]> git.wh0rd.org - dump.git/blob - CHANGES
ee9b9d6a81335ea6e857c55155e8e4e6855620f0
[dump.git] / CHANGES
1 $Id: CHANGES,v 1.163 2002/02/21 09:25:31 stelian Exp $
2
3 Changes between versions 0.4b27 and 0.4b28 (released ?????????????????)
4 =======================================================================
5
6 1. Fixed a bug in the QFA feature which could have caused
7 a bad tape positionning on restore (causing extra delays in
8 restore). Credits go to Uwe Gohlke <uwe@ugsoft.de>.
9
10 Changes between versions 0.4b26 and 0.4b27 (released February 15, 2002)
11 =======================================================================
12
13 1. Fixed behaviour of dump when exceeding resource limits
14 (SIGXFSZ treatment).
15
16 2. Added the -L flag to restore to allow the user to specify a
17 maximal allowed number of miscompares when using restore
18 with the -C option to check the backup.
19
20 3. Detailed the manual entry for the -N option of restore.
21
22 4. Added the -a flag to restore to make able doing unattended
23 restores in -i or -x mode (automatically walks through the
24 multiple dump volumes).
25
26 5. Extended the QFA mode to work with local files and/or
27 remote tapes and files. This way, restore can know in advance
28 the tape number and the offset for the inodes to extract and
29 can minimize the extraction time by seeking directly to the
30 good tape/offset.
31
32 6. Added the -A <archive> option to both dump and restore,
33 which makes dump to archive a dump table-of-contents in
34 the specified file to be used by restore to determine
35 whether a file is in the dump file that is being restored.
36 (the archive file syntax is also compatible with the
37 Solaris ufsdump generated one).
38
39 7. Small fix in restore making it able to read some (broken ?)
40 Solaris ufsdump tapes.
41
42 8. Fixed dump to correctly recognise the root filesystem when using
43 ext2 disk labels (LABEL=/). Thanks to John Yu <jky@it.bu.edu>
44 for reporting this bug.
45
46 9. Added the -P <file> option to restore to create a
47 Quick File Access file from an already made dump. Patch
48 contributed by Uwe Gohlke <uwe@ugsoft.de>.
49
50 10. Made restore compile and run on Solaris, making it a
51 possible replacement for the standard ufsrestore. Port was
52 contributed by Uwe Gohlke <uwe@ugsoft.de>.
53
54 Changes between versions 0.4b25 and 0.4b26 (released January 7, 2002)
55 =====================================================================
56
57 1. Added a set of backup scripts from Eugenio Diaz
58 <getnito@yahoo.com> in the examples section. It features
59 automatic (cron based) full and incremental dumping of
60 several filesystems on a separate filesystem.
61
62 2. Fixed a off-by-one miscalculation which disabled dumping
63 a one letter subdirectory. Thanks to Chris Siebenmann
64 <cks@utcc.utoronto.ca> for reporting the bug.
65
66 3. Fixed several restore bugs occuring when trying to
67 restore some missing files on the tape. Thanks to Chris
68 Siebenmann <cks@utcc.utoronto.ca> for reporting the bug.
69
70 4. Fixed --with-ldopts configure argument passing, installing from
71 a separate object directory, makefile cleanups contributed
72 by <splite@purdue.edu>.
73
74 5. Fix a bug which could caused, in some conditions, the highest
75 number inode of a filesystem, to not be dumped. Many thanks
76 to Chris Siebenmann <cks@utcc.utoronto.ca> for helping me
77 chase this bug.
78
79 Changes between versions 0.4b24 and 0.4b25 (released November 17, 2001)
80 =======================================================================
81
82 1. Added a mini howto from Patrick Walsh in the examples
83 directory.
84
85 2. Minor man pages syntax corrections. Thanks to
86 Chris Danis <screechco@home.com> for reporting the bugs.
87
88 3. Added a script from David B. Peterson <dave@toppledwagon.com>
89 to the examples section. It features dumping several
90 filesystems to a remote tape drive upon ssh.
91
92 4. Added a patch provided by Richard Jones <rich@annexia.org>
93 which allows BRADEMAX (number of read errors tolerated by
94 dump) to be adjusted using the -I option.
95
96 5. Fixed a bug which disabled doing "restore -C -f -". Thanks
97 to Clemens Stiglechner <a9401816@unet.univie.ac.at> for the
98 patch.
99
100 6. Add the -l option to restore to specify if, when doing a
101 remote restore, the file used to access the backup is a
102 regular file (the defaults being a tape drive). Restore needs
103 to know this information when reading a remote compressed
104 dump. Previously, this information was autodetected, but
105 the autodetection code fails (with ioctl: Inappropriate ioctl
106 for device) when using a non Linux remote box. Thanks to
107 many users and especially to Eros Albertazzi
108 <eros@lamel.bo.cnr.it> for reporting this.
109
110 7. Found a workaroung for the dump deadlock problem (3 childs
111 stuck in pause(), father in read()). The workaround seems
112 to work for me and several beta-testers. If it doesn't work
113 for you, please report back.
