[llvm] [LLVM][CodeGen][SVE] Improve lowering of fixed length masked mem ops. (PR #134402)

[David Sherwood via llvm-commits]

================
@@ -28697,17 +28703,36 @@ static SDValue convertFixedMaskToScalableVector(SDValue Mask,
   SDLoc DL(Mask);
   EVT InVT = Mask.getValueType();
   EVT ContainerVT = getContainerForFixedLengthVector(DAG, InVT);
-
-  auto Pg = getPredicateForFixedLengthVector(DAG, DL, InVT);
+  SDValue Pg = getPredicateForFixedLengthVector(DAG, DL, InVT);
 
   if (ISD::isBuildVectorAllOnes(Mask.getNode()))
     return Pg;
 
-  auto Op1 = convertToScalableVector(DAG, ContainerVT, Mask);
-  auto Op2 = DAG.getConstant(0, DL, ContainerVT);
+  bool InvertCond = false;
+  if (isBitwiseNot(Mask)) {
+    InvertCond = true;
+    Mask = Mask.getOperand(0);
+  }
+
+  SDValue Op1, Op2;
+  ISD::CondCode CC;
+
+  // When Mask is the result of a SETCC, it's better to regenerate the compare.
+  if (Mask.getOpcode() == ISD::SETCC) {
+    Op1 = convertToScalableVector(DAG, ContainerVT, Mask.getOperand(0));
----------------
david-arm wrote:

Is `ContainerVT` guaranteed to be correct for the SETCC inputs? It looks like we base `ContainerVT` on the result. I'm not sure if something like this is legal for NEON:

v4i32 = SETCC NE v4i16 %a, v4i16 %b

since v4i16 is also a legal type. I'm just a bit worried that we're effectively promoting a type here and, if so, is that a problem?

https://github.com/llvm/llvm-project/pull/134402