COMPMID-855 - Optimizing im2col on OpenCL (DCHW)
Introduced optimizations for 1x1, 3x3, 5x5 and 11x11
Change-Id: Ibb7f7a9fbec01a7684746ed8513634078126e452
Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/118107
Tested-by: Jenkins <bsgcomp@arm.com>
Reviewed-by: Michalis Spyrou <michalis.spyrou@arm.com>
diff --git a/arm_compute/core/CL/kernels/CLIm2ColKernel.h b/arm_compute/core/CL/kernels/CLIm2ColKernel.h
index 88de1ba..e38e7e8 100644
--- a/arm_compute/core/CL/kernels/CLIm2ColKernel.h
+++ b/arm_compute/core/CL/kernels/CLIm2ColKernel.h
@@ -1,5 +1,5 @@
/*
- * Copyright (c) 2017 ARM Limited.
+ * Copyright (c) 2017-2018 ARM Limited.
*
* SPDX-License-Identifier: MIT
*
@@ -25,11 +25,11 @@
#define __ARM_COMPUTE_CLIM2COLKERNEL_H__
#include "arm_compute/core/CL/ICLKernel.h"
+#include "arm_compute/core/Size2D.h"
namespace arm_compute
{
class ICLTensor;
-class Size2D;
/** Interface for the im2col reshape kernel.
*
@@ -117,6 +117,7 @@
std::pair<unsigned int, unsigned int> _convolved_dims;
unsigned int _num_elems_processed_per_iteration;
Im2ColFunction _run_func;
+ Size2D _kernel_dims;
};
} // namespace arm_compute
#endif /*__ARM_COMPUTE_CLIM2COLKERNEL_H__ */
diff --git a/src/core/CL/CLKernelLibrary.cpp b/src/core/CL/CLKernelLibrary.cpp
index ae35538..d7ae6e3 100644
--- a/src/core/CL/CLKernelLibrary.cpp
+++ b/src/core/CL/CLKernelLibrary.cpp
@@ -170,7 +170,7 @@
{ "combine_gradients_L2", "canny.cl" },
{ "concatenate_depth", "concatenate.cl" },
{ "convolution_rectangle", "convolution_rectangle.cl" },
- { "col2im", "convolution_layer.cl" },
+ { "col2im", "col2im.cl" },
{ "convolution3x3_static", "convolution3x3.cl" },
{ "convolution5x5_static", "convolution5x5.cl" },
{ "convolution7x7_static", "convolution7x7.cl" },
@@ -248,10 +248,13 @@
{ "hog_detector", "hog.cl" },
{ "hog_orientation_binning", "hog.cl" },
{ "hysteresis", "canny.cl" },
- { "im2col_generic", "convolution_layer.cl" },
- { "im2col_generic_padx0_pady0", "convolution_layer.cl" },
- { "im2col_kernel3x3_padx0_pady0", "convolution_layer.cl" },
- { "im2col_reduced", "convolution_layer.cl" },
+ { "im2col1x1_stridex1_dchw", "im2col.cl" },
+ { "im2col3x3_dchw", "im2col.cl" },
+ { "im2col5x5_dchw", "im2col.cl" },
+ { "im2col11x11_padx0_pady0_dchw", "im2col.cl" },
+ { "im2col_generic_dchw", "im2col.cl" },
+ { "im2col_generic_padx0_pady0_dchw", "im2col.cl" },
+ { "im2col_reduced_dchw", "im2col.cl" },
{ "init_level", "optical_flow_pyramid_lk.cl" },
{ "init_level_max", "optical_flow_pyramid_lk.cl" },
{ "init_level_max_initial_estimate", "optical_flow_pyramid_lk.cl" },
@@ -390,6 +393,10 @@
#include "./cl_kernels/channel_extract.clembed"
},
{
+ "col2im.cl",
+#include "./cl_kernels/col2im.clembed"
+ },
+ {
"concatenate.cl",
#include "./cl_kernels/concatenate.clembed"
},
@@ -522,6 +529,10 @@
#include "./cl_kernels/hog.clembed"
},
{
+ "im2col.cl",
+#include "./cl_kernels/im2col.clembed"
+ },
+ {
"integral_image.cl",
#include "./cl_kernels/integral_image.clembed"
},
diff --git a/src/core/CL/cl_kernels/col2im.cl b/src/core/CL/cl_kernels/col2im.cl
new file mode 100644
index 0000000..58fb80a
--- /dev/null
+++ b/src/core/CL/cl_kernels/col2im.cl
@@ -0,0 +1,68 @@
+/*
+ * Copyright (c) 2017-2018 ARM Limited.
+ *
+ * SPDX-License-Identifier: MIT
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to
+ * deal in the Software without restriction, including without limitation the
+ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
+ * sell copies of the Software, and to permit persons to whom the Software is
+ * furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in all
+ * copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+ * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+#include "helpers.h"
+
+#if defined(FIXED_POINT_POSITION)
+#include "fixed_point.h"
+#endif // FIXED_POINT_POSITION
+
+#if defined(DATA_TYPE) && defined(WIDTH_OUTPUT)
+/** This kernel performs a reshaping of the output of the convolution layer.
+ *
+ * @note The data type must be passed at compile time using -DDATA_TYPE: e.g. -DDATA_TYPE=float
+ * @note The width of the output tensor must be passed at compile time using -DWIDTH_OUTPUT: e.g. -DWIDTH_OUTPUT=320
+ *
+ * @param[in] src_ptr Pointer to the source tensor. Supported data types: QS8/QASYMM8/QS16/F16/F32
+ * @param[in] src_stride_x Stride of the source tensor in X dimension (in bytes)
+ * @param[in] src_step_x src_stride_x * number of elements along X processed per workitem(in bytes)
+ * @param[in] src_stride_y Stride of the source tensor in Y dimension (in bytes)
+ * @param[in] src_step_y src_stride_y * number of elements along Y processed per workitem(in bytes)
+ * @param[in] src_stride_z Stride of the source tensor in Z dimension (in bytes)
+ * @param[in] src_step_z src_stride_z * number of elements along Z processed per workitem(in bytes)
+ * @param[in] src_offset_first_element_in_bytes The offset of the first element in the source tensor
+ * @param[out] dst_ptr Pointer to the destination tensor. Supported data types: same as @p src_ptr
+ * @param[in] dst_stride_x Stride of the destination tensor in X dimension (in bytes)
+ * @param[in] dst_step_x dst_stride_x * number of elements along X processed per workitem(in bytes)
+ * @param[in] dst_stride_y Stride of the destination tensor in Y dimension (in bytes)
+ * @param[in] dst_step_y dst_stride_y * number of elements along Y processed per workitem(in bytes)
+ * @param[in] dst_stride_z Stride of the destination tensor in Z dimension (in bytes)
+ * @param[in] dst_step_z dst_stride_z * number of elements along Z processed per workitem(in bytes)
+ * @param[in] dst_offset_first_element_in_bytes The offset of the first element in the destination tensor
+ * @param[in] dst_stride_w Stride of the destination tensor in W dimension (in bytes)
+ */
+__kernel void col2im(
+ TENSOR3D_DECLARATION(src),
+ TENSOR3D_DECLARATION(dst),
+ uint dst_stride_w)
+{
+ Tensor3D src = CONVERT_TO_TENSOR3D_STRUCT(src);
+ Tensor3D dst = CONVERT_TO_TENSOR3D_STRUCT_NO_STEP(dst);
+
+ // Compute output offset
+ int idx = get_global_id(0) * dst.stride_z + (get_global_id(1) / WIDTH_OUTPUT) * dst_stride_y + (get_global_id(1) % WIDTH_OUTPUT) * dst_stride_x + get_global_id(2) * dst_stride_w;
+
+ // Store value
+ *((__global DATA_TYPE *)(dst.ptr + idx)) = *((__global DATA_TYPE *)(src.ptr));
+}
+#endif // defined(DATA_TYPE) && defined(WIDTH_OUTPUT)
\ No newline at end of file
diff --git a/src/core/CL/cl_kernels/convolution_layer.cl b/src/core/CL/cl_kernels/convolution_layer.cl
index 77b9b64..f8e0c27 100644
--- a/src/core/CL/cl_kernels/convolution_layer.cl
+++ b/src/core/CL/cl_kernels/convolution_layer.cl
@@ -1,5 +1,5 @@
/*
- * Copyright (c) 2017 ARM Limited.
+ * Copyright (c) 2017-2018 ARM Limited.
*
* SPDX-License-Identifier: MIT
*
@@ -27,6 +27,7 @@
#include "fixed_point.h"
#endif // FIXED_POINT_POSITION
+#if defined(DATA_TYPE)
/** This kernel reshapes the tensor's low three dimensions to single column
*
* @note Datatype should be given as a preprocessor argument using -DDATA_TYPE=type. e.g. -DDATA_TYPE=short
@@ -96,319 +97,4 @@
}
}
}
-
-#if defined(CONVOLVED_WIDTH) && defined(STRIDE_X) && defined(STRIDE_Y) && defined(PAD_LEFT) && defined(PAD_TOP) && defined(PAD_RIGHT) && defined(PAD_BOTTOM) && defined(KERNEL_WIDTH) && defined(KERNEL_HEIGHT) && defined(KERNEL_DEPTH) && defined(SRC_WIDTH) && defined(SRC_HEIGHT) && defined(PAD_VALUE)
-/** This kernel performs a reshaping of the input tensor to a tensor used to perform convolution using GEMM.
- *
- * @note The data type must be passed at compile time using -DDATA_TYPE: e.g. -DDATA_TYPE=float
- * @note The value to use for the paddings must be passed at compile time using -DPAD_VALUE: e.g. -DPAD_VALUE=0
- * @note In case biases will be added to the convolution -DHAS_BIAS has to be passed to append the final matrix with 1 in each row.
- *
- * @param[in] src_ptr Pointer to the source tensor. Supported data types: QS8/QASYMM8/QS16/F16/F32
- * @param[in] src_stride_x Stride of the source tensor in X dimension (in bytes)
- * @param[in] src_step_x src_stride_x * number of elements along X processed per workitem(in bytes)
- * @param[in] src_stride_y Stride of the source tensor in Y dimension (in bytes)
- * @param[in] src_step_y src_stride_y * number of elements along Y processed per workitem(in bytes)
- * @param[in] src_stride_z Stride of the source tensor in Z dimension (in bytes)
- * @param[in] src_step_z src_stride_z * number of elements along Z processed per workitem(in bytes)
- * @param[in] src_offset_first_element_in_bytes The offset of the first element in the source tensor
- * @param[out] dst_ptr Pointer to the destination tensor. Supported data types: same as @p src_ptr
- * @param[in] dst_stride_x Stride of the destination tensor in X dimension (in bytes)
- * @param[in] dst_step_x dst_stride_x * number of elements along X processed per workitem(in bytes)
- * @param[in] dst_stride_y Stride of the destination tensor in Y dimension (in bytes)
- * @param[in] dst_step_y dst_stride_y * number of elements along Y processed per workitem(in bytes)
- * @param[in] dst_offset_first_element_in_bytes The offset of the first element in the destination tensor
- * @param[in] src_stride_w Stride of the source tensor in W dimension (in bytes).
- * @param[in] dst_stride_w Stride of the destination tensor in W dimension (in bytes).
- */
-__kernel void im2col_generic(
- TENSOR3D_DECLARATION(src),
- IMAGE_DECLARATION(dst),
- uint src_stride_w,
- uint dst_stride_w)
-{
- const int xc = get_global_id(0); // x coordinate in the convolved tensor
- const int yc = get_global_id(1); // y coordinate in the convolved tensor
- const int ch = get_global_id(2) % KERNEL_DEPTH; // input feature map
- const int batch = get_global_id(2) / KERNEL_DEPTH; // batch size
-
- // Calculate input indices
- const int xi = xc * STRIDE_X - PAD_LEFT;
- const int yi = yc * STRIDE_Y - PAD_TOP;
-
- // Calculate output indices
- const int xo = ch * KERNEL_WIDTH * KERNEL_HEIGHT;
- const int yo = xc + yc * CONVOLVED_WIDTH; // Index of the convolution
-
- __global uchar *input_ptr = src_ptr + src_offset_first_element_in_bytes + ch * src_stride_z + batch * src_stride_w;
- __global DATA_TYPE *output_ptr = ((__global DATA_TYPE *)(dst_ptr + dst_offset_first_element_in_bytes + yo * dst_stride_y + batch * dst_stride_w)) + xo;
-
- // Linearize convolution elements
- for(int y = yi, y_e = yi + KERNEL_HEIGHT; y < y_e; ++y)
- {
- for(int x = xi, x_e = xi + KERNEL_WIDTH; x < x_e; ++x, ++output_ptr)
- {
-#if PAD_LEFT == 0 && PAD_TOP == 0 && PAD_RIGHT == 0 && PAD_BOTTOM == 0
- *output_ptr = *((__global DATA_TYPE *)(input_ptr + x * src_stride_x + y * src_stride_y));
-#else // PAD_LEFT == 0 && PAD_TOP == 0 && PAD_RIGHT == 0 && PAD_BOTTOM == 0
- if(x < 0 || x >= SRC_WIDTH || y < 0 || y >= SRC_HEIGHT)
- {
- *output_ptr = PAD_VALUE;
- }
- else
- {
- *output_ptr = *((__global DATA_TYPE *)(input_ptr + x * src_stride_x + y * src_stride_y));
- }
-#endif // PAD_LEFT == 0 && PAD_TOP == 0 && PAD_RIGHT == 0 && PAD_BOTTOM == 0
- }
- }
-
-#ifdef HAS_BIAS
- if(ch == (KERNEL_DEPTH - 1))
- {
-#ifdef FIXED_POINT_POSITION
- *output_ptr = (DATA_TYPE)(1 << FIXED_POINT_POSITION);
-#else // FIXED_POINT_POSITION
- *output_ptr = 1.0f;
-#endif // FIXED_POINT_POSITION
- }
-#endif // HAS_BIAS
-}
-
-/** This kernel performs a reshaping of the input tensor to a tensor used to perform convolution using GEMM when the kernel size is 3x3 and pad_x = pad_y = 0
- *
- * @note The data type must be passed at compile time using -DDATA_TYPE: e.g. -DDATA_TYPE=float
- * @note In case biases will be added to the convolution -DHAS_BIAS has to be passed to append the final matrix with 1 in each row.
- *
- * @param[in] src_ptr Pointer to the source tensor. Supported data types: QS8/QASYMM8/QS16/F16/F32
- * @param[in] src_stride_x Stride of the source tensor in X dimension (in bytes)
- * @param[in] src_step_x src_stride_x * number of elements along X processed per workitem(in bytes)
- * @param[in] src_stride_y Stride of the source tensor in Y dimension (in bytes)
- * @param[in] src_step_y src_stride_y * number of elements along Y processed per workitem(in bytes)
- * @param[in] src_stride_z Stride of the source tensor in Z dimension (in bytes)
- * @param[in] src_step_z src_stride_z * number of elements along Z processed per workitem(in bytes)
- * @param[in] src_offset_first_element_in_bytes The offset of the first element in the source tensor
- * @param[out] dst_ptr Pointer to the destination tensor. Supported data types: same as @p src_ptr
- * @param[in] dst_stride_x Stride of the destination tensor in X dimension (in bytes)
- * @param[in] dst_step_x dst_stride_x * number of elements along X processed per workitem(in bytes)
- * @param[in] dst_stride_y Stride of the destination tensor in Y dimension (in bytes)
- * @param[in] dst_step_y dst_stride_y * number of elements along Y processed per workitem(in bytes)
- * @param[in] dst_offset_first_element_in_bytes The offset of the first element in the destination tensor
- * @param[in] src_stride_w Stride of the source tensor in W dimension (in bytes).
- * @param[in] dst_stride_w Stride of the destination tensor in W dimension (in bytes).
- */
-__kernel void im2col_kernel3x3_padx0_pady0(
- TENSOR3D_DECLARATION(src),
- IMAGE_DECLARATION(dst),
- uint src_stride_w,
- uint dst_stride_w)
-{
- const int xc = get_global_id(0); // x coordinate in the convolved tensor
- const int yc = get_global_id(1); // y coordinate in the convolved tensor
- const int ch = get_global_id(2) % KERNEL_DEPTH; // input feature map
- const int batch = get_global_id(2) / KERNEL_DEPTH; // batch size
-
- // Calculate input indices
- const int xi = xc * STRIDE_X;
- const int yi = yc * STRIDE_Y;
-
- // Calculate output indices
- const int xo = ch * KERNEL_WIDTH * KERNEL_HEIGHT;
- const int yo = xc + yc * CONVOLVED_WIDTH; // Index of the convolution
-
- // Get input and output address
- __global uchar *input_ptr = src_ptr + src_offset_first_element_in_bytes + xi * src_stride_x + yi * src_stride_y + ch * src_stride_z + batch * src_stride_w;
-
- __global DATA_TYPE *output_ptr = (__global DATA_TYPE *)(dst_ptr + dst_offset_first_element_in_bytes + yo * dst_stride_y + batch * dst_stride_w) + xo;
-
- VEC_DATA_TYPE(DATA_TYPE, 3)
- row0 = vload3(0, (__global DATA_TYPE *)(input_ptr + 0 * src_stride_y));
- VEC_DATA_TYPE(DATA_TYPE, 3)
- row1 = vload3(0, (__global DATA_TYPE *)(input_ptr + 1 * src_stride_y));
- VEC_DATA_TYPE(DATA_TYPE, 3)
- row2 = vload3(0, (__global DATA_TYPE *)(input_ptr + 2 * src_stride_y));
-
- vstore8((VEC_DATA_TYPE(DATA_TYPE, 8))(row0.s012, row1.s012, row2.s01), 0, output_ptr);
- *(output_ptr + 8) = row2.s2;
-
-#ifdef HAS_BIAS
- if(ch == (KERNEL_DEPTH - 1))
- {
-#ifdef FIXED_POINT_POSITION
- *(output_ptr + 9) = (DATA_TYPE)(1 << FIXED_POINT_POSITION);
-#else // FIXED_POINT_POSITION
- *(output_ptr + 9) = 1.0f;
-#endif // FIXED_POINT_POSITION
- }
-#endif // HAS_BIAS
-}
-#endif //defined(CONVOLVED_WIDTH) && defined(STRIDE_X) && defined(STRIDE_Y) && defined(PAD_LEFT) && defined(PAD_TOP) && defined(PAD_RIGHT) && defined(PAD_BOTTOM) && defined(KERNEL_WIDTH) && defined(KERNEL_HEIGHT) && defined(KERNEL_DEPTH) && defined(SRC_WIDTH) && defined(SRC_HEIGHT)
-
-#if defined(WIDTH_OUTPUT)
-/** This kernel performs a reshaping of the output of the convolution layer.
- *
- * @note The data type must be passed at compile time using -DDATA_TYPE: e.g. -DDATA_TYPE=float
- *
- * @param[in] src_ptr Pointer to the source tensor. Supported data types: QS8/QASYMM8/QS16/F16/F32
- * @param[in] src_stride_x Stride of the source tensor in X dimension (in bytes)
- * @param[in] src_step_x src_stride_x * number of elements along X processed per workitem(in bytes)
- * @param[in] src_stride_y Stride of the source tensor in Y dimension (in bytes)
- * @param[in] src_step_y src_stride_y * number of elements along Y processed per workitem(in bytes)
- * @param[in] src_stride_z Stride of the source tensor in Z dimension (in bytes)
- * @param[in] src_step_z src_stride_z * number of elements along Z processed per workitem(in bytes)
- * @param[in] src_offset_first_element_in_bytes The offset of the first element in the source tensor
- * @param[out] dst_ptr Pointer to the destination tensor. Supported data types: same as @p src_ptr
- * @param[in] dst_stride_x Stride of the destination tensor in X dimension (in bytes)
- * @param[in] dst_step_x dst_stride_x * number of elements along X processed per workitem(in bytes)
- * @param[in] dst_stride_y Stride of the destination tensor in Y dimension (in bytes)
- * @param[in] dst_step_y dst_stride_y * number of elements along Y processed per workitem(in bytes)
- * @param[in] dst_stride_z Stride of the destination tensor in Z dimension (in bytes)
- * @param[in] dst_step_z dst_stride_z * number of elements along Z processed per workitem(in bytes)
- * @param[in] dst_offset_first_element_in_bytes The offset of the first element in the destination tensor
- * @param[in] dst_stride_w Stride of the destination tensor in W dimension (in bytes)
- */
-__kernel void col2im(
- TENSOR3D_DECLARATION(src),
- TENSOR3D_DECLARATION(dst),
- uint dst_stride_w)
-{
- Tensor3D src = CONVERT_TO_TENSOR3D_STRUCT(src);
- Tensor3D dst = CONVERT_TO_TENSOR3D_STRUCT_NO_STEP(dst);
-
- // Compute output offset
- int idx = get_global_id(0) * dst.stride_z + (get_global_id(1) / WIDTH_OUTPUT) * dst_stride_y + (get_global_id(1) % WIDTH_OUTPUT) * dst_stride_x + get_global_id(2) * dst_stride_w;
-
- // Store value
- *((__global DATA_TYPE *)(dst.ptr + idx)) = *((__global DATA_TYPE *)(src.ptr));
-}
-#endif // defined(WIDTH_OUTPUT)
-
-/** This kernel reshapes the tensor's low three dimensions to single row for GEMM operation
- *
- * @note Datatype should be given as a preprocessor argument using -DDATA_TYPE=type. e.g. -DDATA_TYPE=float
- * @note In case biases will be added in late stage, -DHAS_BIAS has to be passed to append the final matrix with 1 in each row.
- *
- * @param[in] src_ptr Pointer to the source tensor. Supported data types: QS8/QASYMM8/QS16/F16/F32
- * @param[in] src_stride_x Stride of the source tensor in X dimension (in bytes)
- * @param[in] src_step_x src_stride_x * number of elements along X processed per workitem(in bytes)
- * @param[in] src_stride_y Stride of the source tensor in Y dimension (in bytes)
- * @param[in] src_step_y src_stride_y * number of elements along Y processed per workitem(in bytes)
- * @param[in] src_stride_z Stride of the source tensor in Z dimension (in bytes)
- * @param[in] src_step_z src_stride_z * number of elements along Y processed per workitem(in bytes)
- * @param[in] src_offset_first_element_in_bytes The offset of the first element in the source tensor
- * @param[out] dst_ptr Pointer to the destination tensor. Same as @p src_ptr
- * @param[in] dst_stride_x Stride of the destination tensor in X dimension (in bytes)
- * @param[in] dst_step_x dst_stride_x * number of elements along X processed per workitem(in bytes)
- * @param[in] dst_offset_first_element_in_bytes The offset of the first element in the destination tensor
- * @param[in] width The width of the input tensor
- * @param[in] height The height of the input tensor
- */
-__kernel void im2col_reduced(
- TENSOR3D_DECLARATION(src),
- VECTOR_DECLARATION(dst),
- uint width, uint height)
-{
- Tensor3D src = CONVERT_TO_TENSOR3D_STRUCT(src);
-
- const uint image_size = width * height;
-
- __global uchar *tmp_out_ptr = dst_ptr + dst_offset_first_element_in_bytes + (get_global_id(0) + get_global_id(1) * width + get_global_id(2) * image_size) * dst_stride_x;
-
- *((__global DATA_TYPE *)tmp_out_ptr) = *((__global DATA_TYPE *)src.ptr);
-
-#ifdef HAS_BIAS
- // If it is the last thread in the 3 dimensional workgroup
- if(get_global_id(0) == (get_global_size(0) - 1) && get_global_id(1) == (get_global_size(1) - 1) && get_global_id(2) == (get_global_size(2) - 1))
- {
- tmp_out_ptr += dst_stride_x;
-#ifdef FIXED_POINT_POSITION
- *((__global DATA_TYPE *)tmp_out_ptr) = (DATA_TYPE)(1 << FIXED_POINT_POSITION);
-#else // FIXED_POINT_POSITION
- *((__global DATA_TYPE *)tmp_out_ptr) = (DATA_TYPE)1;
-#endif // FIXED_POINT_POSITION
- }
-#endif // HAS_BIAS
-}
-
-#if defined(CONVOLVED_WIDTH) && defined(STRIDE_X) && defined(STRIDE_Y) && defined(PAD_LEFT) && defined(PAD_TOP) && defined(PAD_RIGHT) && defined(PAD_BOTTOM) && defined(KERNEL_WIDTH) && defined(KERNEL_HEIGHT) && defined(KERNEL_DEPTH) && defined(SRC_WIDTH) && defined(SRC_HEIGHT) && defined(VECTOR_SIZE) && defined(WIDTH_MOD_VECTOR_SIZE)
-/** This kernel reshapes the input tensor to a tensor used to perform convolution using GEMM when
- * the kernel width is greater than 1 (except when the kernel size is 3x3) and pad_x == pad_y == 0.
