1 Version 0.92 (2002-03-29)
2 ==========
3 * Minor cleanup. Ran Lindent on the sources.
4
5 Version 0.92 (2002-03-27)
6 ==========
7 * Fixed module makefile problem. It was not co
8 source files!
9 * Removed duplicated function definition
10 * Fixed potential null pointer dereference whe
11
12 Version 0.91 (2002-03-26)
13 ==========
14 * Oy! Fixed stupid bug that would cause an unr
15 Thanks to Laszlo Boszormenyi for point
16
17 Version 0.9 (2002-03-14)
18 ==========
19 * Added Sergey S. Kostyliov's patch to elimina
20 from b+tree operations. Changes the be
21
22 * Segregated the functions that interface dire
23 interface into their own file called l
24
25 Version 0.64 (2002-02-07)
26 ==========
27 * Did the string comparison really right this
28
29 * Fixed up some places where I assumed that a
30 a pointer value. (btree.c) [WD]
31
32 * Andrew Farnham <andrewfarnham@uq.net.au> poin
33 wouldn't work on older (<2.4.10) kerne
34 This is bad, since 2.4.9 is still the
35 a workaround for this problem (compati
36
37 * Sergey S. Kostyliov made befs_find_key() use
38 keys within btree nodes, rather than t
39 before. (btree.c) [Sergey S. Kostyliov<
40
41 * Made a debian package of the source for use
42
43
44 Version 0.63 (2002-01-31)
45 ==========
46 * Fixed bug in befs_find_brun_indirect() that
47 block being read. It was introduced wh
48 0.61. (datastream.c) [WD]
49
50 * Fixed a longstanding bug in befs_find_key()
51 finding the first key that is a substr
52 for. For example, this would cause fil
53 names like file1 and file2 to mysterio
54 (because they have the same inode numb
55 for reporting this serious bug!!!
56 (btree.c) [WD]
57
58 * Added support for long symlinks, after Axel
59 they work. I had forgotten all about t
60
61 * Documentation improvements in source. [WD]
62
63 * Makefile fix for independent module when CON
64 kernel config [Pavel Roskin <proski@gnu
65
66 * Compile warning fix for namei.c. [Sergey S. <
67
68
69 Version 0.62
70 ==========
71 * Fixed makefile for module install [WD]
72
73
74 Version 0.61 (2002-01-20)
75 ==========
76 * Made functions in endian.h to do the correct
77 the arch. [WD]
78
79 * Abbandoned silly checks for a NULL superbloc
80
81 * Misc code cleanups. Also cleanup of this cha
82
83 * Added byteswapping to all metadata reads fro
84 Uses the functions from endian.h [WD]
85
86 * Remove the typedef of struct super_block to
87 certain peoples' aesthetic sense. [WD]
88
89 * Ditto with the befs_read_block() interface.
90
91
92 Version 0.6 (2001-12-15)
93 ==========
94 * Cleanup of NLS functions (util.c) [WD]
95
96 * Make directory lookup/read use the NLS if an
97
98 * Fixed stupid bug where specifying the uid or
99 would result in the filesystem using t
100
101 * Added mount option to control debug printing
102 The option is, simply enough, 'debug'.
103 (super.c, debug.c) [WD]
104
105 * Removed notion of btree handle from btree.c.
106 linux VFS doesn't allow us to keep any
107 dir.c, namei.c befs_fs.h to account fo
108
109 * Improved handleing of overflow nodes when li
110 Now works for overflow nodes hanging o
111 node. This is the cleaner solution to
112
113 * Added new debug/warning/error print function
114 More flexible. Will soon be controllab
115 (see TODO). [WD]
116
117 * Rewrote datastream position lookups.
118 (datastream.c) [WD]
119
120 * Moved the TODO list to its own file.
121
122
123 Version 0.50 (2001-11-13)
124 ==========
125 * Added workaround for mis-understanding of th
126 in directories. A cleaner solution wil
127 for a while. Thanks to Brent Miszalask
128 (btree.c) [WD]
129
130 * Minor cleanups
131
132 * Added test for "impossible" condition of emp
133 seekleaf() in btree.c [WD]
134
135 * Implemented the abstracted read_block() in i
136
137 * Cleaned up the inode validation in inode.c [
138
139 * Anton Altaparmakov figured out (by asking Li
140 hanging disk io problem. It turns out
141 callback defined in your address_space
142 default linux-supplied implementation.
