RELEASE-NOTES revision 511ba985c0821d07a6274515e35cbf5a3c81077d
1E2fsprogs 1.42.7 (January 21, 2013) 2=================================== 3 4Add warnings to mke2fs, resize2fs, and tune2fs that the bigalloc and 5quota features are still under development. For more information 6please see: 7 * https://ext4.wiki.kernel.org/index.php/Bigalloc 8 * https://ext4.wiki.kernel.org/index.php/Quota 9 10Add some new options to filefrag from Lustre's patches to e2fsprogs: 11 * add -k option to print extents in kB-sized units (like df -k) 12 * add -b {blocksize} to print extents in blocksize units 13 * add -e option to print extent format, even when FIBMAP is used 14 * add -X option to print extents in hexadecimal format 15 16Fix resize2fs so that it can handle off-line resizes of file systems 17with the flex_bg feature but without a resize_inode (or if we run out 18of reserved gdt blocks). This also fixes a problem where if the user 19creates a filesystem with a restricted number of reserved gdt blocks, 20an off-line resize which grows the file system could potentially 21result in file system corruption. 22 23Fix a resize2fs bug which could cause it to corrupt bigalloc file 24systems when doing an off-line resize. 25 26Further optimize resize2fs so it doesn't use quite as much CPU when 27resizing very large file systems. 28 29Fixed 32-bit overflow bugs which could cause resize2fs to fail and 30possibly corrupt the file system while resizing 64-bit file systems. 31 32Fix a bug in mke2fs where parsing "-E resize=NNN" will result in a 33mke2fs crash due to a divide-by-zero if the 64bit file system feature 34is enabled. 35 36Add better error checking to mke2fs to check for invalid parameters 37when creating bigalloc file system. 38 39When creating bigalloc filesystems, the -g option to mke2fs will now 40specify the number of clusters per block group. 41 42Add to debugfs the functionality to corrupt a specific file system 43block via the "zap_block" command. 44 45Add to debugfs the functionality to print out a hex dump of a block in 46the file system via the "block_dump" command. 47 48Add to debugfs the functionality to manipulate the extent tree 49directly via the "extent_open" command. 50 51Fixed debugfs's mknod command so that it updates the block group 52statistics. 53 54Fix e2fsck so it can detect and fix inconsistencies in the interior 55nodes of an inode's extent tree. 56 57Fix a potential memory corruption failure in e2fsck's error path if 58the call to ext2fs_open2() fails. 59 60Fix e2fsck if its logging function is enabled in e2fsck.conf, and the 61resulting file name for the log file is longer than 100 bytes, that it 62properly handles this situation instead of crashing. 63 64E2fsck will now report the amount of memory that it attempted to 65allocate when a memory allocation request fails, to make it easier to 66track down the problem. 67 68Fix mke2fs's handling of the mmp_update_interval option. (Addresses 69Lustre Bug: LU-1888) 70 71E2image can now include all data blocks in the e2image output file 72when the user specifies the -a option. 73 74If e2fsprogs is compiled without --enable-quota, make sure that 75tune2fs can not turn on the feature for new-style quota support, since 76afterwards, none of the e2fsprogs tools will be willing to touch that 77file system. 78 79Optimize e2fsck's so that it uses much less CPU for large file 80systems. This can result in significant speedups, especially on 81CPU-constrained systems. This was primarily done by optimizing 82libext2fs's bitmap functions. (Addresses Google Bug: #7534813) 83 84Fix debugfs's htree command so that all its messages are sent through 85the pager. 86 87Fixed debugfs's "dump_file" and "cat" functions work correctly 88on file systems whose block size is greater than 8k. 89 90Fix e2freefrag so it works on 64-bit file systems, and so it uses much 91less memory. (Addresses Google Bug: 7269948) 92 93Update the spd_readdir.c file in the contrib directory to include some 94additions which were made in 2008 that didn't get folded into the 95version which we checked into the e2fsprogs source tree. These 96enhancements include thread safety, support for readdir64_r(), and 97safe_getenv() support in case spd_readdir.so is used for setuid binaries. 98 99Update Czech, Dutch, French, German, Polish, Sweedish, and Vietnamese 100translations 101 102Add a command to debugfs to create symlinks. 103 104Document the bigalloc feature in the mke2fs man page. (Addresses 105Debian Bug: #669730) 106 107 108Programmer's Notes 109------------------ 110 111Fix gcc -Wall nits. 112 113Fix a spelling typo in the libext2fs texinfo documentation. 114 115Change the output from "make check" so that tools such as emacs's "M-x 116compile" does not mistake the output as containing a compiler error. 117 118Export two new functions from libext2fs: ext2fs_extent_node_split() 119and ext2fs_extent_goto2(). 120 121The ext2fs_extents_fix_parents() was fixed so it does not modify the 122location of the extent handle if its current location is at an 123interior node, rather than a leaf node of an inode's extent tree. 124 125Add a regression test to assure that e2fsck can correctly fix an 126inconsistent interior node in an inode's extent tree. 127 128The ext2fs_{mark,unmark,test}_block_bitmap_range2() functions now 129correctly support bigalloc file systems which use store block usage 130information in units of clusters. 131 132Fixed the help text in the configure script for --enable-quota. 133 134The m68k-specific bitops code has been removed since they were 135incorrectly treating bit numbers with the high bit set as signed 136integers. Furthermore, modern compilers do a good enough job 137optimizing the generic code there is no point in having the m68k 138specific asm statements. 139 140Fixed how we link the test programs so they always use the static 141libraries, so that we test using the libraries which we have just 142built. 143 144Update config.guess and config.sub to the latest versions from the GNU 145project. 146 147Fixed the com_err.texinfo file so that it can produce a valid 148postscript/pdf printed output. 149 150Add a regression test which checks debugfs's ability to create 151symlinks, named FIFO's, and device nodes. 152 153Add a function ext2fs_symlink(), which creates symlinks to the 154libext2fs library. 155 156Add debugging code so we can test old kernel interfaces for online 157resize to resize2fs. This backwards compatibility checking is keyed 158off of the RESIZE2FS_KERNEL_VERSION, which is designed to allow us to 159test the functionality of the kernel's older resize ioctls without 160needing to install an old version of resize2fs, and to also test a 161modern resize2fs's ability to work with older kernels without having 162to install an older version of the kernel. 163 164 165E2fsprogs 1.42.6 (September 21, 2012) 166===================================== 167 168When mke2fs creates file systems with lazy itable initialization, the 169progress updates for writing the inode table happens so quickly that 170on a serial console, the time to write the progress updates can be the 171bottleneck. So mke2fs will now only update its progress indicators 172once a second. 173 174Resize2fs will skip initializing the inode tables if the kernel 175supports lazy_itable_init, which speeds up growing off-line growth of 176uninit_bg file systems significantly. Resize2fs will now also 177correctly set the itable_unused field in the block group descriptor to 178speed up the first e2fsck after the file system is grown. 179 180Resize2fs has been fixed so that on-line resizing of meta_bg file 181systems work correctly. This is needed to grow file system which are 182larger than 16T. 183 184Resize2fs will now correctly handle resizing file systems to 16TB on 18532-bit file systems when "16TB" is specified on the command line. 186 187Fix mke2fs so that it will be careful to set the reserved blocks ratio 188larger than 50%; this can happen when creating small file systems and 189when the last block group is dropped because there are not enough 190blocks to support the metadata blocks in the last block group. 191 192Fixed spelling mistake in debugfs's help message. 193 194Fixed a potential seg fault in e2fsck when there is an I/O error while 195reading the superblock. 196 197Fixed various Debian Packaging Issues (Addresses Debian Bug #677497) 198 199Updated/fixed various man pages. (Addresses Sourceforge Bug: 200#3559210) 201 202Programmer's Notes 203------------------ 204 205The configure option --enable-relative-symlink was broken so that it 206needed to be --enable-symlink-relative-symlinks. We will support both 207for at least two years, but then the worng configure option will be 208removed. 209 210Fixed a regression introduced in 1.42.5 so the link order for 211e2fsprogs' libraries will be correct for both static and shared 212linking. (Addresses Sourceforge Bug: #3554345) 213 214Add support for e2fsprogs to be compiled using clang/LLVM. 215 216Fix portability problems on non-Linux systems: avoid compile failures 217on systems that don't have malloc.h 218 219Fix f_mmp regression test suite so that debugfs gets killed if the 220test is interrupted. 221 222 223 224E2fsprogs 1.42.5 (July 29, 2012) 225================================ 226 227Fixed a bug with mke2fs where if there is only 8 inodes per block 228group, the calculation of the number of uninitialized inodes in the 229first block group would go negative. This resulted in "mke2fs -N 256 230-t ext4 /tmp/foo.img 256m" trying to write so many blocks that /tmp 231would run out of space. (Addresses Sourceforge Bug: #3528892) 232 233Fixed a bug in how e2fsck would uniquify directory entry names. 234(AddressesSourceforge Bug: #3540545) 235 236Previously, e2fsck would only allow a mounted file system to be 237checked if it was the root file system and it was mounted read-only. 238Now it will allow any file system mounted read-only to be checked if 239the -f option is specified. This makes it easier to test how e2fsck 240handles checking file systems which are mounted without having to test 241on the root file system. 242 243Fixed a problem if e2fsck where if the root file system is mounted 244read-only, e2fsck would not clear an error indication in the journal 245superblock. Combined with a kernel bug, this would cause the e2fsck 246to check the file system after every single boot. 247 248The e4defrag program can now handle device symlinks, such as 249/dev/mapper/testvg-testlv, instead of insisting on a less 250human-friendly name such as /dev/dm-2. (Addresses Red Hat Bugzilla: 251#707209) 252 253Fixed filefrag so it will not crash with a segfault on files from a 254virtual file system such as /proc. (e.g., "filefrag 255/proc/partitions") 256 257Fixed filefrag so that it correctly reports the number of extents. 258(Addresses Red Hat Bugzilla: #840848) 259 260Fixed a file descriptor leak in logsave which could cause it to hang. 261(Addresses Debian Bug: #682592) 262 263Fixed e2fsck so that the file system is marked as containing an error 264if the user chooses not to fix the quota usage information. 265 266Fixed tune2fs so that it correctly removes the quota feature when the 267last quota inode is removed. 268 269Fix tune2fs so that after removing a quota inode, the block bitmap is 270updated; otherwise, e2fsck would complain after running 'tune2fs -O 271^quota <dev>'. 272 273Fix tune2fs so that when converting a file system from using legacy 274quota files to the new quota file system feature with hidden quota 275files, the accounting for these files is handled correctly so that 276e2fsck doesn't complain. 277 278Improved e2fsck's verbose reporting statistics, and allow the more 279verbose reporting to be enabled via /etc/e2fsck.conf. 280 281Fixed various Debian Packaging Issues (Addresses Debian Bug #678395) 282 283Updated/fixed various man pages. (Addresses Debian Bugs: #680114) 284 285 286Programmer's Notes 287------------------ 288 289Fixed portability problems on other operating systems (e.g., Hurd and 290FreeBsd) caused by the attempted inclusion of <sys/quota.h>. 291 292Make sure that shared libraries link with the shared libraries built 293in the build tree, instead of the system provided libraries. 294Previously, libraries and executables were linked with the system 295libraries if present, and possibly using static archives instead of 296shared libraries. This was also problematic since if libext2fs.so is 297linked with a static libcom_err.a from system, the build system would 298attempt to link without -lpthread. (Addresses Sourceforge Bug: 299#3542572) 300 301 302E2fsprogs 1.42.4 (June 12, 2012) 303================================ 304 305Fixed more 64-bit block number bugs (which could end up corrupting 306file systems!) in e2fsck, debugfs, and libext2fs. 307 308Fixed e2fsck's handling of the journal's s_errno field. E2fsck was 309not properly propagating the journal's s_errno field to the superblock 310field; it was not checking this field if the journal had already been 311replayed, and if the journal *was* being replayed, the "error bit" 312wasn't getting flushed out to disk. 313 314Fixed a false positive complaint by e2fsck if all of the extents in 315the last extent tree block are uninitalized and located after the end 316of the file as defined by i_size. 317 318The dumpe2fs will now display the journal's s_errno field if it is 319non-zero, and it will also display the journal's 64-bit feature flag 320if present. 321 322Fix e2fsck so that it always opens the device file in exclusive mode 323when it might need to modify the file system, and never if the -n 324option is specified. (Previously, there were a few corner cases where 325it might get things wrong in either direction.) 326 327E2fsck now correctly truncates or deallocates extent-mapped inodes on 328the orphan list. The root cause was a bug in libext2fs's block 329iterator which could end up skipping an extent when the last block in 330an extent is removed, causing the current extent to be removed from 331the extent tree. 332 333E2fsck now correctly sets the global free block and inode counts when 334truncating or removing inodes on the orphan list in preen mode. 335Previously, it would leave these values would be set incorrectly, 336which is largely a cosmetic issue since the kernel no longer pays 337attention to those fields, but it can cause spurious complaints in 338subsequent e2fsck runs. 339 340Fix i_blocks accounting when the libext2fs library needs to add or 341remove an extent tree block on bigalloc file systems. 342 343The lsattr and chattr programs now support the No_COW flag for the 344benefit of btrfs. 345 346Debugfs now interprets date strings of the form @ddd as ddd seconds 347after the beginning of the epoch. This is handy when setting an inode 348number into the d_time field when debugging orphan list handling. 349 350Fix a precedence bug with built-in quota support which might result in 351e2fsck paying attention to the quota inode field even if the built-in 352quota feature flag is not set. Fortunately, in practice that 353superblock field should be zero for non-built-in quota file systems, 354so it's unlikely this bug would have caused problems. 355 356Updated/fixed various man pages. (Addresses Debian Bugs: #674453, 357#674694) 358 359Programmer's Notes 360------------------ 361 362The regression test suite can now run the integration tests in the 363tests directory in parallel, via "make -jN check". 364 365Add new test, f_zero_extent_length which tests e2fsck's handling of 366the case where all of the extents in the last extent tree block are 367uninitialized extents after i_size. 368 369Add a new test, f_jnl_errno, which checks handling of an error 370indication set in the journal superblock. 371 372Fix the test f_jnl_64bit so that it properly checks e2fsck's handling 373of a 64-bit journal. 374 375Add two tests, f_orphan_indirect_inode and f_orphan_extent_inode which 376tests e2fsck's handling of orphan inodes in preen mode, and truncation 377of extent inodes on the orphan list. 378 379Fixed more OS X portability issues. 380 381 382E2fsprogs 1.42.3 (May 14, 2012) 383=============================== 384 385Fix a bug in the Unix I/O manager which could cause corruption of file 386systems with more than 16TB when e2fsprogs is compiled in 32-bit mode 387(i.e., when unsigned long is 32-bits). Also fix a bug which caused 388dumpe2fs to incorrectly display block numbers > 32-bits. 389 390Improve the support for integrated quota files (where quota is a first 391class supported feature using hidden files in the ext4 file system). 392Previously the quota file was getting rewritten even when it was not 393necessary, and e2fsck would erroneously try to hide quota files which 394were already hidden. 395 396Quiet complaints in e2fsck when the total free blocks or inodes are 397incorrect in the superblock after an system crash, since we don't 398update nor depend on the superblock summaries at each commit boundary. 399 400Fixed a regression introduced in 1.42.2 which would cause applications 401outside of e2fsprogs which did not pass the EXT2_FLAG_64BIT (and so 402would were still using the legacy 32-bit bitmaps) to crash. This was 403due to missing 32-bit compat code in side the function 404ext2fs_find_first_zero_generic_bmap(). (Addresses Red Hat Bugzilla: 405#808421) 406 407Fix a bug which would cause mke2fs to fail creating the journal if 408/etc/mtab and /proc/mounts are missing. (Addresses Sourceforge Bug: 409#3509398) 410 411Updated/fixed various man pages. 412 413Update Czech, Dutch, French, German, Polish, Sweedish, and Vietnamese 414translations 415 416Fixed various Debian Packaging issues. 417 418 419Programmer's Notes 420------------------ 421 422Change the nonsensically wrong types in the function signature of the 423inline function ext2fs_find_first_zero_block_bitmap2(). This was 424caused by a cut and paste error; fortunately no code in e2fsprogs used 425this inline function, and there are any users of this functions 426outside of e2fsprogs. 427 428Add support for systems which have valloc(), but which do have 429posix_memalign() nor memalign() (such as MacOS 10.5). 430 431Refactor and clean up the allocation of aligned buffers for Direct I/O 432support. Previously some allocations were requesting a greater 433alignment factor that what was strictly necessary. Also optimize 434reading and writing bitmaps using Direct I/O when the size of the 435bitmap did not fully cover the file system blocksize. 436 437Reserve the codepoints for the INCOMPAT features LARGEDATA and 438INLINEDATA. 439 440Improved the regression test suite by adding some new integration 441tests (f_jnl_32bit, f_jnl_64bit) which detect breakage of the on-disk 442jbd2 format, as well as f_eofblocks which tests the new handling of 443uninitialized and initialized blocks beyond i_size. Also add a new 444unit test which verifies 32-bit bitmaps support and the new 445find_first_zero primitives. 446 447Add a few dependencies to fix parallel (make -j) builds. 448 449Removed bash'isms which were breaking the regression test suite on 450systems where /bin/sh is not bash. 451 452The config.guess and config.sub have been updated to the 2012-02-10 453version. 454 455Fix a portability problem caused by assuming the present of mntent.h 456means that setmntent() exists. Instead, explicitly test for this in 457the configure script. 458 459If the sys/signal.h header file does not exist, don't try to include 460it, since it's not available on all systems. 461 462Add support for systems that do not support getpwuid_r() 463 464The configure script now supports a new option, 465--enable-relative-symlinks, which will install relative symlinks for 466the ELF shared library files. (Addresses Sourceforge Bug: #3520767). 467 468When building BSD shared libraries make sure the LDFLAGS variable is 469passed to the linker. Fixing this allows, for example, e2fsprogs to 470be built in 32-bit mode on Mac OS X Lion (Addresses Sourceforge Bug: 471#3517272) 472 473Fix gcc -Wall nitpicks 474 475 476E2fsprogs 1.42.2 (March 27, 2012) 477================================= 478 479The resize2fs program uses much less CPU and is much faster for very 480large file systems. (Addresses Debian Bug: #663237) 481 482The seti and freei commands in debugfs can now take an optional length 483argument to set and clear a contiguous range of inodes. 484 485E2fsck will now make explicit checks for the EOFBLOCKS_FL, since we 486plan to remove support for it from the kernel file system driver. It 487really wasn't very useful and was causing more problems than it 488solves. Since e2fsck will complain if inodes that should have 489EOFBLOCKS_FL do not have the flag set, we are going to remove this 490check from e2fsprogs first, and then only remove the flag from the 491kernel much later. 492 493The mke2fs program can now use direct I/O via "mke2fs -D". This will 494slow down the mke2fs, but it makes it more polite on a loaded server 495by limiting the amount of memory that gets dirtied by mke2fs when it 496is using buffered I/O. 497 498E2fsck was needlessly closing and re-opening the file system as a side 499effect of adding Multiple Mount Protection (MMP). This isn't 500necessary for non-MMP file systems, so drop it. 501 502Print errors returned by ext2fs_open2() and ext2fs_check_desc() so we 503can more easily diagnose memory allocation failures caused by 504insufficient memory. E2fsck will now abort if there are memory 505allocation failures when the file system is initially opened and 506during the block group descriptor checks. (Addresses Google Bug: 507#6208183) 508 509If there are incorrect block group checks, e2fsck will now report the 510incorrect and corrected checksum values. 511 512The e2fsck progam can now write log files containing the details of 513the problems that were found and fixed directly, via configuration 514parameters in /etc/e2fsck.conf. 515 516Added the ability to limit the number of messages reported by e2fsck 517for a given problem type. This avoids a potential bottleneck if there 518is a serial console which can cause a boot sequence to take a long 519time if e2fsck needs to report many, many file system errors. 520 521The dumpe2fs, debugfs, and tune2fs now use rbtree bitmaps, which cause 522them to use much less memory for large file systems. 523 524The dumpe2fs program will now print the expected block group checksum 525if it is incorrect. This helps to diagnose problems caused by 526incorrect block group checksums. 527 528E2fsck now checks for extents with a zero length, since the kernel 529will oops if it comes accross such a corrupted data structure. (See 530https://bugzilla.kernel.org/show_bug.cgi?id=42859) 531 532E2fsck has a number of bugs relating to discard that have been fixed. 533(1) Fixed a bug which could cause e2fsck to discard portions of the 534inode table which were actually in use. (2) E2fsck will now avoid 535using discard if the block device doesn't zero data on discard, since 536otherwise this could cause problems if the file system gets corrupted 537in the future. (3) E2fsck will now avoid using discard when it is run 538in read-only mode. (4) Fixed a bug which caused e2fsck to not issue 539discards in the last block group. 540 541E2fsck's CPU utilization in pass 5 has been optimized, which will 542speed up e2fsck slightly. 543 544E2image will now skip copying uninitialized bitmap and inode table 545blocks. 546 547Fixed mke2fs -S so it does not corrupt the first block group's 548information. 549 550E2fsck will now check the new sysfs interface to determine if we are 551using the battery or AC mains. (Addresses SourceForge Bug: #3439277) 552 553Updated/fixed various man pages. (Addresses Debian Bug: #665427) 554 555Fixed various Debian Packaging issues. (Addresses Debian Bug: #665885) 556 557Programmer's Notes 558------------------ 559 560Fixed various portability issues for non-Linux systems, particularly 561MacOS X, as well as Linux systems running with the just-released glibc 5622.15. 563 564Fix file descriptor leak in ext2fs_close() if the file system with 565uninit_bg is opened read/only with a backup superblock. (Addresses 566SourceForge Bug: #3444351) 567 568Fixed an invalid return in a non-void function in the quota code. 569(Addresses SourceForge Bug: #3468423) 570 571Fixed the debian rules file so that the calls to dpkg-buildflags works 572when the shell is dash. 573 574The debian package build now uses V=1 so that there is more 575information about potential build failures in debian buildd logs. 576 577If the uninit flags get cleared by functions such as 578ext2fs_new_inode() or ext2fs_new_block2(), we now make sure the 579superblock is marked dirty and the block group descriptor checksum is 580updated if necessary. 581 582The debian rules file will now try to load debian/rules.custom of it 583exists. This flie can skip various builds for speed reasons if there 584is no need for the e2fsck-static or udeb packages. Available 585customizations in the rules file includes SKIP_STATIC=yes, 586SKIP_BF=yes, and SKIP_DIETLIBC=yes. 587 588In addition, if the file misc/mke2fs.conf.custom.in exists in the 589source tree, it will be used instead of the standard misc.conf file in 590the upstream sources. This makes it easier for Debian-derived systems 591to distribute a custom mke2fs.conf file without having to worry about 592merge issues if future versions of e2fsprogs makes changes in the 593upstream default version of mke2fs.conf. 594 595 596E2fsprogs 1.42.1 (February 17, 2012) 597=================================== 598 599The mke2fs and e2fsck now use significantly less memory when creating 600or checking very large file systems. This was enabled by adding 601extent-based bitmaps which are stored using a red-block tree, since 602block and inode allocations tend to be contiguous. 603 604The command mke2fs -S is used as a last ditch recovery command to 605write new superblock and block group descriptors, but _not_ to destroy 606the inode table in hopes of recovering from a badly corrupted file 607system. So if the uninit_bg feature is enabled, mke2fs -S will now 608set the unused inodes count field to zero. Otherwise, e2fsck -fy 609after using mke2fs -S would leave the file system completely empty. 610 611Since mke2fs recognizes mke3fs in argv[0] to mean "mkfs.ext3", also 612honor "mke4fs" to work the same as "mke2fs.ext4", since RHEL5 has 613installed an mke2fs binary using that name. 614 615The usage and help messages for the -G, -t and -T options in mke2fs 616have been fixed. 617 618If e2fsck needs to use the backup group descriptors, the 619ext2fs_open2() function clears the UNINIT bits to ensure all of the 620inodes in the file systems get scanned. However, the code which reset 621the UNINIT flags did not also recalculate the checksum, which produced 622many spurious (and scary) e2fsck messages. This has been fixed by 623resetting cheksums when the UNINIT bits are cleared. 624 625Relax a check in e2fsck which required that the block bitmap to be 626initialized when the inode bitmap is in use. This will allow us to 627eventually eliminate code from the kernel which forcibly initialized 628the block bitmap when the inode bitmap is first used, requiring an 629extra journal credit and disk write. (Addresses Google Bug: #5944440) 630 631Make sure rdebugfs (which may be installed setuid or setgid disk) does 632not honor environment variables if euid != uid or egid != gid. 633 634Debugfs's ncheck command has been optimized and now is much more 635robust with faced with corrupted file systems. The ncheck command 636also now has a -c option which will verify the file type information 637in the directory entry to see if matches the inode's mode bits. This 638is extremely useful when trying to use debugsfs to determine which 639parts of the file system metadata can be trusted. 640 641E2image will try to use ftruncate64() to set the i_size for raw 642images, instead of writing a single null byte. This avoid allocating 643an extra block to the raw image, for those file systems and/or 644operating systems that support this. (Linux does.) In addition, fix 645a logic bug that caused the file to not be properly extended if the 646size of the last hole was exactly an multiple of a megabyte. 647 648Fixed a bug in resize2fs where for 1k and 2k file systems, where 649s_first_data_block is non-zero, this wasn't taken into account when 650calculate the minimum file system size for use with the -M option. 651 652Fixed the badblocks program to honor the -s flag when in read-only -t 653mode. (Addresses Debian Bug #646629) 654 655Update Czech, Dutch, French, Polish, and Sweedish translation from the 656Translation Project. 657 658Fixed various Debian Packaging issues so that dpkg-buildflags is used 659if present, which allows e2fsprogs to be built with security hardening 660flags. (Addresses Debian Bugs: #654457) 661 662Programmer's Notes 663------------------ 664 665Fix a bug in ext2fs_clear_generic_bmap() when used for 32-bit bitmaps. 666This was only an issue for programs compiled against e2fsprogs 1.41 667that manipulate bitmaps directly. (Addresses Sourceforge Bugs: 668#3451486) 669 670The libext2fs library now uses sysconf() to fetch the page size, instead 671of the deprecated getpagesize(). 672 673The ext2fs_get_pathname() function will return a partial path if an a 674directory in the path is not a directory, displaying it as an inode 675number in angle brackets instead of giving up and displaying an error. 676This is much more helpful when a user is trying to debug a corrupted 677file system. 678 679Codepoints for the RO_COMPAT_REPLICA feature has been reserved. 680 681Added a new library function, ext2fs_file_get_inode_num(), for use by 682fuse2fs. 683 684Fixed a bug in ext2fs_file_set_size2() so that when it is truncating a 685file, it actually works. 686 687The block iterator now properly honors the BLOCK_ABORT flag for 688extent-based flags. Previously, it didn't, which generally made code 689be less efficient, but it could cause bugs in ext2fs_link(), for 690example, by causing it to insert multiple directory entries. 691 692Fixed an (harmless other than causing a compiler warning) use of an 693uninitialized variable in e2fsck's MMP code. 694 695 696E2fsprogs 1.42 (November 29, 2011) 697================================== 698 699This release of e2fsprogs has support for file systems > 16TB. Online 700resize requires kernel support which will hopefully be in Linux 701version 3.2. Offline support is not yet available for > 16TB file 702systems, but will be coming. 703 704This release of e2fsprogs has support for clustered allocation. This 705reduces the number of block (now cluster) bitmaps by allocating and 706deallocating space in contiguous power-of-2 collections of blocks, 707which are called clustered. This is a file system level feature, 708called 'bigalloc', which must be enabled when the file system is 709initially formatted. It is not backwards compatible with older 710kernels. 711 712Added support for the Multi-Mount Protection (MMP) feature. 713 714E2fsck more efficiently uses scratch files for really big file 715systems. (This is a feature that has to be turned on explicitly; see 716[scratch_files] in the e2fsck.conf man page.) 717 718Fix a bug in e2fsck where if the free blocks and inodes counts are 719incorrect, e2fsck would fix them without printing an error message. 720This would cause a "*** FILE SYSTEM WAS MODIFIED ***" message without 721any explanation of what was fixed. 722 723E2fsck will no longer attempt to clone an extended attribute block in 724pass1b handling if the file system does not support extended 725attributes. 726 727E2fsck will be more careful accidentally asking the user to continue 728if the file system is mounted, so that an escape sequence won't cause 729a false positive. (Addresses Debian Bug: #619859) 730 731E2fsck now uses less cpu time in pass 5 when large portions of the 732bitmaps are uninitialized. 733 734E2fsck will no longer segault when a corrupted file system has a bad 735extent, and removing it leads to a block needing to be deallocated. 736(Addresses SourceForge Bug: #2971800) 737 738E2fsck will catch termination signals (segfaults, bus errors, sigfpe) 739and print debugging information to make it easier to find potential 740problems. 741 742E2fsck will check to see if the bad block inode looks insane, and will 743skip trying to use if it certain fields which should be zero look 744non-zero. This is to avoid a corrupted bad block inode causing e2fsck 745to get confused and do more harm than good. 746 747If e2fsck modifies certain superblock fields which the kernel doesn't 748look at, it will now mark the superblock as dirty without marking the 749file system as changed. This avoids signalling the init scripts that 750a reboot is necessary, since the kernel isn't going to look at those 751fields, so it won't care if they have been changed. 752 753Fixed a bug in the libext2fs library (in the binary search routine of 754the icount abstraction) that could (very, very rarely) cause e2fsck to 755die in the middle of pass 1 or pass 2 processing. 756 757E2fsck will not try to do a discard operation if the -n option was 758specified on the command line. 759 760E2fsck now supports an extended "discard" option which will cause 761e2fsck to attempt discard all unused blocks after a full, successful 762file system check. 763 764E2fsck will check for the bad block inode to make sure it looks sane 765before trusting it, to avoid causing more harm than good to the file 766system. 767 768E2fsck now returns additional status bits in its exit code if it 769aborts early in the e2fsck run. 770 771E2fsck now correctly calculates the maximum file size in the case of 772the huge_file file system feature enabled without extents. 773 774The mke2fs and e2fsck programs now tries to use the punch hole command 775as a "discard" when operating on normal files. 776 777The e2image program now supports the qcow2 format, which is a more 778efficient way of capturing file system dumps. 779 780Mke2fs now supports the [devices] stanza in mke2fs.conf which allows 781per-device defaults to be specified in the configuration file. 782 783Mke2fs now supports the reserved_ratio relation in the [defaults] and 784[fs_types] section in mke2fs.conf. 785 786Mke2fs now creates extent-mapped directories for the root and 787lost+found directories. 788 789Mke2fs will skip zero'ing the journal if the extended option 790"lazy_journal_init" is specified. This can save a lot of time, but it 791does add a small amount of risk if the system crashes before the 792journal is overwritten entirely once. It is epsecially useful for 793testing. 794 795Mke2fs will now create file systems that enable user namespace 796extended attributes and with time- and mount count-based file 797system checks disabled. 798 799Mke2fs will not set a stride or strip size of one block based on block 800device attributes obtained from sysfs. 801 802Mke2fs now displays a progress report during the discard process. 803 804Mke2fs now handles extreme file system parameters correctly which 805previously caused the inodes per group to drop below 8, leading to a 806segfault. (The inodes per group must be a multiple of 8, but the code 807didn't correctly deal with an inodes per group count less than 8.) 808 809Mke2fs and tune2fs previously would give an error if the user tried 810setting the stride and stripe-width parameters to zero; but this is 811necessary to disable the stride and stripe-width settings. So allow 812setting these superblock fields to zero. (Addresses Google Bug: 813#4988557) 814 815Mke2fs now gives a warning if the auto-detected block size exceeds the 816system's page size. 817 818If the enable_periodic_fsck option is false in /etc/mke2fs.conf (which 819is the default), mke2fs will now set the s_max_mnt_count superblock 820field to -1, instead of 0. Kernels older then 3.0 will print a 821spurious message on each mount then they see a s_max_mnt_count set to 8220, which will annoy users. (Addresses Debian Bug: #632637) 823 824The default mke2fs.conf now has entries for "big" and "huge", which 825are needed for very big file systems. 826 827The resize2fs program now has support for a new online resize ioctl 828that can support file systems > 16TB, once it arrives in v3.x kernels. 829 830Fixed bug which caused resize2fs to fail when shrinking an empty file 831system down to its minimal size. (Addresses Sourceforge Bug #3404051) 832 833Fixed tune2fs's mount options parsing. (Addresses Debian Bug: #641667) 834 835Allow tune2fs to remove external journals if the device can not be 836found. 837 838Debugfs's icheck will now correctly find inodes which use the 839searched-for block as an extended attribute block. 840 841Debugfs now has a new "punch" command which remove blocks from the 842middle of an inode. 843 844Debugfs now has a new "e2freefrag" command which analyzes the free 845space fragmentation of the file system, using the same code as the 846e2freefrag program. 847 848Debugfs now has a "filefrag" command which displays information about 849a file's fragmentation. 850 851Add support to build a metadata-only, read-only, stripped-down version 852of debugfs called rdebugfs. 853 854Fixed a potential stack overrun bug in debugfs. 855 856The badblocks program now correctly recovers from I/O errors when 857direct I/O is being used. The badblocks command now also supports a 858-B option which forces the use of buffered I/O, and the -v option will 859provide a more detailed breakdown of read, write, and failed 860comparison errors. 861 862Added e4defrag tool which uses the EXT4_IOC_MOVE_EXT ioctl. 863 864Added support for journals larger than 2GB. 865 866Support using both hard links and symlinks when installing e2fsprogs. 867 868Add overflow checking to tune2fs -i's fsck interval, which must fit in 869a 32-bit field. 870 871The debugfs command now has a new 'blocks' command which prints out 872data blocks of a particular inode in a format which is useful for 873scripting. 874 875Filefrag will report 0 extents correctly in verbose mode. (Addresses 876RedHat Bugzilla: #653234) 877 878Filefrag has been fixed so its -v report prints the correct expected 879block number (previously there had been an off-by-one error). In 880addition, it will now display the number of contiguous extents when -v 881is not specified. This makes it consistent with the number of extents 882printed when the -v option was specified. In addition, the number of 883contiguous extents is far more interesting/useful than the number of 884physical extents for very large files. (Addresses Debian Bug: 885#631498, #644792) 886 887Logsave's usage message has been fixed. (Addresses Debian Bug: 888#619788) 889 890Avoid an infinite loop in ext2fs_find_block_device() if there are 891symlink loops in /dev caused by a buggy udev. 892 893Added a useful "fallocate" program to the contrib directory. 894 895Fixed an ABI compatibility problem in libext2fs which broke the dump 896program. Also added back some macros which dump needed so it could 897compile against the latest version of ext2_fs.h (Addresses Debian Bug: 898#636418) 899 900Fixed parsing of MNTOPT_ options for tune2fs and debugfs (Addresses 901Debian Bug: #641667) 902 903Added internationalization support for libcom_err error table strings. 904 905Fixed various spelling mistakes found in various output strings found 906by I18N translators. 907 908Update translations: French, Chinese, Germany, Indonesian, Swedish, 909Vietnamese, Polish, Dutch, Czech. (Addresses Debian Bugs: #520985, 910#620659) 911 912Fixed various Debian Packaging issues. (Addresses Debian Bugs: 913#614662, #632169, #641838, #627535, #629355) 914 915Updated/clarified man pages. (Addresses Debian Bugs: #639411, 916#642193, #634883) 917 918Programmer's Notes 919------------------ 920 921Initial support for quota as an integrated feature, where the quota 922files are hidden system files that are automatically maintained by 923e2fsck is present, although disabled by default. It must be enabled 924by using the configure option --enable-quota. 925 926Reserved the on-disk fields for the metadata checksum and snapshot 927features, which are currently in development. 928 929The ext2fs library now has the new functions ext2fs_punch(), 930ext2fs_get_memzero() and ext2fs_file_get_inode(). 931 932The ext2fs library now has support for calculating the crc32c checksum 933(via the new functions ext2fs_crc32c_be and ext2fs_crc32c_le). 934 935The I/O manager now supports the discard operation. 936 937Reserved file system code points for new 1st class quota feature. 938 939Shortened the compilation lines by moving the autoconf defines to 940a config.h header file. 941 942Fixed a potential free of an unitialized pointer in 943ext2fs_update_bb_inode(). 944 945Fixed miscellaneous gcc -Wall and coverity warnings. 946 947Fixed portability issues for Mac OS X, Hurd, and FreeBSD. 948 949Fixed a build failure when OMIT_COM_ERR is defined. 950 951Improved error checking and fixed memory leaks caused by error return 952paths. 953 954Add ext2fs_flush2() and ext2fs_close2() which takes a flag to allow 955the fsync() to be skipped. 956 957Added a test for extent-mapped journals by mke2fs and tune2fs. 958 959Added a test for creating a large (over 4GB) journal using mke2fs. 960 961Added a test to make sure the inode size remains constant. 962 963The regression test script now prints the list of which tests failed. 964 965The regression test system now uses /tmp for its scratch files, which 966is often a tmpfs mounted file system and hence much faster. 967 968The i_e2image test was fixed so it works with valgrind. 969 970Tests that rely on debugfs are now skipped if debugfs wasn't built. 971 972Fixed the dependencies for "make check" so all required dependencies 973are built before running the gression tests. 974 975A link to com_err.h is installed in $(includedir) during a "make install". 976 977The po/*.gmo files are automatically rebuilt if they are missing or 978out of date. This helps out Debian packaging. 979 980Allow ext2fs_get_memalign() to compile on systems that don't have 981posix_memalign(). 982 983Fixed a namespace leak in libext2fs (tdb_null). 984 985 986E2fsprogs 1.41.14 (December 22, 2010) 987===================================== 988 989Fix spurious complaint in mke2fs where it would complain if the file 990system type "default" is not defined in mke2fs.conf. 991 992The resize2fs program will no longer clear the resize_inode feature 993when the number reserved group descriptor blocks reaches zero. This 994allows for subsequent shrinks of the file system to work cleanly for 995flex_bg file systems. 996 997The resize2fs program now handles devices which are exactly 16T; 998previously it would give an error saying that the file system was too 999big. 1000 1001E2fsck (and the libext2fs library) will not use the extended rec_len 1002encoding for file systems whose block size is less than 64k, for 1003consistency with the kernel. 1004 1005Programming notes 1006----------------- 1007 1008E2fsprogs 1.41.13 would not compile on big-endian systems. This has 1009been fixed. (Addresses Sourceforge Bug: #3138115) 1010 1011The ext2fs_block_iterator2() function passed an incorrect ref_offset 1012to its callback function in the case of sparse files. (Addresses 1013Sourceforge Bug: #3081087) 1014 1015Fix some type-punning warnings generated by newer versions of gcc. 1016 1017 1018E2fsprogs 1.41.13 (December 13, 2010) 1019===================================== 1020 1021E2fsck now supports the extended option "-E journal_only", which 1022causes it to only do a journal replay. This is useful for scripts 1023that want to first replay the journal and then check to see if it 1024contains errors. 1025 1026E2fsck will now support UUID= and LABEL= specifiers for the -j option 1027(which specifies where to find the external journal). (Addresses 1028Debian Bug #559315) 1029 1030E2fsck now has support for the problems/<problem code>/force_no 1031configuration option in e2fsck.conf, which forces a problem to not be 1032fixed. 1033 1034Dumpe2fs will now avoid printing large negative offsets for the bitmap 1035blocks and instead print a message which is a bit more helpful for 1036flex_bg file systems. 1037 1038Mke2fs will now check the file system type (specified with the -t 1039option) to make sure it is defined in the mke2fs.conf file; if it is 1040not, it will print an error and abort. If the usage type (specified 1041with the -T option) is not defined in mke2fs.conf, mke2fs will print a 1042warning but will continue. (Addresses Debian Bug #594609) 1043 1044Clarified error message from resize2fs clarifying that on-line 1045shrinking is not supported at all. (Addresses Debian Bug #599786) 1046 1047Fix an e2fsck bug that could cause a PROGRAMMING BUG error to be 1048displayed. (Addresses Debian Bug #555456) 1049 1050E2fsck will open the external journal in exclusive mode, to prevent 1051the journal from getting truncated while it is in use due to a user 1052accidentally trying to run e2fsck on a snapshotted file system volume. 1053(Addresses Debian Bug #587531) 1054 1055Fix a bug in e2fsck so it has the correct test for the EOFBLOCKS_FL 1056flag. 1057 1058The tune2fs program can now set the uninit_bg feature without 1059requiring an fsck. 1060 1061The tune2fs, dumpe2fs, and debugfs programs now support the new ext4 1062default mount options settings which were added in 2.6.35. 1063 1064The e2fsck and dumpe2fs programs now support the new ext4 superblock 1065fields which track where and when the first and most recent file 1066system errors occurred. These fields are displayed by dumpe2fs and 1067cleared by e2fsck. These new superblock fields were added in 2.6.36. 1068 1069Debugfs now uses a more concicse format for listing extents in its 1070stat command. This format also includes the interior extent tree 1071blocks, which previously was missing from stat's output for 1072extent-based files. 1073 1074Debugfs has a new option, -D, which will request Direct I/O access of 1075the file system. 1076 1077Mke2fs will skip initializing the inode table if a device supports 1078discard and the discard operation will result in zero'ed blocks. 1079 1080Badblocks will now correctly backspace over UTF-8 characters when 1081displaying its progress bar. (Addresses Gentoo Bug #309909; Addresses 1082Debian Bugs #583782 and #587834) 1083 1084E2freefrag will now display the total number of free extents. 1085 1086Resize2fs -P no longer requires a freshly checked filesystem before 1087printing the minimum resize size. 1088 1089Fixed a floating point precision error in a binary tree search routine 1090that can lead to seg fault in e2fsck and resize2fs. 1091 1092Fixed a bug in e2fsck where if both the original and backup superblock 1093are invalid in some way, e2fsck will fail going back to the original 1094superblock because it didn't close the backup superblock first, and 1095the exclusive open prevented the file system from being reopened. 1096 1097Fixed a big in e2freefrag which caused getopt parsing to fail on 1098architectures with unsigned chars. (Addresses Gentoo Bug: #299386) 1099 1100Clarified an mke2fs error message so a missed common in an -E option 1101(i.e., mke2fs -t ext4 -E stride=128 stripe-width=512 /dev/sda1") 1102results in a more understandable explanation to the user. 1103 1104Mke2fs now displays the correct valid inode ratio range when 1105complaining about an invalid inode ratio specified by the user. 1106 1107Mke2fs now understands the extended option "discard" and "nodiscard", 1108and the older option -K is deprecated. The default of whether 1109discards are enabled by default can be controled by the mke2fs.conf 1110file. 1111 1112Mke2fs's handling of logical and physical sector sizes has been 1113improved to reflect the fact that there will be some SSD's with 8k and 111416k physical sectors arriving soon. Mke2fs will no longer force block 1115size to be the physical sector size, since there will be devices where 1116the physical sector size is larger than the system's page size, and 1117hence larger than the maximal supported block size. In addition, if 1118the minimal and optimal io size are not exported by the device, and 1119the physical sector size is larger than the block size, the physical 1120sector size will be used to set the Raid I/O optimization hints in the 1121superblock. 1122 1123E2fsck will now display a better, more specific error message when the 1124user enters a typo'ed device name, instead of blathering on about 1125alternate superblocks. 1126 1127Fixed various Debian Packaging Issues 1128 1129Updated/clarified man pages (Addresses Debian Bugs: #580236, #594004, 1130#589345, #591083; Addresses Launchpad Bug: #505719) 1131 1132Update the Chinese, Chzech, Dutch, French, Germany, Indonesian, 1133Polish, Swedish, and Vietnamese translations. 1134 1135 1136Programmer's Notes 1137------------------ 1138 1139Fix a dependency definition for the static and profiled blkid 1140libraries which could cause compile failures in some configurations. 1141(Addresses Debian Bug: #604629) 1142 1143Add support for Direct I/O in the Unix I/O access layer. 1144 1145Fixed a memory leak in the Unix I/O layer when changing block sizes. 1146 1147Fixed minor memory leaks in mke2fs. 1148 1149Added a new function to the ext2fs library, ext2fs_get_memalign(). 1150 1151The tst_super_size test program will check to make sure the superblock 1152fields are correctly aligned and will print them out so they can be 1153manually checked for correctness. 1154 1155Fixed some makefile dependencies for test programs in lib/ext2fs. 1156 1157Reserved the feature flags and superblock fields needed for the Next3 1158snapshot feature. 1159 1160Reserved the feature flags for EXT4_FEATURE_INCOMPAT_DIRDATA and 1161EXT4_INCOMPAT_EA_INODE. 1162 1163 1164E2fsprogs 1.41.12 (May 17, 2010) 1165================================ 1166 1167Mke2fs now gives a correct error message if the external journal 1168device is not found. (Addresses Red Hat Bug #572935) 1169 1170Resize2fs -P will now refuse to print a minimum size if the file 1171system is not clean. Previously it would go ahead and print a minimum 1172size anyway, which might not be correct, leading to user confusion. 1173 1174E2fsck now tests for extents that begin at physical block 0 and 1175rejects them as invalid. (Addresses Google Bug: #2573806) 1176 1177Fixed a bug in e2fsck which could cause it to crash when trying to 1178remove an invalid extent and the block bitmaps hadn't yet been loaded. 1179(Addresses SourceForge Bug: #2971800) 1180 1181E2fsck now will completely skip time-based checks if the system clock 1182looks insane or the option broken_system_clock is set in 1183/etc/e2fsck.conf. 1184 1185E2fsck would previously report an i_blocks corruption for a 4T file 1186created using posix_fallocate; this bug has been fixed. 1187 1188E2fsck will now correctly mark a sparse journal as invalid and will 1189delete and recreate the journal to address the problem. 1190 1191E2fsck would previously incorrectly ask the user whether she would 1192like to abort the file system check after finding a problem --- and 1193then abort regardless of the user's answer. This is annoying, and 1194has been fixed. 1195 1196E2fsck can now continue even if it fails to recreate the resize 1197inode; previously it would just abort the file system check 1198altogether. 1199 1200E2fsck could potentially remove directory entries for inodes found in 1201the unused region of the inode table; this would generally happen on 1202ext4 file systems that do not use journaling. This bug has been fixed 1203by not clearing these directory entries once it has been established 1204that bg_unused_inodes may not be trustworthy; once pass #2 has been 1205completed, e2fsck will restart the file system check from the 1206beginning, and then it will be safe to delete any directory entries 1207pointing to inodes that appear to be deleted. (Addresses Google Bug: 1208#2642165) 1209 1210E2fsck will not try to set the block group checksums if the user 1211cancels the fsck with a control-C. It's a bad idea to set the 1212checksums if e2fsck hasn't been completed, and it often results an the 1213error message, "Inode bitmap not loaded while setting block group 1214checksum info". (Addresses Launchpad Bug: #582035) 1215 1216The mke2fs program now queries the kernel for the physical as well as 1217the logical sector size, and will not allow a blocksize below the 1218logical, and will strongly encourage a blocksize at least as big as 1219the physical blocksize. This is needed for 4k sector drives that 1220emulate 512 byte sector sizes. 1221 1222Mke2fs will now allow a flex_bg size of 1. This is unusual, and 1223rarely needed, but it is a legal value. 1224 1225E2fsck will check for cases where the EOFBLOCKS_FL is set when it is 1226not needed, and offer to clear it; this is a sign of a kernel bug, but 1227more importantly, some released kernels may crash when this situation 1228is encountered on ext4 file systems. (Addresses Google Bug: #2604224) 1229 1230E2fsck will use the EOFBLOCKS_FL flag exclusively to check whether 1231i_size is correct. (Kernels starting with 2.6.34 will set 1232EOFBLOCKS_FL.) 1233 1234The com_err library will now only output ^M (a CR character) when the 1235tty is in raw mode. 1236 1237Update the Czech, Chinese, Dutch, French, Germany, Indonesian, Polish, 1238and Vietnamese translations. 1239 1240Fixed various Debian packaging issues --- see debian/changelog for 1241details. (Addresses Debian Bugs: #571247, #563487) 1242 1243 1244Programmer's Notes 1245------------------ 1246 1247The regression test suite now uses its own mke2fs.conf file, so that 1248downstream distributions want change the mke2fs.conf file which is 1249distributed in the RPM or dpkg file, without worrying about screwing 1250up the regression test results. 1251 1252Always build namei.o so that building with configure --disable-debugfs 1253works correctly. Long-term, if we care about reduced e2fsprogs 1254builds, we need a more general solution for deciding what .o files are 1255needed for a particular build. Given that install floppies are going 1256(gone?) the way the dodo bird, we probably don't care, though. 1257(Addresses Sourceforge Bug: #2911433) 1258 1259Add configure options --enable-symlink-build and 1260--enable-symlink-install, which allow e2fsprogs be built using 1261symlinks instad of hard links, and to be installed using symlinks 1262instead of hard links, respectively. It is useful when the file 1263system where the build is taking place, or the file system where 1264e2fsprogs is installed, can't handle hard links for some reason. 1265(Addresses Sourceforge Bug: #1436294) 1266 1267Fixed compile warning in mke2fs.c. 1268 1269 1270E2fsprogs 1.41.11 (March 14, 2010) 1271================================== 1272 1273E2fsck will no longer give a fatal error and abort if the physical 1274device has been resized beyond 2**32 blocks. (Addresses Launchpad 1275Bug: #521648) 1276 1277Debugfs has a bug fixed so that "logdump -b <blk>" now properly shows 1278the allocation status of the block <blk>. (Addresses Debian Bug: 1279#564084) 1280 1281E2fsck now prints a much more emphatic and hopefully scary message 1282when a file system is detected as mounted while doing a read/write 1283check of the filesystem. Hopefully this will dissuade users from 1284thinking, "surely that message doesn't apply to *me*" :-( 1285 1286E2fsck -n will now always open the file system read-only. We now 1287disallow certain combination of options which previously were manual 1288exceptions; this is bad because it causes users to think they are 1289smarter than they really are. So "-n -c", "-n -l", "-n -L", and "-n 1290-D" are no longer supported. (Addresses Launchpad Bug: #537483) 1291 1292In e2fsprogs 1.41.10, mke2fs would ask for confirmation to proceed if 1293it detected a badly aligned partition. Unfortunately, this broke some 1294distribution installation scripts, so it now just prints the warning 1295message and proceeds. (Addresses Red Hat Bug: #569021. Addresses 1296Launchpad Bug: #530071) 1297 1298Mke2fs would take a long time to create very large journal files for 1299ext4. This was caused by a bug in ext2fs_block_iterate2(), which is 1300now fixed. 1301 1302E2fsck now understands the EOFBLOCKS_FL flag which will be used in 13032.6.34 kernels to make e2fsck not complain about blocks deliberately 1304fallocated() beyond an inode's i_size. 1305 1306E2fsprogs 1.41.10 introduced a regression (in commit b71e018) where 1307e2fsck -fD can corrupt non-indexed directories when are exists one or 1308more file names which alphabetically sort before ".". This can happen 1309with ext2 filesystems or for small directories (take less than a lock) 1310which contain filenames that begin with a space or some other 1311punctuation mark. (Addresses Debian Bug: #573923, Addresses Launchpad 1312Bug: #525114) 1313 1314 1315Programmer's Notes 1316------------------ 1317 1318Add new test, f_rehash_dir, which checks to make sure e2fsck -D works 1319correctly. 1320 1321The libcom_err function now has support for Heimdal's com_right_r 1322function(). (Addresses Sourceforge Bug: #2963865, Addresses Debian 1323Bug: #558910) 1324 1325 1326E2fsprogs 1.41.10 (February 7, 2010) 1327==================================== 1328 1329Fix resize2fs bug which causes it to access invalid memory. 1330 1331Add libss support for libreadline.so.6. 1332 1333Fix e2fsck's check for extent-mapped directory with an incorrect file 1334type. 1335 1336Add new e2fsck.conf configuration option, default/broken_system_clock 1337to support systems with broken CMOS hardware clocks. Also, since too 1338many distributions seem to have broken virtualization scripts now, 1339e2fsck will by default accept dates which are off by up to 24 hours by 1340default. (Addresses Debian Bugs: #559776, #557636) 1341 1342Fix a bug where mke2fs may not use the best placement of the inode 1343table when there is only room for a single block group in the last 1344flex_bg. 1345 1346E2fsck is now smarter when it needs to allocate blocks in the course 1347of fixing file system problems. This reduces the number of spurious 1348differences found in pass #5. 1349 1350E2fsck will no longer rehash directories which fit in a single 1351directory block. 1352 1353E2fsck now correctly handles holes in extent-mapped directories (i.e., 1354sparse directories which use extents). 1355 1356Fix big-endian problems with ext2fs_bmap() and ext2fs_bmap2(). 1357 1358Fix a bug in filefrag where on platforms which can allow file systems 1359with 8k blocks, that it doesn't core dump when it sees a file system 1360with 8k block sizes. (Thanks to Mikulas Patocka for pointing this 1361out.) 1362 1363E2fsck will correctly fix directories that are have an inaccurate 1364i_size as well as other problems in a single pass, instead of 1365requiring two e2fsck runs before the file system is fully fixed. 1366 1367Fix e2fsck so it will correctly find and detect duplicate directory 1368entries for non-indexed directories. ( Addresses Sourceforge Bug: 1369#2862551) 1370 1371Mke2fs will use BLKDISCARD to pre-discard all blocks on an SSD or 1372thinly-provisioned storage device. This can be disabled using the -K 1373option. 1374 1375Enhance libext2fs so it works around bug in Linux version 2.6.19 1376and earlier where the /proc/swaps file was missing the header on 1377the first line. 1378 1379Fix bug in Linux version 2.6.19 and earlier where the /proc/swaps file 1380was missing the header on the first line. 1381 1382Fix some big-endian bugs in e2fsck and libext2fs. 1383 1384Fix resize2fs so it works correctly on file systems with external 1385journals instead of failing early with the error "Illegal inode 1386number". 1387 1388Fix libss so that it does not seg fault when using a readline library 1389which does not supply a readline_shutdown() function. 1390 1391Updated dumpe2fs's usage message so it correctly gives the right 1392arguments summary for "-o superblock=<num>" and "-o blocksize=<num>". 1393(Addresses Launchpad Bug: #448099) 1394 1395Teach libext2fs to ignore the high 32 bits of the i_blocks field 1396when huge_file file system feature is set, but the inode does not 1397have the HUGE_FILE_FL flag set. 1398 1399Fix e2fsck's handling of 64-bit i_blocks fields. 1400 1401E2fsck will now print "Illegal indirect block", "Illegal 1402double-indirect block", etc., instead of "Illegal block #-1" or 1403"Illegal block #-2", etc. This makes it easier for users to 1404understand what has gone wrong. (Addresses SourceForge Bug: #2871782) 1405 1406Mke2fs now will obtain get device topology information from blkid and 1407use it to populate the superblock stride and stripe sizes. It will 1408also warn if the block device is misaligned 1409 1410Fix file descriptor leaks in fsck and debugfs. (Addresses Novell Bug: 1411##524526) 1412 1413Fix the libext2fs library code to round up the bitmap size to a 4-byte 1414boundary, to prevent spurious seg faults caused by the x86 1415architecture. This doesn't affect Linux systems, but was a major 1416problem on a number of *BSD systems. (Addresses Sourceforge Bug: 1417#2328708) 1418 1419Fix resize2fs's minimum size required for a file system so it doesn't 1420fail when "resize2fs -M" is run. (Addresses RedHat Bugzilla: #519131) 1421 1422Dumpe2fs now prints summary information about the contents of the 1423journal. 1424 1425Avoid printing scary error messages when e2fsck starts running 1426problems on low-memory systems, as it tends to panic and mislead the 1427user. (Addresses Debian Bug: #509529) 1428 1429Fix blkid's modules.dep parser so it handles compressed (.ko.gz) 1430modules files. (Address Red Hat Bug: #518572) 1431 1432Fix tune2fs so it can add a journal when an extent-enabled file system 1433is mounted. (Addresses Launchpad bug: #416648) 1434 1435Update Czech, Indonesian, Polish and Vietnamese translations (from the 1436Translation Project). 1437 1438Update/clarify man pages. (Addresses Sourceforge Bug: #2822186) 1439 1440Fixed various Debian packaging issues --- see debian/changelog for 1441details. (Addresses Debian Bugs: #540111) 1442 1443Programmer's Notes 1444------------------ 1445 1446The configure script supports the --with-cc, --with-ccopts, and 1447--with-ldopts options. Instead, the more standard use of CC=, 1448CCFLAGS=, and LDFLAGS= in the configure command line is used instead. 1449Also, --with-ld, which never worked, was also removed. 1450(Addresses Sourceforge Bug: #2843248) 1451 1452The in-tree header files are only used if the in-tree uuid or blkid 1453libraries are used. Otherwise, use the system-provided uuid or blkid 1454header files if using the system-provided libraries. 1455 1456Fix some build failures caused by --disable-* configure options. 1457 1458Work around a bug in autoconf 2.64. 1459 1460 1461E2fsprogs 1.41.9 (August 22, 2009) 1462================================== 1463 1464Fix a bug in e2fsck routines for reallocating an inode table which 1465could cause it to loop forever on an ext4 filesystem with the FLEX_BG 1466filesystem feature with a relatively rare (and specific) filesystem 1467corruption. This fix causes e2fsck to try to find space for a new 1468portion of the inode table in the containing flex_bg, and if that 1469fails, the new portion of the inode table will be allocated in any 1470free space available in the filesystem. 1471 1472Make e2fsck less annoying by only asking for permission to relocate a 1473block group's inode table once, instead of for every overlapping 1474block. Similarly, only ask once to recompute the block group 1475checksums, instead of once for each corrupted block group's checksum. 1476 1477Fix filefrag to avoid print the extent header if the FIEMAP ioctl is 1478not present, and it needs to fall back to using the FIBMAP ioctl. 1479 1480Fix filefrag to correctly print the number of extents for zero-length 1481files. (Addresses Debian Bug: #540376) 1482 1483Filefrag now has a -B option which forces the use of the FIBMAP ioctl 1484to more easily debug the FIBMAP code. 1485 1486Fixed filefrag for non-extent based files. 1487 1488Add a new program, e2freefrag, which displays information about the 1489free space fragmentation in an ext2/3/4 filesystem. 1490 1491Fix inode resizing via tune2fs -I so that it works correctly in the 1492face of non-empty bad blocks inodes, and if the filesystem was 1493formatted using the "mke2fs -E stride=N" option for RAID arrays. 1494 1495Fix regression in ext2fs_extent_set_bmap() caused e2fsck -fD to fail 1496and corrupt large directories if the directory needs to shrink by more 1497than one block. (Addresses Debian Bug: #537510) 1498 1499Fix e2fsck's buggy_init_scritps=1 so that the if the last write and/or 1500last mount times are in the future, they are corrected even if 1501buggy_init_scripts is set. This is needed because otherwise resize2fs 1502will refuse to resize the filesystem, even after running "e2fsck -f". 1503(Addresses Launchpad bug: #373409) 1504 1505E2fsck will now print much fuller information when the last mount time 1506or last written time is in the future, since most people can't seem to 1507believe their distribution has buggy init scripts, or they have a 1508failed CMOS/RTS clock battery. 1509 1510Enhance dumpe2fs to dump the extent information via the 'stat' 1511command, and more detailed extent information via the new command 1512'dump_extents'. 1513 1514Update French, Polish, Czech, and Sweedish translation from the 1515Translation Project. 1516 1517Fixed various Debian packaging issues --- see debian/changelog for 1518details. 1519 1520Programmer's Notes 1521------------------ 1522 1523Fixed miscellaneous gcc -Wall warnings. 1524 1525Fixed memory leak in error path in ext2fs_block_iterate2() 1526 1527Fixed non-Linux build of the intl directory by adding support for the 1528E/Q/V macros. 1529 1530The bitmap read/write functions now treat uninitialized bitmaps as 1531unallocated; this fixes a number of problems in all e2fsprogs for ext4 1532filesystems when there is a need to allocate new blocks or inodes, and 1533there aren't any free blocks or inodes in the already-used block 1534groups. 1535 1536Improve ext2fs_extent_set_bmap() to avoid creating new extents which 1537get inserted into the extent tree when they are not needed. 1538 1539 1540E2fsprogs 1.41.8 (July 11, 2009) 1541================================ 1542 1543Fix resize2fs's online resizing, fixing a regression which in 1544e2fpsrogs 1.41.7. (Addresses Debian Bug: #535452) 1545 1546Fix potential filesystem corruptions caused by using resize2fs to 1547shrinking ext4 filesystems with extents enabled. (Addresses Red Hat 1548Bug: #510379) 1549 1550Optimize uuid_generate() to avoid running uuidd if it is not setuid or 1551setgid and the currently running program doesn't have write access to 1552the uuidd work directory. 1553 1554Add safety checks (for non-Linux systems) so that uuidd isn't run with 1555file descriptors 0, 1, and 2 closed; and if they are closed, uuidd 1556will be careful not to close the file descriptor for its unix domain 1557socket when it detaches itself from the controlling tty. Also add 1558safety checks so that if the unix domain socket between the uuid 1559library and uuidd program is closed for any reason, both the library 1560and the uuidd will return an appropriate error code instead of looping 1561in an infinite loop. 1562 1563The e2croncheck script, which creates an LVM snapshot and then checks 1564the ext3/4 filesystem via the LVM snapshot, has been added to the 1565contrib directory. 1566 1567Fix filefrag program for files that have more than 144 extents. 1568 1569Update French, Polish, Czech, Indonsian, and Sweedish translation from 1570the Translation Project. 1571 1572Fixed various Debian packaging issues --- see debian/changelog for 1573details. (Addresses Debian Bug #535530) 1574 1575Update/clarify man pages. 1576 1577Programmer's Notes 1578------------------ 1579 1580Fix compilation problem when configured with --disable-uuid. 1581 1582Don't build uuidgen if configured with --disable-uuid. 1583 1584Add the new library function ext2fs_test_inode_bitmap_range(), and 1585optimized ext2fs_test_block_bitmap_range(), which will be needed for 1586future optimizations for e2fsck. 1587 1588Fix makefile dependencies for libcom_err so that the Makefiles work 1589well on non-GNU make program. (Addresses Sourceforge Patches: #2813809) 1590 1591Enhance the build system so that the full set of commands executed by 1592the Makefiles are displayed, instead of the Linux kernel summary 1593output, if the build was configured with --enable-verbose-makecmds, or 1594if GNU make is in use and the V variable is non-null, i.e., via "make 1595V=1". 1596 1597 1598E2fsprogs 1.41.7 (June 29, 2009) 1599================================ 1600 1601Fix a bug in libext2fs which can cause e2fsck and resize2fs to write 1602uninitalized data into the portion of the inode beyond the first 128 1603bytes when operating on inodes mapped via extents; potentially 1604corrupting filesystems. 1605 1606Fix memory leaks in e2fsprogs, including a very large memory leak 1607which can cause e2fsck to run out of memory when checking very large 1608filesystems using extents. 1609 1610The logsave program will now filter out the ^A and ^B characters when 1611writing to the console. 1612 1613Harden ext2fs_validate_entry() so that lsdel will not read beyond the 1614end of the block even if the directory block is corrupted. 1615 1616Fix debugfs from core dumping if the logdump command fails to open the 1617output file. 1618 1619Enhance badblocks to print the currently tested block number when 1620interrupted with ^C. 1621 1622Fix lsattr to exit with a non-zero status when it encounters errors. 1623(Addresses RedHat Bugzilla #489841) 1624 1625Fix e2fsprogs to use the same encoding for rec_len in 64k (and larger) 1626filesystems as the kernel when encoding rec_len >= 64k. For 64k 1627filesystems (currently all that is supported by e2fsprogs, this is 1628only a minor corner case). 1629 1630Resize2fs will now update the journal backup fileds in the sup0erblock 1631if the journal is moved; this avoids an unnecessary full fsck after 1632resizing the filesystem. (Addresses RedHat Bugzilla: #505339) 1633 1634Fix libext2fs to properly initialize i_extra_size when creating the 1635journal and resize inodes. 1636 1637Change badblocks to allow block sizes larger than 4k. 1638 1639Fix filefrag progam so it correctly checks for errors from the fiemap ioctl. 1640 1641Update Chinese and Czech translation from the Translation Project. 1642 1643Clean up various man pages. (Addresses Debian Bug #531385 and #523063) 1644 1645 1646Programmer's Notes 1647------------------ 1648 1649Add --disable-libuuid option to configure which uses an in-system 1650installed version of libuuid; the private version is enabled by 1651default. 1652 1653Add --valgrind-leakcheck option to the test_script program which runs 1654valgrind with the appropriate options to find memory leaks. 1655 1656 1657E2fsprogs 1.41.6 (May 30, 2009) 1658=============================== 1659 1660Fix a critical regression in e2fsck introduced version 1.41.5 which 1661can undo updates to the block group descriptors after a journal replay. 1662 1663If e2fsck receives an I/O error while replaying the journal, prompt 1664the user whether they want to ignore the error or not; if e2fsck is 1665run in preen mode, force fsck to abort the preen pass. 1666 1667Fix a bug which would cause e2fsck to core dump if the filesystem 1668contains a corrupt extent header, and the user declines to clear the 1669inode. (Addresses Sourceforge Bug: #2791794) 1670 1671Fix e2fsck to restart only once in the case of multiple block groups 1672which inodes in use that are in the uninitialized part of the block 1673group's inode table. 1674 1675To reduce user confusion, if the /etc/mtab file is missing 1676ext2fs_check_mount_point and ext2fs_check_if_mounted will return a 1677new, explicit error code to indicate this case. This will cause 1678e2fsck to give a clearer error message when the user is using buggy 1679rescue CD's that don't properly set up /etc/mtab. (Addresses Debian 1680Bug: #527859) 1681 1682Fix e2fsck so that if the primary group descriptors are corrupted, but 1683the backup superblock is entirely invalid, to go back to using (and 1684fixing) the priary group descriptors instead of completely giving up 1685on the filesystem. (Addresses Debian Bug: #516820) 1686 1687Change e2fsck to not abort a preen pass if an inode is found to have 1688i_file_acl_hi non-zero. Ext3 filesystems don't care, and newer 1689kernels (post 2.6.29) will ignore this field. So let's fix it 1690automatically during the preen pass instead of forcing the user to fix 1691this up manually. (Addresses Debian Bug: #526524) 1692 1693Add resource tracking for e2fsck passes 1b through 1d. 1694 1695Speed up e2fsck by eliminating unnecessary journal checks if the 1696filesystem is already mounted and doesn't need recovery (since the 1697kernel would have run the journal when the filesystem was mounted.) 1698Also speed up e2fsck by avoiding unnecessary block group descriptor 1699updates in ext2fs_close(). 1700 1701Add support to chattr to migrate inodes from using direct/indirect 1702blocks to extents. 1703 1704Avoid corrupting the filesystem if there is an attempt to shrink a 1705filesystem using resize2fs smaller than posible by making 1706ext2fs_set_bmap() more careful not to delete the old block until the 1707new block can be inserted. In addition, fix a bug in how the minimum 1708size of the filesystem (plus a safety margin) is calculated, and 1709modify resize2fs to refuse to shrink the filesystem below that minimum 1710size without the force flag. 1711 1712Teach blkid to try to figure out DM device names using take advantage 1713of information in sysfs when running on kernels (2.6.29 and later) 1714which provide this information in /sys/block/dm-<N>/dm/name; this is 1715much faster than scanning for the devncie number in /dev/mapper. 1716 1717Fix blkid to prefer /dev/mapper/<name> device names instad of the 1718private /dev/dm-X names. 1719 1720Add an -a option to debugfs's close_filesys command which writes any 1721changes to the superblock or block group descriptors to all of the 1722backup superblock locations. 1723 1724Add support to the filefrag program to use the FIEMAP ioctl. 1725 1726Update Chinese translation from the Translation Project. 1727 1728Clean up various man pages. (Addresses Red Hat Bugzilla: #502971 and 1729Launchpad Bug: #381854) 1730 1731Fixed various Debian packaging issues --- see debian/changelog for 1732details. (Addresses Debian Bug #506064) 1733 1734 1735Programmer's Notes 1736------------------ 1737 1738Add test code to make sure e2fsck's problem.c doesn't have two problem 1739codes assigned to duplicate values. 1740 1741Avoid using a hard-coded path for /bin/rm in block's test_probe.in. 1742(Addresses Sourceforge Bug: #2780205) 1743 1744Clean up e2fsck by removing #ifdef RESOURCE_TRACK by adding an empty 1745function declaration for init_resource_track() and 1746print_resource_track() when RESOURCE_TRACK is not defined. 1747 1748The test code which is used to build the tst_csum progam has been 1749moved from from tst_csum.c into csum.c under an #ifdef DEBUG to to 1750avoid compile problems caused by not having a prototype for 1751ext2fs_group_desc_csum(). (Addresses Sourceforge Bug #2484331) 1752 1753Update the config.guess and config.sub file to the latest from the 1754FSF, to allow e2fsprogs to build on the avr32 platform. (Addresses 1755Debian Bug: #528330) 1756 1757Add a new function, ext2fs_extent_open2(), which behaves like 1758pext2fs_extent_open(), but uses the user-supplied inode structure 1759when opening an extent instead of reading the inode from disk. 1760 1761 1762E2fsprogs 1.41.5 (April 23, 2009) 1763================================= 1764 1765Fix a number of filesystem corruption bugs in resize2fs when growing 1766or shrinking ext4 filesystems off-line (i.e., when the ext4 filesystem 1767is not mounted). 1768 1769Debugfs can now set i_file_acl_high via the set_inodes_field command, 1770and print a 64-bit file acl. This is useful for debugging filesystem 1771corruptions where the high bits of i_file_acl_high are set. E2fsck 1772will detect and fix non-zero i_file_acl_high on 32-bit filesystems 1773since some Linux kernel versions pay attention to this field even when 1774they shouldn't. 1775 1776Speed up e2fsck when checking clean filesystems by avoiding 1777unnecessary block reads, and coalescing the block group descriptor 1778blocks so they are read using a single read operation. 1779 1780The libuuid library will now close all file descriptors before running 1781uuidd. This avoids problems when the calling program has open sockets 1782which then never get closed because uuidd is a long-running helper 1783daeomn. (Addresses Launchpad bug: #305057) 1784 1785In order to avoid unnecessary full filesystem checks by e2fsck after 1786an on-line resize, e2fsck will ignore the NEEDS_RECOVERY flag set on 1787the backup superblocks. (Addresses Red Hat Bugzilla: #471925) 1788 1789Mke2fs will avoid trying to create the journal when run in 1790superblock-only mode (mke2fs -S), since the left-over journal in the 1791inode table will cause mke2fs to fail. 1792 1793Fix a bug in libext2fs functions that check to see if a particular 1794device or filesystem image is mounted, which would cause these 1795functions to report that a file identical to the (relative) pathname 1796used by a pseudo-filesystem was mounted when in fact it was not. 1797 1798Update Czech translation from the Translation Project. 1799 1800Add Chinese (simplified) translation from the Translation Project. 1801 1802Fix support for external journals (which was broken in e2fsprogs 18031.41.4). 1804 1805Fix a regression in debugfs where the "stat" command when no 1806filesystem was open would cause debugfs to crash with a segmentation 1807violation. 1808 1809Starting in the 2.6.29 linux kernel, the ext4 filesystem driver can be 1810used to support filesystems without a journal. Update the blkid 1811library so it understands this. 1812 1813The blkid library will remove an entry from the blkid cache 1814representing the entire disk if partitions are found, since presuambly 1815the device previously had no partition table, but has now transitioned 1816to using a partition table. 1817 1818Add a check to mke2fs and tune2fs that the argument to the -m option 1819(which specifies the reserved ratio) must be greater than zero. 1820(Addresses Debian Bug: #517015) 1821 1822Add support for tracking the number kilobytes written to the 1823filesystem via the superblock field s_kbytes_written. It will be 1824updated by the kernel as well as by e2fsprogs programs which write to 1825the filesystem. This is useful for tracking the wear to filesystems 1826on Solid Sstate Drives. 1827 1828Fix compatibility issue in the libext2fs info file and makeinfo 1829version 4.12. (Addresses Red Hat Bugzilla: #481620) 1830 1831Update/clarify man pages. (Addresses Debian Bug: #515693, #365619) 1832 1833Fixed various Debian packaging issues --- see debian/changelog for 1834details. (Addresses Debian Bug: #506279) 1835 1836Programmer's Notes 1837------------------ 1838 1839Fix Hurd compilation problem in e2fsck and tune2fs (Addresses Debian 1840Bug: #521602) 1841 1842Fix various gcc compilation warnings and other programming cleanups. 1843(Addresses Red Hat Bugzilla: #486997) 1844 1845Add support for building the blkid command statically. 1846 1847Add support for disabling the built-in blkid library in favor of a 1848system-installed blkid implementation via the configure option 1849--disable-libblkid. 1850 1851 1852E2fsprogs 1.41.4 (January 27, 2009) 1853=================================== 1854 1855Fixed a bug which could sometimes cause blkid to return an exit value 1856of zero for a non-existent device (Addresses Debian Bug: #502541) 1857 1858Blkid will now recognize ext3 filesystems that have the test_fs flag 1859set as ext3 filesystems. 1860 1861The blkid library will now recognize btrfs filesystems and swap 1862devices currently used by user-level software suspend. 1863 1864Tune2fs now updates the block group checksums when changing the UUID 1865to avoid causing e2fsck to complain vociferously at the next reboot. 1866 1867Tune2fs's inode size resizing algorithms have been fixed so it is not 1868vastly inefficient for moderate-to-large filesystems, due to some 1869O(n**2) and O(n*m) algorithms that didn't scale well at all. 1870 1871Fix tune2fs's inode resizing algorithm so it will not corrupt 1872filesystems laid out for RAID filesystems; in addition, tune2fs will 1873refuse to change the inode size for filesystems that have the flex_bg 1874feature enabled. (This is a limitation in the current implementation 1875of tune2fs -I.) 1876 1877E2fsprogs 1.41 broke debugfs's logdump command for normal ext3/4 1878filesystems with 32-bit block numbers, when the headers for 64-bit 1879block numbers was added. This regression has been fixed. 1880 1881Debugfs's ncheck command has been fixed to avoid printing garbage 1882characters at the end of file names. 1883 1884Fix resize2fs for ext4 filesystems. Some blocks that that need moving 1885when shrinking filesystems with uninit_bg feature would not be moved. 1886In addition, blocks and inode table blocks were not being correctly 1887freed when shrinking filesystems with the flex_bg feable, which caused 1888resize2fs -M to fail. Finally, when blocks are moved, make sure the 1889uninitialized flag in extents is preserved. 1890 1891Fix bug which caused dumpe2fs to abort with an error if run on a 1892filesystem that contained an external journal. 1893 1894Some distributions used "mke3fs" as an alias for "mkfs.ext3"; check 1895for this in argv[0] to provide better legacy support for these 1896distributions. This is a practice that should NOT be continued, 1897however. 1898 1899Mke2fs now has a new option -U, which allows the user to specify the 1900UUID that should be used for the new filesystem. 1901 1902Mke2fs will treat devices that are exactly 16TB as if they were 16TB 1903minus one block. This allows users who have read that ext3 supports 1904up to 16TB filesystems and who create a 16TB LVM to not get confused, 1905since the true limit is really 16TB minus one block. 1906 1907E2fsck will no longer abort an fsck run if block group has an errant 1908INODE_UNINIT flag. 1909 1910E2fsck now distinguishes between fragmented directories and fragmented 1911files in verbose mode statistics and in the fragcheck report. 1912 1913Fix a bug in e2fsck which casued it double count non-contiguous 1914extent-based inodes. 1915 1916E2fsck will leave some slack space when repacking directories to allow 1917room for a few directory entries to be added without causing leaf 1918nodes to be split right away. 1919 1920Fix a bug which caused e2fsck to crash when it comes across a 1921corrupted interior node in an extent tree with the error message: 1922"Error1: Corrupt extent header on inode XXXXXX" 1923 1924E2fsck problem descriptions involving the journal are no longer 1925referred to as "ext3" problems, since ext4 filesystems also have 1926journals. 1927 1928Fix a long-standing bug in e2fsck which would cause it to crash when 1929replying journals for filesystems with block sizes greater than 8k. 1930 1931Update Catalan translation from the Translation Project. 1932 1933Fixed various Debian packaging issues --- see debian/changelog for 1934details. (Addresses Debian Bugs: #503057, #502323, #511207) 1935 1936Programmer's Notes 1937------------------ 1938 1939Fix build of e2fsck.profiled, and add support for building profiled 1940binaries in the misc directory if configured with --enable-profile. 1941 1942The ext2fs_open() function now performs more sanity checks on the 1943superblock to avoid potential divide by zero errors by other parts of 1944library. 1945 1946The ext2fs_read_inode_full() function now has a safety check to avoid 1947a segmentation fault on corrupted filesystems. 1948 1949The ext2fs_new_inode() function now has a sanity check so that if the 1950s_first_inode field in the superblock is insane, it will return 1951EXT2_ET_INODE_ALLOC_FAIL instead of returning an invalid inode number. 1952 1953To avoid segmentation faults, ext2fs_block_alloc_stats() and 1954ext2fs_inode_alloc_stats() now validates the passed inode or block 1955number to avoid overrunning an array boundary. 1956 1957Various signed/unsigned errors for variables containing block numbers 1958have been fixed. 1959 1960Accomodations for gcc's stupidity in not realizing that constant 1961strings that do not contain a '%' character are safe to use in format 1962strings have been made so that distributions that want to compile 1963e2fsprogs with -Werror=format-security have an easier time doing so. 1964 1965Added a new 64-bit getsize interface, ext2fs_get_device_size2(). 1966 1967Added the utility make-sparse.c to the contrib directory. 1968 1969The ext2fs_block_iterate2() function now reflects errors from 1970ext2fs_extent_set_bmap() to the caller, if the callback tries to 1971change a block on an extent-based file, and ext2fs_extent_set_bmap() 1972fails for some reason (for example, there isn't enough disk space to 1973split a node and expand the extent tree. 1974 1975The ext2fs_block_iterate2() function will preserve the uninit flag in 1976extents when the callback function modifies a block in an extent-based 1977file. 1978 1979E2fsck will now flag filesystems that have an insane s_first_ino field 1980in their superblock, and attempt to use a backup superblock to repair 1981the filesystem. 1982 1983 1984E2fsprogs 1.41.3 (October 12, 2008) 1985=================================== 1986 1987E2fsck has been fixed so it prints the correct inode number for 1988uinit_bg related problems. 1989 1990E2fsck will now offer to clear the test_fs flag if the ext4 filesystem 1991is available on linux. This can be disabled via a configuration 1992option in /etc/e2fsck.conf. 1993 1994Fix a file descriptor leak in libblkid when checking to see if an ext4 1995or ext4dev module exists. 1996 1997Fix a bug in e2fsck where in preen mode, if there are disk I/O errors 1998while trying to close a filesystem can lead to infinite loops. 1999(Addresses Red Hat Bugzilla #465679) 2000 2001Fix a bug in resize2fs where passing in a bogus new size of 0 blocks 2002will cause resize2fs to drop into an infinite loop. (Addresses Red 2003Hat Bugzilla: #465984) 2004 2005Add a check in the Unix I/O functions in libext2fs so that when a 2006device is opened read/write, return an error if the device is 2007read-only using the BLKROGET ioctl. 2008 2009Fix debugfs's ncheck command so that it prints all of the names of 2010hardlinks in the same directory. 2011 2012Fix a bug in libblkid so it correctly detects whether the ext4 and 2013ext4dev filesystems are available, so that the ext4dev->ext4 2014fallback code works correctly. 2015 2016Programmer's Notes 2017------------------ 2018 2019Fix a parallel build problem by making sure util/subst is built before 2020trying to build the lib/et directory. (Addresses Sourceforge Bug: 2021#2143281) 2022 2023Updated "make depend" information for crc16.o 2024 2025 2026E2fsprogs 1.41.2 (October 2, 2008) 2027================================== 2028 2029Fix e2fsck's automatic blocksize detection. This fixes a regression 2030from e2fsprogs 1.40.7 which caused e2fsck to fail if the user 2031specifies a block number using the -b option if the blocksize option 2032isn't also specified using -B. Unfortunately, users very commonly 2033invoke e2fsck using "e2fsck -b 32768 /dev/hdXXX" to use the backup 2034superblock; in fack e2fsck will often suggest this kind of command 2035line. Oops. 2036 2037Enhance the debugfs's "ncheck" command so it will print all of the 2038pathnames for the specified inodes. (Previously, in some cases ncheck 2039might not print a pathname for an inode at all if some of the other 2040inodes had multiple hard links.) 2041 2042Enhance debugfs's "hash" command so the hash seed can be specified via 2043a command-line option. In addition, allow the hash algorithm to be 2044specified by name instead of just by number. 2045 2046Fix e2fsck so that we don't accidentally print the translation file's 2047header when asking the user a custom question so there is no prompt 2048defined for a particular problem record. For example, the question 2049"Run journal anyway" will get the PO header tacked on because e2fsck 2050erroneously passed the null string to _(). (Addresses Launchpad Bug: 2051#246892) 2052 2053Enhance badblocks so that it can test a normal file which is greater 2054than 2GB. 2055 2056Enhance the badblocks command so that it displays the time and 2057percentage complete when in verbose mode. (Addresses Debian Bug: 2058#429739) 2059 2060Fix a potential memory leak in a error handling path in debugfs's 2061ncheck function. 2062 2063Fix a potential memory corruption problem if a memory allocation fails 2064in resize2fs. 2065 2066Fix the usage message for debugfs's logdump command to be consistent 2067with its man manpage. 2068 2069Update Polish, French, Vietnamese, Dutch, Indonesian, German, Czech, 2070and Sweedish translation from the Translation Project. 2071 2072Add documentation for the file I/O functions to the libext2fs.texinfo 2073file. (Addresses Debian Bug: #484877) 2074 2075Update and clarified various man pages. (Addresses Launchpad Bug 2076#275272; Addresses Debian Bugs: #498100, #498101, #498102, #498103) 2077 2078Fixed various Debian packaging issues --- see debian/changelog for 2079details. (Addresses Debian Bug: #497619) 2080 2081Programmer's Notes 2082------------------ 2083 2084Fix a potential file descriptor leack in libcom_err by setting the 2085close-on-exec flag for a fd used for debugging. (Addresses Red Hat 2086Bugzilla #464689) 2087 2088Fix a potential race in libcom_err by using sem_post/sem_init. SuSE 2089has been carrying a patch for a long time to prevent a largely 2090theoretical race condition if a multi-threaded application adds and 2091removes error tables in multiple threads. Unfortunately SuSE's 2092approach breaks compatibility by forcing applications to link and 2093compile with the -pthread option; using pthread mutexes has 2094historically been problematic. We fix this by using sem_post/sem_init 2095instead. 2096 2097Fix e2fsprogs-libs build failure due to 'subs' target. (Addresses 2098Sourceforge Bug: #2087502) 2099 2100Avoid linking e2initrd_helper, debugfs, blkid, and fsck with unneeded 2101libraries when using ELF shared libraries. 2102 2103Fix ELF shared library when building on systems that don't already 2104have the e2fsprogs shared libraries already installed. (Addresses 2105Sourceforge Bug: #2088537) 2106 2107Fix the pkg-config files so they work correctly when linking with 2108static libraries and fix the include directory so programs don't have 2109to use #include <ext2fs/ext2fs.h>, but can use #include <ext2fs.h> 2110instead. (Addresses Sourceforge Bug: #2089537) 2111 2112Make sure ext2fs_swab64() is compiled for all platforms, and not just 2113for x86. (Addresses Debian Bug: #497515) 2114 2115Remove the unused ext2fs_find_{first,ext}_bit_set() functions for all 2116non-x86 platforms. (They had been removed for x86 earlier.) 2117 2118Fix diet libc compilation support, which had bitrotted due to lack of 2119TLC. Fixing this improves general portability. 2120 2121When installing the link library when using ELF shared libraries, 2122avoid using absolute pathnames if the link library and the shared 2123library are installed in the same directory. (Addresses Sourceforge 2124Bug: #1782913) 2125 2126Fix gen-tarball so it will work even if the top-level directory has 2127been renamed to something other than "e2fsprogs". Also make 2128gen-tarball print the size of the resulting tar.gz file. 2129 2130 2131E2fsprogs 1.41.1 (September 1, 2008) 2132==================================== 2133 2134Many people are forgetting to update their mke2fs.conf file, and this 2135causes ext3, ext4, and ext4dev filesystems won't get created with the 2136proper features enabled. We address this in two ways. First, mke2fs 2137will issue a warning if there is not definition for an ext3, ext4, or 2138ext4dev filesystem and the user is trying to create such a filesystem 2139type. Secondly, when installing from a source build, "make install" 2140will provide basic configuration file handling for /etc/mke2fs.conf. 2141If it exists, and does not mention ext4dev, it will be moved aside to 2142/etc/mke2fs.conf.e2fpsrogs-old and the new /etc/mke2fs.conf file will 2143be installed. If the existing /etc/mke2fs.conf file does mention 2144ext4dev, then "make install" will install official mke2fs.conf file as 2145/etc/mke2fs.conf.e2fsprogs-new and issue a message to the user that 2146they should look to see if any changes need to be merged. 2147 2148The mke2fs program will now create the journal in the middle of the 2149filesystem, since this minimizes seek times on average for fsync-heavy 2150workloads. In addition, mke2fs will now create journals using extents 2151for filesystems that support them. This results in a more efficient 2152encoding for the journal since it eliminates the need for using 2153indirect blocks. 2154 2155The mke2fs program will avoid allocating an extra block to the 2156journal. (Addresses Sourceforge Bug: #1483791) 2157 2158Mke2fs will correctly enforce the prohibition against features 2159(specifically read-only features) in revision 0 filesystems. (Thanks 2160to Benno Schulenberg for noticing this problem.) 2161 2162Mke2fs previously would occasionaly create some slightly non-optimally 2163placed inode tables; this bug has been fixed. 2164 2165The mke2fs and tune2fs programs now print the correct usage message 2166describing the maximum journal size. (Addresses Debian Bug: #491620) 2167 2168Add support for setting the default hash algorithm used in b-tree 2169directories in tune2fs (from a command-line option) or mke2fs (via 2170mke2fs.conf). In addition, change the default hash algorithm to 2171half_md4, since it is faster and better. 2172 2173The blkid library will now recognize MacOS hfsx filesystems, and 2174correctly extract the label and uuid for hfs, hfsx, and hfsplus 2175filesystems. (Addresses Sourceforge Feature Requests: #2060292) 2176 2177The blkid library has improved detection of JFS and HPFS filesystems. 2178(Addresses Launchpad Bug: #255255) 2179 2180The blkid library is now much more efficiently handling devicemapper 2181devices, mainly by no longer using the devicemapper library. This can 2182speed up access for systems with a large number of device mapper 2183devices. 2184 2185Blkid had a number of cache validation bugs in libblkid that have been 2186fixed. (Addresses Debian Bug: #493216) 2187 2188Resize2fs will now properly close out the "updating inode references" 2189progress bar so there is a newline printed before printing the final 2190"resize is successful" message. 2191 2192Resize2fs will now correctly handle filesystems with extents and/o 2193uninitialized block groups correctly when file/directory blocks need 2194to relocated (i.e., when shrinking a filesystem or if the resize_inode 2195is not present). To support this, the ext2fs library now supports 2196initializing inode and block bitmaps that are not yet initialized when 2197allocating them using ext2fs_new_block() and ext2fs_new_inode(). In 2198addition, e2fs_block_iterate2() can now support changing the location 2199of interior nodes of an extent tree, and ext2fs_extent_set_bmap() has 2200been optimized to avoid creating unnecessary new extents when updating 2201the location of blocks in the extent tree. This will also help out 2202e2fsck's recovery of obscurely corrupted filesystems with extents, 2203when blocks are claimed by multiple inodes. 2204 2205Add support for on-line resizing ext4 filesystem with the flex_bg 2206filesystem feature. The method for doing so is not optimal, but to do 2207a better job will require kernel support. 2208 2209E2fsprogs 1.41.0 intrduced a bug in libext2fs which casued e2image and 2210debugfs programs to not be able to read e2image files; the signed 2211vs. unsigned bug in the code which read bitmaps from the e2image has 2212been fixed. (Addresses Debian Bug: #495830) 2213 2214Resize2fs is now correctly managing the directory in-use counts when 2215shrinking filesystems and directory inodes needed to be moved from one 2216block group to another. This bug has been around since e2fsprogs 22171.26, and is largely harmless, but does cause a filesystem corruption 2218which will be flagged by e2fsck after the filesystem has been shrunk. 2219 2220E2fsck will no longer issue spurious complaints about the inode size 2221caused by very large extent-based files, and by blocks reallocated 2222using fallocate() with the FALLOC_FL_KEEP_SIZE option. (Addresses 2223Kernel Bugzilla: #11341) 2224 2225Mke2fs will now set the creation timestamp on the lost+found directory 2226and the root directory. (More generally, all new inodes created using 2227the ext2fs library will correctly set the creation timestamp.) 2228 2229E2fsck now correctly calculates ind/dind/tind statistics in the 2230presence of extent-based files. In addition, "e2fsck -v" will report 2231statistics of the depth of extent trees in the filesystem. E2fsck can 2232also give an inode fragmentation report using "e2fsck -E fragcheck" 2233which can be useful when debugging the kernel block allocation 2234routines. 2235 2236Fix support for empty directory blocks in ext4 filesystems with 64k 2237blocksize filesystems. 2238 2239E2fsck will now print the depth of corrupt htree directories. 2240 2241Debugfs's htree command now correctly understands extent-based 2242directories. It will also print out the minor hash as well as the 2243major hash. 2244 2245Debugfs has a new command which will print the supported features of 2246e2fsprogs, to enable scripts to know whether the installed version of 2247e2fsprogs can support a specific feature. 2248 2249Debugfs will now write files using extents for filesystems that 2250support them. 2251 2252The error message printed by "tune2fs -I" if the inode size was too 2253small was rather confusing, so it has been improved. Also, we won't 2254try to create an undo log until we know that command-line-specified 2255parameters such as "tune2fs -I <inode size>" are valid. 2256 2257Given some filesystems found "in the wild" that had non-zero block 2258group checksums even though the uninit_bg/gdt_sum feature was not 2259enabled, e2fsck would issue spurious error messages. Teach 2260ext2fs_group_desc_csum_verify() to ignore the block group checksum 2261entirely if the feature flag is not set. (Addresses Debian Bug: 2262#490637) 2263 2264The blkid program will now print out a user-friendly listing of all of 2265the block devices in the system and what they contain when given the 2266-L option. (Addresses Debian Bug: #490527) 2267 2268The filefrag program now has a more accurate calculation for the 2269number of ideal extents. (Addresses Debian Bug: #458306) 2270 2271The test I/O manager is now enabled by default, but its overhead is 2272only incurred when it would be enabled via the TEST_IO_FLAGS or 2273TEST_IO_BLOCK environment variables. 2274 2275Typographical errors in various program strings and usage messages 2276have been fixed; most of these were pointed out by the e2fsprogs 2277message catalog translators. (Thanks, translators!) 2278 2279Update and clarified various man pages, as well as some typographical 2280errors in the libext2fs texinfo file. 2281 2282Fixed various Debian packaging issues --- see debian/changelog for 2283details. 2284 2285Add Indonesian and update French, Polish, Dutch, German, Sweedish, 2286Czech, and Vietnamese Translations. (Addresses Debian Bugs: #313697, 2287#401092) 2288 2289Programmer's Notes 2290------------------ 2291 2292Fix portability problem with the badblocks group; for systems that 2293don't have nanosleep(), try using usleep() instead. 2294 2295The "make check" target in the e2fsck directory now sets 2296LD_LIBRARY_PATH before running the various e2fsck internal library 2297regression tests. 2298 2299The crc32 regression test in the e2fsck library is now portable to 2300greater varienty of environments, including big-endian systems and 2301when cross-building e2fsprogs for embedded systems. (Addresses 2302Sourceforge Bug: #2019287) 2303 2304The ext2fs_extent_set_bmap() had some bugs when setting the first 2305block in a file, or when replacing a single block extent. Those cases 2306fortunately were came up relatively rarely when e2fsck was checking 2307files, but caused some problems when resize2fs was shrinking 2308extent-based files. 2309 2310Fix a potential core-dumping bug in libe2p's iterate_on_dir() 2311function. 2312 2313Various ext2fs library functions --- ext2fs_block_iterate2(), 2314ext2fs_initialize() and ext2fs_extent_open() --- now correctly free 2315allocated memory to avoid memory leaks in all of their error return 2316paths. 2317 2318Ext2ed was failing to build because masix support had been removed in 2319the rest of e2fsprogs, so ext2ed no longer has masix support, either. 2320 2321The configure script now respects the LDFLAGS environment variable if 2322it is set when configure is called. (Addresses Sourceforge Feature 2323Request: #1937287) 2324 2325Libuuid is now more portable to the Windows platform. (Addresses 2326Sourceforge Feature Request: #1937287) 2327 2328The configure script now uses AC_MSG_{RESULT,WARN,ERROR} instead of 2329bare echo commands so that configure flags such as --quiet work 2330correctly. (Addresses Sourceforge Patches: #2058794) 2331 2332A few uses of sprintf have been removed from the ext2fs library to 2333make life easier for bootloaders with a limited libc environment. 2334(Addresses Sourceforge Bug: #2049120) 2335 2336The ext2fs_read_inode() checks the validity of the inode number passed 2337to it earlier, to avoid doing some needless work when it would fail 2338anyway. 2339 2340The ext2fs_open() checks the validity of the blocksize parameter 2341passed to it earlier, to avoid doing some needless work when it would 2342fail anyway. 2343 2344Disable a very annoying automatic "%.sh -> %" GNU make rule in the 2345top-level Makefile. That automatic rule is used to better support 2346SCCS, but it caused problems for a particular niche distribution which 2347likes to use configure.sh files to store the configure options used to 2348build a package. Unfortuntaely GNU make will use the configure.sh to 2349replace the configure script, resulting in a self-inflicted fork bomb 2350leading to an out-of-memory crash. 2351 2352To support old GNU C compilers don't use C99/C++ comments, but only 2353K&R style comments, and don't try to use __builtin_expect if __GNUC__ 2354is less than 3. (__builtin_expect is only supported for gcc versions 23552.96 and up, and it's tricky to check for gcc 2.95 vs gcc 2.96; since 2356this is an optimization, we only try to use __builtin_expect for gcc 3 2357and up.) 2358 2359In e2fsck's crc routines, make sure we use WORDS_BIGENDIAN instead of 2360__LITTLE_ENDIAN, which are only defined by glibc's header files and 2361hence isn't portable. 2362 2363For the convenience for some distributions that need a static tune2fs, 2364the Makefile for misc/ now has a tune2fs.static target. 2365 2366The ext2fs_block_iterate2() function now supports BLOCK_FLAG_APPEND 2367for extent-based files 2368 2369The ext2fs_bmap() function now supports BMAP_ALLOC for extent-based 2370files. 2371 2372All source files no longer have any trailing white space. 2373 2374The io_channel_read_blk64() and io_channel_write_blk64() functions are 2375now functions instead of C preprocessor macros to provide better 2376forward compatibility. 2377 2378The e2fpsrogs translation template now expands the @x abbrevation. 2379 2380Various namespace leackages in libblkid, libe2p, and libext2fs have 2381been fixed. 2382 2383Fix a parallel build problem in e2fsprogs. 2384 2385E2fsprogs is now more portable to Solaris. 2386 * blkid no longer assumes that the TIOCGSIZE and TIOCGWINSZ ioctl's 2387 are always present. 2388 * Scripts do not assume that /bin/true is always in /bin 2389 * Don't use __FUNCTION__ since Solaris's C99 doesn't support it. 2390 * Flush stdio handles before calling setbuf(), since Solaris will 2391 discard any pending output to the stream. 2392 * Define _XOPEN_SOURCE to 600 since Solaris's header files are very 2393 picky about which C compiler can beused for SUSv3 conformance. 2394 Use of C99 is not compatible with SUSv2 (_XOPEN_SOURCE=500), 2395 and C89 is not compatible with SUSv3 (_XOPEN_SOURCE=600). 2396 Since we need some SUSv3 functions, consistently use SUSv3 so 2397 that e2fsprogs will build on Solaris using c99. 2398 * Solaris C99 does not support varargs C preprocessor macros 2399 * Solaris header files pollute the C namespace if in/netinet.h 2400 is included, which conflicts with e2fsprogs' use of the kmem_cache_t 2401 typedef. 2402 * Solaris ships with a pathetically ancient shell in /bin/sh so we 2403 avoid the use of various more avanced shell constructs such as $(). 2404 2405The "make rpm" command will now take some extra configure optiosn from 2406the build environment without needing to patch the source tree. 2407 2408The ext2fs_add_dir_block() function will now grow the dblist more 2409aggressively as an optimization to avoid copying the array too often. 2410 2411The e2fsck_write_bitmaps() will write the block and inode bitmaps 2412together instead of in two passes. 2413 2414 2415E2fsprogs 1.41 (July 10, 2008) 2416============================== 2417 2418Add support for ext4 filesystem features, in particular extents, 2419uninit_bg, flex_bg, huge_file, and dir_nlink features. Also add 2420support for checking journal checksums. Debugfs will print new 2421superblock and inode fields that were defined for ext4. For example, 2422the nanosecond and i_version fields of an inode, and the 2423s_min_extra_isize and s_wanted_extra_isize fields from the superblock. 2424Note: Resize2fs doesn't currently support the combination of flex_bg 2425and !resize_inode. (Addresses Debian Bug: #388452, #425477) 2426 2427Tune2fs can support migrating a filesystem from using 128 byte inodes 2428to 256 byte inodes, so it can take advantage of the full features of 2429ext4. 2430 2431Add support for "undo" support. E2fsck and mke2fs can optionally 2432record an undo log which can replayed by the program e2undo. 2433 2434E2fsck could damage a filesystem by trying to relocate inode tables 2435due to corrupted block group descriptors, where the attempted inode 2436table relocation would do far more harm than good. E2fsck has been 2437fixed to detect this these sorts of corrupted block group descriptors 2438much earlier in e2fsck processing, so it can try to use the backup 2439superblock and block group descriptors first. This should be a much 2440better strategy for recovering these types of corrupted filesystems. 2441(Addresses Sourceforge Bug: #1840291) 2442 2443E2fsck will display a more understandable message when the last check 2444field in the superblock is in the future. (Addresses Debian Bug: 2445#446005). 2446 2447E2fsck now performs more extensive and careful checks of extended 2448attributes stored in the inode. 2449 2450Enhance mke2fs to print a more explanatory error message when 2451ext2fs_get_device_size() returns EFBIG. (Addresses Debian Bug: 2452#488663) 2453 2454Fix mke2fs to use a default block size of 4k when formatting an 2455external journal device. This is done by using a fixed filesystem 2456type list that consists only of the single filesystem type "journal" 2457when looking up configuration keys in /etc/mke2fs.conf. (Addresses 2458Debian Bug: #488663) 2459 2460Speed up how mke2fs writes the journal data blocks by writing the disk 2461blocks in larger chunks. 2462 2463Fix blkid handling of stale devices. Fix a bug which could cause a 2464core dump while garbage collecting the blkid cache, and assure that 2465blkid_find_dev_with_tag() never returns a non-existent device. Also, 2466if a filesystem is found at a new /dev location, eliminate any 2467duplicate stale entries which can not be verified. 2468(Addresses Debian Bugs: #487758, #487783) 2469 2470Add more paranoid checks for LVM volumes and swap partitions in 2471blkid's probe function, to reduce the chances of false positives. 2472 2473The mke2fs program now has a much more sophisticated system for 2474controlling configuration parameters of a newly created filesystem 2475based on a split filesystem and usage type system. The -t option to 2476mke2fs was a deprecated alias to -c; it now specifies a filesystem 2477type (ext2, ext3, ext4, etc.), while the -T option can now be a comma 2478separated usage list. The filesystem type information and type 2479information is used to extract configuration parameters from the 2480/etc/mke2fs.conf file. 2481 2482The mke2fs program will no longer complain and request the -f option 2483when the user tries to create a filesystem with greater than 2**31 2484blocks. 2485 2486When creating a filesystem for the GNU Hurd use a fs-type of Hurd and 2487adjust the mke2fs.conf file so filesystems for the Hurd are created 2488with a blocksize of 4096 and inode size of 128, which is all it knows 2489how to handle. (Addresses Debian Bug: #471977) 2490 2491Mke2fs will always make sure that lost+found always has at least 2 2492blocks, even for filesystems with very large blocksizes (i.e., 64kb). 2493 2494Resize2fs will now print the minimum needed filesystem size if given 2495the -P option, and will resize the filesystem to the smallest possible 2496size if given the -M option. 2497 2498Fix resize2fs to clean up the resize_inode if all of the reserved gdt 2499blocks are consumed during an off-line resize. 2500 2501The "ls" command in debugfs now supports the -p option, which causes 2502it to quote the filenames so that spaces or tabs in directory entries 2503are easily visible. (Addresses Red Hat Bugzilla: #149480; Addresses 2504Sourceforge Feature Request: #1201667) 2505 2506Fix a potential off-by-one buffer oveflow in the fs_device_name in an 2507e2image file. 2508 2509The chattr program will return a non-zero exit code in case of 2510failures, and error messages can be suppressed with the -f option. 2511(Addresses Red Hat Bugzilla: #180596) 2512 2513Fix a bug in badblocks which caused it to overrun an array and likely 2514crash if more than 8 test patterns are specified using the -t option. 2515(Addresses Debian Bug: #487298) 2516 2517Add support to badblocks to limit how quickly it reads from the disk 2518drive (so it can be used for background scrubbing), and so it will 2519abort after finding a given number of errors. 2520 2521Remove support for the legacy big-endian filesystem format which only 2522existed on extremely long-dead PowerPC kernels almost a decade ago. 2523 2524Remove MASIX support from e2fsprogs. 2525 2526Add I/O statistics reporting to e2fsck. 2527 2528Update Vietnamese, Polish, French, Spanish, German, Catalan, Dutch, 2529Czech translations. 2530 2531Fixed various Debian packaging issues --- see debian/changelog for 2532details. (Addresses Debian Bugs: #487443, #487675, #490003) 2533 2534Fixed spelling mistakes, typos, and otherwise clarified man pages. 2535(Addresses Debian Bugs: #393313, #487849, #440983, #440981) 2536 2537Programmer's Notes 2538------------------ 2539 2540Factor out bitmap code in preparation for adding 64-bit new-style 2541bitmaps. 2542 2543Fix gcc -Wall warnings 2544 2545Fix the pkg-config files so that private librares are specified in 2546"Libs.private:". 2547 2548Fix the libext2fs.texinfo manual so it builds with modern versions of 2549texinfo. 2550 2551Silence the makefile from showing the awk command used to build the 2552 2553Clean up the badblocks group so to make it more portable and robust. 2554 2555Avoid using predictable filenames in /tmp in blkid's regression test 2556suite. Also remove bashism's in the regression test script. 2557 2558If the configure script is given --with-diet-libc, don't use thread 2559local storage, since diet libc doesn't support TLS. (Addresses 2560Sourceforge Bug: #2000654) 2561 2562Fix the blkid regression test suite to tolerate older versions of 2563mkswap that don't support the -U option. 2564 2565A few library routines have been converted to support 64-bit block 2566numbers; in particular, the I/O manager functions, the test_io, 2567inode_io, and unix_io managers have all be converted to support 64-bit 2568operation. 2569 2570Debugfs can now be extended for use by test programs. See 2571lib/ext2fs/extents.c for an example for how it can be used. The test 2572program links against the debugfs object files, and provides 2573additional commands by defining an auxiliary libss command table. 2574 2575The lazy_bg filesystem feature, which was only used by developer's 2576testing, has been removed since it has been largely supplanted by 2577uninit_bg. This also simplifies the code. 2578 2579 2580E2fsprogs 1.40.11 (June 17, 2008) 2581================================= 2582 2583Mke2fs, tune2fs, and resize2fs now use floating point to calculate the 2584percentage of reserved blocks. (Addresses Debian Bug: #452639) 2585 2586Updated Spanish and Catalan translations. 2587 2588Fixed various Debian packaging issues --- see debian/changelog for 2589details. (Addresses Debian Bugs: #483962, #483023) 2590 2591Add detection for ZFS volumes to the libblkid library. 2592 2593Fixed spelling mistakes, typos, and otherwise clarified man pages. 2594(Addresses Debian Bug: #486463) 2595 2596Programmer's notes: 2597------------------- 2598 2599Fix marginal C code in probe_lvm2() function to the blkid library more 2600portable for older compilers. 2601 2602Fix build problems on MacOS X. (Addresses Sourceforge Bug: #1972473) 2603 2604Fix ext2fs_swap{16,32,64} functions so they can be used by external 2605applications on big-endian machines. (Addresses Debian Bug: #484879) 2606 2607 2608E2fsprogs 1.40.10 (May 21, 2008) 2609================================ 2610 2611When deciding whether or not to revalidate a blkid cache entry, if the 2612device's mtime is newer than the last time the cached entry was 2613validated, force a revalidation. 2614 2615Fix a potential data corruption bug in e2fsck in the journal replay. 2616The chances of this is happening is extremely remote, especially the 2617default data=ordered or data=writeback modes. However, if a block 2618which has been journalled starts with the first four bytes 0xc03b3998, 2619when e2fsck replays the journal, those four bytes will be replaced 2620with zero's. Fortunately, it is highly, highly unlikely for e2fsck 2621metadata to begin with those fatal 4 byte sequence, and unless 2622data=ordered mode is in use, data blocks are never journaled. 2623 2624Updated German, Dutch, Sweedish, and Vietnamese translations. 2625 2626Programmer's notes: 2627------------------- 2628 2629Fixed various Debian packaging issues --- see debian/changelog for 2630details. 2631 2632Remove default sizes of types when cross compiling, since autoconf 26332.50 can figure this out automatically now. 2634 2635 2636E2fsprogs 1.40.9 (April 27, 2008) 2637================================= 2638 2639SuSE's security team audited uuidd and came up with a few minor 2640issues. None of them are serious given that uuidd runs setuid as a 2641unprivileged user which has no special access other than libuuid 2642directory, but it's good to get them fixed. 2643 2644One additional fix in ext2fs_swap_inode_full() needed for resize2fs to 2645work correctly with in-inode extended attributes. 2646 2647Updated German, Czech, Dutch, French, Polish, Sweedish, and Vietnamese 2648translations. 2649 2650Debugfs will avoid using a pager if the standard output is not a tty. 2651 2652Fix debugfs and tune2fs to correctly handle daylight savings time when 2653parsing a time string. 2654 2655Fixed spelling mistakes, typos, and otherwise clarified man pages. 2656 2657Fix fsck completion bars when multiple filesystems were being checked 2658in parallel. (Addresses Debian Bug: #432865, Addresses Launchpad Bug: 2659#203323, Addresses Sourceforge Bug: #1926023) 2660 2661Fix fsck so that progress information is sent back correctly when 2662multiple filesystems are being check and the output of fsck is being 2663redirected to a file descriptor. Also, include the device name (w/o 2664spaces) in the progress information sent back via a file descriptor. 2665(Addresses Launchpad Bug: #203323, Addresses Sourceforge Bug: 2666#1926023) 2667 2668Teach fsck to treat "ext4" and "ext4dev" as ext* filesystems. 2669 2670If logsave receives a SIGTERM or SIGINT signal, it will now pass that 2671signal to its child process. 2672 2673Fix mke2fs's creation of are resize inode when there is a non-standard 2674s_first_data_block setting. 2675 2676Fix bug in blkid when run by an unprivileged user; most devices were 2677not reported correctly. 9Addresses Launchpad Bug: #220275) 2678 2679Mke2fs will not allow the logically incorect combination of 2680resize_inode and meta_bg, which had previously caused mke2fs to create 2681a corrupt fileystem. 2682 2683Fix fsck in German locales so that a 'j' means yes. 2684(Addresses Sourceforge Bug: #1947683) 2685 2686 2687Programmer's notes: 2688------------------- 2689 2690Fixed various Debian packaging issues --- see debian/changelog for 2691details. 2692 2693Update valgrind options in test_script to work with valgrind 3.2.3 2694 2695Update texinfo.tex to a much newer version from the FSF. 2696 2697Remove bashism for configure script and from the lib/ss Makefile. 2698Addresses Sourceforge Bug: 1921969 2699 2700Fix some silently broken tests: m_no_opt, m_meta_bg, and m_raid_opt. 2701 2702Fix build system so that if texinfo is not installed, it won't print a 2703(harmless) error message. 2704 2705 2706E2fsprogs 1.40.8 (March 13, 2008) 2707================================= 2708 2709Fixed e2image -I so it works on image files which are larger than 2GB. 2710 2711Fixed e2fsck's handling of directory inodes with a corrupt size field. If 2712the size is larger than the number of blocks found in the inode, don't 2713try to allocate extra empty blocks at the end of the directory to make 2714up the difference; there's no point to doing that. In addition, if 2715the size is not a multiple of a blocksize, always fix it. 2716 2717E2fsck handled a pass 2 "should never happen error" by not giving 2718enough information and then core dumping. Unfortunately, it was all 2719too easy to trigger the "should never happen" situation if a 2720directory's inode size was not correct. This has been fixed, but 2721e2fsck has also been taught how to handle this situation more 2722gracefully, by simply removing the inode hash tree information, so 2723that it can be rebuilt again after e2fsck's pass 3. (Addresses 2724Launchpad Bug: #129395) 2725 2726Resize2fs had a bug resizing large inodes with extended attributes 2727that was fixed in 1.40.6; unfortunately, it turned out it wasn't fixed 2728completely on big-endian systems such as PowerPC. The bug should be 2729completely fixed now. Yay for regression test suites. (Addresses Red 2730Hat Bugzilla: #434893) 2731 2732Updated German, Czech, Dutch, Polish, Sweedish, and Vietnamese 2733translations. Many thanks to Philipp Thomas from Novell for stepping 2734up to become the new German translation maintainer! (Addresses Debian 2735Bugs: #302512, #370247, #401092, #412882). 2736 2737When e2fsck is clearing a corrupt inode's HTREE directory information, 2738make it clear that it is just clearing the HTREE information, not the 2739entire inode. 2740 2741Fixed spelling mistakes, typos, and otherwise clarified man pages. 2742 2743Programmer's notes 2744------------------ 2745 2746Add new functions, ext2fs_dblist_get_last() and 2747ext2fs_dblist_drop_last(), which allows the caller to examine the last 2748directory block entry added to the list, and to drop if it necessary. 2749 2750Fixed a portability problem in libblkid with DJGPP. 2751 2752Fix an obvious typo in an "internal error" message in e2fsck. Thanks 2753to Philipp Thomas for pointing this out. 2754 2755If the info files are not built, change "make install" so it doesn't 2756fail with an error code. 2757 2758 2759E2fsprogs 1.40.7 (February 28, 2008) 2760==================================== 2761 2762Remove support for clearing the SPARSE_SUPER feature from tune2fs, and 2763depreciate the -s option, since it can result in filesystems which 2764e2fsck can't fix easily. There are very good reasons for wanting to 2765disable sparse_super; users who wants to turn off sparse_super can use 2766debugfs. (Addresses Sourceforge Bug: #1840286) 2767 2768Add missing options to mke2fs's usage message. (Addresses Sourceforge 2769Bug: #1751393) 2770 2771Fix bug in resize2fs when large (greater than 128 byte) inodes are 2772moved when a filesystem is shrunk; it was only moving the first 128 2773bytes, so extended attributes were not getting moved. (Addresses Red 2774Hat Bugzilla: #434893) 2775 2776E2fsck now prints an explicit message when the bad block inode is 2777updated, to avoid confusion about why the filesystem was modified. 2778(Addresses Sourceforge Bug: #756460) 2779 2780Allow mke2fs and tune2fs manipulate the large_file feature. 2781Previously we just let the kernel and e2fsck do this automatically, 2782but e2fsck will no longer automatically clear the large_file feature. 2783It still isn't really necessary to worry about this feature flag 2784explicitly, but some users seem to care. (Addresses Red Hat Bugzilla: 2785#258381) 2786 2787Suppress message about an old-style fstab if the fstab file is empty. 2788(Addresses Debian Bug: #468176) 2789 2790Fix (really minor) bug in debugfs's find_free_block so it avoids 2791reporting a free block more than once if there are too few free blocks 2792in the filesystem. (Addresses Sourceforge Bug: #1096315) 2793 2794Change e2fsck to no longer clear the LARGE_FILES feature flag 2795automatically, when there are no more > 2GB files in the filesystem. 2796It's been almost a decade since there have been kernels that don't 2797support this flag, and e2fsck clears it quietly without telling the 2798user why the filesystem has been changed. 2799 2800Fix bug which could cause libblkid to seg fault if a device mapper 2801volume disappears while it is being probed. (Addresses RedHat 2802Bugzilla: #433857) 2803 2804Inhance e2fsck's reporting of unsupported filesystem feature flags. 2805(Addresses Sourceforge Feature Request: #1175808) 2806 2807Fix option syntax in dumpe2fs for explicit superblock and blocksize 2808parameters. What was currently documented in the man page has been 2809broken for some time, due to getopt() implementation changes. The 2810option sytax has been changed to one which is can be more portable 2811supported and which is consistent with the format for extended options 2812in mke2fs and tune2fs. (Addresses Sourceforge Bug: #1830994) 2813 2814Add support to tune2fs to clear the resize_inode feature. This 2815requires an fsck afterwards. (Addresses Red Hat Bugzilla: #167816) 2816 2817Teach blkid to detect LVM2 physical volumes. (Addresses Red Hat 2818Bugzilla: #409321) 2819 2820Add support for setting RAID stride and stripe-width via mke2fs and 2821tune2fs. Teach dumpe2fs to print the RAID parameters. 2822 2823Add support for setting new superblock fields to debugfs's 2824set_super_value. 2825 2826Add support for printing "mostly-printable" extended attributes in 2827Debugfs. 2828 2829Add support for the -M option to fsck, which causes it to ignore 2830mounted filesystem. 2831 2832Fix uuidd so that it creates the pid file with the correct pid number. 2833(Addresses Sourceforge Bug: #1893244) 2834 2835Fix various gcc -Wall warnings. 2836 2837Update Czech, Dutch, Polish, Sweedish, and Vietnamese translations 2838 2839Fixed spelling mistakes, typos, and otherwise clarified man pages. 2840(Addresses Sourceforge Patch: #1399325) 2841 2842 2843Programmer's notes: 2844------------------- 2845 2846Fixed various Debian packaging issues --- see debian/changelog for 2847details. (Addresses Debian Bug: #466929) 2848 2849Add new flag EXT2_FLAG_NONFREE_ON_ERROR ext2fs_open2() which returns a 2850partially filled-in filesystem object on an error, so that e2fsck can 2851print more intelligent error messages. 2852 2853Add a new function e2p_edit_feature2() which allows the caller to 2854specify which feature flags are OK to set or clear, and which returns 2855more specific information about feature flags which are not allowed to 2856be set/cleared. 2857 2858Set the C locale in the test_script driver since it uses [A-Za-z]. 2859(Addresses Sourceforge Bug: #1890526) 2860 2861Use fcntl locking instead of lockf in libuuid since Cygwin doesn't 2862support lockf(). 2863 2864Change configure.in to avoid using the 'dc' command unless it is 2865absolutely needed. (i.e., when using parsing a WIP-style version 2866number) (Addresses Sourceforge Bug: #1893024) 2867 2868Add portability checks to support compilation under DJGPP. 2869 2870Update to the latest samba tdb code before the LGPLv3 change, which 2871fixes a realloc() leak on failure. 2872 2873Fix memory leak in ext2fs_alloc_block(). 2874 2875Fix makefile dependency issues for various install targets. 2876(Addresses-Sourceforge-Patches: #1903484, #1903466, #1903456) 2877 2878Improve descriptions for the r_move_itable and r_resize_inode tests. 2879 2880 2881E2fsprogs 1.40.6 (February 9, 2008) 2882=================================== 2883 2884Add support for returning labels for UDF filesystems in the blkid 2885library. 2886 2887Fix bug in the blkid library where cached filesystems was not being 2888flushed when opening USB devices returned the error ENOMEDIUM. 2889(Addresses Debian Bug: #463787) 2890 2891Added logic to the blkid library to automatically choose whether a 2892filesystem should be mounted as ext4 or ext4dev, as appropriate. 2893 2894Allow tune2fs to set and clear the test_fs flag on ext4 filesystems. 2895 2896Fix a bug in e2fsck which caused it to core dump if e2fsprogs had been 2897configured with --enable-jbd-debug. 2898 2899Document the BLKID_FILE environment variable in the libblkid man page 2900 2901Programmer's Notes: 2902------------------- 2903 2904Update e2fsprogs translation template and Vietnamese and Czech translations 2905 2906Fixed various Debian packaging issues --- see debian/changelog for 2907details. (Addresses Debian Bugs: #436058) 2908 2909Don't try to create $DESTDIR/etc/init.d as part of make install as we 2910don't install the init.d script (and it's not the recommended way to 2911start uuidd anyway). (Addresses Sourceforge Bug: #1885085) 2912 2913Use thread local storage to fix a theoretical race condition if two 2914threads try to format an unknown error code at the same time in the 2915com_err library. 2916 2917 2918E2fsprogs 1.40.5 (January 27, 2008) 2919=================================== 2920 2921Fix a potential overflow big in e2image if the device name is too 2922long. 2923 2924Mke2fs will now create new filesystems with 256 byte inodes and the 2925ext_attr feature flag by default. This allows for much better future 2926compatibity with ext4 and speeds up extended attributes even on ext3 2927filesystems. 2928 2929Teach e2fsck to ignore certain "safe" filesystem features which are 2930set automatically by the kernel. Having the kernel set features 2931automagically behind the user's back is a bad idea, and we should try 2932to break the kernel of this habit, especially for the newer ext4 2933feature flags. But for now, we will try to avoid needless full checks 2934which can annoy users who are doing fresh installs. 2935 2936Add support in tune2fs and mke2fs for making a filesystem as being "ok 2937to be used with test kernel code". This will be needed for using test 2938filesystems with the latest ext4 kernel code. 2939 2940Change e2fsck -fD so that it sorts non-htree directories by inode 2941numbers instead of by name, since that optimizes performances much 2942more significantly. (Addresses-Sourceforge-Feature-Request: #532439) 2943 2944If e2image fills the disk, fix it so it exits right away instead of 2945spewing large numbers of error messages. 2946(Addresses-Sourceforge-Feature-Request: #606508) 2947 2948If ftruncate64() is not available for resize2fs, let it use ftrucate() 2949instead, but have it check to see if the size would get truncated, and 2950skip calling ftruncate in that case. 2951 2952Add support for detecting HFS+ filesystems in the blkid library. 2953 2954Add supprt in the blkid library for ext4/ext4dev filesystems. 2955 2956Fix a bug in blkid where it could die on a floating point exception 2957when presented with a corrupt reiserfs image. 2958 2959Fix blkid's handling of ntfs UUID's so that leading zeros are printed 2960such that UUID string is a fixed length. 2961 2962Add sample python bindings for the uuid library to the contrib 2963directory. (Addresses-Sourceforge-Patches: #778817) 2964 2965Fix debugfs's 'lsdel' command so it uses ext2fs_block_iterate2 so it 2966will work with large files. (Addresses Sourceforge Feature Request: 2967#1257500 and Sourceforge Support Request: #1253511) 2968 2969Allow the debugfs 'undel' command to undelete an inode without linking 2970it to a specific destination directory, since this might require 2971allocating blocks that could overwrite some yet-to-be-recovered 2972deleted files. (Addresses-Sourceforge-Feature-Request: #967141) 2973 2974Update Swedish translations from the Translation Project. 2975 2976Programmer's Notes: 2977------------------- 2978 2979Fix configure handling of --sbindir (which should rarely be used, but 2980someone did complain, so let's fix it). (Addresses Sourceforge Bug: 2981#498381) 2982 2983Updated e2fsprogs.spec file to include a new uuidd package 2984 2985Use pkg-config to determine where to find the devmapper library so we 2986can find out where it is located on different distributions. 2987 2988Fix Makefile race so that "make -j3 distclean" works correctly 2989 2990Fix portability problems on non-Linux/non-Hurd/non-Masix systems, 2991especially on MacOS X systems. (Addresses Sourceforge Bugs: #1861633, 2992#1819034, #1863819) 2993 2994Fixed spelling mistakes, typos, and otherwise clarified man pages. 2995 2996Fixed various Debian packaging issues --- see debian/changelog for 2997details. (Addresses Debian Bugs: #459403, #459475, #459614) 2998 2999Remove the --enable-dynamic-static configure option, and build e2fsck 3000dynamically by default. If the user wants e2fsck.static, he/she will 3001need to build it via "make -C e2fsck e2fsck.static" 3002 3003Fix various build warnings due to missing prototypes. 3004(Addresses Sourceforge Patch: #1861663, #1861659) 3005 3006 3007E2fsprogs 1.40.4 (December 31, 2007) 3008==================================== 3009 3010Improve time-based UUID generation. A new daemon uuidd, is started 3011automatically by libuuid if necessary. This daemon is setuid to allow 3012updates to /var/lib/libuuid, so the clock sequence number can be 3013stored and so if the clock is set backwards, it can be detected. 3014(Addresses Sourceforge Bug: #1529672, Addresses Red Hat Bugzilla: 3015#233471) 3016 3017Filter out the NEEDS_RECOVERY feature flag when writing out the backup 3018superblocks. This avoids e2fsck from concluding that a full 3019filesystem check is required before backing up the superblock due to 3020changes in the feature flags. (Addresses Debian Bug: #454926) 3021 3022Fix fsck to only treat the '#' character as a comment at the beginning 3023of the line in /etc/fstab. Otherwise fstabs for the fuse filesystem 3024will cause fsck to issue an bogus warning message. 3025(Addresses Gentoo bug: #195405, Addresses Sourceforge bug: #1826147) 3026 3027Format control characters and characters with the high eighth bit set 3028when printing the contents of the blkid cache, to prevent filesystems 3029with garbage labels from sending escape sequences to the user's screen 3030that might, for example place it in graphics mode. (Addresses Ubuntu 3031Bug: #78087) 3032 3033Fix sign-extension problem on 64-bit systems in in the com_err 3034library. (Addresses Sourceforge Bug: #1809658) 3035 3036Avoid division by zero error when probing an invalid FAT filesystem in 3037the blkid library. (Addresses Sourceforge Bug: #1831627) 3038 3039Update Dutch, Polish, and Vietnamese translations from the Translation 3040Project. Remove the Rwandan translation upon advice of the 3041Translation Project. 3042 3043Programmer's Notes: 3044------------------- 3045 3046Fix the libss "make check" regression test so that it works if the 3047current directory is not in the user's path or if the libss shared 3048library is not installed. (Addresses Sourceforge Bug: #1848974) 3049 3050Fixed spelling mistakes, typos, and otherwise clarified man pages. 3051(Addresses Debian Bugs: #444883, #441872) 3052 3053Fixed various Debian packaging issues --- see debian/changelog for 3054details. (Addresses Debian Bugs: #437720, #451172, #458017) 3055 3056Fix build failure on non-Linux/non-Hurd/non-Masix systems. 3057(Addresses Sourceforge Bug: #1859778) 3058 3059Fix Hurd portability issues. (Addresses Debian Bug: #437720) 3060 3061 3062E2fsprogs 1.40.3 (December 5, 2007) 3063=================================== 3064 3065Fix a potential security vulnerability where an untrusted filesystem 3066can be corrupted in such a way that a program using libext2fs will 3067allocate a buffer which is far too small. This can lead to either a 3068crash or potentially a heap-based buffer overflow crash. No known 3069exploits exist, but main concern is where an untrusted user who 3070possesses privileged access in a guest Xen environment could corrupt a 3071filesystem which is then accessed by the pygrub program, running as 3072root in the dom0 host environment, thus allowing the untrusted user to 3073gain privileged access in the host OS. Thanks to the McAfee AVERT 3074Research group for reporting this issue. (Addresses CVE-2007-5497.) 3075 3076Fix hueristics in blkid which could cause a disk without partitions to 3077be incorrectly skipped when a loopback device is present. (Addresses 3078Red Hat Bugzilla #400321.) 3079 3080Fix e2image so that in raw mode it does not create an image file which 3081is one byte too large. 3082 3083Change mke2fs's usage message so it recommends the preferred -E option 3084instead of the deprecated -R option. 3085 3086Enhance the blkid library so it will recognize squashfs filesystems. 3087(Addresses Red Hat Bugzilla #305151.) 3088 3089Enhance e2fsck so it will force the backup superblocks to be backed up 3090if the filesystem is consistent and key constants have been changed 3091(i.e., by an on-line resize) or by e2fsck in the course of its 3092operations. 3093 3094Enhance blkid's detection of FAT filesystems; so that USB disks with 3095only a single bootable partition will not get missed. 3096 3097E2fsck will no longer mark a filesystem as invalid if it has time 3098errors (i.e., if superblock mount time or last write time is in the 3099future) and the user refuses to fix the problem. 3100 3101The Ubuntu init scripts don't properly set the system time correctly 3102from hardware clock if the hardware clock is configured to tick local 3103time instead of GMT time. Work around this as best as we can by 3104providing an option, buggy_init_scripts, in /etc/e2fsck.conf which can 3105be set on Ubuntu systems. (Addresses Debian Bug #441093, and Ubuntu 3106Bug #131201.) 3107 3108Fix fsck to ignore /etc/fstab entries for bind mounts. (Addresses Red 3109Hat Bugzilla #151533.) 3110 3111Fix e2fsck so that if the superblock is corrupt, but still looks 3112vaguely like an ext2/3/4 superblock, that it automatically tries to 3113fall back to the backup superblock, instead of failing with a hard 3114error. 3115 3116Make the e2fsprogs program more robust so that they will not crash 3117when opening a corrupt filesystem where s_inode_size is zero. 3118 3119Change e2fsck so it uses sscanf() instead of atoi() so it non-numeric 3120arguments are detected as such and the parse error is reported to the 3121user. (Addresses Debian Bug #435381.) 3122 3123Change e2fsck so it will not complain if a file has blocks reallocated 3124up to the next multiple of a system's page size. 3125 3126Fix bug in ext2fs_check_desc() which will cause e2fsck to complain 3127about (valid) filesystems where the inode table extends to the last 3128block of the block group. (Addresses Red Hat Bugzilla #214765.) 3129 3130Fix a bug in ext2fs_initialize() which causes mke2fs to fail while 3131allocating inode tables for some relatively rare odd disk sizes. 3132(Addresses Red Hat Bugzilla #241767.) 3133 3134Add Catalan translation and update Dutch and Swedish translations 3135from the Translation Project. 3136 3137Fix big-endian byte-swapping bug in ext2fs_swap_inode_full(). We 3138still had an issue when trying to figure out whether we need to 3139byte-swap fast symlinks that contained extended attributes. 3140 3141Fixed spelling mistakes, typos, and otherwise clarified man pages. 3142(Addresses SourceForge Bug #1821333.) 3143 3144 3145Programmer's Notes: 3146------------------- 3147 3148Fix mke2fs tests to avoid needing any significant ^M (CR) characters 3149 3150Add "make check" to the RPM spec file 3151 3152Fix "make install" and 'make unstall" in misc/Makefile.in so that it 3153works correctly when the prefix is not the root directory. 3154 3155Fix the resize2fs tests, r_move_itable and r_resize_inode, so they 3156clena up after themselves by deleting the test.img temporary file 3157after completing the test. 3158 3159Fixed a corner case bug ext2fs_unlink() when trying to delete the 3160first directory entry in a directory block and the last directory 3161entry in the previous directory block is not in use. Fortunately 3162ext2fs_unlink() is only used by debugfs and e2fsck, and in e2fsck in 3163places where it is extremely unlikely to run into this corner case. 3164 3165Fix missing dependency which would cuase parallel builds to fail. 3166(Addresses Sourceforge Bug #1842331.) 3167 3168Fix a build error on newer gcc caused by lib/ext2fs/ismounted.c 3169calling open(O_CREATE) with a missing mode parameter. 3170 3171Fix the test_ss.c in lib/ss so it can be used as an example 3172application program for the library as well as a regression test 3173suite. 3174 3175Fix ext2fs_dblist_dir_iterate() so that error codes and abort codes 3176are properly passed back up through the call stack. 3177 3178E2fsprogs 1.40.2 (July 12, 2007) 3179================================ 3180 3181A recent change to e2fsck_add_dir_info() to use tdb files to check 3182filesystems with a very large number of filesystems had a typo which 3183caused us to resize the wrong data structure. This would cause a 3184array overrun leading to malloc pointer corruptions and segfaults. 3185Since we normally can very accurately predict how big the the dirinfo 3186array needs to be, this bug only got triggered on very badly corrupted 3187filesystems. 3188 3189Fix a bug in e2fsck which caused it to incorrectly salvange 3190directories when the last entry's rec_len is bogusly too big. This 3191resulted in a nonsense filesystem corruption to be reported, and 3192required a second run of e2fsck to fully fix up the directory. 3193 3194Update tune2fs man page to include more discussion of reserved blocks 3195(Addresses Launchpad bug #47817) 3196 3197Update Turkish, Polish, Dutch, and Vietnamese PO files from the 3198Translation Project 3199 3200 3201E2fsprogs 1.40.1 (July 7, 2007) 3202=============================== 3203 3204Fix bug which could cause libblkid to loop forever. When revalidating 3205a partition where there is obsolete information in /etc/blkid.tab, we 3206end up freeing a the type tag without clearing dev->bid_type, causing 3207blkid_verify() to loop forever. (Addresses Debian Bug: #432052) 3208 3209The Turkish translation has a bug in it where it has the translation 3210of "E@e '%Dn' in %p (%i)" to "E@E". This causes @E to be expanded at 3211@E, recursively, forever, until the stack fills up and e2fsck core 3212dumps. We fix this by making e2fsck stop @-expansions after a 3213recursive depth of 10, which is far more than we need. 3214(Addresses Sourceforge Bug: #1646081) 3215 3216Compile the default mke2fs.conf into mke2fs program. People are 3217getting surprised by mke2fs creating filesystems with different 3218defaults than earlier versions of mke2fs if mke2fs.conf is not 3219present. So we now create a built in version of mke2fs.conf file 3220which is used by mke2fs if the /etc/mke2fs.conf is not present. 3221(Addresses SourceforgeBug: #1745818) 3222 3223Improve the config/parse_types.sh helper script. Fix a potential 3224security problem if e2fsprogs is built as root (as Gentoo does!). In 3225addition fix the script and how it is called from the configure script 3226so that it does the right thing when cross-compiling. (Fixes Gentoo 3227bug: #146903) 3228 3229Update Vietnamese, French, and Dutch PO files from the Translation 3230Project. Also created a new e2fsprogs.pot file for translator. 3231 3232Fix bogus strip permission errors when building under Debian. When 3233building the e2fsprogs dpkg's, the dh_strip command emits a large 3234number of error messages caused by the permissions not being right. 3235So run dh_fixperms before running dh_strip. 3236 3237Programmer's Notes: 3238------------------- 3239 3240Add new function: profile_set_default(). This function sets the value 3241of the pseudo file "<default>". If the file "<default>" had 3242previously been passed to profile_init(), then def_string parameter 3243will be parsed and used as the profile information for the "<default>" 3244file. 3245 3246Fix mk_cmds's error reporting so that it is unambiguous that it is the 3247mk_cmds script which is generating the error. (Obviates Gentoo patch: 3248e2fsprogs-1.32-mk_cmds-cosmetic.patch) 3249 3250Fix the test suite to use LC_ALL instead of LANG. LC_ALL is the "high 3251priority" environment variable that overrides all others, where as 3252LANG is the lowest priorty environment variable. If LC_ALL is set, it 3253doesn't matter whether LANG, LANGUAGE, LC_COLLATE, LC_MESSAGES, and 3254the all the rest are set. This will assure that the locale when 3255running the test suites is the "C" locale. (Obviates Gentoo patch: 3256e2fsprogs-1.38-tests-locale.patch) 3257 3258 3259E2fsprogs 1.40 (June 29, 2007) 3260============================== 3261 3262Fix divide by zero error in blkid's NTFS probing logic. 3263 3264Add new blkid -g option which causes the blkid cache to be garbage 3265collected. 3266 3267Fix a bug in libblkid which could cause the internal field bid_type to 3268become corrupted. Fortunately bid_type isn't used much, and bid_label 3269and bid_uuid is only used by debugging code, so the impact of this bug 3270was very minor. 3271 3272Mke2fs will now store the RAID stride value when a filesystem is 3273created with a requested RAID stride, and then use it automatically in 3274resize2fs. 3275 3276Mke2fs has a sanity check added to make sure (inode_size * num_inodes) 3277isn't too big. In some cases Lustre users have tried specifying an 3278inode size of 4096 bytes, while keeping an inode ratio of one inode 3279per 4096 bytes. 3280 3281Improve sanity check in e2fsck's algorithm for finding a backup 3282superblock, so that it won't accidentally find a superblock that was 3283located in the journal, and then later reject it as being not a valid 3284backup superblock. 3285 3286Fix e2fsck get_size logic so that it will work with the Linux floppy 3287driver. The Linux floppy driver is a bit different from the other 3288block device drivers, in that if the device has been opened with 3289O_EXCL, it disallows another open(), even if the second open() does 3290not have the O_EXCL flag. (Addresses Debian Bug: #410569) 3291 3292Fix error checking of badblock's last-block and start-block arguments. 3293(Addresses Debian Bug: #416477) 3294 3295Fix e2fsck so that it doesn't overwrite the backup superblocks when 3296recovering a journal until the master superblock has been confirmed as 3297being sane. 3298 3299Change the blkid library to be much more paranoid about concluding 3300that a partition contains an NTFS filesystem, and fetch the UUID and 3301LABEL information from NTFS filesystems. (Addresses Launchpad Bug: 3302#110138) 3303 3304Factor out the code which sets the default journal size and move it 3305into libext2fs. 3306 3307Enhance e2fsck so it will recreate the ext3 journal if the original 3308journal inode was cleared to due it being corrupt after finishing the 3309filesystem check. 3310 3311Fix e2fsck so that it updates the journal inode if it is corrupted and 3312the backup journal information from the superblock was successfully 3313used to recover the filesystem. 3314 3315Fix e2fsck so that it checks all of the blocks in the journal inode 3316for validity. The original code only checked the direct blocks to 3317make sure the journal inode was sane. Unfortunately, if some or all 3318of the indirect or doubly indirect blocks were corrupted, this would 3319not be caught. 3320 3321Add support in blkid to detect LUKS encrypted partitions. 3322 3323Add extra sanity checks for extended attributes in the case where the 3324size is zero but the offset is very large. 3325 3326Fix byte-swapping issues for large inodes in ext2fs_read_inode_full() 3327and ext2fs_get_next_inode_full(). 3328 3329Clarify the copyright licenses used by the various libraries in 3330the top-level COPYING file (Red Hat Bugzilla: 166058) 3331 3332Make mke2fs's defaults when /etc/mke2fs.conf doesn't exist more sane. 3333 3334Fix mke2fs and debugfs to support large (> 16 bit) uid's and gid's. 3335 3336Remove check in e2fsck which requires EA's in inodes to be sorted; 3337they don't need to be sorted, and e2fsck was previously wrongly 3338clearing unsorted EA's stored in the inode structure. 3339 3340Allow mke2fs or tune2fs to create a substantially larger journal (up 3341to 10,240,000 blocks). 3342 3343Fix MD superblock detection, and make sure the correct UUID is 3344reported from the MD superblock. 3345 3346Fix a signed vs. unsigned bug in debugfs. 3347 3348Enhance debugfs's date parser so that it accepts integer values. 3349 3350Fix e2fsck's pass1c accounting so it doesn't terminate too early if a 3351file with multiply claimed blocks is hard linked. or not at all if the 3352root directory contains shared blocks 3353 3354Enhance debugfs so it can modify the block group descriptors using the 3355command set_block_group_descriptor. 3356 3357Improve e2fsck's reporting of I/O errors so it's clearer what it was 3358trying to do when an error happens 3359 3360Fix a bug in in how e2fsprogs byte swaps inodes containing fast 3361symlinks that have extended attributes. (Addresses Red Hat Bugzilla: 3362#232663 and LTC Bugzilla: #27634) 3363 3364Fix potential file descriptor leak in ext2fs_get_device_size() in an 3365error case. 3366 3367Add libreadline.so.5 support to libss. 3368 3369Impove badblocks -n/-w exclusive usage message. 3370 3371Fix dump_unused segfault in debugfs when a filesystem is not open 3372 3373Fix memory leak in blkid library. (Addresses Debian Bug: #413661) 3374 3375Allow the debugfs lcd command to work w/o a filesystem being open. 3376(Addresses LTC Bugzilla #27513) 3377 3378Fix e2fsck to clear i_size for special devices with a bogus i_blocks 3379field on the first pass. 3380 3381Fix e2fsck to set the file type of the '..' entry when connecting 3382a directory to lost+found. (Addresses Lustre Bug: #11645) 3383 3384Enhance e2fsck to recover directories whose modes field were corrupted 3385to look like special files. This is probably only useful in 3386artificial test cases, but it will be useful if we ever do the "inodes 3387in directory" idea for ext4. 3388 3389Allow debugfs to dump (and rdump) > 2GB files. (Addresses Debian Bug: 3390#412614) 3391 3392Fix resize2fs parsing of size parameter (in sector units). This was 3393actually a bug in libe2p's parse_num_blocks() function. When handling 3394the 's' suffix, it was ignoring the blocksize information passed in 3395from the caller and always interpreting the number in terms of a 1k 3396blocksize. (Addresses Debian Bug: #408298) 3397 3398There was a floating point precision error which could cause e2fsck to 3399loop forever on really big filesystems with a large inode count. 3400(Addresses Debian Bug: #411838) 3401 3402Fix memory leak in ext2fs_write_new_inode() 3403 3404Add support for using a scratch files directory to reduce e2fsck's 3405memory utilization on really big filesystems. This uses the TDB 3406library. See the [scratch_files] section of the e2fsck.conf man page 3407for more details. 3408 3409Fixed type-punning bug which caused dumpe2fs to crash on the Arm 3410platform (Addresses Debian Bug: #397044) 3411 3412Add explanatory message to badblocks that -n and -w are mutually exclusive 3413(Addresses Debian Bug: #371869) 3414 3415Allow debugfs and dumpe2fs to support fs features under development. 3416 3417Add support for the new flag EXT2_FLAG_SOFTSUPP_FEATURES flag to 3418ext2fs_open() , which allows application to open filesystes with features 3419which are currently only partially supported by e2fsprogs. 3420 3421Allow unix_io to support offsets greater than 2G (Addresses 3422SourceForge Bug: #1547922) 3423 3424Fixed overflow and signed/unsigned problems caused by the number of 3425blocks or inodes exceeding 2**31 or being close to 2**32-1. 3426 3427Add support for unsigned directory hash calculations with hints in the 3428superblock to fix cross-architectural portability for htree 3429directories with filenames where the high 8th bit is set. (Addresses 3430Debian: #389772) 3431 3432Fix resize2fs so that it gives user-intelligible error messages if the 3433filesystem or the kernel does not support on-line resizing. 3434(Addresses Debian Bug: #380548) 3435 3436Require mke2fs -F -F for really dangerous operations, since -F is 3437needed for less dangerous operations such as creating filesystems 3438images in regular files, or creating filesystems on whole block 3439devices. These relatively innocuous usages should NOT be confused 3440with running mke2fs on an apparently-mounted or in-use filesystem. 3441 3442Allow the default inode size to be specified into the mke2fs.conf 3443file. 3444 3445Make the smallest default journal size is big enough so that on-line 3446resizing should always work. 3447 3448Fix silly spelling error in e2fsck. (Addresses SourceForge bug: 3449#1531372) 3450 3451Fix debugfs coredump when lsdel is run without an open filesystem 3452(Addresses Debian Bug: #378335) 3453 3454Fix debugfs display bug us that bytes that have the high bit set are 3455displayed as "ec" instead of "ffffffec". 3456 3457Add support in lsattr so it will display the EXT4_EXTENTS_FL flag. 3458 3459Device mapper scanning wasn't working in the blkid library because the 3460pathnames had an extra "/dev" when they were being probed. 3461 3462Add GFS/GFS2 support to the blkid library. 3463 3464Fix blkid support of empty FAT filesystem labels. 3465 3466Avoid recursing forever (or for a long time) when the blkid library 3467searches for a device and there are symlinks to directories in /dev. 3468 3469Avoid unaligned halfword access in blkid when accessing FAT 3470superblocks, as this will cause Sparc/Solaris systems to throw a 3471SIGBUS error. 3472 3473The latest devmapper libraries requires pthreads, add -lpthreads to 3474the static link libraries for e2fsck.static if devmapper is enabled. 3475(Addresses Debian bug: #388718) 3476 3477Improve the (non-installed, for experts only) findsuper program by 3478printing the uuid and label from the superblocks, as well as the 3479starting and ending offsets of the filesystem given the information in 3480the superblock. Omit by default printing superblocks that are likely 3481found in located in an ext3 journal unless an explicit -j option is 3482given. 3483 3484Updated Spanish, French and Dutch translations and added Catalan 3485translation. (Addresses Debian bug: #411562) 3486 3487Use FreeBSD's DIOCGMEDIASIZE and DIOCGDINFO ioctls if available when 3488determining a partition's size, since binary searching to determine 3489the device doesn't work on FreeBSD. 3490 3491Documentation about UUID's is available in enough places, and it's 3492awkward to deal with debian-legal's insanities. So I'm caving in the 3493"more-lunatic-than-RMS" wing of Debian by removing RFC-4122 so we 3494don't have do the dfsg tarball. Also remove the rule that only tried 3495to install RFC-4122 on Ubuntu, since Ubuntu seems to want to fetch 3496e2fsprogs exclusively from Debian. (Addresses Debian Bug: #407107) 3497 3498Fix the info-dir line so that the menu name does not contain a .info 3499prefix. First of all, it's ugly, secondly, it causes the install-info 3500command to fail to remove the com_err info file from the 3501/usr/share/info/dir file when the comerr-dev package is removed and 3502purged. (Addresses Debian Bug: #401711) 3503 3504Fixed spelling mistakes, typos, and otherwise clarified man pages. 3505(Addresses Debian Bug: #369761, #373004, #379695) 3506 3507Fixed various Debian packaging issues --- see debian/changelog for 3508details. (Addresses Debian Bugs #389554, #390664, #413208, #419605, 3509#408352, #415560, #399155) 3510 3511 3512Programmer's notes: 3513------------------- 3514 3515E2fsck now supports the %It expansion when printing a problem report. 3516It will print the type of the inode in the problem context. 3517 3518Fix misc/Makefile.in so that it builds even if e2fsck hasn't been built yet 3519(Addresses Sourceforge Bug: #1565561) 3520 3521Remove unused variables and other lint/gcc -Wall cleanups 3522 3523Add check to ext2fs_get_device_size() so it will return EFBIG for for 3524filesystems contained in regular files where the filesystem image size 3525is returned by stat64(). 3526 3527Set local environment variables to C so mk_cmds and compile_et always 3528work. (Addresses SourceForge Bug: #1532177) 3529 3530Added the 64-bit byte swapping function ext2fs_swab64(). 3531 3532Added two new helper functions to prevent 2**31/2**32-1 overflow 3533problems: ext2fs_div_ceil() and e2p_percent(). 3534 3535Create new ext2fs library inline functions ext2fs_group_first_block() 3536and ext2fs_group_last_block() in order to calculate the starting and 3537ending blocks in a block group. 3538 3539Create the generated files read-only to remind developers not to edit them. 3540 3541Add support for autoconf 2.60 (with backwards compatibility for older 3542versions of autoconf). 3543 3544Added an "make rpm" target to top-level Makefile 3545 3546Added various FreeBSD portability fixes. 3547 3548Exclude mercurial files from the RPM build tree to speed up copy/build. 3549 3550Use root_sysconfdir to define the locations of mke2fs.conf and 3551e2fsck.conf instead of using a hard-coded /etc pathname. 3552 3553Prevent e2fsck.h and ext2_ext_attr.h from getting included multiple times. 3554 3555Fixed "make clean" in blkid's Makefile.in file from removing tst_*.c files. 3556 3557If diff -u is supported, use it to report test failures. 3558 3559Updates/improvements to RPM spec file 3560 3561Add on-disk format definitions for the following new features: 3562EXT4_FEATURE_RO_COMPAT_HUGE_FILE, EXT4_FEATURE_RO_COMPAT_GDT_CSUM, 3563EXT4_FEATURE_RO_COMPAT_DIR_NLINK, EXT4_FEATURE_INCOMPAT_64BIT, 3564EXT4_FEATURE_RO_COMPAT_EXTRA_ISIZE 3565 3566Add a new make target "checked-failed" in the tests directory which 3567reruns any failed tests 3568 3569Update draft-leach-uuids-guids-01.txt with rfc4122.txt 3570 3571Fix miscellaneous bugs reported by Coverity: Dead code, potential nul 3572pointer dereferences, memory leaks, etc. None were security-criticial 3573problems. 3574 3575Fix up usage and decrement error messages in the test_icount program 3576 3577Add debugging code to the com_err library; if the environment variable 3578COMERR_DEBUG is set to 1, print out debugging messages as error tables 3579are added and removed from the com_err library. If the 3580COMERR_DEBUG_FILE environment variable is set (and the process is not 3581setuid) the debugging messages may be redirected to a file. 3582 3583Change all of the e2fsprogs programs to use the newer add_error_table() 3584and remove_error_table() interfaces instead of the much older 3585initialize_*_error_table() function. 3586 3587Add TDB support into the ext2fs library. This allows us to have a 3588guaranteed library we can count on always being present so we can 3589store data in an on-disk database. 3590 3591Add support for using TDB to store the icount data, so we don't run out 3592of memory when checking really large filesystems. 3593 3594Change the regression test suite so that it skips empty test directories. 3595 3596Define the l_i_iversion field in ext2_inode. The l_i_version field is 3597now defined from the old l_i_reserved1 field in the ext2 inode. This 3598field will be used to store high 32 bits of the 64-bit inode version 3599number. 3600 3601Add Makefile production rule for e2fsprogs.spec in case it gets 3602deleted. 3603 3604Add new function profile_get_uint() to allow for a clean way to fetch 3605unsigned integers from the context. 3606 3607Add test to make sure the ext2 superblock structure is 1024 bytes. 3608 3609Fix typo in name of f_dup4 regression test 3610 3611Add new function blkid_gc_cache() which performs a garbage collection 3612pass on the /etc/blkid.tab file. 3613 3614The ext2fs_open() function now sets EXT2_FLAG_MASTER_SB_ONLY. In 3615general, only e2fsck (after the filesystem is clean), tune2fs, and 3616resize2fs should change the backup superblocks by default. Most 3617callers of ext2fs_open() should not be touching any superblock fields 3618which require the backups to be touched. 3619 3620Add new function to libext2fs, ext2fs_default_journal_size(), which 3621returns the default journal size. 3622 3623 3624E2fsprogs 1.39 (May 29, 2006) 3625============================= 3626 3627Fix 32-bit cleanliness in e2fsprogs so that we can support filesystems 3628between 2**31 and 2**32 blocks. 3629 3630Change mke2fs to use /etc/mke2fs.conf as a configuration file to 3631configure the filesystem features, blocksize, and inode_ratio for 3632different filesystem types. 3633 3634Mke2fs will now create filesystems hash trees and on-line resizing 3635enabled by default, based on the new /etc/mke2fs.conf file. 3636 3637The e2fsprogs tools (resize2fs, e2fsck, mke2fs) will open the 3638filesystem device node in exclusive mode to prevent accidents by 3639system administrators. In the case of resize2fs and mke2fs, it will 3640only use exclusive mode if the filesystem is not mounted. 3641 3642Fixed a bug in mke2fs which caused it to to fail when creating the 3643resize inode for large filesystems. (Addresses Debian Bug #346580) 3644 3645When allocating space for the RAID filesystems with the stride 3646parameter, mke2fs will now place each portion of the group's inode 3647table right up after the superblock (if present) in order to minimize 3648fragmentation of the freespace. 3649 3650Speed up mke2fs and e2fsck by writing inode and block bitmaps more 3651efficiently by writing the inode and block bitmaps in one pass, thus 3652reducing the number of disk seeks required. 3653 3654Add support for on-line resizing to resize2fs. 3655 3656Fix blkid library so that logic to determine whether or not a device's 3657cached information in /etc/blkid.tab needs to be verified or not 3658doesn't get confused by a system clock which is insane (for example, 3659if the battery is dead on a Macintosh running PPC Linux. (Addresses 3660Red Hat Bug: #182188) 3661 3662The blkid library will now store the UUID of the external journal used 3663by ext3 filesystems, so that in the future, the userspace mount binary 3664can use this to find the location of the external journal and pass 3665this information to the kernel. 3666 3667E2fsck will now consult a configuration file, /etc/e2fsck.conf to 3668control how various options should be handled. See the e2fsck.conf 3669man page for more details. (Addresses Debian Bug: #150295) 3670 3671E2fsck now prints an explanatory message when delaying a filesystem 3672check when the system is running on battery. (Addresses Debian Bug: 3673#350306) 3674 3675E2fsck will detect if the superblock's last mount field or last write 3676field is in the future, and offer to fix if so. (Addresses Debian Bug 3677#327580) These problems will be fixed automatically in preen mode 3678since Debian's boot sequence bogusly doesn't set the time correctly 3679until potentially very late in the bootup process, and this can cause 3680false positives which will cause users' systems to fail to boot. 3681(Addresses Debian Bugs #343662 and #343645) 3682 3683E2fsck now checks to see if the superblock hint for the location of 3684the external journal is incorrect, and if so, offer to update it. 3685(Addresses Debian Bug: #355644) 3686 3687Fix e2fsck from segfaulting on disconnected inodes that contain one or 3688more extended attributes. (Addresses Debian Bug: #316736, #318463) 3689 3690E2fsck will stop and print a warning if the user tries running a 3691read/write badblocks test on a read-only mounted root filesystem. 3692 3693Fix a memory leak in e2fsck's error paths. (Thanks to Michael 3694C. Thompson for pointing these out; they were originally found using 3695Coverity.) 3696 3697When resizing a file containing a filesystem, resize2fs will expand or 3698truncate a file as necessary. (Addresses Debian Bug: #271607) 3699 3700Resize2fs will now automatically determine the RAID stride parameter that 3701had been used to create the filesystem, and use that for newly created 3702block groups. The RAID stride parameter may also be manually specified 3703on the command line using the new -S option to resize2fs. 3704 3705Fix mke2fs so that it correctly creates external journals on 3706big-endian machines (such as a S/390). 3707 3708Fix a bug in the e2p library which could cause dumpe2fs to (rarely) 3709fail to print out the journal or hash seed UUID. (Thanks to Guillaume 3710Chambraud for pointing this out.) 3711 3712Dumpe2fs will now print the size of the journal (if present). 3713 3714Fix debugfs's set_inode_field command so it can properly set the frag, 3715fsize, uid_high, gid_high, and author fields in the inode instead of 3716silently failing, and so that setting the i_size actually sets i_size 3717correctly. 3718 3719Add a new debugfs command, set_current_time, which sets fs->now so 3720that regression test suites can repeatedly modify the filesystem's 3721last_write fields. 3722 3723Fix a bug in debugfs's icheck which would incorrectly report the owner 3724of an extended attribute block. 3725 3726Fix the debugfs commands htree_dump, dx_hash, and list_dir so they print a 3727print a usage message when an illegal option character is given. 3728 3729Fix debugfs's dump_unsued command on filesystems with a 64k blocksize 3730so it won't core dump. (Addresses SourceForge bug #1424311) 3731 3732Fix mklost+found so that it creates a full-sized directory on 3733filesystems with larger block sizes. 3734 3735Fix a file descriptor leak in blkid library. 3736 3737Fix a display bug in "badblocks -sv" so that the done message properly 3738clears the block number at the end of the test. (Addresses Debian Bug 3739#322231) 3740 3741Allow fractional percentages to the -m option in mke2fs and tune2fs 3742(Addresses Debian Bug: #80205) 3743 3744Use fstat/fstat64 in getsize.c if the the target is a regular file, 3745instead of attempting to do a binary search. Fix some fd leaks in 3746error cases. 3747 3748Add support for device mapper library to the blkid library to ensure 3749that the "best" (i.e., leaf) device is probed by the blkid library. 3750 3751Fix the blkid library so that it notices when an ext2 filesystem is 3752upgraded to ext3. 3753 3754Improve the blkid's library VFAT/FAT detection; it now understands 3755labels stored in the root directory, and is more paranoid about 3756checking the FAT superblock values. 3757 3758Fixed a fd leak in the uuid library which was causing problems for the 3759LVM tools. (Addresses Debian Bug: #345832) 3760 3761Add support for the reiser4 and software suspend partitions to the 3762blkid library. Also add support for extract the label from iso9660 3763filesystems. 3764 3765Fix a compile_et bug which miscount the number of error messages if 3766continuations are used in the .et file. 3767 3768Add extra sanity checks to protect users from unusual cirucmstances 3769where /etc/mtab may not be sane, by checking to see if the device is 3770reported busy (works on Linux 2.6) kernels. (Addresses Debian Bug 3771#319002) 3772 3773Updated French, Dutch, Polish, and Swedish translations. (Addresses Debian 3774Bug: #343149, #341911, #300871, #316604, #316782, #330789) 3775 3776Fix use-after-free bug in e2fsck when finishing up the use of the 3777e2fsck context structure. 3778 3779Fixed spelling mistakes, typos, and otherwise clarified man pages and 3780documentation. (Addresses Debian Bugs: #329859, #322188, #316811, 3781#312515, #351268, #357951, #347295, #316040, #368392, #368393, #368394, 3782#368179) 3783 3784Fixed various Debian packaging issues --- see debian/changelog for 3785details. (Addresses Debian Bugs #317862, #320389, #290429, #310950, 3786#310428, #330737, #330736, #329074, #356293, #360046, #366017, #364516, 3787#362544, #362970) 3788 3789 3790Programmer's notes: 3791------------------- 3792 3793Update config.guess and config.sub to latest version (2006-02-23) from 3794FSF. 3795 3796Fix asm_types.h type conflicts on AMD 64 platforms. (Addresses Debian 3797Bugs: #360661, #360317) 3798 3799Fixed the Makefile so that they work correctly on newer versions of 3800GNU make (i.e., 3.81). 3801 3802Add valgrind support to the regression test suites, and eliminate 3803false positives from valgrind. 3804 3805Add a regression test suite for the blkid library. 3806 3807Fix a fencepost error in resize2fs caught by valgrind. 3808 3809Fix compiler warnings about missing memcpy prototypes. 3810 3811We no longer have the sparc assembly code in the header file any more, 3812so we shouldn't set _EXT2_HAVE_HAS_BITOPS_ for the sparc. This would 3813break compiles on the sparc architectures when using gcc. 3814 3815In the libext2fs library, add the new field fs->now which if non-zero 3816is used instead of the system time when setting various filesystem 3817fields (last modified time, last write time, etc.) 3818 3819Fix gcc 4.01 complaints by adding a missing #include <string.h> to 3820ext2fs.h which is needed since the inline functions use memcpy(). 3821(Addresses Sourceforge Bug #1251062) 3822 3823Use BUILD_CFLAGS and BUILD_LDFLAGS instead of CFLAGS and LDFLAGS in 3824the build system when building files in the util directory which are 3825needed during the build process. This avoids potential problems when 3826cross-compiling and some of the options specified in CFLAGS or LDFLAGS 3827are not recognized as valid by the host compiler. (Addresses 3828Sourceforge Bug #1261547) 3829 3830Clean up the blkid library by making the superblock and generic i/o 3831functions to be more generic. Clean up interface to the probe 3832function, and fix memory leak. Finallly, remove an unneeded reference 3833to probe.h in the lib/blkid/resolve.c 3834 3835Add an ext2fs_read_bb_FILE regression test to confirm proper detection 3836of invalid block #'s. 3837 3838The x86 asm constraints for ext2fs_{set/clear}_bit have been fixed to 3839indicate that the the function read/writes the memory location. 3840 3841Fix various gcc -Wall complaints. 3842 3843Add a dependency to make sure that the subdirectories are created 3844before creating all of the object files, in order to address parallel 3845build problem in the library Makefiles. (Addresses Sourceforge Bug: 3846#1261553) 3847 3848Add $(LDFLAGS) to the command line argument when generating an ELF or 3849Solaris shared library, to allow cross-compile and other builds that 3850might need to specify -L paths to needed libraries. (Addresses 3851Sourceforge Bug #1261549) 3852 3853Add a new feature, EXT2_FEATURE_COMPAT_LAZY_BG, which is initially 3854intended for testing purposes. It allows an ext2/ext3 developer to 3855create very large filesystems using sparse files where most of the 3856block groups are not initialized and so do not require much disk 3857space. Eventually it could be used as a way of speeding up mke2fs and 3858e2fsck for large filesystem, but that would be best done by adding an 3859RO_COMPAT extension to the filesystem to allow the inode table to be 3860lazily initialized on a per-block basis, instead of being entirely 3861initialized or entirely unused on a per-blockgroup basis. 3862 3863Fix backwards compatibility so e2fsprogs will better compile on Linux 38642.0.35 systems. 3865 3866Make test scripts more robust against locale-related environment variables 3867 3868Fix type warning problem with sizeof() in ext2fs_open2(). 3869 3870Fix type warning problem with time_t in debugfs. 3871 3872 3873E2fsprogs 1.38 (June 30, 2005) 3874============================== 3875 3876Fix blkid's test programs (built with "make check") compile correctly 3877even without "configure --enable-blkid-debug". 3878 3879Fix ia64 core dump bug caused by e2fsprogs running afoul of C99 strict 3880type aliasing rules on newer gcc compilers. (Addresses Red Hat 3881Bugzilla ##161183.) 3882 3883Fix com_err library to make it more compatible with recent changes 3884made to the com_err library in MIT Kerberos V5 version 1.4. 3885(Addresses Sourcefroge Bug #1150146) 3886 3887General cleanup of messages printed by e2fsprogs programs for grammar, 3888consistency, and to make life easier for translators. Fixed a few 3889strings containing English that had not been marked as needing 3890translations. Removed strings that do not need to be translated, to 3891make life easier for translators. 3892 3893Mke2fs and badblocks will take advantage of a feature in Linux 2.6 to 3894test to see if a device appears to be in use instead of just relying 3895on /proc/mounts and /etc/mtab. (Addresses Debian Bug #308594). 3896 3897Fix portability problem in the filefrag program affecting platforms 3898where the size of an integer is smaller than the size of a long. 3899(Addresses Debian Bug #309655) 3900 3901Mke2fs will now use a larger journal by default for filesystems 3902greater than 4GB. (128 MB instead of 32MB). 3903 3904Mke2fs will refuse to create filesystems greater than 2**31-1 blocks, 3905unless forced. This is to avoid signed vs. unsigned kernel bugs in 3906block numbers that still need to be fixed. 3907 3908The blkid program has a new option which will more efficiently search 3909for device when it is known (or expected) that only one matching 3910device will be found in the system, such as when doing a lookup by 3911UUID. 3912 3913Debian's e2fsprogs-specific initrd fragment will avoid including 3914unnecessary libraries into the initrd ramdisk by unsetting LD_PRELOAD 3915and LD_LIBRARY_PATH, and filtering out libraries found in 3916/etc/ld.so.preload. (Addresses Debian Bug: #304003) 3917 3918Fixed a potential portability issue in the blkid programs for 3919architectures where the char type is unsigned. (Addresses Sourceforge 3920Bug: #1180585) 3921 3922Fix a bug in filefrag so that it doesn't falsely count an extra 3923discontinuity when the first block found is an indirect block. 3924(Addresses Debian Bug #307607). 3925 3926Fix blkid's recognition of cramfs filesystems, and enhance it to be 3927able to handle cramfs labels. 3928 3929Fix debugfs's stat command to not core dump when a filesystem is not 3930open. 3931 3932Fix e2fsck's handling of error conditions caused by the resize inode 3933claiming blocks that are also used by other inodes, a filesystem 3934corruption which was commonly caused by a bug in Fedora Core 3's 3935resize2fs program. 3936 3937Fixed bug in filefrag which caused it to fail on non-ext2/3 3938filesystems. (Addresses Debian Bug: #303509) 3939 3940If the superblock last mount time indicates that the system clock may 3941not be accurate, then e2fsck will omit checking inodes' deletion time 3942field for indications of a potential corrupted orphaned inode list. 3943(Previously e2fsck only ommited these LOW_DTIME checks when the 3944superblock's last write time looked insane.) 3945 3946Fixed a IA64 core dump bug in the e2p library which affected dumpe2fs. 3947(Addresses Debian bug #302200) 3948 3949Make the blkid library more paranoid about being run from setgid 3950programs, and to use __secure_getenv() from libc if it is available. 3951 3952Fixed spelling mistakes, typos, and otherwise clarified man pages. 3953(Addresses Debian Bugs: #304591, #304592, #304594, #304597, #304593 3954and Sourceforge Bug: #1189803) 3955 3956Updated and fixed translations. 3957 3958Fixed various Debian packaging issues --- see debian/changelog for 3959details. 3960 3961Programmer's notes: 3962------------------- 3963 3964Ext2fs_set_bit(), ext2fs_clear_bit(), and ext2fs_test_bit( have been 3965changed to take an unsigned int for the bit number. Negative bit 3966numbers were never allowed (and didn't make any sense), so this should 3967be a safe change. This is needed to allow safe use of block numbers 3968greater than or equal to 2**31. 3969 3970The compile_et program will avoid recreating generated foo_err.c and 3971foo_err.h files if no changes are necessary. The compile_et program 3972will also atomically replace these files to avoid a potential parallel 3973build race problem on SMP systems. (Addresses Sourceforge Bug: 3974#1157933) 3975 3976Added a new function to the blkid library, blkid_probe_all_new(), 3977which only probes newly added disk devices, and change 3978blkid_find_dev_with_tag() to use this function so that when a 3979requested tag is not found, devices that were previously not checked 3980are searched before searching all devices in the system. 3981 3982Added new functions to the blkid library, blkid_dev_set_search() and 3983blkid_dev_has_tag(). 3984 3985E2fsck's problem strings can now use @m and @n as abbrevations for 3986"multiply-claimed" and "invalid", respectively. 3987 3988The e2fsprog.pot file now has an explanation of how the @-expansion 3989and %-expansion works, and strings in e2fsck/problem.c which contain @ 3990characters now have comments in e2fsprogs.pot with the @-expansion to 3991make life easier for translators. 3992 3993Fixed missing return values in the ext2fs library which could cause it 3994to return random garbage in certain error conditions. 3995 3996Allow the current time to be overriden via the E2FSCK_TIME environment 3997variable for use in regression tests. 3998 3999The test scrpit driver program now exits with a non-zero status if 4000there any of its test that it ran failed. 4001 4002Fixed problems with parabuilds on SMP systems. (Addresses Sourceforge 4003Bug: #1157933) 4004 4005Fixed "make check" so that it compiles correctly even when e2fsprogs' 4006header files have not be installed in the system include directories. 4007(Addresses Sourceforge Bug: #1180572) 4008 4009Fixed gcc -Wall nits. 4010 4011 4012E2fsprogs 1.37 (March 21, 2005) 4013=============================== 4014 4015Add support for checking the validity of Extended Attributes stored in 4016inodes to e2fsck. 4017 4018Add support for dumping the contents of large inodes to debugfs, 4019including the extended attributes stored in inodes. 4020 4021Fix mke2fs, e2fsck, debugfs, and the ext2fs_mkdir function so that 4022when we create a new inode we make sure that the extra information in 4023the inode (any extra fields in a large inode and any ea-in-inode 4024information) is initialized correctly. This can take place when 4025mke2fs creates the root and lost+found directory, when e2fsck creates 4026a new root inode or a new lost+found directory, and when the user uses 4027the debugfs write, mknod, or mkdir commands. Otherwise, the newly 4028create inode could inherit garbage (or old EA information) from a 4029previously deleted inode. 4030 4031Fixed a bug in e2fsck so it would notice if a file with an extended 4032attribute block was exactly 2**32 blocks, such that i_blocks wrapped 4033to zero. 4034 4035Added support to filefrag to detect files which are using the new 4036experimental file extents format, and use the non-ext2 algorithm in 4037that case. Fixed a bug to avoid reporting a false discontinuity if 4038there is one or more unallocated blocks at the beginning of a file. 4039 4040Duplicated a check for noticing whether or not the number of blocks 4041(given a certain blocksize) is greater than 2**32 when the 4042BLKGETSIZE64 ioctl is not available to ext2fs_get_device_size(). This 4043allows mke2fs to automatically use a larger blocksize when creating a 4044filesystem on a very large device when run on systems that do not 4045support BLKGETSIZE64. 4046 4047Fix the I18N build which was broken in e2fsprogs 1.36 because the 4048build system had been switched to treat the .gmo files as shipped 4049files (for backwards compatibility with systems that have older GNU 4050I18N tools installed), but the gen_tarball.in script was still 4051removing the .gmo files from the official source distribution. 4052 4053Fixed various Debian packaging issues --- see debian/changelog for 4054details. (Addresses Debian Bugs ##296769, #299341) 4055 4056Programmer's notes: 4057------------------- 4058 4059Added new functions to the e2p library which convert between a string 4060and os_type: e2p_os2string() and e2p_string2os(), and used them to 4061make the generated binaries more compact. 4062 4063Fixed a compile-time error on Darwin systems. 4064 4065Cleaned up the lib/ext2fs Makefile slightly. 4066 4067 4068E2fsprogs 1.36 (February 4, 2005) 4069================================= 4070 4071All of the patches that were applied to Fedore Core 3's 4072e2fsprogs-1.35-11.2 have been integrated, although sometimes with a 4073lot of bug fixes first. Users of Fedora Core 3 are strongly 4074encouraged to upgrade to e2fsprogs 1.36 as soon as possible. 4075 4076Add support for filesystem with the online resizing via resize inode 4077feature. Fixed numerous bugs from the Fedora patches. The Fedora 4078patches also didn't bother to do any consistency checking on the 4079resize inode, or add any tests to the regression test suite. The "-R 4080resize=4g" option to mke2fs was a no-op in the Fedora patches, despite 4081being listed in mke2fs's usage message. All of these shortcomings 4082have been corrected. 4083 4084E2fsck can also also fix filesystems trashed by Fedora's resize2fs 4085program. In order to do this, the user must run the commands: 4086 4087 debugfs -w /dev/hdXXX -R "features ^resize_inode" 4088 e2fsck -f /dev/hdXXX 4089 4090Optionally, the ext2prepare command can be used to re-enable online 4091resizing after the filesystem has been fixed. 4092 4093The fsck program will now accept an optional filedescriptor argument 4094to the -C option. (The Fedora version of this patch would sometimes 4095cause fsck to ignore a parameter on fsck's command line in some rare 4096cases, sigh.) 4097 4098Make sure e2fsprogs doesn't write garbage into the reserved portion of 4099large inodes. 4100 4101Make sure resize2fs releases the blocks belonging to the old inode 4102table blocks when moving the inode table. (Addresses Debian Bug: 4103#290894) 4104 4105Skip the r_resize_inode test if resize2fs is not compiled (due to 4106configure --disable-resizer) 4107 4108E2fsck now checks the summary filesystem accounting information, and 4109if any of the information is obviously wrong, it will force a full 4110filesystem check. (Addresses Debian Bug #291571) 4111 4112Fix e2fsck to not complain when the resize_inode feature is enabled, 4113s_reserved_gdt_blocks is zero, and there is no DIND block allocated in 4114the resize inode. 4115 4116Fix e2fsck to note delete symlinks that contain an extended attribute 4117after the ext_attr feature flag has been cleared. (Addresses Red Hat 4118Bugzilla #146284). 4119 4120Add new utility program, copy_sparse.c, which is very useful 4121for dealing with large sparse files (such as e2image files). 4122 4123Add support for jnl_blocks[] for debugfs's set_super_value. 4124 4125Fix filefrag so that it works correctly with sparse files. 4126 4127Filefrag -v will print first and last blocks. 4128 4129Add interpretation of OS Creator values for FreeBSD and Lites in mke2fs 4130and dumpe2fs. 4131 4132Add mke2fs support so that it can support filesystems larger than 4TB 4133automatically, by retrying with a 4k blocksize if the device size is 4134too big to be expressed using a 1k blocksize. (Addresses Sourceforge 4135bug #1106631) 4136 4137Change blkid to test for NTFS first because Windows sometimes doesn't 4138clear enough of the parition to confuse the probing routines into 4139thinking the old filesystem type is still valid. (Addresses Debian 4140Bug #291990) 4141 4142Add support for swap partition label and uuid's in the blkid library. 4143 4144Add support to the blkid library to recognize Oracle ASM volumes. 4145 4146Make blkid -t display all devices that match the specified criteria, 4147not just the first one, and work more consistently when the blkid 4148cache file is not available or set to /dev/null. (Addresses Debian 4149Bug #290530 and #292425) 4150 4151Badblocks will now correctly display block numbers greater than 4152999,999,999 in its progress display. 4153 4154The tune2fs program will not allow the user from setting a ridiculous 4155number of reserved blocks which would cause e2fsck to assume the 4156superblock was corrupt. E2fsck's standards for what is a ridiculous 4157number of reserved block has also been relaxed to 50% of the blocks in 4158the filesystem. 4159 4160The blkid library will return vfat in preference to msdos, and ext3 in 4161preference to ext2 (if the journalling flag is set) so that mount will 4162do the right thing. (Addresses Debian bug #287455) 4163 4164Mke2fs will now use the -E option for extended options; the old -R 4165(raid options) option is still accepted for backwards compatibility. 4166Fix a double-free problem in resize2fs. (Red Hat Bugzilla #132707) 4167 4168Mke2fs will now accept a size in megabytes, gigabytes, and other units 4169(via "32m" or "4g" on the command line) if the user finds this more 4170convenient than specifying a block count. 4171 4172Fix an obscure, hard-to find bug in "e2fsck -S" caused by an inode 4173cache conherency problem. 4174 4175Debugfs now supports a new command, set_inode_field, which allows a 4176user to manually set a specific inode field more conveniently, as well 4177as set entries in the indirect block map. 4178 4179Debugfs's set_super_value command has been enhanced so that the user 4180can set most superblock fields, including the date/time fields and 4181some of the more newsly added superblock fields. 4182 4183E2fsprogs programs now accept an offset to be passed to the file 4184specifiers, via the syntax: "/tmp/test.img?offset=1024". 4185 4186E2fsprogs programs will now accept blocksizes up to 65536; kernel 4187support on the x86 doesn't exist for now, but it can be useful on 4188other architectures with page sizes greater than 4k. There are 2.6 4189kernel patches out there which enable this, but they are of this 4190writing still experimental. 4191 4192The e2image command now takes the -s option which will scramble 4193directory entries for raw image files. 4194 4195Fix a file descriptor leak in the filefrag program. 4196 4197Make sure e2fsck doesn't crash when /proc/acpi/ac_adapter is not 4198present. 4199 4200Fix bug in debugfs where kill_file would lead to errors when deleting 4201devices and symlinks. (Sourceforge Bugs #954741 and #957244) 4202 4203Fix bug in the blkid library when detecting the ocfs1 filesystem 4204 4205Remove obsolete EVMS 1.x and a.out DLL support. 4206 4207E2fsck will attempt to recover from a journal containing illegal blocks. 4208 4209Fixed two potential ordering constraint problems in e2fsck which might 4210cause the filesystem to be corrupted if e2fsck is interrupted during a 4211(extremely narrow) race window. Thanks to Junfeng Yang from the 4212Stanford Metacompilation group for pointing this out. 4213 4214Fixed bug in e2fsck where it would not accurately detect whether or 4215not the system is running on adaptor if the ACPI device representing 4216the AC adapter didn't correspond to the what was used on IBM 4217Thinkpads. 4218 4219Change e2fsck to accept directories greater than 32MB. 4220 4221Fix e2fsck so that a checkinterval of zero disables a time-based check 4222of the filesystem. 4223 4224Debugfs will check the DEBUGFS_PAGER enviroment variable in preference 4225to the PAGER environment variable. (Addresses Debian Bug #239547) 4226 4227Tune2fs will not mark rewrite the superblock if the feature bitmasks 4228are not modified. 4229 4230The debugfs program will set the filetype information when creating a 4231link. 4232 4233Add debugfs -d option to use a separate source of data blocks when 4234reading from an e2image file. 4235 4236Add e2image -I option which allows the e2image metadata to be 4237installed into a filesystem. 4238 4239Fixed bug in the badblocks program which caused "done" to always 4240appear in english even when a translation was available. (Addresses 4241Debian Bug #252836) 4242 4243The blkid program has a new option -o which controls the output format 4244of the blkid program; this is makes blkid more convenient to use in 4245shell scripts. 4246 4247Fix a minor bug in uuid library, which was not using the full 14 bits 4248of clock sequence when generating UUID's. 4249 4250Fix a Y8.8888K problem in the uuid library. 4251 4252Logsave now creates a new session id for itself to avoid getting 4253killed by init whan transitioning between init levels. 4254 4255Change the licensing of the UUID library to be the 3-clause BSD-style 4256license; this allows Apple to use the uuid library in Darwin. 4257 4258Add ocfs and ocfs2 probe support into the blkid library. 4259 4260Fix a memory and file descriptor leak in the blkid library. 4261 4262The blkid library will revalidate the device if the system time is 4263earlier than last verification time of the device, since that 4264indicates that the system time is probably nottrustworthy. 4265 4266The blkid library will override the default location of the blkid.tab 4267file by the BLKID_FILE environment variable, if it is available. 4268 4269Change the getsize functions to use the BLKGETSIZE64 ioctl on Linux 2.6. 4270 4271Add various portability fixes for lame new versions of glibc, Darwin 4272and GNU/KFreeBSD, as well as removing XSI:ism's. (Addresses Debian 4273Bugs #239934, #264630, #269044, #255589, #289133) 4274 4275Add support for Windows 9x/NT under Cygwin. 4276 4277Updated and clarified various man pages. (Addresses Debian Bugs #236383, 4278 #241940, #238741, #242995, #256669, #268148, #256760, #273679) 4279 4280Updated and fixed translations. (Addresses Debian bugs #244105, #262836) 4281 4282Update the rpm spec files so that it works better with Fedora core 2 4283and RH9. 4284 4285Fixed various Debian packaging issues (see debian/changelog). In 4286particular, fixed the Debian initrd scripts. (#241183, #248050, 4287#253595, #247775) 4288 4289 4290Programmer's notes: 4291------------------- 4292 4293Fixed various gcc -Wall warnings. 4294 4295The uuid library now has new functions uuid_unparse_upper() and 4296uuid_unparse_lower() which forces the case of the hex digits to be 4297upper case, or lower case. 4298 4299The build process has been speeded up by enhancing the subst program 4300to update the modtime on the generated files even when the generated 4301file hasn't changed. 4302 4303The uuid library now uses C99 stdint.h types instead of custom types. 4304 4305Updated config.guess and config.sub with newer versions from the FSF. 4306 4307Removed out of date .cvsignore files from the source distribution. 4308 4309The ext2fs_unlink() function will return an error if both the name and 4310inode number are unspecified, to avoid doing something surprising 4311(such as unconditionally deleting the first directory entry). 4312Directory entries are now deleted by coalescing them with the previous 4313directory entry if possible, to avoid directory fragmentation. This 4314is not an issue with the e2fsprogs suite, but may be a problem for 4315some of the users of libext2fs, such as e2tools. 4316 4317Add support for version numbers of the form "1.36-rc1". 4318 4319Fix build of mke2fs.static. 4320 4321Add basic ext2fs library support for large (EA in inode) inodes. 4322 4323The test_io mechanism can now abort after n reads or writes to a 4324particular block. The block is specified by TEST_IO_BLOCK environment 4325variable, and the read/write count by the TEST_IO_READ_ABORT and 4326TEST_IO_WRITE_ABORT environment variables. The block data is now only 4327dumped if the 0x10 bit is set in TEST_IO_FLAGS. 4328 4329UUID_DEFINE() in the uuid library now creates a static variable, with 4330__attribute__ ((unused)) if we are using GCC, so that UUID_DEFINE can 4331be used in header files. 4332 4333Add support for the install-strip and install-shlibs-strip targets, as 4334suggested by the GNU coding guielines. "make install" no longer 4335strips the binaries which are installed. 4336 4337Remove support for the --enable-old-bitops configure option which was 4338only for very old sparc systems. 4339 4340Remove support for --enable-clear-htree; this was only needed during 4341the early development of the htree patch. 4342 4343Use Linux-kernel-style makefile output so it is easier to see compiler 4344warnings. 4345 4346Update gettext files to version 0.14.1. 4347 4348Update to use autoconf 2.5x. 4349 4350Improved support for compiling e2fsprogs under dietlibc. 4351 4352Make e2fsprogs portable to Solaris and FreeBSD systems. 4353 4354Add blkid_verify(), blkid_get_library_version(), and 4355blkid_parse_version_string() functions to the blkid library. 4356 4357Add pkg-config files for e2fsprogs's libraries. 4358 4359Fix "make uninstall" to so that it removes everything that is installed. 4360 4361Add a configure --enable-maintainer-mode option which enables the 4362makefile rules to rebuild the configure script from configure.in, and 4363to reubuild the .gmo files in po directory. 4364 4365Drop the sparc assembly bitwise operations; it's less efficient 4366than the GCC 3.4 compile code and triggers compiler warnings on 4367sparc64. Thanks to Matthias Andree for his analysis and suggestions. 4368(Addresses Debian Bug #232326) 4369 4370 4371E2fsprogs 1.35 (February 28, 2004) 4372================================== 4373 4374E2fsck has a new -k option, which in conjunction with the -c options, 4375preserves the existing badblocks list. 4376 4377Cleaned up e2fsck's preen-mode messages during the passes 1b, 1c, and 1d. 4378 4379E2fsprogs will now deal correctly with symlinks that contain 4380extended attribute information, which can be created using SE Linux. 4381(Addresses Debian Bug #232328) 4382 4383Remove a double longjump into an invalid stack frame bug in e2fsck. 4384(This was during an abort sequence, which normally worked on Linux and 4385caused a core dump on other operating systems.) 4386 4387Fix NLS bug in e2fsck, by avoiding trying to expand an empty string 4388(the NLS library will replace "" with the .po header information). 4389 4390Fix a bug in mke2fs which caused -T largefile or -T largefile4 to core 4391dump due to a division by zero error. (Addresses Debian bug #207082) 4392 4393Fixed a bug in e2fsck which caused it to incorrectly fix a filesystem 4394when reconnecting a directory requires creating a lost+found 4395directory. (Addresses Debian bug #219640). 4396 4397Fixed a bug where e2fsck would bomb out if a journal needed to be 4398replayed when using an alternate superblock. 4399 4400E2fsck will give an extra grace period before actually forcing a check 4401if the laptop is running on battery. The next time fsck runs while 4402the system is on the AC mains, or after the grace period is exceeded, 4403the filesystem will be checked. (Addresses Debian bug #205177) 4404 4405E2fsck will inform the user when there are 5 or fewer mounts before a 4406filesystem check will be forced. (Addresses Debian bug #157194) 4407 4408Fix e2fsck's handling of corrupted indirect blocks in the bad block. 4409We now correctly handle the case where there is an overlap between a 4410block group descriptor or a superblock and a bad block indirect block. 4411In the case where the indirect block is corrupted, we now suggest 4412"e2fsck -c". 4413 4414Fix byte swap bugs in e2fsck that caused the journal backup location 4415in the superblock and symlinks created by SE Linux to be cleared 4416by e2fsck on big-endian machines. (Addresses Debian bug #228723) 4417 4418E2fsck -c now replaces the current list of bad blocks with the ones 4419found by badblocks. 4420 4421Fix bugs in e2fsck and tune2fs which could cause a core dump if a 4422non-existent LABEL or UUID specifier is to e2fsck or tune2fs. 4423 4424Fix a potential bug in e2fsck which could cause it to core dump when 4425trying to print the location of the backup superblock. 4426 4427Protect against a potential core dump in e2fsck when printing a 4428message about backup superblocks. 4429 4430Add support for backing up the journal inode location in the 4431superblock. E2fsck will automatically save the journal information in 4432the superblock if it is not there already, and will use it if the 4433journal inode appears to be corrupted. ext2fs_add_journal_inode() 4434will also save the backup information, so that new filesystems created 4435by mke2fs and filesystems that have journals added via tune2fs will 4436also have journal location written to the superblock as well. 4437Debugfs's logdump command has been enhanced so that it can use the 4438journal information in the superblock. 4439 4440E2fsck will now update all superblocks when moving the journal inode. 4441 4442Shrink the size of the e2fsck executable by moving some initialized 4443variables to the BSS segment. 4444 4445E2fsck will avoid printing the ^A and ^B characters which bracket the 4446progress bar when stdout and stdin are a tty device instead of a pipe 4447to another program. (Addresses Debian bug #204137) 4448 4449Debugfs's mkdir command will automatically expand the directory if 4450necessary. (Addresses Debian Bug: #217892) 4451 4452Fixed a bug in debugfs so that copying a file from /dev/null uses the 4453correct mode bits. (Addresses Debian Bug: #217456) 4454 4455If the environment variables DEBUFS_PAGER and PAGER are not set, 4456debugfs now searches for the appropriate pager to use, beginning with 4457/usr/bin/pager, and then falling back to 'more' and 'less'. 4458(Addresses Debian bug #221977) 4459 4460Debugfs will now support 2.6 device numbers where the major or minor 4461number may be larger than 255. (Addresses Sourceforge bug #865289) 4462 4463Fix debugging printf in resize2fs. (Addresses Debian Bug #271605) 4464 4465Chattr now stops processing options when it sees '--'. (Addresses 4466Debian bug #225188) 4467 4468Fix regression tests so they work correctly when e2fsprogs is compiled 4469with configure --disable-htree. 4470 4471Fix bug in uuid library when there is no network card and the library 4472is generating a time-based uuid. The random MAC address was not 4473correctly generated to be a multicast address. 4474 4475Add compile_et extensions from Heimdall that were missed the first time 4476around. 4477 4478Fix bug in badblocks when using O_DIRECT; we need to make sure that 4479we're reading from an offset which is page aligned. For read-only and 4480read-write tests, we try to recover after an error so that we can 4481continue reading on page-aligned boundaries. (Addresses Debian Bug 4482#203713) 4483 4484Badblocks now checks 64 blocks at a time instead of 16. (Addresses 4485Debian bug #232240) 4486 4487Updated and clarified various man pages. (Addresses Debian Bug 4488#206845, #222606, #214920, #232406) 4489 4490Updated and fixed translations. (Addresses Debian bugs #200086, #214633) 4491 4492Fixed various Debian packaging issues (see debian/changelog). 4493 4494Programmer's notes: 4495------------------- 4496 4497Fixed a build problem so that e2fsprogs would compile with the 4498--enable-profile option to configure selected. (Addresses Sourceforge 4499bug #811408) 4500 4501Fixed C++ problems with the ext2fs.h header. (Addresses Red Hat 4502Bugzilla Bug #112448) 4503 4504Centralize code which calculates the location of the superblock 4505and block group descriptors so that it is in a single library routine. 4506 4507Added two new functions, ext2fs_file_open2() and 4508ext2fs_inode_io_intern2() which take a pointer to an inode structure. 4509 4510Fix compile_et to output the correct prototype for 4511initialize_xxx_err_table_r() in the header file. (Addresses Debian 4512bug #204332) 4513 4514In the lib/et makefile, make sure com_err.info is deleted on "make clean". 4515 4516Fix 64-bit warnings in e2fsprogs pass1b by using inttypes.h if 4517present. This is for when we try stuffing an int into void * pointer. 4518 4519Fix type-punning which can cause gcc 3.x to miscompile code by getting 4520confused about pointer aliasing. ext2fs_getmem(), ext2fs_free_mem(), 4521and ext2fs_resize_mem() all now take a 'void *' instead of a 'void 4522**'. The EVMS code uses an ugly union approach since we don't want to 4523modify the EVMS interfaces. 4524 4525Make sure all Makefiles use $(MAKE) rather than hardcoded "make", to 4526aid build process on systems can use invoke GNU make as "gmake". 4527 4528Added regression testing for mke2fs. 4529 4530Fixed gcc -Wall nitpicks. 4531 4532Fixed various compiler warnings. 4533 4534Add portability fixes for FreeBSD and for using fsctl under Darwin to 4535support ext2 ioctl's. 4536 4537 4538E2fsprogs 1.34 (July 25, 2003) 4539=============================== 4540 4541Fixed a bug introduced in E2fsprogs 1.30 which caused fsck to spin in 4542a tight loop while waiting for a child fsck to exit in some cases. 4543This burns CPU times which slows down the low-level filesystem check. 4544 4545Added code to mke2fs to assure that the default block size for a 4546filesystem is at least as big as the sector size of the device, if it 4547can be determined. 4548 4549Changed mke2fs and resize2fs to round the default size of a filesystem 4550to be an even multiple of the VM pagesize in order to avoid a Linux 4551kernel bug introduced when the storage of the buffer cache was moved 4552into the page cache. 4553 4554Mke2fs will warn the user when creating a filesystem with journaling 4555and a blocksize greater than 4096. (Addresses Debian bug #193773) 4556 4557Fixed a bug in resize2fs which caused it to fail on filesystems with a 4558non-empty bad block list. Resize2fs now discards any blocks on the 4559badblock list which are no longer part of the filesystem as the result 4560of a filesystem shrink. (Note: this means that shrinking and then 4561enlarging a filesystem is no longer a reversible operation; 4562information about bad blocks in the part of the filesystem which is to 4563be chopped off will be lost.) 4564 4565Changed resize2fs so the user can use prefixes to specify the units of 4566the new filesystem size (sectors, kilobytes, megabytes, or gigabytes), 4567and to make the error and informational messages explicitly display 4568the blocksize used by the filesystem, in order to avoid confusion. 4569(Addresses Debian bug: #189814) 4570 4571Added a new debugfs command, dump_unused, which dumps the contents of 4572all unused blocks to stdout. (Useful as an emergency try-to-find 4573deleted data command.) 4574 4575Added a new debugfs command, imap, which prints the location of a 4576specified inode in the inode table. 4577 4578Fixed a bug in the badblocks program which caused it to use one bit of 4579randomness in its non-destructive read/write test, instead of using a 4580full 8 bits of randomness. 4581 4582Added a new option (-t) to badblocks, which allows the user to control 4583the test pattern(s) used when checking a disk. 4584 4585The blkid probe function now more correctly detects UDF filesystems. 4586 4587Fixed a bug in the blkid library which caused it to not update its 4588cache if a filesystem changes from having a label to not having a 4589label. 4590 4591Fixed a bug in the blkid library wihch could avoid an infinite loop 4592in blkid_find_dev_with_tag() if /proc is not mounted and there the 4593/etc/blkid.tab file has not yet been created. 4594 4595Fixed the badblocks program so that the destructive read/write test 4596honors the -c option, and to use O_DIRECT when possible to avoid 4597thrashing the system block buffer cache. 4598 4599Fixed various NLS issues. 4600 - Added Czech and Swedish translations 4601 - Removed testing NYC translation 4602 - Fixed NLS support for message abbrevations in e2fsck 4603 - Remove de-utf.po, since we shouldn't have two versions using different 4604 charset encodings. 4605 - Used ngettext() (abbreivated with the macro P_(str1, str2, n)) to 4606 simplify the statistics reporting in e2fsck. 4607 4608Changed configure.in so that its defaults for *BSD systems no longer 4609build an fsck wrapper, and not to install in /usr/local by default. 4610 4611Fixed some minor spelling errors/typo's in e2fsck and the configure 4612script. 4613 4614Fixed various Debian packaging issues (see debian/changelog). 4615 4616Updated and clarified man pages. (Addresses Debian Bug #195616) 4617 4618Programmer's notes: 4619------------------- 4620 4621Fix gcc -Wall nitpicks. 4622 4623Updated gettext implementation used by e2fsprogs to 0.11.5, and enable 4624NLS support by default. (Added partial workaround for gettext/Darwin 4625incompatibility problems.) 4626 4627Added full MIT KRB5 and Himdall compaibility support to the com_err 4628library and the compile_et program. (Addresses Debian bug #191900) 4629 4630Added the blkid_known_fstype() function to the blkid library, which 4631returns true if it is passed a filesystem type which is recognized by 4632the blkid probing functions. 4633 4634Improved the documentation for the blkid library. 4635 4636Added the ext2fs_get_device_sectsize() function the the ext2fs library, which 4637returns the hardware sector size of a device, if it is available. 4638 4639Added a dependency in the blkid library's .so file to the uuid 4640library, since the former uses the latter. (Addresses Debian bug 4641#194094) 4642 4643Added --with-diet-libc and --disable-evms to the configure script. 4644 4645Fixed a minor memory leak in the badblocks program. 4646 4647Fixed a portability problem in tune2fs --- not all systems have strptime(). 4648 4649Fixed a portability problem in debugfs with the use of getopt() more 4650than once. Old-style BSD, new-style BSD, and Linux C libraries all do 4651things differently. 4652 4653Add support Windows support to ext2fs_get_device_size(). 4654 4655Added (normally disabled) debugging code to the Unix I/O manager which 4656causes it to disable all userspace caching if the NO_IO_CACHE is 4657defined. 4658 4659Changed the test I/O manager so it can always be linked into e2fsck, 4660mke2fs, and tune2fs if enabled via --enable-test-io-debug to the 4661configure script. The test I/O manager will only print any debugging 4662information if the TEST_IO_FLAGS or TEST_IO_BLOCK environment 4663variables are set, which specifies which I/O operations are logged and 4664a block number to watch, respectively. The log messages are sent to 4665stderr by default, unless a filename is specified via the 4666TEST_IO_LOGFILE environment variable. 4667 4668 4669E2fsprogs 1.33 (April 21, 2003) 4670=============================== 4671 4672Added a new utility program, logsave, which captures the output of a 4673command in a log file, even if the containing directory hasn't been 4674mounted yet (in which case the it saves the output in memory until it 4675can write out the logfile). This is useful for capturing the 4676output of fsck during the boot sequence. 4677 4678Fixed some portability problems that were causing problems under 4679the Cygwin32 environment. 4680 4681Mke2fs now interprets a negative number to the -b option as a minimum 4682block size. 4683 4684Fixed a bug in mke2fs which was incorrectly checking the argument to 4685the -g option if the default block size was used. (Addresses Debian 4686bug #188319) 4687 4688Fsck now explicitly ignores tmpfs and devpts, and it will complain if 4689it can not find filesystem checkers for jfs, reseirfs, and xfs. 4690 4691E2fsck now updates the global free block and inode counters from the 4692block group specific counters quietly. This is needed for an 4693experimental patch which eliminates locking the entire filesystem when 4694allocating blocks or inodes; if the filesystem is not unmounted 4695cleanly, the global counts may not be accurate. 4696 4697Imported bug fixes to the EVMS plugin from the EVMS 2.0 tree. (EVMS 46982.0 is not yet supported; this only pulled in the bug fixes: fixed 4699possible hangs caused by bugs in calling waitpid, and not setting the 4700pipe to non-blocking mode; also fixed a file descriptor leak; made 4701sure all functions call log entry/exit functions.) 4702 4703Badblocks now flushes its output file as bad blocks are discovered. 4704 4705The uuid library is now more paranoid about assuming the correctness 4706of the /dev/random device; it mixes in a stream of bytes from 4707random/srandom, just in case. 4708 4709Update Debian files to reflect the fact that I am now the Debian 4710maintainer of e2fsprogs. Other various Debian-specific packaging 4711cleanups. 4712 4713Move the source tarball generation functions from the top-level 4714makefile to the util/gen-tarball script. 4715 4716Updated the Turkish .po translation file. 4717 4718Added Heimdal and MIT krb5 extensions to the com_err library to make 4719it more compatible with com_err libraries from those distributions. 4720 4721Changed dumpe2fs to always display the superblock fields relating to 4722the journalling and/or directory indexing feature even if those 4723features are not enabled. 4724 4725Updated and clarify copyright statement vis-a-vis alpha releases of 4726e2fsprogs. 4727 4728The ss library will now try to dynamically link to the readline 4729library and use it if it is present in the system. This means that 4730the debugfs program now has line editing and history features. The 4731SS_READLINE_PATH environment variable is used to find a readline or 4732readline-compatible library. 4733 4734E2fsck now finds most duplicate filenames (all when rebuilding all 4735directories via the -D option) and offers to delete or rename 4736duplicate filenames/directory entries. (Addresses Debian Bug #176814). 4737 4738Fix bug in e2image. When writing out a raw image file, include data 4739blocks from symlinks that do not store the symlink within the inode. 4740 4741Fix bug in resize2fs which incorrectly moved the block and inode 4742bitmaps for sparse superblock filesystems and incorrectly marked 4743blocks as in use. (Addresses Debian bug #174766) 4744 4745Added a new shared library, the blkid library, which efficiently 4746allows fsck, mke2fs, e2fsck, and tune2fs to be able to look up LABEL 4747and UUID filesystem specifiers without needing to search all of the 4748devices in the system. Instead, the device is looked up in a cache 4749file, and then verified to make sure the blkid cache is correct. 4750 4751Tune2fs and e2label will accept LABEL=xxx and UID=yyy specifiers for 4752the device name, using the blkid library. (Addresses Debian bugs 4753#166048, #179671) 4754 4755Fsck now supports backslash escapes in /etc/fstab so that \040 can be 4756used for spaces in device labels. 4757 4758Removed 32-bit limitations for debugfs's dump command. 4759 4760If the user specifies a large number of inodes, Mke2fs will 4761automatically adjust the number of blocks per group to find a valid 4762set of filesystem parameters. 4763 4764Add workaround to detect broken MD devices where when some of the 4765underlying devices are marked read-only, writes to the MD device are 4766silently dropped. E2fsck will detect if there is an attempt to run 4767the journal twice, and abort with an error if this is the case. 4768(Addresses IBM Bugzilla bug #1226) 4769 4770E2fsck will print an error if more than one of the -p/-a, -n or -y 4771options are specified. 4772 4773E2fsck will fix HTREE corruptions in preen mode, without stopping the 4774boot process. This is needed because the 2.4 ext2 filesystem 4775accidentally had the INDEX_FL backwards compatibility code removed. 4776 4777Mke2fs no longer creates filesystems with the dir_index flag set by 4778default; the user has to specifically request it. 4779 4780Update and clarified various man pages. (Addresses Debian bugs 4781#173612, #175233, #175113, and #170497, #185945, #188318) 4782 4783Created man page for the mk_cmds program (from the libss library). 4784 4785Programmer's notes: 4786------------------- 4787 4788Fix various gcc -Wall nits. 4789 4790Fixed a lot of portability problems that caused e2fsprogs not to build 4791successfully under Solaris and Apple/Darwin. 4792 4793Fixed a Makefile dependency to allow building e2fsprogs using parallel 4794make jobs. 4795 4796Changes to create a subset distribution which consists only of the 4797et, ss, uuid, and blkid libraries. The configure script and top-level 4798makefile were changed to support working with a subset distribution. 4799 4800Removed EXT2_FEATURE_RO_COMPAT_BTREE_DIR mention of since it's not 4801actually used, and might people who are looking for 4802EXT2_FEATURE_COMPAT_DIR_INDEX, which is in use. 4803 4804Updated debian files to fix a number of Lintian warnings. 4805 4806Updated config.guess and config.sub with newer versions from the FSF. 4807 4808Removed unnecessary libraries from being linked into the fsck, lsattr, 4809chattr, and blkid executables. 4810 4811E2fsprogs 1.32 (Nomvember 9, 2002) 4812================================== 4813 4814Fixed a bug in the Unix I/O routines which caused needless writebacks 4815of clean blocks from the unix_io cache (they were erroneously marked 4816as being dirty, so they were getting written back to disk before 4817getting evicted from the disk cache). This was harmless, but it 4818significantly slowed down e2fsck. 4819 4820Made some other minor optimizations to the Unix I/O routines to save a 4821small amount of CPU time. 4822 4823Updated internationalization files. 4824 4825 4826E2fsprogs 1.31 (Nomvember 8, 2002) 4827================================== 4828 4829Update EVMS ext2fsim plugin with EVMS 1.2. (We still support 4830compiling the fsim plugin with EVMS 1.0 and 1.1.) Add better error 4831handling for child process that die unexpectly. Add a hack to force 4832mkfs to create filesystems that won't cause problems with hardware 4833that has 2k or 4k minimum blocksize requirements. Read from child 4834processes in non-blocking mode, so that the GUI continues to update. 4835 4836Fix e2fsck so that it returns the appropriate exit code when the root 4837filesystem has been changed, so that system's rc scripts will be told that 4838the system needs to be rebooted. 4839 4840Fix a bug in ext2fs_flush/ext2fs_close; when the MASTER_SB_ONLY flag 4841was set, some of the descriptor blocks that should have been written 4842out were getting skipped. 4843 4844Changed e2fsck to force out changes to the backup copies of the 4845superblock and block group descriptors when important changes are made 4846to those data structures. 4847 4848Fix a bug where e2fsck could erroneously mark a filesystem as being 4849clean if a check of dirty filesystem is interrupted with a ^C. (Bug 4850introduced in e2fsprogs 1.28.) 4851 4852If journal debuging is enabled using --enable-jbd-debug, the debugging 4853level is now set via the E2FSCK_JBD_DEBUG environment variable. 4854 4855If byteswapping support is disabled using configure --disable-swapfs, 4856skip the tests which depend on byte-swapping, so that "make check" 4857won't bomb out. 4858 4859Lshattr will now display the indexed directory flag. Also, some of 4860the more esoteric compression flags are supressed unless compression 4861support has been enabled. 4862 4863Update man pages. 4864 4865 4866E2fsprogs 1.30 (October 31, 2002) 4867================================= 4868 4869When resizing a filesystem, and inodes need to moved, resize2fs will 4870update the inode's ctime field, and the mtime field of the containing 4871directories, so that incremental backups using dump/restore will work 4872correctly. 4873 4874In order to avoid spurious errors, e2fsck wil check the last 4875superblock write time to determine whether or not it can safely use 4876the LOW_DTIME checks to determine if there are inodes on the orphan 4877inode linked list that had somehow gotten disconnected. (Addresses 4878Sourceforge bug #620980) 4879 4880Partition in /proc/partitions that being with the string "lvm" are 4881considered real partitions even if they do not end with a number. 4882 4883Fixed a bug in the the uuid generation function, where if /dev/urandom 4884is not present, but /dev/random is, and there isn't sufficient 4885entropy, the get_random_byte function could spin a loop forever. 4886 4887E2fsck, mke2fs, etc. will now reliably notice when image files are 4888mounted using the loopback interface. (Addresses Sourceforge bug 4889#619119) 4890 4891When flushing buffers (as is done by badblocks, for example) check to 4892see if the BLKFLSBUF ioctl succeeds, and if so, avoid using the 4893FDFLUSH ioctl which causes the MD device driver which causes confusing 4894syslog messages about obselete ioctl messages. (Addresses Sourceforge 4895bug #545832). 4896 4897Debugfs's write command now checks to see if the destination filename 4898exists before creating it. (Addresses Sourceforge bug #478195.) 4899 4900When installing man pages, search for compressed man pages using all 4901commonly used compression extensions (i.e., .Z, .gz, .bz2, etc.) 4902 4903Fixed a bug in fsck where multiple filesystems specified on the 4904command were not being checked in parallel due to a logic bug 4905introduced to support the FSCK_MAX_INST environment variable. 4906 4907We have added a new superblock field, s_mkfs_time, which indicates 4908when a filesystem was created. It is set via mke2fs, and printed out 4909by dumpe2fs, but is not actually touched by the kernel. 4910 4911Dumpe2fs has been made more robust by not aborting if there is an 4912error reading the block/inode bitmaps; instead it will still print out 4913the location of the block/inode bitmaps and inode table. 4914 4915Add support for the an alternative block group descriptor layout which 4916allows for on-line resizing without needing to prepare the filesystem 4917in advance. (This is the incomat feature flag meta_bg.) 4918 4919Add support for storing default mount options in the superblock, so 4920that the filesystem can be mounted with specific mount options without 4921needing to specify them on the mount command line or in the /etc/fstab 4922file. 4923 4924Add support for a new inode flag, which is to be used for indicating 4925the top of directory hierarchies for the Orlov block allocator. 4926 4927Fix e2fsck so that if it creates the lost+found directory, it does so 4928with the more apporpriate permissions of 0700. Also change 4929mklost+found so that it also creates the directory 0700. 4930 4931Fixed format bug in e2fsck if NLS is in use. 4932 4933Add a German translation for e2fsprogs's NLS support. 4934 4935Fixed e2fsck so that it more handles BAD_BLOCK_IN_INODE_TABLE even at 4936the beginning of the inode table. This won't matter much, since if 4937there is a bad block at the beginning of the inode table, the root 4938directory won't be available. But at least e2fsck won't crash in this 4939case. 4940 4941Fixed endian problems in the htree code for e2fsck and debugfs. 4942 4943When byte-swapping a filesystem on a PPC architecture, byte-swap the 4944bitmaps since the historical big-endian ext2 variant had byte-swapped 4945bitmaps, and the ext2fs library assumes this. This fixes the 4946regression test suite on PPC machines. 4947 4948Fix e2image so that it handles a bad block in the inode table 4949appropriately. 4950 4951E2fsck now uses a more sophisticated algorithm to salvage corrupted 4952directories that recovers more information in the corrupted directory 4953block. 4954 4955E2fsck now performs additional consistency checks on indexed (HTREE) 4956directories. 4957 4958Fix bug where efsck might get confused about whether a completely 4959empty directory block is an empty leaf block or an interior htree 4960node. This could cause e2fsck to get confused into think that a valid 4961indexed directory was corrupted. 4962 4963E2fsck no longer creates an empty directory entry at the end of a 4964directory block when optimizing a directory. This would cause some 4965earlier versions of the dxdir kernel code to corrupt the directory 4966when attempting to split a node containing an empty directory entry. 4967 4968E2fsck could sometimes in rare circumstances leave the indexed flag 4969set after a small directory was optimized by compressing it instead of 4970indexing it. (This should never happen in real life, since 4971directories that small wouldn't have been indexed, but better safe 4972than sorry.) 4973 4974E2fsck now only updates the master superblock in all cases. This 4975slightly shortens its run time. 4976 4977Ext2ed can deal with directory entries of length 0; previously it 4978would get stuck in an infinite loop. 4979 4980Fsck now has support for reiserfs volumes when parsing LABEL= and UUID= 4981specifiers. (Sourceforge patch #613447) 4982 4983Badblocks will now work correctly on read-only devices such as 4984CD-ROM's. (Sourceforge patch #600451) 4985 4986Updated and clarified man pages. (Addresses Debian bug #167108) 4987 4988 4989E2fsprogs 1.29 (September 24, 2002) 4990=================================== 4991 4992Fixed a bug in e2fsck which could corrupt a directory when optimizing 4993it (via the -D option) or rebuiliding the hash tree index with a 1 in 4994512 probability, due to a fence post error. 4995 4996Fixed a bug in the LVM support code which caused LABEL='xxx' not to 4997work correctly. 4998 4999Mke2fs now enables the directory indexing flag by default. (Since 5000this is a compatible feature flag, it's safe to do so.) 5001 5002Tune2fs will support setting the directory indexing feature flag. It 5003will automatically set up the default hash algorithm and hash seed 5004fields in the superblock. 5005 5006If the bone-headed user enters the root filesystem twice in 5007/etc/fstab, the -R option which skips the root filesystem will skip 5008all of them. (Addresses Debian bug #159423). Note! This is not a 5009precedent for dealing intelligently with any other kind of doubled 5010entry in /etc/fstab! 5011 5012 5013Programmer's notes: 5014------------------- 5015 5016Removed perror declaration in lib/et/internal.h. All modern systems 5017can be expected to define perror() these days. Besides, the lib/et 5018code wasn't using perror at all anyway. :-) 5019 5020 5021E2fsprogs 1.28 (August 31, 2002) 5022================================ 5023 5024Add support for the Hashed-Tree Directory Indexing to e2fsck. Support 5025for setting the htree flag is not included yet, although it can be 5026manually turned on via the debugfs program. 5027 5028Clarified e2fsck error message which is printed when it cannot find 5029sufficient contiguous block when relcating filesystem metadata. 5030 5031Added support for building an EVMS plugin module for ext2/3. This 5032module is substantially the same as the EVMS module shipping with EVMS 50331.1, with one or two bugfixes. E2fsprogs can also build this plugin 5034for use with EVMS 1.0 (which did not include the ext2 plugin module), 5035if the configure --enable-old-evms flag is given. 5036 5037Fsck will search through EVMS volumes when trying to resolve 5038filesystem specifications such as LABEL=xxx or UUID=xxx. 5039 5040Added a new utility program, /sbin/findfs, which will return 5041filesystem specifications such as LABEL=xxx or UUID=xxx, and prints 5042the device name. 5043 5044Update and clarified various man pages. (addresses Debian Bug #145044, 5045#146437, #131350, #151990, #144621, #141938) 5046 5047If there are no filesystems specified on fsck's command line, fsck now 5048treat that as if the -As options were given. Previously it would 5049simply do nothing. (Addresses Debian Bug #153102) 5050 5051Mke2fs no longer treats a failure to be able to clear the MD signature 5052at the end of the filesystem as a fatal error. (Addresses Debian Bug 5053#155007) 5054 5055The e2p library functions (which are used by lsattr and chattr) now 5056double check to make sure the file is a regular file or directory 5057before attempting to use the ext2 ioctls. Some device drivers 5058unfortunately respond to the ext2 ioctl's with unknown behaviour. 5059(Addresses Debian Bug #152029). 5060 5061The extended attribute handling has been updated to correspond with 5062the latest V2 bestbits ACL code. 5063 5064Fixed bug in e2fsck which caused it to not clear the dtime field when 5065processing truncated inodes on the orphan list. This could cause data 5066loss(!) if a filesystem is rebooted before a truncate has been 5067committed. 5068 5069E2fsck now uses red/black trees in pass1b, which removes some O(n**2) 5070algorithms. This makes e2fsck much faster in the case of severely 5071corrupted filesystems where a large number of blocks are claimed by a 5072large number of inodes. (Thanks to the 2.5 IDE device driver for 5073inspiring this work.) 5074 5075Resize2fs has been significantly sped up when shrinking and expanding 5076a filesystem by a very small number of blocks (for example, when EVMS 5077is converting a partition to be an EVMS legacy volume). 5078 5079Added a new option to e2fsck, -D, which will optimize or compress all 5080of the directories in the filesystem. 5081 5082E2fsck now catches SIGINT and SIGTERM to make sure it will can 5083properly clean up and only exit at safe points. Fsck will pass 5084SIGINT/SIGTERM to its child processes, and wait until they have all 5085exited before it exits. 5086 5087The uuid parsing code in the uuid library now properly complains when 5088an illegally formated uuid is presented to it. (Addresses Debian bug 5089#152891) 5090 5091Restrict use of the 2.4 setrlimit ABI f*ckup to kernels between 2.4.10 5092and 2.4.17, since the workaround can cause problems when using a 2.4 5093kernel with an old version of glibc built with the 2.2 headers. 5094 5095Fixed a bug in mke2fs where it wasn't properly clearing the initial 5096superblock used by other filesystems. (Addresses Debian bug #147256.) 5097 5098Added support for the synchronous directory feature written by Andrew 5099Morton. 5100 5101The debugfs program can delete directories using the rmdir command. 5102 5103Add support for 8k and 16k filesystems (for systems with page sizes 5104that are greater or equal to 8k or 16k, respectively). Note that 5105these filesystems can not be mounted on x86 systems, or other systems 5106with only 4k page sizes, due to limitations in the current Linux VM 5107code. 5108 5109Resize2fs requires that the filesystem state be valid and have no 5110errors; otherwise, e2fsck -f must be run first. (Previously it simply 5111required that the last fsck time be greater than the last mount time.) 5112 5113Configure now defaults the man pages directory to /usr/share/man on 5114Linux systems. 5115 5116E2fsck now offers to truncate inodes which contain too many blocks (so 5117that i_blocks would overflow. Also fixed handling of large sparse 5118files. 5119 5120E2fsck now more completely checks for symlink validity, including 5121requiring NULL termination and length checks. 5122 5123E2fsck will offer to try forcing a disk write to remap a bad block 5124after finding a read error when reading a filesystem metadata block. 5125 5126Fixed a bug in debugfs which caused the -b and -s options to crash 5127debugfs, as well as breaking the testb, setb, and clearb functions. 5128 5129Added a bmap command to debugfs which calculates the logical to 5130physical block mapping for a particular inode. 5131 5132Fixed a bug in code which checked to see if a device was mounted which 5133sometimes (rarely) failed in the case of a plain file. 5134 5135Fixed a bug in resize2fs where when it reported an error, it would 5136print a message erroneously indicating that the filesystem had been 5137resized before it aborted. 5138 5139When resizing a plain file which is smaller than the requested size, 5140resize2fs will attempt to extended the file so that programs like 5141e2fsck will not complain that the file is too small. 5142 5143Resize2fs will print the actual new size of the filesystem when it is 5144finished resizing. 5145 5146Fixed a bug in debugfs where "ls -l" would report incorrectl file type 5147information on big-endian systems. 5148 5149 5150Programmer's notes: 5151------------------- 5152 5153Fixed collisions with C++ reserved words. 5154 5155Added portability fixes for building e2fsprogs on the HURD and AIX. 5156 5157Added the ext2ed program for creation of test cases. (ext2ed has many 5158limitations and bugs which make it unsuitable for production use.) 5159 5160The ext2fs_read_dir_block2 and ext2fs_write_dir_block now take a new 5161flag, EXT2_DIRBLOCK_V2_STRUCT, which will will reverse when the 5162name_len field is byte swampped on big-endian machines, since in the 5163V2 structure, name_len is a char field which is doesn't need to be 5164byte swapped --- except if an old-style kernel had byte-swapped the 5165name_len field as part of the V1 structure. 5166 5167 5168E2fsprogs 1.27 (March 8, 2002) 5169============================== 5170 5171The warning messages for mke2fs now go to standard error. 5172 5173Fixed to make sure "make check" runs all of the test programs with 5174LD_LIBRARY_PATH set, so that we test the libraries in the build tree. 5175 5176The mke2fs program checks the boot sector for the BSD disk label, and 5177avoids erasing it if it is there. 5178 5179Fixed a bug in e2fsck which caused it to core dump if the journal 5180inode was missing when it was supposed to be there. 5181 5182E2fsck now prints ranges in pass 5 when printing deltas for the block 5183and inode bitmaps. 5184 5185Debugfs's "ls -l" command now will print out the file type information 5186in the directory entry. 5187 5188Create man pages and hard links for fsck.ext3 and mkfs.ext3. If 5189mke2fs is invoked as mkfs.ext3, create the filesystem with a journal. 5190 5191Debugfs can now examine the experimental directory indexing 5192information. 5193 5194Fixed bug in debugfs which caused it to core dump if modify_inode is 5195called without an open filesystem. 5196 5197The debugfs lsdel command now runs its output through a pager first. 5198 5199When installing manual pages, remove the compressed manual pages first. 5200 5201Synchronized with Debian's packaging information for e2fsprogs-1.26-1. 5202 5203Fix the 2.4 resource limitation workaround introduced in 1.26 which 5204actually broke things on mips32, sparc32, and Alpha platforms. 5205 5206Updated the I18N code so that calls to setlocate(LC_CTYPE, 0) are made 5207(which is required by the newer libintl libraries). 5208 5209Programmer's notes: 5210------------------- 5211 5212Fixed various gcc -Wall complaints. 5213 5214Fixed a few memory leaks in the e2fsck journaling code, and in the 5215ismounted code checking for a swap device. 5216 5217Add new inode I/O abstraction interface which exports an inode as 5218an I/O object. 5219 5220Exported ext2_file_flush as a public interface. 5221 5222In ext2_file_write, we now mark the buffer void * argument as a const, since ext2_file_write doesn't modify the buffer. 5223 5224Lots of small random portability fixes to make e2fsprogs build under 5225AIX --- even without the 5L compatibility toolkit, and even using the 5226uber-crippled AIX native C compiler. 5227 5228 5229E2fsprogs 1.26 (February 3, 2002) 5230================================= 5231 5232Dumpe2fs will keep going now if the bad block inode can't be read. 5233(Previously it stopped with a fatal error.) 5234 5235E2fsck will now give an opportunity to run the journal even if the 5236recovery flag is not set. This is the default behaviour if e2fsck is 5237using a backup superblock, since the needs_recovery flag will never be 5238set in the backup superblock copies. 5239 5240E2fsck now automatically finds the backup superblock/group descriptors 5241even when the primary superblock is completely destroyed for 2k and 4k 5242filesystems. (Previously it just guessed/assumed that we were dealing 5243with a 1k filesystem in that case, and users had to manually specify 5244the backup superblock number.) 5245 5246Fixed a bug in e2fsck where it dereferences a null pointer when there 5247is a problem opening a filesystem in preen mode. 5248 5249E2fsck now handles external journals correctly; previously it trashed 5250the external journal device if the journal needed to be replayed. 5251 5252Work around ulimit incompatibility problem caused by recent 2.4 5253kernels; the unix IO module will automatically try to set any resource 5254limits to be infinite on startup. 5255 5256Fixed a bug in e2fsck where it wasn't allocating a new block or inode 5257bitmap if it wasn't present and the blocksize was 2k or 4k. 5258 (Addresses Debian Bug #116975) 5259 5260E2fsck will check and fix botched ext3 V1 to V2 superblock updates by 5261clearing the new V2 fields if they do not make sense or if the ext3 5262superblock is version 1 superblock. 5263 5264E2fsck will automatically relocate the ext3 journal from a visible 5265file (i.e., /.journal) to an hidden inode if the filesystem has been 5266opened read/write. This allows the users to add a journal while the 5267filesystem is mounted, but the next time the system is rebooted, the 5268journal file will disappear. This avoids problems with backups, 5269stupid operators with superuser bits, etc. 5270 5271Fix a bug in fsck where it would not support a filesystem type of 5272"auto" if the device was specified in terms of LABEL=xxx or UUID=xxx. 5273 5274Fsck now supports fstab entries of the form "ext3,ext2". It will also 5275automatically identify reiserfs filesystems. 5276 5277The number of processes spawned by fsck can now be limited using the 5278FSCK_MAX_INST environment variable. 5279 5280Fsck now searchs the LVM proc hierarchy to find logical volumes which 5281should be searched for UUID and label information. 5282 5283Work around a bug in 2.4.10+ kernels by trying to unset the filesize 5284limit if at all possible, when opening a block device. (The filesize 5285limit shouldn't be applied against writes to a block device.) 5286 5287In mke2fs and e2fsck, specifying the -c option twice will now do 5288a read/write test on the disk. Update the man pages to encourage 5289using the -c option, and to discouraging running badblocks separately, 5290since users tend to forget to set the blocksize when running 5291badblocks. 5292 5293Mke2fs now automatically clears the immutable attribute on a 5294pre-existing .journal file left over from a previous failed attempt to 5295add a journal to an alreadyy-mounted filesystem. 5296 5297Fixed mke2fs's exit codes to consistently indicate when the mke2fs 5298operation failed. 5299 5300Mke2fs now creates the lost+found directory with permissions of 0700, 5301so that files that had previously lived in protected directory are 5302safe if they get dropped in lost+found and the system administrator 5303doesn't deal with immediately. (Addresses Debian #bug 118443) 5304 5305Mke2fs and e2fsck (and all programs that use the 5306ext2fs_check_if_mounted function) will now properly identify that a 5307device is mounted, even in cases where devfs has confused things with 5308multiple devices names with the same identity, or if a dim-witted 5309system administrator has done something stupid like creating device 5310file aliases in their home directory. Also now checks for swap devices 5311by using /proc/swaps. (Addresses Debian bug #106622) 5312 5313Added a new option (-T) to tune2fs which sets the time a filesystem 5314was last checked. 5315 5316Speed up e2image when creating sparse raw image files by optimizing 5317away excess lseek() system calls. 5318 5319Fix support of large (> 2GB) files when used as a filesystem in 5320mke2fs, tune2fs, debugfs, and findsuper. 5321 5322Debugfs's stat and icheck commands now properly deals with large (> 53232GB) files. 5324 5325Debugfs's set_super_value command now prints out the list of valid 5326superblock fields which can be set using the command. 5327 5328Debugfs's rm and kill_file command now updates the superblock free block 5329and inode counts, thus keeping the filesystem consistent. 5330 5331Debugfs's lsdel command now takes an optional argument which allows 5332the user to only see the most recently deleted files. 5333 5334A new command (undel) was added to debugfs which automates 5335undeleting a deleted inode and linking it back to a directory. 5336 5337Debugfs's ls command now takes a new option, -d, which lists 5338deleted directory entries. 5339 5340Debugfs's testb, freeb, setb, and find_free_block commands now take an 5341optional count argument. 5342 5343Add support for a new ext2 file attribute, EXT2_NOTAIL_FL, which will 5344be used to signal that a particular inode should not be eligible for 5345tail-merging --- this is needed for programs like LILO. 5346 5347The findsuper (an unreleased, uninstalled utility program) has been 5348improved to add extra validity checks and to add a progress meter. 5349(It is still an unsupported program; the officially supported way to 5350recover from a trashed partition table is to use gpart; findsuper is 5351for wizards only. :-) 5352 5353Debugfs was fixed to compile with "configure --disable-swapfs". 5354 5355Cleaned up various manual pages. (Addresses Debian bug #119624, #120171) 5356 5357Added new translation file for Turkish. 5358 5359Programmer's notes: 5360------------------- 5361 5362Fix general gcc -Wall complaints. 5363 5364E2fsprogs (mostly) works with the dietlibc. 5365 5366The programming texinfo file has been expanded to include description 5367of additional libext2fs functions. (Still not compeltely done, but a 5368lot of the more important functions have been documented.) 5369 5370Added a umask structure to struct_ext2_filsys, which currently only 5371modifies the behaviour of ext2fs_mkdir(), but if we add any file 5372creation functions to libext2fs, we should also make sure they respect 5373the umask setting. 5374 5375The build-rpm script was fixed to be a bit more safe. 5376 5377The tests' Makefile now has a way of automating test case creation 5378for e2fsck, using "make testnew". 5379 5380Created a new function, ext2fs_dir_iterate2 which passes more 5381information to the callback function (identical to the one used by 5382ext2fs_dblist_dir_iterate). The directory iterator functions take a 5383new flag, DIRENT_FLAG_INCLUDE_REMOVED, which will return deleted 5384directory entries. If the directory entry is deleted, the callback 5385function will be called with the entry paraemter set to 5386DIRENT_DELETED_FILE. 5387 5388Added new functions, ext2fs_inode_alloc_stats and 5389ext2fs_block_alloc_stats, which takes updates block/inode allocation 5390statistics in the bitmaps, block group descriptors, and superblock 5391when inodes or blocks are allocated or deallocated. 5392 5393 5394E2fsprogs 1.25 (September 20, 2001) 5395=================================== 5396 5397This is primarily a bug-fix release; no new features were added, but 5398there are a number of embarassing bug fixes and cleanups applied. 5399 5400Fix a bug mke2fs which causes stack garbage to be written to disk when 5401zapping disk sectors. (This bug could cause mke2fs to core dump for 5402some kernels, I suspect with security enhancement patches.) 5403 5404Remove unneeded #include of <linux/config.h> which was breaking 5405building e2fsprogs on the Hurd. (Addresses Debian bug #112414.) 5406 5407Updated tune2fs man page to reflect the fact that adding or removing a 5408journal doesn't require running e2fsck. 5409 5410Remove use of AC_REQUIRE from autoconf which had been used to prevent 5411AC_CANONICAL_HOST from being called twice; unfortunately this causes 5412recent autoconf to bomb out since they don't allow AC_REQUIRE to be 5413used outside of autoconf macros. Fortunately, it doesn't seem to be 5414necessary any more, anyway. 5415 5416E2fsck will now not fall back to an alternate superblock if the user 5417specifies the superblock location explicitly on the command-line. 5418This allows easier recovery from a situation where the primary 5419superblock and block groups are slightly corrupted, but the backup 5420superblocks are completely broken. 5421 5422Fix problem which caused compile_et and mk_cmds to blow up if 5423e2fsprogs was in a directory with a pathname that contained spaces. 5424 5425We are now more paranoid about checking the entry in /etc/mtab to make 5426sure the filesystem is really mounted, since some broken distributions 5427(read: Slackware) don't initialize /etc/mtab before checking non-root 5428filesystems. (Slackware also doesn't check the root filesystem 5429separately, and reboot if the root filesystem had changes applied to 5430it, which is dangerous and broken, but there's nothing I can do about 5431that.) 5432 5433Make UUID library C++ friendly by adding appropriate extern "C" 5434declarations and using const in the function declarations. 5435 5436Fix up the com_err texinfo file so that it can product a valid info 5437file (previously, it could only be used to generate paper 5438documentation using texinfo.tex). 5439 5440 5441E2fsprogs 1.24a (September 2, 2001) 5442=================================== 5443 5444Fix brown-paper bug in mke2fs which caused it to segfault when 5445printing the version string. 5446 5447 5448E2fsprogs 1.24 (August 30, 2001) 5449================================ 5450 5451Revert the BLKGETSIZE64 support, since for some distributions, this 5452ioctl has been used by an unofficial kernel patch to update the last 5453sector on the disk, and this was causing disk corruption problems as a 5454result. 5455 5456Mke2fs is now more careful about zapping swap space signatures and 5457other filesystem/raid superblock magic values so that programs like 5458mount who try to use hueristics to determine the filesystem type of a 5459partition are less likely to be confused. 5460 5461E2fsck will no longer complain if the the mode of EXT2_RESIZE_INO is a 5462regular file (since Andreas Dilger's on-line resizing tools will set 5463its mode bits). 5464 5465Fixed some minor texinfo, man pages nits for spelling errors, 5466texinfo warnings, etc. (Addresses Debian bug #110621.) 5467 5468E2fsprogs program no longer print the filesystem version number 5469(i.e. 0.5b), since it only confuses people and doesn't serve any real 5470purpose. 5471 5472E2fsck will once again compile under libc5, since it will supply its 5473own version of strnlen if necesssary. 5474 5475mke2fs and tune2fs will allow the use of UUID= or LABEL= specifiers 5476when specifying the external journal device. tune2fs will also search 5477devices looking for the externla journal debice when removing. 5478 5479E2fsprogs 1.23 (August 15, 2001) 5480================================ 5481 5482Add initial support for extended attributes (EA); e2fsck will 5483correctly handle a filesystem with EA's and check the EA blocks for 5484corruptions. 5485 5486E2fsck's symlink sanity checking has been cleaned up. It now checks 5487the i_size value of fast symlinks, and checks for immutable flags 5488being set of symlinks, etc. 5489 5490E2fsck now offers to clear inodes which are special files that have a 5491non-zero i_blocks or i_size field. (The i_size field check was in the 5492previous version of the code, but due to a bug it didn't offer to 5493clear the inode unless i_size and i_size_high were both non-zero.) 5494 5495E2image can now create "raw" image files, which only contain the 5496filesystem metadata placed in a spare file so that e2fsck, dumpe2fs, 5497debugfs, etc., can be run directly on the raw image file. 5498 5499Add support for the 64-bit block device patches. 5500 5501Fixed bugs in creating external journals with a 1k blocksize. 5502 5503Add initial support for external journals (so long as the external 5504journal only supports a single filesystem) in e2fsck. 5505 5506Remove requirement for needing to run fsck on a filesystem after 5507removing a journal (either internal or external). 5508 5509The man pages now document how to create and manage external journals. 5510 5511Speed up the check of an ext3 filesystems by avoiding a needless flush 5512of all of the superblock and block group descriptors. 5513 5514Speed up creating an internal journal using tune2fs in the case where 5515the filesystem has a lot blocks already allocated. 5516 5517Tune2fs has been fixed to make sure that only error messages go to 5518stderr, and normal message go to stdout. (Addresses Debian bug #108555) 5519 5520Fixed a minor bug in mke2fs; if -O none is passed to mke2fs, it will 5521now not set the sparse_super feature. (Addresses Debian bug #108165) 5522 5523Add support in fsck for the filesystem type "auto". 5524 5525Fsck -A will not try to interpret device names for filesystems which 5526have a pass number is 0. (Addresses Debian bug #106696). 5527 5528Fsck prints a warning message if now valid filesystems are passed to 5529it. (Addresses Debian Bug #107458.) 5530 5531E2fsck now gives an explicit warning if there filesystem still has 5532errors at the end of the run. (Addresses Debian bug #104502) 5533 5534E2fsck will set the EXT2_ERROR_FS flag if the journal superblock 5535reflects an error. E2fsck will also not run the orphan list if the 5536filesystem contains errors, since the orphan list might be corrupted. 5537 5538E2fsck now prints the number of large files when given the -v option. 5539 5540Fixed minor memory leaks in e2fsck. 5541 5542Some minor man pages updates. (Addresses Debian bug #30833, #108174) 5543 5544 5545E2fsprogs 1.22 (June 22, 2001) 5546============================== 5547 5548Fixed a bug in e2fsck's handling of orphan inodes which are special 5549files (i.e., block/character device files, named FIFO's, etc.). 5550 5551Fixed a bug accidentally introduced in e2fsprogs 1.21 which caused 5552tune2fs to fail at adding a journal to a mounted filesystem. 5553 5554Fixed a few big-endian bugs in e2fsprogs 5555 * The directory block functions were accidentally reporting 5556 some directories as corrupted when they weren't. 5557 * If e2fsprogs is compiled --disable-swapfs, the C language 5558 equivalents weren't being included for big-endian platforms. 5559 (Fixes Debian bug #101686). 5560 5561Fixed a Hurd compilation problem. (Addresses Debian bug #101361) 5562 5563Programmer's notes: 5564------------------- 5565 5566Use platform independent method of defining the BLKFLSBUF and FDFLUSH 5567ioctls, and allow them to be defined for non-i386 platforms. 5568 5569The uuid.h header file is now protected against multiple inclusions. 5570 5571E2fsprogs is now being developed using BitKeeper. Changed the test 5572scripts to deal with BK's stripping CR characgters from text files, 5573and changed the top-level Makefile.in to avoid including BitKeeper 5574files when generating the source tarball. 5575 5576 5577E2fsprogs 1.21 (June 15, 2001) 5578============================== 5579 5580Added new configure flags which allow a subset e2fsprogs to be built; 5581this is most useful for boot floopies, since the resulting shared 5582libraries and programs are slimmed down by removing features that 5583aren't necessary for a boot floppy. The new flags that were added 5584are: --disable-swapfs, which removes support for byte swapping old 5585legacy PPC/68k filesystems, --disable-debugfs, which removes support 5586for debugfs from the libext2fs library, --disable-imager, which 5587removes support for the e2image program, and --disable-resizer, which 5588removes support for resize2fs. 5589 5590E2fsck now prints the number of mounts or days elapsed since the last 5591check when e2fsck is forced to check an otherwise clean filesystem. 5592 5593Tune2fs now prints an informative message about how often a filesystem 5594will be checked when adding a journal to the filesystem, to remind the 5595user that he/she may want to adjust those parameters using tune2fs 5596-c/-i. 5597 5598Worked around hurd brain-damage which causes e2fsck to sometimes 5599believe a filesystem is the root filesystem based on device numbers 5600(since Hurd doesn't have dev_t's, which is arguably a POSIX.1 5601violation). 5602 5603Fixed a bug introduced in 1.20 which caused e2fsck to abort with an 5604erroneous error with the -F option was specified. 5605 5606Fixed a ext3 recovery bug in the revoke handling; synchronized with 5607ext3 0.7a. 5608 5609Fixed two bugs in e2fsck's handling of dup block handling, dealing 5610with relatively uncommon edge cases: a directory with an indirect 5611block which is claimed by another file, and when the last inode in the 5612filesystem has blocks claimed by another file. 5613 5614E2fsck now checks to see if the i_size field of a fast symlink is too 5615big, and offers to clear the symlink if so. 5616 5617E2fsck now checks to see if i_size_high of special files is non-zero, 5618and offers to clear i_size_high. 5619 5620Fix e2fsck's handling of incompatible journal flags so that the user 5621has chance to abort, and then has the option to clear out the journal 5622entirely. (Addresses Debian bug #98527.) 5623 5624Fixed a bug in fsck which could cause it to core dump if a mix of 5625standard and non-standard device names are used in /etc/fstab. 5626(Debian bug #100559) 5627 5628Fixed a bug in debugfs which caused read errors when copying a file to 5629not be noticed. 5630 5631The debugfs set_super_value command can now modify the s_lastcheck field. 5632 5633Fixed a bug in lsattr and chattr which was accidentally introduced in 56341.20 to support > 2GB files; both lsattr and chattr wasn't reading 5635directories correctly because the change modified the layout of struct 5636dirent to be incompatible with the libe2p shared library. 5637 5638Cleaned up the mke2fs manual page and included a discussion about why 5639it's good to periodically check the filesystem even when journaling is 5640enabled. 5641 5642Programmer's notes: 5643------------------- 5644 5645Fix general gcc -Wall complaints. 5646 5647The types needed by the ext2 header files are now provided by 5648lib/ext2fs/ext2_types.h, instead of include/asm/types.h. 5649 5650Integers are now preferred to longs when trying to find a 32-bit type 5651in ext2_types.h. Also, if linux/types.h has already been defined, 5652don't try to redefine the types. 5653 5654Fixed make depend script so that it automatically corrects the 5655pathname cleanups performed by make -M, so I don't have to fix them up 5656by hand. 5657 5658Fixed the d_loaddump test case to be more robust, and not depend on 5659bash'isms. 5660 5661Removed debugfs's dependence on pread(), which was accidentally 5662intrudced in e2fsprogs 1.20 5663 5664Fixed a performance bug in the libext2fs's icount routine; the size 5665estimate of the icount array was incorrectly being calculated. 5666 5667Removed use of the badblocks compatibility functions in the e2fsprogs 5668programs. 5669 5670Added paranoia code which protects against strange cases where /etc 5671isn't on the root filesystem, or if /etc/mtab doesn't exist. 5672 5673The header file ext2_types.h is now installed. 5674 5675Autoconf is used to determine when we are on big-endian machines, 5676instead of doing run-time tests, to save a few bytes of code. 5677 5678The ext2fs_mark_generic_bitmap and ext2fs_unmark_generic_bitmap 5679functions are no longer inline functions, which saves space and 5680doesn't really cost any real performance. 5681 5682The ext2fs library no longer depends on the e2p library. (What need 5683there was of it --- namely, fsetflags, was coded in-line). 5684 5685Fixed the makefile so that lib/ext2fs/ext2_types.h is generated even 5686when the user is stupid and tries compiling the package using "make 5687install" as root. 5688 5689Miscellaneous code cleanups: 5690 * Added missing files from Makefile.in's SRCS file, so that 5691 their dependencies would be properly calculated. 5692 * Removed redundant code 5693 * Fixed comments in code 5694 * Removed no-longer unneeded argsused #pragma. 5695 5696 5697E2fsprogs 1.20 (May 20, 2001) 5698============================= 5699 5700Add support for replaying the ext3 journal purely in user mode, 5701including handling the orphaned inode list. Used code contributed by 5702Andreas Dilger, with bug fixes and the orphaned inode handling done 5703by Theodore Ts'o. 5704 5705The mke2fs and tune2fs programs can create or modify a filesystem to 5706include an ext3 journal. Tune2fs also can be used to remove an ext3 5707journal from a filesystem. 5708 5709E2fsck will now check for the existence of a linked list of orphan 5710inodes in the superblock, and clear those inodes before starting the 5711rest of the filesystem check (but after the journal playback). 5712 5713E2fsck now validates the file descriptor passed to the -C option, 5714which saves against the completion bar getting written to an 5715unexpected location, such as the disk being checked. (Debian 5716bug/wishlist #55220) 5717 5718E2fsck will now bump the filesystem revision number from zero to one 5719if any of the compatibility bits are set. 5720 5721Fixed a bug where a badly corrupted directory on a big endian system 5722could cause e2fsck to die with a bus error. The 5723ext2fs_read_dir_block() and ext2fs_process_dir_block() functions in 5724the ext2 library now does alignment sanity checks on the rec_len field 5725of the directory entry before using it. 5726 5727The ext2 library has been enhanced to make tune2fs safe to run on 5728mounted filesystems. (Users could usually get away with using tune2fs 5729on mounted filesystems before, but with the advent of ext3 and 5730journaling, it became important to make tune2fs was *really* safe for 5731use even when the filesystem being modified is mounted.) E2label is 5732now implemented by tune2fs using an argv[0] dispatch, so that e2label 5733is also now safe for use on mounted filesystems. 5734 5735Added a new program, e2image, which creates a backup of critical ext2 5736filesystem data structures. The generated image file can be examined 5737using dumpe2fs and debugfs. In the future, e2fsck will be able to use 5738the image file to help recover very badly damaged filesystems. 5739 5740Fixed a number of LFS bugs in e2fsck; very, very large (> 2**42) files 5741no longer cause e2fsck to bomb out. Also treat files > 2GB as being 5742large file when deciding whether or not the filesystem has large files. 5743 5744Fixed lsattr and chattr so that they work correctly on large files. 5745(Fixes Debian bug #72690.) 5746 5747Removed limitation in get_device_size() which imposed a one terrabyte 5748filesystem limitation. (Most 2.2 kernels still have a signed int 5749problem which cause 1 TB block device limitation. Fortunately, the 5750kernel patches to fix this are much easier than fixing the 2TB 5751limitation in the kernel. :-) 5752 5753A max_mount_count of zero is now treated as if no mount count were 5754set. (Previously, no mount count was indicated by using -1, and a 5755mount count of zero caused e2fsck to always be run.) 5756 5757Mke2fs supports two new filesystem types largefile and largefile4. 5758 5759Mke2fs now adds some randomness in s_max_mount_count so that multiple 5760filesystems won't be all checked at the same time under normal 5761operations. 5762 5763Fixed bug in the progress bar printing code which could cause e2fsck 5764to core dump on an illegal filesystem. 5765 5766Fixed bug in fsck which could allow more than one instance of e2fsck 5767to be printing a progress bar. (Debian bug #65267) 5768 5769Fsck using a UUID or a LABEL specifier will work even if devfs is 5770compiled into the kernel and not mounted. If the pathnames in 5771/proc/partitions are incorrect, fsck will search /dev for the correct 5772device (using the new ext2fs_find_block_device library function). 5773Fsck now also checks the RAID devices first so that they are properly 5774found when they are in use. Support has also been added to support 5775additional IDE disks and the DAC 960 device names. (Debian bug #94159) 5776 5777Fixed a bug in fsck which caused it not deal properly with 16 5778byte long filesystem labels. 5779 5780Fsck's -t option has been made a lot more flexible. The semantics for 5781what happens if a comma-separated list to fsck has been regularized, 5782and it is now possible to filter what filesystems will get checked 5783based what is in the filesystem's fstab entry's option field. (Debian 5784bug #89483.) 5785 5786The dumpe2fs program can now print out the group description 5787information in hex, and also prints the location of the superblock and 5788block group descriptor for those block groups that have them. 5789 5790Mke2fs now clears the ext2 superblock before it starts creating a 5791filesystem, so that the superblock magic number is only written if the 5792filesystem creation process successfully completes. 5793 5794The debugfs program's stat command now pretty-prints the blocks used 5795by an inode so that it's more compact and informative. 5796 5797The debugfs stats command now uses the same libe2p code (which is used 5798by dumpe2fs) to print the superblock header information. This is more 5799complete, and it avoids a bit of code duplication. 5800 5801Added a new debugfs command, set_super_value (ssv) which allows the 5802user to set arbitrary superblock fields. 5803 5804Debugfs was extended to support inode numbers in hex (by prefixing 5805them with 0x), and so that modify_inode can set the inode generation 5806number. Also, there is now a new function command called logdump 5807which will dump an ext3 journal. 5808 5809Fixed a bug in debugfs so that quitting out of the pager doesn't kill 5810debugfs. 5811 5812Debugfs's dump command now stops immediately upon reporting a disk 5813read error. (Fixed a bug in ext2fs_file_read library routine which 5814caused debugfs not to stop.) (Debian bug #79163) 5815 5816On systems with /proc/mounts (mainly Linux systems), /proc/mounts is 5817checked before /etc/mtab is used, since /proc/mounts is more likely to 5818be accurate. 5819 5820Added portability fixes for Solaris and Linux/ia64. 5821 5822Various manual pages were clarified and cleaned up. (Fixed debian 5823bugs #63442, #67446, and #87216) 5824 5825 5826Programmer's notes: 5827------------------- 5828 5829The e2fsck message printer now supports %Iu and %Ig, which will print 5830out the inode's user and group owners, respectively. 5831 5832E2fsprogs now includes its own version of include/linux/ext2_fs.h, so 5833that no longer dependent on the system having the correct version of 5834the kernel header files. 5835 5836Added a new function to libext2, ext2fs_find_block_device(), which 5837searches the system (i.e., /dev, /devfs, /devices) for a pathname to a 5838device given its device number. 5839 5840Added a new function to libext2, ext2fs_sync_device, which centralizes 5841all of the places which might try to use the BLKFLSBUF or FDFLUSH 5842ioctls (and usually failing to define them since the system header 5843files don't usually do this for us, and we're trying to avoid usage of 5844kernel include files now). 5845 5846Added new utility programs in tests/progs: random_exercise and 5847hold_inode. They aren't built by default; they're useful for 5848exercising ext3 filesystem code. 5849 5850Added a new ext2 filesystem flag, EXT2_FLAG_SUPER_ONLY, which causes 5851the filesystem close functions to only update the superblock, and to 5852not touch the block group descriptors. Needed by tune2fs when 5853modifying a mounted filesystem. 5854 5855Got rid of struct ext2fs_sb and replaced it with the standard struct 5856ext2_super_block from include/linux/ext2_fs.h. Note: this may break 5857source (but not binary) compatibility of some users of the ext2 5858library. Those applications should just simply do a global search and 5859replace of struct ext2fs_sb with struct ext2_super_block, and include 5860the new header file <ext2fs/ext2_fs.h> which defines it. 5861 5862The ino_t type has been renamed ext2_ino_t to protect applications 5863that attempt to compile -D_FILE_OFFSET_BITS=64, since this 5864inexplicably changes ino_t(!?). So we use ext2_ino_t to avoid an 5865unexpected ABI change. 5866 5867The Makefiles have been reworked so that "make check" can be run from 5868the top-level Makefile. 5869 5870Fix general gcc -Wall complaints and removed dead code. 5871 5872Remove use of NOARGS, because we assume everyone does ANSI C these 5873days. 5874 5875Added build-rpm script from sct. 5876 5877New functions ext2fs_image_{inode,super,bitmap}_{read,write} added 5878to support e2image. 5879 5880New function ext2fs_flush_icache which must be called if the 5881application program modifies the inode table blocks without going 5882through ext2fs_write_inode() interface. 5883 5884New ext2fs_check_mount_point() function, which will return the mount 5885point of a device if mounted. 5886 5887The io_channel abstraction now has an optional interface, 5888io_channel_write_range, which allows specific byte ranges to be 5889written. 5890 5891The unix_io IO channel now supports write-through caching, so that 5892journal creation is more efficient. 5893 5894Added x86 assembly language routines to support byte swapping, to 5895reduce executable size. 5896 5897Fixed bug in the utility program subst so that it's possible to 5898replace a substitution variable with a zero-length string. 5899 5900Fixed numbering e2fsck pass1 problem numbers; an extra zero had 5901slipped into some of the problem number. 5902 5903E2fsprogs 1.19 (July 13, 2000) 5904============================== 5905 5906Release the resize2fs program since the timeout before it could 5907be released under the GPL has finally expired. 5908 5909Add experimental support needed for the ext2 compression patches. 5910This requires compiling e2fsprogs with the --enable-compression flag 5911to the configure script. 5912 5913Added ext3 journalling support. E2fsck will run the journal (if 5914necessary) by temporarily mounting the filesystem. /sbin/fsck.ext3 is 5915installed as a symlink to e2fsck. Fsck has been taught about ext3, 5916and treats it the same as ext2 in terms of the progress bar logic. 5917Dumpe2fs will display the superblock journaling information if the 5918filesystem has a journal. The ext2 library will now permit opening an 5919ext3 filesystem with the recovery flag set. This is necessary for 5920on-line dump's to work correctly, but there may be issues with this 5921working well since ext3 is much less agressive about syncing blocks to 5922the filesystem, since they're safe on the journal. 5923 5924Tune2fs and e2fsck have been changed to allow the mount_count check to 5925be disabled by setting max_mount_count to -1. (This was already 5926supported by the kernel.) 5927 5928Create a symbolic link for fsck.ext3, since the e2fsprogs utilities 5929are used for ext3 as well. 5930 5931Added internationalization support for e2fsprogs; must be enabled 5932by passing --enable-nls to configure. 5933 5934Always use the provided ext2fs header files to insulate ourselves from 5935kernel version changes. Which include files are used by e2fsprogs 5936have also been cleaned up to improve portability. 5937 5938Limit the number of times that e2fsck updates the progress bar so that 5939people who are booting using a 9600 baud console don't get swampped by 5940too many updates. 5941 5942Improved the loop detection algorithm in e2sck's pass #3 so that it is 5943much, much faster for large filesystems with a large number of 5944directories. 5945 5946The memory footprint for e2fsck is now slightly smaller than before. 5947 5948E2fsck now checks if special devices have a non-zero size, and offers 5949to clear the size field if it finds such an inode. 5950 5951E2fsck now checks if special devices have the append-only flag set, 5952and offers to clear the inode. 5953 5954E2fsck now properly handles some "should never fail" cases during a 5955bitmap copy in pass5. 5956 5957E2fsck now properly prints control characters in filenames as ^A .. ^Z. 5958 5959E2fsck now calculates the correct location of the backup superblock in 5960the case of filesystem blocksizes > 1k. 5961 5962Fixed a bug in e2fsck's calculation of the number of inodes_per_block 5963which normally didn't cause problems under most filesystem parameters, 5964but could cause a valid superblock to be rejected in extreme cases. 5965Other checks for validating superblock values were made more 5966stringent. 5967 5968Added non-destructive write testing to the badblocks program, courtesy 5969of David Beattie. The badblocks also now has an option to input the 5970current set of bad blocks, so that known bad blocks are skipped to 5971speed up the badblocks test. There is also a persistent rescan 5972feature which causes badblocks to run until it has completed some 5973number of passes without discovering any new bad blocks. 5974 5975Badblocks now checks to see if the device is mounted and refuses to do 5976the tests involving writing to the device if it is mounted. Also, 5977badblocks now allows the number of blocks to be checked to be 5978defaulted to the size of the partition. 5979 5980Fixed a bug in fsck which didn't allow non-root users to be able to 5981check filesystems if there were any LABEL= or UUID= entries in 5982/etc/fstab. 5983 5984The Hurd doesn't support the filetype filesystem feature. The mke2fs 5985program now makes sure that for the Hurd, the filestype feature is 5986turned off. E2fsck will check to see if the filetype feature is 5987turned on for Hurd filesystems, and offer to turn off the feature. 5988 5989Mke2fs now has a safety check to make sure the number of blocks do not 5990exceed 32 bits even on a 64 bit platform. 5991 5992Really fixed a bug in fsck to allow "fsck -As" to run interactive 5993fsck's. (For those people who like to do interactive fsck's in the 5994/etc/rc scripts!?!) 5995 5996Debugfs has a few new features: the rdump command, which will do a 5997recursive dump of a directory and all of its contents, and the lcd 5998command which does a local chdir (much like the ftp command of the 5999same name). In addition, the debugfs program and the open_filesystem 6000command now takes three new options: -b and -s, which allows the 6001blocksize and superblock location to be specified, and the -c option 6002which is used in catastrophic situations where the block group 6003descriptors are corrupt. If the -c option is specified, debugfs will 6004skip trying to read in the block and inode bitmaps. 6005 6006Debufs's lsdel command was fixed to handle bad blocks in the inode 6007table. 6008 6009A Y2K bug in debugfs's "ls -l" handling was fixed by switching to use 60104 digit years. 6011 6012General improvements in error messages 6013 6014 - Mke2fs prints a sane error message if the partition size is zero 6015 (usually because the partition table wasn't reread by the 6016 kernel due to the partition being busy), instead of "invalid 6017 argument passed to ext2 library while initializing superblock". 6018 6019 - Fsck now prints more self-explanatory message if an invalid UUID= 6020 or LABEL= specification is passed to it. 6021 6022UUID library changed to use the LGPL. 6023 6024Fixed a bug in the UUID library where very rapid calls to the 6025time-based UUID generator could cause duplicate UUID's to be returned. 6026This was not a problem for e2fsprogs, but it could be a problem for 6027other users of the library. 6028 6029Make the UUID library more robust in the face of missing or an 6030improper /dev/urandom or /dev/random files. 6031 6032Added some random portability fixes for Solaris. 6033 6034Some minor man page updates. 6035 6036Fixed a memory leak in the ss library. 6037 6038 6039Programmer's notes: 6040------------------- 6041 6042We now try to use lseek64 and open64 from the LFS if possible. 6043 6044The 3rd parameter in e2p's print_flags is now a flags word, instead of 6045a boolean option. 6046 6047The mark and unmark bitmap functions now return the previous state of 6048the bit that was being changed, which is useful for some speed 6049optimizations. 6050 6051The following functions have been added to enhance the badblocks list 6052handling in libext2fs: ext2fs_write_bb_FILE, ext2fs_read_bb_FILE2, and 6053ext2fs_badblocks_equal. 6054 6055The ext2 header files now have the latest journalling fields to the 6056superblock. 6057 6058The ext2fs_mkdir function in libext2fs now properly backs out of error 6059conditions robustly. 6060 6061Cleaned up makefiles: 6062 - to cleanly compile with the -j flag. 6063 - so distclean removes all generated files. 6064 - so in case of an error while installing header files, the make aborts. 6065 6066Fix test_script so that it works correctly when compiling in the 6067source directory. 6068 6069The random UUID generation routine has been made slightly better in 6070the case where /dev/random doesn't exist. (Use of randomly-based UUID 6071is still not recommended if /dev/random doesn't exist, however; it's 6072better to use the time/ethernet MAC address UUID in this case.) 6073 6074Clean up the build process so it's more friendly in case of missing 6075directories. 6076 6077The ext2fs header file can now be #include'd into C++ programs. 6078 6079The e2p.h header file is now installed. 6080 6081Added workaround to a gawk 3.0.5 bug in lib/ss/mk_cmds. 6082 6083 6084 6085E2fsprogs 1.18 (November 10, 1999) 6086================================== 6087 6088Fix a core dumping bug in e2fsck if an imagic inode is present or 6089(more rarely) if the filesystem is badly corrupted enough that e2fsck 6090has to restart pass 1 processing. E2fsck now closes the filesystem 6091before freeing a large number of its data structures, so in the case 6092of future memory faults, at least the fixed filesystem will be fully 6093written out. 6094 6095If a filesystem doesn't support imagic inodes, and e2fsck discovers an 6096imagic inode, it will offer to clear the imagic flag. 6097 6098E2fsck will now offer to clear the immutable flag on special files 6099(device/socket/fifos) when running it in non-preen mode. 6100 6101E2fsck will now set the filetype when creating /lost+found, and when 6102connected orphaned inodes to /lost+found. 6103 6104Debugfs's ncheck and icheck commands now handles the case where there 6105are bad blocks in the inode table without bombing out. 6106 6107The badblocks list processing code has been made more efficiently for 6108appending a large number of (ordered) badblocks to the badblocks list. 6109 6110Some minor man page updates. 6111 6112Fsck now allows interactive e2fsck's when using fsck -As (not a common 6113mode, but some people like to do this in boot scripts for silly reasons). 6114 6115Programmer's notes: 6116------------------- 6117 6118The internal e2fsck problem code for PR_2_SPLIT_DOT was fixed to meet 6119with the problem code convention. 6120 6121The badblocks list regression test program has been updated to work 6122with previously made API name changes. 6123 6124The ext2fs_free() command now uses the new badblocks API to avoid 6125using the compatibility layer. 6126 6127Added new regression test cases; the run_e2fsck test script now 6128supports the ability for a test case to run a prepratory command 6129before running e2fsck. 6130 6131E2fsprogs 1.17 (October 26, 1999) 6132================================= 6133 6134Fixed nasty typo in fsck which caused parallelized fsck's to go into an 6135infinite loop. 6136 6137Fixed a bug in fsck where it used strncmp to compare a binary UUID, 6138thus potentially causing problems if a binary UUID contained a NULL 6139character. 6140 6141E2fsck now uses stricter checks for directory entries in pass 2: 6142zero-length filenames are not allowed; neither are 8 byte long 6143directory entries. 6144 6145The debugfs "dirty" command now clears the filesystem valid bit. 6146(Previously this just set the dirty-as-in-needs-writing-out-to-disk 6147bit in the in-core superblock image. The new functionality is more 6148what the user expects, and is more useful.) 6149 6150Added a debugging hook to test parallel fsck; if the environment 6151variable FSCK_FORCE_ALL_PARALLEL, then filesystems on the same drive 6152will be checked in parallel when they normally would not be. 6153 6154Programmer's notes: 6155------------------- 6156 6157Fixed some #ifdef's for compilation under the Hurd OS. 6158 6159Fixed minor W2K compatibility problems. 6160 6161Fixed some miscellaneous GCC warnings. 6162 6163 6164E2fsprogs 1.16 (October 22, 1999) 6165================================= 6166 6167Fixed a race condition bug in fsck; when printing a progress bar, if 6168checking multiple filesystems in parallel, it was possible for fsck to 6169send e2fsck a SIGUSR1 signal before e2fsck had installed its signal 6170handler, which would cause it to terminate with a signal 10. 6171 6172E2fsck now properly handles filesystems that have the 6173INCOMPAT_FILETYPE feature turned on. It can be used to convert a 6174filesystem into using or not using FILETYPE feature. 6175 6176E2fsck now properly handles filesystems that have the IMAGIC feature 6177turned on (this is used on Linux AFS servers). 6178 6179The mke2fs program now creates filesystems that have the filetype and 6180sparse_superblock features enabled by default, unless it is run on a 6181pre-2.2 kernel. These features are not supported by a pre-2.2 kernel, 6182so there is now a new flag -O which allows the user to specify with 6183which features she would like to create the filesystem; "mke2fs -O 6184none" will create a filesystsem compatible with 2.0 kernels. 6185 6186The tune2fs program now has a -O option which allows the user to set 6187and reset "safe" filesystem features. Currently, the only ones which 6188allows to be modified are the filetype and sparse_superblock features. 6189Note setting or clearing either feature will require running e2fsck on 6190the filesystem afterwards. (n.b. Clearing the sparse_superblock feature 6191requires that there is enough free space on the filesystem for the 6192extra superblocks which will be created by e2fsck.) 6193 6194Debugfs can now set and print filesystem features in the superblock 6195using the "features" command. Dumpe2fs will print out the complete 6196set of features when listing the superblock. 6197 6198Dumpe2fs has new options -f (force) and -h (header-only). 6199 6200Fixed a bug in e2fsck which could cause the PROGRAMMING ERROR/bonehead 6201message to come up. This could happen when decrementing or 6202incrementing a link count could result in an overflow. 6203 6204Fixed a bug in e2fsck where the block count on the lost+found 6205directory would not be properly incremented when the directory was 6206expanded to the point where an indirect block needed to be allocated. 6207 6208E2fsck now makes some additional sanity checks on the superblock to 6209avoid crashing or giving a memory allocation error if some of the 6210values in the superblock are unresonable (but the superblock otherwise 6211looks valid). 6212 6213Fixed a bug in e2fsck where a very badly corrupted filesystem might 6214require two passes to completely fix the filesystem. This happened if 6215an inode claimed blocks that was part of the filesystem metadata 6216(typically, when garbage was written into an inode table or indirect 6217block, since this kind of filesystem corruption normally doesn't 6218happen otherwise). 6219 6220On the Alpha, glibc declares st_flags although it isn't actually used; 6221the configure script was improved to detect this case so that 6222e2fsprogs can avoid using the non-functional stat field. 6223 6224The manual pages were updated to use a more consistent formatting 6225style consistent with standard Unix man pages. Mke2fs's man page 6226added documentation for a few previously undocumented options. 6227 6228Fixed minor display bugs in tune2fs and mke2fs. 6229 6230Programmer's notes: 6231------------------- 6232 6233Improved portability of e2fsprogs to non-Unix systems (in particular, NT). 6234 6235Added features to parse and print feature strings into the e2p library. 6236(e2p_feature2string, e2p_string2feature, e2p_edit_feature). 6237 6238ext2fs_mkdir() and ext2fs_new_dir_block() now creates directories 6239whose directory entries contain proper filetype information if the 6240filesystem supports it. 6241 6242ext2fs_link() now uses the low 3 bits of its flags parameter to pass 6243the directory entry filetype information. This is used to set the 6244directory entry filetype information if the filesystem supports it. 6245 6246Fixed a bug in ext2fs_expand_dir() where the block count in a 6247directory's inode would not be properly incremented when the directory 6248was expanded to the point where an indirect block needed to be 6249allocated. 6250 6251 6252E2fsprogs 1.15 (July 18, 1999) 6253============================== 6254 6255Add configuration checks so that e2fsprogs will compile cleanly on 6256Linux 2.3 kernels that have renamed i_version to i_generation. 6257 6258E2fsck now prints a progress/completion bar (and not just a simple 6259spinner) if the -C0 option is requested or if it receives a SIGUSR1 6260signal. Fsck will automatically manage the (potentially muliple) 6261e2fsck processes to print completion bars if it is given a -C option, 6262with the right thing happening if multiple filesystems are being 6263checked in parallel. 6264 6265Mke2fs now has better automatic hueristics to determine the filesystem 6266parameters to be used for a particular filesystem. Added a new option 6267-T which allows the user to specify how the filesystem is to be used, 6268which helps mke2fs do a better job selecting the filesystem parameters. 6269 6270Mke2fs now creates revision 1 filesystems by default, and with the 6271sparse superblock feature enabled. The sparse superblock feature is 6272not understood by Linux 2.0 kernels, so they will only allow read-only 6273mounts of filesystems with this sparse superblocks. 6274 6275Fix bug where if /dev/null couldn't be opened (should never happen), 6276e2fsck would hang in a tight loop. 6277 6278Make e2fsck handle the case where /lost+found isn't a directory. 6279 6280E2fsck now uses mallinfo if it exists to get accurate statistics about 6281its memory usage. 6282 6283Fix bug in e2fsck where it wouldn't check to see if a disconnected 6284inode had any problems before connecting it to /lost+found. 6285 6286Add check to e2fsck so it makes sure that total number of inodes in 6287the filesystem is a sane number. 6288 6289Fix fencepost error when clearing an the end of the block bitmap which 6290caused the last block in the bitmap not to get cleared. 6291 6292Cleaned up a number of messages in e2fsck: 6293 * The message "Group's #'s copy of the group descriptor..." 6294 was fixed so that the correct number would be displayed. 6295 * Added missing space in the "disk write-protected" error messsage 6296 * Cleaned up the error message printed when a non-interactive 6297 e2fsck needs to abort a check because the filesystem 6298 appears to be mounted. 6299 6300Added a new command-line utility, uuidgen, which will create and print 6301a UUID. 6302 6303Make debugfs's icheck command more robust by checking to make sure an 6304inode has valid blocks before interarting over the inode's blocks. 6305 6306UUID generation now uses a random-based scheme whenever possible to 6307prevent potential privacy problems. 6308 6309Man pages for all of the UUID functions in the lirbary were added. 6310 6311Fixed bug in fsck so it won't coredump if a filesystem not in 6312/etc/fstab is given to it. 6313 6314Fsck now understands the UUID=xxxx and LABEL=yyyy forms in /etc/fstab 6315that most of the other mount utilities understands. 6316 6317Mke2fs will make a filesystem even if it appears mounted if the force 6318option is given. 6319 6320Dumpe2fs has new command-line options which allow a filesystem expert 6321to specify the superblock and blocksize when opening a filesystem. 6322This is mainly useful when examining the remains of a toasted 6323filesystem. 6324 6325The badblocks program has been updated to display correctly on disks 6326with large block numbers. 6327 6328The badblocks program no longer gives spurious errors when errors 6329occur on non-block boundaries, which is common if the blocksize is 6330larger than 1k. 6331 6332Mke2fs will sync the disk device every MKE2FS_SYNC block groups if the 6333MKE2FS_SYNC environment variable is set. This is to work around a VM 6334bug in the 2.0 kernel. I've heard a report that a RAID user was able 6335to trigger it even using a 2.2 kernel, but hopefully it will not be 6336needed for most Linux 2.2 users. 6337 6338Fixed miscellaneous documentation and man pages. 6339 6340Programmer's notes: 6341------------------- 6342 6343Cleaned up functions such as pass1_get_blocks, pass1_read_inode which 6344in e2fsck's pass1.c really should have been static. 6345 6346The return value of the uuid_compare() function was changed to make it 6347match with the convetions used by strcmp, memcmp, and Paul Leach's 6348UUID sample document. 6349 6350The "make depend" process has now been made more automated; it now 6351automatically word-wraps the dependencies, and only replaces source 6352Makefile.in if there has been a change in the dependencies. Also, a 6353top-level "make depend" now recurses through all the subdirectories 6354automatically. 6355 6356The Makefile in .../util has been changed so that subst is built using 6357the native C compiler during a cross-compilation, since the subst 6358program is only used during the build process. Also add an explicit 6359rule to build util/subst by cd'ing to the correct directory and 6360running Makefile. 6361 6362The man directories are defined in terms mandir, so that the configure 6363script can override the location of the manual pages. 6364 6365The config files have been updated to recognize new machine types for 6366both the i386 and alpha families. 6367 6368Fsck has been modified so that it will accurately create an 6369fsck_instance even when the noexecute flag is set. This allows for 6370accurate debugging of the fsck pass structure. Also, when the verbose 6371flag is given twice, fsck will print debugging information about when 6372fsck is waiting for jobs to finish. 6373 6374 6375E2fsprogs 1.14 (January 9, 1999) 6376================================ 6377 6378Fix the fstab parsing code so that it can handle blank lines and 6379comment characters. Also, missing pass numbers need to be treated as 6380zero. 6381 6382Fixed a bug in e2fsck where under some circumstances (when e2fsck 6383needs to restart processing after fixing an egregious inconsistency) 6384it would try to access already freed memory. 6385 6386E2fsck now prints non-printable characters in directory entries and 6387pathnames using '^' and 'M-' notation. 6388 6389Fixed chattr so that it will ignore symbolic links when doing 6390recursive descent traversals. For both chattr and lsattr, no longer 6391print the version string unless the -V option is given. 6392 6393Allow the system administrator to directly specify the number of 6394inodes desired in the filesystem, for some special cases where this is 6395necessary. 6396 6397Fix portability problems so that e2fsprogs can be compiled under Linux 63981.2 systems and Solaris systems. 6399 6400Update the config.guess file with a more recent version that will 6401identify newer Linux platforms. 6402 6403Programmer's notes 6404------------------ 6405 6406Ext2fs_read_inode and ext2fs_write_inode will now return an error if 6407an inode number of zero is passed to them. 6408 6409E2fsprogs 1.13 (December 15, 1998) 6410================================== 6411 6412Fixed a bug in debugfs where an error messages weren't getting printed 6413when the ext2 library routines to read inodes returned errors in the 6414stat, cmri and rm commands. 6415 6416Fixed a bug in mke2fs so that if a ridiculous inode ratio parameter is 6417provided, it won't create an inode table smaller than the minimum 6418number of inodes required for a proper ext2 filesystem. 6419 6420Fsck now parses the /etc/fstab file directly (instead of using 6421getmntent()), so that it can distinguish between a missing pass number 6422field and pass number field of zero. This caused problems for 6423diskless workstations where all of the filesystems in /etc/fstab have 6424an explicit pass number of zero, and fsck could not distinguish this 6425from a /etc/fstab file with missing pass numbers. 6426 6427E2fsck will create a /lost+found directory if there isn't one in the 6428filesystem, since it's safer to create the lost+found directory before 6429it's needed. 6430 6431Fixed e2fsck so that it would detect bogus immutable inodes which 6432happen to be sockets and FIFO files, and offer to clear them. 6433 6434If a filesystem has multiple reasons why it needs to be checked, and 6435one of the reasons is that it is uncleanly mounted, e2fsck will print 6436that as the reason why the filesystem is being checked. 6437 6438Cleaned up the output routines of mke2fs so that it doesn't overflow 6439an 80 column display when formating really big filesystems. 6440 6441Added a sanity check to e2fsck to make sure that file descriptors 0, 64421, 2 are open before opening the hard disk. This avoids a problem 6443where a broken program might exec e2fsck with those file descriptors 6444closed, which would cause disastrous results if the kernel returns a 6445file descriptor for the block device which is also used by FILE * 6446stdout. 6447 6448Fixed up the e2fsck progress reporting functions so that the values 6449reliably reach 100% at the completion of all of the e2fsck passes. 6450 6451Fixed minor documentation bugs in man pages and usage messages. 6452 6453Programmer's notes: 6454------------------- 6455 6456Fixed a number of lint warnings in the ext2fs library and potential 6457portability problems from other OS's header files that might define 6458CPP macros for names like "max" and "min". 6459 6460ext2fs_badblocks_list_add() has been made more efficient when it needs 6461to grow the bad blocks list. 6462 6463Fixed a bug in e2fsck which caused it to dereference a freed pointer 6464just before exiting. 6465 6466Fixed the substition process for generating the mk_cmds and compile_et 6467scripts so that they will work outside of the build tree. 6468 6469Add sanity check to e2fsck so that if an internal routine 6470(ext2fs_get_dir_info) returns NULL, avoid dereferencing the pointer 6471and causing a core dump. This should never happen, but... 6472 6473E2fsprogs 1.12 (July 9, 1998) 6474================================== 6475 6476E2fsprogs now works with glibc (at least with the version shipped wtih 6477RedHat 5.0). The ext2fs_llseek() function should now work even with 6478i386 ELF shared libraries and if llseek() is not present. We also 6479explicitly do a configure test to see if (a) llseek is in libc, and 6480(b) if llseek is declared in the system header files. (See standard 6481complaints about libc developers don't understand the concept of 6482compatibility with previous versions of libc.) 6483 6484The ext2fs library now writes out the block group number in each of 6485the superblock copies. This makes it easier to automatically 6486determine the starting block group of the filesystem when the block 6487group information is trashed. 6488 6489Added support for the EXT2_FEATURE_INCOMPAT_FILETYPE feature, 6490which means that e2fsprogs will ignore the high 8 bits of the 6491directory entry's name_len field, so that it can be used for other 6492purposes. 6493 6494Added support for the EXT2_FEATURE_RO_COMPAT_LARGE_FILE feature. 6495E2fsprogs will now support filesystems with 64-bit sized files. 6496 6497Added support for the EXT2_FEATURE_COMPAT_DIR_PREALLOC feature. 6498 6499Added new program "e2label", contributed by Andries Brouwer. E2label 6500provides an easy-to-use interface to modify the filesystem label. 6501 6502Fixed bug so that lsattr -v works instead of producing a core dump. 6503 6504Fixed a minor bug in mke2fs so that all groups with bad superblock 6505backup blocks are printed (not just the first one). 6506 6507Mke2fs will check the size of the device, and if the user specifies a 6508filesystem size larger than the apparent size of the device it will 6509print a warning message and ask if the user wants to proceed. 6510 6511E2fsck has a new option -C, which sends completion information to the 6512specified file descriptor. For the most part, this is intended for 6513programs to use, although -C 0 will print a spinning character to the 6514stdout device, which may be useful for users who want to see something 6515happening while e2fsck goes about its business. 6516 6517Fixed a bug in e2fsck which could cause a core dump when it needs to 6518expand the /lost+found directory, and sometimes the bitmaps haven't 6519been merged in. Also fixed a related bug where ext2fs_write_dir_block 6520was used to write out a non-directory block. (Which would be bad on a 6521non-Intel platform with byte swapping going on.) 6522 6523Fixed bug in e2fsck where it would print a "programming error" message 6524instead of correctly identifying where a bad block was in used when 6525the bad block was in a non-primary superblock or block group 6526descriptor. Also fixed a related bug when sparse superblocks are in 6527use and there is a bad block where a superblock or block group 6528descriptor would have been in a group that doesn't include a 6529superblock. 6530 6531Fixed a bug in e2fsck (really in libext2fs's dblist function) where if 6532the block group descriptor table is corrupt, it was possible to try to 6533allocate a huge array, fail, and then abort e2fsck. 6534ext2fs_get_num_dirs() now sanity checks the block group descriptor, 6535and subsitutes reasonable values if the descriptors are obviously bogus. 6536 6537If e2fsck finds a device file which has the immutable flag set and the 6538i_blocks beyond the normal device number are non-zero, e2fsck will 6539offer to remove it, since it's probably caused by garbage in the inode 6540table. 6541 6542When opening a filesystem, e2fsck specially checks for the EROFS error 6543code, and prints a specific error message to the user which is more 6544user friendly. 6545 6546If the filesystem revision is too high, change the e2fsck to print 6547that this is either because e2fsck is out of date, or because the 6548superblock is corrupt. 6549 6550E2fsck now checks for directories that have duplicate '.' and '..' 6551entries, and fixes this corruption. 6552 6553E2fsck no longer forces a sync of the filesystem (with attendant sleep 6554calls) at all times. The ext2fs_flush() function now performs a sync 6555only if it needed to write data blocks to disk. 6556 6557Fixed a minor bug in e2fsck's pass1b's file cloning function, where 6558certain errors would not be properly reported. 6559 6560Updated and expanded a few points in the man pages which users 6561complained wheren't explicit enough. 6562 6563Added special case byte-swapping code if compiling on the PowerPC, to 6564accomodate the strange big-endian variant of the ext2 filesystem that 6565was previously used on the PowerPC port. 6566 6567 6568Programmer's notes: 6569------------------- 6570 6571Removed C++ keywords from the ext2fs libraries so that it could be 6572compiled with C++. 6573 6574E2fsck's internal organization has now been massively reorganized so 6575that pass*.c don't have any printf statements. Instead, all problems 6576are reported through the fix_problem() abstraction interface. E2fsck 6577has also been revamped so that it can be called as a library from a 6578application. 6579 6580Added new fileio primitives in libext2fs for reading and writing 6581files on an unmounted ext2 filesystem. This interface is now used by 6582debugfs. 6583 6584Added a new libext2fs function for mapping logical block numbers of 6585a file to a physical block number. 6586 6587Added a new libext2fs function, ext2fs_alloc_block(), which allocates 6588a block, zeros it, and updates the filesystem accounting records 6589appropriately. 6590 6591Added a new libext2fs function, ext2fs_set_bitmap_padding(), which 6592sets the padding of the bitmap to be all one's. Used by e2fsck pass 5. 6593 6594The libext2fs functions now use a set of memory allocation wrapper 6595functions: ext2fs_get_mem, ext2fs_free_mem, and ext2fs_resize_mem, 6596instead of malloc, free, and resize. This makes it easier for us to 6597be ported to strange environments where malloc, et. al. aren't 6598necessarily available. 6599 6600Change the libext2fs fucntion to return ext2-specific error codes 6601(EXT2_DIR_EXISTS and EXT2_DB_NOT_FOUND, for example) instead of using 6602and depending on the existence of system error codes (such as EEXIST 6603and ENOENT). 6604 6605Renamed io.h to ext2_io.h to avoid collision with other OS's header 6606files. 6607 6608Add protection against ext2_io.h and ext2fs.h being included multiple 6609times. 6610 6611The types used for memory lengths, etc. have been made more portable. 6612In generla, the code has been made 16-bit safe. Added Mark 6613Habersack's contributed DOS disk i/o routines. 6614 6615Miscellaneous portability fixes, including not depending on char's 6616being signed. 6617 6618The io_channel structure has a new element, app_data, which is 6619initialized by the ext2fs routines to contain a copy of the filesystem 6620handle. 6621 6622ext2fs_check_directory()'s callback function may now return the error 6623EXT2_ET_CALLBACK_NOTHANDLED if it wishes ext2fs_check_directory() to 6624really do the checking, despite the presence of the callback function. 6625 6626 6627E2fsprosg 1.11 (June 17, 1997) 6628============================== 6629 6630Fixed e2fsck to detect (previously ignored) conflicts between the 6631superblock or block group descriptors and block bitmaps, inode 6632bitmaps, and inode tables. 6633 6634Fixed bug in e2fsck so that when the message printed out when a block 6635or inode bitmap conflicts with other data, it has the correct group 6636number. 6637 6638Fixed bug in e2fsck and mke2fs where the blocksize wasn't being passed 6639to badblocks. This meant that not all of the filesystem was being 6640tested for bad blocks! 6641 6642Fixed an array boundary overrun case which cropped up in 6643ext2fs_badblocks_list_test when a user tried running "mke2fs -c 6644-b 4096". 6645 6646Adjusted the number of columns printed by mke2fs when displaying the 6647superblock backups to avoid running over 80 columns when making a 6648really big filesystem. 6649 6650Fixed up the man pages for e2fsck, debugfs, badblocks, chattr, 6651dumpe2fs, fsck, mke2fs, and tune2fs (typos and other minor grammar 6652fixes), thanks to some suggestions from Bill Hawes (whawes@star.net). 6653 6654Programmer's notes: 6655------------------- 6656 6657Fixed install rule in lib/ss so that ss_err.h is actually getting 6658installed. 6659 6660Fixed bug in ext2fs_copy_bitmap; the destination bitmap wasn't getting 6661bassed back to the caller. 6662 6663Fixed bug in ext2fs_inode_scan_goto_blockgroup; it had not been 6664setting the current inode number (which meant this function wasn't 6665working at all). 6666 6667Fixed bug in ext2fs_resize_generic_bitmap; it had not be zeroing all 6668blocks in the bitmap when increasing the size of the bitmap. 6669 6670Changed the initial number of blocks allocated by ext2fs_init_dblist() 6671to be more realistic. 6672 6673Added a new function ext2fs_allocate_group_table, which sets up the 6674group descriptor information (and allocates inode and block bitmaps, 6675and inode tables for a particular group). The function was created by 6676factoring out code form ext2fs_allocate_tables(). 6677 6678Added a new function ext2fs_move_blocks which takes a bitmap of the 6679blocks to be moved, and moves them to another location on the 6680boardboard. 6681 6682Make the unix_io channel's io_channel_flush implementation calls sync() 6683to to flush the kernel buffers to disk. 6684 6685Added a new function ext2fs_dblist_count returns the number of 6686directory blocks in dblist. 6687 6688 6689E2fsprogs 1.10 (April 24, 1997) 6690=============================== 6691 6692Mke2fs once again defaults to creating revision #0 filesystems, since 6693people were complaining about breaking compatibility with 1.2 kernels. 6694Warning messages were added to the mke2fs and tune2fs man pages that 6695the sparse superblock option isn't supported by most kernels yet (1.2 6696and 2.0 both don't support parse superblocks.) 6697 6698Added new flag to mke2fs, -R <raid options>, which allows the user to 6699tell mke2fs about the RAID configuration of the filesystem. Currently 6700the only supported raid option is "stride" which specifies the width 6701of the RAID stripe. 6702 6703Fixed bug in e2fsck where pass1b would bomb out if there were any 6704blocks marked bad in the inode table. 6705 6706Fixed rare bug in mke2fs where if the user had a very unlucky number 6707of blocks in a filesystem (probability less than .002) the resulting 6708filesystem would be corrupt in the last block group. 6709 6710Fixed bug where if e2fsck tried to allocate a block to fix a 6711filesystem corruption problem and the filesystem had no free blocks, 6712ext2fs_new_block() would loop forever. 6713 6714The configure script now checks explicitly to see if "-static" works, 6715since that can't be assumed to be true --- RedHat doesn't install 6716libc-static by default. 6717 6718Fixed bug in libext2's block iterator functions where under some 6719cirmcustances, file with holes would cause the bcount parameter to the 6720callback function to be incorrect. This bug didn't affect any of 6721e2fsprogs programs, but it was discovered by Paul Mackerras, the 6722author of the PPC boot loader. 6723 6724Removed use of static variables to store the inode cache in libext2fs. 6725This caused problems if more than one filesystem was accessed via 6726libext2fs (static variables in libraries are generally a bad idea). 6727Again, this didn't affect e2fsprogs programs, but it was discovered by 6728Paul Mackerras. 6729 6730Fixed minor bugs and version code drift to assure that e2fsprogs 1.10 6731will compile cleanly with 1.2.13 kernels (even with a.out shared 6732libraries!) 6733 6734Programmer's notes: 6735------------------- 6736 6737Added new functions to duplicate an ext2 filesystem handle, and its 6738associated substructure. New functions: ext2fs_dup_handle(), 6739ext2fs_copy_dblist(), ext2fs_badblocks_copy(), ext2fs_copy_bitmap(). 6740Other structures, such as the io_channel and the inode_cache, now have 6741a ref count so that they only get freed when they are no longer used 6742by any filesystem handle. (These functions were added as part of the 6743development effort for an ext2 resizer). 6744 6745E2fsprogs 1.09 (April 14, 1997) 6746=============================== 6747 6748Fixed bug in mke2fs (really in lib/ext2fs/initialize.c) which was 6749accidentally introduced in the 1.08 release. The overhead calculation 6750was accidentally removed, which caused ext2fs_initialize() to not 6751notice when the filesystem size needed to be adjusted down because 6752there wasn't enough space in the last block group. 6753 6754Fixed bug in version parsing library routine; it was always parsing 6755the library version string, instead of using the passed-in string. 6756 6757Clarified chattr man page. 6758 6759E2fsprogs 1.08 (April 10, 1997) 6760=============================== 6761 6762E2fsck 1.07 was very slow when checking very large filesystems with a 6763lot of files that had hard links (i.e., news spools). This was fixed 6764by seriously revamping the icount abstraction. Added a formal test 6765suite for the icount abstraction. 6766 6767Debugfs now has a "-l" option to the "ls" command, which lists the 6768inode number, permissions, owner, group, size, and name of the files 6769in the directory. 6770 6771Fix a bug in e2fsck where when a directory had its blocks moved to 6772another location during the pass 1b processing, the directory block 6773list wasn't updated, so pass 2 wouldn't check (and correct) the 6774correct directory block. 6775 6776E2fsck will now treat inodes which contain blocks which are claimed by 6777the filesystem metadata by treating them as multiply claimed blocks. 6778This way, the data in those blocks can be copied to a new block during 6779the pass 1b--1d processing. 6780 6781E2fsck will attempt to determine the correct superblock number and 6782display it in the diagnostic and warning messages if possible. 6783 6784Add support for a new (incompatible) feature, "sparse_super". This 6785feature reduces the number of blocks which contain copies of backup 6786superblocks and block group descriptors. (It is only an incompatible 6787feature because of a bug in ext2_free_blocks.) mke2fs and tune2fs now 6788support a new -s option; e2fsck will recognize filesystems built with 6789this feature turned on. 6790 6791E2fsck now checks the library to make sure is the correct version, 6792using new library functions. (This helps to diagnose incorrectly 6793installed e2fsprogs distributions.) 6794 6795Dumpe2fs now prints more information; its now prints the the 6796filesystem revision number, the filesystem sparse_super feature (if 6797present), the block ranges for each block group, and the offset from 6798the beginning of the block group. 6799 6800Mke2fs now distributes the inode and block bitmap blok so that the 6801won't be concentrated in one or two disks in RAID/striping setups. 6802Also, if the user chooses a 2k or 4k block group, mke2fs will try to 6803choose the largest blocks per group that be chosen. (For 2k blocks, 6804you can have up to 16384 blocks/group; for 4k blocks, you can have up 6805to 32768 blocks/group.) Previously mke2fs would not allow 6806specification of more than 8192 blocks per group, even if you were 6807using a 2k or 4k block group. 6808 6809Programmer's notes: 6810------------------- 6811 6812Added a new function ext2fs_create_icount2() which takes a "hint" 6813argument. This hint argument presets the icount array with the list 6814of inodes which actually need to be in the icount array. This really 6815helps to speed up e2fsck. 6816 6817Added a new function ext2fs_icount_validate() which checks the rep 6818invariant for the icount structure. This is used mostly for testing. 6819 6820The error mesasage given when a bad inode number is passed to 6821test_generic_bitmap to reflect EXT2FS_TEST_ERROR (instead of 6822EXT2FS_UNMARK_ERROR). 6823 6824Added a new function ext2fs_set_dir_block which sets the block of a 6825dblist entry, given the directory inode and blockcnt. 6826 6827Added a new function ext2fs_get_library_version() which returns the 6828current library version, and ext2fs_parse_version_string() which 6829returns a version number based on a e2fsprogs version string. 6830 6831The icount functions will return EINVAL if the passed in inode number 6832is out of bounds. 6833 6834E2fsprogs 1.07 (March 9, 1997) 6835============================== 6836 6837E2fsck is now uses much less memory when checking really large 6838filesystems (or rather, filesystems with a large number of inodes). 6839Previously a filesystem with 1 million inodes required 4 megabytes of 6840memory to store inode count statistics; that storage requirement has 6841now been reduced to roughly half a megabyte. 6842 6843E2fsck can now properly deal with bad blocks appearing inside the 6844inode table. Instead of trying to relocate the inode table (which 6845often failed because there wasn't enough space), the inodes in the bad 6846block are marked as in use. 6847 6848E2fsck will automatically try to use the backup superblocks if the 6849primary superblocks have a bad magic number or have missing meta-data 6850blocks (or meta-data blocks which are out of range). 6851 6852E2fsck's pass 3 has been made more efficient; most noticeable on 6853filesystems with a very large number of directories. 6854 6855Completely revamped e2fsck's system of printing problem reports. It 6856is now table driven, to make them more easily customizeable and 6857extendable. Error messages which can be printed out during preen mode 6858are now one line long. 6859 6860Fixed e2fsck's filesystem swapping code so that it won't try to swap 6861fast symbolic links or deleted files. 6862 6863Fixed e2fsck core dumping when fixing a filesystem which has no 6864directories (not even a root directory). 6865 6866Added a check to e2fsck to make sure that the length of every 6867directory entry is a multiple of 4 (since the kernel complains if it 6868isn't). 6869 6870Added a check to e2fsck to make sure that a directory entry isn't a 6871link to the root directory, since that isn't allowed. 6872 6873Added a check to e2fsk to now make sure the '.' and '..' directory 6874entries are null terminated, since the 2.0 kernel requires it. 6875 6876Added check to write_bitmaps() to make sure the superblock doesn't get 6877trashed if the inode or block bitmap is marked as being block zero. 6878 6879Added checking of the new feature set fields in the superblock, to 6880avoid dealing with new filesystem features that this package wasn't 6881set up to handle. 6882 6883Fixed a fencepost error in ext2fs_new_block() which would occasionally 6884try to allocate a block beyond the end of a filesystem. 6885 6886When the UUID library picks a random IEEE 802 address (because it 6887can't find one from a network card), it sets the multicast bit, to 6888avoid conflicting with a legitimate IEEE 802 address. 6889 6890Mke2fs now sets the root directory's owner to be the real uid of the 6891user running mke2fs. If the real uid is non-zero, it also sets 6892the group ownership of the root directory to be the real group-id of 6893the user running mke2fs. 6894 6895Mke2fs now has more intelligent error checking when it is given a 6896non-existent device. 6897 6898When badblocks is given the -vv option, it now updates the block that 6899it is currently testing on every block. 6900 6901Fixed a bug in fsck where it wouldn't modify the PATH envirnoment 6902currently correctly if PATH wasn't already set. 6903 6904Shared libraries now built with dependencies. This allows the shared 6905library files to be used with dlopen(); it also makes the transition 6906to libc 6 easier, since ld.so can tell which libc a particular shared 6907library expects to use. 6908 6909Programmer's notes: 6910------------------- 6911 6912Added new abstraction (defined in dblist.c) for maintaining a list of 6913blocks which belongs to directories. This is used in e2fsck and other 6914programs which need to iterate over all directories. 6915 6916Added new functions which test to see if a contiguous range of blocks 6917(or inodes) are available. (ext2fs_*_bitmap_range). 6918 6919Added new function (ext2_inode_has_valid_blocks) which returns true if 6920an inode has valid blocks. (moved from e2fsck code). 6921 6922Added new function (ext2fs_allocate_tables) which allocates the 6923meta-data blocks as part of initializing a filesystem. (moved from 6924mke2fs code). 6925 6926Added a new I/O manager for testing purposes. It will either allow a 6927program to intercept I/O requests, or print debugging messages to 6928trace the activity of a program using the I/O manager. 6929 6930The badblocks_list functions now store the bad blocks in a sorted 6931order, and use a binary search to speed up badblocks_list_test. 6932 6933The inode scan function ext2fs_get_next_inode() may now return a soft 6934error returns: MISSING_INODE_TABLE and BAD_BLOCK_IN_INODE_TABLE in 6935those cases where part of an inode table is missing or there is a bad 6936block in the inode table. 6937 6938Added a new function (ext2fs_block_iterate2) which adds new arguments to 6939the callback function to return a pointer (block and offset) to the 6940reference of the block. 6941 6942Added new function (ext2fs_inode_scan_goto_blockgroup) which allows an 6943application to jump to a particular block group while doing an inode 6944scan. 6945 6946The badblocks list functions were renamed from badblocks_* to 6947ext2fs_badblocks_*. Backwards compatibility functions are available 6948for now, but programs should be modified to use the new interface. 6949 6950Some of the library functions were reorganized into separate files to 6951reduce the size of some programs which statically link against the 6952ext2 library. 6953 6954Put in some miscellaneous fixes for the Alpha platform. 6955 6956 6957E2fsprogs 1.06 (October 7, 1996) 6958================================ 6959 6960Fixed serious bug in e2fsck: if the block descriptors are bad, don't 6961smash the backup copies in ext2fs_close(). (The problem was that when 6962e2fsck -p discovered the problem, while it was closing the filesystem 6963and exiting, it was also blowing away the backup superblocks on the 6964disk, which was less than friendly.) We now make it the case that we 6965only write out the backup superblock and the back block descriptors if 6966the filesystem is completely free from problems. 6967 6968Fixed a bug in block_interate in the lib/ext2fs library which caused 6969e2fsck to fail on GNU Hurd-created filesystems. 6970 6971Add support for Linux/FT's bootloader, which actually uses 6972EXT2_BOOT_LOADER, and sets its mode bits which caused e2fsck to want 6973to clear the inode. 6974 6975Add support for the "A" (no atime update) attribute. (Note: this 6976attribute is not yet in production kernels.) 6977 6978The test suite is not automatically run when doing a "make all" from 6979the top level directory. Users should manually run "make check" if 6980they wish to run the test suite. 6981 6982Upon a preenhalt(), make the printed message more explicit that 6983running e2fsck "MANAULLY" means without the -p or -a options. 6984 6985In e2fsck, if a disconnected inode is zero-length, offer to clear it 6986instead of offering to connect it to lost+found. 6987 6988In e2fsck, if a filesystem was just unmounted uncleanly, and needs 6989e2fsck to be run over it, change e2fsck to explicitly display this 6990fact. 6991 6992For dumpe2fs and e2fsck, cause the -V option to print out which 6993version of the ext2fs library is actually getting used. (This will 6994help detect mismatches of using a 1.06 utility with a 1.05 library, 6995etc.) 6996 6997Programmers' notes: 6998------------------- 6999 7000EXT2_SWAP_BYTES was changed to EXT2_FLAG_SWAP_BYTES, which better fits 7001the naming convention. 7002 7003In ext2fs_initialize(), make sure the description for the inode bitmap 7004is correctly initialize. 7005 7006Fixed minor type typo in ext2fs_allocate_generic_bitmap(); 7007 7008E2fsprogs 1.05 (September 7, 1996) 7009================================== 7010 7011Add support for new fields in the ext2 superblock --- volume name, 7012volume UUID, and last mounted field. Dumpe2fs displays these fields, 7013tune2fs and mke2fs allows you to set them. E2fsck will automatically 7014generate a UUID for those volumes that don't have them. 7015 7016Put in support for e2fsck to recognize HURD specific ext2 features --- 7017most notably, the translator block. The e2fsprogs tools will now use 7018the creator_os field in the superblock to correctly handle different 7019OS-specific variants of the ext2 filesystem. 7020 7021E2fsck now fixes inodes which have a the deletion time set, but which 7022have a non-zero i_link_count field by offering to clear the deletion 7023time. Previously e2fsck assumed that the inode was deleted (per 0.3c 7024ext2 kernel behavior) and offered to unlink the file. 7025 7026If e2fsck sets the clean bit, but nothing else, set the exit code 7027FSCK_NONDESTRUCT. After all, e2fsck did fix a filesystem error --- it 7028set the filesystem valid bit when it was previously cleared. :-) This 7029was needed to make the HURD fsck driver happy. 7030 7031If the user refuses to attach an unattached inode, e2fsck will no 7032longer set the inode's link count. Otherwise, the inode would end up 7033getting marked as unused, which might cause loss of data later. 7034 7035Make the message issued by e2fsck when the superblock is corrupt less 7036confusing for users. It now mentions that another reason for the 7037"corrupt superblock" message might be that the partition might not be 7038an ext2 filesystem at all (it might swap, msdos filesystem, ufs, etc.) 7039 7040Make the libext2 library more robuest so that e2fsck won't coredump on 7041an illegal superblock where the blocksize is zero. (f_crashdisk is 7042the test case). 7043 7044By default, create filesystems where the default checkinterval is 6 7045months (180 days). Linux servers can be robust enough that 20 reboots 7046can be a long, long time. 7047 7048Added configure flag --enable-old-bitops, which forces the bitops to 7049use the old (native) bitmask operations. By default on the sparc 7050platform, the standard ext2 bit ordering is now used. 7051 7052Added a new feature to e2fsck to byte-swap filesystems; this can be 7053used to convert old m68k filesystems to use the standard byte-order 7054storage for the superblock, inodes, and directory blocks. This 7055function is invoked by using the '-s' option to e2fsck. 7056 7057Debugfs's "dump" command has been enhanced so that it writes out the 7058exact size of the file so that the nulls at the end of the file are 7059eliminated. The command also accept a new "-p" option which will 7060attempt preserve to preserve the ownernship, permissions, and 7061file modification/access times. 7062 7063Debugfs has two new options, -f and -R. The -R option allows the user 7064to execute a single debugfs command from the command line. The -f 7065option allows the user to specify a "command file" containing debugfs 7066commands which will get executed. 7067 7068Dumpe2fs now pretty prints the check interval, instead of just 7069printing the check interval as a number of seconds. 7070 7071Fix bugs in debugfs: the params command when no filesystem is opened 7072no longer causes a core dump. It is now possible to unlink a file 7073when a pathame containing a '/' is specified. 7074 7075Tune2fs has a new -C option which sets the number of times the 7076filesystem has been mounted. 7077 7078Fix the chattr '-v' option so that it actually works. Chattr was 7079being buggy about the -v option parsing. 7080 7081Programmers' notes: 7082------------------- 7083 7084The directory lib/uuid contains a set of library routines to generate 7085DCE compatible UUIDs. 7086 7087Extended ext2fs_namei() to handle symbolic links. Added new function 7088ext2fs_nami_follow() which will follow last symbolic link in the case 7089where the pathname points to a sym link. 7090 7091The ext2fs_block_iterate function will now return the HURD translator 7092block, if present. The new flag BLOCK_FLAG_DATA_ONLY will cause the 7093iterator to return data blocks only. The ext2fs.h file now defines 7094constants BLOCK_COUNT_IND, BLOCK_COUNT_DIND, BLOCK_COUNT_TIND, and 7095BLOCK_COUNT_TRANSLATOR, which are the magic values passed in the block 7096count field of the iterator callback function. 7097 7098The test script driver now takes an optional second argument, which is 7099the test case to be run. This allows you to run a test case without 7100needing to run the entire test suite. 7101 7102On Linux ELF systems, install the .so files in the correct places 7103(/usr/lib). The .so files must be stored in the same directory as the 7104.a files. 7105 7106Fixed miscellaneous HURD compilation issues with header file being 7107included in the right order. 7108 7109Fixed debugfs so that it resets optind to zero, not one, since setting 7110optind to zero is more correct. 7111 7112 7113E2fsprogs 1.04 (May 16, 1996) 7114============================= 7115 7116First "official" (1.03 was a limited release only) to support building 7117e2fsprogs under Linux 2.0 kernels (as well as late model 1.3 and 1.99 7118kernels). 7119 7120This package includes a RPM specs file, that it can be built using the 7121RedHat Package Manager. 7122 7123E2fsck now prints a hint that if there are lots of incorrectly located 7124inode bitmaps, block bitmaps, and inode table blocks, the user might 7125want to try using e2fsck -b 8193 first, to see if that fares any 7126better. 7127 7128For ext2 filesystem written with the hurd, debugfs will now print out 7129the translator field when printing an inode structure. 7130 7131Lots of miscellaneous linking/installation cleanups: 7132 7133 Libraries are now linked using a relative pathname, instead of 7134 relying on -L working correct. It doesn't, in many cases, including 7135 current versions of GNU ld. This guarantees that the build tree is 7136 linking with the right libraries, instead of the ones installed in 7137 /usr/lib. 7138 7139 Header files, man pages, and the et/ss shell scripts are now 7140 generated using a custom substitution script, instead of relying on 7141 the configure script. This prevents needless recompilation of 7142 files; in addition, the custom substitution script is much faster. 7143 7144 e2fsck may now be linked dynamically, by using the 7145 --enable-dynamic-e2fsck flag to configure. This is not recommended, 7146 since it increases e2fsck's dependence on other files, but some 7147 people need to save disk space, and other critical programs on their 7148 systems were being linked dynamically anyway. 7149 7150 Programs such as fsck which didn't need to be linked against 7151 libext2fs (or mke2fs which didn't need to be linked against libe2p) 7152 only link against libraries they actually need. Otherwise, those 7153 programs would require the presense of libraries that otherwise 7154 could be removed from a rescuse diskette. 7155 7156 The ss include files are now installed correctly so they can 7157 actually be used by another package. 7158 7159 If the profiling libraries are built, they are now installed on a 7160 "make install-libs". 7161 7162 7163E2fsprogs 1.03 (March 27, 1996) 7164=============================== 7165 7166Change the m68k bit numbering for bitmasks to match the bit numbering 7167used by all other ext2 implementations. (This change was requested by 7168the m68k kernel development team.) 7169 7170Support (in-development) filesystem format revision which supports 7171(among other things) dynamically sized inodes. 7172 7173Fixed a bug in the ext2 library so that an intelligent error is 7174returned if mke2fs is run with a ridiculously small number of blocks 7175for a partition. 7176 7177Fixed a bug in the ext2 library which required that the device be 7178openable in read/write mode in order to determine its size. This 7179caused e2fsck -n to require read/write access when it was not 7180previously necessary. 7181 7182Fixed a bug in e2fsck which casued it to occasionally fail the test 7183suite depending on which version of the floating point library it was 7184using. 7185 7186Fixed a bug in e2fsck so that it now halts with a fatal error when 7187certain superblock consistency checks fail. Previously it continued 7188running e2fsck, with some potential confusing/damaging consequences. 7189 7190Added new flag to fsck which allows the root to be checked in parallel 7191with other filesytems. This is not the safest thing in the world to 7192do, but some system administrators really wanted it. 7193 7194Fixed -Wall flames in lib/ss. 7195 7196 7197E2fsprogs 1.02 (January 16, 1996) 7198================================= 7199 7200Fix to allow e2fsprogs to be compiled on recent 1.3 (pl45+) kernels. 7201 7202Change e2fsck to print statistics of how many non-contiguous files are 7203on the system. Note that a file which is larger than 8k blocks, it is 7204guaranteed to be non-contiguous. 7205 7206In mke2fs, print a warning message if a user tries to format a whole 7207disk (/dev/hda versus /dev/hda1). If a user really wants to format a 7208whole disk, the -F (force) option forces mke2fs to format a whole disk 7209as a filesytem. 7210 7211Fix a bug in fsck where in some cases it might start checking 7212partitions in the next pass before it finishes checking partitions in 7213the current pass. This still won't cause two partitions on the same 7214disk will be checked, so it's rarely a problem in real life. 7215 7216Patch lsattr so that it won't hang when checking a named pipe. 7217 7218Minor compilation fixes: 7219 * Fix the order of libraries that were linked in debugfs. 7220 * Allow the sources to be compiled with -ansi turned on. 7221