blob: 029f856c56c04d1c4d660c13ed7e668c82fa5711 [file] [log] [blame]
Linus Torvalds1da177e2005-04-16 15:20:36 -07001ToDo/Notes:
2 - Find and fix bugs.
Linus Torvalds1da177e2005-04-16 15:20:36 -07003 - In between ntfs_prepare/commit_write, need exclusion between
Anton Altaparmakov66129f82004-11-11 12:34:00 +00004 simultaneous file extensions. This is given to us by holding i_sem
5 on the inode. The only places in the kernel when a file is resized
6 are prepare/commit write and truncate for both of which i_sem is
7 held. Just have to be careful in readpage/writepage and all other
8 helpers not running under i_sem that we play nice...
9 Also need to be careful with initialized_size extention in
Linus Torvalds1da177e2005-04-16 15:20:36 -070010 ntfs_prepare_write. Basically, just be _very_ careful in this code...
Anton Altaparmakov66129f82004-11-11 12:34:00 +000011 UPDATE: The only things that need to be checked are read/writepage
12 which do not hold i_sem. Note writepage cannot change i_size but it
13 needs to cope with a concurrent i_size change, just like readpage.
14 Also both need to cope with concurrent changes to the other sizes,
15 i.e. initialized/allocated/compressed size, as well.
Linus Torvalds1da177e2005-04-16 15:20:36 -070016 - Implement mft.c::sync_mft_mirror_umount(). We currently will just
17 leave the volume dirty on umount if the final iput(vol->mft_ino)
18 causes a write of any mirrored mft records due to the mft mirror
19 inode having been discarded already. Whether this can actually ever
20 happen is unclear however so it is worth waiting until someone hits
21 the problem.
22 - Enable the code for setting the NT4 compatibility flag when we start
23 making NTFS 1.2 specific modifications.
24
Anton Altaparmakove7a10332005-09-08 16:12:28 +0100252.1.24-WIP
26
27 - Support journals ($LogFile) which have been modified by chkdsk. This
28 means users can boot into Windows after we marked the volume dirty.
29 The Windows boot will run chkdsk and then reboot. The user can then
30 immediately boot into Linux rather than having to do a full Windows
31 boot first before rebooting into Linux and we will recognize such a
32 journal and empty it as it is clean by definition.
33 - Support journals ($LogFile) with only one restart page as well as
34 journals with two different restart pages. We sanity check both and
35 either use the only sane one or the more recent one of the two in the
36 case that both are valid.
Anton Altaparmakov06d0e3c2005-09-08 16:28:25 +010037 - Modify fs/ntfs/malloc.h::ntfs_malloc_nofs() to do the kmalloc() based
38 allocations with __GFP_HIGHMEM, analogous to how the vmalloc() based
39 allocations are done.
40 - Add fs/ntfs/malloc.h::ntfs_malloc_nofs_nofail() which is analogous to
41 ntfs_malloc_nofs() but it performs allocations with __GFP_NOFAIL and
42 hence cannot fail.
Anton Altaparmakov9529d462005-09-08 16:33:12 +010043 - Use ntfs_malloc_nofs_nofail() in the two critical regions in
44 fs/ntfs/runlist.c::ntfs_runlists_merge(). This means we no longer
45 need to panic() if the allocation fails as it now cannot fail.
Anton Altaparmakov84d6ebe2005-09-08 16:46:55 +010046 - Fix two nasty runlist merging bugs that had gone unnoticed so far.
47 Thanks to Stefano Picerno for the bug report.
Anton Altaparmakov8bb73522005-09-08 16:48:28 +010048 - Remove two bogus BUG_ON()s from fs/ntfs/mft.c.
Anton Altaparmakov2b0ada22005-09-08 16:52:31 +010049 - Fix handling of valid but empty mapping pairs array in
50 fs/ntfs/runlist.c::ntfs_mapping_pairs_decompress().
Anton Altaparmakovf94ad382005-09-08 17:04:11 +010051 - Report unrepresentable inodes during ntfs_readdir() as KERN_WARNING
52 messages and include the inode number. Thanks to Yura Pakhuchiy for
53 pointing this out.
Anton Altaparmakov3ffc5a42005-09-08 20:23:06 +010054 - Change ntfs_rl_truncate_nolock() to throw away the runlist if the new
55 length is zero.
Anton Altaparmakov6e483212005-09-08 20:26:34 +010056 - Add runlist.[hc]::ntfs_rl_punch_nolock() which punches a caller
57 specified hole into a runlist.
Anton Altaparmakov8e08cea2005-09-08 20:29:50 +010058 - Fix a bug in fs/ntfs/index.c::ntfs_index_lookup(). When the returned
59 index entry is in the index root, we forgot to set the @ir pointer in
60 the index context. Thanks to Yura Pakhuchiy for finding this bug.
Anton Altaparmakovf25dfb52005-09-08 20:35:33 +010061 - Remove bogus setting of PageError in ntfs_read_compressed_block().
Anton Altaparmakov0aaccea2005-09-08 20:40:32 +010062 - Add fs/ntfs/attrib.[hc]::ntfs_resident_attr_value_resize().
Anton Altaparmakov2983d1b2005-09-08 20:56:09 +010063 - Fix a bug in ntfs_map_runlist_nolock() where we forgot to protect
64 access to the allocated size in the ntfs inode with the size lock.
65 - Fix ntfs_attr_vcn_to_lcn_nolock() and ntfs_attr_find_vcn_nolock() to
66 return LCN_ENOENT when there is no runlist and the allocated size is
67 zero.
68 - Fix load_attribute_list() to handle the case of a NULL runlist.
Anton Altaparmakov807c4532005-09-08 21:01:17 +010069 - Fix handling of sparse attributes in ntfs_attr_make_non_resident().
70 - Add BUG() checks to ntfs_attr_make_non_resident() and ntfs_attr_set()
71 to ensure that these functions are never called for compressed or
72 encrypted attributes.
Anton Altaparmakovbbf18132005-09-08 21:09:06 +010073 - Fix cluster (de)allocators to work when the runlist is NULL and more
74 importantly to take a locked runlist rather than them locking it
75 which leads to lock reversal.
Anton Altaparmakov1c7d4692005-09-08 21:15:09 +010076 - Truncate {a,c,m}time to the ntfs supported time granularity when
77 updating the times in the inode in ntfs_setattr().
Anton Altaparmakov67bb1032005-09-08 21:19:45 +010078 - Fixup handling of sparse, compressed, and encrypted attributes in
Anton Altaparmakov311120e2005-09-08 22:04:20 +010079 fs/ntfs/inode.c::ntfs_read_locked_{,attr_,index_}inode(),
80 fs/ntfs/aops.c::ntfs_{read,write}page().
Anton Altaparmakov8dcdeba2005-09-08 21:25:48 +010081 - Make ntfs_write_block() not instantiate sparse blocks if they contain
82 only zeroes.
Anton Altaparmakov54b02eb2005-09-08 21:43:47 +010083 - Optimize fs/ntfs/aops.c::ntfs_write_block() by extending the page
84 lock protection over the buffer submission for i/o which allows the
85 removal of the get_bh()/put_bh() pairs for each buffer.
Anton Altaparmakov8273d5d2005-09-08 22:00:33 +010086 - Fix fs/ntfs/aops.c::ntfs_{read,write}_block() to handle the case
87 where a concurrent truncate has truncated the runlist under our feet.
Anton Altaparmakova01ac532005-09-08 22:08:11 +010088 - Fix page_has_buffers()/page_buffers() handling in fs/ntfs/aops.c.
Anton Altaparmakove7a10332005-09-08 16:12:28 +010089
Anton Altaparmakovaf859a42005-06-25 21:07:27 +0100902.1.23 - Implement extension of resident files and make writing safe as well as
91 many bug fixes, cleanups, and enhancements...
Linus Torvalds1da177e2005-04-16 15:20:36 -070092
93 - Add printk rate limiting for ntfs_warning() and ntfs_error() when
94 compiled without debug. This avoids a possible denial of service
95 attack. Thanks to Carl-Daniel Hailfinger from SuSE for pointing this
96 out.
Randy Dunlap89075472005-03-03 11:19:53 +000097 - Fix compilation warnings on ia64. (Randy Dunlap)
Anton Altaparmakovaf859a42005-06-25 21:07:27 +010098 - Use i_size_{read,write}() instead of reading i_size by hand and cache
99 the value where apropriate.
Anton Altaparmakov36763672004-11-18 13:46:45 +0000100 - Add size_lock to the ntfs_inode structure. This is an rw spinlock
101 and it locks against access to the inode sizes. Note, ->size_lock
102 is also accessed from irq context so you must use the _irqsave and
Anton Altaparmakovaf859a42005-06-25 21:07:27 +0100103 _irqrestore lock and unlock functions, respectively. Protect all
104 accesses to allocated_size, initialized_size, and compressed_size.
Anton Altaparmakov218357f2004-11-18 20:34:59 +0000105 - Minor optimization to fs/ntfs/super.c::ntfs_statfs() and its helpers.
Anton Altaparmakov946929d2005-01-13 15:26:29 +0000106 - Implement extension of resident files in the regular file write code
107 paths (fs/ntfs/aops.c::ntfs_{prepare,commit}_write()). At present
108 this only works until the data attribute becomes too big for the mft
109 record after which we abort the write returning -EOPNOTSUPP from
110 ntfs_prepare_write().
Anton Altaparmakovc002f422005-02-03 12:02:56 +0000111 - Add disable_sparse mount option together with a per volume sparse
112 enable bit which is set appropriately and a per inode sparse disable
113 bit which is preset on some system file inodes as appropriate.
114 - Enforce that sparse support is disabled on NTFS volumes pre 3.0.
Anton Altaparmakov1a0df152005-02-03 12:04:36 +0000115 - Fix a bug in fs/ntfs/runlist.c::ntfs_mapping_pairs_decompress() in
116 the creation of the unmapped runlist element for the base attribute
117 extent.
Anton Altaparmakovb6ad6c52005-02-15 10:08:43 +0000118 - Split ntfs_map_runlist() into ntfs_map_runlist() and a non-locking
119 helper ntfs_map_runlist_nolock() which is used by ntfs_map_runlist().
120 This allows us to map runlist fragments with the runlist lock already
121 held without having to drop and reacquire it around the call. Adapt
122 all callers.
123 - Change ntfs_find_vcn() to ntfs_find_vcn_nolock() which takes a locked
124 runlist. This allows us to find runlist elements with the runlist
125 lock already held without having to drop and reacquire it around the
126 call. Adapt all callers.
Anton Altaparmakov37e4c132005-02-18 10:03:13 +0000127 - Change time to u64 in time.h::ntfs2utc() as it otherwise generates a
Anton Altaparmakovd8ec7852005-02-18 09:23:39 +0000128 warning in the do_div() call on sparc32. Thanks to Meelis Roos for
129 the report and analysis of the warning.
Anton Altaparmakov37e4c132005-02-18 10:03:13 +0000130 - Fix a nasty runlist merge bug when merging two holes.
Anton Altaparmakov5ae9fcf2005-03-02 17:03:24 +0000131 - Set the ntfs_inode->allocated_size to the real allocated size in the
132 mft record for resident attributes (fs/ntfs/inode.c).
133 - Small readability cleanup to use "a" instead of "ctx->attr"
134 everywhere (fs/ntfs/inode.c).
Anton Altaparmakov41382682005-03-03 13:44:15 +0000135 - Make fs/ntfs/namei.c::ntfs_get_{parent,dentry} static and move the
136 definition of ntfs_export_ops from fs/ntfs/super.c to namei.c. Also,
137 declare ntfs_export_ops in fs/ntfs/ntfs.h.
Anton Altaparmakov9451f852005-03-03 14:43:43 +0000138 - Correct sparse file handling. The compressed values need to be
139 checked and set in the ntfs inode as done for compressed files and
140 the compressed size needs to be used for vfs inode->i_blocks instead
141 of the allocated size, again, as done for compressed files.
Anton Altaparmakov271849a2005-03-07 21:36:18 +0000142 - Add AT_EA in addition to AT_DATA to whitelist for being allowed to be
143 non-resident in fs/ntfs/attrib.c::ntfs_attr_can_be_non_resident().