143 (file.c) [WD]
144
145 * Anton Altaparmakov and Christoph Hellwig ale
146 filesystem code should be using GFP_NO
147 priority parameter to kmalloc(). Fixed
148 (datastream.c, btree.c super.c inode.c
149
150 * Anton also told me that the blocksize is not
151 the page size in linux, which is 4k i3
152 (blocksize > PAGE_SIZE), and refuse to
153 practically means is that 8k blocksize
154 restructuring of the driver (or an alp
155
156 * Cleaned up the befs_count_blocks() function.
157 And somewhat smaller too. [WD]
158
159 * Made inode allocations use a slab cache
160 (super.c inode.c) [WD]
161
162 * Moved the freeing of the private inode secti
163 clear_inode(). This fixes a potential
164 can be called multiple times for each
165
166 * Converted all non vfs-callback functions to
167 superblock type, rather than struct su
168 portablity. [WD]
169
170 * Fixed a couple of compile warnings due to us
171 is the new way. (inode.c, super.c) [WD
172
173 * Fixed erronous includes of linux/befs_fs_i.h
174 in inode.c [WD]
175
176 Version 0.45 (2001-10-29)
177 ==========
178 * Added functions to get the private superbloc
179 their enclosing public structures. Swi
180 private portions to use them. (many fi
181
182 * Made read_super and read_inode allocate the
183 structures into the generic pointer fi
184 with kmalloc(). put_super and put_inod
185 to have to touch the definitions of th
186 include/linux/fs.h. Also, befs_inode_i
187 symlink string). (super.c, inode.c, be
188
189 * Fixed a thinko that was corrupting file read
190 is done being read. (datastream.c) [WD
191
192 * Removed fsync() hooks, since a read-only fil
193 [Christoph Hellwig].
194
195 * Fixed befs_readlink() (symlink.c) [Christoph
196
197 * Removed all the Read-Write stuff. I'll redo
198 write support (various files) [WD].
199
200 * Removed prototypes for functions who's defin
201 (befs_fs.h) [WD].
202
203
204 Version 0.4 (2001-10-28)
205 ==========
206 * Made it an option to use the old non-pagecac
207 testing purposes. (fs/Config.in)
208
209 * Fixed unused variable warnings when compilin
210
211 * Fixed a bug where the inode and super_block
212 fields set (inode.c and super.c).
213
214 * Release patch version 11. AKA befs-driver ve
215
216 * Thats right. New versioning scheme.
217 I've done some serious testing on it n
218 seems stable and not outragously slow.
219 correct (see TODO list). But it isn't
220 headroom before the big 1.0.
221
222
223 2001-10-26
224 ==========
225 * Fixed date format in this file. Was I smokin
226
227 * Removed old datastream code from file.c, sin
228
229 * Generic_read_file() is now used to read regu
230 It doesn't chew up the buffer cache (i
231 to be about as fast (even though it ha
232 indivdualy). And it knows about doing
233 So it does i/o in much larger chunks.
234 uses befs_get_block() by way of befs_r
235 of blocks, which in turn calls befs_fp
236 the hard work of finding the disk bloc
237
238 * Changed method of checking for a dirty files
239 (super.c). Now we check to see if log_
240 the journal needs to be replayed, and
241
242 * Fixed an extra instance of MOD_DEC_USE_COUNT
243
244 * Fixed a problem with reading the superblock
245 sizes (such as cdroms) on linux 2.4.10
246
247 2001-10-24
248 ==========
249 * Fix nasty bug in converting block numbers to
250 Subtle, because the old version was on
251 Probably responsible for lots of probl
252
253 * Fix bug with reading an empty directory. (bt
254
255 * This one looks good. Release patch version 1
256
257 2001-10-23
258 ==========
259 * Added btree searching function.
260
261 * Use befs_btree_find in befs_lookup (namei.c)
262
263 * Additional comments in btree.c
264
265 2001-10-22
266 ==========
267 * Added B+tree reading functions (in btree.c).
268 Made befs_readdir() use them them inst
269
270 2001-09-11
271 ==========
272 * Converted befs_read_file() to use the new da
273
274 * Finally updated the README file.
275
276 * Added many comments.
277
278 * Posted version 6
279
280 * Removed byte-order conversion code.
281 I have no intention of supporting it,
282 Flow control with #ifdef (ugh). Maybe
283 native byteorder works 100%.
284
285 2001-09-10
286 ==========
287 * Finished implementing read_datastream()
288
289 * made befs_read_brun() more general
290 Supports an offset to start at and a m
291 Added a wrapper function to give the o
292
293 2001-09-30
294 ==========
295 * Discovered that the datastream handleing cod
296 in several respects. For one thing, it
297
298 * Rewrote datastream handleing.
299
300 * Created io.c, for io related functions.
301 Previously, the befs_bread() funtions
302 Created the befs_read_brun() function.
303
304
305 2001-09-07
306 ==========
307 * Made a function to actually count the number
308 And helper functions.
309 (fs/befs/inode.c)
310
311 2001-09-05
312 ==========
313 * Fixed a misunderstanding of the inode fields
314 This fixed the problmem with wrong fil
315 The i_blocks field of the inode struct
316 inode, it is the number of blocks for
317 necessarily the size of the inode, alt
318 Changed to blocksize of filesystem.
319 (fs/befs/inode.c)
320
321 * Permanently removed code that had been provi
322
323 * Since we don't support access time, make tha
324 copying m_time.
325 (fs/befs/inode.c)
326
327 * Added sanity check for inode reading
328 Make sure inode we got was the one we
329 (fs/befs/inode.c)
330
331 * Code cleanup
332 Local pointers to commonly used struct
333 Got rid of abominations befs_iaddr2ino
334 Replaced with single function iaddr2bl
335 (fs/befs/super.c) (fs/befs/inode.c)
336
337 2001-09-01
338 ==========
339 * Fixed the problem with statfs where it would
340 half full, due to improper understandi
341 (fs/befs/super.c)
342
343 * Posted verion 4 of the patch
344
345 2001-09-01
346 ==========
347 * Changed the macros in befs_fs.h to inline fu
348 More readable. Typesafe. Better
349 (include/linux/befs_fs.h)
350
351 * Moved type definitions from befs_fs.h to a n
352 Because befs_fs_i.h and befs_fs_sb.h w
353 typedefs, and they are inlcuded in <li
354 that I want the inline functions in be
355 circularity.
356 (include/linux/befs_fs.h)
357
358 2001-08-30
359 ==========
360 * Cleaned up some wording.
361
362 * Added additional consitency checks on mount
363 Check block_size agrees with block_shi
364 Check flags == BEFS_CLEAN
365 (fs/befs/super.c)
366
367 * Tell the kernel to only mount befs read-only
368 By setting the SB_RDONLY flag in befs_
369 Not that it was possible to write befo
370 (fs/befs/super.c)
371
372 * Got rid of kernel warning on mount.
373 The kernel doesn't like it if you call
374 you have some of its blocks open. Move
375 very end of befs_read_super(), after w
376 (fs/befs/super.c)
377
378 * Fixed wrong number of args bug in befs_dump_
379 (fs/befs/debug.c)
380
381 * Solved lots of type mismatches in kprint()s
382 (everwhere)
383
384 2001-08-27
385 ==========
386 * Cleaned up the fs/Config.in entries a bit, n
387
388 * BeFS depends on NLS, so I made activating Be
389 (fs/nls/Config.in)
390
391 * Added Configure.help entries for CONFIG_BEFS
392 (currently at fs/befs/Kconfig)
393
394 2001-08-??
395 ==========
396 * Removed superblock locking calls in befs_rea
397 hands us a super_block struct that is
398
399 2001-08-13
400 ==========
401 * Will Dyson <will_dyson@pobox.com> is now atte
402 Makoto Kato <m_kato@ga2.so-net.ne.jp> i
403 also did some work on it (fixing it up
404
405 * Fixed compile errors on 2.4.1 kernel (WD)
406 Resolve rejected patches
407 Accommodate changed NLS interface (uti
408 Needed to include <linux/slab.h> in mo
409 Makefile changes
410 fs/Config.in changes
411
412 * Tried to niceify the code using the ext2 fs
413 Declare befs_fs_type using the DECLARE
414
415 * Made it a configure option to turn on debugg
416
417 * Compiles on 2.4.7
Linux® is a registered trademark of Linus Torvalds in the United States and other countries.
TOMOYO® is a registered trademark of NTT DATA CORPORATION.