Fix restoration of extended attributes for fifos and device nodes.
[dump.git] / CHANGES
1 $Id: CHANGES,v 1.306 2009/12/03 12:46:29 stelian Exp $
2
3 Changes between versions 0.4b42 and 0.4b43 (released ?????????????)
4 ===================================================================
5
6 1. Fix a bug in dump making impossible to handle large toc files
7 (> 2 GB). Thanks to X DUGi <xdugi@users.sourceforge.net> for
8 reporting the bug (Sourceforge bug #2820629)
9
10 2. Fix 'restore -x' in multi-volume mode, which caused files being
11 spanned on the first and second volumes to be incorrectly extracted.
12 Thanks to Adam Tkac <atkac@redhat.com> for reporting the bug.
13 (there may be several bug reports about this issue, including
14 Sourceforge bug #2117008, RedHat bugs #232415 and #444958).
15
16 3. Remove -lselinux -lsepol from the standard libraries list,
17 they shouldn't be needed if selinux is not enabled in dump
18 (this also fixes a build failure when selinux libraries are
19 not installed).
20
21 4. Fix restoration of extended attributes of fifos and device nodes.
22 Thanks to Igor Zhbanov <izh1979@gmail.com> for the bug report.
23
24 Changes between versions 0.4b41 and 0.4b42 (released June 18, 2009)
25 ===================================================================
26
27 1. Fix printout of the 'Connection to' message in dump/restore.
28 Thanks to Dale Wiles <dwiles@users.sourceforge.net> for the
29 bug report and its fix.
30
31 2. Fix dump -w/-W output for filesystems having the last backup
32 level equal to 0. Thanks to Pascal Bouchareine
33 <kalou@users.sourceforge.net> for reporting the bug.
34
35 3. Enable restore to handle restoring onto a different SELinux
36 policy from the dump. Thanks to Tony Nelson
37 <tony_nelson@users.sourceforge.net> for the patch.
38 (see http://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=189845
39 for details)
40
41 4. Fix EA set failures when restoring immutable files. Thanks
42 to Andrew Kroeger for the patch.
43
44 5. Fix "mode file too large" errors when restoring huge backups.
45 Thanks to Kenneth Porter <shiva@sewingwitch.com> for the bug
46 report.
47
48 6. Add the ability to link against libtinfo library if present,
49 instead of libtermcap or libncurses libraries. Thanks to
50 Adam Tkac <atkac@redhat.com> for the patch.
51
52 7. When comparing a backup, do not consider to be an error the fact
53 that we encounter files created while dump was in progress.
54 Thanks to Kenneth Porter <shiva@sewingwitch.com> for the bug
55 report.
56
57 8. Force '-a' to be the default value when running 'restore -C'.
58 Thanks to Kenneth Porter <shiva@sewingwitch.com> for the bug
59 report.
60
61 9. Many improvements to the 'cron_dump_to_disk' example.
62 Thanks to Aaron S. Hawley <ashawley@users.sourceforge.net> for
63 the patch.
64
65 10. Fix the QFA generation when extended attributes are backuped
66 along with the inodes. Prior to this fix, the entries for
67 some inodes may get corrupted (duplicated entries or incorrect
68 ones), making the QFA file unusable for repositionning in
69 restore. Many thanks to Kenneth Porter <shiva@sewingwitch.com>
70 for the bug report and the testing of the fix.
71
72 11. Fix EA restoration in debug mode (Sourceforge bug #1986216).
73
74 12. Stefan Auracher <star7@onlinehome.de> noticed that the Tower
75 of Hanoi backup strategy documented in the man page wasn't
76 actually based on the Tower of Hanoi algorithm. The man page
77 was updated accordingly.
78
79 13. Fix missing level information from dump output when using
80 the default level (RedHat bug #493635). Thanks to Adam Tkac
81 <atkac@redhat.com> for the patch.
82
83 14. Fix the dump man page to reflect the fact that the default
84 dump level is 0 not 9 (RedHat bug #356121). Thanks to Adam Tkac
85 <atkac@redhat.com> for the patch.
86
87 15. Fixed a few spelling errors in the man pages (RedHat bug #489853).
88 Thanks to Adam Tkac <atkac@redhat.com> for the patch.
89
90 16. Fixed restore man page and program usage which incorrectly stated
91 that -P and -A were both allowed in the same invocation (RedHat
92 bug #490627). Thanks to Adam Tkac <atkac@redhat.com> for the patch.
93
94 17. Use sys_clone under Linux to share I/O contexts between dump
95 processes, thus drastically increasing the performance of dump
96 under CFQ (which is the default I/O scheduler used in a number of
97 distributions). Thanks to Jeff Moyer <jmoyer@redhat.com> for the
98 patch.
99
100 18. Add (preliminary) ext4 support - thanks to libext2fs which does
101 all the job for us. Thanks to Gertjan van Wingerde
102 <gwingerde@gmail.com> for the patch.
103
104 Changes between versions 0.4b40 and 0.4b41 (released January 2, 2006)
105 =====================================================================
106
107 1. Fix restore of dumped Access Control Lists. The previous
108 code for EA works fine, but ACLs needed conversion from the
109 ext2/3 disk format to posix_acl format before restoring.
110
111 2. Fix some issues with restoration of EA on big endian
112 platforms.
113
114 3. Fix restore when the symtab is over 2GB in size.
115
116 4. Made the directory hash indexing an optional feature,
117 accessible by the '-H' option of restore, and disable it
118 by default.
119
120 5. Fixed dump to not include extended attributes information
121 in the toc (archive) file which confused restore -t.
122
123 Changes between versions 0.4b39 and 0.4b40 (released May 2, 2005)
124 =================================================================
125
126 1. Changed restore to emit warnings (instead of emitting a fatal
127 error) if a file (or a directory) is unavailable for a
128 comparision (if the user doesn't have the necessary permissions
129 to access it for example). Thanks to Kenneth Porter
130 <shiva@sewingwitch.com> for the bug report.
131
132 2. Re-done the 'do not save directory entries to non-dumped inodes
133 (excluded from dump)' feature. The previous implementation
134 worked well for excluded directories but not for regular files.
135 Thanks to Kenneth Porter <shiva@sewingwitch.com> for the bug
136 report.
137
138 3. Fixed a bug in dump where the tape size was miscalculated when
139 the user used -d/-s to specify the tape characteristics. Thanks
140 to Philip Goisman <goisman@physics.arizona.edu> for reporting
141 the bug.
142
143 4. Fixed another bug introduced in restore with the hashtree
144 implementation. This one caused restore to stop saying
145 "removenode: non-empty directory" in some cases.
146
147 5. Added support for dumping and restoring ext2/3 extended
148 attributes (EA), like the access control lists (ACL) or
149 the security labels used by SELinux.
150
151 Changes between versions 0.4b38 and 0.4b39 (released January 21, 2005)
152 ======================================================================
153
154 1. The newly added dump_on_cd_3 example was buggy, replace it
155 with an updated version from Andrew Basterfield
156 <bob@cemetery.homeunix.org>.
157
158 2. Made restore to chdir() back into the initial directory when
159 dumping core while aborting a comparision operation. The
160 previous behaviour was to write the corefile at the root of
161 the directory being compared, which could very well be
162 read only and preventing the corefile generation. Thanks
163 to Kenneth Porter <shiva@sewingwitch.com> for the bug report.
164
165 3. Silenced the failure to call fgetflags() when comparing an
166 entry which has no ext2 attributes (as in lsattr()).
167
168 4. Fix a brown paper bug in restore -C which broke restore and
169 caused modifications on the filesystem being compared
170 (directories containing a file with the same name as the
171 directory get renamed to RSTTMP...). Thanks to Kenneth Porter
172 <shiva@sewingwitch.com> for finding the bug and helping me
173 reproduce it.
174
175 5. Made restore -C force the -N flag (no writing allowed on
176 the disk) in order to prevent more bugs like the above one.
177
178 Changes between versions 0.4b37 and 0.4b38 (released January 7, 2005)
179 =====================================================================
180
181 1. Fix a couple of troff syntax bugs in the man pages.
182 Thanks to Eric Raymond <esr@thyrsus.com> for the patch.
183
184 2. Made restore use either libncurses or libtermcap, depending
185 on which one is available at configure time.
186
187 3. Fixed restore negative size display bug when comparing a
188 dump containing files over 2GB. Thanks to Steve Bonds
189 <sbonds@users.sourceforge.net> for the bug report.
190
191 4. Do not save directory entries to non-dumped inodes
192 (excluded from dump). This will eliminate the 'missing
193 file' warnings when doing 'restore -C'.
194
195 5. Fix dump crash when backuping a huge (2TB) filesystem,
196 due to a bogus calculation on the inode map size.
197 Thanks to Kevin B. Haines <K.B.Haines@rl.ac.uk> for
198 submitting the bug and testing the fix.
199
200 6. Fix a problem in restore where the final \0 in the symbolic
201 link names could have been lost, generating corrupt filenames.
202 Thanks to Kyle Wilson <kyle.wilson@amd.com> for reporting the
203 bug.
204
205 7. Implemented a hash list for the directory names in restore.
206 The linear list used before caused problems in interactive
207 restores when dealing with directories having thousands of
208 entries. Thanks to Brian Ristuccia <bristuccia@starentnetworks.com>
209 for reporting the bug.
210
211 8. Improved restore -C, this time including the directory
212 attributes into the comparision.
213
214 9. Made restore understand tapes containing EA/ACLs (which will
215 be dumped by the next version of dump). In this version
216 extended attributes on the tape are ignored, for full EA/ACL
217 support wait for the next version or try the experimental EA
218 patch.
219
220 Changes between versions 0.4b36 and 0.4b37 (released July 7, 2004)
221 ==================================================================
222
223 1. Added the --enable-staticz configure option which enables
224 dump and restore to be linked with static versions of
225 libz and libbz2 (and dynamic versions of all other libraries).
226 This will make Debian users happy, because libz and libbz2
227 were the only needed libraries living in /usr, all the
228 others live in /lib. In case of system emergency, it is
229 better not to have to rely on an extra filesystem.
230 Thanks to Bdale Garbee <bdale@gag.com> for the suggestion.
231
232 2. Fix compilation on (at least the Linux Debian port to) AMD64.
233 (<ext2fs/ext2_types.h> defines some types (__s64 and __u64)
234 that are also defined by <linux/types.h> (<asm/types.h>) and
235 they conflict).
236
237 3. Make dump's reading of the dumpdates file a bit more robust,
238 preventing dump from crashing when the dumpdates file has
239 been modified by hand.
240
241 4. Fixed some offset calculations in dump code which could
242 lead to "bread lseek errors" on large filesystems. Thanks
243 to Bruce Lowekamp <lowekamp@users.sourceforge.net> for
244 reporting this bug and debugging the issue.
245
246 5. Made dump use the blkid library when searching for devices
247 by label or uuid instead of dump's own routines.
248
249 6. Corrected a bug in dump where a wrong LABEL=... line in
250 /etc/fstab could prevent dump from dumping unrelated
251 filesystems. Thanks to Bruce Lowekamp
252 <lowekamp@users.sourceforge.net> for reporting the bug.
253
254 Changes between versions 0.4b35 and 0.4b36 (released April 21, 2004)
255 ====================================================================
256
257 1. Fixed dump compilation with old gcc versions. Thanks to
258 Mike Castle <dalgoda@users.sourceforge.net> for the patch.
259
260 2. Fixed some warnings (howmany, roundup, powerof2 redefined)
261 when compiling against a recent glibc version.
262
263 3. Fixed a bug in restore preventing the read of a dump tape
264 written with Solaris 7 ufsdump. Thanks to Patrick Higgins
265 <phiggins@transzap.com> for reporting the bug and providing
266 the test case.
267
268 4. Changed dump to enable the creation of volumes of different
269 sizes in a single run (make -B accept a list of values).
270 Patch contributed by Florian Zumbiehl <florz@gmx.de>.
271
272 5. Use the glibc provided minor() and major() macros instead
273 of our own bitmask implementation. This should be safe for
274 when the major/minor namespace will migrate to 32 bits.
275 Thanks to Zhang Jun <zhangjun@nanjing-fnst.com> for reporting
276 the bug.
277
278 6. Made explicit in the dump man page that dump will not create
279 a remote file, it will only write to an already existing one.
280
281 7. Another try at making size estimates better again.
282
283 8. Put back the inconditional running of the end-of-tape
284 script which was changed in 0.434 to be run only when -M
285 or multiple -f were NOT specified. Some users rely on this
286 feature even when it is combined with -M/-f.
287
288 9. Fixed restore when restoring huge backups (where rstdir...
289 temporary files are over 2GB). Thanks to
290 Raphael Jaffey <rjaffey@artic.edu> for reporting this,
291 debugging the issue and making the patch.
292
293 10. Made restore understand FreeBSD UFS2 tapes. Thanks to
294 David <vrtin@users.sourceforge.net> for submitting the bug
295 and providing a test case.
296
297 11. Made dump work with an arbitrary integer level (no more 10
298 levels only limitation). Thanks to Matthew
299 <msvincen@midway.uchicago.edu> for the patch.
300
301 Changes between versions 0.4b34 and 0.4b35 (released December 21, 2003)
302 =======================================================================
303
304 1. Added a note in the dump man page saying that the default
305 blocksize can be 32 if -d is used with a high density argument.
306 Thanks to Antonios Christofides <A.Christofides@itia.ntua.gr>
307 for the patch.
308
309 2. Fixed configure to correctly understand CPPFLAGS, CFLAGS,
310 LDFLAGS environment variables. Thanks to Arcady Genkin
311 <antipode@users.sourceforge.net> for reporting the bug.
312
313 3. Made -e/-E options of dump accept an unlimited number of inodes
314 to be excluded, rather than a hardcoded maximum. Thanks to
315 Dietrich Rothe <d-rothe@users.sourceforge.net> for the patch.
316
317 4. Updated the autoconf system to 2.50. Forced the -D_BSD_SOURCE
318 and -D_USE_BSD_SIGNAL defines in configure in order to solve
319 64bit build problems because quad_t is redifined with a
320 different signature. Thanks to Mike Harris <mharris@redhat.com>
321 for reporting this bug.
322
323 5. Made restore build on Solaris, making possible to
324 restore Linux's "enhanced" tapes. Thanks to Uwe Gohlke
325 <uwe@ugsoft.de> for the patch.
326
327 6. Made an extension in the dump tape format capable of saving
328 MacOSX specific inode extensions. Uwe Gohlke <uwe@ugsoft.de>
329 wrote the extension and contributed the restore code back
330 into this codebase. The same extension mechanism will be
331 used in the future to save ACLs...
332
333 7. Made rmt work correctly with regard to QFA and local/remote
334 files and tapes. The remote access will however work only
335 when the dump provided rmt version is used. If you want to
336 use another rmt server, please do not use the QFA feature.
337 Thanks to Uwe Gohlke <uwe@ugsoft.de> for the patch.
338
339 Changes between versions 0.4b33 and 0.4b34 (released April 18, 2003)
340 ====================================================================
341
342 1. Fixed the INSTALL file to reflect the actual install paths.
343 Thanks to David Ronis <ronis@ronispc.chem.mcgill.ca> for
344 reporting the bug.
345
346 2. Fixed the configure script to only check for headers presence
347 instead of trying to compile them. This should fix issues
348 with old build environments. Thanks to Kari Lempiainen
349 <kari@funky.pp.fi> for reporting the bug.
350
351 3. Fixed restore to correctly ignore sockets when comparing
352 a dump (as socket cannot be properly restored anyway). Thanks
353 to Gunther Reiszig <gunther@mit.edu> for reporting the bug.
354
355 4. Fixed restore to correctly access the archive file (-A argument)
356 even when using a remote tape. Thanks to Fabrice Bellet
357 <fabrice@bellet.info> for reporting the bug.
358
359 5. Fixed (again) handling of long (largefile) seeks in rmt.
360 Thanks to Fabrice Bellet <fabrice@bellet.info> for reporting
361 the bug.
362
363 6. Fixed restore corner case when dealing with large block sizes
364 dump is able to write now (-b 1024). Thanks to Fabrice Bellet
365 <fabrice@bellet.info> for reporting the bug.
366
367 7. Fixed a bug preventing dump to access a filesystem having
368 a label exactly 16 bytes in length. Thanks to <minduni@ti-edu.ch>
369 for reporting the bug.
370
371 8. Made dump store dump dates using explicit timezones, fixing a
372 problem with incremental dumps if the timezone is modified
373 between the dumps. Thanks to Philippe Troin <phil@fifi.org> for
374 the bug report and the patch.
375
376 9. Fixed a bug encountered when dumping individual files (not full
377 filesystems or directories) and dangling symbolic links happen
378 to be in the list of files. For as far as dump is concerned,
379 dangling symbolic links are allowed, and are dumped as is.
380 Thanks to Jin-su Ahn <jsahn@ee.snu.ac.kr> for reporting the
381 bug and providing the fix.
382
383 10. Fixed open and creation modes and permissions for QFA and
384 table-of-contents files in dump and restore. Thanks to
385 Philippe Troin <phil@fifi.org> for the patch.
386
387 11. Fixed the archive file descriptor handling enabling it to be 0.
388 This can happen in some cases when shell redirections are used.
389 Thanks to Philippe Troin <phil@fifi.org> for the patch.
390
391 12. Delayed the opening of archive file until after suid had been
392 dropped (fixing a possible security issue if dump is suid).
393 Thanks to Philippe Troin <phil@fifi.org> for the patch.
394
395 13. Fixed the 'S' command handling in the rmt client part.
