[Bug target/121587] _Float16 vector operations should use addps/subps/mulps/divps if F16C is present

[mkretz at gcc dot gnu.org via Gcc-bugs]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=121587

--- Comment #3 from Matthias Kretz (Vir) <mkretz at gcc dot gnu.org> ---
Hmm, maybe a sequence of 2 vfmadd...ps instructions could do range and
precision reduction with a combined latency of 8.