[llvm] [AMDGPU] Implement IR expansion for frem instruction (PR #130988)

[Matt Arsenault via llvm-commits]

================
@@ -5671,6 +5671,10 @@ class TargetLowering : public TargetLoweringBase {
                                        LoadSDNode *OriginalLoad,
                                        SelectionDAG &DAG) const;
 
+  /// Indicates whether the FRem instruction should be expanded before
+  /// ISel in the LLVM IR.
+  virtual bool shouldExpandFRemInIR() const { return false; };
----------------
arsenm wrote:

Don't introduce a new hook for this. You can check if the operation is going to expand, and then check that there is no call in RuntimeLibcalls (this is essentially how PreISelIntrinsicLowering handles memcpy and co)

https://github.com/llvm/llvm-project/pull/130988