396 Thanks to Philippe Troin <phil@fifi.org> for the patch.
397
398 14. Modified the end-of-tape script handling to print out statistics
399 (and stop the timer) before launching the eot script. Also, the eot
400 script does not get run anymore when using -M (which makes sense) or
401 when multiple tapes are listed on the command line
402 (-f tape0,tape1,tapen) (which also makes sense).
403 Thanks to Philippe Troin <phil@fifi.org> for the patch.
404
405 15. Relicensed dump/restore under the 'revised' BSD license, as per
406 ftp://ftp.cs.berkeley.edu/ucb/4bsd/README.Impt.License.Change.
407
408 16. Added LZO compression to dump. This new compression method has
409 the advantage of being super fast, thus not killing tape streaming
410 on slow machines. Thanks to Helmut Jarausch
411 <jarausch@igpm.rwth-aachen.de> for the patch and to
412 Markus Oberhumer <markus@oberhumer.com> for giving special permission
413 to include his miniLZO project (GPL licensed) in dump/restore.
414
415 17. Some small buffer overruns fixes in rmt. Thanks to Antonomasia
416 <ant@notatla.demon.co.uk> for reporting the bugs.
417
418 18. Added a special rmt version which can do encryption when writing
419 to tape. Read examples/encrypted_rmt/README for details on
420 how to enable and configure it. Thanks to Ken Lalonde
421 <ken@globalremit.com> for the patch.
422
423 19. Made dump work with 2.5 kernel end of tape early warning semantics.
424 Thanks to Kai Makisara <Kai.Makisara@kolumbus.fi> for the patch.
425
426 20. Fixed a bug which caused dump -w|-W not to work anymore, because
427 the fs_freq and fs_passno fields in /etc/mtab are always set
428 to 0 0. Thanks to Trent Piepho <xyzzy@speakeasy.org> for
429 reporting the bug.
430
431 Changes between versions 0.4b32 and 0.4b33 (released February 10, 2003)
432 =======================================================================
433
434 1. Added a note in the restore man page clarifying the question
435 'set the permissions on the current directory ?' asked by
436 restore at the end of treatment in -i and -x modes.
437
438 2. Fixed the endianess issues when reading compressed tapes.
439 Thanks to Dark Force <daq4th@users.sourceforge.net> for
440 reporting this bug and providing test cases.
441
442 3. Fixed the "ACL won't be dumped" warning message (which showed
443 an extra, unrelated error message). Thanks to Dragan Krnic
444 <dkrnic@lycos.com> for reporting this bug.
445
446 4. Made dump look first into /etc/mtab, then into /etc/fstab
447 when searching for filesystem paths. Also fixed some problems
448 caused by binding mounts or multiple block device mounts.
449 Thanks to Matus Uhlar <uhlar@fantomas.sk>, Elliott Mitchell
450 <ehem@m5p.com>, Greg Edwards <gedwards@users.sourceforge.net>,
451 Brian Hoy <brian.hoy@opus.co.nz>. (fixes Debian bugs #147086
452 and #160305, Sourceforge bugs #618699 and #679832).
453
454 5. Made dump's -I option accept the value '0' meaning all the
455 read errors will be ignored. This can be useful when running
456 dump from unattended sessions (like cron jobs). Thanks to
457 John I Wang <jiwang@users.sourceforge.net> for the suggestion.
458
459 6. Fixed the output of dump to indicate 'blocks' instead of
460 'tape blocks' in the various messages (blocks are always
461 1 Kilobyte, tape blocks are 1 BK * '-b' argument), and
462 made it clearly print the current blocksize at the start of
463 a dump. Thanks to Michal Szymanski <msz@astrouw.edu.pl> for
464 the suggestions.
465
466 7. Made rmt understand long (largefiles) seeks.
467
468 8. Fixed build with very old versions of libext2fs, where
469 EXT2_FT_* constants were undefined.
470
471 9. Made dump accept the dumpdates path on the command line
472 (-D file option) instead of using only the hardcoded one.
473 Thanks to Piete Brooks <pb22@users.sourceforge.net> for the
474 suggestion.
475
476 10. Enabled rmt, LFS, readline, QFA options by default in
477 ./configure. Updated the configure process (new versions
478 of config.guess, config.sub etc).
479
480 Changes between versions 0.4b31 and 0.4b32 (released November 15, 2002)
481 =======================================================================
482
483 1. Changed dump to use fcntl(F_SETLK) style locking instead
484 of flock() when locking the dumpdates file. With the old
485 locking scheme, a local user having read rights on the
486 dumpdates file could be able to do a Denial of Service attack
487 on dump. In order to lock the dumpdates file with the new
488 scheme, the user would need to have write access on the file.
489 Thanks to Richard Johnson <Richard.Johnson3@ey.com> for
490 reporting the bug (originally a bugtraq post).
491
492 2. Fixed interactive 'ls' which caused spurious errors warnings
493 about 'undefined filetypes' detected. Thanks to Jorgen Ostling
494 <jorgen_ostling@users.sourceforge.net> for reporting this
495 bug.
496
497 3. Fixed dump's estimate when dealing with sparse inodes.
498
499 4. Modified dump to allow setting a blocksize bigger than 32kB
500 (raised the limit to 1024kB), because newer hardware needs
501 this for proper operation. Thanks to Dirk Traenapp
502 <dtraenapp@users.sourceforge.net> for reporting this.
503
504 5. Fixed a bug causing Dump to stop and report an error if an
505 inode number in the exclude file was followed by some amount
506 of whitespace. Thanks to Jeffrey Sofferin
507 <sofferin@users.sourceforge.net> for reporting this bug.
508
509 6. Fixed a bug which caused restore, in some particular cases,
510 to ask some 'scary' questions and leave a bunch of RSTTMP
511 directories behind when restoring incremental tapes. Thanks
512 to Philippe Troin <phil@fifi.org> for reporting this bug and
513 providing the test cases.
514
515 7. Changed the wording when inodes are excluded from dump:
516 replaced 'Added inode 7 to exclude list' with
517 'Excluding inode 7 (resize inode) from dump', as suggested
518 by Elliott Mitchell <ehem@m5p.com> in a Debian bug report.
519
520 Changes between versions 0.4b30 and 0.4b31 (released July 30, 2002)
521 ===================================================================
522
523 1. Fixed rmt open flags transmission (GNU's symbolic syntax over
524 rmt) which I broke in 0.4b29. Thanks to Eros Albertazzi
525 <eros@lamel.bo.cnr.it> for reporting the bug.
526
527 Changes between versions 0.4b29 and 0.4b30 (released July 25, 2002)
528 ===================================================================
529
530 1. Made dump print out the ext2 logical block number in case of a read
531 error, which can be given as an argument to debugfs' ncheck command,
532 in order to find out the affected inode. Added note about this
533 usage in dump's man page.
534
535 2. Fixed a problem in restore when reading tapes written on big
536 endian machines with very old versions of dump. The patch was
537 contributed by George Helffrich <george@geology.bristol.ac.uk>.
538
539 3. Fixed the tape length calculation when using large tapes
540 and compression. Thanks to Georg Lippold
541 <g_lippold@sourceforge.net> for reporting the bug.
542
543 4. Added a new set of examples in dump_on_cd_2 directory, based
544 on dump_on_cd examples but somewhat enhanced, supporting DVD
545 media, and localized in english and german. Thanks to
546 Georg Lippold <g_lippold@sourceforge.net> for the new scripts.
547
548 5. Made dump save 32 bit UID/GID. Internally, this was achieved
549 by switching from the old BSD inode format to the new BSD
550 inode format, which means that the tape format was changed.
551 However, since all restore versions out there should
552 transparently support both inode formats, the change should
553 have no side effects. Thanks to John Yu <jky@cs.bu.edu> for
554 reporting the bug.
555
556 6. Fixed a lot of warnings in the code shown when compiling
557 with 'gcc -W'. Thanks to Matthias Andree
558 <matthias.andree@stud.uni-dortmund.de> for reporting this.
559
560 7. Fixed a small markup bug in the dump man page. Thanks to
561 Eric S. Raymond <esr@minx.thyrsus.com> for submitting the
562 patch.
563
564 8. Rewrote entirely the man pages using the tmac.an macro
565 package (Linux man page format) instead of the original BSD
566 format. They should be now cleaner and easier to modify.
567
568 Changes between versions 0.4b28 and 0.4b29 (released June 8, 2002)
569 ==================================================================
570
571 1. Fixed a problem in the rmt ioctl command, where ioctl's issued from
572 non Linux clients were misinterpreted. The description of the problem
573 (incompatible numbering in Linux mtio opcodes) is documented at
574 ftp://ftp.fokus.gmd.de/pub/unix/star/README.mtio . Thanks to
575 Jörg Schilling <schilling@fokus.gmd.de> for reporting this bug and
576 providing an excellent, cross-platform replacement for rmt in his
577 star package.
578
579 2. Fixed a bug reported by Andy Mentges <amentges@jumpline.com>
580 which caused restore to fail when the inode maps were not entirely
581 contained on the first volume (when using really small volumes or
582 when dumping a huge number of inodes).
583
584 3. Fixed a problem in dump, where files in subdirectories of directories
585 that have the nodump flag set (or being excluded with -e/-E) were
586 sometimes incorrectly dumped (instead of being excluded). The original
587 fix is from David C Lawrence <tale@dd.org> in the FreeBSD version
588 of dump: http://www.FreeBSD.org/cgi/query-pr.cgi?pr=32414
589 Thanks to Ted Grzesik <tedgyz@roostme.com> for reporting the bug and
590 help testing the patch.
591
592 4. Added some example scripts from Gerd Bavendiek <bav@epost.de>
593 which makes one able to pipe the output of dump, by the net, to
594 a remote CD-burner server.
595
596 5. Made dump use O_CREAT|O_TRUNC both locally and remotely (over rmt),
597 and use GNU's symbolic syntax over rmt instead of numerical values
598 to assure multiple platform compatibility.
599
600 6. Documented the -d option in restore.
601
602 7. Added a -v (verbose) mode to dump. For now it just prints the number
603 of the inode being dumped, but this could evolve in future versions
604 to include interesting debugging output.
605
606 8. Added a -o flag to restore, which automatically restores the current
607 directory permissions in -i or -x mode, without asking the operator
608 whether to do so. Patch submitted by Tony Clayton <tonyc@e-smith.com>
609 and Peter Samuel <peters@e-smith.com>.
610
611 Changes between versions 0.4b27 and 0.4b28 (released April 12, 2002)
612 ====================================================================
613
614 1. Fixed a bug in the QFA feature which could have caused
615 a bad tape positionning on restore (causing extra delays in
616 restore). Credits go to Uwe Gohlke <uwe@ugsoft.de>.
617
618 2. Added a small note in the dump man page specifying that
619 there should be no white space between the option letter and
620 the -j or -z parameter, thanks to Kenneth Porter <shiva@well.com>
621
622 3. Made restore work with older versions of the readline library,
623 by checking in configure for several extended symbols. Restore
624 can now be compiled with a readline as old as the 2.0 release
625 (though it may be a good idea to upgrade it to a more recent
626 version...). Thanks to Andrew Donkin <ard@waikato.ac.nz> for
627 reporting the build failures.
628
629 4. Fixed a performance problem with the QFA file creation in
630 dump, which made unnecessary seeks on the tape slowing down
631 the dump. Thanks to Andrew Donkin <ard@waikato.ac.nz> for
632 reporting this issue.
633
634 5. Removed the inclusion of some kernel headers in the dump
635 source, which prevented the compile in some kernel/glibc
636 headers/architecture combination. Thanks to Bdale Garbee
637 <bdale@gag.com> for reporting the bug.
638
639 6. Added the appropriate error message when dump fails to
640 open the output file for writing. Thanks to Amith Varghese
641 <amithv@yahoo.com> for reporting this bug.
642
643 7. Made restore able to understand large Solaris ufsdump tapes
644 (containing inodes bigger than 4194304). Sun have introduced
645 an "extension" to the dump tape format when dealing with
646 those inodes, which was uncorrectly handled by Linux restore.
647 Thanks to Uwe Gohlke <uwe@ugsoft.de> for reporting the bug and
648 providing a test case.
649
650 8. Added the -m parameter to dump which optimises the output for
651 inodes having been changed but not modified since the last dump
652 ('changed' and 'modified' have the meaning defined in stat(2)).
653 For those inodes, dump will save only the metadata, instead of
654 saving the entire inode contents. Inodes which are either
655 directories or have been modified since the last dump are saved
656 in a regular way. Uses of this flag must be consistent, meaning
657 that either every dump in an incremental dump set have the flag,
658 or no one has it.
659
660 Changes between versions 0.4b26 and 0.4b27 (released February 15, 2002)
661 =======================================================================
662
663 1. Fixed behaviour of dump when exceeding resource limits
664 (SIGXFSZ treatment).
665
666 2. Added the -L flag to restore to allow the user to specify a
667 maximal allowed number of miscompares when using restore
668 with the -C option to check the backup.
669
670 3. Detailed the manual entry for the -N option of restore.
671
672 4. Added the -a flag to restore to make able doing unattended
673 restores in -i or -x mode (automatically walks through the
674 multiple dump volumes).
675
676 5. Extended the QFA mode to work with local files and/or
677 remote tapes and files. This way, restore can know in advance
678 the tape number and the offset for the inodes to extract and
679 can minimize the extraction time by seeking directly to the
680 good tape/offset.
681
682 6. Added the -A <archive> option to both dump and restore,
683 which makes dump to archive a dump table-of-contents in
684 the specified file to be used by restore to determine
685 whether a file is in the dump file that is being restored.
686 (the archive file syntax is also compatible with the
687 Solaris ufsdump generated one).
688
689 7. Small fix in restore making it able to read some (broken ?)
690 Solaris ufsdump tapes.
691
692 8. Fixed dump to correctly recognise the root filesystem when using
693 ext2 disk labels (LABEL=/). Thanks to John Yu <jky@it.bu.edu>
694 for reporting this bug.
695
696 9. Added the -P <file> option to restore to create a
697 Quick File Access file from an already made dump. Patch
698 contributed by Uwe Gohlke <uwe@ugsoft.de>.
699
700 10. Made restore compile and run on Solaris, making it a
701 possible replacement for the standard ufsrestore. Port was
702 contributed by Uwe Gohlke <uwe@ugsoft.de>.
703
704 Changes between versions 0.4b25 and 0.4b26 (released January 7, 2002)
705 =====================================================================
706
707 1. Added a set of backup scripts from Eugenio Diaz
708 <getnito@yahoo.com> in the examples section. It features
709 automatic (cron based) full and incremental dumping of
710 several filesystems on a separate filesystem.
711
712 2. Fixed a off-by-one miscalculation which disabled dumping
713 a one letter subdirectory. Thanks to Chris Siebenmann
714 <cks@utcc.utoronto.ca> for reporting the bug.
715
716 3. Fixed several restore bugs occuring when trying to
717 restore some missing files on the tape. Thanks to Chris
718 Siebenmann <cks@utcc.utoronto.ca> for reporting the bug.
719
720 4. Fixed --with-ldopts configure argument passing, installing from
721 a separate object directory, makefile cleanups contributed
722 by <splite@purdue.edu>.
723
724 5. Fix a bug which could caused, in some conditions, the highest
725 number inode of a filesystem, to not be dumped. Many thanks
726 to Chris Siebenmann <cks@utcc.utoronto.ca> for helping me
727 chase this bug.
728
729 Changes between versions 0.4b24 and 0.4b25 (released November 17, 2001)
730 =======================================================================
731
732 1. Added a mini howto from Patrick Walsh in the examples
733 directory.
734
735 2. Minor man pages syntax corrections. Thanks to
736 Chris Danis <screechco@home.com> for reporting the bugs.
737
738 3. Added a script from David B. Peterson <dave@toppledwagon.com>
739 to the examples section. It features dumping several
740 filesystems to a remote tape drive upon ssh.
741
742 4. Added a patch provided by Richard Jones <rich@annexia.org>
743 which allows BRADEMAX (number of read errors tolerated by
744 dump) to be adjusted using the -I option.
745
746 5. Fixed a bug which disabled doing "restore -C -f -". Thanks
747 to Clemens Stiglechner <a9401816@unet.univie.ac.at> for the
748 patch.
749
750 6. Add the -l option to restore to specify if, when doing a
751 remote restore, the file used to access the backup is a
752 regular file (the defaults being a tape drive). Restore needs
753 to know this information when reading a remote compressed
754 dump. Previously, this information was autodetected, but
755 the autodetection code fails (with ioctl: Inappropriate ioctl
756 for device) when using a non Linux remote box. Thanks to
757 many users and especially to Eros Albertazzi
758 <eros@lamel.bo.cnr.it> for reporting this.
759
760 7. Found a workaroung for the dump deadlock problem (3 childs
761 stuck in pause(), father in read()). The workaround seems
762 to work for me and several beta-testers. If it doesn't work
763 for you, please report back.
764
765 8. Updated the RPM spec file (BuildPrereq, URL etc).
766
767 Changes between versions 0.4b23 and 0.4b24 (released September 12, 2001)
768 ========================================================================
769
770 1. Fixed the permissions of a newly created QFA file by dump.
771
772 2. Cleaned up the source of dump (the external variables
773 definition was a complete mess, making possible to have
774 objects overlap).
775
776 3. Fixed restore to use the full tape volume path when doing
777 a compare (since it changes the working directory to the
778 filesystem being compared in the process).
779
780 4. Added the -q option to dump which makes dump abort
781 whenever it needs operator attention. It should be
782 easier to use dump in scripts with this option.
783
784 5. Detect the use of incompatible options to dump and
785 refuse them (like -a and -B options together).
