RFR 8213150: Add verification for locking by VMThread

[Robbin Ehn]

> MonitorLocker ml(Lock, Monitor::_no_safepoint_check_flag);
> 
> so I didn't.
> 
> open webrev at http://cr.openjdk.java.net/~coleenp/2019/8231472.01/webrev
> bug link https://bugs.openjdk.java.net/browse/JDK-8231472 >
> Worth doing?

Sure, ship it :)

Thanks, Robbin

> 
> thanks,
> Coleen
> 
>>
>> thanks!
>> Coleen
>>
>>>
>>>>
>>>> I'm going to work on the issues with 'special' locks as part of : https://bugs.openjdk.java.net/browse/JDK-8184732 
>>>> next, but the rewriting (below) removed these two uses of this ranking check.
>>>
>>> Ok
>>>
>>>>
>>>>>
>>>>> do_safepoint_check is checked four times (once via not_allowed), can you please re-write this method, or better yet 
>>>>> remove the bool input.
>>>>> The bool is hardcoded so can you just create two methods, with one helper method for the common checks. (yes there 
>>>>> will be more code, but readable :) )
>>>>
>>>> This is a good suggestion.  I've rewritten it to be a lot clearer, and am retesting it now with tier1 (all Oracle 
>>>> platforms) and tier2,3 linux-x64-debug.
>>>>
>>>> incremental webrev at http://cr.openjdk.java.net/~coleenp/2019/8213150.02.incr/webrev
>>>> full webrev at http://cr.openjdk.java.net/~coleenp/2019/8213150.02/webrev
>>>
>>> Thanks!
>>>
>>> /Robbin
>>>
>>>>
>>>> Thanks,
>>>> Coleen
>>>>
>>>>>
>>>>> Thanks, Robbin
>>>>>
>>>>> On 9/20/19 11:36 PM, coleen.phillimore at oracle.com wrote:
>>>>>> Summary: extend verification for all locking not just VMOperations, and fix CLDG lock to not be taken by VM thread.
>>>>>>
>>>>>> See bug comments for more details about this change.
>>>>>>
>>>>>> Tested with tier1 all Oracle platforms, tier2-8 linux-x64-debug.
>>>>>>
>>>>>> open webrev at http://cr.openjdk.java.net/~coleenp/2019/8213150.01/webrev
>>>>>> bug link https://bugs.openjdk.java.net/browse/JDK-8213150
>>>>>>
>>>>>> Thanks,
>>>>>> Coleen
>>>>
>>
>