]> git.wh0rd.org - dump.git/blob - CHANGES
139b14ad16b9c7569c3d5fb761cffa5e4c6c11d0
[dump.git] / CHANGES
1 $Id: CHANGES,v 1.65 2000/06/25 18:42:39 stelian Exp $
2
3 Changes between versions 0.4b17 and 0.4b18 (released ??????????????)
4 ====================================================================
5
6 1. Fixed a potential buffer overflow in restore. Thanks
7 to Stan Bubrouski <satan@fastdial.net> for reporting
8 the bug.
9
10 2. Fixed a readline-related bug which prevented
11 'cat DUMPFILE | restore -i -f -' from working. Thanks
12 to Charles Karney <karney@users.sourceforge.net>
13 for the bug report.
14
15 3. Changed a few "panic" into "exit", causing restore to
16 be more stable against some attacks (like the last one
17 reported on Bugtraq, although the last version of restore
18 was not vulnerable - just dumped core). Thanks to
19 Andreas Hasenack <andreas@conectiva.com.br> for reporting
20 the bugs to me.
21
22 Changes between versions 0.4b16 and 0.4b17 (released June 1st, 2000)
23 ====================================================================
24
25 1. The -F script is called now *only* at the end of a tape,
26 not at the end of the dump. Thanks to Christian Haul
27 <haul@informatik.tu-darmstadt.de> for the bug report.
28
29 Normally, the device name and the current volume number
30 are passed on the command line of the script. However,
31 if you want the old style script (with no arguments
32 passed on the command line), you can enable it in
33 configure with the --enable-oldstylefscript.
34
35 2. Use posix signal handling to preserve dump functionality
36 with libc5. Thanks to Jeff Johnson <jbj@redhat.com> for
37 the patch.
38
39 3. Made the exit code of restore in case of a 'C'ompare
40 command reflect the result of the compare. An exit status
41 of 0 means the dump archive is correct, 1 means tape errors,
42 2 means that some files were modified. Thanks to Kenneth Porter
43 <shiva@well.com> for the suggestion.
44
45 4. Made (finally) quotes work in interactive restore.
46
47 5. Small fixes in order to allow dump to compile with a
48 really old e2fsprogs version. Thanks to Ian Zimmerman
49 <itz@speakeasy.org> for the bug report.
50
51 6. Add GNU readline capabilities to interactive restore.
52 Use configure's --enable-readline flag to enable this feature.
53 Thanks to Patrik Schindler <poc@pocnet.net> for the
54 suggestion.
55
56 7. Do the compare on the fly in restore 'C' mode (this will
57 allow not to exhaust the available /tmp space when
58 ccmparing large files). Thanks to Kenneth Porter
59 <shiva@well.com> for the suggestion.
60
61 Changes between versions 0.4b15 and 0.4b16 (released March 11, 2000)
62 ====================================================================
63
64 1. Fixed some several duplicate 'DUMP: DUMP:' in the
65 output of dump.
66
67 2. Corrected the estimation of blocks to dump. Note that
68 this is still wrong for large directory inodes, due
69 to the size of a BSD directory entry on the tape
70 (estimation is based on the size of the ext2 directory,
71 which is smaller because it doesn't contain the null
72 character at the end).
73
74 3. Made dump report the total number of MB written to
75 tape at the end of run. Thanks to W. Reilly Cooley
76 <wcooley@nakedape.cc> for the patch.
77
78 4. Added the -X option to restore giving the possibility
79 to read the names of the files to be extracted/listed
80 from a text file (in addition of the command line).
81 Thanks to Dejan Muhamedagic <dejan@quant-x.com> for the
82 patch.
83
84 5. Added the device name and the current volume number
85 as arguments to the end of tape script (dump -F option).
86
87 6. Made the multi-volume dump work again (after having
88 broken it in 0.4b15).
89
90 Changes between versions 0.4b14 and 0.4b15 (released March 2, 2000)
91 ===================================================================
92
93 1. Added a prompt command in interactive restore mode. Thanks
94 to Andreas Dilger <adilger@home.com> for the patch.
95
96 2. Fixed a buffer overflow problem in dump (caused by
97 not checking the size of the filesystem parameter).
98 Thanks to Kim Yong-jun <loveyou@hackerslab.org> for
99 reporting this on Bugtraq (and to several dump users
100 who forwarded me his mail).
101
102 3. Added the '-F script' option to dump in order to
103 launch a script at the end of each tape (to be used
104 with a tape changer, or to notify the sysadmin by
105 pager etc.).
106
107 4. Fixed a bug in restore compare code caused by the changes
108 I made in 0.4b14.
109
110 5. Fixed the treatment of options using the old BSD syntax
111 in both dump and restore.
112
113 Changes between versions 0.4b13 and 0.4b14 (released February 10, 2000)
114 =======================================================================
115
116 1. Fixed a bug in dump which may have caused invalid deleted
117 directories to be dumped out if they were deleted after the
118 mapping pass. This could occure on active filesystem and lead
119 to heap corruption (causing dump malfunction in many possible ways).
120 Thanks to Jan Sanislo <oystr@cs.washington.edu> for finding this
121 bug and submitting the patch.
122
123 2. Fixed the handling of the filesystem parameter in dump. A
124 '/mnt/disk' parameter caused the disk contents to be dumped,
125 but a '/mnt/disk/' parameter caused the mountpoint directory
126 to be dumped (generally an empty directory).
127
128 3. Improved the output of dump in order to tell which directory
129 it is currently dumping (when dumping a subtree).
130
131 4. Added the '-e' exclude inode option to dump. Thanks to
132 Isaac Chuang <ike@isl.stanford.edu> for contributing with the patch.
133
134 5. Added a REPORTING-BUGS file in order to provide a guide
135 on how to correctly report a bug in dump/restore.
136
137 6. Fixed a restore problem when restoring a hard link to an inode
138 having the immutable or append-only attribute set. Thanks to
139 Ambrose Li <acli@mingpaoxpress.com> for submitting the patch.
140
141 7. Fixed a compatibility problem between dumps made on little
142 endian machines (the format was correct) and big endian
143 machines (the format was incorrect). This fix break the
144 compatibility with the older dumps made on big endian
145 machines (sparc, m86k, ppc etc). For the first time in
146 linux dump's history (I believe), the dumps made by *BSD,
147 Linux/alpha, Linux/sparc and Linux/x86 are compatible,
148 so interchangeable. Thanks to Rob Cermak
149 <cermak@ahab.rutgers.edu> for submitting the bug and
150 helping me test the fix.
151
152 8. Fixed the way dump reports the remaining percent/time, if
153 the number of blocks actually dumped exceeds the estimated
154 number of blocks. Thanks to Jean-Paul van der Jagt
155 <jeanpaul@dutepp0.et.tudelft.nl> for reporting the bug.
156
157 Changes between versions 0.4b12 and 0.4b13 (released January 21, 2000)
158 ======================================================================
159
160 1. Small Y2K fix in the man pages :). Thanks to Bernhard Sadlowski
161 <sadlowsk@Mathematik.Uni-Bielefeld.DE> for reporting the bug.
162
163 2. Removed the requirement to build the RPM as root from the
164 spec file. Thanks to Christian Weisgerber
165 <naddy@mips.rhein-neckar.de> for submitting this.
166
167 3. Fixed a bug in dump related to the 'filetype' feature of ext2fs,
168 causing dump to block when dumping really huge filesystems.
169 Many thanks to Patrik Schindler <poc@pocnet.net> for
170 helping me find this bug.
171
172 4. Fixed the treatment for an interrupt signal when dump access
173 the remote tape through RSH. Thanks to Christian Weisgerber
174 <naddy@mips.rhein-neckar.de> for providing the patch.
175
176 5. Fixed a bug which was causing dump/restore to display
177 garbage characters instead of the remote host name.
178
179 Changes between versions 0.4b11 and 0.4b12 (released January 8, 2000)
180 =====================================================================
181
182 1. Small fix in the dump man page. Thanks to Thorsten Kukuk
183 <kukuk@suse.de> for submitting the patch.
184
185 2. Fix for the exit code when using the size estimate option of
186 dump. Thanks to Matti Taskinen <mkt@rni.helsinki.fi> for
187 submitting the patch.
188
189 3. Handle EINTR in atomical reads/writes in dump, which was causing
190 dump to fail on some systems. Thanks to Eric Jergensen
191 <eric@dvns.com> for reporting the bug and submitting the patch.
192
193 4. Handle more than 16 characters for the device names in dumpdates.
194 (up to 255 now). Thanks to Rainer Clasen <bj@ncc.cicely.de> for
195 tracking down the problem and proposing the solution.
196
197 5. Fixed a bug in dump which prevented the creation of the
198 dumpdates file when doing a 0-level dump without already
199 having a dumpdates file. Thanks to Patrik Schindler
200 <poc@pocnet.net> for reporting the bug.
201
202 6. Changed the way dump 'S' flag reports the size estimate
203 from number of blocks into bytes (making it compatible
204 with the Solaris version, and simplifying things for
205 amanda users). Thanks to Jason L Tibbitts III
206 <tibbs@math.uh.edu> for reporting the bug.
207
208 7. Fixed a compatibility problem in linux/alpha dump tape format.
209 Now the linux/alpha dump are (again) compatible with the
210 other dump formats. But this breaks compatibility with
211 older dumps made on alpha. Thanks to Mike Tibor
212 <tibor@lib.uaa.alaska.edu> for helping me in finding this bug.
213
214 Changes between versions 0.4b10 and 0.4b11 (released December 5, 1999)
215 ======================================================================
216
217 1. Added a '--enable-kerberos' to configure.
218
219 2. Added a 'S' option to dump which determines the amount of space
220 that is needed to perform the dump without actually doing it, similar
221 to the Sun's ufsdump 'S' option. Patch contributed by Rob Cermak
222 <cermak@ahab.rutgers.edu>.
223
224 3. Added a 'M' multi-volume option to dump and restore which enables
225 dumping to multiple files (useful when dumping to an ext2
226 partition to make several dump files in order to bypass the 2GB
227 file size limitation). The 'f' argument is treated as a prefix and
228 the output files will be named <prefix>001, <prefix>002 etc. With
229 the 'M' flag, restore automatically selects the right file without
230 asking to enter a new tape each time.
231
232 4. Fixed a memory leak which was causing dump to grow very big
233 (270MB when dumping a 10GB filesystem...). Thanks to Jason
234 Fearon <jasonf@netrider.org.au> for reporting the bug.
235
236 Changes between versions 0.4b9 and 0.4b10 (released November 21, 1999)
237 ======================================================================
238
239 1. Make configure test if the system glob routines support
240 extended syntax (ALTDIRFUNC). If not, use the internal glob
241 routines instead of system ones. Thanks to Bernhard Sadlowski
242 <sadlowsk@Mathematik.Uni-Bielefeld.DE> for reporting the bug
243 and helping me resolve this and other minor libc5 compiling
244 glitches.
245
246 2. Fix a problem when dumping a ext2fs with the 'filetype'
247 feature enabled. Thanks to Patrick J. LoPresti
248 <patl@cag.lcs.mit.edu> for reporting the bug and to
249 Theodore Y. Ts'o <tytso@mit.edu> for providing the patch.
250
251 3. Made the nodump flag work on directories. A directory which
252 has the nodump flag gets never dumped, regardless of its
253 contents.
254
255 4. Integrate a patch from Jeremy Fitzhardinge <jeremy@goop.org>
256 which allows dump on an active ext3 filesystem. However, this
257 is a "quick and dirty" patch which enables backup of an ext3
258 filesystem through the ext2 compatibility (by ignoring the
259 NEEDS_RECOVERY bit). The journal file is not recognized and
260 it is dumped (it should not).
261
262 5. Test the superblock compatibility flags when dumping, in order
263 to be sure that we know how to deal with specific features.
264
265 Changes between versions 0.4b8 and 0.4b9 (released November 5, 1999)
266 ====================================================================
267
268 1. Use lchown instead of chown, fixing a possible security problem
269 when restoring symlinks (a malicious user could use this
270 to deliberately corrupt the ownership of important system files).
271 Thanks to Chris Siebenmann <cks@utcc.utoronto.ca> for detecting
272 this and providing the patch.
273
274 Changes between versions 0.4b7 and 0.4b8 (released November 3, 1999)
275 ====================================================================
276
277 1. Put dump sources under CVS, added Id tags in all files so
278 one can use 'ident' on binary files.
279
280 2. Added the dump/restore version in the usage text so one can
281 easily verify the version he is using.
282
283 3. Small patch from Nuno Oliveira <nuno@eq.uc.pt> which fixes
284 a va_start/va_end problem on linux-ppc (always call va_start
285 va_end in pairs each time we use a vararg function).
