blob: cf21e0571b682e69f9a3165fcd098b96ce1709f9 [file] [log] [blame]
Linus Torvalds1da177e2005-04-16 15:20:36 -07001ToDo/Notes:
2 - Find and fix bugs.
3 - Checkpoint or disable the user space journal ($UsnJrnl).
4 - In between ntfs_prepare/commit_write, need exclusion between
Anton Altaparmakov66129f82004-11-11 12:34:00 +00005 simultaneous file extensions. This is given to us by holding i_sem
6 on the inode. The only places in the kernel when a file is resized
7 are prepare/commit write and truncate for both of which i_sem is
8 held. Just have to be careful in readpage/writepage and all other
9 helpers not running under i_sem that we play nice...
10 Also need to be careful with initialized_size extention in
Linus Torvalds1da177e2005-04-16 15:20:36 -070011 ntfs_prepare_write. Basically, just be _very_ careful in this code...
Anton Altaparmakov66129f82004-11-11 12:34:00 +000012 UPDATE: The only things that need to be checked are read/writepage
13 which do not hold i_sem. Note writepage cannot change i_size but it
14 needs to cope with a concurrent i_size change, just like readpage.
15 Also both need to cope with concurrent changes to the other sizes,
16 i.e. initialized/allocated/compressed size, as well.
Linus Torvalds1da177e2005-04-16 15:20:36 -070017 - Implement mft.c::sync_mft_mirror_umount(). We currently will just
18 leave the volume dirty on umount if the final iput(vol->mft_ino)
19 causes a write of any mirrored mft records due to the mft mirror
20 inode having been discarded already. Whether this can actually ever
21 happen is unclear however so it is worth waiting until someone hits
22 the problem.
23 - Enable the code for setting the NT4 compatibility flag when we start
24 making NTFS 1.2 specific modifications.
25
262.1.23-WIP
27
28 - Add printk rate limiting for ntfs_warning() and ntfs_error() when
29 compiled without debug. This avoids a possible denial of service
30 attack. Thanks to Carl-Daniel Hailfinger from SuSE for pointing this
31 out.
Randy Dunlap89075472005-03-03 11:19:53 +000032 - Fix compilation warnings on ia64. (Randy Dunlap)
Anton Altaparmakovda284382004-11-11 11:18:10 +000033 - Use i_size_read() in fs/ntfs/attrib.c::ntfs_attr_set().
Anton Altaparmakov66129f82004-11-11 12:34:00 +000034 - Use i_size_read() in fs/ntfs/logfile.c::ntfs_{check,empty}_logfile().
Anton Altaparmakovdb30d162004-11-11 12:42:00 +000035 - Use i_size_read() once and then use the cached value in
36 fs/ntfs/lcnalloc.c::ntfs_cluster_alloc().
Anton Altaparmakovd4b9ba72004-11-17 15:45:08 +000037 - Use i_size_read() in fs/ntfs/file.c::ntfs_file_open().
Anton Altaparmakov36763672004-11-18 13:46:45 +000038 - Add size_lock to the ntfs_inode structure. This is an rw spinlock
39 and it locks against access to the inode sizes. Note, ->size_lock
40 is also accessed from irq context so you must use the _irqsave and
41 _irqrestore lock and unlock functions, respectively.
42 - Use i_size_read() in fs/ntfs/compress.c at the start of the read and
43 use the cached value afterwards. Cache the initialized_size in the
44 same way and protect access to the two sizes using the size_lock.
Anton Altaparmakov206f9f32004-11-18 15:01:06 +000045 - Use i_size_read() in fs/ntfs/dir.c once and then use the cached
46 value afterwards.
Anton Altaparmakov218357f2004-11-18 20:34:59 +000047 - Use i_size_read() in fs/ntfs/super.c once and then use the cached
48 value afterwards. Cache the initialized_size in the same way and
49 protect access to the two sizes using the size_lock.
50 - Minor optimization to fs/ntfs/super.c::ntfs_statfs() and its helpers.
Anton Altaparmakovf50f3ac2004-11-19 22:16:00 +000051 - Use i_size_read() in fs/ntfs/inode.c once and then use the cached
52 value afterwards when reading the size of the bitmap inode.
Anton Altaparmakov07a4e2d2005-01-12 13:08:26 +000053 - Use i_size_{read,write}() in fs/ntfs/{aops.c,mft.c} and protect
54 access to the i_size and other size fields using the size_lock.
Anton Altaparmakov946929d2005-01-13 15:26:29 +000055 - Implement extension of resident files in the regular file write code
56 paths (fs/ntfs/aops.c::ntfs_{prepare,commit}_write()). At present
57 this only works until the data attribute becomes too big for the mft
58 record after which we abort the write returning -EOPNOTSUPP from
59 ntfs_prepare_write().
Anton Altaparmakovc002f422005-02-03 12:02:56 +000060 - Add disable_sparse mount option together with a per volume sparse
61 enable bit which is set appropriately and a per inode sparse disable
62 bit which is preset on some system file inodes as appropriate.
63 - Enforce that sparse support is disabled on NTFS volumes pre 3.0.
Anton Altaparmakov1a0df152005-02-03 12:04:36 +000064 - Fix a bug in fs/ntfs/runlist.c::ntfs_mapping_pairs_decompress() in
65 the creation of the unmapped runlist element for the base attribute
66 extent.
Anton Altaparmakovb6ad6c52005-02-15 10:08:43 +000067 - Split ntfs_map_runlist() into ntfs_map_runlist() and a non-locking
68 helper ntfs_map_runlist_nolock() which is used by ntfs_map_runlist().
69 This allows us to map runlist fragments with the runlist lock already
70 held without having to drop and reacquire it around the call. Adapt
71 all callers.
72 - Change ntfs_find_vcn() to ntfs_find_vcn_nolock() which takes a locked
73 runlist. This allows us to find runlist elements with the runlist
74 lock already held without having to drop and reacquire it around the
75 call. Adapt all callers.
Anton Altaparmakov37e4c132005-02-18 10:03:13 +000076 - Change time to u64 in time.h::ntfs2utc() as it otherwise generates a
Anton Altaparmakovd8ec7852005-02-18 09:23:39 +000077 warning in the do_div() call on sparc32. Thanks to Meelis Roos for
78 the report and analysis of the warning.
Anton Altaparmakov37e4c132005-02-18 10:03:13 +000079 - Fix a nasty runlist merge bug when merging two holes.
Anton Altaparmakov5ae9fcf2005-03-02 17:03:24 +000080 - Set the ntfs_inode->allocated_size to the real allocated size in the
81 mft record for resident attributes (fs/ntfs/inode.c).
82 - Small readability cleanup to use "a" instead of "ctx->attr"
83 everywhere (fs/ntfs/inode.c).
Anton Altaparmakov41382682005-03-03 13:44:15 +000084 - Make fs/ntfs/namei.c::ntfs_get_{parent,dentry} static and move the
85 definition of ntfs_export_ops from fs/ntfs/super.c to namei.c. Also,
86 declare ntfs_export_ops in fs/ntfs/ntfs.h.
Linus Torvalds1da177e2005-04-16 15:20:36 -070087
882.1.22 - Many bug and race fixes and error handling improvements.
89
90 - Improve error handling in fs/ntfs/inode.c::ntfs_truncate().
91 - Change fs/ntfs/inode.c::ntfs_truncate() to return an error code
92 instead of void and provide a helper ntfs_truncate_vfs() for the
93 vfs ->truncate method.
94 - Add a new ntfs inode flag NInoTruncateFailed() and modify
95 fs/ntfs/inode.c::ntfs_truncate() to set and clear it appropriately.
96 - Fix min_size and max_size definitions in ATTR_DEF structure in
97 fs/ntfs/layout.h to be signed.
98 - Add attribute definition handling helpers to fs/ntfs/attrib.[hc]:
99 ntfs_attr_size_bounds_check(), ntfs_attr_can_be_non_resident(), and
100 ntfs_attr_can_be_resident(), which in turn use the new private helper
101 ntfs_attr_find_in_attrdef().
102 - In fs/ntfs/aops.c::mark_ntfs_record_dirty(), take the
103 mapping->private_lock around the dirtying of the buffer heads
104 analagous to the way it is done in __set_page_dirty_buffers().
105 - Ensure the mft record size does not exceed the PAGE_CACHE_SIZE at
106 mount time as this cannot work with the current implementation.
107 - Check for location of attribute name and improve error handling in
108 general in fs/ntfs/inode.c::ntfs_read_locked_inode() and friends.
109 - In fs/ntfs/aops.c::ntfs_writepage(), if the page is fully outside
110 i_size, i.e. race with truncate, invalidate the buffers on the page
111 so that they become freeable and hence the page does not leak.
112 - Remove unused function fs/ntfs/runlist.c::ntfs_rl_merge(). (Adrian
113 Bunk)
114 - Fix stupid bug in fs/ntfs/attrib.c::ntfs_attr_find() that resulted in
115 a NULL pointer dereference in the error code path when a corrupt
116 attribute was found. (Thanks to Domen Puncer for the bug report.)
117 - Add MODULE_VERSION() to fs/ntfs/super.c.
118 - Make several functions and variables static. (Adrian Bunk)
119 - Modify fs/ntfs/aops.c::mark_ntfs_record_dirty() so it allocates
120 buffers for the page if they are not present and then marks the
121 buffers belonging to the ntfs record dirty. This causes the buffers
122 to become busy and hence they are safe from removal until the page
123 has been written out.
124 - Fix stupid bug in fs/ntfs/attrib.c::ntfs_external_attr_find() in the
125 error handling code path that resulted in a BUG() due to trying to
126 unmap an extent mft record when the mapping of it had failed and it
127 thus was not mapped. (Thanks to Ken MacFerrin for the bug report.)
128 - Drop the runlist lock after the vcn has been read in
129 fs/ntfs/lcnalloc.c::__ntfs_cluster_free().
130 - Rewrite handling of multi sector transfer errors. We now do not set
131 PageError() when such errors are detected in the async i/o handler
132 fs/ntfs/aops.c::ntfs_end_buffer_async_read(). All users of mst
133 protected attributes now check the magic of each ntfs record as they
134 use it and act appropriately. This has the effect of making errors
135 granular per ntfs record rather than per page which solves the case
136 where we cannot access any of the ntfs records in a page when a
137 single one of them had an mst error. (Thanks to Ken MacFerrin for
138 the bug report.)
139 - Fix error handling in fs/ntfs/quota.c::ntfs_mark_quotas_out_of_date()
140 where we failed to release i_sem on the $Quota/$Q attribute inode.
