Skip to content

Conversation

magnumripper
Copy link
Member

This currently applies to AES-XTS, used by diskcryptor and truecrypt.

Also separate Truecrypt kernel from the (potentially) shared PBKDF2-RIPE- MD160 code for good measure.

Nvidia 550.144.03 bugs out on this for truecrypt so we revert that format to bitsliced AES (it doesn't matter for speed).

This currently applies to AES-XTS, used by diskcryptor and truecrypt.

Also separate Truecrypt kernel from the (potentially) shared PBKDF2-RIPE-
MD160 code for good measure.

Nvidia 550.144.03 bugs out on this for truecrypt so we revert that format
to bitsliced AES (it doesn't matter for speed).
@magnumripper
Copy link
Member Author

magnumripper commented Aug 20, 2025

This was originally #5818. After that one was merged I paniced after seeing regressions and crashes so I reverted it in #5820 but I later realised that there was nothing wrong with it - I was juggling with too many branches and got confused.

Anyway this code here (even sligthly improved since #5818) is what I've been running with for my tests since - it works fine and is even tested with debug output for verifying that work sizes end up correct vs. memory constraints.

@magnumripper magnumripper requested a review from solardiz August 20, 2025 23:16
@magnumripper
Copy link
Member Author

Tested a lot, works fine, ready for merge.

Copy link
Member

@solardiz solardiz left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I skimmed, but I mostly trust your testing.

@magnumripper magnumripper merged commit cfe2e31 into openwall:bleeding-jumbo Aug 23, 2025
32 of 33 checks passed
@magnumripper magnumripper deleted the aesxts branch August 23, 2025 07:16
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