Fix SELinux xattr handling.
[dump.git] / CHANGES
1 $Id: CHANGES,v 1.285 2007/02/22 20:12:50 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 Changes between versions 0.4b40 and 0.4b41 (released January 2, 2006)
21 =====================================================================
22
23 1. Fix restore of dumped Access Control Lists. The previous
24 code for EA works fine, but ACLs needed conversion from the
25 ext2/3 disk format to posix_acl format before restoring.
26
27 2. Fix some issues with restoration of EA on big endian
28 platforms.
29
30 3. Fix restore when the symtab is over 2GB in size.
31
32 4. Made the directory hash indexing an optional feature,
33 accessible by the '-H' option of restore, and disable it
34 by default.
35
36 5. Fixed dump to not include extended attributes information
37 in the toc (archive) file which confused restore -t.
38
39 Changes between versions 0.4b39 and 0.4b40 (released May 2, 2005)
40 =================================================================
41
42 1. Changed restore to emit warnings (instead of emitting a fatal
43 error) if a file (or a directory) is unavailable for a
44 comparision (if the user doesn't have the necessary permissions
45 to access it for example). Thanks to Kenneth Porter
46 <shiva@sewingwitch.com> for the bug report.
47
48 2. Re-done the 'do not save directory entries to non-dumped inodes
49 (excluded from dump)' feature. The previous implementation
50 worked well for excluded directories but not for regular files.
51 Thanks to Kenneth Porter <shiva@sewingwitch.com> for the bug
52 report.
53
54 3. Fixed a bug in dump where the tape size was miscalculated when
55 the user used -d/-s to specify the tape characteristics. Thanks
56 to Philip Goisman <goisman@physics.arizona.edu> for reporting
57 the bug.
58
59 4. Fixed another bug introduced in restore with the hashtree
60 implementation. This one caused restore to stop saying
61 "removenode: non-empty directory" in some cases.
62
63 5. Added support for dumping and restoring ext2/3 extended
64 attributes (EA), like the access control lists (ACL) or
65 the security labels used by SELinux.
66
67 Changes between versions 0.4b38 and 0.4b39 (released January 21, 2005)
68 ======================================================================
69
70 1. The newly added dump_on_cd_3 example was buggy, replace it
71 with an updated version from Andrew Basterfield
72 <bob@cemetery.homeunix.org>.
73
74 2. Made restore to chdir() back into the initial directory when
75 dumping core while aborting a comparision operation. The
76 previous behaviour was to write the corefile at the root of
77 the directory being compared, which could very well be
78 read only and preventing the corefile generation. Thanks
79 to Kenneth Porter <shiva@sewingwitch.com> for the bug report.
80
81 3. Silenced the failure to call fgetflags() when comparing an
82 entry which has no ext2 attributes (as in lsattr()).
83
84 4. Fix a brown paper bug in restore -C which broke restore and
85 caused modifications on the filesystem being compared
86 (directories containing a file with the same name as the
87 directory get renamed to RSTTMP...). Thanks to Kenneth Porter
88 <shiva@sewingwitch.com> for finding the bug and helping me
89 reproduce it.
90
91 5. Made restore -C force the -N flag (no writing allowed on
92 the disk) in order to prevent more bugs like the above one.
93
94 Changes between versions 0.4b37 and 0.4b38 (released January 7, 2005)
95 =====================================================================
96
97 1. Fix a couple of troff syntax bugs in the man pages.
98 Thanks to Eric Raymond <esr@thyrsus.com> for the patch.
99
100 2. Made restore use either libncurses or libtermcap, depending
101 on which one is available at configure time.
102
103 3. Fixed restore negative size display bug when comparing a
104 dump containing files over 2GB. Thanks to Steve Bonds
105 <sbonds@users.sourceforge.net> for the bug report.
106
107 4. Do not save directory entries to non-dumped inodes
108 (excluded from dump). This will eliminate the 'missing
109 file' warnings when doing 'restore -C'.
110
111 5. Fix dump crash when backuping a huge (2TB) filesystem,
112 due to a bogus calculation on the inode map size.
113 Thanks to Kevin B. Haines <K.B.Haines@rl.ac.uk> for
114 submitting the bug and testing the fix.
115
116 6. Fix a problem in restore where the final \0 in the symbolic
117 link names could have been lost, generating corrupt filenames.
118 Thanks to Kyle Wilson <kyle.wilson@amd.com> for reporting the
119 bug.
120
121 7. Implemented a hash list for the directory names in restore.
122 The linear list used before caused problems in interactive
123 restores when dealing with directories having thousands of
124 entries. Thanks to Brian Ristuccia <bristuccia@starentnetworks.com>
125 for reporting the bug.
126
127 8. Improved restore -C, this time including the directory
128 attributes into the comparision.
129
130 9. Made restore understand tapes containing EA/ACLs (which will
131 be dumped by the next version of dump). In this version
132 extended attributes on the tape are ignored, for full EA/ACL
133 support wait for the next version or try the experimental EA
134 patch.
135
136 Changes between versions 0.4b36 and 0.4b37 (released July 7, 2004)
137 ==================================================================
138
139 1. Added the --enable-staticz configure option which enables
140 dump and restore to be linked with static versions of
141 libz and libbz2 (and dynamic versions of all other libraries).
142 This will make Debian users happy, because libz and libbz2
143 were the only needed libraries living in /usr, all the
144 others live in /lib. In case of system emergency, it is
145 better not to have to rely on an extra filesystem.
146 Thanks to Bdale Garbee <bdale@gag.com> for the suggestion.
147
148 2. Fix compilation on (at least the Linux Debian port to) AMD64.
149 (<ext2fs/ext2_types.h> defines some types (__s64 and __u64)
150 that are also defined by <linux/types.h> (<asm/types.h>) and
151 they conflict).
152
153 3. Make dump's reading of the dumpdates file a bit more robust,
154 preventing dump from crashing when the dumpdates file has
155 been modified by hand.
156
157 4. Fixed some offset calculations in dump code which could
158 lead to "bread lseek errors" on large filesystems. Thanks
159 to Bruce Lowekamp <lowekamp@users.sourceforge.net> for
160 reporting this bug and debugging the issue.
161
162 5. Made dump use the blkid library when searching for devices
163 by label or uuid instead of dump's own routines.
164
165 6. Corrected a bug in dump where a wrong LABEL=... line in
166 /etc/fstab could prevent dump from dumping unrelated
167 filesystems. Thanks to Bruce Lowekamp
168 <lowekamp@users.sourceforge.net> for reporting the bug.
169
170 Changes between versions 0.4b35 and 0.4b36 (released April 21, 2004)
171 ====================================================================
172
173 1. Fixed dump compilation with old gcc versions. Thanks to
174 Mike Castle <dalgoda@users.sourceforge.net> for the patch.
175
176 2. Fixed some warnings (howmany, roundup, powerof2 redefined)
177 when compiling against a recent glibc version.
178
179 3. Fixed a bug in restore preventing the read of a dump tape
180 written with Solaris 7 ufsdump. Thanks to Patrick Higgins
181 <phiggins@transzap.com> for reporting the bug and providing
182 the test case.
183
184 4. Changed dump to enable the creation of volumes of different
185 sizes in a single run (make -B accept a list of values).
186 Patch contributed by Florian Zumbiehl <florz@gmx.de>.
187
188 5. Use the glibc provided minor() and major() macros instead
189 of our own bitmask implementation. This should be safe for
190 when the major/minor namespace will migrate to 32 bits.
191 Thanks to Zhang Jun <zhangjun@nanjing-fnst.com> for reporting
192 the bug.
193
194 6. Made explicit in the dump man page that dump will not create
195 a remote file, it will only write to an already existing one.
196
197 7. Another try at making size estimates better again.
198
199 8. Put back the inconditional running of the end-of-tape
200 script which was changed in 0.434 to be run only when -M
201 or multiple -f were NOT specified. Some users rely on this
202 feature even when it is combined with -M/-f.
203
204 9. Fixed restore when restoring huge backups (where rstdir...
205 temporary files are over 2GB). Thanks to
206 Raphael Jaffey <rjaffey@artic.edu> for reporting this,
207 debugging the issue and making the patch.
208
209 10. Made restore understand FreeBSD UFS2 tapes. Thanks to
210 David <vrtin@users.sourceforge.net> for submitting the bug
211 and providing a test case.
212
213 11. Made dump work with an arbitrary integer level (no more 10
214 levels only limitation). Thanks to Matthew
215 <msvincen@midway.uchicago.edu> for the patch.
216
217 Changes between versions 0.4b34 and 0.4b35 (released December 21, 2003)
218 =======================================================================
219
220 1. Added a note in the dump man page saying that the default
221 blocksize can be 32 if -d is used with a high density argument.
222 Thanks to Antonios Christofides <A.Christofides@itia.ntua.gr>
223 for the patch.
224
225 2. Fixed configure to correctly understand CPPFLAGS, CFLAGS,
226 LDFLAGS environment variables. Thanks to Arcady Genkin
227 <antipode@users.sourceforge.net> for reporting the bug.
228
229 3. Made -e/-E options of dump accept an unlimited number of inodes
230 to be excluded, rather than a hardcoded maximum. Thanks to
231 Dietrich Rothe <d-rothe@users.sourceforge.net> for the patch.
232
233 4. Updated the autoconf system to 2.50. Forced the -D_BSD_SOURCE
234 and -D_USE_BSD_SIGNAL defines in configure in order to solve
235 64bit build problems because quad_t is redifined with a
236 different signature. Thanks to Mike Harris <mharris@redhat.com>
237 for reporting this bug.
238
239 5. Made restore build on Solaris, making possible to
240 restore Linux's "enhanced" tapes. Thanks to Uwe Gohlke
241 <uwe@ugsoft.de> for the patch.
242
243 6. Made an extension in the dump tape format capable of saving
244 MacOSX specific inode extensions. Uwe Gohlke <uwe@ugsoft.de>
245 wrote the extension and contributed the restore code back
246 into this codebase. The same extension mechanism will be
247 used in the future to save ACLs...
248
249 7. Made rmt work correctly with regard to QFA and local/remote
250 files and tapes. The remote access will however work only
251 when the dump provided rmt version is used. If you want to
252 use another rmt server, please do not use the QFA feature.
253 Thanks to Uwe Gohlke <uwe@ugsoft.de> for the patch.
254
255 Changes between versions 0.4b33 and 0.4b34 (released April 18, 2003)
256 ====================================================================
257
258 1. Fixed the INSTALL file to reflect the actual install paths.
259 Thanks to David Ronis <ronis@ronispc.chem.mcgill.ca> for
260 reporting the bug.
261
262 2. Fixed the configure script to only check for headers presence
263 instead of trying to compile them. This should fix issues
264 with old build environments. Thanks to Kari Lempiainen
265 <kari@funky.pp.fi> for reporting the bug.
266
267 3. Fixed restore to correctly ignore sockets when comparing
268 a dump (as socket cannot be properly restored anyway). Thanks
269 to Gunther Reiszig <gunther@mit.edu> for reporting the bug.
270
271 4. Fixed restore to correctly access the archive file (-A argument)
272 even when using a remote tape. Thanks to Fabrice Bellet
273 <fabrice@bellet.info> for reporting the bug.
274
275 5. Fixed (again) handling of long (largefile) seeks in rmt.
276 Thanks to Fabrice Bellet <fabrice@bellet.info> for reporting
277 the bug.
278
279 6. Fixed restore corner case when dealing with large block sizes
280 dump is able to write now (-b 1024). Thanks to Fabrice Bellet
281 <fabrice@bellet.info> for reporting the bug.
282
283 7. Fixed a bug preventing dump to access a filesystem having
284 a label exactly 16 bytes in length. Thanks to <minduni@ti-edu.ch>
285 for reporting the bug.
286
287 8. Made dump store dump dates using explicit timezones, fixing a
288 problem with incremental dumps if the timezone is modified
289 between the dumps. Thanks to Philippe Troin <phil@fifi.org> for
290 the bug report and the patch.
291
292 9. Fixed a bug encountered when dumping individual files (not full
293 filesystems or directories) and dangling symbolic links happen
294 to be in the list of files. For as far as dump is concerned,
295 dangling symbolic links are allowed, and are dumped as is.
296 Thanks to Jin-su Ahn <jsahn@ee.snu.ac.kr> for reporting the
297 bug and providing the fix.
298
299 10. Fixed open and creation modes and permissions for QFA and
300 table-of-contents files in dump and restore. Thanks to
301 Philippe Troin <phil@fifi.org> for the patch.
302
303 11. Fixed the archive file descriptor handling enabling it to be 0.
304 This can happen in some cases when shell redirections are used.
305 Thanks to Philippe Troin <phil@fifi.org> for the patch.
306
307 12. Delayed the opening of archive file until after suid had been
308 dropped (fixing a possible security issue if dump is suid).
309 Thanks to Philippe Troin <phil@fifi.org> for the patch.
310
311 13. Fixed the 'S' command handling in the rmt client part.
312 Thanks to Philippe Troin <phil@fifi.org> for the patch.
313
314 14. Modified the end-of-tape script handling to print out statistics
315 (and stop the timer) before launching the eot script. Also, the eot
316 script does not get run anymore when using -M (which makes sense) or
317 when multiple tapes are listed on the command line
318 (-f tape0,tape1,tapen) (which also makes sense).
319 Thanks to Philippe Troin <phil@fifi.org> for the patch.
320
321 15. Relicensed dump/restore under the 'revised' BSD license, as per
322 ftp://ftp.cs.berkeley.edu/ucb/4bsd/README.Impt.License.Change.
323
324 16. Added LZO compression to dump. This new compression method has
325 the advantage of being super fast, thus not killing tape streaming
326 on slow machines. Thanks to Helmut Jarausch
327 <jarausch@igpm.rwth-aachen.de> for the patch and to
328 Markus Oberhumer <markus@oberhumer.com> for giving special permission
329 to include his miniLZO project (GPL licensed) in dump/restore.
330
331 17. Some small buffer overruns fixes in rmt. Thanks to Antonomasia
332 <ant@notatla.demon.co.uk> for reporting the bugs.
333
334 18. Added a special rmt version which can do encryption when writing
335 to tape. Read examples/encrypted_rmt/README for details on
336 how to enable and configure it. Thanks to Ken Lalonde
337 <ken@globalremit.com> for the patch.
338
339 19. Made dump work with 2.5 kernel end of tape early warning semantics.
340 Thanks to Kai Makisara <Kai.Makisara@kolumbus.fi> for the patch.
341
342 20. Fixed a bug which caused dump -w|-W not to work anymore, because
343 the fs_freq and fs_passno fields in /etc/mtab are always set
344 to 0 0. Thanks to Trent Piepho <xyzzy@speakeasy.org> for
345 reporting the bug.
346
347 Changes between versions 0.4b32 and 0.4b33 (released February 10, 2003)
348 =======================================================================
349
350 1. Added a note in the restore man page clarifying the question
351 'set the permissions on the current directory ?' asked by
352 restore at the end of treatment in -i and -x modes.
353
354 2. Fixed the endianess issues when reading compressed tapes.
355 Thanks to Dark Force <daq4th@users.sourceforge.net> for
356 reporting this bug and providing test cases.
357
358 3. Fixed the "ACL won't be dumped" warning message (which showed
359 an extra, unrelated error message). Thanks to Dragan Krnic
360 <dkrnic@lycos.com> for reporting this bug.
361
362 4. Made dump look first into /etc/mtab, then into /etc/fstab
363 when searching for filesystem paths. Also fixed some problems
364 caused by binding mounts or multiple block device mounts.
365 Thanks to Matus Uhlar <uhlar@fantomas.sk>, Elliott Mitchell
366 <ehem@m5p.com>, Greg Edwards <gedwards@users.sourceforge.net>,
367 Brian Hoy <brian.hoy@opus.co.nz>. (fixes Debian bugs #147086
368 and #160305, Sourceforge bugs #618699 and #679832).
369
370 5. Made dump's -I option accept the value '0' meaning all the
371 read errors will be ignored. This can be useful when running
372 dump from unattended sessions (like cron jobs). Thanks to
373 John I Wang <jiwang@users.sourceforge.net> for the suggestion.
374
375 6. Fixed the output of dump to indicate 'blocks' instead of
376 'tape blocks' in the various messages (blocks are always
377 1 Kilobyte, tape blocks are 1 BK * '-b' argument), and
378 made it clearly print the current blocksize at the start of
379 a dump. Thanks to Michal Szymanski <msz@astrouw.edu.pl> for
380 the suggestions.
