Fix another hashtree bug in restore causing "removenode: non-empty directory"
[dump.git] / CHANGES
1 $Id: CHANGES,v 1.273 2005/03/18 22:12:54 stelian Exp $
2
3 Changes between versions 0.4b39 and 0.4b40 (released ????????????????)
4 ======================================================================
5
6 1. Changed restore to emit warnings (instead of emitting a fatal
7 error) if a file (or a directory) is unavailable for a
8 comparision (if the user doesn't have the necessary permissions
9 to access it for example). Thanks to Kenneth Porter
10 <shiva@sewingwitch.com> for the bug report.
11
12 2. Re-done the 'do not save directory entries to non-dumped inodes
13 (excluded from dump)' feature. The previous implementation
14 worked well for excluded directories but not for regular files.
15 Thanks to Kenneth Porter <shiva@sewingwitch.com> for the bug
16 report.
17
18 3. Fixed a bug in dump where the tape size was miscalculated when
19 the user used -d/-s to specify the tape characteristics. Thanks
20 to Philip Goisman <goisman@physics.arizona.edu> for reporting
21 the bug.
22
23 4. Fixed another bug introduced in restore with the hashtree
24 implementation. This one caused restore to stop saying
25 "removenode: non-empty directory" in some cases.
26
27 Changes between versions 0.4b38 and 0.4b39 (released January 21, 2005)
28 ======================================================================
29
30 1. The newly added dump_on_cd_3 example was buggy, replace it
31 with an updated version from Andrew Basterfield
32 <bob@cemetery.homeunix.org>.
33
34 2. Made restore to chdir() back into the initial directory when
35 dumping core while aborting a comparision operation. The
36 previous behaviour was to write the corefile at the root of
37 the directory being compared, which could very well be
38 read only and preventing the corefile generation. Thanks
39 to Kenneth Porter <shiva@sewingwitch.com> for the bug report.
40
41 3. Silenced the failure to call fgetflags() when comparing an
42 entry which has no ext2 attributes (as in lsattr()).
43
44 4. Fix a brown paper bug in restore -C which broke restore and
45 caused modifications on the filesystem being compared
46 (directories containing a file with the same name as the
47 directory get renamed to RSTTMP...). Thanks to Kenneth Porter
48 <shiva@sewingwitch.com> for finding the bug and helping me
49 reproduce it.
50
51 5. Made restore -C force the -N flag (no writing allowed on
52 the disk) in order to prevent more bugs like the above one.
53
54 Changes between versions 0.4b37 and 0.4b38 (released January 7, 2005)
55 =====================================================================
56
57 1. Fix a couple of troff syntax bugs in the man pages.
58 Thanks to Eric Raymond <esr@thyrsus.com> for the patch.
59
60 2. Made restore use either libncurses or libtermcap, depending
61 on which one is available at configure time.
62
63 3. Fixed restore negative size display bug when comparing a
64 dump containing files over 2GB. Thanks to Steve Bonds
65 <sbonds@users.sourceforge.net> for the bug report.
66
67 4. Do not save directory entries to non-dumped inodes
68 (excluded from dump). This will eliminate the 'missing
69 file' warnings when doing 'restore -C'.
70
71 5. Fix dump crash when backuping a huge (2TB) filesystem,
72 due to a bogus calculation on the inode map size.
73 Thanks to Kevin B. Haines <K.B.Haines@rl.ac.uk> for
74 submitting the bug and testing the fix.
75
76 6. Fix a problem in restore where the final \0 in the symbolic
77 link names could have been lost, generating corrupt filenames.
78 Thanks to Kyle Wilson <kyle.wilson@amd.com> for reporting the
79 bug.
80
81 7. Implemented a hash list for the directory names in restore.
82 The linear list used before caused problems in interactive
83 restores when dealing with directories having thousands of
84 entries. Thanks to Brian Ristuccia <bristuccia@starentnetworks.com>
85 for reporting the bug.
86
87 8. Improved restore -C, this time including the directory
88 attributes into the comparision.
89
90 9. Made restore understand tapes containing EA/ACLs (which will
91 be dumped by the next version of dump). In this version
92 extended attributes on the tape are ignored, for full EA/ACL
93 support wait for the next version or try the experimental EA
94 patch.
95
96 Changes between versions 0.4b36 and 0.4b37 (released July 7, 2004)
97 ==================================================================
98
99 1. Added the --enable-staticz configure option which enables
100 dump and restore to be linked with static versions of
101 libz and libbz2 (and dynamic versions of all other libraries).
102 This will make Debian users happy, because libz and libbz2
103 were the only needed libraries living in /usr, all the
104 others live in /lib. In case of system emergency, it is
105 better not to have to rely on an extra filesystem.
106 Thanks to Bdale Garbee <bdale@gag.com> for the suggestion.
107
108 2. Fix compilation on (at least the Linux Debian port to) AMD64.
109 (<ext2fs/ext2_types.h> defines some types (__s64 and __u64)
110 that are also defined by <linux/types.h> (<asm/types.h>) and
111 they conflict).
112
113 3. Make dump's reading of the dumpdates file a bit more robust,
114 preventing dump from crashing when the dumpdates file has
115 been modified by hand.
116
117 4. Fixed some offset calculations in dump code which could
118 lead to "bread lseek errors" on large filesystems. Thanks
119 to Bruce Lowekamp <lowekamp@users.sourceforge.net> for
120 reporting this bug and debugging the issue.
121
122 5. Made dump use the blkid library when searching for devices
123 by label or uuid instead of dump's own routines.
124
125 6. Corrected a bug in dump where a wrong LABEL=... line in
126 /etc/fstab could prevent dump from dumping unrelated
127 filesystems. Thanks to Bruce Lowekamp
128 <lowekamp@users.sourceforge.net> for reporting the bug.
129
130 Changes between versions 0.4b35 and 0.4b36 (released April 21, 2004)
131 ====================================================================
132
133 1. Fixed dump compilation with old gcc versions. Thanks to
134 Mike Castle <dalgoda@users.sourceforge.net> for the patch.
135
136 2. Fixed some warnings (howmany, roundup, powerof2 redefined)
137 when compiling against a recent glibc version.
138
139 3. Fixed a bug in restore preventing the read of a dump tape
140 written with Solaris 7 ufsdump. Thanks to Patrick Higgins
141 <phiggins@transzap.com> for reporting the bug and providing
142 the test case.
143
144 4. Changed dump to enable the creation of volumes of different
145 sizes in a single run (make -B accept a list of values).
146 Patch contributed by Florian Zumbiehl <florz@gmx.de>.
147
148 5. Use the glibc provided minor() and major() macros instead
149 of our own bitmask implementation. This should be safe for
150 when the major/minor namespace will migrate to 32 bits.
151 Thanks to Zhang Jun <zhangjun@nanjing-fnst.com> for reporting
152 the bug.
153
154 6. Made explicit in the dump man page that dump will not create
155 a remote file, it will only write to an already existing one.
156
157 7. Another try at making size estimates better again.
158
159 8. Put back the inconditional running of the end-of-tape
160 script which was changed in 0.434 to be run only when -M
161 or multiple -f were NOT specified. Some users rely on this
162 feature even when it is combined with -M/-f.
163
164 9. Fixed restore when restoring huge backups (where rstdir...
165 temporary files are over 2GB). Thanks to
166 Raphael Jaffey <rjaffey@artic.edu> for reporting this,
167 debugging the issue and making the patch.
168
169 10. Made restore understand FreeBSD UFS2 tapes. Thanks to
170 David <vrtin@users.sourceforge.net> for submitting the bug
171 and providing a test case.
172
173 11. Made dump work with an arbitrary integer level (no more 10
174 levels only limitation). Thanks to Matthew
175 <msvincen@midway.uchicago.edu> for the patch.
176
177 Changes between versions 0.4b34 and 0.4b35 (released December 21, 2003)
178 =======================================================================
179
180 1. Added a note in the dump man page saying that the default
181 blocksize can be 32 if -d is used with a high density argument.
182 Thanks to Antonios Christofides <A.Christofides@itia.ntua.gr>
183 for the patch.
184
185 2. Fixed configure to correctly understand CPPFLAGS, CFLAGS,
186 LDFLAGS environment variables. Thanks to Arcady Genkin
187 <antipode@users.sourceforge.net> for reporting the bug.
188
189 3. Made -e/-E options of dump accept an unlimited number of inodes
190 to be excluded, rather than a hardcoded maximum. Thanks to
191 Dietrich Rothe <d-rothe@users.sourceforge.net> for the patch.
192
193 4. Updated the autoconf system to 2.50. Forced the -D_BSD_SOURCE
194 and -D_USE_BSD_SIGNAL defines in configure in order to solve
195 64bit build problems because quad_t is redifined with a
196 different signature. Thanks to Mike Harris <mharris@redhat.com>
197 for reporting this bug.
198
199 5. Made restore build on Solaris, making possible to
200 restore Linux's "enhanced" tapes. Thanks to Uwe Gohlke
201 <uwe@ugsoft.de> for the patch.
202
203 6. Made an extension in the dump tape format capable of saving
204 MacOSX specific inode extensions. Uwe Gohlke <uwe@ugsoft.de>
205 wrote the extension and contributed the restore code back
206 into this codebase. The same extension mechanism will be
207 used in the future to save ACLs...
208
209 7. Made rmt work correctly with regard to QFA and local/remote
210 files and tapes. The remote access will however work only
211 when the dump provided rmt version is used. If you want to
212 use another rmt server, please do not use the QFA feature.
213 Thanks to Uwe Gohlke <uwe@ugsoft.de> for the patch.
214
215 Changes between versions 0.4b33 and 0.4b34 (released April 18, 2003)
216 ====================================================================
217
218 1. Fixed the INSTALL file to reflect the actual install paths.
219 Thanks to David Ronis <ronis@ronispc.chem.mcgill.ca> for
220 reporting the bug.
221
222 2. Fixed the configure script to only check for headers presence
223 instead of trying to compile them. This should fix issues
224 with old build environments. Thanks to Kari Lempiainen
225 <kari@funky.pp.fi> for reporting the bug.
226
227 3. Fixed restore to correctly ignore sockets when comparing
228 a dump (as socket cannot be properly restored anyway). Thanks
229 to Gunther Reiszig <gunther@mit.edu> for reporting the bug.
230
231 4. Fixed restore to correctly access the archive file (-A argument)
232 even when using a remote tape. Thanks to Fabrice Bellet
233 <fabrice@bellet.info> for reporting the bug.
234
235 5. Fixed (again) handling of long (largefile) seeks in rmt.
236 Thanks to Fabrice Bellet <fabrice@bellet.info> for reporting
237 the bug.
238
239 6. Fixed restore corner case when dealing with large block sizes
240 dump is able to write now (-b 1024). Thanks to Fabrice Bellet
241 <fabrice@bellet.info> for reporting the bug.
242
243 7. Fixed a bug preventing dump to access a filesystem having
244 a label exactly 16 bytes in length. Thanks to <minduni@ti-edu.ch>
245 for reporting the bug.
246
247 8. Made dump store dump dates using explicit timezones, fixing a
248 problem with incremental dumps if the timezone is modified
249 between the dumps. Thanks to Philippe Troin <phil@fifi.org> for
250 the bug report and the patch.
251
252 9. Fixed a bug encountered when dumping individual files (not full
253 filesystems or directories) and dangling symbolic links happen
254 to be in the list of files. For as far as dump is concerned,
255 dangling symbolic links are allowed, and are dumped as is.
256 Thanks to Jin-su Ahn <jsahn@ee.snu.ac.kr> for reporting the
257 bug and providing the fix.
258
259 10. Fixed open and creation modes and permissions for QFA and
260 table-of-contents files in dump and restore. Thanks to
261 Philippe Troin <phil@fifi.org> for the patch.
262
263 11. Fixed the archive file descriptor handling enabling it to be 0.
264 This can happen in some cases when shell redirections are used.
265 Thanks to Philippe Troin <phil@fifi.org> for the patch.
266
267 12. Delayed the opening of archive file until after suid had been
268 dropped (fixing a possible security issue if dump is suid).
269 Thanks to Philippe Troin <phil@fifi.org> for the patch.
270
271 13. Fixed the 'S' command handling in the rmt client part.
272 Thanks to Philippe Troin <phil@fifi.org> for the patch.
273
274 14. Modified the end-of-tape script handling to print out statistics
275 (and stop the timer) before launching the eot script. Also, the eot
276 script does not get run anymore when using -M (which makes sense) or
277 when multiple tapes are listed on the command line
278 (-f tape0,tape1,tapen) (which also makes sense).
279 Thanks to Philippe Troin <phil@fifi.org> for the patch.
280
281 15. Relicensed dump/restore under the 'revised' BSD license, as per
282 ftp://ftp.cs.berkeley.edu/ucb/4bsd/README.Impt.License.Change.
283
284 16. Added LZO compression to dump. This new compression method has
285 the advantage of being super fast, thus not killing tape streaming
286 on slow machines. Thanks to Helmut Jarausch
287 <jarausch@igpm.rwth-aachen.de> for the patch and to
288 Markus Oberhumer <markus@oberhumer.com> for giving special permission
289 to include his miniLZO project (GPL licensed) in dump/restore.
290
291 17. Some small buffer overruns fixes in rmt. Thanks to Antonomasia
292 <ant@notatla.demon.co.uk> for reporting the bugs.
293
294 18. Added a special rmt version which can do encryption when writing
295 to tape. Read examples/encrypted_rmt/README for details on
296 how to enable and configure it. Thanks to Ken Lalonde
297 <ken@globalremit.com> for the patch.
298
299 19. Made dump work with 2.5 kernel end of tape early warning semantics.
300 Thanks to Kai Makisara <Kai.Makisara@kolumbus.fi> for the patch.
301
302 20. Fixed a bug which caused dump -w|-W not to work anymore, because
303 the fs_freq and fs_passno fields in /etc/mtab are always set
304 to 0 0. Thanks to Trent Piepho <xyzzy@speakeasy.org> for
305 reporting the bug.
306
307 Changes between versions 0.4b32 and 0.4b33 (released February 10, 2003)
308 =======================================================================
309
310 1. Added a note in the restore man page clarifying the question
311 'set the permissions on the current directory ?' asked by
312 restore at the end of treatment in -i and -x modes.
313
314 2. Fixed the endianess issues when reading compressed tapes.
315 Thanks to Dark Force <daq4th@users.sourceforge.net> for
316 reporting this bug and providing test cases.
317
318 3. Fixed the "ACL won't be dumped" warning message (which showed
319 an extra, unrelated error message). Thanks to Dragan Krnic
320 <dkrnic@lycos.com> for reporting this bug.
321
322 4. Made dump look first into /etc/mtab, then into /etc/fstab
323 when searching for filesystem paths. Also fixed some problems
324 caused by binding mounts or multiple block device mounts.
325 Thanks to Matus Uhlar <uhlar@fantomas.sk>, Elliott Mitchell
326 <ehem@m5p.com>, Greg Edwards <gedwards@users.sourceforge.net>,
327 Brian Hoy <brian.hoy@opus.co.nz>. (fixes Debian bugs #147086
328 and #160305, Sourceforge bugs #618699 and #679832).
329
330 5. Made dump's -I option accept the value '0' meaning all the
331 read errors will be ignored. This can be useful when running
332 dump from unattended sessions (like cron jobs). Thanks to
333 John I Wang <jiwang@users.sourceforge.net> for the suggestion.
334
335 6. Fixed the output of dump to indicate 'blocks' instead of
336 'tape blocks' in the various messages (blocks are always
337 1 Kilobyte, tape blocks are 1 BK * '-b' argument), and
338 made it clearly print the current blocksize at the start of
339 a dump. Thanks to Michal Szymanski <msz@astrouw.edu.pl> for
340 the suggestions.
341
342 7. Made rmt understand long (largefiles) seeks.
343
344 8. Fixed build with very old versions of libext2fs, where
345 EXT2_FT_* constants were undefined.
346
347 9. Made dump accept the dumpdates path on the command line
348 (-D file option) instead of using only the hardcoded one.
349 Thanks to Piete Brooks <pb22@users.sourceforge.net> for the
350 suggestion.
