]> git.wh0rd.org - dump.git/blob - CHANGES
Report any filesystem present in either /etc/fstab with a positive passno or /etc...
[dump.git] / CHANGES
1 $Id: CHANGES,v 1.118 2001/06/18 11:07:45 stelian Exp $
2
3 Changes between versions 0.4b22 and 0.4b23 (released ????????????)
4 ==================================================================
5
6 1. Fixed a buffer overflow in restore/tape.c. Patch provided by
7 Marcus Meissner (Caldera International Security Dept.).
8
9 2. Implement the Sun rmt extended protocol. Patch submitted
10 by Ian Gordon <iangordon@users.sourceforge.net>.
11
12 3. Check for the e2fsprogs header <ext2fs/ext2_fs.h> instead of
13 the linux kernel header. This ensures that dump always has the
14 latest version of this file. Patch submitted by
15 Andreas Dilger <adilger@turbolinux.com>.
16
17 4. Report any filesystem present in either /etc/fstab with a
18 positive passno or /etc/dumpdates in dump -w output.
19 Patch submitted by Andreas Dilger <adilger@turbolinux.com>.
20
21 Changes between versions 0.4b21 and 0.4b22 (released May 12, 2001)
22 ==================================================================
23
24 1. Made dump report the number of blocks dumped per volume.
25 Thanks to Kenneth Porter <shiva@well.com> for the suggestion.
26
27 2. Fix a bug in dump where waiting too much at the 'change volumes'
28 question caused the volume to be rewritten. Thanks to
29 Graham Reed <greed@users.sourceforge.net> for reporting the
30 bug and providing a patch.
31
32 3. Added a compression option to dump, contributed by Jerry
33 Peters <gapeters@worldnet.att.net>.
34
35 WARNING: the compression works only on files/pipes and
36 drives supporting variable block size.
37
38 WARNING2: tapes created using the compression feature are
39 incompatible with the standard dump tape format, and a
40 version of restore >= 0.4b22 will be needed for extraction.
41
42 4. Fixed some compilation problems for glibc 2.2.2 and 64 bit
43 architectures. Thanks to Paul Slootman <paul@debian.org> for
44 the patch and to Bdale Garbee <bdale@gag.com> for forwarding
45 it upstream.
46
47 5. Many cleanups (CPP defines, const char warnings, check of
48 ext2fs COMPAT flags, time_t cleanups, added libext2 version
49 in dump usage text) by Andreas Dilger <adilger@turbolinux.com>.
50
51 6. Made --prefix option work in configure. All the install path
52 are now based on the configure parameters.
53
54 7. Added the Quick File Access mode in dump/restore, contributed
55 by Uwe Gohlke <uwe@ugsoft.de>. In this mode, dump stores in
56 a file tape position for each inode, and this file is used by
57 restore (if called with parameter Q and the filename)
58 to directly position the tape at the file restore is currently
59 working on. This saves hours when restoring single files from
60 large backups, saves the tapes and the drive's head. Use
61 --enable-qfa option of configure to compile in the QFA support.
62
63 8. Added the possibility to dump several files and directories
64 in a single invocation of dump. Thanks to Uwe Gohlke
65 <uwe@ugsoft.de> for implementing this option.
66
67 9. Fixed the dumping and restoring of files with holes
68 (files ending with holes didn't get dumped or restored
69 correctly).
70
71 10. Fixed a socket descriptor leak in dump, which leaved opened
72 3 file descriptors per dump process (and there is one dump
73 process per tape).
74
75 11. Fixed dump large file system support, by explicit use of
76 open64/lseek64/etc functions (explicit use needed because
77 e2fsprogs libraries don't behave well when compiled with
78 FILE_OFFSET_BITS=64).
79
80 Changes between versions 0.4b20 and 0.4b21 (released January 13, 2001)
81 ======================================================================
82
83 1. Fixed some bugs in the dump -w|-W logic introduced by
84 the previous version. Thanks to Andreas Dilger
85 <adilger@turbolinux.com> for his help on this one.
86
87 2. Fixed again a compilation problem when using old e2fs
88 headers (filesystem label related). Thanks to many users
89 who reported this stupid error.
90
91 3. Fixed a build problem on old lib5 based systems dealing with
92 _PATH_MNTTAB being undefined. Thanks to John Adams
93 <johna@onevista.com> for reporting the bug.
94
95 4. Improved the error detection in restore when creating
96 the temporary files in TMPDIR. Now restore will corectly
97 report a 'no space left on device' error instead of
98 strange errors which could imply an archive corruption.
99 Thanks to Gabriel Somlo <somlo@cs.colostate.edu> and
100 bgjenero <bgjenero@sympatico.ca> for reporting the bug.
101
102 5. Added the throughput information in dump's progression
103 messages. Thanks to Andreas Dilger <adilger@turbolinux.com>
104 for the patch.
105
106 6. Use libext2fs's inode scanning routines, which are particularly
107 robust when dealing with errors (especially when having some
108 bad blocks on the filesystem). Many thanks to Colin
109 <colin@horizon.com> for reporting the bug, investigating
110 the issues, finding a workaround, writing the patch and
111 fully testing it... (of course, if this patch will break
112 anything, he is to blame for :-)).
113
114 7. Made dump and restore LFS aware. Dump can dump a filesystem
115 containing large files, generate a large file on output and
116 restore can restore them. This feature is not enabled by
117 default, use --enable-largefile option of configure to enable
118 it (you need to have a LFS aware glibc though). Thanks to
119 Andrea Arcangeli <andrea@suse.de> for submitting the patch,
120 and to Theodore T'so <tytso@valinux.com> for his always
121 useful thoughts.
122
123 8. Made dump ask upon a tape write error if it should rewrite
124 the current volume (assume this is a bad tape) or if it should
125 assume an end-of-tape condition (useful for tape drives which
126 behaves badly at the end of the tape). Thanks to Andreas
127 Dilger <adilger@turbolinux.com> for the suggestion.
128
129 Changes between versions 0.4b19 and 0.4b20 (released November 10, 2000)
130 =======================================================================
131
132 1. Fixed a small compilation problem due to a change
133 in the definintion of the struct sigaction in
134 glibc 2.0 / libc5. Thanks to Gunther Schlegel
135 <schlegel@riege.de> for reporting the bug and to
136 Dave Platt <dplatt@snulbug.mtview.ca.us> for suggesting
137 a fix.
138
139 2. Modified the label and uuid analysis in order to be
140 self-contained (independant of kernel/libc headers). This
141 should fix the compile with older kernel/libc headers and
142 will preserve the functionality. Thanks to Bernhard Erdmann
143 <bernhard.erdmann@gmx.de> for reporting the bug.
144
145 3. The 'exclude inode' option, if applied to a directory
146 inode, excludes now correctly all the leaves of this
147 directory. Thanks to John R. Dennison
148 <gerdesas@users.sourceforge.net> for reporting the bug.
149
150 4. Fixed the '-e' option to disable the possibility
151 to exclude the root inode (which causes the dumps to
152 be unreadable by restore). Prevented array overflow
153 when multiple -e options are used.
154
155 5. Fixed dump to correctly interpret a filesystem argument
156 which is a mountpoint and it is not an absolute path
157 (as specified in the fstab). Thanks to Bernhard R. Erdmann
158 <be@berdmann.de> for reporting the bug.
159
160 6. Made dump able to backup files larger than 2 GB. Note that
161 dump still doesn't cope with files larger than 4 GB.
162
163 7. Restore the real uid and gid before invoking an external
164 RSH program (big hole when dump or restore is suid root!).
165
166 8. Get the values for the default tape device and the location
167 of fstab file from the system headers. Thanks to
168 Andreas Dilger <adilger@turbolinux.com> for the patch.
169
170 9. Made dump -w|-W report all recognized filesystems
171 present in either /etc/fstab or /etc/dumpdates, and present
172 the list in the same order as in fstab file. Thanks
173 to Andreas Dilger <adilger@turbolinux.com> for the patch.
174
175 10. Made dump's -a (automatic end of tape detection) the
176 default. Specifying one of -B, -c, -d or -s options will
177 override the EOM detection. Thanks to Andreas Dilger
178 <adilger@turbolinux.com> for the patch.
179
180 11. Save the ext2 filesystem label into the dump volume label.
181 Specifying a volume label on the command line (-L option)
182 will override this feature. Thanks to Andreas Dilger
183 <adilger@turbolinux.com> for the patch.
184
185 Changes between versions 0.4b18 and 0.4b19 (released August 20, 2000)
186 =====================================================================
187
188 1. Fixed the signal handling in dump (which I broke in 0.4b17)
189 which was causing several strange dump failures (dump
190 hanged or segmentation faults etc).
191
192 2. Specified the default blocksize in the dump man page.
193
194 3. Changed two info messages of restore to be written on stdout
195 instead of stderr in order to leave stderr only for errors
196 or warnings. Thanks to Stephen Atwell
197 <satwell@urbana.css.mot.com> for the suggestion.
198
199 4. Corrected an off by one calculation which prevented
200 dumping a directory having a 1 character name length.
201 Thanks to Bernhard Erdmann <bernhard.erdmann@gmx.de>
202 for reporting the bug.
203
204 5. Reinforce dump in order to allow only level 0 and no
205 -u option when dumping a subdirectory, and document
206 this in the man page. Thanks to Bernhard Erdmann
207 <bernhard.erdmann@gmx.de> for reporting the bug.
208
209 6. Fixed a small harmless bug in restore which caused
210 in some conditions hard links to be restored several
211 times (therefore generation some warning messages).
212 Thanks to Kenneth Porter <shiva@well.com> for
213 reporting the bug.
214
215 7. Updated the RPM spec file to the latest RedHat version,
216 providing FHS packaging and other cosmetic changes.
217 You will need RPM version >= 3.0.5 in order to build the RPM.
218
219 8. Updated the configure script to check for libtermcap
220 before checking for libreadline, because we need this
221 library in order to compile the readline support.
222
223 9. Made dump understand the LABEL= and UUID= notation
224 both in /etc/fstab and on the command line. Note that
225 you will need the /proc filesystem in order to use
226 these notations. Thanks to Erik Troan <ewt@redhat.com>
227 for providing the patch.
228
229 Changes between versions 0.4b17 and 0.4b18 (released June 30, 2000)
230 ===================================================================
231
232 1. Fixed a potential buffer overflow in restore. Thanks
233 to Stan Bubrouski <satan@fastdial.net> for reporting
234 the bug.
235
236 2. Fixed a readline-related bug which prevented
237 'cat DUMPFILE | restore -i -f -' from working. Thanks
238 to Charles Karney <karney@users.sourceforge.net>
239 for the bug report.
240
241 3. Changed a few "panic" into "exit", causing restore to
242 be more stable against some attacks (like the last one
243 reported on Bugtraq, although the last version of restore
244 was not vulnerable - just dumped core). Thanks to
245 Andreas Hasenack <andreas@conectiva.com.br> for reporting
246 the bugs to me.
247
248 4. Removed the suid-root bit on dump and restore in the
249 default build (and generated RPMs). It should be safer
250 now. Users who need the suid-root capabilities in order
251 to make network backups should read first the man page
252 of dump and enable the suid bit manually.
253
254 5. Added -ltermcap to the compile parameters for restore
255 when using readline support, in order to make the compile
256 process work on some systems (SuSE ?). Thanks to
257 Patrik Schindler <poc@pocnet.net> for reporting the bug.
258
259 Changes between versions 0.4b16 and 0.4b17 (released June 1st, 2000)
260 ====================================================================
261
262 1. The -F script is called now *only* at the end of a tape,
263 not at the end of the dump. Thanks to Christian Haul
264 <haul@informatik.tu-darmstadt.de> for the bug report.
265
266 Normally, the device name and the current volume number
267 are passed on the command line of the script. However,
268 if you want the old style script (with no arguments
269 passed on the command line), you can enable it in
270 configure with the --enable-oldstylefscript.
271
272 2. Use posix signal handling to preserve dump functionality
273 with libc5. Thanks to Jeff Johnson <jbj@redhat.com> for
274 the patch.
275
276 3. Made the exit code of restore in case of a 'C'ompare
277 command reflect the result of the compare. An exit status
278 of 0 means the dump archive is correct, 1 means tape errors,
279 2 means that some files were modified. Thanks to Kenneth Porter
280 <shiva@well.com> for the suggestion.
281
282 4. Made (finally) quotes work in interactive restore.
283
284 5. Small fixes in order to allow dump to compile with a
285 really old e2fsprogs version. Thanks to Ian Zimmerman
286 <itz@speakeasy.org> for the bug report.
287
288 6. Add GNU readline capabilities to interactive restore.
289 Use configure's --enable-readline flag to enable this feature.
290 Thanks to Patrik Schindler <poc@pocnet.net> for the
291 suggestion.
292
293 7. Do the compare on the fly in restore 'C' mode (this will
294 allow not to exhaust the available /tmp space when
295 ccmparing large files). Thanks to Kenneth Porter
296 <shiva@well.com> for the suggestion.
297
298 Changes between versions 0.4b15 and 0.4b16 (released March 11, 2000)
299 ====================================================================
300
301 1. Fixed some several duplicate 'DUMP: DUMP:' in the
302 output of dump.
303
304 2. Corrected the estimation of blocks to dump. Note that
305 this is still wrong for large directory inodes, due
306 to the size of a BSD directory entry on the tape
307 (estimation is based on the size of the ext2 directory,
308 which is smaller because it doesn't contain the null
309 character at the end).
310
311 3. Made dump report the total number of MB written to
312 tape at the end of run. Thanks to W. Reilly Cooley
313 <wcooley@nakedape.cc> for the patch.
314
315 4. Added the -X option to restore giving the possibility
316 to read the names of the files to be extracted/listed
317 from a text file (in addition of the command line).
318 Thanks to Dejan Muhamedagic <dejan@quant-x.com> for the
319 patch.
320
321 5. Added the device name and the current volume number
322 as arguments to the end of tape script (dump -F option).
323
324 6. Made the multi-volume dump work again (after having
325 broken it in 0.4b15).
326
327 Changes between versions 0.4b14 and 0.4b15 (released March 2, 2000)
328 ===================================================================
329
330 1. Added a prompt command in interactive restore mode. Thanks
331 to Andreas Dilger <adilger@home.com> for the patch.
332
333 2. Fixed a buffer overflow problem in dump (caused by
334 not checking the size of the filesystem parameter).
335 Thanks to Kim Yong-jun <loveyou@hackerslab.org> for
336 reporting this on Bugtraq (and to several dump users
337 who forwarded me his mail).
338
339 3. Added the '-F script' option to dump in order to
340 launch a script at the end of each tape (to be used
341 with a tape changer, or to notify the sysadmin by
342 pager etc.).
343
344 4. Fixed a bug in restore compare code caused by the changes
345 I made in 0.4b14.
346
347 5. Fixed the treatment of options using the old BSD syntax
348 in both dump and restore.
349
350 Changes between versions 0.4b13 and 0.4b14 (released February 10, 2000)
351 =======================================================================
352
353 1. Fixed a bug in dump which may have caused invalid deleted
354 directories to be dumped out if they were deleted after the
355 mapping pass. This could occure on active filesystem and lead
356 to heap corruption (causing dump malfunction in many possible ways).
357 Thanks to Jan Sanislo <oystr@cs.washington.edu> for finding this
358 bug and submitting the patch.
359
360 2. Fixed the handling of the filesystem parameter in dump. A
361 '/mnt/disk' parameter caused the disk contents to be dumped,
362 but a '/mnt/disk/' parameter caused the mountpoint directory
363 to be dumped (generally an empty directory).
364
365 3. Improved the output of dump in order to tell which directory
366 it is currently dumping (when dumping a subtree).
