aboutsummaryrefslogtreecommitdiff
path: root/src/gpu/cl/kernels/ClGemmLowpQuantizeDownInt32ScaleByFloatKernel.cpp
diff options
context:
space:
mode:
authorGian Marco Iodice <gianmarco.iodice@arm.com>2022-09-05 15:35:35 +0100
committerGian Marco Iodice <gianmarco.iodice@arm.com>2022-09-09 08:37:36 +0000
commitd11de9861e6c32fa389f503e037098f50ffed156 (patch)
treeedbb84b8a5ecdbc7217c7277321e7c61cd01ea5c /src/gpu/cl/kernels/ClGemmLowpQuantizeDownInt32ScaleByFloatKernel.cpp
parente4e3b2ead5b6720af8039f3c9ac15ea6b51b915f (diff)
downloadComputeLibrary-d11de9861e6c32fa389f503e037098f50ffed156.tar.gz
Add a macro guard in all OpenCL kernels in gemmlowp.cl
Resolves COMPMID-5498 Change-Id: I474f3f963257014255d082aab0ccbe3efe5aa067 Signed-off-by: Gian Marco Iodice <gianmarco.iodice@arm.com> Reviewed-on: https://review.mlplatform.org/c/ml/ComputeLibrary/+/8222 Tested-by: Arm Jenkins <bsgcomp@arm.com> Comments-Addressed: Arm Jenkins <bsgcomp@arm.com> Comments-Addressed: Ramy Elgammal <ramy.elgammal@arm.com> Reviewed-by: Ramy Elgammal <ramy.elgammal@arm.com> Reviewed-by: Gunes Bayir <gunes.bayir@arm.com> Benchmark: Arm Jenkins <bsgcomp@arm.com>
Diffstat (limited to 'src/gpu/cl/kernels/ClGemmLowpQuantizeDownInt32ScaleByFloatKernel.cpp')
-rw-r--r--src/gpu/cl/kernels/ClGemmLowpQuantizeDownInt32ScaleByFloatKernel.cpp9
1 files changed, 7 insertions, 2 deletions
diff --git a/src/gpu/cl/kernels/ClGemmLowpQuantizeDownInt32ScaleByFloatKernel.cpp b/src/gpu/cl/kernels/ClGemmLowpQuantizeDownInt32ScaleByFloatKernel.cpp
index c5cea3d17d..8d4cb923d6 100644
--- a/src/gpu/cl/kernels/ClGemmLowpQuantizeDownInt32ScaleByFloatKernel.cpp
+++ b/src/gpu/cl/kernels/ClGemmLowpQuantizeDownInt32ScaleByFloatKernel.cpp
@@ -1,5 +1,5 @@
/*
- * Copyright (c) 2018-2021 Arm Limited.
+ * Copyright (c) 2018-2022 Arm Limited.
*
* SPDX-License-Identifier: MIT
*
@@ -113,8 +113,13 @@ void ClGemmLowpQuantizeDownInt32ScaleByFloatKernel::configure(const CLCompileCon
build_opts.add_option_if((max < 255), "-DMAX_BOUND=" + support::cpp11::to_string(max));
build_opts.add_option_if(bias != nullptr, "-DADD_BIAS");
+ const std::string kernel_name = "gemmlowp_output_stage_quantize_down_float";
+
+ // 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, "gemmlowp_output_stage_quantize_down_float", build_opts.options());
+ _kernel = create_kernel(compile_context, kernel_name, build_opts.options());
// Configure kernel window
Window win = calculate_max_window(*src, Steps(num_elems_processed_per_iteration));