Astra Linux crypt variants using GOST R 34.11-94 or GOST R 34.11-2012

[magnumripper]

The former is 256-bit. The latter is 256 or 512-bit and also known as Streebog. The formats exactly match Drepper's sha256crypt and sha512crypt, with the hash primitive replaced with one of the GOST alternatives.

gost94crypt ($gost94hash$)
streebog256crypt ($gost12256hash$)
streebog512crypt ($gost12512hash$)

[solardiz]

OK, let's get these in, although I doubt all 3 are in active use.

[solardiz]

I guess this unroll produces truly a lot of code. We should be able to optimize this without so much code later (not in this PR).

[magnumripper]

What's with the circleci stall? Should I ignore it and merge? I can't seem to click it for looking into it.

We should be able to optimize this without so much code later (not in this PR)

Yeah, best have a version with readable code in history before going crazy with it. Actually, I based these CPU formats on pristine versions of sha256crypt and sha512crypt formats from our Git history, that was handy for sure (before that, I tried basing it on the current versions with Jim's tweaks but something failed and I lost interest).

[solardiz]

What's with the circleci stall? Should I ignore it and merge? I can't seem to click it for looking into it.

Same here. I think yes, you should ignore it and merge. Thank you!

[claudioandre-br]

What's with the circleci stall?

The "stall" means that an administrator must review the (branch protection) rules. As seen in:

Repository administrators can add constraints like this to branches using branch protection rules. For more information, see "About protected branches."

The ci/circleci: non-SIMD should no longer be required.

[solardiz]

The ci/circleci: non-SIMD should no longer be required.

I've just removed it from required checks, and somehow this resulted in it disappearing from pending checks completely. I have no idea why it was getting stalled lately nor why it disappeared completely when not required.