RQ.- Does -ffast-math flag always enable vectorization of reductions of
floats in GCC, even for scalars?
- Intel(R) Xeon(R) Silver 4216 CPU @ 2.10GHz
- Compiler: GCC 11.2.0, Intel CC
- MARTA Profiler Full Configuration file
-
Profiling
# place at the root of the project marta_profiler examples/autovectorization/config_profiler.yml # or python -m marta.profiler examples/autovectorization/config_profiler.ymlThis step should take many hours.