inode->i_mode & S_ISGID) && !in_group_p(inode->i_gid)) |
c19b3b05a xfs: mode di_mode... |
849 |
inode->i_mode &= ~S_ISGID; |
1da177e4c Linux-2.6.12-rc2 |
850 851 |
ip->i_d.di_size = 0; |
daf83964a xfs: move the per... |
852 |
ip->i_df.if_nextents = 0; |
1da177e4c Linux-2.6.12-rc2 |
853 |
ASSERT(ip->i_d.di_nblocks == 0); |
dff35fd41 [XFS] update time... |
854 |
|
c2050a454 fs: Replace curre... |
855 |
tv = current_time(inode); |
3987848c7 xfs: remove times... |
856 857 858 |
inode->i_mtime = tv; inode->i_atime = tv; inode->i_ctime = tv; |
dff35fd41 [XFS] update time... |
859 |
|
1da177e4c Linux-2.6.12-rc2 |
860 861 862 863 |
ip->i_d.di_extsize = 0; ip->i_d.di_dmevmask = 0; ip->i_d.di_dmstate = 0; ip->i_d.di_flags = 0; |
93848a999 xfs: add version ... |
864 |
|
6471e9c5e xfs: remove the d... |
865 |
if (xfs_sb_version_has_v3inode(&mp->m_sb)) { |
f0e282806 xfs: convert to n... |
866 |
inode_set_iversion(inode, 1); |
f93e5436f xfs: widen ondisk... |
867 |
ip->i_d.di_flags2 = mp->m_ino_geo.new_diflags2; |
f7ca35227 xfs: create a sep... |
868 |
ip->i_d.di_cowextsize = 0; |
8d2d878db xfs: use a struct... |
869 |
ip->i_d.di_crtime = tv; |
93848a999 xfs: add version ... |
870 |
} |
1da177e4c Linux-2.6.12-rc2 |
871 872 873 874 875 876 |
flags = XFS_ILOG_CORE; switch (mode & S_IFMT) { case S_IFIFO: case S_IFCHR: case S_IFBLK: case S_IFSOCK: |
f7e67b20e xfs: move the for... |
877 |
ip->i_df.if_format = XFS_DINODE_FMT_DEV; |
1da177e4c Linux-2.6.12-rc2 |
878 879 880 881 882 |
ip->i_df.if_flags = 0; flags |= XFS_ILOG_DEV; break; case S_IFREG: case S_IFDIR: |
8a569d717 xfs: refactor ino... |
883 884 885 886 |
if (pip && (pip->i_d.di_flags & XFS_DIFLAG_ANY)) xfs_inode_inherit_flags(ip, pip); if (pip && (pip->i_d.di_flags2 & XFS_DIFLAG2_ANY)) xfs_inode_inherit_flags2(ip, pip); |
1da177e4c Linux-2.6.12-rc2 |
887 888 |
/* FALLTHROUGH */ case S_IFLNK: |
f7e67b20e xfs: move the for... |
889 |
ip->i_df.if_format = XFS_DINODE_FMT_EXTENTS; |
1da177e4c Linux-2.6.12-rc2 |
890 |
ip->i_df.if_flags = XFS_IFEXTENTS; |
fcacbc3f5 xfs: remove if_re... |
891 |
ip->i_df.if_bytes = 0; |
6bdcf26ad xfs: use a b+tree... |
892 |
ip->i_df.if_u1.if_root = NULL; |
1da177e4c Linux-2.6.12-rc2 |
893 894 895 896 |
break; default: ASSERT(0); } |
1da177e4c Linux-2.6.12-rc2 |
897 898 899 900 |
/* * Log the new values stuffed into the inode. */ |
ddc3415ab xfs: simplify xfs... |
901 |
xfs_trans_ijoin(tp, ip, XFS_ILOCK_EXCL); |
1da177e4c Linux-2.6.12-rc2 |
902 |
xfs_trans_log_inode(tp, ip, flags); |
58c904734 xfs: inodes are n... |
903 |
/* now that we have an i_mode we can setup the inode structure */ |
41be8bed1 [XFS] sanitize xf... |
904 |
xfs_setup_inode(ip); |
1da177e4c Linux-2.6.12-rc2 |
905 906 907 908 |
*ipp = ip; return 0; } |
e546cb79e xfs: consolidate ... |
909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 |
/* * Allocates a new inode from disk and return a pointer to the * incore copy. This routine will internally commit the current * transaction and allocate a new one if the Space Manager needed * to do an allocation to replenish the inode free-list. * * This routine is designed to be called from xfs_create and * xfs_create_dir. * */ int xfs_dir_ialloc( xfs_trans_t **tpp, /* input: current transaction; output: may be a new transaction. */ xfs_inode_t *dp, /* directory within whose allocate the inode. */ umode_t mode, xfs_nlink_t nlink, |
66f364649 xfs: remove if_rdev |
927 |
dev_t rdev, |
e546cb79e xfs: consolidate ... |
928 |
prid_t prid, /* project id */ |
c959025ed xfs: Remove "comm... |
929 |
xfs_inode_t **ipp) /* pointer to inode; it will be |
e546cb79e xfs: consolidate ... |
930 |
locked. */ |
e546cb79e xfs: consolidate ... |
931 932 |
{ xfs_trans_t *tp; |
e546cb79e xfs: consolidate ... |
933 934 935 |
xfs_inode_t *ip; xfs_buf_t *ialloc_context = NULL; int code; |
e546cb79e xfs: consolidate ... |
936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 |
void *dqinfo; uint tflags; tp = *tpp; ASSERT(tp->t_flags & XFS_TRANS_PERM_LOG_RES); /* * xfs_ialloc will return a pointer to an incore inode if * the Space Manager has an available inode on the free * list. Otherwise, it will do an allocation and replenish * the freelist. Since we can only do one allocation per * transaction without deadlocks, we will need to commit the * current transaction and start a new one. We will then * need to call xfs_ialloc again to get the inode. * * If xfs_ialloc did an allocation to replenish the freelist, * it returns the bp containing the head of the freelist as * ialloc_context. We will hold a lock on it across the * transaction commit so that no other process can steal * the inode(s) that we've just allocated. */ |
f59cf5c29 xfs: remove "no-a... |
957 958 |
code = xfs_ialloc(tp, dp, mode, nlink, rdev, prid, &ialloc_context, &ip); |
e546cb79e xfs: consolidate ... |
959 960 961 962 963 964 965 966 967 968 969 970 |
/* * Return an error if we were unable to allocate a new inode. * This should only happen if we run out of space on disk or * encounter a disk error. */ if (code) { *ipp = NULL; return code; } if (!ialloc_context && !ip) { *ipp = NULL; |
2451337dd xfs: global error... |
971 |
return -ENOSPC; |
e546cb79e xfs: consolidate ... |
972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 |
} /* * If the AGI buffer is non-NULL, then we were unable to get an * inode in one operation. We need to commit the current * transaction and call xfs_ialloc() again. It is guaranteed * to succeed the second time. */ if (ialloc_context) { /* * Normally, xfs_trans_commit releases all the locks. * We call bhold to hang on to the ialloc_context across * the commit. Holding this buffer prevents any other * processes from doing any allocations in this * allocation group. */ xfs_trans_bhold(tp, ialloc_context); |
e546cb79e xfs: consolidate ... |
989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002 |
/* * We want the quota changes to be associated with the next * transaction, NOT this one. So, detach the dqinfo from this * and attach it to the next transaction. */ dqinfo = NULL; tflags = 0; if (tp->t_dqinfo) { dqinfo = (void *)tp->t_dqinfo; tp->t_dqinfo = NULL; tflags = tp->t_flags & XFS_TRANS_DQ_DIRTY; tp->t_flags &= ~(XFS_TRANS_DQ_DIRTY); } |
411350df1 xfs: refactor xfs... |
1003 |
code = xfs_trans_roll(&tp); |
3d3c8b522 xfs: refactor xfs... |
1004 |
|
e546cb79e xfs: consolidate ... |
1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 |
/* * Re-attach the quota info that we detached from prev trx. */ if (dqinfo) { tp->t_dqinfo = dqinfo; tp->t_flags |= tflags; } if (code) { xfs_buf_relse(ialloc_context); |
2e6db6c4c xfs: switch remai... |
1015 |
*tpp = tp; |
e546cb79e xfs: consolidate ... |
1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 |
*ipp = NULL; return code; } xfs_trans_bjoin(tp, ialloc_context); /* * Call ialloc again. Since we've locked out all * other allocations in this allocation group, * this call should always succeed. */ code = xfs_ialloc(tp, dp, mode, nlink, rdev, prid, |
f59cf5c29 xfs: remove "no-a... |
1027 |
&ialloc_context, &ip); |
e546cb79e xfs: consolidate ... |
1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 |
/* * If we get an error at this point, return to the caller * so that the current transaction can be aborted. */ if (code) { *tpp = tp; *ipp = NULL; return code; } ASSERT(!ialloc_context && ip); |
e546cb79e xfs: consolidate ... |
1039 1040 1041 1042 1043 1044 1045 1046 1047 |
} *ipp = ip; *tpp = tp; return 0; } /* |
54d7b5c1d xfs: use vfs inod... |
1048 1049 1050 |
* Decrement the link count on an inode & log the change. If this causes the * link count to go to zero, move the inode to AGI unlinked list so that it can * be freed when the last active reference goes away via xfs_inactive(). |
e546cb79e xfs: consolidate ... |
1051 |
*/ |
0d5a75e9e xfs: make several... |
1052 |
static int /* error */ |
e546cb79e xfs: consolidate ... |
1053 1054 1055 1056 |
xfs_droplink( xfs_trans_t *tp, xfs_inode_t *ip) { |
e546cb79e xfs: consolidate ... |
1057 |
xfs_trans_ichgtime(tp, ip, XFS_ICHGTIME_CHG); |
e546cb79e xfs: consolidate ... |
1058 1059 |
drop_nlink(VFS_I(ip)); xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE); |
54d7b5c1d xfs: use vfs inod... |
1060 1061 1062 1063 |
if (VFS_I(ip)->i_nlink) return 0; return xfs_iunlink(tp, ip); |
e546cb79e xfs: consolidate ... |
1064 1065 1066 |
} /* |
e546cb79e xfs: consolidate ... |
1067 1068 |
* Increment the link count on an inode & log the change. */ |
910832697 xfs: change some ... |
1069 |
static void |
e546cb79e xfs: consolidate ... |
1070 1071 1072 1073 1074 |
xfs_bumplink( xfs_trans_t *tp, xfs_inode_t *ip) { xfs_trans_ichgtime(tp, ip, XFS_ICHGTIME_CHG); |
e546cb79e xfs: consolidate ... |
1075 |
inc_nlink(VFS_I(ip)); |
e546cb79e xfs: consolidate ... |
1076 |
xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE); |
e546cb79e xfs: consolidate ... |
1077 |
} |
c24b5dfad xfs: kill xfs_vno... |
1078 1079 1080 1081 1082 |
int xfs_create( xfs_inode_t *dp, struct xfs_name *name, umode_t mode, |
66f364649 xfs: remove if_rdev |
1083 |
dev_t rdev, |
c24b5dfad xfs: kill xfs_vno... |
1084 1085 1086 1087 1088 1089 1090 |
xfs_inode_t **ipp) { int is_dir = S_ISDIR(mode); struct xfs_mount *mp = dp->i_mount; struct xfs_inode *ip = NULL; struct xfs_trans *tp = NULL; int error; |
c24b5dfad xfs: kill xfs_vno... |
1091 |
bool unlock_dp_on_error = false; |
c24b5dfad xfs: kill xfs_vno... |
1092 1093 1094 1095 |
prid_t prid; struct xfs_dquot *udqp = NULL; struct xfs_dquot *gdqp = NULL; struct xfs_dquot *pdqp = NULL; |
062647a8b xfs: replace on-s... |
1096 |
struct xfs_trans_res *tres; |
c24b5dfad xfs: kill xfs_vno... |
1097 |
uint resblks; |
c24b5dfad xfs: kill xfs_vno... |
1098 1099 1100 1101 |
trace_xfs_create(dp, name); if (XFS_FORCED_SHUTDOWN(mp)) |
2451337dd xfs: global error... |
1102 |
return -EIO; |
c24b5dfad xfs: kill xfs_vno... |
1103 |
|
163467d37 xfs: factor prid ... |
1104 |
prid = xfs_get_initial_prid(dp); |
c24b5dfad xfs: kill xfs_vno... |
1105 1106 1107 1108 |
/* * Make sure that we have allocated dquot(s) on disk. */ |
542951592 xfs: remove the i... |
1109 |
error = xfs_qm_vop_dqalloc(dp, current_fsuid(), current_fsgid(), prid, |
c24b5dfad xfs: kill xfs_vno... |
1110 1111 1112 1113 1114 1115 |
XFS_QMOPT_QUOTALL | XFS_QMOPT_INHERIT, &udqp, &gdqp, &pdqp); if (error) return error; if (is_dir) { |
c24b5dfad xfs: kill xfs_vno... |
1116 |
resblks = XFS_MKDIR_SPACE_RES(mp, name->len); |
062647a8b xfs: replace on-s... |
1117 |
tres = &M_RES(mp)->tr_mkdir; |
c24b5dfad xfs: kill xfs_vno... |
1118 1119 |
} else { resblks = XFS_CREATE_SPACE_RES(mp, name->len); |
062647a8b xfs: replace on-s... |
1120 |
tres = &M_RES(mp)->tr_create; |
c24b5dfad xfs: kill xfs_vno... |
1121 |
} |
c24b5dfad xfs: kill xfs_vno... |
1122 1123 1124 1125 1126 1127 |
/* * Initially assume that the file does not exist and * reserve the resources for that case. If that is not * the case we'll drop the one we have and get a more * appropriate transaction later. */ |
253f4911f xfs: better xfs_t... |
1128 |
error = xfs_trans_alloc(mp, tres, resblks, 0, 0, &tp); |
2451337dd xfs: global error... |
1129 |
if (error == -ENOSPC) { |
c24b5dfad xfs: kill xfs_vno... |
1130 1131 |
/* flush outstanding delalloc blocks and retry */ xfs_flush_inodes(mp); |
253f4911f xfs: better xfs_t... |
1132 |
error = xfs_trans_alloc(mp, tres, resblks, 0, 0, &tp); |
c24b5dfad xfs: kill xfs_vno... |
1133 |
} |
4906e2154 xfs: remove the f... |
1134 |
if (error) |
253f4911f xfs: better xfs_t... |
1135 |
goto out_release_inode; |
c24b5dfad xfs: kill xfs_vno... |
1136 |
|
655232183 xfs: remove i_iol... |
1137 |
xfs_ilock(dp, XFS_ILOCK_EXCL | XFS_ILOCK_PARENT); |
c24b5dfad xfs: kill xfs_vno... |
1138 |
unlock_dp_on_error = true; |
c24b5dfad xfs: kill xfs_vno... |
1139 1140 1141 1142 1143 1144 1145 |
/* * Reserve disk quota and the inode. */ error = xfs_trans_reserve_quota(tp, mp, udqp, gdqp, pdqp, resblks, 1, 0); if (error) goto out_trans_cancel; |
c24b5dfad xfs: kill xfs_vno... |
1146 1147 1148 1149 1150 |
/* * A newly created regular or special file just has one directory * entry pointing to them, but a directory also the "." entry * pointing to itself. */ |
c959025ed xfs: Remove "comm... |
1151 |
error = xfs_dir_ialloc(&tp, dp, mode, is_dir ? 2 : 1, rdev, prid, &ip); |
d6077aa33 xfs: Remove dupli... |
1152 |
if (error) |
4906e2154 xfs: remove the f... |
1153 |
goto out_trans_cancel; |
c24b5dfad xfs: kill xfs_vno... |
1154 1155 1156 1157 1158 1159 1160 1161 |
/* * Now we join the directory inode to the transaction. We do not do it * earlier because xfs_dir_ialloc might commit the previous transaction * (and release all the locks). An error from here on will result in * the transaction cancel unlocking dp so don't do it explicitly in the * error path. */ |
655232183 xfs: remove i_iol... |
1162 |
xfs_trans_ijoin(tp, dp, XFS_ILOCK_EXCL); |
c24b5dfad xfs: kill xfs_vno... |
1163 |
unlock_dp_on_error = false; |
381eee69f xfs: remove first... |
1164 |
error = xfs_dir_createname(tp, dp, name, ip->i_ino, |
63337b63e xfs: remove unnec... |
1165 |
resblks - XFS_IALLOC_SPACE_RES(mp)); |
c24b5dfad xfs: kill xfs_vno... |
1166 |
if (error) { |
2451337dd xfs: global error... |
1167 |
ASSERT(error != -ENOSPC); |
4906e2154 xfs: remove the f... |
1168 |
goto out_trans_cancel; |
c24b5dfad xfs: kill xfs_vno... |
1169 1170 1171 1172 1173 1174 1175 |
} xfs_trans_ichgtime(tp, dp, XFS_ICHGTIME_MOD | XFS_ICHGTIME_CHG); xfs_trans_log_inode(tp, dp, XFS_ILOG_CORE); if (is_dir) { error = xfs_dir_init(tp, ip, dp); if (error) |
c8eac49ef xfs: remove all b... |
1176 |
goto out_trans_cancel; |
c24b5dfad xfs: kill xfs_vno... |
1177 |
|
910832697 xfs: change some ... |
1178 |
xfs_bumplink(tp, dp); |
c24b5dfad xfs: kill xfs_vno... |
1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 |
} /* * If this is a synchronous mount, make sure that the * create transaction goes to disk before returning to * the user. */ if (mp->m_flags & (XFS_MOUNT_WSYNC|XFS_MOUNT_DIRSYNC)) xfs_trans_set_sync(tp); /* * Attach the dquot(s) to the inodes and modify them incore. * These ids of the inode couldn't have changed since the new * inode has been locked ever since it was created. */ xfs_qm_vop_create_dqattach(tp, ip, udqp, gdqp, pdqp); |
70393313d xfs: saner xfs_tr... |
1195 |
error = xfs_trans_commit(tp); |
c24b5dfad xfs: kill xfs_vno... |
1196 1197 1198 1199 1200 1201 1202 1203 1204 |
if (error) goto out_release_inode; xfs_qm_dqrele(udqp); xfs_qm_dqrele(gdqp); xfs_qm_dqrele(pdqp); *ipp = ip; return 0; |
c24b5dfad xfs: kill xfs_vno... |
1205 |
out_trans_cancel: |
4906e2154 xfs: remove the f... |
1206 |
xfs_trans_cancel(tp); |
c24b5dfad xfs: kill xfs_vno... |
1207 1208 |
out_release_inode: /* |
58c904734 xfs: inodes are n... |
1209 1210 1211 |
* Wait until after the current transaction is aborted to finish the * setup of the inode and release the inode. This prevents recursive * transactions and deadlocks from xfs_inactive. |
c24b5dfad xfs: kill xfs_vno... |
1212 |
*/ |
58c904734 xfs: inodes are n... |
1213 1214 |
if (ip) { xfs_finish_inode_setup(ip); |
44a8736bd xfs: clean up IRE... |
1215 |
xfs_irele(ip); |
58c904734 xfs: inodes are n... |
1216 |
} |
c24b5dfad xfs: kill xfs_vno... |
1217 1218 1219 1220 1221 1222 |
xfs_qm_dqrele(udqp); xfs_qm_dqrele(gdqp); xfs_qm_dqrele(pdqp); if (unlock_dp_on_error) |
655232183 xfs: remove i_iol... |
1223 |
xfs_iunlock(dp, XFS_ILOCK_EXCL); |
c24b5dfad xfs: kill xfs_vno... |
1224 1225 1226 1227 |
return error; } int |
99b6436bc xfs: add O_TMPFIL... |
1228 1229 |
xfs_create_tmpfile( struct xfs_inode *dp, |
330033d69 xfs: fix tmpfile/... |
1230 1231 |
umode_t mode, struct xfs_inode **ipp) |
99b6436bc xfs: add O_TMPFIL... |
1232 1233 1234 1235 1236 |
{ struct xfs_mount *mp = dp->i_mount; struct xfs_inode *ip = NULL; struct xfs_trans *tp = NULL; int error; |
