In restore -C, encountering an unexpected file is not an error.
[dump.git] / CHANGES
1 $Id: CHANGES,v 1.289 2008/04/17 15:22:56 stelian Exp $
2
3 Changes between versions 0.4b41 and 0.4b42 (released xxxxxxxxxxxxxxx)
4 =====================================================================
5
6 1. Fix printout of the 'Connection to' message in dump/restore.
7 Thanks to Dale Wiles <dwiles@users.sourceforge.net> for the
8 bug report and its fix.
9
10 2. Fix dump -w/-W output for filesystems having the last backup
11 level equal to 0. Thanks to Pascal Bouchareine
12 <kalou@users.sourceforge.net> for reporting the bug.
13
14 3. Enable restore to handle restoring onto a different SELinux
15 policy from the dump. Thanks to Tony Nelson
16 <tony_nelson@users.sourceforge.net> for the patch.
17 (see http://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=189845
18 for details)
19
20 4. Fix EA set failures when restoring immutable files. Thanks
21 to Andrew Kroeger for the patch.
22
23 5. Fix "mode file too large" errors when restoring huge backups.
24 Thanks to Kenneth Porter <shiva@sewingwitch.com> for the bug
25 report.
26
27 6. Add the ability to link against libtinfo library if present,
28 instead of libtermcap or libncurses libraries. Thanks to
29 Adam Tkac <atkac@redhat.com> for the patch.
30
31 7. When comparing a backup, do not consider to be an error the fact
32 that we encounter files created while dump was in progress.
33 Thanks to Kenneth Porter <shiva@sewingwitch.com> for the bug
34 report.
35
36 Changes between versions 0.4b40 and 0.4b41 (released January 2, 2006)
37 =====================================================================
38
39 1. Fix restore of dumped Access Control Lists. The previous
40 code for EA works fine, but ACLs needed conversion from the
41 ext2/3 disk format to posix_acl format before restoring.
42
43 2. Fix some issues with restoration of EA on big endian
44 platforms.
45
46 3. Fix restore when the symtab is over 2GB in size.
47
48 4. Made the directory hash indexing an optional feature,
49 accessible by the '-H' option of restore, and disable it
50 by default.
51
52 5. Fixed dump to not include extended attributes information
53 in the toc (archive) file which confused restore -t.
54
55 Changes between versions 0.4b39 and 0.4b40 (released May 2, 2005)
56 =================================================================
57
58 1. Changed restore to emit warnings (instead of emitting a fatal
59 error) if a file (or a directory) is unavailable for a
60 comparision (if the user doesn't have the necessary permissions
61 to access it for example). Thanks to Kenneth Porter
62 <shiva@sewingwitch.com> for the bug report.
63
64 2. Re-done the 'do not save directory entries to non-dumped inodes
65 (excluded from dump)' feature. The previous implementation
66 worked well for excluded directories but not for regular files.
67 Thanks to Kenneth Porter <shiva@sewingwitch.com> for the bug
68 report.
69
70 3. Fixed a bug in dump where the tape size was miscalculated when
71 the user used -d/-s to specify the tape characteristics. Thanks
72 to Philip Goisman <goisman@physics.arizona.edu> for reporting
73 the bug.
74
75 4. Fixed another bug introduced in restore with the hashtree
76 implementation. This one caused restore to stop saying
77 "removenode: non-empty directory" in some cases.
78
79 5. Added support for dumping and restoring ext2/3 extended
80 attributes (EA), like the access control lists (ACL) or
81 the security labels used by SELinux.
82
83 Changes between versions 0.4b38 and 0.4b39 (released January 21, 2005)
84 ======================================================================
85
86 1. The newly added dump_on_cd_3 example was buggy, replace it
87 with an updated version from Andrew Basterfield
88 <bob@cemetery.homeunix.org>.
89
90 2. Made restore to chdir() back into the initial directory when
91 dumping core while aborting a comparision operation. The
92 previous behaviour was to write the corefile at the root of
93 the directory being compared, which could very well be
94 read only and preventing the corefile generation. Thanks
95 to Kenneth Porter <shiva@sewingwitch.com> for the bug report.
96
97 3. Silenced the failure to call fgetflags() when comparing an
98 entry which has no ext2 attributes (as in lsattr()).
99
100 4. Fix a brown paper bug in restore -C which broke restore and
101 caused modifications on the filesystem being compared
102 (directories containing a file with the same name as the
103 directory get renamed to RSTTMP...). Thanks to Kenneth Porter
104 <shiva@sewingwitch.com> for finding the bug and helping me
105 reproduce it.
106
107 5. Made restore -C force the -N flag (no writing allowed on
108 the disk) in order to prevent more bugs like the above one.
109
110 Changes between versions 0.4b37 and 0.4b38 (released January 7, 2005)
111 =====================================================================
112
113 1. Fix a couple of troff syntax bugs in the man pages.
114 Thanks to Eric Raymond <esr@thyrsus.com> for the patch.
115
116 2. Made restore use either libncurses or libtermcap, depending
117 on which one is available at configure time.
118
119 3. Fixed restore negative size display bug when comparing a
120 dump containing files over 2GB. Thanks to Steve Bonds
121 <sbonds@users.sourceforge.net> for the bug report.
122
123 4. Do not save directory entries to non-dumped inodes
124 (excluded from dump). This will eliminate the 'missing
125 file' warnings when doing 'restore -C'.
126
127 5. Fix dump crash when backuping a huge (2TB) filesystem,
128 due to a bogus calculation on the inode map size.
129 Thanks to Kevin B. Haines <K.B.Haines@rl.ac.uk> for
130 submitting the bug and testing the fix.
131
132 6. Fix a problem in restore where the final \0 in the symbolic
133 link names could have been lost, generating corrupt filenames.
134 Thanks to Kyle Wilson <kyle.wilson@amd.com> for reporting the
135 bug.
136
137 7. Implemented a hash list for the directory names in restore.
138 The linear list used before caused problems in interactive
139 restores when dealing with directories having thousands of
140 entries. Thanks to Brian Ristuccia <bristuccia@starentnetworks.com>
141 for reporting the bug.
142
143 8. Improved restore -C, this time including the directory
144 attributes into the comparision.
145
146 9. Made restore understand tapes containing EA/ACLs (which will
147 be dumped by the next version of dump). In this version
148 extended attributes on the tape are ignored, for full EA/ACL
149 support wait for the next version or try the experimental EA
150 patch.
151
152 Changes between versions 0.4b36 and 0.4b37 (released July 7, 2004)
153 ==================================================================
154
155 1. Added the --enable-staticz configure option which enables
156 dump and restore to be linked with static versions of
157 libz and libbz2 (and dynamic versions of all other libraries).
158 This will make Debian users happy, because libz and libbz2
159 were the only needed libraries living in /usr, all the
160 others live in /lib. In case of system emergency, it is
161 better not to have to rely on an extra filesystem.
162 Thanks to Bdale Garbee <bdale@gag.com> for the suggestion.
163
164 2. Fix compilation on (at least the Linux Debian port to) AMD64.
165 (<ext2fs/ext2_types.h> defines some types (__s64 and __u64)
166 that are also defined by <linux/types.h> (<asm/types.h>) and
167 they conflict).
168
169 3. Make dump's reading of the dumpdates file a bit more robust,
170 preventing dump from crashing when the dumpdates file has
171 been modified by hand.
172
173 4. Fixed some offset calculations in dump code which could
174 lead to "bread lseek errors" on large filesystems. Thanks
175 to Bruce Lowekamp <lowekamp@users.sourceforge.net> for
176 reporting this bug and debugging the issue.
177
178 5. Made dump use the blkid library when searching for devices
179 by label or uuid instead of dump's own routines.
180
181 6. Corrected a bug in dump where a wrong LABEL=... line in
182 /etc/fstab could prevent dump from dumping unrelated
183 filesystems. Thanks to Bruce Lowekamp
184 <lowekamp@users.sourceforge.net> for reporting the bug.
185
186 Changes between versions 0.4b35 and 0.4b36 (released April 21, 2004)
187 ====================================================================
188
189 1. Fixed dump compilation with old gcc versions. Thanks to
190 Mike Castle <dalgoda@users.sourceforge.net> for the patch.
191
192 2. Fixed some warnings (howmany, roundup, powerof2 redefined)
193 when compiling against a recent glibc version.
194
195 3. Fixed a bug in restore preventing the read of a dump tape
196 written with Solaris 7 ufsdump. Thanks to Patrick Higgins
197 <phiggins@transzap.com> for reporting the bug and providing
198 the test case.
199
200 4. Changed dump to enable the creation of volumes of different
201 sizes in a single run (make -B accept a list of values).
202 Patch contributed by Florian Zumbiehl <florz@gmx.de>.
203
204 5. Use the glibc provided minor() and major() macros instead
205 of our own bitmask implementation. This should be safe for
206 when the major/minor namespace will migrate to 32 bits.
207 Thanks to Zhang Jun <zhangjun@nanjing-fnst.com> for reporting
208 the bug.
209
210 6. Made explicit in the dump man page that dump will not create
211 a remote file, it will only write to an already existing one.
212
213 7. Another try at making size estimates better again.
214
215 8. Put back the inconditional running of the end-of-tape
216 script which was changed in 0.434 to be run only when -M
217 or multiple -f were NOT specified. Some users rely on this
218 feature even when it is combined with -M/-f.
219
220 9. Fixed restore when restoring huge backups (where rstdir...
221 temporary files are over 2GB). Thanks to
222 Raphael Jaffey <rjaffey@artic.edu> for reporting this,
223 debugging the issue and making the patch.
224
225 10. Made restore understand FreeBSD UFS2 tapes. Thanks to
226 David <vrtin@users.sourceforge.net> for submitting the bug
227 and providing a test case.
228
229 11. Made dump work with an arbitrary integer level (no more 10
230 levels only limitation). Thanks to Matthew
231 <msvincen@midway.uchicago.edu> for the patch.
232
233 Changes between versions 0.4b34 and 0.4b35 (released December 21, 2003)
234 =======================================================================
235
236 1. Added a note in the dump man page saying that the default
237 blocksize can be 32 if -d is used with a high density argument.
238 Thanks to Antonios Christofides <A.Christofides@itia.ntua.gr>
239 for the patch.
240
241 2. Fixed configure to correctly understand CPPFLAGS, CFLAGS,
242 LDFLAGS environment variables. Thanks to Arcady Genkin
243 <antipode@users.sourceforge.net> for reporting the bug.
244
245 3. Made -e/-E options of dump accept an unlimited number of inodes
246 to be excluded, rather than a hardcoded maximum. Thanks to
247 Dietrich Rothe <d-rothe@users.sourceforge.net> for the patch.
248
249 4. Updated the autoconf system to 2.50. Forced the -D_BSD_SOURCE
250 and -D_USE_BSD_SIGNAL defines in configure in order to solve
251 64bit build problems because quad_t is redifined with a
252 different signature. Thanks to Mike Harris <mharris@redhat.com>
253 for reporting this bug.
254
255 5. Made restore build on Solaris, making possible to
256 restore Linux's "enhanced" tapes. Thanks to Uwe Gohlke
257 <uwe@ugsoft.de> for the patch.
258
259 6. Made an extension in the dump tape format capable of saving
260 MacOSX specific inode extensions. Uwe Gohlke <uwe@ugsoft.de>
261 wrote the extension and contributed the restore code back
262 into this codebase. The same extension mechanism will be
263 used in the future to save ACLs...
264
265 7. Made rmt work correctly with regard to QFA and local/remote
266 files and tapes. The remote access will however work only
267 when the dump provided rmt version is used. If you want to
268 use another rmt server, please do not use the QFA feature.
269 Thanks to Uwe Gohlke <uwe@ugsoft.de> for the patch.
270
271 Changes between versions 0.4b33 and 0.4b34 (released April 18, 2003)
272 ====================================================================
273
274 1. Fixed the INSTALL file to reflect the actual install paths.
275 Thanks to David Ronis <ronis@ronispc.chem.mcgill.ca> for
276 reporting the bug.
277
278 2. Fixed the configure script to only check for headers presence
279 instead of trying to compile them. This should fix issues
280 with old build environments. Thanks to Kari Lempiainen
281 <kari@funky.pp.fi> for reporting the bug.
282
283 3. Fixed restore to correctly ignore sockets when comparing
284 a dump (as socket cannot be properly restored anyway). Thanks
285 to Gunther Reiszig <gunther@mit.edu> for reporting the bug.
286
287 4. Fixed restore to correctly access the archive file (-A argument)
288 even when using a remote tape. Thanks to Fabrice Bellet
289 <fabrice@bellet.info> for reporting the bug.
290
291 5. Fixed (again) handling of long (largefile) seeks in rmt.
292 Thanks to Fabrice Bellet <fabrice@bellet.info> for reporting
293 the bug.
294
295 6. Fixed restore corner case when dealing with large block sizes
296 dump is able to write now (-b 1024). Thanks to Fabrice Bellet
297 <fabrice@bellet.info> for reporting the bug.
298
299 7. Fixed a bug preventing dump to access a filesystem having
300 a label exactly 16 bytes in length. Thanks to <minduni@ti-edu.ch>
301 for reporting the bug.
302
303 8. Made dump store dump dates using explicit timezones, fixing a
304 problem with incremental dumps if the timezone is modified
305 between the dumps. Thanks to Philippe Troin <phil@fifi.org> for
306 the bug report and the patch.
307
308 9. Fixed a bug encountered when dumping individual files (not full
309 filesystems or directories) and dangling symbolic links happen
310 to be in the list of files. For as far as dump is concerned,
311 dangling symbolic links are allowed, and are dumped as is.
312 Thanks to Jin-su Ahn <jsahn@ee.snu.ac.kr> for reporting the
313 bug and providing the fix.
314
315 10. Fixed open and creation modes and permissions for QFA and
316 table-of-contents files in dump and restore. Thanks to
317 Philippe Troin <phil@fifi.org> for the patch.
318
319 11. Fixed the archive file descriptor handling enabling it to be 0.
320 This can happen in some cases when shell redirections are used.
321 Thanks to Philippe Troin <phil@fifi.org> for the patch.
322
323 12. Delayed the opening of archive file until after suid had been
324 dropped (fixing a possible security issue if dump is suid).
325 Thanks to Philippe Troin <phil@fifi.org> for the patch.
326
327 13. Fixed the 'S' command handling in the rmt client part.
328 Thanks to Philippe Troin <phil@fifi.org> for the patch.
329
330 14. Modified the end-of-tape script handling to print out statistics
331 (and stop the timer) before launching the eot script. Also, the eot
332 script does not get run anymore when using -M (which makes sense) or
333 when multiple tapes are listed on the command line
334 (-f tape0,tape1,tapen) (which also makes sense).
335 Thanks to Philippe Troin <phil@fifi.org> for the patch.
336
337 15. Relicensed dump/restore under the 'revised' BSD license, as per
338 ftp://ftp.cs.berkeley.edu/ucb/4bsd/README.Impt.License.Change.
339
340 16. Added LZO compression to dump. This new compression method has
341 the advantage of being super fast, thus not killing tape streaming
342 on slow machines. Thanks to Helmut Jarausch
343 <jarausch@igpm.rwth-aachen.de> for the patch and to
344 Markus Oberhumer <markus@oberhumer.com> for giving special permission
345 to include his miniLZO project (GPL licensed) in dump/restore.
346
347 17. Some small buffer overruns fixes in rmt. Thanks to Antonomasia
348 <ant@notatla.demon.co.uk> for reporting the bugs.
349
350 18. Added a special rmt version which can do encryption when writing
351 to tape. Read examples/encrypted_rmt/README for details on
352 how to enable and configure it. Thanks to Ken Lalonde
353 <ken@globalremit.com> for the patch.
354
355 19. Made dump work with 2.5 kernel end of tape early warning semantics.
356 Thanks to Kai Makisara <Kai.Makisara@kolumbus.fi> for the patch.
357
358 20. Fixed a bug which caused dump -w|-W not to work anymore, because
359 the fs_freq and fs_passno fields in /etc/mtab are always set
360 to 0 0. Thanks to Trent Piepho <xyzzy@speakeasy.org> for
361 reporting the bug.
362
363 Changes between versions 0.4b32 and 0.4b33 (released February 10, 2003)
364 =======================================================================
365
366 1. Added a note in the restore man page clarifying the question
367 'set the permissions on the current directory ?' asked by
368 restore at the end of treatment in -i and -x modes.
369
370 2. Fixed the endianess issues when reading compressed tapes.
371 Thanks to Dark Force <daq4th@users.sourceforge.net> for
372 reporting this bug and providing test cases.
373
374 3. Fixed the "ACL won't be dumped" warning message (which showed
375 an extra, unrelated error message). Thanks to Dragan Krnic
376 <dkrnic@lycos.com> for reporting this bug.
