enable concat layer fuse for OCL target

Signed-off-by: Li Peng <peng.li@intel.com>

diff --git a/modules/dnn/src/dnn.cpp b/modules/dnn/src/dnn.cpp
index 011631f..db90593 100644 (file)
--- a/modules/dnn/src/dnn.cpp
+++ b/modules/dnn/src/dnn.cpp
@@ -1492,7 +1492,8 @@ struct Net::Impl
             // TODO: OpenCL target support more fusion styles.
             if ( preferableBackend == DNN_BACKEND_OPENCV && IS_DNN_OPENCL_TARGET(preferableTarget) &&
                  (!cv::ocl::useOpenCL() || (ld.layerInstance->type != "Convolution" &&
-                 ld.layerInstance->type != "MVN" && ld.layerInstance->type != "Pooling")) )
+                 ld.layerInstance->type != "MVN" && ld.layerInstance->type != "Pooling" &&
+                 ld.layerInstance->type != "Concat")) )
                 continue;
 
             Ptr<Layer>& currLayer = ld.layerInstance;
@@ -1701,6 +1702,31 @@ struct Net::Impl
                 ld.outputBlobs.size() == 1 )
             {
                 Mat& output = ld.outputBlobs[0];
+                UMat umat_output;
+                if (!ld.outputBlobsWrappers.empty() &&
+                    (preferableBackend == DNN_BACKEND_OPENCV && IS_DNN_OPENCL_TARGET(preferableTarget)))
+                {
+                    size_t i, ninputs = ld.inputBlobsId.size();
+                    bool conv_layer = true;
+                    for( i = 0; i < ninputs; i++ )
+                    {
+                        LayerPin pin = ld.inputBlobsId[i];
+                        LayerData* inp_i_data = &layers[pin.lid];
+                        while(inp_i_data->skip &&
+                              inp_i_data->inputBlobsId.size() == 1 &&
+                              inp_i_data->consumers.size() == 1)
+                        {
+                            pin = inp_i_data->inputBlobsId[0];
+                            inp_i_data = &layers[pin.lid];
+                        }
+                        conv_layer = conv_layer && (inp_i_data->getLayerInstance()->type == "Convolution");
+                    }
+                    if (!conv_layer)
+                        continue;
+                    std::vector<UMat> umat_outputBlobs;
+                    umat_outputBlobs = OpenCLBackendWrapper::getUMatVector(ld.outputBlobsWrappers);
+                    umat_output = umat_outputBlobs[0];
+                }
 
                 // TODO: in general, this optimization can always be done, but
                 // many layers currently check that the input/output blobs are
@@ -1737,6 +1763,14 @@ struct Net::Impl
                         // Allocate new memory to prevent collisions during memory
                         // reusing (see https://github.com/opencv/opencv/pull/10456).
                         output = output.clone();
+                        if (preferableBackend == DNN_BACKEND_OPENCV &&
+                            IS_DNN_OPENCL_TARGET(preferableTarget))
+                        {
+                            std::vector<UMat> umats(1);
+                            umat_output = umat_output.clone();
+                            umats[0] = umat_output;
+                            OpenCLBackendWrapper::update(ld.outputBlobsWrappers, umats);
+                        }
                         Range chrange[] = { Range::all(), Range::all(), Range::all(), Range::all() };
                         int ofs = 0;
                         for( i = 0; i < ninputs; i++ )
@@ -1753,6 +1787,12 @@ struct Net::Impl
                             CV_Assert(output_slice.isContinuous() && output_slice.size == curr_output.size);
                             Mat* oldPtr = &curr_output;
                             curr_output = output_slice;
+                            if (preferableBackend == DNN_BACKEND_OPENCV && IS_DNN_OPENCL_TARGET(preferableTarget))
+                            {
+                                std::vector<UMat> umats(inp_i_data->outputBlobsWrappers.size());
+                                umats[pin.oid] = umat_output(chrange);
+                                OpenCLBackendWrapper::update(inp_i_data->outputBlobsWrappers, umats);
+                            }
                             // Layers that refer old input Mat will refer to the
                             // new data but the same Mat object.
                             CV_Assert(curr_output.data == output_slice.data, oldPtr == &curr_output);

diff --git a/modules/dnn/src/ocl4dnn/src/ocl4dnn_conv_spatial.cpp b/modules/dnn/src/ocl4dnn/src/ocl4dnn_conv_spatial.cpp
index a446f3b..c889c7d 100644 (file)
--- a/modules/dnn/src/ocl4dnn/src/ocl4dnn_conv_spatial.cpp
+++ b/modules/dnn/src/ocl4dnn/src/ocl4dnn_conv_spatial.cpp
@@ -821,7 +821,7 @@ void OCL4DNNConvSpatial<float>::CreateSubBuffer(const UMat& buffer, UMat& sub_bu
     cl_int err;
     size_t element_size = (use_half_) ? sizeof(short) : sizeof(float);
 