381
382 7. Made rmt understand long (largefiles) seeks.
383
384 8. Fixed build with very old versions of libext2fs, where
385 EXT2_FT_* constants were undefined.
386
387 9. Made dump accept the dumpdates path on the command line
388 (-D file option) instead of using only the hardcoded one.
389 Thanks to Piete Brooks <pb22@users.sourceforge.net> for the
390 suggestion.
391
392 10. Enabled rmt, LFS, readline, QFA options by default in
393 ./configure. Updated the configure process (new versions
394 of config.guess, config.sub etc).
395
396 Changes between versions 0.4b31 and 0.4b32 (released November 15, 2002)
397 =======================================================================
398
399 1. Changed dump to use fcntl(F_SETLK) style locking instead
400 of flock() when locking the dumpdates file. With the old
401 locking scheme, a local user having read rights on the
402 dumpdates file could be able to do a Denial of Service attack
403 on dump. In order to lock the dumpdates file with the new
404 scheme, the user would need to have write access on the file.
405 Thanks to Richard Johnson <Richard.Johnson3@ey.com> for
406 reporting the bug (originally a bugtraq post).
407
408 2. Fixed interactive 'ls' which caused spurious errors warnings
409 about 'undefined filetypes' detected. Thanks to Jorgen Ostling
410 <jorgen_ostling@users.sourceforge.net> for reporting this
411 bug.
412
413 3. Fixed dump's estimate when dealing with sparse inodes.
414
415 4. Modified dump to allow setting a blocksize bigger than 32kB
416 (raised the limit to 1024kB), because newer hardware needs
417 this for proper operation. Thanks to Dirk Traenapp
418 <dtraenapp@users.sourceforge.net> for reporting this.
419
420 5. Fixed a bug causing Dump to stop and report an error if an
421 inode number in the exclude file was followed by some amount
422 of whitespace. Thanks to Jeffrey Sofferin
423 <sofferin@users.sourceforge.net> for reporting this bug.
424
425 6. Fixed a bug which caused restore, in some particular cases,
426 to ask some 'scary' questions and leave a bunch of RSTTMP
427 directories behind when restoring incremental tapes. Thanks
428 to Philippe Troin <phil@fifi.org> for reporting this bug and
429 providing the test cases.
430
431 7. Changed the wording when inodes are excluded from dump:
432 replaced 'Added inode 7 to exclude list' with
433 'Excluding inode 7 (resize inode) from dump', as suggested
434 by Elliott Mitchell <ehem@m5p.com> in a Debian bug report.
435
436 Changes between versions 0.4b30 and 0.4b31 (released July 30, 2002)
437 ===================================================================
438
439 1. Fixed rmt open flags transmission (GNU's symbolic syntax over
440 rmt) which I broke in 0.4b29. Thanks to Eros Albertazzi
441 <eros@lamel.bo.cnr.it> for reporting the bug.
442
443 Changes between versions 0.4b29 and 0.4b30 (released July 25, 2002)
444 ===================================================================
445
446 1. Made dump print out the ext2 logical block number in case of a read
447 error, which can be given as an argument to debugfs' ncheck command,
448 in order to find out the affected inode. Added note about this
449 usage in dump's man page.
450
451 2. Fixed a problem in restore when reading tapes written on big
452 endian machines with very old versions of dump. The patch was
453 contributed by George Helffrich <george@geology.bristol.ac.uk>.
454
455 3. Fixed the tape length calculation when using large tapes
456 and compression. Thanks to Georg Lippold
457 <g_lippold@sourceforge.net> for reporting the bug.
458
459 4. Added a new set of examples in dump_on_cd_2 directory, based
460 on dump_on_cd examples but somewhat enhanced, supporting DVD
461 media, and localized in english and german. Thanks to
462 Georg Lippold <g_lippold@sourceforge.net> for the new scripts.
463
464 5. Made dump save 32 bit UID/GID. Internally, this was achieved
465 by switching from the old BSD inode format to the new BSD
466 inode format, which means that the tape format was changed.
467 However, since all restore versions out there should
468 transparently support both inode formats, the change should
469 have no side effects. Thanks to John Yu <jky@cs.bu.edu> for
470 reporting the bug.
471
472 6. Fixed a lot of warnings in the code shown when compiling
473 with 'gcc -W'. Thanks to Matthias Andree
474 <matthias.andree@stud.uni-dortmund.de> for reporting this.
475
476 7. Fixed a small markup bug in the dump man page. Thanks to
477 Eric S. Raymond <esr@minx.thyrsus.com> for submitting the
478 patch.
479
480 8. Rewrote entirely the man pages using the tmac.an macro
481 package (Linux man page format) instead of the original BSD
482 format. They should be now cleaner and easier to modify.
483
484 Changes between versions 0.4b28 and 0.4b29 (released June 8, 2002)
485 ==================================================================
486
487 1. Fixed a problem in the rmt ioctl command, where ioctl's issued from
488 non Linux clients were misinterpreted. The description of the problem
489 (incompatible numbering in Linux mtio opcodes) is documented at
490 ftp://ftp.fokus.gmd.de/pub/unix/star/README.mtio . Thanks to
491 Jörg Schilling <schilling@fokus.gmd.de> for reporting this bug and
492 providing an excellent, cross-platform replacement for rmt in his
493 star package.
494
495 2. Fixed a bug reported by Andy Mentges <amentges@jumpline.com>
496 which caused restore to fail when the inode maps were not entirely
497 contained on the first volume (when using really small volumes or
498 when dumping a huge number of inodes).
499
500 3. Fixed a problem in dump, where files in subdirectories of directories
501 that have the nodump flag set (or being excluded with -e/-E) were
502 sometimes incorrectly dumped (instead of being excluded). The original
503 fix is from David C Lawrence <tale@dd.org> in the FreeBSD version
504 of dump: http://www.FreeBSD.org/cgi/query-pr.cgi?pr=32414
505 Thanks to Ted Grzesik <tedgyz@roostme.com> for reporting the bug and
506 help testing the patch.
507
508 4. Added some example scripts from Gerd Bavendiek <bav@epost.de>
509 which makes one able to pipe the output of dump, by the net, to
510 a remote CD-burner server.
511
512 5. Made dump use O_CREAT|O_TRUNC both locally and remotely (over rmt),
513 and use GNU's symbolic syntax over rmt instead of numerical values
514 to assure multiple platform compatibility.
515
516 6. Documented the -d option in restore.
517
518 7. Added a -v (verbose) mode to dump. For now it just prints the number
519 of the inode being dumped, but this could evolve in future versions
520 to include interesting debugging output.
521
522 8. Added a -o flag to restore, which automatically restores the current
523 directory permissions in -i or -x mode, without asking the operator
524 whether to do so. Patch submitted by Tony Clayton <tonyc@e-smith.com>
525 and Peter Samuel <peters@e-smith.com>.
526
527 Changes between versions 0.4b27 and 0.4b28 (released April 12, 2002)
528 ====================================================================
529
530 1. Fixed a bug in the QFA feature which could have caused
531 a bad tape positionning on restore (causing extra delays in
532 restore). Credits go to Uwe Gohlke <uwe@ugsoft.de>.
533
534 2. Added a small note in the dump man page specifying that
535 there should be no white space between the option letter and
536 the -j or -z parameter, thanks to Kenneth Porter <shiva@well.com>
537
538 3. Made restore work with older versions of the readline library,
539 by checking in configure for several extended symbols. Restore
540 can now be compiled with a readline as old as the 2.0 release
541 (though it may be a good idea to upgrade it to a more recent
542 version...). Thanks to Andrew Donkin <ard@waikato.ac.nz> for
543 reporting the build failures.
544
545 4. Fixed a performance problem with the QFA file creation in
546 dump, which made unnecessary seeks on the tape slowing down
547 the dump. Thanks to Andrew Donkin <ard@waikato.ac.nz> for
548 reporting this issue.
549
550 5. Removed the inclusion of some kernel headers in the dump
551 source, which prevented the compile in some kernel/glibc
552 headers/architecture combination. Thanks to Bdale Garbee
553 <bdale@gag.com> for reporting the bug.
554
555 6. Added the appropriate error message when dump fails to
556 open the output file for writing. Thanks to Amith Varghese
557 <amithv@yahoo.com> for reporting this bug.
