1 $Id: CHANGES,v 1.83 2000/09/26 12:34:51 stelian Exp $
3 Changes between versions 0.4b19 and 0.4b20 (released ???????????????)
4 =====================================================================
6 1. Fixed a small compilation problem due to a change
7 in the definintion of the struct sigaction in
8 glibc 2.0 / libc5. Thanks to Gunther Schlegel
9 <schlegel@riege.de> for reporting the bug and to
10 Dave Platt <dplatt@snulbug.mtview.ca.us> for suggesting
13 2. Modified the label and uuid analysis in order to be
14 self-contained (independant of kernel/libc headers). This
15 should fix the compile with older kernel/libc headers and
16 will preserve the functionality. Thanks to Bernhard Erdmann
17 <bernhard.erdmann@gmx.de> for reporting the bug.
19 3. The 'exclude inode' option, if applied to a directory
20 inode, excludes now correctly all the leaves of this
21 directory. Thanks to John R. Dennison
22 <gerdesas@users.sourceforge.net> for reporting the bug.
24 4. Fixed the '-e' option to disable the possibility
25 to exclude the root inode (which causes the dumps to
26 be unreadable by restore). Prevented array overflow
27 when multiple -e options are used.
29 5. Fixed dump to correctly interpret a filesystem argument
30 which is a mountpoint and it is not an absolute path
31 (as specified in the fstab). Thanks to Bernhard R. Erdmann
32 <be@berdmann.de> for reporting the bug.
34 Changes between versions 0.4b18 and 0.4b19 (released August 20, 2000)
35 =====================================================================
37 1. Fixed the signal handling in dump (which I broke in 0.4b17)
38 which was causing several strange dump failures (dump
39 hanged or segmentation faults etc).
41 2. Specified the default blocksize in the dump man page.
43 3. Changed two info messages of restore to be written on stdout
44 instead of stderr in order to leave stderr only for errors
45 or warnings. Thanks to Stephen Atwell
46 <satwell@urbana.css.mot.com> for the suggestion.
48 4. Corrected an off by one calculation which prevented
49 dumping a directory having a 1 character name length.
50 Thanks to Bernhard Erdmann <bernhard.erdmann@gmx.de>
51 for reporting the bug.
53 5. Reinforce dump in order to allow only level 0 and no
54 -u option when dumping a subdirectory, and document
55 this in the man page. Thanks to Bernhard Erdmann
56 <bernhard.erdmann@gmx.de> for reporting the bug.
58 6. Fixed a small harmless bug in restore which caused
59 in some conditions hard links to be restored several
60 times (therefore generation some warning messages).
61 Thanks to Kenneth Porter <shiva@well.com> for
64 7. Updated the RPM spec file to the latest RedHat version,
65 providing FHS packaging and other cosmetic changes.
66 You will need RPM version >= 3.0.5 in order to build the RPM.
68 8. Updated the configure script to check for libtermcap
69 before checking for libreadline, because we need this
70 library in order to compile the readline support.
72 9. Made dump understand the LABEL= and UUID= notation
73 both in /etc/fstab and on the command line. Note that
74 you will need the /proc filesystem in order to use
75 these notations. Thanks to Erik Troan <ewt@redhat.com>
76 for providing the patch.
78 Changes between versions 0.4b17 and 0.4b18 (released June 30, 2000)
79 ===================================================================
81 1. Fixed a potential buffer overflow in restore. Thanks
82 to Stan Bubrouski <satan@fastdial.net> for reporting
85 2. Fixed a readline-related bug which prevented
86 'cat DUMPFILE | restore -i -f -' from working. Thanks
87 to Charles Karney <karney@users.sourceforge.net>
90 3. Changed a few "panic" into "exit", causing restore to
91 be more stable against some attacks (like the last one
92 reported on Bugtraq, although the last version of restore
93 was not vulnerable - just dumped core). Thanks to
94 Andreas Hasenack <andreas@conectiva.com.br> for reporting
97 4. Removed the suid-root bit on dump and restore in the
98 default build (and generated RPMs). It should be safer
99 now. Users who need the suid-root capabilities in order
100 to make network backups should read first the man page
101 of dump and enable the suid bit manually.