141 - Fix bug in handling of bad inodes in fs/ntfs/namei.c::ntfs_lookup().
142 - Add mapping of unmapped buffers to all remaining code paths, i.e.
143 fs/ntfs/aops.c::ntfs_write_mst_block(), mft.c::ntfs_sync_mft_mirror(),
144 and write_mft_record_nolock(). From now on we require that the
145 complete runlist for the mft mirror is always mapped into memory.
146 - Add creation of buffers to fs/ntfs/mft.c::ntfs_sync_mft_mirror().
147 - Improve error handling in fs/ntfs/aops.c::ntfs_{read,write}_block().
148 - Cleanup fs/ntfs/aops.c::ntfs_{read,write}page() since we know that a
149 resident attribute will be smaller than a page which makes the code
150 simpler. Also make the code more tolerant to concurrent ->truncate.
151
1522.1.21 - Fix some races and bugs, rewrite mft write code, add mft allocator.
153
154 - Implement extent mft record deallocation
155 fs/ntfs/mft.c::ntfs_extent_mft_record_free().
156 - Splitt runlist related functions off from attrib.[hc] to runlist.[hc].
157 - Add vol->mft_data_pos and initialize it at mount time.
158 - Rename init_runlist() to ntfs_init_runlist(), ntfs_vcn_to_lcn() to
159 ntfs_rl_vcn_to_lcn(), decompress_mapping_pairs() to
160 ntfs_mapping_pairs_decompress(), ntfs_merge_runlists() to
161 ntfs_runlists_merge() and adapt all callers.
162 - Add fs/ntfs/runlist.[hc]::ntfs_get_nr_significant_bytes(),
163 ntfs_get_size_for_mapping_pairs(), ntfs_write_significant_bytes(),
164 and ntfs_mapping_pairs_build(), adapted from libntfs.
165 - Make fs/ntfs/lcnalloc.c::ntfs_cluster_free_from_rl_nolock() not
166 static and add a declaration for it to lcnalloc.h.
167 - Add fs/ntfs/lcnalloc.h::ntfs_cluster_free_from_rl() which is a static
168 inline wrapper for ntfs_cluster_free_from_rl_nolock() which takes the
169 cluster bitmap lock for the duration of the call.
170 - Add fs/ntfs/attrib.[hc]::ntfs_attr_record_resize().
171 - Implement the equivalent of memset() for an ntfs attribute in
172 fs/ntfs/attrib.[hc]::ntfs_attr_set() and switch
173 fs/ntfs/logfile.c::ntfs_empty_logfile() to using it.
174 - Remove unnecessary casts from LCN_* constants.
175 - Implement fs/ntfs/runlist.c::ntfs_rl_truncate_nolock().
176 - Add MFT_RECORD_OLD as a copy of MFT_RECORD in fs/ntfs/layout.h and
177 change MFT_RECORD to contain the NTFS 3.1+ specific fields.
178 - Add a helper function fs/ntfs/aops.c::mark_ntfs_record_dirty() which
179 marks all buffers belonging to an ntfs record dirty, followed by
180 marking the page the ntfs record is in dirty and also marking the vfs
181 inode containing the ntfs record dirty (I_DIRTY_PAGES).
182 - Switch fs/ntfs/index.h::ntfs_index_entry_mark_dirty() to using the
183 new helper fs/ntfs/aops.c::mark_ntfs_record_dirty() and remove the no
184 longer needed fs/ntfs/index.[hc]::__ntfs_index_entry_mark_dirty().
185 - Move ntfs_{un,}map_page() from ntfs.h to aops.h and fix resulting
186 include errors.
187 - Move the typedefs for runlist_element and runlist from types.h to
188 runlist.h and fix resulting include errors.
189 - Remove unused {__,}format_mft_record() from fs/ntfs/mft.c.
190 - Modify fs/ntfs/mft.c::__mark_mft_record_dirty() to use the helper
191 mark_ntfs_record_dirty() which also changes the behaviour in that we
192 now set the buffers belonging to the mft record dirty as well as the
193 page itself.
194 - Update fs/ntfs/mft.c::write_mft_record_nolock() and sync_mft_mirror()
195 to cope with the fact that there now are dirty buffers in mft pages.
196 - Update fs/ntfs/inode.c::ntfs_write_inode() to also use the helper
197 mark_ntfs_record_dirty() and thus to set the buffers belonging to the
198 mft record dirty as well as the page itself.
199 - Fix compiler warnings on x86-64 in fs/ntfs/dir.c. (Randy Dunlap,
200 slightly modified by me)
201 - Add fs/ntfs/mft.c::try_map_mft_record() which fails with -EALREADY if
202 the mft record is already locked and otherwise behaves the same way
203 as fs/ntfs/mft.c::map_mft_record().
204 - Modify fs/ntfs/mft.c::write_mft_record_nolock() so that it only
205 writes the mft record if the buffers belonging to it are dirty.
206 Otherwise we assume that it was written out by other means already.
207 - Attempting to write outside initialized size is _not_ a bug so remove
208 the bug check from fs/ntfs/aops.c::ntfs_write_mst_block(). It is in
209 fact required to write outside initialized size when preparing to
210 extend the initialized size.
211 - Map the page instead of using page_address() before writing to it in
212 fs/ntfs/aops.c::ntfs_mft_writepage().
213 - Provide exclusion between opening an inode / mapping an mft record
214 and accessing the mft record in fs/ntfs/mft.c::ntfs_mft_writepage()
215 by setting the page not uptodate throughout ntfs_mft_writepage().
216 - Clear the page uptodate flag in fs/ntfs/aops.c::ntfs_write_mst_block()
217 to ensure noone can see the page whilst the mst fixups are applied.
218 - Add the helper fs/ntfs/mft.c::ntfs_may_write_mft_record() which
219 checks if an mft record may be written out safely obtaining any
220 necessary locks in the process. This is used by
221 fs/ntfs/aops.c::ntfs_write_mst_block().
222 - Modify fs/ntfs/aops.c::ntfs_write_mst_block() to also work for
223 writing mft records and improve its error handling in the process.
224 Now if any of the records in the page fail to be written out, all
225 other records will be written out instead of aborting completely.
226 - Remove ntfs_mft_aops and update all users to use ntfs_mst_aops.
227 - Modify fs/ntfs/inode.c::ntfs_read_locked_inode() to set the
228 ntfs_mst_aops for all inodes which are NInoMstProtected() and
229 ntfs_aops for all other inodes.
230 - Rename fs/ntfs/mft.c::sync_mft_mirror{,_umount}() to
231 ntfs_sync_mft_mirror{,_umount}() and change their parameters so they
232 no longer require an ntfs inode to be present. Update all callers.
233 - Cleanup the error handling in fs/ntfs/mft.c::ntfs_sync_mft_mirror().
234 - Clear the page uptodate flag in fs/ntfs/mft.c::ntfs_sync_mft_mirror()
235 to ensure noone can see the page whilst the mst fixups are applied.
236 - Remove the no longer needed fs/ntfs/mft.c::ntfs_mft_writepage() and
237 fs/ntfs/mft.c::try_map_mft_record().
238 - Fix callers of fs/ntfs/aops.c::mark_ntfs_record_dirty() to call it
239 with the ntfs inode which contains the page rather than the ntfs
240 inode the mft record of which is in the page.
241 - Fix race condition in fs/ntfs/inode.c::ntfs_put_inode() by moving the
242 index inode bitmap inode release code from there to
243 fs/ntfs/inode.c::ntfs_clear_big_inode(). (Thanks to Christoph
244 Hellwig for spotting this.)
245 - Fix race condition in fs/ntfs/inode.c::ntfs_put_inode() by taking the
246 inode semaphore around the code that sets ni->itype.index.bmp_ino to
247 NULL and reorganize the code to optimize it a bit. (Thanks to
248 Christoph Hellwig for spotting this.)
249 - Modify fs/ntfs/aops.c::mark_ntfs_record_dirty() to no longer take the
250 ntfs inode as a parameter as this is confusing and misleading and the
251 needed ntfs inode is available via NTFS_I(page->mapping->host).
252 Adapt all callers to this change.
253 - Modify fs/ntfs/mft.c::write_mft_record_nolock() and
254 fs/ntfs/aops.c::ntfs_write_mst_block() to only check the dirty state
255 of the first buffer in a record and to take this as the ntfs record
256 dirty state. We cannot look at the dirty state for subsequent
257 buffers because we might be racing with
258 fs/ntfs/aops.c::mark_ntfs_record_dirty().
259 - Move the static inline ntfs_init_big_inode() from fs/ntfs/inode.c to
260 inode.h and make fs/ntfs/inode.c::__ntfs_init_inode() non-static and
261 add a declaration for it to inode.h. Fix some compilation issues
262 that resulted due to #includes and header file interdependencies.
263 - Simplify setup of i_mode in fs/ntfs/inode.c::ntfs_read_locked_inode().
264 - Add helpers fs/ntfs/layout.h::MK_MREF() and MK_LE_MREF().
265 - Modify fs/ntfs/mft.c::map_extent_mft_record() to only verify the mft
266 record sequence number if it is specified (i.e. not zero).
267 - Add fs/ntfs/mft.[hc]::ntfs_mft_record_alloc() and various helper
268 functions used by it.
269 - Update Documentation/filesystems/ntfs.txt with instructions on how to
270 use the Device-Mapper driver with NTFS ftdisk/LDM raid. This removes
271 the linear raid problem with the Software RAID / MD driver when one
272 or more of the devices has an odd number of sectors.
273
2742.1.20 - Fix two stupid bugs introduced in 2.1.18 release.
275
276 - Fix stupid bug in fs/ntfs/attrib.c::ntfs_attr_reinit_search_ctx()
277 where we did not clear ctx->al_entry but it was still set due to
278 changes in ntfs_attr_lookup() and ntfs_external_attr_find() in
279 particular.
280 - Fix another stupid bug in fs/ntfs/attrib.c::ntfs_external_attr_find()
281 where we forgot to unmap the extent mft record when we had finished
282 enumerating an attribute which caused a bug check to trigger when the
283 VFS calls ->clear_inode.
284
2852.1.19 - Many cleanups, improvements, and a minor bug fix.
286
287 - Update ->setattr (fs/ntfs/inode.c::ntfs_setattr()) to refuse to
288 change the uid, gid, and mode of an inode as we do not support NTFS
289 ACLs yet.
290 - Remove BKL use from ntfs_setattr() syncing up with the rest of the
291 kernel.
292 - Get rid of the ugly transparent union in fs/ntfs/dir.c::ntfs_readdir()
293 and ntfs_filldir() as per suggestion from Al Viro.
