[llvm] [AMDGPU] Optimize block count calculations to the new ABI (PR #174112)

[Joseph Huber via llvm-commits]

================
@@ -0,0 +1,275 @@
+; NOTE: Assertions have been autogenerated by utils/update_test_checks.py UTC_ARGS: --version 6
+; RUN: opt -mtriple=amdgcn-amd-amdhsa -S -passes=amdgpu-lower-kernel-attributes,instcombine,simplifycfg %s | FileCheck %s
----------------
jhuber6 wrote:

Okay, I just ran those to clean up the leftover instructions that this wouldn't cover since we don't handle removing the old instructions

https://github.com/llvm/llvm-project/pull/174112