diff --git a/tensorflow/compiler/tf2tensorrt/convert/convert_nodes.cc b/tensorflow/compiler/tf2tensorrt/convert/convert_nodes.cc
index f568b947959..21e25fde582 100644
--- a/tensorflow/compiler/tf2tensorrt/convert/convert_nodes.cc
+++ b/tensorflow/compiler/tf2tensorrt/convert/convert_nodes.cc
@@ -790,6 +790,14 @@ class TRT_TensorOrWeights::SimpleITensor : public nvinfer1::ITensor {
   float getDynamicRangeMax() const override { return 0.f; }
 #endif
 
+#if IS_TRT_VERSION_GE(6, 0, 0, 0)
+  void setAllowedFormats(nvinfer1::TensorFormats formats) override {}
+
+  nvinfer1::TensorFormats getAllowedFormats() const override { return 1; }
+
+  bool isShape() const override { return false; }
+#endif
+
  private:
   nvinfer1::DataType trt_dtype_;
   nvinfer1::Dims trt_dims_;
diff --git a/third_party/gpus/find_cuda_config.py b/third_party/gpus/find_cuda_config.py
index 576456e3646..f6c86ad702e 100644
--- a/third_party/gpus/find_cuda_config.py
+++ b/third_party/gpus/find_cuda_config.py
@@ -388,6 +388,12 @@ def _find_tensorrt_config(base_paths, required_version):
   header_path, header_version = _find_header(base_paths, "NvInfer.h",
                                              required_version,
                                              get_header_version)
+
+  if ".." in header_version:
+    header_path, header_version = _find_header(base_paths, "NvInferRTSafe.h",
+                                               required_version,
+                                               get_header_version)
+
   tensorrt_version = header_version.split(".")[0]
 
   library_path = _find_library(base_paths, "nvinfer", tensorrt_version)