367
368 4. Added the '-e' exclude inode option to dump. Thanks to
369 Isaac Chuang <ike@isl.stanford.edu> for contributing with the patch.
370
371 5. Added a REPORTING-BUGS file in order to provide a guide
372 on how to correctly report a bug in dump/restore.
373
374 6. Fixed a restore problem when restoring a hard link to an inode
375 having the immutable or append-only attribute set. Thanks to
376 Ambrose Li <acli@mingpaoxpress.com> for submitting the patch.
377
378 7. Fixed a compatibility problem between dumps made on little
379 endian machines (the format was correct) and big endian
380 machines (the format was incorrect). This fix break the
381 compatibility with the older dumps made on big endian
382 machines (sparc, m86k, ppc etc). For the first time in
383 linux dump's history (I believe), the dumps made by *BSD,
384 Linux/alpha, Linux/sparc and Linux/x86 are compatible,
385 so interchangeable. Thanks to Rob Cermak
386 <cermak@ahab.rutgers.edu> for submitting the bug and
387 helping me test the fix.
388
389 8. Fixed the way dump reports the remaining percent/time, if
390 the number of blocks actually dumped exceeds the estimated
391 number of blocks. Thanks to Jean-Paul van der Jagt
392 <jeanpaul@dutepp0.et.tudelft.nl> for reporting the bug.
393
394 Changes between versions 0.4b12 and 0.4b13 (released January 21, 2000)
395 ======================================================================
396
397 1. Small Y2K fix in the man pages :). Thanks to Bernhard Sadlowski
398 <sadlowsk@Mathematik.Uni-Bielefeld.DE> for reporting the bug.
399
400 2. Removed the requirement to build the RPM as root from the
401 spec file. Thanks to Christian Weisgerber
402 <naddy@mips.rhein-neckar.de> for submitting this.
403
404 3. Fixed a bug in dump related to the 'filetype' feature of ext2fs,
405 causing dump to block when dumping really huge filesystems.
406 Many thanks to Patrik Schindler <poc@pocnet.net> for
407 helping me find this bug.
408
409 4. Fixed the treatment for an interrupt signal when dump access
410 the remote tape through RSH. Thanks to Christian Weisgerber
411 <naddy@mips.rhein-neckar.de> for providing the patch.
412
413 5. Fixed a bug which was causing dump/restore to display
414 garbage characters instead of the remote host name.
415
416 Changes between versions 0.4b11 and 0.4b12 (released January 8, 2000)
417 =====================================================================
418
419 1. Small fix in the dump man page. Thanks to Thorsten Kukuk
420 <kukuk@suse.de> for submitting the patch.
421
422 2. Fix for the exit code when using the size estimate option of
423 dump. Thanks to Matti Taskinen <mkt@rni.helsinki.fi> for
424 submitting the patch.
425
426 3. Handle EINTR in atomical reads/writes in dump, which was causing
427 dump to fail on some systems. Thanks to Eric Jergensen
428 <eric@dvns.com> for reporting the bug and submitting the patch.
429
430 4. Handle more than 16 characters for the device names in dumpdates.
431 (up to 255 now). Thanks to Rainer Clasen <bj@ncc.cicely.de> for
432 tracking down the problem and proposing the solution.
433
434 5. Fixed a bug in dump which prevented the creation of the
435 dumpdates file when doing a 0-level dump without already
436 having a dumpdates file. Thanks to Patrik Schindler
437 <poc@pocnet.net> for reporting the bug.
438
439 6. Changed the way dump 'S' flag reports the size estimate
440 from number of blocks into bytes (making it compatible
441 with the Solaris version, and simplifying things for
442 amanda users). Thanks to Jason L Tibbitts III
443 <tibbs@math.uh.edu> for reporting the bug.
444
445 7. Fixed a compatibility problem in linux/alpha dump tape format.
446 Now the linux/alpha dump are (again) compatible with the
447 other dump formats. But this breaks compatibility with
448 older dumps made on alpha. Thanks to Mike Tibor
449 <tibor@lib.uaa.alaska.edu> for helping me in finding this bug.
