[libcxx-commits] [libcxx] [lldb] [clang] [lld] [libc] [llvm] [mlir] [flang] [openmp] [clang-tools-extra] [SLP]Add support for strided loads. (PR #80310)

[Alexey Bataev via libcxx-commits]

================
@@ -3930,30 +4065,68 @@ static LoadsState canVectorizeLoads(ArrayRef<Value *> VL, const Value *VL0,
       std::optional<int> Diff =
           getPointersDiff(ScalarTy, Ptr0, ScalarTy, PtrN, DL, SE);
       // Check that the sorted loads are consecutive.
-      if (static_cast<unsigned>(*Diff) == VL.size() - 1)
+      if (static_cast<unsigned>(*Diff) == Sz - 1)
         return LoadsState::Vectorize;
       // Simple check if not a strided access - clear order.
-      IsPossibleStrided = *Diff % (VL.size() - 1) == 0;
+      bool IsPossibleStrided = *Diff % (Sz - 1) == 0;
+      // Try to generate strided load node if:
+      // 1. Target with strided load support is detected.
+      // 2. The number of loads is greater than MinProfitableStridedLoads,
+      // or the potential stride <= MaxProfitableLoadStride and the
+      // potential stride is power-of-2 (to avoid perf regressions for the very
+      // small number of loads) and max distance > number of loads, or potential
+      // stride is -1.
+      // 3. The loads are ordered, or number of unordered loads <=
+      // MaxProfitableUnorderedLoads, or loads are in reversed order.
+      // (this check is to avoid extra costs for very expensive shuffles).
+      if (IsPossibleStrided && (((Sz > MinProfitableStridedLoads ||
+                                  (static_cast<unsigned>(std::abs(*Diff)) <=
+                                       MaxProfitableLoadStride * Sz &&
+                                   isPowerOf2_32(std::abs(*Diff)))) &&
+                                 static_cast<unsigned>(std::abs(*Diff)) > Sz) ||
+                                *Diff == -(static_cast<int>(Sz) - 1))) {
+        int Stride = *Diff / static_cast<int>(Sz - 1);
----------------
alexey-bataev wrote:

This is stride in "scalar elements", here the size in bytes is not important, getPointersDiff() handles pointers with different types (sizes). Here we just looking that the pointers have proportional constant distances between them.

https://github.com/llvm/llvm-project/pull/80310