114
115 8. Updated the RPM spec file (BuildPrereq, URL etc).
116
117 Changes between versions 0.4b23 and 0.4b24 (released September 12, 2001)
118 ========================================================================
119
120 1. Fixed the permissions of a newly created QFA file by dump.
121
122 2. Cleaned up the source of dump (the external variables
123 definition was a complete mess, making possible to have
124 objects overlap).
125
126 3. Fixed restore to use the full tape volume path when doing
127 a compare (since it changes the working directory to the
128 filesystem being compared in the process).
129
130 4. Added the -q option to dump which makes dump abort
131 whenever it needs operator attention. It should be
132 easier to use dump in scripts with this option.
133
134 5. Detect the use of incompatible options to dump and
135 refuse them (like -a and -B options together).
136
137 6. Added bzip2 compression to dump/restore (use option -j level
138 to select it). Note that you will need a restore version
139 >= 0.4b24 in order to restore a bzip2 compressed dump.
140 The same warning as for the zlib compression applies:
141 the tape format of a bzip2 dump is not compatible with the
142 original BSD tape format.
143
144 7. Fixed a overflow problem in dump corrupting the dump when
145 very large files were encountered. Thanks to Vladimir Nadvornik
146 for the bug report.
147
148 8. Added a ioctl(BLKFLSBUF) in dump which should flush the
149 kernel buffer/page cache before starting the dump, helping
150 a bit those who use dump on mounted filesystems. Thanks to
151 John Yu <jky@it.bu.edu> and to Theodore T'so <tytso@mit.edu>
152 for this suggestion.
153
154 9. Updated the RPM spec file following the RedHat changes
155 (dynamically linked binaries now in /usr/sbin etc).
156
157 10. Added a patch from Helmut Jarausch <jarausch@igpm.rwth-aachen.de>
158 which enables restore to recognise multi volume compressed dumps
159 done on CD. Included his scripts for dump (which pipe the dump
160 output directly into cdrecord) and restore. There is now
161 possible to to backups to CD on the fly!
162
163 Changes between versions 0.4b22 and 0.4b23 (released July 20, 2001)
164 ===================================================================
165
166 1. Fixed a buffer overflow in restore/tape.c. Patch provided by
167 Marcus Meissner (Caldera International Security Dept.).
168
169 2. Implement the Sun rmt extended protocol. Patch submitted
170 by Ian Gordon <iangordon@users.sourceforge.net>.
171
172 3. Check for the e2fsprogs header <ext2fs/ext2_fs.h> instead of
173 the linux kernel header. This ensures that dump always has the
174 latest version of this file. Patch submitted by
175 Andreas Dilger <adilger@turbolinux.com>.
176
177 4. Report any filesystem present in either /etc/fstab with a
178 positive passno or /etc/dumpdates in dump -w output.
179 Patch submitted by Andreas Dilger <adilger@turbolinux.com>.
180
181 5. Fixed the looping problem in dump introduced in the
182 previous version.
183
184 6. Changed the -B option of dump to limit the size of
185 _compressed_ output per volume if compression is on.
186 Patch contributed by Helmut Jarausch
187 <jarausch@igpm.rwth-aachen.de>. Note however that, since
188 it is impossible to predict the size of the compressed
189 data before doing the compression, the -B limit is a bit
190 conservative.
191
192 7. Fixed a bug in reading the operator typed file/tape path for
193 the next volume in restore.
194
195 8. Implemented a "-F script" option to restore which permits the
196 user to specify a script which will be launched at the
197 beginning of each tape, useful for automatic programming of
198 tape changers for example. See the restore man page for the
199 script parameters and return codes.
200
201 9. Small fix for the QFA routines provided by Uwe Gohlke
202 <uwe@ugsoft.de>, and some recommendations for QFA uses in
203 the man pages.
204
205 10. Fixed the multivolume restoring where making a mistake
206 to the 'Mount next tape' prompt caused several blocks to
207 be lost.
208
209 11. Enhanced the -e option of dump to take as a parameter a
210 comma separated list of inode numbers.
211
212 12. Added the -E option to dump which specify a file containing
213 inode numbers to exclude from the dump.
214
215 13. Fixed the compressed multi-volume dump + restore.
216
217 Changes between versions 0.4b21 and 0.4b22 (released May 12, 2001)
218 ==================================================================
219
220 1. Made dump report the number of blocks dumped per volume.
221 Thanks to Kenneth Porter <shiva@well.com> for the suggestion.
222
223 2. Fix a bug in dump where waiting too much at the 'change volumes'
224 question caused the volume to be rewritten. Thanks to
225 Graham Reed <greed@users.sourceforge.net> for reporting the
226 bug and providing a patch.
227
228 3. Added a compression option to dump, contributed by Jerry
229 Peters <gapeters@worldnet.att.net>.
230
231 WARNING: the compression works only on files/pipes and
232 drives supporting variable block size.
233
234 WARNING2: tapes created using the compression feature are
235 incompatible with the standard dump tape format, and a
236 version of restore >= 0.4b22 will be needed for extraction.
