incorrectly treat as valid symlinks created with SE Linux
(Debian bug #228723) as well as failing the f_journal test case on
big endian systems due to the backup journal blocks not being swapped.
the superblock and block group descriptors into two functions:
ext2fs_reserve_super_and_bgd, found in lib/ext2fs/alloc_sb.c, and
ext2fs_super_and_bgd_lock, found in lib/ext2fs/close.c.
Change e2fsck/pass1.c (mark_table_blocks), lib/ext2fs/closefs.c
(ext2fs_flush), lib/ext2fs/initialize.c (ext2fs_initialize),
and misc/dumpe2fs.c (list_desc) to use these functions.
e2fsck/ChangeLog
pass1.c (mark_table_blocks): Use the new function
ext2fs_reserve_super_and_bgd to calculate the blocks to be
reserved.
lib/ext2fs/ChangeLog
closefs.c (ext2fs_super_and_bgd_loc): New function which
centralizes the calculation of the superblock and block
group descriptors.
(ext2fs_flush): Use ext2fs_super_and_bgd_lock to figure
out where to write the superblock and block group
descriptors.
alloc_sb.c (ext2fs_reserve_super_and_bgd): New function which
reserves space in the block bitmap using
ext2fs_super_and_bgd_loc.
initialize.c (ext2fs_initialize): Use
ext2fs_reserve_super_and_bgd to initialize the block bitmap.
misc/ChangeLog
dumpe2fs.c (list_desc): Use ext2fs_super_and_bgd_loc to
determine the locations of the superblock and block group
descriptors.
superblock. E2fsck will automatically save the journal information
in the superblock if it is not there already, and will use it if the
journal inode appears to be corrupted. ext2fs_add_journal_inode()
will also save the backup information, so that new filesystems
created by mke2fs and filesystems that have journals added via
tune2fs will also have journal location written to the superblock as
well. Debugfs's logdump command has been enhanced so that it can
use the journal information in the superblock.
The debugfs man page has been improved to more fully describe the
logdump command.
Added two new functions, ext2fs_file_open2() and
ext2fs_inode_io_intern2() which take a pointer to an inode structure;
this is needed so that e2fsck and debugfs can synthesize a
fake journal inode and use it to access the journal.
non-empty bad block list. Resize2fs now discards any blocks on the
badblock list which are no longer part of the filesystem as the result
of a filesystem shrink. (Note: this means that shrinking and then
enlarging a filesystem is no longer a reversible operation;
information about bad blocks in the part of the filesystem
which is to be chopped off will be lost.)
device is larger than the default block size, then use the
sector size of the device as the default block size.
getsectsize.c (ext2fs_get_device_sectsize): New function which
returns the hardware sector size (if it is available).
tune2fs to use the test I/O manager.
The test I/O manager has been changed to not do anything extra by
default, unless the TEST_IO_FLAGS and/or TEST_IO_BLOCK environment
variables are set, which controls what I/O operations are logged and
a block number to watch, respectively. The log messages are sent to
stderr by default, unless a filename is specified via the
TEST_IO_LOGFILE environment variable.
Fix typo's in README.subset
Change debian control file so it doesn't bomb out if the EVMS FSIM
is not there, since it is not built on the Hurd. Resolves Debian
bug #189687.
instead of <com_err.h>, so the current version of the header file
is used.
Add a --build-tree option to compile_et to make sure that it uses
the et_?.awk files from the build tree.
Remove legacy support for varargs.h, K&R C, and pre-POSIX signal
support. Also fixed gcc -Wall nits.
large number of inodes, then reduce the number of blocks
per group until we find a workable set of filesystem
parameters.
ext2_err.et.in (EXT2_ET_TOO_MANY_INODES): Add new error code.
unix_write_blk): Optimize routines so that we don't end up
searching the cache twice when a block isn't in the
cache. If reads are larger than READ_DIRECT_SIZE, don't
let them go through the cache.
blocks to be erroneously marked as dirty, so they would
get written back to the disk before they are evicted from
the cache. Harmless, but it slows down e2fsck
significantly.
