Feat/pkcs11 3.0

[Direktor799]

Replace pkcs11.h with https://github.com/latchset/pkcs11-headers/blob/main/public-domain/3.0/pkcs11.h, which was mentioned in #156.

• Use latchset PKCS#11 3.0 header
• Manually specify CK_UNAVAILABLE_INFORMATION due to a bug(i believe) in bindgen
• Add 3.0 macro definations
• Enable PKCS11_DEPRECATED for backward compatibility with 2.4
• Deprecate EC_NAMEDCURVE, replace it with EC_OID

[hug-dev]

Thanks a lot! Excellent as well if this is compatible with 2.4. A few general comments but I am okay otherwise with it!

[hug-dev]

Feel free to resolve the last comment, if it's not something you were planning to investigate. I think it's fine to go as it is, and I can open an issue for this specific item to be looked at later :)

[Direktor799]

Feel free to resolve the last comment, if it's not something you were planning to investigate. I think it's fine to go as it is, and I can open an issue for this specific item to be looked at later :)

I think it's mainly because bindgen doesn't care integer length(only cares about signedness) when generating code for macros, see https://github.com/rust-lang/rust-bindgen/blob/07bbd04ea5f386aa3b83e792548127e2a2e33c43/bindgen/codegen/mod.rs#L705-L709

There are some issues mentioning bindgen generated wrong types with macros, but no issue related to this specific ~0UL case.

• Macro defining a constant with unsigned long suffix generates unsigned integer rust-lang/rust-bindgen#923
• Integer suffixes discarded jethrogb/rust-cexpr#16

Do you want me to open up a new issue?

[Direktor799]

@hug-dev also what do you suggest to do with the misspelling in the new header..

[hug-dev]

Thanks for research! I think it's fine as it is, this PR will be the history in case someone wants to look this up.

@hug-dev also what do you suggest to do with the misspelling in the new header..

I think we can add the folder in the ignored here to make it work!

[wiktor-k]

LGTM 👍