558
559 7. Made restore able to understand large Solaris ufsdump tapes
560 (containing inodes bigger than 4194304). Sun have introduced
561 an "extension" to the dump tape format when dealing with
562 those inodes, which was uncorrectly handled by Linux restore.
563 Thanks to Uwe Gohlke <uwe@ugsoft.de> for reporting the bug and
564 providing a test case.
565
566 8. Added the -m parameter to dump which optimises the output for
567 inodes having been changed but not modified since the last dump
568 ('changed' and 'modified' have the meaning defined in stat(2)).
569 For those inodes, dump will save only the metadata, instead of
570 saving the entire inode contents. Inodes which are either
571 directories or have been modified since the last dump are saved
572 in a regular way. Uses of this flag must be consistent, meaning
573 that either every dump in an incremental dump set have the flag,
574 or no one has it.
575
576 Changes between versions 0.4b26 and 0.4b27 (released February 15, 2002)
577 =======================================================================
578
579 1. Fixed behaviour of dump when exceeding resource limits
580 (SIGXFSZ treatment).
581
582 2. Added the -L flag to restore to allow the user to specify a
583 maximal allowed number of miscompares when using restore
584 with the -C option to check the backup.
585
586 3. Detailed the manual entry for the -N option of restore.
587
588 4. Added the -a flag to restore to make able doing unattended
589 restores in -i or -x mode (automatically walks through the
590 multiple dump volumes).
591
592 5. Extended the QFA mode to work with local files and/or
593 remote tapes and files. This way, restore can know in advance
594 the tape number and the offset for the inodes to extract and
595 can minimize the extraction time by seeking directly to the
596 good tape/offset.
597
598 6. Added the -A <archive> option to both dump and restore,
599 which makes dump to archive a dump table-of-contents in
600 the specified file to be used by restore to determine
601 whether a file is in the dump file that is being restored.
602 (the archive file syntax is also compatible with the
603 Solaris ufsdump generated one).
604
605 7. Small fix in restore making it able to read some (broken ?)
606 Solaris ufsdump tapes.
607
608 8. Fixed dump to correctly recognise the root filesystem when using
609 ext2 disk labels (LABEL=/). Thanks to John Yu <jky@it.bu.edu>
610 for reporting this bug.
611
612 9. Added the -P <file> option to restore to create a
613 Quick File Access file from an already made dump. Patch
614 contributed by Uwe Gohlke <uwe@ugsoft.de>.
615
616 10. Made restore compile and run on Solaris, making it a
617 possible replacement for the standard ufsrestore. Port was
618 contributed by Uwe Gohlke <uwe@ugsoft.de>.
619
620 Changes between versions 0.4b25 and 0.4b26 (released January 7, 2002)
621 =====================================================================
622
623 1. Added a set of backup scripts from Eugenio Diaz
624 <getnito@yahoo.com> in the examples section. It features
625 automatic (cron based) full and incremental dumping of
626 several filesystems on a separate filesystem.
627
628 2. Fixed a off-by-one miscalculation which disabled dumping
629 a one letter subdirectory. Thanks to Chris Siebenmann
630 <cks@utcc.utoronto.ca> for reporting the bug.
631
632 3. Fixed several restore bugs occuring when trying to
633 restore some missing files on the tape. Thanks to Chris
634 Siebenmann <cks@utcc.utoronto.ca> for reporting the bug.
635
636 4. Fixed --with-ldopts configure argument passing, installing from
637 a separate object directory, makefile cleanups contributed
638 by <splite@purdue.edu>.
639
640 5. Fix a bug which could caused, in some conditions, the highest
641 number inode of a filesystem, to not be dumped. Many thanks
642 to Chris Siebenmann <cks@utcc.utoronto.ca> for helping me
643 chase this bug.
644
645 Changes between versions 0.4b24 and 0.4b25 (released November 17, 2001)
646 =======================================================================
647
648 1. Added a mini howto from Patrick Walsh in the examples
649 directory.
650
651 2. Minor man pages syntax corrections. Thanks to
652 Chris Danis <screechco@home.com> for reporting the bugs.
653
654 3. Added a script from David B. Peterson <dave@toppledwagon.com>
655 to the examples section. It features dumping several
656 filesystems to a remote tape drive upon ssh.
657
658 4. Added a patch provided by Richard Jones <rich@annexia.org>
659 which allows BRADEMAX (number of read errors tolerated by
660 dump) to be adjusted using the -I option.
661
662 5. Fixed a bug which disabled doing "restore -C -f -". Thanks
663 to Clemens Stiglechner <a9401816@unet.univie.ac.at> for the
664 patch.
665
666 6. Add the -l option to restore to specify if, when doing a
667 remote restore, the file used to access the backup is a
668 regular file (the defaults being a tape drive). Restore needs
669 to know this information when reading a remote compressed
670 dump. Previously, this information was autodetected, but
671 the autodetection code fails (with ioctl: Inappropriate ioctl
672 for device) when using a non Linux remote box. Thanks to
673 many users and especially to Eros Albertazzi
674 <eros@lamel.bo.cnr.it> for reporting this.
675
676 7. Found a workaroung for the dump deadlock problem (3 childs
677 stuck in pause(), father in read()). The workaround seems
678 to work for me and several beta-testers. If it doesn't work
679 for you, please report back.
680
681 8. Updated the RPM spec file (BuildPrereq, URL etc).
682
683 Changes between versions 0.4b23 and 0.4b24 (released September 12, 2001)
684 ========================================================================
685
686 1. Fixed the permissions of a newly created QFA file by dump.
687
688 2. Cleaned up the source of dump (the external variables
689 definition was a complete mess, making possible to have
690 objects overlap).
691
692 3. Fixed restore to use the full tape volume path when doing
693 a compare (since it changes the working directory to the
694 filesystem being compared in the process).
695
696 4. Added the -q option to dump which makes dump abort
697 whenever it needs operator attention. It should be
698 easier to use dump in scripts with this option.
699
700 5. Detect the use of incompatible options to dump and
701 refuse them (like -a and -B options together).
702
703 6. Added bzip2 compression to dump/restore (use option -j level
704 to select it). Note that you will need a restore version
705 >= 0.4b24 in order to restore a bzip2 compressed dump.
706 The same warning as for the zlib compression applies:
707 the tape format of a bzip2 dump is not compatible with the
708 original BSD tape format.
709
710 7. Fixed a overflow problem in dump corrupting the dump when
711 very large files were encountered. Thanks to Vladimir Nadvornik
712 for the bug report.
713
714 8. Added a ioctl(BLKFLSBUF) in dump which should flush the
715 kernel buffer/page cache before starting the dump, helping
716 a bit those who use dump on mounted filesystems. Thanks to
717 John Yu <jky@it.bu.edu> and to Theodore T'so <tytso@mit.edu>
718 for this suggestion.
719
720 9. Updated the RPM spec file following the RedHat changes
721 (dynamically linked binaries now in /usr/sbin etc).
722
723 10. Added a patch from Helmut Jarausch <jarausch@igpm.rwth-aachen.de>
724 which enables restore to recognise multi volume compressed dumps
725 done on CD. Included his scripts for dump (which pipe the dump
726 output directly into cdrecord) and restore. There is now
727 possible to to backups to CD on the fly!
728
729 Changes between versions 0.4b22 and 0.4b23 (released July 20, 2001)
730 ===================================================================
731
732 1. Fixed a buffer overflow in restore/tape.c. Patch provided by
733 Marcus Meissner (Caldera International Security Dept.).
734
735 2. Implement the Sun rmt extended protocol. Patch submitted
736 by Ian Gordon <iangordon@users.sourceforge.net>.
737
738 3. Check for the e2fsprogs header <ext2fs/ext2_fs.h> instead of
739 the linux kernel header. This ensures that dump always has the
740 latest version of this file. Patch submitted by
741 Andreas Dilger <adilger@turbolinux.com>.
742
743 4. Report any filesystem present in either /etc/fstab with a
744 positive passno or /etc/dumpdates in dump -w output.
745 Patch submitted by Andreas Dilger <adilger@turbolinux.com>.
746
747 5. Fixed the looping problem in dump introduced in the
748 previous version.
749
750 6. Changed the -B option of dump to limit the size of
751 _compressed_ output per volume if compression is on.
