RFR(S): 8233014: Enable ShowCodeDetailsInExceptionMessages per default.

[coleen.phillimore at oracle.com]

On 5/8/20 5:58 AM, David Holmes wrote:
> Hi Goetz,
>
> I have reviewed the CSR.
>
> On 8/05/2020 5:48 pm, Lindenmaier, Goetz wrote:
>> Hi David,
>>
>>> are we ready to flip the flag in 15? We have no feedback on the feature
>>> and would not likely get any for some time.
>> That is basically the point.
>> We have no negative feedback on the issue.
>>
>> As you can see in the bug, the feature is live in SapMachine
>> Since 13.0.2 and 11.0.6. That is quite a time to collect feedback.
>>
>> Also, if we enable it now,
>>    - we have the 15 development frame to get immediate feedback
>>      I don't want to enable it at the end of the development time
>>     (which would give more time to collect feedback from 14).
>>    - there is time to incorporate feedback before LTE 17.
>>      The time frame people are using 15 is long enough before
>>      release of LTE 17 so that I can react on feedback and improve
>>      the implementation.  If I enable it in 16, which will be in use
>>      while 17 is developed and stabilized, this might not be possible.
>>
>> So I think this is the right point of time to enable it.
>
> Okay that sounds quite reasonable.

Thanks Goetz for the justification.  I agree with this.  With the short 
release cycle, we're not likely to get feedback unless a customer is 
staying current on JDK versions and knows about the option. There's also 
probably a lag until people transition from JDK 14 to 15, so that gives 
less time to have users experience the option.  And so far, as you 
report the SAP users have had a positive experiences, and it's something 
we want for the platform.
>
>> For testing, as you know we have quite a test infrastructure.
>> I ran it through all these tests, which incorporate most jck tests
>> and all headless jtreg tests.
>> Also, if I remember correctly, the change has been run through your
>> testing landscape when we first pushed the feature to have it tested
>> in enabled mode.  But I have so many mails on this that I could not
>> find prove of this.
>>
>> So I'd assume there are no correctness problems. There might be a
>> new test somewhere that checks the content of the message
>> (before: null) and which needs to be adapted.
>
> Yeah I don't know what was done before. I'll run it through our 
> testing again while you wait for the CSR approval.
>

I ran our tier1 testing on this patch, but David, please see the JEP, I 
posted internal links to test results for the option on and off.  I also 
completed the security tasks for the JEP which has internal links as well.

https://bugs.openjdk.java.net/browse/JDK-8220715

Thanks,
Coleen


> Thanks,
> David
>
>> Best regards,
>>    Goetz.
>>
>>
>>
>>
>>
>>
>>> -----Original Message-----
>>> From: David Holmes <david.holmes at oracle.com>
>>> Sent: Monday, April 27, 2020 1:08 AM
>>> To: Lindenmaier, Goetz <goetz.lindenmaier at sap.com>; 'hotspot-
>>> dev at openjdk.java.net' <hotspot-dev at openjdk.java.net>
>>> Subject: Re: RFR(S): 8233014: Enable ShowCodeDetailsInExceptionMessages
>>> per default.
>>>
>>> Hi Goetz,
>>>
>>> On 27/04/2020 8:02 am, David Holmes wrote:
>>>> Hi Goetz,
>>>>
>>>> This will need a CSR request to change the default flag value.
>>>
>>> On that front, given this feature only just shipped in 14, on what 
>>> basis
>>> are we ready to flip the flag in 15? We have no feedback on the feature
>>> and would not likely get any for some time.
>>>
>>>> Otherwise change looks okay to me.
>>>
>>> With regards to testing however, please ensure that turning this on 
>>> does
>>> not inadvertently disrupt any tests. Passing jdk-submit is a minimal
>>> requirement here and more extensive testing is desirable.
>>>
>>> Thanks,
>>> David
>>>
>>>> Thanks,
>>>> David
>>>>
>>>> On 25/04/2020 12:36 am, Lindenmaier, Goetz wrote:
>>>>> Hi,
>>>>>
>>>>> I would like to enable JEP 358 Helpful NullPointerExceptions by 
>>>>> default:
>>>>>
>>>>> http://cr.openjdk.java.net/~goetz/wr19/8233014-
>>> enable_NPE_per_default/02/
>>>>>
>>>>> Please review.
>>>>>
>>>>> The change passed our testing and jdk-submit.
>>>>>
>>>>> Best regards,
>>>>>    Goetz.
>>>>>