You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
Repository navigation
Add Argon2 support to System.Security.Cryptography #19933
Reacted by Joseph Musser, Simon Opelt, Tony Ranieri, Ryan Elian, LeaFrock, mojtaba kaviani, Alexandre H.T.R. Bonfitto, Alex Rønne Petersen, Jason Nelson, Laurent Egbakou and 8 more
As far as I can tell, there's no support for Argon2 on Apple and Windows yet, which probably means this issue is dead in the water for now. (IIRC, the rule of thumb is that at least 2 platforms must support an algorithm before support is added to .NET.)
Argon2 depends on BLAKE2b, as well as doing a lot of its own low-level data manipulation (the compression function G and permutation function P, both of which say they're derived from BLAKE2b). So, it's at best a composed algorithm that depends on a primitive we don't have support for; or at worst a cryptographic primitive.
Either way, it's not something that our current cryptographic policy ("don't implement primitives, only use them from the underlying system libraries") permits us to do at this time.
Activity
bartonjs commentedon Jan 17, 2017
There are no plans to do so, no. But we'd be open to contribution (starting with an API proposal)
LeaFrock commentedon Nov 8, 2022
Golang has supported Argon2 since 2017... Maybe someone interested in this issue can refer to it.
https://github.com/golang/crypto/tree/master/argon2
Also, Argon2 has been standardized by RFC9106.
alexandrehtrb commentedon Dec 22, 2022
+1
C reference implementation: https://github.com/P-H-C/phc-winner-argon2
.NET wrapped implementation: https://github.com/alipha/csharp-argon2
alexrp commentedon Dec 25, 2022
Some relevant notes:
bartonjs commentedon Jan 4, 2023
Argon2 depends on BLAKE2b, as well as doing a lot of its own low-level data manipulation (the compression function G and permutation function P, both of which say they're derived from BLAKE2b). So, it's at best a composed algorithm that depends on a primitive we don't have support for; or at worst a cryptographic primitive.
Either way, it's not something that our current cryptographic policy ("don't implement primitives, only use them from the underlying system libraries") permits us to do at this time.
Neustradamus commentedon Jun 15, 2023
To follow
tats-u commentedon Jun 3, 2025
Java is considering to implement Argon2 in 26:
https://bugs.openjdk.org/browse/JDK-8253914
Neustradamus commentedon Jun 3, 2025
@tats-u: Good news! :)