RFR (M) : 8028587 : [TESTBUG] New tests development for intrisics for basic operators - add, neg, inc, dec, sub, mul

[Christian Thalinger]

Oh, good.  Thanks for verifying.

On Dec 4, 2013, at 12:51 AM, Anton Ivanov <anton.ivanov at oracle.com> wrote:

> I've run jtreg with "-conc:auto"
> In parallel mode jtreg uses different directories for different tests (scratch/0, scratch/1, .., scratch/N), so it would not be a problem
> 
> On 04.12.2013 0:23, Christian Thalinger wrote:
>> Could it be a problem that we are using the same log file names for different tests?
>> 
>>   37  * @run main Verifier hs_neg.log hs.log
>> 
>> Maybe when we try to run them in parallel?  Or are different tests operating in different directories?
>> 
>> On Dec 3, 2013, at 8:41 AM, Anton Ivanov <anton.ivanov at oracle.com> wrote:
>> 
>>> Hi all,
>>> 
>>> Please review the patch
>>> 
>>> New tests for intrinsics were developed. It tests that intrinsics are used where it should, and vice versa.
>>> Base class CompilerWhiteBoxTest was refactored.
>>> DecrementExactLongTest.java and IncrementExactLongTest.java fails on amd64 platform. I suppose a product bug (no intrinsic are used where it should be), please check
>>> 
>>> webrev: http://cr.openjdk.java.net/~iignatyev/aaivanov/8028587/webrev.00/ <http://cr.openjdk.java.net/%7Eiignatyev/aaivanov/8028587/webrev.00/>
>>> jbs: https://bugs.openjdk.java.net/browse/JDK-8028587
>>> testing: test/compiler/tiered/, test/compiler/whitebox/, test/compiler/intrinsics/mathexact/sanity/
>>> 
>>> -- Best regards, Anton Ivanov
> 
> -- 
> Best regards,
> Anton Ivanov
>