294 - Change '\0' and L'\0' to simply 0 as per advice from Linus Torvalds.
295 - Update ->truncate (fs/ntfs/inode.c::ntfs_truncate()) to check if the
296 inode size has changed and to only output an error if so.
297 - Rename fs/ntfs/attrib.h::attribute_value_length() to ntfs_attr_size().
298 - Add le{16,32,64} as well as sle{16,32,64} data types to
299 fs/ntfs/types.h.
300 - Change ntfschar to be le16 instead of u16 in fs/ntfs/types.h.
301 - Add le versions of VCN, LCN, and LSN called leVCN, leLCN, and leLSN,
302 respectively, to fs/ntfs/types.h.
303 - Update endianness conversion macros in fs/ntfs/endian.h to use the
304 new types as appropriate.
305 - Do proper type casting when using sle64_to_cpup() in fs/ntfs/dir.c
306 and index.c.
307 - Add leMFT_REF data type to fs/ntfs/layout.h.
308 - Update all NTFS header files with the new little endian data types.
309 Affected files are fs/ntfs/layout.h, logfile.h, and time.h.
310 - Do proper type casting when using ntfs_is_*_recordp() in
311 fs/ntfs/logfile.c, mft.c, and super.c.
312 - Fix all the sparse bitwise warnings. Had to change all the typedef
313 enums storing little endian values to simple enums plus a typedef for
314 the datatype to make sparse happy.
315 - Fix a bug found by the new sparse bitwise warnings where the default
316 upcase table was defined as a pointer to wchar_t rather than ntfschar
317 in fs/ntfs/ntfs.h and super.c.
318 - Change {const_,}cpu_to_le{16,32}(0) to just 0 as suggested by Al Viro.
319
3202.1.18 - Fix scheduling latencies at mount time as well as an endianness bug.
321
322 - Remove vol->nr_mft_records as it was pretty meaningless and optimize
323 the calculation of total/free inodes as used by statfs().
324 - Fix scheduling latencies in ntfs_fill_super() by dropping the BKL
325 because the code itself is using the ntfs_lock semaphore which
326 provides safe locking. (Ingo Molnar)
327 - Fix a potential bug in fs/ntfs/mft.c::map_extent_mft_record() that
328 could occur in the future for when we start closing/freeing extent
329 inodes if we don't set base_ni->ext.extent_ntfs_inos to NULL after
330 we free it.
331 - Rename {find,lookup}_attr() to ntfs_attr_{find,lookup}() as well as
332 find_external_attr() to ntfs_external_attr_find() to cleanup the
333 namespace a bit and to be more consistent with libntfs.
334 - Rename {{re,}init,get,put}_attr_search_ctx() to
335 ntfs_attr_{{re,}init,get,put}_search_ctx() as well as the type
336 attr_search_context to ntfs_attr_search_ctx.
337 - Force use of ntfs_attr_find() in ntfs_attr_lookup() when searching
338 for the attribute list attribute itself.
339 - Fix endianness bug in ntfs_external_attr_find().
340 - Change ntfs_{external_,}attr_find() to return 0 on success, -ENOENT
341 if the attribute is not found, and -EIO on real error. In the case
342 of -ENOENT, the search context is updated to describe the attribute
343 before which the attribute being searched for would need to be
344 inserted if such an action were to be desired and in the case of
345 ntfs_external_attr_find() the search context is also updated to
346 indicate the attribute list entry before which the attribute list
347 entry of the attribute being searched for would need to be inserted
348 if such an action were to be desired. Also make ntfs_find_attr()
349 static and remove its prototype from attrib.h as it is not used
350 anywhere other than attrib.c. Update ntfs_attr_lookup() and all
351 callers of ntfs_{external,}attr_{find,lookup}() for the new return
352 values.
353 - Minor cleanup of fs/ntfs/inode.c::ntfs_init_locked_inode().
354
3552.1.17 - Fix bugs in mount time error code paths and other updates.
356
357 - Implement bitmap modification code (fs/ntfs/bitmap.[hc]). This
358 includes functions to set/clear a single bit or a run of bits.
359 - Add fs/ntfs/attrib.[hc]::ntfs_find_vcn() which returns the locked
360 runlist element containing a particular vcn. It also takes care of
361 mapping any needed runlist fragments.
362 - Implement cluster (de-)allocation code (fs/ntfs/lcnalloc.[hc]).
363 - Load attribute definition table from $AttrDef at mount time.
364 - Fix bugs in mount time error code paths involving (de)allocation of
365 the default and volume upcase tables.
366 - Remove ntfs_nr_mounts as it is no longer used.
367
3682.1.16 - Implement access time updates, file sync, async io, and read/writev.
369
370 - Add support for readv/writev and aio_read/aio_write (fs/ntfs/file.c).
371 This is done by setting the appropriate file operations pointers to
372 the generic helper functions provided by mm/filemap.c.
373 - Implement fsync, fdatasync, and msync both for files (fs/ntfs/file.c)
374 and directories (fs/ntfs/dir.c).
375 - Add support for {a,m,c}time updates to inode.c::ntfs_write_inode().
376 Note, except for the root directory and any other system files opened
377 by the user, the system files will not have their access times
378 updated as they are only accessed at the inode level an hence the
379 file level functions which cause the times to be updated are never
380 invoked.
381
3822.1.15 - Invalidate quotas when (re)mounting read-write.
383
384 - Add new element itype.index.collation_rule to the ntfs inode
385 structure and set it appropriately in ntfs_read_locked_inode().
386 - Implement a new inode type "index" to allow efficient access to the
387 indices found in various system files and adapt inode handling
388 accordingly (fs/ntfs/inode.[hc]). An index inode is essentially an
389 attribute inode (NInoAttr() is true) with an attribute type of
390 AT_INDEX_ALLOCATION. As such, it is no longer allowed to call
391 ntfs_attr_iget() with an attribute type of AT_INDEX_ALLOCATION as
392 there would be no way to distinguish between normal attribute inodes
393 and index inodes. The function to obtain an index inode is
394 ntfs_index_iget() and it uses the helper function
395 ntfs_read_locked_index_inode(). Note, we do not overload
396 ntfs_attr_iget() as indices consist of multiple attributes so using
397 ntfs_attr_iget() to obtain an index inode would be confusing.
398 - Ensure that there is no overflow when doing page->index <<
399 PAGE_CACHE_SHIFT by casting page->index to s64 in fs/ntfs/aops.c.
400 - Use atomic kmap instead of kmap() in fs/ntfs/aops.c::ntfs_read_page()
401 and ntfs_read_block().
402 - Use case sensitive attribute lookups instead of case insensitive ones.
403 - Lock all page cache pages belonging to mst protected attributes while
404 accessing them to ensure we never see corrupt data while the page is
405 under writeout.
406 - Add framework for generic ntfs collation (fs/ntfs/collation.[hc]).
407 We have ntfs_is_collation_rule_supported() to check if the collation
408 rule you want to use is supported and ntfs_collation() which actually
409 collates two data items. We currently only support COLLATION_BINARY
410 and COLLATION_NTOFS_ULONG but support for other collation rules will
411 be added as the need arises.
412 - Add a new type, ntfs_index_context, to allow retrieval of an index
413 entry using the corresponding index key. To get an index context,
414 use ntfs_index_ctx_get() and to release it, use ntfs_index_ctx_put().
415 This also adds a new slab cache for the index contexts. To lookup a
416 key in an index inode, use ntfs_index_lookup(). After modifying an
417 index entry, call ntfs_index_entry_flush_dcache_page() followed by
418 ntfs_index_entry_mark_dirty() to ensure the changes are written out
419 to disk. For details see fs/ntfs/index.[hc]. Note, at present, if
420 an index entry is in the index allocation attribute rather than the
421 index root attribute it will not be written out (you will get a
422 warning message about discarded changes instead).
423 - Load the quota file ($Quota) and check if quota tracking is enabled
424 and if so, mark the quotas out of date. This causes windows to
425 rescan the volume on boot and update all quota entries.
426 - Add a set_page_dirty address space operation for ntfs_m[fs]t_aops.
427 It is simply set to __set_page_dirty_nobuffers() to make sure that
428 running set_page_dirty() on a page containing mft/ntfs records will
429 not affect the dirty state of the page buffers.
430 - Add fs/ntfs/index.c::__ntfs_index_entry_mark_dirty() which sets all
431 buffers that are inside the ntfs record in the page dirty after which
432 it sets the page dirty. This allows ->writepage to only write the
433 dirty index records rather than having to write all the records in
434 the page. Modify fs/ntfs/index.h::ntfs_index_entry_mark_dirty() to
435 use this rather than __set_page_dirty_nobuffers().
436 - Implement fs/ntfs/aops.c::ntfs_write_mst_block() which enables the
437 writing of page cache pages belonging to mst protected attributes
438 like the index allocation attribute in directory indices and other
439 indices like $Quota/$Q, etc. This means that the quota is now marked
440 out of date on all volumes rather than only on ones where the quota
441 defaults entry is in the index root attribute of the $Quota/$Q index.
442
4432.1.14 - Fix an NFSd caused deadlock reported by several users.
444
445 - Modify fs/ntfs/ntfs_readdir() to copy the index root attribute value
446 to a buffer so that we can put the search context and unmap the mft
447 record before calling the filldir() callback. We need to do this
448 because of NFSd which calls ->lookup() from its filldir callback()
449 and this causes NTFS to deadlock as ntfs_lookup() maps the mft record
450 of the directory and since ntfs_readdir() has got it mapped already
451 ntfs_lookup() deadlocks.
452
4532.1.13 - Enable overwriting of resident files and housekeeping of system files.
454
455 - Implement writing of mft records (fs/ntfs/mft.[hc]), which includes
456 keeping the mft mirror in sync with the mft when mirrored mft records
457 are written. The functions are write_mft_record{,_nolock}(). The
458 implementation is quite rudimentary for now with lots of things not
459 implemented yet but I am not sure any of them can actually occur so
460 I will wait for people to hit each one and only then implement it.
461 - Commit open system inodes at umount time. This should make it
462 virtually impossible for sync_mft_mirror_umount() to ever be needed.
463 - Implement ->write_inode (fs/ntfs/inode.c::ntfs_write_inode()) for the
464 ntfs super operations. This gives us inode writing via the VFS inode
465 dirty code paths. Note: Access time updates are not implemented yet.
466 - Implement fs/ntfs/mft.[hc]::{,__}mark_mft_record_dirty() and make
467 fs/ntfs/aops.c::ntfs_writepage() and ntfs_commit_write() use it, thus
468 finally enabling resident file overwrite! (-8 This also includes a
469 placeholder for ->writepage (ntfs_mft_writepage()), which for now
470 just redirties the page and returns. Also, at umount time, we for
471 now throw away all mft data page cache pages after the last call to
472 ntfs_commit_inode() in the hope that all inodes will have been
473 written out by then and hence no dirty (meta)data will be lost. We
474 also check for this case and emit an error message telling the user
475 to run chkdsk.
476 - Use set_page_writeback() and end_page_writeback() in the resident
477 attribute code path of fs/ntfs/aops.c::ntfs_writepage() otherwise
478 the radix-tree tag PAGECACHE_TAG_DIRTY remains set even though the
479 page is clean.
480 - Implement ntfs_mft_writepage() so it now checks if any of the mft
481 records in the page are dirty and if so redirties the page and
482 returns. Otherwise it just returns (after doing set_page_writeback(),
483 unlock_page(), end_page_writeback() or the radix-tree tag
484 PAGECACHE_TAG_DIRTY remains set even though the page is clean), thus
485 alowing the VM to do with the page as it pleases. Also, at umount
486 time, now only throw away dirty mft (meta)data pages if dirty inodes
487 are present and ask the user to email us if they see this happening.
488 - Add functions ntfs_{clear,set}_volume_flags(), to modify the volume
489 information flags (fs/ntfs/super.c).
490 - Mark the volume dirty when (re)mounting read-write and mark it clean
491 when unmounting or remounting read-only. If any volume errors are
492 found, the volume is left marked dirty to force chkdsk to run.
493 - Add code to set the NT4 compatibility flag when (re)mounting
494 read-write for newer NTFS versions but leave it commented out for now
495 since we do not make any modifications that are NTFS 1.2 specific yet
496 and since setting this flag breaks Captive-NTFS which is not nice.
497 This code must be enabled once we start writing NTFS 1.2 specific
498 changes otherwise Windows NTFS driver might crash / cause corruption.
499
5002.1.12 - Fix the second fix to the decompression engine and some cleanups.
501
502 - Add a new address space operations struct, ntfs_mst_aops, for mst
503 protected attributes. This is because the default ntfs_aops do not
504 make sense with mst protected data and were they to write anything to
505 such an attribute they would cause data corruption so we provide
506 ntfs_mst_aops which does not have any write related operations set.
507 - Cleanup dirty ntfs inode handling (fs/ntfs/inode.[hc]) which also
508 includes an adapted ntfs_commit_inode() and an implementation of
509 ntfs_write_inode() which for now just cleans dirty inodes without
510 writing them (it does emit a warning that this is happening).
511 - Undo the second decompression engine fix (see 2.1.9 release ChangeLog
512 entry) as it was only fixing a theoretical bug but at the same time
513 it badly broke the handling of sparse and uncompressed compression
514 blocks.
515
5162.1.11 - Driver internal cleanups.
517
518 - Only build logfile.o if building the driver with read-write support.
519 - Really final white space cleanups.
520 - Use generic_ffs() instead of ffs() in logfile.c which allows the
521 log_page_size variable to be optimized by gcc into a constant.
522 - Rename uchar_t to ntfschar everywhere as uchar_t is unsigned 1-byte
523 char as defined by POSIX and as found on some systems.
524
5252.1.10 - Force read-only (re)mounting of volumes with unsupported volume flags.
526
527 - Finish off the white space cleanups (remove trailing spaces, etc).
528 - Clean up ntfs_fill_super() and ntfs_read_inode_mount() by removing
529 the kludges around the first iget(). Instead of (re)setting ->s_op
530 we have the $MFT inode set up by explicit new_inode() / set ->i_ino /
531 insert_inode_hash() / call ntfs_read_inode_mount() directly. This
532 kills the need for second super_operations and allows to return error
533 from ntfs_read_inode_mount() without resorting to ugly "poisoning"
534 tricks. (Al Viro)
535 - Force read-only (re)mounting if any of the following bits are set in
536 the volume information flags:
537 VOLUME_IS_DIRTY, VOLUME_RESIZE_LOG_FILE,
538 VOLUME_UPGRADE_ON_MOUNT, VOLUME_DELETE_USN_UNDERWAY,
539 VOLUME_REPAIR_OBJECT_ID, VOLUME_MODIFIED_BY_CHKDSK
540 To make this easier we define VOLUME_MUST_MOUNT_RO_MASK with all the
541 above bits set so the test is made easy.
542
5432.1.9 - Fix two bugs in decompression engine.
544
545 - Fix a bug where we would not always detect that we have reached the
546 end of a compression block because we were ending at minus one byte
547 which is effectively the same as being at the end. The fix is to
548 check whether the uncompressed buffer has been fully filled and if so
549 we assume we have reached the end of the compression block. A big
550 thank you to Marcin Gibuła for the bug report, the assistance in
551 tracking down the bug and testing the fix.
552 - Fix a possible bug where when a compressed read is truncated to the
553 end of the file, the offset inside the last page was not truncated.
554
5552.1.8 - Handle $MFT mirror and $LogFile, improve time handling, and cleanups.
556
557 - Use get_bh() instead of manual atomic_inc() in fs/ntfs/compress.c.
558 - Modify fs/ntfs/time.c::ntfs2utc(), get_current_ntfs_time(), and
559 utc2ntfs() to work with struct timespec instead of time_t on the
560 Linux UTC time side thus preserving the full precision of the NTFS
561 time and only loosing up to 99 nano-seconds in the Linux UTC time.
562 - Move fs/ntfs/time.c to fs/ntfs/time.h and make the time functions
563 static inline.
564 - Remove unused ntfs_dirty_inode().
565 - Cleanup super operations declaration in fs/ntfs/super.c.
566 - Wrap flush_dcache_mft_record_page() in #ifdef NTFS_RW.
567 - Add NInoTestSetFoo() and NInoTestClearFoo() macro magic to
568 fs/ntfs/inode.h and use it to declare NInoTest{Set,Clear}Dirty.
569 - Move typedefs for ntfs_attr and test_t from fs/ntfs/inode.c to
570 fs/ntfs/inode.h so they can be used elsewhere.
571 - Determine the mft mirror size as the number of mirrored mft records
572 and store it in ntfs_volume->mftmirr_size (fs/ntfs/super.c).
573 - Load the mft mirror at mount time and compare the mft records stored
574 in it to the ones in the mft. Force a read-only mount if the two do
575 not match (fs/ntfs/super.c).
576 - Fix type casting related warnings on 64-bit architectures. Thanks
577 to Meelis Roos for reporting them.
578 - Move %L to %ll as %L is floating point and %ll is integer which is
579 what we want.
580 - Read the journal ($LogFile) and determine if the volume has been
581 shutdown cleanly and force a read-only mount if not (fs/ntfs/super.c
582 and fs/ntfs/logfile.c). This is a little bit of a crude check in
583 that we only look at the restart areas and not at the actual log
584 records so that there will be a very small number of cases where we
585 think that a volume is dirty when in fact it is clean. This should
586 only affect volumes that have not been shutdown cleanly and did not
587 have any pending, non-check-pointed i/o.
588 - If the $LogFile indicates a clean shutdown and a read-write (re)mount
589 is requested, empty $LogFile by overwriting it with 0xff bytes to
590 ensure that Windows cannot cause data corruption by replaying a stale
591 journal after Linux has written to the volume.
592
5932.1.7 - Enable NFS exporting of mounted NTFS volumes.
594
595 - Set i_generation in the VFS inode from the seq_no of the NTFS inode.
596 - Make ntfs_lookup() NFS export safe, i.e. use d_splice_alias(), etc.
597 - Implement ->get_dentry() in fs/ntfs/namei.c::ntfs_get_dentry() as the
598 default doesn't allow inode number 0 which is a valid inode on NTFS
599 and even if it did allow that it uses iget() instead of ntfs_iget()
600 which makes it useless for us.
601 - Implement ->get_parent() in fs/ntfs/namei.c::ntfs_get_parent() as the
602 default just returns -EACCES which is not very useful.
603 - Define export operations (->s_export_op) for NTFS (ntfs_export_ops)
604 and set them up in the super block at mount time (super.c) this
605 allows mounted NTFS volumes to be exported via NFS.
606 - Add missing return -EOPNOTSUPP; in
607 fs/ntfs/aops.c::ntfs_commit_nonresident_write().
608 - Enforce no atime and no dir atime updates at mount/remount time as
609 they are not implemented yet anyway.
610 - Move a few assignments in fs/ntfs/attrib.c::load_attribute_list() to
611 after a NULL check. Thanks to Dave Jones for pointing this out.
612
6132.1.6 - Fix minor bug in handling of compressed directories.
614
615 - Fix bug in handling of compressed directories. A compressed
616 directory is not really compressed so when we set the ->i_blocks
617 field of a compressed directory inode we were setting it from the
618 non-existing field ni->itype.compressed.size which gave random
619 results... For directories we now always use ni->allocated_size.
620
6212.1.5 - Fix minor bug in attribute list attribute handling.
622
623 - Fix bug in attribute list handling. Actually it is not as much a bug
624 as too much protection in that we were not allowing attribute lists
625 which waste space on disk while Windows XP clearly allows it and in
626 fact creates such attribute lists so our driver was failing.
627 - Update NTFS documentation ready for 2.6 kernel release.
628
6292.1.4 - Reduce compiler requirements.
630
631 - Remove all uses of unnamed structs and unions in the driver to make
632 old and newer gcc versions happy. Makes it a bit uglier IMO but at
633 least people will stop hassling me about it.
634
6352.1.3 - Important bug fixes in corner cases.
636
637 - super.c::parse_ntfs_boot_sector(): Correct the check for 64-bit
638 clusters. (Philipp Thomas)
639 - attrib.c::load_attribute_list(): Fix bug when initialized_size is a
640 multiple of the block_size but not the cluster size. (Szabolcs
641 Szakacsits <szaka@sienet.hu>)
642
6432.1.2 - Important bug fixes aleviating the hangs in statfs.
644
645 - Fix buggy free cluster and free inode determination logic.
646
6472.1.1 - Minor updates.
648
649 - Add handling for initialized_size != data_size in compressed files.
650 - Reduce function local stack usage from 0x3d4 bytes to just noise in
651 fs/ntfs/upcase.c. (Randy Dunlap <rddunlap@osdl.ord>)
652 - Remove compiler warnings for newer gcc.
