Re: RFR: 8248379: Handshake closures for JVMTI monitor functions lack of some validations

[serguei.spit...@oracle.com]

Hi Yasumasa,

I see, some VM_op's also have this check:

1546   ThreadsListHandle tlh;
1547   if (jt != NULL && tlh.includes(jt)

I wonder if it make sense to add as well.

Otherwise, it looks good to me.

Thanks,
Serguei

On 6/26/20 00:03, Yasumasa Suenaga wrote:

Hi all,

Please review this change.

  JBS: https://bugs.openjdk.java.net/browse/JDK-8248379
  webrev: http://cr.openjdk.java.net/~ysuenaga/JDK-8248379/webrev.00/

JDK-8242425 introduces to migrate to thread local handshake from VM operation for GetOwnedMonitorInfo, GetOwnedMonitorStackDepthInfo, and GetCurrentContendedMonitor JVMTI functions. However it lacks of validations for thread state and thread oop of the target.

This change has been tested on submit repo and serviceability/jvmti, serviceability/jdwp vmTestbase/nsk/jvmti, vmTestbase/nsk/jdi vmTestbase/nsk/jdwp.
On submit repo, tools/javac/7118412/ShadowingTest.java and java/foreign/TestMismatch.java were failed (mach5-one-ysuenaga-JDK-8248379-20200626-0503-12110818). However they do not seems to be related to this change.
(Both tests have been passed on my Linux AMD64)


Thanks,

Yasumasa