Hi Sven,
I created https://bugs.openjdk.java.net/browse/JDK-8212795 to track this
issue. NB is using the internal API that is unused in JDK and hence no
test uncovers this regression.
I have a patch that passes a map of key to values to
CompositeDataSupport rather than passing two synchron
I assume it's a simple ThreadInfo then a simple text form would work; if
not, you can use the format NB use if you have the reader.
Mandy
On 10/22/18 7:53 AM, Sven Reimers wrote:
Hi Mandy,
will try to get a sample to you.
Which format do you prefer?
Thanks for helping
-Sven
Mandy Chung ma
Hi Mandy,
will try to get a sample to you.
Which format do you prefer?
Thanks for helping
-Sven
Mandy Chung schrieb am So., 21. Okt. 2018, 23:26:
> Hi Sven,
>
> Do you have a sample ThreadInfo to reproduce this issue? That will be
> helpful. JMX monitoring is not impacted as far as I can
Hi Sven,
Do you have a sample ThreadInfo to reproduce this issue? That will be
helpful. JMX monitoring is not impacted as far as I can tell. NB
profiling is using this internal API
`sun.management.ThreadInfoCompositeData::toCompositeData` which may be
unused since JMX supports MXBean.
Ma
Hi,
one more problem to go
After applying the proposed fix for the
ThreadINfoCompositeData.compositeTyoe() the following exception is raised:
java.lang.reflect.InvocationTargetException
at java.base/jdk.internal.reflect.NativeMethodAccessorImpl.invoke0(Native
Method)
at
java.base/jdk.internal.re