Hey Jeffrey,
I took a look at the implementation and found a bunch of points which I
found optimization worthy. An according PR will be provided shortly.
* ChaCha is capable of using user-defined amounts of rounds. The
current implementation doesn't allow that (as opposed to Salsa20).
I've fixed it by introducing an additional info class that is like
yours and by modifying the existing one to not be templatified by
round number. The benefit of this change is that the user can
dynamically at run-time decide which variant to use rather than at
compile-time (or by using ugly pointers himself)
* You've copy-pasted the definitions for encryption and decryption.
It's easier to just define Encryption to be the same as the
decryption (as was done with Salsa). The benefit of this change is
the increased readability of the code.
I've applied all these changes and it didn't break anything (for me).
BR
JPM
Am 21.04.2016 um 18:26 schrieb Jeffrey Walton:
> Hi Everyone,
>
> We added ChaCha8, ChaCha12 and ChaCha20 this afternoon. I could not
> find test vectors for the seek operation, so its currently disabled
> and throws a NotImplemented() exception.
>
> Commits at:
>
> *
> http://github.com/weidai11/cryptopp/commit/53b51331a4e1d9664bc93a0f1597fc2e70506234
> *
> http://github.com/weidai11/cryptopp/commit/38f6c33789e1a029595564dd5663622b15410d35
> *
> http://github.com/weidai11/cryptopp/commit/90bb31360d556d182bb8e4c9e24901563d64878c
>
> There are no self tests, but there are test vectors. the test vectors
> can be run with:
>
> ./cryptest.exe tv all # All test vectors
> ./cryptest.exe tv chacha # ChaCha8, ChaCha12 and ChaCha20
>
> There's an opportunity for an SSE2 implementation if anyone is
> interested. Salsa20 has and its inline ASM had too many problems with
> the GCC 5.x compilers, so I did not want to copy/paste it and then
> endure the subtle bugs. If anyone has time for an SSE2 implementation
> using intrinsics (_mm_X), then I'd be happy to check it in.
>
> Jeff
>
> --
> --
> You received this message because you are subscribed to the "Crypto++
> Users" Google Group.
> To unsubscribe, send an email to
> [email protected].
> More information about Crypto++ and this group is available at
> http://www.cryptopp.com.
> ---
> You received this message because you are subscribed to the Google
> Groups "Crypto++ Users" group.
> To unsubscribe from this group and stop receiving emails from it, send
> an email to [email protected]
> <mailto:[email protected]>.
> For more options, visit https://groups.google.com/d/optout.
--
--
You received this message because you are subscribed to the "Crypto++ Users"
Google Group.
To unsubscribe, send an email to [email protected].
More information about Crypto++ and this group is available at
http://www.cryptopp.com.
---
You received this message because you are subscribed to the Google Groups
"Crypto++ Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
For more options, visit https://groups.google.com/d/optout.
