Hi to all,
I am testing a in-house CFD code with massive floating point operations in a system with the Ryzen 7 5800X CPU. Using gnu compiler, it works flawlessly. Under AOCC 2.3 compiler, in debug mode (O0 optimization) it also works properly. When I try to use O2 or O3 options, the code segafaults or gives NaN errors. In O1 it works again, but with poor performance. There are no other option in compilation. Some hints? Thank you.