debug: errorlevel conditionals #2285
Conversation
jharlow-intel
requested review from
antonwolfy,
AlexanderKalistratov,
vlad-perevezentsev and
vtavana
as code owners
|
View rendered docs @ https://intelpython.github.io/dpnp/pull/2285/index.html |
|
Array API standard conformance tests for dpnp=0.17.0dev5=py312he4f9c94_13 ran successfully. |
antonwolfy
dismissed
their stale review
It seems doesn't work properly.
All tests passed there:
https://cje-fm-owrp-prod04.devtools.intel.com/satg-dap-intelpython/job/intel-packages/job/dpnp/job/dev-windows-py3.11/job/test/1493/cloudbees-pipeline-explorer/
but job returned error status.
|
Hmm, I was a little afraid of that. I guess the tests can all "pass" - but due to how some tests are muted I guess it doesn't always return an exit code of 0 even when it passes. I'll need to experiment more, and I'll share my findings with you in Teams. We at least now know it is indeed the edit: |
|
Looks like it's passing and failing when appropriate now 😄 |
|
It looks like my conditional is basically identical to the ones used previously, but their is a nuance to TLDR; these variables are technically not the same and Links I read: I also peeked at conda-forge recipes and they also seem to lean into the |
Wheels tests were failing because some python abrupt failures (or windows access violations) would use an exit code that was both non-zero and non-1. This should capture everything that isn't an exit code of 0. 91161a8
Wheels tests were failing because some python abrupt failures (or windows access violations) would use an exit code that was both non-zero and non-1.
This should capture everything that isn't an exit code of 0