Fix LLVM-12 concat_split_op.h error (pytorch#66060)

Summary:
Pull Request resolved: pytorch#66060

Fixes
```
testTumHistoryAdditionalLaser (caffe2.caffe2.fb.layers.tests.tum_history_test.TestTumHistory) ... caffe2/caffe2/operators/concat_split_op.h:363:74: runtime error: applying non-zero offset 8 to null pointer
    #0 0x7f8f39d29795 in caffe2::ConcatOp<caffe2::CPUContext>::RunOnDevice() caffe2/caffe2/operators/concat_split_op.h:363
    #1 0x7f8f39c4978d in caffe2::Operator<caffe2::CPUContext>::Run(int) caffe2/caffe2/core/operator.h:987
    #2 0x7f8f381fe9c9 in caffe2::SimpleNet::Run() caffe2/caffe2/core/net_simple.cc:67
    #3 0x7f8f38ee488e in caffe2::Workspace::RunNet(std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&) caffe2/caffe2/core/workspace.cc:289
```

Test Plan: Sandcastle

Reviewed By: dzhulgakov, xush6528

Differential Revision: D31366205

fbshipit-source-id: 566aa519677c9d371189e4b1f81d595732861efc

Author: r-barnes

caffe2/operators/concat_split_op.h

@@ -6,6 +6,7 @@
 #include "caffe2/core/types.h"
 #include "caffe2/utils/math.h"
 #include "caffe2/utils/string_utils.h"
+#include <c10/util/accumulate.h>
 
 namespace caffe2 {
 
@@ -277,13 +278,13 @@ bool SplitByLengthsOp<Context>::RunOnDevice() {
 
 template <class Context>
 bool ConcatOp<Context>::RunOnDevice() {
-  auto* output = Output(0);
+  auto *const output = Output(0);
 
   // We can override default options(Context::GetDeviceType())
   // by explicitly passing in device type we want
-  Tensor* split = Output(
+  Tensor *const split = Output(
       1, at::IntArrayRef({InputSize()}), at::dtype<int>().device(CPU));
-  int* axis_data = split->template mutable_data<int>();
+  int *const axis_data = split->template mutable_data<int>();
   auto& input_zero = Input(0);
   int adj_size = input_zero.dim() + (add_axis_ ? 1 : 0);
   int canonical_axis = canonical_axis_index_(axis_, adj_size);
@@ -347,7 +348,13 @@ bool ConcatOp<Context>::RunOnDevice() {
   } else {
     output_dims[canonical_axis] = output_channels;
   }
+
   output->Resize(output_dims);
+  auto *const output_ptr = static_cast<char*>(output->raw_mutable_data(input_zero.dtype()));
+  if(output_ptr == nullptr){
+    return true;
+  }
+
   size_t output_offset = 0;
   for (int i = 0; i < InputSize(); ++i) {
     auto& input = Input(i);
@@ -358,8 +365,7 @@ bool ConcatOp<Context>::RunOnDevice() {
         axis_dim * after,
         input.raw_data(),
         axis_dim * after,
-        static_cast<char*>(output->raw_mutable_data(input_zero.dtype())) +
-            output_offset,
+        output_ptr + output_offset,
         output_channels * after,
         &context_,
         input_zero.dtype().copy());