[llvm] [RISCV] Form VFWMUL.VF when fp_extend is scalar and then splatted. (PR #98590)

[Craig Topper via llvm-commits]

topperc wrote:

> LGTM
> 
> Just for the record, this is making an assumption that we have a vf variant for all widening vv instructions, and wf variant for all wv instructions supported by this transform. This appears to hold as far as I can tell.

I guess I should add an FADD test too.

https://github.com/llvm/llvm-project/pull/98590