blob: 3f49f2d4ec1646b90c94fc45748d9a29176e2b39 [file] [log] [blame]
Theodore Ts'obc210071997-06-17 05:41:36 +00001E2fsprosg 1.11 (June 17, 1997)
2==============================
3
4Fixed e2fsck to detect (previously ignored) conflicts between the
5superblock or block group descriptors and block bitmaps, inode
6bitmaps, and inode tables.
7
8Fixed bug in e2fsck so that when the message printed out when a block
9or inode bitmap conflicts with other data, it has the correct group
10number.
11
12Fixed bug in e2fsck and mke2fs where the blocksize wasn't being passed
13to badblocks. This meant that not all of the filesystem was being
14tested for bad blocks!
15
16Fixed an array boundary overrun case which cropped up in
17ext2fs_badblocks_list_test when a user tried running "mke2fs -c
18-b 4096".
19
20Adjusted the number of columns printed by mke2fs when displaying the
21superblock backups to avoid running over 80 columns when making a
22really filesystem.
23
24Fixed up the man pages for e2fsck, debugfs, badblocks, chattr,
25dumpe2fs, fsck, mke2fs, and tune2fs (typos and other minor grammar
26fixes), thanks to some suggestions from Bill Hawes (whawes@star.net).
27
28Programmer's notes:
29-------------------
30
31Fixed install rule in lib/ss so that ss_err.h is actually getting
32installed.
33
34Fixed bug in ext2fs_copy_bitmap; the destination bitmap wasn't getting
35bassed back to the caller.
36
37Fixed bug in ext2fs_inode_scan_goto_blockgroup; it had not been
38setting the current inode number (which meant this function wasn't
39working at all).
40
41Fixed bug in ext2fs_resize_generic_bitmap; it had not be zeroing all
42blocks in the bitmap when increasing the size of the bitmap.
43
44Changed the initial number of blocks allocated by ext2fs_init_dblist()
45to be more realistic.
46
47Added a new function ext2fs_allocate_group_table, which sets up the
48group descriptor information (and allocates inode and block bitmaps,
49and inode tables for a particular group). The function was created by
50factoring out code form ext2fs_allocate_tables().
51
52Added a new function ext2fs_move_blocks which takes a bitmap of the
53blocks to be moved, and moves them to another location on the
54boardboard.
55
56Make the unix_io channel's io_channel_flush implementation calls sync()
57to to flush the kernel buffers to disk.
58
59Added a new function ext2fs_dblist_count returns the number of
60directory blocks in dblist.
61
62
Theodore Ts'oae851481997-04-29 18:13:24 +000063E2fsprogs 1.10 (April 24, 1997)
64===============================
65
66Mke2fs once again defaults to creating revision #0 filesystems, since
67people were complaining about breaking compatibility with 1.2 kernels.
68Warning messages were added to the mke2fs and tune2fs man pages that
69the sparse superblock option isn't supported by most kernels yet (1.2
70and 2.0 both don't support parse superblocks.)
71
72Added new flag to mke2fs, -R <raid options>, which allows the user to
73tell mke2fs about the RAID configuration of the filesystem. Currently
74the only supported raid option is "stride" which specifies the width
75of the RAID stripe.
76
77Fixed bug in e2fsck where pass1b would bomb out if there were any
78blocks marked bad in the inode table.
79
80Fixed rare bug in mke2fs where if the user had a very unlucky number
81of blocks in a filesystem (probability less than .002) the resulting
82filesystem would be corrupt in the last block group.
83
84Fixed bug where if e2fsck tried to allocate a block to fix a
85filesystem corruption problem and the filesystem had no free blocks,
86ext2fs_new_block() would loop forever.
87
88The configure script now checks explicitly to see if "-static" works,
89since that can't be assumed to be true --- RedHat doesn't install
90libc-static by default.
91
92Fixed bug in libext2's block iterator functions where under some
93cirmcustances, file with holes would cause the bcount parameter to the
94callback function to be incorrect. This bug didn't affect any of
95e2fsprogs programs, but it was discovered by Paul Mackerras, the
96author of the PPC boot loader.
