[llvm] [AMDGPU][SIPreEmitPeephole] mustRetainExeczBranch: use BranchProbability and TargetSchedmodel (PR #109818)

[Juan Manuel Martinez Caamaño via llvm-commits]

================
@@ -1,95 +0,0 @@
-# NOTE: Assertions have been autogenerated by utils/update_mir_test_checks.py
-# RUN: llc -mtriple=amdgcn -mcpu=polaris10 -run-pass si-pre-emit-peephole -amdgpu-skip-threshold=1 -verify-machineinstrs  %s -o - | FileCheck %s
----------------
jmmartinez wrote:

This test was used to check that `SIPreEmitPeephole` did not optimize vmem memory accesses; but now they are handled.

I've added a commit before the main commit of this patch, to be submitted separately.
I restored it, renamed it, and added some branch weights such that the transformation is profitable.

https://github.com/llvm/llvm-project/pull/109818