b99fb6b1187137f66365d6827d2afceac503c54d
[dump.git] / CHANGES
1 $Id: CHANGES,v 1.291 2008/04/17 15:29:47 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 8. Force '-a' to be the default value when running 'restore -C'.
37 Thanks to Kenneth Porter <shiva@sewingwitch.com> for the bug
38 report.
39
40 9. Cleanup the output of the 'cron_dump_to_disk' example.
41 Thanks to Aaron S. Hawley <ashawley@users.sourceforge.net> for
42 the patch.
43
44 Changes between versions 0.4b40 and 0.4b41 (released January 2, 2006)
45 =====================================================================
46
47 1. Fix restore of dumped Access Control Lists. The previous
48 code for EA works fine, but ACLs needed conversion from the
49 ext2/3 disk format to posix_acl format before restoring.
50
51 2. Fix some issues with restoration of EA on big endian
52 platforms.
53
54 3. Fix restore when the symtab is over 2GB in size.
55
56 4. Made the directory hash indexing an optional feature,
57 accessible by the '-H' option of restore, and disable it
58 by default.
59
60 5. Fixed dump to not include extended attributes information
61 in the toc (archive) file which confused restore -t.
62
63 Changes between versions 0.4b39 and 0.4b40 (released May 2, 2005)
64 =================================================================
65
66 1. Changed restore to emit warnings (instead of emitting a fatal
67 error) if a file (or a directory) is unavailable for a
68 comparision (if the user doesn't have the necessary permissions
69 to access it for example). Thanks to Kenneth Porter
70 <shiva@sewingwitch.com> for the bug report.
71
72 2. Re-done the 'do not save directory entries to non-dumped inodes
73 (excluded from dump)' feature. The previous implementation
74 worked well for excluded directories but not for regular files.
75 Thanks to Kenneth Porter <shiva@sewingwitch.com> for the bug
76 report.
77
78 3. Fixed a bug in dump where the tape size was miscalculated when
79 the user used -d/-s to specify the tape characteristics. Thanks
80 to Philip Goisman <goisman@physics.arizona.edu> for reporting
81 the bug.
82
83 4. Fixed another bug introduced in restore with the hashtree
84 implementation. This one caused restore to stop saying
85 "removenode: non-empty directory" in some cases.
86
87 5. Added support for dumping and restoring ext2/3 extended
88 attributes (EA), like the access control lists (ACL) or
89 the security labels used by SELinux.
90
91 Changes between versions 0.4b38 and 0.4b39 (released January 21, 2005)
92 ======================================================================
93
94 1. The newly added dump_on_cd_3 example was buggy, replace it
95 with an updated version from Andrew Basterfield
96 <bob@cemetery.homeunix.org>.
97
98 2. Made restore to chdir() back into the initial directory when
99 dumping core while aborting a comparision operation. The
100 previous behaviour was to write the corefile at the root of
101 the directory being compared, which could very well be
102 read only and preventing the corefile generation. Thanks
103 to Kenneth Porter <shiva@sewingwitch.com> for the bug report.
104
105 3. Silenced the failure to call fgetflags() when comparing an
106 entry which has no ext2 attributes (as in lsattr()).
107
108 4. Fix a brown paper bug in restore -C which broke restore and
109 caused modifications on the filesystem being compared
110 (directories containing a file with the same name as the
111 directory get renamed to RSTTMP...). Thanks to Kenneth Porter
112 <shiva@sewingwitch.com> for finding the bug and helping me
113 reproduce it.
114
115 5. Made restore -C force the -N flag (no writing allowed on
116 the disk) in order to prevent more bugs like the above one.
117
118 Changes between versions 0.4b37 and 0.4b38 (released January 7, 2005)
119 =====================================================================
120
121 1. Fix a couple of troff syntax bugs in the man pages.
122 Thanks to Eric Raymond <esr@thyrsus.com> for the patch.
123
124 2. Made restore use either libncurses or libtermcap, depending
125 on which one is available at configure time.
126
127 3. Fixed restore negative size display bug when comparing a
128 dump containing files over 2GB. Thanks to Steve Bonds
129 <sbonds@users.sourceforge.net> for the bug report.
130
131 4. Do not save directory entries to non-dumped inodes
132 (excluded from dump). This will eliminate the 'missing
133 file' warnings when doing 'restore -C'.
134
135 5. Fix dump crash when backuping a huge (2TB) filesystem,
136 due to a bogus calculation on the inode map size.
137 Thanks to Kevin B. Haines <K.B.Haines@rl.ac.uk> for
138 submitting the bug and testing the fix.
139
140 6. Fix a problem in restore where the final \0 in the symbolic
141 link names could have been lost, generating corrupt filenames.
142 Thanks to Kyle Wilson <kyle.wilson@amd.com> for reporting the
143 bug.
144
145 7. Implemented a hash list for the directory names in restore.
146 The linear list used before caused problems in interactive
147 restores when dealing with directories having thousands of
148 entries. Thanks to Brian Ristuccia <bristuccia@starentnetworks.com>
149 for reporting the bug.
150
151 8. Improved restore -C, this time including the directory
152 attributes into the comparision.
153
154 9. Made restore understand tapes containing EA/ACLs (which will
155 be dumped by the next version of dump). In this version
156 extended attributes on the tape are ignored, for full EA/ACL
157 support wait for the next version or try the experimental EA
158 patch.
159
160 Changes between versions 0.4b36 and 0.4b37 (released July 7, 2004)
161 ==================================================================
162
163 1. Added the --enable-staticz configure option which enables
164 dump and restore to be linked with static versions of
165 libz and libbz2 (and dynamic versions of all other libraries).
166 This will make Debian users happy, because libz and libbz2
167 were the only needed libraries living in /usr, all the
168 others live in /lib. In case of system emergency, it is
169 better not to have to rely on an extra filesystem.
170 Thanks to Bdale Garbee <bdale@gag.com> for the suggestion.
171
172 2. Fix compilation on (at least the Linux Debian port to) AMD64.
173 (<ext2fs/ext2_types.h> defines some types (__s64 and __u64)
174 that are also defined by <linux/types.h> (<asm/types.h>) and
175 they conflict).
176
177 3. Make dump's reading of the dumpdates file a bit more robust,
178 preventing dump from crashing when the dumpdates file has
179 been modified by hand.
180
181 4. Fixed some offset calculations in dump code which could
182 lead to "bread lseek errors" on large filesystems. Thanks
183 to Bruce Lowekamp <lowekamp@users.sourceforge.net> for
184 reporting this bug and debugging the issue.
185
186 5. Made dump use the blkid library when searching for devices
187 by label or uuid instead of dump's own routines.
188
189 6. Corrected a bug in dump where a wrong LABEL=... line in
190 /etc/fstab could prevent dump from dumping unrelated
191 filesystems. Thanks to Bruce Lowekamp
192 <lowekamp@users.sourceforge.net> for reporting the bug.
193
194 Changes between versions 0.4b35 and 0.4b36 (released April 21, 2004)
195 ====================================================================
196
197 1. Fixed dump compilation with old gcc versions. Thanks to
198 Mike Castle <dalgoda@users.sourceforge.net> for the patch.
199
200 2. Fixed some warnings (howmany, roundup, powerof2 redefined)
201 when compiling against a recent glibc version.
202
203 3. Fixed a bug in restore preventing the read of a dump tape
204 written with Solaris 7 ufsdump. Thanks to Patrick Higgins
205 <phiggins@transzap.com> for reporting the bug and providing
206 the test case.
207
208 4. Changed dump to enable the creation of volumes of different
209 sizes in a single run (make -B accept a list of values).
210 Patch contributed by Florian Zumbiehl <florz@gmx.de>.
211
212 5. Use the glibc provided minor() and major() macros instead
213 of our own bitmask implementation. This should be safe for
214 when the major/minor namespace will migrate to 32 bits.
215 Thanks to Zhang Jun <zhangjun@nanjing-fnst.com> for reporting
216 the bug.
217
218 6. Made explicit in the dump man page that dump will not create
219 a remote file, it will only write to an already existing one.
220
221 7. Another try at making size estimates better again.
222
223 8. Put back the inconditional running of the end-of-tape
224 script which was changed in 0.434 to be run only when -M
225 or multiple -f were NOT specified. Some users rely on this
226 feature even when it is combined with -M/-f.
227
228 9. Fixed restore when restoring huge backups (where rstdir...
