On Tue, 10 Nov 2026 08:18:11 GMT, Tobias Hartmann <[email protected]> wrote:

> I see this failure in our testing on macosx-aarch64-debug and on Linux x64:
> 
> ```
> compiler/intrinsics/sha/sanity/TestSHA3ParallelIntrinsics.java
> -ea -esa -XX:CompileThreshold=100 -XX:+UnlockExperimentalVMOptions -server 
> -XX:-TieredCompilation
> 
> ----------System.out:(5/108)----------
> class = sun.security.provider.SHA3Parallel
> method = doubleKeccak
> TestInvoke Error: 
> expected = 0
> actual = 1
> ----------System.err:(17/1127)----------
> java.lang.Exception: Exception: java.lang.Exception: TestInvoke Error
>       at compiler.intrinsics.sha.TestInvoke.testInvoke(TestInvoke.java:98)
>       at 
> compiler.intrinsics.sha.sanity.InvokeSanityTestBase.test(InvokeSanityTestBase.java:81)
>       at 
> compiler.intrinsics.sha.sanity.TestSHA3ParallelIntrinsics.main(TestSHA3ParallelIntrinsics.java:86)
>       at 
> java.base/jdk.internal.reflect.DirectMethodHandleAccessor.invoke(DirectMethodHandleAccessor.java:104)
>       at java.base/java.lang.reflect.Method.invoke(Method.java:583)
>       at 
> com.sun.javatest.regtest.agent.MainMethodHelper.executeModernMainClass(MainMethodHelper.java:56)
>       at 
> com.sun.javatest.regtest.agent.MainWrapper$MainTask.run(MainWrapper.java:140)
>       at java.base/java.lang.Thread.run(Thread.java:1552)
> Caused by: java.lang.Exception: TestInvoke Error
>       at compiler.intrinsics.sha.TestInvoke.testInvoke(TestInvoke.java:95)
>       ... 7 more
> ```

There appears to be a clash between -XX:CompileThreshold and 
-XX:-TieredCompilation flags. I removed -XX:-TieredCompilation since I don't 
strictly need to disable it. Also, I added a guard in 
TestSHA3ParallelIntrinsics that skips the test if tiered compilation is 
disabled.

> src/hotspot/share/opto/c2compiler.cpp line 651:
> 
>> 649:   case vmIntrinsics::_quad_keccak:
>> 650:     if (StubRoutines::quad_keccak() == nullptr) return false;
>> 651:     break;
> 
> This is out of scope of this test enhancement and should be filed/fixed 
> separately.

Ok, I reverted the change here and created 
[JDK-8397115](https://bugs.openjdk.org/browse/JDK-8397115) to track it.

-------------

PR Comment: https://git.openjdk.org/jdk/pull/33709#issuecomment-6779742674
PR Review Comment: https://git.openjdk.org/jdk/pull/33709#discussion_r4530733819

Reply via email to