diff options
author | Viet-Hoa Do <viet-hoa.do@arm.com> | 2024-01-16 16:23:24 +0000 |
---|---|---|
committer | Viet-Hoa Do <viet-hoa.do@arm.com> | 2024-01-18 11:32:09 +0000 |
commit | 6829e0201e886cfd311e39f1d88f7452894bdfe5 (patch) | |
tree | ac1b97df4b76d0d09652ef3dbdce2be3d222df3c /src/core/CL/cl_kernels/common/instance_normalization.cl | |
parent | 8896cf7cb7df34c699e7453a0f0c683d1202ed15 (diff) | |
download | ComputeLibrary-6829e0201e886cfd311e39f1d88f7452894bdfe5.tar.gz |
Fix divide-by-zero compilation error
* CONVERT_TO_TENSOR4D_STRUCT_NO_STEP is implemented and used
in some CL kernels in the way that causes divide-by-zero issue.
- Since the steps are all zeros, the issue might have been
ignored by the compiler.
Resolves: COMPMID-6795
Signed-off-by: Viet-Hoa Do <viet-hoa.do@arm.com>
Change-Id: I0fb38fc62d63671b8abefa39b3d9b3ca6f49c7fe
Reviewed-on: https://review.mlplatform.org/c/ml/ComputeLibrary/+/10967
Reviewed-by: Gunes Bayir <gunes.bayir@arm.com>
Tested-by: Arm Jenkins <bsgcomp@arm.com>
Comments-Addressed: Arm Jenkins <bsgcomp@arm.com>
Benchmark: Arm Jenkins <bsgcomp@arm.com>
Diffstat (limited to 'src/core/CL/cl_kernels/common/instance_normalization.cl')
-rw-r--r-- | src/core/CL/cl_kernels/common/instance_normalization.cl | 8 |
1 files changed, 4 insertions, 4 deletions
diff --git a/src/core/CL/cl_kernels/common/instance_normalization.cl b/src/core/CL/cl_kernels/common/instance_normalization.cl index adfbebd67d..f9b3cd3620 100644 --- a/src/core/CL/cl_kernels/common/instance_normalization.cl +++ b/src/core/CL/cl_kernels/common/instance_normalization.cl @@ -1,5 +1,5 @@ /* - * Copyright (c) 2019-2021 Arm Limited. + * Copyright (c) 2019-2021, 2024 Arm Limited. * * SPDX-License-Identifier: MIT * @@ -53,7 +53,7 @@ __kernel void compute_mean_var( TENSOR4D_DECLARATION(input), TENSOR3D_DECLARATION(output)) { - Tensor4D in = CONVERT_TO_TENSOR4D_STRUCT_NO_STEP(input, 0); + Tensor4D in = CONVERT_TO_TENSOR4D_STRUCT_NO_STEP(input); Tensor3D out = CONVERT_TO_TENSOR3D_STRUCT_NO_STEP(output); #if defined(NHWC) @@ -176,10 +176,10 @@ __kernel void instance_normalization( #endif /* IN_PLACE */ ) { - Tensor4D in = CONVERT_TO_TENSOR4D_STRUCT_NO_STEP(input, 0); + Tensor4D in = CONVERT_TO_TENSOR4D_STRUCT_NO_STEP(input); Tensor3D mean_var = CONVERT_TO_TENSOR3D_STRUCT_NO_STEP(mean_var); #ifndef IN_PLACE - Tensor4D out = CONVERT_TO_TENSOR4D_STRUCT_NO_STEP(output, 0); + Tensor4D out = CONVERT_TO_TENSOR4D_STRUCT_NO_STEP(output); #endif /* IN_PLACE */ #if defined(NHWC) |