]> git.wh0rd.org - dump.git/blob - CHANGES
Added -a flag to restore (automatic walk through volumes)
[dump.git] / CHANGES
1 $Id: CHANGES,v 1.156 2002/01/16 10:29:23 stelian Exp $
2
3 Changes between versions 0.4b26 and 0.4b27 (released ???????????????)
4 =====================================================================
5
6 1. Fixed behaviour of dump when exceeding resource limits
7 (SIGXFSZ treatment).
8
9 2. Added the -L flag to restore to allow the user to specify a
10 maximal allowed number of miscompares when using restore
11 with the -C option to check the backup.
12
13 3. Detailed the manual entry for the -N option of restore.
14
15 4. Added the -a flag to restore to make able doing unattended
16 restores in -i or -x mode (automatically walks through the
17 multiple dump volumes).
18
19 Changes between versions 0.4b25 and 0.4b26 (released January 7, 2002)
20 =====================================================================
21
22 1. Added a set of backup scripts from Eugenio Diaz
23 <getnito@yahoo.com> in the examples section. It features
24 automatic (cron based) full and incremental dumping of
25 several filesystems on a separate filesystem.
26
27 2. Fixed a off-by-one miscalculation which disabled dumping
28 a one letter subdirectory. Thanks to Chris Siebenmann
29 <cks@utcc.utoronto.ca> for reporting the bug.
30
31 3. Fixed several restore bugs occuring when trying to
32 restore some missing files on the tape. Thanks to Chris
33 Siebenmann <cks@utcc.utoronto.ca> for reporting the bug.
34
35 4. Fixed --with-ldopts configure argument passing, installing from
36 a separate object directory, makefile cleanups contributed
37 by <splite@purdue.edu>.
38
39 5. Fix a bug which could caused, in some conditions, the highest
40 number inode of a filesystem, to not be dumped. Many thanks
41 to Chris Siebenmann <cks@utcc.utoronto.ca> for helping me
42 chase this bug.
43
44 Changes between versions 0.4b24 and 0.4b25 (released November 17, 2001)
45 =======================================================================
46
47 1. Added a mini howto from Patrick Walsh in the examples
48 directory.
49
50 2. Minor man pages syntax corrections. Thanks to
51 Chris Danis <screechco@home.com> for reporting the bugs.
52
53 3. Added a script from David B. Peterson <dave@toppledwagon.com>
54 to the examples section. It features dumping several
55 filesystems to a remote tape drive upon ssh.
56
57 4. Added a patch provided by Richard Jones <rich@annexia.org>
58 which allows BRADEMAX (number of read errors tolerated by
59 dump) to be adjusted using the -I option.
60
61 5. Fixed a bug which disabled doing "restore -C -f -". Thanks
62 to Clemens Stiglechner <a9401816@unet.univie.ac.at> for the
63 patch.
64
65 6. Add the -l option to restore to specify if, when doing a
66 remote restore, the file used to access the backup is a
67 regular file (the defaults being a tape drive). Restore needs
68 to know this information when reading a remote compressed
69 dump. Previously, this information was autodetected, but
70 the autodetection code fails (with ioctl: Inappropriate ioctl
71 for device) when using a non Linux remote box. Thanks to
72 many users and especially to Eros Albertazzi
73 <eros@lamel.bo.cnr.it> for reporting this.
74
75 7. Found a workaroung for the dump deadlock problem (3 childs
76 stuck in pause(), father in read()). The workaround seems
77 to work for me and several beta-testers. If it doesn't work
78 for you, please report back.
79
80 8. Updated the RPM spec file (BuildPrereq, URL etc).
81
82 Changes between versions 0.4b23 and 0.4b24 (released September 12, 2001)
83 ========================================================================
84
85 1. Fixed the permissions of a newly created QFA file by dump.
86
87 2. Cleaned up the source of dump (the external variables
88 definition was a complete mess, making possible to have
89 objects overlap).
90
91 3. Fixed restore to use the full tape volume path when doing
92 a compare (since it changes the working directory to the
93 filesystem being compared in the process).
94
95 4. Added the -q option to dump which makes dump abort
96 whenever it needs operator attention. It should be
97 easier to use dump in scripts with this option.
98
99 5. Detect the use of incompatible options to dump and
100 refuse them (like -a and -B options together).
101
102 6. Added bzip2 compression to dump/restore (use option -j level
103 to select it). Note that you will need a restore version
104 >= 0.4b24 in order to restore a bzip2 compressed dump.
105 The same warning as for the zlib compression applies:
106 the tape format of a bzip2 dump is not compatible with the
107 original BSD tape format.
108
109 7. Fixed a overflow problem in dump corrupting the dump when
110 very large files were encountered. Thanks to Vladimir Nadvornik
111 for the bug report.
112
113 8. Added a ioctl(BLKFLSBUF) in dump which should flush the
114 kernel buffer/page cache before starting the dump, helping
115 a bit those who use dump on mounted filesystems. Thanks to
116 John Yu <jky@it.bu.edu> and to Theodore T'so <tytso@mit.edu>
117 for this suggestion.
118
119 9. Updated the RPM spec file following the RedHat changes
120 (dynamically linked binaries now in /usr/sbin etc).
121
122 10. Added a patch from Helmut Jarausch <jarausch@igpm.rwth-aachen.de>
123 which enables restore to recognise multi volume compressed dumps
124 done on CD. Included his scripts for dump (which pipe the dump
125 output directly into cdrecord) and restore. There is now
126 possible to to backups to CD on the fly!
127
128 Changes between versions 0.4b22 and 0.4b23 (released July 20, 2001)
129 ===================================================================
130
131 1. Fixed a buffer overflow in restore/tape.c. Patch provided by
132 Marcus Meissner (Caldera International Security Dept.).
133
134 2. Implement the Sun rmt extended protocol. Patch submitted
135 by Ian Gordon <iangordon@users.sourceforge.net>.
136
137 3. Check for the e2fsprogs header <ext2fs/ext2_fs.h> instead of
138 the linux kernel header. This ensures that dump always has the
139 latest version of this file. Patch submitted by
140 Andreas Dilger <adilger@turbolinux.com>.
141
142 4. Report any filesystem present in either /etc/fstab with a
143 positive passno or /etc/dumpdates in dump -w output.
144 Patch submitted by Andreas Dilger <adilger@turbolinux.com>.
145
146 5. Fixed the looping problem in dump introduced in the
147 previous version.
148
149 6. Changed the -B option of dump to limit the size of
150 _compressed_ output per volume if compression is on.
151 Patch contributed by Helmut Jarausch
152 <jarausch@igpm.rwth-aachen.de>. Note however that, since
153 it is impossible to predict the size of the compressed
154 data before doing the compression, the -B limit is a bit
155 conservative.
156
157 7. Fixed a bug in reading the operator typed file/tape path for
158 the next volume in restore.
159
160 8. Implemented a "-F script" option to restore which permits the
161 user to specify a script which will be launched at the
162 beginning of each tape, useful for automatic programming of
163 tape changers for example. See the restore man page for the
164 script parameters and return codes.
165
166 9. Small fix for the QFA routines provided by Uwe Gohlke
167 <uwe@ugsoft.de>, and some recommendations for QFA uses in
168 the man pages.
169
170 10. Fixed the multivolume restoring where making a mistake
171 to the 'Mount next tape' prompt caused several blocks to
172 be lost.
173
174 11. Enhanced the -e option of dump to take as a parameter a
175 comma separated list of inode numbers.
176
177 12. Added the -E option to dump which specify a file containing
178 inode numbers to exclude from the dump.
179
180 13. Fixed the compressed multi-volume dump + restore.
181
182 Changes between versions 0.4b21 and 0.4b22 (released May 12, 2001)
183 ==================================================================
184
185 1. Made dump report the number of blocks dumped per volume.
186 Thanks to Kenneth Porter <shiva@well.com> for the suggestion.
187
188 2. Fix a bug in dump where waiting too much at the 'change volumes'
189 question caused the volume to be rewritten. Thanks to
190 Graham Reed <greed@users.sourceforge.net> for reporting the
191 bug and providing a patch.
192
193 3. Added a compression option to dump, contributed by Jerry
194 Peters <gapeters@worldnet.att.net>.
195
196 WARNING: the compression works only on files/pipes and
197 drives supporting variable block size.
198
199 WARNING2: tapes created using the compression feature are
200 incompatible with the standard dump tape format, and a
201 version of restore >= 0.4b22 will be needed for extraction.
202
203 4. Fixed some compilation problems for glibc 2.2.2 and 64 bit
204 architectures. Thanks to Paul Slootman <paul@debian.org> for
205 the patch and to Bdale Garbee <bdale@gag.com> for forwarding
206 it upstream.
207
208 5. Many cleanups (CPP defines, const char warnings, check of
209 ext2fs COMPAT flags, time_t cleanups, added libext2 version
210 in dump usage text) by Andreas Dilger <adilger@turbolinux.com>.
211
212 6. Made --prefix option work in configure. All the install path
213 are now based on the configure parameters.
214
215 7. Added the Quick File Access mode in dump/restore, contributed
216 by Uwe Gohlke <uwe@ugsoft.de>. In this mode, dump stores in
217 a file tape position for each inode, and this file is used by
218 restore (if called with parameter Q and the filename)
219 to directly position the tape at the file restore is currently
220 working on. This saves hours when restoring single files from
221 large backups, saves the tapes and the drive's head. Use
222 --enable-qfa option of configure to compile in the QFA support.
223
224 8. Added the possibility to dump several files and directories
225 in a single invocation of dump. Thanks to Uwe Gohlke
226 <uwe@ugsoft.de> for implementing this option.
227
228 9. Fixed the dumping and restoring of files with holes
229 (files ending with holes didn't get dumped or restored
230 correctly).
231
232 10. Fixed a socket descriptor leak in dump, which leaved opened
233 3 file descriptors per dump process (and there is one dump
234 process per tape).
235
236 11. Fixed dump large file system support, by explicit use of
237 open64/lseek64/etc functions (explicit use needed because
238 e2fsprogs libraries don't behave well when compiled with
239 FILE_OFFSET_BITS=64).
240
241 Changes between versions 0.4b20 and 0.4b21 (released January 13, 2001)
242 ======================================================================
243
244 1. Fixed some bugs in the dump -w|-W logic introduced by
245 the previous version. Thanks to Andreas Dilger
246 <adilger@turbolinux.com> for his help on this one.
247
248 2. Fixed again a compilation problem when using old e2fs
249 headers (filesystem label related). Thanks to many users
250 who reported this stupid error.
251
252 3. Fixed a build problem on old lib5 based systems dealing with
253 _PATH_MNTTAB being undefined. Thanks to John Adams
254 <johna@onevista.com> for reporting the bug.
255
256 4. Improved the error detection in restore when creating
257 the temporary files in TMPDIR. Now restore will corectly
258 report a 'no space left on device' error instead of
259 strange errors which could imply an archive corruption.
260 Thanks to Gabriel Somlo <somlo@cs.colostate.edu> and
261 bgjenero <bgjenero@sympatico.ca> for reporting the bug.
262
263 5. Added the throughput information in dump's progression
264 messages. Thanks to Andreas Dilger <adilger@turbolinux.com>
265 for the patch.
266
267 6. Use libext2fs's inode scanning routines, which are particularly
268 robust when dealing with errors (especially when having some
269 bad blocks on the filesystem). Many thanks to Colin
270 <colin@horizon.com> for reporting the bug, investigating
271 the issues, finding a workaround, writing the patch and
272 fully testing it... (of course, if this patch will break
273 anything, he is to blame for :-)).
274
275 7. Made dump and restore LFS aware. Dump can dump a filesystem
276 containing large files, generate a large file on output and
277 restore can restore them. This feature is not enabled by
278 default, use --enable-largefile option of configure to enable
279 it (you need to have a LFS aware glibc though). Thanks to
280 Andrea Arcangeli <andrea@suse.de> for submitting the patch,
281 and to Theodore T'so <tytso@valinux.com> for his always
282 useful thoughts.
283
284 8. Made dump ask upon a tape write error if it should rewrite
285 the current volume (assume this is a bad tape) or if it should
286 assume an end-of-tape condition (useful for tape drives which
287 behaves badly at the end of the tape). Thanks to Andreas
288 Dilger <adilger@turbolinux.com> for the suggestion.
289
290 Changes between versions 0.4b19 and 0.4b20 (released November 10, 2000)
291 =======================================================================
292
293 1. Fixed a small compilation problem due to a change
294 in the definintion of the struct sigaction in
295 glibc 2.0 / libc5. Thanks to Gunther Schlegel
296 <schlegel@riege.de> for reporting the bug and to
297 Dave Platt <dplatt@snulbug.mtview.ca.us> for suggesting
298 a fix.
299
300 2. Modified the label and uuid analysis in order to be
301 self-contained (independant of kernel/libc headers). This
302 should fix the compile with older kernel/libc headers and
303 will preserve the functionality. Thanks to Bernhard Erdmann
304 <bernhard.erdmann@gmx.de> for reporting the bug.
305
306 3. The 'exclude inode' option, if applied to a directory
307 inode, excludes now correctly all the leaves of this
308 directory. Thanks to John R. Dennison
309 <gerdesas@users.sourceforge.net> for reporting the bug.
310
311 4. Fixed the '-e' option to disable the possibility
312 to exclude the root inode (which causes the dumps to
313 be unreadable by restore). Prevented array overflow
314 when multiple -e options are used.
315
316 5. Fixed dump to correctly interpret a filesystem argument
317 which is a mountpoint and it is not an absolute path
318 (as specified in the fstab). Thanks to Bernhard R. Erdmann
319 <be@berdmann.de> for reporting the bug.
320
321 6. Made dump able to backup files larger than 2 GB. Note that
322 dump still doesn't cope with files larger than 4 GB.
323
324 7. Restore the real uid and gid before invoking an external
325 RSH program (big hole when dump or restore is suid root!).
326
327 8. Get the values for the default tape device and the location
328 of fstab file from the system headers. Thanks to
329 Andreas Dilger <adilger@turbolinux.com> for the patch.
330
331 9. Made dump -w|-W report all recognized filesystems
332 present in either /etc/fstab or /etc/dumpdates, and present
333 the list in the same order as in fstab file. Thanks
334 to Andreas Dilger <adilger@turbolinux.com> for the patch.
