1 $Id: CHANGES,v 1.93 2000/12/04 14:00:14 stelian Exp $
3 Changes between versions 0.4b20 and 0.4b21 (released ?????????????????)
4 =======================================================================
6 1. Fixed some bugs in the dump -w|-W logic introduced by
7 the previous version. Thanks to Andreas Dilger
8 <adilger@turbolinux.com> for his help on this one.
10 2. Fixed again a compilation problem when using old e2fs
11 headers (filesystem label related). Thanks to many users
12 who reported this stupid error.
14 3. Fixed a build problem on old lib5 based systems dealing with
15 _PATH_MNTTAB being undefined. Thanks to <johna@onevista.com>
16 for reporting the bug.
18 Changes between versions 0.4b19 and 0.4b20 (released November 10, 2000)
19 =======================================================================
21 1. Fixed a small compilation problem due to a change
22 in the definintion of the struct sigaction in
23 glibc 2.0 / libc5. Thanks to Gunther Schlegel
24 <schlegel@riege.de> for reporting the bug and to
25 Dave Platt <dplatt@snulbug.mtview.ca.us> for suggesting
28 2. Modified the label and uuid analysis in order to be
29 self-contained (independant of kernel/libc headers). This
30 should fix the compile with older kernel/libc headers and
31 will preserve the functionality. Thanks to Bernhard Erdmann
32 <bernhard.erdmann@gmx.de> for reporting the bug.
34 3. The 'exclude inode' option, if applied to a directory
35 inode, excludes now correctly all the leaves of this
36 directory. Thanks to John R. Dennison
37 <gerdesas@users.sourceforge.net> for reporting the bug.
39 4. Fixed the '-e' option to disable the possibility
40 to exclude the root inode (which causes the dumps to
41 be unreadable by restore). Prevented array overflow
42 when multiple -e options are used.
44 5. Fixed dump to correctly interpret a filesystem argument
45 which is a mountpoint and it is not an absolute path
46 (as specified in the fstab). Thanks to Bernhard R. Erdmann
47 <be@berdmann.de> for reporting the bug.
49 6. Made dump able to backup files larger than 2 GB. Note that
50 dump still doesn't cope with files larger than 4 GB.
52 7. Restore the real uid and gid before invoking an external
53 RSH program (big hole when dump or restore is suid root!).
55 8. Get the values for the default tape device and the location
56 of fstab file from the system headers. Thanks to
57 Andreas Dilger <adilger@turbolinux.com> for the patch.
59 9. Made dump -w|-W report all recognized filesystems
60 present in either /etc/fstab or /etc/dumpdates, and present
61 the list in the same order as in fstab file. Thanks
62 to Andreas Dilger <adilger@turbolinux.com> for the patch.
64 10. Made dump's -a (automatic end of tape detection) the
65 default. Specifying one of -B, -c, -d or -s options will
66 override the EOM detection. Thanks to Andreas Dilger
67 <adilger@turbolinux.com> for the patch.
69 11. Save the ext2 filesystem label into the dump volume label.
70 Specifying a volume label on the command line (-L option)
71 will override this feature. Thanks to Andreas Dilger
72 <adilger@turbolinux.com> for the patch.
74 Changes between versions 0.4b18 and 0.4b19 (released August 20, 2000)
75 =====================================================================
77 1. Fixed the signal handling in dump (which I broke in 0.4b17)
78 which was causing several strange dump failures (dump
79 hanged or segmentation faults etc).
81 2. Specified the default blocksize in the dump man page.
83 3. Changed two info messages of restore to be written on stdout
84 instead of stderr in order to leave stderr only for errors
85 or warnings. Thanks to Stephen Atwell
86 <satwell@urbana.css.mot.com> for the suggestion.
88 4. Corrected an off by one calculation which prevented
89 dumping a directory having a 1 character name length.
90 Thanks to Bernhard Erdmann <bernhard.erdmann@gmx.de>
91 for reporting the bug.
93 5. Reinforce dump in order to allow only level 0 and no
94 -u option when dumping a subdirectory, and document
95 this in the man page. Thanks to Bernhard Erdmann
96 <bernhard.erdmann@gmx.de> for reporting the bug.
98 6. Fixed a small harmless bug in restore which caused
99 in some conditions hard links to be restored several
100 times (therefore generation some warning messages).
101 Thanks to Kenneth Porter <shiva@well.com> for
104 7. Updated the RPM spec file to the latest RedHat version,
105 providing FHS packaging and other cosmetic changes.
106 You will need RPM version >= 3.0.5 in order to build the RPM.
108 8. Updated the configure script to check for libtermcap
109 before checking for libreadline, because we need this
110 library in order to compile the readline support.
112 9. Made dump understand the LABEL= and UUID= notation
113 both in /etc/fstab and on the command line. Note that
114 you will need the /proc filesystem in order to use
115 these notations. Thanks to Erik Troan <ewt@redhat.com>
116 for providing the patch.
118 Changes between versions 0.4b17 and 0.4b18 (released June 30, 2000)
119 ===================================================================
121 1. Fixed a potential buffer overflow in restore. Thanks
122 to Stan Bubrouski <satan@fastdial.net> for reporting
125 2. Fixed a readline-related bug which prevented
126 'cat DUMPFILE | restore -i -f -' from working. Thanks
127 to Charles Karney <karney@users.sourceforge.net>
130 3. Changed a few "panic" into "exit", causing restore to
131 be more stable against some attacks (like the last one
132 reported on Bugtraq, although the last version of restore
133 was not vulnerable - just dumped core). Thanks to
134 Andreas Hasenack <andreas@conectiva.com.br> for reporting
137 4. Removed the suid-root bit on dump and restore in the
138 default build (and generated RPMs). It should be safer
139 now. Users who need the suid-root capabilities in order
140 to make network backups should read first the man page
141 of dump and enable the suid bit manually.
143 5. Added -ltermcap to the compile parameters for restore
144 when using readline support, in order to make the compile
145 process work on some systems (SuSE ?). Thanks to
146 Patrik Schindler <poc@pocnet.net> for reporting the bug.
148 Changes between versions 0.4b16 and 0.4b17 (released June 1st, 2000)
149 ====================================================================
151 1. The -F script is called now *only* at the end of a tape,
152 not at the end of the dump. Thanks to Christian Haul
153 <haul@informatik.tu-darmstadt.de> for the bug report.
155 Normally, the device name and the current volume number
156 are passed on the command line of the script. However,
157 if you want the old style script (with no arguments
158 passed on the command line), you can enable it in
159 configure with the --enable-oldstylefscript.
161 2. Use posix signal handling to preserve dump functionality
162 with libc5. Thanks to Jeff Johnson <jbj@redhat.com> for
165 3. Made the exit code of restore in case of a 'C'ompare
166 command reflect the result of the compare. An exit status
167 of 0 means the dump archive is correct, 1 means tape errors,
168 2 means that some files were modified. Thanks to Kenneth Porter
169 <shiva@well.com> for the suggestion.
171 4. Made (finally) quotes work in interactive restore.
173 5. Small fixes in order to allow dump to compile with a
174 really old e2fsprogs version. Thanks to Ian Zimmerman
175 <itz@speakeasy.org> for the bug report.
177 6. Add GNU readline capabilities to interactive restore.
178 Use configure's --enable-readline flag to enable this feature.
179 Thanks to Patrik Schindler <poc@pocnet.net> for the
182 7. Do the compare on the fly in restore 'C' mode (this will
183 allow not to exhaust the available /tmp space when
184 ccmparing large files). Thanks to Kenneth Porter
185 <shiva@well.com> for the suggestion.
187 Changes between versions 0.4b15 and 0.4b16 (released March 11, 2000)
188 ====================================================================
190 1. Fixed some several duplicate 'DUMP: DUMP:' in the
193 2. Corrected the estimation of blocks to dump. Note that
194 this is still wrong for large directory inodes, due
195 to the size of a BSD directory entry on the tape
196 (estimation is based on the size of the ext2 directory,
197 which is smaller because it doesn't contain the null
198 character at the end).
200 3. Made dump report the total number of MB written to
201 tape at the end of run. Thanks to W. Reilly Cooley
202 <wcooley@nakedape.cc> for the patch.
204 4. Added the -X option to restore giving the possibility
205 to read the names of the files to be extracted/listed
206 from a text file (in addition of the command line).
207 Thanks to Dejan Muhamedagic <dejan@quant-x.com> for the
210 5. Added the device name and the current volume number
211 as arguments to the end of tape script (dump -F option).