237
238 4. Fixed some compilation problems for glibc 2.2.2 and 64 bit
239 architectures. Thanks to Paul Slootman <paul@debian.org> for
240 the patch and to Bdale Garbee <bdale@gag.com> for forwarding
241 it upstream.
242
243 5. Many cleanups (CPP defines, const char warnings, check of
244 ext2fs COMPAT flags, time_t cleanups, added libext2 version
245 in dump usage text) by Andreas Dilger <adilger@turbolinux.com>.
246
247 6. Made --prefix option work in configure. All the install path
248 are now based on the configure parameters.
249
250 7. Added the Quick File Access mode in dump/restore, contributed
251 by Uwe Gohlke <uwe@ugsoft.de>. In this mode, dump stores in
252 a file tape position for each inode, and this file is used by
253 restore (if called with parameter Q and the filename)
254 to directly position the tape at the file restore is currently
255 working on. This saves hours when restoring single files from
256 large backups, saves the tapes and the drive's head. Use
257 --enable-qfa option of configure to compile in the QFA support.
258
259 8. Added the possibility to dump several files and directories
260 in a single invocation of dump. Thanks to Uwe Gohlke
261 <uwe@ugsoft.de> for implementing this option.
262
263 9. Fixed the dumping and restoring of files with holes
264 (files ending with holes didn't get dumped or restored
265 correctly).
266
267 10. Fixed a socket descriptor leak in dump, which leaved opened
268 3 file descriptors per dump process (and there is one dump
269 process per tape).
270
271 11. Fixed dump large file system support, by explicit use of
272 open64/lseek64/etc functions (explicit use needed because
273 e2fsprogs libraries don't behave well when compiled with
274 FILE_OFFSET_BITS=64).
275
276 Changes between versions 0.4b20 and 0.4b21 (released January 13, 2001)
277 ======================================================================
278
279 1. Fixed some bugs in the dump -w|-W logic introduced by
280 the previous version. Thanks to Andreas Dilger
281 <adilger@turbolinux.com> for his help on this one.
282
283 2. Fixed again a compilation problem when using old e2fs
284 headers (filesystem label related). Thanks to many users
285 who reported this stupid error.
286
287 3. Fixed a build problem on old lib5 based systems dealing with
288 _PATH_MNTTAB being undefined. Thanks to John Adams
289 <johna@onevista.com> for reporting the bug.
290
291 4. Improved the error detection in restore when creating
292 the temporary files in TMPDIR. Now restore will corectly
293 report a 'no space left on device' error instead of
294 strange errors which could imply an archive corruption.
295 Thanks to Gabriel Somlo <somlo@cs.colostate.edu> and
296 bgjenero <bgjenero@sympatico.ca> for reporting the bug.
297
298 5. Added the throughput information in dump's progression
299 messages. Thanks to Andreas Dilger <adilger@turbolinux.com>
300 for the patch.
301
302 6. Use libext2fs's inode scanning routines, which are particularly
303 robust when dealing with errors (especially when having some
304 bad blocks on the filesystem). Many thanks to Colin
305 <colin@horizon.com> for reporting the bug, investigating
306 the issues, finding a workaround, writing the patch and
307 fully testing it... (of course, if this patch will break
308 anything, he is to blame for :-)).
309
310 7. Made dump and restore LFS aware. Dump can dump a filesystem
311 containing large files, generate a large file on output and
312 restore can restore them. This feature is not enabled by
313 default, use --enable-largefile option of configure to enable
314 it (you need to have a LFS aware glibc though). Thanks to
315 Andrea Arcangeli <andrea@suse.de> for submitting the patch,
316 and to Theodore T'so <tytso@valinux.com> for his always
317 useful thoughts.
318
319 8. Made dump ask upon a tape write error if it should rewrite
320 the current volume (assume this is a bad tape) or if it should
321 assume an end-of-tape condition (useful for tape drives which
322 behaves badly at the end of the tape). Thanks to Andreas
323 Dilger <adilger@turbolinux.com> for the suggestion.
324
325 Changes between versions 0.4b19 and 0.4b20 (released November 10, 2000)
326 =======================================================================
327
328 1. Fixed a small compilation problem due to a change
329 in the definintion of the struct sigaction in
330 glibc 2.0 / libc5. Thanks to Gunther Schlegel
331 <schlegel@riege.de> for reporting the bug and to
332 Dave Platt <dplatt@snulbug.mtview.ca.us> for suggesting
333 a fix.
334
335 2. Modified the label and uuid analysis in order to be
336 self-contained (independant of kernel/libc headers). This
337 should fix the compile with older kernel/libc headers and
338 will preserve the functionality. Thanks to Bernhard Erdmann
339 <bernhard.erdmann@gmx.de> for reporting the bug.
340
341 3. The 'exclude inode' option, if applied to a directory
342 inode, excludes now correctly all the leaves of this
343 directory. Thanks to John R. Dennison
344 <gerdesas@users.sourceforge.net> for reporting the bug.
