RELEASE-NOTES revision 5a9c036bf3d48ca622b08286b5e82bec2cbb6b7a
1E2fsprogs 1.21-WIP (June 1, 2001) 2================================= 3 4Fixed two bugs in e2fsck's handling of dup block handling, dealing 5with relatively uncommon edge cases: a directory with an indirect 6block which is claimed by another file, and when the last inode in the 7filesystem has blocks claimed by another file. 8 9E2fsck now checks to see if the i_size field of a fast symlink is too 10big, and offers to clear the symlink if so. 11 12E2fsck now checks to see if i_size_high of special files is non-zero, 13and offers to clear i_size_high. 14 15Fixed byte-ordering conversion bug which caused e2fsck's revoke 16handling to essentially not work. 17 18Fix e2fsck's handling of incompatible journal flags so that the user 19has chance to abort, and then has the option to clear out the journal 20entirely. 21 22Programmer's notes: 23------------------- 24 25Fix general gcc -Wall complaints. 26 27The types needed by the ext2 header files are now provided by 28lib/ext2fs/ext2_types.h, instead of include/asm/types.h. 29 30Fixed make depend script so that it automatically corrects the 31pathname cleanups performed by make -M, so I don't have to fix them up 32by hand. 33 34Fixed the d_loaddump test case to be more robust, and not depend on 35bash'isms. 36 37Removed debugfs's dependence on pread(), which was accidentally 38intrudced in e2fsprogs 1.20 39 40 41E2fsprogs 1.20 (May 20, 2001) 42============================= 43 44Add support for replaying the ext3 journal purely in user mode, 45including handling the orphaned inode list. Used code contributed by 46Andreas Dilger, with bug fixes and the orphaned inode handling done 47by Theodore Ts'o. 48 49The mke2fs and tune2fs programs can create or modify a filesystem to 50include an ext3 journal. Tune2fs also can be used to remove an ext3 51journal from a filesystem. 52 53E2fsck will now check for the existence of a linked list of orphan 54inodes in the superblock, and clear those inodes before starting the 55rest of the filesystem check (but after the journal playback). 56 57E2fsck now validates the file descriptor passed to the -C option, 58which saves against the completion bar getting written to an 59unexpected location, such as the disk being checked. (Debian 60bug/wishlist #55220) 61 62E2fsck will now bump the filesystem revision number from zero to one 63if any of the compatibility bits are set. 64 65Fixed a bug where a badly corrupted directory on a big endian system 66could cause e2fsck to die with a bus error. The 67ext2fs_read_dir_block() and ext2fs_process_dir_block() functions in 68the ext2 library now does alignment sanity checks on the rec_len field 69of the directory entry before using it. 70 71The ext2 library has been enhanced to make tune2fs safe to run on 72mounted filesystems. (Users could usually get away with using tune2fs 73on mounted filesystems before, but with the advent of ext3 and 74journaling, it became important to make tune2fs was *really* safe for 75use even when the filesystem being modified is mounted.) E2label is 76now implemented by tune2fs using an argv[0] dispatch, so that e2label 77is also now safe for use on mounted filesystems. 78 79Added a new program, e2image, which creates a backup of critical ext2 80filesystem data structures. The generated image file can be examined 81using dumpe2fs and debugfs. In the future, e2fsck will be able to use 82the image file to help recover very badly damaged filesystems. 83 84Fixed a number of LFS bugs in e2fsck; very, very large (> 2**42) files 85no longer cause e2fsck to bomb out. Also treat files > 2GB as being 86large file when deciding whether or not the filesystem has large files. 87 88Fixed lsattr and chattr so that they work correctly on large files. 89(Fixes Debian bug #72690.) 90 91Removed limitation in get_device_size() which imposed a one terrabyte 92filesystem limitation. (Most 2.2 kernels still have a signed int 93problem which cause 1 TB block device limitation. Fortunately, the 94kernel patches to fix this are much easier than fixing the 2TB 95limitation in the kernel. :-) 96 97A max_mount_count of zero is now treated as if no mount count were 98set. (Previously, no mount count was indicated by using -1, and a 99mount count of zero caused e2fsck to always be run.) 100 101Mke2fs supports two new filesystem types largefile and largefile4. 102 103Mke2fs now adds some randomness in s_max_mount_count so that multiple 104filesystems won't be all checked at the same time under normal 105operations. 106 107Fixed bug in the progress bar printing code which could cause e2fsck 108to core dump on an illegal filesystem. 109 110Fixed bug in fsck which could allow more than one instance of e2fsck 111to be printing a progress bar. (Debian bug #65267) 112 113Fsck using a UUID or a LABEL specifier will work even if devfs is 114compiled into the kernel and not mounted. If the pathnames in 115/proc/partitions are incorrect, fsck will search /dev for the correct 116device (using the new ext2fs_find_block_device library function). 117Fsck now also checks the RAID devices first so that they are properly 118found when they are in use. Support has also been added to support 119additional IDE disks and the DAC 960 device names. (Debian bug #94159) 120 121Fixed a bug in fsck which caused it not deal properly with 16 122byte long filesystem labels. 123 124Fsck's -t option has been made a lot more flexible. The semantics for 125what happens if a comma-separated list to fsck has been regularized, 126and it is now possible to filter what filesystems will get checked 127based what is in the filesystem's fstab entry's option field. (Debian 128bug #89483.) 129 130The dumpe2fs program can now print out the group description 131information in hex, and also prints the location of the superblock and 132block group descriptor for those block groups that have them. 133 134Mke2fs now clears the ext2 superblock before it starts creating a 135filesystem, so that the superblock magic number is only written if the 136filesystem creation process successfully completes. 137 138The debugfs program's stat command now pretty-prints the blocks used 139by an inode so that it's more compact and informative. 140 141The debugfs stats command now uses the same libe2p code (which is used 142by dumpe2fs) to print the superblock header information. This is more 143complete, and it avoids a bit of code duplication. 144 145Added a new debugfs command, set_super_value (ssv) which allows the 146user to set arbitrary superblock fields. 147 148Debugfs was extended to support inode numbers in hex (by prefixing 149them with 0x), and so that modify_inode can set the inode generation 150number. Also, there is now a new function command called logdump 151which will dump an ext3 journal. 152 153Fixed a bug in debugfs so that quitting out of the pager doesn't kill 154debugfs. 155 156Debugfs's dump command now stops immediately upon reporting a disk 157read error. (Fixed a bug in ext2fs_file_read library routine which 158caused debugfs not to stop.) (Debian bug #79163) 159 160On systems with /proc/mounts (mainly Linux systems), /proc/mounts is 161checked before /etc/mtab is used, since /proc/mounts is more likely to 162be accurate. 163 164Added portability fixes for Solaris and Linux/ia64. 165 166Various manual pages were clarified and cleaned up. (Fixed debian 167bugs #63442, #67446, and #87216) 168 169 170Programmer's notes: 171------------------- 172 173The e2fsck message printer now supports %Iu and %Ig, which will print 174out the inode's user and group owners, respectively. 175 176E2fsprogs now includes its own version of include/linux/ext2_fs.h, so 177that no longer dependent on the system having the correct version of 178the kernel header files. 179 180Added a new function to libext2, ext2fs_find_block_device(), which 181searches the system (i.e., /dev, /devfs, /devices) for a pathname to a 182device given its device number. 183 184Added a new function to libext2, ext2fs_sync_device, which centralizes 185all of the places which might try to use the BLKFLSBUF or FDFLUSH 186ioctls (and usually failing to define them since the system header 187files don't usually do this for us, and we're trying to avoid usage of 188kernel include files now). 189 190Added new utility programs in tests/progs: random_exercise and 191hold_inode. They aren't built by default; they're useful for 192exercising ext3 filesystem code. 193 194Added a new ext2 filesystem flag, EXT2_FLAG_SUPER_ONLY, which causes 195the filesystem close functions to only update the superblock, and to 196not touch the block group descriptors. Needed by tune2fs when 197modifying a mounted filesystem. 198 199Got rid of struct ext2fs_sb and replaced it with the standard struct 200ext2_super_block from include/linux/ext2_fs.h. Note: this may break 201source (but not binary) compatibility of some users of the ext2 202library. Those applications should just simply do a global search and 203replace of struct ext2fs_sb with struct ext2_super_block, and use 204their own private copy of ext2_fs.h if they aren't already. 205 206The ino_t type has been renamed ext2_ino_t to protect applications 207that attempt to compile -D_FILE_OFFSET_BITS=64, since this 208inexplicably changes ino_t(!?). So we use ext2_ino_t to avoid an 209unexpected ABI change. 210 211The Makefiles have been reworked so that "make check" can be run from 212the top-level Makefile. 213 214Fix general gcc -Wall complaints and removed dead code. 215 216Remove use of NOARGS, because we assume everyone does ANSI C these 217days. 218 219Added build-rpm script from sct. 220 221New functions ext2fs_image_{inode,super,bitmap}_{read,write} added 222to support e2image. 223 224New function ext2fs_flush_icache which must be called if the 225application program modifies the inode table blocks without going 226through ext2fs_write_inode() interface. 227 228New ext2fs_check_mount_point() function, which will return the mount 229point of a device if mounted. 230 231The io_channel abstraction now has an optional interface, 232io_channel_write_range, which allows specific byte ranges to be 233written. 234 235The unix_io IO channel now supports write-through caching, so that 236journal creation is more efficient. 237 238Added x86 assembly language routines to support byte swapping, to 239reduce executable size. 240 241Fixed bug in the utility program subst so that it's possible to 242replace a substitution variable with a zero-length string. 243 244Fixed numbering e2fsck pass1 problem numbers; an extra zero had 245slipped into some of the problem number. 246 247E2fsprogs 1.19 (July 13, 2000) 248============================== 249 250Release the resize2fs program since the timeout before it could 251be released under the GPL has finally expired. 252 253Add experimental support needed for the ext2 compression patches. 254This requires compiling e2fsprogs with the --enable-compression flag 255to the configure script. 256 257Added ext3 journalling support. E2fsck will run the journal (if 258necessary) by temporarily mounting the filesystem. /sbin/fsck.ext3 is 259installed as a symlink to e2fsck. Fsck has been taught about ext3, 260and treats it the same as ext2 in terms of the progress bar logic. 261Dumpe2fs will display the superblock journaling information if the 262filesystem has a journal. The ext2 library will now permit opening an 263ext3 filesystem with the recovery flag set. This is necessary for 264on-line dump's to work correctly, but there may be issues with this 265working well since ext3 is much less agressive about syncing blocks to 266the filesystem, since they're safe on the journal. 267 268Tune2fs and e2fsck have been changed to allow the mount_count check to 269be disabled by setting max_mount_count to -1. (This was already 270supported by the kernel.) 271 272Create a symbolic link for fsck.ext3, since the e2fsprogs utilities 273are used for ext3 as well. 274 275Added internationalization support for e2fsprogs; must be enabled 276by passing --enable-nls to configure. 277 278Always use the provided ext2fs header files to insulate ourselves from 279kernel version changes. Which include files are used by e2fsprogs 280have also been cleaned up to improve portability. 281 282Limit the number of times that e2fsck updates the progress bar so that 283people who are booting using a 9600 baud console don't get swampped by 284too many updates. 285 286Improved the loop detection algorithm in e2sck's pass #3 so that it is 287much, much faster for large filesystems with a large number of 288directories. 289 290The memory footprint for e2fsck is now slightly smaller than before. 291 292E2fsck now checks if special devices have a non-zero size, and offers 293to clear the size field if it finds such an inode. 294 295E2fsck now checks if special devices have the append-only flag set, 296and offers to clear the inode. 297 298E2fsck now properly handles some "should never fail" cases during a 299bitmap copy in pass5. 300 301E2fsck now properly prints control characters in filenames as ^A .. ^Z. 302 303E2fsck now calculates the correct location of the backup superblock in 304the case of filesystem blocksizes > 1k. 305 306Fixed a bug in e2fsck's calculation of the number of inodes_per_block 307which normally didn't cause problems under most filesystem parameters, 308but could cause a valid superblock to be rejected in extreme cases. 309Other checks for validating superblock values were made more 310stringent. 311 312Added non-destructive write testing to the badblocks program, courtesy 313of David Beattie. The badblocks also now has an option to input the 314current set of bad blocks, so that known bad blocks are skipped to 315speed up the badblocks test. There is also a persistent rescan 316feature which causes badblocks to run until it has completed some 317number of passes without discovering any new bad blocks. 318 319Badblocks now checks to see if the device is mounted and refuses to do 320the tests involving writing to the device if it is mounted. Also, 321badblocks now allows the number of blocks to be checked to be 322defaulted to the size of the partition. 323 324Fixed a bug in fsck which didn't allow non-root users to be able to 325check filesystems if there were any LABEL= or UUID= entries in 326/etc/fstab. 327 328The Hurd doesn't support the filetype filesystem feature. The mke2fs 329program now makes sure that for the Hurd, the filestype feature is 330turned off. E2fsck will check to see if the filetype feature is 331turned on for Hurd filesystems, and offer to turn off the feature. 332 333Mke2fs now has a safety check to make sure the number of blocks do not 334exceed 32 bits even on a 64 bit platform. 335 336Really fixed a bug in fsck to allow "fsck -As" to run interactive 337fsck's. (For those people who like to do interactive fsck's in the 338/etc/rc scripts!?!) 339 340Debugfs has a few new features: the rdump command, which will do a 341recursive dump of a directory and all of its contents, and the lcd 342command which does a local chdir (much like the ftp command of the 343same name). In addition, the debugfs program and the open_filesystem 344command now takes three new options: -b and -s, which allows the 345blocksize and superblock location to be specified, and the -c option 346which is used in catastrophic situations where the block group 347descriptors are corrupt. If the -c option is specified, debugfs will 348skip trying to read in the block and inode bitmaps. 349 350Debufs's lsdel command was fixed to handle bad blocks in the inode 351table. 352 353A Y2K bug in debugfs's "ls -l" handling was fixed by switching to use 3544 digit years. 355 356General improvements in error messages 357 358 - Mke2fs prints a sane error message if the partition size is zero 359 (usually because the partition table wasn't reread by the 360 kernel due to the partition being busy), instead of "invalid 361 argument passed to ext2 library while initializing superblock". 362 363 - Fsck now prints more self-explanatory message if an invalid UUID= 364 or LABEL= specification is passed to it. 365 366UUID library changed to use the LGPL. 367 368Fixed a bug in the UUID library where very rapid calls to the 369time-based UUID generator could cause duplicate UUID's to be returned. 370This was not a problem for e2fsprogs, but it could be a problem for 371other users of the library. 372 373Make the UUID library more robust in the face of missing or an 374improper /dev/urandom or /dev/random files. 375 376Added some random portability fixes for Solaris. 377 378Some minor man page updates. 379 380Fixed a memory leak in the ss library. 381 382 383Programmer's notes: 384------------------- 385 386We now try to use lseek64 and open64 from the LFS if possible. 387 388The 3rd parameter in e2p's print_flags is now a flags word, instead of 389a boolean option. 390 391The mark and unmark bitmap functions now return the previous state of 392the bit that was being changed, which is useful for some speed 393optimizations. 394 395The following functions have been added to enhance the badblocks list 396handling in libext2fs: ext2fs_write_bb_FILE, ext2fs_read_bb_FILE2, and 397ext2fs_badblocks_equal. 398 399The ext2 header files now have the latest journalling fields to the 400superblock. 401 402The ext2fs_mkdir function in libext2fs now properly backs out of error 403conditions robustly. 404 405Cleaned up makefiles: 406 - to cleanly compile with the -j flag. 407 - so distclean removes all generated files. 408 - so in case of an error while installing header files, the make aborts. 409 410Fix test_script so that it works correctly when compiling in the 411source directory. 412 413The random UUID generation routine has been made slightly better in 414the case where /dev/random doesn't exist. (Use of randomly-based UUID 415is still not recommended if /dev/random doesn't exist, however; it's 416better to use the time/ethernet MAC address UUID in this case.) 417 418Clean up the build process so it's more friendly in case of missing 419directories. 420 421The ext2fs header file can now be #include'd into C++ programs. 422 423The e2p.h header file is now installed. 424 425Added workaround to a gawk 3.0.5 bug in lib/ss/mk_cmds. 426 427 428 429E2fsprogs 1.18 (November 10, 1999) 430================================== 431 432Fix a core dumping bug in e2fsck if an imagic inode is present or 433(more rarely) if the filesystem is badly corrupted enough that e2fsck 434has to restart pass 1 processing. E2fsck now closes the filesystem 435before freeing a large number of its data structures, so in the case 436of future memory faults, at least the fixed filesystem will be fully 437written out. 438 439If a filesystem doesn't support imagic inodes, and e2fsck discovers an 440imagic inode, it will offer to clear the imagic flag. 441 442E2fsck will now offer to clear the immutable flag on special files 443(device/socket/fifos) when running it in non-preen mode. 444 445E2fsck will now set the filetype when creating /lost+found, and when 446connected orphaned inodes to /lost+found. 447 448Debugfs's ncheck and icheck commands now handles the case where there 449are bad blocks in the inode table without bombing out. 450 451The badblocks list processing code has been made more efficiently for 452appending a large number of (ordered) badblocks to the badblocks list. 453 454Some minor man page updates. 455 456Fsck now allows interactive e2fsck's when using fsck -As (not a common 457mode, but some people like to do this in boot scripts for silly reasons). 458 459Programmer's notes: 460------------------- 461 462The internal e2fsck problem code for PR_2_SPLIT_DOT was fixed to meet 463with the problem code convention. 464 465The badblocks list regression test program has been updated to work 466with previously made API name changes. 467 468The ext2fs_free() command now uses the new badblocks API to avoid 469using the compatibility layer. 470 471Added new regression test cases; the run_e2fsck test script now 472supports the ability for a test case to run a prepratory command 473before running e2fsck. 474 475E2fsprogs 1.17 (October 26, 1999) 476================================= 477 478Fixed nasty typo in fsck which caused parallelized fsck's to go into an 479infinite loop. 480 481Fixed a bug in fsck where it used strncmp to compare a binary UUID, 482thus potentially causing problems if a binary UUID contained a NULL 483character. 484 485E2fsck now uses stricter checks for directory entries in pass 2: 486zero-length filenames are not allowed; neither are 8 byte long 487directory entries. 488 489The debugfs "dirty" command now clears the filesystem valid bit. 490(Previously this just set the dirty-as-in-needs-writing-out-to-disk 491bit in the in-core superblock image. The new functionality is more 492what the user expects, and is more useful.) 493 494Added a debugging hook to test parallel fsck; if the environment 495variable FSCK_FORCE_ALL_PARALLEL, then filesystems on the same drive 496will be checked in parallel when they normally would not be. 497 498Programmer's notes: 499------------------- 500 501Fixed some #ifdef's for compilation under the Hurd OS. 502 503Fixed minor W2K compatibility problems. 504 505Fixed some miscellaneous GCC warnings. 506 507 508E2fsprogs 1.16 (October 22, 1999) 509================================= 510 511Fixed a race condition bug in fsck; when printing a progress bar, if 512checking multiple filesystems in parallel, it was possible for fsck to 513send e2fsck a SIGUSR1 signal before e2fsck had installed its signal 514handler, which would cause it to terminate with a signal 10. 515 516E2fsck now properly handles filesystems that have the 517INCOMPAT_FILETYPE feature turned on. It can be used to convert a 518filesystem into using or not using FILETYPE feature. 519 520E2fsck now properly handles filesystems that have the IMAGIC feature 521turned on (this is used on Linux AFS servers). 522 523The mke2fs program now creates filesystems that have the filetype and 524sparse_superblock features enabled by default, unless it is run on a 525pre-2.2 kernel. These features are not supported by a pre-2.2 kernel, 526so there is now a new flag -O which allows the user to specify with 527which features she would like to create the filesystem; "mke2fs -O 528none" will create a filesystsem compatible with 2.0 kernels. 529 530The tune2fs program now has a -O option which allows the user to set 531and reset "safe" filesystem features. Currently, the only ones which 532allows to be modified are the filetype and sparse_superblock features. 533Note setting or clearing either feature will require running e2fsck on 534the filesystem afterwards. (n.b. Clearing the sparse_superblock feature 535requires that there is enough free space on the filesystem for the 536extra superblocks which will be created by e2fsck.) 537 538Debugfs can now set and print filesystem features in the superblock 539using the "features" command. Dumpe2fs will print out the complete 540set of features when listing the superblock. 541 542Dumpe2fs has new options -f (force) and -h (header-only). 543 544Fixed a bug in e2fsck which could cause the PROGRAMMING ERROR/bonehead 545message to come up. This could happen when decrementing or 546incrementing a link count could result in an overflow. 547 548Fixed a bug in e2fsck where the block count on the lost+found 549directory would not be properly incremented when the directory was 550expanded to the point where an indirect block needed to be allocated. 551 552E2fsck now makes some additional sanity checks on the superblock to 553avoid crashing or giving a memory allocation error if some of the 554values in the superblock are unresonable (but the superblock otherwise 555looks valid). 556 557Fixed a bug in e2fsck where a very badly corrupted filesystem might 558require two passes to completely fix the filesystem. This happened if 559an inode claimed blocks that was part of the filesystem metadata 560(typically, when garbage was written into an inode table or indirect 561block, since this kind of filesystem corruption normally doesn't 562happen otherwise). 563 564On the Alpha, glibc declares st_flags although it isn't actually used; 565the configure script was improved to detect this case so that 566e2fsprogs can avoid using the non-functional stat field. 567 568The manual pages were updated to use a more consistent formatting 569style consistent with standard Unix man pages. Mke2fs's man page 570added documentation for a few previously undocumented options. 571 572Fixed minor display bugs in tune2fs and mke2fs. 573 574Programmer's notes: 575------------------- 576 577Improved portability of e2fsprogs to non-Unix systems (in particular, NT). 578 579Added features to parse and print feature strings into the e2p library. 580(e2p_feature2string, e2p_string2feature, e2p_edit_feature). 581 582ext2fs_mkdir() and ext2fs_new_dir_block() now creates directories 583whose directory entries contain proper filetype information if the 584filesystem supports it. 585 586ext2fs_link() now uses the low 3 bits of its flags parameter to pass 587the directory entry filetype information. This is used to set the 588directory entry filetype information if the filesystem supports it. 589 590Fixed a bug in ext2fs_expand_dir() where the block count in a 591directory's inode would not be properly incremented when the directory 592was expanded to the point where an indirect block needed to be 593allocated. 594 595 596E2fsprogs 1.15 (July 18, 1999) 597============================== 598 599Add configuration checks so that e2fsprogs will compile cleanly on 600Linux 2.3 kernels that have renamed i_version to i_generation. 601 602E2fsck now prints a progress/completion bar (and not just a simple 603spinner) if the -C0 option is requested or if it receives a SIGUSR1 604signal. Fsck will automatically manage the (potentially muliple) 605e2fsck processes to print completion bars if it is given a -C option, 606with the right thing happening if multiple filesystems are being 607checked in parallel. 608 609Mke2fs now has better automatic hueristics to determine the filesystem 610parameters to be used for a particular filesystem. Added a new option 611-T which allows the user to specify how the filesystem is to be used, 612which helps mke2fs do a better job selecting the filesystem parameters. 613 614Mke2fs now creates revision 1 filesystems by default, and with the 615sparse superblock feature enabled. The sparse superblock feature is 616not understood by Linux 2.0 kernels, so they will only allow read-only 617mounts of filesystems with this sparse superblocks. 618 619Fix bug where if /dev/null couldn't be opened (should never happen), 620e2fsck would hang in a tight loop. 621 622Make e2fsck handle the case where /lost+found isn't a directory. 623 624E2fsck now uses mallinfo if it exists to get accurate statistics about 625its memory usage. 626 627Fix bug in e2fsck where it wouldn't check to see if a disconnected 628inode had any problems before connecting it to /lost+found. 629 630Add check to e2fsck so it makes sure that total number of inodes in 631the filesystem is a sane number. 632 633Fix fencepost error when clearing an the end of the block bitmap which 634caused the last block in the bitmap not to get cleared. 635 636Cleaned up a number of messages in e2fsck: 637 * The message "Group's #'s copy of the group descriptor..." 638 was fixed so that the correct number would be displayed. 639 * Added missing space in the "disk write-protected" error messsage 640 * Cleaned up the error message printed when a non-interactive 641 e2fsck needs to abort a check because the filesystem 642 appears to be mounted. 643 644Added a new command-line utility, uuidgen, which will create and print 645a UUID. 646 647Make debugfs's icheck command more robust by checking to make sure an 648inode has valid blocks before interarting over the inode's blocks. 649 650UUID generation now uses a random-based scheme whenever possible to 651prevent potential privacy problems. 652 653Man pages for all of the UUID functions in the lirbary were added. 654 655Fixed bug in fsck so it won't coredump if a filesystem not in 656/etc/fstab is given to it. 657 658Fsck now understands the UUID=xxxx and LABEL=yyyy forms in /etc/fstab 659that most of the other mount utilities understands. 660 661Mke2fs will make a filesystem even if it appears mounted if the force 662option is given. 663 664Dumpe2fs has new command-line options which allow a filesystem expert 665to specify the superblock and blocksize when opening a filesystem. 666This is mainly useful when examining the remains of a toasted 667filesystem. 668 669The badblocks program has been updated to display correctly on disks 670with large block numbers. 671 672The badblocks program no longer gives spurious errors when errors 673occur on non-block boundaries, which is common if the blocksize is 674larger than 1k. 675 676Mke2fs will sync the disk device every MKE2FS_SYNC block groups if the 677MKE2FS_SYNC environment variable is set. This is to work around a VM 678bug in the 2.0 kernel. I've heard a report that a RAID user was able 679to trigger it even using a 2.2 kernel, but hopefully it will not be 680needed for most Linux 2.2 users. 681 682Fixed miscellaneous documentation and man pages. 683 684Programmer's notes: 685------------------- 686 687Cleaned up functions such as pass1_get_blocks, pass1_read_inode which 688in e2fsck's pass1.c really should have been static. 689 690The return value of the uuid_compare() function was changed to make it 691match with the convetions used by strcmp, memcmp, and Paul Leach's 692UUID sample document. 693 694The "make depend" process has now been made more automated; it now 695automatically word-wraps the dependencies, and only replaces source 696Makefile.in if there has been a change in the dependencies. Also, a 697top-level "make depend" now recurses through all the subdirectories 698automatically. 699 700The Makefile in .../util has been changed so that subst is built using 701the native C compiler during a cross-compilation, since the subst 702program is only used during the build process. Also add an explicit 703rule to build util/subst by cd'ing to the correct directory and 704running Makefile. 705 706The man directories are defined in terms mandir, so that the configure 707script can override the location of the manual pages. 708 709The config files have been updated to recognize new machine types for 710both the i386 and alpha families. 711 712Fsck has been modified so that it will accurately create an 713fsck_instance even when the noexecute flag is set. This allows for 714accurate debugging of the fsck pass structure. Also, when the verbose 715flag is given twice, fsck will print debugging information about when 716fsck is waiting for jobs to finish. 717 718 719E2fsprogs 1.14 (January 9, 1999) 720================================ 721 722Fix the fstab parsing code so that it can handle blank lines and 723comment characters. Also, missing pass numbers need to be treated as 724zero. 725 726Fixed a bug in e2fsck where under some circumstances (when e2fsck 727needs to restart processing after fixing an egregious inconsistency) 728it would try to access already freed memory. 729 730E2fsck now prints non-printable characters in directory entries and 731pathnames using '^' and 'M-' notation. 732 733Fixed chattr so that it will ignore symbolic links when doing 734recursive descent traversals. For both chattr and lsattr, no longer 735print the version string unless the -V option is given. 736 737Allow the system administrator to directly specify the number of 738inodes desired in the filesystem, for some special cases where this is 739necessary. 740 741Fix portability problems so that e2fsprogs can be compiled under Linux 7421.2 systems and Solaris systems. 743 744Update the config.guess file with a more recent version that will 745identify newer Linux platforms. 746 747Programmer's notes 748------------------ 749 750Ext2fs_read_inode and ext2fs_write_inode will now return an error if 751an inode number of zero is passed to them. 752 753E2fsprogs 1.13 (December 15, 1998) 754================================== 755 756Fixed a bug in debugfs where an error messages weren't getting printed 757when the ext2 library routines to read inodes returned errors in the 758stat, cmri and rm commands. 759 760Fixed a bug in mke2fs so that if a ridiculous inode ratio parameter is 761provided, it won't create an inode table smaller than the minimum 762number of inodes required for a proper ext2 filesystem. 763 764Fsck now parses the /etc/fstab file directly (instead of using 765getmntent()), so that it can distinguish between a missing pass number 766field and pass number field of zero. This caused problems for 767diskless workstations where all of the filesystems in /etc/fstab have 768an explicit pass number of zero, and fsck could not distinguish this 769from a /etc/fstab file with missing pass numbers. 770 771E2fsck will create a /lost+found directory if there isn't one in the 772filesystem, since it's safer to create the lost+found directory before 773it's needed. 774 775Fixed e2fsck so that it would detect bogus immutable inodes which 776happen to be sockets and FIFO files, and offer to clear them. 777 778If a filesystem has multiple reasons why it needs to be checked, and 779one of the reasons is that it is uncleanly mounted, e2fsck will print 780that as the reason why the filesystem is being checked. 781 782Cleaned up the output routines of mke2fs so that it doesn't overflow 783an 80 column display when formating really big filesystems. 784 785Added a sanity check to e2fsck to make sure that file descriptors 0, 7861, 2 are open before opening the hard disk. This avoids a problem 787where a broken program might exec e2fsck with those file descriptors 788closed, which would cause disastrous results if the kernel returns a 789file descriptor for the block device which is also used by FILE * 790stdout. 791 792Fixed up the e2fsck progress reporting functions so that the values 793reliably reach 100% at the completion of all of the e2fsck passes. 794 795Fixed minor documentation bugs in man pages and usage messages. 