144 - Add fs/ntfs/attrib.c::ntfs_attr_vcn_to_lcn_nolock() used by the new
145 write code.
Anton Altaparmakovc0c1cc02005-03-07 21:43:38 +0000146 - Fix bug in fs/ntfs/attrib.c::ntfs_find_vcn_nolock() where after
147 dropping the read lock and taking the write lock we were not checking
148 whether someone else did not already do the work we wanted to do.
149 - Rename fs/ntfs/attrib.c::ntfs_find_vcn_nolock() to
150 ntfs_attr_find_vcn_nolock() and update all callers.
Anton Altaparmakov2bfb4ff2005-03-09 15:15:06 +0000151 - Add fs/ntfs/attrib.[hc]::ntfs_attr_make_non_resident().
Anton Altaparmakov43b01fd2005-03-09 15:18:43 +0000152 - Fix sign of various error return values to be negative in
153 fs/ntfs/lcnalloc.c.
Anton Altaparmakov905685f2005-03-10 11:06:19 +0000154 - Modify ->readpage and ->writepage (fs/ntfs/aops.c) so they detect and
155 handle the case where an attribute is converted from resident to
156 non-resident by a concurrent file write.
Jesper Juhl251c8422005-04-04 14:59:56 +0100157 - Remove checks for NULL before calling kfree() since kfree() does the
158 checking itself. (Jesper Juhl)
Anton Altaparmakovb0d23742005-04-04 16:20:14 +0100159 - Some utilities modify the boot sector but do not update the checksum.
160 Thus, relax the checking in fs/ntfs/super.c::is_boot_sector_ntfs() to
161 only emit a warning when the checksum is incorrect rather than
162 refusing the mount. Thanks to Bernd Casimir for pointing this
163 problem out.
Anton Altaparmakovbb3cf332005-04-06 13:34:31 +0100164 - Update attribute definition handling.
Anton Altaparmakov7fafb8b2005-04-06 16:09:21 +0100165 - Add NTFS_MAX_CLUSTER_SIZE and NTFS_MAX_PAGES_PER_CLUSTER constants.
166 - Use NTFS_MAX_CLUSTER_SIZE in super.c instead of hard coding 0x10000.
Anton Altaparmakovd53ee322005-04-06 16:11:20 +0100167 - Use MAX_BUF_PER_PAGE instead of variable sized array allocation for
168 better code generation and one less sparse warning in fs/ntfs/aops.c.
Pekka Enberg2fb21db2005-05-25 21:15:34 +0300169 - Remove spurious void pointer casts from fs/ntfs/. (Pekka Enberg)
Anton Altaparmakov442d2072005-05-27 16:42:56 +0100170 - Use C99 style structure initialization after memory allocation where
171 possible (fs/ntfs/{attrib.c,index.c,super.c}). Thanks to Al Viro and
172 Pekka Enberg.
Anton Altaparmakov3f2faef2005-06-25 15:28:56 +0100173 - Stamp the transaction log ($UsnJrnl), aka user space journal, if it
174 is active on the volume and we are mounting read-write or remounting
175 from read-only to read-write.
Anton Altaparmakov9f993fe2005-06-25 16:15:36 +0100176 - Fix a bug in address space operations error recovery code paths where
177 if the runlist was not mapped at all and a mapping error occured we
178 would leave the runlist locked on exit to the function so that the
179 next access to the same file would try to take the lock and deadlock.
Anton Altaparmakovca8fd7a2005-06-25 16:31:27 +0100180 - Detect the case when Windows has been suspended to disk on the volume
181 to be mounted and if this is the case do not allow (re)mounting
182 read-write. This is done by parsing hiberfil.sys if present.
Anton Altaparmakov3bd1f4a2005-06-25 16:51:58 +0100183 - Fix several occurences of a bug where we would perform 'var & ~const'
184 with a 64-bit variable and a int, i.e. 32-bit, constant. This causes
185 the higher order 32-bits of the 64-bit variable to be zeroed. To fix
186 this cast the 'const' to the same 64-bit type as 'var'.
Anton Altaparmakov1d58b272005-06-25 17:04:55 +0100187 - Change the runlist terminator of the newly allocated cluster(s) to
188 LCN_ENOENT in ntfs_attr_make_non_resident(). Otherwise the runlist
189 code gets confused.
Anton Altaparmakovfa3be922005-06-25 17:15:36 +0100190 - Add an extra parameter @last_vcn to ntfs_get_size_for_mapping_pairs()
191 and ntfs_mapping_pairs_build() to allow the runlist encoding to be
192 partial which is desirable when filling holes in sparse attributes.
193 Update all callers.
Anton Altaparmakov4757d7df2005-06-25 17:24:08 +0100194 - Change ntfs_map_runlist_nolock() to only decompress the mapping pairs
195 if the requested vcn is inside it. Otherwise we get into problems
196 when we try to map an out of bounds vcn because we then try to map
197 the already mapped runlist fragment which causes
198 ntfs_mapping_pairs_decompress() to fail and return error. Update
199 ntfs_attr_find_vcn_nolock() accordingly.
Anton Altaparmakovba6d2372005-06-26 22:12:02 +0100200 - Fix a nasty deadlock that appeared in recent kernels.
201 The situation: VFS inode X on a mounted ntfs volume is dirty. For
202 same inode X, the ntfs_inode is dirty and thus corresponding on-disk
203 inode, i.e. mft record, which is in a dirty PAGE_CACHE_PAGE belonging
204 to the table of inodes, i.e. $MFT, inode 0.
205 What happens:
206 Process 1: sys_sync()/umount()/whatever... calls
207 __sync_single_inode() for $MFT -> do_writepages() -> write_page for
208 the dirty page containing the on-disk inode X, the page is now locked
209 -> ntfs_write_mst_block() which clears PageUptodate() on the page to
210 prevent anyone else getting hold of it whilst it does the write out.
211 This is necessary as the on-disk inode needs "fixups" applied before
212 the write to disk which are removed again after the write and
213 PageUptodate is then set again. It then analyses the page looking
214 for dirty on-disk inodes and when it finds one it calls
215 ntfs_may_write_mft_record() to see if it is safe to write this
216 on-disk inode. This then calls ilookup5() to check if the
217 corresponding VFS inode is in icache(). This in turn calls ifind()
218 which waits on the inode lock via wait_on_inode whilst holding the
219 global inode_lock.
220 Process 2: pdflush results in a call to __sync_single_inode for the
221 same VFS inode X on the ntfs volume. This locks the inode (I_LOCK)
222 then calls write-inode -> ntfs_write_inode -> map_mft_record() ->
223 read_cache_page() for the page (in page cache of table of inodes
224 $MFT, inode 0) containing the on-disk inode. This page has
225 PageUptodate() clear because of Process 1 (see above) so
226 read_cache_page() blocks when it tries to take the page lock for the
227 page so it can call ntfs_read_page().
228 Thus Process 1 is holding the page lock on the page containing the
229 on-disk inode X and it is waiting on the inode X to be unlocked in
230 ifind() so it can write the page out and then unlock the page.
231 And Process 2 is holding the inode lock on inode X and is waiting for
232 the page to be unlocked so it can call ntfs_readpage() or discover
233 that Process 1 set PageUptodate() again and use the page.
234 Thus we have a deadlock due to ifind() waiting on the inode lock.
235 The solution: The fix is to use the newly introduced
236 ilookup5_nowait() which does not wait on the inode's lock and hence
237 avoids the deadlock. This is safe as we do not care about the VFS
238 inode and only use the fact that it is in the VFS inode cache and the
239 fact that the vfs and ntfs inodes are one struct in memory to find
240 the ntfs inode in memory if present. Also, the ntfs inode has its
241 own locking so it does not matter if the vfs inode is locked.
Anton Altaparmakove74589a2005-08-16 16:38:28 +0100242 - Fix bug in mft record writing where we forgot to set the device in
Anton Altaparmakov481d0372005-08-16 19:42:56 +0100243 the buffers when mapping them after the VM had discarded them.
Anton Altaparmakove74589a2005-08-16 16:38:28 +0100244 Thanks to Martin MOKREJÅ  for the bug report.
Linus Torvalds1da177e2005-04-16 15:20:36 -0700245
2462.1.22 - Many bug and race fixes and error handling improvements.
247
248 - Improve error handling in fs/ntfs/inode.c::ntfs_truncate().
249 - Change fs/ntfs/inode.c::ntfs_truncate() to return an error code
250 instead of void and provide a helper ntfs_truncate_vfs() for the
251 vfs ->truncate method.
252 - Add a new ntfs inode flag NInoTruncateFailed() and modify
253 fs/ntfs/inode.c::ntfs_truncate() to set and clear it appropriately.
254 - Fix min_size and max_size definitions in ATTR_DEF structure in
255 fs/ntfs/layout.h to be signed.
256 - Add attribute definition handling helpers to fs/ntfs/attrib.[hc]:
257 ntfs_attr_size_bounds_check(), ntfs_attr_can_be_non_resident(), and
258 ntfs_attr_can_be_resident(), which in turn use the new private helper
259 ntfs_attr_find_in_attrdef().
260 - In fs/ntfs/aops.c::mark_ntfs_record_dirty(), take the
261 mapping->private_lock around the dirtying of the buffer heads
262 analagous to the way it is done in __set_page_dirty_buffers().
263 - Ensure the mft record size does not exceed the PAGE_CACHE_SIZE at
264 mount time as this cannot work with the current implementation.
265 - Check for location of attribute name and improve error handling in
266 general in fs/ntfs/inode.c::ntfs_read_locked_inode() and friends.
267 - In fs/ntfs/aops.c::ntfs_writepage(), if the page is fully outside
268 i_size, i.e. race with truncate, invalidate the buffers on the page
269 so that they become freeable and hence the page does not leak.
270 - Remove unused function fs/ntfs/runlist.c::ntfs_rl_merge(). (Adrian
271 Bunk)
272 - Fix stupid bug in fs/ntfs/attrib.c::ntfs_attr_find() that resulted in
273 a NULL pointer dereference in the error code path when a corrupt
274 attribute was found. (Thanks to Domen Puncer for the bug report.)
275 - Add MODULE_VERSION() to fs/ntfs/super.c.
276 - Make several functions and variables static. (Adrian Bunk)
277 - Modify fs/ntfs/aops.c::mark_ntfs_record_dirty() so it allocates
278 buffers for the page if they are not present and then marks the
279 buffers belonging to the ntfs record dirty. This causes the buffers
280 to become busy and hence they are safe from removal until the page
281 has been written out.
282 - Fix stupid bug in fs/ntfs/attrib.c::ntfs_external_attr_find() in the
283 error handling code path that resulted in a BUG() due to trying to
284 unmap an extent mft record when the mapping of it had failed and it
285 thus was not mapped. (Thanks to Ken MacFerrin for the bug report.)
286 - Drop the runlist lock after the vcn has been read in
287 fs/ntfs/lcnalloc.c::__ntfs_cluster_free().
288 - Rewrite handling of multi sector transfer errors. We now do not set
289 PageError() when such errors are detected in the async i/o handler
290 fs/ntfs/aops.c::ntfs_end_buffer_async_read(). All users of mst
291 protected attributes now check the magic of each ntfs record as they
292 use it and act appropriately. This has the effect of making errors
293 granular per ntfs record rather than per page which solves the case
294 where we cannot access any of the ntfs records in a page when a
295 single one of them had an mst error. (Thanks to Ken MacFerrin for
296 the bug report.)
297 - Fix error handling in fs/ntfs/quota.c::ntfs_mark_quotas_out_of_date()
298 where we failed to release i_sem on the $Quota/$Q attribute inode.
299 - Fix bug in handling of bad inodes in fs/ntfs/namei.c::ntfs_lookup().
300 - Add mapping of unmapped buffers to all remaining code paths, i.e.
301 fs/ntfs/aops.c::ntfs_write_mst_block(), mft.c::ntfs_sync_mft_mirror(),
302 and write_mft_record_nolock(). From now on we require that the
303 complete runlist for the mft mirror is always mapped into memory.
304 - Add creation of buffers to fs/ntfs/mft.c::ntfs_sync_mft_mirror().
305 - Improve error handling in fs/ntfs/aops.c::ntfs_{read,write}_block().
306 - Cleanup fs/ntfs/aops.c::ntfs_{read,write}page() since we know that a
307 resident attribute will be smaller than a page which makes the code
308 simpler. Also make the code more tolerant to concurrent ->truncate.
309
3102.1.21 - Fix some races and bugs, rewrite mft write code, add mft allocator.
311
312 - Implement extent mft record deallocation
313 fs/ntfs/mft.c::ntfs_extent_mft_record_free().
314 - Splitt runlist related functions off from attrib.[hc] to runlist.[hc].
315 - Add vol->mft_data_pos and initialize it at mount time.
316 - Rename init_runlist() to ntfs_init_runlist(), ntfs_vcn_to_lcn() to
317 ntfs_rl_vcn_to_lcn(), decompress_mapping_pairs() to
318 ntfs_mapping_pairs_decompress(), ntfs_merge_runlists() to
319 ntfs_runlists_merge() and adapt all callers.
320 - Add fs/ntfs/runlist.[hc]::ntfs_get_nr_significant_bytes(),
321 ntfs_get_size_for_mapping_pairs(), ntfs_write_significant_bytes(),
322 and ntfs_mapping_pairs_build(), adapted from libntfs.
323 - Make fs/ntfs/lcnalloc.c::ntfs_cluster_free_from_rl_nolock() not
324 static and add a declaration for it to lcnalloc.h.
325 - Add fs/ntfs/lcnalloc.h::ntfs_cluster_free_from_rl() which is a static
326 inline wrapper for ntfs_cluster_free_from_rl_nolock() which takes the
327 cluster bitmap lock for the duration of the call.
328 - Add fs/ntfs/attrib.[hc]::ntfs_attr_record_resize().
329 - Implement the equivalent of memset() for an ntfs attribute in
330 fs/ntfs/attrib.[hc]::ntfs_attr_set() and switch
331 fs/ntfs/logfile.c::ntfs_empty_logfile() to using it.
332 - Remove unnecessary casts from LCN_* constants.
333 - Implement fs/ntfs/runlist.c::ntfs_rl_truncate_nolock().
334 - Add MFT_RECORD_OLD as a copy of MFT_RECORD in fs/ntfs/layout.h and
335 change MFT_RECORD to contain the NTFS 3.1+ specific fields.
336 - Add a helper function fs/ntfs/aops.c::mark_ntfs_record_dirty() which
337 marks all buffers belonging to an ntfs record dirty, followed by
338 marking the page the ntfs record is in dirty and also marking the vfs
339 inode containing the ntfs record dirty (I_DIRTY_PAGES).
340 - Switch fs/ntfs/index.h::ntfs_index_entry_mark_dirty() to using the
341 new helper fs/ntfs/aops.c::mark_ntfs_record_dirty() and remove the no
342 longer needed fs/ntfs/index.[hc]::__ntfs_index_entry_mark_dirty().
343 - Move ntfs_{un,}map_page() from ntfs.h to aops.h and fix resulting
344 include errors.
345 - Move the typedefs for runlist_element and runlist from types.h to
346 runlist.h and fix resulting include errors.
347 - Remove unused {__,}format_mft_record() from fs/ntfs/mft.c.
348 - Modify fs/ntfs/mft.c::__mark_mft_record_dirty() to use the helper
349 mark_ntfs_record_dirty() which also changes the behaviour in that we
350 now set the buffers belonging to the mft record dirty as well as the
351 page itself.
352 - Update fs/ntfs/mft.c::write_mft_record_nolock() and sync_mft_mirror()
353 to cope with the fact that there now are dirty buffers in mft pages.
354 - Update fs/ntfs/inode.c::ntfs_write_inode() to also use the helper
355 mark_ntfs_record_dirty() and thus to set the buffers belonging to the
356 mft record dirty as well as the page itself.
357 - Fix compiler warnings on x86-64 in fs/ntfs/dir.c. (Randy Dunlap,
358 slightly modified by me)
359 - Add fs/ntfs/mft.c::try_map_mft_record() which fails with -EALREADY if
360 the mft record is already locked and otherwise behaves the same way
361 as fs/ntfs/mft.c::map_mft_record().
362 - Modify fs/ntfs/mft.c::write_mft_record_nolock() so that it only
363 writes the mft record if the buffers belonging to it are dirty.
364 Otherwise we assume that it was written out by other means already.
365 - Attempting to write outside initialized size is _not_ a bug so remove
366 the bug check from fs/ntfs/aops.c::ntfs_write_mst_block(). It is in
367 fact required to write outside initialized size when preparing to
368 extend the initialized size.
369 - Map the page instead of using page_address() before writing to it in
370 fs/ntfs/aops.c::ntfs_mft_writepage().
371 - Provide exclusion between opening an inode / mapping an mft record
372 and accessing the mft record in fs/ntfs/mft.c::ntfs_mft_writepage()
373 by setting the page not uptodate throughout ntfs_mft_writepage().
374 - Clear the page uptodate flag in fs/ntfs/aops.c::ntfs_write_mst_block()
375 to ensure noone can see the page whilst the mst fixups are applied.
376 - Add the helper fs/ntfs/mft.c::ntfs_may_write_mft_record() which
377 checks if an mft record may be written out safely obtaining any
378 necessary locks in the process. This is used by
379 fs/ntfs/aops.c::ntfs_write_mst_block().
380 - Modify fs/ntfs/aops.c::ntfs_write_mst_block() to also work for
381 writing mft records and improve its error handling in the process.
382 Now if any of the records in the page fail to be written out, all
383 other records will be written out instead of aborting completely.
384 - Remove ntfs_mft_aops and update all users to use ntfs_mst_aops.
385 - Modify fs/ntfs/inode.c::ntfs_read_locked_inode() to set the
386 ntfs_mst_aops for all inodes which are NInoMstProtected() and
387 ntfs_aops for all other inodes.
388 - Rename fs/ntfs/mft.c::sync_mft_mirror{,_umount}() to
389 ntfs_sync_mft_mirror{,_umount}() and change their parameters so they
390 no longer require an ntfs inode to be present. Update all callers.
391 - Cleanup the error handling in fs/ntfs/mft.c::ntfs_sync_mft_mirror().
392 - Clear the page uptodate flag in fs/ntfs/mft.c::ntfs_sync_mft_mirror()
393 to ensure noone can see the page whilst the mst fixups are applied.
394 - Remove the no longer needed fs/ntfs/mft.c::ntfs_mft_writepage() and
395 fs/ntfs/mft.c::try_map_mft_record().
396 - Fix callers of fs/ntfs/aops.c::mark_ntfs_record_dirty() to call it
397 with the ntfs inode which contains the page rather than the ntfs
398 inode the mft record of which is in the page.
399 - Fix race condition in fs/ntfs/inode.c::ntfs_put_inode() by moving the
400 index inode bitmap inode release code from there to
401 fs/ntfs/inode.c::ntfs_clear_big_inode(). (Thanks to Christoph
402 Hellwig for spotting this.)
403 - Fix race condition in fs/ntfs/inode.c::ntfs_put_inode() by taking the
404 inode semaphore around the code that sets ni->itype.index.bmp_ino to
405 NULL and reorganize the code to optimize it a bit. (Thanks to
406 Christoph Hellwig for spotting this.)
407 - Modify fs/ntfs/aops.c::mark_ntfs_record_dirty() to no longer take the
408 ntfs inode as a parameter as this is confusing and misleading and the
409 needed ntfs inode is available via NTFS_I(page->mapping->host).
410 Adapt all callers to this change.
411 - Modify fs/ntfs/mft.c::write_mft_record_nolock() and
412 fs/ntfs/aops.c::ntfs_write_mst_block() to only check the dirty state
413 of the first buffer in a record and to take this as the ntfs record
414 dirty state. We cannot look at the dirty state for subsequent
415 buffers because we might be racing with
416 fs/ntfs/aops.c::mark_ntfs_record_dirty().
417 - Move the static inline ntfs_init_big_inode() from fs/ntfs/inode.c to
418 inode.h and make fs/ntfs/inode.c::__ntfs_init_inode() non-static and
419 add a declaration for it to inode.h. Fix some compilation issues
420 that resulted due to #includes and header file interdependencies.
421 - Simplify setup of i_mode in fs/ntfs/inode.c::ntfs_read_locked_inode().
422 - Add helpers fs/ntfs/layout.h::MK_MREF() and MK_LE_MREF().
423 - Modify fs/ntfs/mft.c::map_extent_mft_record() to only verify the mft
424 record sequence number if it is specified (i.e. not zero).
425 - Add fs/ntfs/mft.[hc]::ntfs_mft_record_alloc() and various helper
426 functions used by it.
427 - Update Documentation/filesystems/ntfs.txt with instructions on how to
428 use the Device-Mapper driver with NTFS ftdisk/LDM raid. This removes
429 the linear raid problem with the Software RAID / MD driver when one
430 or more of the devices has an odd number of sectors.
431
4322.1.20 - Fix two stupid bugs introduced in 2.1.18 release.
433
434 - Fix stupid bug in fs/ntfs/attrib.c::ntfs_attr_reinit_search_ctx()
435 where we did not clear ctx->al_entry but it was still set due to
436 changes in ntfs_attr_lookup() and ntfs_external_attr_find() in
437 particular.
438 - Fix another stupid bug in fs/ntfs/attrib.c::ntfs_external_attr_find()
439 where we forgot to unmap the extent mft record when we had finished
440 enumerating an attribute which caused a bug check to trigger when the
441 VFS calls ->clear_inode.
442
4432.1.19 - Many cleanups, improvements, and a minor bug fix.
444
445 - Update ->setattr (fs/ntfs/inode.c::ntfs_setattr()) to refuse to
446 change the uid, gid, and mode of an inode as we do not support NTFS
447 ACLs yet.
448 - Remove BKL use from ntfs_setattr() syncing up with the rest of the
449 kernel.
450 - Get rid of the ugly transparent union in fs/ntfs/dir.c::ntfs_readdir()
451 and ntfs_filldir() as per suggestion from Al Viro.
452 - Change '\0' and L'\0' to simply 0 as per advice from Linus Torvalds.
453 - Update ->truncate (fs/ntfs/inode.c::ntfs_truncate()) to check if the
454 inode size has changed and to only output an error if so.
455 - Rename fs/ntfs/attrib.h::attribute_value_length() to ntfs_attr_size().
456 - Add le{16,32,64} as well as sle{16,32,64} data types to
457 fs/ntfs/types.h.
458 - Change ntfschar to be le16 instead of u16 in fs/ntfs/types.h.
459 - Add le versions of VCN, LCN, and LSN called leVCN, leLCN, and leLSN,
460 respectively, to fs/ntfs/types.h.
461 - Update endianness conversion macros in fs/ntfs/endian.h to use the
462 new types as appropriate.
463 - Do proper type casting when using sle64_to_cpup() in fs/ntfs/dir.c
464 and index.c.
465 - Add leMFT_REF data type to fs/ntfs/layout.h.
466 - Update all NTFS header files with the new little endian data types.
467 Affected files are fs/ntfs/layout.h, logfile.h, and time.h.
468 - Do proper type casting when using ntfs_is_*_recordp() in
469 fs/ntfs/logfile.c, mft.c, and super.c.
470 - Fix all the sparse bitwise warnings. Had to change all the typedef
471 enums storing little endian values to simple enums plus a typedef for
472 the datatype to make sparse happy.
473 - Fix a bug found by the new sparse bitwise warnings where the default
474 upcase table was defined as a pointer to wchar_t rather than ntfschar
475 in fs/ntfs/ntfs.h and super.c.
476 - Change {const_,}cpu_to_le{16,32}(0) to just 0 as suggested by Al Viro.
477
4782.1.18 - Fix scheduling latencies at mount time as well as an endianness bug.
479
480 - Remove vol->nr_mft_records as it was pretty meaningless and optimize
481 the calculation of total/free inodes as used by statfs().
482 - Fix scheduling latencies in ntfs_fill_super() by dropping the BKL
483 because the code itself is using the ntfs_lock semaphore which
484 provides safe locking. (Ingo Molnar)
485 - Fix a potential bug in fs/ntfs/mft.c::map_extent_mft_record() that
486 could occur in the future for when we start closing/freeing extent
487 inodes if we don't set base_ni->ext.extent_ntfs_inos to NULL after
488 we free it.
