PierUgit

Would the new layout be enabled by the user (e.g. `complex(layout=split)`) or automatically by the compiler?

> Calling any fortran experts to help debug or diagnose... It's actually not specific to Fortran, but to the Intel compilers when OpenMP is used. I can reproduce the same...

Looks like it has been fixed with the 2024 version if the Intel compilers 2023: https://godbolt.org/z/W9x7WTEfM 2024: https://godbolt.org/z/dxeheGdez

> so if I summarize, we have: > > * the approach with the @4, which is discouraged. > > * the possibility to bind the deprecated '_sleep'. If I...

It's definitely OS-dependent, and even worse maybe compiler-dependent. I assume that on Posix systems all compilers will provide the `mmap` features, but on Windows I don't know. The Intel compiler...

I will try porting the demo code to Windows and see... I have to set up a dev environment on my multimedia mini-PC first (the one that sits below the...

An example code (even with some dust :)) could definitely help. And you're also welcome if you want to help.

I could install the MSYS2 environment on my office laptop, and start porting the code to Windows. I have a rough version of the anonymous mapping and it looks OK,...

The code now runs on Windows :)

Excellent approach! I'm wondering if the function versions should have an error handler at all, or just abort in case of any error? The problem with the error handler is...