RFR: 8203410: Zero: Disable jfr feature by default

Erik Joelsson erik.joelsson at oracle.com
Fri May 18 16:50:35 UTC 2018


Looks good (assuming you have verified that it behaves as it should)

/Erik


On 2018-05-18 09:27, Severin Gehwolf wrote:
> Hi Erik,
>
> On Fri, 2018-05-18 at 08:41 -0700, Erik Joelsson wrote:
>> The change makes sense, but I think I would prefer if the conditional
>> was based on jvm variant rather than the target cpu, like this:
>>
>> if ! HOTSPOT_CHECK_JVM_VARIANT(zero); then
> Thanks for the review!
>
> New webrev:
> http://cr.openjdk.java.net/~sgehwolf/webrevs/JDK-8203410/webrev.02/
>
> Cheers,
> Severin
>
>> /Erik
>>
>>
>> On 2018-05-18 06:57, Severin Gehwolf wrote:
>>> Hi,
>>>
>>> Currently the jfr (Flight Recorder) feature is being built by default
>>> for Zero JVMs. However, it's not clear whether there will be jfr
>>> support for the Zero variant JVM. At this point, when
>>> StartFlightRecording option is being used for a Zero JVM it
>>> asserts/crashes. It seems more appropriate to disable the feature for a
>>> default build and require one to enable the feature explicitly via:
>>>
>>> configure --with-jvm-feature=jfr --with-jvm-variants=zero
>>>
>>> We can revisit this change of default once there is good Zero+JFR
>>> support, which currently isn't the case.
>>>
>>> Bug: https://bugs.openjdk.java.net/browse/JDK-8203410
>>> webrev: http://cr.openjdk.java.net/~sgehwolf/webrevs/JDK-8203410/webrev.01/
>>>
>>> Testing: Verified that the StartFlightRecording option is unrecognized
>>> for a default --with-jvm-variants=zero build.
>>>
>>> Thoughts?
>>>
>>> Thanks,
>>> Severin
>>




More information about the build-dev mailing list