RFR(S) [15] : 8208243 : vmTestbase/gc/lock/jni/jnilock002/TestDescription.java fails in jdk/hs nightly

Thomas Schatzl thomas.schatzl at oracle.com
Tue Jul 7 09:13:10 UTC 2020


Hi,

On 07.07.20 04:57, Igor Ignatyev wrote:
> ping?
> -- Igor
> 
>> On Jul 2, 2020, at 11:02 AM, Igor Ignatyev <igor.ignatyev at oracle.com> wrote:
>>
>> http://cr.openjdk.java.net/~iignatyev//8208243/webrev.00
>>> 325 lines changed: 18 ins; 0 del; 307 mod;
>>
>> Hi all,
>>
>> could you please review the patch for nsk.share.gc.lock.jni.*CriticalLocker.cpp files which:
>> - uses difftime to compare current_time and start_time;
>> - calls FatalError if Get*Critical returns NULL so the test would fail w/ a more obvious error than 'Native hash: 0 != Java hash:...' and won't hide this problem in cases when Java hash happens to be 0 (which is currently always true for all but float/double arrays);
>> - resets initial hash value in each iteration of while-loop;
>> - aligns whitespace w/ usual hotspot code style.
>>
>> testing: vmTestbase/gc/lock/jni on {linux,windows,macosx}-x64-{product,fastedebug} (w/ jnilock002 being removed from the problem-list)
>> JBS: https://bugs.openjdk.java.net/browse/JDK-8208243
>> webrev: http://cr.openjdk.java.net/~iignatyev//8208243/webrev.00
>>
>> Thanks,
>> -- Igor
> 

- in the CR there is a concern about the objFieldId static which seems 
not an issue after all.

Looks good.

Thomas


More information about the hotspot-gc-dev mailing list