aricer123

What are the next steps for me? Should I switch to single precision mode and commit to test on the GPUs? I noticed it is also showing that the code...

@sbryngelson Wouldn't it be better to replace the 1d-6 with a parametrized type like real(wp), parameter :: dflt_real = (-1_wp * (10._wp ** 6)) instead of 1e-6. I have finished...

@sbryngelson incorporating the patch file seems to be failing tests, i simply incorporated the changes along with the fixes to m_precision_select. Did this happen on your system? What could be...

This was the command i used to try to replace all instances. This did a good job of replacing all instances. find src -type f \( -name "*.f90" -o -name...

It succeeded in replacing all instances as far as I could see and the code built. But, it failed tests. I stopped working on it after we decided to go...

> The precision is not agnostic; it defaults to float precision (single). I don't know what case you are referring to for failing tests. My tests pass in double precision...

Its okay I think I'll try taking a look at this on my own for now because I now have a suspicion of what the error may be. I'll let...

find src -type f \( -name "*.f90" -o -name "*.fpp" \) -exec sed -i '' -E \ 's/([+-]?[0-9]+(\.[0-9]*)?)[dD]([+-]?[0-9]+)/\1e\3_wp/g' {} + replaced all intances. I fixed the bug that was causing...

By the way, is it possible for me to trigger the full test suite runs for future commits or do you only have access to that?

Ok. Keep me posted on what you think of needs to be added prior to merging. Perhaps the flags could be a different PR/issue.