[9] RFR 8043349: Consider adding aliases for Ucrypto algorithm-only Cipher transformations.

Valerie Peng valerie.peng at oracle.com
Wed Dec 17 23:58:11 UTC 2014


Thanks for the review~

Well, as for allowing the missing options, I am not sure how useful or 
how frequently people use them. As for me, I find them a bit too 
ambiguous to my liking. I think there are other more useful bugs to fix 
than spending time on this. So far, no bug report regarding support for 
missing options, thus, no plan for looking into this (yet).

As for not including the Supported* values, it's mostly due to 
priorities and effort trade-offs. For example, updating the Ucrypto 
provider w/ latest list of supported Solaris algos would happen before 
adding them...

Thanks,
Valerie

On 12/17/2014 3:36 PM, Bradford Wetmore wrote:
> I think this is ok.
>
> I have a recollection our Cipher.getInstance() provider selection 
> mechanism (getTransforms()) allows for missing options: 
> "AES//NoPadding"  "AES/ECB/"  But it's been a while since I've looked 
> at this.  These ucrypto values look like they must be completely 
> specified.  Is that something to look into for down the road?
>
> One other point, is there a reason why we're not including the 
> Supported* values in ucrypto?
>
> Brad
>
>
>
> On 12/17/2014 3:18 PM, Valerie Peng wrote:
>> Hi, Brad,
>>
>> Can you please review this straightforward Ucrypto fix? This is about
>> adding aliases to the AES and RSA ciphers of OracleUcrypto provider.
>>
>> Webrev: http://cr.openjdk.java.net/~valeriep/8043349/webrev.00/
>> Bug: https://bugs.openjdk.java.net/browse/JDK-8043349
>>
>> Thanks,
>> Valerie



More information about the security-dev mailing list