752 Patch contributed by Helmut Jarausch
753 <jarausch@igpm.rwth-aachen.de>. Note however that, since
754 it is impossible to predict the size of the compressed
755 data before doing the compression, the -B limit is a bit
756 conservative.
757
758 7. Fixed a bug in reading the operator typed file/tape path for
759 the next volume in restore.
760
761 8. Implemented a "-F script" option to restore which permits the
762 user to specify a script which will be launched at the
763 beginning of each tape, useful for automatic programming of
764 tape changers for example. See the restore man page for the
765 script parameters and return codes.
766
767 9. Small fix for the QFA routines provided by Uwe Gohlke
768 <uwe@ugsoft.de>, and some recommendations for QFA uses in
769 the man pages.
770
771 10. Fixed the multivolume restoring where making a mistake
772 to the 'Mount next tape' prompt caused several blocks to
773 be lost.
774
775 11. Enhanced the -e option of dump to take as a parameter a
776 comma separated list of inode numbers.
777
778 12. Added the -E option to dump which specify a file containing
779 inode numbers to exclude from the dump.
780
781 13. Fixed the compressed multi-volume dump + restore.
782
783 Changes between versions 0.4b21 and 0.4b22 (released May 12, 2001)
784 ==================================================================
785
786 1. Made dump report the number of blocks dumped per volume.
787 Thanks to Kenneth Porter <shiva@well.com> for the suggestion.
788
789 2. Fix a bug in dump where waiting too much at the 'change volumes'
790 question caused the volume to be rewritten. Thanks to
791 Graham Reed <greed@users.sourceforge.net> for reporting the
792 bug and providing a patch.
793
794 3. Added a compression option to dump, contributed by Jerry
795 Peters <gapeters@worldnet.att.net>.
796
797 WARNING: the compression works only on files/pipes and
798 drives supporting variable block size.
799
800 WARNING2: tapes created using the compression feature are
801 incompatible with the standard dump tape format, and a
802 version of restore >= 0.4b22 will be needed for extraction.
803
804 4. Fixed some compilation problems for glibc 2.2.2 and 64 bit
805 architectures. Thanks to Paul Slootman <paul@debian.org> for
806 the patch and to Bdale Garbee <bdale@gag.com> for forwarding
807 it upstream.
808
809 5. Many cleanups (CPP defines, const char warnings, check of
810 ext2fs COMPAT flags, time_t cleanups, added libext2 version
811 in dump usage text) by Andreas Dilger <adilger@turbolinux.com>.
812
813 6. Made --prefix option work in configure. All the install path
814 are now based on the configure parameters.
815
816 7. Added the Quick File Access mode in dump/restore, contributed
817 by Uwe Gohlke <uwe@ugsoft.de>. In this mode, dump stores in
818 a file tape position for each inode, and this file is used by
819 restore (if called with parameter Q and the filename)
820 to directly position the tape at the file restore is currently
821 working on. This saves hours when restoring single files from
822 large backups, saves the tapes and the drive's head. Use
823 --enable-qfa option of configure to compile in the QFA support.
824
825 8. Added the possibility to dump several files and directories
826 in a single invocation of dump. Thanks to Uwe Gohlke
827 <uwe@ugsoft.de> for implementing this option.
828
829 9. Fixed the dumping and restoring of files with holes
830 (files ending with holes didn't get dumped or restored
831 correctly).
832
833 10. Fixed a socket descriptor leak in dump, which leaved opened
834 3 file descriptors per dump process (and there is one dump
835 process per tape).
836
837 11. Fixed dump large file system support, by explicit use of
838 open64/lseek64/etc functions (explicit use needed because
839 e2fsprogs libraries don't behave well when compiled with
840 FILE_OFFSET_BITS=64).
841
842 Changes between versions 0.4b20 and 0.4b21 (released January 13, 2001)
843 ======================================================================
844
845 1. Fixed some bugs in the dump -w|-W logic introduced by
846 the previous version. Thanks to Andreas Dilger
847 <adilger@turbolinux.com> for his help on this one.
848
849 2. Fixed again a compilation problem when using old e2fs
850 headers (filesystem label related). Thanks to many users
851 who reported this stupid error.
852
853 3. Fixed a build problem on old lib5 based systems dealing with
854 _PATH_MNTTAB being undefined. Thanks to John Adams
855 <johna@onevista.com> for reporting the bug.
856
857 4. Improved the error detection in restore when creating
858 the temporary files in TMPDIR. Now restore will corectly
859 report a 'no space left on device' error instead of
860 strange errors which could imply an archive corruption.
861 Thanks to Gabriel Somlo <somlo@cs.colostate.edu> and
862 bgjenero <bgjenero@sympatico.ca> for reporting the bug.
863
864 5. Added the throughput information in dump's progression
865 messages. Thanks to Andreas Dilger <adilger@turbolinux.com>
866 for the patch.
867
868 6. Use libext2fs's inode scanning routines, which are particularly
869 robust when dealing with errors (especially when having some
870 bad blocks on the filesystem). Many thanks to Colin
871 <colin@horizon.com> for reporting the bug, investigating
872 the issues, finding a workaround, writing the patch and
873 fully testing it... (of course, if this patch will break
874 anything, he is to blame for :-)).
875
876 7. Made dump and restore LFS aware. Dump can dump a filesystem
877 containing large files, generate a large file on output and
878 restore can restore them. This feature is not enabled by
879 default, use --enable-largefile option of configure to enable
880 it (you need to have a LFS aware glibc though). Thanks to
881 Andrea Arcangeli <andrea@suse.de> for submitting the patch,
882 and to Theodore T'so <tytso@valinux.com> for his always
883 useful thoughts.
884
885 8. Made dump ask upon a tape write error if it should rewrite
886 the current volume (assume this is a bad tape) or if it should
887 assume an end-of-tape condition (useful for tape drives which
888 behaves badly at the end of the tape). Thanks to Andreas
889 Dilger <adilger@turbolinux.com> for the suggestion.
890
891 Changes between versions 0.4b19 and 0.4b20 (released November 10, 2000)
892 =======================================================================
893
894 1. Fixed a small compilation problem due to a change
895 in the definintion of the struct sigaction in
896 glibc 2.0 / libc5. Thanks to Gunther Schlegel
897 <schlegel@riege.de> for reporting the bug and to
898 Dave Platt <dplatt@snulbug.mtview.ca.us> for suggesting
899 a fix.
900
901 2. Modified the label and uuid analysis in order to be
902 self-contained (independant of kernel/libc headers). This
903 should fix the compile with older kernel/libc headers and
904 will preserve the functionality. Thanks to Bernhard Erdmann
905 <bernhard.erdmann@gmx.de> for reporting the bug.
906
907 3. The 'exclude inode' option, if applied to a directory
908 inode, excludes now correctly all the leaves of this
909 directory. Thanks to John R. Dennison
910 <gerdesas@users.sourceforge.net> for reporting the bug.
911
912 4. Fixed the '-e' option to disable the possibility
913 to exclude the root inode (which causes the dumps to
914 be unreadable by restore). Prevented array overflow
915 when multiple -e options are used.
916
917 5. Fixed dump to correctly interpret a filesystem argument
918 which is a mountpoint and it is not an absolute path
919 (as specified in the fstab). Thanks to Bernhard R. Erdmann
920 <be@berdmann.de> for reporting the bug.
921
922 6. Made dump able to backup files larger than 2 GB. Note that
923 dump still doesn't cope with files larger than 4 GB.
924
925 7. Restore the real uid and gid before invoking an external
926 RSH program (big hole when dump or restore is suid root!).
927
928 8. Get the values for the default tape device and the location
929 of fstab file from the system headers. Thanks to
930 Andreas Dilger <adilger@turbolinux.com> for the patch.
931
932 9. Made dump -w|-W report all recognized filesystems
933 present in either /etc/fstab or /etc/dumpdates, and present
934 the list in the same order as in fstab file. Thanks
935 to Andreas Dilger <adilger@turbolinux.com> for the patch.
936
937 10. Made dump's -a (automatic end of tape detection) the
938 default. Specifying one of -B, -c, -d or -s options will
939 override the EOM detection. Thanks to Andreas Dilger
940 <adilger@turbolinux.com> for the patch.
941
942 11. Save the ext2 filesystem label into the dump volume label.
