[llvm] [NVPTX] support packed f32 instructions for sm_100+ (PR #126337)

[Alex MacLean via llvm-commits]

================
@@ -5098,7 +5148,7 @@ static SDValue combinePackingMovIntoStore(SDNode *N,
   // Get the type of the operands being stored.
   EVT ElementVT = N->getOperand(Front).getValueType();
 
-  if (!Isv2x16VT(ElementVT))
+  if (!isPackedVectorTy(ElementVT))
----------------
AlexMaclean wrote:

Do we want to do this for v4i8? The current check seems tautological given the above check that ensures we're after legalization. 

https://github.com/llvm/llvm-project/pull/126337