aboutsummaryrefslogtreecommitdiff
path: root/src/core/CL/ICLKernel.cpp
diff options
context:
space:
mode:
authorSang-Hoon Park <sang-hoon.park@arm.com>2019-09-17 08:59:09 +0100
committerGeorgios Pinitas <georgios.pinitas@arm.com>2019-09-30 15:33:01 +0000
commitcecb0a75a9de0a12d30f8fabbe16a656c722afa1 (patch)
tree9517b7fd1fa31126c5fbd3e66f4fb9c7e0b9af1b /src/core/CL/ICLKernel.cpp
parent879d1313ba69d9ced8424f54ffeea6a3c60496f0 (diff)
downloadComputeLibrary-cecb0a75a9de0a12d30f8fabbe16a656c722afa1.tar.gz
COMPMID-2637 [CL] fix broadcast pixel-wise multiplication with 5D tensors
Broadcast pixel-wise multiplication with 5D tensors is fixed by adding information whether a dimension has been broadcasted to compute correct start offset when adding 3D tensor argument. The testcase that failed is added to the validation test suite. Change-Id: I320876f507012c27b39daae1316f9b69138ed204 Signed-off-by: Sang-Hoon Park <sang-hoon.park@arm.com> Reviewed-on: https://review.mlplatform.org/c/1994 Tested-by: Arm Jenkins <bsgcomp@arm.com> Comments-Addressed: Arm Jenkins <bsgcomp@arm.com> Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com>
Diffstat (limited to 'src/core/CL/ICLKernel.cpp')
-rw-r--r--src/core/CL/ICLKernel.cpp2
1 files changed, 1 insertions, 1 deletions
diff --git a/src/core/CL/ICLKernel.cpp b/src/core/CL/ICLKernel.cpp
index 2d28a496c9..d81ad46b29 100644
--- a/src/core/CL/ICLKernel.cpp
+++ b/src/core/CL/ICLKernel.cpp
@@ -98,7 +98,7 @@ void ICLKernel::add_tensor_argument(unsigned &idx, const ICLTensor *tensor, cons
for(unsigned int n = 0; n < info->num_dimensions(); ++n)
{
- offset_first_element += window[n].start() * strides[n];
+ offset_first_element += (window.is_broadcasted(n) ? 0 : window[n].start()) * strides[n];
}
unsigned int idx_start = idx;