Disable certain tests that depend on the byteswapping functions
being present so that "make check" succeeds even if --disable-swapfs
is passed to the configure script.
support for loopback-mounted filesystems. Check /etc/mtab
if /proc/mounts doesn't turn up any mount flags, since
/etc/mtab has the loopback image filename, instead of
/dev/loop0. Also, check based on st_dev and st_ino, so
that if a relative pathname or a pathnames using symbolic
links are used, we can detect the the filesystem correctly
in those cases. (Addresses Sourceforge bug #619119)
don't try the FDFLUSH ioctl that was required for floppies
with older kernels. This avoids needless whining from the
MD device driver. (Addresses Sourceforge bug #545832).
wrong group_block to ext2fs_descriptor_block_loc if we're
using the backup superblock/block group descriptors.
(ext2fs_descriptor_block_loc): If we're using the backup
superblock descriptors, use the backup descriptor block in
the next block group.
the filesystem was created. It is set via mke2fs, and printed
via list_super2() (which is called by dumpe2fs).
Also, list_super2() will now print "n/a" if the last mount time
(s_mtime) is zero.
When byte-swapping a filesystem on a PPC architecture, byte-swap
the bitmaps since the historical big-endian ext2 variant had
byte-swapped bitmaps, and the ext2fs library assumes this. Otherwise
the regression test suite will fail...
using a non-zero hash version (i.e., half MD4 or TEA hash).
The hash version wasn't getting copied into dx_dir->hashversion and
this caused the kernel to treat all directories if they were using the
legacy hash, which was Bad.
Revert the shlibs hack.
Makefile.in: Remove inode_io.o from the standard object files
in libext2fs, and only build it if debugfs is enabled
(it requires fileio.o, which is only built if
--disable-debugfs isn't specified to configure).
a backwards incompatible change, but this should be OK since the kernel
MD4 ext3 code was only in the CVS tree, and hasn't generally escaped.
Add support for the TEA hash.
the kernel version is 2.4.10 -- 2.4.17, since otherwise an
old version of glibc (built against 2.2 headers) will
interact badly with the workaround to actually cause more
problems. I hate it when the glibc folks think they're
being smarter than the kernel....
8192-byte and 16384-byte blocksize filesystems.
Change the default bytes-per-inode ratio of a new filesystem to be at most
one inode per block for large blocksizes.
the EXT2_SB() macro instead. Remove kernel function
prototypes also. This matches the 2.5 kernel, and
is also cleaner for other reasons. Whitespace cleanup.
mapping for a particular inode.
Fixed a bug in the libext2 library which broke ext2fs_bmap if no inode structre
was passed inside for here.
Fixed bad calling parameters to parse_ulong which broken the -b and -s
options to debugfs, as well as do_init, and the testb, setb, clearb functions.
file we are checking is a block device file before looking
at st_rdev, since it's not valid for normal files.
(is_swap_device): Move so that it is outside the
HAVE_MNTENT_H, so that it is always built.
functions which take an extra flags argument. The flag
EXT2_DIRBLOCK_V2_STRUCT will reverse when the name_len
field is byte swampped on big-endian machines, since in
the V2 structure, name_len is a char field which is
doesn't need to be byte swapped --- except if an
old-style kernel had byte-swapped the name_len field
as part of the V1 structure.
Also fixed a bug in debugfs which used ext2_dir_entry_2 without
worrying about the above issue, with the net result that "ls -l"
would print an incorrect file type on big-endian systems.
an I/O object.
Export ext2_file_flush as a public interface.
Also minor cleanups to tighten code in other I/O abstractions, and to
mark a void * pointer as const in the ext2_file_write interface.
Add support for a new flag, DIRENT_FLAG_INCLUDE_REMOVED,
which will return deleted directory entries.
ext2fs_dir_iterate2 takes a new callback function which
is identical with the one used by
ext2fs_dblist_dir_iterate(). If the directory entry is
deleted, the callback function will be called with the
entry paraemter set to DIRENT_DELETED_FILE.
Makefile.in, alloc_stats.c (ext2fs_inode_alloc_stats,
ext2fs_block_alloc_stats): New functions which update
block/inode allocation statistics in the bitmaps, block
group descriptors, and superblock.
mkjournal.c (mkjournal_proc), mkdir.c (ext2fs_mkdir),
expanddir.c (expand_dir_proc), bb_inode.c
(clear_bad_block_proc, set_bad_block_proc,
ext2fs_update_bb_inode), alloc.c (ext2fs_alloc_block):
Update to use new block/inode allocation statistics.
ext2fs_check_if_mounted and ext2fs_check_mount_point which
determines whether or not the specified device is a swap
device by using /proc/swaps. More bulletproofing for
idiotic/careless system administrators!
since files which get dropped into that directory may have come
from a protected directory, and the system administrator may not
deal with immediately. (Addresses Debian bug #118443)
- Renamed linux/list.h to be linux/linked_list.h to work around a
problem caused by diet libc insistence to search the kernel
header files ahead of all other files in the include path,
including the user specified include files.
- Worked around a bug in diet libc which core dumps when using
putc with stderr by using fputs instead. As a bonus, this
also shaved a few bytes off of com_err.o.
- Fixed a real bug in debugfs which was detected because diet libc
was more sensitive than glibc when incorrectly using fclose()
where pclose() is required.
trying to unset the filesize limit if at all possible,
if a block device is getting opened. (The filesize limit
shouldn't be applied against writes to a block device, but
starting in 2.4.10, the kernel is doing this.)
file on adding a journal to an already-mounted filesystem,
try to clear the ext2 file attributes on an already
existing .journal file so that we don't fail if on a
partially added journal to the filesystem.
which signals that a particular inode should not have the
last bits of data (the "tail") be merged with another
file. This is necessary to keep programs like LILO happy.
/etc/mtab to make sure the filesystem is really mounted,
since some broken distributions (read: Slackware) have
buggy boot scripts that don't initialize /etc/mtab before
checking non-root filesystems. (Slackware also doesn't
check the root filesystem separately, and reboot if the
root filesystem had changes applied to it, which is
dangerous and broken.)
unconditional ext2_flush() call. Also made ext2_flush() more
efficient when the superblock is modified after the filesystem has
been flushed, and make it clear the superblock dirty flag
after doing the superblock/block group descriptor flush.
included if EXT2FS_ENABLE_SWAPFS is defined, since we need to support
byte swapping even if we don't support the conversion functions.
(Fixes Debian bug #101686).
the byte-swapping code all assumes the ext2_dir_entry structure.
(It's a question of whether or not name_len should be byte-swapped or
not, and whether it's a 16 bit or 8 bit field.)
Makefile.in: Fix bug in installation of ext2_types.h. It is a file
which is generated and is therefore found in the build directory, not
source directory.
ismounted.c (check_mntent_file): Use a test file in / to check to see
if the root filesystem is mounted read-only. This protects against
the case where /etc might not be on /, as well as the case where
/etc/mtab doesn't exist. (Both are should-never happen scenarios, but
you never know...)
ext2_types.h.in: If linux/types.h has been defined, then don't try to
redefine the typedefs.
ChangeLog, Makefile.in:
Makefile.in: Make the install target depend on $(SUBS) to accomodate
the fools who want to compile and install e2fsprogs as root using just
one command.
setsuper.c: Add s_lastcheck field to the fields which can be modified
using set_super_value. (Suggested by Andreas Dilger)
ChangeLog, Makefile.in:
Makefile.in (HFILES): Add ext2_types.h to the list of files which
should be installed.
Theodore Ts'o