-
Notifications
You must be signed in to change notification settings - Fork 722
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
OpenJDK jdk_vector_double128, jdk_vector_int128, jdk_vector_long128 passed but failed #17185
Comments
The tests check if vectorization happened by using grep. Vectorization was never enabled on x86. So I am actually wondering how this could work before… @BradleyWood |
These tests passed in openjdk run, but we add the extra step |
I thought that step was there for a while. Was it added recently for some tests? @llxia I am just wondering how "VECTOR API" could be produced on x86 before. |
@r30shah fyi |
No, there were no recent changes to these tests. Also, ran
|
https://openj9-jenkins.osuosl.org/job/Test_openjdk20_j9_sanity.openjdk_x86-64_windows_Nightly/43
|
https://openj9-jenkins.osuosl.org/job/Test_openjdk20_j9_sanity.openjdk_x86-64_windows_Nightly/48
|
https://openj9-jenkins.osuosl.org/job/Test_openjdk20_j9_sanity.openjdk_x86-64_windows_Nightly/50
|
https://openj9-jenkins.osuosl.org/job/Test_openjdk20_j9_sanity.openjdk_x86-64_windows_Nightly/52 jdk_vector_float128_j9_0
jdk_vector_short128_j9_0
|
|
As JDK20 is now part of the OMR Acceptance build and this failure has appeared in the OMR Acceptance build the priority of resolving this should be higher so that it doesn't block promotion. @gita-omr : it is not clear who owns the next step on this one, but can you assign please. |
I can see that Scalarization happens randomly on these tests. Investigating. |
Seems that it only failed on Windows. Please update if you happen to see failures on other platforms. |
FYI they still occur fairly frequently (on Windows), I just stopped reporting them. |
Interesting... so it does look like only Windows problem. |
@gita-omr : any updates on this issue? It is presently targeted to 0.40 so we will need to make a determination on it soon. |
I think we should move it to the next release. This is incubator and it’s not a functional failure: test case passes but it’s not scalarized. The failure is very random and seem to happen exclusively on Windows. I actually suspect some environment issues since the same tests passes on Windows internally. |
@pshipton : are you ok with deferring this? I know it is impacting the OMR Acceptance build which means you (personally) have to manually ignore the failures when they occur to promote the build. |
Yes it's fine, just one of many. |
In OMR acceptance. Likely occurs on all platforms, but they aren't yet all completed. https://openj9-jenkins.osuosl.org/job/Test_openjdk20_j9_sanity.openjdk_aarch64_linux_OMR/31 |
The previous comment can be ignored as those failures aren't related to this issue. This issue does occur regularly in the builds, but I'm not going to report them all. The latest being |
|
|
Still happening. |
I think it's a low priority since the test case produces correct result. |
Vector is still an incubator so I've moved it forward. |
|
This hasn't been seen for some time, closing assuming fixed. |
https://openj9-jenkins.osuosl.org/job/Test_openjdk20_j9_sanity.openjdk_x86-64_windows_Nightly/34/
win2012-x86-1
jdk_vector_double128_j9_0
jdk_vector_int128_j9_0
jdk_vector_long128_j9_0
Changes from previous build
f9950fe...4a2643c
adoptium/aqa-tests@44ea825...4fd9cfa
The text was updated successfully, but these errors were encountered: