-
Notifications
You must be signed in to change notification settings - Fork 675
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
test failures in 1.0.30 on s390x running Fedora (36) #2070
Comments
Do you actually have wasm2c running in production on s390x? Or is this just for CI/testing? |
At Fedora, we use |
I that case I suggest we simply declare s390x and unsupported architecture for wasm2c. |
If by "s390x" you mean "big-endian", then it should be easy to disable wasm2c build and testing there. However, a lot of effort went into fixing wabt to pass its testsuite on s390x in 1.0.24 (see #1670 and #1697), so I'd hate to see s390x support start to bit-rot again. @ngzhian would you be able to help here? |
I think its fine to say that wasm2c is only fully functional (spec conforming) on a fixed set of known/tested architectures. I don't see any reason to try to grow that list unless we have specific users who are asking for such support. Even if we fix these remaining tests, and get the test suite passing, I don't think we want to be spending future time/effort on supporting architectures that nobody is using production. Anybody who wants to work on fixing these is issues is obviously free to and I think, within reason, any patches would be most welcome. But I'd be hesitant to add it to CI since that would then burden all developers with the task of keeping all those tests passing which, again, I think we should do if it we have real users. |
I don't work directly on WebAssembly anymore so I won't be able to help there. |
I think we have the PR ready to add CI for s390x (#1971), and 100% agreed that it seems fine to skip the wasm2c tests when running on big-endian architectures. However, there is a failing unit test (unrelated to wasm2c) from an endian issue that is currently blocking this (#1972). If somebody can fix #1972, which appears like it might be pretty simple, I think it would be pretty simple to merge the s390x parts of #1971. |
I'm trying to release 1.0.30 packages for Fedora and I'm getting these new test failures on s390x:
The text was updated successfully, but these errors were encountered: