1 $Id: CHANGES,v 1.307 2010/03/08 10:40:52 stelian Exp $
3 Changes between versions 0.4b42 and 0.4b43 (released ?????????????)
4 ===================================================================
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)
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).
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
21 4. Fix restoration of extended attributes of fifos and device nodes.
22 Thanks to Igor Zhbanov <izh1979@gmail.com> for the bug report.
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.
28 Changes between versions 0.4b41 and 0.4b42 (released June 18, 2009)
29 ===================================================================
31 1. Fix printout of the 'Connection to' message in dump/restore.
32 Thanks to Dale Wiles <dwiles@users.sourceforge.net> for the
33 bug report and its fix.
35 2. Fix dump -w/-W output for filesystems having the last backup
36 level equal to 0. Thanks to Pascal Bouchareine
37 <kalou@users.sourceforge.net> for reporting the bug.
39 3. Enable restore to handle restoring onto a different SELinux
40 policy from the dump. Thanks to Tony Nelson
41 <tony_nelson@users.sourceforge.net> for the patch.
42 (see http://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=189845
45 4. Fix EA set failures when restoring immutable files. Thanks
46 to Andrew Kroeger for the patch.
48 5. Fix "mode file too large" errors when restoring huge backups.
49 Thanks to Kenneth Porter <shiva@sewingwitch.com> for the bug
52 6. Add the ability to link against libtinfo library if present,
53 instead of libtermcap or libncurses libraries. Thanks to
54 Adam Tkac <atkac@redhat.com> for the patch.
56 7. When comparing a backup, do not consider to be an error the fact
57 that we encounter files created while dump was in progress.
58 Thanks to Kenneth Porter <shiva@sewingwitch.com> for the bug
61 8. Force '-a' to be the default value when running 'restore -C'.
62 Thanks to Kenneth Porter <shiva@sewingwitch.com> for the bug
65 9. Many improvements to the 'cron_dump_to_disk' example.
66 Thanks to Aaron S. Hawley <ashawley@users.sourceforge.net> for
69 10. Fix the QFA generation when extended attributes are backuped
70 along with the inodes. Prior to this fix, the entries for
71 some inodes may get corrupted (duplicated entries or incorrect
72 ones), making the QFA file unusable for repositionning in
73 restore. Many thanks to Kenneth Porter <shiva@sewingwitch.com>
74 for the bug report and the testing of the fix.
76 11. Fix EA restoration in debug mode (Sourceforge bug #1986216).
78 12. Stefan Auracher <star7@onlinehome.de> noticed that the Tower
79 of Hanoi backup strategy documented in the man page wasn't
80 actually based on the Tower of Hanoi algorithm. The man page
81 was updated accordingly.
83 13. Fix missing level information from dump output when using
84 the default level (RedHat bug #493635). Thanks to Adam Tkac
85 <atkac@redhat.com> for the patch.
87 14. Fix the dump man page to reflect the fact that the default
88 dump level is 0 not 9 (RedHat bug #356121). Thanks to Adam Tkac
89 <atkac@redhat.com> for the patch.
91 15. Fixed a few spelling errors in the man pages (RedHat bug #489853).
92 Thanks to Adam Tkac <atkac@redhat.com> for the patch.
94 16. Fixed restore man page and program usage which incorrectly stated
95 that -P and -A were both allowed in the same invocation (RedHat
96 bug #490627). Thanks to Adam Tkac <atkac@redhat.com> for the patch.
98 17. Use sys_clone under Linux to share I/O contexts between dump
99 processes, thus drastically increasing the performance of dump
100 under CFQ (which is the default I/O scheduler used in a number of
101 distributions). Thanks to Jeff Moyer <jmoyer@redhat.com> for the
104 18. Add (preliminary) ext4 support - thanks to libext2fs which does
105 all the job for us. Thanks to Gertjan van Wingerde
106 <gwingerde@gmail.com> for the patch.
108 Changes between versions 0.4b40 and 0.4b41 (released January 2, 2006)
109 =====================================================================
111 1. Fix restore of dumped Access Control Lists. The previous
112 code for EA works fine, but ACLs needed conversion from the
113 ext2/3 disk format to posix_acl format before restoring.
115 2. Fix some issues with restoration of EA on big endian
118 3. Fix restore when the symtab is over 2GB in size.
120 4. Made the directory hash indexing an optional feature,
121 accessible by the '-H' option of restore, and disable it
124 5. Fixed dump to not include extended attributes information
125 in the toc (archive) file which confused restore -t.
127 Changes between versions 0.4b39 and 0.4b40 (released May 2, 2005)
128 =================================================================
130 1. Changed restore to emit warnings (instead of emitting a fatal
131 error) if a file (or a directory) is unavailable for a
132 comparision (if the user doesn't have the necessary permissions
133 to access it for example). Thanks to Kenneth Porter
134 <shiva@sewingwitch.com> for the bug report.
136 2. Re-done the 'do not save directory entries to non-dumped inodes
137 (excluded from dump)' feature. The previous implementation
138 worked well for excluded directories but not for regular files.
139 Thanks to Kenneth Porter <shiva@sewingwitch.com> for the bug
142 3. Fixed a bug in dump where the tape size was miscalculated when
143 the user used -d/-s to specify the tape characteristics. Thanks
144 to Philip Goisman <goisman@physics.arizona.edu> for reporting
147 4. Fixed another bug introduced in restore with the hashtree
148 implementation. This one caused restore to stop saying
149 "removenode: non-empty directory" in some cases.
151 5. Added support for dumping and restoring ext2/3 extended
152 attributes (EA), like the access control lists (ACL) or
153 the security labels used by SELinux.
155 Changes between versions 0.4b38 and 0.4b39 (released January 21, 2005)
156 ======================================================================
158 1. The newly added dump_on_cd_3 example was buggy, replace it
159 with an updated version from Andrew Basterfield
160 <bob@cemetery.homeunix.org>.
162 2. Made restore to chdir() back into the initial directory when
163 dumping core while aborting a comparision operation. The
164 previous behaviour was to write the corefile at the root of
165 the directory being compared, which could very well be
166 read only and preventing the corefile generation. Thanks
167 to Kenneth Porter <shiva@sewingwitch.com> for the bug report.
169 3. Silenced the failure to call fgetflags() when comparing an
170 entry which has no ext2 attributes (as in lsattr()).
172 4. Fix a brown paper bug in restore -C which broke restore and
173 caused modifications on the filesystem being compared
174 (directories containing a file with the same name as the
175 directory get renamed to RSTTMP...). Thanks to Kenneth Porter
176 <shiva@sewingwitch.com> for finding the bug and helping me
179 5. Made restore -C force the -N flag (no writing allowed on
180 the disk) in order to prevent more bugs like the above one.
182 Changes between versions 0.4b37 and 0.4b38 (released January 7, 2005)
183 =====================================================================
185 1. Fix a couple of troff syntax bugs in the man pages.
186 Thanks to Eric Raymond <esr@thyrsus.com> for the patch.
188 2. Made restore use either libncurses or libtermcap, depending
189 on which one is available at configure time.
191 3. Fixed restore negative size display bug when comparing a
192 dump containing files over 2GB. Thanks to Steve Bonds
193 <sbonds@users.sourceforge.net> for the bug report.
195 4. Do not save directory entries to non-dumped inodes
196 (excluded from dump). This will eliminate the 'missing
197 file' warnings when doing 'restore -C'.
199 5. Fix dump crash when backuping a huge (2TB) filesystem,
200 due to a bogus calculation on the inode map size.
201 Thanks to Kevin B. Haines <K.B.Haines@rl.ac.uk> for
202 submitting the bug and testing the fix.
204 6. Fix a problem in restore where the final \0 in the symbolic
205 link names could have been lost, generating corrupt filenames.
206 Thanks to Kyle Wilson <kyle.wilson@amd.com> for reporting the
209 7. Implemented a hash list for the directory names in restore.
210 The linear list used before caused problems in interactive
211 restores when dealing with directories having thousands of
212 entries. Thanks to Brian Ristuccia <bristuccia@starentnetworks.com>
213 for reporting the bug.
215 8. Improved restore -C, this time including the directory
216 attributes into the comparision.
218 9. Made restore understand tapes containing EA/ACLs (which will
219 be dumped by the next version of dump). In this version
220 extended attributes on the tape are ignored, for full EA/ACL
221 support wait for the next version or try the experimental EA
224 Changes between versions 0.4b36 and 0.4b37 (released July 7, 2004)
225 ==================================================================
227 1. Added the --enable-staticz configure option which enables
228 dump and restore to be linked with static versions of
229 libz and libbz2 (and dynamic versions of all other libraries).
230 This will make Debian users happy, because libz and libbz2
231 were the only needed libraries living in /usr, all the
232 others live in /lib. In case of system emergency, it is
233 better not to have to rely on an extra filesystem.
234 Thanks to Bdale Garbee <bdale@gag.com> for the suggestion.
236 2. Fix compilation on (at least the Linux Debian port to) AMD64.
237 (<ext2fs/ext2_types.h> defines some types (__s64 and __u64)
238 that are also defined by <linux/types.h> (<asm/types.h>) and
241 3. Make dump's reading of the dumpdates file a bit more robust,
242 preventing dump from crashing when the dumpdates file has
243 been modified by hand.
245 4. Fixed some offset calculations in dump code which could
246 lead to "bread lseek errors" on large filesystems. Thanks
247 to Bruce Lowekamp <lowekamp@users.sourceforge.net> for
248 reporting this bug and debugging the issue.
250 5. Made dump use the blkid library when searching for devices
251 by label or uuid instead of dump's own routines.
253 6. Corrected a bug in dump where a wrong LABEL=... line in
254 /etc/fstab could prevent dump from dumping unrelated
255 filesystems. Thanks to Bruce Lowekamp
256 <lowekamp@users.sourceforge.net> for reporting the bug.
258 Changes between versions 0.4b35 and 0.4b36 (released April 21, 2004)
259 ====================================================================
261 1. Fixed dump compilation with old gcc versions. Thanks to
262 Mike Castle <dalgoda@users.sourceforge.net> for the patch.
264 2. Fixed some warnings (howmany, roundup, powerof2 redefined)
265 when compiling against a recent glibc version.
267 3. Fixed a bug in restore preventing the read of a dump tape
268 written with Solaris 7 ufsdump. Thanks to Patrick Higgins
269 <phiggins@transzap.com> for reporting the bug and providing
272 4. Changed dump to enable the creation of volumes of different
273 sizes in a single run (make -B accept a list of values).
274 Patch contributed by Florian Zumbiehl <florz@gmx.de>.
276 5. Use the glibc provided minor() and major() macros instead
277 of our own bitmask implementation. This should be safe for
278 when the major/minor namespace will migrate to 32 bits.
279 Thanks to Zhang Jun <zhangjun@nanjing-fnst.com> for reporting
282 6. Made explicit in the dump man page that dump will not create
283 a remote file, it will only write to an already existing one.
285 7. Another try at making size estimates better again.
287 8. Put back the inconditional running of the end-of-tape
288 script which was changed in 0.434 to be run only when -M
289 or multiple -f were NOT specified. Some users rely on this
290 feature even when it is combined with -M/-f.
292 9. Fixed restore when restoring huge backups (where rstdir...
293 temporary files are over 2GB). Thanks to
294 Raphael Jaffey <rjaffey@artic.edu> for reporting this,
295 debugging the issue and making the patch.
297 10. Made restore understand FreeBSD UFS2 tapes. Thanks to
298 David <vrtin@users.sourceforge.net> for submitting the bug
299 and providing a test case.
301 11. Made dump work with an arbitrary integer level (no more 10
302 levels only limitation). Thanks to Matthew
303 <msvincen@midway.uchicago.edu> for the patch.
305 Changes between versions 0.4b34 and 0.4b35 (released December 21, 2003)
306 =======================================================================
308 1. Added a note in the dump man page saying that the default
309 blocksize can be 32 if -d is used with a high density argument.
310 Thanks to Antonios Christofides <A.Christofides@itia.ntua.gr>
313 2. Fixed configure to correctly understand CPPFLAGS, CFLAGS,
314 LDFLAGS environment variables. Thanks to Arcady Genkin
315 <antipode@users.sourceforge.net> for reporting the bug.
317 3. Made -e/-E options of dump accept an unlimited number of inodes
318 to be excluded, rather than a hardcoded maximum. Thanks to
319 Dietrich Rothe <d-rothe@users.sourceforge.net> for the patch.
321 4. Updated the autoconf system to 2.50. Forced the -D_BSD_SOURCE
322 and -D_USE_BSD_SIGNAL defines in configure in order to solve
323 64bit build problems because quad_t is redifined with a
324 different signature. Thanks to Mike Harris <mharris@redhat.com>
325 for reporting this bug.
327 5. Made restore build on Solaris, making possible to
328 restore Linux's "enhanced" tapes. Thanks to Uwe Gohlke
329 <uwe@ugsoft.de> for the patch.
331 6. Made an extension in the dump tape format capable of saving
332 MacOSX specific inode extensions. Uwe Gohlke <uwe@ugsoft.de>
333 wrote the extension and contributed the restore code back
334 into this codebase. The same extension mechanism will be
335 used in the future to save ACLs...
337 7. Made rmt work correctly with regard to QFA and local/remote
338 files and tapes. The remote access will however work only
339 when the dump provided rmt version is used. If you want to
340 use another rmt server, please do not use the QFA feature.
341 Thanks to Uwe Gohlke <uwe@ugsoft.de> for the patch.
343 Changes between versions 0.4b33 and 0.4b34 (released April 18, 2003)
344 ====================================================================
346 1. Fixed the INSTALL file to reflect the actual install paths.
347 Thanks to David Ronis <ronis@ronispc.chem.mcgill.ca> for
350 2. Fixed the configure script to only check for headers presence
351 instead of trying to compile them. This should fix issues
352 with old build environments. Thanks to Kari Lempiainen
353 <kari@funky.pp.fi> for reporting the bug.
355 3. Fixed restore to correctly ignore sockets when comparing
356 a dump (as socket cannot be properly restored anyway). Thanks
357 to Gunther Reiszig <gunther@mit.edu> for reporting the bug.
359 4. Fixed restore to correctly access the archive file (-A argument)
360 even when using a remote tape. Thanks to Fabrice Bellet
361 <fabrice@bellet.info> for reporting the bug.
363 5. Fixed (again) handling of long (largefile) seeks in rmt.
364 Thanks to Fabrice Bellet <fabrice@bellet.info> for reporting
367 6. Fixed restore corner case when dealing with large block sizes
368 dump is able to write now (-b 1024). Thanks to Fabrice Bellet
369 <fabrice@bellet.info> for reporting the bug.
371 7. Fixed a bug preventing dump to access a filesystem having
372 a label exactly 16 bytes in length. Thanks to <minduni@ti-edu.ch>
373 for reporting the bug.
375 8. Made dump store dump dates using explicit timezones, fixing a
376 problem with incremental dumps if the timezone is modified
377 between the dumps. Thanks to Philippe Troin <phil@fifi.org> for
378 the bug report and the patch.
380 9. Fixed a bug encountered when dumping individual files (not full
381 filesystems or directories) and dangling symbolic links happen
382 to be in the list of files. For as far as dump is concerned,
383 dangling symbolic links are allowed, and are dumped as is.
384 Thanks to Jin-su Ahn <jsahn@ee.snu.ac.kr> for reporting the
385 bug and providing the fix.
387 10. Fixed open and creation modes and permissions for QFA and
388 table-of-contents files in dump and restore. Thanks to
389 Philippe Troin <phil@fifi.org> for the patch.
391 11. Fixed the archive file descriptor handling enabling it to be 0.
392 This can happen in some cases when shell redirections are used.
393 Thanks to Philippe Troin <phil@fifi.org> for the patch.
395 12. Delayed the opening of archive file until after suid had been
396 dropped (fixing a possible security issue if dump is suid).
397 Thanks to Philippe Troin <phil@fifi.org> for the patch.
399 13. Fixed the 'S' command handling in the rmt client part.
400 Thanks to Philippe Troin <phil@fifi.org> for the patch.
402 14. Modified the end-of-tape script handling to print out statistics
403 (and stop the timer) before launching the eot script. Also, the eot
404 script does not get run anymore when using -M (which makes sense) or
405 when multiple tapes are listed on the command line
406 (-f tape0,tape1,tapen) (which also makes sense).
407 Thanks to Philippe Troin <phil@fifi.org> for the patch.
409 15. Relicensed dump/restore under the 'revised' BSD license, as per
410 ftp://ftp.cs.berkeley.edu/ucb/4bsd/README.Impt.License.Change.
412 16. Added LZO compression to dump. This new compression method has
413 the advantage of being super fast, thus not killing tape streaming
414 on slow machines. Thanks to Helmut Jarausch
415 <jarausch@igpm.rwth-aachen.de> for the patch and to
416 Markus Oberhumer <markus@oberhumer.com> for giving special permission
417 to include his miniLZO project (GPL licensed) in dump/restore.
419 17. Some small buffer overruns fixes in rmt. Thanks to Antonomasia
420 <ant@notatla.demon.co.uk> for reporting the bugs.
422 18. Added a special rmt version which can do encryption when writing
423 to tape. Read examples/encrypted_rmt/README for details on
424 how to enable and configure it. Thanks to Ken Lalonde
425 <ken@globalremit.com> for the patch.
427 19. Made dump work with 2.5 kernel end of tape early warning semantics.
428 Thanks to Kai Makisara <Kai.Makisara@kolumbus.fi> for the patch.
430 20. Fixed a bug which caused dump -w|-W not to work anymore, because
431 the fs_freq and fs_passno fields in /etc/mtab are always set
432 to 0 0. Thanks to Trent Piepho <xyzzy@speakeasy.org> for
435 Changes between versions 0.4b32 and 0.4b33 (released February 10, 2003)
436 =======================================================================
438 1. Added a note in the restore man page clarifying the question
439 'set the permissions on the current directory ?' asked by
440 restore at the end of treatment in -i and -x modes.
442 2. Fixed the endianess issues when reading compressed tapes.
443 Thanks to Dark Force <daq4th@users.sourceforge.net> for
444 reporting this bug and providing test cases.
446 3. Fixed the "ACL won't be dumped" warning message (which showed
447 an extra, unrelated error message). Thanks to Dragan Krnic
448 <dkrnic@lycos.com> for reporting this bug.
450 4. Made dump look first into /etc/mtab, then into /etc/fstab
451 when searching for filesystem paths. Also fixed some problems
452 caused by binding mounts or multiple block device mounts.
453 Thanks to Matus Uhlar <uhlar@fantomas.sk>, Elliott Mitchell
454 <ehem@m5p.com>, Greg Edwards <gedwards@users.sourceforge.net>,
455 Brian Hoy <brian.hoy@opus.co.nz>. (fixes Debian bugs #147086
456 and #160305, Sourceforge bugs #618699 and #679832).
458 5. Made dump's -I option accept the value '0' meaning all the
459 read errors will be ignored. This can be useful when running
460 dump from unattended sessions (like cron jobs). Thanks to
461 John I Wang <jiwang@users.sourceforge.net> for the suggestion.
463 6. Fixed the output of dump to indicate 'blocks' instead of
464 'tape blocks' in the various messages (blocks are always
465 1 Kilobyte, tape blocks are 1 BK * '-b' argument), and
466 made it clearly print the current blocksize at the start of
467 a dump. Thanks to Michal Szymanski <msz@astrouw.edu.pl> for
470 7. Made rmt understand long (largefiles) seeks.
472 8. Fixed build with very old versions of libext2fs, where
473 EXT2_FT_* constants were undefined.
475 9. Made dump accept the dumpdates path on the command line
476 (-D file option) instead of using only the hardcoded one.
477 Thanks to Piete Brooks <pb22@users.sourceforge.net> for the
480 10. Enabled rmt, LFS, readline, QFA options by default in
481 ./configure. Updated the configure process (new versions
482 of config.guess, config.sub etc).
484 Changes between versions 0.4b31 and 0.4b32 (released November 15, 2002)
485 =======================================================================
487 1. Changed dump to use fcntl(F_SETLK) style locking instead
488 of flock() when locking the dumpdates file. With the old
489 locking scheme, a local user having read rights on the
490 dumpdates file could be able to do a Denial of Service attack
491 on dump. In order to lock the dumpdates file with the new
492 scheme, the user would need to have write access on the file.
493 Thanks to Richard Johnson <Richard.Johnson3@ey.com> for
494 reporting the bug (originally a bugtraq post).
496 2. Fixed interactive 'ls' which caused spurious errors warnings
497 about 'undefined filetypes' detected. Thanks to Jorgen Ostling
498 <jorgen_ostling@users.sourceforge.net> for reporting this
501 3. Fixed dump's estimate when dealing with sparse inodes.
503 4. Modified dump to allow setting a blocksize bigger than 32kB
504 (raised the limit to 1024kB), because newer hardware needs
505 this for proper operation. Thanks to Dirk Traenapp
506 <dtraenapp@users.sourceforge.net> for reporting this.
508 5. Fixed a bug causing Dump to stop and report an error if an
509 inode number in the exclude file was followed by some amount
510 of whitespace. Thanks to Jeffrey Sofferin
511 <sofferin@users.sourceforge.net> for reporting this bug.
513 6. Fixed a bug which caused restore, in some particular cases,
514 to ask some 'scary' questions and leave a bunch of RSTTMP
515 directories behind when restoring incremental tapes. Thanks
516 to Philippe Troin <phil@fifi.org> for reporting this bug and
517 providing the test cases.
519 7. Changed the wording when inodes are excluded from dump:
520 replaced 'Added inode 7 to exclude list' with
521 'Excluding inode 7 (resize inode) from dump', as suggested
522 by Elliott Mitchell <ehem@m5p.com> in a Debian bug report.
524 Changes between versions 0.4b30 and 0.4b31 (released July 30, 2002)
525 ===================================================================
527 1. Fixed rmt open flags transmission (GNU's symbolic syntax over
528 rmt) which I broke in 0.4b29. Thanks to Eros Albertazzi
529 <eros@lamel.bo.cnr.it> for reporting the bug.
531 Changes between versions 0.4b29 and 0.4b30 (released July 25, 2002)
532 ===================================================================
534 1. Made dump print out the ext2 logical block number in case of a read
535 error, which can be given as an argument to debugfs' ncheck command,
536 in order to find out the affected inode. Added note about this
537 usage in dump's man page.
539 2. Fixed a problem in restore when reading tapes written on big
540 endian machines with very old versions of dump. The patch was
541 contributed by George Helffrich <george@geology.bristol.ac.uk>.
543 3. Fixed the tape length calculation when using large tapes
544 and compression. Thanks to Georg Lippold
545 <g_lippold@sourceforge.net> for reporting the bug.
547 4. Added a new set of examples in dump_on_cd_2 directory, based
548 on dump_on_cd examples but somewhat enhanced, supporting DVD
549 media, and localized in english and german. Thanks to
550 Georg Lippold <g_lippold@sourceforge.net> for the new scripts.
552 5. Made dump save 32 bit UID/GID. Internally, this was achieved
553 by switching from the old BSD inode format to the new BSD
554 inode format, which means that the tape format was changed.
555 However, since all restore versions out there should
556 transparently support both inode formats, the change should
557 have no side effects. Thanks to John Yu <jky@cs.bu.edu> for
560 6. Fixed a lot of warnings in the code shown when compiling
561 with 'gcc -W'. Thanks to Matthias Andree
562 <matthias.andree@stud.uni-dortmund.de> for reporting this.
564 7. Fixed a small markup bug in the dump man page. Thanks to
565 Eric S. Raymond <esr@minx.thyrsus.com> for submitting the
568 8. Rewrote entirely the man pages using the tmac.an macro
569 package (Linux man page format) instead of the original BSD
570 format. They should be now cleaner and easier to modify.
572 Changes between versions 0.4b28 and 0.4b29 (released June 8, 2002)
573 ==================================================================
575 1. Fixed a problem in the rmt ioctl command, where ioctl's issued from
576 non Linux clients were misinterpreted. The description of the problem
577 (incompatible numbering in Linux mtio opcodes) is documented at
578 ftp://ftp.fokus.gmd.de/pub/unix/star/README.mtio . Thanks to
579 Jörg Schilling <schilling@fokus.gmd.de> for reporting this bug and
580 providing an excellent, cross-platform replacement for rmt in his
583 2. Fixed a bug reported by Andy Mentges <amentges@jumpline.com>
584 which caused restore to fail when the inode maps were not entirely
585 contained on the first volume (when using really small volumes or
586 when dumping a huge number of inodes).
