[llvm] [RISCV][VLOPT] Add support for more floating point instructions (PR #122326)

[Craig Topper via llvm-commits]

================
@@ -2925,3 +2925,123 @@ define <vscale x 4 x i32> @vid.v(<vscale x 4 x i32> %c, iXLen %vl) {
   %2 = call <vscale x 4 x i32> @llvm.riscv.vadd.nxv4i32.nxv4i32(<vscale x 4 x i32> poison, <vscale x 4 x i32> %1, <vscale x 4 x i32> %c, iXLen %vl)
   ret <vscale x 4 x i32> %2
 }
+
+define <vscale x 4 x float> @vfadd_vv(<vscale x 4 x float> %a, <vscale x 4 x float> %b, iXLen %vl) {
+; NOVLOPT-LABEL: vfadd_vv:
+; NOVLOPT:       # %bb.0:
+; NOVLOPT-NEXT:    fsrmi a1, 0
----------------
topperc wrote:

This fsrmi doesn't look right. Your rounding mode operand is 7 which shouldn't have an fsrmi.

https://github.com/llvm/llvm-project/pull/122326