- *
- * @note The data type must be passed at compile time using -DDATA_TYPE e.g. -DDATA_TYPE=float.
- * @note The vector size must be passed at compile time using -DVECTOR_SIZE e.g. -DVECTOR_SIZE=4.
- * @note The width modulo vector size must be passed at compile time using -DWIDTH_MOD_VECTOR_SIZE e.g. -DWIDTH_MOD_VECTOR_SIZE=3.
- * @note In case biases will be added to the convolution -DHAS_BIAS has to be passed to append the final matrix with 1 in each row.
- *
- * @param[in] src_ptr Pointer to the source tensor. Supported data types: QS8/QS16/F16/F32
- * @param[in] src_stride_x Stride of the source tensor in X dimension (in bytes)
- * @param[in] src_step_x src_stride_x * number of elements along X processed per workitem(in bytes)
- * @param[in] src_stride_y Stride of the source tensor in Y dimension (in bytes)
- * @param[in] src_step_y src_stride_y * number of elements along Y processed per workitem(in bytes)
- * @param[in] src_stride_z Stride of the source tensor in Z dimension (in bytes)
- * @param[in] src_step_z src_stride_z * number of elements along Z processed per workitem(in bytes)
- * @param[in] src_offset_first_element_in_bytes The offset of the first element in the source tensor
- * @param[out] dst_ptr Pointer to the destination tensor. Supported data types: same as @p src_ptr
- * @param[in] dst_stride_x Stride of the destination tensor in X dimension (in bytes)
- * @param[in] dst_step_x dst_stride_x * number of elements along X processed per workitem(in bytes)
- * @param[in] dst_stride_y Stride of the destination tensor in Y dimension (in bytes)
- * @param[in] dst_step_y dst_stride_y * number of elements along Y processed per workitem(in bytes)
- * @param[in] dst_offset_first_element_in_bytes The offset of the first element in the destination tensor
- * @param[in] src_stride_w Stride of the source tensor in W dimension (in bytes).
- * @param[in] dst_stride_w Stride of the destination tensor in W dimension (in bytes).
- */
-__kernel void im2col_generic_padx0_pady0(
- TENSOR3D_DECLARATION(src),
- IMAGE_DECLARATION(dst),
- uint src_stride_w,
- uint dst_stride_w)
-{
- const int xc = get_global_id(0); // x coordinate in the convolved tensor
- const int yc = get_global_id(1); // y coordinate in the convolved tensor
- const int ch = get_global_id(2) % KERNEL_DEPTH; // input feature map
- const int batch = get_global_id(2) / KERNEL_DEPTH; // batch size
-
- // Calculate input indices
- const int xi = xc * STRIDE_X;
- const int yi = yc * STRIDE_Y;
- // Calculate output indices
- const int xo = ch * KERNEL_WIDTH * KERNEL_HEIGHT;
- const int yo = xc + yc * CONVOLVED_WIDTH; // Index of the convolution
- __global uchar *input_ptr = src_ptr + src_offset_first_element_in_bytes + ch * src_stride_z + batch * src_stride_w;
- __global DATA_TYPE *output_ptr = ((__global DATA_TYPE *)(dst_ptr + dst_offset_first_element_in_bytes + yo * dst_stride_y + batch * dst_stride_w)) + xo;
- // Linearize convolution elements
- for(int y = yi, y_e = yi + KERNEL_HEIGHT; y < y_e; ++y)
- {
- int last_x = 0;
- for(int x = xi, x_e = xi + KERNEL_WIDTH; x + VECTOR_SIZE <= x_e; x += VECTOR_SIZE, output_ptr += VECTOR_SIZE)
- {
- VEC_DATA_TYPE(DATA_TYPE, VECTOR_SIZE)
- row = VLOAD(VECTOR_SIZE)(0, (__global DATA_TYPE *)(input_ptr + x * src_stride_x + y * src_stride_y));
- VSTORE(VECTOR_SIZE)
- (row, 0, output_ptr);
- last_x = x;
- }
- // Copy the remainder of the row by doing VLOAD(WIDTH_MOD_VECTOR_SIZE) and VSTORE(WIDTH_MOD_VECTOR_SIZE).
- // Note that x and output_ptr have already been incremented by VECTOR_SIZE by the loop just before exit.
-#if WIDTH_MOD_VECTOR_SIZE == 1
- *output_ptr = *((__global DATA_TYPE *)(input_ptr + (last_x + VECTOR_SIZE) * src_stride_x + y * src_stride_y));
-#elif WIDTH_MOD_VECTOR_SIZE > 1
- VEC_DATA_TYPE(DATA_TYPE, WIDTH_MOD_VECTOR_SIZE)
- row = VLOAD(WIDTH_MOD_VECTOR_SIZE)(0, (__global DATA_TYPE *)(input_ptr + (last_x + VECTOR_SIZE) * src_stride_x + y * src_stride_y));
- VSTORE(WIDTH_MOD_VECTOR_SIZE)
- (row, 0, output_ptr);
-#endif /* WIDTH_MOD_VECTOR_SIZE */
- output_ptr += WIDTH_MOD_VECTOR_SIZE;
- } /* End of loop over KERNEL_HEIGHT */
-
-#ifdef HAS_BIAS
- if(ch == (KERNEL_DEPTH - 1))
- {
-#ifdef FIXED_POINT_POSITION
- *output_ptr = (DATA_TYPE)(1 << FIXED_POINT_POSITION);
-#else // FIXED_POINT_POSITION
- *output_ptr = 1.0f;
-#endif // FIXED_POINT_POSITION
- }
-#endif // HAS_BIAS
-}
-#endif //defined(CONVOLVED_WIDTH) && defined(STRIDE_X) && defined(STRIDE_Y) && defined(PAD_LEFT) && defined(PAD_TOP) && defined(PAD_RIGHT) && defined(PAD_BOTTOM) && defined(KERNEL_WIDTH) && defined(KERNEL_HEIGHT) && defined(KERNEL_DEPTH) && defined(SRC_WIDTH) && defined(SRC_HEIGHT) && defined(VECTOR_SIZE) && defined(WIDTH_MOD_VECTOR_SIZE)
+#endif // defined(DATA_TYPE)
\ No newline at end of file
diff --git a/src/core/CL/cl_kernels/fixed_point.h b/src/core/CL/cl_kernels/fixed_point.h
index d55346b..46fa645 100644
--- a/src/core/CL/cl_kernels/fixed_point.h
+++ b/src/core/CL/cl_kernels/fixed_point.h
@@ -1,5 +1,5 @@
/*
- * Copyright (c) 2017 ARM Limited.
+ * Copyright (c) 2017-2018 ARM Limited.
*
* SPDX-License-Identifier: MIT
*
@@ -476,19 +476,19 @@
#define floatx16 float16
#define float16_TYPE float16
-#define CONVERTQ_DOWN_IMPL(in_type, out_type) \
- inline out_type convert_##out_type##_##in_type(in_type a, int fixed_point_position) \
- { \
- return CONVERT(a * (1 << fixed_point_position) + select((in_type)-0.5, (in_type)0.5, isgreater(a, (in_type)0)), out_type); \
+#define CONVERTQ_DOWN_IMPL(in_type, out_type) \
+ inline out_type convert_##out_type##_##in_type(in_type a, int fixed_point_position) \
+ { \
+ return CONVERT(a * (1 << fixed_point_position) + select((in_type)-0.5f, (in_type)0.5f, isgreater(a, (in_type)0)), out_type); \
}
CONVERTQ_DOWN_IMPL(float16, qs8x16)
CONVERTQ_DOWN_IMPL(float16, qs16x16)
-#define CONVERTQ_DOWN_SAT_IMPL(in_type, out_type) \
- inline out_type convert_##out_type##_##in_type##_sat(in_type a, int fixed_point_position) \
- { \
- return CONVERT_SAT(a * (1 << fixed_point_position) + select((in_type)-0.5, (in_type)0.5, isgreater(a, (in_type)0)), out_type); \
+#define CONVERTQ_DOWN_SAT_IMPL(in_type, out_type) \
+ inline out_type convert_##out_type##_##in_type##_sat(in_type a, int fixed_point_position) \
+ { \
+ return CONVERT_SAT(a * (1 << fixed_point_position) + select((in_type)-0.5f, (in_type)0.5f, isgreater(a, (in_type)0)), out_type); \
}
CONVERTQ_DOWN_SAT_IMPL(float16, qs8x16)
diff --git a/src/core/CL/cl_kernels/im2col.cl b/src/core/CL/cl_kernels/im2col.cl
new file mode 100644
index 0000000..75d99bd
--- /dev/null
+++ b/src/core/CL/cl_kernels/im2col.cl
@@ -0,0 +1,804 @@
+/*
+ * Copyright (c) 2018 ARM Limited.
+ *
+ * SPDX-License-Identifier: MIT
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to
+ * deal in the Software without restriction, including without limitation the
+ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
+ * sell copies of the Software, and to permit persons to whom the Software is
+ * furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in all
+ * copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+ * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+#include "helpers.h"
+
+#if defined(FIXED_POINT_POSITION)
+#include "fixed_point.h"
+#endif // FIXED_POINT_POSITION
+
+#if defined(DATA_TYPE) && defined(ELEMENT_SIZE)
+#if !defined(FIXED_POINT_POSITION)
+
+#if ELEMENT_SIZE == 1
+#define COND_DATA_TYPE char
+#elif ELEMENT_SIZE == 2
+#define COND_DATA_TYPE short
+#elif ELEMENT_SIZE == 4
+#define COND_DATA_TYPE int
+#else // ELEMENT_SIZE
+#error "Element size not support"
+#endif // ELEMENT_SIZE
+
+#if defined(CONVOLVED_WIDTH) && defined(STRIDE_Y) && defined(KERNEL_DEPTH)
+/** This kernel performs a reshaping of the input tensor to a tensor used to perform convolution using GEMM when the kernel size is 1x1 and the stride_x = 1
+ *
+ * @note This kernel computes 4 elements
+ * @note The data type must be passed at compile time using -DDATA_TYPE: e.g. -DDATA_TYPE=float
+ * @note The width of output tensor after matrix multiplication must be passed at compile time using -DCONVOLVED_WIDTH: e.g. -DCONVOLVED_WIDTH=34
+ * @note The kernel depth must be passed at compile time using -DKERNEL_DEPTH: e.g. -DKERNEL_DEPTH=3
+ * @note The stride along the Y direction must be passed at compile time using -DSTRIDE_Y: e.g. -DSTRIDE_Y=1
+ * @note In case biases will be added to the convolution -DHAS_BIAS has to be passed to append the final matrix with 1 in each row.
+ *
+ * @param[in] src_ptr Pointer to the source tensor. Supported data types: QS8/QASYMM8/QS16/F16/F32
+ * @param[in] src_stride_x Stride of the source tensor in X dimension (in bytes)
+ * @param[in] src_step_x src_stride_x * number of elements along X processed per workitem(in bytes)
+ * @param[in] src_stride_y Stride of the source tensor in Y dimension (in bytes)
+ * @param[in] src_step_y src_stride_y * number of elements along Y processed per workitem(in bytes)
+ * @param[in] src_stride_z Stride of the source tensor in Z dimension (in bytes)
+ * @param[in] src_step_z src_stride_z * number of elements along Z processed per workitem(in bytes)
+ * @param[in] src_offset_first_element_in_bytes The offset of the first element in the source tensor
+ * @param[out] dst_ptr Pointer to the destination tensor. Supported data types: same as @p src_ptr
+ * @param[in] dst_stride_x Stride of the destination tensor in X dimension (in bytes)
+ * @param[in] dst_step_x dst_stride_x * number of elements along X processed per workitem(in bytes)
+ * @param[in] dst_stride_y Stride of the destination tensor in Y dimension (in bytes)
+ * @param[in] dst_step_y dst_stride_y * number of elements along Y processed per workitem(in bytes)
+ * @param[in] dst_offset_first_element_in_bytes The offset of the first element in the destination tensor
+ * @param[in] src_stride_w Stride of the source tensor in W dimension (in bytes).
+ * @param[in] dst_stride_w Stride of the destination tensor in W dimension (in bytes).
+ */
+__kernel void im2col1x1_stridex1_dchw(
+ TENSOR3D_DECLARATION(src),
+ IMAGE_DECLARATION(dst),
+ uint src_stride_w,
+ uint dst_stride_w)
+{
+ const uint xc = get_global_id(0) * 4; // x coordinate in the convolved tensor
+ const uint yc = get_global_id(1); // y coordinate in the convolved tensor
+ const uint ch = get_global_id(2) % KERNEL_DEPTH; // input feature map
+ const uint batch = get_global_id(2) / KERNEL_DEPTH; // batch size
+
+ // Clamp xc
+ // The strategy clamps at "xc" as it will be a valid value for sure
+ uint4 xc_clamped = xc + (uint4)(0, 1, 2, 3);
+
+ // Check which values are valid
+ const VEC_DATA_TYPE(COND_DATA_TYPE, 4) cond0 = CONVERT((xc_clamped < SRC_WIDTH), VEC_DATA_TYPE(COND_DATA_TYPE, 4));
+
+ xc_clamped = select((uint4)xc, xc_clamped, convert_int4(cond0));
+
+ // Calculate input indices
+ const uint xi = xc;
+ const uint yi = yc * STRIDE_Y;
+
+ // Calculate output indices
+ const uint xo = ch;
+ const uint4 yo = xc_clamped + yc * CONVOLVED_WIDTH; // Index of the convolution
+
+ // Get input and output address
+ __global uchar *input_ptr = src_ptr + src_offset_first_element_in_bytes + xi * src_stride_x + yi * src_stride_y + ch * src_stride_z + batch * src_stride_w;
+
+ __global uchar *output_ptr = dst_ptr + dst_offset_first_element_in_bytes + xo * dst_stride_x + batch * dst_stride_w;
+
+ VEC_DATA_TYPE(DATA_TYPE, 4)
+ data = vload4(0, (__global DATA_TYPE *)input_ptr);
+
+ // If out-of-bound, overwrite with the first element
+ data = select((VEC_DATA_TYPE(DATA_TYPE, 4))data.s0, data, cond0);
+
+ *(__global DATA_TYPE *)(output_ptr + yo.s0 * dst_stride_y) = data.s0;
+ *(__global DATA_TYPE *)(output_ptr + yo.s1 * dst_stride_y) = data.s1;
+ *(__global DATA_TYPE *)(output_ptr + yo.s2 * dst_stride_y) = data.s2;
+ *(__global DATA_TYPE *)(output_ptr + yo.s3 * dst_stride_y) = data.s3;
+
+#ifdef HAS_BIAS
+ if(ch == (KERNEL_DEPTH - 1))
+ {
+ *((__global DATA_TYPE *)(output_ptr + yo.s0 * dst_stride_y) + 1) = 1.0f;
+ *((__global DATA_TYPE *)(output_ptr + yo.s1 * dst_stride_y) + 1) = 1.0f;
+ *((__global DATA_TYPE *)(output_ptr + yo.s2 * dst_stride_y) + 1) = 1.0f;
+ *((__global DATA_TYPE *)(output_ptr + yo.s3 * dst_stride_y) + 1) = 1.0f;
+ }
+#endif // HAS_BIAS
+}
+#endif // defined(CONVOLVED_WIDTH) && defined(STRIDE_Y) && defined(KERNEL_DEPTH)
+
+#if defined(CONVOLVED_WIDTH) && defined(SRC_WIDTH) && defined(SRC_HEIGHT) && defined(STRIDE_X) && defined(STRIDE_Y) && defined(KERNEL_DEPTH) && defined(PAD_LEFT) && defined(PAD_RIGHT) && defined(PAD_TOP) && defined(PAD_BOTTOM) && defined(PAD_VALUE)
+/** This kernel performs a reshaping of the input tensor to a tensor used to perform convolution using GEMM when the kernel size is 3x3
+ *
+ * @note The data type must be passed at compile time using -DDATA_TYPE: e.g. -DDATA_TYPE=float
+ * @note The width and height of the input tensor must be passed at compile time using -DSRC_WIDTH and -DSRC_HEIGHT: e.g. -DSRC_WIDTH=128 and -DSRC_HEIGHT=128
+ * @note The width of output tensor after matrix multiplication must be passed at compile time using -DCONVOLVED_WIDTH: e.g. -DCONVOLVED_WIDTH=34
+ * @note The kernel depth must be passed at compile time using -DKERNEL_DEPTH: e.g. -DKERNEL_DEPTH=3
+ * @note The pad_left, pad_right, pad_top and pad_bottom must be passed at compile time using -DPAD_LEFT, -DPAD_RIGHT, -DPAD_TOP and -DPAD_BOTTOM: e.g. -DPAD_LEFT=1, -DPAD_RIGHT=2, -DPAD_TOP=3 and -DPAD_BOTTOM=2
+ * @note The zero value to store in case we load values out-of-bounds must be passed at compile time using -DPAD_VALUE: e.g. -DPAD_VALUE=0.0
+ * @note The stride along the X and Y directions must be passed at compile time using -DSTRIDE_X and -DSTRIDE_Y: e.g. -DSTRIDE_X=1 and -DSTRIDE_Y=1
+ * @note In case biases will be added to the convolution -DHAS_BIAS has to be passed to append the final matrix with 1 in each row.
+ *
+ * @param[in] src_ptr Pointer to the source tensor. Supported data types: QS8/QASYMM8/QS16/F16/F32
+ * @param[in] src_stride_x Stride of the source tensor in X dimension (in bytes)
+ * @param[in] src_step_x src_stride_x * number of elements along X processed per workitem(in bytes)
+ * @param[in] src_stride_y Stride of the source tensor in Y dimension (in bytes)
+ * @param[in] src_step_y src_stride_y * number of elements along Y processed per workitem(in bytes)
+ * @param[in] src_stride_z Stride of the source tensor in Z dimension (in bytes)
+ * @param[in] src_step_z src_stride_z * number of elements along Z processed per workitem(in bytes)
+ * @param[in] src_offset_first_element_in_bytes The offset of the first element in the source tensor
+ * @param[out] dst_ptr Pointer to the destination tensor. Supported data types: same as @p src_ptr
+ * @param[in] dst_stride_x Stride of the destination tensor in X dimension (in bytes)
+ * @param[in] dst_step_x dst_stride_x * number of elements along X processed per workitem(in bytes)
+ * @param[in] dst_stride_y Stride of the destination tensor in Y dimension (in bytes)
+ * @param[in] dst_step_y dst_stride_y * number of elements along Y processed per workitem(in bytes)
+ * @param[in] dst_offset_first_element_in_bytes The offset of the first element in the destination tensor
+ * @param[in] src_stride_w Stride of the source tensor in W dimension (in bytes).
+ * @param[in] dst_stride_w Stride of the destination tensor in W dimension (in bytes).
+ */
+__kernel void im2col3x3_dchw(
+ TENSOR3D_DECLARATION(src),
+ IMAGE_DECLARATION(dst),
+ uint src_stride_w,
+ uint dst_stride_w)
+{
+ const int xc = get_global_id(0); // x coordinate in the convolved tensor
+ const int yc = get_global_id(1); // y coordinate in the convolved tensor
+ const int ch = get_global_id(2) % KERNEL_DEPTH; // input feature map
+ const int batch = get_global_id(2) / KERNEL_DEPTH; // batch size
+
+ // Calculate input indices
+ const int xi = xc * STRIDE_X - PAD_LEFT;
+ const int yi = yc * STRIDE_Y - PAD_TOP;
+
+ // Calculate output indices
+ const int xo = ch * 9; // 3x3
+ const int yo = xc + yc * CONVOLVED_WIDTH; // Index of the convolution
+
+ // Get input and output address
+ __global uchar *input_ptr = src_ptr + src_offset_first_element_in_bytes + xi * (int)src_stride_x + yi * (int)src_stride_y + ch * src_stride_z + batch * src_stride_w;
+
+ __global uchar *output_ptr = dst_ptr + dst_offset_first_element_in_bytes + xo * dst_stride_x + yo * dst_stride_y + batch * dst_stride_w;
+
+ VEC_DATA_TYPE(DATA_TYPE, 3)
+ row0 = vload3(0, (__global DATA_TYPE *)(input_ptr + 0 * src_stride_y));
+ VEC_DATA_TYPE(DATA_TYPE, 3)
+ row1 = vload3(0, (__global DATA_TYPE *)(input_ptr + 1 * src_stride_y));
+ VEC_DATA_TYPE(DATA_TYPE, 3)
+ row2 = vload3(0, (__global DATA_TYPE *)(input_ptr + 2 * src_stride_y));
+
+#if PAD_LEFT != 0 || PAD_TOP != 0 || PAD_RIGHT != 0 || PAD_BOTTOM != 0
+ // Put 0 if the value is out-of-bound
+ int3 x = (int3)xi + (int3)(0, 1, 2);
+ int3 y = (int3)yi + (int3)(0, 1, 2);
+
+ VEC_DATA_TYPE(COND_DATA_TYPE, 3)
+ cond0 = CONVERT((x >= (int3)0 && x < (int3)SRC_WIDTH && (int3)(y.s0 >= 0 && y.s0 < SRC_HEIGHT)), VEC_DATA_TYPE(COND_DATA_TYPE, 3));
+ VEC_DATA_TYPE(COND_DATA_TYPE, 3)
+ cond1 = CONVERT((x >= (int3)0 && x < (int3)SRC_WIDTH && (int3)(y.s1 >= 0 && y.s1 < SRC_HEIGHT)), VEC_DATA_TYPE(COND_DATA_TYPE, 3));
+ VEC_DATA_TYPE(COND_DATA_TYPE, 3)
+ cond2 = CONVERT((x >= (int3)0 && x < (int3)SRC_WIDTH && (int3)(y.s2 >= 0 && y.s2 < SRC_HEIGHT)), VEC_DATA_TYPE(COND_DATA_TYPE, 3));
+
+ row0 = select((VEC_DATA_TYPE(DATA_TYPE, 3))PAD_VALUE, row0, cond0);
+ row1 = select((VEC_DATA_TYPE(DATA_TYPE, 3))PAD_VALUE, row1, cond1);
+ row2 = select((VEC_DATA_TYPE(DATA_TYPE, 3))PAD_VALUE, row2, cond2);
+#endif // PAD_LEFT != 0 || PAD_TOP != 0 || PAD_RIGHT != 0 || PAD_BOTTOM != 0
+
+ vstore8((VEC_DATA_TYPE(DATA_TYPE, 8))(row0.s012, row1.s012, row2.s01), 0, (__global DATA_TYPE *)output_ptr);
+ *((__global DATA_TYPE *)output_ptr + 8) = row2.s2;
+
+#ifdef HAS_BIAS
+ if(ch == (KERNEL_DEPTH - 1))
+ {
+ *((__global DATA_TYPE *)output_ptr + 9) = 1.0f;
+ }
+#endif // HAS_BIAS
+}
+
+/** This kernel performs a reshaping of the input tensor to a tensor used to perform convolution using GEMM when the kernel size is 5x5
+ *
+ * @note The data type must be passed at compile time using -DDATA_TYPE: e.g. -DDATA_TYPE=float
+ * @note The width and height of the input tensor must be passed at compile time using -DSRC_WIDTH and -DSRC_HEIGHT: e.g. -DSRC_WIDTH=128 and -DSRC_HEIGHT=128
+ * @note The width of output tensor after matrix multiplication must be passed at compile time using -DCONVOLVED_WIDTH: e.g. -DCONVOLVED_WIDTH=34
+ * @note The kernel depth must be passed at compile time using -DKERNEL_DEPTH: e.g. -DKERNEL_DEPTH=3
+ * @note The pad_left, pad_right, pad_top and pad_bottom must be passed at compile time using -DPAD_LEFT, -DPAD_RIGHT, -DPAD_TOP and -DPAD_BOTTOM: e.g. -DPAD_LEFT=1, -DPAD_RIGHT=2, -DPAD_TOP=3 and -DPAD_BOTTOM=2
+ * @note The zero value to store in case we load values out-of-bounds must be passed at compile time using -DPAD_VALUE: e.g. -DPAD_VALUE=0.0
+ * @note The stride along the X and Y directions must be passed at compile time using -DSTRIDE_X and -DSTRIDE_Y: e.g. -DSTRIDE_X=1 and -DSTRIDE_Y=1
+ * @note In case biases will be added to the convolution -DHAS_BIAS has to be passed to append the final matrix with 1 in each row.
+ *
+ * @param[in] src_ptr Pointer to the source tensor. Supported data types: QS8/QASYMM8/QS16/F16/F32
+ * @param[in] src_stride_x Stride of the source tensor in X dimension (in bytes)
+ * @param[in] src_step_x src_stride_x * number of elements along X processed per workitem(in bytes)
+ * @param[in] src_stride_y Stride of the source tensor in Y dimension (in bytes)
+ * @param[in] src_step_y src_stride_y * number of elements along Y processed per workitem(in bytes)
+ * @param[in] src_stride_z Stride of the source tensor in Z dimension (in bytes)
+ * @param[in] src_step_z src_stride_z * number of elements along Z processed per workitem(in bytes)
+ * @param[in] src_offset_first_element_in_bytes The offset of the first element in the source tensor
+ * @param[out] dst_ptr Pointer to the destination tensor. Supported data types: same as @p src_ptr
+ * @param[in] dst_stride_x Stride of the destination tensor in X dimension (in bytes)
+ * @param[in] dst_step_x dst_stride_x * number of elements along X processed per workitem(in bytes)
+ * @param[in] dst_stride_y Stride of the destination tensor in Y dimension (in bytes)
+ * @param[in] dst_step_y dst_stride_y * number of elements along Y processed per workitem(in bytes)
+ * @param[in] dst_offset_first_element_in_bytes The offset of the first element in the destination tensor
+ * @param[in] src_stride_w Stride of the source tensor in W dimension (in bytes).
+ * @param[in] dst_stride_w Stride of the destination tensor in W dimension (in bytes).
+ */
+__kernel void im2col5x5_dchw(
+ TENSOR3D_DECLARATION(src),
+ IMAGE_DECLARATION(dst),
+ uint src_stride_w,
+ uint dst_stride_w)
+{
+ const int xc = get_global_id(0); // x coordinate in the convolved tensor
+ const int yc = get_global_id(1); // y coordinate in the convolved tensor
+ const int ch = get_global_id(2) % KERNEL_DEPTH; // input feature map
+ const int batch = get_global_id(2) / KERNEL_DEPTH; // batch size
+
+ // Calculate input indices
+ const int xi = xc * STRIDE_X - PAD_LEFT;
+ const int yi = yc * STRIDE_Y - PAD_TOP;
+
+ // Calculate output indices
+ const int xo = ch * 25; // 5x5
+ const int yo = xc + yc * CONVOLVED_WIDTH; // Index of the convolution
+
+#if PAD_LEFT != 0 || PAD_TOP != 0 || PAD_RIGHT != 0 || PAD_BOTTOM != 0
+ // Put 0 if the value is out-of-bound
+ int4 x0 = (int4)xi + (int4)(0, 1, 2, 3);
+ int4 y0 = (int4)yi + (int4)(0, 1, 2, 3);
+ int x1 = xi + 4;
+ int y1 = yi + 4;
+
+ // Check if we could have out-of-bounds elements in the x direction
+ VEC_DATA_TYPE(COND_DATA_TYPE, 4)
+ x0_condition = CONVERT((x0 >= (int4)0 && x0 < (int4)SRC_WIDTH), VEC_DATA_TYPE(COND_DATA_TYPE, 4));
+ VEC_DATA_TYPE(COND_DATA_TYPE, 4)
+ y0_condition = CONVERT((y0 >= (int4)0 && y0 < (int4)SRC_HEIGHT), VEC_DATA_TYPE(COND_DATA_TYPE, 4));
+ COND_DATA_TYPE x1_condition = (COND_DATA_TYPE)(x1 >= 0 && x1 < SRC_WIDTH);
+ COND_DATA_TYPE y1_condition = (COND_DATA_TYPE)(y1 >= 0 && y1 < SRC_HEIGHT);
+#endif // PAD_LEFT != 0 || PAD_TOP != 0 || PAD_RIGHT != 0 || PAD_BOTTOM != 0
+
+ // Get input and output address
+ __global uchar *input_ptr = src_ptr + src_offset_first_element_in_bytes + xi * (int)src_stride_x + yi * (int)src_stride_y + ch * src_stride_z + batch * src_stride_w;
+
+ __global uchar *output_ptr = dst_ptr + dst_offset_first_element_in_bytes + xo * dst_stride_x + yo * dst_stride_y + batch * dst_stride_w;
+
+ {
+ VEC_DATA_TYPE(DATA_TYPE, 4)
+ row00 = vload4(0, (__global DATA_TYPE *)input_ptr);
+ DATA_TYPE
+ row01 = *((__global DATA_TYPE *)input_ptr + 4);
+
+ input_ptr += src_stride_y;
+
+ VEC_DATA_TYPE(DATA_TYPE, 4)
+ row10 = vload4(0, (__global DATA_TYPE *)input_ptr);
+ DATA_TYPE
+ row11 = *((__global DATA_TYPE *)input_ptr + 4);
+
+#if PAD_LEFT != 0 || PAD_TOP != 0 || PAD_RIGHT != 0 || PAD_BOTTOM != 0
+ VEC_DATA_TYPE(COND_DATA_TYPE, 4)
+ cond00 = x0_condition && (VEC_DATA_TYPE(COND_DATA_TYPE, 4))y0_condition.s0;
+ VEC_DATA_TYPE(COND_DATA_TYPE, 4)
+ cond10 = x0_condition && (VEC_DATA_TYPE(COND_DATA_TYPE, 4))y0_condition.s1;
+ COND_DATA_TYPE cond01 = (COND_DATA_TYPE)(x1_condition && y0_condition.s0);
+ COND_DATA_TYPE cond11 = (COND_DATA_TYPE)(x1_condition && y0_condition.s1);
+
+ // Replace with 0 if the value is not valid
+ row00 = select((VEC_DATA_TYPE(DATA_TYPE, 4))PAD_VALUE, row00, cond00);
+ row10 = select((VEC_DATA_TYPE(DATA_TYPE, 4))PAD_VALUE, row10, cond10);
+ row01 = select((DATA_TYPE)PAD_VALUE, row01, cond01);
+ row11 = select((DATA_TYPE)PAD_VALUE, row11, cond11);
+#endif // PAD_LEFT != 0 || PAD_TOP != 0 || PAD_RIGHT != 0 || PAD_BOTTOM != 0
+
+ vstore8((VEC_DATA_TYPE(DATA_TYPE, 8))(row00.s0123, row01,
+ row10.s012),
+ 0, (__global DATA_TYPE *)output_ptr);
+ vstore2((VEC_DATA_TYPE(DATA_TYPE, 2))(row10.s3, row11), 0, (__global DATA_TYPE *)output_ptr + 8);
+
+ input_ptr += src_stride_y;
+ output_ptr += 10 * dst_stride_x;
+ }
+
+ {
+ VEC_DATA_TYPE(DATA_TYPE, 4)
+ row00 = vload4(0, (__global DATA_TYPE *)input_ptr);
+ DATA_TYPE
+ row01 = *((__global DATA_TYPE *)input_ptr + 4);
+
+ input_ptr += src_stride_y;
+
+ VEC_DATA_TYPE(DATA_TYPE, 4)
+ row10 = vload4(0, (__global DATA_TYPE *)input_ptr);
+ DATA_TYPE
+ row11 = *((__global DATA_TYPE *)input_ptr + 4);
+
+#if PAD_LEFT != 0 || PAD_TOP != 0 || PAD_RIGHT != 0 || PAD_BOTTOM != 0
+ VEC_DATA_TYPE(COND_DATA_TYPE, 4)
+ cond00 = x0_condition && (VEC_DATA_TYPE(COND_DATA_TYPE, 4))y0_condition.s2;
+ VEC_DATA_TYPE(COND_DATA_TYPE, 4)
+ cond10 = x0_condition && (VEC_DATA_TYPE(COND_DATA_TYPE, 4))y0_condition.s3;
+ COND_DATA_TYPE cond01 = (COND_DATA_TYPE)(x1_condition && y0_condition.s2);
+ COND_DATA_TYPE cond11 = (COND_DATA_TYPE)(x1_condition && y0_condition.s3);
+
+ // Replace with 0 if the value is not valid
+ row00 = select((VEC_DATA_TYPE(DATA_TYPE, 4))PAD_VALUE, row00, cond00);
+ row10 = select((VEC_DATA_TYPE(DATA_TYPE, 4))PAD_VALUE, row10, cond10);
+ row01 = select((DATA_TYPE)PAD_VALUE, row01, cond01);
+ row11 = select((DATA_TYPE)PAD_VALUE, row11, cond11);
+#endif // PAD_LEFT != 0 || PAD_TOP != 0 || PAD_RIGHT != 0 || PAD_BOTTOM != 0
+
+ vstore8((VEC_DATA_TYPE(DATA_TYPE, 8))(row00.s0123, row01,
+ row10.s012),
+ 0, (__global DATA_TYPE *)output_ptr);
+ vstore2((VEC_DATA_TYPE(DATA_TYPE, 2))(row10.s3, row11), 0, (__global DATA_TYPE *)output_ptr + 8);
+
+ input_ptr += src_stride_y;
+ output_ptr += 10 * dst_stride_x;
+ }
+
+ {
+ VEC_DATA_TYPE(DATA_TYPE, 4)
+ row00 = vload4(0, (__global DATA_TYPE *)input_ptr);
+ DATA_TYPE
+ row01 = *((__global DATA_TYPE *)input_ptr + 4);
+
+ input_ptr += src_stride_y;
+
+#if PAD_LEFT != 0 || PAD_TOP != 0 || PAD_RIGHT != 0 || PAD_BOTTOM != 0
+ VEC_DATA_TYPE(COND_DATA_TYPE, 4)
+ cond00 = x0_condition && (VEC_DATA_TYPE(COND_DATA_TYPE, 4))y1_condition;
+ COND_DATA_TYPE cond01 = (COND_DATA_TYPE)(x1_condition && y1_condition);
+
+ // Replace with 0 if the value is not valid
+ row00 = select((VEC_DATA_TYPE(DATA_TYPE, 4))PAD_VALUE, row00, cond00);
+ row01 = select((DATA_TYPE)PAD_VALUE, row01, cond01);
+#endif // PAD_LEFT != 0 || PAD_TOP != 0 || PAD_RIGHT != 0 || PAD_BOTTOM != 0
+
+ vstore4(row00, 0, (__global DATA_TYPE *)output_ptr);
+ *((__global DATA_TYPE *)output_ptr + 4) = row01;
+
+ output_ptr += 5 * dst_stride_x;
+ }
+
+#ifdef HAS_BIAS
+ if(ch == (KERNEL_DEPTH - 1))
+ {
+ *((__global DATA_TYPE *)output_ptr) = 1.0f;
+ }
+#endif // HAS_BIAS
+}
+#endif // defined(CONVOLVED_WIDTH) && defined(SRC_WIDTH) && defined(SRC_HEIGHT) && defined(STRIDE_X) && defined(STRIDE_Y) && defined(KERNEL_DEPTH) && defined(PAD_LEFT) && defined(PAD_RIGHT) && defined(PAD_TOP) && defined(PAD_BOTTOM) && defined(PAD_VALUE)
+
+#if defined(CONVOLVED_WIDTH) && defined(STRIDE_X) && defined(STRIDE_Y) && defined(KERNEL_DEPTH)
+/** This kernel performs a reshaping of the input tensor to a tensor used to perform convolution using GEMM when the kernel size is 11x11
+ *
+ * @note The data type must be passed at compile time using -DDATA_TYPE: e.g. -DDATA_TYPE=float
+ * @note The width of output tensor after matrix multiplication must be passed at compile time using -DCONVOLVED_WIDTH: e.g. -DCONVOLVED_WIDTH=34
+ * @note The kernel depth must be passed at compile time using -DKERNEL_DEPTH: e.g. -DKERNEL_DEPTH=3
+ * @note The stride along the X and Y directions must be passed at compile time using -DSTRIDE_X and -DSTRIDE_Y: e.g. -DSTRIDE_X=1 and -DSTRIDE_Y=1
+ * @note In case biases will be added to the convolution -DHAS_BIAS has to be passed to append the final matrix with 1 in each row.