97
98Removed use of static variables to store the inode cache in libext2fs.
99This caused problems if more than one filesystem was accessed via
100libext2fs (static variables in libraries are generally a bad idea).
101Again, this didn't affect e2fsprogs programs, but it was discovered by
102Paul Mackerras.
103
104Fixed minor bugs and version code drift to assure that e2fsprogs 1.10
105will compile cleanly with 1.2.13 kernels (even with a.out shared
106libraries!)
107
108Programmer's notes:
109-------------------
110
111Added new functions to duplicate an ext2 filesystem handle, and its
112associated substructure. New functions: ext2fs_dup_handle(),
113ext2fs_copy_dblist(), ext2fs_badblocks_copy(), ext2fs_copy_bitmap().
114Other structures, such as the io_channel and the inode_cache, now have
115a ref count so that they only get freed when they are no longer used
116by any filesystem handle. (These functions were added as part of the
117development effort for an ext2 resizer).
118
Theodore Ts'o2ecc6fe1997-04-29 17:57:00 +0000119E2fsprogs 1.09 (April 14, 1997)
120===============================
121
122Fixed bug in mke2fs (really in lib/ext2fs/initialize.c) which was
123accidentally introduced in the 1.08 release. The overhead calculation
124was accidentally removed, which caused ext2fs_initialize() to not
125notice when the filesystem size needed to be adjusted down because
126there wasn't enough space in the last block group.
127
128Fixed bug in version parsing library routine; it was always parsing
129the library version string, instead of using the passed-in string.
130
131Clarified chattr man page.
132
Theodore Ts'o24757fa1997-04-29 17:39:27 +0000133E2fsprogs 1.08 (April 10, 1997)
134===============================
135
136E2fsck 1.07 was very slow when checking very large filesystems with a
137lot of files that had hard links (i.e., news spools). This was fixed
138by seriously revamping the icount abstraction. Added a formal test
139suite for the icount abstraction.
140
141Debugfs now has a "-l" option to the "ls" command, which lists the
142inode number, permissions, owner, group, size, and name of the files
143in the directory.
144
145Fix a bug in e2fsck where when a directory had its blocks moved to
146another location during the pass 1b processing, the directory block
147list wasn't updated, so pass 2 wouldn't check (and correct) the
148correct directory block.
149
150E2fsck will now treat inodes which contain blocks which are claimed by
151the filesystem metadata by treating them as multiply claimed blocks.
152This way, the data in those blocks can be copied to a new block during
153the pass 1b--1d processing.
154
155E2fsck will attempt to determine the correct superblock number and
156display it in the diagnostic and warning messages if possible.
157
158Add support for a new (incompatible) feature, "sparse_super". This
159feature reduces the number of blocks which contain copies of backup
160superblocks and block group descriptors. (It is only an incompatible
161feature because of a bug in ext2_free_blocks.) mke2fs and tune2fs now
162support a new -s option; e2fsck will recognize filesystems built with
163this feature turned on.
164
165E2fsck now checks the library to make sure is the correct version,
166using new library functions. (This helps to diagnose incorrectly
167installed e2fsprogs distributions.)
168
169Dumpe2fs now prints more information; its now prints the the
170filesystem revision number, the filesystem sparse_super feature (if
171present), the block ranges for each block group, and the offset from
172the beginning of the block group.
173
174Mke2fs now distributes the inode and block bitmap blok so that the
175won't be concentrated in one or two disks in RAID/striping setups.
176Also, if the user chooses a 2k or 4k block group, mke2fs will try to
177choose the largest blocks per group that be chosen. (For 2k blocks,
178you can have up to 16384 blocks/group; for 4k blocks, you can have up
179to 32768 blocks/group.) Previously mke2fs would not allow
180specification of more than 8192 blocks per group, even if you were
181using a 2k or 4k block group.
182
183Programmer's notes:
184-------------------
185
186Added a new function ext2fs_create_icount2() which takes a "hint"
187argument. This hint argument presets the icount array with the list
188of inodes which actually need to be in the icount array. This really
189helps to speed up e2fsck.
190
191Added a new function ext2fs_icount_validate() which checks the rep
192invariant for the icount structure. This is used mostly for testing.
193
194The error mesasage given when a bad inode number is passed to
195test_generic_bitmap to reflect EXT2FS_TEST_ERROR (instead of
196EXT2FS_UNMARK_ERROR).
197
198Added a new function ext2fs_set_dir_block which sets the block of a
199dblist entry, given the directory inode and blockcnt.
200
201Added a new function ext2fs_get_library_version() which returns the
202current library version, and ext2fs_parse_version_string() which
203returns a version number based on a e2fsprogs version string.
204
205The icount functions will return EINVAL if the passed in inode number
206is out of bounds.
207
Theodore Ts'o21c84b71997-04-29 16:15:03 +0000208E2fsprogs 1.07 (March 9, 1997)
209==============================
210
211E2fsck is now uses much less memory when checking really large
212filesystems (or rather, filesystems with a large number of inodes).
213Previously a filesystem with 1 million inodes required 4 megabytes of
214memory to store inode count statistics; that storage requirement has
215now been reduced to roughly half a megabyte.
216
217E2fsck can now properly deal with bad blocks appearing inside the
218inode table. Instead of trying to relocate the inode table (which
219often failed because there wasn't enough space), the inodes in the bad
220block are marked as in use.
221
222E2fsck will automatically try to use the backup superblocks if the
223primary superblocks have a bad magic number or have missing meta-data
224blocks (or meta-data blocks which are out of range).
225
226E2fsck's pass 3 has been made more efficient; most noticeable on
227filesystems with a very large number of directories.
228
229Completely revamped e2fsck's system of printing problem reports. It
230is now table driven, to make them more easily customizeable and
231extendable. Error messages which can be printed out during preen mode
232are now one line long.
233
234Fixed e2fsck's filesystem swapping code so that it won't try to swap
235fast symbolic links or deleted files.
236
237Fixed e2fsck core dumping when fixing a filesystem which has no
238directories (not even a root directory).
239
240Added a check to e2fsck to make sure that the length of every
241directory entry is a multiple of 4 (since the kernel complains if it
242isn't).
243
244Added a check to e2fsck to make sure that a directory entry isn't a
245link to the root directory, since that isn't allowed.
246
247Added a check to e2fsk to now make sure the '.' and '..' directory
248entries are null terminated, since the 2.0 kernel requires it.
249
250Added check to write_bitmaps() to make sure the superblock doesn't get
251trashed if the inode or block bitmap is marked as being block zero.
252
253Added checking of the new feature set fields in the superblock, to
254avoid dealing with new filesystem features that this package wasn't
255set up to handle.
256
257Fixed a fencepost error in ext2fs_new_block() which would occasionally
258try to allocate a block beyond the end of a filesystem.
259
260When the UUID library picks a random IEEE 802 address (because it
261can't find one from a network card), it sets the multicast bit, to
262avoid conflicting with a legitimate IEEE 802 address.
263
264Mke2fs now sets the root directory's owner to be the real uid of the
265user running mke2fs. If the real uid is non-zero, it also sets
266the group ownership of the root directory to be the real group-id of
267the user running mke2fs.
268
269Mke2fs now has more intelligent error checking when it is given a
270non-existent device.
271
272When badblocks is given the -vv option, it now updates the block that
273it is currently testing on every block.
274
275Fixed a bug in fsck where it wouldn't modify the PATH envirnoment
276currently correctly if PATH wasn't already set.
277
278Shared libraries now built with dependencies. This allows the shared
279library files to be used with dlopen(); it also makes the transition
280to libc 6 easier, since ld.so can tell which libc a particular shared
281library expects to use.
282
283Programmer's notes:
284-------------------
285
286Added new abstraction (defined in dblist.c) for maintaining a list of
287blocks which belongs to directories. This is used in e2fsck and other
288programs which need to iterate over all directories.
289
290Added new functions which test to see if a contiguous range of blocks
291(or inodes) are available. (ext2fs_*_bitmap_range).
292
293Added new function (ext2_inode_has_valid_blocks) which returns true if
294an inode has valid blocks. (moved from e2fsck code).
295
296Added new function (ext2fs_allocate_tables) which allocates the
297meta-data blocks as part of initializing a filesystem. (moved from
298mke2fs code).
299
300Added a new I/O manager for testing purposes. It will either allow a
301program to intercept I/O requests, or print debugging messages to
302trace the activity of a program using the I/O manager.
303
304The badblocks_list functions now store the bad blocks in a sorted
305order, and use a binary search to speed up badblocks_list_test.
306
307The inode scan function ext2fs_get_next_inode() may now return a soft
308error returns: MISSING_INODE_TABLE and BAD_BLOCK_IN_INODE_TABLE in
309those cases where part of an inode table is missing or there is a bad
310block in the inode table.
311
312Added a new function (ext2fs_block_iterate2) which adds new arguments to
313the callback function to return a pointer (block and offset) to the
314reference of the block.
315
316Added new function (ext2fs_inode_scan_goto_blockgroup) which allows an
317application to jump to a particular block group while doing an inode
318scan.
319
320The badblocks list functions were renamed from badblocks_* to
321ext2fs_badblocks_*. Backwards compatibility functions are available
322for now, but programs should be modified to use the new interface.
323
324Some of the library functions were reorganized into separate files to
325reduce the size of some programs which statically link against the
326ext2 library.
327
328Put in some miscellaneous fixes for the Alpha platform.
329
Theodore Ts'o5c576471997-04-29 15:29:49 +0000330
331E2fsprogs 1.06 (October 7, 1996)
332================================
333
334Fixed serious bug in e2fsck: if the block descriptors are bad, don't
335smash the backup copies in ext2fs_close(). (The problem was that when
336e2fsck -p discovered the problem, while it was closing the filesystem
337and exiting, it was also blowing away the backup superblocks on the
338disk, which was less than friendly.) We now make it the case that we
339only write out the backup superblock and the back block descriptors if
340the filesystem is completely free from problems.
341
342Fixed a bug in block_interate in the lib/ext2fs library which caused
343e2fsck to fail on GNU Hurd-created filesystems.
344
345Add support for Linux/FT's bootloader, which actually uses
346EXT2_BOOT_LOADER, and sets its mode bits which caused e2fsck to want
347to clear the inode.
348
349Add support for the "A" (no atime update) attribute. (Note: this
350attribute is not yet in production kernels.)
351
352The test suite is not automatically run when doing a "make all" from
353the top level directory. Users should manually run "make check" if
354they wish to run the test suite.
355
356Upon a preenhalt(), make the printed message more explicit that
357running e2fsck "MANAULLY" means without the -p or -a options.
358
359In e2fsck, if a disconnected inode is zero-length, offer to clear it
360instead of offering to connect it to lost+found.
361
362In e2fsck, if a filesystem was just unmounted uncleanly, and needs
363e2fsck to be run over it, change e2fsck to explicitly display this
364fact.
365
366For dumpe2fs and e2fsck, cause the -V option to print out which
367version of the ext2fs library is actually getting used. (This will
368help detect mismatches of using a 1.06 utility with a 1.05 library,
369etc.)
370
371Programmers' notes:
372-------------------
373
374EXT2_SWAP_BYTES was changed to EXT2_FLAG_SWAP_BYTES, which better fits
375the naming convention.
376
377In ext2fs_initialize(), make sure the description for the inode bitmap
378is correctly initialize.
379
380Fixed minor type typo in ext2fs_allocate_generic_bitmap();
381
Theodore Ts'ofc6d9d51997-04-29 14:51:31 +0000382E2fsprogs 1.05 (September 7, 1996)
383==================================
384
385Add support for new fields in the ext2 superblock --- volume name,
386volume UUID, and last mounted field. Dumpe2fs displays these fields,
387tune2fs and mke2fs allows you to set them. E2fsck will automatically
388generate a UUID for those volumes that don't have them.
389
390Put in support for e2fsck to recognize HURD specific ext2 features ---
391most notably, the translator block. The e2fsprogs tools will now use
392the creator_os field in the superblock to correctly handle different
393OS-specific variants of the ext2 filesystem.
394
395E2fsck now fixes inodes which have a the deletion time set, but which
396have a non-zero i_link_count field by offering to clear the deletion
397time. Previously e2fsck assumed that the inode was deleted (per 0.3c
398ext2 kernel behavior) and offered to unlink the file.
399
400If e2fsck sets the clean bit, but nothing else, set the exit code
401FSCK_NONDESTRUCT. After all, e2fsck did fix a filesystem error --- it
402set the filesystem valid bit when it was previously cleared. :-) This
403was needed to make the HURD fsck driver happy.
404
405If the user refuses to attach an unattached inode, e2fsck will no
406longer set the inode's link count. Otherwise, the inode would end up
407getting marked as unused, which might cause loss of data later.
408
409Make the message issued by e2fsck when the superblock is corrupt less
410confusing for users. It now mentions that another reason for the
411"corrupt superblock" message might be that the partition might not be
412an ext2 filesystem at all (it might swap, msdos filesystem, ufs, etc.)
413
414Make the libext2 library more robuest so that e2fsck won't coredump on
415an illegal superblock where the blocksize is zero. (f_crashdisk is
416the test case).
417
418By default, create filesystems where the default checkinterval is 6
419months (180 days). Linux servers can be robust enough that 20 reboots
420can be a long, long time.
421
422Added configure flag --enable-old-bitops, which forces the bitops to
423use the old (native) bitmask operations. By default on the sparc
424platform, the standard ext2 bit ordering is now used.
425
426Added a new feature to e2fsck to byte-swap filesystems; this can be
427used to convert old m68k filesystems to use the standard byte-order
428storage for the superblock, inodes, and directory blocks. This
429function is invoked by using the '-s' option to e2fsck.
430
431Debugfs's "dump" command has been enhanced so that it writes out the
432exact size of the file so that the nulls at the end of the file are
433eliminated. The command also accept a new "-p" option which will
434attempt preserve to preserve the ownernship, permissions, and
435file modification/access times.
436
437Debugfs has two new options, -f and -R. The -R option allows the user
438to execute a single debugfs command from the command line. The -f
439option allows the user to specify a "command file" containing debugfs
440commands which will get executed.
441
442Dumpe2fs now pretty prints the check interval, instead of just
443printing the check interval as a number of seconds.
444
445Fix bugs in debugfs: the params command when no filesystem is opened
446no longer causes a core dump. It is now possible to unlink a file
447when a pathame containing a '/' is specified.
448
449Tune2fs has a new -C option which sets the number of times the
450filesystem has been mounted.
451
452Fix the chattr '-v' option so that it actually works. Chattr was
453being buggy about the -v option parsing.
454
455Programmers' notes:
456-------------------
457
458The directory lib/uuid contains a set of library routines to generate
459DCE compatible UUIDs.
460
461Extended ext2fs_namei() to handle symbolic links. Added new function
462ext2fs_nami_follow() which will follow last symbolic link in the case
463where the pathname points to a sym link.
464
465The ext2fs_block_iterate function will now return the HURD translator
466block, if present. The new flag BLOCK_FLAG_DATA_ONLY will cause the
467iterator to return data blocks only. The ext2fs.h file now defines
468constants BLOCK_COUNT_IND, BLOCK_COUNT_DIND, BLOCK_COUNT_TIND, and
469BLOCK_COUNT_TRANSLATOR, which are the magic values passed in the block
470count field of the iterator callback function.
471
472The test script driver now takes an optional second argument, which is
473the test case to be run. This allows you to run a test case without
474needing to run the entire test suite.
475
476On Linux ELF systems, install the .so files in the correct places
477(/usr/lib). The .so files must be stored in the same directory as the
478.a files.
479
480Fixed miscellaneous HURD compilation issues with header file being
481included in the right order.
482
483Fixed debugfs so that it resets optind to zero, not one, since setting
484optind to zero is more correct.
485
486
Theodore Ts'oa4d09611997-04-29 14:28:00 +0000487E2fsprogs 1.04 (May 16, 1996)
488=============================
489
490First "official" (1.03 was a limited release only) to support building
491e2fsprogs under Linux 2.0 kernels (as well as late model 1.3 and 1.99
492kernels).
