Fix invalid EA metainformation in dump, add a workaround to restore to
[dump.git] / CHANGES
1 $Id: CHANGES,v 1.309 2010/03/22 15:40:55 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
37 Changes between versions 0.4b41 and 0.4b42 (released June 18, 2009)
38 ===================================================================
39
40 1. Fix printout of the 'Connection to' message in dump/restore.
41 Thanks to Dale Wiles <dwiles@users.sourceforge.net> for the
42 bug report and its fix.
43
44 2. Fix dump -w/-W output for filesystems having the last backup
45 level equal to 0. Thanks to Pascal Bouchareine
46 <kalou@users.sourceforge.net> for reporting the bug.
47
48 3. Enable restore to handle restoring onto a different SELinux
49 policy from the dump. Thanks to Tony Nelson
50 <tony_nelson@users.sourceforge.net> for the patch.
51 (see http://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=189845
52 for details)
53
54 4. Fix EA set failures when restoring immutable files. Thanks
55 to Andrew Kroeger for the patch.
56
57 5. Fix "mode file too large" errors when restoring huge backups.
58 Thanks to Kenneth Porter <shiva@sewingwitch.com> for the bug
59 report.
60
61 6. Add the ability to link against libtinfo library if present,
62 instead of libtermcap or libncurses libraries. Thanks to
63 Adam Tkac <atkac@redhat.com> for the patch.
64
65 7. When comparing a backup, do not consider to be an error the fact
66 that we encounter files created while dump was in progress.
67 Thanks to Kenneth Porter <shiva@sewingwitch.com> for the bug
68 report.
69
70 8. Force '-a' to be the default value when running 'restore -C'.
71 Thanks to Kenneth Porter <shiva@sewingwitch.com> for the bug
72 report.
73
74 9. Many improvements to the 'cron_dump_to_disk' example.
75 Thanks to Aaron S. Hawley <ashawley@users.sourceforge.net> for
76 the patch.
77
78 10. Fix the QFA generation when extended attributes are backuped
79 along with the inodes. Prior to this fix, the entries for
80 some inodes may get corrupted (duplicated entries or incorrect
81 ones), making the QFA file unusable for repositionning in
82 restore. Many thanks to Kenneth Porter <shiva@sewingwitch.com>
83 for the bug report and the testing of the fix.
84
85 11. Fix EA restoration in debug mode (Sourceforge bug #1986216).
86
87 12. Stefan Auracher <star7@onlinehome.de> noticed that the Tower
88 of Hanoi backup strategy documented in the man page wasn't
89 actually based on the Tower of Hanoi algorithm. The man page
90 was updated accordingly.
91
92 13. Fix missing level information from dump output when using
93 the default level (RedHat bug #493635). Thanks to Adam Tkac
94 <atkac@redhat.com> for the patch.
95
96 14. Fix the dump man page to reflect the fact that the default
97 dump level is 0 not 9 (RedHat bug #356121). Thanks to Adam Tkac
98 <atkac@redhat.com> for the patch.
99
100 15. Fixed a few spelling errors in the man pages (RedHat bug #489853).
101 Thanks to Adam Tkac <atkac@redhat.com> for the patch.
102
103 16. Fixed restore man page and program usage which incorrectly stated
104 that -P and -A were both allowed in the same invocation (RedHat
105 bug #490627). Thanks to Adam Tkac <atkac@redhat.com> for the patch.
106
107 17. Use sys_clone under Linux to share I/O contexts between dump
108 processes, thus drastically increasing the performance of dump
109 under CFQ (which is the default I/O scheduler used in a number of
110 distributions). Thanks to Jeff Moyer <jmoyer@redhat.com> for the
111 patch.
112
113 18. Add (preliminary) ext4 support - thanks to libext2fs which does
114 all the job for us. Thanks to Gertjan van Wingerde
115 <gwingerde@gmail.com> for the patch.
116
117 Changes between versions 0.4b40 and 0.4b41 (released January 2, 2006)
118 =====================================================================
119
120 1. Fix restore of dumped Access Control Lists. The previous
121 code for EA works fine, but ACLs needed conversion from the
122 ext2/3 disk format to posix_acl format before restoring.
123
124 2. Fix some issues with restoration of EA on big endian
125 platforms.
126
127 3. Fix restore when the symtab is over 2GB in size.
128
129 4. Made the directory hash indexing an optional feature,
130 accessible by the '-H' option of restore, and disable it
131 by default.
132
133 5. Fixed dump to not include extended attributes information
134 in the toc (archive) file which confused restore -t.
135
136 Changes between versions 0.4b39 and 0.4b40 (released May 2, 2005)
137 =================================================================
138
139 1. Changed restore to emit warnings (instead of emitting a fatal
140 error) if a file (or a directory) is unavailable for a
141 comparision (if the user doesn't have the necessary permissions
142 to access it for example). Thanks to Kenneth Porter
143 <shiva@sewingwitch.com> for the bug report.
144
145 2. Re-done the 'do not save directory entries to non-dumped inodes
146 (excluded from dump)' feature. The previous implementation
147 worked well for excluded directories but not for regular files.
148 Thanks to Kenneth Porter <shiva@sewingwitch.com> for the bug
149 report.
150
151 3. Fixed a bug in dump where the tape size was miscalculated when
152 the user used -d/-s to specify the tape characteristics. Thanks
153 to Philip Goisman <goisman@physics.arizona.edu> for reporting
154 the bug.
155
156 4. Fixed another bug introduced in restore with the hashtree
157 implementation. This one caused restore to stop saying
158 "removenode: non-empty directory" in some cases.
159
160 5. Added support for dumping and restoring ext2/3 extended
161 attributes (EA), like the access control lists (ACL) or
162 the security labels used by SELinux.
163
164 Changes between versions 0.4b38 and 0.4b39 (released January 21, 2005)
165 ======================================================================
166
167 1. The newly added dump_on_cd_3 example was buggy, replace it
168 with an updated version from Andrew Basterfield
169 <bob@cemetery.homeunix.org>.
170
171 2. Made restore to chdir() back into the initial directory when
172 dumping core while aborting a comparision operation. The
173 previous behaviour was to write the corefile at the root of
174 the directory being compared, which could very well be
175 read only and preventing the corefile generation. Thanks
176 to Kenneth Porter <shiva@sewingwitch.com> for the bug report.
177
178 3. Silenced the failure to call fgetflags() when comparing an
179 entry which has no ext2 attributes (as in lsattr()).
180
181 4. Fix a brown paper bug in restore -C which broke restore and
182 caused modifications on the filesystem being compared
183 (directories containing a file with the same name as the
184 directory get renamed to RSTTMP...). Thanks to Kenneth Porter
185 <shiva@sewingwitch.com> for finding the bug and helping me
186 reproduce it.
187
188 5. Made restore -C force the -N flag (no writing allowed on
189 the disk) in order to prevent more bugs like the above one.
190
191 Changes between versions 0.4b37 and 0.4b38 (released January 7, 2005)
192 =====================================================================
193
194 1. Fix a couple of troff syntax bugs in the man pages.
195 Thanks to Eric Raymond <esr@thyrsus.com> for the patch.
196
197 2. Made restore use either libncurses or libtermcap, depending
198 on which one is available at configure time.
199
200 3. Fixed restore negative size display bug when comparing a
201 dump containing files over 2GB. Thanks to Steve Bonds
202 <sbonds@users.sourceforge.net> for the bug report.
203
204 4. Do not save directory entries to non-dumped inodes
205 (excluded from dump). This will eliminate the 'missing
206 file' warnings when doing 'restore -C'.
207
208 5. Fix dump crash when backuping a huge (2TB) filesystem,
209 due to a bogus calculation on the inode map size.
210 Thanks to Kevin B. Haines <K.B.Haines@rl.ac.uk> for
211 submitting the bug and testing the fix.
212
213 6. Fix a problem in restore where the final \0 in the symbolic
214 link names could have been lost, generating corrupt filenames.
215 Thanks to Kyle Wilson <kyle.wilson@amd.com> for reporting the
216 bug.
217
218 7. Implemented a hash list for the directory names in restore.
219 The linear list used before caused problems in interactive
220 restores when dealing with directories having thousands of
221 entries. Thanks to Brian Ristuccia <bristuccia@starentnetworks.com>
222 for reporting the bug.
223
224 8. Improved restore -C, this time including the directory
225 attributes into the comparision.
226
227 9. Made restore understand tapes containing EA/ACLs (which will
228 be dumped by the next version of dump). In this version
229 extended attributes on the tape are ignored, for full EA/ACL
230 support wait for the next version or try the experimental EA
231 patch.
232
233 Changes between versions 0.4b36 and 0.4b37 (released July 7, 2004)
234 ==================================================================
235
236 1. Added the --enable-staticz configure option which enables
237 dump and restore to be linked with static versions of
238 libz and libbz2 (and dynamic versions of all other libraries).