377
378 4. Made dump look first into /etc/mtab, then into /etc/fstab
379 when searching for filesystem paths. Also fixed some problems
380 caused by binding mounts or multiple block device mounts.
381 Thanks to Matus Uhlar <uhlar@fantomas.sk>, Elliott Mitchell
382 <ehem@m5p.com>, Greg Edwards <gedwards@users.sourceforge.net>,
383 Brian Hoy <brian.hoy@opus.co.nz>. (fixes Debian bugs #147086
384 and #160305, Sourceforge bugs #618699 and #679832).
385
386 5. Made dump's -I option accept the value '0' meaning all the
387 read errors will be ignored. This can be useful when running
388 dump from unattended sessions (like cron jobs). Thanks to
389 John I Wang <jiwang@users.sourceforge.net> for the suggestion.
390
391 6. Fixed the output of dump to indicate 'blocks' instead of
392 'tape blocks' in the various messages (blocks are always
393 1 Kilobyte, tape blocks are 1 BK * '-b' argument), and
394 made it clearly print the current blocksize at the start of
395 a dump. Thanks to Michal Szymanski <msz@astrouw.edu.pl> for
396 the suggestions.
397
398 7. Made rmt understand long (largefiles) seeks.
399
400 8. Fixed build with very old versions of libext2fs, where
401 EXT2_FT_* constants were undefined.
402
403 9. Made dump accept the dumpdates path on the command line
404 (-D file option) instead of using only the hardcoded one.
405 Thanks to Piete Brooks <pb22@users.sourceforge.net> for the
406 suggestion.
407
408 10. Enabled rmt, LFS, readline, QFA options by default in
409 ./configure. Updated the configure process (new versions
410 of config.guess, config.sub etc).
411
412 Changes between versions 0.4b31 and 0.4b32 (released November 15, 2002)
413 =======================================================================
414
415 1. Changed dump to use fcntl(F_SETLK) style locking instead
416 of flock() when locking the dumpdates file. With the old
417 locking scheme, a local user having read rights on the
418 dumpdates file could be able to do a Denial of Service attack
419 on dump. In order to lock the dumpdates file with the new
420 scheme, the user would need to have write access on the file.
421 Thanks to Richard Johnson <Richard.Johnson3@ey.com> for
422 reporting the bug (originally a bugtraq post).
423
424 2. Fixed interactive 'ls' which caused spurious errors warnings
425 about 'undefined filetypes' detected. Thanks to Jorgen Ostling
426 <jorgen_ostling@users.sourceforge.net> for reporting this
427 bug.
428
429 3. Fixed dump's estimate when dealing with sparse inodes.
430
431 4. Modified dump to allow setting a blocksize bigger than 32kB
432 (raised the limit to 1024kB), because newer hardware needs
433 this for proper operation. Thanks to Dirk Traenapp
434 <dtraenapp@users.sourceforge.net> for reporting this.
435
436 5. Fixed a bug causing Dump to stop and report an error if an
437 inode number in the exclude file was followed by some amount
438 of whitespace. Thanks to Jeffrey Sofferin
439 <sofferin@users.sourceforge.net> for reporting this bug.
440
441 6. Fixed a bug which caused restore, in some particular cases,
442 to ask some 'scary' questions and leave a bunch of RSTTMP
443 directories behind when restoring incremental tapes. Thanks
444 to Philippe Troin <phil@fifi.org> for reporting this bug and
445 providing the test cases.
446
447 7. Changed the wording when inodes are excluded from dump:
448 replaced 'Added inode 7 to exclude list' with
449 'Excluding inode 7 (resize inode) from dump', as suggested
450 by Elliott Mitchell <ehem@m5p.com> in a Debian bug report.
451
452 Changes between versions 0.4b30 and 0.4b31 (released July 30, 2002)
453 ===================================================================
454
455 1. Fixed rmt open flags transmission (GNU's symbolic syntax over
456 rmt) which I broke in 0.4b29. Thanks to Eros Albertazzi
457 <eros@lamel.bo.cnr.it> for reporting the bug.
458
459 Changes between versions 0.4b29 and 0.4b30 (released July 25, 2002)
460 ===================================================================
461
462 1. Made dump print out the ext2 logical block number in case of a read
463 error, which can be given as an argument to debugfs' ncheck command,
464 in order to find out the affected inode. Added note about this
465 usage in dump's man page.
466
467 2. Fixed a problem in restore when reading tapes written on big
468 endian machines with very old versions of dump. The patch was
469 contributed by George Helffrich <george@geology.bristol.ac.uk>.
470
471 3. Fixed the tape length calculation when using large tapes
472 and compression. Thanks to Georg Lippold
473 <g_lippold@sourceforge.net> for reporting the bug.
474
475 4. Added a new set of examples in dump_on_cd_2 directory, based
476 on dump_on_cd examples but somewhat enhanced, supporting DVD
477 media, and localized in english and german. Thanks to
478 Georg Lippold <g_lippold@sourceforge.net> for the new scripts.
479
480 5. Made dump save 32 bit UID/GID. Internally, this was achieved
481 by switching from the old BSD inode format to the new BSD
482 inode format, which means that the tape format was changed.
483 However, since all restore versions out there should
484 transparently support both inode formats, the change should
485 have no side effects. Thanks to John Yu <jky@cs.bu.edu> for
486 reporting the bug.
487
488 6. Fixed a lot of warnings in the code shown when compiling
489 with 'gcc -W'. Thanks to Matthias Andree
490 <matthias.andree@stud.uni-dortmund.de> for reporting this.
491
492 7. Fixed a small markup bug in the dump man page. Thanks to
493 Eric S. Raymond <esr@minx.thyrsus.com> for submitting the
494 patch.
495
496 8. Rewrote entirely the man pages using the tmac.an macro
497 package (Linux man page format) instead of the original BSD
498 format. They should be now cleaner and easier to modify.
499
500 Changes between versions 0.4b28 and 0.4b29 (released June 8, 2002)
501 ==================================================================
502
503 1. Fixed a problem in the rmt ioctl command, where ioctl's issued from
504 non Linux clients were misinterpreted. The description of the problem
505 (incompatible numbering in Linux mtio opcodes) is documented at
506 ftp://ftp.fokus.gmd.de/pub/unix/star/README.mtio . Thanks to
507 Jörg Schilling <schilling@fokus.gmd.de> for reporting this bug and
508 providing an excellent, cross-platform replacement for rmt in his
509 star package.
510
511 2. Fixed a bug reported by Andy Mentges <amentges@jumpline.com>
512 which caused restore to fail when the inode maps were not entirely
513 contained on the first volume (when using really small volumes or
514 when dumping a huge number of inodes).
515
516 3. Fixed a problem in dump, where files in subdirectories of directories
517 that have the nodump flag set (or being excluded with -e/-E) were
518 sometimes incorrectly dumped (instead of being excluded). The original
519 fix is from David C Lawrence <tale@dd.org> in the FreeBSD version
520 of dump: http://www.FreeBSD.org/cgi/query-pr.cgi?pr=32414
521 Thanks to Ted Grzesik <tedgyz@roostme.com> for reporting the bug and
522 help testing the patch.
523
524 4. Added some example scripts from Gerd Bavendiek <bav@epost.de>
525 which makes one able to pipe the output of dump, by the net, to
526 a remote CD-burner server.
527
528 5. Made dump use O_CREAT|O_TRUNC both locally and remotely (over rmt),
529 and use GNU's symbolic syntax over rmt instead of numerical values
530 to assure multiple platform compatibility.
531
532 6. Documented the -d option in restore.
533
534 7. Added a -v (verbose) mode to dump. For now it just prints the number
535 of the inode being dumped, but this could evolve in future versions
536 to include interesting debugging output.
537
538 8. Added a -o flag to restore, which automatically restores the current
539 directory permissions in -i or -x mode, without asking the operator
540 whether to do so. Patch submitted by Tony Clayton <tonyc@e-smith.com>
541 and Peter Samuel <peters@e-smith.com>.
542
543 Changes between versions 0.4b27 and 0.4b28 (released April 12, 2002)
544 ====================================================================
545
546 1. Fixed a bug in the QFA feature which could have caused
547 a bad tape positionning on restore (causing extra delays in
548 restore). Credits go to Uwe Gohlke <uwe@ugsoft.de>.
549
550 2. Added a small note in the dump man page specifying that
551 there should be no white space between the option letter and
552 the -j or -z parameter, thanks to Kenneth Porter <shiva@well.com>
553
554 3. Made restore work with older versions of the readline library,
555 by checking in configure for several extended symbols. Restore
556 can now be compiled with a readline as old as the 2.0 release
557 (though it may be a good idea to upgrade it to a more recent
558 version...). Thanks to Andrew Donkin <ard@waikato.ac.nz> for
559 reporting the build failures.
560
561 4. Fixed a performance problem with the QFA file creation in
562 dump, which made unnecessary seeks on the tape slowing down
563 the dump. Thanks to Andrew Donkin <ard@waikato.ac.nz> for
564 reporting this issue.
565
566 5. Removed the inclusion of some kernel headers in the dump
567 source, which prevented the compile in some kernel/glibc
568 headers/architecture combination. Thanks to Bdale Garbee
569 <bdale@gag.com> for reporting the bug.
570
571 6. Added the appropriate error message when dump fails to
572 open the output file for writing. Thanks to Amith Varghese
573 <amithv@yahoo.com> for reporting this bug.
574
575 7. Made restore able to understand large Solaris ufsdump tapes
576 (containing inodes bigger than 4194304). Sun have introduced
577 an "extension" to the dump tape format when dealing with
578 those inodes, which was uncorrectly handled by Linux restore.
579 Thanks to Uwe Gohlke <uwe@ugsoft.de> for reporting the bug and
580 providing a test case.
581
582 8. Added the -m parameter to dump which optimises the output for
583 inodes having been changed but not modified since the last dump
584 ('changed' and 'modified' have the meaning defined in stat(2)).
585 For those inodes, dump will save only the metadata, instead of
586 saving the entire inode contents. Inodes which are either
587 directories or have been modified since the last dump are saved
588 in a regular way. Uses of this flag must be consistent, meaning
589 that either every dump in an incremental dump set have the flag,
590 or no one has it.
591
592 Changes between versions 0.4b26 and 0.4b27 (released February 15, 2002)
593 =======================================================================
594
595 1. Fixed behaviour of dump when exceeding resource limits
596 (SIGXFSZ treatment).
597
598 2. Added the -L flag to restore to allow the user to specify a
599 maximal allowed number of miscompares when using restore
600 with the -C option to check the backup.
601
602 3. Detailed the manual entry for the -N option of restore.
603
604 4. Added the -a flag to restore to make able doing unattended
605 restores in -i or -x mode (automatically walks through the
606 multiple dump volumes).
607
608 5. Extended the QFA mode to work with local files and/or
609 remote tapes and files. This way, restore can know in advance
610 the tape number and the offset for the inodes to extract and
611 can minimize the extraction time by seeking directly to the
612 good tape/offset.
613
614 6. Added the -A <archive> option to both dump and restore,
615 which makes dump to archive a dump table-of-contents in
616 the specified file to be used by restore to determine
617 whether a file is in the dump file that is being restored.
618 (the archive file syntax is also compatible with the
619 Solaris ufsdump generated one).
620
621 7. Small fix in restore making it able to read some (broken ?)
622 Solaris ufsdump tapes.
623
624 8. Fixed dump to correctly recognise the root filesystem when using
625 ext2 disk labels (LABEL=/). Thanks to John Yu <jky@it.bu.edu>
626 for reporting this bug.
627
628 9. Added the -P <file> option to restore to create a
629 Quick File Access file from an already made dump. Patch
630 contributed by Uwe Gohlke <uwe@ugsoft.de>.
631
632 10. Made restore compile and run on Solaris, making it a
633 possible replacement for the standard ufsrestore. Port was
634 contributed by Uwe Gohlke <uwe@ugsoft.de>.
635
636 Changes between versions 0.4b25 and 0.4b26 (released January 7, 2002)
637 =====================================================================
638
639 1. Added a set of backup scripts from Eugenio Diaz
640 <getnito@yahoo.com> in the examples section. It features
641 automatic (cron based) full and incremental dumping of
642 several filesystems on a separate filesystem.
643
644 2. Fixed a off-by-one miscalculation which disabled dumping
645 a one letter subdirectory. Thanks to Chris Siebenmann
646 <cks@utcc.utoronto.ca> for reporting the bug.
647
648 3. Fixed several restore bugs occuring when trying to
649 restore some missing files on the tape. Thanks to Chris
650 Siebenmann <cks@utcc.utoronto.ca> for reporting the bug.
651
652 4. Fixed --with-ldopts configure argument passing, installing from
653 a separate object directory, makefile cleanups contributed
654 by <splite@purdue.edu>.
655
656 5. Fix a bug which could caused, in some conditions, the highest
657 number inode of a filesystem, to not be dumped. Many thanks
658 to Chris Siebenmann <cks@utcc.utoronto.ca> for helping me
659 chase this bug.
660
661 Changes between versions 0.4b24 and 0.4b25 (released November 17, 2001)
662 =======================================================================
663
664 1. Added a mini howto from Patrick Walsh in the examples
665 directory.
666
667 2. Minor man pages syntax corrections. Thanks to
668 Chris Danis <screechco@home.com> for reporting the bugs.
669
670 3. Added a script from David B. Peterson <dave@toppledwagon.com>
671 to the examples section. It features dumping several
672 filesystems to a remote tape drive upon ssh.
673
674 4. Added a patch provided by Richard Jones <rich@annexia.org>
675 which allows BRADEMAX (number of read errors tolerated by
676 dump) to be adjusted using the -I option.
677
678 5. Fixed a bug which disabled doing "restore -C -f -". Thanks
679 to Clemens Stiglechner <a9401816@unet.univie.ac.at> for the
680 patch.
681
682 6. Add the -l option to restore to specify if, when doing a
683 remote restore, the file used to access the backup is a
684 regular file (the defaults being a tape drive). Restore needs
685 to know this information when reading a remote compressed
686 dump. Previously, this information was autodetected, but
687 the autodetection code fails (with ioctl: Inappropriate ioctl
688 for device) when using a non Linux remote box. Thanks to
689 many users and especially to Eros Albertazzi
690 <eros@lamel.bo.cnr.it> for reporting this.
691
692 7. Found a workaroung for the dump deadlock problem (3 childs
693 stuck in pause(), father in read()). The workaround seems
694 to work for me and several beta-testers. If it doesn't work
695 for you, please report back.
696
697 8. Updated the RPM spec file (BuildPrereq, URL etc).
698
699 Changes between versions 0.4b23 and 0.4b24 (released September 12, 2001)
700 ========================================================================
701
702 1. Fixed the permissions of a newly created QFA file by dump.
703
704 2. Cleaned up the source of dump (the external variables
705 definition was a complete mess, making possible to have
706 objects overlap).
707
708 3. Fixed restore to use the full tape volume path when doing
709 a compare (since it changes the working directory to the
710 filesystem being compared in the process).
711
712 4. Added the -q option to dump which makes dump abort
713 whenever it needs operator attention. It should be
714 easier to use dump in scripts with this option.
715
716 5. Detect the use of incompatible options to dump and
717 refuse them (like -a and -B options together).
718
719 6. Added bzip2 compression to dump/restore (use option -j level
720 to select it). Note that you will need a restore version
721 >= 0.4b24 in order to restore a bzip2 compressed dump.
722 The same warning as for the zlib compression applies:
723 the tape format of a bzip2 dump is not compatible with the
724 original BSD tape format.
725
726 7. Fixed a overflow problem in dump corrupting the dump when
727 very large files were encountered. Thanks to Vladimir Nadvornik
728 for the bug report.
729
730 8. Added a ioctl(BLKFLSBUF) in dump which should flush the
731 kernel buffer/page cache before starting the dump, helping
732 a bit those who use dump on mounted filesystems. Thanks to
733 John Yu <jky@it.bu.edu> and to Theodore T'so <tytso@mit.edu>
734 for this suggestion.
735
736 9. Updated the RPM spec file following the RedHat changes
737 (dynamically linked binaries now in /usr/sbin etc).
738
739 10. Added a patch from Helmut Jarausch <jarausch@igpm.rwth-aachen.de>
740 which enables restore to recognise multi volume compressed dumps
741 done on CD. Included his scripts for dump (which pipe the dump
742 output directly into cdrecord) and restore. There is now
743 possible to to backups to CD on the fly!
744
745 Changes between versions 0.4b22 and 0.4b23 (released July 20, 2001)
746 ===================================================================
747
748 1. Fixed a buffer overflow in restore/tape.c. Patch provided by
749 Marcus Meissner (Caldera International Security Dept.).
750
751 2. Implement the Sun rmt extended protocol. Patch submitted
752 by Ian Gordon <iangordon@users.sourceforge.net>.
