[PATCH] Disallow use of DES encryption functions in new programs.

[Joseph Myers]

On Thu, 28 Jun 2018, Florian Weimer wrote:

> Joseph, do you think the above changes are correct?  Or was crypt a
> non-optional part of any of the standards at one point?

As far as I can tell it's always been in an optional CRYPT group.  It's 
less clear whether in the absence of support for that group some standards 
might have expected the interfaces to be present as ENOSYS stubs.

We don't have a well-defined way of handling optional feature groups in 
the conform/ tests (i.e. a way of saying "this is part of group X" and the 
tests verifying consistency with whether group X is declared to be 
supported).  Sometimes we just list them in the expectations as required 
(if glibc supports that group), sometimes as optional-* (which is more or 
less equivalent to not listing them), sometimes don't list them at all.

-- 
Joseph S. Myers
joseph@codesourcery.com