Improve level 1 mapdirs speed.
[dump.git] / CHANGES
1 $Id: CHANGES,v 1.312 2010/06/10 12:01:30 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 5. Don't attempt to set extended attributes on files that were not
25 extracted in 'restore -N' mode. Thanks to Jan Görig
26 <jgorig@users.sourceforge.net> for the patch.
27
28 6. Make the dump man page explicit about ext4 support.
29
30 7. Fix a bug in dump which caused EA entries to be unrecognizable by
31 restore in some corner cases. Add a workaround to restore making
32 it able to properly deal with those corrupted EA entries. Thanks
33 to John Austin <jaustin1@users.sourceforge.net> for reporting
34 the bug.
35
36 8. Fix some issues when restoring a dump which was generated using
37 exclusion patterns (either via -e or via the nodump attribute)
38 (see the Debian bug #574667 for details. Thanks to Frédéric
39 Brière <fbriere@fbriere.net> for the bug report and the associated
40 patch.
41
42 9. Fix progress/estimated blocks display in dump when doing really
43 huge backups. Thanks to Steve Bonds <sbonds@users.sourceforge.net>
44 for the bug report and tests. (Sourceforge bug #2987758)
45
46 10. Improve level 1 dump speed by rearanging the "mapdirs" code.
47 Thanks to Andreas Kies <andikies@t-online.de> for finding the
48 bottleneck and sending a patch.
49
50 Changes between versions 0.4b41 and 0.4b42 (released June 18, 2009)
51 ===================================================================
52
53 1. Fix printout of the 'Connection to' message in dump/restore.
54 Thanks to Dale Wiles <dwiles@users.sourceforge.net> for the
55 bug report and its fix.
56
57 2. Fix dump -w/-W output for filesystems having the last backup
58 level equal to 0. Thanks to Pascal Bouchareine
59 <kalou@users.sourceforge.net> for reporting the bug.
60
61 3. Enable restore to handle restoring onto a different SELinux
62 policy from the dump. Thanks to Tony Nelson
63 <tony_nelson@users.sourceforge.net> for the patch.
64 (see http://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=189845
65 for details)
66
67 4. Fix EA set failures when restoring immutable files. Thanks
68 to Andrew Kroeger for the patch.
69
70 5. Fix "mode file too large" errors when restoring huge backups.
71 Thanks to Kenneth Porter <shiva@sewingwitch.com> for the bug
72 report.
73
74 6. Add the ability to link against libtinfo library if present,
75 instead of libtermcap or libncurses libraries. Thanks to
76 Adam Tkac <atkac@redhat.com> for the patch.
77
78 7. When comparing a backup, do not consider to be an error the fact
79 that we encounter files created while dump was in progress.
80 Thanks to Kenneth Porter <shiva@sewingwitch.com> for the bug
81 report.
82
83 8. Force '-a' to be the default value when running 'restore -C'.
84 Thanks to Kenneth Porter <shiva@sewingwitch.com> for the bug
85 report.
86
87 9. Many improvements to the 'cron_dump_to_disk' example.
88 Thanks to Aaron S. Hawley <ashawley@users.sourceforge.net> for
89 the patch.
90
91 10. Fix the QFA generation when extended attributes are backuped
92 along with the inodes. Prior to this fix, the entries for
93 some inodes may get corrupted (duplicated entries or incorrect
94 ones), making the QFA file unusable for repositionning in
95 restore. Many thanks to Kenneth Porter <shiva@sewingwitch.com>
96 for the bug report and the testing of the fix.
97
98 11. Fix EA restoration in debug mode (Sourceforge bug #1986216).
99
100 12. Stefan Auracher <star7@onlinehome.de> noticed that the Tower
101 of Hanoi backup strategy documented in the man page wasn't
102 actually based on the Tower of Hanoi algorithm. The man page
103 was updated accordingly.
104
105 13. Fix missing level information from dump output when using
106 the default level (RedHat bug #493635). Thanks to Adam Tkac
107 <atkac@redhat.com> for the patch.
108
109 14. Fix the dump man page to reflect the fact that the default
110 dump level is 0 not 9 (RedHat bug #356121). Thanks to Adam Tkac
111 <atkac@redhat.com> for the patch.
112
113 15. Fixed a few spelling errors in the man pages (RedHat bug #489853).
114 Thanks to Adam Tkac <atkac@redhat.com> for the patch.
115
116 16. Fixed restore man page and program usage which incorrectly stated
117 that -P and -A were both allowed in the same invocation (RedHat
118 bug #490627). Thanks to Adam Tkac <atkac@redhat.com> for the patch.
119
120 17. Use sys_clone under Linux to share I/O contexts between dump
121 processes, thus drastically increasing the performance of dump
122 under CFQ (which is the default I/O scheduler used in a number of
123 distributions). Thanks to Jeff Moyer <jmoyer@redhat.com> for the
124 patch.
125
126 18. Add (preliminary) ext4 support - thanks to libext2fs which does
127 all the job for us. Thanks to Gertjan van Wingerde
128 <gwingerde@gmail.com> for the patch.
129
130 Changes between versions 0.4b40 and 0.4b41 (released January 2, 2006)
131 =====================================================================
132
133 1. Fix restore of dumped Access Control Lists. The previous
134 code for EA works fine, but ACLs needed conversion from the
135 ext2/3 disk format to posix_acl format before restoring.
136
137 2. Fix some issues with restoration of EA on big endian
138 platforms.
139
140 3. Fix restore when the symtab is over 2GB in size.
141
142 4. Made the directory hash indexing an optional feature,
143 accessible by the '-H' option of restore, and disable it
144 by default.
145
146 5. Fixed dump to not include extended attributes information
147 in the toc (archive) file which confused restore -t.
148
149 Changes between versions 0.4b39 and 0.4b40 (released May 2, 2005)
150 =================================================================
151
152 1. Changed restore to emit warnings (instead of emitting a fatal
153 error) if a file (or a directory) is unavailable for a
154 comparision (if the user doesn't have the necessary permissions
155 to access it for example). Thanks to Kenneth Porter
156 <shiva@sewingwitch.com> for the bug report.
157
158 2. Re-done the 'do not save directory entries to non-dumped inodes
159 (excluded from dump)' feature. The previous implementation
160 worked well for excluded directories but not for regular files.
161 Thanks to Kenneth Porter <shiva@sewingwitch.com> for the bug
162 report.
163
164 3. Fixed a bug in dump where the tape size was miscalculated when
165 the user used -d/-s to specify the tape characteristics. Thanks
166 to Philip Goisman <goisman@physics.arizona.edu> for reporting
167 the bug.
168
169 4. Fixed another bug introduced in restore with the hashtree
170 implementation. This one caused restore to stop saying
171 "removenode: non-empty directory" in some cases.
172
173 5. Added support for dumping and restoring ext2/3 extended
174 attributes (EA), like the access control lists (ACL) or
175 the security labels used by SELinux.
176
177 Changes between versions 0.4b38 and 0.4b39 (released January 21, 2005)
178 ======================================================================
179
180 1. The newly added dump_on_cd_3 example was buggy, replace it
181 with an updated version from Andrew Basterfield
182 <bob@cemetery.homeunix.org>.
183
184 2. Made restore to chdir() back into the initial directory when
185 dumping core while aborting a comparision operation. The
186 previous behaviour was to write the corefile at the root of
187 the directory being compared, which could very well be
188 read only and preventing the corefile generation. Thanks
189 to Kenneth Porter <shiva@sewingwitch.com> for the bug report.
190
191 3. Silenced the failure to call fgetflags() when comparing an
192 entry which has no ext2 attributes (as in lsattr()).
193
194 4. Fix a brown paper bug in restore -C which broke restore and
195 caused modifications on the filesystem being compared
196 (directories containing a file with the same name as the
197 directory get renamed to RSTTMP...). Thanks to Kenneth Porter
198 <shiva@sewingwitch.com> for finding the bug and helping me
199 reproduce it.
200
201 5. Made restore -C force the -N flag (no writing allowed on
202 the disk) in order to prevent more bugs like the above one.
203
204 Changes between versions 0.4b37 and 0.4b38 (released January 7, 2005)
205 =====================================================================
206
207 1. Fix a couple of troff syntax bugs in the man pages.
208 Thanks to Eric Raymond <esr@thyrsus.com> for the patch.
209
210 2. Made restore use either libncurses or libtermcap, depending
211 on which one is available at configure time.
212
213 3. Fixed restore negative size display bug when comparing a
214 dump containing files over 2GB. Thanks to Steve Bonds
215 <sbonds@users.sourceforge.net> for the bug report.
216
217 4. Do not save directory entries to non-dumped inodes
218 (excluded from dump). This will eliminate the 'missing
219 file' warnings when doing 'restore -C'.
220
221 5. Fix dump crash when backuping a huge (2TB) filesystem,
222 due to a bogus calculation on the inode map size.
223 Thanks to Kevin B. Haines <K.B.Haines@rl.ac.uk> for
224 submitting the bug and testing the fix.
225
226 6. Fix a problem in restore where the final \0 in the symbolic
227 link names could have been lost, generating corrupt filenames.
228 Thanks to Kyle Wilson <kyle.wilson@amd.com> for reporting the
229 bug.
230
231 7. Implemented a hash list for the directory names in restore.
232 The linear list used before caused problems in interactive
233 restores when dealing with directories having thousands of
234 entries. Thanks to Brian Ristuccia <bristuccia@starentnetworks.com>
235 for reporting the bug.
236
237 8. Improved restore -C, this time including the directory
238 attributes into the comparision.
239
240 9. Made restore understand tapes containing EA/ACLs (which will
241 be dumped by the next version of dump). In this version
242 extended attributes on the tape are ignored, for full EA/ACL
243 support wait for the next version or try the experimental EA
244 patch.
245
246 Changes between versions 0.4b36 and 0.4b37 (released July 7, 2004)
247 ==================================================================
248
249 1. Added the --enable-staticz configure option which enables
250 dump and restore to be linked with static versions of
251 libz and libbz2 (and dynamic versions of all other libraries).
252 This will make Debian users happy, because libz and libbz2
253 were the only needed libraries living in /usr, all the
254 others live in /lib. In case of system emergency, it is
255 better not to have to rely on an extra filesystem.
256 Thanks to Bdale Garbee <bdale@gag.com> for the suggestion.
257
258 2. Fix compilation on (at least the Linux Debian port to) AMD64.
259 (<ext2fs/ext2_types.h> defines some types (__s64 and __u64)
260 that are also defined by <linux/types.h> (<asm/types.h>) and
261 they conflict).
262
263 3. Make dump's reading of the dumpdates file a bit more robust,
264 preventing dump from crashing when the dumpdates file has
265 been modified by hand.
266
267 4. Fixed some offset calculations in dump code which could
268 lead to "bread lseek errors" on large filesystems. Thanks
269 to Bruce Lowekamp <lowekamp@users.sourceforge.net> for
270 reporting this bug and debugging the issue.
271
272 5. Made dump use the blkid library when searching for devices
273 by label or uuid instead of dump's own routines.
274
275 6. Corrected a bug in dump where a wrong LABEL=... line in
276 /etc/fstab could prevent dump from dumping unrelated
277 filesystems. Thanks to Bruce Lowekamp
278 <lowekamp@users.sourceforge.net> for reporting the bug.
279
280 Changes between versions 0.4b35 and 0.4b36 (released April 21, 2004)
281 ====================================================================
282
283 1. Fixed dump compilation with old gcc versions. Thanks to
284 Mike Castle <dalgoda@users.sourceforge.net> for the patch.
285
286 2. Fixed some warnings (howmany, roundup, powerof2 redefined)
287 when compiling against a recent glibc version.
288
289 3. Fixed a bug in restore preventing the read of a dump tape
290 written with Solaris 7 ufsdump. Thanks to Patrick Higgins
291 <phiggins@transzap.com> for reporting the bug and providing
292 the test case.
293
294 4. Changed dump to enable the creation of volumes of different
295 sizes in a single run (make -B accept a list of values).
296 Patch contributed by Florian Zumbiehl <florz@gmx.de>.
297
298 5. Use the glibc provided minor() and major() macros instead
299 of our own bitmask implementation. This should be safe for
300 when the major/minor namespace will migrate to 32 bits.
301 Thanks to Zhang Jun <zhangjun@nanjing-fnst.com> for reporting
302 the bug.
303
304 6. Made explicit in the dump man page that dump will not create
305 a remote file, it will only write to an already existing one.
306
307 7. Another try at making size estimates better again.
308
309 8. Put back the inconditional running of the end-of-tape
310 script which was changed in 0.434 to be run only when -M
311 or multiple -f were NOT specified. Some users rely on this
312 feature even when it is combined with -M/-f.
313
314 9. Fixed restore when restoring huge backups (where rstdir...
315 temporary files are over 2GB). Thanks to
316 Raphael Jaffey <rjaffey@artic.edu> for reporting this,
317 debugging the issue and making the patch.
318
319 10. Made restore understand FreeBSD UFS2 tapes. Thanks to
320 David <vrtin@users.sourceforge.net> for submitting the bug
321 and providing a test case.
322
323 11. Made dump work with an arbitrary integer level (no more 10
324 levels only limitation). Thanks to Matthew
325 <msvincen@midway.uchicago.edu> for the patch.
326
327 Changes between versions 0.4b34 and 0.4b35 (released December 21, 2003)
328 =======================================================================
329
330 1. Added a note in the dump man page saying that the default
331 blocksize can be 32 if -d is used with a high density argument.
332 Thanks to Antonios Christofides <A.Christofides@itia.ntua.gr>
333 for the patch.
334
335 2. Fixed configure to correctly understand CPPFLAGS, CFLAGS,
336 LDFLAGS environment variables. Thanks to Arcady Genkin
337 <antipode@users.sourceforge.net> for reporting the bug.
338
339 3. Made -e/-E options of dump accept an unlimited number of inodes
340 to be excluded, rather than a hardcoded maximum. Thanks to
341 Dietrich Rothe <d-rothe@users.sourceforge.net> for the patch.
342
343 4. Updated the autoconf system to 2.50. Forced the -D_BSD_SOURCE
344 and -D_USE_BSD_SIGNAL defines in configure in order to solve
345 64bit build problems because quad_t is redifined with a
346 different signature. Thanks to Mike Harris <mharris@redhat.com>
347 for reporting this bug.
348
349 5. Made restore build on Solaris, making possible to
350 restore Linux's "enhanced" tapes. Thanks to Uwe Gohlke
351 <uwe@ugsoft.de> for the patch.
352
353 6. Made an extension in the dump tape format capable of saving
354 MacOSX specific inode extensions. Uwe Gohlke <uwe@ugsoft.de>
355 wrote the extension and contributed the restore code back
356 into this codebase. The same extension mechanism will be
357 used in the future to save ACLs...
358
359 7. Made rmt work correctly with regard to QFA and local/remote
360 files and tapes. The remote access will however work only
361 when the dump provided rmt version is used. If you want to
362 use another rmt server, please do not use the QFA feature.
363 Thanks to Uwe Gohlke <uwe@ugsoft.de> for the patch.
364
365 Changes between versions 0.4b33 and 0.4b34 (released April 18, 2003)
366 ====================================================================
367
368 1. Fixed the INSTALL file to reflect the actual install paths.
369 Thanks to David Ronis <ronis@ronispc.chem.mcgill.ca> for
370 reporting the bug.
371
372 2. Fixed the configure script to only check for headers presence
373 instead of trying to compile them. This should fix issues
374 with old build environments. Thanks to Kari Lempiainen
375 <kari@funky.pp.fi> for reporting the bug.
376
377 3. Fixed restore to correctly ignore sockets when comparing
378 a dump (as socket cannot be properly restored anyway). Thanks
379 to Gunther Reiszig <gunther@mit.edu> for reporting the bug.
380
381 4. Fixed restore to correctly access the archive file (-A argument)
382 even when using a remote tape. Thanks to Fabrice Bellet
383 <fabrice@bellet.info> for reporting the bug.
384
385 5. Fixed (again) handling of long (largefile) seeks in rmt.
386 Thanks to Fabrice Bellet <fabrice@bellet.info> for reporting
387 the bug.
388
389 6. Fixed restore corner case when dealing with large block sizes
390 dump is able to write now (-b 1024). Thanks to Fabrice Bellet
391 <fabrice@bellet.info> for reporting the bug.
392
393 7. Fixed a bug preventing dump to access a filesystem having
394 a label exactly 16 bytes in length. Thanks to <minduni@ti-edu.ch>
395 for reporting the bug.
396
397 8. Made dump store dump dates using explicit timezones, fixing a
398 problem with incremental dumps if the timezone is modified
399 between the dumps. Thanks to Philippe Troin <phil@fifi.org> for
400 the bug report and the patch.
401
402 9. Fixed a bug encountered when dumping individual files (not full
403 filesystems or directories) and dangling symbolic links happen
404 to be in the list of files. For as far as dump is concerned,
405 dangling symbolic links are allowed, and are dumped as is.
406 Thanks to Jin-su Ahn <jsahn@ee.snu.ac.kr> for reporting the
407 bug and providing the fix.
408
409 10. Fixed open and creation modes and permissions for QFA and
410 table-of-contents files in dump and restore. Thanks to
411 Philippe Troin <phil@fifi.org> for the patch.
412
413 11. Fixed the archive file descriptor handling enabling it to be 0.
414 This can happen in some cases when shell redirections are used.
415 Thanks to Philippe Troin <phil@fifi.org> for the patch.
416
417 12. Delayed the opening of archive file until after suid had been
418 dropped (fixing a possible security issue if dump is suid).
419 Thanks to Philippe Troin <phil@fifi.org> for the patch.
420
421 13. Fixed the 'S' command handling in the rmt client part.