213 6. Made the multi-volume dump work again (after having
214 broken it in 0.4b15).
216 Changes between versions 0.4b14 and 0.4b15 (released March 2, 2000)
217 ===================================================================
219 1. Added a prompt command in interactive restore mode. Thanks
220 to Andreas Dilger <adilger@home.com> for the patch.
222 2. Fixed a buffer overflow problem in dump (caused by
223 not checking the size of the filesystem parameter).
224 Thanks to Kim Yong-jun <loveyou@hackerslab.org> for
225 reporting this on Bugtraq (and to several dump users
226 who forwarded me his mail).
228 3. Added the '-F script' option to dump in order to
229 launch a script at the end of each tape (to be used
230 with a tape changer, or to notify the sysadmin by
233 4. Fixed a bug in restore compare code caused by the changes
236 5. Fixed the treatment of options using the old BSD syntax
237 in both dump and restore.
239 Changes between versions 0.4b13 and 0.4b14 (released February 10, 2000)
240 =======================================================================
242 1. Fixed a bug in dump which may have caused invalid deleted
243 directories to be dumped out if they were deleted after the
244 mapping pass. This could occure on active filesystem and lead
245 to heap corruption (causing dump malfunction in many possible ways).
246 Thanks to Jan Sanislo <oystr@cs.washington.edu> for finding this
247 bug and submitting the patch.
249 2. Fixed the handling of the filesystem parameter in dump. A
250 '/mnt/disk' parameter caused the disk contents to be dumped,
251 but a '/mnt/disk/' parameter caused the mountpoint directory
252 to be dumped (generally an empty directory).
254 3. Improved the output of dump in order to tell which directory
255 it is currently dumping (when dumping a subtree).
257 4. Added the '-e' exclude inode option to dump. Thanks to
258 Isaac Chuang <ike@isl.stanford.edu> for contributing with the patch.
260 5. Added a REPORTING-BUGS file in order to provide a guide
261 on how to correctly report a bug in dump/restore.
263 6. Fixed a restore problem when restoring a hard link to an inode
264 having the immutable or append-only attribute set. Thanks to
265 Ambrose Li <acli@mingpaoxpress.com> for submitting the patch.
267 7. Fixed a compatibility problem between dumps made on little
268 endian machines (the format was correct) and big endian
269 machines (the format was incorrect). This fix break the
270 compatibility with the older dumps made on big endian
271 machines (sparc, m86k, ppc etc). For the first time in
272 linux dump's history (I believe), the dumps made by *BSD,
273 Linux/alpha, Linux/sparc and Linux/x86 are compatible,
274 so interchangeable. Thanks to Rob Cermak
275 <cermak@ahab.rutgers.edu> for submitting the bug and
276 helping me test the fix.
278 8. Fixed the way dump reports the remaining percent/time, if
279 the number of blocks actually dumped exceeds the estimated
280 number of blocks. Thanks to Jean-Paul van der Jagt
281 <jeanpaul@dutepp0.et.tudelft.nl> for reporting the bug.
283 Changes between versions 0.4b12 and 0.4b13 (released January 21, 2000)
284 ======================================================================
286 1. Small Y2K fix in the man pages :). Thanks to Bernhard Sadlowski
287 <sadlowsk@Mathematik.Uni-Bielefeld.DE> for reporting the bug.
289 2. Removed the requirement to build the RPM as root from the
290 spec file. Thanks to Christian Weisgerber
291 <naddy@mips.rhein-neckar.de> for submitting this.
293 3. Fixed a bug in dump related to the 'filetype' feature of ext2fs,
294 causing dump to block when dumping really huge filesystems.
295 Many thanks to Patrik Schindler <poc@pocnet.net> for
296 helping me find this bug.
298 4. Fixed the treatment for an interrupt signal when dump access
299 the remote tape through RSH. Thanks to Christian Weisgerber
300 <naddy@mips.rhein-neckar.de> for providing the patch.
302 5. Fixed a bug which was causing dump/restore to display
303 garbage characters instead of the remote host name.
305 Changes between versions 0.4b11 and 0.4b12 (released January 8, 2000)
306 =====================================================================
308 1. Small fix in the dump man page. Thanks to Thorsten Kukuk
309 <kukuk@suse.de> for submitting the patch.
