RFR(S): [TESTBUG]runtime/ErrorHandling fails when using jtreg 4.1 b13 (original) (raw)

David Holmes david.holmes at oracle.com
Tue Oct 30 01:26:22 UTC 2018

On 29/10/2018 6:34 PM, Vaibhav Choudhary wrote:

Thank you Igor and David for the review. This make sense to remove @build. Here goes the next webrev :-

http://cr.openjdk.java.net/~rpatil/8151295/webrev.01/

The removal seem okay - as long as the required test library classes end up getting built. Neither myself nor Igor are clear on the correct approach in 8u.

This test is failing from quite a long time but its getting ignored as known failure.

You mean it's been failing for a long time due to a switch to jtreg 4.1 b13?

Thanks, David

Thanks, Vaibhav Choudhary vaibhav.x.choudhary at oracle.com https://blogs.oracle.com/vaibhav

On 25-Oct-2018, at 11:54 PM, Igor Ignatyev <igor.ignatyev at oracle.com> wrote: Hi David, On Oct 24, 2018, at 10:14 PM, David Holmes <david.holmes at oracle.com> wrote:

Hi Igor, On 25/10/2018 3:11 PM, Igor Ignatyev wrote: IIRC, prior jtreg 4.1b13 it was fine to have nonexistent classes in @build actions, so the fix is good regardless of used version of jtreg, but are necessary for jtreg 4.1b13+. Okay, just wanted to be sure that 8u testing still expects use of hotspot testlibrary implementation in hotspot tests. Though if the @build is actually a no-op is it even necessary to fix it rather than just delete it? Sorry I've lost track of where we stand with needing to explicitly build the test library in older versions. Unfortunately, I can't recall what the guidelines were for jdk8, but I agree it'd be safer to just remove @build actions here if these tests worked fine w/ jtreg4.1b12 (where these @builds were simply ignored) Cheers, -- Igor Thanks, David

-- Igor On Oct 24, 2018, at 10:06 PM, David Holmes <david.holmes at oracle.com> wrote:

Hi Vaibhav, cc'ing jdk8u-dev to get some input from the maintainers. I was very confused until I realized this RFR was for 8u! :) Even so, after reading the bug report I'm still quite confused. This has been "broken" since the test was added in January 2016, but only causes a failure if jtreg 4.1b13 is used - is that correct? What is the default version of jtreg being used for 8u today? I'd be happy to review this but given I have no idea how the 8u testing is being handled in general I can't say whether this is the right fix ?? Thanks, David On 25/10/2018 2:18 PM, Vaibhav Choudhary wrote: Hi, Please review this tiny fix. Bug : https://bugs.openjdk.java.net/browse/JDK-8151295 Fix: http://cr.openjdk.java.net/~rpatil/8151295/webrev.00/ Desc: In stead of @build jdk.test.lib.* we need to use @build com.oracle.java.testlibrary.* These two tests are failing in nightly. Thanks, Vaibhav

More information about the hotspot-runtime-dev mailing list