588 3. Fixed a problem in dump, where files in subdirectories of directories
589 that have the nodump flag set (or being excluded with -e/-E) were
590 sometimes incorrectly dumped (instead of being excluded). The original
591 fix is from David C Lawrence <tale@dd.org> in the FreeBSD version
592 of dump: http://www.FreeBSD.org/cgi/query-pr.cgi?pr=32414
593 Thanks to Ted Grzesik <tedgyz@roostme.com> for reporting the bug and
594 help testing the patch.
596 4. Added some example scripts from Gerd Bavendiek <bav@epost.de>
597 which makes one able to pipe the output of dump, by the net, to
598 a remote CD-burner server.
600 5. Made dump use O_CREAT|O_TRUNC both locally and remotely (over rmt),
601 and use GNU's symbolic syntax over rmt instead of numerical values
602 to assure multiple platform compatibility.
604 6. Documented the -d option in restore.
606 7. Added a -v (verbose) mode to dump. For now it just prints the number
607 of the inode being dumped, but this could evolve in future versions
608 to include interesting debugging output.
610 8. Added a -o flag to restore, which automatically restores the current
611 directory permissions in -i or -x mode, without asking the operator
612 whether to do so. Patch submitted by Tony Clayton <tonyc@e-smith.com>
613 and Peter Samuel <peters@e-smith.com>.
615 Changes between versions 0.4b27 and 0.4b28 (released April 12, 2002)
616 ====================================================================
618 1. Fixed a bug in the QFA feature which could have caused
619 a bad tape positionning on restore (causing extra delays in
620 restore). Credits go to Uwe Gohlke <uwe@ugsoft.de>.
622 2. Added a small note in the dump man page specifying that
623 there should be no white space between the option letter and
624 the -j or -z parameter, thanks to Kenneth Porter <shiva@well.com>
626 3. Made restore work with older versions of the readline library,
627 by checking in configure for several extended symbols. Restore
628 can now be compiled with a readline as old as the 2.0 release
629 (though it may be a good idea to upgrade it to a more recent
630 version...). Thanks to Andrew Donkin <ard@waikato.ac.nz> for
631 reporting the build failures.
633 4. Fixed a performance problem with the QFA file creation in
634 dump, which made unnecessary seeks on the tape slowing down
635 the dump. Thanks to Andrew Donkin <ard@waikato.ac.nz> for
636 reporting this issue.
638 5. Removed the inclusion of some kernel headers in the dump
639 source, which prevented the compile in some kernel/glibc
640 headers/architecture combination. Thanks to Bdale Garbee
641 <bdale@gag.com> for reporting the bug.
643 6. Added the appropriate error message when dump fails to
644 open the output file for writing. Thanks to Amith Varghese
645 <amithv@yahoo.com> for reporting this bug.
647 7. Made restore able to understand large Solaris ufsdump tapes
648 (containing inodes bigger than 4194304). Sun have introduced
649 an "extension" to the dump tape format when dealing with
650 those inodes, which was uncorrectly handled by Linux restore.
651 Thanks to Uwe Gohlke <uwe@ugsoft.de> for reporting the bug and
652 providing a test case.
654 8. Added the -m parameter to dump which optimises the output for
655 inodes having been changed but not modified since the last dump
656 ('changed' and 'modified' have the meaning defined in stat(2)).
657 For those inodes, dump will save only the metadata, instead of
658 saving the entire inode contents. Inodes which are either
659 directories or have been modified since the last dump are saved
660 in a regular way. Uses of this flag must be consistent, meaning
661 that either every dump in an incremental dump set have the flag,
664 Changes between versions 0.4b26 and 0.4b27 (released February 15, 2002)
665 =======================================================================
667 1. Fixed behaviour of dump when exceeding resource limits
670 2. Added the -L flag to restore to allow the user to specify a
671 maximal allowed number of miscompares when using restore
672 with the -C option to check the backup.
674 3. Detailed the manual entry for the -N option of restore.
676 4. Added the -a flag to restore to make able doing unattended
677 restores in -i or -x mode (automatically walks through the
678 multiple dump volumes).
680 5. Extended the QFA mode to work with local files and/or
681 remote tapes and files. This way, restore can know in advance
682 the tape number and the offset for the inodes to extract and
683 can minimize the extraction time by seeking directly to the
686 6. Added the -A <archive> option to both dump and restore,
687 which makes dump to archive a dump table-of-contents in
688 the specified file to be used by restore to determine
689 whether a file is in the dump file that is being restored.
690 (the archive file syntax is also compatible with the
691 Solaris ufsdump generated one).
693 7. Small fix in restore making it able to read some (broken ?)
694 Solaris ufsdump tapes.
696 8. Fixed dump to correctly recognise the root filesystem when using
697 ext2 disk labels (LABEL=/). Thanks to John Yu <jky@it.bu.edu>
698 for reporting this bug.
700 9. Added the -P <file> option to restore to create a
701 Quick File Access file from an already made dump. Patch
702 contributed by Uwe Gohlke <uwe@ugsoft.de>.
704 10. Made restore compile and run on Solaris, making it a
705 possible replacement for the standard ufsrestore. Port was
706 contributed by Uwe Gohlke <uwe@ugsoft.de>.
708 Changes between versions 0.4b25 and 0.4b26 (released January 7, 2002)
709 =====================================================================
711 1. Added a set of backup scripts from Eugenio Diaz
712 <getnito@yahoo.com> in the examples section. It features
713 automatic (cron based) full and incremental dumping of
714 several filesystems on a separate filesystem.
716 2. Fixed a off-by-one miscalculation which disabled dumping
717 a one letter subdirectory. Thanks to Chris Siebenmann
718 <cks@utcc.utoronto.ca> for reporting the bug.
720 3. Fixed several restore bugs occuring when trying to
721 restore some missing files on the tape. Thanks to Chris
722 Siebenmann <cks@utcc.utoronto.ca> for reporting the bug.
724 4. Fixed --with-ldopts configure argument passing, installing from
725 a separate object directory, makefile cleanups contributed
726 by <splite@purdue.edu>.
728 5. Fix a bug which could caused, in some conditions, the highest
729 number inode of a filesystem, to not be dumped. Many thanks
730 to Chris Siebenmann <cks@utcc.utoronto.ca> for helping me
733 Changes between versions 0.4b24 and 0.4b25 (released November 17, 2001)
734 =======================================================================
736 1. Added a mini howto from Patrick Walsh in the examples
739 2. Minor man pages syntax corrections. Thanks to
740 Chris Danis <screechco@home.com> for reporting the bugs.
742 3. Added a script from David B. Peterson <dave@toppledwagon.com>
743 to the examples section. It features dumping several
744 filesystems to a remote tape drive upon ssh.
746 4. Added a patch provided by Richard Jones <rich@annexia.org>
747 which allows BRADEMAX (number of read errors tolerated by
748 dump) to be adjusted using the -I option.
750 5. Fixed a bug which disabled doing "restore -C -f -". Thanks
751 to Clemens Stiglechner <a9401816@unet.univie.ac.at> for the
754 6. Add the -l option to restore to specify if, when doing a
755 remote restore, the file used to access the backup is a
756 regular file (the defaults being a tape drive). Restore needs
757 to know this information when reading a remote compressed
758 dump. Previously, this information was autodetected, but
759 the autodetection code fails (with ioctl: Inappropriate ioctl
760 for device) when using a non Linux remote box. Thanks to
761 many users and especially to Eros Albertazzi
762 <eros@lamel.bo.cnr.it> for reporting this.
764 7. Found a workaroung for the dump deadlock problem (3 childs
765 stuck in pause(), father in read()). The workaround seems
766 to work for me and several beta-testers. If it doesn't work
767 for you, please report back.
769 8. Updated the RPM spec file (BuildPrereq, URL etc).
771 Changes between versions 0.4b23 and 0.4b24 (released September 12, 2001)
772 ========================================================================
774 1. Fixed the permissions of a newly created QFA file by dump.
776 2. Cleaned up the source of dump (the external variables
777 definition was a complete mess, making possible to have
780 3. Fixed restore to use the full tape volume path when doing
781 a compare (since it changes the working directory to the
782 filesystem being compared in the process).
784 4. Added the -q option to dump which makes dump abort
785 whenever it needs operator attention. It should be
786 easier to use dump in scripts with this option.
788 5. Detect the use of incompatible options to dump and
789 refuse them (like -a and -B options together).