796 797Programmer's notes: 798------------------- 799 800Fixed a number of lint warnings in the ext2fs library and potential 801portability problems from other OS's header files that might define 802CPP macros for names like "max" and "min". 803 804ext2fs_badblocks_list_add() has been made more efficient when it needs 805to grow the bad blocks list. 806 807Fixed a bug in e2fsck which caused it to dereference a freed pointer 808just before exiting. 809 810Fixed the substition process for generating the mk_cmds and compile_et 811scripts so that they will work outside of the build tree. 812 813Add sanity check to e2fsck so that if an internal routine 814(ext2fs_get_dir_info) returns NULL, avoid dereferencing the pointer 815and causing a core dump. This should never happen, but... 816 817E2fsprogs 1.12 (July 9, 1998) 818================================== 819 820E2fsprogs now works with glibc (at least with the version shipped wtih 821RedHat 5.0). The ext2fs_llseek() function should now work even with 822i386 ELF shared libraries and if llseek() is not present. We also 823explicitly do a configure test to see if (a) llseek is in libc, and 824(b) if llseek is declared in the system header files. (See standard 825complaints about libc developers don't understand the concept of 826compatibility with previous versions of libc.) 827 828The ext2fs library now writes out the block group number in each of 829the superblock copies. This makes it easier to automatically 830determine the starting block group of the filesystem when the block 831group information is trashed. 832 833Added support for the EXT2_FEATURE_INCOMPAT_FILETYPE feature, 834which means that e2fsprogs will ignore the high 8 bits of the 835directory entry's name_len field, so that it can be used for other 836purposes. 837 838Added support for the EXT2_FEATURE_RO_COMPAT_LARGE_FILE feature. 839E2fsprogs will now support filesystems with 64-bit sized files. 840 841Added support for the EXT2_FEATURE_COMPAT_DIR_PREALLOC feature. 842 843Added new program "e2label", contributed by Andries Brouwer. E2label 844provides an easy-to-use interface to modify the filesystem label. 845 846Fixed bug so that lsattr -v works instead of producing a core dump. 847 848Fixed a minor bug in mke2fs so that all groups with bad superblock 849backup blocks are printed (not just the first one). 850 851Mke2fs will check the size of the device, and if the user specifies a 852filesystem size larger than the apparent size of the device it will 853print a warning message and ask if the user wants to proceed. 854 855E2fsck has a new option -C, which sends completion information to the 856specified file descriptor. For the most part, this is intended for 857programs to use, although -C 0 will print a spinning character to the 858stdout device, which may be useful for users who want to see something 859happening while e2fsck goes about its business. 860 861Fixed a bug in e2fsck which could cause a core dump when it needs to 862expand the /lost+found directory, and sometimes the bitmaps haven't 863been merged in. Also fixed a related bug where ext2fs_write_dir_block 864was used to write out a non-directory block. (Which would be bad on a 865non-Intel platform with byte swapping going on.) 866 867Fixed bug in e2fsck where it would print a "programming error" message 868instead of correctly identifying where a bad block was in used when 869the bad block was in a non-primary superblock or block group 870descriptor. Also fixed a related bug when sparse superblocks are in 871use and there is a bad block where a superblock or block group 872descriptor would have been in a group that doesn't include a 873superblock. 874 875Fixed a bug in e2fsck (really in libext2fs's dblist function) where if 876the block group descriptor table is corrupt, it was possible to try to 877allocate a huge array, fail, and then abort e2fsck. 878ext2fs_get_num_dirs() now sanity checks the block group descriptor, 879and subsitutes reasonable values if the descriptors are obviously bogus. 880 881If e2fsck finds a device file which has the immutable flag set and the 882i_blocks beyond the normal device number are non-zero, e2fsck will 883offer to remove it, since it's probably caused by garbage in the inode 884table. 885 886When opening a filesystem, e2fsck specially checks for the EROFS error 887code, and prints a specific error message to the user which is more 888user friendly. 889 890If the filesystem revision is too high, change the e2fsck to print 891that this is either because e2fsck is out of date, or because the 892superblock is corrupt. 893 894E2fsck now checks for directories that have duplicate '.' and '..' 895entries, and fixes this corruption. 896 897E2fsck no longer forces a sync of the filesystem (with attendant sleep 898calls) at all times. The ext2fs_flush() function now performs a sync 899only if it needed to write data blocks to disk. 900 901Fixed a minor bug in e2fsck's pass1b's file cloning function, where 902certain errors would not be properly reported. 903 904Updated and expanded a few points in the man pages which users 905complained wheren't explicit enough. 906 907Added special case byte-swapping code if compiling on the PowerPC, to 908accomodate the strange big-endian variant of the ext2 filesystem that 909was previously used on the PowerPC port. 910 911 912Programmer's notes: 913------------------- 914 915Removed C++ keywords from the ext2fs libraries so that it could be 916compiled with C++. 917 918E2fsck's internal organization has now been massively reorganized so 919that pass*.c don't have any printf statements. Instead, all problems 920are reported through the fix_problem() abstraction interface. E2fsck 921has also been revamped so that it can be called as a library from a 922application. 923 924Added new fileio primitives in libext2fs for reading and writing 925files on an unmounted ext2 filesystem. This interface is now used by 926debugfs. 927 928Added a new libext2fs function for mapping logical block numbers of 929a file to a physical block number. 930 931Added a new libext2fs function, ext2fs_alloc_block(), which allocates 932a block, zeros it, and updates the filesystem accounting records 933appropriately. 934 935Added a new libext2fs function, ext2fs_set_bitmap_padding(), which 936sets the padding of the bitmap to be all one's. Used by e2fsck pass 5. 937 938The libext2fs functions now use a set of memory allocation wrapper 939functions: ext2fs_get_mem, ext2fs_free_mem, and ext2fs_resize_mem, 940instead of malloc, free, and resize. This makes it easier for us to 941be ported to strange environments where malloc, et. al. aren't 942necessarily available. 943 944Change the libext2fs fucntion to return ext2-specific error codes 945(EXT2_DIR_EXISTS and EXT2_DB_NOT_FOUND, for example) instead of using 946and depending on the existence of system error codes (such as EEXIST 947and ENOENT). 948 949Renamed io.h to ext2_io.h to avoid collision with other OS's header 950files. 951 952Add protection against ext2_io.h and ext2fs.h being included multiple 953times. 954 955The types used for memory lengths, etc. have been made more portable. 956In generla, the code has been made 16-bit safe. Added Mark 957Habersack's contributed DOS disk i/o routines. 958 959Miscellaneous portability fixes, including not depending on char's 960being signed. 961 962The io_channel structure has a new element, app_data, which is 963initialized by the ext2fs routines to contain a copy of the filesystem 964handle. 965 966ext2fs_check_directory()'s callback function may now return the error 967EXT2_ET_CALLBACK_NOTHANDLED if it wishes ext2fs_check_directory() to 968really do the checking, despite the presence of the callback function. 969 970 971E2fsprosg 1.11 (June 17, 1997) 972============================== 973 974Fixed e2fsck to detect (previously ignored) conflicts between the 975superblock or block group descriptors and block bitmaps, inode 976bitmaps, and inode tables. 977 978Fixed bug in e2fsck so that when the message printed out when a block 979or inode bitmap conflicts with other data, it has the correct group 980number. 981 982Fixed bug in e2fsck and mke2fs where the blocksize wasn't being passed 983to badblocks. This meant that not all of the filesystem was being 984tested for bad blocks! 985 986Fixed an array boundary overrun case which cropped up in 987ext2fs_badblocks_list_test when a user tried running "mke2fs -c 988-b 4096". 989 990Adjusted the number of columns printed by mke2fs when displaying the 991superblock backups to avoid running over 80 columns when making a 992really big filesystem. 993 994Fixed up the man pages for e2fsck, debugfs, badblocks, chattr, 995dumpe2fs, fsck, mke2fs, and tune2fs (typos and other minor grammar 996fixes), thanks to some suggestions from Bill Hawes (whawes@star.net). 997 998Programmer's notes: 999------------------- 1000 1001Fixed install rule in lib/ss so that ss_err.h is actually getting 1002installed. 1003 1004Fixed bug in ext2fs_copy_bitmap; the destination bitmap wasn't getting 1005bassed back to the caller. 1006 1007Fixed bug in ext2fs_inode_scan_goto_blockgroup; it had not been 1008setting the current inode number (which meant this function wasn't 1009working at all). 1010 1011Fixed bug in ext2fs_resize_generic_bitmap; it had not be zeroing all 1012blocks in the bitmap when increasing the size of the bitmap. 1013 1014Changed the initial number of blocks allocated by ext2fs_init_dblist() 1015to be more realistic. 1016 1017Added a new function ext2fs_allocate_group_table, which sets up the 1018group descriptor information (and allocates inode and block bitmaps, 1019and inode tables for a particular group). The function was created by 1020factoring out code form ext2fs_allocate_tables(). 1021 1022Added a new function ext2fs_move_blocks which takes a bitmap of the 1023blocks to be moved, and moves them to another location on the 1024boardboard. 1025 1026Make the unix_io channel's io_channel_flush implementation calls sync() 1027to to flush the kernel buffers to disk. 1028 1029Added a new function ext2fs_dblist_count returns the number of 1030directory blocks in dblist. 1031 1032 1033E2fsprogs 1.10 (April 24, 1997) 1034=============================== 1035 1036Mke2fs once again defaults to creating revision #0 filesystems, since 1037people were complaining about breaking compatibility with 1.2 kernels. 1038Warning messages were added to the mke2fs and tune2fs man pages that 1039the sparse superblock option isn't supported by most kernels yet (1.2 1040and 2.0 both don't support parse superblocks.) 1041 1042Added new flag to mke2fs, -R <raid options>, which allows the user to 1043tell mke2fs about the RAID configuration of the filesystem. Currently 1044the only supported raid option is "stride" which specifies the width 1045of the RAID stripe. 1046 1047Fixed bug in e2fsck where pass1b would bomb out if there were any 1048blocks marked bad in the inode table. 1049 1050Fixed rare bug in mke2fs where if the user had a very unlucky number 1051of blocks in a filesystem (probability less than .002) the resulting 1052filesystem would be corrupt in the last block group. 1053 1054Fixed bug where if e2fsck tried to allocate a block to fix a 1055filesystem corruption problem and the filesystem had no free blocks, 1056ext2fs_new_block() would loop forever. 1057 1058The configure script now checks explicitly to see if "-static" works, 1059since that can't be assumed to be true --- RedHat doesn't install 1060libc-static by default. 1061 1062Fixed bug in libext2's block iterator functions where under some 1063cirmcustances, file with holes would cause the bcount parameter to the 1064callback function to be incorrect. This bug didn't affect any of 1065e2fsprogs programs, but it was discovered by Paul Mackerras, the 1066author of the PPC boot loader. 1067 1068Removed use of static variables to store the inode cache in libext2fs. 1069This caused problems if more than one filesystem was accessed via 1070libext2fs (static variables in libraries are generally a bad idea). 1071Again, this didn't affect e2fsprogs programs, but it was discovered by 1072Paul Mackerras. 1073 1074Fixed minor bugs and version code drift to assure that e2fsprogs 1.10 1075will compile cleanly with 1.2.13 kernels (even with a.out shared 1076libraries!) 1077 1078Programmer's notes: 1079------------------- 1080 1081Added new functions to duplicate an ext2 filesystem handle, and its 1082associated substructure. New functions: ext2fs_dup_handle(), 1083ext2fs_copy_dblist(), ext2fs_badblocks_copy(), ext2fs_copy_bitmap(). 1084Other structures, such as the io_channel and the inode_cache, now have 1085a ref count so that they only get freed when they are no longer used 1086by any filesystem handle. (These functions were added as part of the 1087development effort for an ext2 resizer). 1088 1089E2fsprogs 1.09 (April 14, 1997) 1090=============================== 1091 1092Fixed bug in mke2fs (really in lib/ext2fs/initialize.c) which was 1093accidentally introduced in the 1.08 release. The overhead calculation 1094was accidentally removed, which caused ext2fs_initialize() to not 1095notice when the filesystem size needed to be adjusted down because 1096there wasn't enough space in the last block group. 1097 1098Fixed bug in version parsing library routine; it was always parsing 1099the library version string, instead of using the passed-in string. 1100 1101Clarified chattr man page. 1102 1103E2fsprogs 1.08 (April 10, 1997) 1104=============================== 1105 1106E2fsck 1.07 was very slow when checking very large filesystems with a 1107lot of files that had hard links (i.e., news spools). This was fixed 1108by seriously revamping the icount abstraction. Added a formal test 1109suite for the icount abstraction. 1110 1111Debugfs now has a "-l" option to the "ls" command, which lists the 1112inode number, permissions, owner, group, size, and name of the files 1113in the directory. 1114 1115Fix a bug in e2fsck where when a directory had its blocks moved to 1116another location during the pass 1b processing, the directory block 1117list wasn't updated, so pass 2 wouldn't check (and correct) the 1118correct directory block. 1119 1120E2fsck will now treat inodes which contain blocks which are claimed by 1121the filesystem metadata by treating them as multiply claimed blocks. 1122This way, the data in those blocks can be copied to a new block during 1123the pass 1b--1d processing. 1124 1125E2fsck will attempt to determine the correct superblock number and 1126display it in the diagnostic and warning messages if possible. 1127 1128Add support for a new (incompatible) feature, "sparse_super". This 1129feature reduces the number of blocks which contain copies of backup 1130superblocks and block group descriptors. (It is only an incompatible 1131feature because of a bug in ext2_free_blocks.) mke2fs and tune2fs now 1132support a new -s option; e2fsck will recognize filesystems built with 1133this feature turned on. 1134 1135E2fsck now checks the library to make sure is the correct version, 1136using new library functions. (This helps to diagnose incorrectly 1137installed e2fsprogs distributions.) 1138 1139Dumpe2fs now prints more information; its now prints the the 1140filesystem revision number, the filesystem sparse_super feature (if 1141present), the block ranges for each block group, and the offset from 1142the beginning of the block group. 1143 1144Mke2fs now distributes the inode and block bitmap blok so that the 1145won't be concentrated in one or two disks in RAID/striping setups. 1146Also, if the user chooses a 2k or 4k block group, mke2fs will try to 1147choose the largest blocks per group that be chosen. (For 2k blocks, 1148you can have up to 16384 blocks/group; for 4k blocks, you can have up 1149to 32768 blocks/group.) Previously mke2fs would not allow 1150specification of more than 8192 blocks per group, even if you were 1151using a 2k or 4k block group. 1152 1153Programmer's notes: 1154------------------- 1155 1156Added a new function ext2fs_create_icount2() which takes a "hint" 1157argument. This hint argument presets the icount array with the list 1158of inodes which actually need to be in the icount array. This really 1159helps to speed up e2fsck. 1160 1161Added a new function ext2fs_icount_validate() which checks the rep 1162invariant for the icount structure. This is used mostly for testing. 1163 1164The error mesasage given when a bad inode number is passed to 1165test_generic_bitmap to reflect EXT2FS_TEST_ERROR (instead of 1166EXT2FS_UNMARK_ERROR). 1167 1168Added a new function ext2fs_set_dir_block which sets the block of a 1169dblist entry, given the directory inode and blockcnt. 1170 1171Added a new function ext2fs_get_library_version() which returns the 1172current library version, and ext2fs_parse_version_string() which 1173returns a version number based on a e2fsprogs version string. 1174 1175The icount functions will return EINVAL if the passed in inode number 1176is out of bounds. 1177 1178E2fsprogs 1.07 (March 9, 1997) 1179============================== 1180 1181E2fsck is now uses much less memory when checking really large 1182filesystems (or rather, filesystems with a large number of inodes). 