Nils Friess

> I only found some UB in the tests, but I don't think the group tests were among the problematic tests. But I'll check again Checked again, there is actually...

How have you built LLVM and AdaptiveCpp? I remember vaguely that the optimisation levels between LLVM and AdaptiveCpp have to match on Windows, maybe this is a similar problem?

Thanks for the suggestions. I did some quick testing and I also think this is definitely something that can be implemented using `std::source_location` (or something similar in C++17). For this...

In my quick test I did something like this in `handler::single_task` ```c++ template #ifndef NDEBUG void single_task(KernelType kernelFunc, nostd::source_location loc = nostd::source_location::current()) { #else void single_task(KernelType kernelFunc) { #endif this->submit_kernel(...

Sorry, completely forgot to reply. We had some internal discussion about this, I don't remember exactly now but I think we came up with the following compromise: - Simply remove...

Note sure why tests are failing again, maybe it's something like in #1173, looking into it.

The segfault seems to occur in the group function tests. @DieGoldeneEnte I vaguely remember that you had some problems with weird segfaults at some point, maybe this is the same...

I think we only did it for `marray`. But this should require similar changes, I‘ll have a look.