653 - Pages are no longer kmapped by mm/filemap.c::generic_file_write()
654 around calls to ->{prepare,commit}_write. Adapt NTFS appropriately
655 in fs/ntfs/aops.c::ntfs_prepare_nonresident_write() by using
656 kmap_atomic(KM_USER0).
657
6582.1.0 - First steps towards write support: implement file overwrite.
659
660 - Add configuration option for developmental write support with an
661 appropriately scary configuration help text.
662 - Initial implementation of fs/ntfs/aops.c::ntfs_writepage() and its
663 helper fs/ntfs/aops.c::ntfs_write_block(). This enables mmap(2) based
664 overwriting of existing files on ntfs. Note: Resident files are
665 only written into memory, and not written out to disk at present, so
666 avoid writing to files smaller than about 1kiB.
667 - Initial implementation of fs/ntfs/aops.c::ntfs_prepare_write(), its
668 helper fs/ntfs/aops.c::ntfs_prepare_nonresident_write() and their
669 counterparts, fs/ntfs/aops.c::ntfs_commit_write(), and
670 fs/ntfs/aops.c::ntfs_commit_nonresident_write(), respectively. Also,
671 add generic_file_write() to the ntfs file operations (fs/ntfs/file.c).
672 This enables write(2) based overwriting of existing files on ntfs.
673 Note: As with mmap(2) based overwriting, resident files are only
674 written into memory, and not written out to disk at present, so avoid
675 writing to files smaller than about 1kiB.
676 - Implement ->truncate (fs/ntfs/inode.c::ntfs_truncate()) and
677 ->setattr() (fs/ntfs/inode.c::ntfs_setattr()) inode operations for
678 files with the purpose of intercepting and aborting all i_size
679 changes which we do not support yet. ntfs_truncate() actually only
680 emits a warning message but AFAICS our interception of i_size changes
681 elsewhere means ntfs_truncate() never gets called for i_size changes.
682 It is only called from generic_file_write() when we fail in
683 ntfs_prepare_{,nonresident_}write() in order to discard any
684 instantiated buffers beyond i_size. Thus i_size is not actually
685 changed so our warning message is enough. Unfortunately it is not
686 possible to easily determine if i_size is being changed or not hence
687 we just emit an appropriately worded error message.
688
6892.0.25 - Small bug fixes and cleanups.
690
691 - Unlock the page in an out of memory error code path in
692 fs/ntfs/aops.c::ntfs_read_block().
693 - If fs/ntfs/aops.c::ntfs_read_page() is called on an uptodate page,
694 just unlock the page and return. (This can happen due to ->writepage
695 clearing PageUptodate() during write out of MstProtected()
696 attributes.
697 - Remove leaked write code again.
698
6992.0.24 - Cleanups.
700
701 - Treat BUG_ON() as ASSERT() not VERIFY(), i.e. do not use side effects
702 inside BUG_ON(). (Adam J. Richter)
703 - Split logical OR expressions inside BUG_ON() into individual BUG_ON()
704 calls for improved debugging. (Adam J. Richter)
705 - Add errors flag to the ntfs volume state, accessed via
706 NVol{,Set,Clear}Errors(vol).
707 - Do not allow read-write remounts of read-only volumes with errors.
708 - Clarify comment for ntfs file operation sendfile which was added by
709 Christoph Hellwig a while ago (just using generic_file_sendfile())
710 to say that ntfs ->sendfile is only used for the case where the
711 source data is on the ntfs partition and the destination is
712 somewhere else, i.e. nothing we need to concern ourselves with.
713 - Add generic_file_write() as our ntfs file write operation.
714
7152.0.23 - Major bug fixes (races, deadlocks, non-i386 architectures).
716
717 - Massive internal locking changes to mft record locking. Fixes lock
718 recursion and replaces the mrec_lock read/write semaphore with a
719 mutex. Also removes the now superfluous mft_count. This fixes several
720 race conditions and deadlocks, especially in the future write code.
721 - Fix ntfs over loopback for compressed files by adding an
722 optimization barrier. (gcc was screwing up otherwise ?)
723 - Miscellaneous cleanups all over the code and a fix or two in error
724 handling code paths.
725 Thanks go to Christoph Hellwig for pointing out the following two:
726 - Remove now unused function fs/ntfs/malloc.h::vmalloc_nofs().
727 - Fix ntfs_free() for ia64 and parisc by checking for VMALLOC_END, too.
728
7292.0.22 - Cleanups, mainly to ntfs_readdir(), and use C99 initializers.
730
731 - Change fs/ntfs/dir.c::ntfs_reddir() to only read/write ->f_pos once
732 at entry/exit respectively.
733 - Use C99 initializers for structures.
734 - Remove unused variable blocks from fs/ntfs/aops.c::ntfs_read_block().
735
7362.0.21 - Check for, and refuse to work with too large files/directories/volumes.
737
738 - Limit volume size at mount time to 2TiB on architectures where
739 unsigned long is 32-bits (fs/ntfs/super.c::parse_ntfs_boot_sector()).
740 This is the most we can do without overflowing the 32-bit limit of
741 the block device size imposed on us by sb_bread() and sb_getblk()
742 for the time being.
743 - Limit file/directory size at open() time to 16TiB on architectures
744 where unsigned long is 32-bits (fs/ntfs/file.c::ntfs_file_open() and
745 fs/ntfs/dir.c::ntfs_dir_open()). This is the most we can do without
746 overflowing the page cache page index.
747
7482.0.20 - Support non-resident directory index bitmaps, fix page leak in readdir.
749
750 - Move the directory index bitmap to use an attribute inode instead of
751 having special fields for it inside the ntfs inode structure. This
752 means that the index bitmaps now use the page cache for i/o, too,
753 and also as a side effect we get support for non-resident index
754 bitmaps for free.
755 - Simplify/cleanup error handling in fs/ntfs/dir.c::ntfs_readdir() and
756 fix a page leak that manifested itself in some cases.
757 - Add fs/ntfs/inode.c::ntfs_put_inode(), which we need to release the
758 index bitmap inode on the final iput().
759
7602.0.19 - Fix race condition, improvements, and optimizations in i/o interface.
761
762 - Apply block optimization added to fs/ntfs/aops.c::ntfs_read_block()
763 to fs/ntfs/compress.c::ntfs_file_read_compressed_block() as well.
764 - Drop the "file" from ntfs_file_read_compressed_block().
765 - Rename fs/ntfs/aops.c::ntfs_enb_buffer_read_async() to
766 ntfs_end_buffer_async_read() (more like the fs/buffer.c counterpart).
767 - Update ntfs_end_buffer_async_read() with the improved logic from
768 its updated counterpart fs/buffer.c::end_buffer_async_read(). Apply
769 further logic improvements to better determine when we set PageError.
770 - Update submission of buffers in fs/ntfs/aops.c::ntfs_read_block() to
771 check for the buffers being uptodate first in line with the updated
772 fs/buffer.c::block_read_full_page(). This plugs a small race
773 condition.
774
7752.0.18 - Fix race condition in reading of compressed files.
776
777 - There was a narrow window between checking a buffer head for being
778 uptodate and locking it in ntfs_file_read_compressed_block(). We now
779 lock the buffer and then check whether it is uptodate or not.
780
7812.0.17 - Cleanups and optimizations - shrinking the ToDo list.
782
783 - Modify fs/ntfs/inode.c::ntfs_read_locked_inode() to return an error
784 code and update callers, i.e. ntfs_iget(), to pass that error code
785 up instead of just using -EIO.
786 - Modifications to super.c to ensure that both mount and remount
787 cannot set any write related options when the driver is compiled
788 read-only.
789 - Optimize block resolution in fs/ntfs/aops.c::ntfs_read_block() to
790 cache the current runlist element. This should improve performance
791 when reading very large and/or very fragmented data.
792
7932.0.16 - Convert access to $MFT/$BITMAP to attribute inode API.
794
795 - Fix a stupid bug introduced in 2.0.15 where we were unmapping the
796 wrong inode in fs/ntfs/inode.c::ntfs_attr_iget().
797 - Fix debugging check in fs/ntfs/aops.c::ntfs_read_block().
798 - Convert $MFT/$BITMAP access to attribute inode API and remove all
799 remnants of the ugly mftbmp address space and operations hack. This
800 means we finally have only one readpage function as well as only one
801 async io completion handler. Yey! The mft bitmap is now just an
802 attribute inode and is accessed from vol->mftbmp_ino just as if it
803 were a normal file. Fake inodes rule. (-:
804
8052.0.15 - Fake inodes based attribute i/o via the pagecache, fixes and cleanups.
806
807 - Fix silly bug in fs/ntfs/super.c::parse_options() which was causing
808 remounts to fail when the partition had an entry in /etc/fstab and
809 the entry specified the nls= option.
810 - Apply same macro magic used in fs/ntfs/inode.h to fs/ntfs/volume.h to
811 expand all the helper functions NVolFoo(), NVolSetFoo(), and
812 NVolClearFoo().
813 - Move copyright statement from driver initialisation message to
814 module description (fs/super.c). This makes the initialisation
815 message fit on one line and fits in better with rest of kernel.
816 - Update fs/ntfs/attrib.c::map_run_list() to work on both real and
817 attribute inodes, and both for files and directories.
818 - Implement fake attribute inodes allowing all attribute i/o to go via
819 the page cache and to use all the normal vfs/mm functionality:
820 - Add ntfs_attr_iget() and its helper ntfs_read_locked_attr_inode()
821 to fs/ntfs/inode.c.
822 - Add needed cleanup code to ntfs_clear_big_inode().
823 - Merge address space operations for files and directories (aops.c),
824 now just have ntfs_aops:
825 - Rename:
826 end_buffer_read_attr_async() -> ntfs_end_buffer_read_async(),
827 ntfs_attr_read_block() -> ntfs_read_block(),
828 ntfs_file_read_page() -> ntfs_readpage().
829 - Rewrite fs/ntfs/aops.c::ntfs_readpage() to work on both real and
830 attribute inodes, and both for files and directories.
831 - Remove obsolete fs/ntfs/aops.c::ntfs_mst_readpage().
832
8332.0.14 - Run list merging code cleanup, minor locking changes, typo fixes.
834
835 - Change fs/ntfs/super.c::ntfs_statfs() to not rely on BKL by moving
836 the locking out of super.c::get_nr_free_mft_records() and taking and
837 dropping the mftbmp_lock rw_semaphore in ntfs_statfs() itself.
838 - Bring attribute runlist merging code (fs/ntfs/attrib.c) in sync with
839 current userspace ntfs library code. This means that if a merge
840 fails the original runlists are always left unmodified instead of
841 being silently corrupted.
842 - Misc typo fixes.
