aboutsummaryrefslogtreecommitdiff
path: root/src/core/CL/kernels/CLDerivativeKernel.cpp
diff options
context:
space:
mode:
Diffstat (limited to 'src/core/CL/kernels/CLDerivativeKernel.cpp')
-rw-r--r--src/core/CL/kernels/CLDerivativeKernel.cpp16
1 files changed, 14 insertions, 2 deletions
diff --git a/src/core/CL/kernels/CLDerivativeKernel.cpp b/src/core/CL/kernels/CLDerivativeKernel.cpp
index af7df14359..f55a4cc464 100644
--- a/src/core/CL/kernels/CLDerivativeKernel.cpp
+++ b/src/core/CL/kernels/CLDerivativeKernel.cpp
@@ -1,5 +1,5 @@
/*
- * Copyright (c) 2016-2018 ARM Limited.
+ * Copyright (c) 2016-2019 ARM Limited.
*
* SPDX-License-Identifier: MIT
*
@@ -83,7 +83,8 @@ void CLDerivativeKernel::configure(const ICLTensor *input, ICLTensor *output_x,
}
// Create kernel
- _kernel = static_cast<cl::Kernel>(CLKernelLibrary::get().create_kernel("derivative", build_opts));
+ const std::string kernel_name = std::string("derivative");
+ _kernel = static_cast<cl::Kernel>(CLKernelLibrary::get().create_kernel(kernel_name, build_opts));
// Configure kernel window
constexpr unsigned int num_elems_processed_per_iteration = 16;
@@ -118,6 +119,17 @@ void CLDerivativeKernel::configure(const ICLTensor *input, ICLTensor *output_x,
output_y_access.set_valid_region(win, input->info()->valid_region(), border_undefined, border_size());
ICLKernel::configure_internal(win);
+
+ // Set config_id for enabling LWS tuning
+ _config_id = kernel_name;
+ _config_id += "_";
+ _config_id += lower_string(string_from_data_type(input->info()->data_type()));
+ _config_id += "_";
+ _config_id += support::cpp11::to_string(input->info()->dimension(0));
+ _config_id += "_";
+ _config_id += support::cpp11::to_string(input->info()->dimension(1));
+ _config_id += "_";
+ _config_id += support::cpp11::to_string(border_undefined);
}
void CLDerivativeKernel::run(const Window &window, cl::CommandQueue &queue)