KAS ECC Ar3 Full MQV "Specified argument was out of the range of valid values." after addition of KMAC in KC list

[mwcw]

Hi,

We are getting:
[{"acvVersion":"1.0"},{"error":"Specified argument was out of the range of valid values. (Parameter \u0027capacity\u0027)"}]
After adding:

"KMAC-256": {
                  "macLen": 512,
                  "keyLen": 256
                },

and

"KMAC-128": {
                  "macLen": 256,
                  "keyLen": 128
                },

to the capabilities request.

I could not find "capacity" in the spec.

I have attached the request and responses.

Archive.zip

Please let me know if you need any more information.

MW

[Kritner]

I'll take a look at this shortly, I have a pretty good idea of where the issue is. I put some additional guards into our kmac implementation because of a way a Kas kdf was invoking it incorrectly, apparently that same incorrect behavior existed for key confirmation as well. I'll put a correction in and update when the issue is resolved. Thanks!

…

On Thu, Dec 10, 2020, 12:28 AM MW ***@***.***> wrote: Hi, We are getting: [{"acvVersion":"1.0"},{"error":"Specified argument was out of the range of valid values. (Parameter \u0027capacity\u0027)"}] After adding: "KMAC-256": { "macLen": 512, "keyLen": 256 }, and "KMAC-128": { "macLen": 256, "keyLen": 128 }, to the capabilities request. I could not find " I have attached the request and responses. Archive.zip <https://github.com/usnistgov/ACVP/files/5670448/Archive.zip> Please let me know if you need any more information. MW — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#1090>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ACDIDB5RHHUPR43VK2PAAG3SUBL7FANCNFSM4UUP7PTA> .

[Kritner]

Looks like my suspicion was correct. The fix put in (and subsequent guards within our KMAC implementation) from usnistgov/ACVP-Server#43 caused this response from the server. Rather than create incorrect values due to a bad KMAC construction it threw an exception - overall better behavior IMO.

Thanks for the issue report, I'll get a fix in.

[Kritner]

The change is on demo, I'll update again once on production. I re-ran the generation request from the vector set in the original issue, and the files have been generated.

[Kritner]

This change is now on demo and production in release v1.1.0.14 hotfix 2