On 05/17/2016 07:50 PM, shilpi.rast...@oracle.com wrote: > On 5/17/2016 7:44 PM, Alan Bateman wrote: >> >> On 17/05/2016 14:34, Martin Buchholz wrote: >>> In jdk8 the /othervm is missing. >>> http://mail.openjdk.java.net/pipermail/jtreg-use/2016-May/000478.html >>> >> Ah, I assumed we were discussing the version in jdk9/dev. >> >> BTW: My interest here is to make sure we didn't break anything with >> the module system refresh two weeks ago. We are currently joined at >> the hip to jtreg and so had to force an update to jtreg as part of the >> update. We updated many tests as part of this and I don't recall any >> tests failing when we pushed. > > Could someone please explain why it is not required for this test as i > saw we are updating other tests ( > http://hg.openjdk.java.net/jdk9/hs/hotspot/rev/0e2a2be60453 ) ?
Look, the change above adds "/othervm" (I think those tests already have "/bootclasspath" because they access WhiteBox API): http://hg.openjdk.java.net/jdk9/hs/hotspot/rev/0e2a2be60453 Your change, OTOH, adds "/bootclasspath": http://cr.openjdk.java.net/~srastogi/8155791/webrev.00/test/java/lang/invoke/CustomizedLambdaFormTest.java.sdiff.html Is this the intent? Does the test fail without "/bootclasspath"? Thanks, -Aleksey
