RFR (S) 8030955: assert(_prologue != NULL) failed: prologue pointer must be initialized

[harold seigel]

Hi Zhengyu,

Thanks for the review.

I could limit to only one call to PerfMemory::destroy(), but I think 
that would be a buggier fix, especially if someday someone adds another 
call to PerfMemory::destroy().

Note that PerfMemory::initialize() can be called multiple times.

Thanks, Harold

On 12/31/2013 11:48 AM, Zhengyu Gu wrote:
> Can you limit the call to PerfMemory::destroy() once? the assertion 
> should be valid for the first call.
>
> Thanks,
>
> -Zhengyu
>
>
> On 12/30/2013 9:27 AM, harold seigel wrote:
>> Hi,
>>
>> Please review this small fix for bug 8030955.  The assert occurred 
>> because PerfMemory::destroy() was called multiple times.  The fix 
>> replaces the assert with a check that immediately returns if 
>> PerfMemory::destroy() has already been called.
>>
>> The fix was tested with the accompanying test, JCK lang and vm tests, 
>> hotspot jtreg tests, and ute quick tests.
>>
>> Open webrev at: http://cr.openjdk.java.net/~hseigel/bug_8030955/ 
>> <http://cr.openjdk.java.net/%7Ehseigel/bug_8030955/>
>>
>> Bug: https://bugs.openjdk.java.net/browse/JDK-8030955
>>
>> Thanks! Harold
>