843
8442.0.13 - Use iget5_locked() in preparation for fake inodes and small cleanups.
845
846 - Remove nr_mft_bits and the now superfluous union with nr_mft_records
847 from ntfs_volume structure.
848 - Remove nr_lcn_bits and the now superfluous union with nr_clusters
849 from ntfs_volume structure.
850 - Use iget5_locked() and friends instead of conventional iget(). Wrap
851 the call in fs/ntfs/inode.c::ntfs_iget() and update callers of iget()
852 to use ntfs_iget(). Leave only one iget() call at mount time so we
853 don't need an ntfs_iget_mount().
854 - Change fs/ntfs/inode.c::ntfs_new_extent_inode() to take mft_no as an
855 additional argument.
856
8572.0.12 - Initial cleanup of address space operations following 2.0.11 changes.
858
859 - Merge fs/ntfs/aops.c::end_buffer_read_mst_async() and
860 fs/ntfs/aops.c::end_buffer_read_file_async() into one function
861 fs/ntfs/aops.c::end_buffer_read_attr_async() using NInoMstProtected()
862 to determine whether to apply mst fixups or not.
863 - Above change allows merging fs/ntfs/aops.c::ntfs_file_read_block()
864 and fs/ntfs/aops.c::ntfs_mst_readpage() into one function
865 fs/ntfs/aops.c::ntfs_attr_read_block(). Also, create a tiny wrapper
866 fs/ntfs/aops.c::ntfs_mst_readpage() to transform the parameters from
867 the VFS readpage function prototype to the ntfs_attr_read_block()
868 function prototype.
869
8702.0.11 - Initial preparations for fake inode based attribute i/o.
871
872 - Move definition of ntfs_inode_state_bits to fs/ntfs/inode.h and
873 do some macro magic (adapted from include/linux/buffer_head.h) to
874 expand all the helper functions NInoFoo(), NInoSetFoo(), and
875 NInoClearFoo().
876 - Add new flag to ntfs_inode_state_bits: NI_Sparse.
877 - Add new fields to ntfs_inode structure to allow use of fake inodes
878 for attribute i/o: type, name, name_len. Also add new state bits:
879 NI_Attr, which, if set, indicates the inode is a fake inode, and
880 NI_MstProtected, which, if set, indicates the attribute uses multi
881 sector transfer protection, i.e. fixups need to be applied after
882 reads and before/after writes.
883 - Rename fs/ntfs/inode.c::ntfs_{new,clear,destroy}_inode() to
884 ntfs_{new,clear,destroy}_extent_inode() and update callers.
885 - Use ntfs_clear_extent_inode() in fs/ntfs/inode.c::__ntfs_clear_inode()
886 instead of ntfs_destroy_extent_inode().
887 - Cleanup memory deallocations in {__,}ntfs_clear_{,big_}inode().
888 - Make all operations on ntfs inode state bits use the NIno* functions.
889 - Set up the new ntfs inode fields and state bits in
890 fs/ntfs/inode.c::ntfs_read_inode() and add appropriate cleanup of
891 allocated memory to __ntfs_clear_inode().
892 - Cleanup ntfs_inode structure a bit for better ordering of elements
893 w.r.t. their size to allow better packing of the structure in memory.
894
8952.0.10 - There can only be 2^32 - 1 inodes on an NTFS volume.
896
897 - Add check at mount time to verify that the number of inodes on the
898 volume does not exceed 2^32 - 1, which is the maximum allowed for
899 NTFS according to Microsoft.
900 - Change mft_no member of ntfs_inode structure to be unsigned long.
901 Update all users. This makes ntfs_inode->mft_no just a copy of struct
902 inode->i_ino. But we can't just always use struct inode->i_ino and
903 remove mft_no because extent inodes do not have an attached struct
904 inode.
905
9062.0.9 - Decompression engine now uses a single buffer and other cleanups.
907
908 - Change decompression engine to use a single buffer protected by a
909 spin lock instead of per-CPU buffers. (Rusty Russell)
910 - Do not update cb_pos when handling a partial final page during
911 decompression of a sparse compression block, as the value is later
912 reset without being read/used. (Rusty Russell)
913 - Switch to using the new KM_BIO_SRC_IRQ for atomic kmap()s. (Andrew
914 Morton)
915 - Change buffer size in ntfs_readdir()/ntfs_filldir() to use
916 NLS_MAX_CHARSET_SIZE which makes the buffers almost 1kiB each but
917 it also makes everything safer so it is a good thing.
918 - Miscellaneous minor cleanups to comments.
919
9202.0.8 - Major updates for handling of case sensitivity and dcache aliasing.
921
922 Big thanks go to Al Viro and other inhabitants of #kernel for investing
923 their time to discuss the case sensitivity and dcache aliasing issues.
924
925 - Remove unused source file fs/ntfs/attraops.c.
926 - Remove show_inodes mount option(s), thus dropping support for
927 displaying of short file names.
928 - Remove deprecated mount option posix.
929 - Restore show_sys_files mount option.
930 - Add new mount option case_sensitive, to determine if the driver
931 treats file names as case sensitive or not. If case sensitive, create
932 file names in the POSIX namespace. Otherwise create file names in the
933 LONG/WIN32 namespace. Note, files remain accessible via their short
934 file name, if it exists.
935 - Remove really dumb logic bug in boot sector recovery code.
936 - Fix dcache aliasing issues wrt short/long file names via changes
937 to fs/ntfs/dir.c::ntfs_lookup_inode_by_name() and
938 fs/ntfs/namei.c::ntfs_lookup():
939 - Add additional argument to ntfs_lookup_inode_by_name() in which we
940 return information about the matching file name if the case is not
941 matching or the match is a short file name. See comments above the
942 function definition for details.
943 - Change ntfs_lookup() to only create dcache entries for the correctly
944 cased file name and only for the WIN32 namespace counterpart of DOS
945 namespace file names. This ensures we have only one dentry per
946 directory and also removes all dcache aliasing issues between short
947 and long file names once we add write support. See comments above
948 function for details.
949 - Fix potential 1 byte overflow in fs/ntfs/unistr.c::ntfs_ucstonls().
950
9512.0.7 - Minor cleanups and updates for changes in core kernel code.
952
953 - Remove much of the NULL struct element initializers.
954 - Various updates to make compatible with recent kernels.
955 - Remove defines of MAX_BUF_PER_PAGE and include linux/buffer_head.h
956 in fs/ntfs/ntfs.h instead.
957 - Remove no longer needed KERNEL_VERSION checks. We are now in the
958 kernel proper so they are no longer needed.
959
9602.0.6 - Major bugfix to make compatible with other kernel changes.
961
962 - Initialize the mftbmp address space properly now that there are more
963 fields in the struct address_space. This was leading to hangs and
964 oopses on umount since 2.5.12 because of changes to other parts of
965 the kernel. We probably want a kernel generic init_address_space()
966 function...
967 - Drop BKL from ntfs_readdir() after consultation with Al Viro. The
968 only caller of ->readdir() is vfs_readdir() which holds i_sem during
969 the call, and i_sem is sufficient protection against changes in the
970 directory inode (including ->i_size).
971 - Use generic_file_llseek() for directories (as opposed to
972 default_llseek()) as this downs i_sem instead of the BKL which is
973 what we now need for exclusion against ->f_pos changes considering we
974 no longer take the BKL in ntfs_readdir().
975
9762.0.5 - Major bugfix. Buffer overflow in extent inode handling.
977
978 - No need to set old blocksize in super.c::ntfs_fill_super() as the
979 VFS does so via invocation of deactivate_super() calling
980 fs->fill_super() calling block_kill_super() which does it.
981 - BKL moved from VFS into dir.c::ntfs_readdir(). (Linus Torvalds)
982 -> Do we really need it? I don't think so as we have exclusion on
983 the directory ntfs_inode rw_semaphore mrec_lock. We mmight have to
984 move the ->f_pos accesses under the mrec_lock though. Check this...
985 - Fix really, really, really stupid buffer overflow in extent inode
986 handling in mft.c::map_extent_mft_record().
987
9882.0.4 - Cleanups and updates for kernel 2.5.11.
989
990 - Add documentation on how to use the MD driver to be able to use NTFS
991 stripe and volume sets in Linux and generally cleanup documentation
992 a bit.
993 Remove all uses of kdev_t in favour of struct block_device *:
994 - Change compress.c::ntfs_file_read_compressed_block() to use
995 sb_getblk() instead of getblk().
996 - Change super.c::ntfs_fill_super() to use bdev_hardsect_size() instead
997 of get_hardsect_size().
998 - No need to get old blocksize in super.c::ntfs_fill_super() as
999 fs/super.c::get_sb_bdev() already does this.
1000 - Set bh->b_bdev instead of bh->b_dev throughout aops.c.
1001
10022.0.3 - Small bug fixes, cleanups, and performance improvements.
1003
1004 - Remove some dead code from mft.c.
1005 - Optimize readpage and read_block functions throughout aops.c so that
1006 only initialized blocks are read. Non-initialized ones have their
1007 buffer head mapped, zeroed, and set up to date, without scheduling
1008 any i/o. Thanks to Al Viro for advice on how to avoid the device i/o.
1009 Thanks go to Andrew Morton for spotting the below:
1010 - Fix buglet in allocate_compression_buffers() error code path.
1011 - Call flush_dcache_page() after modifying page cache page contents in
1012 ntfs_file_readpage().
1013 - Check for existence of page buffers throughout aops.c before calling
1014 create_empty_buffers(). This happens when an I/O error occurs and the
1015 read is retried. (It also happens once writing is implemented so that
1016 needed doing anyway but I had left it for later...)
1017 - Don't BUG_ON() uptodate and/or mapped buffers throughout aops.c in
1018 readpage and read_block functions. Reasoning same as above (i.e. I/O
1019 error retries and future write code paths.)
1020
10212.0.2 - Minor updates and cleanups.
1022
1023 - Cleanup: rename mst.c::__post_read_mst_fixup to post_write_mst_fixup
1024 and cleanup the code a bit, removing the unused size parameter.
1025 - Change default fmask to 0177 and update documentation.
1026 - Change attrib.c::get_attr_search_ctx() to return the search context
1027 directly instead of taking the address of a pointer. A return value
1028 of NULL means the allocation failed. Updated all callers
1029 appropriately.
1030 - Update to 2.5.9 kernel (preserving backwards compatibility) by
1031 replacing all occurences of page->buffers with page_buffers(page).
1032 - Fix minor bugs in runlist merging, also minor cleanup.
1033 - Updates to bootsector layout and mft mirror contents descriptions.