786
787 6. Added bzip2 compression to dump/restore (use option -j level
788 to select it). Note that you will need a restore version
789 >= 0.4b24 in order to restore a bzip2 compressed dump.
790 The same warning as for the zlib compression applies:
791 the tape format of a bzip2 dump is not compatible with the
792 original BSD tape format.
793
794 7. Fixed a overflow problem in dump corrupting the dump when
795 very large files were encountered. Thanks to Vladimir Nadvornik
796 for the bug report.
797
798 8. Added a ioctl(BLKFLSBUF) in dump which should flush the
799 kernel buffer/page cache before starting the dump, helping
800 a bit those who use dump on mounted filesystems. Thanks to
801 John Yu <jky@it.bu.edu> and to Theodore T'so <tytso@mit.edu>
802 for this suggestion.
803
804 9. Updated the RPM spec file following the RedHat changes
805 (dynamically linked binaries now in /usr/sbin etc).
806
807 10. Added a patch from Helmut Jarausch <jarausch@igpm.rwth-aachen.de>
808 which enables restore to recognise multi volume compressed dumps
809 done on CD. Included his scripts for dump (which pipe the dump
810 output directly into cdrecord) and restore. There is now
811 possible to to backups to CD on the fly!
812
813 Changes between versions 0.4b22 and 0.4b23 (released July 20, 2001)
814 ===================================================================
815
816 1. Fixed a buffer overflow in restore/tape.c. Patch provided by
817 Marcus Meissner (Caldera International Security Dept.).
818
819 2. Implement the Sun rmt extended protocol. Patch submitted
820 by Ian Gordon <iangordon@users.sourceforge.net>.
821
822 3. Check for the e2fsprogs header <ext2fs/ext2_fs.h> instead of
823 the linux kernel header. This ensures that dump always has the
824 latest version of this file. Patch submitted by
825 Andreas Dilger <adilger@turbolinux.com>.
826
827 4. Report any filesystem present in either /etc/fstab with a
828 positive passno or /etc/dumpdates in dump -w output.
829 Patch submitted by Andreas Dilger <adilger@turbolinux.com>.
830
831 5. Fixed the looping problem in dump introduced in the
832 previous version.
833
834 6. Changed the -B option of dump to limit the size of
835 _compressed_ output per volume if compression is on.
836 Patch contributed by Helmut Jarausch
837 <jarausch@igpm.rwth-aachen.de>. Note however that, since
838 it is impossible to predict the size of the compressed
839 data before doing the compression, the -B limit is a bit
840 conservative.
841
842 7. Fixed a bug in reading the operator typed file/tape path for
843 the next volume in restore.
844
845 8. Implemented a "-F script" option to restore which permits the
846 user to specify a script which will be launched at the
847 beginning of each tape, useful for automatic programming of
848 tape changers for example. See the restore man page for the
849 script parameters and return codes.
850
851 9. Small fix for the QFA routines provided by Uwe Gohlke
852 <uwe@ugsoft.de>, and some recommendations for QFA uses in
853 the man pages.
854
855 10. Fixed the multivolume restoring where making a mistake
856 to the 'Mount next tape' prompt caused several blocks to
857 be lost.
858
859 11. Enhanced the -e option of dump to take as a parameter a
860 comma separated list of inode numbers.
861
862 12. Added the -E option to dump which specify a file containing
863 inode numbers to exclude from the dump.
864
865 13. Fixed the compressed multi-volume dump + restore.
866
867 Changes between versions 0.4b21 and 0.4b22 (released May 12, 2001)
868 ==================================================================
869
870 1. Made dump report the number of blocks dumped per volume.
871 Thanks to Kenneth Porter <shiva@well.com> for the suggestion.
872
873 2. Fix a bug in dump where waiting too much at the 'change volumes'
874 question caused the volume to be rewritten. Thanks to
875 Graham Reed <greed@users.sourceforge.net> for reporting the
876 bug and providing a patch.
877
878 3. Added a compression option to dump, contributed by Jerry
879 Peters <gapeters@worldnet.att.net>.
880
881 WARNING: the compression works only on files/pipes and
882 drives supporting variable block size.
883
884 WARNING2: tapes created using the compression feature are
885 incompatible with the standard dump tape format, and a
886 version of restore >= 0.4b22 will be needed for extraction.
887
888 4. Fixed some compilation problems for glibc 2.2.2 and 64 bit
889 architectures. Thanks to Paul Slootman <paul@debian.org> for
890 the patch and to Bdale Garbee <bdale@gag.com> for forwarding
891 it upstream.
892
893 5. Many cleanups (CPP defines, const char warnings, check of
894 ext2fs COMPAT flags, time_t cleanups, added libext2 version
895 in dump usage text) by Andreas Dilger <adilger@turbolinux.com>.
896
897 6. Made --prefix option work in configure. All the install path
898 are now based on the configure parameters.
899
900 7. Added the Quick File Access mode in dump/restore, contributed
901 by Uwe Gohlke <uwe@ugsoft.de>. In this mode, dump stores in
902 a file tape position for each inode, and this file is used by
903 restore (if called with parameter Q and the filename)
904 to directly position the tape at the file restore is currently
905 working on. This saves hours when restoring single files from
906 large backups, saves the tapes and the drive's head. Use
907 --enable-qfa option of configure to compile in the QFA support.
908
909 8. Added the possibility to dump several files and directories
910 in a single invocation of dump. Thanks to Uwe Gohlke
911 <uwe@ugsoft.de> for implementing this option.
912
913 9. Fixed the dumping and restoring of files with holes
914 (files ending with holes didn't get dumped or restored
915 correctly).
916
917 10. Fixed a socket descriptor leak in dump, which leaved opened
918 3 file descriptors per dump process (and there is one dump
919 process per tape).
920
921 11. Fixed dump large file system support, by explicit use of
922 open64/lseek64/etc functions (explicit use needed because
923 e2fsprogs libraries don't behave well when compiled with
924 FILE_OFFSET_BITS=64).
925
926 Changes between versions 0.4b20 and 0.4b21 (released January 13, 2001)
927 ======================================================================
928
929 1. Fixed some bugs in the dump -w|-W logic introduced by
930 the previous version. Thanks to Andreas Dilger
931 <adilger@turbolinux.com> for his help on this one.
932
933 2. Fixed again a compilation problem when using old e2fs
934 headers (filesystem label related). Thanks to many users
935 who reported this stupid error.
936
937 3. Fixed a build problem on old lib5 based systems dealing with
938 _PATH_MNTTAB being undefined. Thanks to John Adams
939 <johna@onevista.com> for reporting the bug.
940
941 4. Improved the error detection in restore when creating
942 the temporary files in TMPDIR. Now restore will corectly
943 report a 'no space left on device' error instead of
944 strange errors which could imply an archive corruption.
945 Thanks to Gabriel Somlo <somlo@cs.colostate.edu> and
946 bgjenero <bgjenero@sympatico.ca> for reporting the bug.
947
948 5. Added the throughput information in dump's progression
949 messages. Thanks to Andreas Dilger <adilger@turbolinux.com>
950 for the patch.
951
952 6. Use libext2fs's inode scanning routines, which are particularly
953 robust when dealing with errors (especially when having some
954 bad blocks on the filesystem). Many thanks to Colin
955 <colin@horizon.com> for reporting the bug, investigating
956 the issues, finding a workaround, writing the patch and
957 fully testing it... (of course, if this patch will break
958 anything, he is to blame for :-)).
959
960 7. Made dump and restore LFS aware. Dump can dump a filesystem
961 containing large files, generate a large file on output and
962 restore can restore them. This feature is not enabled by
963 default, use --enable-largefile option of configure to enable
964 it (you need to have a LFS aware glibc though). Thanks to
965 Andrea Arcangeli <andrea@suse.de> for submitting the patch,
966 and to Theodore T'so <tytso@valinux.com> for his always
967 useful thoughts.
968
969 8. Made dump ask upon a tape write error if it should rewrite
970 the current volume (assume this is a bad tape) or if it should
971 assume an end-of-tape condition (useful for tape drives which
972 behaves badly at the end of the tape). Thanks to Andreas
973 Dilger <adilger@turbolinux.com> for the suggestion.
974
975 Changes between versions 0.4b19 and 0.4b20 (released November 10, 2000)
976 =======================================================================
977
978 1. Fixed a small compilation problem due to a change
979 in the definintion of the struct sigaction in
980 glibc 2.0 / libc5. Thanks to Gunther Schlegel
981 <schlegel@riege.de> for reporting the bug and to
982 Dave Platt <dplatt@snulbug.mtview.ca.us> for suggesting
983 a fix.
984
985 2. Modified the label and uuid analysis in order to be
986 self-contained (independant of kernel/libc headers). This
987 should fix the compile with older kernel/libc headers and
988 will preserve the functionality. Thanks to Bernhard Erdmann
989 <bernhard.erdmann@gmx.de> for reporting the bug.
