temporary benchmark, <vvaddint32> working

rvvector/benchmark_temp/vx_vec_saxpy_float.s

@@ -0,0 +1,28 @@
+.type vx_vec_saxpy, @function
+.global vx_vec_saxpy_float
+# void
+# saxpy(size_t n, const float a, const float *x, float *y)
+# {
+#   size_t i;
+#   for (i=0; i<n; i++)
+#     y[i] = a * x[i] + y[i];
+# }
+#
+# register arguments:
+#     a0      n
+#     fa0     a
+#     a1      x
+#     a2      y
+vx_vec_saxpy_float:
+    vsetvli a4, a0, e32, m8
+saxpy:
+    vlw.v v0, (a1)
+    sub a0, a0, a4
+    slli a4, a4, 2
+    add a1, a1, a4
+    vlw.v v8, (a2)
+    vfmacc.vf v8, fa0, v0
+    vsw.v v8, (a2)
+    add a2, a2, a4
+    bnez a0, saxpy
+    ret