99b6436bc xfs: add O_TMPFIL... |
1237 1238 1239 1240 1241 1242 1243 1244 |
prid_t prid; struct xfs_dquot *udqp = NULL; struct xfs_dquot *gdqp = NULL; struct xfs_dquot *pdqp = NULL; struct xfs_trans_res *tres; uint resblks; if (XFS_FORCED_SHUTDOWN(mp)) |
2451337dd xfs: global error... |
1245 |
return -EIO; |
99b6436bc xfs: add O_TMPFIL... |
1246 1247 1248 1249 1250 1251 |
prid = xfs_get_initial_prid(dp); /* * Make sure that we have allocated dquot(s) on disk. */ |
542951592 xfs: remove the i... |
1252 |
error = xfs_qm_vop_dqalloc(dp, current_fsuid(), current_fsgid(), prid, |
99b6436bc xfs: add O_TMPFIL... |
1253 1254 1255 1256 1257 1258 |
XFS_QMOPT_QUOTALL | XFS_QMOPT_INHERIT, &udqp, &gdqp, &pdqp); if (error) return error; resblks = XFS_IALLOC_SPACE_RES(mp); |
99b6436bc xfs: add O_TMPFIL... |
1259 |
tres = &M_RES(mp)->tr_create_tmpfile; |
253f4911f xfs: better xfs_t... |
1260 1261 |
error = xfs_trans_alloc(mp, tres, resblks, 0, 0, &tp); |
4906e2154 xfs: remove the f... |
1262 |
if (error) |
253f4911f xfs: better xfs_t... |
1263 |
goto out_release_inode; |
99b6436bc xfs: add O_TMPFIL... |
1264 1265 1266 1267 1268 |
error = xfs_trans_reserve_quota(tp, mp, udqp, gdqp, pdqp, resblks, 1, 0); if (error) goto out_trans_cancel; |
c4a6bf7f6 xfs: don't ever p... |
1269 |
error = xfs_dir_ialloc(&tp, dp, mode, 0, 0, prid, &ip); |
d6077aa33 xfs: Remove dupli... |
1270 |
if (error) |
4906e2154 xfs: remove the f... |
1271 |
goto out_trans_cancel; |
99b6436bc xfs: add O_TMPFIL... |
1272 1273 1274 1275 1276 1277 1278 1279 1280 1281 |
if (mp->m_flags & XFS_MOUNT_WSYNC) xfs_trans_set_sync(tp); /* * Attach the dquot(s) to the inodes and modify them incore. * These ids of the inode couldn't have changed since the new * inode has been locked ever since it was created. */ xfs_qm_vop_create_dqattach(tp, ip, udqp, gdqp, pdqp); |
99b6436bc xfs: add O_TMPFIL... |
1282 1283 |
error = xfs_iunlink(tp, ip); if (error) |
4906e2154 xfs: remove the f... |
1284 |
goto out_trans_cancel; |
99b6436bc xfs: add O_TMPFIL... |
1285 |
|
70393313d xfs: saner xfs_tr... |
1286 |
error = xfs_trans_commit(tp); |
99b6436bc xfs: add O_TMPFIL... |
1287 1288 1289 1290 1291 1292 |
if (error) goto out_release_inode; xfs_qm_dqrele(udqp); xfs_qm_dqrele(gdqp); xfs_qm_dqrele(pdqp); |
330033d69 xfs: fix tmpfile/... |
1293 |
*ipp = ip; |
99b6436bc xfs: add O_TMPFIL... |
1294 |
return 0; |
99b6436bc xfs: add O_TMPFIL... |
1295 |
out_trans_cancel: |
4906e2154 xfs: remove the f... |
1296 |
xfs_trans_cancel(tp); |
99b6436bc xfs: add O_TMPFIL... |
1297 1298 |
out_release_inode: /* |
58c904734 xfs: inodes are n... |
1299 1300 1301 |
* Wait until after the current transaction is aborted to finish the * setup of the inode and release the inode. This prevents recursive * transactions and deadlocks from xfs_inactive. |
99b6436bc xfs: add O_TMPFIL... |
1302 |
*/ |
58c904734 xfs: inodes are n... |
1303 1304 |
if (ip) { xfs_finish_inode_setup(ip); |
44a8736bd xfs: clean up IRE... |
1305 |
xfs_irele(ip); |
58c904734 xfs: inodes are n... |
1306 |
} |
99b6436bc xfs: add O_TMPFIL... |
1307 1308 1309 1310 1311 1312 1313 1314 1315 |
xfs_qm_dqrele(udqp); xfs_qm_dqrele(gdqp); xfs_qm_dqrele(pdqp); return error; } int |
c24b5dfad xfs: kill xfs_vno... |
1316 1317 1318 1319 1320 1321 1322 1323 |
xfs_link( xfs_inode_t *tdp, xfs_inode_t *sip, struct xfs_name *target_name) { xfs_mount_t *mp = tdp->i_mount; xfs_trans_t *tp; int error; |
c24b5dfad xfs: kill xfs_vno... |
1324 1325 1326 |
int resblks; trace_xfs_link(tdp, target_name); |
c19b3b05a xfs: mode di_mode... |
1327 |
ASSERT(!S_ISDIR(VFS_I(sip)->i_mode)); |
c24b5dfad xfs: kill xfs_vno... |
1328 1329 |
if (XFS_FORCED_SHUTDOWN(mp)) |
2451337dd xfs: global error... |
1330 |
return -EIO; |
c24b5dfad xfs: kill xfs_vno... |
1331 |
|
c14cfccab xfs: remove unnec... |
1332 |
error = xfs_qm_dqattach(sip); |
c24b5dfad xfs: kill xfs_vno... |
1333 1334 |
if (error) goto std_return; |
c14cfccab xfs: remove unnec... |
1335 |
error = xfs_qm_dqattach(tdp); |
c24b5dfad xfs: kill xfs_vno... |
1336 1337 |
if (error) goto std_return; |
c24b5dfad xfs: kill xfs_vno... |
1338 |
resblks = XFS_LINK_SPACE_RES(mp, target_name->len); |
253f4911f xfs: better xfs_t... |
1339 |
error = xfs_trans_alloc(mp, &M_RES(mp)->tr_link, resblks, 0, 0, &tp); |
2451337dd xfs: global error... |
1340 |
if (error == -ENOSPC) { |
c24b5dfad xfs: kill xfs_vno... |
1341 |
resblks = 0; |
253f4911f xfs: better xfs_t... |
1342 |
error = xfs_trans_alloc(mp, &M_RES(mp)->tr_link, 0, 0, 0, &tp); |
c24b5dfad xfs: kill xfs_vno... |
1343 |
} |
4906e2154 xfs: remove the f... |
1344 |
if (error) |
253f4911f xfs: better xfs_t... |
1345 |
goto std_return; |
c24b5dfad xfs: kill xfs_vno... |
1346 |
|
7c2d238ac xfs: allow xfs_lo... |
1347 |
xfs_lock_two_inodes(sip, XFS_ILOCK_EXCL, tdp, XFS_ILOCK_EXCL); |
c24b5dfad xfs: kill xfs_vno... |
1348 1349 |
xfs_trans_ijoin(tp, sip, XFS_ILOCK_EXCL); |
655232183 xfs: remove i_iol... |
1350 |
xfs_trans_ijoin(tp, tdp, XFS_ILOCK_EXCL); |
c24b5dfad xfs: kill xfs_vno... |
1351 1352 1353 1354 1355 1356 1357 |
/* * If we are using project inheritance, we only allow hard link * creation in our tree when the project IDs are the same; else * the tree quota mechanism could be circumvented. */ if (unlikely((tdp->i_d.di_flags & XFS_DIFLAG_PROJINHERIT) && |
de7a866fd xfs: merge the pr... |
1358 |
tdp->i_d.di_projid != sip->i_d.di_projid)) { |
2451337dd xfs: global error... |
1359 |
error = -EXDEV; |
c24b5dfad xfs: kill xfs_vno... |
1360 1361 |
goto error_return; } |
94f3cad55 xfs: check resblk... |
1362 1363 1364 1365 1366 |
if (!resblks) { error = xfs_dir_canenter(tp, tdp, target_name); if (error) goto error_return; } |
c24b5dfad xfs: kill xfs_vno... |
1367 |
|
54d7b5c1d xfs: use vfs inod... |
1368 1369 1370 1371 |
/* * Handle initial link state of O_TMPFILE inode */ if (VFS_I(sip)->i_nlink == 0) { |
ab2974311 xfs: allow linkat... |
1372 1373 |
error = xfs_iunlink_remove(tp, sip); if (error) |
4906e2154 xfs: remove the f... |
1374 |
goto error_return; |
ab2974311 xfs: allow linkat... |
1375 |
} |
c24b5dfad xfs: kill xfs_vno... |
1376 |
error = xfs_dir_createname(tp, tdp, target_name, sip->i_ino, |
381eee69f xfs: remove first... |
1377 |
resblks); |
c24b5dfad xfs: kill xfs_vno... |
1378 |
if (error) |
4906e2154 xfs: remove the f... |
1379 |
goto error_return; |
c24b5dfad xfs: kill xfs_vno... |
1380 1381 |
xfs_trans_ichgtime(tp, tdp, XFS_ICHGTIME_MOD | XFS_ICHGTIME_CHG); xfs_trans_log_inode(tp, tdp, XFS_ILOG_CORE); |
910832697 xfs: change some ... |
1382 |
xfs_bumplink(tp, sip); |
c24b5dfad xfs: kill xfs_vno... |
1383 1384 1385 1386 1387 1388 |
/* * If this is a synchronous mount, make sure that the * link transaction goes to disk before returning to * the user. */ |
f6106efae xfs: eliminate co... |
1389 |
if (mp->m_flags & (XFS_MOUNT_WSYNC|XFS_MOUNT_DIRSYNC)) |
c24b5dfad xfs: kill xfs_vno... |
1390 |
xfs_trans_set_sync(tp); |
c24b5dfad xfs: kill xfs_vno... |
1391 |
|
70393313d xfs: saner xfs_tr... |
1392 |
return xfs_trans_commit(tp); |
c24b5dfad xfs: kill xfs_vno... |
1393 |
|
c24b5dfad xfs: kill xfs_vno... |
1394 |
error_return: |
4906e2154 xfs: remove the f... |
1395 |
xfs_trans_cancel(tp); |
c24b5dfad xfs: kill xfs_vno... |
1396 1397 1398 |
std_return: return error; } |
363e59baa xfs: don't be so ... |
1399 1400 1401 1402 1403 1404 1405 1406 1407 1408 1409 1410 1411 1412 1413 1414 1415 |
/* Clear the reflink flag and the cowblocks tag if possible. */ static void xfs_itruncate_clear_reflink_flags( struct xfs_inode *ip) { struct xfs_ifork *dfork; struct xfs_ifork *cfork; if (!xfs_is_reflink_inode(ip)) return; dfork = XFS_IFORK_PTR(ip, XFS_DATA_FORK); cfork = XFS_IFORK_PTR(ip, XFS_COW_FORK); if (dfork->if_bytes == 0 && cfork->if_bytes == 0) ip->i_d.di_flags2 &= ~XFS_DIFLAG2_REFLINK; if (cfork->if_bytes == 0) xfs_inode_clear_cowblocks_tag(ip); } |
1da177e4c Linux-2.6.12-rc2 |
1416 |
/* |
8f04c47aa xfs: split xfs_it... |
1417 1418 1419 |
* Free up the underlying blocks past new_size. The new size must be smaller * than the current size. This routine can be used both for the attribute and * data fork, and does not modify the inode size, which is left to the caller. |
1da177e4c Linux-2.6.12-rc2 |
1420 |
* |
f6485057c [XFS] Ensure the ... |
1421 1422 1423 1424 1425 1426 1427 1428 1429 |
* The transaction passed to this routine must have made a permanent log * reservation of at least XFS_ITRUNCATE_LOG_RES. This routine may commit the * given transaction and start new ones, so make sure everything involved in * the transaction is tidy before calling here. Some transaction will be * returned to the caller to be committed. The incoming transaction must * already include the inode, and both inode locks must be held exclusively. * The inode must also be "held" within the transaction. On return the inode * will be "held" within the returned transaction. This routine does NOT * require any disk space to be reserved for it within the transaction. |
1da177e4c Linux-2.6.12-rc2 |
1430 |
* |
f6485057c [XFS] Ensure the ... |
1431 1432 1433 1434 1435 |
* If we get an error, we must return with the inode locked and linked into the * current transaction. This keeps things simple for the higher level code, * because it always knows that the inode is locked and held in the transaction * that returns to it whether errors occur or not. We don't mark the inode * dirty on error so that transactions can be easily aborted if possible. |
1da177e4c Linux-2.6.12-rc2 |
1436 1437 |
*/ int |
4e529339a xfs: factor out n... |
1438 |
xfs_itruncate_extents_flags( |
8f04c47aa xfs: split xfs_it... |
1439 1440 1441 |
struct xfs_trans **tpp, struct xfs_inode *ip, int whichfork, |
13b86fc33 xfs: skip online ... |
1442 |
xfs_fsize_t new_size, |
4e529339a xfs: factor out n... |
1443 |
int flags) |
1da177e4c Linux-2.6.12-rc2 |
1444 |
{ |
8f04c47aa xfs: split xfs_it... |
1445 1446 |
struct xfs_mount *mp = ip->i_mount; struct xfs_trans *tp = *tpp; |
8f04c47aa xfs: split xfs_it... |
1447 |
xfs_fileoff_t first_unmap_block; |
8f04c47aa xfs: split xfs_it... |
1448 |
xfs_filblks_t unmap_len; |
8f04c47aa xfs: split xfs_it... |
1449 |
int error = 0; |
1da177e4c Linux-2.6.12-rc2 |
1450 |
|
0b56185b0 xfs: do not take ... |
1451 1452 1453 |
ASSERT(xfs_isilocked(ip, XFS_ILOCK_EXCL)); ASSERT(!atomic_read(&VFS_I(ip)->i_count) || xfs_isilocked(ip, XFS_IOLOCK_EXCL)); |
ce7ae151d xfs: remove the i... |
1454 |
ASSERT(new_size <= XFS_ISIZE(ip)); |
8f04c47aa xfs: split xfs_it... |
1455 |
ASSERT(tp->t_flags & XFS_TRANS_PERM_LOG_RES); |
1da177e4c Linux-2.6.12-rc2 |
1456 |
ASSERT(ip->i_itemp != NULL); |
898621d5a xfs: simplify ino... |
1457 |
ASSERT(ip->i_itemp->ili_lock_flags == 0); |
8f04c47aa xfs: split xfs_it... |
1458 |
ASSERT(!XFS_NOT_DQATTACHED(mp, ip)); |
1da177e4c Linux-2.6.12-rc2 |
1459 |
|
673e8e597 xfs: remove xfs_i... |
1460 |
trace_xfs_itruncate_extents_start(ip, new_size); |
4e529339a xfs: factor out n... |
1461 |
flags |= xfs_bmapi_aflag(whichfork); |
13b86fc33 xfs: skip online ... |
1462 |
|
1da177e4c Linux-2.6.12-rc2 |
1463 1464 1465 1466 1467 |
/* * Since it is possible for space to become allocated beyond * the end of the file (in a crash where the space is allocated * but the inode size is not yet updated), simply remove any * blocks which show up between the new EOF and the maximum |
4bbb04abb xfs: truncate sho... |
1468 1469 1470 1471 |
* possible file size. * * We have to free all the blocks to the bmbt maximum offset, even if * the page cache can't scale that far. |
1da177e4c Linux-2.6.12-rc2 |
1472 |
*/ |
8f04c47aa xfs: split xfs_it... |
1473 |
first_unmap_block = XFS_B_TO_FSB(mp, (xfs_ufsize_t)new_size); |
4bbb04abb xfs: truncate sho... |
1474 1475 |
if (first_unmap_block >= XFS_MAX_FILEOFF) { WARN_ON_ONCE(first_unmap_block > XFS_MAX_FILEOFF); |
8f04c47aa xfs: split xfs_it... |
1476 |
return 0; |
4bbb04abb xfs: truncate sho... |
1477 |
} |
8f04c47aa xfs: split xfs_it... |
1478 |
|
4bbb04abb xfs: truncate sho... |
1479 1480 |
unmap_len = XFS_MAX_FILEOFF - first_unmap_block + 1; while (unmap_len > 0) { |
02dff7bf8 xfs: pull up dfop... |
1481 |
ASSERT(tp->t_firstblock == NULLFSBLOCK); |
4bbb04abb xfs: truncate sho... |
1482 1483 |
error = __xfs_bunmapi(tp, ip, first_unmap_block, &unmap_len, flags, XFS_ITRUNC_MAX_EXTENTS); |
8f04c47aa xfs: split xfs_it... |
1484 |
if (error) |
d5a2e2893 xfs: remove last ... |
1485 |
goto out; |
1da177e4c Linux-2.6.12-rc2 |
1486 |
|
6dd379c7f xfs: drop extra t... |
1487 |
/* free the just unmapped extents */ |
9e28a242b xfs: drop unneces... |
1488 |
error = xfs_defer_finish(&tp); |
8f04c47aa xfs: split xfs_it... |
1489 |
if (error) |
9b1f4e983 xfs: cancel dfops... |
1490 |
goto out; |
1da177e4c Linux-2.6.12-rc2 |
1491 |
} |
8f04c47aa xfs: split xfs_it... |
1492 |
|
4919d42ab xfs: only cancel ... |
1493 1494 1495 |
if (whichfork == XFS_DATA_FORK) { /* Remove all pending CoW reservations. */ error = xfs_reflink_cancel_cow_blocks(ip, &tp, |
4bbb04abb xfs: truncate sho... |
1496 |
first_unmap_block, XFS_MAX_FILEOFF, true); |
4919d42ab xfs: only cancel ... |
1497 1498 |
if (error) goto out; |
aa8968f22 xfs: cancel CoW r... |
1499 |
|
4919d42ab xfs: only cancel ... |
1500 1501 |
xfs_itruncate_clear_reflink_flags(ip); } |
aa8968f22 xfs: cancel CoW r... |
1502 |
|
673e8e597 xfs: remove xfs_i... |
1503 1504 1505 1506 1507 1508 1509 |
/* * Always re-log the inode so that our permanent transaction can keep * on rolling it forward in the log. */ xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE); trace_xfs_itruncate_extents_end(ip, new_size); |
8f04c47aa xfs: split xfs_it... |
1510 1511 1512 |
out: *tpp = tp; return error; |
8f04c47aa xfs: split xfs_it... |
1513 |
} |
c24b5dfad xfs: kill xfs_vno... |
1514 1515 1516 1517 1518 1519 |
int xfs_release( xfs_inode_t *ip) { xfs_mount_t *mp = ip->i_mount; int error; |
c19b3b05a xfs: mode di_mode... |
1520 |
if (!S_ISREG(VFS_I(ip)->i_mode) || (VFS_I(ip)->i_mode == 0)) |
c24b5dfad xfs: kill xfs_vno... |
1521 1522 1523 1524 1525 1526 1527 1528 1529 1530 |
return 0; /* If this is a read-only mount, don't do this (would generate I/O) */ if (mp->m_flags & XFS_MOUNT_RDONLY) return 0; if (!XFS_FORCED_SHUTDOWN(mp)) { int truncated; /* |
c24b5dfad xfs: kill xfs_vno... |
1531 1532 1533 1534 1535 1536 1537 1538 1539 1540 1541 1542 |
* If we previously truncated this file and removed old data * in the process, we want to initiate "early" writeout on * the last close. This is an attempt to combat the notorious * NULL files problem which is particularly noticeable from a * truncate down, buffered (re-)write (delalloc), followed by * a crash. What we are effectively doing here is * significantly reducing the time window where we'd otherwise * be exposed to that problem. */ truncated = xfs_iflags_test_and_clear(ip, XFS_ITRUNCATED); if (truncated) { xfs_iflags_clear(ip, XFS_IDIRTY_RELEASE); |
eac152b47 xfs: kill VN_DIRTY() |
1543 |
if (ip->i_delayed_blks > 0) { |
2451337dd xfs: global error... |
1544 |
error = filemap_flush(VFS_I(ip)->i_mapping); |
c24b5dfad xfs: kill xfs_vno... |
1545 1546 1547 1548 1549 |
if (error) return error; } } } |
54d7b5c1d xfs: use vfs inod... |
1550 |
if (VFS_I(ip)->i_nlink == 0) |
c24b5dfad xfs: kill xfs_vno... |
1551 1552 1553 1554 1555 |
return 0; if (xfs_can_free_eofblocks(ip, false)) { /* |
a36b92618 xfs: pull up iolo... |
1556 1557 1558 1559 1560 1561 1562 1563 1564 1565 1566 1567 1568 1569 1570 1571 |
* Check if the inode is being opened, written and closed * frequently and we have delayed allocation blocks outstanding * (e.g. streaming writes from the NFS server), truncating the * blocks past EOF will cause fragmentation to occur. * * In this case don't do the truncation, but we have to be * careful how we detect this case. Blocks beyond EOF show up as * i_delayed_blks even when the inode is clean, so we need to * truncate them away first before checking for a dirty release. * Hence on the first dirty close we will still remove the * speculative allocation, but after that we will leave it in * place. */ if (xfs_iflags_test(ip, XFS_IDIRTY_RELEASE)) return 0; /* |