1034 - Small bug fix in error detection in unistr.c and some cleanups.
1035 - Grow name buffer allocations in unistr.c in aligned mutlipled of 64
1036 bytes.
1037
10382.0.1 - Minor updates.
1039
1040 - Make default umask correspond to documentation.
1041 - Improve documentation.
1042 - Set default mode to include execute bit. The {u,f,d}mask can be used
1043 to take it away if desired. This allows binaries to be executed from
1044 a mounted ntfs partition.
1045
10462.0.0 - New version number. Remove TNG from the name. Now in the kernel.
1047
1048 - Add kill_super, just keeping up with the vfs changes in the kernel.
1049 - Repeat some changes from tng-0.0.8 that somehow got lost on the way
1050 from the CVS import into BitKeeper.
1051 - Begin to implement proper handling of allocated_size vs
1052 initialized_size vs data_size (i.e. i_size). Done are
1053 mft.c::ntfs_mft_readpage(), aops.c::end_buffer_read_index_async(),
1054 and attrib.c::load_attribute_list().
1055 - Lock the runlist in attrib.c::load_attribute_list() while using it.
1056 - Fix memory leak in ntfs_file_read_compressed_block() and generally
1057 clean up compress.c a little, removing some uncommented/unused debug
1058 code.
1059 - Tidy up dir.c a little bit.
1060 - Don't bother getting the runlist in inode.c::ntfs_read_inode().
1061 - Merge mft.c::ntfs_mft_readpage() and aops.c::ntfs_index_readpage()
1062 creating aops.c::ntfs_mst_readpage(), improving the handling of
1063 holes and overflow in the process and implementing the correct
1064 equivalent of ntfs_file_get_block() in ntfs_mst_readpage() itself.
1065 I am aiming for correctness at the moment. Modularisation can come
1066 later.
1067 - Rename aops.c::end_buffer_read_index_async() to
1068 end_buffer_read_mst_async() and optimize the overflow checking and
1069 handling.
1070 - Use the host of the mftbmp address space mapping to hold the ntfs
1071 volume. This is needed so the async i/o completion handler can
1072 retrieve a pointer to the volume. Hopefully this will not cause
1073 problems elsewhere in the kernel... Otherwise will need to use a
1074 fake inode.
1075 - Complete implementation of proper handling of allocated_size vs
1076 initialized_size vs data_size (i.e. i_size) in whole driver.
1077 Basically aops.c is now completely rewritten.
1078 - Change NTFS driver name to just NTFS and set version number to 2.0.0
1079 to make a clear distinction from the old driver which is still on
1080 version 1.1.22.
1081
1082tng-0.0.8 - 08/03/2002 - Now using BitKeeper, http://linux-ntfs.bkbits.net/
1083
1084 - Replace bdevname(sb->s_dev) with sb->s_id.
1085 - Remove now superfluous new-line characters in all callers of
1086 ntfs_debug().
1087 - Apply kludge in ntfs_read_inode(), setting i_nlink to 1 for
1088 directories. Without this the "find" utility gets very upset which is
1089 fair enough as Linux/Unix do not support directory hard links.
1090 - Further runlist merging work. (Richard Russon)
1091 - Backwards compatibility for gcc-2.95. (Richard Russon)
1092 - Update to kernel 2.5.5-pre1 and rediff the now tiny patch.
Anton Altaparmakovc002f422005-02-03 12:02:56 +00001093 - Convert to new filesystem declaration using ->ntfs_get_sb() and
Linus Torvalds1da177e2005-04-16 15:20:36 -07001094 replacing ntfs_read_super() with ntfs_fill_super().
1095 - Set s_maxbytes to MAX_LFS_FILESIZE to avoid page cache page index
1096 overflow on 32-bit architectures.
1097 - Cleanup upcase loading code to use ntfs_(un)map_page().
1098 - Disable/reenable preemtion in critical sections of compession engine.
1099 - Replace device size determination in ntfs_fill_super() with
1100 sb->s_bdev->bd_inode->i_size (in bytes) and remove now superfluous
1101 function super.c::get_nr_blocks().
1102 - Implement a mount time option (show_inodes) allowing choice of which
1103 types of inode names readdir() returns and modify ntfs_filldir()
1104 accordingly. There are several parameters to show_inodes:
1105 system: system files
1106 win32: long file names (including POSIX file names) [DEFAULT]
1107 long: same as win32
1108 dos: short file names only (excluding POSIX file names)
1109 short: same as dos
1110 posix: same as both win32 and dos
1111 all: all file names
1112 Note that the options are additive, i.e. specifying:
1113 -o show_inodes=system,show_inodes=win32,show_inodes=dos
1114 is the same as specifying:
1115 -o show_inodes=all
1116 Note that the "posix" and "all" options will show all directory
1117 names, BUT the link count on each directory inode entry is set to 1,
1118 due to Linux not supporting directory hard links. This may well
1119 confuse some userspace applications, since the directory names will
1120 have the same inode numbers. Thus it is NOT advisable to use the
1121 "posix" or "all" options. We provide them only for completeness sake.
1122 - Add copies of allocated_size, initialized_size, and compressed_size to
1123 the ntfs inode structure and set them up in
1124 inode.c::ntfs_read_inode(). These reflect the unnamed data attribute
1125 for files and the index allocation attribute for directories.
1126 - Add copies of allocated_size and initialized_size to ntfs inode for
1127 $BITMAP attribute of large directories and set them up in
1128 inode.c::ntfs_read_inode().
1129 - Add copies of allocated_size and initialized_size to ntfs volume for
1130 $BITMAP attribute of $MFT and set them up in
1131 super.c::load_system_files().
1132 - Parse deprecated ntfs driver options (iocharset, show_sys_files,
1133 posix, and utf8) and tell user what the new options to use are. Note
1134 we still do support them but they will be removed with kernel 2.7.x.
1135 - Change all occurences of integer long long printf formatting to hex
1136 as printk() will not support long long integer format if/when the
1137 div64 patch goes into the kernel.
1138 - Make slab caches have stable names and change the names to what they
1139 were intended to be. These changes are required/made possible by the
1140 new slab cache name handling which removes the length limitation by
1141 requiring the caller of kmem_cache_create() to supply a stable name
1142 which is then referenced but not copied.
1143 - Rename run_list structure to run_list_element and create a new
1144 run_list structure containing a pointer to a run_list_element
1145 structure and a read/write semaphore. Adapt all users of runlists
1146 to new scheme and take and release the lock as needed. This fixes a
1147 nasty race as the run_list changes even when inodes are locked for
1148 reading and even when the inode isn't locked at all, so we really
1149 needed the serialization. We use a semaphore rather than a spinlock
1150 as memory allocations can sleep and doing everything GFP_ATOMIC
1151 would be silly.
1152 - Cleanup read_inode() removing all code checking for lowest_vcn != 0.
1153 This can never happen due to the nature of lookup_attr() and how we
1154 support attribute lists. If it did happen it would imply the inode
1155 being corrupt.
1156 - Check for lowest_vcn != 0 in ntfs_read_inode() and mark the inode as
1157 bad if found.
1158 - Update to 2.5.6-pre2 changes in struct address_space.
1159 - Use parent_ino() when accessing d_parent inode number in dir.c.
1160 - Import Sourceforge CVS repository into BitKeeper repository:
1161 http://linux-ntfs.bkbits.net/ntfs-tng-2.5
1162 - Update fs/Makefile, fs/Config.help, fs/Config.in, and
1163 Documentation/filesystems/ntfs.txt for NTFS TNG.
1164 - Create kernel configuration option controlling whether debugging
1165 is enabled or not.
1166 - Add the required export of end_buffer_io_sync() from the patches
1167 directory to the kernel code.
1168 - Update inode.c::ntfs_show_options() with show_inodes mount option.
1169 - Update errors mount option.
1170
1171tng-0.0.7 - 13/02/2002 - The driver is now feature complete for read-only!
1172
1173 - Cleanup mft.c and it's debug/error output in particular. Fix a minor
1174 bug in mapping of extent inodes. Update all the comments to fit all
1175 the recent code changes.
1176 - Modify vcn_to_lcn() to cope with entirely unmapped runlists.
1177 - Cleanups in compress.c, mostly comments and folding help.
1178 - Implement attrib.c::map_run_list() as a generic helper.
1179 - Make compress.c::ntfs_file_read_compressed_block() use map_run_list()
1180 thus making code shorter and enabling attribute list support.
1181 - Cleanup incorrect use of [su]64 with %L printf format specifier in
1182 all source files. Type casts to [unsigned] long long added to correct
1183 the mismatches (important for architectures which have long long not
1184 being 64 bits).
1185 - Merge async io completion handlers for directory indexes and $MFT
1186 data into one by setting the index_block_size{_bits} of the ntfs
1187 inode for $MFT to the mft_record_size{_bits} of the ntfs_volume.
1188 - Cleanup aops.c, update comments.
1189 - Make ntfs_file_get_block() use map_run_list() so all files now
1190 support attribute lists.
1191 - Make ntfs_dir_readpage() almost verbatim copy of
1192 block_read_full_page() by using ntfs_file_get_block() with only real
1193 difference being the use of our own async io completion handler
1194 rather than the default one, thus reducing the amount of code and
1195 automatically enabling attribute list support for directory indices.
1196 - Fix bug in load_attribute_list() - forgot to call brelse in error
1197 code path.
1198 - Change parameters to find_attr() and lookup_attr(). We no longer
1199 pass in the upcase table and its length. These can be gotten from
1200 ctx->ntfs_ino->vol->upcase{_len}. Update all callers.
1201 - Cleanups in attrib.c.
1202 - Implement merging of runlists, attrib.c::merge_run_lists() and its
1203 helpers. (Richard Russon)
1204 - Attribute lists part 2, attribute extents and multi part runlists:
1205 enable proper support for LCN_RL_NOT_MAPPED and automatic mapping of
1206 further runlist parts via attrib.c::map_run_list().
1207 - Tiny endianness bug fix in decompress_mapping_pairs().
1208
1209tng-0.0.6 - Encrypted directories, bug fixes, cleanups, debugging enhancements.
1210
1211 - Enable encrypted directories. (Their index root is marked encrypted
1212 to indicate that new files in that directory should be created
1213 encrypted.)
1214 - Fix bug in NInoBmpNonResident() macro. (Cut and paste error.)
1215 - Enable $Extend system directory. Most (if not all) extended system
1216 files do not have unnamed data attributes so ntfs_read_inode() had to
1217 special case them but that is ok, as the special casing recovery
1218 happens inside an error code path so there is zero slow down in the
1219 normal fast path. The special casing is done by introducing a new
1220 function inode.c::ntfs_is_extended_system_file() which checks if any
1221 of the hard links in the inode point to $Extend as being their parent
1222 directory and if they do we assume this is an extended system file.