422 Thanks to Philippe Troin <phil@fifi.org> for the patch.
423
424 14. Modified the end-of-tape script handling to print out statistics
425 (and stop the timer) before launching the eot script. Also, the eot
426 script does not get run anymore when using -M (which makes sense) or
427 when multiple tapes are listed on the command line
428 (-f tape0,tape1,tapen) (which also makes sense).
429 Thanks to Philippe Troin <phil@fifi.org> for the patch.
430
431 15. Relicensed dump/restore under the 'revised' BSD license, as per
432 ftp://ftp.cs.berkeley.edu/ucb/4bsd/README.Impt.License.Change.
433
434 16. Added LZO compression to dump. This new compression method has
435 the advantage of being super fast, thus not killing tape streaming
436 on slow machines. Thanks to Helmut Jarausch
437 <jarausch@igpm.rwth-aachen.de> for the patch and to
438 Markus Oberhumer <markus@oberhumer.com> for giving special permission
439 to include his miniLZO project (GPL licensed) in dump/restore.
440
441 17. Some small buffer overruns fixes in rmt. Thanks to Antonomasia
442 <ant@notatla.demon.co.uk> for reporting the bugs.
443
444 18. Added a special rmt version which can do encryption when writing
445 to tape. Read examples/encrypted_rmt/README for details on
446 how to enable and configure it. Thanks to Ken Lalonde
447 <ken@globalremit.com> for the patch.
448
449 19. Made dump work with 2.5 kernel end of tape early warning semantics.
450 Thanks to Kai Makisara <Kai.Makisara@kolumbus.fi> for the patch.
451
452 20. Fixed a bug which caused dump -w|-W not to work anymore, because
453 the fs_freq and fs_passno fields in /etc/mtab are always set
454 to 0 0. Thanks to Trent Piepho <xyzzy@speakeasy.org> for
455 reporting the bug.
456
457 Changes between versions 0.4b32 and 0.4b33 (released February 10, 2003)
458 =======================================================================
459
460 1. Added a note in the restore man page clarifying the question
461 'set the permissions on the current directory ?' asked by
462 restore at the end of treatment in -i and -x modes.
463
464 2. Fixed the endianess issues when reading compressed tapes.
465 Thanks to Dark Force <daq4th@users.sourceforge.net> for
466 reporting this bug and providing test cases.
467
468 3. Fixed the "ACL won't be dumped" warning message (which showed
469 an extra, unrelated error message). Thanks to Dragan Krnic
470 <dkrnic@lycos.com> for reporting this bug.
471
472 4. Made dump look first into /etc/mtab, then into /etc/fstab
473 when searching for filesystem paths. Also fixed some problems
474 caused by binding mounts or multiple block device mounts.
475 Thanks to Matus Uhlar <uhlar@fantomas.sk>, Elliott Mitchell
476 <ehem@m5p.com>, Greg Edwards <gedwards@users.sourceforge.net>,
477 Brian Hoy <brian.hoy@opus.co.nz>. (fixes Debian bugs #147086
478 and #160305, Sourceforge bugs #618699 and #679832).
479
480 5. Made dump's -I option accept the value '0' meaning all the
481 read errors will be ignored. This can be useful when running
482 dump from unattended sessions (like cron jobs). Thanks to
483 John I Wang <jiwang@users.sourceforge.net> for the suggestion.
484
485 6. Fixed the output of dump to indicate 'blocks' instead of
486 'tape blocks' in the various messages (blocks are always
487 1 Kilobyte, tape blocks are 1 BK * '-b' argument), and
488 made it clearly print the current blocksize at the start of
489 a dump. Thanks to Michal Szymanski <msz@astrouw.edu.pl> for
490 the suggestions.
491
492 7. Made rmt understand long (largefiles) seeks.
493
494 8. Fixed build with very old versions of libext2fs, where
495 EXT2_FT_* constants were undefined.
496
497 9. Made dump accept the dumpdates path on the command line
498 (-D file option) instead of using only the hardcoded one.
499 Thanks to Piete Brooks <pb22@users.sourceforge.net> for the
500 suggestion.
501
502 10. Enabled rmt, LFS, readline, QFA options by default in
503 ./configure. Updated the configure process (new versions
504 of config.guess, config.sub etc).
505
506 Changes between versions 0.4b31 and 0.4b32 (released November 15, 2002)
507 =======================================================================
508
509 1. Changed dump to use fcntl(F_SETLK) style locking instead
510 of flock() when locking the dumpdates file. With the old
511 locking scheme, a local user having read rights on the
512 dumpdates file could be able to do a Denial of Service attack
513 on dump. In order to lock the dumpdates file with the new
514 scheme, the user would need to have write access on the file.
515 Thanks to Richard Johnson <Richard.Johnson3@ey.com> for
516 reporting the bug (originally a bugtraq post).
517
518 2. Fixed interactive 'ls' which caused spurious errors warnings
519 about 'undefined filetypes' detected. Thanks to Jorgen Ostling
520 <jorgen_ostling@users.sourceforge.net> for reporting this
521 bug.
522
523 3. Fixed dump's estimate when dealing with sparse inodes.
524
525 4. Modified dump to allow setting a blocksize bigger than 32kB
526 (raised the limit to 1024kB), because newer hardware needs
527 this for proper operation. Thanks to Dirk Traenapp
528 <dtraenapp@users.sourceforge.net> for reporting this.
529
530 5. Fixed a bug causing Dump to stop and report an error if an
531 inode number in the exclude file was followed by some amount
532 of whitespace. Thanks to Jeffrey Sofferin
533 <sofferin@users.sourceforge.net> for reporting this bug.
534
535 6. Fixed a bug which caused restore, in some particular cases,
536 to ask some 'scary' questions and leave a bunch of RSTTMP
537 directories behind when restoring incremental tapes. Thanks
538 to Philippe Troin <phil@fifi.org> for reporting this bug and
539 providing the test cases.
540
541 7. Changed the wording when inodes are excluded from dump:
542 replaced 'Added inode 7 to exclude list' with
543 'Excluding inode 7 (resize inode) from dump', as suggested
544 by Elliott Mitchell <ehem@m5p.com> in a Debian bug report.
545
546 Changes between versions 0.4b30 and 0.4b31 (released July 30, 2002)
547 ===================================================================
548
549 1. Fixed rmt open flags transmission (GNU's symbolic syntax over
550 rmt) which I broke in 0.4b29. Thanks to Eros Albertazzi
551 <eros@lamel.bo.cnr.it> for reporting the bug.
552
553 Changes between versions 0.4b29 and 0.4b30 (released July 25, 2002)
554 ===================================================================
555
556 1. Made dump print out the ext2 logical block number in case of a read
557 error, which can be given as an argument to debugfs' ncheck command,
558 in order to find out the affected inode. Added note about this
559 usage in dump's man page.
560
561 2. Fixed a problem in restore when reading tapes written on big
562 endian machines with very old versions of dump. The patch was
563 contributed by George Helffrich <george@geology.bristol.ac.uk>.
564
565 3. Fixed the tape length calculation when using large tapes
566 and compression. Thanks to Georg Lippold
567 <g_lippold@sourceforge.net> for reporting the bug.
568
569 4. Added a new set of examples in dump_on_cd_2 directory, based
570 on dump_on_cd examples but somewhat enhanced, supporting DVD
571 media, and localized in english and german. Thanks to
572 Georg Lippold <g_lippold@sourceforge.net> for the new scripts.
573
574 5. Made dump save 32 bit UID/GID. Internally, this was achieved
575 by switching from the old BSD inode format to the new BSD
576 inode format, which means that the tape format was changed.
577 However, since all restore versions out there should
578 transparently support both inode formats, the change should
579 have no side effects. Thanks to John Yu <jky@cs.bu.edu> for
580 reporting the bug.
581
582 6. Fixed a lot of warnings in the code shown when compiling
583 with 'gcc -W'. Thanks to Matthias Andree
584 <matthias.andree@stud.uni-dortmund.de> for reporting this.
585
586 7. Fixed a small markup bug in the dump man page. Thanks to
587 Eric S. Raymond <esr@minx.thyrsus.com> for submitting the
588 patch.
589
590 8. Rewrote entirely the man pages using the tmac.an macro
591 package (Linux man page format) instead of the original BSD
592 format. They should be now cleaner and easier to modify.
593
594 Changes between versions 0.4b28 and 0.4b29 (released June 8, 2002)
595 ==================================================================
596
597 1. Fixed a problem in the rmt ioctl command, where ioctl's issued from
598 non Linux clients were misinterpreted. The description of the problem
599 (incompatible numbering in Linux mtio opcodes) is documented at
600 ftp://ftp.fokus.gmd.de/pub/unix/star/README.mtio . Thanks to
601 Jörg Schilling <schilling@fokus.gmd.de> for reporting this bug and
602 providing an excellent, cross-platform replacement for rmt in his
603 star package.