311 2. Fix for the exit code when using the size estimate option of
312 dump. Thanks to Matti Taskinen <mkt@rni.helsinki.fi> for
313 submitting the patch.
315 3. Handle EINTR in atomical reads/writes in dump, which was causing
316 dump to fail on some systems. Thanks to Eric Jergensen
317 <eric@dvns.com> for reporting the bug and submitting the patch.
319 4. Handle more than 16 characters for the device names in dumpdates.
320 (up to 255 now). Thanks to Rainer Clasen <bj@ncc.cicely.de> for
321 tracking down the problem and proposing the solution.
323 5. Fixed a bug in dump which prevented the creation of the
324 dumpdates file when doing a 0-level dump without already
325 having a dumpdates file. Thanks to Patrik Schindler
326 <poc@pocnet.net> for reporting the bug.
328 6. Changed the way dump 'S' flag reports the size estimate
329 from number of blocks into bytes (making it compatible
330 with the Solaris version, and simplifying things for
331 amanda users). Thanks to Jason L Tibbitts III
332 <tibbs@math.uh.edu> for reporting the bug.
334 7. Fixed a compatibility problem in linux/alpha dump tape format.
335 Now the linux/alpha dump are (again) compatible with the
336 other dump formats. But this breaks compatibility with
337 older dumps made on alpha. Thanks to Mike Tibor
338 <tibor@lib.uaa.alaska.edu> for helping me in finding this bug.
340 Changes between versions 0.4b10 and 0.4b11 (released December 5, 1999)
341 ======================================================================
343 1. Added a '--enable-kerberos' to configure.
345 2. Added a 'S' option to dump which determines the amount of space
346 that is needed to perform the dump without actually doing it, similar
347 to the Sun's ufsdump 'S' option. Patch contributed by Rob Cermak
348 <cermak@ahab.rutgers.edu>.
350 3. Added a 'M' multi-volume option to dump and restore which enables
351 dumping to multiple files (useful when dumping to an ext2
352 partition to make several dump files in order to bypass the 2GB
353 file size limitation). The 'f' argument is treated as a prefix and
354 the output files will be named <prefix>001, <prefix>002 etc. With
355 the 'M' flag, restore automatically selects the right file without
356 asking to enter a new tape each time.
358 4. Fixed a memory leak which was causing dump to grow very big
359 (270MB when dumping a 10GB filesystem...). Thanks to Jason
360 Fearon <jasonf@netrider.org.au> for reporting the bug.
362 Changes between versions 0.4b9 and 0.4b10 (released November 21, 1999)
363 ======================================================================
365 1. Make configure test if the system glob routines support
366 extended syntax (ALTDIRFUNC). If not, use the internal glob
367 routines instead of system ones. Thanks to Bernhard Sadlowski
368 <sadlowsk@Mathematik.Uni-Bielefeld.DE> for reporting the bug
369 and helping me resolve this and other minor libc5 compiling
372 2. Fix a problem when dumping a ext2fs with the 'filetype'
373 feature enabled. Thanks to Patrick J. LoPresti
374 <patl@cag.lcs.mit.edu> for reporting the bug and to
375 Theodore Y. Ts'o <tytso@mit.edu> for providing the patch.
377 3. Made the nodump flag work on directories. A directory which
378 has the nodump flag gets never dumped, regardless of its
381 4. Integrate a patch from Jeremy Fitzhardinge <jeremy@goop.org>
382 which allows dump on an active ext3 filesystem. However, this
383 is a "quick and dirty" patch which enables backup of an ext3
384 filesystem through the ext2 compatibility (by ignoring the
385 NEEDS_RECOVERY bit). The journal file is not recognized and
386 it is dumped (it should not).
388 5. Test the superblock compatibility flags when dumping, in order
389 to be sure that we know how to deal with specific features.
391 Changes between versions 0.4b8 and 0.4b9 (released November 5, 1999)
392 ====================================================================
394 1. Use lchown instead of chown, fixing a possible security problem
395 when restoring symlinks (a malicious user could use this
396 to deliberately corrupt the ownership of important system files).
397 Thanks to Chris Siebenmann <cks@utcc.utoronto.ca> for detecting
398 this and providing the patch.