229 temporary files are over 2GB). Thanks to
230 Raphael Jaffey <rjaffey@artic.edu> for reporting this,
231 debugging the issue and making the patch.
232
233 10. Made restore understand FreeBSD UFS2 tapes. Thanks to
234 David <vrtin@users.sourceforge.net> for submitting the bug
235 and providing a test case.
236
237 11. Made dump work with an arbitrary integer level (no more 10
238 levels only limitation). Thanks to Matthew
239 <msvincen@midway.uchicago.edu> for the patch.
240
241 Changes between versions 0.4b34 and 0.4b35 (released December 21, 2003)
242 =======================================================================
243
244 1. Added a note in the dump man page saying that the default
245 blocksize can be 32 if -d is used with a high density argument.
246 Thanks to Antonios Christofides <A.Christofides@itia.ntua.gr>
247 for the patch.
248
249 2. Fixed configure to correctly understand CPPFLAGS, CFLAGS,
250 LDFLAGS environment variables. Thanks to Arcady Genkin
251 <antipode@users.sourceforge.net> for reporting the bug.
252
253 3. Made -e/-E options of dump accept an unlimited number of inodes
254 to be excluded, rather than a hardcoded maximum. Thanks to
255 Dietrich Rothe <d-rothe@users.sourceforge.net> for the patch.
256
257 4. Updated the autoconf system to 2.50. Forced the -D_BSD_SOURCE
258 and -D_USE_BSD_SIGNAL defines in configure in order to solve
259 64bit build problems because quad_t is redifined with a
260 different signature. Thanks to Mike Harris <mharris@redhat.com>
261 for reporting this bug.
262
263 5. Made restore build on Solaris, making possible to
264 restore Linux's "enhanced" tapes. Thanks to Uwe Gohlke
265 <uwe@ugsoft.de> for the patch.
266
267 6. Made an extension in the dump tape format capable of saving
268 MacOSX specific inode extensions. Uwe Gohlke <uwe@ugsoft.de>
269 wrote the extension and contributed the restore code back
270 into this codebase. The same extension mechanism will be
271 used in the future to save ACLs...
272
273 7. Made rmt work correctly with regard to QFA and local/remote
274 files and tapes. The remote access will however work only
275 when the dump provided rmt version is used. If you want to
276 use another rmt server, please do not use the QFA feature.
277 Thanks to Uwe Gohlke <uwe@ugsoft.de> for the patch.
278
279 Changes between versions 0.4b33 and 0.4b34 (released April 18, 2003)
280 ====================================================================
281
282 1. Fixed the INSTALL file to reflect the actual install paths.
283 Thanks to David Ronis <ronis@ronispc.chem.mcgill.ca> for
284 reporting the bug.
285
286 2. Fixed the configure script to only check for headers presence
287 instead of trying to compile them. This should fix issues
288 with old build environments. Thanks to Kari Lempiainen
289 <kari@funky.pp.fi> for reporting the bug.
290
291 3. Fixed restore to correctly ignore sockets when comparing
292 a dump (as socket cannot be properly restored anyway). Thanks
293 to Gunther Reiszig <gunther@mit.edu> for reporting the bug.
294
295 4. Fixed restore to correctly access the archive file (-A argument)
296 even when using a remote tape. Thanks to Fabrice Bellet
297 <fabrice@bellet.info> for reporting the bug.
298
299 5. Fixed (again) handling of long (largefile) seeks in rmt.
300 Thanks to Fabrice Bellet <fabrice@bellet.info> for reporting
301 the bug.
302
303 6. Fixed restore corner case when dealing with large block sizes
304 dump is able to write now (-b 1024). Thanks to Fabrice Bellet
305 <fabrice@bellet.info> for reporting the bug.
306
307 7. Fixed a bug preventing dump to access a filesystem having
308 a label exactly 16 bytes in length. Thanks to <minduni@ti-edu.ch>
309 for reporting the bug.
310
311 8. Made dump store dump dates using explicit timezones, fixing a
312 problem with incremental dumps if the timezone is modified
313 between the dumps. Thanks to Philippe Troin <phil@fifi.org> for
314 the bug report and the patch.
315
316 9. Fixed a bug encountered when dumping individual files (not full
317 filesystems or directories) and dangling symbolic links happen
318 to be in the list of files. For as far as dump is concerned,
319 dangling symbolic links are allowed, and are dumped as is.
320 Thanks to Jin-su Ahn <jsahn@ee.snu.ac.kr> for reporting the
321 bug and providing the fix.
322
323 10. Fixed open and creation modes and permissions for QFA and
324 table-of-contents files in dump and restore. Thanks to
325 Philippe Troin <phil@fifi.org> for the patch.
326
327 11. Fixed the archive file descriptor handling enabling it to be 0.
328 This can happen in some cases when shell redirections are used.
329 Thanks to Philippe Troin <phil@fifi.org> for the patch.
330
331 12. Delayed the opening of archive file until after suid had been
332 dropped (fixing a possible security issue if dump is suid).
333 Thanks to Philippe Troin <phil@fifi.org> for the patch.
334
335 13. Fixed the 'S' command handling in the rmt client part.
336 Thanks to Philippe Troin <phil@fifi.org> for the patch.
337
338 14. Modified the end-of-tape script handling to print out statistics
339 (and stop the timer) before launching the eot script. Also, the eot
340 script does not get run anymore when using -M (which makes sense) or
341 when multiple tapes are listed on the command line
342 (-f tape0,tape1,tapen) (which also makes sense).
343 Thanks to Philippe Troin <phil@fifi.org> for the patch.
344
345 15. Relicensed dump/restore under the 'revised' BSD license, as per
346 ftp://ftp.cs.berkeley.edu/ucb/4bsd/README.Impt.License.Change.
347
348 16. Added LZO compression to dump. This new compression method has
349 the advantage of being super fast, thus not killing tape streaming
350 on slow machines. Thanks to Helmut Jarausch
351 <jarausch@igpm.rwth-aachen.de> for the patch and to
352 Markus Oberhumer <markus@oberhumer.com> for giving special permission
353 to include his miniLZO project (GPL licensed) in dump/restore.
354
355 17. Some small buffer overruns fixes in rmt. Thanks to Antonomasia
356 <ant@notatla.demon.co.uk> for reporting the bugs.
357
358 18. Added a special rmt version which can do encryption when writing
359 to tape. Read examples/encrypted_rmt/README for details on
360 how to enable and configure it. Thanks to Ken Lalonde
361 <ken@globalremit.com> for the patch.
362
363 19. Made dump work with 2.5 kernel end of tape early warning semantics.
364 Thanks to Kai Makisara <Kai.Makisara@kolumbus.fi> for the patch.
365
366 20. Fixed a bug which caused dump -w|-W not to work anymore, because
367 the fs_freq and fs_passno fields in /etc/mtab are always set
368 to 0 0. Thanks to Trent Piepho <xyzzy@speakeasy.org> for
369 reporting the bug.
370
371 Changes between versions 0.4b32 and 0.4b33 (released February 10, 2003)
372 =======================================================================
373
374 1. Added a note in the restore man page clarifying the question
375 'set the permissions on the current directory ?' asked by
376 restore at the end of treatment in -i and -x modes.
377
378 2. Fixed the endianess issues when reading compressed tapes.
379 Thanks to Dark Force <daq4th@users.sourceforge.net> for
380 reporting this bug and providing test cases.
381
382 3. Fixed the "ACL won't be dumped" warning message (which showed
383 an extra, unrelated error message). Thanks to Dragan Krnic
384 <dkrnic@lycos.com> for reporting this bug.
385
386 4. Made dump look first into /etc/mtab, then into /etc/fstab
387 when searching for filesystem paths. Also fixed some problems
388 caused by binding mounts or multiple block device mounts.
389 Thanks to Matus Uhlar <uhlar@fantomas.sk>, Elliott Mitchell
390 <ehem@m5p.com>, Greg Edwards <gedwards@users.sourceforge.net>,
391 Brian Hoy <brian.hoy@opus.co.nz>. (fixes Debian bugs #147086
392 and #160305, Sourceforge bugs #618699 and #679832).
393
394 5. Made dump's -I option accept the value '0' meaning all the
395 read errors will be ignored. This can be useful when running
396 dump from unattended sessions (like cron jobs). Thanks to
397 John I Wang <jiwang@users.sourceforge.net> for the suggestion.
398
399 6. Fixed the output of dump to indicate 'blocks' instead of
400 'tape blocks' in the various messages (blocks are always
401 1 Kilobyte, tape blocks are 1 BK * '-b' argument), and
402 made it clearly print the current blocksize at the start of
403 a dump. Thanks to Michal Szymanski <msz@astrouw.edu.pl> for
404 the suggestions.
405
406 7. Made rmt understand long (largefiles) seeks.
407
408 8. Fixed build with very old versions of libext2fs, where
409 EXT2_FT_* constants were undefined.
410
411 9. Made dump accept the dumpdates path on the command line
412 (-D file option) instead of using only the hardcoded one.
413 Thanks to Piete Brooks <pb22@users.sourceforge.net> for the
414 suggestion.
415
416 10. Enabled rmt, LFS, readline, QFA options by default in
417 ./configure. Updated the configure process (new versions
418 of config.guess, config.sub etc).
419
420 Changes between versions 0.4b31 and 0.4b32 (released November 15, 2002)
421 =======================================================================
422
423 1. Changed dump to use fcntl(F_SETLK) style locking instead
424 of flock() when locking the dumpdates file. With the old
425 locking scheme, a local user having read rights on the
426 dumpdates file could be able to do a Denial of Service attack
427 on dump. In order to lock the dumpdates file with the new
428 scheme, the user would need to have write access on the file.
429 Thanks to Richard Johnson <Richard.Johnson3@ey.com> for
430 reporting the bug (originally a bugtraq post).
431
432 2. Fixed interactive 'ls' which caused spurious errors warnings
433 about 'undefined filetypes' detected. Thanks to Jorgen Ostling
434 <jorgen_ostling@users.sourceforge.net> for reporting this
435 bug.
436
437 3. Fixed dump's estimate when dealing with sparse inodes.
438
439 4. Modified dump to allow setting a blocksize bigger than 32kB
440 (raised the limit to 1024kB), because newer hardware needs
441 this for proper operation. Thanks to Dirk Traenapp
442 <dtraenapp@users.sourceforge.net> for reporting this.
443
444 5. Fixed a bug causing Dump to stop and report an error if an
445 inode number in the exclude file was followed by some amount
446 of whitespace. Thanks to Jeffrey Sofferin
447 <sofferin@users.sourceforge.net> for reporting this bug.
448
449 6. Fixed a bug which caused restore, in some particular cases,
450 to ask some 'scary' questions and leave a bunch of RSTTMP
451 directories behind when restoring incremental tapes. Thanks
452 to Philippe Troin <phil@fifi.org> for reporting this bug and
453 providing the test cases.
454
455 7. Changed the wording when inodes are excluded from dump:
456 replaced 'Added inode 7 to exclude list' with
457 'Excluding inode 7 (resize inode) from dump', as suggested
458 by Elliott Mitchell <ehem@m5p.com> in a Debian bug report.
459
460 Changes between versions 0.4b30 and 0.4b31 (released July 30, 2002)
461 ===================================================================
462
463 1. Fixed rmt open flags transmission (GNU's symbolic syntax over
464 rmt) which I broke in 0.4b29. Thanks to Eros Albertazzi
465 <eros@lamel.bo.cnr.it> for reporting the bug.
466
467 Changes between versions 0.4b29 and 0.4b30 (released July 25, 2002)
468 ===================================================================
469
470 1. Made dump print out the ext2 logical block number in case of a read
471 error, which can be given as an argument to debugfs' ncheck command,
472 in order to find out the affected inode. Added note about this
473 usage in dump's man page.
474
475 2. Fixed a problem in restore when reading tapes written on big
476 endian machines with very old versions of dump. The patch was
477 contributed by George Helffrich <george@geology.bristol.ac.uk>.
478
479 3. Fixed the tape length calculation when using large tapes
480 and compression. Thanks to Georg Lippold
481 <g_lippold@sourceforge.net> for reporting the bug.
482
483 4. Added a new set of examples in dump_on_cd_2 directory, based
484 on dump_on_cd examples but somewhat enhanced, supporting DVD
485 media, and localized in english and german. Thanks to
486 Georg Lippold <g_lippold@sourceforge.net> for the new scripts.
487
488 5. Made dump save 32 bit UID/GID. Internally, this was achieved
489 by switching from the old BSD inode format to the new BSD
490 inode format, which means that the tape format was changed.
491 However, since all restore versions out there should
492 transparently support both inode formats, the change should
493 have no side effects. Thanks to John Yu <jky@cs.bu.edu> for
494 reporting the bug.
495
496 6. Fixed a lot of warnings in the code shown when compiling
497 with 'gcc -W'. Thanks to Matthias Andree
498 <matthias.andree@stud.uni-dortmund.de> for reporting this.
499
500 7. Fixed a small markup bug in the dump man page. Thanks to
501 Eric S. Raymond <esr@minx.thyrsus.com> for submitting the
502 patch.
503
504 8. Rewrote entirely the man pages using the tmac.an macro
505 package (Linux man page format) instead of the original BSD
506 format. They should be now cleaner and easier to modify.
507
508 Changes between versions 0.4b28 and 0.4b29 (released June 8, 2002)
509 ==================================================================
510
511 1. Fixed a problem in the rmt ioctl command, where ioctl's issued from
512 non Linux clients were misinterpreted. The description of the problem
513 (incompatible numbering in Linux mtio opcodes) is documented at
514 ftp://ftp.fokus.gmd.de/pub/unix/star/README.mtio . Thanks to
515 Jörg Schilling <schilling@fokus.gmd.de> for reporting this bug and
516 providing an excellent, cross-platform replacement for rmt in his
517 star package.
518
519 2. Fixed a bug reported by Andy Mentges <amentges@jumpline.com>
520 which caused restore to fail when the inode maps were not entirely
521 contained on the first volume (when using really small volumes or
522 when dumping a huge number of inodes).
523
524 3. Fixed a problem in dump, where files in subdirectories of directories
525 that have the nodump flag set (or being excluded with -e/-E) were
526 sometimes incorrectly dumped (instead of being excluded). The original
527 fix is from David C Lawrence <tale@dd.org> in the FreeBSD version
528 of dump: http://www.FreeBSD.org/cgi/query-pr.cgi?pr=32414
529 Thanks to Ted Grzesik <tedgyz@roostme.com> for reporting the bug and
530 help testing the patch.
531
532 4. Added some example scripts from Gerd Bavendiek <bav@epost.de>
533 which makes one able to pipe the output of dump, by the net, to
534 a remote CD-burner server.
535
536 5. Made dump use O_CREAT|O_TRUNC both locally and remotely (over rmt),
537 and use GNU's symbolic syntax over rmt instead of numerical values
538 to assure multiple platform compatibility.
539
540 6. Documented the -d option in restore.
541
542 7. Added a -v (verbose) mode to dump. For now it just prints the number
543 of the inode being dumped, but this could evolve in future versions
544 to include interesting debugging output.
545
546 8. Added a -o flag to restore, which automatically restores the current
547 directory permissions in -i or -x mode, without asking the operator
548 whether to do so. Patch submitted by Tony Clayton <tonyc@e-smith.com>
549 and Peter Samuel <peters@e-smith.com>.
550
551 Changes between versions 0.4b27 and 0.4b28 (released April 12, 2002)
552 ====================================================================
553
554 1. Fixed a bug in the QFA feature which could have caused
555 a bad tape positionning on restore (causing extra delays in
556 restore). Credits go to Uwe Gohlke <uwe@ugsoft.de>.
557
558 2. Added a small note in the dump man page specifying that
559 there should be no white space between the option letter and
560 the -j or -z parameter, thanks to Kenneth Porter <shiva@well.com>
561
562 3. Made restore work with older versions of the readline library,
563 by checking in configure for several extended symbols. Restore
564 can now be compiled with a readline as old as the 2.0 release
565 (though it may be a good idea to upgrade it to a more recent
566 version...). Thanks to Andrew Donkin <ard@waikato.ac.nz> for
567 reporting the build failures.
568
569 4. Fixed a performance problem with the QFA file creation in
570 dump, which made unnecessary seeks on the tape slowing down
571 the dump. Thanks to Andrew Donkin <ard@waikato.ac.nz> for
572 reporting this issue.
573
574 5. Removed the inclusion of some kernel headers in the dump
575 source, which prevented the compile in some kernel/glibc
576 headers/architecture combination. Thanks to Bdale Garbee
577 <bdale@gag.com> for reporting the bug.
578
579 6. Added the appropriate error message when dump fails to
580 open the output file for writing. Thanks to Amith Varghese
581 <amithv@yahoo.com> for reporting this bug.
582
583 7. Made restore able to understand large Solaris ufsdump tapes
584 (containing inodes bigger than 4194304). Sun have introduced
585 an "extension" to the dump tape format when dealing with
586 those inodes, which was uncorrectly handled by Linux restore.
587 Thanks to Uwe Gohlke <uwe@ugsoft.de> for reporting the bug and
588 providing a test case.
589
590 8. Added the -m parameter to dump which optimises the output for
591 inodes having been changed but not modified since the last dump
592 ('changed' and 'modified' have the meaning defined in stat(2)).
593 For those inodes, dump will save only the metadata, instead of
594 saving the entire inode contents. Inodes which are either
595 directories or have been modified since the last dump are saved
596 in a regular way. Uses of this flag must be consistent, meaning
597 that either every dump in an incremental dump set have the flag,
598 or no one has it.
599
600 Changes between versions 0.4b26 and 0.4b27 (released February 15, 2002)
601 =======================================================================
602
603 1. Fixed behaviour of dump when exceeding resource limits
604 (SIGXFSZ treatment).
605
606 2. Added the -L flag to restore to allow the user to specify a
607 maximal allowed number of miscompares when using restore
608 with the -C option to check the backup.
609
610 3. Detailed the manual entry for the -N option of restore.
611
612 4. Added the -a flag to restore to make able doing unattended
613 restores in -i or -x mode (automatically walks through the
614 multiple dump volumes).
615
616 5. Extended the QFA mode to work with local files and/or
617 remote tapes and files. This way, restore can know in advance
618 the tape number and the offset for the inodes to extract and
619 can minimize the extraction time by seeking directly to the
620 good tape/offset.
621
622 6. Added the -A <archive> option to both dump and restore,
623 which makes dump to archive a dump table-of-contents in
624 the specified file to be used by restore to determine
625 whether a file is in the dump file that is being restored.
626 (the archive file syntax is also compatible with the
627 Solaris ufsdump generated one).
628
629 7. Small fix in restore making it able to read some (broken ?)
630 Solaris ufsdump tapes.
631
632 8. Fixed dump to correctly recognise the root filesystem when using
633 ext2 disk labels (LABEL=/). Thanks to John Yu <jky@it.bu.edu>
634 for reporting this bug.
635
636 9. Added the -P <file> option to restore to create a
637 Quick File Access file from an already made dump. Patch
638 contributed by Uwe Gohlke <uwe@ugsoft.de>.
639
640 10. Made restore compile and run on Solaris, making it a
641 possible replacement for the standard ufsrestore. Port was
642 contributed by Uwe Gohlke <uwe@ugsoft.de>.
643
644 Changes between versions 0.4b25 and 0.4b26 (released January 7, 2002)
645 =====================================================================
646
647 1. Added a set of backup scripts from Eugenio Diaz
648 <getnito@yahoo.com> in the examples section. It features
649 automatic (cron based) full and incremental dumping of
650 several filesystems on a separate filesystem.
651
652 2. Fixed a off-by-one miscalculation which disabled dumping
653 a one letter subdirectory. Thanks to Chris Siebenmann
654 <cks@utcc.utoronto.ca> for reporting the bug.
655
656 3. Fixed several restore bugs occuring when trying to
657 restore some missing files on the tape. Thanks to Chris
658 Siebenmann <cks@utcc.utoronto.ca> for reporting the bug.
659
660 4. Fixed --with-ldopts configure argument passing, installing from
661 a separate object directory, makefile cleanups contributed
662 by <splite@purdue.edu>.
663
664 5. Fix a bug which could caused, in some conditions, the highest
665 number inode of a filesystem, to not be dumped. Many thanks
666 to Chris Siebenmann <cks@utcc.utoronto.ca> for helping me
667 chase this bug.
668
669 Changes between versions 0.4b24 and 0.4b25 (released November 17, 2001)
670 =======================================================================
671
672 1. Added a mini howto from Patrick Walsh in the examples
673 directory.
674
675 2. Minor man pages syntax corrections. Thanks to
676 Chris Danis <screechco@home.com> for reporting the bugs.
677
678 3. Added a script from David B. Peterson <dave@toppledwagon.com>
679 to the examples section. It features dumping several
680 filesystems to a remote tape drive upon ssh.
681
682 4. Added a patch provided by Richard Jones <rich@annexia.org>
683 which allows BRADEMAX (number of read errors tolerated by
684 dump) to be adjusted using the -I option.
685
686 5. Fixed a bug which disabled doing "restore -C -f -". Thanks
687 to Clemens Stiglechner <a9401816@unet.univie.ac.at> for the
688 patch.
689
690 6. Add the -l option to restore to specify if, when doing a
691 remote restore, the file used to access the backup is a
692 regular file (the defaults being a tape drive). Restore needs
693 to know this information when reading a remote compressed
694 dump. Previously, this information was autodetected, but
695 the autodetection code fails (with ioctl: Inappropriate ioctl
696 for device) when using a non Linux remote box. Thanks to
697 many users and especially to Eros Albertazzi
698 <eros@lamel.bo.cnr.it> for reporting this.
699
700 7. Found a workaroung for the dump deadlock problem (3 childs
701 stuck in pause(), father in read()). The workaround seems
702 to work for me and several beta-testers. If it doesn't work
703 for you, please report back.
704
705 8. Updated the RPM spec file (BuildPrereq, URL etc).
706
707 Changes between versions 0.4b23 and 0.4b24 (released September 12, 2001)
708 ========================================================================
709
710 1. Fixed the permissions of a newly created QFA file by dump.
711
712 2. Cleaned up the source of dump (the external variables
713 definition was a complete mess, making possible to have
714 objects overlap).
715
716 3. Fixed restore to use the full tape volume path when doing
717 a compare (since it changes the working directory to the
718 filesystem being compared in the process).
719
720 4. Added the -q option to dump which makes dump abort
721 whenever it needs operator attention. It should be
722 easier to use dump in scripts with this option.
723
724 5. Detect the use of incompatible options to dump and
725 refuse them (like -a and -B options together).
726
727 6. Added bzip2 compression to dump/restore (use option -j level
728 to select it). Note that you will need a restore version
729 >= 0.4b24 in order to restore a bzip2 compressed dump.
730 The same warning as for the zlib compression applies:
731 the tape format of a bzip2 dump is not compatible with the
732 original BSD tape format.
733
734 7. Fixed a overflow problem in dump corrupting the dump when
735 very large files were encountered. Thanks to Vladimir Nadvornik
736 for the bug report.
737
738 8. Added a ioctl(BLKFLSBUF) in dump which should flush the
739 kernel buffer/page cache before starting the dump, helping
740 a bit those who use dump on mounted filesystems. Thanks to
741 John Yu <jky@it.bu.edu> and to Theodore T'so <tytso@mit.edu>
742 for this suggestion.
743
744 9. Updated the RPM spec file following the RedHat changes
745 (dynamically linked binaries now in /usr/sbin etc).
746
747 10. Added a patch from Helmut Jarausch <jarausch@igpm.rwth-aachen.de>
748 which enables restore to recognise multi volume compressed dumps
749 done on CD. Included his scripts for dump (which pipe the dump
750 output directly into cdrecord) and restore. There is now
751 possible to to backups to CD on the fly!
752
753 Changes between versions 0.4b22 and 0.4b23 (released July 20, 2001)
754 ===================================================================
755
756 1. Fixed a buffer overflow in restore/tape.c. Patch provided by
757 Marcus Meissner (Caldera International Security Dept.).
758
759 2. Implement the Sun rmt extended protocol. Patch submitted
760 by Ian Gordon <iangordon@users.sourceforge.net>.
761
762 3. Check for the e2fsprogs header <ext2fs/ext2_fs.h> instead of
763 the linux kernel header. This ensures that dump always has the
764 latest version of this file. Patch submitted by
765 Andreas Dilger <adilger@turbolinux.com>.
766
767 4. Report any filesystem present in either /etc/fstab with a
768 positive passno or /etc/dumpdates in dump -w output.
