[llvm] [LAA] Add stencil group merging to reduce runtime pointer checks (PR #187252)

[Florian Hahn via llvm-commits]

================
@@ -737,6 +752,367 @@ void RuntimePointerChecking::groupChecks(
   }
 }
 
+/// Result of decomposing a SCEV expression into stencil offset form:
+///   Offset = Constant + sum(Coefficients[stride] * stride)
+/// where each stride is a loop-invariant SCEV expression.
+struct StencilDecomposition {
+  int64_t Constant = 0;
+  /// Map from loop-invariant stride SCEV to its integer coefficient.
+  SmallMapVector<const SCEV *, int64_t, 4> Coefficients;
+};
+
+/// Try to decompose \p Expr into a stencil offset function of loop-invariant
+/// strides: C + a1*s1 + a2*s2 + ...
+/// Relies on SCEV's canonical form: AddExpr operands are flattened (N-ary),
+/// MulExpr has the constant operand first when present.
+/// Returns std::nullopt if the expression contains non-stencil terms or any
+/// SCEV constant doesn't fit in int64_t (we commit to the signed
+/// interpretation; values that need more than 64 significant bits are
+/// out of scope).
+static std::optional<StencilDecomposition>
+decomposeStencilOffset(const SCEV *Expr, ScalarEvolution &SE, const Loop &L) {
+  StencilDecomposition D;
+
+  auto ToInt64 = [](const APInt &V) -> std::optional<int64_t> {
+    if (V.getSignificantBits() > 64)
+      return std::nullopt;
----------------
fhahn wrote:

could you add a test that hits this codepath? Something like an address space with i128 index width 

https://github.com/llvm/llvm-project/pull/187252