RFR (XS) 8206931: Misleading "COMPILE SKIPPED: invalid non-klass dependency" compile log

Vladimir Ivanov vladimir.x.ivanov at oracle.com
Tue Jul 10 11:19:12 UTC 2018


src/hotspot/share/ci/ciEnv.cpp:

IMO the code is more readable when the check is left as is, but the 
messages are swapped instead:

      } else if (Dependencies::is_klass_type(result)) {
        record_failure("concurrent class loading");
      } else {
        record_failure("invalid non-klass dependency");
      }

Otherwise, looks good!

Best regards,
Vladimir Ivanov

On 10/07/2018 10:09, Aleksey Shipilev wrote:
> Bug:
>    https://bugs.openjdk.java.net/browse/JDK-8206931
> 
> It seems that JDK-8191052 made a misleading change:
>    http://hg.openjdk.java.net/jdk/hs/rev/e8f5fc8f5f67
> 
> We used to print "invalid non-klass dependency" when we have detected the dependency that failed
> is_klass_type() check. New code has Dependencies::validate_dependencies to reply with klass
> dependency when SystemDictionary was changed and non-klass dependency when such dependency was
> detected. But caller in ciEnv only prints "invalid non-klass dependency" when dependency _is_ the
> klass dependency. jvmciEnv uses it inappropriately too, even though without observable effects.
> 
> Fix:
>    http://cr.openjdk.java.net/~shade/8206931/webrev.01/
> 
> Testing: tier1_compiler, eyeballing compiler logs
> 
> Thanks,
> -Aleksey
> 


More information about the hotspot-compiler-dev mailing list