-
Notifications
You must be signed in to change notification settings - Fork 278
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
Hide windows x86 random failures on CI #1655
Conversation
Codecov Report
@@ Coverage Diff @@
## main #1655 +/- ##
=======================================
Coverage 66.92% 66.92%
=======================================
Files 151 151
Lines 20807 20807
=======================================
Hits 13925 13925
Misses 6882 6882 Continue to review full report at Codecov.
|
I initially thought the initial loop conversion PR was the cause. But reverting it doesn't seem to do anything. I don't have a MSVC setup either. |
5066ed5
to
cc9c644
Compare
67aa85d
to
5b7be5f
Compare
8d659e7
to
f71413c
Compare
On CI, pythonTests fail more or less consistently in the configuration Arch:x86, BuildType:Relase, Shared:ON. I tried to reproduce this locally, but I could not. Since Win-x86 is marginal for Exiv2, we disable it from CI until somebody raise concerns about it.
The idea is to generate a conan cache on main that could be reused in future PRs to main.
Note that I have updated the description with information about the changes done and the decision taken to not run the tests on Win x86. |
Works for me. x86 windows is a dying platform. X86 Linux and BSD are already not a thing. |
After few days investigating the random windows x86 failures on CI (see link), I decided to give up since I could not reproduce them locally.
Anyways, the investigation bringts some improvements that I list here:
/Ox /Zo
).cmake_find_package
generator we can simplify few things.tests/bugfixes/github/test_issue_1472.py
was failing on windows without a bash terminal.main
on push events.Finally I decided to not run the tests on windows x86 architectures to hide the failures. I think we do not lose too much since Windows x86 is marginal for Exiv2 and furthermore, I could not reproduce the failures locally.