RELEASE-NOTES revision d6a3e41172ce43b4a92fae0791beb7592a4575b1
1E2fsprogs 1.42.4 (June 12, 2012) 2================================ 3 4Fixed more 64-bit block number bugs (which could end up corrupting 5file systems!) in e2fsck, debugfs, and libext2fs. 6 7Fixed e2fsck's handling of the journal's s_errno field. E2fsck was 8not properly propagating the journal's s_errno field to the superblock 9field; it was not checking this field if the journal had already been 10replayed, and if the journal *was* being replayed, the "error bit" 11wasn't getting flushed out to disk. 12 13Fixed a false positive complaint by e2fsck if all of the extents in 14the last extent tree block are uninitalized and located after the end 15of the file as defined by i_size. 16 17The dumpe2fs will now display the journal's s_errno field if it is 18non-zero, and it will also display the journal's 64-bit feature flag 19if present. 20 21Fix e2fsck so that it always opens the device file in exclusive mode 22when it might need to modify the file system, and never if the -n 23option is specified. (Previously, there were a few corner cases where 24it might get things wrong in either direction.) 25 26E2fsck now correctly truncates or deallocates extent-mapped inodes on 27the orphan list. The root cause was a bug in libext2fs's block 28iterator which could end up skipping an extent when the last block in 29an extent is removed, causing the current extent to be removed from 30the extent tree. 31 32E2fsck now correctly sets the global free block and inode counts when 33truncating or removing inodes on the orphan list in preen mode. 34Previously, it would leave these values would be set incorrectly, 35which is largely a cosmetic issue since the kernel no longer pays 36attention to those fields, but it can cause spurious complaints in 37subsequent e2fsck runs. 38 39Fix i_blocks accounting when the libext2fs library needs to add or 40remove an extent tree block on bigalloc file systems. 41 42The lsattr and chattr programs now support the No_COW flag for the 43benefit of btrfs. 44 45Debugfs now interprets date strings of the form @ddd as ddd seconds 46after the beginning of the epoch. This is handy when setting an inode 47number into the d_time field when debugging orphan list handling. 48 49Fix a precedence bug with built-in quota support which might result in 50e2fsck paying attention to the quota inode field even if the built-in 51quota feature flag is not set. Fortunately, in practice that 52superblock field should be zero for non-built-in quota file systems, 53so it's unlikely this bug would have caused problems. 54 55Updated/fixed various man pages. (Addresses Debian Bugs: #674453, 56#674694) 57 58Programmer's Notes 59------------------ 60 61The regression test suite can now run the integration tests in the 62tests directory in parallel, via "make -jN check". 63 64Add new test, f_zero_extent_length which tests e2fsck's handling of 65the case where all of the extents in the last extent tree block are 66uninitialized extents after i_size. 67 68Add a new test, f_jnl_errno, which checks handling of an error 69indication set in the journal superblock. 70 71Fix the test f_jnl_64bit so that it properly checks e2fsck's handling 72of a 64-bit journal. 73 74Add two tests, f_orphan_indirect_inode and f_orphan_extent_inode which 75tests e2fsck's handling of orphan inodes in preen mode, and truncation 76of extent inodes on the orphan list. 77 78Fixed more OS X portability issues. 79 80 81E2fsprogs 1.42.3 (May 14, 2012) 82=============================== 83 84Fix a bug in the Unix I/O manager which could cause corruption of file 85systems with more than 16TB when e2fsprogs is compiled in 32-bit mode 86(i.e., when unsigned long is 32-bits). Also fix a bug which caused 87dumpe2fs to incorrectly display block numbers > 32-bits. 88 89Improve the support for integrated quota files (where quota is a first 90class supported feature using hidden files in the ext4 file system). 91Previously the quota file was getting rewritten even when it was not 92necessary, and e2fsck would erroneously try to hide quota files which 93were already hidden. 94 95Quiet complaints in e2fsck when the total free blocks or inodes are 96incorrect in the superblock after an system crash, since we don't 97update nor depend on the superblock summaries at each commit boundary. 98 99Fixed a regression introduced in 1.42.2 which would cause applications 100outside of e2fsprogs which did not pass the EXT2_FLAG_64BIT (and so 101would were still using the legacy 32-bit bitmaps) to crash. This was 102due to missing 32-bit compat code in side the function 103ext2fs_find_first_zero_generic_bmap(). (Addresses Red Hat Bugzilla: 104#808421) 105 106Fix a bug which would cause mke2fs to fail creating the journal if 107/etc/mtab and /proc/mounts are missing. (Addresses Sourceforge Bug: 108#3509398) 109 110Updated/fixed various man pages. 111 112Update Czech, Dutch, French, German, Polish, Sweedish, and Vietnamese 113translations 114 115Fixed various Debian Packaging issues. 116 117 118Programmer's Notes 119------------------ 120 121Change the nonsensically wrong types in the function signature of the 122inline function ext2fs_find_first_zero_block_bitmap2(). This was 123caused by a cut and paste error; fortunately no code in e2fsprogs used 124this inline function, and there are any users of this functions 125outside of e2fsprogs. 126 127Add support for systems which have valloc(), but which do have 128posix_memalign() nor memalign() (such as MacOS 10.5). 129 130Refactor and clean up the allocation of aligned buffers for Direct I/O 131support. Previously some allocations were requesting a greater 132alignment factor that what was strictly necessary. Also optimize 133reading and writing bitmaps using Direct I/O when the size of the 134bitmap did not fully cover the file system blocksize. 135 136Reserve the codepoints for the INCOMPAT features LARGEDATA and 137INLINEDATA. 138 139Improved the regression test suite by adding some new integration 140tests (f_jnl_32bit, f_jnl_64bit) which detect breakage of the on-disk 141jbd2 format, as well as f_eofblocks which tests the new handling of 142uninitialized and initialized blocks beyond i_size. Also add a new 143unit test which verifies 32-bit bitmaps support and the new 144find_first_zero primitives. 145 146Add a few dependencies to fix parallel (make -j) builds. 147 148Removed bash'isms which were breaking the regression test suite on 149systems where /bin/sh is not bash. 150 151The config.guess and config.sub have been updated to the 2012-02-10 152version. 153 154Fix a portability problem caused by assuming the present of mntent.h 155means that setmntent() exists. Instead, explicitly test for this in 156the configure script. 157 158If the sys/signal.h header file does not exist, don't try to include 159it, since it's not available on all systems. 160 161Add support for systems that do not support getpwuid_r() 162 163The configure script now supports a new option, 164--enable-relative-symlinks, which will install relative symlinks for 165the ELF shared library files. (Addresses Sourceforge Bug: #3520767). 166 167When building BSD shared libraries make sure the LDFLAGS variable is 168passed to the linker. Fixing this allows, for example, e2fsprogs to 169be built in 32-bit mode on Mac OS X Lion (Addresses Sourceforge Bug: 170#3517272) 171 172Fix gcc -Wall nitpicks 173 174 175E2fsprogs 1.42.2 (March 27, 2012) 176================================= 177 178The resize2fs program uses much less CPU and is much faster for very 179large file systems. (Addresses Debian Bug: #663237) 180 181The seti and freei commands in debugfs can now take an optional length 182argument to set and clear a contiguous range of inodes. 183 184E2fsck will now make explicit checks for the EOFBLOCKS_FL, since we 185plan to remove support for it from the kernel file system driver. It 186really wasn't very useful and was causing more problems than it 187solves. Since e2fsck will complain if inodes that should have 188EOFBLOCKS_FL do not have the flag set, we are going to remove this 189check from e2fsprogs first, and then only remove the flag from the 190kernel much later. 191 192The mke2fs program can now use direct I/O via "mke2fs -D". This will 193slow down the mke2fs, but it makes it more polite on a loaded server 194by limiting the amount of memory that gets dirtied by mke2fs when it 195is using buffered I/O. 196 197E2fsck was needlessly closing and re-opening the file system as a side 198effect of adding Multiple Mount Protection (MMP). This isn't 199necessary for non-MMP file systems, so drop it. 200 201Print errors returned by ext2fs_open2() and ext2fs_check_desc() so we 202can more easily diagnose memory allocation failures caused by 203insufficient memory. E2fsck will now abort if there are memory 204allocation failures when the file system is initially opened and 205during the block group descriptor checks. (Addresses Google Bug: 206#6208183) 207 208If there are incorrect block group checks, e2fsck will now report the 209incorrect and corrected checksum values. 210 211The e2fsck progam can now write log files containing the details of 212the problems that were found and fixed directly, via configuration 213parameters in /etc/e2fsck.conf. 214 215Added the ability to limit the number of messages reported by e2fsck 216for a given problem type. This avoids a potential bottleneck if there 217is a serial console which can cause a boot sequence to take a long 218time if e2fsck needs to report many, many file system errors. 219 220The dumpe2fs, debugfs, and tune2fs now use rbtree bitmaps, which cause 221them to use much less memory for large file systems. 222 223The dumpe2fs program will now print the expected block group checksum 224if it is incorrect. This helps to diagnose problems caused by 225incorrect block group checksums. 226 227E2fsck now checks for extents with a zero length, since the kernel 228will oops if it comes accross such a corrupted data structure. (See 229https://bugzilla.kernel.org/show_bug.cgi?id=42859) 230 231E2fsck has a number of bugs relating to discard that have been fixed. 232(1) Fixed a bug which could cause e2fsck to discard portions of the 233inode table which were actually in use. (2) E2fsck will now avoid 234using discard if the block device doesn't zero data on discard, since 235otherwise this could cause problems if the file system gets corrupted 236in the future. (3) E2fsck will now avoid using discard when it is run 237in read-only mode. (4) Fixed a bug which caused e2fsck to not issue 238discards in the last block group. 239 240E2fsck's CPU utilization in pass 5 has been optimized, which will 241speed up e2fsck slightly. 242 243E2image will now skip copying uninitialized bitmap and inode table 244blocks. 245 246Fixed mke2fs -S so it does not corrupt the first block group's 247information. 248 249E2fsck will now check the new sysfs interface to determine if we are 250using the battery or AC mains. (Addresses SourceForge Bug: #3439277) 251 252Updated/fixed various man pages. (Addresses Debian Bug: #665427) 253 254Fixed various Debian Packaging issues. (Addresses Debian Bug: #665885) 255 256Programmer's Notes 257------------------ 258 259Fixed various portability issues for non-Linux systems, particularly 260MacOS X, as well as Linux systems running with the just-released glibc 2612.15. 262 263Fix file descriptor leak in ext2fs_close() if the file system with 264uninit_bg is opened read/only with a backup superblock. (Addresses 265SourceForge Bug: #3444351) 266 267Fixed an invalid return in a non-void function in the quota code. 268(Addresses SourceForge Bug: #3468423) 269 270Fixed the debian rules file so that the calls to dpkg-buildflags works 271when the shell is dash. 272 273The debian package build now uses V=1 so that there is more 274information about potential build failures in debian buildd logs. 275 276If the uninit flags get cleared by functions such as 277ext2fs_new_inode() or ext2fs_new_block2(), we now make sure the 278superblock is marked dirty and the block group descriptor checksum is 279updated if necessary. 280 281The debian rules file will now try to load debian/rules.custom of it 282exists. This flie can skip various builds for speed reasons if there 283is no need for the e2fsck-static or udeb packages. Available 284customizations in the rules file includes SKIP_STATIC=yes, 285SKIP_BF=yes, and SKIP_DIETLIBC=yes. 286 287In addition, if the file misc/mke2fs.conf.custom.in exists in the 288source tree, it will be used instead of the standard misc.conf file in 289the upstream sources. This makes it easier for Debian-derived systems 290to distribute a custom mke2fs.conf file without having to worry about 291merge issues if future versions of e2fsprogs makes changes in the 292upstream default version of mke2fs.conf. 293 294 295E2fsprogs 1.42.1 (February 17, 2012) 296=================================== 297 298The mke2fs and e2fsck now use significantly less memory when creating 299or checking very large file systems. This was enabled by adding 300extent-based bitmaps which are stored using a red-block tree, since 301block and inode allocations tend to be contiguous. 302 303The command mke2fs -S is used as a last ditch recovery command to 304write new superblock and block group descriptors, but _not_ to destroy 305the inode table in hopes of recovering from a badly corrupted file 306system. So if the uninit_bg feature is enabled, mke2fs -S will now 307set the unused inodes count field to zero. Otherwise, e2fsck -fy 308after using mke2fs -S would leave the file system completely empty. 309 310Since mke2fs recognizes mke3fs in argv[0] to mean "mkfs.ext3", also 311honor "mke4fs" to work the same as "mke2fs.ext4", since RHEL5 has 312installed an mke2fs binary using that name. 313 314The usage and help messages for the -G, -t and -T options in mke2fs 315have been fixed. 316 317If e2fsck needs to use the backup group descriptors, the 318ext2fs_open2() function clears the UNINIT bits to ensure all of the 319inodes in the file systems get scanned. However, the code which reset 320the UNINIT flags did not also recalculate the checksum, which produced 321many spurious (and scary) e2fsck messages. This has been fixed by 322resetting cheksums when the UNINIT bits are cleared. 323 324Relax a check in e2fsck which required that the block bitmap to be 325initialized when the inode bitmap is in use. This will allow us to 326eventually eliminate code from the kernel which forcibly initialized 327the block bitmap when the inode bitmap is first used, requiring an 328extra journal credit and disk write. (Addresses Google Bug: #5944440) 329 330Make sure rdebugfs (which may be installed setuid or setgid disk) does 331not honor environment variables if euid != uid or egid != gid. 332 333Debugfs's ncheck command has been optimized and now is much more 334robust with faced with corrupted file systems. The ncheck command 335also now has a -c option which will verify the file type information 336in the directory entry to see if matches the inode's mode bits. This 337is extremely useful when trying to use debugsfs to determine which 338parts of the file system metadata can be trusted. 339 340E2image will try to use ftruncate64() to set the i_size for raw 341images, instead of writing a single null byte. This avoid allocating 342an extra block to the raw image, for those file systems and/or 343operating systems that support this. (Linux does.) In addition, fix 344a logic bug that caused the file to not be properly extended if the 345size of the last hole was exactly an multiple of a megabyte. 346 347Fixed a bug in resize2fs where for 1k and 2k file systems, where 348s_first_data_block is non-zero, this wasn't taken into account when 349calculate the minimum file system size for use with the -M option. 350 351Fixed the badblocks program to honor the -s flag when in read-only -t 352mode. (Addresses Debian Bug #646629) 353 354Update Czech, Dutch, French, Polish, and Sweedish translation from the 355Translation Project. 356 357Fixed various Debian Packaging issues so that dpkg-buildflags is used 358if present, which allows e2fsprogs to be built with security hardening 359flags. (Addresses Debian Bugs: #654457) 360 361Programmer's Notes 362------------------ 363 364Fix a bug in ext2fs_clear_generic_bmap() when used for 32-bit bitmaps. 365This was only an issue for programs compiled against e2fsprogs 1.41 366that manipulate bitmaps directly. (Addresses Sourceforge Bugs: 367#3451486) 368 369The libext2fs library now uses sysconf() to fetch the page size, instead 370of the deprecated getpagesize(). 371 372The ext2fs_get_pathname() function will return a partial path if an a 373directory in the path is not a directory, displaying it as an inode 374number in angle brackets instead of giving up and displaying an error. 375This is much more helpful when a user is trying to debug a corrupted 376file system. 377 378Codepoints for the RO_COMPAT_REPLICA feature has been reserved. 379 380Added a new library function, ext2fs_file_get_inode_num(), for use by 381fuse2fs. 382 383Fixed a bug in ext2fs_file_set_size2() so that when it is truncating a 384file, it actually works. 385 386The block iterator now properly honors the BLOCK_ABORT flag for 387extent-based flags. Previously, it didn't, which generally made code 388be less efficient, but it could cause bugs in ext2fs_link(), for 389example, by causing it to insert multiple directory entries. 390 391Fixed an (harmless other than causing a compiler warning) use of an 392uninitialized variable in e2fsck's MMP code. 393 394 395E2fsprogs 1.42 (November 29, 2011) 396================================== 397 398This release of e2fsprogs has support for file systems > 16TB. Online 399resize requires kernel support which will hopefully be in Linux 400version 3.2. Offline support is not yet available for > 16TB file 401systems, but will be coming. 402 403This release of e2fsprogs has support for clustered allocation. This 404reduces the number of block (now cluster) bitmaps by allocating and 405deallocating space in contiguous power-of-2 collections of blocks, 406which are called clustered. This is a file system level feature, 407called 'bigalloc', which must be enabled when the file system is 408initially formatted. It is not backwards compatible with older 409kernels. 410 411Added support for the Multi-Mount Protection (MMP) feature. 412 413E2fsck more efficiently uses scratch files for really big file 414systems. (This is a feature that has to be turned on explicitly; see 415[scratch_files] in the e2fsck.conf man page.) 416 417Fix a bug in e2fsck where if the free blocks and inodes counts are 418incorrect, e2fsck would fix them without printing an error message. 419This would cause a "*** FILE SYSTEM WAS MODIFIED ***" message without 420any explanation of what was fixed. 421 422E2fsck will no longer attempt to clone an extended attribute block in 423pass1b handling if the file system does not support extended 424attributes. 425 426E2fsck will be more careful accidentally asking the user to continue 427if the file system is mounted, so that an escape sequence won't cause 428a false positive. (Addresses Debian Bug: #619859) 429 430E2fsck now uses less cpu time in pass 5 when large portions of the 431bitmaps are uninitialized. 432 433E2fsck will no longer segault when a corrupted file system has a bad 434extent, and removing it leads to a block needing to be deallocated. 435(Addresses SourceForge Bug: #2971800) 436 437E2fsck will catch termination signals (segfaults, bus errors, sigfpe) 438and print debugging information to make it easier to find potential 439problems. 440 441E2fsck will check to see if the bad block inode looks insane, and will 442skip trying to use if it certain fields which should be zero look 443non-zero. This is to avoid a corrupted bad block inode causing e2fsck 444to get confused and do more harm than good. 445 446If e2fsck modifies certain superblock fields which the kernel doesn't 447look at, it will now mark the superblock as dirty without marking the 448file system as changed. This avoids signalling the init scripts that 449a reboot is necessary, since the kernel isn't going to look at those 450fields, so it won't care if they have been changed. 451 452Fixed a bug in the libext2fs library (in the binary search routine of 453the icount abstraction) that could (very, very rarely) cause e2fsck to 454die in the middle of pass 1 or pass 2 processing. 455 456E2fsck will not try to do a discard operation if the -n option was 457specified on the command line. 458 459E2fsck now supports an extended "discard" option which will cause 460e2fsck to attempt discard all unused blocks after a full, successful 461file system check. 462 463E2fsck will check for the bad block inode to make sure it looks sane 464before trusting it, to avoid causing more harm than good to the file 465system. 466 467E2fsck now returns additional status bits in its exit code if it 468aborts early in the e2fsck run. 469 470E2fsck now correctly calculates the maximum file size in the case of 471the huge_file file system feature enabled without extents. 472 473The mke2fs and e2fsck programs now tries to use the punch hole command 474as a "discard" when operating on normal files. 475 476The e2image program now supports the qcow2 format, which is a more 477efficient way of capturing file system dumps. 478 479Mke2fs now supports the [devices] stanza in mke2fs.conf which allows 480per-device defaults to be specified in the configuration file. 481 482Mke2fs now supports the reserved_ratio relation in the [defaults] and 483[fs_types] section in mke2fs.conf. 484 485Mke2fs now creates extent-mapped directories for the root and 486lost+found directories. 487 488Mke2fs will skip zero'ing the journal if the extended option 489"lazy_journal_init" is specified. This can save a lot of time, but it 490does add a small amount of risk if the system crashes before the 491journal is overwritten entirely once. It is epsecially useful for 492testing. 493 494Mke2fs will now create file systems that enable user namespace 495extended attributes and with time- and mount count-based file 496system checks disabled. 497 498Mke2fs will not set a stride or strip size of one block based on block 499device attributes obtained from sysfs. 500 501Mke2fs now displays a progress report during the discard process. 502 503Mke2fs now handles extreme file system parameters correctly which 504previously caused the inodes per group to drop below 8, leading to a 505segfault. (The inodes per group must be a multiple of 8, but the code 506didn't correctly deal with an inodes per group count less than 8.) 507 508Mke2fs and tune2fs previously would give an error if the user tried 509setting the stride and stripe-width parameters to zero; but this is 510necessary to disable the stride and stripe-width settings. So allow 511setting these superblock fields to zero. (Addresses Google Bug: 512#4988557) 513 514Mke2fs now gives a warning if the auto-detected block size exceeds the 515system's page size. 516 517If the enable_periodic_fsck option is false in /etc/mke2fs.conf (which 518is the default), mke2fs will now set the s_max_mnt_count superblock 519field to -1, instead of 0. Kernels older then 3.0 will print a 520spurious message on each mount then they see a s_max_mnt_count set to 5210, which will annoy users. (Addresses Debian Bug: #632637) 522 523The default mke2fs.conf now has entries for "big" and "huge", which 524are needed for very big file systems. 525 526The resize2fs program now has support for a new online resize ioctl 527that can support file systems > 16TB, once it arrives in v3.x kernels. 528 529Fixed bug which caused resize2fs to fail when shrinking an empty file 530system down to its minimal size. (Addresses Sourceforge Bug #3404051) 531 532Fixed tune2fs's mount options parsing. (Addresses Debian Bug: #641667) 533 534Allow tune2fs to remove external journals if the device can not be 535found. 536 537Debugfs's icheck will now correctly find inodes which use the 538searched-for block as an extended attribute block. 539 540Debugfs now has a new "punch" command which remove blocks from the 541middle of an inode. 542 543Debugfs now has a new "e2freefrag" command which analyzes the free 544space fragmentation of the file system, using the same code as the 545e2freefrag program. 546 547Debugfs now has a "filefrag" command which displays information about 548a file's fragmentation. 549 550Add support to build a metadata-only, read-only, stripped-down version 551of debugfs called rdebugfs. 552 553Fixed a potential stack overrun bug in debugfs. 554 555The badblocks program now correctly recovers from I/O errors when 556direct I/O is being used. The badblocks command now also supports a 557-B option which forces the use of buffered I/O, and the -v option will 558provide a more detailed breakdown of read, write, and failed 559comparison errors. 560 561Added e4defrag tool which uses the EXT4_IOC_MOVE_EXT ioctl. 562 563Added support for journals larger than 2GB. 564 565Support using both hard links and symlinks when installing e2fsprogs. 566 567Add overflow checking to tune2fs -i's fsck interval, which must fit in 568a 32-bit field. 569 570The debugfs command now has a new 'blocks' command which prints out 571data blocks of a particular inode in a format which is useful for 572scripting. 573 574Filefrag will report 0 extents correctly in verbose mode. (Addresses 575RedHat Bugzilla: #653234) 576 577Filefrag has been fixed so its -v report prints the correct expected 578block number (previously there had been an off-by-one error). In 579addition, it will now display the number of contiguous extents when -v 580is not specified. This makes it consistent with the number of extents 581printed when the -v option was specified. In addition, the number of 582contiguous extents is far more interesting/useful than the number of 583physical extents for very large files. (Addresses Debian Bug: 584#631498, #644792) 585 586Logsave's usage message has been fixed. (Addresses Debian Bug: 587#619788) 588 589Avoid an infinite loop in ext2fs_find_block_device() if there are 590symlink loops in /dev caused by a buggy udev. 591 592Added a useful "fallocate" program to the contrib directory. 593 594Fixed an ABI compatibility problem in libext2fs which broke the dump 595program. Also added back some macros which dump needed so it could 596compile against the latest version of ext2_fs.h (Addresses Debian Bug: 597#636418) 598 599Fixed parsing of MNTOPT_ options for tune2fs and debugfs (Addresses 600Debian Bug: #641667) 601 602Added internationalization support for libcom_err error table strings. 603 604Fixed various spelling mistakes found in various output strings found 605by I18N translators. 606 607Update translations: French, Chinese, Germany, Indonesian, Swedish, 608Vietnamese, Polish, Dutch, Czech. (Addresses Debian Bugs: #520985, 609#620659) 610 611Fixed various Debian Packaging issues. (Addresses Debian Bugs: 612#614662, #632169, #641838, #627535, #629355) 613 614Updated/clarified man pages. (Addresses Debian Bugs: #639411, 615#642193, #634883) 616 617Programmer's Notes 618------------------ 619 620Initial support for quota as an integrated feature, where the quota 621files are hidden system files that are automatically maintained by 622e2fsck is present, although disabled by default. It must be enabled 623by using the configure option --enable-quota. 624 625Reserved the on-disk fields for the metadata checksum and snapshot 626features, which are currently in development. 627 628The ext2fs library now has the new functions ext2fs_punch(), 629ext2fs_get_memzero() and ext2fs_file_get_inode(). 630 631The ext2fs library now has support for calculating the crc32c checksum 632(via the new functions ext2fs_crc32c_be and ext2fs_crc32c_le). 633 634The I/O manager now supports the discard operation. 635 636Reserved file system code points for new 1st class quota feature. 637 638Shortened the compilation lines by moving the autoconf defines to 639a config.h header file. 640 641Fixed a potential free of an unitialized pointer in 642ext2fs_update_bb_inode(). 643 644Fixed miscellaneous gcc -Wall and coverity warnings. 645 646Fixed portability issues for Mac OS X, Hurd, and FreeBSD. 647 648Fixed a build failure when OMIT_COM_ERR is defined. 649 650Improved error checking and fixed memory leaks caused by error return 651paths. 652 653Add ext2fs_flush2() and ext2fs_close2() which takes a flag to allow 654the fsync() to be skipped. 655 656Added a test for extent-mapped journals by mke2fs and tune2fs. 657 658Added a test for creating a large (over 4GB) journal using mke2fs. 659 660Added a test to make sure the inode size remains constant. 661 662The regression test script now prints the list of which tests failed. 663 664The regression test system now uses /tmp for its scratch files, which 665is often a tmpfs mounted file system and hence much faster. 666 667The i_e2image test was fixed so it works with valgrind. 668 669Tests that rely on debugfs are now skipped if debugfs wasn't built. 670 671Fixed the dependencies for "make check" so all required dependencies 672are built before running the gression tests. 673 674A link to com_err.h is installed in $(includedir) during a "make install". 675 676The po/*.gmo files are automatically rebuilt if they are missing or 677out of date. This helps out Debian packaging. 678 679Allow ext2fs_get_memalign() to compile on systems that don't have 680posix_memalign(). 681 682Fixed a namespace leak in libext2fs (tdb_null). 683 684 685E2fsprogs 1.41.14 (December 22, 2010) 686===================================== 687 688Fix spurious complaint in mke2fs where it would complain if the file 689system type "default" is not defined in mke2fs.conf. 690 691The resize2fs program will no longer clear the resize_inode feature 692when the number reserved group descriptor blocks reaches zero. This 693allows for subsequent shrinks of the file system to work cleanly for 694flex_bg file systems. 695 696The resize2fs program now handles devices which are exactly 16T; 697previously it would give an error saying that the file system was too 698big. 699 700E2fsck (and the libext2fs library) will not use the extended rec_len 701encoding for file systems whose block size is less than 64k, for 702consistency with the kernel. 703 704Programming notes 705----------------- 706 707E2fsprogs 1.41.13 would not compile on big-endian systems. This has 708been fixed. (Addresses Sourceforge Bug: #3138115) 709 710The ext2fs_block_iterator2() function passed an incorrect ref_offset 711to its callback function in the case of sparse files. (Addresses 712Sourceforge Bug: #3081087) 713 714Fix some type-punning warnings generated by newer versions of gcc. 715 716 717E2fsprogs 1.41.13 (December 13, 2010) 718===================================== 719 720E2fsck now supports the extended option "-E journal_only", which 721causes it to only do a journal replay. This is useful for scripts 722that want to first replay the journal and then check to see if it 723contains errors. 724 725E2fsck will now support UUID= and LABEL= specifiers for the -j option 726(which specifies where to find the external journal). (Addresses 727Debian Bug #559315) 728 729E2fsck now has support for the problems/<problem code>/force_no 730configuration option in e2fsck.conf, which forces a problem to not be 731fixed. 732 733Dumpe2fs will now avoid printing large negative offsets for the bitmap 734blocks and instead print a message which is a bit more helpful for 735flex_bg file systems. 736 737Mke2fs will now check the file system type (specified with the -t 738option) to make sure it is defined in the mke2fs.conf file; if it is 739not, it will print an error and abort. If the usage type (specified 740with the -T option) is not defined in mke2fs.conf, mke2fs will print a 741warning but will continue. (Addresses Debian Bug #594609) 742 743Clarified error message from resize2fs clarifying that on-line 744shrinking is not supported at all. (Addresses Debian Bug #599786) 745 746Fix an e2fsck bug that could cause a PROGRAMMING BUG error to be 747displayed. (Addresses Debian Bug #555456) 748 749E2fsck will open the external journal in exclusive mode, to prevent 750the journal from getting truncated while it is in use due to a user 751accidentally trying to run e2fsck on a snapshotted file system volume. 752(Addresses Debian Bug #587531) 753 754Fix a bug in e2fsck so it has the correct test for the EOFBLOCKS_FL 755flag. 756 757The tune2fs program can now set the uninit_bg feature without 758requiring an fsck. 759 760The tune2fs, dumpe2fs, and debugfs programs now support the new ext4 761default mount options settings which were added in 2.6.35. 762 763The e2fsck and dumpe2fs programs now support the new ext4 superblock 764fields which track where and when the first and most recent file 765system errors occurred. These fields are displayed by dumpe2fs and 766cleared by e2fsck. These new superblock fields were added in 2.6.36. 767 768Debugfs now uses a more concicse format for listing extents in its 769stat command. This format also includes the interior extent tree 770blocks, which previously was missing from stat's output for 771extent-based files. 772 773Debugfs has a new option, -D, which will request Direct I/O access of 774the file system. 775 776Mke2fs will skip initializing the inode table if a device supports 777discard and the discard operation will result in zero'ed blocks. 778 779Badblocks will now correctly backspace over UTF-8 characters when 780displaying its progress bar. (Addresses Gentoo Bug #309909; Addresses 781Debian Bugs #583782 and #587834) 782 783E2freefrag will now display the total number of free extents. 784 785Resize2fs -P no longer requires a freshly checked filesystem before 786printing the minimum resize size. 787 788Fixed a floating point precision error in a binary tree search routine 789that can lead to seg fault in e2fsck and resize2fs. 790 791Fixed a bug in e2fsck where if both the original and backup superblock 792are invalid in some way, e2fsck will fail going back to the original 793superblock because it didn't close the backup superblock first, and 794the exclusive open prevented the file system from being reopened. 795 796Fixed a big in e2freefrag which caused getopt parsing to fail on 797architectures with unsigned chars. (Addresses Gentoo Bug: #299386) 798 799Clarified an mke2fs error message so a missed common in an -E option 800(i.e., mke2fs -t ext4 -E stride=128 stripe-width=512 /dev/sda1") 801results in a more understandable explanation to the user. 802 803Mke2fs now displays the correct valid inode ratio range when 804complaining about an invalid inode ratio specified by the user. 805 806Mke2fs now understands the extended option "discard" and "nodiscard", 807and the older option -K is deprecated. The default of whether 808discards are enabled by default can be controled by the mke2fs.conf 809file. 810 811Mke2fs's handling of logical and physical sector sizes has been 812improved to reflect the fact that there will be some SSD's with 8k and 81316k physical sectors arriving soon. Mke2fs will no longer force block 814size to be the physical sector size, since there will be devices where 815the physical sector size is larger than the system's page size, and 816hence larger than the maximal supported block size. In addition, if 817the minimal and optimal io size are not exported by the device, and 818the physical sector size is larger than the block size, the physical 819sector size will be used to set the Raid I/O optimization hints in the 820superblock. 821 822E2fsck will now display a better, more specific error message when the 823user enters a typo'ed device name, instead of blathering on about 824alternate superblocks. 825 826Fixed various Debian Packaging Issues 827 828Updated/clarified man pages (Addresses Debian Bugs: #580236, #594004, 829#589345, #591083; Addresses Launchpad Bug: #505719) 830 831Update the Chinese, Chzech, Dutch, French, Germany, Indonesian, 832Polish, Swedish, and Vietnamese translations. 833 834 835Programmer's Notes 836------------------ 837 838Fix a dependency definition for the static and profiled blkid 839libraries which could cause compile failures in some configurations. 840(Addresses Debian Bug: #604629) 841 842Add support for Direct I/O in the Unix I/O access layer. 843 844Fixed a memory leak in the Unix I/O layer when changing block sizes. 845 846Fixed minor memory leaks in mke2fs. 847 848Added a new function to the ext2fs library, ext2fs_get_memalign(). 849 850The tst_super_size test program will check to make sure the superblock 851fields are correctly aligned and will print them out so they can be 852manually checked for correctness. 853 854Fixed some makefile dependencies for test programs in lib/ext2fs. 855 856Reserved the feature flags and superblock fields needed for the Next3 857snapshot feature. 858 859Reserved the feature flags for EXT4_FEATURE_INCOMPAT_DIRDATA and 860EXT4_INCOMPAT_EA_INODE. 861 862 863E2fsprogs 1.41.12 (May 17, 2010) 864================================ 865 866Mke2fs now gives a correct error message if the external journal 867device is not found. (Addresses Red Hat Bug #572935) 868 869Resize2fs -P will now refuse to print a minimum size if the file 870system is not clean. Previously it would go ahead and print a minimum 871size anyway, which might not be correct, leading to user confusion. 872 873E2fsck now tests for extents that begin at physical block 0 and 874rejects them as invalid. (Addresses Google Bug: #2573806) 875 876Fixed a bug in e2fsck which could cause it to crash when trying to 877remove an invalid extent and the block bitmaps hadn't yet been loaded. 878(Addresses SourceForge Bug: #2971800) 879 880E2fsck now will completely skip time-based checks if the system clock 881looks insane or the option broken_system_clock is set in 882/etc/e2fsck.conf. 883 884E2fsck would previously report an i_blocks corruption for a 4T file 885created using posix_fallocate; this bug has been fixed. 886 887E2fsck will now correctly mark a sparse journal as invalid and will 888delete and recreate the journal to address the problem. 889 890E2fsck would previously incorrectly ask the user whether she would 891like to abort the file system check after finding a problem --- and 892then abort regardless of the user's answer. This is annoying, and 893has been fixed. 894 895E2fsck can now continue even if it fails to recreate the resize 896inode; previously it would just abort the file system check 897altogether. 898 899E2fsck could potentially remove directory entries for inodes found in 900the unused region of the inode table; this would generally happen on 901ext4 file systems that do not use journaling. This bug has been fixed 902by not clearing these directory entries once it has been established 903that bg_unused_inodes may not be trustworthy; once pass #2 has been 904completed, e2fsck will restart the file system check from the 905beginning, and then it will be safe to delete any directory entries 906pointing to inodes that appear to be deleted. (Addresses Google Bug: 907#2642165) 908 909E2fsck will not try to set the block group checksums if the user 910cancels the fsck with a control-C. It's a bad idea to set the 911checksums if e2fsck hasn't been completed, and it often results an the 912error message, "Inode bitmap not loaded while setting block group 913checksum info". (Addresses Launchpad Bug: #582035) 914 915The mke2fs program now queries the kernel for the physical as well as 916the logical sector size, and will not allow a blocksize below the 917logical, and will strongly encourage a blocksize at least as big as 918the physical blocksize. This is needed for 4k sector drives that 919emulate 512 byte sector sizes. 920 921Mke2fs will now allow a flex_bg size of 1. This is unusual, and 922rarely needed, but it is a legal value. 923 924E2fsck will check for cases where the EOFBLOCKS_FL is set when it is 925not needed, and offer to clear it; this is a sign of a kernel bug, but 926more importantly, some released kernels may crash when this situation 927is encountered on ext4 file systems. (Addresses Google Bug: #2604224) 928 929E2fsck will use the EOFBLOCKS_FL flag exclusively to check whether 930i_size is correct. (Kernels starting with 2.6.34 will set 931EOFBLOCKS_FL.) 932 933The com_err library will now only output ^M (a CR character) when the 934tty is in raw mode. 935 936Update the Czech, Chinese, Dutch, French, Germany, Indonesian, Polish, 937and Vietnamese translations. 938 939Fixed various Debian packaging issues --- see debian/changelog for 940details. (Addresses Debian Bugs: #571247, #563487) 941 942 943Programmer's Notes 944------------------ 945 946The regression test suite now uses its own mke2fs.conf file, so that 947downstream distributions want change the mke2fs.conf file which is 948distributed in the RPM or dpkg file, without worrying about screwing 949up the regression test results. 950 951Always build namei.o so that building with configure --disable-debugfs 952works correctly. Long-term, if we care about reduced e2fsprogs 953builds, we need a more general solution for deciding what .o files are 954needed for a particular build. Given that install floppies are going 955(gone?) the way the dodo bird, we probably don't care, though. 956(Addresses Sourceforge Bug: #2911433) 957 958Add configure options --enable-symlink-build and 959--enable-symlink-install, which allow e2fsprogs be built using 960symlinks instad of hard links, and to be installed using symlinks 961instead of hard links, respectively. It is useful when the file 962system where the build is taking place, or the file system where 963e2fsprogs is installed, can't handle hard links for some reason. 964(Addresses Sourceforge Bug: #1436294) 965 966Fixed compile warning in mke2fs.c. 967 968 969E2fsprogs 1.41.11 (March 14, 2010) 970================================== 971 972E2fsck will no longer give a fatal error and abort if the physical 973device has been resized beyond 2**32 blocks. (Addresses Launchpad 974Bug: #521648) 975 976Debugfs has a bug fixed so that "logdump -b <blk>" now properly shows 977the allocation status of the block <blk>. (Addresses Debian Bug: 978#564084) 979 980E2fsck now prints a much more emphatic and hopefully scary message 981when a file system is detected as mounted while doing a read/write 982check of the filesystem. Hopefully this will dissuade users from 983thinking, "surely that message doesn't apply to *me*" :-( 984 985E2fsck -n will now always open the file system read-only. We now 986disallow certain combination of options which previously were manual 987exceptions; this is bad because it causes users to think they are 988smarter than they really are. So "-n -c", "-n -l", "-n -L", and "-n 989-D" are no longer supported. (Addresses Launchpad Bug: #537483) 990 991In e2fsprogs 1.41.10, mke2fs would ask for confirmation to proceed if 992it detected a badly aligned partition. Unfortunately, this broke some 993distribution installation scripts, so it now just prints the warning 994message and proceeds. (Addresses Red Hat Bug: #569021. Addresses 995Launchpad Bug: #530071) 996 997Mke2fs would take a long time to create very large journal files for 998ext4. This was caused by a bug in ext2fs_block_iterate2(), which is 999now fixed. 1000 1001E2fsck now understands the EOFBLOCKS_FL flag which will be used in 10022.6.34 kernels to make e2fsck not complain about blocks deliberately 1003fallocated() beyond an inode's i_size. 1004 1005E2fsprogs 1.41.10 introduced a regression (in commit b71e018) where 1006e2fsck -fD can corrupt non-indexed directories when are exists one or 1007more file names which alphabetically sort before ".". This can happen 1008with ext2 filesystems or for small directories (take less than a lock) 1009which contain filenames that begin with a space or some other 1010punctuation mark. (Addresses Debian Bug: #573923, Addresses Launchpad 1011Bug: #525114) 1012 1013 1014Programmer's Notes 1015------------------ 1016 1017Add new test, f_rehash_dir, which checks to make sure e2fsck -D works 1018correctly. 1019 1020The libcom_err function now has support for Heimdal's com_right_r 1021function(). (Addresses Sourceforge Bug: #2963865, Addresses Debian 1022Bug: #558910) 1023 1024 1025E2fsprogs 1.41.10 (February 7, 2010) 1026==================================== 1027 1028Fix resize2fs bug which causes it to access invalid memory. 1029 1030Add libss support for libreadline.so.6. 1031 1032Fix e2fsck's check for extent-mapped directory with an incorrect file 1033type. 1034 1035Add new e2fsck.conf configuration option, default/broken_system_clock 1036to support systems with broken CMOS hardware clocks. Also, since too 1037many distributions seem to have broken virtualization scripts now, 1038e2fsck will by default accept dates which are off by up to 24 hours by 1039default. (Addresses Debian Bugs: #559776, #557636) 1040 1041Fix a bug where mke2fs may not use the best placement of the inode 1042table when there is only room for a single block group in the last 1043flex_bg. 1044 1045E2fsck is now smarter when it needs to allocate blocks in the course 1046of fixing file system problems. This reduces the number of spurious 1047differences found in pass #5. 1048 1049E2fsck will no longer rehash directories which fit in a single 1050directory block. 1051 1052E2fsck now correctly handles holes in extent-mapped directories (i.e., 1053sparse directories which use extents). 1054 1055Fix big-endian problems with ext2fs_bmap() and ext2fs_bmap2(). 1056 1057Fix a bug in filefrag where on platforms which can allow file systems 1058with 8k blocks, that it doesn't core dump when it sees a file system 1059with 8k block sizes. (Thanks to Mikulas Patocka for pointing this 1060out.) 1061 1062E2fsck will correctly fix directories that are have an inaccurate 1063i_size as well as other problems in a single pass, instead of 1064requiring two e2fsck runs before the file system is fully fixed. 1065 1066Fix e2fsck so it will correctly find and detect duplicate directory 1067entries for non-indexed directories. ( Addresses Sourceforge Bug: 1068#2862551) 1069 1070Mke2fs will use BLKDISCARD to pre-discard all blocks on an SSD or 1071thinly-provisioned storage device. This can be disabled using the -K 1072option. 1073 1074Enhance libext2fs so it works around bug in Linux version 2.6.19 1075and earlier where the /proc/swaps file was missing the header on 1076the first line. 1077 1078Fix bug in Linux version 2.6.19 and earlier where the /proc/swaps file 1079was missing the header on the first line. 1080 1081Fix some big-endian bugs in e2fsck and libext2fs. 1082 1083Fix resize2fs so it works correctly on file systems with external 1084journals instead of failing early with the error "Illegal inode 1085number". 1086 1087Fix libss so that it does not seg fault when using a readline library 1088which does not supply a readline_shutdown() function. 1089 1090Updated dumpe2fs's usage message so it correctly gives the right 1091arguments summary for "-o superblock=<num>" and "-o blocksize=<num>". 1092(Addresses Launchpad Bug: #448099) 1093 1094Teach libext2fs to ignore the high 32 bits of the i_blocks field 1095when huge_file file system feature is set, but the inode does not 1096have the HUGE_FILE_FL flag set. 1097 1098Fix e2fsck's handling of 64-bit i_blocks fields. 1099 1100E2fsck will now print "Illegal indirect block", "Illegal 1101double-indirect block", etc., instead of "Illegal block #-1" or 1102"Illegal block #-2", etc. This makes it easier for users to 1103understand what has gone wrong. (Addresses SourceForge Bug: #2871782) 1104 1105Mke2fs now will obtain get device topology information from blkid and 1106use it to populate the superblock stride and stripe sizes. It will 1107also warn if the block device is misaligned 1108 1109Fix file descriptor leaks in fsck and debugfs. (Addresses Novell Bug: 1110##524526) 1111 1112Fix the libext2fs library code to round up the bitmap size to a 4-byte 1113boundary, to prevent spurious seg faults caused by the x86 1114architecture. This doesn't affect Linux systems, but was a major 1115problem on a number of *BSD systems. (Addresses Sourceforge Bug: 1116#2328708) 1117 1118Fix resize2fs's minimum size required for a file system so it doesn't 1119fail when "resize2fs -M" is run. (Addresses RedHat Bugzilla: #519131) 1120 1121Dumpe2fs now prints summary information about the contents of the 1122journal. 1123 1124Avoid printing scary error messages when e2fsck starts running 1125problems on low-memory systems, as it tends to panic and mislead the 1126user. (Addresses Debian Bug: #509529) 1127 1128Fix blkid's modules.dep parser so it handles compressed (.ko.gz) 1129modules files. (Address Red Hat Bug: #518572) 1130 1131Fix tune2fs so it can add a journal when an extent-enabled file system 1132is mounted. (Addresses Launchpad bug: #416648) 1133 1134Update Czech, Indonesian, Polish and Vietnamese translations (from the 1135Translation Project). 1136 1137Update/clarify man pages. (Addresses Sourceforge Bug: #2822186) 1138 1139Fixed various Debian packaging issues --- see debian/changelog for 1140details. (Addresses Debian Bugs: #540111) 1141 1142Programmer's Notes 1143------------------ 1144 1145The configure script supports the --with-cc, --with-ccopts, and 1146--with-ldopts options. Instead, the more standard use of CC=, 1147CCFLAGS=, and LDFLAGS= in the configure command line is used instead. 1148Also, --with-ld, which never worked, was also removed. 1149(Addresses Sourceforge Bug: #2843248) 1150 1151The in-tree header files are only used if the in-tree uuid or blkid 1152libraries are used. Otherwise, use the system-provided uuid or blkid 1153header files if using the system-provided libraries. 1154 1155Fix some build failures caused by --disable-* configure options. 1156 1157Work around a bug in autoconf 2.64. 1158 1159 1160E2fsprogs 1.41.9 (August 22, 2009) 1161================================== 1162 1163Fix a bug in e2fsck routines for reallocating an inode table which 1164could cause it to loop forever on an ext4 filesystem with the FLEX_BG 1165filesystem feature with a relatively rare (and specific) filesystem 1166corruption. This fix causes e2fsck to try to find space for a new 1167portion of the inode table in the containing flex_bg, and if that 1168fails, the new portion of the inode table will be allocated in any 1169free space available in the filesystem. 1170 1171Make e2fsck less annoying by only asking for permission to relocate a 1172block group's inode table once, instead of for every overlapping 1173block. Similarly, only ask once to recompute the block group 1174checksums, instead of once for each corrupted block group's checksum. 1175 1176Fix filefrag to avoid print the extent header if the FIEMAP ioctl is 1177not present, and it needs to fall back to using the FIBMAP ioctl. 1178 1179Fix filefrag to correctly print the number of extents for zero-length 1180files. (Addresses Debian Bug: #540376) 1181 1182Filefrag now has a -B option which forces the use of the FIBMAP ioctl 1183to more easily debug the FIBMAP code. 1184 1185Fixed filefrag for non-extent based files. 1186 1187Add a new program, e2freefrag, which displays information about the 1188free space fragmentation in an ext2/3/4 filesystem. 1189 1190Fix inode resizing via tune2fs -I so that it works correctly in the 1191face of non-empty bad blocks inodes, and if the filesystem was 1192formatted using the "mke2fs -E stride=N" option for RAID arrays. 1193 1194Fix regression in ext2fs_extent_set_bmap() caused e2fsck -fD to fail 1195and corrupt large directories if the directory needs to shrink by more 1196than one block. (Addresses Debian Bug: #537510) 1197 1198Fix e2fsck's buggy_init_scritps=1 so that the if the last write and/or 1199last mount times are in the future, they are corrected even if 1200buggy_init_scripts is set. This is needed because otherwise resize2fs 1201will refuse to resize the filesystem, even after running "e2fsck -f". 1202(Addresses Launchpad bug: #373409) 1203 1204E2fsck will now print much fuller information when the last mount time 1205or last written time is in the future, since most people can't seem to 1206believe their distribution has buggy init scripts, or they have a 1207failed CMOS/RTS clock battery. 1208 1209Enhance dumpe2fs to dump the extent information via the 'stat' 1210command, and more detailed extent information via the new command 1211'dump_extents'. 1212 1213Update French, Polish, Czech, and Sweedish translation from the 1214Translation Project. 1215 1216Fixed various Debian packaging issues --- see debian/changelog for 1217details. 1218 1219Programmer's Notes 1220------------------ 1221 1222Fixed miscellaneous gcc -Wall warnings. 1223 1224Fixed memory leak in error path in ext2fs_block_iterate2() 1225 1226Fixed non-Linux build of the intl directory by adding support for the 1227E/Q/V macros. 1228 1229The bitmap read/write functions now treat uninitialized bitmaps as 1230unallocated; this fixes a number of problems in all e2fsprogs for ext4 1231filesystems when there is a need to allocate new blocks or inodes, and 1232there aren't any free blocks or inodes in the already-used block 1233groups. 1234 1235Improve ext2fs_extent_set_bmap() to avoid creating new extents which 1236get inserted into the extent tree when they are not needed. 1237 1238 1239E2fsprogs 1.41.8 (July 11, 2009) 1240================================ 1241 1242Fix resize2fs's online resizing, fixing a regression which in 1243e2fpsrogs 1.41.7. (Addresses Debian Bug: #535452) 1244 1245Fix potential filesystem corruptions caused by using resize2fs to 1246shrinking ext4 filesystems with extents enabled. (Addresses Red Hat 1247Bug: #510379) 1248 1249Optimize uuid_generate() to avoid running uuidd if it is not setuid or 1250setgid and the currently running program doesn't have write access to 1251the uuidd work directory. 1252 1253Add safety checks (for non-Linux systems) so that uuidd isn't run with 1254file descriptors 0, 1, and 2 closed; and if they are closed, uuidd 1255will be careful not to close the file descriptor for its unix domain 1256socket when it detaches itself from the controlling tty. Also add 1257safety checks so that if the unix domain socket between the uuid 1258library and uuidd program is closed for any reason, both the library 1259and the uuidd will return an appropriate error code instead of looping 1260in an infinite loop. 1261 1262The e2croncheck script, which creates an LVM snapshot and then checks 1263the ext3/4 filesystem via the LVM snapshot, has been added to the 1264contrib directory. 1265 1266Fix filefrag program for files that have more than 144 extents. 1267 1268Update French, Polish, Czech, Indonsian, and Sweedish translation from 1269the Translation Project. 1270 1271Fixed various Debian packaging issues --- see debian/changelog for 1272details. (Addresses Debian Bug #535530) 1273 1274Update/clarify man pages. 1275 1276Programmer's Notes 1277------------------ 1278 1279Fix compilation problem when configured with --disable-uuid. 1280 1281Don't build uuidgen if configured with --disable-uuid. 1282 1283Add the new library function ext2fs_test_inode_bitmap_range(), and 1284optimized ext2fs_test_block_bitmap_range(), which will be needed for 1285future optimizations for e2fsck. 1286 1287Fix makefile dependencies for libcom_err so that the Makefiles work 1288well on non-GNU make program. (Addresses Sourceforge Patches: #2813809) 1289 1290Enhance the build system so that the full set of commands executed by 1291the Makefiles are displayed, instead of the Linux kernel summary 1292output, if the build was configured with --enable-verbose-makecmds, or 1293if GNU make is in use and the V variable is non-null, i.e., via "make 1294V=1". 1295 1296 1297E2fsprogs 1.41.7 (June 29, 2009) 1298================================ 1299 1300Fix a bug in libext2fs which can cause e2fsck and resize2fs to write 1301uninitalized data into the portion of the inode beyond the first 128 1302bytes when operating on inodes mapped via extents; potentially 1303corrupting filesystems. 1304 1305Fix memory leaks in e2fsprogs, including a very large memory leak 1306which can cause e2fsck to run out of memory when checking very large 1307filesystems using extents. 1308 1309The logsave program will now filter out the ^A and ^B characters when 1310writing to the console. 1311 1312Harden ext2fs_validate_entry() so that lsdel will not read beyond the 1313end of the block even if the directory block is corrupted. 1314 1315Fix debugfs from core dumping if the logdump command fails to open the 1316output file. 1317 1318Enhance badblocks to print the currently tested block number when 1319interrupted with ^C. 1320 1321Fix lsattr to exit with a non-zero status when it encounters errors. 1322(Addresses RedHat Bugzilla #489841) 1323 1324Fix e2fsprogs to use the same encoding for rec_len in 64k (and larger) 1325filesystems as the kernel when encoding rec_len >= 64k. For 64k 1326filesystems (currently all that is supported by e2fsprogs, this is 1327only a minor corner case). 1328 1329Resize2fs will now update the journal backup fileds in the sup0erblock 1330if the journal is moved; this avoids an unnecessary full fsck after 1331resizing the filesystem. (Addresses RedHat Bugzilla: #505339) 1332 1333Fix libext2fs to properly initialize i_extra_size when creating the 1334journal and resize inodes. 1335 1336Change badblocks to allow block sizes larger than 4k. 1337 1338Fix filefrag progam so it correctly checks for errors from the fiemap ioctl. 1339 1340Update Chinese and Czech translation from the Translation Project. 1341 1342Clean up various man pages. (Addresses Debian Bug #531385 and #523063) 1343 1344 1345Programmer's Notes 1346------------------ 1347 1348Add --disable-libuuid option to configure which uses an in-system 1349installed version of libuuid; the private version is enabled by 1350default. 1351 1352Add --valgrind-leakcheck option to the test_script program which runs 1353valgrind with the appropriate options to find memory leaks. 1354 1355 1356E2fsprogs 1.41.6 (May 30, 2009) 1357=============================== 1358 1359Fix a critical regression in e2fsck introduced version 1.41.5 which 1360can undo updates to the block group descriptors after a journal replay. 1361 1362If e2fsck receives an I/O error while replaying the journal, prompt 1363the user whether they want to ignore the error or not; if e2fsck is 1364run in preen mode, force fsck to abort the preen pass. 1365 1366Fix a bug which would cause e2fsck to core dump if the filesystem 1367contains a corrupt extent header, and the user declines to clear the 1368inode. (Addresses Sourceforge Bug: #2791794) 1369 1370Fix e2fsck to restart only once in the case of multiple block groups 1371which inodes in use that are in the uninitialized part of the block 1372group's inode table. 1373 1374To reduce user confusion, if the /etc/mtab file is missing 1375ext2fs_check_mount_point and ext2fs_check_if_mounted will return a 1376new, explicit error code to indicate this case. This will cause 1377e2fsck to give a clearer error message when the user is using buggy 1378rescue CD's that don't properly set up /etc/mtab. (Addresses Debian 1379Bug: #527859) 1380 1381Fix e2fsck so that if the primary group descriptors are corrupted, but 1382the backup superblock is entirely invalid, to go back to using (and 1383fixing) the priary group descriptors instead of completely giving up 1384on the filesystem. (Addresses Debian Bug: #516820) 1385 1386Change e2fsck to not abort a preen pass if an inode is found to have 1387i_file_acl_hi non-zero. Ext3 filesystems don't care, and newer 1388kernels (post 2.6.29) will ignore this field. So let's fix it 1389automatically during the preen pass instead of forcing the user to fix 1390this up manually. (Addresses Debian Bug: #526524) 1391 1392Add resource tracking for e2fsck passes 1b through 1d. 1393 1394Speed up e2fsck by eliminating unnecessary journal checks if the 1395filesystem is already mounted and doesn't need recovery (since the 1396kernel would have run the journal when the filesystem was mounted.) 1397Also speed up e2fsck by avoiding unnecessary block group descriptor 1398updates in ext2fs_close(). 1399 1400Add support to chattr to migrate inodes from using direct/indirect 1401blocks to extents. 1402 1403Avoid corrupting the filesystem if there is an attempt to shrink a 1404filesystem using resize2fs smaller than posible by making 1405ext2fs_set_bmap() more careful not to delete the old block until the 1406new block can be inserted. In addition, fix a bug in how the minimum 1407size of the filesystem (plus a safety margin) is calculated, and 1408modify resize2fs to refuse to shrink the filesystem below that minimum 1409size without the force flag. 1410 1411Teach blkid to try to figure out DM device names using take advantage 1412of information in sysfs when running on kernels (2.6.29 and later) 1413which provide this information in /sys/block/dm-<N>/dm/name; this is 1414much faster than scanning for the devncie number in /dev/mapper. 1415 1416Fix blkid to prefer /dev/mapper/<name> device names instad of the 1417private /dev/dm-X names. 1418 1419Add an -a option to debugfs's close_filesys command which writes any 1420changes to the superblock or block group descriptors to all of the 1421backup superblock locations. 1422 1423Add support to the filefrag program to use the FIEMAP ioctl. 1424 1425Update Chinese translation from the Translation Project. 1426 1427Clean up various man pages. (Addresses Red Hat Bugzilla: #502971 and 1428Launchpad Bug: #381854) 1429 1430Fixed various Debian packaging issues --- see debian/changelog for 1431details. (Addresses Debian Bug #506064) 1432 1433 1434Programmer's Notes 1435------------------ 1436 1437Add test code to make sure e2fsck's problem.c doesn't have two problem 1438codes assigned to duplicate values. 1439 1440Avoid using a hard-coded path for /bin/rm in block's test_probe.in. 1441(Addresses Sourceforge Bug: #2780205) 1442 1443Clean up e2fsck by removing #ifdef RESOURCE_TRACK by adding an empty 1444function declaration for init_resource_track() and 1445print_resource_track() when RESOURCE_TRACK is not defined. 1446 1447The test code which is used to build the tst_csum progam has been 1448moved from from tst_csum.c into csum.c under an #ifdef DEBUG to to 1449avoid compile problems caused by not having a prototype for 1450ext2fs_group_desc_csum(). (Addresses Sourceforge Bug #2484331) 1451 1452Update the config.guess and config.sub file to the latest from the 1453FSF, to allow e2fsprogs to build on the avr32 platform. (Addresses 1454Debian Bug: #528330) 1455 1456Add a new function, ext2fs_extent_open2(), which behaves like 1457pext2fs_extent_open(), but uses the user-supplied inode structure 1458when opening an extent instead of reading the inode from disk. 1459 1460 1461E2fsprogs 1.41.5 (April 23, 2009) 1462================================= 1463 1464Fix a number of filesystem corruption bugs in resize2fs when growing 1465or shrinking ext4 filesystems off-line (i.e., when the ext4 filesystem 1466is not mounted). 1467 1468Debugfs can now set i_file_acl_high via the set_inodes_field command, 1469and print a 64-bit file acl. This is useful for debugging filesystem 1470corruptions where the high bits of i_file_acl_high are set. E2fsck 1471will detect and fix non-zero i_file_acl_high on 32-bit filesystems 1472since some Linux kernel versions pay attention to this field even when 1473they shouldn't. 1474 1475Speed up e2fsck when checking clean filesystems by avoiding 1476unnecessary block reads, and coalescing the block group descriptor 1477blocks so they are read using a single read operation. 1478 1479The libuuid library will now close all file descriptors before running 1480uuidd. This avoids problems when the calling program has open sockets 1481which then never get closed because uuidd is a long-running helper 1482daeomn. (Addresses Launchpad bug: #305057) 1483 1484In order to avoid unnecessary full filesystem checks by e2fsck after 1485an on-line resize, e2fsck will ignore the NEEDS_RECOVERY flag set on 1486the backup superblocks. (Addresses Red Hat Bugzilla: #471925) 1487 1488Mke2fs will avoid trying to create the journal when run in 1489superblock-only mode (mke2fs -S), since the left-over journal in the 1490inode table will cause mke2fs to fail. 1491 1492Fix a bug in libext2fs functions that check to see if a particular 1493device or filesystem image is mounted, which would cause these 1494functions to report that a file identical to the (relative) pathname 1495used by a pseudo-filesystem was mounted when in fact it was not. 1496 1497Update Czech translation from the Translation Project. 1498 1499Add Chinese (simplified) translation from the Translation Project. 1500 1501Fix support for external journals (which was broken in e2fsprogs 15021.41.4). 1503 1504Fix a regression in debugfs where the "stat" command when no 1505filesystem was open would cause debugfs to crash with a segmentation 1506violation. 1507 1508Starting in the 2.6.29 linux kernel, the ext4 filesystem driver can be 1509used to support filesystems without a journal. Update the blkid 1510library so it understands this. 1511 1512The blkid library will remove an entry from the blkid cache 1513representing the entire disk if partitions are found, since presuambly 1514the device previously had no partition table, but has now transitioned 1515to using a partition table. 1516 1517Add a check to mke2fs and tune2fs that the argument to the -m option 1518(which specifies the reserved ratio) must be greater than zero. 1519(Addresses Debian Bug: #517015) 1520 1521Add support for tracking the number kilobytes written to the 1522filesystem via the superblock field s_kbytes_written. It will be 1523updated by the kernel as well as by e2fsprogs programs which write to 1524the filesystem. This is useful for tracking the wear to filesystems 1525on Solid Sstate Drives. 1526 1527Fix compatibility issue in the libext2fs info file and makeinfo 1528version 4.12. (Addresses Red Hat Bugzilla: #481620) 1529 1530Update/clarify man pages. (Addresses Debian Bug: #515693, #365619) 1531 1532Fixed various Debian packaging issues --- see debian/changelog for 1533details. (Addresses Debian Bug: #506279) 1534 1535Programmer's Notes 1536------------------ 1537 1538Fix Hurd compilation problem in e2fsck and tune2fs (Addresses Debian 1539Bug: #521602) 1540 1541Fix various gcc compilation warnings and other programming cleanups. 1542(Addresses Red Hat Bugzilla: #486997) 1543 1544Add support for building the blkid command statically. 1545 1546Add support for disabling the built-in blkid library in favor of a 1547system-installed blkid implementation via the configure option 1548--disable-libblkid. 1549 1550 1551E2fsprogs 1.41.4 (January 27, 2009) 1552=================================== 1553 1554Fixed a bug which could sometimes cause blkid to return an exit value 1555of zero for a non-existent device (Addresses Debian Bug: #502541) 1556 1557Blkid will now recognize ext3 filesystems that have the test_fs flag 1558set as ext3 filesystems. 1559 1560The blkid library will now recognize btrfs filesystems and swap 1561devices currently used by user-level software suspend. 1562 1563Tune2fs now updates the block group checksums when changing the UUID 1564to avoid causing e2fsck to complain vociferously at the next reboot. 1565 1566Tune2fs's inode size resizing algorithms have been fixed so it is not 1567vastly inefficient for moderate-to-large filesystems, due to some 1568O(n**2) and O(n*m) algorithms that didn't scale well at all. 1569 1570Fix tune2fs's inode resizing algorithm so it will not corrupt 1571filesystems laid out for RAID filesystems; in addition, tune2fs will 1572refuse to change the inode size for filesystems that have the flex_bg 1573feature enabled. (This is a limitation in the current implementation 1574of tune2fs -I.) 1575 1576E2fsprogs 1.41 broke debugfs's logdump command for normal ext3/4 1577filesystems with 32-bit block numbers, when the headers for 64-bit 1578block numbers was added. This regression has been fixed. 1579 1580Debugfs's ncheck command has been fixed to avoid printing garbage 1581characters at the end of file names. 1582 1583Fix resize2fs for ext4 filesystems. Some blocks that that need moving 1584when shrinking filesystems with uninit_bg feature would not be moved. 1585In addition, blocks and inode table blocks were not being correctly 1586freed when shrinking filesystems with the flex_bg feable, which caused 1587resize2fs -M to fail. Finally, when blocks are moved, make sure the 1588uninitialized flag in extents is preserved. 1589 1590Fix bug which caused dumpe2fs to abort with an error if run on a 1591filesystem that contained an external journal. 1592 1593Some distributions used "mke3fs" as an alias for "mkfs.ext3"; check 1594for this in argv[0] to provide better legacy support for these 1595distributions. This is a practice that should NOT be continued, 1596however. 1597 1598Mke2fs now has a new option -U, which allows the user to specify the 1599UUID that should be used for the new filesystem. 1600 1601Mke2fs will treat devices that are exactly 16TB as if they were 16TB 1602minus one block. This allows users who have read that ext3 supports 1603up to 16TB filesystems and who create a 16TB LVM to not get confused, 1604since the true limit is really 16TB minus one block. 1605 1606E2fsck will no longer abort an fsck run if block group has an errant 1607INODE_UNINIT flag. 1608 1609E2fsck now distinguishes between fragmented directories and fragmented 1610files in verbose mode statistics and in the fragcheck report. 1611 1612Fix a bug in e2fsck which casued it double count non-contiguous 1613extent-based inodes. 1614 1615E2fsck will leave some slack space when repacking directories to allow 1616room for a few directory entries to be added without causing leaf 1617nodes to be split right away. 1618 1619Fix a bug which caused e2fsck to crash when it comes across a 1620corrupted interior node in an extent tree with the error message: 1621"Error1: Corrupt extent header on inode XXXXXX" 1622 1623E2fsck problem descriptions involving the journal are no longer 1624referred to as "ext3" problems, since ext4 filesystems also have 1625journals. 1626 1627Fix a long-standing bug in e2fsck which would cause it to crash when 1628replying journals for filesystems with block sizes greater than 8k. 1629 1630Update Catalan translation from the Translation Project. 1631 1632Fixed various Debian packaging issues --- see debian/changelog for 1633details. (Addresses Debian Bugs: #503057, #502323, #511207) 1634 1635Programmer's Notes 1636------------------ 1637 1638Fix build of e2fsck.profiled, and add support for building profiled 1639binaries in the misc directory if configured with --enable-profile. 1640 1641The ext2fs_open() function now performs more sanity checks on the 1642superblock to avoid potential divide by zero errors by other parts of 1643library. 1644 1645The ext2fs_read_inode_full() function now has a safety check to avoid 1646a segmentation fault on corrupted filesystems. 1647 1648The ext2fs_new_inode() function now has a sanity check so that if the 1649s_first_inode field in the superblock is insane, it will return 1650EXT2_ET_INODE_ALLOC_FAIL instead of returning an invalid inode number. 1651 1652To avoid segmentation faults, ext2fs_block_alloc_stats() and 1653ext2fs_inode_alloc_stats() now validates the passed inode or block 1654number to avoid overrunning an array boundary. 1655 1656Various signed/unsigned errors for variables containing block numbers 1657have been fixed. 1658 1659Accomodations for gcc's stupidity in not realizing that constant 1660strings that do not contain a '%' character are safe to use in format 1661strings have been made so that distributions that want to compile 1662e2fsprogs with -Werror=format-security have an easier time doing so. 1663 1664Added a new 64-bit getsize interface, ext2fs_get_device_size2(). 1665 1666Added the utility make-sparse.c to the contrib directory. 1667 1668The ext2fs_block_iterate2() function now reflects errors from 1669ext2fs_extent_set_bmap() to the caller, if the callback tries to 1670change a block on an extent-based file, and ext2fs_extent_set_bmap() 1671fails for some reason (for example, there isn't enough disk space to 1672split a node and expand the extent tree. 1673 1674The ext2fs_block_iterate2() function will preserve the uninit flag in 1675extents when the callback function modifies a block in an extent-based 1676file. 1677 1678E2fsck will now flag filesystems that have an insane s_first_ino field 1679in their superblock, and attempt to use a backup superblock to repair 1680the filesystem. 1681 1682 1683E2fsprogs 1.41.3 (October 12, 2008) 1684=================================== 1685 1686E2fsck has been fixed so it prints the correct inode number for 1687uinit_bg related problems. 1688 1689E2fsck will now offer to clear the test_fs flag if the ext4 filesystem 1690is available on linux. This can be disabled via a configuration 1691option in /etc/e2fsck.conf. 1692 1693Fix a file descriptor leak in libblkid when checking to see if an ext4 1694or ext4dev module exists. 1695 1696Fix a bug in e2fsck where in preen mode, if there are disk I/O errors 1697while trying to close a filesystem can lead to infinite loops. 1698(Addresses Red Hat Bugzilla #465679) 1699 1700Fix a bug in resize2fs where passing in a bogus new size of 0 blocks 1701will cause resize2fs to drop into an infinite loop. (Addresses Red 1702Hat Bugzilla: #465984) 1703 1704Add a check in the Unix I/O functions in libext2fs so that when a 1705device is opened read/write, return an error if the device is 1706read-only using the BLKROGET ioctl. 1707 1708Fix debugfs's ncheck command so that it prints all of the names of 1709hardlinks in the same directory. 1710 1711Fix a bug in libblkid so it correctly detects whether the ext4 and 1712ext4dev filesystems are available, so that the ext4dev->ext4 1713fallback code works correctly. 1714 1715Programmer's Notes 1716------------------ 1717 1718Fix a parallel build problem by making sure util/subst is built before 1719trying to build the lib/et directory. (Addresses Sourceforge Bug: 1720#2143281) 1721 1722Updated "make depend" information for crc16.o 1723 1724 1725E2fsprogs 1.41.2 (October 2, 2008) 1726================================== 1727 1728Fix e2fsck's automatic blocksize detection. This fixes a regression 1729from e2fsprogs 1.40.7 which caused e2fsck to fail if the user 1730specifies a block number using the -b option if the blocksize option 1731isn't also specified using -B. Unfortunately, users very commonly 1732invoke e2fsck using "e2fsck -b 32768 /dev/hdXXX" to use the backup 1733superblock; in fack e2fsck will often suggest this kind of command 1734line. Oops. 1735 1736Enhance the debugfs's "ncheck" command so it will print all of the 1737pathnames for the specified inodes. (Previously, in some cases ncheck 1738might not print a pathname for an inode at all if some of the other 1739inodes had multiple hard links.) 1740 1741Enhance debugfs's "hash" command so the hash seed can be specified via 1742a command-line option. In addition, allow the hash algorithm to be 1743specified by name instead of just by number. 1744 1745Fix e2fsck so that we don't accidentally print the translation file's 1746header when asking the user a custom question so there is no prompt 1747defined for a particular problem record. For example, the question 1748"Run journal anyway" will get the PO header tacked on because e2fsck 1749erroneously passed the null string to _(). (Addresses Launchpad Bug: 1750#246892) 1751 1752Enhance badblocks so that it can test a normal file which is greater 1753than 2GB. 1754 1755Enhance the badblocks command so that it displays the time and 1756percentage complete when in verbose mode. (Addresses Debian Bug: 1757#429739) 1758 1759Fix a potential memory leak in a error handling path in debugfs's 1760ncheck function. 1761 1762Fix a potential memory corruption problem if a memory allocation fails 1763in resize2fs. 1764 1765Fix the usage message for debugfs's logdump command to be consistent 1766with its man manpage. 1767 1768Update Polish, French, Vietnamese, Dutch, Indonesian, German, Czech, 1769and Sweedish translation from the Translation Project. 1770 1771Add documentation for the file I/O functions to the libext2fs.texinfo 1772file. (Addresses Debian Bug: #484877) 1773 1774Update and clarified various man pages. (Addresses Launchpad Bug 1775#275272; Addresses Debian Bugs: #498100, #498101, #498102, #498103) 1776 1777Fixed various Debian packaging issues --- see debian/changelog for 1778details. (Addresses Debian Bug: #497619) 1779 1780Programmer's Notes 1781------------------ 1782 1783Fix a potential file descriptor leack in libcom_err by setting the 1784close-on-exec flag for a fd used for debugging. (Addresses Red Hat 1785Bugzilla #464689) 1786 1787Fix a potential race in libcom_err by using sem_post/sem_init. SuSE 1788has been carrying a patch for a long time to prevent a largely 1789theoretical race condition if a multi-threaded application adds and 1790removes error tables in multiple threads. Unfortunately SuSE's 1791approach breaks compatibility by forcing applications to link and 1792compile with the -pthread option; using pthread mutexes has 1793historically been problematic. We fix this by using sem_post/sem_init 1794instead. 1795 1796Fix e2fsprogs-libs build failure due to 'subs' target. (Addresses 1797Sourceforge Bug: #2087502) 1798 1799Avoid linking e2initrd_helper, debugfs, blkid, and fsck with unneeded 1800libraries when using ELF shared libraries. 1801 1802Fix ELF shared library when building on systems that don't already 1803have the e2fsprogs shared libraries already installed. (Addresses 1804Sourceforge Bug: #2088537) 1805 1806Fix the pkg-config files so they work correctly when linking with 1807static libraries and fix the include directory so programs don't have 1808to use #include <ext2fs/ext2fs.h>, but can use #include <ext2fs.h> 1809instead. (Addresses Sourceforge Bug: #2089537) 1810 1811Make sure ext2fs_swab64() is compiled for all platforms, and not just 1812for x86. (Addresses Debian Bug: #497515) 1813 1814Remove the unused ext2fs_find_{first,ext}_bit_set() functions for all 1815non-x86 platforms. (They had been removed for x86 earlier.) 1816 1817Fix diet libc compilation support, which had bitrotted due to lack of 1818TLC. Fixing this improves general portability. 1819 1820When installing the link library when using ELF shared libraries, 1821avoid using absolute pathnames if the link library and the shared 1822library are installed in the same directory. (Addresses Sourceforge 1823Bug: #1782913) 1824 1825Fix gen-tarball so it will work even if the top-level directory has 1826been renamed to something other than "e2fsprogs". Also make 1827gen-tarball print the size of the resulting tar.gz file. 1828 1829 1830E2fsprogs 1.41.1 (September 1, 2008) 1831==================================== 1832 1833Many people are forgetting to update their mke2fs.conf file, and this 1834causes ext3, ext4, and ext4dev filesystems won't get created with the 1835proper features enabled. We address this in two ways. First, mke2fs 1836will issue a warning if there is not definition for an ext3, ext4, or 1837ext4dev filesystem and the user is trying to create such a filesystem 1838type. Secondly, when installing from a source build, "make install" 1839will provide basic configuration file handling for /etc/mke2fs.conf. 1840If it exists, and does not mention ext4dev, it will be moved aside to 1841/etc/mke2fs.conf.e2fpsrogs-old and the new /etc/mke2fs.conf file will 1842be installed. If the existing /etc/mke2fs.conf file does mention 1843ext4dev, then "make install" will install official mke2fs.conf file as 1844/etc/mke2fs.conf.e2fsprogs-new and issue a message to the user that 1845they should look to see if any changes need to be merged. 1846 1847The mke2fs program will now create the journal in the middle of the 1848filesystem, since this minimizes seek times on average for fsync-heavy 1849workloads. In addition, mke2fs will now create journals using extents 1850for filesystems that support them. This results in a more efficient 1851encoding for the journal since it eliminates the need for using 1852indirect blocks. 1853 1854The mke2fs program will avoid allocating an extra block to the 1855journal. (Addresses Sourceforge Bug: #1483791) 1856 1857Mke2fs will correctly enforce the prohibition against features 1858(specifically read-only features) in revision 0 filesystems. (Thanks 1859to Benno Schulenberg for noticing this problem.) 1860 1861Mke2fs previously would occasionaly create some slightly non-optimally 1862placed inode tables; this bug has been fixed. 1863 1864The mke2fs and tune2fs programs now print the correct usage message 1865describing the maximum journal size. (Addresses Debian Bug: #491620) 1866 1867Add support for setting the default hash algorithm used in b-tree 1868directories in tune2fs (from a command-line option) or mke2fs (via 1869mke2fs.conf). In addition, change the default hash algorithm to 1870half_md4, since it is faster and better. 1871 1872The blkid library will now recognize MacOS hfsx filesystems, and 1873correctly extract the label and uuid for hfs, hfsx, and hfsplus 1874filesystems. (Addresses Sourceforge Feature Requests: #2060292) 1875 1876The blkid library has improved detection of JFS and HPFS filesystems. 1877(Addresses Launchpad Bug: #255255) 1878 1879The blkid library is now much more efficiently handling devicemapper 1880devices, mainly by no longer using the devicemapper library. This can 1881speed up access for systems with a large number of device mapper 1882devices. 1883 1884Blkid had a number of cache validation bugs in libblkid that have been 1885fixed. (Addresses Debian Bug: #493216) 1886 1887Resize2fs will now properly close out the "updating inode references" 1888progress bar so there is a newline printed before printing the final 1889"resize is successful" message. 1890 1891Resize2fs will now correctly handle filesystems with extents and/o 1892uninitialized block groups correctly when file/directory blocks need 1893to relocated (i.e., when shrinking a filesystem or if the resize_inode 1894is not present). To support this, the ext2fs library now supports 1895initializing inode and block bitmaps that are not yet initialized when 1896allocating them using ext2fs_new_block() and ext2fs_new_inode(). In 1897addition, e2fs_block_iterate2() can now support changing the location 1898of interior nodes of an extent tree, and ext2fs_extent_set_bmap() has 1899been optimized to avoid creating unnecessary new extents when updating 1900the location of blocks in the extent tree. This will also help out 1901e2fsck's recovery of obscurely corrupted filesystems with extents, 1902when blocks are claimed by multiple inodes. 1903 1904Add support for on-line resizing ext4 filesystem with the flex_bg 1905filesystem feature. The method for doing so is not optimal, but to do 1906a better job will require kernel support. 1907 1908E2fsprogs 1.41.0 intrduced a bug in libext2fs which casued e2image and 1909debugfs programs to not be able to read e2image files; the signed 1910vs. unsigned bug in the code which read bitmaps from the e2image has 1911been fixed. (Addresses Debian Bug: #495830) 1912 1913Resize2fs is now correctly managing the directory in-use counts when 1914shrinking filesystems and directory inodes needed to be moved from one 1915block group to another. This bug has been around since e2fsprogs 19161.26, and is largely harmless, but does cause a filesystem corruption 1917which will be flagged by e2fsck after the filesystem has been shrunk. 1918 1919E2fsck will no longer issue spurious complaints about the inode size 1920caused by very large extent-based files, and by blocks reallocated 1921using fallocate() with the FALLOC_FL_KEEP_SIZE option. (Addresses 1922Kernel Bugzilla: #11341) 1923 1924Mke2fs will now set the creation timestamp on the lost+found directory 1925and the root directory. (More generally, all new inodes created using 1926the ext2fs library will correctly set the creation timestamp.) 1927 1928E2fsck now correctly calculates ind/dind/tind statistics in the 1929presence of extent-based files. In addition, "e2fsck -v" will report 1930statistics of the depth of extent trees in the filesystem. E2fsck can 1931also give an inode fragmentation report using "e2fsck -E fragcheck" 1932which can be useful when debugging the kernel block allocation 1933routines. 1934 1935Fix support for empty directory blocks in ext4 filesystems with 64k 1936blocksize filesystems. 1937 1938E2fsck will now print the depth of corrupt htree directories. 1939 1940Debugfs's htree command now correctly understands extent-based 1941directories. It will also print out the minor hash as well as the 1942major hash. 1943 1944Debugfs has a new command which will print the supported features of 1945e2fsprogs, to enable scripts to know whether the installed version of 1946e2fsprogs can support a specific feature. 1947 1948Debugfs will now write files using extents for filesystems that 1949support them. 1950 1951The error message printed by "tune2fs -I" if the inode size was too 1952small was rather confusing, so it has been improved. Also, we won't 1953try to create an undo log until we know that command-line-specified 1954parameters such as "tune2fs -I <inode size>" are valid. 1955 1956Given some filesystems found "in the wild" that had non-zero block 1957group checksums even though the uninit_bg/gdt_sum feature was not 1958enabled, e2fsck would issue spurious error messages. Teach 1959ext2fs_group_desc_csum_verify() to ignore the block group checksum 1960entirely if the feature flag is not set. (Addresses Debian Bug: 1961#490637) 1962 1963The blkid program will now print out a user-friendly listing of all of 1964the block devices in the system and what they contain when given the 1965-L option. (Addresses Debian Bug: #490527) 1966 1967The filefrag program now has a more accurate calculation for the 1968number of ideal extents. (Addresses Debian Bug: #458306) 1969 1970The test I/O manager is now enabled by default, but its overhead is 1971only incurred when it would be enabled via the TEST_IO_FLAGS or 1972TEST_IO_BLOCK environment variables. 1973 1974Typographical errors in various program strings and usage messages 1975have been fixed; most of these were pointed out by the e2fsprogs 1976message catalog translators. (Thanks, translators!) 1977 1978Update and clarified various man pages, as well as some typographical 1979errors in the libext2fs texinfo file. 1980 1981Fixed various Debian packaging issues --- see debian/changelog for 1982details. 1983 1984Add Indonesian and update French, Polish, Dutch, German, Sweedish, 1985Czech, and Vietnamese Translations. (Addresses Debian Bugs: #313697, 1986#401092) 1987 1988Programmer's Notes 1989------------------ 1990 1991Fix portability problem with the badblocks group; for systems that 1992don't have nanosleep(), try using usleep() instead. 1993 1994The "make check" target in the e2fsck directory now sets 1995LD_LIBRARY_PATH before running the various e2fsck internal library 1996regression tests. 1997 1998The crc32 regression test in the e2fsck library is now portable to 1999greater varienty of environments, including big-endian systems and 2000when cross-building e2fsprogs for embedded systems. (Addresses 2001Sourceforge Bug: #2019287) 2002 2003The ext2fs_extent_set_bmap() had some bugs when setting the first 2004block in a file, or when replacing a single block extent. Those cases 2005fortunately were came up relatively rarely when e2fsck was checking 2006files, but caused some problems when resize2fs was shrinking 2007extent-based files. 2008 2009Fix a potential core-dumping bug in libe2p's iterate_on_dir() 2010function. 2011 2012Various ext2fs library functions --- ext2fs_block_iterate2(), 2013ext2fs_initialize() and ext2fs_extent_open() --- now correctly free 2014allocated memory to avoid memory leaks in all of their error return 2015paths. 2016 2017Ext2ed was failing to build because masix support had been removed in 2018the rest of e2fsprogs, so ext2ed no longer has masix support, either. 2019 2020The configure script now respects the LDFLAGS environment variable if 2021it is set when configure is called. (Addresses Sourceforge Feature 2022Request: #1937287) 2023 2024Libuuid is now more portable to the Windows platform. (Addresses 2025Sourceforge Feature Request: #1937287) 2026 2027The configure script now uses AC_MSG_{RESULT,WARN,ERROR} instead of 2028bare echo commands so that configure flags such as --quiet work 2029correctly. (Addresses Sourceforge Patches: #2058794) 2030 2031A few uses of sprintf have been removed from the ext2fs library to 2032make life easier for bootloaders with a limited libc environment. 2033(Addresses Sourceforge Bug: #2049120) 2034 2035The ext2fs_read_inode() checks the validity of the inode number passed 2036to it earlier, to avoid doing some needless work when it would fail 2037anyway. 2038 2039The ext2fs_open() checks the validity of the blocksize parameter 2040passed to it earlier, to avoid doing some needless work when it would 2041fail anyway. 2042 2043Disable a very annoying automatic "%.sh -> %" GNU make rule in the 2044top-level Makefile. That automatic rule is used to better support 2045SCCS, but it caused problems for a particular niche distribution which 2046likes to use configure.sh files to store the configure options used to 2047build a package. Unfortuntaely GNU make will use the configure.sh to 2048replace the configure script, resulting in a self-inflicted fork bomb 2049leading to an out-of-memory crash. 2050 2051To support old GNU C compilers don't use C99/C++ comments, but only 2052K&R style comments, and don't try to use __builtin_expect if __GNUC__ 2053is less than 3. (__builtin_expect is only supported for gcc versions 20542.96 and up, and it's tricky to check for gcc 2.95 vs gcc 2.96; since 2055this is an optimization, we only try to use __builtin_expect for gcc 3 2056and up.) 2057 2058In e2fsck's crc routines, make sure we use WORDS_BIGENDIAN instead of 2059__LITTLE_ENDIAN, which are only defined by glibc's header files and 2060hence isn't portable. 2061 2062For the convenience for some distributions that need a static tune2fs, 2063the Makefile for misc/ now has a tune2fs.static target. 2064 2065The ext2fs_block_iterate2() function now supports BLOCK_FLAG_APPEND 2066for extent-based files 2067 2068The ext2fs_bmap() function now supports BMAP_ALLOC for extent-based 2069files. 2070 2071All source files no longer have any trailing white space. 2072 2073The io_channel_read_blk64() and io_channel_write_blk64() functions are 2074now functions instead of C preprocessor macros to provide better 2075forward compatibility. 2076 2077The e2fpsrogs translation template now expands the @x abbrevation. 2078 2079Various namespace leackages in libblkid, libe2p, and libext2fs have 2080been fixed. 2081 2082Fix a parallel build problem in e2fsprogs. 2083 2084E2fsprogs is now more portable to Solaris. 2085 * blkid no longer assumes that the TIOCGSIZE and TIOCGWINSZ ioctl's 2086 are always present. 2087 * Scripts do not assume that /bin/true is always in /bin 2088 * Don't use __FUNCTION__ since Solaris's C99 doesn't support it. 2089 * Flush stdio handles before calling setbuf(), since Solaris will 2090 discard any pending output to the stream. 2091 * Define _XOPEN_SOURCE to 600 since Solaris's header files are very 2092 picky about which C compiler can beused for SUSv3 conformance. 2093 Use of C99 is not compatible with SUSv2 (_XOPEN_SOURCE=500), 2094 and C89 is not compatible with SUSv3 (_XOPEN_SOURCE=600). 2095 Since we need some SUSv3 functions, consistently use SUSv3 so 2096 that e2fsprogs will build on Solaris using c99. 2097 * Solaris C99 does not support varargs C preprocessor macros 2098 * Solaris header files pollute the C namespace if in/netinet.h 2099 is included, which conflicts with e2fsprogs' use of the kmem_cache_t 2100 typedef. 2101 * Solaris ships with a pathetically ancient shell in /bin/sh so we 2102 avoid the use of various more avanced shell constructs such as $(). 2103 2104The "make rpm" command will now take some extra configure optiosn from 2105the build environment without needing to patch the source tree. 2106 2107The ext2fs_add_dir_block() function will now grow the dblist more 2108aggressively as an optimization to avoid copying the array too often. 2109 2110The e2fsck_write_bitmaps() will write the block and inode bitmaps 2111together instead of in two passes. 2112 2113 2114E2fsprogs 1.41 (July 10, 2008) 2115============================== 2116 2117Add support for ext4 filesystem features, in particular extents, 2118uninit_bg, flex_bg, huge_file, and dir_nlink features. Also add 2119support for checking journal checksums. Debugfs will print new 2120superblock and inode fields that were defined for ext4. For example, 2121the nanosecond and i_version fields of an inode, and the 2122s_min_extra_isize and s_wanted_extra_isize fields from the superblock. 2123Note: Resize2fs doesn't currently support the combination of flex_bg 2124and !resize_inode. (Addresses Debian Bug: #388452, #425477) 2125 2126Tune2fs can support migrating a filesystem from using 128 byte inodes 2127to 256 byte inodes, so it can take advantage of the full features of 2128ext4. 2129 2130Add support for "undo" support. E2fsck and mke2fs can optionally 2131record an undo log which can replayed by the program e2undo. 2132 2133E2fsck could damage a filesystem by trying to relocate inode tables 2134due to corrupted block group descriptors, where the attempted inode 2135table relocation would do far more harm than good. E2fsck has been 2136fixed to detect this these sorts of corrupted block group descriptors 2137much earlier in e2fsck processing, so it can try to use the backup 2138superblock and block group descriptors first. This should be a much 2139better strategy for recovering these types of corrupted filesystems. 2140(Addresses Sourceforge Bug: #1840291) 2141 2142E2fsck will display a more understandable message when the last check 2143field in the superblock is in the future. (Addresses Debian Bug: 2144#446005). 2145 2146E2fsck now performs more extensive and careful checks of extended 2147attributes stored in the inode. 2148 2149Enhance mke2fs to print a more explanatory error message when 2150ext2fs_get_device_size() returns EFBIG. (Addresses Debian Bug: 2151#488663) 2152 2153Fix mke2fs to use a default block size of 4k when formatting an 2154external journal device. This is done by using a fixed filesystem 2155type list that consists only of the single filesystem type "journal" 2156when looking up configuration keys in /etc/mke2fs.conf. (Addresses 2157Debian Bug: #488663) 2158 2159Speed up how mke2fs writes the journal data blocks by writing the disk 2160blocks in larger chunks. 2161 2162Fix blkid handling of stale devices. Fix a bug which could cause a 2163core dump while garbage collecting the blkid cache, and assure that 2164blkid_find_dev_with_tag() never returns a non-existent device. Also, 2165if a filesystem is found at a new /dev location, eliminate any 2166duplicate stale entries which can not be verified. 2167(Addresses Debian Bugs: #487758, #487783) 2168 2169Add more paranoid checks for LVM volumes and swap partitions in 2170blkid's probe function, to reduce the chances of false positives. 2171 2172The mke2fs program now has a much more sophisticated system for 2173controlling configuration parameters of a newly created filesystem 2174based on a split filesystem and usage type system. The -t option to 2175mke2fs was a deprecated alias to -c; it now specifies a filesystem 2176type (ext2, ext3, ext4, etc.), while the -T option can now be a comma 2177separated usage list. The filesystem type information and type 2178information is used to extract configuration parameters from the 2179/etc/mke2fs.conf file. 2180 2181The mke2fs program will no longer complain and request the -f option 2182when the user tries to create a filesystem with greater than 2**31 2183blocks. 2184 2185When creating a filesystem for the GNU Hurd use a fs-type of Hurd and 2186adjust the mke2fs.conf file so filesystems for the Hurd are created 2187with a blocksize of 4096 and inode size of 128, which is all it knows 2188how to handle. (Addresses Debian Bug: #471977) 2189 2190Mke2fs will always make sure that lost+found always has at least 2 2191blocks, even for filesystems with very large blocksizes (i.e., 64kb). 2192 2193Resize2fs will now print the minimum needed filesystem size if given 2194the -P option, and will resize the filesystem to the smallest possible 2195size if given the -M option. 2196 2197Fix resize2fs to clean up the resize_inode if all of the reserved gdt 2198blocks are consumed during an off-line resize. 2199 2200The "ls" command in debugfs now supports the -p option, which causes 2201it to quote the filenames so that spaces or tabs in directory entries 2202are easily visible. (Addresses Red Hat Bugzilla: #149480; Addresses 2203Sourceforge Feature Request: #1201667) 2204 2205Fix a potential off-by-one buffer oveflow in the fs_device_name in an 2206e2image file. 2207 2208The chattr program will return a non-zero exit code in case of 2209failures, and error messages can be suppressed with the -f option. 2210(Addresses Red Hat Bugzilla: #180596) 2211 2212Fix a bug in badblocks which caused it to overrun an array and likely 2213crash if more than 8 test patterns are specified using the -t option. 2214(Addresses Debian Bug: #487298) 2215 2216Add support to badblocks to limit how quickly it reads from the disk 2217drive (so it can be used for background scrubbing), and so it will 2218abort after finding a given number of errors. 2219 2220Remove support for the legacy big-endian filesystem format which only 2221existed on extremely long-dead PowerPC kernels almost a decade ago. 2222 2223Remove MASIX support from e2fsprogs. 2224 2225Add I/O statistics reporting to e2fsck. 2226 2227Update Vietnamese, Polish, French, Spanish, German, Catalan, Dutch, 2228Czech translations. 2229 2230Fixed various Debian packaging issues --- see debian/changelog for 2231details. (Addresses Debian Bugs: #487443, #487675, #490003) 2232 2233Fixed spelling mistakes, typos, and otherwise clarified man pages. 2234(Addresses Debian Bugs: #393313, #487849, #440983, #440981) 2235 2236Programmer's Notes 2237------------------ 2238 2239Factor out bitmap code in preparation for adding 64-bit new-style 2240bitmaps. 2241 2242Fix gcc -Wall warnings 2243 2244Fix the pkg-config files so that private librares are specified in 2245"Libs.private:". 2246 2247Fix the libext2fs.texinfo manual so it builds with modern versions of 2248texinfo. 2249 2250Silence the makefile from showing the awk command used to build the 2251 2252Clean up the badblocks group so to make it more portable and robust. 2253 2254Avoid using predictable filenames in /tmp in blkid's regression test 2255suite. Also remove bashism's in the regression test script. 2256 2257If the configure script is given --with-diet-libc, don't use thread 2258local storage, since diet libc doesn't support TLS. (Addresses 2259Sourceforge Bug: #2000654) 2260 2261Fix the blkid regression test suite to tolerate older versions of 2262mkswap that don't support the -U option. 2263 2264A few library routines have been converted to support 64-bit block 2265numbers; in particular, the I/O manager functions, the test_io, 2266inode_io, and unix_io managers have all be converted to support 64-bit 2267operation. 2268 2269Debugfs can now be extended for use by test programs. See 2270lib/ext2fs/extents.c for an example for how it can be used. The test 2271program links against the debugfs object files, and provides 2272additional commands by defining an auxiliary libss command table. 2273 2274The lazy_bg filesystem feature, which was only used by developer's 2275testing, has been removed since it has been largely supplanted by 2276uninit_bg. This also simplifies the code. 2277 2278 2279E2fsprogs 1.40.11 (June 17, 2008) 2280================================= 2281 2282Mke2fs, tune2fs, and resize2fs now use floating point to calculate the 2283percentage of reserved blocks. (Addresses Debian Bug: #452639) 2284 2285Updated Spanish and Catalan translations. 2286 2287Fixed various Debian packaging issues --- see debian/changelog for 2288details. (Addresses Debian Bugs: #483962, #483023) 2289 2290Add detection for ZFS volumes to the libblkid library. 2291 2292Fixed spelling mistakes, typos, and otherwise clarified man pages. 2293(Addresses Debian Bug: #486463) 2294 2295Programmer's notes: 2296------------------- 2297 2298Fix marginal C code in probe_lvm2() function to the blkid library more 2299portable for older compilers. 2300 2301Fix build problems on MacOS X. (Addresses Sourceforge Bug: #1972473) 2302 2303Fix ext2fs_swap{16,32,64} functions so they can be used by external 2304applications on big-endian machines. (Addresses Debian Bug: #484879) 2305 2306 2307E2fsprogs 1.40.10 (May 21, 2008) 2308================================ 2309 2310When deciding whether or not to revalidate a blkid cache entry, if the 2311device's mtime is newer than the last time the cached entry was 2312validated, force a revalidation. 2313 2314Fix a potential data corruption bug in e2fsck in the journal replay. 2315The chances of this is happening is extremely remote, especially the 2316default data=ordered or data=writeback modes. However, if a block 2317which has been journalled starts with the first four bytes 0xc03b3998, 2318when e2fsck replays the journal, those four bytes will be replaced 2319with zero's. Fortunately, it is highly, highly unlikely for e2fsck 2320metadata to begin with those fatal 4 byte sequence, and unless 2321data=ordered mode is in use, data blocks are never journaled. 2322 2323Updated German, Dutch, Sweedish, and Vietnamese translations. 2324 2325Programmer's notes: 2326------------------- 2327 2328Fixed various Debian packaging issues --- see debian/changelog for 2329details. 2330 2331Remove default sizes of types when cross compiling, since autoconf 23322.50 can figure this out automatically now. 2333 2334 2335E2fsprogs 1.40.9 (April 27, 2008) 2336================================= 2337 2338SuSE's security team audited uuidd and came up with a few minor 2339issues. None of them are serious given that uuidd runs setuid as a 2340unprivileged user which has no special access other than libuuid 2341directory, but it's good to get them fixed. 2342 2343One additional fix in ext2fs_swap_inode_full() needed for resize2fs to 2344work correctly with in-inode extended attributes. 2345 2346Updated German, Czech, Dutch, French, Polish, Sweedish, and Vietnamese 2347translations. 2348 2349Debugfs will avoid using a pager if the standard output is not a tty. 2350 2351Fix debugfs and tune2fs to correctly handle daylight savings time when 2352parsing a time string. 2353 2354Fixed spelling mistakes, typos, and otherwise clarified man pages. 2355 2356Fix fsck completion bars when multiple filesystems were being checked 2357in parallel. (Addresses Debian Bug: #432865, Addresses Launchpad Bug: 2358#203323, Addresses Sourceforge Bug: #1926023) 2359 2360Fix fsck so that progress information is sent back correctly when 2361multiple filesystems are being check and the output of fsck is being 2362redirected to a file descriptor. Also, include the device name (w/o 2363spaces) in the progress information sent back via a file descriptor. 2364(Addresses Launchpad Bug: #203323, Addresses Sourceforge Bug: 2365#1926023) 2366 2367Teach fsck to treat "ext4" and "ext4dev" as ext* filesystems. 2368 2369If logsave receives a SIGTERM or SIGINT signal, it will now pass that 2370signal to its child process. 2371 2372Fix mke2fs's creation of are resize inode when there is a non-standard 2373s_first_data_block setting. 2374 2375Fix bug in blkid when run by an unprivileged user; most devices were 2376not reported correctly. 9Addresses Launchpad Bug: #220275) 2377 2378Mke2fs will not allow the logically incorect combination of 2379resize_inode and meta_bg, which had previously caused mke2fs to create 2380a corrupt fileystem. 2381 2382Fix fsck in German locales so that a 'j' means yes. 2383(Addresses Sourceforge Bug: #1947683) 2384 2385 2386Programmer's notes: 2387------------------- 2388 2389Fixed various Debian packaging issues --- see debian/changelog for 2390details. 2391 2392Update valgrind options in test_script to work with valgrind 3.2.3 2393 2394Update texinfo.tex to a much newer version from the FSF. 2395 2396Remove bashism for configure script and from the lib/ss Makefile. 2397Addresses Sourceforge Bug: 1921969 2398 2399Fix some silently broken tests: m_no_opt, m_meta_bg, and m_raid_opt. 2400 2401Fix build system so that if texinfo is not installed, it won't print a 2402(harmless) error message. 2403 2404 2405E2fsprogs 1.40.8 (March 13, 2008) 2406================================= 2407 2408Fixed e2image -I so it works on image files which are larger than 2GB. 2409 2410Fixed e2fsck's handling of directory inodes with a corrupt size field. If 2411the size is larger than the number of blocks found in the inode, don't 2412try to allocate extra empty blocks at the end of the directory to make 2413up the difference; there's no point to doing that. In addition, if 2414the size is not a multiple of a blocksize, always fix it. 2415 2416E2fsck handled a pass 2 "should never happen error" by not giving 2417enough information and then core dumping. Unfortunately, it was all 2418too easy to trigger the "should never happen" situation if a 2419directory's inode size was not correct. This has been fixed, but 2420e2fsck has also been taught how to handle this situation more 2421gracefully, by simply removing the inode hash tree information, so 2422that it can be rebuilt again after e2fsck's pass 3. (Addresses 2423Launchpad Bug: #129395) 2424 2425Resize2fs had a bug resizing large inodes with extended attributes 2426that was fixed in 1.40.6; unfortunately, it turned out it wasn't fixed 2427completely on big-endian systems such as PowerPC. The bug should be 2428completely fixed now. Yay for regression test suites. (Addresses Red 2429Hat Bugzilla: #434893) 2430 2431Updated German, Czech, Dutch, Polish, Sweedish, and Vietnamese 2432translations. Many thanks to Philipp Thomas from Novell for stepping 2433up to become the new German translation maintainer! (Addresses Debian 2434Bugs: #302512, #370247, #401092, #412882). 2435 2436When e2fsck is clearing a corrupt inode's HTREE directory information, 2437make it clear that it is just clearing the HTREE information, not the 2438entire inode. 2439 2440Fixed spelling mistakes, typos, and otherwise clarified man pages. 2441 2442Programmer's notes 2443------------------ 2444 2445Add new functions, ext2fs_dblist_get_last() and 2446ext2fs_dblist_drop_last(), which allows the caller to examine the last 2447directory block entry added to the list, and to drop if it necessary. 2448 2449Fixed a portability problem in libblkid with DJGPP. 2450 2451Fix an obvious typo in an "internal error" message in e2fsck. Thanks 2452to Philipp Thomas for pointing this out. 2453 2454If the info files are not built, change "make install" so it doesn't 2455fail with an error code. 2456 2457 2458E2fsprogs 1.40.7 (February 28, 2008) 2459==================================== 2460 2461Remove support for clearing the SPARSE_SUPER feature from tune2fs, and 2462depreciate the -s option, since it can result in filesystems which 2463e2fsck can't fix easily. There are very good reasons for wanting to 2464disable sparse_super; users who wants to turn off sparse_super can use 2465debugfs. (Addresses Sourceforge Bug: #1840286) 2466 2467Add missing options to mke2fs's usage message. (Addresses Sourceforge 2468Bug: #1751393) 2469 2470Fix bug in resize2fs when large (greater than 128 byte) inodes are 2471moved when a filesystem is shrunk; it was only moving the first 128 2472bytes, so extended attributes were not getting moved. (Addresses Red 2473Hat Bugzilla: #434893) 2474 2475E2fsck now prints an explicit message when the bad block inode is 2476updated, to avoid confusion about why the filesystem was modified. 2477(Addresses Sourceforge Bug: #756460) 2478 2479Allow mke2fs and tune2fs manipulate the large_file feature. 2480Previously we just let the kernel and e2fsck do this automatically, 2481but e2fsck will no longer automatically clear the large_file feature. 2482It still isn't really necessary to worry about this feature flag 2483explicitly, but some users seem to care. (Addresses Red Hat Bugzilla: 2484#258381) 2485 2486Suppress message about an old-style fstab if the fstab file is empty. 2487(Addresses Debian Bug: #468176) 2488 2489Fix (really minor) bug in debugfs's find_free_block so it avoids 2490reporting a free block more than once if there are too few free blocks 2491in the filesystem. (Addresses Sourceforge Bug: #1096315) 2492 2493Change e2fsck to no longer clear the LARGE_FILES feature flag 2494automatically, when there are no more > 2GB files in the filesystem. 2495It's been almost a decade since there have been kernels that don't 2496support this flag, and e2fsck clears it quietly without telling the 2497user why the filesystem has been changed. 2498 2499Fix bug which could cause libblkid to seg fault if a device mapper 2500volume disappears while it is being probed. (Addresses RedHat 2501Bugzilla: #433857) 2502 2503Inhance e2fsck's reporting of unsupported filesystem feature flags. 2504(Addresses Sourceforge Feature Request: #1175808) 2505 2506Fix option syntax in dumpe2fs for explicit superblock and blocksize 2507parameters. What was currently documented in the man page has been 2508broken for some time, due to getopt() implementation changes. The 2509option sytax has been changed to one which is can be more portable 2510supported and which is consistent with the format for extended options 2511in mke2fs and tune2fs. (Addresses Sourceforge Bug: #1830994) 2512 2513Add support to tune2fs to clear the resize_inode feature. This 2514requires an fsck afterwards. (Addresses Red Hat Bugzilla: #167816) 2515 2516Teach blkid to detect LVM2 physical volumes. (Addresses Red Hat 2517Bugzilla: #409321) 2518 2519Add support for setting RAID stride and stripe-width via mke2fs and 2520tune2fs. Teach dumpe2fs to print the RAID parameters. 2521 2522Add support for setting new superblock fields to debugfs's 2523set_super_value. 2524 2525Add support for printing "mostly-printable" extended attributes in 2526Debugfs. 2527 2528Add support for the -M option to fsck, which causes it to ignore 2529mounted filesystem. 2530 2531Fix uuidd so that it creates the pid file with the correct pid number. 2532(Addresses Sourceforge Bug: #1893244) 2533 2534Fix various gcc -Wall warnings. 2535 2536Update Czech, Dutch, Polish, Sweedish, and Vietnamese translations 2537 2538Fixed spelling mistakes, typos, and otherwise clarified man pages. 2539(Addresses Sourceforge Patch: #1399325) 2540 2541 2542Programmer's notes: 2543------------------- 2544 2545Fixed various Debian packaging issues --- see debian/changelog for 2546details. (Addresses Debian Bug: #466929) 2547 2548Add new flag EXT2_FLAG_NONFREE_ON_ERROR ext2fs_open2() which returns a 2549partially filled-in filesystem object on an error, so that e2fsck can 2550print more intelligent error messages. 2551 2552Add a new function e2p_edit_feature2() which allows the caller to 2553specify which feature flags are OK to set or clear, and which returns 2554more specific information about feature flags which are not allowed to 2555be set/cleared. 2556 2557Set the C locale in the test_script driver since it uses [A-Za-z]. 2558(Addresses Sourceforge Bug: #1890526) 2559 2560Use fcntl locking instead of lockf in libuuid since Cygwin doesn't 2561support lockf(). 2562 2563Change configure.in to avoid using the 'dc' command unless it is 2564absolutely needed. (i.e., when using parsing a WIP-style version 2565number) (Addresses Sourceforge Bug: #1893024) 2566 2567Add portability checks to support compilation under DJGPP. 2568 2569Update to the latest samba tdb code before the LGPLv3 change, which 2570fixes a realloc() leak on failure. 2571 2572Fix memory leak in ext2fs_alloc_block(). 2573 2574Fix makefile dependency issues for various install targets. 2575(Addresses-Sourceforge-Patches: #1903484, #1903466, #1903456) 2576 2577Improve descriptions for the r_move_itable and r_resize_inode tests. 2578 2579 2580E2fsprogs 1.40.6 (February 9, 2008) 2581=================================== 2582 2583Add support for returning labels for UDF filesystems in the blkid 2584library. 2585 2586Fix bug in the blkid library where cached filesystems was not being 2587flushed when opening USB devices returned the error ENOMEDIUM. 2588(Addresses Debian Bug: #463787) 2589 2590Added logic to the blkid library to automatically choose whether a 2591filesystem should be mounted as ext4 or ext4dev, as appropriate. 2592 2593Allow tune2fs to set and clear the test_fs flag on ext4 filesystems. 2594 2595Fix a bug in e2fsck which caused it to core dump if e2fsprogs had been 2596configured with --enable-jbd-debug. 2597 2598Document the BLKID_FILE environment variable in the libblkid man page 2599 2600Programmer's Notes: 2601------------------- 2602 2603Update e2fsprogs translation template and Vietnamese and Czech translations 2604 2605Fixed various Debian packaging issues --- see debian/changelog for 2606details. (Addresses Debian Bugs: #436058) 2607 2608Don't try to create $DESTDIR/etc/init.d as part of make install as we 2609don't install the init.d script (and it's not the recommended way to 2610start uuidd anyway). (Addresses Sourceforge Bug: #1885085) 2611 2612Use thread local storage to fix a theoretical race condition if two 2613threads try to format an unknown error code at the same time in the 2614com_err library. 