791 6. Added bzip2 compression to dump/restore (use option -j level
792 to select it). Note that you will need a restore version
793 >= 0.4b24 in order to restore a bzip2 compressed dump.
794 The same warning as for the zlib compression applies:
795 the tape format of a bzip2 dump is not compatible with the
796 original BSD tape format.
798 7. Fixed a overflow problem in dump corrupting the dump when
799 very large files were encountered. Thanks to Vladimir Nadvornik
802 8. Added a ioctl(BLKFLSBUF) in dump which should flush the
803 kernel buffer/page cache before starting the dump, helping
804 a bit those who use dump on mounted filesystems. Thanks to
805 John Yu <jky@it.bu.edu> and to Theodore T'so <tytso@mit.edu>
808 9. Updated the RPM spec file following the RedHat changes
809 (dynamically linked binaries now in /usr/sbin etc).
811 10. Added a patch from Helmut Jarausch <jarausch@igpm.rwth-aachen.de>
812 which enables restore to recognise multi volume compressed dumps
813 done on CD. Included his scripts for dump (which pipe the dump
814 output directly into cdrecord) and restore. There is now
815 possible to to backups to CD on the fly!
817 Changes between versions 0.4b22 and 0.4b23 (released July 20, 2001)
818 ===================================================================
820 1. Fixed a buffer overflow in restore/tape.c. Patch provided by
821 Marcus Meissner (Caldera International Security Dept.).
823 2. Implement the Sun rmt extended protocol. Patch submitted
824 by Ian Gordon <iangordon@users.sourceforge.net>.
826 3. Check for the e2fsprogs header <ext2fs/ext2_fs.h> instead of
827 the linux kernel header. This ensures that dump always has the
828 latest version of this file. Patch submitted by
829 Andreas Dilger <adilger@turbolinux.com>.
831 4. Report any filesystem present in either /etc/fstab with a
832 positive passno or /etc/dumpdates in dump -w output.
833 Patch submitted by Andreas Dilger <adilger@turbolinux.com>.
835 5. Fixed the looping problem in dump introduced in the
838 6. Changed the -B option of dump to limit the size of
839 _compressed_ output per volume if compression is on.
840 Patch contributed by Helmut Jarausch
841 <jarausch@igpm.rwth-aachen.de>. Note however that, since
842 it is impossible to predict the size of the compressed
843 data before doing the compression, the -B limit is a bit
846 7. Fixed a bug in reading the operator typed file/tape path for
847 the next volume in restore.
849 8. Implemented a "-F script" option to restore which permits the
850 user to specify a script which will be launched at the
851 beginning of each tape, useful for automatic programming of
852 tape changers for example. See the restore man page for the
853 script parameters and return codes.
855 9. Small fix for the QFA routines provided by Uwe Gohlke
856 <uwe@ugsoft.de>, and some recommendations for QFA uses in
859 10. Fixed the multivolume restoring where making a mistake
860 to the 'Mount next tape' prompt caused several blocks to
863 11. Enhanced the -e option of dump to take as a parameter a
864 comma separated list of inode numbers.
866 12. Added the -E option to dump which specify a file containing
867 inode numbers to exclude from the dump.
869 13. Fixed the compressed multi-volume dump + restore.
871 Changes between versions 0.4b21 and 0.4b22 (released May 12, 2001)
872 ==================================================================
874 1. Made dump report the number of blocks dumped per volume.
875 Thanks to Kenneth Porter <shiva@well.com> for the suggestion.
877 2. Fix a bug in dump where waiting too much at the 'change volumes'
878 question caused the volume to be rewritten. Thanks to
879 Graham Reed <greed@users.sourceforge.net> for reporting the
880 bug and providing a patch.
882 3. Added a compression option to dump, contributed by Jerry
883 Peters <gapeters@worldnet.att.net>.
885 WARNING: the compression works only on files/pipes and
886 drives supporting variable block size.
888 WARNING2: tapes created using the compression feature are
889 incompatible with the standard dump tape format, and a
890 version of restore >= 0.4b22 will be needed for extraction.
892 4. Fixed some compilation problems for glibc 2.2.2 and 64 bit
893 architectures. Thanks to Paul Slootman <paul@debian.org> for
894 the patch and to Bdale Garbee <bdale@gag.com> for forwarding
897 5. Many cleanups (CPP defines, const char warnings, check of
898 ext2fs COMPAT flags, time_t cleanups, added libext2 version
899 in dump usage text) by Andreas Dilger <adilger@turbolinux.com>.
901 6. Made --prefix option work in configure. All the install path
902 are now based on the configure parameters.
904 7. Added the Quick File Access mode in dump/restore, contributed
905 by Uwe Gohlke <uwe@ugsoft.de>. In this mode, dump stores in
906 a file tape position for each inode, and this file is used by
907 restore (if called with parameter Q and the filename)
908 to directly position the tape at the file restore is currently
909 working on. This saves hours when restoring single files from
910 large backups, saves the tapes and the drive's head. Use
911 --enable-qfa option of configure to compile in the QFA support.
913 8. Added the possibility to dump several files and directories
914 in a single invocation of dump. Thanks to Uwe Gohlke
915 <uwe@ugsoft.de> for implementing this option.
917 9. Fixed the dumping and restoring of files with holes
918 (files ending with holes didn't get dumped or restored
921 10. Fixed a socket descriptor leak in dump, which leaved opened
922 3 file descriptors per dump process (and there is one dump
925 11. Fixed dump large file system support, by explicit use of
926 open64/lseek64/etc functions (explicit use needed because
927 e2fsprogs libraries don't behave well when compiled with
928 FILE_OFFSET_BITS=64).
930 Changes between versions 0.4b20 and 0.4b21 (released January 13, 2001)
931 ======================================================================
933 1. Fixed some bugs in the dump -w|-W logic introduced by
934 the previous version. Thanks to Andreas Dilger
935 <adilger@turbolinux.com> for his help on this one.
937 2. Fixed again a compilation problem when using old e2fs
938 headers (filesystem label related). Thanks to many users
939 who reported this stupid error.
941 3. Fixed a build problem on old lib5 based systems dealing with
942 _PATH_MNTTAB being undefined. Thanks to John Adams
943 <johna@onevista.com> for reporting the bug.
945 4. Improved the error detection in restore when creating
946 the temporary files in TMPDIR. Now restore will corectly
947 report a 'no space left on device' error instead of
948 strange errors which could imply an archive corruption.
949 Thanks to Gabriel Somlo <somlo@cs.colostate.edu> and
950 bgjenero <bgjenero@sympatico.ca> for reporting the bug.
952 5. Added the throughput information in dump's progression
953 messages. Thanks to Andreas Dilger <adilger@turbolinux.com>
956 6. Use libext2fs's inode scanning routines, which are particularly
957 robust when dealing with errors (especially when having some
958 bad blocks on the filesystem). Many thanks to Colin
959 <colin@horizon.com> for reporting the bug, investigating
960 the issues, finding a workaround, writing the patch and
961 fully testing it... (of course, if this patch will break
962 anything, he is to blame for :-)).
964 7. Made dump and restore LFS aware. Dump can dump a filesystem
965 containing large files, generate a large file on output and
966 restore can restore them. This feature is not enabled by
967 default, use --enable-largefile option of configure to enable
968 it (you need to have a LFS aware glibc though). Thanks to
969 Andrea Arcangeli <andrea@suse.de> for submitting the patch,
970 and to Theodore T'so <tytso@valinux.com> for his always
973 8. Made dump ask upon a tape write error if it should rewrite
974 the current volume (assume this is a bad tape) or if it should
975 assume an end-of-tape condition (useful for tape drives which
976 behaves badly at the end of the tape). Thanks to Andreas
977 Dilger <adilger@turbolinux.com> for the suggestion.
979 Changes between versions 0.4b19 and 0.4b20 (released November 10, 2000)
980 =======================================================================
982 1. Fixed a small compilation problem due to a change
983 in the definintion of the struct sigaction in
984 glibc 2.0 / libc5. Thanks to Gunther Schlegel
985 <schlegel@riege.de> for reporting the bug and to
986 Dave Platt <dplatt@snulbug.mtview.ca.us> for suggesting
989 2. Modified the label and uuid analysis in order to be
990 self-contained (independant of kernel/libc headers). This
991 should fix the compile with older kernel/libc headers and
992 will preserve the functionality. Thanks to Bernhard Erdmann
993 <bernhard.erdmann@gmx.de> for reporting the bug.
