[llvm] [AMDGPU] Convert more 64-bit lshr to 32-bit if shift amt>=32 (PR #138204)

[via llvm-commits]

================
@@ -0,0 +1,561 @@
+; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py UTC_ARGS: --version 5
+;; Test reduction of:
+;;
+;;   DST = lshr i64 X, Y
+;;
+;; where Y is in the range [63-32] to:
+;;
+;;   DST = [srl i32 X, (Y & 0x1F), 0]
+
+; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx900 < %s | FileCheck %s
+
+;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
+; Test range with metadata
+;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
+
+define i64 @srl_metadata(i64 %arg0, ptr %arg1.ptr) {
----------------
LU-JOHN wrote:

Test pre-committed in https://github.com/llvm/llvm-project/pull/138281

https://github.com/llvm/llvm-project/pull/138204