604
605 2. Fixed a bug reported by Andy Mentges <amentges@jumpline.com>
606 which caused restore to fail when the inode maps were not entirely
607 contained on the first volume (when using really small volumes or
608 when dumping a huge number of inodes).
609
610 3. Fixed a problem in dump, where files in subdirectories of directories
611 that have the nodump flag set (or being excluded with -e/-E) were
612 sometimes incorrectly dumped (instead of being excluded). The original
613 fix is from David C Lawrence <tale@dd.org> in the FreeBSD version
614 of dump: http://www.FreeBSD.org/cgi/query-pr.cgi?pr=32414
615 Thanks to Ted Grzesik <tedgyz@roostme.com> for reporting the bug and
616 help testing the patch.
617
618 4. Added some example scripts from Gerd Bavendiek <bav@epost.de>
619 which makes one able to pipe the output of dump, by the net, to
620 a remote CD-burner server.
621
622 5. Made dump use O_CREAT|O_TRUNC both locally and remotely (over rmt),
623 and use GNU's symbolic syntax over rmt instead of numerical values
624 to assure multiple platform compatibility.
625
626 6. Documented the -d option in restore.
627
628 7. Added a -v (verbose) mode to dump. For now it just prints the number
629 of the inode being dumped, but this could evolve in future versions
630 to include interesting debugging output.
631
632 8. Added a -o flag to restore, which automatically restores the current
633 directory permissions in -i or -x mode, without asking the operator
634 whether to do so. Patch submitted by Tony Clayton <tonyc@e-smith.com>
635 and Peter Samuel <peters@e-smith.com>.
636
637 Changes between versions 0.4b27 and 0.4b28 (released April 12, 2002)
638 ====================================================================
639
640 1. Fixed a bug in the QFA feature which could have caused
641 a bad tape positionning on restore (causing extra delays in
642 restore). Credits go to Uwe Gohlke <uwe@ugsoft.de>.
643
644 2. Added a small note in the dump man page specifying that
645 there should be no white space between the option letter and
646 the -j or -z parameter, thanks to Kenneth Porter <shiva@well.com>
647
648 3. Made restore work with older versions of the readline library,
649 by checking in configure for several extended symbols. Restore
650 can now be compiled with a readline as old as the 2.0 release
651 (though it may be a good idea to upgrade it to a more recent
652 version...). Thanks to Andrew Donkin <ard@waikato.ac.nz> for
653 reporting the build failures.
654
655 4. Fixed a performance problem with the QFA file creation in
656 dump, which made unnecessary seeks on the tape slowing down
657 the dump. Thanks to Andrew Donkin <ard@waikato.ac.nz> for
658 reporting this issue.
659
660 5. Removed the inclusion of some kernel headers in the dump
661 source, which prevented the compile in some kernel/glibc
662 headers/architecture combination. Thanks to Bdale Garbee
663 <bdale@gag.com> for reporting the bug.
664
665 6. Added the appropriate error message when dump fails to
666 open the output file for writing. Thanks to Amith Varghese
667 <amithv@yahoo.com> for reporting this bug.
668
669 7. Made restore able to understand large Solaris ufsdump tapes
670 (containing inodes bigger than 4194304). Sun have introduced
671 an "extension" to the dump tape format when dealing with
672 those inodes, which was uncorrectly handled by Linux restore.
673 Thanks to Uwe Gohlke <uwe@ugsoft.de> for reporting the bug and
674 providing a test case.
675
676 8. Added the -m parameter to dump which optimises the output for
677 inodes having been changed but not modified since the last dump
678 ('changed' and 'modified' have the meaning defined in stat(2)).
679 For those inodes, dump will save only the metadata, instead of
680 saving the entire inode contents. Inodes which are either
681 directories or have been modified since the last dump are saved
682 in a regular way. Uses of this flag must be consistent, meaning
683 that either every dump in an incremental dump set have the flag,
684 or no one has it.
685
686 Changes between versions 0.4b26 and 0.4b27 (released February 15, 2002)
687 =======================================================================
688
689 1. Fixed behaviour of dump when exceeding resource limits
690 (SIGXFSZ treatment).
691
692 2. Added the -L flag to restore to allow the user to specify a
693 maximal allowed number of miscompares when using restore
694 with the -C option to check the backup.
695
696 3. Detailed the manual entry for the -N option of restore.
697
698 4. Added the -a flag to restore to make able doing unattended
699 restores in -i or -x mode (automatically walks through the
700 multiple dump volumes).
701
702 5. Extended the QFA mode to work with local files and/or
703 remote tapes and files. This way, restore can know in advance
704 the tape number and the offset for the inodes to extract and
705 can minimize the extraction time by seeking directly to the
706 good tape/offset.
707
708 6. Added the -A <archive> option to both dump and restore,
709 which makes dump to archive a dump table-of-contents in
710 the specified file to be used by restore to determine
711 whether a file is in the dump file that is being restored.
712 (the archive file syntax is also compatible with the
713 Solaris ufsdump generated one).
714
715 7. Small fix in restore making it able to read some (broken ?)
716 Solaris ufsdump tapes.
717
718 8. Fixed dump to correctly recognise the root filesystem when using
719 ext2 disk labels (LABEL=/). Thanks to John Yu <jky@it.bu.edu>
720 for reporting this bug.
721
722 9. Added the -P <file> option to restore to create a
723 Quick File Access file from an already made dump. Patch
724 contributed by Uwe Gohlke <uwe@ugsoft.de>.
725
726 10. Made restore compile and run on Solaris, making it a
727 possible replacement for the standard ufsrestore. Port was
728 contributed by Uwe Gohlke <uwe@ugsoft.de>.
729
730 Changes between versions 0.4b25 and 0.4b26 (released January 7, 2002)
731 =====================================================================
732
733 1. Added a set of backup scripts from Eugenio Diaz
734 <getnito@yahoo.com> in the examples section. It features
735 automatic (cron based) full and incremental dumping of
736 several filesystems on a separate filesystem.
737
738 2. Fixed a off-by-one miscalculation which disabled dumping
739 a one letter subdirectory. Thanks to Chris Siebenmann
740 <cks@utcc.utoronto.ca> for reporting the bug.
741
742 3. Fixed several restore bugs occuring when trying to
743 restore some missing files on the tape. Thanks to Chris
744 Siebenmann <cks@utcc.utoronto.ca> for reporting the bug.
745
746 4. Fixed --with-ldopts configure argument passing, installing from
747 a separate object directory, makefile cleanups contributed
748 by <splite@purdue.edu>.
749
750 5. Fix a bug which could caused, in some conditions, the highest
751 number inode of a filesystem, to not be dumped. Many thanks
752 to Chris Siebenmann <cks@utcc.utoronto.ca> for helping me
753 chase this bug.
754
755 Changes between versions 0.4b24 and 0.4b25 (released November 17, 2001)
756 =======================================================================
757
758 1. Added a mini howto from Patrick Walsh in the examples
759 directory.
760
761 2. Minor man pages syntax corrections. Thanks to
762 Chris Danis <screechco@home.com> for reporting the bugs.
763
764 3. Added a script from David B. Peterson <dave@toppledwagon.com>
765 to the examples section. It features dumping several
766 filesystems to a remote tape drive upon ssh.
767
768 4. Added a patch provided by Richard Jones <rich@annexia.org>
769 which allows BRADEMAX (number of read errors tolerated by
770 dump) to be adjusted using the -I option.
771
772 5. Fixed a bug which disabled doing "restore -C -f -". Thanks
773 to Clemens Stiglechner <a9401816@unet.univie.ac.at> for the
774 patch.
775
776 6. Add the -l option to restore to specify if, when doing a
777 remote restore, the file used to access the backup is a
778 regular file (the defaults being a tape drive). Restore needs
779 to know this information when reading a remote compressed
780 dump. Previously, this information was autodetected, but
781 the autodetection code fails (with ioctl: Inappropriate ioctl
782 for device) when using a non Linux remote box. Thanks to
783 many users and especially to Eros Albertazzi
784 <eros@lamel.bo.cnr.it> for reporting this.
785
786 7. Found a workaroung for the dump deadlock problem (3 childs
787 stuck in pause(), father in read()). The workaround seems
788 to work for me and several beta-testers. If it doesn't work
789 for you, please report back.
790
791 8. Updated the RPM spec file (BuildPrereq, URL etc).
792
793 Changes between versions 0.4b23 and 0.4b24 (released September 12, 2001)
794 ========================================================================
795
796 1. Fixed the permissions of a newly created QFA file by dump.
797
798 2. Cleaned up the source of dump (the external variables
799 definition was a complete mess, making possible to have
800 objects overlap).
801
802 3. Fixed restore to use the full tape volume path when doing
803 a compare (since it changes the working directory to the
804 filesystem being compared in the process).
805
806 4. Added the -q option to dump which makes dump abort
807 whenever it needs operator attention. It should be
808 easier to use dump in scripts with this option.
