[JDK 8]Code review request 8022228, Intermittent test failures in sun/security/ssl/javax/net/ssl/NewAPIs

[Xuelei Fan]

On 8/22/2013 10:09 AM, Weijun Wang wrote:
> 
> 
> On 8/22/13 9:46 AM, Xuelei Fan wrote:
>> new webrev: http://cr.openjdk.java.net/~xuelei/8022228/webrev.01/
>>
>> On 8/22/2013 9:30 AM, Weijun Wang wrote:
>>> SessionTimeOutTests.java:
>>>
>>> 411             local.initCause(remote);
>>>
>>> What if local already has a cause? Will this overwrite it?
>>>
>> Yes.  It's intent to override the current cause, but with the current
>> stacks.  It's a approach to show some information of both side.
> 
> I don't know if the current stack is enough for you to debug. Why not
> also use addSuppressed here? Or you can just printStackTrace.
> 
Good suggestions.  This template is wildly used in JSSE testing.  I will
consider it later.

Thanks,
Xuelei

> --Max
> 
>>
>>> 427                 exception.addSuppressed(startException);
>>>
>>> Is it possible startException is null?
>>>
>> Good catch!  It's a safer code to consider this case.
>>
>>> Same for the other test.
>>>
>>> Also, if you have to do all these checks in every SSL test, can they be
>>> added into a library?
>>>
>> Maybe we can design a super class for this kind of tests later.
>>
>> Thanks,
>> Xuelei
>>
>>> Thanks
>>> Max
>>>
>>> On 8/22/13 9:18 AM, Xuelei Fan wrote:
>>>> Hi Weijun,
>>>>
>>>> Please review this regression test bug fix for JDK 8.
>>>>
>>>> Webrev: http://cr.openjdk.java.net/~xuelei/8022228/webrev.00/
>>>>
>>>> The timeout issue of SessionTimeOutTests.java is still unknown.  This
>>>> fix is just try to use new SSL testing template so as to expose more
>>>> information.
>>>>
>>>> The issue of SessionCacheSizeTests is that the client side may kick off
>>>> a connection before server is ready to listen on a particular port.
>>>>
>>>> Thanks,
>>>> Xuelei
>>>>
>>