769 Patch submitted by Andreas Dilger <adilger@turbolinux.com>.
770
771 5. Fixed the looping problem in dump introduced in the
772 previous version.
773
774 6. Changed the -B option of dump to limit the size of
775 _compressed_ output per volume if compression is on.
776 Patch contributed by Helmut Jarausch
777 <jarausch@igpm.rwth-aachen.de>. Note however that, since
778 it is impossible to predict the size of the compressed
779 data before doing the compression, the -B limit is a bit
780 conservative.
781
782 7. Fixed a bug in reading the operator typed file/tape path for
783 the next volume in restore.
784
785 8. Implemented a "-F script" option to restore which permits the
786 user to specify a script which will be launched at the
787 beginning of each tape, useful for automatic programming of
788 tape changers for example. See the restore man page for the
789 script parameters and return codes.
790
791 9. Small fix for the QFA routines provided by Uwe Gohlke
792 <uwe@ugsoft.de>, and some recommendations for QFA uses in
793 the man pages.
794
795 10. Fixed the multivolume restoring where making a mistake
796 to the 'Mount next tape' prompt caused several blocks to
797 be lost.
798
799 11. Enhanced the -e option of dump to take as a parameter a
800 comma separated list of inode numbers.
801
802 12. Added the -E option to dump which specify a file containing
803 inode numbers to exclude from the dump.
804
805 13. Fixed the compressed multi-volume dump + restore.
806
807 Changes between versions 0.4b21 and 0.4b22 (released May 12, 2001)
808 ==================================================================
809
810 1. Made dump report the number of blocks dumped per volume.
811 Thanks to Kenneth Porter <shiva@well.com> for the suggestion.
812
813 2. Fix a bug in dump where waiting too much at the 'change volumes'
814 question caused the volume to be rewritten. Thanks to
815 Graham Reed <greed@users.sourceforge.net> for reporting the
816 bug and providing a patch.
817
818 3. Added a compression option to dump, contributed by Jerry
819 Peters <gapeters@worldnet.att.net>.
820
821 WARNING: the compression works only on files/pipes and
822 drives supporting variable block size.
823
824 WARNING2: tapes created using the compression feature are
825 incompatible with the standard dump tape format, and a
826 version of restore >= 0.4b22 will be needed for extraction.
827
828 4. Fixed some compilation problems for glibc 2.2.2 and 64 bit
829 architectures. Thanks to Paul Slootman <paul@debian.org> for
830 the patch and to Bdale Garbee <bdale@gag.com> for forwarding
831 it upstream.
832
833 5. Many cleanups (CPP defines, const char warnings, check of
834 ext2fs COMPAT flags, time_t cleanups, added libext2 version
835 in dump usage text) by Andreas Dilger <adilger@turbolinux.com>.
836
837 6. Made --prefix option work in configure. All the install path
838 are now based on the configure parameters.
839
840 7. Added the Quick File Access mode in dump/restore, contributed
841 by Uwe Gohlke <uwe@ugsoft.de>. In this mode, dump stores in
842 a file tape position for each inode, and this file is used by
843 restore (if called with parameter Q and the filename)
844 to directly position the tape at the file restore is currently
845 working on. This saves hours when restoring single files from
846 large backups, saves the tapes and the drive's head. Use
847 --enable-qfa option of configure to compile in the QFA support.
848
849 8. Added the possibility to dump several files and directories
850 in a single invocation of dump. Thanks to Uwe Gohlke
851 <uwe@ugsoft.de> for implementing this option.
852
853 9. Fixed the dumping and restoring of files with holes
854 (files ending with holes didn't get dumped or restored
855 correctly).
856
857 10. Fixed a socket descriptor leak in dump, which leaved opened
858 3 file descriptors per dump process (and there is one dump
859 process per tape).
860
861 11. Fixed dump large file system support, by explicit use of
862 open64/lseek64/etc functions (explicit use needed because
863 e2fsprogs libraries don't behave well when compiled with
864 FILE_OFFSET_BITS=64).
865
866 Changes between versions 0.4b20 and 0.4b21 (released January 13, 2001)
867 ======================================================================
868
869 1. Fixed some bugs in the dump -w|-W logic introduced by
870 the previous version. Thanks to Andreas Dilger
871 <adilger@turbolinux.com> for his help on this one.
872
873 2. Fixed again a compilation problem when using old e2fs
874 headers (filesystem label related). Thanks to many users
875 who reported this stupid error.
876
877 3. Fixed a build problem on old lib5 based systems dealing with
878 _PATH_MNTTAB being undefined. Thanks to John Adams
879 <johna@onevista.com> for reporting the bug.
880
881 4. Improved the error detection in restore when creating
882 the temporary files in TMPDIR. Now restore will corectly
883 report a 'no space left on device' error instead of
884 strange errors which could imply an archive corruption.
885 Thanks to Gabriel Somlo <somlo@cs.colostate.edu> and
886 bgjenero <bgjenero@sympatico.ca> for reporting the bug.
887
888 5. Added the throughput information in dump's progression
889 messages. Thanks to Andreas Dilger <adilger@turbolinux.com>
890 for the patch.
891
892 6. Use libext2fs's inode scanning routines, which are particularly
893 robust when dealing with errors (especially when having some
894 bad blocks on the filesystem). Many thanks to Colin
895 <colin@horizon.com> for reporting the bug, investigating
896 the issues, finding a workaround, writing the patch and
897 fully testing it... (of course, if this patch will break
898 anything, he is to blame for :-)).
899
900 7. Made dump and restore LFS aware. Dump can dump a filesystem
901 containing large files, generate a large file on output and
902 restore can restore them. This feature is not enabled by
903 default, use --enable-largefile option of configure to enable
904 it (you need to have a LFS aware glibc though). Thanks to
905 Andrea Arcangeli <andrea@suse.de> for submitting the patch,
906 and to Theodore T'so <tytso@valinux.com> for his always
907 useful thoughts.
908
909 8. Made dump ask upon a tape write error if it should rewrite
910 the current volume (assume this is a bad tape) or if it should
911 assume an end-of-tape condition (useful for tape drives which
912 behaves badly at the end of the tape). Thanks to Andreas
913 Dilger <adilger@turbolinux.com> for the suggestion.
914
915 Changes between versions 0.4b19 and 0.4b20 (released November 10, 2000)
916 =======================================================================
917
918 1. Fixed a small compilation problem due to a change
919 in the definintion of the struct sigaction in
920 glibc 2.0 / libc5. Thanks to Gunther Schlegel
921 <schlegel@riege.de> for reporting the bug and to
922 Dave Platt <dplatt@snulbug.mtview.ca.us> for suggesting
923 a fix.
924
925 2. Modified the label and uuid analysis in order to be
926 self-contained (independant of kernel/libc headers). This
927 should fix the compile with older kernel/libc headers and
928 will preserve the functionality. Thanks to Bernhard Erdmann
929 <bernhard.erdmann@gmx.de> for reporting the bug.
930
931 3. The 'exclude inode' option, if applied to a directory
932 inode, excludes now correctly all the leaves of this
933 directory. Thanks to John R. Dennison
934 <gerdesas@users.sourceforge.net> for reporting the bug.
935
936 4. Fixed the '-e' option to disable the possibility
937 to exclude the root inode (which causes the dumps to
938 be unreadable by restore). Prevented array overflow
939 when multiple -e options are used.
940
941 5. Fixed dump to correctly interpret a filesystem argument
942 which is a mountpoint and it is not an absolute path
943 (as specified in the fstab). Thanks to Bernhard R. Erdmann
944 <be@berdmann.de> for reporting the bug.
945
946 6. Made dump able to backup files larger than 2 GB. Note that
947 dump still doesn't cope with files larger than 4 GB.
948
949 7. Restore the real uid and gid before invoking an external
950 RSH program (big hole when dump or restore is suid root!).
951
952 8. Get the values for the default tape device and the location
953 of fstab file from the system headers. Thanks to
954 Andreas Dilger <adilger@turbolinux.com> for the patch.
955
956 9. Made dump -w|-W report all recognized filesystems
957 present in either /etc/fstab or /etc/dumpdates, and present
958 the list in the same order as in fstab file. Thanks
959 to Andreas Dilger <adilger@turbolinux.com> for the patch.
960
961 10. Made dump's -a (automatic end of tape detection) the
962 default. Specifying one of -B, -c, -d or -s options will
963 override the EOM detection. Thanks to Andreas Dilger
964 <adilger@turbolinux.com> for the patch.
