Log Message:
Updated sysutils/e2fsprogs to 1.43.4
-------------------------------------
    (from RELEASE-NOTE, some 170 lines, sorry)
-----------------------------------------
E2fsprogs 1.43.4 (January 31, 2017)
===================================

Fix e2fsck to require that the system.data extended attribute is
always present even for files smaller than 60 bytes, to be consistent
with the kernel.

Fix a bug which was causing mke2fs -d and fuse2fs to improperly handle
Posix ACL's.

Fix a bug which could cause mke2fs -d to fail if there is an
zero-length file in source directory tree.

Clarify the default for the "proceed?" question printed by mke2fs and
tune2fs.  (Addresses Debian Bug: #852727)

Debugfs will now display project quota information.

Debugfs's do_set_xattr now uses C strings to parse extended attribute
values, and will print the extended attributes using either C strings
or hex bytes when using debugfs's do_get_xattr command.  It will now
allow developers to see the contents of the system.data extended
attribute.

Fix a bug in mke2fs where I/O errors weren't getting properly reported
to the user.

Fix multiblock writes in the unaligned Direct I/O fallback code.
(Which is rarely needed on Linux systems, but which is required on
Freebsd systems.)

Make sure the default mke2fs.conf file includes the uninit_bg feature
flag.  (Which was accidentally dropped in v1.43).

Fix a bug in resize2fs when operating on very large file systems which
have a block size different from the VM page size.

If the reported device size is absurdly large, mke2fs will now report
an error instead looping forever or crashing.

Fix various Coverity warnings and other memory leaks in fuse2fs and
extended attribute manipulation functions.

Replace a test file but which had a "non-commercial use-only"
copyright permission file with a newer version from the Cyrus imapd
package which now has a 4-clause BSD license, which was making some
lawyers nervous, even though the test file in question was only used
in lib/et's regression testing and was never included in any compiled
binary.  (Addresses Debian Bug: #840733)

Stop installing mkfs.ext4dev and fsck.ext4dev

Update Chinese, Czech, Polish, Serbian, and Spanish translations and
add the Finnish and Malay translation.  (Addresses-Debian-Bug: #774379)

Update various man pages (Addresses-Debian-Bug: #852726)

Programming notes
-----------------

Fix more FreeBSD-11 compatibility bugs, including some unmasked by
FreeBSD 11-rc2.

Fix the Mac build.

Fix build failure on RHEL 5.x systems caused by an old version of
libmagic.

Fix a number of bugs reported by ASAN that can cause some (mostly
harmless) memory dereferences beyond allocated memory.

E2fsprogs 1.43.3 (September 4, 2016)
====================================

Fix e2fsck's handling of timestamps on 32-bit systems.

E2fsck will now check, and if necessary repair the extra isize fields
in the inode and superblock.

Fix crashes on architectures such as sparc64 that are sensitive to
unaligned pointer derferences in the journal recovery code when
journal checksums are enabled.

Programming notes
-----------------

Support reproducible builds by not capturing the build directory into
the mk_cmds and compile_et scripts.  Also fix debian build rules to
ensure build reproducibility.

Fix debian build rules to ensure build reproducibility and to avoid
hiding the linker flags for e2fsck.static so the build hardening log
scanner can properly audit the build.

Fix compatibility with FreeBSD's pmake and teach the configure script
to force the creation of pmake-compatible Makefiles if the
FORCE_NATIVE_MAKE environment variable is set to a non-empty value.

E2fsprogs 1.43.2 (September 1, 2016)
====================================

Fix resize2fs so it will not crash if there is an extended attribute
block but it doesn't need to migate any blocks during an off-line
resize

Fix a crash when mke2fs or debugfs tries to copy in a directory
hierarchy containing an empty directory

Mke2fs will now use a larger journal by default for filesystems
greater than 128 GB.  (1GB instead of 128 MB.)

Fix an alignment bug in e2fsck which caused sparc64 architectures to
crash when replaying the journal on file systems with a 64-bit block
number.

Clarify the message printed by tune2fs message when the user needs to
run e2fsck so it's clear that the -f flag to e2fsck is needed to
force a full e2fsck scan.  (Addresses Debian Bug: #828022)

Fix a bug in e2fsck caused by a power failure during e2fsck's journal
replay could leave the file system in a state where if the file system
is mounted without doing a full e2fsck scan, the file system could get
corrupted

Fix the logic in e2fsck which decides when to repair legacy negative
timestamp encodings.

Add a command to debugfs to copy the inode structure from one inode to
another.

Fix a typo in debugfs's stat command when printing out the dtime field
on file systems with an extended timestamp.

Fix big endian bugs in the e2undo program.  (Addresses Red Hat Bug
#1344636)

Debugfs's logdump can now properly handle journals larger than 2GB.

Avoid installing the man page for fuse2fs if it has not been built.

Update the Catalan, Chinese, Danish, Dutch, French, German, Polish,
Swedish, Ukrainian translations and added new translations for
Hugarian and Serbian.

Programming notes
-----------------

Fix portability problems in fuse2fs.  Previously it wouldn't build on
systems with older glibc versions where clock_gettime() is only
available in the librt library, and if libintl is not bundled into the
C library.

Remove complicated logic which caused a static code analyzer to flag a
false positive.  (A static code analyzer also found a valid bug in
deciding when to repair a legacy negative timestamp encoding, so
eliminating false positives is important.)

Fixed a bug where the ext2fs library cloud provoke when a
extfs2_zero_blocks() is used (via fallocation, initializing a file
system, uninitialized uninitialized inode table blocks) after a
different file system which also used ext2fs_zero_blocks().

Enable the unix_io manager in the ext2fs library so it can accept the
use of a file descriptor.  This is helpful in cases where the file
descriptor comes from temporary file created using O_TMPFILE, or
passed in from a unix domain socket.

Fix a Windows64 portability bug.