[llvm] [AMDGPU][True16] si-fold-operand selecting srcidx for v_mov_b16_t16_e64 (PR #162101)

[Brox Chen via llvm-commits]

================
@@ -3990,8 +3966,8 @@ static bool getFoldableImm(Register Reg, const MachineRegisterInfo &MRI,
   return false;
 }
 
-static bool getFoldableImm(const MachineOperand *MO, int64_t &Imm,
-                           MachineInstr **DefMI = nullptr) {
+bool SIInstrInfo::getFoldableImm(const MachineOperand *MO, int64_t &Imm,
----------------
broxigarchen wrote:

This will leads to a chain of functions to be converted to static so I decided to do the opposite

https://github.com/llvm/llvm-project/pull/162101