990
991 3. The 'exclude inode' option, if applied to a directory
992 inode, excludes now correctly all the leaves of this
993 directory. Thanks to John R. Dennison
994 <gerdesas@users.sourceforge.net> for reporting the bug.
995
996 4. Fixed the '-e' option to disable the possibility
997 to exclude the root inode (which causes the dumps to
998 be unreadable by restore). Prevented array overflow
999 when multiple -e options are used.
1000
1001 5. Fixed dump to correctly interpret a filesystem argument
1002 which is a mountpoint and it is not an absolute path
1003 (as specified in the fstab). Thanks to Bernhard R. Erdmann
1004 <be@berdmann.de> for reporting the bug.
1005
1006 6. Made dump able to backup files larger than 2 GB. Note that
1007 dump still doesn't cope with files larger than 4 GB.
1008
1009 7. Restore the real uid and gid before invoking an external
1010 RSH program (big hole when dump or restore is suid root!).
1011
1012 8. Get the values for the default tape device and the location
1013 of fstab file from the system headers. Thanks to
1014 Andreas Dilger <adilger@turbolinux.com> for the patch.
1015
1016 9. Made dump -w|-W report all recognized filesystems
1017 present in either /etc/fstab or /etc/dumpdates, and present
1018 the list in the same order as in fstab file. Thanks
1019 to Andreas Dilger <adilger@turbolinux.com> for the patch.
1020
1021 10. Made dump's -a (automatic end of tape detection) the
1022 default. Specifying one of -B, -c, -d or -s options will
1023 override the EOM detection. Thanks to Andreas Dilger
1024 <adilger@turbolinux.com> for the patch.
1025
1026 11. Save the ext2 filesystem label into the dump volume label.
1027 Specifying a volume label on the command line (-L option)
1028 will override this feature. Thanks to Andreas Dilger
1029 <adilger@turbolinux.com> for the patch.
1030
1031 Changes between versions 0.4b18 and 0.4b19 (released August 20, 2000)
1032 =====================================================================
1033
1034 1. Fixed the signal handling in dump (which I broke in 0.4b17)
1035 which was causing several strange dump failures (dump
1036 hanged or segmentation faults etc).
1037
1038 2. Specified the default blocksize in the dump man page.
1039
1040 3. Changed two info messages of restore to be written on stdout
1041 instead of stderr in order to leave stderr only for errors
1042 or warnings. Thanks to Stephen Atwell
1043 <satwell@urbana.css.mot.com> for the suggestion.
1044
1045 4. Corrected an off by one calculation which prevented
1046 dumping a directory having a 1 character name length.
1047 Thanks to Bernhard Erdmann <bernhard.erdmann@gmx.de>
1048 for reporting the bug.
1049
1050 5. Reinforce dump in order to allow only level 0 and no
1051 -u option when dumping a subdirectory, and document
1052 this in the man page. Thanks to Bernhard Erdmann
1053 <bernhard.erdmann@gmx.de> for reporting the bug.
1054
1055 6. Fixed a small harmless bug in restore which caused
1056 in some conditions hard links to be restored several
1057 times (therefore generation some warning messages).
1058 Thanks to Kenneth Porter <shiva@well.com> for
1059 reporting the bug.
1060
1061 7. Updated the RPM spec file to the latest RedHat version,
1062 providing FHS packaging and other cosmetic changes.
1063 You will need RPM version >= 3.0.5 in order to build the RPM.
1064
1065 8. Updated the configure script to check for libtermcap
1066 before checking for libreadline, because we need this
1067 library in order to compile the readline support.
1068
1069 9. Made dump understand the LABEL= and UUID= notation
1070 both in /etc/fstab and on the command line. Note that
1071 you will need the /proc filesystem in order to use
1072 these notations. Thanks to Erik Troan <ewt@redhat.com>
1073 for providing the patch.
1074
1075 Changes between versions 0.4b17 and 0.4b18 (released June 30, 2000)
1076 ===================================================================
1077
1078 1. Fixed a potential buffer overflow in restore. Thanks
1079 to Stan Bubrouski <satan@fastdial.net> for reporting
1080 the bug.
1081
1082 2. Fixed a readline-related bug which prevented
1083 'cat DUMPFILE | restore -i -f -' from working. Thanks
1084 to Charles Karney <karney@users.sourceforge.net>
1085 for the bug report.
1086
1087 3. Changed a few "panic" into "exit", causing restore to
1088 be more stable against some attacks (like the last one
1089 reported on Bugtraq, although the last version of restore
1090 was not vulnerable - just dumped core). Thanks to
1091 Andreas Hasenack <andreas@conectiva.com.br> for reporting
1092 the bugs to me.
1093
1094 4. Removed the suid-root bit on dump and restore in the
1095 default build (and generated RPMs). It should be safer
1096 now. Users who need the suid-root capabilities in order
1097 to make network backups should read first the man page
1098 of dump and enable the suid bit manually.
1099
1100 5. Added -ltermcap to the compile parameters for restore
1101 when using readline support, in order to make the compile
1102 process work on some systems (SuSE ?). Thanks to
1103 Patrik Schindler <poc@pocnet.net> for reporting the bug.
1104
1105 Changes between versions 0.4b16 and 0.4b17 (released June 1st, 2000)
1106 ====================================================================
1107
1108 1. The -F script is called now *only* at the end of a tape,
1109 not at the end of the dump. Thanks to Christian Haul
1110 <haul@informatik.tu-darmstadt.de> for the bug report.
1111
1112 Normally, the device name and the current volume number
1113 are passed on the command line of the script. However,
1114 if you want the old style script (with no arguments
1115 passed on the command line), you can enable it in
1116 configure with the --enable-oldstylefscript.
1117
1118 2. Use posix signal handling to preserve dump functionality
1119 with libc5. Thanks to Jeff Johnson <jbj@redhat.com> for
1120 the patch.
1121
1122 3. Made the exit code of restore in case of a 'C'ompare
1123 command reflect the result of the compare. An exit status
1124 of 0 means the dump archive is correct, 1 means tape errors,
1125 2 means that some files were modified. Thanks to Kenneth Porter
1126 <shiva@well.com> for the suggestion.
1127
1128 4. Made (finally) quotes work in interactive restore.
1129
1130 5. Small fixes in order to allow dump to compile with a
1131 really old e2fsprogs version. Thanks to Ian Zimmerman
1132 <itz@speakeasy.org> for the bug report.
1133
1134 6. Add GNU readline capabilities to interactive restore.
1135 Use configure's --enable-readline flag to enable this feature.
1136 Thanks to Patrik Schindler <poc@pocnet.net> for the
1137 suggestion.
1138
1139 7. Do the compare on the fly in restore 'C' mode (this will
1140 allow not to exhaust the available /tmp space when
1141 ccmparing large files). Thanks to Kenneth Porter
1142 <shiva@well.com> for the suggestion.
1143
1144 Changes between versions 0.4b15 and 0.4b16 (released March 11, 2000)
1145 ====================================================================
1146
1147 1. Fixed some several duplicate 'DUMP: DUMP:' in the
1148 output of dump.
1149
1150 2. Corrected the estimation of blocks to dump. Note that
1151 this is still wrong for large directory inodes, due
1152 to the size of a BSD directory entry on the tape
1153 (estimation is based on the size of the ext2 directory,
1154 which is smaller because it doesn't contain the null
1155 character at the end).
1156
1157 3. Made dump report the total number of MB written to
1158 tape at the end of run. Thanks to W. Reilly Cooley
1159 <wcooley@nakedape.cc> for the patch.
1160
1161 4. Added the -X option to restore giving the possibility
1162 to read the names of the files to be extracted/listed
1163 from a text file (in addition of the command line).
1164 Thanks to Dejan Muhamedagic <dejan@quant-x.com> for the
1165 patch.
1166
1167 5. Added the device name and the current volume number
1168 as arguments to the end of tape script (dump -F option).
1169
1170 6. Made the multi-volume dump work again (after having
1171 broken it in 0.4b15).
1172
1173 Changes between versions 0.4b14 and 0.4b15 (released March 2, 2000)
1174 ===================================================================
1175
1176 1. Added a prompt command in interactive restore mode. Thanks
1177 to Andreas Dilger <adilger@home.com> for the patch.
1178
1179 2. Fixed a buffer overflow problem in dump (caused by
1180 not checking the size of the filesystem parameter).
1181 Thanks to Kim Yong-jun <loveyou@hackerslab.org> for
1182 reporting this on Bugtraq (and to several dump users
1183 who forwarded me his mail).
1184
1185 3. Added the '-F script' option to dump in order to
1186 launch a script at the end of each tape (to be used
1187 with a tape changer, or to notify the sysadmin by
1188 pager etc.).
1189
1190 4. Fixed a bug in restore compare code caused by the changes
1191 I made in 0.4b14.
1192
1193 5. Fixed the treatment of options using the old BSD syntax
1194 in both dump and restore.