351
352 10. Enabled rmt, LFS, readline, QFA options by default in
353 ./configure. Updated the configure process (new versions
354 of config.guess, config.sub etc).
355
356 Changes between versions 0.4b31 and 0.4b32 (released November 15, 2002)
357 =======================================================================
358
359 1. Changed dump to use fcntl(F_SETLK) style locking instead
360 of flock() when locking the dumpdates file. With the old
361 locking scheme, a local user having read rights on the
362 dumpdates file could be able to do a Denial of Service attack
363 on dump. In order to lock the dumpdates file with the new
364 scheme, the user would need to have write access on the file.
365 Thanks to Richard Johnson <Richard.Johnson3@ey.com> for
366 reporting the bug (originally a bugtraq post).
367
368 2. Fixed interactive 'ls' which caused spurious errors warnings
369 about 'undefined filetypes' detected. Thanks to Jorgen Ostling
370 <jorgen_ostling@users.sourceforge.net> for reporting this
371 bug.
372
373 3. Fixed dump's estimate when dealing with sparse inodes.
374
375 4. Modified dump to allow setting a blocksize bigger than 32kB
376 (raised the limit to 1024kB), because newer hardware needs
377 this for proper operation. Thanks to Dirk Traenapp
378 <dtraenapp@users.sourceforge.net> for reporting this.
379
380 5. Fixed a bug causing Dump to stop and report an error if an
381 inode number in the exclude file was followed by some amount
382 of whitespace. Thanks to Jeffrey Sofferin
383 <sofferin@users.sourceforge.net> for reporting this bug.
384
385 6. Fixed a bug which caused restore, in some particular cases,
386 to ask some 'scary' questions and leave a bunch of RSTTMP
387 directories behind when restoring incremental tapes. Thanks
388 to Philippe Troin <phil@fifi.org> for reporting this bug and
389 providing the test cases.
390
391 7. Changed the wording when inodes are excluded from dump:
392 replaced 'Added inode 7 to exclude list' with
393 'Excluding inode 7 (resize inode) from dump', as suggested
394 by Elliott Mitchell <ehem@m5p.com> in a Debian bug report.
395
396 Changes between versions 0.4b30 and 0.4b31 (released July 30, 2002)
397 ===================================================================
398
399 1. Fixed rmt open flags transmission (GNU's symbolic syntax over
400 rmt) which I broke in 0.4b29. Thanks to Eros Albertazzi
401 <eros@lamel.bo.cnr.it> for reporting the bug.
402
403 Changes between versions 0.4b29 and 0.4b30 (released July 25, 2002)
404 ===================================================================
405
406 1. Made dump print out the ext2 logical block number in case of a read
407 error, which can be given as an argument to debugfs' ncheck command,
408 in order to find out the affected inode. Added note about this
409 usage in dump's man page.
410
411 2. Fixed a problem in restore when reading tapes written on big
412 endian machines with very old versions of dump. The patch was
413 contributed by George Helffrich <george@geology.bristol.ac.uk>.
414
415 3. Fixed the tape length calculation when using large tapes
416 and compression. Thanks to Georg Lippold
417 <g_lippold@sourceforge.net> for reporting the bug.
418
419 4. Added a new set of examples in dump_on_cd_2 directory, based
420 on dump_on_cd examples but somewhat enhanced, supporting DVD
421 media, and localized in english and german. Thanks to
422 Georg Lippold <g_lippold@sourceforge.net> for the new scripts.
423
424 5. Made dump save 32 bit UID/GID. Internally, this was achieved
425 by switching from the old BSD inode format to the new BSD
426 inode format, which means that the tape format was changed.
427 However, since all restore versions out there should
428 transparently support both inode formats, the change should
429 have no side effects. Thanks to John Yu <jky@cs.bu.edu> for
430 reporting the bug.
431
432 6. Fixed a lot of warnings in the code shown when compiling
433 with 'gcc -W'. Thanks to Matthias Andree
434 <matthias.andree@stud.uni-dortmund.de> for reporting this.
435
436 7. Fixed a small markup bug in the dump man page. Thanks to
437 Eric S. Raymond <esr@minx.thyrsus.com> for submitting the
438 patch.
439
440 8. Rewrote entirely the man pages using the tmac.an macro
441 package (Linux man page format) instead of the original BSD
442 format. They should be now cleaner and easier to modify.
443
444 Changes between versions 0.4b28 and 0.4b29 (released June 8, 2002)
445 ==================================================================
446
447 1. Fixed a problem in the rmt ioctl command, where ioctl's issued from
448 non Linux clients were misinterpreted. The description of the problem
449 (incompatible numbering in Linux mtio opcodes) is documented at
450 ftp://ftp.fokus.gmd.de/pub/unix/star/README.mtio . Thanks to
451 Jörg Schilling <schilling@fokus.gmd.de> for reporting this bug and
452 providing an excellent, cross-platform replacement for rmt in his
453 star package.
454
455 2. Fixed a bug reported by Andy Mentges <amentges@jumpline.com>
456 which caused restore to fail when the inode maps were not entirely
457 contained on the first volume (when using really small volumes or
458 when dumping a huge number of inodes).
459
460 3. Fixed a problem in dump, where files in subdirectories of directories
461 that have the nodump flag set (or being excluded with -e/-E) were
462 sometimes incorrectly dumped (instead of being excluded). The original
463 fix is from David C Lawrence <tale@dd.org> in the FreeBSD version
464 of dump: http://www.FreeBSD.org/cgi/query-pr.cgi?pr=32414
465 Thanks to Ted Grzesik <tedgyz@roostme.com> for reporting the bug and
466 help testing the patch.
467
468 4. Added some example scripts from Gerd Bavendiek <bav@epost.de>
469 which makes one able to pipe the output of dump, by the net, to
470 a remote CD-burner server.
471
472 5. Made dump use O_CREAT|O_TRUNC both locally and remotely (over rmt),
473 and use GNU's symbolic syntax over rmt instead of numerical values
474 to assure multiple platform compatibility.
475
476 6. Documented the -d option in restore.
477
478 7. Added a -v (verbose) mode to dump. For now it just prints the number
479 of the inode being dumped, but this could evolve in future versions
480 to include interesting debugging output.
481
482 8. Added a -o flag to restore, which automatically restores the current
483 directory permissions in -i or -x mode, without asking the operator
484 whether to do so. Patch submitted by Tony Clayton <tonyc@e-smith.com>
485 and Peter Samuel <peters@e-smith.com>.
486
487 Changes between versions 0.4b27 and 0.4b28 (released April 12, 2002)
488 ====================================================================
489
490 1. Fixed a bug in the QFA feature which could have caused
491 a bad tape positionning on restore (causing extra delays in
492 restore). Credits go to Uwe Gohlke <uwe@ugsoft.de>.
493
494 2. Added a small note in the dump man page specifying that
495 there should be no white space between the option letter and
496 the -j or -z parameter, thanks to Kenneth Porter <shiva@well.com>
497
498 3. Made restore work with older versions of the readline library,
499 by checking in configure for several extended symbols. Restore
500 can now be compiled with a readline as old as the 2.0 release
501 (though it may be a good idea to upgrade it to a more recent
502 version...). Thanks to Andrew Donkin <ard@waikato.ac.nz> for
503 reporting the build failures.
504
505 4. Fixed a performance problem with the QFA file creation in
506 dump, which made unnecessary seeks on the tape slowing down
507 the dump. Thanks to Andrew Donkin <ard@waikato.ac.nz> for
508 reporting this issue.
509
510 5. Removed the inclusion of some kernel headers in the dump
511 source, which prevented the compile in some kernel/glibc
512 headers/architecture combination. Thanks to Bdale Garbee
513 <bdale@gag.com> for reporting the bug.
514
515 6. Added the appropriate error message when dump fails to
516 open the output file for writing. Thanks to Amith Varghese
517 <amithv@yahoo.com> for reporting this bug.
518
519 7. Made restore able to understand large Solaris ufsdump tapes
520 (containing inodes bigger than 4194304). Sun have introduced
521 an "extension" to the dump tape format when dealing with
522 those inodes, which was uncorrectly handled by Linux restore.
523 Thanks to Uwe Gohlke <uwe@ugsoft.de> for reporting the bug and
524 providing a test case.
525
526 8. Added the -m parameter to dump which optimises the output for
527 inodes having been changed but not modified since the last dump
528 ('changed' and 'modified' have the meaning defined in stat(2)).
529 For those inodes, dump will save only the metadata, instead of
530 saving the entire inode contents. Inodes which are either
531 directories or have been modified since the last dump are saved
532 in a regular way. Uses of this flag must be consistent, meaning
533 that either every dump in an incremental dump set have the flag,
534 or no one has it.
535
536 Changes between versions 0.4b26 and 0.4b27 (released February 15, 2002)
537 =======================================================================
538
539 1. Fixed behaviour of dump when exceeding resource limits
540 (SIGXFSZ treatment).
541
542 2. Added the -L flag to restore to allow the user to specify a
543 maximal allowed number of miscompares when using restore
544 with the -C option to check the backup.
545
546 3. Detailed the manual entry for the -N option of restore.
547
548 4. Added the -a flag to restore to make able doing unattended
549 restores in -i or -x mode (automatically walks through the
550 multiple dump volumes).
551
552 5. Extended the QFA mode to work with local files and/or
553 remote tapes and files. This way, restore can know in advance
554 the tape number and the offset for the inodes to extract and
555 can minimize the extraction time by seeking directly to the
556 good tape/offset.
557
558 6. Added the -A <archive> option to both dump and restore,
559 which makes dump to archive a dump table-of-contents in
560 the specified file to be used by restore to determine
561 whether a file is in the dump file that is being restored.
562 (the archive file syntax is also compatible with the
563 Solaris ufsdump generated one).
564
565 7. Small fix in restore making it able to read some (broken ?)
566 Solaris ufsdump tapes.
567
568 8. Fixed dump to correctly recognise the root filesystem when using
569 ext2 disk labels (LABEL=/). Thanks to John Yu <jky@it.bu.edu>
570 for reporting this bug.
571
572 9. Added the -P <file> option to restore to create a
573 Quick File Access file from an already made dump. Patch
574 contributed by Uwe Gohlke <uwe@ugsoft.de>.
575
576 10. Made restore compile and run on Solaris, making it a
577 possible replacement for the standard ufsrestore. Port was
578 contributed by Uwe Gohlke <uwe@ugsoft.de>.
579
580 Changes between versions 0.4b25 and 0.4b26 (released January 7, 2002)
581 =====================================================================
582
583 1. Added a set of backup scripts from Eugenio Diaz
584 <getnito@yahoo.com> in the examples section. It features
585 automatic (cron based) full and incremental dumping of
586 several filesystems on a separate filesystem.
587
588 2. Fixed a off-by-one miscalculation which disabled dumping
589 a one letter subdirectory. Thanks to Chris Siebenmann
590 <cks@utcc.utoronto.ca> for reporting the bug.
591
592 3. Fixed several restore bugs occuring when trying to
593 restore some missing files on the tape. Thanks to Chris
594 Siebenmann <cks@utcc.utoronto.ca> for reporting the bug.
595
596 4. Fixed --with-ldopts configure argument passing, installing from
597 a separate object directory, makefile cleanups contributed
598 by <splite@purdue.edu>.
599
600 5. Fix a bug which could caused, in some conditions, the highest
601 number inode of a filesystem, to not be dumped. Many thanks
602 to Chris Siebenmann <cks@utcc.utoronto.ca> for helping me
603 chase this bug.
604
605 Changes between versions 0.4b24 and 0.4b25 (released November 17, 2001)
606 =======================================================================
607
608 1. Added a mini howto from Patrick Walsh in the examples
609 directory.
610
611 2. Minor man pages syntax corrections. Thanks to
612 Chris Danis <screechco@home.com> for reporting the bugs.
613
614 3. Added a script from David B. Peterson <dave@toppledwagon.com>
615 to the examples section. It features dumping several
616 filesystems to a remote tape drive upon ssh.
617
618 4. Added a patch provided by Richard Jones <rich@annexia.org>
619 which allows BRADEMAX (number of read errors tolerated by
620 dump) to be adjusted using the -I option.
621
622 5. Fixed a bug which disabled doing "restore -C -f -". Thanks
623 to Clemens Stiglechner <a9401816@unet.univie.ac.at> for the
624 patch.
625
626 6. Add the -l option to restore to specify if, when doing a
627 remote restore, the file used to access the backup is a
628 regular file (the defaults being a tape drive). Restore needs
629 to know this information when reading a remote compressed
630 dump. Previously, this information was autodetected, but
631 the autodetection code fails (with ioctl: Inappropriate ioctl
632 for device) when using a non Linux remote box. Thanks to
633 many users and especially to Eros Albertazzi
634 <eros@lamel.bo.cnr.it> for reporting this.
635
636 7. Found a workaroung for the dump deadlock problem (3 childs
637 stuck in pause(), father in read()). The workaround seems
638 to work for me and several beta-testers. If it doesn't work
639 for you, please report back.
640
641 8. Updated the RPM spec file (BuildPrereq, URL etc).
642
643 Changes between versions 0.4b23 and 0.4b24 (released September 12, 2001)
644 ========================================================================
645
646 1. Fixed the permissions of a newly created QFA file by dump.
647
648 2. Cleaned up the source of dump (the external variables
649 definition was a complete mess, making possible to have
650 objects overlap).
651
652 3. Fixed restore to use the full tape volume path when doing
653 a compare (since it changes the working directory to the
654 filesystem being compared in the process).
655
656 4. Added the -q option to dump which makes dump abort
657 whenever it needs operator attention. It should be
658 easier to use dump in scripts with this option.
659
660 5. Detect the use of incompatible options to dump and
661 refuse them (like -a and -B options together).
662
663 6. Added bzip2 compression to dump/restore (use option -j level
664 to select it). Note that you will need a restore version
665 >= 0.4b24 in order to restore a bzip2 compressed dump.
666 The same warning as for the zlib compression applies:
667 the tape format of a bzip2 dump is not compatible with the
668 original BSD tape format.
669
670 7. Fixed a overflow problem in dump corrupting the dump when
671 very large files were encountered. Thanks to Vladimir Nadvornik
672 for the bug report.
673
674 8. Added a ioctl(BLKFLSBUF) in dump which should flush the
675 kernel buffer/page cache before starting the dump, helping
676 a bit those who use dump on mounted filesystems. Thanks to
677 John Yu <jky@it.bu.edu> and to Theodore T'so <tytso@mit.edu>
678 for this suggestion.
679
680 9. Updated the RPM spec file following the RedHat changes
681 (dynamically linked binaries now in /usr/sbin etc).
682
683 10. Added a patch from Helmut Jarausch <jarausch@igpm.rwth-aachen.de>
684 which enables restore to recognise multi volume compressed dumps
685 done on CD. Included his scripts for dump (which pipe the dump
686 output directly into cdrecord) and restore. There is now
687 possible to to backups to CD on the fly!
688
689 Changes between versions 0.4b22 and 0.4b23 (released July 20, 2001)
690 ===================================================================
691
692 1. Fixed a buffer overflow in restore/tape.c. Patch provided by
693 Marcus Meissner (Caldera International Security Dept.).
694
695 2. Implement the Sun rmt extended protocol. Patch submitted
696 by Ian Gordon <iangordon@users.sourceforge.net>.
697
698 3. Check for the e2fsprogs header <ext2fs/ext2_fs.h> instead of
699 the linux kernel header. This ensures that dump always has the
700 latest version of this file. Patch submitted by
701 Andreas Dilger <adilger@turbolinux.com>.
702
703 4. Report any filesystem present in either /etc/fstab with a
704 positive passno or /etc/dumpdates in dump -w output.
705 Patch submitted by Andreas Dilger <adilger@turbolinux.com>.
706
707 5. Fixed the looping problem in dump introduced in the
708 previous version.
709
710 6. Changed the -B option of dump to limit the size of
711 _compressed_ output per volume if compression is on.
712 Patch contributed by Helmut Jarausch
713 <jarausch@igpm.rwth-aachen.de>. Note however that, since
714 it is impossible to predict the size of the compressed
715 data before doing the compression, the -B limit is a bit
716 conservative.
717
718 7. Fixed a bug in reading the operator typed file/tape path for
719 the next volume in restore.