345
346 4. Fixed the '-e' option to disable the possibility
347 to exclude the root inode (which causes the dumps to
348 be unreadable by restore). Prevented array overflow
349 when multiple -e options are used.
350
351 5. Fixed dump to correctly interpret a filesystem argument
352 which is a mountpoint and it is not an absolute path
353 (as specified in the fstab). Thanks to Bernhard R. Erdmann
354 <be@berdmann.de> for reporting the bug.
355
356 6. Made dump able to backup files larger than 2 GB. Note that
357 dump still doesn't cope with files larger than 4 GB.
358
359 7. Restore the real uid and gid before invoking an external
360 RSH program (big hole when dump or restore is suid root!).
361
362 8. Get the values for the default tape device and the location
363 of fstab file from the system headers. Thanks to
364 Andreas Dilger <adilger@turbolinux.com> for the patch.
365
366 9. Made dump -w|-W report all recognized filesystems
367 present in either /etc/fstab or /etc/dumpdates, and present
368 the list in the same order as in fstab file. Thanks
369 to Andreas Dilger <adilger@turbolinux.com> for the patch.
370
371 10. Made dump's -a (automatic end of tape detection) the
372 default. Specifying one of -B, -c, -d or -s options will
373 override the EOM detection. Thanks to Andreas Dilger
374 <adilger@turbolinux.com> for the patch.
375
376 11. Save the ext2 filesystem label into the dump volume label.
377 Specifying a volume label on the command line (-L option)
378 will override this feature. Thanks to Andreas Dilger
379 <adilger@turbolinux.com> for the patch.
380
381 Changes between versions 0.4b18 and 0.4b19 (released August 20, 2000)
382 =====================================================================
383
384 1. Fixed the signal handling in dump (which I broke in 0.4b17)
385 which was causing several strange dump failures (dump
386 hanged or segmentation faults etc).
387
388 2. Specified the default blocksize in the dump man page.
389
390 3. Changed two info messages of restore to be written on stdout
391 instead of stderr in order to leave stderr only for errors
392 or warnings. Thanks to Stephen Atwell
393 <satwell@urbana.css.mot.com> for the suggestion.
394
395 4. Corrected an off by one calculation which prevented
396 dumping a directory having a 1 character name length.
397 Thanks to Bernhard Erdmann <bernhard.erdmann@gmx.de>
398 for reporting the bug.
399
400 5. Reinforce dump in order to allow only level 0 and no
401 -u option when dumping a subdirectory, and document
402 this in the man page. Thanks to Bernhard Erdmann
403 <bernhard.erdmann@gmx.de> for reporting the bug.
404
405 6. Fixed a small harmless bug in restore which caused
406 in some conditions hard links to be restored several
407 times (therefore generation some warning messages).
408 Thanks to Kenneth Porter <shiva@well.com> for
409 reporting the bug.
410
411 7. Updated the RPM spec file to the latest RedHat version,
412 providing FHS packaging and other cosmetic changes.
413 You will need RPM version >= 3.0.5 in order to build the RPM.
414
415 8. Updated the configure script to check for libtermcap
416 before checking for libreadline, because we need this
417 library in order to compile the readline support.
418
419 9. Made dump understand the LABEL= and UUID= notation
420 both in /etc/fstab and on the command line. Note that
421 you will need the /proc filesystem in order to use
422 these notations. Thanks to Erik Troan <ewt@redhat.com>
423 for providing the patch.
424
425 Changes between versions 0.4b17 and 0.4b18 (released June 30, 2000)
426 ===================================================================
427
428 1. Fixed a potential buffer overflow in restore. Thanks
429 to Stan Bubrouski <satan@fastdial.net> for reporting
430 the bug.
431
432 2. Fixed a readline-related bug which prevented
433 'cat DUMPFILE | restore -i -f -' from working. Thanks
434 to Charles Karney <karney@users.sourceforge.net>
435 for the bug report.
436
437 3. Changed a few "panic" into "exit", causing restore to
438 be more stable against some attacks (like the last one
439 reported on Bugtraq, although the last version of restore
440 was not vulnerable - just dumped core). Thanks to
441 Andreas Hasenack <andreas@conectiva.com.br> for reporting
442 the bugs to me.
443
444 4. Removed the suid-root bit on dump and restore in the
445 default build (and generated RPMs). It should be safer
446 now. Users who need the suid-root capabilities in order
447 to make network backups should read first the man page
448 of dump and enable the suid bit manually.
449
450 5. Added -ltermcap to the compile parameters for restore
451 when using readline support, in order to make the compile
452 process work on some systems (SuSE ?). Thanks to
453 Patrik Schindler <poc@pocnet.net> for reporting the bug.
454
455 Changes between versions 0.4b16 and 0.4b17 (released June 1st, 2000)
456 ====================================================================
457
458 1. The -F script is called now *only* at the end of a tape,
459 not at the end of the dump. Thanks to Christian Haul
460 <haul@informatik.tu-darmstadt.de> for the bug report.
461
462 Normally, the device name and the current volume number
463 are passed on the command line of the script. However,
464 if you want the old style script (with no arguments
465 passed on the command line), you can enable it in
466 configure with the --enable-oldstylefscript.
