Closed abkarcher closed 4 months ago
Hi @abkarcher, I'll take a look into this, thanks for letting us know.
Hi @abkarcher, do you know if this is a new behavior as of release v1.1.0.34? Or pre-existing issue?
Also, @szendros seems to have been able to test against ECDSA siggen/sigver using SHAKE successfully I believe. See https://github.com/usnistgov/ACVP-Server/issues/301#issuecomment-2133352998
Hi,
Sorry, I seem to have missed the notification for this. I am not aware if this issue existed prior to .34, I did not notice it until after.
I am seeing all sorts of failures testing siggen/sigver with a basic implementation with SHAKE, with some specific test groups working fine. I will continue to look into it on my end.
Thanks, Andrew
K, thanks Andrew.
Hi @abkarcher,
This is indeed a bug I added with the last hot fix, I have refactored and fixed the bug. It will go out with the next release.
As far as usage goes, this is actually just an issue with it being set incorrecly, but not used. So, in this case, the setting is showing as True, but it's actually False. So, the test data should be fine.
Thanks, Joel
Thanks!
environment Demo
testSessionId 514488
vsId 2358750
This test session was requested with componentTest: false for all vector sets. However, the ECDSA sigver test groups have true for componentTest when using SHAKE hash algs.
Might anyone be able to confirm if they have been able to pass ECDSA siggen/sigver testing using SHAKE in general?
Thanks, Andrew