COMPMID-2451: Use kernel lws_hint() on enqueue

Avoid querying device's default lws on kernel enqueue as this is already
cached in the kernel during configuration.

Change-Id: Ia26ecb712caeb8f042356815e0cfd23522764d27
Signed-off-by: Georgios Pinitas <georgios.pinitas@arm.com>
Reviewed-on: https://review.mlplatform.org/c/1803
Reviewed-by: Michele Di Giorgio <michele.digiorgio@arm.com>
Comments-Addressed: Arm Jenkins <bsgcomp@arm.com>
Tested-by: Arm Jenkins <bsgcomp@arm.com>

56 files changed, 86 insertions, 86 deletions

diff --git a/src/core/CL/kernels/CLAbsoluteDifferenceKernel.cpp b/src/core/CL/kernels/CLAbsoluteDifferenceKernel.cpp
index 0c1206adfb..557046e831 100644
--- a/src/core/CL/kernels/CLAbsoluteDifferenceKernel.cpp
+++ b/src/core/CL/kernels/CLAbsoluteDifferenceKernel.cpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2016-2018 ARM Limited.
+ * Copyright (c) 2016-2019 ARM Limited.
  *
  * SPDX-License-Identifier: MIT
  *
@@ -96,7 +96,7 @@ void CLAbsoluteDifferenceKernel::run(const Window &window, cl::CommandQueue &que
         add_2D_tensor_argument(idx, _input1, slice);
         add_2D_tensor_argument(idx, _input2, slice);
         add_2D_tensor_argument(idx, _output, slice);
-        enqueue(queue, *this, slice);
+        enqueue(queue, *this, slice, lws_hint());
     }
     while(window.slide_window_slice_2D(slice));
 }
diff --git a/src/core/CL/kernels/CLBatchToSpaceLayerKernel.cpp b/src/core/CL/kernels/CLBatchToSpaceLayerKernel.cpp
index aa06d3af8c..075a7b1b3e 100644
--- a/src/core/CL/kernels/CLBatchToSpaceLayerKernel.cpp
+++ b/src/core/CL/kernels/CLBatchToSpaceLayerKernel.cpp
@@ -173,7 +173,7 @@ void CLBatchToSpaceLayerKernel::run(const Window &window, cl::CommandQueue &queu
             add_1D_tensor_argument(idx, _block_shape, vector_slice);
         }
         add_4D_tensor_argument(idx, _output, slice_out);
-        enqueue(queue, *this, slice_in);
+        enqueue(queue, *this, slice_in, lws_hint());
 
         ++batch_id;
     }
diff --git a/src/core/CL/kernels/CLBitwiseAndKernel.cpp b/src/core/CL/kernels/CLBitwiseAndKernel.cpp
index dd301cd02e..2d05f2e2bd 100644
--- a/src/core/CL/kernels/CLBitwiseAndKernel.cpp
+++ b/src/core/CL/kernels/CLBitwiseAndKernel.cpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2016-2018 ARM Limited.
+ * Copyright (c) 2016-2019 ARM Limited.
  *
  * SPDX-License-Identifier: MIT
  *
@@ -82,7 +82,7 @@ void CLBitwiseAndKernel::run(const Window &window, cl::CommandQueue &queue)
         add_2D_tensor_argument(idx, _input1, slice);
         add_2D_tensor_argument(idx, _input2, slice);
         add_2D_tensor_argument(idx, _output, slice);
-        enqueue(queue, *this, slice);
+        enqueue(queue, *this, slice, lws_hint());
     }
     while(window.slide_window_slice_2D(slice));
 }
diff --git a/src/core/CL/kernels/CLBitwiseOrKernel.cpp b/src/core/CL/kernels/CLBitwiseOrKernel.cpp
index aa84618258..b3efab8b1f 100644
--- a/src/core/CL/kernels/CLBitwiseOrKernel.cpp
+++ b/src/core/CL/kernels/CLBitwiseOrKernel.cpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2016-2018 ARM Limited.
+ * Copyright (c) 2016-2019 ARM Limited.
  *
  * SPDX-License-Identifier: MIT
  *
@@ -83,7 +83,7 @@ void CLBitwiseOrKernel::run(const Window &window, cl::CommandQueue &queue)
         add_2D_tensor_argument(idx, _input1, slice);
         add_2D_tensor_argument(idx, _input2, slice);
         add_2D_tensor_argument(idx, _output, slice);
-        enqueue(queue, *this, slice);
+        enqueue(queue, *this, slice, lws_hint());
     }
     while(window.slide_window_slice_2D(slice));
 }
diff --git a/src/core/CL/kernels/CLBitwiseXorKernel.cpp b/src/core/CL/kernels/CLBitwiseXorKernel.cpp
index ad1f923253..d8ac486d0f 100644
--- a/src/core/CL/kernels/CLBitwiseXorKernel.cpp
+++ b/src/core/CL/kernels/CLBitwiseXorKernel.cpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2016-2018 ARM Limited.
+ * Copyright (c) 2016-2019 ARM Limited.
  *
  * SPDX-License-Identifier: MIT
  *
@@ -83,7 +83,7 @@ void CLBitwiseXorKernel::run(const Window &window, cl::CommandQueue &queue)
         add_2D_tensor_argument(idx, _input1, slice);
         add_2D_tensor_argument(idx, _input2, slice);
         add_2D_tensor_argument(idx, _output, slice);
-        enqueue(queue, *this, slice);
+        enqueue(queue, *this, slice, lws_hint());
     }
     while(window.slide_window_slice_2D(slice));
 }
diff --git a/src/core/CL/kernels/CLBoundingBoxTransformKernel.cpp b/src/core/CL/kernels/CLBoundingBoxTransformKernel.cpp
index 7c30a942db..045bd02d73 100644
--- a/src/core/CL/kernels/CLBoundingBoxTransformKernel.cpp
+++ b/src/core/CL/kernels/CLBoundingBoxTransformKernel.cpp
@@ -127,6 +127,6 @@ void CLBoundingBoxTransformKernel::run(const Window &window, cl::CommandQueue &q
     add_2D_tensor_argument(idx, _deltas, slice);
 
     // Note that we don't need to loop over the slices, as we are sure that we are dealing with all 2D tensors
-    enqueue(queue, *this, slice);
+    enqueue(queue, *this, slice, lws_hint());
 }
 } // namespace arm_compute
diff --git a/src/core/CL/kernels/CLChannelCombineKernel.cpp b/src/core/CL/kernels/CLChannelCombineKernel.cpp
index 126614a599..d029efe110 100644
--- a/src/core/CL/kernels/CLChannelCombineKernel.cpp
+++ b/src/core/CL/kernels/CLChannelCombineKernel.cpp
@@ -281,7 +281,7 @@ void CLChannelCombineKernel::run(const Window &window, cl::CommandQueue &queue)
             _kernel.setArg(idx++, slice.y().end());
         }
 
-        enqueue(queue, *this, slice);
+        enqueue(queue, *this, slice, lws_hint());
     }
     while(window.slide_window_slice_2D(slice));
 }
diff --git a/src/core/CL/kernels/CLChannelExtractKernel.cpp b/src/core/CL/kernels/CLChannelExtractKernel.cpp
index 8bddba837a..d2a0f984da 100644
--- a/src/core/CL/kernels/CLChannelExtractKernel.cpp
+++ b/src/core/CL/kernels/CLChannelExtractKernel.cpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2016-2018 ARM Limited.
+ * Copyright (c) 2016-2019 ARM Limited.
  *
  * SPDX-License-Identifier: MIT
  *
@@ -181,7 +181,7 @@ void CLChannelExtractKernel::run(const Window &window, cl::CommandQueue &queue)
         unsigned int idx = 0;
         add_2D_tensor_argument(idx, _input, slice);
         add_2D_tensor_argument(idx, _output, win_sub);
-        enqueue(queue, *this, slice);
+        enqueue(queue, *this, slice, lws_hint());
     }
     while(window.slide_window_slice_2D(slice));
 }
diff --git a/src/core/CL/kernels/CLConvertFullyConnectedWeightsKernel.cpp b/src/core/CL/kernels/CLConvertFullyConnectedWeightsKernel.cpp
index ace3fd5840..81856769b2 100644
--- a/src/core/CL/kernels/CLConvertFullyConnectedWeightsKernel.cpp
+++ b/src/core/CL/kernels/CLConvertFullyConnectedWeightsKernel.cpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2018 ARM Limited.
+ * Copyright (c) 2018-2019 ARM Limited.
  *
  * SPDX-License-Identifier: MIT
  *
@@ -107,6 +107,6 @@ void CLConvertFullyConnectedWeightsKernel::run(const Window &window, cl::Command
     unsigned int idx = 0;
     add_2D_tensor_argument(idx, _input, window);
     add_2D_tensor_argument(idx, _output, window);
-    enqueue(queue, *this, window);
+    enqueue(queue, *this, window, lws_hint());
 }
 } // namespace arm_compute
diff --git a/src/core/CL/kernels/CLCopyKernel.cpp b/src/core/CL/kernels/CLCopyKernel.cpp
index c87768aeb7..c49251ba45 100644
--- a/src/core/CL/kernels/CLCopyKernel.cpp
+++ b/src/core/CL/kernels/CLCopyKernel.cpp
@@ -259,7 +259,7 @@ void CLCopyKernel::run(const Window &window, cl::CommandQueue &queue)
             unsigned int idx = 0;
             add_3D_tensor_argument(idx, _input, slice);
             add_3D_tensor_argument(idx, _output, out_slice);
-            enqueue(queue, *this, slice);
+            enqueue(queue, *this, slice, lws_hint());
         }
         while(window.slide_window_slice_3D(slice) && _output_window.slide_window_slice_3D(out_slice));
     }
@@ -272,7 +272,7 @@ void CLCopyKernel::run(const Window &window, cl::CommandQueue &queue)
             unsigned int idx = 0;
             add_3D_tensor_argument(idx, _input, slice);
             add_3D_tensor_argument(idx, _output, slice);
-            enqueue(queue, *this, slice);
+            enqueue(queue, *this, slice, lws_hint());
         }
         while(collapsed.slide_window_slice_3D(slice));
     }
diff --git a/src/core/CL/kernels/CLCropKernel.cpp b/src/core/CL/kernels/CLCropKernel.cpp
index f8a2456d4a..90e054b047 100644
--- a/src/core/CL/kernels/CLCropKernel.cpp
+++ b/src/core/CL/kernels/CLCropKernel.cpp
@@ -127,6 +127,6 @@ void CLCropKernel::run(const Window &window, cl::CommandQueue &queue)
     add_3D_tensor_argument(idx, _output, window);
     add_argument(idx, _start.x);
     add_argument(idx, _start.y);
-    enqueue(queue, *this, window);
+    enqueue(queue, *this, window, lws_hint());
 }
 } // namespace arm_compute
diff --git a/src/core/CL/kernels/CLDeconvolutionLayerUpsampleKernel.cpp b/src/core/CL/kernels/CLDeconvolutionLayerUpsampleKernel.cpp
index 295fb5c997..50f654680c 100644
--- a/src/core/CL/kernels/CLDeconvolutionLayerUpsampleKernel.cpp
+++ b/src/core/CL/kernels/CLDeconvolutionLayerUpsampleKernel.cpp
@@ -129,7 +129,7 @@ void CLDeconvolutionLayerUpsampleKernel::run(const Window &window, cl::CommandQu
                 unsigned int idx = 0;
                 add_3D_tensor_argument(idx, _input, slice_in);
                 add_3D_tensor_argument(idx, _output, slice_out);
-                enqueue(queue, *this, slice_out);
+                enqueue(queue, *this, slice_out, lws_hint());
             }
             while(collapsed.slide_window_slice_3D(slice_in) && collapsed.slide_window_slice_3D(slice_out));
             break;
@@ -148,7 +148,7 @@ void CLDeconvolutionLayerUpsampleKernel::run(const Window &window, cl::CommandQu
                 unsigned int idx = 0;
                 add_3D_tensor_argument(idx, _input, slice_in);
                 add_3D_tensor_argument(idx, _output, slice_out);
-                enqueue(queue, *this, slice_out);
+                enqueue(queue, *this, slice_out, lws_hint());
             }
             while(window.slide_window_slice_3D(slice_in) && window.slide_window_slice_3D(slice_out));
             break;
diff --git a/src/core/CL/kernels/CLDepthConcatenateLayerKernel.cpp b/src/core/CL/kernels/CLDepthConcatenateLayerKernel.cpp
index 2d8de755b3..c985811fbc 100644
--- a/src/core/CL/kernels/CLDepthConcatenateLayerKernel.cpp
+++ b/src/core/CL/kernels/CLDepthConcatenateLayerKernel.cpp
@@ -147,7 +147,7 @@ void CLDepthConcatenateLayerKernel::run(const Window &window, cl::CommandQueue &
         unsigned int idx = 0;
         add_3D_tensor_argument(idx, _input, slice);
         add_3D_tensor_argument(idx, _output, slice);
-        enqueue(queue, *this, slice);
+        enqueue(queue, *this, slice, lws_hint());
     }
     while(window.slide_window_slice_3D(slice));
 }
diff --git a/src/core/CL/kernels/CLDepthToSpaceLayerKernel.cpp b/src/core/CL/kernels/CLDepthToSpaceLayerKernel.cpp
index f2d65e813c..49a5590231 100644
--- a/src/core/CL/kernels/CLDepthToSpaceLayerKernel.cpp
+++ b/src/core/CL/kernels/CLDepthToSpaceLayerKernel.cpp
@@ -119,7 +119,7 @@ void CLDepthToSpaceLayerKernel::run(const Window &window, cl::CommandQueue &queu
         add_3D_tensor_argument(idx, _input, slice_in);
         add_argument(idx, batch_id);
         add_4D_tensor_argument(idx, _output, slice_out);
-        enqueue(queue, *this, slice_in);
+        enqueue(queue, *this, slice_in, lws_hint());
 
         ++batch_id;
     }
diff --git a/src/core/CL/kernels/CLDepthwiseConvolutionLayerReshapeWeightsGenericKernel.cpp b/src/core/CL/kernels/CLDepthwiseConvolutionLayerReshapeWeightsGenericKernel.cpp
index b73ccf5ddc..582c600c61 100644
--- a/src/core/CL/kernels/CLDepthwiseConvolutionLayerReshapeWeightsGenericKernel.cpp
+++ b/src/core/CL/kernels/CLDepthwiseConvolutionLayerReshapeWeightsGenericKernel.cpp
@@ -141,7 +141,7 @@ void CLDepthwiseConvolutionLayerReshapeWeightsGenericKernel::run(const Window &w
         unsigned int idx = 0;
         add_3D_tensor_argument(idx, _input, slice);
         add_2D_tensor_argument(idx, _output, slice_out);
-        enqueue(queue, *this, slice);
+        enqueue(queue, *this, slice, lws_hint());
     }
     while(window.slide_window_slice_3D(slice) && window.slide_window_slice_2D(slice_out));
 }
diff --git a/src/core/CL/kernels/CLDepthwiseConvolutionLayerReshapeWeightsKernel.cpp b/src/core/CL/kernels/CLDepthwiseConvolutionLayerReshapeWeightsKernel.cpp
index 6b6438abef..1fd6312295 100644
--- a/src/core/CL/kernels/CLDepthwiseConvolutionLayerReshapeWeightsKernel.cpp
+++ b/src/core/CL/kernels/CLDepthwiseConvolutionLayerReshapeWeightsKernel.cpp
@@ -121,6 +121,6 @@ void CLDepthwiseConvolutionLayerReshapeWeightsKernel::run(const Window &window,
     unsigned int idx = 0;
     add_3D_tensor_argument(idx, _input, window);
     add_2D_tensor_argument(idx, _output, window);
-    enqueue(queue, *this, window);
+    enqueue(queue, *this, window, lws_hint());
 }
 } // namespace arm_compute
diff --git a/src/core/CL/kernels/CLDepthwiseVectorToTensorKernel.cpp b/src/core/CL/kernels/CLDepthwiseVectorToTensorKernel.cpp
index 2dad72944e..0f029fda74 100644
--- a/src/core/CL/kernels/CLDepthwiseVectorToTensorKernel.cpp
+++ b/src/core/CL/kernels/CLDepthwiseVectorToTensorKernel.cpp
@@ -119,7 +119,7 @@ void CLDepthwiseVectorToTensorKernel::run(const Window &window, cl::CommandQueue
         unsigned int idx = 0;
         add_1D_tensor_argument(idx, _input, slice);
         add_3D_tensor_argument(idx, _output, slice_out);
-        enqueue(queue, *this, slice);
+        enqueue(queue, *this, slice, lws_hint());
     }
     while(window.slide_window_slice_1D(slice) && window.slide_window_slice_3D(slice_out));
 }
diff --git a/src/core/CL/kernels/CLDequantizationLayerKernel.cpp b/src/core/CL/kernels/CLDequantizationLayerKernel.cpp
index 12d36cdb9f..10a2878be7 100644
--- a/src/core/CL/kernels/CLDequantizationLayerKernel.cpp
+++ b/src/core/CL/kernels/CLDequantizationLayerKernel.cpp
@@ -131,7 +131,7 @@ void CLDequantizationLayerKernel::run(const Window &window, cl::CommandQueue &qu
         unsigned int idx = 0;
         add_3D_tensor_argument(idx, _input, slice);
         add_3D_tensor_argument(idx, _output, slice);
-        enqueue(queue, *this, slice);
+        enqueue(queue, *this, slice, lws_hint());
     }
     while(window_collapsed.slide_window_slice_3D(slice));
 }
diff --git a/src/core/CL/kernels/CLElementWiseUnaryLayerKernel.cpp b/src/core/CL/kernels/CLElementWiseUnaryLayerKernel.cpp
index 03edcef781..c4ab50411a 100644
--- a/src/core/CL/kernels/CLElementWiseUnaryLayerKernel.cpp
+++ b/src/core/CL/kernels/CLElementWiseUnaryLayerKernel.cpp
@@ -128,7 +128,7 @@ void CLElementWiseUnaryLayerKernel::run(const Window &window, cl::CommandQueue &
         unsigned int idx = 0;
         add_1D_tensor_argument(idx, _input, collapsed);
         add_1D_tensor_argument(idx, _output, collapsed);
-        enqueue(queue, *this, collapsed);
+        enqueue(queue, *this, collapsed, lws_hint());
     }
     while(window.slide_window_slice_1D(collapsed));
 }
diff --git a/src/core/CL/kernels/CLFillBorderKernel.cpp b/src/core/CL/kernels/CLFillBorderKernel.cpp
index c63b3be9af..475a52b3db 100644
--- a/src/core/CL/kernels/CLFillBorderKernel.cpp
+++ b/src/core/CL/kernels/CLFillBorderKernel.cpp
@@ -180,7 +180,7 @@ void CLFillBorderKernel::run(const Window &window, cl::CommandQueue &queue)
     {
         unsigned int idx = 0;
         add_3D_tensor_argument(idx, _tensor, slice);
-        enqueue(queue, *this, slice, cl::NullRange);
+        enqueue(queue, *this, slice, lws_hint());
     }
     while(collapsed.slide_window_slice_3D(slice));
 }
diff --git a/src/core/CL/kernels/CLFloorKernel.cpp b/src/core/CL/kernels/CLFloorKernel.cpp
index 831173d076..7c39d87512 100644
--- a/src/core/CL/kernels/CLFloorKernel.cpp
+++ b/src/core/CL/kernels/CLFloorKernel.cpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2017-2018 ARM Limited.
+ * Copyright (c) 2017-2019 ARM Limited.
  *
  * SPDX-License-Identifier: MIT
  *
@@ -124,7 +124,7 @@ void CLFloorKernel::run(const Window &window, cl::CommandQueue &queue)
         unsigned int idx = 0;
         add_3D_tensor_argument(idx, _input, slice);
         add_3D_tensor_argument(idx, _output, slice);
-        enqueue(queue, *this, slice);
+        enqueue(queue, *this, slice, lws_hint());
     }
     while(collapsed.slide_window_slice_3D(slice));
 }
diff --git a/src/core/CL/kernels/CLFuseBatchNormalizationKernel.cpp b/src/core/CL/kernels/CLFuseBatchNormalizationKernel.cpp
index bf827bf2c2..423e583059 100644
--- a/src/core/CL/kernels/CLFuseBatchNormalizationKernel.cpp
+++ b/src/core/CL/kernels/CLFuseBatchNormalizationKernel.cpp
@@ -209,6 +209,6 @@ void CLFuseBatchNormalizationKernel::run(const arm_compute::Window &window, cl::
     {
         add_1D_tensor_argument(idx, _bn_gamma, slice_1d);
     }
-    enqueue(queue, *this, slice_3d);
+    enqueue(queue, *this, slice_3d, lws_hint());
 }
 } // namespace arm_compute
diff --git a/src/core/CL/kernels/CLGEMMLowpQuantizeDownInt32ToInt16ScaleByFixedPointKernel.cpp b/src/core/CL/kernels/CLGEMMLowpQuantizeDownInt32ToInt16ScaleByFixedPointKernel.cpp
index 557e82dc50..8720123366 100644
--- a/src/core/CL/kernels/CLGEMMLowpQuantizeDownInt32ToInt16ScaleByFixedPointKernel.cpp
+++ b/src/core/CL/kernels/CLGEMMLowpQuantizeDownInt32ToInt16ScaleByFixedPointKernel.cpp
@@ -174,7 +174,7 @@ void CLGEMMLowpQuantizeDownInt32ToInt16ScaleByFixedPointKernel::run(const Window
         unsigned int idx = 0;
         add_3D_tensor_argument(idx, _input, slice);
         add_3D_tensor_argument(idx1, _output, slice);
-        enqueue(queue, *this, slice);
+        enqueue(queue, *this, slice, lws_hint());
     }
     while(collapsed.slide_window_slice_3D(slice));
 }
diff --git a/src/core/CL/kernels/CLGEMMLowpQuantizeDownInt32ToUint8ScaleByFixedPointKernel.cpp b/src/core/CL/kernels/CLGEMMLowpQuantizeDownInt32ToUint8ScaleByFixedPointKernel.cpp
index b7eff0f8ec..7c066381ce 100644
--- a/src/core/CL/kernels/CLGEMMLowpQuantizeDownInt32ToUint8ScaleByFixedPointKernel.cpp
+++ b/src/core/CL/kernels/CLGEMMLowpQuantizeDownInt32ToUint8ScaleByFixedPointKernel.cpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2017-2018 ARM Limited.
+ * Copyright (c) 2017-2019 ARM Limited.
  *
  * SPDX-License-Identifier: MIT
  *
@@ -176,7 +176,7 @@ void CLGEMMLowpQuantizeDownInt32ToUint8ScaleByFixedPointKernel::run(const Window
         unsigned int idx = 0;
         add_3D_tensor_argument(idx, _input, slice);
         add_3D_tensor_argument(idx1, _output, slice);
-        enqueue(queue, *this, slice);
+        enqueue(queue, *this, slice, lws_hint());
     }
     while(collapsed.slide_window_slice_3D(slice));
 }
diff --git a/src/core/CL/kernels/CLGEMMLowpQuantizeDownInt32ToUint8ScaleByFloatKernel.cpp b/src/core/CL/kernels/CLGEMMLowpQuantizeDownInt32ToUint8ScaleByFloatKernel.cpp
index b7730d5060..2967a73866 100644
--- a/src/core/CL/kernels/CLGEMMLowpQuantizeDownInt32ToUint8ScaleByFloatKernel.cpp
+++ b/src/core/CL/kernels/CLGEMMLowpQuantizeDownInt32ToUint8ScaleByFloatKernel.cpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2018 ARM Limited.
+ * Copyright (c) 2018-2019 ARM Limited.
  *
  * SPDX-License-Identifier: MIT
  *
@@ -173,7 +173,7 @@ void CLGEMMLowpQuantizeDownInt32ToUint8ScaleByFloatKernel::run(const Window &win
         unsigned int idx = 0;
         add_3D_tensor_argument(idx, _input, slice);
         add_3D_tensor_argument(idx1, _output, slice);
-        enqueue(queue, *this, slice);
+        enqueue(queue, *this, slice, lws_hint());
     }
     while(collapsed.slide_window_slice_3D(slice));
 }
diff --git a/src/core/CL/kernels/CLGEMMLowpQuantizeDownInt32ToUint8ScaleKernel.cpp b/src/core/CL/kernels/CLGEMMLowpQuantizeDownInt32ToUint8ScaleKernel.cpp
index 621bd2b54b..8175f60275 100644
--- a/src/core/CL/kernels/CLGEMMLowpQuantizeDownInt32ToUint8ScaleKernel.cpp
+++ b/src/core/CL/kernels/CLGEMMLowpQuantizeDownInt32ToUint8ScaleKernel.cpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2017-2018 ARM Limited.
+ * Copyright (c) 2017-2019 ARM Limited.
  *
  * SPDX-License-Identifier: MIT
  *
@@ -170,7 +170,7 @@ void CLGEMMLowpQuantizeDownInt32ToUint8ScaleKernel::run(const Window &window, cl
         unsigned int idx = 0;
         add_3D_tensor_argument(idx, _input, slice);
         add_3D_tensor_argument(idx1, _output, slice);
-        enqueue(queue, *this, slice);
+        enqueue(queue, *this, slice, lws_hint());
     }
     while(collapsed.slide_window_slice_3D(slice));
 }
diff --git a/src/core/CL/kernels/CLGEMMLowpReductionKernel.cpp b/src/core/CL/kernels/CLGEMMLowpReductionKernel.cpp
index 0c0b0ec817..3a59b43823 100644
--- a/src/core/CL/kernels/CLGEMMLowpReductionKernel.cpp
+++ b/src/core/CL/kernels/CLGEMMLowpReductionKernel.cpp
@@ -214,7 +214,7 @@ void CLGEMMLowpMatrixBReductionKernel::run(const Window &window, cl::CommandQueu
         unsigned int idx = 0;
         add_3D_tensor_argument(idx, _input, slice_in);
         add_2D_tensor_argument(idx, _output, slice_out);
-        enqueue(queue, *this, slice_out);
+        enqueue(queue, *this, slice_out, lws_hint());
     }
     while(collapsed.slide_window_slice_2D(slice_out));
 }
diff --git a/src/core/CL/kernels/CLGEMMReshapeRHSMatrixKernel.cpp b/src/core/CL/kernels/CLGEMMReshapeRHSMatrixKernel.cpp
index 5b9e68d2bd..2ca4132b15 100644
--- a/src/core/CL/kernels/CLGEMMReshapeRHSMatrixKernel.cpp
+++ b/src/core/CL/kernels/CLGEMMReshapeRHSMatrixKernel.cpp
@@ -166,7 +166,7 @@ void CLGEMMReshapeRHSMatrixKernel::run(const Window &window, cl::CommandQueue &q
         unsigned int idx = 0;
         add_3D_tensor_argument(idx, _input, slice);
         add_3D_tensor_argument(idx, _output, slice);
-        enqueue(queue, *this, slice);
+        enqueue(queue, *this, slice, lws_hint());
     }
     while(window.slide_window_slice_3D(slice));
 }
\ No newline at end of file
diff --git a/src/core/CL/kernels/CLGatherKernel.cpp b/src/core/CL/kernels/CLGatherKernel.cpp
index 412821b8bd..4e91cf2cc9 100644
--- a/src/core/CL/kernels/CLGatherKernel.cpp
+++ b/src/core/CL/kernels/CLGatherKernel.cpp
@@ -131,6 +131,6 @@ void CLGatherKernel::run(const Window &window, cl::CommandQueue &queue)
     add_4D_tensor_argument(idx, _input, window_collapsed);
     add_1D_tensor_argument(idx, _indices, window_collapsed);
     add_4D_tensor_argument(idx, _output, window_collapsed);
-    enqueue(queue, *this, window_collapsed);
+    enqueue(queue, *this, window_collapsed, lws_hint());
 }
 } // namespace arm_compute
diff --git a/src/core/CL/kernels/CLGenerateProposalsLayerKernel.cpp b/src/core/CL/kernels/CLGenerateProposalsLayerKernel.cpp
index ab95ddca54..79e364caf7 100644
--- a/src/core/CL/kernels/CLGenerateProposalsLayerKernel.cpp
+++ b/src/core/CL/kernels/CLGenerateProposalsLayerKernel.cpp
@@ -124,6 +124,6 @@ void CLComputeAllAnchorsKernel::run(const Window &window, cl::CommandQueue &queu
 
     // Note that we don't need to loop over the slices, as we are launching exactly
     // as many threads as all the anchors generated
-    enqueue(queue, *this, collapsed);
+    enqueue(queue, *this, collapsed, lws_hint());
 }
 } // namespace arm_compute
diff --git a/src/core/CL/kernels/CLHeightConcatenateLayerKernel.cpp b/src/core/CL/kernels/CLHeightConcatenateLayerKernel.cpp
index 70af541ef1..ea292c0b7b 100644
--- a/src/core/CL/kernels/CLHeightConcatenateLayerKernel.cpp
+++ b/src/core/CL/kernels/CLHeightConcatenateLayerKernel.cpp
@@ -162,5 +162,5 @@ void CLHeightConcatenateLayerKernel::run(const Window &window, cl::CommandQueue
     unsigned int idx = 0;
     add_4D_tensor_argument(idx, _input, window);
     add_4D_tensor_argument(idx, _output, window);
-    enqueue(queue, *this, window);
+    enqueue(queue, *this, window, lws_hint());
 }
diff --git a/src/core/CL/kernels/CLL2NormalizeLayerKernel.cpp b/src/core/CL/kernels/CLL2NormalizeLayerKernel.cpp
index 8816138e2e..2ef977bebf 100644
--- a/src/core/CL/kernels/CLL2NormalizeLayerKernel.cpp
+++ b/src/core/CL/kernels/CLL2NormalizeLayerKernel.cpp
@@ -178,7 +178,7 @@ void CLL2NormalizeLayerKernel::run(const Window &window, cl::CommandQueue &queue
                 add_2D_tensor_argument(idx, _input, in_slice);
                 add_2D_tensor_argument(idx, _sum, sum_slice);
                 add_2D_tensor_argument(idx, _output, in_slice);
-                enqueue(queue, *this, in_slice);
+                enqueue(queue, *this, in_slice, lws_hint());
             }
             while(window.slide_window_slice_2D(in_slice) && window.slide_window_slice_2D(sum_slice));
         }
@@ -194,7 +194,7 @@ void CLL2NormalizeLayerKernel::run(const Window &window, cl::CommandQueue &queue
                 add_2D_tensor_argument(idx, _input, in_slice);
                 add_2D_tensor_argument(idx, _sum, sum_slice);
                 add_2D_tensor_argument(idx, _output, in_slice);
-                enqueue(queue, *this, in_slice);
+                enqueue(queue, *this, in_slice, lws_hint());
             }
             while(window.slide_window_slice_2D(in_slice) && window.slide_window_slice_2D(sum_slice));
         }
@@ -210,7 +210,7 @@ void CLL2NormalizeLayerKernel::run(const Window &window, cl::CommandQueue &queue
                 add_3D_tensor_argument(idx, _input, in_slice);
                 add_3D_tensor_argument(idx, _sum, sum_slice);
                 add_3D_tensor_argument(idx, _output, in_slice);
-                enqueue(queue, *this, in_slice);
+                enqueue(queue, *this, in_slice, lws_hint());
             }
             while(window.slide_window_slice_3D(in_slice) && window.slide_window_slice_3D(sum_slice));
         }
diff --git a/src/core/CL/kernels/CLLKTrackerKernel.cpp b/src/core/CL/kernels/CLLKTrackerKernel.cpp
index 40ed630c89..68a210c115 100644
--- a/src/core/CL/kernels/CLLKTrackerKernel.cpp
+++ b/src/core/CL/kernels/CLLKTrackerKernel.cpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2017-2018 ARM Limited.
+ * Copyright (c) 2017-2019 ARM Limited.
  *
  * SPDX-License-Identifier: MIT
  *
@@ -83,7 +83,7 @@ void CLLKTrackerInitKernel::run(const Window &window, cl::CommandQueue &queue)
     ARM_COMPUTE_ERROR_ON_UNCONFIGURED_KERNEL(this);
     ARM_COMPUTE_ERROR_ON_INVALID_SUBWINDOW(ICLKernel::window(), window);
 
