COMPMID-417: Port DepthConcatenate to QS8/QS16 for NEON/CL.

Change-Id: I3dddae63043c7aa18d908a4fc8abacf3c64f98ca
Reviewed-on: http://mpd-gerrit.cambridge.arm.com/80081
Tested-by: Kaizen <jeremy.johnson+kaizengerrit@arm.com>
Reviewed-by: Steven Niu <steven.niu@arm.com>

1 files changed, 11 insertions, 7 deletions

diff --git a/arm_compute/core/NEON/kernels/NEDepthConcatenateKernel.h b/arm_compute/core/NEON/kernels/NEDepthConcatenateKernel.h
index 7384cd1f02..b22d37bfe6 100644
--- a/arm_compute/core/NEON/kernels/NEDepthConcatenateKernel.h
+++ b/arm_compute/core/NEON/kernels/NEDepthConcatenateKernel.h
@@ -51,9 +51,9 @@ public:
     ~NEDepthConcatenateKernel() = default;
     /** Initialise the kernel's inputs and output
      *
-     * @param[in]     input        Input tensor. Data types supported: F32.
+     * @param[in]     input        Input tensor. Data types supported: QS8/QS16/F16/F32.
      * @param[in]     depth_offset The offset on the Z axis.
-     * @param[in,out] output       Output tensor. Data types supported: F32.
+     * @param[in,out] output       Output tensor. Data types supported: Same as @p input.
      *
      * @note: The output tensor's low two dimensions can't be smaller than the input one's.
      * @note: The gaps between the two lowest dimensions of input and output need to be divisible by 2.
@@ -66,11 +66,15 @@ public:
     BorderSize border_size() const override;
 
 private:
-    const ITensor *_input;
-    ITensor       *_output;
-    int            _top_bottom;
-    int            _left_right;
-    unsigned int   _depth_offset;
+    using DepthConcatFunction = void(const ITensor *in, ITensor *out, std::pair<int, int> start_xy, int depth_offset, const Window &window);
+
+private:
+    DepthConcatFunction *_func;
+    const ITensor       *_input;
+    ITensor             *_output;
+    int                  _top_bottom;
+    int                  _left_right;
+    unsigned int         _depth_offset;
 };
 }
 #endif /* __ARM_COMPUTE_NEDEPTHCONCATENATEKERNEL_H__ */