Decouple CpuElementwiseKernel

1- reorganize the folders struct according the new definition
2- separate between unary and binary implementations
3- decuple kernels - unary , binary op and binary comparision

Resolves COMPMID-4634
Change-Id: I0195846cc372e74a63c659069a4508de53a22110
Signed-off-by: Dana Zlotnik <dana.zlotnik@arm.com>
Reviewed-on: https://review.mlplatform.org/c/ml/ComputeLibrary/+/6860
Tested-by: Arm Jenkins <bsgcomp@arm.com>
Reviewed-by: Giorgio Arena <giorgio.arena@arm.com>
Comments-Addressed: Arm Jenkins <bsgcomp@arm.com>

1 files changed, 58 insertions, 0 deletions

diff --git a/src/cpu/kernels/elementwise_binary/generic/neon/fp32.cpp b/src/cpu/kernels/elementwise_binary/generic/neon/fp32.cpp
new file mode 100644
index 0000000000..2d8fec91c5
--- /dev/null
+++ b/src/cpu/kernels/elementwise_binary/generic/neon/fp32.cpp
@@ -0,0 +1,58 @@
+/*
+ * Copyright (c) 2022 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.
+ */
+#include "arm_compute/core/Helpers.h"
+#include "src/cpu/kernels/elementwise_binary/generic/neon/impl.h"
+
+namespace arm_compute
+{
+namespace cpu
+{
+template <ArithmeticOperation op>
+void neon_fp32_elementwise_binary(const ITensor *in1, const ITensor *in2, ITensor *out, const Window &window)
+{
+    return elementwise_arithm_op<op, typename wrapper::traits::neon_vector<float, 4>>(in1, in2, out, window);
+}
+
+template void neon_fp32_elementwise_binary<ArithmeticOperation::ADD>(const ITensor *in1, const ITensor *in2, ITensor *out, const Window &window);
+template void neon_fp32_elementwise_binary<ArithmeticOperation::SUB>(const ITensor *in1, const ITensor *in2, ITensor *out, const Window &window);
+template void neon_fp32_elementwise_binary<ArithmeticOperation::DIV>(const ITensor *in1, const ITensor *in2, ITensor *out, const Window &window);
+template void neon_fp32_elementwise_binary<ArithmeticOperation::MIN>(const ITensor *in1, const ITensor *in2, ITensor *out, const Window &window);
+template void neon_fp32_elementwise_binary<ArithmeticOperation::MAX>(const ITensor *in1, const ITensor *in2, ITensor *out, const Window &window);
+template void neon_fp32_elementwise_binary<ArithmeticOperation::SQUARED_DIFF>(const ITensor *in1, const ITensor *in2, ITensor *out, const Window &window);
+template void neon_fp32_elementwise_binary<ArithmeticOperation::POWER>(const ITensor *in1, const ITensor *in2, ITensor *out, const Window &window);
+template void neon_fp32_elementwise_binary<ArithmeticOperation::PRELU>(const ITensor *in1, const ITensor *in2, ITensor *out, const Window &window);
+
+template <ComparisonOperation op>
+void neon_fp32_comparison_elementwise_binary(const ITensor *in1, const ITensor *in2, ITensor *out, const Window &window)
+{
+    return elementwise_comp_op_32<op, float, float32x4_t>(in1, in2, out, window);
+}
+template void neon_fp32_comparison_elementwise_binary<ComparisonOperation::Equal>(const ITensor *in1, const ITensor *in2, ITensor *out, const Window &window);
+template void neon_fp32_comparison_elementwise_binary<ComparisonOperation::NotEqual>(const ITensor *in1, const ITensor *in2, ITensor *out, const Window &window);
+template void neon_fp32_comparison_elementwise_binary<ComparisonOperation::Greater>(const ITensor *in1, const ITensor *in2, ITensor *out, const Window &window);
+template void neon_fp32_comparison_elementwise_binary<ComparisonOperation::GreaterEqual>(const ITensor *in1, const ITensor *in2, ITensor *out, const Window &window);
+template void neon_fp32_comparison_elementwise_binary<ComparisonOperation::Less>(const ITensor *in1, const ITensor *in2, ITensor *out, const Window &window);
+template void neon_fp32_comparison_elementwise_binary<ComparisonOperation::LessEqual>(const ITensor *in1, const ITensor *in2, ITensor *out, const Window &window);
+}
+} // namespace arm_compute