Conversation
encukou
left a comment
There was a problem hiding this comment.
Right, a test for a C-level crash should ignore Python-level exceptions.
Sorry, something went wrong.
vstinner
left a comment
There was a problem hiding this comment.
LGTM, I just have a question/suggestion on the comment.
Sorry, something went wrong.
|
I proposed a different approach, explicitly reject exit code 1: PR #7147. |
Sorry, something went wrong.
vstinner
left a comment
There was a problem hiding this comment.
LGTM.
Sorry, something went wrong.
|
Thanks @Traceur759 for the PR, and @encukou for merging it 🌮🎉.. I'm working now to backport this PR to: 3.7. |
Sorry, something went wrong.
|
The change should also be backported to Python 3.6, no? @encukou |
Sorry, something went wrong.
|
@vstinner Python 3.6 does not have features which the test is written against. |
Sorry, something went wrong.
I'm talking about this code in the 3.6 branch: It's the commit 1da0479 of bpo-32374. Which feature is missing from 3.6? |
Sorry, something went wrong.
|
@vstinner Sorry, my bad, I had outdated local branch and didn't see the code there. Backporting is a valid question. |
Sorry, something went wrong.
|
Thanks @Traceur759 for the PR, and @encukou for merging it 🌮🎉.. I'm working now to backport this PR to: 3.6. |
Sorry, something went wrong.
I asked @miss-islington to create a PR, I tested it: it works as expected, so I approved the PR #7155. I will be merged as soon as tests pass. |
Sorry, something went wrong.
Previous test assuring that incorrect handling of m_traverse "shouts loudly" on crash
passed also when Python exception was raised.
Fix using try-except block is proposed.
https://bugs.python.org/issue32374