* "avoid copy back to cpu"

* "add infinity support"

* "fix ci"

* "add cpu macro"

* rerun ci; test=develop

* "fix api"
test=develop

* test=develop

* test=develop

* test=develop

* test=develop

* test=develop

test=develop

test=develop

test=develop

test=develop

test=develop

- add naive executor
- fix concurrency performance issue

  extended test_text_classification ot use new op

Also fix MKL-DNN pooling integration for ceil mode

* add guide

* "fix doc"

* Update windows_inference.md

Looks like there is a little problem in markdown format writing of head lines

* hotfix for PaddleTensor buffer.