995 3. The 'exclude inode' option, if applied to a directory
996 inode, excludes now correctly all the leaves of this
997 directory. Thanks to John R. Dennison
998 <gerdesas@users.sourceforge.net> for reporting the bug.
1000 4. Fixed the '-e' option to disable the possibility
1001 to exclude the root inode (which causes the dumps to
1002 be unreadable by restore). Prevented array overflow
1003 when multiple -e options are used.
1005 5. Fixed dump to correctly interpret a filesystem argument
1006 which is a mountpoint and it is not an absolute path
1007 (as specified in the fstab). Thanks to Bernhard R. Erdmann
1008 <be@berdmann.de> for reporting the bug.
1010 6. Made dump able to backup files larger than 2 GB. Note that
1011 dump still doesn't cope with files larger than 4 GB.
1013 7. Restore the real uid and gid before invoking an external
1014 RSH program (big hole when dump or restore is suid root!).
1016 8. Get the values for the default tape device and the location
1017 of fstab file from the system headers. Thanks to
1018 Andreas Dilger <adilger@turbolinux.com> for the patch.
1020 9. Made dump -w|-W report all recognized filesystems
1021 present in either /etc/fstab or /etc/dumpdates, and present
1022 the list in the same order as in fstab file. Thanks
1023 to Andreas Dilger <adilger@turbolinux.com> for the patch.
1025 10. Made dump's -a (automatic end of tape detection) the
1026 default. Specifying one of -B, -c, -d or -s options will
1027 override the EOM detection. Thanks to Andreas Dilger
1028 <adilger@turbolinux.com> for the patch.
1030 11. Save the ext2 filesystem label into the dump volume label.
1031 Specifying a volume label on the command line (-L option)
1032 will override this feature. Thanks to Andreas Dilger
1033 <adilger@turbolinux.com> for the patch.
1035 Changes between versions 0.4b18 and 0.4b19 (released August 20, 2000)
1036 =====================================================================
1038 1. Fixed the signal handling in dump (which I broke in 0.4b17)
1039 which was causing several strange dump failures (dump
1040 hanged or segmentation faults etc).
1042 2. Specified the default blocksize in the dump man page.
1044 3. Changed two info messages of restore to be written on stdout
1045 instead of stderr in order to leave stderr only for errors
1046 or warnings. Thanks to Stephen Atwell
1047 <satwell@urbana.css.mot.com> for the suggestion.
1049 4. Corrected an off by one calculation which prevented
1050 dumping a directory having a 1 character name length.
1051 Thanks to Bernhard Erdmann <bernhard.erdmann@gmx.de>
1052 for reporting the bug.
1054 5. Reinforce dump in order to allow only level 0 and no
1055 -u option when dumping a subdirectory, and document
1056 this in the man page. Thanks to Bernhard Erdmann
1057 <bernhard.erdmann@gmx.de> for reporting the bug.
1059 6. Fixed a small harmless bug in restore which caused
1060 in some conditions hard links to be restored several
1061 times (therefore generation some warning messages).
1062 Thanks to Kenneth Porter <shiva@well.com> for
1065 7. Updated the RPM spec file to the latest RedHat version,
1066 providing FHS packaging and other cosmetic changes.
1067 You will need RPM version >= 3.0.5 in order to build the RPM.
1069 8. Updated the configure script to check for libtermcap
1070 before checking for libreadline, because we need this
1071 library in order to compile the readline support.
1073 9. Made dump understand the LABEL= and UUID= notation
1074 both in /etc/fstab and on the command line. Note that
1075 you will need the /proc filesystem in order to use
1076 these notations. Thanks to Erik Troan <ewt@redhat.com>
1077 for providing the patch.
1079 Changes between versions 0.4b17 and 0.4b18 (released June 30, 2000)
1080 ===================================================================
1082 1. Fixed a potential buffer overflow in restore. Thanks
1083 to Stan Bubrouski <satan@fastdial.net> for reporting
1086 2. Fixed a readline-related bug which prevented
1087 'cat DUMPFILE | restore -i -f -' from working. Thanks
1088 to Charles Karney <karney@users.sourceforge.net>
1091 3. Changed a few "panic" into "exit", causing restore to
1092 be more stable against some attacks (like the last one
1093 reported on Bugtraq, although the last version of restore
1094 was not vulnerable - just dumped core). Thanks to
1095 Andreas Hasenack <andreas@conectiva.com.br> for reporting
1098 4. Removed the suid-root bit on dump and restore in the
1099 default build (and generated RPMs). It should be safer
1100 now. Users who need the suid-root capabilities in order
1101 to make network backups should read first the man page
1102 of dump and enable the suid bit manually.
1104 5. Added -ltermcap to the compile parameters for restore
1105 when using readline support, in order to make the compile
1106 process work on some systems (SuSE ?). Thanks to
1107 Patrik Schindler <poc@pocnet.net> for reporting the bug.
1109 Changes between versions 0.4b16 and 0.4b17 (released June 1st, 2000)
1110 ====================================================================
1112 1. The -F script is called now *only* at the end of a tape,
1113 not at the end of the dump. Thanks to Christian Haul
1114 <haul@informatik.tu-darmstadt.de> for the bug report.
1116 Normally, the device name and the current volume number
1117 are passed on the command line of the script. However,
1118 if you want the old style script (with no arguments
1119 passed on the command line), you can enable it in
1120 configure with the --enable-oldstylefscript.
1122 2. Use posix signal handling to preserve dump functionality
1123 with libc5. Thanks to Jeff Johnson <jbj@redhat.com> for
1126 3. Made the exit code of restore in case of a 'C'ompare
1127 command reflect the result of the compare. An exit status
1128 of 0 means the dump archive is correct, 1 means tape errors,
1129 2 means that some files were modified. Thanks to Kenneth Porter
1130 <shiva@well.com> for the suggestion.
1132 4. Made (finally) quotes work in interactive restore.
1134 5. Small fixes in order to allow dump to compile with a
1135 really old e2fsprogs version. Thanks to Ian Zimmerman
1136 <itz@speakeasy.org> for the bug report.
1138 6. Add GNU readline capabilities to interactive restore.
1139 Use configure's --enable-readline flag to enable this feature.
1140 Thanks to Patrik Schindler <poc@pocnet.net> for the
1143 7. Do the compare on the fly in restore 'C' mode (this will
1144 allow not to exhaust the available /tmp space when
1145 ccmparing large files). Thanks to Kenneth Porter
1146 <shiva@well.com> for the suggestion.
1148 Changes between versions 0.4b15 and 0.4b16 (released March 11, 2000)
1149 ====================================================================
1151 1. Fixed some several duplicate 'DUMP: DUMP:' in the
1154 2. Corrected the estimation of blocks to dump. Note that
1155 this is still wrong for large directory inodes, due
1156 to the size of a BSD directory entry on the tape
1157 (estimation is based on the size of the ext2 directory,
1158 which is smaller because it doesn't contain the null
1159 character at the end).
1161 3. Made dump report the total number of MB written to
1162 tape at the end of run. Thanks to W. Reilly Cooley
1163 <wcooley@nakedape.cc> for the patch.
1165 4. Added the -X option to restore giving the possibility
1166 to read the names of the files to be extracted/listed
1167 from a text file (in addition of the command line).
1168 Thanks to Dejan Muhamedagic <dejan@quant-x.com> for the
1171 5. Added the device name and the current volume number
1172 as arguments to the end of tape script (dump -F option).
1174 6. Made the multi-volume dump work again (after having
1175 broken it in 0.4b15).
1177 Changes between versions 0.4b14 and 0.4b15 (released March 2, 2000)
1178 ===================================================================
1180 1. Added a prompt command in interactive restore mode. Thanks
1181 to Andreas Dilger <adilger@home.com> for the patch.
1183 2. Fixed a buffer overflow problem in dump (caused by
1184 not checking the size of the filesystem parameter).
1185 Thanks to Kim Yong-jun <loveyou@hackerslab.org> for
1186 reporting this on Bugtraq (and to several dump users
1187 who forwarded me his mail).
1189 3. Added the '-F script' option to dump in order to
1190 launch a script at the end of each tape (to be used
1191 with a tape changer, or to notify the sysadmin by
1194 4. Fixed a bug in restore compare code caused by the changes
1197 5. Fixed the treatment of options using the old BSD syntax
1198 in both dump and restore.
