RFR(S) 8211743: [AOT] crash in ScopeDesc::decode_body() when JVMTI walks AOT frames

[dean.long at oracle.com]

Thanks Vladimir.

dl

On 10/30/18 8:48 AM, Vladimir Kozlov wrote:
> Good.
>
> Thanks,
> Vladimir
>
> On 10/29/18 11:02 PM, dean.long at oracle.com wrote:
>> Yes, vm/jvmti/Breakpoint/brkp001/brkp00102/brkp00102.html still passes.
>>
>> dl
>>
>> On 10/29/18 5:51 PM, Vladimir Kozlov wrote:
>>> Dean,
>>>
>>> Did you run tests from JDK-8209574 with your fix?
>>>
>>> Changes are good.
>>>
>>> Thanks,
>>> Vladimir
>>>
>>> On 10/29/18 5:14 PM, dean.long at oracle.com wrote:
>>>> https://bugs.openjdk.java.net/browse/JDK-8211743
>>>> http://cr.openjdk.java.net/~dlong/8211743/webrev/
>>>>
>>>> The fix for JDK-8209574 was a little too aggressive.  We can't turn 
>>>> off UseAOT after JVM startup if there is AOT code already loaded, 
>>>> unless we deoptimize that code first.  This fix is a redo of that 
>>>> fix, but only applied for breakpoints (and not hotswap) and only 
>>>> during startup.
>>>>
>>>> dl
>>