1223 - Create a sysctl/proc interface to allow {dis,en}abling of debug output
1224 when compiled with -DDEBUG. Default is debug messages to be disabled.
1225 To enable them, one writes a non-zero value to /proc/sys/fs/ntfs-debug
1226 (if /proc is enabled) or uses sysctl(2) to effect the same (if sysctl
1227 interface is enabled). Inspired by old ntfs driver.
1228 - Add debug_msgs insmod/kernel boot parameter to set whether debug
1229 messages are {dis,en}abled. This is useful to enable debug messages
1230 during ntfs initialization and is the only way to activate debugging
1231 when the sysctl interface is not enabled.
1232 - Cleanup debug output in various places.
1233 - Remove all dollar signs ($) from the source (except comments) to
1234 enable compilation on architectures whose gcc compiler does not
1235 support dollar signs in the names of variables/constants. Attribute
1236 types now start with AT_ instead of $ and $I30 is now just I30.
1237 - Cleanup ntfs_lookup() and add consistency check of sequence numbers.
1238 - Load complete runlist for $MFT/$BITMAP during mount and cleanup
1239 access functions. This means we now cope with $MFT/$BITMAP being
1240 spread accross several mft records.
1241 - Disable modification of mft_zone_multiplier on remount. We can always
1242 reenable this later on if we really want to, but we will need to make
1243 sure we readjust the mft_zone size / layout accordingly.
1244
1245tng-0.0.5 - Modernize for 2.5.x and further in line-ing with Al Viro's comments.
1246
1247 - Use sb_set_blocksize() instead of set_blocksize() and verify the
1248 return value.
1249 - Use sb_bread() instead of bread() throughout.
1250 - Add index_vcn_size{_bits} to ntfs_inode structure to store the size
1251 of a directory index block vcn. Apply resulting simplifications in
1252 dir.c everywhere.
1253 - Fix a small bug somewhere (but forgot what it was).
1254 - Change ntfs_{debug,error,warning} to enable gcc to do type checking
1255 on the printf-format parameter list and fix bugs reported by gcc
1256 as a result. (Richard Russon)
1257 - Move inode allocation strategy to Al's new stuff but maintain the
1258 divorce of ntfs_inode from struct inode. To achieve this we have two
1259 separate slab caches, one for big ntfs inodes containing a struct
1260 inode and pure ntfs inodes and at the same time fix some faulty
1261 error code paths in ntfs_read_inode().
1262 - Show mount options in proc (inode.c::ntfs_show_options()).
1263
1264tng-0.0.4 - Big changes, getting in line with Al Viro's comments.
1265
1266 - Modified (un)map_mft_record functions to be common for read and write
1267 case. To specify which is which, added extra parameter at front of
1268 parameter list. Pass either READ or WRITE to this, each has the
1269 obvious meaning.
1270 - General cleanups to allow for easier folding in vi.
1271 - attrib.c::decompress_mapping_pairs() now accepts the old runlist
1272 argument, and invokes attrib.c::merge_run_lists() to merge the old
1273 and the new runlists.
1274 - Removed attrib.c::find_first_attr().
1275 - Implemented loading of attribute list and complete runlist for $MFT.
1276 This means we now cope with $MFT being spread across several mft
1277 records.
1278 - Adapt to 2.5.2-pre9 and the changed create_empty_buffers() syntax.
1279 - Adapt major/minor/kdev_t/[bk]devname stuff to new 2.5.x kernels.
1280 - Make ntfs_volume be allocated via kmalloc() instead of using a slab
1281 cache. There are too little ntfs_volume structures at any one time
1282 to justify a private slab cache.
1283 - Fix bogus kmap() use in async io completion. Now use kmap_atomic().
1284 Use KM_BIO_IRQ on advice from IRC/kernel...
1285 - Use ntfs_map_page() in map_mft_record() and create ->readpage method
1286 for reading $MFT (ntfs_mft_readpage). In the process create dedicated
1287 address space operations (ntfs_mft_aops) for $MFT inode mapping. Also
1288 removed the now superfluous exports from the kernel core patch.
1289 - Fix a bug where kfree() was used insted of ntfs_free().
1290 - Change map_mft_record() to take ntfs_inode as argument instead of
1291 vfs inode. Dito for unmap_mft_record(). Adapt all callers.
1292 - Add pointer to ntfs_volume to ntfs_inode.
1293 - Add mft record number and sequence number to ntfs_inode. Stop using
1294 i_ino and i_generation for in-driver purposes.
1295 - Implement attrib.c::merge_run_lists(). (Richard Russon)
1296 - Remove use of proper inodes by extent inodes. Move i_ino and
1297 i_generation to ntfs_inode to do this. Apply simplifications that
1298 result and remove iget_no_wait(), etc.
1299 - Pass ntfs_inode everywhere in the driver (used to be struct inode).
1300 - Add reference counting in ntfs_inode for the ntfs inode itself and
1301 for the mapped mft record.
1302 - Extend mft record mapping so we can (un)map extent mft records (new
1303 functions (un)map_extent_mft_record), and so mappings are reference
1304 counted and don't have to happen twice if already mapped - just ref
1305 count increases.
1306 - Add -o iocharset as alias to -o nls for backwards compatibility.
1307 - The latest core patch is now tiny. In fact just a single additional
1308 export is necessary over the base kernel.
1309
1310tng-0.0.3 - Cleanups, enhancements, bug fixes.
1311
1312 - Work on attrib.c::decompress_mapping_pairs() to detect base extents
1313 and setup the runlist appropriately using knowledge provided by the
1314 sizes in the base attribute record.
1315 - Balance the get_/put_attr_search_ctx() calls so we don't leak memory
1316 any more.
1317 - Introduce ntfs_malloc_nofs() and ntfs_free() to allocate/free a single
1318 page or use vmalloc depending on the amount of memory requested.
1319 - Cleanup error output. The __FUNCTION__ "(): " is now added
1320 automatically. Introduced a new header file debug.h to support this
1321 and also moved ntfs_debug() function into it.
1322 - Make reading of compressed files more intelligent and especially get
1323 rid of the vmalloc_nofs() from readpage(). This now uses per CPU
1324 buffers (allocated at first mount with cluster size <= 4kiB and
1325 deallocated on last umount with cluster size <= 4kiB), and
1326 asynchronous io for the compressed data using a list of buffer heads.
1327 Er, we use synchronous io as async io only works on whole pages
1328 covered by buffers and not on individual buffer heads...
1329 - Bug fix for reading compressed files with sparse compression blocks.
1330
1331tng-0.0.2 - Now handles larger/fragmented/compressed volumes/files/dirs.
1332
1333 - Fixed handling of directories when cluster size exceeds index block
1334 size.
1335 - Hide DOS only name space directory entries from readdir() but allow
1336 them in lookup(). This should fix the problem that Linux doesn't
1337 support directory hard links, while still allowing access to entries
1338 via their short file name. This also has the benefit of mimicking
1339 what Windows users are used to, so it is the ideal solution.
1340 - Implemented sync_page everywhere so no more hangs in D state when
1341 waiting for a page.
1342 - Stop using bforget() in favour of brelse().
1343 - Stop locking buffers unnecessarily.
1344 - Implemented compressed files (inode->mapping contains uncompressed
1345 data, raw compressed data is currently bread() into a vmalloc()ed
1346 memory buffer).
1347 - Enable compressed directories. (Their index root is marked compressed
1348 to indicate that new files in that directory should be created
1349 compressed.)
1350 - Use vsnprintf rather than vsprintf in the ntfs_error and ntfs_warning
1351 functions. (Thanks to Will Dyson for pointing this out.)
1352 - Moved the ntfs_inode and ntfs_volume (the former ntfs_inode_info and
1353 ntfs_sb_info) out of the common inode and super_block structures and
1354 started using the generic_ip and generic_sbp pointers instead. This
1355 makes ntfs entirely private with respect to the kernel tree.
1356 - Detect compiler version and abort with error message if gcc less than
1357 2.96 is used.
1358 - Fix bug in name comparison function in unistr.c.
1359 - Implement attribute lists part 1, the infrastructure: search contexts
1360 and operations, find_external_attr(), lookup_attr()) and make the
1361 code use the infrastructure.
1362 - Fix stupid buffer overflow bug that became apparent on larger run
1363 list containing attributes.
1364 - Fix bugs in readdir() that became apparent on larger directories.
1365
1366 The driver is now really useful and survives the test
1367 find . -type f -exec md5sum "{}" \;
1368 without any error messages on a over 1GiB sized partition with >16k
1369 files on it, including compressed files and directories and many files
1370 and directories with attribute lists.
1371
1372tng-0.0.1 - The first useful version.
1373
1374 - Added ntfs_lookup().
1375 - Added default upcase generation and handling.
1376 - Added compile options to be shown on module init.
1377 - Many bug fixes that were "hidden" before.
1378 - Update to latest kernel.
1379 - Added ntfs_readdir().
1380 - Added file operations for mmap(), read(), open() and llseek(). We just
1381 use the generic ones. The whole point of going through implementing
1382 readpage() methods and where possible get_block() call backs is that
1383 this allows us to make use of the generic high level methods provided
1384 by the kernel.
1385
1386 The driver is now actually useful! Yey. (-: It undoubtedly has got bugs
1387 though and it doesn't implement accesssing compressed files yet. Also,
1388 accessing files with attribute list attributes is not implemented yet
Anton Altaparmakovc002f422005-02-03 12:02:56 +00001389 either. But for small or simple filesystems it should work and allow
Linus Torvalds1da177e2005-04-16 15:20:36 -07001390 you to list directories, use stat on directory entries and the file
1391 system, open, read, mmap and llseek around in files. A big mile stone
1392 has been reached!
1393
1394tng-0.0.0 - Initial version tag.
1395
1396 Initial driver implementation. The driver can mount and umount simple
Anton Altaparmakovc002f422005-02-03 12:02:56 +00001397 NTFS filesystems (i.e. ones without attribute lists in the system
Linus Torvalds1da177e2005-04-16 15:20:36 -07001398 files). If the mount fails there might be problems in the error handling
1399 code paths, so be warned. Otherwise it seems to be loading the system
1400 files nicely and the mft record read mapping/unmapping seems to be
1401 working nicely, too. Proof of inode metadata in the page cache and non-
1402 resident file unnamed stream data in the page cache concepts is thus
1403 complete.