c24b5dfad xfs: kill xfs_vno... |
1572 |
* If we can't get the iolock just skip truncating the blocks |
c1e8d7c6a mmap locking API:... |
1573 |
* past EOF because we could deadlock with the mmap_lock |
a36b92618 xfs: pull up iolo... |
1574 |
* otherwise. We'll get another chance to drop them once the |
c24b5dfad xfs: kill xfs_vno... |
1575 1576 |
* last reference to the inode is dropped, so we'll never leak * blocks permanently. |
c24b5dfad xfs: kill xfs_vno... |
1577 |
*/ |
a36b92618 xfs: pull up iolo... |
1578 1579 1580 1581 1582 1583 |
if (xfs_ilock_nowait(ip, XFS_IOLOCK_EXCL)) { error = xfs_free_eofblocks(ip); xfs_iunlock(ip, XFS_IOLOCK_EXCL); if (error) return error; } |
c24b5dfad xfs: kill xfs_vno... |
1584 1585 1586 1587 1588 1589 1590 1591 1592 |
/* delalloc blocks after truncation means it really is dirty */ if (ip->i_delayed_blks) xfs_iflags_set(ip, XFS_IDIRTY_RELEASE); } return 0; } /* |
f7be2d7f5 xfs: push down in... |
1593 1594 1595 1596 1597 1598 1599 1600 1601 1602 1603 |
* xfs_inactive_truncate * * Called to perform a truncate when an inode becomes unlinked. */ STATIC int xfs_inactive_truncate( struct xfs_inode *ip) { struct xfs_mount *mp = ip->i_mount; struct xfs_trans *tp; int error; |
253f4911f xfs: better xfs_t... |
1604 |
error = xfs_trans_alloc(mp, &M_RES(mp)->tr_itruncate, 0, 0, 0, &tp); |
f7be2d7f5 xfs: push down in... |
1605 1606 |
if (error) { ASSERT(XFS_FORCED_SHUTDOWN(mp)); |
f7be2d7f5 xfs: push down in... |
1607 1608 |
return error; } |
f7be2d7f5 xfs: push down in... |
1609 1610 1611 1612 1613 1614 |
xfs_ilock(ip, XFS_ILOCK_EXCL); xfs_trans_ijoin(tp, ip, 0); /* * Log the inode size first to prevent stale data exposure in the event * of a system crash before the truncate completes. See the related |
69bca8074 xfs: Propagate de... |
1615 |
* comment in xfs_vn_setattr_size() for details. |
f7be2d7f5 xfs: push down in... |
1616 1617 1618 1619 1620 1621 1622 |
*/ ip->i_d.di_size = 0; xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE); error = xfs_itruncate_extents(&tp, ip, XFS_DATA_FORK, 0); if (error) goto error_trans_cancel; |
daf83964a xfs: move the per... |
1623 |
ASSERT(ip->i_df.if_nextents == 0); |
f7be2d7f5 xfs: push down in... |
1624 |
|
70393313d xfs: saner xfs_tr... |
1625 |
error = xfs_trans_commit(tp); |
f7be2d7f5 xfs: push down in... |
1626 1627 1628 1629 1630 1631 1632 |
if (error) goto error_unlock; xfs_iunlock(ip, XFS_ILOCK_EXCL); return 0; error_trans_cancel: |
4906e2154 xfs: remove the f... |
1633 |
xfs_trans_cancel(tp); |
f7be2d7f5 xfs: push down in... |
1634 1635 1636 1637 1638 1639 |
error_unlock: xfs_iunlock(ip, XFS_ILOCK_EXCL); return error; } /* |
88877d2b9 xfs: push down in... |
1640 1641 1642 1643 1644 1645 1646 1647 |
* xfs_inactive_ifree() * * Perform the inode free when an inode is unlinked. */ STATIC int xfs_inactive_ifree( struct xfs_inode *ip) { |
88877d2b9 xfs: push down in... |
1648 1649 1650 |
struct xfs_mount *mp = ip->i_mount; struct xfs_trans *tp; int error; |
9d43b180a xfs: update inode... |
1651 |
/* |
76d771b4c xfs: use per-AG r... |
1652 1653 1654 1655 1656 |
* We try to use a per-AG reservation for any block needed by the finobt * tree, but as the finobt feature predates the per-AG reservation * support a degraded file system might not have enough space for the * reservation at mount time. In that case try to dip into the reserved * pool and pray. |
9d43b180a xfs: update inode... |
1657 1658 1659 1660 1661 |
* * Send a warning if the reservation does happen to fail, as the inode * now remains allocated and sits on the unlinked list until the fs is * repaired. */ |
e1f6ca113 xfs: rename m_ino... |
1662 |
if (unlikely(mp->m_finobt_nores)) { |
76d771b4c xfs: use per-AG r... |
1663 1664 1665 1666 1667 1668 |
error = xfs_trans_alloc(mp, &M_RES(mp)->tr_ifree, XFS_IFREE_SPACE_RES(mp), 0, XFS_TRANS_RESERVE, &tp); } else { error = xfs_trans_alloc(mp, &M_RES(mp)->tr_ifree, 0, 0, 0, &tp); } |
88877d2b9 xfs: push down in... |
1669 |
if (error) { |
2451337dd xfs: global error... |
1670 |
if (error == -ENOSPC) { |
9d43b180a xfs: update inode... |
1671 1672 1673 1674 1675 1676 |
xfs_warn_ratelimited(mp, "Failed to remove inode(s) from unlinked list. " "Please free space, unmount and run xfs_repair."); } else { ASSERT(XFS_FORCED_SHUTDOWN(mp)); } |
88877d2b9 xfs: push down in... |
1677 1678 |
return error; } |
96355d5a1 xfs: Don't allow ... |
1679 1680 1681 1682 1683 1684 1685 1686 1687 1688 1689 1690 1691 1692 1693 1694 1695 1696 1697 1698 |
/* * We do not hold the inode locked across the entire rolling transaction * here. We only need to hold it for the first transaction that * xfs_ifree() builds, which may mark the inode XFS_ISTALE if the * underlying cluster buffer is freed. Relogging an XFS_ISTALE inode * here breaks the relationship between cluster buffer invalidation and * stale inode invalidation on cluster buffer item journal commit * completion, and can result in leaving dirty stale inodes hanging * around in memory. * * We have no need for serialising this inode operation against other * operations - we freed the inode and hence reallocation is required * and that will serialise on reallocating the space the deferops need * to free. Hence we can unlock the inode on the first commit of * the transaction rather than roll it right through the deferops. This * avoids relogging the XFS_ISTALE inode. * * We check that xfs_ifree() hasn't grown an internal transaction roll * by asserting that the inode is still locked when it returns. */ |
88877d2b9 xfs: push down in... |
1699 |
xfs_ilock(ip, XFS_ILOCK_EXCL); |
96355d5a1 xfs: Don't allow ... |
1700 |
xfs_trans_ijoin(tp, ip, XFS_ILOCK_EXCL); |
88877d2b9 xfs: push down in... |
1701 |
|
0e0417f3e xfs: remove dfops... |
1702 |
error = xfs_ifree(tp, ip); |
96355d5a1 xfs: Don't allow ... |
1703 |
ASSERT(xfs_isilocked(ip, XFS_ILOCK_EXCL)); |
88877d2b9 xfs: push down in... |
1704 1705 1706 1707 1708 1709 1710 1711 1712 1713 1714 |
if (error) { /* * If we fail to free the inode, shut down. The cancel * might do that, we need to make sure. Otherwise the * inode might be lost for a long time or forever. */ if (!XFS_FORCED_SHUTDOWN(mp)) { xfs_notice(mp, "%s: xfs_ifree returned error %d", __func__, error); xfs_force_shutdown(mp, SHUTDOWN_META_IO_ERROR); } |
4906e2154 xfs: remove the f... |
1715 |
xfs_trans_cancel(tp); |
88877d2b9 xfs: push down in... |
1716 1717 1718 1719 1720 1721 1722 1723 1724 |
return error; } /* * Credit the quota account(s). The inode is gone. */ xfs_trans_mod_dquot_byino(tp, ip, XFS_TRANS_DQ_ICOUNT, -1); /* |
d4a97a042 xfs: add missing ... |
1725 1726 |
* Just ignore errors at this point. There is nothing we can do except * to try to keep going. Make sure it's not a silent error. |
88877d2b9 xfs: push down in... |
1727 |
*/ |
70393313d xfs: saner xfs_tr... |
1728 |
error = xfs_trans_commit(tp); |
88877d2b9 xfs: push down in... |
1729 1730 1731 |
if (error) xfs_notice(mp, "%s: xfs_trans_commit returned error %d", __func__, error); |
88877d2b9 xfs: push down in... |
1732 1733 1734 1735 |
return 0; } /* |
c24b5dfad xfs: kill xfs_vno... |
1736 1737 1738 1739 1740 1741 1742 |
* xfs_inactive * * This is called when the vnode reference count for the vnode * goes to zero. If the file has been unlinked, then it must * now be truncated. Also, we clear all of the read-ahead state * kept for the inode here since the file is now closed. */ |
74564fb48 xfs: clean up xfs... |
1743 |
void |
c24b5dfad xfs: kill xfs_vno... |
1744 1745 1746 |
xfs_inactive( xfs_inode_t *ip) { |
3d3c8b522 xfs: refactor xfs... |
1747 |
struct xfs_mount *mp; |
3d3c8b522 xfs: refactor xfs... |
1748 1749 |
int error; int truncate = 0; |
c24b5dfad xfs: kill xfs_vno... |
1750 1751 1752 1753 1754 |
/* * If the inode is already free, then there can be nothing * to clean up here. */ |
c19b3b05a xfs: mode di_mode... |
1755 |
if (VFS_I(ip)->i_mode == 0) { |
c24b5dfad xfs: kill xfs_vno... |
1756 |
ASSERT(ip->i_df.if_broot_bytes == 0); |
74564fb48 xfs: clean up xfs... |
1757 |
return; |
c24b5dfad xfs: kill xfs_vno... |
1758 1759 1760 |
} mp = ip->i_mount; |
17c12bcd3 xfs: when replayi... |
1761 |
ASSERT(!xfs_iflags_test(ip, XFS_IRECOVERY)); |
c24b5dfad xfs: kill xfs_vno... |
1762 |
|
c24b5dfad xfs: kill xfs_vno... |
1763 1764 |
/* If this is a read-only mount, don't do this (would generate I/O) */ if (mp->m_flags & XFS_MOUNT_RDONLY) |
74564fb48 xfs: clean up xfs... |
1765 |
return; |
c24b5dfad xfs: kill xfs_vno... |
1766 |
|
6231848c3 xfs: check for co... |
1767 |
/* Try to clean out the cow blocks if there are any. */ |
51d626903 xfs: introduce a ... |
1768 |
if (xfs_inode_has_cow_data(ip)) |
6231848c3 xfs: check for co... |
1769 |
xfs_reflink_cancel_cow_range(ip, 0, NULLFILEOFF, true); |
54d7b5c1d xfs: use vfs inod... |
1770 |
if (VFS_I(ip)->i_nlink != 0) { |
c24b5dfad xfs: kill xfs_vno... |
1771 1772 1773 1774 |
/* * force is true because we are evicting an inode from the * cache. Post-eof blocks must be freed, lest we end up with * broken free space accounting. |
3b4683c29 xfs: drop iolock ... |
1775 1776 1777 1778 |
* * Note: don't bother with iolock here since lockdep complains * about acquiring it in reclaim context. We have the only * reference to the inode at this point anyways. |
c24b5dfad xfs: kill xfs_vno... |
1779 |
*/ |
3b4683c29 xfs: drop iolock ... |
1780 |
if (xfs_can_free_eofblocks(ip, true)) |
a36b92618 xfs: pull up iolo... |
1781 |
xfs_free_eofblocks(ip); |
74564fb48 xfs: clean up xfs... |
1782 1783 |
return; |
c24b5dfad xfs: kill xfs_vno... |
1784 |
} |
c19b3b05a xfs: mode di_mode... |
1785 |
if (S_ISREG(VFS_I(ip)->i_mode) && |
c24b5dfad xfs: kill xfs_vno... |
1786 |
(ip->i_d.di_size != 0 || XFS_ISIZE(ip) != 0 || |
daf83964a xfs: move the per... |
1787 |
ip->i_df.if_nextents > 0 || ip->i_delayed_blks > 0)) |
c24b5dfad xfs: kill xfs_vno... |
1788 |
truncate = 1; |
c14cfccab xfs: remove unnec... |
1789 |
error = xfs_qm_dqattach(ip); |
c24b5dfad xfs: kill xfs_vno... |
1790 |
if (error) |
74564fb48 xfs: clean up xfs... |
1791 |
return; |
c24b5dfad xfs: kill xfs_vno... |
1792 |
|
c19b3b05a xfs: mode di_mode... |
1793 |
if (S_ISLNK(VFS_I(ip)->i_mode)) |
36b21dde6 xfs: push down in... |
1794 |
error = xfs_inactive_symlink(ip); |
f7be2d7f5 xfs: push down in... |
1795 1796 1797 |
else if (truncate) error = xfs_inactive_truncate(ip); if (error) |
74564fb48 xfs: clean up xfs... |
1798 |
return; |
c24b5dfad xfs: kill xfs_vno... |
1799 1800 1801 1802 |
/* * If there are attributes associated with the file then blow them away * now. The code calls a routine that recursively deconstructs the |
6dfe5a049 xfs: xfs_attr_ina... |
1803 |
* attribute fork. If also blows away the in-core attribute fork. |
c24b5dfad xfs: kill xfs_vno... |
1804 |
*/ |
6dfe5a049 xfs: xfs_attr_ina... |
1805 |
if (XFS_IFORK_Q(ip)) { |
c24b5dfad xfs: kill xfs_vno... |
1806 1807 |
error = xfs_attr_inactive(ip); if (error) |
74564fb48 xfs: clean up xfs... |
1808 |
return; |
c24b5dfad xfs: kill xfs_vno... |
1809 |
} |
6dfe5a049 xfs: xfs_attr_ina... |
1810 |
ASSERT(!ip->i_afp); |
6dfe5a049 xfs: xfs_attr_ina... |
1811 |
ASSERT(ip->i_d.di_forkoff == 0); |
c24b5dfad xfs: kill xfs_vno... |
1812 1813 1814 1815 |
/* * Free the inode. */ |
88877d2b9 xfs: push down in... |
1816 1817 |
error = xfs_inactive_ifree(ip); if (error) |
74564fb48 xfs: clean up xfs... |
1818 |
return; |
c24b5dfad xfs: kill xfs_vno... |
1819 1820 1821 1822 1823 |
/* * Release the dquots held by inode, if any. */ xfs_qm_dqdetach(ip); |
c24b5dfad xfs: kill xfs_vno... |
1824 |
} |
1da177e4c Linux-2.6.12-rc2 |
1825 |
/* |
9b2471797 xfs: cache unlink... |
1826 1827 1828 1829 1830 1831 1832 1833 1834 1835 1836 1837 1838 1839 1840 1841 1842 1843 1844 1845 1846 1847 1848 1849 1850 1851 1852 1853 1854 1855 1856 1857 1858 1859 1860 1861 1862 1863 1864 1865 1866 1867 1868 1869 1870 1871 1872 1873 1874 1875 1876 1877 1878 1879 1880 1881 1882 1883 1884 1885 1886 1887 1888 1889 1890 1891 1892 1893 1894 1895 1896 1897 1898 1899 1900 1901 1902 1903 1904 1905 1906 1907 1908 1909 1910 1911 1912 1913 1914 1915 1916 1917 1918 1919 1920 1921 1922 1923 1924 1925 1926 1927 1928 1929 1930 1931 1932 1933 1934 1935 1936 1937 1938 1939 1940 1941 1942 1943 1944 1945 1946 |
* In-Core Unlinked List Lookups * ============================= * * Every inode is supposed to be reachable from some other piece of metadata * with the exception of the root directory. Inodes with a connection to a * file descriptor but not linked from anywhere in the on-disk directory tree * are collectively known as unlinked inodes, though the filesystem itself * maintains links to these inodes so that on-disk metadata are consistent. * * XFS implements a per-AG on-disk hash table of unlinked inodes. The AGI * header contains a number of buckets that point to an inode, and each inode * record has a pointer to the next inode in the hash chain. This * singly-linked list causes scaling problems in the iunlink remove function * because we must walk that list to find the inode that points to the inode * being removed from the unlinked hash bucket list. * * What if we modelled the unlinked list as a collection of records capturing * "X.next_unlinked = Y" relations? If we indexed those records on Y, we'd * have a fast way to look up unlinked list predecessors, which avoids the * slow list walk. That's exactly what we do here (in-core) with a per-AG * rhashtable. * * Because this is a backref cache, we ignore operational failures since the * iunlink code can fall back to the slow bucket walk. The only errors that * should bubble out are for obviously incorrect situations. * * All users of the backref cache MUST hold the AGI buffer lock to serialize * access or have otherwise provided for concurrency control. */ /* Capture a "X.next_unlinked = Y" relationship. */ struct xfs_iunlink { struct rhash_head iu_rhash_head; xfs_agino_t iu_agino; /* X */ xfs_agino_t iu_next_unlinked; /* Y */ }; /* Unlinked list predecessor lookup hashtable construction */ static int xfs_iunlink_obj_cmpfn( struct rhashtable_compare_arg *arg, const void *obj) { const xfs_agino_t *key = arg->key; const struct xfs_iunlink *iu = obj; if (iu->iu_next_unlinked != *key) return 1; return 0; } static const struct rhashtable_params xfs_iunlink_hash_params = { .min_size = XFS_AGI_UNLINKED_BUCKETS, .key_len = sizeof(xfs_agino_t), .key_offset = offsetof(struct xfs_iunlink, iu_next_unlinked), .head_offset = offsetof(struct xfs_iunlink, iu_rhash_head), .automatic_shrinking = true, .obj_cmpfn = xfs_iunlink_obj_cmpfn, }; /* * Return X, where X.next_unlinked == @agino. Returns NULLAGINO if no such * relation is found. */ static xfs_agino_t xfs_iunlink_lookup_backref( struct xfs_perag *pag, xfs_agino_t agino) { struct xfs_iunlink *iu; iu = rhashtable_lookup_fast(&pag->pagi_unlinked_hash, &agino, xfs_iunlink_hash_params); return iu ? iu->iu_agino : NULLAGINO; } /* * Take ownership of an iunlink cache entry and insert it into the hash table. * If successful, the entry will be owned by the cache; if not, it is freed. * Either way, the caller does not own @iu after this call. */ static int xfs_iunlink_insert_backref( struct xfs_perag *pag, struct xfs_iunlink *iu) { int error; error = rhashtable_insert_fast(&pag->pagi_unlinked_hash, &iu->iu_rhash_head, xfs_iunlink_hash_params); /* * Fail loudly if there already was an entry because that's a sign of * corruption of in-memory data. Also fail loudly if we see an error * code we didn't anticipate from the rhashtable code. Currently we * only anticipate ENOMEM. */ if (error) { WARN(error != -ENOMEM, "iunlink cache insert error %d", error); kmem_free(iu); } /* * Absorb any runtime errors that aren't a result of corruption because * this is a cache and we can always fall back to bucket list scanning. */ if (error != 0 && error != -EEXIST) error = 0; return error; } /* Remember that @prev_agino.next_unlinked = @this_agino. */ static int xfs_iunlink_add_backref( struct xfs_perag *pag, xfs_agino_t prev_agino, xfs_agino_t this_agino) { struct xfs_iunlink *iu; if (XFS_TEST_ERROR(false, pag->pag_mount, XFS_ERRTAG_IUNLINK_FALLBACK)) return 0; |
707e0ddaf fs: xfs: Remove K... |
1947 |
iu = kmem_zalloc(sizeof(*iu), KM_NOFS); |
9b2471797 xfs: cache unlink... |
1948 1949 1950 1951 1952 1953 1954 1955 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 |
iu->iu_agino = prev_agino; iu->iu_next_unlinked = this_agino; return xfs_iunlink_insert_backref(pag, iu); } /* * Replace X.next_unlinked = @agino with X.next_unlinked = @next_unlinked. * If @next_unlinked is NULLAGINO, we drop the backref and exit. If there * wasn't any such entry then we don't bother. */ static int xfs_iunlink_change_backref( struct xfs_perag *pag, xfs_agino_t agino, xfs_agino_t next_unlinked) { struct xfs_iunlink *iu; int error; /* Look up the old entry; if there wasn't one then exit. */ iu = rhashtable_lookup_fast(&pag->pagi_unlinked_hash, &agino, xfs_iunlink_hash_params); if (!iu) return 0; /* * Remove the entry. This shouldn't ever return an error, but if we * couldn't remove the old entry we don't want to add it again to the * hash table, and if the entry disappeared on us then someone's * violated the locking rules and we need to fail loudly. Either way * we cannot remove the inode because internal state is or would have * been corrupt. */ error = rhashtable_remove_fast(&pag->pagi_unlinked_hash, &iu->iu_rhash_head, xfs_iunlink_hash_params); if (error) return error; /* If there is no new next entry just free our item and return. */ if (next_unlinked == NULLAGINO) { kmem_free(iu); return 0; } /* Update the entry and re-add it to the hash table. */ iu->iu_next_unlinked = next_unlinked; return xfs_iunlink_insert_backref(pag, iu); } /* Set up the in-core predecessor structures. */ int xfs_iunlink_init( struct xfs_perag *pag) { return rhashtable_init(&pag->pagi_unlinked_hash, &xfs_iunlink_hash_params); } /* Free the in-core predecessor structures. */ static void xfs_iunlink_free_item( void *ptr, void *arg) { struct xfs_iunlink *iu = ptr; bool *freed_anything = arg; *freed_anything = true; kmem_free(iu); } void xfs_iunlink_destroy( struct xfs_perag *pag) { bool freed_anything = false; rhashtable_free_and_destroy(&pag->pagi_unlinked_hash, xfs_iunlink_free_item, &freed_anything); ASSERT(freed_anything == false || XFS_FORCED_SHUTDOWN(pag->pag_mount)); } /* |
9a4a51186 xfs: refactor AGI... |
2033 2034 2035 2036 2037 2038 2039 2040 2041 2042 2043 |
* Point the AGI unlinked bucket at an inode and log the results. The caller * is responsible for validating the old value. */ STATIC int xfs_iunlink_update_bucket( struct xfs_trans *tp, xfs_agnumber_t agno, struct xfs_buf *agibp, unsigned int bucket_index, xfs_agino_t new_agino) { |
370c782b9 xfs: remove XFS_B... |
2044 |
struct xfs_agi *agi = agibp->b_addr; |
9a4a51186 xfs: refactor AGI... |
2045 2046 2047 2048 2049 2050 2051 2052 2053 2054 2055 2056 2057 2058 |
xfs_agino_t old_value; int offset; ASSERT(xfs_verify_agino_or_null(tp->t_mountp, agno, new_agino)); old_value = be32_to_cpu(agi->agi_unlinked[bucket_index]); trace_xfs_iunlink_update_bucket(tp->t_mountp, agno, bucket_index, old_value, new_agino); /* * We should never find the head of the list already set to the value * passed in because either we're adding or removing ourselves from the * head of the list. */ |
a5155b870 xfs: always log c... |
2059 |
if (old_value == new_agino) { |
8d57c2160 xfs: add a functi... |
2060 |
xfs_buf_mark_corrupt(agibp); |
9a4a51186 xfs: refactor AGI... |
2061 |
return -EFSCORRUPTED; |
a5155b870 xfs: always log c... |
2062 |
} |
9a4a51186 xfs: refactor AGI... |
2063 2064 2065 2066 2067 2068 2069 |
agi->agi_unlinked[bucket_index] = cpu_to_be32(new_agino); offset = offsetof(struct xfs_agi, agi_unlinked) + (sizeof(xfs_agino_t) * bucket_index); xfs_trans_log_buf(tp, agibp, offset, offset + sizeof(xfs_agino_t) - 1); return 0; } |
f2fc16a3d xfs: refactor ino... |
2070 2071 2072 2073 2074 2075 2076 2077 2078 2079 2080 2081 2082 2083 2084 2085 2086 2087 2088 2089 2090 2091 2092 2093 2094 2095 2096 |
/* Set an on-disk inode's next_unlinked pointer. */ STATIC void xfs_iunlink_update_dinode( struct xfs_trans *tp, xfs_agnumber_t agno, xfs_agino_t agino, struct xfs_buf *ibp, struct xfs_dinode *dip, struct xfs_imap *imap, xfs_agino_t next_agino) { struct xfs_mount *mp = tp->t_mountp; int offset; ASSERT(xfs_verify_agino_or_null(mp, agno, next_agino)); trace_xfs_iunlink_update_dinode(mp, agno, agino, be32_to_cpu(dip->di_next_unlinked), next_agino); dip->di_next_unlinked = cpu_to_be32(next_agino); offset = imap->im_boffset + offsetof(struct xfs_dinode, di_next_unlinked); /* need to recalc the inode CRC if appropriate */ xfs_dinode_calc_crc(mp, dip); xfs_trans_inode_buf(tp, ibp); xfs_trans_log_buf(tp, ibp, offset, offset + sizeof(xfs_agino_t) - 1); |
f2fc16a3d xfs: refactor ino... |
2097 2098 2099 2100 2101 2102 2103 2104 2105 2106 2107 2108 2109 2110 2111 2112 2113 2114 |
} /* Set an in-core inode's unlinked pointer and return the old value. */ STATIC int xfs_iunlink_update_inode( struct xfs_trans *tp, struct xfs_inode *ip, xfs_agnumber_t agno, xfs_agino_t next_agino, xfs_agino_t *old_next_agino) { struct xfs_mount *mp = tp->t_mountp; struct xfs_dinode *dip; struct xfs_buf *ibp; xfs_agino_t old_value; int error; ASSERT(xfs_verify_agino_or_null(mp, agno, next_agino)); |
c19950799 xfs: remove unuse... |
2115 |
error = xfs_imap_to_bp(mp, tp, &ip->i_imap, &dip, &ibp, 0); |
f2fc16a3d xfs: refactor ino... |
2116 2117 2118 2119 2120 2121 |
if (error) return error; /* Make sure the old pointer isn't garbage. */ old_value = be32_to_cpu(dip->di_next_unlinked); if (!xfs_verify_agino_or_null(mp, agno, old_value)) { |
a5155b870 xfs: always log c... |
2122 2123 |
xfs_inode_verifier_error(ip, -EFSCORRUPTED, __func__, dip, sizeof(*dip), __this_address); |
f2fc16a3d xfs: refactor ino... |
2124 2125 2126 2127 2128 2129 2130 2131 2132 2133 2134 |
error = -EFSCORRUPTED; goto out; } /* * Since we're updating a linked list, we should never find that the * current pointer is the same as the new value, unless we're * terminating the list. */ *old_next_agino = old_value; if (old_value == next_agino) { |
a5155b870 xfs: always log c... |
2135 2136 2137 |
if (next_agino != NULLAGINO) { xfs_inode_verifier_error(ip, -EFSCORRUPTED, __func__, dip, sizeof(*dip), __this_address); |
f2fc16a3d xfs: refactor ino... |
2138 |
error = -EFSCORRUPTED; |
a5155b870 xfs: always log c... |
2139 |
} |
f2fc16a3d xfs: refactor ino... |
2140 2141 2142 2143 2144 2145 2146 2147 2148 2149 2150 |
goto out; } /* Ok, update the new pointer. */ xfs_iunlink_update_dinode(tp, agno, XFS_INO_TO_AGINO(mp, ip->i_ino), ibp, dip, &ip->i_imap, next_agino); return 0; out: xfs_trans_brelse(tp, ibp); return error; } |
9a4a51186 xfs: refactor AGI... |
2151 |
/* |
c4a6bf7f6 xfs: don't ever p... |
2152 2153 |
* This is called when the inode's link count has gone to 0 or we are creating * a tmpfile via O_TMPFILE. The inode @ip must have nlink == 0. |
54d7b5c1d xfs: use vfs inod... |
2154 2155 2156 |
* * We place the on-disk inode on a list in the AGI. It will be pulled from this * list when the inode is freed. |
1da177e4c Linux-2.6.12-rc2 |
2157 |
*/ |
54d7b5c1d xfs: use vfs inod... |
2158 |
STATIC int |
1da177e4c Linux-2.6.12-rc2 |
2159 |
xfs_iunlink( |
5837f6259 xfs: clean up iun... |
2160 2161 |
struct xfs_trans *tp, struct xfs_inode *ip) |
1da177e4c Linux-2.6.12-rc2 |
2162 |
{ |
5837f6259 xfs: clean up iun... |
2163 2164 |
struct xfs_mount *mp = tp->t_mountp; struct xfs_agi *agi; |
5837f6259 xfs: clean up iun... |
2165 |
struct xfs_buf *agibp; |
86bfd3750 xfs: strengthen A... |
2166 |
xfs_agino_t next_agino; |
5837f6259 xfs: clean up iun... |
2167 2168 2169 |
xfs_agnumber_t agno = XFS_INO_TO_AGNO(mp, ip->i_ino); xfs_agino_t agino = XFS_INO_TO_AGINO(mp, ip->i_ino); short bucket_index = agino % XFS_AGI_UNLINKED_BUCKETS; |
5837f6259 xfs: clean up iun... |
2170 |
int error; |
1da177e4c Linux-2.6.12-rc2 |
2171 |
|
c4a6bf7f6 xfs: don't ever p... |
2172 |
ASSERT(VFS_I(ip)->i_nlink == 0); |
c19b3b05a xfs: mode di_mode... |
2173 |
ASSERT(VFS_I(ip)->i_mode != 0); |
4664c66c9 xfs: add tracepoi... |
2174 |
trace_xfs_iunlink(ip); |
1da177e4c Linux-2.6.12-rc2 |
2175 |
|
5837f6259 xfs: clean up iun... |
2176 2177 |
/* Get the agi buffer first. It ensures lock ordering on the list. */ error = xfs_read_agi(mp, tp, agno, &agibp); |
859d71827 [XFS] get_bulkall... |
2178 |
if (error) |
1da177e4c Linux-2.6.12-rc2 |
2179 |
return error; |
370c782b9 xfs: remove XFS_B... |
2180 |
agi = agibp->b_addr; |
5e1be0fb1 [XFS] factor out ... |
2181 |
|
1da177e4c Linux-2.6.12-rc2 |
2182 |
/* |
86bfd3750 xfs: strengthen A... |
2183 2184 2185 |
* Get the index into the agi hash table for the list this inode will * go on. Make sure the pointer isn't garbage and that this inode * isn't already on the list. |
1da177e4c Linux-2.6.12-rc2 |
2186 |
*/ |
86bfd3750 xfs: strengthen A... |
2187 2188 |
next_agino = be32_to_cpu(agi->agi_unlinked[bucket_index]); if (next_agino == agino || |
a5155b870 xfs: always log c... |
2189 |
!xfs_verify_agino_or_null(mp, agno, next_agino)) { |
8d57c2160 xfs: add a functi... |
2190 |
xfs_buf_mark_corrupt(agibp); |
86bfd3750 xfs: strengthen A... |
2191 |
return -EFSCORRUPTED; |
a5155b870 xfs: always log c... |
2192 |
} |
1da177e4c Linux-2.6.12-rc2 |
2193 |
|
86bfd3750 xfs: strengthen A... |
2194 |
if (next_agino != NULLAGINO) { |
9b2471797 xfs: cache unlink... |
2195 |
xfs_agino_t old_agino; |
f2fc16a3d xfs: refactor ino... |
2196 |
|
1da177e4c Linux-2.6.12-rc2 |
2197 |
/* |
f2fc16a3d xfs: refactor ino... |
2198 2199 |
* There is already another inode in the bucket, so point this * inode to the current head of the list. |
1da177e4c Linux-2.6.12-rc2 |
2200 |
*/ |
f2fc16a3d xfs: refactor ino... |
2201 2202 |
error = xfs_iunlink_update_inode(tp, ip, agno, next_agino, &old_agino); |
c319b58b1 [XFS] Make xfs_bu... |
2203 2204 |
if (error) return error; |
f2fc16a3d xfs: refactor ino... |
2205 |
ASSERT(old_agino == NULLAGINO); |
9b2471797 xfs: cache unlink... |
2206 2207 2208 2209 2210 |
/* * agino has been unlinked, add a backref from the next inode * back to agino. */ |
92a005448 xfs: get rid of u... |
2211 |
error = xfs_iunlink_add_backref(agibp->b_pag, agino, next_agino); |
9b2471797 xfs: cache unlink... |
2212 2213 |
if (error) return error; |
1da177e4c Linux-2.6.12-rc2 |
2214 |
} |
9a4a51186 xfs: refactor AGI... |
2215 2216 |
/* Point the head of the list to point to this inode. */ return xfs_iunlink_update_bucket(tp, agno, agibp, bucket_index, agino); |
1da177e4c Linux-2.6.12-rc2 |
2217 |
} |
23ffa52cc xfs: refactor unl... |
2218 2219 2220 2221 2222 2223 2224 2225 2226 2227 2228 2229 2230 2231 2232 2233 2234 2235 2236 2237 |
/* Return the imap, dinode pointer, and buffer for an inode. */ STATIC int xfs_iunlink_map_ino( struct xfs_trans *tp, xfs_agnumber_t agno, xfs_agino_t agino, struct xfs_imap *imap, struct xfs_dinode **dipp, struct xfs_buf **bpp) { struct xfs_mount *mp = tp->t_mountp; int error; imap->im_blkno = 0; error = xfs_imap(mp, tp, XFS_AGINO_TO_INO(mp, agno, agino), imap, 0); if (error) { xfs_warn(mp, "%s: xfs_imap returned error %d.", __func__, error); return error; } |
c19950799 xfs: remove unuse... |
2238 |
error = xfs_imap_to_bp(mp, tp, imap, dipp, bpp, 0); |
23ffa52cc xfs: refactor unl... |
2239 2240 2241 2242 2243 2244 2245 2246 2247 2248 2249 2250 2251 2252 2253 2254 2255 2256 2257 2258 2259 2260 2261 2262 2263 2264 2265 2266 |
if (error) { xfs_warn(mp, "%s: xfs_imap_to_bp returned error %d.", __func__, error); return error; } return 0; } /* * Walk the unlinked chain from @head_agino until we find the inode that * points to @target_agino. Return the inode number, map, dinode pointer, * and inode cluster buffer of that inode as @agino, @imap, @dipp, and @bpp. * * @tp, @pag, @head_agino, and @target_agino are input parameters. * @agino, @imap, @dipp, and @bpp are all output parameters. * * Do not call this function if @target_agino is the head of the list. */ STATIC int xfs_iunlink_map_prev( struct xfs_trans *tp, xfs_agnumber_t agno, xfs_agino_t head_agino, xfs_agino_t target_agino, xfs_agino_t *agino, struct xfs_imap *imap, struct xfs_dinode **dipp, |
9b2471797 xfs: cache unlink... |
2267 2268 |
struct xfs_buf **bpp, struct xfs_perag *pag) |
23ffa52cc xfs: refactor unl... |
2269 2270 2271 2272 2273 2274 2275 |
{ struct xfs_mount *mp = tp->t_mountp; xfs_agino_t next_agino; int error; ASSERT(head_agino != target_agino); *bpp = NULL; |
9b2471797 xfs: cache unlink... |
2276 2277 2278 2279 2280 2281 2282 2283 2284 2285 2286 2287 2288 2289 2290 2291 2292 2293 2294 2295 2296 2297 |
/* See if our backref cache can find it faster. */ *agino = xfs_iunlink_lookup_backref(pag, target_agino); if (*agino != NULLAGINO) { error = xfs_iunlink_map_ino(tp, agno, *agino, imap, dipp, bpp); if (error) return error; if (be32_to_cpu((*dipp)->di_next_unlinked) == target_agino) return 0; /* * If we get here the cache contents were corrupt, so drop the * buffer and fall back to walking the bucket list. */ xfs_trans_brelse(tp, *bpp); *bpp = NULL; WARN_ON_ONCE(1); } trace_xfs_iunlink_map_prev_fallback(mp, agno); /* Otherwise, walk the entire bucket until we find it. */ |
23ffa52cc xfs: refactor unl... |
2298 2299 2300 2301 2302 2303 2304 2305 2306 2307 2308 2309 2310 2311 2312 2313 2314 2315 2316 2317 2318 2319 2320 2321 2322 2323 2324 2325 2326 2327 2328 |
next_agino = head_agino; while (next_agino != target_agino) { xfs_agino_t unlinked_agino; if (*bpp) xfs_trans_brelse(tp, *bpp); *agino = next_agino; error = xfs_iunlink_map_ino(tp, agno, next_agino, imap, dipp, bpp); if (error) return error; unlinked_agino = be32_to_cpu((*dipp)->di_next_unlinked); /* * Make sure this pointer is valid and isn't an obvious * infinite loop. */ if (!xfs_verify_agino(mp, agno, unlinked_agino) || next_agino == unlinked_agino) { XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp, *dipp, sizeof(**dipp)); error = -EFSCORRUPTED; return error; } next_agino = unlinked_agino; } return 0; } |
1da177e4c Linux-2.6.12-rc2 |
2329 2330 2331 2332 2333 |
/* * Pull the on-disk inode from the AGI unlinked list. */ STATIC int xfs_iunlink_remove( |
5837f6259 xfs: clean up iun... |
2334 2335 |
struct xfs_trans *tp, struct xfs_inode *ip) |
1da177e4c Linux-2.6.12-rc2 |
2336 |
{ |
5837f6259 xfs: clean up iun... |
2337 2338 |
struct xfs_mount *mp = tp->t_mountp; struct xfs_agi *agi; |
5837f6259 xfs: clean up iun... |
2339 |
struct xfs_buf *agibp; |
5837f6259 xfs: clean up iun... |
2340 2341 |
struct xfs_buf *last_ibp; struct xfs_dinode *last_dip = NULL; |
5837f6259 xfs: clean up iun... |
2342 2343 2344 |
xfs_agnumber_t agno = XFS_INO_TO_AGNO(mp, ip->i_ino); xfs_agino_t agino = XFS_INO_TO_AGINO(mp, ip->i_ino); xfs_agino_t next_agino; |
b1d2a068e xfs: refactor ino... |
2345 |
xfs_agino_t head_agino; |
5837f6259 xfs: clean up iun... |
2346 |
short bucket_index = agino % XFS_AGI_UNLINKED_BUCKETS; |
5837f6259 xfs: clean up iun... |
2347 |
int error; |
1da177e4c Linux-2.6.12-rc2 |
2348 |
|
4664c66c9 xfs: add tracepoi... |
2349 |
trace_xfs_iunlink_remove(ip); |
5837f6259 xfs: clean up iun... |
2350 |
/* Get the agi buffer first. It ensures lock ordering on the list. */ |
5e1be0fb1 [XFS] factor out ... |
2351 2352 |
error = xfs_read_agi(mp, tp, agno, &agibp); if (error) |
1da177e4c Linux-2.6.12-rc2 |
2353 |
return error; |
370c782b9 xfs: remove XFS_B... |
2354 |
