]> git.wh0rd.org - dump.git/blob - CHANGES
*** empty log message ***
[dump.git] / CHANGES
1 $Id: CHANGES,v 1.61 2000/06/01 18:51:22 stelian Exp $
2
3 Changes between versions 0.4b16 and 0.4b17 (released ??????????????)
4 ====================================================================
5
6 1. The -F script is called now *only* at the end of a tape,
7 not at the end of the dump. Thanks to Christian Haul
8 <haul@informatik.tu-darmstadt.de> for the bug report.
9
10 Normally, the device name and the current volume number
11 are passed on the command line of the script. However,
12 if you want the old style script (with no arguments
13 passed on the command line), you can enable it in
14 configure with the --enable-oldstylefscript.
15
16 2. Use posix signal handling to preserve dump functionality
17 with libc5. Thanks to Jeff Johnson <jbj@redhat.com> for
18 the patch.
19
20 3. Made the exit code of restore in case of a 'C'ompare
21 command reflect the result of the compare. An exit status
22 of 0 means the dump archive is correct, 1 means tape errors,
23 2 means that some files were modified. Thanks to Kenneth Porter
24 <shiva@well.com> for the suggestion.
25
26 4. Made (finally) quotes work in interactive restore.
27
28 5. Small fixes in order to allow dump to compile with a
29 really old e2fsprogs version. Thanks to Ian Zimmerman
30 <itz@speakeasy.org> for the bug report.
31
32 6. Add GNU readline capabilities to interactive restore.
33 Use configure's --enable-readline flag to enable this feature.
34 Thanks to Patrik Schindler <poc@pocnet.net> for the
35 suggestion.
36
37 7. Do the compare on the fly in restore 'C' mode (this will
38 allow not to exhaust the available /tmp space when
39 ccmparing large files). Thanks to Kenneth Porter
40 <shiva@well.com> for the suggestion.
41
42 Changes between versions 0.4b15 and 0.4b16 (released March 11, 2000)
43 ====================================================================
44
45 1. Fixed some several duplicate 'DUMP: DUMP:' in the
46 output of dump.
47
48 2. Corrected the estimation of blocks to dump. Note that
49 this is still wrong for large directory inodes, due
50 to the size of a BSD directory entry on the tape
51 (estimation is based on the size of the ext2 directory,
52 which is smaller because it doesn't contain the null
53 character at the end).
54
55 3. Made dump report the total number of MB written to
56 tape at the end of run. Thanks to W. Reilly Cooley
57 <wcooley@nakedape.cc> for the patch.
58
59 4. Added the -X option to restore giving the possibility
60 to read the names of the files to be extracted/listed
61 from a text file (in addition of the command line).
62 Thanks to Dejan Muhamedagic <dejan@quant-x.com> for the
63 patch.
64
65 5. Added the device name and the current volume number
66 as arguments to the end of tape script (dump -F option).
67
68 6. Made the multi-volume dump work again (after having
69 broken it in 0.4b15).
70
71 Changes between versions 0.4b14 and 0.4b15 (released March 2, 2000)
72 ===================================================================
73
74 1. Added a prompt command in interactive restore mode. Thanks
75 to Andreas Dilger <adilger@home.com> for the patch.
76
77 2. Fixed a buffer overflow problem in dump (caused by
78 not checking the size of the filesystem parameter).
79 Thanks to Kim Yong-jun <loveyou@hackerslab.org> for
80 reporting this on Bugtraq (and to several dump users
81 who forwarded me his mail).
82
83 3. Added the '-F script' option to dump in order to
84 launch a script at the end of each tape (to be used
85 with a tape changer, or to notify the sysadmin by
86 pager etc.).
87
88 4. Fixed a bug in restore compare code caused by the changes
89 I made in 0.4b14.
90
91 5. Fixed the treatment of options using the old BSD syntax
92 in both dump and restore.
93
94 Changes between versions 0.4b13 and 0.4b14 (released February 10, 2000)
95 =======================================================================
96
97 1. Fixed a bug in dump which may have caused invalid deleted
98 directories to be dumped out if they were deleted after the
99 mapping pass. This could occure on active filesystem and lead
100 to heap corruption (causing dump malfunction in many possible ways).
101 Thanks to Jan Sanislo <oystr@cs.washington.edu> for finding this
102 bug and submitting the patch.
103
104 2. Fixed the handling of the filesystem parameter in dump. A
105 '/mnt/disk' parameter caused the disk contents to be dumped,
106 but a '/mnt/disk/' parameter caused the mountpoint directory
107 to be dumped (generally an empty directory).
108
109 3. Improved the output of dump in order to tell which directory
110 it is currently dumping (when dumping a subtree).
111
112 4. Added the '-e' exclude inode option to dump. Thanks to
113 Isaac Chuang <ike@isl.stanford.edu> for contributing with the patch.
114
115 5. Added a REPORTING-BUGS file in order to provide a guide
116 on how to correctly report a bug in dump/restore.
117
118 6. Fixed a restore problem when restoring a hard link to an inode
119 having the immutable or append-only attribute set. Thanks to
120 Ambrose Li <acli@mingpaoxpress.com> for submitting the patch.
121
122 7. Fixed a compatibility problem between dumps made on little
123 endian machines (the format was correct) and big endian
124 machines (the format was incorrect). This fix break the
125 compatibility with the older dumps made on big endian
126 machines (sparc, m86k, ppc etc). For the first time in
127 linux dump's history (I believe), the dumps made by *BSD,
128 Linux/alpha, Linux/sparc and Linux/x86 are compatible,
129 so interchangeable. Thanks to Rob Cermak
130 <cermak@ahab.rutgers.edu> for submitting the bug and
131 helping me test the fix.
132
133 8. Fixed the way dump reports the remaining percent/time, if
134 the number of blocks actually dumped exceeds the estimated
135 number of blocks. Thanks to Jean-Paul van der Jagt
136 <jeanpaul@dutepp0.et.tudelft.nl> for reporting the bug.
137
138 Changes between versions 0.4b12 and 0.4b13 (released January 21, 2000)
139 ======================================================================
140
141 1. Small Y2K fix in the man pages :). Thanks to Bernhard Sadlowski
142 <sadlowsk@Mathematik.Uni-Bielefeld.DE> for reporting the bug.
143
144 2. Removed the requirement to build the RPM as root from the
145 spec file. Thanks to Christian Weisgerber
146 <naddy@mips.rhein-neckar.de> for submitting this.
147
148 3. Fixed a bug in dump related to the 'filetype' feature of ext2fs,
149 causing dump to block when dumping really huge filesystems.
150 Many thanks to Patrik Schindler <poc@pocnet.net> for
151 helping me find this bug.
152
153 4. Fixed the treatment for an interrupt signal when dump access
154 the remote tape through RSH. Thanks to Christian Weisgerber
155 <naddy@mips.rhein-neckar.de> for providing the patch.
156
157 5. Fixed a bug which was causing dump/restore to display
158 garbage characters instead of the remote host name.
159
160 Changes between versions 0.4b11 and 0.4b12 (released January 8, 2000)
161 =====================================================================
162
163 1. Small fix in the dump man page. Thanks to Thorsten Kukuk
164 <kukuk@suse.de> for submitting the patch.
165
166 2. Fix for the exit code when using the size estimate option of
167 dump. Thanks to Matti Taskinen <mkt@rni.helsinki.fi> for
168 submitting the patch.
169
170 3. Handle EINTR in atomical reads/writes in dump, which was causing
171 dump to fail on some systems. Thanks to Eric Jergensen
172 <eric@dvns.com> for reporting the bug and submitting the patch.
173
174 4. Handle more than 16 characters for the device names in dumpdates.
175 (up to 255 now). Thanks to Rainer Clasen <bj@ncc.cicely.de> for
176 tracking down the problem and proposing the solution.
177
178 5. Fixed a bug in dump which prevented the creation of the
179 dumpdates file when doing a 0-level dump without already
180 having a dumpdates file. Thanks to Patrik Schindler
181 <poc@pocnet.net> for reporting the bug.
182
183 6. Changed the way dump 'S' flag reports the size estimate
184 from number of blocks into bytes (making it compatible
185 with the Solaris version, and simplifying things for
186 amanda users). Thanks to Jason L Tibbitts III
187 <tibbs@math.uh.edu> for reporting the bug.
188
189 7. Fixed a compatibility problem in linux/alpha dump tape format.
190 Now the linux/alpha dump are (again) compatible with the
191 other dump formats. But this breaks compatibility with
192 older dumps made on alpha. Thanks to Mike Tibor
193 <tibor@lib.uaa.alaska.edu> for helping me in finding this bug.
194
195 Changes between versions 0.4b10 and 0.4b11 (released December 5, 1999)
196 ======================================================================
197
198 1. Added a '--enable-kerberos' to configure.
199
200 2. Added a 'S' option to dump which determines the amount of space
201 that is needed to perform the dump without actually doing it, similar
202 to the Sun's ufsdump 'S' option. Patch contributed by Rob Cermak
203 <cermak@ahab.rutgers.edu>.
204
205 3. Added a 'M' multi-volume option to dump and restore which enables
206 dumping to multiple files (useful when dumping to an ext2
207 partition to make several dump files in order to bypass the 2GB
208 file size limitation). The 'f' argument is treated as a prefix and
209 the output files will be named <prefix>001, <prefix>002 etc. With
210 the 'M' flag, restore automatically selects the right file without
211 asking to enter a new tape each time.
212
213 4. Fixed a memory leak which was causing dump to grow very big
214 (270MB when dumping a 10GB filesystem...). Thanks to Jason
215 Fearon <jasonf@netrider.org.au> for reporting the bug.
216
217 Changes between versions 0.4b9 and 0.4b10 (released November 21, 1999)
218 ======================================================================
219
220 1. Make configure test if the system glob routines support
221 extended syntax (ALTDIRFUNC). If not, use the internal glob
222 routines instead of system ones. Thanks to Bernhard Sadlowski
223 <sadlowsk@Mathematik.Uni-Bielefeld.DE> for reporting the bug
224 and helping me resolve this and other minor libc5 compiling
225 glitches.
226
227 2. Fix a problem when dumping a ext2fs with the 'filetype'
228 feature enabled. Thanks to Patrick J. LoPresti
229 <patl@cag.lcs.mit.edu> for reporting the bug and to
230 Theodore Y. Ts'o <tytso@mit.edu> for providing the patch.
231
232 3. Made the nodump flag work on directories. A directory which
233 has the nodump flag gets never dumped, regardless of its
234 contents.
235
236 4. Integrate a patch from Jeremy Fitzhardinge <jeremy@goop.org>
237 which allows dump on an active ext3 filesystem. However, this
238 is a "quick and dirty" patch which enables backup of an ext3
239 filesystem through the ext2 compatibility (by ignoring the
240 NEEDS_RECOVERY bit). The journal file is not recognized and
241 it is dumped (it should not).
242
243 5. Test the superblock compatibility flags when dumping, in order
244 to be sure that we know how to deal with specific features.
245
246 Changes between versions 0.4b8 and 0.4b9 (released November 5, 1999)
247 ====================================================================
248
249 1. Use lchown instead of chown, fixing a possible security problem
250 when restoring symlinks (a malicious user could use this
251 to deliberately corrupt the ownership of important system files).
252 Thanks to Chris Siebenmann <cks@utcc.utoronto.ca> for detecting
253 this and providing the patch.
254
255 Changes between versions 0.4b7 and 0.4b8 (released November 3, 1999)
256 ====================================================================
257
258 1. Put dump sources under CVS, added Id tags in all files so
259 one can use 'ident' on binary files.
260
261 2. Added the dump/restore version in the usage text so one can
262 easily verify the version he is using.
263
264 3. Small patch from Nuno Oliveira <nuno@eq.uc.pt> which fixes
265 a va_start/va_end problem on linux-ppc (always call va_start
266 va_end in pairs each time we use a vararg function).
267
268 4. Added again the DT_* constants because old libc does not
269 contain them :(. Thanks to Eric Maisonobe <virnet@nat.fr>
270 for submitting the bug report.
271
272 5. Use ext2fs_llseek instead of llseek. With recent e2fsprogs
273 this should enable dumping big (huge) filesystems.
