ot_present: fix ot_present_decrypt

[luismarques]

No description provided.

[AlexJones0]

This looks correct to me - decrypting block ciphers should apply the inverse of the the P-Box, and then the inverse of the S-box.

As a nit, I think you want the PR title to be

ot_present: fix ot_present_decrypt

or similar to match the convention we've been using in this repo recently.

[AlexJones0]

For now you can ignore any Earlgrey regression failures on rsa_3072_verify_functest_hardcoded_* by the way, I have a PR to fix this (#251) but that itself is pending on upstream OpenTitan fixes.

[rivos-eblot]

LGTM. Have you checked whether the same fix is needed for otptool.py?
If not I'll do it, I do not remember which tool I cross checked with another :-(

[AlexJones0]

Have you checked whether the same fix is needed for otptool.py?
If not I'll do it, I do not remember which tool I cross checked with another :-(

As far as I can tell otptool.py uses python/qemu/ot/util/present.py which is correct.

[rivos-eblot]

Have you checked whether the same fix is needed for otptool.py?
If not I'll do it, I do not remember which tool I cross checked with another :-(

As far as I can tell otptool.py uses python/qemu/ot/util/present.py which is correct.

Ok thanks for checking.