720
721 8. Implemented a "-F script" option to restore which permits the
722 user to specify a script which will be launched at the
723 beginning of each tape, useful for automatic programming of
724 tape changers for example. See the restore man page for the
725 script parameters and return codes.
726
727 9. Small fix for the QFA routines provided by Uwe Gohlke
728 <uwe@ugsoft.de>, and some recommendations for QFA uses in
729 the man pages.
730
731 10. Fixed the multivolume restoring where making a mistake
732 to the 'Mount next tape' prompt caused several blocks to
733 be lost.
734
735 11. Enhanced the -e option of dump to take as a parameter a
736 comma separated list of inode numbers.
737
738 12. Added the -E option to dump which specify a file containing
739 inode numbers to exclude from the dump.
740
741 13. Fixed the compressed multi-volume dump + restore.
742
743 Changes between versions 0.4b21 and 0.4b22 (released May 12, 2001)
744 ==================================================================
745
746 1. Made dump report the number of blocks dumped per volume.
747 Thanks to Kenneth Porter <shiva@well.com> for the suggestion.
748
749 2. Fix a bug in dump where waiting too much at the 'change volumes'
750 question caused the volume to be rewritten. Thanks to
751 Graham Reed <greed@users.sourceforge.net> for reporting the
752 bug and providing a patch.
753
754 3. Added a compression option to dump, contributed by Jerry
755 Peters <gapeters@worldnet.att.net>.
756
757 WARNING: the compression works only on files/pipes and
758 drives supporting variable block size.
759
760 WARNING2: tapes created using the compression feature are
761 incompatible with the standard dump tape format, and a
762 version of restore >= 0.4b22 will be needed for extraction.
763
764 4. Fixed some compilation problems for glibc 2.2.2 and 64 bit
765 architectures. Thanks to Paul Slootman <paul@debian.org> for
766 the patch and to Bdale Garbee <bdale@gag.com> for forwarding
767 it upstream.
768
769 5. Many cleanups (CPP defines, const char warnings, check of
770 ext2fs COMPAT flags, time_t cleanups, added libext2 version
771 in dump usage text) by Andreas Dilger <adilger@turbolinux.com>.
772
773 6. Made --prefix option work in configure. All the install path
774 are now based on the configure parameters.
775
776 7. Added the Quick File Access mode in dump/restore, contributed
777 by Uwe Gohlke <uwe@ugsoft.de>. In this mode, dump stores in
778 a file tape position for each inode, and this file is used by
779 restore (if called with parameter Q and the filename)
780 to directly position the tape at the file restore is currently
781 working on. This saves hours when restoring single files from
782 large backups, saves the tapes and the drive's head. Use
783 --enable-qfa option of configure to compile in the QFA support.
784
785 8. Added the possibility to dump several files and directories
786 in a single invocation of dump. Thanks to Uwe Gohlke
787 <uwe@ugsoft.de> for implementing this option.
788
789 9. Fixed the dumping and restoring of files with holes
790 (files ending with holes didn't get dumped or restored
791 correctly).
792
793 10. Fixed a socket descriptor leak in dump, which leaved opened
794 3 file descriptors per dump process (and there is one dump
795 process per tape).
796
797 11. Fixed dump large file system support, by explicit use of
798 open64/lseek64/etc functions (explicit use needed because
799 e2fsprogs libraries don't behave well when compiled with
800 FILE_OFFSET_BITS=64).
801
802 Changes between versions 0.4b20 and 0.4b21 (released January 13, 2001)
803 ======================================================================
804
805 1. Fixed some bugs in the dump -w|-W logic introduced by
806 the previous version. Thanks to Andreas Dilger
807 <adilger@turbolinux.com> for his help on this one.
808
809 2. Fixed again a compilation problem when using old e2fs
810 headers (filesystem label related). Thanks to many users
811 who reported this stupid error.
812
813 3. Fixed a build problem on old lib5 based systems dealing with
814 _PATH_MNTTAB being undefined. Thanks to John Adams
815 <johna@onevista.com> for reporting the bug.
816
817 4. Improved the error detection in restore when creating
818 the temporary files in TMPDIR. Now restore will corectly
819 report a 'no space left on device' error instead of
820 strange errors which could imply an archive corruption.
821 Thanks to Gabriel Somlo <somlo@cs.colostate.edu> and
822 bgjenero <bgjenero@sympatico.ca> for reporting the bug.
823
824 5. Added the throughput information in dump's progression
825 messages. Thanks to Andreas Dilger <adilger@turbolinux.com>
826 for the patch.
827
828 6. Use libext2fs's inode scanning routines, which are particularly
829 robust when dealing with errors (especially when having some
830 bad blocks on the filesystem). Many thanks to Colin
831 <colin@horizon.com> for reporting the bug, investigating
832 the issues, finding a workaround, writing the patch and
833 fully testing it... (of course, if this patch will break
834 anything, he is to blame for :-)).
835
836 7. Made dump and restore LFS aware. Dump can dump a filesystem
837 containing large files, generate a large file on output and
838 restore can restore them. This feature is not enabled by
839 default, use --enable-largefile option of configure to enable
840 it (you need to have a LFS aware glibc though). Thanks to
841 Andrea Arcangeli <andrea@suse.de> for submitting the patch,
842 and to Theodore T'so <tytso@valinux.com> for his always
843 useful thoughts.
844
845 8. Made dump ask upon a tape write error if it should rewrite
846 the current volume (assume this is a bad tape) or if it should
847 assume an end-of-tape condition (useful for tape drives which
848 behaves badly at the end of the tape). Thanks to Andreas
849 Dilger <adilger@turbolinux.com> for the suggestion.
850
851 Changes between versions 0.4b19 and 0.4b20 (released November 10, 2000)
852 =======================================================================
853
854 1. Fixed a small compilation problem due to a change
855 in the definintion of the struct sigaction in
856 glibc 2.0 / libc5. Thanks to Gunther Schlegel
857 <schlegel@riege.de> for reporting the bug and to
858 Dave Platt <dplatt@snulbug.mtview.ca.us> for suggesting
859 a fix.
860
861 2. Modified the label and uuid analysis in order to be
862 self-contained (independant of kernel/libc headers). This
863 should fix the compile with older kernel/libc headers and
864 will preserve the functionality. Thanks to Bernhard Erdmann
865 <bernhard.erdmann@gmx.de> for reporting the bug.
866
867 3. The 'exclude inode' option, if applied to a directory
868 inode, excludes now correctly all the leaves of this
869 directory. Thanks to John R. Dennison
870 <gerdesas@users.sourceforge.net> for reporting the bug.
871
872 4. Fixed the '-e' option to disable the possibility
873 to exclude the root inode (which causes the dumps to
874 be unreadable by restore). Prevented array overflow
875 when multiple -e options are used.
876
877 5. Fixed dump to correctly interpret a filesystem argument
878 which is a mountpoint and it is not an absolute path
879 (as specified in the fstab). Thanks to Bernhard R. Erdmann
880 <be@berdmann.de> for reporting the bug.
881
882 6. Made dump able to backup files larger than 2 GB. Note that
883 dump still doesn't cope with files larger than 4 GB.
884
885 7. Restore the real uid and gid before invoking an external
886 RSH program (big hole when dump or restore is suid root!).
887
888 8. Get the values for the default tape device and the location
889 of fstab file from the system headers. Thanks to
890 Andreas Dilger <adilger@turbolinux.com> for the patch.
891
892 9. Made dump -w|-W report all recognized filesystems
893 present in either /etc/fstab or /etc/dumpdates, and present
894 the list in the same order as in fstab file. Thanks
895 to Andreas Dilger <adilger@turbolinux.com> for the patch.
896
897 10. Made dump's -a (automatic end of tape detection) the
898 default. Specifying one of -B, -c, -d or -s options will
899 override the EOM detection. Thanks to Andreas Dilger
900 <adilger@turbolinux.com> for the patch.
901
902 11. Save the ext2 filesystem label into the dump volume label.
903 Specifying a volume label on the command line (-L option)
904 will override this feature. Thanks to Andreas Dilger
905 <adilger@turbolinux.com> for the patch.
