[ONERT] OpenCL Tuner for acl_cl backend

[intom]

I applied OpenCL Tuner to the acl_cl backend. OpenCL Tuner finds the optimal local work size for GPU acceleration tuning parameters. https://developer.arm.com/solutions/machine-learning-on-arm/developer-material/how-to-guides/implement-a-neural-style-transfer-on-android-with-arm-nn-apis/tuning-performance-with-opencl-tuner

I am going to introduce the OpenCL Tuner with Draft PR

Refer to the below test result. Model : mnasnet https://tfhub.dev/tensorflow/lite-model/mnasnet_1.0_224/1/default/1 CL_Tuner

Package Filename /root/mnasnet_1.0_224_1_default_1/
===================================
MODEL_LOAD   takes 27.815 ms
PREPARE      takes 3220.900 ms
EXECUTE      takes 16.578 ms
- MEAN     :  16.578 ms
- MAX      :  28.282 ms
- MIN      :  14.836 ms
- GEOMEAN  :  16.434 ms
===================================

No CL_Tuner

Package Filename /root/mnasnet_1.0_224_1_default_1/
===================================
MODEL_LOAD   takes 11.660 ms
PREPARE      takes 3142.837 ms
EXECUTE      takes 19.295 ms
- MEAN     :  19.295 ms
- MAX      :  25.660 ms
- MIN      :  17.280 ms
- GEOMEAN  :  19.198 ms
===================================

[intom]

cc @ragmani