cryptsetup-2.8.1

[Douglas R. Reno]

New point version

Not tagged, so it can go into 12.4.

[Bruce Dubbs]

I've been using it, but have not gotten around to pushing the tag. Moving to 12.5.

[Douglas R. Reno]

I would like to request moving it back to 12.4 and updating it, per upstream recommendation:

Cryptsetup 2.8.1 Release Notes
==============================
Stable bug-fix release with minor extensions.

All users of cryptsetup 2.8.0 must upgrade to this version.

Changes since version 2.8.0
~~~~~~~~~~~~~~~~~~~~~~~~~~~

* Fix status and deactivation of TCRYPT (VeraCrypt compatible) devices that use chained 
ciphers.

* Fix unlocking BITLK (BitLocker compatible) devices with multibyte UTF8 characters in 
the passphrase.

* Do not allow activation of the LUKS2 device if the used keyslot is not encrypted (it 
uses a null cipher).

  Such a configuration cannot be created by cryptsetup, but can be crafted outside of 
it.
  Null cipher is sometimes used to create an empty container for later reencryption.
  Only an empty passphrase can activate such a container (the same as in LUKS1).

* Do not silently decrease PBKDF parallel cost (threads) if set by an option.
  The maximum parallel cost is limited to 4 threads.

* Fixes to configuration and installation scripts.

  Meson and autoconf tools now properly support --prefix option for temporary directory 
installation.
  Multiple fixes and cleanups to config.h for compatibility between Meson and autoconf.
  Fix the luks2-external-tokens-path Meson option to work the same as in autoconf.
  Fix Meson install for tool binaries, install fvault2Open man page and include test/
fuzz/meson.build in release.

* Major update to manual pages.

  Try to explain the PBKDF hardcoded limits.
  Add a better explanation for automatic integrity tag recalculation.
  Mention crypt/verity/integritytab.
  Remove or reformulate some misleading warnings present only with old and no longer 
supported kernels.
  Clarify that some commands do not wipe data and unify OPAL reset wording.
  Clarify the --label option.
  There are also many other grammar and stylistic fixes to unify the man-page style.

* Fixes for false-positive and annoying (optional) warnings added in recent compilers.

[Bruce Dubbs]

OK, I'll do it in a few minutes.

[Douglas R. Reno]

Bruce updated it at bf37902623e605fb9fa14840a6cf2b4eb60b37b2 but had me run the tests due to some missing kernel configuration (done at 0107d57857d09896fa079bc461e3ae4667f8591d)