RFR: 8330842: Support AES CBC with Ciphertext Stealing (CTS) in SunPKCS11 [v3]

Francisco Ferrari Bihurriet fferrari at openjdk.org
Tue Jun 4 12:53:11 UTC 2024


On Tue, 4 Jun 2024 02:55:30 GMT, Martin Balao <mbalao at openjdk.org> wrote:

>> src/jdk.crypto.cryptoki/share/classes/sun/security/pkcs11/P11Cipher.java line 922:
>> 
>>> 920:                             0, out, outOfs, outLen);
>>> 921:                 }
>>> 922:                 if (paddingObj != null) {
>> 
>> Why not put this if-block inside an else-block of the `if (blockMode == Mode.CTS),` so it's clear that CTS mode won't use paddingObj?
>
> I'd rather not add a level of indentation to the else-block, but perhaps we can add an else-if block to the `paddingObj != null` block. @franferrax, what do you think?

I wouldn't increase indentation either. Regarding adding an `else if`, I'm neither for nor against it, so I can do that if that's your preference.

I agree that added here in line **922** of encryption case, it would make it clearer that CTS mode won't use `paddingObj`:

https://github.com/openjdk/jdk/blob/997777e86c6fa03f070dcf0f219813c11cb480ce/src/jdk.crypto.cryptoki/share/classes/sun/security/pkcs11/P11Cipher.java#L917-L940

However, it would also introduce an asymmetry with the decryption case in line **965**, where we can't do the same, since the code inside the `else` of line **983** must also be executed in CTS mode:

https://github.com/openjdk/jdk/blob/997777e86c6fa03f070dcf0f219813c11cb480ce/src/jdk.crypto.cryptoki/share/classes/sun/security/pkcs11/P11Cipher.java#L957-L987

----

NOTE: this discussion also applies to the [same block in the `ByteBuffer` version of `P11Cipher::implDoFinal()`](https://github.com/openjdk/jdk/blob/997777e86c6fa03f070dcf0f219813c11cb480ce/src/jdk.crypto.cryptoki/share/classes/sun/security/pkcs11/P11Cipher.java#L1030-L1104): an `else if` can be introduced line **1036** but not in line **1079**.

-------------

PR Review Comment: https://git.openjdk.org/jdk/pull/18898#discussion_r1625948565



More information about the security-dev mailing list