906
907 Changes between versions 0.4b18 and 0.4b19 (released August 20, 2000)
908 =====================================================================
909
910 1. Fixed the signal handling in dump (which I broke in 0.4b17)
911 which was causing several strange dump failures (dump
912 hanged or segmentation faults etc).
913
914 2. Specified the default blocksize in the dump man page.
915
916 3. Changed two info messages of restore to be written on stdout
917 instead of stderr in order to leave stderr only for errors
918 or warnings. Thanks to Stephen Atwell
919 <satwell@urbana.css.mot.com> for the suggestion.
920
921 4. Corrected an off by one calculation which prevented
922 dumping a directory having a 1 character name length.
923 Thanks to Bernhard Erdmann <bernhard.erdmann@gmx.de>
924 for reporting the bug.
925
926 5. Reinforce dump in order to allow only level 0 and no
927 -u option when dumping a subdirectory, and document
928 this in the man page. Thanks to Bernhard Erdmann
929 <bernhard.erdmann@gmx.de> for reporting the bug.
930
931 6. Fixed a small harmless bug in restore which caused
932 in some conditions hard links to be restored several
933 times (therefore generation some warning messages).
934 Thanks to Kenneth Porter <shiva@well.com> for
935 reporting the bug.
936
937 7. Updated the RPM spec file to the latest RedHat version,
938 providing FHS packaging and other cosmetic changes.
939 You will need RPM version >= 3.0.5 in order to build the RPM.
940
941 8. Updated the configure script to check for libtermcap
942 before checking for libreadline, because we need this
943 library in order to compile the readline support.
944
945 9. Made dump understand the LABEL= and UUID= notation
946 both in /etc/fstab and on the command line. Note that
947 you will need the /proc filesystem in order to use
948 these notations. Thanks to Erik Troan <ewt@redhat.com>
949 for providing the patch.
950
951 Changes between versions 0.4b17 and 0.4b18 (released June 30, 2000)
952 ===================================================================
953
954 1. Fixed a potential buffer overflow in restore. Thanks
955 to Stan Bubrouski <satan@fastdial.net> for reporting
956 the bug.
957
958 2. Fixed a readline-related bug which prevented
959 'cat DUMPFILE | restore -i -f -' from working. Thanks
960 to Charles Karney <karney@users.sourceforge.net>
961 for the bug report.
962
963 3. Changed a few "panic" into "exit", causing restore to
964 be more stable against some attacks (like the last one
965 reported on Bugtraq, although the last version of restore
966 was not vulnerable - just dumped core). Thanks to
967 Andreas Hasenack <andreas@conectiva.com.br> for reporting
968 the bugs to me.
969
970 4. Removed the suid-root bit on dump and restore in the
971 default build (and generated RPMs). It should be safer
972 now. Users who need the suid-root capabilities in order
973 to make network backups should read first the man page
974 of dump and enable the suid bit manually.
975
976 5. Added -ltermcap to the compile parameters for restore
977 when using readline support, in order to make the compile
978 process work on some systems (SuSE ?). Thanks to
979 Patrik Schindler <poc@pocnet.net> for reporting the bug.
980
981 Changes between versions 0.4b16 and 0.4b17 (released June 1st, 2000)
982 ====================================================================
983
984 1. The -F script is called now *only* at the end of a tape,
985 not at the end of the dump. Thanks to Christian Haul
986 <haul@informatik.tu-darmstadt.de> for the bug report.
987
988 Normally, the device name and the current volume number
989 are passed on the command line of the script. However,
990 if you want the old style script (with no arguments
991 passed on the command line), you can enable it in
992 configure with the --enable-oldstylefscript.
993
994 2. Use posix signal handling to preserve dump functionality
995 with libc5. Thanks to Jeff Johnson <jbj@redhat.com> for
996 the patch.
997
998 3. Made the exit code of restore in case of a 'C'ompare
999 command reflect the result of the compare. An exit status
1000 of 0 means the dump archive is correct, 1 means tape errors,
1001 2 means that some files were modified. Thanks to Kenneth Porter
1002 <shiva@well.com> for the suggestion.
1003
1004 4. Made (finally) quotes work in interactive restore.
1005
1006 5. Small fixes in order to allow dump to compile with a
1007 really old e2fsprogs version. Thanks to Ian Zimmerman
1008 <itz@speakeasy.org> for the bug report.
1009
1010 6. Add GNU readline capabilities to interactive restore.
1011 Use configure's --enable-readline flag to enable this feature.
1012 Thanks to Patrik Schindler <poc@pocnet.net> for the
1013 suggestion.
1014
1015 7. Do the compare on the fly in restore 'C' mode (this will
1016 allow not to exhaust the available /tmp space when
1017 ccmparing large files). Thanks to Kenneth Porter
1018 <shiva@well.com> for the suggestion.
1019
1020 Changes between versions 0.4b15 and 0.4b16 (released March 11, 2000)
1021 ====================================================================
1022
1023 1. Fixed some several duplicate 'DUMP: DUMP:' in the
1024 output of dump.
1025
1026 2. Corrected the estimation of blocks to dump. Note that
1027 this is still wrong for large directory inodes, due
1028 to the size of a BSD directory entry on the tape
1029 (estimation is based on the size of the ext2 directory,
1030 which is smaller because it doesn't contain the null
1031 character at the end).
1032
1033 3. Made dump report the total number of MB written to
1034 tape at the end of run. Thanks to W. Reilly Cooley
1035 <wcooley@nakedape.cc> for the patch.
1036
1037 4. Added the -X option to restore giving the possibility
1038 to read the names of the files to be extracted/listed
1039 from a text file (in addition of the command line).
1040 Thanks to Dejan Muhamedagic <dejan@quant-x.com> for the
1041 patch.
1042
1043 5. Added the device name and the current volume number
1044 as arguments to the end of tape script (dump -F option).
1045
1046 6. Made the multi-volume dump work again (after having
1047 broken it in 0.4b15).
1048
1049 Changes between versions 0.4b14 and 0.4b15 (released March 2, 2000)
1050 ===================================================================
1051
1052 1. Added a prompt command in interactive restore mode. Thanks
1053 to Andreas Dilger <adilger@home.com> for the patch.
1054
1055 2. Fixed a buffer overflow problem in dump (caused by
1056 not checking the size of the filesystem parameter).
1057 Thanks to Kim Yong-jun <loveyou@hackerslab.org> for
1058 reporting this on Bugtraq (and to several dump users
1059 who forwarded me his mail).
1060
1061 3. Added the '-F script' option to dump in order to
1062 launch a script at the end of each tape (to be used
1063 with a tape changer, or to notify the sysadmin by
1064 pager etc.).
1065
1066 4. Fixed a bug in restore compare code caused by the changes
1067 I made in 0.4b14.
1068
1069 5. Fixed the treatment of options using the old BSD syntax
1070 in both dump and restore.
1071
1072 Changes between versions 0.4b13 and 0.4b14 (released February 10, 2000)
1073 =======================================================================
1074
1075 1. Fixed a bug in dump which may have caused invalid deleted
1076 directories to be dumped out if they were deleted after the
1077 mapping pass. This could occure on active filesystem and lead
1078 to heap corruption (causing dump malfunction in many possible ways).
1079 Thanks to Jan Sanislo <oystr@cs.washington.edu> for finding this
1080 bug and submitting the patch.
1081
1082 2. Fixed the handling of the filesystem parameter in dump. A
1083 '/mnt/disk' parameter caused the disk contents to be dumped,
1084 but a '/mnt/disk/' parameter caused the mountpoint directory
1085 to be dumped (generally an empty directory).
1086
1087 3. Improved the output of dump in order to tell which directory
1088 it is currently dumping (when dumping a subtree).
1089
1090 4. Added the '-e' exclude inode option to dump. Thanks to
1091 Isaac Chuang <ike@isl.stanford.edu> for contributing with the patch.
1092
1093 5. Added a REPORTING-BUGS file in order to provide a guide
1094 on how to correctly report a bug in dump/restore.
1095
1096 6. Fixed a restore problem when restoring a hard link to an inode
1097 having the immutable or append-only attribute set. Thanks to
1098 Ambrose Li <acli@mingpaoxpress.com> for submitting the patch.
1099
1100 7. Fixed a compatibility problem between dumps made on little
1101 endian machines (the format was correct) and big endian
1102 machines (the format was incorrect). This fix break the
1103 compatibility with the older dumps made on big endian
1104 machines (sparc, m86k, ppc etc). For the first time in
1105 linux dump's history (I believe), the dumps made by *BSD,
1106 Linux/alpha, Linux/sparc and Linux/x86 are compatible,
1107 so interchangeable. Thanks to Rob Cermak
1108 <cermak@ahab.rutgers.edu> for submitting the bug and
1109 helping me test the fix.
1110
1111 8. Fixed the way dump reports the remaining percent/time, if
1112 the number of blocks actually dumped exceeds the estimated
1113 number of blocks. Thanks to Jean-Paul van der Jagt
1114 <jeanpaul@dutepp0.et.tudelft.nl> for reporting the bug.
1115
1116 Changes between versions 0.4b12 and 0.4b13 (released January 21, 2000)
1117 ======================================================================
1118
1119 1. Small Y2K fix in the man pages :). Thanks to Bernhard Sadlowski
1120 <sadlowsk@Mathematik.Uni-Bielefeld.DE> for reporting the bug.
1121
1122 2. Removed the requirement to build the RPM as root from the
1123 spec file. Thanks to Christian Weisgerber
1124 <naddy@mips.rhein-neckar.de> for submitting this.
1125
1126 3. Fixed a bug in dump related to the 'filetype' feature of ext2fs,
1127 causing dump to block when dumping really huge filesystems.
1128 Many thanks to Patrik Schindler <poc@pocnet.net> for
1129 helping me find this bug.
1130
1131 4. Fixed the treatment for an interrupt signal when dump access
1132 the remote tape through RSH. Thanks to Christian Weisgerber
1133 <naddy@mips.rhein-neckar.de> for providing the patch.
1134
1135 5. Fixed a bug which was causing dump/restore to display
1136 garbage characters instead of the remote host name.
1137
1138 Changes between versions 0.4b11 and 0.4b12 (released January 8, 2000)
1139 =====================================================================
1140
1141 1. Small fix in the dump man page. Thanks to Thorsten Kukuk
1142 <kukuk@suse.de> for submitting the patch.
1143
1144 2. Fix for the exit code when using the size estimate option of
1145 dump. Thanks to Matti Taskinen <mkt@rni.helsinki.fi> for
1146 submitting the patch.
1147
1148 3. Handle EINTR in atomical reads/writes in dump, which was causing
1149 dump to fail on some systems. Thanks to Eric Jergensen
1150 <eric@dvns.com> for reporting the bug and submitting the patch.
1151
1152 4. Handle more than 16 characters for the device names in dumpdates.
1153 (up to 255 now). Thanks to Rainer Clasen <bj@ncc.cicely.de> for
1154 tracking down the problem and proposing the solution.
1155
1156 5. Fixed a bug in dump which prevented the creation of the
1157 dumpdates file when doing a 0-level dump without already
1158 having a dumpdates file. Thanks to Patrik Schindler
1159 <poc@pocnet.net> for reporting the bug.
1160
1161 6. Changed the way dump 'S' flag reports the size estimate
1162 from number of blocks into bytes (making it compatible
1163 with the Solaris version, and simplifying things for
1164 amanda users). Thanks to Jason L Tibbitts III
1165 <tibbs@math.uh.edu> for reporting the bug.
1166
1167 7. Fixed a compatibility problem in linux/alpha dump tape format.
1168 Now the linux/alpha dump are (again) compatible with the
1169 other dump formats. But this breaks compatibility with
1170 older dumps made on alpha. Thanks to Mike Tibor
1171 <tibor@lib.uaa.alaska.edu> for helping me in finding this bug.
1172
1173 Changes between versions 0.4b10 and 0.4b11 (released December 5, 1999)
1174 ======================================================================
1175
1176 1. Added a '--enable-kerberos' to configure.
1177
1178 2. Added a 'S' option to dump which determines the amount of space
1179 that is needed to perform the dump without actually doing it, similar
1180 to the Sun's ufsdump 'S' option. Patch contributed by Rob Cermak
1181 <cermak@ahab.rutgers.edu>.
1182
1183 3. Added a 'M' multi-volume option to dump and restore which enables
1184 dumping to multiple files (useful when dumping to an ext2
1185 partition to make several dump files in order to bypass the 2GB
1186 file size limitation). The 'f' argument is treated as a prefix and
1187 the output files will be named <prefix>001, <prefix>002 etc. With
1188 the 'M' flag, restore automatically selects the right file without
1189 asking to enter a new tape each time.
1190
1191 4. Fixed a memory leak which was causing dump to grow very big
1192 (270MB when dumping a 10GB filesystem...). Thanks to Jason
1193 Fearon <jasonf@netrider.org.au> for reporting the bug.
1194
1195 Changes between versions 0.4b9 and 0.4b10 (released November 21, 1999)
1196 ======================================================================
1197
1198 1. Make configure test if the system glob routines support
1199 extended syntax (ALTDIRFUNC). If not, use the internal glob
1200 routines instead of system ones. Thanks to Bernhard Sadlowski
1201 <sadlowsk@Mathematik.Uni-Bielefeld.DE> for reporting the bug
1202 and helping me resolve this and other minor libc5 compiling
1203 glitches.
1204
1205 2. Fix a problem when dumping a ext2fs with the 'filetype'
1206 feature enabled. Thanks to Patrick J. LoPresti
1207 <patl@cag.lcs.mit.edu> for reporting the bug and to
1208 Theodore Y. Ts'o <tytso@mit.edu> for providing the patch.
1209
1210 3. Made the nodump flag work on directories. A directory which
1211 has the nodump flag gets never dumped, regardless of its
1212 contents.
1213
1214 4. Integrate a patch from Jeremy Fitzhardinge <jeremy@goop.org>
1215 which allows dump on an active ext3 filesystem. However, this
1216 is a "quick and dirty" patch which enables backup of an ext3
1217 filesystem through the ext2 compatibility (by ignoring the
1218 NEEDS_RECOVERY bit). The journal file is not recognized and
1219 it is dumped (it should not).
1220
1221 5. Test the superblock compatibility flags when dumping, in order
1222 to be sure that we know how to deal with specific features.
1223
1224 Changes between versions 0.4b8 and 0.4b9 (released November 5, 1999)
1225 ====================================================================
1226
1227 1. Use lchown instead of chown, fixing a possible security problem
1228 when restoring symlinks (a malicious user could use this
1229 to deliberately corrupt the ownership of important system files).
1230 Thanks to Chris Siebenmann <cks@utcc.utoronto.ca> for detecting
1231 this and providing the patch.
1232
1233 Changes between versions 0.4b7 and 0.4b8 (released November 3, 1999)
1234 ====================================================================
1235
1236 1. Put dump sources under CVS, added Id tags in all files so
1237 one can use 'ident' on binary files.
1238
1239 2. Added the dump/restore version in the usage text so one can
1240 easily verify the version he is using.
1241
1242 3. Small patch from Nuno Oliveira <nuno@eq.uc.pt> which fixes
1243 a va_start/va_end problem on linux-ppc (always call va_start
1244 va_end in pairs each time we use a vararg function).
1245
1246 4. Added again the DT_* constants because old libc does not
1247 contain them :(. Thanks to Eric Maisonobe <virnet@nat.fr>
1248 for submitting the bug report.
1249
1250 5. Use ext2fs_llseek instead of llseek. With recent e2fsprogs
1251 this should enable dumping big (huge) filesystems.
1252
1253 6. Added the RSH environment variable in order to be able to
1254 use a rsh replacement like ssh when doing remote backups (and
1255 bypass the security limitations of rcmd). Now you can do remote
1256 backups without being root (or making dump setuid root).
1257
1258 7. Modified again the way dumpdates works. For incremental dumps,
1259 we need to read dumpdates even if we are not using 'u' option.
1260 Thanks to Bdale Garbee <bdale@gag.com> for his ideas on how
1261 this should work.
1262
1263 Changes between versions 0.4b6 and 0.4b7 (released October 8, 1999)
1264 ===================================================================
1265
1266 1. Removed the 'k' flag from the restore 'about' text if kerberos
1267 was not compiled in.