450
451 Changes between versions 0.4b10 and 0.4b11 (released December 5, 1999)
452 ======================================================================
453
454 1. Added a '--enable-kerberos' to configure.
455
456 2. Added a 'S' option to dump which determines the amount of space
457 that is needed to perform the dump without actually doing it, similar
458 to the Sun's ufsdump 'S' option. Patch contributed by Rob Cermak
459 <cermak@ahab.rutgers.edu>.
460
461 3. Added a 'M' multi-volume option to dump and restore which enables
462 dumping to multiple files (useful when dumping to an ext2
463 partition to make several dump files in order to bypass the 2GB
464 file size limitation). The 'f' argument is treated as a prefix and
465 the output files will be named <prefix>001, <prefix>002 etc. With
466 the 'M' flag, restore automatically selects the right file without
467 asking to enter a new tape each time.
468
469 4. Fixed a memory leak which was causing dump to grow very big
470 (270MB when dumping a 10GB filesystem...). Thanks to Jason
471 Fearon <jasonf@netrider.org.au> for reporting the bug.
472
473 Changes between versions 0.4b9 and 0.4b10 (released November 21, 1999)
474 ======================================================================
475
476 1. Make configure test if the system glob routines support
477 extended syntax (ALTDIRFUNC). If not, use the internal glob
478 routines instead of system ones. Thanks to Bernhard Sadlowski
479 <sadlowsk@Mathematik.Uni-Bielefeld.DE> for reporting the bug
480 and helping me resolve this and other minor libc5 compiling
481 glitches.
482
483 2. Fix a problem when dumping a ext2fs with the 'filetype'
484 feature enabled. Thanks to Patrick J. LoPresti
485 <patl@cag.lcs.mit.edu> for reporting the bug and to
486 Theodore Y. Ts'o <tytso@mit.edu> for providing the patch.
487
488 3. Made the nodump flag work on directories. A directory which
489 has the nodump flag gets never dumped, regardless of its
490 contents.
491
492 4. Integrate a patch from Jeremy Fitzhardinge <jeremy@goop.org>
493 which allows dump on an active ext3 filesystem. However, this
494 is a "quick and dirty" patch which enables backup of an ext3
495 filesystem through the ext2 compatibility (by ignoring the
496 NEEDS_RECOVERY bit). The journal file is not recognized and
497 it is dumped (it should not).
498
499 5. Test the superblock compatibility flags when dumping, in order
500 to be sure that we know how to deal with specific features.
501
502 Changes between versions 0.4b8 and 0.4b9 (released November 5, 1999)
503 ====================================================================
504
505 1. Use lchown instead of chown, fixing a possible security problem
506 when restoring symlinks (a malicious user could use this
507 to deliberately corrupt the ownership of important system files).
508 Thanks to Chris Siebenmann <cks@utcc.utoronto.ca> for detecting
509 this and providing the patch.
510
511 Changes between versions 0.4b7 and 0.4b8 (released November 3, 1999)
512 ====================================================================
513
514 1. Put dump sources under CVS, added Id tags in all files so
515 one can use 'ident' on binary files.
516
517 2. Added the dump/restore version in the usage text so one can
518 easily verify the version he is using.
519
520 3. Small patch from Nuno Oliveira <nuno@eq.uc.pt> which fixes
521 a va_start/va_end problem on linux-ppc (always call va_start
522 va_end in pairs each time we use a vararg function).
523
524 4. Added again the DT_* constants because old libc does not
525 contain them :(. Thanks to Eric Maisonobe <virnet@nat.fr>
526 for submitting the bug report.
527
528 5. Use ext2fs_llseek instead of llseek. With recent e2fsprogs
529 this should enable dumping big (huge) filesystems.
530
531 6. Added the RSH environment variable in order to be able to
532 use a rsh replacement like ssh when doing remote backups (and
533 bypass the security limitations of rcmd). Now you can do remote
534 backups without being root (or making dump setuid root).
535
536 7. Modified again the way dumpdates works. For incremental dumps,
537 we need to read dumpdates even if we are not using 'u' option.