239 This will make Debian users happy, because libz and libbz2
240 were the only needed libraries living in /usr, all the
241 others live in /lib. In case of system emergency, it is
242 better not to have to rely on an extra filesystem.
243 Thanks to Bdale Garbee <bdale@gag.com> for the suggestion.
244
245 2. Fix compilation on (at least the Linux Debian port to) AMD64.
246 (<ext2fs/ext2_types.h> defines some types (__s64 and __u64)
247 that are also defined by <linux/types.h> (<asm/types.h>) and
248 they conflict).
249
250 3. Make dump's reading of the dumpdates file a bit more robust,
251 preventing dump from crashing when the dumpdates file has
252 been modified by hand.
253
254 4. Fixed some offset calculations in dump code which could
255 lead to "bread lseek errors" on large filesystems. Thanks
256 to Bruce Lowekamp <lowekamp@users.sourceforge.net> for
257 reporting this bug and debugging the issue.
258
259 5. Made dump use the blkid library when searching for devices
260 by label or uuid instead of dump's own routines.
261
262 6. Corrected a bug in dump where a wrong LABEL=... line in
263 /etc/fstab could prevent dump from dumping unrelated
264 filesystems. Thanks to Bruce Lowekamp
265 <lowekamp@users.sourceforge.net> for reporting the bug.
266
267 Changes between versions 0.4b35 and 0.4b36 (released April 21, 2004)
268 ====================================================================
269
270 1. Fixed dump compilation with old gcc versions. Thanks to
271 Mike Castle <dalgoda@users.sourceforge.net> for the patch.
272
273 2. Fixed some warnings (howmany, roundup, powerof2 redefined)
274 when compiling against a recent glibc version.
275
276 3. Fixed a bug in restore preventing the read of a dump tape
277 written with Solaris 7 ufsdump. Thanks to Patrick Higgins
278 <phiggins@transzap.com> for reporting the bug and providing
279 the test case.
280
281 4. Changed dump to enable the creation of volumes of different
282 sizes in a single run (make -B accept a list of values).
283 Patch contributed by Florian Zumbiehl <florz@gmx.de>.
284
285 5. Use the glibc provided minor() and major() macros instead
286 of our own bitmask implementation. This should be safe for
287 when the major/minor namespace will migrate to 32 bits.
288 Thanks to Zhang Jun <zhangjun@nanjing-fnst.com> for reporting
289 the bug.
290
291 6. Made explicit in the dump man page that dump will not create
292 a remote file, it will only write to an already existing one.
293
294 7. Another try at making size estimates better again.
295
296 8. Put back the inconditional running of the end-of-tape
297 script which was changed in 0.434 to be run only when -M
298 or multiple -f were NOT specified. Some users rely on this
299 feature even when it is combined with -M/-f.
300
301 9. Fixed restore when restoring huge backups (where rstdir...
302 temporary files are over 2GB). Thanks to
303 Raphael Jaffey <rjaffey@artic.edu> for reporting this,
304 debugging the issue and making the patch.
305
306 10. Made restore understand FreeBSD UFS2 tapes. Thanks to
307 David <vrtin@users.sourceforge.net> for submitting the bug
308 and providing a test case.
309
310 11. Made dump work with an arbitrary integer level (no more 10
311 levels only limitation). Thanks to Matthew
312 <msvincen@midway.uchicago.edu> for the patch.
313
314 Changes between versions 0.4b34 and 0.4b35 (released December 21, 2003)
315 =======================================================================
316
317 1. Added a note in the dump man page saying that the default
318 blocksize can be 32 if -d is used with a high density argument.
319 Thanks to Antonios Christofides <A.Christofides@itia.ntua.gr>
320 for the patch.
321
322 2. Fixed configure to correctly understand CPPFLAGS, CFLAGS,
323 LDFLAGS environment variables. Thanks to Arcady Genkin
324 <antipode@users.sourceforge.net> for reporting the bug.
325
326 3. Made -e/-E options of dump accept an unlimited number of inodes
327 to be excluded, rather than a hardcoded maximum. Thanks to
328 Dietrich Rothe <d-rothe@users.sourceforge.net> for the patch.
329
330 4. Updated the autoconf system to 2.50. Forced the -D_BSD_SOURCE
331 and -D_USE_BSD_SIGNAL defines in configure in order to solve
332 64bit build problems because quad_t is redifined with a
333 different signature. Thanks to Mike Harris <mharris@redhat.com>
334 for reporting this bug.
335
336 5. Made restore build on Solaris, making possible to
337 restore Linux's "enhanced" tapes. Thanks to Uwe Gohlke
338 <uwe@ugsoft.de> for the patch.
339
340 6. Made an extension in the dump tape format capable of saving
341 MacOSX specific inode extensions. Uwe Gohlke <uwe@ugsoft.de>
342 wrote the extension and contributed the restore code back
343 into this codebase. The same extension mechanism will be
344 used in the future to save ACLs...
345
346 7. Made rmt work correctly with regard to QFA and local/remote
347 files and tapes. The remote access will however work only
348 when the dump provided rmt version is used. If you want to
349 use another rmt server, please do not use the QFA feature.
350 Thanks to Uwe Gohlke <uwe@ugsoft.de> for the patch.
351
352 Changes between versions 0.4b33 and 0.4b34 (released April 18, 2003)
353 ====================================================================
354
355 1. Fixed the INSTALL file to reflect the actual install paths.
356 Thanks to David Ronis <ronis@ronispc.chem.mcgill.ca> for
357 reporting the bug.
358
359 2. Fixed the configure script to only check for headers presence
360 instead of trying to compile them. This should fix issues
361 with old build environments. Thanks to Kari Lempiainen
362 <kari@funky.pp.fi> for reporting the bug.
363
364 3. Fixed restore to correctly ignore sockets when comparing
365 a dump (as socket cannot be properly restored anyway). Thanks
366 to Gunther Reiszig <gunther@mit.edu> for reporting the bug.
367
368 4. Fixed restore to correctly access the archive file (-A argument)
369 even when using a remote tape. Thanks to Fabrice Bellet
370 <fabrice@bellet.info> for reporting the bug.
371
372 5. Fixed (again) handling of long (largefile) seeks in rmt.
373 Thanks to Fabrice Bellet <fabrice@bellet.info> for reporting
374 the bug.
375
376 6. Fixed restore corner case when dealing with large block sizes
377 dump is able to write now (-b 1024). Thanks to Fabrice Bellet
378 <fabrice@bellet.info> for reporting the bug.
379
380 7. Fixed a bug preventing dump to access a filesystem having
381 a label exactly 16 bytes in length. Thanks to <minduni@ti-edu.ch>
382 for reporting the bug.
383
384 8. Made dump store dump dates using explicit timezones, fixing a
385 problem with incremental dumps if the timezone is modified
386 between the dumps. Thanks to Philippe Troin <phil@fifi.org> for
387 the bug report and the patch.
388
389 9. Fixed a bug encountered when dumping individual files (not full
390 filesystems or directories) and dangling symbolic links happen
391 to be in the list of files. For as far as dump is concerned,
392 dangling symbolic links are allowed, and are dumped as is.
393 Thanks to Jin-su Ahn <jsahn@ee.snu.ac.kr> for reporting the
394 bug and providing the fix.
395
396 10. Fixed open and creation modes and permissions for QFA and
397 table-of-contents files in dump and restore. Thanks to
398 Philippe Troin <phil@fifi.org> for the patch.
399
400 11. Fixed the archive file descriptor handling enabling it to be 0.
401 This can happen in some cases when shell redirections are used.
402 Thanks to Philippe Troin <phil@fifi.org> for the patch.
403
404 12. Delayed the opening of archive file until after suid had been
405 dropped (fixing a possible security issue if dump is suid).
406 Thanks to Philippe Troin <phil@fifi.org> for the patch.
407
408 13. Fixed the 'S' command handling in the rmt client part.
409 Thanks to Philippe Troin <phil@fifi.org> for the patch.
410
411 14. Modified the end-of-tape script handling to print out statistics
412 (and stop the timer) before launching the eot script. Also, the eot
413 script does not get run anymore when using -M (which makes sense) or
414 when multiple tapes are listed on the command line
415 (-f tape0,tape1,tapen) (which also makes sense).
416 Thanks to Philippe Troin <phil@fifi.org> for the patch.
417
418 15. Relicensed dump/restore under the 'revised' BSD license, as per
419 ftp://ftp.cs.berkeley.edu/ucb/4bsd/README.Impt.License.Change.
420
421 16. Added LZO compression to dump. This new compression method has
422 the advantage of being super fast, thus not killing tape streaming
423 on slow machines. Thanks to Helmut Jarausch
424 <jarausch@igpm.rwth-aachen.de> for the patch and to
425 Markus Oberhumer <markus@oberhumer.com> for giving special permission
426 to include his miniLZO project (GPL licensed) in dump/restore.
