[llvm] [AMDGPU] Disable inline constants for pseudo scalar transcendentals (PR #104395)

[Matt Arsenault via llvm-commits]

================
@@ -946,6 +946,14 @@ class SIInstrInfo final : public AMDGPUGenInstrInfo {
            Opcode == AMDGPU::DS_GWS_BARRIER;
   }
 
+  bool isF16PseudoScalarTrans(unsigned Opcode) const {
----------------
arsenm wrote:

this can be static 

https://github.com/llvm/llvm-project/pull/104395