538 Thanks to Bdale Garbee <bdale@gag.com> for his ideas on how
539 this should work.
540
541 Changes between versions 0.4b6 and 0.4b7 (released October 8, 1999)
542 ===================================================================
543
544 1. Removed the 'k' flag from the restore 'about' text if kerberos
545 was not compiled in.
546
547 2. Prototyped (f)setflags from e2fsprogs and corrected the calls
548 to them (fsetflags takes a char*, setflags an open fd!).
549
550 3. (f)setflags is called only if the flags aren't empty. If the
551 file is a special file, a warning is printed, because changing
552 flags implies opening the device. Normally, a special file
553 should not have any flag... (Debian bug #29775, patch provided
554 by Abhijit Dasgupta <abhijit@ans.net>).
555
556 4. Made possible to dump a file system not mentioned in /etc/fstab.
557 (Debian bug #11904, patch provided by Eirik Fuller <eirik@netcom.com>).
558
559 5. Changed the default behaviour to not create dumpdates
560 unless 'u' option is specified. Removed the old "debian-patch"
561 which provided the same thing. (Debian bug #38136, #33818).
562
563 6. Removed all those dump*announce, since they were getting old...
564
565 7. Added warning messages if dumpdates does not exist and
566 when an inferior level dump does not exist (except for a level 0
567 dump).
568
569 8. Debugged the glob calls in interactive mode: restore used a
570 dirent struct which was different from the /usr/include/dirent.h
571 one (this used to work, is it a glibc2 change?), so none of the
572 compat glob (which used /usr/include/dirent.h) or the system glob
573 worked. Restore use now the system dirent (and the system
574 DT_* constants), which are compatible with BSD ones.
575
576 9. Added a configure flag (--with-dumpdatespath) to specify
577 the location of dumpdates file. By default, it is
578 /etc/dumpdates.
579
580 10. Added the "AUTHOR" and "AVAILABILITY" sections and
581 included the current date/version in man pages.
582
583 11. Corrected the estimation of remaining time when
584 the operator doesn't change the tapes quickly enough. This
585 was an old bug, I thought I corrected it, and discovered
586 that in fact it was corrected in two different places, so
587 the results canceled each other...
588
589 Changes between versions 0.4b5 and 0.4b6 (released October 1, 1999)
590 ===================================================================
591
592 1. Integrated multiple patches from RedHat, Debian and SuSE:
593
594 - tweak dump/itime.c to not try to read dumpdates if the 'u' option
595 isn't specified.
596 - several fixes in the man pages.
597 - update the default tape device to /dev/st0.
598 - many updates for Linux Alpha (byte ordering, size_t etc).
599 - buffer overruns.
600 - use environment variable for TMPDIR (instead of /tmp).
601 - use sigjmp_buf instead of jmp_buf (RedHat bug #3260).
602 - workaround egcs bug (RedHat bugs #4281 and #2989).
603 - wire $(OPT) throughout Makefile's.
604
605 2. Upgrade the dump revision to 1, making possible to dump filesystems
606 made with e2fsprogs-1.15 or newer. Nothing seems to break...
607
608 3. Fix some compile warnings, prototype all functions.
609
610 4. Use glibc err/glob instead of internal compatibility
611 routines (only if available).
612
613 5. Fix a compile error on Linux 2.2.7 / libc5 (5.4.44) (patch provided
614 by Bernhard Sadlowski <sadlowsk@mathematik.uni-bielefeld.de>).
615
616 Changes between versions 0.4b4 and 0.4b5 (released September 22, 1999)
617 ======================================================================
618
619 1. Integrated the changes from FreeBSD-3.1-RELEASE
620 (mostly bug fixes, buffer overruns, dump has now an "automatic
621 tape length calculation" flag, dump/restore can use kerberos now
622 (this is NOT tested), use environment variables for TAPE and
623 RMT etc.).
624
625 2. Integrated three RedHat patches ("glibc", "kernel" and "bread" patches)
626
627 3. Corrected a bug in restore when using 'C' option with multi-volumes
628 tapes (files splited accros two tapes give "size changed" errors
629 when comparing).