943 Specifying a volume label on the command line (-L option)
944 will override this feature. Thanks to Andreas Dilger
945 <adilger@turbolinux.com> for the patch.
946
947 Changes between versions 0.4b18 and 0.4b19 (released August 20, 2000)
948 =====================================================================
949
950 1. Fixed the signal handling in dump (which I broke in 0.4b17)
951 which was causing several strange dump failures (dump
952 hanged or segmentation faults etc).
953
954 2. Specified the default blocksize in the dump man page.
955
956 3. Changed two info messages of restore to be written on stdout
957 instead of stderr in order to leave stderr only for errors
958 or warnings. Thanks to Stephen Atwell
959 <satwell@urbana.css.mot.com> for the suggestion.
960
961 4. Corrected an off by one calculation which prevented
962 dumping a directory having a 1 character name length.
963 Thanks to Bernhard Erdmann <bernhard.erdmann@gmx.de>
964 for reporting the bug.
965
966 5. Reinforce dump in order to allow only level 0 and no
967 -u option when dumping a subdirectory, and document
968 this in the man page. Thanks to Bernhard Erdmann
969 <bernhard.erdmann@gmx.de> for reporting the bug.
970
971 6. Fixed a small harmless bug in restore which caused
972 in some conditions hard links to be restored several
973 times (therefore generation some warning messages).
974 Thanks to Kenneth Porter <shiva@well.com> for
975 reporting the bug.
976
977 7. Updated the RPM spec file to the latest RedHat version,
978 providing FHS packaging and other cosmetic changes.
979 You will need RPM version >= 3.0.5 in order to build the RPM.
980
981 8. Updated the configure script to check for libtermcap
982 before checking for libreadline, because we need this
983 library in order to compile the readline support.
984
985 9. Made dump understand the LABEL= and UUID= notation
986 both in /etc/fstab and on the command line. Note that
987 you will need the /proc filesystem in order to use
988 these notations. Thanks to Erik Troan <ewt@redhat.com>
989 for providing the patch.
990
991 Changes between versions 0.4b17 and 0.4b18 (released June 30, 2000)
992 ===================================================================
993
994 1. Fixed a potential buffer overflow in restore. Thanks
995 to Stan Bubrouski <satan@fastdial.net> for reporting
996 the bug.
997
998 2. Fixed a readline-related bug which prevented
999 'cat DUMPFILE | restore -i -f -' from working. Thanks
1000 to Charles Karney <karney@users.sourceforge.net>
1001 for the bug report.
1002
1003 3. Changed a few "panic" into "exit", causing restore to
1004 be more stable against some attacks (like the last one
1005 reported on Bugtraq, although the last version of restore
1006 was not vulnerable - just dumped core). Thanks to
1007 Andreas Hasenack <andreas@conectiva.com.br> for reporting
1008 the bugs to me.
1009
1010 4. Removed the suid-root bit on dump and restore in the
1011 default build (and generated RPMs). It should be safer
1012 now. Users who need the suid-root capabilities in order
1013 to make network backups should read first the man page
1014 of dump and enable the suid bit manually.
1015
1016 5. Added -ltermcap to the compile parameters for restore
1017 when using readline support, in order to make the compile
1018 process work on some systems (SuSE ?). Thanks to
1019 Patrik Schindler <poc@pocnet.net> for reporting the bug.
1020
1021 Changes between versions 0.4b16 and 0.4b17 (released June 1st, 2000)
1022 ====================================================================
1023
1024 1. The -F script is called now *only* at the end of a tape,
1025 not at the end of the dump. Thanks to Christian Haul
1026 <haul@informatik.tu-darmstadt.de> for the bug report.
1027
1028 Normally, the device name and the current volume number
1029 are passed on the command line of the script. However,
1030 if you want the old style script (with no arguments
1031 passed on the command line), you can enable it in
1032 configure with the --enable-oldstylefscript.
1033
1034 2. Use posix signal handling to preserve dump functionality
1035 with libc5. Thanks to Jeff Johnson <jbj@redhat.com> for
1036 the patch.
1037
1038 3. Made the exit code of restore in case of a 'C'ompare
1039 command reflect the result of the compare. An exit status
1040 of 0 means the dump archive is correct, 1 means tape errors,
1041 2 means that some files were modified. Thanks to Kenneth Porter
1042 <shiva@well.com> for the suggestion.
1043
1044 4. Made (finally) quotes work in interactive restore.
1045
1046 5. Small fixes in order to allow dump to compile with a
1047 really old e2fsprogs version. Thanks to Ian Zimmerman
1048 <itz@speakeasy.org> for the bug report.
1049
1050 6. Add GNU readline capabilities to interactive restore.
1051 Use configure's --enable-readline flag to enable this feature.
1052 Thanks to Patrik Schindler <poc@pocnet.net> for the
1053 suggestion.
1054
1055 7. Do the compare on the fly in restore 'C' mode (this will
1056 allow not to exhaust the available /tmp space when
1057 ccmparing large files). Thanks to Kenneth Porter
1058 <shiva@well.com> for the suggestion.
1059
1060 Changes between versions 0.4b15 and 0.4b16 (released March 11, 2000)
1061 ====================================================================
1062
1063 1. Fixed some several duplicate 'DUMP: DUMP:' in the
1064 output of dump.
1065
1066 2. Corrected the estimation of blocks to dump. Note that
1067 this is still wrong for large directory inodes, due
1068 to the size of a BSD directory entry on the tape
1069 (estimation is based on the size of the ext2 directory,
1070 which is smaller because it doesn't contain the null
1071 character at the end).
1072
1073 3. Made dump report the total number of MB written to
1074 tape at the end of run. Thanks to W. Reilly Cooley
1075 <wcooley@nakedape.cc> for the patch.
1076
1077 4. Added the -X option to restore giving the possibility
1078 to read the names of the files to be extracted/listed
1079 from a text file (in addition of the command line).
1080 Thanks to Dejan Muhamedagic <dejan@quant-x.com> for the
1081 patch.
1082
1083 5. Added the device name and the current volume number
1084 as arguments to the end of tape script (dump -F option).
1085
1086 6. Made the multi-volume dump work again (after having
1087 broken it in 0.4b15).
1088
1089 Changes between versions 0.4b14 and 0.4b15 (released March 2, 2000)
1090 ===================================================================
1091
1092 1. Added a prompt command in interactive restore mode. Thanks
1093 to Andreas Dilger <adilger@home.com> for the patch.
1094
1095 2. Fixed a buffer overflow problem in dump (caused by
1096 not checking the size of the filesystem parameter).
1097 Thanks to Kim Yong-jun <loveyou@hackerslab.org> for
1098 reporting this on Bugtraq (and to several dump users
1099 who forwarded me his mail).
1100
1101 3. Added the '-F script' option to dump in order to
1102 launch a script at the end of each tape (to be used
1103 with a tape changer, or to notify the sysadmin by
1104 pager etc.).
1105
1106 4. Fixed a bug in restore compare code caused by the changes
1107 I made in 0.4b14.
1108
1109 5. Fixed the treatment of options using the old BSD syntax
1110 in both dump and restore.
1111
1112 Changes between versions 0.4b13 and 0.4b14 (released February 10, 2000)
1113 =======================================================================
1114
1115 1. Fixed a bug in dump which may have caused invalid deleted
1116 directories to be dumped out if they were deleted after the
1117 mapping pass. This could occure on active filesystem and lead
1118 to heap corruption (causing dump malfunction in many possible ways).
1119 Thanks to Jan Sanislo <oystr@cs.washington.edu> for finding this
1120 bug and submitting the patch.
1121
1122 2. Fixed the handling of the filesystem parameter in dump. A
1123 '/mnt/disk' parameter caused the disk contents to be dumped,
1124 but a '/mnt/disk/' parameter caused the mountpoint directory
1125 to be dumped (generally an empty directory).
1126
1127 3. Improved the output of dump in order to tell which directory
1128 it is currently dumping (when dumping a subtree).
1129
1130 4. Added the '-e' exclude inode option to dump. Thanks to
1131 Isaac Chuang <ike@isl.stanford.edu> for contributing with the patch.
1132
1133 5. Added a REPORTING-BUGS file in order to provide a guide
1134 on how to correctly report a bug in dump/restore.
