]> git.wh0rd.org - dump.git/blob - CHANGES
755423f67fa321f2e67983b2502a009e8906f64d
[dump.git] / CHANGES
1 $Id: CHANGES,v 1.140 2001/10/02 09:41:24 stelian Exp $
2
3 Changes between versions 0.4b24 and 0.4b25 (released ??????????????????)
4 ========================================================================
5
6 1. Added a mini howto from Patrick Walsh in the examples
7 directory.
8
9 Changes between versions 0.4b23 and 0.4b24 (released September 12, 2001)
10 ========================================================================
11
12 1. Fixed the permissions of a newly created QFA file by dump.
13
14 2. Cleaned up the source of dump (the external variables
15 definition was a complete mess, making possible to have
16 objects overlap).
17
18 3. Fixed restore to use the full tape volume path when doing
19 a compare (since it changes the working directory to the
20 filesystem being compared in the process).
21
22 4. Added the -q option to dump which makes dump abort
23 whenever it needs operator attention. It should be
24 easier to use dump in scripts with this option.
25
26 5. Detect the use of incompatible options to dump and
27 refuse them (like -a and -B options together).
28
29 6. Added bzip2 compression to dump/restore (use option -j level
30 to select it). Note that you will need a restore version
31 >= 0.4b24 in order to restore a bzip2 compressed dump.
32 The same warning as for the zlib compression applies:
33 the tape format of a bzip2 dump is not compatible with the
34 original BSD tape format.
35
36 7. Fixed a overflow problem in dump corrupting the dump when
37 very large files were encountered. Thanks to Vladimir Nadvornik
38 for the bug report.
39
40 8. Added a ioctl(BLKFLSBUF) in dump which should flush the
41 kernel buffer/page cache before starting the dump, helping
42 a bit those who use dump on mounted filesystems. Thanks to
43 John Yu <jky@it.bu.edu> and to Theodore T'so <tytso@mit.edu>
44 for this suggestion.
45
46 9. Updated the RPM spec file following the RedHat changes
47 (dynamically linked binaries now in /usr/sbin etc).
48
49 10. Added a patch from Helmut Jarausch <jarausch@igpm.rwth-aachen.de>
50 which enables restore to recognise multi volume compressed dumps
51 done on CD. Included his scripts for dump (which pipe the dump
52 output directly into cdrecord) and restore. There is now
53 possible to to backups to CD on the fly!
54
55 Changes between versions 0.4b22 and 0.4b23 (released July 20, 2001)
56 ===================================================================
57
58 1. Fixed a buffer overflow in restore/tape.c. Patch provided by
59 Marcus Meissner (Caldera International Security Dept.).
60
61 2. Implement the Sun rmt extended protocol. Patch submitted
62 by Ian Gordon <iangordon@users.sourceforge.net>.
63
64 3. Check for the e2fsprogs header <ext2fs/ext2_fs.h> instead of
65 the linux kernel header. This ensures that dump always has the
66 latest version of this file. Patch submitted by
67 Andreas Dilger <adilger@turbolinux.com>.
68
69 4. Report any filesystem present in either /etc/fstab with a
70 positive passno or /etc/dumpdates in dump -w output.
71 Patch submitted by Andreas Dilger <adilger@turbolinux.com>.
72
73 5. Fixed the looping problem in dump introduced in the
74 previous version.
75
76 6. Changed the -B option of dump to limit the size of
77 _compressed_ output per volume if compression is on.
78 Patch contributed by Helmut Jarausch
79 <jarausch@igpm.rwth-aachen.de>. Note however that, since
80 it is impossible to predict the size of the compressed
81 data before doing the compression, the -B limit is a bit
82 conservative.
83
84 7. Fixed a bug in reading the operator typed file/tape path for
85 the next volume in restore.
86
87 8. Implemented a "-F script" option to restore which permits the
88 user to specify a script which will be launched at the
89 beginning of each tape, useful for automatic programming of
90 tape changers for example. See the restore man page for the
91 script parameters and return codes.
92
93 9. Small fix for the QFA routines provided by Uwe Gohlke
94 <uwe@ugsoft.de>, and some recommendations for QFA uses in
95 the man pages.
96
97 10. Fixed the multivolume restoring where making a mistake
98 to the 'Mount next tape' prompt caused several blocks to
99 be lost.
100
101 11. Enhanced the -e option of dump to take as a parameter a
102 comma separated list of inode numbers.
103
104 12. Added the -E option to dump which specify a file containing
105 inode numbers to exclude from the dump.
106
107 13. Fixed the compressed multi-volume dump + restore.
108
109 Changes between versions 0.4b21 and 0.4b22 (released May 12, 2001)
110 ==================================================================
111
112 1. Made dump report the number of blocks dumped per volume.
113 Thanks to Kenneth Porter <shiva@well.com> for the suggestion.
114
115 2. Fix a bug in dump where waiting too much at the 'change volumes'
116 question caused the volume to be rewritten. Thanks to
117 Graham Reed <greed@users.sourceforge.net> for reporting the
118 bug and providing a patch.
119
120 3. Added a compression option to dump, contributed by Jerry
121 Peters <gapeters@worldnet.att.net>.
122
123 WARNING: the compression works only on files/pipes and
124 drives supporting variable block size.
125
126 WARNING2: tapes created using the compression feature are
127 incompatible with the standard dump tape format, and a
128 version of restore >= 0.4b22 will be needed for extraction.
129
130 4. Fixed some compilation problems for glibc 2.2.2 and 64 bit
131 architectures. Thanks to Paul Slootman <paul@debian.org> for
132 the patch and to Bdale Garbee <bdale@gag.com> for forwarding
133 it upstream.
134
135 5. Many cleanups (CPP defines, const char warnings, check of
136 ext2fs COMPAT flags, time_t cleanups, added libext2 version
137 in dump usage text) by Andreas Dilger <adilger@turbolinux.com>.
138
139 6. Made --prefix option work in configure. All the install path
140 are now based on the configure parameters.
141
142 7. Added the Quick File Access mode in dump/restore, contributed
143 by Uwe Gohlke <uwe@ugsoft.de>. In this mode, dump stores in
144 a file tape position for each inode, and this file is used by
145 restore (if called with parameter Q and the filename)
146 to directly position the tape at the file restore is currently
147 working on. This saves hours when restoring single files from
148 large backups, saves the tapes and the drive's head. Use
149 --enable-qfa option of configure to compile in the QFA support.
150
151 8. Added the possibility to dump several files and directories
152 in a single invocation of dump. Thanks to Uwe Gohlke
153 <uwe@ugsoft.de> for implementing this option.
154
155 9. Fixed the dumping and restoring of files with holes
156 (files ending with holes didn't get dumped or restored
157 correctly).
158
159 10. Fixed a socket descriptor leak in dump, which leaved opened
160 3 file descriptors per dump process (and there is one dump
161 process per tape).
162
163 11. Fixed dump large file system support, by explicit use of
164 open64/lseek64/etc functions (explicit use needed because
165 e2fsprogs libraries don't behave well when compiled with
166 FILE_OFFSET_BITS=64).
167
168 Changes between versions 0.4b20 and 0.4b21 (released January 13, 2001)
169 ======================================================================
170
171 1. Fixed some bugs in the dump -w|-W logic introduced by
172 the previous version. Thanks to Andreas Dilger
173 <adilger@turbolinux.com> for his help on this one.
174
175 2. Fixed again a compilation problem when using old e2fs
176 headers (filesystem label related). Thanks to many users
177 who reported this stupid error.
178
179 3. Fixed a build problem on old lib5 based systems dealing with
180 _PATH_MNTTAB being undefined. Thanks to John Adams
181 <johna@onevista.com> for reporting the bug.
182
183 4. Improved the error detection in restore when creating
184 the temporary files in TMPDIR. Now restore will corectly
185 report a 'no space left on device' error instead of
186 strange errors which could imply an archive corruption.
187 Thanks to Gabriel Somlo <somlo@cs.colostate.edu> and
188 bgjenero <bgjenero@sympatico.ca> for reporting the bug.
189
190 5. Added the throughput information in dump's progression
191 messages. Thanks to Andreas Dilger <adilger@turbolinux.com>
192 for the patch.
193
194 6. Use libext2fs's inode scanning routines, which are particularly
195 robust when dealing with errors (especially when having some
196 bad blocks on the filesystem). Many thanks to Colin
197 <colin@horizon.com> for reporting the bug, investigating
198 the issues, finding a workaround, writing the patch and
199 fully testing it... (of course, if this patch will break
200 anything, he is to blame for :-)).
201
202 7. Made dump and restore LFS aware. Dump can dump a filesystem
203 containing large files, generate a large file on output and
204 restore can restore them. This feature is not enabled by
205 default, use --enable-largefile option of configure to enable
206 it (you need to have a LFS aware glibc though). Thanks to
207 Andrea Arcangeli <andrea@suse.de> for submitting the patch,
208 and to Theodore T'so <tytso@valinux.com> for his always
209 useful thoughts.