400 Changes between versions 0.4b7 and 0.4b8 (released November 3, 1999)
401 ====================================================================
403 1. Put dump sources under CVS, added Id tags in all files so
404 one can use 'ident' on binary files.
406 2. Added the dump/restore version in the usage text so one can
407 easily verify the version he is using.
409 3. Small patch from Nuno Oliveira <nuno@eq.uc.pt> which fixes
410 a va_start/va_end problem on linux-ppc (always call va_start
411 va_end in pairs each time we use a vararg function).
413 4. Added again the DT_* constants because old libc does not
414 contain them :(. Thanks to Eric Maisonobe <virnet@nat.fr>
415 for submitting the bug report.
417 5. Use ext2fs_llseek instead of llseek. With recent e2fsprogs
418 this should enable dumping big (huge) filesystems.
420 6. Added the RSH environment variable in order to be able to
421 use a rsh replacement like ssh when doing remote backups (and
422 bypass the security limitations of rcmd). Now you can do remote
423 backups without being root (or making dump setuid root).
425 7. Modified again the way dumpdates works. For incremental dumps,
426 we need to read dumpdates even if we are not using 'u' option.
427 Thanks to Bdale Garbee <bdale@gag.com> for his ideas on how
430 Changes between versions 0.4b6 and 0.4b7 (released October 8, 1999)
431 ===================================================================
433 1. Removed the 'k' flag from the restore 'about' text if kerberos
436 2. Prototyped (f)setflags from e2fsprogs and corrected the calls
437 to them (fsetflags takes a char*, setflags an open fd!).
439 3. (f)setflags is called only if the flags aren't empty. If the
440 file is a special file, a warning is printed, because changing
441 flags implies opening the device. Normally, a special file
442 should not have any flag... (Debian bug #29775, patch provided
443 by Abhijit Dasgupta <abhijit@ans.net>).
445 4. Made possible to dump a file system not mentioned in /etc/fstab.
446 (Debian bug #11904, patch provided by Eirik Fuller <eirik@netcom.com>).
448 5. Changed the default behaviour to not create dumpdates
449 unless 'u' option is specified. Removed the old "debian-patch"
450 which provided the same thing. (Debian bug #38136, #33818).
452 6. Removed all those dump*announce, since they were getting old...
454 7. Added warning messages if dumpdates does not exist and
455 when an inferior level dump does not exist (except for a level 0
458 8. Debugged the glob calls in interactive mode: restore used a
459 dirent struct which was different from the /usr/include/dirent.h
460 one (this used to work, is it a glibc2 change?), so none of the
461 compat glob (which used /usr/include/dirent.h) or the system glob
462 worked. Restore use now the system dirent (and the system
463 DT_* constants), which are compatible with BSD ones.
465 9. Added a configure flag (--with-dumpdatespath) to specify
466 the location of dumpdates file. By default, it is
469 10. Added the "AUTHOR" and "AVAILABILITY" sections and
470 included the current date/version in man pages.
472 11. Corrected the estimation of remaining time when
473 the operator doesn't change the tapes quickly enough. This
474 was an old bug, I thought I corrected it, and discovered
475 that in fact it was corrected in two different places, so
476 the results canceled each other...
478 Changes between versions 0.4b5 and 0.4b6 (released October 1, 1999)
479 ===================================================================
481 1. Integrated multiple patches from RedHat, Debian and SuSE:
483 - tweak dump/itime.c to not try to read dumpdates if the 'u' option
485 - several fixes in the man pages.
486 - update the default tape device to /dev/st0.
487 - many updates for Linux Alpha (byte ordering, size_t etc).
489 - use environment variable for TMPDIR (instead of /tmp).
490 - use sigjmp_buf instead of jmp_buf (RedHat bug #3260).
491 - workaround egcs bug (RedHat bugs #4281 and #2989).
492 - wire $(OPT) throughout Makefile's.
494 2. Upgrade the dump revision to 1, making possible to dump filesystems
495 made with e2fsprogs-1.15 or newer. Nothing seems to break...
497 3. Fix some compile warnings, prototype all functions.
499 4. Use glibc err/glob instead of internal compatibility
500 routines (only if available).
502 5. Fix a compile error on Linux 2.2.7 / libc5 (5.4.44) (patch provided
503 by Bernhard Sadlowski <sadlowsk@mathematik.uni-bielefeld.de>).