489 - Rename {find,lookup}_attr() to ntfs_attr_{find,lookup}() as well as
490 find_external_attr() to ntfs_external_attr_find() to cleanup the
491 namespace a bit and to be more consistent with libntfs.
492 - Rename {{re,}init,get,put}_attr_search_ctx() to
493 ntfs_attr_{{re,}init,get,put}_search_ctx() as well as the type
494 attr_search_context to ntfs_attr_search_ctx.
495 - Force use of ntfs_attr_find() in ntfs_attr_lookup() when searching
496 for the attribute list attribute itself.
497 - Fix endianness bug in ntfs_external_attr_find().
498 - Change ntfs_{external_,}attr_find() to return 0 on success, -ENOENT
499 if the attribute is not found, and -EIO on real error. In the case
500 of -ENOENT, the search context is updated to describe the attribute
501 before which the attribute being searched for would need to be
502 inserted if such an action were to be desired and in the case of
503 ntfs_external_attr_find() the search context is also updated to
504 indicate the attribute list entry before which the attribute list
505 entry of the attribute being searched for would need to be inserted
506 if such an action were to be desired. Also make ntfs_find_attr()
507 static and remove its prototype from attrib.h as it is not used
508 anywhere other than attrib.c. Update ntfs_attr_lookup() and all
509 callers of ntfs_{external,}attr_{find,lookup}() for the new return
510 values.
511 - Minor cleanup of fs/ntfs/inode.c::ntfs_init_locked_inode().
512
5132.1.17 - Fix bugs in mount time error code paths and other updates.
514
515 - Implement bitmap modification code (fs/ntfs/bitmap.[hc]). This
516 includes functions to set/clear a single bit or a run of bits.
517 - Add fs/ntfs/attrib.[hc]::ntfs_find_vcn() which returns the locked
518 runlist element containing a particular vcn. It also takes care of
519 mapping any needed runlist fragments.
520 - Implement cluster (de-)allocation code (fs/ntfs/lcnalloc.[hc]).
521 - Load attribute definition table from $AttrDef at mount time.
522 - Fix bugs in mount time error code paths involving (de)allocation of
523 the default and volume upcase tables.
524 - Remove ntfs_nr_mounts as it is no longer used.
525
5262.1.16 - Implement access time updates, file sync, async io, and read/writev.
527
528 - Add support for readv/writev and aio_read/aio_write (fs/ntfs/file.c).
529 This is done by setting the appropriate file operations pointers to
530 the generic helper functions provided by mm/filemap.c.
531 - Implement fsync, fdatasync, and msync both for files (fs/ntfs/file.c)
532 and directories (fs/ntfs/dir.c).
533 - Add support for {a,m,c}time updates to inode.c::ntfs_write_inode().
534 Note, except for the root directory and any other system files opened
535 by the user, the system files will not have their access times
536 updated as they are only accessed at the inode level an hence the
537 file level functions which cause the times to be updated are never
538 invoked.
539
5402.1.15 - Invalidate quotas when (re)mounting read-write.
541
542 - Add new element itype.index.collation_rule to the ntfs inode
543 structure and set it appropriately in ntfs_read_locked_inode().
544 - Implement a new inode type "index" to allow efficient access to the
545 indices found in various system files and adapt inode handling
546 accordingly (fs/ntfs/inode.[hc]). An index inode is essentially an
547 attribute inode (NInoAttr() is true) with an attribute type of
548 AT_INDEX_ALLOCATION. As such, it is no longer allowed to call
549 ntfs_attr_iget() with an attribute type of AT_INDEX_ALLOCATION as
550 there would be no way to distinguish between normal attribute inodes
551 and index inodes. The function to obtain an index inode is
552 ntfs_index_iget() and it uses the helper function
553 ntfs_read_locked_index_inode(). Note, we do not overload
554 ntfs_attr_iget() as indices consist of multiple attributes so using
555 ntfs_attr_iget() to obtain an index inode would be confusing.
556 - Ensure that there is no overflow when doing page->index <<
557 PAGE_CACHE_SHIFT by casting page->index to s64 in fs/ntfs/aops.c.
558 - Use atomic kmap instead of kmap() in fs/ntfs/aops.c::ntfs_read_page()
559 and ntfs_read_block().
560 - Use case sensitive attribute lookups instead of case insensitive ones.
561 - Lock all page cache pages belonging to mst protected attributes while
562 accessing them to ensure we never see corrupt data while the page is
563 under writeout.
564 - Add framework for generic ntfs collation (fs/ntfs/collation.[hc]).
565 We have ntfs_is_collation_rule_supported() to check if the collation
566 rule you want to use is supported and ntfs_collation() which actually
567 collates two data items. We currently only support COLLATION_BINARY
568 and COLLATION_NTOFS_ULONG but support for other collation rules will
569 be added as the need arises.
570 - Add a new type, ntfs_index_context, to allow retrieval of an index
571 entry using the corresponding index key. To get an index context,
572 use ntfs_index_ctx_get() and to release it, use ntfs_index_ctx_put().
573 This also adds a new slab cache for the index contexts. To lookup a
574 key in an index inode, use ntfs_index_lookup(). After modifying an
575 index entry, call ntfs_index_entry_flush_dcache_page() followed by
576 ntfs_index_entry_mark_dirty() to ensure the changes are written out
577 to disk. For details see fs/ntfs/index.[hc]. Note, at present, if
578 an index entry is in the index allocation attribute rather than the
579 index root attribute it will not be written out (you will get a
580 warning message about discarded changes instead).
581 - Load the quota file ($Quota) and check if quota tracking is enabled
582 and if so, mark the quotas out of date. This causes windows to
583 rescan the volume on boot and update all quota entries.
584 - Add a set_page_dirty address space operation for ntfs_m[fs]t_aops.
585 It is simply set to __set_page_dirty_nobuffers() to make sure that
586 running set_page_dirty() on a page containing mft/ntfs records will
587 not affect the dirty state of the page buffers.
588 - Add fs/ntfs/index.c::__ntfs_index_entry_mark_dirty() which sets all
589 buffers that are inside the ntfs record in the page dirty after which
590 it sets the page dirty. This allows ->writepage to only write the
591 dirty index records rather than having to write all the records in
592 the page. Modify fs/ntfs/index.h::ntfs_index_entry_mark_dirty() to
593 use this rather than __set_page_dirty_nobuffers().
594 - Implement fs/ntfs/aops.c::ntfs_write_mst_block() which enables the
595 writing of page cache pages belonging to mst protected attributes
596 like the index allocation attribute in directory indices and other
597 indices like $Quota/$Q, etc. This means that the quota is now marked
598 out of date on all volumes rather than only on ones where the quota
599 defaults entry is in the index root attribute of the $Quota/$Q index.
600
6012.1.14 - Fix an NFSd caused deadlock reported by several users.
602
603 - Modify fs/ntfs/ntfs_readdir() to copy the index root attribute value
604 to a buffer so that we can put the search context and unmap the mft
605 record before calling the filldir() callback. We need to do this
606 because of NFSd which calls ->lookup() from its filldir callback()
607 and this causes NTFS to deadlock as ntfs_lookup() maps the mft record
608 of the directory and since ntfs_readdir() has got it mapped already
609 ntfs_lookup() deadlocks.
610
6112.1.13 - Enable overwriting of resident files and housekeeping of system files.
612
613 - Implement writing of mft records (fs/ntfs/mft.[hc]), which includes
614 keeping the mft mirror in sync with the mft when mirrored mft records
615 are written. The functions are write_mft_record{,_nolock}(). The
616 implementation is quite rudimentary for now with lots of things not
617 implemented yet but I am not sure any of them can actually occur so
618 I will wait for people to hit each one and only then implement it.
619 - Commit open system inodes at umount time. This should make it
620 virtually impossible for sync_mft_mirror_umount() to ever be needed.
621 - Implement ->write_inode (fs/ntfs/inode.c::ntfs_write_inode()) for the
622 ntfs super operations. This gives us inode writing via the VFS inode
623 dirty code paths. Note: Access time updates are not implemented yet.
624 - Implement fs/ntfs/mft.[hc]::{,__}mark_mft_record_dirty() and make
625 fs/ntfs/aops.c::ntfs_writepage() and ntfs_commit_write() use it, thus
626 finally enabling resident file overwrite! (-8 This also includes a
627 placeholder for ->writepage (ntfs_mft_writepage()), which for now
628 just redirties the page and returns. Also, at umount time, we for
629 now throw away all mft data page cache pages after the last call to
630 ntfs_commit_inode() in the hope that all inodes will have been
631 written out by then and hence no dirty (meta)data will be lost. We
632 also check for this case and emit an error message telling the user
633 to run chkdsk.
634 - Use set_page_writeback() and end_page_writeback() in the resident
635 attribute code path of fs/ntfs/aops.c::ntfs_writepage() otherwise
636 the radix-tree tag PAGECACHE_TAG_DIRTY remains set even though the
637 page is clean.
638 - Implement ntfs_mft_writepage() so it now checks if any of the mft
639 records in the page are dirty and if so redirties the page and
640 returns. Otherwise it just returns (after doing set_page_writeback(),
641 unlock_page(), end_page_writeback() or the radix-tree tag
642 PAGECACHE_TAG_DIRTY remains set even though the page is clean), thus
643 alowing the VM to do with the page as it pleases. Also, at umount
644 time, now only throw away dirty mft (meta)data pages if dirty inodes
645 are present and ask the user to email us if they see this happening.
646 - Add functions ntfs_{clear,set}_volume_flags(), to modify the volume
647 information flags (fs/ntfs/super.c).
648 - Mark the volume dirty when (re)mounting read-write and mark it clean
649 when unmounting or remounting read-only. If any volume errors are
650 found, the volume is left marked dirty to force chkdsk to run.
651 - Add code to set the NT4 compatibility flag when (re)mounting
652 read-write for newer NTFS versions but leave it commented out for now
653 since we do not make any modifications that are NTFS 1.2 specific yet
654 and since setting this flag breaks Captive-NTFS which is not nice.
655 This code must be enabled once we start writing NTFS 1.2 specific
656 changes otherwise Windows NTFS driver might crash / cause corruption.
657
6582.1.12 - Fix the second fix to the decompression engine and some cleanups.
659
660 - Add a new address space operations struct, ntfs_mst_aops, for mst
661 protected attributes. This is because the default ntfs_aops do not
662 make sense with mst protected data and were they to write anything to
663 such an attribute they would cause data corruption so we provide
664 ntfs_mst_aops which does not have any write related operations set.
665 - Cleanup dirty ntfs inode handling (fs/ntfs/inode.[hc]) which also
666 includes an adapted ntfs_commit_inode() and an implementation of
667 ntfs_write_inode() which for now just cleans dirty inodes without
668 writing them (it does emit a warning that this is happening).
669 - Undo the second decompression engine fix (see 2.1.9 release ChangeLog
670 entry) as it was only fixing a theoretical bug but at the same time
671 it badly broke the handling of sparse and uncompressed compression
672 blocks.
673
6742.1.11 - Driver internal cleanups.
675
676 - Only build logfile.o if building the driver with read-write support.
677 - Really final white space cleanups.
678 - Use generic_ffs() instead of ffs() in logfile.c which allows the
679 log_page_size variable to be optimized by gcc into a constant.
680 - Rename uchar_t to ntfschar everywhere as uchar_t is unsigned 1-byte
681 char as defined by POSIX and as found on some systems.
682
6832.1.10 - Force read-only (re)mounting of volumes with unsupported volume flags.
684
685 - Finish off the white space cleanups (remove trailing spaces, etc).
686 - Clean up ntfs_fill_super() and ntfs_read_inode_mount() by removing
687 the kludges around the first iget(). Instead of (re)setting ->s_op
688 we have the $MFT inode set up by explicit new_inode() / set ->i_ino /
689 insert_inode_hash() / call ntfs_read_inode_mount() directly. This
690 kills the need for second super_operations and allows to return error
691 from ntfs_read_inode_mount() without resorting to ugly "poisoning"
692 tricks. (Al Viro)
693 - Force read-only (re)mounting if any of the following bits are set in
694 the volume information flags:
695 VOLUME_IS_DIRTY, VOLUME_RESIZE_LOG_FILE,
696 VOLUME_UPGRADE_ON_MOUNT, VOLUME_DELETE_USN_UNDERWAY,
697 VOLUME_REPAIR_OBJECT_ID, VOLUME_MODIFIED_BY_CHKDSK
698 To make this easier we define VOLUME_MUST_MOUNT_RO_MASK with all the
699 above bits set so the test is made easy.