2615 2616 2617E2fsprogs 1.40.5 (January 27, 2008) 2618=================================== 2619 2620Fix a potential overflow big in e2image if the device name is too 2621long. 2622 2623Mke2fs will now create new filesystems with 256 byte inodes and the 2624ext_attr feature flag by default. This allows for much better future 2625compatibity with ext4 and speeds up extended attributes even on ext3 2626filesystems. 2627 2628Teach e2fsck to ignore certain "safe" filesystem features which are 2629set automatically by the kernel. Having the kernel set features 2630automagically behind the user's back is a bad idea, and we should try 2631to break the kernel of this habit, especially for the newer ext4 2632feature flags. But for now, we will try to avoid needless full checks 2633which can annoy users who are doing fresh installs. 2634 2635Add support in tune2fs and mke2fs for making a filesystem as being "ok 2636to be used with test kernel code". This will be needed for using test 2637filesystems with the latest ext4 kernel code. 2638 2639Change e2fsck -fD so that it sorts non-htree directories by inode 2640numbers instead of by name, since that optimizes performances much 2641more significantly. (Addresses-Sourceforge-Feature-Request: #532439) 2642 2643If e2image fills the disk, fix it so it exits right away instead of 2644spewing large numbers of error messages. 2645(Addresses-Sourceforge-Feature-Request: #606508) 2646 2647If ftruncate64() is not available for resize2fs, let it use ftrucate() 2648instead, but have it check to see if the size would get truncated, and 2649skip calling ftruncate in that case. 2650 2651Add support for detecting HFS+ filesystems in the blkid library. 2652 2653Add supprt in the blkid library for ext4/ext4dev filesystems. 2654 2655Fix a bug in blkid where it could die on a floating point exception 2656when presented with a corrupt reiserfs image. 2657 2658Fix blkid's handling of ntfs UUID's so that leading zeros are printed 2659such that UUID string is a fixed length. 2660 2661Add sample python bindings for the uuid library to the contrib 2662directory. (Addresses-Sourceforge-Patches: #778817) 2663 2664Fix debugfs's 'lsdel' command so it uses ext2fs_block_iterate2 so it 2665will work with large files. (Addresses Sourceforge Feature Request: 2666#1257500 and Sourceforge Support Request: #1253511) 2667 2668Allow the debugfs 'undel' command to undelete an inode without linking 2669it to a specific destination directory, since this might require 2670allocating blocks that could overwrite some yet-to-be-recovered 2671deleted files. (Addresses-Sourceforge-Feature-Request: #967141) 2672 2673Update Swedish translations from the Translation Project. 2674 2675Programmer's Notes: 2676------------------- 2677 2678Fix configure handling of --sbindir (which should rarely be used, but 2679someone did complain, so let's fix it). (Addresses Sourceforge Bug: 2680#498381) 2681 2682Updated e2fsprogs.spec file to include a new uuidd package 2683 2684Use pkg-config to determine where to find the devmapper library so we 2685can find out where it is located on different distributions. 2686 2687Fix Makefile race so that "make -j3 distclean" works correctly 2688 2689Fix portability problems on non-Linux/non-Hurd/non-Masix systems, 2690especially on MacOS X systems. (Addresses Sourceforge Bugs: #1861633, 2691#1819034, #1863819) 2692 2693Fixed spelling mistakes, typos, and otherwise clarified man pages. 2694 2695Fixed various Debian packaging issues --- see debian/changelog for 2696details. (Addresses Debian Bugs: #459403, #459475, #459614) 2697 2698Remove the --enable-dynamic-static configure option, and build e2fsck 2699dynamically by default. If the user wants e2fsck.static, he/she will 2700need to build it via "make -C e2fsck e2fsck.static" 2701 2702Fix various build warnings due to missing prototypes. 2703(Addresses Sourceforge Patch: #1861663, #1861659) 2704 2705 2706E2fsprogs 1.40.4 (December 31, 2007) 2707==================================== 2708 2709Improve time-based UUID generation. A new daemon uuidd, is started 2710automatically by libuuid if necessary. This daemon is setuid to allow 2711updates to /var/lib/libuuid, so the clock sequence number can be 2712stored and so if the clock is set backwards, it can be detected. 2713(Addresses Sourceforge Bug: #1529672, Addresses Red Hat Bugzilla: 2714#233471) 2715 2716Filter out the NEEDS_RECOVERY feature flag when writing out the backup 2717superblocks. This avoids e2fsck from concluding that a full 2718filesystem check is required before backing up the superblock due to 2719changes in the feature flags. (Addresses Debian Bug: #454926) 2720 2721Fix fsck to only treat the '#' character as a comment at the beginning 2722of the line in /etc/fstab. Otherwise fstabs for the fuse filesystem 2723will cause fsck to issue an bogus warning message. 2724(Addresses Gentoo bug: #195405, Addresses Sourceforge bug: #1826147) 2725 2726Format control characters and characters with the high eighth bit set 2727when printing the contents of the blkid cache, to prevent filesystems 2728with garbage labels from sending escape sequences to the user's screen 2729that might, for example place it in graphics mode. (Addresses Ubuntu 2730Bug: #78087) 2731 2732Fix sign-extension problem on 64-bit systems in in the com_err 2733library. (Addresses Sourceforge Bug: #1809658) 2734 2735Avoid division by zero error when probing an invalid FAT filesystem in 2736the blkid library. (Addresses Sourceforge Bug: #1831627) 2737 2738Update Dutch, Polish, and Vietnamese translations from the Translation 2739Project. Remove the Rwandan translation upon advice of the 2740Translation Project. 2741 2742Programmer's Notes: 2743------------------- 2744 2745Fix the libss "make check" regression test so that it works if the 2746current directory is not in the user's path or if the libss shared 2747library is not installed. (Addresses Sourceforge Bug: #1848974) 2748 2749Fixed spelling mistakes, typos, and otherwise clarified man pages. 2750(Addresses Debian Bugs: #444883, #441872) 2751 2752Fixed various Debian packaging issues --- see debian/changelog for 2753details. (Addresses Debian Bugs: #437720, #451172, #458017) 2754 2755Fix build failure on non-Linux/non-Hurd/non-Masix systems. 2756(Addresses Sourceforge Bug: #1859778) 2757 2758Fix Hurd portability issues. (Addresses Debian Bug: #437720) 2759 2760 2761E2fsprogs 1.40.3 (December 5, 2007) 2762=================================== 2763 2764Fix a potential security vulnerability where an untrusted filesystem 2765can be corrupted in such a way that a program using libext2fs will 2766allocate a buffer which is far too small. This can lead to either a 2767crash or potentially a heap-based buffer overflow crash. No known 2768exploits exist, but main concern is where an untrusted user who 2769possesses privileged access in a guest Xen environment could corrupt a 2770filesystem which is then accessed by the pygrub program, running as 2771root in the dom0 host environment, thus allowing the untrusted user to 2772gain privileged access in the host OS. Thanks to the McAfee AVERT 2773Research group for reporting this issue. (Addresses CVE-2007-5497.) 2774 2775Fix hueristics in blkid which could cause a disk without partitions to 2776be incorrectly skipped when a loopback device is present. (Addresses 2777Red Hat Bugzilla #400321.) 2778 2779Fix e2image so that in raw mode it does not create an image file which 2780is one byte too large. 2781 2782Change mke2fs's usage message so it recommends the preferred -E option 2783instead of the deprecated -R option. 2784 2785Enhance the blkid library so it will recognize squashfs filesystems. 2786(Addresses Red Hat Bugzilla #305151.) 2787 2788Enhance e2fsck so it will force the backup superblocks to be backed up 2789if the filesystem is consistent and key constants have been changed 2790(i.e., by an on-line resize) or by e2fsck in the course of its 2791operations. 2792 2793Enhance blkid's detection of FAT filesystems; so that USB disks with 2794only a single bootable partition will not get missed. 2795 2796E2fsck will no longer mark a filesystem as invalid if it has time 2797errors (i.e., if superblock mount time or last write time is in the 2798future) and the user refuses to fix the problem. 2799 2800The Ubuntu init scripts don't properly set the system time correctly 2801from hardware clock if the hardware clock is configured to tick local 2802time instead of GMT time. Work around this as best as we can by 2803providing an option, buggy_init_scripts, in /etc/e2fsck.conf which can 2804be set on Ubuntu systems. (Addresses Debian Bug #441093, and Ubuntu 2805Bug #131201.) 2806 2807Fix fsck to ignore /etc/fstab entries for bind mounts. (Addresses Red 2808Hat Bugzilla #151533.) 2809 2810Fix e2fsck so that if the superblock is corrupt, but still looks 2811vaguely like an ext2/3/4 superblock, that it automatically tries to 2812fall back to the backup superblock, instead of failing with a hard 2813error. 2814 2815Make the e2fsprogs program more robust so that they will not crash 2816when opening a corrupt filesystem where s_inode_size is zero. 2817 2818Change e2fsck so it uses sscanf() instead of atoi() so it non-numeric 2819arguments are detected as such and the parse error is reported to the 2820user. (Addresses Debian Bug #435381.) 2821 2822Change e2fsck so it will not complain if a file has blocks reallocated 2823up to the next multiple of a system's page size. 2824 2825Fix bug in ext2fs_check_desc() which will cause e2fsck to complain 2826about (valid) filesystems where the inode table extends to the last 2827block of the block group. (Addresses Red Hat Bugzilla #214765.) 2828 2829Fix a bug in ext2fs_initialize() which causes mke2fs to fail while 2830allocating inode tables for some relatively rare odd disk sizes. 2831(Addresses Red Hat Bugzilla #241767.) 2832 2833Add Catalan translation and update Dutch and Swedish translations 2834from the Translation Project. 2835 2836Fix big-endian byte-swapping bug in ext2fs_swap_inode_full(). We 2837still had an issue when trying to figure out whether we need to 2838byte-swap fast symlinks that contained extended attributes. 2839 2840Fixed spelling mistakes, typos, and otherwise clarified man pages. 2841(Addresses SourceForge Bug #1821333.) 2842 2843 2844Programmer's Notes: 2845------------------- 2846 2847Fix mke2fs tests to avoid needing any significant ^M (CR) characters 2848 2849Add "make check" to the RPM spec file 2850 2851Fix "make install" and 'make unstall" in misc/Makefile.in so that it 2852works correctly when the prefix is not the root directory. 2853 2854Fix the resize2fs tests, r_move_itable and r_resize_inode, so they 2855clena up after themselves by deleting the test.img temporary file 2856after completing the test. 2857 2858Fixed a corner case bug ext2fs_unlink() when trying to delete the 2859first directory entry in a directory block and the last directory 2860entry in the previous directory block is not in use. Fortunately 2861ext2fs_unlink() is only used by debugfs and e2fsck, and in e2fsck in 2862places where it is extremely unlikely to run into this corner case. 2863 2864Fix missing dependency which would cuase parallel builds to fail. 2865(Addresses Sourceforge Bug #1842331.) 2866 2867Fix a build error on newer gcc caused by lib/ext2fs/ismounted.c 2868calling open(O_CREATE) with a missing mode parameter. 2869 2870Fix the test_ss.c in lib/ss so it can be used as an example 2871application program for the library as well as a regression test 2872suite. 2873 2874Fix ext2fs_dblist_dir_iterate() so that error codes and abort codes 2875are properly passed back up through the call stack. 2876 2877E2fsprogs 1.40.2 (July 12, 2007) 2878================================ 2879 2880A recent change to e2fsck_add_dir_info() to use tdb files to check 2881filesystems with a very large number of filesystems had a typo which 2882caused us to resize the wrong data structure. This would cause a 2883array overrun leading to malloc pointer corruptions and segfaults. 2884Since we normally can very accurately predict how big the the dirinfo 2885array needs to be, this bug only got triggered on very badly corrupted 2886filesystems. 2887 2888Fix a bug in e2fsck which caused it to incorrectly salvange 2889directories when the last entry's rec_len is bogusly too big. This 2890resulted in a nonsense filesystem corruption to be reported, and 2891required a second run of e2fsck to fully fix up the directory. 2892 2893Update tune2fs man page to include more discussion of reserved blocks 2894(Addresses Launchpad bug #47817) 2895 2896Update Turkish, Polish, Dutch, and Vietnamese PO files from the 2897Translation Project 2898 2899 2900E2fsprogs 1.40.1 (July 7, 2007) 2901=============================== 2902 2903Fix bug which could cause libblkid to loop forever. When revalidating 2904a partition where there is obsolete information in /etc/blkid.tab, we 2905end up freeing a the type tag without clearing dev->bid_type, causing 2906blkid_verify() to loop forever. (Addresses Debian Bug: #432052) 2907 2908The Turkish translation has a bug in it where it has the translation 2909of "E@e '%Dn' in %p (%i)" to "E@E". This causes @E to be expanded at 2910@E, recursively, forever, until the stack fills up and e2fsck core 2911dumps. We fix this by making e2fsck stop @-expansions after a 2912recursive depth of 10, which is far more than we need. 2913(Addresses Sourceforge Bug: #1646081) 2914 2915Compile the default mke2fs.conf into mke2fs program. People are 2916getting surprised by mke2fs creating filesystems with different 2917defaults than earlier versions of mke2fs if mke2fs.conf is not 2918present. So we now create a built in version of mke2fs.conf file 2919which is used by mke2fs if the /etc/mke2fs.conf is not present. 2920(Addresses SourceforgeBug: #1745818) 2921 2922Improve the config/parse_types.sh helper script. Fix a potential 2923security problem if e2fsprogs is built as root (as Gentoo does!). In 2924addition fix the script and how it is called from the configure script 2925so that it does the right thing when cross-compiling. (Fixes Gentoo 2926bug: #146903) 2927 2928Update Vietnamese, French, and Dutch PO files from the Translation 2929Project. Also created a new e2fsprogs.pot file for translator. 2930 2931Fix bogus strip permission errors when building under Debian. When 2932building the e2fsprogs dpkg's, the dh_strip command emits a large 2933number of error messages caused by the permissions not being right. 2934So run dh_fixperms before running dh_strip. 2935 2936Programmer's Notes: 2937------------------- 2938 2939Add new function: profile_set_default(). This function sets the value 2940of the pseudo file "<default>". If the file "<default>" had 2941previously been passed to profile_init(), then def_string parameter 2942will be parsed and used as the profile information for the "<default>" 2943file. 2944 2945Fix mk_cmds's error reporting so that it is unambiguous that it is the 2946mk_cmds script which is generating the error. (Obviates Gentoo patch: 2947e2fsprogs-1.32-mk_cmds-cosmetic.patch) 2948 2949Fix the test suite to use LC_ALL instead of LANG. LC_ALL is the "high 2950priority" environment variable that overrides all others, where as 2951LANG is the lowest priorty environment variable. If LC_ALL is set, it 2952doesn't matter whether LANG, LANGUAGE, LC_COLLATE, LC_MESSAGES, and 2953the all the rest are set. This will assure that the locale when 2954running the test suites is the "C" locale. (Obviates Gentoo patch: 2955e2fsprogs-1.38-tests-locale.patch) 2956 2957 2958E2fsprogs 1.40 (June 29, 2007) 2959============================== 2960 2961Fix divide by zero error in blkid's NTFS probing logic. 2962 2963Add new blkid -g option which causes the blkid cache to be garbage 2964collected. 2965 2966Fix a bug in libblkid which could cause the internal field bid_type to 2967become corrupted. Fortunately bid_type isn't used much, and bid_label 2968and bid_uuid is only used by debugging code, so the impact of this bug 2969was very minor. 2970 2971Mke2fs will now store the RAID stride value when a filesystem is 2972created with a requested RAID stride, and then use it automatically in 2973resize2fs. 2974 2975Mke2fs has a sanity check added to make sure (inode_size * num_inodes) 2976isn't too big. In some cases Lustre users have tried specifying an 2977inode size of 4096 bytes, while keeping an inode ratio of one inode 2978per 4096 bytes. 2979 2980Improve sanity check in e2fsck's algorithm for finding a backup 2981superblock, so that it won't accidentally find a superblock that was 2982located in the journal, and then later reject it as being not a valid 2983backup superblock. 2984 2985Fix e2fsck get_size logic so that it will work with the Linux floppy 2986driver. The Linux floppy driver is a bit different from the other 2987block device drivers, in that if the device has been opened with 2988O_EXCL, it disallows another open(), even if the second open() does 2989not have the O_EXCL flag. (Addresses Debian Bug: #410569) 2990 2991Fix error checking of badblock's last-block and start-block arguments. 2992(Addresses Debian Bug: #416477) 2993 2994Fix e2fsck so that it doesn't overwrite the backup superblocks when 2995recovering a journal until the master superblock has been confirmed as 2996being sane. 2997 2998Change the blkid library to be much more paranoid about concluding 2999that a partition contains an NTFS filesystem, and fetch the UUID and 3000LABEL information from NTFS filesystems. (Addresses Launchpad Bug: 3001#110138) 3002 3003Factor out the code which sets the default journal size and move it 3004into libext2fs. 3005 3006Enhance e2fsck so it will recreate the ext3 journal if the original 3007journal inode was cleared to due it being corrupt after finishing the 3008filesystem check. 3009 3010Fix e2fsck so that it updates the journal inode if it is corrupted and 3011the backup journal information from the superblock was successfully 3012used to recover the filesystem. 3013 3014Fix e2fsck so that it checks all of the blocks in the journal inode 3015for validity. The original code only checked the direct blocks to 3016make sure the journal inode was sane. Unfortunately, if some or all 3017of the indirect or doubly indirect blocks were corrupted, this would 3018not be caught. 3019 3020Add support in blkid to detect LUKS encrypted partitions. 3021 3022Add extra sanity checks for extended attributes in the case where the 3023size is zero but the offset is very large. 3024 3025Fix byte-swapping issues for large inodes in ext2fs_read_inode_full() 3026and ext2fs_get_next_inode_full(). 3027 3028Clarify the copyright licenses used by the various libraries in 3029the top-level COPYING file (Red Hat Bugzilla: 166058) 3030 3031Make mke2fs's defaults when /etc/mke2fs.conf doesn't exist more sane. 3032 3033Fix mke2fs and debugfs to support large (> 16 bit) uid's and gid's. 3034 3035Remove check in e2fsck which requires EA's in inodes to be sorted; 3036they don't need to be sorted, and e2fsck was previously wrongly 3037clearing unsorted EA's stored in the inode structure. 3038 3039Allow mke2fs or tune2fs to create a substantially larger journal (up 3040to 10,240,000 blocks). 3041 3042Fix MD superblock detection, and make sure the correct UUID is 3043reported from the MD superblock. 3044 3045Fix a signed vs. unsigned bug in debugfs. 3046 3047Enhance debugfs's date parser so that it accepts integer values. 3048 3049Fix e2fsck's pass1c accounting so it doesn't terminate too early if a 3050file with multiply claimed blocks is hard linked. or not at all if the 3051root directory contains shared blocks 3052 3053Enhance debugfs so it can modify the block group descriptors using the 3054command set_block_group_descriptor. 3055 3056Improve e2fsck's reporting of I/O errors so it's clearer what it was 3057trying to do when an error happens 3058 3059Fix a bug in in how e2fsprogs byte swaps inodes containing fast 3060symlinks that have extended attributes. (Addresses Red Hat Bugzilla: 3061#232663 and LTC Bugzilla: #27634) 3062 3063Fix potential file descriptor leak in ext2fs_get_device_size() in an 3064error case. 3065 3066Add libreadline.so.5 support to libss. 3067 3068Impove badblocks -n/-w exclusive usage message. 3069 3070Fix dump_unused segfault in debugfs when a filesystem is not open 3071 3072Fix memory leak in blkid library. (Addresses Debian Bug: #413661) 3073 3074Allow the debugfs lcd command to work w/o a filesystem being open. 3075(Addresses LTC Bugzilla #27513) 3076 3077Fix e2fsck to clear i_size for special devices with a bogus i_blocks 3078field on the first pass. 3079 3080Fix e2fsck to set the file type of the '..' entry when connecting 3081a directory to lost+found. (Addresses Lustre Bug: #11645) 3082 3083Enhance e2fsck to recover directories whose modes field were corrupted 3084to look like special files. This is probably only useful in 3085artificial test cases, but it will be useful if we ever do the "inodes 3086in directory" idea for ext4. 3087 3088Allow debugfs to dump (and rdump) > 2GB files. (Addresses Debian Bug: 3089#412614) 3090 3091Fix resize2fs parsing of size parameter (in sector units). This was 3092actually a bug in libe2p's parse_num_blocks() function. When handling 3093the 's' suffix, it was ignoring the blocksize information passed in 3094from the caller and always interpreting the number in terms of a 1k 3095blocksize. (Addresses Debian Bug: #408298) 3096 3097There was a floating point precision error which could cause e2fsck to 3098loop forever on really big filesystems with a large inode count. 3099(Addresses Debian Bug: #411838) 3100 3101Fix memory leak in ext2fs_write_new_inode() 3102 3103Add support for using a scratch files directory to reduce e2fsck's 3104memory utilization on really big filesystems. This uses the TDB 3105library. See the [scratch_files] section of the e2fsck.conf man page 3106for more details. 3107 3108Fixed type-punning bug which caused dumpe2fs to crash on the Arm 3109platform (Addresses Debian Bug: #397044) 3110 3111Add explanatory message to badblocks that -n and -w are mutually exclusive 3112(Addresses Debian Bug: #371869) 3113 3114Allow debugfs and dumpe2fs to support fs features under development. 3115 3116Add support for the new flag EXT2_FLAG_SOFTSUPP_FEATURES flag to 3117ext2fs_open() , which allows application to open filesystes with features 3118which are currently only partially supported by e2fsprogs. 3119 3120Allow unix_io to support offsets greater than 2G (Addresses 3121SourceForge Bug: #1547922) 3122 3123Fixed overflow and signed/unsigned problems caused by the number of 3124blocks or inodes exceeding 2**31 or being close to 2**32-1. 3125 3126Add support for unsigned directory hash calculations with hints in the 3127superblock to fix cross-architectural portability for htree 3128directories with filenames where the high 8th bit is set. (Addresses 3129Debian: #389772) 3130 3131Fix resize2fs so that it gives user-intelligible error messages if the 3132filesystem or the kernel does not support on-line resizing. 3133(Addresses Debian Bug: #380548) 3134 3135Require mke2fs -F -F for really dangerous operations, since -F is 3136needed for less dangerous operations such as creating filesystems 3137images in regular files, or creating filesystems on whole block 3138devices. These relatively innocuous usages should NOT be confused 3139with running mke2fs on an apparently-mounted or in-use filesystem. 3140 3141Allow the default inode size to be specified into the mke2fs.conf 3142file. 3143 3144Make the smallest default journal size is big enough so that on-line 3145resizing should always work. 3146 3147Fix silly spelling error in e2fsck. (Addresses SourceForge bug: 3148#1531372) 3149 3150Fix debugfs coredump when lsdel is run without an open filesystem 3151(Addresses Debian Bug: #378335) 3152 3153Fix debugfs display bug us that bytes that have the high bit set are 3154displayed as "ec" instead of "ffffffec". 3155 3156Add support in lsattr so it will display the EXT4_EXTENTS_FL flag. 3157 3158Device mapper scanning wasn't working in the blkid library because the 3159pathnames had an extra "/dev" when they were being probed. 3160 3161Add GFS/GFS2 support to the blkid library. 3162 3163Fix blkid support of empty FAT filesystem labels. 3164 3165Avoid recursing forever (or for a long time) when the blkid library 3166searches for a device and there are symlinks to directories in /dev. 3167 3168Avoid unaligned halfword access in blkid when accessing FAT 3169superblocks, as this will cause Sparc/Solaris systems to throw a 3170SIGBUS error. 3171 3172The latest devmapper libraries requires pthreads, add -lpthreads to 3173the static link libraries for e2fsck.static if devmapper is enabled. 3174(Addresses Debian bug: #388718) 3175 3176Improve the (non-installed, for experts only) findsuper program by 3177printing the uuid and label from the superblocks, as well as the 3178starting and ending offsets of the filesystem given the information in 3179the superblock. Omit by default printing superblocks that are likely 3180found in located in an ext3 journal unless an explicit -j option is 3181given. 3182 3183Updated Spanish, French and Dutch translations and added Catalan 3184translation. (Addresses Debian bug: #411562) 3185 3186Use FreeBSD's DIOCGMEDIASIZE and DIOCGDINFO ioctls if available when 3187determining a partition's size, since binary searching to determine 3188the device doesn't work on FreeBSD. 3189 3190Documentation about UUID's is available in enough places, and it's 3191awkward to deal with debian-legal's insanities. So I'm caving in the 3192"more-lunatic-than-RMS" wing of Debian by removing RFC-4122 so we 3193don't have do the dfsg tarball. Also remove the rule that only tried 3194to install RFC-4122 on Ubuntu, since Ubuntu seems to want to fetch 3195e2fsprogs exclusively from Debian. (Addresses Debian Bug: #407107) 3196 3197Fix the info-dir line so that the menu name does not contain a .info 3198prefix. First of all, it's ugly, secondly, it causes the install-info 3199command to fail to remove the com_err info file from the 3200/usr/share/info/dir file when the comerr-dev package is removed and 3201purged. (Addresses Debian Bug: #401711) 3202 3203Fixed spelling mistakes, typos, and otherwise clarified man pages. 3204(Addresses Debian Bug: #369761, #373004, #379695) 3205 3206Fixed various Debian packaging issues --- see debian/changelog for 3207details. (Addresses Debian Bugs #389554, #390664, #413208, #419605, 3208#408352, #415560, #399155) 3209 3210 3211Programmer's notes: 3212------------------- 3213 3214E2fsck now supports the %It expansion when printing a problem report. 3215It will print the type of the inode in the problem context. 3216 3217Fix misc/Makefile.in so that it builds even if e2fsck hasn't been built yet 3218(Addresses Sourceforge Bug: #1565561) 3219 3220Remove unused variables and other lint/gcc -Wall cleanups 3221 3222Add check to ext2fs_get_device_size() so it will return EFBIG for for 3223filesystems contained in regular files where the filesystem image size 3224is returned by stat64(). 3225 3226Set local environment variables to C so mk_cmds and compile_et always 3227work. (Addresses SourceForge Bug: #1532177) 3228 3229Added the 64-bit byte swapping function ext2fs_swab64(). 3230 3231Added two new helper functions to prevent 2**31/2**32-1 overflow 3232problems: ext2fs_div_ceil() and e2p_percent(). 3233 3234Create new ext2fs library inline functions ext2fs_group_first_block() 3235and ext2fs_group_last_block() in order to calculate the starting and 3236ending blocks in a block group. 3237 3238Create the generated files read-only to remind developers not to edit them. 3239 3240Add support for autoconf 2.60 (with backwards compatibility for older 3241versions of autoconf). 3242 3243Added an "make rpm" target to top-level Makefile 3244 3245Added various FreeBSD portability fixes. 3246 3247Exclude mercurial files from the RPM build tree to speed up copy/build. 3248 3249Use root_sysconfdir to define the locations of mke2fs.conf and 3250e2fsck.conf instead of using a hard-coded /etc pathname. 3251 3252Prevent e2fsck.h and ext2_ext_attr.h from getting included multiple times. 3253 3254Fixed "make clean" in blkid's Makefile.in file from removing tst_*.c files. 3255 3256If diff -u is supported, use it to report test failures. 3257 3258Updates/improvements to RPM spec file 3259 3260Add on-disk format definitions for the following new features: 3261EXT4_FEATURE_RO_COMPAT_HUGE_FILE, EXT4_FEATURE_RO_COMPAT_GDT_CSUM, 3262EXT4_FEATURE_RO_COMPAT_DIR_NLINK, EXT4_FEATURE_INCOMPAT_64BIT, 3263EXT4_FEATURE_RO_COMPAT_EXTRA_ISIZE 3264 3265Add a new make target "checked-failed" in the tests directory which 3266reruns any failed tests 3267 3268Update draft-leach-uuids-guids-01.txt with rfc4122.txt 3269 3270Fix miscellaneous bugs reported by Coverity: Dead code, potential nul 3271pointer dereferences, memory leaks, etc. None were security-criticial 3272problems. 3273 3274Fix up usage and decrement error messages in the test_icount program 3275 3276Add debugging code to the com_err library; if the environment variable 3277COMERR_DEBUG is set to 1, print out debugging messages as error tables 3278are added and removed from the com_err library. If the 3279COMERR_DEBUG_FILE environment variable is set (and the process is not 3280setuid) the debugging messages may be redirected to a file. 3281 3282Change all of the e2fsprogs programs to use the newer add_error_table() 3283and remove_error_table() interfaces instead of the much older 3284initialize_*_error_table() function. 3285 3286Add TDB support into the ext2fs library. This allows us to have a 3287guaranteed library we can count on always being present so we can 3288store data in an on-disk database. 3289 3290Add support for using TDB to store the icount data, so we don't run out 3291of memory when checking really large filesystems. 3292 3293Change the regression test suite so that it skips empty test directories. 3294 3295Define the l_i_iversion field in ext2_inode. The l_i_version field is 3296now defined from the old l_i_reserved1 field in the ext2 inode. This 3297field will be used to store high 32 bits of the 64-bit inode version 3298number. 3299 3300Add Makefile production rule for e2fsprogs.spec in case it gets 3301deleted. 3302 3303Add new function profile_get_uint() to allow for a clean way to fetch 3304unsigned integers from the context. 3305 3306Add test to make sure the ext2 superblock structure is 1024 bytes. 3307 3308Fix typo in name of f_dup4 regression test 3309 3310Add new function blkid_gc_cache() which performs a garbage collection 3311pass on the /etc/blkid.tab file. 3312 3313The ext2fs_open() function now sets EXT2_FLAG_MASTER_SB_ONLY. In 3314general, only e2fsck (after the filesystem is clean), tune2fs, and 3315resize2fs should change the backup superblocks by default. Most 3316callers of ext2fs_open() should not be touching any superblock fields 3317which require the backups to be touched. 3318 3319Add new function to libext2fs, ext2fs_default_journal_size(), which 3320returns the default journal size. 3321 3322 3323E2fsprogs 1.39 (May 29, 2006) 3324============================= 3325 3326Fix 32-bit cleanliness in e2fsprogs so that we can support filesystems 3327between 2**31 and 2**32 blocks. 3328 3329Change mke2fs to use /etc/mke2fs.conf as a configuration file to 3330configure the filesystem features, blocksize, and inode_ratio for 3331different filesystem types. 3332 3333Mke2fs will now create filesystems hash trees and on-line resizing 3334enabled by default, based on the new /etc/mke2fs.conf file. 3335 3336The e2fsprogs tools (resize2fs, e2fsck, mke2fs) will open the 3337filesystem device node in exclusive mode to prevent accidents by 3338system administrators. In the case of resize2fs and mke2fs, it will 3339only use exclusive mode if the filesystem is not mounted. 3340 3341Fixed a bug in mke2fs which caused it to to fail when creating the 3342resize inode for large filesystems. (Addresses Debian Bug #346580) 3343 3344When allocating space for the RAID filesystems with the stride 3345parameter, mke2fs will now place each portion of the group's inode 3346table right up after the superblock (if present) in order to minimize 3347fragmentation of the freespace. 3348 3349Speed up mke2fs and e2fsck by writing inode and block bitmaps more 3350efficiently by writing the inode and block bitmaps in one pass, thus 3351reducing the number of disk seeks required. 3352 3353Add support for on-line resizing to resize2fs. 3354 3355Fix blkid library so that logic to determine whether or not a device's 3356cached information in /etc/blkid.tab needs to be verified or not 3357doesn't get confused by a system clock which is insane (for example, 3358if the battery is dead on a Macintosh running PPC Linux. (Addresses 3359Red Hat Bug: #182188) 3360 3361The blkid library will now store the UUID of the external journal used 3362by ext3 filesystems, so that in the future, the userspace mount binary 3363can use this to find the location of the external journal and pass 3364this information to the kernel. 3365 3366E2fsck will now consult a configuration file, /etc/e2fsck.conf to 3367control how various options should be handled. See the e2fsck.conf 3368man page for more details. (Addresses Debian Bug: #150295) 3369 3370E2fsck now prints an explanatory message when delaying a filesystem 3371check when the system is running on battery. (Addresses Debian Bug: 3372#350306) 3373 3374E2fsck will detect if the superblock's last mount field or last write 3375field is in the future, and offer to fix if so. (Addresses Debian Bug 3376#327580) These problems will be fixed automatically in preen mode 3377since Debian's boot sequence bogusly doesn't set the time correctly 3378until potentially very late in the bootup process, and this can cause 3379false positives which will cause users' systems to fail to boot. 3380(Addresses Debian Bugs #343662 and #343645) 3381 3382E2fsck now checks to see if the superblock hint for the location of 3383the external journal is incorrect, and if so, offer to update it. 3384(Addresses Debian Bug: #355644) 3385 3386Fix e2fsck from segfaulting on disconnected inodes that contain one or 3387more extended attributes. (Addresses Debian Bug: #316736, #318463) 3388 3389E2fsck will stop and print a warning if the user tries running a 3390read/write badblocks test on a read-only mounted root filesystem. 3391 3392Fix a memory leak in e2fsck's error paths. (Thanks to Michael 3393C. Thompson for pointing these out; they were originally found using 3394Coverity.) 3395 3396When resizing a file containing a filesystem, resize2fs will expand or 3397truncate a file as necessary. (Addresses Debian Bug: #271607) 3398 3399Resize2fs will now automatically determine the RAID stride parameter that 3400had been used to create the filesystem, and use that for newly created 3401block groups. The RAID stride parameter may also be manually specified 3402on the command line using the new -S option to resize2fs. 3403 3404Fix mke2fs so that it correctly creates external journals on 3405big-endian machines (such as a S/390). 3406 3407Fix a bug in the e2p library which could cause dumpe2fs to (rarely) 3408fail to print out the journal or hash seed UUID. (Thanks to Guillaume 3409Chambraud for pointing this out.) 3410 3411Dumpe2fs will now print the size of the journal (if present). 3412 3413Fix debugfs's set_inode_field command so it can properly set the frag, 3414fsize, uid_high, gid_high, and author fields in the inode instead of 3415silently failing, and so that setting the i_size actually sets i_size 3416correctly. 3417 3418Add a new debugfs command, set_current_time, which sets fs->now so 3419that regression test suites can repeatedly modify the filesystem's 3420last_write fields. 3421 3422Fix a bug in debugfs's icheck which would incorrectly report the owner 3423of an extended attribute block. 3424 3425Fix the debugfs commands htree_dump, dx_hash, and list_dir so they print a 3426print a usage message when an illegal option character is given. 3427 3428Fix debugfs's dump_unsued command on filesystems with a 64k blocksize 3429so it won't core dump. (Addresses SourceForge bug #1424311) 3430 3431Fix mklost+found so that it creates a full-sized directory on 3432filesystems with larger block sizes. 3433 3434Fix a file descriptor leak in blkid library. 3435 3436Fix a display bug in "badblocks -sv" so that the done message properly 3437clears the block number at the end of the test. (Addresses Debian Bug 3438#322231) 3439 3440Allow fractional percentages to the -m option in mke2fs and tune2fs 3441(Addresses Debian Bug: #80205) 3442 3443Use fstat/fstat64 in getsize.c if the the target is a regular file, 3444instead of attempting to do a binary search. Fix some fd leaks in 3445error cases. 3446 3447Add support for device mapper library to the blkid library to ensure 3448that the "best" (i.e., leaf) device is probed by the blkid library. 3449 3450Fix the blkid library so that it notices when an ext2 filesystem is 3451upgraded to ext3. 3452 3453Improve the blkid's library VFAT/FAT detection; it now understands 3454labels stored in the root directory, and is more paranoid about 3455checking the FAT superblock values. 3456 3457Fixed a fd leak in the uuid library which was causing problems for the 3458LVM tools. (Addresses Debian Bug: #345832) 3459 3460Add support for the reiser4 and software suspend partitions to the 3461blkid library. Also add support for extract the label from iso9660 3462filesystems. 3463 3464Fix a compile_et bug which miscount the number of error messages if 3465continuations are used in the .et file. 3466 3467Add extra sanity checks to protect users from unusual cirucmstances 3468where /etc/mtab may not be sane, by checking to see if the device is 3469reported busy (works on Linux 2.6) kernels. (Addresses Debian Bug 3470#319002) 3471 3472Updated French, Dutch, Polish, and Swedish translations. (Addresses Debian 3473Bug: #343149, #341911, #300871, #316604, #316782, #330789) 3474 3475Fix use-after-free bug in e2fsck when finishing up the use of the 3476e2fsck context structure. 