Code review request, CR 7174244, NPE in Krb5ProxyImpl.getServerKeys()

Weijun Wang weijun.wang at oracle.com
Wed Jun 6 06:11:49 PDT 2012


Hi Xuelei

I don't quite understand the bug report. Certainly it's good to prevent 
that NPE, but why must the cipher suites be ordered? It seems it's not 
ordered before the new HashMap. Is this a regression or an RFE? You'd 
better add some explanation in the bug report.

Thanks
Max

On 06/06/2012 08:27 PM, Xuelei Fan wrote:
> Hi Weijun,
>
> Would you please review the fix for CR 7174244.
>      http://cr.openjdk.java.net./~xuelei/7174244/webrev.00/
>
> There are mainly two updates in the fix. One is to have
> Krb5ProxyImpl.getServerKeys()  to check null return value of
> Krb5Util.getServiceCreds(). The other one is to have cipher suite order
> to be independent from HashMap implementation.  TreeSet is used to
> replace ArrayList so that the cipher suites order are sorted automatically.
>
> Thanks,
> Xuelei



More information about the security-dev mailing list