630
631 4. Corrected the long standing bug when dumping multiple tapes.
632 This works for me, needs further testing.
633
634 Changes between versions 0.4b3 and 0.4b4 (released January 17, 1997)
635 ====================================================================
636
637 1. Dump now runs correctly on kernels 2.1.x
638 Fix made by Gerald Peters <gapeters@worldnet.att.net>
639
640 Changes between versions 0.4b2 and 0.4b3
641 ========================================
642
643 1. Use realpath() if available
644
645 2. Report statistics
646
647 Changes between versions 0.4b1 and 0.4b2
648 ========================================
649
650 1. Fixed the bug fix from Greg Lutz (I had made a mistake when integrating
651 the patch)
652
653 2. Fixed restore to make it able to read FreeBSD 2.x dumps again
654
655 3. Fixed configure.in to correctly handle --enable-rmt
656
657 Changes between versions 0.3 and 0.4b1
658 ======================================
659
660 1. Integrated the changes from 4.4BSD-Lite2
661
662 2. Integrated the patches from Debian and Red Hat
663
664 3. Portability changes: use the __u32, __u16, __s32, and __s16 types
665
666 4. Changed dump to use the Ext2fs library to get block addresses. This
667 should solve the endianness problem on SparcLinux.
668
669 5. Created a configure.in file (shamelessly stolen from the e2fsprogs
670 distribution's one) to use autoconf
671
672 6. Fixed a few minor bugs
673
674 Changes between versions 0.2e and 0.2f
675 ======================================
676
677 1. Added the creation of named pipes (fifos) in restore.
678
679 2. Added the -N flag in the restore manual page.
680
681 3. Added the file kernel.patch which contains the llseek() optimization
682 patch for 1.2.x kernels.
683
684 4. Fixed a bug in the restoration of symbolic links: owner and group were
685 not restored.
686
687 5. Integrated some changes from FreeBSD 2.2.
688
689 6. Added a call to ftruncate() after restoring each file to restore
690 correctly files ending by a hole.
691
692 Changes between versions 0.2d and 0.2e
693 ======================================
694
695 1. Fixed a bug in the "set owner/mode" process. Every file was restored
696 with owner = root (0) and group = root/wheel/whatever (0).
697
698 Changes between versions 0.2c and 0.2d
699 ======================================
700
701 1. Dump is now able to backup 2GB+ filesystems.
702
703 2. Dump and restore can now be linked as static binaries.
704
705 Changes between versions 0.2b and 0.2c
706 ======================================
707
708 1. Fixed a bug when dumping ``slow'' (i.e. normal) symbolic links.
709
710 Changes between versions 0.2a and 0.2b
711 ======================================
712
713 1. Really fixed the bug that I should have corrected in 0.2a.
714
715 2. Enabled optimization again.
716
717 Changes between versions 0.2 and 0.2a
718 =====================================
719
720 1. Disabled the optimization during compilation.
721
722 Changes between versions 0.1 and 0.2
723 ====================================
724
725 1. Fixed a bug in fstab.c which caused a null pointer to be stored in
726 the fs_type field (actually, I modified the file fstab.c to make it
727 use the mntent functions).
728
729 2. Dump and restore now use a 4.3 BSD compatible dump format. Backups
730 made by dump should be readable by the BSD restore and backups made
731 by the BSD dump should be readable by restore. Unfortunately, this
732 means that the dump format has changed between version 0.1 and version
733 0.2 :-(
734
735 3. Dump is now able to backup a subtree, it is no longer limited to whole
736 filesystems like the BSD version.
737
738 4. Dump now uses ext2_llseek() so it is able to backup filesystems bigger
739 than 2 GB.
740
741 Changes between versions 0.0 and 0.1
742 ====================================
743
744 1. Now create links rdump and rrestore during the `make install' step.
745
746 2. Linux port specific bugs added to the manual pages
747
748 3. Incorrect estimation of the number of tapes blocks fixed when doing
749 incremental backups.
750
751 4. Better ls-like format in restore in interactive mode.