[llvm] [InstCombine] Constant fold binops through `vector.insert` (PR #164624)

[Paul Walker via llvm-commits]

================
@@ -0,0 +1,139 @@
+; NOTE: Assertions have been autogenerated by utils/update_test_checks.py
+; RUN: opt -S -passes=instcombine %s | FileCheck %s
----------------
paulwalker-arm wrote:

Please add a RUN line containing the relevant `-use-constant-*-splat` flags. Ideally the behaviour will be the same but that's not required.

Perhaps worth adding a floating-point test?


https://github.com/llvm/llvm-project/pull/164624