jmhCompile should not compile with compiler arg -Werror

[DotSpy]

Jmh fail on compile time because of -Werror flag

[jkwatson]

we should add jmhCompile to the CI steps, so this doesn't happen again. Can you add this to the circle configuration, or to the Makefile, @DotSpy ?

[DotSpy]

i would like to get into CI, i'll do it

[DotSpy]

@jkwatson done

[thisthat]

I would create a different Makefile target for the benchmarks. This way, we can decouple benchmarks from compilation in CircleCI

[DotSpy]

agree

[DotSpy]

Any ideas why on CI
https://circleci.com/gh/open-telemetry/opentelemetry-java/2008?utm_campaign=vcs-integration-link&utm_medium=referral&utm_source=github-build-link

Task :opentelemetry-api:compileJmhJava SKIPPED ?

[codecov-io]

Codecov Report

Merging #866 into master will increase coverage by 1.8%.
The diff coverage is 84.15%.

@@             Coverage Diff             @@
##             master     #866     +/-   ##
===========================================
+ Coverage     78.94%   80.75%   +1.8%     
- Complexity      789      820     +31     
===========================================
  Files           105      108      +3     
  Lines          2845     2899     +54     
  Branches        273      265      -8     
===========================================
+ Hits           2246     2341     +95     
+ Misses          494      452     -42     
- Partials        105      106      +1

Impacted Files	Coverage Δ	Complexity Δ
...ntelemetry/sdk/trace/RecordEventsReadableSpan.java	91.51% <ø> (ø)	45 <0> (ø)	⬇️
...ntelemetry/exporters/inmemory/InMemoryTracing.java	85.71% <ø> (ø)	5 <0> (ø)	⬇️
...lemetry/sdk/trace/export/SimpleSpansProcessor.java	87.5% <ø> (ø)	7 <0> (ø)	⬇️
...metry/exporters/jaeger/JaegerGrpcSpanExporter.java	67.79% <ø> (ø)	4 <0> (ø)	⬇️
...ava/io/opentelemetry/sdk/trace/SpanBuilderSdk.java	93.87% <ø> (ø)	36 <0> (ø)	⬇️
...o/opentelemetry/sdk/trace/export/SpanExporter.java	100% <ø> (ø)	0 <0> (ø)	⬇️
...elemetry/sdk/trace/export/BatchSpansProcessor.java	89.65% <ø> (+0.86%)	7 <0> (ø)	⬇️
...ava/io/opentelemetry/exporters/jaeger/Adapter.java	96.2% <ø> (ø)	16 <0> (ø)	⬇️
...ntelemetry/sdk/trace/export/MultiSpanExporter.java	100% <ø> (ø)	12 <0> (ø)	⬇️
...java/io/opentelemetry/sdk/trace/data/SpanData.java	100% <ø> (ø)	2 <0> (?)
... and 33 more

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update dc5cfb7...b826c8d. Read the comment docs.

[DotSpy]

Any ideas why on CI
circleci.com/gh/open-telemetry/opentelemetry-java/2008?utm_campaign=vcs-integration-link&utm_medium=referral&utm_source=github-build-link

Task :opentelemetry-api:compileJmhJava SKIPPED ?

In the link you provide, you can find the compileJmhJava task be executed after the build task.

yeap it executed in Build section, but i think that it should be executed in Compile JMH section and not executed in Build

[thisthat]

Sorry, I wrongly read your comment. It is skipped because there is no /src/jmh/java directory in the API module.

[DotSpy]

still if u look at build section it calling > Task :opentelemetry-sdk:compileJmhJava

[DotSpy]

Found: gradle assemble calling :opentelemetry-sdk:compileJmhJava

[DotSpy]

@jkwatson as u can see here https://circleci.com/gh/open-telemetry/opentelemetry-java/2008#tests/containers/0
gradle assemble calling > Task :opentelemetry-sdk:compileJmhJava
so now we have jmhCompile in our CI, and after i will update #815 from master it will fail.
So question is : Should we move out jmhCompile from gradle assemble and build step in CI in this PR or should we not touch it for now?

[jkwatson]

This seems fine to me.

[DotSpy]

in current state there is no sense for Compile JMH step, in it we will get

Task :opentelemetry-sdk:compileJmhJava UP-TO-DATE

because it's called at Build step