aboutsummaryrefslogtreecommitdiff
path: root/src/core/gpu
diff options
context:
space:
mode:
Diffstat (limited to 'src/core/gpu')
-rw-r--r--src/core/gpu/cl/kernels/ClDirectConv2dKernel.cpp (renamed from src/core/gpu/cl/kernels/ClDirectConvolutionKernel.cpp)14
-rw-r--r--src/core/gpu/cl/kernels/ClDirectConv2dKernel.h (renamed from src/core/gpu/cl/kernels/ClDirectConvolutionKernel.h)26
-rw-r--r--src/core/gpu/cl/kernels/ClPool2dKernel.cpp (renamed from src/core/gpu/cl/kernels/ClPoolingKernel.cpp)12
-rw-r--r--src/core/gpu/cl/kernels/ClPool2dKernel.h (renamed from src/core/gpu/cl/kernels/ClPoolingKernel.h)19
4 files changed, 29 insertions, 42 deletions
diff --git a/src/core/gpu/cl/kernels/ClDirectConvolutionKernel.cpp b/src/core/gpu/cl/kernels/ClDirectConv2dKernel.cpp
index 0a5101f564..2c9a4f301b 100644
--- a/src/core/gpu/cl/kernels/ClDirectConvolutionKernel.cpp
+++ b/src/core/gpu/cl/kernels/ClDirectConv2dKernel.cpp
@@ -21,7 +21,7 @@
* OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
* SOFTWARE.
*/
-#include "src/core/gpu/cl/kernels/ClDirectConvolutionKernel.h"
+#include "src/core/gpu/cl/kernels/ClDirectConv2dKernel.h"
#include "arm_compute/core/CL/CLHelpers.h"
#include "arm_compute/core/CL/CLKernelLibrary.h"
@@ -369,13 +369,13 @@ bool export_to_cl_image_support(ITensorInfo *tensor, GPUTarget gpu_target, DataL
} // namespace
-BorderSize ClDirectConvolutionKernel::border_size() const
+BorderSize ClDirectConv2dKernel::border_size() const
{
return _border_size;
}
-void ClDirectConvolutionKernel::configure(const CLCompileContext &compile_context, ITensorInfo *src, ITensorInfo *weights, ITensorInfo *biases, ITensorInfo *dst,
- const PadStrideInfo &conv_info)
+void ClDirectConv2dKernel::configure(const CLCompileContext &compile_context, ITensorInfo *src, ITensorInfo *weights, ITensorInfo *biases, ITensorInfo *dst,
+ const PadStrideInfo &conv_info)
{
ARM_COMPUTE_ERROR_ON_NULLPTR(src, weights, dst);
@@ -564,8 +564,8 @@ void ClDirectConvolutionKernel::configure(const CLCompileContext &compile_contex
_config_id += lower_string(string_from_data_layout(_data_layout));
}
-Status ClDirectConvolutionKernel::validate(const ITensorInfo *src, const ITensorInfo *weights, const ITensorInfo *biases, const ITensorInfo *dst, const PadStrideInfo &conv_info,
- const GPUTarget target)
+Status ClDirectConv2dKernel::validate(const ITensorInfo *src, const ITensorInfo *weights, const ITensorInfo *biases, const ITensorInfo *dst, const PadStrideInfo &conv_info,
+ const GPUTarget target)
{
ARM_COMPUTE_RETURN_ON_ERROR(validate_arguments(src, weights, biases, dst, conv_info));
ARM_COMPUTE_RETURN_ON_ERROR(validate_and_configure_window(src->clone().get(), weights->clone().get(), dst->clone().get(), conv_info, target).first);
@@ -573,7 +573,7 @@ Status ClDirectConvolutionKernel::validate(const ITensorInfo *src, const ITensor
return Status{};
}
-void ClDirectConvolutionKernel::run_op(ITensorPack &tensors, const Window &window, cl::CommandQueue &queue)
+void ClDirectConv2dKernel::run_op(ITensorPack &tensors, const Window &window, cl::CommandQueue &queue)
{
ARM_COMPUTE_ERROR_ON_UNCONFIGURED_KERNEL(this);
ARM_COMPUTE_ERROR_ON_INVALID_SUBWINDOW(IKernel::window(), window);
diff --git a/src/core/gpu/cl/kernels/ClDirectConvolutionKernel.h b/src/core/gpu/cl/kernels/ClDirectConv2dKernel.h
index 384b561003..ec76624e5c 100644
--- a/src/core/gpu/cl/kernels/ClDirectConvolutionKernel.h
+++ b/src/core/gpu/cl/kernels/ClDirectConv2dKernel.h
@@ -21,8 +21,8 @@
* OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
* SOFTWARE.
*/
-#ifndef ARM_COMPUTE_CL_DIRECT_CONVOLUTION_KERNEL_H
-#define ARM_COMPUTE_CL_DIRECT_CONVOLUTION_KERNEL_H
+#ifndef ARM_COMPUTE_CL_DIRECT_CONV2D_KERNEL_H
+#define ARM_COMPUTE_CL_DIRECT_CONV2D_KERNEL_H
#include "src/core/common/Macros.h"
#include "src/core/gpu/cl/ClCompileContext.h"
@@ -36,11 +36,11 @@ namespace kernels
{
/** Interface for the direct convolution kernel.
*/
-class ClDirectConvolutionKernel : public IClKernel
+class ClDirectConv2dKernel : public IClKernel
{
public:
- ClDirectConvolutionKernel() = default;
- ARM_COMPUTE_DISALLOW_COPY_ALLOW_MOVE(ClDirectConvolutionKernel);
+ ClDirectConv2dKernel() = default;
+ ARM_COMPUTE_DISALLOW_COPY_ALLOW_MOVE(ClDirectConv2dKernel);
/** Set the src, weights, biases and dst tensors info.
*
* @note: Due to set_valid_region(), thus src/weights/biases cannot be const. Need to change this once the set_valid_region() is removed.
@@ -64,19 +64,9 @@ public:
* @param[in] conv_info Contains padding and stride information described in @ref PadStrideInfo.
*/
void configure(const CLCompileContext &compile_context, ITensorInfo *src, ITensorInfo *weights, ITensorInfo *biases, ITensorInfo *dst, const PadStrideInfo &conv_info);
- /** Static function to check if given info will lead to a valid configuration of @ref ClDirectConvolutionKernel
+ /** Static function to check if given info will lead to a valid configuration
*
- * @param[in] src The src tensor info to convolve. 3 lower dimensions represent a single src [width, height, IFM],
- * while every optional dimension from 4 and above represent a batch of inputs. Data types supported: QASYMM8_SIGNED/QASYMM8/F16/F32.
- * @param[in] weights Weights tensor info. Weights are 4D tensor with dimensions [kernel_x, kernel_y, IFM, OFM].
- * The 3rd dimension must be the same as the src's volume 3rd dimension.
- * Data type supported:Same as @p src.
- * @param[in] biases Biases tensor info. Biases are 1D tensor with dimension [OFM].
- * Data type supported: Should match @p src data type, except for src of QASYMM8 and QASYMM8_SIGNED type where biases should be of S32 type.
- * @param[in] dst Output tensor info.
- * The 3rd dimensions must be equal to the 4th dimension of the @p kernels tensor. Data types supported: Same as @p src.
- * @param[in] conv_info Contains padding and stride information described in @ref PadStrideInfo.
- * @param[in] target Target GPU architecture.
+ * Similar to ClDirectConv2dKernel::configure()
*
* @return a status
*/
@@ -94,4 +84,4 @@ public:
} // namespace kernels
} // namespace opencl
} // namespace arm_compute
-#endif /*ARM_COMPUTE_CL_DIRECT_CONVOLUTION_KERNEL_H */
+#endif /*ARM_COMPUTE_CL_DIRECT_CONV2D_KERNEL_H */
diff --git a/src/core/gpu/cl/kernels/ClPoolingKernel.cpp b/src/core/gpu/cl/kernels/ClPool2dKernel.cpp
index 08a3ce3784..0e15bffd14 100644
--- a/src/core/gpu/cl/kernels/ClPoolingKernel.cpp
+++ b/src/core/gpu/cl/kernels/ClPool2dKernel.cpp
@@ -21,7 +21,7 @@
* OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
* SOFTWARE.
*/
-#include "src/core/gpu/cl/kernels/ClPoolingKernel.h"
+#include "src/core/gpu/cl/kernels/ClPool2dKernel.h"
#include "arm_compute/core/CL/CLHelpers.h"
#include "arm_compute/core/CL/CLKernelLibrary.h"
@@ -202,17 +202,17 @@ std::tuple<Status, Window, ClPoolingConfig> validate_and_configure_window(ITenso
}
} // namespace
-ClPoolingKernel::ClPoolingKernel()
+ClPool2dKernel::ClPool2dKernel()
: _pool_info(), _data_layout(DataLayout::UNKNOWN), _border_size(0), _num_elems_processed_per_iteration(1)
{
}
-BorderSize ClPoolingKernel::border_size() const
+BorderSize ClPool2dKernel::border_size() const
{
return _border_size;
}
-void ClPoolingKernel::configure(const ClCompileContext &compile_context, ITensorInfo *src, ITensorInfo *dst, const PoolingLayerInfo &pool_info, ITensorInfo *indices)
+void ClPool2dKernel::configure(const ClCompileContext &compile_context, ITensorInfo *src, ITensorInfo *dst, const PoolingLayerInfo &pool_info, ITensorInfo *indices)
{
ARM_COMPUTE_ERROR_ON_NULLPTR(src, dst);
@@ -422,7 +422,7 @@ void ClPoolingKernel::configure(const ClCompileContext &compile_context, ITensor
ARM_COMPUTE_ERROR_ON(src->data_layout() == DataLayout::NHWC && has_padding_changed(padding_info));
}
-Status ClPoolingKernel::validate(const ITensorInfo *src, const ITensorInfo *dst, const PoolingLayerInfo &pool_info, const ITensorInfo *indices)
+Status ClPool2dKernel::validate(const ITensorInfo *src, const ITensorInfo *dst, const PoolingLayerInfo &pool_info, const ITensorInfo *indices)
{
ARM_COMPUTE_RETURN_ON_ERROR(validate_arguments(src, dst, pool_info, indices));
ARM_COMPUTE_RETURN_ON_ERROR(std::get<0>(validate_and_configure_window(src->clone().get(), dst->clone().get(), pool_info)));
@@ -430,7 +430,7 @@ Status ClPoolingKernel::validate(const ITensorInfo *src, const ITensorInfo *dst,
return Status{};
}
-void ClPoolingKernel::run_op(ITensorPack &tensors, const Window &window, cl::CommandQueue &queue)
+void ClPool2dKernel::run_op(ITensorPack &tensors, const Window &window, cl::CommandQueue &queue)
{
ARM_COMPUTE_ERROR_ON_UNCONFIGURED_KERNEL(this);
ARM_COMPUTE_ERROR_ON_INVALID_SUBWINDOW(ICLKernel::window(), window);
diff --git a/src/core/gpu/cl/kernels/ClPoolingKernel.h b/src/core/gpu/cl/kernels/ClPool2dKernel.h
index c1ce859e2c..8ecb8eb7b7 100644
--- a/src/core/gpu/cl/kernels/ClPoolingKernel.h
+++ b/src/core/gpu/cl/kernels/ClPool2dKernel.h
@@ -21,8 +21,8 @@
* OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
* SOFTWARE.
*/
-#ifndef ARM_COMPUTE_CL_POOLING_KERNEL_H
-#define ARM_COMPUTE_CL_POOLING_KERNEL_H
+#ifndef ARM_COMPUTE_CL_POOL2D_KERNEL_H
+#define ARM_COMPUTE_CL_POOL2D_KERNEL_H
#include "src/core/common/Macros.h"
#include "src/core/gpu/cl/ClCompileContext.h"
@@ -35,12 +35,12 @@ namespace opencl
namespace kernels
{
/** Interface for the pooling layer kernel */
-class ClPoolingKernel : public IClKernel
+class ClPool2dKernel : public IClKernel
{
public:
/** Default constructor */
- ClPoolingKernel();
- ARM_COMPUTE_DISALLOW_COPY_ALLOW_MOVE(ClPoolingKernel);
+ ClPool2dKernel();
+ ARM_COMPUTE_DISALLOW_COPY_ALLOW_MOVE(ClPool2dKernel);
/** Configure kernel for a given list of arguments
*
@@ -52,12 +52,9 @@ public:
* @param[out] indices (optional) The indices of the maximal values. Data type supported: U32.
*/
void configure(const ClCompileContext &compile_context, ITensorInfo *src, ITensorInfo *dst, const PoolingLayerInfo &pool_info, ITensorInfo *indices = nullptr);
- /** Static function to check if given info will lead to a valid configuration of @ref ClPoolingKernel
+ /** Static function to check if given info will lead to a valid configuration
*
- * @param[in] src Source tensor info. Data types supported: QASYMM8/QASYMM8_SIGNED/F16/F32.
- * @param[in] dst Destination tensor info. Data types supported: same as @p src.
- * @param[in] pool_info Contains pooling operation information described in @ref PoolingLayerInfo.
- * @param[in] indices (optional) The indices of the maximal values. Data type supported: U32.
+ * Similar to ClPool2dKernel::configure()
*
* @return a status
*/
@@ -76,4 +73,4 @@ public:
} // namespace kernels
} // namespace opencl
} // namespace arm_compute
-#endif /*ARM_COMPUTE_CL_POOLING_KERNEL_H */
+#endif /* ARM_COMPUTE_CL_POOL2D_KERNEL_H */
generated by cgit v1.2.1 (git 2.23.0) at 2024-06-11 22:29:46 +0000