Compiling failing when WOLFSSH_NO_NISTP256_MLKEM768_SHA256 is defined.

[tjko]

I noticed that if WOLFSSH_NO_NISTP256_MLKEM768_SHA256 is defined but WOLFSSH_NO_CURVE25519_MLKEM768_SHA256 is not defined (when one wants to only use "mlkem768x25519-sha256" algorithm), compiling of wolfSSH is failing:

/home/tjko/git/fanpico/libs/wolfssh-dev/src/internal.c: In function 'SendKexDhReply':
/home/tjko/git/fanpico/libs/wolfssh-dev/src/internal.c:12560:25: error: 'MSGID_KEXKEM_REPLY' undeclared (first use in this function); did you mean 'MSGID_KEXDH_REPLY'?
12560 |                 msgId = MSGID_KEXKEM_REPLY;
      |                         ^~~~~~~~~~~~~~~~~~
      |                         MSGID_KEXDH_REPLY
/home/tjko/git/fanpico/libs/wolfssh-dev/src/internal.c:12560:25: note: each undeclared identifier is reported only once for each function it appears in
/home/tjko/git/fanpico/libs/wolfssh-dev/src/internal.c: In function 'SendKexDhInit':
/home/tjko/git/fanpico/libs/wolfssh-dev/src/internal.c:13166:21: error: 'MSGID_KEXKEM_INIT' undeclared (first use in this function); did you mean 'MSGID_KEXDH_INIT'?
13166 |             msgId = MSGID_KEXKEM_INIT;
      |                     ^~~~~~~~~~~~~~~~~
      |                     MSGID_KEXDH_INIT

[ejohnstown]

This is a guard check I've been thinking of removing. There's plenty of other guarded features whose message IDs are in that list. Those message IDs in particular are duplicates. They just look right reading them in the code that sends messages.

[tjko]

@ejohnstown, PR updated to remove the guard checks.