Sync tolerance number of dynamic fusion direct conv2d with the current library
Add descriptions and pointers in the tests to document the differences
in test coverage between dynamic fusion and the current library, and
most importantly, why the differences. This will come in handy when we
want to quickly check if all old tests have been migrated so that we can
safely deprecate / remove them.
Resolves COMPMID-5840
Signed-off-by: SiCong Li <sicong.li@arm.com>
Change-Id: Ie6227098979e51d7921810288f594beac19bce6f
Reviewed-on: https://review.mlplatform.org/c/ml/ComputeLibrary/+/9043
Comments-Addressed: Arm Jenkins <bsgcomp@arm.com>
Reviewed-by: Viet-Hoa Do <viet-hoa.do@arm.com>
Tested-by: Arm Jenkins <bsgcomp@arm.com>
Benchmark: Arm Jenkins <bsgcomp@arm.com>
diff --git a/tests/validation/CL/DirectConvolutionLayer.cpp b/tests/validation/CL/DirectConvolutionLayer.cpp
index 512c2ca..342a093 100644
--- a/tests/validation/CL/DirectConvolutionLayer.cpp
+++ b/tests/validation/CL/DirectConvolutionLayer.cpp
@@ -35,6 +35,9 @@
#include "tests/validation/Validation.h"
#include "tests/validation/fixtures/DirectConvolutionLayerFixture.h"
+/** Synced with tests/validation/dynamic_fusion/gpu/cl/DirectConv2d.cpp
+ * Please check there for any differences in the coverage
+ */
namespace arm_compute
{
namespace test