1183Previously a filesystem with 1 million inodes required 4 megabytes of 1184memory to store inode count statistics; that storage requirement has 1185now been reduced to roughly half a megabyte. 1186 1187E2fsck can now properly deal with bad blocks appearing inside the 1188inode table. Instead of trying to relocate the inode table (which 1189often failed because there wasn't enough space), the inodes in the bad 1190block are marked as in use. 1191 1192E2fsck will automatically try to use the backup superblocks if the 1193primary superblocks have a bad magic number or have missing meta-data 1194blocks (or meta-data blocks which are out of range). 1195 1196E2fsck's pass 3 has been made more efficient; most noticeable on 1197filesystems with a very large number of directories. 1198 1199Completely revamped e2fsck's system of printing problem reports. It 1200is now table driven, to make them more easily customizeable and 1201extendable. Error messages which can be printed out during preen mode 1202are now one line long. 1203 1204Fixed e2fsck's filesystem swapping code so that it won't try to swap 1205fast symbolic links or deleted files. 1206 1207Fixed e2fsck core dumping when fixing a filesystem which has no 1208directories (not even a root directory). 1209 1210Added a check to e2fsck to make sure that the length of every 1211directory entry is a multiple of 4 (since the kernel complains if it 1212isn't). 1213 1214Added a check to e2fsck to make sure that a directory entry isn't a 1215link to the root directory, since that isn't allowed. 1216 1217Added a check to e2fsk to now make sure the '.' and '..' directory 1218entries are null terminated, since the 2.0 kernel requires it. 1219 1220Added check to write_bitmaps() to make sure the superblock doesn't get 1221trashed if the inode or block bitmap is marked as being block zero. 1222 1223Added checking of the new feature set fields in the superblock, to 1224avoid dealing with new filesystem features that this package wasn't 1225set up to handle. 1226 1227Fixed a fencepost error in ext2fs_new_block() which would occasionally 1228try to allocate a block beyond the end of a filesystem. 1229 1230When the UUID library picks a random IEEE 802 address (because it 1231can't find one from a network card), it sets the multicast bit, to 1232avoid conflicting with a legitimate IEEE 802 address. 1233 1234Mke2fs now sets the root directory's owner to be the real uid of the 1235user running mke2fs. If the real uid is non-zero, it also sets 1236the group ownership of the root directory to be the real group-id of 1237the user running mke2fs. 1238 1239Mke2fs now has more intelligent error checking when it is given a 1240non-existent device. 1241 1242When badblocks is given the -vv option, it now updates the block that 1243it is currently testing on every block. 1244 1245Fixed a bug in fsck where it wouldn't modify the PATH envirnoment 1246currently correctly if PATH wasn't already set. 1247 1248Shared libraries now built with dependencies. This allows the shared 1249library files to be used with dlopen(); it also makes the transition 1250to libc 6 easier, since ld.so can tell which libc a particular shared 1251library expects to use. 1252 1253Programmer's notes: 1254------------------- 1255 1256Added new abstraction (defined in dblist.c) for maintaining a list of 1257blocks which belongs to directories. This is used in e2fsck and other 1258programs which need to iterate over all directories. 1259 1260Added new functions which test to see if a contiguous range of blocks 1261(or inodes) are available. (ext2fs_*_bitmap_range). 1262 1263Added new function (ext2_inode_has_valid_blocks) which returns true if 1264an inode has valid blocks. (moved from e2fsck code). 1265 1266Added new function (ext2fs_allocate_tables) which allocates the 1267meta-data blocks as part of initializing a filesystem. (moved from 1268mke2fs code). 1269 1270Added a new I/O manager for testing purposes. It will either allow a 1271program to intercept I/O requests, or print debugging messages to 1272trace the activity of a program using the I/O manager. 1273 1274The badblocks_list functions now store the bad blocks in a sorted 1275order, and use a binary search to speed up badblocks_list_test. 1276 1277The inode scan function ext2fs_get_next_inode() may now return a soft 1278error returns: MISSING_INODE_TABLE and BAD_BLOCK_IN_INODE_TABLE in 1279those cases where part of an inode table is missing or there is a bad 1280block in the inode table. 1281 1282Added a new function (ext2fs_block_iterate2) which adds new arguments to 1283the callback function to return a pointer (block and offset) to the 1284reference of the block. 1285 1286Added new function (ext2fs_inode_scan_goto_blockgroup) which allows an 1287application to jump to a particular block group while doing an inode 1288scan. 1289 1290The badblocks list functions were renamed from badblocks_* to 1291ext2fs_badblocks_*. Backwards compatibility functions are available 1292for now, but programs should be modified to use the new interface. 1293 1294Some of the library functions were reorganized into separate files to 1295reduce the size of some programs which statically link against the 1296ext2 library. 1297 1298Put in some miscellaneous fixes for the Alpha platform. 1299 1300 1301E2fsprogs 1.06 (October 7, 1996) 1302================================ 1303 1304Fixed serious bug in e2fsck: if the block descriptors are bad, don't 1305smash the backup copies in ext2fs_close(). (The problem was that when 1306e2fsck -p discovered the problem, while it was closing the filesystem 1307and exiting, it was also blowing away the backup superblocks on the 1308disk, which was less than friendly.) We now make it the case that we 1309only write out the backup superblock and the back block descriptors if 1310the filesystem is completely free from problems. 1311 1312Fixed a bug in block_interate in the lib/ext2fs library which caused 1313e2fsck to fail on GNU Hurd-created filesystems. 1314 1315Add support for Linux/FT's bootloader, which actually uses 1316EXT2_BOOT_LOADER, and sets its mode bits which caused e2fsck to want 1317to clear the inode. 1318 1319Add support for the "A" (no atime update) attribute. (Note: this 1320attribute is not yet in production kernels.) 1321 1322The test suite is not automatically run when doing a "make all" from 1323the top level directory. Users should manually run "make check" if 1324they wish to run the test suite. 1325 1326Upon a preenhalt(), make the printed message more explicit that 1327running e2fsck "MANAULLY" means without the -p or -a options. 1328 1329In e2fsck, if a disconnected inode is zero-length, offer to clear it 1330instead of offering to connect it to lost+found. 1331 1332In e2fsck, if a filesystem was just unmounted uncleanly, and needs 1333e2fsck to be run over it, change e2fsck to explicitly display this 1334fact. 1335 1336For dumpe2fs and e2fsck, cause the -V option to print out which 1337version of the ext2fs library is actually getting used. (This will 1338help detect mismatches of using a 1.06 utility with a 1.05 library, 1339etc.) 1340 1341Programmers' notes: 1342------------------- 1343 1344EXT2_SWAP_BYTES was changed to EXT2_FLAG_SWAP_BYTES, which better fits 1345the naming convention. 1346 1347In ext2fs_initialize(), make sure the description for the inode bitmap 1348is correctly initialize. 1349 1350Fixed minor type typo in ext2fs_allocate_generic_bitmap(); 1351 1352E2fsprogs 1.05 (September 7, 1996) 1353================================== 1354 1355Add support for new fields in the ext2 superblock --- volume name, 1356volume UUID, and last mounted field. Dumpe2fs displays these fields, 1357tune2fs and mke2fs allows you to set them. E2fsck will automatically 1358generate a UUID for those volumes that don't have them. 1359 1360Put in support for e2fsck to recognize HURD specific ext2 features --- 1361most notably, the translator block. The e2fsprogs tools will now use 1362the creator_os field in the superblock to correctly handle different 1363OS-specific variants of the ext2 filesystem. 1364 1365E2fsck now fixes inodes which have a the deletion time set, but which 1366have a non-zero i_link_count field by offering to clear the deletion 1367time. Previously e2fsck assumed that the inode was deleted (per 0.3c 1368ext2 kernel behavior) and offered to unlink the file. 1369 1370If e2fsck sets the clean bit, but nothing else, set the exit code 1371FSCK_NONDESTRUCT. After all, e2fsck did fix a filesystem error --- it 1372set the filesystem valid bit when it was previously cleared. :-) This 1373was needed to make the HURD fsck driver happy. 1374 1375If the user refuses to attach an unattached inode, e2fsck will no 1376longer set the inode's link count. Otherwise, the inode would end up 1377getting marked as unused, which might cause loss of data later. 1378 1379Make the message issued by e2fsck when the superblock is corrupt less 1380confusing for users. It now mentions that another reason for the 1381"corrupt superblock" message might be that the partition might not be 1382an ext2 filesystem at all (it might swap, msdos filesystem, ufs, etc.) 1383 1384Make the libext2 library more robuest so that e2fsck won't coredump on 1385an illegal superblock where the blocksize is zero. (f_crashdisk is 1386the test case). 1387 1388By default, create filesystems where the default checkinterval is 6 1389months (180 days). Linux servers can be robust enough that 20 reboots 1390can be a long, long time. 1391 1392Added configure flag --enable-old-bitops, which forces the bitops to 1393use the old (native) bitmask operations. By default on the sparc 1394platform, the standard ext2 bit ordering is now used. 1395 1396Added a new feature to e2fsck to byte-swap filesystems; this can be 1397used to convert old m68k filesystems to use the standard byte-order 1398storage for the superblock, inodes, and directory blocks. This 1399function is invoked by using the '-s' option to e2fsck. 1400 1401Debugfs's "dump" command has been enhanced so that it writes out the 1402exact size of the file so that the nulls at the end of the file are 1403eliminated. The command also accept a new "-p" option which will 1404attempt preserve to preserve the ownernship, permissions, and 1405file modification/access times. 1406 1407Debugfs has two new options, -f and -R. The -R option allows the user 1408to execute a single debugfs command from the command line. The -f 1409option allows the user to specify a "command file" containing debugfs 1410commands which will get executed. 1411 1412Dumpe2fs now pretty prints the check interval, instead of just 1413printing the check interval as a number of seconds. 1414 1415Fix bugs in debugfs: the params command when no filesystem is opened 1416no longer causes a core dump. It is now possible to unlink a file 1417when a pathame containing a '/' is specified. 1418 1419Tune2fs has a new -C option which sets the number of times the 1420filesystem has been mounted. 1421 1422Fix the chattr '-v' option so that it actually works. Chattr was 1423being buggy about the -v option parsing. 1424 1425Programmers' notes: 1426------------------- 1427 1428The directory lib/uuid contains a set of library routines to generate 1429DCE compatible UUIDs. 1430 1431Extended ext2fs_namei() to handle symbolic links. Added new function 1432ext2fs_nami_follow() which will follow last symbolic link in the case 1433where the pathname points to a sym link. 1434 1435The ext2fs_block_iterate function will now return the HURD translator 1436block, if present. The new flag BLOCK_FLAG_DATA_ONLY will cause the 1437iterator to return data blocks only. The ext2fs.h file now defines 1438constants BLOCK_COUNT_IND, BLOCK_COUNT_DIND, BLOCK_COUNT_TIND, and 1439BLOCK_COUNT_TRANSLATOR, which are the magic values passed in the block 1440count field of the iterator callback function. 1441 1442The test script driver now takes an optional second argument, which is 1443the test case to be run. This allows you to run a test case without 1444needing to run the entire test suite. 1445 1446On Linux ELF systems, install the .so files in the correct places 1447(/usr/lib). The .so files must be stored in the same directory as the 1448.a files. 1449 1450Fixed miscellaneous HURD compilation issues with header file being 1451included in the right order. 1452 1453Fixed debugfs so that it resets optind to zero, not one, since setting 1454optind to zero is more correct. 1455 1456 1457E2fsprogs 1.04 (May 16, 1996) 1458============================= 1459 1460First "official" (1.03 was a limited release only) to support building 1461e2fsprogs under Linux 2.0 kernels (as well as late model 1.3 and 1.99 1462kernels). 1463 1464This package includes a RPM specs file, that it can be built using the 1465RedHat Package Manager. 1466 1467E2fsck now prints a hint that if there are lots of incorrectly located 1468inode bitmaps, block bitmaps, and inode table blocks, the user might 1469want to try using e2fsck -b 8193 first, to see if that fares any 1470better. 1471 1472For ext2 filesystem written with the hurd, debugfs will now print out 1473the translator field when printing an inode structure. 1474 1475Lots of miscellaneous linking/installation cleanups: 1476 1477 Libraries are now linked using a relative pathname, instead of 1478 relying on -L working correct. It doesn't, in many cases, including 1479 current versions of GNU ld. This guarantees that the build tree is 1480 linking with the right libraries, instead of the ones installed in 1481 /usr/lib. 1482 1483 Header files, man pages, and the et/ss shell scripts are now 1484 generated using a custom substitution script, instead of relying on 1485 the configure script. This prevents needless recompilation of 1486 files; in addition, the custom substitution script is much faster. 1487 1488 e2fsck may now be linked dynamically, by using the 1489 --enable-dynamic-e2fsck flag to configure. This is not recommended, 1490 since it increases e2fsck's dependence on other files, but some 1491 people need to save disk space, and other critical programs on their 1492 systems were being linked dynamically anyway. 1493 1494 Programs such as fsck which didn't need to be linked against 1495 libext2fs (or mke2fs which didn't need to be linked against libe2p) 1496 only link against libraries they actually need. Otherwise, those 1497 programs would require the presense of libraries that otherwise 1498 could be removed from a rescuse diskette. 1499 1500 The ss include files are now installed correctly so they can 1501 actually be used by another package. 1502 1503 If the profiling libraries are built, they are now installed on a 1504 "make install-libs". 1505 1506 1507E2fsprogs 1.03 (March 27, 1996) 1508=============================== 1509 1510Change the m68k bit numbering for bitmasks to match the bit numbering 1511used by all other ext2 implementations. (This change was requested by 1512the m68k kernel development team.) 1513 1514Support (in-development) filesystem format revision which supports 1515(among other things) dynamically sized inodes. 1516 1517Fixed a bug in the ext2 library so that an intelligent error is 1518returned if mke2fs is run with a ridiculously small number of blocks 1519for a partition. 1520 1521Fixed a bug in the ext2 library which required that the device be 1522openable in read/write mode in order to determine its size. This 1523caused e2fsck -n to require read/write access when it was not 1524previously necessary. 1525 1526Fixed a bug in e2fsck which casued it to occasionally fail the test 1527suite depending on which version of the floating point library it was 1528using. 1529 1530Fixed a bug in e2fsck so that it now halts with a fatal error when 1531certain superblock consistency checks fail. Previously it continued 1532running e2fsck, with some potential confusing/damaging consequences. 1533 1534Added new flag to fsck which allows the root to be checked in parallel 1535with other filesytems. This is not the safest thing in the world to 1536do, but some system administrators really wanted it. 1537 1538Fixed -Wall flames in lib/ss. 1539 1540 1541E2fsprogs 1.02 (January 16, 1996) 1542================================= 1543 1544Fix to allow e2fsprogs to be compiled on recent 1.3 (pl45+) kernels. 1545 1546Change e2fsck to print statistics of how many non-contiguous files are 1547on the system. Note that a file which is larger than 8k blocks, it is 1548guaranteed to be non-contiguous. 1549 1550In mke2fs, print a warning message if a user tries to format a whole 1551disk (/dev/hda versus /dev/hda1). If a user really wants to format a 1552whole disk, the -F (force) option forces mke2fs to format a whole disk 1553as a filesytem. 1554 1555Fix a bug in fsck where in some cases it might start checking 1556partitions in the next pass before it finishes checking partitions in 1557the current pass. This still won't cause two partitions on the same 1558disk will be checked, so it's rarely a problem in real life. 1559 1560Patch lsattr so that it won't hang when checking a named pipe. 1561 1562Minor compilation fixes: 1563 * Fix the order of libraries that were linked in debugfs. 1564 * Allow the sources to be compiled with -ansi turned on. 1565