-
-
Notifications
You must be signed in to change notification settings - Fork 10.9k
Complex math errors compiling under clang 5 / Windows #8768
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
Comments
@matthew-brett Have you tried this anytime recently? I've been trying this on some extensions that need uint128 and have had good success. Do you think that a ClangClCompiler would be worth trying? |
Hi - sorry - no - I haven't compiled with Clang on Windows recently - but yes - it would be very good to have a CI matrix entry for Clang on Windows. |
Master doesn't build right now due to differences in how intrinsics are included in MSVC and Clang. 1.16.x passes all tests with clang-cl, although I didn't bother building OpenBlas. |
See #13816 |
Going to close this, there is a new/remaining issue around einsum, but I think this is just outdated now (and the linked PRs give you to more work). Yes: CI matrix might be nice though. |
Here I have compiled with the clang-cl.exe frontend replacing the standard visual studio 2015
cl.exe
binary. I first installed http://llvm.org/pre-releases/win-snapshots/LLVM-5.0.0-r295586-win64.exe from http://llvm.org/builds , then:This results in the following test failures:
There are no test errors compiling and testing the same numpy checkout with Visual C++
cl.exe
(numpy commit 9bba99d). Any hints as to what might be going on?The text was updated successfully, but these errors were encountered: