Re: 6.17 Regression: loading trusted.ko with fips=1 fails due to crypto/testmgr.c: desupport SHA-1 for FIPS 140

["Theodore Ts'o" <tytso@mit.edu>]

On Sat, Oct 04, 2025 at 04:24:51PM -0700, Eric Biggers wrote:
> But for future reference, if the people doing FIPS certifications of the
> whole kernel actually determine that a particular kernel feature(s) that
> use SHA-1 *must* be disabled when fips_enabled=1, then of course they'll
> need to do that properly by submitting a tested and well-justified patch
> for each feature that carefully disables the correct functionality.

There's a hidden philosopical question here, which is whether "FIPS
certification of the whole kernel" is actually a good thing.
Personally, I don't think it is, but if booting with fips=1 neuters a
whole bunch of kernel features, and that is considered "working as
intended", to the extent that it discourages the use of FIPS mode,
maybe it's not such a bad thing.  :-)

But with that said, I suspect one of the things that distributions
might find useful is per-kernel subsystem fips enablement.  (e.g.,
dm_crypt.fips=1 which might make a whole bunch of existing users' file
systems become useless, precepitating a whole bunch of angry inquiries
to the distrobution's help desk, but maybe a particular user only needs
ipsec.fips=1)

						- Ted