Enabling clean PRs in Leyden

[ioi.lam at oracle.com]

On 8/19/24 12:29 PM, ioi.lam at oracle.com wrote:
>>> 3. Code changes to make most (all?) builds clean:
>>>   (this one is progress, I need to confirm all GHA jobs are green)
>>>   https://github.com/openjdk/leyden/pull/10
>>
>> This progresses nicely, and needs another thing: a merge from current 
>> openjdk mainline, so that we get x86_32 test jobs disabled 
>> (https://bugs.openjdk.org/browse/JDK-8338286).
>>
>> Besides that, there are two DeterministicDump test failures on 
>> macos-x64 and windows-x64. Interestingly, there is no test failure on 
>> macos-aarch64, linux-aarch64, or linux-x64. I have not looked into 
>> that in any detail, maybe someone knows what is up with those:
>>
> I am working on Windows failures now (e.g., JDK-8338604) and I've seen 
> this also. If I have time I'll take a look at the deterministic dump, 
> but it would be low priority.
>

The problem with deterministic build is fixed in this commit

https://github.com/openjdk/leyden/commit/a5daf90b0e818194fd11e74569e72f6dc70c105c


> Thanks
>
> - Ioi
>
>
>> Fails like this:
>>
>> ----------System.err:(12/677)----------
>> java.lang.RuntimeException: File content different at byte #4, b0 = 
>> -92, b1 = 113
>>     at DeterministicDump.compare(DeterministicDump.java:114)
>>     at DeterministicDump.doTest(DeterministicDump.java:66)
>>     at DeterministicDump.main(DeterministicDump.java:46)
>>     at 
>> java.base/jdk.internal.reflect.DirectMethodHandleAccessor.invoke(DirectMethodHandleAccessor.java:103)
>>     at java.base/java.lang.reflect.Method.invoke(Method.java:588)
>>     at 
>> com.sun.javatest.regtest.agent.MainActionHelper$AgentVMRunnable.run(MainActionHelper.java:333)
>>     at java.base/java.lang.Thread.run(Thread.java:1575)
>>
>> Example:
>>   https://github.com/shipilev/leyden/actions/runs/10415862715
>>
>> -Aleksey