427
428 17. Some small buffer overruns fixes in rmt. Thanks to Antonomasia
429 <ant@notatla.demon.co.uk> for reporting the bugs.
430
431 18. Added a special rmt version which can do encryption when writing
432 to tape. Read examples/encrypted_rmt/README for details on
433 how to enable and configure it. Thanks to Ken Lalonde
434 <ken@globalremit.com> for the patch.
435
436 19. Made dump work with 2.5 kernel end of tape early warning semantics.
437 Thanks to Kai Makisara <Kai.Makisara@kolumbus.fi> for the patch.
438
439 20. Fixed a bug which caused dump -w|-W not to work anymore, because
440 the fs_freq and fs_passno fields in /etc/mtab are always set
441 to 0 0. Thanks to Trent Piepho <xyzzy@speakeasy.org> for
442 reporting the bug.
443
444 Changes between versions 0.4b32 and 0.4b33 (released February 10, 2003)
445 =======================================================================
446
447 1. Added a note in the restore man page clarifying the question
448 'set the permissions on the current directory ?' asked by
449 restore at the end of treatment in -i and -x modes.
450
451 2. Fixed the endianess issues when reading compressed tapes.
452 Thanks to Dark Force <daq4th@users.sourceforge.net> for
453 reporting this bug and providing test cases.
454
455 3. Fixed the "ACL won't be dumped" warning message (which showed
456 an extra, unrelated error message). Thanks to Dragan Krnic
457 <dkrnic@lycos.com> for reporting this bug.
458
459 4. Made dump look first into /etc/mtab, then into /etc/fstab
460 when searching for filesystem paths. Also fixed some problems
461 caused by binding mounts or multiple block device mounts.
462 Thanks to Matus Uhlar <uhlar@fantomas.sk>, Elliott Mitchell
463 <ehem@m5p.com>, Greg Edwards <gedwards@users.sourceforge.net>,
464 Brian Hoy <brian.hoy@opus.co.nz>. (fixes Debian bugs #147086
465 and #160305, Sourceforge bugs #618699 and #679832).
466
467 5. Made dump's -I option accept the value '0' meaning all the
468 read errors will be ignored. This can be useful when running
469 dump from unattended sessions (like cron jobs). Thanks to
470 John I Wang <jiwang@users.sourceforge.net> for the suggestion.
471
472 6. Fixed the output of dump to indicate 'blocks' instead of
473 'tape blocks' in the various messages (blocks are always
474 1 Kilobyte, tape blocks are 1 BK * '-b' argument), and
475 made it clearly print the current blocksize at the start of
476 a dump. Thanks to Michal Szymanski <msz@astrouw.edu.pl> for
477 the suggestions.
478
479 7. Made rmt understand long (largefiles) seeks.
480
481 8. Fixed build with very old versions of libext2fs, where
482 EXT2_FT_* constants were undefined.
483
484 9. Made dump accept the dumpdates path on the command line
485 (-D file option) instead of using only the hardcoded one.
486 Thanks to Piete Brooks <pb22@users.sourceforge.net> for the
487 suggestion.
488
489 10. Enabled rmt, LFS, readline, QFA options by default in
490 ./configure. Updated the configure process (new versions
491 of config.guess, config.sub etc).
492
493 Changes between versions 0.4b31 and 0.4b32 (released November 15, 2002)
494 =======================================================================
495
496 1. Changed dump to use fcntl(F_SETLK) style locking instead
497 of flock() when locking the dumpdates file. With the old
498 locking scheme, a local user having read rights on the
499 dumpdates file could be able to do a Denial of Service attack
500 on dump. In order to lock the dumpdates file with the new
501 scheme, the user would need to have write access on the file.
502 Thanks to Richard Johnson <Richard.Johnson3@ey.com> for
503 reporting the bug (originally a bugtraq post).
504
505 2. Fixed interactive 'ls' which caused spurious errors warnings
506 about 'undefined filetypes' detected. Thanks to Jorgen Ostling
507 <jorgen_ostling@users.sourceforge.net> for reporting this
508 bug.
509
510 3. Fixed dump's estimate when dealing with sparse inodes.
511
512 4. Modified dump to allow setting a blocksize bigger than 32kB
513 (raised the limit to 1024kB), because newer hardware needs
514 this for proper operation. Thanks to Dirk Traenapp
515 <dtraenapp@users.sourceforge.net> for reporting this.
516
517 5. Fixed a bug causing Dump to stop and report an error if an
518 inode number in the exclude file was followed by some amount
519 of whitespace. Thanks to Jeffrey Sofferin
520 <sofferin@users.sourceforge.net> for reporting this bug.
521
522 6. Fixed a bug which caused restore, in some particular cases,
523 to ask some 'scary' questions and leave a bunch of RSTTMP
524 directories behind when restoring incremental tapes. Thanks
525 to Philippe Troin <phil@fifi.org> for reporting this bug and
526 providing the test cases.
527
528 7. Changed the wording when inodes are excluded from dump:
529 replaced 'Added inode 7 to exclude list' with
530 'Excluding inode 7 (resize inode) from dump', as suggested
531 by Elliott Mitchell <ehem@m5p.com> in a Debian bug report.
532
533 Changes between versions 0.4b30 and 0.4b31 (released July 30, 2002)
534 ===================================================================
535
536 1. Fixed rmt open flags transmission (GNU's symbolic syntax over
537 rmt) which I broke in 0.4b29. Thanks to Eros Albertazzi
538 <eros@lamel.bo.cnr.it> for reporting the bug.
539
540 Changes between versions 0.4b29 and 0.4b30 (released July 25, 2002)
541 ===================================================================
542
543 1. Made dump print out the ext2 logical block number in case of a read
544 error, which can be given as an argument to debugfs' ncheck command,
545 in order to find out the affected inode. Added note about this
546 usage in dump's man page.
547
548 2. Fixed a problem in restore when reading tapes written on big
549 endian machines with very old versions of dump. The patch was
550 contributed by George Helffrich <george@geology.bristol.ac.uk>.
551
552 3. Fixed the tape length calculation when using large tapes
553 and compression. Thanks to Georg Lippold
554 <g_lippold@sourceforge.net> for reporting the bug.
555
556 4. Added a new set of examples in dump_on_cd_2 directory, based
557 on dump_on_cd examples but somewhat enhanced, supporting DVD
558 media, and localized in english and german. Thanks to
559 Georg Lippold <g_lippold@sourceforge.net> for the new scripts.
560
561 5. Made dump save 32 bit UID/GID. Internally, this was achieved
562 by switching from the old BSD inode format to the new BSD
563 inode format, which means that the tape format was changed.
564 However, since all restore versions out there should
565 transparently support both inode formats, the change should
566 have no side effects. Thanks to John Yu <jky@cs.bu.edu> for
567 reporting the bug.
568
569 6. Fixed a lot of warnings in the code shown when compiling
570 with 'gcc -W'. Thanks to Matthias Andree
571 <matthias.andree@stud.uni-dortmund.de> for reporting this.
572
573 7. Fixed a small markup bug in the dump man page. Thanks to
574 Eric S. Raymond <esr@minx.thyrsus.com> for submitting the
575 patch.
576
577 8. Rewrote entirely the man pages using the tmac.an macro
578 package (Linux man page format) instead of the original BSD
579 format. They should be now cleaner and easier to modify.
580
581 Changes between versions 0.4b28 and 0.4b29 (released June 8, 2002)
582 ==================================================================
583
584 1. Fixed a problem in the rmt ioctl command, where ioctl's issued from
585 non Linux clients were misinterpreted. The description of the problem
586 (incompatible numbering in Linux mtio opcodes) is documented at
587 ftp://ftp.fokus.gmd.de/pub/unix/star/README.mtio . Thanks to
588 Jörg Schilling <schilling@fokus.gmd.de> for reporting this bug and
589 providing an excellent, cross-platform replacement for rmt in his
590 star package.
591
592 2. Fixed a bug reported by Andy Mentges <amentges@jumpline.com>
593 which caused restore to fail when the inode maps were not entirely
594 contained on the first volume (when using really small volumes or
595 when dumping a huge number of inodes).
596
597 3. Fixed a problem in dump, where files in subdirectories of directories
598 that have the nodump flag set (or being excluded with -e/-E) were
599 sometimes incorrectly dumped (instead of being excluded). The original
600 fix is from David C Lawrence <tale@dd.org> in the FreeBSD version
601 of dump: http://www.FreeBSD.org/cgi/query-pr.cgi?pr=32414
602 Thanks to Ted Grzesik <tedgyz@roostme.com> for reporting the bug and
603 help testing the patch.
