-
-
Notifications
You must be signed in to change notification settings - Fork 2.7k
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
Selftests sometimes fail on pypy: TypeError: expected some sort of stmt, but got <_ast.Load object ...>
#11771
Comments
I have same failure on older versions as well. E.g. 7.4.3: This happens for all tested in the mpmath pypy versions (3.8-3.10). But most often on 3.10. |
Another example: #11168 (comment) @cfbolz, I don't see a bugreport in pypy issues. Is there any chance this might be related to #11168 (an fixed in pypy)? |
@skirpichev it's not impossible that it's the same problem, yes. You can try a nightly PyPy build (or wait for the next release but that is still a bit out). |
@cfbolz, thanks for a suggestion. I did tests for nightly builds (on 3.8-3.10) and, after ~20 re-runs, I got a TypeError on pypy3.8, which is similar to this issue:
Summary page: https://github.com/mpmath/mpmath/actions/runs/8491432157?pr=772
|
Yes, the 3.8 variants are not maintained any more, you need to try 3.9 or 3.10. Thanks for trying this out! |
Well, I did ~50 re-runs of mpmath's tests on pypy3.9 and pypy3.10 - without any failures. Given the previous statistics, probably there is a chance that this issue was fixed in PyPy's nightly builds. I'll switch regular tests for pypy3.9/10 to them. Thanks for bugfix! |
thank you for trying this, that's very valuable feedback! |
During the sprints, we were seeing 3–4 variants of different AST object reprs there. Hopefully, bumping PyPy in CI will fix this. Ronny suggested this, but nobody got to actually doing it... |
FTR @LilyFoote linked https://www.pypy.org/posts/2024/03/fixing-bug-incremental-gc.html in #11168 (comment) but that reference never got cross-posted into this issue. |
Action item for whoever gets to prepare the PR:
|
Working on this issue |
PyPy 3.8 has a flaky bug in the garbage collector that is not going to be fixed there since that version is EoL, but newer version have it addressed. Here's the problem we're seeing in the CI - TypeError: expected some sort of stmt, but got <_ast.Load object ...> Here's the Link to the explanation - pytest-dev#11771 (comment) Resolved pytest-dev#11771
I even saw a case of E NameError: name '@py_assert2' is not defined (https://github.com/pytest-dev/pytest/actions/runs/9746885084/job/26898335196#step:6:97) |
PyPy 3.8 has a flaky bug in the garbage collector that is not going to be fixed there since that version is EoL, but newer version have it addressed. Here's the problem we're seeing in the CI ```TypeError: expected some sort of stmt, but got <_ast.Load object ...> ``` Here's the Link to the explanation Resolves pytest-dev#11771
PyPy 3.8 has a flaky bug in the garbage collector that is not going to be fixed there since that version is EoL, but newer version have it addressed. Here's the problem we're seeing in the CI: ``` TypeError: expected some sort of stmt, but got <_ast.Load object ...> ``` Here's the Link to the explanation Resolves pytest-dev#11771
PyPy 3.8 has a flaky bug in the garbage collector that is not going to be fixed there since that version is EoL, but newer version have it addressed. Here's the problem we're seeing in the CI: "TypeError: expected some sort of stmt, but got <_ast.Load object ...>" Here's the Link to the explanation Resolves pytest-dev#11771
PyPy 3.8 has a flaky bug in the garbage collector that is not going to be fixed there since that version is EoL, but newer versions have it addressed. The problem manifests itself in CI as follows: ```console TypeError: expected some sort of stmt, but got <_ast.Load object ...> ``` This patch makes use of the PyPy version that is still supported and contains a fix for the said bug. Ref: https://www.pypy.org/posts/2024/03/fixing-bug-incremental-gc.html Resolves pytest-dev#11771.
PyPy 3.8 has a flaky bug in the garbage collector that is not going to be fixed there since that version is EoL, but newer versions have it addressed. The problem manifests itself in CI as follows: ```console TypeError: expected some sort of stmt, but got <_ast.Load object ...> ``` This patch makes use of the PyPy version that is still supported and contains a fix for the said bug. Ref: https://www.pypy.org/posts/2024/03/fixing-bug-incremental-gc.html Resolves pytest-dev#11771.
Co-authored-by: Sviatoslav Sydorenko <[email protected]>
This is necessary because: * `pypy-3.8` is EOL * `pypy-3.8` has flaky SEGFAULTs on import [[1]] [[2]] [[3]] due to a bug in their GC that is fixed in `pypy-3.9` [1]: https://github.com/jazzband/pip-tools/actions/runs/12162197242/job/33918558133?pr=2106#step:8:59 [2]: pytest-dev/pytest#11771 (comment) [3]: https://pypy.org/posts/2024/03/fixing-bug-incremental-gc.html
This is necessary because: * `pypy-3.8` is EOL * `pypy-3.8` has flaky SEGFAULTs on import [[1]] [[2]] [[3]] due to a bug in their GC that is fixed in `pypy-3.9` * `pypy-3.9` has a bug that does not have a backport with the fix [[4]]. Furthermore, PyPy maintainers recommend using 3.10 [[5]]. [1]: https://github.com/jazzband/pip-tools/actions/runs/12162197242/job/33918558133?pr=2106#step:8:59 [2]: pytest-dev/pytest#11771 (comment) [3]: https://pypy.org/posts/2024/03/fixing-bug-incremental-gc.html [4]: tox-dev/tox#3284 [5]: pypy/pypy#4958 (comment)
Example: https://github.com/pytest-dev/pytest/actions/runs/7408595776/job/20157160097
This doesn't always happen, maybe 50% of the time. Started in the last few days. I checked if the PyPy version changed during these days, but seems not (it's 3.8.16).
The error is:
Sometimes it says
TypeError: expected some sort of stmt, but got <_ast.Load object at 0x000000000365d040>
instead.The text was updated successfully, but these errors were encountered: