Split op oneDNN AVX2 fix (#33944)

* added checking if md uses blocking format

* minor change

* removed unnecessary line

paddle/fluid/operators/split_op.cc

@@ -78,6 +78,15 @@ class SplitOp : public framework::OperatorWithKernel {
 
 #ifdef PADDLE_WITH_MKLDNN
     if (this->CanMKLDNNBeUsed(ctx, input_data_type)) {
+      // OneDNN uses blocking format, which cannot be always
+      // supported with reorders, because if blocked dimension is not divisible
+      // by
+      // 8 or 16(depending on which blocking format is used) submemory cannot be
+      // created, so in that scenario a fallback is needed
+      auto tmp_md = dnnl::memory::desc(
+          framework::vectorize(ctx.Input<Tensor>("X")->dims()),
+          dnnl::memory::data_type::f32, ctx.Input<Tensor>("X")->format());
+      if (tmp_md.data.format_desc.blocking.inner_nblks == 0)
         return framework::OpKernelType(input_data_type, ctx.GetPlace(),
                                        framework::DataLayout::kMKLDNN,
                                        framework::LibraryType::kMKLDNN);