467
468 2. Use posix signal handling to preserve dump functionality
469 with libc5. Thanks to Jeff Johnson <jbj@redhat.com> for
470 the patch.
471
472 3. Made the exit code of restore in case of a 'C'ompare
473 command reflect the result of the compare. An exit status
474 of 0 means the dump archive is correct, 1 means tape errors,
475 2 means that some files were modified. Thanks to Kenneth Porter
476 <shiva@well.com> for the suggestion.
477
478 4. Made (finally) quotes work in interactive restore.
479
480 5. Small fixes in order to allow dump to compile with a
481 really old e2fsprogs version. Thanks to Ian Zimmerman
482 <itz@speakeasy.org> for the bug report.
483
484 6. Add GNU readline capabilities to interactive restore.
485 Use configure's --enable-readline flag to enable this feature.
486 Thanks to Patrik Schindler <poc@pocnet.net> for the
487 suggestion.
488
489 7. Do the compare on the fly in restore 'C' mode (this will
490 allow not to exhaust the available /tmp space when
491 ccmparing large files). Thanks to Kenneth Porter
492 <shiva@well.com> for the suggestion.
493
494 Changes between versions 0.4b15 and 0.4b16 (released March 11, 2000)
495 ====================================================================
496
497 1. Fixed some several duplicate 'DUMP: DUMP:' in the
498 output of dump.
499
500 2. Corrected the estimation of blocks to dump. Note that
501 this is still wrong for large directory inodes, due
502 to the size of a BSD directory entry on the tape
503 (estimation is based on the size of the ext2 directory,
504 which is smaller because it doesn't contain the null
505 character at the end).
506
507 3. Made dump report the total number of MB written to
508 tape at the end of run. Thanks to W. Reilly Cooley
509 <wcooley@nakedape.cc> for the patch.
510
511 4. Added the -X option to restore giving the possibility
512 to read the names of the files to be extracted/listed
513 from a text file (in addition of the command line).
514 Thanks to Dejan Muhamedagic <dejan@quant-x.com> for the
515 patch.
516
517 5. Added the device name and the current volume number
518 as arguments to the end of tape script (dump -F option).
519
520 6. Made the multi-volume dump work again (after having
521 broken it in 0.4b15).
522
523 Changes between versions 0.4b14 and 0.4b15 (released March 2, 2000)
524 ===================================================================
525
526 1. Added a prompt command in interactive restore mode. Thanks
527 to Andreas Dilger <adilger@home.com> for the patch.
528
529 2. Fixed a buffer overflow problem in dump (caused by
530 not checking the size of the filesystem parameter).
531 Thanks to Kim Yong-jun <loveyou@hackerslab.org> for
532 reporting this on Bugtraq (and to several dump users
533 who forwarded me his mail).
534
535 3. Added the '-F script' option to dump in order to
536 launch a script at the end of each tape (to be used
537 with a tape changer, or to notify the sysadmin by
538 pager etc.).
539
540 4. Fixed a bug in restore compare code caused by the changes
541 I made in 0.4b14.
542
543 5. Fixed the treatment of options using the old BSD syntax
544 in both dump and restore.
545
546 Changes between versions 0.4b13 and 0.4b14 (released February 10, 2000)
547 =======================================================================
548
549 1. Fixed a bug in dump which may have caused invalid deleted
550 directories to be dumped out if they were deleted after the
551 mapping pass. This could occure on active filesystem and lead
552 to heap corruption (causing dump malfunction in many possible ways).
553 Thanks to Jan Sanislo <oystr@cs.washington.edu> for finding this
554 bug and submitting the patch.
555
556 2. Fixed the handling of the filesystem parameter in dump. A
557 '/mnt/disk' parameter caused the disk contents to be dumped,
558 but a '/mnt/disk/' parameter caused the mountpoint directory
559 to be dumped (generally an empty directory).
560
561 3. Improved the output of dump in order to tell which directory
562 it is currently dumping (when dumping a subtree).
563
564 4. Added the '-e' exclude inode option to dump. Thanks to
565 Isaac Chuang <ike@isl.stanford.edu> for contributing with the patch.
566
567 5. Added a REPORTING-BUGS file in order to provide a guide
568 on how to correctly report a bug in dump/restore.
569
570 6. Fixed a restore problem when restoring a hard link to an inode
571 having the immutable or append-only attribute set. Thanks to
572 Ambrose Li <acli@mingpaoxpress.com> for submitting the patch.
573
574 7. Fixed a compatibility problem between dumps made on little
575 endian machines (the format was correct) and big endian
576 machines (the format was incorrect). This fix break the
577 compatibility with the older dumps made on big endian
578 machines (sparc, m86k, ppc etc). For the first time in
579 linux dump's history (I believe), the dumps made by *BSD,
580 Linux/alpha, Linux/sparc and Linux/x86 are compatible,
581 so interchangeable. Thanks to Rob Cermak
582 <cermak@ahab.rutgers.edu> for submitting the bug and
583 helping me test the fix.
584
585 8. Fixed the way dump reports the remaining percent/time, if
586 the number of blocks actually dumped exceeds the estimated
587 number of blocks. Thanks to Jean-Paul van der Jagt
588 <jeanpaul@dutepp0.et.tudelft.nl> for reporting the bug.
589
590 Changes between versions 0.4b12 and 0.4b13 (released January 21, 2000)
591 ======================================================================
592
593 1. Small Y2K fix in the man pages :). Thanks to Bernhard Sadlowski
594 <sadlowsk@Mathematik.Uni-Bielefeld.DE> for reporting the bug.
595
596 2. Removed the requirement to build the RPM as root from the
597 spec file. Thanks to Christian Weisgerber
598 <naddy@mips.rhein-neckar.de> for submitting this.
599
600 3. Fixed a bug in dump related to the 'filetype' feature of ext2fs,
601 causing dump to block when dumping really huge filesystems.
602 Many thanks to Patrik Schindler <poc@pocnet.net> for
603 helping me find this bug.
604
605 4. Fixed the treatment for an interrupt signal when dump access
606 the remote tape through RSH. Thanks to Christian Weisgerber
607 <naddy@mips.rhein-neckar.de> for providing the patch.
608
609 5. Fixed a bug which was causing dump/restore to display
610 garbage characters instead of the remote host name.
611
612 Changes between versions 0.4b11 and 0.4b12 (released January 8, 2000)
613 =====================================================================
614
615 1. Small fix in the dump man page. Thanks to Thorsten Kukuk
616 <kukuk@suse.de> for submitting the patch.
617
618 2. Fix for the exit code when using the size estimate option of
619 dump. Thanks to Matti Taskinen <mkt@rni.helsinki.fi> for
620 submitting the patch.
621
622 3. Handle EINTR in atomical reads/writes in dump, which was causing
623 dump to fail on some systems. Thanks to Eric Jergensen
624 <eric@dvns.com> for reporting the bug and submitting the patch.
625
626 4. Handle more than 16 characters for the device names in dumpdates.
627 (up to 255 now). Thanks to Rainer Clasen <bj@ncc.cicely.de> for
628 tracking down the problem and proposing the solution.
629
630 5. Fixed a bug in dump which prevented the creation of the
631 dumpdates file when doing a 0-level dump without already
632 having a dumpdates file. Thanks to Patrik Schindler
633 <poc@pocnet.net> for reporting the bug.
634
635 6. Changed the way dump 'S' flag reports the size estimate
636 from number of blocks into bytes (making it compatible
637 with the Solaris version, and simplifying things for
638 amanda users). Thanks to Jason L Tibbitts III
639 <tibbs@math.uh.edu> for reporting the bug.
640
641 7. Fixed a compatibility problem in linux/alpha dump tape format.
642 Now the linux/alpha dump are (again) compatible with the
643 other dump formats. But this breaks compatibility with
644 older dumps made on alpha. Thanks to Mike Tibor
645 <tibor@lib.uaa.alaska.edu> for helping me in finding this bug.
646
647 Changes between versions 0.4b10 and 0.4b11 (released December 5, 1999)
648 ======================================================================
649
650 1. Added a '--enable-kerberos' to configure.
651
652 2. Added a 'S' option to dump which determines the amount of space
653 that is needed to perform the dump without actually doing it, similar
654 to the Sun's ufsdump 'S' option. Patch contributed by Rob Cermak
655 <cermak@ahab.rutgers.edu>.
656
657 3. Added a 'M' multi-volume option to dump and restore which enables
658 dumping to multiple files (useful when dumping to an ext2
659 partition to make several dump files in order to bypass the 2GB
660 file size limitation). The 'f' argument is treated as a prefix and
661 the output files will be named <prefix>001, <prefix>002 etc. With
662 the 'M' flag, restore automatically selects the right file without
663 asking to enter a new tape each time.
664
665 4. Fixed a memory leak which was causing dump to grow very big
666 (270MB when dumping a 10GB filesystem...). Thanks to Jason
667 Fearon <jasonf@netrider.org.au> for reporting the bug.
668
669 Changes between versions 0.4b9 and 0.4b10 (released November 21, 1999)
670 ======================================================================
671
672 1. Make configure test if the system glob routines support
673 extended syntax (ALTDIRFUNC). If not, use the internal glob
674 routines instead of system ones. Thanks to Bernhard Sadlowski
675 <sadlowsk@Mathematik.Uni-Bielefeld.DE> for reporting the bug
676 and helping me resolve this and other minor libc5 compiling
677 glitches.
678
679 2. Fix a problem when dumping a ext2fs with the 'filetype'
680 feature enabled. Thanks to Patrick J. LoPresti
681 <patl@cag.lcs.mit.edu> for reporting the bug and to
682 Theodore Y. Ts'o <tytso@mit.edu> for providing the patch.
683
684 3. Made the nodump flag work on directories. A directory which
685 has the nodump flag gets never dumped, regardless of its
686 contents.
687
688 4. Integrate a patch from Jeremy Fitzhardinge <jeremy@goop.org>
689 which allows dump on an active ext3 filesystem. However, this
690 is a "quick and dirty" patch which enables backup of an ext3
691 filesystem through the ext2 compatibility (by ignoring the
692 NEEDS_RECOVERY bit). The journal file is not recognized and
693 it is dumped (it should not).
694
695 5. Test the superblock compatibility flags when dumping, in order
696 to be sure that we know how to deal with specific features.
697
698 Changes between versions 0.4b8 and 0.4b9 (released November 5, 1999)
699 ====================================================================
700
701 1. Use lchown instead of chown, fixing a possible security problem
702 when restoring symlinks (a malicious user could use this
703 to deliberately corrupt the ownership of important system files).
704 Thanks to Chris Siebenmann <cks@utcc.utoronto.ca> for detecting
705 this and providing the patch.
706
707 Changes between versions 0.4b7 and 0.4b8 (released November 3, 1999)
708 ====================================================================
709
710 1. Put dump sources under CVS, added Id tags in all files so
711 one can use 'ident' on binary files.
712
713 2. Added the dump/restore version in the usage text so one can
714 easily verify the version he is using.
715
716 3. Small patch from Nuno Oliveira <nuno@eq.uc.pt> which fixes
717 a va_start/va_end problem on linux-ppc (always call va_start
718 va_end in pairs each time we use a vararg function).
719
720 4. Added again the DT_* constants because old libc does not
721 contain them :(. Thanks to Eric Maisonobe <virnet@nat.fr>
722 for submitting the bug report.
723
724 5. Use ext2fs_llseek instead of llseek. With recent e2fsprogs
725 this should enable dumping big (huge) filesystems.
726
727 6. Added the RSH environment variable in order to be able to
728 use a rsh replacement like ssh when doing remote backups (and
729 bypass the security limitations of rcmd). Now you can do remote
730 backups without being root (or making dump setuid root).
731
732 7. Modified again the way dumpdates works. For incremental dumps,
733 we need to read dumpdates even if we are not using 'u' option.
734 Thanks to Bdale Garbee <bdale@gag.com> for his ideas on how
735 this should work.
736
737 Changes between versions 0.4b6 and 0.4b7 (released October 8, 1999)
738 ===================================================================
739
740 1. Removed the 'k' flag from the restore 'about' text if kerberos
741 was not compiled in.
742
743 2. Prototyped (f)setflags from e2fsprogs and corrected the calls
744 to them (fsetflags takes a char*, setflags an open fd!).
745
746 3. (f)setflags is called only if the flags aren't empty. If the
747 file is a special file, a warning is printed, because changing
748 flags implies opening the device. Normally, a special file
749 should not have any flag... (Debian bug #29775, patch provided
750 by Abhijit Dasgupta <abhijit@ans.net>).
751
752 4. Made possible to dump a file system not mentioned in /etc/fstab.
753 (Debian bug #11904, patch provided by Eirik Fuller <eirik@netcom.com>).
754
755 5. Changed the default behaviour to not create dumpdates
756 unless 'u' option is specified. Removed the old "debian-patch"
757 which provided the same thing. (Debian bug #38136, #33818).
758
759 6. Removed all those dump*announce, since they were getting old...
760
761 7. Added warning messages if dumpdates does not exist and
762 when an inferior level dump does not exist (except for a level 0
763 dump).
764
765 8. Debugged the glob calls in interactive mode: restore used a
766 dirent struct which was different from the /usr/include/dirent.h
767 one (this used to work, is it a glibc2 change?), so none of the
768 compat glob (which used /usr/include/dirent.h) or the system glob
769 worked. Restore use now the system dirent (and the system
770 DT_* constants), which are compatible with BSD ones.
771
772 9. Added a configure flag (--with-dumpdatespath) to specify
773 the location of dumpdates file. By default, it is
774 /etc/dumpdates.
775
776 10. Added the "AUTHOR" and "AVAILABILITY" sections and
777 included the current date/version in man pages.
778
779 11. Corrected the estimation of remaining time when
780 the operator doesn't change the tapes quickly enough. This
781 was an old bug, I thought I corrected it, and discovered
782 that in fact it was corrected in two different places, so
783 the results canceled each other...
784
785 Changes between versions 0.4b5 and 0.4b6 (released October 1, 1999)
786 ===================================================================
787
788 1. Integrated multiple patches from RedHat, Debian and SuSE:
789
790 - tweak dump/itime.c to not try to read dumpdates if the 'u' option
791 isn't specified.
792 - several fixes in the man pages.
793 - update the default tape device to /dev/st0.
794 - many updates for Linux Alpha (byte ordering, size_t etc).
795 - buffer overruns.
796 - use environment variable for TMPDIR (instead of /tmp).
797 - use sigjmp_buf instead of jmp_buf (RedHat bug #3260).
798 - workaround egcs bug (RedHat bugs #4281 and #2989).
799 - wire $(OPT) throughout Makefile's.
800
801 2. Upgrade the dump revision to 1, making possible to dump filesystems
802 made with e2fsprogs-1.15 or newer. Nothing seems to break...
803
804 3. Fix some compile warnings, prototype all functions.
805
806 4. Use glibc err/glob instead of internal compatibility
807 routines (only if available).
808
809 5. Fix a compile error on Linux 2.2.7 / libc5 (5.4.44) (patch provided
810 by Bernhard Sadlowski <sadlowsk@mathematik.uni-bielefeld.de>).
811
812 Changes between versions 0.4b4 and 0.4b5 (released September 22, 1999)
813 ======================================================================
814
815 1. Integrated the changes from FreeBSD-3.1-RELEASE
816 (mostly bug fixes, buffer overruns, dump has now an "automatic
817 tape length calculation" flag, dump/restore can use kerberos now
818 (this is NOT tested), use environment variables for TAPE and
819 RMT etc.).
820
821 2. Integrated three RedHat patches ("glibc", "kernel" and "bread" patches)
822
823 3. Corrected a bug in restore when using 'C' option with multi-volumes
824 tapes (files splited accros two tapes give "size changed" errors
825 when comparing).
826
827 4. Corrected the long standing bug when dumping multiple tapes.
828 This works for me, needs further testing.
829
830 Changes between versions 0.4b3 and 0.4b4 (released January 17, 1997)
831 ====================================================================
832
833 1. Dump now runs correctly on kernels 2.1.x
834 Fix made by Gerald Peters <gapeters@worldnet.att.net>
835
836 Changes between versions 0.4b2 and 0.4b3
837 ========================================
838
839 1. Use realpath() if available
840
841 2. Report statistics
842
843 Changes between versions 0.4b1 and 0.4b2
844 ========================================
845
846 1. Fixed the bug fix from Greg Lutz (I had made a mistake when integrating
847 the patch)
848
849 2. Fixed restore to make it able to read FreeBSD 2.x dumps again
850
851 3. Fixed configure.in to correctly handle --enable-rmt
852
853 Changes between versions 0.3 and 0.4b1
854 ======================================
855
856 1. Integrated the changes from 4.4BSD-Lite2
857
858 2. Integrated the patches from Debian and Red Hat
859
860 3. Portability changes: use the __u32, __u16, __s32, and __s16 types
861
862 4. Changed dump to use the Ext2fs library to get block addresses. This
863 should solve the endianness problem on SparcLinux.
864
865 5. Created a configure.in file (shamelessly stolen from the e2fsprogs
866 distribution's one) to use autoconf
867
868 6. Fixed a few minor bugs
869
870 Changes between versions 0.2e and 0.2f
871 ======================================
872
873 1. Added the creation of named pipes (fifos) in restore.
874
875 2. Added the -N flag in the restore manual page.
876
877 3. Added the file kernel.patch which contains the llseek() optimization
878 patch for 1.2.x kernels.
879
880 4. Fixed a bug in the restoration of symbolic links: owner and group were
881 not restored.
882
883 5. Integrated some changes from FreeBSD 2.2.
884
885 6. Added a call to ftruncate() after restoring each file to restore
886 correctly files ending by a hole.
887
888 Changes between versions 0.2d and 0.2e
889 ======================================
890
891 1. Fixed a bug in the "set owner/mode" process. Every file was restored
892 with owner = root (0) and group = root/wheel/whatever (0).
893
894 Changes between versions 0.2c and 0.2d
895 ======================================
896
897 1. Dump is now able to backup 2GB+ filesystems.
898
899 2. Dump and restore can now be linked as static binaries.
900
901 Changes between versions 0.2b and 0.2c
902 ======================================
903
904 1. Fixed a bug when dumping ``slow'' (i.e. normal) symbolic links.
905
906 Changes between versions 0.2a and 0.2b
907 ======================================
908
909 1. Really fixed the bug that I should have corrected in 0.2a.
910
911 2. Enabled optimization again.
912
913 Changes between versions 0.2 and 0.2a
914 =====================================
915
916 1. Disabled the optimization during compilation.
917
918 Changes between versions 0.1 and 0.2
919 ====================================
920
921 1. Fixed a bug in fstab.c which caused a null pointer to be stored in
922 the fs_type field (actually, I modified the file fstab.c to make it
923 use the mntent functions).
924
925 2. Dump and restore now use a 4.3 BSD compatible dump format. Backups
926 made by dump should be readable by the BSD restore and backups made
927 by the BSD dump should be readable by restore. Unfortunately, this
928 means that the dump format has changed between version 0.1 and version
929 0.2 :-(
930
931 3. Dump is now able to backup a subtree, it is no longer limited to whole
932 filesystems like the BSD version.
933
934 4. Dump now uses ext2_llseek() so it is able to backup filesystems bigger
935 than 2 GB.
936
937 Changes between versions 0.0 and 0.1
938 ====================================
939
940 1. Now create links rdump and rrestore during the `make install' step.
941
942 2. Linux port specific bugs added to the manual pages
943
944 3. Incorrect estimation of the number of tapes blocks fixed when doing
945 incremental backups.
946
947 4. Better ls-like format in restore in interactive mode.