-    region.origin = offset * element_size;
+    region.origin = offset * element_size + buffer.offset;
     region.size = size * element_size;
     sub_mem = clCreateSubBuffer((cl_mem)buffer.handle(ACCESS_READ),
                                 write_only ? CL_MEM_WRITE_ONLY : CL_MEM_READ_ONLY,
@@ -853,6 +853,7 @@ bool OCL4DNNConvSpatial<float>::convolve(const UMat &bottom, UMat &top,
         return false;
 
     int32_t bias_offset;
+    int32_t element_size = use_half_ ? sizeof(short) : sizeof(float);
 
     if (config->kernelType == KERNEL_TYPE_INTEL_IDLF) {
         if (!swizzleWeight(weight, config->workItem_output[2], false))
@@ -931,10 +932,12 @@ bool OCL4DNNConvSpatial<float>::convolve(const UMat &bottom, UMat &top,
                     return false;
 
                 kernel.set(argIdx++, ocl::KernelArg::PtrWriteOnly(out_buffer));
+                kernel.set(argIdx++, (int)(out_buffer.offset / element_size));
             }
             else
             {
                 kernel.set(argIdx++, ocl::KernelArg::PtrWriteOnly(top));
+                kernel.set(argIdx++, (int)(top.offset / element_size));
             }
 
             kernel.set(argIdx++, (uint16_t)width_);
@@ -1024,10 +1027,12 @@ bool OCL4DNNConvSpatial<float>::convolve(const UMat &bottom, UMat &top,
                     return false;
 
                 kernel.set(argIdx++, ocl::KernelArg::PtrWriteOnly(out_buffer));
+                kernel.set(argIdx++, (int)(out_buffer.offset / element_size));
             }
             else
             {
                 kernel.set(argIdx++, ocl::KernelArg::PtrWriteOnly(top));
+                kernel.set(argIdx++, (int)(top.offset / element_size));
             }
 
             kernel.set(argIdx++, (uint16_t)width_);
@@ -1079,6 +1084,7 @@ bool OCL4DNNConvSpatial<float>::convolve(const UMat &bottom, UMat &top,
         if (bias_term_)
             kernel.set(argIdx++, ocl::KernelArg::PtrReadOnly(bias));
         kernel.set(argIdx++, ocl::KernelArg::PtrWriteOnly(top));
+        kernel.set(argIdx++, (int)(top.offset / element_size));
         kernel.set(argIdx++, (uint16_t)width_);
         kernel.set(argIdx++, (uint16_t)height_);
         kernel.set(argIdx++, (uint16_t)output_w_);
@@ -1126,6 +1132,7 @@ bool OCL4DNNConvSpatial<float>::convolve(const UMat &bottom, UMat &top,
                     kernel.set(argIdx++, (void *)NULL);
                 kernel.set(argIdx++, bias_offset);
                 kernel.set(argIdx++, ocl::KernelArg::PtrWriteOnly(top));
+                kernel.set(argIdx++, (int)(top.offset / element_size));
                 kernel.set(argIdx++, output_image_offset);
                 kernel.set(argIdx++, (uint16_t)width_);
                 kernel.set(argIdx++, (uint16_t)height_);

