Move to user_settings and WOLFHSM_CFG_*.

[billphipps]

Further isolation of compile-time configuration values and moving to using user_settings.h to be included in those headers that support reconfiguration (wh_client.h, wh_server.h, wh_nvm.h, wh_she_common.h).

A number of additional typos and minor doc edits as well.

[bigbrett]

A few random things I noticed but good to merge as is.

Something that I do want to discuss though (also mentioned inline in a comment below):

• I don't think we want to establish the paradigm of putting wolfHSM configuration defines in a wolfSSL user_settings.h
• We should retain a separate user configuration header for wolfHSM, as supplying defines on CLI would get unwieldy.

What was wrong with wh_config.h?

[bigbrett]

small opinion/nit, feel free to disregard: I kind of like keeping the "allow" word in these defines, because it really indicates the number of addresses in the allowlist.

[bigbrett]

LGTM nice work. I'm wondering if we want to be including wh_settings.h in every header when it isn't needed externally - think it would be best to keep that inclusion in the *.c file and only export it if it is needed for the external API. But that is not a critical item.

[billphipps]

---I'm wondering if we want to be including wh_settings.h in every header when it isn't needed externally

I decided to err on the side of caution. If EVERY .h file in the base library includes wh_settings.h first, then IF some change later needs a new CFG value, it will already be there. This also means if a customer includes a random wh_XXX.h file, we are assured those configs will be included.