aboutsummaryrefslogtreecommitdiff
path: root/src/core/CL/kernels/CLGEMMMatrixMultiplyKernel.cpp
diff options
context:
space:
mode:
authorAnthony Barbier <anthony.barbier@arm.com>2018-08-08 13:20:04 +0100
committerAnthony Barbier <anthony.barbier@arm.com>2018-11-02 16:54:54 +0000
commitb6eb35371d222c6b7f61210d97ebd7dd9e197458 (patch)
treeaf89729ad68d665916c37abb5fd49e512fa40614 /src/core/CL/kernels/CLGEMMMatrixMultiplyKernel.cpp
parent1d1f32ce7ef6acea4afd4cf6a929436640b72ccd (diff)
downloadComputeLibrary-b6eb35371d222c6b7f61210d97ebd7dd9e197458.tar.gz
COMPMID-1478: Stop relying on static default OpenCL objects in cl2.hpp
This causes problems when ACL is used as a shared library on Android. Fixes some problems related to creation / destruction order between the Graph's CL backend and core / runtime Change-Id: I716d63fd42f4586df1ffbb6fa97e4db06d3a781b Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/143228 Tested-by: Jenkins <bsgcomp@arm.com> Reviewed-by: Michele DiGiorgio <michele.digiorgio@arm.com> Reviewed-by: Gian Marco Iodice <gianmarco.iodice@arm.com>
Diffstat (limited to 'src/core/CL/kernels/CLGEMMMatrixMultiplyKernel.cpp')
-rw-r--r--src/core/CL/kernels/CLGEMMMatrixMultiplyKernel.cpp6
1 files changed, 3 insertions, 3 deletions
diff --git a/src/core/CL/kernels/CLGEMMMatrixMultiplyKernel.cpp b/src/core/CL/kernels/CLGEMMMatrixMultiplyKernel.cpp
index 79e2f8b11a..8530ed2fd3 100644
--- a/src/core/CL/kernels/CLGEMMMatrixMultiplyKernel.cpp
+++ b/src/core/CL/kernels/CLGEMMMatrixMultiplyKernel.cpp
@@ -253,7 +253,7 @@ void CLGEMMMatrixMultiplyKernel::configure(const ICLTensor *input0, const ICLTen
// Configure kernel window
auto win_config = validate_and_configure_window(input0->info(), input1->info(), output->info(), is_interleaved_transposed, reshape_info, gpu_target, num_elements_processed);
ARM_COMPUTE_ERROR_THROW_ON(win_config.first);
- ICLKernel::configure(win_config.second);
+ ICLKernel::configure_internal(win_config.second);
// Create build options
CLBuildOptions build_opts;
@@ -316,7 +316,7 @@ void CLGEMMMatrixMultiplyKernel::configure(const ICLTensor *input0, const ICLTen
// The work-group size equal to the Bifrost quad size has been proved to be optimal for these kernels
// via exhaustive autotuning over a range of representative layer configurations.
- _lws_hint = cl::NDRange(4);
+ set_lws_hint(cl::NDRange(4));
}
else // (MIDGARD and F32) or (F16)
{
@@ -416,7 +416,7 @@ void CLGEMMMatrixMultiplyKernel::run(const Window &window, cl::CommandQueue &que
_kernel.setArg<cl_uint>(idx++, static_cast<unsigned int>(_input0->info()->strides_in_bytes()[2]));
_kernel.setArg<cl_uint>(idx++, static_cast<unsigned int>(_input1->info()->strides_in_bytes()[2]));
_kernel.setArg<cl_uint>(idx++, static_cast<unsigned int>(_output->info()->strides_in_bytes()[2]));
- enqueue(queue, *this, slice, _lws_hint);
+ enqueue(queue, *this, slice, lws_hint());
}
while(window.slide_window_slice_3D(slice));
}