test=develop

* Remove the memory copy for feeding data in C++ inference API
* Fix compling dependence
* Fix compling in ONLY_CPU mode

test=develop

test=develop

the parameters will load from CPUPlace, that will keep copying data
between CPU and GPU places.

test=develop

Fix code to support cpplint syntax check

test=develop

This patch also refactors conv+bn (includes changes from PR
https://github.com/PaddlePaddle/Paddle/pull/13926)
updated to use the mkldnn-placement-pass.

test=develop

After the MKL-DNN placement pass there is no need to enable MKL-DNN
in operators via executor

test=develop

* add MKL-DNN placement pass

This patch also refactors conv+bn (includes changes from PR
https://github.com/PaddlePaddle/Paddle/pull/13926)
updated to use the mkldnn-placement-pass.

test=develop

* remove redundant pass list

* add comment on the default first pass

* fix test for conv+relu mkldnn fuse

test=develop

test=develop

test=develop

test=develop

- add naive executor
- fix concurrency performance issue

Also fix MKL-DNN pooling integration for ceil mode