1200 Changes between versions 0.4b13 and 0.4b14 (released February 10, 2000)
1201 =======================================================================
1203 1. Fixed a bug in dump which may have caused invalid deleted
1204 directories to be dumped out if they were deleted after the
1205 mapping pass. This could occure on active filesystem and lead
1206 to heap corruption (causing dump malfunction in many possible ways).
1207 Thanks to Jan Sanislo <oystr@cs.washington.edu> for finding this
1208 bug and submitting the patch.
1210 2. Fixed the handling of the filesystem parameter in dump. A
1211 '/mnt/disk' parameter caused the disk contents to be dumped,
1212 but a '/mnt/disk/' parameter caused the mountpoint directory
1213 to be dumped (generally an empty directory).
1215 3. Improved the output of dump in order to tell which directory
1216 it is currently dumping (when dumping a subtree).
1218 4. Added the '-e' exclude inode option to dump. Thanks to
1219 Isaac Chuang <ike@isl.stanford.edu> for contributing with the patch.
1221 5. Added a REPORTING-BUGS file in order to provide a guide
1222 on how to correctly report a bug in dump/restore.
1224 6. Fixed a restore problem when restoring a hard link to an inode
1225 having the immutable or append-only attribute set. Thanks to
1226 Ambrose Li <acli@mingpaoxpress.com> for submitting the patch.
1228 7. Fixed a compatibility problem between dumps made on little
1229 endian machines (the format was correct) and big endian
1230 machines (the format was incorrect). This fix break the
1231 compatibility with the older dumps made on big endian
1232 machines (sparc, m86k, ppc etc). For the first time in
1233 linux dump's history (I believe), the dumps made by *BSD,
1234 Linux/alpha, Linux/sparc and Linux/x86 are compatible,
1235 so interchangeable. Thanks to Rob Cermak
1236 <cermak@ahab.rutgers.edu> for submitting the bug and
1237 helping me test the fix.
1239 8. Fixed the way dump reports the remaining percent/time, if
1240 the number of blocks actually dumped exceeds the estimated
1241 number of blocks. Thanks to Jean-Paul van der Jagt
1242 <jeanpaul@dutepp0.et.tudelft.nl> for reporting the bug.
1244 Changes between versions 0.4b12 and 0.4b13 (released January 21, 2000)
1245 ======================================================================
1247 1. Small Y2K fix in the man pages :). Thanks to Bernhard Sadlowski
1248 <sadlowsk@Mathematik.Uni-Bielefeld.DE> for reporting the bug.
1250 2. Removed the requirement to build the RPM as root from the
1251 spec file. Thanks to Christian Weisgerber
1252 <naddy@mips.rhein-neckar.de> for submitting this.
1254 3. Fixed a bug in dump related to the 'filetype' feature of ext2fs,
1255 causing dump to block when dumping really huge filesystems.
1256 Many thanks to Patrik Schindler <poc@pocnet.net> for
1257 helping me find this bug.
1259 4. Fixed the treatment for an interrupt signal when dump access
1260 the remote tape through RSH. Thanks to Christian Weisgerber
1261 <naddy@mips.rhein-neckar.de> for providing the patch.
1263 5. Fixed a bug which was causing dump/restore to display
1264 garbage characters instead of the remote host name.
1266 Changes between versions 0.4b11 and 0.4b12 (released January 8, 2000)
1267 =====================================================================
1269 1. Small fix in the dump man page. Thanks to Thorsten Kukuk
1270 <kukuk@suse.de> for submitting the patch.
1272 2. Fix for the exit code when using the size estimate option of
1273 dump. Thanks to Matti Taskinen <mkt@rni.helsinki.fi> for
1274 submitting the patch.
1276 3. Handle EINTR in atomical reads/writes in dump, which was causing
1277 dump to fail on some systems. Thanks to Eric Jergensen
1278 <eric@dvns.com> for reporting the bug and submitting the patch.
1280 4. Handle more than 16 characters for the device names in dumpdates.
1281 (up to 255 now). Thanks to Rainer Clasen <bj@ncc.cicely.de> for
1282 tracking down the problem and proposing the solution.
1284 5. Fixed a bug in dump which prevented the creation of the
1285 dumpdates file when doing a 0-level dump without already
1286 having a dumpdates file. Thanks to Patrik Schindler
1287 <poc@pocnet.net> for reporting the bug.
1289 6. Changed the way dump 'S' flag reports the size estimate
1290 from number of blocks into bytes (making it compatible
1291 with the Solaris version, and simplifying things for
1292 amanda users). Thanks to Jason L Tibbitts III
1293 <tibbs@math.uh.edu> for reporting the bug.
1295 7. Fixed a compatibility problem in linux/alpha dump tape format.
1296 Now the linux/alpha dump are (again) compatible with the
1297 other dump formats. But this breaks compatibility with
1298 older dumps made on alpha. Thanks to Mike Tibor
1299 <tibor@lib.uaa.alaska.edu> for helping me in finding this bug.
1301 Changes between versions 0.4b10 and 0.4b11 (released December 5, 1999)
1302 ======================================================================
1304 1. Added a '--enable-kerberos' to configure.
1306 2. Added a 'S' option to dump which determines the amount of space
1307 that is needed to perform the dump without actually doing it, similar
1308 to the Sun's ufsdump 'S' option. Patch contributed by Rob Cermak
1309 <cermak@ahab.rutgers.edu>.
1311 3. Added a 'M' multi-volume option to dump and restore which enables
1312 dumping to multiple files (useful when dumping to an ext2
1313 partition to make several dump files in order to bypass the 2GB
1314 file size limitation). The 'f' argument is treated as a prefix and
1315 the output files will be named <prefix>001, <prefix>002 etc. With
1316 the 'M' flag, restore automatically selects the right file without
1317 asking to enter a new tape each time.
1319 4. Fixed a memory leak which was causing dump to grow very big
1320 (270MB when dumping a 10GB filesystem...). Thanks to Jason
1321 Fearon <jasonf@netrider.org.au> for reporting the bug.
1323 Changes between versions 0.4b9 and 0.4b10 (released November 21, 1999)
1324 ======================================================================
1326 1. Make configure test if the system glob routines support
1327 extended syntax (ALTDIRFUNC). If not, use the internal glob
1328 routines instead of system ones. Thanks to Bernhard Sadlowski
1329 <sadlowsk@Mathematik.Uni-Bielefeld.DE> for reporting the bug
1330 and helping me resolve this and other minor libc5 compiling
1333 2. Fix a problem when dumping a ext2fs with the 'filetype'
1334 feature enabled. Thanks to Patrick J. LoPresti
1335 <patl@cag.lcs.mit.edu> for reporting the bug and to
1336 Theodore Y. Ts'o <tytso@mit.edu> for providing the patch.
1338 3. Made the nodump flag work on directories. A directory which
1339 has the nodump flag gets never dumped, regardless of its
1342 4. Integrate a patch from Jeremy Fitzhardinge <jeremy@goop.org>
1343 which allows dump on an active ext3 filesystem. However, this
1344 is a "quick and dirty" patch which enables backup of an ext3
1345 filesystem through the ext2 compatibility (by ignoring the
1346 NEEDS_RECOVERY bit). The journal file is not recognized and
1347 it is dumped (it should not).
1349 5. Test the superblock compatibility flags when dumping, in order
1350 to be sure that we know how to deal with specific features.
1352 Changes between versions 0.4b8 and 0.4b9 (released November 5, 1999)
1353 ====================================================================
1355 1. Use lchown instead of chown, fixing a possible security problem
1356 when restoring symlinks (a malicious user could use this
1357 to deliberately corrupt the ownership of important system files).
1358 Thanks to Chris Siebenmann <cks@utcc.utoronto.ca> for detecting
1359 this and providing the patch.
1361 Changes between versions 0.4b7 and 0.4b8 (released November 3, 1999)
1362 ====================================================================
1364 1. Put dump sources under CVS, added Id tags in all files so
1365 one can use 'ident' on binary files.
1367 2. Added the dump/restore version in the usage text so one can
1368 easily verify the version he is using.
1370 3. Small patch from Nuno Oliveira <nuno@eq.uc.pt> which fixes
1371 a va_start/va_end problem on linux-ppc (always call va_start
1372 va_end in pairs each time we use a vararg function).