agi = agibp->b_addr; |
5e1be0fb1 [XFS] factor out ... |
2355 |
|
1da177e4c Linux-2.6.12-rc2 |
2356 |
/* |
86bfd3750 xfs: strengthen A... |
2357 2358 |
* Get the index into the agi hash table for the list this inode will * go on. Make sure the head pointer isn't garbage. |
1da177e4c Linux-2.6.12-rc2 |
2359 |
*/ |
b1d2a068e xfs: refactor ino... |
2360 2361 |
head_agino = be32_to_cpu(agi->agi_unlinked[bucket_index]); if (!xfs_verify_agino(mp, agno, head_agino)) { |
d2e736654 xfs: don't assert... |
2362 2363 2364 2365 |
XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp, agi, sizeof(*agi)); return -EFSCORRUPTED; } |
1da177e4c Linux-2.6.12-rc2 |
2366 |
|
b1d2a068e xfs: refactor ino... |
2367 2368 2369 2370 2371 2372 2373 2374 |
/* * Set our inode's next_unlinked pointer to NULL and then return * the old pointer value so that we can update whatever was previous * to us in the list to point to whatever was next in the list. */ error = xfs_iunlink_update_inode(tp, ip, agno, NULLAGINO, &next_agino); if (error) return error; |
9a4a51186 xfs: refactor AGI... |
2375 |
|
9b2471797 xfs: cache unlink... |
2376 2377 2378 2379 2380 2381 2382 2383 |
/* * If there was a backref pointing from the next inode back to this * one, remove it because we've removed this inode from the list. * * Later, if this inode was in the middle of the list we'll update * this inode's backref to point from the next inode. */ if (next_agino != NULLAGINO) { |
92a005448 xfs: get rid of u... |
2384 |
error = xfs_iunlink_change_backref(agibp->b_pag, next_agino, |
9b2471797 xfs: cache unlink... |
2385 2386 |
NULLAGINO); if (error) |
92a005448 xfs: get rid of u... |
2387 |
return error; |
9b2471797 xfs: cache unlink... |
2388 |
} |
92a005448 xfs: get rid of u... |
2389 |
if (head_agino != agino) { |
f2fc16a3d xfs: refactor ino... |
2390 2391 |
struct xfs_imap imap; xfs_agino_t prev_agino; |
23ffa52cc xfs: refactor unl... |
2392 |
/* We need to search the list for the inode being freed. */ |
b1d2a068e xfs: refactor ino... |
2393 |
error = xfs_iunlink_map_prev(tp, agno, head_agino, agino, |
9b2471797 xfs: cache unlink... |
2394 |
&prev_agino, &imap, &last_dip, &last_ibp, |
92a005448 xfs: get rid of u... |
2395 |
agibp->b_pag); |
23ffa52cc xfs: refactor unl... |
2396 |
if (error) |
92a005448 xfs: get rid of u... |
2397 |
return error; |
475ee413f xfs: merge xfs_it... |
2398 |
|
f2fc16a3d xfs: refactor ino... |
2399 2400 2401 |
/* Point the previous inode on the list to the next inode. */ xfs_iunlink_update_dinode(tp, agno, prev_agino, last_ibp, last_dip, &imap, next_agino); |
9b2471797 xfs: cache unlink... |
2402 2403 2404 2405 2406 2407 2408 2409 2410 |
/* * Now we deal with the backref for this inode. If this inode * pointed at a real inode, change the backref that pointed to * us to point to our old next. If this inode was the end of * the list, delete the backref that pointed to us. Note that * change_backref takes care of deleting the backref if * next_agino is NULLAGINO. */ |
92a005448 xfs: get rid of u... |
2411 2412 |
return xfs_iunlink_change_backref(agibp->b_pag, agino, next_agino); |
1da177e4c Linux-2.6.12-rc2 |
2413 |
} |
9b2471797 xfs: cache unlink... |
2414 |
|
92a005448 xfs: get rid of u... |
2415 2416 2417 |
/* Point the head of the list to the next unlinked inode. */ return xfs_iunlink_update_bucket(tp, agno, agibp, bucket_index, next_agino); |
1da177e4c Linux-2.6.12-rc2 |
2418 |
} |
5b3eed756 xfs: ensure we ma... |
2419 |
/* |
71e3e3564 xfs: rework stale... |
2420 2421 2422 |
* Look up the inode number specified and if it is not already marked XFS_ISTALE * mark it stale. We should only find clean inodes in this lookup that aren't * already stale. |
5806165a6 xfs: factor inode... |
2423 |
*/ |
71e3e3564 xfs: rework stale... |
2424 2425 2426 |
static void xfs_ifree_mark_inode_stale( struct xfs_buf *bp, |
5806165a6 xfs: factor inode... |
2427 |
struct xfs_inode *free_ip, |
d9fdd0adf xfs: fix inode nu... |
2428 |
xfs_ino_t inum) |
5806165a6 xfs: factor inode... |
2429 |
{ |
71e3e3564 xfs: rework stale... |
2430 2431 2432 |
struct xfs_mount *mp = bp->b_mount; struct xfs_perag *pag = bp->b_pag; struct xfs_inode_log_item *iip; |
5806165a6 xfs: factor inode... |
2433 2434 2435 2436 2437 2438 2439 |
struct xfs_inode *ip; retry: rcu_read_lock(); ip = radix_tree_lookup(&pag->pag_ici_root, XFS_INO_TO_AGINO(mp, inum)); /* Inode not in memory, nothing to do */ |
71e3e3564 xfs: rework stale... |
2440 2441 2442 2443 |
if (!ip) { rcu_read_unlock(); return; } |
5806165a6 xfs: factor inode... |
2444 2445 2446 2447 2448 2449 2450 2451 |
/* * because this is an RCU protected lookup, we could find a recently * freed or even reallocated inode during the lookup. We need to check * under the i_flags_lock for a valid inode here. Skip it if it is not * valid, the wrong inode or stale. */ spin_lock(&ip->i_flags_lock); |
718ecc503 xfs: xfs_iflock i... |
2452 2453 |
if (ip->i_ino != inum || __xfs_iflags_test(ip, XFS_ISTALE)) goto out_iflags_unlock; |
5806165a6 xfs: factor inode... |
2454 2455 2456 2457 2458 2459 2460 2461 2462 |
/* * Don't try to lock/unlock the current inode, but we _cannot_ skip the * other inodes that we did not find in the list attached to the buffer * and are not already marked stale. If we can't lock it, back off and * retry. */ if (ip != free_ip) { if (!xfs_ilock_nowait(ip, XFS_ILOCK_EXCL)) { |
71e3e3564 xfs: rework stale... |
2463 |
spin_unlock(&ip->i_flags_lock); |
5806165a6 xfs: factor inode... |
2464 2465 2466 2467 |
rcu_read_unlock(); delay(1); goto retry; } |
5806165a6 xfs: factor inode... |
2468 |
} |
71e3e3564 xfs: rework stale... |
2469 |
ip->i_flags |= XFS_ISTALE; |
5806165a6 xfs: factor inode... |
2470 |
|
71e3e3564 xfs: rework stale... |
2471 |
/* |
718ecc503 xfs: xfs_iflock i... |
2472 |
* If the inode is flushing, it is already attached to the buffer. All |
71e3e3564 xfs: rework stale... |
2473 2474 2475 2476 |
* we needed to do here is mark the inode stale so buffer IO completion * will remove it from the AIL. */ iip = ip->i_itemp; |
718ecc503 xfs: xfs_iflock i... |
2477 |
if (__xfs_iflags_test(ip, XFS_IFLUSHING)) { |
71e3e3564 xfs: rework stale... |
2478 2479 2480 2481 |
ASSERT(!list_empty(&iip->ili_item.li_bio_list)); ASSERT(iip->ili_last_fields); goto out_iunlock; } |
5806165a6 xfs: factor inode... |
2482 2483 |
/* |
48d55e2ae xfs: attach inode... |
2484 2485 2486 2487 |
* Inodes not attached to the buffer can be released immediately. * Everything else has to go through xfs_iflush_abort() on journal * commit as the flock synchronises removal of the inode from the * cluster buffer against inode reclaim. |
5806165a6 xfs: factor inode... |
2488 |
*/ |
718ecc503 xfs: xfs_iflock i... |
2489 |
if (!iip || list_empty(&iip->ili_item.li_bio_list)) |
71e3e3564 xfs: rework stale... |
2490 |
goto out_iunlock; |
718ecc503 xfs: xfs_iflock i... |
2491 2492 2493 2494 |
__xfs_iflags_set(ip, XFS_IFLUSHING); spin_unlock(&ip->i_flags_lock); rcu_read_unlock(); |
5806165a6 xfs: factor inode... |
2495 |
|
71e3e3564 xfs: rework stale... |
2496 |
/* we have a dirty inode in memory that has not yet been flushed. */ |
71e3e3564 xfs: rework stale... |
2497 2498 2499 2500 2501 |
spin_lock(&iip->ili_lock); iip->ili_last_fields = iip->ili_fields; iip->ili_fields = 0; iip->ili_fsync_fields = 0; spin_unlock(&iip->ili_lock); |
71e3e3564 xfs: rework stale... |
2502 |
ASSERT(iip->ili_last_fields); |
718ecc503 xfs: xfs_iflock i... |
2503 2504 2505 |
if (ip != free_ip) xfs_iunlock(ip, XFS_ILOCK_EXCL); return; |
71e3e3564 xfs: rework stale... |
2506 2507 2508 |
out_iunlock: if (ip != free_ip) xfs_iunlock(ip, XFS_ILOCK_EXCL); |
718ecc503 xfs: xfs_iflock i... |
2509 2510 2511 |
out_iflags_unlock: spin_unlock(&ip->i_flags_lock); rcu_read_unlock(); |
5806165a6 xfs: factor inode... |
2512 2513 2514 |
} /* |
0b8182dba xfs: fix the comm... |
2515 |
* A big issue when freeing the inode cluster is that we _cannot_ skip any |
5b3eed756 xfs: ensure we ma... |
2516 2517 2518 |
* inodes that are in memory - they all must be marked stale and attached to * the cluster buffer. */ |
2a30f36d9 xfs: Check the re... |
2519 |
STATIC int |
1da177e4c Linux-2.6.12-rc2 |
2520 |
xfs_ifree_cluster( |
71e3e3564 xfs: rework stale... |
2521 2522 |
struct xfs_inode *free_ip, struct xfs_trans *tp, |
09b566041 xfs: skip unalloc... |
2523 |
struct xfs_icluster *xic) |
1da177e4c Linux-2.6.12-rc2 |
2524 |
{ |
71e3e3564 xfs: rework stale... |
2525 2526 2527 2528 2529 |
struct xfs_mount *mp = free_ip->i_mount; struct xfs_ino_geometry *igeo = M_IGEO(mp); struct xfs_buf *bp; xfs_daddr_t blkno; xfs_ino_t inum = xic->first_ino; |
1da177e4c Linux-2.6.12-rc2 |
2530 |
int nbufs; |
5b257b4a1 xfs: fix race in ... |
2531 |
int i, j; |
3cdaa1898 xfs: fix sparse i... |
2532 |
int ioffset; |
ce92464c1 xfs: make xfs_tra... |
2533 |
int error; |
1da177e4c Linux-2.6.12-rc2 |
2534 |
|
ef3259599 xfs: separate ino... |
2535 |
nbufs = igeo->ialloc_blks / igeo->blocks_per_cluster; |
1da177e4c Linux-2.6.12-rc2 |
2536 |
|
ef3259599 xfs: separate ino... |
2537 |
for (j = 0; j < nbufs; j++, inum += igeo->inodes_per_cluster) { |
09b566041 xfs: skip unalloc... |
2538 2539 2540 2541 2542 |
/* * The allocation bitmap tells us which inodes of the chunk were * physically allocated. Skip the cluster if an inode falls into * a sparse region. */ |
3cdaa1898 xfs: fix sparse i... |
2543 2544 |
ioffset = inum - xic->first_ino; if ((xic->alloc & XFS_INOBT_MASK(ioffset)) == 0) { |
ef3259599 xfs: separate ino... |
2545 |
ASSERT(ioffset % igeo->inodes_per_cluster == 0); |
09b566041 xfs: skip unalloc... |
2546 2547 |
continue; } |
1da177e4c Linux-2.6.12-rc2 |
2548 2549 |
blkno = XFS_AGB_TO_DADDR(mp, XFS_INO_TO_AGNO(mp, inum), XFS_INO_TO_AGBNO(mp, inum)); |
5b257b4a1 xfs: fix race in ... |
2550 2551 |
/* * We obtain and lock the backing buffer first in the process |
718ecc503 xfs: xfs_iflock i... |
2552 2553 2554 |
* here to ensure dirty inodes attached to the buffer remain in * the flushing state while we mark them stale. * |
5b257b4a1 xfs: fix race in ... |
2555 2556 2557 2558 |
* If we scan the in-memory inodes first, then buffer IO can * complete before we get a lock on it, and hence we may fail * to mark all the active inodes on the buffer stale. */ |
ce92464c1 xfs: make xfs_tra... |
2559 2560 2561 |
error = xfs_trans_get_buf(tp, mp->m_ddev_targp, blkno, mp->m_bsize * igeo->blocks_per_cluster, XBF_UNMAPPED, &bp); |
71e3e3564 xfs: rework stale... |
2562 |
if (error) |
ce92464c1 xfs: make xfs_tra... |
2563 |
return error; |
b0f539de9 xfs: connect up w... |
2564 2565 2566 2567 2568 2569 2570 2571 2572 2573 |
/* * This buffer may not have been correctly initialised as we * didn't read it from disk. That's not important because we are * only using to mark the buffer as stale in the log, and to * attach stale cached inodes on it. That means it will never be * dispatched for IO. If it is, we want to know about it, and we * want it to fail. We can acheive this by adding a write * verifier to the buffer. */ |
8c4ce794e xfs: clean up ind... |
2574 |
bp->b_ops = &xfs_inode_buf_ops; |
b0f539de9 xfs: connect up w... |
2575 |
|
5b257b4a1 xfs: fix race in ... |
2576 |
/* |
71e3e3564 xfs: rework stale... |
2577 2578 2579 |
* Now we need to set all the cached clean inodes as XFS_ISTALE, * too. This requires lookups, and will skip inodes that we've * already marked XFS_ISTALE. |
1da177e4c Linux-2.6.12-rc2 |
2580 |
*/ |
71e3e3564 xfs: rework stale... |
2581 2582 |
for (i = 0; i < igeo->inodes_per_cluster; i++) xfs_ifree_mark_inode_stale(bp, free_ip, inum + i); |
1da177e4c Linux-2.6.12-rc2 |
2583 |
|
5b3eed756 xfs: ensure we ma... |
2584 |
xfs_trans_stale_inode_buf(tp, bp); |
1da177e4c Linux-2.6.12-rc2 |
2585 2586 |
xfs_trans_binval(tp, bp); } |
2a30f36d9 xfs: Check the re... |
2587 |
return 0; |
1da177e4c Linux-2.6.12-rc2 |
2588 2589 2590 2591 2592 2593 2594 2595 2596 2597 2598 2599 2600 2601 |
} /* * This is called to return an inode to the inode free list. * The inode should already be truncated to 0 length and have * no pages associated with it. This routine also assumes that * the inode is already a part of the transaction. * * The on-disk copy of the inode will have been added to the list * of unlinked inodes in the AGI. We need to remove the inode from * that list atomically with respect to freeing it here. */ int xfs_ifree( |
0e0417f3e xfs: remove dfops... |
2602 2603 |
struct xfs_trans *tp, struct xfs_inode *ip) |
1da177e4c Linux-2.6.12-rc2 |
2604 2605 |
{ int error; |
09b566041 xfs: skip unalloc... |
2606 |
struct xfs_icluster xic = { 0 }; |
1319ebefd xfs: add an inode... |
2607 |
struct xfs_inode_log_item *iip = ip->i_itemp; |
1da177e4c Linux-2.6.12-rc2 |
2608 |
|
579aa9caf [XFS] shrink mrlo... |
2609 |
ASSERT(xfs_isilocked(ip, XFS_ILOCK_EXCL)); |
54d7b5c1d xfs: use vfs inod... |
2610 |
ASSERT(VFS_I(ip)->i_nlink == 0); |
daf83964a xfs: move the per... |
2611 |
ASSERT(ip->i_df.if_nextents == 0); |
c19b3b05a xfs: mode di_mode... |
2612 |
ASSERT(ip->i_d.di_size == 0 || !S_ISREG(VFS_I(ip)->i_mode)); |
1da177e4c Linux-2.6.12-rc2 |
2613 2614 2615 2616 2617 2618 |
ASSERT(ip->i_d.di_nblocks == 0); /* * Pull the on-disk inode from the AGI unlinked list. */ error = xfs_iunlink_remove(tp, ip); |
1baaed8fa xfs: xfs_ifree do... |
2619 |
if (error) |
1da177e4c Linux-2.6.12-rc2 |
2620 |
return error; |
1da177e4c Linux-2.6.12-rc2 |
2621 |
|
0e0417f3e xfs: remove dfops... |
2622 |
error = xfs_difree(tp, ip->i_ino, &xic); |
1baaed8fa xfs: xfs_ifree do... |
2623 |
if (error) |
1da177e4c Linux-2.6.12-rc2 |
2624 |
return error; |
1baaed8fa xfs: xfs_ifree do... |
2625 |
|
b2c20045b xfs: remove xfs_i... |
2626 2627 2628 2629 2630 |
/* * Free any local-format data sitting around before we reset the * data fork to extents format. Note that the attr fork data has * already been freed by xfs_attr_inactive. */ |
f7e67b20e xfs: move the for... |
2631 |
if (ip->i_df.if_format == XFS_DINODE_FMT_LOCAL) { |
b2c20045b xfs: remove xfs_i... |
2632 2633 2634 2635 |
kmem_free(ip->i_df.if_u1.if_data); ip->i_df.if_u1.if_data = NULL; ip->i_df.if_bytes = 0; } |
98c4f78dc xfs: always free ... |
2636 |
|
c19b3b05a xfs: mode di_mode... |
2637 |
VFS_I(ip)->i_mode = 0; /* mark incore inode as free */ |
1da177e4c Linux-2.6.12-rc2 |
2638 |
ip->i_d.di_flags = 0; |
f93e5436f xfs: widen ondisk... |
2639 |
ip->i_d.di_flags2 = ip->i_mount->m_ino_geo.new_diflags2; |
1da177e4c Linux-2.6.12-rc2 |
2640 2641 |
ip->i_d.di_dmevmask = 0; ip->i_d.di_forkoff = 0; /* mark the attr fork not in use */ |
f7e67b20e xfs: move the for... |
2642 |
ip->i_df.if_format = XFS_DINODE_FMT_EXTENTS; |
dc1baa715 xfs: do not log/r... |
2643 2644 |
/* Don't attempt to replay owner changes for a deleted inode */ |
1319ebefd xfs: add an inode... |
2645 2646 2647 |
spin_lock(&iip->ili_lock); iip->ili_fields &= ~(XFS_ILOG_AOWNER | XFS_ILOG_DOWNER); spin_unlock(&iip->ili_lock); |
dc1baa715 xfs: do not log/r... |
2648 |
|
1da177e4c Linux-2.6.12-rc2 |
2649 2650 2651 2652 |
/* * Bump the generation count so no one will be confused * by reincarnations of this inode. */ |
9e9a2674e xfs: move inode g... |
2653 |
VFS_I(ip)->i_generation++; |
1da177e4c Linux-2.6.12-rc2 |
2654 |
xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE); |
09b566041 xfs: skip unalloc... |
2655 2656 |
if (xic.deleted) error = xfs_ifree_cluster(ip, tp, &xic); |
1da177e4c Linux-2.6.12-rc2 |
2657 |
|
2a30f36d9 xfs: Check the re... |
2658 |
return error; |
1da177e4c Linux-2.6.12-rc2 |
2659 2660 2661 |
} /* |
60ec67837 xfs: cleanup xfs_... |
2662 2663 2664 |
* This is called to unpin an inode. The caller must have the inode locked * in at least shared mode so that the buffer cannot be subsequently pinned * once someone is waiting for it to be unpinned. |
1da177e4c Linux-2.6.12-rc2 |
2665 |
*/ |
60ec67837 xfs: cleanup xfs_... |
2666 |
static void |