1268
1269 2. Prototyped (f)setflags from e2fsprogs and corrected the calls
1270 to them (fsetflags takes a char*, setflags an open fd!).
1271
1272 3. (f)setflags is called only if the flags aren't empty. If the
1273 file is a special file, a warning is printed, because changing
1274 flags implies opening the device. Normally, a special file
1275 should not have any flag... (Debian bug #29775, patch provided
1276 by Abhijit Dasgupta <abhijit@ans.net>).
1277
1278 4. Made possible to dump a file system not mentioned in /etc/fstab.
1279 (Debian bug #11904, patch provided by Eirik Fuller <eirik@netcom.com>).
1280
1281 5. Changed the default behaviour to not create dumpdates
1282 unless 'u' option is specified. Removed the old "debian-patch"
1283 which provided the same thing. (Debian bug #38136, #33818).
1284
1285 6. Removed all those dump*announce, since they were getting old...
1286
1287 7. Added warning messages if dumpdates does not exist and
1288 when an inferior level dump does not exist (except for a level 0
1289 dump).
1290
1291 8. Debugged the glob calls in interactive mode: restore used a
1292 dirent struct which was different from the /usr/include/dirent.h
1293 one (this used to work, is it a glibc2 change?), so none of the
1294 compat glob (which used /usr/include/dirent.h) or the system glob
1295 worked. Restore use now the system dirent (and the system
1296 DT_* constants), which are compatible with BSD ones.
1297
1298 9. Added a configure flag (--with-dumpdatespath) to specify
1299 the location of dumpdates file. By default, it is
1300 /etc/dumpdates.
1301
1302 10. Added the "AUTHOR" and "AVAILABILITY" sections and
1303 included the current date/version in man pages.
1304
1305 11. Corrected the estimation of remaining time when
1306 the operator doesn't change the tapes quickly enough. This
1307 was an old bug, I thought I corrected it, and discovered
1308 that in fact it was corrected in two different places, so
1309 the results canceled each other...
1310
1311 Changes between versions 0.4b5 and 0.4b6 (released October 1, 1999)
1312 ===================================================================
1313
1314 1. Integrated multiple patches from RedHat, Debian and SuSE:
1315
1316 - tweak dump/itime.c to not try to read dumpdates if the 'u' option
1317 isn't specified.
1318 - several fixes in the man pages.
1319 - update the default tape device to /dev/st0.
1320 - many updates for Linux Alpha (byte ordering, size_t etc).
1321 - buffer overruns.
1322 - use environment variable for TMPDIR (instead of /tmp).
1323 - use sigjmp_buf instead of jmp_buf (RedHat bug #3260).
1324 - workaround egcs bug (RedHat bugs #4281 and #2989).
1325 - wire $(OPT) throughout Makefile's.
1326
1327 2. Upgrade the dump revision to 1, making possible to dump filesystems
1328 made with e2fsprogs-1.15 or newer. Nothing seems to break...
1329
1330 3. Fix some compile warnings, prototype all functions.
1331
1332 4. Use glibc err/glob instead of internal compatibility
1333 routines (only if available).
1334
1335 5. Fix a compile error on Linux 2.2.7 / libc5 (5.4.44) (patch provided
1336 by Bernhard Sadlowski <sadlowsk@mathematik.uni-bielefeld.de>).
1337
1338 Changes between versions 0.4b4 and 0.4b5 (released September 22, 1999)
1339 ======================================================================
1340
1341 1. Integrated the changes from FreeBSD-3.1-RELEASE
1342 (mostly bug fixes, buffer overruns, dump has now an "automatic
1343 tape length calculation" flag, dump/restore can use kerberos now
1344 (this is NOT tested), use environment variables for TAPE and
1345 RMT etc.).
1346
1347 2. Integrated three RedHat patches ("glibc", "kernel" and "bread" patches)
1348
1349 3. Corrected a bug in restore when using 'C' option with multi-volumes
1350 tapes (files splited accros two tapes give "size changed" errors
1351 when comparing).
1352
1353 4. Corrected the long standing bug when dumping multiple tapes.
1354 This works for me, needs further testing.
1355
1356 Changes between versions 0.4b3 and 0.4b4 (released January 17, 1997)
1357 ====================================================================
1358
1359 1. Dump now runs correctly on kernels 2.1.x
1360 Fix made by Gerald Peters <gapeters@worldnet.att.net>
1361
1362 Changes between versions 0.4b2 and 0.4b3
1363 ========================================
1364
1365 1. Use realpath() if available
1366
1367 2. Report statistics
1368
1369 Changes between versions 0.4b1 and 0.4b2
1370 ========================================
1371
1372 1. Fixed the bug fix from Greg Lutz (I had made a mistake when integrating
1373 the patch)
1374
1375 2. Fixed restore to make it able to read FreeBSD 2.x dumps again
1376
1377 3. Fixed configure.in to correctly handle --enable-rmt
1378
1379 Changes between versions 0.3 and 0.4b1
1380 ======================================
1381
1382 1. Integrated the changes from 4.4BSD-Lite2
1383
1384 2. Integrated the patches from Debian and Red Hat
1385
1386 3. Portability changes: use the __u32, __u16, __s32, and __s16 types
1387
1388 4. Changed dump to use the Ext2fs library to get block addresses. This
1389 should solve the endianness problem on SparcLinux.
1390
1391 5. Created a configure.in file (shamelessly stolen from the e2fsprogs
1392 distribution's one) to use autoconf
1393
1394 6. Fixed a few minor bugs
1395
1396 Changes between versions 0.2e and 0.2f
1397 ======================================
1398
1399 1. Added the creation of named pipes (fifos) in restore.
1400
1401 2. Added the -N flag in the restore manual page.
1402
1403 3. Added the file kernel.patch which contains the llseek() optimization
1404 patch for 1.2.x kernels.
1405
1406 4. Fixed a bug in the restoration of symbolic links: owner and group were
1407 not restored.
1408
1409 5. Integrated some changes from FreeBSD 2.2.
1410
1411 6. Added a call to ftruncate() after restoring each file to restore
1412 correctly files ending by a hole.
1413
1414 Changes between versions 0.2d and 0.2e
1415 ======================================
1416
1417 1. Fixed a bug in the "set owner/mode" process. Every file was restored
1418 with owner = root (0) and group = root/wheel/whatever (0).
1419
1420 Changes between versions 0.2c and 0.2d
1421 ======================================
1422
1423 1. Dump is now able to backup 2GB+ filesystems.
1424
1425 2. Dump and restore can now be linked as static binaries.
1426
1427 Changes between versions 0.2b and 0.2c
1428 ======================================
1429
1430 1. Fixed a bug when dumping ``slow'' (i.e. normal) symbolic links.
1431
1432 Changes between versions 0.2a and 0.2b
1433 ======================================
1434
1435 1. Really fixed the bug that I should have corrected in 0.2a.
1436
1437 2. Enabled optimization again.
1438
1439 Changes between versions 0.2 and 0.2a
1440 =====================================
1441
1442 1. Disabled the optimization during compilation.
1443
1444 Changes between versions 0.1 and 0.2
1445 ====================================
1446
1447 1. Fixed a bug in fstab.c which caused a null pointer to be stored in
1448 the fs_type field (actually, I modified the file fstab.c to make it
1449 use the mntent functions).
1450
1451 2. Dump and restore now use a 4.3 BSD compatible dump format. Backups
1452 made by dump should be readable by the BSD restore and backups made
1453 by the BSD dump should be readable by restore. Unfortunately, this
1454 means that the dump format has changed between version 0.1 and version
1455 0.2 :-(
1456
1457 3. Dump is now able to backup a subtree, it is no longer limited to whole
1458 filesystems like the BSD version.
1459
1460 4. Dump now uses ext2_llseek() so it is able to backup filesystems bigger
1461 than 2 GB.
1462
1463 Changes between versions 0.0 and 0.1
1464 ====================================
1465
1466 1. Now create links rdump and rrestore during the `make install' step.
1467
1468 2. Linux port specific bugs added to the manual pages
1469
1470 3. Incorrect estimation of the number of tapes blocks fixed when doing
1471 incremental backups.
1472
1473 4. Better ls-like format in restore in interactive mode.