RFR (XS): 8025842: Convert warning("Thread holding lock at safepoint that vm can block on", ...) to fatal(...)

Vladimir Ivanov vladimir.x.ivanov at oracle.com
Wed Oct 2 12:18:47 PDT 2013


http://cr.openjdk.java.net/~vlivanov/8025842/webrev.00/
1 line changed: 0 ins; 0 del; 1 mod

Warnings are mostly ignored during testing - tests rarely looks at VM 
output. But this check may reveal a serious problem which is hard to 
diagnose (possible deadlock between VM thread & suspended Java thread).

If it is converted to an assert, the crash log will contain enough 
information to find the place where VM lock is acquired, since the check 
is performed on a thread which hold the lock.

The change doesn't affect product - the check is guarded by #ifdef ASSERT.

Inspired by JDK-8023461 [1]

Testing: JPRT, VM testbase.

Reviewed-by: ?

Thanks!

Best regards,
Vladimir Ivanov

[1] https://bugs.openjdk.java.net/browse/JDK-8023461


More information about the hotspot-dev mailing list