]> git.wh0rd.org - dump.git/blob - CHANGES
Man pages rewrite.
[dump.git] / CHANGES
1 $Id: CHANGES,v 1.188 2002/07/24 14:11:59 stelian Exp $
2
3 Changes between versions 0.4b29 and 0.4b30 (released ????????????????)
4 ======================================================================
5
6 1. Made dump print out the ext2 logical block number in case of a read
7 error, which can be given as an argument to debugfs' ncheck command,
8 in order to find out the affected inode. Added note about this
9 usage in dump's man page.
10
11 2. Fixed a problem in restore when reading tapes written on big
12 endian machines with very old versions of dump. The patch was
13 contributed by George Helffrich <george@geology.bristol.ac.uk>.
14
15 3. Fixed the tape length calculation when using large tapes
16 and compression. Thanks to Georg Lippold
17 <g_lippold@sourceforge.net> for reporting the bug.
18
19 4. Added a new set of examples in dump_on_cd_2 directory, based
20 on dump_on_cd examples but somewhat enhanced, supporting DVD
21 media, and localized in english and german. Thanks to
22 Georg Lippold <g_lippold@sourceforge.net> for the new scripts.
23
24 5. Made dump save 32 bit UID/GID. Internally, this was achieved
25 by switching from the old BSD inode format to the new BSD
26 inode format, which means that the tape format was changed.
27 However, since all restore versions out there should
28 transparently support both inode formats, the change should
29 have no side effects. Thanks to John Yu <jky@cs.bu.edu> for
30 reporting the bug.
31
32 6. Fixed a lot of warnings in the code shown when compiling
33 with 'gcc -W'. Thanks to Matthias Andree
34 <matthias.andree@stud.uni-dortmund.de> for reporting this.
35
36 7. Fixed a small markup bug in the dump man page. Thanks to
37 Eric S. Raymond <esr@minx.thyrsus.com> for submitting the
38 patch.
39
40 8. Rewrote entirely the man pages using the tmac.an macro
41 package (Linux man page format) instead of the original BSD
42 format. They should be now cleaner and easier to modify.
43
44 Changes between versions 0.4b28 and 0.4b29 (released June 8, 2002)
45 ==================================================================
46
47 1. Fixed a problem in the rmt ioctl command, where ioctl's issued from
48 non Linux clients were misinterpreted. The description of the problem
49 (incompatible numbering in Linux mtio opcodes) is documented at
50 ftp://ftp.fokus.gmd.de/pub/unix/star/README.mtio . Thanks to
51 Jörg Schilling <schilling@fokus.gmd.de> for reporting this bug and
52 providing an excellent, cross-platform replacement for rmt in his
53 star package.
54
55 2. Fixed a bug reported by Andy Mentges <amentges@jumpline.com>
56 which caused restore to fail when the inode maps were not entirely
57 contained on the first volume (when using really small volumes or
58 when dumping a huge number of inodes).
59
60 3. Fixed a problem in dump, where files in subdirectories of directories
61 that have the nodump flag set (or being excluded with -e/-E) were
62 sometimes incorrectly dumped (instead of being excluded). The original
63 fix is from David C Lawrence <tale@dd.org> in the FreeBSD version
64 of dump: http://www.FreeBSD.org/cgi/query-pr.cgi?pr=32414
65 Thanks to Ted Grzesik <tedgyz@roostme.com> for reporting the bug and
66 help testing the patch.
67
68 4. Added some example scripts from Gerd Bavendiek <bav@epost.de>
69 which makes one able to pipe the output of dump, by the net, to
70 a remote CD-burner server.
71
72 5. Made dump use O_CREAT|O_TRUNC both locally and remotely (over rmt),
73 and use GNU's symbolic syntax over rmt instead of numerical values
74 to assure multiple platform compatibility.
75
76 6. Documented the -d option in restore.
77
78 7. Added a -v (verbose) mode to dump. For now it just prints the number
79 of the inode being dumped, but this could evolve in future versions
80 to include interesting debugging output.
81
82 8. Added a -o flag to restore, which automatically restores the current
83 directory permissions in -i or -x mode, without asking the operator
84 whether to do so. Patch submitted by Tony Clayton <tonyc@e-smith.com>
85 and Peter Samuel <peters@e-smith.com>.
86
87 Changes between versions 0.4b27 and 0.4b28 (released April 12, 2002)
88 ====================================================================
89
90 1. Fixed a bug in the QFA feature which could have caused
91 a bad tape positionning on restore (causing extra delays in
92 restore). Credits go to Uwe Gohlke <uwe@ugsoft.de>.
93
94 2. Added a small note in the dump man page specifying that
95 there should be no white space between the option letter and
96 the -j or -z parameter, thanks to Kenneth Porter <shiva@well.com>
97
98 3. Made restore work with older versions of the readline library,
99 by checking in configure for several extended symbols. Restore
100 can now be compiled with a readline as old as the 2.0 release
101 (though it may be a good idea to upgrade it to a more recent
102 version...). Thanks to Andrew Donkin <ard@waikato.ac.nz> for
103 reporting the build failures.
104
105 4. Fixed a performance problem with the QFA file creation in
106 dump, which made unnecessary seeks on the tape slowing down
107 the dump. Thanks to Andrew Donkin <ard@waikato.ac.nz> for
108 reporting this issue.
109
110 5. Removed the inclusion of some kernel headers in the dump
111 source, which prevented the compile in some kernel/glibc
112 headers/architecture combination. Thanks to Bdale Garbee
113 <bdale@gag.com> for reporting the bug.
114
115 6. Added the appropriate error message when dump fails to
116 open the output file for writing. Thanks to Amith Varghese
117 <amithv@yahoo.com> for reporting this bug.
