Kelly (KT) Thompson

I am also struggling this as I try to move from MS-MPI 8 to version 10. Is the `cfg_stub.c` work around the only path forward?

This works for me: ```console C:\>mpiexec Microsoft MPI Startup Program [Version 10.0.12498.5] Launches an application on multiple hosts. Usage: mpiexec [options] executable [args] [ : [options] exe [args] : ......

👍 Thanks for putting this fix together. I'm seeing the same issues with gfortran.

@scivision @albertziegenhagel Do you know of a gfortran compiler flag or code pragma that will suppress this warning? I would like to silence this warning in my Appveyor CI. Thanks!

Here is the full patch file that I am using: ```diff diff --git a/include/experimental/__p0009_bits/config.hpp b/include/experimental/__p0009_bits/config.hpp index 2f9031e..265bd39 100644 --- a/include/experimental/__p0009_bits/config.hpp +++ b/include/experimental/__p0009_bits/config.hpp @@ -65,6 +65,14 @@ # define _MDSPAN_COMPILER_APPLECLANG #endif...

> Can you check out the https://github.com/kokkos/mdspan/tree/fix-msvc-gtest-tests branch and see if it works for you? At least for my installation, it doesn't work with `-DMDSPAN_ENABLE_TESTS=ON` ```console >------ Build All started:...

@crtrott I appreciate you looking at this. I have an appveyor MSVC build that shows the build issues I'm hitting at https://github.com/lanl/Draco/pull/814. Scroll down to the "AppVeyor build failed" block...

I've created a mini-repository that shows the build issues we are having with MSVC 2019 (our compiles are limited to C++14, see `CMakeLists.txt`) at https://github.com/KineticTheory/mdspan-msvc2019. This simple test trips over...

I'm having a similar pain-point. `jscpd` is taking about 15 minutes to parse through 100k LOC. I'm running in GitHub's [super-linter image](https://github.com/github/super-linter).

@sebaFlame Thank you for adding the notes above. Your experience provided a hit that fixed my issue. I removed all files from `%TEMP%`, rebooted and now `winget` is working again....