Skip to content

Add ASCON#953

Open
xmamalou wants to merge 1 commit intocisco:libacvp_2_3_0-throttlefrom
xmamalou:main
Open

Add ASCON#953
xmamalou wants to merge 1 commit intocisco:libacvp_2_3_0-throttlefrom
xmamalou:main

Conversation

@xmamalou
Copy link

@xmamalou xmamalou commented Feb 6, 2026

Add handling for ASCON

@abkarcher
Copy link
Contributor

Hello,
Thank you for the contribution! Before I take a look at this, could you please remove the formatting/style changes unrelated to ASCON, and ensure consistent code style is used in the ASCON additions?

And then, if you could base this from the "libacvp_2_3_0-throttle" branch instead of main, that would be needed too.

Thanks again,
Andrew

@xmamalou xmamalou force-pushed the main branch 2 times, most recently from 50bda28 to 60eff14 Compare February 11, 2026 17:00
@xmamalou xmamalou force-pushed the main branch 6 times, most recently from 9aca533 to 392d1d4 Compare February 21, 2026 20:03
@xmamalou
Copy link
Author

xmamalou commented Feb 21, 2026

Hello, sorry for the long wait.

I have done the changes as requested. Sorry again for the formatting, Emacs was more zealous with the autoformatting than I expected and I hadn't noticed it had changed the entire file.

Best regards,
Christoforos

@xmamalou xmamalou changed the base branch from main to libacvp_2_3_0-throttle February 21, 2026 20:09
@abkarcher
Copy link
Contributor

Thanks! Taking a look now.

@xmamalou xmamalou force-pushed the main branch 3 times, most recently from 75a17fd to b757c54 Compare February 25, 2026 11:26
@abkarcher
Copy link
Contributor

Thank you! I pulled a branch from this PR to do some more testing, and noted a few more things.

  • I am hitting ACVP_INVALID_ARG errors when trying to build a registration for all ASCON algs. Something about acvp_cap_ascon_set_domain() is not functioning as expected. Instead of using switch case fall-throughs, can we restructure this to handle each PARM value for each cipher, similar to other algorithms?
  • I do not think the ACVP_ASCON_MODE parameters or enums are needed; in each case we always have the ACVP_CIPHER value, unless I am mistaken
  • The set_parm() and set_domain() ascon APIs for setting values should have bounds checks against ASCON_*_MAX values in acvp_lcl.h.
  • Nit: (Sorry for not catching this the first go around): can we rename the ASCON params to have "parm" in the middle instead of the end? e.g. instead of ACVP_ASCON_PAYLEN_PARM, maybe ACVP_ASCON_PARM_PAYLEN? to be more consistent with other algorithms

@xmamalou
Copy link
Author

Hello! I have implemented the changes you have requested.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants