RFR (S): [TESTBUG] gc/g1/TestGCLogMessages.java fails on 32bit arm after changes for JDK-8155094

Thomas Schatzl thomas.schatzl at oracle.com
Wed Mar 29 07:23:36 UTC 2017


Hi all,

  can I have a second "R"eviewer review for this small test change?

Thanks,
  Thomas

On Mon, 2017-03-20 at 21:17 +0100, Thomas Schatzl wrote:
> Hi all,
> 
>   can I have reviews for this change that fixes the
> gc/g1/TestGCLogMessages test for (atm unsupported but available) 32
> bit
> arm?
> 
> The reason is that the log message introduced with JDK-8155094
> depends
> on either C2 or JVMCI (Graal) available, which is not the case on 32
> bit arm.
> 
> The fix adds a check for that log message to not require the log
> message in this case to be present (actually it checks that it is not
> there in this case).
> 
> The reason for this bug is that 32 bit ARM is not supported (but
> available) so the test is not run as often, particularly not during
> jprt (also for performance reasons).
> 
> CR:
> https://bugs.openjdk.java.net/browse/JDK-8177059
> Webrev:
> http://cr.openjdk.java.net/~tschatzl/8177059/webrev/
> Testing:
> rbt, jprt
> 
> Thanks,
>   Thomas
> 


More information about the hotspot-gc-dev mailing list