[aarch64-port-dev ] Failed hotspot tests on aarch64

[Alexeev, Alexander]

Andrew

This run was against http://hg.openjdk.java.net/jdk9/dev
Actually, after that I've updated repository again and rerun the suite

alexander at crb3:~/build-open-jdk/dev/jdk9/dev/hotspot$ hg id --num
8306

BTW DekkerTest isn't in the list anymore.

Now, with pended patch results are the following:
Failed:
compiler/codegen/7184394/TestAESMain.java
compiler/intrinsics/classcast/NullCheckDroppingsTest.java
compiler/intrinsics/mathexact/sanity/SubtractExactIntTest.java
compiler/intrinsics/mathexact/sanity/SubtractExactLongTest.java
compiler/intrinsics/multiplytolen/TestMultiplyToLen.java
compiler/intrinsics/sha/cli/TestUseSHA256IntrinsicsOptionOnUnsupportedCPU.java
compiler/intrinsics/sha/cli/TestUseSHAOptionOnUnsupportedCPU.java
runtime/7107135/Test7107135.sh
serviceability/sa/jmap-hashcode/Test8028623.java

Error:
native_sanity/JniVersion.java

And for repository "as is" without the patch

Failed:
compiler/codegen/7184394/TestAESMain.java
compiler/intrinsics/classcast/NullCheckDroppingsTest.java
compiler/intrinsics/mathexact/sanity/AddExactIntTest.java
compiler/intrinsics/mathexact/sanity/AddExactLongTest.java
compiler/intrinsics/mathexact/sanity/DecrementExactIntTest.java
compiler/intrinsics/mathexact/sanity/DecrementExactLongTest.java
compiler/intrinsics/mathexact/sanity/IncrementExactIntTest.java
compiler/intrinsics/mathexact/sanity/IncrementExactLongTest.java
compiler/intrinsics/mathexact/sanity/MultiplyExactIntTest.java
compiler/intrinsics/mathexact/sanity/MultiplyExactLongTest.java
compiler/intrinsics/mathexact/sanity/NegateExactIntTest.java
compiler/intrinsics/mathexact/sanity/NegateExactLongTest.java
compiler/intrinsics/mathexact/sanity/SubtractExactIntTest.java
compiler/intrinsics/mathexact/sanity/SubtractExactLongTest.java
compiler/intrinsics/multiplytolen/TestMultiplyToLen.java
compiler/intrinsics/sha/cli/TestUseSHA1IntrinsicsOptionOnUnsupportedCPU.java
compiler/intrinsics/sha/cli/TestUseSHA256IntrinsicsOptionOnUnsupportedCPU.java
compiler/intrinsics/sha/cli/TestUseSHAOptionOnUnsupportedCPU.java
compiler/intrinsics/sha/sanity/TestSHA1Intrinsics.java
compiler/intrinsics/sha/sanity/TestSHA1MultiBlockIntrinsics.java
compiler/intrinsics/sha/sanity/TestSHA256Intrinsics.java
compiler/intrinsics/sha/sanity/TestSHA256MultiBlockIntrinsics.java
compiler/stable/TestStableBoolean.java
compiler/stable/TestStableByte.java
compiler/stable/TestStableChar.java
compiler/stable/TestStableDouble.java
compiler/stable/TestStableFloat.java
compiler/stable/TestStableInt.java
compiler/stable/TestStableLong.java
compiler/stable/TestStableObject.java
compiler/stable/TestStableShort.java
runtime/7107135/Test7107135.sh
serviceability/sa/jmap-hashcode/Test8028623.java
testlibrary_tests/TestMutuallyExclusivePlatformPredicates.java

Error:
native_sanity/JniVersion.java


________________________________________
From: Andrew Haley [aph at redhat.com]
Sent: 16 May 2015 11:49
To: Alexeev, Alexander; aarch64-port-dev at openjdk.java.net
Subject: Re: [aarch64-port-dev ] Failed hotspot tests on aarch64

Hi,

On 15/05/15 14:41, Alexeev, Alexander wrote:

> Currently there are six hotspot tests that are failed on aarch64 and don't have fixes on the way, at least that I am aware of.
> Those all test are successfully passed on x86.
>
> Tests:
>
>
> compiler/codegen/7184394/TestAESMain.java

Ed -- this one is yours, I think.

> compiler/intrinsics/classcast/NullCheckDroppingsTest.java
>
> compiler/intrinsics/multiplytolen/TestMultiplyToLen.java
>
> compiler/membars/DekkerTest.java

This one is surprising and interesting.

> runtime/7107135/Test7107135.sh
>
> serviceability/sa/jmap-hashcode/Test8028623.java
>
> Does somebody has any extra information about those failures which worth to know before dealing with them?

Thanks.

I'll have a look at these. Which version of OpenJDK did you test?

Andrew.