Llvm-args=-force-vector-width does vectorization ignoring the specified addition order

An example of summing a slice compiled with the "-C llvm-args=-force-vector-width=8" flag. The code is vectorized, ignoring the order of floating point operations. In this case, the separate instruction "-C llvm-args=-ffast-math" does not lead to any result.
Question: Does the -force-vector-width flag include -ffast-math? and why does vectorization not work when -ffast-math is specified separately (without -force-vector-width)?

Seems to be [RISCV] When enabling `-force-vector-width`, RVV vectorization has produced wrong results · Issue #77044 · llvm/llvm-project · GitHub

This topic was automatically closed 90 days after the last reply. We invite you to open a new topic if you have further questions or comments.