RFR 8183938: [MVT] Add test for mock vector API
Maurizio Cimadamore
maurizio.cimadamore at oracle.com
Thu Jul 6 14:27:24 UTC 2017
Nope - there's something else going on here - probably some bad
optimization? I can only reproduce the problem with -Xcomp and if I'm
_not_ debugging.
Maurizio
On 06/07/17 15:15, Maurizio Cimadamore wrote:
> Something is probably off with clasloaders - and that makes the test
> System.isValueClass (and Class.isValueClass) to fail. Investigating.
>
> Maurizio
>
>
> On 06/07/17 15:08, Maurizio Cimadamore wrote:
>> -Xcomp doesn't work - I get this:
>>
>> org.testng.TestNGException:
>> An error occurred while instantiating class VectorTest: Long2$Value
>> not a value type
>> at
>> org.testng.internal.ClassHelper.createInstance1(ClassHelper.java:393)
>> at
>> org.testng.internal.ClassHelper.createInstance(ClassHelper.java:295)
>> at
>> org.testng.internal.ClassImpl.getDefaultInstance(ClassImpl.java:117)
>> at org.testng.internal.ClassImpl.getInstances(ClassImpl.java:183)
>> at
>> org.testng.internal.TestNGClassFinder.<init>(TestNGClassFinder.java:128)
>> at org.testng.TestRunner.initMethods(TestRunner.java:415)
>> at org.testng.TestRunner.init(TestRunner.java:241)
>> at org.testng.TestRunner.init(TestRunner.java:211)
>> at org.testng.TestRunner.<init>(TestRunner.java:158)
>> at
>> org.testng.SuiteRunner$DefaultTestRunnerFactory.newTestRunner(SuiteRunner.java:545)
>> at org.testng.SuiteRunner.init(SuiteRunner.java:159)
>> at org.testng.SuiteRunner.<init>(SuiteRunner.java:114)
>> at org.testng.TestNG.createSuiteRunner(TestNG.java:1260)
>> at org.testng.TestNG.createSuiteRunners(TestNG.java:1247)
>> at org.testng.TestNG.runSuitesLocally(TestNG.java:1101)
>> at org.testng.TestNG.run(TestNG.java:1018)
>> at
>> com.sun.javatest.regtest.agent.TestNGRunner.main(TestNGRunner.java:94)
>> at
>> com.sun.javatest.regtest.agent.TestNGRunner.main(TestNGRunner.java:54)
>> at
>> java.base/jdk.internal.reflect.NativeMethodAccessorImpl.invoke0(Native
>> Method)
>> at
>> java.base/jdk.internal.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:62)
>> at
>> java.base/jdk.internal.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43)
>> at java.base/java.lang.reflect.Method.invoke(Method.java:563)
>> at
>> com.sun.javatest.regtest.agent.MainWrapper$MainThread.run(MainWrapper.java:115)
>> at java.base/java.lang.Thread.run(Thread.java:844)
>> Caused by: java.lang.InternalError: Long2$Value not a value type
>> at java.base/java.lang.System$2.loadValueTypeClass(System.java:2174)
>> at
>> java.base/valhalla.shady.MinimalValueTypes_1_0.loadValueTypeClass(MinimalValueTypes_1_0.java:114)
>> at
>> java.base/valhalla.shady.MinimalValueTypes_1_0.getValueTypeClass(MinimalValueTypes_1_0.java:107)
>> at
>> java.base/jdk.experimental.value.ValueType.forClass(ValueType.java:158)
>> at VectorTest.<clinit>(VectorTest.java:18)
>> at
>> java.base/jdk.internal.reflect.NativeConstructorAccessorImpl.newInstance0(Native
>> Method)
>> at
>> java.base/jdk.internal.reflect.NativeConstructorAccessorImpl.newInstance(NativeConstructorAccessorImpl.java:62)
>> at
>> java.base/jdk.internal.reflect.DelegatingConstructorAccessorImpl.newInstance(DelegatingConstructorAccessorImpl.java:45)
>> at
>> java.base/java.lang.reflect.Constructor.newInstance(Constructor.java:487)
>> at
>> org.testng.internal.ObjectFactoryImpl.newInstance(ObjectFactoryImpl.java:29)
>> at
>> org.testng.internal.ClassHelper.createInstance1(ClassHelper.java:382)
>> ... 23 more
>>
>>
>> Without -Xcomp, or with -Xint everything works.
>>
>> Any clues?
>>
>> Maurizio
>>
>>
>> On 06/07/17 15:03, Maurizio Cimadamore wrote:
>>>
>>>
>>> On 06/07/17 14:51, Tobias Hartmann wrote:
>>>> On 06.07.2017 15:48, Maurizio Cimadamore wrote:
>>>>> Not really, that's something that came from before your fix of
>>>>> JDK-8183129 now it can be safely removed if preferred.
>>>>>
>>>>> I will also remove -Xint flag from the test that I've already pushed.
>>>> Okay, thanks! I think it would be best to have one run with -Xint
>>>> and one with -Xcomp.
>>>>
>>>> Is there a reason you are using testng instead of a normal jtreg test?
>>> testng makes it easier to add new test cases as you go along - with
>>> plain jtreg you really only have a main method. It is also easier to
>>> create things like data providers - e.g. create a bunch of objects
>>> and execute a given test method against each possible object in the
>>> data set. In other words, testng is typically more suited for
>>> writing unit tests, and the test report will be generally more
>>> useful e.g :
>>>
>>> test VectorTest.testConditional(): success
>>> test VectorTest.testSumArray(1): success
>>> test VectorTest.testSumArray(5): success
>>> test VectorTest.testSumArray(10): success
>>> test VectorTest.testSumArray(0): success
>>>
>>> ===============================================
>>> valhalla/mvt/vector/VectorTest.java
>>> Total tests run: 5, Failures: 0, Skips: 0
>>> ===============================================
>>>
>>> You can of course do all this with jtreg, but you have to do it
>>> manually, with the risk of replicating routines across multiple
>>> tests (e.g. assertEquals, etc.)
>>>
>>>
>>> Maurizio
>>>
>>>>
>>>> Best regards,
>>>> Tobias
>>>>
>>>>> On 06/07/17 14:40, Tobias Hartmann wrote:
>>>>>> Hi Maurizio,
>>>>>>
>>>>>> is there a reason that the test is only executed with -Xint?
>>>>>>
>>>>>> Thanks,
>>>>>> Tobias
>>>>>>
>>>>>> On 06.07.2017 15:28, Maurizio Cimadamore wrote:
>>>>>>> Hi,
>>>>>>> this changeset adds the mock vector API test that Vlad has
>>>>>>> worked on. I cleaned it up a bit and converted to a working
>>>>>>> testng test.
>>>>>>>
>>>>>>> http://cr.openjdk.java.net/~mcimadamore/8183938/
>>>>>>>
>>>>>>> Cheers
>>>>>>> Maurizio
>>>>>>>
>>>
>>
>
More information about the valhalla-dev
mailing list