+ *
+ * @param[in] src_ptr Pointer to the source tensor. Supported data types: QS8/QASYMM8/QS16/F16/F32
+ * @param[in] src_stride_x Stride of the source tensor in X dimension (in bytes)
+ * @param[in] src_step_x src_stride_x * number of elements along X processed per workitem(in bytes)
+ * @param[in] src_stride_y Stride of the source tensor in Y dimension (in bytes)
+ * @param[in] src_step_y src_stride_y * number of elements along Y processed per workitem(in bytes)
+ * @param[in] src_stride_z Stride of the source tensor in Z dimension (in bytes)
+ * @param[in] src_step_z src_stride_z * number of elements along Z processed per workitem(in bytes)
+ * @param[in] src_offset_first_element_in_bytes The offset of the first element in the source tensor
+ * @param[out] dst_ptr Pointer to the destination tensor. Supported data types: same as @p src_ptr
+ * @param[in] dst_stride_x Stride of the destination tensor in X dimension (in bytes)
+ * @param[in] dst_step_x dst_stride_x * number of elements along X processed per workitem(in bytes)
+ * @param[in] dst_stride_y Stride of the destination tensor in Y dimension (in bytes)
+ * @param[in] dst_step_y dst_stride_y * number of elements along Y processed per workitem(in bytes)
+ * @param[in] dst_offset_first_element_in_bytes The offset of the first element in the destination tensor
+ * @param[in] src_stride_w Stride of the source tensor in W dimension (in bytes).
+ * @param[in] dst_stride_w Stride of the destination tensor in W dimension (in bytes).
+ */
+__kernel void im2col11x11_padx0_pady0_dchw(
+ TENSOR3D_DECLARATION(src),
+ IMAGE_DECLARATION(dst),
+ uint src_stride_w,
+ uint dst_stride_w)
+{
+ const int xc = get_global_id(0); // x coordinate in the convolved tensor
+ const int yc = get_global_id(1); // y coordinate in the convolved tensor
+ const int ch = get_global_id(2) % KERNEL_DEPTH; // input feature map
+ const int batch = get_global_id(2) / KERNEL_DEPTH; // batch size
+
+ // Calculate input indices
+ const int xi = xc * STRIDE_X;
+ const int yi = yc * STRIDE_Y;
+
+ // Calculate output indices
+ const int xo = ch * 121; // 11x11
+ const int yo = xc + yc * CONVOLVED_WIDTH; // Index of the convolution
+
+ // Get input and output address
+ __global uchar *input_ptr = src_ptr + src_offset_first_element_in_bytes + xi * src_stride_x + yi * src_stride_y + ch * src_stride_z + batch * src_stride_w;
+
+ __global uchar *output_ptr = dst_ptr + dst_offset_first_element_in_bytes + xo * dst_stride_x + yo * dst_stride_y + batch * dst_stride_w;
+ {
+ VEC_DATA_TYPE(DATA_TYPE, 8)
+ row00 = vload8(0, (__global DATA_TYPE *)(input_ptr));
+ VEC_DATA_TYPE(DATA_TYPE, 3)
+ row01 = vload3(0, (__global DATA_TYPE *)(input_ptr) + 8);
+
+ vstore8((VEC_DATA_TYPE(DATA_TYPE, 8))(row00.s01234567), 0, (__global DATA_TYPE *)output_ptr);
+ vstore3((VEC_DATA_TYPE(DATA_TYPE, 3))(row01.s012), 0, (__global DATA_TYPE *)output_ptr + 8);
+
+ input_ptr += src_stride_y;
+ output_ptr += 11 * src_stride_x;
+ }
+
+ {
+ VEC_DATA_TYPE(DATA_TYPE, 8)
+ row00 = vload8(0, (__global DATA_TYPE *)(input_ptr));
+ VEC_DATA_TYPE(DATA_TYPE, 3)
+ row01 = vload3(0, (__global DATA_TYPE *)(input_ptr) + 8);
+
+ vstore8((VEC_DATA_TYPE(DATA_TYPE, 8))(row00.s01234567), 0, (__global DATA_TYPE *)output_ptr);
+ vstore3((VEC_DATA_TYPE(DATA_TYPE, 3))(row01.s012), 0, (__global DATA_TYPE *)output_ptr + 8);
+
+ input_ptr += src_stride_y;
+ output_ptr += 11 * src_stride_x;
+ }
+
+ {
+ VEC_DATA_TYPE(DATA_TYPE, 8)
+ row00 = vload8(0, (__global DATA_TYPE *)(input_ptr));
+ VEC_DATA_TYPE(DATA_TYPE, 3)
+ row01 = vload3(0, (__global DATA_TYPE *)(input_ptr) + 8);
+
+ vstore8((VEC_DATA_TYPE(DATA_TYPE, 8))(row00.s01234567), 0, (__global DATA_TYPE *)output_ptr);
+ vstore3((VEC_DATA_TYPE(DATA_TYPE, 3))(row01.s012), 0, (__global DATA_TYPE *)output_ptr + 8);
+
+ input_ptr += src_stride_y;
+ output_ptr += 11 * src_stride_x;
+ }
+
+ {
+ VEC_DATA_TYPE(DATA_TYPE, 8)
+ row00 = vload8(0, (__global DATA_TYPE *)(input_ptr));
+ VEC_DATA_TYPE(DATA_TYPE, 3)
+ row01 = vload3(0, (__global DATA_TYPE *)(input_ptr) + 8);
+
+ vstore8((VEC_DATA_TYPE(DATA_TYPE, 8))(row00.s01234567), 0, (__global DATA_TYPE *)output_ptr);
+ vstore3((VEC_DATA_TYPE(DATA_TYPE, 3))(row01.s012), 0, (__global DATA_TYPE *)output_ptr + 8);
+
+ input_ptr += src_stride_y;
+ output_ptr += 11 * src_stride_x;
+ }
+
+ {
+ VEC_DATA_TYPE(DATA_TYPE, 8)
+ row00 = vload8(0, (__global DATA_TYPE *)(input_ptr));
+ VEC_DATA_TYPE(DATA_TYPE, 3)
+ row01 = vload3(0, (__global DATA_TYPE *)(input_ptr) + 8);
+
+ vstore8((VEC_DATA_TYPE(DATA_TYPE, 8))(row00.s01234567), 0, (__global DATA_TYPE *)output_ptr);
+ vstore3((VEC_DATA_TYPE(DATA_TYPE, 3))(row01.s012), 0, (__global DATA_TYPE *)output_ptr + 8);
+
+ input_ptr += src_stride_y;
+ output_ptr += 11 * src_stride_x;
+ }
+
+ {
+ VEC_DATA_TYPE(DATA_TYPE, 8)
+ row00 = vload8(0, (__global DATA_TYPE *)(input_ptr));
+ VEC_DATA_TYPE(DATA_TYPE, 3)
+ row01 = vload3(0, (__global DATA_TYPE *)(input_ptr) + 8);
+
+ vstore8((VEC_DATA_TYPE(DATA_TYPE, 8))(row00.s01234567), 0, (__global DATA_TYPE *)output_ptr);
+ vstore3((VEC_DATA_TYPE(DATA_TYPE, 3))(row01.s012), 0, (__global DATA_TYPE *)output_ptr + 8);
+
+ input_ptr += src_stride_y;
+ output_ptr += 11 * src_stride_x;
+ }
+
+ {
+ VEC_DATA_TYPE(DATA_TYPE, 8)
+ row00 = vload8(0, (__global DATA_TYPE *)(input_ptr));
+ VEC_DATA_TYPE(DATA_TYPE, 3)
+ row01 = vload3(0, (__global DATA_TYPE *)(input_ptr) + 8);
+
+ vstore8((VEC_DATA_TYPE(DATA_TYPE, 8))(row00.s01234567), 0, (__global DATA_TYPE *)output_ptr);
+ vstore3((VEC_DATA_TYPE(DATA_TYPE, 3))(row01.s012), 0, (__global DATA_TYPE *)output_ptr + 8);
+
+ input_ptr += src_stride_y;
+ output_ptr += 11 * src_stride_x;
+ }
+
+ {
+ VEC_DATA_TYPE(DATA_TYPE, 8)
+ row00 = vload8(0, (__global DATA_TYPE *)(input_ptr));
+ VEC_DATA_TYPE(DATA_TYPE, 3)
+ row01 = vload3(0, (__global DATA_TYPE *)(input_ptr) + 8);
+
+ vstore8((VEC_DATA_TYPE(DATA_TYPE, 8))(row00.s01234567), 0, (__global DATA_TYPE *)output_ptr);
+ vstore3((VEC_DATA_TYPE(DATA_TYPE, 3))(row01.s012), 0, (__global DATA_TYPE *)output_ptr + 8);
+
+ input_ptr += src_stride_y;
+ output_ptr += 11 * src_stride_x;
+ }
+
+ {
+ VEC_DATA_TYPE(DATA_TYPE, 8)
+ row00 = vload8(0, (__global DATA_TYPE *)(input_ptr));
+ VEC_DATA_TYPE(DATA_TYPE, 3)
+ row01 = vload3(0, (__global DATA_TYPE *)(input_ptr) + 8);
+
+ vstore8((VEC_DATA_TYPE(DATA_TYPE, 8))(row00.s01234567), 0, (__global DATA_TYPE *)output_ptr);
+ vstore3((VEC_DATA_TYPE(DATA_TYPE, 3))(row01.s012), 0, (__global DATA_TYPE *)output_ptr + 8);
+
+ input_ptr += src_stride_y;
+ output_ptr += 11 * src_stride_x;
+ }
+
+ {
+ VEC_DATA_TYPE(DATA_TYPE, 8)
+ row00 = vload8(0, (__global DATA_TYPE *)(input_ptr));
+ VEC_DATA_TYPE(DATA_TYPE, 3)
+ row01 = vload3(0, (__global DATA_TYPE *)(input_ptr) + 8);
+
+ vstore8((VEC_DATA_TYPE(DATA_TYPE, 8))(row00.s01234567), 0, (__global DATA_TYPE *)output_ptr);
+ vstore3((VEC_DATA_TYPE(DATA_TYPE, 3))(row01.s012), 0, (__global DATA_TYPE *)output_ptr + 8);
+
+ input_ptr += src_stride_y;
+ output_ptr += 11 * src_stride_x;
+ }
+
+ {
+ VEC_DATA_TYPE(DATA_TYPE, 8)
+ row00 = vload8(0, (__global DATA_TYPE *)(input_ptr));
+ VEC_DATA_TYPE(DATA_TYPE, 3)
+ row01 = vload3(0, (__global DATA_TYPE *)(input_ptr) + 8);
+
+ vstore8((VEC_DATA_TYPE(DATA_TYPE, 8))(row00.s01234567), 0, (__global DATA_TYPE *)output_ptr);
+ vstore3((VEC_DATA_TYPE(DATA_TYPE, 3))(row01.s012), 0, (__global DATA_TYPE *)output_ptr + 8);
+
+ output_ptr += 11 * src_stride_x;
+ }
+
+#ifdef HAS_BIAS
+ if(ch == (KERNEL_DEPTH - 1))
+ {
+ *((__global DATA_TYPE *)output_ptr) = 1.0f;
+ }
+#endif // HAS_BIAS
+}
+#endif // defined(CONVOLVED_WIDTH) && defined(STRIDE_X) && defined(STRIDE_Y) && defined(KERNEL_DEPTH)
+#endif // !defined(FIXED_POINT_POSITION)
+
+#if defined(CONVOLVED_WIDTH) && defined(STRIDE_X) && defined(STRIDE_Y) && defined(KERNEL_WIDTH) && defined(KERNEL_HEIGHT) && defined(KERNEL_DEPTH) && defined(SRC_WIDTH) && defined(SRC_HEIGHT) && defined(VECTOR_SIZE) && defined(WIDTH_MOD_VECTOR_SIZE)
+/** This kernel reshapes the input tensor to a tensor used to perform convolution using GEMM when
+ * the kernel width is greater than 1 (except when the kernel size is 3x3) and pad_x == pad_y == 0.
+ *
+ * @note The data type must be passed at compile time using -DDATA_TYPE e.g. -DDATA_TYPE=float.
+ * @note The vector size must be passed at compile time using -DVECTOR_SIZE e.g. -DVECTOR_SIZE=4.
+ * @note The width modulo vector size must be passed at compile time using -DWIDTH_MOD_VECTOR_SIZE e.g. -DWIDTH_MOD_VECTOR_SIZE=3.
+ * @note In case biases will be added to the convolution -DHAS_BIAS has to be passed to append the final matrix with 1 in each row.
+ *
+ * @param[in] src_ptr Pointer to the source tensor. Supported data types: QS8/QS16/F16/F32
+ * @param[in] src_stride_x Stride of the source tensor in X dimension (in bytes)
+ * @param[in] src_step_x src_stride_x * number of elements along X processed per workitem(in bytes)
+ * @param[in] src_stride_y Stride of the source tensor in Y dimension (in bytes)
+ * @param[in] src_step_y src_stride_y * number of elements along Y processed per workitem(in bytes)
+ * @param[in] src_stride_z Stride of the source tensor in Z dimension (in bytes)
+ * @param[in] src_step_z src_stride_z * number of elements along Z processed per workitem(in bytes)
+ * @param[in] src_offset_first_element_in_bytes The offset of the first element in the source tensor
+ * @param[out] dst_ptr Pointer to the destination tensor. Supported data types: same as @p src_ptr
+ * @param[in] dst_stride_x Stride of the destination tensor in X dimension (in bytes)
+ * @param[in] dst_step_x dst_stride_x * number of elements along X processed per workitem(in bytes)
+ * @param[in] dst_stride_y Stride of the destination tensor in Y dimension (in bytes)
+ * @param[in] dst_step_y dst_stride_y * number of elements along Y processed per workitem(in bytes)
+ * @param[in] dst_offset_first_element_in_bytes The offset of the first element in the destination tensor
+ * @param[in] src_stride_w Stride of the source tensor in W dimension (in bytes).
+ * @param[in] dst_stride_w Stride of the destination tensor in W dimension (in bytes).
+ */
+__kernel void im2col_generic_padx0_pady0_dchw(
+ TENSOR3D_DECLARATION(src),
+ IMAGE_DECLARATION(dst),
+ uint src_stride_w,
+ uint dst_stride_w)
+{
+ const int xc = get_global_id(0); // x coordinate in the convolved tensor
+ const int yc = get_global_id(1); // y coordinate in the convolved tensor
+ const int ch = get_global_id(2) % KERNEL_DEPTH; // input feature map
+ const int batch = get_global_id(2) / KERNEL_DEPTH; // batch size
+
+ // Calculate input indices
+ const int xi = xc * STRIDE_X;
+ const int yi = yc * STRIDE_Y;
+ // Calculate output indices
+ const int xo = ch * KERNEL_WIDTH * KERNEL_HEIGHT;
+ const int yo = xc + yc * CONVOLVED_WIDTH; // Index of the convolution
+ __global uchar *input_ptr = src_ptr + src_offset_first_element_in_bytes + ch * src_stride_z + batch * src_stride_w;
+ __global DATA_TYPE *output_ptr = ((__global DATA_TYPE *)(dst_ptr + dst_offset_first_element_in_bytes + yo * dst_stride_y + batch * dst_stride_w)) + xo;
+ // Linearize convolution elements
+ for(int y = yi, y_e = yi + KERNEL_HEIGHT; y < y_e; ++y)
+ {
+ int last_x = 0;
+ for(int x = xi, x_e = xi + KERNEL_WIDTH; x + VECTOR_SIZE <= x_e; x += VECTOR_SIZE, output_ptr += VECTOR_SIZE)
+ {
+ VEC_DATA_TYPE(DATA_TYPE, VECTOR_SIZE)
+ row = VLOAD(VECTOR_SIZE)(0, (__global DATA_TYPE *)(input_ptr + x * src_stride_x + y * src_stride_y));
+ VSTORE(VECTOR_SIZE)
+ (row, 0, output_ptr);
+ last_x = x;
+ }
+ // Copy the remainder of the row by doing VLOAD(WIDTH_MOD_VECTOR_SIZE) and VSTORE(WIDTH_MOD_VECTOR_SIZE).
+ // Note that x and output_ptr have already been incremented by VECTOR_SIZE by the loop just before exit.
+#if WIDTH_MOD_VECTOR_SIZE == 1
+ *output_ptr = *((__global DATA_TYPE *)(input_ptr + (last_x + VECTOR_SIZE) * src_stride_x + y * src_stride_y));
+#elif WIDTH_MOD_VECTOR_SIZE > 1
+ VEC_DATA_TYPE(DATA_TYPE, WIDTH_MOD_VECTOR_SIZE)
+ row = VLOAD(WIDTH_MOD_VECTOR_SIZE)(0, (__global DATA_TYPE *)(input_ptr + (last_x + VECTOR_SIZE) * src_stride_x + y * src_stride_y));
+ VSTORE(WIDTH_MOD_VECTOR_SIZE)
+ (row, 0, output_ptr);
+#endif /* WIDTH_MOD_VECTOR_SIZE */
+ output_ptr += WIDTH_MOD_VECTOR_SIZE;
+ } /* End of loop over KERNEL_HEIGHT */
+
+#ifdef HAS_BIAS
+ if(ch == (KERNEL_DEPTH - 1))
+ {
+#ifdef FIXED_POINT_POSITION
+ *output_ptr = (DATA_TYPE)(1 << FIXED_POINT_POSITION);
+#else // FIXED_POINT_POSITION
+ *output_ptr = 1.0f;
+#endif // FIXED_POINT_POSITION
+ }
+#endif // HAS_BIAS
+}
+#endif //defined(CONVOLVED_WIDTH) && defined(STRIDE_X) && defined(STRIDE_Y) && defined(PAD_LEFT) && defined(PAD_TOP) && defined(PAD_RIGHT) && defined(PAD_BOTTOM) && defined(KERNEL_WIDTH) && defined(KERNEL_HEIGHT) && defined(KERNEL_DEPTH) && defined(SRC_WIDTH) && defined(SRC_HEIGHT) && defined(VECTOR_SIZE) && defined(WIDTH_MOD_VECTOR_SIZE)
+
+#if defined(CONVOLVED_WIDTH) && defined(SRC_WIDTH) && defined(SRC_HEIGHT) && defined(STRIDE_X) && defined(STRIDE_Y) && defined(KERNEL_WIDTH) && defined(KERNEL_HEIGHT) && defined(KERNEL_DEPTH) && defined(PAD_LEFT) && defined(PAD_RIGHT) && defined(PAD_TOP) && defined(PAD_BOTTOM) && defined(PAD_VALUE)
+/** This kernel performs a reshaping of the input tensor to a tensor used to perform convolution using GEMM.
+ *
+ * @note The data type must be passed at compile time using -DDATA_TYPE: e.g. -DDATA_TYPE=float
+ * @note The width and height of the input tensor must be passed at compile time using -DSRC_WIDTH and -DSRC_HEIGHT: e.g. -DSRC_WIDTH=128 and -DSRC_HEIGHT=128
+ * @note The width of output tensor after matrix multiplication must be passed at compile time using -DCONVOLVED_WIDTH: e.g. -DCONVOLVED_WIDTH=34
+ * @note The kernel width, height and depth must be passed at compile time using -DKERNEL_WIDTH, -DKERNEL_HEIGHT and -DKERNEL_DEPTH: e.g. -DKERNEL_WIDTH=3, -DKERNEL_HEIGHT=3 and -DKERNEL_DEPTH=64
+ * @note The pad_left, pad_right, pad_top and pad_bottom must be passed at compile time using -DPAD_LEFT, -DPAD_RIGHT, -DPAD_TOP and -DPAD_BOTTOM: e.g. -DPAD_LEFT=1, -DPAD_RIGHT=2, -DPAD_TOP=3 and -DPAD_BOTTOM=2
+ * @note The zero value to store in case we load values out-of-bounds must be passed at compile time using -DPAD_VALUE: e.g. -DPAD_VALUE=0.0
+ * @note The stride along the X and Y directions must be passed at compile time using -DSTRIDE_X and -DSTRIDE_Y: e.g. -DSTRIDE_X=1 and -DSTRIDE_Y=1
+ * @note In case biases will be added to the convolution -DHAS_BIAS has to be passed to append the final matrix with 1 in each row.
+ *
+ * @param[in] src_ptr Pointer to the source tensor. Supported data types: QS8/QASYMM8/QS16/F16/F32
+ * @param[in] src_stride_x Stride of the source tensor in X dimension (in bytes)
+ * @param[in] src_step_x src_stride_x * number of elements along X processed per workitem(in bytes)
+ * @param[in] src_stride_y Stride of the source tensor in Y dimension (in bytes)
+ * @param[in] src_step_y src_stride_y * number of elements along Y processed per workitem(in bytes)
+ * @param[in] src_stride_z Stride of the source tensor in Z dimension (in bytes)
+ * @param[in] src_step_z src_stride_z * number of elements along Z processed per workitem(in bytes)
+ * @param[in] src_offset_first_element_in_bytes The offset of the first element in the source tensor
+ * @param[out] dst_ptr Pointer to the destination tensor. Supported data types: same as @p src_ptr
+ * @param[in] dst_stride_x Stride of the destination tensor in X dimension (in bytes)
+ * @param[in] dst_step_x dst_stride_x * number of elements along X processed per workitem(in bytes)
+ * @param[in] dst_stride_y Stride of the destination tensor in Y dimension (in bytes)
+ * @param[in] dst_step_y dst_stride_y * number of elements along Y processed per workitem(in bytes)
+ * @param[in] dst_offset_first_element_in_bytes The offset of the first element in the destination tensor
+ * @param[in] src_stride_w Stride of the source tensor in W dimension (in bytes).
+ * @param[in] dst_stride_w Stride of the destination tensor in W dimension (in bytes).
+ */
+__kernel void im2col_generic_dchw(
+ TENSOR3D_DECLARATION(src),
+ IMAGE_DECLARATION(dst),
+ uint src_stride_w,
+ uint dst_stride_w)
+{
+ const int xc = get_global_id(0); // x coordinate in the convolved tensor
+ const int yc = get_global_id(1); // y coordinate in the convolved tensor
+ const int ch = get_global_id(2) % KERNEL_DEPTH; // input feature map
+ const int batch = get_global_id(2) / KERNEL_DEPTH; // batch size
+
+ // Calculate input indices
+ const int xi = xc * STRIDE_X - PAD_LEFT;
+ const int yi = yc * STRIDE_Y - PAD_TOP;
+
+ // Calculate output indices
+ const int xo = ch * KERNEL_WIDTH * KERNEL_HEIGHT;
+ const int yo = xc + yc * CONVOLVED_WIDTH; // Index of the convolution
+
+ __global uchar *input_ptr = src_ptr + src_offset_first_element_in_bytes + ch * src_stride_z + batch * src_stride_w;
+ __global DATA_TYPE *output_ptr = ((__global DATA_TYPE *)(dst_ptr + dst_offset_first_element_in_bytes + yo * dst_stride_y + batch * dst_stride_w)) + xo;
+
+ // Linearize convolution elements
+ for(int y = yi, y_e = yi + KERNEL_HEIGHT; y < y_e; ++y)
+ {
+ for(int x = xi, x_e = xi + KERNEL_WIDTH; x < x_e; ++x, ++output_ptr)
+ {
+#if PAD_LEFT == 0 && PAD_TOP == 0 && PAD_RIGHT == 0 && PAD_BOTTOM == 0
+ *output_ptr = *((__global DATA_TYPE *)(input_ptr + x * src_stride_x + y * src_stride_y));
+#else // PAD_LEFT == 0 && PAD_TOP == 0 && PAD_RIGHT == 0 && PAD_BOTTOM == 0
+ if(x < 0 || x >= SRC_WIDTH || y < 0 || y >= SRC_HEIGHT)
+ {
+ *output_ptr = PAD_VALUE;
+ }
+ else
+ {
+ *output_ptr = *((__global DATA_TYPE *)(input_ptr + x * src_stride_x + y * src_stride_y));
+ }
+#endif // PAD_LEFT == 0 && PAD_TOP == 0 && PAD_RIGHT == 0 && PAD_BOTTOM == 0
+ }
+ }
+
+#ifdef HAS_BIAS
+ if(ch == (KERNEL_DEPTH - 1))
+ {
+#ifdef FIXED_POINT_POSITION
+ *output_ptr = (DATA_TYPE)(1 << FIXED_POINT_POSITION);
+#else // FIXED_POINT_POSITION
+ *output_ptr = 1.0f;
+#endif // FIXED_POINT_POSITION
+ }
+#endif // HAS_BIAS
+}
+#endif // defined(CONVOLVED_WIDTH) && defined(SRC_WIDTH) && defined(SRC_HEIGHT) && defined(STRIDE_X) && defined(STRIDE_Y) && defined(KERNEL_WIDTH) && defined(KERNEL_HEIGHT) && defined(KERNEL_DEPTH) && defined(PAD_LEFT) && defined(PAD_RIGHT) && defined(PAD_TOP) && defined(PAD_BOTTOM) && defined(PAD_VALUE)
+
+/**This kernel reshapes the input tensor to a tensor used to perform convolution using GEMM when
+ * the kernel width and height are the same of width and height of the input tensor
+ *
+ * @note Datatype should be given as a preprocessor argument using -DDATA_TYPE=type. e.g. -DDATA_TYPE=float
+ * @note In case biases will be added in late stage, -DHAS_BIAS has to be passed to append the final matrix with 1 in each row.
