[llvm] [InstCombine] Transform `vector.reduce.add` and `splat` into multiplication (PR #161020)

[Gábor Spaits via llvm-commits]

================
@@ -308,3 +308,93 @@ define i32 @diff_of_sums_type_mismatch2(<8 x i32> %v0, <4 x i32> %v1) {
   %r = sub i32 %r0, %r1
   ret i32 %r
 }
+
----------------
spaits wrote:

It seems like in the case of i1 we won't even get to the point where this new optimization would be activated, because the i1 array will be changed to `i8`. I will check with other small types.

https://github.com/llvm/llvm-project/pull/161020