f392e6319 xfs: replace i_pi... |
2667 |
xfs_iunpin( |
60ec67837 xfs: cleanup xfs_... |
2668 |
struct xfs_inode *ip) |
1da177e4c Linux-2.6.12-rc2 |
2669 |
{ |
579aa9caf [XFS] shrink mrlo... |
2670 |
ASSERT(xfs_isilocked(ip, XFS_ILOCK_EXCL|XFS_ILOCK_SHARED)); |
1da177e4c Linux-2.6.12-rc2 |
2671 |
|
4aaf15d1a xfs: Add inode pi... |
2672 |
trace_xfs_inode_unpin_nowait(ip, _RET_IP_); |
a3f74ffb6 [XFS] Don't block... |
2673 |
/* Give the log a push to start the unpinning I/O */ |
656de4ffa xfs: merge _xfs_l... |
2674 |
xfs_log_force_lsn(ip->i_mount, ip->i_itemp->ili_last_lsn, 0, NULL); |
a14a348bf xfs: cleanup up x... |
2675 |
|
a3f74ffb6 [XFS] Don't block... |
2676 |
} |
1da177e4c Linux-2.6.12-rc2 |
2677 |
|
f392e6319 xfs: replace i_pi... |
2678 2679 2680 2681 2682 2683 2684 2685 2686 2687 |
static void __xfs_iunpin_wait( struct xfs_inode *ip) { wait_queue_head_t *wq = bit_waitqueue(&ip->i_flags, __XFS_IPINNED_BIT); DEFINE_WAIT_BIT(wait, &ip->i_flags, __XFS_IPINNED_BIT); xfs_iunpin(ip); do { |
214171361 sched/wait: Stand... |
2688 |
prepare_to_wait(wq, &wait.wq_entry, TASK_UNINTERRUPTIBLE); |
f392e6319 xfs: replace i_pi... |
2689 2690 2691 |
if (xfs_ipincount(ip)) io_schedule(); } while (xfs_ipincount(ip)); |
214171361 sched/wait: Stand... |
2692 |
finish_wait(wq, &wait.wq_entry); |
f392e6319 xfs: replace i_pi... |
2693 |
} |
777df5afd xfs: Make inode r... |
2694 |
void |
a3f74ffb6 [XFS] Don't block... |
2695 |
xfs_iunpin_wait( |
60ec67837 xfs: cleanup xfs_... |
2696 |
struct xfs_inode *ip) |
a3f74ffb6 [XFS] Don't block... |
2697 |
{ |
f392e6319 xfs: replace i_pi... |
2698 2699 |
if (xfs_ipincount(ip)) __xfs_iunpin_wait(ip); |
1da177e4c Linux-2.6.12-rc2 |
2700 |
} |
273203699 xfs: xfs_remove d... |
2701 2702 2703 2704 2705 2706 2707 2708 2709 2710 2711 2712 2713 2714 2715 2716 2717 2718 2719 2720 2721 2722 |
/* * Removing an inode from the namespace involves removing the directory entry * and dropping the link count on the inode. Removing the directory entry can * result in locking an AGF (directory blocks were freed) and removing a link * count can result in placing the inode on an unlinked list which results in * locking an AGI. * * The big problem here is that we have an ordering constraint on AGF and AGI * locking - inode allocation locks the AGI, then can allocate a new extent for * new inodes, locking the AGF after the AGI. Similarly, freeing the inode * removes the inode from the unlinked list, requiring that we lock the AGI * first, and then freeing the inode can result in an inode chunk being freed * and hence freeing disk space requiring that we lock an AGF. * * Hence the ordering that is imposed by other parts of the code is AGI before * AGF. This means we cannot remove the directory entry before we drop the inode * reference count and put it on the unlinked list as this results in a lock * order of AGF then AGI, and this can deadlock against inode allocation and * freeing. Therefore we must drop the link counts before we remove the * directory entry. * * This is still safe from a transactional point of view - it is not until we |
310a75a3c xfs: change xfs_b... |
2723 |
* get to xfs_defer_finish() that we have the possibility of multiple |
273203699 xfs: xfs_remove d... |
2724 2725 2726 2727 |
* transactions in this operation. Hence as long as we remove the directory * entry and drop the link count in the first transaction of the remove * operation, there are no transactional constraints on the ordering here. */ |
c24b5dfad xfs: kill xfs_vno... |
2728 2729 2730 2731 2732 2733 2734 2735 |
int xfs_remove( xfs_inode_t *dp, struct xfs_name *name, xfs_inode_t *ip) { xfs_mount_t *mp = dp->i_mount; xfs_trans_t *tp = NULL; |
c19b3b05a xfs: mode di_mode... |
2736 |
int is_dir = S_ISDIR(VFS_I(ip)->i_mode); |
c24b5dfad xfs: kill xfs_vno... |
2737 |
int error = 0; |
c24b5dfad xfs: kill xfs_vno... |
2738 |
uint resblks; |
c24b5dfad xfs: kill xfs_vno... |
2739 2740 2741 2742 |
trace_xfs_remove(dp, name); if (XFS_FORCED_SHUTDOWN(mp)) |
2451337dd xfs: global error... |
2743 |
return -EIO; |
c24b5dfad xfs: kill xfs_vno... |
2744 |
|
c14cfccab xfs: remove unnec... |
2745 |
error = xfs_qm_dqattach(dp); |
c24b5dfad xfs: kill xfs_vno... |
2746 2747 |
if (error) goto std_return; |
c14cfccab xfs: remove unnec... |
2748 |
error = xfs_qm_dqattach(ip); |
c24b5dfad xfs: kill xfs_vno... |
2749 2750 |
if (error) goto std_return; |
c24b5dfad xfs: kill xfs_vno... |
2751 2752 2753 2754 2755 2756 2757 2758 2759 2760 |
/* * We try to get the real space reservation first, * allowing for directory btree deletion(s) implying * possible bmap insert(s). If we can't get the space * reservation then we use 0 instead, and avoid the bmap * btree insert(s) in the directory code by, if the bmap * insert tries to happen, instead trimming the LAST * block from the directory. */ resblks = XFS_REMOVE_SPACE_RES(mp); |
253f4911f xfs: better xfs_t... |
2761 |
error = xfs_trans_alloc(mp, &M_RES(mp)->tr_remove, resblks, 0, 0, &tp); |
2451337dd xfs: global error... |
2762 |
if (error == -ENOSPC) { |
c24b5dfad xfs: kill xfs_vno... |
2763 |
resblks = 0; |
253f4911f xfs: better xfs_t... |
2764 2765 |
error = xfs_trans_alloc(mp, &M_RES(mp)->tr_remove, 0, 0, 0, &tp); |
c24b5dfad xfs: kill xfs_vno... |
2766 2767 |
} if (error) { |
2451337dd xfs: global error... |
2768 |
ASSERT(error != -ENOSPC); |
253f4911f xfs: better xfs_t... |
2769 |
goto std_return; |
c24b5dfad xfs: kill xfs_vno... |
2770 |
} |
7c2d238ac xfs: allow xfs_lo... |
2771 |
xfs_lock_two_inodes(dp, XFS_ILOCK_EXCL, ip, XFS_ILOCK_EXCL); |
c24b5dfad xfs: kill xfs_vno... |
2772 |
|
655232183 xfs: remove i_iol... |
2773 |
xfs_trans_ijoin(tp, dp, XFS_ILOCK_EXCL); |
c24b5dfad xfs: kill xfs_vno... |
2774 2775 2776 2777 2778 2779 |
xfs_trans_ijoin(tp, ip, XFS_ILOCK_EXCL); /* * If we're removing a directory perform some additional validation. */ if (is_dir) { |
54d7b5c1d xfs: use vfs inod... |
2780 2781 |
ASSERT(VFS_I(ip)->i_nlink >= 2); if (VFS_I(ip)->i_nlink != 2) { |
2451337dd xfs: global error... |
2782 |
error = -ENOTEMPTY; |
c24b5dfad xfs: kill xfs_vno... |
2783 2784 2785 |
goto out_trans_cancel; } if (!xfs_dir_isempty(ip)) { |
2451337dd xfs: global error... |
2786 |
error = -ENOTEMPTY; |
c24b5dfad xfs: kill xfs_vno... |
2787 2788 |
goto out_trans_cancel; } |
c24b5dfad xfs: kill xfs_vno... |
2789 |
|
273203699 xfs: xfs_remove d... |
2790 |
/* Drop the link from ip's "..". */ |
c24b5dfad xfs: kill xfs_vno... |
2791 2792 |
error = xfs_droplink(tp, dp); if (error) |
273203699 xfs: xfs_remove d... |
2793 |
goto out_trans_cancel; |
c24b5dfad xfs: kill xfs_vno... |
2794 |
|
273203699 xfs: xfs_remove d... |
2795 |
/* Drop the "." link from ip to self. */ |
c24b5dfad xfs: kill xfs_vno... |
2796 2797 |
error = xfs_droplink(tp, ip); if (error) |
273203699 xfs: xfs_remove d... |
2798 |
goto out_trans_cancel; |
c24b5dfad xfs: kill xfs_vno... |
2799 2800 2801 2802 2803 2804 2805 2806 |
} else { /* * When removing a non-directory we need to log the parent * inode here. For a directory this is done implicitly * by the xfs_droplink call for the ".." entry. */ xfs_trans_log_inode(tp, dp, XFS_ILOG_CORE); } |
273203699 xfs: xfs_remove d... |
2807 |
xfs_trans_ichgtime(tp, dp, XFS_ICHGTIME_MOD | XFS_ICHGTIME_CHG); |
c24b5dfad xfs: kill xfs_vno... |
2808 |
|
273203699 xfs: xfs_remove d... |
2809 |
/* Drop the link from dp to ip. */ |
c24b5dfad xfs: kill xfs_vno... |
2810 2811 |
error = xfs_droplink(tp, ip); if (error) |
273203699 xfs: xfs_remove d... |
2812 |
goto out_trans_cancel; |
c24b5dfad xfs: kill xfs_vno... |
2813 |
|
381eee69f xfs: remove first... |
2814 |
error = xfs_dir_removename(tp, dp, name, ip->i_ino, resblks); |
273203699 xfs: xfs_remove d... |
2815 |
if (error) { |
2451337dd xfs: global error... |
2816 |
ASSERT(error != -ENOENT); |
c8eac49ef xfs: remove all b... |
2817 |
goto out_trans_cancel; |
273203699 xfs: xfs_remove d... |
2818 |
} |
c24b5dfad xfs: kill xfs_vno... |
2819 2820 2821 2822 2823 2824 2825 |
/* * If this is a synchronous mount, make sure that the * remove transaction goes to disk before returning to * the user. */ if (mp->m_flags & (XFS_MOUNT_WSYNC|XFS_MOUNT_DIRSYNC)) xfs_trans_set_sync(tp); |
70393313d xfs: saner xfs_tr... |
2826 |
error = xfs_trans_commit(tp); |
c24b5dfad xfs: kill xfs_vno... |
2827 2828 |
if (error) goto std_return; |
2cd2ef6a3 xfs: rewrite the ... |
2829 |
if (is_dir && xfs_inode_is_filestream(ip)) |
c24b5dfad xfs: kill xfs_vno... |
2830 2831 2832 |
xfs_filestream_deassociate(ip); return 0; |
c24b5dfad xfs: kill xfs_vno... |
2833 |
out_trans_cancel: |
4906e2154 xfs: remove the f... |
2834 |
xfs_trans_cancel(tp); |
c24b5dfad xfs: kill xfs_vno... |
2835 2836 2837 |
std_return: return error; } |
f6bba2017 xfs: consolidate ... |
2838 2839 2840 |
/* * Enter all inodes for a rename transaction into a sorted array. */ |
95afcf5c7 xfs: clean up ino... |
2841 |
#define __XFS_SORT_INODES 5 |
f6bba2017 xfs: consolidate ... |
2842 2843 |
STATIC void xfs_sort_for_rename( |
95afcf5c7 xfs: clean up ino... |
2844 2845 2846 2847 2848 2849 2850 |
struct xfs_inode *dp1, /* in: old (source) directory inode */ struct xfs_inode *dp2, /* in: new (target) directory inode */ struct xfs_inode *ip1, /* in: inode of old entry */ struct xfs_inode *ip2, /* in: inode of new entry */ struct xfs_inode *wip, /* in: whiteout inode */ struct xfs_inode **i_tab,/* out: sorted array of inodes */ int *num_inodes) /* in/out: inodes in array */ |
f6bba2017 xfs: consolidate ... |
2851 |
{ |
f6bba2017 xfs: consolidate ... |
2852 |
int i, j; |
95afcf5c7 xfs: clean up ino... |
2853 2854 |
ASSERT(*num_inodes == __XFS_SORT_INODES); memset(i_tab, 0, *num_inodes * sizeof(struct xfs_inode *)); |
f6bba2017 xfs: consolidate ... |
2855 2856 2857 2858 2859 2860 2861 |
/* * i_tab contains a list of pointers to inodes. We initialize * the table here & we'll sort it. We will then use it to * order the acquisition of the inode locks. * * Note that the table may contain duplicates. e.g., dp1 == dp2. */ |
95afcf5c7 xfs: clean up ino... |
2862 2863 2864 2865 2866 2867 2868 2869 2870 |
i = 0; i_tab[i++] = dp1; i_tab[i++] = dp2; i_tab[i++] = ip1; if (ip2) i_tab[i++] = ip2; if (wip) i_tab[i++] = wip; *num_inodes = i; |
f6bba2017 xfs: consolidate ... |
2871 2872 2873 |
/* * Sort the elements via bubble sort. (Remember, there are at |
95afcf5c7 xfs: clean up ino... |
2874 |
* most 5 elements to sort, so this is adequate.) |
f6bba2017 xfs: consolidate ... |
2875 2876 2877 2878 |
*/ for (i = 0; i < *num_inodes; i++) { for (j = 1; j < *num_inodes; j++) { if (i_tab[j]->i_ino < i_tab[j-1]->i_ino) { |
95afcf5c7 xfs: clean up ino... |
2879 |
struct xfs_inode *temp = i_tab[j]; |
f6bba2017 xfs: consolidate ... |
2880 2881 2882 2883 2884 2885 |
i_tab[j] = i_tab[j-1]; i_tab[j-1] = temp; } } } } |
310606b0c xfs: factor out x... |
2886 2887 |
static int xfs_finish_rename( |
c9cfdb381 xfs: remove dfops... |
2888 |
struct xfs_trans *tp) |
310606b0c xfs: factor out x... |
2889 |
{ |
310606b0c xfs: factor out x... |
2890 2891 2892 2893 2894 2895 |
/* * If this is a synchronous mount, make sure that the rename transaction * goes to disk before returning to the user. */ if (tp->t_mountp->m_flags & (XFS_MOUNT_WSYNC|XFS_MOUNT_DIRSYNC)) xfs_trans_set_sync(tp); |
70393313d xfs: saner xfs_tr... |
2896 |
return xfs_trans_commit(tp); |
310606b0c xfs: factor out x... |
2897 |
} |
f6bba2017 xfs: consolidate ... |
2898 |
/* |
d31a18254 xfs: Add support ... |
2899 2900 2901 2902 2903 2904 2905 2906 2907 2908 2909 2910 2911 |
* xfs_cross_rename() * * responsible for handling RENAME_EXCHANGE flag in renameat2() sytemcall */ STATIC int xfs_cross_rename( struct xfs_trans *tp, struct xfs_inode *dp1, struct xfs_name *name1, struct xfs_inode *ip1, struct xfs_inode *dp2, struct xfs_name *name2, struct xfs_inode *ip2, |
d31a18254 xfs: Add support ... |
2912 2913 2914 2915 2916 2917 2918 2919 |
int spaceres) { int error = 0; int ip1_flags = 0; int ip2_flags = 0; int dp2_flags = 0; /* Swap inode number for dirent in first parent */ |
381eee69f xfs: remove first... |
2920 |
error = xfs_dir_replace(tp, dp1, name1, ip2->i_ino, spaceres); |
d31a18254 xfs: Add support ... |
2921 |
if (error) |
eeacd3217 xfs: make xfs_cro... |
2922 |
goto out_trans_abort; |
d31a18254 xfs: Add support ... |
2923 2924 |
/* Swap inode number for dirent in second parent */ |
381eee69f xfs: remove first... |
2925 |
error = xfs_dir_replace(tp, dp2, name2, ip1->i_ino, spaceres); |
d31a18254 xfs: Add support ... |
2926 |
if (error) |
eeacd3217 xfs: make xfs_cro... |
2927 |
goto out_trans_abort; |
d31a18254 xfs: Add support ... |
2928 2929 2930 2931 2932 2933 2934 2935 |
/* * If we're renaming one or more directories across different parents, * update the respective ".." entries (and link counts) to match the new * parents. */ if (dp1 != dp2) { dp2_flags = XFS_ICHGTIME_MOD | XFS_ICHGTIME_CHG; |
c19b3b05a xfs: mode di_mode... |
2936 |
if (S_ISDIR(VFS_I(ip2)->i_mode)) { |
d31a18254 xfs: Add support ... |
2937 |
error = xfs_dir_replace(tp, ip2, &xfs_name_dotdot, |
381eee69f xfs: remove first... |
2938 |
dp1->i_ino, spaceres); |
d31a18254 xfs: Add support ... |
2939 |
if (error) |
eeacd3217 xfs: make xfs_cro... |
2940 |
goto out_trans_abort; |
d31a18254 xfs: Add support ... |
2941 2942 |
/* transfer ip2 ".." reference to dp1 */ |
c19b3b05a xfs: mode di_mode... |
2943 |
if (!S_ISDIR(VFS_I(ip1)->i_mode)) { |
d31a18254 xfs: Add support ... |
2944 2945 |
error = xfs_droplink(tp, dp2); if (error) |
eeacd3217 xfs: make xfs_cro... |
2946 |
goto out_trans_abort; |
910832697 xfs: change some ... |
2947 |
xfs_bumplink(tp, dp1); |
d31a18254 xfs: Add support ... |
2948 2949 2950 2951 2952 2953 2954 2955 2956 2957 2958 |
} /* * Although ip1 isn't changed here, userspace needs * to be warned about the change, so that applications * relying on it (like backup ones), will properly * notify the change */ ip1_flags |= XFS_ICHGTIME_CHG; ip2_flags |= XFS_ICHGTIME_MOD | XFS_ICHGTIME_CHG; } |
c19b3b05a xfs: mode di_mode... |
2959 |
if (S_ISDIR(VFS_I(ip1)->i_mode)) { |
d31a18254 xfs: Add support ... |
2960 |
error = xfs_dir_replace(tp, ip1, &xfs_name_dotdot, |
381eee69f xfs: remove first... |
2961 |
dp2->i_ino, spaceres); |
d31a18254 xfs: Add support ... |
2962 |
if (error) |
eeacd3217 xfs: make xfs_cro... |
2963 |
goto out_trans_abort; |
d31a18254 xfs: Add support ... |
2964 2965 |
/* transfer ip1 ".." reference to dp2 */ |
c19b3b05a xfs: mode di_mode... |
2966 |
if (!S_ISDIR(VFS_I(ip2)->i_mode)) { |
d31a18254 xfs: Add support ... |
2967 2968 |
error = xfs_droplink(tp, dp1); if (error) |
eeacd3217 xfs: make xfs_cro... |
2969 |
goto out_trans_abort; |
910832697 xfs: change some ... |
2970 |
xfs_bumplink(tp, dp2); |
d31a18254 xfs: Add support ... |
2971 2972 2973 2974 2975 2976 2977 2978 2979 2980 2981 2982 2983 2984 2985 2986 2987 2988 2989 2990 2991 2992 2993 2994 2995 2996 2997 |
} /* * Although ip2 isn't changed here, userspace needs * to be warned about the change, so that applications * relying on it (like backup ones), will properly * notify the change */ ip1_flags |= XFS_ICHGTIME_MOD | XFS_ICHGTIME_CHG; ip2_flags |= XFS_ICHGTIME_CHG; } } if (ip1_flags) { xfs_trans_ichgtime(tp, ip1, ip1_flags); xfs_trans_log_inode(tp, ip1, XFS_ILOG_CORE); } if (ip2_flags) { xfs_trans_ichgtime(tp, ip2, ip2_flags); xfs_trans_log_inode(tp, ip2, XFS_ILOG_CORE); } if (dp2_flags) { xfs_trans_ichgtime(tp, dp2, dp2_flags); xfs_trans_log_inode(tp, dp2, XFS_ILOG_CORE); } xfs_trans_ichgtime(tp, dp1, XFS_ICHGTIME_MOD | XFS_ICHGTIME_CHG); xfs_trans_log_inode(tp, dp1, XFS_ILOG_CORE); |
c9cfdb381 xfs: remove dfops... |
2998 |
return xfs_finish_rename(tp); |
eeacd3217 xfs: make xfs_cro... |
2999 3000 |
out_trans_abort: |
4906e2154 xfs: remove the f... |
3001 |
xfs_trans_cancel(tp); |
d31a18254 xfs: Add support ... |
3002 3003 3004 3005 |