604
605 4. Added some example scripts from Gerd Bavendiek <bav@epost.de>
606 which makes one able to pipe the output of dump, by the net, to
607 a remote CD-burner server.
608
609 5. Made dump use O_CREAT|O_TRUNC both locally and remotely (over rmt),
610 and use GNU's symbolic syntax over rmt instead of numerical values
611 to assure multiple platform compatibility.
612
613 6. Documented the -d option in restore.
614
615 7. Added a -v (verbose) mode to dump. For now it just prints the number
616 of the inode being dumped, but this could evolve in future versions
617 to include interesting debugging output.
618
619 8. Added a -o flag to restore, which automatically restores the current
620 directory permissions in -i or -x mode, without asking the operator
621 whether to do so. Patch submitted by Tony Clayton <tonyc@e-smith.com>
622 and Peter Samuel <peters@e-smith.com>.
623
624 Changes between versions 0.4b27 and 0.4b28 (released April 12, 2002)
625 ====================================================================
626
627 1. Fixed a bug in the QFA feature which could have caused
628 a bad tape positionning on restore (causing extra delays in
629 restore). Credits go to Uwe Gohlke <uwe@ugsoft.de>.
630
631 2. Added a small note in the dump man page specifying that
632 there should be no white space between the option letter and
633 the -j or -z parameter, thanks to Kenneth Porter <shiva@well.com>
634
635 3. Made restore work with older versions of the readline library,
636 by checking in configure for several extended symbols. Restore
637 can now be compiled with a readline as old as the 2.0 release
638 (though it may be a good idea to upgrade it to a more recent
639 version...). Thanks to Andrew Donkin <ard@waikato.ac.nz> for
640 reporting the build failures.
641
642 4. Fixed a performance problem with the QFA file creation in
643 dump, which made unnecessary seeks on the tape slowing down
644 the dump. Thanks to Andrew Donkin <ard@waikato.ac.nz> for
645 reporting this issue.
646
647 5. Removed the inclusion of some kernel headers in the dump
648 source, which prevented the compile in some kernel/glibc
649 headers/architecture combination. Thanks to Bdale Garbee
650 <bdale@gag.com> for reporting the bug.
651
652 6. Added the appropriate error message when dump fails to
653 open the output file for writing. Thanks to Amith Varghese
654 <amithv@yahoo.com> for reporting this bug.
655
656 7. Made restore able to understand large Solaris ufsdump tapes
657 (containing inodes bigger than 4194304). Sun have introduced
658 an "extension" to the dump tape format when dealing with
659 those inodes, which was uncorrectly handled by Linux restore.
660 Thanks to Uwe Gohlke <uwe@ugsoft.de> for reporting the bug and
661 providing a test case.
662
663 8. Added the -m parameter to dump which optimises the output for
664 inodes having been changed but not modified since the last dump
665 ('changed' and 'modified' have the meaning defined in stat(2)).
666 For those inodes, dump will save only the metadata, instead of
667 saving the entire inode contents. Inodes which are either
668 directories or have been modified since the last dump are saved
669 in a regular way. Uses of this flag must be consistent, meaning
670 that either every dump in an incremental dump set have the flag,
671 or no one has it.
672
673 Changes between versions 0.4b26 and 0.4b27 (released February 15, 2002)
674 =======================================================================
675
676 1. Fixed behaviour of dump when exceeding resource limits
677 (SIGXFSZ treatment).
678
679 2. Added the -L flag to restore to allow the user to specify a
680 maximal allowed number of miscompares when using restore
681 with the -C option to check the backup.
682
683 3. Detailed the manual entry for the -N option of restore.
684
685 4. Added the -a flag to restore to make able doing unattended
686 restores in -i or -x mode (automatically walks through the
687 multiple dump volumes).
688
689 5. Extended the QFA mode to work with local files and/or
690 remote tapes and files. This way, restore can know in advance
691 the tape number and the offset for the inodes to extract and
692 can minimize the extraction time by seeking directly to the
693 good tape/offset.
694
695 6. Added the -A <archive> option to both dump and restore,
696 which makes dump to archive a dump table-of-contents in
697 the specified file to be used by restore to determine
698 whether a file is in the dump file that is being restored.
699 (the archive file syntax is also compatible with the
700 Solaris ufsdump generated one).
701
702 7. Small fix in restore making it able to read some (broken ?)
703 Solaris ufsdump tapes.
704
705 8. Fixed dump to correctly recognise the root filesystem when using
706 ext2 disk labels (LABEL=/). Thanks to John Yu <jky@it.bu.edu>
707 for reporting this bug.
708
709 9. Added the -P <file> option to restore to create a
710 Quick File Access file from an already made dump. Patch
711 contributed by Uwe Gohlke <uwe@ugsoft.de>.
712
713 10. Made restore compile and run on Solaris, making it a
714 possible replacement for the standard ufsrestore. Port was
715 contributed by Uwe Gohlke <uwe@ugsoft.de>.
716
717 Changes between versions 0.4b25 and 0.4b26 (released January 7, 2002)
718 =====================================================================
719
720 1. Added a set of backup scripts from Eugenio Diaz
721 <getnito@yahoo.com> in the examples section. It features
722 automatic (cron based) full and incremental dumping of
723 several filesystems on a separate filesystem.
724
725 2. Fixed a off-by-one miscalculation which disabled dumping
726 a one letter subdirectory. Thanks to Chris Siebenmann
727 <cks@utcc.utoronto.ca> for reporting the bug.
728
729 3. Fixed several restore bugs occuring when trying to
730 restore some missing files on the tape. Thanks to Chris
731 Siebenmann <cks@utcc.utoronto.ca> for reporting the bug.
732
733 4. Fixed --with-ldopts configure argument passing, installing from
734 a separate object directory, makefile cleanups contributed
735 by <splite@purdue.edu>.
736
737 5. Fix a bug which could caused, in some conditions, the highest
738 number inode of a filesystem, to not be dumped. Many thanks
739 to Chris Siebenmann <cks@utcc.utoronto.ca> for helping me
740 chase this bug.
741
742 Changes between versions 0.4b24 and 0.4b25 (released November 17, 2001)
743 =======================================================================
744
745 1. Added a mini howto from Patrick Walsh in the examples
746 directory.
747
748 2. Minor man pages syntax corrections. Thanks to
749 Chris Danis <screechco@home.com> for reporting the bugs.
750
751 3. Added a script from David B. Peterson <dave@toppledwagon.com>
752 to the examples section. It features dumping several
753 filesystems to a remote tape drive upon ssh.
754
755 4. Added a patch provided by Richard Jones <rich@annexia.org>
756 which allows BRADEMAX (number of read errors tolerated by
757 dump) to be adjusted using the -I option.
758
759 5. Fixed a bug which disabled doing "restore -C -f -". Thanks
760 to Clemens Stiglechner <a9401816@unet.univie.ac.at> for the
761 patch.
762
763 6. Add the -l option to restore to specify if, when doing a
764 remote restore, the file used to access the backup is a
765 regular file (the defaults being a tape drive). Restore needs
766 to know this information when reading a remote compressed
767 dump. Previously, this information was autodetected, but
768 the autodetection code fails (with ioctl: Inappropriate ioctl
769 for device) when using a non Linux remote box. Thanks to
770 many users and especially to Eros Albertazzi
771 <eros@lamel.bo.cnr.it> for reporting this.
772
773 7. Found a workaroung for the dump deadlock problem (3 childs
774 stuck in pause(), father in read()). The workaround seems
775 to work for me and several beta-testers. If it doesn't work
776 for you, please report back.
777
778 8. Updated the RPM spec file (BuildPrereq, URL etc).
779
780 Changes between versions 0.4b23 and 0.4b24 (released September 12, 2001)
781 ========================================================================
782
783 1. Fixed the permissions of a newly created QFA file by dump.
784
785 2. Cleaned up the source of dump (the external variables
786 definition was a complete mess, making possible to have
787 objects overlap).
788
789 3. Fixed restore to use the full tape volume path when doing
790 a compare (since it changes the working directory to the
791 filesystem being compared in the process).
792
793 4. Added the -q option to dump which makes dump abort
794 whenever it needs operator attention. It should be
795 easier to use dump in scripts with this option.
796
797 5. Detect the use of incompatible options to dump and
798 refuse them (like -a and -B options together).
799
800 6. Added bzip2 compression to dump/restore (use option -j level
801 to select it). Note that you will need a restore version
802 >= 0.4b24 in order to restore a bzip2 compressed dump.
803 The same warning as for the zlib compression applies:
804 the tape format of a bzip2 dump is not compatible with the
805 original BSD tape format.
806
807 7. Fixed a overflow problem in dump corrupting the dump when
808 very large files were encountered. Thanks to Vladimir Nadvornik
809 for the bug report.