965
966 11. Save the ext2 filesystem label into the dump volume label.
967 Specifying a volume label on the command line (-L option)
968 will override this feature. Thanks to Andreas Dilger
969 <adilger@turbolinux.com> for the patch.
970
971 Changes between versions 0.4b18 and 0.4b19 (released August 20, 2000)
972 =====================================================================
973
974 1. Fixed the signal handling in dump (which I broke in 0.4b17)
975 which was causing several strange dump failures (dump
976 hanged or segmentation faults etc).
977
978 2. Specified the default blocksize in the dump man page.
979
980 3. Changed two info messages of restore to be written on stdout
981 instead of stderr in order to leave stderr only for errors
982 or warnings. Thanks to Stephen Atwell
983 <satwell@urbana.css.mot.com> for the suggestion.
984
985 4. Corrected an off by one calculation which prevented
986 dumping a directory having a 1 character name length.
987 Thanks to Bernhard Erdmann <bernhard.erdmann@gmx.de>
988 for reporting the bug.
989
990 5. Reinforce dump in order to allow only level 0 and no
991 -u option when dumping a subdirectory, and document
992 this in the man page. Thanks to Bernhard Erdmann
993 <bernhard.erdmann@gmx.de> for reporting the bug.
994
995 6. Fixed a small harmless bug in restore which caused
996 in some conditions hard links to be restored several
997 times (therefore generation some warning messages).
998 Thanks to Kenneth Porter <shiva@well.com> for
999 reporting the bug.
1000
1001 7. Updated the RPM spec file to the latest RedHat version,
1002 providing FHS packaging and other cosmetic changes.
1003 You will need RPM version >= 3.0.5 in order to build the RPM.
1004
1005 8. Updated the configure script to check for libtermcap
1006 before checking for libreadline, because we need this
1007 library in order to compile the readline support.
1008
1009 9. Made dump understand the LABEL= and UUID= notation
1010 both in /etc/fstab and on the command line. Note that
1011 you will need the /proc filesystem in order to use
1012 these notations. Thanks to Erik Troan <ewt@redhat.com>
1013 for providing the patch.
1014
1015 Changes between versions 0.4b17 and 0.4b18 (released June 30, 2000)
1016 ===================================================================
1017
1018 1. Fixed a potential buffer overflow in restore. Thanks
1019 to Stan Bubrouski <satan@fastdial.net> for reporting
1020 the bug.
1021
1022 2. Fixed a readline-related bug which prevented
1023 'cat DUMPFILE | restore -i -f -' from working. Thanks
1024 to Charles Karney <karney@users.sourceforge.net>
1025 for the bug report.
1026
1027 3. Changed a few "panic" into "exit", causing restore to
1028 be more stable against some attacks (like the last one
1029 reported on Bugtraq, although the last version of restore
1030 was not vulnerable - just dumped core). Thanks to
1031 Andreas Hasenack <andreas@conectiva.com.br> for reporting
1032 the bugs to me.
1033
1034 4. Removed the suid-root bit on dump and restore in the
1035 default build (and generated RPMs). It should be safer
1036 now. Users who need the suid-root capabilities in order
1037 to make network backups should read first the man page
1038 of dump and enable the suid bit manually.
1039
1040 5. Added -ltermcap to the compile parameters for restore
1041 when using readline support, in order to make the compile
1042 process work on some systems (SuSE ?). Thanks to
1043 Patrik Schindler <poc@pocnet.net> for reporting the bug.
1044
1045 Changes between versions 0.4b16 and 0.4b17 (released June 1st, 2000)
1046 ====================================================================
1047
1048 1. The -F script is called now *only* at the end of a tape,
1049 not at the end of the dump. Thanks to Christian Haul
1050 <haul@informatik.tu-darmstadt.de> for the bug report.
1051
1052 Normally, the device name and the current volume number
1053 are passed on the command line of the script. However,
1054 if you want the old style script (with no arguments
1055 passed on the command line), you can enable it in
1056 configure with the --enable-oldstylefscript.
1057
1058 2. Use posix signal handling to preserve dump functionality
1059 with libc5. Thanks to Jeff Johnson <jbj@redhat.com> for
1060 the patch.
1061
1062 3. Made the exit code of restore in case of a 'C'ompare
1063 command reflect the result of the compare. An exit status
1064 of 0 means the dump archive is correct, 1 means tape errors,
1065 2 means that some files were modified. Thanks to Kenneth Porter
1066 <shiva@well.com> for the suggestion.
1067
1068 4. Made (finally) quotes work in interactive restore.
1069
1070 5. Small fixes in order to allow dump to compile with a
1071 really old e2fsprogs version. Thanks to Ian Zimmerman
1072 <itz@speakeasy.org> for the bug report.
1073
1074 6. Add GNU readline capabilities to interactive restore.
1075 Use configure's --enable-readline flag to enable this feature.
1076 Thanks to Patrik Schindler <poc@pocnet.net> for the
1077 suggestion.
1078
1079 7. Do the compare on the fly in restore 'C' mode (this will
1080 allow not to exhaust the available /tmp space when
1081 ccmparing large files). Thanks to Kenneth Porter
1082 <shiva@well.com> for the suggestion.
1083
1084 Changes between versions 0.4b15 and 0.4b16 (released March 11, 2000)
1085 ====================================================================
1086
1087 1. Fixed some several duplicate 'DUMP: DUMP:' in the
1088 output of dump.
1089
1090 2. Corrected the estimation of blocks to dump. Note that
1091 this is still wrong for large directory inodes, due
1092 to the size of a BSD directory entry on the tape
1093 (estimation is based on the size of the ext2 directory,
1094 which is smaller because it doesn't contain the null
1095 character at the end).
1096
1097 3. Made dump report the total number of MB written to
1098 tape at the end of run. Thanks to W. Reilly Cooley
1099 <wcooley@nakedape.cc> for the patch.
1100
1101 4. Added the -X option to restore giving the possibility
1102 to read the names of the files to be extracted/listed
1103 from a text file (in addition of the command line).
1104 Thanks to Dejan Muhamedagic <dejan@quant-x.com> for the
1105 patch.
1106
1107 5. Added the device name and the current volume number
1108 as arguments to the end of tape script (dump -F option).
1109
1110 6. Made the multi-volume dump work again (after having
1111 broken it in 0.4b15).
1112
1113 Changes between versions 0.4b14 and 0.4b15 (released March 2, 2000)
1114 ===================================================================
1115
1116 1. Added a prompt command in interactive restore mode. Thanks
1117 to Andreas Dilger <adilger@home.com> for the patch.
1118
1119 2. Fixed a buffer overflow problem in dump (caused by
1120 not checking the size of the filesystem parameter).
1121 Thanks to Kim Yong-jun <loveyou@hackerslab.org> for
1122 reporting this on Bugtraq (and to several dump users
1123 who forwarded me his mail).
1124
1125 3. Added the '-F script' option to dump in order to
1126 launch a script at the end of each tape (to be used
1127 with a tape changer, or to notify the sysadmin by
1128 pager etc.).
1129
1130 4. Fixed a bug in restore compare code caused by the changes
1131 I made in 0.4b14.
1132
1133 5. Fixed the treatment of options using the old BSD syntax
1134 in both dump and restore.
1135
1136 Changes between versions 0.4b13 and 0.4b14 (released February 10, 2000)
1137 =======================================================================
1138
1139 1. Fixed a bug in dump which may have caused invalid deleted
1140 directories to be dumped out if they were deleted after the
1141 mapping pass. This could occure on active filesystem and lead
1142 to heap corruption (causing dump malfunction in many possible ways).
1143 Thanks to Jan Sanislo <oystr@cs.washington.edu> for finding this
1144 bug and submitting the patch.
1145
1146 2. Fixed the handling of the filesystem parameter in dump. A
1147 '/mnt/disk' parameter caused the disk contents to be dumped,
1148 but a '/mnt/disk/' parameter caused the mountpoint directory
1149 to be dumped (generally an empty directory).
1150
1151 3. Improved the output of dump in order to tell which directory
1152 it is currently dumping (when dumping a subtree).
1153
1154 4. Added the '-e' exclude inode option to dump. Thanks to
1155 Isaac Chuang <ike@isl.stanford.edu> for contributing with the patch.
1156
1157 5. Added a REPORTING-BUGS file in order to provide a guide
1158 on how to correctly report a bug in dump/restore.