3477 3478Fixed spelling mistakes, typos, and otherwise clarified man pages and 3479documentation. (Addresses Debian Bugs: #329859, #322188, #316811, 3480#312515, #351268, #357951, #347295, #316040, #368392, #368393, #368394, 3481#368179) 3482 3483Fixed various Debian packaging issues --- see debian/changelog for 3484details. (Addresses Debian Bugs #317862, #320389, #290429, #310950, 3485#310428, #330737, #330736, #329074, #356293, #360046, #366017, #364516, 3486#362544, #362970) 3487 3488 3489Programmer's notes: 3490------------------- 3491 3492Update config.guess and config.sub to latest version (2006-02-23) from 3493FSF. 3494 3495Fix asm_types.h type conflicts on AMD 64 platforms. (Addresses Debian 3496Bugs: #360661, #360317) 3497 3498Fixed the Makefile so that they work correctly on newer versions of 3499GNU make (i.e., 3.81). 3500 3501Add valgrind support to the regression test suites, and eliminate 3502false positives from valgrind. 3503 3504Add a regression test suite for the blkid library. 3505 3506Fix a fencepost error in resize2fs caught by valgrind. 3507 3508Fix compiler warnings about missing memcpy prototypes. 3509 3510We no longer have the sparc assembly code in the header file any more, 3511so we shouldn't set _EXT2_HAVE_HAS_BITOPS_ for the sparc. This would 3512break compiles on the sparc architectures when using gcc. 3513 3514In the libext2fs library, add the new field fs->now which if non-zero 3515is used instead of the system time when setting various filesystem 3516fields (last modified time, last write time, etc.) 3517 3518Fix gcc 4.01 complaints by adding a missing #include <string.h> to 3519ext2fs.h which is needed since the inline functions use memcpy(). 3520(Addresses Sourceforge Bug #1251062) 3521 3522Use BUILD_CFLAGS and BUILD_LDFLAGS instead of CFLAGS and LDFLAGS in 3523the build system when building files in the util directory which are 3524needed during the build process. This avoids potential problems when 3525cross-compiling and some of the options specified in CFLAGS or LDFLAGS 3526are not recognized as valid by the host compiler. (Addresses 3527Sourceforge Bug #1261547) 3528 3529Clean up the blkid library by making the superblock and generic i/o 3530functions to be more generic. Clean up interface to the probe 3531function, and fix memory leak. Finallly, remove an unneeded reference 3532to probe.h in the lib/blkid/resolve.c 3533 3534Add an ext2fs_read_bb_FILE regression test to confirm proper detection 3535of invalid block #'s. 3536 3537The x86 asm constraints for ext2fs_{set/clear}_bit have been fixed to 3538indicate that the the function read/writes the memory location. 3539 3540Fix various gcc -Wall complaints. 3541 3542Add a dependency to make sure that the subdirectories are created 3543before creating all of the object files, in order to address parallel 3544build problem in the library Makefiles. (Addresses Sourceforge Bug: 3545#1261553) 3546 3547Add $(LDFLAGS) to the command line argument when generating an ELF or 3548Solaris shared library, to allow cross-compile and other builds that 3549might need to specify -L paths to needed libraries. (Addresses 3550Sourceforge Bug #1261549) 3551 3552Add a new feature, EXT2_FEATURE_COMPAT_LAZY_BG, which is initially 3553intended for testing purposes. It allows an ext2/ext3 developer to 3554create very large filesystems using sparse files where most of the 3555block groups are not initialized and so do not require much disk 3556space. Eventually it could be used as a way of speeding up mke2fs and 3557e2fsck for large filesystem, but that would be best done by adding an 3558RO_COMPAT extension to the filesystem to allow the inode table to be 3559lazily initialized on a per-block basis, instead of being entirely 3560initialized or entirely unused on a per-blockgroup basis. 3561 3562Fix backwards compatibility so e2fsprogs will better compile on Linux 35632.0.35 systems. 3564 3565Make test scripts more robust against locale-related environment variables 3566 3567Fix type warning problem with sizeof() in ext2fs_open2(). 3568 3569Fix type warning problem with time_t in debugfs. 3570 3571 3572E2fsprogs 1.38 (June 30, 2005) 3573============================== 3574 3575Fix blkid's test programs (built with "make check") compile correctly 3576even without "configure --enable-blkid-debug". 3577 3578Fix ia64 core dump bug caused by e2fsprogs running afoul of C99 strict 3579type aliasing rules on newer gcc compilers. (Addresses Red Hat 3580Bugzilla ##161183.) 3581 3582Fix com_err library to make it more compatible with recent changes 3583made to the com_err library in MIT Kerberos V5 version 1.4. 3584(Addresses Sourcefroge Bug #1150146) 3585 3586General cleanup of messages printed by e2fsprogs programs for grammar, 3587consistency, and to make life easier for translators. Fixed a few 3588strings containing English that had not been marked as needing 3589translations. Removed strings that do not need to be translated, to 3590make life easier for translators. 3591 3592Mke2fs and badblocks will take advantage of a feature in Linux 2.6 to 3593test to see if a device appears to be in use instead of just relying 3594on /proc/mounts and /etc/mtab. (Addresses Debian Bug #308594). 3595 3596Fix portability problem in the filefrag program affecting platforms 3597where the size of an integer is smaller than the size of a long. 3598(Addresses Debian Bug #309655) 3599 3600Mke2fs will now use a larger journal by default for filesystems 3601greater than 4GB. (128 MB instead of 32MB). 3602 3603Mke2fs will refuse to create filesystems greater than 2**31-1 blocks, 3604unless forced. This is to avoid signed vs. unsigned kernel bugs in 3605block numbers that still need to be fixed. 3606 3607The blkid program has a new option which will more efficiently search 3608for device when it is known (or expected) that only one matching 3609device will be found in the system, such as when doing a lookup by 3610UUID. 3611 3612Debian's e2fsprogs-specific initrd fragment will avoid including 3613unnecessary libraries into the initrd ramdisk by unsetting LD_PRELOAD 3614and LD_LIBRARY_PATH, and filtering out libraries found in 3615/etc/ld.so.preload. (Addresses Debian Bug: #304003) 3616 3617Fixed a potential portability issue in the blkid programs for 3618architectures where the char type is unsigned. (Addresses Sourceforge 3619Bug: #1180585) 3620 3621Fix a bug in filefrag so that it doesn't falsely count an extra 3622discontinuity when the first block found is an indirect block. 3623(Addresses Debian Bug #307607). 3624 3625Fix blkid's recognition of cramfs filesystems, and enhance it to be 3626able to handle cramfs labels. 3627 3628Fix debugfs's stat command to not core dump when a filesystem is not 3629open. 3630 3631Fix e2fsck's handling of error conditions caused by the resize inode 3632claiming blocks that are also used by other inodes, a filesystem 3633corruption which was commonly caused by a bug in Fedora Core 3's 3634resize2fs program. 3635 3636Fixed bug in filefrag which caused it to fail on non-ext2/3 3637filesystems. (Addresses Debian Bug: #303509) 3638 3639If the superblock last mount time indicates that the system clock may 3640not be accurate, then e2fsck will omit checking inodes' deletion time 3641field for indications of a potential corrupted orphaned inode list. 3642(Previously e2fsck only ommited these LOW_DTIME checks when the 3643superblock's last write time looked insane.) 3644 3645Fixed a IA64 core dump bug in the e2p library which affected dumpe2fs. 3646(Addresses Debian bug #302200) 3647 3648Make the blkid library more paranoid about being run from setgid 3649programs, and to use __secure_getenv() from libc if it is available. 3650 3651Fixed spelling mistakes, typos, and otherwise clarified man pages. 3652(Addresses Debian Bugs: #304591, #304592, #304594, #304597, #304593 3653and Sourceforge Bug: #1189803) 3654 3655Updated and fixed translations. 3656 3657Fixed various Debian packaging issues --- see debian/changelog for 3658details. 3659 3660Programmer's notes: 3661------------------- 3662 3663Ext2fs_set_bit(), ext2fs_clear_bit(), and ext2fs_test_bit( have been 3664changed to take an unsigned int for the bit number. Negative bit 3665numbers were never allowed (and didn't make any sense), so this should 3666be a safe change. This is needed to allow safe use of block numbers 3667greater than or equal to 2**31. 3668 3669The compile_et program will avoid recreating generated foo_err.c and 3670foo_err.h files if no changes are necessary. The compile_et program 3671will also atomically replace these files to avoid a potential parallel 3672build race problem on SMP systems. (Addresses Sourceforge Bug: 3673#1157933) 3674 3675Added a new function to the blkid library, blkid_probe_all_new(), 3676which only probes newly added disk devices, and change 3677blkid_find_dev_with_tag() to use this function so that when a 3678requested tag is not found, devices that were previously not checked 3679are searched before searching all devices in the system. 3680 3681Added new functions to the blkid library, blkid_dev_set_search() and 3682blkid_dev_has_tag(). 3683 3684E2fsck's problem strings can now use @m and @n as abbrevations for 3685"multiply-claimed" and "invalid", respectively. 3686 3687The e2fsprog.pot file now has an explanation of how the @-expansion 3688and %-expansion works, and strings in e2fsck/problem.c which contain @ 3689characters now have comments in e2fsprogs.pot with the @-expansion to 3690make life easier for translators. 3691 3692Fixed missing return values in the ext2fs library which could cause it 3693to return random garbage in certain error conditions. 3694 3695Allow the current time to be overriden via the E2FSCK_TIME environment 3696variable for use in regression tests. 3697 3698The test scrpit driver program now exits with a non-zero status if 3699there any of its test that it ran failed. 3700 3701Fixed problems with parabuilds on SMP systems. (Addresses Sourceforge 3702Bug: #1157933) 3703 3704Fixed "make check" so that it compiles correctly even when e2fsprogs' 3705header files have not be installed in the system include directories. 3706(Addresses Sourceforge Bug: #1180572) 3707 3708Fixed gcc -Wall nits. 3709 3710 3711E2fsprogs 1.37 (March 21, 2005) 3712=============================== 3713 3714Add support for checking the validity of Extended Attributes stored in 3715inodes to e2fsck. 3716 3717Add support for dumping the contents of large inodes to debugfs, 3718including the extended attributes stored in inodes. 3719 3720Fix mke2fs, e2fsck, debugfs, and the ext2fs_mkdir function so that 3721when we create a new inode we make sure that the extra information in 3722the inode (any extra fields in a large inode and any ea-in-inode 3723information) is initialized correctly. This can take place when 3724mke2fs creates the root and lost+found directory, when e2fsck creates 3725a new root inode or a new lost+found directory, and when the user uses 3726the debugfs write, mknod, or mkdir commands. Otherwise, the newly 3727create inode could inherit garbage (or old EA information) from a 3728previously deleted inode. 3729 3730Fixed a bug in e2fsck so it would notice if a file with an extended 3731attribute block was exactly 2**32 blocks, such that i_blocks wrapped 3732to zero. 3733 3734Added support to filefrag to detect files which are using the new 3735experimental file extents format, and use the non-ext2 algorithm in 3736that case. Fixed a bug to avoid reporting a false discontinuity if 3737there is one or more unallocated blocks at the beginning of a file. 3738 3739Duplicated a check for noticing whether or not the number of blocks 3740(given a certain blocksize) is greater than 2**32 when the 3741BLKGETSIZE64 ioctl is not available to ext2fs_get_device_size(). This 3742allows mke2fs to automatically use a larger blocksize when creating a 3743filesystem on a very large device when run on systems that do not 3744support BLKGETSIZE64. 3745 3746Fix the I18N build which was broken in e2fsprogs 1.36 because the 3747build system had been switched to treat the .gmo files as shipped 3748files (for backwards compatibility with systems that have older GNU 3749I18N tools installed), but the gen_tarball.in script was still 3750removing the .gmo files from the official source distribution. 3751 3752Fixed various Debian packaging issues --- see debian/changelog for 3753details. (Addresses Debian Bugs ##296769, #299341) 3754 3755Programmer's notes: 3756------------------- 3757 3758Added new functions to the e2p library which convert between a string 3759and os_type: e2p_os2string() and e2p_string2os(), and used them to 3760make the generated binaries more compact. 3761 3762Fixed a compile-time error on Darwin systems. 3763 3764Cleaned up the lib/ext2fs Makefile slightly. 3765 3766 3767E2fsprogs 1.36 (February 4, 2005) 3768================================= 3769 3770All of the patches that were applied to Fedore Core 3's 3771e2fsprogs-1.35-11.2 have been integrated, although sometimes with a 3772lot of bug fixes first. Users of Fedora Core 3 are strongly 3773encouraged to upgrade to e2fsprogs 1.36 as soon as possible. 3774 3775Add support for filesystem with the online resizing via resize inode 3776feature. Fixed numerous bugs from the Fedora patches. The Fedora 3777patches also didn't bother to do any consistency checking on the 3778resize inode, or add any tests to the regression test suite. The "-R 3779resize=4g" option to mke2fs was a no-op in the Fedora patches, despite 3780being listed in mke2fs's usage message. All of these shortcomings 3781have been corrected. 3782 3783E2fsck can also also fix filesystems trashed by Fedora's resize2fs 3784program. In order to do this, the user must run the commands: 3785 3786 debugfs -w /dev/hdXXX -R "features ^resize_inode" 3787 e2fsck -f /dev/hdXXX 3788 3789Optionally, the ext2prepare command can be used to re-enable online 3790resizing after the filesystem has been fixed. 3791 3792The fsck program will now accept an optional filedescriptor argument 3793to the -C option. (The Fedora version of this patch would sometimes 3794cause fsck to ignore a parameter on fsck's command line in some rare 3795cases, sigh.) 3796 3797Make sure e2fsprogs doesn't write garbage into the reserved portion of 3798large inodes. 3799 3800Make sure resize2fs releases the blocks belonging to the old inode 3801table blocks when moving the inode table. (Addresses Debian Bug: 3802#290894) 3803 3804Skip the r_resize_inode test if resize2fs is not compiled (due to 3805configure --disable-resizer) 3806 3807E2fsck now checks the summary filesystem accounting information, and 3808if any of the information is obviously wrong, it will force a full 3809filesystem check. (Addresses Debian Bug #291571) 3810 3811Fix e2fsck to not complain when the resize_inode feature is enabled, 3812s_reserved_gdt_blocks is zero, and there is no DIND block allocated in 3813the resize inode. 3814 3815Fix e2fsck to note delete symlinks that contain an extended attribute 3816after the ext_attr feature flag has been cleared. (Addresses Red Hat 3817Bugzilla #146284). 3818 3819Add new utility program, copy_sparse.c, which is very useful 3820for dealing with large sparse files (such as e2image files). 3821 3822Add support for jnl_blocks[] for debugfs's set_super_value. 3823 3824Fix filefrag so that it works correctly with sparse files. 3825 3826Filefrag -v will print first and last blocks. 3827 3828Add interpretation of OS Creator values for FreeBSD and Lites in mke2fs 3829and dumpe2fs. 3830 3831Add mke2fs support so that it can support filesystems larger than 4TB 3832automatically, by retrying with a 4k blocksize if the device size is 3833too big to be expressed using a 1k blocksize. (Addresses Sourceforge 3834bug #1106631) 3835 3836Change blkid to test for NTFS first because Windows sometimes doesn't 3837clear enough of the parition to confuse the probing routines into 3838thinking the old filesystem type is still valid. (Addresses Debian 3839Bug #291990) 3840 3841Add support for swap partition label and uuid's in the blkid library. 3842 3843Add support to the blkid library to recognize Oracle ASM volumes. 3844 3845Make blkid -t display all devices that match the specified criteria, 3846not just the first one, and work more consistently when the blkid 3847cache file is not available or set to /dev/null. (Addresses Debian 3848Bug #290530 and #292425) 3849 3850Badblocks will now correctly display block numbers greater than 3851999,999,999 in its progress display. 3852 3853The tune2fs program will not allow the user from setting a ridiculous 3854number of reserved blocks which would cause e2fsck to assume the 3855superblock was corrupt. E2fsck's standards for what is a ridiculous 3856number of reserved block has also been relaxed to 50% of the blocks in 3857the filesystem. 3858 3859The blkid library will return vfat in preference to msdos, and ext3 in 3860preference to ext2 (if the journalling flag is set) so that mount will 3861do the right thing. (Addresses Debian bug #287455) 3862 3863Mke2fs will now use the -E option for extended options; the old -R 3864(raid options) option is still accepted for backwards compatibility. 3865Fix a double-free problem in resize2fs. (Red Hat Bugzilla #132707) 3866 3867Mke2fs will now accept a size in megabytes, gigabytes, and other units 3868(via "32m" or "4g" on the command line) if the user finds this more 3869convenient than specifying a block count. 3870 3871Fix an obscure, hard-to find bug in "e2fsck -S" caused by an inode 3872cache conherency problem. 3873 3874Debugfs now supports a new command, set_inode_field, which allows a 3875user to manually set a specific inode field more conveniently, as well 3876as set entries in the indirect block map. 3877 3878Debugfs's set_super_value command has been enhanced so that the user 3879can set most superblock fields, including the date/time fields and 3880some of the more newsly added superblock fields. 3881 3882E2fsprogs programs now accept an offset to be passed to the file 3883specifiers, via the syntax: "/tmp/test.img?offset=1024". 3884 3885E2fsprogs programs will now accept blocksizes up to 65536; kernel 3886support on the x86 doesn't exist for now, but it can be useful on 3887other architectures with page sizes greater than 4k. There are 2.6 3888kernel patches out there which enable this, but they are of this 3889writing still experimental. 3890 3891The e2image command now takes the -s option which will scramble 3892directory entries for raw image files. 3893 3894Fix a file descriptor leak in the filefrag program. 3895 3896Make sure e2fsck doesn't crash when /proc/acpi/ac_adapter is not 3897present. 3898 3899Fix bug in debugfs where kill_file would lead to errors when deleting 3900devices and symlinks. (Sourceforge Bugs #954741 and #957244) 3901 3902Fix bug in the blkid library when detecting the ocfs1 filesystem 3903 3904Remove obsolete EVMS 1.x and a.out DLL support. 3905 3906E2fsck will attempt to recover from a journal containing illegal blocks. 3907 3908Fixed two potential ordering constraint problems in e2fsck which might 3909cause the filesystem to be corrupted if e2fsck is interrupted during a 3910(extremely narrow) race window. Thanks to Junfeng Yang from the 3911Stanford Metacompilation group for pointing this out. 3912 3913Fixed bug in e2fsck where it would not accurately detect whether or 3914not the system is running on adaptor if the ACPI device representing 3915the AC adapter didn't correspond to the what was used on IBM 3916Thinkpads. 3917 3918Change e2fsck to accept directories greater than 32MB. 3919 3920Fix e2fsck so that a checkinterval of zero disables a time-based check 3921of the filesystem. 3922 3923Debugfs will check the DEBUGFS_PAGER enviroment variable in preference 3924to the PAGER environment variable. (Addresses Debian Bug #239547) 3925 3926Tune2fs will not mark rewrite the superblock if the feature bitmasks 3927are not modified. 3928 3929The debugfs program will set the filetype information when creating a 3930link. 3931 3932Add debugfs -d option to use a separate source of data blocks when 3933reading from an e2image file. 3934 3935Add e2image -I option which allows the e2image metadata to be 3936installed into a filesystem. 3937 3938Fixed bug in the badblocks program which caused "done" to always 3939appear in english even when a translation was available. (Addresses 3940Debian Bug #252836) 3941 3942The blkid program has a new option -o which controls the output format 3943of the blkid program; this is makes blkid more convenient to use in 3944shell scripts. 3945 3946Fix a minor bug in uuid library, which was not using the full 14 bits 3947of clock sequence when generating UUID's. 3948 3949Fix a Y8.8888K problem in the uuid library. 3950 3951Logsave now creates a new session id for itself to avoid getting 3952killed by init whan transitioning between init levels. 3953 3954Change the licensing of the UUID library to be the 3-clause BSD-style 3955license; this allows Apple to use the uuid library in Darwin. 3956 3957Add ocfs and ocfs2 probe support into the blkid library. 3958 3959Fix a memory and file descriptor leak in the blkid library. 3960 3961The blkid library will revalidate the device if the system time is 3962earlier than last verification time of the device, since that 3963indicates that the system time is probably nottrustworthy. 3964 3965The blkid library will override the default location of the blkid.tab 3966file by the BLKID_FILE environment variable, if it is available. 3967 3968Change the getsize functions to use the BLKGETSIZE64 ioctl on Linux 2.6. 3969 3970Add various portability fixes for lame new versions of glibc, Darwin 3971and GNU/KFreeBSD, as well as removing XSI:ism's. (Addresses Debian 3972Bugs #239934, #264630, #269044, #255589, #289133) 3973 3974Add support for Windows 9x/NT under Cygwin. 3975 3976Updated and clarified various man pages. (Addresses Debian Bugs #236383, 3977 #241940, #238741, #242995, #256669, #268148, #256760, #273679) 3978 3979Updated and fixed translations. (Addresses Debian bugs #244105, #262836) 3980 3981Update the rpm spec files so that it works better with Fedora core 2 3982and RH9. 3983 3984Fixed various Debian packaging issues (see debian/changelog). In 3985particular, fixed the Debian initrd scripts. (#241183, #248050, 3986#253595, #247775) 3987 3988 3989Programmer's notes: 3990------------------- 3991 3992Fixed various gcc -Wall warnings. 3993 3994The uuid library now has new functions uuid_unparse_upper() and 3995uuid_unparse_lower() which forces the case of the hex digits to be 3996upper case, or lower case. 3997 3998The build process has been speeded up by enhancing the subst program 3999to update the modtime on the generated files even when the generated 4000file hasn't changed. 4001 4002The uuid library now uses C99 stdint.h types instead of custom types. 4003 4004Updated config.guess and config.sub with newer versions from the FSF. 4005 4006Removed out of date .cvsignore files from the source distribution. 4007 4008The ext2fs_unlink() function will return an error if both the name and 4009inode number are unspecified, to avoid doing something surprising 4010(such as unconditionally deleting the first directory entry). 4011Directory entries are now deleted by coalescing them with the previous 4012directory entry if possible, to avoid directory fragmentation. This 4013is not an issue with the e2fsprogs suite, but may be a problem for 4014some of the users of libext2fs, such as e2tools. 4015 4016Add support for version numbers of the form "1.36-rc1". 4017 4018Fix build of mke2fs.static. 4019 4020Add basic ext2fs library support for large (EA in inode) inodes. 4021 4022The test_io mechanism can now abort after n reads or writes to a 4023particular block. The block is specified by TEST_IO_BLOCK environment 4024variable, and the read/write count by the TEST_IO_READ_ABORT and 4025TEST_IO_WRITE_ABORT environment variables. The block data is now only 4026dumped if the 0x10 bit is set in TEST_IO_FLAGS. 4027 4028UUID_DEFINE() in the uuid library now creates a static variable, with 4029__attribute__ ((unused)) if we are using GCC, so that UUID_DEFINE can 4030be used in header files. 4031 4032Add support for the install-strip and install-shlibs-strip targets, as 4033suggested by the GNU coding guielines. "make install" no longer 4034strips the binaries which are installed. 4035 4036Remove support for the --enable-old-bitops configure option which was 4037only for very old sparc systems. 4038 4039Remove support for --enable-clear-htree; this was only needed during 4040the early development of the htree patch. 4041 4042Use Linux-kernel-style makefile output so it is easier to see compiler 4043warnings. 4044 4045Update gettext files to version 0.14.1. 4046 4047Update to use autoconf 2.5x. 4048 4049Improved support for compiling e2fsprogs under dietlibc. 4050 4051Make e2fsprogs portable to Solaris and FreeBSD systems. 4052 4053Add blkid_verify(), blkid_get_library_version(), and 4054blkid_parse_version_string() functions to the blkid library. 4055 4056Add pkg-config files for e2fsprogs's libraries. 4057 4058Fix "make uninstall" to so that it removes everything that is installed. 4059 4060Add a configure --enable-maintainer-mode option which enables the 4061makefile rules to rebuild the configure script from configure.in, and 4062to reubuild the .gmo files in po directory. 4063 4064Drop the sparc assembly bitwise operations; it's less efficient 4065than the GCC 3.4 compile code and triggers compiler warnings on 4066sparc64. Thanks to Matthias Andree for his analysis and suggestions. 4067(Addresses Debian Bug #232326) 4068 4069 4070E2fsprogs 1.35 (February 28, 2004) 4071================================== 4072 4073E2fsck has a new -k option, which in conjunction with the -c options, 4074preserves the existing badblocks list. 4075 4076Cleaned up e2fsck's preen-mode messages during the passes 1b, 1c, and 1d. 4077 4078E2fsprogs will now deal correctly with symlinks that contain 4079extended attribute information, which can be created using SE Linux. 4080(Addresses Debian Bug #232328) 4081 4082Remove a double longjump into an invalid stack frame bug in e2fsck. 4083(This was during an abort sequence, which normally worked on Linux and 4084caused a core dump on other operating systems.) 4085 4086Fix NLS bug in e2fsck, by avoiding trying to expand an empty string 4087(the NLS library will replace "" with the .po header information). 4088 4089Fix a bug in mke2fs which caused -T largefile or -T largefile4 to core 4090dump due to a division by zero error. (Addresses Debian bug #207082) 4091 4092Fixed a bug in e2fsck which caused it to incorrectly fix a filesystem 4093when reconnecting a directory requires creating a lost+found 4094directory. (Addresses Debian bug #219640). 4095 4096Fixed a bug where e2fsck would bomb out if a journal needed to be 4097replayed when using an alternate superblock. 4098 4099E2fsck will give an extra grace period before actually forcing a check 4100if the laptop is running on battery. The next time fsck runs while 4101the system is on the AC mains, or after the grace period is exceeded, 4102the filesystem will be checked. (Addresses Debian bug #205177) 4103 4104E2fsck will inform the user when there are 5 or fewer mounts before a 4105filesystem check will be forced. (Addresses Debian bug #157194) 4106 4107Fix e2fsck's handling of corrupted indirect blocks in the bad block. 4108We now correctly handle the case where there is an overlap between a 4109block group descriptor or a superblock and a bad block indirect block. 4110In the case where the indirect block is corrupted, we now suggest 4111"e2fsck -c". 4112 4113Fix byte swap bugs in e2fsck that caused the journal backup location 4114in the superblock and symlinks created by SE Linux to be cleared 4115by e2fsck on big-endian machines. (Addresses Debian bug #228723) 4116 4117E2fsck -c now replaces the current list of bad blocks with the ones 4118found by badblocks. 4119 4120Fix bugs in e2fsck and tune2fs which could cause a core dump if a 4121non-existent LABEL or UUID specifier is to e2fsck or tune2fs. 4122 4123Fix a potential bug in e2fsck which could cause it to core dump when 4124trying to print the location of the backup superblock. 4125 4126Protect against a potential core dump in e2fsck when printing a 4127message about backup superblocks. 4128 4129Add support for backing up the journal inode location in the 4130superblock. E2fsck will automatically save the journal information in 4131the superblock if it is not there already, and will use it if the 4132journal inode appears to be corrupted. ext2fs_add_journal_inode() 4133will also save the backup information, so that new filesystems created 4134by mke2fs and filesystems that have journals added via tune2fs will 4135also have journal location written to the superblock as well. 4136Debugfs's logdump command has been enhanced so that it can use the 4137journal information in the superblock. 4138 4139E2fsck will now update all superblocks when moving the journal inode. 4140 4141Shrink the size of the e2fsck executable by moving some initialized 4142variables to the BSS segment. 4143 4144E2fsck will avoid printing the ^A and ^B characters which bracket the 4145progress bar when stdout and stdin are a tty device instead of a pipe 4146to another program. (Addresses Debian bug #204137) 4147 4148Debugfs's mkdir command will automatically expand the directory if 4149necessary. (Addresses Debian Bug: #217892) 4150 4151Fixed a bug in debugfs so that copying a file from /dev/null uses the 4152correct mode bits. (Addresses Debian Bug: #217456) 4153 4154If the environment variables DEBUFS_PAGER and PAGER are not set, 4155debugfs now searches for the appropriate pager to use, beginning with 4156/usr/bin/pager, and then falling back to 'more' and 'less'. 4157(Addresses Debian bug #221977) 4158 4159Debugfs will now support 2.6 device numbers where the major or minor 4160number may be larger than 255. (Addresses Sourceforge bug #865289) 4161 4162Fix debugging printf in resize2fs. (Addresses Debian Bug #271605) 4163 4164Chattr now stops processing options when it sees '--'. (Addresses 4165Debian bug #225188) 4166 4167Fix regression tests so they work correctly when e2fsprogs is compiled 4168with configure --disable-htree. 4169 4170Fix bug in uuid library when there is no network card and the library 4171is generating a time-based uuid. The random MAC address was not 4172correctly generated to be a multicast address. 4173 4174Add compile_et extensions from Heimdall that were missed the first time 4175around. 4176 4177Fix bug in badblocks when using O_DIRECT; we need to make sure that 4178we're reading from an offset which is page aligned. For read-only and 4179read-write tests, we try to recover after an error so that we can 4180continue reading on page-aligned boundaries. (Addresses Debian Bug 4181#203713) 4182 4183Badblocks now checks 64 blocks at a time instead of 16. (Addresses 4184Debian bug #232240) 4185 4186Updated and clarified various man pages. (Addresses Debian Bug 4187#206845, #222606, #214920, #232406) 4188 4189Updated and fixed translations. (Addresses Debian bugs #200086, #214633) 4190 4191Fixed various Debian packaging issues (see debian/changelog). 4192 4193Programmer's notes: 4194------------------- 4195 4196Fixed a build problem so that e2fsprogs would compile with the 4197--enable-profile option to configure selected. (Addresses Sourceforge 4198bug #811408) 4199 4200Fixed C++ problems with the ext2fs.h header. (Addresses Red Hat 4201Bugzilla Bug #112448) 4202 4203Centralize code which calculates the location of the superblock 4204and block group descriptors so that it is in a single library routine. 4205 4206Added two new functions, ext2fs_file_open2() and 4207ext2fs_inode_io_intern2() which take a pointer to an inode structure. 4208 4209Fix compile_et to output the correct prototype for 4210initialize_xxx_err_table_r() in the header file. (Addresses Debian 4211bug #204332) 4212 4213In the lib/et makefile, make sure com_err.info is deleted on "make clean". 4214 4215Fix 64-bit warnings in e2fsprogs pass1b by using inttypes.h if 4216present. This is for when we try stuffing an int into void * pointer. 4217 4218Fix type-punning which can cause gcc 3.x to miscompile code by getting 4219confused about pointer aliasing. ext2fs_getmem(), ext2fs_free_mem(), 4220and ext2fs_resize_mem() all now take a 'void *' instead of a 'void 4221**'. The EVMS code uses an ugly union approach since we don't want to 4222modify the EVMS interfaces. 4223 4224Make sure all Makefiles use $(MAKE) rather than hardcoded "make", to 4225aid build process on systems can use invoke GNU make as "gmake". 4226 4227Added regression testing for mke2fs. 4228 4229Fixed gcc -Wall nitpicks. 4230 4231Fixed various compiler warnings. 4232 4233Add portability fixes for FreeBSD and for using fsctl under Darwin to 4234support ext2 ioctl's. 4235 4236 4237E2fsprogs 1.34 (July 25, 2003) 4238=============================== 4239 4240Fixed a bug introduced in E2fsprogs 1.30 which caused fsck to spin in 4241a tight loop while waiting for a child fsck to exit in some cases. 4242This burns CPU times which slows down the low-level filesystem check. 4243 4244Added code to mke2fs to assure that the default block size for a 4245filesystem is at least as big as the sector size of the device, if it 4246can be determined. 4247 4248Changed mke2fs and resize2fs to round the default size of a filesystem 4249to be an even multiple of the VM pagesize in order to avoid a Linux 4250kernel bug introduced when the storage of the buffer cache was moved 4251into the page cache. 4252 4253Mke2fs will warn the user when creating a filesystem with journaling 4254and a blocksize greater than 4096. (Addresses Debian bug #193773) 4255 4256Fixed a bug in resize2fs which caused it to fail on filesystems with a 4257non-empty bad block list. Resize2fs now discards any blocks on the 4258badblock list which are no longer part of the filesystem as the result 4259of a filesystem shrink. (Note: this means that shrinking and then 4260enlarging a filesystem is no longer a reversible operation; 4261information about bad blocks in the part of the filesystem which is to 4262be chopped off will be lost.) 4263 4264Changed resize2fs so the user can use prefixes to specify the units of 4265the new filesystem size (sectors, kilobytes, megabytes, or gigabytes), 4266and to make the error and informational messages explicitly display 4267the blocksize used by the filesystem, in order to avoid confusion. 4268(Addresses Debian bug: #189814) 4269 4270Added a new debugfs command, dump_unused, which dumps the contents of 4271all unused blocks to stdout. (Useful as an emergency try-to-find 4272deleted data command.) 4273 4274Added a new debugfs command, imap, which prints the location of a 4275specified inode in the inode table. 4276 4277Fixed a bug in the badblocks program which caused it to use one bit of 4278randomness in its non-destructive read/write test, instead of using a 4279full 8 bits of randomness. 4280 4281Added a new option (-t) to badblocks, which allows the user to control 4282the test pattern(s) used when checking a disk. 4283 4284The blkid probe function now more correctly detects UDF filesystems. 4285 4286Fixed a bug in the blkid library which caused it to not update its 4287cache if a filesystem changes from having a label to not having a 4288label. 4289 4290Fixed a bug in the blkid library wihch could avoid an infinite loop 4291in blkid_find_dev_with_tag() if /proc is not mounted and there the 4292/etc/blkid.tab file has not yet been created. 4293 4294Fixed the badblocks program so that the destructive read/write test 4295honors the -c option, and to use O_DIRECT when possible to avoid 4296thrashing the system block buffer cache. 4297 4298Fixed various NLS issues. 4299 - Added Czech and Swedish translations 4300 - Removed testing NYC translation 4301 - Fixed NLS support for message abbrevations in e2fsck 4302 - Remove de-utf.po, since we shouldn't have two versions using different 4303 charset encodings. 4304 - Used ngettext() (abbreivated with the macro P_(str1, str2, n)) to 4305 simplify the statistics reporting in e2fsck. 4306 4307Changed configure.in so that its defaults for *BSD systems no longer 4308build an fsck wrapper, and not to install in /usr/local by default. 4309 4310Fixed some minor spelling errors/typo's in e2fsck and the configure 4311script. 4312 4313Fixed various Debian packaging issues (see debian/changelog). 4314 4315Updated and clarified man pages. (Addresses Debian Bug #195616) 4316 4317Programmer's notes: 4318------------------- 4319 4320Fix gcc -Wall nitpicks. 4321 4322Updated gettext implementation used by e2fsprogs to 0.11.5, and enable 4323NLS support by default. (Added partial workaround for gettext/Darwin 4324incompatibility problems.) 4325 4326Added full MIT KRB5 and Himdall compaibility support to the com_err 4327library and the compile_et program. (Addresses Debian bug #191900) 4328 4329Added the blkid_known_fstype() function to the blkid library, which 4330returns true if it is passed a filesystem type which is recognized by 4331the blkid probing functions. 4332 4333Improved the documentation for the blkid library. 4334 4335Added the ext2fs_get_device_sectsize() function the the ext2fs library, which 4336returns the hardware sector size of a device, if it is available. 4337 4338Added a dependency in the blkid library's .so file to the uuid 4339library, since the former uses the latter. (Addresses Debian bug 4340#194094) 4341 4342Added --with-diet-libc and --disable-evms to the configure script. 4343 4344Fixed a minor memory leak in the badblocks program. 4345 4346Fixed a portability problem in tune2fs --- not all systems have strptime(). 4347 4348Fixed a portability problem in debugfs with the use of getopt() more 4349than once. Old-style BSD, new-style BSD, and Linux C libraries all do 4350things differently. 4351 4352Add support Windows support to ext2fs_get_device_size(). 4353 4354Added (normally disabled) debugging code to the Unix I/O manager which 4355causes it to disable all userspace caching if the NO_IO_CACHE is 4356defined. 4357 4358Changed the test I/O manager so it can always be linked into e2fsck, 4359mke2fs, and tune2fs if enabled via --enable-test-io-debug to the 4360configure script. The test I/O manager will only print any debugging 4361information if the TEST_IO_FLAGS or TEST_IO_BLOCK environment 4362variables are set, which specifies which I/O operations are logged and 4363a block number to watch, respectively. The log messages are sent to 4364stderr by default, unless a filename is specified via the 4365TEST_IO_LOGFILE environment variable. 4366 4367 4368E2fsprogs 1.33 (April 21, 2003) 4369=============================== 4370 4371Added a new utility program, logsave, which captures the output of a 4372command in a log file, even if the containing directory hasn't been 4373mounted yet (in which case the it saves the output in memory until it 4374can write out the logfile). This is useful for capturing the 4375output of fsck during the boot sequence. 4376 4377Fixed some portability problems that were causing problems under 4378the Cygwin32 environment. 4379 4380Mke2fs now interprets a negative number to the -b option as a minimum 4381block size. 4382 4383Fixed a bug in mke2fs which was incorrectly checking the argument to 4384the -g option if the default block size was used. (Addresses Debian 4385bug #188319) 4386 4387Fsck now explicitly ignores tmpfs and devpts, and it will complain if 4388it can not find filesystem checkers for jfs, reseirfs, and xfs. 4389 4390E2fsck now updates the global free block and inode counters from the 4391block group specific counters quietly. This is needed for an 4392experimental patch which eliminates locking the entire filesystem when 4393allocating blocks or inodes; if the filesystem is not unmounted 4394cleanly, the global counts may not be accurate. 4395 4396Imported bug fixes to the EVMS plugin from the EVMS 2.0 tree. (EVMS 43972.0 is not yet supported; this only pulled in the bug fixes: fixed 4398possible hangs caused by bugs in calling waitpid, and not setting the 4399pipe to non-blocking mode; also fixed a file descriptor leak; made 4400sure all functions call log entry/exit functions.) 4401 4402Badblocks now flushes its output file as bad blocks are discovered. 4403 4404The uuid library is now more paranoid about assuming the correctness 4405of the /dev/random device; it mixes in a stream of bytes from 4406random/srandom, just in case. 4407 4408Update Debian files to reflect the fact that I am now the Debian 4409maintainer of e2fsprogs. Other various Debian-specific packaging 4410cleanups. 4411 4412Move the source tarball generation functions from the top-level 4413makefile to the util/gen-tarball script. 4414 4415Updated the Turkish .po translation file. 4416 4417Added Heimdal and MIT krb5 extensions to the com_err library to make 4418it more compatible with com_err libraries from those distributions. 4419 4420Changed dumpe2fs to always display the superblock fields relating to 4421the journalling and/or directory indexing feature even if those 4422features are not enabled. 4423 4424Updated and clarify copyright statement vis-a-vis alpha releases of 4425e2fsprogs. 4426 4427The ss library will now try to dynamically link to the readline 4428library and use it if it is present in the system. This means that 4429the debugfs program now has line editing and history features. The 4430SS_READLINE_PATH environment variable is used to find a readline or 4431readline-compatible library. 4432 4433E2fsck now finds most duplicate filenames (all when rebuilding all 4434directories via the -D option) and offers to delete or rename 4435duplicate filenames/directory entries. (Addresses Debian Bug #176814). 4436 4437Fix bug in e2image. When writing out a raw image file, include data 4438blocks from symlinks that do not store the symlink within the inode. 4439 4440Fix bug in resize2fs which incorrectly moved the block and inode 4441bitmaps for sparse superblock filesystems and incorrectly marked 4442blocks as in use. (Addresses Debian bug #174766) 4443 4444Added a new shared library, the blkid library, which efficiently 4445allows fsck, mke2fs, e2fsck, and tune2fs to be able to look up LABEL 4446and UUID filesystem specifiers without needing to search all of the 4447devices in the system. Instead, the device is looked up in a cache 4448file, and then verified to make sure the blkid cache is correct. 4449 4450Tune2fs and e2label will accept LABEL=xxx and UID=yyy specifiers for 4451the device name, using the blkid library. (Addresses Debian bugs 4452#166048, #179671) 4453 4454Fsck now supports backslash escapes in /etc/fstab so that \040 can be 4455used for spaces in device labels. 4456 4457Removed 32-bit limitations for debugfs's dump command. 4458 4459If the user specifies a large number of inodes, Mke2fs will 4460automatically adjust the number of blocks per group to find a valid 4461set of filesystem parameters. 4462 4463Add workaround to detect broken MD devices where when some of the 4464underlying devices are marked read-only, writes to the MD device are 4465silently dropped. E2fsck will detect if there is an attempt to run 4466the journal twice, and abort with an error if this is the case. 4467(Addresses IBM Bugzilla bug #1226) 4468 4469E2fsck will print an error if more than one of the -p/-a, -n or -y 4470options are specified. 4471 4472E2fsck will fix HTREE corruptions in preen mode, without stopping the 4473boot process. This is needed because the 2.4 ext2 filesystem 4474accidentally had the INDEX_FL backwards compatibility code removed. 4475 4476Mke2fs no longer creates filesystems with the dir_index flag set by 4477default; the user has to specifically request it. 4478 4479Update and clarified various man pages. (Addresses Debian bugs 4480#173612, #175233, #175113, and #170497, #185945, #188318) 4481 4482Created man page for the mk_cmds program (from the libss library). 4483 4484Programmer's notes: 4485------------------- 4486 4487Fix various gcc -Wall nits. 4488 4489Fixed a lot of portability problems that caused e2fsprogs not to build 4490successfully under Solaris and Apple/Darwin. 4491 4492Fixed a Makefile dependency to allow building e2fsprogs using parallel 4493make jobs. 4494 4495Changes to create a subset distribution which consists only of the 4496et, ss, uuid, and blkid libraries. The configure script and top-level 4497makefile were changed to support working with a subset distribution. 4498 4499Removed EXT2_FEATURE_RO_COMPAT_BTREE_DIR mention of since it's not 4500actually used, and might people who are looking for 4501EXT2_FEATURE_COMPAT_DIR_INDEX, which is in use. 4502 4503Updated debian files to fix a number of Lintian warnings. 4504 4505Updated config.guess and config.sub with newer versions from the FSF. 4506 4507Removed unnecessary libraries from being linked into the fsck, lsattr, 4508chattr, and blkid executables. 4509 4510E2fsprogs 1.32 (Nomvember 9, 2002) 4511================================== 4512 4513Fixed a bug in the Unix I/O routines which caused needless writebacks 4514of clean blocks from the unix_io cache (they were erroneously marked 4515as being dirty, so they were getting written back to disk before 4516getting evicted from the disk cache). This was harmless, but it 4517significantly slowed down e2fsck. 4518 4519Made some other minor optimizations to the Unix I/O routines to save a 4520small amount of CPU time. 4521 4522Updated internationalization files. 4523 4524 4525E2fsprogs 1.31 (Nomvember 8, 2002) 4526================================== 4527 4528Update EVMS ext2fsim plugin with EVMS 1.2. (We still support 4529compiling the fsim plugin with EVMS 1.0 and 1.1.) Add better error 4530handling for child process that die unexpectly. Add a hack to force 4531mkfs to create filesystems that won't cause problems with hardware 4532that has 2k or 4k minimum blocksize requirements. Read from child 4533processes in non-blocking mode, so that the GUI continues to update. 4534 4535Fix e2fsck so that it returns the appropriate exit code when the root 4536filesystem has been changed, so that system's rc scripts will be told that 4537the system needs to be rebooted. 4538 4539Fix a bug in ext2fs_flush/ext2fs_close; when the MASTER_SB_ONLY flag 4540was set, some of the descriptor blocks that should have been written 4541out were getting skipped. 4542 4543Changed e2fsck to force out changes to the backup copies of the 4544superblock and block group descriptors when important changes are made 4545to those data structures. 4546 4547Fix a bug where e2fsck could erroneously mark a filesystem as being 4548clean if a check of dirty filesystem is interrupted with a ^C. (Bug 4549introduced in e2fsprogs 1.28.) 4550 4551If journal debuging is enabled using --enable-jbd-debug, the debugging 4552level is now set via the E2FSCK_JBD_DEBUG environment variable. 4553 4554If byteswapping support is disabled using configure --disable-swapfs, 4555skip the tests which depend on byte-swapping, so that "make check" 4556won't bomb out. 4557 4558Lshattr will now display the indexed directory flag. Also, some of 4559the more esoteric compression flags are supressed unless compression 4560support has been enabled. 4561 4562Update man pages. 4563 4564 4565E2fsprogs 1.30 (October 31, 2002) 4566================================= 4567 4568When resizing a filesystem, and inodes need to moved, resize2fs will 4569update the inode's ctime field, and the mtime field of the containing 4570directories, so that incremental backups using dump/restore will work 4571correctly. 4572 4573In order to avoid spurious errors, e2fsck wil check the last 4574superblock write time to determine whether or not it can safely use 4575the LOW_DTIME checks to determine if there are inodes on the orphan 4576inode linked list that had somehow gotten disconnected. (Addresses 4577Sourceforge bug #620980) 4578 4579Partition in /proc/partitions that being with the string "lvm" are 4580considered real partitions even if they do not end with a number. 4581 4582Fixed a bug in the the uuid generation function, where if /dev/urandom 4583is not present, but /dev/random is, and there isn't sufficient 4584entropy, the get_random_byte function could spin a loop forever. 4585 4586E2fsck, mke2fs, etc. will now reliably notice when image files are 4587mounted using the loopback interface. (Addresses Sourceforge bug 4588#619119) 4589 4590When flushing buffers (as is done by badblocks, for example) check to 4591see if the BLKFLSBUF ioctl succeeds, and if so, avoid using the 4592FDFLUSH ioctl which causes the MD device driver which causes confusing 4593syslog messages about obselete ioctl messages. (Addresses Sourceforge 4594bug #545832). 4595 4596Debugfs's write command now checks to see if the destination filename 4597exists before creating it. (Addresses Sourceforge bug #478195.) 4598 4599When installing man pages, search for compressed man pages using all 4600commonly used compression extensions (i.e., .Z, .gz, .bz2, etc.) 4601 4602Fixed a bug in fsck where multiple filesystems specified on the 4603command were not being checked in parallel due to a logic bug 4604introduced to support the FSCK_MAX_INST environment variable. 4605 4606We have added a new superblock field, s_mkfs_time, which indicates 4607when a filesystem was created. It is set via mke2fs, and printed out 4608by dumpe2fs, but is not actually touched by the kernel. 4609 4610Dumpe2fs has been made more robust by not aborting if there is an 4611error reading the block/inode bitmaps; instead it will still print out 4612the location of the block/inode bitmaps and inode table. 4613 4614Add support for the an alternative block group descriptor layout which 4615allows for on-line resizing without needing to prepare the filesystem 4616in advance. (This is the incomat feature flag meta_bg.) 4617 4618Add support for storing default mount options in the superblock, so 4619that the filesystem can be mounted with specific mount options without 4620needing to specify them on the mount command line or in the /etc/fstab 4621file. 4622 4623Add support for a new inode flag, which is to be used for indicating 4624the top of directory hierarchies for the Orlov block allocator. 4625 4626Fix e2fsck so that if it creates the lost+found directory, it does so 4627with the more apporpriate permissions of 0700. Also change 4628mklost+found so that it also creates the directory 0700. 4629 4630Fixed format bug in e2fsck if NLS is in use. 4631 4632Add a German translation for e2fsprogs's NLS support. 4633 4634Fixed e2fsck so that it more handles BAD_BLOCK_IN_INODE_TABLE even at 4635the beginning of the inode table. This won't matter much, since if 4636there is a bad block at the beginning of the inode table, the root 4637directory won't be available. But at least e2fsck won't crash in this 4638case. 4639 4640Fixed endian problems in the htree code for e2fsck and debugfs. 4641 4642When byte-swapping a filesystem on a PPC architecture, byte-swap the 4643bitmaps since the historical big-endian ext2 variant had byte-swapped 4644bitmaps, and the ext2fs library assumes this. This fixes the 4645regression test suite on PPC machines. 4646 4647Fix e2image so that it handles a bad block in the inode table 4648appropriately. 4649 4650E2fsck now uses a more sophisticated algorithm to salvage corrupted 4651directories that recovers more information in the corrupted directory 4652block. 4653 4654E2fsck now performs additional consistency checks on indexed (HTREE) 4655directories. 4656 4657Fix bug where efsck might get confused about whether a completely 4658empty directory block is an empty leaf block or an interior htree 4659node. This could cause e2fsck to get confused into think that a valid 4660indexed directory was corrupted. 4661 4662E2fsck no longer creates an empty directory entry at the end of a 4663directory block when optimizing a directory. This would cause some 4664earlier versions of the dxdir kernel code to corrupt the directory 4665when attempting to split a node containing an empty directory entry. 4666 4667E2fsck could sometimes in rare circumstances leave the indexed flag 4668set after a small directory was optimized by compressing it instead of 4669indexing it. (This should never happen in real life, since 4670directories that small wouldn't have been indexed, but better safe 4671than sorry.) 4672 4673E2fsck now only updates the master superblock in all cases. This 4674slightly shortens its run time. 4675 4676Ext2ed can deal with directory entries of length 0; previously it 4677would get stuck in an infinite loop. 4678 4679Fsck now has support for reiserfs volumes when parsing LABEL= and UUID= 4680specifiers. (Sourceforge patch #613447) 4681 4682Badblocks will now work correctly on read-only devices such as 4683CD-ROM's. (Sourceforge patch #600451) 4684 4685Updated and clarified man pages. (Addresses Debian bug #167108) 4686 4687 4688E2fsprogs 1.29 (September 24, 2002) 4689=================================== 4690 4691Fixed a bug in e2fsck which could corrupt a directory when optimizing 4692it (via the -D option) or rebuiliding the hash tree index with a 1 in 4693512 probability, due to a fence post error. 4694 4695Fixed a bug in the LVM support code which caused LABEL='xxx' not to 4696work correctly. 4697 4698Mke2fs now enables the directory indexing flag by default. (Since 4699this is a compatible feature flag, it's safe to do so.) 4700 4701Tune2fs will support setting the directory indexing feature flag. It 4702will automatically set up the default hash algorithm and hash seed 4703fields in the superblock. 4704 4705If the bone-headed user enters the root filesystem twice in 4706/etc/fstab, the -R option which skips the root filesystem will skip 4707all of them. (Addresses Debian bug #159423). Note! This is not a 4708precedent for dealing intelligently with any other kind of doubled 4709entry in /etc/fstab! 4710 4711 4712Programmer's notes: 4713------------------- 4714 4715Removed perror declaration in lib/et/internal.h. All modern systems 4716can be expected to define perror() these days. Besides, the lib/et 4717code wasn't using perror at all anyway. :-) 4718 4719 4720E2fsprogs 1.28 (August 31, 2002) 4721================================ 4722 4723Add support for the Hashed-Tree Directory Indexing to e2fsck. Support 4724for setting the htree flag is not included yet, although it can be 4725manually turned on via the debugfs program. 4726 4727Clarified e2fsck error message which is printed when it cannot find 4728sufficient contiguous block when relcating filesystem metadata. 4729 4730Added support for building an EVMS plugin module for ext2/3. This 4731module is substantially the same as the EVMS module shipping with EVMS 47321.1, with one or two bugfixes. E2fsprogs can also build this plugin 4733for use with EVMS 1.0 (which did not include the ext2 plugin module), 4734if the configure --enable-old-evms flag is given. 4735 4736Fsck will search through EVMS volumes when trying to resolve 4737filesystem specifications such as LABEL=xxx or UUID=xxx. 4738 4739Added a new utility program, /sbin/findfs, which will return 4740filesystem specifications such as LABEL=xxx or UUID=xxx, and prints 4741the device name. 4742 4743Update and clarified various man pages. (addresses Debian Bug #145044, 4744#146437, #131350, #151990, #144621, #141938) 4745 4746If there are no filesystems specified on fsck's command line, fsck now 4747treat that as if the -As options were given. Previously it would 4748simply do nothing. (Addresses Debian Bug #153102) 4749 4750Mke2fs no longer treats a failure to be able to clear the MD signature 4751at the end of the filesystem as a fatal error. (Addresses Debian Bug 4752#155007) 4753 4754The e2p library functions (which are used by lsattr and chattr) now 4755double check to make sure the file is a regular file or directory 4756before attempting to use the ext2 ioctls. Some device drivers 4757unfortunately respond to the ext2 ioctl's with unknown behaviour. 4758(Addresses Debian Bug #152029). 4759 4760The extended attribute handling has been updated to correspond with 4761the latest V2 bestbits ACL code. 4762 4763Fixed bug in e2fsck which caused it to not clear the dtime field when 4764processing truncated inodes on the orphan list. This could cause data 4765loss(!) if a filesystem is rebooted before a truncate has been 4766committed. 4767 4768E2fsck now uses red/black trees in pass1b, which removes some O(n**2) 4769algorithms. This makes e2fsck much faster in the case of severely 4770corrupted filesystems where a large number of blocks are claimed by a 4771large number of inodes. (Thanks to the 2.5 IDE device driver for 4772inspiring this work.) 4773 4774Resize2fs has been significantly sped up when shrinking and expanding 4775a filesystem by a very small number of blocks (for example, when EVMS 4776is converting a partition to be an EVMS legacy volume). 4777 4778Added a new option to e2fsck, -D, which will optimize or compress all 4779of the directories in the filesystem. 4780 4781E2fsck now catches SIGINT and SIGTERM to make sure it will can 4782properly clean up and only exit at safe points. Fsck will pass 4783SIGINT/SIGTERM to its child processes, and wait until they have all 4784exited before it exits. 4785 4786The uuid parsing code in the uuid library now properly complains when 4787an illegally formated uuid is presented to it. (Addresses Debian bug 4788#152891) 4789 4790Restrict use of the 2.4 setrlimit ABI f*ckup to kernels between 2.4.10 4791and 2.4.17, since the workaround can cause problems when using a 2.4 4792kernel with an old version of glibc built with the 2.2 headers. 4793 4794Fixed a bug in mke2fs where it wasn't properly clearing the initial 4795superblock used by other filesystems. (Addresses Debian bug #147256.) 4796 4797Added support for the synchronous directory feature written by Andrew 4798Morton. 4799 4800The debugfs program can delete directories using the rmdir command. 4801 4802Add support for 8k and 16k filesystems (for systems with page sizes 4803that are greater or equal to 8k or 16k, respectively). Note that 4804these filesystems can not be mounted on x86 systems, or other systems 4805with only 4k page sizes, due to limitations in the current Linux VM 4806code. 4807 4808Resize2fs requires that the filesystem state be valid and have no 4809errors; otherwise, e2fsck -f must be run first. (Previously it simply 4810required that the last fsck time be greater than the last mount time.) 4811 4812Configure now defaults the man pages directory to /usr/share/man on 4813Linux systems. 4814 4815E2fsck now offers to truncate inodes which contain too many blocks (so 4816that i_blocks would overflow. Also fixed handling of large sparse 4817files. 4818 4819E2fsck now more completely checks for symlink validity, including 4820requiring NULL termination and length checks. 4821 4822E2fsck will offer to try forcing a disk write to remap a bad block 4823after finding a read error when reading a filesystem metadata block. 4824 4825Fixed a bug in debugfs which caused the -b and -s options to crash 4826debugfs, as well as breaking the testb, setb, and clearb functions. 4827 4828Added a bmap command to debugfs which calculates the logical to 4829physical block mapping for a particular inode. 4830 4831Fixed a bug in code which checked to see if a device was mounted which 4832sometimes (rarely) failed in the case of a plain file. 4833 4834Fixed a bug in resize2fs where when it reported an error, it would 4835print a message erroneously indicating that the filesystem had been 4836resized before it aborted. 4837 4838When resizing a plain file which is smaller than the requested size, 4839resize2fs will attempt to extended the file so that programs like 4840e2fsck will not complain that the file is too small. 4841 4842Resize2fs will print the actual new size of the filesystem when it is 4843finished resizing. 4844 4845Fixed a bug in debugfs where "ls -l" would report incorrectl file type 4846information on big-endian systems. 4847 4848 4849Programmer's notes: 4850------------------- 4851 4852Fixed collisions with C++ reserved words. 4853 4854Added portability fixes for building e2fsprogs on the HURD and AIX. 4855 4856Added the ext2ed program for creation of test cases. (ext2ed has many 4857limitations and bugs which make it unsuitable for production use.) 4858 4859The ext2fs_read_dir_block2 and ext2fs_write_dir_block now take a new 4860flag, EXT2_DIRBLOCK_V2_STRUCT, which will will reverse when the 4861name_len field is byte swampped on big-endian machines, since in the 4862V2 structure, name_len is a char field which is doesn't need to be 4863byte swapped --- except if an old-style kernel had byte-swapped the 4864name_len field as part of the V1 structure. 4865 4866 4867E2fsprogs 1.27 (March 8, 2002) 4868============================== 4869 4870The warning messages for mke2fs now go to standard error. 4871 4872Fixed to make sure "make check" runs all of the test programs with 4873LD_LIBRARY_PATH set, so that we test the libraries in the build tree. 4874 4875The mke2fs program checks the boot sector for the BSD disk label, and 4876avoids erasing it if it is there. 4877 4878Fixed a bug in e2fsck which caused it to core dump if the journal 4879inode was missing when it was supposed to be there. 4880 4881E2fsck now prints ranges in pass 5 when printing deltas for the block 4882and inode bitmaps. 4883 4884Debugfs's "ls -l" command now will print out the file type information 4885in the directory entry. 4886 4887Create man pages and hard links for fsck.ext3 and mkfs.ext3. If 4888mke2fs is invoked as mkfs.ext3, create the filesystem with a journal. 4889 4890Debugfs can now examine the experimental directory indexing 4891information. 4892 4893Fixed bug in debugfs which caused it to core dump if modify_inode is 4894called without an open filesystem. 4895 4896The debugfs lsdel command now runs its output through a pager first. 4897 4898When installing manual pages, remove the compressed manual pages first. 4899 4900Synchronized with Debian's packaging information for e2fsprogs-1.26-1. 4901 4902Fix the 2.4 resource limitation workaround introduced in 1.26 which 4903actually broke things on mips32, sparc32, and Alpha platforms. 4904 4905Updated the I18N code so that calls to setlocate(LC_CTYPE, 0) are made 4906(which is required by the newer libintl libraries). 4907 4908Programmer's notes: 4909------------------- 4910 4911Fixed various gcc -Wall complaints. 4912 4913Fixed a few memory leaks in the e2fsck journaling code, and in the 4914ismounted code checking for a swap device. 4915 4916Add new inode I/O abstraction interface which exports an inode as 4917an I/O object. 4918 4919Exported ext2_file_flush as a public interface. 4920 4921In ext2_file_write, we now mark the buffer void * argument as a const, since ext2_file_write doesn't modify the buffer. 4922 4923Lots of small random portability fixes to make e2fsprogs build under 4924AIX --- even without the 5L compatibility toolkit, and even using the 4925uber-crippled AIX native C compiler. 4926 4927 4928E2fsprogs 1.26 (February 3, 2002) 4929================================= 4930 4931Dumpe2fs will keep going now if the bad block inode can't be read. 4932(Previously it stopped with a fatal error.) 4933 4934E2fsck will now give an opportunity to run the journal even if the 4935recovery flag is not set. This is the default behaviour if e2fsck is 4936using a backup superblock, since the needs_recovery flag will never be 4937set in the backup superblock copies. 4938 4939E2fsck now automatically finds the backup superblock/group descriptors 4940even when the primary superblock is completely destroyed for 2k and 4k 4941filesystems. (Previously it just guessed/assumed that we were dealing 4942with a 1k filesystem in that case, and users had to manually specify 4943the backup superblock number.) 4944 4945Fixed a bug in e2fsck where it dereferences a null pointer when there 4946is a problem opening a filesystem in preen mode. 4947 4948E2fsck now handles external journals correctly; previously it trashed 4949the external journal device if the journal needed to be replayed. 4950 4951Work around ulimit incompatibility problem caused by recent 2.4 4952kernels; the unix IO module will automatically try to set any resource 4953limits to be infinite on startup. 4954 4955Fixed a bug in e2fsck where it wasn't allocating a new block or inode 4956bitmap if it wasn't present and the blocksize was 2k or 4k. 4957 (Addresses Debian Bug #116975) 4958 4959E2fsck will check and fix botched ext3 V1 to V2 superblock updates by 4960clearing the new V2 fields if they do not make sense or if the ext3 4961superblock is version 1 superblock. 4962 4963E2fsck will automatically relocate the ext3 journal from a visible 4964file (i.e., /.journal) to an hidden inode if the filesystem has been 4965opened read/write. This allows the users to add a journal while the 4966filesystem is mounted, but the next time the system is rebooted, the 4967journal file will disappear. This avoids problems with backups, 4968stupid operators with superuser bits, etc. 4969 4970Fix a bug in fsck where it would not support a filesystem type of 4971"auto" if the device was specified in terms of LABEL=xxx or UUID=xxx. 4972 4973Fsck now supports fstab entries of the form "ext3,ext2". It will also 4974automatically identify reiserfs filesystems. 4975 4976The number of processes spawned by fsck can now be limited using the 4977FSCK_MAX_INST environment variable. 4978 4979Fsck now searchs the LVM proc hierarchy to find logical volumes which 4980should be searched for UUID and label information. 4981 4982Work around a bug in 2.4.10+ kernels by trying to unset the filesize 4983limit if at all possible, when opening a block device. (The filesize 4984limit shouldn't be applied against writes to a block device.) 4985 4986In mke2fs and e2fsck, specifying the -c option twice will now do 4987a read/write test on the disk. Update the man pages to encourage 4988using the -c option, and to discouraging running badblocks separately, 4989since users tend to forget to set the blocksize when running 4990badblocks. 4991 4992Mke2fs now automatically clears the immutable attribute on a 4993pre-existing .journal file left over from a previous failed attempt to 4994add a journal to an alreadyy-mounted filesystem. 4995 4996Fixed mke2fs's exit codes to consistently indicate when the mke2fs 4997operation failed. 4998 4999Mke2fs now creates the lost+found directory with permissions of 0700, 5000so that files that had previously lived in protected directory are 5001safe if they get dropped in lost+found and the system administrator 5002doesn't deal with immediately. (Addresses Debian #bug 118443) 5003 5004Mke2fs and e2fsck (and all programs that use the 5005ext2fs_check_if_mounted function) will now properly identify that a 5006device is mounted, even in cases where devfs has confused things with 5007multiple devices names with the same identity, or if a dim-witted 5008system administrator has done something stupid like creating device 5009file aliases in their home directory. Also now checks for swap devices 5010by using /proc/swaps. (Addresses Debian bug #106622) 5011 5012Added a new option (-T) to tune2fs which sets the time a filesystem 5013was last checked. 5014 5015Speed up e2image when creating sparse raw image files by optimizing 5016away excess lseek() system calls. 5017 5018Fix support of large (> 2GB) files when used as a filesystem in 5019mke2fs, tune2fs, debugfs, and findsuper. 5020 5021Debugfs's stat and icheck commands now properly deals with large (> 50222GB) files. 5023 5024Debugfs's set_super_value command now prints out the list of valid 5025superblock fields which can be set using the command. 5026 5027Debugfs's rm and kill_file command now updates the superblock free block 5028and inode counts, thus keeping the filesystem consistent. 5029 5030Debugfs's lsdel command now takes an optional argument which allows 5031the user to only see the most recently deleted files. 5032 5033A new command (undel) was added to debugfs which automates 5034undeleting a deleted inode and linking it back to a directory. 5035 5036Debugfs's ls command now takes a new option, -d, which lists 5037deleted directory entries. 5038 5039Debugfs's testb, freeb, setb, and find_free_block commands now take an 5040optional count argument. 5041 5042Add support for a new ext2 file attribute, EXT2_NOTAIL_FL, which will 5043be used to signal that a particular inode should not be eligible for 5044tail-merging --- this is needed for programs like LILO. 5045 5046The findsuper (an unreleased, uninstalled utility program) has been 5047improved to add extra validity checks and to add a progress meter. 5048(It is still an unsupported program; the officially supported way to 5049recover from a trashed partition table is to use gpart; findsuper is 5050for wizards only. :-) 5051 5052Debugfs was fixed to compile with "configure --disable-swapfs". 5053 5054Cleaned up various manual pages. (Addresses Debian bug #119624, #120171) 5055 5056Added new translation file for Turkish. 5057 5058Programmer's notes: 5059------------------- 5060 5061Fix general gcc -Wall complaints. 5062 5063E2fsprogs (mostly) works with the dietlibc. 5064 5065The programming texinfo file has been expanded to include description 5066of additional libext2fs functions. (Still not compeltely done, but a 5067lot of the more important functions have been documented.) 5068 5069Added a umask structure to struct_ext2_filsys, which currently only 5070modifies the behaviour of ext2fs_mkdir(), but if we add any file 5071creation functions to libext2fs, we should also make sure they respect 5072the umask setting. 5073 5074The build-rpm script was fixed to be a bit more safe. 5075 5076The tests' Makefile now has a way of automating test case creation 5077for e2fsck, using "make testnew". 5078 5079Created a new function, ext2fs_dir_iterate2 which passes more 5080information to the callback function (identical to the one used by 5081ext2fs_dblist_dir_iterate). The directory iterator functions take a 5082new flag, DIRENT_FLAG_INCLUDE_REMOVED, which will return deleted 5083directory entries. If the directory entry is deleted, the callback 5084function will be called with the entry paraemter set to 5085DIRENT_DELETED_FILE. 5086 5087Added new functions, ext2fs_inode_alloc_stats and 5088ext2fs_block_alloc_stats, which takes updates block/inode allocation 5089statistics in the bitmaps, block group descriptors, and superblock 5090when inodes or blocks are allocated or deallocated. 5091 5092 5093E2fsprogs 1.25 (September 20, 2001) 5094=================================== 5095 5096This is primarily a bug-fix release; no new features were added, but 5097there are a number of embarassing bug fixes and cleanups applied. 5098 5099Fix a bug mke2fs which causes stack garbage to be written to disk when 5100zapping disk sectors. (This bug could cause mke2fs to core dump for 5101some kernels, I suspect with security enhancement patches.) 5102 5103Remove unneeded #include of <linux/config.h> which was breaking 5104building e2fsprogs on the Hurd. (Addresses Debian bug #112414.) 5105 5106Updated tune2fs man page to reflect the fact that adding or removing a 5107journal doesn't require running e2fsck. 5108 5109Remove use of AC_REQUIRE from autoconf which had been used to prevent 5110AC_CANONICAL_HOST from being called twice; unfortunately this causes 5111recent autoconf to bomb out since they don't allow AC_REQUIRE to be 5112used outside of autoconf macros. Fortunately, it doesn't seem to be 5113necessary any more, anyway. 5114 5115E2fsck will now not fall back to an alternate superblock if the user 5116specifies the superblock location explicitly on the command-line. 5117This allows easier recovery from a situation where the primary 5118superblock and block groups are slightly corrupted, but the backup 5119superblocks are completely broken. 5120 5121Fix problem which caused compile_et and mk_cmds to blow up if 5122e2fsprogs was in a directory with a pathname that contained spaces. 5123 5124We are now more paranoid about checking the entry in /etc/mtab to make 5125sure the filesystem is really mounted, since some broken distributions 5126(read: Slackware) don't initialize /etc/mtab before checking non-root 5127filesystems. (Slackware also doesn't check the root filesystem 5128separately, and reboot if the root filesystem had changes applied to 5129it, which is dangerous and broken, but there's nothing I can do about 5130that.) 5131 5132Make UUID library C++ friendly by adding appropriate extern "C" 5133declarations and using const in the function declarations. 5134 5135Fix up the com_err texinfo file so that it can product a valid info 5136file (previously, it could only be used to generate paper 5137documentation using texinfo.tex). 5138 5139 5140E2fsprogs 1.24a (September 2, 2001) 5141=================================== 5142 5143Fix brown-paper bug in mke2fs which caused it to segfault when 5144printing the version string. 5145 5146 5147E2fsprogs 1.24 (August 30, 2001) 5148================================ 5149 5150Revert the BLKGETSIZE64 support, since for some distributions, this 5151ioctl has been used by an unofficial kernel patch to update the last 5152sector on the disk, and this was causing disk corruption problems as a 5153result. 5154 5155Mke2fs is now more careful about zapping swap space signatures and 5156other filesystem/raid superblock magic values so that programs like 5157mount who try to use hueristics to determine the filesystem type of a 5158partition are less likely to be confused. 5159 5160E2fsck will no longer complain if the the mode of EXT2_RESIZE_INO is a 5161regular file (since Andreas Dilger's on-line resizing tools will set 5162its mode bits). 5163 5164Fixed some minor texinfo, man pages nits for spelling errors, 5165texinfo warnings, etc. (Addresses Debian bug #110621.) 5166 5167E2fsprogs program no longer print the filesystem version number 5168(i.e. 0.5b), since it only confuses people and doesn't serve any real 5169purpose. 5170 5171E2fsck will once again compile under libc5, since it will supply its 5172own version of strnlen if necesssary. 5173 5174mke2fs and tune2fs will allow the use of UUID= or LABEL= specifiers 5175when specifying the external journal device. tune2fs will also search 5176devices looking for the externla journal debice when removing. 5177 5178E2fsprogs 1.23 (August 15, 2001) 5179================================ 5180 5181Add initial support for extended attributes (EA); e2fsck will 5182correctly handle a filesystem with EA's and check the EA blocks for 5183corruptions. 5184 5185E2fsck's symlink sanity checking has been cleaned up. It now checks 5186the i_size value of fast symlinks, and checks for immutable flags 5187being set of symlinks, etc. 5188 5189E2fsck now offers to clear inodes which are special files that have a 5190non-zero i_blocks or i_size field. (The i_size field check was in the 5191previous version of the code, but due to a bug it didn't offer to 5192clear the inode unless i_size and i_size_high were both non-zero.) 5193 5194E2image can now create "raw" image files, which only contain the 5195filesystem metadata placed in a spare file so that e2fsck, dumpe2fs, 5196debugfs, etc., can be run directly on the raw image file. 5197 5198Add support for the 64-bit block device patches. 5199 5200Fixed bugs in creating external journals with a 1k blocksize. 5201 5202Add initial support for external journals (so long as the external 5203journal only supports a single filesystem) in e2fsck. 5204 5205Remove requirement for needing to run fsck on a filesystem after 5206removing a journal (either internal or external). 5207 5208The man pages now document how to create and manage external journals. 5209 5210Speed up the check of an ext3 filesystems by avoiding a needless flush 5211of all of the superblock and block group descriptors. 5212 5213Speed up creating an internal journal using tune2fs in the case where 5214the filesystem has a lot blocks already allocated. 5215 5216Tune2fs has been fixed to make sure that only error messages go to 5217stderr, and normal message go to stdout. (Addresses Debian bug #108555) 5218 5219Fixed a minor bug in mke2fs; if -O none is passed to mke2fs, it will 5220now not set the sparse_super feature. (Addresses Debian bug #108165) 5221 5222Add support in fsck for the filesystem type "auto". 5223 5224Fsck -A will not try to interpret device names for filesystems which 5225have a pass number is 0. (Addresses Debian bug #106696). 5226 5227Fsck prints a warning message if now valid filesystems are passed to 5228it. (Addresses Debian Bug #107458.) 5229 5230E2fsck now gives an explicit warning if there filesystem still has 5231errors at the end of the run. (Addresses Debian bug #104502) 5232 5233E2fsck will set the EXT2_ERROR_FS flag if the journal superblock 5234reflects an error. E2fsck will also not run the orphan list if the 5235filesystem contains errors, since the orphan list might be corrupted. 5236 5237E2fsck now prints the number of large files when given the -v option. 5238 5239Fixed minor memory leaks in e2fsck. 5240 5241Some minor man pages updates. (Addresses Debian bug #30833, #108174) 5242 5243 5244E2fsprogs 1.22 (June 22, 2001) 5245============================== 5246 5247Fixed a bug in e2fsck's handling of orphan inodes which are special 5248files (i.e., block/character device files, named FIFO's, etc.). 5249 5250Fixed a bug accidentally introduced in e2fsprogs 1.21 which caused 5251tune2fs to fail at adding a journal to a mounted filesystem. 5252 5253Fixed a few big-endian bugs in e2fsprogs 5254 * The directory block functions were accidentally reporting 5255 some directories as corrupted when they weren't. 5256 * If e2fsprogs is compiled --disable-swapfs, the C language 5257 equivalents weren't being included for big-endian platforms. 5258 (Fixes Debian bug #101686). 5259 5260Fixed a Hurd compilation problem. (Addresses Debian bug #101361) 5261 5262Programmer's notes: 5263------------------- 5264 5265Use platform independent method of defining the BLKFLSBUF and FDFLUSH 5266ioctls, and allow them to be defined for non-i386 platforms. 5267 5268The uuid.h header file is now protected against multiple inclusions. 5269 5270E2fsprogs is now being developed using BitKeeper. Changed the test 5271scripts to deal with BK's stripping CR characgters from text files, 5272and changed the top-level Makefile.in to avoid including BitKeeper 5273files when generating the source tarball. 5274 5275 5276E2fsprogs 1.21 (June 15, 2001) 5277============================== 5278 5279Added new configure flags which allow a subset e2fsprogs to be built; 5280this is most useful for boot floopies, since the resulting shared 5281libraries and programs are slimmed down by removing features that 5282aren't necessary for a boot floppy. The new flags that were added 5283are: --disable-swapfs, which removes support for byte swapping old 5284legacy PPC/68k filesystems, --disable-debugfs, which removes support 5285for debugfs from the libext2fs library, --disable-imager, which 5286removes support for the e2image program, and --disable-resizer, which 5287removes support for resize2fs. 5288 5289E2fsck now prints the number of mounts or days elapsed since the last 5290check when e2fsck is forced to check an otherwise clean filesystem. 5291 5292Tune2fs now prints an informative message about how often a filesystem 5293will be checked when adding a journal to the filesystem, to remind the 5294user that he/she may want to adjust those parameters using tune2fs 5295-c/-i. 5296 5297Worked around hurd brain-damage which causes e2fsck to sometimes 5298believe a filesystem is the root filesystem based on device numbers 5299(since Hurd doesn't have dev_t's, which is arguably a POSIX.1 5300violation). 5301 5302Fixed a bug introduced in 1.20 which caused e2fsck to abort with an 5303erroneous error with the -F option was specified. 5304 5305Fixed a ext3 recovery bug in the revoke handling; synchronized with 5306ext3 0.7a. 5307 5308Fixed two bugs in e2fsck's handling of dup block handling, dealing 5309with relatively uncommon edge cases: a directory with an indirect 5310block which is claimed by another file, and when the last inode in the 5311filesystem has blocks claimed by another file. 5312 5313E2fsck now checks to see if the i_size field of a fast symlink is too 5314big, and offers to clear the symlink if so. 5315 5316E2fsck now checks to see if i_size_high of special files is non-zero, 5317and offers to clear i_size_high. 5318 5319Fix e2fsck's handling of incompatible journal flags so that the user 5320has chance to abort, and then has the option to clear out the journal 5321entirely. (Addresses Debian bug #98527.) 5322 5323Fixed a bug in fsck which could cause it to core dump if a mix of 5324standard and non-standard device names are used in /etc/fstab. 5325(Debian bug #100559) 5326 5327Fixed a bug in debugfs which caused read errors when copying a file to 5328not be noticed. 5329 5330The debugfs set_super_value command can now modify the s_lastcheck field. 5331 5332Fixed a bug in lsattr and chattr which was accidentally introduced in 53331.20 to support > 2GB files; both lsattr and chattr wasn't reading 5334directories correctly because the change modified the layout of struct 5335dirent to be incompatible with the libe2p shared library. 5336 5337Cleaned up the mke2fs manual page and included a discussion about why 5338it's good to periodically check the filesystem even when journaling is 5339enabled. 5340 5341Programmer's notes: 5342------------------- 5343 5344Fix general gcc -Wall complaints. 5345 5346The types needed by the ext2 header files are now provided by 5347lib/ext2fs/ext2_types.h, instead of include/asm/types.h. 5348 5349Integers are now preferred to longs when trying to find a 32-bit type 5350in ext2_types.h. Also, if linux/types.h has already been defined, 5351don't try to redefine the types. 5352 5353Fixed make depend script so that it automatically corrects the 5354pathname cleanups performed by make -M, so I don't have to fix them up 5355by hand. 5356 5357Fixed the d_loaddump test case to be more robust, and not depend on 5358bash'isms. 5359 5360Removed debugfs's dependence on pread(), which was accidentally 5361intrudced in e2fsprogs 1.20 5362 5363Fixed a performance bug in the libext2fs's icount routine; the size 5364estimate of the icount array was incorrectly being calculated. 5365 5366Removed use of the badblocks compatibility functions in the e2fsprogs 5367programs. 5368 5369Added paranoia code which protects against strange cases where /etc 5370isn't on the root filesystem, or if /etc/mtab doesn't exist. 5371 5372The header file ext2_types.h is now installed. 5373 5374Autoconf is used to determine when we are on big-endian machines, 5375instead of doing run-time tests, to save a few bytes of code. 5376 5377The ext2fs_mark_generic_bitmap and ext2fs_unmark_generic_bitmap 5378functions are no longer inline functions, which saves space and 5379doesn't really cost any real performance. 5380 5381The ext2fs library no longer depends on the e2p library. (What need 5382there was of it --- namely, fsetflags, was coded in-line). 5383 5384Fixed the makefile so that lib/ext2fs/ext2_types.h is generated even 5385when the user is stupid and tries compiling the package using "make 5386install" as root. 5387 5388Miscellaneous code cleanups: 5389 * Added missing files from Makefile.in's SRCS file, so that 5390 their dependencies would be properly calculated. 5391 * Removed redundant code 5392 * Fixed comments in code 5393 * Removed no-longer unneeded argsused #pragma. 5394 5395 5396E2fsprogs 1.20 (May 20, 2001) 5397============================= 5398 5399Add support for replaying the ext3 journal purely in user mode, 5400including handling the orphaned inode list. Used code contributed by 5401Andreas Dilger, with bug fixes and the orphaned inode handling done 5402by Theodore Ts'o. 5403 5404The mke2fs and tune2fs programs can create or modify a filesystem to 5405include an ext3 journal. Tune2fs also can be used to remove an ext3 5406journal from a filesystem. 5407 5408E2fsck will now check for the existence of a linked list of orphan 5409inodes in the superblock, and clear those inodes before starting the 5410rest of the filesystem check (but after the journal playback). 5411 5412E2fsck now validates the file descriptor passed to the -C option, 5413which saves against the completion bar getting written to an 5414unexpected location, such as the disk being checked. (Debian 5415bug/wishlist #55220) 5416 5417E2fsck will now bump the filesystem revision number from zero to one 5418if any of the compatibility bits are set. 5419 5420Fixed a bug where a badly corrupted directory on a big endian system 5421could cause e2fsck to die with a bus error. The 5422ext2fs_read_dir_block() and ext2fs_process_dir_block() functions in 5423the ext2 library now does alignment sanity checks on the rec_len field 5424of the directory entry before using it. 5425 5426The ext2 library has been enhanced to make tune2fs safe to run on 5427mounted filesystems. (Users could usually get away with using tune2fs 5428on mounted filesystems before, but with the advent of ext3 and 5429journaling, it became important to make tune2fs was *really* safe for 5430use even when the filesystem being modified is mounted.) E2label is 5431now implemented by tune2fs using an argv[0] dispatch, so that e2label 5432is also now safe for use on mounted filesystems. 5433 5434Added a new program, e2image, which creates a backup of critical ext2 5435filesystem data structures. The generated image file can be examined 5436using dumpe2fs and debugfs. In the future, e2fsck will be able to use 5437the image file to help recover very badly damaged filesystems. 5438 5439Fixed a number of LFS bugs in e2fsck; very, very large (> 2**42) files 5440no longer cause e2fsck to bomb out. Also treat files > 2GB as being 5441large file when deciding whether or not the filesystem has large files. 5442 5443Fixed lsattr and chattr so that they work correctly on large files. 5444(Fixes Debian bug #72690.) 5445 5446Removed limitation in get_device_size() which imposed a one terrabyte 5447filesystem limitation. (Most 2.2 kernels still have a signed int 5448problem which cause 1 TB block device limitation. Fortunately, the 5449kernel patches to fix this are much easier than fixing the 2TB 5450limitation in the kernel. :-) 5451 5452A max_mount_count of zero is now treated as if no mount count were 5453set. (Previously, no mount count was indicated by using -1, and a 5454mount count of zero caused e2fsck to always be run.) 5455 5456Mke2fs supports two new filesystem types largefile and largefile4. 5457 5458Mke2fs now adds some randomness in s_max_mount_count so that multiple 5459filesystems won't be all checked at the same time under normal 5460operations. 5461 5462Fixed bug in the progress bar printing code which could cause e2fsck 5463to core dump on an illegal filesystem. 5464 5465Fixed bug in fsck which could allow more than one instance of e2fsck 5466to be printing a progress bar. (Debian bug #65267) 5467 5468Fsck using a UUID or a LABEL specifier will work even if devfs is 5469compiled into the kernel and not mounted. If the pathnames in 5470/proc/partitions are incorrect, fsck will search /dev for the correct 5471device (using the new ext2fs_find_block_device library function). 5472Fsck now also checks the RAID devices first so that they are properly 5473found when they are in use. Support has also been added to support 5474additional IDE disks and the DAC 960 device names. (Debian bug #94159) 5475 5476Fixed a bug in fsck which caused it not deal properly with 16 5477byte long filesystem labels. 5478 5479Fsck's -t option has been made a lot more flexible. The semantics for 5480what happens if a comma-separated list to fsck has been regularized, 5481and it is now possible to filter what filesystems will get checked 5482based what is in the filesystem's fstab entry's option field. (Debian 5483bug #89483.) 5484 5485The dumpe2fs program can now print out the group description 5486information in hex, and also prints the location of the superblock and 5487block group descriptor for those block groups that have them. 5488 5489Mke2fs now clears the ext2 superblock before it starts creating a 5490filesystem, so that the superblock magic number is only written if the 5491filesystem creation process successfully completes. 5492 5493The debugfs program's stat command now pretty-prints the blocks used 5494by an inode so that it's more compact and informative. 5495 5496The debugfs stats command now uses the same libe2p code (which is used 5497by dumpe2fs) to print the superblock header information. This is more 5498complete, and it avoids a bit of code duplication. 5499 5500Added a new debugfs command, set_super_value (ssv) which allows the 5501user to set arbitrary superblock fields. 5502 5503Debugfs was extended to support inode numbers in hex (by prefixing 5504them with 0x), and so that modify_inode can set the inode generation 5505number. Also, there is now a new function command called logdump 5506which will dump an ext3 journal. 5507 5508Fixed a bug in debugfs so that quitting out of the pager doesn't kill 5509debugfs. 5510 5511Debugfs's dump command now stops immediately upon reporting a disk 5512read error. (Fixed a bug in ext2fs_file_read library routine which 5513caused debugfs not to stop.) (Debian bug #79163) 5514 5515On systems with /proc/mounts (mainly Linux systems), /proc/mounts is 5516checked before /etc/mtab is used, since /proc/mounts is more likely to 5517be accurate. 5518 5519Added portability fixes for Solaris and Linux/ia64. 5520 5521Various manual pages were clarified and cleaned up. (Fixed debian 5522bugs #63442, #67446, and #87216) 5523 5524 5525Programmer's notes: 5526------------------- 5527 5528The e2fsck message printer now supports %Iu and %Ig, which will print 5529out the inode's user and group owners, respectively. 5530 5531E2fsprogs now includes its own version of include/linux/ext2_fs.h, so 5532that no longer dependent on the system having the correct version of 5533the kernel header files. 5534 5535Added a new function to libext2, ext2fs_find_block_device(), which 5536searches the system (i.e., /dev, /devfs, /devices) for a pathname to a 5537device given its device number. 5538 5539Added a new function to libext2, ext2fs_sync_device, which centralizes 5540all of the places which might try to use the BLKFLSBUF or FDFLUSH 5541ioctls (and usually failing to define them since the system header 5542files don't usually do this for us, and we're trying to avoid usage of 5543kernel include files now). 5544 5545Added new utility programs in tests/progs: random_exercise and 5546hold_inode. They aren't built by default; they're useful for 5547exercising ext3 filesystem code. 5548 5549Added a new ext2 filesystem flag, EXT2_FLAG_SUPER_ONLY, which causes 5550the filesystem close functions to only update the superblock, and to 5551not touch the block group descriptors. Needed by tune2fs when 5552modifying a mounted filesystem. 5553 5554Got rid of struct ext2fs_sb and replaced it with the standard struct 5555ext2_super_block from include/linux/ext2_fs.h. Note: this may break 5556source (but not binary) compatibility of some users of the ext2 5557library. Those applications should just simply do a global search and 5558replace of struct ext2fs_sb with struct ext2_super_block, and include 5559the new header file <ext2fs/ext2_fs.h> which defines it. 5560 5561The ino_t type has been renamed ext2_ino_t to protect applications 5562that attempt to compile -D_FILE_OFFSET_BITS=64, since this 5563inexplicably changes ino_t(!?). So we use ext2_ino_t to avoid an 5564unexpected ABI change. 5565 5566The Makefiles have been reworked so that "make check" can be run from 5567the top-level Makefile. 5568 5569Fix general gcc -Wall complaints and removed dead code. 5570 5571Remove use of NOARGS, because we assume everyone does ANSI C these 5572days. 5573 5574Added build-rpm script from sct. 5575 5576New functions ext2fs_image_{inode,super,bitmap}_{read,write} added 5577to support e2image. 5578 5579New function ext2fs_flush_icache which must be called if the 5580application program modifies the inode table blocks without going 5581through ext2fs_write_inode() interface. 5582 5583New ext2fs_check_mount_point() function, which will return the mount 5584point of a device if mounted. 5585 5586The io_channel abstraction now has an optional interface, 5587io_channel_write_range, which allows specific byte ranges to be 5588written. 5589 5590The unix_io IO channel now supports write-through caching, so that 5591journal creation is more efficient. 5592 5593Added x86 assembly language routines to support byte swapping, to 5594reduce executable size. 5595 5596Fixed bug in the utility program subst so that it's possible to 5597replace a substitution variable with a zero-length string. 5598 5599Fixed numbering e2fsck pass1 problem numbers; an extra zero had 5600slipped into some of the problem number. 5601 5602E2fsprogs 1.19 (July 13, 2000) 5603============================== 5604 5605Release the resize2fs program since the timeout before it could 5606be released under the GPL has finally expired. 5607 5608Add experimental support needed for the ext2 compression patches. 5609This requires compiling e2fsprogs with the --enable-compression flag 5610to the configure script. 5611 5612Added ext3 journalling support. E2fsck will run the journal (if 5613necessary) by temporarily mounting the filesystem. /sbin/fsck.ext3 is 5614installed as a symlink to e2fsck. Fsck has been taught about ext3, 5615and treats it the same as ext2 in terms of the progress bar logic. 5616Dumpe2fs will display the superblock journaling information if the 5617filesystem has a journal. The ext2 library will now permit opening an 5618ext3 filesystem with the recovery flag set. This is necessary for 5619on-line dump's to work correctly, but there may be issues with this 5620working well since ext3 is much less agressive about syncing blocks to 5621the filesystem, since they're safe on the journal. 5622 5623Tune2fs and e2fsck have been changed to allow the mount_count check to 5624be disabled by setting max_mount_count to -1. (This was already 5625supported by the kernel.) 5626 5627Create a symbolic link for fsck.ext3, since the e2fsprogs utilities 5628are used for ext3 as well. 5629 5630Added internationalization support for e2fsprogs; must be enabled 5631by passing --enable-nls to configure. 5632 5633Always use the provided ext2fs header files to insulate ourselves from 5634kernel version changes. Which include files are used by e2fsprogs 5635have also been cleaned up to improve portability. 5636 5637Limit the number of times that e2fsck updates the progress bar so that 5638people who are booting using a 9600 baud console don't get swampped by 5639too many updates. 5640 5641Improved the loop detection algorithm in e2sck's pass #3 so that it is 5642much, much faster for large filesystems with a large number of 5643directories. 5644 5645The memory footprint for e2fsck is now slightly smaller than before. 5646 5647E2fsck now checks if special devices have a non-zero size, and offers 5648to clear the size field if it finds such an inode. 5649 5650E2fsck now checks if special devices have the append-only flag set, 5651and offers to clear the inode. 5652 5653E2fsck now properly handles some "should never fail" cases during a 5654bitmap copy in pass5. 5655 5656E2fsck now properly prints control characters in filenames as ^A .. ^Z. 5657 5658E2fsck now calculates the correct location of the backup superblock in 5659the case of filesystem blocksizes > 1k. 5660 5661Fixed a bug in e2fsck's calculation of the number of inodes_per_block 5662which normally didn't cause problems under most filesystem parameters, 5663but could cause a valid superblock to be rejected in extreme cases. 5664Other checks for validating superblock values were made more 5665stringent. 5666 5667Added non-destructive write testing to the badblocks program, courtesy 5668of David Beattie. The badblocks also now has an option to input the 5669current set of bad blocks, so that known bad blocks are skipped to 5670speed up the badblocks test. There is also a persistent rescan 5671feature which causes badblocks to run until it has completed some 5672number of passes without discovering any new bad blocks. 5673 5674Badblocks now checks to see if the device is mounted and refuses to do 5675the tests involving writing to the device if it is mounted. Also, 5676badblocks now allows the number of blocks to be checked to be 5677defaulted to the size of the partition. 5678 5679Fixed a bug in fsck which didn't allow non-root users to be able to 5680check filesystems if there were any LABEL= or UUID= entries in 5681/etc/fstab. 5682 5683The Hurd doesn't support the filetype filesystem feature. The mke2fs 5684program now makes sure that for the Hurd, the filestype feature is 5685turned off. E2fsck will check to see if the filetype feature is 5686turned on for Hurd filesystems, and offer to turn off the feature. 5687 5688Mke2fs now has a safety check to make sure the number of blocks do not 5689exceed 32 bits even on a 64 bit platform. 5690 5691Really fixed a bug in fsck to allow "fsck -As" to run interactive 5692fsck's. (For those people who like to do interactive fsck's in the 5693/etc/rc scripts!?!) 5694 5695Debugfs has a few new features: the rdump command, which will do a 5696recursive dump of a directory and all of its contents, and the lcd 5697command which does a local chdir (much like the ftp command of the 5698same name). In addition, the debugfs program and the open_filesystem 5699command now takes three new options: -b and -s, which allows the 5700blocksize and superblock location to be specified, and the -c option 5701which is used in catastrophic situations where the block group 5702descriptors are corrupt. If the -c option is specified, debugfs will 5703skip trying to read in the block and inode bitmaps. 5704 5705Debufs's lsdel command was fixed to handle bad blocks in the inode 5706table. 5707 5708A Y2K bug in debugfs's "ls -l" handling was fixed by switching to use 57094 digit years. 5710 5711General improvements in error messages 5712 5713 - Mke2fs prints a sane error message if the partition size is zero 5714 (usually because the partition table wasn't reread by the 5715 kernel due to the partition being busy), instead of "invalid 5716 argument passed to ext2 library while initializing superblock". 5717 5718 - Fsck now prints more self-explanatory message if an invalid UUID= 5719 or LABEL= specification is passed to it. 5720 5721UUID library changed to use the LGPL. 5722 5723Fixed a bug in the UUID library where very rapid calls to the 5724time-based UUID generator could cause duplicate UUID's to be returned. 5725This was not a problem for e2fsprogs, but it could be a problem for 5726other users of the library. 5727 5728Make the UUID library more robust in the face of missing or an 5729improper /dev/urandom or /dev/random files. 5730 5731Added some random portability fixes for Solaris. 5732 5733Some minor man page updates. 5734 5735Fixed a memory leak in the ss library. 5736 5737 5738Programmer's notes: 5739------------------- 5740 5741We now try to use lseek64 and open64 from the LFS if possible. 5742 5743The 3rd parameter in e2p's print_flags is now a flags word, instead of 5744a boolean option. 5745 5746The mark and unmark bitmap functions now return the previous state of 5747the bit that was being changed, which is useful for some speed 5748optimizations. 5749 5750The following functions have been added to enhance the badblocks list 5751handling in libext2fs: ext2fs_write_bb_FILE, ext2fs_read_bb_FILE2, and 5752ext2fs_badblocks_equal. 5753 5754The ext2 header files now have the latest journalling fields to the 5755superblock. 5756 5757The ext2fs_mkdir function in libext2fs now properly backs out of error 5758conditions robustly. 5759 5760Cleaned up makefiles: 5761 - to cleanly compile with the -j flag. 5762 - so distclean removes all generated files. 5763 - so in case of an error while installing header files, the make aborts. 5764 5765Fix test_script so that it works correctly when compiling in the 5766source directory. 5767 5768The random UUID generation routine has been made slightly better in 5769the case where /dev/random doesn't exist. (Use of randomly-based UUID 5770is still not recommended if /dev/random doesn't exist, however; it's 5771better to use the time/ethernet MAC address UUID in this case.) 5772 5773Clean up the build process so it's more friendly in case of missing 5774directories. 5775 5776The ext2fs header file can now be #include'd into C++ programs. 5777 5778The e2p.h header file is now installed. 5779 5780Added workaround to a gawk 3.0.5 bug in lib/ss/mk_cmds. 5781 5782 5783 5784E2fsprogs 1.18 (November 10, 1999) 5785================================== 5786 5787Fix a core dumping bug in e2fsck if an imagic inode is present or 5788(more rarely) if the filesystem is badly corrupted enough that e2fsck 5789has to restart pass 1 processing. E2fsck now closes the filesystem 5790before freeing a large number of its data structures, so in the case 5791of future memory faults, at least the fixed filesystem will be fully 5792written out. 5793 5794If a filesystem doesn't support imagic inodes, and e2fsck discovers an 5795imagic inode, it will offer to clear the imagic flag. 5796 5797E2fsck will now offer to clear the immutable flag on special files 5798(device/socket/fifos) when running it in non-preen mode. 5799 5800E2fsck will now set the filetype when creating /lost+found, and when 5801connected orphaned inodes to /lost+found. 5802 5803Debugfs's ncheck and icheck commands now handles the case where there 5804are bad blocks in the inode table without bombing out. 5805 5806The badblocks list processing code has been made more efficiently for 5807appending a large number of (ordered) badblocks to the badblocks list. 5808 5809Some minor man page updates. 5810 5811Fsck now allows interactive e2fsck's when using fsck -As (not a common 5812mode, but some people like to do this in boot scripts for silly reasons). 5813 5814Programmer's notes: 5815------------------- 5816 5817The internal e2fsck problem code for PR_2_SPLIT_DOT was fixed to meet 5818with the problem code convention. 5819 5820The badblocks list regression test program has been updated to work 5821with previously made API name changes. 5822 5823The ext2fs_free() command now uses the new badblocks API to avoid 5824using the compatibility layer. 5825 5826Added new regression test cases; the run_e2fsck test script now 5827supports the ability for a test case to run a prepratory command 5828before running e2fsck. 5829 5830E2fsprogs 1.17 (October 26, 1999) 5831================================= 5832 5833Fixed nasty typo in fsck which caused parallelized fsck's to go into an 5834infinite loop. 5835 5836Fixed a bug in fsck where it used strncmp to compare a binary UUID, 5837thus potentially causing problems if a binary UUID contained a NULL 5838character. 5839 5840E2fsck now uses stricter checks for directory entries in pass 2: 5841zero-length filenames are not allowed; neither are 8 byte long 5842directory entries. 5843 5844The debugfs "dirty" command now clears the filesystem valid bit. 5845(Previously this just set the dirty-as-in-needs-writing-out-to-disk 5846bit in the in-core superblock image. The new functionality is more 5847what the user expects, and is more useful.) 5848 5849Added a debugging hook to test parallel fsck; if the environment 5850variable FSCK_FORCE_ALL_PARALLEL, then filesystems on the same drive 5851will be checked in parallel when they normally would not be. 5852 5853Programmer's notes: 5854------------------- 5855 5856Fixed some #ifdef's for compilation under the Hurd OS. 5857 5858Fixed minor W2K compatibility problems. 5859 5860Fixed some miscellaneous GCC warnings. 5861 5862 5863E2fsprogs 1.16 (October 22, 1999) 5864================================= 5865 5866Fixed a race condition bug in fsck; when printing a progress bar, if 5867checking multiple filesystems in parallel, it was possible for fsck to 5868send e2fsck a SIGUSR1 signal before e2fsck had installed its signal 5869handler, which would cause it to terminate with a signal 10. 5870 5871E2fsck now properly handles filesystems that have the 5872INCOMPAT_FILETYPE feature turned on. It can be used to convert a 5873filesystem into using or not using FILETYPE feature. 5874 5875E2fsck now properly handles filesystems that have the IMAGIC feature 5876turned on (this is used on Linux AFS servers). 5877 5878The mke2fs program now creates filesystems that have the filetype and 5879sparse_superblock features enabled by default, unless it is run on a 5880pre-2.2 kernel. These features are not supported by a pre-2.2 kernel, 5881so there is now a new flag -O which allows the user to specify with 5882which features she would like to create the filesystem; "mke2fs -O 5883none" will create a filesystsem compatible with 2.0 kernels. 5884 5885The tune2fs program now has a -O option which allows the user to set 5886and reset "safe" filesystem features. Currently, the only ones which 5887allows to be modified are the filetype and sparse_superblock features. 5888Note setting or clearing either feature will require running e2fsck on 5889the filesystem afterwards. (n.b. Clearing the sparse_superblock feature 5890requires that there is enough free space on the filesystem for the 5891extra superblocks which will be created by e2fsck.) 5892 5893Debugfs can now set and print filesystem features in the superblock 5894using the "features" command. Dumpe2fs will print out the complete 5895set of features when listing the superblock. 5896 5897Dumpe2fs has new options -f (force) and -h (header-only). 5898 5899Fixed a bug in e2fsck which could cause the PROGRAMMING ERROR/bonehead 5900message to come up. This could happen when decrementing or 5901incrementing a link count could result in an overflow. 5902 5903Fixed a bug in e2fsck where the block count on the lost+found 5904directory would not be properly incremented when the directory was 5905expanded to the point where an indirect block needed to be allocated. 5906 5907E2fsck now makes some additional sanity checks on the superblock to 5908avoid crashing or giving a memory allocation error if some of the 5909values in the superblock are unresonable (but the superblock otherwise 5910looks valid). 5911 5912Fixed a bug in e2fsck where a very badly corrupted filesystem might 5913require two passes to completely fix the filesystem. This happened if 5914an inode claimed blocks that was part of the filesystem metadata 5915(typically, when garbage was written into an inode table or indirect 5916block, since this kind of filesystem corruption normally doesn't 5917happen otherwise). 5918 5919On the Alpha, glibc declares st_flags although it isn't actually used; 5920the configure script was improved to detect this case so that 5921e2fsprogs can avoid using the non-functional stat field. 5922 5923The manual pages were updated to use a more consistent formatting 5924style consistent with standard Unix man pages. Mke2fs's man page 5925added documentation for a few previously undocumented options. 5926 5927Fixed minor display bugs in tune2fs and mke2fs. 5928 5929Programmer's notes: 5930------------------- 5931 5932Improved portability of e2fsprogs to non-Unix systems (in particular, NT). 5933 5934Added features to parse and print feature strings into the e2p library. 5935(e2p_feature2string, e2p_string2feature, e2p_edit_feature). 5936 5937ext2fs_mkdir() and ext2fs_new_dir_block() now creates directories 5938whose directory entries contain proper filetype information if the 5939filesystem supports it. 5940 5941ext2fs_link() now uses the low 3 bits of its flags parameter to pass 5942the directory entry filetype information. This is used to set the 5943directory entry filetype information if the filesystem supports it. 5944 5945Fixed a bug in ext2fs_expand_dir() where the block count in a 5946directory's inode would not be properly incremented when the directory 5947was expanded to the point where an indirect block needed to be 5948allocated. 5949 5950 5951E2fsprogs 1.15 (July 18, 1999) 5952============================== 5953 5954Add configuration checks so that e2fsprogs will compile cleanly on 5955Linux 2.3 kernels that have renamed i_version to i_generation. 5956 5957E2fsck now prints a progress/completion bar (and not just a simple 5958spinner) if the -C0 option is requested or if it receives a SIGUSR1 5959signal. Fsck will automatically manage the (potentially muliple) 5960e2fsck processes to print completion bars if it is given a -C option, 5961with the right thing happening if multiple filesystems are being 5962checked in parallel. 5963 5964Mke2fs now has better automatic hueristics to determine the filesystem 5965parameters to be used for a particular filesystem. Added a new option 5966-T which allows the user to specify how the filesystem is to be used, 5967which helps mke2fs do a better job selecting the filesystem parameters. 5968 5969Mke2fs now creates revision 1 filesystems by default, and with the 5970sparse superblock feature enabled. The sparse superblock feature is 5971not understood by Linux 2.0 kernels, so they will only allow read-only 5972mounts of filesystems with this sparse superblocks. 5973 5974Fix bug where if /dev/null couldn't be opened (should never happen), 5975e2fsck would hang in a tight loop. 5976 5977Make e2fsck handle the case where /lost+found isn't a directory. 5978 5979E2fsck now uses mallinfo if it exists to get accurate statistics about 5980its memory usage. 5981 5982Fix bug in e2fsck where it wouldn't check to see if a disconnected 5983inode had any problems before connecting it to /lost+found. 5984 5985Add check to e2fsck so it makes sure that total number of inodes in 5986the filesystem is a sane number. 5987 5988Fix fencepost error when clearing an the end of the block bitmap which 5989caused the last block in the bitmap not to get cleared. 5990 5991Cleaned up a number of messages in e2fsck: 5992 * The message "Group's #'s copy of the group descriptor..." 5993 was fixed so that the correct number would be displayed. 5994 * Added missing space in the "disk write-protected" error messsage 5995 * Cleaned up the error message printed when a non-interactive 5996 e2fsck needs to abort a check because the filesystem 5997 appears to be mounted. 5998 5999Added a new command-line utility, uuidgen, which will create and print 6000a UUID. 6001 6002Make debugfs's icheck command more robust by checking to make sure an 6003inode has valid blocks before interarting over the inode's blocks. 6004 6005UUID generation now uses a random-based scheme whenever possible to 6006prevent potential privacy problems. 6007 6008Man pages for all of the UUID functions in the lirbary were added. 6009 6010Fixed bug in fsck so it won't coredump if a filesystem not in 6011/etc/fstab is given to it. 6012 6013Fsck now understands the UUID=xxxx and LABEL=yyyy forms in /etc/fstab 6014that most of the other mount utilities understands. 6015 6016Mke2fs will make a filesystem even if it appears mounted if the force 6017option is given. 6018 6019Dumpe2fs has new command-line options which allow a filesystem expert 6020to specify the superblock and blocksize when opening a filesystem. 6021This is mainly useful when examining the remains of a toasted 6022filesystem. 6023 6024The badblocks program has been updated to display correctly on disks 6025with large block numbers. 6026 6027The badblocks program no longer gives spurious errors when errors 6028occur on non-block boundaries, which is common if the blocksize is 6029larger than 1k. 6030 6031Mke2fs will sync the disk device every MKE2FS_SYNC block groups if the 6032MKE2FS_SYNC environment variable is set. This is to work around a VM 6033bug in the 2.0 kernel. I've heard a report that a RAID user was able 6034to trigger it even using a 2.2 kernel, but hopefully it will not be 6035needed for most Linux 2.2 users. 6036 6037Fixed miscellaneous documentation and man pages. 6038 6039Programmer's notes: 6040------------------- 6041 6042Cleaned up functions such as pass1_get_blocks, pass1_read_inode which 6043in e2fsck's pass1.c really should have been static. 6044 6045The return value of the uuid_compare() function was changed to make it 6046match with the convetions used by strcmp, memcmp, and Paul Leach's 6047UUID sample document. 6048 6049The "make depend" process has now been made more automated; it now 6050automatically word-wraps the dependencies, and only replaces source 6051Makefile.in if there has been a change in the dependencies. Also, a 6052top-level "make depend" now recurses through all the subdirectories 6053automatically. 6054 6055The Makefile in .../util has been changed so that subst is built using 6056the native C compiler during a cross-compilation, since the subst 6057program is only used during the build process. Also add an explicit 6058rule to build util/subst by cd'ing to the correct directory and 6059running Makefile. 6060 6061The man directories are defined in terms mandir, so that the configure 6062script can override the location of the manual pages. 6063 6064The config files have been updated to recognize new machine types for 6065both the i386 and alpha families. 6066 6067Fsck has been modified so that it will accurately create an 6068fsck_instance even when the noexecute flag is set. This allows for 6069accurate debugging of the fsck pass structure. Also, when the verbose 6070flag is given twice, fsck will print debugging information about when 6071fsck is waiting for jobs to finish. 6072 6073 6074E2fsprogs 1.14 (January 9, 1999) 6075================================ 6076 6077Fix the fstab parsing code so that it can handle blank lines and 6078comment characters. Also, missing pass numbers need to be treated as 6079zero. 6080 6081Fixed a bug in e2fsck where under some circumstances (when e2fsck 6082needs to restart processing after fixing an egregious inconsistency) 6083it would try to access already freed memory. 6084 6085E2fsck now prints non-printable characters in directory entries and 6086pathnames using '^' and 'M-' notation. 6087 6088Fixed chattr so that it will ignore symbolic links when doing 6089recursive descent traversals. For both chattr and lsattr, no longer 6090print the version string unless the -V option is given. 6091 6092Allow the system administrator to directly specify the number of 6093inodes desired in the filesystem, for some special cases where this is 6094necessary. 6095 6096Fix portability problems so that e2fsprogs can be compiled under Linux 60971.2 systems and Solaris systems. 6098 6099Update the config.guess file with a more recent version that will 6100identify newer Linux platforms. 6101 6102Programmer's notes 6103------------------ 6104 6105Ext2fs_read_inode and ext2fs_write_inode will now return an error if 6106an inode number of zero is passed to them. 6107 6108E2fsprogs 1.13 (December 15, 1998) 6109================================== 6110 6111Fixed a bug in debugfs where an error messages weren't getting printed 6112when the ext2 library routines to read inodes returned errors in the 6113stat, cmri and rm commands. 6114 6115Fixed a bug in mke2fs so that if a ridiculous inode ratio parameter is 6116provided, it won't create an inode table smaller than the minimum 6117number of inodes required for a proper ext2 filesystem. 6118 6119Fsck now parses the /etc/fstab file directly (instead of using 6120getmntent()), so that it can distinguish between a missing pass number 6121field and pass number field of zero. This caused problems for 6122diskless workstations where all of the filesystems in /etc/fstab have 6123an explicit pass number of zero, and fsck could not distinguish this 6124from a /etc/fstab file with missing pass numbers. 6125 6126E2fsck will create a /lost+found directory if there isn't one in the 6127filesystem, since it's safer to create the lost+found directory before 6128it's needed. 6129 6130Fixed e2fsck so that it would detect bogus immutable inodes which 6131happen to be sockets and FIFO files, and offer to clear them. 6132 6133If a filesystem has multiple reasons why it needs to be checked, and 6134one of the reasons is that it is uncleanly mounted, e2fsck will print 6135that as the reason why the filesystem is being checked. 6136 6137Cleaned up the output routines of mke2fs so that it doesn't overflow 6138an 80 column display when formating really big filesystems. 6139 6140Added a sanity check to e2fsck to make sure that file descriptors 0, 61411, 2 are open before opening the hard disk. This avoids a problem 6142where a broken program might exec e2fsck with those file descriptors 6143closed, which would cause disastrous results if the kernel returns a 6144file descriptor for the block device which is also used by FILE * 6145stdout. 6146 6147Fixed up the e2fsck progress reporting functions so that the values 6148reliably reach 100% at the completion of all of the e2fsck passes. 6149 6150Fixed minor documentation bugs in man pages and usage messages. 6151 6152Programmer's notes: 6153------------------- 6154 6155Fixed a number of lint warnings in the ext2fs library and potential 6156portability problems from other OS's header files that might define 6157CPP macros for names like "max" and "min". 6158 6159ext2fs_badblocks_list_add() has been made more efficient when it needs 6160to grow the bad blocks list. 6161 6162Fixed a bug in e2fsck which caused it to dereference a freed pointer 6163just before exiting. 6164 6165Fixed the substition process for generating the mk_cmds and compile_et 6166scripts so that they will work outside of the build tree. 6167 6168Add sanity check to e2fsck so that if an internal routine 6169(ext2fs_get_dir_info) returns NULL, avoid dereferencing the pointer 6170and causing a core dump. This should never happen, but... 6171 6172E2fsprogs 1.12 (July 9, 1998) 6173================================== 6174 6175E2fsprogs now works with glibc (at least with the version shipped wtih 6176RedHat 5.0). The ext2fs_llseek() function should now work even with 6177i386 ELF shared libraries and if llseek() is not present. We also 6178explicitly do a configure test to see if (a) llseek is in libc, and 6179(b) if llseek is declared in the system header files. (See standard 6180complaints about libc developers don't understand the concept of 6181compatibility with previous versions of libc.) 6182 6183The ext2fs library now writes out the block group number in each of 6184the superblock copies. This makes it easier to automatically 6185determine the starting block group of the filesystem when the block 6186group information is trashed. 6187 6188Added support for the EXT2_FEATURE_INCOMPAT_FILETYPE feature, 6189which means that e2fsprogs will ignore the high 8 bits of the 6190directory entry's name_len field, so that it can be used for other 6191purposes. 6192 6193Added support for the EXT2_FEATURE_RO_COMPAT_LARGE_FILE feature. 6194E2fsprogs will now support filesystems with 64-bit sized files. 6195 6196Added support for the EXT2_FEATURE_COMPAT_DIR_PREALLOC feature. 6197 6198Added new program "e2label", contributed by Andries Brouwer. E2label 6199provides an easy-to-use interface to modify the filesystem label. 6200 6201Fixed bug so that lsattr -v works instead of producing a core dump. 6202 6203Fixed a minor bug in mke2fs so that all groups with bad superblock 6204backup blocks are printed (not just the first one). 6205 6206Mke2fs will check the size of the device, and if the user specifies a 6207filesystem size larger than the apparent size of the device it will 6208print a warning message and ask if the user wants to proceed. 6209 6210E2fsck has a new option -C, which sends completion information to the 6211specified file descriptor. For the most part, this is intended for 6212programs to use, although -C 0 will print a spinning character to the 6213stdout device, which may be useful for users who want to see something 6214happening while e2fsck goes about its business. 6215 6216Fixed a bug in e2fsck which could cause a core dump when it needs to 6217expand the /lost+found directory, and sometimes the bitmaps haven't 6218been merged in. Also fixed a related bug where ext2fs_write_dir_block 6219was used to write out a non-directory block. (Which would be bad on a 6220non-Intel platform with byte swapping going on.) 6221 6222Fixed bug in e2fsck where it would print a "programming error" message 6223instead of correctly identifying where a bad block was in used when 6224the bad block was in a non-primary superblock or block group 6225descriptor. Also fixed a related bug when sparse superblocks are in 6226use and there is a bad block where a superblock or block group 6227descriptor would have been in a group that doesn't include a 6228superblock. 6229 6230Fixed a bug in e2fsck (really in libext2fs's dblist function) where if 6231the block group descriptor table is corrupt, it was possible to try to 6232allocate a huge array, fail, and then abort e2fsck. 6233ext2fs_get_num_dirs() now sanity checks the block group descriptor, 6234and subsitutes reasonable values if the descriptors are obviously bogus. 6235 6236If e2fsck finds a device file which has the immutable flag set and the 6237i_blocks beyond the normal device number are non-zero, e2fsck will 6238offer to remove it, since it's probably caused by garbage in the inode 6239table. 6240 6241When opening a filesystem, e2fsck specially checks for the EROFS error 6242code, and prints a specific error message to the user which is more 6243user friendly. 6244 6245If the filesystem revision is too high, change the e2fsck to print 6246that this is either because e2fsck is out of date, or because the 6247superblock is corrupt. 6248 6249E2fsck now checks for directories that have duplicate '.' and '..' 6250entries, and fixes this corruption. 6251 6252E2fsck no longer forces a sync of the filesystem (with attendant sleep 6253calls) at all times. The ext2fs_flush() function now performs a sync 6254only if it needed to write data blocks to disk. 6255 6256Fixed a minor bug in e2fsck's pass1b's file cloning function, where 6257certain errors would not be properly reported. 6258 6259Updated and expanded a few points in the man pages which users 6260complained wheren't explicit enough. 6261 6262Added special case byte-swapping code if compiling on the PowerPC, to 6263accomodate the strange big-endian variant of the ext2 filesystem that 6264was previously used on the PowerPC port. 6265 6266 6267Programmer's notes: 6268------------------- 6269 6270Removed C++ keywords from the ext2fs libraries so that it could be 6271compiled with C++. 6272 6273E2fsck's internal organization has now been massively reorganized so 6274that pass*.c don't have any printf statements. Instead, all problems 6275are reported through the fix_problem() abstraction interface. E2fsck 6276has also been revamped so that it can be called as a library from a 6277application. 6278 6279Added new fileio primitives in libext2fs for reading and writing 6280files on an unmounted ext2 filesystem. This interface is now used by 6281debugfs. 6282 6283Added a new libext2fs function for mapping logical block numbers of 6284a file to a physical block number. 6285 6286Added a new libext2fs function, ext2fs_alloc_block(), which allocates 6287a block, zeros it, and updates the filesystem accounting records 6288appropriately. 6289 6290Added a new libext2fs function, ext2fs_set_bitmap_padding(), which 6291sets the padding of the bitmap to be all one's. Used by e2fsck pass 5. 6292 6293The libext2fs functions now use a set of memory allocation wrapper 6294functions: ext2fs_get_mem, ext2fs_free_mem, and ext2fs_resize_mem, 6295instead of malloc, free, and resize. This makes it easier for us to 6296be ported to strange environments where malloc, et. al. aren't 6297necessarily available. 6298 6299Change the libext2fs fucntion to return ext2-specific error codes 6300(EXT2_DIR_EXISTS and EXT2_DB_NOT_FOUND, for example) instead of using 6301and depending on the existence of system error codes (such as EEXIST 6302and ENOENT). 6303 6304Renamed io.h to ext2_io.h to avoid collision with other OS's header 6305files. 6306 6307Add protection against ext2_io.h and ext2fs.h being included multiple 6308times. 6309 6310The types used for memory lengths, etc. have been made more portable. 6311In generla, the code has been made 16-bit safe. Added Mark 6312Habersack's contributed DOS disk i/o routines. 6313 6314Miscellaneous portability fixes, including not depending on char's 6315being signed. 6316 6317The io_channel structure has a new element, app_data, which is 6318initialized by the ext2fs routines to contain a copy of the filesystem 6319handle. 6320 6321ext2fs_check_directory()'s callback function may now return the error 6322EXT2_ET_CALLBACK_NOTHANDLED if it wishes ext2fs_check_directory() to 6323really do the checking, despite the presence of the callback function. 6324 6325 6326E2fsprosg 1.11 (June 17, 1997) 6327============================== 6328 6329Fixed e2fsck to detect (previously ignored) conflicts between the 6330superblock or block group descriptors and block bitmaps, inode 6331bitmaps, and inode tables. 6332 6333Fixed bug in e2fsck so that when the message printed out when a block 6334or inode bitmap conflicts with other data, it has the correct group 6335number. 6336 6337Fixed bug in e2fsck and mke2fs where the blocksize wasn't being passed 6338to badblocks. This meant that not all of the filesystem was being 6339tested for bad blocks! 6340 6341Fixed an array boundary overrun case which cropped up in 6342ext2fs_badblocks_list_test when a user tried running "mke2fs -c 6343-b 4096". 6344 6345Adjusted the number of columns printed by mke2fs when displaying the 6346superblock backups to avoid running over 80 columns when making a 6347really big filesystem. 6348 6349Fixed up the man pages for e2fsck, debugfs, badblocks, chattr, 6350dumpe2fs, fsck, mke2fs, and tune2fs (typos and other minor grammar 6351fixes), thanks to some suggestions from Bill Hawes (whawes@star.net). 6352 6353Programmer's notes: 6354------------------- 6355 6356Fixed install rule in lib/ss so that ss_err.h is actually getting 6357installed. 6358 6359Fixed bug in ext2fs_copy_bitmap; the destination bitmap wasn't getting 6360bassed back to the caller. 6361 6362Fixed bug in ext2fs_inode_scan_goto_blockgroup; it had not been 6363setting the current inode number (which meant this function wasn't 6364working at all). 6365 6366Fixed bug in ext2fs_resize_generic_bitmap; it had not be zeroing all 6367blocks in the bitmap when increasing the size of the bitmap. 6368 6369Changed the initial number of blocks allocated by ext2fs_init_dblist() 6370to be more realistic. 6371 6372Added a new function ext2fs_allocate_group_table, which sets up the 6373group descriptor information (and allocates inode and block bitmaps, 6374and inode tables for a particular group). The function was created by 6375factoring out code form ext2fs_allocate_tables(). 6376 6377Added a new function ext2fs_move_blocks which takes a bitmap of the 6378blocks to be moved, and moves them to another location on the 6379boardboard. 6380 6381Make the unix_io channel's io_channel_flush implementation calls sync() 6382to to flush the kernel buffers to disk. 6383 6384Added a new function ext2fs_dblist_count returns the number of 6385directory blocks in dblist. 6386 6387 6388E2fsprogs 1.10 (April 24, 1997) 6389=============================== 6390 6391Mke2fs once again defaults to creating revision #0 filesystems, since 6392people were complaining about breaking compatibility with 1.2 kernels. 6393Warning messages were added to the mke2fs and tune2fs man pages that 6394the sparse superblock option isn't supported by most kernels yet (1.2 6395and 2.0 both don't support parse superblocks.) 6396 6397Added new flag to mke2fs, -R <raid options>, which allows the user to 6398tell mke2fs about the RAID configuration of the filesystem. Currently 6399the only supported raid option is "stride" which specifies the width 6400of the RAID stripe. 6401 6402Fixed bug in e2fsck where pass1b would bomb out if there were any 6403blocks marked bad in the inode table. 6404 6405Fixed rare bug in mke2fs where if the user had a very unlucky number 6406of blocks in a filesystem (probability less than .002) the resulting 6407filesystem would be corrupt in the last block group. 6408 6409Fixed bug where if e2fsck tried to allocate a block to fix a 6410filesystem corruption problem and the filesystem had no free blocks, 6411ext2fs_new_block() would loop forever. 6412 6413The configure script now checks explicitly to see if "-static" works, 6414since that can't be assumed to be true --- RedHat doesn't install 6415libc-static by default. 6416 6417Fixed bug in libext2's block iterator functions where under some 6418cirmcustances, file with holes would cause the bcount parameter to the 6419callback function to be incorrect. This bug didn't affect any of 6420e2fsprogs programs, but it was discovered by Paul Mackerras, the 6421author of the PPC boot loader. 6422 6423Removed use of static variables to store the inode cache in libext2fs. 6424This caused problems if more than one filesystem was accessed via 6425libext2fs (static variables in libraries are generally a bad idea). 6426Again, this didn't affect e2fsprogs programs, but it was discovered by 6427Paul Mackerras. 6428 6429Fixed minor bugs and version code drift to assure that e2fsprogs 1.10 6430will compile cleanly with 1.2.13 kernels (even with a.out shared 6431libraries!) 6432 6433Programmer's notes: 6434------------------- 6435 6436Added new functions to duplicate an ext2 filesystem handle, and its 6437associated substructure. New functions: ext2fs_dup_handle(), 6438ext2fs_copy_dblist(), ext2fs_badblocks_copy(), ext2fs_copy_bitmap(). 6439Other structures, such as the io_channel and the inode_cache, now have 6440a ref count so that they only get freed when they are no longer used 6441by any filesystem handle. (These functions were added as part of the 6442development effort for an ext2 resizer). 6443 6444E2fsprogs 1.09 (April 14, 1997) 6445=============================== 6446 6447Fixed bug in mke2fs (really in lib/ext2fs/initialize.c) which was 6448accidentally introduced in the 1.08 release. The overhead calculation 6449was accidentally removed, which caused ext2fs_initialize() to not 6450notice when the filesystem size needed to be adjusted down because 6451there wasn't enough space in the last block group. 6452 6453Fixed bug in version parsing library routine; it was always parsing 6454the library version string, instead of using the passed-in string. 6455 6456Clarified chattr man page. 6457 6458E2fsprogs 1.08 (April 10, 1997) 6459=============================== 6460 6461E2fsck 1.07 was very slow when checking very large filesystems with a 6462lot of files that had hard links (i.e., news spools). This was fixed 6463by seriously revamping the icount abstraction. Added a formal test 6464suite for the icount abstraction. 6465 6466Debugfs now has a "-l" option to the "ls" command, which lists the 6467inode number, permissions, owner, group, size, and name of the files 6468in the directory. 6469 6470Fix a bug in e2fsck where when a directory had its blocks moved to 6471another location during the pass 1b processing, the directory block 6472list wasn't updated, so pass 2 wouldn't check (and correct) the 6473correct directory block. 6474 6475E2fsck will now treat inodes which contain blocks which are claimed by 6476the filesystem metadata by treating them as multiply claimed blocks. 6477This way, the data in those blocks can be copied to a new block during 6478the pass 1b--1d processing. 6479 6480E2fsck will attempt to determine the correct superblock number and 6481display it in the diagnostic and warning messages if possible. 6482 6483Add support for a new (incompatible) feature, "sparse_super". This 6484feature reduces the number of blocks which contain copies of backup 6485superblocks and block group descriptors. (It is only an incompatible 6486feature because of a bug in ext2_free_blocks.) mke2fs and tune2fs now 6487support a new -s option; e2fsck will recognize filesystems built with 6488this feature turned on. 6489 6490E2fsck now checks the library to make sure is the correct version, 6491using new library functions. (This helps to diagnose incorrectly 6492installed e2fsprogs distributions.) 6493 6494Dumpe2fs now prints more information; its now prints the the 6495filesystem revision number, the filesystem sparse_super feature (if 6496present), the block ranges for each block group, and the offset from 6497the beginning of the block group. 6498 6499Mke2fs now distributes the inode and block bitmap blok so that the 6500won't be concentrated in one or two disks in RAID/striping setups. 6501Also, if the user chooses a 2k or 4k block group, mke2fs will try to 6502choose the largest blocks per group that be chosen. (For 2k blocks, 6503you can have up to 16384 blocks/group; for 4k blocks, you can have up 6504to 32768 blocks/group.) Previously mke2fs would not allow 6505specification of more than 8192 blocks per group, even if you were 6506using a 2k or 4k block group. 6507 6508Programmer's notes: 6509------------------- 6510 6511Added a new function ext2fs_create_icount2() which takes a "hint" 6512argument. This hint argument presets the icount array with the list 6513of inodes which actually need to be in the icount array. This really 6514helps to speed up e2fsck. 6515 6516Added a new function ext2fs_icount_validate() which checks the rep 6517invariant for the icount structure. This is used mostly for testing. 6518 6519The error mesasage given when a bad inode number is passed to 6520test_generic_bitmap to reflect EXT2FS_TEST_ERROR (instead of 6521EXT2FS_UNMARK_ERROR). 6522 6523Added a new function ext2fs_set_dir_block which sets the block of a 6524dblist entry, given the directory inode and blockcnt. 6525 6526Added a new function ext2fs_get_library_version() which returns the 6527current library version, and ext2fs_parse_version_string() which 6528returns a version number based on a e2fsprogs version string. 6529 6530The icount functions will return EINVAL if the passed in inode number 6531is out of bounds. 6532 6533E2fsprogs 1.07 (March 9, 1997) 6534============================== 6535 6536E2fsck is now uses much less memory when checking really large 6537filesystems (or rather, filesystems with a large number of inodes). 6538Previously a filesystem with 1 million inodes required 4 megabytes of 6539memory to store inode count statistics; that storage requirement has 6540now been reduced to roughly half a megabyte. 6541 6542E2fsck can now properly deal with bad blocks appearing inside the 6543inode table. Instead of trying to relocate the inode table (which 6544often failed because there wasn't enough space), the inodes in the bad 6545block are marked as in use. 6546 6547E2fsck will automatically try to use the backup superblocks if the 6548primary superblocks have a bad magic number or have missing meta-data 6549blocks (or meta-data blocks which are out of range). 6550 6551E2fsck's pass 3 has been made more efficient; most noticeable on 6552filesystems with a very large number of directories. 6553 6554Completely revamped e2fsck's system of printing problem reports. It 6555is now table driven, to make them more easily customizeable and 6556extendable. Error messages which can be printed out during preen mode 6557are now one line long. 6558 6559Fixed e2fsck's filesystem swapping code so that it won't try to swap 6560fast symbolic links or deleted files. 6561 6562Fixed e2fsck core dumping when fixing a filesystem which has no 6563directories (not even a root directory). 6564 6565Added a check to e2fsck to make sure that the length of every 6566directory entry is a multiple of 4 (since the kernel complains if it 6567isn't). 6568 6569Added a check to e2fsck to make sure that a directory entry isn't a 6570link to the root directory, since that isn't allowed. 6571 6572Added a check to e2fsk to now make sure the '.' and '..' directory 6573entries are null terminated, since the 2.0 kernel requires it. 6574 6575Added check to write_bitmaps() to make sure the superblock doesn't get 6576trashed if the inode or block bitmap is marked as being block zero. 6577 6578Added checking of the new feature set fields in the superblock, to 6579avoid dealing with new filesystem features that this package wasn't 6580set up to handle. 6581 6582Fixed a fencepost error in ext2fs_new_block() which would occasionally 6583try to allocate a block beyond the end of a filesystem. 6584 6585When the UUID library picks a random IEEE 802 address (because it 6586can't find one from a network card), it sets the multicast bit, to 6587avoid conflicting with a legitimate IEEE 802 address. 6588 6589Mke2fs now sets the root directory's owner to be the real uid of the 6590user running mke2fs. If the real uid is non-zero, it also sets 6591the group ownership of the root directory to be the real group-id of 6592the user running mke2fs. 6593 6594Mke2fs now has more intelligent error checking when it is given a 6595non-existent device. 6596 6597When badblocks is given the -vv option, it now updates the block that 6598it is currently testing on every block. 6599 6600Fixed a bug in fsck where it wouldn't modify the PATH envirnoment 6601currently correctly if PATH wasn't already set. 6602 6603Shared libraries now built with dependencies. This allows the shared 6604library files to be used with dlopen(); it also makes the transition 6605to libc 6 easier, since ld.so can tell which libc a particular shared 6606library expects to use. 6607 6608Programmer's notes: 6609------------------- 6610 6611Added new abstraction (defined in dblist.c) for maintaining a list of 6612blocks which belongs to directories. This is used in e2fsck and other 6613programs which need to iterate over all directories. 6614 6615Added new functions which test to see if a contiguous range of blocks 6616(or inodes) are available. (ext2fs_*_bitmap_range). 6617 6618Added new function (ext2_inode_has_valid_blocks) which returns true if 6619an inode has valid blocks. (moved from e2fsck code). 6620 6621Added new function (ext2fs_allocate_tables) which allocates the 6622meta-data blocks as part of initializing a filesystem. (moved from 6623mke2fs code). 6624 6625Added a new I/O manager for testing purposes. It will either allow a 6626program to intercept I/O requests, or print debugging messages to 6627trace the activity of a program using the I/O manager. 6628 6629The badblocks_list functions now store the bad blocks in a sorted 6630order, and use a binary search to speed up badblocks_list_test. 6631 6632The inode scan function ext2fs_get_next_inode() may now return a soft 6633error returns: MISSING_INODE_TABLE and BAD_BLOCK_IN_INODE_TABLE in 6634those cases where part of an inode table is missing or there is a bad 6635block in the inode table. 6636 6637Added a new function (ext2fs_block_iterate2) which adds new arguments to 6638the callback function to return a pointer (block and offset) to the 6639reference of the block. 6640 6641Added new function (ext2fs_inode_scan_goto_blockgroup) which allows an 6642application to jump to a particular block group while doing an inode 6643scan. 6644 6645The badblocks list functions were renamed from badblocks_* to 6646ext2fs_badblocks_*. Backwards compatibility functions are available 6647for now, but programs should be modified to use the new interface. 6648 6649Some of the library functions were reorganized into separate files to 6650reduce the size of some programs which statically link against the 6651ext2 library. 6652 6653Put in some miscellaneous fixes for the Alpha platform. 6654 6655 6656E2fsprogs 1.06 (October 7, 1996) 6657================================ 6658 6659Fixed serious bug in e2fsck: if the block descriptors are bad, don't 6660smash the backup copies in ext2fs_close(). (The problem was that when 6661e2fsck -p discovered the problem, while it was closing the filesystem 6662and exiting, it was also blowing away the backup superblocks on the 6663disk, which was less than friendly.) We now make it the case that we 6664only write out the backup superblock and the back block descriptors if 6665the filesystem is completely free from problems. 6666 6667Fixed a bug in block_interate in the lib/ext2fs library which caused 6668e2fsck to fail on GNU Hurd-created filesystems. 6669 6670Add support for Linux/FT's bootloader, which actually uses 6671EXT2_BOOT_LOADER, and sets its mode bits which caused e2fsck to want 6672to clear the inode. 6673 6674Add support for the "A" (no atime update) attribute. (Note: this 6675attribute is not yet in production kernels.) 6676 6677The test suite is not automatically run when doing a "make all" from 6678the top level directory. Users should manually run "make check" if 6679they wish to run the test suite. 6680 6681Upon a preenhalt(), make the printed message more explicit that 6682running e2fsck "MANAULLY" means without the -p or -a options. 6683 6684In e2fsck, if a disconnected inode is zero-length, offer to clear it 6685instead of offering to connect it to lost+found. 6686 6687In e2fsck, if a filesystem was just unmounted uncleanly, and needs 6688e2fsck to be run over it, change e2fsck to explicitly display this 6689fact. 6690 6691For dumpe2fs and e2fsck, cause the -V option to print out which 6692version of the ext2fs library is actually getting used. (This will 6693help detect mismatches of using a 1.06 utility with a 1.05 library, 6694etc.) 6695 6696Programmers' notes: 6697------------------- 6698 6699EXT2_SWAP_BYTES was changed to EXT2_FLAG_SWAP_BYTES, which better fits 6700the naming convention. 6701 6702In ext2fs_initialize(), make sure the description for the inode bitmap 6703is correctly initialize. 6704 6705Fixed minor type typo in ext2fs_allocate_generic_bitmap(); 6706 6707E2fsprogs 1.05 (September 7, 1996) 6708================================== 6709 6710Add support for new fields in the ext2 superblock --- volume name, 6711volume UUID, and last mounted field. Dumpe2fs displays these fields, 6712tune2fs and mke2fs allows you to set them. E2fsck will automatically 6713generate a UUID for those volumes that don't have them. 6714 6715Put in support for e2fsck to recognize HURD specific ext2 features --- 6716most notably, the translator block. The e2fsprogs tools will now use 6717the creator_os field in the superblock to correctly handle different 6718OS-specific variants of the ext2 filesystem. 6719 6720E2fsck now fixes inodes which have a the deletion time set, but which 6721have a non-zero i_link_count field by offering to clear the deletion 6722time. Previously e2fsck assumed that the inode was deleted (per 0.3c 6723ext2 kernel behavior) and offered to unlink the file. 6724 6725If e2fsck sets the clean bit, but nothing else, set the exit code 6726FSCK_NONDESTRUCT. After all, e2fsck did fix a filesystem error --- it 6727set the filesystem valid bit when it was previously cleared. :-) This 6728was needed to make the HURD fsck driver happy. 6729 6730If the user refuses to attach an unattached inode, e2fsck will no 6731longer set the inode's link count. Otherwise, the inode would end up 6732getting marked as unused, which might cause loss of data later. 6733 6734Make the message issued by e2fsck when the superblock is corrupt less 6735confusing for users. It now mentions that another reason for the 6736"corrupt superblock" message might be that the partition might not be 6737an ext2 filesystem at all (it might swap, msdos filesystem, ufs, etc.) 6738 6739Make the libext2 library more robuest so that e2fsck won't coredump on 6740an illegal superblock where the blocksize is zero. (f_crashdisk is 6741the test case). 6742 6743By default, create filesystems where the default checkinterval is 6 6744months (180 days). Linux servers can be robust enough that 20 reboots 6745can be a long, long time. 6746 6747Added configure flag --enable-old-bitops, which forces the bitops to 6748use the old (native) bitmask operations. By default on the sparc 6749platform, the standard ext2 bit ordering is now used. 6750 6751Added a new feature to e2fsck to byte-swap filesystems; this can be 6752used to convert old m68k filesystems to use the standard byte-order 6753storage for the superblock, inodes, and directory blocks. This 6754function is invoked by using the '-s' option to e2fsck. 6755 6756Debugfs's "dump" command has been enhanced so that it writes out the 6757exact size of the file so that the nulls at the end of the file are 6758eliminated. The command also accept a new "-p" option which will 6759attempt preserve to preserve the ownernship, permissions, and 6760file modification/access times. 6761 6762Debugfs has two new options, -f and -R. The -R option allows the user 6763to execute a single debugfs command from the command line. The -f 6764option allows the user to specify a "command file" containing debugfs 6765commands which will get executed. 6766 6767Dumpe2fs now pretty prints the check interval, instead of just 6768printing the check interval as a number of seconds. 6769 6770Fix bugs in debugfs: the params command when no filesystem is opened 6771no longer causes a core dump. It is now possible to unlink a file 6772when a pathame containing a '/' is specified. 6773 6774Tune2fs has a new -C option which sets the number of times the 6775filesystem has been mounted. 6776 6777Fix the chattr '-v' option so that it actually works. Chattr was 6778being buggy about the -v option parsing. 6779 6780Programmers' notes: 6781------------------- 6782 6783The directory lib/uuid contains a set of library routines to generate 6784DCE compatible UUIDs. 6785 6786Extended ext2fs_namei() to handle symbolic links. Added new function 6787ext2fs_nami_follow() which will follow last symbolic link in the case 6788where the pathname points to a sym link. 6789 6790The ext2fs_block_iterate function will now return the HURD translator 6791block, if present. The new flag BLOCK_FLAG_DATA_ONLY will cause the 6792iterator to return data blocks only. The ext2fs.h file now defines 6793constants BLOCK_COUNT_IND, BLOCK_COUNT_DIND, BLOCK_COUNT_TIND, and 6794BLOCK_COUNT_TRANSLATOR, which are the magic values passed in the block 6795count field of the iterator callback function. 6796 6797The test script driver now takes an optional second argument, which is 6798the test case to be run. This allows you to run a test case without 6799needing to run the entire test suite. 6800 6801On Linux ELF systems, install the .so files in the correct places 6802(/usr/lib). The .so files must be stored in the same directory as the 6803.a files. 6804 6805Fixed miscellaneous HURD compilation issues with header file being 6806included in the right order. 6807 6808Fixed debugfs so that it resets optind to zero, not one, since setting 6809optind to zero is more correct. 6810 6811 6812E2fsprogs 1.04 (May 16, 1996) 6813============================= 6814 6815First "official" (1.03 was a limited release only) to support building 6816e2fsprogs under Linux 2.0 kernels (as well as late model 1.3 and 1.99 6817kernels). 6818 6819This package includes a RPM specs file, that it can be built using the 6820RedHat Package Manager. 6821 6822E2fsck now prints a hint that if there are lots of incorrectly located 6823inode bitmaps, block bitmaps, and inode table blocks, the user might 6824want to try using e2fsck -b 8193 first, to see if that fares any 6825better. 6826 6827For ext2 filesystem written with the hurd, debugfs will now print out 6828the translator field when printing an inode structure. 6829 6830Lots of miscellaneous linking/installation cleanups: 6831 6832 Libraries are now linked using a relative pathname, instead of 6833 relying on -L working correct. It doesn't, in many cases, including 6834 current versions of GNU ld. This guarantees that the build tree is 6835 linking with the right libraries, instead of the ones installed in 6836 /usr/lib. 6837 6838 Header files, man pages, and the et/ss shell scripts are now 6839 generated using a custom substitution script, instead of relying on 6840 the configure script. This prevents needless recompilation of 6841 files; in addition, the custom substitution script is much faster. 6842 6843 e2fsck may now be linked dynamically, by using the 6844 --enable-dynamic-e2fsck flag to configure. This is not recommended, 6845 since it increases e2fsck's dependence on other files, but some 6846 people need to save disk space, and other critical programs on their 6847 systems were being linked dynamically anyway. 6848 6849 Programs such as fsck which didn't need to be linked against 6850 libext2fs (or mke2fs which didn't need to be linked against libe2p) 6851 only link against libraries they actually need. Otherwise, those 6852 programs would require the presense of libraries that otherwise 6853 could be removed from a rescuse diskette. 6854 6855 The ss include files are now installed correctly so they can 6856 actually be used by another package. 6857 6858 If the profiling libraries are built, they are now installed on a 6859 "make install-libs". 6860 6861 6862E2fsprogs 1.03 (March 27, 1996) 6863=============================== 6864 6865Change the m68k bit numbering for bitmasks to match the bit numbering 6866used by all other ext2 implementations. (This change was requested by 6867the m68k kernel development team.) 6868 6869Support (in-development) filesystem format revision which supports 6870(among other things) dynamically sized inodes. 6871 6872Fixed a bug in the ext2 library so that an intelligent error is 6873returned if mke2fs is run with a ridiculously small number of blocks 6874for a partition. 6875 6876Fixed a bug in the ext2 library which required that the device be 6877openable in read/write mode in order to determine its size. This 6878caused e2fsck -n to require read/write access when it was not 6879previously necessary. 6880 6881Fixed a bug in e2fsck which casued it to occasionally fail the test 6882suite depending on which version of the floating point library it was 6883using. 6884 6885Fixed a bug in e2fsck so that it now halts with a fatal error when 6886certain superblock consistency checks fail. Previously it continued 6887running e2fsck, with some potential confusing/damaging consequences. 6888 6889Added new flag to fsck which allows the root to be checked in parallel 6890with other filesytems. This is not the safest thing in the world to 6891do, but some system administrators really wanted it. 6892 6893Fixed -Wall flames in lib/ss. 6894 6895 6896E2fsprogs 1.02 (January 16, 1996) 6897================================= 6898 6899Fix to allow e2fsprogs to be compiled on recent 1.3 (pl45+) kernels. 6900 6901Change e2fsck to print statistics of how many non-contiguous files are 6902on the system. Note that a file which is larger than 8k blocks, it is 6903guaranteed to be non-contiguous. 6904 6905In mke2fs, print a warning message if a user tries to format a whole 6906disk (/dev/hda versus /dev/hda1). If a user really wants to format a 6907whole disk, the -F (force) option forces mke2fs to format a whole disk 6908as a filesytem. 6909 6910Fix a bug in fsck where in some cases it might start checking 6911partitions in the next pass before it finishes checking partitions in 6912the current pass. This still won't cause two partitions on the same 6913disk will be checked, so it's rarely a problem in real life. 6914 6915Patch lsattr so that it won't hang when checking a named pipe. 6916 6917Minor compilation fixes: 6918 * Fix the order of libraries that were linked in debugfs. 6919 * Allow the sources to be compiled with -ansi turned on. 6920