810
811 8. Added a ioctl(BLKFLSBUF) in dump which should flush the
812 kernel buffer/page cache before starting the dump, helping
813 a bit those who use dump on mounted filesystems. Thanks to
814 John Yu <jky@it.bu.edu> and to Theodore T'so <tytso@mit.edu>
815 for this suggestion.
816
817 9. Updated the RPM spec file following the RedHat changes
818 (dynamically linked binaries now in /usr/sbin etc).
819
820 10. Added a patch from Helmut Jarausch <jarausch@igpm.rwth-aachen.de>
821 which enables restore to recognise multi volume compressed dumps
822 done on CD. Included his scripts for dump (which pipe the dump
823 output directly into cdrecord) and restore. There is now
824 possible to to backups to CD on the fly!
825
826 Changes between versions 0.4b22 and 0.4b23 (released July 20, 2001)
827 ===================================================================
828
829 1. Fixed a buffer overflow in restore/tape.c. Patch provided by
830 Marcus Meissner (Caldera International Security Dept.).
831
832 2. Implement the Sun rmt extended protocol. Patch submitted
833 by Ian Gordon <iangordon@users.sourceforge.net>.
834
835 3. Check for the e2fsprogs header <ext2fs/ext2_fs.h> instead of
836 the linux kernel header. This ensures that dump always has the
837 latest version of this file. Patch submitted by
838 Andreas Dilger <adilger@turbolinux.com>.
839
840 4. Report any filesystem present in either /etc/fstab with a
841 positive passno or /etc/dumpdates in dump -w output.
842 Patch submitted by Andreas Dilger <adilger@turbolinux.com>.
843
844 5. Fixed the looping problem in dump introduced in the
845 previous version.
846
847 6. Changed the -B option of dump to limit the size of
848 _compressed_ output per volume if compression is on.
849 Patch contributed by Helmut Jarausch
850 <jarausch@igpm.rwth-aachen.de>. Note however that, since
851 it is impossible to predict the size of the compressed
852 data before doing the compression, the -B limit is a bit
853 conservative.
854
855 7. Fixed a bug in reading the operator typed file/tape path for
856 the next volume in restore.
857
858 8. Implemented a "-F script" option to restore which permits the
859 user to specify a script which will be launched at the
860 beginning of each tape, useful for automatic programming of
861 tape changers for example. See the restore man page for the
862 script parameters and return codes.
863
864 9. Small fix for the QFA routines provided by Uwe Gohlke
865 <uwe@ugsoft.de>, and some recommendations for QFA uses in
866 the man pages.
867
868 10. Fixed the multivolume restoring where making a mistake
869 to the 'Mount next tape' prompt caused several blocks to
870 be lost.
871
872 11. Enhanced the -e option of dump to take as a parameter a
873 comma separated list of inode numbers.
874
875 12. Added the -E option to dump which specify a file containing
876 inode numbers to exclude from the dump.
877
878 13. Fixed the compressed multi-volume dump + restore.
879
880 Changes between versions 0.4b21 and 0.4b22 (released May 12, 2001)
881 ==================================================================
882
883 1. Made dump report the number of blocks dumped per volume.
884 Thanks to Kenneth Porter <shiva@well.com> for the suggestion.
885
886 2. Fix a bug in dump where waiting too much at the 'change volumes'
887 question caused the volume to be rewritten. Thanks to
888 Graham Reed <greed@users.sourceforge.net> for reporting the
889 bug and providing a patch.
890
891 3. Added a compression option to dump, contributed by Jerry
892 Peters <gapeters@worldnet.att.net>.
893
894 WARNING: the compression works only on files/pipes and
895 drives supporting variable block size.
896
897 WARNING2: tapes created using the compression feature are
898 incompatible with the standard dump tape format, and a
899 version of restore >= 0.4b22 will be needed for extraction.
900
901 4. Fixed some compilation problems for glibc 2.2.2 and 64 bit
902 architectures. Thanks to Paul Slootman <paul@debian.org> for
903 the patch and to Bdale Garbee <bdale@gag.com> for forwarding
904 it upstream.
905
906 5. Many cleanups (CPP defines, const char warnings, check of
907 ext2fs COMPAT flags, time_t cleanups, added libext2 version
908 in dump usage text) by Andreas Dilger <adilger@turbolinux.com>.
909
910 6. Made --prefix option work in configure. All the install path
911 are now based on the configure parameters.
912
913 7. Added the Quick File Access mode in dump/restore, contributed
914 by Uwe Gohlke <uwe@ugsoft.de>. In this mode, dump stores in
915 a file tape position for each inode, and this file is used by
916 restore (if called with parameter Q and the filename)
917 to directly position the tape at the file restore is currently
918 working on. This saves hours when restoring single files from
919 large backups, saves the tapes and the drive's head. Use
920 --enable-qfa option of configure to compile in the QFA support.
921
922 8. Added the possibility to dump several files and directories
923 in a single invocation of dump. Thanks to Uwe Gohlke
924 <uwe@ugsoft.de> for implementing this option.
925
926 9. Fixed the dumping and restoring of files with holes
927 (files ending with holes didn't get dumped or restored
928 correctly).
929
930 10. Fixed a socket descriptor leak in dump, which leaved opened
931 3 file descriptors per dump process (and there is one dump
932 process per tape).
933
934 11. Fixed dump large file system support, by explicit use of
935 open64/lseek64/etc functions (explicit use needed because
936 e2fsprogs libraries don't behave well when compiled with
937 FILE_OFFSET_BITS=64).
938
939 Changes between versions 0.4b20 and 0.4b21 (released January 13, 2001)
940 ======================================================================
941
942 1. Fixed some bugs in the dump -w|-W logic introduced by
943 the previous version. Thanks to Andreas Dilger
944 <adilger@turbolinux.com> for his help on this one.
945
946 2. Fixed again a compilation problem when using old e2fs
947 headers (filesystem label related). Thanks to many users
948 who reported this stupid error.
949
950 3. Fixed a build problem on old lib5 based systems dealing with
951 _PATH_MNTTAB being undefined. Thanks to John Adams
952 <johna@onevista.com> for reporting the bug.
953
954 4. Improved the error detection in restore when creating
955 the temporary files in TMPDIR. Now restore will corectly
956 report a 'no space left on device' error instead of
957 strange errors which could imply an archive corruption.
958 Thanks to Gabriel Somlo <somlo@cs.colostate.edu> and
959 bgjenero <bgjenero@sympatico.ca> for reporting the bug.
960
961 5. Added the throughput information in dump's progression
962 messages. Thanks to Andreas Dilger <adilger@turbolinux.com>
963 for the patch.
964
965 6. Use libext2fs's inode scanning routines, which are particularly
966 robust when dealing with errors (especially when having some
967 bad blocks on the filesystem). Many thanks to Colin
968 <colin@horizon.com> for reporting the bug, investigating
969 the issues, finding a workaround, writing the patch and
970 fully testing it... (of course, if this patch will break
971 anything, he is to blame for :-)).
972
973 7. Made dump and restore LFS aware. Dump can dump a filesystem
974 containing large files, generate a large file on output and
975 restore can restore them. This feature is not enabled by
976 default, use --enable-largefile option of configure to enable
977 it (you need to have a LFS aware glibc though). Thanks to
978 Andrea Arcangeli <andrea@suse.de> for submitting the patch,
979 and to Theodore T'so <tytso@valinux.com> for his always
980 useful thoughts.
981
982 8. Made dump ask upon a tape write error if it should rewrite
983 the current volume (assume this is a bad tape) or if it should
984 assume an end-of-tape condition (useful for tape drives which
985 behaves badly at the end of the tape). Thanks to Andreas
986 Dilger <adilger@turbolinux.com> for the suggestion.
987
988 Changes between versions 0.4b19 and 0.4b20 (released November 10, 2000)
989 =======================================================================
990
991 1. Fixed a small compilation problem due to a change
992 in the definintion of the struct sigaction in
993 glibc 2.0 / libc5. Thanks to Gunther Schlegel
994 <schlegel@riege.de> for reporting the bug and to
995 Dave Platt <dplatt@snulbug.mtview.ca.us> for suggesting
996 a fix.
997
998 2. Modified the label and uuid analysis in order to be
999 self-contained (independant of kernel/libc headers). This
1000 should fix the compile with older kernel/libc headers and
1001 will preserve the functionality. Thanks to Bernhard Erdmann
1002 <bernhard.erdmann@gmx.de> for reporting the bug.
1003
1004 3. The 'exclude inode' option, if applied to a directory
1005 inode, excludes now correctly all the leaves of this
1006 directory. Thanks to John R. Dennison
1007 <gerdesas@users.sourceforge.net> for reporting the bug.
1008
1009 4. Fixed the '-e' option to disable the possibility
1010 to exclude the root inode (which causes the dumps to
1011 be unreadable by restore). Prevented array overflow
1012 when multiple -e options are used.
1013
1014 5. Fixed dump to correctly interpret a filesystem argument
1015 which is a mountpoint and it is not an absolute path
1016 (as specified in the fstab). Thanks to Bernhard R. Erdmann
1017 <be@berdmann.de> for reporting the bug.
1018
1019 6. Made dump able to backup files larger than 2 GB. Note that
1020 dump still doesn't cope with files larger than 4 GB.
1021
1022 7. Restore the real uid and gid before invoking an external
1023 RSH program (big hole when dump or restore is suid root!).
1024
1025 8. Get the values for the default tape device and the location
1026 of fstab file from the system headers. Thanks to
1027 Andreas Dilger <adilger@turbolinux.com> for the patch.
1028
1029 9. Made dump -w|-W report all recognized filesystems
1030 present in either /etc/fstab or /etc/dumpdates, and present
1031 the list in the same order as in fstab file. Thanks
1032 to Andreas Dilger <adilger@turbolinux.com> for the patch.
1033
1034 10. Made dump's -a (automatic end of tape detection) the
1035 default. Specifying one of -B, -c, -d or -s options will
1036 override the EOM detection. Thanks to Andreas Dilger
1037 <adilger@turbolinux.com> for the patch.
1038
1039 11. Save the ext2 filesystem label into the dump volume label.
1040 Specifying a volume label on the command line (-L option)
1041 will override this feature. Thanks to Andreas Dilger
1042 <adilger@turbolinux.com> for the patch.
1043
1044 Changes between versions 0.4b18 and 0.4b19 (released August 20, 2000)
1045 =====================================================================
1046
1047 1. Fixed the signal handling in dump (which I broke in 0.4b17)
1048 which was causing several strange dump failures (dump
1049 hanged or segmentation faults etc).
1050
1051 2. Specified the default blocksize in the dump man page.
1052
1053 3. Changed two info messages of restore to be written on stdout
1054 instead of stderr in order to leave stderr only for errors
1055 or warnings. Thanks to Stephen Atwell
1056 <satwell@urbana.css.mot.com> for the suggestion.
1057
1058 4. Corrected an off by one calculation which prevented
1059 dumping a directory having a 1 character name length.
1060 Thanks to Bernhard Erdmann <bernhard.erdmann@gmx.de>
1061 for reporting the bug.
1062
1063 5. Reinforce dump in order to allow only level 0 and no
1064 -u option when dumping a subdirectory, and document
1065 this in the man page. Thanks to Bernhard Erdmann
1066 <bernhard.erdmann@gmx.de> for reporting the bug.
1067
1068 6. Fixed a small harmless bug in restore which caused
1069 in some conditions hard links to be restored several
1070 times (therefore generation some warning messages).
1071 Thanks to Kenneth Porter <shiva@well.com> for
1072 reporting the bug.
1073
1074 7. Updated the RPM spec file to the latest RedHat version,
1075 providing FHS packaging and other cosmetic changes.
1076 You will need RPM version >= 3.0.5 in order to build the RPM.
1077
1078 8. Updated the configure script to check for libtermcap
1079 before checking for libreadline, because we need this
1080 library in order to compile the readline support.
1081
1082 9. Made dump understand the LABEL= and UUID= notation
1083 both in /etc/fstab and on the command line. Note that
1084 you will need the /proc filesystem in order to use
1085 these notations. Thanks to Erik Troan <ewt@redhat.com>
1086 for providing the patch.
1087
1088 Changes between versions 0.4b17 and 0.4b18 (released June 30, 2000)
1089 ===================================================================
1090
1091 1. Fixed a potential buffer overflow in restore. Thanks
1092 to Stan Bubrouski <satan@fastdial.net> for reporting
1093 the bug.
1094
1095 2. Fixed a readline-related bug which prevented
1096 'cat DUMPFILE | restore -i -f -' from working. Thanks
1097 to Charles Karney <karney@users.sourceforge.net>
1098 for the bug report.
1099
1100 3. Changed a few "panic" into "exit", causing restore to
1101 be more stable against some attacks (like the last one
1102 reported on Bugtraq, although the last version of restore
1103 was not vulnerable - just dumped core). Thanks to
1104 Andreas Hasenack <andreas@conectiva.com.br> for reporting
1105 the bugs to me.
1106
1107 4. Removed the suid-root bit on dump and restore in the
1108 default build (and generated RPMs). It should be safer
1109 now. Users who need the suid-root capabilities in order
1110 to make network backups should read first the man page
1111 of dump and enable the suid bit manually.
1112
1113 5. Added -ltermcap to the compile parameters for restore
1114 when using readline support, in order to make the compile
1115 process work on some systems (SuSE ?). Thanks to
1116 Patrik Schindler <poc@pocnet.net> for reporting the bug.
1117
1118 Changes between versions 0.4b16 and 0.4b17 (released June 1st, 2000)
1119 ====================================================================
1120
1121 1. The -F script is called now *only* at the end of a tape,
1122 not at the end of the dump. Thanks to Christian Haul
1123 <haul@informatik.tu-darmstadt.de> for the bug report.
1124
1125 Normally, the device name and the current volume number
1126 are passed on the command line of the script. However,
1127 if you want the old style script (with no arguments
1128 passed on the command line), you can enable it in
1129 configure with the --enable-oldstylefscript.
1130
1131 2. Use posix signal handling to preserve dump functionality
1132 with libc5. Thanks to Jeff Johnson <jbj@redhat.com> for
1133 the patch.
1134
1135 3. Made the exit code of restore in case of a 'C'ompare
1136 command reflect the result of the compare. An exit status
1137 of 0 means the dump archive is correct, 1 means tape errors,
1138 2 means that some files were modified. Thanks to Kenneth Porter
1139 <shiva@well.com> for the suggestion.
1140
1141 4. Made (finally) quotes work in interactive restore.
1142
1143 5. Small fixes in order to allow dump to compile with a
1144 really old e2fsprogs version. Thanks to Ian Zimmerman
1145 <itz@speakeasy.org> for the bug report.
1146
1147 6. Add GNU readline capabilities to interactive restore.
1148 Use configure's --enable-readline flag to enable this feature.
1149 Thanks to Patrik Schindler <poc@pocnet.net> for the
1150 suggestion.
1151
1152 7. Do the compare on the fly in restore 'C' mode (this will
1153 allow not to exhaust the available /tmp space when
1154 ccmparing large files). Thanks to Kenneth Porter
1155 <shiva@well.com> for the suggestion.
1156
1157 Changes between versions 0.4b15 and 0.4b16 (released March 11, 2000)
1158 ====================================================================
1159
1160 1. Fixed some several duplicate 'DUMP: DUMP:' in the
1161 output of dump.
1162
1163 2. Corrected the estimation of blocks to dump. Note that
1164 this is still wrong for large directory inodes, due
1165 to the size of a BSD directory entry on the tape
1166 (estimation is based on the size of the ext2 directory,
1167 which is smaller because it doesn't contain the null
1168 character at the end).
1169
1170 3. Made dump report the total number of MB written to
1171 tape at the end of run. Thanks to W. Reilly Cooley
1172 <wcooley@nakedape.cc> for the patch.
1173
1174 4. Added the -X option to restore giving the possibility
1175 to read the names of the files to be extracted/listed
1176 from a text file (in addition of the command line).
1177 Thanks to Dejan Muhamedagic <dejan@quant-x.com> for the
1178 patch.
1179
1180 5. Added the device name and the current volume number
1181 as arguments to the end of tape script (dump -F option).
1182
1183 6. Made the multi-volume dump work again (after having
1184 broken it in 0.4b15).
1185
1186 Changes between versions 0.4b14 and 0.4b15 (released March 2, 2000)
1187 ===================================================================
1188
1189 1. Added a prompt command in interactive restore mode. Thanks
1190 to Andreas Dilger <adilger@home.com> for the patch.
1191
1192 2. Fixed a buffer overflow problem in dump (caused by
1193 not checking the size of the filesystem parameter).
1194 Thanks to Kim Yong-jun <loveyou@hackerslab.org> for
1195 reporting this on Bugtraq (and to several dump users
1196 who forwarded me his mail).
1197
1198 3. Added the '-F script' option to dump in order to
1199 launch a script at the end of each tape (to be used
1200 with a tape changer, or to notify the sysadmin by
1201 pager etc.).
1202
1203 4. Fixed a bug in restore compare code caused by the changes
1204 I made in 0.4b14.
1205
1206 5. Fixed the treatment of options using the old BSD syntax
1207 in both dump and restore.
1208
1209 Changes between versions 0.4b13 and 0.4b14 (released February 10, 2000)
1210 =======================================================================
1211
1212 1. Fixed a bug in dump which may have caused invalid deleted
1213 directories to be dumped out if they were deleted after the
1214 mapping pass. This could occure on active filesystem and lead
1215 to heap corruption (causing dump malfunction in many possible ways).
1216 Thanks to Jan Sanislo <oystr@cs.washington.edu> for finding this
1217 bug and submitting the patch.
1218
1219 2. Fixed the handling of the filesystem parameter in dump. A
1220 '/mnt/disk' parameter caused the disk contents to be dumped,
1221 but a '/mnt/disk/' parameter caused the mountpoint directory
1222 to be dumped (generally an empty directory).
1223
1224 3. Improved the output of dump in order to tell which directory
1225 it is currently dumping (when dumping a subtree).
1226
1227 4. Added the '-e' exclude inode option to dump. Thanks to
1228 Isaac Chuang <ike@isl.stanford.edu> for contributing with the patch.
1229
1230 5. Added a REPORTING-BUGS file in order to provide a guide
1231 on how to correctly report a bug in dump/restore.
1232
1233 6. Fixed a restore problem when restoring a hard link to an inode
1234 having the immutable or append-only attribute set. Thanks to
1235 Ambrose Li <acli@mingpaoxpress.com> for submitting the patch.
1236
1237 7. Fixed a compatibility problem between dumps made on little
1238 endian machines (the format was correct) and big endian
1239 machines (the format was incorrect). This fix break the
1240 compatibility with the older dumps made on big endian
1241 machines (sparc, m86k, ppc etc). For the first time in
1242 linux dump's history (I believe), the dumps made by *BSD,
1243 Linux/alpha, Linux/sparc and Linux/x86 are compatible,
1244 so interchangeable. Thanks to Rob Cermak
1245 <cermak@ahab.rutgers.edu> for submitting the bug and
1246 helping me test the fix.
1247
1248 8. Fixed the way dump reports the remaining percent/time, if
1249 the number of blocks actually dumped exceeds the estimated
1250 number of blocks. Thanks to Jean-Paul van der Jagt
1251 <jeanpaul@dutepp0.et.tudelft.nl> for reporting the bug.
1252
1253 Changes between versions 0.4b12 and 0.4b13 (released January 21, 2000)
1254 ======================================================================
1255
1256 1. Small Y2K fix in the man pages :). Thanks to Bernhard Sadlowski
1257 <sadlowsk@Mathematik.Uni-Bielefeld.DE> for reporting the bug.
1258
1259 2. Removed the requirement to build the RPM as root from the
1260 spec file. Thanks to Christian Weisgerber
1261 <naddy@mips.rhein-neckar.de> for submitting this.
1262
1263 3. Fixed a bug in dump related to the 'filetype' feature of ext2fs,
1264 causing dump to block when dumping really huge filesystems.
1265 Many thanks to Patrik Schindler <poc@pocnet.net> for
1266 helping me find this bug.
1267
1268 4. Fixed the treatment for an interrupt signal when dump access
1269 the remote tape through RSH. Thanks to Christian Weisgerber
1270 <naddy@mips.rhein-neckar.de> for providing the patch.
1271
1272 5. Fixed a bug which was causing dump/restore to display
1273 garbage characters instead of the remote host name.
1274
1275 Changes between versions 0.4b11 and 0.4b12 (released January 8, 2000)
1276 =====================================================================
1277
1278 1. Small fix in the dump man page. Thanks to Thorsten Kukuk
1279 <kukuk@suse.de> for submitting the patch.
1280
1281 2. Fix for the exit code when using the size estimate option of
1282 dump. Thanks to Matti Taskinen <mkt@rni.helsinki.fi> for
1283 submitting the patch.
1284
1285 3. Handle EINTR in atomical reads/writes in dump, which was causing
1286 dump to fail on some systems. Thanks to Eric Jergensen
1287 <eric@dvns.com> for reporting the bug and submitting the patch.
1288
1289 4. Handle more than 16 characters for the device names in dumpdates.
1290 (up to 255 now). Thanks to Rainer Clasen <bj@ncc.cicely.de> for
1291 tracking down the problem and proposing the solution.
1292
1293 5. Fixed a bug in dump which prevented the creation of the
1294 dumpdates file when doing a 0-level dump without already
1295 having a dumpdates file. Thanks to Patrik Schindler
1296 <poc@pocnet.net> for reporting the bug.
1297
1298 6. Changed the way dump 'S' flag reports the size estimate
1299 from number of blocks into bytes (making it compatible
1300 with the Solaris version, and simplifying things for
1301 amanda users). Thanks to Jason L Tibbitts III
1302 <tibbs@math.uh.edu> for reporting the bug.
1303
1304 7. Fixed a compatibility problem in linux/alpha dump tape format.
1305 Now the linux/alpha dump are (again) compatible with the
1306 other dump formats. But this breaks compatibility with
1307 older dumps made on alpha. Thanks to Mike Tibor
1308 <tibor@lib.uaa.alaska.edu> for helping me in finding this bug.
1309
1310 Changes between versions 0.4b10 and 0.4b11 (released December 5, 1999)
1311 ======================================================================
1312
1313 1. Added a '--enable-kerberos' to configure.
1314
1315 2. Added a 'S' option to dump which determines the amount of space
1316 that is needed to perform the dump without actually doing it, similar
1317 to the Sun's ufsdump 'S' option. Patch contributed by Rob Cermak
1318 <cermak@ahab.rutgers.edu>.
1319
1320 3. Added a 'M' multi-volume option to dump and restore which enables
1321 dumping to multiple files (useful when dumping to an ext2
1322 partition to make several dump files in order to bypass the 2GB
1323 file size limitation). The 'f' argument is treated as a prefix and
1324 the output files will be named <prefix>001, <prefix>002 etc. With
1325 the 'M' flag, restore automatically selects the right file without
1326 asking to enter a new tape each time.
1327
1328 4. Fixed a memory leak which was causing dump to grow very big
1329 (270MB when dumping a 10GB filesystem...). Thanks to Jason
1330 Fearon <jasonf@netrider.org.au> for reporting the bug.
1331
1332 Changes between versions 0.4b9 and 0.4b10 (released November 21, 1999)
1333 ======================================================================
1334
1335 1. Make configure test if the system glob routines support
1336 extended syntax (ALTDIRFUNC). If not, use the internal glob
1337 routines instead of system ones. Thanks to Bernhard Sadlowski
1338 <sadlowsk@Mathematik.Uni-Bielefeld.DE> for reporting the bug
1339 and helping me resolve this and other minor libc5 compiling
1340 glitches.
1341
1342 2. Fix a problem when dumping a ext2fs with the 'filetype'
1343 feature enabled. Thanks to Patrick J. LoPresti
1344 <patl@cag.lcs.mit.edu> for reporting the bug and to
1345 Theodore Y. Ts'o <tytso@mit.edu> for providing the patch.
1346
1347 3. Made the nodump flag work on directories. A directory which
1348 has the nodump flag gets never dumped, regardless of its
1349 contents.
1350
1351 4. Integrate a patch from Jeremy Fitzhardinge <jeremy@goop.org>
1352 which allows dump on an active ext3 filesystem. However, this
1353 is a "quick and dirty" patch which enables backup of an ext3
1354 filesystem through the ext2 compatibility (by ignoring the
1355 NEEDS_RECOVERY bit). The journal file is not recognized and
1356 it is dumped (it should not).
1357
1358 5. Test the superblock compatibility flags when dumping, in order
1359 to be sure that we know how to deal with specific features.
1360
1361 Changes between versions 0.4b8 and 0.4b9 (released November 5, 1999)
1362 ====================================================================
1363
1364 1. Use lchown instead of chown, fixing a possible security problem
1365 when restoring symlinks (a malicious user could use this
1366 to deliberately corrupt the ownership of important system files).
1367 Thanks to Chris Siebenmann <cks@utcc.utoronto.ca> for detecting
1368 this and providing the patch.
1369
1370 Changes between versions 0.4b7 and 0.4b8 (released November 3, 1999)
1371 ====================================================================
1372
1373 1. Put dump sources under CVS, added Id tags in all files so
1374 one can use 'ident' on binary files.
1375
1376 2. Added the dump/restore version in the usage text so one can
1377 easily verify the version he is using.
1378
1379 3. Small patch from Nuno Oliveira <nuno@eq.uc.pt> which fixes
1380 a va_start/va_end problem on linux-ppc (always call va_start
1381 va_end in pairs each time we use a vararg function).
1382
1383 4. Added again the DT_* constants because old libc does not
1384 contain them :(. Thanks to Eric Maisonobe <virnet@nat.fr>
1385 for submitting the bug report.
1386
1387 5. Use ext2fs_llseek instead of llseek. With recent e2fsprogs
1388 this should enable dumping big (huge) filesystems.
1389
1390 6. Added the RSH environment variable in order to be able to
1391 use a rsh replacement like ssh when doing remote backups (and
1392 bypass the security limitations of rcmd). Now you can do remote
1393 backups without being root (or making dump setuid root).
1394
1395 7. Modified again the way dumpdates works. For incremental dumps,
1396 we need to read dumpdates even if we are not using 'u' option.
1397 Thanks to Bdale Garbee <bdale@gag.com> for his ideas on how
1398 this should work.
1399
1400 Changes between versions 0.4b6 and 0.4b7 (released October 8, 1999)
1401 ===================================================================
1402
1403 1. Removed the 'k' flag from the restore 'about' text if kerberos
1404 was not compiled in.
1405
1406 2. Prototyped (f)setflags from e2fsprogs and corrected the calls
1407 to them (fsetflags takes a char*, setflags an open fd!).
1408
1409 3. (f)setflags is called only if the flags aren't empty. If the
1410 file is a special file, a warning is printed, because changing
1411 flags implies opening the device. Normally, a special file
1412 should not have any flag... (Debian bug #29775, patch provided
1413 by Abhijit Dasgupta <abhijit@ans.net>).
1414
1415 4. Made possible to dump a file system not mentioned in /etc/fstab.
1416 (Debian bug #11904, patch provided by Eirik Fuller <eirik@netcom.com>).
1417
1418 5. Changed the default behaviour to not create dumpdates
1419 unless 'u' option is specified. Removed the old "debian-patch"
1420 which provided the same thing. (Debian bug #38136, #33818).
1421
1422 6. Removed all those dump*announce, since they were getting old...
1423
1424 7. Added warning messages if dumpdates does not exist and
1425 when an inferior level dump does not exist (except for a level 0
1426 dump).
1427
1428 8. Debugged the glob calls in interactive mode: restore used a
1429 dirent struct which was different from the /usr/include/dirent.h
1430 one (this used to work, is it a glibc2 change?), so none of the
1431 compat glob (which used /usr/include/dirent.h) or the system glob
1432 worked. Restore use now the system dirent (and the system
1433 DT_* constants), which are compatible with BSD ones.
1434
1435 9. Added a configure flag (--with-dumpdatespath) to specify
1436 the location of dumpdates file. By default, it is
1437 /etc/dumpdates.
1438
1439 10. Added the "AUTHOR" and "AVAILABILITY" sections and
1440 included the current date/version in man pages.
1441
1442 11. Corrected the estimation of remaining time when
1443 the operator doesn't change the tapes quickly enough. This
1444 was an old bug, I thought I corrected it, and discovered
1445 that in fact it was corrected in two different places, so
1446 the results canceled each other...
1447
1448 Changes between versions 0.4b5 and 0.4b6 (released October 1, 1999)
1449 ===================================================================
1450
1451 1. Integrated multiple patches from RedHat, Debian and SuSE:
1452
1453 - tweak dump/itime.c to not try to read dumpdates if the 'u' option
1454 isn't specified.
1455 - several fixes in the man pages.
1456 - update the default tape device to /dev/st0.
1457 - many updates for Linux Alpha (byte ordering, size_t etc).
1458 - buffer overruns.
1459 - use environment variable for TMPDIR (instead of /tmp).
1460 - use sigjmp_buf instead of jmp_buf (RedHat bug #3260).
1461 - workaround egcs bug (RedHat bugs #4281 and #2989).
1462 - wire $(OPT) throughout Makefile's.
1463
1464 2. Upgrade the dump revision to 1, making possible to dump filesystems
1465 made with e2fsprogs-1.15 or newer. Nothing seems to break...
1466
1467 3. Fix some compile warnings, prototype all functions.
1468
1469 4. Use glibc err/glob instead of internal compatibility
1470 routines (only if available).
1471
1472 5. Fix a compile error on Linux 2.2.7 / libc5 (5.4.44) (patch provided
1473 by Bernhard Sadlowski <sadlowsk@mathematik.uni-bielefeld.de>).
1474
1475 Changes between versions 0.4b4 and 0.4b5 (released September 22, 1999)
1476 ======================================================================
1477
1478 1. Integrated the changes from FreeBSD-3.1-RELEASE
1479 (mostly bug fixes, buffer overruns, dump has now an "automatic
1480 tape length calculation" flag, dump/restore can use kerberos now
1481 (this is NOT tested), use environment variables for TAPE and
1482 RMT etc.).
1483
1484 2. Integrated three RedHat patches ("glibc", "kernel" and "bread" patches)
1485
1486 3. Corrected a bug in restore when using 'C' option with multi-volumes
1487 tapes (files splited accros two tapes give "size changed" errors
1488 when comparing).
1489
1490 4. Corrected the long standing bug when dumping multiple tapes.
1491 This works for me, needs further testing.
1492
1493 Changes between versions 0.4b3 and 0.4b4 (released January 17, 1997)
1494 ====================================================================
1495
1496 1. Dump now runs correctly on kernels 2.1.x
1497 Fix made by Gerald Peters <gapeters@worldnet.att.net>
1498
1499 Changes between versions 0.4b2 and 0.4b3
1500 ========================================
1501
1502 1. Use realpath() if available
1503
1504 2. Report statistics
1505
1506 Changes between versions 0.4b1 and 0.4b2
1507 ========================================
1508
1509 1. Fixed the bug fix from Greg Lutz (I had made a mistake when integrating
1510 the patch)
1511
1512 2. Fixed restore to make it able to read FreeBSD 2.x dumps again
1513
1514 3. Fixed configure.in to correctly handle --enable-rmt
1515
1516 Changes between versions 0.3 and 0.4b1
1517 ======================================
1518
1519 1. Integrated the changes from 4.4BSD-Lite2
1520
1521 2. Integrated the patches from Debian and Red Hat
1522
1523 3. Portability changes: use the __u32, __u16, __s32, and __s16 types
1524
1525 4. Changed dump to use the Ext2fs library to get block addresses. This
1526 should solve the endianness problem on SparcLinux.
1527
1528 5. Created a configure.in file (shamelessly stolen from the e2fsprogs
1529 distribution's one) to use autoconf
1530
1531 6. Fixed a few minor bugs
1532
1533 Changes between versions 0.2e and 0.2f
1534 ======================================
1535
1536 1. Added the creation of named pipes (fifos) in restore.
1537
1538 2. Added the -N flag in the restore manual page.
1539
1540 3. Added the file kernel.patch which contains the llseek() optimization
1541 patch for 1.2.x kernels.
1542
1543 4. Fixed a bug in the restoration of symbolic links: owner and group were
1544 not restored.
1545
1546 5. Integrated some changes from FreeBSD 2.2.
1547
1548 6. Added a call to ftruncate() after restoring each file to restore
1549 correctly files ending by a hole.
1550
1551 Changes between versions 0.2d and 0.2e
1552 ======================================
1553
1554 1. Fixed a bug in the "set owner/mode" process. Every file was restored
1555 with owner = root (0) and group = root/wheel/whatever (0).
1556
1557 Changes between versions 0.2c and 0.2d
1558 ======================================
1559
1560 1. Dump is now able to backup 2GB+ filesystems.
1561
1562 2. Dump and restore can now be linked as static binaries.
1563
1564 Changes between versions 0.2b and 0.2c
1565 ======================================
1566
1567 1. Fixed a bug when dumping ``slow'' (i.e. normal) symbolic links.
1568
1569 Changes between versions 0.2a and 0.2b
1570 ======================================
1571
1572 1. Really fixed the bug that I should have corrected in 0.2a.
1573
1574 2. Enabled optimization again.
1575
1576 Changes between versions 0.2 and 0.2a
1577 =====================================
1578
1579 1. Disabled the optimization during compilation.
1580
1581 Changes between versions 0.1 and 0.2
1582 ====================================
1583
1584 1. Fixed a bug in fstab.c which caused a null pointer to be stored in
1585 the fs_type field (actually, I modified the file fstab.c to make it
1586 use the mntent functions).
1587
1588 2. Dump and restore now use a 4.3 BSD compatible dump format. Backups
1589 made by dump should be readable by the BSD restore and backups made
1590 by the BSD dump should be readable by restore. Unfortunately, this
1591 means that the dump format has changed between version 0.1 and version
1592 0.2 :-(
1593
1594 3. Dump is now able to backup a subtree, it is no longer limited to whole
1595 filesystems like the BSD version.
1596
1597 4. Dump now uses ext2_llseek() so it is able to backup filesystems bigger
1598 than 2 GB.
1599
1600 Changes between versions 0.0 and 0.1
1601 ====================================
1602
1603 1. Now create links rdump and rrestore during the `make install' step.
1604
1605 2. Linux port specific bugs added to the manual pages
1606
1607 3. Incorrect estimation of the number of tapes blocks fixed when doing
1608 incremental backups.
1609
1610 4. Better ls-like format in restore in interactive mode.