1374 4. Added again the DT_* constants because old libc does not
1375 contain them :(. Thanks to Eric Maisonobe <virnet@nat.fr>
1376 for submitting the bug report.
1378 5. Use ext2fs_llseek instead of llseek. With recent e2fsprogs
1379 this should enable dumping big (huge) filesystems.
1381 6. Added the RSH environment variable in order to be able to
1382 use a rsh replacement like ssh when doing remote backups (and
1383 bypass the security limitations of rcmd). Now you can do remote
1384 backups without being root (or making dump setuid root).
1386 7. Modified again the way dumpdates works. For incremental dumps,
1387 we need to read dumpdates even if we are not using 'u' option.
1388 Thanks to Bdale Garbee <bdale@gag.com> for his ideas on how
1391 Changes between versions 0.4b6 and 0.4b7 (released October 8, 1999)
1392 ===================================================================
1394 1. Removed the 'k' flag from the restore 'about' text if kerberos
1395 was not compiled in.
1397 2. Prototyped (f)setflags from e2fsprogs and corrected the calls
1398 to them (fsetflags takes a char*, setflags an open fd!).
1400 3. (f)setflags is called only if the flags aren't empty. If the
1401 file is a special file, a warning is printed, because changing
1402 flags implies opening the device. Normally, a special file
1403 should not have any flag... (Debian bug #29775, patch provided
1404 by Abhijit Dasgupta <abhijit@ans.net>).
1406 4. Made possible to dump a file system not mentioned in /etc/fstab.
1407 (Debian bug #11904, patch provided by Eirik Fuller <eirik@netcom.com>).
1409 5. Changed the default behaviour to not create dumpdates
1410 unless 'u' option is specified. Removed the old "debian-patch"
1411 which provided the same thing. (Debian bug #38136, #33818).
1413 6. Removed all those dump*announce, since they were getting old...
1415 7. Added warning messages if dumpdates does not exist and
1416 when an inferior level dump does not exist (except for a level 0
1419 8. Debugged the glob calls in interactive mode: restore used a
1420 dirent struct which was different from the /usr/include/dirent.h
1421 one (this used to work, is it a glibc2 change?), so none of the
1422 compat glob (which used /usr/include/dirent.h) or the system glob
1423 worked. Restore use now the system dirent (and the system
1424 DT_* constants), which are compatible with BSD ones.
1426 9. Added a configure flag (--with-dumpdatespath) to specify
1427 the location of dumpdates file. By default, it is
1430 10. Added the "AUTHOR" and "AVAILABILITY" sections and
1431 included the current date/version in man pages.
1433 11. Corrected the estimation of remaining time when
1434 the operator doesn't change the tapes quickly enough. This
1435 was an old bug, I thought I corrected it, and discovered
1436 that in fact it was corrected in two different places, so
1437 the results canceled each other...
1439 Changes between versions 0.4b5 and 0.4b6 (released October 1, 1999)
1440 ===================================================================
1442 1. Integrated multiple patches from RedHat, Debian and SuSE:
1444 - tweak dump/itime.c to not try to read dumpdates if the 'u' option
1446 - several fixes in the man pages.
1447 - update the default tape device to /dev/st0.
1448 - many updates for Linux Alpha (byte ordering, size_t etc).
1450 - use environment variable for TMPDIR (instead of /tmp).
1451 - use sigjmp_buf instead of jmp_buf (RedHat bug #3260).
1452 - workaround egcs bug (RedHat bugs #4281 and #2989).
1453 - wire $(OPT) throughout Makefile's.
1455 2. Upgrade the dump revision to 1, making possible to dump filesystems
1456 made with e2fsprogs-1.15 or newer. Nothing seems to break...
1458 3. Fix some compile warnings, prototype all functions.
1460 4. Use glibc err/glob instead of internal compatibility
1461 routines (only if available).
1463 5. Fix a compile error on Linux 2.2.7 / libc5 (5.4.44) (patch provided
1464 by Bernhard Sadlowski <sadlowsk@mathematik.uni-bielefeld.de>).
1466 Changes between versions 0.4b4 and 0.4b5 (released September 22, 1999)
1467 ======================================================================
1469 1. Integrated the changes from FreeBSD-3.1-RELEASE
1470 (mostly bug fixes, buffer overruns, dump has now an "automatic
1471 tape length calculation" flag, dump/restore can use kerberos now
1472 (this is NOT tested), use environment variables for TAPE and
1475 2. Integrated three RedHat patches ("glibc", "kernel" and "bread" patches)
1477 3. Corrected a bug in restore when using 'C' option with multi-volumes
1478 tapes (files splited accros two tapes give "size changed" errors
1481 4. Corrected the long standing bug when dumping multiple tapes.
1482 This works for me, needs further testing.
1484 Changes between versions 0.4b3 and 0.4b4 (released January 17, 1997)
1485 ====================================================================
1487 1. Dump now runs correctly on kernels 2.1.x
1488 Fix made by Gerald Peters <gapeters@worldnet.att.net>
1490 Changes between versions 0.4b2 and 0.4b3
1491 ========================================
1493 1. Use realpath() if available
1495 2. Report statistics
1497 Changes between versions 0.4b1 and 0.4b2
1498 ========================================
1500 1. Fixed the bug fix from Greg Lutz (I had made a mistake when integrating
1503 2. Fixed restore to make it able to read FreeBSD 2.x dumps again
1505 3. Fixed configure.in to correctly handle --enable-rmt
1507 Changes between versions 0.3 and 0.4b1
1508 ======================================
1510 1. Integrated the changes from 4.4BSD-Lite2
1512 2. Integrated the patches from Debian and Red Hat
1514 3. Portability changes: use the __u32, __u16, __s32, and __s16 types
1516 4. Changed dump to use the Ext2fs library to get block addresses. This
1517 should solve the endianness problem on SparcLinux.
1519 5. Created a configure.in file (shamelessly stolen from the e2fsprogs
1520 distribution's one) to use autoconf
1522 6. Fixed a few minor bugs
1524 Changes between versions 0.2e and 0.2f
1525 ======================================
1527 1. Added the creation of named pipes (fifos) in restore.
1529 2. Added the -N flag in the restore manual page.
1531 3. Added the file kernel.patch which contains the llseek() optimization
1532 patch for 1.2.x kernels.
1534 4. Fixed a bug in the restoration of symbolic links: owner and group were
1537 5. Integrated some changes from FreeBSD 2.2.
1539 6. Added a call to ftruncate() after restoring each file to restore
1540 correctly files ending by a hole.
1542 Changes between versions 0.2d and 0.2e
1543 ======================================
1545 1. Fixed a bug in the "set owner/mode" process. Every file was restored
1546 with owner = root (0) and group = root/wheel/whatever (0).
1548 Changes between versions 0.2c and 0.2d
1549 ======================================
1551 1. Dump is now able to backup 2GB+ filesystems.
1553 2. Dump and restore can now be linked as static binaries.
1555 Changes between versions 0.2b and 0.2c
1556 ======================================
1558 1. Fixed a bug when dumping ``slow'' (i.e. normal) symbolic links.
1560 Changes between versions 0.2a and 0.2b
1561 ======================================
1563 1. Really fixed the bug that I should have corrected in 0.2a.
1565 2. Enabled optimization again.
1567 Changes between versions 0.2 and 0.2a
1568 =====================================
1570 1. Disabled the optimization during compilation.
1572 Changes between versions 0.1 and 0.2
1573 ====================================
1575 1. Fixed a bug in fstab.c which caused a null pointer to be stored in
1576 the fs_type field (actually, I modified the file fstab.c to make it
1577 use the mntent functions).
1579 2. Dump and restore now use a 4.3 BSD compatible dump format. Backups
1580 made by dump should be readable by the BSD restore and backups made
1581 by the BSD dump should be readable by restore. Unfortunately, this
1582 means that the dump format has changed between version 0.1 and version
1585 3. Dump is now able to backup a subtree, it is no longer limited to whole
1586 filesystems like the BSD version.
1588 4. Dump now uses ext2_llseek() so it is able to backup filesystems bigger
1591 Changes between versions 0.0 and 0.1
1592 ====================================
1594 1. Now create links rdump and rrestore during the `make install' step.
1596 2. Linux port specific bugs added to the manual pages
1598 3. Incorrect estimation of the number of tapes blocks fixed when doing
1599 incremental backups.
1601 4. Better ls-like format in restore in interactive mode.