IVGCVSW-2957 MergerLayer subtensor optimization now backend agnostic
+ Update clframework pin
+ Cl and Neon Merger workloads updated to use MemoryLayout agnostic API
+ Workloads only use sub-tensor optimization if ALL input tensors are sub-tensors
+ Refactor LayerSupportCommon code to be a bit more succinct
Change-Id: Ib61ad4ccbd767e924dff07e61022e0cda4069828
Signed-off-by: Derek Lamberti <derek.lamberti@arm.com>
diff --git a/src/backends/cl/workloads/ClMergerWorkload.hpp b/src/backends/cl/workloads/ClMergerWorkload.hpp
index 8189a1b..1c2f823 100644
--- a/src/backends/cl/workloads/ClMergerWorkload.hpp
+++ b/src/backends/cl/workloads/ClMergerWorkload.hpp
@@ -24,8 +24,7 @@
void Execute() const override;
private:
- mutable arm_compute::CLConcatenateLayer m_Layer;
- bool m_Execute;
+ mutable std::unique_ptr<arm_compute::CLConcatenateLayer> m_Layer;
};
} //namespace armnn