Fix encryption and decyption for streams

[lo1ol]

Hello,

Our package depends on your package v. 4.1.1. Recently, we noticed that encryption and decryption doesn't work properly for streams. After viewing the sources, we've found the incorrect using of C_EncryptUpdate function inside Encrypt/Decrypt override for streams. The problem arises when length of output buffer doesn't equal to length of input buffer (such as inside my pkcs11 lib librtpkcs11ecp). So, I've creating a patch which resolve this problem by getting the length of the output buffer before encryption and decryption.

If the patch is right, I can create the same patch for 5.1.1 version (problem doesn't still resolve for it). But I'm waiting for that you will releases a new version 4.1.2 with this patch.

[jariq]

Thanks for the PR. I'll take a look at it soon:tm: 😉

[lo1ol]

Hello, guys!

Did you not forget about issue?...

[jariq]

Nope. I'm on it 👍🏻

[jariq]

The problem arises...

@lo1ol can you please describe the problem in more details?
What encryption mechanism were you using?
Did C_EncryptUpdate throw an exception?
Was ciphertext (encryptedPart) longer than plaintext (part) ?

[lo1ol]

What encryption mechanism were you using?

The problem appears when I'm using CKM_GOST28147 mechanism and pass data using streams.

Was ciphertext (encryptedPart) longer than plaintext (part) ?

No.

Did C_EncryptUpdate throw an exception?

Yes. First iteration return zero length output buffer. And next iteration return the buffer with length more then partLen. I don’t know the reason of this behaviour, but the implementation of CKM_GOST28147 works this way.

[jariq]

So C_EncryptUpdate returned CKR_BUFFER_TOO_SMALL ?

[lo1ol]

I don't remember exact return code, but probably that.

[lo1ol]

You solution is more efficient:)

[jariq]

Can you please test my alternative implementation from stream-encryption branch and check whether it solves your problem with CKM_GOST28147? It should be more efficient for other mechanisms than code in this PR.

[lo1ol]

No problem. I'm not close to a computer right now. I will give you feedback tomorrow.

[lo1ol]

I have tested your solution. It's work!

[jariq]

Thanks for testing. I'll release fix asap ™️ 😉

[lo1ol]

Thank you so much!

[jariq]

Closing this PR as the problem was resolved in #172.

[jariq]

@lo1ol the fix is available in freshly published Pkcs11Interop 4.1.2