753
754 3. Check for the e2fsprogs header <ext2fs/ext2_fs.h> instead of
755 the linux kernel header. This ensures that dump always has the
756 latest version of this file. Patch submitted by
757 Andreas Dilger <adilger@turbolinux.com>.
758
759 4. Report any filesystem present in either /etc/fstab with a
760 positive passno or /etc/dumpdates in dump -w output.
761 Patch submitted by Andreas Dilger <adilger@turbolinux.com>.
762
763 5. Fixed the looping problem in dump introduced in the
764 previous version.
765
766 6. Changed the -B option of dump to limit the size of
767 _compressed_ output per volume if compression is on.
768 Patch contributed by Helmut Jarausch
769 <jarausch@igpm.rwth-aachen.de>. Note however that, since
770 it is impossible to predict the size of the compressed
771 data before doing the compression, the -B limit is a bit
772 conservative.
773
774 7. Fixed a bug in reading the operator typed file/tape path for
775 the next volume in restore.
776
777 8. Implemented a "-F script" option to restore which permits the
778 user to specify a script which will be launched at the
779 beginning of each tape, useful for automatic programming of
780 tape changers for example. See the restore man page for the
781 script parameters and return codes.
782
783 9. Small fix for the QFA routines provided by Uwe Gohlke
784 <uwe@ugsoft.de>, and some recommendations for QFA uses in
785 the man pages.
786
787 10. Fixed the multivolume restoring where making a mistake
788 to the 'Mount next tape' prompt caused several blocks to
789 be lost.
790
791 11. Enhanced the -e option of dump to take as a parameter a
792 comma separated list of inode numbers.
793
794 12. Added the -E option to dump which specify a file containing
795 inode numbers to exclude from the dump.
796
797 13. Fixed the compressed multi-volume dump + restore.
798
799 Changes between versions 0.4b21 and 0.4b22 (released May 12, 2001)
800 ==================================================================
801
802 1. Made dump report the number of blocks dumped per volume.
803 Thanks to Kenneth Porter <shiva@well.com> for the suggestion.
804
805 2. Fix a bug in dump where waiting too much at the 'change volumes'
806 question caused the volume to be rewritten. Thanks to
807 Graham Reed <greed@users.sourceforge.net> for reporting the
808 bug and providing a patch.
809
810 3. Added a compression option to dump, contributed by Jerry
811 Peters <gapeters@worldnet.att.net>.
812
813 WARNING: the compression works only on files/pipes and
814 drives supporting variable block size.
815
816 WARNING2: tapes created using the compression feature are
817 incompatible with the standard dump tape format, and a
818 version of restore >= 0.4b22 will be needed for extraction.
819
820 4. Fixed some compilation problems for glibc 2.2.2 and 64 bit
821 architectures. Thanks to Paul Slootman <paul@debian.org> for
822 the patch and to Bdale Garbee <bdale@gag.com> for forwarding
823 it upstream.
824
825 5. Many cleanups (CPP defines, const char warnings, check of
826 ext2fs COMPAT flags, time_t cleanups, added libext2 version
827 in dump usage text) by Andreas Dilger <adilger@turbolinux.com>.
828
829 6. Made --prefix option work in configure. All the install path
830 are now based on the configure parameters.
831
832 7. Added the Quick File Access mode in dump/restore, contributed
833 by Uwe Gohlke <uwe@ugsoft.de>. In this mode, dump stores in
834 a file tape position for each inode, and this file is used by
835 restore (if called with parameter Q and the filename)
836 to directly position the tape at the file restore is currently
837 working on. This saves hours when restoring single files from
838 large backups, saves the tapes and the drive's head. Use
839 --enable-qfa option of configure to compile in the QFA support.
840
841 8. Added the possibility to dump several files and directories
842 in a single invocation of dump. Thanks to Uwe Gohlke
843 <uwe@ugsoft.de> for implementing this option.
844
845 9. Fixed the dumping and restoring of files with holes
846 (files ending with holes didn't get dumped or restored
847 correctly).
848
849 10. Fixed a socket descriptor leak in dump, which leaved opened
850 3 file descriptors per dump process (and there is one dump
851 process per tape).
852
853 11. Fixed dump large file system support, by explicit use of
854 open64/lseek64/etc functions (explicit use needed because
855 e2fsprogs libraries don't behave well when compiled with
856 FILE_OFFSET_BITS=64).
857
858 Changes between versions 0.4b20 and 0.4b21 (released January 13, 2001)
859 ======================================================================
860
861 1. Fixed some bugs in the dump -w|-W logic introduced by
862 the previous version. Thanks to Andreas Dilger
863 <adilger@turbolinux.com> for his help on this one.
864
865 2. Fixed again a compilation problem when using old e2fs
866 headers (filesystem label related). Thanks to many users
867 who reported this stupid error.
868
869 3. Fixed a build problem on old lib5 based systems dealing with
870 _PATH_MNTTAB being undefined. Thanks to John Adams
871 <johna@onevista.com> for reporting the bug.
872
873 4. Improved the error detection in restore when creating
874 the temporary files in TMPDIR. Now restore will corectly
875 report a 'no space left on device' error instead of
876 strange errors which could imply an archive corruption.
877 Thanks to Gabriel Somlo <somlo@cs.colostate.edu> and
878 bgjenero <bgjenero@sympatico.ca> for reporting the bug.
879
880 5. Added the throughput information in dump's progression
881 messages. Thanks to Andreas Dilger <adilger@turbolinux.com>
882 for the patch.
883
884 6. Use libext2fs's inode scanning routines, which are particularly
885 robust when dealing with errors (especially when having some
886 bad blocks on the filesystem). Many thanks to Colin
887 <colin@horizon.com> for reporting the bug, investigating
888 the issues, finding a workaround, writing the patch and
889 fully testing it... (of course, if this patch will break
890 anything, he is to blame for :-)).
891
892 7. Made dump and restore LFS aware. Dump can dump a filesystem
893 containing large files, generate a large file on output and
894 restore can restore them. This feature is not enabled by
895 default, use --enable-largefile option of configure to enable
896 it (you need to have a LFS aware glibc though). Thanks to
897 Andrea Arcangeli <andrea@suse.de> for submitting the patch,
898 and to Theodore T'so <tytso@valinux.com> for his always
899 useful thoughts.
900
901 8. Made dump ask upon a tape write error if it should rewrite
902 the current volume (assume this is a bad tape) or if it should
903 assume an end-of-tape condition (useful for tape drives which
904 behaves badly at the end of the tape). Thanks to Andreas
905 Dilger <adilger@turbolinux.com> for the suggestion.
906
907 Changes between versions 0.4b19 and 0.4b20 (released November 10, 2000)
908 =======================================================================
909
910 1. Fixed a small compilation problem due to a change
911 in the definintion of the struct sigaction in
912 glibc 2.0 / libc5. Thanks to Gunther Schlegel
913 <schlegel@riege.de> for reporting the bug and to
914 Dave Platt <dplatt@snulbug.mtview.ca.us> for suggesting
915 a fix.
916
917 2. Modified the label and uuid analysis in order to be
918 self-contained (independant of kernel/libc headers). This
919 should fix the compile with older kernel/libc headers and
920 will preserve the functionality. Thanks to Bernhard Erdmann
921 <bernhard.erdmann@gmx.de> for reporting the bug.
922
923 3. The 'exclude inode' option, if applied to a directory
924 inode, excludes now correctly all the leaves of this
925 directory. Thanks to John R. Dennison
926 <gerdesas@users.sourceforge.net> for reporting the bug.
927
928 4. Fixed the '-e' option to disable the possibility
929 to exclude the root inode (which causes the dumps to
930 be unreadable by restore). Prevented array overflow
931 when multiple -e options are used.
932
933 5. Fixed dump to correctly interpret a filesystem argument
934 which is a mountpoint and it is not an absolute path
935 (as specified in the fstab). Thanks to Bernhard R. Erdmann
936 <be@berdmann.de> for reporting the bug.
937
938 6. Made dump able to backup files larger than 2 GB. Note that
939 dump still doesn't cope with files larger than 4 GB.
940
941 7. Restore the real uid and gid before invoking an external
942 RSH program (big hole when dump or restore is suid root!).
943
944 8. Get the values for the default tape device and the location
945 of fstab file from the system headers. Thanks to
946 Andreas Dilger <adilger@turbolinux.com> for the patch.
947
948 9. Made dump -w|-W report all recognized filesystems
949 present in either /etc/fstab or /etc/dumpdates, and present
950 the list in the same order as in fstab file. Thanks
951 to Andreas Dilger <adilger@turbolinux.com> for the patch.
952
953 10. Made dump's -a (automatic end of tape detection) the
954 default. Specifying one of -B, -c, -d or -s options will
955 override the EOM detection. Thanks to Andreas Dilger
956 <adilger@turbolinux.com> for the patch.
957
958 11. Save the ext2 filesystem label into the dump volume label.
959 Specifying a volume label on the command line (-L option)
960 will override this feature. Thanks to Andreas Dilger
961 <adilger@turbolinux.com> for the patch.
962
963 Changes between versions 0.4b18 and 0.4b19 (released August 20, 2000)
964 =====================================================================
965
966 1. Fixed the signal handling in dump (which I broke in 0.4b17)
967 which was causing several strange dump failures (dump
968 hanged or segmentation faults etc).
969
970 2. Specified the default blocksize in the dump man page.
971
972 3. Changed two info messages of restore to be written on stdout
973 instead of stderr in order to leave stderr only for errors
974 or warnings. Thanks to Stephen Atwell
975 <satwell@urbana.css.mot.com> for the suggestion.
976
977 4. Corrected an off by one calculation which prevented
978 dumping a directory having a 1 character name length.
979 Thanks to Bernhard Erdmann <bernhard.erdmann@gmx.de>
980 for reporting the bug.
981
982 5. Reinforce dump in order to allow only level 0 and no
983 -u option when dumping a subdirectory, and document
984 this in the man page. Thanks to Bernhard Erdmann
985 <bernhard.erdmann@gmx.de> for reporting the bug.
986
987 6. Fixed a small harmless bug in restore which caused
988 in some conditions hard links to be restored several
989 times (therefore generation some warning messages).
990 Thanks to Kenneth Porter <shiva@well.com> for
991 reporting the bug.
992
993 7. Updated the RPM spec file to the latest RedHat version,
994 providing FHS packaging and other cosmetic changes.
995 You will need RPM version >= 3.0.5 in order to build the RPM.
996
997 8. Updated the configure script to check for libtermcap
998 before checking for libreadline, because we need this
999 library in order to compile the readline support.
1000
1001 9. Made dump understand the LABEL= and UUID= notation
1002 both in /etc/fstab and on the command line. Note that
1003 you will need the /proc filesystem in order to use
1004 these notations. Thanks to Erik Troan <ewt@redhat.com>
1005 for providing the patch.
1006
1007 Changes between versions 0.4b17 and 0.4b18 (released June 30, 2000)
1008 ===================================================================
1009
1010 1. Fixed a potential buffer overflow in restore. Thanks
1011 to Stan Bubrouski <satan@fastdial.net> for reporting
1012 the bug.
1013
1014 2. Fixed a readline-related bug which prevented
1015 'cat DUMPFILE | restore -i -f -' from working. Thanks
1016 to Charles Karney <karney@users.sourceforge.net>
1017 for the bug report.
1018
1019 3. Changed a few "panic" into "exit", causing restore to
1020 be more stable against some attacks (like the last one
1021 reported on Bugtraq, although the last version of restore
1022 was not vulnerable - just dumped core). Thanks to
1023 Andreas Hasenack <andreas@conectiva.com.br> for reporting
1024 the bugs to me.
1025
1026 4. Removed the suid-root bit on dump and restore in the
1027 default build (and generated RPMs). It should be safer
1028 now. Users who need the suid-root capabilities in order
1029 to make network backups should read first the man page
1030 of dump and enable the suid bit manually.
1031
1032 5. Added -ltermcap to the compile parameters for restore
1033 when using readline support, in order to make the compile
1034 process work on some systems (SuSE ?). Thanks to
1035 Patrik Schindler <poc@pocnet.net> for reporting the bug.
1036
1037 Changes between versions 0.4b16 and 0.4b17 (released June 1st, 2000)
1038 ====================================================================
1039
1040 1. The -F script is called now *only* at the end of a tape,
1041 not at the end of the dump. Thanks to Christian Haul
1042 <haul@informatik.tu-darmstadt.de> for the bug report.
1043
1044 Normally, the device name and the current volume number
1045 are passed on the command line of the script. However,
1046 if you want the old style script (with no arguments
1047 passed on the command line), you can enable it in
1048 configure with the --enable-oldstylefscript.
1049
1050 2. Use posix signal handling to preserve dump functionality
1051 with libc5. Thanks to Jeff Johnson <jbj@redhat.com> for
1052 the patch.
1053
1054 3. Made the exit code of restore in case of a 'C'ompare
1055 command reflect the result of the compare. An exit status
1056 of 0 means the dump archive is correct, 1 means tape errors,
1057 2 means that some files were modified. Thanks to Kenneth Porter
1058 <shiva@well.com> for the suggestion.
1059
1060 4. Made (finally) quotes work in interactive restore.
1061
1062 5. Small fixes in order to allow dump to compile with a
1063 really old e2fsprogs version. Thanks to Ian Zimmerman
1064 <itz@speakeasy.org> for the bug report.
1065
1066 6. Add GNU readline capabilities to interactive restore.
1067 Use configure's --enable-readline flag to enable this feature.
1068 Thanks to Patrik Schindler <poc@pocnet.net> for the
1069 suggestion.
1070
1071 7. Do the compare on the fly in restore 'C' mode (this will
1072 allow not to exhaust the available /tmp space when
1073 ccmparing large files). Thanks to Kenneth Porter
1074 <shiva@well.com> for the suggestion.
1075
1076 Changes between versions 0.4b15 and 0.4b16 (released March 11, 2000)
1077 ====================================================================
1078
1079 1. Fixed some several duplicate 'DUMP: DUMP:' in the
1080 output of dump.
1081
1082 2. Corrected the estimation of blocks to dump. Note that
1083 this is still wrong for large directory inodes, due
1084 to the size of a BSD directory entry on the tape
1085 (estimation is based on the size of the ext2 directory,
1086 which is smaller because it doesn't contain the null
1087 character at the end).
1088
1089 3. Made dump report the total number of MB written to
1090 tape at the end of run. Thanks to W. Reilly Cooley
1091 <wcooley@nakedape.cc> for the patch.
1092
1093 4. Added the -X option to restore giving the possibility
1094 to read the names of the files to be extracted/listed
1095 from a text file (in addition of the command line).
1096 Thanks to Dejan Muhamedagic <dejan@quant-x.com> for the
1097 patch.
1098
1099 5. Added the device name and the current volume number
1100 as arguments to the end of tape script (dump -F option).
1101
1102 6. Made the multi-volume dump work again (after having
1103 broken it in 0.4b15).
1104
1105 Changes between versions 0.4b14 and 0.4b15 (released March 2, 2000)
1106 ===================================================================
1107
1108 1. Added a prompt command in interactive restore mode. Thanks
1109 to Andreas Dilger <adilger@home.com> for the patch.
1110
1111 2. Fixed a buffer overflow problem in dump (caused by
1112 not checking the size of the filesystem parameter).
1113 Thanks to Kim Yong-jun <loveyou@hackerslab.org> for
1114 reporting this on Bugtraq (and to several dump users
1115 who forwarded me his mail).
1116
1117 3. Added the '-F script' option to dump in order to
1118 launch a script at the end of each tape (to be used
1119 with a tape changer, or to notify the sysadmin by
1120 pager etc.).
1121
1122 4. Fixed a bug in restore compare code caused by the changes
1123 I made in 0.4b14.
1124
1125 5. Fixed the treatment of options using the old BSD syntax
1126 in both dump and restore.
1127
1128 Changes between versions 0.4b13 and 0.4b14 (released February 10, 2000)
1129 =======================================================================
1130
1131 1. Fixed a bug in dump which may have caused invalid deleted
1132 directories to be dumped out if they were deleted after the
1133 mapping pass. This could occure on active filesystem and lead
1134 to heap corruption (causing dump malfunction in many possible ways).
1135 Thanks to Jan Sanislo <oystr@cs.washington.edu> for finding this
1136 bug and submitting the patch.
1137
1138 2. Fixed the handling of the filesystem parameter in dump. A
1139 '/mnt/disk' parameter caused the disk contents to be dumped,
1140 but a '/mnt/disk/' parameter caused the mountpoint directory
1141 to be dumped (generally an empty directory).
1142
1143 3. Improved the output of dump in order to tell which directory
1144 it is currently dumping (when dumping a subtree).
1145
1146 4. Added the '-e' exclude inode option to dump. Thanks to
1147 Isaac Chuang <ike@isl.stanford.edu> for contributing with the patch.
1148
1149 5. Added a REPORTING-BUGS file in order to provide a guide
1150 on how to correctly report a bug in dump/restore.
1151
1152 6. Fixed a restore problem when restoring a hard link to an inode
1153 having the immutable or append-only attribute set. Thanks to
1154 Ambrose Li <acli@mingpaoxpress.com> for submitting the patch.
1155
1156 7. Fixed a compatibility problem between dumps made on little
1157 endian machines (the format was correct) and big endian
1158 machines (the format was incorrect). This fix break the
1159 compatibility with the older dumps made on big endian
1160 machines (sparc, m86k, ppc etc). For the first time in
1161 linux dump's history (I believe), the dumps made by *BSD,
1162 Linux/alpha, Linux/sparc and Linux/x86 are compatible,
1163 so interchangeable. Thanks to Rob Cermak
1164 <cermak@ahab.rutgers.edu> for submitting the bug and
1165 helping me test the fix.
1166
1167 8. Fixed the way dump reports the remaining percent/time, if
1168 the number of blocks actually dumped exceeds the estimated
1169 number of blocks. Thanks to Jean-Paul van der Jagt
1170 <jeanpaul@dutepp0.et.tudelft.nl> for reporting the bug.
1171
1172 Changes between versions 0.4b12 and 0.4b13 (released January 21, 2000)
1173 ======================================================================
1174
1175 1. Small Y2K fix in the man pages :). Thanks to Bernhard Sadlowski
1176 <sadlowsk@Mathematik.Uni-Bielefeld.DE> for reporting the bug.
1177
1178 2. Removed the requirement to build the RPM as root from the
1179 spec file. Thanks to Christian Weisgerber
1180 <naddy@mips.rhein-neckar.de> for submitting this.
1181
1182 3. Fixed a bug in dump related to the 'filetype' feature of ext2fs,
1183 causing dump to block when dumping really huge filesystems.
1184 Many thanks to Patrik Schindler <poc@pocnet.net> for
1185 helping me find this bug.
1186
1187 4. Fixed the treatment for an interrupt signal when dump access
1188 the remote tape through RSH. Thanks to Christian Weisgerber
1189 <naddy@mips.rhein-neckar.de> for providing the patch.
1190
1191 5. Fixed a bug which was causing dump/restore to display
1192 garbage characters instead of the remote host name.
1193
1194 Changes between versions 0.4b11 and 0.4b12 (released January 8, 2000)
1195 =====================================================================
1196
1197 1. Small fix in the dump man page. Thanks to Thorsten Kukuk
1198 <kukuk@suse.de> for submitting the patch.
1199
1200 2. Fix for the exit code when using the size estimate option of
1201 dump. Thanks to Matti Taskinen <mkt@rni.helsinki.fi> for
1202 submitting the patch.
1203
1204 3. Handle EINTR in atomical reads/writes in dump, which was causing
1205 dump to fail on some systems. Thanks to Eric Jergensen
1206 <eric@dvns.com> for reporting the bug and submitting the patch.
1207
1208 4. Handle more than 16 characters for the device names in dumpdates.
1209 (up to 255 now). Thanks to Rainer Clasen <bj@ncc.cicely.de> for
1210 tracking down the problem and proposing the solution.
1211
1212 5. Fixed a bug in dump which prevented the creation of the
1213 dumpdates file when doing a 0-level dump without already
1214 having a dumpdates file. Thanks to Patrik Schindler
1215 <poc@pocnet.net> for reporting the bug.
1216
1217 6. Changed the way dump 'S' flag reports the size estimate
1218 from number of blocks into bytes (making it compatible
1219 with the Solaris version, and simplifying things for
1220 amanda users). Thanks to Jason L Tibbitts III
1221 <tibbs@math.uh.edu> for reporting the bug.
1222
1223 7. Fixed a compatibility problem in linux/alpha dump tape format.
1224 Now the linux/alpha dump are (again) compatible with the
1225 other dump formats. But this breaks compatibility with
1226 older dumps made on alpha. Thanks to Mike Tibor
1227 <tibor@lib.uaa.alaska.edu> for helping me in finding this bug.
1228
1229 Changes between versions 0.4b10 and 0.4b11 (released December 5, 1999)
1230 ======================================================================
1231
1232 1. Added a '--enable-kerberos' to configure.
1233
1234 2. Added a 'S' option to dump which determines the amount of space
1235 that is needed to perform the dump without actually doing it, similar
1236 to the Sun's ufsdump 'S' option. Patch contributed by Rob Cermak
1237 <cermak@ahab.rutgers.edu>.
1238
1239 3. Added a 'M' multi-volume option to dump and restore which enables
1240 dumping to multiple files (useful when dumping to an ext2
1241 partition to make several dump files in order to bypass the 2GB
1242 file size limitation). The 'f' argument is treated as a prefix and
1243 the output files will be named <prefix>001, <prefix>002 etc. With
1244 the 'M' flag, restore automatically selects the right file without
1245 asking to enter a new tape each time.
1246
1247 4. Fixed a memory leak which was causing dump to grow very big
1248 (270MB when dumping a 10GB filesystem...). Thanks to Jason
1249 Fearon <jasonf@netrider.org.au> for reporting the bug.
1250
1251 Changes between versions 0.4b9 and 0.4b10 (released November 21, 1999)
1252 ======================================================================
1253
1254 1. Make configure test if the system glob routines support
1255 extended syntax (ALTDIRFUNC). If not, use the internal glob
1256 routines instead of system ones. Thanks to Bernhard Sadlowski
1257 <sadlowsk@Mathematik.Uni-Bielefeld.DE> for reporting the bug
1258 and helping me resolve this and other minor libc5 compiling
1259 glitches.
1260
1261 2. Fix a problem when dumping a ext2fs with the 'filetype'
1262 feature enabled. Thanks to Patrick J. LoPresti
1263 <patl@cag.lcs.mit.edu> for reporting the bug and to
1264 Theodore Y. Ts'o <tytso@mit.edu> for providing the patch.
1265
1266 3. Made the nodump flag work on directories. A directory which
1267 has the nodump flag gets never dumped, regardless of its
1268 contents.
1269
1270 4. Integrate a patch from Jeremy Fitzhardinge <jeremy@goop.org>
1271 which allows dump on an active ext3 filesystem. However, this
1272 is a "quick and dirty" patch which enables backup of an ext3
1273 filesystem through the ext2 compatibility (by ignoring the
1274 NEEDS_RECOVERY bit). The journal file is not recognized and
1275 it is dumped (it should not).
1276
1277 5. Test the superblock compatibility flags when dumping, in order
1278 to be sure that we know how to deal with specific features.
1279
1280 Changes between versions 0.4b8 and 0.4b9 (released November 5, 1999)
1281 ====================================================================
1282
1283 1. Use lchown instead of chown, fixing a possible security problem
1284 when restoring symlinks (a malicious user could use this
1285 to deliberately corrupt the ownership of important system files).
1286 Thanks to Chris Siebenmann <cks@utcc.utoronto.ca> for detecting
1287 this and providing the patch.
1288
1289 Changes between versions 0.4b7 and 0.4b8 (released November 3, 1999)
1290 ====================================================================
1291
1292 1. Put dump sources under CVS, added Id tags in all files so
1293 one can use 'ident' on binary files.
1294
1295 2. Added the dump/restore version in the usage text so one can
1296 easily verify the version he is using.
1297
1298 3. Small patch from Nuno Oliveira <nuno@eq.uc.pt> which fixes
1299 a va_start/va_end problem on linux-ppc (always call va_start
1300 va_end in pairs each time we use a vararg function).
1301
1302 4. Added again the DT_* constants because old libc does not
1303 contain them :(. Thanks to Eric Maisonobe <virnet@nat.fr>
1304 for submitting the bug report.
1305
1306 5. Use ext2fs_llseek instead of llseek. With recent e2fsprogs
1307 this should enable dumping big (huge) filesystems.
1308
1309 6. Added the RSH environment variable in order to be able to
1310 use a rsh replacement like ssh when doing remote backups (and
1311 bypass the security limitations of rcmd). Now you can do remote
1312 backups without being root (or making dump setuid root).
1313
1314 7. Modified again the way dumpdates works. For incremental dumps,
1315 we need to read dumpdates even if we are not using 'u' option.
