-
Notifications
You must be signed in to change notification settings - Fork 33
Promote AdoptOpenJDK Version jdk8u282-b08 #197
Comments
@AdoptOpenJDK/tsc - Can I get some approvals to promote these? |
+1 |
+1 to release 👍 Nice assessment @karianna ACTION: exclude the tests that have known/reported upstream issues. |
JDK 8 HotSpot shipped https://github.com/AdoptOpenJDK/openjdk8-binaries/releases/tag/jdk8u282-b08 |
Keeping open until we have arm32 and aarch64 out |
Aarch64 LinuxTRSS link Test_openjdk8_hs_extended.openjdk_aarch64_linux |
aarch64 linux shipped |
So just 32 to go? |
Arm32 LinuxTRSS link Test_openjdk8_hs_sanity.openjdk_arm_linux |
The LambdaForms bug should be reported upstream but isn't a blocker (it's an optimization). The util issues need investigation but again shouldn't block the release IMO |
I have raised adoptium/aqa-tests#2263 for the 3 native test cases (java.util.stream.xxxxOpTest) failing in the jdk_util_2 target, and running a Grinder to see if last release also shows these or if it is a regression. |
Update the 3 failing java.util.stream tests do not fail when run against the last arm_linux release (Grinder/6710). So it is a regression from last release... have not yet looked in JBS to see if there are matching behaviours described java/util/stream/DistinctOpTest.java - ArrayOutOfBoundsException (see below)
|
I am hoping that someone other than I will look at the failures more closely. They appear to be regressions (related: adoptium/aqa-tests#2061 not seen in previous releases). Next actions could be to find a different distro of this platform and run the tests against it, report upstream if found to be an issue there. We can decide whether we block on these failures or not, once someone does some additional analysis. |
Grinders for ARM32 with Azul Zulu:
Same problem is present:
I looked through the test and didn't see any unusal dependency on native code that might cause troubles on our test machines (but not elsewhere). So my best guess: This is a problem in OpenJDK and not limited to our build. What makes me a bit nervous is that Azul Zulu 8u275 has it, too. So I'm running another Grinder with AdoptOpenJDK 8u275 at https://ci.adoptopenjdk.net/job/Grinder/7072/. |
The outcome of that Grinder with AdoptOpenJDK 8u275 was that the problem was apparently present in previous releases, too:
So that's the best I can offer. |
Thanks @aahlenst for your investigation... I must not have picked up the last release in the Grinder I ran to check if it was present or not. Given we see these failures in both last release and Azul's build, we will not consider them blocking (FYI @gdams). I guess we best try to confirm if we have a machine/test config issue here or an actual product issue and get a defect raised to the appropriate repo. |
@gdams do we need to do anything now to promote this last binary? |
Released linux arm32 |
JVM:
@AdoptOpenJDK/tsc please can somebody +1 this request?
The text was updated successfully, but these errors were encountered: