On Tue, 10 Oct 2023 14:21:40 GMT, Jaikiran Pai <j...@openjdk.org> wrote:

> Can I please get a review of this change which proposes to fix jmh test 
> launch failures noted in https://bugs.openjdk.org/browse/JDK-8317802?
> 
> jmh apparently relies on annotation processors during compilation of a 
> benchmark. When annotation processing is disabled, the generated benchmark 
> jar is unusable when the benchmark is launched and fails with:
> 
> 
> Exception in thread "main" java.lang.RuntimeException: ERROR: Unable to find 
> the resource: /META-INF/BenchmarkList
>       at 
> org.openjdk.jmh.runner.AbstractResourceReader.getReaders(AbstractResourceReader.java:98)
>       at org.openjdk.jmh.runner.BenchmarkList.find(BenchmarkList.java:124)
>       at org.openjdk.jmh.runner.Runner.internalRun(Runner.java:252)
>       at org.openjdk.jmh.runner.Runner.run(Runner.java:208)
>       at org.openjdk.jmh.Main.main(Main.java:71)
> 
> After the integration of https://bugs.openjdk.org/browse/JDK-8306819 recently 
> in JDK mainline, `javac` no longer runs annotation processors by default. 
> These jmh tests that are compiled as part of the JDK build will now have to 
> enable annotation processing explicitly.
> 
> The commit in this PR enables annotation processing by setting `-proc:full` 
> (implying run annotation processors as well as compile the files) when 
> compiling the benchmarks.
> 
> I tested this change locally. Without this change, the following command 
> fails:
> 
> 
> make test TEST="micro:java.lang.ArraysSort"                                   
>                             
> ...
> Test selection 'micro:java.lang.ArraysSort', will run:
> * micro:java.lang.ArraysSort
> 
> Running test 'micro:java.lang.ArraysSort'
> Exception in thread "main" java.lang.RuntimeException: ERROR: Unable to find 
> the resource: /META-INF/BenchmarkList
>       at 
> org.openjdk.jmh.runner.AbstractResourceReader.getReaders(AbstractResourceReader.java:98)
>       at org.openjdk.jmh.runner.BenchmarkList.find(BenchmarkList.java:124)
>       at org.openjdk.jmh.runner.Runner.internalRun(Runner.java:252)
>       at org.openjdk.jmh.runner.Runner.run(Runner.java:208)
>       at org.openjdk.jmh.Main.main(Main.java:71)
> Finished running test 'micro:java.lang.ArraysSort'
> 
> 
> 
> After the change in this PR, the same command works fine and the jmh 
> benchmark is run.

Magnus noted in the JBS issue that it would be good to enable only the specific 
JMH annotation processor instead of using `-proc:full`. I agree with Magnus.

I looked at the annotation processor that jmh jar ships with (in 
`META-INF/services`) and I see that it's 
`org.openjdk.jmh.generators.BenchmarkProcessor`. I've now updated this PR to 
use `-processor` option instead of `-proc:full`.

With this latest change, I cleaned up the local JDK build and rebuilt 
everything fresh and used the same command as previously to run the micro 
benchmark. It continues to pass with this new change.

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

PR Comment: https://git.openjdk.org/jdk/pull/16122#issuecomment-1755567425

Reply via email to