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
[DO NOT SUBMIT] Reproduce MacOS + Py3.8 mozinstall failure #24813
Conversation
b279d98
to
4e25089
Compare
At this point I'm fairly confident that this is some very, very strange issue where we seem to somehow be running old mozinstall code, even as the filesystem tells us new code is there (seriously, https://dev.azure.com/web-platform-tests/b14026b4-9423-4454-858f-bf76cf6d1faa/_apis/build/builds/51979/logs/213 is a fun read... an error thrown in a file that when later cat'd has different content). I have been digging in on-and-off via runs on the bots (oh, to own a mac device...), will summarize my findings so far tomorrow. EDIT: But a note for myself for tomorrow - in https://dev.azure.com/web-platform-tests/b14026b4-9423-4454-858f-bf76cf6d1faa/_apis/build/builds/51986/logs/206 it is the old code in _venv3 post-test !!! |
I've gone down the rabbit hole and found the root cause. First of all, some of wpt/tools/wpt/tests/test_install.py Line 89 in bd98c68
(Some other tests create venvs themselves, e.g. test_run.py .)
Line 150 in bd98c68
This means that only the test cases use the dependencies installed by tox; everything inside Now to the root: we manage Lines 128 to 132 in bd98c68
The newest release of Finally, the Lines 54 to 55 in bd98c68
unlike in tools/wptrunner/requirements_firefox.txt .
Therefore, when you run |
@jgraham has published the wheel for mozinstall 2.0.0. Let's try again now. |
The tests in question indeed pass now. Filed #24829 to keep track of the venv management issue. |
No description provided.