210
211 8. Made dump ask upon a tape write error if it should rewrite
212 the current volume (assume this is a bad tape) or if it should
213 assume an end-of-tape condition (useful for tape drives which
214 behaves badly at the end of the tape). Thanks to Andreas
215 Dilger <adilger@turbolinux.com> for the suggestion.
216
217 Changes between versions 0.4b19 and 0.4b20 (released November 10, 2000)
218 =======================================================================
219
220 1. Fixed a small compilation problem due to a change
221 in the definintion of the struct sigaction in
222 glibc 2.0 / libc5. Thanks to Gunther Schlegel
223 <schlegel@riege.de> for reporting the bug and to
224 Dave Platt <dplatt@snulbug.mtview.ca.us> for suggesting
225 a fix.
226
227 2. Modified the label and uuid analysis in order to be
228 self-contained (independant of kernel/libc headers). This
229 should fix the compile with older kernel/libc headers and
230 will preserve the functionality. Thanks to Bernhard Erdmann
231 <bernhard.erdmann@gmx.de> for reporting the bug.
232
233 3. The 'exclude inode' option, if applied to a directory
234 inode, excludes now correctly all the leaves of this
235 directory. Thanks to John R. Dennison
236 <gerdesas@users.sourceforge.net> for reporting the bug.
237
238 4. Fixed the '-e' option to disable the possibility
239 to exclude the root inode (which causes the dumps to
240 be unreadable by restore). Prevented array overflow
241 when multiple -e options are used.
242
243 5. Fixed dump to correctly interpret a filesystem argument
244 which is a mountpoint and it is not an absolute path
245 (as specified in the fstab). Thanks to Bernhard R. Erdmann
246 <be@berdmann.de> for reporting the bug.
247
248 6. Made dump able to backup files larger than 2 GB. Note that
249 dump still doesn't cope with files larger than 4 GB.
250
251 7. Restore the real uid and gid before invoking an external
252 RSH program (big hole when dump or restore is suid root!).
253
254 8. Get the values for the default tape device and the location
255 of fstab file from the system headers. Thanks to
256 Andreas Dilger <adilger@turbolinux.com> for the patch.
257
258 9. Made dump -w|-W report all recognized filesystems
259 present in either /etc/fstab or /etc/dumpdates, and present
260 the list in the same order as in fstab file. Thanks
261 to Andreas Dilger <adilger@turbolinux.com> for the patch.
262
263 10. Made dump's -a (automatic end of tape detection) the
264 default. Specifying one of -B, -c, -d or -s options will
265 override the EOM detection. Thanks to Andreas Dilger
266 <adilger@turbolinux.com> for the patch.
267
268 11. Save the ext2 filesystem label into the dump volume label.
269 Specifying a volume label on the command line (-L option)
270 will override this feature. Thanks to Andreas Dilger
271 <adilger@turbolinux.com> for the patch.
272
273 Changes between versions 0.4b18 and 0.4b19 (released August 20, 2000)
274 =====================================================================
275
276 1. Fixed the signal handling in dump (which I broke in 0.4b17)
277 which was causing several strange dump failures (dump
278 hanged or segmentation faults etc).
279
280 2. Specified the default blocksize in the dump man page.
281
282 3. Changed two info messages of restore to be written on stdout
283 instead of stderr in order to leave stderr only for errors
284 or warnings. Thanks to Stephen Atwell
285 <satwell@urbana.css.mot.com> for the suggestion.
286
287 4. Corrected an off by one calculation which prevented
288 dumping a directory having a 1 character name length.
289 Thanks to Bernhard Erdmann <bernhard.erdmann@gmx.de>
290 for reporting the bug.
291
292 5. Reinforce dump in order to allow only level 0 and no
293 -u option when dumping a subdirectory, and document
294 this in the man page. Thanks to Bernhard Erdmann
295 <bernhard.erdmann@gmx.de> for reporting the bug.
296
297 6. Fixed a small harmless bug in restore which caused
298 in some conditions hard links to be restored several
299 times (therefore generation some warning messages).
300 Thanks to Kenneth Porter <shiva@well.com> for
301 reporting the bug.
302
303 7. Updated the RPM spec file to the latest RedHat version,
304 providing FHS packaging and other cosmetic changes.