-    enqueue(queue, *this, window);
+    enqueue(queue, *this, window, lws_hint());
 }
 
 void CLLKTrackerFinalizeKernel::configure(ICLLKInternalKeypointArray *new_points_internal, ICLKeyPointArray *new_points)
@@ -112,7 +112,7 @@ void CLLKTrackerFinalizeKernel::run(const Window &window, cl::CommandQueue &queu
     ARM_COMPUTE_ERROR_ON_UNCONFIGURED_KERNEL(this);
     ARM_COMPUTE_ERROR_ON_INVALID_SUBWINDOW(ICLKernel::window(), window);
 
-    enqueue(queue, *this, window);
+    enqueue(queue, *this, window, lws_hint());
 }
 
 CLLKTrackerStage0Kernel::CLLKTrackerStage0Kernel()
@@ -202,7 +202,7 @@ void CLLKTrackerStage0Kernel::run(const Window &window, cl::CommandQueue &queue)
     add_2D_tensor_argument(idx, _old_scharr_gx, window);
     add_2D_tensor_argument(idx, _old_scharr_gy, window);
 
-    enqueue(queue, *this, window);
+    enqueue(queue, *this, window, lws_hint());
 }
 
 CLLKTrackerStage1Kernel::CLLKTrackerStage1Kernel()
@@ -284,5 +284,5 @@ void CLLKTrackerStage1Kernel::run(const Window &window, cl::CommandQueue &queue)
     unsigned int idx = 0;
     add_2D_tensor_argument(idx, _new_input, window);
 
-    enqueue(queue, *this, window);
+    enqueue(queue, *this, window, lws_hint());
 }
diff --git a/src/core/CL/kernels/CLMeanStdDevKernel.cpp b/src/core/CL/kernels/CLMeanStdDevKernel.cpp
index 0cde9c5fe6..7bfd6d6e53 100644
--- a/src/core/CL/kernels/CLMeanStdDevKernel.cpp
+++ b/src/core/CL/kernels/CLMeanStdDevKernel.cpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2016-2018 ARM Limited.
+ * Copyright (c) 2016-2019 ARM Limited.
  *
  * SPDX-License-Identifier: MIT
  *
@@ -132,7 +132,7 @@ void CLMeanStdDevKernel::run(const Window &window, cl::CommandQueue &queue)
         // Set slice step equal to height to force gws[1] to 1,
         // as each thread calculates the sum across all rows and columns equal to the number of elements processed by each work-item
         slice.set_dimension_step(Window::DimY, _input->info()->dimension(1));
-        enqueue(queue, *this, slice);
+        enqueue(queue, *this, slice, lws_hint());
     }
     while(window.slide_window_slice_2D(slice));
 
diff --git a/src/core/CL/kernels/CLMemsetKernel.cpp b/src/core/CL/kernels/CLMemsetKernel.cpp
index 80caf9406e..142642097f 100644
--- a/src/core/CL/kernels/CLMemsetKernel.cpp
+++ b/src/core/CL/kernels/CLMemsetKernel.cpp
@@ -103,7 +103,7 @@ void CLMemsetKernel::run(const Window &window, cl::CommandQueue &queue)
     {
         unsigned int idx = 0;
         add_3D_tensor_argument(idx, _tensor, slice);
-        enqueue(queue, *this, slice);
+        enqueue(queue, *this, slice, lws_hint());
     }
     while(collapsed.slide_window_slice_3D(slice));
 }
diff --git a/src/core/CL/kernels/CLMinMaxLayerKernel.cpp b/src/core/CL/kernels/CLMinMaxLayerKernel.cpp
index 92b5f8d505..05912743b3 100644
--- a/src/core/CL/kernels/CLMinMaxLayerKernel.cpp
+++ b/src/core/CL/kernels/CLMinMaxLayerKernel.cpp
@@ -157,7 +157,7 @@ void CLMinMaxLayerKernel::run(const Window &window, cl::CommandQueue &queue)
         // Set inputs
         add_3D_tensor_argument(idx, _input, slice);
         add_1D_tensor_argument(idx, _output, output_slice);
-        enqueue(queue, *this, slice);
+        enqueue(queue, *this, slice, lws_hint());
     }
     while(window_collapsed.slide_window_slice_3D(slice));
 }
diff --git a/src/core/CL/kernels/CLMinMaxLocationKernel.cpp b/src/core/CL/kernels/CLMinMaxLocationKernel.cpp
index 0c7f3bc070..e865b45e73 100644
--- a/src/core/CL/kernels/CLMinMaxLocationKernel.cpp
+++ b/src/core/CL/kernels/CLMinMaxLocationKernel.cpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2017-2018 ARM Limited.
+ * Copyright (c) 2017-2019 ARM Limited.
  *
  * SPDX-License-Identifier: MIT
  *
@@ -134,7 +134,7 @@ void CLMinMaxKernel::run(const Window &window, cl::CommandQueue &queue)
     {
         unsigned int idx = 0;
         add_2D_tensor_argument(idx, _input, slice);
-        enqueue(queue, *this, slice);
+        enqueue(queue, *this, slice, lws_hint());
     }
     while(window.slide_window_slice_2D(slice));
 
@@ -226,7 +226,7 @@ void CLMinMaxLocationKernel::run(const Window &window, cl::CommandQueue &queue)
     {
         unsigned int idx = 0;
         add_2D_tensor_argument(idx, _input, slice);
-        enqueue(queue, *this, slice);
+        enqueue(queue, *this, slice, lws_hint());
     }
     while(window.slide_window_slice_2D(slice));
 }
diff --git a/src/core/CL/kernels/CLPadLayerKernel.cpp b/src/core/CL/kernels/CLPadLayerKernel.cpp
index d35164d309..5270e31bbe 100644
--- a/src/core/CL/kernels/CLPadLayerKernel.cpp
+++ b/src/core/CL/kernels/CLPadLayerKernel.cpp
@@ -136,7 +136,7 @@ void CLPadLayerKernel::run(const Window &window, cl::CommandQueue &queue)
         add_3D_tensor_argument(idx, _input, slice_in);
         add_3D_tensor_argument(idx, _output, slice_out);
 
-        enqueue(queue, *this, slice_out);
+        enqueue(queue, *this, slice_out, lws_hint());
     }
     while(window.slide_window_slice_3D(slice_out) && win_in.slide_window_slice_3D(slice_in));
 }
diff --git a/src/core/CL/kernels/CLPermuteKernel.cpp b/src/core/CL/kernels/CLPermuteKernel.cpp
index a5fc1a705f..9cb72b3c04 100644
--- a/src/core/CL/kernels/CLPermuteKernel.cpp
+++ b/src/core/CL/kernels/CLPermuteKernel.cpp
@@ -141,7 +141,7 @@ void CLPermuteKernel::run(const Window &window, cl::CommandQueue &queue)
         unsigned int idx = 0;
         add_4D_tensor_argument(idx, _input, slice_in);
         add_4D_tensor_argument(idx, _output, slice_out);
-        enqueue(queue, *this, slice_in);
+        enqueue(queue, *this, slice_in, lws_hint());
     }
     while(window.slide_window_slice_4D(slice_in) && window.slide_window_slice_4D(slice_out));
 }
diff --git a/src/core/CL/kernels/CLPixelWiseMultiplicationKernel.cpp b/src/core/CL/kernels/CLPixelWiseMultiplicationKernel.cpp
index 76403d18d4..50cdc9c7f4 100644
--- a/src/core/CL/kernels/CLPixelWiseMultiplicationKernel.cpp
+++ b/src/core/CL/kernels/CLPixelWiseMultiplicationKernel.cpp
@@ -274,7 +274,7 @@ void CLPixelWiseMultiplicationKernel::run(const Window &window, cl::CommandQueue
         add_3D_tensor_argument(idx, _input1, slice_input1);
         add_3D_tensor_argument(idx, _input2, slice_input2);
         add_3D_tensor_argument(idx, _output, slice);
-        enqueue(queue, *this, slice);
+        enqueue(queue, *this, slice, lws_hint());
 
         ARM_COMPUTE_UNUSED(collapsed.slide_window_slice_3D(slice_input1));
         ARM_COMPUTE_UNUSED(collapsed.slide_window_slice_3D(slice_input2));
@@ -409,7 +409,7 @@ void CLComplexPixelWiseMultiplicationKernel::run(const Window &window, cl::Comma
         add_3D_tensor_argument(idx, _input1, slice_input1);
         add_3D_tensor_argument(idx, _input2, slice_input2);
         add_3D_tensor_argument(idx, _output, slice);
-        enqueue(queue, *this, slice);
+        enqueue(queue, *this, slice, lws_hint());
 
         ARM_COMPUTE_UNUSED(collapsed.slide_window_slice_3D(slice_input1));
         ARM_COMPUTE_UNUSED(collapsed.slide_window_slice_3D(slice_input2));
diff --git a/src/core/CL/kernels/CLPriorBoxLayerKernel.cpp b/src/core/CL/kernels/CLPriorBoxLayerKernel.cpp
index c76d839551..12ea19394b 100644
--- a/src/core/CL/kernels/CLPriorBoxLayerKernel.cpp
+++ b/src/core/CL/kernels/CLPriorBoxLayerKernel.cpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2018 ARM Limited.
+ * Copyright (c) 2018-2019 ARM Limited.
  *
  * SPDX-License-Identifier: MIT
  *
@@ -214,6 +214,6 @@ void CLPriorBoxLayerKernel::run(const Window &window, cl::CommandQueue &queue)
 