286
287 4. Added again the DT_* constants because old libc does not
288 contain them :(. Thanks to Eric Maisonobe <virnet@nat.fr>
289 for submitting the bug report.
290
291 5. Use ext2fs_llseek instead of llseek. With recent e2fsprogs
292 this should enable dumping big (huge) filesystems.
293
294 6. Added the RSH environment variable in order to be able to
295 use a rsh replacement like ssh when doing remote backups (and
296 bypass the security limitations of rcmd). Now you can do remote
297 backups without being root (or making dump setuid root).
298
299 7. Modified again the way dumpdates works. For incremental dumps,
300 we need to read dumpdates even if we are not using 'u' option.
301 Thanks to Bdale Garbee <bdale@gag.com> for his ideas on how
302 this should work.
303
304 Changes between versions 0.4b6 and 0.4b7 (released October 8, 1999)
305 ===================================================================
306
307 1. Removed the 'k' flag from the restore 'about' text if kerberos
308 was not compiled in.
309
310 2. Prototyped (f)setflags from e2fsprogs and corrected the calls
311 to them (fsetflags takes a char*, setflags an open fd!).
312
313 3. (f)setflags is called only if the flags aren't empty. If the
314 file is a special file, a warning is printed, because changing
315 flags implies opening the device. Normally, a special file
316 should not have any flag... (Debian bug #29775, patch provided
317 by Abhijit Dasgupta <abhijit@ans.net>).
318
319 4. Made possible to dump a file system not mentioned in /etc/fstab.
320 (Debian bug #11904, patch provided by Eirik Fuller <eirik@netcom.com>).
321
322 5. Changed the default behaviour to not create dumpdates
323 unless 'u' option is specified. Removed the old "debian-patch"
324 which provided the same thing. (Debian bug #38136, #33818).
325
326 6. Removed all those dump*announce, since they were getting old...
327
328 7. Added warning messages if dumpdates does not exist and
329 when an inferior level dump does not exist (except for a level 0
330 dump).
331
332 8. Debugged the glob calls in interactive mode: restore used a
333 dirent struct which was different from the /usr/include/dirent.h
334 one (this used to work, is it a glibc2 change?), so none of the
335 compat glob (which used /usr/include/dirent.h) or the system glob
336 worked. Restore use now the system dirent (and the system
337 DT_* constants), which are compatible with BSD ones.
338
339 9. Added a configure flag (--with-dumpdatespath) to specify
340 the location of dumpdates file. By default, it is
341 /etc/dumpdates.
342
343 10. Added the "AUTHOR" and "AVAILABILITY" sections and
344 included the current date/version in man pages.
345
346 11. Corrected the estimation of remaining time when
347 the operator doesn't change the tapes quickly enough. This
348 was an old bug, I thought I corrected it, and discovered
349 that in fact it was corrected in two different places, so
350 the results canceled each other...
351
352 Changes between versions 0.4b5 and 0.4b6 (released October 1, 1999)
353 ===================================================================
354
355 1. Integrated multiple patches from RedHat, Debian and SuSE:
356
357 - tweak dump/itime.c to not try to read dumpdates if the 'u' option
358 isn't specified.
359 - several fixes in the man pages.
360 - update the default tape device to /dev/st0.
361 - many updates for Linux Alpha (byte ordering, size_t etc).
362 - buffer overruns.
363 - use environment variable for TMPDIR (instead of /tmp).
364 - use sigjmp_buf instead of jmp_buf (RedHat bug #3260).
365 - workaround egcs bug (RedHat bugs #4281 and #2989).
366 - wire $(OPT) throughout Makefile's.
367
368 2. Upgrade the dump revision to 1, making possible to dump filesystems
369 made with e2fsprogs-1.15 or newer. Nothing seems to break...
370
371 3. Fix some compile warnings, prototype all functions.
372
373 4. Use glibc err/glob instead of internal compatibility
374 routines (only if available).
375
376 5. Fix a compile error on Linux 2.2.7 / libc5 (5.4.44) (patch provided
377 by Bernhard Sadlowski <sadlowsk@mathematik.uni-bielefeld.de>).
378
379 Changes between versions 0.4b4 and 0.4b5 (released September 22, 1999)
380 ======================================================================
381
382 1. Integrated the changes from FreeBSD-3.1-RELEASE
383 (mostly bug fixes, buffer overruns, dump has now an "automatic
384 tape length calculation" flag, dump/restore can use kerberos now
385 (this is NOT tested), use environment variables for TAPE and
386 RMT etc.).
387
388 2. Integrated three RedHat patches ("glibc", "kernel" and "bread" patches)
389
390 3. Corrected a bug in restore when using 'C' option with multi-volumes
391 tapes (files splited accros two tapes give "size changed" errors
392 when comparing).
393
394 4. Corrected the long standing bug when dumping multiple tapes.
395 This works for me, needs further testing.
396
397 Changes between versions 0.4b3 and 0.4b4 (released January 17, 1997)
398 ====================================================================
399
400 1. Dump now runs correctly on kernels 2.1.x
401 Fix made by Gerald Peters <gapeters@worldnet.att.net>
402
403 Changes between versions 0.4b2 and 0.4b3
404 ========================================
405
406 1. Use realpath() if available
407
408 2. Report statistics
409
410 Changes between versions 0.4b1 and 0.4b2
411 ========================================
412
413 1. Fixed the bug fix from Greg Lutz (I had made a mistake when integrating
414 the patch)
415
416 2. Fixed restore to make it able to read FreeBSD 2.x dumps again
417
418 3. Fixed configure.in to correctly handle --enable-rmt
419
420 Changes between versions 0.3 and 0.4b1
421 ======================================
422
423 1. Integrated the changes from 4.4BSD-Lite2
424
425 2. Integrated the patches from Debian and Red Hat
426
427 3. Portability changes: use the __u32, __u16, __s32, and __s16 types
428
429 4. Changed dump to use the Ext2fs library to get block addresses. This
430 should solve the endianness problem on SparcLinux.
431
432 5. Created a configure.in file (shamelessly stolen from the e2fsprogs
433 distribution's one) to use autoconf
434
435 6. Fixed a few minor bugs
436
437 Changes between versions 0.2e and 0.2f
438 ======================================
439
440 1. Added the creation of named pipes (fifos) in restore.
441
442 2. Added the -N flag in the restore manual page.
443
444 3. Added the file kernel.patch which contains the llseek() optimization
445 patch for 1.2.x kernels.
446
447 4. Fixed a bug in the restoration of symbolic links: owner and group were
448 not restored.
449
450 5. Integrated some changes from FreeBSD 2.2.
451
452 6. Added a call to ftruncate() after restoring each file to restore
453 correctly files ending by a hole.
454
455 Changes between versions 0.2d and 0.2e
456 ======================================
457
458 1. Fixed a bug in the "set owner/mode" process. Every file was restored
459 with owner = root (0) and group = root/wheel/whatever (0).
460
461 Changes between versions 0.2c and 0.2d
462 ======================================
463
464 1. Dump is now able to backup 2GB+ filesystems.
465
466 2. Dump and restore can now be linked as static binaries.
467
468 Changes between versions 0.2b and 0.2c
469 ======================================
470
471 1. Fixed a bug when dumping ``slow'' (i.e. normal) symbolic links.
472
473 Changes between versions 0.2a and 0.2b
474 ======================================
475
476 1. Really fixed the bug that I should have corrected in 0.2a.
477
478 2. Enabled optimization again.
479
480 Changes between versions 0.2 and 0.2a
481 =====================================
482
483 1. Disabled the optimization during compilation.
484
485 Changes between versions 0.1 and 0.2
486 ====================================
487
488 1. Fixed a bug in fstab.c which caused a null pointer to be stored in
489 the fs_type field (actually, I modified the file fstab.c to make it
490 use the mntent functions).
491
492 2. Dump and restore now use a 4.3 BSD compatible dump format. Backups
493 made by dump should be readable by the BSD restore and backups made
494 by the BSD dump should be readable by restore. Unfortunately, this
495 means that the dump format has changed between version 0.1 and version
496 0.2 :-(
497
498 3. Dump is now able to backup a subtree, it is no longer limited to whole
499 filesystems like the BSD version.
500
501 4. Dump now uses ext2_llseek() so it is able to backup filesystems bigger
502 than 2 GB.
503
504 Changes between versions 0.0 and 0.1
505 ====================================
506
507 1. Now create links rdump and rrestore during the `make install' step.
508
509 2. Linux port specific bugs added to the manual pages
510
511 3. Incorrect estimation of the number of tapes blocks fixed when doing
512 incremental backups.
513
514 4. Better ls-like format in restore in interactive mode.