2.add flatten2 for gpu.
3.add concat 4 inputs size for gpu.
4.fix pool.
5.fix transpose2
test=develop

* [test=develop]1.fix crash when gpu op scale&elementwise_add input dim size equal 2
2.add gpu op mul

* fix code style test=develop

* Catch mobile exceptions test=develop

* code style format test=develop

* enable fast test compilation && push opencl kernels by run.py

* merge cl kernels into so

* restore pre-commit