1135
1136 6. Fixed a restore problem when restoring a hard link to an inode
1137 having the immutable or append-only attribute set. Thanks to
1138 Ambrose Li <acli@mingpaoxpress.com> for submitting the patch.
1139
1140 7. Fixed a compatibility problem between dumps made on little
1141 endian machines (the format was correct) and big endian
1142 machines (the format was incorrect). This fix break the
1143 compatibility with the older dumps made on big endian
1144 machines (sparc, m86k, ppc etc). For the first time in
1145 linux dump's history (I believe), the dumps made by *BSD,
1146 Linux/alpha, Linux/sparc and Linux/x86 are compatible,
1147 so interchangeable. Thanks to Rob Cermak
1148 <cermak@ahab.rutgers.edu> for submitting the bug and
1149 helping me test the fix.
1150
1151 8. Fixed the way dump reports the remaining percent/time, if
1152 the number of blocks actually dumped exceeds the estimated
1153 number of blocks. Thanks to Jean-Paul van der Jagt
1154 <jeanpaul@dutepp0.et.tudelft.nl> for reporting the bug.
1155
1156 Changes between versions 0.4b12 and 0.4b13 (released January 21, 2000)
1157 ======================================================================
1158
1159 1. Small Y2K fix in the man pages :). Thanks to Bernhard Sadlowski
1160 <sadlowsk@Mathematik.Uni-Bielefeld.DE> for reporting the bug.
1161
1162 2. Removed the requirement to build the RPM as root from the
1163 spec file. Thanks to Christian Weisgerber
1164 <naddy@mips.rhein-neckar.de> for submitting this.
1165
1166 3. Fixed a bug in dump related to the 'filetype' feature of ext2fs,
1167 causing dump to block when dumping really huge filesystems.
1168 Many thanks to Patrik Schindler <poc@pocnet.net> for
1169 helping me find this bug.
1170
1171 4. Fixed the treatment for an interrupt signal when dump access
1172 the remote tape through RSH. Thanks to Christian Weisgerber
1173 <naddy@mips.rhein-neckar.de> for providing the patch.
1174
1175 5. Fixed a bug which was causing dump/restore to display
1176 garbage characters instead of the remote host name.
1177
1178 Changes between versions 0.4b11 and 0.4b12 (released January 8, 2000)
1179 =====================================================================
1180
1181 1. Small fix in the dump man page. Thanks to Thorsten Kukuk
1182 <kukuk@suse.de> for submitting the patch.
1183
1184 2. Fix for the exit code when using the size estimate option of
1185 dump. Thanks to Matti Taskinen <mkt@rni.helsinki.fi> for
1186 submitting the patch.
1187
1188 3. Handle EINTR in atomical reads/writes in dump, which was causing
1189 dump to fail on some systems. Thanks to Eric Jergensen
1190 <eric@dvns.com> for reporting the bug and submitting the patch.
1191
1192 4. Handle more than 16 characters for the device names in dumpdates.
1193 (up to 255 now). Thanks to Rainer Clasen <bj@ncc.cicely.de> for
1194 tracking down the problem and proposing the solution.
1195
1196 5. Fixed a bug in dump which prevented the creation of the
1197 dumpdates file when doing a 0-level dump without already
1198 having a dumpdates file. Thanks to Patrik Schindler
1199 <poc@pocnet.net> for reporting the bug.
1200
1201 6. Changed the way dump 'S' flag reports the size estimate
1202 from number of blocks into bytes (making it compatible
1203 with the Solaris version, and simplifying things for
1204 amanda users). Thanks to Jason L Tibbitts III
1205 <tibbs@math.uh.edu> for reporting the bug.
1206
1207 7. Fixed a compatibility problem in linux/alpha dump tape format.
1208 Now the linux/alpha dump are (again) compatible with the
1209 other dump formats. But this breaks compatibility with
1210 older dumps made on alpha. Thanks to Mike Tibor
1211 <tibor@lib.uaa.alaska.edu> for helping me in finding this bug.
1212
1213 Changes between versions 0.4b10 and 0.4b11 (released December 5, 1999)
1214 ======================================================================
1215
1216 1. Added a '--enable-kerberos' to configure.
1217
1218 2. Added a 'S' option to dump which determines the amount of space
1219 that is needed to perform the dump without actually doing it, similar
1220 to the Sun's ufsdump 'S' option. Patch contributed by Rob Cermak
1221 <cermak@ahab.rutgers.edu>.
1222
1223 3. Added a 'M' multi-volume option to dump and restore which enables
1224 dumping to multiple files (useful when dumping to an ext2
1225 partition to make several dump files in order to bypass the 2GB
1226 file size limitation). The 'f' argument is treated as a prefix and
1227 the output files will be named <prefix>001, <prefix>002 etc. With
1228 the 'M' flag, restore automatically selects the right file without
1229 asking to enter a new tape each time.
1230
1231 4. Fixed a memory leak which was causing dump to grow very big
1232 (270MB when dumping a 10GB filesystem...). Thanks to Jason
1233 Fearon <jasonf@netrider.org.au> for reporting the bug.
1234
1235 Changes between versions 0.4b9 and 0.4b10 (released November 21, 1999)
1236 ======================================================================
1237
1238 1. Make configure test if the system glob routines support
1239 extended syntax (ALTDIRFUNC). If not, use the internal glob
1240 routines instead of system ones. Thanks to Bernhard Sadlowski
1241 <sadlowsk@Mathematik.Uni-Bielefeld.DE> for reporting the bug
1242 and helping me resolve this and other minor libc5 compiling
1243 glitches.
1244
1245 2. Fix a problem when dumping a ext2fs with the 'filetype'
1246 feature enabled. Thanks to Patrick J. LoPresti
1247 <patl@cag.lcs.mit.edu> for reporting the bug and to
1248 Theodore Y. Ts'o <tytso@mit.edu> for providing the patch.
1249
1250 3. Made the nodump flag work on directories. A directory which
1251 has the nodump flag gets never dumped, regardless of its
1252 contents.
1253
1254 4. Integrate a patch from Jeremy Fitzhardinge <jeremy@goop.org>
1255 which allows dump on an active ext3 filesystem. However, this
1256 is a "quick and dirty" patch which enables backup of an ext3
1257 filesystem through the ext2 compatibility (by ignoring the
1258 NEEDS_RECOVERY bit). The journal file is not recognized and
1259 it is dumped (it should not).
1260
1261 5. Test the superblock compatibility flags when dumping, in order
1262 to be sure that we know how to deal with specific features.
1263
1264 Changes between versions 0.4b8 and 0.4b9 (released November 5, 1999)
1265 ====================================================================
1266
1267 1. Use lchown instead of chown, fixing a possible security problem
1268 when restoring symlinks (a malicious user could use this
1269 to deliberately corrupt the ownership of important system files).
1270 Thanks to Chris Siebenmann <cks@utcc.utoronto.ca> for detecting
1271 this and providing the patch.
1272
1273 Changes between versions 0.4b7 and 0.4b8 (released November 3, 1999)
1274 ====================================================================
1275
1276 1. Put dump sources under CVS, added Id tags in all files so
1277 one can use 'ident' on binary files.
1278
1279 2. Added the dump/restore version in the usage text so one can
1280 easily verify the version he is using.
1281
1282 3. Small patch from Nuno Oliveira <nuno@eq.uc.pt> which fixes
1283 a va_start/va_end problem on linux-ppc (always call va_start
1284 va_end in pairs each time we use a vararg function).
1285
1286 4. Added again the DT_* constants because old libc does not
1287 contain them :(. Thanks to Eric Maisonobe <virnet@nat.fr>
1288 for submitting the bug report.
1289
1290 5. Use ext2fs_llseek instead of llseek. With recent e2fsprogs
1291 this should enable dumping big (huge) filesystems.
1292
1293 6. Added the RSH environment variable in order to be able to
1294 use a rsh replacement like ssh when doing remote backups (and
1295 bypass the security limitations of rcmd). Now you can do remote
1296 backups without being root (or making dump setuid root).
1297
1298 7. Modified again the way dumpdates works. For incremental dumps,
1299 we need to read dumpdates even if we are not using 'u' option.
1300 Thanks to Bdale Garbee <bdale@gag.com> for his ideas on how
1301 this should work.
1302
1303 Changes between versions 0.4b6 and 0.4b7 (released October 8, 1999)
1304 ===================================================================
1305
1306 1. Removed the 'k' flag from the restore 'about' text if kerberos
1307 was not compiled in.
1308
1309 2. Prototyped (f)setflags from e2fsprogs and corrected the calls
1310 to them (fsetflags takes a char*, setflags an open fd!).
1311
1312 3. (f)setflags is called only if the flags aren't empty. If the
1313 file is a special file, a warning is printed, because changing
1314 flags implies opening the device. Normally, a special file
1315 should not have any flag... (Debian bug #29775, patch provided
1316 by Abhijit Dasgupta <abhijit@ans.net>).
1317
1318 4. Made possible to dump a file system not mentioned in /etc/fstab.
1319 (Debian bug #11904, patch provided by Eirik Fuller <eirik@netcom.com>).
1320
1321 5. Changed the default behaviour to not create dumpdates
1322 unless 'u' option is specified. Removed the old "debian-patch"
1323 which provided the same thing. (Debian bug #38136, #33818).
1324
1325 6. Removed all those dump*announce, since they were getting old...
1326
1327 7. Added warning messages if dumpdates does not exist and
1328 when an inferior level dump does not exist (except for a level 0
1329 dump).
1330
1331 8. Debugged the glob calls in interactive mode: restore used a
1332 dirent struct which was different from the /usr/include/dirent.h
1333 one (this used to work, is it a glibc2 change?), so none of the
1334 compat glob (which used /usr/include/dirent.h) or the system glob
1335 worked. Restore use now the system dirent (and the system
1336 DT_* constants), which are compatible with BSD ones.
1337
1338 9. Added a configure flag (--with-dumpdatespath) to specify
1339 the location of dumpdates file. By default, it is
1340 /etc/dumpdates.
1341
1342 10. Added the "AUTHOR" and "AVAILABILITY" sections and
1343 included the current date/version in man pages.
1344
1345 11. Corrected the estimation of remaining time when
1346 the operator doesn't change the tapes quickly enough. This
1347 was an old bug, I thought I corrected it, and discovered
1348 that in fact it was corrected in two different places, so
1349 the results canceled each other...
1350
1351 Changes between versions 0.4b5 and 0.4b6 (released October 1, 1999)
1352 ===================================================================
1353
1354 1. Integrated multiple patches from RedHat, Debian and SuSE:
1355
1356 - tweak dump/itime.c to not try to read dumpdates if the 'u' option
1357 isn't specified.
1358 - several fixes in the man pages.
1359 - update the default tape device to /dev/st0.
1360 - many updates for Linux Alpha (byte ordering, size_t etc).
1361 - buffer overruns.
1362 - use environment variable for TMPDIR (instead of /tmp).
1363 - use sigjmp_buf instead of jmp_buf (RedHat bug #3260).
1364 - workaround egcs bug (RedHat bugs #4281 and #2989).
1365 - wire $(OPT) throughout Makefile's.
1366
1367 2. Upgrade the dump revision to 1, making possible to dump filesystems
1368 made with e2fsprogs-1.15 or newer. Nothing seems to break...
1369
1370 3. Fix some compile warnings, prototype all functions.
1371
1372 4. Use glibc err/glob instead of internal compatibility
1373 routines (only if available).
1374
1375 5. Fix a compile error on Linux 2.2.7 / libc5 (5.4.44) (patch provided
1376 by Bernhard Sadlowski <sadlowsk@mathematik.uni-bielefeld.de>).
1377
1378 Changes between versions 0.4b4 and 0.4b5 (released September 22, 1999)
1379 ======================================================================
1380
1381 1. Integrated the changes from FreeBSD-3.1-RELEASE
1382 (mostly bug fixes, buffer overruns, dump has now an "automatic
1383 tape length calculation" flag, dump/restore can use kerberos now
1384 (this is NOT tested), use environment variables for TAPE and
1385 RMT etc.).
1386
1387 2. Integrated three RedHat patches ("glibc", "kernel" and "bread" patches)
1388
1389 3. Corrected a bug in restore when using 'C' option with multi-volumes
1390 tapes (files splited accros two tapes give "size changed" errors
1391 when comparing).
1392
1393 4. Corrected the long standing bug when dumping multiple tapes.
1394 This works for me, needs further testing.
1395
1396 Changes between versions 0.4b3 and 0.4b4 (released January 17, 1997)
1397 ====================================================================
1398
1399 1. Dump now runs correctly on kernels 2.1.x
1400 Fix made by Gerald Peters <gapeters@worldnet.att.net>
1401
1402 Changes between versions 0.4b2 and 0.4b3
1403 ========================================
1404
1405 1. Use realpath() if available
1406
1407 2. Report statistics
1408
1409 Changes between versions 0.4b1 and 0.4b2
1410 ========================================
1411
1412 1. Fixed the bug fix from Greg Lutz (I had made a mistake when integrating
1413 the patch)
1414
1415 2. Fixed restore to make it able to read FreeBSD 2.x dumps again
1416
1417 3. Fixed configure.in to correctly handle --enable-rmt
1418
1419 Changes between versions 0.3 and 0.4b1
1420 ======================================
1421
1422 1. Integrated the changes from 4.4BSD-Lite2
1423
1424 2. Integrated the patches from Debian and Red Hat
1425
1426 3. Portability changes: use the __u32, __u16, __s32, and __s16 types
1427
1428 4. Changed dump to use the Ext2fs library to get block addresses. This
1429 should solve the endianness problem on SparcLinux.
1430
1431 5. Created a configure.in file (shamelessly stolen from the e2fsprogs
1432 distribution's one) to use autoconf
1433
1434 6. Fixed a few minor bugs
1435
1436 Changes between versions 0.2e and 0.2f
1437 ======================================
1438
1439 1. Added the creation of named pipes (fifos) in restore.
1440
1441 2. Added the -N flag in the restore manual page.
1442
1443 3. Added the file kernel.patch which contains the llseek() optimization
1444 patch for 1.2.x kernels.
1445
1446 4. Fixed a bug in the restoration of symbolic links: owner and group were
1447 not restored.
1448
1449 5. Integrated some changes from FreeBSD 2.2.
1450
1451 6. Added a call to ftruncate() after restoring each file to restore
1452 correctly files ending by a hole.
1453
1454 Changes between versions 0.2d and 0.2e
1455 ======================================
1456
1457 1. Fixed a bug in the "set owner/mode" process. Every file was restored
1458 with owner = root (0) and group = root/wheel/whatever (0).
1459
1460 Changes between versions 0.2c and 0.2d
1461 ======================================
1462
1463 1. Dump is now able to backup 2GB+ filesystems.
1464
1465 2. Dump and restore can now be linked as static binaries.
1466
1467 Changes between versions 0.2b and 0.2c
1468 ======================================
1469
1470 1. Fixed a bug when dumping ``slow'' (i.e. normal) symbolic links.
1471
1472 Changes between versions 0.2a and 0.2b
1473 ======================================
1474
1475 1. Really fixed the bug that I should have corrected in 0.2a.
1476
1477 2. Enabled optimization again.
1478
1479 Changes between versions 0.2 and 0.2a
1480 =====================================
1481
1482 1. Disabled the optimization during compilation.
1483
1484 Changes between versions 0.1 and 0.2
1485 ====================================
1486
1487 1. Fixed a bug in fstab.c which caused a null pointer to be stored in
1488 the fs_type field (actually, I modified the file fstab.c to make it
1489 use the mntent functions).
1490
1491 2. Dump and restore now use a 4.3 BSD compatible dump format. Backups
1492 made by dump should be readable by the BSD restore and backups made
1493 by the BSD dump should be readable by restore. Unfortunately, this
1494 means that the dump format has changed between version 0.1 and version
1495 0.2 :-(
1496
1497 3. Dump is now able to backup a subtree, it is no longer limited to whole
1498 filesystems like the BSD version.
1499
1500 4. Dump now uses ext2_llseek() so it is able to backup filesystems bigger
1501 than 2 GB.
1502
1503 Changes between versions 0.0 and 0.1
1504 ====================================
1505
1506 1. Now create links rdump and rrestore during the `make install' step.
1507
1508 2. Linux port specific bugs added to the manual pages
1509
1510 3. Incorrect estimation of the number of tapes blocks fixed when doing
1511 incremental backups.
1512
1513 4. Better ls-like format in restore in interactive mode.