Make variable::GetMutable robust

Explain LoD and a few other concepts

fix demo_ci bug in vis_demo.cc

* fix layers.uniform_random

* fix uniform_random
test=develop

* remove var type set
test=develop

* fix similar error
test=develop

test=develop

test=develop

This can cause a lot of problem:
1. Wrong operator implementation, Op can get a wrong type without failure.
2. Anytype can be Get without defined in VarType.

Also fix wrong STEP_SCOPE usage.

test=develop

* optimize depthwise conv by register memory
* test=develop

* profiler infor merge thread statistic information

* update profiler

* fix bug

* add merge thread msg to report

* optimize report

* statistic the time of ops in each thread but not all

* optimize report format

* optimize profile report

* optimize profile report
test=develop

Merge selected_rows for clip_by_norm op

Revert "[MKLDNN] Pass: Fuse Conv + Bias"

test=develop

[MKLDNN] Pass: Fuse Conv + Bias

* Fix truncated normal.

* test=develop

Sfraczek/conv bn fuse pass

add trt api lib to paddle_fluid lib

test=develop

test=develop

Unit test to Fused_embedding_fc_lstm op

Fix flowers data read in python3

Make cmake support compile in MacOSX 10.14

test=develop

test=develop

Revert "optimize pyreader"

test=develop

test=develop

* set the right shape of selected_rows
test=develop

* enhance check

* fix activation_op

* remove cast

* use ShareDimInfo replace SetDim and ShareLod

* use ShareDimAndLod
test=develop

* follow comment

test=develop

* check whether the input has lod
test=develop

* Split ShareDimAndLod

test=develop

* checkout clip.py
test=develop

fix lod tensor

test=develop

test=develop

test=develop

test=develop

* Fix bug in reduce_op caused by PR #13534
* Fix output shape and enhance unit test.
test=develop

revert with_fast_math to ON

test=develop