Closed abkarcher closed 1 year ago
Hi @abkarcher,
At first look, I think you might be right on this. But let me dig deeper to verify that I'm understanding correctly what I'm looking at.
Thanks,
Ben
Hi,
Is there any update on this? I only ask because it will require me to redo some libacvp code in the coming month or two if the raw Z value is required.
Thanks, Andrew
Yes! We plan to have a fix for this in the next week or two. We'd like to send a new release to Demo next week if possible. Sorry for the wait on this.
Thanks for the update 😄
Andrew
environment Demo
testSessionId 259262
Hello,
I have been trying to debug a harness implementation of KAS-IFC, specifically an issue where some VAL tests with a "responder" role were failing with the reason "FailChangedZ". Libacvp is detecting that the tests passed.
I think I may have isolated a possible cause, related to (pending) changes I made after https://github.com/usnistgov/ACVP/issues/1338.
After that conversation, I have changed the library side of libacvp to make comparisons against hashZ when a hash function is provided, and ignore the provided Z value; the Z value seemed to be redundant and caused extra complexity in the client.
However, with this particular failure reaason "FailChangedZ", my idea is that the server is altering the Z value to cause a failure in the VAL test, but not altering the hashZ value. Thus, I am not able to pass the test properly. If this is the case, could the behavior be modified so the hashZ value also gets modified?
I could be wrong; I have been looking at this code for a bit too long now :) Any help isolating this issue would be greatly appreciated!
Thanks, Andrew