[jdk9] RFR: 8158802: com.sun.jndi.ldap.SimpleClientId produces wrong hash code

Ivan Gerasimov ivan.gerasimov at oracle.com
Thu Jun 9 14:40:19 UTC 2016


Thanks Seán!


On 09.06.2016 17:22, Seán Coffey wrote:
> Looks fine to me Ivan. Nice find.
>
> Regards,
> Sean.
>
> On 09/06/2016 14:52, Ivan Gerasimov wrote:
>> Could someone please help review this?
>>
>> The essential part of the fix is quite simple:  i.e. replacing 
>> char/byte[].hashCode() with Arrays.hashCode(char/byte[]).
>> The regression test confirms that the change is good.
>>
>> With kind regards,
>> Ivan
>>
>>
>> On 06.06.2016 19:58, Ivan Gerasimov wrote:
>>> Hello!
>>>
>>> It was noticed that objects of an internal class SimpleClientId that 
>>> are equal (in terms of equals() method) can have different hash codes.
>>> This happens when the password is represented with an array, so 
>>> passwd.hashCode() produces unique value even when the content is same.
>>>
>>> Some minor cleanup was done in the sibling file to remove an 
>>> intermediate hash value.
>>>
>>> Would you please help review the fix?
>>>
>>> BUGURL: https://bugs.openjdk.java.net/browse/JDK-8158802
>>> WEBREV: http://cr.openjdk.java.net/~igerasim/8158802/00/webrev/
>>>
>>> With kind regards,
>>> Ivan
>>>
>>>
>>
>
>



More information about the core-libs-dev mailing list