Re: [CFRG] XDR in RFC8391

Hi David,

I agree, the superfluous commas and strange identifiers are easy to fix. Also having multiple discriminant values for one union arm is not a problem.

But RFC4506 requests at the end of a "case-spec" a full "declaration" and not only a single value. (only "void" would be allowed here). Therefore I ask for the binary encoding of such definition.

Thanks,
Andreas

From: David McGrew (mcgrew) <mcgrew@cisco.com>
Sent: Mittwoch, 2. Dezember 2020 22:03
To: Kretschmer, Andreas (T RDA CST SEA-DE) <andreas.kretschmer@siemens.com>
Cc: cfrg@irtf.org
Subject: Re: [CFRG] XDR in RFC8391

Hi Andreas,

I’m not an author, but I think I can help, please see inline:

On Dec 2, 2020, at 4:42 AM, Kretschmer, Andreas <mailto:andreas.kretschmer@siemens.com> wrote:

Hello,

I tried to use the XDR definitions given in RFC8391 Appendix A,B,C for generating a parser skeleton. Unfortunately I found some parts of the XDR definitions not compliant to the referred RFC4506:

- some Identifiers contain "/" and "-", RFC4506 allows only letter, digits and underbars

This is unfortunate, but it can be worked around easily by replacing the nonconformant characters with underbars, or some similar change.

- some enum bodies end with  ",}", RFC4506 requests "}" here
- some union definitions have incomplete declarations in the case-spec, e.g. the union xmss_ots_signature refers to the wotsp-sha2_256 without giving a type.

The meaning of the following snippet:

      union xmss_ots_signature switch (xmss_algorithm_type type) {
        case xmss-sha2_10_256:
        case xmss-sha2_16_256:
        case xmss-sha2_20_256:
          wotsp-sha2_256;
      }
Has the same meaning as

      union xmss_ots_signature switch (xmss_algorithm_type type) {
        case xmss-sha2_10_256:
          wotsp-sha2_256;
        case xmss-sha2_16_256:
          wotsp-sha2_256;
        case xmss-sha2_20_256:
          wotsp-sha2_256;
      }

That is, multiple discriminant values correspond to the same implied arm.

Now my questions:

- Is there a fixed formal correct version of the  RFC8391 XDR definitions available?

Not that I’m aware of.

-  Whats the binary representations of an union holding incomplete  declarations?

Could somebody please give me a  hint or a pointer to the answers?

Hope that helps.

David

Regards,
Andreas

_______________________________________________
CFRG mailing list
mailto:CFRG@irtf.org
https://www.irtf.org/mailman/listinfo/cfrg