Use tensorflow::se instead of perftools::gputools for StreamExecutor.

PiperOrigin-RevId: 194010749

1 files changed, 5 insertions, 6 deletions

diff --git a/tensorflow/contrib/rnn/kernels/blas_gemm.cc b/tensorflow/contrib/rnn/kernels/blas_gemm.cc
index 03006dab32..45d22b739b 100644
--- a/tensorflow/contrib/rnn/kernels/blas_gemm.cc
+++ b/tensorflow/contrib/rnn/kernels/blas_gemm.cc
@@ -26,9 +26,9 @@ namespace tensorflow {
 #if GOOGLE_CUDA
 namespace {
 template <typename T>
-perftools::gputools::DeviceMemory<T> AsDeviceMemory(const T* cuda_memory) {
-  perftools::gputools::DeviceMemoryBase wrapped(const_cast<T*>(cuda_memory));
-  perftools::gputools::DeviceMemory<T> typed(wrapped);
+se::DeviceMemory<T> AsDeviceMemory(const T* cuda_memory) {
+  se::DeviceMemoryBase wrapped(const_cast<T*>(cuda_memory));
+  se::DeviceMemory<T> typed(wrapped);
   return typed;
 }
 }  // namespace
@@ -41,9 +41,8 @@ void TensorCuBlasGemm<T>::operator()(OpKernelContext* ctx, bool transa,
                                      T alpha, const T* a, int lda, const T* b,
                                      int ldb, T beta, T* c, int ldc) {
 #if GOOGLE_CUDA
-  perftools::gputools::blas::Transpose trans[] = {
-      perftools::gputools::blas::Transpose::kNoTranspose,
-      perftools::gputools::blas::Transpose::kTranspose};
+  se::blas::Transpose trans[] = {se::blas::Transpose::kNoTranspose,
+                                 se::blas::Transpose::kTranspose};
 
   auto a_ptr = AsDeviceMemory(a);
   auto b_ptr = AsDeviceMemory(b);