Improve start-up time for concatenation layers
- pass tensor's dimensions at runtime rather than compile time
- Add guard macro to compile only kernel of internest
Resolves: COMPMID-5121
Signed-off-by: Ramy Elgammal <ramy.elgammal@arm.com>
Change-Id: I76b7c0cf56d803f58ebff5494c904ace2a86ef5a
Reviewed-on: https://review.mlplatform.org/c/ml/ComputeLibrary/+/7097
Tested-by: Arm Jenkins <bsgcomp@arm.com>
Reviewed-by: Gian Marco Iodice <gianmarco.iodice@arm.com>
Comments-Addressed: Arm Jenkins <bsgcomp@arm.com>
diff --git a/src/gpu/cl/kernels/ClBatchConcatenateKernel.cpp b/src/gpu/cl/kernels/ClBatchConcatenateKernel.cpp
index 8c2af5f..62040df 100644
--- a/src/gpu/cl/kernels/ClBatchConcatenateKernel.cpp
+++ b/src/gpu/cl/kernels/ClBatchConcatenateKernel.cpp
@@ -1,5 +1,5 @@
/*
- * Copyright (c) 2019-2021 Arm Limited.
+ * Copyright (c) 2019-2022 Arm Limited.
*
* SPDX-License-Identifier: MIT
*
@@ -91,8 +91,13 @@
build_opts.add_option("-DSCALE_OUT=" + float_to_string_with_full_precision(oq_info.scale));
}
+ std::string kernel_name = "concatenate";
+
+ // A macro guard to compile ONLY the kernel of interest
+ build_opts.add_option("-D" + upper_string(kernel_name));
+
// Create kernel
- _kernel = create_kernel(compile_context, "concatenate", build_opts.options());
+ _kernel = create_kernel(compile_context, kernel_name, build_opts.options());
// Configure kernel window
auto win = calculate_max_window(*dst, Steps(num_elems_processed_per_iteration));