[llvm] [AMDGPU] Reject non-fp16 image sample data in D16 detection while lowering image (PR #213272)

[Matt Arsenault via llvm-commits]

================
@@ -10440,6 +10440,11 @@ SDValue SITargetLowering::lowerImage(SDValue Op,
       VData = Op.getOperand(2);
 
       MVT StoreVT = VData.getSimpleValueType();
+      if (StoreVT.getScalarType().getSizeInBits() == 16 &&
----------------
arsenm wrote:

Drop the size check?

https://github.com/llvm/llvm-project/pull/213272