118
119 7. Made restore able to understand large Solaris ufsdump tapes
120 (containing inodes bigger than 4194304). Sun have introduced
121 an "extension" to the dump tape format when dealing with
122 those inodes, which was uncorrectly handled by Linux restore.
123 Thanks to Uwe Gohlke <uwe@ugsoft.de> for reporting the bug and
124 providing a test case.
125
126 8. Added the -m parameter to dump which optimises the output for
127 inodes having been changed but not modified since the last dump
128 ('changed' and 'modified' have the meaning defined in stat(2)).
129 For those inodes, dump will save only the metadata, instead of
130 saving the entire inode contents. Inodes which are either
131 directories or have been modified since the last dump are saved
132 in a regular way. Uses of this flag must be consistent, meaning
133 that either every dump in an incremental dump set have the flag,
134 or no one has it.
135
136 Changes between versions 0.4b26 and 0.4b27 (released February 15, 2002)
137 =======================================================================
138
139 1. Fixed behaviour of dump when exceeding resource limits
140 (SIGXFSZ treatment).
141
142 2. Added the -L flag to restore to allow the user to specify a
143 maximal allowed number of miscompares when using restore
144 with the -C option to check the backup.
145
146 3. Detailed the manual entry for the -N option of restore.
147
148 4. Added the -a flag to restore to make able doing unattended
149 restores in -i or -x mode (automatically walks through the
150 multiple dump volumes).
151
152 5. Extended the QFA mode to work with local files and/or
153 remote tapes and files. This way, restore can know in advance
154 the tape number and the offset for the inodes to extract and
155 can minimize the extraction time by seeking directly to the
156 good tape/offset.
157
158 6. Added the -A <archive> option to both dump and restore,
159 which makes dump to archive a dump table-of-contents in
160 the specified file to be used by restore to determine
161 whether a file is in the dump file that is being restored.
162 (the archive file syntax is also compatible with the
163 Solaris ufsdump generated one).
164
165 7. Small fix in restore making it able to read some (broken ?)
166 Solaris ufsdump tapes.
167
168 8. Fixed dump to correctly recognise the root filesystem when using
169 ext2 disk labels (LABEL=/). Thanks to John Yu <jky@it.bu.edu>
170 for reporting this bug.
171
172 9. Added the -P <file> option to restore to create a
173 Quick File Access file from an already made dump. Patch
174 contributed by Uwe Gohlke <uwe@ugsoft.de>.
175
176 10. Made restore compile and run on Solaris, making it a
177 possible replacement for the standard ufsrestore. Port was
178 contributed by Uwe Gohlke <uwe@ugsoft.de>.
179
180 Changes between versions 0.4b25 and 0.4b26 (released January 7, 2002)
181 =====================================================================
182
183 1. Added a set of backup scripts from Eugenio Diaz
184 <getnito@yahoo.com> in the examples section. It features
185 automatic (cron based) full and incremental dumping of
186 several filesystems on a separate filesystem.
187
188 2. Fixed a off-by-one miscalculation which disabled dumping
189 a one letter subdirectory. Thanks to Chris Siebenmann
190 <cks@utcc.utoronto.ca> for reporting the bug.
191
192 3. Fixed several restore bugs occuring when trying to
193 restore some missing files on the tape. Thanks to Chris
194 Siebenmann <cks@utcc.utoronto.ca> for reporting the bug.
195
196 4. Fixed --with-ldopts configure argument passing, installing from
197 a separate object directory, makefile cleanups contributed
198 by <splite@purdue.edu>.
199
200 5. Fix a bug which could caused, in some conditions, the highest
201 number inode of a filesystem, to not be dumped. Many thanks
202 to Chris Siebenmann <cks@utcc.utoronto.ca> for helping me
203 chase this bug.
204
205 Changes between versions 0.4b24 and 0.4b25 (released November 17, 2001)
206 =======================================================================
207
208 1. Added a mini howto from Patrick Walsh in the examples
209 directory.
210
211 2. Minor man pages syntax corrections. Thanks to
212 Chris Danis <screechco@home.com> for reporting the bugs.
213
214 3. Added a script from David B. Peterson <dave@toppledwagon.com>
215 to the examples section. It features dumping several
216 filesystems to a remote tape drive upon ssh.
217
218 4. Added a patch provided by Richard Jones <rich@annexia.org>
219 which allows BRADEMAX (number of read errors tolerated by
220 dump) to be adjusted using the -I option.
221
222 5. Fixed a bug which disabled doing "restore -C -f -". Thanks
223 to Clemens Stiglechner <a9401816@unet.univie.ac.at> for the
224 patch.
225
226 6. Add the -l option to restore to specify if, when doing a
227 remote restore, the file used to access the backup is a
228 regular file (the defaults being a tape drive). Restore needs
229 to know this information when reading a remote compressed
230 dump. Previously, this information was autodetected, but
231 the autodetection code fails (with ioctl: Inappropriate ioctl
232 for device) when using a non Linux remote box. Thanks to
233 many users and especially to Eros Albertazzi
234 <eros@lamel.bo.cnr.it> for reporting this.
235
236 7. Found a workaroung for the dump deadlock problem (3 childs
237 stuck in pause(), father in read()). The workaround seems
238 to work for me and several beta-testers. If it doesn't work
239 for you, please report back.
240
241 8. Updated the RPM spec file (BuildPrereq, URL etc).
242
243 Changes between versions 0.4b23 and 0.4b24 (released September 12, 2001)
244 ========================================================================
245
246 1. Fixed the permissions of a newly created QFA file by dump.
247
248 2. Cleaned up the source of dump (the external variables
249 definition was a complete mess, making possible to have
250 objects overlap).
251
252 3. Fixed restore to use the full tape volume path when doing
253 a compare (since it changes the working directory to the
254 filesystem being compared in the process).
255
256 4. Added the -q option to dump which makes dump abort
257 whenever it needs operator attention. It should be
258 easier to use dump in scripts with this option.
259
260 5. Detect the use of incompatible options to dump and
261 refuse them (like -a and -B options together).
262
263 6. Added bzip2 compression to dump/restore (use option -j level
264 to select it). Note that you will need a restore version
265 >= 0.4b24 in order to restore a bzip2 compressed dump.
266 The same warning as for the zlib compression applies:
267 the tape format of a bzip2 dump is not compatible with the
268 original BSD tape format.
269
270 7. Fixed a overflow problem in dump corrupting the dump when
271 very large files were encountered. Thanks to Vladimir Nadvornik
272 for the bug report.
273
274 8. Added a ioctl(BLKFLSBUF) in dump which should flush the
275 kernel buffer/page cache before starting the dump, helping
276 a bit those who use dump on mounted filesystems. Thanks to
277 John Yu <jky@it.bu.edu> and to Theodore T'so <tytso@mit.edu>
278 for this suggestion.
279
280 9. Updated the RPM spec file following the RedHat changes
281 (dynamically linked binaries now in /usr/sbin etc).
282
283 10. Added a patch from Helmut Jarausch <jarausch@igpm.rwth-aachen.de>
284 which enables restore to recognise multi volume compressed dumps
285 done on CD. Included his scripts for dump (which pipe the dump
286 output directly into cdrecord) and restore. There is now
287 possible to to backups to CD on the fly!
288
289 Changes between versions 0.4b22 and 0.4b23 (released July 20, 2001)
290 ===================================================================
291
292 1. Fixed a buffer overflow in restore/tape.c. Patch provided by
293 Marcus Meissner (Caldera International Security Dept.).
294
295 2. Implement the Sun rmt extended protocol. Patch submitted
296 by Ian Gordon <iangordon@users.sourceforge.net>.
297
298 3. Check for the e2fsprogs header <ext2fs/ext2_fs.h> instead of
299 the linux kernel header. This ensures that dump always has the
300 latest version of this file. Patch submitted by
301 Andreas Dilger <adilger@turbolinux.com>.
302
303 4. Report any filesystem present in either /etc/fstab with a
304 positive passno or /etc/dumpdates in dump -w output.
305 Patch submitted by Andreas Dilger <adilger@turbolinux.com>.
306
307 5. Fixed the looping problem in dump introduced in the
308 previous version.
309
310 6. Changed the -B option of dump to limit the size of
311 _compressed_ output per volume if compression is on.
312 Patch contributed by Helmut Jarausch
313 <jarausch@igpm.rwth-aachen.de>. Note however that, since
314 it is impossible to predict the size of the compressed
315 data before doing the compression, the -B limit is a bit
316 conservative.
317
318 7. Fixed a bug in reading the operator typed file/tape path for
319 the next volume in restore.
320
321 8. Implemented a "-F script" option to restore which permits the
322 user to specify a script which will be launched at the
323 beginning of each tape, useful for automatic programming of
324 tape changers for example. See the restore man page for the
325 script parameters and return codes.
326
327 9. Small fix for the QFA routines provided by Uwe Gohlke
328 <uwe@ugsoft.de>, and some recommendations for QFA uses in
329 the man pages.
330
331 10. Fixed the multivolume restoring where making a mistake
332 to the 'Mount next tape' prompt caused several blocks to
333 be lost.
334
335 11. Enhanced the -e option of dump to take as a parameter a
336 comma separated list of inode numbers.
337
338 12. Added the -E option to dump which specify a file containing
339 inode numbers to exclude from the dump.
340
341 13. Fixed the compressed multi-volume dump + restore.
342
343 Changes between versions 0.4b21 and 0.4b22 (released May 12, 2001)
344 ==================================================================
345
346 1. Made dump report the number of blocks dumped per volume.
347 Thanks to Kenneth Porter <shiva@well.com> for the suggestion.
348
349 2. Fix a bug in dump where waiting too much at the 'change volumes'
350 question caused the volume to be rewritten. Thanks to
351 Graham Reed <greed@users.sourceforge.net> for reporting the
352 bug and providing a patch.
353
354 3. Added a compression option to dump, contributed by Jerry
355 Peters <gapeters@worldnet.att.net>.
356
357 WARNING: the compression works only on files/pipes and
358 drives supporting variable block size.
359
360 WARNING2: tapes created using the compression feature are
361 incompatible with the standard dump tape format, and a
362 version of restore >= 0.4b22 will be needed for extraction.
363
364 4. Fixed some compilation problems for glibc 2.2.2 and 64 bit
365 architectures. Thanks to Paul Slootman <paul@debian.org> for
366 the patch and to Bdale Garbee <bdale@gag.com> for forwarding
367 it upstream.
368
369 5. Many cleanups (CPP defines, const char warnings, check of
370 ext2fs COMPAT flags, time_t cleanups, added libext2 version
371 in dump usage text) by Andreas Dilger <adilger@turbolinux.com>.
372
373 6. Made --prefix option work in configure. All the install path
374 are now based on the configure parameters.
375
376 7. Added the Quick File Access mode in dump/restore, contributed
377 by Uwe Gohlke <uwe@ugsoft.de>. In this mode, dump stores in
378 a file tape position for each inode, and this file is used by
379 restore (if called with parameter Q and the filename)
380 to directly position the tape at the file restore is currently
381 working on. This saves hours when restoring single files from
382 large backups, saves the tapes and the drive's head. Use
383 --enable-qfa option of configure to compile in the QFA support.
384
385 8. Added the possibility to dump several files and directories
386 in a single invocation of dump. Thanks to Uwe Gohlke
387 <uwe@ugsoft.de> for implementing this option.
388
389 9. Fixed the dumping and restoring of files with holes
390 (files ending with holes didn't get dumped or restored
391 correctly).
392
393 10. Fixed a socket descriptor leak in dump, which leaved opened
394 3 file descriptors per dump process (and there is one dump
395 process per tape).
396
397 11. Fixed dump large file system support, by explicit use of
398 open64/lseek64/etc functions (explicit use needed because
399 e2fsprogs libraries don't behave well when compiled with
400 FILE_OFFSET_BITS=64).
401
402 Changes between versions 0.4b20 and 0.4b21 (released January 13, 2001)
403 ======================================================================
404
405 1. Fixed some bugs in the dump -w|-W logic introduced by
406 the previous version. Thanks to Andreas Dilger
407 <adilger@turbolinux.com> for his help on this one.
408
409 2. Fixed again a compilation problem when using old e2fs
410 headers (filesystem label related). Thanks to many users
411 who reported this stupid error.
412
413 3. Fixed a build problem on old lib5 based systems dealing with
414 _PATH_MNTTAB being undefined. Thanks to John Adams
415 <johna@onevista.com> for reporting the bug.
416
417 4. Improved the error detection in restore when creating
418 the temporary files in TMPDIR. Now restore will corectly
419 report a 'no space left on device' error instead of
420 strange errors which could imply an archive corruption.
421 Thanks to Gabriel Somlo <somlo@cs.colostate.edu> and
422 bgjenero <bgjenero@sympatico.ca> for reporting the bug.
423
424 5. Added the throughput information in dump's progression
425 messages. Thanks to Andreas Dilger <adilger@turbolinux.com>
426 for the patch.
427
428 6. Use libext2fs's inode scanning routines, which are particularly
429 robust when dealing with errors (especially when having some
430 bad blocks on the filesystem). Many thanks to Colin
431 <colin@horizon.com> for reporting the bug, investigating
432 the issues, finding a workaround, writing the patch and
433 fully testing it... (of course, if this patch will break
434 anything, he is to blame for :-)).
435
436 7. Made dump and restore LFS aware. Dump can dump a filesystem
437 containing large files, generate a large file on output and
438 restore can restore them. This feature is not enabled by
439 default, use --enable-largefile option of configure to enable
440 it (you need to have a LFS aware glibc though). Thanks to
441 Andrea Arcangeli <andrea@suse.de> for submitting the patch,
442 and to Theodore T'so <tytso@valinux.com> for his always
443 useful thoughts.
444
445 8. Made dump ask upon a tape write error if it should rewrite
446 the current volume (assume this is a bad tape) or if it should
447 assume an end-of-tape condition (useful for tape drives which
448 behaves badly at the end of the tape). Thanks to Andreas
449 Dilger <adilger@turbolinux.com> for the suggestion.
450
451 Changes between versions 0.4b19 and 0.4b20 (released November 10, 2000)
452 =======================================================================
453
454 1. Fixed a small compilation problem due to a change
455 in the definintion of the struct sigaction in
456 glibc 2.0 / libc5. Thanks to Gunther Schlegel
457 <schlegel@riege.de> for reporting the bug and to
458 Dave Platt <dplatt@snulbug.mtview.ca.us> for suggesting
459 a fix.
460
461 2. Modified the label and uuid analysis in order to be
462 self-contained (independant of kernel/libc headers). This
463 should fix the compile with older kernel/libc headers and
464 will preserve the functionality. Thanks to Bernhard Erdmann
465 <bernhard.erdmann@gmx.de> for reporting the bug.
466
467 3. The 'exclude inode' option, if applied to a directory
468 inode, excludes now correctly all the leaves of this
469 directory. Thanks to John R. Dennison
470 <gerdesas@users.sourceforge.net> for reporting the bug.
471
472 4. Fixed the '-e' option to disable the possibility
473 to exclude the root inode (which causes the dumps to
474 be unreadable by restore). Prevented array overflow
475 when multiple -e options are used.
476
477 5. Fixed dump to correctly interpret a filesystem argument
478 which is a mountpoint and it is not an absolute path
479 (as specified in the fstab). Thanks to Bernhard R. Erdmann
480 <be@berdmann.de> for reporting the bug.
481
482 6. Made dump able to backup files larger than 2 GB. Note that
483 dump still doesn't cope with files larger than 4 GB.
484
485 7. Restore the real uid and gid before invoking an external
486 RSH program (big hole when dump or restore is suid root!).
487
488 8. Get the values for the default tape device and the location
489 of fstab file from the system headers. Thanks to
490 Andreas Dilger <adilger@turbolinux.com> for the patch.
491
492 9. Made dump -w|-W report all recognized filesystems
493 present in either /etc/fstab or /etc/dumpdates, and present
494 the list in the same order as in fstab file. Thanks
495 to Andreas Dilger <adilger@turbolinux.com> for the patch.
496
497 10. Made dump's -a (automatic end of tape detection) the
498 default. Specifying one of -B, -c, -d or -s options will
499 override the EOM detection. Thanks to Andreas Dilger
500 <adilger@turbolinux.com> for the patch.
501
502 11. Save the ext2 filesystem label into the dump volume label.
503 Specifying a volume label on the command line (-L option)
504 will override this feature. Thanks to Andreas Dilger
505 <adilger@turbolinux.com> for the patch.
506
507 Changes between versions 0.4b18 and 0.4b19 (released August 20, 2000)
508 =====================================================================
509
510 1. Fixed the signal handling in dump (which I broke in 0.4b17)
511 which was causing several strange dump failures (dump
512 hanged or segmentation faults etc).
513
514 2. Specified the default blocksize in the dump man page.
515
516 3. Changed two info messages of restore to be written on stdout
517 instead of stderr in order to leave stderr only for errors
518 or warnings. Thanks to Stephen Atwell
519 <satwell@urbana.css.mot.com> for the suggestion.
520
521 4. Corrected an off by one calculation which prevented
522 dumping a directory having a 1 character name length.
523 Thanks to Bernhard Erdmann <bernhard.erdmann@gmx.de>
524 for reporting the bug.
525
526 5. Reinforce dump in order to allow only level 0 and no
527 -u option when dumping a subdirectory, and document
528 this in the man page. Thanks to Bernhard Erdmann
529 <bernhard.erdmann@gmx.de> for reporting the bug.
530
531 6. Fixed a small harmless bug in restore which caused
532 in some conditions hard links to be restored several
533 times (therefore generation some warning messages).
534 Thanks to Kenneth Porter <shiva@well.com> for
535 reporting the bug.
536
537 7. Updated the RPM spec file to the latest RedHat version,
538 providing FHS packaging and other cosmetic changes.
539 You will need RPM version >= 3.0.5 in order to build the RPM.
540
541 8. Updated the configure script to check for libtermcap
542 before checking for libreadline, because we need this
543 library in order to compile the readline support.
544
545 9. Made dump understand the LABEL= and UUID= notation
546 both in /etc/fstab and on the command line. Note that
547 you will need the /proc filesystem in order to use
548 these notations. Thanks to Erik Troan <ewt@redhat.com>
549 for providing the patch.
550
551 Changes between versions 0.4b17 and 0.4b18 (released June 30, 2000)
552 ===================================================================
553
554 1. Fixed a potential buffer overflow in restore. Thanks
555 to Stan Bubrouski <satan@fastdial.net> for reporting
556 the bug.
557
558 2. Fixed a readline-related bug which prevented
559 'cat DUMPFILE | restore -i -f -' from working. Thanks
560 to Charles Karney <karney@users.sourceforge.net>
561 for the bug report.
562
563 3. Changed a few "panic" into "exit", causing restore to
564 be more stable against some attacks (like the last one
565 reported on Bugtraq, although the last version of restore
566 was not vulnerable - just dumped core). Thanks to
567 Andreas Hasenack <andreas@conectiva.com.br> for reporting
568 the bugs to me.
569
570 4. Removed the suid-root bit on dump and restore in the
571 default build (and generated RPMs). It should be safer
572 now. Users who need the suid-root capabilities in order
573 to make network backups should read first the man page
574 of dump and enable the suid bit manually.
575
576 5. Added -ltermcap to the compile parameters for restore
577 when using readline support, in order to make the compile
578 process work on some systems (SuSE ?). Thanks to
579 Patrik Schindler <poc@pocnet.net> for reporting the bug.
580
581 Changes between versions 0.4b16 and 0.4b17 (released June 1st, 2000)
582 ====================================================================
583
584 1. The -F script is called now *only* at the end of a tape,
585 not at the end of the dump. Thanks to Christian Haul
586 <haul@informatik.tu-darmstadt.de> for the bug report.
587
588 Normally, the device name and the current volume number
589 are passed on the command line of the script. However,
590 if you want the old style script (with no arguments
591 passed on the command line), you can enable it in
592 configure with the --enable-oldstylefscript.
593
594 2. Use posix signal handling to preserve dump functionality
595 with libc5. Thanks to Jeff Johnson <jbj@redhat.com> for
596 the patch.
597
598 3. Made the exit code of restore in case of a 'C'ompare
599 command reflect the result of the compare. An exit status
600 of 0 means the dump archive is correct, 1 means tape errors,
601 2 means that some files were modified. Thanks to Kenneth Porter
602 <shiva@well.com> for the suggestion.
603
604 4. Made (finally) quotes work in interactive restore.
605
606 5. Small fixes in order to allow dump to compile with a
607 really old e2fsprogs version. Thanks to Ian Zimmerman
608 <itz@speakeasy.org> for the bug report.
609
610 6. Add GNU readline capabilities to interactive restore.
611 Use configure's --enable-readline flag to enable this feature.
612 Thanks to Patrik Schindler <poc@pocnet.net> for the
613 suggestion.
614
615 7. Do the compare on the fly in restore 'C' mode (this will
616 allow not to exhaust the available /tmp space when
617 ccmparing large files). Thanks to Kenneth Porter
618 <shiva@well.com> for the suggestion.
619
620 Changes between versions 0.4b15 and 0.4b16 (released March 11, 2000)
621 ====================================================================
622
623 1. Fixed some several duplicate 'DUMP: DUMP:' in the
624 output of dump.
625
626 2. Corrected the estimation of blocks to dump. Note that
627 this is still wrong for large directory inodes, due
628 to the size of a BSD directory entry on the tape
629 (estimation is based on the size of the ext2 directory,
630 which is smaller because it doesn't contain the null
631 character at the end).
632
633 3. Made dump report the total number of MB written to
634 tape at the end of run. Thanks to W. Reilly Cooley
635 <wcooley@nakedape.cc> for the patch.
636
637 4. Added the -X option to restore giving the possibility
638 to read the names of the files to be extracted/listed
639 from a text file (in addition of the command line).
640 Thanks to Dejan Muhamedagic <dejan@quant-x.com> for the
641 patch.
642
643 5. Added the device name and the current volume number
644 as arguments to the end of tape script (dump -F option).
645
646 6. Made the multi-volume dump work again (after having
647 broken it in 0.4b15).
648
649 Changes between versions 0.4b14 and 0.4b15 (released March 2, 2000)
650 ===================================================================
651
652 1. Added a prompt command in interactive restore mode. Thanks
653 to Andreas Dilger <adilger@home.com> for the patch.
654
655 2. Fixed a buffer overflow problem in dump (caused by
656 not checking the size of the filesystem parameter).
657 Thanks to Kim Yong-jun <loveyou@hackerslab.org> for
658 reporting this on Bugtraq (and to several dump users
659 who forwarded me his mail).
660
661 3. Added the '-F script' option to dump in order to
662 launch a script at the end of each tape (to be used
663 with a tape changer, or to notify the sysadmin by
664 pager etc.).
665
666 4. Fixed a bug in restore compare code caused by the changes
667 I made in 0.4b14.
668
669 5. Fixed the treatment of options using the old BSD syntax
670 in both dump and restore.
671
672 Changes between versions 0.4b13 and 0.4b14 (released February 10, 2000)
673 =======================================================================
674
675 1. Fixed a bug in dump which may have caused invalid deleted
676 directories to be dumped out if they were deleted after the
677 mapping pass. This could occure on active filesystem and lead
678 to heap corruption (causing dump malfunction in many possible ways).
679 Thanks to Jan Sanislo <oystr@cs.washington.edu> for finding this
680 bug and submitting the patch.
681
682 2. Fixed the handling of the filesystem parameter in dump. A
683 '/mnt/disk' parameter caused the disk contents to be dumped,
684 but a '/mnt/disk/' parameter caused the mountpoint directory
685 to be dumped (generally an empty directory).
686
687 3. Improved the output of dump in order to tell which directory
688 it is currently dumping (when dumping a subtree).
689
690 4. Added the '-e' exclude inode option to dump. Thanks to
691 Isaac Chuang <ike@isl.stanford.edu> for contributing with the patch.
692
693 5. Added a REPORTING-BUGS file in order to provide a guide
694 on how to correctly report a bug in dump/restore.
695
696 6. Fixed a restore problem when restoring a hard link to an inode
697 having the immutable or append-only attribute set. Thanks to
698 Ambrose Li <acli@mingpaoxpress.com> for submitting the patch.
699
700 7. Fixed a compatibility problem between dumps made on little
701 endian machines (the format was correct) and big endian
702 machines (the format was incorrect). This fix break the
703 compatibility with the older dumps made on big endian
704 machines (sparc, m86k, ppc etc). For the first time in
705 linux dump's history (I believe), the dumps made by *BSD,
706 Linux/alpha, Linux/sparc and Linux/x86 are compatible,
707 so interchangeable. Thanks to Rob Cermak
708 <cermak@ahab.rutgers.edu> for submitting the bug and
709 helping me test the fix.
710
711 8. Fixed the way dump reports the remaining percent/time, if
712 the number of blocks actually dumped exceeds the estimated
713 number of blocks. Thanks to Jean-Paul van der Jagt
714 <jeanpaul@dutepp0.et.tudelft.nl> for reporting the bug.
715
716 Changes between versions 0.4b12 and 0.4b13 (released January 21, 2000)
717 ======================================================================
718
719 1. Small Y2K fix in the man pages :). Thanks to Bernhard Sadlowski
720 <sadlowsk@Mathematik.Uni-Bielefeld.DE> for reporting the bug.
721
722 2. Removed the requirement to build the RPM as root from the
723 spec file. Thanks to Christian Weisgerber
724 <naddy@mips.rhein-neckar.de> for submitting this.
725
726 3. Fixed a bug in dump related to the 'filetype' feature of ext2fs,
727 causing dump to block when dumping really huge filesystems.
728 Many thanks to Patrik Schindler <poc@pocnet.net> for
729 helping me find this bug.
730
731 4. Fixed the treatment for an interrupt signal when dump access
732 the remote tape through RSH. Thanks to Christian Weisgerber
733 <naddy@mips.rhein-neckar.de> for providing the patch.
734
735 5. Fixed a bug which was causing dump/restore to display
736 garbage characters instead of the remote host name.
737
738 Changes between versions 0.4b11 and 0.4b12 (released January 8, 2000)
739 =====================================================================
740
741 1. Small fix in the dump man page. Thanks to Thorsten Kukuk
742 <kukuk@suse.de> for submitting the patch.
743
744 2. Fix for the exit code when using the size estimate option of
745 dump. Thanks to Matti Taskinen <mkt@rni.helsinki.fi> for
746 submitting the patch.
747
748 3. Handle EINTR in atomical reads/writes in dump, which was causing
749 dump to fail on some systems. Thanks to Eric Jergensen
750 <eric@dvns.com> for reporting the bug and submitting the patch.
751
752 4. Handle more than 16 characters for the device names in dumpdates.
753 (up to 255 now). Thanks to Rainer Clasen <bj@ncc.cicely.de> for
754 tracking down the problem and proposing the solution.
755
756 5. Fixed a bug in dump which prevented the creation of the
757 dumpdates file when doing a 0-level dump without already
758 having a dumpdates file. Thanks to Patrik Schindler
759 <poc@pocnet.net> for reporting the bug.
760
761 6. Changed the way dump 'S' flag reports the size estimate
762 from number of blocks into bytes (making it compatible
763 with the Solaris version, and simplifying things for
764 amanda users). Thanks to Jason L Tibbitts III
765 <tibbs@math.uh.edu> for reporting the bug.
766
767 7. Fixed a compatibility problem in linux/alpha dump tape format.
768 Now the linux/alpha dump are (again) compatible with the
769 other dump formats. But this breaks compatibility with
770 older dumps made on alpha. Thanks to Mike Tibor
771 <tibor@lib.uaa.alaska.edu> for helping me in finding this bug.
772
773 Changes between versions 0.4b10 and 0.4b11 (released December 5, 1999)
774 ======================================================================
775
776 1. Added a '--enable-kerberos' to configure.
777
778 2. Added a 'S' option to dump which determines the amount of space
779 that is needed to perform the dump without actually doing it, similar
780 to the Sun's ufsdump 'S' option. Patch contributed by Rob Cermak
781 <cermak@ahab.rutgers.edu>.
782
783 3. Added a 'M' multi-volume option to dump and restore which enables
784 dumping to multiple files (useful when dumping to an ext2
785 partition to make several dump files in order to bypass the 2GB
786 file size limitation). The 'f' argument is treated as a prefix and
787 the output files will be named <prefix>001, <prefix>002 etc. With
788 the 'M' flag, restore automatically selects the right file without
789 asking to enter a new tape each time.
790
791 4. Fixed a memory leak which was causing dump to grow very big
792 (270MB when dumping a 10GB filesystem...). Thanks to Jason
793 Fearon <jasonf@netrider.org.au> for reporting the bug.
794
795 Changes between versions 0.4b9 and 0.4b10 (released November 21, 1999)
796 ======================================================================
797
798 1. Make configure test if the system glob routines support
799 extended syntax (ALTDIRFUNC). If not, use the internal glob
800 routines instead of system ones. Thanks to Bernhard Sadlowski
801 <sadlowsk@Mathematik.Uni-Bielefeld.DE> for reporting the bug
802 and helping me resolve this and other minor libc5 compiling
803 glitches.
804
805 2. Fix a problem when dumping a ext2fs with the 'filetype'
806 feature enabled. Thanks to Patrick J. LoPresti
807 <patl@cag.lcs.mit.edu> for reporting the bug and to
808 Theodore Y. Ts'o <tytso@mit.edu> for providing the patch.
809
810 3. Made the nodump flag work on directories. A directory which
811 has the nodump flag gets never dumped, regardless of its
812 contents.
813
814 4. Integrate a patch from Jeremy Fitzhardinge <jeremy@goop.org>
815 which allows dump on an active ext3 filesystem. However, this
816 is a "quick and dirty" patch which enables backup of an ext3
817 filesystem through the ext2 compatibility (by ignoring the
818 NEEDS_RECOVERY bit). The journal file is not recognized and
819 it is dumped (it should not).
820
821 5. Test the superblock compatibility flags when dumping, in order
822 to be sure that we know how to deal with specific features.
823
824 Changes between versions 0.4b8 and 0.4b9 (released November 5, 1999)
825 ====================================================================
826
827 1. Use lchown instead of chown, fixing a possible security problem
828 when restoring symlinks (a malicious user could use this
829 to deliberately corrupt the ownership of important system files).
830 Thanks to Chris Siebenmann <cks@utcc.utoronto.ca> for detecting
831 this and providing the patch.
832
833 Changes between versions 0.4b7 and 0.4b8 (released November 3, 1999)
834 ====================================================================
835
836 1. Put dump sources under CVS, added Id tags in all files so
837 one can use 'ident' on binary files.
838
839 2. Added the dump/restore version in the usage text so one can
840 easily verify the version he is using.
841
842 3. Small patch from Nuno Oliveira <nuno@eq.uc.pt> which fixes
843 a va_start/va_end problem on linux-ppc (always call va_start
844 va_end in pairs each time we use a vararg function).
845
846 4. Added again the DT_* constants because old libc does not
847 contain them :(. Thanks to Eric Maisonobe <virnet@nat.fr>
848 for submitting the bug report.
849
850 5. Use ext2fs_llseek instead of llseek. With recent e2fsprogs
851 this should enable dumping big (huge) filesystems.
852
853 6. Added the RSH environment variable in order to be able to
854 use a rsh replacement like ssh when doing remote backups (and
855 bypass the security limitations of rcmd). Now you can do remote
856 backups without being root (or making dump setuid root).
857
858 7. Modified again the way dumpdates works. For incremental dumps,
859 we need to read dumpdates even if we are not using 'u' option.
860 Thanks to Bdale Garbee <bdale@gag.com> for his ideas on how
861 this should work.
862
863 Changes between versions 0.4b6 and 0.4b7 (released October 8, 1999)
864 ===================================================================
865
866 1. Removed the 'k' flag from the restore 'about' text if kerberos
867 was not compiled in.
868
869 2. Prototyped (f)setflags from e2fsprogs and corrected the calls
870 to them (fsetflags takes a char*, setflags an open fd!).
871
872 3. (f)setflags is called only if the flags aren't empty. If the
873 file is a special file, a warning is printed, because changing
874 flags implies opening the device. Normally, a special file
875 should not have any flag... (Debian bug #29775, patch provided
876 by Abhijit Dasgupta <abhijit@ans.net>).
877
878 4. Made possible to dump a file system not mentioned in /etc/fstab.
879 (Debian bug #11904, patch provided by Eirik Fuller <eirik@netcom.com>).
880
881 5. Changed the default behaviour to not create dumpdates
882 unless 'u' option is specified. Removed the old "debian-patch"
883 which provided the same thing. (Debian bug #38136, #33818).
884
885 6. Removed all those dump*announce, since they were getting old...
886
887 7. Added warning messages if dumpdates does not exist and
888 when an inferior level dump does not exist (except for a level 0
889 dump).
890
891 8. Debugged the glob calls in interactive mode: restore used a
892 dirent struct which was different from the /usr/include/dirent.h
893 one (this used to work, is it a glibc2 change?), so none of the
894 compat glob (which used /usr/include/dirent.h) or the system glob
895 worked. Restore use now the system dirent (and the system
896 DT_* constants), which are compatible with BSD ones.
897
898 9. Added a configure flag (--with-dumpdatespath) to specify
899 the location of dumpdates file. By default, it is
900 /etc/dumpdates.
901
902 10. Added the "AUTHOR" and "AVAILABILITY" sections and
903 included the current date/version in man pages.
904
905 11. Corrected the estimation of remaining time when
906 the operator doesn't change the tapes quickly enough. This
907 was an old bug, I thought I corrected it, and discovered
908 that in fact it was corrected in two different places, so
909 the results canceled each other...
910
911 Changes between versions 0.4b5 and 0.4b6 (released October 1, 1999)
912 ===================================================================
913
914 1. Integrated multiple patches from RedHat, Debian and SuSE:
915
916 - tweak dump/itime.c to not try to read dumpdates if the 'u' option
917 isn't specified.
918 - several fixes in the man pages.
919 - update the default tape device to /dev/st0.
920 - many updates for Linux Alpha (byte ordering, size_t etc).
921 - buffer overruns.
922 - use environment variable for TMPDIR (instead of /tmp).
923 - use sigjmp_buf instead of jmp_buf (RedHat bug #3260).
924 - workaround egcs bug (RedHat bugs #4281 and #2989).
925 - wire $(OPT) throughout Makefile's.
926
927 2. Upgrade the dump revision to 1, making possible to dump filesystems
928 made with e2fsprogs-1.15 or newer. Nothing seems to break...
929
930 3. Fix some compile warnings, prototype all functions.
931
932 4. Use glibc err/glob instead of internal compatibility
933 routines (only if available).
934
935 5. Fix a compile error on Linux 2.2.7 / libc5 (5.4.44) (patch provided
936 by Bernhard Sadlowski <sadlowsk@mathematik.uni-bielefeld.de>).
937
938 Changes between versions 0.4b4 and 0.4b5 (released September 22, 1999)
939 ======================================================================
940
941 1. Integrated the changes from FreeBSD-3.1-RELEASE
942 (mostly bug fixes, buffer overruns, dump has now an "automatic
943 tape length calculation" flag, dump/restore can use kerberos now
944 (this is NOT tested), use environment variables for TAPE and
945 RMT etc.).
946
947 2. Integrated three RedHat patches ("glibc", "kernel" and "bread" patches)
948
949 3. Corrected a bug in restore when using 'C' option with multi-volumes
950 tapes (files splited accros two tapes give "size changed" errors
951 when comparing).
952
953 4. Corrected the long standing bug when dumping multiple tapes.
954 This works for me, needs further testing.
955
956 Changes between versions 0.4b3 and 0.4b4 (released January 17, 1997)
957 ====================================================================
958
959 1. Dump now runs correctly on kernels 2.1.x
960 Fix made by Gerald Peters <gapeters@worldnet.att.net>
961
962 Changes between versions 0.4b2 and 0.4b3
963 ========================================
964
965 1. Use realpath() if available
966
967 2. Report statistics
968
969 Changes between versions 0.4b1 and 0.4b2
970 ========================================
971
972 1. Fixed the bug fix from Greg Lutz (I had made a mistake when integrating
973 the patch)
974
975 2. Fixed restore to make it able to read FreeBSD 2.x dumps again
976
977 3. Fixed configure.in to correctly handle --enable-rmt
978
979 Changes between versions 0.3 and 0.4b1
980 ======================================
981
982 1. Integrated the changes from 4.4BSD-Lite2
983
984 2. Integrated the patches from Debian and Red Hat
985
986 3. Portability changes: use the __u32, __u16, __s32, and __s16 types
987
988 4. Changed dump to use the Ext2fs library to get block addresses. This
989 should solve the endianness problem on SparcLinux.
990
991 5. Created a configure.in file (shamelessly stolen from the e2fsprogs
992 distribution's one) to use autoconf
993
994 6. Fixed a few minor bugs
995
996 Changes between versions 0.2e and 0.2f
997 ======================================
998
999 1. Added the creation of named pipes (fifos) in restore.
1000
1001 2. Added the -N flag in the restore manual page.
1002
1003 3. Added the file kernel.patch which contains the llseek() optimization
1004 patch for 1.2.x kernels.
1005
1006 4. Fixed a bug in the restoration of symbolic links: owner and group were
1007 not restored.
1008
1009 5. Integrated some changes from FreeBSD 2.2.
1010
1011 6. Added a call to ftruncate() after restoring each file to restore
1012 correctly files ending by a hole.
1013
1014 Changes between versions 0.2d and 0.2e
1015 ======================================
1016
1017 1. Fixed a bug in the "set owner/mode" process. Every file was restored
1018 with owner = root (0) and group = root/wheel/whatever (0).
1019
1020 Changes between versions 0.2c and 0.2d
1021 ======================================
1022
1023 1. Dump is now able to backup 2GB+ filesystems.
1024
1025 2. Dump and restore can now be linked as static binaries.
1026
1027 Changes between versions 0.2b and 0.2c
1028 ======================================
1029
1030 1. Fixed a bug when dumping ``slow'' (i.e. normal) symbolic links.
1031
1032 Changes between versions 0.2a and 0.2b
1033 ======================================
1034
1035 1. Really fixed the bug that I should have corrected in 0.2a.
1036
1037 2. Enabled optimization again.
1038
1039 Changes between versions 0.2 and 0.2a
1040 =====================================
1041
1042 1. Disabled the optimization during compilation.
1043
1044 Changes between versions 0.1 and 0.2
1045 ====================================
1046
1047 1. Fixed a bug in fstab.c which caused a null pointer to be stored in
1048 the fs_type field (actually, I modified the file fstab.c to make it
1049 use the mntent functions).
1050
1051 2. Dump and restore now use a 4.3 BSD compatible dump format. Backups
1052 made by dump should be readable by the BSD restore and backups made
1053 by the BSD dump should be readable by restore. Unfortunately, this
1054 means that the dump format has changed between version 0.1 and version
1055 0.2 :-(
1056
1057 3. Dump is now able to backup a subtree, it is no longer limited to whole
1058 filesystems like the BSD version.
1059
1060 4. Dump now uses ext2_llseek() so it is able to backup filesystems bigger
1061 than 2 GB.
1062
1063 Changes between versions 0.0 and 0.1
1064 ====================================
1065
1066 1. Now create links rdump and rrestore during the `make install' step.
1067
1068 2. Linux port specific bugs added to the manual pages
1069
1070 3. Incorrect estimation of the number of tapes blocks fixed when doing
1071 incremental backups.
1072
1073 4. Better ls-like format in restore in interactive mode.