493
494This package includes a RPM specs file, that it can be built using the
495RedHat Package Manager.
496
497E2fsck now prints a hint that if there are lots of incorrectly located
498inode bitmaps, block bitmaps, and inode table blocks, the user might
499want to try using e2fsck -b 8193 first, to see if that fares any
500better.
501
502For ext2 filesystem written with the hurd, debugfs will now print out
503the translator field when printing an inode structure.
504
505Lots of miscellaneous linking/installation cleanups:
506
507 Libraries are now linked using a relative pathname, instead of
508 relying on -L working correct. It doesn't, in many cases, including
509 current versions of GNU ld. This guarantees that the build tree is
510 linking with the right libraries, instead of the ones installed in
511 /usr/lib.
512
513 Header files, man pages, and the et/ss shell scripts are now
514 generated using a custom substitution script, instead of relying on
515 the configure script. This prevents needless recompilation of
516 files; in addition, the custom substitution script is much faster.
517
518 e2fsck may now be linked dynamically, by using the
519 --enable-dynamic-e2fsck flag to configure. This is not recommended,
520 since it increases e2fsck's dependence on other files, but some
521 people need to save disk space, and other critical programs on their
522 systems were being linked dynamically anyway.
523
524 Programs such as fsck which didn't need to be linked against
525 libext2fs (or mke2fs which didn't need to be linked against libe2p)
526 only link against libraries they actually need. Otherwise, those
527 programs would require the presense of libraries that otherwise
528 could be removed from a rescuse diskette.
529
530 The ss include files are now installed correctly so they can
531 actually be used by another package.
532
533 If the profiling libraries are built, they are now installed on a
534 "make install-libs".
535
536
Theodore Ts'o7f88b041997-04-26 14:48:50 +0000537E2fsprogs 1.03 (March 27, 1996)
538===============================
539
540Change the m68k bit numbering for bitmasks to match the bit numbering
541used by all other ext2 implementations. (This change was requested by
542the m68k kernel development team.)
543
544Support (in-development) filesystem format revision which supports
545(among other things) dynamically sized inodes.
546
547Fixed a bug in the ext2 library so that an intelligent error is
548returned if mke2fs is run with a ridiculously small number of blocks
549for a partition.
550
551Fixed a bug in the ext2 library which required that the device be
552openable in read/write mode in order to determine its size. This
553caused e2fsck -n to require read/write access when it was not
554previously necessary.
555
556Fixed a bug in e2fsck which casued it to occasionally fail the test
557suite depending on which version of the floating point library it was
558using.
559
560Fixed a bug in e2fsck so that it now halts with a fatal error when
561certain superblock consistency checks fail. Previously it continued
562running e2fsck, with some potential confusing/damaging consequences.
563
564Added new flag to fsck which allows the root to be checked in parallel
565with other filesytems. This is not the safest thing in the world to
566do, but some system administrators really wanted it.
567
Theodore Ts'ofc6d9d51997-04-29 14:51:31 +0000568Fixed -Wall flames in lib/ss.
569
Theodore Ts'o74becf31997-04-26 14:37:06 +0000570
571E2fsprogs 1.02 (January 16, 1996)
572=================================
573
574Fix to allow e2fsprogs to be compiled on recent 1.3 (pl45+) kernels.
575
576Change e2fsck to print statistics of how many non-contiguous files are
577on the system. Note that a file which is larger than 8k blocks, it is
578guaranteed to be non-contiguous.
579
580In mke2fs, print a warning message if a user tries to format a whole
581disk (/dev/hda versus /dev/hda1). If a user really wants to format a
582whole disk, the -F (force) option forces mke2fs to format a whole disk
583as a filesytem.
584
585Fix a bug in fsck where in some cases it might start checking
586partitions in the next pass before it finishes checking partitions in
587the current pass. This still won't cause two partitions on the same
588disk will be checked, so it's rarely a problem in real life.
589
590Patch lsattr so that it won't hang when checking a named pipe.
591
592Minor compilation fixes:
593 * Fix the order of libraries that were linked in debugfs.
594 * Allow the sources to be compiled with -ansi turned on.