809
810 5. Detect the use of incompatible options to dump and
811 refuse them (like -a and -B options together).
812
813 6. Added bzip2 compression to dump/restore (use option -j level
814 to select it). Note that you will need a restore version
815 >= 0.4b24 in order to restore a bzip2 compressed dump.
816 The same warning as for the zlib compression applies:
817 the tape format of a bzip2 dump is not compatible with the
818 original BSD tape format.
819
820 7. Fixed a overflow problem in dump corrupting the dump when
821 very large files were encountered. Thanks to Vladimir Nadvornik
822 for the bug report.
823
824 8. Added a ioctl(BLKFLSBUF) in dump which should flush the
825 kernel buffer/page cache before starting the dump, helping
826 a bit those who use dump on mounted filesystems. Thanks to
827 John Yu <jky@it.bu.edu> and to Theodore T'so <tytso@mit.edu>
828 for this suggestion.
829
830 9. Updated the RPM spec file following the RedHat changes
831 (dynamically linked binaries now in /usr/sbin etc).
832
833 10. Added a patch from Helmut Jarausch <jarausch@igpm.rwth-aachen.de>
834 which enables restore to recognise multi volume compressed dumps
835 done on CD. Included his scripts for dump (which pipe the dump
836 output directly into cdrecord) and restore. There is now
837 possible to to backups to CD on the fly!
838
839 Changes between versions 0.4b22 and 0.4b23 (released July 20, 2001)
840 ===================================================================
841
842 1. Fixed a buffer overflow in restore/tape.c. Patch provided by
843 Marcus Meissner (Caldera International Security Dept.).
844
845 2. Implement the Sun rmt extended protocol. Patch submitted
846 by Ian Gordon <iangordon@users.sourceforge.net>.
847
848 3. Check for the e2fsprogs header <ext2fs/ext2_fs.h> instead of
849 the linux kernel header. This ensures that dump always has the
850 latest version of this file. Patch submitted by
851 Andreas Dilger <adilger@turbolinux.com>.
852
853 4. Report any filesystem present in either /etc/fstab with a
854 positive passno or /etc/dumpdates in dump -w output.
855 Patch submitted by Andreas Dilger <adilger@turbolinux.com>.
856
857 5. Fixed the looping problem in dump introduced in the
858 previous version.
859
860 6. Changed the -B option of dump to limit the size of
861 _compressed_ output per volume if compression is on.
862 Patch contributed by Helmut Jarausch
863 <jarausch@igpm.rwth-aachen.de>. Note however that, since
864 it is impossible to predict the size of the compressed
865 data before doing the compression, the -B limit is a bit
866 conservative.
867
868 7. Fixed a bug in reading the operator typed file/tape path for
869 the next volume in restore.
870
871 8. Implemented a "-F script" option to restore which permits the
872 user to specify a script which will be launched at the
873 beginning of each tape, useful for automatic programming of
874 tape changers for example. See the restore man page for the
875 script parameters and return codes.
876
877 9. Small fix for the QFA routines provided by Uwe Gohlke
878 <uwe@ugsoft.de>, and some recommendations for QFA uses in
879 the man pages.
880
881 10. Fixed the multivolume restoring where making a mistake
882 to the 'Mount next tape' prompt caused several blocks to
883 be lost.
884
885 11. Enhanced the -e option of dump to take as a parameter a
886 comma separated list of inode numbers.
887
888 12. Added the -E option to dump which specify a file containing
889 inode numbers to exclude from the dump.
890
891 13. Fixed the compressed multi-volume dump + restore.
892
893 Changes between versions 0.4b21 and 0.4b22 (released May 12, 2001)
894 ==================================================================
895
896 1. Made dump report the number of blocks dumped per volume.
897 Thanks to Kenneth Porter <shiva@well.com> for the suggestion.
898
899 2. Fix a bug in dump where waiting too much at the 'change volumes'
900 question caused the volume to be rewritten. Thanks to
901 Graham Reed <greed@users.sourceforge.net> for reporting the
902 bug and providing a patch.
903
904 3. Added a compression option to dump, contributed by Jerry
905 Peters <gapeters@worldnet.att.net>.
906
907 WARNING: the compression works only on files/pipes and
908 drives supporting variable block size.
909
910 WARNING2: tapes created using the compression feature are
911 incompatible with the standard dump tape format, and a
912 version of restore >= 0.4b22 will be needed for extraction.
913
914 4. Fixed some compilation problems for glibc 2.2.2 and 64 bit
915 architectures. Thanks to Paul Slootman <paul@debian.org> for
916 the patch and to Bdale Garbee <bdale@gag.com> for forwarding
917 it upstream.
918
919 5. Many cleanups (CPP defines, const char warnings, check of
920 ext2fs COMPAT flags, time_t cleanups, added libext2 version
921 in dump usage text) by Andreas Dilger <adilger@turbolinux.com>.
922
923 6. Made --prefix option work in configure. All the install path
924 are now based on the configure parameters.
925
926 7. Added the Quick File Access mode in dump/restore, contributed
927 by Uwe Gohlke <uwe@ugsoft.de>. In this mode, dump stores in
928 a file tape position for each inode, and this file is used by
929 restore (if called with parameter Q and the filename)
930 to directly position the tape at the file restore is currently
931 working on. This saves hours when restoring single files from
932 large backups, saves the tapes and the drive's head. Use
933 --enable-qfa option of configure to compile in the QFA support.
934
935 8. Added the possibility to dump several files and directories
936 in a single invocation of dump. Thanks to Uwe Gohlke
937 <uwe@ugsoft.de> for implementing this option.
938
939 9. Fixed the dumping and restoring of files with holes
940 (files ending with holes didn't get dumped or restored
941 correctly).
942
943 10. Fixed a socket descriptor leak in dump, which leaved opened
944 3 file descriptors per dump process (and there is one dump
945 process per tape).
946
947 11. Fixed dump large file system support, by explicit use of
948 open64/lseek64/etc functions (explicit use needed because
949 e2fsprogs libraries don't behave well when compiled with
950 FILE_OFFSET_BITS=64).
951
952 Changes between versions 0.4b20 and 0.4b21 (released January 13, 2001)
953 ======================================================================
954
955 1. Fixed some bugs in the dump -w|-W logic introduced by
956 the previous version. Thanks to Andreas Dilger
957 <adilger@turbolinux.com> for his help on this one.
958
959 2. Fixed again a compilation problem when using old e2fs
960 headers (filesystem label related). Thanks to many users
961 who reported this stupid error.
962
963 3. Fixed a build problem on old lib5 based systems dealing with
964 _PATH_MNTTAB being undefined. Thanks to John Adams
965 <johna@onevista.com> for reporting the bug.
966
967 4. Improved the error detection in restore when creating
968 the temporary files in TMPDIR. Now restore will corectly
969 report a 'no space left on device' error instead of
970 strange errors which could imply an archive corruption.
971 Thanks to Gabriel Somlo <somlo@cs.colostate.edu> and
972 bgjenero <bgjenero@sympatico.ca> for reporting the bug.
973
974 5. Added the throughput information in dump's progression
975 messages. Thanks to Andreas Dilger <adilger@turbolinux.com>
976 for the patch.
977
978 6. Use libext2fs's inode scanning routines, which are particularly
979 robust when dealing with errors (especially when having some
980 bad blocks on the filesystem). Many thanks to Colin
981 <colin@horizon.com> for reporting the bug, investigating
982 the issues, finding a workaround, writing the patch and
983 fully testing it... (of course, if this patch will break
984 anything, he is to blame for :-)).
985
986 7. Made dump and restore LFS aware. Dump can dump a filesystem
987 containing large files, generate a large file on output and
988 restore can restore them. This feature is not enabled by
989 default, use --enable-largefile option of configure to enable
990 it (you need to have a LFS aware glibc though). Thanks to
991 Andrea Arcangeli <andrea@suse.de> for submitting the patch,
992 and to Theodore T'so <tytso@valinux.com> for his always
993 useful thoughts.
994
995 8. Made dump ask upon a tape write error if it should rewrite
996 the current volume (assume this is a bad tape) or if it should
997 assume an end-of-tape condition (useful for tape drives which
998 behaves badly at the end of the tape). Thanks to Andreas
999 Dilger <adilger@turbolinux.com> for the suggestion.
1000
1001 Changes between versions 0.4b19 and 0.4b20 (released November 10, 2000)
1002 =======================================================================
1003
1004 1. Fixed a small compilation problem due to a change
1005 in the definintion of the struct sigaction in
1006 glibc 2.0 / libc5. Thanks to Gunther Schlegel
1007 <schlegel@riege.de> for reporting the bug and to
1008 Dave Platt <dplatt@snulbug.mtview.ca.us> for suggesting
1009 a fix.
1010
1011 2. Modified the label and uuid analysis in order to be
1012 self-contained (independant of kernel/libc headers). This
1013 should fix the compile with older kernel/libc headers and
1014 will preserve the functionality. Thanks to Bernhard Erdmann
1015 <bernhard.erdmann@gmx.de> for reporting the bug.