1316 Thanks to Bdale Garbee <bdale@gag.com> for his ideas on how
1317 this should work.
1318
1319 Changes between versions 0.4b6 and 0.4b7 (released October 8, 1999)
1320 ===================================================================
1321
1322 1. Removed the 'k' flag from the restore 'about' text if kerberos
1323 was not compiled in.
1324
1325 2. Prototyped (f)setflags from e2fsprogs and corrected the calls
1326 to them (fsetflags takes a char*, setflags an open fd!).
1327
1328 3. (f)setflags is called only if the flags aren't empty. If the
1329 file is a special file, a warning is printed, because changing
1330 flags implies opening the device. Normally, a special file
1331 should not have any flag... (Debian bug #29775, patch provided
1332 by Abhijit Dasgupta <abhijit@ans.net>).
1333
1334 4. Made possible to dump a file system not mentioned in /etc/fstab.
1335 (Debian bug #11904, patch provided by Eirik Fuller <eirik@netcom.com>).
1336
1337 5. Changed the default behaviour to not create dumpdates
1338 unless 'u' option is specified. Removed the old "debian-patch"
1339 which provided the same thing. (Debian bug #38136, #33818).
1340
1341 6. Removed all those dump*announce, since they were getting old...
1342
1343 7. Added warning messages if dumpdates does not exist and
1344 when an inferior level dump does not exist (except for a level 0
1345 dump).
1346
1347 8. Debugged the glob calls in interactive mode: restore used a
1348 dirent struct which was different from the /usr/include/dirent.h
1349 one (this used to work, is it a glibc2 change?), so none of the
1350 compat glob (which used /usr/include/dirent.h) or the system glob
1351 worked. Restore use now the system dirent (and the system
1352 DT_* constants), which are compatible with BSD ones.
1353
1354 9. Added a configure flag (--with-dumpdatespath) to specify
1355 the location of dumpdates file. By default, it is
1356 /etc/dumpdates.
1357
1358 10. Added the "AUTHOR" and "AVAILABILITY" sections and
1359 included the current date/version in man pages.
1360
1361 11. Corrected the estimation of remaining time when
1362 the operator doesn't change the tapes quickly enough. This
1363 was an old bug, I thought I corrected it, and discovered
1364 that in fact it was corrected in two different places, so
1365 the results canceled each other...
1366
1367 Changes between versions 0.4b5 and 0.4b6 (released October 1, 1999)
1368 ===================================================================
1369
1370 1. Integrated multiple patches from RedHat, Debian and SuSE:
1371
1372 - tweak dump/itime.c to not try to read dumpdates if the 'u' option
1373 isn't specified.
1374 - several fixes in the man pages.
1375 - update the default tape device to /dev/st0.
1376 - many updates for Linux Alpha (byte ordering, size_t etc).
1377 - buffer overruns.
1378 - use environment variable for TMPDIR (instead of /tmp).
1379 - use sigjmp_buf instead of jmp_buf (RedHat bug #3260).
1380 - workaround egcs bug (RedHat bugs #4281 and #2989).
1381 - wire $(OPT) throughout Makefile's.
1382
1383 2. Upgrade the dump revision to 1, making possible to dump filesystems
1384 made with e2fsprogs-1.15 or newer. Nothing seems to break...
1385
1386 3. Fix some compile warnings, prototype all functions.
1387
1388 4. Use glibc err/glob instead of internal compatibility
1389 routines (only if available).
1390
1391 5. Fix a compile error on Linux 2.2.7 / libc5 (5.4.44) (patch provided
1392 by Bernhard Sadlowski <sadlowsk@mathematik.uni-bielefeld.de>).
1393
1394 Changes between versions 0.4b4 and 0.4b5 (released September 22, 1999)
1395 ======================================================================
1396
1397 1. Integrated the changes from FreeBSD-3.1-RELEASE
1398 (mostly bug fixes, buffer overruns, dump has now an "automatic
1399 tape length calculation" flag, dump/restore can use kerberos now
1400 (this is NOT tested), use environment variables for TAPE and
1401 RMT etc.).
1402
1403 2. Integrated three RedHat patches ("glibc", "kernel" and "bread" patches)
1404
1405 3. Corrected a bug in restore when using 'C' option with multi-volumes
1406 tapes (files splited accros two tapes give "size changed" errors
1407 when comparing).
1408
1409 4. Corrected the long standing bug when dumping multiple tapes.
1410 This works for me, needs further testing.
1411
1412 Changes between versions 0.4b3 and 0.4b4 (released January 17, 1997)
1413 ====================================================================
1414
1415 1. Dump now runs correctly on kernels 2.1.x
1416 Fix made by Gerald Peters <gapeters@worldnet.att.net>
1417
1418 Changes between versions 0.4b2 and 0.4b3
1419 ========================================
1420
1421 1. Use realpath() if available
1422
1423 2. Report statistics
1424
1425 Changes between versions 0.4b1 and 0.4b2
1426 ========================================
1427
1428 1. Fixed the bug fix from Greg Lutz (I had made a mistake when integrating
1429 the patch)
1430
1431 2. Fixed restore to make it able to read FreeBSD 2.x dumps again
1432
1433 3. Fixed configure.in to correctly handle --enable-rmt
1434
1435 Changes between versions 0.3 and 0.4b1
1436 ======================================
1437
1438 1. Integrated the changes from 4.4BSD-Lite2
1439
1440 2. Integrated the patches from Debian and Red Hat
1441
1442 3. Portability changes: use the __u32, __u16, __s32, and __s16 types
1443
1444 4. Changed dump to use the Ext2fs library to get block addresses. This
1445 should solve the endianness problem on SparcLinux.
1446
1447 5. Created a configure.in file (shamelessly stolen from the e2fsprogs
1448 distribution's one) to use autoconf
1449
1450 6. Fixed a few minor bugs
1451
1452 Changes between versions 0.2e and 0.2f
1453 ======================================
1454
1455 1. Added the creation of named pipes (fifos) in restore.
1456
1457 2. Added the -N flag in the restore manual page.
1458
1459 3. Added the file kernel.patch which contains the llseek() optimization
1460 patch for 1.2.x kernels.
1461
1462 4. Fixed a bug in the restoration of symbolic links: owner and group were
1463 not restored.
1464
1465 5. Integrated some changes from FreeBSD 2.2.
1466
1467 6. Added a call to ftruncate() after restoring each file to restore
1468 correctly files ending by a hole.
1469
1470 Changes between versions 0.2d and 0.2e
1471 ======================================
1472
1473 1. Fixed a bug in the "set owner/mode" process. Every file was restored
1474 with owner = root (0) and group = root/wheel/whatever (0).
1475
1476 Changes between versions 0.2c and 0.2d
1477 ======================================
1478
1479 1. Dump is now able to backup 2GB+ filesystems.
1480
1481 2. Dump and restore can now be linked as static binaries.
1482
1483 Changes between versions 0.2b and 0.2c
1484 ======================================
1485
1486 1. Fixed a bug when dumping ``slow'' (i.e. normal) symbolic links.
1487
1488 Changes between versions 0.2a and 0.2b
1489 ======================================
1490
1491 1. Really fixed the bug that I should have corrected in 0.2a.
1492
1493 2. Enabled optimization again.
1494
1495 Changes between versions 0.2 and 0.2a
1496 =====================================
1497
1498 1. Disabled the optimization during compilation.
1499
1500 Changes between versions 0.1 and 0.2
1501 ====================================
1502
1503 1. Fixed a bug in fstab.c which caused a null pointer to be stored in
1504 the fs_type field (actually, I modified the file fstab.c to make it
1505 use the mntent functions).
1506
1507 2. Dump and restore now use a 4.3 BSD compatible dump format. Backups
1508 made by dump should be readable by the BSD restore and backups made
1509 by the BSD dump should be readable by restore. Unfortunately, this
1510 means that the dump format has changed between version 0.1 and version
1511 0.2 :-(
1512
1513 3. Dump is now able to backup a subtree, it is no longer limited to whole
1514 filesystems like the BSD version.
1515
1516 4. Dump now uses ext2_llseek() so it is able to backup filesystems bigger
1517 than 2 GB.
1518
1519 Changes between versions 0.0 and 0.1
1520 ====================================
1521
1522 1. Now create links rdump and rrestore during the `make install' step.
1523
1524 2. Linux port specific bugs added to the manual pages
1525
1526 3. Incorrect estimation of the number of tapes blocks fixed when doing
1527 incremental backups.
1528
1529 4. Better ls-like format in restore in interactive mode.