103 5. Added -ltermcap to the compile parameters for restore
104 when using readline support, in order to make the compile
105 process work on some systems (SuSE ?). Thanks to
106 Patrik Schindler <poc@pocnet.net> for reporting the bug.
108 Changes between versions 0.4b16 and 0.4b17 (released June 1st, 2000)
109 ====================================================================
111 1. The -F script is called now *only* at the end of a tape,
112 not at the end of the dump. Thanks to Christian Haul
113 <haul@informatik.tu-darmstadt.de> for the bug report.
115 Normally, the device name and the current volume number
116 are passed on the command line of the script. However,
117 if you want the old style script (with no arguments
118 passed on the command line), you can enable it in
119 configure with the --enable-oldstylefscript.
121 2. Use posix signal handling to preserve dump functionality
122 with libc5. Thanks to Jeff Johnson <jbj@redhat.com> for
125 3. Made the exit code of restore in case of a 'C'ompare
126 command reflect the result of the compare. An exit status
127 of 0 means the dump archive is correct, 1 means tape errors,
128 2 means that some files were modified. Thanks to Kenneth Porter
129 <shiva@well.com> for the suggestion.
131 4. Made (finally) quotes work in interactive restore.
133 5. Small fixes in order to allow dump to compile with a
134 really old e2fsprogs version. Thanks to Ian Zimmerman
135 <itz@speakeasy.org> for the bug report.
137 6. Add GNU readline capabilities to interactive restore.
138 Use configure's --enable-readline flag to enable this feature.
139 Thanks to Patrik Schindler <poc@pocnet.net> for the
142 7. Do the compare on the fly in restore 'C' mode (this will
143 allow not to exhaust the available /tmp space when
144 ccmparing large files). Thanks to Kenneth Porter
145 <shiva@well.com> for the suggestion.
147 Changes between versions 0.4b15 and 0.4b16 (released March 11, 2000)
148 ====================================================================
150 1. Fixed some several duplicate 'DUMP: DUMP:' in the
153 2. Corrected the estimation of blocks to dump. Note that
154 this is still wrong for large directory inodes, due
155 to the size of a BSD directory entry on the tape
156 (estimation is based on the size of the ext2 directory,
157 which is smaller because it doesn't contain the null
158 character at the end).
160 3. Made dump report the total number of MB written to
161 tape at the end of run. Thanks to W. Reilly Cooley
162 <wcooley@nakedape.cc> for the patch.
164 4. Added the -X option to restore giving the possibility
165 to read the names of the files to be extracted/listed
166 from a text file (in addition of the command line).
167 Thanks to Dejan Muhamedagic <dejan@quant-x.com> for the
170 5. Added the device name and the current volume number
171 as arguments to the end of tape script (dump -F option).
173 6. Made the multi-volume dump work again (after having
174 broken it in 0.4b15).
176 Changes between versions 0.4b14 and 0.4b15 (released March 2, 2000)
177 ===================================================================
179 1. Added a prompt command in interactive restore mode. Thanks
180 to Andreas Dilger <adilger@home.com> for the patch.
182 2. Fixed a buffer overflow problem in dump (caused by
183 not checking the size of the filesystem parameter).
184 Thanks to Kim Yong-jun <loveyou@hackerslab.org> for
185 reporting this on Bugtraq (and to several dump users
186 who forwarded me his mail).
188 3. Added the '-F script' option to dump in order to
189 launch a script at the end of each tape (to be used
190 with a tape changer, or to notify the sysadmin by
193 4. Fixed a bug in restore compare code caused by the changes
196 5. Fixed the treatment of options using the old BSD syntax
197 in both dump and restore.
199 Changes between versions 0.4b13 and 0.4b14 (released February 10, 2000)
200 =======================================================================
202 1. Fixed a bug in dump which may have caused invalid deleted
203 directories to be dumped out if they were deleted after the
204 mapping pass. This could occure on active filesystem and lead
205 to heap corruption (causing dump malfunction in many possible ways).
206 Thanks to Jan Sanislo <oystr@cs.washington.edu> for finding this
207 bug and submitting the patch.
209 2. Fixed the handling of the filesystem parameter in dump. A
210 '/mnt/disk' parameter caused the disk contents to be dumped,
211 but a '/mnt/disk/' parameter caused the mountpoint directory
212 to be dumped (generally an empty directory).
214 3. Improved the output of dump in order to tell which directory
215 it is currently dumping (when dumping a subtree).
217 4. Added the '-e' exclude inode option to dump. Thanks to
218 Isaac Chuang <ike@isl.stanford.edu> for contributing with the patch.
220 5. Added a REPORTING-BUGS file in order to provide a guide
221 on how to correctly report a bug in dump/restore.
223 6. Fixed a restore problem when restoring a hard link to an inode
224 having the immutable or append-only attribute set. Thanks to
225 Ambrose Li <acli@mingpaoxpress.com> for submitting the patch.
227 7. Fixed a compatibility problem between dumps made on little
228 endian machines (the format was correct) and big endian
229 machines (the format was incorrect). This fix break the
230 compatibility with the older dumps made on big endian
231 machines (sparc, m86k, ppc etc). For the first time in
232 linux dump's history (I believe), the dumps made by *BSD,
233 Linux/alpha, Linux/sparc and Linux/x86 are compatible,
234 so interchangeable. Thanks to Rob Cermak
235 <cermak@ahab.rutgers.edu> for submitting the bug and
236 helping me test the fix.
238 8. Fixed the way dump reports the remaining percent/time, if
239 the number of blocks actually dumped exceeds the estimated
240 number of blocks. Thanks to Jean-Paul van der Jagt
241 <jeanpaul@dutepp0.et.tudelft.nl> for reporting the bug.
243 Changes between versions 0.4b12 and 0.4b13 (released January 21, 2000)
244 ======================================================================
246 1. Small Y2K fix in the man pages :). Thanks to Bernhard Sadlowski
247 <sadlowsk@Mathematik.Uni-Bielefeld.DE> for reporting the bug.
249 2. Removed the requirement to build the RPM as root from the
250 spec file. Thanks to Christian Weisgerber
251 <naddy@mips.rhein-neckar.de> for submitting this.
253 3. Fixed a bug in dump related to the 'filetype' feature of ext2fs,
254 causing dump to block when dumping really huge filesystems.
255 Many thanks to Patrik Schindler <poc@pocnet.net> for
256 helping me find this bug.
258 4. Fixed the treatment for an interrupt signal when dump access
259 the remote tape through RSH. Thanks to Christian Weisgerber
260 <naddy@mips.rhein-neckar.de> for providing the patch.
262 5. Fixed a bug which was causing dump/restore to display
263 garbage characters instead of the remote host name.
265 Changes between versions 0.4b11 and 0.4b12 (released January 8, 2000)
266 =====================================================================
268 1. Small fix in the dump man page. Thanks to Thorsten Kukuk
269 <kukuk@suse.de> for submitting the patch.
271 2. Fix for the exit code when using the size estimate option of
272 dump. Thanks to Matti Taskinen <mkt@rni.helsinki.fi> for
273 submitting the patch.
275 3. Handle EINTR in atomical reads/writes in dump, which was causing
276 dump to fail on some systems. Thanks to Eric Jergensen
277 <eric@dvns.com> for reporting the bug and submitting the patch.
279 4. Handle more than 16 characters for the device names in dumpdates.
280 (up to 255 now). Thanks to Rainer Clasen <bj@ncc.cicely.de> for
281 tracking down the problem and proposing the solution.
283 5. Fixed a bug in dump which prevented the creation of the
284 dumpdates file when doing a 0-level dump without already
285 having a dumpdates file. Thanks to Patrik Schindler
286 <poc@pocnet.net> for reporting the bug.
288 6. Changed the way dump 'S' flag reports the size estimate
289 from number of blocks into bytes (making it compatible
290 with the Solaris version, and simplifying things for
291 amanda users). Thanks to Jason L Tibbitts III
292 <tibbs@math.uh.edu> for reporting the bug.
294 7. Fixed a compatibility problem in linux/alpha dump tape format.
295 Now the linux/alpha dump are (again) compatible with the
296 other dump formats. But this breaks compatibility with
297 older dumps made on alpha. Thanks to Mike Tibor
298 <tibor@lib.uaa.alaska.edu> for helping me in finding this bug.
300 Changes between versions 0.4b10 and 0.4b11 (released December 5, 1999)
301 ======================================================================
303 1. Added a '--enable-kerberos' to configure.
305 2. Added a 'S' option to dump which determines the amount of space
306 that is needed to perform the dump without actually doing it, similar
307 to the Sun's ufsdump 'S' option. Patch contributed by Rob Cermak
308 <cermak@ahab.rutgers.edu>.
310 3. Added a 'M' multi-volume option to dump and restore which enables
311 dumping to multiple files (useful when dumping to an ext2
312 partition to make several dump files in order to bypass the 2GB
313 file size limitation). The 'f' argument is treated as a prefix and
314 the output files will be named <prefix>001, <prefix>002 etc. With
315 the 'M' flag, restore automatically selects the right file without
316 asking to enter a new tape each time.
318 4. Fixed a memory leak which was causing dump to grow very big
319 (270MB when dumping a 10GB filesystem...). Thanks to Jason
320 Fearon <jasonf@netrider.org.au> for reporting the bug.
322 Changes between versions 0.4b9 and 0.4b10 (released November 21, 1999)
323 ======================================================================
325 1. Make configure test if the system glob routines support
326 extended syntax (ALTDIRFUNC). If not, use the internal glob
327 routines instead of system ones. Thanks to Bernhard Sadlowski
328 <sadlowsk@Mathematik.Uni-Bielefeld.DE> for reporting the bug
329 and helping me resolve this and other minor libc5 compiling
332 2. Fix a problem when dumping a ext2fs with the 'filetype'
333 feature enabled. Thanks to Patrick J. LoPresti
334 <patl@cag.lcs.mit.edu> for reporting the bug and to
335 Theodore Y. Ts'o <tytso@mit.edu> for providing the patch.
337 3. Made the nodump flag work on directories. A directory which
338 has the nodump flag gets never dumped, regardless of its
341 4. Integrate a patch from Jeremy Fitzhardinge <jeremy@goop.org>
342 which allows dump on an active ext3 filesystem. However, this
343 is a "quick and dirty" patch which enables backup of an ext3
344 filesystem through the ext2 compatibility (by ignoring the
345 NEEDS_RECOVERY bit). The journal file is not recognized and
346 it is dumped (it should not).
348 5. Test the superblock compatibility flags when dumping, in order
349 to be sure that we know how to deal with specific features.
351 Changes between versions 0.4b8 and 0.4b9 (released November 5, 1999)
352 ====================================================================
354 1. Use lchown instead of chown, fixing a possible security problem
355 when restoring symlinks (a malicious user could use this
356 to deliberately corrupt the ownership of important system files).
357 Thanks to Chris Siebenmann <cks@utcc.utoronto.ca> for detecting
358 this and providing the patch.
360 Changes between versions 0.4b7 and 0.4b8 (released November 3, 1999)
361 ====================================================================
363 1. Put dump sources under CVS, added Id tags in all files so
364 one can use 'ident' on binary files.
366 2. Added the dump/restore version in the usage text so one can
367 easily verify the version he is using.
369 3. Small patch from Nuno Oliveira <nuno@eq.uc.pt> which fixes
370 a va_start/va_end problem on linux-ppc (always call va_start
371 va_end in pairs each time we use a vararg function).
373 4. Added again the DT_* constants because old libc does not
374 contain them :(. Thanks to Eric Maisonobe <virnet@nat.fr>
375 for submitting the bug report.
377 5. Use ext2fs_llseek instead of llseek. With recent e2fsprogs
378 this should enable dumping big (huge) filesystems.
380 6. Added the RSH environment variable in order to be able to
381 use a rsh replacement like ssh when doing remote backups (and
382 bypass the security limitations of rcmd). Now you can do remote
383 backups without being root (or making dump setuid root).
385 7. Modified again the way dumpdates works. For incremental dumps,
386 we need to read dumpdates even if we are not using 'u' option.
387 Thanks to Bdale Garbee <bdale@gag.com> for his ideas on how
390 Changes between versions 0.4b6 and 0.4b7 (released October 8, 1999)
391 ===================================================================
393 1. Removed the 'k' flag from the restore 'about' text if kerberos
396 2. Prototyped (f)setflags from e2fsprogs and corrected the calls
397 to them (fsetflags takes a char*, setflags an open fd!).
399 3. (f)setflags is called only if the flags aren't empty. If the
400 file is a special file, a warning is printed, because changing
401 flags implies opening the device. Normally, a special file
402 should not have any flag... (Debian bug #29775, patch provided
403 by Abhijit Dasgupta <abhijit@ans.net>).
405 4. Made possible to dump a file system not mentioned in /etc/fstab.
406 (Debian bug #11904, patch provided by Eirik Fuller <eirik@netcom.com>).
408 5. Changed the default behaviour to not create dumpdates
409 unless 'u' option is specified. Removed the old "debian-patch"
410 which provided the same thing. (Debian bug #38136, #33818).
412 6. Removed all those dump*announce, since they were getting old...
414 7. Added warning messages if dumpdates does not exist and
415 when an inferior level dump does not exist (except for a level 0
418 8. Debugged the glob calls in interactive mode: restore used a
419 dirent struct which was different from the /usr/include/dirent.h
420 one (this used to work, is it a glibc2 change?), so none of the
421 compat glob (which used /usr/include/dirent.h) or the system glob
422 worked. Restore use now the system dirent (and the system
423 DT_* constants), which are compatible with BSD ones.
425 9. Added a configure flag (--with-dumpdatespath) to specify
426 the location of dumpdates file. By default, it is
429 10. Added the "AUTHOR" and "AVAILABILITY" sections and
430 included the current date/version in man pages.
432 11. Corrected the estimation of remaining time when
433 the operator doesn't change the tapes quickly enough. This
434 was an old bug, I thought I corrected it, and discovered
435 that in fact it was corrected in two different places, so
436 the results canceled each other...
438 Changes between versions 0.4b5 and 0.4b6 (released October 1, 1999)
439 ===================================================================
441 1. Integrated multiple patches from RedHat, Debian and SuSE:
443 - tweak dump/itime.c to not try to read dumpdates if the 'u' option
445 - several fixes in the man pages.
446 - update the default tape device to /dev/st0.
447 - many updates for Linux Alpha (byte ordering, size_t etc).
449 - use environment variable for TMPDIR (instead of /tmp).
450 - use sigjmp_buf instead of jmp_buf (RedHat bug #3260).
451 - workaround egcs bug (RedHat bugs #4281 and #2989).
452 - wire $(OPT) throughout Makefile's.
454 2. Upgrade the dump revision to 1, making possible to dump filesystems
455 made with e2fsprogs-1.15 or newer. Nothing seems to break...
457 3. Fix some compile warnings, prototype all functions.
459 4. Use glibc err/glob instead of internal compatibility
460 routines (only if available).
462 5. Fix a compile error on Linux 2.2.7 / libc5 (5.4.44) (patch provided
463 by Bernhard Sadlowski <sadlowsk@mathematik.uni-bielefeld.de>).
465 Changes between versions 0.4b4 and 0.4b5 (released September 22, 1999)
466 ======================================================================
468 1. Integrated the changes from FreeBSD-3.1-RELEASE
469 (mostly bug fixes, buffer overruns, dump has now an "automatic
470 tape length calculation" flag, dump/restore can use kerberos now
471 (this is NOT tested), use environment variables for TAPE and
474 2. Integrated three RedHat patches ("glibc", "kernel" and "bread" patches)
476 3. Corrected a bug in restore when using 'C' option with multi-volumes
477 tapes (files splited accros two tapes give "size changed" errors
480 4. Corrected the long standing bug when dumping multiple tapes.
481 This works for me, needs further testing.
483 Changes between versions 0.4b3 and 0.4b4 (released January 17, 1997)
484 ====================================================================
486 1. Dump now runs correctly on kernels 2.1.x
487 Fix made by Gerald Peters <gapeters@worldnet.att.net>
489 Changes between versions 0.4b2 and 0.4b3
490 ========================================
492 1. Use realpath() if available
496 Changes between versions 0.4b1 and 0.4b2
497 ========================================
499 1. Fixed the bug fix from Greg Lutz (I had made a mistake when integrating
502 2. Fixed restore to make it able to read FreeBSD 2.x dumps again
504 3. Fixed configure.in to correctly handle --enable-rmt
506 Changes between versions 0.3 and 0.4b1
507 ======================================
509 1. Integrated the changes from 4.4BSD-Lite2
511 2. Integrated the patches from Debian and Red Hat
513 3. Portability changes: use the __u32, __u16, __s32, and __s16 types
515 4. Changed dump to use the Ext2fs library to get block addresses. This
516 should solve the endianness problem on SparcLinux.
518 5. Created a configure.in file (shamelessly stolen from the e2fsprogs
519 distribution's one) to use autoconf
521 6. Fixed a few minor bugs
523 Changes between versions 0.2e and 0.2f
524 ======================================
526 1. Added the creation of named pipes (fifos) in restore.
528 2. Added the -N flag in the restore manual page.
530 3. Added the file kernel.patch which contains the llseek() optimization
531 patch for 1.2.x kernels.
533 4. Fixed a bug in the restoration of symbolic links: owner and group were
536 5. Integrated some changes from FreeBSD 2.2.
538 6. Added a call to ftruncate() after restoring each file to restore
539 correctly files ending by a hole.
541 Changes between versions 0.2d and 0.2e
542 ======================================
544 1. Fixed a bug in the "set owner/mode" process. Every file was restored
545 with owner = root (0) and group = root/wheel/whatever (0).
547 Changes between versions 0.2c and 0.2d
548 ======================================
550 1. Dump is now able to backup 2GB+ filesystems.
552 2. Dump and restore can now be linked as static binaries.
554 Changes between versions 0.2b and 0.2c
555 ======================================
557 1. Fixed a bug when dumping ``slow'' (i.e. normal) symbolic links.
559 Changes between versions 0.2a and 0.2b
560 ======================================
562 1. Really fixed the bug that I should have corrected in 0.2a.
564 2. Enabled optimization again.
566 Changes between versions 0.2 and 0.2a
567 =====================================
569 1. Disabled the optimization during compilation.
571 Changes between versions 0.1 and 0.2
572 ====================================
574 1. Fixed a bug in fstab.c which caused a null pointer to be stored in
575 the fs_type field (actually, I modified the file fstab.c to make it
576 use the mntent functions).
578 2. Dump and restore now use a 4.3 BSD compatible dump format. Backups
579 made by dump should be readable by the BSD restore and backups made
580 by the BSD dump should be readable by restore. Unfortunately, this
581 means that the dump format has changed between version 0.1 and version
584 3. Dump is now able to backup a subtree, it is no longer limited to whole
585 filesystems like the BSD version.
587 4. Dump now uses ext2_llseek() so it is able to backup filesystems bigger
590 Changes between versions 0.0 and 0.1
591 ====================================
593 1. Now create links rdump and rrestore during the `make install' step.
595 2. Linux port specific bugs added to the manual pages
597 3. Incorrect estimation of the number of tapes blocks fixed when doing
600 4. Better ls-like format in restore in interactive mode.