diff --git a/modules/dnn/src/opencl/conv_layer_spatial.cl b/modules/dnn/src/opencl/conv_layer_spatial.cl
index 2cc161d..adeb385 100644 (file)
--- a/modules/dnn/src/opencl/conv_layer_spatial.cl
+++ b/modules/dnn/src/opencl/conv_layer_spatial.cl
@@ -136,7 +136,8 @@ __kernel void ConvolveBasic(
     int kernel_offset,
     __global Dtype* bias,
     const int bias_offset,
-    __global Dtype* convolved_image,
+    __global Dtype* convolved_image_base,
+    const int convolved_image_base_offset,
     const int convolved_image_offset,
     const ushort input_width,
     const ushort input_height,
@@ -146,6 +147,7 @@ __kernel void ConvolveBasic(
     const ushort pad_h
 )
 {
+    __global Dtype* convolved_image = convolved_image_base + convolved_image_base_offset;
     const int outputX = get_global_id(0);
     const int outputY = get_global_id(1);
     const int kernelNum = get_global_id(2) * ZPAR;
@@ -220,12 +222,14 @@ convolve_simd(
     __global Dtype* inputs,
     __global Dtype* weights,
     BIAS_KERNEL_ARG
-    __global Dtype* outputs,
+    __global Dtype* outputs_base,
+    const int outputs_offset,
     const ushort input_width,
     const ushort input_height,
     const ushort output_width,
     const ushort output_height)
 {
+  __global Dtype* outputs = outputs_base + outputs_offset;
   unsigned int oc = get_global_id(0) * OUT_BLOCK_WIDTH;  // oc = Output Column
   unsigned int or = get_global_id(1) * OUT_BLOCK_HEIGHT; // or = Output Row
   unsigned int fm = get_global_id(2);                    // fm = Feature Map = od = Output Depth
@@ -395,7 +399,8 @@ typedef struct half0 { half s0; } half0; //never used but makes compiler happy.
     const __global Dtype *src0,   \
     const __global Dtype *src1,   \
     BIAS_KERNEL_ARG               \
-    __global Dtype *dst,          \
+    __global Dtype *dst_base,     \
+    const int dst_offset,         \
     const ushort input_width,     \
     const ushort input_height,    \
     const ushort output_width,    \
@@ -425,6 +430,7 @@ typedef struct half0 { half s0; } half0; //never used but makes compiler happy.
 __attribute__((intel_reqd_sub_group_size(8)))
 __kernel void Conv_Interleaved(GEMM_LIKE_KERNEL_ARGS)
 {
+    __global Dtype *dst = dst_base + dst_offset;
     const int group_x = get_group_id(0);
     const int group_y = get_group_id(1);
     const int global_x = get_global_id(0);
@@ -813,6 +819,7 @@ __kernel void Conv_Interleaved(GEMM_LIKE_KERNEL_ARGS)
 __attribute__((intel_reqd_sub_group_size(8)))
 __kernel void Conv_Interleaved(GEMM_LIKE_KERNEL_ARGS)
 {
+    __global Dtype *dst = dst_base + dst_offset;
     const int group_x = get_group_id(0);
     const int group_y = get_group_id(1);
     const int global_x = get_global_id(0);
@@ -1374,6 +1381,7 @@ __kernel void Conv_Interleaved(GEMM_LIKE_KERNEL_ARGS)
 __attribute__((intel_reqd_sub_group_size(16)))
 __kernel void Conv_Interleaved(GEMM_LIKE_KERNEL_ARGS)
 {
+    __global Dtype *dst = dst_base + dst_offset;
     const int group_x = get_group_id(0);
     const int group_y = get_group_id(1);
     const int global_x = get_global_id(0);
@@ -1559,6 +1567,7 @@ __kernel void Conv_Interleaved(GEMM_LIKE_KERNEL_ARGS)
 __attribute__((intel_reqd_sub_group_size(16)))
 __kernel void Conv_Interleaved(GEMM_LIKE_KERNEL_ARGS)
 {
+    __global Dtype *dst = dst_base + dst_offset;
     const int group_x = get_group_id(0);
     const int group_y = get_group_id(1);
     const int global_x = get_global_id(0);
@@ -1770,12 +1779,13 @@ __kernel void DWCONV(
     __global Dtype* image_data,
     __global Dtype* kernel_data,
     BIAS_KERNEL_ARG
-    __global Dtype* convolved_image,
+    __global Dtype* convolved_image_base,
+    const int convolved_image_offset,
     const ushort input_width,
     const ushort input_height,
     const ushort output_width,
     const ushort output_height) {
-
+  __global Dtype* convolved_image = convolved_image_base + convolved_image_offset;
   const int outputX = get_global_id(0);
   const int outputY = get_global_id(1);
   const int outputZ = get_global_id(2);