505 Changes between versions 0.4b4 and 0.4b5 (released September 22, 1999)
506 ======================================================================
508 1. Integrated the changes from FreeBSD-3.1-RELEASE
509 (mostly bug fixes, buffer overruns, dump has now an "automatic
510 tape length calculation" flag, dump/restore can use kerberos now
511 (this is NOT tested), use environment variables for TAPE and
514 2. Integrated three RedHat patches ("glibc", "kernel" and "bread" patches)
516 3. Corrected a bug in restore when using 'C' option with multi-volumes
517 tapes (files splited accros two tapes give "size changed" errors
520 4. Corrected the long standing bug when dumping multiple tapes.
521 This works for me, needs further testing.
523 Changes between versions 0.4b3 and 0.4b4 (released January 17, 1997)
524 ====================================================================
526 1. Dump now runs correctly on kernels 2.1.x
527 Fix made by Gerald Peters <gapeters@worldnet.att.net>
529 Changes between versions 0.4b2 and 0.4b3
530 ========================================
532 1. Use realpath() if available
536 Changes between versions 0.4b1 and 0.4b2
537 ========================================
539 1. Fixed the bug fix from Greg Lutz (I had made a mistake when integrating
542 2. Fixed restore to make it able to read FreeBSD 2.x dumps again
544 3. Fixed configure.in to correctly handle --enable-rmt
546 Changes between versions 0.3 and 0.4b1
547 ======================================
549 1. Integrated the changes from 4.4BSD-Lite2
551 2. Integrated the patches from Debian and Red Hat
553 3. Portability changes: use the __u32, __u16, __s32, and __s16 types
555 4. Changed dump to use the Ext2fs library to get block addresses. This
556 should solve the endianness problem on SparcLinux.
558 5. Created a configure.in file (shamelessly stolen from the e2fsprogs
559 distribution's one) to use autoconf
561 6. Fixed a few minor bugs
563 Changes between versions 0.2e and 0.2f
564 ======================================
566 1. Added the creation of named pipes (fifos) in restore.
568 2. Added the -N flag in the restore manual page.
570 3. Added the file kernel.patch which contains the llseek() optimization
571 patch for 1.2.x kernels.
573 4. Fixed a bug in the restoration of symbolic links: owner and group were
576 5. Integrated some changes from FreeBSD 2.2.
578 6. Added a call to ftruncate() after restoring each file to restore
579 correctly files ending by a hole.
581 Changes between versions 0.2d and 0.2e
582 ======================================
584 1. Fixed a bug in the "set owner/mode" process. Every file was restored
585 with owner = root (0) and group = root/wheel/whatever (0).
587 Changes between versions 0.2c and 0.2d
588 ======================================
590 1. Dump is now able to backup 2GB+ filesystems.
592 2. Dump and restore can now be linked as static binaries.
594 Changes between versions 0.2b and 0.2c
595 ======================================
597 1. Fixed a bug when dumping ``slow'' (i.e. normal) symbolic links.
599 Changes between versions 0.2a and 0.2b
600 ======================================
602 1. Really fixed the bug that I should have corrected in 0.2a.
604 2. Enabled optimization again.
606 Changes between versions 0.2 and 0.2a
607 =====================================
609 1. Disabled the optimization during compilation.
611 Changes between versions 0.1 and 0.2
612 ====================================
614 1. Fixed a bug in fstab.c which caused a null pointer to be stored in
615 the fs_type field (actually, I modified the file fstab.c to make it
616 use the mntent functions).
618 2. Dump and restore now use a 4.3 BSD compatible dump format. Backups
619 made by dump should be readable by the BSD restore and backups made
620 by the BSD dump should be readable by restore. Unfortunately, this
621 means that the dump format has changed between version 0.1 and version
624 3. Dump is now able to backup a subtree, it is no longer limited to whole
625 filesystems like the BSD version.
627 4. Dump now uses ext2_llseek() so it is able to backup filesystems bigger
630 Changes between versions 0.0 and 0.1
631 ====================================
633 1. Now create links rdump and rrestore during the `make install' step.
635 2. Linux port specific bugs added to the manual pages
637 3. Incorrect estimation of the number of tapes blocks fixed when doing
640 4. Better ls-like format in restore in interactive mode.