RFR: 8214552: Resolve clash between 4947890 and 8130266
Phil Race
philip.race at oracle.com
Fri Nov 30 20:15:16 UTC 2018
On 11/30/18 11:54 AM, Roger Riggs wrote:
> Hi Phil,
>
> The patch looks fine.
Ok, thanks.
>
> Where does it die? SystemProps should skip null values in putIfAbsent().
You are right it checks and that explains why I didn't see it when I tested.
I saw the NPE only after I noticed the code and it seems was just from
my debugging I
quickly added since System.out is likely not defined at this point ...
Anyway we still need this fix.
Thanks,
-phil.
>
> Sorry for the turbulence.
>
> Roger
>
>
>
> On 11/30/2018 02:43 PM, Phil Race wrote:
>> Roger,
>>
>> I just pushed 8130266 which had the CSR approved a couple of days ago
>> only
>> to find that it clashed badly with the change you made in 4947890 to
>> reduce
>> the cost of system property initialisation.
>> It appears that the new code there is not happy if there is no value
>> for the
>> system property.
>>
>> I think java.awt.printerjob just needs to be removed from the new code :
>>
>> If this looks right to you I need to push ASAP :
>>
>> Bug: https://bugs.openjdk.java.net/browse/JDK-8214552
>> Webrev: http://cr.openjdk.java.net/~prr/8214552/
>>
>> Separately I think the SystemProps code ought to consider whether
>> it ought to error out the launcher if a property is not defined ..
>>
>> -phil.
>>
>>
>
More information about the core-libs-dev
mailing list