RFR 8237218: Support NIST Curves verification in java implementation

Anthony Scarpino anthony.scarpino at oracle.com
Mon Feb 17 18:09:36 UTC 2020


The change looks fine.  I'm trusting that the existing Known Answer 
Tests are proving your verifySignedDigest() is correct.

You may want to comment in the code that your test depends on these 
method names.  I was going to suggest simplifying the all the 
verifySigned*() methods until I saw the test was dependent on it.

Tony


On 2/13/20 3:06 AM, Weijun Wang wrote:
> Webrev updated at
> 
>     http://cr.openjdk.java.net/~weijun/8237218/webrev.03/
> 
> The test is modified. Instead of adding a hacked ECDSASignature I'm using JDI to detect if the Java impl or the native impl is used. Two method names in ECDSASignature are modified to ease method lookup in the test.
> 
> Thanks,
> Max
> 
>> On Feb 11, 2020, at 7:52 PM, Weijun Wang <weijun.wang at oracle.com> wrote:
>>
>> Please take a review at
>>
>>    http://cr.openjdk.java.net/~weijun/8237218/webrev.02/
>>
>> A test is added that uses a patched ECDSASignature.java that exposes how the signature is verified.
>>
>> BTW, I also updated ECDSASignature.java a little to accept non SunEC keys, so that I can do some interop testing. If you believe this is unnecessary I can revert the change.
>>
>> Thanks,
>> Max
>>
> 



More information about the security-dev mailing list