Dimitri John Ledkov

```console make test TESTS=test_sysdefault ... make[2]: Entering directory '/home/xnox/upstream/openssl' ( SRCTOP=. \ BLDTOP=. \ PERL="/usr/bin/perl" \ FIPSKEY="f4556650ac31d35461610bac4ed81b1a181b2d8a43ea2854cbae22ca74560813" \ EXE_EXT= \ /usr/bin/perl ./test/run_tests.pl test_sysdefault ) 00-prep_fipsmodule_cnf.t .. ok All tests successful....

> I strongly believe that for this kind of nonsensical FIPS requirements that need to have clear exceptions we should use only explicit indicators. I am still perplexed by this....

```gdb (gdb) run kdf -keylen 32 -kdfopt key:16charlongsecret -kdfopt digest:SHA2-256 -kdfopt mode:EXTRACT_ONLY TLS13-KDF Starting program: /home/xnox/upstream/openssl/apps/openssl kdf -keylen 32 -kdfopt key:16charlongsecret -kdfopt digest:SHA2-256 -kdfopt mode:EXTRACT_ONLY TLS13-KDF Breakpoint 1, ossl_mac_check_key (min=0,...

Let's see if https://github.com/openssl/openssl/pull/24204 resolves the zero-length mac key during HKDF.

> How does this patch interact with PBKDF2? That algorithm uses HMAC internally on a password which may not be 112 bits long. I've updated current status on that in...

> My though would be to ban this by default (implicit indicator) but have the option of a bypass which sets an explicit indicator. > > This is what @slontis...

> My though would be to ban this by default (implicit indicator) but have the option of a bypass which sets an explicit indicator. > > This is what @slontis...

> I am curious about how this patch interact with a two-step KDF (e.g. HKDF)? > > As per the response from our lab and your statement in PR description,...

> IMO this clearly shows that many FIPS requirements are nonsense that sometimes cannot even be reasonably obeyed. > > The minimum MAC key length is a clear example. For...

@t8m can i join OTC meeting or some such to discuss future plans for fips, such that I can make progress towards end goals; in piece meal upgrades. Because current...