305 You will need RPM version >= 3.0.5 in order to build the RPM.
306
307 8. Updated the configure script to check for libtermcap
308 before checking for libreadline, because we need this
309 library in order to compile the readline support.
310
311 9. Made dump understand the LABEL= and UUID= notation
312 both in /etc/fstab and on the command line. Note that
313 you will need the /proc filesystem in order to use
314 these notations. Thanks to Erik Troan <ewt@redhat.com>
315 for providing the patch.
316
317 Changes between versions 0.4b17 and 0.4b18 (released June 30, 2000)
318 ===================================================================
319
320 1. Fixed a potential buffer overflow in restore. Thanks
321 to Stan Bubrouski <satan@fastdial.net> for reporting
322 the bug.
323
324 2. Fixed a readline-related bug which prevented
325 'cat DUMPFILE | restore -i -f -' from working. Thanks
326 to Charles Karney <karney@users.sourceforge.net>
327 for the bug report.
328
329 3. Changed a few "panic" into "exit", causing restore to
330 be more stable against some attacks (like the last one
331 reported on Bugtraq, although the last version of restore
332 was not vulnerable - just dumped core). Thanks to
333 Andreas Hasenack <andreas@conectiva.com.br> for reporting
334 the bugs to me.
335
336 4. Removed the suid-root bit on dump and restore in the
337 default build (and generated RPMs). It should be safer
338 now. Users who need the suid-root capabilities in order
339 to make network backups should read first the man page
340 of dump and enable the suid bit manually.
341
342 5. Added -ltermcap to the compile parameters for restore
343 when using readline support, in order to make the compile
344 process work on some systems (SuSE ?). Thanks to
345 Patrik Schindler <poc@pocnet.net> for reporting the bug.
346
347 Changes between versions 0.4b16 and 0.4b17 (released June 1st, 2000)
348 ====================================================================
349
350 1. The -F script is called now *only* at the end of a tape,
351 not at the end of the dump. Thanks to Christian Haul
352 <haul@informatik.tu-darmstadt.de> for the bug report.
353
354 Normally, the device name and the current volume number
355 are passed on the command line of the script. However,
356 if you want the old style script (with no arguments
357 passed on the command line), you can enable it in
358 configure with the --enable-oldstylefscript.
359
360 2. Use posix signal handling to preserve dump functionality
361 with libc5. Thanks to Jeff Johnson <jbj@redhat.com> for
362 the patch.
363
364 3. Made the exit code of restore in case of a 'C'ompare
365 command reflect the result of the compare. An exit status
366 of 0 means the dump archive is correct, 1 means tape errors,
367 2 means that some files were modified. Thanks to Kenneth Porter
368 <shiva@well.com> for the suggestion.
369
370 4. Made (finally) quotes work in interactive restore.
371
372 5. Small fixes in order to allow dump to compile with a
373 really old e2fsprogs version. Thanks to Ian Zimmerman
374 <itz@speakeasy.org> for the bug report.
375
376 6. Add GNU readline capabilities to interactive restore.
377 Use configure's --enable-readline flag to enable this feature.
378 Thanks to Patrik Schindler <poc@pocnet.net> for the
379 suggestion.
380
381 7. Do the compare on the fly in restore 'C' mode (this will
382 allow not to exhaust the available /tmp space when
383 ccmparing large files). Thanks to Kenneth Porter
384 <shiva@well.com> for the suggestion.
385
386 Changes between versions 0.4b15 and 0.4b16 (released March 11, 2000)
387 ====================================================================
388
389 1. Fixed some several duplicate 'DUMP: DUMP:' in the
390 output of dump.
391
392 2. Corrected the estimation of blocks to dump. Note that
393 this is still wrong for large directory inodes, due
394 to the size of a BSD directory entry on the tape
395 (estimation is based on the size of the ext2 directory,
396 which is smaller because it doesn't contain the null
397 character at the end).
398
399 3. Made dump report the total number of MB written to
400 tape at the end of run. Thanks to W. Reilly Cooley
401 <wcooley@nakedape.cc> for the patch.
402
403 4. Added the -X option to restore giving the possibility
404 to read the names of the files to be extracted/listed
405 from a text file (in addition of the command line).
406 Thanks to Dejan Muhamedagic <dejan@quant-x.com> for the
407 patch.
408
409 5. Added the device name and the current volume number
410 as arguments to the end of tape script (dump -F option).
411
412 6. Made the multi-volume dump work again (after having
413 broken it in 0.4b15).
414
415 Changes between versions 0.4b14 and 0.4b15 (released March 2, 2000)
416 ===================================================================
417
418 1. Added a prompt command in interactive restore mode. Thanks
419 to Andreas Dilger <adilger@home.com> for the patch.
420
421 2. Fixed a buffer overflow problem in dump (caused by
422 not checking the size of the filesystem parameter).
423 Thanks to Kim Yong-jun <loveyou@hackerslab.org> for
424 reporting this on Bugtraq (and to several dump users
425 who forwarded me his mail).
426
427 3. Added the '-F script' option to dump in order to
428 launch a script at the end of each tape (to be used
429 with a tape changer, or to notify the sysadmin by
430 pager etc.).
431
432 4. Fixed a bug in restore compare code caused by the changes
433 I made in 0.4b14.
434
435 5. Fixed the treatment of options using the old BSD syntax
436 in both dump and restore.
437
438 Changes between versions 0.4b13 and 0.4b14 (released February 10, 2000)
439 =======================================================================
440
441 1. Fixed a bug in dump which may have caused invalid deleted
442 directories to be dumped out if they were deleted after the
443 mapping pass. This could occure on active filesystem and lead
444 to heap corruption (causing dump malfunction in many possible ways).
445 Thanks to Jan Sanislo <oystr@cs.washington.edu> for finding this
446 bug and submitting the patch.
447
448 2. Fixed the handling of the filesystem parameter in dump. A
449 '/mnt/disk' parameter caused the disk contents to be dumped,
450 but a '/mnt/disk/' parameter caused the mountpoint directory
451 to be dumped (generally an empty directory).
452
453 3. Improved the output of dump in order to tell which directory
454 it is currently dumping (when dumping a subtree).
455
456 4. Added the '-e' exclude inode option to dump. Thanks to
457 Isaac Chuang <ike@isl.stanford.edu> for contributing with the patch.
458
459 5. Added a REPORTING-BUGS file in order to provide a guide
460 on how to correctly report a bug in dump/restore.
461
462 6. Fixed a restore problem when restoring a hard link to an inode
463 having the immutable or append-only attribute set. Thanks to
464 Ambrose Li <acli@mingpaoxpress.com> for submitting the patch.
465
466 7. Fixed a compatibility problem between dumps made on little
467 endian machines (the format was correct) and big endian
468 machines (the format was incorrect). This fix break the
469 compatibility with the older dumps made on big endian
470 machines (sparc, m86k, ppc etc). For the first time in
471 linux dump's history (I believe), the dumps made by *BSD,
472 Linux/alpha, Linux/sparc and Linux/x86 are compatible,
473 so interchangeable. Thanks to Rob Cermak
474 <cermak@ahab.rutgers.edu> for submitting the bug and
475 helping me test the fix.
476
477 8. Fixed the way dump reports the remaining percent/time, if
478 the number of blocks actually dumped exceeds the estimated
479 number of blocks. Thanks to Jean-Paul van der Jagt
480 <jeanpaul@dutepp0.et.tudelft.nl> for reporting the bug.
481
482 Changes between versions 0.4b12 and 0.4b13 (released January 21, 2000)
483 ======================================================================
484
485 1. Small Y2K fix in the man pages :). Thanks to Bernhard Sadlowski
486 <sadlowsk@Mathematik.Uni-Bielefeld.DE> for reporting the bug.
487
488 2. Removed the requirement to build the RPM as root from the
489 spec file. Thanks to Christian Weisgerber
490 <naddy@mips.rhein-neckar.de> for submitting this.
491
492 3. Fixed a bug in dump related to the 'filetype' feature of ext2fs,
493 causing dump to block when dumping really huge filesystems.
494 Many thanks to Patrik Schindler <poc@pocnet.net> for
495 helping me find this bug.
496
497 4. Fixed the treatment for an interrupt signal when dump access
498 the remote tape through RSH. Thanks to Christian Weisgerber
499 <naddy@mips.rhein-neckar.de> for providing the patch.
500
501 5. Fixed a bug which was causing dump/restore to display
502 garbage characters instead of the remote host name.
503
504 Changes between versions 0.4b11 and 0.4b12 (released January 8, 2000)
505 =====================================================================
506
507 1. Small fix in the dump man page. Thanks to Thorsten Kukuk
508 <kukuk@suse.de> for submitting the patch.
509
510 2. Fix for the exit code when using the size estimate option of
511 dump. Thanks to Matti Taskinen <mkt@rni.helsinki.fi> for
512 submitting the patch.
513
514 3. Handle EINTR in atomical reads/writes in dump, which was causing
515 dump to fail on some systems. Thanks to Eric Jergensen
516 <eric@dvns.com> for reporting the bug and submitting the patch.
517
518 4. Handle more than 16 characters for the device names in dumpdates.
519 (up to 255 now). Thanks to Rainer Clasen <bj@ncc.cicely.de> for
520 tracking down the problem and proposing the solution.
521
522 5. Fixed a bug in dump which prevented the creation of the
523 dumpdates file when doing a 0-level dump without already
524 having a dumpdates file. Thanks to Patrik Schindler
525 <poc@pocnet.net> for reporting the bug.
526
527 6. Changed the way dump 'S' flag reports the size estimate
528 from number of blocks into bytes (making it compatible
529 with the Solaris version, and simplifying things for
530 amanda users). Thanks to Jason L Tibbitts III
531 <tibbs@math.uh.edu> for reporting the bug.
532
533 7. Fixed a compatibility problem in linux/alpha dump tape format.
534 Now the linux/alpha dump are (again) compatible with the
535 other dump formats. But this breaks compatibility with
536 older dumps made on alpha. Thanks to Mike Tibor
537 <tibor@lib.uaa.alaska.edu> for helping me in finding this bug.
538
539 Changes between versions 0.4b10 and 0.4b11 (released December 5, 1999)
540 ======================================================================
541
542 1. Added a '--enable-kerberos' to configure.
543
544 2. Added a 'S' option to dump which determines the amount of space
545 that is needed to perform the dump without actually doing it, similar
546 to the Sun's ufsdump 'S' option. Patch contributed by Rob Cermak
547 <cermak@ahab.rutgers.edu>.
548
549 3. Added a 'M' multi-volume option to dump and restore which enables
550 dumping to multiple files (useful when dumping to an ext2
551 partition to make several dump files in order to bypass the 2GB
552 file size limitation). The 'f' argument is treated as a prefix and
553 the output files will be named <prefix>001, <prefix>002 etc. With
554 the 'M' flag, restore automatically selects the right file without
555 asking to enter a new tape each time.
556
557 4. Fixed a memory leak which was causing dump to grow very big
558 (270MB when dumping a 10GB filesystem...). Thanks to Jason
559 Fearon <jasonf@netrider.org.au> for reporting the bug.
560
561 Changes between versions 0.4b9 and 0.4b10 (released November 21, 1999)
562 ======================================================================
563
564 1. Make configure test if the system glob routines support
565 extended syntax (ALTDIRFUNC). If not, use the internal glob
566 routines instead of system ones. Thanks to Bernhard Sadlowski
567 <sadlowsk@Mathematik.Uni-Bielefeld.DE> for reporting the bug
568 and helping me resolve this and other minor libc5 compiling
569 glitches.
570
571 2. Fix a problem when dumping a ext2fs with the 'filetype'
572 feature enabled. Thanks to Patrick J. LoPresti
573 <patl@cag.lcs.mit.edu> for reporting the bug and to
574 Theodore Y. Ts'o <tytso@mit.edu> for providing the patch.
575
576 3. Made the nodump flag work on directories. A directory which
577 has the nodump flag gets never dumped, regardless of its
578 contents.
579
580 4. Integrate a patch from Jeremy Fitzhardinge <jeremy@goop.org>
581 which allows dump on an active ext3 filesystem. However, this
582 is a "quick and dirty" patch which enables backup of an ext3
583 filesystem through the ext2 compatibility (by ignoring the
584 NEEDS_RECOVERY bit). The journal file is not recognized and
585 it is dumped (it should not).
586
587 5. Test the superblock compatibility flags when dumping, in order
588 to be sure that we know how to deal with specific features.
589
590 Changes between versions 0.4b8 and 0.4b9 (released November 5, 1999)
591 ====================================================================
592
593 1. Use lchown instead of chown, fixing a possible security problem
594 when restoring symlinks (a malicious user could use this
595 to deliberately corrupt the ownership of important system files).
596 Thanks to Chris Siebenmann <cks@utcc.utoronto.ca> for detecting
597 this and providing the patch.
598
599 Changes between versions 0.4b7 and 0.4b8 (released November 3, 1999)
600 ====================================================================
601
602 1. Put dump sources under CVS, added Id tags in all files so
603 one can use 'ident' on binary files.
604
605 2. Added the dump/restore version in the usage text so one can
606 easily verify the version he is using.
607
608 3. Small patch from Nuno Oliveira <nuno@eq.uc.pt> which fixes
609 a va_start/va_end problem on linux-ppc (always call va_start
610 va_end in pairs each time we use a vararg function).
611
612 4. Added again the DT_* constants because old libc does not
613 contain them :(. Thanks to Eric Maisonobe <virnet@nat.fr>
614 for submitting the bug report.
615
616 5. Use ext2fs_llseek instead of llseek. With recent e2fsprogs
617 this should enable dumping big (huge) filesystems.
618
619 6. Added the RSH environment variable in order to be able to
620 use a rsh replacement like ssh when doing remote backups (and
621 bypass the security limitations of rcmd). Now you can do remote
622 backups without being root (or making dump setuid root).
623
624 7. Modified again the way dumpdates works. For incremental dumps,
625 we need to read dumpdates even if we are not using 'u' option.
626 Thanks to Bdale Garbee <bdale@gag.com> for his ideas on how
627 this should work.
628
629 Changes between versions 0.4b6 and 0.4b7 (released October 8, 1999)
630 ===================================================================
631
632 1. Removed the 'k' flag from the restore 'about' text if kerberos
633 was not compiled in.
634
635 2. Prototyped (f)setflags from e2fsprogs and corrected the calls
636 to them (fsetflags takes a char*, setflags an open fd!).
637
638 3. (f)setflags is called only if the flags aren't empty. If the
639 file is a special file, a warning is printed, because changing
640 flags implies opening the device. Normally, a special file
641 should not have any flag... (Debian bug #29775, patch provided
642 by Abhijit Dasgupta <abhijit@ans.net>).
643
644 4. Made possible to dump a file system not mentioned in /etc/fstab.
645 (Debian bug #11904, patch provided by Eirik Fuller <eirik@netcom.com>).
646
647 5. Changed the default behaviour to not create dumpdates
648 unless 'u' option is specified. Removed the old "debian-patch"
649 which provided the same thing. (Debian bug #38136, #33818).
650
651 6. Removed all those dump*announce, since they were getting old...
652
653 7. Added warning messages if dumpdates does not exist and
654 when an inferior level dump does not exist (except for a level 0
655 dump).
656
657 8. Debugged the glob calls in interactive mode: restore used a
658 dirent struct which was different from the /usr/include/dirent.h
659 one (this used to work, is it a glibc2 change?), so none of the
660 compat glob (which used /usr/include/dirent.h) or the system glob
661 worked. Restore use now the system dirent (and the system
662 DT_* constants), which are compatible with BSD ones.
663
664 9. Added a configure flag (--with-dumpdatespath) to specify
665 the location of dumpdates file. By default, it is
666 /etc/dumpdates.
667
668 10. Added the "AUTHOR" and "AVAILABILITY" sections and
669 included the current date/version in man pages.
670
671 11. Corrected the estimation of remaining time when
672 the operator doesn't change the tapes quickly enough. This
673 was an old bug, I thought I corrected it, and discovered
674 that in fact it was corrected in two different places, so
675 the results canceled each other...
676
677 Changes between versions 0.4b5 and 0.4b6 (released October 1, 1999)
678 ===================================================================
679
680 1. Integrated multiple patches from RedHat, Debian and SuSE:
681
682 - tweak dump/itime.c to not try to read dumpdates if the 'u' option
683 isn't specified.
684 - several fixes in the man pages.
685 - update the default tape device to /dev/st0.
686 - many updates for Linux Alpha (byte ordering, size_t etc).
687 - buffer overruns.
688 - use environment variable for TMPDIR (instead of /tmp).
689 - use sigjmp_buf instead of jmp_buf (RedHat bug #3260).
690 - workaround egcs bug (RedHat bugs #4281 and #2989).
691 - wire $(OPT) throughout Makefile's.
692
693 2. Upgrade the dump revision to 1, making possible to dump filesystems
694 made with e2fsprogs-1.15 or newer. Nothing seems to break...
695
696 3. Fix some compile warnings, prototype all functions.
697
698 4. Use glibc err/glob instead of internal compatibility
699 routines (only if available).
700
701 5. Fix a compile error on Linux 2.2.7 / libc5 (5.4.44) (patch provided
702 by Bernhard Sadlowski <sadlowsk@mathematik.uni-bielefeld.de>).
703
704 Changes between versions 0.4b4 and 0.4b5 (released September 22, 1999)
705 ======================================================================
706
707 1. Integrated the changes from FreeBSD-3.1-RELEASE
708 (mostly bug fixes, buffer overruns, dump has now an "automatic
709 tape length calculation" flag, dump/restore can use kerberos now
710 (this is NOT tested), use environment variables for TAPE and
711 RMT etc.).
712
713 2. Integrated three RedHat patches ("glibc", "kernel" and "bread" patches)
714
715 3. Corrected a bug in restore when using 'C' option with multi-volumes
716 tapes (files splited accros two tapes give "size changed" errors
717 when comparing).
718
719 4. Corrected the long standing bug when dumping multiple tapes.
720 This works for me, needs further testing.
721
722 Changes between versions 0.4b3 and 0.4b4 (released January 17, 1997)
723 ====================================================================
724
725 1. Dump now runs correctly on kernels 2.1.x
726 Fix made by Gerald Peters <gapeters@worldnet.att.net>
727
728 Changes between versions 0.4b2 and 0.4b3
729 ========================================
730
731 1. Use realpath() if available
732
733 2. Report statistics
734
735 Changes between versions 0.4b1 and 0.4b2
736 ========================================
737
738 1. Fixed the bug fix from Greg Lutz (I had made a mistake when integrating
739 the patch)
740
741 2. Fixed restore to make it able to read FreeBSD 2.x dumps again
742
743 3. Fixed configure.in to correctly handle --enable-rmt
744
745 Changes between versions 0.3 and 0.4b1
746 ======================================
747
748 1. Integrated the changes from 4.4BSD-Lite2
749
750 2. Integrated the patches from Debian and Red Hat
751
752 3. Portability changes: use the __u32, __u16, __s32, and __s16 types
753
754 4. Changed dump to use the Ext2fs library to get block addresses. This
755 should solve the endianness problem on SparcLinux.
756
757 5. Created a configure.in file (shamelessly stolen from the e2fsprogs
758 distribution's one) to use autoconf
759
760 6. Fixed a few minor bugs
761
762 Changes between versions 0.2e and 0.2f
763 ======================================
764
765 1. Added the creation of named pipes (fifos) in restore.
766
767 2. Added the -N flag in the restore manual page.
768
769 3. Added the file kernel.patch which contains the llseek() optimization
770 patch for 1.2.x kernels.
771
772 4. Fixed a bug in the restoration of symbolic links: owner and group were
773 not restored.
774
775 5. Integrated some changes from FreeBSD 2.2.
776
777 6. Added a call to ftruncate() after restoring each file to restore
778 correctly files ending by a hole.
779
780 Changes between versions 0.2d and 0.2e
781 ======================================
782
783 1. Fixed a bug in the "set owner/mode" process. Every file was restored
784 with owner = root (0) and group = root/wheel/whatever (0).
785
786 Changes between versions 0.2c and 0.2d
787 ======================================
788
789 1. Dump is now able to backup 2GB+ filesystems.
790
791 2. Dump and restore can now be linked as static binaries.
792
793 Changes between versions 0.2b and 0.2c
794 ======================================
795
796 1. Fixed a bug when dumping ``slow'' (i.e. normal) symbolic links.
797
798 Changes between versions 0.2a and 0.2b
799 ======================================
800
801 1. Really fixed the bug that I should have corrected in 0.2a.
802
803 2. Enabled optimization again.
804
805 Changes between versions 0.2 and 0.2a
806 =====================================
807
808 1. Disabled the optimization during compilation.
809
810 Changes between versions 0.1 and 0.2
811 ====================================
812
813 1. Fixed a bug in fstab.c which caused a null pointer to be stored in
814 the fs_type field (actually, I modified the file fstab.c to make it
815 use the mntent functions).
816
817 2. Dump and restore now use a 4.3 BSD compatible dump format. Backups
818 made by dump should be readable by the BSD restore and backups made
819 by the BSD dump should be readable by restore. Unfortunately, this
820 means that the dump format has changed between version 0.1 and version
821 0.2 :-(
822
823 3. Dump is now able to backup a subtree, it is no longer limited to whole
824 filesystems like the BSD version.
825
826 4. Dump now uses ext2_llseek() so it is able to backup filesystems bigger
827 than 2 GB.
828
829 Changes between versions 0.0 and 0.1
830 ====================================
831
832 1. Now create links rdump and rrestore during the `make install' step.
833
834 2. Linux port specific bugs added to the manual pages
835
836 3. Incorrect estimation of the number of tapes blocks fixed when doing
837 incremental backups.
838
839 4. Better ls-like format in restore in interactive mode.