COMPMID-3638: Move NEON kernels

Signed-off-by: Michalis Spyrou <michalis.spyrou@arm.com>
Change-Id: Ieed3e4bc8be7fef80c90c5094599b477a56fc473
Reviewed-on: https://review.mlplatform.org/c/ml/ComputeLibrary/+/4285
Comments-Addressed: Arm Jenkins <bsgcomp@arm.com>
Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com>
Tested-by: Arm Jenkins <bsgcomp@arm.com>

1 files changed, 89 insertions, 0 deletions

diff --git a/src/core/NEON/kernels/NEBatchConcatenateLayerKernel.h b/src/core/NEON/kernels/NEBatchConcatenateLayerKernel.h
new file mode 100644
index 0000000000..b74a94805d
--- /dev/null
+++ b/src/core/NEON/kernels/NEBatchConcatenateLayerKernel.h
@@ -0,0 +1,89 @@
+/*
+ * Copyright (c) 2019-2020 Arm Limited.
+ *
+ * SPDX-License-Identifier: MIT
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to
+ * deal in the Software without restriction, including without limitation the
+ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
+ * sell copies of the Software, and to permit persons to whom the Software is
+ * furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in all
+ * copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+ * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+
+#ifndef ARM_COMPUTE_NEBATCHCONCATENATEKERNEL_H
+#define ARM_COMPUTE_NEBATCHCONCATENATEKERNEL_H
+
+#include "src/core/NEON/INEKernel.h"
+
+namespace arm_compute
+{
+// Forward declarations
+class ITensor;
+
+/** Interface for the batch concatenate kernel.
+ *  The input tensor will be concatenated into the output tensor.
+ */
+class NEBatchConcatenateLayerKernel : public INEKernel
+{
+public:
+    const char *name() const override
+    {
+        return "NEBatchConcatenateLayerKernel";
+    }
+    /** Default constructor */
+    NEBatchConcatenateLayerKernel();
+    /** Prevent instances of this class from being copied (As this class contains pointers) */
+    NEBatchConcatenateLayerKernel(const NEBatchConcatenateLayerKernel &) = delete;
+    /** Prevent instances of this class from being copied (As this class contains pointers) */
+    NEBatchConcatenateLayerKernel &operator=(const NEBatchConcatenateLayerKernel &) = delete;
+    /** Allow instances of this class to be moved */
+    NEBatchConcatenateLayerKernel(NEBatchConcatenateLayerKernel &&) = default;
+    /** Allow instances of this class to be moved */
+    NEBatchConcatenateLayerKernel &operator=(NEBatchConcatenateLayerKernel &&) = default;
+    /** Default destructor */
+    ~NEBatchConcatenateLayerKernel() = default;
+    /** Initialise the kernel's inputs and output
+     *
+     * @param[in]     input        Input tensor info. Data types supported: All.
+     * @param[in]     batch_offset The offset on axis # 3.
+     * @param[in,out] output       Output tensor info. 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.
+     *
+     */
+    void configure(const ITensorInfo *input, unsigned int batch_offset, ITensorInfo *output);
+    /**  Static function to check if given info will lead to a valid configuration of @ref NEBatchConcatenateLayerKernel
+     *
+     * @param[in] input        Input tensor info. Data types supported: All.
+     * @param[in] batch_offset The offset on axis # 3.
+     * @param[in] output       Output tensor info. Data types supported: Same as @p input.
+     *
+     * @return a status
+     */
+    static Status validate(const ITensorInfo *input, unsigned int batch_offset, const ITensorInfo *output);
+
+    // Inherited methods overridden:
+    void run_op(ITensorPack &tensors, const Window &window, const ThreadInfo &info) override;
+
+private:
+    using BatchConcatFunction = void(const ITensor *in, ITensor *out, unsigned int batch_offset, const Window &window);
+
+private:
+    BatchConcatFunction *_func;
+    unsigned int         _batch_offset;
+};
+} // namespace arm_compute
+#endif /* ARM_COMPUTE_NEBATCHCONCATENATEKERNEL_H */