!272 Increase the max tensor size

Merge pull request !272 from jiangzhenguang/Increase-the-max-size-of-tensor

mindspore/ccsrc/kernel/aicpu/aicpu_kernel_build.cc

@@ -65,16 +65,16 @@ bool SetIOIputSize(const std::shared_ptr<AnfNode> &anf_node, const size_t &input
     } else {
       auto type_ptr = TypeIdToType(AnfAlgo::GetInputDeviceDataType(anf_node, i));
       MS_EXCEPTION_IF_NULL(type_ptr);
-      int size_i = 1;
+      int64_t size_i = 1;
       for (size_t j = 0; j < shape_i.size(); j++) {
-        IntMulWithOverflowCheck(size_i, static_cast<int>(shape_i[j]), &size_i);
+        size_i = LongMulWithOverflowCheck(size_i, static_cast<int>(shape_i[j]));
       }
       size_t type_byte = GetTypeByte(type_ptr);
       if (type_byte == 0) {
         return false;
       }
-      IntMulWithOverflowCheck(size_i, SizeToInt(type_byte), &size_i);
-      input_size_list->push_back(IntToSize(size_i));
+      size_i = LongMulWithOverflowCheck(size_i, SizeToInt(type_byte));
+      input_size_list->push_back(LongToSize(size_i));
     }
   }
   return true;
@@ -97,16 +97,16 @@ bool SetIOSize(const std::shared_ptr<AnfNode> &anf_node, const std::shared_ptr<A
     std::vector<size_t> shape_i = AnfAlgo::GetOutputDeviceShape(anf_node, i);
     TypePtr type_ptr = TypeIdToType(AnfAlgo::GetOutputDeviceDataType(anf_node, i));
     MS_EXCEPTION_IF_NULL(type_ptr);
-    int size_i = 1;
+    int64_t size_i = 1;
     for (size_t j = 0; j < shape_i.size(); j++) {
-      IntMulWithOverflowCheck(size_i, static_cast<int>(shape_i[j]), &size_i);
+      size_i = LongMulWithOverflowCheck(size_i, static_cast<int>(shape_i[j]));
     }
     size_t type_byte = GetTypeByte(type_ptr);
     if (type_byte == 0) {
       return false;
     }
-    IntMulWithOverflowCheck(size_i, SizeToInt(type_byte), &size_i);
-    output_size_list.push_back(IntToSize(size_i));
+    size_i = LongMulWithOverflowCheck(size_i, SizeToInt(type_byte));
+    output_size_list.push_back(LongToSize(size_i));
   }
   kernel_mod_ptr->SetOutputSizeList(output_size_list);
   return true;

mindspore/ccsrc/operator/prim_structures.cc

@@ -582,7 +582,7 @@ AbstractBasePtr InferImplShapeMul(const AnalysisEnginePtr &, const PrimitivePtr
   int result = 1;
   for (size_t i = 0; i < shpx_data.size(); i++) {
     int value = GetValue<int>(shpx_data[i]);
-    IntMulWithOverflowCheck(result, value, &result);
+    result = IntMulWithOverflowCheck(result, value);
   }
 
   auto result_v = MakeValue(result);

mindspore/ccsrc/utils/convert_utils_base.h

@@ -91,15 +91,26 @@ inline unsigned int UlongToUint(size_t u) {
   return static_cast<unsigned int>(u);
 }
 
-inline void IntMulWithOverflowCheck(int a, int b, int *c) {
+inline int IntMulWithOverflowCheck(int a, int b) {
   int out = a * b;
   if (a != 0) {
-    bool ok = ((out / a) != b);
-    if (ok) {
+    bool overflow = ((out / a) != b);
+    if (overflow) {
       MS_LOG(EXCEPTION) << "Mul: a(" << a << ") * b(" << b << ") result is overflow";
     }
   }
-  *c = out;
+  return out;
+}
+
+inline int64_t LongMulWithOverflowCheck(int64_t a, int64_t b) {
+  int64_t out = a * b;
+  if (a != 0) {
+    bool overflow = ((out / a) != b);
+    if (overflow) {
+      MS_LOG(EXCEPTION) << "Mul: a(" << a << ") * b(" << b << ") result is overflow";
+    }
+  }
+  return out;
 }
 
 inline size_t SizetMulWithOverflowCheck(size_t a, size_t b) {