274
275 6. Added the RSH environment variable in order to be able to
276 use a rsh replacement like ssh when doing remote backups (and
277 bypass the security limitations of rcmd). Now you can do remote
278 backups without being root (or making dump setuid root).
279
280 7. Modified again the way dumpdates works. For incremental dumps,
281 we need to read dumpdates even if we are not using 'u' option.
282 Thanks to Bdale Garbee <bdale@gag.com> for his ideas on how
283 this should work.
284
285 Changes between versions 0.4b6 and 0.4b7 (released October 8, 1999)
286 ===================================================================
287
288 1. Removed the 'k' flag from the restore 'about' text if kerberos
289 was not compiled in.
290
291 2. Prototyped (f)setflags from e2fsprogs and corrected the calls
292 to them (fsetflags takes a char*, setflags an open fd!).
293
294 3. (f)setflags is called only if the flags aren't empty. If the
295 file is a special file, a warning is printed, because changing
296 flags implies opening the device. Normally, a special file
297 should not have any flag... (Debian bug #29775, patch provided
298 by Abhijit Dasgupta <abhijit@ans.net>).
299
300 4. Made possible to dump a file system not mentioned in /etc/fstab.
301 (Debian bug #11904, patch provided by Eirik Fuller <eirik@netcom.com>).
302
303 5. Changed the default behaviour to not create dumpdates
304 unless 'u' option is specified. Removed the old "debian-patch"
305 which provided the same thing. (Debian bug #38136, #33818).
306
307 6. Removed all those dump*announce, since they were getting old...
308
309 7. Added warning messages if dumpdates does not exist and
310 when an inferior level dump does not exist (except for a level 0
311 dump).
312
313 8. Debugged the glob calls in interactive mode: restore used a
314 dirent struct which was different from the /usr/include/dirent.h
315 one (this used to work, is it a glibc2 change?), so none of the
316 compat glob (which used /usr/include/dirent.h) or the system glob
317 worked. Restore use now the system dirent (and the system
318 DT_* constants), which are compatible with BSD ones.
319
320 9. Added a configure flag (--with-dumpdatespath) to specify
321 the location of dumpdates file. By default, it is
322 /etc/dumpdates.
323
324 10. Added the "AUTHOR" and "AVAILABILITY" sections and
325 included the current date/version in man pages.
326
327 11. Corrected the estimation of remaining time when
328 the operator doesn't change the tapes quickly enough. This
329 was an old bug, I thought I corrected it, and discovered
330 that in fact it was corrected in two different places, so
331 the results canceled each other...
332
333 Changes between versions 0.4b5 and 0.4b6 (released October 1, 1999)
334 ===================================================================
335
336 1. Integrated multiple patches from RedHat, Debian and SuSE:
337
338 - tweak dump/itime.c to not try to read dumpdates if the 'u' option
339 isn't specified.
340 - several fixes in the man pages.
341 - update the default tape device to /dev/st0.
342 - many updates for Linux Alpha (byte ordering, size_t etc).
343 - buffer overruns.
344 - use environment variable for TMPDIR (instead of /tmp).
345 - use sigjmp_buf instead of jmp_buf (RedHat bug #3260).
346 - workaround egcs bug (RedHat bugs #4281 and #2989).
347 - wire $(OPT) throughout Makefile's.
348
349 2. Upgrade the dump revision to 1, making possible to dump filesystems
350 made with e2fsprogs-1.15 or newer. Nothing seems to break...
351
352 3. Fix some compile warnings, prototype all functions.
353
354 4. Use glibc err/glob instead of internal compatibility
355 routines (only if available).
356
357 5. Fix a compile error on Linux 2.2.7 / libc5 (5.4.44) (patch provided
358 by Bernhard Sadlowski <sadlowsk@mathematik.uni-bielefeld.de>).
359
360 Changes between versions 0.4b4 and 0.4b5 (released September 22, 1999)
361 ======================================================================
362
363 1. Integrated the changes from FreeBSD-3.1-RELEASE
364 (mostly bug fixes, buffer overruns, dump has now an "automatic
365 tape length calculation" flag, dump/restore can use kerberos now
366 (this is NOT tested), use environment variables for TAPE and
367 RMT etc.).
368
369 2. Integrated three RedHat patches ("glibc", "kernel" and "bread" patches)
370
371 3. Corrected a bug in restore when using 'C' option with multi-volumes
372 tapes (files splited accros two tapes give "size changed" errors
373 when comparing).
374
375 4. Corrected the long standing bug when dumping multiple tapes.
376 This works for me, needs further testing.
377
378 Changes between versions 0.4b3 and 0.4b4 (released January 17, 1997)
379 ====================================================================
380
381 1. Dump now runs correctly on kernels 2.1.x
382 Fix made by Gerald Peters <gapeters@worldnet.att.net>
383
384 Changes between versions 0.4b2 and 0.4b3
385 ========================================
386
387 1. Use realpath() if available
388
389 2. Report statistics
390
391 Changes between versions 0.4b1 and 0.4b2
392 ========================================
393
394 1. Fixed the bug fix from Greg Lutz (I had made a mistake when integrating
395 the patch)
396
397 2. Fixed restore to make it able to read FreeBSD 2.x dumps again
398
399 3. Fixed configure.in to correctly handle --enable-rmt
400
401 Changes between versions 0.3 and 0.4b1
402 ======================================
403
404 1. Integrated the changes from 4.4BSD-Lite2
405
406 2. Integrated the patches from Debian and Red Hat
407
408 3. Portability changes: use the __u32, __u16, __s32, and __s16 types
409
410 4. Changed dump to use the Ext2fs library to get block addresses. This
411 should solve the endianness problem on SparcLinux.
412
413 5. Created a configure.in file (shamelessly stolen from the e2fsprogs
414 distribution's one) to use autoconf
415
416 6. Fixed a few minor bugs
417
418 Changes between versions 0.2e and 0.2f
419 ======================================
420
421 1. Added the creation of named pipes (fifos) in restore.
422
423 2. Added the -N flag in the restore manual page.
424
425 3. Added the file kernel.patch which contains the llseek() optimization
426 patch for 1.2.x kernels.
427
428 4. Fixed a bug in the restoration of symbolic links: owner and group were
429 not restored.
430
431 5. Integrated some changes from FreeBSD 2.2.
432
433 6. Added a call to ftruncate() after restoring each file to restore
434 correctly files ending by a hole.
435
436 Changes between versions 0.2d and 0.2e
437 ======================================
438
439 1. Fixed a bug in the "set owner/mode" process. Every file was restored
440 with owner = root (0) and group = root/wheel/whatever (0).
441
442 Changes between versions 0.2c and 0.2d
443 ======================================
444
445 1. Dump is now able to backup 2GB+ filesystems.
446
447 2. Dump and restore can now be linked as static binaries.
448
449 Changes between versions 0.2b and 0.2c
450 ======================================
451
452 1. Fixed a bug when dumping ``slow'' (i.e. normal) symbolic links.
453
454 Changes between versions 0.2a and 0.2b
455 ======================================
456
457 1. Really fixed the bug that I should have corrected in 0.2a.
458
459 2. Enabled optimization again.
460
461 Changes between versions 0.2 and 0.2a
462 =====================================
463
464 1. Disabled the optimization during compilation.
465
466 Changes between versions 0.1 and 0.2
467 ====================================
468
469 1. Fixed a bug in fstab.c which caused a null pointer to be stored in
470 the fs_type field (actually, I modified the file fstab.c to make it
471 use the mntent functions).
472
473 2. Dump and restore now use a 4.3 BSD compatible dump format. Backups
474 made by dump should be readable by the BSD restore and backups made
475 by the BSD dump should be readable by restore. Unfortunately, this
476 means that the dump format has changed between version 0.1 and version
477 0.2 :-(
478
479 3. Dump is now able to backup a subtree, it is no longer limited to whole
480 filesystems like the BSD version.
481
482 4. Dump now uses ext2_llseek() so it is able to backup filesystems bigger
483 than 2 GB.
484
485 Changes between versions 0.0 and 0.1
486 ====================================
487
488 1. Now create links rdump and rrestore during the `make install' step.
489
490 2. Linux port specific bugs added to the manual pages
491
492 3. Incorrect estimation of the number of tapes blocks fixed when doing
493 incremental backups.
494
495 4. Better ls-like format in restore in interactive mode.