RFR(XS): 8204563: UseAppCDS obsolete message confusing

David Holmes david.holmes at oracle.com
Mon Jun 11 21:38:26 UTC 2018

On 12/06/2018 2:17 AM, Calvin Cheung wrote:
> Hi David,
> Thanks for your review.
> On 6/10/18, 6:14 AM, David Holmes wrote:
>> On 9/06/2018 4:50 AM, Calvin Cheung wrote:
>>> bug: https://bugs.openjdk.java.net/browse/JDK-8204563
>>> webrev: http://cr.openjdk.java.net/~ccheung/8204563/webrev.00/
>>> This change is for clarifying the obsolete message for the UseAppCDS 
>>> option.
>> Nit: not withstanding the edits made to the bug report the correct way 
>> to refer to this is as the "obsoletion message".
> I didn't change the synopsis of the bug but the word "obsoletion" 
> couldn't be found in dictionary.com.

Obsoletion is pefectly correct:


> Maybe it's more correct to say something like "the message for 
> obsoleting the UseAppCDS option is confusing" ?

Rephrasing that way is also correct.

>>> Existing:
>>>      Ignoring option UseAppCDS; support was removed in 11.0
>>> New:
>>>      Ignoring obsolete option UseAppCDS; AppCDS is automatically enabled
>>> Ran hs-tier1 and hs-tier2 testing successfully.
>> I'd normally object to special-casing this way but as this will only 
>> need to be present for 4 weeks or so it's okay. In the future we may 
>> need a way to customize messages in a more generic way.
> Agreed. BTW, the process_argument() is for handling the -XX:xxxx 
> options. The obsolete -Xprof is being handled in a separate block of 
> code in arguments.cpp.

Yes we only deal with -XX deprecation/obsoletion/expiration. Any -X flag 
that isn't handled by the launcher has to be handled as a special case.


> thanks,
> Calvin
>> Looks good.
>> Thanks,
>> David
>>> thanks,
>>> Calvin

More information about the hotspot-runtime-dev mailing list