Remove padding from ClScaleKernel

- Merge quantized kernels with fp for bilinear interpolation (both NCHW and NHWC)
- Pass dimensions at compile time rather than at run time
- Use tile-based approach to rework the NCHW kernels
- Remove unused functions/files

Resolve COMPMID-4723

Signed-off-by: Giorgio Arena <giorgio.arena@arm.com>
Change-Id: Ifcdf02beb9daa9f318395751b3c85eb2fe874082
Reviewed-on: https://review.mlplatform.org/c/ml/ComputeLibrary/+/6138
Tested-by: Arm Jenkins <bsgcomp@arm.com>
Comments-Addressed: Arm Jenkins <bsgcomp@arm.com>
Reviewed-by: Michele Di Giorgio <michele.digiorgio@arm.com>

1 files changed, 0 insertions, 4 deletions

diff --git a/src/runtime/gpu/cl/operators/ClScale.h b/src/runtime/gpu/cl/operators/ClScale.h
index 905c43a41c..0ff78640f7 100644
--- a/src/runtime/gpu/cl/operators/ClScale.h
+++ b/src/runtime/gpu/cl/operators/ClScale.h
@@ -34,7 +34,6 @@ namespace opencl
 {
 /** Basic function to simulate a scale layer. This function calls the following OpenCL kernels:
  *
- * -# @ref CLFillBorderKernel (executed if padding size is different from zero)
  * -# @ref kernels::ClScaleKernel
  */
 class ClScale : public IClOperator
@@ -61,9 +60,6 @@ public:
 
     // Inherited method overridden
     void run(ITensorPack &tensors) override;
-
-protected:
-    std::unique_ptr<ICLKernel> _border_handler{ nullptr };
 };
 } // namespace opencl
 } // namespace arm_compute