return error; } /* |
7dcf5c3e4 xfs: add RENAME_W... |
3006 3007 |
* xfs_rename_alloc_whiteout() * |
b63da6c8d xfs: delete dupli... |
3008 |
* Return a referenced, unlinked, unlocked inode that can be used as a |
7dcf5c3e4 xfs: add RENAME_W... |
3009 3010 3011 3012 3013 3014 3015 3016 3017 3018 3019 |
* whiteout in a rename transaction. We use a tmpfile inode here so that if we * crash between allocating the inode and linking it into the rename transaction * recovery will free the inode and we won't leak it. */ static int xfs_rename_alloc_whiteout( struct xfs_inode *dp, struct xfs_inode **wip) { struct xfs_inode *tmpfile; int error; |
a1f69417c xfs: non-scrub - ... |
3020 |
error = xfs_create_tmpfile(dp, S_IFCHR | WHITEOUT_MODE, &tmpfile); |
7dcf5c3e4 xfs: add RENAME_W... |
3021 3022 |
if (error) return error; |
22419ac9f xfs: fix broken i... |
3023 3024 |
/* * Prepare the tmpfile inode as if it were created through the VFS. |
c4a6bf7f6 xfs: don't ever p... |
3025 3026 |
* Complete the inode setup and flag it as linkable. nlink is already * zero, so we can skip the drop_nlink. |
22419ac9f xfs: fix broken i... |
3027 |
*/ |
2b3d1d41b xfs: set up inode... |
3028 |
xfs_setup_iops(tmpfile); |
7dcf5c3e4 xfs: add RENAME_W... |
3029 3030 3031 3032 3033 3034 3035 3036 |
xfs_finish_inode_setup(tmpfile); VFS_I(tmpfile)->i_state |= I_LINKABLE; *wip = tmpfile; return 0; } /* |
f6bba2017 xfs: consolidate ... |
3037 3038 3039 3040 |
* xfs_rename */ int xfs_rename( |
7dcf5c3e4 xfs: add RENAME_W... |
3041 3042 3043 3044 3045 3046 3047 |
struct xfs_inode *src_dp, struct xfs_name *src_name, struct xfs_inode *src_ip, struct xfs_inode *target_dp, struct xfs_name *target_name, struct xfs_inode *target_ip, unsigned int flags) |
f6bba2017 xfs: consolidate ... |
3048 |
{ |
7dcf5c3e4 xfs: add RENAME_W... |
3049 3050 |
struct xfs_mount *mp = src_dp->i_mount; struct xfs_trans *tp; |
7dcf5c3e4 xfs: add RENAME_W... |
3051 3052 |
struct xfs_inode *wip = NULL; /* whiteout inode */ struct xfs_inode *inodes[__XFS_SORT_INODES]; |
93597ae8d xfs: Fix deadlock... |
3053 |
struct xfs_buf *agibp; |
7dcf5c3e4 xfs: add RENAME_W... |
3054 |
int num_inodes = __XFS_SORT_INODES; |
2b93681f5 Merge branch 'xfs... |
3055 |
bool new_parent = (src_dp != target_dp); |
c19b3b05a xfs: mode di_mode... |
3056 |
bool src_is_directory = S_ISDIR(VFS_I(src_ip)->i_mode); |
7dcf5c3e4 xfs: add RENAME_W... |
3057 3058 |
int spaceres; int error; |
f6bba2017 xfs: consolidate ... |
3059 3060 |
trace_xfs_rename(src_dp, target_dp, src_name, target_name); |
eeacd3217 xfs: make xfs_cro... |
3061 3062 |
if ((flags & RENAME_EXCHANGE) && !target_ip) return -EINVAL; |
7dcf5c3e4 xfs: add RENAME_W... |
3063 3064 3065 3066 3067 3068 3069 3070 3071 3072 3073 3074 3075 3076 |
/* * If we are doing a whiteout operation, allocate the whiteout inode * we will be placing at the target and ensure the type is set * appropriately. */ if (flags & RENAME_WHITEOUT) { ASSERT(!(flags & (RENAME_NOREPLACE | RENAME_EXCHANGE))); error = xfs_rename_alloc_whiteout(target_dp, &wip); if (error) return error; /* setup target dirent info as whiteout */ src_name->type = XFS_DIR3_FT_CHRDEV; } |
f6bba2017 xfs: consolidate ... |
3077 |
|
7dcf5c3e4 xfs: add RENAME_W... |
3078 |
xfs_sort_for_rename(src_dp, target_dp, src_ip, target_ip, wip, |
f6bba2017 xfs: consolidate ... |
3079 |
inodes, &num_inodes); |
f6bba2017 xfs: consolidate ... |
3080 |
spaceres = XFS_RENAME_SPACE_RES(mp, target_name->len); |
253f4911f xfs: better xfs_t... |
3081 |
error = xfs_trans_alloc(mp, &M_RES(mp)->tr_rename, spaceres, 0, 0, &tp); |
2451337dd xfs: global error... |
3082 |
if (error == -ENOSPC) { |
f6bba2017 xfs: consolidate ... |
3083 |
spaceres = 0; |
253f4911f xfs: better xfs_t... |
3084 3085 |
error = xfs_trans_alloc(mp, &M_RES(mp)->tr_rename, 0, 0, 0, &tp); |
f6bba2017 xfs: consolidate ... |
3086 |
} |
445883e81 xfs: cleanup xfs_... |
3087 |
if (error) |
253f4911f xfs: better xfs_t... |
3088 |
goto out_release_wip; |
f6bba2017 xfs: consolidate ... |
3089 3090 3091 3092 3093 |
/* * Attach the dquots to the inodes */ error = xfs_qm_vop_rename_dqattach(inodes); |
445883e81 xfs: cleanup xfs_... |
3094 3095 |
if (error) goto out_trans_cancel; |
f6bba2017 xfs: consolidate ... |
3096 3097 3098 3099 3100 3101 3102 3103 3104 3105 3106 3107 3108 3109 |
/* * Lock all the participating inodes. Depending upon whether * the target_name exists in the target directory, and * whether the target directory is the same as the source * directory, we can lock from 2 to 4 inodes. */ xfs_lock_inodes(inodes, num_inodes, XFS_ILOCK_EXCL); /* * Join all the inodes to the transaction. From this point on, * we can rely on either trans_commit or trans_cancel to unlock * them. */ |
655232183 xfs: remove i_iol... |
3110 |
xfs_trans_ijoin(tp, src_dp, XFS_ILOCK_EXCL); |
f6bba2017 xfs: consolidate ... |
3111 |
if (new_parent) |
655232183 xfs: remove i_iol... |
3112 |
xfs_trans_ijoin(tp, target_dp, XFS_ILOCK_EXCL); |
f6bba2017 xfs: consolidate ... |
3113 3114 3115 |
xfs_trans_ijoin(tp, src_ip, XFS_ILOCK_EXCL); if (target_ip) xfs_trans_ijoin(tp, target_ip, XFS_ILOCK_EXCL); |
7dcf5c3e4 xfs: add RENAME_W... |
3116 3117 |
if (wip) xfs_trans_ijoin(tp, wip, XFS_ILOCK_EXCL); |
f6bba2017 xfs: consolidate ... |
3118 3119 3120 3121 3122 3123 3124 |
/* * If we are using project inheritance, we only allow renames * into our tree when the project IDs are the same; else the * tree quota mechanism would be circumvented. */ if (unlikely((target_dp->i_d.di_flags & XFS_DIFLAG_PROJINHERIT) && |
de7a866fd xfs: merge the pr... |
3125 |
target_dp->i_d.di_projid != src_ip->i_d.di_projid)) { |
2451337dd xfs: global error... |
3126 |
error = -EXDEV; |
445883e81 xfs: cleanup xfs_... |
3127 |
goto out_trans_cancel; |
f6bba2017 xfs: consolidate ... |
3128 |
} |
eeacd3217 xfs: make xfs_cro... |
3129 3130 3131 3132 |
/* RENAME_EXCHANGE is unique from here on. */ if (flags & RENAME_EXCHANGE) return xfs_cross_rename(tp, src_dp, src_name, src_ip, target_dp, target_name, target_ip, |
f16dea54b xfs: use ->t_firs... |
3133 |
spaceres); |
d31a18254 xfs: Add support ... |
3134 3135 |
/* |
bc56ad8c7 xfs: Fix deadlock... |
3136 3137 |
* Check for expected errors before we dirty the transaction * so we can return an error without a transaction abort. |
f6bba2017 xfs: consolidate ... |
3138 3139 3140 3141 3142 3143 |
*/ if (target_ip == NULL) { /* * If there's no space reservation, check the entry will * fit before actually inserting it. */ |
94f3cad55 xfs: check resblk... |
3144 3145 3146 |
if (!spaceres) { error = xfs_dir_canenter(tp, target_dp, target_name); if (error) |
445883e81 xfs: cleanup xfs_... |
3147 |
goto out_trans_cancel; |
94f3cad55 xfs: check resblk... |
3148 |
} |
bc56ad8c7 xfs: Fix deadlock... |
3149 3150 3151 3152 3153 3154 3155 3156 3157 3158 3159 3160 3161 3162 3163 3164 3165 3166 3167 3168 3169 3170 3171 3172 3173 3174 3175 3176 3177 3178 3179 3180 |
} else { /* * If target exists and it's a directory, check that whether * it can be destroyed. */ if (S_ISDIR(VFS_I(target_ip)->i_mode) && (!xfs_dir_isempty(target_ip) || (VFS_I(target_ip)->i_nlink > 2))) { error = -EEXIST; goto out_trans_cancel; } } /* * Directory entry creation below may acquire the AGF. Remove * the whiteout from the unlinked list first to preserve correct * AGI/AGF locking order. This dirties the transaction so failures * after this point will abort and log recovery will clean up the * mess. * * For whiteouts, we need to bump the link count on the whiteout * inode. After this point, we have a real link, clear the tmpfile * state flag from the inode so it doesn't accidentally get misused * in future. */ if (wip) { ASSERT(VFS_I(wip)->i_nlink == 0); error = xfs_iunlink_remove(tp, wip); if (error) goto out_trans_cancel; xfs_bumplink(tp, wip); |
bc56ad8c7 xfs: Fix deadlock... |
3181 3182 3183 3184 3185 3186 3187 |
VFS_I(wip)->i_state &= ~I_LINKABLE; } /* * Set up the target. */ if (target_ip == NULL) { |
f6bba2017 xfs: consolidate ... |
3188 3189 3190 3191 3192 3193 |
/* * If target does not exist and the rename crosses * directories, adjust the target directory link count * to account for the ".." reference from the new entry. */ error = xfs_dir_createname(tp, target_dp, target_name, |
381eee69f xfs: remove first... |
3194 |
src_ip->i_ino, spaceres); |
f6bba2017 xfs: consolidate ... |
3195 |
if (error) |
c8eac49ef xfs: remove all b... |
3196 |
goto out_trans_cancel; |
f6bba2017 xfs: consolidate ... |
3197 3198 3199 3200 3201 |
xfs_trans_ichgtime(tp, target_dp, XFS_ICHGTIME_MOD | XFS_ICHGTIME_CHG); if (new_parent && src_is_directory) { |
910832697 xfs: change some ... |
3202 |
xfs_bumplink(tp, target_dp); |
f6bba2017 xfs: consolidate ... |
3203 3204 3205 |
} } else { /* target_ip != NULL */ /* |
f6bba2017 xfs: consolidate ... |
3206 3207 3208 3209 3210 3211 3212 3213 |
* Link the source inode under the target name. * If the source inode is a directory and we are moving * it across directories, its ".." entry will be * inconsistent until we replace that down below. * * In case there is already an entry with the same * name at the destination directory, remove it first. */ |
93597ae8d xfs: Fix deadlock... |
3214 3215 3216 3217 3218 3219 3220 3221 3222 3223 3224 3225 3226 3227 3228 |
/* * Check whether the replace operation will need to allocate * blocks. This happens when the shortform directory lacks * space and we have to convert it to a block format directory. * When more blocks are necessary, we must lock the AGI first * to preserve locking order (AGI -> AGF). */ if (xfs_dir2_sf_replace_needblock(target_dp, src_ip->i_ino)) { error = xfs_read_agi(mp, tp, XFS_INO_TO_AGNO(mp, target_ip->i_ino), &agibp); if (error) goto out_trans_cancel; } |
f6bba2017 xfs: consolidate ... |
3229 |
error = xfs_dir_replace(tp, target_dp, target_name, |
381eee69f xfs: remove first... |
3230 |
src_ip->i_ino, spaceres); |
f6bba2017 xfs: consolidate ... |
3231 |
if (error) |
c8eac49ef xfs: remove all b... |
3232 |
goto out_trans_cancel; |
f6bba2017 xfs: consolidate ... |
3233 3234 3235 3236 3237 3238 3239 3240 3241 3242 |
xfs_trans_ichgtime(tp, target_dp, XFS_ICHGTIME_MOD | XFS_ICHGTIME_CHG); /* * Decrement the link count on the target since the target * dir no longer points to it. */ error = xfs_droplink(tp, target_ip); if (error) |
c8eac49ef xfs: remove all b... |
3243 |
goto out_trans_cancel; |
f6bba2017 xfs: consolidate ... |
3244 3245 3246 3247 3248 3249 3250 |
if (src_is_directory) { /* * Drop the link from the old "." entry. */ error = xfs_droplink(tp, target_ip); if (error) |
c8eac49ef xfs: remove all b... |
3251 |
goto out_trans_cancel; |
f6bba2017 xfs: consolidate ... |
3252 3253 3254 3255 3256 3257 3258 3259 3260 3261 3262 3263 |
} } /* target_ip != NULL */ /* * Remove the source. */ if (new_parent && src_is_directory) { /* * Rewrite the ".." entry to point to the new * directory. */ error = xfs_dir_replace(tp, src_ip, &xfs_name_dotdot, |
381eee69f xfs: remove first... |
3264 |
target_dp->i_ino, spaceres); |
2451337dd xfs: global error... |
3265 |
ASSERT(error != -EEXIST); |
f6bba2017 xfs: consolidate ... |
3266 |
if (error) |
c8eac49ef xfs: remove all b... |
3267 |
goto out_trans_cancel; |
f6bba2017 xfs: consolidate ... |
3268 3269 3270 3271 3272 3273 3274 3275 3276 3277 3278 3279 3280 3281 3282 3283 3284 3285 3286 3287 3288 3289 3290 3291 3292 |
} /* * We always want to hit the ctime on the source inode. * * This isn't strictly required by the standards since the source * inode isn't really being changed, but old unix file systems did * it and some incremental backup programs won't work without it. */ xfs_trans_ichgtime(tp, src_ip, XFS_ICHGTIME_CHG); xfs_trans_log_inode(tp, src_ip, XFS_ILOG_CORE); /* * Adjust the link count on src_dp. This is necessary when * renaming a directory, either within one parent when * the target existed, or across two parent directories. */ if (src_is_directory && (new_parent || target_ip != NULL)) { /* * Decrement link count on src_directory since the * entry that's moved no longer points to it. */ error = xfs_droplink(tp, src_dp); if (error) |
c8eac49ef xfs: remove all b... |
3293 |
goto out_trans_cancel; |
f6bba2017 xfs: consolidate ... |
3294 |
} |
7dcf5c3e4 xfs: add RENAME_W... |
3295 3296 3297 3298 3299 3300 3301 |
/* * For whiteouts, we only need to update the source dirent with the * inode number of the whiteout inode rather than removing it * altogether. */ if (wip) { error = xfs_dir_replace(tp, src_dp, src_name, wip->i_ino, |
381eee69f xfs: remove first... |
3302 |
spaceres); |
7dcf5c3e4 xfs: add RENAME_W... |
3303 3304 |
} else error = xfs_dir_removename(tp, src_dp, src_name, src_ip->i_ino, |
381eee69f xfs: remove first... |
3305 |
spaceres); |
f6bba2017 xfs: consolidate ... |
3306 |
if (error) |
c8eac49ef xfs: remove all b... |
3307 |
goto out_trans_cancel; |
f6bba2017 xfs: consolidate ... |
3308 |
|
f6bba2017 xfs: consolidate ... |
3309 3310 3311 3312 |
xfs_trans_ichgtime(tp, src_dp, XFS_ICHGTIME_MOD | XFS_ICHGTIME_CHG); xfs_trans_log_inode(tp, src_dp, XFS_ILOG_CORE); if (new_parent) xfs_trans_log_inode(tp, target_dp, XFS_ILOG_CORE); |
f6bba2017 xfs: consolidate ... |
3313 |
|
c9cfdb381 xfs: remove dfops... |
3314 |
error = xfs_finish_rename(tp); |
7dcf5c3e4 xfs: add RENAME_W... |
3315 |
if (wip) |
44a8736bd xfs: clean up IRE... |
3316 |
xfs_irele(wip); |
7dcf5c3e4 xfs: add RENAME_W... |
3317 |
return error; |
f6bba2017 xfs: consolidate ... |
3318 |
|
445883e81 xfs: cleanup xfs_... |
3319 |
out_trans_cancel: |
4906e2154 xfs: remove the f... |
3320 |
xfs_trans_cancel(tp); |
253f4911f xfs: better xfs_t... |
3321 |
out_release_wip: |
7dcf5c3e4 xfs: add RENAME_W... |
3322 |
if (wip) |
44a8736bd xfs: clean up IRE... |
3323 |
xfs_irele(wip); |
f6bba2017 xfs: consolidate ... |
3324 3325 |
return error; } |
e6187b344 xfs: rename xfs_i... |
3326 3327 |
static int xfs_iflush( |
93848a999 xfs: add version ... |
3328 3329 |
struct xfs_inode *ip, struct xfs_buf *bp) |
1da177e4c Linux-2.6.12-rc2 |
3330 |
{ |
93848a999 xfs: add version ... |
3331 3332 3333 |
struct xfs_inode_log_item *iip = ip->i_itemp; struct xfs_dinode *dip; struct xfs_mount *mp = ip->i_mount; |
f20192991 xfs: simplify ino... |
3334 |
int error; |
1da177e4c Linux-2.6.12-rc2 |
3335 |
|
579aa9caf [XFS] shrink mrlo... |
3336 |
ASSERT(xfs_isilocked(ip, XFS_ILOCK_EXCL|XFS_ILOCK_SHARED)); |
718ecc503 xfs: xfs_iflock i... |
3337 |
ASSERT(xfs_iflags_test(ip, XFS_IFLUSHING)); |
f7e67b20e xfs: move the for... |
3338 |
ASSERT(ip->i_df.if_format != XFS_DINODE_FMT_BTREE || |
daf83964a xfs: move the per... |
3339 |
ip->i_df.if_nextents > XFS_IFORK_MAXEXT(ip, XFS_DATA_FORK)); |
90c60e164 xfs: xfs_iflush()... |
3340 |
ASSERT(iip->ili_item.li_buf == bp); |
1da177e4c Linux-2.6.12-rc2 |
3341 |
|
88ee2df7f xfs: return a voi... |
3342 |
dip = xfs_buf_offset(bp, ip->i_imap.im_boffset); |
1da177e4c Linux-2.6.12-rc2 |
3343 |
|
f20192991 xfs: simplify ino... |
3344 3345 3346 3347 3348 3349 3350 |
/* * We don't flush the inode if any of the following checks fail, but we * do still update the log item and attach to the backing buffer as if * the flush happened. This is a formality to facilitate predictable * error handling as the caller will shutdown and fail the buffer. */ error = -EFSCORRUPTED; |
69ef921b5 xfs: byteswap con... |
3351 |
if (XFS_TEST_ERROR(dip->di_magic != cpu_to_be16(XFS_DINODE_MAGIC), |
9e24cfd04 xfs: remove unnee... |
3352 |
mp, XFS_ERRTAG_IFLUSH_1)) { |
6a19d9393 xfs: convert xfs_... |
3353 |
xfs_alert_tag(mp, XFS_PTAG_IFLUSH, |
c96900435 xfs: use %px for ... |
3354 |
"%s: Bad inode %Lu magic number 0x%x, ptr "PTR_FMT, |
6a19d9393 xfs: convert xfs_... |
3355 |
__func__, ip->i_ino, be16_to_cpu(dip->di_magic), dip); |
f20192991 xfs: simplify ino... |
3356 |
goto flush_out; |
1da177e4c Linux-2.6.12-rc2 |
3357 |
} |
c19b3b05a xfs: mode di_mode... |
3358 |
if (S_ISREG(VFS_I(ip)->i_mode)) { |
1da177e4c Linux-2.6.12-rc2 |
3359 |
if (XFS_TEST_ERROR( |
f7e67b20e xfs: move the for... |
3360 3361 |
ip->i_df.if_format != XFS_DINODE_FMT_EXTENTS && ip->i_df.if_format != XFS_DINODE_FMT_BTREE, |
9e24cfd04 xfs: remove unnee... |
3362 |
mp, XFS_ERRTAG_IFLUSH_3)) { |
6a19d9393 xfs: convert xfs_... |
3363 |
xfs_alert_tag(mp, XFS_PTAG_IFLUSH, |
c96900435 xfs: use %px for ... |
3364 |
"%s: Bad regular inode %Lu, ptr "PTR_FMT, |
6a19d9393 xfs: convert xfs_... |
3365 |
__func__, ip->i_ino, ip); |
f20192991 xfs: simplify ino... |
3366 |
goto flush_out; |
1da177e4c Linux-2.6.12-rc2 |
3367 |
} |
c19b3b05a xfs: mode di_mode... |
3368 |
} else if (S_ISDIR(VFS_I(ip)->i_mode)) { |
1da177e4c Linux-2.6.12-rc2 |
3369 |
if (XFS_TEST_ERROR( |
f7e67b20e xfs: move the for... |
3370 3371 3372 |
ip->i_df.if_format != XFS_DINODE_FMT_EXTENTS && ip->i_df.if_format != XFS_DINODE_FMT_BTREE && ip->i_df.if_format != XFS_DINODE_FMT_LOCAL, |
9e24cfd04 xfs: remove unnee... |
3373 |
mp, XFS_ERRTAG_IFLUSH_4)) { |
6a19d9393 xfs: convert xfs_... |
3374 |
xfs_alert_tag(mp, XFS_PTAG_IFLUSH, |
c96900435 xfs: use %px for ... |
3375 |
"%s: Bad directory inode %Lu, ptr "PTR_FMT, |
6a19d9393 xfs: convert xfs_... |
3376 |
__func__, ip->i_ino, ip); |
f20192991 xfs: simplify ino... |
3377 |
goto flush_out; |
1da177e4c Linux-2.6.12-rc2 |
3378 3379 |
} } |
daf83964a xfs: move the per... |
3380 |
if (XFS_TEST_ERROR(ip->i_df.if_nextents + xfs_ifork_nextents(ip->i_afp) > |
9e24cfd04 xfs: remove unnee... |
3381 |
ip->i_d.di_nblocks, mp, XFS_ERRTAG_IFLUSH_5)) { |
6a19d9393 xfs: convert xfs_... |
3382 3383 |
xfs_alert_tag(mp, XFS_PTAG_IFLUSH, "%s: detected corrupt incore inode %Lu, " |
c96900435 xfs: use %px for ... |
3384 |
"total extents = %d, nblocks = %Ld, ptr "PTR_FMT, |
6a19d9393 xfs: convert xfs_... |
3385 |
__func__, ip->i_ino, |
daf83964a xfs: move the per... |
3386 |
ip->i_df.if_nextents + xfs_ifork_nextents(ip->i_afp), |
6a19d9393 xfs: convert xfs_... |
3387 |
ip->i_d.di_nblocks, ip); |
f20192991 xfs: simplify ino... |
3388 |
goto flush_out; |
1da177e4c Linux-2.6.12-rc2 |
3389 3390 |
} if (XFS_TEST_ERROR(ip->i_d.di_forkoff > mp->m_sb.sb_inodesize, |
9e24cfd04 xfs: remove unnee... |
3391 |
mp, XFS_ERRTAG_IFLUSH_6)) { |
6a19d9393 xfs: convert xfs_... |
3392 |
xfs_alert_tag(mp, XFS_PTAG_IFLUSH, |
c96900435 xfs: use %px for ... |
3393 |
"%s: bad inode %Lu, forkoff 0x%x, ptr "PTR_FMT, |
6a19d9393 xfs: convert xfs_... |
3394 |
__func__, ip->i_ino, ip->i_d.di_forkoff, ip); |
f20192991 xfs: simplify ino... |
3395 |
goto flush_out; |
1da177e4c Linux-2.6.12-rc2 |
3396 |
} |
e60896d8f xfs: di_flushiter... |
3397 |
|
1da177e4c Linux-2.6.12-rc2 |
3398 |
/* |
263997a68 xfs: turn NLINK f... |
3399 |
* Inode item log recovery for v2 inodes are dependent on the |
e60896d8f xfs: di_flushiter... |
3400 3401 3402 3403 3404 3405 |
* di_flushiter count for correct sequencing. We bump the flush * iteration count so we can detect flushes which postdate a log record * during recovery. This is redundant as we now log every change and * hence this can't happen but we need to still do it to ensure * backwards compatibility with old kernels that predate logging all * inode changes. |
1da177e4c Linux-2.6.12-rc2 |
3406 |
*/ |
6471e9c5e xfs: remove the d... |
3407 |
if (!xfs_sb_version_has_v3inode(&mp->m_sb)) |
e60896d8f xfs: di_flushiter... |
3408 |
ip->i_d.di_flushiter++; |
1da177e4c Linux-2.6.12-rc2 |
3409 |
|
0f45a1b20 xfs: improve loca... |
3410 3411 3412 3413 |
/* * If there are inline format data / attr forks attached to this inode, * make sure they are not corrupt. */ |
f7e67b20e xfs: move the for... |
3414 |
if (ip->i_df.if_format == XFS_DINODE_FMT_LOCAL && |
0f45a1b20 xfs: improve loca... |
3415 3416 |
xfs_ifork_verify_local_data(ip)) goto flush_out; |
f7e67b20e xfs: move the for... |
3417 |
if (ip->i_afp && ip->i_afp->if_format == XFS_DINODE_FMT_LOCAL && |
0f45a1b20 xfs: improve loca... |
3418 |
xfs_ifork_verify_local_attr(ip)) |
f20192991 xfs: simplify ino... |
3419 |
goto flush_out; |
005c5db8f xfs: rework the i... |
3420 |
|
1da177e4c Linux-2.6.12-rc2 |
3421 |
/* |
3987848c7 xfs: remove times... |
3422 3423 3424 |
* Copy the dirty parts of the inode into the on-disk inode. We always * copy out the core of the inode, because if the inode is dirty at all * the core must be. |
1da177e4c Linux-2.6.12-rc2 |
3425 |
*/ |
93f958f9c xfs: cull unneces... |
3426 |
xfs_inode_to_disk(ip, dip, iip->ili_item.li_lsn); |
1da177e4c Linux-2.6.12-rc2 |
3427 3428 3429 3430 |
/* Wrap, we never let the log put out DI_MAX_FLUSH */ if (ip->i_d.di_flushiter == DI_MAX_FLUSH) ip->i_d.di_flushiter = 0; |
005c5db8f xfs: rework the i... |
3431 3432 3433 |
xfs_iflush_fork(ip, dip, iip, XFS_DATA_FORK); if (XFS_IFORK_Q(ip)) xfs_iflush_fork(ip, dip, iip, XFS_ATTR_FORK); |
1da177e4c Linux-2.6.12-rc2 |
3434 3435 |
/* |
f5d8d5c4b xfs: split in-cor... |
3436 3437 3438 3439 3440 3441 3442 |
* We've recorded everything logged in the inode, so we'd like to clear * the ili_fields bits so we don't log and flush things unnecessarily. * However, we can't stop logging all this information until the data * we've copied into the disk buffer is written to disk. If we did we * might overwrite the copy of the inode in the log with all the data * after re-logging only part of it, and in the face of a crash we * wouldn't have all the data we need to recover. |
1da177e4c Linux-2.6.12-rc2 |
3443 |
* |
f5d8d5c4b xfs: split in-cor... |
3444 3445 |
* What we do is move the bits to the ili_last_fields field. When * logging the inode, these bits are moved back to the ili_fields field. |
664ffb8a4 xfs: move the buf... |
3446 3447 |
* In the xfs_buf_inode_iodone() routine we clear ili_last_fields, since * we know that the information those bits represent is permanently on |
f5d8d5c4b xfs: split in-cor... |
3448 3449 |
* disk. As long as the flush completes before the inode is logged * again, then both ili_fields and ili_last_fields will be cleared. |
1da177e4c Linux-2.6.12-rc2 |
3450 |
*/ |
f20192991 xfs: simplify ino... |
3451 3452 |
error = 0; flush_out: |
1319ebefd xfs: add an inode... |
3453 |
spin_lock(&iip->ili_lock); |
93848a999 xfs: add version ... |
3454 3455 |
iip->ili_last_fields = iip->ili_fields; iip->ili_fields = 0; |
fc0561cef xfs: optimise awa... |
3456 |
iip->ili_fsync_fields = 0; |
1319ebefd xfs: add an inode... |
3457 |
spin_unlock(&iip->ili_lock); |
1da177e4c Linux-2.6.12-rc2 |
3458 |
|
1319ebefd xfs: add an inode... |
3459 3460 |
/* * Store the current LSN of the inode so that we can tell whether the |
664ffb8a4 xfs: move the buf... |
3461 |
* item has moved in the AIL from xfs_buf_inode_iodone(). |
1319ebefd xfs: add an inode... |
3462 |
*/ |
93848a999 xfs: add version ... |
3463 3464 |
xfs_trans_ail_copy_lsn(mp->m_ail, &iip->ili_flush_lsn, &iip->ili_item.li_lsn); |
1da177e4c Linux-2.6.12-rc2 |
3465 |
|
93848a999 xfs: add version ... |
3466 3467 |
/* generate the checksum. */ xfs_dinode_calc_crc(mp, dip); |
f20192991 xfs: simplify ino... |
3468 |
return error; |
1da177e4c Linux-2.6.12-rc2 |
3469 |
} |
44a8736bd xfs: clean up IRE... |
3470 |
|
e6187b344 xfs: rename xfs_i... |
3471 3472 3473 3474 3475 3476 3477 |
/* * Non-blocking flush of dirty inode metadata into the backing buffer. * * The caller must have a reference to the inode and hold the cluster buffer * locked. The function will walk across all the inodes on the cluster buffer it * can find and lock without blocking, and flush them to the cluster buffer. * |
5717ea4d5 xfs: rework xfs_i... |
3478 3479 3480 3481 3482 |
* On successful flushing of at least one inode, the caller must write out the * buffer and release it. If no inodes are flushed, -EAGAIN will be returned and * the caller needs to release the buffer. On failure, the filesystem will be * shut down, the buffer will have been unlocked and released, and EFSCORRUPTED * will be returned. |
e6187b344 xfs: rename xfs_i... |
3483 3484 3485 |
*/ int xfs_iflush_cluster( |
e6187b344 xfs: rename xfs_i... |
3486 3487 |
struct xfs_buf *bp) { |
5717ea4d5 xfs: rework xfs_i... |
3488 3489 3490 3491 |
struct xfs_mount *mp = bp->b_mount; struct xfs_log_item *lip, *n; struct xfs_inode *ip; struct xfs_inode_log_item *iip; |
e6187b344 xfs: rename xfs_i... |
3492 |
int clcount = 0; |
5717ea4d5 xfs: rework xfs_i... |
3493 |
int error = 0; |
e6187b344 xfs: rename xfs_i... |
3494 |
|
5717ea4d5 xfs: rework xfs_i... |
3495 3496 3497 3498 3499 3500 3501 |
/* * We must use the safe variant here as on shutdown xfs_iflush_abort() * can remove itself from the list. */ list_for_each_entry_safe(lip, n, &bp->b_li_list, li_bio_list) { iip = (struct xfs_inode_log_item *)lip; ip = iip->ili_inode; |
e6187b344 xfs: rename xfs_i... |
3502 3503 |
/* |
5717ea4d5 xfs: rework xfs_i... |
3504 |
* Quick and dirty check to avoid locks if possible. |
e6187b344 xfs: rename xfs_i... |
3505 |
*/ |
718ecc503 xfs: xfs_iflock i... |
3506 |
if (__xfs_iflags_test(ip, XFS_IRECLAIM | XFS_IFLUSHING)) |
5717ea4d5 xfs: rework xfs_i... |
3507 3508 |
continue; if (xfs_ipincount(ip)) |
e6187b344 xfs: rename xfs_i... |
3509 |
continue; |
e6187b344 xfs: rename xfs_i... |
3510 3511 |
/* |
5717ea4d5 xfs: rework xfs_i... |
3512 3513 3514 3515 3516 |
* The inode is still attached to the buffer, which means it is * dirty but reclaim might try to grab it. Check carefully for * that, and grab the ilock while still holding the i_flags_lock * to guarantee reclaim will not be able to reclaim this inode * once we drop the i_flags_lock. |
e6187b344 xfs: rename xfs_i... |
3517 |
*/ |
5717ea4d5 xfs: rework xfs_i... |
3518 3519 |
spin_lock(&ip->i_flags_lock); ASSERT(!__xfs_iflags_test(ip, XFS_ISTALE)); |
718ecc503 xfs: xfs_iflock i... |
3520 |
if (__xfs_iflags_test(ip, XFS_IRECLAIM | XFS_IFLUSHING)) { |
5717ea4d5 xfs: rework xfs_i... |
3521 3522 |
spin_unlock(&ip->i_flags_lock); continue; |
e6187b344 xfs: rename xfs_i... |
3523 |
} |
e6187b344 xfs: rename xfs_i... |
3524 3525 |
/* |
5717ea4d5 xfs: rework xfs_i... |
3526 3527 |
* ILOCK will pin the inode against reclaim and prevent * concurrent transactions modifying the inode while we are |
718ecc503 xfs: xfs_iflock i... |
3528 3529 |
* flushing the inode. If we get the lock, set the flushing * state before we drop the i_flags_lock. |
e6187b344 xfs: rename xfs_i... |
3530 |
*/ |
5717ea4d5 xfs: rework xfs_i... |
3531 3532 |
if (!xfs_ilock_nowait(ip, XFS_ILOCK_SHARED)) { spin_unlock(&ip->i_flags_lock); |
e6187b344 xfs: rename xfs_i... |
3533 |
continue; |
5717ea4d5 xfs: rework xfs_i... |
3534 |
} |
718ecc503 xfs: xfs_iflock i... |
3535 |
__xfs_iflags_set(ip, XFS_IFLUSHING); |
5717ea4d5 xfs: rework xfs_i... |
3536 |
spin_unlock(&ip->i_flags_lock); |
e6187b344 xfs: rename xfs_i... |
3537 3538 |
/* |
5717ea4d5 xfs: rework xfs_i... |
3539 3540 3541 3542 3543 |
* Abort flushing this inode if we are shut down because the * inode may not currently be in the AIL. This can occur when * log I/O failure unpins the inode without inserting into the * AIL, leaving a dirty/unpinned inode attached to the buffer * that otherwise looks like it should be flushed. |
e6187b344 xfs: rename xfs_i... |
3544 |
*/ |
5717ea4d5 xfs: rework xfs_i... |
3545 3546 |
if (XFS_FORCED_SHUTDOWN(mp)) { xfs_iunpin_wait(ip); |
5717ea4d5 xfs: rework xfs_i... |
3547 3548 3549 |
xfs_iflush_abort(ip); xfs_iunlock(ip, XFS_ILOCK_SHARED); error = -EIO; |
e6187b344 xfs: rename xfs_i... |
3550 3551 |
continue; } |
5717ea4d5 xfs: rework xfs_i... |
3552 3553 |
/* don't block waiting on a log force to unpin dirty inodes */ if (xfs_ipincount(ip)) { |
718ecc503 xfs: xfs_iflock i... |
3554 |
xfs_iflags_clear(ip, XFS_IFLUSHING); |
5717ea4d5 xfs: rework xfs_i... |
3555 3556 |
xfs_iunlock(ip, XFS_ILOCK_SHARED); continue; |
e6187b344 xfs: rename xfs_i... |
3557 |
} |
e6187b344 xfs: rename xfs_i... |
3558 |
|
5717ea4d5 xfs: rework xfs_i... |
3559 3560 3561 |
if (!xfs_inode_clean(ip)) error = xfs_iflush(ip, bp); else |
718ecc503 xfs: xfs_iflock i... |
3562 |
xfs_iflags_clear(ip, XFS_IFLUSHING); |
5717ea4d5 xfs: rework xfs_i... |
3563 3564 3565 3566 |
xfs_iunlock(ip, XFS_ILOCK_SHARED); if (error) break; clcount++; |
e6187b344 xfs: rename xfs_i... |
3567 |
} |
e6187b344 xfs: rename xfs_i... |
3568 3569 3570 3571 |
if (error) { bp->b_flags |= XBF_ASYNC; xfs_buf_ioend_fail(bp); xfs_force_shutdown(mp, SHUTDOWN_CORRUPT_INCORE); |
5717ea4d5 xfs: rework xfs_i... |
3572 |
return error; |
e6187b344 xfs: rename xfs_i... |
3573 |
} |
5717ea4d5 xfs: rework xfs_i... |
3574 3575 3576 3577 3578 3579 3580 |
if (!clcount) return -EAGAIN; XFS_STATS_INC(mp, xs_icluster_flushcnt); XFS_STATS_ADD(mp, xs_icluster_flushinode, clcount); return 0; |
e6187b344 xfs: rename xfs_i... |
3581 |
} |
44a8736bd xfs: clean up IRE... |
3582 3583 3584 3585 3586 3587 3588 3589 |
/* Release an inode. */ void xfs_irele( struct xfs_inode *ip) { trace_xfs_irele(ip, _RET_IP_); iput(VFS_I(ip)); } |
54fbdd103 xfs: factor out a... |
3590 3591 3592 3593 3594 3595 3596 3597 3598 3599 3600 3601 3602 3603 3604 3605 3606 3607 3608 |
/* * Ensure all commited transactions touching the inode are written to the log. */ int xfs_log_force_inode( struct xfs_inode *ip) { xfs_lsn_t lsn = 0; xfs_ilock(ip, XFS_ILOCK_SHARED); if (xfs_ipincount(ip)) lsn = ip->i_itemp->ili_last_lsn; xfs_iunlock(ip, XFS_ILOCK_SHARED); if (!lsn) return 0; return xfs_log_force_lsn(ip->i_mount, lsn, XFS_LOG_SYNC, NULL); } |
e2aaee9cd xfs: move helpers... |
3609 3610 3611 3612 3613 3614 3615 3616 3617 3618 3619 3620 3621 3622 3623 3624 3625 3626 3627 3628 3629 3630 3631 3632 3633 3634 3635 3636 3637 3638 3639 3640 3641 3642 3643 3644 3645 3646 3647 3648 3649 3650 3651 3652 3653 3654 3655 3656 3657 3658 3659 3660 3661 3662 3663 3664 3665 3666 3667 3668 3669 3670 3671 3672 3673 3674 3675 3676 3677 3678 3679 3680 3681 3682 3683 3684 3685 3686 3687 3688 3689 3690 3691 3692 3693 3694 3695 3696 3697 3698 3699 3700 3701 |
/* * Grab the exclusive iolock for a data copy from src to dest, making sure to * abide vfs locking order (lowest pointer value goes first) and breaking the * layout leases before proceeding. The loop is needed because we cannot call * the blocking break_layout() with the iolocks held, and therefore have to * back out both locks. */ static int xfs_iolock_two_inodes_and_break_layout( struct inode *src, struct inode *dest) { int error; if (src > dest) swap(src, dest); retry: /* Wait to break both inodes' layouts before we start locking. */ error = break_layout(src, true); if (error) return error; if (src != dest) { error = break_layout(dest, true); if (error) return error; } /* Lock one inode and make sure nobody got in and leased it. */ inode_lock(src); error = break_layout(src, false); if (error) { inode_unlock(src); if (error == -EWOULDBLOCK) goto retry; return error; } if (src == dest) return 0; /* Lock the other inode and make sure nobody got in and leased it. */ inode_lock_nested(dest, I_MUTEX_NONDIR2); error = break_layout(dest, false); if (error) { inode_unlock(src); inode_unlock(dest); if (error == -EWOULDBLOCK) goto retry; return error; } return 0; } /* * Lock two inodes so that userspace cannot initiate I/O via file syscalls or * mmap activity. */ int xfs_ilock2_io_mmap( struct xfs_inode *ip1, struct xfs_inode *ip2) { int ret; ret = xfs_iolock_two_inodes_and_break_layout(VFS_I(ip1), VFS_I(ip2)); if (ret) return ret; if (ip1 == ip2) xfs_ilock(ip1, XFS_MMAPLOCK_EXCL); else xfs_lock_two_inodes(ip1, XFS_MMAPLOCK_EXCL, ip2, XFS_MMAPLOCK_EXCL); return 0; } /* Unlock both inodes to allow IO and mmap activity. */ void xfs_iunlock2_io_mmap( struct xfs_inode *ip1, struct xfs_inode *ip2) { bool same_inode = (ip1 == ip2); xfs_iunlock(ip2, XFS_MMAPLOCK_EXCL); if (!same_inode) xfs_iunlock(ip1, XFS_MMAPLOCK_EXCL); inode_unlock(VFS_I(ip2)); if (!same_inode) inode_unlock(VFS_I(ip1)); } |