[llvm] [NFC][AMDGPU] Add SLP to asm tests for int to fp casts (PR #227344)

[Matt Arsenault via llvm-commits]

================
@@ -0,0 +1,858 @@
+; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py UTC_ARGS: --version 6
+; RUN: opt -mtriple=amdgpu9.00-amd-amdhsa -passes=slp-vectorizer < %s | llc -mtriple=amdgpu9.00-amd-amdhsa | FileCheck -check-prefix=GFX900 %s
----------------
arsenm wrote:

doesn't SLP vectorizer run as part of the codegen pipeline? Why does it need the pre-opt run? 

https://github.com/llvm/llvm-project/pull/227344