1 2 3 4 5
// vfrdiv.vf vd, vs2, rs1, vm # scalar-vector, vd[i] = f[rs1]/vs2[i] VI_VFP_VF_LOOP ({ vd = f32_div(rs1, vs2); })