335
336 10. Made dump's -a (automatic end of tape detection) the
337 default. Specifying one of -B, -c, -d or -s options will
338 override the EOM detection. Thanks to Andreas Dilger
339 <adilger@turbolinux.com> for the patch.
340
341 11. Save the ext2 filesystem label into the dump volume label.
342 Specifying a volume label on the command line (-L option)
343 will override this feature. Thanks to Andreas Dilger
344 <adilger@turbolinux.com> for the patch.
345
346 Changes between versions 0.4b18 and 0.4b19 (released August 20, 2000)
347 =====================================================================
348
349 1. Fixed the signal handling in dump (which I broke in 0.4b17)
350 which was causing several strange dump failures (dump
351 hanged or segmentation faults etc).
352
353 2. Specified the default blocksize in the dump man page.
354
355 3. Changed two info messages of restore to be written on stdout
356 instead of stderr in order to leave stderr only for errors
357 or warnings. Thanks to Stephen Atwell
358 <satwell@urbana.css.mot.com> for the suggestion.
359
360 4. Corrected an off by one calculation which prevented
361 dumping a directory having a 1 character name length.
362 Thanks to Bernhard Erdmann <bernhard.erdmann@gmx.de>
363 for reporting the bug.
364
365 5. Reinforce dump in order to allow only level 0 and no
366 -u option when dumping a subdirectory, and document
367 this in the man page. Thanks to Bernhard Erdmann
368 <bernhard.erdmann@gmx.de> for reporting the bug.
369
370 6. Fixed a small harmless bug in restore which caused
371 in some conditions hard links to be restored several
372 times (therefore generation some warning messages).
373 Thanks to Kenneth Porter <shiva@well.com> for
374 reporting the bug.
375
376 7. Updated the RPM spec file to the latest RedHat version,
377 providing FHS packaging and other cosmetic changes.
378 You will need RPM version >= 3.0.5 in order to build the RPM.
379
380 8. Updated the configure script to check for libtermcap
381 before checking for libreadline, because we need this
382 library in order to compile the readline support.
383
384 9. Made dump understand the LABEL= and UUID= notation
385 both in /etc/fstab and on the command line. Note that
386 you will need the /proc filesystem in order to use
387 these notations. Thanks to Erik Troan <ewt@redhat.com>
388 for providing the patch.
389
390 Changes between versions 0.4b17 and 0.4b18 (released June 30, 2000)
391 ===================================================================
392
393 1. Fixed a potential buffer overflow in restore. Thanks
394 to Stan Bubrouski <satan@fastdial.net> for reporting
395 the bug.
396
397 2. Fixed a readline-related bug which prevented
398 'cat DUMPFILE | restore -i -f -' from working. Thanks
399 to Charles Karney <karney@users.sourceforge.net>
400 for the bug report.
401
402 3. Changed a few "panic" into "exit", causing restore to
403 be more stable against some attacks (like the last one
404 reported on Bugtraq, although the last version of restore
405 was not vulnerable - just dumped core). Thanks to
406 Andreas Hasenack <andreas@conectiva.com.br> for reporting
407 the bugs to me.
408
409 4. Removed the suid-root bit on dump and restore in the
410 default build (and generated RPMs). It should be safer
411 now. Users who need the suid-root capabilities in order
412 to make network backups should read first the man page
413 of dump and enable the suid bit manually.
414
415 5. Added -ltermcap to the compile parameters for restore
416 when using readline support, in order to make the compile
417 process work on some systems (SuSE ?). Thanks to
418 Patrik Schindler <poc@pocnet.net> for reporting the bug.
419
420 Changes between versions 0.4b16 and 0.4b17 (released June 1st, 2000)
421 ====================================================================
422
423 1. The -F script is called now *only* at the end of a tape,
424 not at the end of the dump. Thanks to Christian Haul
425 <haul@informatik.tu-darmstadt.de> for the bug report.
426
427 Normally, the device name and the current volume number
428 are passed on the command line of the script. However,
429 if you want the old style script (with no arguments
430 passed on the command line), you can enable it in
431 configure with the --enable-oldstylefscript.
432
433 2. Use posix signal handling to preserve dump functionality
434 with libc5. Thanks to Jeff Johnson <jbj@redhat.com> for
435 the patch.
436
437 3. Made the exit code of restore in case of a 'C'ompare
438 command reflect the result of the compare. An exit status
439 of 0 means the dump archive is correct, 1 means tape errors,
440 2 means that some files were modified. Thanks to Kenneth Porter
441 <shiva@well.com> for the suggestion.
442
443 4. Made (finally) quotes work in interactive restore.
444
445 5. Small fixes in order to allow dump to compile with a
446 really old e2fsprogs version. Thanks to Ian Zimmerman
447 <itz@speakeasy.org> for the bug report.
448
449 6. Add GNU readline capabilities to interactive restore.
450 Use configure's --enable-readline flag to enable this feature.
451 Thanks to Patrik Schindler <poc@pocnet.net> for the
452 suggestion.
453
454 7. Do the compare on the fly in restore 'C' mode (this will
455 allow not to exhaust the available /tmp space when
456 ccmparing large files). Thanks to Kenneth Porter
457 <shiva@well.com> for the suggestion.
458
459 Changes between versions 0.4b15 and 0.4b16 (released March 11, 2000)
460 ====================================================================
461
462 1. Fixed some several duplicate 'DUMP: DUMP:' in the
463 output of dump.
464
465 2. Corrected the estimation of blocks to dump. Note that
466 this is still wrong for large directory inodes, due
467 to the size of a BSD directory entry on the tape
468 (estimation is based on the size of the ext2 directory,
469 which is smaller because it doesn't contain the null
470 character at the end).
471
472 3. Made dump report the total number of MB written to
473 tape at the end of run. Thanks to W. Reilly Cooley
474 <wcooley@nakedape.cc> for the patch.
475
476 4. Added the -X option to restore giving the possibility
477 to read the names of the files to be extracted/listed
478 from a text file (in addition of the command line).
479 Thanks to Dejan Muhamedagic <dejan@quant-x.com> for the
480 patch.
481
482 5. Added the device name and the current volume number
483 as arguments to the end of tape script (dump -F option).
484
485 6. Made the multi-volume dump work again (after having
486 broken it in 0.4b15).
487
488 Changes between versions 0.4b14 and 0.4b15 (released March 2, 2000)
489 ===================================================================
490
491 1. Added a prompt command in interactive restore mode. Thanks
492 to Andreas Dilger <adilger@home.com> for the patch.
493
494 2. Fixed a buffer overflow problem in dump (caused by
495 not checking the size of the filesystem parameter).
496 Thanks to Kim Yong-jun <loveyou@hackerslab.org> for
497 reporting this on Bugtraq (and to several dump users
498 who forwarded me his mail).
499
500 3. Added the '-F script' option to dump in order to
501 launch a script at the end of each tape (to be used
502 with a tape changer, or to notify the sysadmin by
503 pager etc.).
504
505 4. Fixed a bug in restore compare code caused by the changes
506 I made in 0.4b14.
507
508 5. Fixed the treatment of options using the old BSD syntax
509 in both dump and restore.
510
511 Changes between versions 0.4b13 and 0.4b14 (released February 10, 2000)
512 =======================================================================
513
514 1. Fixed a bug in dump which may have caused invalid deleted
515 directories to be dumped out if they were deleted after the
516 mapping pass. This could occure on active filesystem and lead
517 to heap corruption (causing dump malfunction in many possible ways).
518 Thanks to Jan Sanislo <oystr@cs.washington.edu> for finding this
519 bug and submitting the patch.
520
521 2. Fixed the handling of the filesystem parameter in dump. A
522 '/mnt/disk' parameter caused the disk contents to be dumped,
523 but a '/mnt/disk/' parameter caused the mountpoint directory
524 to be dumped (generally an empty directory).
525
526 3. Improved the output of dump in order to tell which directory
527 it is currently dumping (when dumping a subtree).
528
529 4. Added the '-e' exclude inode option to dump. Thanks to
530 Isaac Chuang <ike@isl.stanford.edu> for contributing with the patch.
531
532 5. Added a REPORTING-BUGS file in order to provide a guide
533 on how to correctly report a bug in dump/restore.
534
535 6. Fixed a restore problem when restoring a hard link to an inode
536 having the immutable or append-only attribute set. Thanks to
537 Ambrose Li <acli@mingpaoxpress.com> for submitting the patch.
538
539 7. Fixed a compatibility problem between dumps made on little
540 endian machines (the format was correct) and big endian
541 machines (the format was incorrect). This fix break the
542 compatibility with the older dumps made on big endian
543 machines (sparc, m86k, ppc etc). For the first time in
544 linux dump's history (I believe), the dumps made by *BSD,
545 Linux/alpha, Linux/sparc and Linux/x86 are compatible,
546 so interchangeable. Thanks to Rob Cermak
547 <cermak@ahab.rutgers.edu> for submitting the bug and
548 helping me test the fix.
549
550 8. Fixed the way dump reports the remaining percent/time, if
551 the number of blocks actually dumped exceeds the estimated
552 number of blocks. Thanks to Jean-Paul van der Jagt
553 <jeanpaul@dutepp0.et.tudelft.nl> for reporting the bug.
554
555 Changes between versions 0.4b12 and 0.4b13 (released January 21, 2000)
556 ======================================================================
557
558 1. Small Y2K fix in the man pages :). Thanks to Bernhard Sadlowski
559 <sadlowsk@Mathematik.Uni-Bielefeld.DE> for reporting the bug.
560
561 2. Removed the requirement to build the RPM as root from the
562 spec file. Thanks to Christian Weisgerber
563 <naddy@mips.rhein-neckar.de> for submitting this.
564
565 3. Fixed a bug in dump related to the 'filetype' feature of ext2fs,
566 causing dump to block when dumping really huge filesystems.
567 Many thanks to Patrik Schindler <poc@pocnet.net> for
568 helping me find this bug.
569
570 4. Fixed the treatment for an interrupt signal when dump access
571 the remote tape through RSH. Thanks to Christian Weisgerber
572 <naddy@mips.rhein-neckar.de> for providing the patch.
573
574 5. Fixed a bug which was causing dump/restore to display
575 garbage characters instead of the remote host name.
576
577 Changes between versions 0.4b11 and 0.4b12 (released January 8, 2000)
578 =====================================================================
579
580 1. Small fix in the dump man page. Thanks to Thorsten Kukuk
581 <kukuk@suse.de> for submitting the patch.
582
583 2. Fix for the exit code when using the size estimate option of
584 dump. Thanks to Matti Taskinen <mkt@rni.helsinki.fi> for
585 submitting the patch.
586
587 3. Handle EINTR in atomical reads/writes in dump, which was causing
588 dump to fail on some systems. Thanks to Eric Jergensen
589 <eric@dvns.com> for reporting the bug and submitting the patch.
590
591 4. Handle more than 16 characters for the device names in dumpdates.
592 (up to 255 now). Thanks to Rainer Clasen <bj@ncc.cicely.de> for
593 tracking down the problem and proposing the solution.
594
595 5. Fixed a bug in dump which prevented the creation of the
596 dumpdates file when doing a 0-level dump without already
597 having a dumpdates file. Thanks to Patrik Schindler
598 <poc@pocnet.net> for reporting the bug.
599
600 6. Changed the way dump 'S' flag reports the size estimate
601 from number of blocks into bytes (making it compatible
602 with the Solaris version, and simplifying things for
603 amanda users). Thanks to Jason L Tibbitts III
604 <tibbs@math.uh.edu> for reporting the bug.
605
606 7. Fixed a compatibility problem in linux/alpha dump tape format.
607 Now the linux/alpha dump are (again) compatible with the
608 other dump formats. But this breaks compatibility with
609 older dumps made on alpha. Thanks to Mike Tibor
610 <tibor@lib.uaa.alaska.edu> for helping me in finding this bug.
611
612 Changes between versions 0.4b10 and 0.4b11 (released December 5, 1999)
613 ======================================================================
614
615 1. Added a '--enable-kerberos' to configure.
616
617 2. Added a 'S' option to dump which determines the amount of space
618 that is needed to perform the dump without actually doing it, similar
619 to the Sun's ufsdump 'S' option. Patch contributed by Rob Cermak
620 <cermak@ahab.rutgers.edu>.
621
622 3. Added a 'M' multi-volume option to dump and restore which enables
623 dumping to multiple files (useful when dumping to an ext2
624 partition to make several dump files in order to bypass the 2GB
625 file size limitation). The 'f' argument is treated as a prefix and
626 the output files will be named <prefix>001, <prefix>002 etc. With
627 the 'M' flag, restore automatically selects the right file without
628 asking to enter a new tape each time.
629
630 4. Fixed a memory leak which was causing dump to grow very big
631 (270MB when dumping a 10GB filesystem...). Thanks to Jason
632 Fearon <jasonf@netrider.org.au> for reporting the bug.
633
634 Changes between versions 0.4b9 and 0.4b10 (released November 21, 1999)
635 ======================================================================
636
637 1. Make configure test if the system glob routines support
638 extended syntax (ALTDIRFUNC). If not, use the internal glob
639 routines instead of system ones. Thanks to Bernhard Sadlowski
640 <sadlowsk@Mathematik.Uni-Bielefeld.DE> for reporting the bug
641 and helping me resolve this and other minor libc5 compiling
642 glitches.
643
644 2. Fix a problem when dumping a ext2fs with the 'filetype'
645 feature enabled. Thanks to Patrick J. LoPresti
646 <patl@cag.lcs.mit.edu> for reporting the bug and to
647 Theodore Y. Ts'o <tytso@mit.edu> for providing the patch.
648
649 3. Made the nodump flag work on directories. A directory which
650 has the nodump flag gets never dumped, regardless of its
651 contents.
652
653 4. Integrate a patch from Jeremy Fitzhardinge <jeremy@goop.org>
654 which allows dump on an active ext3 filesystem. However, this
655 is a "quick and dirty" patch which enables backup of an ext3
656 filesystem through the ext2 compatibility (by ignoring the
657 NEEDS_RECOVERY bit). The journal file is not recognized and
658 it is dumped (it should not).
659
660 5. Test the superblock compatibility flags when dumping, in order
661 to be sure that we know how to deal with specific features.
