* [LITE][OPENCL] Change fp32 fc to fp16's. test=develop

* fix act in conv3x3opt opencl kernel. test=develop

* [LITE][OPENCL] Fix issue(previous definition in gcc.a) armv7 gcc compile for opencl. test=develop

* add mean op, test=develop

* split forward and backward, test=develop

* set arm_lang default is off. test=develop

* fix resize error, test-develop

* [CORE] Support Clang compiling for Android
test=develop

* fix lto for jni so
test=develop

* disable assembly code
test=develop

* disable layout.cc
test=develop

* refine
test=develop

* Enable LTO on Clang
test=develop

* Disable LTO when clang is used.
test=develop

* Remove space
test=develop

* [Core] enable clang in build.sh
test=develop

* [Core] Remove the workaround for clang compiling
test=develop

* [LITE][CMAKE] support clang compile for armv8 cpu. test=develop

* add build for clang. test=develop

* fix cmake format. test=develop

* [MLU] support resnet50 on MLU

* [MLU] support resnet50 on MLU

* test=develop

* test=develop

add opencl kernel build list; set num cores to build opencl in ci_build.sh and build.sh. test=develop (#3081)

* test=devellop

* add cl file into resulted lib test=develop

* test=develop

* test=develop

* modify opt name test=develop

* fix code style test=develop

* test=develop

* Support bitman backend

* upgrade benchmark 2, test=develop

* update benchmark, test=develop

remove test_models ci projects, because these project hass been removed in ci test test=develop (#2669)

[LITE][NPU][XPU] Refine subgraph pass, and support NPU/XPU model generation at execution time (#2576)

修改了选kernel的逻辑，默认从模型文件中读取出lod_tensor的data type，在static_kernel_pick pass中如果kernel输入输出的类型与读取的data type完全一致，则选择该Kernel的概率增大。

- 增加 从模型文件__model__读取lod_tensor的data type到cpp::vardesc

- program中增加unordered_map<string, type>字段，并在 Program::PrepareWorkspace中对该字段赋值

- 修改了node.h文件，将const Type* 更改为Type*，并在SSAGraph::Build过程中为符合条件的type*赋值

- static_kernel_pick_pass中添加新规则，如果kernel的输入类型输出类型与__model__中存储的类型的一致，则score*=2。

- 支持模型中用到sequence_reverse_float kernel（输入输出均为float）和sequence_reverse_int64 kernel（输入输出均为int64），能够根据输入输出type选kernel

* fix cuda building, test=develop

* remove sequence_pool from cmake because build error, test=develop

* support build C++ cuda shared lib

* add cuda ci building, test=develop

* update comments, test=develop

* fix conv kernel bugs and open mobilenet ci test=develop

* add cv image process

* fix arm liunx build error

* add LITE_WITH_CV defien to make cv, test=develop

* fix cv format, annd add describe in utils/cv

* delete some Meaningless comments, test=develop

* set LITE_WITH_CV=OFF in build.sh, test=develop

* delete cv_enum.h in utils/cv, push the contents in cv_ennum.h to paddle_image_preprocess.h, test=develop

* according to reviews to redefine paddle_image_preprocess.h, test=develop

* add detailed note of flipParam, test=develop

* fix format in paddle_image_preprocess.h, test=develop

* fix error when build x86. test=develop

* lite_with_X86 does not contain lite_with_cv

* move some basic ops into extra type to reduce library size test=develop (#2347)