fix

tensorflow/c/eager/dlpack.cc

@@ -40,6 +40,7 @@ struct TfDlManagedTensorCtx {
       : reference(ref), shape(), tensor() {}
 };
 
+// Get tensor from eager tensor handle
 const Tensor* GetTensorFromHandle(TFE_TensorHandle* h, TF_Status* status) {
   if (h == nullptr || !h->handle->IsValid(&status->status)) {
     status->status = tensorflow::errors::InvalidArgument(
@@ -63,6 +64,7 @@ const Tensor* GetTensorFromHandle(TFE_TensorHandle* h, TF_Status* status) {
   return tensor;
 };
 
+// Deleter for DLManagedTensor
 void DLManagedTensorDeleter(DLManagedTensor* arg) {
   TfDlManagedTensorCtx* owner =
       static_cast<TfDlManagedTensorCtx*>(arg->manager_ctx);
@@ -129,47 +131,7 @@ DLContext GetDLContext(TFE_TensorHandle* h, TF_Status* status) {
   return ctx;
 }
 
-DLManagedTensor* TFEHandleToTfDlManagedTensorCtx(TFE_TensorHandle* h,
+// Convert DLContext to TF device name
-                                                 TF_Status* status) {
-  const Tensor* tensor = GetTensorFromHandle(h, status);
-  TF_DataType data_type = static_cast<TF_DataType>(tensor->dtype());
-  TensorReference tensor_ref(*tensor);  // This will call buf_->Ref()
-
-  auto* tf_dlm_tensor_ctx = new TfDlManagedTensorCtx(tensor_ref);
-  tf_dlm_tensor_ctx->reference = tensor_ref;
-
-  DLManagedTensor* dlm_tensor = &tf_dlm_tensor_ctx->tensor;
-  dlm_tensor->manager_ctx = tf_dlm_tensor_ctx;
-  dlm_tensor->deleter = &DLManagedTensorDeleter;
-  dlm_tensor->dl_tensor.ctx = GetDLContext(h, status);
-  int ndim = tensor->dims();
-  dlm_tensor->dl_tensor.ndim = ndim;
-  dlm_tensor->dl_tensor.data = TFE_TensorHandleDevicePointer(h, status);
-  dlm_tensor->dl_tensor.dtype = GetDLDataType(data_type, status);
-
-  std::vector<int64_t>* shape_arr = &tf_dlm_tensor_ctx->shape;
-  std::vector<int64_t>* stride_arr = &tf_dlm_tensor_ctx->strides;
-  shape_arr->resize(ndim);
-  stride_arr->resize(ndim, 1);
-  for (int i = 0; i < ndim; i++) {
-    (*shape_arr)[i] = tensor->dim_size(i);
-  }
-  for (int i = ndim - 2; i >= 0; --i) {
-    (*stride_arr)[i] = (*shape_arr)[i + 1] * (*stride_arr)[i + 1];
-  }
-
-  dlm_tensor->dl_tensor.shape = &(*shape_arr)[0];
-  // There are two ways to represent compact row-major data
-  // 1) nullptr indicates tensor is compact and row-majored.
-  // 2) fill in the strides array as the real case for compact row-major data
-  // Here we choose option 2, since some framework didn't handle the strides
-  // argument properly
-  dlm_tensor->dl_tensor.strides = &(*stride_arr)[0];
-  dlm_tensor->dl_tensor.byte_offset =
-      0;  // TF doesn't handle the strides and byte_offsets here
-  return &tf_dlm_tensor_ctx->tensor;
-}
-
 absl::optional<std::string> DeviceNameFromDlContext(const DLContext& ctx,
                                                     TF_Status* status) {
   switch (ctx.device_type) {
@@ -181,6 +143,8 @@ absl::optional<std::string> DeviceNameFromDlContext(const DLContext& ctx,
       return absl::nullopt;
   };
 }
+
+// Convert DLPack data type to TF_DATATYPE
 TF_DataType TfDataTypeFormDlDataType(const DLDataType& dtype,
                                      TF_Status* status) {
   TF_DataType tf_dtype;
@@ -257,11 +221,16 @@ TF_DataType TfDataTypeFormDlDataType(const DLDataType& dtype,
   return tf_dtype;
 }
 
+// Wrapper function to match the function signature
+// TFE_NewTensorHandleFromDeviceMemory, calling the deleter of the
+// DLManagedTensor
 void DeallocatorWrapperFunc(void* data, size_t len, void* dlmt_vptr) {
   DLManagedTensor* dlmt = static_cast<DLManagedTensor*>(dlmt_vptr);
   dlmt->deleter(const_cast<DLManagedTensor*>(dlmt));
 }
 
+// Check whether the stride array matches the layout of compact, row-majored
+// data
 bool IsValidStrideCompactRowMajorData(int64_t* shape_arr, int64_t* stride_arr,
                                       int ndim) {
   if (ndim >= 1 && stride_arr[ndim - 1] != 1) {
@@ -284,8 +253,43 @@ void TFE_CallDLManagedTensorDeleter(void* dlm_ptr) {
 }
 
 void* TFE_HandleToDLPack(TFE_TensorHandle* h, TF_Status* status) {
-  DLManagedTensor* tfdlmtensor = TFEHandleToTfDlManagedTensorCtx(h, status);
+  const Tensor* tensor = GetTensorFromHandle(h, status);
-  return static_cast<void*>(tfdlmtensor);
+  TF_DataType data_type = static_cast<TF_DataType>(tensor->dtype());
+  TensorReference tensor_ref(*tensor);  // This will call buf_->Ref()
+
+  auto* tf_dlm_tensor_ctx = new TfDlManagedTensorCtx(tensor_ref);
+  tf_dlm_tensor_ctx->reference = tensor_ref;
+
+  DLManagedTensor* dlm_tensor = &tf_dlm_tensor_ctx->tensor;
+  dlm_tensor->manager_ctx = tf_dlm_tensor_ctx;
+  dlm_tensor->deleter = &DLManagedTensorDeleter;
+  dlm_tensor->dl_tensor.ctx = GetDLContext(h, status);
+  int ndim = tensor->dims();
+  dlm_tensor->dl_tensor.ndim = ndim;
+  dlm_tensor->dl_tensor.data = TFE_TensorHandleDevicePointer(h, status);
+  dlm_tensor->dl_tensor.dtype = GetDLDataType(data_type, status);
+
+  std::vector<int64_t>* shape_arr = &tf_dlm_tensor_ctx->shape;
+  std::vector<int64_t>* stride_arr = &tf_dlm_tensor_ctx->strides;
+  shape_arr->resize(ndim);
+  stride_arr->resize(ndim, 1);
+  for (int i = 0; i < ndim; i++) {
+    (*shape_arr)[i] = tensor->dim_size(i);
+  }
+  for (int i = ndim - 2; i >= 0; --i) {
+    (*stride_arr)[i] = (*shape_arr)[i + 1] * (*stride_arr)[i + 1];
+  }
+
+  dlm_tensor->dl_tensor.shape = &(*shape_arr)[0];
+  // There are two ways to represent compact row-major data
+  // 1) nullptr indicates tensor is compact and row-majored.
+  // 2) fill in the strides array as the real case for compact row-major data
+  // Here we choose option 2, since some framework didn't handle the strides
+  // argument properly
+  dlm_tensor->dl_tensor.strides = &(*stride_arr)[0];
+  dlm_tensor->dl_tensor.byte_offset =
+      0;  // TF doesn't handle the strides and byte_offsets here
+  return static_cast<void*>(dlm_tensor);
 }
 
 TFE_TensorHandle* TFE_HandleFromDLPack(void* dlm, TF_Status* status) {

tensorflow/c/eager/dlpack.h

@@ -13,7 +13,6 @@ See the License for the specific language governing permissions and
 limitations under the License.
 ==============================================================================*/
 
-
 #ifndef TENSORFLOW_C_DLPACK_H_
 #define TENSORFLOW_C_DLPACK_H_
 
@@ -22,14 +21,20 @@ limitations under the License.
 
 namespace tensorflow {
 
+// PyCapsule name for DLPack Tensor
 const char* const kDlTensorCapsuleName = "dltensor";
 
-TF_CAPI_EXPORT extern void* TFE_HandleToDLPack(TFE_TensorHandle* h, TF_Status* status);
+// Convert eager tensor handle to DLPack (DLManagedTensor*), and return the
+// void* for further PyCapsule construction
+TF_CAPI_EXPORT extern void* TFE_HandleToDLPack(TFE_TensorHandle* h,
+                                               TF_Status* status);
 
-TF_CAPI_EXPORT extern TFE_TensorHandle* TFE_HandleFromDLPack(void* dlm, TF_Status* status);
+// Convert DLPack (DLManagedTensor*) to eager tensor handle
+TF_CAPI_EXPORT extern TFE_TensorHandle* TFE_HandleFromDLPack(void* dlm,
+                                                             TF_Status* status);
 
+// Call the destructor of DLManagedTensor, used in the destructor of PyCapsule
 TF_CAPI_EXPORT extern void TFE_CallDLManagedTensorDeleter(void* dlm_ptr);
 }  // namespace tensorflow
 
-
 #endif  // TENSORFLOW_C_DLPACK_H_