Changelog
cryptsetup (2:1.7.3-3) unstable; urgency=medium
[ Jonas Meurer ]
* debian/scripts/decrypt_ssl: fix script to actually output the decrypted
key. Apparently this script has been broken since June 2008. Doesn't seem
like anybody is using it. Thanks to g1 for spotting and reporting the
error. (Closes: #844050)
* debian/initramfs/cryptroot-script:
+ limit the sleep after max passphrase attempts to devices for the rootfs.
This mitigates the negative impact in case of broken keyscripts etc.
+ add $crypttarget to each message to provide more context.
* debian/initramfs/cryptroot-hook: fix sanity check for key files on root
fs in get_device_opts(): detect if processed device is a root (parent)
device even for LVM setups. (closes: #842951)
* debian/README.initramfs: minor fix to the decrypt_derived keyscript
section: now that systemd is standard, 'cryptdisks_start' should be used
instead of '/etc/init.d/cryptdisks start'.
* debian/manpages/crypttab.xml: add a warning to the 'keyscript' option
that systemd doesn't support the option (yet) and mention the possible
workaround to process the devices in question in the initramfs.
[ Guilhem Moulin ]
* add debian/gbp.conf to set the upstream tag to "v%(version%.%_)s". As
this enables git-buildpackage >= 0.8.7 to automatically generate
orig.tar.gz, step nr. 5 is now removed from debian/README.source.
* debian/compat: bump debhelper compatibility version to 9.
* debian/initramfs/cryptroot-hook:
+ fix tab damage for consistency with the rest of the code
+ better warning for deprecated settings
+ fix sanity check for key files in get_device_opts(): print a warning if
the key file isn't on the root FS, or if the root device is not
encrypted, even for LVM setups.
+ fix sanity check for key files in get_device_opts(): print a warning if
the processed device is a resume device, even for LVM setups.
+ fix runtime error in get_lvm_deps() if the first argument is either
missing or the empty string.
+ reset IFS after processing $rootopts in get_device_opts(); the missing
linefeed in $IFS caused LVM logical volumes spaning over multiple PVs
not to have their parent devices detected correctly.
-- Jonas Meurer <email address hidden> Fri, 09 Dec 2016 01:18:17 +0100