     unsigned int idx = 0;
     add_2D_tensor_argument(idx, _output, slice);
-    enqueue(queue, *this, slice);
+    enqueue(queue, *this, slice, lws_hint());
 }
 } // namespace arm_compute
diff --git a/src/core/CL/kernels/CLQuantizationLayerKernel.cpp b/src/core/CL/kernels/CLQuantizationLayerKernel.cpp
index 22d4e3345f..493255f1cd 100644
--- a/src/core/CL/kernels/CLQuantizationLayerKernel.cpp
+++ b/src/core/CL/kernels/CLQuantizationLayerKernel.cpp
@@ -126,7 +126,7 @@ void CLQuantizationLayerKernel::run(const Window &window, cl::CommandQueue &queu
         unsigned int idx = 0;
         add_3D_tensor_argument(idx, _input, slice);
         add_3D_tensor_argument(idx, _output, slice);
-        enqueue(queue, *this, slice);
+        enqueue(queue, *this, slice, lws_hint());
     }
     while(window_collapsed.slide_window_slice_3D(slice));
 }
diff --git a/src/core/CL/kernels/CLROIAlignLayerKernel.cpp b/src/core/CL/kernels/CLROIAlignLayerKernel.cpp
index 47dc62c4f4..50729f2421 100644
--- a/src/core/CL/kernels/CLROIAlignLayerKernel.cpp
+++ b/src/core/CL/kernels/CLROIAlignLayerKernel.cpp
@@ -149,6 +149,6 @@ void CLROIAlignLayerKernel::run(const Window &window, cl::CommandQueue &queue)
     add_argument<cl_uint>(idx, _input->info()->strides_in_bytes()[3]);
     add_argument<cl_uint>(idx, _output->info()->strides_in_bytes()[3]);
 
-    enqueue(queue, *this, slice);
+    enqueue(queue, *this, slice, lws_hint());
 }
 } // namespace arm_compute
diff --git a/src/core/CL/kernels/CLROIPoolingLayerKernel.cpp b/src/core/CL/kernels/CLROIPoolingLayerKernel.cpp
index c32ec1bb2b..9858a4de36 100644
--- a/src/core/CL/kernels/CLROIPoolingLayerKernel.cpp
+++ b/src/core/CL/kernels/CLROIPoolingLayerKernel.cpp
@@ -143,6 +143,6 @@ void CLROIPoolingLayerKernel::run(const Window &window, cl::CommandQueue &queue)
     add_argument<cl_uint>(idx, _input->info()->strides_in_bytes()[3]);
     add_argument<cl_uint>(idx, _output->info()->strides_in_bytes()[3]);
 
-    enqueue(queue, *this, slice);
+    enqueue(queue, *this, slice, lws_hint());
 }
 } // namespace arm_compute
diff --git a/src/core/CL/kernels/CLReductionOperationKernel.cpp b/src/core/CL/kernels/CLReductionOperationKernel.cpp
index 9db8ae6cde..ee33a72792 100644
--- a/src/core/CL/kernels/CLReductionOperationKernel.cpp
+++ b/src/core/CL/kernels/CLReductionOperationKernel.cpp
@@ -285,7 +285,7 @@ void CLReductionOperationKernel::run(const Window &window, cl::CommandQueue &que
                     unsigned int idx = 0;
                     add_1D_tensor_argument(idx, _input, in_slice);
                     add_1D_tensor_argument(idx, _output, out_slice);
-                    enqueue(queue, *this, in_slice);
+                    enqueue(queue, *this, in_slice, lws_hint());
                 }
                 while(window_in.slide_window_slice_1D(in_slice) && window.slide_window_slice_1D(out_slice));
             }
@@ -331,7 +331,7 @@ void CLReductionOperationKernel::run(const Window &window, cl::CommandQueue &que
                 unsigned int idx = 0;
                 add_2D_tensor_argument(idx, _input, in_slice);
                 add_2D_tensor_argument(idx, _output, out_slice);
-                enqueue(queue, *this, in_slice);
+                enqueue(queue, *this, in_slice, lws_hint());
             }
             while(window_in.slide_window_slice_2D(in_slice) && window.slide_window_slice_2D(out_slice));
         }
@@ -349,7 +349,7 @@ void CLReductionOperationKernel::run(const Window &window, cl::CommandQueue &que
                 unsigned int idx = 0;
                 add_3D_tensor_argument(idx, _input, in_slice);
                 add_3D_tensor_argument(idx, _output, out_slice);
-                enqueue(queue, *this, in_slice);
+                enqueue(queue, *this, in_slice, lws_hint());
             }
             while(window_in.slide_window_slice_3D(in_slice) && window.slide_window_slice_3D(out_slice));
         }
@@ -367,7 +367,7 @@ void CLReductionOperationKernel::run(const Window &window, cl::CommandQueue &que
                 unsigned int idx = 0;
                 add_4D_tensor_argument(idx, _input, in_slice);
                 add_4D_tensor_argument(idx, _output, out_slice);
-                enqueue(queue, *this, in_slice);
+                enqueue(queue, *this, in_slice, lws_hint());
             }
             while(window_in.slide_window_slice_4D(in_slice) && window.slide_window_slice_4D(out_slice));
         }
diff --git a/src/core/CL/kernels/CLRemapKernel.cpp b/src/core/CL/kernels/CLRemapKernel.cpp
index 33c5f2d402..12161fcd70 100644
--- a/src/core/CL/kernels/CLRemapKernel.cpp
+++ b/src/core/CL/kernels/CLRemapKernel.cpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2017-2018 ARM Limited.
+ * Copyright (c) 2017-2019 ARM Limited.
  *
  * SPDX-License-Identifier: MIT
  *
@@ -105,7 +105,7 @@ void CLRemapKernel::run(const Window &window, cl::CommandQueue &queue)
         add_2D_tensor_argument(idx, _output, slice);
         add_2D_tensor_argument(idx, _map_x, slice);
         add_2D_tensor_argument(idx, _map_y, slice);
-        enqueue(queue, *this, slice);
+        enqueue(queue, *this, slice, lws_hint());
     }
     while(window.slide_window_slice_2D(slice));
 }
diff --git a/src/core/CL/kernels/CLReshapeLayerKernel.cpp b/src/core/CL/kernels/CLReshapeLayerKernel.cpp
index aa1339dd1e..040e442845 100644
--- a/src/core/CL/kernels/CLReshapeLayerKernel.cpp
+++ b/src/core/CL/kernels/CLReshapeLayerKernel.cpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2017-2018 ARM Limited.
+ * Copyright (c) 2017-2019 ARM Limited.
  *
  * SPDX-License-Identifier: MIT
  *
@@ -122,6 +122,6 @@ void CLReshapeLayerKernel::run(const Window &window, cl::CommandQueue &queue)
     unsigned int idx = 0;
     add_3D_tensor_argument(idx, _input, window_collapsed);
     add_3D_tensor_argument(idx, _output, window_collapsed);
-    enqueue(queue, *this, slice);
+    enqueue(queue, *this, slice, lws_hint());
 }
 /** [CLReshapeLayerKernel Kernel] **/
diff --git a/src/core/CL/kernels/CLScharr3x3Kernel.cpp b/src/core/CL/kernels/CLScharr3x3Kernel.cpp
index 1c1fedca15..94b0d38c52 100644
--- a/src/core/CL/kernels/CLScharr3x3Kernel.cpp
+++ b/src/core/CL/kernels/CLScharr3x3Kernel.cpp
@@ -118,7 +118,7 @@ void CLScharr3x3Kernel::run(const Window &window, cl::CommandQueue &queue)
         add_2D_tensor_argument_if((_run_scharr_x), idx, _output_x, slice);
         add_2D_tensor_argument_if((_run_scharr_y), idx, _output_y, slice);
 
-        enqueue(queue, *this, slice);
+        enqueue(queue, *this, slice, lws_hint());
     }
     while(window.slide_window_slice_2D(slice));
 }
diff --git a/src/core/CL/kernels/CLSpaceToBatchLayerKernel.cpp b/src/core/CL/kernels/CLSpaceToBatchLayerKernel.cpp
index d2f676b00c..a3441cd5bf 100644
--- a/src/core/CL/kernels/CLSpaceToBatchLayerKernel.cpp
+++ b/src/core/CL/kernels/CLSpaceToBatchLayerKernel.cpp
@@ -201,7 +201,7 @@ void CLSpaceToBatchLayerKernel::run(const Window &window, cl::CommandQueue &queu
 
         add_argument(idx, batch_id);
         add_3D_tensor_argument(idx, _output, slice_out);
-        enqueue(queue, *this, slice_out);
+        enqueue(queue, *this, slice_out, lws_hint());
         ++batch_id;
     }
     while(window.slide_window_slice_3D(slice_out));
diff --git a/src/core/CL/kernels/CLSpaceToDepthLayerKernel.cpp b/src/core/CL/kernels/CLSpaceToDepthLayerKernel.cpp
index 57d8305e47..20454b32a8 100644
--- a/src/core/CL/kernels/CLSpaceToDepthLayerKernel.cpp
+++ b/src/core/CL/kernels/CLSpaceToDepthLayerKernel.cpp
@@ -120,7 +120,7 @@ void CLSpaceToDepthLayerKernel::run(const Window &window, cl::CommandQueue &queu
         add_4D_tensor_argument(idx, _input, slice_in);
         add_argument(idx, batch_id);
         add_3D_tensor_argument(idx, _output, slice_out);
-        enqueue(queue, *this, slice_out);
+        enqueue(queue, *this, slice_out, lws_hint());
 
         ++batch_id;
     }
diff --git a/src/core/CL/kernels/CLStackLayerKernel.cpp b/src/core/CL/kernels/CLStackLayerKernel.cpp
index ccbe1fc016..d5bbae6d4d 100644
--- a/src/core/CL/kernels/CLStackLayerKernel.cpp
+++ b/src/core/CL/kernels/CLStackLayerKernel.cpp
@@ -132,5 +132,5 @@ void CLStackLayerKernel::run(const Window &window, cl::CommandQueue &queue)
     unsigned int idx = 0;
     add_4D_tensor_argument(idx, _input, slice_in);
     add_4D_tensor_argument(idx, _output, slice_out);
-    enqueue(queue, *this, slice_in);
+    enqueue(queue, *this, slice_in, lws_hint());
 }
diff --git a/src/core/CL/kernels/CLTileKernel.cpp b/src/core/CL/kernels/CLTileKernel.cpp
index 7559e7ae72..bcd1bdc5f9 100644
--- a/src/core/CL/kernels/CLTileKernel.cpp
+++ b/src/core/CL/kernels/CLTileKernel.cpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2018 ARM Limited.
+ * Copyright (c) 2018-2019 ARM Limited.
  *
  * SPDX-License-Identifier: MIT
  *
@@ -145,7 +145,7 @@ void CLTileKernel::run(const Window &window, cl::CommandQueue &queue)
         unsigned int idx = 0;
         add_4D_tensor_argument(idx, _input, slice);
         add_4D_tensor_argument(idx, _output, slice);
-        enqueue(queue, *this, slice);
+        enqueue(queue, *this, slice, lws_hint());
     }
     while(collapsed.slide_window_slice_4D(slice));
 }
diff --git a/src/core/CL/kernels/CLUpsampleLayerKernel.cpp b/src/core/CL/kernels/CLUpsampleLayerKernel.cpp
index ce5ed86332..331b02d2e8 100644
--- a/src/core/CL/kernels/CLUpsampleLayerKernel.cpp
+++ b/src/core/CL/kernels/CLUpsampleLayerKernel.cpp
@@ -160,7 +160,7 @@ void CLUpsampleLayerKernel::run(const Window &window, cl::CommandQueue &queue)
         unsigned int idx = 0;
         add_3D_tensor_argument(idx, _input, slice_in);
         add_3D_tensor_argument(idx, _output, slice_out);
-        enqueue(queue, *this, slice_out);
+        enqueue(queue, *this, slice_out, lws_hint());
     }
     while(collapsed_window.slide_window_slice_3D(slice_out) && collapsed_window.slide_window_slice_3D(slice_in));
 }
diff --git a/src/core/CL/kernels/CLWeightsReshapeKernel.cpp b/src/core/CL/kernels/CLWeightsReshapeKernel.cpp
index 590af4d184..9330b3b8a1 100644
--- a/src/core/CL/kernels/CLWeightsReshapeKernel.cpp
+++ b/src/core/CL/kernels/CLWeightsReshapeKernel.cpp
@@ -156,7 +156,7 @@ void CLWeightsReshapeKernel::run(const Window &window, cl::CommandQueue &queue)
         }
 
         // Run kernel
-        enqueue(queue, *this, in_slice);
+        enqueue(queue, *this, in_slice, lws_hint());
     }
     while(window.slide_window_slice_4D(in_slice) && out_window.slide_window_slice_2D(out_slice));
 }
diff --git a/src/core/CL/kernels/CLWidthConcatenateLayerKernel.cpp b/src/core/CL/kernels/CLWidthConcatenateLayerKernel.cpp
index 88cdde7935..92ffc9fa90 100644
--- a/src/core/CL/kernels/CLWidthConcatenateLayerKernel.cpp
+++ b/src/core/CL/kernels/CLWidthConcatenateLayerKernel.cpp
@@ -137,6 +137,6 @@ void CLWidthConcatenateLayerKernel::run(const Window &window, cl::CommandQueue &
     unsigned int idx = 0;
     add_4D_tensor_argument(idx, _input, window);
     add_4D_tensor_argument(idx, _output, window);
-    enqueue(queue, *this, window);
+    enqueue(queue, *this, window, lws_hint());
 }
 } // namespace arm_compute
diff --git a/src/core/CL/kernels/CLWinogradFilterTransformKernel.cpp b/src/core/CL/kernels/CLWinogradFilterTransformKernel.cpp
index 55cc465292..5a508e4b3a 100644
--- a/src/core/CL/kernels/CLWinogradFilterTransformKernel.cpp
+++ b/src/core/CL/kernels/CLWinogradFilterTransformKernel.cpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2018 ARM Limited.
+ * Copyright (c) 2018-2019 ARM Limited.
  *
  * SPDX-License-Identifier: MIT
  *
@@ -150,5 +150,5 @@ void CLWinogradFilterTransformKernel::run(const Window &window, cl::CommandQueue
     unsigned int idx = 0;
     add_4D_tensor_argument(idx, _input, window);
     add_3D_tensor_argument(idx, _output, window_out);
-    enqueue(queue, *this, window);
+    enqueue(queue, *this, window, lws_hint());
 }