-
Notifications
You must be signed in to change notification settings - Fork 10.8k
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
flang sometimes fails to compile gfortran test is_iostat_end_eor_1 on AArch64 buildbots #63263
Comments
@llvm/issue-subscribers-backend-aarch64 |
@luporl Can you take a look please? I am new to fortran so most importantly, is this error ever valid? I tried it with gfortran and it was fine: https://godbolt.org/z/bEfj8f8nz |
The problem is that there is a conflict between these tests:
iso_fortran_env_1.f90 defines an iso_fortran_env module, that gets created in the current dir as iso_fortran_env.mod. Fortunately, patch https://reviews.llvm.org/D152307 already addresses it. It's already accepted and should land soon. |
Use a dedicated module directory for each test. Any .mod files generated by the test at build-time will be written into this directory. The same directory is also used as the working directory for the the tests at test-time. Any files created/read/written by the test at test-time reside in that directory. This ensures that the tests don't race on the file system when run in parallel. The tests that were previously disabled because of these races have been re-enabled. This fixes PR #63263 (llvm/llvm-project#63263) Differential Revision: https://reviews.llvm.org/D152307
No failures seen since the patch landed. |
In the following builds, https://github.com/llvm/llvm-test-suite/blob/main/Fortran/gfortran/regression/is_iostat_end_eor_1.f90 fails to compile with this error:
Selection of examples:
https://lab.llvm.org/buildbot/#/builders/179/builds/6396
https://lab.llvm.org/buildbot/#/builders/179/builds/6262
https://lab.llvm.org/buildbot/#/builders/179/builds/6231
https://lab.llvm.org/buildbot/#/builders/184/builds/4299
https://lab.llvm.org/buildbot/#/builders/184/builds/4283
https://lab.llvm.org/buildbot/#/builders/197/builds/6532
https://lab.llvm.org/buildbot/#/builders/176/builds/2666
The failure happens intermittently across 1 and 2 stage bots, SVE and non-SVE, vector length agnostic and vector length specific.
In each build there is no obvious change that could have caused it.
The text was updated successfully, but these errors were encountered: