[llvm] [Analysis][RISCV] More accurately estimate the cost of strided vector loads (PR #175135)

[Ryan Buchner via llvm-commits]

================
@@ -1274,6 +1274,25 @@ RISCVTTIImpl::getStridedMemoryOpCost(const MemIntrinsicCostAttributes &MICA,
       getMemoryOpCost(Opcode, VTy.getElementType(), Alignment, 0, CostKind,
                       {TTI::OK_AnyValue, TTI::OP_None}, I);
   unsigned NumLoads = getEstimatedVLFor(&VTy);
+  // Performant implementations of the vector extension will attempt to re-use
+  // elements if they fall on the same cache line
+  uint64_t CacheLineBytes = ST->getCacheLineSize();
+  if (!CacheLineBytes) // If no value, use default value of 64
+    CacheLineBytes = 64;
+  unsigned EltsPerCL = (CacheLineBytes * 8) / DataTy->getScalarSizeInBits();
+  if (const ConstantInt *StrideCI =
+          dyn_cast_or_null<ConstantInt>(MICA.getStrideVal())) {
+    uint64_t AbsStride = (uint64_t)std::abs(StrideCI->getSExtValue());
+    if (AbsStride < EltsPerCL) {
----------------
bababuck wrote:

Thanks. I mistakenly thought that `stride` was in terms of number of elements, but per [LangRef](https://llvm.org/docs/LangRef.html#llvm-experimental-vp-strided-load-intrinsic).
```
The ‘llvm.experimental.vp.strided.load’ intrinsic loads, into a vector,
scalar values from memory locations evenly spaced apart by
‘stride’ number of bytes, starting from ‘ptr’.
```

https://github.com/llvm/llvm-project/pull/175135