1195
1196 Changes between versions 0.4b13 and 0.4b14 (released February 10, 2000)
1197 =======================================================================
1198
1199 1. Fixed a bug in dump which may have caused invalid deleted
1200 directories to be dumped out if they were deleted after the
1201 mapping pass. This could occure on active filesystem and lead
1202 to heap corruption (causing dump malfunction in many possible ways).
1203 Thanks to Jan Sanislo <oystr@cs.washington.edu> for finding this
1204 bug and submitting the patch.
1205
1206 2. Fixed the handling of the filesystem parameter in dump. A
1207 '/mnt/disk' parameter caused the disk contents to be dumped,
1208 but a '/mnt/disk/' parameter caused the mountpoint directory
1209 to be dumped (generally an empty directory).
1210
1211 3. Improved the output of dump in order to tell which directory
1212 it is currently dumping (when dumping a subtree).
1213
1214 4. Added the '-e' exclude inode option to dump. Thanks to
1215 Isaac Chuang <ike@isl.stanford.edu> for contributing with the patch.
1216
1217 5. Added a REPORTING-BUGS file in order to provide a guide
1218 on how to correctly report a bug in dump/restore.
1219
1220 6. Fixed a restore problem when restoring a hard link to an inode
1221 having the immutable or append-only attribute set. Thanks to
1222 Ambrose Li <acli@mingpaoxpress.com> for submitting the patch.
1223
1224 7. Fixed a compatibility problem between dumps made on little
1225 endian machines (the format was correct) and big endian
1226 machines (the format was incorrect). This fix break the
1227 compatibility with the older dumps made on big endian
1228 machines (sparc, m86k, ppc etc). For the first time in
1229 linux dump's history (I believe), the dumps made by *BSD,
1230 Linux/alpha, Linux/sparc and Linux/x86 are compatible,
1231 so interchangeable. Thanks to Rob Cermak
1232 <cermak@ahab.rutgers.edu> for submitting the bug and
1233 helping me test the fix.
1234
1235 8. Fixed the way dump reports the remaining percent/time, if
1236 the number of blocks actually dumped exceeds the estimated
1237 number of blocks. Thanks to Jean-Paul van der Jagt
1238 <jeanpaul@dutepp0.et.tudelft.nl> for reporting the bug.
1239
1240 Changes between versions 0.4b12 and 0.4b13 (released January 21, 2000)
1241 ======================================================================
1242
1243 1. Small Y2K fix in the man pages :). Thanks to Bernhard Sadlowski
1244 <sadlowsk@Mathematik.Uni-Bielefeld.DE> for reporting the bug.
1245
1246 2. Removed the requirement to build the RPM as root from the
1247 spec file. Thanks to Christian Weisgerber
1248 <naddy@mips.rhein-neckar.de> for submitting this.
1249
1250 3. Fixed a bug in dump related to the 'filetype' feature of ext2fs,
1251 causing dump to block when dumping really huge filesystems.
1252 Many thanks to Patrik Schindler <poc@pocnet.net> for
1253 helping me find this bug.
1254
1255 4. Fixed the treatment for an interrupt signal when dump access
1256 the remote tape through RSH. Thanks to Christian Weisgerber
1257 <naddy@mips.rhein-neckar.de> for providing the patch.
1258
1259 5. Fixed a bug which was causing dump/restore to display
1260 garbage characters instead of the remote host name.
1261
1262 Changes between versions 0.4b11 and 0.4b12 (released January 8, 2000)
1263 =====================================================================
1264
1265 1. Small fix in the dump man page. Thanks to Thorsten Kukuk
1266 <kukuk@suse.de> for submitting the patch.
1267
1268 2. Fix for the exit code when using the size estimate option of
1269 dump. Thanks to Matti Taskinen <mkt@rni.helsinki.fi> for
1270 submitting the patch.
1271
1272 3. Handle EINTR in atomical reads/writes in dump, which was causing
1273 dump to fail on some systems. Thanks to Eric Jergensen
1274 <eric@dvns.com> for reporting the bug and submitting the patch.
1275
1276 4. Handle more than 16 characters for the device names in dumpdates.
1277 (up to 255 now). Thanks to Rainer Clasen <bj@ncc.cicely.de> for
1278 tracking down the problem and proposing the solution.
1279
1280 5. Fixed a bug in dump which prevented the creation of the
1281 dumpdates file when doing a 0-level dump without already
1282 having a dumpdates file. Thanks to Patrik Schindler
1283 <poc@pocnet.net> for reporting the bug.
1284
1285 6. Changed the way dump 'S' flag reports the size estimate
1286 from number of blocks into bytes (making it compatible
1287 with the Solaris version, and simplifying things for
1288 amanda users). Thanks to Jason L Tibbitts III
1289 <tibbs@math.uh.edu> for reporting the bug.
1290
1291 7. Fixed a compatibility problem in linux/alpha dump tape format.
1292 Now the linux/alpha dump are (again) compatible with the
1293 other dump formats. But this breaks compatibility with
1294 older dumps made on alpha. Thanks to Mike Tibor
1295 <tibor@lib.uaa.alaska.edu> for helping me in finding this bug.
1296
1297 Changes between versions 0.4b10 and 0.4b11 (released December 5, 1999)
1298 ======================================================================
1299
1300 1. Added a '--enable-kerberos' to configure.
1301
1302 2. Added a 'S' option to dump which determines the amount of space
1303 that is needed to perform the dump without actually doing it, similar
1304 to the Sun's ufsdump 'S' option. Patch contributed by Rob Cermak
1305 <cermak@ahab.rutgers.edu>.
1306
1307 3. Added a 'M' multi-volume option to dump and restore which enables
1308 dumping to multiple files (useful when dumping to an ext2
1309 partition to make several dump files in order to bypass the 2GB
1310 file size limitation). The 'f' argument is treated as a prefix and
1311 the output files will be named <prefix>001, <prefix>002 etc. With
1312 the 'M' flag, restore automatically selects the right file without
1313 asking to enter a new tape each time.
1314
1315 4. Fixed a memory leak which was causing dump to grow very big
1316 (270MB when dumping a 10GB filesystem...). Thanks to Jason
1317 Fearon <jasonf@netrider.org.au> for reporting the bug.
1318
1319 Changes between versions 0.4b9 and 0.4b10 (released November 21, 1999)
1320 ======================================================================
1321
1322 1. Make configure test if the system glob routines support
1323 extended syntax (ALTDIRFUNC). If not, use the internal glob
1324 routines instead of system ones. Thanks to Bernhard Sadlowski
1325 <sadlowsk@Mathematik.Uni-Bielefeld.DE> for reporting the bug
1326 and helping me resolve this and other minor libc5 compiling
1327 glitches.
1328
1329 2. Fix a problem when dumping a ext2fs with the 'filetype'
1330 feature enabled. Thanks to Patrick J. LoPresti
1331 <patl@cag.lcs.mit.edu> for reporting the bug and to
1332 Theodore Y. Ts'o <tytso@mit.edu> for providing the patch.
1333
1334 3. Made the nodump flag work on directories. A directory which
1335 has the nodump flag gets never dumped, regardless of its
1336 contents.
1337
1338 4. Integrate a patch from Jeremy Fitzhardinge <jeremy@goop.org>
1339 which allows dump on an active ext3 filesystem. However, this
1340 is a "quick and dirty" patch which enables backup of an ext3
1341 filesystem through the ext2 compatibility (by ignoring the
1342 NEEDS_RECOVERY bit). The journal file is not recognized and
1343 it is dumped (it should not).
1344
1345 5. Test the superblock compatibility flags when dumping, in order
1346 to be sure that we know how to deal with specific features.
1347
1348 Changes between versions 0.4b8 and 0.4b9 (released November 5, 1999)
1349 ====================================================================
1350
1351 1. Use lchown instead of chown, fixing a possible security problem
1352 when restoring symlinks (a malicious user could use this
1353 to deliberately corrupt the ownership of important system files).
1354 Thanks to Chris Siebenmann <cks@utcc.utoronto.ca> for detecting
1355 this and providing the patch.
1356
1357 Changes between versions 0.4b7 and 0.4b8 (released November 3, 1999)
1358 ====================================================================
1359
1360 1. Put dump sources under CVS, added Id tags in all files so
1361 one can use 'ident' on binary files.
1362
1363 2. Added the dump/restore version in the usage text so one can
1364 easily verify the version he is using.
1365
1366 3. Small patch from Nuno Oliveira <nuno@eq.uc.pt> which fixes
1367 a va_start/va_end problem on linux-ppc (always call va_start
1368 va_end in pairs each time we use a vararg function).
1369
1370 4. Added again the DT_* constants because old libc does not
1371 contain them :(. Thanks to Eric Maisonobe <virnet@nat.fr>
1372 for submitting the bug report.
1373
1374 5. Use ext2fs_llseek instead of llseek. With recent e2fsprogs
1375 this should enable dumping big (huge) filesystems.
1376
1377 6. Added the RSH environment variable in order to be able to
1378 use a rsh replacement like ssh when doing remote backups (and
1379 bypass the security limitations of rcmd). Now you can do remote
1380 backups without being root (or making dump setuid root).
1381
1382 7. Modified again the way dumpdates works. For incremental dumps,
1383 we need to read dumpdates even if we are not using 'u' option.
1384 Thanks to Bdale Garbee <bdale@gag.com> for his ideas on how
1385 this should work.
1386
1387 Changes between versions 0.4b6 and 0.4b7 (released October 8, 1999)
1388 ===================================================================
1389
1390 1. Removed the 'k' flag from the restore 'about' text if kerberos
1391 was not compiled in.
1392
1393 2. Prototyped (f)setflags from e2fsprogs and corrected the calls
1394 to them (fsetflags takes a char*, setflags an open fd!).
1395
1396 3. (f)setflags is called only if the flags aren't empty. If the
1397 file is a special file, a warning is printed, because changing
1398 flags implies opening the device. Normally, a special file
1399 should not have any flag... (Debian bug #29775, patch provided
1400 by Abhijit Dasgupta <abhijit@ans.net>).
1401
1402 4. Made possible to dump a file system not mentioned in /etc/fstab.
1403 (Debian bug #11904, patch provided by Eirik Fuller <eirik@netcom.com>).
1404
1405 5. Changed the default behaviour to not create dumpdates
1406 unless 'u' option is specified. Removed the old "debian-patch"
1407 which provided the same thing. (Debian bug #38136, #33818).
1408
1409 6. Removed all those dump*announce, since they were getting old...
1410
1411 7. Added warning messages if dumpdates does not exist and
1412 when an inferior level dump does not exist (except for a level 0
1413 dump).
1414
1415 8. Debugged the glob calls in interactive mode: restore used a
1416 dirent struct which was different from the /usr/include/dirent.h
1417 one (this used to work, is it a glibc2 change?), so none of the
1418 compat glob (which used /usr/include/dirent.h) or the system glob
1419 worked. Restore use now the system dirent (and the system
1420 DT_* constants), which are compatible with BSD ones.
1421
1422 9. Added a configure flag (--with-dumpdatespath) to specify
1423 the location of dumpdates file. By default, it is
1424 /etc/dumpdates.
1425
1426 10. Added the "AUTHOR" and "AVAILABILITY" sections and
1427 included the current date/version in man pages.
1428
1429 11. Corrected the estimation of remaining time when
1430 the operator doesn't change the tapes quickly enough. This
1431 was an old bug, I thought I corrected it, and discovered
1432 that in fact it was corrected in two different places, so
1433 the results canceled each other...
1434
1435 Changes between versions 0.4b5 and 0.4b6 (released October 1, 1999)
1436 ===================================================================
1437
1438 1. Integrated multiple patches from RedHat, Debian and SuSE:
1439
1440 - tweak dump/itime.c to not try to read dumpdates if the 'u' option
1441 isn't specified.
1442 - several fixes in the man pages.
1443 - update the default tape device to /dev/st0.
1444 - many updates for Linux Alpha (byte ordering, size_t etc).
1445 - buffer overruns.
1446 - use environment variable for TMPDIR (instead of /tmp).
1447 - use sigjmp_buf instead of jmp_buf (RedHat bug #3260).
1448 - workaround egcs bug (RedHat bugs #4281 and #2989).
1449 - wire $(OPT) throughout Makefile's.
1450
1451 2. Upgrade the dump revision to 1, making possible to dump filesystems
1452 made with e2fsprogs-1.15 or newer. Nothing seems to break...
1453
1454 3. Fix some compile warnings, prototype all functions.
1455
1456 4. Use glibc err/glob instead of internal compatibility
1457 routines (only if available).
1458
1459 5. Fix a compile error on Linux 2.2.7 / libc5 (5.4.44) (patch provided
1460 by Bernhard Sadlowski <sadlowsk@mathematik.uni-bielefeld.de>).
1461
1462 Changes between versions 0.4b4 and 0.4b5 (released September 22, 1999)
1463 ======================================================================
1464
1465 1. Integrated the changes from FreeBSD-3.1-RELEASE
1466 (mostly bug fixes, buffer overruns, dump has now an "automatic
1467 tape length calculation" flag, dump/restore can use kerberos now
1468 (this is NOT tested), use environment variables for TAPE and
1469 RMT etc.).
1470
1471 2. Integrated three RedHat patches ("glibc", "kernel" and "bread" patches)
1472
1473 3. Corrected a bug in restore when using 'C' option with multi-volumes
1474 tapes (files splited accros two tapes give "size changed" errors
1475 when comparing).
1476
1477 4. Corrected the long standing bug when dumping multiple tapes.
1478 This works for me, needs further testing.
1479
1480 Changes between versions 0.4b3 and 0.4b4 (released January 17, 1997)
1481 ====================================================================
1482
1483 1. Dump now runs correctly on kernels 2.1.x
1484 Fix made by Gerald Peters <gapeters@worldnet.att.net>
1485
1486 Changes between versions 0.4b2 and 0.4b3
1487 ========================================
1488
1489 1. Use realpath() if available
1490
1491 2. Report statistics
1492
1493 Changes between versions 0.4b1 and 0.4b2
1494 ========================================
1495
1496 1. Fixed the bug fix from Greg Lutz (I had made a mistake when integrating
1497 the patch)
1498
1499 2. Fixed restore to make it able to read FreeBSD 2.x dumps again
1500
1501 3. Fixed configure.in to correctly handle --enable-rmt
1502
1503 Changes between versions 0.3 and 0.4b1
1504 ======================================
1505
1506 1. Integrated the changes from 4.4BSD-Lite2
1507
1508 2. Integrated the patches from Debian and Red Hat
1509
1510 3. Portability changes: use the __u32, __u16, __s32, and __s16 types
1511
1512 4. Changed dump to use the Ext2fs library to get block addresses. This
1513 should solve the endianness problem on SparcLinux.
1514
1515 5. Created a configure.in file (shamelessly stolen from the e2fsprogs
1516 distribution's one) to use autoconf
1517
1518 6. Fixed a few minor bugs
1519
1520 Changes between versions 0.2e and 0.2f
1521 ======================================
1522
1523 1. Added the creation of named pipes (fifos) in restore.
1524
1525 2. Added the -N flag in the restore manual page.
1526
1527 3. Added the file kernel.patch which contains the llseek() optimization
1528 patch for 1.2.x kernels.
1529
1530 4. Fixed a bug in the restoration of symbolic links: owner and group were
1531 not restored.
1532
1533 5. Integrated some changes from FreeBSD 2.2.
1534
1535 6. Added a call to ftruncate() after restoring each file to restore
1536 correctly files ending by a hole.
1537
1538 Changes between versions 0.2d and 0.2e
1539 ======================================
1540
1541 1. Fixed a bug in the "set owner/mode" process. Every file was restored
1542 with owner = root (0) and group = root/wheel/whatever (0).
1543
1544 Changes between versions 0.2c and 0.2d
1545 ======================================
1546
1547 1. Dump is now able to backup 2GB+ filesystems.
1548
1549 2. Dump and restore can now be linked as static binaries.
1550
1551 Changes between versions 0.2b and 0.2c
1552 ======================================
1553
1554 1. Fixed a bug when dumping ``slow'' (i.e. normal) symbolic links.
1555
1556 Changes between versions 0.2a and 0.2b
1557 ======================================
1558
1559 1. Really fixed the bug that I should have corrected in 0.2a.
1560
1561 2. Enabled optimization again.
1562
1563 Changes between versions 0.2 and 0.2a
1564 =====================================
1565
1566 1. Disabled the optimization during compilation.
1567
1568 Changes between versions 0.1 and 0.2
1569 ====================================
1570
1571 1. Fixed a bug in fstab.c which caused a null pointer to be stored in
1572 the fs_type field (actually, I modified the file fstab.c to make it
1573 use the mntent functions).
1574
1575 2. Dump and restore now use a 4.3 BSD compatible dump format. Backups
1576 made by dump should be readable by the BSD restore and backups made
1577 by the BSD dump should be readable by restore. Unfortunately, this
1578 means that the dump format has changed between version 0.1 and version
1579 0.2 :-(
1580
1581 3. Dump is now able to backup a subtree, it is no longer limited to whole
1582 filesystems like the BSD version.
1583
1584 4. Dump now uses ext2_llseek() so it is able to backup filesystems bigger
1585 than 2 GB.
1586
1587 Changes between versions 0.0 and 0.1
1588 ====================================
1589
1590 1. Now create links rdump and rrestore during the `make install' step.
1591
1592 2. Linux port specific bugs added to the manual pages
1593
1594 3. Incorrect estimation of the number of tapes blocks fixed when doing
1595 incremental backups.
1596
1597 4. Better ls-like format in restore in interactive mode.