[clang] [Clang] VectorExprEvaluator::VisitCallExpr / InterpretBuiltin - allow AVX/AVX512 subvector insertion intrinsics to be used in constexpr #157709 (PR #158778)

[Simon Pilgrim via cfe-commits]

================
@@ -2937,6 +2937,50 @@ static bool interp__builtin_elementwise_triop(
   return true;
 }
 
+static bool interp__builtin_x86_insert_subvector(InterpState &S, CodePtr OpPC,
+                                                 const CallExpr *Call,
+                                                 unsigned ID) {
+  assert(Call->getNumArgs() == 3);
+
+  PrimType ImmPT = *S.getContext().classify(Call->getArg(2));
+  APSInt ImmAPS = popToAPSInt(S.Stk, ImmPT);
+  uint64_t Index = ImmAPS.getZExtValue();
+
+  const Pointer &SubVec = S.Stk.pop<Pointer>();
+  if (!SubVec.getFieldDesc()->isPrimitiveArray())
+    return false;
+
+  const Pointer &DstVec = S.Stk.pop<Pointer>();
+  if (!DstVec.getFieldDesc()->isPrimitiveArray())
+    return false;
+
+  const Pointer &Result = S.Stk.peek<Pointer>();
+
+  unsigned DstElements = DstVec.getNumElems();
+  unsigned SubElements = SubVec.getNumElems();
+
+  if (SubElements == 0 || DstElements == 0 || (DstElements % SubElements) != 0)
----------------
RKSimon wrote:

This can probably be an assert - sema checking should catch this for use anyhow

https://github.com/llvm/llvm-project/pull/158778