================
@@ -185,4 +185,174 @@ exit:
ret i32 %sub
}
+define float @fsub_reduction(float %startval, ptr %src1, ptr %src2) #0 {
----------------
sdesmalen-arm wrote:
Could you also add a test with both fadd and fsub?
https://github.com/llvm/llvm-project/pull/191186