1016
1017 3. The 'exclude inode' option, if applied to a directory
1018 inode, excludes now correctly all the leaves of this
1019 directory. Thanks to John R. Dennison
1020 <gerdesas@users.sourceforge.net> for reporting the bug.
1021
1022 4. Fixed the '-e' option to disable the possibility
1023 to exclude the root inode (which causes the dumps to
1024 be unreadable by restore). Prevented array overflow
1025 when multiple -e options are used.
1026
1027 5. Fixed dump to correctly interpret a filesystem argument
1028 which is a mountpoint and it is not an absolute path
1029 (as specified in the fstab). Thanks to Bernhard R. Erdmann
1030 <be@berdmann.de> for reporting the bug.
1031
1032 6. Made dump able to backup files larger than 2 GB. Note that
1033 dump still doesn't cope with files larger than 4 GB.
1034
1035 7. Restore the real uid and gid before invoking an external
1036 RSH program (big hole when dump or restore is suid root!).
1037
1038 8. Get the values for the default tape device and the location
1039 of fstab file from the system headers. Thanks to
1040 Andreas Dilger <adilger@turbolinux.com> for the patch.
1041
1042 9. Made dump -w|-W report all recognized filesystems
1043 present in either /etc/fstab or /etc/dumpdates, and present
1044 the list in the same order as in fstab file. Thanks
1045 to Andreas Dilger <adilger@turbolinux.com> for the patch.
1046
1047 10. Made dump's -a (automatic end of tape detection) the
1048 default. Specifying one of -B, -c, -d or -s options will
1049 override the EOM detection. Thanks to Andreas Dilger
1050 <adilger@turbolinux.com> for the patch.
1051
1052 11. Save the ext2 filesystem label into the dump volume label.
1053 Specifying a volume label on the command line (-L option)
1054 will override this feature. Thanks to Andreas Dilger
1055 <adilger@turbolinux.com> for the patch.
1056
1057 Changes between versions 0.4b18 and 0.4b19 (released August 20, 2000)
1058 =====================================================================
1059
1060 1. Fixed the signal handling in dump (which I broke in 0.4b17)
1061 which was causing several strange dump failures (dump
1062 hanged or segmentation faults etc).
1063
1064 2. Specified the default blocksize in the dump man page.
1065
1066 3. Changed two info messages of restore to be written on stdout
1067 instead of stderr in order to leave stderr only for errors
1068 or warnings. Thanks to Stephen Atwell
1069 <satwell@urbana.css.mot.com> for the suggestion.
1070
1071 4. Corrected an off by one calculation which prevented
1072 dumping a directory having a 1 character name length.
1073 Thanks to Bernhard Erdmann <bernhard.erdmann@gmx.de>
1074 for reporting the bug.
1075
1076 5. Reinforce dump in order to allow only level 0 and no
1077 -u option when dumping a subdirectory, and document
1078 this in the man page. Thanks to Bernhard Erdmann
1079 <bernhard.erdmann@gmx.de> for reporting the bug.
1080
1081 6. Fixed a small harmless bug in restore which caused
1082 in some conditions hard links to be restored several
1083 times (therefore generation some warning messages).
1084 Thanks to Kenneth Porter <shiva@well.com> for
1085 reporting the bug.
1086
1087 7. Updated the RPM spec file to the latest RedHat version,
1088 providing FHS packaging and other cosmetic changes.
1089 You will need RPM version >= 3.0.5 in order to build the RPM.
1090
1091 8. Updated the configure script to check for libtermcap
1092 before checking for libreadline, because we need this
1093 library in order to compile the readline support.
1094
1095 9. Made dump understand the LABEL= and UUID= notation
1096 both in /etc/fstab and on the command line. Note that
1097 you will need the /proc filesystem in order to use
1098 these notations. Thanks to Erik Troan <ewt@redhat.com>
1099 for providing the patch.
1100
1101 Changes between versions 0.4b17 and 0.4b18 (released June 30, 2000)
1102 ===================================================================
1103
1104 1. Fixed a potential buffer overflow in restore. Thanks
1105 to Stan Bubrouski <satan@fastdial.net> for reporting
1106 the bug.
1107
1108 2. Fixed a readline-related bug which prevented
1109 'cat DUMPFILE | restore -i -f -' from working. Thanks
1110 to Charles Karney <karney@users.sourceforge.net>
1111 for the bug report.
1112
1113 3. Changed a few "panic" into "exit", causing restore to
1114 be more stable against some attacks (like the last one
1115 reported on Bugtraq, although the last version of restore
1116 was not vulnerable - just dumped core). Thanks to
1117 Andreas Hasenack <andreas@conectiva.com.br> for reporting
1118 the bugs to me.
1119
1120 4. Removed the suid-root bit on dump and restore in the
1121 default build (and generated RPMs). It should be safer
1122 now. Users who need the suid-root capabilities in order
1123 to make network backups should read first the man page
1124 of dump and enable the suid bit manually.
1125
1126 5. Added -ltermcap to the compile parameters for restore
1127 when using readline support, in order to make the compile
1128 process work on some systems (SuSE ?). Thanks to
1129 Patrik Schindler <poc@pocnet.net> for reporting the bug.
1130
1131 Changes between versions 0.4b16 and 0.4b17 (released June 1st, 2000)
1132 ====================================================================
1133
1134 1. The -F script is called now *only* at the end of a tape,
1135 not at the end of the dump. Thanks to Christian Haul
1136 <haul@informatik.tu-darmstadt.de> for the bug report.
1137
1138 Normally, the device name and the current volume number
1139 are passed on the command line of the script. However,
1140 if you want the old style script (with no arguments
1141 passed on the command line), you can enable it in
1142 configure with the --enable-oldstylefscript.
1143
1144 2. Use posix signal handling to preserve dump functionality
1145 with libc5. Thanks to Jeff Johnson <jbj@redhat.com> for
1146 the patch.
1147
1148 3. Made the exit code of restore in case of a 'C'ompare
1149 command reflect the result of the compare. An exit status
1150 of 0 means the dump archive is correct, 1 means tape errors,
1151 2 means that some files were modified. Thanks to Kenneth Porter
1152 <shiva@well.com> for the suggestion.
1153
1154 4. Made (finally) quotes work in interactive restore.
1155
1156 5. Small fixes in order to allow dump to compile with a
1157 really old e2fsprogs version. Thanks to Ian Zimmerman
1158 <itz@speakeasy.org> for the bug report.
1159
1160 6. Add GNU readline capabilities to interactive restore.
1161 Use configure's --enable-readline flag to enable this feature.
1162 Thanks to Patrik Schindler <poc@pocnet.net> for the
1163 suggestion.
1164
1165 7. Do the compare on the fly in restore 'C' mode (this will
1166 allow not to exhaust the available /tmp space when
1167 ccmparing large files). Thanks to Kenneth Porter
1168 <shiva@well.com> for the suggestion.
1169
1170 Changes between versions 0.4b15 and 0.4b16 (released March 11, 2000)
1171 ====================================================================
1172
1173 1. Fixed some several duplicate 'DUMP: DUMP:' in the
1174 output of dump.
1175
1176 2. Corrected the estimation of blocks to dump. Note that
1177 this is still wrong for large directory inodes, due
1178 to the size of a BSD directory entry on the tape
1179 (estimation is based on the size of the ext2 directory,
1180 which is smaller because it doesn't contain the null
1181 character at the end).
1182
1183 3. Made dump report the total number of MB written to
1184 tape at the end of run. Thanks to W. Reilly Cooley
1185 <wcooley@nakedape.cc> for the patch.
1186
1187 4. Added the -X option to restore giving the possibility
1188 to read the names of the files to be extracted/listed
1189 from a text file (in addition of the command line).
1190 Thanks to Dejan Muhamedagic <dejan@quant-x.com> for the
1191 patch.
1192
1193 5. Added the device name and the current volume number
1194 as arguments to the end of tape script (dump -F option).
1195
1196 6. Made the multi-volume dump work again (after having
1197 broken it in 0.4b15).
1198
1199 Changes between versions 0.4b14 and 0.4b15 (released March 2, 2000)
1200 ===================================================================
1201
1202 1. Added a prompt command in interactive restore mode. Thanks
1203 to Andreas Dilger <adilger@home.com> for the patch.
1204
1205 2. Fixed a buffer overflow problem in dump (caused by
1206 not checking the size of the filesystem parameter).
1207 Thanks to Kim Yong-jun <loveyou@hackerslab.org> for
1208 reporting this on Bugtraq (and to several dump users
1209 who forwarded me his mail).
1210
1211 3. Added the '-F script' option to dump in order to
1212 launch a script at the end of each tape (to be used
1213 with a tape changer, or to notify the sysadmin by
1214 pager etc.).
1215
1216 4. Fixed a bug in restore compare code caused by the changes
1217 I made in 0.4b14.
1218
1219 5. Fixed the treatment of options using the old BSD syntax
1220 in both dump and restore.
1221
1222 Changes between versions 0.4b13 and 0.4b14 (released February 10, 2000)
1223 =======================================================================
1224
1225 1. Fixed a bug in dump which may have caused invalid deleted
1226 directories to be dumped out if they were deleted after the
1227 mapping pass. This could occure on active filesystem and lead
1228 to heap corruption (causing dump malfunction in many possible ways).
1229 Thanks to Jan Sanislo <oystr@cs.washington.edu> for finding this
1230 bug and submitting the patch.
1231
1232 2. Fixed the handling of the filesystem parameter in dump. A
1233 '/mnt/disk' parameter caused the disk contents to be dumped,
1234 but a '/mnt/disk/' parameter caused the mountpoint directory
1235 to be dumped (generally an empty directory).
1236
1237 3. Improved the output of dump in order to tell which directory
1238 it is currently dumping (when dumping a subtree).
1239
1240 4. Added the '-e' exclude inode option to dump. Thanks to
1241 Isaac Chuang <ike@isl.stanford.edu> for contributing with the patch.
1242
1243 5. Added a REPORTING-BUGS file in order to provide a guide
1244 on how to correctly report a bug in dump/restore.
1245
1246 6. Fixed a restore problem when restoring a hard link to an inode
1247 having the immutable or append-only attribute set. Thanks to
1248 Ambrose Li <acli@mingpaoxpress.com> for submitting the patch.
1249
1250 7. Fixed a compatibility problem between dumps made on little
1251 endian machines (the format was correct) and big endian
1252 machines (the format was incorrect). This fix break the
1253 compatibility with the older dumps made on big endian
1254 machines (sparc, m86k, ppc etc). For the first time in
1255 linux dump's history (I believe), the dumps made by *BSD,
1256 Linux/alpha, Linux/sparc and Linux/x86 are compatible,
1257 so interchangeable. Thanks to Rob Cermak
1258 <cermak@ahab.rutgers.edu> for submitting the bug and
1259 helping me test the fix.
1260
1261 8. Fixed the way dump reports the remaining percent/time, if
1262 the number of blocks actually dumped exceeds the estimated
1263 number of blocks. Thanks to Jean-Paul van der Jagt
1264 <jeanpaul@dutepp0.et.tudelft.nl> for reporting the bug.
1265
1266 Changes between versions 0.4b12 and 0.4b13 (released January 21, 2000)
1267 ======================================================================
1268
1269 1. Small Y2K fix in the man pages :). Thanks to Bernhard Sadlowski
1270 <sadlowsk@Mathematik.Uni-Bielefeld.DE> for reporting the bug.
1271
1272 2. Removed the requirement to build the RPM as root from the
1273 spec file. Thanks to Christian Weisgerber
1274 <naddy@mips.rhein-neckar.de> for submitting this.
1275
1276 3. Fixed a bug in dump related to the 'filetype' feature of ext2fs,
1277 causing dump to block when dumping really huge filesystems.
1278 Many thanks to Patrik Schindler <poc@pocnet.net> for
1279 helping me find this bug.
1280
1281 4. Fixed the treatment for an interrupt signal when dump access
1282 the remote tape through RSH. Thanks to Christian Weisgerber
1283 <naddy@mips.rhein-neckar.de> for providing the patch.
1284
1285 5. Fixed a bug which was causing dump/restore to display
1286 garbage characters instead of the remote host name.
1287
1288 Changes between versions 0.4b11 and 0.4b12 (released January 8, 2000)
1289 =====================================================================
1290
1291 1. Small fix in the dump man page. Thanks to Thorsten Kukuk
1292 <kukuk@suse.de> for submitting the patch.
1293
1294 2. Fix for the exit code when using the size estimate option of
1295 dump. Thanks to Matti Taskinen <mkt@rni.helsinki.fi> for
1296 submitting the patch.
1297
1298 3. Handle EINTR in atomical reads/writes in dump, which was causing
1299 dump to fail on some systems. Thanks to Eric Jergensen
1300 <eric@dvns.com> for reporting the bug and submitting the patch.
1301
1302 4. Handle more than 16 characters for the device names in dumpdates.
1303 (up to 255 now). Thanks to Rainer Clasen <bj@ncc.cicely.de> for
1304 tracking down the problem and proposing the solution.
1305
1306 5. Fixed a bug in dump which prevented the creation of the
1307 dumpdates file when doing a 0-level dump without already
1308 having a dumpdates file. Thanks to Patrik Schindler
1309 <poc@pocnet.net> for reporting the bug.
1310
1311 6. Changed the way dump 'S' flag reports the size estimate
1312 from number of blocks into bytes (making it compatible
1313 with the Solaris version, and simplifying things for
1314 amanda users). Thanks to Jason L Tibbitts III
1315 <tibbs@math.uh.edu> for reporting the bug.
1316
1317 7. Fixed a compatibility problem in linux/alpha dump tape format.
1318 Now the linux/alpha dump are (again) compatible with the
1319 other dump formats. But this breaks compatibility with
1320 older dumps made on alpha. Thanks to Mike Tibor
1321 <tibor@lib.uaa.alaska.edu> for helping me in finding this bug.
1322
1323 Changes between versions 0.4b10 and 0.4b11 (released December 5, 1999)
1324 ======================================================================
1325
1326 1. Added a '--enable-kerberos' to configure.
1327
1328 2. Added a 'S' option to dump which determines the amount of space
1329 that is needed to perform the dump without actually doing it, similar
1330 to the Sun's ufsdump 'S' option. Patch contributed by Rob Cermak
1331 <cermak@ahab.rutgers.edu>.
1332
1333 3. Added a 'M' multi-volume option to dump and restore which enables
1334 dumping to multiple files (useful when dumping to an ext2
1335 partition to make several dump files in order to bypass the 2GB
1336 file size limitation). The 'f' argument is treated as a prefix and
1337 the output files will be named <prefix>001, <prefix>002 etc. With
1338 the 'M' flag, restore automatically selects the right file without
1339 asking to enter a new tape each time.
1340
1341 4. Fixed a memory leak which was causing dump to grow very big
1342 (270MB when dumping a 10GB filesystem...). Thanks to Jason
1343 Fearon <jasonf@netrider.org.au> for reporting the bug.
1344
1345 Changes between versions 0.4b9 and 0.4b10 (released November 21, 1999)
1346 ======================================================================
1347
1348 1. Make configure test if the system glob routines support
1349 extended syntax (ALTDIRFUNC). If not, use the internal glob
1350 routines instead of system ones. Thanks to Bernhard Sadlowski
1351 <sadlowsk@Mathematik.Uni-Bielefeld.DE> for reporting the bug
1352 and helping me resolve this and other minor libc5 compiling
1353 glitches.
1354
1355 2. Fix a problem when dumping a ext2fs with the 'filetype'
1356 feature enabled. Thanks to Patrick J. LoPresti
1357 <patl@cag.lcs.mit.edu> for reporting the bug and to
1358 Theodore Y. Ts'o <tytso@mit.edu> for providing the patch.
1359
1360 3. Made the nodump flag work on directories. A directory which
1361 has the nodump flag gets never dumped, regardless of its
1362 contents.
1363
1364 4. Integrate a patch from Jeremy Fitzhardinge <jeremy@goop.org>
1365 which allows dump on an active ext3 filesystem. However, this
1366 is a "quick and dirty" patch which enables backup of an ext3
1367 filesystem through the ext2 compatibility (by ignoring the
1368 NEEDS_RECOVERY bit). The journal file is not recognized and
1369 it is dumped (it should not).
1370
1371 5. Test the superblock compatibility flags when dumping, in order
1372 to be sure that we know how to deal with specific features.
1373
1374 Changes between versions 0.4b8 and 0.4b9 (released November 5, 1999)
1375 ====================================================================
1376
1377 1. Use lchown instead of chown, fixing a possible security problem
1378 when restoring symlinks (a malicious user could use this
1379 to deliberately corrupt the ownership of important system files).
1380 Thanks to Chris Siebenmann <cks@utcc.utoronto.ca> for detecting
1381 this and providing the patch.
1382
1383 Changes between versions 0.4b7 and 0.4b8 (released November 3, 1999)
1384 ====================================================================
1385
1386 1. Put dump sources under CVS, added Id tags in all files so
1387 one can use 'ident' on binary files.
1388
1389 2. Added the dump/restore version in the usage text so one can
1390 easily verify the version he is using.
1391
1392 3. Small patch from Nuno Oliveira <nuno@eq.uc.pt> which fixes
1393 a va_start/va_end problem on linux-ppc (always call va_start
1394 va_end in pairs each time we use a vararg function).
1395
1396 4. Added again the DT_* constants because old libc does not
1397 contain them :(. Thanks to Eric Maisonobe <virnet@nat.fr>
1398 for submitting the bug report.
1399
1400 5. Use ext2fs_llseek instead of llseek. With recent e2fsprogs
1401 this should enable dumping big (huge) filesystems.
1402
1403 6. Added the RSH environment variable in order to be able to
1404 use a rsh replacement like ssh when doing remote backups (and
1405 bypass the security limitations of rcmd). Now you can do remote
1406 backups without being root (or making dump setuid root).
1407
1408 7. Modified again the way dumpdates works. For incremental dumps,
1409 we need to read dumpdates even if we are not using 'u' option.
1410 Thanks to Bdale Garbee <bdale@gag.com> for his ideas on how
1411 this should work.
1412
1413 Changes between versions 0.4b6 and 0.4b7 (released October 8, 1999)
1414 ===================================================================
1415
1416 1. Removed the 'k' flag from the restore 'about' text if kerberos
1417 was not compiled in.
1418
1419 2. Prototyped (f)setflags from e2fsprogs and corrected the calls
1420 to them (fsetflags takes a char*, setflags an open fd!).
1421
1422 3. (f)setflags is called only if the flags aren't empty. If the
1423 file is a special file, a warning is printed, because changing
1424 flags implies opening the device. Normally, a special file
1425 should not have any flag... (Debian bug #29775, patch provided
1426 by Abhijit Dasgupta <abhijit@ans.net>).
1427
1428 4. Made possible to dump a file system not mentioned in /etc/fstab.
1429 (Debian bug #11904, patch provided by Eirik Fuller <eirik@netcom.com>).
1430
1431 5. Changed the default behaviour to not create dumpdates
1432 unless 'u' option is specified. Removed the old "debian-patch"
1433 which provided the same thing. (Debian bug #38136, #33818).
1434
1435 6. Removed all those dump*announce, since they were getting old...
1436
1437 7. Added warning messages if dumpdates does not exist and
1438 when an inferior level dump does not exist (except for a level 0
1439 dump).
1440
1441 8. Debugged the glob calls in interactive mode: restore used a
1442 dirent struct which was different from the /usr/include/dirent.h
1443 one (this used to work, is it a glibc2 change?), so none of the
1444 compat glob (which used /usr/include/dirent.h) or the system glob
1445 worked. Restore use now the system dirent (and the system
1446 DT_* constants), which are compatible with BSD ones.
1447
1448 9. Added a configure flag (--with-dumpdatespath) to specify
1449 the location of dumpdates file. By default, it is
1450 /etc/dumpdates.
1451
1452 10. Added the "AUTHOR" and "AVAILABILITY" sections and
1453 included the current date/version in man pages.
1454
1455 11. Corrected the estimation of remaining time when
1456 the operator doesn't change the tapes quickly enough. This
1457 was an old bug, I thought I corrected it, and discovered
1458 that in fact it was corrected in two different places, so
1459 the results canceled each other...
1460
1461 Changes between versions 0.4b5 and 0.4b6 (released October 1, 1999)
1462 ===================================================================
1463
1464 1. Integrated multiple patches from RedHat, Debian and SuSE:
1465
1466 - tweak dump/itime.c to not try to read dumpdates if the 'u' option
1467 isn't specified.
1468 - several fixes in the man pages.
1469 - update the default tape device to /dev/st0.
1470 - many updates for Linux Alpha (byte ordering, size_t etc).
1471 - buffer overruns.
1472 - use environment variable for TMPDIR (instead of /tmp).
1473 - use sigjmp_buf instead of jmp_buf (RedHat bug #3260).
1474 - workaround egcs bug (RedHat bugs #4281 and #2989).
1475 - wire $(OPT) throughout Makefile's.
1476
1477 2. Upgrade the dump revision to 1, making possible to dump filesystems
1478 made with e2fsprogs-1.15 or newer. Nothing seems to break...
1479
1480 3. Fix some compile warnings, prototype all functions.
1481
1482 4. Use glibc err/glob instead of internal compatibility
1483 routines (only if available).
1484
1485 5. Fix a compile error on Linux 2.2.7 / libc5 (5.4.44) (patch provided
1486 by Bernhard Sadlowski <sadlowsk@mathematik.uni-bielefeld.de>).
1487
1488 Changes between versions 0.4b4 and 0.4b5 (released September 22, 1999)
1489 ======================================================================
1490
1491 1. Integrated the changes from FreeBSD-3.1-RELEASE
1492 (mostly bug fixes, buffer overruns, dump has now an "automatic
1493 tape length calculation" flag, dump/restore can use kerberos now
1494 (this is NOT tested), use environment variables for TAPE and
1495 RMT etc.).
1496
1497 2. Integrated three RedHat patches ("glibc", "kernel" and "bread" patches)
1498
1499 3. Corrected a bug in restore when using 'C' option with multi-volumes
1500 tapes (files splited accros two tapes give "size changed" errors
1501 when comparing).
1502
1503 4. Corrected the long standing bug when dumping multiple tapes.
1504 This works for me, needs further testing.
1505
1506 Changes between versions 0.4b3 and 0.4b4 (released January 17, 1997)
1507 ====================================================================
1508
1509 1. Dump now runs correctly on kernels 2.1.x
1510 Fix made by Gerald Peters <gapeters@worldnet.att.net>
1511
1512 Changes between versions 0.4b2 and 0.4b3
1513 ========================================
1514
1515 1. Use realpath() if available
1516
1517 2. Report statistics
1518
1519 Changes between versions 0.4b1 and 0.4b2
1520 ========================================
1521
1522 1. Fixed the bug fix from Greg Lutz (I had made a mistake when integrating
1523 the patch)
1524
1525 2. Fixed restore to make it able to read FreeBSD 2.x dumps again
1526
1527 3. Fixed configure.in to correctly handle --enable-rmt
1528
1529 Changes between versions 0.3 and 0.4b1
1530 ======================================
1531
1532 1. Integrated the changes from 4.4BSD-Lite2
1533
1534 2. Integrated the patches from Debian and Red Hat
1535
1536 3. Portability changes: use the __u32, __u16, __s32, and __s16 types
1537
1538 4. Changed dump to use the Ext2fs library to get block addresses. This
1539 should solve the endianness problem on SparcLinux.
1540
1541 5. Created a configure.in file (shamelessly stolen from the e2fsprogs
1542 distribution's one) to use autoconf
1543
1544 6. Fixed a few minor bugs
1545
1546 Changes between versions 0.2e and 0.2f
1547 ======================================
1548
1549 1. Added the creation of named pipes (fifos) in restore.
1550
1551 2. Added the -N flag in the restore manual page.
1552
1553 3. Added the file kernel.patch which contains the llseek() optimization
1554 patch for 1.2.x kernels.
1555
1556 4. Fixed a bug in the restoration of symbolic links: owner and group were
1557 not restored.
1558
1559 5. Integrated some changes from FreeBSD 2.2.
1560
1561 6. Added a call to ftruncate() after restoring each file to restore
1562 correctly files ending by a hole.
1563
1564 Changes between versions 0.2d and 0.2e
1565 ======================================
1566
1567 1. Fixed a bug in the "set owner/mode" process. Every file was restored
1568 with owner = root (0) and group = root/wheel/whatever (0).
1569
1570 Changes between versions 0.2c and 0.2d
1571 ======================================
1572
1573 1. Dump is now able to backup 2GB+ filesystems.
1574
1575 2. Dump and restore can now be linked as static binaries.
1576
1577 Changes between versions 0.2b and 0.2c
1578 ======================================
1579
1580 1. Fixed a bug when dumping ``slow'' (i.e. normal) symbolic links.
1581
1582 Changes between versions 0.2a and 0.2b
1583 ======================================
1584
1585 1. Really fixed the bug that I should have corrected in 0.2a.
1586
1587 2. Enabled optimization again.
1588
1589 Changes between versions 0.2 and 0.2a
1590 =====================================
1591
1592 1. Disabled the optimization during compilation.
1593
1594 Changes between versions 0.1 and 0.2
1595 ====================================
1596
1597 1. Fixed a bug in fstab.c which caused a null pointer to be stored in
1598 the fs_type field (actually, I modified the file fstab.c to make it
1599 use the mntent functions).
1600
1601 2. Dump and restore now use a 4.3 BSD compatible dump format. Backups
1602 made by dump should be readable by the BSD restore and backups made
1603 by the BSD dump should be readable by restore. Unfortunately, this
1604 means that the dump format has changed between version 0.1 and version
1605 0.2 :-(
1606
1607 3. Dump is now able to backup a subtree, it is no longer limited to whole
1608 filesystems like the BSD version.
1609
1610 4. Dump now uses ext2_llseek() so it is able to backup filesystems bigger
1611 than 2 GB.
1612
1613 Changes between versions 0.0 and 0.1
1614 ====================================
1615
1616 1. Now create links rdump and rrestore during the `make install' step.
1617
1618 2. Linux port specific bugs added to the manual pages
1619
1620 3. Incorrect estimation of the number of tapes blocks fixed when doing
1621 incremental backups.
1622
1623 4. Better ls-like format in restore in interactive mode.