[llvm] [X86] Support vectorized llvm.fmaximum/fminimum.vXf16 lowering (PR #120988)

[Evgenii Kudriashov via llvm-commits]

e-kud wrote:

The origin of the change is here https://github.com/llvm/llvm-project/pull/93579#discussion_r1618076003

https://github.com/llvm/llvm-project/pull/120988