700
7012.1.9 - Fix two bugs in decompression engine.
702
703 - Fix a bug where we would not always detect that we have reached the
704 end of a compression block because we were ending at minus one byte
705 which is effectively the same as being at the end. The fix is to
706 check whether the uncompressed buffer has been fully filled and if so
707 we assume we have reached the end of the compression block. A big
708 thank you to Marcin Gibuła for the bug report, the assistance in
709 tracking down the bug and testing the fix.
710 - Fix a possible bug where when a compressed read is truncated to the
711 end of the file, the offset inside the last page was not truncated.
712
7132.1.8 - Handle $MFT mirror and $LogFile, improve time handling, and cleanups.
714
715 - Use get_bh() instead of manual atomic_inc() in fs/ntfs/compress.c.
716 - Modify fs/ntfs/time.c::ntfs2utc(), get_current_ntfs_time(), and
717 utc2ntfs() to work with struct timespec instead of time_t on the
718 Linux UTC time side thus preserving the full precision of the NTFS
719 time and only loosing up to 99 nano-seconds in the Linux UTC time.
720 - Move fs/ntfs/time.c to fs/ntfs/time.h and make the time functions
721 static inline.
722 - Remove unused ntfs_dirty_inode().
723 - Cleanup super operations declaration in fs/ntfs/super.c.
724 - Wrap flush_dcache_mft_record_page() in #ifdef NTFS_RW.
725 - Add NInoTestSetFoo() and NInoTestClearFoo() macro magic to
726 fs/ntfs/inode.h and use it to declare NInoTest{Set,Clear}Dirty.
727 - Move typedefs for ntfs_attr and test_t from fs/ntfs/inode.c to
728 fs/ntfs/inode.h so they can be used elsewhere.
729 - Determine the mft mirror size as the number of mirrored mft records
730 and store it in ntfs_volume->mftmirr_size (fs/ntfs/super.c).
731 - Load the mft mirror at mount time and compare the mft records stored
732 in it to the ones in the mft. Force a read-only mount if the two do
733 not match (fs/ntfs/super.c).
734 - Fix type casting related warnings on 64-bit architectures. Thanks
735 to Meelis Roos for reporting them.
736 - Move %L to %ll as %L is floating point and %ll is integer which is
737 what we want.
738 - Read the journal ($LogFile) and determine if the volume has been
739 shutdown cleanly and force a read-only mount if not (fs/ntfs/super.c
740 and fs/ntfs/logfile.c). This is a little bit of a crude check in
741 that we only look at the restart areas and not at the actual log
742 records so that there will be a very small number of cases where we
743 think that a volume is dirty when in fact it is clean. This should
744 only affect volumes that have not been shutdown cleanly and did not
745 have any pending, non-check-pointed i/o.
746 - If the $LogFile indicates a clean shutdown and a read-write (re)mount
747 is requested, empty $LogFile by overwriting it with 0xff bytes to
748 ensure that Windows cannot cause data corruption by replaying a stale
749 journal after Linux has written to the volume.
750
7512.1.7 - Enable NFS exporting of mounted NTFS volumes.
752
753 - Set i_generation in the VFS inode from the seq_no of the NTFS inode.
754 - Make ntfs_lookup() NFS export safe, i.e. use d_splice_alias(), etc.
755 - Implement ->get_dentry() in fs/ntfs/namei.c::ntfs_get_dentry() as the
756 default doesn't allow inode number 0 which is a valid inode on NTFS
757 and even if it did allow that it uses iget() instead of ntfs_iget()
758 which makes it useless for us.
759 - Implement ->get_parent() in fs/ntfs/namei.c::ntfs_get_parent() as the
760 default just returns -EACCES which is not very useful.
761 - Define export operations (->s_export_op) for NTFS (ntfs_export_ops)
762 and set them up in the super block at mount time (super.c) this
763 allows mounted NTFS volumes to be exported via NFS.
764 - Add missing return -EOPNOTSUPP; in
765 fs/ntfs/aops.c::ntfs_commit_nonresident_write().
766 - Enforce no atime and no dir atime updates at mount/remount time as
767 they are not implemented yet anyway.
768 - Move a few assignments in fs/ntfs/attrib.c::load_attribute_list() to
769 after a NULL check. Thanks to Dave Jones for pointing this out.
770
7712.1.6 - Fix minor bug in handling of compressed directories.
772
773 - Fix bug in handling of compressed directories. A compressed
774 directory is not really compressed so when we set the ->i_blocks
775 field of a compressed directory inode we were setting it from the
776 non-existing field ni->itype.compressed.size which gave random
777 results... For directories we now always use ni->allocated_size.
778
7792.1.5 - Fix minor bug in attribute list attribute handling.
780
781 - Fix bug in attribute list handling. Actually it is not as much a bug
782 as too much protection in that we were not allowing attribute lists
783 which waste space on disk while Windows XP clearly allows it and in
784 fact creates such attribute lists so our driver was failing.
785 - Update NTFS documentation ready for 2.6 kernel release.
786
7872.1.4 - Reduce compiler requirements.
788
789 - Remove all uses of unnamed structs and unions in the driver to make
790 old and newer gcc versions happy. Makes it a bit uglier IMO but at
791 least people will stop hassling me about it.
792
7932.1.3 - Important bug fixes in corner cases.
794
795 - super.c::parse_ntfs_boot_sector(): Correct the check for 64-bit
796 clusters. (Philipp Thomas)
797 - attrib.c::load_attribute_list(): Fix bug when initialized_size is a
798 multiple of the block_size but not the cluster size. (Szabolcs
799 Szakacsits <szaka@sienet.hu>)
800
8012.1.2 - Important bug fixes aleviating the hangs in statfs.
802
803 - Fix buggy free cluster and free inode determination logic.
804
8052.1.1 - Minor updates.
806
807 - Add handling for initialized_size != data_size in compressed files.
808 - Reduce function local stack usage from 0x3d4 bytes to just noise in
809 fs/ntfs/upcase.c. (Randy Dunlap <rddunlap@osdl.ord>)
810 - Remove compiler warnings for newer gcc.
811 - Pages are no longer kmapped by mm/filemap.c::generic_file_write()
812 around calls to ->{prepare,commit}_write. Adapt NTFS appropriately
813 in fs/ntfs/aops.c::ntfs_prepare_nonresident_write() by using
814 kmap_atomic(KM_USER0).
815
8162.1.0 - First steps towards write support: implement file overwrite.
817
818 - Add configuration option for developmental write support with an
819 appropriately scary configuration help text.
820 - Initial implementation of fs/ntfs/aops.c::ntfs_writepage() and its
821 helper fs/ntfs/aops.c::ntfs_write_block(). This enables mmap(2) based
822 overwriting of existing files on ntfs. Note: Resident files are
823 only written into memory, and not written out to disk at present, so
824 avoid writing to files smaller than about 1kiB.
825 - Initial implementation of fs/ntfs/aops.c::ntfs_prepare_write(), its
826 helper fs/ntfs/aops.c::ntfs_prepare_nonresident_write() and their
827 counterparts, fs/ntfs/aops.c::ntfs_commit_write(), and
828 fs/ntfs/aops.c::ntfs_commit_nonresident_write(), respectively. Also,
829 add generic_file_write() to the ntfs file operations (fs/ntfs/file.c).
830 This enables write(2) based overwriting of existing files on ntfs.
831 Note: As with mmap(2) based overwriting, resident files are only
832 written into memory, and not written out to disk at present, so avoid
833 writing to files smaller than about 1kiB.
834 - Implement ->truncate (fs/ntfs/inode.c::ntfs_truncate()) and
835 ->setattr() (fs/ntfs/inode.c::ntfs_setattr()) inode operations for
836 files with the purpose of intercepting and aborting all i_size
837 changes which we do not support yet. ntfs_truncate() actually only
838 emits a warning message but AFAICS our interception of i_size changes
839 elsewhere means ntfs_truncate() never gets called for i_size changes.
840 It is only called from generic_file_write() when we fail in
841 ntfs_prepare_{,nonresident_}write() in order to discard any
842 instantiated buffers beyond i_size. Thus i_size is not actually
843 changed so our warning message is enough. Unfortunately it is not
844 possible to easily determine if i_size is being changed or not hence
845 we just emit an appropriately worded error message.
846
8472.0.25 - Small bug fixes and cleanups.
848
849 - Unlock the page in an out of memory error code path in
850 fs/ntfs/aops.c::ntfs_read_block().
851 - If fs/ntfs/aops.c::ntfs_read_page() is called on an uptodate page,
852 just unlock the page and return. (This can happen due to ->writepage
853 clearing PageUptodate() during write out of MstProtected()
854 attributes.
855 - Remove leaked write code again.
856
8572.0.24 - Cleanups.
858
859 - Treat BUG_ON() as ASSERT() not VERIFY(), i.e. do not use side effects
860 inside BUG_ON(). (Adam J. Richter)
861 - Split logical OR expressions inside BUG_ON() into individual BUG_ON()
862 calls for improved debugging. (Adam J. Richter)
863 - Add errors flag to the ntfs volume state, accessed via
864 NVol{,Set,Clear}Errors(vol).
865 - Do not allow read-write remounts of read-only volumes with errors.
866 - Clarify comment for ntfs file operation sendfile which was added by
867 Christoph Hellwig a while ago (just using generic_file_sendfile())
868 to say that ntfs ->sendfile is only used for the case where the
869 source data is on the ntfs partition and the destination is
870 somewhere else, i.e. nothing we need to concern ourselves with.
871 - Add generic_file_write() as our ntfs file write operation.
872
8732.0.23 - Major bug fixes (races, deadlocks, non-i386 architectures).
874
875 - Massive internal locking changes to mft record locking. Fixes lock
876 recursion and replaces the mrec_lock read/write semaphore with a
877 mutex. Also removes the now superfluous mft_count. This fixes several
878 race conditions and deadlocks, especially in the future write code.
879 - Fix ntfs over loopback for compressed files by adding an
880 optimization barrier. (gcc was screwing up otherwise ?)
881 - Miscellaneous cleanups all over the code and a fix or two in error
882 handling code paths.
883 Thanks go to Christoph Hellwig for pointing out the following two:
884 - Remove now unused function fs/ntfs/malloc.h::vmalloc_nofs().
885 - Fix ntfs_free() for ia64 and parisc by checking for VMALLOC_END, too.
886
8872.0.22 - Cleanups, mainly to ntfs_readdir(), and use C99 initializers.
888
889 - Change fs/ntfs/dir.c::ntfs_reddir() to only read/write ->f_pos once
890 at entry/exit respectively.
891 - Use C99 initializers for structures.
892 - Remove unused variable blocks from fs/ntfs/aops.c::ntfs_read_block().
893
8942.0.21 - Check for, and refuse to work with too large files/directories/volumes.
895
896 - Limit volume size at mount time to 2TiB on architectures where
897 unsigned long is 32-bits (fs/ntfs/super.c::parse_ntfs_boot_sector()).
898 This is the most we can do without overflowing the 32-bit limit of
899 the block device size imposed on us by sb_bread() and sb_getblk()
900 for the time being.
901 - Limit file/directory size at open() time to 16TiB on architectures
902 where unsigned long is 32-bits (fs/ntfs/file.c::ntfs_file_open() and
903 fs/ntfs/dir.c::ntfs_dir_open()). This is the most we can do without
904 overflowing the page cache page index.
905
9062.0.20 - Support non-resident directory index bitmaps, fix page leak in readdir.
907
908 - Move the directory index bitmap to use an attribute inode instead of
909 having special fields for it inside the ntfs inode structure. This
910 means that the index bitmaps now use the page cache for i/o, too,
911 and also as a side effect we get support for non-resident index
912 bitmaps for free.
913 - Simplify/cleanup error handling in fs/ntfs/dir.c::ntfs_readdir() and
914 fix a page leak that manifested itself in some cases.
915 - Add fs/ntfs/inode.c::ntfs_put_inode(), which we need to release the
916 index bitmap inode on the final iput().
917
9182.0.19 - Fix race condition, improvements, and optimizations in i/o interface.
919
920 - Apply block optimization added to fs/ntfs/aops.c::ntfs_read_block()
921 to fs/ntfs/compress.c::ntfs_file_read_compressed_block() as well.
922 - Drop the "file" from ntfs_file_read_compressed_block().
923 - Rename fs/ntfs/aops.c::ntfs_enb_buffer_read_async() to
924 ntfs_end_buffer_async_read() (more like the fs/buffer.c counterpart).
925 - Update ntfs_end_buffer_async_read() with the improved logic from
926 its updated counterpart fs/buffer.c::end_buffer_async_read(). Apply
927 further logic improvements to better determine when we set PageError.
928 - Update submission of buffers in fs/ntfs/aops.c::ntfs_read_block() to
929 check for the buffers being uptodate first in line with the updated
930 fs/buffer.c::block_read_full_page(). This plugs a small race
931 condition.
932
9332.0.18 - Fix race condition in reading of compressed files.
934
935 - There was a narrow window between checking a buffer head for being
936 uptodate and locking it in ntfs_file_read_compressed_block(). We now
937 lock the buffer and then check whether it is uptodate or not.
938
9392.0.17 - Cleanups and optimizations - shrinking the ToDo list.
940
941 - Modify fs/ntfs/inode.c::ntfs_read_locked_inode() to return an error
942 code and update callers, i.e. ntfs_iget(), to pass that error code
943 up instead of just using -EIO.
944 - Modifications to super.c to ensure that both mount and remount
945 cannot set any write related options when the driver is compiled
946 read-only.
947 - Optimize block resolution in fs/ntfs/aops.c::ntfs_read_block() to
948 cache the current runlist element. This should improve performance
949 when reading very large and/or very fragmented data.
950
9512.0.16 - Convert access to $MFT/$BITMAP to attribute inode API.
952
953 - Fix a stupid bug introduced in 2.0.15 where we were unmapping the
954 wrong inode in fs/ntfs/inode.c::ntfs_attr_iget().
955 - Fix debugging check in fs/ntfs/aops.c::ntfs_read_block().
956 - Convert $MFT/$BITMAP access to attribute inode API and remove all
957 remnants of the ugly mftbmp address space and operations hack. This
958 means we finally have only one readpage function as well as only one
959 async io completion handler. Yey! The mft bitmap is now just an
960 attribute inode and is accessed from vol->mftbmp_ino just as if it
961 were a normal file. Fake inodes rule. (-:
962
9632.0.15 - Fake inodes based attribute i/o via the pagecache, fixes and cleanups.
964
965 - Fix silly bug in fs/ntfs/super.c::parse_options() which was causing
966 remounts to fail when the partition had an entry in /etc/fstab and
967 the entry specified the nls= option.
968 - Apply same macro magic used in fs/ntfs/inode.h to fs/ntfs/volume.h to
969 expand all the helper functions NVolFoo(), NVolSetFoo(), and
970 NVolClearFoo().
971 - Move copyright statement from driver initialisation message to
972 module description (fs/super.c). This makes the initialisation
973 message fit on one line and fits in better with rest of kernel.
974 - Update fs/ntfs/attrib.c::map_run_list() to work on both real and
975 attribute inodes, and both for files and directories.
976 - Implement fake attribute inodes allowing all attribute i/o to go via
977 the page cache and to use all the normal vfs/mm functionality:
978 - Add ntfs_attr_iget() and its helper ntfs_read_locked_attr_inode()
979 to fs/ntfs/inode.c.
980 - Add needed cleanup code to ntfs_clear_big_inode().
981 - Merge address space operations for files and directories (aops.c),
982 now just have ntfs_aops:
983 - Rename:
984 end_buffer_read_attr_async() -> ntfs_end_buffer_read_async(),
985 ntfs_attr_read_block() -> ntfs_read_block(),
986 ntfs_file_read_page() -> ntfs_readpage().
987 - Rewrite fs/ntfs/aops.c::ntfs_readpage() to work on both real and
988 attribute inodes, and both for files and directories.
989 - Remove obsolete fs/ntfs/aops.c::ntfs_mst_readpage().
990
9912.0.14 - Run list merging code cleanup, minor locking changes, typo fixes.
992
993 - Change fs/ntfs/super.c::ntfs_statfs() to not rely on BKL by moving
994 the locking out of super.c::get_nr_free_mft_records() and taking and
995 dropping the mftbmp_lock rw_semaphore in ntfs_statfs() itself.
996 - Bring attribute runlist merging code (fs/ntfs/attrib.c) in sync with
997 current userspace ntfs library code. This means that if a merge
998 fails the original runlists are always left unmodified instead of
999 being silently corrupted.
1000 - Misc typo fixes.
1001
10022.0.13 - Use iget5_locked() in preparation for fake inodes and small cleanups.
1003
1004 - Remove nr_mft_bits and the now superfluous union with nr_mft_records
1005 from ntfs_volume structure.
1006 - Remove nr_lcn_bits and the now superfluous union with nr_clusters
1007 from ntfs_volume structure.
1008 - Use iget5_locked() and friends instead of conventional iget(). Wrap
1009 the call in fs/ntfs/inode.c::ntfs_iget() and update callers of iget()
1010 to use ntfs_iget(). Leave only one iget() call at mount time so we
1011 don't need an ntfs_iget_mount().
1012 - Change fs/ntfs/inode.c::ntfs_new_extent_inode() to take mft_no as an
1013 additional argument.
1014
10152.0.12 - Initial cleanup of address space operations following 2.0.11 changes.
1016
1017 - Merge fs/ntfs/aops.c::end_buffer_read_mst_async() and
1018 fs/ntfs/aops.c::end_buffer_read_file_async() into one function
1019 fs/ntfs/aops.c::end_buffer_read_attr_async() using NInoMstProtected()
1020 to determine whether to apply mst fixups or not.
1021 - Above change allows merging fs/ntfs/aops.c::ntfs_file_read_block()
1022 and fs/ntfs/aops.c::ntfs_mst_readpage() into one function
1023 fs/ntfs/aops.c::ntfs_attr_read_block(). Also, create a tiny wrapper
1024 fs/ntfs/aops.c::ntfs_mst_readpage() to transform the parameters from
1025 the VFS readpage function prototype to the ntfs_attr_read_block()
1026 function prototype.
1027
10282.0.11 - Initial preparations for fake inode based attribute i/o.
1029
1030 - Move definition of ntfs_inode_state_bits to fs/ntfs/inode.h and
1031 do some macro magic (adapted from include/linux/buffer_head.h) to
1032 expand all the helper functions NInoFoo(), NInoSetFoo(), and
1033 NInoClearFoo().
1034 - Add new flag to ntfs_inode_state_bits: NI_Sparse.
1035 - Add new fields to ntfs_inode structure to allow use of fake inodes
1036 for attribute i/o: type, name, name_len. Also add new state bits:
1037 NI_Attr, which, if set, indicates the inode is a fake inode, and
1038 NI_MstProtected, which, if set, indicates the attribute uses multi
1039 sector transfer protection, i.e. fixups need to be applied after
1040 reads and before/after writes.
1041 - Rename fs/ntfs/inode.c::ntfs_{new,clear,destroy}_inode() to
1042 ntfs_{new,clear,destroy}_extent_inode() and update callers.
1043 - Use ntfs_clear_extent_inode() in fs/ntfs/inode.c::__ntfs_clear_inode()
1044 instead of ntfs_destroy_extent_inode().
1045 - Cleanup memory deallocations in {__,}ntfs_clear_{,big_}inode().
1046 - Make all operations on ntfs inode state bits use the NIno* functions.
1047 - Set up the new ntfs inode fields and state bits in
1048 fs/ntfs/inode.c::ntfs_read_inode() and add appropriate cleanup of
1049 allocated memory to __ntfs_clear_inode().
1050 - Cleanup ntfs_inode structure a bit for better ordering of elements
1051 w.r.t. their size to allow better packing of the structure in memory.
1052
10532.0.10 - There can only be 2^32 - 1 inodes on an NTFS volume.
1054
1055 - Add check at mount time to verify that the number of inodes on the
1056 volume does not exceed 2^32 - 1, which is the maximum allowed for
1057 NTFS according to Microsoft.
1058 - Change mft_no member of ntfs_inode structure to be unsigned long.
1059 Update all users. This makes ntfs_inode->mft_no just a copy of struct
1060 inode->i_ino. But we can't just always use struct inode->i_ino and
1061 remove mft_no because extent inodes do not have an attached struct
1062 inode.
1063
10642.0.9 - Decompression engine now uses a single buffer and other cleanups.
1065
1066 - Change decompression engine to use a single buffer protected by a
1067 spin lock instead of per-CPU buffers. (Rusty Russell)
1068 - Do not update cb_pos when handling a partial final page during
1069 decompression of a sparse compression block, as the value is later
1070 reset without being read/used. (Rusty Russell)
1071 - Switch to using the new KM_BIO_SRC_IRQ for atomic kmap()s. (Andrew
1072 Morton)
1073 - Change buffer size in ntfs_readdir()/ntfs_filldir() to use
1074 NLS_MAX_CHARSET_SIZE which makes the buffers almost 1kiB each but
1075 it also makes everything safer so it is a good thing.
1076 - Miscellaneous minor cleanups to comments.
1077
10782.0.8 - Major updates for handling of case sensitivity and dcache aliasing.
1079
1080 Big thanks go to Al Viro and other inhabitants of #kernel for investing
1081 their time to discuss the case sensitivity and dcache aliasing issues.
1082
1083 - Remove unused source file fs/ntfs/attraops.c.
1084 - Remove show_inodes mount option(s), thus dropping support for
1085 displaying of short file names.
1086 - Remove deprecated mount option posix.
1087 - Restore show_sys_files mount option.
1088 - Add new mount option case_sensitive, to determine if the driver
1089 treats file names as case sensitive or not. If case sensitive, create
1090 file names in the POSIX namespace. Otherwise create file names in the
1091 LONG/WIN32 namespace. Note, files remain accessible via their short
1092 file name, if it exists.
1093 - Remove really dumb logic bug in boot sector recovery code.
1094 - Fix dcache aliasing issues wrt short/long file names via changes
1095 to fs/ntfs/dir.c::ntfs_lookup_inode_by_name() and
1096 fs/ntfs/namei.c::ntfs_lookup():
1097 - Add additional argument to ntfs_lookup_inode_by_name() in which we
1098 return information about the matching file name if the case is not
1099 matching or the match is a short file name. See comments above the
1100 function definition for details.
1101 - Change ntfs_lookup() to only create dcache entries for the correctly
1102 cased file name and only for the WIN32 namespace counterpart of DOS
1103 namespace file names. This ensures we have only one dentry per
1104 directory and also removes all dcache aliasing issues between short
1105 and long file names once we add write support. See comments above
1106 function for details.
1107 - Fix potential 1 byte overflow in fs/ntfs/unistr.c::ntfs_ucstonls().
1108
11092.0.7 - Minor cleanups and updates for changes in core kernel code.
1110
1111 - Remove much of the NULL struct element initializers.
1112 - Various updates to make compatible with recent kernels.
1113 - Remove defines of MAX_BUF_PER_PAGE and include linux/buffer_head.h
1114 in fs/ntfs/ntfs.h instead.
1115 - Remove no longer needed KERNEL_VERSION checks. We are now in the
1116 kernel proper so they are no longer needed.
1117
11182.0.6 - Major bugfix to make compatible with other kernel changes.
1119
1120 - Initialize the mftbmp address space properly now that there are more
1121 fields in the struct address_space. This was leading to hangs and
1122 oopses on umount since 2.5.12 because of changes to other parts of
1123 the kernel. We probably want a kernel generic init_address_space()
1124 function...
1125 - Drop BKL from ntfs_readdir() after consultation with Al Viro. The
1126 only caller of ->readdir() is vfs_readdir() which holds i_sem during
1127 the call, and i_sem is sufficient protection against changes in the
1128 directory inode (including ->i_size).
1129 - Use generic_file_llseek() for directories (as opposed to
1130 default_llseek()) as this downs i_sem instead of the BKL which is
1131 what we now need for exclusion against ->f_pos changes considering we
1132 no longer take the BKL in ntfs_readdir().
1133
11342.0.5 - Major bugfix. Buffer overflow in extent inode handling.
1135
1136 - No need to set old blocksize in super.c::ntfs_fill_super() as the
1137 VFS does so via invocation of deactivate_super() calling
1138 fs->fill_super() calling block_kill_super() which does it.
1139 - BKL moved from VFS into dir.c::ntfs_readdir(). (Linus Torvalds)
1140 -> Do we really need it? I don't think so as we have exclusion on
1141 the directory ntfs_inode rw_semaphore mrec_lock. We mmight have to
1142 move the ->f_pos accesses under the mrec_lock though. Check this...
1143 - Fix really, really, really stupid buffer overflow in extent inode
1144 handling in mft.c::map_extent_mft_record().
1145
11462.0.4 - Cleanups and updates for kernel 2.5.11.
1147
1148 - Add documentation on how to use the MD driver to be able to use NTFS
1149 stripe and volume sets in Linux and generally cleanup documentation
1150 a bit.
1151 Remove all uses of kdev_t in favour of struct block_device *:
1152 - Change compress.c::ntfs_file_read_compressed_block() to use
1153 sb_getblk() instead of getblk().
1154 - Change super.c::ntfs_fill_super() to use bdev_hardsect_size() instead
1155 of get_hardsect_size().
1156 - No need to get old blocksize in super.c::ntfs_fill_super() as
1157 fs/super.c::get_sb_bdev() already does this.
1158 - Set bh->b_bdev instead of bh->b_dev throughout aops.c.
1159
11602.0.3 - Small bug fixes, cleanups, and performance improvements.
1161
1162 - Remove some dead code from mft.c.
1163 - Optimize readpage and read_block functions throughout aops.c so that
1164 only initialized blocks are read. Non-initialized ones have their
1165 buffer head mapped, zeroed, and set up to date, without scheduling
1166 any i/o. Thanks to Al Viro for advice on how to avoid the device i/o.
1167 Thanks go to Andrew Morton for spotting the below:
1168 - Fix buglet in allocate_compression_buffers() error code path.
1169 - Call flush_dcache_page() after modifying page cache page contents in
1170 ntfs_file_readpage().
1171 - Check for existence of page buffers throughout aops.c before calling
1172 create_empty_buffers(). This happens when an I/O error occurs and the
1173 read is retried. (It also happens once writing is implemented so that
1174 needed doing anyway but I had left it for later...)
1175 - Don't BUG_ON() uptodate and/or mapped buffers throughout aops.c in
1176 readpage and read_block functions. Reasoning same as above (i.e. I/O
1177 error retries and future write code paths.)
1178
11792.0.2 - Minor updates and cleanups.
1180
1181 - Cleanup: rename mst.c::__post_read_mst_fixup to post_write_mst_fixup
1182 and cleanup the code a bit, removing the unused size parameter.
1183 - Change default fmask to 0177 and update documentation.
1184 - Change attrib.c::get_attr_search_ctx() to return the search context
1185 directly instead of taking the address of a pointer. A return value
1186 of NULL means the allocation failed. Updated all callers
1187 appropriately.
1188 - Update to 2.5.9 kernel (preserving backwards compatibility) by
1189 replacing all occurences of page->buffers with page_buffers(page).
1190 - Fix minor bugs in runlist merging, also minor cleanup.
1191 - Updates to bootsector layout and mft mirror contents descriptions.
1192 - Small bug fix in error detection in unistr.c and some cleanups.
1193 - Grow name buffer allocations in unistr.c in aligned mutlipled of 64
1194 bytes.
1195
11962.0.1 - Minor updates.
1197
1198 - Make default umask correspond to documentation.
1199 - Improve documentation.
1200 - Set default mode to include execute bit. The {u,f,d}mask can be used
1201 to take it away if desired. This allows binaries to be executed from
1202 a mounted ntfs partition.
1203
12042.0.0 - New version number. Remove TNG from the name. Now in the kernel.
1205
1206 - Add kill_super, just keeping up with the vfs changes in the kernel.
1207 - Repeat some changes from tng-0.0.8 that somehow got lost on the way
1208 from the CVS import into BitKeeper.
1209 - Begin to implement proper handling of allocated_size vs
1210 initialized_size vs data_size (i.e. i_size). Done are
1211 mft.c::ntfs_mft_readpage(), aops.c::end_buffer_read_index_async(),
1212 and attrib.c::load_attribute_list().
1213 - Lock the runlist in attrib.c::load_attribute_list() while using it.
1214 - Fix memory leak in ntfs_file_read_compressed_block() and generally
1215 clean up compress.c a little, removing some uncommented/unused debug
1216 code.
1217 - Tidy up dir.c a little bit.
1218 - Don't bother getting the runlist in inode.c::ntfs_read_inode().
1219 - Merge mft.c::ntfs_mft_readpage() and aops.c::ntfs_index_readpage()
1220 creating aops.c::ntfs_mst_readpage(), improving the handling of
1221 holes and overflow in the process and implementing the correct
1222 equivalent of ntfs_file_get_block() in ntfs_mst_readpage() itself.
1223 I am aiming for correctness at the moment. Modularisation can come
1224 later.
1225 - Rename aops.c::end_buffer_read_index_async() to
1226 end_buffer_read_mst_async() and optimize the overflow checking and
1227 handling.
1228 - Use the host of the mftbmp address space mapping to hold the ntfs
1229 volume. This is needed so the async i/o completion handler can
1230 retrieve a pointer to the volume. Hopefully this will not cause
1231 problems elsewhere in the kernel... Otherwise will need to use a
1232 fake inode.
1233 - Complete implementation of proper handling of allocated_size vs
1234 initialized_size vs data_size (i.e. i_size) in whole driver.
1235 Basically aops.c is now completely rewritten.
1236 - Change NTFS driver name to just NTFS and set version number to 2.0.0
1237 to make a clear distinction from the old driver which is still on
1238 version 1.1.22.
1239
1240tng-0.0.8 - 08/03/2002 - Now using BitKeeper, http://linux-ntfs.bkbits.net/
1241
1242 - Replace bdevname(sb->s_dev) with sb->s_id.
1243 - Remove now superfluous new-line characters in all callers of
1244 ntfs_debug().
1245 - Apply kludge in ntfs_read_inode(), setting i_nlink to 1 for
1246 directories. Without this the "find" utility gets very upset which is
1247 fair enough as Linux/Unix do not support directory hard links.
1248 - Further runlist merging work. (Richard Russon)
1249 - Backwards compatibility for gcc-2.95. (Richard Russon)
1250 - Update to kernel 2.5.5-pre1 and rediff the now tiny patch.
Anton Altaparmakovc002f422005-02-03 12:02:56 +00001251 - Convert to new filesystem declaration using ->ntfs_get_sb() and
Linus Torvalds1da177e2005-04-16 15:20:36 -07001252 replacing ntfs_read_super() with ntfs_fill_super().
1253 - Set s_maxbytes to MAX_LFS_FILESIZE to avoid page cache page index
1254 overflow on 32-bit architectures.
1255 - Cleanup upcase loading code to use ntfs_(un)map_page().
1256 - Disable/reenable preemtion in critical sections of compession engine.
1257 - Replace device size determination in ntfs_fill_super() with
1258 sb->s_bdev->bd_inode->i_size (in bytes) and remove now superfluous
1259 function super.c::get_nr_blocks().
1260 - Implement a mount time option (show_inodes) allowing choice of which
1261 types of inode names readdir() returns and modify ntfs_filldir()
1262 accordingly. There are several parameters to show_inodes:
1263 system: system files
1264 win32: long file names (including POSIX file names) [DEFAULT]
1265 long: same as win32
1266 dos: short file names only (excluding POSIX file names)
1267 short: same as dos
1268 posix: same as both win32 and dos
1269 all: all file names
1270 Note that the options are additive, i.e. specifying:
1271 -o show_inodes=system,show_inodes=win32,show_inodes=dos
1272 is the same as specifying:
1273 -o show_inodes=all
1274 Note that the "posix" and "all" options will show all directory
1275 names, BUT the link count on each directory inode entry is set to 1,
1276 due to Linux not supporting directory hard links. This may well
1277 confuse some userspace applications, since the directory names will
1278 have the same inode numbers. Thus it is NOT advisable to use the
1279 "posix" or "all" options. We provide them only for completeness sake.
1280 - Add copies of allocated_size, initialized_size, and compressed_size to
1281 the ntfs inode structure and set them up in
1282 inode.c::ntfs_read_inode(). These reflect the unnamed data attribute
1283 for files and the index allocation attribute for directories.
1284 - Add copies of allocated_size and initialized_size to ntfs inode for
1285 $BITMAP attribute of large directories and set them up in
1286 inode.c::ntfs_read_inode().
1287 - Add copies of allocated_size and initialized_size to ntfs volume for
1288 $BITMAP attribute of $MFT and set them up in
1289 super.c::load_system_files().
1290 - Parse deprecated ntfs driver options (iocharset, show_sys_files,
1291 posix, and utf8) and tell user what the new options to use are. Note
1292 we still do support them but they will be removed with kernel 2.7.x.
1293 - Change all occurences of integer long long printf formatting to hex
1294 as printk() will not support long long integer format if/when the
1295 div64 patch goes into the kernel.
1296 - Make slab caches have stable names and change the names to what they
1297 were intended to be. These changes are required/made possible by the
1298 new slab cache name handling which removes the length limitation by
1299 requiring the caller of kmem_cache_create() to supply a stable name
1300 which is then referenced but not copied.
1301 - Rename run_list structure to run_list_element and create a new
1302 run_list structure containing a pointer to a run_list_element
1303 structure and a read/write semaphore. Adapt all users of runlists
1304 to new scheme and take and release the lock as needed. This fixes a
1305 nasty race as the run_list changes even when inodes are locked for
1306 reading and even when the inode isn't locked at all, so we really
1307 needed the serialization. We use a semaphore rather than a spinlock
1308 as memory allocations can sleep and doing everything GFP_ATOMIC
1309 would be silly.
1310 - Cleanup read_inode() removing all code checking for lowest_vcn != 0.
1311 This can never happen due to the nature of lookup_attr() and how we
1312 support attribute lists. If it did happen it would imply the inode
1313 being corrupt.
1314 - Check for lowest_vcn != 0 in ntfs_read_inode() and mark the inode as
1315 bad if found.
1316 - Update to 2.5.6-pre2 changes in struct address_space.
1317 - Use parent_ino() when accessing d_parent inode number in dir.c.
1318 - Import Sourceforge CVS repository into BitKeeper repository:
1319 http://linux-ntfs.bkbits.net/ntfs-tng-2.5
1320 - Update fs/Makefile, fs/Config.help, fs/Config.in, and
1321 Documentation/filesystems/ntfs.txt for NTFS TNG.
1322 - Create kernel configuration option controlling whether debugging
1323 is enabled or not.
1324 - Add the required export of end_buffer_io_sync() from the patches
1325 directory to the kernel code.
1326 - Update inode.c::ntfs_show_options() with show_inodes mount option.
1327 - Update errors mount option.
1328
1329tng-0.0.7 - 13/02/2002 - The driver is now feature complete for read-only!
1330
1331 - Cleanup mft.c and it's debug/error output in particular. Fix a minor
1332 bug in mapping of extent inodes. Update all the comments to fit all
1333 the recent code changes.
1334 - Modify vcn_to_lcn() to cope with entirely unmapped runlists.
1335 - Cleanups in compress.c, mostly comments and folding help.
1336 - Implement attrib.c::map_run_list() as a generic helper.
1337 - Make compress.c::ntfs_file_read_compressed_block() use map_run_list()
1338 thus making code shorter and enabling attribute list support.
1339 - Cleanup incorrect use of [su]64 with %L printf format specifier in
1340 all source files. Type casts to [unsigned] long long added to correct
1341 the mismatches (important for architectures which have long long not
1342 being 64 bits).
1343 - Merge async io completion handlers for directory indexes and $MFT
1344 data into one by setting the index_block_size{_bits} of the ntfs
1345 inode for $MFT to the mft_record_size{_bits} of the ntfs_volume.
1346 - Cleanup aops.c, update comments.
1347 - Make ntfs_file_get_block() use map_run_list() so all files now
1348 support attribute lists.
1349 - Make ntfs_dir_readpage() almost verbatim copy of
1350 block_read_full_page() by using ntfs_file_get_block() with only real
1351 difference being the use of our own async io completion handler
1352 rather than the default one, thus reducing the amount of code and
1353 automatically enabling attribute list support for directory indices.
1354 - Fix bug in load_attribute_list() - forgot to call brelse in error
1355 code path.
1356 - Change parameters to find_attr() and lookup_attr(). We no longer
1357 pass in the upcase table and its length. These can be gotten from
1358 ctx->ntfs_ino->vol->upcase{_len}. Update all callers.
1359 - Cleanups in attrib.c.
1360 - Implement merging of runlists, attrib.c::merge_run_lists() and its
1361 helpers. (Richard Russon)
1362 - Attribute lists part 2, attribute extents and multi part runlists:
1363 enable proper support for LCN_RL_NOT_MAPPED and automatic mapping of
1364 further runlist parts via attrib.c::map_run_list().
1365 - Tiny endianness bug fix in decompress_mapping_pairs().
1366
1367tng-0.0.6 - Encrypted directories, bug fixes, cleanups, debugging enhancements.
1368
1369 - Enable encrypted directories. (Their index root is marked encrypted
1370 to indicate that new files in that directory should be created
1371 encrypted.)
1372 - Fix bug in NInoBmpNonResident() macro. (Cut and paste error.)
1373 - Enable $Extend system directory. Most (if not all) extended system
1374 files do not have unnamed data attributes so ntfs_read_inode() had to
1375 special case them but that is ok, as the special casing recovery
1376 happens inside an error code path so there is zero slow down in the
1377 normal fast path. The special casing is done by introducing a new
1378 function inode.c::ntfs_is_extended_system_file() which checks if any
1379 of the hard links in the inode point to $Extend as being their parent
1380 directory and if they do we assume this is an extended system file.
1381 - Create a sysctl/proc interface to allow {dis,en}abling of debug output
1382 when compiled with -DDEBUG. Default is debug messages to be disabled.
1383 To enable them, one writes a non-zero value to /proc/sys/fs/ntfs-debug
1384 (if /proc is enabled) or uses sysctl(2) to effect the same (if sysctl
1385 interface is enabled). Inspired by old ntfs driver.
1386 - Add debug_msgs insmod/kernel boot parameter to set whether debug
1387 messages are {dis,en}abled. This is useful to enable debug messages
1388 during ntfs initialization and is the only way to activate debugging
1389 when the sysctl interface is not enabled.
1390 - Cleanup debug output in various places.
1391 - Remove all dollar signs ($) from the source (except comments) to
1392 enable compilation on architectures whose gcc compiler does not
1393 support dollar signs in the names of variables/constants. Attribute
1394 types now start with AT_ instead of $ and $I30 is now just I30.
1395 - Cleanup ntfs_lookup() and add consistency check of sequence numbers.
1396 - Load complete runlist for $MFT/$BITMAP during mount and cleanup
1397 access functions. This means we now cope with $MFT/$BITMAP being
1398 spread accross several mft records.
1399 - Disable modification of mft_zone_multiplier on remount. We can always
1400 reenable this later on if we really want to, but we will need to make
1401 sure we readjust the mft_zone size / layout accordingly.
1402
1403tng-0.0.5 - Modernize for 2.5.x and further in line-ing with Al Viro's comments.
1404
1405 - Use sb_set_blocksize() instead of set_blocksize() and verify the
1406 return value.
1407 - Use sb_bread() instead of bread() throughout.
1408 - Add index_vcn_size{_bits} to ntfs_inode structure to store the size
1409 of a directory index block vcn. Apply resulting simplifications in
1410 dir.c everywhere.
1411 - Fix a small bug somewhere (but forgot what it was).
1412 - Change ntfs_{debug,error,warning} to enable gcc to do type checking
1413 on the printf-format parameter list and fix bugs reported by gcc
1414 as a result. (Richard Russon)
1415 - Move inode allocation strategy to Al's new stuff but maintain the
1416 divorce of ntfs_inode from struct inode. To achieve this we have two
1417 separate slab caches, one for big ntfs inodes containing a struct
1418 inode and pure ntfs inodes and at the same time fix some faulty
1419 error code paths in ntfs_read_inode().
1420 - Show mount options in proc (inode.c::ntfs_show_options()).
1421
1422tng-0.0.4 - Big changes, getting in line with Al Viro's comments.
1423
1424 - Modified (un)map_mft_record functions to be common for read and write
1425 case. To specify which is which, added extra parameter at front of
1426 parameter list. Pass either READ or WRITE to this, each has the
1427 obvious meaning.
1428 - General cleanups to allow for easier folding in vi.
1429 - attrib.c::decompress_mapping_pairs() now accepts the old runlist
1430 argument, and invokes attrib.c::merge_run_lists() to merge the old
1431 and the new runlists.
1432 - Removed attrib.c::find_first_attr().
1433 - Implemented loading of attribute list and complete runlist for $MFT.
1434 This means we now cope with $MFT being spread across several mft
1435 records.
1436 - Adapt to 2.5.2-pre9 and the changed create_empty_buffers() syntax.
1437 - Adapt major/minor/kdev_t/[bk]devname stuff to new 2.5.x kernels.
1438 - Make ntfs_volume be allocated via kmalloc() instead of using a slab
1439 cache. There are too little ntfs_volume structures at any one time
1440 to justify a private slab cache.
1441 - Fix bogus kmap() use in async io completion. Now use kmap_atomic().
1442 Use KM_BIO_IRQ on advice from IRC/kernel...
1443 - Use ntfs_map_page() in map_mft_record() and create ->readpage method
1444 for reading $MFT (ntfs_mft_readpage). In the process create dedicated
1445 address space operations (ntfs_mft_aops) for $MFT inode mapping. Also
1446 removed the now superfluous exports from the kernel core patch.
1447 - Fix a bug where kfree() was used insted of ntfs_free().
1448 - Change map_mft_record() to take ntfs_inode as argument instead of
1449 vfs inode. Dito for unmap_mft_record(). Adapt all callers.
1450 - Add pointer to ntfs_volume to ntfs_inode.
1451 - Add mft record number and sequence number to ntfs_inode. Stop using
1452 i_ino and i_generation for in-driver purposes.
1453 - Implement attrib.c::merge_run_lists(). (Richard Russon)
1454 - Remove use of proper inodes by extent inodes. Move i_ino and
1455 i_generation to ntfs_inode to do this. Apply simplifications that
1456 result and remove iget_no_wait(), etc.
1457 - Pass ntfs_inode everywhere in the driver (used to be struct inode).
1458 - Add reference counting in ntfs_inode for the ntfs inode itself and
1459 for the mapped mft record.
1460 - Extend mft record mapping so we can (un)map extent mft records (new
1461 functions (un)map_extent_mft_record), and so mappings are reference
1462 counted and don't have to happen twice if already mapped - just ref
1463 count increases.
1464 - Add -o iocharset as alias to -o nls for backwards compatibility.
1465 - The latest core patch is now tiny. In fact just a single additional
1466 export is necessary over the base kernel.
1467
1468tng-0.0.3 - Cleanups, enhancements, bug fixes.
1469
1470 - Work on attrib.c::decompress_mapping_pairs() to detect base extents
1471 and setup the runlist appropriately using knowledge provided by the
1472 sizes in the base attribute record.
1473 - Balance the get_/put_attr_search_ctx() calls so we don't leak memory
1474 any more.
1475 - Introduce ntfs_malloc_nofs() and ntfs_free() to allocate/free a single
1476 page or use vmalloc depending on the amount of memory requested.
1477 - Cleanup error output. The __FUNCTION__ "(): " is now added
1478 automatically. Introduced a new header file debug.h to support this
1479 and also moved ntfs_debug() function into it.
1480 - Make reading of compressed files more intelligent and especially get
1481 rid of the vmalloc_nofs() from readpage(). This now uses per CPU
1482 buffers (allocated at first mount with cluster size <= 4kiB and
1483 deallocated on last umount with cluster size <= 4kiB), and
1484 asynchronous io for the compressed data using a list of buffer heads.
1485 Er, we use synchronous io as async io only works on whole pages
1486 covered by buffers and not on individual buffer heads...
1487 - Bug fix for reading compressed files with sparse compression blocks.
1488
1489tng-0.0.2 - Now handles larger/fragmented/compressed volumes/files/dirs.
1490
1491 - Fixed handling of directories when cluster size exceeds index block
1492 size.
1493 - Hide DOS only name space directory entries from readdir() but allow
1494 them in lookup(). This should fix the problem that Linux doesn't
1495 support directory hard links, while still allowing access to entries
1496 via their short file name. This also has the benefit of mimicking
1497 what Windows users are used to, so it is the ideal solution.
1498 - Implemented sync_page everywhere so no more hangs in D state when
1499 waiting for a page.
1500 - Stop using bforget() in favour of brelse().
1501 - Stop locking buffers unnecessarily.
1502 - Implemented compressed files (inode->mapping contains uncompressed
1503 data, raw compressed data is currently bread() into a vmalloc()ed
1504 memory buffer).
1505 - Enable compressed directories. (Their index root is marked compressed
1506 to indicate that new files in that directory should be created
1507 compressed.)
1508 - Use vsnprintf rather than vsprintf in the ntfs_error and ntfs_warning
1509 functions. (Thanks to Will Dyson for pointing this out.)
1510 - Moved the ntfs_inode and ntfs_volume (the former ntfs_inode_info and
1511 ntfs_sb_info) out of the common inode and super_block structures and
1512 started using the generic_ip and generic_sbp pointers instead. This
1513 makes ntfs entirely private with respect to the kernel tree.
1514 - Detect compiler version and abort with error message if gcc less than
1515 2.96 is used.
1516 - Fix bug in name comparison function in unistr.c.
1517 - Implement attribute lists part 1, the infrastructure: search contexts
1518 and operations, find_external_attr(), lookup_attr()) and make the
1519 code use the infrastructure.
1520 - Fix stupid buffer overflow bug that became apparent on larger run
1521 list containing attributes.
1522 - Fix bugs in readdir() that became apparent on larger directories.
1523
1524 The driver is now really useful and survives the test
1525 find . -type f -exec md5sum "{}" \;
1526 without any error messages on a over 1GiB sized partition with >16k
1527 files on it, including compressed files and directories and many files
1528 and directories with attribute lists.
1529
1530tng-0.0.1 - The first useful version.
1531
1532 - Added ntfs_lookup().
1533 - Added default upcase generation and handling.
1534 - Added compile options to be shown on module init.
1535 - Many bug fixes that were "hidden" before.
1536 - Update to latest kernel.
1537 - Added ntfs_readdir().
1538 - Added file operations for mmap(), read(), open() and llseek(). We just
1539 use the generic ones. The whole point of going through implementing
1540 readpage() methods and where possible get_block() call backs is that
1541 this allows us to make use of the generic high level methods provided
1542 by the kernel.
1543
1544 The driver is now actually useful! Yey. (-: It undoubtedly has got bugs
1545 though and it doesn't implement accesssing compressed files yet. Also,
1546 accessing files with attribute list attributes is not implemented yet
Anton Altaparmakovc002f422005-02-03 12:02:56 +00001547 either. But for small or simple filesystems it should work and allow
Linus Torvalds1da177e2005-04-16 15:20:36 -07001548 you to list directories, use stat on directory entries and the file
1549 system, open, read, mmap and llseek around in files. A big mile stone
1550 has been reached!
1551
1552tng-0.0.0 - Initial version tag.
1553
1554 Initial driver implementation. The driver can mount and umount simple
Anton Altaparmakovc002f422005-02-03 12:02:56 +00001555 NTFS filesystems (i.e. ones without attribute lists in the system
Linus Torvalds1da177e2005-04-16 15:20:36 -07001556 files). If the mount fails there might be problems in the error handling
1557 code paths, so be warned. Otherwise it seems to be loading the system
1558 files nicely and the mft record read mapping/unmapping seems to be
1559 working nicely, too. Proof of inode metadata in the page cache and non-
1560 resident file unnamed stream data in the page cache concepts is thus
1561 complete.