1159
1160 6. Fixed a restore problem when restoring a hard link to an inode
1161 having the immutable or append-only attribute set. Thanks to
1162 Ambrose Li <acli@mingpaoxpress.com> for submitting the patch.
1163
1164 7. Fixed a compatibility problem between dumps made on little
1165 endian machines (the format was correct) and big endian
1166 machines (the format was incorrect). This fix break the
1167 compatibility with the older dumps made on big endian
1168 machines (sparc, m86k, ppc etc). For the first time in
1169 linux dump's history (I believe), the dumps made by *BSD,
1170 Linux/alpha, Linux/sparc and Linux/x86 are compatible,
1171 so interchangeable. Thanks to Rob Cermak
1172 <cermak@ahab.rutgers.edu> for submitting the bug and
1173 helping me test the fix.
1174
1175 8. Fixed the way dump reports the remaining percent/time, if
1176 the number of blocks actually dumped exceeds the estimated
1177 number of blocks. Thanks to Jean-Paul van der Jagt
1178 <jeanpaul@dutepp0.et.tudelft.nl> for reporting the bug.
1179
1180 Changes between versions 0.4b12 and 0.4b13 (released January 21, 2000)
1181 ======================================================================
1182
1183 1. Small Y2K fix in the man pages :). Thanks to Bernhard Sadlowski
1184 <sadlowsk@Mathematik.Uni-Bielefeld.DE> for reporting the bug.
1185
1186 2. Removed the requirement to build the RPM as root from the
1187 spec file. Thanks to Christian Weisgerber
1188 <naddy@mips.rhein-neckar.de> for submitting this.
1189
1190 3. Fixed a bug in dump related to the 'filetype' feature of ext2fs,
1191 causing dump to block when dumping really huge filesystems.
1192 Many thanks to Patrik Schindler <poc@pocnet.net> for
1193 helping me find this bug.
1194
1195 4. Fixed the treatment for an interrupt signal when dump access
1196 the remote tape through RSH. Thanks to Christian Weisgerber
1197 <naddy@mips.rhein-neckar.de> for providing the patch.
1198
1199 5. Fixed a bug which was causing dump/restore to display
1200 garbage characters instead of the remote host name.
1201
1202 Changes between versions 0.4b11 and 0.4b12 (released January 8, 2000)
1203 =====================================================================
1204
1205 1. Small fix in the dump man page. Thanks to Thorsten Kukuk
1206 <kukuk@suse.de> for submitting the patch.
1207
1208 2. Fix for the exit code when using the size estimate option of
1209 dump. Thanks to Matti Taskinen <mkt@rni.helsinki.fi> for
1210 submitting the patch.
1211
1212 3. Handle EINTR in atomical reads/writes in dump, which was causing
1213 dump to fail on some systems. Thanks to Eric Jergensen
1214 <eric@dvns.com> for reporting the bug and submitting the patch.
1215
1216 4. Handle more than 16 characters for the device names in dumpdates.
1217 (up to 255 now). Thanks to Rainer Clasen <bj@ncc.cicely.de> for
1218 tracking down the problem and proposing the solution.
1219
1220 5. Fixed a bug in dump which prevented the creation of the
1221 dumpdates file when doing a 0-level dump without already
1222 having a dumpdates file. Thanks to Patrik Schindler
1223 <poc@pocnet.net> for reporting the bug.
1224
1225 6. Changed the way dump 'S' flag reports the size estimate
1226 from number of blocks into bytes (making it compatible
1227 with the Solaris version, and simplifying things for
1228 amanda users). Thanks to Jason L Tibbitts III
1229 <tibbs@math.uh.edu> for reporting the bug.
1230
1231 7. Fixed a compatibility problem in linux/alpha dump tape format.
1232 Now the linux/alpha dump are (again) compatible with the
1233 other dump formats. But this breaks compatibility with
1234 older dumps made on alpha. Thanks to Mike Tibor
1235 <tibor@lib.uaa.alaska.edu> for helping me in finding this bug.
1236
1237 Changes between versions 0.4b10 and 0.4b11 (released December 5, 1999)
1238 ======================================================================
1239
1240 1. Added a '--enable-kerberos' to configure.
1241
1242 2. Added a 'S' option to dump which determines the amount of space
1243 that is needed to perform the dump without actually doing it, similar
1244 to the Sun's ufsdump 'S' option. Patch contributed by Rob Cermak
1245 <cermak@ahab.rutgers.edu>.
1246
1247 3. Added a 'M' multi-volume option to dump and restore which enables
1248 dumping to multiple files (useful when dumping to an ext2
1249 partition to make several dump files in order to bypass the 2GB
1250 file size limitation). The 'f' argument is treated as a prefix and
1251 the output files will be named <prefix>001, <prefix>002 etc. With
1252 the 'M' flag, restore automatically selects the right file without
1253 asking to enter a new tape each time.
1254
1255 4. Fixed a memory leak which was causing dump to grow very big
1256 (270MB when dumping a 10GB filesystem...). Thanks to Jason
1257 Fearon <jasonf@netrider.org.au> for reporting the bug.
1258
1259 Changes between versions 0.4b9 and 0.4b10 (released November 21, 1999)
1260 ======================================================================
1261
1262 1. Make configure test if the system glob routines support
1263 extended syntax (ALTDIRFUNC). If not, use the internal glob
1264 routines instead of system ones. Thanks to Bernhard Sadlowski
1265 <sadlowsk@Mathematik.Uni-Bielefeld.DE> for reporting the bug
1266 and helping me resolve this and other minor libc5 compiling
1267 glitches.
1268
1269 2. Fix a problem when dumping a ext2fs with the 'filetype'
1270 feature enabled. Thanks to Patrick J. LoPresti
1271 <patl@cag.lcs.mit.edu> for reporting the bug and to
1272 Theodore Y. Ts'o <tytso@mit.edu> for providing the patch.
1273
1274 3. Made the nodump flag work on directories. A directory which
1275 has the nodump flag gets never dumped, regardless of its
1276 contents.
1277
1278 4. Integrate a patch from Jeremy Fitzhardinge <jeremy@goop.org>
1279 which allows dump on an active ext3 filesystem. However, this
1280 is a "quick and dirty" patch which enables backup of an ext3
1281 filesystem through the ext2 compatibility (by ignoring the
1282 NEEDS_RECOVERY bit). The journal file is not recognized and
1283 it is dumped (it should not).
1284
1285 5. Test the superblock compatibility flags when dumping, in order
1286 to be sure that we know how to deal with specific features.
1287
1288 Changes between versions 0.4b8 and 0.4b9 (released November 5, 1999)
1289 ====================================================================
1290
1291 1. Use lchown instead of chown, fixing a possible security problem
1292 when restoring symlinks (a malicious user could use this
1293 to deliberately corrupt the ownership of important system files).
1294 Thanks to Chris Siebenmann <cks@utcc.utoronto.ca> for detecting
1295 this and providing the patch.
1296
1297 Changes between versions 0.4b7 and 0.4b8 (released November 3, 1999)
1298 ====================================================================
1299
1300 1. Put dump sources under CVS, added Id tags in all files so
1301 one can use 'ident' on binary files.
1302
1303 2. Added the dump/restore version in the usage text so one can
1304 easily verify the version he is using.
1305
1306 3. Small patch from Nuno Oliveira <nuno@eq.uc.pt> which fixes
1307 a va_start/va_end problem on linux-ppc (always call va_start
1308 va_end in pairs each time we use a vararg function).
1309
1310 4. Added again the DT_* constants because old libc does not
1311 contain them :(. Thanks to Eric Maisonobe <virnet@nat.fr>
1312 for submitting the bug report.
1313
1314 5. Use ext2fs_llseek instead of llseek. With recent e2fsprogs
1315 this should enable dumping big (huge) filesystems.
1316
1317 6. Added the RSH environment variable in order to be able to
1318 use a rsh replacement like ssh when doing remote backups (and
1319 bypass the security limitations of rcmd). Now you can do remote
1320 backups without being root (or making dump setuid root).
1321
1322 7. Modified again the way dumpdates works. For incremental dumps,
1323 we need to read dumpdates even if we are not using 'u' option.
1324 Thanks to Bdale Garbee <bdale@gag.com> for his ideas on how
1325 this should work.
1326
1327 Changes between versions 0.4b6 and 0.4b7 (released October 8, 1999)
1328 ===================================================================
1329
1330 1. Removed the 'k' flag from the restore 'about' text if kerberos
1331 was not compiled in.
1332
1333 2. Prototyped (f)setflags from e2fsprogs and corrected the calls
1334 to them (fsetflags takes a char*, setflags an open fd!).
1335
1336 3. (f)setflags is called only if the flags aren't empty. If the
1337 file is a special file, a warning is printed, because changing
1338 flags implies opening the device. Normally, a special file
1339 should not have any flag... (Debian bug #29775, patch provided
1340 by Abhijit Dasgupta <abhijit@ans.net>).
1341
1342 4. Made possible to dump a file system not mentioned in /etc/fstab.
1343 (Debian bug #11904, patch provided by Eirik Fuller <eirik@netcom.com>).
1344
1345 5. Changed the default behaviour to not create dumpdates
1346 unless 'u' option is specified. Removed the old "debian-patch"
1347 which provided the same thing. (Debian bug #38136, #33818).
1348
1349 6. Removed all those dump*announce, since they were getting old...
1350
1351 7. Added warning messages if dumpdates does not exist and
1352 when an inferior level dump does not exist (except for a level 0
1353 dump).
1354
1355 8. Debugged the glob calls in interactive mode: restore used a
1356 dirent struct which was different from the /usr/include/dirent.h
1357 one (this used to work, is it a glibc2 change?), so none of the
1358 compat glob (which used /usr/include/dirent.h) or the system glob
1359 worked. Restore use now the system dirent (and the system
1360 DT_* constants), which are compatible with BSD ones.
1361
1362 9. Added a configure flag (--with-dumpdatespath) to specify
1363 the location of dumpdates file. By default, it is
1364 /etc/dumpdates.
1365
1366 10. Added the "AUTHOR" and "AVAILABILITY" sections and
1367 included the current date/version in man pages.
1368
1369 11. Corrected the estimation of remaining time when
1370 the operator doesn't change the tapes quickly enough. This
1371 was an old bug, I thought I corrected it, and discovered
1372 that in fact it was corrected in two different places, so
1373 the results canceled each other...
1374
1375 Changes between versions 0.4b5 and 0.4b6 (released October 1, 1999)
1376 ===================================================================
1377
1378 1. Integrated multiple patches from RedHat, Debian and SuSE:
1379
1380 - tweak dump/itime.c to not try to read dumpdates if the 'u' option
1381 isn't specified.
1382 - several fixes in the man pages.
1383 - update the default tape device to /dev/st0.
1384 - many updates for Linux Alpha (byte ordering, size_t etc).
1385 - buffer overruns.
1386 - use environment variable for TMPDIR (instead of /tmp).
1387 - use sigjmp_buf instead of jmp_buf (RedHat bug #3260).
1388 - workaround egcs bug (RedHat bugs #4281 and #2989).
1389 - wire $(OPT) throughout Makefile's.
1390
1391 2. Upgrade the dump revision to 1, making possible to dump filesystems
1392 made with e2fsprogs-1.15 or newer. Nothing seems to break...
1393
1394 3. Fix some compile warnings, prototype all functions.
1395
1396 4. Use glibc err/glob instead of internal compatibility
1397 routines (only if available).
1398
1399 5. Fix a compile error on Linux 2.2.7 / libc5 (5.4.44) (patch provided
1400 by Bernhard Sadlowski <sadlowsk@mathematik.uni-bielefeld.de>).
1401
1402 Changes between versions 0.4b4 and 0.4b5 (released September 22, 1999)
1403 ======================================================================
1404
1405 1. Integrated the changes from FreeBSD-3.1-RELEASE
1406 (mostly bug fixes, buffer overruns, dump has now an "automatic
1407 tape length calculation" flag, dump/restore can use kerberos now
1408 (this is NOT tested), use environment variables for TAPE and
1409 RMT etc.).
1410
1411 2. Integrated three RedHat patches ("glibc", "kernel" and "bread" patches)
1412
1413 3. Corrected a bug in restore when using 'C' option with multi-volumes
1414 tapes (files splited accros two tapes give "size changed" errors
1415 when comparing).
1416
1417 4. Corrected the long standing bug when dumping multiple tapes.
1418 This works for me, needs further testing.
1419
1420 Changes between versions 0.4b3 and 0.4b4 (released January 17, 1997)
1421 ====================================================================
1422
1423 1. Dump now runs correctly on kernels 2.1.x
1424 Fix made by Gerald Peters <gapeters@worldnet.att.net>
1425
1426 Changes between versions 0.4b2 and 0.4b3
1427 ========================================
1428
1429 1. Use realpath() if available
1430
1431 2. Report statistics
1432
1433 Changes between versions 0.4b1 and 0.4b2
1434 ========================================
1435
1436 1. Fixed the bug fix from Greg Lutz (I had made a mistake when integrating
1437 the patch)
1438
1439 2. Fixed restore to make it able to read FreeBSD 2.x dumps again
1440
1441 3. Fixed configure.in to correctly handle --enable-rmt
1442
1443 Changes between versions 0.3 and 0.4b1
1444 ======================================
1445
1446 1. Integrated the changes from 4.4BSD-Lite2
1447
1448 2. Integrated the patches from Debian and Red Hat
1449
1450 3. Portability changes: use the __u32, __u16, __s32, and __s16 types
1451
1452 4. Changed dump to use the Ext2fs library to get block addresses. This
1453 should solve the endianness problem on SparcLinux.
1454
1455 5. Created a configure.in file (shamelessly stolen from the e2fsprogs
1456 distribution's one) to use autoconf
1457
1458 6. Fixed a few minor bugs
1459
1460 Changes between versions 0.2e and 0.2f
1461 ======================================
1462
1463 1. Added the creation of named pipes (fifos) in restore.
1464
1465 2. Added the -N flag in the restore manual page.
1466
1467 3. Added the file kernel.patch which contains the llseek() optimization
1468 patch for 1.2.x kernels.
1469
1470 4. Fixed a bug in the restoration of symbolic links: owner and group were
1471 not restored.
1472
1473 5. Integrated some changes from FreeBSD 2.2.
1474
1475 6. Added a call to ftruncate() after restoring each file to restore
1476 correctly files ending by a hole.
1477
1478 Changes between versions 0.2d and 0.2e
1479 ======================================
1480
1481 1. Fixed a bug in the "set owner/mode" process. Every file was restored
1482 with owner = root (0) and group = root/wheel/whatever (0).
1483
1484 Changes between versions 0.2c and 0.2d
1485 ======================================
1486
1487 1. Dump is now able to backup 2GB+ filesystems.
1488
1489 2. Dump and restore can now be linked as static binaries.
1490
1491 Changes between versions 0.2b and 0.2c
1492 ======================================
1493
1494 1. Fixed a bug when dumping ``slow'' (i.e. normal) symbolic links.
1495
1496 Changes between versions 0.2a and 0.2b
1497 ======================================
1498
1499 1. Really fixed the bug that I should have corrected in 0.2a.
1500
1501 2. Enabled optimization again.
1502
1503 Changes between versions 0.2 and 0.2a
1504 =====================================
1505
1506 1. Disabled the optimization during compilation.
1507
1508 Changes between versions 0.1 and 0.2
1509 ====================================
1510
1511 1. Fixed a bug in fstab.c which caused a null pointer to be stored in
1512 the fs_type field (actually, I modified the file fstab.c to make it
1513 use the mntent functions).
1514
1515 2. Dump and restore now use a 4.3 BSD compatible dump format. Backups
1516 made by dump should be readable by the BSD restore and backups made
1517 by the BSD dump should be readable by restore. Unfortunately, this
1518 means that the dump format has changed between version 0.1 and version
1519 0.2 :-(
1520
1521 3. Dump is now able to backup a subtree, it is no longer limited to whole
1522 filesystems like the BSD version.
1523
1524 4. Dump now uses ext2_llseek() so it is able to backup filesystems bigger
1525 than 2 GB.
1526
1527 Changes between versions 0.0 and 0.1
1528 ====================================
1529
1530 1. Now create links rdump and rrestore during the `make install' step.
1531
1532 2. Linux port specific bugs added to the manual pages
1533
1534 3. Incorrect estimation of the number of tapes blocks fixed when doing
1535 incremental backups.
1536
1537 4. Better ls-like format in restore in interactive mode.