dnn(opencl): fix convolution kernel w/o bias with activation

diff --git a/modules/dnn/src/ocl4dnn/src/ocl4dnn_conv_spatial.cpp b/modules/dnn/src/ocl4dnn/src/ocl4dnn_conv_spatial.cpp
index 3cca52d..3707a31 100644 (file)
--- a/modules/dnn/src/ocl4dnn/src/ocl4dnn_conv_spatial.cpp
+++ b/modules/dnn/src/ocl4dnn/src/ocl4dnn_conv_spatial.cpp
@@ -607,6 +607,7 @@ void OCL4DNNConvSpatial<Dtype>::calculateBenchmark(const UMat &bottom, UMat &ver
 {
     options_.str(""); options_.clear(); // clear contents and state flags
     createBasicKernel(1, 1, 1);
+    CV_Assert(!kernelQueue.empty());  // basic kernel must be available
     kernel_index_ = kernelQueue.size() - 1;
     convolve(bottom, verifyTop, weight, bias, numImages, kernelQueue[kernel_index_]);
     CV_Assert(phash.find(kernelQueue[kernel_index_]->kernelName) != phash.end());
@@ -1713,6 +1714,7 @@ void OCL4DNNConvSpatial<float>::useFirstAvailable(const UMat &bottom,
                                     tunerItems[i]->blockHeight,
                                     tunerItems[i]->blockDepth))
         {
+            CV_Assert(!kernelQueue.empty());  // basic kernel must be available
             int kernelIdx = kernelQueue.size() - 1;
             kernelConfig* config = kernelQueue[kernelIdx].get();
             bool failed = false;
@@ -1883,6 +1885,7 @@ void OCL4DNNConvSpatial<float>::setupConvolution(const UMat &bottom,
         CV_LOG_INFO(NULL, "fallback to basic kernel");
         options_.str(""); options_.clear(); // clear contents and state flags
         createBasicKernel(1, 1, 1);
+        CV_Assert(!kernelQueue.empty());  // basic kernel must be available
         kernel_index_ = kernelQueue.size() - 1;
     }
     this->bestKernelConfig = kernelQueue[kernel_index_];

diff --git a/modules/dnn/src/opencl/conv_layer_spatial.cl b/modules/dnn/src/opencl/conv_layer_spatial.cl
index 5d4d6f3..236e8d0 100644 (file)
--- a/modules/dnn/src/opencl/conv_layer_spatial.cl
+++ b/modules/dnn/src/opencl/conv_layer_spatial.cl
@@ -205,7 +205,7 @@ __kernel void ConvolveBasic(
 #if APPLY_BIAS
                 ACTIVATION_FUNCTION(convolved_image, offset, sum[kern] + bias[biasIndex + kern], biasIndex + kern);
 #else
-                ACTIVATION_FUNCTION(convolved_image, offset, sum[kern], biasIndex + kern);
+                ACTIVATION_FUNCTION(convolved_image, offset, sum[kern], kernelNum + kern);
 #endif
             }
         }

diff --git a/modules/dnn/test/test_layers.cpp b/modules/dnn/test/test_layers.cpp
index f9b0f62..085e5a5 100644 (file)
--- a/modules/dnn/test/test_layers.cpp
+++ b/modules/dnn/test/test_layers.cpp
@@ -2223,13 +2223,6 @@ TEST_P(ConvolutionActivationFusion, Accuracy)
     if (actType == "Power" && backendId == DNN_BACKEND_OPENCV && (targetId == DNN_TARGET_OPENCL || targetId == DNN_TARGET_OPENCL_FP16))
         applyTestTag(CV_TEST_TAG_DNN_SKIP_OPENCL);
 
-    // bug: https://github.com/opencv/opencv/issues/17953
-    if (actType == "ChannelsPReLU" && bias_term == false &&
-        backendId == DNN_BACKEND_OPENCV && (targetId == DNN_TARGET_OPENCL || targetId == DNN_TARGET_OPENCL_FP16))
-    {
-        applyTestTag(CV_TEST_TAG_DNN_SKIP_OPENCL);
-    }
-
     Net net;
     int convId = net.addLayer(convParams.name, convParams.type, convParams);
     int activId = net.addLayerToPrev(activationParams.name, activationParams.type, activationParams);