Wrong argument names for .vf widening multiply-add intrinsics

[dzaima]

e.g.

https://github.com/riscv-non-isa/rvv-intrinsic-doc/blob/207ac85b7d07df8bf268ee78aa2bb26a350f743f/auto-generated/bfloat16/intrinsic_funcs.adoc?plain=1#L1427-L1428

__bf16 vs1 should be __bf16 rs1.

Similarly, auto-generated/intrinsic_funcs.adoc has:

vfloat32mf2_t __riscv_vfwmacc_vf_f32mf2(vfloat32mf2_t vd, _Float16 vs1,
                                        vfloat16mf4_t vs2, size_t vl);
...
vfloat64m1_t __riscv_vfwnmsac_vf_f64m1(vfloat64m1_t vd, float vs1,
                                       vfloat32mf2_t vs2, size_t vl);