662
663 Changes between versions 0.4b8 and 0.4b9 (released November 5, 1999)
664 ====================================================================
665
666 1. Use lchown instead of chown, fixing a possible security problem
667 when restoring symlinks (a malicious user could use this
668 to deliberately corrupt the ownership of important system files).
669 Thanks to Chris Siebenmann <cks@utcc.utoronto.ca> for detecting
670 this and providing the patch.
671
672 Changes between versions 0.4b7 and 0.4b8 (released November 3, 1999)
673 ====================================================================
674
675 1. Put dump sources under CVS, added Id tags in all files so
676 one can use 'ident' on binary files.
677
678 2. Added the dump/restore version in the usage text so one can
679 easily verify the version he is using.
680
681 3. Small patch from Nuno Oliveira <nuno@eq.uc.pt> which fixes
682 a va_start/va_end problem on linux-ppc (always call va_start
683 va_end in pairs each time we use a vararg function).
684
685 4. Added again the DT_* constants because old libc does not
686 contain them :(. Thanks to Eric Maisonobe <virnet@nat.fr>
687 for submitting the bug report.
688
689 5. Use ext2fs_llseek instead of llseek. With recent e2fsprogs
690 this should enable dumping big (huge) filesystems.
691
692 6. Added the RSH environment variable in order to be able to
693 use a rsh replacement like ssh when doing remote backups (and
694 bypass the security limitations of rcmd). Now you can do remote
695 backups without being root (or making dump setuid root).
696
697 7. Modified again the way dumpdates works. For incremental dumps,
698 we need to read dumpdates even if we are not using 'u' option.
699 Thanks to Bdale Garbee <bdale@gag.com> for his ideas on how
700 this should work.
701
702 Changes between versions 0.4b6 and 0.4b7 (released October 8, 1999)
703 ===================================================================
704
705 1. Removed the 'k' flag from the restore 'about' text if kerberos
706 was not compiled in.
707
708 2. Prototyped (f)setflags from e2fsprogs and corrected the calls
709 to them (fsetflags takes a char*, setflags an open fd!).
710
711 3. (f)setflags is called only if the flags aren't empty. If the
712 file is a special file, a warning is printed, because changing
713 flags implies opening the device. Normally, a special file
714 should not have any flag... (Debian bug #29775, patch provided
715 by Abhijit Dasgupta <abhijit@ans.net>).
716
717 4. Made possible to dump a file system not mentioned in /etc/fstab.
718 (Debian bug #11904, patch provided by Eirik Fuller <eirik@netcom.com>).
719
720 5. Changed the default behaviour to not create dumpdates
721 unless 'u' option is specified. Removed the old "debian-patch"
722 which provided the same thing. (Debian bug #38136, #33818).
723
724 6. Removed all those dump*announce, since they were getting old...
725
726 7. Added warning messages if dumpdates does not exist and
727 when an inferior level dump does not exist (except for a level 0
728 dump).
729
730 8. Debugged the glob calls in interactive mode: restore used a
731 dirent struct which was different from the /usr/include/dirent.h
732 one (this used to work, is it a glibc2 change?), so none of the
733 compat glob (which used /usr/include/dirent.h) or the system glob
734 worked. Restore use now the system dirent (and the system
735 DT_* constants), which are compatible with BSD ones.
736
737 9. Added a configure flag (--with-dumpdatespath) to specify
738 the location of dumpdates file. By default, it is
739 /etc/dumpdates.
740
741 10. Added the "AUTHOR" and "AVAILABILITY" sections and
742 included the current date/version in man pages.
743
744 11. Corrected the estimation of remaining time when
745 the operator doesn't change the tapes quickly enough. This
746 was an old bug, I thought I corrected it, and discovered
747 that in fact it was corrected in two different places, so
748 the results canceled each other...
749
750 Changes between versions 0.4b5 and 0.4b6 (released October 1, 1999)
751 ===================================================================
752
753 1. Integrated multiple patches from RedHat, Debian and SuSE:
754
755 - tweak dump/itime.c to not try to read dumpdates if the 'u' option
756 isn't specified.
757 - several fixes in the man pages.
758 - update the default tape device to /dev/st0.
759 - many updates for Linux Alpha (byte ordering, size_t etc).
760 - buffer overruns.
761 - use environment variable for TMPDIR (instead of /tmp).
762 - use sigjmp_buf instead of jmp_buf (RedHat bug #3260).
763 - workaround egcs bug (RedHat bugs #4281 and #2989).
764 - wire $(OPT) throughout Makefile's.
765
766 2. Upgrade the dump revision to 1, making possible to dump filesystems
767 made with e2fsprogs-1.15 or newer. Nothing seems to break...
768
769 3. Fix some compile warnings, prototype all functions.
770
771 4. Use glibc err/glob instead of internal compatibility
772 routines (only if available).
773
774 5. Fix a compile error on Linux 2.2.7 / libc5 (5.4.44) (patch provided
775 by Bernhard Sadlowski <sadlowsk@mathematik.uni-bielefeld.de>).
776
777 Changes between versions 0.4b4 and 0.4b5 (released September 22, 1999)
778 ======================================================================
779
780 1. Integrated the changes from FreeBSD-3.1-RELEASE
781 (mostly bug fixes, buffer overruns, dump has now an "automatic
782 tape length calculation" flag, dump/restore can use kerberos now
783 (this is NOT tested), use environment variables for TAPE and
784 RMT etc.).
785
786 2. Integrated three RedHat patches ("glibc", "kernel" and "bread" patches)
787
788 3. Corrected a bug in restore when using 'C' option with multi-volumes
789 tapes (files splited accros two tapes give "size changed" errors
790 when comparing).
791
792 4. Corrected the long standing bug when dumping multiple tapes.
793 This works for me, needs further testing.
794
795 Changes between versions 0.4b3 and 0.4b4 (released January 17, 1997)
796 ====================================================================
797
798 1. Dump now runs correctly on kernels 2.1.x
799 Fix made by Gerald Peters <gapeters@worldnet.att.net>
800
801 Changes between versions 0.4b2 and 0.4b3
802 ========================================
803
804 1. Use realpath() if available
805
806 2. Report statistics
807
808 Changes between versions 0.4b1 and 0.4b2
809 ========================================
810
811 1. Fixed the bug fix from Greg Lutz (I had made a mistake when integrating
812 the patch)
813
814 2. Fixed restore to make it able to read FreeBSD 2.x dumps again
815
816 3. Fixed configure.in to correctly handle --enable-rmt
817
818 Changes between versions 0.3 and 0.4b1
819 ======================================
820
821 1. Integrated the changes from 4.4BSD-Lite2
822
823 2. Integrated the patches from Debian and Red Hat
824
825 3. Portability changes: use the __u32, __u16, __s32, and __s16 types
826
827 4. Changed dump to use the Ext2fs library to get block addresses. This
828 should solve the endianness problem on SparcLinux.
829
830 5. Created a configure.in file (shamelessly stolen from the e2fsprogs
831 distribution's one) to use autoconf
832
833 6. Fixed a few minor bugs
834
835 Changes between versions 0.2e and 0.2f
836 ======================================
837
838 1. Added the creation of named pipes (fifos) in restore.
839
840 2. Added the -N flag in the restore manual page.
841
842 3. Added the file kernel.patch which contains the llseek() optimization
843 patch for 1.2.x kernels.
844
845 4. Fixed a bug in the restoration of symbolic links: owner and group were
846 not restored.
847
848 5. Integrated some changes from FreeBSD 2.2.
849
850 6. Added a call to ftruncate() after restoring each file to restore
851 correctly files ending by a hole.
852
853 Changes between versions 0.2d and 0.2e
854 ======================================
855
856 1. Fixed a bug in the "set owner/mode" process. Every file was restored
857 with owner = root (0) and group = root/wheel/whatever (0).
858
859 Changes between versions 0.2c and 0.2d
860 ======================================
861
862 1. Dump is now able to backup 2GB+ filesystems.
863
864 2. Dump and restore can now be linked as static binaries.
865
866 Changes between versions 0.2b and 0.2c
867 ======================================
868
869 1. Fixed a bug when dumping ``slow'' (i.e. normal) symbolic links.
870
871 Changes between versions 0.2a and 0.2b
872 ======================================
873
874 1. Really fixed the bug that I should have corrected in 0.2a.
875
876 2. Enabled optimization again.
877
878 Changes between versions 0.2 and 0.2a
879 =====================================
880
881 1. Disabled the optimization during compilation.
882
883 Changes between versions 0.1 and 0.2
884 ====================================
885
886 1. Fixed a bug in fstab.c which caused a null pointer to be stored in
887 the fs_type field (actually, I modified the file fstab.c to make it
888 use the mntent functions).
889
890 2. Dump and restore now use a 4.3 BSD compatible dump format. Backups
891 made by dump should be readable by the BSD restore and backups made
892 by the BSD dump should be readable by restore. Unfortunately, this
893 means that the dump format has changed between version 0.1 and version
894 0.2 :-(
895
896 3. Dump is now able to backup a subtree, it is no longer limited to whole
897 filesystems like the BSD version.
898
899 4. Dump now uses ext2_llseek() so it is able to backup filesystems bigger
900 than 2 GB.
901
902 Changes between versions 0.0 and 0.1
903 ====================================
904
905 1. Now create links rdump and rrestore during the `make install' step.
906
907 2. Linux port specific bugs added to the manual pages
908
909 3. Incorrect estimation of the number of tapes blocks fixed when doing
910 incremental backups.
911
912 4. Better ls-like format in restore in interactive mode.