RFR: 8208583: Better management of internal KeyStore buffers

[Weijun Wang]

KeyProtector.java:

 113         pbeKeySpec.clearPassword();

 You can also put this into the finally block.

 189             Arrays.fill(plain, (byte) 0x00);

 Can this be in finally?

JavaKeyStore.java:

 149         Arrays.fill(passwordBytes, (byte) 0x00);

 In other cases, you call it in a finally block. Unnecessary here?

 (Oops, every comment is about finally.)

KeyProtector.java:

 123     public KeyProtector(byte[] password)

 How about just "public KeyProtector(byte[] passwordBytes)"?

> On Aug 2, 2018, at 7:38 PM, Seán Coffey <sean.coffey at oracle.com> wrote:
> 
> No - no problem at all. Some extra exception handling but probably best for the long run.

I wonder why DestroyedFailedException was a checked exception, what can we do if it's thrown?

Thanks
Max

> 
> http://cr.openjdk.java.net/~coffeys/webrev.8208583.v3/webrev/index.html
> 
> regards,
> Sean.
> 
> On 02/08/2018 02:13, Weijun Wang wrote:
>>> 1.
>>> 
>>> I wasn't able to rename to destroy since that method is reserved for the Destroyable interface. I've gone with destroyKey.
>>> 
>> Sorry I wasn't clear but this is exactly what I meant. SecretKey implements Destroyable so you don't need to define sKey as PBEKey. Does it make any problem?
>> 
>> 
>