+ *
+ * @param[in] src_ptr Pointer to the source tensor. Supported data types: QS8/QASYMM8/QS16/F16/F32
+ * @param[in] src_stride_x Stride of the source tensor in X dimension (in bytes)
+ * @param[in] src_step_x src_stride_x * number of elements along X processed per workitem(in bytes)
+ * @param[in] src_stride_y Stride of the source tensor in Y dimension (in bytes)
+ * @param[in] src_step_y src_stride_y * number of elements along Y processed per workitem(in bytes)
+ * @param[in] src_stride_z Stride of the source tensor in Z dimension (in bytes)
+ * @param[in] src_step_z src_stride_z * number of elements along Y processed per workitem(in bytes)
+ * @param[in] src_offset_first_element_in_bytes The offset of the first element in the source tensor
+ * @param[out] dst_ptr Pointer to the destination tensor. Same as @p src_ptr
+ * @param[in] dst_stride_x Stride of the destination tensor in X dimension (in bytes)
+ * @param[in] dst_step_x dst_stride_x * number of elements along X processed per workitem(in bytes)
+ * @param[in] dst_offset_first_element_in_bytes The offset of the first element in the destination tensor
+ * @param[in] width The width of the input tensor
+ * @param[in] height The height of the input tensor
+ */
+__kernel void im2col_reduced_dchw(
+ TENSOR3D_DECLARATION(src),
+ VECTOR_DECLARATION(dst),
+ uint width, uint height)
+{
+ Tensor3D src = CONVERT_TO_TENSOR3D_STRUCT(src);
+
+ const uint image_size = width * height;
+
+ __global uchar *tmp_out_ptr = dst_ptr + dst_offset_first_element_in_bytes + (get_global_id(0) + get_global_id(1) * width + get_global_id(2) * image_size) * dst_stride_x;
+
+ *((__global DATA_TYPE *)tmp_out_ptr) = *((__global DATA_TYPE *)src.ptr);
+
+#ifdef HAS_BIAS
+ // If it is the last thread in the 3 dimensional workgroup
+ if(get_global_id(0) == (get_global_size(0) - 1) && get_global_id(1) == (get_global_size(1) - 1) && get_global_id(2) == (get_global_size(2) - 1))
+ {
+ tmp_out_ptr += dst_stride_x;
+#ifdef FIXED_POINT_POSITION
+ *((__global DATA_TYPE *)tmp_out_ptr) = (DATA_TYPE)(1 << FIXED_POINT_POSITION);
+#else // FIXED_POINT_POSITION
+ *((__global DATA_TYPE *)tmp_out_ptr) = (DATA_TYPE)1.0f;
+#endif // FIXED_POINT_POSITION
+ }
+#endif // HAS_BIAS
+}
+#endif // defined(DATA_TYPE) && defined(ELEMENT_SIZE)
\ No newline at end of file
diff --git a/src/core/CL/kernels/CLIm2ColKernel.cpp b/src/core/CL/kernels/CLIm2ColKernel.cpp
index 4f69318..d1fc503 100644
--- a/src/core/CL/kernels/CLIm2ColKernel.cpp
+++ b/src/core/CL/kernels/CLIm2ColKernel.cpp
@@ -59,7 +59,7 @@
} // namespace
CLIm2ColKernel::CLIm2ColKernel()
- : _input(nullptr), _output(nullptr), _convolved_dims(), _num_elems_processed_per_iteration(1), _run_func(nullptr)
+ : _input(nullptr), _output(nullptr), _convolved_dims(), _num_elems_processed_per_iteration(1), _run_func(nullptr), _kernel_dims()
{
}
@@ -70,8 +70,9 @@
// Perform validation step
ARM_COMPUTE_ERROR_THROW_ON(validate_arguments(input->info(), output->info(), has_bias));
- _input = input;
- _output = output;
+ _input = input;
+ _output = output;
+ _kernel_dims = kernel_dims;
const DataType data_type = input->info()->data_type();
const GPUTarget gpu_target = get_arch_from_target(get_target());
@@ -79,6 +80,7 @@
// Create kernel
CLBuildOptions build_opts;
build_opts.add_option(("-DDATA_TYPE=" + get_cl_type_from_data_type(data_type)));
+ build_opts.add_option("-DELEMENT_SIZE=" + support::cpp11::to_string(input->info()->element_size()));
build_opts.add_option_if(has_bias, "-DHAS_BIAS");
build_opts.add_option_if(is_data_type_fixed_point(data_type), "-DFIXED_POINT_POSITION=" + support::cpp11::to_string(input->info()->fixed_point_position()));
@@ -93,13 +95,19 @@
output->info()->tensor_shape().cbegin() + 1))
&& ((stride_x == 1) && (stride_y == 1) && !conv_info.has_padding());
- std::string kernel_name = "im2col_generic";
+ bool is_optimized_path = false;
+
+ _num_elems_processed_per_iteration = 1;
+
+ std::string kernel_name;
if(!run_img2col_reduced)
{
+ // Default kernel name
+ kernel_name = "im2col_generic_dchw";
+
_convolved_dims = scaled_dimensions(input->info()->dimension(0), input->info()->dimension(1),
kernel_dims.width, kernel_dims.height,
conv_info);
- _num_elems_processed_per_iteration = output->info()->dimension(0);
build_opts.add_option("-DKERNEL_WIDTH=" + support::cpp11::to_string(kernel_dims.width));
build_opts.add_option("-DKERNEL_HEIGHT=" + support::cpp11::to_string(kernel_dims.height));
@@ -116,19 +124,50 @@
build_opts.add_option("-DSRC_HEIGHT=" + support::cpp11::to_string(input->info()->dimension(1)));
build_opts.add_option_if_else(is_data_type_quantized(data_type), "-DPAD_VALUE=" + support::cpp11::to_string(input->info()->quantization_info().offset), "-DPAD_VALUE=0");
- if(kernel_dims.width == 3 && kernel_dims.height == 3 && !conv_info.has_padding())
- {
- kernel_name = "im2col_kernel3x3_padx0_pady0";
+ const bool squared_im2col = kernel_dims.width == kernel_dims.height;
- // Local work size optimized for the 3x3 MobileNets convolution on Bifrost.
- if(gpu_target == GPUTarget::BIFROST && input->info()->dimension(0) == 224)
+ if(squared_im2col && !is_data_type_fixed_point(data_type))
+ {
+ // Check if we can run an optimized im2col
+ switch(kernel_dims.width)
{
- _lws_hint = cl::NDRange(2, 3, 3);
+ case 1:
+ // Optimized im2col1x1 if stride_x = 1 and conv_info.has_padding() = false
+ if(conv_info.stride().first == 1 && !conv_info.has_padding())
+ {
+ _num_elems_processed_per_iteration = 4;
+ is_optimized_path = true;
+ kernel_name = "im2col1x1_stridex1_dchw";
+ }
+ break;
+ case 3:
+ _num_elems_processed_per_iteration = 1;
+ is_optimized_path = true;
+ kernel_name = "im2col3x3_dchw";
+ break;
+ case 5:
+ _num_elems_processed_per_iteration = 1;
+ is_optimized_path = true;
+ kernel_name = "im2col5x5_dchw";
+ break;
+ case 11:
+ // Optimized im2col11x11 if pad_x = pad_y = 0
+ if(!conv_info.has_padding())
+ {
+ _num_elems_processed_per_iteration = 1;
+ is_optimized_path = true;
+ kernel_name = "im2col11x11_padx0_pady0_dchw";
+ }
+ break;
+ default:
+ is_optimized_path = false;
+ break;
}
}
else if(kernel_dims.width > 1 && !conv_info.has_padding())
{
- kernel_name = "im2col_generic_padx0_pady0";
+ _num_elems_processed_per_iteration = 1;
+ kernel_name = "im2col_generic_padx0_pady0_dchw";
// Optimized im2col is performed using one or more vector operations with the specified vector size
// and a remainder. For example, for 5x5 convolutions, im2col is performed using vectors of size 4
@@ -152,30 +191,12 @@
build_opts.add_option("-DVECTOR_SIZE=" + support::cpp11::to_string(vector_size));
build_opts.add_option("-DWIDTH_MOD_VECTOR_SIZE=" + support::cpp11::to_string(width_mod_vector_size));
}
- else
- {
- if(gpu_target == GPUTarget::BIFROST)
- {
- const size_t input_channels = input->info()->dimension(2);
- if((input_channels & (input_channels - 1)) == 0)
- {
- // input_channels is a power of two
- _lws_hint = cl::NDRange(1, 1, 4);
- }
- else if(input_channels < 192 && (input_channels % 4) == 0)
- {
- // input_channels is less than 192 and is a multiple of 4
- _lws_hint = cl::NDRange(1, 1, 2);
- }
- // otherwise the default is optimal
- }
- }
_run_func = &CLIm2ColKernel::run_generic;
}
else
{
- kernel_name = "im2col_reduced";
_num_elems_processed_per_iteration = 1;
+ kernel_name = "im2col_reduced_dchw";
_run_func = &CLIm2ColKernel::run_reduced;
}
@@ -183,8 +204,30 @@
_kernel = static_cast<cl::Kernel>(CLKernelLibrary::get().create_kernel(kernel_name, build_opts.options()));
// Configure kernel window
- Window win = calculate_max_window(*input->info(), Steps());
- // The CLIm2ColKernel doesn't need padding so update_window_and_padding() can be skipped
+ Window win;
+ if(is_optimized_path)
+ {
+ win = calculate_max_window(*input->info(),
+ Steps(_num_elems_processed_per_iteration),
+ false,
+ BorderSize(conv_info.pad_top(), conv_info.pad_right(), conv_info.pad_bottom(), conv_info.pad_left()));
+
+ const int x = -conv_info.pad_left();
+ const int y = -conv_info.pad_top();
+ const int w = kernel_dims.width * _num_elems_processed_per_iteration;
+ const int h = kernel_dims.height;
+
+ AccessWindowRectangle input_access(input->info(), x, y, w, h);
+
+ update_window_and_padding(win, input_access);
+ }
+ else
+ {
+ // For the generic case, CLIm2ColKernel doesn't need padding (we do not read out-of-bounds elements) so
+ // update_window_and_padding() can be skipped
+ win = calculate_max_window(*input->info(), Steps());
+ }
+
output->info()->set_valid_region(ValidRegion(Coordinates(), output->info()->tensor_shape()));
if(!run_img2col_reduced)
{
@@ -195,8 +238,8 @@
ICLKernel::configure(win);
// Set config_id for enabling LWS tuning
- _config_id = "im2col_";
- _config_id += (run_img2col_reduced ? "reduced_" : "");
+ _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(output->info()->dimension(0));
@@ -233,9 +276,15 @@
Window slice_in = window_collapsed.first_slice_window_3D();
Window slice_out = window_collapsed.first_slice_window_3D();
- // Setup slice
- slice.set(Window::DimX, Window::Dimension(0, static_cast<int>(_convolved_dims.first), 1));
- slice.set(Window::DimY, Window::Dimension(0, static_cast<int>(_convolved_dims.second), 1));
+ // Setup slice if stride_x != 0 or stride_y != 0
+ if(_convolved_dims.first != _input->info()->dimension(0) || _convolved_dims.second != _input->info()->dimension(1))
+ {
+ // If the stride_x or stride_y are not 1, the output tensor of matrix multiply (Convolved tensor) will not
+ // have the same shape of the im2col input tensor
+ // In this case we need to re-compute the window using the shape of the tensor after matrix multiply (convolved_dims)
+ slice.set(Window::DimX, Window::Dimension(0, static_cast<int>(_convolved_dims.first), 1));
+ slice.set(Window::DimY, Window::Dimension(0, static_cast<int>(_convolved_dims.second), 1));
+ }
// Setup input slice
// The first three dimensions of the input are increased by the inner loops
@@ -244,7 +293,7 @@
slice_in.set(Window::DimZ, Window::Dimension(0, 0, 0));
// Setup output slice
- slice_out.set(Window::DimX, Window::Dimension(0, _output->info()->dimension(0), _num_elems_processed_per_iteration));
+ slice_out.set(Window::DimX, Window::Dimension(0, _output->info()->dimension(0), _kernel_dims.area()));
slice_out.set(Window::DimY, Window::Dimension(0, _output->info()->dimension(1), 1));
slice_out.set(Window::DimZ, Window::Dimension(0, 1, 1));