RFR (S) 8206960: [Graal] serviceability/jvmti/HeapMonitor/MyPackage/HeapMonitor tests fail

JC Beyler jcbeyler at google.com
Wed Jul 11 17:04:17 UTC 2018


Hi all,

Could someone review the small-ish webrev for the bug:
https://bugs.openjdk.java.net/browse/JDK-8206960

The webrev is here:
http://cr.openjdk.java.net/~jcbeyler/8206960/webrev.00/

Basically, the tests were failing for two reasons:
  - VMEventTest was failing because Graal does not support DisableIntrinsic
required by the test, I disabled testing the test with Graal in this case
  - The other tests were failing because the BCI <-> source code line
numbers are not always correct when using Graal via uncommon traps;
therefore the tests now check if Graal is being used and, if so, only
checks the method names. This allows us to still have tests working with
Graal, albeit a bit more coarse.

This passes all the HeapMonitor tests
with -vmoptions:"-XX:+UnlockExperimentalVMOptions -XX:+EnableJVMCI
-XX:+TieredCompilation -XX:+UseJVMCICompiler -Djvmci.Compiler=graal"

(Except the GCCMS one which is being fixed via the one-liner for
JDK-8205643).

Let me know what you think,
Jc
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.openjdk.java.net/pipermail/serviceability-dev/attachments/20180711/407bcfd5/attachment.html>


More information about the serviceability-dev mailing list