[llvm] [AMDGPU][True16][CodeGen] gisel true16 for G_Merge and ICMP (PR #128913)

[Brox Chen via llvm-commits]

================
@@ -639,7 +639,7 @@ bool AMDGPUInstructionSelector::selectG_MERGE_VALUES(MachineInstr &MI) const {
   LLT SrcTy = MRI->getType(MI.getOperand(1).getReg());
 
   const unsigned SrcSize = SrcTy.getSizeInBits();
-  if (SrcSize < 32)
+  if (SrcSize < 16)
----------------
broxigarchen wrote:

Since the the ICMP one has doubt, I'll move this to a seperate PR

https://github.com/llvm/llvm-project/pull/128913